HDDM-0.2/0000775000102500010250000000000011727455647012651 5ustar wieckiwiecki00000000000000HDDM-0.2/setupegg.py0000664000102500010250000000022511713025572015026 0ustar wieckiwiecki00000000000000#!/usr/bin/env python """ A setup.py script to use setuptools, which gives egg goodness, etc. """ from setuptools import setup execfile('setup.py') HDDM-0.2/docs/0000775000102500010250000000000011727455647013601 5ustar wieckiwiecki00000000000000HDDM-0.2/docs/tutorial_config_subjects.rst0000664000102500010250000004171411726176127021424 0ustar wieckiwiecki00000000000000.. index:: Tutorial .. _chap_tutorial_config_subjects: *********************************** Creating a hierarchical group model *********************************** Up until now, we have been looking at data that was generated from the same set of parameters. However, in most experiments, we test multiple subjects and may only gather relatively few trials per subject. Traditionally, we would either fit a separate model to each individual subject or fit one model to all subjects. Neither of these approaches are ideal as we will see below. We can expect that subjects will be similar in many ways yet have non-negligable individual differences. If we fit separate models we ignore their similarities and need much more data per subject to make useful inference. If we fit one model to all subjects we ignore their differences and may get a model fit that does not well characterize any individual. It would be best if the model fitting could determine to what extent all subjects are similar to the others in some respects, and use this information as a prior to draw inferences about the paramters of any given subject. The hierarchical approach optimally allocates the information from the group vs the individual depending on the statistics of the data. To illustrate this point, consider the following example: we tested 40 subjects on the above task with the easy and hard condition. For practical reasons, however, we only collected 20 trials per condition. As an example of what happens when trying to fit separate models to each subject, we will run HDDM on the first subject. The file simple_difficulty_subjs_single.csv only contains data from the first subject. Lets run our model and see what happens: :: hddm_fit.py simple_difficulty.conf simple_subjs_difficulty_single.csv Creating model... Sampling: 100% [0000000000000000000000000000000000] Iterations: 10000 name mean std 2.5q 25q 50q 75q 97.5 mc_err a : 1.936 0.152 1.665 1.830 1.926 2.030 2.259 0.007 t : 0.314 0.044 0.214 0.288 0.318 0.347 0.386 0.002 v('easy',): 0.468 0.248 -0.017 0.298 0.473 0.642 0.929 0.009 v('hard',): 0.426 0.234 -0.028 0.268 0.428 0.571 0.869 0.008 logp: -60.214194 DIC: 113.099890 As you can see, the estimates are far worse (especially in the hard condition) and the posterior distributions are much wider indicating a lack of confidence in the estimates. Looking at the posterior predictive and ill-shaped RT distribution makes obvious why fitting a DDM to 10 trials is a fruitless attempt. However, what about the data from the 39 other subjects? We certainly wouldn't expect everyone to have the exact same parameters, but they should be fairly similar. Couldn't we combine the data? This is where the hierarchical approach becomes useful -- we can estimate individual parameters, but at the same time have the parameters feed into a group distribution so that the tiny bit we learn from each subject can be pooled together and constrain the subject fits. Unfortunately, such a model is a little bit more difficult to create in general. Fortunately, HDDM does all of this automatically. Simply running hddm_fit.py on a data file that has a column named 'subj_idx' will make HDDM create a hierarchical model where each subject gets assigned its own subject distribution which is assumed to be distributed around a normal group distribution. Running this example produces the following output: :: >>hddm_fit.py simple_difficulty.conf simple_subjs_difficulty.csv Creating model... Sampling: 100% [00000000000000000000000000000000] Iterations: 20000 name mean std 2.5q 25q 50q 75q 97.5 mc_err a : 2.000 0.026 1.950 1.985 1.996 2.018 2.059 0.002 a0 : 2.002 0.053 1.889 1.977 1.996 2.032 2.120 0.003 a1 : 1.987 0.051 1.870 1.966 1.990 2.011 2.084 0.002 a10 : 2.007 0.049 1.918 1.981 1.997 2.032 2.123 0.003 a11 : 2.009 0.050 1.916 1.981 1.998 2.033 2.122 0.003 a12 : 2.001 0.052 1.898 1.977 1.994 2.028 2.111 0.002 a13 : 2.000 0.052 1.890 1.977 1.994 2.027 2.112 0.002 a14 : 1.995 0.051 1.885 1.973 1.993 2.020 2.104 0.002 a15 : 1.978 0.058 1.829 1.956 1.986 2.009 2.082 0.003 a16 : 2.000 0.050 1.900 1.976 1.995 2.023 2.121 0.002 a17 : 2.011 0.053 1.915 1.980 1.999 2.040 2.142 0.003 a18 : 2.013 0.057 1.906 1.983 2.000 2.045 2.149 0.003 a19 : 2.009 0.054 1.910 1.981 1.999 2.038 2.127 0.003 a2 : 2.003 0.051 1.900 1.980 1.996 2.029 2.119 0.003 a20 : 2.041 0.069 1.951 1.991 2.022 2.076 2.213 0.005 a21 : 1.987 0.054 1.856 1.966 1.990 2.015 2.094 0.003 a22 : 2.003 0.053 1.894 1.976 1.995 2.030 2.112 0.002 a23 : 2.002 0.054 1.886 1.978 1.995 2.029 2.125 0.002 a24 : 1.999 0.049 1.903 1.976 1.994 2.026 2.107 0.002 a25 : 2.012 0.055 1.914 1.983 2.000 2.041 2.139 0.003 a26 : 2.006 0.052 1.906 1.981 1.997 2.033 2.122 0.002 a27 : 1.988 0.053 1.865 1.968 1.991 2.012 2.087 0.003 a28 : 2.015 0.055 1.923 1.983 2.001 2.045 2.147 0.003 a29 : 1.982 0.054 1.853 1.961 1.988 2.010 2.082 0.003 a3 : 1.981 0.056 1.842 1.960 1.988 2.008 2.081 0.003 a30 : 2.013 0.056 1.910 1.983 1.999 2.042 2.147 0.003 a31 : 1.999 0.048 1.901 1.976 1.993 2.024 2.101 0.002 a32 : 1.999 0.050 1.899 1.977 1.994 2.024 2.111 0.002 a33 : 2.017 0.053 1.929 1.985 2.004 2.044 2.153 0.003 a34 : 2.001 0.052 1.898 1.978 1.994 2.023 2.120 0.002 a35 : 2.002 0.052 1.895 1.978 1.995 2.027 2.117 0.002 a36 : 1.983 0.052 1.849 1.964 1.988 2.010 2.077 0.003 a37 : 2.002 0.054 1.890 1.977 1.994 2.029 2.127 0.003 a38 : 1.984 0.055 1.845 1.965 1.991 2.011 2.083 0.003 a39 : 1.975 0.058 1.826 1.954 1.985 2.007 2.071 0.003 a4 : 1.998 0.052 1.886 1.975 1.994 2.025 2.105 0.002 a5 : 2.024 0.056 1.930 1.987 2.008 2.053 2.164 0.003 a6 : 2.003 0.051 1.900 1.980 1.996 2.029 2.120 0.002 a7 : 2.004 0.051 1.901 1.978 1.996 2.030 2.119 0.002 a8 : 1.989 0.051 1.879 1.967 1.990 2.015 2.089 0.002 a9 : 1.984 0.055 1.847 1.964 1.990 2.014 2.083 0.003 avar : 0.042 0.030 0.005 0.016 0.038 0.065 0.103 0.003 t : 0.305 0.006 0.293 0.302 0.305 0.309 0.316 0.000 t0 : 0.308 0.012 0.286 0.301 0.307 0.314 0.337 0.001 t1 : 0.304 0.012 0.279 0.298 0.305 0.311 0.328 0.000 t10 : 0.308 0.012 0.284 0.301 0.307 0.314 0.334 0.001 t11 : 0.308 0.013 0.284 0.301 0.307 0.314 0.341 0.001 t12 : 0.306 0.012 0.281 0.300 0.306 0.312 0.331 0.000 t13 : 0.309 0.013 0.286 0.301 0.307 0.314 0.340 0.001 t14 : 0.304 0.011 0.277 0.299 0.305 0.311 0.327 0.000 t15 : 0.305 0.011 0.279 0.299 0.305 0.311 0.326 0.001 t16 : 0.305 0.012 0.281 0.299 0.305 0.311 0.328 0.000 t17 : 0.307 0.012 0.284 0.300 0.306 0.312 0.333 0.001 t18 : 0.306 0.011 0.283 0.300 0.306 0.312 0.331 0.001 t19 : 0.307 0.012 0.285 0.301 0.307 0.313 0.335 0.001 t2 : 0.306 0.012 0.281 0.299 0.306 0.312 0.329 0.001 t20 : 0.309 0.014 0.287 0.301 0.308 0.315 0.343 0.001 t21 : 0.303 0.011 0.275 0.298 0.304 0.310 0.323 0.001 t22 : 0.306 0.011 0.282 0.300 0.306 0.312 0.331 0.001 t23 : 0.307 0.012 0.284 0.301 0.306 0.313 0.335 0.001 t24 : 0.303 0.012 0.275 0.298 0.304 0.310 0.325 0.001 t25 : 0.305 0.012 0.281 0.299 0.306 0.312 0.330 0.001 t26 : 0.308 0.012 0.286 0.301 0.307 0.313 0.338 0.001 t27 : 0.302 0.012 0.273 0.296 0.303 0.310 0.323 0.001 t28 : 0.307 0.011 0.285 0.300 0.306 0.313 0.332 0.001 t29 : 0.300 0.011 0.272 0.295 0.302 0.308 0.319 0.001 t3 : 0.303 0.011 0.278 0.298 0.304 0.310 0.323 0.001 t30 : 0.308 0.012 0.286 0.301 0.307 0.314 0.338 0.001 t31 : 0.304 0.011 0.280 0.298 0.305 0.311 0.325 0.001 t32 : 0.305 0.011 0.278 0.299 0.305 0.311 0.328 0.001 t33 : 0.304 0.011 0.278 0.298 0.304 0.310 0.325 0.001 t34 : 0.306 0.012 0.281 0.299 0.306 0.312 0.330 0.000 t35 : 0.304 0.012 0.275 0.298 0.304 0.311 0.327 0.001 t36 : 0.299 0.011 0.270 0.293 0.301 0.307 0.316 0.001 t37 : 0.307 0.012 0.283 0.300 0.307 0.313 0.335 0.001 t38 : 0.302 0.011 0.276 0.296 0.303 0.309 0.321 0.001 t39 : 0.306 0.012 0.282 0.299 0.305 0.312 0.332 0.001 t4 : 0.306 0.011 0.283 0.300 0.305 0.312 0.329 0.000 t5 : 0.305 0.011 0.280 0.299 0.305 0.311 0.327 0.001 t6 : 0.306 0.012 0.285 0.300 0.306 0.312 0.334 0.001 t7 : 0.308 0.012 0.285 0.301 0.307 0.313 0.337 0.001 t8 : 0.305 0.012 0.281 0.299 0.306 0.312 0.329 0.001 t9 : 0.299 0.012 0.269 0.294 0.301 0.307 0.318 0.001 tvar : 0.009 0.006 0.002 0.004 0.008 0.013 0.025 0.001 v('easy',) : 0.401 0.045 0.317 0.370 0.400 0.430 0.500 0.003 v('easy',)0 : 0.488 0.138 0.278 0.387 0.464 0.565 0.803 0.007 v('easy',)1 : 0.424 0.119 0.203 0.354 0.411 0.489 0.695 0.004 v('easy',)10 : 0.333 0.115 0.075 0.273 0.344 0.406 0.551 0.005 v('easy',)11 : 0.393 0.111 0.181 0.322 0.387 0.461 0.626 0.004 v('easy',)12 : 0.400 0.109 0.183 0.336 0.393 0.466 0.632 0.003 v('easy',)13 : 0.444 0.120 0.232 0.365 0.430 0.506 0.727 0.005 v('easy',)14 : 0.385 0.109 0.162 0.321 0.380 0.447 0.618 0.003 v('easy',)15 : 0.387 0.116 0.157 0.320 0.383 0.454 0.629 0.004 v('easy',)16 : 0.379 0.111 0.150 0.313 0.375 0.448 0.614 0.004 v('easy',)17 : 0.435 0.119 0.238 0.350 0.421 0.502 0.717 0.005 v('easy',)18 : 0.374 0.109 0.140 0.313 0.373 0.438 0.597 0.003 v('easy',)19 : 0.423 0.112 0.212 0.350 0.417 0.486 0.654 0.005 v('easy',)2 : 0.396 0.117 0.148 0.327 0.391 0.463 0.653 0.004 v('easy',)20 : 0.422 0.109 0.230 0.351 0.412 0.485 0.675 0.004 v('easy',)21 : 0.437 0.120 0.229 0.357 0.423 0.510 0.705 0.005 v('easy',)22 : 0.414 0.116 0.196 0.342 0.406 0.479 0.671 0.004 v('easy',)23 : 0.375 0.110 0.140 0.312 0.376 0.438 0.590 0.004 v('easy',)24 : 0.303 0.130 -0.010 0.240 0.323 0.388 0.522 0.006 v('easy',)25 : 0.465 0.129 0.248 0.377 0.447 0.546 0.767 0.007 v('easy',)26 : 0.398 0.116 0.167 0.330 0.395 0.460 0.638 0.004 v('easy',)27 : 0.437 0.117 0.223 0.360 0.423 0.501 0.697 0.005 v('easy',)28 : 0.347 0.117 0.081 0.284 0.357 0.420 0.557 0.004 v('easy',)29 : 0.405 0.116 0.202 0.332 0.396 0.471 0.653 0.004 v('easy',)3 : 0.420 0.117 0.193 0.344 0.412 0.489 0.682 0.004 v('easy',)30 : 0.357 0.107 0.111 0.296 0.359 0.421 0.564 0.004 v('easy',)31 : 0.395 0.108 0.182 0.328 0.392 0.454 0.626 0.004 v('easy',)32 : 0.338 0.115 0.079 0.274 0.347 0.410 0.561 0.004 v('easy',)33 : 0.389 0.112 0.159 0.324 0.389 0.456 0.629 0.004 v('easy',)34 : 0.450 0.124 0.254 0.366 0.437 0.519 0.727 0.005 v('easy',)35 : 0.394 0.112 0.172 0.327 0.387 0.461 0.630 0.004 v('easy',)36 : 0.326 0.120 0.045 0.266 0.342 0.397 0.537 0.004 v('easy',)37 : 0.393 0.111 0.164 0.327 0.388 0.459 0.636 0.004 v('easy',)38 : 0.433 0.118 0.223 0.356 0.418 0.497 0.704 0.005 v('easy',)39 : 0.424 0.119 0.202 0.349 0.418 0.490 0.680 0.005 v('easy',)4 : 0.430 0.126 0.201 0.355 0.420 0.499 0.718 0.005 v('easy',)5 : 0.380 0.110 0.144 0.319 0.377 0.446 0.603 0.004 v('easy',)6 : 0.357 0.115 0.114 0.295 0.362 0.423 0.586 0.004 v('easy',)7 : 0.449 0.124 0.238 0.366 0.434 0.527 0.720 0.006 v('easy',)8 : 0.425 0.122 0.214 0.349 0.411 0.494 0.706 0.005 v('easy',)9 : 0.476 0.133 0.264 0.383 0.456 0.553 0.796 0.007 v('hard',) : 0.966 0.050 0.863 0.934 0.969 0.998 1.057 0.004 v('hard',)0 : 0.970 0.092 0.786 0.920 0.971 1.017 1.190 0.004 v('hard',)1 : 0.983 0.091 0.810 0.929 0.983 1.031 1.180 0.005 v('hard',)10 : 0.964 0.093 0.762 0.915 0.970 1.013 1.159 0.004 v('hard',)11 : 0.989 0.093 0.806 0.936 0.988 1.038 1.195 0.005 v('hard',)12 : 0.962 0.085 0.786 0.914 0.967 1.012 1.125 0.004 v('hard',)13 : 0.964 0.090 0.764 0.915 0.971 1.017 1.153 0.005 v('hard',)14 : 0.962 0.089 0.770 0.918 0.967 1.010 1.136 0.004 v('hard',)15 : 0.963 0.090 0.774 0.916 0.967 1.013 1.151 0.004 v('hard',)16 : 0.949 0.091 0.739 0.902 0.960 1.005 1.119 0.004 v('hard',)17 : 0.969 0.089 0.784 0.919 0.968 1.016 1.160 0.005 v('hard',)18 : 0.978 0.089 0.806 0.926 0.979 1.026 1.176 0.004 v('hard',)19 : 0.933 0.101 0.687 0.887 0.950 0.997 1.098 0.006 v('hard',)2 : 0.989 0.093 0.821 0.933 0.983 1.035 1.204 0.005 v('hard',)20 : 0.981 0.094 0.798 0.927 0.978 1.024 1.197 0.005 v('hard',)21 : 0.995 0.096 0.816 0.937 0.990 1.042 1.213 0.005 v('hard',)22 : 0.936 0.097 0.695 0.888 0.952 1.001 1.095 0.005 v('hard',)23 : 0.958 0.088 0.755 0.912 0.968 1.010 1.129 0.005 v('hard',)24 : 0.971 0.085 0.788 0.925 0.974 1.019 1.147 0.004 v('hard',)25 : 0.961 0.089 0.773 0.909 0.968 1.011 1.134 0.004 v('hard',)26 : 0.977 0.087 0.794 0.930 0.977 1.025 1.168 0.004 v('hard',)27 : 0.995 0.096 0.830 0.939 0.987 1.037 1.234 0.005 v('hard',)28 : 0.957 0.090 0.756 0.909 0.966 1.009 1.133 0.004 v('hard',)29 : 0.963 0.093 0.765 0.917 0.970 1.014 1.135 0.005 v('hard',)3 : 0.970 0.093 0.784 0.919 0.972 1.018 1.163 0.004 v('hard',)30 : 0.950 0.093 0.748 0.896 0.961 1.007 1.117 0.005 v('hard',)31 : 0.959 0.087 0.780 0.910 0.965 1.011 1.134 0.004 v('hard',)32 : 0.958 0.087 0.770 0.907 0.965 1.011 1.128 0.004 v('hard',)33 : 0.983 0.093 0.816 0.930 0.982 1.029 1.188 0.005 v('hard',)34 : 0.969 0.088 0.785 0.921 0.973 1.016 1.144 0.004 v('hard',)35 : 0.965 0.086 0.789 0.913 0.971 1.015 1.142 0.004 v('hard',)36 : 0.949 0.090 0.751 0.901 0.961 1.001 1.103 0.005 v('hard',)37 : 0.908 0.111 0.639 0.856 0.933 0.982 1.067 0.007 v('hard',)38 : 0.938 0.091 0.727 0.894 0.954 0.997 1.089 0.005 v('hard',)39 : 0.969 0.091 0.786 0.922 0.975 1.016 1.150 0.005 v('hard',)4 : 0.983 0.090 0.807 0.934 0.979 1.029 1.180 0.004 v('hard',)5 : 0.965 0.087 0.788 0.915 0.969 1.013 1.152 0.004 v('hard',)6 : 0.961 0.087 0.782 0.909 0.966 1.013 1.138 0.004 v('hard',)7 : 0.973 0.088 0.794 0.924 0.975 1.018 1.173 0.004 v('hard',)8 : 0.989 0.095 0.817 0.935 0.982 1.028 1.230 0.005 v('hard',)9 : 0.969 0.089 0.780 0.918 0.972 1.021 1.148 0.004 vvar('easy',): 0.111 0.053 0.030 0.068 0.107 0.145 0.221 0.004 vvar('hard',): 0.070 0.047 0.017 0.033 0.058 0.097 0.185 0.004 logp: -1161.693344 DIC: 2882.445038 Plotting t Plotting a Plotting v('hard',) Plotting v('easy',) Plotting posterior predictive... The first you can see when examining the recovered parameter values is that the mean of the group distributions (i.e. a, t, v('hard',) and v('easy',)) is that they match very well the parameters we used to generate the data from. So by pooling the very little data we had on each subject we can make useful inference about the group parameters. The second thing you can see is that individual subject parameters (ending with the index of the subject) are very close to the group mean (this is also indicated by the fact that the var posteriors, the variance of the group distribution -- representing the spread of the individual subject parameters). This property is called *shrinkage*. Intuitively, if we can not make meaningful inference about individual subjects we will assume that they are distributed as the rest of the group. HDDM-0.2/docs/tutorial_config_difficulty.rst0000664000102500010250000002107011727451403021727 0ustar wieckiwiecki00000000000000.. index:: Tutorial .. _chap_tutorial_config: ***************** A note of caution ***************** Although HDDM tries to make hierarchical Bayesian estimation as easy and automatic as possible, the statistical methods used to estimate the posterior (i.e. Markov-Chain Monte Carlo) rely on certain assumptions (e.g. chain-convergence). Although we encourage everyone to try this software, we would like to stress that you are responsible for making sure that the model actually works. In order to assess whether the necessary conditions for interpreting your results are met you have to have some basic grasp on what those methods do and on what assumptions they rest. There are multiple books introducing you to the world of hierarchical Bayesian estimation. Two of which we like are: `A Practical Course in Bayesian Graphical Modeling`_ by E.J. Wagenmakers and M. Lee `Doing Bayesian Data Analysis\: A Tutorial with R and BUGS`_ by J. Kruschke **************************************** Getting started: Creating a simple model **************************************** Imagine that we collected data from one subject on the moving dots or coherent motion task (e.g., Roitman & Shadlen, 2002). In this task, participants indicate via keypress in which of two directions dots are moving on a screen. Only some of the dots are moving in a coherent direction; depending on the difficulty, more or less of the dots are moving in random directions (see the figure). In our working example, we used two conditions, an easy and a hard condition. In the following, we will walk through the steps on creating a model in HDDM to estimate the underlying psychological decision making parameters of this task. .. figure:: moving_dots.jpg :scale: 20% The easiest way to use HDDM is to create a configuration file. First, you have to prepare your data to be in a specific format (csv). The data that we use here were generated from simulated DDM processes (i.e. they are not real data), so that we know the true underlying generative parameters. The data file can be found in the examples directory and is named simple_difficulty.csv (under Windows, these files can probably be found in ``C:\Python27\Lib\site-packages\hddm\examples``). Lets take a look at what it looks like: .. literalinclude :: ../hddm/examples/simple_difficulty.csv :lines: 1,101-105,579-582 The first line contains the header and specifies which columns contain which data. IMPORTANT: There must be one column named 'rt' and one named 'response'. The rows following the header contain the response made (e.g. 1=correct, 0=error or 1=left, 0=right), followed by the reaction time of the trial, followed by the difficulty of the trial. Note, that 'difficulty' is just an example, you can call them whatever you like. The following configuration file specifies a model in which drift-rate depends on difficulty: .. literalinclude :: ../hddm/examples/simple_difficulty.conf The (optional) tag [depends] specifies DDM parameters that depend on data. In our case, we want to estimate separate drift-rates (v) for the conditions found in the data column 'difficulty'. The optional [mcmc] tag specifies parameter of the Markov chain Monte-Carlo estimation such as how many samples to draw from the posterior and how many samples to discard as burn-in (as in any MCMC case, often it takes the MCMC chains some time to converge to the true posterior so that one would not want to use the initial samples to draw inferences about the true parameters; for details please read up on MCMC approaches). Note that you can also specify these parameters via the command line. Our model specification is now complete and we can fit the model by calling hddmfit: :: hddm_fit.py simple_difficulty.conf simple_difficulty.csv The first argument tells HDDM which model specification to use, the second argument is the data file to apply the model to. Calling hddmfit in this way will generate the following output (note that the numbers will be slightly different each time you run this): :: Creating model... Sampling: 100% [0000000000000000000000000000000000] Iterations: 10000 name mean std 2.5q 25q 50q 75q 97.5 mc_err a : 2.029 0.034 1.953 2.009 2.028 2.049 2.090 0.002 t : 0.297 0.007 0.282 0.292 0.297 0.302 0.311 0.001 v('easy',): 0.992 0.051 0.902 0.953 0.987 1.028 1.102 0.003 v('hard',): 0.522 0.049 0.429 0.485 0.514 0.561 0.612 0.002 logp: -1171.276303 DIC: 2329.069932 The parameters of DDM are usually abbreviated and have the following meaning: * a: threshold * t: non-decision time * v: drift-rate * z: bias (optional) * V: inter-trial variability in drift-rate (optional) * Z: inter-trial variability in bias (optional) * T: inter-trial variability in non-decision time (optional) Because we used simulated data in this example, we know the true parameters that generated the data (i.e. a=2, t=0.3, v_easy=1, v_hard=0.5). As you can see, the mean posterior values are very close to the true parameters -- our estimation worked! However, often we are not only interested in the best fitting value but also how confident we are in that estimation and how good other values are fitting. This is one of advantages of the Bayesian approach -- it gives us the complete posterior distribution, rather than just a single best guess (in principle the maximum likelihood estimate could be only slightly better fit than a range of other values, in which case the posterior would have high variance). As such the next columns are statistics on the shape of the distribution, such as the standard deviation and different quantiles to give you a feel for how certain you can be in the estimates. Lastly, logp and DIC give you a measure of how well the model fits the data. These values are not all that useful if looked at in isolation but they provide a tool to do model comparison. Logp is the summed log-likelihood of the best-fitting values (higher is better). DIC stands for deviance information criterion and is a measure that takes model complexity into account for bayesian models, similar to BIC or AIC (see here for the relationship http://www.mrc-bsu.cam.ac.uk/bugs/winbugs/dicpage.shtml). Lower values are better. :Exercise: Create a new model that ignores the different difficulties (i.e. only estimate a single drift-rate). Compare the resulting DIC score with that of the previous model -- does the increased complexity of the first model result in a sufficient increase in model fit to justify using it? Why does the drift-rate estimate of the second model make sense? Output plots ************ In addition, HDDM generates some useful plots such as the posterior predictive probability density on top of the normalized RT distribution for each condition: .. figure:: ../hddm/examples/plots/simple_difficulty_easy.png :scale: 40% .. figure:: ../hddm/examples/plots/simple_difficulty_hard.png :scale: 40% Note that error responses have been mirrored along the y-axis (on the left) to display both RT distributions in one plot. These plots allow you to see how good the estimation fits your data. Here, we also see that our subject makes more errors and are slower in the difficult condition. This combination is well captured by the reduced estimated drift-rate parameter in this condition. Moreover, HDDM generates the trace and histogram of the posterior samples. As pointed out in the introduction, we can rarely compute the posterior analytically so we have to estimate it. One standard method is MCMC which allows you to draw samples from the posterior. On the left side of the plot we see the trace of this sampling. The main thing to look out for is if the chain drifts around such that the mean value is not stable (i.e. has not converged) or if there are periods where it seems stuck in one place (i.e. proposal distribution too wide). In our case the chain of the parameter "a" (threshold) seems to have converged nicely to the correct value. This is also illustrated in the right side plot which is the histogram of the trace and gives a feel for how to the posterior distribution looks like. In our case, it looks like a normal distribution centered around a value close to 2. .. figure:: ../hddm/examples/plots/simple_difficulty_trace_a.png :scale: 40% Now we are ready for :ref:`part two of the tutorial `. .. _A Practical Course in Bayesian Graphical Modeling: http://www.ejwagenmakers.com/BayesCourse/BayesBook.html .. _Doing Bayesian Data Analysis\: A Tutorial with R and BUGS: http://www.indiana.edu/~kruschke/DoingBayesianDataAnalysis/ HDDM-0.2/docs/manual.rst0000664000102500010250000000123711713025571015573 0ustar wieckiwiecki00000000000000hddm Package ============ :mod:`hddm` Package ------------------- .. automodule:: hddm :members: :undoc-members: :show-inheritance: :mod:`model` Module ------------------- .. automodule:: hddm.model :members: :undoc-members: :show-inheritance: :mod:`diag` Module ------------------ .. automodule:: hddm.diag :members: :mod:`generate` Module ---------------------- .. automodule:: hddm.generate :members: :mod:`likelihoods` Module ------------------------- .. automodule:: hddm.likelihoods :members: :undoc-members: :show-inheritance: :mod:`utils` Module ------------------- .. automodule:: hddm.utils :members: HDDM-0.2/docs/tutorial.rst0000664000102500010250000000025711713025571016162 0ustar wieckiwiecki00000000000000.. index:: Tutorial .. _chap_tutorial: ######## Tutorial ######## .. toctree:: :maxdepth: 2 tutorial_config_difficulty tutorial_config_subjects tutorial_python HDDM-0.2/docs/abstract.rst0000664000102500010250000000153311713025571016120 0ustar wieckiwiecki00000000000000Abstract ======== Diffusion models are used to infer psychological processes underlying decision making. Efficient open source software to fit these models to data has been made available in different programming languages. However, current estimation methods require an abundance of reaction time measurements to recover meaningful parameters. Here, we present HDDM which stands for Hierarchical Drift Diffusion Modeling and allows fast and easy bayesian estimation of the model parameters using PyMC. Instead of estimating parameters for each subject individually like other tools, HDDM allows simultanious estimation of group and subject parameters and thus requires less data per subject. HDDM does not only provide best fitting parameters, but rather the whole posterior distribution of each parameter which allows intuitive goodness-of-fit statistics. HDDM-0.2/docs/intro.rst0000664000102500010250000001677611727453546015503 0ustar wieckiwiecki00000000000000============ Introduction ============ *NOTE*: This document is still under development. Diffusion models have established themselves as the de-facto standard for modeling data from simple decision making tasks. Each decision is modeled as a drift process which generates a response once a boundary is crossed. This simple assumption about the underlying psychological process has the intriguing property of reproducing reaction time distributions and choice probability in simple two-choice decision making tasks. Hierarchical Bayesian methods are quickly gaining popularity in cognitive sciences. Traditionally, psychological models where either fit separately to individual subjects (thus not taking similarities of subjects into account) or to the whole group (thus not taking differences of subjects into account). Hierarchical Bayesian methods provide a remedy for this problem by allowing group and subject parameters to be estimated simultaniously at different hierarchies. In essence, subject parameters are assumed to come from a group distribution. In addition, because these methods are Bayesian they deal naturally with uncertainty and variability in the parameter estimations. HDDM_ (Hierarchical Drift Diffusion Modeling) is an open-source software package written in Python_ which allows (i) the construction of hierarchical Bayesian drift models and (ii) the estimation of its posterior parameter distributions via PyMC_. For efficiency, all runtime critical functions are coded in Cython_, heavily optimized and compiled natively. User-defined models can be constructed via a simple configuration files or directly via HDDM library calls. To assess model fit, HDDM generates different statistics and comes with various plotting capabilities. For illustrative purposes, HDDM includes a graphical demo applet which simulates individual drift processes under user-specified parameter combinations. The code is test-covered to assure correct function and is properly documented. Online documentation and tutorials are provided. In sum, the presented software allows researches to construct and fit complex, user-specified models using state-of-the-art estimation methods without requiring a strong computer science or math background. ---------------- Diffusion Models ---------------- HDDM implements two types of drift models, (i) the Ratcliff drift diffusion model (DDM) and (ii) the linear ballistic accumulator (LBA). Both of these models implement decision making as an evidence accumulation process that executes a response upon crossing one of two decision boundaries. The speed of which the accumulation process approaches one of the two boundaries is called the drift rate *v* and influences how swiftly a particular reponse is executed. The distance between the two boundaries (i.e. threshold *a*) influences how much evidence must be accumulated until a reponse is executed. A lower threshold makes responding faster in general but more random while a higher threshold leads to more cautious responding. Reaction time, however, is not solely comprised of the decision making process -- perception, movement initiation and execution all take time and are summarized into one variable called non-decision time *ter*. Ratcliff Drift Diffusion Model ------------------------------ The Ratcliff DDM models decision making in two-choice tasks. Each choice is represented as and upper and lower boundary. A drift process accumulates evidence over time until it crosses one of the two boundaries and initiates the corresponding response. Because there is noise in the drift process, the time of the boundary crossing and the selected response will vary between trials. The starting point of the drift process relative to the two boundaries can influence if one response has a prepotent bias *z*. This pattern gives rise to the reaction time distributions of both choices and will henceforth be called the simple DDM (see figure 1). .. figure:: DDM_drifts_w_labels.svg Later on, it was noticed that this simple DDM could not account for two phenomena observed in decision making tasks -- early and late errors. This lead to the inclusion of inter-trial variability in the drift-rate, the non-decision time and the starting point. Models that take this into account are henceforth called full DDM. Linear Ballistic Accumulator ---------------------------- While the Linear Ballistic Accumulator (LBA) model also belongs to the class of drift models, it separates itself from the Ratcliff DDM in multiple ways. First, instead of one drift process and two boundaries, the LBA contains one drift process for each response. Thus, the LBA can model decision making when more than two responses are possible Moreover, unlike the DDM, the LBA drift process has no intra-trial variance. RT variability is obtained by including inter-trial variability in the drift-rate and the starting point distribution. ------------------------------ Hierarchical Bayesian Modeling ------------------------------ Bayesian estimation of posterior distributions differs from other approaches such as maximum likelihood (ML) by not only finding the one parameter value with the highest probability, but instead estimating the likelihood of each parameter value (i.e. the posterior distribution). The posterior can be computed with Bayes formula: .. math:: P(\theta|data) = \frac{P(data|\theta) * P(\theta)}{P(data)} Where P(data|\theta) is the likelihood and P(\theta) is the prior probability. To compute P(data) we have to integrate (or sum in the discrete case) across all parameter values. Note that in most scenarios this integral is analytically intractable. Sampling methods like Markov-Chain Monte Carlo (MCMC) have been developed that instead approximate the posterior directly by drawing samples from it, rather than analytically solving the above formula. These methods have been used with great success in many different scenarios. The theoretical background of these sampling methods is out of the scope of this paper. Another nice property of the Bayesian method is that it lends itself naturally to a hierarchical design. In such a design, parameters for one distribution can themselves come from a different distribution which allows chaining together of distributions of arbitrary complexity and map the structure of the data onto the model. This hierarchical property has a particular benefit to cognitive modeling where data is often scarce. Traditionally, multiple subjects are tested on the same task and a model is fit to explain the subject's behavior. Using maximum likelihood we can either fit individual models to individual subjects and neglect that they are most likely more similar than not and that what we learn from one subject could be used for inference on other subjects. Alternatively, we could fit one model to all subjects and neglect the individual differences that most likely exist. A hierarchical model helps us to capture this specific design of our data: we can create one distribution for each individual subject parameter but let the parameters for these distributions be modeled according to a group distribution. This way, the individual subject distributions restrict the group distribution and vice versa. ---------------------------------------------- Hierarchical Bayesian Drift Diffusion Modeling ---------------------------------------------- The structure of our simple hierarchical DDM can be appreciated in figure 2. .. figure:: hier_model.svg .. _HDDM: http://github.com/twiecki/hddm .. _Python: http://www.python.org/ .. _PyMC: http://code.google.com/p/pymc/ .. _Cython: http://www.cython.org/ HDDM-0.2/docs/index.rst0000664000102500010250000000071211713025571015422 0ustar wieckiwiecki00000000000000.. HDDM documentation master file, created by sphinx-quickstart on Sat Sep 11 16:47:10 2010. You can adapt this file completely to your liking, but it should at least contain the root `toctree` directive. .. include :: ../README.rst Table of Contents: ================== .. toctree:: :maxdepth: 2 abstract intro tutorial manual Indices and tables ================== * :ref:`genindex` * :ref:`modindex` * :ref:`search` HDDM-0.2/docs/tutorial_python.rst0000664000102500010250000000614011726176127017570 0ustar wieckiwiecki00000000000000.. index:: Tutorial .. _chap_tutorial_python: **************************** Specifiying Models in Python **************************** As an alternative to the configuration file, HDDM offers model specification directly from Python. For this, you first import hddm: .. literalinclude :: ../hddm/examples/simple_model.py :lines: 1 Next, we have to load the data into Python. HDDM expects a NumPy structured array which you can either create yourself or load it from a csv file. Information on how to create a proper structured NumPy array can be found here. If you want to load a csv file make sure it is in the proper format outlined above. You can then load the data as follows: .. literalinclude :: ../hddm/examples/simple_model.py :lines: 4 After you loaded the data you can create the model object which is called Multi because it allows you to dynamically create multiple HDDM models depending on your data. In the simplest case, you'll want to create a simple DDM (default): .. literalinclude :: ../hddm/examples/simple_model.py :lines: 7 You may then sample from the posterior distribution by calling: .. literalinclude :: ../hddm/examples/simple_model.py :lines: 10 Depending on the model and amount of data this can take some time. After enough samples were generated, you may want to print some statistics on the screen: .. literalinclude :: ../hddm/examples/simple_model.py :lines: 13 You can currently generate two plots to examine model fit. If you want to see if your chains converged and what the posteriors for each parameter look like you can call: .. literalinclude :: ../hddm/examples/simple_model.py :lines: 16 To see how well the RT distributions are fit by the mean of the posterior distribution we can plot the theoretical RT distribution on top of our empirical RT distribution by calling: .. literalinclude :: ../hddm/examples/simple_model.py :lines: 17 The closer the two distributions look like, the better the fit. Note that the RT distribution for the second response is mirrored on the y-axis. The final program then looks as follows: .. literalinclude :: ../hddm/examples/simple_model.py More complex models can be generated by specifiying different paramters during model creation. Say we wanted to create a model where each subject receives its own set of parameters which are themselves sampled from a group parameter distribution. Morever, as in the example above, we have two trial types in our data, easy and hard. Based on previous research, we assume that difficulty affects drift-rate 'v'. Thus, we want to fit different drift rate parameters for those two conditions while keeping the other parameters fixed across conditions. Finally, we want to use the full DDM with inter-trial variability for drift, non-decision time ter and starting point z. The full model requires integration of these variability parameters and is hence much slower. The model creation and sampling then might look like this (assuming we imported hddm and loaded the data as above): >>> model = hddm.HDDM(data, include=('V','Z','T'), bias=True, depends_on={'v':'difficulty'}) >>> model.sample(10000, burn=5000) HDDM-0.2/src/0000775000102500010250000000000011727455647013440 5ustar wieckiwiecki00000000000000HDDM-0.2/src/pdf.pxi0000664000102500010250000001156411713025572014723 0ustar wieckiwiecki00000000000000#cython: embedsignature=True #cython: cdivision=True #cython: wraparound=False #cython: boundscheck=False cimport cython include "integrate.pxi" #from libc.math cimport tan, sin, cos, log, exp, sqrt, fmax, pow, ceil, floor, fabs, M_PI cdef extern from "math.h" nogil: double sin(double) double cos(double) double log(double) double exp(double) double sqrt(double) double fmax(double, double) double pow(double, double) double ceil(double) double floor(double) double fabs(double) double M_PI cdef double ftt_01w(double tt, double w, double err) nogil: """Compute f(t|0,1,w) for the likelihood of the drift diffusion model using the method and implementation of Navarro & Fuss, 2009. """ cdef double kl, ks, p cdef int k, K, lower, upper # calculate number of terms needed for large t if M_PI*tt*err<1: # if error threshold is set low enough kl=sqrt(-2*log(M_PI*tt*err)/(M_PI**2*tt)) # bound kl=fmax(kl,1./(M_PI*sqrt(tt))) # ensure boundary conditions met else: # if error threshold set too high kl=1./(M_PI*sqrt(tt)) # set to boundary condition # calculate number of terms needed for small t if 2*sqrt(2*M_PI*tt)*err<1: # if error threshold is set low enough ks=2+sqrt(-2*tt*log(2*sqrt(2*M_PI*tt)*err)) # bound ks=fmax(ks,sqrt(tt)+1) # ensure boundary conditions are met else: # if error threshold was set too high ks=2 # minimal kappa for that case # compute f(tt|0,1,w) p=0 #initialize density if ks(ceil(ks)) # round to smallest integer meeting error lower = (-floor((K-1)/2.)) upper = (ceil((K-1)/2.)) for k from lower <= k <= upper: # loop over k p+=(w+2*k)*exp(-(pow((w+2*k),2))/2/tt) # increment sum p/=sqrt(2*M_PI*pow(tt,3)) # add constant term else: # if large t is better... K=(ceil(kl)) # round to smallest integer meeting error for k from 1 <= k <= K: p+=k*exp(-(pow(k,2))*(M_PI**2)*tt/2)*sin(k*M_PI*w) # increment sum p*=M_PI # add constant term return p cdef inline double prob_ub(double v, double a, double z) nogil: """Probability of hitting upper boundary.""" return (exp(-2*a*z*v) - 1) / (exp(-2*a*v) - 1) cdef double pdf(double x, double v, double a, double w, double err) nogil: """Compute the likelihood of the drift diffusion model f(t|v,a,z) using the method and implementation of Navarro & Fuss, 2009. """ if x <= 0: return 0 cdef double tt = x/a**2 # use normalized time cdef double p = ftt_01w(tt, w, err) #get f(t|0,1,w) # convert to f(t|v,a,w) return p*exp(-v*a*w -(pow(v,2))*x/2.)/(pow(a,2)) cdef double pdf_V(double x, double v, double V, double a, double z, double err) nogil: """Compute the likelihood of the drift diffusion model f(t|v,a,z,V) using the method and implementation of Navarro & Fuss, 2009. V is the std of the drift rate """ if x <= 0: return 0 if V==0: return pdf(x, v, a, z, err) cdef double tt = x/(pow(a,2)) # use normalized time cdef double p = ftt_01w(tt, z, err) #get f(t|0,1,w) # convert to f(t|v,a,w) return exp(log(p) + ((a*z*V)**2 - 2*a*v*z - (v**2)*x)/(2*(V**2)*x+2))/sqrt((V**2)*x+1)/(a**2) cpdef double full_pdf(double x, double v, double V, double a, double z, double Z, double t, double T, double err, int nT=2, int nZ=2, bint use_adaptive=1, double simps_err=1e-3) nogil: """full pdf""" # Check if parpameters are valid if z<0 or z>1 or a<0 or ((fabs(x)-(t-T/2.))<0) or (z+Z/2.>1) or (z-Z/2.<0) or (t-T/2.<0) or (t<0): return 0 # transform x,v,z if x is upper bound response if x > 0: v = -v z = 1.-z x = fabs(x) if T<1e-3: T = 0 if Z <1e-3: Z = 0 if (Z==0): if (T==0): #V=0,Z=0,T=0 return pdf_V(x - t, v, V, a, z, err) else: #V=0,Z=0,T=$ if use_adaptive>0: return adaptiveSimpsons_1D(x, v, V, a, z, t, err, z, z, t-T/2., t+T/2., simps_err, nT) else: return simpson_1D(x, v, V, a, z, t, err, z, z, 0, t-T/2., t+T/2., nT) else: #Z=$ if (T==0): #V=0,Z=$,T=0 if use_adaptive: return adaptiveSimpsons_1D(x, v, V, a, z, t, err, z-Z/2., z+Z/2., t, t, simps_err, nZ) else: return simpson_1D(x, v, V, a, z, t, err, z-Z/2., z+Z/2., nZ, t, t , 0) else: #V=0,Z=$,T=$ if use_adaptive: return adaptiveSimpsons_2D(x, v, V, a, z, t, err, z-Z/2., z+Z/2., t-T/2., t+T/2., simps_err, nZ, nT) else: return simpson_2D(x, v, V, a, z, t, err, z-Z/2., z+Z/2., nZ, t-T/2., t+T/2., nT) HDDM-0.2/src/wfpt_full.pyx0000664000102500010250000001312011713025572016162 0ustar wieckiwiecki00000000000000#!/usr/bin/python # # Cython version of the Navarro & Fuss, 2009 DDM PDF. Based directly # on the following code by Navarro & Fuss: # http://www.psychocmath.logy.adelaide.edu.au/personalpages/staff/danielnavarro/resources/wfpt.m # # This implementation is about 170 times faT than the matlab # reference version. # # Copyleft Thomas Wiecki (thomas_wiecki[at]brown.edu), 2010 # GPLv3 from copy import copy import numpy as np cimport numpy as np cimport cython from cython.parallel import * include "pdf.pxi" @cython.wraparound(False) @cython.boundscheck(False) # turn of bounds-checking for entire function def wiener_like_full_intrp(np.ndarray[double, ndim=1] x, double v, double V, double a, double z, double Z, double t, double T, double err, int nT= 10, int nZ=10, bint use_adaptive=1, double simps_err=1e-8): cdef Py_ssize_t i cdef double p cdef double sum_logp = 0 for i in prange(x.shape[0], nogil=True): p = full_pdf(x[i], v, V, a, z, Z, t, T, err, nT, nZ, use_adaptive, simps_err) # If one probability = 0, the log sum will be -Inf if p == 0: with gil: return -np.inf sum_logp += log(p) return sum_logp @cython.wraparound(False) @cython.boundscheck(False) # turn of bounds-checking for entire function def wiener_like_full(np.ndarray[double, ndim=1] x, np.ndarray[double, ndim=1] v, np.ndarray[double, ndim=1] a, np.ndarray[double, ndim=1] z, np.ndarray[double, ndim=1] t, err): cdef Py_ssize_t i cdef double p cdef double sum_logp = 0 for i in prange(x.shape[0], nogil=True): p = pdf_sign(x[i], v[i], a[i], z[i], t[i], err) # If one probability = 0, the log sum will be -Inf if p == 0: with gil: return -np.inf sum_logp += log(p) return sum_logp @cython.wraparound(False) @cython.boundscheck(False) # turn of bounds-checking for entire function def wiener_like_full_multi(np.ndarray[double, ndim=1] x, v, V, a, z, Z, t, T, double err, multi=None): cdef unsigned int size = x.shape[0] cdef unsigned int i cdef double p = 0 if multi is None: return wiener_like_full_intrp(x, v, V, a, z, Z, t, T, err) else: params = {'v':v, 'z':z, 't':t, 'a':a, 'V':V, 'Z':Z, 'T':T} params_iter = copy(params) for i from 0 <= i < size: for param in multi: params_iter[param] = params[param][i] p += log(full_pdf(x[i], params_iter['v'], params_iter['V'], params_iter['a'], params_iter['z'], params_iter['Z'], params_iter['t'], params_iter['T'], err)) return p @cython.wraparound(False) @cython.boundscheck(False) # turn of bounds-checking for entire function def gen_rts_from_cdf(double v, double V, double a, double z, double Z, double t, \ double T, int samples=1000, double cdf_lb = -6, double cdf_ub = 6, double dt=1e-2): cdef np.ndarray[double, ndim=1] x = np.arange(cdf_lb, cdf_ub, dt) cdef np.ndarray[double, ndim=1] l_cdf = np.empty(x.shape[0], dtype=np.double) cdef double pdf, rt cdef Py_ssize_t i, j cdef int idx l_cdf[0] = 0 for i in range(x.shape[0]): pdf = full_pdf(x[i], v, V, a, z, Z, 0, 0, 1e-4) l_cdf[i] = l_cdf[i-1] + pdf l_cdf /= l_cdf[x.shape[0]-1] cdef np.ndarray[double, ndim=1] rts = np.empty(samples, dtype=np.double) cdef np.ndarray[double, ndim=1] f = np.random.rand(samples) cdef np.ndarray[double, ndim=1] delay if T!=0: delay = (np.random.rand(samples)*T + (t - T/2.)) for i from 0 <= i < samples: idx = np.searchsorted(l_cdf, f[i]) rt = x[idx] if T==0: rt = rt + np.sign(rt)*t else: rt = rt + np.sign(rt)*delay[i] rts[i] = rt return rts @cython.wraparound(False) @cython.boundscheck(False) # turn of bounds-checking for entire function def wiener_like_full_contaminant(np.ndarray[double, ndim=1] value, np.ndarray[int, ndim=1] cont_x, double gamma, double v, double V, double a, double z, double Z, double t, double T, double t_min, double t_max, double err): """Wiener likelihood function where RTs could come from a separate, uniform contaminant distribution. Reference: Lee, Vandekerckhove, Navarro, & Tuernlinckx (2007) """ cdef Py_ssize_t i cdef double p cdef double sum_logp = 0 cdef int n_cont = np.sum(cont_x) cdef int pos_cont = 0 for i in prange(value.shape[0], nogil=True): if cont_x[i] == 0: p = full_pdf(value[i], v, V, a, z, Z, t, T, err) sum_logp += log(p) elif value[i]>0: pos_cont += 1 # add the log likelihood of the contaminations #first the guesses sum_logp += n_cont*log(gamma*(0.5 * 1./(t_max-t_min))) #then the positive prob_boundary sum_logp += pos_cont*log((1-gamma) * prob_ub(v, a, z) * 1./(t_max-t_min)) #and the negative prob_boundary sum_logp += (n_cont - pos_cont)*log((1-gamma)*(1-prob_ub(v, a, z)) * 1./(t_max-t_min)) return sum_logp def wiener_like_full_single(double x, double v, double V, double a, double z, double Z, double t, double T, double err, int nT=2, int nZ=2, bint use_adaptive=1, double simps_err=1e-3): # Wrapper for c-only full_pdf function, only required for unittest return full_pdf(x, v, V, a, z, Z, t, T, err, nT, nZ, use_adaptive, simps_err) HDDM-0.2/src/lba.pyx0000664000102500010250000001017611713025572014726 0ustar wieckiwiecki00000000000000#!/usr/bin/python # # Cython version of the Navarro & Fuss, 2009 DDM PDF. Based directly # on the following code by Navarro & Fuss: # http://www.psychocmath.logy.adelaide.edu.au/personalpages/staff/danielnavarro/resources/wfpt.m # # # Copyleft Thomas Wiecki (thomas_wiecki[at]brown.edu), 2010 # GPLv3 from __future__ import division from copy import copy import numpy as np cimport numpy as cnp import scipy as sp import scipy.stats from scipy.stats.distributions import norm cimport cython # Define data type DTYPE = np.double ctypedef cnp.double_t DTYPE_t cdef double PI = 3.1415926535897 cdef extern from "math.h": double sin(double) double log(double) double exp(double) double sqrt(double) double fmax(double, double) double pow(double, double) double ceil(double) double floor(double) double fabs(double) double erf(double) cdef inline double norm_pdf(double x): return 0.3989422804014327 * exp(-(x**2/2)) cdef inline double norm_cdf(double x): return .5 * (1 + erf((x)/1.4142135623730951)) cdef DTYPE_t fptcdf_single(DTYPE_t t, double z, double a, double driftrate, double sddrift): cdef double zs=t*sddrift cdef double zu=t*driftrate cdef double aminuszu=a-zu cdef double xx=aminuszu-z cdef double azu=aminuszu/zs cdef double azumax=xx/zs cdef double tmp1 = zs * (norm_pdf(azumax) - norm_pdf(azu)) cdef double tmp2 = (xx * norm_cdf(azumax)) - (aminuszu * norm_cdf(azu)) return 1+(tmp1+tmp2)/z cdef DTYPE_t fptpdf_single(DTYPE_t t, double z, double a, double driftrate, double sddrift): cdef double zs=t*sddrift cdef double zu=t*driftrate cdef double aminuszu=a-zu cdef double azu=aminuszu/zs cdef double azumax=(aminuszu-z)/zs return (driftrate*(norm_cdf(azu)-norm_cdf(azumax)) + sddrift*(norm_pdf(azumax)-norm_pdf(azu)))/z @cython.wraparound(False) @cython.boundscheck(False) # turn of bounds-checking for entire function def lba_like(cnp.ndarray[DTYPE_t, ndim=1] value, double z, double a, double ter, double sv, double v0, double v1, unsigned int logp=0, unsigned int normalize_v=0): cdef cnp.ndarray[DTYPE_t, ndim=1] rt = (np.abs(value) - ter) cdef unsigned int nresp = value.shape[0] cdef unsigned int i = 0 cdef cnp.ndarray[DTYPE_t, ndim=1] probs = np.empty_like(value) assert sv >= 0, "sv must be larger than 0" if normalize_v == 1: v1 = 1 - v0 if a <= z: #print "Starting point larger than threshold!" return -np.Inf #print "z: %f, a: %f, ter: %i, v: %f, sv: %f" % (z, a, ter, v[0], sv) for i from 0 <= i < nresp: if value[i] > 0: probs[i] = (1-fptcdf_single(rt[i], z, a, v1, sv)) * \ fptpdf_single(rt[i], z, a, v0, sv) else: probs[i] = (1-fptcdf_single(rt[i], z, a, v0, sv)) * \ fptpdf_single(rt[i], z, a, v1, sv) if logp == 1: return np.sum(np.log(probs)) else: return probs # Backed up if in the future multiple responses are to be supported @cython.boundscheck(False) # turn of bounds-checking for entire function def lba_like_old(cnp.ndarray[DTYPE_t, ndim=1] value, cnp.ndarray[DTYPE_t, ndim=1] resps, double z, double a, double ter, cnp.ndarray[DTYPE_t, ndim=1] drift, double sv, unsigned int logp=0): # Rescale parameters so as to fit in the same range as the DDM cdef cnp.ndarray[DTYPE_t, ndim=1] rt = (np.abs(value) - ter) cdef unsigned int nresp = value.shape[0] cdef unsigned int i cdef cnp.ndarray[DTYPE_t, ndim=1] probs = np.empty(nresp, dtype=DTYPE) if a <= z: #print "Starting point larger than threshold!" return -np.Inf #print "z: %f, a: %f, ter: %i, v: %f, sv: %f" % (z, a, ter, v[0], sv) for i from 0 <= i < nresp: if resps[i] == 1: probs[i] = (1-fptcdf_single(rt[i], z, a, drift[1], sv)) * \ fptpdf_single(rt[i], z, a, drift[0], sv) else: probs[i] = (1-fptcdf_single(rt[i], z, a, drift[0], sv)) * \ fptpdf_single(rt[i], z, a, drift[1], sv) if logp == 1: return np.sum(np.log(probs)) else: return probs HDDM-0.2/src/wfpt.c0000664000102500010250000203174711726417112014562 0ustar wieckiwiecki00000000000000/* Generated by Cython 0.15.1+ on Fri Mar 9 10:14:18 2012 */ #define PY_SSIZE_T_CLEAN #include "Python.h" #ifndef Py_PYTHON_H #error Python headers needed to compile C extensions, please install development version of Python. #elif PY_VERSION_HEX < 0x02040000 #error Cython requires Python 2.4+. #else #include /* For offsetof */ #ifndef offsetof #define offsetof(type, member) ( (size_t) & ((type*)0) -> member ) #endif #if !defined(WIN32) && !defined(MS_WINDOWS) #ifndef __stdcall #define __stdcall #endif #ifndef __cdecl #define __cdecl #endif #ifndef __fastcall #define __fastcall #endif #endif #ifndef DL_IMPORT #define DL_IMPORT(t) t #endif #ifndef DL_EXPORT #define DL_EXPORT(t) t #endif #ifndef PY_LONG_LONG #define PY_LONG_LONG LONG_LONG #endif #if PY_VERSION_HEX < 0x02050000 typedef int Py_ssize_t; #define PY_SSIZE_T_MAX INT_MAX #define PY_SSIZE_T_MIN INT_MIN #define PY_FORMAT_SIZE_T "" #define PyInt_FromSsize_t(z) PyInt_FromLong(z) #define PyInt_AsSsize_t(o) __Pyx_PyInt_AsInt(o) #define PyNumber_Index(o) PyNumber_Int(o) #define PyIndex_Check(o) PyNumber_Check(o) #define PyErr_WarnEx(category, message, stacklevel) PyErr_Warn(category, message) #endif #if PY_VERSION_HEX < 0x02060000 #define Py_REFCNT(ob) (((PyObject*)(ob))->ob_refcnt) #define Py_TYPE(ob) (((PyObject*)(ob))->ob_type) #define Py_SIZE(ob) (((PyVarObject*)(ob))->ob_size) #define PyVarObject_HEAD_INIT(type, size) \ PyObject_HEAD_INIT(type) size, #define PyType_Modified(t) typedef struct { void *buf; PyObject *obj; Py_ssize_t len; Py_ssize_t itemsize; int readonly; int ndim; char *format; Py_ssize_t *shape; Py_ssize_t *strides; Py_ssize_t *suboffsets; void *internal; } Py_buffer; #define PyBUF_SIMPLE 0 #define PyBUF_WRITABLE 0x0001 #define PyBUF_FORMAT 0x0004 #define PyBUF_ND 0x0008 #define PyBUF_STRIDES (0x0010 | PyBUF_ND) #define PyBUF_C_CONTIGUOUS (0x0020 | PyBUF_STRIDES) #define PyBUF_F_CONTIGUOUS (0x0040 | PyBUF_STRIDES) #define PyBUF_ANY_CONTIGUOUS (0x0080 | PyBUF_STRIDES) #define PyBUF_INDIRECT (0x0100 | PyBUF_STRIDES) #define PyBUF_RECORDS (PyBUF_STRIDES | PyBUF_FORMAT | PyBUF_WRITABLE) #define PyBUF_FULL (PyBUF_INDIRECT | PyBUF_FORMAT | PyBUF_WRITABLE) #endif #if PY_MAJOR_VERSION < 3 #define __Pyx_BUILTIN_MODULE_NAME "__builtin__" #define __Pyx_PyIdentifier_FromString(s) PyString_FromString(s) #define __Pyx_PyCode_New(a, k, l, s, f, code, c, n, v, fv, cell, fn, name, fline, lnos) \ PyCode_New(a, l, s, f, code, c, n, v, fv, cell, fn, name, fline, lnos) #else #define __Pyx_BUILTIN_MODULE_NAME "builtins" #define __Pyx_PyIdentifier_FromString(s) PyUnicode_FromString(s) #define __Pyx_PyCode_New(a, k, l, s, f, code, c, n, v, fv, cell, fn, name, fline, lnos) \ PyCode_New(a, k, l, s, f, code, c, n, v, fv, cell, fn, name, fline, lnos) #endif #if PY_MAJOR_VERSION < 3 && PY_MINOR_VERSION < 6 #define PyUnicode_FromString(s) PyUnicode_Decode(s, strlen(s), "UTF-8", "strict") #endif #if PY_MAJOR_VERSION >= 3 #define Py_TPFLAGS_CHECKTYPES 0 #define Py_TPFLAGS_HAVE_INDEX 0 #endif #if (PY_VERSION_HEX < 0x02060000) || (PY_MAJOR_VERSION >= 3) #define Py_TPFLAGS_HAVE_NEWBUFFER 0 #endif #if PY_VERSION_HEX > 0x03030000 && defined(PyUnicode_GET_LENGTH) #define CYTHON_PEP393_ENABLED #define __Pyx_PyUnicode_GET_LENGTH(u) PyUnicode_GET_LENGTH(u) #define __Pyx_PyUnicode_READ_CHAR(u, i) PyUnicode_READ_CHAR(u, i) #else #define __Pyx_PyUnicode_GET_LENGTH(u) PyUnicode_GET_SIZE(u) #define __Pyx_PyUnicode_READ_CHAR(u, i) ((Py_UCS4)(PyUnicode_AS_UNICODE(u)[i])) #endif #if PY_MAJOR_VERSION >= 3 #define PyBaseString_Type PyUnicode_Type #define PyStringObject PyUnicodeObject #define PyString_Type PyUnicode_Type #define PyString_Check PyUnicode_Check #define PyString_CheckExact PyUnicode_CheckExact #endif #if PY_VERSION_HEX < 0x02060000 #define PyBytesObject PyStringObject #define PyBytes_Type PyString_Type #define PyBytes_Check PyString_Check #define PyBytes_CheckExact PyString_CheckExact #define PyBytes_FromString PyString_FromString #define PyBytes_FromStringAndSize PyString_FromStringAndSize #define PyBytes_FromFormat PyString_FromFormat #define PyBytes_DecodeEscape PyString_DecodeEscape #define PyBytes_AsString PyString_AsString #define PyBytes_AsStringAndSize PyString_AsStringAndSize #define PyBytes_Size PyString_Size #define PyBytes_AS_STRING PyString_AS_STRING #define PyBytes_GET_SIZE PyString_GET_SIZE #define PyBytes_Repr PyString_Repr #define PyBytes_Concat PyString_Concat #define PyBytes_ConcatAndDel PyString_ConcatAndDel #endif #if PY_VERSION_HEX < 0x02060000 #define PySet_Check(obj) PyObject_TypeCheck(obj, &PySet_Type) #define PyFrozenSet_Check(obj) PyObject_TypeCheck(obj, &PyFrozenSet_Type) #endif #ifndef PySet_CheckExact #define PySet_CheckExact(obj) (Py_TYPE(obj) == &PySet_Type) #endif #define __Pyx_TypeCheck(obj, type) PyObject_TypeCheck(obj, (PyTypeObject *)type) #if PY_MAJOR_VERSION >= 3 #define PyIntObject PyLongObject #define PyInt_Type PyLong_Type #define PyInt_Check(op) PyLong_Check(op) #define PyInt_CheckExact(op) PyLong_CheckExact(op) #define PyInt_FromString PyLong_FromString #define PyInt_FromUnicode PyLong_FromUnicode #define PyInt_FromLong PyLong_FromLong #define PyInt_FromSize_t PyLong_FromSize_t #define PyInt_FromSsize_t PyLong_FromSsize_t #define PyInt_AsLong PyLong_AsLong #define PyInt_AS_LONG PyLong_AS_LONG #define PyInt_AsSsize_t PyLong_AsSsize_t #define PyInt_AsUnsignedLongMask PyLong_AsUnsignedLongMask #define PyInt_AsUnsignedLongLongMask PyLong_AsUnsignedLongLongMask #endif #if PY_MAJOR_VERSION >= 3 #define PyBoolObject PyLongObject #endif #if PY_VERSION_HEX < 0x03020000 typedef long Py_hash_t; #define __Pyx_PyInt_FromHash_t PyInt_FromLong #define __Pyx_PyInt_AsHash_t PyInt_AsLong #else #define __Pyx_PyInt_FromHash_t PyInt_FromSsize_t #define __Pyx_PyInt_AsHash_t PyInt_AsSsize_t #endif #if (PY_MAJOR_VERSION < 3) || (PY_VERSION_HEX >= 0x03010300) #define __Pyx_PySequence_GetSlice(obj, a, b) PySequence_GetSlice(obj, a, b) #define __Pyx_PySequence_SetSlice(obj, a, b, value) PySequence_SetSlice(obj, a, b, value) #define __Pyx_PySequence_DelSlice(obj, a, b) PySequence_DelSlice(obj, a, b) #else #define __Pyx_PySequence_GetSlice(obj, a, b) (unlikely(!(obj)) ? \ (PyErr_SetString(PyExc_SystemError, "null argument to internal routine"), (PyObject*)0) : \ (likely((obj)->ob_type->tp_as_mapping) ? (PySequence_GetSlice(obj, a, b)) : \ (PyErr_Format(PyExc_TypeError, "'%.200s' object is unsliceable", (obj)->ob_type->tp_name), (PyObject*)0))) #define __Pyx_PySequence_SetSlice(obj, a, b, value) (unlikely(!(obj)) ? \ (PyErr_SetString(PyExc_SystemError, "null argument to internal routine"), -1) : \ (likely((obj)->ob_type->tp_as_mapping) ? (PySequence_SetSlice(obj, a, b, value)) : \ (PyErr_Format(PyExc_TypeError, "'%.200s' object doesn't support slice assignment", (obj)->ob_type->tp_name), -1))) #define __Pyx_PySequence_DelSlice(obj, a, b) (unlikely(!(obj)) ? \ (PyErr_SetString(PyExc_SystemError, "null argument to internal routine"), -1) : \ (likely((obj)->ob_type->tp_as_mapping) ? (PySequence_DelSlice(obj, a, b)) : \ (PyErr_Format(PyExc_TypeError, "'%.200s' object doesn't support slice deletion", (obj)->ob_type->tp_name), -1))) #endif #if PY_MAJOR_VERSION >= 3 #define PyMethod_New(func, self, klass) ((self) ? PyMethod_New(func, self) : PyInstanceMethod_New(func)) #endif #if PY_VERSION_HEX < 0x02050000 #define __Pyx_GetAttrString(o,n) PyObject_GetAttrString((o),((char *)(n))) #define __Pyx_SetAttrString(o,n,a) PyObject_SetAttrString((o),((char *)(n)),(a)) #define __Pyx_DelAttrString(o,n) PyObject_DelAttrString((o),((char *)(n))) #else #define __Pyx_GetAttrString(o,n) PyObject_GetAttrString((o),(n)) #define __Pyx_SetAttrString(o,n,a) PyObject_SetAttrString((o),(n),(a)) #define __Pyx_DelAttrString(o,n) PyObject_DelAttrString((o),(n)) #endif #if PY_VERSION_HEX < 0x02050000 #define __Pyx_NAMESTR(n) ((char *)(n)) #define __Pyx_DOCSTR(n) ((char *)(n)) #else #define __Pyx_NAMESTR(n) (n) #define __Pyx_DOCSTR(n) (n) #endif #if PY_MAJOR_VERSION >= 3 #define __Pyx_PyNumber_Divide(x,y) PyNumber_TrueDivide(x,y) #define __Pyx_PyNumber_InPlaceDivide(x,y) PyNumber_InPlaceTrueDivide(x,y) #else #define __Pyx_PyNumber_Divide(x,y) PyNumber_Divide(x,y) #define __Pyx_PyNumber_InPlaceDivide(x,y) PyNumber_InPlaceDivide(x,y) #endif #ifndef __PYX_EXTERN_C #ifdef __cplusplus #define __PYX_EXTERN_C extern "C" #else #define __PYX_EXTERN_C extern #endif #endif #if defined(WIN32) || defined(MS_WINDOWS) #define _USE_MATH_DEFINES #endif #include #define __PYX_HAVE__wfpt #define __PYX_HAVE_API__wfpt #include "stdio.h" #include "stdlib.h" #include "numpy/arrayobject.h" #include "numpy/ufuncobject.h" #include "math.h" #ifdef _OPENMP #include #endif /* _OPENMP */ #ifdef PYREX_WITHOUT_ASSERTIONS #define CYTHON_WITHOUT_ASSERTIONS #endif /* inline attribute */ #ifndef CYTHON_INLINE #if defined(__GNUC__) #define CYTHON_INLINE __inline__ #elif defined(_MSC_VER) #define CYTHON_INLINE __inline #elif defined (__STDC_VERSION__) && __STDC_VERSION__ >= 199901L #define CYTHON_INLINE inline #else #define CYTHON_INLINE #endif #endif /* unused attribute */ #ifndef CYTHON_UNUSED # if defined(__GNUC__) # if !(defined(__cplusplus)) || (__GNUC__ > 3 || (__GNUC__ == 3 && __GNUC_MINOR__ >= 4)) # define CYTHON_UNUSED __attribute__ ((__unused__)) # else # define CYTHON_UNUSED # endif # elif defined(__ICC) || (defined(__INTEL_COMPILER) && !defined(_MSC_VER)) # define CYTHON_UNUSED __attribute__ ((__unused__)) # else # define CYTHON_UNUSED # endif #endif typedef struct {PyObject **p; char *s; const long n; const char* encoding; const char is_unicode; const char is_str; const char intern; } __Pyx_StringTabEntry; /*proto*/ /* Type Conversion Predeclarations */ #define __Pyx_PyBytes_FromUString(s) PyBytes_FromString((char*)s) #define __Pyx_PyBytes_AsUString(s) ((unsigned char*) PyBytes_AsString(s)) #define __Pyx_Owned_Py_None(b) (Py_INCREF(Py_None), Py_None) #define __Pyx_PyBool_FromLong(b) ((b) ? (Py_INCREF(Py_True), Py_True) : (Py_INCREF(Py_False), Py_False)) static CYTHON_INLINE int __Pyx_PyObject_IsTrue(PyObject*); static CYTHON_INLINE PyObject* __Pyx_PyNumber_Int(PyObject* x); static CYTHON_INLINE Py_ssize_t __Pyx_PyIndex_AsSsize_t(PyObject*); static CYTHON_INLINE PyObject * __Pyx_PyInt_FromSize_t(size_t); static CYTHON_INLINE size_t __Pyx_PyInt_AsSize_t(PyObject*); #define __pyx_PyFloat_AsDouble(x) (PyFloat_CheckExact(x) ? PyFloat_AS_DOUBLE(x) : PyFloat_AsDouble(x)) #define __pyx_PyFloat_AsFloat(x) ((float) __pyx_PyFloat_AsDouble(x)) #ifdef __GNUC__ /* Test for GCC > 2.95 */ #if __GNUC__ > 2 || (__GNUC__ == 2 && (__GNUC_MINOR__ > 95)) #define likely(x) __builtin_expect(!!(x), 1) #define unlikely(x) __builtin_expect(!!(x), 0) #else /* __GNUC__ > 2 ... */ #define likely(x) (x) #define unlikely(x) (x) #endif /* __GNUC__ > 2 ... */ #else /* __GNUC__ */ #define likely(x) (x) #define unlikely(x) (x) #endif /* __GNUC__ */ static PyObject *__pyx_m; static PyObject *__pyx_b; static PyObject *__pyx_empty_tuple; static PyObject *__pyx_empty_bytes; static int __pyx_lineno; static int __pyx_clineno = 0; static const char * __pyx_cfilenm= __FILE__; static const char *__pyx_filename; #if !defined(CYTHON_CCOMPLEX) #if defined(__cplusplus) #define CYTHON_CCOMPLEX 1 #elif defined(_Complex_I) #define CYTHON_CCOMPLEX 1 #else #define CYTHON_CCOMPLEX 0 #endif #endif #if CYTHON_CCOMPLEX #ifdef __cplusplus #include #else #include #endif #endif #if CYTHON_CCOMPLEX && !defined(__cplusplus) && defined(__sun__) && defined(__GNUC__) #undef _Complex_I #define _Complex_I 1.0fj #endif static const char *__pyx_f[] = { "wfpt.pyx", "pdf.pxi", "numpy.pxd", "integrate.pxi", }; /* "numpy.pxd":722 * # in Cython to enable them only on the right systems. * * ctypedef npy_int8 int8_t # <<<<<<<<<<<<<< * ctypedef npy_int16 int16_t * ctypedef npy_int32 int32_t */ typedef npy_int8 __pyx_t_5numpy_int8_t; /* "numpy.pxd":723 * * ctypedef npy_int8 int8_t * ctypedef npy_int16 int16_t # <<<<<<<<<<<<<< * ctypedef npy_int32 int32_t * ctypedef npy_int64 int64_t */ typedef npy_int16 __pyx_t_5numpy_int16_t; /* "numpy.pxd":724 * ctypedef npy_int8 int8_t * ctypedef npy_int16 int16_t * ctypedef npy_int32 int32_t # <<<<<<<<<<<<<< * ctypedef npy_int64 int64_t * #ctypedef npy_int96 int96_t */ typedef npy_int32 __pyx_t_5numpy_int32_t; /* "numpy.pxd":725 * ctypedef npy_int16 int16_t * ctypedef npy_int32 int32_t * ctypedef npy_int64 int64_t # <<<<<<<<<<<<<< * #ctypedef npy_int96 int96_t * #ctypedef npy_int128 int128_t */ typedef npy_int64 __pyx_t_5numpy_int64_t; /* "numpy.pxd":729 * #ctypedef npy_int128 int128_t * * ctypedef npy_uint8 uint8_t # <<<<<<<<<<<<<< * ctypedef npy_uint16 uint16_t * ctypedef npy_uint32 uint32_t */ typedef npy_uint8 __pyx_t_5numpy_uint8_t; /* "numpy.pxd":730 * * ctypedef npy_uint8 uint8_t * ctypedef npy_uint16 uint16_t # <<<<<<<<<<<<<< * ctypedef npy_uint32 uint32_t * ctypedef npy_uint64 uint64_t */ typedef npy_uint16 __pyx_t_5numpy_uint16_t; /* "numpy.pxd":731 * ctypedef npy_uint8 uint8_t * ctypedef npy_uint16 uint16_t * ctypedef npy_uint32 uint32_t # <<<<<<<<<<<<<< * ctypedef npy_uint64 uint64_t * #ctypedef npy_uint96 uint96_t */ typedef npy_uint32 __pyx_t_5numpy_uint32_t; /* "numpy.pxd":732 * ctypedef npy_uint16 uint16_t * ctypedef npy_uint32 uint32_t * ctypedef npy_uint64 uint64_t # <<<<<<<<<<<<<< * #ctypedef npy_uint96 uint96_t * #ctypedef npy_uint128 uint128_t */ typedef npy_uint64 __pyx_t_5numpy_uint64_t; /* "numpy.pxd":736 * #ctypedef npy_uint128 uint128_t * * ctypedef npy_float32 float32_t # <<<<<<<<<<<<<< * ctypedef npy_float64 float64_t * #ctypedef npy_float80 float80_t */ typedef npy_float32 __pyx_t_5numpy_float32_t; /* "numpy.pxd":737 * * ctypedef npy_float32 float32_t * ctypedef npy_float64 float64_t # <<<<<<<<<<<<<< * #ctypedef npy_float80 float80_t * #ctypedef npy_float128 float128_t */ typedef npy_float64 __pyx_t_5numpy_float64_t; /* "numpy.pxd":746 * # The int types are mapped a bit surprising -- * # numpy.int corresponds to 'l' and numpy.long to 'q' * ctypedef npy_long int_t # <<<<<<<<<<<<<< * ctypedef npy_longlong long_t * ctypedef npy_longlong longlong_t */ typedef npy_long __pyx_t_5numpy_int_t; /* "numpy.pxd":747 * # numpy.int corresponds to 'l' and numpy.long to 'q' * ctypedef npy_long int_t * ctypedef npy_longlong long_t # <<<<<<<<<<<<<< * ctypedef npy_longlong longlong_t * */ typedef npy_longlong __pyx_t_5numpy_long_t; /* "numpy.pxd":748 * ctypedef npy_long int_t * ctypedef npy_longlong long_t * ctypedef npy_longlong longlong_t # <<<<<<<<<<<<<< * * ctypedef npy_ulong uint_t */ typedef npy_longlong __pyx_t_5numpy_longlong_t; /* "numpy.pxd":750 * ctypedef npy_longlong longlong_t * * ctypedef npy_ulong uint_t # <<<<<<<<<<<<<< * ctypedef npy_ulonglong ulong_t * ctypedef npy_ulonglong ulonglong_t */ typedef npy_ulong __pyx_t_5numpy_uint_t; /* "numpy.pxd":751 * * ctypedef npy_ulong uint_t * ctypedef npy_ulonglong ulong_t # <<<<<<<<<<<<<< * ctypedef npy_ulonglong ulonglong_t * */ typedef npy_ulonglong __pyx_t_5numpy_ulong_t; /* "numpy.pxd":752 * ctypedef npy_ulong uint_t * ctypedef npy_ulonglong ulong_t * ctypedef npy_ulonglong ulonglong_t # <<<<<<<<<<<<<< * * ctypedef npy_intp intp_t */ typedef npy_ulonglong __pyx_t_5numpy_ulonglong_t; /* "numpy.pxd":754 * ctypedef npy_ulonglong ulonglong_t * * ctypedef npy_intp intp_t # <<<<<<<<<<<<<< * ctypedef npy_uintp uintp_t * */ typedef npy_intp __pyx_t_5numpy_intp_t; /* "numpy.pxd":755 * * ctypedef npy_intp intp_t * ctypedef npy_uintp uintp_t # <<<<<<<<<<<<<< * * ctypedef npy_double float_t */ typedef npy_uintp __pyx_t_5numpy_uintp_t; /* "numpy.pxd":757 * ctypedef npy_uintp uintp_t * * ctypedef npy_double float_t # <<<<<<<<<<<<<< * ctypedef npy_double double_t * ctypedef npy_longdouble longdouble_t */ typedef npy_double __pyx_t_5numpy_float_t; /* "numpy.pxd":758 * * ctypedef npy_double float_t * ctypedef npy_double double_t # <<<<<<<<<<<<<< * ctypedef npy_longdouble longdouble_t * */ typedef npy_double __pyx_t_5numpy_double_t; /* "numpy.pxd":759 * ctypedef npy_double float_t * ctypedef npy_double double_t * ctypedef npy_longdouble longdouble_t # <<<<<<<<<<<<<< * * ctypedef npy_cfloat cfloat_t */ typedef npy_longdouble __pyx_t_5numpy_longdouble_t; #if CYTHON_CCOMPLEX #ifdef __cplusplus typedef ::std::complex< float > __pyx_t_float_complex; #else typedef float _Complex __pyx_t_float_complex; #endif #else typedef struct { float real, imag; } __pyx_t_float_complex; #endif #if CYTHON_CCOMPLEX #ifdef __cplusplus typedef ::std::complex< double > __pyx_t_double_complex; #else typedef double _Complex __pyx_t_double_complex; #endif #else typedef struct { double real, imag; } __pyx_t_double_complex; #endif /*--- Type declarations ---*/ /* "numpy.pxd":761 * ctypedef npy_longdouble longdouble_t * * ctypedef npy_cfloat cfloat_t # <<<<<<<<<<<<<< * ctypedef npy_cdouble cdouble_t * ctypedef npy_clongdouble clongdouble_t */ typedef npy_cfloat __pyx_t_5numpy_cfloat_t; /* "numpy.pxd":762 * * ctypedef npy_cfloat cfloat_t * ctypedef npy_cdouble cdouble_t # <<<<<<<<<<<<<< * ctypedef npy_clongdouble clongdouble_t * */ typedef npy_cdouble __pyx_t_5numpy_cdouble_t; /* "numpy.pxd":763 * ctypedef npy_cfloat cfloat_t * ctypedef npy_cdouble cdouble_t * ctypedef npy_clongdouble clongdouble_t # <<<<<<<<<<<<<< * * ctypedef npy_cdouble complex_t */ typedef npy_clongdouble __pyx_t_5numpy_clongdouble_t; /* "numpy.pxd":765 * ctypedef npy_clongdouble clongdouble_t * * ctypedef npy_cdouble complex_t # <<<<<<<<<<<<<< * * cdef inline object PyArray_MultiIterNew1(a): */ typedef npy_cdouble __pyx_t_5numpy_complex_t; struct __pyx_opt_args_4wfpt_full_pdf; /* "/home/whyking/working/projects/hddm/src/pdf.pxi":98 * return exp(log(p) + ((a*z*V)**2 - 2*a*v*z - (v**2)*x)/(2*(V**2)*x+2))/sqrt((V**2)*x+1)/(a**2) * * cpdef double full_pdf(double x, double v, double V, double a, double # <<<<<<<<<<<<<< * z, double Z, double t, double T, double err, int * nT=2, int nZ=2, bint use_adaptive=1, double */ struct __pyx_opt_args_4wfpt_full_pdf { int __pyx_n; int nT; int nZ; int use_adaptive; double simps_err; }; #ifndef CYTHON_REFNANNY #define CYTHON_REFNANNY 0 #endif #if CYTHON_REFNANNY typedef struct { void (*INCREF)(void*, PyObject*, int); void (*DECREF)(void*, PyObject*, int); void (*GOTREF)(void*, PyObject*, int); void (*GIVEREF)(void*, PyObject*, int); void* (*SetupContext)(const char*, int, const char*); void (*FinishContext)(void**); } __Pyx_RefNannyAPIStruct; static __Pyx_RefNannyAPIStruct *__Pyx_RefNanny = NULL; static __Pyx_RefNannyAPIStruct *__Pyx_RefNannyImportAPI(const char *modname); /*proto*/ #define __Pyx_RefNannyDeclarations void *__pyx_refnanny = NULL; #define __Pyx_RefNannySetupContext(name) __pyx_refnanny = __Pyx_RefNanny->SetupContext((name), __LINE__, __FILE__) #define __Pyx_RefNannyFinishContext() __Pyx_RefNanny->FinishContext(&__pyx_refnanny) #define __Pyx_INCREF(r) __Pyx_RefNanny->INCREF(__pyx_refnanny, (PyObject *)(r), __LINE__) #define __Pyx_DECREF(r) __Pyx_RefNanny->DECREF(__pyx_refnanny, (PyObject *)(r), __LINE__) #define __Pyx_GOTREF(r) __Pyx_RefNanny->GOTREF(__pyx_refnanny, (PyObject *)(r), __LINE__) #define __Pyx_GIVEREF(r) __Pyx_RefNanny->GIVEREF(__pyx_refnanny, (PyObject *)(r), __LINE__) #define __Pyx_XINCREF(r) do { if((r) != NULL) {__Pyx_INCREF(r); }} while(0) #define __Pyx_XDECREF(r) do { if((r) != NULL) {__Pyx_DECREF(r); }} while(0) #define __Pyx_XGOTREF(r) do { if((r) != NULL) {__Pyx_GOTREF(r); }} while(0) #define __Pyx_XGIVEREF(r) do { if((r) != NULL) {__Pyx_GIVEREF(r);}} while(0) #else #define __Pyx_RefNannyDeclarations #define __Pyx_RefNannySetupContext(name) #define __Pyx_RefNannyFinishContext() #define __Pyx_INCREF(r) Py_INCREF(r) #define __Pyx_DECREF(r) Py_DECREF(r) #define __Pyx_GOTREF(r) #define __Pyx_GIVEREF(r) #define __Pyx_XINCREF(r) Py_XINCREF(r) #define __Pyx_XDECREF(r) Py_XDECREF(r) #define __Pyx_XGOTREF(r) #define __Pyx_XGIVEREF(r) #endif /* CYTHON_REFNANNY */ #define __Pyx_CLEAR(r) do { PyObject* tmp = ((PyObject*)(r)); r = NULL; __Pyx_DECREF(tmp);} while(0) #define __Pyx_XCLEAR(r) do { if((r) != NULL) {PyObject* tmp = ((PyObject*)(r)); r = NULL; __Pyx_DECREF(tmp);}} while(0) static PyObject *__Pyx_GetName(PyObject *dict, PyObject *name); /*proto*/ static void __Pyx_RaiseArgtupleInvalid(const char* func_name, int exact, Py_ssize_t num_min, Py_ssize_t num_max, Py_ssize_t num_found); /*proto*/ static void __Pyx_RaiseDoubleKeywordsError(const char* func_name, PyObject* kw_name); /*proto*/ static int __Pyx_ParseOptionalKeywords(PyObject *kwds, PyObject **argnames[], \ PyObject *kwds2, PyObject *values[], Py_ssize_t num_pos_args, \ const char* function_name); /*proto*/ static int __Pyx_ArgTypeTest(PyObject *obj, PyTypeObject *type, int none_allowed, const char *name, int exact); /*proto*/ #define IS_UNSIGNED(type) (((type) -1) > 0) struct __Pyx_StructField_; #define __PYX_BUF_FLAGS_PACKED_STRUCT (1 << 0) typedef struct { const char* name; /* for error messages only */ struct __Pyx_StructField_* fields; size_t size; /* sizeof(type) */ size_t arraysize[8]; /* length of array in each dimension */ int ndim; char typegroup; /* _R_eal, _C_omplex, Signed _I_nt, _U_nsigned int, _S_truct, _P_ointer, _O_bject */ char is_unsigned; int flags; } __Pyx_TypeInfo; typedef struct __Pyx_StructField_ { __Pyx_TypeInfo* type; const char* name; size_t offset; } __Pyx_StructField; typedef struct { __Pyx_StructField* field; size_t parent_offset; } __Pyx_BufFmt_StackElem; typedef struct { __Pyx_StructField root; __Pyx_BufFmt_StackElem* head; size_t fmt_offset; size_t new_count, enc_count; size_t struct_alignment; int is_complex; char enc_type; char new_packmode; char enc_packmode; char is_valid_array; } __Pyx_BufFmt_Context; static CYTHON_INLINE int __Pyx_GetBufferAndValidate(Py_buffer* buf, PyObject* obj, __Pyx_TypeInfo* dtype, int flags, int nd, int cast, __Pyx_BufFmt_StackElem* stack); static CYTHON_INLINE void __Pyx_SafeReleaseBuffer(Py_buffer* info); static CYTHON_INLINE int __Pyx_TypeTest(PyObject *obj, PyTypeObject *type); /*proto*/ #define __Pyx_BufPtrStrided1d(type, buf, i0, s0) (type)((char*)buf + i0 * s0) static CYTHON_INLINE void __Pyx_ErrRestore(PyObject *type, PyObject *value, PyObject *tb); /*proto*/ static CYTHON_INLINE void __Pyx_ErrFetch(PyObject **type, PyObject **value, PyObject **tb); /*proto*/ #include void __pyx_init_nan(void); static float __PYX_NAN; #ifndef __PYX_FORCE_INIT_THREADS #define __PYX_FORCE_INIT_THREADS 0 #endif static CYTHON_INLINE void __Pyx_RaiseNoneIndexingError(void); #if PY_MAJOR_VERSION >= 3 static PyObject *__Pyx_PyDict_GetItem(PyObject *d, PyObject* key) { PyObject *value; if (unlikely(d == Py_None)) { __Pyx_RaiseNoneIndexingError(); return NULL; } value = PyDict_GetItemWithError(d, key); if (unlikely(!value)) { if (!PyErr_Occurred()) PyErr_SetObject(PyExc_KeyError, key); return NULL; } Py_INCREF(value); return value; } #else #define __Pyx_PyDict_GetItem(d, key) PyObject_GetItem(d, key) #endif static CYTHON_INLINE PyObject *__Pyx_GetItemInt_Generic(PyObject *o, PyObject* j) { PyObject *r; if (!j) return NULL; r = PyObject_GetItem(o, j); Py_DECREF(j); return r; } #define __Pyx_GetItemInt_List(o, i, size, to_py_func) (((size) <= sizeof(Py_ssize_t)) ? \ __Pyx_GetItemInt_List_Fast(o, i) : \ __Pyx_GetItemInt_Generic(o, to_py_func(i))) static CYTHON_INLINE PyObject *__Pyx_GetItemInt_List_Fast(PyObject *o, Py_ssize_t i) { if (likely(o != Py_None)) { if (likely((0 <= i) & (i < PyList_GET_SIZE(o)))) { PyObject *r = PyList_GET_ITEM(o, i); Py_INCREF(r); return r; } else if ((-PyList_GET_SIZE(o) <= i) & (i < 0)) { PyObject *r = PyList_GET_ITEM(o, PyList_GET_SIZE(o) + i); Py_INCREF(r); return r; } } return __Pyx_GetItemInt_Generic(o, PyInt_FromSsize_t(i)); } #define __Pyx_GetItemInt_Tuple(o, i, size, to_py_func) (((size) <= sizeof(Py_ssize_t)) ? \ __Pyx_GetItemInt_Tuple_Fast(o, i) : \ __Pyx_GetItemInt_Generic(o, to_py_func(i))) static CYTHON_INLINE PyObject *__Pyx_GetItemInt_Tuple_Fast(PyObject *o, Py_ssize_t i) { if (likely(o != Py_None)) { if (likely((0 <= i) & (i < PyTuple_GET_SIZE(o)))) { PyObject *r = PyTuple_GET_ITEM(o, i); Py_INCREF(r); return r; } else if ((-PyTuple_GET_SIZE(o) <= i) & (i < 0)) { PyObject *r = PyTuple_GET_ITEM(o, PyTuple_GET_SIZE(o) + i); Py_INCREF(r); return r; } } return __Pyx_GetItemInt_Generic(o, PyInt_FromSsize_t(i)); } #define __Pyx_GetItemInt(o, i, size, to_py_func) (((size) <= sizeof(Py_ssize_t)) ? \ __Pyx_GetItemInt_Fast(o, i) : \ __Pyx_GetItemInt_Generic(o, to_py_func(i))) static CYTHON_INLINE PyObject *__Pyx_GetItemInt_Fast(PyObject *o, Py_ssize_t i) { if (PyList_CheckExact(o)) { Py_ssize_t n = (likely(i >= 0)) ? i : i + PyList_GET_SIZE(o); if (likely((n >= 0) & (n < PyList_GET_SIZE(o)))) { PyObject *r = PyList_GET_ITEM(o, n); Py_INCREF(r); return r; } } else if (PyTuple_CheckExact(o)) { Py_ssize_t n = (likely(i >= 0)) ? i : i + PyTuple_GET_SIZE(o); if (likely((n >= 0) & (n < PyTuple_GET_SIZE(o)))) { PyObject *r = PyTuple_GET_ITEM(o, n); Py_INCREF(r); return r; } } else if (likely(i >= 0)) { PySequenceMethods *m = Py_TYPE(o)->tp_as_sequence; if (likely(m && m->sq_item)) { return m->sq_item(o, i); } } return __Pyx_GetItemInt_Generic(o, PyInt_FromSsize_t(i)); } static void __Pyx_RaiseBufferFallbackError(void); /*proto*/ static void __Pyx_Raise(PyObject *type, PyObject *value, PyObject *tb, PyObject *cause); /*proto*/ static CYTHON_INLINE void __Pyx_RaiseNeedMoreValuesError(Py_ssize_t index); static CYTHON_INLINE void __Pyx_RaiseTooManyValuesError(Py_ssize_t expected); static CYTHON_INLINE void __Pyx_RaiseNoneNotIterableError(void); static void __Pyx_UnpackTupleError(PyObject *, Py_ssize_t index); /*proto*/ typedef struct { Py_ssize_t shape, strides, suboffsets; } __Pyx_Buf_DimInfo; typedef struct { size_t refcount; Py_buffer pybuffer; } __Pyx_Buffer; typedef struct { __Pyx_Buffer *rcbuffer; char *data; __Pyx_Buf_DimInfo diminfo[8]; } __Pyx_LocalBuf_ND; #if PY_MAJOR_VERSION < 3 static int __Pyx_GetBuffer(PyObject *obj, Py_buffer *view, int flags); static void __Pyx_ReleaseBuffer(Py_buffer *view); #else #define __Pyx_GetBuffer PyObject_GetBuffer #define __Pyx_ReleaseBuffer PyBuffer_Release #endif Py_ssize_t __Pyx_zeros[] = {0, 0, 0, 0, 0, 0, 0, 0}; Py_ssize_t __Pyx_minusones[] = {-1, -1, -1, -1, -1, -1, -1, -1}; static PyObject *__Pyx_Import(PyObject *name, PyObject *from_list, long level); /*proto*/ static CYTHON_INLINE void __Pyx_RaiseImportError(PyObject *name); static CYTHON_INLINE PyObject *__Pyx_PyInt_to_py_Py_intptr_t(Py_intptr_t); #if CYTHON_CCOMPLEX #ifdef __cplusplus #define __Pyx_CREAL(z) ((z).real()) #define __Pyx_CIMAG(z) ((z).imag()) #else #define __Pyx_CREAL(z) (__real__(z)) #define __Pyx_CIMAG(z) (__imag__(z)) #endif #else #define __Pyx_CREAL(z) ((z).real) #define __Pyx_CIMAG(z) ((z).imag) #endif #if defined(_WIN32) && defined(__cplusplus) && CYTHON_CCOMPLEX #define __Pyx_SET_CREAL(z,x) ((z).real(x)) #define __Pyx_SET_CIMAG(z,y) ((z).imag(y)) #else #define __Pyx_SET_CREAL(z,x) __Pyx_CREAL(z) = (x) #define __Pyx_SET_CIMAG(z,y) __Pyx_CIMAG(z) = (y) #endif static CYTHON_INLINE __pyx_t_float_complex __pyx_t_float_complex_from_parts(float, float); #if CYTHON_CCOMPLEX #define __Pyx_c_eqf(a, b) ((a)==(b)) #define __Pyx_c_sumf(a, b) ((a)+(b)) #define __Pyx_c_difff(a, b) ((a)-(b)) #define __Pyx_c_prodf(a, b) ((a)*(b)) #define __Pyx_c_quotf(a, b) ((a)/(b)) #define __Pyx_c_negf(a) (-(a)) #ifdef __cplusplus #define __Pyx_c_is_zerof(z) ((z)==(float)0) #define __Pyx_c_conjf(z) (::std::conj(z)) #if 1 #define __Pyx_c_absf(z) (::std::abs(z)) #define __Pyx_c_powf(a, b) (::std::pow(a, b)) #endif #else #define __Pyx_c_is_zerof(z) ((z)==0) #define __Pyx_c_conjf(z) (conjf(z)) #if 1 #define __Pyx_c_absf(z) (cabsf(z)) #define __Pyx_c_powf(a, b) (cpowf(a, b)) #endif #endif #else static CYTHON_INLINE int __Pyx_c_eqf(__pyx_t_float_complex, __pyx_t_float_complex); static CYTHON_INLINE __pyx_t_float_complex __Pyx_c_sumf(__pyx_t_float_complex, __pyx_t_float_complex); static CYTHON_INLINE __pyx_t_float_complex __Pyx_c_difff(__pyx_t_float_complex, __pyx_t_float_complex); static CYTHON_INLINE __pyx_t_float_complex __Pyx_c_prodf(__pyx_t_float_complex, __pyx_t_float_complex); static CYTHON_INLINE __pyx_t_float_complex __Pyx_c_quotf(__pyx_t_float_complex, __pyx_t_float_complex); static CYTHON_INLINE __pyx_t_float_complex __Pyx_c_negf(__pyx_t_float_complex); static CYTHON_INLINE int __Pyx_c_is_zerof(__pyx_t_float_complex); static CYTHON_INLINE __pyx_t_float_complex __Pyx_c_conjf(__pyx_t_float_complex); #if 1 static CYTHON_INLINE float __Pyx_c_absf(__pyx_t_float_complex); static CYTHON_INLINE __pyx_t_float_complex __Pyx_c_powf(__pyx_t_float_complex, __pyx_t_float_complex); #endif #endif static CYTHON_INLINE __pyx_t_double_complex __pyx_t_double_complex_from_parts(double, double); #if CYTHON_CCOMPLEX #define __Pyx_c_eq(a, b) ((a)==(b)) #define __Pyx_c_sum(a, b) ((a)+(b)) #define __Pyx_c_diff(a, b) ((a)-(b)) #define __Pyx_c_prod(a, b) ((a)*(b)) #define __Pyx_c_quot(a, b) ((a)/(b)) #define __Pyx_c_neg(a) (-(a)) #ifdef __cplusplus #define __Pyx_c_is_zero(z) ((z)==(double)0) #define __Pyx_c_conj(z) (::std::conj(z)) #if 1 #define __Pyx_c_abs(z) (::std::abs(z)) #define __Pyx_c_pow(a, b) (::std::pow(a, b)) #endif #else #define __Pyx_c_is_zero(z) ((z)==0) #define __Pyx_c_conj(z) (conj(z)) #if 1 #define __Pyx_c_abs(z) (cabs(z)) #define __Pyx_c_pow(a, b) (cpow(a, b)) #endif #endif #else static CYTHON_INLINE int __Pyx_c_eq(__pyx_t_double_complex, __pyx_t_double_complex); static CYTHON_INLINE __pyx_t_double_complex __Pyx_c_sum(__pyx_t_double_complex, __pyx_t_double_complex); static CYTHON_INLINE __pyx_t_double_complex __Pyx_c_diff(__pyx_t_double_complex, __pyx_t_double_complex); static CYTHON_INLINE __pyx_t_double_complex __Pyx_c_prod(__pyx_t_double_complex, __pyx_t_double_complex); static CYTHON_INLINE __pyx_t_double_complex __Pyx_c_quot(__pyx_t_double_complex, __pyx_t_double_complex); static CYTHON_INLINE __pyx_t_double_complex __Pyx_c_neg(__pyx_t_double_complex); static CYTHON_INLINE int __Pyx_c_is_zero(__pyx_t_double_complex); static CYTHON_INLINE __pyx_t_double_complex __Pyx_c_conj(__pyx_t_double_complex); #if 1 static CYTHON_INLINE double __Pyx_c_abs(__pyx_t_double_complex); static CYTHON_INLINE __pyx_t_double_complex __Pyx_c_pow(__pyx_t_double_complex, __pyx_t_double_complex); #endif #endif static CYTHON_INLINE unsigned char __Pyx_PyInt_AsUnsignedChar(PyObject *); static CYTHON_INLINE unsigned short __Pyx_PyInt_AsUnsignedShort(PyObject *); static CYTHON_INLINE unsigned int __Pyx_PyInt_AsUnsignedInt(PyObject *); static CYTHON_INLINE char __Pyx_PyInt_AsChar(PyObject *); static CYTHON_INLINE short __Pyx_PyInt_AsShort(PyObject *); static CYTHON_INLINE int __Pyx_PyInt_AsInt(PyObject *); static CYTHON_INLINE signed char __Pyx_PyInt_AsSignedChar(PyObject *); static CYTHON_INLINE signed short __Pyx_PyInt_AsSignedShort(PyObject *); static CYTHON_INLINE signed int __Pyx_PyInt_AsSignedInt(PyObject *); static CYTHON_INLINE int __Pyx_PyInt_AsLongDouble(PyObject *); static CYTHON_INLINE unsigned long __Pyx_PyInt_AsUnsignedLong(PyObject *); static CYTHON_INLINE unsigned PY_LONG_LONG __Pyx_PyInt_AsUnsignedLongLong(PyObject *); static CYTHON_INLINE long __Pyx_PyInt_AsLong(PyObject *); static CYTHON_INLINE PY_LONG_LONG __Pyx_PyInt_AsLongLong(PyObject *); static CYTHON_INLINE signed long __Pyx_PyInt_AsSignedLong(PyObject *); static CYTHON_INLINE signed PY_LONG_LONG __Pyx_PyInt_AsSignedLongLong(PyObject *); static int __Pyx_check_binary_version(void); static PyTypeObject *__Pyx_ImportType(const char *module_name, const char *class_name, size_t size, int strict); /*proto*/ static PyObject *__Pyx_ImportModule(const char *name); /*proto*/ typedef struct { int code_line; PyCodeObject* code_object; } __Pyx_CodeObjectCacheEntry; struct __Pyx_CodeObjectCache { int count; int max_count; __Pyx_CodeObjectCacheEntry* entries; }; static struct __Pyx_CodeObjectCache __pyx_code_cache = {0,0,NULL}; static int __pyx_bisect_code_objects(__Pyx_CodeObjectCacheEntry* entries, int count, int code_line); static void __pyx_clear_code_object_cache(void); static PyCodeObject *__pyx_find_code_object(int code_line); static void __pyx_insert_code_object(int code_line, PyCodeObject* code_object); static void __Pyx_AddTraceback(const char *funcname, int __pyx_clineno, int __pyx_lineno, const char *__pyx_filename); /*proto*/ static int __Pyx_InitStrings(__Pyx_StringTabEntry *t); /*proto*/ /* Module declarations from 'cpython.buffer' */ /* Module declarations from 'cpython.ref' */ /* Module declarations from 'libc.stdio' */ /* Module declarations from 'cpython.object' */ /* Module declarations from 'libc.stdlib' */ /* Module declarations from 'numpy' */ /* Module declarations from 'numpy' */ static PyTypeObject *__pyx_ptype_5numpy_dtype = 0; static PyTypeObject *__pyx_ptype_5numpy_flatiter = 0; static PyTypeObject *__pyx_ptype_5numpy_broadcast = 0; static PyTypeObject *__pyx_ptype_5numpy_ndarray = 0; static PyTypeObject *__pyx_ptype_5numpy_ufunc = 0; static CYTHON_INLINE char *__pyx_f_5numpy__util_dtypestring(PyArray_Descr *, char *, char *, int *); /*proto*/ /* Module declarations from 'cython' */ /* Module declarations from 'wfpt' */ static double __pyx_f_4wfpt_simpson_1D(double, double, double, double, double, double, double, double, double, int, double, double, int); /*proto*/ static double __pyx_f_4wfpt_simpson_2D(double, double, double, double, double, double, double, double, double, int, double, double, int); /*proto*/ static double __pyx_f_4wfpt_adaptiveSimpsonsAux(double, double, double, double, double, double, double, double, double, double, double, double, double, double, double, double, double, int); /*proto*/ static double __pyx_f_4wfpt_adaptiveSimpsons_1D(double, double, double, double, double, double, double, double, double, double, double, double, int); /*proto*/ static double __pyx_f_4wfpt_adaptiveSimpsonsAux_2D(double, double, double, double, double, double, double, double, double, double, double, double, double, double, double, double, double, double, int, int); /*proto*/ static double __pyx_f_4wfpt_adaptiveSimpsons_2D(double, double, double, double, double, double, double, double, double, double, double, double, int, int); /*proto*/ static double __pyx_f_4wfpt_ftt_01w(double, double, double); /*proto*/ static double __pyx_f_4wfpt_pdf(double, double, double, double, double); /*proto*/ static double __pyx_f_4wfpt_pdf_V(double, double, double, double, double, double); /*proto*/ static double __pyx_f_4wfpt_full_pdf(double, double, double, double, double, double, double, double, double, int __pyx_skip_dispatch, struct __pyx_opt_args_4wfpt_full_pdf *__pyx_optional_args); /*proto*/ static __Pyx_TypeInfo __Pyx_TypeInfo_double = { "double", NULL, sizeof(double), { 0 }, 0, 'R', 0, 0 }; static __Pyx_TypeInfo __Pyx_TypeInfo_int = { "int", NULL, sizeof(int), { 0 }, 0, 'I', IS_UNSIGNED(int), 0 }; #define __Pyx_MODULE_NAME "wfpt" int __pyx_module_is_main_wfpt = 0; /* Implementation of 'wfpt' */ static PyObject *__pyx_builtin_range; static PyObject *__pyx_builtin_ValueError; static PyObject *__pyx_builtin_RuntimeError; static PyObject *__pyx_pf_4wfpt_full_pdf(CYTHON_UNUSED PyObject *__pyx_self, double __pyx_v_x, double __pyx_v_v, double __pyx_v_V, double __pyx_v_a, double __pyx_v_z, double __pyx_v_Z, double __pyx_v_t, double __pyx_v_T, double __pyx_v_err, int __pyx_v_nT, int __pyx_v_nZ, int __pyx_v_use_adaptive, double __pyx_v_simps_err); /* proto */ static PyObject *__pyx_pf_4wfpt_2pdf_array(CYTHON_UNUSED PyObject *__pyx_self, PyArrayObject *__pyx_v_x, double __pyx_v_v, double __pyx_v_V, double __pyx_v_a, double __pyx_v_z, double __pyx_v_Z, double __pyx_v_t, double __pyx_v_T, double __pyx_v_err, int __pyx_v_logp, int __pyx_v_nT, int __pyx_v_nZ, int __pyx_v_use_adaptive, double __pyx_v_simps_err); /* proto */ static PyObject *__pyx_pf_4wfpt_4wiener_like(CYTHON_UNUSED PyObject *__pyx_self, PyArrayObject *__pyx_v_x, double __pyx_v_v, double __pyx_v_V, double __pyx_v_a, double __pyx_v_z, double __pyx_v_Z, double __pyx_v_t, double __pyx_v_T, double __pyx_v_err, int __pyx_v_nT, int __pyx_v_nZ, int __pyx_v_use_adaptive, double __pyx_v_simps_err); /* proto */ static PyObject *__pyx_pf_4wfpt_6wiener_like_array(CYTHON_UNUSED PyObject *__pyx_self, PyArrayObject *__pyx_v_x, double __pyx_v_v, double __pyx_v_V, double __pyx_v_a, double __pyx_v_z, double __pyx_v_Z, double __pyx_v_t, double __pyx_v_T, double __pyx_v_err, int __pyx_v_nT, int __pyx_v_nZ, int __pyx_v_use_adaptive, double __pyx_v_simps_err); /* proto */ static PyObject *__pyx_pf_4wfpt_8wiener_like_multi(CYTHON_UNUSED PyObject *__pyx_self, PyArrayObject *__pyx_v_x, PyObject *__pyx_v_v, PyObject *__pyx_v_V, PyObject *__pyx_v_a, PyObject *__pyx_v_z, PyObject *__pyx_v_Z, PyObject *__pyx_v_t, PyObject *__pyx_v_T, double __pyx_v_err, PyObject *__pyx_v_multi); /* proto */ static PyObject *__pyx_pf_4wfpt_10gen_rts_from_cdf(CYTHON_UNUSED PyObject *__pyx_self, double __pyx_v_v, double __pyx_v_V, double __pyx_v_a, double __pyx_v_z, double __pyx_v_Z, double __pyx_v_t, double __pyx_v_T, int __pyx_v_samples, double __pyx_v_cdf_lb, double __pyx_v_cdf_ub, double __pyx_v_dt); /* proto */ static PyObject *__pyx_pf_4wfpt_12wiener_like_contaminant(CYTHON_UNUSED PyObject *__pyx_self, PyArrayObject *__pyx_v_x, PyArrayObject *__pyx_v_cont_x, double __pyx_v_v, double __pyx_v_V, double __pyx_v_a, double __pyx_v_z, double __pyx_v_Z, double __pyx_v_t, double __pyx_v_T, double __pyx_v_t_min, double __pyx_v_t_max, double __pyx_v_err, int __pyx_v_nT, int __pyx_v_nZ, int __pyx_v_use_adaptive, double __pyx_v_simps_err); /* proto */ static int __pyx_pf_5numpy_7ndarray___getbuffer__(PyArrayObject *__pyx_v_self, Py_buffer *__pyx_v_info, int __pyx_v_flags); /* proto */ static void __pyx_pf_5numpy_7ndarray_2__releasebuffer__(PyArrayObject *__pyx_v_self, Py_buffer *__pyx_v_info); /* proto */ static char __pyx_k_1[] = "ndarray is not C contiguous"; static char __pyx_k_3[] = "ndarray is not Fortran contiguous"; static char __pyx_k_5[] = "Non-native byte order not supported"; static char __pyx_k_7[] = "unknown dtype code in numpy.pxd (%d)"; static char __pyx_k_8[] = "Format string allocated too short, see comment in numpy.pxd"; static char __pyx_k_11[] = "Format string allocated too short."; static char __pyx_k_15[] = "/home/whyking/working/projects/hddm/src/wfpt.pyx"; static char __pyx_k_26[] = "wiener_like_contaminant"; static char __pyx_k__B[] = "B"; static char __pyx_k__H[] = "H"; static char __pyx_k__I[] = "I"; static char __pyx_k__L[] = "L"; static char __pyx_k__O[] = "O"; static char __pyx_k__Q[] = "Q"; static char __pyx_k__T[] = "T"; static char __pyx_k__V[] = "V"; static char __pyx_k__Z[] = "Z"; static char __pyx_k__a[] = "a"; static char __pyx_k__b[] = "b"; static char __pyx_k__d[] = "d"; static char __pyx_k__f[] = "f"; static char __pyx_k__g[] = "g"; static char __pyx_k__h[] = "h"; static char __pyx_k__i[] = "i"; static char __pyx_k__j[] = "j"; static char __pyx_k__l[] = "l"; static char __pyx_k__p[] = "p"; static char __pyx_k__q[] = "q"; static char __pyx_k__t[] = "t"; static char __pyx_k__v[] = "v"; static char __pyx_k__x[] = "x"; static char __pyx_k__y[] = "y"; static char __pyx_k__z[] = "z"; static char __pyx_k__Zd[] = "Zd"; static char __pyx_k__Zf[] = "Zf"; static char __pyx_k__Zg[] = "Zg"; static char __pyx_k__dt[] = "dt"; static char __pyx_k__nT[] = "nT"; static char __pyx_k__nZ[] = "nZ"; static char __pyx_k__np[] = "np"; static char __pyx_k__rt[] = "rt"; static char __pyx_k__err[] = "err"; static char __pyx_k__idx[] = "idx"; static char __pyx_k__inf[] = "inf"; static char __pyx_k__log[] = "log"; static char __pyx_k__pdf[] = "pdf"; static char __pyx_k__rts[] = "rts"; static char __pyx_k__sum[] = "sum"; static char __pyx_k__copy[] = "copy"; static char __pyx_k__hddm[] = "hddm"; static char __pyx_k__logp[] = "logp"; static char __pyx_k__rand[] = "rand"; static char __pyx_k__sign[] = "sign"; static char __pyx_k__size[] = "size"; static char __pyx_k__wfpt[] = "wfpt"; static char __pyx_k__delay[] = "delay"; static char __pyx_k__dtype[] = "dtype"; static char __pyx_k__empty[] = "empty"; static char __pyx_k__float[] = "float"; static char __pyx_k__l_cdf[] = "l_cdf"; static char __pyx_k__multi[] = "multi"; static char __pyx_k__numpy[] = "numpy"; static char __pyx_k__param[] = "param"; static char __pyx_k__range[] = "range"; static char __pyx_k__t_max[] = "t_max"; static char __pyx_k__t_min[] = "t_min"; static char __pyx_k__arange[] = "arange"; static char __pyx_k__cdf_lb[] = "cdf_lb"; static char __pyx_k__cdf_ub[] = "cdf_ub"; static char __pyx_k__cont_x[] = "cont_x"; static char __pyx_k__double[] = "double"; static char __pyx_k__n_cont[] = "n_cont"; static char __pyx_k__params[] = "params"; static char __pyx_k__random[] = "random"; static char __pyx_k__p_array[] = "p_array"; static char __pyx_k__samples[] = "samples"; static char __pyx_k____main__[] = "__main__"; static char __pyx_k____test__[] = "__test__"; static char __pyx_k__pos_cont[] = "pos_cont"; static char __pyx_k__sum_logp[] = "sum_logp"; static char __pyx_k__pdf_array[] = "pdf_array"; static char __pyx_k__simps_err[] = "simps_err"; static char __pyx_k__ValueError[] = "ValueError"; static char __pyx_k__params_iter[] = "params_iter"; static char __pyx_k__wiener_like[] = "wiener_like"; static char __pyx_k__RuntimeError[] = "RuntimeError"; static char __pyx_k__searchsorted[] = "searchsorted"; static char __pyx_k__use_adaptive[] = "use_adaptive"; static char __pyx_k__gen_rts_from_cdf[] = "gen_rts_from_cdf"; static char __pyx_k__wiener_like_array[] = "wiener_like_array"; static char __pyx_k__wiener_like_multi[] = "wiener_like_multi"; static PyObject *__pyx_kp_u_1; static PyObject *__pyx_kp_u_11; static PyObject *__pyx_kp_s_15; static PyObject *__pyx_n_s_26; static PyObject *__pyx_kp_u_3; static PyObject *__pyx_kp_u_5; static PyObject *__pyx_kp_u_7; static PyObject *__pyx_kp_u_8; static PyObject *__pyx_n_s__RuntimeError; static PyObject *__pyx_n_s__T; static PyObject *__pyx_n_s__V; static PyObject *__pyx_n_s__ValueError; static PyObject *__pyx_n_s__Z; static PyObject *__pyx_n_s____main__; static PyObject *__pyx_n_s____test__; static PyObject *__pyx_n_s__a; static PyObject *__pyx_n_s__arange; static PyObject *__pyx_n_s__cdf_lb; static PyObject *__pyx_n_s__cdf_ub; static PyObject *__pyx_n_s__cont_x; static PyObject *__pyx_n_s__copy; static PyObject *__pyx_n_s__delay; static PyObject *__pyx_n_s__double; static PyObject *__pyx_n_s__dt; static PyObject *__pyx_n_s__dtype; static PyObject *__pyx_n_s__empty; static PyObject *__pyx_n_s__err; static PyObject *__pyx_n_s__f; static PyObject *__pyx_n_s__float; static PyObject *__pyx_n_s__gen_rts_from_cdf; static PyObject *__pyx_n_s__hddm; static PyObject *__pyx_n_s__i; static PyObject *__pyx_n_s__idx; static PyObject *__pyx_n_s__inf; static PyObject *__pyx_n_s__j; static PyObject *__pyx_n_s__l_cdf; static PyObject *__pyx_n_s__log; static PyObject *__pyx_n_s__logp; static PyObject *__pyx_n_s__multi; static PyObject *__pyx_n_s__nT; static PyObject *__pyx_n_s__nZ; static PyObject *__pyx_n_s__n_cont; static PyObject *__pyx_n_s__np; static PyObject *__pyx_n_s__numpy; static PyObject *__pyx_n_s__p; static PyObject *__pyx_n_s__p_array; static PyObject *__pyx_n_s__param; static PyObject *__pyx_n_s__params; static PyObject *__pyx_n_s__params_iter; static PyObject *__pyx_n_s__pdf; static PyObject *__pyx_n_s__pdf_array; static PyObject *__pyx_n_s__pos_cont; static PyObject *__pyx_n_s__rand; static PyObject *__pyx_n_s__random; static PyObject *__pyx_n_s__range; static PyObject *__pyx_n_s__rt; static PyObject *__pyx_n_s__rts; static PyObject *__pyx_n_s__samples; static PyObject *__pyx_n_s__searchsorted; static PyObject *__pyx_n_s__sign; static PyObject *__pyx_n_s__simps_err; static PyObject *__pyx_n_s__size; static PyObject *__pyx_n_s__sum; static PyObject *__pyx_n_s__sum_logp; static PyObject *__pyx_n_s__t; static PyObject *__pyx_n_s__t_max; static PyObject *__pyx_n_s__t_min; static PyObject *__pyx_n_s__use_adaptive; static PyObject *__pyx_n_s__v; static PyObject *__pyx_n_s__wfpt; static PyObject *__pyx_n_s__wiener_like; static PyObject *__pyx_n_s__wiener_like_array; static PyObject *__pyx_n_s__wiener_like_multi; static PyObject *__pyx_n_s__x; static PyObject *__pyx_n_s__y; static PyObject *__pyx_n_s__z; static PyObject *__pyx_int_15; static PyObject *__pyx_k_tuple_2; static PyObject *__pyx_k_tuple_4; static PyObject *__pyx_k_tuple_6; static PyObject *__pyx_k_tuple_9; static PyObject *__pyx_k_tuple_10; static PyObject *__pyx_k_tuple_12; static PyObject *__pyx_k_tuple_13; static PyObject *__pyx_k_tuple_16; static PyObject *__pyx_k_tuple_18; static PyObject *__pyx_k_tuple_20; static PyObject *__pyx_k_tuple_22; static PyObject *__pyx_k_tuple_24; static PyObject *__pyx_k_codeobj_14; static PyObject *__pyx_k_codeobj_17; static PyObject *__pyx_k_codeobj_19; static PyObject *__pyx_k_codeobj_21; static PyObject *__pyx_k_codeobj_23; static PyObject *__pyx_k_codeobj_25; /* "/home/whyking/working/projects/hddm/src/integrate.pxi":10 * cimport cython * * cdef double simpson_1D(double x, double v, double V, double a, double z, double t, double err, # <<<<<<<<<<<<<< * double lb_z, double ub_z, int nZ, double lb_t, double ub_t, int nT) nogil: * #assert ((nZ&1)==0 and (nT&1)==0), "nT and nZ have to be even" */ static double __pyx_f_4wfpt_simpson_1D(double __pyx_v_x, double __pyx_v_v, double __pyx_v_V, double __pyx_v_a, double __pyx_v_z, double __pyx_v_t, double __pyx_v_err, double __pyx_v_lb_z, double __pyx_v_ub_z, int __pyx_v_nZ, double __pyx_v_lb_t, double __pyx_v_ub_t, int __pyx_v_nT) { double __pyx_v_ht; double __pyx_v_hz; int __pyx_v_n; double __pyx_v_S; double __pyx_v_z_tag; double __pyx_v_t_tag; double __pyx_v_y; int __pyx_v_i; double __pyx_r; int __pyx_t_1; int __pyx_t_2; int __pyx_t_3; int __pyx_t_4; long __pyx_t_5; /* "/home/whyking/working/projects/hddm/src/integrate.pxi":16 * * cdef double ht, hz * cdef int n = max(nT,nZ) # <<<<<<<<<<<<<< * if nT==0: #integration over z * hz = (ub_z-lb_z)/n */ __pyx_t_1 = __pyx_v_nZ; __pyx_t_2 = __pyx_v_nT; if ((__pyx_t_1 > __pyx_t_2)) { __pyx_t_3 = __pyx_t_1; } else { __pyx_t_3 = __pyx_t_2; } __pyx_v_n = __pyx_t_3; /* "/home/whyking/working/projects/hddm/src/integrate.pxi":17 * cdef double ht, hz * cdef int n = max(nT,nZ) * if nT==0: #integration over z # <<<<<<<<<<<<<< * hz = (ub_z-lb_z)/n * ht = 0 */ __pyx_t_4 = (__pyx_v_nT == 0); if (__pyx_t_4) { /* "/home/whyking/working/projects/hddm/src/integrate.pxi":18 * cdef int n = max(nT,nZ) * if nT==0: #integration over z * hz = (ub_z-lb_z)/n # <<<<<<<<<<<<<< * ht = 0 * lb_t = t */ __pyx_v_hz = ((__pyx_v_ub_z - __pyx_v_lb_z) / __pyx_v_n); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":19 * if nT==0: #integration over z * hz = (ub_z-lb_z)/n * ht = 0 # <<<<<<<<<<<<<< * lb_t = t * ub_t = t */ __pyx_v_ht = 0.0; /* "/home/whyking/working/projects/hddm/src/integrate.pxi":20 * hz = (ub_z-lb_z)/n * ht = 0 * lb_t = t # <<<<<<<<<<<<<< * ub_t = t * else: #integration over t */ __pyx_v_lb_t = __pyx_v_t; /* "/home/whyking/working/projects/hddm/src/integrate.pxi":21 * ht = 0 * lb_t = t * ub_t = t # <<<<<<<<<<<<<< * else: #integration over t * hz = 0 */ __pyx_v_ub_t = __pyx_v_t; goto __pyx_L3; } /*else*/ { /* "/home/whyking/working/projects/hddm/src/integrate.pxi":23 * ub_t = t * else: #integration over t * hz = 0 # <<<<<<<<<<<<<< * ht = (ub_t-lb_t)/n * lb_z = z */ __pyx_v_hz = 0.0; /* "/home/whyking/working/projects/hddm/src/integrate.pxi":24 * else: #integration over t * hz = 0 * ht = (ub_t-lb_t)/n # <<<<<<<<<<<<<< * lb_z = z * ub_z = z */ __pyx_v_ht = ((__pyx_v_ub_t - __pyx_v_lb_t) / __pyx_v_n); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":25 * hz = 0 * ht = (ub_t-lb_t)/n * lb_z = z # <<<<<<<<<<<<<< * ub_z = z * */ __pyx_v_lb_z = __pyx_v_z; /* "/home/whyking/working/projects/hddm/src/integrate.pxi":26 * ht = (ub_t-lb_t)/n * lb_z = z * ub_z = z # <<<<<<<<<<<<<< * * cdef double S = pdf_V(x - lb_t, v, V, a, lb_z, err) */ __pyx_v_ub_z = __pyx_v_z; } __pyx_L3:; /* "/home/whyking/working/projects/hddm/src/integrate.pxi":28 * ub_z = z * * cdef double S = pdf_V(x - lb_t, v, V, a, lb_z, err) # <<<<<<<<<<<<<< * cdef double z_tag, t_tag, y * cdef int i */ __pyx_v_S = __pyx_f_4wfpt_pdf_V((__pyx_v_x - __pyx_v_lb_t), __pyx_v_v, __pyx_v_V, __pyx_v_a, __pyx_v_lb_z, __pyx_v_err); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":32 * cdef int i * * for i from 1 <= i <= n: # <<<<<<<<<<<<<< * z_tag = lb_z + hz * i * t_tag = lb_t + ht * i */ __pyx_t_3 = __pyx_v_n; for (__pyx_v_i = 1; __pyx_v_i <= __pyx_t_3; __pyx_v_i++) { /* "/home/whyking/working/projects/hddm/src/integrate.pxi":33 * * for i from 1 <= i <= n: * z_tag = lb_z + hz * i # <<<<<<<<<<<<<< * t_tag = lb_t + ht * i * y = pdf_V(x - t_tag, v, V, a, z_tag, err) */ __pyx_v_z_tag = (__pyx_v_lb_z + (__pyx_v_hz * __pyx_v_i)); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":34 * for i from 1 <= i <= n: * z_tag = lb_z + hz * i * t_tag = lb_t + ht * i # <<<<<<<<<<<<<< * y = pdf_V(x - t_tag, v, V, a, z_tag, err) * if i&1: #check if i is odd */ __pyx_v_t_tag = (__pyx_v_lb_t + (__pyx_v_ht * __pyx_v_i)); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":35 * z_tag = lb_z + hz * i * t_tag = lb_t + ht * i * y = pdf_V(x - t_tag, v, V, a, z_tag, err) # <<<<<<<<<<<<<< * if i&1: #check if i is odd * S += (4 * y) */ __pyx_v_y = __pyx_f_4wfpt_pdf_V((__pyx_v_x - __pyx_v_t_tag), __pyx_v_v, __pyx_v_V, __pyx_v_a, __pyx_v_z_tag, __pyx_v_err); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":36 * t_tag = lb_t + ht * i * y = pdf_V(x - t_tag, v, V, a, z_tag, err) * if i&1: #check if i is odd # <<<<<<<<<<<<<< * S += (4 * y) * else: */ __pyx_t_5 = (__pyx_v_i & 1); if (__pyx_t_5) { /* "/home/whyking/working/projects/hddm/src/integrate.pxi":37 * y = pdf_V(x - t_tag, v, V, a, z_tag, err) * if i&1: #check if i is odd * S += (4 * y) # <<<<<<<<<<<<<< * else: * S += (2 * y) */ __pyx_v_S = (__pyx_v_S + (4.0 * __pyx_v_y)); goto __pyx_L6; } /*else*/ { /* "/home/whyking/working/projects/hddm/src/integrate.pxi":39 * S += (4 * y) * else: * S += (2 * y) # <<<<<<<<<<<<<< * S = S - y #the last term should be f(b) and not 2*f(b) so we subtract y * S = S / ((ub_t-lb_t)+(ub_z-lb_z)) #the right function if pdf_V()/Z or pdf_V()/T */ __pyx_v_S = (__pyx_v_S + (2.0 * __pyx_v_y)); } __pyx_L6:; } /* "/home/whyking/working/projects/hddm/src/integrate.pxi":40 * else: * S += (2 * y) * S = S - y #the last term should be f(b) and not 2*f(b) so we subtract y # <<<<<<<<<<<<<< * S = S / ((ub_t-lb_t)+(ub_z-lb_z)) #the right function if pdf_V()/Z or pdf_V()/T * */ __pyx_v_S = (__pyx_v_S - __pyx_v_y); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":41 * S += (2 * y) * S = S - y #the last term should be f(b) and not 2*f(b) so we subtract y * S = S / ((ub_t-lb_t)+(ub_z-lb_z)) #the right function if pdf_V()/Z or pdf_V()/T # <<<<<<<<<<<<<< * * return ((ht+hz) * S / 3) */ __pyx_v_S = (__pyx_v_S / ((__pyx_v_ub_t - __pyx_v_lb_t) + (__pyx_v_ub_z - __pyx_v_lb_z))); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":43 * S = S / ((ub_t-lb_t)+(ub_z-lb_z)) #the right function if pdf_V()/Z or pdf_V()/T * * return ((ht+hz) * S / 3) # <<<<<<<<<<<<<< * * cdef double simpson_2D(double x, double v, double V, double a, double z, double t, double err, double lb_z, double ub_z, int nZ, double lb_t, double ub_t, int nT) nogil: */ __pyx_r = (((__pyx_v_ht + __pyx_v_hz) * __pyx_v_S) / 3.0); goto __pyx_L0; __pyx_r = 0; __pyx_L0:; return __pyx_r; } /* "/home/whyking/working/projects/hddm/src/integrate.pxi":45 * return ((ht+hz) * S / 3) * * cdef double simpson_2D(double x, double v, double V, double a, double z, double t, double err, double lb_z, double ub_z, int nZ, double lb_t, double ub_t, int nT) nogil: # <<<<<<<<<<<<<< * #assert ((nZ&1)==0 and (nT&1)==0), "nT and nZ have to be even" * #assert ((ub_t-lb_t)*(ub_z-lb_z)>0 and (nZ*nT)>0), "the function is defined for 2D-integration only, lb_t: %f, ub_t %f, lb_z %f, ub_z %f, nZ: %d, nT %d" % (lb_t, ub_t, lb_z, ub_z, nZ, nT) */ static double __pyx_f_4wfpt_simpson_2D(double __pyx_v_x, double __pyx_v_v, double __pyx_v_V, double __pyx_v_a, double __pyx_v_z, CYTHON_UNUSED double __pyx_v_t, double __pyx_v_err, double __pyx_v_lb_z, double __pyx_v_ub_z, int __pyx_v_nZ, double __pyx_v_lb_t, double __pyx_v_ub_t, int __pyx_v_nT) { double __pyx_v_ht; double __pyx_v_S; double __pyx_v_t_tag; double __pyx_v_y; int __pyx_v_i_t; double __pyx_r; int __pyx_t_1; long __pyx_t_2; /* "/home/whyking/working/projects/hddm/src/integrate.pxi":54 * cdef int i_t * * ht = (ub_t-lb_t)/nT # <<<<<<<<<<<<<< * * S = simpson_1D(x, v, V, a, z, lb_t, err, lb_z, ub_z, nZ, 0, 0, 0) */ __pyx_v_ht = ((__pyx_v_ub_t - __pyx_v_lb_t) / __pyx_v_nT); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":56 * ht = (ub_t-lb_t)/nT * * S = simpson_1D(x, v, V, a, z, lb_t, err, lb_z, ub_z, nZ, 0, 0, 0) # <<<<<<<<<<<<<< * * for i_t from 1 <= i_t <= nT: */ __pyx_v_S = __pyx_f_4wfpt_simpson_1D(__pyx_v_x, __pyx_v_v, __pyx_v_V, __pyx_v_a, __pyx_v_z, __pyx_v_lb_t, __pyx_v_err, __pyx_v_lb_z, __pyx_v_ub_z, __pyx_v_nZ, 0.0, 0.0, 0); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":58 * S = simpson_1D(x, v, V, a, z, lb_t, err, lb_z, ub_z, nZ, 0, 0, 0) * * for i_t from 1 <= i_t <= nT: # <<<<<<<<<<<<<< * t_tag = lb_t + ht * i_t * y = simpson_1D(x, v, V, a, z, t_tag, err, lb_z, ub_z, nZ, 0, 0, 0) */ __pyx_t_1 = __pyx_v_nT; for (__pyx_v_i_t = 1; __pyx_v_i_t <= __pyx_t_1; __pyx_v_i_t++) { /* "/home/whyking/working/projects/hddm/src/integrate.pxi":59 * * for i_t from 1 <= i_t <= nT: * t_tag = lb_t + ht * i_t # <<<<<<<<<<<<<< * y = simpson_1D(x, v, V, a, z, t_tag, err, lb_z, ub_z, nZ, 0, 0, 0) * if i_t&1: #check if i is odd */ __pyx_v_t_tag = (__pyx_v_lb_t + (__pyx_v_ht * __pyx_v_i_t)); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":60 * for i_t from 1 <= i_t <= nT: * t_tag = lb_t + ht * i_t * y = simpson_1D(x, v, V, a, z, t_tag, err, lb_z, ub_z, nZ, 0, 0, 0) # <<<<<<<<<<<<<< * if i_t&1: #check if i is odd * S += (4 * y) */ __pyx_v_y = __pyx_f_4wfpt_simpson_1D(__pyx_v_x, __pyx_v_v, __pyx_v_V, __pyx_v_a, __pyx_v_z, __pyx_v_t_tag, __pyx_v_err, __pyx_v_lb_z, __pyx_v_ub_z, __pyx_v_nZ, 0.0, 0.0, 0); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":61 * t_tag = lb_t + ht * i_t * y = simpson_1D(x, v, V, a, z, t_tag, err, lb_z, ub_z, nZ, 0, 0, 0) * if i_t&1: #check if i is odd # <<<<<<<<<<<<<< * S += (4 * y) * else: */ __pyx_t_2 = (__pyx_v_i_t & 1); if (__pyx_t_2) { /* "/home/whyking/working/projects/hddm/src/integrate.pxi":62 * y = simpson_1D(x, v, V, a, z, t_tag, err, lb_z, ub_z, nZ, 0, 0, 0) * if i_t&1: #check if i is odd * S += (4 * y) # <<<<<<<<<<<<<< * else: * S += (2 * y) */ __pyx_v_S = (__pyx_v_S + (4.0 * __pyx_v_y)); goto __pyx_L5; } /*else*/ { /* "/home/whyking/working/projects/hddm/src/integrate.pxi":64 * S += (4 * y) * else: * S += (2 * y) # <<<<<<<<<<<<<< * S = S - y #the last term should be f(b) and not 2*f(b) so we subtract y * S = S/ (ub_t-lb_t) */ __pyx_v_S = (__pyx_v_S + (2.0 * __pyx_v_y)); } __pyx_L5:; } /* "/home/whyking/working/projects/hddm/src/integrate.pxi":65 * else: * S += (2 * y) * S = S - y #the last term should be f(b) and not 2*f(b) so we subtract y # <<<<<<<<<<<<<< * S = S/ (ub_t-lb_t) * */ __pyx_v_S = (__pyx_v_S - __pyx_v_y); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":66 * S += (2 * y) * S = S - y #the last term should be f(b) and not 2*f(b) so we subtract y * S = S/ (ub_t-lb_t) # <<<<<<<<<<<<<< * * return (ht * S / 3) */ __pyx_v_S = (__pyx_v_S / (__pyx_v_ub_t - __pyx_v_lb_t)); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":68 * S = S/ (ub_t-lb_t) * * return (ht * S / 3) # <<<<<<<<<<<<<< * * cdef double adaptiveSimpsonsAux(double x, double v, double V, double a, double z, double t, double pdf_err, */ __pyx_r = ((__pyx_v_ht * __pyx_v_S) / 3.0); goto __pyx_L0; __pyx_r = 0; __pyx_L0:; return __pyx_r; } /* "/home/whyking/working/projects/hddm/src/integrate.pxi":70 * return (ht * S / 3) * * cdef double adaptiveSimpsonsAux(double x, double v, double V, double a, double z, double t, double pdf_err, # <<<<<<<<<<<<<< * double lb_z, double ub_z, double lb_t, double ub_t, double ZT, double simps_err, * double S, double f_beg, double f_end, double f_mid, int bottom) nogil: */ static double __pyx_f_4wfpt_adaptiveSimpsonsAux(double __pyx_v_x, double __pyx_v_v, double __pyx_v_V, double __pyx_v_a, double __pyx_v_z, double __pyx_v_t, double __pyx_v_pdf_err, double __pyx_v_lb_z, double __pyx_v_ub_z, double __pyx_v_lb_t, double __pyx_v_ub_t, double __pyx_v_ZT, double __pyx_v_simps_err, double __pyx_v_S, double __pyx_v_f_beg, double __pyx_v_f_end, double __pyx_v_f_mid, int __pyx_v_bottom) { double __pyx_v_z_c; double __pyx_v_z_d; double __pyx_v_z_e; double __pyx_v_t_c; double __pyx_v_t_d; double __pyx_v_t_e; double __pyx_v_h; double __pyx_v_fd; double __pyx_v_fe; double __pyx_v_Sleft; double __pyx_v_Sright; double __pyx_v_S2; double __pyx_r; int __pyx_t_1; int __pyx_t_2; int __pyx_t_3; /* "/home/whyking/working/projects/hddm/src/integrate.pxi":79 * #print "in AdaptiveSimpsAux: lb_z: %f, ub_z: %f, lb_t %f, ub_t %f, f_beg: %f, f_end: %f, bottom: %d" % (lb_z, ub_z, lb_t, ub_t, f_beg, f_end, bottom) * * if (ub_t-lb_t) == 0: #integration over Z # <<<<<<<<<<<<<< * h = ub_z - lb_z * z_c = (ub_z + lb_z)/2. */ __pyx_t_1 = ((__pyx_v_ub_t - __pyx_v_lb_t) == 0.0); if (__pyx_t_1) { /* "/home/whyking/working/projects/hddm/src/integrate.pxi":80 * * if (ub_t-lb_t) == 0: #integration over Z * h = ub_z - lb_z # <<<<<<<<<<<<<< * z_c = (ub_z + lb_z)/2. * z_d = (lb_z + z_c)/2. */ __pyx_v_h = (__pyx_v_ub_z - __pyx_v_lb_z); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":81 * if (ub_t-lb_t) == 0: #integration over Z * h = ub_z - lb_z * z_c = (ub_z + lb_z)/2. # <<<<<<<<<<<<<< * z_d = (lb_z + z_c)/2. * z_e = (z_c + ub_z)/2. */ __pyx_v_z_c = ((__pyx_v_ub_z + __pyx_v_lb_z) / 2.); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":82 * h = ub_z - lb_z * z_c = (ub_z + lb_z)/2. * z_d = (lb_z + z_c)/2. # <<<<<<<<<<<<<< * z_e = (z_c + ub_z)/2. * t_c = t */ __pyx_v_z_d = ((__pyx_v_lb_z + __pyx_v_z_c) / 2.); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":83 * z_c = (ub_z + lb_z)/2. * z_d = (lb_z + z_c)/2. * z_e = (z_c + ub_z)/2. # <<<<<<<<<<<<<< * t_c = t * t_d = t */ __pyx_v_z_e = ((__pyx_v_z_c + __pyx_v_ub_z) / 2.); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":84 * z_d = (lb_z + z_c)/2. * z_e = (z_c + ub_z)/2. * t_c = t # <<<<<<<<<<<<<< * t_d = t * t_e = t */ __pyx_v_t_c = __pyx_v_t; /* "/home/whyking/working/projects/hddm/src/integrate.pxi":85 * z_e = (z_c + ub_z)/2. * t_c = t * t_d = t # <<<<<<<<<<<<<< * t_e = t * */ __pyx_v_t_d = __pyx_v_t; /* "/home/whyking/working/projects/hddm/src/integrate.pxi":86 * t_c = t * t_d = t * t_e = t # <<<<<<<<<<<<<< * * else: #integration over t */ __pyx_v_t_e = __pyx_v_t; goto __pyx_L3; } /*else*/ { /* "/home/whyking/working/projects/hddm/src/integrate.pxi":89 * * else: #integration over t * h = ub_t - lb_t # <<<<<<<<<<<<<< * t_c = (ub_t + lb_t)/2. * t_d = (lb_t + t_c)/2. */ __pyx_v_h = (__pyx_v_ub_t - __pyx_v_lb_t); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":90 * else: #integration over t * h = ub_t - lb_t * t_c = (ub_t + lb_t)/2. # <<<<<<<<<<<<<< * t_d = (lb_t + t_c)/2. * t_e = (t_c + ub_t)/2. */ __pyx_v_t_c = ((__pyx_v_ub_t + __pyx_v_lb_t) / 2.); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":91 * h = ub_t - lb_t * t_c = (ub_t + lb_t)/2. * t_d = (lb_t + t_c)/2. # <<<<<<<<<<<<<< * t_e = (t_c + ub_t)/2. * z_c = z */ __pyx_v_t_d = ((__pyx_v_lb_t + __pyx_v_t_c) / 2.); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":92 * t_c = (ub_t + lb_t)/2. * t_d = (lb_t + t_c)/2. * t_e = (t_c + ub_t)/2. # <<<<<<<<<<<<<< * z_c = z * z_d = z */ __pyx_v_t_e = ((__pyx_v_t_c + __pyx_v_ub_t) / 2.); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":93 * t_d = (lb_t + t_c)/2. * t_e = (t_c + ub_t)/2. * z_c = z # <<<<<<<<<<<<<< * z_d = z * z_e = z */ __pyx_v_z_c = __pyx_v_z; /* "/home/whyking/working/projects/hddm/src/integrate.pxi":94 * t_e = (t_c + ub_t)/2. * z_c = z * z_d = z # <<<<<<<<<<<<<< * z_e = z * */ __pyx_v_z_d = __pyx_v_z; /* "/home/whyking/working/projects/hddm/src/integrate.pxi":95 * z_c = z * z_d = z * z_e = z # <<<<<<<<<<<<<< * * fd = pdf_V(x - t_d, v, V, a, z_d, pdf_err)/ZT */ __pyx_v_z_e = __pyx_v_z; } __pyx_L3:; /* "/home/whyking/working/projects/hddm/src/integrate.pxi":97 * z_e = z * * fd = pdf_V(x - t_d, v, V, a, z_d, pdf_err)/ZT # <<<<<<<<<<<<<< * fe = pdf_V(x - t_e, v, V, a, z_e, pdf_err)/ZT * */ __pyx_v_fd = (__pyx_f_4wfpt_pdf_V((__pyx_v_x - __pyx_v_t_d), __pyx_v_v, __pyx_v_V, __pyx_v_a, __pyx_v_z_d, __pyx_v_pdf_err) / __pyx_v_ZT); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":98 * * fd = pdf_V(x - t_d, v, V, a, z_d, pdf_err)/ZT * fe = pdf_V(x - t_e, v, V, a, z_e, pdf_err)/ZT # <<<<<<<<<<<<<< * * Sleft = (h/12)*(f_beg + 4*fd + f_mid) */ __pyx_v_fe = (__pyx_f_4wfpt_pdf_V((__pyx_v_x - __pyx_v_t_e), __pyx_v_v, __pyx_v_V, __pyx_v_a, __pyx_v_z_e, __pyx_v_pdf_err) / __pyx_v_ZT); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":100 * fe = pdf_V(x - t_e, v, V, a, z_e, pdf_err)/ZT * * Sleft = (h/12)*(f_beg + 4*fd + f_mid) # <<<<<<<<<<<<<< * Sright = (h/12)*(f_mid + 4*fe + f_end) * S2 = Sleft + Sright */ __pyx_v_Sleft = ((__pyx_v_h / 12.0) * ((__pyx_v_f_beg + (4.0 * __pyx_v_fd)) + __pyx_v_f_mid)); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":101 * * Sleft = (h/12)*(f_beg + 4*fd + f_mid) * Sright = (h/12)*(f_mid + 4*fe + f_end) # <<<<<<<<<<<<<< * S2 = Sleft + Sright * if (bottom <= 0 or fabs(S2 - S) <= 15*simps_err): */ __pyx_v_Sright = ((__pyx_v_h / 12.0) * ((__pyx_v_f_mid + (4.0 * __pyx_v_fe)) + __pyx_v_f_end)); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":102 * Sleft = (h/12)*(f_beg + 4*fd + f_mid) * Sright = (h/12)*(f_mid + 4*fe + f_end) * S2 = Sleft + Sright # <<<<<<<<<<<<<< * if (bottom <= 0 or fabs(S2 - S) <= 15*simps_err): * return S2 + (S2 - S)/15 */ __pyx_v_S2 = (__pyx_v_Sleft + __pyx_v_Sright); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":103 * Sright = (h/12)*(f_mid + 4*fe + f_end) * S2 = Sleft + Sright * if (bottom <= 0 or fabs(S2 - S) <= 15*simps_err): # <<<<<<<<<<<<<< * return S2 + (S2 - S)/15 * return adaptiveSimpsonsAux(x, v, V, a, z, t, pdf_err, */ __pyx_t_1 = (__pyx_v_bottom <= 0); if (!__pyx_t_1) { __pyx_t_2 = (fabs((__pyx_v_S2 - __pyx_v_S)) <= (15.0 * __pyx_v_simps_err)); __pyx_t_3 = __pyx_t_2; } else { __pyx_t_3 = __pyx_t_1; } if (__pyx_t_3) { /* "/home/whyking/working/projects/hddm/src/integrate.pxi":104 * S2 = Sleft + Sright * if (bottom <= 0 or fabs(S2 - S) <= 15*simps_err): * return S2 + (S2 - S)/15 # <<<<<<<<<<<<<< * return adaptiveSimpsonsAux(x, v, V, a, z, t, pdf_err, * lb_z, z_c, lb_t, t_c, ZT, simps_err/2, */ __pyx_r = (__pyx_v_S2 + ((__pyx_v_S2 - __pyx_v_S) / 15.0)); goto __pyx_L0; goto __pyx_L4; } __pyx_L4:; /* "/home/whyking/working/projects/hddm/src/integrate.pxi":110 * adaptiveSimpsonsAux(x, v, V, a, z, t, pdf_err, * z_c, ub_z, t_c, ub_t, ZT, simps_err/2, * Sright, f_mid, f_end, fe, bottom-1) # <<<<<<<<<<<<<< * * cdef double adaptiveSimpsons_1D(double x, double v, double V, double a, double z, double t, */ __pyx_r = (__pyx_f_4wfpt_adaptiveSimpsonsAux(__pyx_v_x, __pyx_v_v, __pyx_v_V, __pyx_v_a, __pyx_v_z, __pyx_v_t, __pyx_v_pdf_err, __pyx_v_lb_z, __pyx_v_z_c, __pyx_v_lb_t, __pyx_v_t_c, __pyx_v_ZT, (__pyx_v_simps_err / 2.0), __pyx_v_Sleft, __pyx_v_f_beg, __pyx_v_f_mid, __pyx_v_fd, (__pyx_v_bottom - 1)) + __pyx_f_4wfpt_adaptiveSimpsonsAux(__pyx_v_x, __pyx_v_v, __pyx_v_V, __pyx_v_a, __pyx_v_z, __pyx_v_t, __pyx_v_pdf_err, __pyx_v_z_c, __pyx_v_ub_z, __pyx_v_t_c, __pyx_v_ub_t, __pyx_v_ZT, (__pyx_v_simps_err / 2.0), __pyx_v_Sright, __pyx_v_f_mid, __pyx_v_f_end, __pyx_v_fe, (__pyx_v_bottom - 1))); goto __pyx_L0; __pyx_r = 0; __pyx_L0:; return __pyx_r; } /* "/home/whyking/working/projects/hddm/src/integrate.pxi":112 * Sright, f_mid, f_end, fe, bottom-1) * * cdef double adaptiveSimpsons_1D(double x, double v, double V, double a, double z, double t, # <<<<<<<<<<<<<< * double pdf_err, double lb_z, double ub_z, double lb_t, double ub_t, * double simps_err, int maxRecursionDepth) nogil: */ static double __pyx_f_4wfpt_adaptiveSimpsons_1D(double __pyx_v_x, double __pyx_v_v, double __pyx_v_V, double __pyx_v_a, double __pyx_v_z, double __pyx_v_t, double __pyx_v_pdf_err, double __pyx_v_lb_z, double __pyx_v_ub_z, double __pyx_v_lb_t, double __pyx_v_ub_t, double __pyx_v_simps_err, int __pyx_v_maxRecursionDepth) { double __pyx_v_h; double __pyx_v_ZT; double __pyx_v_c_t; double __pyx_v_c_z; double __pyx_v_f_beg; double __pyx_v_f_end; double __pyx_v_f_mid; double __pyx_v_S; double __pyx_v_res; double __pyx_r; int __pyx_t_1; /* "/home/whyking/working/projects/hddm/src/integrate.pxi":118 * cdef double h * * if (ub_t - lb_t) == 0: #integration over z # <<<<<<<<<<<<<< * lb_t = t * ub_t = t */ __pyx_t_1 = ((__pyx_v_ub_t - __pyx_v_lb_t) == 0.0); if (__pyx_t_1) { /* "/home/whyking/working/projects/hddm/src/integrate.pxi":119 * * if (ub_t - lb_t) == 0: #integration over z * lb_t = t # <<<<<<<<<<<<<< * ub_t = t * h = ub_z - lb_z */ __pyx_v_lb_t = __pyx_v_t; /* "/home/whyking/working/projects/hddm/src/integrate.pxi":120 * if (ub_t - lb_t) == 0: #integration over z * lb_t = t * ub_t = t # <<<<<<<<<<<<<< * h = ub_z - lb_z * else: #integration over t */ __pyx_v_ub_t = __pyx_v_t; /* "/home/whyking/working/projects/hddm/src/integrate.pxi":121 * lb_t = t * ub_t = t * h = ub_z - lb_z # <<<<<<<<<<<<<< * else: #integration over t * h = (ub_t-lb_t) */ __pyx_v_h = (__pyx_v_ub_z - __pyx_v_lb_z); goto __pyx_L3; } /*else*/ { /* "/home/whyking/working/projects/hddm/src/integrate.pxi":123 * h = ub_z - lb_z * else: #integration over t * h = (ub_t-lb_t) # <<<<<<<<<<<<<< * lb_z = z * ub_z = z */ __pyx_v_h = (__pyx_v_ub_t - __pyx_v_lb_t); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":124 * else: #integration over t * h = (ub_t-lb_t) * lb_z = z # <<<<<<<<<<<<<< * ub_z = z * */ __pyx_v_lb_z = __pyx_v_z; /* "/home/whyking/working/projects/hddm/src/integrate.pxi":125 * h = (ub_t-lb_t) * lb_z = z * ub_z = z # <<<<<<<<<<<<<< * * cdef double ZT = h */ __pyx_v_ub_z = __pyx_v_z; } __pyx_L3:; /* "/home/whyking/working/projects/hddm/src/integrate.pxi":127 * ub_z = z * * cdef double ZT = h # <<<<<<<<<<<<<< * cdef double c_t = (lb_t + ub_t)/2. * cdef double c_z = (lb_z + ub_z)/2. */ __pyx_v_ZT = __pyx_v_h; /* "/home/whyking/working/projects/hddm/src/integrate.pxi":128 * * cdef double ZT = h * cdef double c_t = (lb_t + ub_t)/2. # <<<<<<<<<<<<<< * cdef double c_z = (lb_z + ub_z)/2. * */ __pyx_v_c_t = ((__pyx_v_lb_t + __pyx_v_ub_t) / 2.); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":129 * cdef double ZT = h * cdef double c_t = (lb_t + ub_t)/2. * cdef double c_z = (lb_z + ub_z)/2. # <<<<<<<<<<<<<< * * cdef double f_beg, f_end, f_mid, S */ __pyx_v_c_z = ((__pyx_v_lb_z + __pyx_v_ub_z) / 2.); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":132 * * cdef double f_beg, f_end, f_mid, S * f_beg = pdf_V(x - lb_t, v, V, a, lb_z, pdf_err)/ZT # <<<<<<<<<<<<<< * f_end = pdf_V(x - ub_t, v, V, a, ub_z, pdf_err)/ZT * f_mid = pdf_V(x - c_t, v, V, a, c_z, pdf_err)/ZT */ __pyx_v_f_beg = (__pyx_f_4wfpt_pdf_V((__pyx_v_x - __pyx_v_lb_t), __pyx_v_v, __pyx_v_V, __pyx_v_a, __pyx_v_lb_z, __pyx_v_pdf_err) / __pyx_v_ZT); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":133 * cdef double f_beg, f_end, f_mid, S * f_beg = pdf_V(x - lb_t, v, V, a, lb_z, pdf_err)/ZT * f_end = pdf_V(x - ub_t, v, V, a, ub_z, pdf_err)/ZT # <<<<<<<<<<<<<< * f_mid = pdf_V(x - c_t, v, V, a, c_z, pdf_err)/ZT * S = (h/6)*(f_beg + 4*f_mid + f_end) */ __pyx_v_f_end = (__pyx_f_4wfpt_pdf_V((__pyx_v_x - __pyx_v_ub_t), __pyx_v_v, __pyx_v_V, __pyx_v_a, __pyx_v_ub_z, __pyx_v_pdf_err) / __pyx_v_ZT); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":134 * f_beg = pdf_V(x - lb_t, v, V, a, lb_z, pdf_err)/ZT * f_end = pdf_V(x - ub_t, v, V, a, ub_z, pdf_err)/ZT * f_mid = pdf_V(x - c_t, v, V, a, c_z, pdf_err)/ZT # <<<<<<<<<<<<<< * S = (h/6)*(f_beg + 4*f_mid + f_end) * cdef double res = adaptiveSimpsonsAux(x, v, V, a, z, t, pdf_err, */ __pyx_v_f_mid = (__pyx_f_4wfpt_pdf_V((__pyx_v_x - __pyx_v_c_t), __pyx_v_v, __pyx_v_V, __pyx_v_a, __pyx_v_c_z, __pyx_v_pdf_err) / __pyx_v_ZT); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":135 * f_end = pdf_V(x - ub_t, v, V, a, ub_z, pdf_err)/ZT * f_mid = pdf_V(x - c_t, v, V, a, c_z, pdf_err)/ZT * S = (h/6)*(f_beg + 4*f_mid + f_end) # <<<<<<<<<<<<<< * cdef double res = adaptiveSimpsonsAux(x, v, V, a, z, t, pdf_err, * lb_z, ub_z, lb_t, ub_t, ZT, simps_err, */ __pyx_v_S = ((__pyx_v_h / 6.0) * ((__pyx_v_f_beg + (4.0 * __pyx_v_f_mid)) + __pyx_v_f_end)); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":138 * cdef double res = adaptiveSimpsonsAux(x, v, V, a, z, t, pdf_err, * lb_z, ub_z, lb_t, ub_t, ZT, simps_err, * S, f_beg, f_end, f_mid, maxRecursionDepth) # <<<<<<<<<<<<<< * return res * */ __pyx_v_res = __pyx_f_4wfpt_adaptiveSimpsonsAux(__pyx_v_x, __pyx_v_v, __pyx_v_V, __pyx_v_a, __pyx_v_z, __pyx_v_t, __pyx_v_pdf_err, __pyx_v_lb_z, __pyx_v_ub_z, __pyx_v_lb_t, __pyx_v_ub_t, __pyx_v_ZT, __pyx_v_simps_err, __pyx_v_S, __pyx_v_f_beg, __pyx_v_f_end, __pyx_v_f_mid, __pyx_v_maxRecursionDepth); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":139 * lb_z, ub_z, lb_t, ub_t, ZT, simps_err, * S, f_beg, f_end, f_mid, maxRecursionDepth) * return res # <<<<<<<<<<<<<< * * cdef double adaptiveSimpsonsAux_2D(double x, double v, double V, */ __pyx_r = __pyx_v_res; goto __pyx_L0; __pyx_r = 0; __pyx_L0:; return __pyx_r; } /* "/home/whyking/working/projects/hddm/src/integrate.pxi":141 * return res * * cdef double adaptiveSimpsonsAux_2D(double x, double v, double V, # <<<<<<<<<<<<<< * double a, double z, double t, double * pdf_err, double err_1d, double lb_z, */ static double __pyx_f_4wfpt_adaptiveSimpsonsAux_2D(double __pyx_v_x, double __pyx_v_v, double __pyx_v_V, double __pyx_v_a, double __pyx_v_z, double __pyx_v_t, double __pyx_v_pdf_err, double __pyx_v_err_1d, double __pyx_v_lb_z, double __pyx_v_ub_z, double __pyx_v_lb_t, double __pyx_v_ub_t, double __pyx_v_T, double __pyx_v_err_2d, double __pyx_v_S, double __pyx_v_f_beg, double __pyx_v_f_end, double __pyx_v_f_mid, int __pyx_v_maxRecursionDepth_Z, int __pyx_v_bottom) { double __pyx_v_fd; double __pyx_v_fe; double __pyx_v_Sleft; double __pyx_v_Sright; double __pyx_v_S2; double __pyx_v_t_c; double __pyx_v_t_d; double __pyx_v_t_e; double __pyx_v_h; double __pyx_r; int __pyx_t_1; int __pyx_t_2; int __pyx_t_3; /* "/home/whyking/working/projects/hddm/src/integrate.pxi":154 * #print "in AdaptiveSimpsAux_2D: lb_z: %f, ub_z: %f, lb_t %f, ub_t %f, f_beg: %f, f_end: %f, bottom: %d" % (lb_z, ub_z, lb_t, ub_t, f_beg, f_end, bottom) * * cdef double t_c = (ub_t + lb_t)/2. # <<<<<<<<<<<<<< * cdef double t_d = (lb_t + t_c)/2. * cdef double t_e = (t_c + ub_t)/2. */ __pyx_v_t_c = ((__pyx_v_ub_t + __pyx_v_lb_t) / 2.); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":155 * * cdef double t_c = (ub_t + lb_t)/2. * cdef double t_d = (lb_t + t_c)/2. # <<<<<<<<<<<<<< * cdef double t_e = (t_c + ub_t)/2. * cdef double h = ub_t - lb_t */ __pyx_v_t_d = ((__pyx_v_lb_t + __pyx_v_t_c) / 2.); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":156 * cdef double t_c = (ub_t + lb_t)/2. * cdef double t_d = (lb_t + t_c)/2. * cdef double t_e = (t_c + ub_t)/2. # <<<<<<<<<<<<<< * cdef double h = ub_t - lb_t * */ __pyx_v_t_e = ((__pyx_v_t_c + __pyx_v_ub_t) / 2.); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":157 * cdef double t_d = (lb_t + t_c)/2. * cdef double t_e = (t_c + ub_t)/2. * cdef double h = ub_t - lb_t # <<<<<<<<<<<<<< * * fd = adaptiveSimpsons_1D(x, v, V, a, z, t_d, pdf_err, lb_z, ub_z, */ __pyx_v_h = (__pyx_v_ub_t - __pyx_v_lb_t); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":160 * * fd = adaptiveSimpsons_1D(x, v, V, a, z, t_d, pdf_err, lb_z, ub_z, * 0, 0, err_1d, maxRecursionDepth_Z)/T # <<<<<<<<<<<<<< * fe = adaptiveSimpsons_1D(x, v, V, a, z, t_e, pdf_err, lb_z, ub_z, * 0, 0, err_1d, maxRecursionDepth_Z)/T */ __pyx_v_fd = (__pyx_f_4wfpt_adaptiveSimpsons_1D(__pyx_v_x, __pyx_v_v, __pyx_v_V, __pyx_v_a, __pyx_v_z, __pyx_v_t_d, __pyx_v_pdf_err, __pyx_v_lb_z, __pyx_v_ub_z, 0.0, 0.0, __pyx_v_err_1d, __pyx_v_maxRecursionDepth_Z) / __pyx_v_T); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":162 * 0, 0, err_1d, maxRecursionDepth_Z)/T * fe = adaptiveSimpsons_1D(x, v, V, a, z, t_e, pdf_err, lb_z, ub_z, * 0, 0, err_1d, maxRecursionDepth_Z)/T # <<<<<<<<<<<<<< * * Sleft = (h/12)*(f_beg + 4*fd + f_mid) */ __pyx_v_fe = (__pyx_f_4wfpt_adaptiveSimpsons_1D(__pyx_v_x, __pyx_v_v, __pyx_v_V, __pyx_v_a, __pyx_v_z, __pyx_v_t_e, __pyx_v_pdf_err, __pyx_v_lb_z, __pyx_v_ub_z, 0.0, 0.0, __pyx_v_err_1d, __pyx_v_maxRecursionDepth_Z) / __pyx_v_T); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":164 * 0, 0, err_1d, maxRecursionDepth_Z)/T * * Sleft = (h/12)*(f_beg + 4*fd + f_mid) # <<<<<<<<<<<<<< * Sright = (h/12)*(f_mid + 4*fe + f_end) * S2 = Sleft + Sright */ __pyx_v_Sleft = ((__pyx_v_h / 12.0) * ((__pyx_v_f_beg + (4.0 * __pyx_v_fd)) + __pyx_v_f_mid)); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":165 * * Sleft = (h/12)*(f_beg + 4*fd + f_mid) * Sright = (h/12)*(f_mid + 4*fe + f_end) # <<<<<<<<<<<<<< * S2 = Sleft + Sright * */ __pyx_v_Sright = ((__pyx_v_h / 12.0) * ((__pyx_v_f_mid + (4.0 * __pyx_v_fe)) + __pyx_v_f_end)); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":166 * Sleft = (h/12)*(f_beg + 4*fd + f_mid) * Sright = (h/12)*(f_mid + 4*fe + f_end) * S2 = Sleft + Sright # <<<<<<<<<<<<<< * * if (bottom <= 0 or fabs(S2 - S) <= 15*err_2d): */ __pyx_v_S2 = (__pyx_v_Sleft + __pyx_v_Sright); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":168 * S2 = Sleft + Sright * * if (bottom <= 0 or fabs(S2 - S) <= 15*err_2d): # <<<<<<<<<<<<<< * return S2 + (S2 - S)/15; * */ __pyx_t_1 = (__pyx_v_bottom <= 0); if (!__pyx_t_1) { __pyx_t_2 = (fabs((__pyx_v_S2 - __pyx_v_S)) <= (15.0 * __pyx_v_err_2d)); __pyx_t_3 = __pyx_t_2; } else { __pyx_t_3 = __pyx_t_1; } if (__pyx_t_3) { /* "/home/whyking/working/projects/hddm/src/integrate.pxi":169 * * if (bottom <= 0 or fabs(S2 - S) <= 15*err_2d): * return S2 + (S2 - S)/15; # <<<<<<<<<<<<<< * * return adaptiveSimpsonsAux_2D(x, v, V, a, z, t, pdf_err, err_1d, */ __pyx_r = (__pyx_v_S2 + ((__pyx_v_S2 - __pyx_v_S) / 15.0)); goto __pyx_L0; goto __pyx_L3; } __pyx_L3:; /* "/home/whyking/working/projects/hddm/src/integrate.pxi":176 * adaptiveSimpsonsAux_2D(x, v, V, a, z, t, pdf_err, err_1d, * lb_z, ub_z, t_c, ub_t, T, err_2d/2, * Sright, f_mid, f_end, fe, maxRecursionDepth_Z, bottom-1) # <<<<<<<<<<<<<< * * */ __pyx_r = (__pyx_f_4wfpt_adaptiveSimpsonsAux_2D(__pyx_v_x, __pyx_v_v, __pyx_v_V, __pyx_v_a, __pyx_v_z, __pyx_v_t, __pyx_v_pdf_err, __pyx_v_err_1d, __pyx_v_lb_z, __pyx_v_ub_z, __pyx_v_lb_t, __pyx_v_t_c, __pyx_v_T, (__pyx_v_err_2d / 2.0), __pyx_v_Sleft, __pyx_v_f_beg, __pyx_v_f_mid, __pyx_v_fd, __pyx_v_maxRecursionDepth_Z, (__pyx_v_bottom - 1)) + __pyx_f_4wfpt_adaptiveSimpsonsAux_2D(__pyx_v_x, __pyx_v_v, __pyx_v_V, __pyx_v_a, __pyx_v_z, __pyx_v_t, __pyx_v_pdf_err, __pyx_v_err_1d, __pyx_v_lb_z, __pyx_v_ub_z, __pyx_v_t_c, __pyx_v_ub_t, __pyx_v_T, (__pyx_v_err_2d / 2.0), __pyx_v_Sright, __pyx_v_f_mid, __pyx_v_f_end, __pyx_v_fe, __pyx_v_maxRecursionDepth_Z, (__pyx_v_bottom - 1))); goto __pyx_L0; __pyx_r = 0; __pyx_L0:; return __pyx_r; } /* "/home/whyking/working/projects/hddm/src/integrate.pxi":179 * * * cdef double adaptiveSimpsons_2D(double x, double v, double V, double a, double z, double t, # <<<<<<<<<<<<<< * double pdf_err, double lb_z, double ub_z, double lb_t, double ub_t, * double simps_err, int maxRecursionDepth_Z, int maxRecursionDepth_T) nogil: */ static double __pyx_f_4wfpt_adaptiveSimpsons_2D(double __pyx_v_x, double __pyx_v_v, double __pyx_v_V, double __pyx_v_a, double __pyx_v_z, double __pyx_v_t, double __pyx_v_pdf_err, double __pyx_v_lb_z, double __pyx_v_ub_z, double __pyx_v_lb_t, double __pyx_v_ub_t, double __pyx_v_simps_err, int __pyx_v_maxRecursionDepth_Z, int __pyx_v_maxRecursionDepth_T) { double __pyx_v_h; double __pyx_v_T; CYTHON_UNUSED double __pyx_v_c_t; CYTHON_UNUSED double __pyx_v_c_z; double __pyx_v_f_beg; double __pyx_v_f_end; double __pyx_v_f_mid; double __pyx_v_S; double __pyx_v_err_1d; double __pyx_v_err_2d; double __pyx_v_res; double __pyx_r; /* "/home/whyking/working/projects/hddm/src/integrate.pxi":183 * double simps_err, int maxRecursionDepth_Z, int maxRecursionDepth_T) nogil: * * cdef double h = (ub_t-lb_t) # <<<<<<<<<<<<<< * * cdef double T = (ub_t - lb_t) */ __pyx_v_h = (__pyx_v_ub_t - __pyx_v_lb_t); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":185 * cdef double h = (ub_t-lb_t) * * cdef double T = (ub_t - lb_t) # <<<<<<<<<<<<<< * cdef double c_t = (lb_t + ub_t)/2. * cdef double c_z = (lb_z + ub_z)/2. */ __pyx_v_T = (__pyx_v_ub_t - __pyx_v_lb_t); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":186 * * cdef double T = (ub_t - lb_t) * cdef double c_t = (lb_t + ub_t)/2. # <<<<<<<<<<<<<< * cdef double c_z = (lb_z + ub_z)/2. * */ __pyx_v_c_t = ((__pyx_v_lb_t + __pyx_v_ub_t) / 2.); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":187 * cdef double T = (ub_t - lb_t) * cdef double c_t = (lb_t + ub_t)/2. * cdef double c_z = (lb_z + ub_z)/2. # <<<<<<<<<<<<<< * * cdef double f_beg, f_end, f_mid, S */ __pyx_v_c_z = ((__pyx_v_lb_z + __pyx_v_ub_z) / 2.); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":190 * * cdef double f_beg, f_end, f_mid, S * cdef double err_1d = simps_err # <<<<<<<<<<<<<< * cdef double err_2d = simps_err * */ __pyx_v_err_1d = __pyx_v_simps_err; /* "/home/whyking/working/projects/hddm/src/integrate.pxi":191 * cdef double f_beg, f_end, f_mid, S * cdef double err_1d = simps_err * cdef double err_2d = simps_err # <<<<<<<<<<<<<< * * f_beg = adaptiveSimpsons_1D(x, v, V, a, z, lb_t, pdf_err, lb_z, ub_z, */ __pyx_v_err_2d = __pyx_v_simps_err; /* "/home/whyking/working/projects/hddm/src/integrate.pxi":194 * * f_beg = adaptiveSimpsons_1D(x, v, V, a, z, lb_t, pdf_err, lb_z, ub_z, * 0, 0, err_1d, maxRecursionDepth_Z)/T # <<<<<<<<<<<<<< * * f_end = adaptiveSimpsons_1D(x, v, V, a, z, ub_t, pdf_err, lb_z, ub_z, */ __pyx_v_f_beg = (__pyx_f_4wfpt_adaptiveSimpsons_1D(__pyx_v_x, __pyx_v_v, __pyx_v_V, __pyx_v_a, __pyx_v_z, __pyx_v_lb_t, __pyx_v_pdf_err, __pyx_v_lb_z, __pyx_v_ub_z, 0.0, 0.0, __pyx_v_err_1d, __pyx_v_maxRecursionDepth_Z) / __pyx_v_T); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":197 * * f_end = adaptiveSimpsons_1D(x, v, V, a, z, ub_t, pdf_err, lb_z, ub_z, * 0, 0, err_1d, maxRecursionDepth_Z)/T # <<<<<<<<<<<<<< * f_mid = adaptiveSimpsons_1D(x, v, V, a, z, (lb_t+ub_t)/2, pdf_err, lb_z, ub_z, * 0, 0, err_1d, maxRecursionDepth_Z)/T */ __pyx_v_f_end = (__pyx_f_4wfpt_adaptiveSimpsons_1D(__pyx_v_x, __pyx_v_v, __pyx_v_V, __pyx_v_a, __pyx_v_z, __pyx_v_ub_t, __pyx_v_pdf_err, __pyx_v_lb_z, __pyx_v_ub_z, 0.0, 0.0, __pyx_v_err_1d, __pyx_v_maxRecursionDepth_Z) / __pyx_v_T); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":199 * 0, 0, err_1d, maxRecursionDepth_Z)/T * f_mid = adaptiveSimpsons_1D(x, v, V, a, z, (lb_t+ub_t)/2, pdf_err, lb_z, ub_z, * 0, 0, err_1d, maxRecursionDepth_Z)/T # <<<<<<<<<<<<<< * S = (h/6)*(f_beg + 4*f_mid + f_end) * cdef double res = adaptiveSimpsonsAux_2D(x, v, V, a, z, t, pdf_err, err_1d, */ __pyx_v_f_mid = (__pyx_f_4wfpt_adaptiveSimpsons_1D(__pyx_v_x, __pyx_v_v, __pyx_v_V, __pyx_v_a, __pyx_v_z, ((__pyx_v_lb_t + __pyx_v_ub_t) / 2.0), __pyx_v_pdf_err, __pyx_v_lb_z, __pyx_v_ub_z, 0.0, 0.0, __pyx_v_err_1d, __pyx_v_maxRecursionDepth_Z) / __pyx_v_T); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":200 * f_mid = adaptiveSimpsons_1D(x, v, V, a, z, (lb_t+ub_t)/2, pdf_err, lb_z, ub_z, * 0, 0, err_1d, maxRecursionDepth_Z)/T * S = (h/6)*(f_beg + 4*f_mid + f_end) # <<<<<<<<<<<<<< * cdef double res = adaptiveSimpsonsAux_2D(x, v, V, a, z, t, pdf_err, err_1d, * lb_z, ub_z, lb_t, ub_t, T, err_2d, */ __pyx_v_S = ((__pyx_v_h / 6.0) * ((__pyx_v_f_beg + (4.0 * __pyx_v_f_mid)) + __pyx_v_f_end)); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":203 * cdef double res = adaptiveSimpsonsAux_2D(x, v, V, a, z, t, pdf_err, err_1d, * lb_z, ub_z, lb_t, ub_t, T, err_2d, * S, f_beg, f_end, f_mid, maxRecursionDepth_Z, maxRecursionDepth_T) # <<<<<<<<<<<<<< * return res */ __pyx_v_res = __pyx_f_4wfpt_adaptiveSimpsonsAux_2D(__pyx_v_x, __pyx_v_v, __pyx_v_V, __pyx_v_a, __pyx_v_z, __pyx_v_t, __pyx_v_pdf_err, __pyx_v_err_1d, __pyx_v_lb_z, __pyx_v_ub_z, __pyx_v_lb_t, __pyx_v_ub_t, __pyx_v_T, __pyx_v_err_2d, __pyx_v_S, __pyx_v_f_beg, __pyx_v_f_end, __pyx_v_f_mid, __pyx_v_maxRecursionDepth_Z, __pyx_v_maxRecursionDepth_T); /* "/home/whyking/working/projects/hddm/src/integrate.pxi":204 * lb_z, ub_z, lb_t, ub_t, T, err_2d, * S, f_beg, f_end, f_mid, maxRecursionDepth_Z, maxRecursionDepth_T) * return res # <<<<<<<<<<<<<< */ __pyx_r = __pyx_v_res; goto __pyx_L0; __pyx_r = 0; __pyx_L0:; return __pyx_r; } /* "/home/whyking/working/projects/hddm/src/pdf.pxi":25 * double M_PI * * cdef double ftt_01w(double tt, double w, double err) nogil: # <<<<<<<<<<<<<< * """Compute f(t|0,1,w) for the likelihood of the drift diffusion model using the method * and implementation of Navarro & Fuss, 2009. */ static double __pyx_f_4wfpt_ftt_01w(double __pyx_v_tt, double __pyx_v_w, double __pyx_v_err) { double __pyx_v_kl; double __pyx_v_ks; double __pyx_v_p; int __pyx_v_k; int __pyx_v_K; int __pyx_v_lower; int __pyx_v_upper; double __pyx_r; int __pyx_t_1; int __pyx_t_2; /* "/home/whyking/working/projects/hddm/src/pdf.pxi":33 * * # calculate number of terms needed for large t * if M_PI*tt*err<1: # if error threshold is set low enough # <<<<<<<<<<<<<< * kl=sqrt(-2*log(M_PI*tt*err)/(M_PI**2*tt)) # bound * kl=fmax(kl,1./(M_PI*sqrt(tt))) # ensure boundary conditions met */ __pyx_t_1 = (((M_PI * __pyx_v_tt) * __pyx_v_err) < 1.0); if (__pyx_t_1) { /* "/home/whyking/working/projects/hddm/src/pdf.pxi":34 * # calculate number of terms needed for large t * if M_PI*tt*err<1: # if error threshold is set low enough * kl=sqrt(-2*log(M_PI*tt*err)/(M_PI**2*tt)) # bound # <<<<<<<<<<<<<< * kl=fmax(kl,1./(M_PI*sqrt(tt))) # ensure boundary conditions met * else: # if error threshold set too high */ __pyx_v_kl = sqrt(((-2.0 * log(((M_PI * __pyx_v_tt) * __pyx_v_err))) / (pow(M_PI, 2.0) * __pyx_v_tt))); /* "/home/whyking/working/projects/hddm/src/pdf.pxi":35 * if M_PI*tt*err<1: # if error threshold is set low enough * kl=sqrt(-2*log(M_PI*tt*err)/(M_PI**2*tt)) # bound * kl=fmax(kl,1./(M_PI*sqrt(tt))) # ensure boundary conditions met # <<<<<<<<<<<<<< * else: # if error threshold set too high * kl=1./(M_PI*sqrt(tt)) # set to boundary condition */ __pyx_v_kl = fmax(__pyx_v_kl, (1. / (M_PI * sqrt(__pyx_v_tt)))); goto __pyx_L3; } /*else*/ { /* "/home/whyking/working/projects/hddm/src/pdf.pxi":37 * kl=fmax(kl,1./(M_PI*sqrt(tt))) # ensure boundary conditions met * else: # if error threshold set too high * kl=1./(M_PI*sqrt(tt)) # set to boundary condition # <<<<<<<<<<<<<< * * # calculate number of terms needed for small t */ __pyx_v_kl = (1. / (M_PI * sqrt(__pyx_v_tt))); } __pyx_L3:; /* "/home/whyking/working/projects/hddm/src/pdf.pxi":40 * * # calculate number of terms needed for small t * if 2*sqrt(2*M_PI*tt)*err<1: # if error threshold is set low enough # <<<<<<<<<<<<<< * ks=2+sqrt(-2*tt*log(2*sqrt(2*M_PI*tt)*err)) # bound * ks=fmax(ks,sqrt(tt)+1) # ensure boundary conditions are met */ __pyx_t_1 = (((2.0 * sqrt(((2.0 * M_PI) * __pyx_v_tt))) * __pyx_v_err) < 1.0); if (__pyx_t_1) { /* "/home/whyking/working/projects/hddm/src/pdf.pxi":41 * # calculate number of terms needed for small t * if 2*sqrt(2*M_PI*tt)*err<1: # if error threshold is set low enough * ks=2+sqrt(-2*tt*log(2*sqrt(2*M_PI*tt)*err)) # bound # <<<<<<<<<<<<<< * ks=fmax(ks,sqrt(tt)+1) # ensure boundary conditions are met * else: # if error threshold was set too high */ __pyx_v_ks = (2.0 + sqrt(((-2.0 * __pyx_v_tt) * log(((2.0 * sqrt(((2.0 * M_PI) * __pyx_v_tt))) * __pyx_v_err))))); /* "/home/whyking/working/projects/hddm/src/pdf.pxi":42 * if 2*sqrt(2*M_PI*tt)*err<1: # if error threshold is set low enough * ks=2+sqrt(-2*tt*log(2*sqrt(2*M_PI*tt)*err)) # bound * ks=fmax(ks,sqrt(tt)+1) # ensure boundary conditions are met # <<<<<<<<<<<<<< * else: # if error threshold was set too high * ks=2 # minimal kappa for that case */ __pyx_v_ks = fmax(__pyx_v_ks, (sqrt(__pyx_v_tt) + 1.0)); goto __pyx_L4; } /*else*/ { /* "/home/whyking/working/projects/hddm/src/pdf.pxi":44 * ks=fmax(ks,sqrt(tt)+1) # ensure boundary conditions are met * else: # if error threshold was set too high * ks=2 # minimal kappa for that case # <<<<<<<<<<<<<< * * # compute f(tt|0,1,w) */ __pyx_v_ks = 2.0; } __pyx_L4:; /* "/home/whyking/working/projects/hddm/src/pdf.pxi":47 * * # compute f(tt|0,1,w) * p=0 #initialize density # <<<<<<<<<<<<<< * if ks(ceil(ks)) # round to smallest integer meeting error */ __pyx_v_p = 0.0; /* "/home/whyking/working/projects/hddm/src/pdf.pxi":48 * # compute f(tt|0,1,w) * p=0 #initialize density * if ks(ceil(ks)) # round to smallest integer meeting error * lower = (-floor((K-1)/2.)) */ __pyx_t_1 = (__pyx_v_ks < __pyx_v_kl); if (__pyx_t_1) { /* "/home/whyking/working/projects/hddm/src/pdf.pxi":49 * p=0 #initialize density * if ks(ceil(ks)) # round to smallest integer meeting error # <<<<<<<<<<<<<< * lower = (-floor((K-1)/2.)) * upper = (ceil((K-1)/2.)) */ __pyx_v_K = ((int)ceil(__pyx_v_ks)); /* "/home/whyking/working/projects/hddm/src/pdf.pxi":50 * if ks(ceil(ks)) # round to smallest integer meeting error * lower = (-floor((K-1)/2.)) # <<<<<<<<<<<<<< * upper = (ceil((K-1)/2.)) * for k from lower <= k <= upper: # loop over k */ __pyx_v_lower = ((int)(-floor(((__pyx_v_K - 1) / 2.)))); /* "/home/whyking/working/projects/hddm/src/pdf.pxi":51 * K=(ceil(ks)) # round to smallest integer meeting error * lower = (-floor((K-1)/2.)) * upper = (ceil((K-1)/2.)) # <<<<<<<<<<<<<< * for k from lower <= k <= upper: # loop over k * p+=(w+2*k)*exp(-(pow((w+2*k),2))/2/tt) # increment sum */ __pyx_v_upper = ((int)ceil(((__pyx_v_K - 1) / 2.))); /* "/home/whyking/working/projects/hddm/src/pdf.pxi":52 * lower = (-floor((K-1)/2.)) * upper = (ceil((K-1)/2.)) * for k from lower <= k <= upper: # loop over k # <<<<<<<<<<<<<< * p+=(w+2*k)*exp(-(pow((w+2*k),2))/2/tt) # increment sum * p/=sqrt(2*M_PI*pow(tt,3)) # add constant term */ __pyx_t_2 = __pyx_v_upper; for (__pyx_v_k = __pyx_v_lower; __pyx_v_k <= __pyx_t_2; __pyx_v_k++) { /* "/home/whyking/working/projects/hddm/src/pdf.pxi":53 * upper = (ceil((K-1)/2.)) * for k from lower <= k <= upper: # loop over k * p+=(w+2*k)*exp(-(pow((w+2*k),2))/2/tt) # increment sum # <<<<<<<<<<<<<< * p/=sqrt(2*M_PI*pow(tt,3)) # add constant term * */ __pyx_v_p = (__pyx_v_p + ((__pyx_v_w + (2 * __pyx_v_k)) * exp((((-pow((__pyx_v_w + (2 * __pyx_v_k)), 2.0)) / 2.0) / __pyx_v_tt)))); } /* "/home/whyking/working/projects/hddm/src/pdf.pxi":54 * for k from lower <= k <= upper: # loop over k * p+=(w+2*k)*exp(-(pow((w+2*k),2))/2/tt) # increment sum * p/=sqrt(2*M_PI*pow(tt,3)) # add constant term # <<<<<<<<<<<<<< * * else: # if large t is better... */ __pyx_v_p = (__pyx_v_p / sqrt(((2.0 * M_PI) * pow(__pyx_v_tt, 3.0)))); goto __pyx_L5; } /*else*/ { /* "/home/whyking/working/projects/hddm/src/pdf.pxi":57 * * else: # if large t is better... * K=(ceil(kl)) # round to smallest integer meeting error # <<<<<<<<<<<<<< * for k from 1 <= k <= K: * p+=k*exp(-(pow(k,2))*(M_PI**2)*tt/2)*sin(k*M_PI*w) # increment sum */ __pyx_v_K = ((int)ceil(__pyx_v_kl)); /* "/home/whyking/working/projects/hddm/src/pdf.pxi":58 * else: # if large t is better... * K=(ceil(kl)) # round to smallest integer meeting error * for k from 1 <= k <= K: # <<<<<<<<<<<<<< * p+=k*exp(-(pow(k,2))*(M_PI**2)*tt/2)*sin(k*M_PI*w) # increment sum * p*=M_PI # add constant term */ __pyx_t_2 = __pyx_v_K; for (__pyx_v_k = 1; __pyx_v_k <= __pyx_t_2; __pyx_v_k++) { /* "/home/whyking/working/projects/hddm/src/pdf.pxi":59 * K=(ceil(kl)) # round to smallest integer meeting error * for k from 1 <= k <= K: * p+=k*exp(-(pow(k,2))*(M_PI**2)*tt/2)*sin(k*M_PI*w) # increment sum # <<<<<<<<<<<<<< * p*=M_PI # add constant term * */ __pyx_v_p = (__pyx_v_p + ((__pyx_v_k * exp(((((-pow(__pyx_v_k, 2.0)) * pow(M_PI, 2.0)) * __pyx_v_tt) / 2.0))) * sin(((__pyx_v_k * M_PI) * __pyx_v_w)))); } /* "/home/whyking/working/projects/hddm/src/pdf.pxi":60 * for k from 1 <= k <= K: * p+=k*exp(-(pow(k,2))*(M_PI**2)*tt/2)*sin(k*M_PI*w) # increment sum * p*=M_PI # add constant term # <<<<<<<<<<<<<< * * return p */ __pyx_v_p = (__pyx_v_p * M_PI); } __pyx_L5:; /* "/home/whyking/working/projects/hddm/src/pdf.pxi":62 * p*=M_PI # add constant term * * return p # <<<<<<<<<<<<<< * * cdef inline double prob_ub(double v, double a, double z) nogil: */ __pyx_r = __pyx_v_p; goto __pyx_L0; __pyx_r = 0; __pyx_L0:; return __pyx_r; } /* "/home/whyking/working/projects/hddm/src/pdf.pxi":64 * return p * * cdef inline double prob_ub(double v, double a, double z) nogil: # <<<<<<<<<<<<<< * """Probability of hitting upper boundary.""" * return (exp(-2*a*z*v) - 1) / (exp(-2*a*v) - 1) */ static CYTHON_INLINE double __pyx_f_4wfpt_prob_ub(double __pyx_v_v, double __pyx_v_a, double __pyx_v_z) { double __pyx_r; /* "/home/whyking/working/projects/hddm/src/pdf.pxi":66 * cdef inline double prob_ub(double v, double a, double z) nogil: * """Probability of hitting upper boundary.""" * return (exp(-2*a*z*v) - 1) / (exp(-2*a*v) - 1) # <<<<<<<<<<<<<< * * cdef double pdf(double x, double v, double a, double w, double err) nogil: */ __pyx_r = ((exp((((-2.0 * __pyx_v_a) * __pyx_v_z) * __pyx_v_v)) - 1.0) / (exp(((-2.0 * __pyx_v_a) * __pyx_v_v)) - 1.0)); goto __pyx_L0; __pyx_r = 0; __pyx_L0:; return __pyx_r; } /* "/home/whyking/working/projects/hddm/src/pdf.pxi":68 * return (exp(-2*a*z*v) - 1) / (exp(-2*a*v) - 1) * * cdef double pdf(double x, double v, double a, double w, double err) nogil: # <<<<<<<<<<<<<< * """Compute the likelihood of the drift diffusion model f(t|v,a,z) using the method * and implementation of Navarro & Fuss, 2009. */ static double __pyx_f_4wfpt_pdf(double __pyx_v_x, double __pyx_v_v, double __pyx_v_a, double __pyx_v_w, double __pyx_v_err) { double __pyx_v_tt; double __pyx_v_p; double __pyx_r; int __pyx_t_1; /* "/home/whyking/working/projects/hddm/src/pdf.pxi":72 * and implementation of Navarro & Fuss, 2009. * """ * if x <= 0: # <<<<<<<<<<<<<< * return 0 * */ __pyx_t_1 = (__pyx_v_x <= 0.0); if (__pyx_t_1) { /* "/home/whyking/working/projects/hddm/src/pdf.pxi":73 * """ * if x <= 0: * return 0 # <<<<<<<<<<<<<< * * cdef double tt = x/a**2 # use normalized time */ __pyx_r = 0.0; goto __pyx_L0; goto __pyx_L3; } __pyx_L3:; /* "/home/whyking/working/projects/hddm/src/pdf.pxi":75 * return 0 * * cdef double tt = x/a**2 # use normalized time # <<<<<<<<<<<<<< * cdef double p = ftt_01w(tt, w, err) #get f(t|0,1,w) * */ __pyx_v_tt = (__pyx_v_x / pow(__pyx_v_a, 2.0)); /* "/home/whyking/working/projects/hddm/src/pdf.pxi":76 * * cdef double tt = x/a**2 # use normalized time * cdef double p = ftt_01w(tt, w, err) #get f(t|0,1,w) # <<<<<<<<<<<<<< * * # convert to f(t|v,a,w) */ __pyx_v_p = __pyx_f_4wfpt_ftt_01w(__pyx_v_tt, __pyx_v_w, __pyx_v_err); /* "/home/whyking/working/projects/hddm/src/pdf.pxi":79 * * # convert to f(t|v,a,w) * return p*exp(-v*a*w -(pow(v,2))*x/2.)/(pow(a,2)) # <<<<<<<<<<<<<< * * cdef double pdf_V(double x, double v, double V, double a, double z, double err) nogil: */ __pyx_r = ((__pyx_v_p * exp(((((-__pyx_v_v) * __pyx_v_a) * __pyx_v_w) - ((pow(__pyx_v_v, 2.0) * __pyx_v_x) / 2.)))) / pow(__pyx_v_a, 2.0)); goto __pyx_L0; __pyx_r = 0; __pyx_L0:; return __pyx_r; } /* "/home/whyking/working/projects/hddm/src/pdf.pxi":81 * return p*exp(-v*a*w -(pow(v,2))*x/2.)/(pow(a,2)) * * cdef double pdf_V(double x, double v, double V, double a, double z, double err) nogil: # <<<<<<<<<<<<<< * """Compute the likelihood of the drift diffusion model f(t|v,a,z,V) using the method * and implementation of Navarro & Fuss, 2009. */ static double __pyx_f_4wfpt_pdf_V(double __pyx_v_x, double __pyx_v_v, double __pyx_v_V, double __pyx_v_a, double __pyx_v_z, double __pyx_v_err) { double __pyx_v_tt; double __pyx_v_p; double __pyx_r; int __pyx_t_1; /* "/home/whyking/working/projects/hddm/src/pdf.pxi":86 * V is the std of the drift rate * """ * if x <= 0: # <<<<<<<<<<<<<< * return 0 * */ __pyx_t_1 = (__pyx_v_x <= 0.0); if (__pyx_t_1) { /* "/home/whyking/working/projects/hddm/src/pdf.pxi":87 * """ * if x <= 0: * return 0 # <<<<<<<<<<<<<< * * if V==0: */ __pyx_r = 0.0; goto __pyx_L0; goto __pyx_L3; } __pyx_L3:; /* "/home/whyking/working/projects/hddm/src/pdf.pxi":89 * return 0 * * if V==0: # <<<<<<<<<<<<<< * return pdf(x, v, a, z, err) * */ __pyx_t_1 = (__pyx_v_V == 0.0); if (__pyx_t_1) { /* "/home/whyking/working/projects/hddm/src/pdf.pxi":90 * * if V==0: * return pdf(x, v, a, z, err) # <<<<<<<<<<<<<< * * cdef double tt = x/(pow(a,2)) # use normalized time */ __pyx_r = __pyx_f_4wfpt_pdf(__pyx_v_x, __pyx_v_v, __pyx_v_a, __pyx_v_z, __pyx_v_err); goto __pyx_L0; goto __pyx_L4; } __pyx_L4:; /* "/home/whyking/working/projects/hddm/src/pdf.pxi":92 * return pdf(x, v, a, z, err) * * cdef double tt = x/(pow(a,2)) # use normalized time # <<<<<<<<<<<<<< * cdef double p = ftt_01w(tt, z, err) #get f(t|0,1,w) * */ __pyx_v_tt = (__pyx_v_x / pow(__pyx_v_a, 2.0)); /* "/home/whyking/working/projects/hddm/src/pdf.pxi":93 * * cdef double tt = x/(pow(a,2)) # use normalized time * cdef double p = ftt_01w(tt, z, err) #get f(t|0,1,w) # <<<<<<<<<<<<<< * * # convert to f(t|v,a,w) */ __pyx_v_p = __pyx_f_4wfpt_ftt_01w(__pyx_v_tt, __pyx_v_z, __pyx_v_err); /* "/home/whyking/working/projects/hddm/src/pdf.pxi":96 * * # convert to f(t|v,a,w) * return exp(log(p) + ((a*z*V)**2 - 2*a*v*z - (v**2)*x)/(2*(V**2)*x+2))/sqrt((V**2)*x+1)/(a**2) # <<<<<<<<<<<<<< * * cpdef double full_pdf(double x, double v, double V, double a, double */ __pyx_r = ((exp((log(__pyx_v_p) + (((pow(((__pyx_v_a * __pyx_v_z) * __pyx_v_V), 2.0) - (((2.0 * __pyx_v_a) * __pyx_v_v) * __pyx_v_z)) - (pow(__pyx_v_v, 2.0) * __pyx_v_x)) / (((2.0 * pow(__pyx_v_V, 2.0)) * __pyx_v_x) + 2.0)))) / sqrt(((pow(__pyx_v_V, 2.0) * __pyx_v_x) + 1.0))) / pow(__pyx_v_a, 2.0)); goto __pyx_L0; __pyx_r = 0; __pyx_L0:; return __pyx_r; } /* "/home/whyking/working/projects/hddm/src/pdf.pxi":98 * return exp(log(p) + ((a*z*V)**2 - 2*a*v*z - (v**2)*x)/(2*(V**2)*x+2))/sqrt((V**2)*x+1)/(a**2) * * cpdef double full_pdf(double x, double v, double V, double a, double # <<<<<<<<<<<<<< * z, double Z, double t, double T, double err, int * nT=2, int nZ=2, bint use_adaptive=1, double */ static PyObject *__pyx_pw_4wfpt_1full_pdf(PyObject *__pyx_self, PyObject *__pyx_args, PyObject *__pyx_kwds); /*proto*/ static double __pyx_f_4wfpt_full_pdf(double __pyx_v_x, double __pyx_v_v, double __pyx_v_V, double __pyx_v_a, double __pyx_v_z, double __pyx_v_Z, double __pyx_v_t, double __pyx_v_T, double __pyx_v_err, CYTHON_UNUSED int __pyx_skip_dispatch, struct __pyx_opt_args_4wfpt_full_pdf *__pyx_optional_args) { int __pyx_v_nT = ((int)2); int __pyx_v_nZ = ((int)2); int __pyx_v_use_adaptive = ((int)1); /* "/home/whyking/working/projects/hddm/src/pdf.pxi":101 * z, double Z, double t, double T, double err, int * nT=2, int nZ=2, bint use_adaptive=1, double * simps_err=1e-3) nogil: # <<<<<<<<<<<<<< * """full pdf""" * */ double __pyx_v_simps_err = ((double)1e-3); double __pyx_r; int __pyx_t_1; int __pyx_t_2; int __pyx_t_3; int __pyx_t_4; int __pyx_t_5; int __pyx_t_6; int __pyx_t_7; int __pyx_t_8; int __pyx_t_9; if (__pyx_optional_args) { if (__pyx_optional_args->__pyx_n > 0) { __pyx_v_nT = __pyx_optional_args->nT; if (__pyx_optional_args->__pyx_n > 1) { __pyx_v_nZ = __pyx_optional_args->nZ; if (__pyx_optional_args->__pyx_n > 2) { __pyx_v_use_adaptive = __pyx_optional_args->use_adaptive; if (__pyx_optional_args->__pyx_n > 3) { __pyx_v_simps_err = __pyx_optional_args->simps_err; } } } } } /* "/home/whyking/working/projects/hddm/src/pdf.pxi":105 * * # Check if parpameters are valid * if z<0 or z>1 or a<0 or ((fabs(x)-(t-T/2.))<0) or (z+Z/2.>1) or (z-Z/2.<0) or (t-T/2.<0) or (t<0): # <<<<<<<<<<<<<< * return 0 * */ __pyx_t_1 = (__pyx_v_z < 0.0); if (!__pyx_t_1) { __pyx_t_2 = (__pyx_v_z > 1.0); if (!__pyx_t_2) { __pyx_t_3 = (__pyx_v_a < 0.0); if (!__pyx_t_3) { __pyx_t_4 = ((fabs(__pyx_v_x) - (__pyx_v_t - (__pyx_v_T / 2.))) < 0.0); if (!__pyx_t_4) { __pyx_t_5 = ((__pyx_v_z + (__pyx_v_Z / 2.)) > 1.0); if (!__pyx_t_5) { __pyx_t_6 = ((__pyx_v_z - (__pyx_v_Z / 2.)) < 0.0); if (!__pyx_t_6) { __pyx_t_7 = ((__pyx_v_t - (__pyx_v_T / 2.)) < 0.0); if (!__pyx_t_7) { __pyx_t_8 = (__pyx_v_t < 0.0); __pyx_t_9 = __pyx_t_8; } else { __pyx_t_9 = __pyx_t_7; } __pyx_t_7 = __pyx_t_9; } else { __pyx_t_7 = __pyx_t_6; } __pyx_t_6 = __pyx_t_7; } else { __pyx_t_6 = __pyx_t_5; } __pyx_t_5 = __pyx_t_6; } else { __pyx_t_5 = __pyx_t_4; } __pyx_t_4 = __pyx_t_5; } else { __pyx_t_4 = __pyx_t_3; } __pyx_t_3 = __pyx_t_4; } else { __pyx_t_3 = __pyx_t_2; } __pyx_t_2 = __pyx_t_3; } else { __pyx_t_2 = __pyx_t_1; } if (__pyx_t_2) { /* "/home/whyking/working/projects/hddm/src/pdf.pxi":106 * # Check if parpameters are valid * if z<0 or z>1 or a<0 or ((fabs(x)-(t-T/2.))<0) or (z+Z/2.>1) or (z-Z/2.<0) or (t-T/2.<0) or (t<0): * return 0 # <<<<<<<<<<<<<< * * # transform x,v,z if x is upper bound response */ __pyx_r = 0.0; goto __pyx_L0; goto __pyx_L3; } __pyx_L3:; /* "/home/whyking/working/projects/hddm/src/pdf.pxi":109 * * # transform x,v,z if x is upper bound response * if x > 0: # <<<<<<<<<<<<<< * v = -v * z = 1.-z */ __pyx_t_2 = (__pyx_v_x > 0.0); if (__pyx_t_2) { /* "/home/whyking/working/projects/hddm/src/pdf.pxi":110 * # transform x,v,z if x is upper bound response * if x > 0: * v = -v # <<<<<<<<<<<<<< * z = 1.-z * */ __pyx_v_v = (-__pyx_v_v); /* "/home/whyking/working/projects/hddm/src/pdf.pxi":111 * if x > 0: * v = -v * z = 1.-z # <<<<<<<<<<<<<< * * x = fabs(x) */ __pyx_v_z = (1. - __pyx_v_z); goto __pyx_L4; } __pyx_L4:; /* "/home/whyking/working/projects/hddm/src/pdf.pxi":113 * z = 1.-z * * x = fabs(x) # <<<<<<<<<<<<<< * * if T<1e-3: */ __pyx_v_x = fabs(__pyx_v_x); /* "/home/whyking/working/projects/hddm/src/pdf.pxi":115 * x = fabs(x) * * if T<1e-3: # <<<<<<<<<<<<<< * T = 0 * if Z <1e-3: */ __pyx_t_2 = (__pyx_v_T < 1e-3); if (__pyx_t_2) { /* "/home/whyking/working/projects/hddm/src/pdf.pxi":116 * * if T<1e-3: * T = 0 # <<<<<<<<<<<<<< * if Z <1e-3: * Z = 0 */ __pyx_v_T = 0.0; goto __pyx_L5; } __pyx_L5:; /* "/home/whyking/working/projects/hddm/src/pdf.pxi":117 * if T<1e-3: * T = 0 * if Z <1e-3: # <<<<<<<<<<<<<< * Z = 0 * */ __pyx_t_2 = (__pyx_v_Z < 1e-3); if (__pyx_t_2) { /* "/home/whyking/working/projects/hddm/src/pdf.pxi":118 * T = 0 * if Z <1e-3: * Z = 0 # <<<<<<<<<<<<<< * * if (Z==0): */ __pyx_v_Z = 0.0; goto __pyx_L6; } __pyx_L6:; /* "/home/whyking/working/projects/hddm/src/pdf.pxi":120 * Z = 0 * * if (Z==0): # <<<<<<<<<<<<<< * if (T==0): #V=0,Z=0,T=0 * return pdf_V(x - t, v, V, a, z, err) */ __pyx_t_2 = (__pyx_v_Z == 0.0); if (__pyx_t_2) { /* "/home/whyking/working/projects/hddm/src/pdf.pxi":121 * * if (Z==0): * if (T==0): #V=0,Z=0,T=0 # <<<<<<<<<<<<<< * return pdf_V(x - t, v, V, a, z, err) * else: #V=0,Z=0,T=$ */ __pyx_t_2 = (__pyx_v_T == 0.0); if (__pyx_t_2) { /* "/home/whyking/working/projects/hddm/src/pdf.pxi":122 * if (Z==0): * if (T==0): #V=0,Z=0,T=0 * return pdf_V(x - t, v, V, a, z, err) # <<<<<<<<<<<<<< * else: #V=0,Z=0,T=$ * if use_adaptive>0: */ __pyx_r = __pyx_f_4wfpt_pdf_V((__pyx_v_x - __pyx_v_t), __pyx_v_v, __pyx_v_V, __pyx_v_a, __pyx_v_z, __pyx_v_err); goto __pyx_L0; goto __pyx_L8; } /*else*/ { /* "/home/whyking/working/projects/hddm/src/pdf.pxi":124 * return pdf_V(x - t, v, V, a, z, err) * else: #V=0,Z=0,T=$ * if use_adaptive>0: # <<<<<<<<<<<<<< * return adaptiveSimpsons_1D(x, v, V, a, z, t, err, z, z, t-T/2., t+T/2., simps_err, nT) * else: */ __pyx_t_2 = (__pyx_v_use_adaptive > 0); if (__pyx_t_2) { /* "/home/whyking/working/projects/hddm/src/pdf.pxi":125 * else: #V=0,Z=0,T=$ * if use_adaptive>0: * return adaptiveSimpsons_1D(x, v, V, a, z, t, err, z, z, t-T/2., t+T/2., simps_err, nT) # <<<<<<<<<<<<<< * else: * return simpson_1D(x, v, V, a, z, t, err, z, z, 0, t-T/2., t+T/2., nT) */ __pyx_r = __pyx_f_4wfpt_adaptiveSimpsons_1D(__pyx_v_x, __pyx_v_v, __pyx_v_V, __pyx_v_a, __pyx_v_z, __pyx_v_t, __pyx_v_err, __pyx_v_z, __pyx_v_z, (__pyx_v_t - (__pyx_v_T / 2.)), (__pyx_v_t + (__pyx_v_T / 2.)), __pyx_v_simps_err, __pyx_v_nT); goto __pyx_L0; goto __pyx_L9; } /*else*/ { /* "/home/whyking/working/projects/hddm/src/pdf.pxi":127 * return adaptiveSimpsons_1D(x, v, V, a, z, t, err, z, z, t-T/2., t+T/2., simps_err, nT) * else: * return simpson_1D(x, v, V, a, z, t, err, z, z, 0, t-T/2., t+T/2., nT) # <<<<<<<<<<<<<< * * else: #Z=$ */ __pyx_r = __pyx_f_4wfpt_simpson_1D(__pyx_v_x, __pyx_v_v, __pyx_v_V, __pyx_v_a, __pyx_v_z, __pyx_v_t, __pyx_v_err, __pyx_v_z, __pyx_v_z, 0, (__pyx_v_t - (__pyx_v_T / 2.)), (__pyx_v_t + (__pyx_v_T / 2.)), __pyx_v_nT); goto __pyx_L0; } __pyx_L9:; } __pyx_L8:; goto __pyx_L7; } /*else*/ { /* "/home/whyking/working/projects/hddm/src/pdf.pxi":130 * * else: #Z=$ * if (T==0): #V=0,Z=$,T=0 # <<<<<<<<<<<<<< * if use_adaptive: * return adaptiveSimpsons_1D(x, v, V, a, z, t, err, z-Z/2., z+Z/2., t, t, simps_err, nZ) */ __pyx_t_2 = (__pyx_v_T == 0.0); if (__pyx_t_2) { /* "/home/whyking/working/projects/hddm/src/pdf.pxi":131 * else: #Z=$ * if (T==0): #V=0,Z=$,T=0 * if use_adaptive: # <<<<<<<<<<<<<< * return adaptiveSimpsons_1D(x, v, V, a, z, t, err, z-Z/2., z+Z/2., t, t, simps_err, nZ) * else: */ if (__pyx_v_use_adaptive) { /* "/home/whyking/working/projects/hddm/src/pdf.pxi":132 * if (T==0): #V=0,Z=$,T=0 * if use_adaptive: * return adaptiveSimpsons_1D(x, v, V, a, z, t, err, z-Z/2., z+Z/2., t, t, simps_err, nZ) # <<<<<<<<<<<<<< * else: * return simpson_1D(x, v, V, a, z, t, err, z-Z/2., z+Z/2., nZ, t, t , 0) */ __pyx_r = __pyx_f_4wfpt_adaptiveSimpsons_1D(__pyx_v_x, __pyx_v_v, __pyx_v_V, __pyx_v_a, __pyx_v_z, __pyx_v_t, __pyx_v_err, (__pyx_v_z - (__pyx_v_Z / 2.)), (__pyx_v_z + (__pyx_v_Z / 2.)), __pyx_v_t, __pyx_v_t, __pyx_v_simps_err, __pyx_v_nZ); goto __pyx_L0; goto __pyx_L11; } /*else*/ { /* "/home/whyking/working/projects/hddm/src/pdf.pxi":134 * return adaptiveSimpsons_1D(x, v, V, a, z, t, err, z-Z/2., z+Z/2., t, t, simps_err, nZ) * else: * return simpson_1D(x, v, V, a, z, t, err, z-Z/2., z+Z/2., nZ, t, t , 0) # <<<<<<<<<<<<<< * else: #V=0,Z=$,T=$ * if use_adaptive: */ __pyx_r = __pyx_f_4wfpt_simpson_1D(__pyx_v_x, __pyx_v_v, __pyx_v_V, __pyx_v_a, __pyx_v_z, __pyx_v_t, __pyx_v_err, (__pyx_v_z - (__pyx_v_Z / 2.)), (__pyx_v_z + (__pyx_v_Z / 2.)), __pyx_v_nZ, __pyx_v_t, __pyx_v_t, 0); goto __pyx_L0; } __pyx_L11:; goto __pyx_L10; } /*else*/ { /* "/home/whyking/working/projects/hddm/src/pdf.pxi":136 * return simpson_1D(x, v, V, a, z, t, err, z-Z/2., z+Z/2., nZ, t, t , 0) * else: #V=0,Z=$,T=$ * if use_adaptive: # <<<<<<<<<<<<<< * return adaptiveSimpsons_2D(x, v, V, a, z, t, err, z-Z/2., z+Z/2., t-T/2., t+T/2., simps_err, nZ, nT) * else: */ if (__pyx_v_use_adaptive) { /* "/home/whyking/working/projects/hddm/src/pdf.pxi":137 * else: #V=0,Z=$,T=$ * if use_adaptive: * return adaptiveSimpsons_2D(x, v, V, a, z, t, err, z-Z/2., z+Z/2., t-T/2., t+T/2., simps_err, nZ, nT) # <<<<<<<<<<<<<< * else: * return simpson_2D(x, v, V, a, z, t, err, z-Z/2., z+Z/2., nZ, t-T/2., t+T/2., nT) */ __pyx_r = __pyx_f_4wfpt_adaptiveSimpsons_2D(__pyx_v_x, __pyx_v_v, __pyx_v_V, __pyx_v_a, __pyx_v_z, __pyx_v_t, __pyx_v_err, (__pyx_v_z - (__pyx_v_Z / 2.)), (__pyx_v_z + (__pyx_v_Z / 2.)), (__pyx_v_t - (__pyx_v_T / 2.)), (__pyx_v_t + (__pyx_v_T / 2.)), __pyx_v_simps_err, __pyx_v_nZ, __pyx_v_nT); goto __pyx_L0; goto __pyx_L12; } /*else*/ { /* "/home/whyking/working/projects/hddm/src/pdf.pxi":139 * return adaptiveSimpsons_2D(x, v, V, a, z, t, err, z-Z/2., z+Z/2., t-T/2., t+T/2., simps_err, nZ, nT) * else: * return simpson_2D(x, v, V, a, z, t, err, z-Z/2., z+Z/2., nZ, t-T/2., t+T/2., nT) # <<<<<<<<<<<<<< */ __pyx_r = __pyx_f_4wfpt_simpson_2D(__pyx_v_x, __pyx_v_v, __pyx_v_V, __pyx_v_a, __pyx_v_z, __pyx_v_t, __pyx_v_err, (__pyx_v_z - (__pyx_v_Z / 2.)), (__pyx_v_z + (__pyx_v_Z / 2.)), __pyx_v_nZ, (__pyx_v_t - (__pyx_v_T / 2.)), (__pyx_v_t + (__pyx_v_T / 2.)), __pyx_v_nT); goto __pyx_L0; } __pyx_L12:; } __pyx_L10:; } __pyx_L7:; __pyx_r = 0; __pyx_L0:; return __pyx_r; } /* Python wrapper */ static PyObject *__pyx_pw_4wfpt_1full_pdf(PyObject *__pyx_self, PyObject *__pyx_args, PyObject *__pyx_kwds); /*proto*/ static char __pyx_doc_4wfpt_full_pdf[] = "full_pdf(double x, double v, double V, double a, double z, double Z, double t, double T, double err, int nT=2, int nZ=2, int use_adaptive=1, double simps_err=0.001) -> double\nfull pdf"; static PyObject *__pyx_pw_4wfpt_1full_pdf(PyObject *__pyx_self, PyObject *__pyx_args, PyObject *__pyx_kwds) { double __pyx_v_x; double __pyx_v_v; double __pyx_v_V; double __pyx_v_a; double __pyx_v_z; double __pyx_v_Z; double __pyx_v_t; double __pyx_v_T; double __pyx_v_err; int __pyx_v_nT; int __pyx_v_nZ; int __pyx_v_use_adaptive; double __pyx_v_simps_err; static PyObject **__pyx_pyargnames[] = {&__pyx_n_s__x,&__pyx_n_s__v,&__pyx_n_s__V,&__pyx_n_s__a,&__pyx_n_s__z,&__pyx_n_s__Z,&__pyx_n_s__t,&__pyx_n_s__T,&__pyx_n_s__err,&__pyx_n_s__nT,&__pyx_n_s__nZ,&__pyx_n_s__use_adaptive,&__pyx_n_s__simps_err,0}; PyObject *__pyx_r = 0; __Pyx_RefNannyDeclarations __Pyx_RefNannySetupContext("full_pdf (wrapper)"); __pyx_self = __pyx_self; { PyObject* values[13] = {0,0,0,0,0,0,0,0,0,0,0,0,0}; if (unlikely(__pyx_kwds)) { Py_ssize_t kw_args; const Py_ssize_t pos_args = PyTuple_GET_SIZE(__pyx_args); switch (pos_args) { case 13: values[12] = PyTuple_GET_ITEM(__pyx_args, 12); case 12: values[11] = PyTuple_GET_ITEM(__pyx_args, 11); case 11: values[10] = PyTuple_GET_ITEM(__pyx_args, 10); case 10: values[9] = PyTuple_GET_ITEM(__pyx_args, 9); case 9: values[8] = PyTuple_GET_ITEM(__pyx_args, 8); case 8: values[7] = PyTuple_GET_ITEM(__pyx_args, 7); case 7: values[6] = PyTuple_GET_ITEM(__pyx_args, 6); case 6: values[5] = PyTuple_GET_ITEM(__pyx_args, 5); case 5: values[4] = PyTuple_GET_ITEM(__pyx_args, 4); case 4: values[3] = PyTuple_GET_ITEM(__pyx_args, 3); case 3: values[2] = PyTuple_GET_ITEM(__pyx_args, 2); case 2: values[1] = PyTuple_GET_ITEM(__pyx_args, 1); case 1: values[0] = PyTuple_GET_ITEM(__pyx_args, 0); case 0: break; default: goto __pyx_L5_argtuple_error; } kw_args = PyDict_Size(__pyx_kwds); switch (pos_args) { case 0: values[0] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__x); if (likely(values[0])) kw_args--; else goto __pyx_L5_argtuple_error; case 1: values[1] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__v); if (likely(values[1])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("full_pdf", 0, 9, 13, 1); {__pyx_filename = __pyx_f[1]; __pyx_lineno = 98; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 2: values[2] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__V); if (likely(values[2])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("full_pdf", 0, 9, 13, 2); {__pyx_filename = __pyx_f[1]; __pyx_lineno = 98; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 3: values[3] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__a); if (likely(values[3])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("full_pdf", 0, 9, 13, 3); {__pyx_filename = __pyx_f[1]; __pyx_lineno = 98; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 4: values[4] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__z); if (likely(values[4])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("full_pdf", 0, 9, 13, 4); {__pyx_filename = __pyx_f[1]; __pyx_lineno = 98; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 5: values[5] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__Z); if (likely(values[5])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("full_pdf", 0, 9, 13, 5); {__pyx_filename = __pyx_f[1]; __pyx_lineno = 98; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 6: values[6] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__t); if (likely(values[6])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("full_pdf", 0, 9, 13, 6); {__pyx_filename = __pyx_f[1]; __pyx_lineno = 98; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 7: values[7] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__T); if (likely(values[7])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("full_pdf", 0, 9, 13, 7); {__pyx_filename = __pyx_f[1]; __pyx_lineno = 98; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 8: values[8] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__err); if (likely(values[8])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("full_pdf", 0, 9, 13, 8); {__pyx_filename = __pyx_f[1]; __pyx_lineno = 98; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 9: if (kw_args > 0) { PyObject* value = PyDict_GetItem(__pyx_kwds, __pyx_n_s__nT); if (value) { values[9] = value; kw_args--; } } case 10: if (kw_args > 0) { PyObject* value = PyDict_GetItem(__pyx_kwds, __pyx_n_s__nZ); if (value) { values[10] = value; kw_args--; } } case 11: if (kw_args > 0) { PyObject* value = PyDict_GetItem(__pyx_kwds, __pyx_n_s__use_adaptive); if (value) { values[11] = value; kw_args--; } } case 12: if (kw_args > 0) { PyObject* value = PyDict_GetItem(__pyx_kwds, __pyx_n_s__simps_err); if (value) { values[12] = value; kw_args--; } } } if (unlikely(kw_args > 0)) { if (unlikely(__Pyx_ParseOptionalKeywords(__pyx_kwds, __pyx_pyargnames, 0, values, pos_args, "full_pdf") < 0)) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 98; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } __pyx_v_x = __pyx_PyFloat_AsDouble(values[0]); if (unlikely((__pyx_v_x == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 98; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_v = __pyx_PyFloat_AsDouble(values[1]); if (unlikely((__pyx_v_v == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 98; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_V = __pyx_PyFloat_AsDouble(values[2]); if (unlikely((__pyx_v_V == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 98; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_a = __pyx_PyFloat_AsDouble(values[3]); if (unlikely((__pyx_v_a == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 98; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_z = __pyx_PyFloat_AsDouble(values[4]); if (unlikely((__pyx_v_z == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 98; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_Z = __pyx_PyFloat_AsDouble(values[5]); if (unlikely((__pyx_v_Z == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 99; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_t = __pyx_PyFloat_AsDouble(values[6]); if (unlikely((__pyx_v_t == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 99; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_T = __pyx_PyFloat_AsDouble(values[7]); if (unlikely((__pyx_v_T == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 99; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_err = __pyx_PyFloat_AsDouble(values[8]); if (unlikely((__pyx_v_err == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 99; __pyx_clineno = __LINE__; goto __pyx_L3_error;} if (values[9]) { __pyx_v_nT = __Pyx_PyInt_AsInt(values[9]); if (unlikely((__pyx_v_nT == (int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 99; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } else { __pyx_v_nT = ((int)2); } if (values[10]) { __pyx_v_nZ = __Pyx_PyInt_AsInt(values[10]); if (unlikely((__pyx_v_nZ == (int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 100; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } else { __pyx_v_nZ = ((int)2); } if (values[11]) { __pyx_v_use_adaptive = __Pyx_PyObject_IsTrue(values[11]); if (unlikely((__pyx_v_use_adaptive == (int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 100; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } else { __pyx_v_use_adaptive = ((int)1); } if (values[12]) { __pyx_v_simps_err = __pyx_PyFloat_AsDouble(values[12]); if (unlikely((__pyx_v_simps_err == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 100; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } else { /* "/home/whyking/working/projects/hddm/src/pdf.pxi":101 * z, double Z, double t, double T, double err, int * nT=2, int nZ=2, bint use_adaptive=1, double * simps_err=1e-3) nogil: # <<<<<<<<<<<<<< * """full pdf""" * */ __pyx_v_simps_err = ((double)1e-3); } } else { switch (PyTuple_GET_SIZE(__pyx_args)) { case 13: values[12] = PyTuple_GET_ITEM(__pyx_args, 12); case 12: values[11] = PyTuple_GET_ITEM(__pyx_args, 11); case 11: values[10] = PyTuple_GET_ITEM(__pyx_args, 10); case 10: values[9] = PyTuple_GET_ITEM(__pyx_args, 9); case 9: values[8] = PyTuple_GET_ITEM(__pyx_args, 8); values[7] = PyTuple_GET_ITEM(__pyx_args, 7); values[6] = PyTuple_GET_ITEM(__pyx_args, 6); values[5] = PyTuple_GET_ITEM(__pyx_args, 5); values[4] = PyTuple_GET_ITEM(__pyx_args, 4); values[3] = PyTuple_GET_ITEM(__pyx_args, 3); values[2] = PyTuple_GET_ITEM(__pyx_args, 2); values[1] = PyTuple_GET_ITEM(__pyx_args, 1); values[0] = PyTuple_GET_ITEM(__pyx_args, 0); break; default: goto __pyx_L5_argtuple_error; } } __pyx_v_x = __pyx_PyFloat_AsDouble(values[0]); if (unlikely((__pyx_v_x == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 98; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_v = __pyx_PyFloat_AsDouble(values[1]); if (unlikely((__pyx_v_v == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 98; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_V = __pyx_PyFloat_AsDouble(values[2]); if (unlikely((__pyx_v_V == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 98; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_a = __pyx_PyFloat_AsDouble(values[3]); if (unlikely((__pyx_v_a == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 98; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_z = __pyx_PyFloat_AsDouble(values[4]); if (unlikely((__pyx_v_z == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 98; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_Z = __pyx_PyFloat_AsDouble(values[5]); if (unlikely((__pyx_v_Z == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 99; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_t = __pyx_PyFloat_AsDouble(values[6]); if (unlikely((__pyx_v_t == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 99; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_T = __pyx_PyFloat_AsDouble(values[7]); if (unlikely((__pyx_v_T == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 99; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_err = __pyx_PyFloat_AsDouble(values[8]); if (unlikely((__pyx_v_err == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 99; __pyx_clineno = __LINE__; goto __pyx_L3_error;} if (values[9]) { __pyx_v_nT = __Pyx_PyInt_AsInt(values[9]); if (unlikely((__pyx_v_nT == (int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 99; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } else { __pyx_v_nT = ((int)2); } if (values[10]) { __pyx_v_nZ = __Pyx_PyInt_AsInt(values[10]); if (unlikely((__pyx_v_nZ == (int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 100; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } else { __pyx_v_nZ = ((int)2); } if (values[11]) { __pyx_v_use_adaptive = __Pyx_PyObject_IsTrue(values[11]); if (unlikely((__pyx_v_use_adaptive == (int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 100; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } else { __pyx_v_use_adaptive = ((int)1); } if (values[12]) { __pyx_v_simps_err = __pyx_PyFloat_AsDouble(values[12]); if (unlikely((__pyx_v_simps_err == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 100; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } else { __pyx_v_simps_err = ((double)1e-3); } } goto __pyx_L4_argument_unpacking_done; __pyx_L5_argtuple_error:; __Pyx_RaiseArgtupleInvalid("full_pdf", 0, 9, 13, PyTuple_GET_SIZE(__pyx_args)); {__pyx_filename = __pyx_f[1]; __pyx_lineno = 98; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_L3_error:; __Pyx_AddTraceback("wfpt.full_pdf", __pyx_clineno, __pyx_lineno, __pyx_filename); __Pyx_RefNannyFinishContext(); return NULL; __pyx_L4_argument_unpacking_done:; __pyx_r = __pyx_pf_4wfpt_full_pdf(__pyx_self, __pyx_v_x, __pyx_v_v, __pyx_v_V, __pyx_v_a, __pyx_v_z, __pyx_v_Z, __pyx_v_t, __pyx_v_T, __pyx_v_err, __pyx_v_nT, __pyx_v_nZ, __pyx_v_use_adaptive, __pyx_v_simps_err); __Pyx_RefNannyFinishContext(); return __pyx_r; } /* "/home/whyking/working/projects/hddm/src/pdf.pxi":98 * return exp(log(p) + ((a*z*V)**2 - 2*a*v*z - (v**2)*x)/(2*(V**2)*x+2))/sqrt((V**2)*x+1)/(a**2) * * cpdef double full_pdf(double x, double v, double V, double a, double # <<<<<<<<<<<<<< * z, double Z, double t, double T, double err, int * nT=2, int nZ=2, bint use_adaptive=1, double */ static PyObject *__pyx_pf_4wfpt_full_pdf(CYTHON_UNUSED PyObject *__pyx_self, double __pyx_v_x, double __pyx_v_v, double __pyx_v_V, double __pyx_v_a, double __pyx_v_z, double __pyx_v_Z, double __pyx_v_t, double __pyx_v_T, double __pyx_v_err, int __pyx_v_nT, int __pyx_v_nZ, int __pyx_v_use_adaptive, double __pyx_v_simps_err) { PyObject *__pyx_r = NULL; __Pyx_RefNannyDeclarations double __pyx_t_1; struct __pyx_opt_args_4wfpt_full_pdf __pyx_t_2; PyObject *__pyx_t_3 = NULL; int __pyx_lineno = 0; const char *__pyx_filename = NULL; int __pyx_clineno = 0; __Pyx_RefNannySetupContext("full_pdf"); __Pyx_XDECREF(__pyx_r); __pyx_t_2.__pyx_n = 4; __pyx_t_2.nT = __pyx_v_nT; __pyx_t_2.nZ = __pyx_v_nZ; __pyx_t_2.use_adaptive = __pyx_v_use_adaptive; __pyx_t_2.simps_err = __pyx_v_simps_err; __pyx_t_1 = __pyx_f_4wfpt_full_pdf(__pyx_v_x, __pyx_v_v, __pyx_v_V, __pyx_v_a, __pyx_v_z, __pyx_v_Z, __pyx_v_t, __pyx_v_T, __pyx_v_err, 0, &__pyx_t_2); __pyx_t_3 = PyFloat_FromDouble(__pyx_t_1); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 98; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); __pyx_r = __pyx_t_3; __pyx_t_3 = 0; goto __pyx_L0; __pyx_r = Py_None; __Pyx_INCREF(Py_None); goto __pyx_L0; __pyx_L1_error:; __Pyx_XDECREF(__pyx_t_3); __Pyx_AddTraceback("wfpt.full_pdf", __pyx_clineno, __pyx_lineno, __pyx_filename); __pyx_r = NULL; __pyx_L0:; __Pyx_XGIVEREF(__pyx_r); __Pyx_RefNannyFinishContext(); return __pyx_r; } /* Python wrapper */ static PyObject *__pyx_pw_4wfpt_3pdf_array(PyObject *__pyx_self, PyObject *__pyx_args, PyObject *__pyx_kwds); /*proto*/ static char __pyx_doc_4wfpt_2pdf_array[] = "pdf_array(ndarray x, double v, double V, double a, double z, double Z, double t, double T, double err, int logp=0, int nT=2, int nZ=2, int use_adaptive=1, double simps_err=0.001)"; static PyMethodDef __pyx_mdef_4wfpt_3pdf_array = {__Pyx_NAMESTR("pdf_array"), (PyCFunction)__pyx_pw_4wfpt_3pdf_array, METH_VARARGS|METH_KEYWORDS, __Pyx_DOCSTR(__pyx_doc_4wfpt_2pdf_array)}; static PyObject *__pyx_pw_4wfpt_3pdf_array(PyObject *__pyx_self, PyObject *__pyx_args, PyObject *__pyx_kwds) { PyArrayObject *__pyx_v_x = 0; double __pyx_v_v; double __pyx_v_V; double __pyx_v_a; double __pyx_v_z; double __pyx_v_Z; double __pyx_v_t; double __pyx_v_T; double __pyx_v_err; int __pyx_v_logp; int __pyx_v_nT; int __pyx_v_nZ; int __pyx_v_use_adaptive; double __pyx_v_simps_err; static PyObject **__pyx_pyargnames[] = {&__pyx_n_s__x,&__pyx_n_s__v,&__pyx_n_s__V,&__pyx_n_s__a,&__pyx_n_s__z,&__pyx_n_s__Z,&__pyx_n_s__t,&__pyx_n_s__T,&__pyx_n_s__err,&__pyx_n_s__logp,&__pyx_n_s__nT,&__pyx_n_s__nZ,&__pyx_n_s__use_adaptive,&__pyx_n_s__simps_err,0}; PyObject *__pyx_r = 0; __Pyx_RefNannyDeclarations __Pyx_RefNannySetupContext("pdf_array (wrapper)"); __pyx_self = __pyx_self; { PyObject* values[14] = {0,0,0,0,0,0,0,0,0,0,0,0,0,0}; if (unlikely(__pyx_kwds)) { Py_ssize_t kw_args; const Py_ssize_t pos_args = PyTuple_GET_SIZE(__pyx_args); switch (pos_args) { case 14: values[13] = PyTuple_GET_ITEM(__pyx_args, 13); case 13: values[12] = PyTuple_GET_ITEM(__pyx_args, 12); case 12: values[11] = PyTuple_GET_ITEM(__pyx_args, 11); case 11: values[10] = PyTuple_GET_ITEM(__pyx_args, 10); case 10: values[9] = PyTuple_GET_ITEM(__pyx_args, 9); case 9: values[8] = PyTuple_GET_ITEM(__pyx_args, 8); case 8: values[7] = PyTuple_GET_ITEM(__pyx_args, 7); case 7: values[6] = PyTuple_GET_ITEM(__pyx_args, 6); case 6: values[5] = PyTuple_GET_ITEM(__pyx_args, 5); case 5: values[4] = PyTuple_GET_ITEM(__pyx_args, 4); case 4: values[3] = PyTuple_GET_ITEM(__pyx_args, 3); case 3: values[2] = PyTuple_GET_ITEM(__pyx_args, 2); case 2: values[1] = PyTuple_GET_ITEM(__pyx_args, 1); case 1: values[0] = PyTuple_GET_ITEM(__pyx_args, 0); case 0: break; default: goto __pyx_L5_argtuple_error; } kw_args = PyDict_Size(__pyx_kwds); switch (pos_args) { case 0: values[0] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__x); if (likely(values[0])) kw_args--; else goto __pyx_L5_argtuple_error; case 1: values[1] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__v); if (likely(values[1])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("pdf_array", 0, 9, 14, 1); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 28; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 2: values[2] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__V); if (likely(values[2])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("pdf_array", 0, 9, 14, 2); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 28; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 3: values[3] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__a); if (likely(values[3])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("pdf_array", 0, 9, 14, 3); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 28; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 4: values[4] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__z); if (likely(values[4])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("pdf_array", 0, 9, 14, 4); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 28; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 5: values[5] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__Z); if (likely(values[5])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("pdf_array", 0, 9, 14, 5); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 28; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 6: values[6] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__t); if (likely(values[6])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("pdf_array", 0, 9, 14, 6); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 28; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 7: values[7] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__T); if (likely(values[7])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("pdf_array", 0, 9, 14, 7); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 28; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 8: values[8] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__err); if (likely(values[8])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("pdf_array", 0, 9, 14, 8); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 28; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 9: if (kw_args > 0) { PyObject* value = PyDict_GetItem(__pyx_kwds, __pyx_n_s__logp); if (value) { values[9] = value; kw_args--; } } case 10: if (kw_args > 0) { PyObject* value = PyDict_GetItem(__pyx_kwds, __pyx_n_s__nT); if (value) { values[10] = value; kw_args--; } } case 11: if (kw_args > 0) { PyObject* value = PyDict_GetItem(__pyx_kwds, __pyx_n_s__nZ); if (value) { values[11] = value; kw_args--; } } case 12: if (kw_args > 0) { PyObject* value = PyDict_GetItem(__pyx_kwds, __pyx_n_s__use_adaptive); if (value) { values[12] = value; kw_args--; } } case 13: if (kw_args > 0) { PyObject* value = PyDict_GetItem(__pyx_kwds, __pyx_n_s__simps_err); if (value) { values[13] = value; kw_args--; } } } if (unlikely(kw_args > 0)) { if (unlikely(__Pyx_ParseOptionalKeywords(__pyx_kwds, __pyx_pyargnames, 0, values, pos_args, "pdf_array") < 0)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 28; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } __pyx_v_x = ((PyArrayObject *)values[0]); __pyx_v_v = __pyx_PyFloat_AsDouble(values[1]); if (unlikely((__pyx_v_v == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 28; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_V = __pyx_PyFloat_AsDouble(values[2]); if (unlikely((__pyx_v_V == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 28; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_a = __pyx_PyFloat_AsDouble(values[3]); if (unlikely((__pyx_v_a == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 28; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_z = __pyx_PyFloat_AsDouble(values[4]); if (unlikely((__pyx_v_z == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 28; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_Z = __pyx_PyFloat_AsDouble(values[5]); if (unlikely((__pyx_v_Z == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 28; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_t = __pyx_PyFloat_AsDouble(values[6]); if (unlikely((__pyx_v_t == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 28; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_T = __pyx_PyFloat_AsDouble(values[7]); if (unlikely((__pyx_v_T == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 28; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_err = __pyx_PyFloat_AsDouble(values[8]); if (unlikely((__pyx_v_err == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 28; __pyx_clineno = __LINE__; goto __pyx_L3_error;} if (values[9]) { __pyx_v_logp = __Pyx_PyObject_IsTrue(values[9]); if (unlikely((__pyx_v_logp == (int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 28; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } else { __pyx_v_logp = ((int)0); } if (values[10]) { __pyx_v_nT = __Pyx_PyInt_AsInt(values[10]); if (unlikely((__pyx_v_nT == (int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 28; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } else { __pyx_v_nT = ((int)2); } if (values[11]) { __pyx_v_nZ = __Pyx_PyInt_AsInt(values[11]); if (unlikely((__pyx_v_nZ == (int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 28; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } else { __pyx_v_nZ = ((int)2); } if (values[12]) { __pyx_v_use_adaptive = __Pyx_PyObject_IsTrue(values[12]); if (unlikely((__pyx_v_use_adaptive == (int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 28; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } else { __pyx_v_use_adaptive = ((int)1); } if (values[13]) { __pyx_v_simps_err = __pyx_PyFloat_AsDouble(values[13]); if (unlikely((__pyx_v_simps_err == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 28; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } else { /* "wfpt.pyx":28 * include "pdf.pxi" * * def pdf_array(np.ndarray[double, ndim=1] x, double v, double V, double a, double z, double Z, double t, double T, double err, bint logp=0, int nT=2, int nZ=2, bint use_adaptive=1, double simps_err=1e-3): # <<<<<<<<<<<<<< * cdef Py_ssize_t size = x.shape[0] * cdef Py_ssize_t i */ __pyx_v_simps_err = ((double)1e-3); } } else { switch (PyTuple_GET_SIZE(__pyx_args)) { case 14: values[13] = PyTuple_GET_ITEM(__pyx_args, 13); case 13: values[12] = PyTuple_GET_ITEM(__pyx_args, 12); case 12: values[11] = PyTuple_GET_ITEM(__pyx_args, 11); case 11: values[10] = PyTuple_GET_ITEM(__pyx_args, 10); case 10: values[9] = PyTuple_GET_ITEM(__pyx_args, 9); case 9: values[8] = PyTuple_GET_ITEM(__pyx_args, 8); values[7] = PyTuple_GET_ITEM(__pyx_args, 7); values[6] = PyTuple_GET_ITEM(__pyx_args, 6); values[5] = PyTuple_GET_ITEM(__pyx_args, 5); values[4] = PyTuple_GET_ITEM(__pyx_args, 4); values[3] = PyTuple_GET_ITEM(__pyx_args, 3); values[2] = PyTuple_GET_ITEM(__pyx_args, 2); values[1] = PyTuple_GET_ITEM(__pyx_args, 1); values[0] = PyTuple_GET_ITEM(__pyx_args, 0); break; default: goto __pyx_L5_argtuple_error; } } __pyx_v_x = ((PyArrayObject *)values[0]); __pyx_v_v = __pyx_PyFloat_AsDouble(values[1]); if (unlikely((__pyx_v_v == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 28; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_V = __pyx_PyFloat_AsDouble(values[2]); if (unlikely((__pyx_v_V == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 28; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_a = __pyx_PyFloat_AsDouble(values[3]); if (unlikely((__pyx_v_a == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 28; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_z = __pyx_PyFloat_AsDouble(values[4]); if (unlikely((__pyx_v_z == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 28; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_Z = __pyx_PyFloat_AsDouble(values[5]); if (unlikely((__pyx_v_Z == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 28; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_t = __pyx_PyFloat_AsDouble(values[6]); if (unlikely((__pyx_v_t == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 28; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_T = __pyx_PyFloat_AsDouble(values[7]); if (unlikely((__pyx_v_T == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 28; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_err = __pyx_PyFloat_AsDouble(values[8]); if (unlikely((__pyx_v_err == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 28; __pyx_clineno = __LINE__; goto __pyx_L3_error;} if (values[9]) { __pyx_v_logp = __Pyx_PyObject_IsTrue(values[9]); if (unlikely((__pyx_v_logp == (int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 28; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } else { __pyx_v_logp = ((int)0); } if (values[10]) { __pyx_v_nT = __Pyx_PyInt_AsInt(values[10]); if (unlikely((__pyx_v_nT == (int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 28; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } else { __pyx_v_nT = ((int)2); } if (values[11]) { __pyx_v_nZ = __Pyx_PyInt_AsInt(values[11]); if (unlikely((__pyx_v_nZ == (int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 28; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } else { __pyx_v_nZ = ((int)2); } if (values[12]) { __pyx_v_use_adaptive = __Pyx_PyObject_IsTrue(values[12]); if (unlikely((__pyx_v_use_adaptive == (int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 28; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } else { __pyx_v_use_adaptive = ((int)1); } if (values[13]) { __pyx_v_simps_err = __pyx_PyFloat_AsDouble(values[13]); if (unlikely((__pyx_v_simps_err == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 28; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } else { __pyx_v_simps_err = ((double)1e-3); } } goto __pyx_L4_argument_unpacking_done; __pyx_L5_argtuple_error:; __Pyx_RaiseArgtupleInvalid("pdf_array", 0, 9, 14, PyTuple_GET_SIZE(__pyx_args)); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 28; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_L3_error:; __Pyx_AddTraceback("wfpt.pdf_array", __pyx_clineno, __pyx_lineno, __pyx_filename); __Pyx_RefNannyFinishContext(); return NULL; __pyx_L4_argument_unpacking_done:; if (unlikely(!__Pyx_ArgTypeTest(((PyObject *)__pyx_v_x), __pyx_ptype_5numpy_ndarray, 1, "x", 0))) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 28; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __pyx_r = __pyx_pf_4wfpt_2pdf_array(__pyx_self, __pyx_v_x, __pyx_v_v, __pyx_v_V, __pyx_v_a, __pyx_v_z, __pyx_v_Z, __pyx_v_t, __pyx_v_T, __pyx_v_err, __pyx_v_logp, __pyx_v_nT, __pyx_v_nZ, __pyx_v_use_adaptive, __pyx_v_simps_err); goto __pyx_L0; __pyx_L1_error:; __pyx_L0:; __Pyx_RefNannyFinishContext(); return __pyx_r; } static PyObject *__pyx_pf_4wfpt_2pdf_array(CYTHON_UNUSED PyObject *__pyx_self, PyArrayObject *__pyx_v_x, double __pyx_v_v, double __pyx_v_V, double __pyx_v_a, double __pyx_v_z, double __pyx_v_Z, double __pyx_v_t, double __pyx_v_T, double __pyx_v_err, int __pyx_v_logp, int __pyx_v_nT, int __pyx_v_nZ, int __pyx_v_use_adaptive, double __pyx_v_simps_err) { Py_ssize_t __pyx_v_size; Py_ssize_t __pyx_v_i; PyArrayObject *__pyx_v_y = 0; __Pyx_LocalBuf_ND __pyx_pybuffernd_y; __Pyx_Buffer __pyx_pybuffer_y; __Pyx_LocalBuf_ND __pyx_pybuffernd_x; __Pyx_Buffer __pyx_pybuffer_x; PyObject *__pyx_r = NULL; __Pyx_RefNannyDeclarations PyObject *__pyx_t_1 = NULL; PyObject *__pyx_t_2 = NULL; PyObject *__pyx_t_3 = NULL; PyObject *__pyx_t_4 = NULL; PyObject *__pyx_t_5 = NULL; PyArrayObject *__pyx_t_6 = NULL; Py_ssize_t __pyx_t_7; Py_ssize_t __pyx_t_8; Py_ssize_t __pyx_t_9; Py_ssize_t __pyx_t_10; double __pyx_t_11; struct __pyx_opt_args_4wfpt_full_pdf __pyx_t_12; Py_ssize_t __pyx_t_13; int __pyx_t_14; int __pyx_lineno = 0; const char *__pyx_filename = NULL; int __pyx_clineno = 0; __Pyx_RefNannySetupContext("pdf_array"); __pyx_pybuffer_y.pybuffer.buf = NULL; __pyx_pybuffer_y.refcount = 0; __pyx_pybuffernd_y.data = NULL; __pyx_pybuffernd_y.rcbuffer = &__pyx_pybuffer_y; __pyx_pybuffer_x.pybuffer.buf = NULL; __pyx_pybuffer_x.refcount = 0; __pyx_pybuffernd_x.data = NULL; __pyx_pybuffernd_x.rcbuffer = &__pyx_pybuffer_x; { __Pyx_BufFmt_StackElem __pyx_stack[1]; if (unlikely(__Pyx_GetBufferAndValidate(&__pyx_pybuffernd_x.rcbuffer->pybuffer, (PyObject*)__pyx_v_x, &__Pyx_TypeInfo_double, PyBUF_FORMAT| PyBUF_STRIDES, 1, 0, __pyx_stack) == -1)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 28; __pyx_clineno = __LINE__; goto __pyx_L1_error;} } __pyx_pybuffernd_x.diminfo[0].strides = __pyx_pybuffernd_x.rcbuffer->pybuffer.strides[0]; __pyx_pybuffernd_x.diminfo[0].shape = __pyx_pybuffernd_x.rcbuffer->pybuffer.shape[0]; /* "wfpt.pyx":29 * * def pdf_array(np.ndarray[double, ndim=1] x, double v, double V, double a, double z, double Z, double t, double T, double err, bint logp=0, int nT=2, int nZ=2, bint use_adaptive=1, double simps_err=1e-3): * cdef Py_ssize_t size = x.shape[0] # <<<<<<<<<<<<<< * cdef Py_ssize_t i * cdef np.ndarray[double, ndim=1] y = np.empty(size, dtype=np.double) */ __pyx_v_size = (__pyx_v_x->dimensions[0]); /* "wfpt.pyx":31 * cdef Py_ssize_t size = x.shape[0] * cdef Py_ssize_t i * cdef np.ndarray[double, ndim=1] y = np.empty(size, dtype=np.double) # <<<<<<<<<<<<<< * * for i in prange(size, nogil=True): */ __pyx_t_1 = __Pyx_GetName(__pyx_m, __pyx_n_s__np); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 31; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_1); __pyx_t_2 = PyObject_GetAttr(__pyx_t_1, __pyx_n_s__empty); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 31; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_2); __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; __pyx_t_1 = PyInt_FromSsize_t(__pyx_v_size); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 31; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_1); __pyx_t_3 = PyTuple_New(1); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 31; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); PyTuple_SET_ITEM(__pyx_t_3, 0, __pyx_t_1); __Pyx_GIVEREF(__pyx_t_1); __pyx_t_1 = 0; __pyx_t_1 = PyDict_New(); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 31; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(((PyObject *)__pyx_t_1)); __pyx_t_4 = __Pyx_GetName(__pyx_m, __pyx_n_s__np); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 31; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_4); __pyx_t_5 = PyObject_GetAttr(__pyx_t_4, __pyx_n_s__double); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 31; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_5); __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0; if (PyDict_SetItem(__pyx_t_1, ((PyObject *)__pyx_n_s__dtype), __pyx_t_5) < 0) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 31; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0; __pyx_t_5 = PyObject_Call(__pyx_t_2, ((PyObject *)__pyx_t_3), ((PyObject *)__pyx_t_1)); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 31; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_5); __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; __Pyx_DECREF(((PyObject *)__pyx_t_3)); __pyx_t_3 = 0; __Pyx_DECREF(((PyObject *)__pyx_t_1)); __pyx_t_1 = 0; if (!(likely(((__pyx_t_5) == Py_None) || likely(__Pyx_TypeTest(__pyx_t_5, __pyx_ptype_5numpy_ndarray))))) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 31; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __pyx_t_6 = ((PyArrayObject *)__pyx_t_5); { __Pyx_BufFmt_StackElem __pyx_stack[1]; if (unlikely(__Pyx_GetBufferAndValidate(&__pyx_pybuffernd_y.rcbuffer->pybuffer, (PyObject*)__pyx_t_6, &__Pyx_TypeInfo_double, PyBUF_FORMAT| PyBUF_STRIDES| PyBUF_WRITABLE, 1, 0, __pyx_stack) == -1)) { __pyx_v_y = ((PyArrayObject *)Py_None); __Pyx_INCREF(Py_None); __pyx_pybuffernd_y.rcbuffer->pybuffer.buf = NULL; {__pyx_filename = __pyx_f[0]; __pyx_lineno = 31; __pyx_clineno = __LINE__; goto __pyx_L1_error;} } else {__pyx_pybuffernd_y.diminfo[0].strides = __pyx_pybuffernd_y.rcbuffer->pybuffer.strides[0]; __pyx_pybuffernd_y.diminfo[0].shape = __pyx_pybuffernd_y.rcbuffer->pybuffer.shape[0]; } } __pyx_t_6 = 0; __pyx_v_y = ((PyArrayObject *)__pyx_t_5); __pyx_t_5 = 0; /* "wfpt.pyx":33 * cdef np.ndarray[double, ndim=1] y = np.empty(size, dtype=np.double) * * for i in prange(size, nogil=True): # <<<<<<<<<<<<<< * y[i] = full_pdf(x[i], v, V, a, z, Z, t, T, err, nT, nZ, use_adaptive, simps_err) * */ { #ifdef WITH_THREAD PyThreadState *_save = NULL; #endif Py_UNBLOCK_THREADS /*try:*/ { __pyx_t_7 = __pyx_v_size; if (1 == 0) abort(); { __pyx_t_9 = (__pyx_t_7 - 0) / 1; if (__pyx_t_9 > 0) { __pyx_v_i = 0; #ifdef _OPENMP #pragma omp parallel private(__pyx_t_10, __pyx_t_13, __pyx_t_11, __pyx_t_12) #endif /* _OPENMP */ { #ifdef _OPENMP #pragma omp for firstprivate(__pyx_v_i) lastprivate(__pyx_v_i) #endif /* _OPENMP */ for (__pyx_t_8 = 0; __pyx_t_8 < __pyx_t_9; __pyx_t_8++){ { __pyx_v_i = 0 + 1 * __pyx_t_8; /* "wfpt.pyx":34 * * for i in prange(size, nogil=True): * y[i] = full_pdf(x[i], v, V, a, z, Z, t, T, err, nT, nZ, use_adaptive, simps_err) # <<<<<<<<<<<<<< * * if logp==1: */ __pyx_t_10 = __pyx_v_i; __pyx_t_12.__pyx_n = 4; __pyx_t_12.nT = __pyx_v_nT; __pyx_t_12.nZ = __pyx_v_nZ; __pyx_t_12.use_adaptive = __pyx_v_use_adaptive; __pyx_t_12.simps_err = __pyx_v_simps_err; __pyx_t_11 = __pyx_f_4wfpt_full_pdf((*__Pyx_BufPtrStrided1d(double *, __pyx_pybuffernd_x.rcbuffer->pybuffer.buf, __pyx_t_10, __pyx_pybuffernd_x.diminfo[0].strides)), __pyx_v_v, __pyx_v_V, __pyx_v_a, __pyx_v_z, __pyx_v_Z, __pyx_v_t, __pyx_v_T, __pyx_v_err, 0, &__pyx_t_12); __pyx_t_13 = __pyx_v_i; *__Pyx_BufPtrStrided1d(double *, __pyx_pybuffernd_y.rcbuffer->pybuffer.buf, __pyx_t_13, __pyx_pybuffernd_y.diminfo[0].strides) = __pyx_t_11; } } } } } } /* "wfpt.pyx":33 * cdef np.ndarray[double, ndim=1] y = np.empty(size, dtype=np.double) * * for i in prange(size, nogil=True): # <<<<<<<<<<<<<< * y[i] = full_pdf(x[i], v, V, a, z, Z, t, T, err, nT, nZ, use_adaptive, simps_err) * */ /*finally:*/ { Py_BLOCK_THREADS } } /* "wfpt.pyx":36 * y[i] = full_pdf(x[i], v, V, a, z, Z, t, T, err, nT, nZ, use_adaptive, simps_err) * * if logp==1: # <<<<<<<<<<<<<< * return np.log(y) * else: */ __pyx_t_14 = (__pyx_v_logp == 1); if (__pyx_t_14) { /* "wfpt.pyx":37 * * if logp==1: * return np.log(y) # <<<<<<<<<<<<<< * else: * return y */ __Pyx_XDECREF(__pyx_r); __pyx_t_5 = __Pyx_GetName(__pyx_m, __pyx_n_s__np); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 37; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_5); __pyx_t_1 = PyObject_GetAttr(__pyx_t_5, __pyx_n_s__log); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 37; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_1); __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0; __pyx_t_5 = PyTuple_New(1); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 37; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_5); __Pyx_INCREF(((PyObject *)__pyx_v_y)); PyTuple_SET_ITEM(__pyx_t_5, 0, ((PyObject *)__pyx_v_y)); __Pyx_GIVEREF(((PyObject *)__pyx_v_y)); __pyx_t_3 = PyObject_Call(__pyx_t_1, ((PyObject *)__pyx_t_5), NULL); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 37; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; __Pyx_DECREF(((PyObject *)__pyx_t_5)); __pyx_t_5 = 0; __pyx_r = __pyx_t_3; __pyx_t_3 = 0; goto __pyx_L0; goto __pyx_L12; } /*else*/ { /* "wfpt.pyx":39 * return np.log(y) * else: * return y # <<<<<<<<<<<<<< * * def wiener_like(np.ndarray[double, ndim=1] x, double v, double V, double a, double z, double Z, double t, */ __Pyx_XDECREF(__pyx_r); __Pyx_INCREF(((PyObject *)__pyx_v_y)); __pyx_r = ((PyObject *)__pyx_v_y); goto __pyx_L0; } __pyx_L12:; __pyx_r = Py_None; __Pyx_INCREF(Py_None); goto __pyx_L0; __pyx_L1_error:; __Pyx_XDECREF(__pyx_t_1); __Pyx_XDECREF(__pyx_t_2); __Pyx_XDECREF(__pyx_t_3); __Pyx_XDECREF(__pyx_t_4); __Pyx_XDECREF(__pyx_t_5); { PyObject *__pyx_type, *__pyx_value, *__pyx_tb; __Pyx_ErrFetch(&__pyx_type, &__pyx_value, &__pyx_tb); __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd_y.rcbuffer->pybuffer); __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd_x.rcbuffer->pybuffer); __Pyx_ErrRestore(__pyx_type, __pyx_value, __pyx_tb);} __Pyx_AddTraceback("wfpt.pdf_array", __pyx_clineno, __pyx_lineno, __pyx_filename); __pyx_r = NULL; goto __pyx_L2; __pyx_L0:; __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd_y.rcbuffer->pybuffer); __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd_x.rcbuffer->pybuffer); __pyx_L2:; __Pyx_XDECREF((PyObject *)__pyx_v_y); __Pyx_XGIVEREF(__pyx_r); __Pyx_RefNannyFinishContext(); return __pyx_r; } /* Python wrapper */ static PyObject *__pyx_pw_4wfpt_5wiener_like(PyObject *__pyx_self, PyObject *__pyx_args, PyObject *__pyx_kwds); /*proto*/ static char __pyx_doc_4wfpt_4wiener_like[] = "wiener_like(ndarray x, double v, double V, double a, double z, double Z, double t, double T, double err, int nT=10, int nZ=10, int use_adaptive=1, double simps_err=1e-08)"; static PyMethodDef __pyx_mdef_4wfpt_5wiener_like = {__Pyx_NAMESTR("wiener_like"), (PyCFunction)__pyx_pw_4wfpt_5wiener_like, METH_VARARGS|METH_KEYWORDS, __Pyx_DOCSTR(__pyx_doc_4wfpt_4wiener_like)}; static PyObject *__pyx_pw_4wfpt_5wiener_like(PyObject *__pyx_self, PyObject *__pyx_args, PyObject *__pyx_kwds) { PyArrayObject *__pyx_v_x = 0; double __pyx_v_v; double __pyx_v_V; double __pyx_v_a; double __pyx_v_z; double __pyx_v_Z; double __pyx_v_t; double __pyx_v_T; double __pyx_v_err; int __pyx_v_nT; int __pyx_v_nZ; int __pyx_v_use_adaptive; double __pyx_v_simps_err; static PyObject **__pyx_pyargnames[] = {&__pyx_n_s__x,&__pyx_n_s__v,&__pyx_n_s__V,&__pyx_n_s__a,&__pyx_n_s__z,&__pyx_n_s__Z,&__pyx_n_s__t,&__pyx_n_s__T,&__pyx_n_s__err,&__pyx_n_s__nT,&__pyx_n_s__nZ,&__pyx_n_s__use_adaptive,&__pyx_n_s__simps_err,0}; PyObject *__pyx_r = 0; __Pyx_RefNannyDeclarations __Pyx_RefNannySetupContext("wiener_like (wrapper)"); __pyx_self = __pyx_self; { PyObject* values[13] = {0,0,0,0,0,0,0,0,0,0,0,0,0}; if (unlikely(__pyx_kwds)) { Py_ssize_t kw_args; const Py_ssize_t pos_args = PyTuple_GET_SIZE(__pyx_args); switch (pos_args) { case 13: values[12] = PyTuple_GET_ITEM(__pyx_args, 12); case 12: values[11] = PyTuple_GET_ITEM(__pyx_args, 11); case 11: values[10] = PyTuple_GET_ITEM(__pyx_args, 10); case 10: values[9] = PyTuple_GET_ITEM(__pyx_args, 9); case 9: values[8] = PyTuple_GET_ITEM(__pyx_args, 8); case 8: values[7] = PyTuple_GET_ITEM(__pyx_args, 7); case 7: values[6] = PyTuple_GET_ITEM(__pyx_args, 6); case 6: values[5] = PyTuple_GET_ITEM(__pyx_args, 5); case 5: values[4] = PyTuple_GET_ITEM(__pyx_args, 4); case 4: values[3] = PyTuple_GET_ITEM(__pyx_args, 3); case 3: values[2] = PyTuple_GET_ITEM(__pyx_args, 2); case 2: values[1] = PyTuple_GET_ITEM(__pyx_args, 1); case 1: values[0] = PyTuple_GET_ITEM(__pyx_args, 0); case 0: break; default: goto __pyx_L5_argtuple_error; } kw_args = PyDict_Size(__pyx_kwds); switch (pos_args) { case 0: values[0] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__x); if (likely(values[0])) kw_args--; else goto __pyx_L5_argtuple_error; case 1: values[1] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__v); if (likely(values[1])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("wiener_like", 0, 9, 13, 1); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 41; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 2: values[2] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__V); if (likely(values[2])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("wiener_like", 0, 9, 13, 2); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 41; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 3: values[3] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__a); if (likely(values[3])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("wiener_like", 0, 9, 13, 3); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 41; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 4: values[4] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__z); if (likely(values[4])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("wiener_like", 0, 9, 13, 4); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 41; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 5: values[5] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__Z); if (likely(values[5])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("wiener_like", 0, 9, 13, 5); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 41; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 6: values[6] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__t); if (likely(values[6])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("wiener_like", 0, 9, 13, 6); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 41; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 7: values[7] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__T); if (likely(values[7])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("wiener_like", 0, 9, 13, 7); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 41; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 8: values[8] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__err); if (likely(values[8])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("wiener_like", 0, 9, 13, 8); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 41; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 9: if (kw_args > 0) { PyObject* value = PyDict_GetItem(__pyx_kwds, __pyx_n_s__nT); if (value) { values[9] = value; kw_args--; } } case 10: if (kw_args > 0) { PyObject* value = PyDict_GetItem(__pyx_kwds, __pyx_n_s__nZ); if (value) { values[10] = value; kw_args--; } } case 11: if (kw_args > 0) { PyObject* value = PyDict_GetItem(__pyx_kwds, __pyx_n_s__use_adaptive); if (value) { values[11] = value; kw_args--; } } case 12: if (kw_args > 0) { PyObject* value = PyDict_GetItem(__pyx_kwds, __pyx_n_s__simps_err); if (value) { values[12] = value; kw_args--; } } } if (unlikely(kw_args > 0)) { if (unlikely(__Pyx_ParseOptionalKeywords(__pyx_kwds, __pyx_pyargnames, 0, values, pos_args, "wiener_like") < 0)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 41; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } __pyx_v_x = ((PyArrayObject *)values[0]); __pyx_v_v = __pyx_PyFloat_AsDouble(values[1]); if (unlikely((__pyx_v_v == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 41; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_V = __pyx_PyFloat_AsDouble(values[2]); if (unlikely((__pyx_v_V == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 41; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_a = __pyx_PyFloat_AsDouble(values[3]); if (unlikely((__pyx_v_a == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 41; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_z = __pyx_PyFloat_AsDouble(values[4]); if (unlikely((__pyx_v_z == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 41; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_Z = __pyx_PyFloat_AsDouble(values[5]); if (unlikely((__pyx_v_Z == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 41; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_t = __pyx_PyFloat_AsDouble(values[6]); if (unlikely((__pyx_v_t == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 41; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_T = __pyx_PyFloat_AsDouble(values[7]); if (unlikely((__pyx_v_T == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 42; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_err = __pyx_PyFloat_AsDouble(values[8]); if (unlikely((__pyx_v_err == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 42; __pyx_clineno = __LINE__; goto __pyx_L3_error;} if (values[9]) { __pyx_v_nT = __Pyx_PyInt_AsInt(values[9]); if (unlikely((__pyx_v_nT == (int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 42; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } else { __pyx_v_nT = ((int)10); } if (values[10]) { __pyx_v_nZ = __Pyx_PyInt_AsInt(values[10]); if (unlikely((__pyx_v_nZ == (int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 42; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } else { __pyx_v_nZ = ((int)10); } if (values[11]) { __pyx_v_use_adaptive = __Pyx_PyObject_IsTrue(values[11]); if (unlikely((__pyx_v_use_adaptive == (int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 42; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } else { __pyx_v_use_adaptive = ((int)1); } if (values[12]) { __pyx_v_simps_err = __pyx_PyFloat_AsDouble(values[12]); if (unlikely((__pyx_v_simps_err == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 42; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } else { /* "wfpt.pyx":42 * * def wiener_like(np.ndarray[double, ndim=1] x, double v, double V, double a, double z, double Z, double t, * double T, double err, int nT=10, int nZ=10, bint use_adaptive=1, double simps_err=1e-8): # <<<<<<<<<<<<<< * cdef Py_ssize_t size = x.shape[0] * cdef Py_ssize_t i */ __pyx_v_simps_err = ((double)1e-8); } } else { switch (PyTuple_GET_SIZE(__pyx_args)) { case 13: values[12] = PyTuple_GET_ITEM(__pyx_args, 12); case 12: values[11] = PyTuple_GET_ITEM(__pyx_args, 11); case 11: values[10] = PyTuple_GET_ITEM(__pyx_args, 10); case 10: values[9] = PyTuple_GET_ITEM(__pyx_args, 9); case 9: values[8] = PyTuple_GET_ITEM(__pyx_args, 8); values[7] = PyTuple_GET_ITEM(__pyx_args, 7); values[6] = PyTuple_GET_ITEM(__pyx_args, 6); values[5] = PyTuple_GET_ITEM(__pyx_args, 5); values[4] = PyTuple_GET_ITEM(__pyx_args, 4); values[3] = PyTuple_GET_ITEM(__pyx_args, 3); values[2] = PyTuple_GET_ITEM(__pyx_args, 2); values[1] = PyTuple_GET_ITEM(__pyx_args, 1); values[0] = PyTuple_GET_ITEM(__pyx_args, 0); break; default: goto __pyx_L5_argtuple_error; } } __pyx_v_x = ((PyArrayObject *)values[0]); __pyx_v_v = __pyx_PyFloat_AsDouble(values[1]); if (unlikely((__pyx_v_v == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 41; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_V = __pyx_PyFloat_AsDouble(values[2]); if (unlikely((__pyx_v_V == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 41; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_a = __pyx_PyFloat_AsDouble(values[3]); if (unlikely((__pyx_v_a == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 41; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_z = __pyx_PyFloat_AsDouble(values[4]); if (unlikely((__pyx_v_z == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 41; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_Z = __pyx_PyFloat_AsDouble(values[5]); if (unlikely((__pyx_v_Z == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 41; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_t = __pyx_PyFloat_AsDouble(values[6]); if (unlikely((__pyx_v_t == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 41; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_T = __pyx_PyFloat_AsDouble(values[7]); if (unlikely((__pyx_v_T == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 42; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_err = __pyx_PyFloat_AsDouble(values[8]); if (unlikely((__pyx_v_err == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 42; __pyx_clineno = __LINE__; goto __pyx_L3_error;} if (values[9]) { __pyx_v_nT = __Pyx_PyInt_AsInt(values[9]); if (unlikely((__pyx_v_nT == (int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 42; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } else { __pyx_v_nT = ((int)10); } if (values[10]) { __pyx_v_nZ = __Pyx_PyInt_AsInt(values[10]); if (unlikely((__pyx_v_nZ == (int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 42; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } else { __pyx_v_nZ = ((int)10); } if (values[11]) { __pyx_v_use_adaptive = __Pyx_PyObject_IsTrue(values[11]); if (unlikely((__pyx_v_use_adaptive == (int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 42; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } else { __pyx_v_use_adaptive = ((int)1); } if (values[12]) { __pyx_v_simps_err = __pyx_PyFloat_AsDouble(values[12]); if (unlikely((__pyx_v_simps_err == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 42; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } else { __pyx_v_simps_err = ((double)1e-8); } } goto __pyx_L4_argument_unpacking_done; __pyx_L5_argtuple_error:; __Pyx_RaiseArgtupleInvalid("wiener_like", 0, 9, 13, PyTuple_GET_SIZE(__pyx_args)); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 41; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_L3_error:; __Pyx_AddTraceback("wfpt.wiener_like", __pyx_clineno, __pyx_lineno, __pyx_filename); __Pyx_RefNannyFinishContext(); return NULL; __pyx_L4_argument_unpacking_done:; if (unlikely(!__Pyx_ArgTypeTest(((PyObject *)__pyx_v_x), __pyx_ptype_5numpy_ndarray, 1, "x", 0))) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 41; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __pyx_r = __pyx_pf_4wfpt_4wiener_like(__pyx_self, __pyx_v_x, __pyx_v_v, __pyx_v_V, __pyx_v_a, __pyx_v_z, __pyx_v_Z, __pyx_v_t, __pyx_v_T, __pyx_v_err, __pyx_v_nT, __pyx_v_nZ, __pyx_v_use_adaptive, __pyx_v_simps_err); goto __pyx_L0; __pyx_L1_error:; __pyx_L0:; __Pyx_RefNannyFinishContext(); return __pyx_r; } /* "wfpt.pyx":41 * return y * * def wiener_like(np.ndarray[double, ndim=1] x, double v, double V, double a, double z, double Z, double t, # <<<<<<<<<<<<<< * double T, double err, int nT=10, int nZ=10, bint use_adaptive=1, double simps_err=1e-8): * cdef Py_ssize_t size = x.shape[0] */ static PyObject *__pyx_pf_4wfpt_4wiener_like(CYTHON_UNUSED PyObject *__pyx_self, PyArrayObject *__pyx_v_x, double __pyx_v_v, double __pyx_v_V, double __pyx_v_a, double __pyx_v_z, double __pyx_v_Z, double __pyx_v_t, double __pyx_v_T, double __pyx_v_err, int __pyx_v_nT, int __pyx_v_nZ, int __pyx_v_use_adaptive, double __pyx_v_simps_err) { CYTHON_UNUSED Py_ssize_t __pyx_v_size; Py_ssize_t __pyx_v_i; double __pyx_v_p; double __pyx_v_sum_logp; __Pyx_LocalBuf_ND __pyx_pybuffernd_x; __Pyx_Buffer __pyx_pybuffer_x; PyObject *__pyx_r = NULL; __Pyx_RefNannyDeclarations Py_ssize_t __pyx_t_1; Py_ssize_t __pyx_t_2; Py_ssize_t __pyx_t_3; Py_ssize_t __pyx_t_4; double __pyx_t_5; struct __pyx_opt_args_4wfpt_full_pdf __pyx_t_6; int __pyx_t_7; PyObject *__pyx_t_8 = NULL; PyObject *__pyx_t_9 = NULL; int __pyx_lineno = 0; const char *__pyx_filename = NULL; int __pyx_clineno = 0; __Pyx_RefNannySetupContext("wiener_like"); __pyx_pybuffer_x.pybuffer.buf = NULL; __pyx_pybuffer_x.refcount = 0; __pyx_pybuffernd_x.data = NULL; __pyx_pybuffernd_x.rcbuffer = &__pyx_pybuffer_x; { __Pyx_BufFmt_StackElem __pyx_stack[1]; if (unlikely(__Pyx_GetBufferAndValidate(&__pyx_pybuffernd_x.rcbuffer->pybuffer, (PyObject*)__pyx_v_x, &__Pyx_TypeInfo_double, PyBUF_FORMAT| PyBUF_STRIDES, 1, 0, __pyx_stack) == -1)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 41; __pyx_clineno = __LINE__; goto __pyx_L1_error;} } __pyx_pybuffernd_x.diminfo[0].strides = __pyx_pybuffernd_x.rcbuffer->pybuffer.strides[0]; __pyx_pybuffernd_x.diminfo[0].shape = __pyx_pybuffernd_x.rcbuffer->pybuffer.shape[0]; /* "wfpt.pyx":43 * def wiener_like(np.ndarray[double, ndim=1] x, double v, double V, double a, double z, double Z, double t, * double T, double err, int nT=10, int nZ=10, bint use_adaptive=1, double simps_err=1e-8): * cdef Py_ssize_t size = x.shape[0] # <<<<<<<<<<<<<< * cdef Py_ssize_t i * cdef double p */ __pyx_v_size = (__pyx_v_x->dimensions[0]); /* "wfpt.pyx":46 * cdef Py_ssize_t i * cdef double p * cdef double sum_logp = 0 # <<<<<<<<<<<<<< * * #if num_threads != 0: */ __pyx_v_sum_logp = 0.0; /* "wfpt.pyx":51 * # openmp.omp_set_num_threads(num_threads) * * for i in prange(size, nogil=True, schedule='dynamic'): # <<<<<<<<<<<<<< * p = full_pdf(x[i], v, V, a, z, Z, t, T, err, nT, nZ, use_adaptive, simps_err) * # If one probability = 0, the log sum will be -Inf */ { #ifdef WITH_THREAD PyThreadState *_save = NULL; #endif Py_UNBLOCK_THREADS /*try:*/ { __pyx_t_1 = __pyx_v_size; if (1 == 0) abort(); { Py_ssize_t __pyx_parallel_temp0 = 0xbad0bad0; double __pyx_parallel_temp1 = __PYX_NAN; double __pyx_parallel_temp2 = __PYX_NAN; const char *__pyx_parallel_filename; int __pyx_parallel_lineno, __pyx_parallel_clineno; __pyx_parallel_filename = NULL; __pyx_parallel_lineno = __pyx_parallel_clineno = 0; PyObject *__pyx_parallel_exc_type = NULL, *__pyx_parallel_exc_value = NULL, *__pyx_parallel_exc_tb = NULL; int __pyx_parallel_why; __pyx_parallel_why = 0; __pyx_t_3 = (__pyx_t_1 - 0) / 1; if (__pyx_t_3 > 0) { __pyx_v_i = 0; #ifdef _OPENMP #pragma omp parallel reduction(+:__pyx_v_sum_logp) private(__pyx_t_5, __pyx_t_7, __pyx_t_4, __pyx_t_6) firstprivate(__pyx_t_8, __pyx_t_9) private(__pyx_filename, __pyx_lineno, __pyx_clineno) shared(__pyx_parallel_why, __pyx_parallel_exc_type, __pyx_parallel_exc_value, __pyx_parallel_exc_tb) #endif /* _OPENMP */ { #ifdef _OPENMP #ifdef WITH_THREAD PyGILState_STATE __pyx_gilstate_save = PyGILState_Ensure(); #endif Py_BEGIN_ALLOW_THREADS #endif /* _OPENMP */ #ifdef _OPENMP #pragma omp for firstprivate(__pyx_v_i) lastprivate(__pyx_v_i) lastprivate(__pyx_v_p) schedule(dynamic) #endif /* _OPENMP */ for (__pyx_t_2 = 0; __pyx_t_2 < __pyx_t_3; __pyx_t_2++){ if (__pyx_parallel_why < 2) { __pyx_v_i = 0 + 1 * __pyx_t_2; /* Initialize private variables to invalid values */ __pyx_v_p = ((double)__PYX_NAN); /* "wfpt.pyx":52 * * for i in prange(size, nogil=True, schedule='dynamic'): * p = full_pdf(x[i], v, V, a, z, Z, t, T, err, nT, nZ, use_adaptive, simps_err) # <<<<<<<<<<<<<< * # If one probability = 0, the log sum will be -Inf * if p == 0: */ __pyx_t_4 = __pyx_v_i; __pyx_t_6.__pyx_n = 4; __pyx_t_6.nT = __pyx_v_nT; __pyx_t_6.nZ = __pyx_v_nZ; __pyx_t_6.use_adaptive = __pyx_v_use_adaptive; __pyx_t_6.simps_err = __pyx_v_simps_err; __pyx_t_5 = __pyx_f_4wfpt_full_pdf((*__Pyx_BufPtrStrided1d(double *, __pyx_pybuffernd_x.rcbuffer->pybuffer.buf, __pyx_t_4, __pyx_pybuffernd_x.diminfo[0].strides)), __pyx_v_v, __pyx_v_V, __pyx_v_a, __pyx_v_z, __pyx_v_Z, __pyx_v_t, __pyx_v_T, __pyx_v_err, 0, &__pyx_t_6); __pyx_v_p = __pyx_t_5; /* "wfpt.pyx":54 * p = full_pdf(x[i], v, V, a, z, Z, t, T, err, nT, nZ, use_adaptive, simps_err) * # If one probability = 0, the log sum will be -Inf * if p == 0: # <<<<<<<<<<<<<< * with gil: * return -np.inf */ __pyx_t_7 = (__pyx_v_p == 0.0); if (__pyx_t_7) { /* "wfpt.pyx":55 * # If one probability = 0, the log sum will be -Inf * if p == 0: * with gil: # <<<<<<<<<<<<<< * return -np.inf * */ { #ifdef WITH_THREAD PyGILState_STATE __pyx_gilstate_save = PyGILState_Ensure(); #endif /*try:*/ { /* "wfpt.pyx":56 * if p == 0: * with gil: * return -np.inf # <<<<<<<<<<<<<< * * sum_logp += log(p) */ __Pyx_XDECREF(__pyx_r); __pyx_t_8 = __Pyx_GetName(__pyx_m, __pyx_n_s__np); if (unlikely(!__pyx_t_8)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 56; __pyx_clineno = __LINE__; goto __pyx_L14;} __Pyx_GOTREF(__pyx_t_8); __pyx_t_9 = PyObject_GetAttr(__pyx_t_8, __pyx_n_s__inf); if (unlikely(!__pyx_t_9)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 56; __pyx_clineno = __LINE__; goto __pyx_L14;} __Pyx_GOTREF(__pyx_t_9); __Pyx_DECREF(__pyx_t_8); __pyx_t_8 = 0; __pyx_t_8 = PyNumber_Negative(__pyx_t_9); if (unlikely(!__pyx_t_8)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 56; __pyx_clineno = __LINE__; goto __pyx_L14;} __Pyx_GOTREF(__pyx_t_8); __Pyx_DECREF(__pyx_t_9); __pyx_t_9 = 0; __pyx_r = __pyx_t_8; __pyx_t_8 = 0; goto __pyx_L13; } /* "wfpt.pyx":55 * # If one probability = 0, the log sum will be -Inf * if p == 0: * with gil: # <<<<<<<<<<<<<< * return -np.inf * */ /*finally:*/ { int __pyx_why; __pyx_why = 0; goto __pyx_L15; __pyx_L13: __pyx_why = 3; goto __pyx_L15; __pyx_L14: __pyx_why = 4; goto __pyx_L15; __pyx_L15:; #ifdef WITH_THREAD PyGILState_Release(__pyx_gilstate_save); #endif switch (__pyx_why) { case 3: goto __pyx_L9_return; case 4: goto __pyx_L8_error; } } } goto __pyx_L10; } __pyx_L10:; /* "wfpt.pyx":58 * return -np.inf * * sum_logp += log(p) # <<<<<<<<<<<<<< * * return sum_logp */ __pyx_v_sum_logp = (__pyx_v_sum_logp + log(__pyx_v_p)); goto __pyx_L18; __pyx_L9_return:; __pyx_parallel_why = 3; goto __pyx_L17; __pyx_L8_error:; { #ifdef WITH_THREAD PyGILState_STATE __pyx_gilstate_save = PyGILState_Ensure(); #endif #ifdef _OPENMP #pragma omp flush(__pyx_parallel_exc_type) #endif /* _OPENMP */ if (!__pyx_parallel_exc_type) { __Pyx_ErrFetch(&__pyx_parallel_exc_type, &__pyx_parallel_exc_value, &__pyx_parallel_exc_tb); __pyx_parallel_filename = __pyx_filename; __pyx_parallel_lineno = __pyx_lineno; __pyx_parallel_clineno = __pyx_clineno; __Pyx_GOTREF(__pyx_parallel_exc_type); } #ifdef WITH_THREAD PyGILState_Release(__pyx_gilstate_save); #endif } __pyx_parallel_why = 4; goto __pyx_L17; __pyx_L17:; #ifdef _OPENMP #pragma omp critical(__pyx_parallel_lastprivates0) #endif /* _OPENMP */ { __pyx_parallel_temp0 = __pyx_v_i; __pyx_parallel_temp1 = __pyx_v_p; __pyx_parallel_temp2 = __pyx_v_sum_logp; } __pyx_L18:; #ifdef _OPENMP #pragma omp flush(__pyx_parallel_why) #endif /* _OPENMP */ } } #ifdef _OPENMP Py_END_ALLOW_THREADS #ifdef WITH_THREAD PyGILState_Release(__pyx_gilstate_save); #endif #endif /* _OPENMP */ } } if (__pyx_parallel_exc_type) { /* This may have been overridden by a continue, break or return in another thread. Prefer the error. */ __pyx_parallel_why = 4; } if (__pyx_parallel_why) { __pyx_v_i = __pyx_parallel_temp0; __pyx_v_p = __pyx_parallel_temp1; __pyx_v_sum_logp = __pyx_parallel_temp2; switch (__pyx_parallel_why) { case 3: goto __pyx_L3; case 4: { #ifdef WITH_THREAD PyGILState_STATE __pyx_gilstate_save = PyGILState_Ensure(); #endif __Pyx_ErrRestore(__pyx_parallel_exc_type, __pyx_parallel_exc_value, __pyx_parallel_exc_tb); __pyx_filename = __pyx_parallel_filename; __pyx_lineno = __pyx_parallel_lineno; __pyx_clineno = __pyx_parallel_clineno; __Pyx_GIVEREF(__pyx_parallel_exc_type); #ifdef WITH_THREAD PyGILState_Release(__pyx_gilstate_save); #endif } goto __pyx_L4; } } } } /* "wfpt.pyx":51 * # openmp.omp_set_num_threads(num_threads) * * for i in prange(size, nogil=True, schedule='dynamic'): # <<<<<<<<<<<<<< * p = full_pdf(x[i], v, V, a, z, Z, t, T, err, nT, nZ, use_adaptive, simps_err) * # If one probability = 0, the log sum will be -Inf */ /*finally:*/ { int __pyx_why; __pyx_why = 0; goto __pyx_L5; __pyx_L3: __pyx_why = 3; goto __pyx_L5; __pyx_L4: __pyx_why = 4; goto __pyx_L5; __pyx_L5:; Py_BLOCK_THREADS switch (__pyx_why) { case 3: goto __pyx_L0; case 4: goto __pyx_L1_error; } } } /* "wfpt.pyx":60 * sum_logp += log(p) * * return sum_logp # <<<<<<<<<<<<<< * * def wiener_like_array(np.ndarray[double, ndim=1] x, double v, double V, double a, double z, double Z, double t, */ __Pyx_XDECREF(__pyx_r); __pyx_t_8 = PyFloat_FromDouble(__pyx_v_sum_logp); if (unlikely(!__pyx_t_8)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 60; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_8); __pyx_r = __pyx_t_8; __pyx_t_8 = 0; goto __pyx_L0; __pyx_r = Py_None; __Pyx_INCREF(Py_None); goto __pyx_L0; __pyx_L1_error:; __Pyx_XDECREF(__pyx_t_8); __Pyx_XDECREF(__pyx_t_9); { PyObject *__pyx_type, *__pyx_value, *__pyx_tb; __Pyx_ErrFetch(&__pyx_type, &__pyx_value, &__pyx_tb); __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd_x.rcbuffer->pybuffer); __Pyx_ErrRestore(__pyx_type, __pyx_value, __pyx_tb);} __Pyx_AddTraceback("wfpt.wiener_like", __pyx_clineno, __pyx_lineno, __pyx_filename); __pyx_r = NULL; goto __pyx_L2; __pyx_L0:; __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd_x.rcbuffer->pybuffer); __pyx_L2:; __Pyx_XGIVEREF(__pyx_r); __Pyx_RefNannyFinishContext(); return __pyx_r; } /* Python wrapper */ static PyObject *__pyx_pw_4wfpt_7wiener_like_array(PyObject *__pyx_self, PyObject *__pyx_args, PyObject *__pyx_kwds); /*proto*/ static char __pyx_doc_4wfpt_6wiener_like_array[] = "wiener_like_array(ndarray x, double v, double V, double a, double z, double Z, double t, double T, double err, int nT=10, int nZ=10, int use_adaptive=1, double simps_err=1e-08)"; static PyMethodDef __pyx_mdef_4wfpt_7wiener_like_array = {__Pyx_NAMESTR("wiener_like_array"), (PyCFunction)__pyx_pw_4wfpt_7wiener_like_array, METH_VARARGS|METH_KEYWORDS, __Pyx_DOCSTR(__pyx_doc_4wfpt_6wiener_like_array)}; static PyObject *__pyx_pw_4wfpt_7wiener_like_array(PyObject *__pyx_self, PyObject *__pyx_args, PyObject *__pyx_kwds) { PyArrayObject *__pyx_v_x = 0; double __pyx_v_v; double __pyx_v_V; double __pyx_v_a; double __pyx_v_z; double __pyx_v_Z; double __pyx_v_t; double __pyx_v_T; double __pyx_v_err; int __pyx_v_nT; int __pyx_v_nZ; int __pyx_v_use_adaptive; double __pyx_v_simps_err; static PyObject **__pyx_pyargnames[] = {&__pyx_n_s__x,&__pyx_n_s__v,&__pyx_n_s__V,&__pyx_n_s__a,&__pyx_n_s__z,&__pyx_n_s__Z,&__pyx_n_s__t,&__pyx_n_s__T,&__pyx_n_s__err,&__pyx_n_s__nT,&__pyx_n_s__nZ,&__pyx_n_s__use_adaptive,&__pyx_n_s__simps_err,0}; PyObject *__pyx_r = 0; __Pyx_RefNannyDeclarations __Pyx_RefNannySetupContext("wiener_like_array (wrapper)"); __pyx_self = __pyx_self; { PyObject* values[13] = {0,0,0,0,0,0,0,0,0,0,0,0,0}; if (unlikely(__pyx_kwds)) { Py_ssize_t kw_args; const Py_ssize_t pos_args = PyTuple_GET_SIZE(__pyx_args); switch (pos_args) { case 13: values[12] = PyTuple_GET_ITEM(__pyx_args, 12); case 12: values[11] = PyTuple_GET_ITEM(__pyx_args, 11); case 11: values[10] = PyTuple_GET_ITEM(__pyx_args, 10); case 10: values[9] = PyTuple_GET_ITEM(__pyx_args, 9); case 9: values[8] = PyTuple_GET_ITEM(__pyx_args, 8); case 8: values[7] = PyTuple_GET_ITEM(__pyx_args, 7); case 7: values[6] = PyTuple_GET_ITEM(__pyx_args, 6); case 6: values[5] = PyTuple_GET_ITEM(__pyx_args, 5); case 5: values[4] = PyTuple_GET_ITEM(__pyx_args, 4); case 4: values[3] = PyTuple_GET_ITEM(__pyx_args, 3); case 3: values[2] = PyTuple_GET_ITEM(__pyx_args, 2); case 2: values[1] = PyTuple_GET_ITEM(__pyx_args, 1); case 1: values[0] = PyTuple_GET_ITEM(__pyx_args, 0); case 0: break; default: goto __pyx_L5_argtuple_error; } kw_args = PyDict_Size(__pyx_kwds); switch (pos_args) { case 0: values[0] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__x); if (likely(values[0])) kw_args--; else goto __pyx_L5_argtuple_error; case 1: values[1] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__v); if (likely(values[1])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("wiener_like_array", 0, 9, 13, 1); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 62; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 2: values[2] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__V); if (likely(values[2])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("wiener_like_array", 0, 9, 13, 2); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 62; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 3: values[3] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__a); if (likely(values[3])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("wiener_like_array", 0, 9, 13, 3); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 62; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 4: values[4] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__z); if (likely(values[4])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("wiener_like_array", 0, 9, 13, 4); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 62; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 5: values[5] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__Z); if (likely(values[5])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("wiener_like_array", 0, 9, 13, 5); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 62; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 6: values[6] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__t); if (likely(values[6])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("wiener_like_array", 0, 9, 13, 6); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 62; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 7: values[7] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__T); if (likely(values[7])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("wiener_like_array", 0, 9, 13, 7); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 62; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 8: values[8] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__err); if (likely(values[8])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("wiener_like_array", 0, 9, 13, 8); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 62; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 9: if (kw_args > 0) { PyObject* value = PyDict_GetItem(__pyx_kwds, __pyx_n_s__nT); if (value) { values[9] = value; kw_args--; } } case 10: if (kw_args > 0) { PyObject* value = PyDict_GetItem(__pyx_kwds, __pyx_n_s__nZ); if (value) { values[10] = value; kw_args--; } } case 11: if (kw_args > 0) { PyObject* value = PyDict_GetItem(__pyx_kwds, __pyx_n_s__use_adaptive); if (value) { values[11] = value; kw_args--; } } case 12: if (kw_args > 0) { PyObject* value = PyDict_GetItem(__pyx_kwds, __pyx_n_s__simps_err); if (value) { values[12] = value; kw_args--; } } } if (unlikely(kw_args > 0)) { if (unlikely(__Pyx_ParseOptionalKeywords(__pyx_kwds, __pyx_pyargnames, 0, values, pos_args, "wiener_like_array") < 0)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 62; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } __pyx_v_x = ((PyArrayObject *)values[0]); __pyx_v_v = __pyx_PyFloat_AsDouble(values[1]); if (unlikely((__pyx_v_v == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 62; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_V = __pyx_PyFloat_AsDouble(values[2]); if (unlikely((__pyx_v_V == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 62; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_a = __pyx_PyFloat_AsDouble(values[3]); if (unlikely((__pyx_v_a == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 62; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_z = __pyx_PyFloat_AsDouble(values[4]); if (unlikely((__pyx_v_z == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 62; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_Z = __pyx_PyFloat_AsDouble(values[5]); if (unlikely((__pyx_v_Z == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 62; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_t = __pyx_PyFloat_AsDouble(values[6]); if (unlikely((__pyx_v_t == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 62; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_T = __pyx_PyFloat_AsDouble(values[7]); if (unlikely((__pyx_v_T == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 63; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_err = __pyx_PyFloat_AsDouble(values[8]); if (unlikely((__pyx_v_err == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 63; __pyx_clineno = __LINE__; goto __pyx_L3_error;} if (values[9]) { __pyx_v_nT = __Pyx_PyInt_AsInt(values[9]); if (unlikely((__pyx_v_nT == (int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 63; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } else { __pyx_v_nT = ((int)10); } if (values[10]) { __pyx_v_nZ = __Pyx_PyInt_AsInt(values[10]); if (unlikely((__pyx_v_nZ == (int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 63; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } else { __pyx_v_nZ = ((int)10); } if (values[11]) { __pyx_v_use_adaptive = __Pyx_PyObject_IsTrue(values[11]); if (unlikely((__pyx_v_use_adaptive == (int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 63; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } else { __pyx_v_use_adaptive = ((int)1); } if (values[12]) { __pyx_v_simps_err = __pyx_PyFloat_AsDouble(values[12]); if (unlikely((__pyx_v_simps_err == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 63; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } else { /* "wfpt.pyx":63 * * def wiener_like_array(np.ndarray[double, ndim=1] x, double v, double V, double a, double z, double Z, double t, * double T, double err, int nT=10, int nZ=10, bint use_adaptive=1, double simps_err=1e-8): # <<<<<<<<<<<<<< * cdef Py_ssize_t size = x.shape[0] * cdef Py_ssize_t i */ __pyx_v_simps_err = ((double)1e-8); } } else { switch (PyTuple_GET_SIZE(__pyx_args)) { case 13: values[12] = PyTuple_GET_ITEM(__pyx_args, 12); case 12: values[11] = PyTuple_GET_ITEM(__pyx_args, 11); case 11: values[10] = PyTuple_GET_ITEM(__pyx_args, 10); case 10: values[9] = PyTuple_GET_ITEM(__pyx_args, 9); case 9: values[8] = PyTuple_GET_ITEM(__pyx_args, 8); values[7] = PyTuple_GET_ITEM(__pyx_args, 7); values[6] = PyTuple_GET_ITEM(__pyx_args, 6); values[5] = PyTuple_GET_ITEM(__pyx_args, 5); values[4] = PyTuple_GET_ITEM(__pyx_args, 4); values[3] = PyTuple_GET_ITEM(__pyx_args, 3); values[2] = PyTuple_GET_ITEM(__pyx_args, 2); values[1] = PyTuple_GET_ITEM(__pyx_args, 1); values[0] = PyTuple_GET_ITEM(__pyx_args, 0); break; default: goto __pyx_L5_argtuple_error; } } __pyx_v_x = ((PyArrayObject *)values[0]); __pyx_v_v = __pyx_PyFloat_AsDouble(values[1]); if (unlikely((__pyx_v_v == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 62; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_V = __pyx_PyFloat_AsDouble(values[2]); if (unlikely((__pyx_v_V == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 62; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_a = __pyx_PyFloat_AsDouble(values[3]); if (unlikely((__pyx_v_a == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 62; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_z = __pyx_PyFloat_AsDouble(values[4]); if (unlikely((__pyx_v_z == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 62; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_Z = __pyx_PyFloat_AsDouble(values[5]); if (unlikely((__pyx_v_Z == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 62; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_t = __pyx_PyFloat_AsDouble(values[6]); if (unlikely((__pyx_v_t == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 62; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_T = __pyx_PyFloat_AsDouble(values[7]); if (unlikely((__pyx_v_T == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 63; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_err = __pyx_PyFloat_AsDouble(values[8]); if (unlikely((__pyx_v_err == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 63; __pyx_clineno = __LINE__; goto __pyx_L3_error;} if (values[9]) { __pyx_v_nT = __Pyx_PyInt_AsInt(values[9]); if (unlikely((__pyx_v_nT == (int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 63; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } else { __pyx_v_nT = ((int)10); } if (values[10]) { __pyx_v_nZ = __Pyx_PyInt_AsInt(values[10]); if (unlikely((__pyx_v_nZ == (int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 63; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } else { __pyx_v_nZ = ((int)10); } if (values[11]) { __pyx_v_use_adaptive = __Pyx_PyObject_IsTrue(values[11]); if (unlikely((__pyx_v_use_adaptive == (int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 63; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } else { __pyx_v_use_adaptive = ((int)1); } if (values[12]) { __pyx_v_simps_err = __pyx_PyFloat_AsDouble(values[12]); if (unlikely((__pyx_v_simps_err == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 63; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } else { __pyx_v_simps_err = ((double)1e-8); } } goto __pyx_L4_argument_unpacking_done; __pyx_L5_argtuple_error:; __Pyx_RaiseArgtupleInvalid("wiener_like_array", 0, 9, 13, PyTuple_GET_SIZE(__pyx_args)); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 62; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_L3_error:; __Pyx_AddTraceback("wfpt.wiener_like_array", __pyx_clineno, __pyx_lineno, __pyx_filename); __Pyx_RefNannyFinishContext(); return NULL; __pyx_L4_argument_unpacking_done:; if (unlikely(!__Pyx_ArgTypeTest(((PyObject *)__pyx_v_x), __pyx_ptype_5numpy_ndarray, 1, "x", 0))) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 62; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __pyx_r = __pyx_pf_4wfpt_6wiener_like_array(__pyx_self, __pyx_v_x, __pyx_v_v, __pyx_v_V, __pyx_v_a, __pyx_v_z, __pyx_v_Z, __pyx_v_t, __pyx_v_T, __pyx_v_err, __pyx_v_nT, __pyx_v_nZ, __pyx_v_use_adaptive, __pyx_v_simps_err); goto __pyx_L0; __pyx_L1_error:; __pyx_L0:; __Pyx_RefNannyFinishContext(); return __pyx_r; } /* "wfpt.pyx":62 * return sum_logp * * def wiener_like_array(np.ndarray[double, ndim=1] x, double v, double V, double a, double z, double Z, double t, # <<<<<<<<<<<<<< * double T, double err, int nT=10, int nZ=10, bint use_adaptive=1, double simps_err=1e-8): * cdef Py_ssize_t size = x.shape[0] */ static PyObject *__pyx_pf_4wfpt_6wiener_like_array(CYTHON_UNUSED PyObject *__pyx_self, PyArrayObject *__pyx_v_x, double __pyx_v_v, double __pyx_v_V, double __pyx_v_a, double __pyx_v_z, double __pyx_v_Z, double __pyx_v_t, double __pyx_v_T, double __pyx_v_err, int __pyx_v_nT, int __pyx_v_nZ, int __pyx_v_use_adaptive, double __pyx_v_simps_err) { Py_ssize_t __pyx_v_size; Py_ssize_t __pyx_v_i; double __pyx_v_sum_logp; PyArrayObject *__pyx_v_p_array = 0; __Pyx_LocalBuf_ND __pyx_pybuffernd_p_array; __Pyx_Buffer __pyx_pybuffer_p_array; __Pyx_LocalBuf_ND __pyx_pybuffernd_x; __Pyx_Buffer __pyx_pybuffer_x; PyObject *__pyx_r = NULL; __Pyx_RefNannyDeclarations PyObject *__pyx_t_1 = NULL; PyObject *__pyx_t_2 = NULL; PyObject *__pyx_t_3 = NULL; PyObject *__pyx_t_4 = NULL; PyObject *__pyx_t_5 = NULL; PyArrayObject *__pyx_t_6 = NULL; Py_ssize_t __pyx_t_7; Py_ssize_t __pyx_t_8; Py_ssize_t __pyx_t_9; Py_ssize_t __pyx_t_10; double __pyx_t_11; struct __pyx_opt_args_4wfpt_full_pdf __pyx_t_12; Py_ssize_t __pyx_t_13; Py_ssize_t __pyx_t_14; int __pyx_t_15; Py_ssize_t __pyx_t_16; int __pyx_lineno = 0; const char *__pyx_filename = NULL; int __pyx_clineno = 0; __Pyx_RefNannySetupContext("wiener_like_array"); __pyx_pybuffer_p_array.pybuffer.buf = NULL; __pyx_pybuffer_p_array.refcount = 0; __pyx_pybuffernd_p_array.data = NULL; __pyx_pybuffernd_p_array.rcbuffer = &__pyx_pybuffer_p_array; __pyx_pybuffer_x.pybuffer.buf = NULL; __pyx_pybuffer_x.refcount = 0; __pyx_pybuffernd_x.data = NULL; __pyx_pybuffernd_x.rcbuffer = &__pyx_pybuffer_x; { __Pyx_BufFmt_StackElem __pyx_stack[1]; if (unlikely(__Pyx_GetBufferAndValidate(&__pyx_pybuffernd_x.rcbuffer->pybuffer, (PyObject*)__pyx_v_x, &__Pyx_TypeInfo_double, PyBUF_FORMAT| PyBUF_STRIDES, 1, 0, __pyx_stack) == -1)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 62; __pyx_clineno = __LINE__; goto __pyx_L1_error;} } __pyx_pybuffernd_x.diminfo[0].strides = __pyx_pybuffernd_x.rcbuffer->pybuffer.strides[0]; __pyx_pybuffernd_x.diminfo[0].shape = __pyx_pybuffernd_x.rcbuffer->pybuffer.shape[0]; /* "wfpt.pyx":64 * def wiener_like_array(np.ndarray[double, ndim=1] x, double v, double V, double a, double z, double Z, double t, * double T, double err, int nT=10, int nZ=10, bint use_adaptive=1, double simps_err=1e-8): * cdef Py_ssize_t size = x.shape[0] # <<<<<<<<<<<<<< * cdef Py_ssize_t i * cdef double sum_logp = 0 */ __pyx_v_size = (__pyx_v_x->dimensions[0]); /* "wfpt.pyx":66 * cdef Py_ssize_t size = x.shape[0] * cdef Py_ssize_t i * cdef double sum_logp = 0 # <<<<<<<<<<<<<< * * cdef np.ndarray[double, ndim=1] p_array = np.empty(size, dtype=np.float) */ __pyx_v_sum_logp = 0.0; /* "wfpt.pyx":68 * cdef double sum_logp = 0 * * cdef np.ndarray[double, ndim=1] p_array = np.empty(size, dtype=np.float) # <<<<<<<<<<<<<< * * #if num_threads != 0: */ __pyx_t_1 = __Pyx_GetName(__pyx_m, __pyx_n_s__np); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 68; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_1); __pyx_t_2 = PyObject_GetAttr(__pyx_t_1, __pyx_n_s__empty); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 68; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_2); __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; __pyx_t_1 = PyInt_FromSsize_t(__pyx_v_size); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 68; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_1); __pyx_t_3 = PyTuple_New(1); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 68; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); PyTuple_SET_ITEM(__pyx_t_3, 0, __pyx_t_1); __Pyx_GIVEREF(__pyx_t_1); __pyx_t_1 = 0; __pyx_t_1 = PyDict_New(); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 68; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(((PyObject *)__pyx_t_1)); __pyx_t_4 = __Pyx_GetName(__pyx_m, __pyx_n_s__np); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 68; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_4); __pyx_t_5 = PyObject_GetAttr(__pyx_t_4, __pyx_n_s__float); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 68; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_5); __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0; if (PyDict_SetItem(__pyx_t_1, ((PyObject *)__pyx_n_s__dtype), __pyx_t_5) < 0) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 68; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0; __pyx_t_5 = PyObject_Call(__pyx_t_2, ((PyObject *)__pyx_t_3), ((PyObject *)__pyx_t_1)); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 68; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_5); __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; __Pyx_DECREF(((PyObject *)__pyx_t_3)); __pyx_t_3 = 0; __Pyx_DECREF(((PyObject *)__pyx_t_1)); __pyx_t_1 = 0; if (!(likely(((__pyx_t_5) == Py_None) || likely(__Pyx_TypeTest(__pyx_t_5, __pyx_ptype_5numpy_ndarray))))) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 68; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __pyx_t_6 = ((PyArrayObject *)__pyx_t_5); { __Pyx_BufFmt_StackElem __pyx_stack[1]; if (unlikely(__Pyx_GetBufferAndValidate(&__pyx_pybuffernd_p_array.rcbuffer->pybuffer, (PyObject*)__pyx_t_6, &__Pyx_TypeInfo_double, PyBUF_FORMAT| PyBUF_STRIDES| PyBUF_WRITABLE, 1, 0, __pyx_stack) == -1)) { __pyx_v_p_array = ((PyArrayObject *)Py_None); __Pyx_INCREF(Py_None); __pyx_pybuffernd_p_array.rcbuffer->pybuffer.buf = NULL; {__pyx_filename = __pyx_f[0]; __pyx_lineno = 68; __pyx_clineno = __LINE__; goto __pyx_L1_error;} } else {__pyx_pybuffernd_p_array.diminfo[0].strides = __pyx_pybuffernd_p_array.rcbuffer->pybuffer.strides[0]; __pyx_pybuffernd_p_array.diminfo[0].shape = __pyx_pybuffernd_p_array.rcbuffer->pybuffer.shape[0]; } } __pyx_t_6 = 0; __pyx_v_p_array = ((PyArrayObject *)__pyx_t_5); __pyx_t_5 = 0; /* "wfpt.pyx":73 * # openmp.omp_set_num_threads(num_threads) * * for i in prange(size, nogil=True, schedule='dynamic'): # <<<<<<<<<<<<<< * p_array[i] = full_pdf(x[i], v, V, a, z, Z, t, T, err, nT, nZ, use_adaptive, simps_err) * # If one probability = 0, the log sum will be -Inf */ { #ifdef WITH_THREAD PyThreadState *_save = NULL; #endif Py_UNBLOCK_THREADS /*try:*/ { __pyx_t_7 = __pyx_v_size; if (1 == 0) abort(); { Py_ssize_t __pyx_parallel_temp0 = 0xbad0bad0; const char *__pyx_parallel_filename; int __pyx_parallel_lineno, __pyx_parallel_clineno; __pyx_parallel_filename = NULL; __pyx_parallel_lineno = __pyx_parallel_clineno = 0; PyObject *__pyx_parallel_exc_type = NULL, *__pyx_parallel_exc_value = NULL, *__pyx_parallel_exc_tb = NULL; int __pyx_parallel_why; __pyx_parallel_why = 0; __pyx_t_9 = (__pyx_t_7 - 0) / 1; if (__pyx_t_9 > 0) { __pyx_v_i = 0; #ifdef _OPENMP #pragma omp parallel private(__pyx_t_12, __pyx_t_13, __pyx_t_11, __pyx_t_14, __pyx_t_10, __pyx_t_15) firstprivate(__pyx_t_1, __pyx_t_5) private(__pyx_filename, __pyx_lineno, __pyx_clineno) shared(__pyx_parallel_why, __pyx_parallel_exc_type, __pyx_parallel_exc_value, __pyx_parallel_exc_tb) #endif /* _OPENMP */ { #ifdef _OPENMP #ifdef WITH_THREAD PyGILState_STATE __pyx_gilstate_save = PyGILState_Ensure(); #endif Py_BEGIN_ALLOW_THREADS #endif /* _OPENMP */ #ifdef _OPENMP #pragma omp for firstprivate(__pyx_v_i) lastprivate(__pyx_v_i) schedule(dynamic) #endif /* _OPENMP */ for (__pyx_t_8 = 0; __pyx_t_8 < __pyx_t_9; __pyx_t_8++){ if (__pyx_parallel_why < 2) { __pyx_v_i = 0 + 1 * __pyx_t_8; /* "wfpt.pyx":74 * * for i in prange(size, nogil=True, schedule='dynamic'): * p_array[i] = full_pdf(x[i], v, V, a, z, Z, t, T, err, nT, nZ, use_adaptive, simps_err) # <<<<<<<<<<<<<< * # If one probability = 0, the log sum will be -Inf * if p_array[i] == 0: */ __pyx_t_10 = __pyx_v_i; __pyx_t_12.__pyx_n = 4; __pyx_t_12.nT = __pyx_v_nT; __pyx_t_12.nZ = __pyx_v_nZ; __pyx_t_12.use_adaptive = __pyx_v_use_adaptive; __pyx_t_12.simps_err = __pyx_v_simps_err; __pyx_t_11 = __pyx_f_4wfpt_full_pdf((*__Pyx_BufPtrStrided1d(double *, __pyx_pybuffernd_x.rcbuffer->pybuffer.buf, __pyx_t_10, __pyx_pybuffernd_x.diminfo[0].strides)), __pyx_v_v, __pyx_v_V, __pyx_v_a, __pyx_v_z, __pyx_v_Z, __pyx_v_t, __pyx_v_T, __pyx_v_err, 0, &__pyx_t_12); __pyx_t_13 = __pyx_v_i; *__Pyx_BufPtrStrided1d(double *, __pyx_pybuffernd_p_array.rcbuffer->pybuffer.buf, __pyx_t_13, __pyx_pybuffernd_p_array.diminfo[0].strides) = __pyx_t_11; /* "wfpt.pyx":76 * p_array[i] = full_pdf(x[i], v, V, a, z, Z, t, T, err, nT, nZ, use_adaptive, simps_err) * # If one probability = 0, the log sum will be -Inf * if p_array[i] == 0: # <<<<<<<<<<<<<< * with gil: * return -np.inf */ __pyx_t_14 = __pyx_v_i; __pyx_t_15 = ((*__Pyx_BufPtrStrided1d(double *, __pyx_pybuffernd_p_array.rcbuffer->pybuffer.buf, __pyx_t_14, __pyx_pybuffernd_p_array.diminfo[0].strides)) == 0.0); if (__pyx_t_15) { /* "wfpt.pyx":77 * # If one probability = 0, the log sum will be -Inf * if p_array[i] == 0: * with gil: # <<<<<<<<<<<<<< * return -np.inf * */ { #ifdef WITH_THREAD PyGILState_STATE __pyx_gilstate_save = PyGILState_Ensure(); #endif /*try:*/ { /* "wfpt.pyx":78 * if p_array[i] == 0: * with gil: * return -np.inf # <<<<<<<<<<<<<< * * for i in prange(size, nogil=True, schedule='dynamic'): */ __Pyx_XDECREF(__pyx_r); __pyx_t_5 = __Pyx_GetName(__pyx_m, __pyx_n_s__np); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 78; __pyx_clineno = __LINE__; goto __pyx_L14;} __Pyx_GOTREF(__pyx_t_5); __pyx_t_1 = PyObject_GetAttr(__pyx_t_5, __pyx_n_s__inf); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 78; __pyx_clineno = __LINE__; goto __pyx_L14;} __Pyx_GOTREF(__pyx_t_1); __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0; __pyx_t_5 = PyNumber_Negative(__pyx_t_1); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 78; __pyx_clineno = __LINE__; goto __pyx_L14;} __Pyx_GOTREF(__pyx_t_5); __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; __pyx_r = __pyx_t_5; __pyx_t_5 = 0; goto __pyx_L13; } /* "wfpt.pyx":77 * # If one probability = 0, the log sum will be -Inf * if p_array[i] == 0: * with gil: # <<<<<<<<<<<<<< * return -np.inf * */ /*finally:*/ { int __pyx_why; __pyx_why = 0; goto __pyx_L15; __pyx_L13: __pyx_why = 3; goto __pyx_L15; __pyx_L14: __pyx_why = 4; goto __pyx_L15; __pyx_L15:; #ifdef WITH_THREAD PyGILState_Release(__pyx_gilstate_save); #endif switch (__pyx_why) { case 3: goto __pyx_L9_return; case 4: goto __pyx_L8_error; } } } goto __pyx_L10; } __pyx_L10:; goto __pyx_L18; __pyx_L9_return:; __pyx_parallel_why = 3; goto __pyx_L17; __pyx_L8_error:; { #ifdef WITH_THREAD PyGILState_STATE __pyx_gilstate_save = PyGILState_Ensure(); #endif #ifdef _OPENMP #pragma omp flush(__pyx_parallel_exc_type) #endif /* _OPENMP */ if (!__pyx_parallel_exc_type) { __Pyx_ErrFetch(&__pyx_parallel_exc_type, &__pyx_parallel_exc_value, &__pyx_parallel_exc_tb); __pyx_parallel_filename = __pyx_filename; __pyx_parallel_lineno = __pyx_lineno; __pyx_parallel_clineno = __pyx_clineno; __Pyx_GOTREF(__pyx_parallel_exc_type); } #ifdef WITH_THREAD PyGILState_Release(__pyx_gilstate_save); #endif } __pyx_parallel_why = 4; goto __pyx_L17; __pyx_L17:; #ifdef _OPENMP #pragma omp critical(__pyx_parallel_lastprivates1) #endif /* _OPENMP */ { __pyx_parallel_temp0 = __pyx_v_i; } __pyx_L18:; #ifdef _OPENMP #pragma omp flush(__pyx_parallel_why) #endif /* _OPENMP */ } } #ifdef _OPENMP Py_END_ALLOW_THREADS #ifdef WITH_THREAD PyGILState_Release(__pyx_gilstate_save); #endif #endif /* _OPENMP */ } } if (__pyx_parallel_exc_type) { /* This may have been overridden by a continue, break or return in another thread. Prefer the error. */ __pyx_parallel_why = 4; } if (__pyx_parallel_why) { __pyx_v_i = __pyx_parallel_temp0; switch (__pyx_parallel_why) { case 3: goto __pyx_L3; case 4: { #ifdef WITH_THREAD PyGILState_STATE __pyx_gilstate_save = PyGILState_Ensure(); #endif __Pyx_ErrRestore(__pyx_parallel_exc_type, __pyx_parallel_exc_value, __pyx_parallel_exc_tb); __pyx_filename = __pyx_parallel_filename; __pyx_lineno = __pyx_parallel_lineno; __pyx_clineno = __pyx_parallel_clineno; __Pyx_GIVEREF(__pyx_parallel_exc_type); #ifdef WITH_THREAD PyGILState_Release(__pyx_gilstate_save); #endif } goto __pyx_L4; } } } } /* "wfpt.pyx":73 * # openmp.omp_set_num_threads(num_threads) * * for i in prange(size, nogil=True, schedule='dynamic'): # <<<<<<<<<<<<<< * p_array[i] = full_pdf(x[i], v, V, a, z, Z, t, T, err, nT, nZ, use_adaptive, simps_err) * # If one probability = 0, the log sum will be -Inf */ /*finally:*/ { int __pyx_why; __pyx_why = 0; goto __pyx_L5; __pyx_L3: __pyx_why = 3; goto __pyx_L5; __pyx_L4: __pyx_why = 4; goto __pyx_L5; __pyx_L5:; Py_BLOCK_THREADS switch (__pyx_why) { case 3: goto __pyx_L0; case 4: goto __pyx_L1_error; } } } /* "wfpt.pyx":80 * return -np.inf * * for i in prange(size, nogil=True, schedule='dynamic'): # <<<<<<<<<<<<<< * sum_logp += log(p_array[i]) * */ { #ifdef WITH_THREAD PyThreadState *_save = NULL; #endif Py_UNBLOCK_THREADS /*try:*/ { __pyx_t_9 = __pyx_v_size; if (1 == 0) abort(); { __pyx_t_7 = (__pyx_t_9 - 0) / 1; if (__pyx_t_7 > 0) { __pyx_v_i = 0; #ifdef _OPENMP #pragma omp parallel reduction(+:__pyx_v_sum_logp) private(__pyx_t_16) #endif /* _OPENMP */ { #ifdef _OPENMP #pragma omp for firstprivate(__pyx_v_i) lastprivate(__pyx_v_i) schedule(dynamic) #endif /* _OPENMP */ for (__pyx_t_8 = 0; __pyx_t_8 < __pyx_t_7; __pyx_t_8++){ { __pyx_v_i = 0 + 1 * __pyx_t_8; /* "wfpt.pyx":81 * * for i in prange(size, nogil=True, schedule='dynamic'): * sum_logp += log(p_array[i]) # <<<<<<<<<<<<<< * * return sum_logp */ __pyx_t_16 = __pyx_v_i; __pyx_v_sum_logp = (__pyx_v_sum_logp + log((*__Pyx_BufPtrStrided1d(double *, __pyx_pybuffernd_p_array.rcbuffer->pybuffer.buf, __pyx_t_16, __pyx_pybuffernd_p_array.diminfo[0].strides)))); } } } } } } /* "wfpt.pyx":80 * return -np.inf * * for i in prange(size, nogil=True, schedule='dynamic'): # <<<<<<<<<<<<<< * sum_logp += log(p_array[i]) * */ /*finally:*/ { Py_BLOCK_THREADS } } /* "wfpt.pyx":83 * sum_logp += log(p_array[i]) * * return sum_logp # <<<<<<<<<<<<<< * * def wiener_like_multi(np.ndarray[double, ndim=1] x, v, V, a, z, Z, t, T, double err, multi=None): */ __Pyx_XDECREF(__pyx_r); __pyx_t_5 = PyFloat_FromDouble(__pyx_v_sum_logp); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 83; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_5); __pyx_r = __pyx_t_5; __pyx_t_5 = 0; goto __pyx_L0; __pyx_r = Py_None; __Pyx_INCREF(Py_None); goto __pyx_L0; __pyx_L1_error:; __Pyx_XDECREF(__pyx_t_1); __Pyx_XDECREF(__pyx_t_2); __Pyx_XDECREF(__pyx_t_3); __Pyx_XDECREF(__pyx_t_4); __Pyx_XDECREF(__pyx_t_5); { PyObject *__pyx_type, *__pyx_value, *__pyx_tb; __Pyx_ErrFetch(&__pyx_type, &__pyx_value, &__pyx_tb); __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd_p_array.rcbuffer->pybuffer); __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd_x.rcbuffer->pybuffer); __Pyx_ErrRestore(__pyx_type, __pyx_value, __pyx_tb);} __Pyx_AddTraceback("wfpt.wiener_like_array", __pyx_clineno, __pyx_lineno, __pyx_filename); __pyx_r = NULL; goto __pyx_L2; __pyx_L0:; __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd_p_array.rcbuffer->pybuffer); __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd_x.rcbuffer->pybuffer); __pyx_L2:; __Pyx_XDECREF((PyObject *)__pyx_v_p_array); __Pyx_XGIVEREF(__pyx_r); __Pyx_RefNannyFinishContext(); return __pyx_r; } /* Python wrapper */ static PyObject *__pyx_pw_4wfpt_9wiener_like_multi(PyObject *__pyx_self, PyObject *__pyx_args, PyObject *__pyx_kwds); /*proto*/ static char __pyx_doc_4wfpt_8wiener_like_multi[] = "wiener_like_multi(ndarray x, v, V, a, z, Z, t, T, double err, multi=None)"; static PyMethodDef __pyx_mdef_4wfpt_9wiener_like_multi = {__Pyx_NAMESTR("wiener_like_multi"), (PyCFunction)__pyx_pw_4wfpt_9wiener_like_multi, METH_VARARGS|METH_KEYWORDS, __Pyx_DOCSTR(__pyx_doc_4wfpt_8wiener_like_multi)}; static PyObject *__pyx_pw_4wfpt_9wiener_like_multi(PyObject *__pyx_self, PyObject *__pyx_args, PyObject *__pyx_kwds) { PyArrayObject *__pyx_v_x = 0; PyObject *__pyx_v_v = 0; PyObject *__pyx_v_V = 0; PyObject *__pyx_v_a = 0; PyObject *__pyx_v_z = 0; PyObject *__pyx_v_Z = 0; PyObject *__pyx_v_t = 0; PyObject *__pyx_v_T = 0; double __pyx_v_err; PyObject *__pyx_v_multi = 0; static PyObject **__pyx_pyargnames[] = {&__pyx_n_s__x,&__pyx_n_s__v,&__pyx_n_s__V,&__pyx_n_s__a,&__pyx_n_s__z,&__pyx_n_s__Z,&__pyx_n_s__t,&__pyx_n_s__T,&__pyx_n_s__err,&__pyx_n_s__multi,0}; PyObject *__pyx_r = 0; __Pyx_RefNannyDeclarations __Pyx_RefNannySetupContext("wiener_like_multi (wrapper)"); __pyx_self = __pyx_self; { PyObject* values[10] = {0,0,0,0,0,0,0,0,0,0}; /* "wfpt.pyx":85 * return sum_logp * * def wiener_like_multi(np.ndarray[double, ndim=1] x, v, V, a, z, Z, t, T, double err, multi=None): # <<<<<<<<<<<<<< * cdef Py_ssize_t size = x.shape[0] * cdef Py_ssize_t i */ values[9] = ((PyObject *)Py_None); if (unlikely(__pyx_kwds)) { Py_ssize_t kw_args; const Py_ssize_t pos_args = PyTuple_GET_SIZE(__pyx_args); switch (pos_args) { case 10: values[9] = PyTuple_GET_ITEM(__pyx_args, 9); case 9: values[8] = PyTuple_GET_ITEM(__pyx_args, 8); case 8: values[7] = PyTuple_GET_ITEM(__pyx_args, 7); case 7: values[6] = PyTuple_GET_ITEM(__pyx_args, 6); case 6: values[5] = PyTuple_GET_ITEM(__pyx_args, 5); case 5: values[4] = PyTuple_GET_ITEM(__pyx_args, 4); case 4: values[3] = PyTuple_GET_ITEM(__pyx_args, 3); case 3: values[2] = PyTuple_GET_ITEM(__pyx_args, 2); case 2: values[1] = PyTuple_GET_ITEM(__pyx_args, 1); case 1: values[0] = PyTuple_GET_ITEM(__pyx_args, 0); case 0: break; default: goto __pyx_L5_argtuple_error; } kw_args = PyDict_Size(__pyx_kwds); switch (pos_args) { case 0: values[0] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__x); if (likely(values[0])) kw_args--; else goto __pyx_L5_argtuple_error; case 1: values[1] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__v); if (likely(values[1])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("wiener_like_multi", 0, 9, 10, 1); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 85; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 2: values[2] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__V); if (likely(values[2])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("wiener_like_multi", 0, 9, 10, 2); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 85; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 3: values[3] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__a); if (likely(values[3])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("wiener_like_multi", 0, 9, 10, 3); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 85; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 4: values[4] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__z); if (likely(values[4])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("wiener_like_multi", 0, 9, 10, 4); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 85; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 5: values[5] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__Z); if (likely(values[5])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("wiener_like_multi", 0, 9, 10, 5); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 85; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 6: values[6] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__t); if (likely(values[6])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("wiener_like_multi", 0, 9, 10, 6); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 85; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 7: values[7] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__T); if (likely(values[7])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("wiener_like_multi", 0, 9, 10, 7); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 85; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 8: values[8] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__err); if (likely(values[8])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("wiener_like_multi", 0, 9, 10, 8); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 85; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 9: if (kw_args > 0) { PyObject* value = PyDict_GetItem(__pyx_kwds, __pyx_n_s__multi); if (value) { values[9] = value; kw_args--; } } } if (unlikely(kw_args > 0)) { if (unlikely(__Pyx_ParseOptionalKeywords(__pyx_kwds, __pyx_pyargnames, 0, values, pos_args, "wiener_like_multi") < 0)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 85; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } __pyx_v_x = ((PyArrayObject *)values[0]); __pyx_v_v = values[1]; __pyx_v_V = values[2]; __pyx_v_a = values[3]; __pyx_v_z = values[4]; __pyx_v_Z = values[5]; __pyx_v_t = values[6]; __pyx_v_T = values[7]; __pyx_v_err = __pyx_PyFloat_AsDouble(values[8]); if (unlikely((__pyx_v_err == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 85; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_multi = values[9]; } else { switch (PyTuple_GET_SIZE(__pyx_args)) { case 10: values[9] = PyTuple_GET_ITEM(__pyx_args, 9); case 9: values[8] = PyTuple_GET_ITEM(__pyx_args, 8); values[7] = PyTuple_GET_ITEM(__pyx_args, 7); values[6] = PyTuple_GET_ITEM(__pyx_args, 6); values[5] = PyTuple_GET_ITEM(__pyx_args, 5); values[4] = PyTuple_GET_ITEM(__pyx_args, 4); values[3] = PyTuple_GET_ITEM(__pyx_args, 3); values[2] = PyTuple_GET_ITEM(__pyx_args, 2); values[1] = PyTuple_GET_ITEM(__pyx_args, 1); values[0] = PyTuple_GET_ITEM(__pyx_args, 0); break; default: goto __pyx_L5_argtuple_error; } } __pyx_v_x = ((PyArrayObject *)values[0]); __pyx_v_v = values[1]; __pyx_v_V = values[2]; __pyx_v_a = values[3]; __pyx_v_z = values[4]; __pyx_v_Z = values[5]; __pyx_v_t = values[6]; __pyx_v_T = values[7]; __pyx_v_err = __pyx_PyFloat_AsDouble(values[8]); if (unlikely((__pyx_v_err == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 85; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_multi = values[9]; } goto __pyx_L4_argument_unpacking_done; __pyx_L5_argtuple_error:; __Pyx_RaiseArgtupleInvalid("wiener_like_multi", 0, 9, 10, PyTuple_GET_SIZE(__pyx_args)); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 85; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_L3_error:; __Pyx_AddTraceback("wfpt.wiener_like_multi", __pyx_clineno, __pyx_lineno, __pyx_filename); __Pyx_RefNannyFinishContext(); return NULL; __pyx_L4_argument_unpacking_done:; if (unlikely(!__Pyx_ArgTypeTest(((PyObject *)__pyx_v_x), __pyx_ptype_5numpy_ndarray, 1, "x", 0))) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 85; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __pyx_r = __pyx_pf_4wfpt_8wiener_like_multi(__pyx_self, __pyx_v_x, __pyx_v_v, __pyx_v_V, __pyx_v_a, __pyx_v_z, __pyx_v_Z, __pyx_v_t, __pyx_v_T, __pyx_v_err, __pyx_v_multi); goto __pyx_L0; __pyx_L1_error:; __pyx_L0:; __Pyx_RefNannyFinishContext(); return __pyx_r; } static PyObject *__pyx_pf_4wfpt_8wiener_like_multi(CYTHON_UNUSED PyObject *__pyx_self, PyArrayObject *__pyx_v_x, PyObject *__pyx_v_v, PyObject *__pyx_v_V, PyObject *__pyx_v_a, PyObject *__pyx_v_z, PyObject *__pyx_v_Z, PyObject *__pyx_v_t, PyObject *__pyx_v_T, double __pyx_v_err, PyObject *__pyx_v_multi) { Py_ssize_t __pyx_v_size; Py_ssize_t __pyx_v_i; double __pyx_v_p; PyObject *__pyx_v_params = NULL; PyObject *__pyx_v_params_iter = NULL; PyObject *__pyx_v_param = NULL; __Pyx_LocalBuf_ND __pyx_pybuffernd_x; __Pyx_Buffer __pyx_pybuffer_x; PyObject *__pyx_r = NULL; __Pyx_RefNannyDeclarations int __pyx_t_1; double __pyx_t_2; double __pyx_t_3; double __pyx_t_4; double __pyx_t_5; double __pyx_t_6; double __pyx_t_7; double __pyx_t_8; double __pyx_t_9; PyObject *__pyx_t_10 = NULL; PyObject *__pyx_t_11 = NULL; PyObject *__pyx_t_12 = NULL; Py_ssize_t __pyx_t_13; Py_ssize_t __pyx_t_14; PyObject *(*__pyx_t_15)(PyObject *); int __pyx_lineno = 0; const char *__pyx_filename = NULL; int __pyx_clineno = 0; __Pyx_RefNannySetupContext("wiener_like_multi"); __pyx_pybuffer_x.pybuffer.buf = NULL; __pyx_pybuffer_x.refcount = 0; __pyx_pybuffernd_x.data = NULL; __pyx_pybuffernd_x.rcbuffer = &__pyx_pybuffer_x; { __Pyx_BufFmt_StackElem __pyx_stack[1]; if (unlikely(__Pyx_GetBufferAndValidate(&__pyx_pybuffernd_x.rcbuffer->pybuffer, (PyObject*)__pyx_v_x, &__Pyx_TypeInfo_double, PyBUF_FORMAT| PyBUF_STRIDES, 1, 0, __pyx_stack) == -1)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 85; __pyx_clineno = __LINE__; goto __pyx_L1_error;} } __pyx_pybuffernd_x.diminfo[0].strides = __pyx_pybuffernd_x.rcbuffer->pybuffer.strides[0]; __pyx_pybuffernd_x.diminfo[0].shape = __pyx_pybuffernd_x.rcbuffer->pybuffer.shape[0]; /* "wfpt.pyx":86 * * def wiener_like_multi(np.ndarray[double, ndim=1] x, v, V, a, z, Z, t, T, double err, multi=None): * cdef Py_ssize_t size = x.shape[0] # <<<<<<<<<<<<<< * cdef Py_ssize_t i * cdef double p = 0 */ __pyx_v_size = (__pyx_v_x->dimensions[0]); /* "wfpt.pyx":88 * cdef Py_ssize_t size = x.shape[0] * cdef Py_ssize_t i * cdef double p = 0 # <<<<<<<<<<<<<< * * if multi is None: */ __pyx_v_p = 0.0; /* "wfpt.pyx":90 * cdef double p = 0 * * if multi is None: # <<<<<<<<<<<<<< * return full_pdf(x, v, V, a, z, Z, t, T, err) * else: */ __pyx_t_1 = (__pyx_v_multi == Py_None); if (__pyx_t_1) { /* "wfpt.pyx":91 * * if multi is None: * return full_pdf(x, v, V, a, z, Z, t, T, err) # <<<<<<<<<<<<<< * else: * params = {'v':v, 'z':z, 't':t, 'a':a, 'V':V, 'Z':Z, 'T':T} */ __Pyx_XDECREF(__pyx_r); __pyx_t_2 = __pyx_PyFloat_AsDouble(((PyObject *)__pyx_v_x)); if (unlikely((__pyx_t_2 == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 91; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __pyx_t_3 = __pyx_PyFloat_AsDouble(__pyx_v_v); if (unlikely((__pyx_t_3 == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 91; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __pyx_t_4 = __pyx_PyFloat_AsDouble(__pyx_v_V); if (unlikely((__pyx_t_4 == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 91; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __pyx_t_5 = __pyx_PyFloat_AsDouble(__pyx_v_a); if (unlikely((__pyx_t_5 == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 91; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __pyx_t_6 = __pyx_PyFloat_AsDouble(__pyx_v_z); if (unlikely((__pyx_t_6 == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 91; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __pyx_t_7 = __pyx_PyFloat_AsDouble(__pyx_v_Z); if (unlikely((__pyx_t_7 == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 91; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __pyx_t_8 = __pyx_PyFloat_AsDouble(__pyx_v_t); if (unlikely((__pyx_t_8 == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 91; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __pyx_t_9 = __pyx_PyFloat_AsDouble(__pyx_v_T); if (unlikely((__pyx_t_9 == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 91; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __pyx_t_10 = PyFloat_FromDouble(__pyx_f_4wfpt_full_pdf(__pyx_t_2, __pyx_t_3, __pyx_t_4, __pyx_t_5, __pyx_t_6, __pyx_t_7, __pyx_t_8, __pyx_t_9, __pyx_v_err, 0, NULL)); if (unlikely(!__pyx_t_10)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 91; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_10); __pyx_r = __pyx_t_10; __pyx_t_10 = 0; goto __pyx_L0; goto __pyx_L3; } /*else*/ { /* "wfpt.pyx":93 * return full_pdf(x, v, V, a, z, Z, t, T, err) * else: * params = {'v':v, 'z':z, 't':t, 'a':a, 'V':V, 'Z':Z, 'T':T} # <<<<<<<<<<<<<< * params_iter = copy(params) * for i from 0 <= i < size: */ __pyx_t_10 = PyDict_New(); if (unlikely(!__pyx_t_10)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 93; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(((PyObject *)__pyx_t_10)); if (PyDict_SetItem(__pyx_t_10, ((PyObject *)__pyx_n_s__v), __pyx_v_v) < 0) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 93; __pyx_clineno = __LINE__; goto __pyx_L1_error;} if (PyDict_SetItem(__pyx_t_10, ((PyObject *)__pyx_n_s__z), __pyx_v_z) < 0) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 93; __pyx_clineno = __LINE__; goto __pyx_L1_error;} if (PyDict_SetItem(__pyx_t_10, ((PyObject *)__pyx_n_s__t), __pyx_v_t) < 0) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 93; __pyx_clineno = __LINE__; goto __pyx_L1_error;} if (PyDict_SetItem(__pyx_t_10, ((PyObject *)__pyx_n_s__a), __pyx_v_a) < 0) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 93; __pyx_clineno = __LINE__; goto __pyx_L1_error;} if (PyDict_SetItem(__pyx_t_10, ((PyObject *)__pyx_n_s__V), __pyx_v_V) < 0) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 93; __pyx_clineno = __LINE__; goto __pyx_L1_error;} if (PyDict_SetItem(__pyx_t_10, ((PyObject *)__pyx_n_s__Z), __pyx_v_Z) < 0) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 93; __pyx_clineno = __LINE__; goto __pyx_L1_error;} if (PyDict_SetItem(__pyx_t_10, ((PyObject *)__pyx_n_s__T), __pyx_v_T) < 0) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 93; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __pyx_v_params = __pyx_t_10; __pyx_t_10 = 0; /* "wfpt.pyx":94 * else: * params = {'v':v, 'z':z, 't':t, 'a':a, 'V':V, 'Z':Z, 'T':T} * params_iter = copy(params) # <<<<<<<<<<<<<< * for i from 0 <= i < size: * for param in multi: */ __pyx_t_10 = __Pyx_GetName(__pyx_m, __pyx_n_s__copy); if (unlikely(!__pyx_t_10)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 94; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_10); __pyx_t_11 = PyTuple_New(1); if (unlikely(!__pyx_t_11)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 94; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_11); __Pyx_INCREF(((PyObject *)__pyx_v_params)); PyTuple_SET_ITEM(__pyx_t_11, 0, ((PyObject *)__pyx_v_params)); __Pyx_GIVEREF(((PyObject *)__pyx_v_params)); __pyx_t_12 = PyObject_Call(__pyx_t_10, ((PyObject *)__pyx_t_11), NULL); if (unlikely(!__pyx_t_12)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 94; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_12); __Pyx_DECREF(__pyx_t_10); __pyx_t_10 = 0; __Pyx_DECREF(((PyObject *)__pyx_t_11)); __pyx_t_11 = 0; __pyx_v_params_iter = __pyx_t_12; __pyx_t_12 = 0; /* "wfpt.pyx":95 * params = {'v':v, 'z':z, 't':t, 'a':a, 'V':V, 'Z':Z, 'T':T} * params_iter = copy(params) * for i from 0 <= i < size: # <<<<<<<<<<<<<< * for param in multi: * params_iter[param] = params[param][i] */ __pyx_t_13 = __pyx_v_size; for (__pyx_v_i = 0; __pyx_v_i < __pyx_t_13; __pyx_v_i++) { /* "wfpt.pyx":96 * params_iter = copy(params) * for i from 0 <= i < size: * for param in multi: # <<<<<<<<<<<<<< * params_iter[param] = params[param][i] * */ if (PyList_CheckExact(__pyx_v_multi) || PyTuple_CheckExact(__pyx_v_multi)) { __pyx_t_12 = __pyx_v_multi; __Pyx_INCREF(__pyx_t_12); __pyx_t_14 = 0; __pyx_t_15 = NULL; } else { __pyx_t_14 = -1; __pyx_t_12 = PyObject_GetIter(__pyx_v_multi); if (unlikely(!__pyx_t_12)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 96; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_12); __pyx_t_15 = Py_TYPE(__pyx_t_12)->tp_iternext; } for (;;) { if (!__pyx_t_15 && PyList_CheckExact(__pyx_t_12)) { if (__pyx_t_14 >= PyList_GET_SIZE(__pyx_t_12)) break; __pyx_t_11 = PyList_GET_ITEM(__pyx_t_12, __pyx_t_14); __Pyx_INCREF(__pyx_t_11); __pyx_t_14++; } else if (!__pyx_t_15 && PyTuple_CheckExact(__pyx_t_12)) { if (__pyx_t_14 >= PyTuple_GET_SIZE(__pyx_t_12)) break; __pyx_t_11 = PyTuple_GET_ITEM(__pyx_t_12, __pyx_t_14); __Pyx_INCREF(__pyx_t_11); __pyx_t_14++; } else { __pyx_t_11 = __pyx_t_15(__pyx_t_12); if (unlikely(!__pyx_t_11)) { if (PyErr_Occurred()) { if (likely(PyErr_ExceptionMatches(PyExc_StopIteration))) PyErr_Clear(); else {__pyx_filename = __pyx_f[0]; __pyx_lineno = 96; __pyx_clineno = __LINE__; goto __pyx_L1_error;} } break; } __Pyx_GOTREF(__pyx_t_11); } __Pyx_XDECREF(__pyx_v_param); __pyx_v_param = __pyx_t_11; __pyx_t_11 = 0; /* "wfpt.pyx":97 * for i from 0 <= i < size: * for param in multi: * params_iter[param] = params[param][i] # <<<<<<<<<<<<<< * * p += log(full_pdf(x[i], params_iter['v'], */ __pyx_t_11 = __Pyx_PyDict_GetItem(((PyObject *)__pyx_v_params), __pyx_v_param); if (!__pyx_t_11) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 97; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_11); __pyx_t_10 = __Pyx_GetItemInt(__pyx_t_11, __pyx_v_i, sizeof(Py_ssize_t), PyInt_FromSsize_t); if (!__pyx_t_10) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 97; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_10); __Pyx_DECREF(__pyx_t_11); __pyx_t_11 = 0; if (PyObject_SetItem(__pyx_v_params_iter, __pyx_v_param, __pyx_t_10) < 0) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 97; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_10); __pyx_t_10 = 0; } __Pyx_DECREF(__pyx_t_12); __pyx_t_12 = 0; /* "wfpt.pyx":99 * params_iter[param] = params[param][i] * * p += log(full_pdf(x[i], params_iter['v'], # <<<<<<<<<<<<<< * params_iter['V'], params_iter['a'], params_iter['z'], * params_iter['Z'], params_iter['t'], params_iter['T'], */ __pyx_t_14 = __pyx_v_i; __pyx_t_12 = PyObject_GetItem(__pyx_v_params_iter, ((PyObject *)__pyx_n_s__v)); if (!__pyx_t_12) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 99; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_12); __pyx_t_9 = __pyx_PyFloat_AsDouble(__pyx_t_12); if (unlikely((__pyx_t_9 == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 99; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_12); __pyx_t_12 = 0; /* "wfpt.pyx":100 * * p += log(full_pdf(x[i], params_iter['v'], * params_iter['V'], params_iter['a'], params_iter['z'], # <<<<<<<<<<<<<< * params_iter['Z'], params_iter['t'], params_iter['T'], * err)) */ __pyx_t_12 = PyObject_GetItem(__pyx_v_params_iter, ((PyObject *)__pyx_n_s__V)); if (!__pyx_t_12) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 100; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_12); __pyx_t_8 = __pyx_PyFloat_AsDouble(__pyx_t_12); if (unlikely((__pyx_t_8 == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 100; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_12); __pyx_t_12 = 0; __pyx_t_12 = PyObject_GetItem(__pyx_v_params_iter, ((PyObject *)__pyx_n_s__a)); if (!__pyx_t_12) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 100; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_12); __pyx_t_7 = __pyx_PyFloat_AsDouble(__pyx_t_12); if (unlikely((__pyx_t_7 == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 100; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_12); __pyx_t_12 = 0; __pyx_t_12 = PyObject_GetItem(__pyx_v_params_iter, ((PyObject *)__pyx_n_s__z)); if (!__pyx_t_12) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 100; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_12); __pyx_t_6 = __pyx_PyFloat_AsDouble(__pyx_t_12); if (unlikely((__pyx_t_6 == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 100; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_12); __pyx_t_12 = 0; /* "wfpt.pyx":101 * p += log(full_pdf(x[i], params_iter['v'], * params_iter['V'], params_iter['a'], params_iter['z'], * params_iter['Z'], params_iter['t'], params_iter['T'], # <<<<<<<<<<<<<< * err)) * return p */ __pyx_t_12 = PyObject_GetItem(__pyx_v_params_iter, ((PyObject *)__pyx_n_s__Z)); if (!__pyx_t_12) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 101; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_12); __pyx_t_5 = __pyx_PyFloat_AsDouble(__pyx_t_12); if (unlikely((__pyx_t_5 == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 101; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_12); __pyx_t_12 = 0; __pyx_t_12 = PyObject_GetItem(__pyx_v_params_iter, ((PyObject *)__pyx_n_s__t)); if (!__pyx_t_12) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 101; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_12); __pyx_t_4 = __pyx_PyFloat_AsDouble(__pyx_t_12); if (unlikely((__pyx_t_4 == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 101; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_12); __pyx_t_12 = 0; __pyx_t_12 = PyObject_GetItem(__pyx_v_params_iter, ((PyObject *)__pyx_n_s__T)); if (!__pyx_t_12) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 101; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_12); __pyx_t_3 = __pyx_PyFloat_AsDouble(__pyx_t_12); if (unlikely((__pyx_t_3 == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 101; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_12); __pyx_t_12 = 0; /* "wfpt.pyx":102 * params_iter['V'], params_iter['a'], params_iter['z'], * params_iter['Z'], params_iter['t'], params_iter['T'], * err)) # <<<<<<<<<<<<<< * return p * */ __pyx_v_p = (__pyx_v_p + log(__pyx_f_4wfpt_full_pdf((*__Pyx_BufPtrStrided1d(double *, __pyx_pybuffernd_x.rcbuffer->pybuffer.buf, __pyx_t_14, __pyx_pybuffernd_x.diminfo[0].strides)), __pyx_t_9, __pyx_t_8, __pyx_t_7, __pyx_t_6, __pyx_t_5, __pyx_t_4, __pyx_t_3, __pyx_v_err, 0, NULL))); } /* "wfpt.pyx":103 * params_iter['Z'], params_iter['t'], params_iter['T'], * err)) * return p # <<<<<<<<<<<<<< * * def gen_rts_from_cdf(double v, double V, double a, double z, double Z, double t, \ */ __Pyx_XDECREF(__pyx_r); __pyx_t_12 = PyFloat_FromDouble(__pyx_v_p); if (unlikely(!__pyx_t_12)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 103; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_12); __pyx_r = __pyx_t_12; __pyx_t_12 = 0; goto __pyx_L0; } __pyx_L3:; __pyx_r = Py_None; __Pyx_INCREF(Py_None); goto __pyx_L0; __pyx_L1_error:; __Pyx_XDECREF(__pyx_t_10); __Pyx_XDECREF(__pyx_t_11); __Pyx_XDECREF(__pyx_t_12); { PyObject *__pyx_type, *__pyx_value, *__pyx_tb; __Pyx_ErrFetch(&__pyx_type, &__pyx_value, &__pyx_tb); __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd_x.rcbuffer->pybuffer); __Pyx_ErrRestore(__pyx_type, __pyx_value, __pyx_tb);} __Pyx_AddTraceback("wfpt.wiener_like_multi", __pyx_clineno, __pyx_lineno, __pyx_filename); __pyx_r = NULL; goto __pyx_L2; __pyx_L0:; __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd_x.rcbuffer->pybuffer); __pyx_L2:; __Pyx_XDECREF(__pyx_v_params); __Pyx_XDECREF(__pyx_v_params_iter); __Pyx_XDECREF(__pyx_v_param); __Pyx_XGIVEREF(__pyx_r); __Pyx_RefNannyFinishContext(); return __pyx_r; } /* Python wrapper */ static PyObject *__pyx_pw_4wfpt_11gen_rts_from_cdf(PyObject *__pyx_self, PyObject *__pyx_args, PyObject *__pyx_kwds); /*proto*/ static char __pyx_doc_4wfpt_10gen_rts_from_cdf[] = "gen_rts_from_cdf(double v, double V, double a, double z, double Z, double t, double T, int samples=1000, double cdf_lb=-6, double cdf_ub=6, double dt=0.01)"; static PyMethodDef __pyx_mdef_4wfpt_11gen_rts_from_cdf = {__Pyx_NAMESTR("gen_rts_from_cdf"), (PyCFunction)__pyx_pw_4wfpt_11gen_rts_from_cdf, METH_VARARGS|METH_KEYWORDS, __Pyx_DOCSTR(__pyx_doc_4wfpt_10gen_rts_from_cdf)}; static PyObject *__pyx_pw_4wfpt_11gen_rts_from_cdf(PyObject *__pyx_self, PyObject *__pyx_args, PyObject *__pyx_kwds) { double __pyx_v_v; double __pyx_v_V; double __pyx_v_a; double __pyx_v_z; double __pyx_v_Z; double __pyx_v_t; double __pyx_v_T; int __pyx_v_samples; double __pyx_v_cdf_lb; double __pyx_v_cdf_ub; double __pyx_v_dt; static PyObject **__pyx_pyargnames[] = {&__pyx_n_s__v,&__pyx_n_s__V,&__pyx_n_s__a,&__pyx_n_s__z,&__pyx_n_s__Z,&__pyx_n_s__t,&__pyx_n_s__T,&__pyx_n_s__samples,&__pyx_n_s__cdf_lb,&__pyx_n_s__cdf_ub,&__pyx_n_s__dt,0}; PyObject *__pyx_r = 0; __Pyx_RefNannyDeclarations __Pyx_RefNannySetupContext("gen_rts_from_cdf (wrapper)"); __pyx_self = __pyx_self; { PyObject* values[11] = {0,0,0,0,0,0,0,0,0,0,0}; if (unlikely(__pyx_kwds)) { Py_ssize_t kw_args; const Py_ssize_t pos_args = PyTuple_GET_SIZE(__pyx_args); switch (pos_args) { case 11: values[10] = PyTuple_GET_ITEM(__pyx_args, 10); case 10: values[9] = PyTuple_GET_ITEM(__pyx_args, 9); case 9: values[8] = PyTuple_GET_ITEM(__pyx_args, 8); case 8: values[7] = PyTuple_GET_ITEM(__pyx_args, 7); case 7: values[6] = PyTuple_GET_ITEM(__pyx_args, 6); case 6: values[5] = PyTuple_GET_ITEM(__pyx_args, 5); case 5: values[4] = PyTuple_GET_ITEM(__pyx_args, 4); case 4: values[3] = PyTuple_GET_ITEM(__pyx_args, 3); case 3: values[2] = PyTuple_GET_ITEM(__pyx_args, 2); case 2: values[1] = PyTuple_GET_ITEM(__pyx_args, 1); case 1: values[0] = PyTuple_GET_ITEM(__pyx_args, 0); case 0: break; default: goto __pyx_L5_argtuple_error; } kw_args = PyDict_Size(__pyx_kwds); switch (pos_args) { case 0: values[0] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__v); if (likely(values[0])) kw_args--; else goto __pyx_L5_argtuple_error; case 1: values[1] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__V); if (likely(values[1])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("gen_rts_from_cdf", 0, 7, 11, 1); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 105; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 2: values[2] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__a); if (likely(values[2])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("gen_rts_from_cdf", 0, 7, 11, 2); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 105; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 3: values[3] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__z); if (likely(values[3])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("gen_rts_from_cdf", 0, 7, 11, 3); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 105; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 4: values[4] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__Z); if (likely(values[4])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("gen_rts_from_cdf", 0, 7, 11, 4); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 105; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 5: values[5] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__t); if (likely(values[5])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("gen_rts_from_cdf", 0, 7, 11, 5); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 105; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 6: values[6] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__T); if (likely(values[6])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("gen_rts_from_cdf", 0, 7, 11, 6); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 105; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 7: if (kw_args > 0) { PyObject* value = PyDict_GetItem(__pyx_kwds, __pyx_n_s__samples); if (value) { values[7] = value; kw_args--; } } case 8: if (kw_args > 0) { PyObject* value = PyDict_GetItem(__pyx_kwds, __pyx_n_s__cdf_lb); if (value) { values[8] = value; kw_args--; } } case 9: if (kw_args > 0) { PyObject* value = PyDict_GetItem(__pyx_kwds, __pyx_n_s__cdf_ub); if (value) { values[9] = value; kw_args--; } } case 10: if (kw_args > 0) { PyObject* value = PyDict_GetItem(__pyx_kwds, __pyx_n_s__dt); if (value) { values[10] = value; kw_args--; } } } if (unlikely(kw_args > 0)) { if (unlikely(__Pyx_ParseOptionalKeywords(__pyx_kwds, __pyx_pyargnames, 0, values, pos_args, "gen_rts_from_cdf") < 0)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 105; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } __pyx_v_v = __pyx_PyFloat_AsDouble(values[0]); if (unlikely((__pyx_v_v == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 105; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_V = __pyx_PyFloat_AsDouble(values[1]); if (unlikely((__pyx_v_V == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 105; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_a = __pyx_PyFloat_AsDouble(values[2]); if (unlikely((__pyx_v_a == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 105; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_z = __pyx_PyFloat_AsDouble(values[3]); if (unlikely((__pyx_v_z == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 105; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_Z = __pyx_PyFloat_AsDouble(values[4]); if (unlikely((__pyx_v_Z == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 105; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_t = __pyx_PyFloat_AsDouble(values[5]); if (unlikely((__pyx_v_t == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 105; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_T = __pyx_PyFloat_AsDouble(values[6]); if (unlikely((__pyx_v_T == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 106; __pyx_clineno = __LINE__; goto __pyx_L3_error;} if (values[7]) { __pyx_v_samples = __Pyx_PyInt_AsInt(values[7]); if (unlikely((__pyx_v_samples == (int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 106; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } else { __pyx_v_samples = ((int)1000); } if (values[8]) { __pyx_v_cdf_lb = __pyx_PyFloat_AsDouble(values[8]); if (unlikely((__pyx_v_cdf_lb == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 106; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } else { /* "wfpt.pyx":106 * * def gen_rts_from_cdf(double v, double V, double a, double z, double Z, double t, \ * double T, int samples=1000, double cdf_lb=-6, double cdf_ub=6, double dt=1e-2): # <<<<<<<<<<<<<< * * cdef np.ndarray[double, ndim=1] x = np.arange(cdf_lb, cdf_ub, dt) */ __pyx_v_cdf_lb = ((double)-6.0); } if (values[9]) { __pyx_v_cdf_ub = __pyx_PyFloat_AsDouble(values[9]); if (unlikely((__pyx_v_cdf_ub == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 106; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } else { __pyx_v_cdf_ub = ((double)6.0); } if (values[10]) { __pyx_v_dt = __pyx_PyFloat_AsDouble(values[10]); if (unlikely((__pyx_v_dt == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 106; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } else { __pyx_v_dt = ((double)1e-2); } } else { switch (PyTuple_GET_SIZE(__pyx_args)) { case 11: values[10] = PyTuple_GET_ITEM(__pyx_args, 10); case 10: values[9] = PyTuple_GET_ITEM(__pyx_args, 9); case 9: values[8] = PyTuple_GET_ITEM(__pyx_args, 8); case 8: values[7] = PyTuple_GET_ITEM(__pyx_args, 7); case 7: values[6] = PyTuple_GET_ITEM(__pyx_args, 6); values[5] = PyTuple_GET_ITEM(__pyx_args, 5); values[4] = PyTuple_GET_ITEM(__pyx_args, 4); values[3] = PyTuple_GET_ITEM(__pyx_args, 3); values[2] = PyTuple_GET_ITEM(__pyx_args, 2); values[1] = PyTuple_GET_ITEM(__pyx_args, 1); values[0] = PyTuple_GET_ITEM(__pyx_args, 0); break; default: goto __pyx_L5_argtuple_error; } } __pyx_v_v = __pyx_PyFloat_AsDouble(values[0]); if (unlikely((__pyx_v_v == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 105; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_V = __pyx_PyFloat_AsDouble(values[1]); if (unlikely((__pyx_v_V == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 105; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_a = __pyx_PyFloat_AsDouble(values[2]); if (unlikely((__pyx_v_a == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 105; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_z = __pyx_PyFloat_AsDouble(values[3]); if (unlikely((__pyx_v_z == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 105; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_Z = __pyx_PyFloat_AsDouble(values[4]); if (unlikely((__pyx_v_Z == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 105; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_t = __pyx_PyFloat_AsDouble(values[5]); if (unlikely((__pyx_v_t == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 105; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_T = __pyx_PyFloat_AsDouble(values[6]); if (unlikely((__pyx_v_T == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 106; __pyx_clineno = __LINE__; goto __pyx_L3_error;} if (values[7]) { __pyx_v_samples = __Pyx_PyInt_AsInt(values[7]); if (unlikely((__pyx_v_samples == (int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 106; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } else { __pyx_v_samples = ((int)1000); } if (values[8]) { __pyx_v_cdf_lb = __pyx_PyFloat_AsDouble(values[8]); if (unlikely((__pyx_v_cdf_lb == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 106; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } else { __pyx_v_cdf_lb = ((double)-6.0); } if (values[9]) { __pyx_v_cdf_ub = __pyx_PyFloat_AsDouble(values[9]); if (unlikely((__pyx_v_cdf_ub == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 106; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } else { __pyx_v_cdf_ub = ((double)6.0); } if (values[10]) { __pyx_v_dt = __pyx_PyFloat_AsDouble(values[10]); if (unlikely((__pyx_v_dt == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 106; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } else { __pyx_v_dt = ((double)1e-2); } } goto __pyx_L4_argument_unpacking_done; __pyx_L5_argtuple_error:; __Pyx_RaiseArgtupleInvalid("gen_rts_from_cdf", 0, 7, 11, PyTuple_GET_SIZE(__pyx_args)); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 105; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_L3_error:; __Pyx_AddTraceback("wfpt.gen_rts_from_cdf", __pyx_clineno, __pyx_lineno, __pyx_filename); __Pyx_RefNannyFinishContext(); return NULL; __pyx_L4_argument_unpacking_done:; __pyx_r = __pyx_pf_4wfpt_10gen_rts_from_cdf(__pyx_self, __pyx_v_v, __pyx_v_V, __pyx_v_a, __pyx_v_z, __pyx_v_Z, __pyx_v_t, __pyx_v_T, __pyx_v_samples, __pyx_v_cdf_lb, __pyx_v_cdf_ub, __pyx_v_dt); __Pyx_RefNannyFinishContext(); return __pyx_r; } /* "wfpt.pyx":105 * return p * * def gen_rts_from_cdf(double v, double V, double a, double z, double Z, double t, \ # <<<<<<<<<<<<<< * double T, int samples=1000, double cdf_lb=-6, double cdf_ub=6, double dt=1e-2): * */ static PyObject *__pyx_pf_4wfpt_10gen_rts_from_cdf(CYTHON_UNUSED PyObject *__pyx_self, double __pyx_v_v, double __pyx_v_V, double __pyx_v_a, double __pyx_v_z, double __pyx_v_Z, double __pyx_v_t, double __pyx_v_T, int __pyx_v_samples, double __pyx_v_cdf_lb, double __pyx_v_cdf_ub, double __pyx_v_dt) { PyArrayObject *__pyx_v_x = 0; PyArrayObject *__pyx_v_l_cdf = 0; double __pyx_v_pdf; double __pyx_v_rt; Py_ssize_t __pyx_v_size; Py_ssize_t __pyx_v_i; int __pyx_v_idx; PyArrayObject *__pyx_v_rts = 0; PyArrayObject *__pyx_v_f = 0; PyArrayObject *__pyx_v_delay = 0; __Pyx_LocalBuf_ND __pyx_pybuffernd_rts; __Pyx_Buffer __pyx_pybuffer_rts; __Pyx_LocalBuf_ND __pyx_pybuffernd_x; __Pyx_Buffer __pyx_pybuffer_x; __Pyx_LocalBuf_ND __pyx_pybuffernd_delay; __Pyx_Buffer __pyx_pybuffer_delay; __Pyx_LocalBuf_ND __pyx_pybuffernd_l_cdf; __Pyx_Buffer __pyx_pybuffer_l_cdf; __Pyx_LocalBuf_ND __pyx_pybuffernd_f; __Pyx_Buffer __pyx_pybuffer_f; PyObject *__pyx_r = NULL; __Pyx_RefNannyDeclarations PyObject *__pyx_t_1 = NULL; PyObject *__pyx_t_2 = NULL; PyObject *__pyx_t_3 = NULL; PyObject *__pyx_t_4 = NULL; PyObject *__pyx_t_5 = NULL; PyArrayObject *__pyx_t_6 = NULL; PyArrayObject *__pyx_t_7 = NULL; long __pyx_t_8; Py_ssize_t __pyx_t_9; Py_ssize_t __pyx_t_10; Py_ssize_t __pyx_t_11; Py_ssize_t __pyx_t_12; Py_ssize_t __pyx_t_13; long __pyx_t_14; int __pyx_t_15; PyObject *__pyx_t_16 = NULL; PyObject *__pyx_t_17 = NULL; PyObject *__pyx_t_18 = NULL; PyArrayObject *__pyx_t_19 = NULL; PyArrayObject *__pyx_t_20 = NULL; int __pyx_t_21; PyArrayObject *__pyx_t_22 = NULL; int __pyx_t_23; double __pyx_t_24; Py_ssize_t __pyx_t_25; int __pyx_lineno = 0; const char *__pyx_filename = NULL; int __pyx_clineno = 0; __Pyx_RefNannySetupContext("gen_rts_from_cdf"); __pyx_pybuffer_x.pybuffer.buf = NULL; __pyx_pybuffer_x.refcount = 0; __pyx_pybuffernd_x.data = NULL; __pyx_pybuffernd_x.rcbuffer = &__pyx_pybuffer_x; __pyx_pybuffer_l_cdf.pybuffer.buf = NULL; __pyx_pybuffer_l_cdf.refcount = 0; __pyx_pybuffernd_l_cdf.data = NULL; __pyx_pybuffernd_l_cdf.rcbuffer = &__pyx_pybuffer_l_cdf; __pyx_pybuffer_rts.pybuffer.buf = NULL; __pyx_pybuffer_rts.refcount = 0; __pyx_pybuffernd_rts.data = NULL; __pyx_pybuffernd_rts.rcbuffer = &__pyx_pybuffer_rts; __pyx_pybuffer_f.pybuffer.buf = NULL; __pyx_pybuffer_f.refcount = 0; __pyx_pybuffernd_f.data = NULL; __pyx_pybuffernd_f.rcbuffer = &__pyx_pybuffer_f; __pyx_pybuffer_delay.pybuffer.buf = NULL; __pyx_pybuffer_delay.refcount = 0; __pyx_pybuffernd_delay.data = NULL; __pyx_pybuffernd_delay.rcbuffer = &__pyx_pybuffer_delay; /* "wfpt.pyx":108 * double T, int samples=1000, double cdf_lb=-6, double cdf_ub=6, double dt=1e-2): * * cdef np.ndarray[double, ndim=1] x = np.arange(cdf_lb, cdf_ub, dt) # <<<<<<<<<<<<<< * cdef np.ndarray[double, ndim=1] l_cdf = np.empty(x.shape[0], dtype=np.double) * cdef double pdf, rt */ __pyx_t_1 = __Pyx_GetName(__pyx_m, __pyx_n_s__np); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 108; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_1); __pyx_t_2 = PyObject_GetAttr(__pyx_t_1, __pyx_n_s__arange); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 108; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_2); __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; __pyx_t_1 = PyFloat_FromDouble(__pyx_v_cdf_lb); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 108; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_1); __pyx_t_3 = PyFloat_FromDouble(__pyx_v_cdf_ub); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 108; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); __pyx_t_4 = PyFloat_FromDouble(__pyx_v_dt); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 108; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_4); __pyx_t_5 = PyTuple_New(3); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 108; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_5); PyTuple_SET_ITEM(__pyx_t_5, 0, __pyx_t_1); __Pyx_GIVEREF(__pyx_t_1); PyTuple_SET_ITEM(__pyx_t_5, 1, __pyx_t_3); __Pyx_GIVEREF(__pyx_t_3); PyTuple_SET_ITEM(__pyx_t_5, 2, __pyx_t_4); __Pyx_GIVEREF(__pyx_t_4); __pyx_t_1 = 0; __pyx_t_3 = 0; __pyx_t_4 = 0; __pyx_t_4 = PyObject_Call(__pyx_t_2, ((PyObject *)__pyx_t_5), NULL); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 108; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_4); __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; __Pyx_DECREF(((PyObject *)__pyx_t_5)); __pyx_t_5 = 0; if (!(likely(((__pyx_t_4) == Py_None) || likely(__Pyx_TypeTest(__pyx_t_4, __pyx_ptype_5numpy_ndarray))))) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 108; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __pyx_t_6 = ((PyArrayObject *)__pyx_t_4); { __Pyx_BufFmt_StackElem __pyx_stack[1]; if (unlikely(__Pyx_GetBufferAndValidate(&__pyx_pybuffernd_x.rcbuffer->pybuffer, (PyObject*)__pyx_t_6, &__Pyx_TypeInfo_double, PyBUF_FORMAT| PyBUF_STRIDES, 1, 0, __pyx_stack) == -1)) { __pyx_v_x = ((PyArrayObject *)Py_None); __Pyx_INCREF(Py_None); __pyx_pybuffernd_x.rcbuffer->pybuffer.buf = NULL; {__pyx_filename = __pyx_f[0]; __pyx_lineno = 108; __pyx_clineno = __LINE__; goto __pyx_L1_error;} } else {__pyx_pybuffernd_x.diminfo[0].strides = __pyx_pybuffernd_x.rcbuffer->pybuffer.strides[0]; __pyx_pybuffernd_x.diminfo[0].shape = __pyx_pybuffernd_x.rcbuffer->pybuffer.shape[0]; } } __pyx_t_6 = 0; __pyx_v_x = ((PyArrayObject *)__pyx_t_4); __pyx_t_4 = 0; /* "wfpt.pyx":109 * * cdef np.ndarray[double, ndim=1] x = np.arange(cdf_lb, cdf_ub, dt) * cdef np.ndarray[double, ndim=1] l_cdf = np.empty(x.shape[0], dtype=np.double) # <<<<<<<<<<<<<< * cdef double pdf, rt * cdef Py_ssize_t size = x.shape[0] */ __pyx_t_4 = __Pyx_GetName(__pyx_m, __pyx_n_s__np); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 109; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_4); __pyx_t_5 = PyObject_GetAttr(__pyx_t_4, __pyx_n_s__empty); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 109; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_5); __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0; __pyx_t_4 = __Pyx_PyInt_to_py_Py_intptr_t((__pyx_v_x->dimensions[0])); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 109; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_4); __pyx_t_2 = PyTuple_New(1); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 109; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_2); PyTuple_SET_ITEM(__pyx_t_2, 0, __pyx_t_4); __Pyx_GIVEREF(__pyx_t_4); __pyx_t_4 = 0; __pyx_t_4 = PyDict_New(); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 109; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(((PyObject *)__pyx_t_4)); __pyx_t_3 = __Pyx_GetName(__pyx_m, __pyx_n_s__np); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 109; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); __pyx_t_1 = PyObject_GetAttr(__pyx_t_3, __pyx_n_s__double); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 109; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_1); __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; if (PyDict_SetItem(__pyx_t_4, ((PyObject *)__pyx_n_s__dtype), __pyx_t_1) < 0) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 109; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; __pyx_t_1 = PyObject_Call(__pyx_t_5, ((PyObject *)__pyx_t_2), ((PyObject *)__pyx_t_4)); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 109; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_1); __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0; __Pyx_DECREF(((PyObject *)__pyx_t_2)); __pyx_t_2 = 0; __Pyx_DECREF(((PyObject *)__pyx_t_4)); __pyx_t_4 = 0; if (!(likely(((__pyx_t_1) == Py_None) || likely(__Pyx_TypeTest(__pyx_t_1, __pyx_ptype_5numpy_ndarray))))) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 109; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __pyx_t_7 = ((PyArrayObject *)__pyx_t_1); { __Pyx_BufFmt_StackElem __pyx_stack[1]; if (unlikely(__Pyx_GetBufferAndValidate(&__pyx_pybuffernd_l_cdf.rcbuffer->pybuffer, (PyObject*)__pyx_t_7, &__Pyx_TypeInfo_double, PyBUF_FORMAT| PyBUF_STRIDES| PyBUF_WRITABLE, 1, 0, __pyx_stack) == -1)) { __pyx_v_l_cdf = ((PyArrayObject *)Py_None); __Pyx_INCREF(Py_None); __pyx_pybuffernd_l_cdf.rcbuffer->pybuffer.buf = NULL; {__pyx_filename = __pyx_f[0]; __pyx_lineno = 109; __pyx_clineno = __LINE__; goto __pyx_L1_error;} } else {__pyx_pybuffernd_l_cdf.diminfo[0].strides = __pyx_pybuffernd_l_cdf.rcbuffer->pybuffer.strides[0]; __pyx_pybuffernd_l_cdf.diminfo[0].shape = __pyx_pybuffernd_l_cdf.rcbuffer->pybuffer.shape[0]; } } __pyx_t_7 = 0; __pyx_v_l_cdf = ((PyArrayObject *)__pyx_t_1); __pyx_t_1 = 0; /* "wfpt.pyx":111 * cdef np.ndarray[double, ndim=1] l_cdf = np.empty(x.shape[0], dtype=np.double) * cdef double pdf, rt * cdef Py_ssize_t size = x.shape[0] # <<<<<<<<<<<<<< * cdef Py_ssize_t i, j * cdef int idx */ __pyx_v_size = (__pyx_v_x->dimensions[0]); /* "wfpt.pyx":115 * cdef int idx * * l_cdf[0] = 0 # <<<<<<<<<<<<<< * for i in range(size): * pdf = full_pdf(x[i], v, V, a, z, Z, 0, 0, 1e-4) */ __pyx_t_8 = 0; *__Pyx_BufPtrStrided1d(double *, __pyx_pybuffernd_l_cdf.rcbuffer->pybuffer.buf, __pyx_t_8, __pyx_pybuffernd_l_cdf.diminfo[0].strides) = 0.0; /* "wfpt.pyx":116 * * l_cdf[0] = 0 * for i in range(size): # <<<<<<<<<<<<<< * pdf = full_pdf(x[i], v, V, a, z, Z, 0, 0, 1e-4) * l_cdf[i] = l_cdf[i-1] + pdf */ __pyx_t_9 = __pyx_v_size; for (__pyx_t_10 = 0; __pyx_t_10 < __pyx_t_9; __pyx_t_10+=1) { __pyx_v_i = __pyx_t_10; /* "wfpt.pyx":117 * l_cdf[0] = 0 * for i in range(size): * pdf = full_pdf(x[i], v, V, a, z, Z, 0, 0, 1e-4) # <<<<<<<<<<<<<< * l_cdf[i] = l_cdf[i-1] + pdf * */ __pyx_t_11 = __pyx_v_i; __pyx_v_pdf = __pyx_f_4wfpt_full_pdf((*__Pyx_BufPtrStrided1d(double *, __pyx_pybuffernd_x.rcbuffer->pybuffer.buf, __pyx_t_11, __pyx_pybuffernd_x.diminfo[0].strides)), __pyx_v_v, __pyx_v_V, __pyx_v_a, __pyx_v_z, __pyx_v_Z, 0.0, 0.0, 1e-4, 0, NULL); /* "wfpt.pyx":118 * for i in range(size): * pdf = full_pdf(x[i], v, V, a, z, Z, 0, 0, 1e-4) * l_cdf[i] = l_cdf[i-1] + pdf # <<<<<<<<<<<<<< * * l_cdf /= l_cdf[x.shape[0]-1] */ __pyx_t_12 = (__pyx_v_i - 1); __pyx_t_13 = __pyx_v_i; *__Pyx_BufPtrStrided1d(double *, __pyx_pybuffernd_l_cdf.rcbuffer->pybuffer.buf, __pyx_t_13, __pyx_pybuffernd_l_cdf.diminfo[0].strides) = ((*__Pyx_BufPtrStrided1d(double *, __pyx_pybuffernd_l_cdf.rcbuffer->pybuffer.buf, __pyx_t_12, __pyx_pybuffernd_l_cdf.diminfo[0].strides)) + __pyx_v_pdf); } /* "wfpt.pyx":120 * l_cdf[i] = l_cdf[i-1] + pdf * * l_cdf /= l_cdf[x.shape[0]-1] # <<<<<<<<<<<<<< * * cdef np.ndarray[double, ndim=1] rts = np.empty(samples, dtype=np.double) */ __pyx_t_14 = ((__pyx_v_x->dimensions[0]) - 1); __pyx_t_1 = PyFloat_FromDouble((*__Pyx_BufPtrStrided1d(double *, __pyx_pybuffernd_l_cdf.rcbuffer->pybuffer.buf, __pyx_t_14, __pyx_pybuffernd_l_cdf.diminfo[0].strides))); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 120; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_1); __pyx_t_4 = __Pyx_PyNumber_InPlaceDivide(((PyObject *)__pyx_v_l_cdf), __pyx_t_1); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 120; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_4); __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; if (!(likely(((__pyx_t_4) == Py_None) || likely(__Pyx_TypeTest(__pyx_t_4, __pyx_ptype_5numpy_ndarray))))) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 120; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __pyx_t_7 = ((PyArrayObject *)__pyx_t_4); { __Pyx_BufFmt_StackElem __pyx_stack[1]; __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd_l_cdf.rcbuffer->pybuffer); __pyx_t_15 = __Pyx_GetBufferAndValidate(&__pyx_pybuffernd_l_cdf.rcbuffer->pybuffer, (PyObject*)__pyx_t_7, &__Pyx_TypeInfo_double, PyBUF_FORMAT| PyBUF_STRIDES| PyBUF_WRITABLE, 1, 0, __pyx_stack); if (unlikely(__pyx_t_15 < 0)) { PyErr_Fetch(&__pyx_t_16, &__pyx_t_17, &__pyx_t_18); if (unlikely(__Pyx_GetBufferAndValidate(&__pyx_pybuffernd_l_cdf.rcbuffer->pybuffer, (PyObject*)__pyx_v_l_cdf, &__Pyx_TypeInfo_double, PyBUF_FORMAT| PyBUF_STRIDES| PyBUF_WRITABLE, 1, 0, __pyx_stack) == -1)) { Py_XDECREF(__pyx_t_16); Py_XDECREF(__pyx_t_17); Py_XDECREF(__pyx_t_18); __Pyx_RaiseBufferFallbackError(); } else { PyErr_Restore(__pyx_t_16, __pyx_t_17, __pyx_t_18); } } __pyx_pybuffernd_l_cdf.diminfo[0].strides = __pyx_pybuffernd_l_cdf.rcbuffer->pybuffer.strides[0]; __pyx_pybuffernd_l_cdf.diminfo[0].shape = __pyx_pybuffernd_l_cdf.rcbuffer->pybuffer.shape[0]; if (unlikely(__pyx_t_15 < 0)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 120; __pyx_clineno = __LINE__; goto __pyx_L1_error;} } __pyx_t_7 = 0; __Pyx_DECREF(((PyObject *)__pyx_v_l_cdf)); __pyx_v_l_cdf = ((PyArrayObject *)__pyx_t_4); __pyx_t_4 = 0; /* "wfpt.pyx":122 * l_cdf /= l_cdf[x.shape[0]-1] * * cdef np.ndarray[double, ndim=1] rts = np.empty(samples, dtype=np.double) # <<<<<<<<<<<<<< * cdef np.ndarray[double, ndim=1] f = np.random.rand(samples) * cdef np.ndarray[double, ndim=1] delay */ __pyx_t_4 = __Pyx_GetName(__pyx_m, __pyx_n_s__np); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 122; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_4); __pyx_t_1 = PyObject_GetAttr(__pyx_t_4, __pyx_n_s__empty); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 122; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_1); __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0; __pyx_t_4 = PyInt_FromLong(__pyx_v_samples); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 122; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_4); __pyx_t_2 = PyTuple_New(1); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 122; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_2); PyTuple_SET_ITEM(__pyx_t_2, 0, __pyx_t_4); __Pyx_GIVEREF(__pyx_t_4); __pyx_t_4 = 0; __pyx_t_4 = PyDict_New(); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 122; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(((PyObject *)__pyx_t_4)); __pyx_t_5 = __Pyx_GetName(__pyx_m, __pyx_n_s__np); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 122; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_5); __pyx_t_3 = PyObject_GetAttr(__pyx_t_5, __pyx_n_s__double); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 122; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0; if (PyDict_SetItem(__pyx_t_4, ((PyObject *)__pyx_n_s__dtype), __pyx_t_3) < 0) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 122; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; __pyx_t_3 = PyObject_Call(__pyx_t_1, ((PyObject *)__pyx_t_2), ((PyObject *)__pyx_t_4)); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 122; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; __Pyx_DECREF(((PyObject *)__pyx_t_2)); __pyx_t_2 = 0; __Pyx_DECREF(((PyObject *)__pyx_t_4)); __pyx_t_4 = 0; if (!(likely(((__pyx_t_3) == Py_None) || likely(__Pyx_TypeTest(__pyx_t_3, __pyx_ptype_5numpy_ndarray))))) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 122; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __pyx_t_19 = ((PyArrayObject *)__pyx_t_3); { __Pyx_BufFmt_StackElem __pyx_stack[1]; if (unlikely(__Pyx_GetBufferAndValidate(&__pyx_pybuffernd_rts.rcbuffer->pybuffer, (PyObject*)__pyx_t_19, &__Pyx_TypeInfo_double, PyBUF_FORMAT| PyBUF_STRIDES| PyBUF_WRITABLE, 1, 0, __pyx_stack) == -1)) { __pyx_v_rts = ((PyArrayObject *)Py_None); __Pyx_INCREF(Py_None); __pyx_pybuffernd_rts.rcbuffer->pybuffer.buf = NULL; {__pyx_filename = __pyx_f[0]; __pyx_lineno = 122; __pyx_clineno = __LINE__; goto __pyx_L1_error;} } else {__pyx_pybuffernd_rts.diminfo[0].strides = __pyx_pybuffernd_rts.rcbuffer->pybuffer.strides[0]; __pyx_pybuffernd_rts.diminfo[0].shape = __pyx_pybuffernd_rts.rcbuffer->pybuffer.shape[0]; } } __pyx_t_19 = 0; __pyx_v_rts = ((PyArrayObject *)__pyx_t_3); __pyx_t_3 = 0; /* "wfpt.pyx":123 * * cdef np.ndarray[double, ndim=1] rts = np.empty(samples, dtype=np.double) * cdef np.ndarray[double, ndim=1] f = np.random.rand(samples) # <<<<<<<<<<<<<< * cdef np.ndarray[double, ndim=1] delay * */ __pyx_t_3 = __Pyx_GetName(__pyx_m, __pyx_n_s__np); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 123; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); __pyx_t_4 = PyObject_GetAttr(__pyx_t_3, __pyx_n_s__random); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 123; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_4); __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; __pyx_t_3 = PyObject_GetAttr(__pyx_t_4, __pyx_n_s__rand); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 123; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0; __pyx_t_4 = PyInt_FromLong(__pyx_v_samples); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 123; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_4); __pyx_t_2 = PyTuple_New(1); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 123; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_2); PyTuple_SET_ITEM(__pyx_t_2, 0, __pyx_t_4); __Pyx_GIVEREF(__pyx_t_4); __pyx_t_4 = 0; __pyx_t_4 = PyObject_Call(__pyx_t_3, ((PyObject *)__pyx_t_2), NULL); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 123; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_4); __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; __Pyx_DECREF(((PyObject *)__pyx_t_2)); __pyx_t_2 = 0; if (!(likely(((__pyx_t_4) == Py_None) || likely(__Pyx_TypeTest(__pyx_t_4, __pyx_ptype_5numpy_ndarray))))) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 123; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __pyx_t_20 = ((PyArrayObject *)__pyx_t_4); { __Pyx_BufFmt_StackElem __pyx_stack[1]; if (unlikely(__Pyx_GetBufferAndValidate(&__pyx_pybuffernd_f.rcbuffer->pybuffer, (PyObject*)__pyx_t_20, &__Pyx_TypeInfo_double, PyBUF_FORMAT| PyBUF_STRIDES, 1, 0, __pyx_stack) == -1)) { __pyx_v_f = ((PyArrayObject *)Py_None); __Pyx_INCREF(Py_None); __pyx_pybuffernd_f.rcbuffer->pybuffer.buf = NULL; {__pyx_filename = __pyx_f[0]; __pyx_lineno = 123; __pyx_clineno = __LINE__; goto __pyx_L1_error;} } else {__pyx_pybuffernd_f.diminfo[0].strides = __pyx_pybuffernd_f.rcbuffer->pybuffer.strides[0]; __pyx_pybuffernd_f.diminfo[0].shape = __pyx_pybuffernd_f.rcbuffer->pybuffer.shape[0]; } } __pyx_t_20 = 0; __pyx_v_f = ((PyArrayObject *)__pyx_t_4); __pyx_t_4 = 0; /* "wfpt.pyx":126 * cdef np.ndarray[double, ndim=1] delay * * if T!=0: # <<<<<<<<<<<<<< * delay = (np.random.rand(samples)*T + (t - T/2.)) * for i from 0 <= i < samples: */ __pyx_t_21 = (__pyx_v_T != 0.0); if (__pyx_t_21) { /* "wfpt.pyx":127 * * if T!=0: * delay = (np.random.rand(samples)*T + (t - T/2.)) # <<<<<<<<<<<<<< * for i from 0 <= i < samples: * idx = np.searchsorted(l_cdf, f[i]) */ __pyx_t_4 = __Pyx_GetName(__pyx_m, __pyx_n_s__np); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 127; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_4); __pyx_t_2 = PyObject_GetAttr(__pyx_t_4, __pyx_n_s__random); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 127; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_2); __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0; __pyx_t_4 = PyObject_GetAttr(__pyx_t_2, __pyx_n_s__rand); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 127; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_4); __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; __pyx_t_2 = PyInt_FromLong(__pyx_v_samples); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 127; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_2); __pyx_t_3 = PyTuple_New(1); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 127; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); PyTuple_SET_ITEM(__pyx_t_3, 0, __pyx_t_2); __Pyx_GIVEREF(__pyx_t_2); __pyx_t_2 = 0; __pyx_t_2 = PyObject_Call(__pyx_t_4, ((PyObject *)__pyx_t_3), NULL); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 127; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_2); __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0; __Pyx_DECREF(((PyObject *)__pyx_t_3)); __pyx_t_3 = 0; __pyx_t_3 = PyFloat_FromDouble(__pyx_v_T); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 127; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); __pyx_t_4 = PyNumber_Multiply(__pyx_t_2, __pyx_t_3); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 127; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_4); __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; __pyx_t_3 = PyFloat_FromDouble((__pyx_v_t - (__pyx_v_T / 2.))); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 127; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); __pyx_t_2 = PyNumber_Add(__pyx_t_4, __pyx_t_3); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 127; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_2); __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0; __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; if (!(likely(((__pyx_t_2) == Py_None) || likely(__Pyx_TypeTest(__pyx_t_2, __pyx_ptype_5numpy_ndarray))))) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 127; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __pyx_t_22 = ((PyArrayObject *)__pyx_t_2); { __Pyx_BufFmt_StackElem __pyx_stack[1]; __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd_delay.rcbuffer->pybuffer); __pyx_t_15 = __Pyx_GetBufferAndValidate(&__pyx_pybuffernd_delay.rcbuffer->pybuffer, (PyObject*)__pyx_t_22, &__Pyx_TypeInfo_double, PyBUF_FORMAT| PyBUF_STRIDES, 1, 0, __pyx_stack); if (unlikely(__pyx_t_15 < 0)) { PyErr_Fetch(&__pyx_t_18, &__pyx_t_17, &__pyx_t_16); if (unlikely(__Pyx_GetBufferAndValidate(&__pyx_pybuffernd_delay.rcbuffer->pybuffer, (PyObject*)__pyx_v_delay, &__Pyx_TypeInfo_double, PyBUF_FORMAT| PyBUF_STRIDES, 1, 0, __pyx_stack) == -1)) { Py_XDECREF(__pyx_t_18); Py_XDECREF(__pyx_t_17); Py_XDECREF(__pyx_t_16); __Pyx_RaiseBufferFallbackError(); } else { PyErr_Restore(__pyx_t_18, __pyx_t_17, __pyx_t_16); } } __pyx_pybuffernd_delay.diminfo[0].strides = __pyx_pybuffernd_delay.rcbuffer->pybuffer.strides[0]; __pyx_pybuffernd_delay.diminfo[0].shape = __pyx_pybuffernd_delay.rcbuffer->pybuffer.shape[0]; if (unlikely(__pyx_t_15 < 0)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 127; __pyx_clineno = __LINE__; goto __pyx_L1_error;} } __pyx_t_22 = 0; __pyx_v_delay = ((PyArrayObject *)__pyx_t_2); __pyx_t_2 = 0; goto __pyx_L5; } __pyx_L5:; /* "wfpt.pyx":128 * if T!=0: * delay = (np.random.rand(samples)*T + (t - T/2.)) * for i from 0 <= i < samples: # <<<<<<<<<<<<<< * idx = np.searchsorted(l_cdf, f[i]) * rt = x[idx] */ __pyx_t_15 = __pyx_v_samples; for (__pyx_v_i = 0; __pyx_v_i < __pyx_t_15; __pyx_v_i++) { /* "wfpt.pyx":129 * delay = (np.random.rand(samples)*T + (t - T/2.)) * for i from 0 <= i < samples: * idx = np.searchsorted(l_cdf, f[i]) # <<<<<<<<<<<<<< * rt = x[idx] * if T==0: */ __pyx_t_2 = __Pyx_GetName(__pyx_m, __pyx_n_s__np); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 129; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_2); __pyx_t_3 = PyObject_GetAttr(__pyx_t_2, __pyx_n_s__searchsorted); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 129; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; __pyx_t_9 = __pyx_v_i; __pyx_t_2 = PyFloat_FromDouble((*__Pyx_BufPtrStrided1d(double *, __pyx_pybuffernd_f.rcbuffer->pybuffer.buf, __pyx_t_9, __pyx_pybuffernd_f.diminfo[0].strides))); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 129; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_2); __pyx_t_4 = PyTuple_New(2); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 129; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_4); __Pyx_INCREF(((PyObject *)__pyx_v_l_cdf)); PyTuple_SET_ITEM(__pyx_t_4, 0, ((PyObject *)__pyx_v_l_cdf)); __Pyx_GIVEREF(((PyObject *)__pyx_v_l_cdf)); PyTuple_SET_ITEM(__pyx_t_4, 1, __pyx_t_2); __Pyx_GIVEREF(__pyx_t_2); __pyx_t_2 = 0; __pyx_t_2 = PyObject_Call(__pyx_t_3, ((PyObject *)__pyx_t_4), NULL); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 129; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_2); __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; __Pyx_DECREF(((PyObject *)__pyx_t_4)); __pyx_t_4 = 0; __pyx_t_23 = __Pyx_PyInt_AsInt(__pyx_t_2); if (unlikely((__pyx_t_23 == (int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 129; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; __pyx_v_idx = __pyx_t_23; /* "wfpt.pyx":130 * for i from 0 <= i < samples: * idx = np.searchsorted(l_cdf, f[i]) * rt = x[idx] # <<<<<<<<<<<<<< * if T==0: * rt = rt + np.sign(rt)*t */ __pyx_t_23 = __pyx_v_idx; __pyx_v_rt = (*__Pyx_BufPtrStrided1d(double *, __pyx_pybuffernd_x.rcbuffer->pybuffer.buf, __pyx_t_23, __pyx_pybuffernd_x.diminfo[0].strides)); /* "wfpt.pyx":131 * idx = np.searchsorted(l_cdf, f[i]) * rt = x[idx] * if T==0: # <<<<<<<<<<<<<< * rt = rt + np.sign(rt)*t * else: */ __pyx_t_21 = (__pyx_v_T == 0.0); if (__pyx_t_21) { /* "wfpt.pyx":132 * rt = x[idx] * if T==0: * rt = rt + np.sign(rt)*t # <<<<<<<<<<<<<< * else: * rt = rt + np.sign(rt)*delay[i] */ __pyx_t_2 = PyFloat_FromDouble(__pyx_v_rt); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 132; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_2); __pyx_t_4 = __Pyx_GetName(__pyx_m, __pyx_n_s__np); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 132; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_4); __pyx_t_3 = PyObject_GetAttr(__pyx_t_4, __pyx_n_s__sign); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 132; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0; __pyx_t_4 = PyFloat_FromDouble(__pyx_v_rt); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 132; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_4); __pyx_t_1 = PyTuple_New(1); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 132; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_1); PyTuple_SET_ITEM(__pyx_t_1, 0, __pyx_t_4); __Pyx_GIVEREF(__pyx_t_4); __pyx_t_4 = 0; __pyx_t_4 = PyObject_Call(__pyx_t_3, ((PyObject *)__pyx_t_1), NULL); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 132; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_4); __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; __Pyx_DECREF(((PyObject *)__pyx_t_1)); __pyx_t_1 = 0; __pyx_t_1 = PyFloat_FromDouble(__pyx_v_t); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 132; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_1); __pyx_t_3 = PyNumber_Multiply(__pyx_t_4, __pyx_t_1); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 132; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0; __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; __pyx_t_1 = PyNumber_Add(__pyx_t_2, __pyx_t_3); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 132; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_1); __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; __pyx_t_24 = __pyx_PyFloat_AsDouble(__pyx_t_1); if (unlikely((__pyx_t_24 == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 132; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; __pyx_v_rt = __pyx_t_24; goto __pyx_L8; } /*else*/ { /* "wfpt.pyx":134 * rt = rt + np.sign(rt)*t * else: * rt = rt + np.sign(rt)*delay[i] # <<<<<<<<<<<<<< * rts[i] = rt * return rts */ __pyx_t_1 = PyFloat_FromDouble(__pyx_v_rt); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 134; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_1); __pyx_t_3 = __Pyx_GetName(__pyx_m, __pyx_n_s__np); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 134; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); __pyx_t_2 = PyObject_GetAttr(__pyx_t_3, __pyx_n_s__sign); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 134; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_2); __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; __pyx_t_3 = PyFloat_FromDouble(__pyx_v_rt); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 134; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); __pyx_t_4 = PyTuple_New(1); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 134; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_4); PyTuple_SET_ITEM(__pyx_t_4, 0, __pyx_t_3); __Pyx_GIVEREF(__pyx_t_3); __pyx_t_3 = 0; __pyx_t_3 = PyObject_Call(__pyx_t_2, ((PyObject *)__pyx_t_4), NULL); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 134; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; __Pyx_DECREF(((PyObject *)__pyx_t_4)); __pyx_t_4 = 0; __pyx_t_10 = __pyx_v_i; __pyx_t_4 = PyFloat_FromDouble((*__Pyx_BufPtrStrided1d(double *, __pyx_pybuffernd_delay.rcbuffer->pybuffer.buf, __pyx_t_10, __pyx_pybuffernd_delay.diminfo[0].strides))); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 134; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_4); __pyx_t_2 = PyNumber_Multiply(__pyx_t_3, __pyx_t_4); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 134; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_2); __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0; __pyx_t_4 = PyNumber_Add(__pyx_t_1, __pyx_t_2); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 134; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_4); __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; __pyx_t_24 = __pyx_PyFloat_AsDouble(__pyx_t_4); if (unlikely((__pyx_t_24 == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 134; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0; __pyx_v_rt = __pyx_t_24; } __pyx_L8:; /* "wfpt.pyx":135 * else: * rt = rt + np.sign(rt)*delay[i] * rts[i] = rt # <<<<<<<<<<<<<< * return rts * */ __pyx_t_25 = __pyx_v_i; *__Pyx_BufPtrStrided1d(double *, __pyx_pybuffernd_rts.rcbuffer->pybuffer.buf, __pyx_t_25, __pyx_pybuffernd_rts.diminfo[0].strides) = __pyx_v_rt; } /* "wfpt.pyx":136 * rt = rt + np.sign(rt)*delay[i] * rts[i] = rt * return rts # <<<<<<<<<<<<<< * * def wiener_like_contaminant(np.ndarray[double, ndim=1] x, np.ndarray[int, ndim=1] cont_x, double v, \ */ __Pyx_XDECREF(__pyx_r); __Pyx_INCREF(((PyObject *)__pyx_v_rts)); __pyx_r = ((PyObject *)__pyx_v_rts); goto __pyx_L0; __pyx_r = Py_None; __Pyx_INCREF(Py_None); goto __pyx_L0; __pyx_L1_error:; __Pyx_XDECREF(__pyx_t_1); __Pyx_XDECREF(__pyx_t_2); __Pyx_XDECREF(__pyx_t_3); __Pyx_XDECREF(__pyx_t_4); __Pyx_XDECREF(__pyx_t_5); { PyObject *__pyx_type, *__pyx_value, *__pyx_tb; __Pyx_ErrFetch(&__pyx_type, &__pyx_value, &__pyx_tb); __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd_rts.rcbuffer->pybuffer); __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd_x.rcbuffer->pybuffer); __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd_delay.rcbuffer->pybuffer); __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd_l_cdf.rcbuffer->pybuffer); __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd_f.rcbuffer->pybuffer); __Pyx_ErrRestore(__pyx_type, __pyx_value, __pyx_tb);} __Pyx_AddTraceback("wfpt.gen_rts_from_cdf", __pyx_clineno, __pyx_lineno, __pyx_filename); __pyx_r = NULL; goto __pyx_L2; __pyx_L0:; __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd_rts.rcbuffer->pybuffer); __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd_x.rcbuffer->pybuffer); __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd_delay.rcbuffer->pybuffer); __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd_l_cdf.rcbuffer->pybuffer); __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd_f.rcbuffer->pybuffer); __pyx_L2:; __Pyx_XDECREF((PyObject *)__pyx_v_x); __Pyx_XDECREF((PyObject *)__pyx_v_l_cdf); __Pyx_XDECREF((PyObject *)__pyx_v_rts); __Pyx_XDECREF((PyObject *)__pyx_v_f); __Pyx_XDECREF((PyObject *)__pyx_v_delay); __Pyx_XGIVEREF(__pyx_r); __Pyx_RefNannyFinishContext(); return __pyx_r; } /* Python wrapper */ static PyObject *__pyx_pw_4wfpt_13wiener_like_contaminant(PyObject *__pyx_self, PyObject *__pyx_args, PyObject *__pyx_kwds); /*proto*/ static char __pyx_doc_4wfpt_12wiener_like_contaminant[] = "wiener_like_contaminant(ndarray x, ndarray cont_x, double v, double V, double a, double z, double Z, double t, double T, double t_min, double t_max, double err, int nT=10, int nZ=10, int use_adaptive=1, double simps_err=1e-08)\nWiener likelihood function where RTs could come from a\n separate, uniform contaminant distribution.\n\n Reference: Lee, Vandekerckhove, Navarro, & Tuernlinckx (2007)\n "; static PyMethodDef __pyx_mdef_4wfpt_13wiener_like_contaminant = {__Pyx_NAMESTR("wiener_like_contaminant"), (PyCFunction)__pyx_pw_4wfpt_13wiener_like_contaminant, METH_VARARGS|METH_KEYWORDS, __Pyx_DOCSTR(__pyx_doc_4wfpt_12wiener_like_contaminant)}; static PyObject *__pyx_pw_4wfpt_13wiener_like_contaminant(PyObject *__pyx_self, PyObject *__pyx_args, PyObject *__pyx_kwds) { PyArrayObject *__pyx_v_x = 0; PyArrayObject *__pyx_v_cont_x = 0; double __pyx_v_v; double __pyx_v_V; double __pyx_v_a; double __pyx_v_z; double __pyx_v_Z; double __pyx_v_t; double __pyx_v_T; double __pyx_v_t_min; double __pyx_v_t_max; double __pyx_v_err; int __pyx_v_nT; int __pyx_v_nZ; int __pyx_v_use_adaptive; double __pyx_v_simps_err; static PyObject **__pyx_pyargnames[] = {&__pyx_n_s__x,&__pyx_n_s__cont_x,&__pyx_n_s__v,&__pyx_n_s__V,&__pyx_n_s__a,&__pyx_n_s__z,&__pyx_n_s__Z,&__pyx_n_s__t,&__pyx_n_s__T,&__pyx_n_s__t_min,&__pyx_n_s__t_max,&__pyx_n_s__err,&__pyx_n_s__nT,&__pyx_n_s__nZ,&__pyx_n_s__use_adaptive,&__pyx_n_s__simps_err,0}; PyObject *__pyx_r = 0; __Pyx_RefNannyDeclarations __Pyx_RefNannySetupContext("wiener_like_contaminant (wrapper)"); __pyx_self = __pyx_self; { PyObject* values[16] = {0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0}; if (unlikely(__pyx_kwds)) { Py_ssize_t kw_args; const Py_ssize_t pos_args = PyTuple_GET_SIZE(__pyx_args); switch (pos_args) { case 16: values[15] = PyTuple_GET_ITEM(__pyx_args, 15); case 15: values[14] = PyTuple_GET_ITEM(__pyx_args, 14); case 14: values[13] = PyTuple_GET_ITEM(__pyx_args, 13); case 13: values[12] = PyTuple_GET_ITEM(__pyx_args, 12); case 12: values[11] = PyTuple_GET_ITEM(__pyx_args, 11); case 11: values[10] = PyTuple_GET_ITEM(__pyx_args, 10); case 10: values[9] = PyTuple_GET_ITEM(__pyx_args, 9); case 9: values[8] = PyTuple_GET_ITEM(__pyx_args, 8); case 8: values[7] = PyTuple_GET_ITEM(__pyx_args, 7); case 7: values[6] = PyTuple_GET_ITEM(__pyx_args, 6); case 6: values[5] = PyTuple_GET_ITEM(__pyx_args, 5); case 5: values[4] = PyTuple_GET_ITEM(__pyx_args, 4); case 4: values[3] = PyTuple_GET_ITEM(__pyx_args, 3); case 3: values[2] = PyTuple_GET_ITEM(__pyx_args, 2); case 2: values[1] = PyTuple_GET_ITEM(__pyx_args, 1); case 1: values[0] = PyTuple_GET_ITEM(__pyx_args, 0); case 0: break; default: goto __pyx_L5_argtuple_error; } kw_args = PyDict_Size(__pyx_kwds); switch (pos_args) { case 0: values[0] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__x); if (likely(values[0])) kw_args--; else goto __pyx_L5_argtuple_error; case 1: values[1] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__cont_x); if (likely(values[1])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("wiener_like_contaminant", 0, 12, 16, 1); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 138; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 2: values[2] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__v); if (likely(values[2])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("wiener_like_contaminant", 0, 12, 16, 2); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 138; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 3: values[3] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__V); if (likely(values[3])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("wiener_like_contaminant", 0, 12, 16, 3); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 138; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 4: values[4] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__a); if (likely(values[4])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("wiener_like_contaminant", 0, 12, 16, 4); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 138; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 5: values[5] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__z); if (likely(values[5])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("wiener_like_contaminant", 0, 12, 16, 5); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 138; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 6: values[6] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__Z); if (likely(values[6])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("wiener_like_contaminant", 0, 12, 16, 6); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 138; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 7: values[7] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__t); if (likely(values[7])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("wiener_like_contaminant", 0, 12, 16, 7); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 138; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 8: values[8] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__T); if (likely(values[8])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("wiener_like_contaminant", 0, 12, 16, 8); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 138; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 9: values[9] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__t_min); if (likely(values[9])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("wiener_like_contaminant", 0, 12, 16, 9); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 138; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 10: values[10] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__t_max); if (likely(values[10])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("wiener_like_contaminant", 0, 12, 16, 10); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 138; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 11: values[11] = PyDict_GetItem(__pyx_kwds, __pyx_n_s__err); if (likely(values[11])) kw_args--; else { __Pyx_RaiseArgtupleInvalid("wiener_like_contaminant", 0, 12, 16, 11); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 138; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } case 12: if (kw_args > 0) { PyObject* value = PyDict_GetItem(__pyx_kwds, __pyx_n_s__nT); if (value) { values[12] = value; kw_args--; } } case 13: if (kw_args > 0) { PyObject* value = PyDict_GetItem(__pyx_kwds, __pyx_n_s__nZ); if (value) { values[13] = value; kw_args--; } } case 14: if (kw_args > 0) { PyObject* value = PyDict_GetItem(__pyx_kwds, __pyx_n_s__use_adaptive); if (value) { values[14] = value; kw_args--; } } case 15: if (kw_args > 0) { PyObject* value = PyDict_GetItem(__pyx_kwds, __pyx_n_s__simps_err); if (value) { values[15] = value; kw_args--; } } } if (unlikely(kw_args > 0)) { if (unlikely(__Pyx_ParseOptionalKeywords(__pyx_kwds, __pyx_pyargnames, 0, values, pos_args, "wiener_like_contaminant") < 0)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 138; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } __pyx_v_x = ((PyArrayObject *)values[0]); __pyx_v_cont_x = ((PyArrayObject *)values[1]); __pyx_v_v = __pyx_PyFloat_AsDouble(values[2]); if (unlikely((__pyx_v_v == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 138; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_V = __pyx_PyFloat_AsDouble(values[3]); if (unlikely((__pyx_v_V == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 139; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_a = __pyx_PyFloat_AsDouble(values[4]); if (unlikely((__pyx_v_a == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 139; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_z = __pyx_PyFloat_AsDouble(values[5]); if (unlikely((__pyx_v_z == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 139; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_Z = __pyx_PyFloat_AsDouble(values[6]); if (unlikely((__pyx_v_Z == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 139; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_t = __pyx_PyFloat_AsDouble(values[7]); if (unlikely((__pyx_v_t == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 139; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_T = __pyx_PyFloat_AsDouble(values[8]); if (unlikely((__pyx_v_T == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 139; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_t_min = __pyx_PyFloat_AsDouble(values[9]); if (unlikely((__pyx_v_t_min == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 139; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_t_max = __pyx_PyFloat_AsDouble(values[10]); if (unlikely((__pyx_v_t_max == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 140; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_err = __pyx_PyFloat_AsDouble(values[11]); if (unlikely((__pyx_v_err == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 140; __pyx_clineno = __LINE__; goto __pyx_L3_error;} if (values[12]) { __pyx_v_nT = __Pyx_PyInt_AsInt(values[12]); if (unlikely((__pyx_v_nT == (int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 140; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } else { __pyx_v_nT = ((int)10); } if (values[13]) { __pyx_v_nZ = __Pyx_PyInt_AsInt(values[13]); if (unlikely((__pyx_v_nZ == (int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 140; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } else { __pyx_v_nZ = ((int)10); } if (values[14]) { __pyx_v_use_adaptive = __Pyx_PyObject_IsTrue(values[14]); if (unlikely((__pyx_v_use_adaptive == (int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 140; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } else { __pyx_v_use_adaptive = ((int)1); } if (values[15]) { __pyx_v_simps_err = __pyx_PyFloat_AsDouble(values[15]); if (unlikely((__pyx_v_simps_err == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 141; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } else { /* "wfpt.pyx":141 * double V, double a, double z, double Z, double t, double T, double t_min, \ * double t_max, double err, int nT= 10, int nZ=10, bint use_adaptive=1, \ * double simps_err=1e-8): # <<<<<<<<<<<<<< * """Wiener likelihood function where RTs could come from a * separate, uniform contaminant distribution. */ __pyx_v_simps_err = ((double)1e-8); } } else { switch (PyTuple_GET_SIZE(__pyx_args)) { case 16: values[15] = PyTuple_GET_ITEM(__pyx_args, 15); case 15: values[14] = PyTuple_GET_ITEM(__pyx_args, 14); case 14: values[13] = PyTuple_GET_ITEM(__pyx_args, 13); case 13: values[12] = PyTuple_GET_ITEM(__pyx_args, 12); case 12: values[11] = PyTuple_GET_ITEM(__pyx_args, 11); values[10] = PyTuple_GET_ITEM(__pyx_args, 10); values[9] = PyTuple_GET_ITEM(__pyx_args, 9); values[8] = PyTuple_GET_ITEM(__pyx_args, 8); values[7] = PyTuple_GET_ITEM(__pyx_args, 7); values[6] = PyTuple_GET_ITEM(__pyx_args, 6); values[5] = PyTuple_GET_ITEM(__pyx_args, 5); values[4] = PyTuple_GET_ITEM(__pyx_args, 4); values[3] = PyTuple_GET_ITEM(__pyx_args, 3); values[2] = PyTuple_GET_ITEM(__pyx_args, 2); values[1] = PyTuple_GET_ITEM(__pyx_args, 1); values[0] = PyTuple_GET_ITEM(__pyx_args, 0); break; default: goto __pyx_L5_argtuple_error; } } __pyx_v_x = ((PyArrayObject *)values[0]); __pyx_v_cont_x = ((PyArrayObject *)values[1]); __pyx_v_v = __pyx_PyFloat_AsDouble(values[2]); if (unlikely((__pyx_v_v == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 138; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_V = __pyx_PyFloat_AsDouble(values[3]); if (unlikely((__pyx_v_V == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 139; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_a = __pyx_PyFloat_AsDouble(values[4]); if (unlikely((__pyx_v_a == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 139; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_z = __pyx_PyFloat_AsDouble(values[5]); if (unlikely((__pyx_v_z == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 139; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_Z = __pyx_PyFloat_AsDouble(values[6]); if (unlikely((__pyx_v_Z == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 139; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_t = __pyx_PyFloat_AsDouble(values[7]); if (unlikely((__pyx_v_t == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 139; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_T = __pyx_PyFloat_AsDouble(values[8]); if (unlikely((__pyx_v_T == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 139; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_t_min = __pyx_PyFloat_AsDouble(values[9]); if (unlikely((__pyx_v_t_min == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 139; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_t_max = __pyx_PyFloat_AsDouble(values[10]); if (unlikely((__pyx_v_t_max == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 140; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_v_err = __pyx_PyFloat_AsDouble(values[11]); if (unlikely((__pyx_v_err == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 140; __pyx_clineno = __LINE__; goto __pyx_L3_error;} if (values[12]) { __pyx_v_nT = __Pyx_PyInt_AsInt(values[12]); if (unlikely((__pyx_v_nT == (int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 140; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } else { __pyx_v_nT = ((int)10); } if (values[13]) { __pyx_v_nZ = __Pyx_PyInt_AsInt(values[13]); if (unlikely((__pyx_v_nZ == (int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 140; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } else { __pyx_v_nZ = ((int)10); } if (values[14]) { __pyx_v_use_adaptive = __Pyx_PyObject_IsTrue(values[14]); if (unlikely((__pyx_v_use_adaptive == (int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 140; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } else { __pyx_v_use_adaptive = ((int)1); } if (values[15]) { __pyx_v_simps_err = __pyx_PyFloat_AsDouble(values[15]); if (unlikely((__pyx_v_simps_err == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 141; __pyx_clineno = __LINE__; goto __pyx_L3_error;} } else { __pyx_v_simps_err = ((double)1e-8); } } goto __pyx_L4_argument_unpacking_done; __pyx_L5_argtuple_error:; __Pyx_RaiseArgtupleInvalid("wiener_like_contaminant", 0, 12, 16, PyTuple_GET_SIZE(__pyx_args)); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 138; __pyx_clineno = __LINE__; goto __pyx_L3_error;} __pyx_L3_error:; __Pyx_AddTraceback("wfpt.wiener_like_contaminant", __pyx_clineno, __pyx_lineno, __pyx_filename); __Pyx_RefNannyFinishContext(); return NULL; __pyx_L4_argument_unpacking_done:; if (unlikely(!__Pyx_ArgTypeTest(((PyObject *)__pyx_v_x), __pyx_ptype_5numpy_ndarray, 1, "x", 0))) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 138; __pyx_clineno = __LINE__; goto __pyx_L1_error;} if (unlikely(!__Pyx_ArgTypeTest(((PyObject *)__pyx_v_cont_x), __pyx_ptype_5numpy_ndarray, 1, "cont_x", 0))) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 138; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __pyx_r = __pyx_pf_4wfpt_12wiener_like_contaminant(__pyx_self, __pyx_v_x, __pyx_v_cont_x, __pyx_v_v, __pyx_v_V, __pyx_v_a, __pyx_v_z, __pyx_v_Z, __pyx_v_t, __pyx_v_T, __pyx_v_t_min, __pyx_v_t_max, __pyx_v_err, __pyx_v_nT, __pyx_v_nZ, __pyx_v_use_adaptive, __pyx_v_simps_err); goto __pyx_L0; __pyx_L1_error:; __pyx_L0:; __Pyx_RefNannyFinishContext(); return __pyx_r; } /* "wfpt.pyx":138 * return rts * * def wiener_like_contaminant(np.ndarray[double, ndim=1] x, np.ndarray[int, ndim=1] cont_x, double v, \ # <<<<<<<<<<<<<< * double V, double a, double z, double Z, double t, double T, double t_min, \ * double t_max, double err, int nT= 10, int nZ=10, bint use_adaptive=1, \ */ static PyObject *__pyx_pf_4wfpt_12wiener_like_contaminant(CYTHON_UNUSED PyObject *__pyx_self, PyArrayObject *__pyx_v_x, PyArrayObject *__pyx_v_cont_x, double __pyx_v_v, double __pyx_v_V, double __pyx_v_a, double __pyx_v_z, double __pyx_v_Z, double __pyx_v_t, double __pyx_v_T, double __pyx_v_t_min, double __pyx_v_t_max, double __pyx_v_err, int __pyx_v_nT, int __pyx_v_nZ, int __pyx_v_use_adaptive, double __pyx_v_simps_err) { CYTHON_UNUSED Py_ssize_t __pyx_v_size; Py_ssize_t __pyx_v_i; double __pyx_v_p; double __pyx_v_sum_logp; int __pyx_v_n_cont; CYTHON_UNUSED int __pyx_v_pos_cont; __Pyx_LocalBuf_ND __pyx_pybuffernd_cont_x; __Pyx_Buffer __pyx_pybuffer_cont_x; __Pyx_LocalBuf_ND __pyx_pybuffernd_x; __Pyx_Buffer __pyx_pybuffer_x; PyObject *__pyx_r = NULL; __Pyx_RefNannyDeclarations PyObject *__pyx_t_1 = NULL; PyObject *__pyx_t_2 = NULL; PyObject *__pyx_t_3 = NULL; int __pyx_t_4; Py_ssize_t __pyx_t_5; Py_ssize_t __pyx_t_6; Py_ssize_t __pyx_t_7; Py_ssize_t __pyx_t_8; int __pyx_t_9; Py_ssize_t __pyx_t_10; double __pyx_t_11; struct __pyx_opt_args_4wfpt_full_pdf __pyx_t_12; int __pyx_lineno = 0; const char *__pyx_filename = NULL; int __pyx_clineno = 0; __Pyx_RefNannySetupContext("wiener_like_contaminant"); __pyx_pybuffer_x.pybuffer.buf = NULL; __pyx_pybuffer_x.refcount = 0; __pyx_pybuffernd_x.data = NULL; __pyx_pybuffernd_x.rcbuffer = &__pyx_pybuffer_x; __pyx_pybuffer_cont_x.pybuffer.buf = NULL; __pyx_pybuffer_cont_x.refcount = 0; __pyx_pybuffernd_cont_x.data = NULL; __pyx_pybuffernd_cont_x.rcbuffer = &__pyx_pybuffer_cont_x; { __Pyx_BufFmt_StackElem __pyx_stack[1]; if (unlikely(__Pyx_GetBufferAndValidate(&__pyx_pybuffernd_x.rcbuffer->pybuffer, (PyObject*)__pyx_v_x, &__Pyx_TypeInfo_double, PyBUF_FORMAT| PyBUF_STRIDES, 1, 0, __pyx_stack) == -1)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 138; __pyx_clineno = __LINE__; goto __pyx_L1_error;} } __pyx_pybuffernd_x.diminfo[0].strides = __pyx_pybuffernd_x.rcbuffer->pybuffer.strides[0]; __pyx_pybuffernd_x.diminfo[0].shape = __pyx_pybuffernd_x.rcbuffer->pybuffer.shape[0]; { __Pyx_BufFmt_StackElem __pyx_stack[1]; if (unlikely(__Pyx_GetBufferAndValidate(&__pyx_pybuffernd_cont_x.rcbuffer->pybuffer, (PyObject*)__pyx_v_cont_x, &__Pyx_TypeInfo_int, PyBUF_FORMAT| PyBUF_STRIDES, 1, 0, __pyx_stack) == -1)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 138; __pyx_clineno = __LINE__; goto __pyx_L1_error;} } __pyx_pybuffernd_cont_x.diminfo[0].strides = __pyx_pybuffernd_cont_x.rcbuffer->pybuffer.strides[0]; __pyx_pybuffernd_cont_x.diminfo[0].shape = __pyx_pybuffernd_cont_x.rcbuffer->pybuffer.shape[0]; /* "wfpt.pyx":147 * Reference: Lee, Vandekerckhove, Navarro, & Tuernlinckx (2007) * """ * cdef Py_ssize_t size = x.shape[0] # <<<<<<<<<<<<<< * cdef Py_ssize_t i * cdef double p */ __pyx_v_size = (__pyx_v_x->dimensions[0]); /* "wfpt.pyx":150 * cdef Py_ssize_t i * cdef double p * cdef double sum_logp = 0 # <<<<<<<<<<<<<< * cdef int n_cont = np.sum(cont_x) * cdef int pos_cont = 0 */ __pyx_v_sum_logp = 0.0; /* "wfpt.pyx":151 * cdef double p * cdef double sum_logp = 0 * cdef int n_cont = np.sum(cont_x) # <<<<<<<<<<<<<< * cdef int pos_cont = 0 * */ __pyx_t_1 = __Pyx_GetName(__pyx_m, __pyx_n_s__np); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 151; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_1); __pyx_t_2 = PyObject_GetAttr(__pyx_t_1, __pyx_n_s__sum); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 151; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_2); __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; __pyx_t_1 = PyTuple_New(1); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 151; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_1); __Pyx_INCREF(((PyObject *)__pyx_v_cont_x)); PyTuple_SET_ITEM(__pyx_t_1, 0, ((PyObject *)__pyx_v_cont_x)); __Pyx_GIVEREF(((PyObject *)__pyx_v_cont_x)); __pyx_t_3 = PyObject_Call(__pyx_t_2, ((PyObject *)__pyx_t_1), NULL); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 151; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; __Pyx_DECREF(((PyObject *)__pyx_t_1)); __pyx_t_1 = 0; __pyx_t_4 = __Pyx_PyInt_AsInt(__pyx_t_3); if (unlikely((__pyx_t_4 == (int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 151; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; __pyx_v_n_cont = __pyx_t_4; /* "wfpt.pyx":152 * cdef double sum_logp = 0 * cdef int n_cont = np.sum(cont_x) * cdef int pos_cont = 0 # <<<<<<<<<<<<<< * * for i in prange(size, nogil=True): */ __pyx_v_pos_cont = 0; /* "wfpt.pyx":154 * cdef int pos_cont = 0 * * for i in prange(size, nogil=True): # <<<<<<<<<<<<<< * if cont_x[i] == 0: * p = full_pdf(x[i], v, V, a, z, Z, t, T, err, nT, nZ, use_adaptive, simps_err) */ { #ifdef WITH_THREAD PyThreadState *_save = NULL; #endif Py_UNBLOCK_THREADS /*try:*/ { __pyx_t_5 = __pyx_v_size; if (1 == 0) abort(); { double __pyx_parallel_temp0 = __PYX_NAN; Py_ssize_t __pyx_parallel_temp1 = 0xbad0bad0; double __pyx_parallel_temp2 = __PYX_NAN; const char *__pyx_parallel_filename; int __pyx_parallel_lineno, __pyx_parallel_clineno; __pyx_parallel_filename = NULL; __pyx_parallel_lineno = __pyx_parallel_clineno = 0; PyObject *__pyx_parallel_exc_type = NULL, *__pyx_parallel_exc_value = NULL, *__pyx_parallel_exc_tb = NULL; int __pyx_parallel_why; __pyx_parallel_why = 0; __pyx_t_7 = (__pyx_t_5 - 0) / 1; if (__pyx_t_7 > 0) { __pyx_v_i = 0; #ifdef _OPENMP #pragma omp parallel reduction(+:__pyx_v_sum_logp) private(__pyx_t_8, __pyx_t_12, __pyx_t_11, __pyx_t_9, __pyx_t_10) firstprivate(__pyx_t_1, __pyx_t_3) private(__pyx_filename, __pyx_lineno, __pyx_clineno) shared(__pyx_parallel_why, __pyx_parallel_exc_type, __pyx_parallel_exc_value, __pyx_parallel_exc_tb) #endif /* _OPENMP */ { #ifdef _OPENMP #ifdef WITH_THREAD PyGILState_STATE __pyx_gilstate_save = PyGILState_Ensure(); #endif Py_BEGIN_ALLOW_THREADS #endif /* _OPENMP */ #ifdef _OPENMP #pragma omp for firstprivate(__pyx_v_i) lastprivate(__pyx_v_i) lastprivate(__pyx_v_p) #endif /* _OPENMP */ for (__pyx_t_6 = 0; __pyx_t_6 < __pyx_t_7; __pyx_t_6++){ if (__pyx_parallel_why < 2) { __pyx_v_i = 0 + 1 * __pyx_t_6; /* Initialize private variables to invalid values */ __pyx_v_p = ((double)__PYX_NAN); /* "wfpt.pyx":155 * * for i in prange(size, nogil=True): * if cont_x[i] == 0: # <<<<<<<<<<<<<< * p = full_pdf(x[i], v, V, a, z, Z, t, T, err, nT, nZ, use_adaptive, simps_err) * if p == 0: */ __pyx_t_8 = __pyx_v_i; __pyx_t_9 = ((*__Pyx_BufPtrStrided1d(int *, __pyx_pybuffernd_cont_x.rcbuffer->pybuffer.buf, __pyx_t_8, __pyx_pybuffernd_cont_x.diminfo[0].strides)) == 0); if (__pyx_t_9) { /* "wfpt.pyx":156 * for i in prange(size, nogil=True): * if cont_x[i] == 0: * p = full_pdf(x[i], v, V, a, z, Z, t, T, err, nT, nZ, use_adaptive, simps_err) # <<<<<<<<<<<<<< * if p == 0: * with gil: */ __pyx_t_10 = __pyx_v_i; __pyx_t_12.__pyx_n = 4; __pyx_t_12.nT = __pyx_v_nT; __pyx_t_12.nZ = __pyx_v_nZ; __pyx_t_12.use_adaptive = __pyx_v_use_adaptive; __pyx_t_12.simps_err = __pyx_v_simps_err; __pyx_t_11 = __pyx_f_4wfpt_full_pdf((*__Pyx_BufPtrStrided1d(double *, __pyx_pybuffernd_x.rcbuffer->pybuffer.buf, __pyx_t_10, __pyx_pybuffernd_x.diminfo[0].strides)), __pyx_v_v, __pyx_v_V, __pyx_v_a, __pyx_v_z, __pyx_v_Z, __pyx_v_t, __pyx_v_T, __pyx_v_err, 0, &__pyx_t_12); __pyx_v_p = __pyx_t_11; /* "wfpt.pyx":157 * if cont_x[i] == 0: * p = full_pdf(x[i], v, V, a, z, Z, t, T, err, nT, nZ, use_adaptive, simps_err) * if p == 0: # <<<<<<<<<<<<<< * with gil: * return -np.inf */ __pyx_t_9 = (__pyx_v_p == 0.0); if (__pyx_t_9) { /* "wfpt.pyx":158 * p = full_pdf(x[i], v, V, a, z, Z, t, T, err, nT, nZ, use_adaptive, simps_err) * if p == 0: * with gil: # <<<<<<<<<<<<<< * return -np.inf * sum_logp += log(p) */ { #ifdef WITH_THREAD PyGILState_STATE __pyx_gilstate_save = PyGILState_Ensure(); #endif /*try:*/ { /* "wfpt.pyx":159 * if p == 0: * with gil: * return -np.inf # <<<<<<<<<<<<<< * sum_logp += log(p) * # If one probability = 0, the log sum will be -Inf */ __Pyx_XDECREF(__pyx_r); __pyx_t_3 = __Pyx_GetName(__pyx_m, __pyx_n_s__np); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 159; __pyx_clineno = __LINE__; goto __pyx_L15;} __Pyx_GOTREF(__pyx_t_3); __pyx_t_1 = PyObject_GetAttr(__pyx_t_3, __pyx_n_s__inf); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 159; __pyx_clineno = __LINE__; goto __pyx_L15;} __Pyx_GOTREF(__pyx_t_1); __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; __pyx_t_3 = PyNumber_Negative(__pyx_t_1); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 159; __pyx_clineno = __LINE__; goto __pyx_L15;} __Pyx_GOTREF(__pyx_t_3); __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; __pyx_r = __pyx_t_3; __pyx_t_3 = 0; goto __pyx_L14; } /* "wfpt.pyx":158 * p = full_pdf(x[i], v, V, a, z, Z, t, T, err, nT, nZ, use_adaptive, simps_err) * if p == 0: * with gil: # <<<<<<<<<<<<<< * return -np.inf * sum_logp += log(p) */ /*finally:*/ { int __pyx_why; __pyx_why = 0; goto __pyx_L16; __pyx_L14: __pyx_why = 3; goto __pyx_L16; __pyx_L15: __pyx_why = 4; goto __pyx_L16; __pyx_L16:; #ifdef WITH_THREAD PyGILState_Release(__pyx_gilstate_save); #endif switch (__pyx_why) { case 3: goto __pyx_L9_return; case 4: goto __pyx_L8_error; } } } goto __pyx_L11; } __pyx_L11:; /* "wfpt.pyx":160 * with gil: * return -np.inf * sum_logp += log(p) # <<<<<<<<<<<<<< * # If one probability = 0, the log sum will be -Inf * */ __pyx_v_sum_logp = (__pyx_v_sum_logp + log(__pyx_v_p)); goto __pyx_L10; } __pyx_L10:; goto __pyx_L19; __pyx_L9_return:; __pyx_parallel_why = 3; goto __pyx_L18; __pyx_L8_error:; { #ifdef WITH_THREAD PyGILState_STATE __pyx_gilstate_save = PyGILState_Ensure(); #endif #ifdef _OPENMP #pragma omp flush(__pyx_parallel_exc_type) #endif /* _OPENMP */ if (!__pyx_parallel_exc_type) { __Pyx_ErrFetch(&__pyx_parallel_exc_type, &__pyx_parallel_exc_value, &__pyx_parallel_exc_tb); __pyx_parallel_filename = __pyx_filename; __pyx_parallel_lineno = __pyx_lineno; __pyx_parallel_clineno = __pyx_clineno; __Pyx_GOTREF(__pyx_parallel_exc_type); } #ifdef WITH_THREAD PyGILState_Release(__pyx_gilstate_save); #endif } __pyx_parallel_why = 4; goto __pyx_L18; __pyx_L18:; #ifdef _OPENMP #pragma omp critical(__pyx_parallel_lastprivates2) #endif /* _OPENMP */ { __pyx_parallel_temp0 = __pyx_v_sum_logp; __pyx_parallel_temp1 = __pyx_v_i; __pyx_parallel_temp2 = __pyx_v_p; } __pyx_L19:; #ifdef _OPENMP #pragma omp flush(__pyx_parallel_why) #endif /* _OPENMP */ } } #ifdef _OPENMP Py_END_ALLOW_THREADS #ifdef WITH_THREAD PyGILState_Release(__pyx_gilstate_save); #endif #endif /* _OPENMP */ } } if (__pyx_parallel_exc_type) { /* This may have been overridden by a continue, break or return in another thread. Prefer the error. */ __pyx_parallel_why = 4; } if (__pyx_parallel_why) { __pyx_v_sum_logp = __pyx_parallel_temp0; __pyx_v_i = __pyx_parallel_temp1; __pyx_v_p = __pyx_parallel_temp2; switch (__pyx_parallel_why) { case 3: goto __pyx_L3; case 4: { #ifdef WITH_THREAD PyGILState_STATE __pyx_gilstate_save = PyGILState_Ensure(); #endif __Pyx_ErrRestore(__pyx_parallel_exc_type, __pyx_parallel_exc_value, __pyx_parallel_exc_tb); __pyx_filename = __pyx_parallel_filename; __pyx_lineno = __pyx_parallel_lineno; __pyx_clineno = __pyx_parallel_clineno; __Pyx_GIVEREF(__pyx_parallel_exc_type); #ifdef WITH_THREAD PyGILState_Release(__pyx_gilstate_save); #endif } goto __pyx_L4; } } } } /* "wfpt.pyx":154 * cdef int pos_cont = 0 * * for i in prange(size, nogil=True): # <<<<<<<<<<<<<< * if cont_x[i] == 0: * p = full_pdf(x[i], v, V, a, z, Z, t, T, err, nT, nZ, use_adaptive, simps_err) */ /*finally:*/ { int __pyx_why; __pyx_why = 0; goto __pyx_L5; __pyx_L3: __pyx_why = 3; goto __pyx_L5; __pyx_L4: __pyx_why = 4; goto __pyx_L5; __pyx_L5:; Py_BLOCK_THREADS switch (__pyx_why) { case 3: goto __pyx_L0; case 4: goto __pyx_L1_error; } } } /* "wfpt.pyx":165 * * # add the log likelihood of the contaminations * sum_logp += n_cont*log(0.5 * 1./(t_max-t_min)) # <<<<<<<<<<<<<< * * return sum_logp */ __pyx_v_sum_logp = (__pyx_v_sum_logp + (__pyx_v_n_cont * log(((0.5 * 1.) / (__pyx_v_t_max - __pyx_v_t_min))))); /* "wfpt.pyx":167 * sum_logp += n_cont*log(0.5 * 1./(t_max-t_min)) * * return sum_logp # <<<<<<<<<<<<<< * */ __Pyx_XDECREF(__pyx_r); __pyx_t_3 = PyFloat_FromDouble(__pyx_v_sum_logp); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 167; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); __pyx_r = __pyx_t_3; __pyx_t_3 = 0; goto __pyx_L0; __pyx_r = Py_None; __Pyx_INCREF(Py_None); goto __pyx_L0; __pyx_L1_error:; __Pyx_XDECREF(__pyx_t_1); __Pyx_XDECREF(__pyx_t_2); __Pyx_XDECREF(__pyx_t_3); { PyObject *__pyx_type, *__pyx_value, *__pyx_tb; __Pyx_ErrFetch(&__pyx_type, &__pyx_value, &__pyx_tb); __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd_cont_x.rcbuffer->pybuffer); __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd_x.rcbuffer->pybuffer); __Pyx_ErrRestore(__pyx_type, __pyx_value, __pyx_tb);} __Pyx_AddTraceback("wfpt.wiener_like_contaminant", __pyx_clineno, __pyx_lineno, __pyx_filename); __pyx_r = NULL; goto __pyx_L2; __pyx_L0:; __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd_cont_x.rcbuffer->pybuffer); __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd_x.rcbuffer->pybuffer); __pyx_L2:; __Pyx_XGIVEREF(__pyx_r); __Pyx_RefNannyFinishContext(); return __pyx_r; } /* Python wrapper */ static int __pyx_pw_5numpy_7ndarray_1__getbuffer__(PyObject *__pyx_v_self, Py_buffer *__pyx_v_info, int __pyx_v_flags); /*proto*/ static int __pyx_pw_5numpy_7ndarray_1__getbuffer__(PyObject *__pyx_v_self, Py_buffer *__pyx_v_info, int __pyx_v_flags) { int __pyx_r; __Pyx_RefNannyDeclarations __Pyx_RefNannySetupContext("__getbuffer__ (wrapper)"); __pyx_r = __pyx_pf_5numpy_7ndarray___getbuffer__(((PyArrayObject *)__pyx_v_self), ((Py_buffer *)__pyx_v_info), ((int)__pyx_v_flags)); __Pyx_RefNannyFinishContext(); return __pyx_r; } /* "numpy.pxd":193 * # experimental exception made for __getbuffer__ and __releasebuffer__ * # -- the details of this may change. * def __getbuffer__(ndarray self, Py_buffer* info, int flags): # <<<<<<<<<<<<<< * # This implementation of getbuffer is geared towards Cython * # requirements, and does not yet fullfill the PEP. */ static int __pyx_pf_5numpy_7ndarray___getbuffer__(PyArrayObject *__pyx_v_self, Py_buffer *__pyx_v_info, int __pyx_v_flags) { int __pyx_v_copy_shape; int __pyx_v_i; int __pyx_v_ndim; int __pyx_v_endian_detector; int __pyx_v_little_endian; int __pyx_v_t; char *__pyx_v_f; PyArray_Descr *__pyx_v_descr = 0; int __pyx_v_offset; int __pyx_v_hasfields; int __pyx_r; __Pyx_RefNannyDeclarations int __pyx_t_1; int __pyx_t_2; int __pyx_t_3; PyObject *__pyx_t_4 = NULL; int __pyx_t_5; int __pyx_t_6; int __pyx_t_7; PyObject *__pyx_t_8 = NULL; char *__pyx_t_9; int __pyx_lineno = 0; const char *__pyx_filename = NULL; int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__getbuffer__"); if (__pyx_v_info != NULL) { __pyx_v_info->obj = Py_None; __Pyx_INCREF(Py_None); __Pyx_GIVEREF(__pyx_v_info->obj); } /* "numpy.pxd":199 * # of flags * * if info == NULL: return # <<<<<<<<<<<<<< * * cdef int copy_shape, i, ndim */ __pyx_t_1 = (__pyx_v_info == NULL); if (__pyx_t_1) { __pyx_r = 0; goto __pyx_L0; goto __pyx_L3; } __pyx_L3:; /* "numpy.pxd":202 * * cdef int copy_shape, i, ndim * cdef int endian_detector = 1 # <<<<<<<<<<<<<< * cdef bint little_endian = ((&endian_detector)[0] != 0) * */ __pyx_v_endian_detector = 1; /* "numpy.pxd":203 * cdef int copy_shape, i, ndim * cdef int endian_detector = 1 * cdef bint little_endian = ((&endian_detector)[0] != 0) # <<<<<<<<<<<<<< * * ndim = PyArray_NDIM(self) */ __pyx_v_little_endian = ((((char *)(&__pyx_v_endian_detector))[0]) != 0); /* "numpy.pxd":205 * cdef bint little_endian = ((&endian_detector)[0] != 0) * * ndim = PyArray_NDIM(self) # <<<<<<<<<<<<<< * * if sizeof(npy_intp) != sizeof(Py_ssize_t): */ __pyx_v_ndim = PyArray_NDIM(__pyx_v_self); /* "numpy.pxd":207 * ndim = PyArray_NDIM(self) * * if sizeof(npy_intp) != sizeof(Py_ssize_t): # <<<<<<<<<<<<<< * copy_shape = 1 * else: */ __pyx_t_1 = ((sizeof(npy_intp)) != (sizeof(Py_ssize_t))); if (__pyx_t_1) { /* "numpy.pxd":208 * * if sizeof(npy_intp) != sizeof(Py_ssize_t): * copy_shape = 1 # <<<<<<<<<<<<<< * else: * copy_shape = 0 */ __pyx_v_copy_shape = 1; goto __pyx_L4; } /*else*/ { /* "numpy.pxd":210 * copy_shape = 1 * else: * copy_shape = 0 # <<<<<<<<<<<<<< * * if ((flags & pybuf.PyBUF_C_CONTIGUOUS == pybuf.PyBUF_C_CONTIGUOUS) */ __pyx_v_copy_shape = 0; } __pyx_L4:; /* "numpy.pxd":212 * copy_shape = 0 * * if ((flags & pybuf.PyBUF_C_CONTIGUOUS == pybuf.PyBUF_C_CONTIGUOUS) # <<<<<<<<<<<<<< * and not PyArray_CHKFLAGS(self, NPY_C_CONTIGUOUS)): * raise ValueError(u"ndarray is not C contiguous") */ __pyx_t_1 = ((__pyx_v_flags & PyBUF_C_CONTIGUOUS) == PyBUF_C_CONTIGUOUS); if (__pyx_t_1) { /* "numpy.pxd":213 * * if ((flags & pybuf.PyBUF_C_CONTIGUOUS == pybuf.PyBUF_C_CONTIGUOUS) * and not PyArray_CHKFLAGS(self, NPY_C_CONTIGUOUS)): # <<<<<<<<<<<<<< * raise ValueError(u"ndarray is not C contiguous") * */ __pyx_t_2 = (!PyArray_CHKFLAGS(__pyx_v_self, NPY_C_CONTIGUOUS)); __pyx_t_3 = __pyx_t_2; } else { __pyx_t_3 = __pyx_t_1; } if (__pyx_t_3) { /* "numpy.pxd":214 * if ((flags & pybuf.PyBUF_C_CONTIGUOUS == pybuf.PyBUF_C_CONTIGUOUS) * and not PyArray_CHKFLAGS(self, NPY_C_CONTIGUOUS)): * raise ValueError(u"ndarray is not C contiguous") # <<<<<<<<<<<<<< * * if ((flags & pybuf.PyBUF_F_CONTIGUOUS == pybuf.PyBUF_F_CONTIGUOUS) */ __pyx_t_4 = PyObject_Call(__pyx_builtin_ValueError, ((PyObject *)__pyx_k_tuple_2), NULL); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 214; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_4); __Pyx_Raise(__pyx_t_4, 0, 0, 0); __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0; {__pyx_filename = __pyx_f[2]; __pyx_lineno = 214; __pyx_clineno = __LINE__; goto __pyx_L1_error;} goto __pyx_L5; } __pyx_L5:; /* "numpy.pxd":216 * raise ValueError(u"ndarray is not C contiguous") * * if ((flags & pybuf.PyBUF_F_CONTIGUOUS == pybuf.PyBUF_F_CONTIGUOUS) # <<<<<<<<<<<<<< * and not PyArray_CHKFLAGS(self, NPY_F_CONTIGUOUS)): * raise ValueError(u"ndarray is not Fortran contiguous") */ __pyx_t_3 = ((__pyx_v_flags & PyBUF_F_CONTIGUOUS) == PyBUF_F_CONTIGUOUS); if (__pyx_t_3) { /* "numpy.pxd":217 * * if ((flags & pybuf.PyBUF_F_CONTIGUOUS == pybuf.PyBUF_F_CONTIGUOUS) * and not PyArray_CHKFLAGS(self, NPY_F_CONTIGUOUS)): # <<<<<<<<<<<<<< * raise ValueError(u"ndarray is not Fortran contiguous") * */ __pyx_t_1 = (!PyArray_CHKFLAGS(__pyx_v_self, NPY_F_CONTIGUOUS)); __pyx_t_2 = __pyx_t_1; } else { __pyx_t_2 = __pyx_t_3; } if (__pyx_t_2) { /* "numpy.pxd":218 * if ((flags & pybuf.PyBUF_F_CONTIGUOUS == pybuf.PyBUF_F_CONTIGUOUS) * and not PyArray_CHKFLAGS(self, NPY_F_CONTIGUOUS)): * raise ValueError(u"ndarray is not Fortran contiguous") # <<<<<<<<<<<<<< * * info.buf = PyArray_DATA(self) */ __pyx_t_4 = PyObject_Call(__pyx_builtin_ValueError, ((PyObject *)__pyx_k_tuple_4), NULL); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 218; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_4); __Pyx_Raise(__pyx_t_4, 0, 0, 0); __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0; {__pyx_filename = __pyx_f[2]; __pyx_lineno = 218; __pyx_clineno = __LINE__; goto __pyx_L1_error;} goto __pyx_L6; } __pyx_L6:; /* "numpy.pxd":220 * raise ValueError(u"ndarray is not Fortran contiguous") * * info.buf = PyArray_DATA(self) # <<<<<<<<<<<<<< * info.ndim = ndim * if copy_shape: */ __pyx_v_info->buf = PyArray_DATA(__pyx_v_self); /* "numpy.pxd":221 * * info.buf = PyArray_DATA(self) * info.ndim = ndim # <<<<<<<<<<<<<< * if copy_shape: * # Allocate new buffer for strides and shape info. */ __pyx_v_info->ndim = __pyx_v_ndim; /* "numpy.pxd":222 * info.buf = PyArray_DATA(self) * info.ndim = ndim * if copy_shape: # <<<<<<<<<<<<<< * # Allocate new buffer for strides and shape info. * # This is allocated as one block, strides first. */ if (__pyx_v_copy_shape) { /* "numpy.pxd":225 * # Allocate new buffer for strides and shape info. * # This is allocated as one block, strides first. * info.strides = stdlib.malloc(sizeof(Py_ssize_t) * ndim * 2) # <<<<<<<<<<<<<< * info.shape = info.strides + ndim * for i in range(ndim): */ __pyx_v_info->strides = ((Py_ssize_t *)malloc((((sizeof(Py_ssize_t)) * ((size_t)__pyx_v_ndim)) * 2))); /* "numpy.pxd":226 * # This is allocated as one block, strides first. * info.strides = stdlib.malloc(sizeof(Py_ssize_t) * ndim * 2) * info.shape = info.strides + ndim # <<<<<<<<<<<<<< * for i in range(ndim): * info.strides[i] = PyArray_STRIDES(self)[i] */ __pyx_v_info->shape = (__pyx_v_info->strides + __pyx_v_ndim); /* "numpy.pxd":227 * info.strides = stdlib.malloc(sizeof(Py_ssize_t) * ndim * 2) * info.shape = info.strides + ndim * for i in range(ndim): # <<<<<<<<<<<<<< * info.strides[i] = PyArray_STRIDES(self)[i] * info.shape[i] = PyArray_DIMS(self)[i] */ __pyx_t_5 = __pyx_v_ndim; for (__pyx_t_6 = 0; __pyx_t_6 < __pyx_t_5; __pyx_t_6+=1) { __pyx_v_i = __pyx_t_6; /* "numpy.pxd":228 * info.shape = info.strides + ndim * for i in range(ndim): * info.strides[i] = PyArray_STRIDES(self)[i] # <<<<<<<<<<<<<< * info.shape[i] = PyArray_DIMS(self)[i] * else: */ (__pyx_v_info->strides[__pyx_v_i]) = (PyArray_STRIDES(__pyx_v_self)[__pyx_v_i]); /* "numpy.pxd":229 * for i in range(ndim): * info.strides[i] = PyArray_STRIDES(self)[i] * info.shape[i] = PyArray_DIMS(self)[i] # <<<<<<<<<<<<<< * else: * info.strides = PyArray_STRIDES(self) */ (__pyx_v_info->shape[__pyx_v_i]) = (PyArray_DIMS(__pyx_v_self)[__pyx_v_i]); } goto __pyx_L7; } /*else*/ { /* "numpy.pxd":231 * info.shape[i] = PyArray_DIMS(self)[i] * else: * info.strides = PyArray_STRIDES(self) # <<<<<<<<<<<<<< * info.shape = PyArray_DIMS(self) * info.suboffsets = NULL */ __pyx_v_info->strides = ((Py_ssize_t *)PyArray_STRIDES(__pyx_v_self)); /* "numpy.pxd":232 * else: * info.strides = PyArray_STRIDES(self) * info.shape = PyArray_DIMS(self) # <<<<<<<<<<<<<< * info.suboffsets = NULL * info.itemsize = PyArray_ITEMSIZE(self) */ __pyx_v_info->shape = ((Py_ssize_t *)PyArray_DIMS(__pyx_v_self)); } __pyx_L7:; /* "numpy.pxd":233 * info.strides = PyArray_STRIDES(self) * info.shape = PyArray_DIMS(self) * info.suboffsets = NULL # <<<<<<<<<<<<<< * info.itemsize = PyArray_ITEMSIZE(self) * info.readonly = not PyArray_ISWRITEABLE(self) */ __pyx_v_info->suboffsets = NULL; /* "numpy.pxd":234 * info.shape = PyArray_DIMS(self) * info.suboffsets = NULL * info.itemsize = PyArray_ITEMSIZE(self) # <<<<<<<<<<<<<< * info.readonly = not PyArray_ISWRITEABLE(self) * */ __pyx_v_info->itemsize = PyArray_ITEMSIZE(__pyx_v_self); /* "numpy.pxd":235 * info.suboffsets = NULL * info.itemsize = PyArray_ITEMSIZE(self) * info.readonly = not PyArray_ISWRITEABLE(self) # <<<<<<<<<<<<<< * * cdef int t */ __pyx_v_info->readonly = (!PyArray_ISWRITEABLE(__pyx_v_self)); /* "numpy.pxd":238 * * cdef int t * cdef char* f = NULL # <<<<<<<<<<<<<< * cdef dtype descr = self.descr * cdef list stack */ __pyx_v_f = NULL; /* "numpy.pxd":239 * cdef int t * cdef char* f = NULL * cdef dtype descr = self.descr # <<<<<<<<<<<<<< * cdef list stack * cdef int offset */ __Pyx_INCREF(((PyObject *)__pyx_v_self->descr)); __pyx_v_descr = __pyx_v_self->descr; /* "numpy.pxd":243 * cdef int offset * * cdef bint hasfields = PyDataType_HASFIELDS(descr) # <<<<<<<<<<<<<< * * if not hasfields and not copy_shape: */ __pyx_v_hasfields = PyDataType_HASFIELDS(__pyx_v_descr); /* "numpy.pxd":245 * cdef bint hasfields = PyDataType_HASFIELDS(descr) * * if not hasfields and not copy_shape: # <<<<<<<<<<<<<< * # do not call releasebuffer * info.obj = None */ __pyx_t_2 = (!__pyx_v_hasfields); if (__pyx_t_2) { __pyx_t_3 = (!__pyx_v_copy_shape); __pyx_t_1 = __pyx_t_3; } else { __pyx_t_1 = __pyx_t_2; } if (__pyx_t_1) { /* "numpy.pxd":247 * if not hasfields and not copy_shape: * # do not call releasebuffer * info.obj = None # <<<<<<<<<<<<<< * else: * # need to call releasebuffer */ __Pyx_INCREF(Py_None); __Pyx_GIVEREF(Py_None); __Pyx_GOTREF(__pyx_v_info->obj); __Pyx_DECREF(__pyx_v_info->obj); __pyx_v_info->obj = Py_None; goto __pyx_L10; } /*else*/ { /* "numpy.pxd":250 * else: * # need to call releasebuffer * info.obj = self # <<<<<<<<<<<<<< * * if not hasfields: */ __Pyx_INCREF(((PyObject *)__pyx_v_self)); __Pyx_GIVEREF(((PyObject *)__pyx_v_self)); __Pyx_GOTREF(__pyx_v_info->obj); __Pyx_DECREF(__pyx_v_info->obj); __pyx_v_info->obj = ((PyObject *)__pyx_v_self); } __pyx_L10:; /* "numpy.pxd":252 * info.obj = self * * if not hasfields: # <<<<<<<<<<<<<< * t = descr.type_num * if ((descr.byteorder == '>' and little_endian) or */ __pyx_t_1 = (!__pyx_v_hasfields); if (__pyx_t_1) { /* "numpy.pxd":253 * * if not hasfields: * t = descr.type_num # <<<<<<<<<<<<<< * if ((descr.byteorder == '>' and little_endian) or * (descr.byteorder == '<' and not little_endian)): */ __pyx_v_t = __pyx_v_descr->type_num; /* "numpy.pxd":254 * if not hasfields: * t = descr.type_num * if ((descr.byteorder == '>' and little_endian) or # <<<<<<<<<<<<<< * (descr.byteorder == '<' and not little_endian)): * raise ValueError(u"Non-native byte order not supported") */ __pyx_t_1 = (__pyx_v_descr->byteorder == '>'); if (__pyx_t_1) { __pyx_t_2 = __pyx_v_little_endian; } else { __pyx_t_2 = __pyx_t_1; } if (!__pyx_t_2) { /* "numpy.pxd":255 * t = descr.type_num * if ((descr.byteorder == '>' and little_endian) or * (descr.byteorder == '<' and not little_endian)): # <<<<<<<<<<<<<< * raise ValueError(u"Non-native byte order not supported") * if t == NPY_BYTE: f = "b" */ __pyx_t_1 = (__pyx_v_descr->byteorder == '<'); if (__pyx_t_1) { __pyx_t_3 = (!__pyx_v_little_endian); __pyx_t_7 = __pyx_t_3; } else { __pyx_t_7 = __pyx_t_1; } __pyx_t_1 = __pyx_t_7; } else { __pyx_t_1 = __pyx_t_2; } if (__pyx_t_1) { /* "numpy.pxd":256 * if ((descr.byteorder == '>' and little_endian) or * (descr.byteorder == '<' and not little_endian)): * raise ValueError(u"Non-native byte order not supported") # <<<<<<<<<<<<<< * if t == NPY_BYTE: f = "b" * elif t == NPY_UBYTE: f = "B" */ __pyx_t_4 = PyObject_Call(__pyx_builtin_ValueError, ((PyObject *)__pyx_k_tuple_6), NULL); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 256; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_4); __Pyx_Raise(__pyx_t_4, 0, 0, 0); __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0; {__pyx_filename = __pyx_f[2]; __pyx_lineno = 256; __pyx_clineno = __LINE__; goto __pyx_L1_error;} goto __pyx_L12; } __pyx_L12:; /* "numpy.pxd":257 * (descr.byteorder == '<' and not little_endian)): * raise ValueError(u"Non-native byte order not supported") * if t == NPY_BYTE: f = "b" # <<<<<<<<<<<<<< * elif t == NPY_UBYTE: f = "B" * elif t == NPY_SHORT: f = "h" */ __pyx_t_1 = (__pyx_v_t == NPY_BYTE); if (__pyx_t_1) { __pyx_v_f = __pyx_k__b; goto __pyx_L13; } /* "numpy.pxd":258 * raise ValueError(u"Non-native byte order not supported") * if t == NPY_BYTE: f = "b" * elif t == NPY_UBYTE: f = "B" # <<<<<<<<<<<<<< * elif t == NPY_SHORT: f = "h" * elif t == NPY_USHORT: f = "H" */ __pyx_t_1 = (__pyx_v_t == NPY_UBYTE); if (__pyx_t_1) { __pyx_v_f = __pyx_k__B; goto __pyx_L13; } /* "numpy.pxd":259 * if t == NPY_BYTE: f = "b" * elif t == NPY_UBYTE: f = "B" * elif t == NPY_SHORT: f = "h" # <<<<<<<<<<<<<< * elif t == NPY_USHORT: f = "H" * elif t == NPY_INT: f = "i" */ __pyx_t_1 = (__pyx_v_t == NPY_SHORT); if (__pyx_t_1) { __pyx_v_f = __pyx_k__h; goto __pyx_L13; } /* "numpy.pxd":260 * elif t == NPY_UBYTE: f = "B" * elif t == NPY_SHORT: f = "h" * elif t == NPY_USHORT: f = "H" # <<<<<<<<<<<<<< * elif t == NPY_INT: f = "i" * elif t == NPY_UINT: f = "I" */ __pyx_t_1 = (__pyx_v_t == NPY_USHORT); if (__pyx_t_1) { __pyx_v_f = __pyx_k__H; goto __pyx_L13; } /* "numpy.pxd":261 * elif t == NPY_SHORT: f = "h" * elif t == NPY_USHORT: f = "H" * elif t == NPY_INT: f = "i" # <<<<<<<<<<<<<< * elif t == NPY_UINT: f = "I" * elif t == NPY_LONG: f = "l" */ __pyx_t_1 = (__pyx_v_t == NPY_INT); if (__pyx_t_1) { __pyx_v_f = __pyx_k__i; goto __pyx_L13; } /* "numpy.pxd":262 * elif t == NPY_USHORT: f = "H" * elif t == NPY_INT: f = "i" * elif t == NPY_UINT: f = "I" # <<<<<<<<<<<<<< * elif t == NPY_LONG: f = "l" * elif t == NPY_ULONG: f = "L" */ __pyx_t_1 = (__pyx_v_t == NPY_UINT); if (__pyx_t_1) { __pyx_v_f = __pyx_k__I; goto __pyx_L13; } /* "numpy.pxd":263 * elif t == NPY_INT: f = "i" * elif t == NPY_UINT: f = "I" * elif t == NPY_LONG: f = "l" # <<<<<<<<<<<<<< * elif t == NPY_ULONG: f = "L" * elif t == NPY_LONGLONG: f = "q" */ __pyx_t_1 = (__pyx_v_t == NPY_LONG); if (__pyx_t_1) { __pyx_v_f = __pyx_k__l; goto __pyx_L13; } /* "numpy.pxd":264 * elif t == NPY_UINT: f = "I" * elif t == NPY_LONG: f = "l" * elif t == NPY_ULONG: f = "L" # <<<<<<<<<<<<<< * elif t == NPY_LONGLONG: f = "q" * elif t == NPY_ULONGLONG: f = "Q" */ __pyx_t_1 = (__pyx_v_t == NPY_ULONG); if (__pyx_t_1) { __pyx_v_f = __pyx_k__L; goto __pyx_L13; } /* "numpy.pxd":265 * elif t == NPY_LONG: f = "l" * elif t == NPY_ULONG: f = "L" * elif t == NPY_LONGLONG: f = "q" # <<<<<<<<<<<<<< * elif t == NPY_ULONGLONG: f = "Q" * elif t == NPY_FLOAT: f = "f" */ __pyx_t_1 = (__pyx_v_t == NPY_LONGLONG); if (__pyx_t_1) { __pyx_v_f = __pyx_k__q; goto __pyx_L13; } /* "numpy.pxd":266 * elif t == NPY_ULONG: f = "L" * elif t == NPY_LONGLONG: f = "q" * elif t == NPY_ULONGLONG: f = "Q" # <<<<<<<<<<<<<< * elif t == NPY_FLOAT: f = "f" * elif t == NPY_DOUBLE: f = "d" */ __pyx_t_1 = (__pyx_v_t == NPY_ULONGLONG); if (__pyx_t_1) { __pyx_v_f = __pyx_k__Q; goto __pyx_L13; } /* "numpy.pxd":267 * elif t == NPY_LONGLONG: f = "q" * elif t == NPY_ULONGLONG: f = "Q" * elif t == NPY_FLOAT: f = "f" # <<<<<<<<<<<<<< * elif t == NPY_DOUBLE: f = "d" * elif t == NPY_LONGDOUBLE: f = "g" */ __pyx_t_1 = (__pyx_v_t == NPY_FLOAT); if (__pyx_t_1) { __pyx_v_f = __pyx_k__f; goto __pyx_L13; } /* "numpy.pxd":268 * elif t == NPY_ULONGLONG: f = "Q" * elif t == NPY_FLOAT: f = "f" * elif t == NPY_DOUBLE: f = "d" # <<<<<<<<<<<<<< * elif t == NPY_LONGDOUBLE: f = "g" * elif t == NPY_CFLOAT: f = "Zf" */ __pyx_t_1 = (__pyx_v_t == NPY_DOUBLE); if (__pyx_t_1) { __pyx_v_f = __pyx_k__d; goto __pyx_L13; } /* "numpy.pxd":269 * elif t == NPY_FLOAT: f = "f" * elif t == NPY_DOUBLE: f = "d" * elif t == NPY_LONGDOUBLE: f = "g" # <<<<<<<<<<<<<< * elif t == NPY_CFLOAT: f = "Zf" * elif t == NPY_CDOUBLE: f = "Zd" */ __pyx_t_1 = (__pyx_v_t == NPY_LONGDOUBLE); if (__pyx_t_1) { __pyx_v_f = __pyx_k__g; goto __pyx_L13; } /* "numpy.pxd":270 * elif t == NPY_DOUBLE: f = "d" * elif t == NPY_LONGDOUBLE: f = "g" * elif t == NPY_CFLOAT: f = "Zf" # <<<<<<<<<<<<<< * elif t == NPY_CDOUBLE: f = "Zd" * elif t == NPY_CLONGDOUBLE: f = "Zg" */ __pyx_t_1 = (__pyx_v_t == NPY_CFLOAT); if (__pyx_t_1) { __pyx_v_f = __pyx_k__Zf; goto __pyx_L13; } /* "numpy.pxd":271 * elif t == NPY_LONGDOUBLE: f = "g" * elif t == NPY_CFLOAT: f = "Zf" * elif t == NPY_CDOUBLE: f = "Zd" # <<<<<<<<<<<<<< * elif t == NPY_CLONGDOUBLE: f = "Zg" * elif t == NPY_OBJECT: f = "O" */ __pyx_t_1 = (__pyx_v_t == NPY_CDOUBLE); if (__pyx_t_1) { __pyx_v_f = __pyx_k__Zd; goto __pyx_L13; } /* "numpy.pxd":272 * elif t == NPY_CFLOAT: f = "Zf" * elif t == NPY_CDOUBLE: f = "Zd" * elif t == NPY_CLONGDOUBLE: f = "Zg" # <<<<<<<<<<<<<< * elif t == NPY_OBJECT: f = "O" * else: */ __pyx_t_1 = (__pyx_v_t == NPY_CLONGDOUBLE); if (__pyx_t_1) { __pyx_v_f = __pyx_k__Zg; goto __pyx_L13; } /* "numpy.pxd":273 * elif t == NPY_CDOUBLE: f = "Zd" * elif t == NPY_CLONGDOUBLE: f = "Zg" * elif t == NPY_OBJECT: f = "O" # <<<<<<<<<<<<<< * else: * raise ValueError(u"unknown dtype code in numpy.pxd (%d)" % t) */ __pyx_t_1 = (__pyx_v_t == NPY_OBJECT); if (__pyx_t_1) { __pyx_v_f = __pyx_k__O; goto __pyx_L13; } /*else*/ { /* "numpy.pxd":275 * elif t == NPY_OBJECT: f = "O" * else: * raise ValueError(u"unknown dtype code in numpy.pxd (%d)" % t) # <<<<<<<<<<<<<< * info.format = f * return */ __pyx_t_4 = PyInt_FromLong(__pyx_v_t); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 275; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_4); __pyx_t_8 = PyNumber_Remainder(((PyObject *)__pyx_kp_u_7), __pyx_t_4); if (unlikely(!__pyx_t_8)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 275; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(((PyObject *)__pyx_t_8)); __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0; __pyx_t_4 = PyTuple_New(1); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 275; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_4); PyTuple_SET_ITEM(__pyx_t_4, 0, ((PyObject *)__pyx_t_8)); __Pyx_GIVEREF(((PyObject *)__pyx_t_8)); __pyx_t_8 = 0; __pyx_t_8 = PyObject_Call(__pyx_builtin_ValueError, ((PyObject *)__pyx_t_4), NULL); if (unlikely(!__pyx_t_8)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 275; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_8); __Pyx_DECREF(((PyObject *)__pyx_t_4)); __pyx_t_4 = 0; __Pyx_Raise(__pyx_t_8, 0, 0, 0); __Pyx_DECREF(__pyx_t_8); __pyx_t_8 = 0; {__pyx_filename = __pyx_f[2]; __pyx_lineno = 275; __pyx_clineno = __LINE__; goto __pyx_L1_error;} } __pyx_L13:; /* "numpy.pxd":276 * else: * raise ValueError(u"unknown dtype code in numpy.pxd (%d)" % t) * info.format = f # <<<<<<<<<<<<<< * return * else: */ __pyx_v_info->format = __pyx_v_f; /* "numpy.pxd":277 * raise ValueError(u"unknown dtype code in numpy.pxd (%d)" % t) * info.format = f * return # <<<<<<<<<<<<<< * else: * info.format = stdlib.malloc(_buffer_format_string_len) */ __pyx_r = 0; goto __pyx_L0; goto __pyx_L11; } /*else*/ { /* "numpy.pxd":279 * return * else: * info.format = stdlib.malloc(_buffer_format_string_len) # <<<<<<<<<<<<<< * info.format[0] = '^' # Native data types, manual alignment * offset = 0 */ __pyx_v_info->format = ((char *)malloc(255)); /* "numpy.pxd":280 * else: * info.format = stdlib.malloc(_buffer_format_string_len) * info.format[0] = '^' # Native data types, manual alignment # <<<<<<<<<<<<<< * offset = 0 * f = _util_dtypestring(descr, info.format + 1, */ (__pyx_v_info->format[0]) = '^'; /* "numpy.pxd":281 * info.format = stdlib.malloc(_buffer_format_string_len) * info.format[0] = '^' # Native data types, manual alignment * offset = 0 # <<<<<<<<<<<<<< * f = _util_dtypestring(descr, info.format + 1, * info.format + _buffer_format_string_len, */ __pyx_v_offset = 0; /* "numpy.pxd":284 * f = _util_dtypestring(descr, info.format + 1, * info.format + _buffer_format_string_len, * &offset) # <<<<<<<<<<<<<< * f[0] = 0 # Terminate format string * */ __pyx_t_9 = __pyx_f_5numpy__util_dtypestring(__pyx_v_descr, (__pyx_v_info->format + 1), (__pyx_v_info->format + 255), (&__pyx_v_offset)); if (unlikely(__pyx_t_9 == NULL)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 282; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __pyx_v_f = __pyx_t_9; /* "numpy.pxd":285 * info.format + _buffer_format_string_len, * &offset) * f[0] = 0 # Terminate format string # <<<<<<<<<<<<<< * * def __releasebuffer__(ndarray self, Py_buffer* info): */ (__pyx_v_f[0]) = 0; } __pyx_L11:; __pyx_r = 0; goto __pyx_L0; __pyx_L1_error:; __Pyx_XDECREF(__pyx_t_4); __Pyx_XDECREF(__pyx_t_8); __Pyx_AddTraceback("numpy.ndarray.__getbuffer__", __pyx_clineno, __pyx_lineno, __pyx_filename); __pyx_r = -1; if (__pyx_v_info != NULL && __pyx_v_info->obj != NULL) { __Pyx_GOTREF(__pyx_v_info->obj); __Pyx_DECREF(__pyx_v_info->obj); __pyx_v_info->obj = NULL; } goto __pyx_L2; __pyx_L0:; if (__pyx_v_info != NULL && __pyx_v_info->obj == Py_None) { __Pyx_GOTREF(Py_None); __Pyx_DECREF(Py_None); __pyx_v_info->obj = NULL; } __pyx_L2:; __Pyx_XDECREF((PyObject *)__pyx_v_descr); __Pyx_RefNannyFinishContext(); return __pyx_r; } /* Python wrapper */ static void __pyx_pw_5numpy_7ndarray_3__releasebuffer__(PyObject *__pyx_v_self, Py_buffer *__pyx_v_info); /*proto*/ static void __pyx_pw_5numpy_7ndarray_3__releasebuffer__(PyObject *__pyx_v_self, Py_buffer *__pyx_v_info) { __Pyx_RefNannyDeclarations __Pyx_RefNannySetupContext("__releasebuffer__ (wrapper)"); __pyx_pf_5numpy_7ndarray_2__releasebuffer__(((PyArrayObject *)__pyx_v_self), ((Py_buffer *)__pyx_v_info)); __Pyx_RefNannyFinishContext(); } /* "numpy.pxd":287 * f[0] = 0 # Terminate format string * * def __releasebuffer__(ndarray self, Py_buffer* info): # <<<<<<<<<<<<<< * if PyArray_HASFIELDS(self): * stdlib.free(info.format) */ static void __pyx_pf_5numpy_7ndarray_2__releasebuffer__(PyArrayObject *__pyx_v_self, Py_buffer *__pyx_v_info) { __Pyx_RefNannyDeclarations int __pyx_t_1; __Pyx_RefNannySetupContext("__releasebuffer__"); /* "numpy.pxd":288 * * def __releasebuffer__(ndarray self, Py_buffer* info): * if PyArray_HASFIELDS(self): # <<<<<<<<<<<<<< * stdlib.free(info.format) * if sizeof(npy_intp) != sizeof(Py_ssize_t): */ __pyx_t_1 = PyArray_HASFIELDS(__pyx_v_self); if (__pyx_t_1) { /* "numpy.pxd":289 * def __releasebuffer__(ndarray self, Py_buffer* info): * if PyArray_HASFIELDS(self): * stdlib.free(info.format) # <<<<<<<<<<<<<< * if sizeof(npy_intp) != sizeof(Py_ssize_t): * stdlib.free(info.strides) */ free(__pyx_v_info->format); goto __pyx_L3; } __pyx_L3:; /* "numpy.pxd":290 * if PyArray_HASFIELDS(self): * stdlib.free(info.format) * if sizeof(npy_intp) != sizeof(Py_ssize_t): # <<<<<<<<<<<<<< * stdlib.free(info.strides) * # info.shape was stored after info.strides in the same block */ __pyx_t_1 = ((sizeof(npy_intp)) != (sizeof(Py_ssize_t))); if (__pyx_t_1) { /* "numpy.pxd":291 * stdlib.free(info.format) * if sizeof(npy_intp) != sizeof(Py_ssize_t): * stdlib.free(info.strides) # <<<<<<<<<<<<<< * # info.shape was stored after info.strides in the same block * */ free(__pyx_v_info->strides); goto __pyx_L4; } __pyx_L4:; __Pyx_RefNannyFinishContext(); } /* "numpy.pxd":767 * ctypedef npy_cdouble complex_t * * cdef inline object PyArray_MultiIterNew1(a): # <<<<<<<<<<<<<< * return PyArray_MultiIterNew(1, a) * */ static CYTHON_INLINE PyObject *__pyx_f_5numpy_PyArray_MultiIterNew1(PyObject *__pyx_v_a) { PyObject *__pyx_r = NULL; __Pyx_RefNannyDeclarations PyObject *__pyx_t_1 = NULL; int __pyx_lineno = 0; const char *__pyx_filename = NULL; int __pyx_clineno = 0; __Pyx_RefNannySetupContext("PyArray_MultiIterNew1"); /* "numpy.pxd":768 * * cdef inline object PyArray_MultiIterNew1(a): * return PyArray_MultiIterNew(1, a) # <<<<<<<<<<<<<< * * cdef inline object PyArray_MultiIterNew2(a, b): */ __Pyx_XDECREF(__pyx_r); __pyx_t_1 = PyArray_MultiIterNew(1, ((void *)__pyx_v_a)); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 768; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_1); __pyx_r = __pyx_t_1; __pyx_t_1 = 0; goto __pyx_L0; __pyx_r = Py_None; __Pyx_INCREF(Py_None); goto __pyx_L0; __pyx_L1_error:; __Pyx_XDECREF(__pyx_t_1); __Pyx_AddTraceback("numpy.PyArray_MultiIterNew1", __pyx_clineno, __pyx_lineno, __pyx_filename); __pyx_r = 0; __pyx_L0:; __Pyx_XGIVEREF(__pyx_r); __Pyx_RefNannyFinishContext(); return __pyx_r; } /* "numpy.pxd":770 * return PyArray_MultiIterNew(1, a) * * cdef inline object PyArray_MultiIterNew2(a, b): # <<<<<<<<<<<<<< * return PyArray_MultiIterNew(2, a, b) * */ static CYTHON_INLINE PyObject *__pyx_f_5numpy_PyArray_MultiIterNew2(PyObject *__pyx_v_a, PyObject *__pyx_v_b) { PyObject *__pyx_r = NULL; __Pyx_RefNannyDeclarations PyObject *__pyx_t_1 = NULL; int __pyx_lineno = 0; const char *__pyx_filename = NULL; int __pyx_clineno = 0; __Pyx_RefNannySetupContext("PyArray_MultiIterNew2"); /* "numpy.pxd":771 * * cdef inline object PyArray_MultiIterNew2(a, b): * return PyArray_MultiIterNew(2, a, b) # <<<<<<<<<<<<<< * * cdef inline object PyArray_MultiIterNew3(a, b, c): */ __Pyx_XDECREF(__pyx_r); __pyx_t_1 = PyArray_MultiIterNew(2, ((void *)__pyx_v_a), ((void *)__pyx_v_b)); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 771; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_1); __pyx_r = __pyx_t_1; __pyx_t_1 = 0; goto __pyx_L0; __pyx_r = Py_None; __Pyx_INCREF(Py_None); goto __pyx_L0; __pyx_L1_error:; __Pyx_XDECREF(__pyx_t_1); __Pyx_AddTraceback("numpy.PyArray_MultiIterNew2", __pyx_clineno, __pyx_lineno, __pyx_filename); __pyx_r = 0; __pyx_L0:; __Pyx_XGIVEREF(__pyx_r); __Pyx_RefNannyFinishContext(); return __pyx_r; } /* "numpy.pxd":773 * return PyArray_MultiIterNew(2, a, b) * * cdef inline object PyArray_MultiIterNew3(a, b, c): # <<<<<<<<<<<<<< * return PyArray_MultiIterNew(3, a, b, c) * */ static CYTHON_INLINE PyObject *__pyx_f_5numpy_PyArray_MultiIterNew3(PyObject *__pyx_v_a, PyObject *__pyx_v_b, PyObject *__pyx_v_c) { PyObject *__pyx_r = NULL; __Pyx_RefNannyDeclarations PyObject *__pyx_t_1 = NULL; int __pyx_lineno = 0; const char *__pyx_filename = NULL; int __pyx_clineno = 0; __Pyx_RefNannySetupContext("PyArray_MultiIterNew3"); /* "numpy.pxd":774 * * cdef inline object PyArray_MultiIterNew3(a, b, c): * return PyArray_MultiIterNew(3, a, b, c) # <<<<<<<<<<<<<< * * cdef inline object PyArray_MultiIterNew4(a, b, c, d): */ __Pyx_XDECREF(__pyx_r); __pyx_t_1 = PyArray_MultiIterNew(3, ((void *)__pyx_v_a), ((void *)__pyx_v_b), ((void *)__pyx_v_c)); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 774; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_1); __pyx_r = __pyx_t_1; __pyx_t_1 = 0; goto __pyx_L0; __pyx_r = Py_None; __Pyx_INCREF(Py_None); goto __pyx_L0; __pyx_L1_error:; __Pyx_XDECREF(__pyx_t_1); __Pyx_AddTraceback("numpy.PyArray_MultiIterNew3", __pyx_clineno, __pyx_lineno, __pyx_filename); __pyx_r = 0; __pyx_L0:; __Pyx_XGIVEREF(__pyx_r); __Pyx_RefNannyFinishContext(); return __pyx_r; } /* "numpy.pxd":776 * return PyArray_MultiIterNew(3, a, b, c) * * cdef inline object PyArray_MultiIterNew4(a, b, c, d): # <<<<<<<<<<<<<< * return PyArray_MultiIterNew(4, a, b, c, d) * */ static CYTHON_INLINE PyObject *__pyx_f_5numpy_PyArray_MultiIterNew4(PyObject *__pyx_v_a, PyObject *__pyx_v_b, PyObject *__pyx_v_c, PyObject *__pyx_v_d) { PyObject *__pyx_r = NULL; __Pyx_RefNannyDeclarations PyObject *__pyx_t_1 = NULL; int __pyx_lineno = 0; const char *__pyx_filename = NULL; int __pyx_clineno = 0; __Pyx_RefNannySetupContext("PyArray_MultiIterNew4"); /* "numpy.pxd":777 * * cdef inline object PyArray_MultiIterNew4(a, b, c, d): * return PyArray_MultiIterNew(4, a, b, c, d) # <<<<<<<<<<<<<< * * cdef inline object PyArray_MultiIterNew5(a, b, c, d, e): */ __Pyx_XDECREF(__pyx_r); __pyx_t_1 = PyArray_MultiIterNew(4, ((void *)__pyx_v_a), ((void *)__pyx_v_b), ((void *)__pyx_v_c), ((void *)__pyx_v_d)); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 777; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_1); __pyx_r = __pyx_t_1; __pyx_t_1 = 0; goto __pyx_L0; __pyx_r = Py_None; __Pyx_INCREF(Py_None); goto __pyx_L0; __pyx_L1_error:; __Pyx_XDECREF(__pyx_t_1); __Pyx_AddTraceback("numpy.PyArray_MultiIterNew4", __pyx_clineno, __pyx_lineno, __pyx_filename); __pyx_r = 0; __pyx_L0:; __Pyx_XGIVEREF(__pyx_r); __Pyx_RefNannyFinishContext(); return __pyx_r; } /* "numpy.pxd":779 * return PyArray_MultiIterNew(4, a, b, c, d) * * cdef inline object PyArray_MultiIterNew5(a, b, c, d, e): # <<<<<<<<<<<<<< * return PyArray_MultiIterNew(5, a, b, c, d, e) * */ static CYTHON_INLINE PyObject *__pyx_f_5numpy_PyArray_MultiIterNew5(PyObject *__pyx_v_a, PyObject *__pyx_v_b, PyObject *__pyx_v_c, PyObject *__pyx_v_d, PyObject *__pyx_v_e) { PyObject *__pyx_r = NULL; __Pyx_RefNannyDeclarations PyObject *__pyx_t_1 = NULL; int __pyx_lineno = 0; const char *__pyx_filename = NULL; int __pyx_clineno = 0; __Pyx_RefNannySetupContext("PyArray_MultiIterNew5"); /* "numpy.pxd":780 * * cdef inline object PyArray_MultiIterNew5(a, b, c, d, e): * return PyArray_MultiIterNew(5, a, b, c, d, e) # <<<<<<<<<<<<<< * * cdef inline char* _util_dtypestring(dtype descr, char* f, char* end, int* offset) except NULL: */ __Pyx_XDECREF(__pyx_r); __pyx_t_1 = PyArray_MultiIterNew(5, ((void *)__pyx_v_a), ((void *)__pyx_v_b), ((void *)__pyx_v_c), ((void *)__pyx_v_d), ((void *)__pyx_v_e)); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 780; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_1); __pyx_r = __pyx_t_1; __pyx_t_1 = 0; goto __pyx_L0; __pyx_r = Py_None; __Pyx_INCREF(Py_None); goto __pyx_L0; __pyx_L1_error:; __Pyx_XDECREF(__pyx_t_1); __Pyx_AddTraceback("numpy.PyArray_MultiIterNew5", __pyx_clineno, __pyx_lineno, __pyx_filename); __pyx_r = 0; __pyx_L0:; __Pyx_XGIVEREF(__pyx_r); __Pyx_RefNannyFinishContext(); return __pyx_r; } /* "numpy.pxd":782 * return PyArray_MultiIterNew(5, a, b, c, d, e) * * cdef inline char* _util_dtypestring(dtype descr, char* f, char* end, int* offset) except NULL: # <<<<<<<<<<<<<< * # Recursive utility function used in __getbuffer__ to get format * # string. The new location in the format string is returned. */ static CYTHON_INLINE char *__pyx_f_5numpy__util_dtypestring(PyArray_Descr *__pyx_v_descr, char *__pyx_v_f, char *__pyx_v_end, int *__pyx_v_offset) { PyArray_Descr *__pyx_v_child = 0; int __pyx_v_endian_detector; int __pyx_v_little_endian; PyObject *__pyx_v_fields = 0; PyObject *__pyx_v_childname = NULL; PyObject *__pyx_v_new_offset = NULL; PyObject *__pyx_v_t = NULL; char *__pyx_r; __Pyx_RefNannyDeclarations PyObject *__pyx_t_1 = NULL; Py_ssize_t __pyx_t_2; PyObject *__pyx_t_3 = NULL; PyObject *__pyx_t_4 = NULL; PyObject *__pyx_t_5 = NULL; int __pyx_t_6; int __pyx_t_7; int __pyx_t_8; int __pyx_t_9; long __pyx_t_10; char *__pyx_t_11; int __pyx_lineno = 0; const char *__pyx_filename = NULL; int __pyx_clineno = 0; __Pyx_RefNannySetupContext("_util_dtypestring"); /* "numpy.pxd":789 * cdef int delta_offset * cdef tuple i * cdef int endian_detector = 1 # <<<<<<<<<<<<<< * cdef bint little_endian = ((&endian_detector)[0] != 0) * cdef tuple fields */ __pyx_v_endian_detector = 1; /* "numpy.pxd":790 * cdef tuple i * cdef int endian_detector = 1 * cdef bint little_endian = ((&endian_detector)[0] != 0) # <<<<<<<<<<<<<< * cdef tuple fields * */ __pyx_v_little_endian = ((((char *)(&__pyx_v_endian_detector))[0]) != 0); /* "numpy.pxd":793 * cdef tuple fields * * for childname in descr.names: # <<<<<<<<<<<<<< * fields = descr.fields[childname] * child, new_offset = fields */ if (unlikely(((PyObject *)__pyx_v_descr->names) == Py_None)) { PyErr_SetString(PyExc_TypeError, "'NoneType' object is not iterable"); {__pyx_filename = __pyx_f[2]; __pyx_lineno = 793; __pyx_clineno = __LINE__; goto __pyx_L1_error;} } __pyx_t_1 = ((PyObject *)__pyx_v_descr->names); __Pyx_INCREF(__pyx_t_1); __pyx_t_2 = 0; for (;;) { if (__pyx_t_2 >= PyTuple_GET_SIZE(__pyx_t_1)) break; __pyx_t_3 = PyTuple_GET_ITEM(__pyx_t_1, __pyx_t_2); __Pyx_INCREF(__pyx_t_3); __pyx_t_2++; __Pyx_XDECREF(__pyx_v_childname); __pyx_v_childname = __pyx_t_3; __pyx_t_3 = 0; /* "numpy.pxd":794 * * for childname in descr.names: * fields = descr.fields[childname] # <<<<<<<<<<<<<< * child, new_offset = fields * */ __pyx_t_3 = PyObject_GetItem(__pyx_v_descr->fields, __pyx_v_childname); if (!__pyx_t_3) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 794; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); if (!(likely(PyTuple_CheckExact(__pyx_t_3))||((__pyx_t_3) == Py_None)||(PyErr_Format(PyExc_TypeError, "Expected tuple, got %.200s", Py_TYPE(__pyx_t_3)->tp_name), 0))) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 794; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_XDECREF(((PyObject *)__pyx_v_fields)); __pyx_v_fields = ((PyObject*)__pyx_t_3); __pyx_t_3 = 0; /* "numpy.pxd":795 * for childname in descr.names: * fields = descr.fields[childname] * child, new_offset = fields # <<<<<<<<<<<<<< * * if (end - f) - (new_offset - offset[0]) < 15: */ if (likely(PyTuple_CheckExact(((PyObject *)__pyx_v_fields)))) { PyObject* sequence = ((PyObject *)__pyx_v_fields); if (unlikely(PyTuple_GET_SIZE(sequence) != 2)) { if (PyTuple_GET_SIZE(sequence) > 2) __Pyx_RaiseTooManyValuesError(2); else __Pyx_RaiseNeedMoreValuesError(PyTuple_GET_SIZE(sequence)); {__pyx_filename = __pyx_f[2]; __pyx_lineno = 795; __pyx_clineno = __LINE__; goto __pyx_L1_error;} } __pyx_t_3 = PyTuple_GET_ITEM(sequence, 0); __pyx_t_4 = PyTuple_GET_ITEM(sequence, 1); __Pyx_INCREF(__pyx_t_3); __Pyx_INCREF(__pyx_t_4); } else { __Pyx_UnpackTupleError(((PyObject *)__pyx_v_fields), 2); {__pyx_filename = __pyx_f[2]; __pyx_lineno = 795; __pyx_clineno = __LINE__; goto __pyx_L1_error;} } if (!(likely(((__pyx_t_3) == Py_None) || likely(__Pyx_TypeTest(__pyx_t_3, __pyx_ptype_5numpy_dtype))))) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 795; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_XDECREF(((PyObject *)__pyx_v_child)); __pyx_v_child = ((PyArray_Descr *)__pyx_t_3); __pyx_t_3 = 0; __Pyx_XDECREF(__pyx_v_new_offset); __pyx_v_new_offset = __pyx_t_4; __pyx_t_4 = 0; /* "numpy.pxd":797 * child, new_offset = fields * * if (end - f) - (new_offset - offset[0]) < 15: # <<<<<<<<<<<<<< * raise RuntimeError(u"Format string allocated too short, see comment in numpy.pxd") * */ __pyx_t_4 = PyInt_FromLong((__pyx_v_end - __pyx_v_f)); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 797; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_4); __pyx_t_3 = PyInt_FromLong((__pyx_v_offset[0])); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 797; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); __pyx_t_5 = PyNumber_Subtract(__pyx_v_new_offset, __pyx_t_3); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 797; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_5); __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; __pyx_t_3 = PyNumber_Subtract(__pyx_t_4, __pyx_t_5); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 797; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0; __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0; __pyx_t_5 = PyObject_RichCompare(__pyx_t_3, __pyx_int_15, Py_LT); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 797; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_5); __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; __pyx_t_6 = __Pyx_PyObject_IsTrue(__pyx_t_5); if (unlikely(__pyx_t_6 < 0)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 797; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0; if (__pyx_t_6) { /* "numpy.pxd":798 * * if (end - f) - (new_offset - offset[0]) < 15: * raise RuntimeError(u"Format string allocated too short, see comment in numpy.pxd") # <<<<<<<<<<<<<< * * if ((child.byteorder == '>' and little_endian) or */ __pyx_t_5 = PyObject_Call(__pyx_builtin_RuntimeError, ((PyObject *)__pyx_k_tuple_9), NULL); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 798; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_5); __Pyx_Raise(__pyx_t_5, 0, 0, 0); __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0; {__pyx_filename = __pyx_f[2]; __pyx_lineno = 798; __pyx_clineno = __LINE__; goto __pyx_L1_error;} goto __pyx_L5; } __pyx_L5:; /* "numpy.pxd":800 * raise RuntimeError(u"Format string allocated too short, see comment in numpy.pxd") * * if ((child.byteorder == '>' and little_endian) or # <<<<<<<<<<<<<< * (child.byteorder == '<' and not little_endian)): * raise ValueError(u"Non-native byte order not supported") */ __pyx_t_6 = (__pyx_v_child->byteorder == '>'); if (__pyx_t_6) { __pyx_t_7 = __pyx_v_little_endian; } else { __pyx_t_7 = __pyx_t_6; } if (!__pyx_t_7) { /* "numpy.pxd":801 * * if ((child.byteorder == '>' and little_endian) or * (child.byteorder == '<' and not little_endian)): # <<<<<<<<<<<<<< * raise ValueError(u"Non-native byte order not supported") * # One could encode it in the format string and have Cython */ __pyx_t_6 = (__pyx_v_child->byteorder == '<'); if (__pyx_t_6) { __pyx_t_8 = (!__pyx_v_little_endian); __pyx_t_9 = __pyx_t_8; } else { __pyx_t_9 = __pyx_t_6; } __pyx_t_6 = __pyx_t_9; } else { __pyx_t_6 = __pyx_t_7; } if (__pyx_t_6) { /* "numpy.pxd":802 * if ((child.byteorder == '>' and little_endian) or * (child.byteorder == '<' and not little_endian)): * raise ValueError(u"Non-native byte order not supported") # <<<<<<<<<<<<<< * # One could encode it in the format string and have Cython * # complain instead, BUT: < and > in format strings also imply */ __pyx_t_5 = PyObject_Call(__pyx_builtin_ValueError, ((PyObject *)__pyx_k_tuple_10), NULL); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 802; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_5); __Pyx_Raise(__pyx_t_5, 0, 0, 0); __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0; {__pyx_filename = __pyx_f[2]; __pyx_lineno = 802; __pyx_clineno = __LINE__; goto __pyx_L1_error;} goto __pyx_L6; } __pyx_L6:; /* "numpy.pxd":812 * * # Output padding bytes * while offset[0] < new_offset: # <<<<<<<<<<<<<< * f[0] = 120 # "x"; pad byte * f += 1 */ while (1) { __pyx_t_5 = PyInt_FromLong((__pyx_v_offset[0])); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 812; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_5); __pyx_t_3 = PyObject_RichCompare(__pyx_t_5, __pyx_v_new_offset, Py_LT); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 812; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0; __pyx_t_6 = __Pyx_PyObject_IsTrue(__pyx_t_3); if (unlikely(__pyx_t_6 < 0)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 812; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; if (!__pyx_t_6) break; /* "numpy.pxd":813 * # Output padding bytes * while offset[0] < new_offset: * f[0] = 120 # "x"; pad byte # <<<<<<<<<<<<<< * f += 1 * offset[0] += 1 */ (__pyx_v_f[0]) = 120; /* "numpy.pxd":814 * while offset[0] < new_offset: * f[0] = 120 # "x"; pad byte * f += 1 # <<<<<<<<<<<<<< * offset[0] += 1 * */ __pyx_v_f = (__pyx_v_f + 1); /* "numpy.pxd":815 * f[0] = 120 # "x"; pad byte * f += 1 * offset[0] += 1 # <<<<<<<<<<<<<< * * offset[0] += child.itemsize */ __pyx_t_10 = 0; (__pyx_v_offset[__pyx_t_10]) = ((__pyx_v_offset[__pyx_t_10]) + 1); } /* "numpy.pxd":817 * offset[0] += 1 * * offset[0] += child.itemsize # <<<<<<<<<<<<<< * * if not PyDataType_HASFIELDS(child): */ __pyx_t_10 = 0; (__pyx_v_offset[__pyx_t_10]) = ((__pyx_v_offset[__pyx_t_10]) + __pyx_v_child->elsize); /* "numpy.pxd":819 * offset[0] += child.itemsize * * if not PyDataType_HASFIELDS(child): # <<<<<<<<<<<<<< * t = child.type_num * if end - f < 5: */ __pyx_t_6 = (!PyDataType_HASFIELDS(__pyx_v_child)); if (__pyx_t_6) { /* "numpy.pxd":820 * * if not PyDataType_HASFIELDS(child): * t = child.type_num # <<<<<<<<<<<<<< * if end - f < 5: * raise RuntimeError(u"Format string allocated too short.") */ __pyx_t_3 = PyInt_FromLong(__pyx_v_child->type_num); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 820; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); __Pyx_XDECREF(__pyx_v_t); __pyx_v_t = __pyx_t_3; __pyx_t_3 = 0; /* "numpy.pxd":821 * if not PyDataType_HASFIELDS(child): * t = child.type_num * if end - f < 5: # <<<<<<<<<<<<<< * raise RuntimeError(u"Format string allocated too short.") * */ __pyx_t_6 = ((__pyx_v_end - __pyx_v_f) < 5); if (__pyx_t_6) { /* "numpy.pxd":822 * t = child.type_num * if end - f < 5: * raise RuntimeError(u"Format string allocated too short.") # <<<<<<<<<<<<<< * * # Until ticket #99 is fixed, use integers to avoid warnings */ __pyx_t_3 = PyObject_Call(__pyx_builtin_RuntimeError, ((PyObject *)__pyx_k_tuple_12), NULL); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 822; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); __Pyx_Raise(__pyx_t_3, 0, 0, 0); __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; {__pyx_filename = __pyx_f[2]; __pyx_lineno = 822; __pyx_clineno = __LINE__; goto __pyx_L1_error;} goto __pyx_L10; } __pyx_L10:; /* "numpy.pxd":825 * * # Until ticket #99 is fixed, use integers to avoid warnings * if t == NPY_BYTE: f[0] = 98 #"b" # <<<<<<<<<<<<<< * elif t == NPY_UBYTE: f[0] = 66 #"B" * elif t == NPY_SHORT: f[0] = 104 #"h" */ __pyx_t_3 = PyInt_FromLong(NPY_BYTE); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 825; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); __pyx_t_5 = PyObject_RichCompare(__pyx_v_t, __pyx_t_3, Py_EQ); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 825; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_5); __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; __pyx_t_6 = __Pyx_PyObject_IsTrue(__pyx_t_5); if (unlikely(__pyx_t_6 < 0)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 825; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0; if (__pyx_t_6) { (__pyx_v_f[0]) = 98; goto __pyx_L11; } /* "numpy.pxd":826 * # Until ticket #99 is fixed, use integers to avoid warnings * if t == NPY_BYTE: f[0] = 98 #"b" * elif t == NPY_UBYTE: f[0] = 66 #"B" # <<<<<<<<<<<<<< * elif t == NPY_SHORT: f[0] = 104 #"h" * elif t == NPY_USHORT: f[0] = 72 #"H" */ __pyx_t_5 = PyInt_FromLong(NPY_UBYTE); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 826; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_5); __pyx_t_3 = PyObject_RichCompare(__pyx_v_t, __pyx_t_5, Py_EQ); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 826; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0; __pyx_t_6 = __Pyx_PyObject_IsTrue(__pyx_t_3); if (unlikely(__pyx_t_6 < 0)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 826; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; if (__pyx_t_6) { (__pyx_v_f[0]) = 66; goto __pyx_L11; } /* "numpy.pxd":827 * if t == NPY_BYTE: f[0] = 98 #"b" * elif t == NPY_UBYTE: f[0] = 66 #"B" * elif t == NPY_SHORT: f[0] = 104 #"h" # <<<<<<<<<<<<<< * elif t == NPY_USHORT: f[0] = 72 #"H" * elif t == NPY_INT: f[0] = 105 #"i" */ __pyx_t_3 = PyInt_FromLong(NPY_SHORT); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 827; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); __pyx_t_5 = PyObject_RichCompare(__pyx_v_t, __pyx_t_3, Py_EQ); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 827; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_5); __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; __pyx_t_6 = __Pyx_PyObject_IsTrue(__pyx_t_5); if (unlikely(__pyx_t_6 < 0)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 827; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0; if (__pyx_t_6) { (__pyx_v_f[0]) = 104; goto __pyx_L11; } /* "numpy.pxd":828 * elif t == NPY_UBYTE: f[0] = 66 #"B" * elif t == NPY_SHORT: f[0] = 104 #"h" * elif t == NPY_USHORT: f[0] = 72 #"H" # <<<<<<<<<<<<<< * elif t == NPY_INT: f[0] = 105 #"i" * elif t == NPY_UINT: f[0] = 73 #"I" */ __pyx_t_5 = PyInt_FromLong(NPY_USHORT); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 828; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_5); __pyx_t_3 = PyObject_RichCompare(__pyx_v_t, __pyx_t_5, Py_EQ); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 828; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0; __pyx_t_6 = __Pyx_PyObject_IsTrue(__pyx_t_3); if (unlikely(__pyx_t_6 < 0)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 828; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; if (__pyx_t_6) { (__pyx_v_f[0]) = 72; goto __pyx_L11; } /* "numpy.pxd":829 * elif t == NPY_SHORT: f[0] = 104 #"h" * elif t == NPY_USHORT: f[0] = 72 #"H" * elif t == NPY_INT: f[0] = 105 #"i" # <<<<<<<<<<<<<< * elif t == NPY_UINT: f[0] = 73 #"I" * elif t == NPY_LONG: f[0] = 108 #"l" */ __pyx_t_3 = PyInt_FromLong(NPY_INT); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 829; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); __pyx_t_5 = PyObject_RichCompare(__pyx_v_t, __pyx_t_3, Py_EQ); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 829; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_5); __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; __pyx_t_6 = __Pyx_PyObject_IsTrue(__pyx_t_5); if (unlikely(__pyx_t_6 < 0)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 829; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0; if (__pyx_t_6) { (__pyx_v_f[0]) = 105; goto __pyx_L11; } /* "numpy.pxd":830 * elif t == NPY_USHORT: f[0] = 72 #"H" * elif t == NPY_INT: f[0] = 105 #"i" * elif t == NPY_UINT: f[0] = 73 #"I" # <<<<<<<<<<<<<< * elif t == NPY_LONG: f[0] = 108 #"l" * elif t == NPY_ULONG: f[0] = 76 #"L" */ __pyx_t_5 = PyInt_FromLong(NPY_UINT); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 830; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_5); __pyx_t_3 = PyObject_RichCompare(__pyx_v_t, __pyx_t_5, Py_EQ); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 830; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0; __pyx_t_6 = __Pyx_PyObject_IsTrue(__pyx_t_3); if (unlikely(__pyx_t_6 < 0)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 830; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; if (__pyx_t_6) { (__pyx_v_f[0]) = 73; goto __pyx_L11; } /* "numpy.pxd":831 * elif t == NPY_INT: f[0] = 105 #"i" * elif t == NPY_UINT: f[0] = 73 #"I" * elif t == NPY_LONG: f[0] = 108 #"l" # <<<<<<<<<<<<<< * elif t == NPY_ULONG: f[0] = 76 #"L" * elif t == NPY_LONGLONG: f[0] = 113 #"q" */ __pyx_t_3 = PyInt_FromLong(NPY_LONG); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 831; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); __pyx_t_5 = PyObject_RichCompare(__pyx_v_t, __pyx_t_3, Py_EQ); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 831; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_5); __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; __pyx_t_6 = __Pyx_PyObject_IsTrue(__pyx_t_5); if (unlikely(__pyx_t_6 < 0)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 831; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0; if (__pyx_t_6) { (__pyx_v_f[0]) = 108; goto __pyx_L11; } /* "numpy.pxd":832 * elif t == NPY_UINT: f[0] = 73 #"I" * elif t == NPY_LONG: f[0] = 108 #"l" * elif t == NPY_ULONG: f[0] = 76 #"L" # <<<<<<<<<<<<<< * elif t == NPY_LONGLONG: f[0] = 113 #"q" * elif t == NPY_ULONGLONG: f[0] = 81 #"Q" */ __pyx_t_5 = PyInt_FromLong(NPY_ULONG); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 832; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_5); __pyx_t_3 = PyObject_RichCompare(__pyx_v_t, __pyx_t_5, Py_EQ); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 832; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0; __pyx_t_6 = __Pyx_PyObject_IsTrue(__pyx_t_3); if (unlikely(__pyx_t_6 < 0)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 832; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; if (__pyx_t_6) { (__pyx_v_f[0]) = 76; goto __pyx_L11; } /* "numpy.pxd":833 * elif t == NPY_LONG: f[0] = 108 #"l" * elif t == NPY_ULONG: f[0] = 76 #"L" * elif t == NPY_LONGLONG: f[0] = 113 #"q" # <<<<<<<<<<<<<< * elif t == NPY_ULONGLONG: f[0] = 81 #"Q" * elif t == NPY_FLOAT: f[0] = 102 #"f" */ __pyx_t_3 = PyInt_FromLong(NPY_LONGLONG); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 833; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); __pyx_t_5 = PyObject_RichCompare(__pyx_v_t, __pyx_t_3, Py_EQ); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 833; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_5); __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; __pyx_t_6 = __Pyx_PyObject_IsTrue(__pyx_t_5); if (unlikely(__pyx_t_6 < 0)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 833; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0; if (__pyx_t_6) { (__pyx_v_f[0]) = 113; goto __pyx_L11; } /* "numpy.pxd":834 * elif t == NPY_ULONG: f[0] = 76 #"L" * elif t == NPY_LONGLONG: f[0] = 113 #"q" * elif t == NPY_ULONGLONG: f[0] = 81 #"Q" # <<<<<<<<<<<<<< * elif t == NPY_FLOAT: f[0] = 102 #"f" * elif t == NPY_DOUBLE: f[0] = 100 #"d" */ __pyx_t_5 = PyInt_FromLong(NPY_ULONGLONG); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 834; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_5); __pyx_t_3 = PyObject_RichCompare(__pyx_v_t, __pyx_t_5, Py_EQ); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 834; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0; __pyx_t_6 = __Pyx_PyObject_IsTrue(__pyx_t_3); if (unlikely(__pyx_t_6 < 0)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 834; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; if (__pyx_t_6) { (__pyx_v_f[0]) = 81; goto __pyx_L11; } /* "numpy.pxd":835 * elif t == NPY_LONGLONG: f[0] = 113 #"q" * elif t == NPY_ULONGLONG: f[0] = 81 #"Q" * elif t == NPY_FLOAT: f[0] = 102 #"f" # <<<<<<<<<<<<<< * elif t == NPY_DOUBLE: f[0] = 100 #"d" * elif t == NPY_LONGDOUBLE: f[0] = 103 #"g" */ __pyx_t_3 = PyInt_FromLong(NPY_FLOAT); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 835; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); __pyx_t_5 = PyObject_RichCompare(__pyx_v_t, __pyx_t_3, Py_EQ); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 835; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_5); __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; __pyx_t_6 = __Pyx_PyObject_IsTrue(__pyx_t_5); if (unlikely(__pyx_t_6 < 0)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 835; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0; if (__pyx_t_6) { (__pyx_v_f[0]) = 102; goto __pyx_L11; } /* "numpy.pxd":836 * elif t == NPY_ULONGLONG: f[0] = 81 #"Q" * elif t == NPY_FLOAT: f[0] = 102 #"f" * elif t == NPY_DOUBLE: f[0] = 100 #"d" # <<<<<<<<<<<<<< * elif t == NPY_LONGDOUBLE: f[0] = 103 #"g" * elif t == NPY_CFLOAT: f[0] = 90; f[1] = 102; f += 1 # Zf */ __pyx_t_5 = PyInt_FromLong(NPY_DOUBLE); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 836; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_5); __pyx_t_3 = PyObject_RichCompare(__pyx_v_t, __pyx_t_5, Py_EQ); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 836; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0; __pyx_t_6 = __Pyx_PyObject_IsTrue(__pyx_t_3); if (unlikely(__pyx_t_6 < 0)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 836; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; if (__pyx_t_6) { (__pyx_v_f[0]) = 100; goto __pyx_L11; } /* "numpy.pxd":837 * elif t == NPY_FLOAT: f[0] = 102 #"f" * elif t == NPY_DOUBLE: f[0] = 100 #"d" * elif t == NPY_LONGDOUBLE: f[0] = 103 #"g" # <<<<<<<<<<<<<< * elif t == NPY_CFLOAT: f[0] = 90; f[1] = 102; f += 1 # Zf * elif t == NPY_CDOUBLE: f[0] = 90; f[1] = 100; f += 1 # Zd */ __pyx_t_3 = PyInt_FromLong(NPY_LONGDOUBLE); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 837; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); __pyx_t_5 = PyObject_RichCompare(__pyx_v_t, __pyx_t_3, Py_EQ); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 837; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_5); __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; __pyx_t_6 = __Pyx_PyObject_IsTrue(__pyx_t_5); if (unlikely(__pyx_t_6 < 0)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 837; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0; if (__pyx_t_6) { (__pyx_v_f[0]) = 103; goto __pyx_L11; } /* "numpy.pxd":838 * elif t == NPY_DOUBLE: f[0] = 100 #"d" * elif t == NPY_LONGDOUBLE: f[0] = 103 #"g" * elif t == NPY_CFLOAT: f[0] = 90; f[1] = 102; f += 1 # Zf # <<<<<<<<<<<<<< * elif t == NPY_CDOUBLE: f[0] = 90; f[1] = 100; f += 1 # Zd * elif t == NPY_CLONGDOUBLE: f[0] = 90; f[1] = 103; f += 1 # Zg */ __pyx_t_5 = PyInt_FromLong(NPY_CFLOAT); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 838; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_5); __pyx_t_3 = PyObject_RichCompare(__pyx_v_t, __pyx_t_5, Py_EQ); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 838; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0; __pyx_t_6 = __Pyx_PyObject_IsTrue(__pyx_t_3); if (unlikely(__pyx_t_6 < 0)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 838; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; if (__pyx_t_6) { (__pyx_v_f[0]) = 90; (__pyx_v_f[1]) = 102; __pyx_v_f = (__pyx_v_f + 1); goto __pyx_L11; } /* "numpy.pxd":839 * elif t == NPY_LONGDOUBLE: f[0] = 103 #"g" * elif t == NPY_CFLOAT: f[0] = 90; f[1] = 102; f += 1 # Zf * elif t == NPY_CDOUBLE: f[0] = 90; f[1] = 100; f += 1 # Zd # <<<<<<<<<<<<<< * elif t == NPY_CLONGDOUBLE: f[0] = 90; f[1] = 103; f += 1 # Zg * elif t == NPY_OBJECT: f[0] = 79 #"O" */ __pyx_t_3 = PyInt_FromLong(NPY_CDOUBLE); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 839; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); __pyx_t_5 = PyObject_RichCompare(__pyx_v_t, __pyx_t_3, Py_EQ); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 839; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_5); __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; __pyx_t_6 = __Pyx_PyObject_IsTrue(__pyx_t_5); if (unlikely(__pyx_t_6 < 0)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 839; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0; if (__pyx_t_6) { (__pyx_v_f[0]) = 90; (__pyx_v_f[1]) = 100; __pyx_v_f = (__pyx_v_f + 1); goto __pyx_L11; } /* "numpy.pxd":840 * elif t == NPY_CFLOAT: f[0] = 90; f[1] = 102; f += 1 # Zf * elif t == NPY_CDOUBLE: f[0] = 90; f[1] = 100; f += 1 # Zd * elif t == NPY_CLONGDOUBLE: f[0] = 90; f[1] = 103; f += 1 # Zg # <<<<<<<<<<<<<< * elif t == NPY_OBJECT: f[0] = 79 #"O" * else: */ __pyx_t_5 = PyInt_FromLong(NPY_CLONGDOUBLE); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 840; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_5); __pyx_t_3 = PyObject_RichCompare(__pyx_v_t, __pyx_t_5, Py_EQ); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 840; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0; __pyx_t_6 = __Pyx_PyObject_IsTrue(__pyx_t_3); if (unlikely(__pyx_t_6 < 0)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 840; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; if (__pyx_t_6) { (__pyx_v_f[0]) = 90; (__pyx_v_f[1]) = 103; __pyx_v_f = (__pyx_v_f + 1); goto __pyx_L11; } /* "numpy.pxd":841 * elif t == NPY_CDOUBLE: f[0] = 90; f[1] = 100; f += 1 # Zd * elif t == NPY_CLONGDOUBLE: f[0] = 90; f[1] = 103; f += 1 # Zg * elif t == NPY_OBJECT: f[0] = 79 #"O" # <<<<<<<<<<<<<< * else: * raise ValueError(u"unknown dtype code in numpy.pxd (%d)" % t) */ __pyx_t_3 = PyInt_FromLong(NPY_OBJECT); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 841; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); __pyx_t_5 = PyObject_RichCompare(__pyx_v_t, __pyx_t_3, Py_EQ); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 841; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_5); __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; __pyx_t_6 = __Pyx_PyObject_IsTrue(__pyx_t_5); if (unlikely(__pyx_t_6 < 0)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 841; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0; if (__pyx_t_6) { (__pyx_v_f[0]) = 79; goto __pyx_L11; } /*else*/ { /* "numpy.pxd":843 * elif t == NPY_OBJECT: f[0] = 79 #"O" * else: * raise ValueError(u"unknown dtype code in numpy.pxd (%d)" % t) # <<<<<<<<<<<<<< * f += 1 * else: */ __pyx_t_5 = PyNumber_Remainder(((PyObject *)__pyx_kp_u_7), __pyx_v_t); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 843; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(((PyObject *)__pyx_t_5)); __pyx_t_3 = PyTuple_New(1); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 843; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); PyTuple_SET_ITEM(__pyx_t_3, 0, ((PyObject *)__pyx_t_5)); __Pyx_GIVEREF(((PyObject *)__pyx_t_5)); __pyx_t_5 = 0; __pyx_t_5 = PyObject_Call(__pyx_builtin_ValueError, ((PyObject *)__pyx_t_3), NULL); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 843; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_5); __Pyx_DECREF(((PyObject *)__pyx_t_3)); __pyx_t_3 = 0; __Pyx_Raise(__pyx_t_5, 0, 0, 0); __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0; {__pyx_filename = __pyx_f[2]; __pyx_lineno = 843; __pyx_clineno = __LINE__; goto __pyx_L1_error;} } __pyx_L11:; /* "numpy.pxd":844 * else: * raise ValueError(u"unknown dtype code in numpy.pxd (%d)" % t) * f += 1 # <<<<<<<<<<<<<< * else: * # Cython ignores struct boundary information ("T{...}"), */ __pyx_v_f = (__pyx_v_f + 1); goto __pyx_L9; } /*else*/ { /* "numpy.pxd":848 * # Cython ignores struct boundary information ("T{...}"), * # so don't output it * f = _util_dtypestring(child, f, end, offset) # <<<<<<<<<<<<<< * return f * */ __pyx_t_11 = __pyx_f_5numpy__util_dtypestring(__pyx_v_child, __pyx_v_f, __pyx_v_end, __pyx_v_offset); if (unlikely(__pyx_t_11 == NULL)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 848; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __pyx_v_f = __pyx_t_11; } __pyx_L9:; } __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; /* "numpy.pxd":849 * # so don't output it * f = _util_dtypestring(child, f, end, offset) * return f # <<<<<<<<<<<<<< * * */ __pyx_r = __pyx_v_f; goto __pyx_L0; __pyx_r = 0; goto __pyx_L0; __pyx_L1_error:; __Pyx_XDECREF(__pyx_t_1); __Pyx_XDECREF(__pyx_t_3); __Pyx_XDECREF(__pyx_t_4); __Pyx_XDECREF(__pyx_t_5); __Pyx_AddTraceback("numpy._util_dtypestring", __pyx_clineno, __pyx_lineno, __pyx_filename); __pyx_r = NULL; __pyx_L0:; __Pyx_XDECREF((PyObject *)__pyx_v_child); __Pyx_XDECREF(__pyx_v_fields); __Pyx_XDECREF(__pyx_v_childname); __Pyx_XDECREF(__pyx_v_new_offset); __Pyx_XDECREF(__pyx_v_t); __Pyx_RefNannyFinishContext(); return __pyx_r; } /* "numpy.pxd":964 * * * cdef inline void set_array_base(ndarray arr, object base): # <<<<<<<<<<<<<< * cdef PyObject* baseptr * if base is None: */ static CYTHON_INLINE void __pyx_f_5numpy_set_array_base(PyArrayObject *__pyx_v_arr, PyObject *__pyx_v_base) { PyObject *__pyx_v_baseptr; __Pyx_RefNannyDeclarations int __pyx_t_1; __Pyx_RefNannySetupContext("set_array_base"); /* "numpy.pxd":966 * cdef inline void set_array_base(ndarray arr, object base): * cdef PyObject* baseptr * if base is None: # <<<<<<<<<<<<<< * baseptr = NULL * else: */ __pyx_t_1 = (__pyx_v_base == Py_None); if (__pyx_t_1) { /* "numpy.pxd":967 * cdef PyObject* baseptr * if base is None: * baseptr = NULL # <<<<<<<<<<<<<< * else: * Py_INCREF(base) # important to do this before decref below! */ __pyx_v_baseptr = NULL; goto __pyx_L3; } /*else*/ { /* "numpy.pxd":969 * baseptr = NULL * else: * Py_INCREF(base) # important to do this before decref below! # <<<<<<<<<<<<<< * baseptr = base * Py_XDECREF(arr.base) */ Py_INCREF(__pyx_v_base); /* "numpy.pxd":970 * else: * Py_INCREF(base) # important to do this before decref below! * baseptr = base # <<<<<<<<<<<<<< * Py_XDECREF(arr.base) * arr.base = baseptr */ __pyx_v_baseptr = ((PyObject *)__pyx_v_base); } __pyx_L3:; /* "numpy.pxd":971 * Py_INCREF(base) # important to do this before decref below! * baseptr = base * Py_XDECREF(arr.base) # <<<<<<<<<<<<<< * arr.base = baseptr * */ Py_XDECREF(__pyx_v_arr->base); /* "numpy.pxd":972 * baseptr = base * Py_XDECREF(arr.base) * arr.base = baseptr # <<<<<<<<<<<<<< * * cdef inline object get_array_base(ndarray arr): */ __pyx_v_arr->base = __pyx_v_baseptr; __Pyx_RefNannyFinishContext(); } /* "numpy.pxd":974 * arr.base = baseptr * * cdef inline object get_array_base(ndarray arr): # <<<<<<<<<<<<<< * if arr.base is NULL: * return None */ static CYTHON_INLINE PyObject *__pyx_f_5numpy_get_array_base(PyArrayObject *__pyx_v_arr) { PyObject *__pyx_r = NULL; __Pyx_RefNannyDeclarations int __pyx_t_1; __Pyx_RefNannySetupContext("get_array_base"); /* "numpy.pxd":975 * * cdef inline object get_array_base(ndarray arr): * if arr.base is NULL: # <<<<<<<<<<<<<< * return None * else: */ __pyx_t_1 = (__pyx_v_arr->base == NULL); if (__pyx_t_1) { /* "numpy.pxd":976 * cdef inline object get_array_base(ndarray arr): * if arr.base is NULL: * return None # <<<<<<<<<<<<<< * else: * return arr.base */ __Pyx_XDECREF(__pyx_r); __Pyx_INCREF(Py_None); __pyx_r = Py_None; goto __pyx_L0; goto __pyx_L3; } /*else*/ { /* "numpy.pxd":978 * return None * else: * return arr.base # <<<<<<<<<<<<<< */ __Pyx_XDECREF(__pyx_r); __Pyx_INCREF(((PyObject *)__pyx_v_arr->base)); __pyx_r = ((PyObject *)__pyx_v_arr->base); goto __pyx_L0; } __pyx_L3:; __pyx_r = Py_None; __Pyx_INCREF(Py_None); __pyx_L0:; __Pyx_XGIVEREF(__pyx_r); __Pyx_RefNannyFinishContext(); return __pyx_r; } static PyMethodDef __pyx_methods[] = { {__Pyx_NAMESTR("full_pdf"), (PyCFunction)__pyx_pw_4wfpt_1full_pdf, METH_VARARGS|METH_KEYWORDS, __Pyx_DOCSTR(__pyx_doc_4wfpt_full_pdf)}, {0, 0, 0, 0} }; #if PY_MAJOR_VERSION >= 3 static struct PyModuleDef __pyx_moduledef = { PyModuleDef_HEAD_INIT, __Pyx_NAMESTR("wfpt"), 0, /* m_doc */ -1, /* m_size */ __pyx_methods /* m_methods */, NULL, /* m_reload */ NULL, /* m_traverse */ NULL, /* m_clear */ NULL /* m_free */ }; #endif static __Pyx_StringTabEntry __pyx_string_tab[] = { {&__pyx_kp_u_1, __pyx_k_1, sizeof(__pyx_k_1), 0, 1, 0, 0}, {&__pyx_kp_u_11, __pyx_k_11, sizeof(__pyx_k_11), 0, 1, 0, 0}, {&__pyx_kp_s_15, __pyx_k_15, sizeof(__pyx_k_15), 0, 0, 1, 0}, {&__pyx_n_s_26, __pyx_k_26, sizeof(__pyx_k_26), 0, 0, 1, 1}, {&__pyx_kp_u_3, __pyx_k_3, sizeof(__pyx_k_3), 0, 1, 0, 0}, {&__pyx_kp_u_5, __pyx_k_5, sizeof(__pyx_k_5), 0, 1, 0, 0}, {&__pyx_kp_u_7, __pyx_k_7, sizeof(__pyx_k_7), 0, 1, 0, 0}, {&__pyx_kp_u_8, __pyx_k_8, sizeof(__pyx_k_8), 0, 1, 0, 0}, {&__pyx_n_s__RuntimeError, __pyx_k__RuntimeError, sizeof(__pyx_k__RuntimeError), 0, 0, 1, 1}, {&__pyx_n_s__T, __pyx_k__T, sizeof(__pyx_k__T), 0, 0, 1, 1}, {&__pyx_n_s__V, __pyx_k__V, sizeof(__pyx_k__V), 0, 0, 1, 1}, {&__pyx_n_s__ValueError, __pyx_k__ValueError, sizeof(__pyx_k__ValueError), 0, 0, 1, 1}, {&__pyx_n_s__Z, __pyx_k__Z, sizeof(__pyx_k__Z), 0, 0, 1, 1}, {&__pyx_n_s____main__, __pyx_k____main__, sizeof(__pyx_k____main__), 0, 0, 1, 1}, {&__pyx_n_s____test__, __pyx_k____test__, sizeof(__pyx_k____test__), 0, 0, 1, 1}, {&__pyx_n_s__a, __pyx_k__a, sizeof(__pyx_k__a), 0, 0, 1, 1}, {&__pyx_n_s__arange, __pyx_k__arange, sizeof(__pyx_k__arange), 0, 0, 1, 1}, {&__pyx_n_s__cdf_lb, __pyx_k__cdf_lb, sizeof(__pyx_k__cdf_lb), 0, 0, 1, 1}, {&__pyx_n_s__cdf_ub, __pyx_k__cdf_ub, sizeof(__pyx_k__cdf_ub), 0, 0, 1, 1}, {&__pyx_n_s__cont_x, __pyx_k__cont_x, sizeof(__pyx_k__cont_x), 0, 0, 1, 1}, {&__pyx_n_s__copy, __pyx_k__copy, sizeof(__pyx_k__copy), 0, 0, 1, 1}, {&__pyx_n_s__delay, __pyx_k__delay, sizeof(__pyx_k__delay), 0, 0, 1, 1}, {&__pyx_n_s__double, __pyx_k__double, sizeof(__pyx_k__double), 0, 0, 1, 1}, {&__pyx_n_s__dt, __pyx_k__dt, sizeof(__pyx_k__dt), 0, 0, 1, 1}, {&__pyx_n_s__dtype, __pyx_k__dtype, sizeof(__pyx_k__dtype), 0, 0, 1, 1}, {&__pyx_n_s__empty, __pyx_k__empty, sizeof(__pyx_k__empty), 0, 0, 1, 1}, {&__pyx_n_s__err, __pyx_k__err, sizeof(__pyx_k__err), 0, 0, 1, 1}, {&__pyx_n_s__f, __pyx_k__f, sizeof(__pyx_k__f), 0, 0, 1, 1}, {&__pyx_n_s__float, __pyx_k__float, sizeof(__pyx_k__float), 0, 0, 1, 1}, {&__pyx_n_s__gen_rts_from_cdf, __pyx_k__gen_rts_from_cdf, sizeof(__pyx_k__gen_rts_from_cdf), 0, 0, 1, 1}, {&__pyx_n_s__hddm, __pyx_k__hddm, sizeof(__pyx_k__hddm), 0, 0, 1, 1}, {&__pyx_n_s__i, __pyx_k__i, sizeof(__pyx_k__i), 0, 0, 1, 1}, {&__pyx_n_s__idx, __pyx_k__idx, sizeof(__pyx_k__idx), 0, 0, 1, 1}, {&__pyx_n_s__inf, __pyx_k__inf, sizeof(__pyx_k__inf), 0, 0, 1, 1}, {&__pyx_n_s__j, __pyx_k__j, sizeof(__pyx_k__j), 0, 0, 1, 1}, {&__pyx_n_s__l_cdf, __pyx_k__l_cdf, sizeof(__pyx_k__l_cdf), 0, 0, 1, 1}, {&__pyx_n_s__log, __pyx_k__log, sizeof(__pyx_k__log), 0, 0, 1, 1}, {&__pyx_n_s__logp, __pyx_k__logp, sizeof(__pyx_k__logp), 0, 0, 1, 1}, {&__pyx_n_s__multi, __pyx_k__multi, sizeof(__pyx_k__multi), 0, 0, 1, 1}, {&__pyx_n_s__nT, __pyx_k__nT, sizeof(__pyx_k__nT), 0, 0, 1, 1}, {&__pyx_n_s__nZ, __pyx_k__nZ, sizeof(__pyx_k__nZ), 0, 0, 1, 1}, {&__pyx_n_s__n_cont, __pyx_k__n_cont, sizeof(__pyx_k__n_cont), 0, 0, 1, 1}, {&__pyx_n_s__np, __pyx_k__np, sizeof(__pyx_k__np), 0, 0, 1, 1}, {&__pyx_n_s__numpy, __pyx_k__numpy, sizeof(__pyx_k__numpy), 0, 0, 1, 1}, {&__pyx_n_s__p, __pyx_k__p, sizeof(__pyx_k__p), 0, 0, 1, 1}, {&__pyx_n_s__p_array, __pyx_k__p_array, sizeof(__pyx_k__p_array), 0, 0, 1, 1}, {&__pyx_n_s__param, __pyx_k__param, sizeof(__pyx_k__param), 0, 0, 1, 1}, {&__pyx_n_s__params, __pyx_k__params, sizeof(__pyx_k__params), 0, 0, 1, 1}, {&__pyx_n_s__params_iter, __pyx_k__params_iter, sizeof(__pyx_k__params_iter), 0, 0, 1, 1}, {&__pyx_n_s__pdf, __pyx_k__pdf, sizeof(__pyx_k__pdf), 0, 0, 1, 1}, {&__pyx_n_s__pdf_array, __pyx_k__pdf_array, sizeof(__pyx_k__pdf_array), 0, 0, 1, 1}, {&__pyx_n_s__pos_cont, __pyx_k__pos_cont, sizeof(__pyx_k__pos_cont), 0, 0, 1, 1}, {&__pyx_n_s__rand, __pyx_k__rand, sizeof(__pyx_k__rand), 0, 0, 1, 1}, {&__pyx_n_s__random, __pyx_k__random, sizeof(__pyx_k__random), 0, 0, 1, 1}, {&__pyx_n_s__range, __pyx_k__range, sizeof(__pyx_k__range), 0, 0, 1, 1}, {&__pyx_n_s__rt, __pyx_k__rt, sizeof(__pyx_k__rt), 0, 0, 1, 1}, {&__pyx_n_s__rts, __pyx_k__rts, sizeof(__pyx_k__rts), 0, 0, 1, 1}, {&__pyx_n_s__samples, __pyx_k__samples, sizeof(__pyx_k__samples), 0, 0, 1, 1}, {&__pyx_n_s__searchsorted, __pyx_k__searchsorted, sizeof(__pyx_k__searchsorted), 0, 0, 1, 1}, {&__pyx_n_s__sign, __pyx_k__sign, sizeof(__pyx_k__sign), 0, 0, 1, 1}, {&__pyx_n_s__simps_err, __pyx_k__simps_err, sizeof(__pyx_k__simps_err), 0, 0, 1, 1}, {&__pyx_n_s__size, __pyx_k__size, sizeof(__pyx_k__size), 0, 0, 1, 1}, {&__pyx_n_s__sum, __pyx_k__sum, sizeof(__pyx_k__sum), 0, 0, 1, 1}, {&__pyx_n_s__sum_logp, __pyx_k__sum_logp, sizeof(__pyx_k__sum_logp), 0, 0, 1, 1}, {&__pyx_n_s__t, __pyx_k__t, sizeof(__pyx_k__t), 0, 0, 1, 1}, {&__pyx_n_s__t_max, __pyx_k__t_max, sizeof(__pyx_k__t_max), 0, 0, 1, 1}, {&__pyx_n_s__t_min, __pyx_k__t_min, sizeof(__pyx_k__t_min), 0, 0, 1, 1}, {&__pyx_n_s__use_adaptive, __pyx_k__use_adaptive, sizeof(__pyx_k__use_adaptive), 0, 0, 1, 1}, {&__pyx_n_s__v, __pyx_k__v, sizeof(__pyx_k__v), 0, 0, 1, 1}, {&__pyx_n_s__wfpt, __pyx_k__wfpt, sizeof(__pyx_k__wfpt), 0, 0, 1, 1}, {&__pyx_n_s__wiener_like, __pyx_k__wiener_like, sizeof(__pyx_k__wiener_like), 0, 0, 1, 1}, {&__pyx_n_s__wiener_like_array, __pyx_k__wiener_like_array, sizeof(__pyx_k__wiener_like_array), 0, 0, 1, 1}, {&__pyx_n_s__wiener_like_multi, __pyx_k__wiener_like_multi, sizeof(__pyx_k__wiener_like_multi), 0, 0, 1, 1}, {&__pyx_n_s__x, __pyx_k__x, sizeof(__pyx_k__x), 0, 0, 1, 1}, {&__pyx_n_s__y, __pyx_k__y, sizeof(__pyx_k__y), 0, 0, 1, 1}, {&__pyx_n_s__z, __pyx_k__z, sizeof(__pyx_k__z), 0, 0, 1, 1}, {0, 0, 0, 0, 0, 0, 0} }; static int __Pyx_InitCachedBuiltins(void) { __pyx_builtin_range = __Pyx_GetName(__pyx_b, __pyx_n_s__range); if (!__pyx_builtin_range) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 116; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __pyx_builtin_ValueError = __Pyx_GetName(__pyx_b, __pyx_n_s__ValueError); if (!__pyx_builtin_ValueError) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 214; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __pyx_builtin_RuntimeError = __Pyx_GetName(__pyx_b, __pyx_n_s__RuntimeError); if (!__pyx_builtin_RuntimeError) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 798; __pyx_clineno = __LINE__; goto __pyx_L1_error;} return 0; __pyx_L1_error:; return -1; } static int __Pyx_InitCachedConstants(void) { __Pyx_RefNannyDeclarations __Pyx_RefNannySetupContext("__Pyx_InitCachedConstants"); /* "numpy.pxd":214 * if ((flags & pybuf.PyBUF_C_CONTIGUOUS == pybuf.PyBUF_C_CONTIGUOUS) * and not PyArray_CHKFLAGS(self, NPY_C_CONTIGUOUS)): * raise ValueError(u"ndarray is not C contiguous") # <<<<<<<<<<<<<< * * if ((flags & pybuf.PyBUF_F_CONTIGUOUS == pybuf.PyBUF_F_CONTIGUOUS) */ __pyx_k_tuple_2 = PyTuple_New(1); if (unlikely(!__pyx_k_tuple_2)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 214; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_k_tuple_2); __Pyx_INCREF(((PyObject *)__pyx_kp_u_1)); PyTuple_SET_ITEM(__pyx_k_tuple_2, 0, ((PyObject *)__pyx_kp_u_1)); __Pyx_GIVEREF(((PyObject *)__pyx_kp_u_1)); __Pyx_GIVEREF(((PyObject *)__pyx_k_tuple_2)); /* "numpy.pxd":218 * if ((flags & pybuf.PyBUF_F_CONTIGUOUS == pybuf.PyBUF_F_CONTIGUOUS) * and not PyArray_CHKFLAGS(self, NPY_F_CONTIGUOUS)): * raise ValueError(u"ndarray is not Fortran contiguous") # <<<<<<<<<<<<<< * * info.buf = PyArray_DATA(self) */ __pyx_k_tuple_4 = PyTuple_New(1); if (unlikely(!__pyx_k_tuple_4)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 218; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_k_tuple_4); __Pyx_INCREF(((PyObject *)__pyx_kp_u_3)); PyTuple_SET_ITEM(__pyx_k_tuple_4, 0, ((PyObject *)__pyx_kp_u_3)); __Pyx_GIVEREF(((PyObject *)__pyx_kp_u_3)); __Pyx_GIVEREF(((PyObject *)__pyx_k_tuple_4)); /* "numpy.pxd":256 * if ((descr.byteorder == '>' and little_endian) or * (descr.byteorder == '<' and not little_endian)): * raise ValueError(u"Non-native byte order not supported") # <<<<<<<<<<<<<< * if t == NPY_BYTE: f = "b" * elif t == NPY_UBYTE: f = "B" */ __pyx_k_tuple_6 = PyTuple_New(1); if (unlikely(!__pyx_k_tuple_6)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 256; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_k_tuple_6); __Pyx_INCREF(((PyObject *)__pyx_kp_u_5)); PyTuple_SET_ITEM(__pyx_k_tuple_6, 0, ((PyObject *)__pyx_kp_u_5)); __Pyx_GIVEREF(((PyObject *)__pyx_kp_u_5)); __Pyx_GIVEREF(((PyObject *)__pyx_k_tuple_6)); /* "numpy.pxd":798 * * if (end - f) - (new_offset - offset[0]) < 15: * raise RuntimeError(u"Format string allocated too short, see comment in numpy.pxd") # <<<<<<<<<<<<<< * * if ((child.byteorder == '>' and little_endian) or */ __pyx_k_tuple_9 = PyTuple_New(1); if (unlikely(!__pyx_k_tuple_9)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 798; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_k_tuple_9); __Pyx_INCREF(((PyObject *)__pyx_kp_u_8)); PyTuple_SET_ITEM(__pyx_k_tuple_9, 0, ((PyObject *)__pyx_kp_u_8)); __Pyx_GIVEREF(((PyObject *)__pyx_kp_u_8)); __Pyx_GIVEREF(((PyObject *)__pyx_k_tuple_9)); /* "numpy.pxd":802 * if ((child.byteorder == '>' and little_endian) or * (child.byteorder == '<' and not little_endian)): * raise ValueError(u"Non-native byte order not supported") # <<<<<<<<<<<<<< * # One could encode it in the format string and have Cython * # complain instead, BUT: < and > in format strings also imply */ __pyx_k_tuple_10 = PyTuple_New(1); if (unlikely(!__pyx_k_tuple_10)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 802; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_k_tuple_10); __Pyx_INCREF(((PyObject *)__pyx_kp_u_5)); PyTuple_SET_ITEM(__pyx_k_tuple_10, 0, ((PyObject *)__pyx_kp_u_5)); __Pyx_GIVEREF(((PyObject *)__pyx_kp_u_5)); __Pyx_GIVEREF(((PyObject *)__pyx_k_tuple_10)); /* "numpy.pxd":822 * t = child.type_num * if end - f < 5: * raise RuntimeError(u"Format string allocated too short.") # <<<<<<<<<<<<<< * * # Until ticket #99 is fixed, use integers to avoid warnings */ __pyx_k_tuple_12 = PyTuple_New(1); if (unlikely(!__pyx_k_tuple_12)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 822; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_k_tuple_12); __Pyx_INCREF(((PyObject *)__pyx_kp_u_11)); PyTuple_SET_ITEM(__pyx_k_tuple_12, 0, ((PyObject *)__pyx_kp_u_11)); __Pyx_GIVEREF(((PyObject *)__pyx_kp_u_11)); __Pyx_GIVEREF(((PyObject *)__pyx_k_tuple_12)); /* "wfpt.pyx":28 * include "pdf.pxi" * * def pdf_array(np.ndarray[double, ndim=1] x, double v, double V, double a, double z, double Z, double t, double T, double err, bint logp=0, int nT=2, int nZ=2, bint use_adaptive=1, double simps_err=1e-3): # <<<<<<<<<<<<<< * cdef Py_ssize_t size = x.shape[0] * cdef Py_ssize_t i */ __pyx_k_tuple_13 = PyTuple_New(17); if (unlikely(!__pyx_k_tuple_13)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 28; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_k_tuple_13); __Pyx_INCREF(((PyObject *)__pyx_n_s__x)); PyTuple_SET_ITEM(__pyx_k_tuple_13, 0, ((PyObject *)__pyx_n_s__x)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__x)); __Pyx_INCREF(((PyObject *)__pyx_n_s__v)); PyTuple_SET_ITEM(__pyx_k_tuple_13, 1, ((PyObject *)__pyx_n_s__v)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__v)); __Pyx_INCREF(((PyObject *)__pyx_n_s__V)); PyTuple_SET_ITEM(__pyx_k_tuple_13, 2, ((PyObject *)__pyx_n_s__V)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__V)); __Pyx_INCREF(((PyObject *)__pyx_n_s__a)); PyTuple_SET_ITEM(__pyx_k_tuple_13, 3, ((PyObject *)__pyx_n_s__a)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__a)); __Pyx_INCREF(((PyObject *)__pyx_n_s__z)); PyTuple_SET_ITEM(__pyx_k_tuple_13, 4, ((PyObject *)__pyx_n_s__z)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__z)); __Pyx_INCREF(((PyObject *)__pyx_n_s__Z)); PyTuple_SET_ITEM(__pyx_k_tuple_13, 5, ((PyObject *)__pyx_n_s__Z)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__Z)); __Pyx_INCREF(((PyObject *)__pyx_n_s__t)); PyTuple_SET_ITEM(__pyx_k_tuple_13, 6, ((PyObject *)__pyx_n_s__t)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__t)); __Pyx_INCREF(((PyObject *)__pyx_n_s__T)); PyTuple_SET_ITEM(__pyx_k_tuple_13, 7, ((PyObject *)__pyx_n_s__T)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__T)); __Pyx_INCREF(((PyObject *)__pyx_n_s__err)); PyTuple_SET_ITEM(__pyx_k_tuple_13, 8, ((PyObject *)__pyx_n_s__err)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__err)); __Pyx_INCREF(((PyObject *)__pyx_n_s__logp)); PyTuple_SET_ITEM(__pyx_k_tuple_13, 9, ((PyObject *)__pyx_n_s__logp)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__logp)); __Pyx_INCREF(((PyObject *)__pyx_n_s__nT)); PyTuple_SET_ITEM(__pyx_k_tuple_13, 10, ((PyObject *)__pyx_n_s__nT)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__nT)); __Pyx_INCREF(((PyObject *)__pyx_n_s__nZ)); PyTuple_SET_ITEM(__pyx_k_tuple_13, 11, ((PyObject *)__pyx_n_s__nZ)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__nZ)); __Pyx_INCREF(((PyObject *)__pyx_n_s__use_adaptive)); PyTuple_SET_ITEM(__pyx_k_tuple_13, 12, ((PyObject *)__pyx_n_s__use_adaptive)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__use_adaptive)); __Pyx_INCREF(((PyObject *)__pyx_n_s__simps_err)); PyTuple_SET_ITEM(__pyx_k_tuple_13, 13, ((PyObject *)__pyx_n_s__simps_err)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__simps_err)); __Pyx_INCREF(((PyObject *)__pyx_n_s__size)); PyTuple_SET_ITEM(__pyx_k_tuple_13, 14, ((PyObject *)__pyx_n_s__size)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__size)); __Pyx_INCREF(((PyObject *)__pyx_n_s__i)); PyTuple_SET_ITEM(__pyx_k_tuple_13, 15, ((PyObject *)__pyx_n_s__i)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__i)); __Pyx_INCREF(((PyObject *)__pyx_n_s__y)); PyTuple_SET_ITEM(__pyx_k_tuple_13, 16, ((PyObject *)__pyx_n_s__y)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__y)); __Pyx_GIVEREF(((PyObject *)__pyx_k_tuple_13)); __pyx_k_codeobj_14 = (PyObject*)__Pyx_PyCode_New(14, 0, 17, 0, 0, __pyx_empty_bytes, __pyx_empty_tuple, __pyx_empty_tuple, __pyx_k_tuple_13, __pyx_empty_tuple, __pyx_empty_tuple, __pyx_kp_s_15, __pyx_n_s__pdf_array, 28, __pyx_empty_bytes); if (unlikely(!__pyx_k_codeobj_14)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 28; __pyx_clineno = __LINE__; goto __pyx_L1_error;} /* "wfpt.pyx":41 * return y * * def wiener_like(np.ndarray[double, ndim=1] x, double v, double V, double a, double z, double Z, double t, # <<<<<<<<<<<<<< * double T, double err, int nT=10, int nZ=10, bint use_adaptive=1, double simps_err=1e-8): * cdef Py_ssize_t size = x.shape[0] */ __pyx_k_tuple_16 = PyTuple_New(17); if (unlikely(!__pyx_k_tuple_16)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 41; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_k_tuple_16); __Pyx_INCREF(((PyObject *)__pyx_n_s__x)); PyTuple_SET_ITEM(__pyx_k_tuple_16, 0, ((PyObject *)__pyx_n_s__x)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__x)); __Pyx_INCREF(((PyObject *)__pyx_n_s__v)); PyTuple_SET_ITEM(__pyx_k_tuple_16, 1, ((PyObject *)__pyx_n_s__v)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__v)); __Pyx_INCREF(((PyObject *)__pyx_n_s__V)); PyTuple_SET_ITEM(__pyx_k_tuple_16, 2, ((PyObject *)__pyx_n_s__V)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__V)); __Pyx_INCREF(((PyObject *)__pyx_n_s__a)); PyTuple_SET_ITEM(__pyx_k_tuple_16, 3, ((PyObject *)__pyx_n_s__a)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__a)); __Pyx_INCREF(((PyObject *)__pyx_n_s__z)); PyTuple_SET_ITEM(__pyx_k_tuple_16, 4, ((PyObject *)__pyx_n_s__z)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__z)); __Pyx_INCREF(((PyObject *)__pyx_n_s__Z)); PyTuple_SET_ITEM(__pyx_k_tuple_16, 5, ((PyObject *)__pyx_n_s__Z)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__Z)); __Pyx_INCREF(((PyObject *)__pyx_n_s__t)); PyTuple_SET_ITEM(__pyx_k_tuple_16, 6, ((PyObject *)__pyx_n_s__t)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__t)); __Pyx_INCREF(((PyObject *)__pyx_n_s__T)); PyTuple_SET_ITEM(__pyx_k_tuple_16, 7, ((PyObject *)__pyx_n_s__T)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__T)); __Pyx_INCREF(((PyObject *)__pyx_n_s__err)); PyTuple_SET_ITEM(__pyx_k_tuple_16, 8, ((PyObject *)__pyx_n_s__err)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__err)); __Pyx_INCREF(((PyObject *)__pyx_n_s__nT)); PyTuple_SET_ITEM(__pyx_k_tuple_16, 9, ((PyObject *)__pyx_n_s__nT)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__nT)); __Pyx_INCREF(((PyObject *)__pyx_n_s__nZ)); PyTuple_SET_ITEM(__pyx_k_tuple_16, 10, ((PyObject *)__pyx_n_s__nZ)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__nZ)); __Pyx_INCREF(((PyObject *)__pyx_n_s__use_adaptive)); PyTuple_SET_ITEM(__pyx_k_tuple_16, 11, ((PyObject *)__pyx_n_s__use_adaptive)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__use_adaptive)); __Pyx_INCREF(((PyObject *)__pyx_n_s__simps_err)); PyTuple_SET_ITEM(__pyx_k_tuple_16, 12, ((PyObject *)__pyx_n_s__simps_err)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__simps_err)); __Pyx_INCREF(((PyObject *)__pyx_n_s__size)); PyTuple_SET_ITEM(__pyx_k_tuple_16, 13, ((PyObject *)__pyx_n_s__size)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__size)); __Pyx_INCREF(((PyObject *)__pyx_n_s__i)); PyTuple_SET_ITEM(__pyx_k_tuple_16, 14, ((PyObject *)__pyx_n_s__i)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__i)); __Pyx_INCREF(((PyObject *)__pyx_n_s__p)); PyTuple_SET_ITEM(__pyx_k_tuple_16, 15, ((PyObject *)__pyx_n_s__p)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__p)); __Pyx_INCREF(((PyObject *)__pyx_n_s__sum_logp)); PyTuple_SET_ITEM(__pyx_k_tuple_16, 16, ((PyObject *)__pyx_n_s__sum_logp)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__sum_logp)); __Pyx_GIVEREF(((PyObject *)__pyx_k_tuple_16)); __pyx_k_codeobj_17 = (PyObject*)__Pyx_PyCode_New(13, 0, 17, 0, 0, __pyx_empty_bytes, __pyx_empty_tuple, __pyx_empty_tuple, __pyx_k_tuple_16, __pyx_empty_tuple, __pyx_empty_tuple, __pyx_kp_s_15, __pyx_n_s__wiener_like, 41, __pyx_empty_bytes); if (unlikely(!__pyx_k_codeobj_17)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 41; __pyx_clineno = __LINE__; goto __pyx_L1_error;} /* "wfpt.pyx":62 * return sum_logp * * def wiener_like_array(np.ndarray[double, ndim=1] x, double v, double V, double a, double z, double Z, double t, # <<<<<<<<<<<<<< * double T, double err, int nT=10, int nZ=10, bint use_adaptive=1, double simps_err=1e-8): * cdef Py_ssize_t size = x.shape[0] */ __pyx_k_tuple_18 = PyTuple_New(17); if (unlikely(!__pyx_k_tuple_18)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 62; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_k_tuple_18); __Pyx_INCREF(((PyObject *)__pyx_n_s__x)); PyTuple_SET_ITEM(__pyx_k_tuple_18, 0, ((PyObject *)__pyx_n_s__x)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__x)); __Pyx_INCREF(((PyObject *)__pyx_n_s__v)); PyTuple_SET_ITEM(__pyx_k_tuple_18, 1, ((PyObject *)__pyx_n_s__v)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__v)); __Pyx_INCREF(((PyObject *)__pyx_n_s__V)); PyTuple_SET_ITEM(__pyx_k_tuple_18, 2, ((PyObject *)__pyx_n_s__V)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__V)); __Pyx_INCREF(((PyObject *)__pyx_n_s__a)); PyTuple_SET_ITEM(__pyx_k_tuple_18, 3, ((PyObject *)__pyx_n_s__a)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__a)); __Pyx_INCREF(((PyObject *)__pyx_n_s__z)); PyTuple_SET_ITEM(__pyx_k_tuple_18, 4, ((PyObject *)__pyx_n_s__z)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__z)); __Pyx_INCREF(((PyObject *)__pyx_n_s__Z)); PyTuple_SET_ITEM(__pyx_k_tuple_18, 5, ((PyObject *)__pyx_n_s__Z)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__Z)); __Pyx_INCREF(((PyObject *)__pyx_n_s__t)); PyTuple_SET_ITEM(__pyx_k_tuple_18, 6, ((PyObject *)__pyx_n_s__t)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__t)); __Pyx_INCREF(((PyObject *)__pyx_n_s__T)); PyTuple_SET_ITEM(__pyx_k_tuple_18, 7, ((PyObject *)__pyx_n_s__T)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__T)); __Pyx_INCREF(((PyObject *)__pyx_n_s__err)); PyTuple_SET_ITEM(__pyx_k_tuple_18, 8, ((PyObject *)__pyx_n_s__err)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__err)); __Pyx_INCREF(((PyObject *)__pyx_n_s__nT)); PyTuple_SET_ITEM(__pyx_k_tuple_18, 9, ((PyObject *)__pyx_n_s__nT)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__nT)); __Pyx_INCREF(((PyObject *)__pyx_n_s__nZ)); PyTuple_SET_ITEM(__pyx_k_tuple_18, 10, ((PyObject *)__pyx_n_s__nZ)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__nZ)); __Pyx_INCREF(((PyObject *)__pyx_n_s__use_adaptive)); PyTuple_SET_ITEM(__pyx_k_tuple_18, 11, ((PyObject *)__pyx_n_s__use_adaptive)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__use_adaptive)); __Pyx_INCREF(((PyObject *)__pyx_n_s__simps_err)); PyTuple_SET_ITEM(__pyx_k_tuple_18, 12, ((PyObject *)__pyx_n_s__simps_err)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__simps_err)); __Pyx_INCREF(((PyObject *)__pyx_n_s__size)); PyTuple_SET_ITEM(__pyx_k_tuple_18, 13, ((PyObject *)__pyx_n_s__size)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__size)); __Pyx_INCREF(((PyObject *)__pyx_n_s__i)); PyTuple_SET_ITEM(__pyx_k_tuple_18, 14, ((PyObject *)__pyx_n_s__i)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__i)); __Pyx_INCREF(((PyObject *)__pyx_n_s__sum_logp)); PyTuple_SET_ITEM(__pyx_k_tuple_18, 15, ((PyObject *)__pyx_n_s__sum_logp)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__sum_logp)); __Pyx_INCREF(((PyObject *)__pyx_n_s__p_array)); PyTuple_SET_ITEM(__pyx_k_tuple_18, 16, ((PyObject *)__pyx_n_s__p_array)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__p_array)); __Pyx_GIVEREF(((PyObject *)__pyx_k_tuple_18)); __pyx_k_codeobj_19 = (PyObject*)__Pyx_PyCode_New(13, 0, 17, 0, 0, __pyx_empty_bytes, __pyx_empty_tuple, __pyx_empty_tuple, __pyx_k_tuple_18, __pyx_empty_tuple, __pyx_empty_tuple, __pyx_kp_s_15, __pyx_n_s__wiener_like_array, 62, __pyx_empty_bytes); if (unlikely(!__pyx_k_codeobj_19)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 62; __pyx_clineno = __LINE__; goto __pyx_L1_error;} /* "wfpt.pyx":85 * return sum_logp * * def wiener_like_multi(np.ndarray[double, ndim=1] x, v, V, a, z, Z, t, T, double err, multi=None): # <<<<<<<<<<<<<< * cdef Py_ssize_t size = x.shape[0] * cdef Py_ssize_t i */ __pyx_k_tuple_20 = PyTuple_New(16); if (unlikely(!__pyx_k_tuple_20)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 85; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_k_tuple_20); __Pyx_INCREF(((PyObject *)__pyx_n_s__x)); PyTuple_SET_ITEM(__pyx_k_tuple_20, 0, ((PyObject *)__pyx_n_s__x)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__x)); __Pyx_INCREF(((PyObject *)__pyx_n_s__v)); PyTuple_SET_ITEM(__pyx_k_tuple_20, 1, ((PyObject *)__pyx_n_s__v)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__v)); __Pyx_INCREF(((PyObject *)__pyx_n_s__V)); PyTuple_SET_ITEM(__pyx_k_tuple_20, 2, ((PyObject *)__pyx_n_s__V)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__V)); __Pyx_INCREF(((PyObject *)__pyx_n_s__a)); PyTuple_SET_ITEM(__pyx_k_tuple_20, 3, ((PyObject *)__pyx_n_s__a)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__a)); __Pyx_INCREF(((PyObject *)__pyx_n_s__z)); PyTuple_SET_ITEM(__pyx_k_tuple_20, 4, ((PyObject *)__pyx_n_s__z)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__z)); __Pyx_INCREF(((PyObject *)__pyx_n_s__Z)); PyTuple_SET_ITEM(__pyx_k_tuple_20, 5, ((PyObject *)__pyx_n_s__Z)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__Z)); __Pyx_INCREF(((PyObject *)__pyx_n_s__t)); PyTuple_SET_ITEM(__pyx_k_tuple_20, 6, ((PyObject *)__pyx_n_s__t)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__t)); __Pyx_INCREF(((PyObject *)__pyx_n_s__T)); PyTuple_SET_ITEM(__pyx_k_tuple_20, 7, ((PyObject *)__pyx_n_s__T)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__T)); __Pyx_INCREF(((PyObject *)__pyx_n_s__err)); PyTuple_SET_ITEM(__pyx_k_tuple_20, 8, ((PyObject *)__pyx_n_s__err)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__err)); __Pyx_INCREF(((PyObject *)__pyx_n_s__multi)); PyTuple_SET_ITEM(__pyx_k_tuple_20, 9, ((PyObject *)__pyx_n_s__multi)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__multi)); __Pyx_INCREF(((PyObject *)__pyx_n_s__size)); PyTuple_SET_ITEM(__pyx_k_tuple_20, 10, ((PyObject *)__pyx_n_s__size)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__size)); __Pyx_INCREF(((PyObject *)__pyx_n_s__i)); PyTuple_SET_ITEM(__pyx_k_tuple_20, 11, ((PyObject *)__pyx_n_s__i)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__i)); __Pyx_INCREF(((PyObject *)__pyx_n_s__p)); PyTuple_SET_ITEM(__pyx_k_tuple_20, 12, ((PyObject *)__pyx_n_s__p)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__p)); __Pyx_INCREF(((PyObject *)__pyx_n_s__params)); PyTuple_SET_ITEM(__pyx_k_tuple_20, 13, ((PyObject *)__pyx_n_s__params)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__params)); __Pyx_INCREF(((PyObject *)__pyx_n_s__params_iter)); PyTuple_SET_ITEM(__pyx_k_tuple_20, 14, ((PyObject *)__pyx_n_s__params_iter)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__params_iter)); __Pyx_INCREF(((PyObject *)__pyx_n_s__param)); PyTuple_SET_ITEM(__pyx_k_tuple_20, 15, ((PyObject *)__pyx_n_s__param)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__param)); __Pyx_GIVEREF(((PyObject *)__pyx_k_tuple_20)); __pyx_k_codeobj_21 = (PyObject*)__Pyx_PyCode_New(10, 0, 16, 0, 0, __pyx_empty_bytes, __pyx_empty_tuple, __pyx_empty_tuple, __pyx_k_tuple_20, __pyx_empty_tuple, __pyx_empty_tuple, __pyx_kp_s_15, __pyx_n_s__wiener_like_multi, 85, __pyx_empty_bytes); if (unlikely(!__pyx_k_codeobj_21)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 85; __pyx_clineno = __LINE__; goto __pyx_L1_error;} /* "wfpt.pyx":105 * return p * * def gen_rts_from_cdf(double v, double V, double a, double z, double Z, double t, \ # <<<<<<<<<<<<<< * double T, int samples=1000, double cdf_lb=-6, double cdf_ub=6, double dt=1e-2): * */ __pyx_k_tuple_22 = PyTuple_New(22); if (unlikely(!__pyx_k_tuple_22)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 105; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_k_tuple_22); __Pyx_INCREF(((PyObject *)__pyx_n_s__v)); PyTuple_SET_ITEM(__pyx_k_tuple_22, 0, ((PyObject *)__pyx_n_s__v)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__v)); __Pyx_INCREF(((PyObject *)__pyx_n_s__V)); PyTuple_SET_ITEM(__pyx_k_tuple_22, 1, ((PyObject *)__pyx_n_s__V)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__V)); __Pyx_INCREF(((PyObject *)__pyx_n_s__a)); PyTuple_SET_ITEM(__pyx_k_tuple_22, 2, ((PyObject *)__pyx_n_s__a)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__a)); __Pyx_INCREF(((PyObject *)__pyx_n_s__z)); PyTuple_SET_ITEM(__pyx_k_tuple_22, 3, ((PyObject *)__pyx_n_s__z)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__z)); __Pyx_INCREF(((PyObject *)__pyx_n_s__Z)); PyTuple_SET_ITEM(__pyx_k_tuple_22, 4, ((PyObject *)__pyx_n_s__Z)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__Z)); __Pyx_INCREF(((PyObject *)__pyx_n_s__t)); PyTuple_SET_ITEM(__pyx_k_tuple_22, 5, ((PyObject *)__pyx_n_s__t)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__t)); __Pyx_INCREF(((PyObject *)__pyx_n_s__T)); PyTuple_SET_ITEM(__pyx_k_tuple_22, 6, ((PyObject *)__pyx_n_s__T)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__T)); __Pyx_INCREF(((PyObject *)__pyx_n_s__samples)); PyTuple_SET_ITEM(__pyx_k_tuple_22, 7, ((PyObject *)__pyx_n_s__samples)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__samples)); __Pyx_INCREF(((PyObject *)__pyx_n_s__cdf_lb)); PyTuple_SET_ITEM(__pyx_k_tuple_22, 8, ((PyObject *)__pyx_n_s__cdf_lb)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__cdf_lb)); __Pyx_INCREF(((PyObject *)__pyx_n_s__cdf_ub)); PyTuple_SET_ITEM(__pyx_k_tuple_22, 9, ((PyObject *)__pyx_n_s__cdf_ub)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__cdf_ub)); __Pyx_INCREF(((PyObject *)__pyx_n_s__dt)); PyTuple_SET_ITEM(__pyx_k_tuple_22, 10, ((PyObject *)__pyx_n_s__dt)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__dt)); __Pyx_INCREF(((PyObject *)__pyx_n_s__x)); PyTuple_SET_ITEM(__pyx_k_tuple_22, 11, ((PyObject *)__pyx_n_s__x)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__x)); __Pyx_INCREF(((PyObject *)__pyx_n_s__l_cdf)); PyTuple_SET_ITEM(__pyx_k_tuple_22, 12, ((PyObject *)__pyx_n_s__l_cdf)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__l_cdf)); __Pyx_INCREF(((PyObject *)__pyx_n_s__pdf)); PyTuple_SET_ITEM(__pyx_k_tuple_22, 13, ((PyObject *)__pyx_n_s__pdf)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__pdf)); __Pyx_INCREF(((PyObject *)__pyx_n_s__rt)); PyTuple_SET_ITEM(__pyx_k_tuple_22, 14, ((PyObject *)__pyx_n_s__rt)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__rt)); __Pyx_INCREF(((PyObject *)__pyx_n_s__size)); PyTuple_SET_ITEM(__pyx_k_tuple_22, 15, ((PyObject *)__pyx_n_s__size)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__size)); __Pyx_INCREF(((PyObject *)__pyx_n_s__i)); PyTuple_SET_ITEM(__pyx_k_tuple_22, 16, ((PyObject *)__pyx_n_s__i)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__i)); __Pyx_INCREF(((PyObject *)__pyx_n_s__j)); PyTuple_SET_ITEM(__pyx_k_tuple_22, 17, ((PyObject *)__pyx_n_s__j)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__j)); __Pyx_INCREF(((PyObject *)__pyx_n_s__idx)); PyTuple_SET_ITEM(__pyx_k_tuple_22, 18, ((PyObject *)__pyx_n_s__idx)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__idx)); __Pyx_INCREF(((PyObject *)__pyx_n_s__rts)); PyTuple_SET_ITEM(__pyx_k_tuple_22, 19, ((PyObject *)__pyx_n_s__rts)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__rts)); __Pyx_INCREF(((PyObject *)__pyx_n_s__f)); PyTuple_SET_ITEM(__pyx_k_tuple_22, 20, ((PyObject *)__pyx_n_s__f)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__f)); __Pyx_INCREF(((PyObject *)__pyx_n_s__delay)); PyTuple_SET_ITEM(__pyx_k_tuple_22, 21, ((PyObject *)__pyx_n_s__delay)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__delay)); __Pyx_GIVEREF(((PyObject *)__pyx_k_tuple_22)); __pyx_k_codeobj_23 = (PyObject*)__Pyx_PyCode_New(11, 0, 22, 0, 0, __pyx_empty_bytes, __pyx_empty_tuple, __pyx_empty_tuple, __pyx_k_tuple_22, __pyx_empty_tuple, __pyx_empty_tuple, __pyx_kp_s_15, __pyx_n_s__gen_rts_from_cdf, 105, __pyx_empty_bytes); if (unlikely(!__pyx_k_codeobj_23)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 105; __pyx_clineno = __LINE__; goto __pyx_L1_error;} /* "wfpt.pyx":138 * return rts * * def wiener_like_contaminant(np.ndarray[double, ndim=1] x, np.ndarray[int, ndim=1] cont_x, double v, \ # <<<<<<<<<<<<<< * double V, double a, double z, double Z, double t, double T, double t_min, \ * double t_max, double err, int nT= 10, int nZ=10, bint use_adaptive=1, \ */ __pyx_k_tuple_24 = PyTuple_New(22); if (unlikely(!__pyx_k_tuple_24)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 138; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_k_tuple_24); __Pyx_INCREF(((PyObject *)__pyx_n_s__x)); PyTuple_SET_ITEM(__pyx_k_tuple_24, 0, ((PyObject *)__pyx_n_s__x)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__x)); __Pyx_INCREF(((PyObject *)__pyx_n_s__cont_x)); PyTuple_SET_ITEM(__pyx_k_tuple_24, 1, ((PyObject *)__pyx_n_s__cont_x)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__cont_x)); __Pyx_INCREF(((PyObject *)__pyx_n_s__v)); PyTuple_SET_ITEM(__pyx_k_tuple_24, 2, ((PyObject *)__pyx_n_s__v)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__v)); __Pyx_INCREF(((PyObject *)__pyx_n_s__V)); PyTuple_SET_ITEM(__pyx_k_tuple_24, 3, ((PyObject *)__pyx_n_s__V)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__V)); __Pyx_INCREF(((PyObject *)__pyx_n_s__a)); PyTuple_SET_ITEM(__pyx_k_tuple_24, 4, ((PyObject *)__pyx_n_s__a)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__a)); __Pyx_INCREF(((PyObject *)__pyx_n_s__z)); PyTuple_SET_ITEM(__pyx_k_tuple_24, 5, ((PyObject *)__pyx_n_s__z)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__z)); __Pyx_INCREF(((PyObject *)__pyx_n_s__Z)); PyTuple_SET_ITEM(__pyx_k_tuple_24, 6, ((PyObject *)__pyx_n_s__Z)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__Z)); __Pyx_INCREF(((PyObject *)__pyx_n_s__t)); PyTuple_SET_ITEM(__pyx_k_tuple_24, 7, ((PyObject *)__pyx_n_s__t)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__t)); __Pyx_INCREF(((PyObject *)__pyx_n_s__T)); PyTuple_SET_ITEM(__pyx_k_tuple_24, 8, ((PyObject *)__pyx_n_s__T)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__T)); __Pyx_INCREF(((PyObject *)__pyx_n_s__t_min)); PyTuple_SET_ITEM(__pyx_k_tuple_24, 9, ((PyObject *)__pyx_n_s__t_min)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__t_min)); __Pyx_INCREF(((PyObject *)__pyx_n_s__t_max)); PyTuple_SET_ITEM(__pyx_k_tuple_24, 10, ((PyObject *)__pyx_n_s__t_max)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__t_max)); __Pyx_INCREF(((PyObject *)__pyx_n_s__err)); PyTuple_SET_ITEM(__pyx_k_tuple_24, 11, ((PyObject *)__pyx_n_s__err)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__err)); __Pyx_INCREF(((PyObject *)__pyx_n_s__nT)); PyTuple_SET_ITEM(__pyx_k_tuple_24, 12, ((PyObject *)__pyx_n_s__nT)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__nT)); __Pyx_INCREF(((PyObject *)__pyx_n_s__nZ)); PyTuple_SET_ITEM(__pyx_k_tuple_24, 13, ((PyObject *)__pyx_n_s__nZ)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__nZ)); __Pyx_INCREF(((PyObject *)__pyx_n_s__use_adaptive)); PyTuple_SET_ITEM(__pyx_k_tuple_24, 14, ((PyObject *)__pyx_n_s__use_adaptive)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__use_adaptive)); __Pyx_INCREF(((PyObject *)__pyx_n_s__simps_err)); PyTuple_SET_ITEM(__pyx_k_tuple_24, 15, ((PyObject *)__pyx_n_s__simps_err)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__simps_err)); __Pyx_INCREF(((PyObject *)__pyx_n_s__size)); PyTuple_SET_ITEM(__pyx_k_tuple_24, 16, ((PyObject *)__pyx_n_s__size)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__size)); __Pyx_INCREF(((PyObject *)__pyx_n_s__i)); PyTuple_SET_ITEM(__pyx_k_tuple_24, 17, ((PyObject *)__pyx_n_s__i)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__i)); __Pyx_INCREF(((PyObject *)__pyx_n_s__p)); PyTuple_SET_ITEM(__pyx_k_tuple_24, 18, ((PyObject *)__pyx_n_s__p)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__p)); __Pyx_INCREF(((PyObject *)__pyx_n_s__sum_logp)); PyTuple_SET_ITEM(__pyx_k_tuple_24, 19, ((PyObject *)__pyx_n_s__sum_logp)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__sum_logp)); __Pyx_INCREF(((PyObject *)__pyx_n_s__n_cont)); PyTuple_SET_ITEM(__pyx_k_tuple_24, 20, ((PyObject *)__pyx_n_s__n_cont)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__n_cont)); __Pyx_INCREF(((PyObject *)__pyx_n_s__pos_cont)); PyTuple_SET_ITEM(__pyx_k_tuple_24, 21, ((PyObject *)__pyx_n_s__pos_cont)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__pos_cont)); __Pyx_GIVEREF(((PyObject *)__pyx_k_tuple_24)); __pyx_k_codeobj_25 = (PyObject*)__Pyx_PyCode_New(16, 0, 22, 0, 0, __pyx_empty_bytes, __pyx_empty_tuple, __pyx_empty_tuple, __pyx_k_tuple_24, __pyx_empty_tuple, __pyx_empty_tuple, __pyx_kp_s_15, __pyx_n_s_26, 138, __pyx_empty_bytes); if (unlikely(!__pyx_k_codeobj_25)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 138; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_RefNannyFinishContext(); return 0; __pyx_L1_error:; __Pyx_RefNannyFinishContext(); return -1; } static int __Pyx_InitGlobals(void) { /* Initialize NaN. The sign is irrelevant, an exponent with all bits 1 and a nonzero mantissa means NaN. If the first bit in the mantissa is 1, it is a quiet NaN. */ memset(&__PYX_NAN, 0xFF, sizeof(__PYX_NAN)); PyEval_InitThreads(); if (__Pyx_InitStrings(__pyx_string_tab) < 0) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 1; __pyx_clineno = __LINE__; goto __pyx_L1_error;}; __pyx_int_15 = PyInt_FromLong(15); if (unlikely(!__pyx_int_15)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 1; __pyx_clineno = __LINE__; goto __pyx_L1_error;}; return 0; __pyx_L1_error:; return -1; } #if PY_MAJOR_VERSION < 3 PyMODINIT_FUNC initwfpt(void); /*proto*/ PyMODINIT_FUNC initwfpt(void) #else PyMODINIT_FUNC PyInit_wfpt(void); /*proto*/ PyMODINIT_FUNC PyInit_wfpt(void) #endif { PyObject *__pyx_t_1 = NULL; PyObject *__pyx_t_2 = NULL; __Pyx_RefNannyDeclarations #if CYTHON_REFNANNY __Pyx_RefNanny = __Pyx_RefNannyImportAPI("refnanny"); if (!__Pyx_RefNanny) { PyErr_Clear(); __Pyx_RefNanny = __Pyx_RefNannyImportAPI("Cython.Runtime.refnanny"); if (!__Pyx_RefNanny) Py_FatalError("failed to import 'refnanny' module"); } #endif __Pyx_RefNannySetupContext("PyMODINIT_FUNC PyInit_wfpt(void)"); if ( __Pyx_check_binary_version() < 0) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 1; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __pyx_empty_tuple = PyTuple_New(0); if (unlikely(!__pyx_empty_tuple)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 1; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __pyx_empty_bytes = PyBytes_FromStringAndSize("", 0); if (unlikely(!__pyx_empty_bytes)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 1; __pyx_clineno = __LINE__; goto __pyx_L1_error;} #ifdef __Pyx_CyFunction_USED if (__Pyx_CyFunction_init() < 0) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 1; __pyx_clineno = __LINE__; goto __pyx_L1_error;} #endif #ifdef __Pyx_FusedFunction_USED if (__pyx_FusedFunction_init() < 0) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 1; __pyx_clineno = __LINE__; goto __pyx_L1_error;} #endif #ifdef __Pyx_Generator_USED if (__pyx_Generator_init() < 0) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 1; __pyx_clineno = __LINE__; goto __pyx_L1_error;} #endif /*--- Library function declarations ---*/ /*--- Threads initialization code ---*/ #if defined(__PYX_FORCE_INIT_THREADS) && __PYX_FORCE_INIT_THREADS #ifdef WITH_THREAD /* Python build with threading support? */ PyEval_InitThreads(); #endif #endif /*--- Module creation code ---*/ #if PY_MAJOR_VERSION < 3 __pyx_m = Py_InitModule4(__Pyx_NAMESTR("wfpt"), __pyx_methods, 0, 0, PYTHON_API_VERSION); #else __pyx_m = PyModule_Create(&__pyx_moduledef); #endif if (!__pyx_m) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 1; __pyx_clineno = __LINE__; goto __pyx_L1_error;}; #if PY_MAJOR_VERSION < 3 Py_INCREF(__pyx_m); #endif __pyx_b = PyImport_AddModule(__Pyx_NAMESTR(__Pyx_BUILTIN_MODULE_NAME)); if (!__pyx_b) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 1; __pyx_clineno = __LINE__; goto __pyx_L1_error;}; if (__Pyx_SetAttrString(__pyx_m, "__builtins__", __pyx_b) < 0) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 1; __pyx_clineno = __LINE__; goto __pyx_L1_error;}; /*--- Initialize various global constants etc. ---*/ if (unlikely(__Pyx_InitGlobals() < 0)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 1; __pyx_clineno = __LINE__; goto __pyx_L1_error;} if (__pyx_module_is_main_wfpt) { if (__Pyx_SetAttrString(__pyx_m, "__name__", __pyx_n_s____main__) < 0) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 1; __pyx_clineno = __LINE__; goto __pyx_L1_error;}; } /*--- Builtin init code ---*/ if (unlikely(__Pyx_InitCachedBuiltins() < 0)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 1; __pyx_clineno = __LINE__; goto __pyx_L1_error;} /*--- Constants init code ---*/ if (unlikely(__Pyx_InitCachedConstants() < 0)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 1; __pyx_clineno = __LINE__; goto __pyx_L1_error;} /*--- Global init code ---*/ /*--- Variable export code ---*/ /*--- Function export code ---*/ /*--- Type init code ---*/ /*--- Type import code ---*/ __pyx_ptype_5numpy_dtype = __Pyx_ImportType("numpy", "dtype", sizeof(PyArray_Descr), 0); if (unlikely(!__pyx_ptype_5numpy_dtype)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 154; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __pyx_ptype_5numpy_flatiter = __Pyx_ImportType("numpy", "flatiter", sizeof(PyArrayIterObject), 0); if (unlikely(!__pyx_ptype_5numpy_flatiter)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 164; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __pyx_ptype_5numpy_broadcast = __Pyx_ImportType("numpy", "broadcast", sizeof(PyArrayMultiIterObject), 0); if (unlikely(!__pyx_ptype_5numpy_broadcast)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 168; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __pyx_ptype_5numpy_ndarray = __Pyx_ImportType("numpy", "ndarray", sizeof(PyArrayObject), 0); if (unlikely(!__pyx_ptype_5numpy_ndarray)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 177; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __pyx_ptype_5numpy_ufunc = __Pyx_ImportType("numpy", "ufunc", sizeof(PyUFuncObject), 0); if (unlikely(!__pyx_ptype_5numpy_ufunc)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 860; __pyx_clineno = __LINE__; goto __pyx_L1_error;} /*--- Variable import code ---*/ /*--- Function import code ---*/ /*--- Execution code ---*/ /* "wfpt.pyx":15 * # GPLv3 * * import hddm # <<<<<<<<<<<<<< * * from copy import copy */ __pyx_t_1 = __Pyx_Import(((PyObject *)__pyx_n_s__hddm), 0, -1); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 15; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_1); if (PyObject_SetAttr(__pyx_m, __pyx_n_s__hddm, __pyx_t_1) < 0) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 15; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; /* "wfpt.pyx":17 * import hddm * * from copy import copy # <<<<<<<<<<<<<< * import numpy as np * cimport numpy as np */ __pyx_t_1 = PyList_New(1); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 17; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_1); __Pyx_INCREF(((PyObject *)__pyx_n_s__copy)); PyList_SET_ITEM(__pyx_t_1, 0, ((PyObject *)__pyx_n_s__copy)); __Pyx_GIVEREF(((PyObject *)__pyx_n_s__copy)); __pyx_t_2 = __Pyx_Import(((PyObject *)__pyx_n_s__copy), ((PyObject *)__pyx_t_1), -1); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 17; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_2); __Pyx_DECREF(((PyObject *)__pyx_t_1)); __pyx_t_1 = 0; __pyx_t_1 = PyObject_GetAttr(__pyx_t_2, __pyx_n_s__copy); if (__pyx_t_1 == NULL) { if (PyErr_ExceptionMatches(PyExc_AttributeError)) __Pyx_RaiseImportError(__pyx_n_s__copy); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 17; __pyx_clineno = __LINE__; goto __pyx_L1_error;} } __Pyx_GOTREF(__pyx_t_1); if (PyObject_SetAttr(__pyx_m, __pyx_n_s__copy, __pyx_t_1) < 0) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 17; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; /* "wfpt.pyx":18 * * from copy import copy * import numpy as np # <<<<<<<<<<<<<< * cimport numpy as np * */ __pyx_t_2 = __Pyx_Import(((PyObject *)__pyx_n_s__numpy), 0, -1); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 18; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_2); if (PyObject_SetAttr(__pyx_m, __pyx_n_s__np, __pyx_t_2) < 0) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 18; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; /* "/home/whyking/working/projects/hddm/src/integrate.pxi":6 * #cython: boundscheck=False * * import numpy as np # <<<<<<<<<<<<<< * cimport numpy as np * cimport cython */ __pyx_t_2 = __Pyx_Import(((PyObject *)__pyx_n_s__numpy), 0, -1); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[3]; __pyx_lineno = 6; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_2); if (PyObject_SetAttr(__pyx_m, __pyx_n_s__np, __pyx_t_2) < 0) {__pyx_filename = __pyx_f[3]; __pyx_lineno = 6; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; /* "wfpt.pyx":28 * include "pdf.pxi" * * def pdf_array(np.ndarray[double, ndim=1] x, double v, double V, double a, double z, double Z, double t, double T, double err, bint logp=0, int nT=2, int nZ=2, bint use_adaptive=1, double simps_err=1e-3): # <<<<<<<<<<<<<< * cdef Py_ssize_t size = x.shape[0] * cdef Py_ssize_t i */ __pyx_t_2 = PyCFunction_NewEx(&__pyx_mdef_4wfpt_3pdf_array, NULL, __pyx_n_s__wfpt); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 28; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_2); if (PyObject_SetAttr(__pyx_m, __pyx_n_s__pdf_array, __pyx_t_2) < 0) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 28; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; /* "wfpt.pyx":41 * return y * * def wiener_like(np.ndarray[double, ndim=1] x, double v, double V, double a, double z, double Z, double t, # <<<<<<<<<<<<<< * double T, double err, int nT=10, int nZ=10, bint use_adaptive=1, double simps_err=1e-8): * cdef Py_ssize_t size = x.shape[0] */ __pyx_t_2 = PyCFunction_NewEx(&__pyx_mdef_4wfpt_5wiener_like, NULL, __pyx_n_s__wfpt); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 41; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_2); if (PyObject_SetAttr(__pyx_m, __pyx_n_s__wiener_like, __pyx_t_2) < 0) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 41; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; /* "wfpt.pyx":62 * return sum_logp * * def wiener_like_array(np.ndarray[double, ndim=1] x, double v, double V, double a, double z, double Z, double t, # <<<<<<<<<<<<<< * double T, double err, int nT=10, int nZ=10, bint use_adaptive=1, double simps_err=1e-8): * cdef Py_ssize_t size = x.shape[0] */ __pyx_t_2 = PyCFunction_NewEx(&__pyx_mdef_4wfpt_7wiener_like_array, NULL, __pyx_n_s__wfpt); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 62; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_2); if (PyObject_SetAttr(__pyx_m, __pyx_n_s__wiener_like_array, __pyx_t_2) < 0) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 62; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; /* "wfpt.pyx":85 * return sum_logp * * def wiener_like_multi(np.ndarray[double, ndim=1] x, v, V, a, z, Z, t, T, double err, multi=None): # <<<<<<<<<<<<<< * cdef Py_ssize_t size = x.shape[0] * cdef Py_ssize_t i */ __pyx_t_2 = PyCFunction_NewEx(&__pyx_mdef_4wfpt_9wiener_like_multi, NULL, __pyx_n_s__wfpt); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 85; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_2); if (PyObject_SetAttr(__pyx_m, __pyx_n_s__wiener_like_multi, __pyx_t_2) < 0) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 85; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; /* "wfpt.pyx":105 * return p * * def gen_rts_from_cdf(double v, double V, double a, double z, double Z, double t, \ # <<<<<<<<<<<<<< * double T, int samples=1000, double cdf_lb=-6, double cdf_ub=6, double dt=1e-2): * */ __pyx_t_2 = PyCFunction_NewEx(&__pyx_mdef_4wfpt_11gen_rts_from_cdf, NULL, __pyx_n_s__wfpt); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 105; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_2); if (PyObject_SetAttr(__pyx_m, __pyx_n_s__gen_rts_from_cdf, __pyx_t_2) < 0) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 105; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; /* "wfpt.pyx":138 * return rts * * def wiener_like_contaminant(np.ndarray[double, ndim=1] x, np.ndarray[int, ndim=1] cont_x, double v, \ # <<<<<<<<<<<<<< * double V, double a, double z, double Z, double t, double T, double t_min, \ * double t_max, double err, int nT= 10, int nZ=10, bint use_adaptive=1, \ */ __pyx_t_2 = PyCFunction_NewEx(&__pyx_mdef_4wfpt_13wiener_like_contaminant, NULL, __pyx_n_s__wfpt); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 138; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_2); if (PyObject_SetAttr(__pyx_m, __pyx_n_s_26, __pyx_t_2) < 0) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 138; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; /* "wfpt.pyx":1 * #cython: embedsignature=True # <<<<<<<<<<<<<< * #cython: cdivision=True * #cython: wraparound=False */ __pyx_t_2 = PyDict_New(); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 1; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(((PyObject *)__pyx_t_2)); if (PyObject_SetAttr(__pyx_m, __pyx_n_s____test__, ((PyObject *)__pyx_t_2)) < 0) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 1; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(((PyObject *)__pyx_t_2)); __pyx_t_2 = 0; /* "numpy.pxd":974 * arr.base = baseptr * * cdef inline object get_array_base(ndarray arr): # <<<<<<<<<<<<<< * if arr.base is NULL: * return None */ goto __pyx_L0; __pyx_L1_error:; __Pyx_XDECREF(__pyx_t_1); __Pyx_XDECREF(__pyx_t_2); if (__pyx_m) { __Pyx_AddTraceback("init wfpt", __pyx_clineno, __pyx_lineno, __pyx_filename); Py_DECREF(__pyx_m); __pyx_m = 0; } else if (!PyErr_Occurred()) { PyErr_SetString(PyExc_ImportError, "init wfpt"); } __pyx_L0:; __Pyx_RefNannyFinishContext(); #if PY_MAJOR_VERSION < 3 return; #else return __pyx_m; #endif } /* Runtime support code */ #if CYTHON_REFNANNY static __Pyx_RefNannyAPIStruct *__Pyx_RefNannyImportAPI(const char *modname) { PyObject *m = NULL, *p = NULL; void *r = NULL; m = PyImport_ImportModule((char *)modname); if (!m) goto end; p = PyObject_GetAttrString(m, (char *)"RefNannyAPI"); if (!p) goto end; r = PyLong_AsVoidPtr(p); end: Py_XDECREF(p); Py_XDECREF(m); return (__Pyx_RefNannyAPIStruct *)r; } #endif /* CYTHON_REFNANNY */ static PyObject *__Pyx_GetName(PyObject *dict, PyObject *name) { PyObject *result; result = PyObject_GetAttr(dict, name); if (!result) { if (dict != __pyx_b) { PyErr_Clear(); result = PyObject_GetAttr(__pyx_b, name); } if (!result) { PyErr_SetObject(PyExc_NameError, name); } } return result; } static void __Pyx_RaiseArgtupleInvalid( const char* func_name, int exact, Py_ssize_t num_min, Py_ssize_t num_max, Py_ssize_t num_found) { Py_ssize_t num_expected; const char *more_or_less; if (num_found < num_min) { num_expected = num_min; more_or_less = "at least"; } else { num_expected = num_max; more_or_less = "at most"; } if (exact) { more_or_less = "exactly"; } PyErr_Format(PyExc_TypeError, "%s() takes %s %"PY_FORMAT_SIZE_T"d positional argument%s (%"PY_FORMAT_SIZE_T"d given)", func_name, more_or_less, num_expected, (num_expected == 1) ? "" : "s", num_found); } static void __Pyx_RaiseDoubleKeywordsError( const char* func_name, PyObject* kw_name) { PyErr_Format(PyExc_TypeError, #if PY_MAJOR_VERSION >= 3 "%s() got multiple values for keyword argument '%U'", func_name, kw_name); #else "%s() got multiple values for keyword argument '%s'", func_name, PyString_AS_STRING(kw_name)); #endif } static int __Pyx_ParseOptionalKeywords( PyObject *kwds, PyObject **argnames[], PyObject *kwds2, PyObject *values[], Py_ssize_t num_pos_args, const char* function_name) { PyObject *key = 0, *value = 0; Py_ssize_t pos = 0; PyObject*** name; PyObject*** first_kw_arg = argnames + num_pos_args; while (PyDict_Next(kwds, &pos, &key, &value)) { name = first_kw_arg; while (*name && (**name != key)) name++; if (*name) { values[name-argnames] = value; } else { #if PY_MAJOR_VERSION < 3 if (unlikely(!PyString_CheckExact(key)) && unlikely(!PyString_Check(key))) { #else if (unlikely(!PyUnicode_CheckExact(key)) && unlikely(!PyUnicode_Check(key))) { #endif goto invalid_keyword_type; } else { for (name = first_kw_arg; *name; name++) { #if PY_MAJOR_VERSION >= 3 if (PyUnicode_GET_SIZE(**name) == PyUnicode_GET_SIZE(key) && PyUnicode_Compare(**name, key) == 0) break; #else if (PyString_GET_SIZE(**name) == PyString_GET_SIZE(key) && _PyString_Eq(**name, key)) break; #endif } if (*name) { values[name-argnames] = value; } else { for (name=argnames; name != first_kw_arg; name++) { if (**name == key) goto arg_passed_twice; #if PY_MAJOR_VERSION >= 3 if (PyUnicode_GET_SIZE(**name) == PyUnicode_GET_SIZE(key) && PyUnicode_Compare(**name, key) == 0) goto arg_passed_twice; #else if (PyString_GET_SIZE(**name) == PyString_GET_SIZE(key) && _PyString_Eq(**name, key)) goto arg_passed_twice; #endif } if (kwds2) { if (unlikely(PyDict_SetItem(kwds2, key, value))) goto bad; } else { goto invalid_keyword; } } } } } return 0; arg_passed_twice: __Pyx_RaiseDoubleKeywordsError(function_name, **name); goto bad; invalid_keyword_type: PyErr_Format(PyExc_TypeError, "%s() keywords must be strings", function_name); goto bad; invalid_keyword: PyErr_Format(PyExc_TypeError, #if PY_MAJOR_VERSION < 3 "%s() got an unexpected keyword argument '%s'", function_name, PyString_AsString(key)); #else "%s() got an unexpected keyword argument '%U'", function_name, key); #endif bad: return -1; } static int __Pyx_ArgTypeTest(PyObject *obj, PyTypeObject *type, int none_allowed, const char *name, int exact) { if (!type) { PyErr_Format(PyExc_SystemError, "Missing type object"); return 0; } if (none_allowed && obj == Py_None) return 1; else if (exact) { if (Py_TYPE(obj) == type) return 1; } else { if (PyObject_TypeCheck(obj, type)) return 1; } PyErr_Format(PyExc_TypeError, "Argument '%s' has incorrect type (expected %s, got %s)", name, type->tp_name, Py_TYPE(obj)->tp_name); return 0; } static CYTHON_INLINE int __Pyx_IsLittleEndian(void) { unsigned int n = 1; return *(unsigned char*)(&n) != 0; } static void __Pyx_BufFmt_Init(__Pyx_BufFmt_Context* ctx, __Pyx_BufFmt_StackElem* stack, __Pyx_TypeInfo* type) { stack[0].field = &ctx->root; stack[0].parent_offset = 0; ctx->root.type = type; ctx->root.name = "buffer dtype"; ctx->root.offset = 0; ctx->head = stack; ctx->head->field = &ctx->root; ctx->fmt_offset = 0; ctx->head->parent_offset = 0; ctx->new_packmode = '@'; ctx->enc_packmode = '@'; ctx->new_count = 1; ctx->enc_count = 0; ctx->enc_type = 0; ctx->is_complex = 0; ctx->is_valid_array = 0; ctx->struct_alignment = 0; while (type->typegroup == 'S') { ++ctx->head; ctx->head->field = type->fields; ctx->head->parent_offset = 0; type = type->fields->type; } } static int __Pyx_BufFmt_ParseNumber(const char** ts) { int count; const char* t = *ts; if (*t < '0' || *t > '9') { return -1; } else { count = *t++ - '0'; while (*t >= '0' && *t < '9') { count *= 10; count += *t++ - '0'; } } *ts = t; return count; } static int __Pyx_BufFmt_ExpectNumber(const char **ts) { int number = __Pyx_BufFmt_ParseNumber(ts); if (number == -1) /* First char was not a digit */ PyErr_Format(PyExc_ValueError,\ "Does not understand character buffer dtype format string ('%c')", **ts); return number; } static void __Pyx_BufFmt_RaiseUnexpectedChar(char ch) { PyErr_Format(PyExc_ValueError, "Unexpected format string character: '%c'", ch); } static const char* __Pyx_BufFmt_DescribeTypeChar(char ch, int is_complex) { switch (ch) { case 'b': return "'char'"; case 'B': return "'unsigned char'"; case 'h': return "'short'"; case 'H': return "'unsigned short'"; case 'i': return "'int'"; case 'I': return "'unsigned int'"; case 'l': return "'long'"; case 'L': return "'unsigned long'"; case 'q': return "'long long'"; case 'Q': return "'unsigned long long'"; case 'f': return (is_complex ? "'complex float'" : "'float'"); case 'd': return (is_complex ? "'complex double'" : "'double'"); case 'g': return (is_complex ? "'complex long double'" : "'long double'"); case 'T': return "a struct"; case 'O': return "Python object"; case 'P': return "a pointer"; case 's': case 'p': return "a string"; case 0: return "end"; default: return "unparseable format string"; } } static size_t __Pyx_BufFmt_TypeCharToStandardSize(char ch, int is_complex) { switch (ch) { case '?': case 'c': case 'b': case 'B': case 's': case 'p': return 1; case 'h': case 'H': return 2; case 'i': case 'I': case 'l': case 'L': return 4; case 'q': case 'Q': return 8; case 'f': return (is_complex ? 8 : 4); case 'd': return (is_complex ? 16 : 8); case 'g': { PyErr_SetString(PyExc_ValueError, "Python does not define a standard format string size for long double ('g').."); return 0; } case 'O': case 'P': return sizeof(void*); default: __Pyx_BufFmt_RaiseUnexpectedChar(ch); return 0; } } static size_t __Pyx_BufFmt_TypeCharToNativeSize(char ch, int is_complex) { switch (ch) { case 'c': case 'b': case 'B': case 's': case 'p': return 1; case 'h': case 'H': return sizeof(short); case 'i': case 'I': return sizeof(int); case 'l': case 'L': return sizeof(long); #ifdef HAVE_LONG_LONG case 'q': case 'Q': return sizeof(PY_LONG_LONG); #endif case 'f': return sizeof(float) * (is_complex ? 2 : 1); case 'd': return sizeof(double) * (is_complex ? 2 : 1); case 'g': return sizeof(long double) * (is_complex ? 2 : 1); case 'O': case 'P': return sizeof(void*); default: { __Pyx_BufFmt_RaiseUnexpectedChar(ch); return 0; } } } typedef struct { char c; short x; } __Pyx_st_short; typedef struct { char c; int x; } __Pyx_st_int; typedef struct { char c; long x; } __Pyx_st_long; typedef struct { char c; float x; } __Pyx_st_float; typedef struct { char c; double x; } __Pyx_st_double; typedef struct { char c; long double x; } __Pyx_st_longdouble; typedef struct { char c; void *x; } __Pyx_st_void_p; #ifdef HAVE_LONG_LONG typedef struct { char c; PY_LONG_LONG x; } __Pyx_st_longlong; #endif static size_t __Pyx_BufFmt_TypeCharToAlignment(char ch, int is_complex) { switch (ch) { case '?': case 'c': case 'b': case 'B': case 's': case 'p': return 1; case 'h': case 'H': return sizeof(__Pyx_st_short) - sizeof(short); case 'i': case 'I': return sizeof(__Pyx_st_int) - sizeof(int); case 'l': case 'L': return sizeof(__Pyx_st_long) - sizeof(long); #ifdef HAVE_LONG_LONG case 'q': case 'Q': return sizeof(__Pyx_st_longlong) - sizeof(PY_LONG_LONG); #endif case 'f': return sizeof(__Pyx_st_float) - sizeof(float); case 'd': return sizeof(__Pyx_st_double) - sizeof(double); case 'g': return sizeof(__Pyx_st_longdouble) - sizeof(long double); case 'P': case 'O': return sizeof(__Pyx_st_void_p) - sizeof(void*); default: __Pyx_BufFmt_RaiseUnexpectedChar(ch); return 0; } } /* These are for computing the padding at the end of the struct to align on the first member of the struct. This will probably the same as above, but we don't have any guarantees. */ typedef struct { short x; char c; } __Pyx_pad_short; typedef struct { int x; char c; } __Pyx_pad_int; typedef struct { long x; char c; } __Pyx_pad_long; typedef struct { float x; char c; } __Pyx_pad_float; typedef struct { double x; char c; } __Pyx_pad_double; typedef struct { long double x; char c; } __Pyx_pad_longdouble; typedef struct { void *x; char c; } __Pyx_pad_void_p; #ifdef HAVE_LONG_LONG typedef struct { PY_LONG_LONG x; char c; } __Pyx_pad_longlong; #endif static size_t __Pyx_BufFmt_TypeCharToPadding(char ch, int is_complex) { switch (ch) { case '?': case 'c': case 'b': case 'B': case 's': case 'p': return 1; case 'h': case 'H': return sizeof(__Pyx_pad_short) - sizeof(short); case 'i': case 'I': return sizeof(__Pyx_pad_int) - sizeof(int); case 'l': case 'L': return sizeof(__Pyx_pad_long) - sizeof(long); #ifdef HAVE_LONG_LONG case 'q': case 'Q': return sizeof(__Pyx_pad_longlong) - sizeof(PY_LONG_LONG); #endif case 'f': return sizeof(__Pyx_pad_float) - sizeof(float); case 'd': return sizeof(__Pyx_pad_double) - sizeof(double); case 'g': return sizeof(__Pyx_pad_longdouble) - sizeof(long double); case 'P': case 'O': return sizeof(__Pyx_pad_void_p) - sizeof(void*); default: __Pyx_BufFmt_RaiseUnexpectedChar(ch); return 0; } } static char __Pyx_BufFmt_TypeCharToGroup(char ch, int is_complex) { switch (ch) { case 'c': case 'b': case 'h': case 'i': case 'l': case 'q': case 's': case 'p': return 'I'; case 'B': case 'H': case 'I': case 'L': case 'Q': return 'U'; case 'f': case 'd': case 'g': return (is_complex ? 'C' : 'R'); case 'O': return 'O'; case 'P': return 'P'; default: { __Pyx_BufFmt_RaiseUnexpectedChar(ch); return 0; } } } static void __Pyx_BufFmt_RaiseExpected(__Pyx_BufFmt_Context* ctx) { if (ctx->head == NULL || ctx->head->field == &ctx->root) { const char* expected; const char* quote; if (ctx->head == NULL) { expected = "end"; quote = ""; } else { expected = ctx->head->field->type->name; quote = "'"; } PyErr_Format(PyExc_ValueError, "Buffer dtype mismatch, expected %s%s%s but got %s", quote, expected, quote, __Pyx_BufFmt_DescribeTypeChar(ctx->enc_type, ctx->is_complex)); } else { __Pyx_StructField* field = ctx->head->field; __Pyx_StructField* parent = (ctx->head - 1)->field; PyErr_Format(PyExc_ValueError, "Buffer dtype mismatch, expected '%s' but got %s in '%s.%s'", field->type->name, __Pyx_BufFmt_DescribeTypeChar(ctx->enc_type, ctx->is_complex), parent->type->name, field->name); } } static int __Pyx_BufFmt_ProcessTypeChunk(__Pyx_BufFmt_Context* ctx) { char group; size_t size, offset, arraysize = 1; if (ctx->enc_type == 0) return 0; if (ctx->head->field->type->arraysize[0]) { int i, ndim = 0; if (ctx->enc_type == 's' || ctx->enc_type == 'p') { ctx->is_valid_array = ctx->head->field->type->ndim == 1; ndim = 1; if (ctx->enc_count != ctx->head->field->type->arraysize[0]) { PyErr_Format(PyExc_ValueError, "Expected a dimension of size %zu, got %zu", ctx->head->field->type->arraysize[0], ctx->enc_count); return -1; } } if (!ctx->is_valid_array) { PyErr_Format(PyExc_ValueError, "Expected %d dimensions, got %d", ctx->head->field->type->ndim, ndim); return -1; } for (i = 0; i < ctx->head->field->type->ndim; i++) { arraysize *= ctx->head->field->type->arraysize[i]; } ctx->is_valid_array = 0; ctx->enc_count = 1; } group = __Pyx_BufFmt_TypeCharToGroup(ctx->enc_type, ctx->is_complex); do { __Pyx_StructField* field = ctx->head->field; __Pyx_TypeInfo* type = field->type; if (ctx->enc_packmode == '@' || ctx->enc_packmode == '^') { size = __Pyx_BufFmt_TypeCharToNativeSize(ctx->enc_type, ctx->is_complex); } else { size = __Pyx_BufFmt_TypeCharToStandardSize(ctx->enc_type, ctx->is_complex); } if (ctx->enc_packmode == '@') { size_t align_at = __Pyx_BufFmt_TypeCharToAlignment(ctx->enc_type, ctx->is_complex); size_t align_mod_offset; if (align_at == 0) return -1; align_mod_offset = ctx->fmt_offset % align_at; if (align_mod_offset > 0) ctx->fmt_offset += align_at - align_mod_offset; if (ctx->struct_alignment == 0) ctx->struct_alignment = __Pyx_BufFmt_TypeCharToPadding(ctx->enc_type, ctx->is_complex); } if (type->size != size || type->typegroup != group) { if (type->typegroup == 'C' && type->fields != NULL) { size_t parent_offset = ctx->head->parent_offset + field->offset; ++ctx->head; ctx->head->field = type->fields; ctx->head->parent_offset = parent_offset; continue; } __Pyx_BufFmt_RaiseExpected(ctx); return -1; } offset = ctx->head->parent_offset + field->offset; if (ctx->fmt_offset != offset) { PyErr_Format(PyExc_ValueError, "Buffer dtype mismatch; next field is at offset %"PY_FORMAT_SIZE_T"d but %"PY_FORMAT_SIZE_T"d expected", (Py_ssize_t)ctx->fmt_offset, (Py_ssize_t)offset); return -1; } ctx->fmt_offset += size; if (arraysize) ctx->fmt_offset += (arraysize - 1) * size; --ctx->enc_count; /* Consume from buffer string */ while (1) { if (field == &ctx->root) { ctx->head = NULL; if (ctx->enc_count != 0) { __Pyx_BufFmt_RaiseExpected(ctx); return -1; } break; /* breaks both loops as ctx->enc_count == 0 */ } ctx->head->field = ++field; if (field->type == NULL) { --ctx->head; field = ctx->head->field; continue; } else if (field->type->typegroup == 'S') { size_t parent_offset = ctx->head->parent_offset + field->offset; if (field->type->fields->type == NULL) continue; /* empty struct */ field = field->type->fields; ++ctx->head; ctx->head->field = field; ctx->head->parent_offset = parent_offset; break; } else { break; } } } while (ctx->enc_count); ctx->enc_type = 0; ctx->is_complex = 0; return 0; } static CYTHON_INLINE PyObject * __pyx_buffmt_parse_array(__Pyx_BufFmt_Context* ctx, const char** tsp) { const char *ts = *tsp; int i = 0, number; int ndim = ctx->head->field->type->ndim; ; ++ts; if (ctx->new_count != 1) { PyErr_SetString(PyExc_ValueError, "Cannot handle repeated arrays in format string"); return NULL; } if (__Pyx_BufFmt_ProcessTypeChunk(ctx) == -1) return NULL; while (*ts && *ts != ')') { if (isspace(*ts)) continue; number = __Pyx_BufFmt_ExpectNumber(&ts); if (number == -1) return NULL; if (i < ndim && (size_t) number != ctx->head->field->type->arraysize[i]) return PyErr_Format(PyExc_ValueError, "Expected a dimension of size %zu, got %d", ctx->head->field->type->arraysize[i], number); if (*ts != ',' && *ts != ')') return PyErr_Format(PyExc_ValueError, "Expected a comma in format string, got '%c'", *ts); if (*ts == ',') ts++; i++; } if (i != ndim) return PyErr_Format(PyExc_ValueError, "Expected %d dimension(s), got %d", ctx->head->field->type->ndim, i); if (!*ts) { PyErr_SetString(PyExc_ValueError, "Unexpected end of format string, expected ')'"); return NULL; } ctx->is_valid_array = 1; ctx->new_count = 1; *tsp = ++ts; return Py_None; } static const char* __Pyx_BufFmt_CheckString(__Pyx_BufFmt_Context* ctx, const char* ts) { int got_Z = 0; while (1) { switch(*ts) { case 0: if (ctx->enc_type != 0 && ctx->head == NULL) { __Pyx_BufFmt_RaiseExpected(ctx); return NULL; } if (__Pyx_BufFmt_ProcessTypeChunk(ctx) == -1) return NULL; if (ctx->head != NULL) { __Pyx_BufFmt_RaiseExpected(ctx); return NULL; } return ts; case ' ': case 10: case 13: ++ts; break; case '<': if (!__Pyx_IsLittleEndian()) { PyErr_SetString(PyExc_ValueError, "Little-endian buffer not supported on big-endian compiler"); return NULL; } ctx->new_packmode = '='; ++ts; break; case '>': case '!': if (__Pyx_IsLittleEndian()) { PyErr_SetString(PyExc_ValueError, "Big-endian buffer not supported on little-endian compiler"); return NULL; } ctx->new_packmode = '='; ++ts; break; case '=': case '@': case '^': ctx->new_packmode = *ts++; break; case 'T': /* substruct */ { const char* ts_after_sub; size_t i, struct_count = ctx->new_count; size_t struct_alignment = ctx->struct_alignment; ctx->new_count = 1; ++ts; if (*ts != '{') { PyErr_SetString(PyExc_ValueError, "Buffer acquisition: Expected '{' after 'T'"); return NULL; } if (__Pyx_BufFmt_ProcessTypeChunk(ctx) == -1) return NULL; ctx->enc_type = 0; /* Erase processed last struct element */ ctx->enc_count = 0; ctx->struct_alignment = 0; ++ts; ts_after_sub = ts; for (i = 0; i != struct_count; ++i) { ts_after_sub = __Pyx_BufFmt_CheckString(ctx, ts); if (!ts_after_sub) return NULL; } ts = ts_after_sub; if (struct_alignment) ctx->struct_alignment = struct_alignment; } break; case '}': /* end of substruct; either repeat or move on */ { size_t alignment = ctx->struct_alignment; ++ts; if (__Pyx_BufFmt_ProcessTypeChunk(ctx) == -1) return NULL; ctx->enc_type = 0; /* Erase processed last struct element */ if (alignment && ctx->fmt_offset % alignment) { ctx->fmt_offset += alignment - (ctx->fmt_offset % alignment); } } return ts; case 'x': if (__Pyx_BufFmt_ProcessTypeChunk(ctx) == -1) return NULL; ctx->fmt_offset += ctx->new_count; ctx->new_count = 1; ctx->enc_count = 0; ctx->enc_type = 0; ctx->enc_packmode = ctx->new_packmode; ++ts; break; case 'Z': got_Z = 1; ++ts; if (*ts != 'f' && *ts != 'd' && *ts != 'g') { __Pyx_BufFmt_RaiseUnexpectedChar('Z'); return NULL; } /* fall through */ case 'c': case 'b': case 'B': case 'h': case 'H': case 'i': case 'I': case 'l': case 'L': case 'q': case 'Q': case 'f': case 'd': case 'g': case 'O': case 's': case 'p': if (ctx->enc_type == *ts && got_Z == ctx->is_complex && ctx->enc_packmode == ctx->new_packmode) { ctx->enc_count += ctx->new_count; } else { if (__Pyx_BufFmt_ProcessTypeChunk(ctx) == -1) return NULL; ctx->enc_count = ctx->new_count; ctx->enc_packmode = ctx->new_packmode; ctx->enc_type = *ts; ctx->is_complex = got_Z; } ++ts; ctx->new_count = 1; got_Z = 0; break; case ':': ++ts; while(*ts != ':') ++ts; ++ts; break; case '(': if (!__pyx_buffmt_parse_array(ctx, &ts)) return NULL; break; default: { int number = __Pyx_BufFmt_ExpectNumber(&ts); if (number == -1) return NULL; ctx->new_count = (size_t)number; } } } } static CYTHON_INLINE void __Pyx_ZeroBuffer(Py_buffer* buf) { buf->buf = NULL; buf->obj = NULL; buf->strides = __Pyx_zeros; buf->shape = __Pyx_zeros; buf->suboffsets = __Pyx_minusones; } static CYTHON_INLINE int __Pyx_GetBufferAndValidate( Py_buffer* buf, PyObject* obj, __Pyx_TypeInfo* dtype, int flags, int nd, int cast, __Pyx_BufFmt_StackElem* stack) { if (obj == Py_None || obj == NULL) { __Pyx_ZeroBuffer(buf); return 0; } buf->buf = NULL; if (__Pyx_GetBuffer(obj, buf, flags) == -1) goto fail; if (buf->ndim != nd) { PyErr_Format(PyExc_ValueError, "Buffer has wrong number of dimensions (expected %d, got %d)", nd, buf->ndim); goto fail; } if (!cast) { __Pyx_BufFmt_Context ctx; __Pyx_BufFmt_Init(&ctx, stack, dtype); if (!__Pyx_BufFmt_CheckString(&ctx, buf->format)) goto fail; } if ((unsigned)buf->itemsize != dtype->size) { PyErr_Format(PyExc_ValueError, "Item size of buffer (%"PY_FORMAT_SIZE_T"d byte%s) does not match size of '%s' (%"PY_FORMAT_SIZE_T"d byte%s)", buf->itemsize, (buf->itemsize > 1) ? "s" : "", dtype->name, (Py_ssize_t)dtype->size, (dtype->size > 1) ? "s" : ""); goto fail; } if (buf->suboffsets == NULL) buf->suboffsets = __Pyx_minusones; return 0; fail:; __Pyx_ZeroBuffer(buf); return -1; } static CYTHON_INLINE void __Pyx_SafeReleaseBuffer(Py_buffer* info) { if (info->buf == NULL) return; if (info->suboffsets == __Pyx_minusones) info->suboffsets = NULL; __Pyx_ReleaseBuffer(info); } static CYTHON_INLINE int __Pyx_TypeTest(PyObject *obj, PyTypeObject *type) { if (unlikely(!type)) { PyErr_Format(PyExc_SystemError, "Missing type object"); return 0; } if (likely(PyObject_TypeCheck(obj, type))) return 1; PyErr_Format(PyExc_TypeError, "Cannot convert %.200s to %.200s", Py_TYPE(obj)->tp_name, type->tp_name); return 0; } static CYTHON_INLINE void __Pyx_ErrRestore(PyObject *type, PyObject *value, PyObject *tb) { PyObject *tmp_type, *tmp_value, *tmp_tb; PyThreadState *tstate = PyThreadState_GET(); tmp_type = tstate->curexc_type; tmp_value = tstate->curexc_value; tmp_tb = tstate->curexc_traceback; tstate->curexc_type = type; tstate->curexc_value = value; tstate->curexc_traceback = tb; Py_XDECREF(tmp_type); Py_XDECREF(tmp_value); Py_XDECREF(tmp_tb); } static CYTHON_INLINE void __Pyx_ErrFetch(PyObject **type, PyObject **value, PyObject **tb) { PyThreadState *tstate = PyThreadState_GET(); *type = tstate->curexc_type; *value = tstate->curexc_value; *tb = tstate->curexc_traceback; tstate->curexc_type = 0; tstate->curexc_value = 0; tstate->curexc_traceback = 0; } static CYTHON_INLINE void __Pyx_RaiseNoneIndexingError(void) { PyErr_SetString(PyExc_TypeError, "'NoneType' object is unsubscriptable"); } static void __Pyx_RaiseBufferFallbackError(void) { PyErr_Format(PyExc_ValueError, "Buffer acquisition failed on assignment; and then reacquiring the old buffer failed too!"); } #if PY_MAJOR_VERSION < 3 static void __Pyx_Raise(PyObject *type, PyObject *value, PyObject *tb, CYTHON_UNUSED PyObject *cause) { /* cause is unused */ Py_XINCREF(type); Py_XINCREF(value); Py_XINCREF(tb); /* First, check the traceback argument, replacing None with NULL. */ if (tb == Py_None) { Py_DECREF(tb); tb = 0; } else if (tb != NULL && !PyTraceBack_Check(tb)) { PyErr_SetString(PyExc_TypeError, "raise: arg 3 must be a traceback or None"); goto raise_error; } /* Next, replace a missing value with None */ if (value == NULL) { value = Py_None; Py_INCREF(value); } #if PY_VERSION_HEX < 0x02050000 if (!PyClass_Check(type)) #else if (!PyType_Check(type)) #endif { /* Raising an instance. The value should be a dummy. */ if (value != Py_None) { PyErr_SetString(PyExc_TypeError, "instance exception may not have a separate value"); goto raise_error; } /* Normalize to raise , */ Py_DECREF(value); value = type; #if PY_VERSION_HEX < 0x02050000 if (PyInstance_Check(type)) { type = (PyObject*) ((PyInstanceObject*)type)->in_class; Py_INCREF(type); } else { type = 0; PyErr_SetString(PyExc_TypeError, "raise: exception must be an old-style class or instance"); goto raise_error; } #else type = (PyObject*) Py_TYPE(type); Py_INCREF(type); if (!PyType_IsSubtype((PyTypeObject *)type, (PyTypeObject *)PyExc_BaseException)) { PyErr_SetString(PyExc_TypeError, "raise: exception class must be a subclass of BaseException"); goto raise_error; } #endif } __Pyx_ErrRestore(type, value, tb); return; raise_error: Py_XDECREF(value); Py_XDECREF(type); Py_XDECREF(tb); return; } #else /* Python 3+ */ static void __Pyx_Raise(PyObject *type, PyObject *value, PyObject *tb, PyObject *cause) { if (tb == Py_None) { tb = 0; } else if (tb && !PyTraceBack_Check(tb)) { PyErr_SetString(PyExc_TypeError, "raise: arg 3 must be a traceback or None"); goto bad; } if (value == Py_None) value = 0; if (PyExceptionInstance_Check(type)) { if (value) { PyErr_SetString(PyExc_TypeError, "instance exception may not have a separate value"); goto bad; } value = type; type = (PyObject*) Py_TYPE(value); } else if (!PyExceptionClass_Check(type)) { PyErr_SetString(PyExc_TypeError, "raise: exception class must be a subclass of BaseException"); goto bad; } if (cause) { PyObject *fixed_cause; if (PyExceptionClass_Check(cause)) { fixed_cause = PyObject_CallObject(cause, NULL); if (fixed_cause == NULL) goto bad; } else if (PyExceptionInstance_Check(cause)) { fixed_cause = cause; Py_INCREF(fixed_cause); } else { PyErr_SetString(PyExc_TypeError, "exception causes must derive from " "BaseException"); goto bad; } if (!value) { value = PyObject_CallObject(type, NULL); } PyException_SetCause(value, fixed_cause); } PyErr_SetObject(type, value); if (tb) { PyThreadState *tstate = PyThreadState_GET(); PyObject* tmp_tb = tstate->curexc_traceback; if (tb != tmp_tb) { Py_INCREF(tb); tstate->curexc_traceback = tb; Py_XDECREF(tmp_tb); } } bad: return; } #endif static CYTHON_INLINE void __Pyx_RaiseNeedMoreValuesError(Py_ssize_t index) { PyErr_Format(PyExc_ValueError, "need more than %"PY_FORMAT_SIZE_T"d value%s to unpack", index, (index == 1) ? "" : "s"); } static CYTHON_INLINE void __Pyx_RaiseTooManyValuesError(Py_ssize_t expected) { PyErr_Format(PyExc_ValueError, "too many values to unpack (expected %"PY_FORMAT_SIZE_T"d)", expected); } static CYTHON_INLINE void __Pyx_RaiseNoneNotIterableError(void) { PyErr_SetString(PyExc_TypeError, "'NoneType' object is not iterable"); } static void __Pyx_UnpackTupleError(PyObject *t, Py_ssize_t index) { if (t == Py_None) { __Pyx_RaiseNoneNotIterableError(); } else if (PyTuple_GET_SIZE(t) < index) { __Pyx_RaiseNeedMoreValuesError(PyTuple_GET_SIZE(t)); } else { __Pyx_RaiseTooManyValuesError(index); } } #if PY_MAJOR_VERSION < 3 static int __Pyx_GetBuffer(PyObject *obj, Py_buffer *view, int flags) { #if PY_VERSION_HEX >= 0x02060000 if (PyObject_CheckBuffer(obj)) return PyObject_GetBuffer(obj, view, flags); #endif if (PyObject_TypeCheck(obj, __pyx_ptype_5numpy_ndarray)) return __pyx_pw_5numpy_7ndarray_1__getbuffer__(obj, view, flags); else { PyErr_Format(PyExc_TypeError, "'%100s' does not have the buffer interface", Py_TYPE(obj)->tp_name); return -1; } } static void __Pyx_ReleaseBuffer(Py_buffer *view) { PyObject* obj = view->obj; if (obj) { #if PY_VERSION_HEX >= 0x02060000 if (PyObject_CheckBuffer(obj)) {PyBuffer_Release(view); return;} #endif if (PyObject_TypeCheck(obj, __pyx_ptype_5numpy_ndarray)) __pyx_pw_5numpy_7ndarray_3__releasebuffer__(obj, view); Py_DECREF(obj); view->obj = NULL; } } #endif static PyObject *__Pyx_Import(PyObject *name, PyObject *from_list, long level) { PyObject *py_import = 0; PyObject *empty_list = 0; PyObject *module = 0; PyObject *global_dict = 0; PyObject *empty_dict = 0; PyObject *list; py_import = __Pyx_GetAttrString(__pyx_b, "__import__"); if (!py_import) goto bad; if (from_list) list = from_list; else { empty_list = PyList_New(0); if (!empty_list) goto bad; list = empty_list; } global_dict = PyModule_GetDict(__pyx_m); if (!global_dict) goto bad; empty_dict = PyDict_New(); if (!empty_dict) goto bad; #if PY_VERSION_HEX >= 0x02050000 { PyObject *py_level = PyInt_FromLong(level); if (!py_level) goto bad; module = PyObject_CallFunctionObjArgs(py_import, name, global_dict, empty_dict, list, py_level, NULL); Py_DECREF(py_level); } #else if (level>0) { PyErr_SetString(PyExc_RuntimeError, "Relative import is not supported for Python <=2.4."); goto bad; } module = PyObject_CallFunctionObjArgs(py_import, name, global_dict, empty_dict, list, NULL); #endif bad: Py_XDECREF(empty_list); Py_XDECREF(py_import); Py_XDECREF(empty_dict); return module; } static CYTHON_INLINE void __Pyx_RaiseImportError(PyObject *name) { #if PY_MAJOR_VERSION < 3 PyErr_Format(PyExc_ImportError, "cannot import name %.230s", PyString_AsString(name)); #else PyErr_Format(PyExc_ImportError, "cannot import name %S", name); #endif } static CYTHON_INLINE PyObject *__Pyx_PyInt_to_py_Py_intptr_t(Py_intptr_t val) { const Py_intptr_t neg_one = (Py_intptr_t)-1, const_zero = (Py_intptr_t)0; const int is_unsigned = const_zero < neg_one; if ((sizeof(Py_intptr_t) == sizeof(char)) || (sizeof(Py_intptr_t) == sizeof(short))) { return PyInt_FromLong((long)val); } else if ((sizeof(Py_intptr_t) == sizeof(int)) || (sizeof(Py_intptr_t) == sizeof(long))) { if (is_unsigned) return PyLong_FromUnsignedLong((unsigned long)val); else return PyInt_FromLong((long)val); } else if (sizeof(Py_intptr_t) == sizeof(PY_LONG_LONG)) { if (is_unsigned) return PyLong_FromUnsignedLongLong((unsigned PY_LONG_LONG)val); else return PyLong_FromLongLong((PY_LONG_LONG)val); } else { int one = 1; int little = (int)*(unsigned char *)&one; unsigned char *bytes = (unsigned char *)&val; return _PyLong_FromByteArray(bytes, sizeof(Py_intptr_t), little, !is_unsigned); } } #if CYTHON_CCOMPLEX #ifdef __cplusplus static CYTHON_INLINE __pyx_t_float_complex __pyx_t_float_complex_from_parts(float x, float y) { return ::std::complex< float >(x, y); } #else static CYTHON_INLINE __pyx_t_float_complex __pyx_t_float_complex_from_parts(float x, float y) { return x + y*(__pyx_t_float_complex)_Complex_I; } #endif #else static CYTHON_INLINE __pyx_t_float_complex __pyx_t_float_complex_from_parts(float x, float y) { __pyx_t_float_complex z; z.real = x; z.imag = y; return z; } #endif #if CYTHON_CCOMPLEX #else static CYTHON_INLINE int __Pyx_c_eqf(__pyx_t_float_complex a, __pyx_t_float_complex b) { return (a.real == b.real) && (a.imag == b.imag); } static CYTHON_INLINE __pyx_t_float_complex __Pyx_c_sumf(__pyx_t_float_complex a, __pyx_t_float_complex b) { __pyx_t_float_complex z; z.real = a.real + b.real; z.imag = a.imag + b.imag; return z; } static CYTHON_INLINE __pyx_t_float_complex __Pyx_c_difff(__pyx_t_float_complex a, __pyx_t_float_complex b) { __pyx_t_float_complex z; z.real = a.real - b.real; z.imag = a.imag - b.imag; return z; } static CYTHON_INLINE __pyx_t_float_complex __Pyx_c_prodf(__pyx_t_float_complex a, __pyx_t_float_complex b) { __pyx_t_float_complex z; z.real = a.real * b.real - a.imag * b.imag; z.imag = a.real * b.imag + a.imag * b.real; return z; } static CYTHON_INLINE __pyx_t_float_complex __Pyx_c_quotf(__pyx_t_float_complex a, __pyx_t_float_complex b) { __pyx_t_float_complex z; float denom = b.real * b.real + b.imag * b.imag; z.real = (a.real * b.real + a.imag * b.imag) / denom; z.imag = (a.imag * b.real - a.real * b.imag) / denom; return z; } static CYTHON_INLINE __pyx_t_float_complex __Pyx_c_negf(__pyx_t_float_complex a) { __pyx_t_float_complex z; z.real = -a.real; z.imag = -a.imag; return z; } static CYTHON_INLINE int __Pyx_c_is_zerof(__pyx_t_float_complex a) { return (a.real == 0) && (a.imag == 0); } static CYTHON_INLINE __pyx_t_float_complex __Pyx_c_conjf(__pyx_t_float_complex a) { __pyx_t_float_complex z; z.real = a.real; z.imag = -a.imag; return z; } #if 1 static CYTHON_INLINE float __Pyx_c_absf(__pyx_t_float_complex z) { #if !defined(HAVE_HYPOT) || defined(_MSC_VER) return sqrtf(z.real*z.real + z.imag*z.imag); #else return hypotf(z.real, z.imag); #endif } static CYTHON_INLINE __pyx_t_float_complex __Pyx_c_powf(__pyx_t_float_complex a, __pyx_t_float_complex b) { __pyx_t_float_complex z; float r, lnr, theta, z_r, z_theta; if (b.imag == 0 && b.real == (int)b.real) { if (b.real < 0) { float denom = a.real * a.real + a.imag * a.imag; a.real = a.real / denom; a.imag = -a.imag / denom; b.real = -b.real; } switch ((int)b.real) { case 0: z.real = 1; z.imag = 0; return z; case 1: return a; case 2: z = __Pyx_c_prodf(a, a); return __Pyx_c_prodf(a, a); case 3: z = __Pyx_c_prodf(a, a); return __Pyx_c_prodf(z, a); case 4: z = __Pyx_c_prodf(a, a); return __Pyx_c_prodf(z, z); } } if (a.imag == 0) { if (a.real == 0) { return a; } r = a.real; theta = 0; } else { r = __Pyx_c_absf(a); theta = atan2f(a.imag, a.real); } lnr = logf(r); z_r = expf(lnr * b.real - theta * b.imag); z_theta = theta * b.real + lnr * b.imag; z.real = z_r * cosf(z_theta); z.imag = z_r * sinf(z_theta); return z; } #endif #endif #if CYTHON_CCOMPLEX #ifdef __cplusplus static CYTHON_INLINE __pyx_t_double_complex __pyx_t_double_complex_from_parts(double x, double y) { return ::std::complex< double >(x, y); } #else static CYTHON_INLINE __pyx_t_double_complex __pyx_t_double_complex_from_parts(double x, double y) { return x + y*(__pyx_t_double_complex)_Complex_I; } #endif #else static CYTHON_INLINE __pyx_t_double_complex __pyx_t_double_complex_from_parts(double x, double y) { __pyx_t_double_complex z; z.real = x; z.imag = y; return z; } #endif #if CYTHON_CCOMPLEX #else static CYTHON_INLINE int __Pyx_c_eq(__pyx_t_double_complex a, __pyx_t_double_complex b) { return (a.real == b.real) && (a.imag == b.imag); } static CYTHON_INLINE __pyx_t_double_complex __Pyx_c_sum(__pyx_t_double_complex a, __pyx_t_double_complex b) { __pyx_t_double_complex z; z.real = a.real + b.real; z.imag = a.imag + b.imag; return z; } static CYTHON_INLINE __pyx_t_double_complex __Pyx_c_diff(__pyx_t_double_complex a, __pyx_t_double_complex b) { __pyx_t_double_complex z; z.real = a.real - b.real; z.imag = a.imag - b.imag; return z; } static CYTHON_INLINE __pyx_t_double_complex __Pyx_c_prod(__pyx_t_double_complex a, __pyx_t_double_complex b) { __pyx_t_double_complex z; z.real = a.real * b.real - a.imag * b.imag; z.imag = a.real * b.imag + a.imag * b.real; return z; } static CYTHON_INLINE __pyx_t_double_complex __Pyx_c_quot(__pyx_t_double_complex a, __pyx_t_double_complex b) { __pyx_t_double_complex z; double denom = b.real * b.real + b.imag * b.imag; z.real = (a.real * b.real + a.imag * b.imag) / denom; z.imag = (a.imag * b.real - a.real * b.imag) / denom; return z; } static CYTHON_INLINE __pyx_t_double_complex __Pyx_c_neg(__pyx_t_double_complex a) { __pyx_t_double_complex z; z.real = -a.real; z.imag = -a.imag; return z; } static CYTHON_INLINE int __Pyx_c_is_zero(__pyx_t_double_complex a) { return (a.real == 0) && (a.imag == 0); } static CYTHON_INLINE __pyx_t_double_complex __Pyx_c_conj(__pyx_t_double_complex a) { __pyx_t_double_complex z; z.real = a.real; z.imag = -a.imag; return z; } #if 1 static CYTHON_INLINE double __Pyx_c_abs(__pyx_t_double_complex z) { #if !defined(HAVE_HYPOT) || defined(_MSC_VER) return sqrt(z.real*z.real + z.imag*z.imag); #else return hypot(z.real, z.imag); #endif } static CYTHON_INLINE __pyx_t_double_complex __Pyx_c_pow(__pyx_t_double_complex a, __pyx_t_double_complex b) { __pyx_t_double_complex z; double r, lnr, theta, z_r, z_theta; if (b.imag == 0 && b.real == (int)b.real) { if (b.real < 0) { double denom = a.real * a.real + a.imag * a.imag; a.real = a.real / denom; a.imag = -a.imag / denom; b.real = -b.real; } switch ((int)b.real) { case 0: z.real = 1; z.imag = 0; return z; case 1: return a; case 2: z = __Pyx_c_prod(a, a); return __Pyx_c_prod(a, a); case 3: z = __Pyx_c_prod(a, a); return __Pyx_c_prod(z, a); case 4: z = __Pyx_c_prod(a, a); return __Pyx_c_prod(z, z); } } if (a.imag == 0) { if (a.real == 0) { return a; } r = a.real; theta = 0; } else { r = __Pyx_c_abs(a); theta = atan2(a.imag, a.real); } lnr = log(r); z_r = exp(lnr * b.real - theta * b.imag); z_theta = theta * b.real + lnr * b.imag; z.real = z_r * cos(z_theta); z.imag = z_r * sin(z_theta); return z; } #endif #endif static CYTHON_INLINE unsigned char __Pyx_PyInt_AsUnsignedChar(PyObject* x) { const unsigned char neg_one = (unsigned char)-1, const_zero = 0; const int is_unsigned = neg_one > const_zero; if (sizeof(unsigned char) < sizeof(long)) { long val = __Pyx_PyInt_AsLong(x); if (unlikely(val != (long)(unsigned char)val)) { if (!unlikely(val == -1 && PyErr_Occurred())) { PyErr_SetString(PyExc_OverflowError, (is_unsigned && unlikely(val < 0)) ? "can't convert negative value to unsigned char" : "value too large to convert to unsigned char"); } return (unsigned char)-1; } return (unsigned char)val; } return (unsigned char)__Pyx_PyInt_AsUnsignedLong(x); } static CYTHON_INLINE unsigned short __Pyx_PyInt_AsUnsignedShort(PyObject* x) { const unsigned short neg_one = (unsigned short)-1, const_zero = 0; const int is_unsigned = neg_one > const_zero; if (sizeof(unsigned short) < sizeof(long)) { long val = __Pyx_PyInt_AsLong(x); if (unlikely(val != (long)(unsigned short)val)) { if (!unlikely(val == -1 && PyErr_Occurred())) { PyErr_SetString(PyExc_OverflowError, (is_unsigned && unlikely(val < 0)) ? "can't convert negative value to unsigned short" : "value too large to convert to unsigned short"); } return (unsigned short)-1; } return (unsigned short)val; } return (unsigned short)__Pyx_PyInt_AsUnsignedLong(x); } static CYTHON_INLINE unsigned int __Pyx_PyInt_AsUnsignedInt(PyObject* x) { const unsigned int neg_one = (unsigned int)-1, const_zero = 0; const int is_unsigned = neg_one > const_zero; if (sizeof(unsigned int) < sizeof(long)) { long val = __Pyx_PyInt_AsLong(x); if (unlikely(val != (long)(unsigned int)val)) { if (!unlikely(val == -1 && PyErr_Occurred())) { PyErr_SetString(PyExc_OverflowError, (is_unsigned && unlikely(val < 0)) ? "can't convert negative value to unsigned int" : "value too large to convert to unsigned int"); } return (unsigned int)-1; } return (unsigned int)val; } return (unsigned int)__Pyx_PyInt_AsUnsignedLong(x); } static CYTHON_INLINE char __Pyx_PyInt_AsChar(PyObject* x) { const char neg_one = (char)-1, const_zero = 0; const int is_unsigned = neg_one > const_zero; if (sizeof(char) < sizeof(long)) { long val = __Pyx_PyInt_AsLong(x); if (unlikely(val != (long)(char)val)) { if (!unlikely(val == -1 && PyErr_Occurred())) { PyErr_SetString(PyExc_OverflowError, (is_unsigned && unlikely(val < 0)) ? "can't convert negative value to char" : "value too large to convert to char"); } return (char)-1; } return (char)val; } return (char)__Pyx_PyInt_AsLong(x); } static CYTHON_INLINE short __Pyx_PyInt_AsShort(PyObject* x) { const short neg_one = (short)-1, const_zero = 0; const int is_unsigned = neg_one > const_zero; if (sizeof(short) < sizeof(long)) { long val = __Pyx_PyInt_AsLong(x); if (unlikely(val != (long)(short)val)) { if (!unlikely(val == -1 && PyErr_Occurred())) { PyErr_SetString(PyExc_OverflowError, (is_unsigned && unlikely(val < 0)) ? "can't convert negative value to short" : "value too large to convert to short"); } return (short)-1; } return (short)val; } return (short)__Pyx_PyInt_AsLong(x); } static CYTHON_INLINE int __Pyx_PyInt_AsInt(PyObject* x) { const int neg_one = (int)-1, const_zero = 0; const int is_unsigned = neg_one > const_zero; if (sizeof(int) < sizeof(long)) { long val = __Pyx_PyInt_AsLong(x); if (unlikely(val != (long)(int)val)) { if (!unlikely(val == -1 && PyErr_Occurred())) { PyErr_SetString(PyExc_OverflowError, (is_unsigned && unlikely(val < 0)) ? "can't convert negative value to int" : "value too large to convert to int"); } return (int)-1; } return (int)val; } return (int)__Pyx_PyInt_AsLong(x); } static CYTHON_INLINE signed char __Pyx_PyInt_AsSignedChar(PyObject* x) { const signed char neg_one = (signed char)-1, const_zero = 0; const int is_unsigned = neg_one > const_zero; if (sizeof(signed char) < sizeof(long)) { long val = __Pyx_PyInt_AsLong(x); if (unlikely(val != (long)(signed char)val)) { if (!unlikely(val == -1 && PyErr_Occurred())) { PyErr_SetString(PyExc_OverflowError, (is_unsigned && unlikely(val < 0)) ? "can't convert negative value to signed char" : "value too large to convert to signed char"); } return (signed char)-1; } return (signed char)val; } return (signed char)__Pyx_PyInt_AsSignedLong(x); } static CYTHON_INLINE signed short __Pyx_PyInt_AsSignedShort(PyObject* x) { const signed short neg_one = (signed short)-1, const_zero = 0; const int is_unsigned = neg_one > const_zero; if (sizeof(signed short) < sizeof(long)) { long val = __Pyx_PyInt_AsLong(x); if (unlikely(val != (long)(signed short)val)) { if (!unlikely(val == -1 && PyErr_Occurred())) { PyErr_SetString(PyExc_OverflowError, (is_unsigned && unlikely(val < 0)) ? "can't convert negative value to signed short" : "value too large to convert to signed short"); } return (signed short)-1; } return (signed short)val; } return (signed short)__Pyx_PyInt_AsSignedLong(x); } static CYTHON_INLINE signed int __Pyx_PyInt_AsSignedInt(PyObject* x) { const signed int neg_one = (signed int)-1, const_zero = 0; const int is_unsigned = neg_one > const_zero; if (sizeof(signed int) < sizeof(long)) { long val = __Pyx_PyInt_AsLong(x); if (unlikely(val != (long)(signed int)val)) { if (!unlikely(val == -1 && PyErr_Occurred())) { PyErr_SetString(PyExc_OverflowError, (is_unsigned && unlikely(val < 0)) ? "can't convert negative value to signed int" : "value too large to convert to signed int"); } return (signed int)-1; } return (signed int)val; } return (signed int)__Pyx_PyInt_AsSignedLong(x); } static CYTHON_INLINE int __Pyx_PyInt_AsLongDouble(PyObject* x) { const int neg_one = (int)-1, const_zero = 0; const int is_unsigned = neg_one > const_zero; if (sizeof(int) < sizeof(long)) { long val = __Pyx_PyInt_AsLong(x); if (unlikely(val != (long)(int)val)) { if (!unlikely(val == -1 && PyErr_Occurred())) { PyErr_SetString(PyExc_OverflowError, (is_unsigned && unlikely(val < 0)) ? "can't convert negative value to int" : "value too large to convert to int"); } return (int)-1; } return (int)val; } return (int)__Pyx_PyInt_AsLong(x); } static CYTHON_INLINE unsigned long __Pyx_PyInt_AsUnsignedLong(PyObject* x) { const unsigned long neg_one = (unsigned long)-1, const_zero = 0; const int is_unsigned = neg_one > const_zero; #if PY_VERSION_HEX < 0x03000000 if (likely(PyInt_Check(x))) { long val = PyInt_AS_LONG(x); if (is_unsigned && unlikely(val < 0)) { PyErr_SetString(PyExc_OverflowError, "can't convert negative value to unsigned long"); return (unsigned long)-1; } return (unsigned long)val; } else #endif if (likely(PyLong_Check(x))) { if (is_unsigned) { if (unlikely(Py_SIZE(x) < 0)) { PyErr_SetString(PyExc_OverflowError, "can't convert negative value to unsigned long"); return (unsigned long)-1; } return (unsigned long)PyLong_AsUnsignedLong(x); } else { return (unsigned long)PyLong_AsLong(x); } } else { unsigned long val; PyObject *tmp = __Pyx_PyNumber_Int(x); if (!tmp) return (unsigned long)-1; val = __Pyx_PyInt_AsUnsignedLong(tmp); Py_DECREF(tmp); return val; } } static CYTHON_INLINE unsigned PY_LONG_LONG __Pyx_PyInt_AsUnsignedLongLong(PyObject* x) { const unsigned PY_LONG_LONG neg_one = (unsigned PY_LONG_LONG)-1, const_zero = 0; const int is_unsigned = neg_one > const_zero; #if PY_VERSION_HEX < 0x03000000 if (likely(PyInt_Check(x))) { long val = PyInt_AS_LONG(x); if (is_unsigned && unlikely(val < 0)) { PyErr_SetString(PyExc_OverflowError, "can't convert negative value to unsigned PY_LONG_LONG"); return (unsigned PY_LONG_LONG)-1; } return (unsigned PY_LONG_LONG)val; } else #endif if (likely(PyLong_Check(x))) { if (is_unsigned) { if (unlikely(Py_SIZE(x) < 0)) { PyErr_SetString(PyExc_OverflowError, "can't convert negative value to unsigned PY_LONG_LONG"); return (unsigned PY_LONG_LONG)-1; } return (unsigned PY_LONG_LONG)PyLong_AsUnsignedLongLong(x); } else { return (unsigned PY_LONG_LONG)PyLong_AsLongLong(x); } } else { unsigned PY_LONG_LONG val; PyObject *tmp = __Pyx_PyNumber_Int(x); if (!tmp) return (unsigned PY_LONG_LONG)-1; val = __Pyx_PyInt_AsUnsignedLongLong(tmp); Py_DECREF(tmp); return val; } } static CYTHON_INLINE long __Pyx_PyInt_AsLong(PyObject* x) { const long neg_one = (long)-1, const_zero = 0; const int is_unsigned = neg_one > const_zero; #if PY_VERSION_HEX < 0x03000000 if (likely(PyInt_Check(x))) { long val = PyInt_AS_LONG(x); if (is_unsigned && unlikely(val < 0)) { PyErr_SetString(PyExc_OverflowError, "can't convert negative value to long"); return (long)-1; } return (long)val; } else #endif if (likely(PyLong_Check(x))) { if (is_unsigned) { if (unlikely(Py_SIZE(x) < 0)) { PyErr_SetString(PyExc_OverflowError, "can't convert negative value to long"); return (long)-1; } return (long)PyLong_AsUnsignedLong(x); } else { return (long)PyLong_AsLong(x); } } else { long val; PyObject *tmp = __Pyx_PyNumber_Int(x); if (!tmp) return (long)-1; val = __Pyx_PyInt_AsLong(tmp); Py_DECREF(tmp); return val; } } static CYTHON_INLINE PY_LONG_LONG __Pyx_PyInt_AsLongLong(PyObject* x) { const PY_LONG_LONG neg_one = (PY_LONG_LONG)-1, const_zero = 0; const int is_unsigned = neg_one > const_zero; #if PY_VERSION_HEX < 0x03000000 if (likely(PyInt_Check(x))) { long val = PyInt_AS_LONG(x); if (is_unsigned && unlikely(val < 0)) { PyErr_SetString(PyExc_OverflowError, "can't convert negative value to PY_LONG_LONG"); return (PY_LONG_LONG)-1; } return (PY_LONG_LONG)val; } else #endif if (likely(PyLong_Check(x))) { if (is_unsigned) { if (unlikely(Py_SIZE(x) < 0)) { PyErr_SetString(PyExc_OverflowError, "can't convert negative value to PY_LONG_LONG"); return (PY_LONG_LONG)-1; } return (PY_LONG_LONG)PyLong_AsUnsignedLongLong(x); } else { return (PY_LONG_LONG)PyLong_AsLongLong(x); } } else { PY_LONG_LONG val; PyObject *tmp = __Pyx_PyNumber_Int(x); if (!tmp) return (PY_LONG_LONG)-1; val = __Pyx_PyInt_AsLongLong(tmp); Py_DECREF(tmp); return val; } } static CYTHON_INLINE signed long __Pyx_PyInt_AsSignedLong(PyObject* x) { const signed long neg_one = (signed long)-1, const_zero = 0; const int is_unsigned = neg_one > const_zero; #if PY_VERSION_HEX < 0x03000000 if (likely(PyInt_Check(x))) { long val = PyInt_AS_LONG(x); if (is_unsigned && unlikely(val < 0)) { PyErr_SetString(PyExc_OverflowError, "can't convert negative value to signed long"); return (signed long)-1; } return (signed long)val; } else #endif if (likely(PyLong_Check(x))) { if (is_unsigned) { if (unlikely(Py_SIZE(x) < 0)) { PyErr_SetString(PyExc_OverflowError, "can't convert negative value to signed long"); return (signed long)-1; } return (signed long)PyLong_AsUnsignedLong(x); } else { return (signed long)PyLong_AsLong(x); } } else { signed long val; PyObject *tmp = __Pyx_PyNumber_Int(x); if (!tmp) return (signed long)-1; val = __Pyx_PyInt_AsSignedLong(tmp); Py_DECREF(tmp); return val; } } static CYTHON_INLINE signed PY_LONG_LONG __Pyx_PyInt_AsSignedLongLong(PyObject* x) { const signed PY_LONG_LONG neg_one = (signed PY_LONG_LONG)-1, const_zero = 0; const int is_unsigned = neg_one > const_zero; #if PY_VERSION_HEX < 0x03000000 if (likely(PyInt_Check(x))) { long val = PyInt_AS_LONG(x); if (is_unsigned && unlikely(val < 0)) { PyErr_SetString(PyExc_OverflowError, "can't convert negative value to signed PY_LONG_LONG"); return (signed PY_LONG_LONG)-1; } return (signed PY_LONG_LONG)val; } else #endif if (likely(PyLong_Check(x))) { if (is_unsigned) { if (unlikely(Py_SIZE(x) < 0)) { PyErr_SetString(PyExc_OverflowError, "can't convert negative value to signed PY_LONG_LONG"); return (signed PY_LONG_LONG)-1; } return (signed PY_LONG_LONG)PyLong_AsUnsignedLongLong(x); } else { return (signed PY_LONG_LONG)PyLong_AsLongLong(x); } } else { signed PY_LONG_LONG val; PyObject *tmp = __Pyx_PyNumber_Int(x); if (!tmp) return (signed PY_LONG_LONG)-1; val = __Pyx_PyInt_AsSignedLongLong(tmp); Py_DECREF(tmp); return val; } } static int __Pyx_check_binary_version(void) { char ctversion[4], rtversion[4]; PyOS_snprintf(ctversion, 4, "%d.%d", PY_MAJOR_VERSION, PY_MINOR_VERSION); PyOS_snprintf(rtversion, 4, "%s", Py_GetVersion()); if (ctversion[0] != rtversion[0] || ctversion[2] != rtversion[2]) { char message[200]; PyOS_snprintf(message, sizeof(message), "compiletime version %s of module '%.100s' " "does not match runtime version %s", ctversion, __Pyx_MODULE_NAME, rtversion); #if PY_VERSION_HEX < 0x02050000 return PyErr_Warn(NULL, message); #else return PyErr_WarnEx(NULL, message, 1); #endif } return 0; } #ifndef __PYX_HAVE_RT_ImportType #define __PYX_HAVE_RT_ImportType static PyTypeObject *__Pyx_ImportType(const char *module_name, const char *class_name, size_t size, int strict) { PyObject *py_module = 0; PyObject *result = 0; PyObject *py_name = 0; char warning[200]; py_module = __Pyx_ImportModule(module_name); if (!py_module) goto bad; py_name = __Pyx_PyIdentifier_FromString(class_name); if (!py_name) goto bad; result = PyObject_GetAttr(py_module, py_name); Py_DECREF(py_name); py_name = 0; Py_DECREF(py_module); py_module = 0; if (!result) goto bad; if (!PyType_Check(result)) { PyErr_Format(PyExc_TypeError, "%s.%s is not a type object", module_name, class_name); goto bad; } if (!strict && ((PyTypeObject *)result)->tp_basicsize > (Py_ssize_t)size) { PyOS_snprintf(warning, sizeof(warning), "%s.%s size changed, may indicate binary incompatibility", module_name, class_name); #if PY_VERSION_HEX < 0x02050000 if (PyErr_Warn(NULL, warning) < 0) goto bad; #else if (PyErr_WarnEx(NULL, warning, 0) < 0) goto bad; #endif } else if (((PyTypeObject *)result)->tp_basicsize != (Py_ssize_t)size) { PyErr_Format(PyExc_ValueError, "%s.%s has the wrong size, try recompiling", module_name, class_name); goto bad; } return (PyTypeObject *)result; bad: Py_XDECREF(py_module); Py_XDECREF(result); return NULL; } #endif #ifndef __PYX_HAVE_RT_ImportModule #define __PYX_HAVE_RT_ImportModule static PyObject *__Pyx_ImportModule(const char *name) { PyObject *py_name = 0; PyObject *py_module = 0; py_name = __Pyx_PyIdentifier_FromString(name); if (!py_name) goto bad; py_module = PyImport_Import(py_name); Py_DECREF(py_name); return py_module; bad: Py_XDECREF(py_name); return 0; } #endif static void __pyx_clear_code_object_cache(void) { __Pyx_CodeObjectCacheEntry* entries = __pyx_code_cache.entries; int count = __pyx_code_cache.count; int i; if (entries == NULL) { return; } __pyx_code_cache.count = 0; __pyx_code_cache.max_count = 0; __pyx_code_cache.entries = NULL; for (i=0; i= 0 && code_line > entries[end].code_line) { return count; } while (start < end) { mid = (start + end) / 2; if (code_line < entries[mid].code_line) { end = mid; } else if (code_line > entries[mid].code_line) { start = mid + 1; } else { return mid; } } if (code_line <= entries[mid].code_line) { return mid; } else { return mid + 1; } } static PyCodeObject *__pyx_find_code_object(int code_line) { PyCodeObject* code_object; int pos; if (unlikely(!code_line) || unlikely(!__pyx_code_cache.entries)) { return NULL; } pos = __pyx_bisect_code_objects(__pyx_code_cache.entries, __pyx_code_cache.count, code_line); if (unlikely(pos >= __pyx_code_cache.count) || unlikely(__pyx_code_cache.entries[pos].code_line != code_line)) { return NULL; } code_object = __pyx_code_cache.entries[pos].code_object; Py_INCREF(code_object); return code_object; } static void __pyx_insert_code_object(int code_line, PyCodeObject* code_object) { int pos, i; __Pyx_CodeObjectCacheEntry* entries = __pyx_code_cache.entries; if (unlikely(!code_line)) { return; } if (unlikely(!entries)) { entries = (__Pyx_CodeObjectCacheEntry*)PyMem_Malloc(64*sizeof(__Pyx_CodeObjectCacheEntry)); if (likely(entries)) { __pyx_code_cache.entries = entries; __pyx_code_cache.max_count = 64; __pyx_code_cache.count = 1; entries[0].code_line = code_line; entries[0].code_object = code_object; Py_INCREF(code_object); } return; } pos = __pyx_bisect_code_objects(__pyx_code_cache.entries, __pyx_code_cache.count, code_line); if ((pos < __pyx_code_cache.count) && unlikely(__pyx_code_cache.entries[pos].code_line == code_line)) { PyCodeObject* tmp = entries[pos].code_object; entries[pos].code_object = code_object; Py_DECREF(tmp); return; } if (__pyx_code_cache.count == __pyx_code_cache.max_count) { int new_max = __pyx_code_cache.max_count + 64; entries = (__Pyx_CodeObjectCacheEntry*)PyMem_Realloc( __pyx_code_cache.entries, new_max*sizeof(__Pyx_CodeObjectCacheEntry)); if (unlikely(!entries)) { return; } __pyx_code_cache.entries = entries; __pyx_code_cache.max_count = new_max; } for (i=__pyx_code_cache.count; i>pos; i--) { entries[i] = entries[i-1]; } entries[pos].code_line = code_line; entries[pos].code_object = code_object; __pyx_code_cache.count++; Py_INCREF(code_object); } #include "compile.h" #include "frameobject.h" #include "traceback.h" static PyCodeObject* __Pyx_CreateCodeObjectForTraceback( const char *funcname, int __pyx_clineno, int __pyx_lineno, const char *__pyx_filename) { PyCodeObject *py_code = 0; PyObject *py_srcfile = 0; PyObject *py_funcname = 0; #if PY_MAJOR_VERSION < 3 py_srcfile = PyString_FromString(__pyx_filename); #else py_srcfile = PyUnicode_FromString(__pyx_filename); #endif if (!py_srcfile) goto bad; if (__pyx_clineno) { #if PY_MAJOR_VERSION < 3 py_funcname = PyString_FromFormat( "%s (%s:%d)", funcname, __pyx_cfilenm, __pyx_clineno); #else py_funcname = PyUnicode_FromFormat( "%s (%s:%d)", funcname, __pyx_cfilenm, __pyx_clineno); #endif } else { #if PY_MAJOR_VERSION < 3 py_funcname = PyString_FromString(funcname); #else py_funcname = PyUnicode_FromString(funcname); #endif } if (!py_funcname) goto bad; py_code = __Pyx_PyCode_New( 0, /*int argcount,*/ 0, /*int kwonlyargcount,*/ 0, /*int nlocals,*/ 0, /*int stacksize,*/ 0, /*int flags,*/ __pyx_empty_bytes, /*PyObject *code,*/ __pyx_empty_tuple, /*PyObject *consts,*/ __pyx_empty_tuple, /*PyObject *names,*/ __pyx_empty_tuple, /*PyObject *varnames,*/ __pyx_empty_tuple, /*PyObject *freevars,*/ __pyx_empty_tuple, /*PyObject *cellvars,*/ py_srcfile, /*PyObject *filename,*/ py_funcname, /*PyObject *name,*/ __pyx_lineno, /*int firstlineno,*/ __pyx_empty_bytes /*PyObject *lnotab*/ ); Py_DECREF(py_srcfile); Py_DECREF(py_funcname); return py_code; bad: Py_XDECREF(py_srcfile); Py_XDECREF(py_funcname); return NULL; } static void __Pyx_AddTraceback(const char *funcname, int __pyx_clineno, int __pyx_lineno, const char *__pyx_filename) { PyCodeObject *py_code = 0; PyObject *py_globals = 0; PyFrameObject *py_frame = 0; py_code = __pyx_find_code_object(__pyx_clineno ? __pyx_clineno : __pyx_lineno); if (!py_code) { py_code = __Pyx_CreateCodeObjectForTraceback( funcname, __pyx_clineno, __pyx_lineno, __pyx_filename); if (!py_code) goto bad; __pyx_insert_code_object(__pyx_clineno ? __pyx_clineno : __pyx_lineno, py_code); } py_globals = PyModule_GetDict(__pyx_m); if (!py_globals) goto bad; py_frame = PyFrame_New( PyThreadState_GET(), /*PyThreadState *tstate,*/ py_code, /*PyCodeObject *code,*/ py_globals, /*PyObject *globals,*/ 0 /*PyObject *locals*/ ); if (!py_frame) goto bad; py_frame->f_lineno = __pyx_lineno; PyTraceBack_Here(py_frame); bad: Py_XDECREF(py_code); Py_XDECREF(py_frame); } static int __Pyx_InitStrings(__Pyx_StringTabEntry *t) { while (t->p) { #if PY_MAJOR_VERSION < 3 if (t->is_unicode) { *t->p = PyUnicode_DecodeUTF8(t->s, t->n - 1, NULL); } else if (t->intern) { *t->p = PyString_InternFromString(t->s); } else { *t->p = PyString_FromStringAndSize(t->s, t->n - 1); } #else /* Python 3+ has unicode identifiers */ if (t->is_unicode | t->is_str) { if (t->intern) { *t->p = PyUnicode_InternFromString(t->s); } else if (t->encoding) { *t->p = PyUnicode_Decode(t->s, t->n - 1, t->encoding, NULL); } else { *t->p = PyUnicode_FromStringAndSize(t->s, t->n - 1); } } else { *t->p = PyBytes_FromStringAndSize(t->s, t->n - 1); } #endif if (!*t->p) return -1; ++t; } return 0; } /* Type Conversion Functions */ static CYTHON_INLINE int __Pyx_PyObject_IsTrue(PyObject* x) { int is_true = x == Py_True; if (is_true | (x == Py_False) | (x == Py_None)) return is_true; else return PyObject_IsTrue(x); } static CYTHON_INLINE PyObject* __Pyx_PyNumber_Int(PyObject* x) { PyNumberMethods *m; const char *name = NULL; PyObject *res = NULL; #if PY_VERSION_HEX < 0x03000000 if (PyInt_Check(x) || PyLong_Check(x)) #else if (PyLong_Check(x)) #endif return Py_INCREF(x), x; m = Py_TYPE(x)->tp_as_number; #if PY_VERSION_HEX < 0x03000000 if (m && m->nb_int) { name = "int"; res = PyNumber_Int(x); } else if (m && m->nb_long) { name = "long"; res = PyNumber_Long(x); } #else if (m && m->nb_int) { name = "int"; res = PyNumber_Long(x); } #endif if (res) { #if PY_VERSION_HEX < 0x03000000 if (!PyInt_Check(res) && !PyLong_Check(res)) { #else if (!PyLong_Check(res)) { #endif PyErr_Format(PyExc_TypeError, "__%s__ returned non-%s (type %.200s)", name, name, Py_TYPE(res)->tp_name); Py_DECREF(res); return NULL; } } else if (!PyErr_Occurred()) { PyErr_SetString(PyExc_TypeError, "an integer is required"); } return res; } static CYTHON_INLINE Py_ssize_t __Pyx_PyIndex_AsSsize_t(PyObject* b) { Py_ssize_t ival; PyObject* x = PyNumber_Index(b); if (!x) return -1; ival = PyInt_AsSsize_t(x); Py_DECREF(x); return ival; } static CYTHON_INLINE PyObject * __Pyx_PyInt_FromSize_t(size_t ival) { #if PY_VERSION_HEX < 0x02050000 if (ival <= LONG_MAX) return PyInt_FromLong((long)ival); else { unsigned char *bytes = (unsigned char *) &ival; int one = 1; int little = (int)*(unsigned char*)&one; return _PyLong_FromByteArray(bytes, sizeof(size_t), little, 0); } #else return PyInt_FromSize_t(ival); #endif } static CYTHON_INLINE size_t __Pyx_PyInt_AsSize_t(PyObject* x) { unsigned PY_LONG_LONG val = __Pyx_PyInt_AsUnsignedLongLong(x); if (unlikely(val == (unsigned PY_LONG_LONG)-1 && PyErr_Occurred())) { return (size_t)-1; } else if (unlikely(val != (unsigned PY_LONG_LONG)(size_t)val)) { PyErr_SetString(PyExc_OverflowError, "value too large to convert to size_t"); return (size_t)-1; } return (size_t)val; } #endif /* Py_PYTHON_H */ HDDM-0.2/src/wfpt_switch.pyx0000664000102500010250000002300311725664521016530 0ustar wieckiwiecki00000000000000#cython: embedsignature=True #cython: cdivision=True #cython: wraparound=False #cython: boundscheck=False # # Copyleft Thomas Wiecki (thomas_wiecki[at]brown.edu), 2012 # GPLv3 from copy import copy import numpy as np cimport numpy as np cimport cython from cython.parallel import * from cython_gsl cimport * include "pdf.pxi" ctypedef double * double_ptr ctypedef void * void_ptr from libc.math cimport * cdef extern from "stdlib.h": void free(void* ptr) void* malloc(size_t size) void* realloc(void* ptr, size_t size) cdef double wfpt_gsl(double x, void * params) nogil: cdef double rt, v, v_switch, a, z, t, t_switch, f, T rt = ( params)[0] v = ( params)[1] v_switch = ( params)[2] V_switch = ( params)[3] a = ( params)[4] z = ( params)[5] t = ( params)[6] t_switch = ( params)[7] T = ( params)[8] f = full_pdf(rt, v_switch, V_switch, a, x, 0, t+t_switch, 0, 1e-4) * calc_drift_dens_T(x*a, t_switch, v, a, z*a, T) return f cdef double calc_drift_dens_T(double x, double t, double v, double a, double z, double T) nogil: if T < 1e-4: return calc_drift_dens(x, t, v, a, z,False) else: return 1/T * (calc_drift_dens(x, t+T/2, v, a, z,True) - calc_drift_dens(x, t-T/2, v, a, z, True)) cpdef double calc_drift_dens(double x, double t, double v, double a, double z, bint integrate_t) nogil: cdef int N=40 cdef int n=0 cdef int got_zero = 0 cdef double term = 0 cdef double summed = 0 cdef double divisor # Compute sum minimum of 60 evals or until 5 zeroes have been encountered. while(got_zero < 10 and n < N): if not integrate_t: # Ratcliff 1980 Equation 12 term = sin(n*M_PI*z/a) * sin(n*M_PI*x/a) * exp(-.5*(v**2 + (n**2*M_PI**2)/a**2)*t) else: # Indefinite integral over t divisor = (-0.5*M_PI**2*n**2/a**2 - 0.5*v**2) #if divisor == 0: # print n, a, v term = sin(n*M_PI*z/a) * sin(n*M_PI*x/a) * (exp(-.5*(v**2 + (n**2*M_PI**2)/a**2)*t) / divisor) # Start counting after N terms if fabs(term) < 1e-6 and n > N: got_zero+=1 summed += term n+=1 #if term == -np.inf or term == +np.inf: #print x, t, v, a, z, n, term #return 0 return 2 * exp(v*(x-z)) * summed cdef double pdf_post_switch(double rt, double v, double v_switch, double V_switch, double a, double z, double t, double t_switch, double T, double err) nogil: cdef double alpha, result, error, expected cdef gsl_integration_workspace * W W = gsl_integration_workspace_alloc(1000) cdef gsl_function F cdef double params[9] cdef size_t neval params[0] = rt params[1] = v params[2] = v_switch params[3] = V_switch params[4] = a params[5] = z params[6] = t params[7] = t_switch params[8] = T F.function = &wfpt_gsl F.params = params gsl_integration_qag(&F, 0, 1, 1e-2, 1e-2, 1000, GSL_INTEG_GAUSS15, W, &result, &error) gsl_integration_workspace_free(W) return result cpdef double pdf_switch(double rt, double v, double v_switch, double V_switch, double a, double z, double t, double t_switch, double T, double err) nogil: cdef double p if fabs(rt) < t-T/2 or t < T/2 or t_switch < T/2 or t<0 or t_switch<0 or T<0 or a<=0 or z<=0 or z>=1 or T>.5: return 0 if (fabs(rt) <= t+t_switch): # Prosaccade or pre-switch p = full_pdf(rt, v, 0, a, z, 0, t, T, 1e-4, 2, 2, True, 1e-3) elif t_switch < 0.001: # Use regular likelihood p = full_pdf(rt, v_switch, 0, a, z, 0, t, T, 1e-4, 2, 2, True, 1e-3) else: # post-switch antisaccade p = pdf_post_switch(rt, v, v_switch, V_switch, a, z, t, t_switch, T, err) return p def wiener_like_antisaccade(np.ndarray[double, ndim=1] rt, double v, double v_switch, double V_switch, double a, double z, double t, double t_switch, double T, double err): cdef Py_ssize_t size = rt.shape[0] cdef Py_ssize_t i cdef double p cdef double sum_logp = 0 if np.any(np.abs(rt) < t-T/2) or t < T/2 or t_switch < T/2 or t<0 or t_switch<0 or T<0 or a<=0 or z<=0 or z>=1 or T>.5: return -np.inf for i in prange(size, nogil=True): p = pdf_switch(rt[i], v, v_switch, V_switch, a, z, t, t_switch, T, err) if p == 0: with gil: return -np.inf #if p < 0: # print rt[i], instruct[i], v, v_switch, V_switch, a, z, t, t_switch, T sum_logp += log(p) return sum_logp #################################################################### # Functions to compute antisaccade likelihood with precomputing the # drift density # Global variables for density and interpolation cdef double *drift_density cdef double *eval_dens cdef gsl_interp_accel *acc cdef gsl_spline *spline @cython.wraparound(False) @cython.boundscheck(False) # turn of bounds-checking for entire function def wiener_like_antisaccade_precomp(np.ndarray[double, ndim=1] rt, double v, double v_switch, double V_switch, double a, double z, double t, double t_switch, double T, double err, int evals=40, double t_switch_cutoff=.02): cdef Py_ssize_t size = rt.shape[0] cdef Py_ssize_t i cdef Py_ssize_t x cdef double p cdef double sum_logp = 0 if np.any(np.abs(rt) < t-T/2) or t < T/2 or t_switch == 0 or t_switch < T/2 or t<0 or t_switch<0 or T<0 or a<=0 or z<=0 or z>=1 or T>.5: return -np.inf ############################# # Precompute drift density # Initialize cubic spline gsl variables global drift_density, eval_dens, acc, spline drift_density = malloc(evals * sizeof(double)) eval_dens = malloc(evals * sizeof(double)) acc = gsl_interp_accel_alloc() spline = gsl_spline_alloc(gsl_interp_cspline, evals) # Compute density for x in range(evals): eval_dens[x] = a * (x/(evals-1)) #if x == evals-1: # eval_dens[x] = a - 1e-3 if t_switch < t_switch_cutoff: # If too small, approximate drift-density with normal distribution drift_density[x] = gsl_ran_gaussian_pdf(eval_dens[x] - (t_switch*v + z*a), sqrt(t_switch)) else: drift_density[x] = calc_drift_dens_T(eval_dens[x], t_switch, v, a, z*a, T) if np.isnan(drift_density[x]) or drift_density[x] < 0 or fabs(drift_density[x]) > 100: if (drift_density[x] < 0 and drift_density[x] > -1e-2) or x == 1: drift_density[x] = 0 else: print eval_dens[x], drift_density[x], x, t_switch, v, a, z*a, T # Ran into numerical stability issues, abort. free(drift_density) free(eval_dens) gsl_spline_free (spline) gsl_interp_accel_free (acc) return -np.inf # return 0 # Init spline gsl_spline_init(spline, eval_dens, drift_density, evals) for i in range(size): p = pdf_switch_precomp(rt[i], v, v_switch, V_switch, a, z, t, t_switch, T, err) if np.isnan(log(p)): print p, rt[i], v, v_switch, V_switch, a, z, t, t_switch, T, err if p == 0: sum_logp = -np.inf break sum_logp += log(p) free(drift_density) free(eval_dens) gsl_spline_free (spline) gsl_interp_accel_free (acc) return sum_logp cdef double pdf_switch_precomp(double rt, double v, double v_switch, double V_switch, double a, double z, double t, double t_switch, double T, double err) nogil: cdef double p if fabs(rt) < t-T/2 or t < T/2 or t_switch < T/2 or t<0 or t_switch<0 or T<0 or a<=0 or z<=0 or z>=1 or T>.5: return 0 if (fabs(rt) <= t+t_switch): # Prosaccade or pre-switch p = full_pdf(rt, v, 0, a, z, 0, t, T, 1e-4, 2, 2, True, 1e-3) #elif t_switch < 0.05: # Use regular likelihood #p = full_pdf(rt, v_switch, 0, a, z, 0, t+t_switch, T, 1e-4, 2, 2, True, 1e-3) else: # post-switch antisaccade p = pdf_post_switch_precomp(rt, v, v_switch, V_switch, a, z, t, t_switch, err) return p cdef double pdf_post_switch_precomp(double rt, double v, double v_switch, double V_switch, double a, double z, double t, double t_switch, double err) nogil: cdef double alpha, result, error, expected cdef gsl_integration_workspace * W W = gsl_integration_workspace_alloc(1000) cdef gsl_function F cdef double params[8] cdef size_t neval params[0] = rt params[1] = v params[2] = v_switch params[3] = V_switch params[4] = a params[5] = z params[6] = t params[7] = t_switch F.function = &wfpt_gsl_precomp F.params = params gsl_integration_qag(&F, 0, 1, 1e-3, 1e-3, 1000, GSL_INTEG_GAUSS15, W, &result, &error) gsl_integration_workspace_free(W) return result cdef double wfpt_gsl_precomp(double x, void * params) nogil: cdef double rt, v, v_switch, V_switch, a, z, t, t_switch, f, interp rt = ( params)[0] v = ( params)[1] v_switch = ( params)[2] V_switch = ( params)[3] a = ( params)[4] z = ( params)[5] t = ( params)[6] t_switch = ( params)[7] global acc, spline interp = gsl_spline_eval(spline, x*a, acc) if interp <= 0: return 0 f = full_pdf(rt, v_switch, V_switch, a, x, 0, t+t_switch, 0, 1e-4) * interp return f HDDM-0.2/src/integrate.pxi0000664000102500010250000002040311713025572016124 0ustar wieckiwiecki00000000000000#cython: embedsignature=True #cython: cdivision=True #cython: wraparound=False #cython: boundscheck=False import numpy as np cimport numpy as np cimport cython cdef double simpson_1D(double x, double v, double V, double a, double z, double t, double err, double lb_z, double ub_z, int nZ, double lb_t, double ub_t, int nT) nogil: #assert ((nZ&1)==0 and (nT&1)==0), "nT and nZ have to be even" #assert ((ub_t-lb_t)*(ub_z-lb_z)==0 and (nZ*nT)==0), "the function is defined for 1D-integration only" cdef double ht, hz cdef int n = max(nT,nZ) if nT==0: #integration over z hz = (ub_z-lb_z)/n ht = 0 lb_t = t ub_t = t else: #integration over t hz = 0 ht = (ub_t-lb_t)/n lb_z = z ub_z = z cdef double S = pdf_V(x - lb_t, v, V, a, lb_z, err) cdef double z_tag, t_tag, y cdef int i for i from 1 <= i <= n: z_tag = lb_z + hz * i t_tag = lb_t + ht * i y = pdf_V(x - t_tag, v, V, a, z_tag, err) if i&1: #check if i is odd S += (4 * y) else: S += (2 * y) S = S - y #the last term should be f(b) and not 2*f(b) so we subtract y S = S / ((ub_t-lb_t)+(ub_z-lb_z)) #the right function if pdf_V()/Z or pdf_V()/T return ((ht+hz) * S / 3) cdef double simpson_2D(double x, double v, double V, double a, double z, double t, double err, double lb_z, double ub_z, int nZ, double lb_t, double ub_t, int nT) nogil: #assert ((nZ&1)==0 and (nT&1)==0), "nT and nZ have to be even" #assert ((ub_t-lb_t)*(ub_z-lb_z)>0 and (nZ*nT)>0), "the function is defined for 2D-integration only, lb_t: %f, ub_t %f, lb_z %f, ub_z %f, nZ: %d, nT %d" % (lb_t, ub_t, lb_z, ub_z, nZ, nT) cdef double ht cdef double S cdef double t_tag, y cdef int i_t ht = (ub_t-lb_t)/nT S = simpson_1D(x, v, V, a, z, lb_t, err, lb_z, ub_z, nZ, 0, 0, 0) for i_t from 1 <= i_t <= nT: t_tag = lb_t + ht * i_t y = simpson_1D(x, v, V, a, z, t_tag, err, lb_z, ub_z, nZ, 0, 0, 0) if i_t&1: #check if i is odd S += (4 * y) else: S += (2 * y) S = S - y #the last term should be f(b) and not 2*f(b) so we subtract y S = S/ (ub_t-lb_t) return (ht * S / 3) cdef double adaptiveSimpsonsAux(double x, double v, double V, double a, double z, double t, double pdf_err, double lb_z, double ub_z, double lb_t, double ub_t, double ZT, double simps_err, double S, double f_beg, double f_end, double f_mid, int bottom) nogil: cdef double z_c, z_d, z_e, t_c, t_d, t_e, h cdef double fd, fe cdef double Sleft, Sright, S2 #print "in AdaptiveSimpsAux: lb_z: %f, ub_z: %f, lb_t %f, ub_t %f, f_beg: %f, f_end: %f, bottom: %d" % (lb_z, ub_z, lb_t, ub_t, f_beg, f_end, bottom) if (ub_t-lb_t) == 0: #integration over Z h = ub_z - lb_z z_c = (ub_z + lb_z)/2. z_d = (lb_z + z_c)/2. z_e = (z_c + ub_z)/2. t_c = t t_d = t t_e = t else: #integration over t h = ub_t - lb_t t_c = (ub_t + lb_t)/2. t_d = (lb_t + t_c)/2. t_e = (t_c + ub_t)/2. z_c = z z_d = z z_e = z fd = pdf_V(x - t_d, v, V, a, z_d, pdf_err)/ZT fe = pdf_V(x - t_e, v, V, a, z_e, pdf_err)/ZT Sleft = (h/12)*(f_beg + 4*fd + f_mid) Sright = (h/12)*(f_mid + 4*fe + f_end) S2 = Sleft + Sright if (bottom <= 0 or fabs(S2 - S) <= 15*simps_err): return S2 + (S2 - S)/15 return adaptiveSimpsonsAux(x, v, V, a, z, t, pdf_err, lb_z, z_c, lb_t, t_c, ZT, simps_err/2, Sleft, f_beg, f_mid, fd, bottom-1) + \ adaptiveSimpsonsAux(x, v, V, a, z, t, pdf_err, z_c, ub_z, t_c, ub_t, ZT, simps_err/2, Sright, f_mid, f_end, fe, bottom-1) cdef double adaptiveSimpsons_1D(double x, double v, double V, double a, double z, double t, double pdf_err, double lb_z, double ub_z, double lb_t, double ub_t, double simps_err, int maxRecursionDepth) nogil: cdef double h if (ub_t - lb_t) == 0: #integration over z lb_t = t ub_t = t h = ub_z - lb_z else: #integration over t h = (ub_t-lb_t) lb_z = z ub_z = z cdef double ZT = h cdef double c_t = (lb_t + ub_t)/2. cdef double c_z = (lb_z + ub_z)/2. cdef double f_beg, f_end, f_mid, S f_beg = pdf_V(x - lb_t, v, V, a, lb_z, pdf_err)/ZT f_end = pdf_V(x - ub_t, v, V, a, ub_z, pdf_err)/ZT f_mid = pdf_V(x - c_t, v, V, a, c_z, pdf_err)/ZT S = (h/6)*(f_beg + 4*f_mid + f_end) cdef double res = adaptiveSimpsonsAux(x, v, V, a, z, t, pdf_err, lb_z, ub_z, lb_t, ub_t, ZT, simps_err, S, f_beg, f_end, f_mid, maxRecursionDepth) return res cdef double adaptiveSimpsonsAux_2D(double x, double v, double V, double a, double z, double t, double pdf_err, double err_1d, double lb_z, double ub_z, double lb_t, double ub_t, double T, double err_2d, double S, double f_beg, double f_end, double f_mid, int maxRecursionDepth_Z, int bottom) nogil: cdef double fd, fe cdef double Sleft, Sright, S2 #print "in AdaptiveSimpsAux_2D: lb_z: %f, ub_z: %f, lb_t %f, ub_t %f, f_beg: %f, f_end: %f, bottom: %d" % (lb_z, ub_z, lb_t, ub_t, f_beg, f_end, bottom) cdef double t_c = (ub_t + lb_t)/2. cdef double t_d = (lb_t + t_c)/2. cdef double t_e = (t_c + ub_t)/2. cdef double h = ub_t - lb_t fd = adaptiveSimpsons_1D(x, v, V, a, z, t_d, pdf_err, lb_z, ub_z, 0, 0, err_1d, maxRecursionDepth_Z)/T fe = adaptiveSimpsons_1D(x, v, V, a, z, t_e, pdf_err, lb_z, ub_z, 0, 0, err_1d, maxRecursionDepth_Z)/T Sleft = (h/12)*(f_beg + 4*fd + f_mid) Sright = (h/12)*(f_mid + 4*fe + f_end) S2 = Sleft + Sright if (bottom <= 0 or fabs(S2 - S) <= 15*err_2d): return S2 + (S2 - S)/15; return adaptiveSimpsonsAux_2D(x, v, V, a, z, t, pdf_err, err_1d, lb_z, ub_z, lb_t, t_c, T, err_2d/2, Sleft, f_beg, f_mid, fd, maxRecursionDepth_Z, bottom-1) + \ adaptiveSimpsonsAux_2D(x, v, V, a, z, t, pdf_err, err_1d, lb_z, ub_z, t_c, ub_t, T, err_2d/2, Sright, f_mid, f_end, fe, maxRecursionDepth_Z, bottom-1) cdef double adaptiveSimpsons_2D(double x, double v, double V, double a, double z, double t, double pdf_err, double lb_z, double ub_z, double lb_t, double ub_t, double simps_err, int maxRecursionDepth_Z, int maxRecursionDepth_T) nogil: cdef double h = (ub_t-lb_t) cdef double T = (ub_t - lb_t) cdef double c_t = (lb_t + ub_t)/2. cdef double c_z = (lb_z + ub_z)/2. cdef double f_beg, f_end, f_mid, S cdef double err_1d = simps_err cdef double err_2d = simps_err f_beg = adaptiveSimpsons_1D(x, v, V, a, z, lb_t, pdf_err, lb_z, ub_z, 0, 0, err_1d, maxRecursionDepth_Z)/T f_end = adaptiveSimpsons_1D(x, v, V, a, z, ub_t, pdf_err, lb_z, ub_z, 0, 0, err_1d, maxRecursionDepth_Z)/T f_mid = adaptiveSimpsons_1D(x, v, V, a, z, (lb_t+ub_t)/2, pdf_err, lb_z, ub_z, 0, 0, err_1d, maxRecursionDepth_Z)/T S = (h/6)*(f_beg + 4*f_mid + f_end) cdef double res = adaptiveSimpsonsAux_2D(x, v, V, a, z, t, pdf_err, err_1d, lb_z, ub_z, lb_t, ub_t, T, err_2d, S, f_beg, f_end, f_mid, maxRecursionDepth_Z, maxRecursionDepth_T) return res HDDM-0.2/src/wfpt.pyx0000664000102500010250000001327511720307732015152 0ustar wieckiwiecki00000000000000#cython: embedsignature=True #cython: cdivision=True #cython: wraparound=False #cython: boundscheck=False # # Cython version of the Navarro & Fuss, 2009 DDM PDF. Based on the following code by Navarro & Fuss: # http://www.psychocmath.logy.adelaide.edu.au/personalpages/staff/danielnavarro/resources/wfpt.m # # This implementation is about 170 times faster than the matlab # reference version. # # Copyleft Thomas Wiecki (thomas_wiecki[at]brown.edu) & Imri Sofer, 2011 # GPLv3 import hddm from copy import copy import numpy as np cimport numpy as np cimport cython from cython.parallel import * #cimport openmp include "pdf.pxi" def pdf_array(np.ndarray[double, ndim=1] x, double v, double V, double a, double z, double Z, double t, double T, double err, bint logp=0, int nT=2, int nZ=2, bint use_adaptive=1, double simps_err=1e-3): cdef Py_ssize_t size = x.shape[0] cdef Py_ssize_t i cdef np.ndarray[double, ndim=1] y = np.empty(size, dtype=np.double) for i in prange(size, nogil=True): y[i] = full_pdf(x[i], v, V, a, z, Z, t, T, err, nT, nZ, use_adaptive, simps_err) if logp==1: return np.log(y) else: return y def wiener_like(np.ndarray[double, ndim=1] x, double v, double V, double a, double z, double Z, double t, double T, double err, int nT=10, int nZ=10, bint use_adaptive=1, double simps_err=1e-8): cdef Py_ssize_t size = x.shape[0] cdef Py_ssize_t i cdef double p cdef double sum_logp = 0 #if num_threads != 0: # openmp.omp_set_num_threads(num_threads) for i in prange(size, nogil=True, schedule='dynamic'): p = full_pdf(x[i], v, V, a, z, Z, t, T, err, nT, nZ, use_adaptive, simps_err) # If one probability = 0, the log sum will be -Inf if p == 0: with gil: return -np.inf sum_logp += log(p) return sum_logp def wiener_like_array(np.ndarray[double, ndim=1] x, double v, double V, double a, double z, double Z, double t, double T, double err, int nT=10, int nZ=10, bint use_adaptive=1, double simps_err=1e-8): cdef Py_ssize_t size = x.shape[0] cdef Py_ssize_t i cdef double sum_logp = 0 cdef np.ndarray[double, ndim=1] p_array = np.empty(size, dtype=np.float) #if num_threads != 0: # openmp.omp_set_num_threads(num_threads) for i in prange(size, nogil=True, schedule='dynamic'): p_array[i] = full_pdf(x[i], v, V, a, z, Z, t, T, err, nT, nZ, use_adaptive, simps_err) # If one probability = 0, the log sum will be -Inf if p_array[i] == 0: with gil: return -np.inf for i in prange(size, nogil=True, schedule='dynamic'): sum_logp += log(p_array[i]) return sum_logp def wiener_like_multi(np.ndarray[double, ndim=1] x, v, V, a, z, Z, t, T, double err, multi=None): cdef Py_ssize_t size = x.shape[0] cdef Py_ssize_t i cdef double p = 0 if multi is None: return full_pdf(x, v, V, a, z, Z, t, T, err) else: params = {'v':v, 'z':z, 't':t, 'a':a, 'V':V, 'Z':Z, 'T':T} params_iter = copy(params) for i from 0 <= i < size: for param in multi: params_iter[param] = params[param][i] p += log(full_pdf(x[i], params_iter['v'], params_iter['V'], params_iter['a'], params_iter['z'], params_iter['Z'], params_iter['t'], params_iter['T'], err)) return p def gen_rts_from_cdf(double v, double V, double a, double z, double Z, double t, \ double T, int samples=1000, double cdf_lb=-6, double cdf_ub=6, double dt=1e-2): cdef np.ndarray[double, ndim=1] x = np.arange(cdf_lb, cdf_ub, dt) cdef np.ndarray[double, ndim=1] l_cdf = np.empty(x.shape[0], dtype=np.double) cdef double pdf, rt cdef Py_ssize_t size = x.shape[0] cdef Py_ssize_t i, j cdef int idx l_cdf[0] = 0 for i in range(size): pdf = full_pdf(x[i], v, V, a, z, Z, 0, 0, 1e-4) l_cdf[i] = l_cdf[i-1] + pdf l_cdf /= l_cdf[x.shape[0]-1] cdef np.ndarray[double, ndim=1] rts = np.empty(samples, dtype=np.double) cdef np.ndarray[double, ndim=1] f = np.random.rand(samples) cdef np.ndarray[double, ndim=1] delay if T!=0: delay = (np.random.rand(samples)*T + (t - T/2.)) for i from 0 <= i < samples: idx = np.searchsorted(l_cdf, f[i]) rt = x[idx] if T==0: rt = rt + np.sign(rt)*t else: rt = rt + np.sign(rt)*delay[i] rts[i] = rt return rts def wiener_like_contaminant(np.ndarray[double, ndim=1] x, np.ndarray[int, ndim=1] cont_x, double v, \ double V, double a, double z, double Z, double t, double T, double t_min, \ double t_max, double err, int nT= 10, int nZ=10, bint use_adaptive=1, \ double simps_err=1e-8): """Wiener likelihood function where RTs could come from a separate, uniform contaminant distribution. Reference: Lee, Vandekerckhove, Navarro, & Tuernlinckx (2007) """ cdef Py_ssize_t size = x.shape[0] cdef Py_ssize_t i cdef double p cdef double sum_logp = 0 cdef int n_cont = np.sum(cont_x) cdef int pos_cont = 0 for i in prange(size, nogil=True): if cont_x[i] == 0: p = full_pdf(x[i], v, V, a, z, Z, t, T, err, nT, nZ, use_adaptive, simps_err) if p == 0: with gil: return -np.inf sum_logp += log(p) # If one probability = 0, the log sum will be -Inf # add the log likelihood of the contaminations sum_logp += n_cont*log(0.5 * 1./(t_max-t_min)) return sum_logp HDDM-0.2/setup.py0000664000102500010250000000243711727455600014356 0ustar wieckiwiecki00000000000000from distutils.core import setup from distutils.extension import Extension import numpy as np import os setup( name="HDDM", version="0.2", author="Thomas V. Wiecki, Imri Sofer, Michael J. Frank", author_email="thomas_wiecki@brown.edu", url="http://github.com/hddm-devs/hddm", packages=["hddm", "hddm.tests", "hddm.sandbox"], package_data={"hddm":["examples/*.csv", "examples/*.conf"]}, scripts=["scripts/hddm_fit.py", "scripts/hddm_demo.py"], description="HDDM is a python module that implements Hierarchical Bayesian parameter estimation of Drift Diffusion Models (via PyMC).", install_requires=['NumPy >=1.5.0', 'SciPy >= 0.6.0', 'kabuki >= 0.2', 'PyMC >= 2.0'], setup_requires=['NumPy >=1.5.0', 'SciPy >= 0.6.0', 'kabuki >= 0.2', 'PyMC >= 2.0'], include_dirs = [np.get_include()], classifiers=[ 'Development Status :: 4 - Beta', 'Environment :: Console', 'Operating System :: OS Independent', 'Intended Audience :: Science/Research', 'License :: OSI Approved :: GNU General Public License (GPL)', 'Programming Language :: Python', 'Topic :: Scientific/Engineering', ], ext_modules = [Extension("wfpt", ["src/wfpt.c"])] ) HDDM-0.2/scripts/0000775000102500010250000000000011727455647014340 5ustar wieckiwiecki00000000000000HDDM-0.2/scripts/hddm_demo.py0000775000102500010250000003007311725646100016620 0ustar wieckiwiecki00000000000000#!/usr/bin/env python """ brownian() implements one dimensional Brownian motion (i.e. the Wiener process). """ import wx import matplotlib # We want matplotlib to use a wxPython backend if __name__ == "__main__": matplotlib.use('WXAgg') import matplotlib.pyplot as plt from matplotlib.figure import Figure from traits.api import HasTraits, Instance, Range,\ Array, on_trait_change, Property,\ cached_property, Bool, Button, Tuple,\ Any, Int, Str, Float, Delegate, Dict from traitsui.view import View, Item import numpy as np import time #import enthought.traits.ui try: from IPython.Debugger import Tracer; debug_here = Tracer() except: pass import hddm def timer(method): def time_me(self, *args,**kwargs): t0 = time.clock() res = method(self,*args,**kwargs) print "%s took %f secs" % (method.__name__, time.clock() - t0) return res return time_me class DDM(HasTraits): """Drift diffusion model""" # Paremeters z = Range(0, 1., .5) sz = Range(0, 1., .0) v = Range(-40.,40.,-2.) sv = Range(0.0,2.,0.0) ter = Range(0,2.,.3) ster = Range(0,2.,.0) a = Range(0.,10.,2.) switch = Bool(False) t_switch = Range(0,2.,.3) v_switch = Range(-20.,20.,1.) intra_sv = Range(0.1,10.,1.) urgency = Range(.1,10.,1.) params = Property(Array, depends_on=['z', 'sz', 'v', 'sv', 'ter', 'ster', 'a', 'switch', 't_switch', 'v_switch', 'intra_sv']) # Distributions drifts = Property(Tuple, depends_on=['params']) rts = Property(Tuple, depends_on=['drifts']) params_dict = Property(Dict) histo = Property(Array, depends_on=['params']) # Total time. T = Float(5.0) # Time step size dt = Float(1e-4) # Number of realizations to generate. num_samples = Int(5000) # Number of drifts to plot iter_plot = Int(50) # Number of histogram bins bins = Int(200) view = View('z', 'sz', 'v', 'sv', 'ter', 'ster', 'a', 'num_samples', 'iter_plot', 'switch', 't_switch', 'v_switch', 'intra_sv', 'T') def _get_params_dict(self): d = {'v': self.v, 'V': self.sv, 'z': self.z, 'Z': self.sz, 't': self.ter, 'T': self.ster, 'a': self.a} if self.switch: d['v_switch'] = self.v_switch d['t_switch'] = self.t_switch d['V_switch'] = self.sv return d @cached_property def _get_drifts(self): return hddm.generate._gen_rts_from_simulated_drift(self.params_dict, samples=self.iter_plot, dt=self.dt, intra_sv=self.intra_sv)[1] @cached_property def _get_rts(self): if not self.switch: # Use faster cdf method return hddm.generate.gen_rts(self.params_dict, samples=self.num_samples, range_=(-self.T, self.T)) else: # Simulate individual drifts return hddm.generate._gen_rts_from_simulated_drift(self.params_dict, samples=self.num_samples, dt=self.dt, intra_sv=self.intra_sv)[0] #return hddm.generate.gen_rts(self.params_dict, samples=self.num_samples, range_=(-5, 5), method='drift') @cached_property def _get_histo(self): n, bins = hddm.utils.histogram(self.rts, bins=2 * self.bins, range=(-self.T, self.T)) db = np.array(np.diff(bins), float) return n / db / (n.sum()) def _get_params(self): return np.array([self.a, self.v, self.ter, self.sz, self.sv, self.ster]) class DDMPlot(HasTraits): from hddm.MPLTraits import MPLFigureEditor figure = Instance(Figure, ()) parameters = Instance(DDM, ()) plot_histogram = Bool(True) plot_simple = Bool(False) plot_wiener = Bool(False) plot_lba = Bool(False) plot_drifts = Bool(True) #plot_data = Bool(False) #plot_density = Bool(False) plot_true_density = Bool(False) #plot_density_dist = Bool(False) plot_mean_rt = Bool(False) plot_switch = Bool(True) color_errors = Bool(True) x_analytical = Property(Array) wiener = Property(Array) simple = Property(Array) lba = Property(Array) switch = Property(Array) x_raster = Int(100) data = Array() # Can be set externally external_params = Dict() go = Button('Go') view = View(Item('figure', editor=MPLFigureEditor(), show_label=False), Item('parameters'), Item('plot_histogram'), Item('plot_drifts'), Item('plot_wiener'), #Item('plot_lba'), #Item('plot_data'), #Item('plot_density'), #Item('plot_true_density'), #Item('plot_density_dist'), Item('plot_switch'), Item('color_errors'), Item('go'), #style='custom', width=800, height=600, resizable=True) def __init__(self, data=None, params=None): super(DDMPlot, self).__init__() plt.hot() # Create plot self.num_axes = 3 self.ax1 = self.figure.add_subplot(411) self.ax2 = self.figure.add_subplot(412, sharex=self.ax1) self.ax3 = self.figure.add_subplot(413, sharex=self.ax1) self.set_figure() #self.update_plot() # Set external data if necessary if data is not None: self.data = data if params is not None: self._set_external_params(params) def _data_changed(self): self.plot_data = True @cached_property def _get_x_analytical(self): return np.linspace(-self.parameters.T, self.parameters.T, self.x_raster) @timer def _get_switch(self): from hddm.sandbox.model import wfpt_switch_like pdf = wfpt_switch_like.rv.pdf(self.x_analytical, self.parameters.v, self.parameters.v_switch, self.parameters.sv, self.parameters.a, self.parameters.z, self.parameters.ter, self.parameters.t_switch, self.parameters.ster) return pdf @timer def _get_wiener(self): from hddm.likelihoods import wfpt_like pdf = wfpt_like.rv.pdf(self.x_analytical, self.parameters.v, self.parameters.sv, self.parameters.a, self.parameters.z, self.parameters.sz, self.parameters.ter, self.parameters.ster) return pdf @timer def _get_lba(self): return hddm.likelihoods.LBA_like(self.x_analytical, a=self.parameters.a, z=self.parameters.z, v0=self.parameters.v, v1=self.parameters.sz, t=self.parameters.ter, V=self.parameters.sv, logp=False) def _go_fired(self): self.update_plot() def plot_histo(self, x, y, color, max_perc=None): # y consists of lower and upper boundary responses # mid point tells us where to split assert y.shape[0] % 2 == 0, "x_analytical has to be even. Shape is %s " % str(y.shape) mid = y.shape[0] / 2 self.figure.axes[0].plot(x, y[mid:], color=color, lw=2.) # Compute correct EV mean_correct_rt = np.sum(x * y[mid:]) / np.sum(y[mid:]) # [::-1] -> reverse ordering if self.color_errors: color='r' self.figure.axes[2].plot(x, -y[:mid][::-1], color=color, lw=2.) # Compute error EV mean_error_rt = np.sum(x * y[:mid][::-1]) / np.sum(y[:mid][::-1]) if self.plot_mean_rt: self.figure.axes[0].axvline(mean_correct_rt, color=color) if self.color_errors: self.figure.axes[2].axvline(mean_error_rt, color='r') def plot_dens(self): t = np.linspace(0.0, self.parameters.T, self.x_raster) xm, ym = np.meshgrid(t, np.linspace(0, self.parameters.a, 100)) zm = np.zeros_like(xm) # np.zeros((t, 100), dtype=np.float) #print zs i = 0 for xs, ys in zip(xm, ym): j = 0 for x, y in zip(xs, ys): if x <= self.parameters.ter + .05: zm[i, j] = 1. else: zm[i, j] = hddm.wfpt_switch.drift_dens(y, x - self.parameters.ter, self.parameters.v, self.parameters.a, self.parameters.z * self.parameters.a) j += 1 self.figure.axes[1].contourf(xm, ym, zm) @on_trait_change('parameters.params') def update_plot(self): if self.figure.canvas is None: return # Clear all plots for i in range(self.num_axes): self.figure.axes[i].clear() # Set x axis values x = np.linspace(0, self.parameters.T, self.parameters.bins) # Set x axis values for analyticals x_anal = np.linspace(0, self.parameters.T, self.x_raster / 2) # Plot normalized histograms of simulated data if self.plot_histogram: self.plot_histo(x, self.parameters.histo, color='b', max_perc=.99) # Plot normalized histograms of empirical data #if self.plot_data: # histo = hddm.utils.histogram(self.data, bins=2*self.parameters.bins, range=(-self.parameters.T, self.parameters.T), dens=True)[0] # self.plot_histo(x, histo, color='y') # Plot analytical simple likelihood function if self.plot_switch and self.parameters.switch: self.plot_histo(x_anal, self.switch, color='r') if self.plot_lba: self.plot_histo(x_anal, self.lba, color='k') if self.plot_wiener: self.plot_histo(x_anal, self.wiener, color='y') if self.plot_true_density: self.plot_dens() if self.plot_drifts: t = np.linspace(0.0, self.parameters.T, self.parameters.T / 1e-4) drifts = self.parameters.drifts for drift in drifts: # Make sure drift is not longer than x-axis drift_len = len(drift) if len(drift) < len(t) else len(t) if self.color_errors: if drift[drift_len-1] < 0.1: color = 'r' else: color = 'b' else: color = 'b' self.figure.axes[1].plot(t[:drift_len], drift[:drift_len], color, alpha=.5) # Plot analytical simple likelihood function if self.parameters.switch: self.figure.axes[1].axvline(self.parameters.ter+self.parameters.t_switch, color='k', linestyle='--', lw=2) self.set_figure() wx.CallAfter(self.figure.canvas.draw) def set_figure(self): # Set axes limits # TODO: Fix boundary now that we are using true densities total_max = 1. if len(self.figure.axes[0].get_lines()) != 0: for line in self.figure.axes[0].get_lines(): max_point = np.max(line.get_ydata()) if max_point > total_max: total_max = max_point for line in self.figure.axes[2].get_lines(): max_point = np.abs(np.min(line.get_ydata())) if max_point > total_max: total_max = max_point self.figure.axes[0].set_ylim((0, total_max)) self.figure.axes[2].set_ylim((-total_max, 0)) self.figure.axes[0].set_xlim((0, self.parameters.T)) self.figure.axes[2].set_xlim((0, self.parameters.T)) self.figure.axes[1].set_ylim((0, self.parameters.a)) self.figure.subplots_adjust(hspace=0) for ax in self.ax1, self.ax2: plt.setp(ax.get_xticklabels(), visible=False) plt.setp(ax.get_yticklabels(), visible=False) plt.setp(self.ax3.get_yticklabels(), visible=False) self.ax3.set_xlabel('time (secs)') if __name__ == "__main__": ddm = DDMPlot() ddm.configure_traits() HDDM-0.2/scripts/hddm_fit.py0000775000102500010250000000404011726176127016461 0ustar wieckiwiecki00000000000000#!/usr/bin/env python import sys import hddm import os try: import argparse except: # fall back to version not using argparse if __name__=='__main__': if len(sys.argv) == 1 or sys.argv[1] == '-h' or sys.argv[1] == '--help': print "hddmfit model.cfg [data.csv]" sys.exit(-1) config_fname = sys.argv[1] data_fname = None if len(sys.argv) == 3: data_fname = sys.argv[2] hddm.utils.parse_config_file(config_fname, data=data_fname) sys.exit(0) if __name__=='__main__': parser = argparse.ArgumentParser(description='HDDM: Hierarchical estimation of drift diffusion models.', add_help=True, version='0.2a') parser.add_argument('--verbose', action="store_true", dest="verbose", help='more verbosity') parser.add_argument('-m', '--map', action="store_true", dest="map", help='find a starting point via MAP estimation') parser.add_argument('-s', '--samples', action="store", dest="samples", type=int, help='samples to generate') parser.add_argument('-b', '--burn', action="store", dest="burn", type=int, help='initial samples to discard as burn-in') parser.add_argument('-t', '--thin', action="store", dest="thin", type=int, help='keep every nth sample') parser.add_argument('-og', action="store_true", dest="only_group_stats", default=False, help='Output only group stats.') parser.add_argument('-np', action='store_true', dest='no_plots', default=False, help='Do not generate output plots.') parser.add_argument('model') parser.add_argument('data') results = parser.parse_args() hddm.utils.parse_config_file(results.model, data=results.data, samples=results.samples, burn=results.burn, thin=results.thin, map=results.map, only_group_stats=results.only_group_stats, plot=not(results.no_plots), verbose=results.verbose) sys.exit(0) HDDM-0.2/hddm/0000775000102500010250000000000011727455647013565 5ustar wieckiwiecki00000000000000HDDM-0.2/hddm/table_print.py0000664000102500010250000000717511713025572016435 0ustar wieckiwiecki00000000000000import cStringIO,operator import re import math def indent(rows, hasHeader=False, headerChar='-', delim=' | ', justify='left', separateRows=False, prefix='', postfix='', wrapfunc=lambda x:x): """Indents a table by column. - rows: A sequence of sequences of items, one sequence per row. - hasHeader: True if the first row consists of the columns' names. - headerChar: Character to be used for the row separator line (if hasHeader==True or separateRows==True). - delim: The column delimiter. - justify: Determines how are data justified in their column. Valid values are 'left','right' and 'center'. - separateRows: True if rows are to be separated by a line of 'headerChar's. - prefix: A string prepended to each printed row. - postfix: A string appended to each printed row. - wrapfunc: A function f(text) for wrapping text; each element in the table is first wrapped by this function. copied from http://code.activestate.com/recipes/267662/ """ # closure for breaking logical rows to physical, using wrapfunc def rowWrapper(row): newRows = [wrapfunc(item).split('\n') for item in row] return [[substr or '' for substr in item] for item in map(None,*newRows)] # break each logical row into one or more physical ones logicalRows = [rowWrapper(row) for row in rows] # columns of physical rows columns = map(None,*reduce(operator.add,logicalRows)) # get the maximum of each column by the string length of its items maxWidths = [max([len(str(item)) for item in column]) for column in columns] rowSeparator = headerChar * (len(prefix) + len(postfix) + sum(maxWidths) + \ len(delim)*(len(maxWidths)-1)) # select the appropriate justify method justify = {'center':str.center, 'right':str.rjust, 'left':str.ljust}[justify.lower()] output=cStringIO.StringIO() if separateRows: print >> output, rowSeparator for physicalRows in logicalRows: for row in physicalRows: print >> output, \ prefix \ + delim.join([justify(str(item),width) for (item,width) in zip(row,maxWidths)]) \ + postfix if separateRows or hasHeader: print >> output, rowSeparator; hasHeader=False return output.getvalue() def wrap_onspace(text, width): """ A word-wrap function that preserves existing line breaks and most spaces in the text. Expects that existing line breaks are posix newlines (\n). """ return reduce(lambda line, word, width=width: '%s%s%s' % (line, ' \n'[(len(line[line.rfind('\n')+1:]) + len(word.split('\n',1)[0] ) >= width)], word), text.split(' ') ) def wrap_onspace_strict(text, width): """Similar to wrap_onspace, but enforces the width constraint: words longer than width are split.""" wordRegex = re.compile(r'\S{'+str(width)+r',}') return wrap_onspace(wordRegex.sub(lambda m: wrap_always(m.group(),width),text),width) def wrap_always(text, width): """A simple word-wrap function that wraps text on exactly width characters. It doesn't split the text in words.""" return '\n'.join([ text[width*i:width*(i+1)] \ for i in xrange(int(math.ceil(1.*len(text)/width))) ]) data = \ '''John,Smith,24,Software Engineer Mary,Brohowski,23,Sales Manager Aristidis,Papageorgopoulos,28,Senior Reseacher''' rows = [row.strip().split(',') for row in data.splitlines()] HDDM-0.2/hddm/examples/0000775000102500010250000000000011727455647015403 5ustar wieckiwiecki00000000000000HDDM-0.2/hddm/examples/simple_subjs_difficulty.csv0000664000102500010250000006467011726176127023046 0ustar wieckiwiecki00000000000000response,rt,difficulty,subj_idx 1.0,1.929,easy,0 1.0,1.709,easy,0 0.0,0.917999999998,easy,0 1.0,0.559000000002,easy,0 1.0,1.132,easy,0 1.0,0.815000000002,easy,0 1.0,0.737000000002,easy,0 1.0,0.794000000002,easy,0 1.0,0.661000000002,easy,0 1.0,0.733000000002,easy,0 1.0,1.005,easy,0 1.0,0.836000000002,easy,0 0.0,0.612999999998,easy,0 1.0,1.325,easy,0 1.0,0.681000000002,easy,0 1.0,1.18,hard,0 1.0,2.631,hard,0 1.0,1.547,hard,0 1.0,1.561,hard,0 1.0,1.121,hard,0 1.0,1.462,hard,0 1.0,0.618000000002,hard,0 1.0,0.710000000002,hard,0 0.0,1.209,hard,0 1.0,0.486000000002,hard,0 1.0,0.830000000002,hard,0 1.0,0.822000000002,hard,0 1.0,1.606,hard,0 1.0,0.637000000002,hard,0 1.0,0.758000000002,hard,0 1.0,0.650000000002,easy,1 1.0,0.915000000002,easy,1 1.0,0.425000000002,easy,1 1.0,0.505000000002,easy,1 1.0,0.894000000002,easy,1 1.0,0.654000000002,easy,1 0.0,0.988999999998,easy,1 1.0,0.613000000002,easy,1 0.0,2.416,easy,1 1.0,1.077,easy,1 1.0,1.423,easy,1 1.0,0.772000000002,easy,1 1.0,0.916000000002,easy,1 0.0,1.879,easy,1 0.0,1.818,easy,1 1.0,0.480000000002,hard,1 1.0,0.661000000002,hard,1 1.0,1.275,hard,1 1.0,0.537000000002,hard,1 0.0,0.860999999998,hard,1 1.0,0.953000000002,hard,1 1.0,1.45,hard,1 1.0,2.347,hard,1 1.0,0.905000000002,hard,1 1.0,0.525000000002,hard,1 1.0,2.191,hard,1 1.0,0.668000000002,hard,1 1.0,0.743000000002,hard,1 1.0,0.640000000002,hard,1 1.0,0.667000000002,hard,1 1.0,0.501000000002,easy,2 0.0,0.620999999998,easy,2 1.0,0.614000000002,easy,2 1.0,1.179,easy,2 0.0,0.818999999998,easy,2 1.0,0.441000000002,easy,2 0.0,1.547,easy,2 0.0,1.92,easy,2 1.0,0.919000000002,easy,2 1.0,2.613,easy,2 1.0,1.019,easy,2 0.0,0.789999999998,easy,2 1.0,1.937,easy,2 1.0,0.710000000002,easy,2 1.0,2.316,easy,2 1.0,0.781000000002,hard,2 1.0,1.149,hard,2 1.0,1.442,hard,2 1.0,1.186,hard,2 1.0,0.743000000002,hard,2 1.0,0.556000000002,hard,2 1.0,0.555000000002,hard,2 1.0,1.262,hard,2 1.0,0.768000000002,hard,2 1.0,1.393,hard,2 1.0,1.774,hard,2 1.0,0.583000000002,hard,2 1.0,1.035,hard,2 1.0,1.937,hard,2 1.0,1.099,hard,2 1.0,1.706,easy,3 1.0,0.425000000002,easy,3 1.0,0.660000000002,easy,3 1.0,0.764000000002,easy,3 1.0,0.796000000002,easy,3 1.0,1.123,easy,3 0.0,0.517999999998,easy,3 1.0,1.218,easy,3 1.0,1.427,easy,3 1.0,2.575,easy,3 0.0,1.74,easy,3 1.0,0.861000000002,easy,3 1.0,1.285,easy,3 0.0,0.950999999998,easy,3 0.0,1.283,easy,3 1.0,1.656,hard,3 0.0,0.676999999998,hard,3 0.0,0.971999999998,hard,3 1.0,0.496000000002,hard,3 1.0,0.634000000002,hard,3 1.0,0.868000000002,hard,3 1.0,1.103,hard,3 1.0,0.582000000002,hard,3 1.0,0.941000000002,hard,3 0.0,0.434999999998,hard,3 1.0,0.735000000002,hard,3 1.0,1.117,hard,3 1.0,0.804000000002,hard,3 1.0,1.162,hard,3 1.0,0.600000000002,hard,3 1.0,1.265,easy,4 0.0,1.086,easy,4 1.0,0.491000000002,easy,4 0.0,1.198,easy,4 1.0,0.956000000002,easy,4 1.0,1.269,easy,4 1.0,1.192,easy,4 0.0,1.94,easy,4 0.0,0.680999999998,easy,4 1.0,0.719000000002,easy,4 1.0,1.009,easy,4 1.0,0.617000000002,easy,4 1.0,0.997000000002,easy,4 1.0,0.716000000002,easy,4 1.0,0.888000000002,easy,4 1.0,0.997000000002,hard,4 1.0,0.897000000002,hard,4 1.0,1.117,hard,4 1.0,0.664000000002,hard,4 1.0,1.903,hard,4 1.0,0.695000000002,hard,4 1.0,2.255,hard,4 1.0,0.989000000002,hard,4 1.0,1.38,hard,4 1.0,2.095,hard,4 1.0,1.256,hard,4 1.0,0.673000000002,hard,4 1.0,0.534000000002,hard,4 1.0,1.001,hard,4 1.0,0.439000000002,hard,4 1.0,1.042,easy,5 0.0,2.024,easy,5 1.0,0.589000000002,easy,5 1.0,1.393,easy,5 0.0,0.645999999998,easy,5 1.0,1.467,easy,5 1.0,1.945,easy,5 1.0,2.366,easy,5 1.0,0.954000000002,easy,5 0.0,2.691,easy,5 1.0,1.237,easy,5 1.0,0.413000000002,easy,5 0.0,0.761999999998,easy,5 1.0,1.011,easy,5 0.0,3.145,easy,5 1.0,1.42,hard,5 1.0,0.886000000002,hard,5 1.0,1.739,hard,5 1.0,0.627000000002,hard,5 1.0,1.829,hard,5 1.0,0.606000000002,hard,5 1.0,0.753000000002,hard,5 1.0,1.144,hard,5 1.0,0.630000000002,hard,5 1.0,0.612000000002,hard,5 1.0,1.331,hard,5 1.0,2.723,hard,5 1.0,0.616000000002,hard,5 1.0,3.61,hard,5 1.0,2.198,hard,5 0.0,1.298,easy,6 0.0,0.745999999998,easy,6 1.0,1.577,easy,6 0.0,0.873999999998,easy,6 0.0,1.72,easy,6 1.0,0.608000000002,easy,6 0.0,0.834999999998,easy,6 1.0,0.589000000002,easy,6 0.0,0.840999999998,easy,6 1.0,3.229,easy,6 1.0,0.682000000002,easy,6 1.0,1.501,easy,6 1.0,1.355,easy,6 1.0,0.806000000002,easy,6 1.0,4.593,easy,6 1.0,1.395,hard,6 1.0,0.639000000002,hard,6 1.0,0.548000000002,hard,6 1.0,0.603000000002,hard,6 1.0,2.141,hard,6 1.0,0.480000000002,hard,6 1.0,1.486,hard,6 0.0,0.534999999998,hard,6 1.0,0.548000000002,hard,6 1.0,0.598000000002,hard,6 1.0,1.413,hard,6 0.0,1.127,hard,6 0.0,1.067,hard,6 1.0,1.175,hard,6 1.0,1.553,hard,6 1.0,1.73,easy,7 1.0,0.711000000002,easy,7 1.0,0.753000000002,easy,7 1.0,0.637000000002,easy,7 1.0,0.510000000002,easy,7 1.0,1.063,easy,7 1.0,1.914,easy,7 1.0,1.075,easy,7 0.0,0.807999999998,easy,7 1.0,0.879000000002,easy,7 1.0,0.959000000002,easy,7 1.0,1.56,easy,7 1.0,1.622,easy,7 0.0,0.736999999998,easy,7 0.0,1.359,easy,7 1.0,0.728000000002,hard,7 1.0,0.833000000002,hard,7 1.0,2.302,hard,7 1.0,0.962000000002,hard,7 1.0,0.571000000002,hard,7 1.0,0.606000000002,hard,7 1.0,0.850000000002,hard,7 0.0,0.712999999998,hard,7 1.0,0.517000000002,hard,7 1.0,0.824000000002,hard,7 1.0,0.684000000002,hard,7 1.0,1.538,hard,7 1.0,1.113,hard,7 1.0,1.21,hard,7 1.0,3.112,hard,7 1.0,1.143,easy,8 1.0,0.941000000002,easy,8 1.0,0.490000000002,easy,8 1.0,0.567000000002,easy,8 0.0,0.808999999998,easy,8 1.0,0.532000000002,easy,8 1.0,0.720000000002,easy,8 1.0,0.930000000002,easy,8 0.0,0.490999999998,easy,8 1.0,0.623000000002,easy,8 0.0,0.731999999998,easy,8 1.0,2.149,easy,8 0.0,1.728,easy,8 1.0,1.794,easy,8 1.0,2.254,easy,8 1.0,1.971,hard,8 1.0,0.647000000002,hard,8 1.0,0.607000000002,hard,8 1.0,1.218,hard,8 0.0,0.593999999998,hard,8 1.0,0.886000000002,hard,8 1.0,0.905000000002,hard,8 1.0,1.525,hard,8 1.0,1.346,hard,8 1.0,0.531000000002,hard,8 1.0,0.716000000002,hard,8 1.0,0.542000000002,hard,8 1.0,1.781,hard,8 1.0,0.531000000002,hard,8 1.0,0.803000000002,hard,8 1.0,0.647000000002,easy,9 1.0,0.890000000002,easy,9 1.0,0.491000000002,easy,9 1.0,0.517000000002,easy,9 1.0,1.414,easy,9 1.0,1.819,easy,9 1.0,0.804000000002,easy,9 1.0,0.638000000002,easy,9 0.0,2.661,easy,9 1.0,0.672000000002,easy,9 0.0,0.595999999998,easy,9 1.0,2.143,easy,9 1.0,0.621000000002,easy,9 1.0,0.604000000002,easy,9 1.0,1.307,easy,9 1.0,1.245,hard,9 1.0,1.665,hard,9 1.0,0.804000000002,hard,9 1.0,2.741,hard,9 1.0,0.432000000002,hard,9 1.0,1.066,hard,9 0.0,1.169,hard,9 1.0,1.879,hard,9 0.0,0.805999999998,hard,9 1.0,0.787000000002,hard,9 1.0,0.483000000002,hard,9 1.0,0.384000000002,hard,9 1.0,0.709000000002,hard,9 1.0,1.365,hard,9 1.0,0.819000000002,hard,9 0.0,0.737999999998,easy,10 1.0,3.151,easy,10 1.0,2.237,easy,10 1.0,1.36,easy,10 0.0,2.569,easy,10 1.0,1.231,easy,10 1.0,0.463000000002,easy,10 1.0,1.274,easy,10 0.0,0.599999999998,easy,10 1.0,1.191,easy,10 0.0,2.339,easy,10 0.0,0.784999999998,easy,10 1.0,1.911,easy,10 0.0,0.745999999998,easy,10 0.0,2.293,easy,10 0.0,1.076,hard,10 1.0,0.809000000002,hard,10 1.0,1.956,hard,10 1.0,1.086,hard,10 1.0,0.719000000002,hard,10 1.0,0.959000000002,hard,10 1.0,0.969000000002,hard,10 1.0,1.127,hard,10 1.0,1.273,hard,10 0.0,0.670999999998,hard,10 1.0,0.625000000002,hard,10 1.0,0.603000000002,hard,10 1.0,0.889000000002,hard,10 1.0,0.685000000002,hard,10 0.0,1.032,hard,10 0.0,0.805999999998,easy,11 0.0,0.589999999998,easy,11 0.0,1.181,easy,11 1.0,0.750000000002,easy,11 1.0,2.041,easy,11 1.0,0.509000000002,easy,11 1.0,1.606,easy,11 1.0,0.812000000002,easy,11 1.0,0.643000000002,easy,11 1.0,1.751,easy,11 1.0,0.641000000002,easy,11 1.0,1.922,easy,11 1.0,1.776,easy,11 0.0,1.892,easy,11 0.0,0.853999999998,easy,11 1.0,0.900000000002,hard,11 1.0,1.984,hard,11 1.0,0.710000000002,hard,11 1.0,1.356,hard,11 1.0,0.559000000002,hard,11 1.0,1.252,hard,11 1.0,0.885000000002,hard,11 1.0,1.21,hard,11 1.0,1.414,hard,11 1.0,0.768000000002,hard,11 1.0,0.773000000002,hard,11 1.0,0.993000000002,hard,11 1.0,1.302,hard,11 1.0,1.453,hard,11 1.0,0.924000000002,hard,11 0.0,1.028,easy,12 1.0,0.722000000002,easy,12 0.0,0.691999999998,easy,12 1.0,1.267,easy,12 0.0,0.713999999998,easy,12 1.0,0.783000000002,easy,12 1.0,1.501,easy,12 1.0,1.872,easy,12 0.0,0.607999999998,easy,12 0.0,0.736999999998,easy,12 1.0,0.770000000002,easy,12 1.0,1.817,easy,12 1.0,2.469,easy,12 1.0,1.223,easy,12 1.0,2.462,easy,12 1.0,0.597000000002,hard,12 1.0,1.489,hard,12 1.0,2.76,hard,12 0.0,0.616999999998,hard,12 1.0,0.465000000002,hard,12 1.0,0.768000000002,hard,12 1.0,0.640000000002,hard,12 1.0,2.047,hard,12 0.0,1.087,hard,12 1.0,1.763,hard,12 1.0,0.695000000002,hard,12 1.0,0.945000000002,hard,12 1.0,0.636000000002,hard,12 1.0,0.474000000002,hard,12 1.0,1.677,hard,12 0.0,1.125,easy,13 1.0,0.828000000002,easy,13 1.0,0.971000000002,easy,13 1.0,0.921000000002,easy,13 1.0,1.35,easy,13 1.0,0.975000000002,easy,13 1.0,0.638000000002,easy,13 1.0,0.698000000002,easy,13 1.0,1.384,easy,13 1.0,0.967000000002,easy,13 0.0,0.875999999998,easy,13 1.0,1.426,easy,13 1.0,1.92,easy,13 1.0,1.221,easy,13 0.0,2.049,easy,13 1.0,0.510000000002,hard,13 1.0,1.081,hard,13 0.0,0.629999999998,hard,13 1.0,0.680000000002,hard,13 0.0,0.753999999998,hard,13 0.0,1.234,hard,13 1.0,1.02,hard,13 1.0,0.982000000002,hard,13 1.0,0.555000000002,hard,13 1.0,2.331,hard,13 1.0,0.562000000002,hard,13 1.0,0.931000000002,hard,13 1.0,1.187,hard,13 1.0,1.168,hard,13 1.0,0.691000000002,hard,13 1.0,0.648000000002,easy,14 1.0,1.066,easy,14 1.0,3.594,easy,14 1.0,0.778000000002,easy,14 1.0,1.319,easy,14 0.0,1.755,easy,14 1.0,1.207,easy,14 1.0,0.584000000002,easy,14 0.0,0.874999999998,easy,14 0.0,1.004,easy,14 0.0,0.567999999998,easy,14 1.0,1.224,easy,14 0.0,0.965999999998,easy,14 1.0,2.981,easy,14 1.0,1.275,easy,14 1.0,0.535000000002,hard,14 1.0,0.420000000002,hard,14 0.0,1.722,hard,14 1.0,1.315,hard,14 1.0,0.463000000002,hard,14 1.0,1.401,hard,14 1.0,0.779000000002,hard,14 1.0,0.882000000002,hard,14 1.0,0.786000000002,hard,14 0.0,0.645999999998,hard,14 1.0,1.032,hard,14 1.0,0.736000000002,hard,14 1.0,1.312,hard,14 1.0,1.902,hard,14 0.0,0.617999999998,hard,14 1.0,1.085,easy,15 1.0,0.889000000002,easy,15 1.0,0.675000000002,easy,15 1.0,0.659000000002,easy,15 0.0,1.096,easy,15 1.0,1.07,easy,15 0.0,0.893999999998,easy,15 0.0,0.517999999998,easy,15 1.0,0.968000000002,easy,15 0.0,1.604,easy,15 0.0,2.38,easy,15 1.0,0.997000000002,easy,15 0.0,0.826999999998,easy,15 1.0,0.602000000002,easy,15 1.0,0.992000000002,easy,15 1.0,0.765000000002,hard,15 1.0,1.242,hard,15 1.0,1.538,hard,15 1.0,0.623000000002,hard,15 0.0,1.22,hard,15 1.0,1.209,hard,15 1.0,0.830000000002,hard,15 1.0,0.904000000002,hard,15 1.0,0.502000000002,hard,15 0.0,1.1,hard,15 1.0,1.174,hard,15 0.0,0.409999999998,hard,15 1.0,0.812000000002,hard,15 1.0,0.569000000002,hard,15 1.0,0.897000000002,hard,15 1.0,0.915000000002,easy,16 0.0,0.989999999998,easy,16 0.0,2.629,easy,16 1.0,0.686000000002,easy,16 1.0,2.01,easy,16 1.0,1.192,easy,16 1.0,1.493,easy,16 1.0,1.001,easy,16 1.0,0.751000000002,easy,16 1.0,2.83,easy,16 1.0,0.925000000002,easy,16 1.0,1.302,easy,16 0.0,3.572,easy,16 0.0,0.547999999998,easy,16 0.0,1.885,easy,16 1.0,0.554000000002,hard,16 0.0,0.815999999998,hard,16 0.0,1.751,hard,16 1.0,0.809000000002,hard,16 1.0,0.455000000002,hard,16 1.0,1.923,hard,16 0.0,0.573999999998,hard,16 1.0,0.464000000002,hard,16 1.0,0.679000000002,hard,16 1.0,1.026,hard,16 1.0,0.567000000002,hard,16 1.0,1.906,hard,16 1.0,1.092,hard,16 1.0,0.662000000002,hard,16 0.0,1.464,hard,16 1.0,0.510000000002,easy,17 0.0,0.647999999998,easy,17 1.0,1.454,easy,17 1.0,2.544,easy,17 1.0,1.659,easy,17 1.0,1.317,easy,17 1.0,0.679000000002,easy,17 1.0,2.143,easy,17 1.0,0.845000000002,easy,17 0.0,1.32,easy,17 1.0,0.806000000002,easy,17 0.0,0.915999999998,easy,17 1.0,2.036,easy,17 1.0,0.652000000002,easy,17 1.0,1.022,easy,17 1.0,0.520000000002,hard,17 1.0,0.625000000002,hard,17 1.0,0.610000000002,hard,17 1.0,0.529000000002,hard,17 1.0,1.754,hard,17 0.0,5.254,hard,17 1.0,0.658000000002,hard,17 1.0,0.815000000002,hard,17 1.0,1.351,hard,17 1.0,0.993000000002,hard,17 1.0,1.308,hard,17 1.0,0.528000000002,hard,17 1.0,0.661000000002,hard,17 1.0,1.11,hard,17 1.0,0.779000000002,hard,17 1.0,1.273,easy,18 0.0,0.874999999998,easy,18 1.0,0.708000000002,easy,18 0.0,1.455,easy,18 0.0,3.374,easy,18 1.0,1.257,easy,18 1.0,0.989000000002,easy,18 0.0,0.946999999998,easy,18 1.0,2.648,easy,18 1.0,0.594000000002,easy,18 1.0,3.25,easy,18 1.0,0.572000000002,easy,18 1.0,0.899000000002,easy,18 0.0,1.642,easy,18 1.0,1.762,easy,18 1.0,0.485000000002,hard,18 1.0,0.875000000002,hard,18 1.0,0.449000000002,hard,18 1.0,0.711000000002,hard,18 1.0,0.627000000002,hard,18 0.0,0.755999999998,hard,18 1.0,0.625000000002,hard,18 1.0,1.973,hard,18 1.0,1.274,hard,18 1.0,0.597000000002,hard,18 1.0,2.016,hard,18 1.0,1.593,hard,18 1.0,0.894000000002,hard,18 1.0,0.637000000002,hard,18 1.0,2.203,hard,18 1.0,1.091,easy,19 1.0,0.575000000002,easy,19 1.0,1.332,easy,19 1.0,1.302,easy,19 0.0,2.734,easy,19 1.0,1.678,easy,19 0.0,1.361,easy,19 1.0,0.896000000002,easy,19 1.0,0.810000000002,easy,19 1.0,1.706,easy,19 0.0,2.464,easy,19 1.0,1.577,easy,19 1.0,3.081,easy,19 1.0,0.774000000002,easy,19 1.0,1.291,easy,19 0.0,0.822999999998,hard,19 0.0,0.654999999998,hard,19 0.0,0.451999999998,hard,19 1.0,0.856000000002,hard,19 1.0,0.952000000002,hard,19 1.0,2.852,hard,19 1.0,1.15,hard,19 1.0,1.377,hard,19 1.0,0.899000000002,hard,19 1.0,0.715000000002,hard,19 1.0,0.644000000002,hard,19 0.0,1.016,hard,19 0.0,1.164,hard,19 1.0,0.928000000002,hard,19 1.0,0.701000000002,hard,19 1.0,1.443,easy,20 1.0,2.136,easy,20 1.0,3.172,easy,20 1.0,2.659,easy,20 1.0,1.208,easy,20 1.0,0.933000000002,easy,20 0.0,1.855,easy,20 0.0,0.798999999998,easy,20 1.0,0.957000000002,easy,20 1.0,0.728000000002,easy,20 1.0,4.398,easy,20 1.0,1.237,easy,20 1.0,1.328,easy,20 1.0,0.738000000002,easy,20 1.0,2.476,easy,20 1.0,0.656000000002,hard,20 1.0,1.111,hard,20 1.0,1.191,hard,20 1.0,1.152,hard,20 1.0,0.781000000002,hard,20 1.0,2.422,hard,20 1.0,1.614,hard,20 1.0,1.188,hard,20 1.0,1.137,hard,20 1.0,0.563000000002,hard,20 1.0,1.398,hard,20 1.0,1.771,hard,20 1.0,0.717000000002,hard,20 1.0,0.753000000002,hard,20 1.0,1.409,hard,20 1.0,0.993000000002,easy,21 1.0,0.984000000002,easy,21 0.0,2.231,easy,21 1.0,1.802,easy,21 1.0,1.007,easy,21 1.0,0.734000000002,easy,21 1.0,2.006,easy,21 1.0,0.689000000002,easy,21 1.0,0.601000000002,easy,21 1.0,1.562,easy,21 1.0,0.456000000002,easy,21 1.0,2.114,easy,21 0.0,2.265,easy,21 0.0,0.782999999998,easy,21 1.0,0.522000000002,easy,21 1.0,0.495000000002,hard,21 1.0,1.325,hard,21 1.0,0.579000000002,hard,21 1.0,0.429000000002,hard,21 1.0,1.384,hard,21 0.0,0.651999999998,hard,21 1.0,0.943000000002,hard,21 1.0,1.069,hard,21 1.0,0.779000000002,hard,21 1.0,0.703000000002,hard,21 1.0,0.866000000002,hard,21 1.0,1.748,hard,21 1.0,0.595000000002,hard,21 1.0,0.735000000002,hard,21 1.0,0.475000000002,hard,21 1.0,1.181,easy,22 1.0,0.923000000002,easy,22 1.0,0.585000000002,easy,22 0.0,1.238,easy,22 0.0,1.585,easy,22 1.0,2.754,easy,22 1.0,0.747000000002,easy,22 1.0,1.744,easy,22 1.0,0.780000000002,easy,22 1.0,2.278,easy,22 0.0,0.612999999998,easy,22 1.0,1.346,easy,22 0.0,1.817,easy,22 1.0,0.845000000002,easy,22 1.0,0.877000000002,easy,22 0.0,1.009,hard,22 1.0,1.232,hard,22 0.0,0.882999999998,hard,22 1.0,0.733000000002,hard,22 1.0,4.287,hard,22 1.0,0.683000000002,hard,22 1.0,3.147,hard,22 1.0,0.425000000002,hard,22 1.0,0.563000000002,hard,22 0.0,0.650999999998,hard,22 0.0,1.155,hard,22 1.0,0.759000000002,hard,22 1.0,0.999000000002,hard,22 1.0,0.554000000002,hard,22 1.0,0.841000000002,hard,22 1.0,0.827000000002,easy,23 1.0,0.994000000002,easy,23 1.0,0.439000000002,easy,23 1.0,2.235,easy,23 1.0,0.772000000002,easy,23 0.0,2.432,easy,23 1.0,0.966000000002,easy,23 0.0,0.972999999998,easy,23 1.0,1.105,easy,23 0.0,1.47,easy,23 1.0,0.923000000002,easy,23 0.0,0.883999999998,easy,23 0.0,1.171,easy,23 1.0,1.758,easy,23 0.0,0.617999999998,easy,23 1.0,1.034,hard,23 1.0,0.877000000002,hard,23 1.0,0.914000000002,hard,23 1.0,0.816000000002,hard,23 1.0,1.326,hard,23 1.0,1.399,hard,23 1.0,0.797000000002,hard,23 0.0,1.536,hard,23 1.0,2.966,hard,23 1.0,0.841000000002,hard,23 0.0,0.602999999998,hard,23 1.0,0.625000000002,hard,23 1.0,0.624000000002,hard,23 1.0,0.844000000002,hard,23 1.0,1.561,hard,23 0.0,0.561999999998,easy,24 0.0,0.669999999998,easy,24 0.0,1.285,easy,24 1.0,0.893000000002,easy,24 1.0,1.147,easy,24 0.0,3.599,easy,24 0.0,1.625,easy,24 1.0,0.554000000002,easy,24 1.0,0.983000000002,easy,24 0.0,0.533999999998,easy,24 1.0,0.843000000002,easy,24 1.0,0.457000000002,easy,24 0.0,0.638999999998,easy,24 0.0,0.647999999998,easy,24 0.0,2.903,easy,24 1.0,0.870000000002,hard,24 1.0,0.636000000002,hard,24 1.0,2.555,hard,24 1.0,0.475000000002,hard,24 1.0,0.928000000002,hard,24 1.0,0.735000000002,hard,24 1.0,3.81,hard,24 1.0,1.848,hard,24 1.0,1.383,hard,24 1.0,2.225,hard,24 1.0,0.779000000002,hard,24 1.0,0.456000000002,hard,24 1.0,0.975000000002,hard,24 1.0,1.244,hard,24 1.0,0.505000000002,hard,24 0.0,1.274,easy,25 1.0,1.76,easy,25 1.0,0.809000000002,easy,25 1.0,0.672000000002,easy,25 0.0,0.858999999998,easy,25 1.0,2.405,easy,25 1.0,0.908000000002,easy,25 1.0,0.833000000002,easy,25 1.0,0.765000000002,easy,25 1.0,0.457000000002,easy,25 1.0,2.19,easy,25 1.0,0.492000000002,easy,25 1.0,0.674000000002,easy,25 1.0,2.005,easy,25 1.0,1.429,easy,25 1.0,1.03,hard,25 1.0,1.863,hard,25 1.0,0.873000000002,hard,25 1.0,1.316,hard,25 1.0,0.882000000002,hard,25 1.0,0.595000000002,hard,25 1.0,0.453000000002,hard,25 1.0,1.258,hard,25 1.0,1.066,hard,25 1.0,0.790000000002,hard,25 1.0,3.038,hard,25 1.0,3.006,hard,25 1.0,1.197,hard,25 1.0,0.739000000002,hard,25 0.0,1.055,hard,25 1.0,1.845,easy,26 1.0,1.606,easy,26 1.0,0.731000000002,easy,26 0.0,2.069,easy,26 1.0,0.698000000002,easy,26 1.0,0.687000000002,easy,26 1.0,1.142,easy,26 0.0,1.069,easy,26 0.0,0.557999999998,easy,26 1.0,3.259,easy,26 0.0,1.225,easy,26 1.0,0.517000000002,easy,26 1.0,0.763000000002,easy,26 0.0,0.652999999998,easy,26 1.0,0.582000000002,easy,26 1.0,1.152,hard,26 1.0,1.611,hard,26 1.0,1.293,hard,26 1.0,1.525,hard,26 1.0,1.141,hard,26 1.0,0.801000000002,hard,26 1.0,1.483,hard,26 1.0,0.521000000002,hard,26 1.0,0.755000000002,hard,26 1.0,0.979000000002,hard,26 0.0,1.055,hard,26 1.0,1.285,hard,26 1.0,0.706000000002,hard,26 1.0,2.032,hard,26 1.0,0.722000000002,hard,26 1.0,0.615000000002,easy,27 1.0,0.877000000002,easy,27 0.0,1.754,easy,27 1.0,0.453000000002,easy,27 1.0,2.256,easy,27 1.0,0.985000000002,easy,27 1.0,2.245,easy,27 0.0,0.594999999998,easy,27 1.0,0.822000000002,easy,27 1.0,1.048,easy,27 0.0,1.381,easy,27 1.0,2.367,easy,27 1.0,0.924000000002,easy,27 1.0,0.644000000002,easy,27 1.0,2.323,easy,27 1.0,0.650000000002,hard,27 1.0,1.699,hard,27 1.0,0.467000000002,hard,27 1.0,0.686000000002,hard,27 1.0,0.763000000002,hard,27 1.0,0.489000000002,hard,27 1.0,1.228,hard,27 1.0,0.745000000002,hard,27 1.0,1.297,hard,27 1.0,1.272,hard,27 1.0,1.494,hard,27 1.0,0.777000000002,hard,27 1.0,0.438000000002,hard,27 0.0,0.654999999998,hard,27 1.0,0.429000000002,hard,27 1.0,1.95,easy,28 1.0,1.184,easy,28 1.0,1.477,easy,28 0.0,2.345,easy,28 0.0,0.949999999998,easy,28 0.0,0.546999999998,easy,28 1.0,1.764,easy,28 0.0,0.748999999998,easy,28 1.0,1.09,easy,28 0.0,1.393,easy,28 0.0,1.069,easy,28 1.0,0.696000000002,easy,28 1.0,1.422,easy,28 1.0,0.451000000002,easy,28 0.0,1.751,easy,28 1.0,0.474000000002,hard,28 1.0,0.874000000002,hard,28 0.0,2.442,hard,28 1.0,0.643000000002,hard,28 1.0,1.424,hard,28 1.0,1.283,hard,28 1.0,1.71,hard,28 1.0,1.465,hard,28 1.0,1.076,hard,28 1.0,0.907000000002,hard,28 1.0,1.567,hard,28 1.0,3.206,hard,28 1.0,0.709000000002,hard,28 1.0,1.061,hard,28 1.0,0.847000000002,hard,28 0.0,1.307,easy,29 1.0,0.468000000002,easy,29 1.0,1.205,easy,29 1.0,0.589000000002,easy,29 1.0,0.689000000002,easy,29 1.0,1.582,easy,29 1.0,0.620000000002,easy,29 1.0,2.019,easy,29 0.0,1.864,easy,29 1.0,2.421,easy,29 0.0,0.967999999998,easy,29 1.0,0.669000000002,easy,29 1.0,0.950000000002,easy,29 0.0,0.547999999998,easy,29 0.0,0.379999999998,easy,29 1.0,0.683000000002,hard,29 1.0,0.659000000002,hard,29 1.0,1.115,hard,29 1.0,1.834,hard,29 0.0,1.407,hard,29 1.0,0.762000000002,hard,29 1.0,1.978,hard,29 1.0,0.829000000002,hard,29 1.0,0.882000000002,hard,29 1.0,0.784000000002,hard,29 1.0,0.894000000002,hard,29 1.0,0.991000000002,hard,29 0.0,0.443999999998,hard,29 1.0,1.644,hard,29 1.0,0.655000000002,hard,29 1.0,0.779000000002,easy,30 1.0,0.824000000002,easy,30 1.0,1.277,easy,30 0.0,1.124,easy,30 1.0,0.660000000002,easy,30 0.0,1.234,easy,30 0.0,2.407,easy,30 1.0,1.271,easy,30 0.0,0.733999999998,easy,30 1.0,2.299,easy,30 0.0,2.073,easy,30 0.0,1.274,easy,30 1.0,2.958,easy,30 1.0,3.011,easy,30 1.0,0.565000000002,easy,30 1.0,0.764000000002,hard,30 1.0,1.13,hard,30 1.0,1.07,hard,30 1.0,0.539000000002,hard,30 1.0,0.598000000002,hard,30 0.0,0.519999999998,hard,30 1.0,0.584000000002,hard,30 0.0,2.743,hard,30 1.0,2.139,hard,30 1.0,0.736000000002,hard,30 1.0,1.31,hard,30 1.0,0.601000000002,hard,30 1.0,1.038,hard,30 0.0,0.789999999998,hard,30 1.0,1.876,hard,30 1.0,0.755000000002,easy,31 1.0,0.619000000002,easy,31 1.0,0.790000000002,easy,31 1.0,1.479,easy,31 1.0,5.281,easy,31 0.0,0.461999999998,easy,31 0.0,0.484999999998,easy,31 1.0,1.212,easy,31 1.0,0.880000000002,easy,31 1.0,0.877000000002,easy,31 1.0,0.935000000002,easy,31 0.0,0.606999999998,easy,31 1.0,0.991000000002,easy,31 0.0,2.033,easy,31 0.0,0.845999999998,easy,31 1.0,0.903000000002,hard,31 0.0,1.399,hard,31 1.0,1.178,hard,31 1.0,0.917000000002,hard,31 1.0,0.439000000002,hard,31 1.0,0.927000000002,hard,31 1.0,0.947000000002,hard,31 1.0,0.873000000002,hard,31 1.0,0.501000000002,hard,31 1.0,3.431,hard,31 1.0,1.554,hard,31 1.0,0.615000000002,hard,31 0.0,0.961999999998,hard,31 1.0,1.003,hard,31 1.0,1.303,hard,31 0.0,1.034,easy,32 0.0,1.382,easy,32 1.0,1.926,easy,32 1.0,0.837000000002,easy,32 0.0,1.126,easy,32 1.0,1.425,easy,32 1.0,0.946000000002,easy,32 1.0,2.884,easy,32 0.0,1.71,easy,32 1.0,1.123,easy,32 0.0,0.841999999998,easy,32 0.0,2.775,easy,32 1.0,1.091,easy,32 1.0,1.961,easy,32 0.0,0.464999999998,easy,32 1.0,1.945,hard,32 1.0,0.628000000002,hard,32 1.0,2.079,hard,32 1.0,0.967000000002,hard,32 1.0,0.465000000002,hard,32 1.0,0.445000000002,hard,32 1.0,0.816000000002,hard,32 1.0,0.893000000002,hard,32 0.0,1.694,hard,32 1.0,2.176,hard,32 1.0,0.835000000002,hard,32 1.0,0.564000000002,hard,32 0.0,0.739999999998,hard,32 0.0,0.944999999998,hard,32 1.0,0.538000000002,hard,32 1.0,2.653,easy,33 0.0,2.494,easy,33 1.0,1.601,easy,33 1.0,3.168,easy,33 1.0,0.771000000002,easy,33 1.0,2.08,easy,33 0.0,1.697,easy,33 1.0,1.316,easy,33 1.0,2.156,easy,33 0.0,0.605999999998,easy,33 1.0,1.605,easy,33 1.0,0.755000000002,easy,33 1.0,1.104,easy,33 1.0,1.002,easy,33 0.0,1.429,easy,33 1.0,0.469000000002,hard,33 1.0,1.784,hard,33 1.0,0.417000000002,hard,33 1.0,1.01,hard,33 1.0,1.662,hard,33 0.0,2.765,hard,33 1.0,0.637000000002,hard,33 1.0,0.471000000002,hard,33 1.0,0.963000000002,hard,33 1.0,0.646000000002,hard,33 1.0,0.775000000002,hard,33 1.0,0.627000000002,hard,33 1.0,0.805000000002,hard,33 1.0,1.042,hard,33 1.0,1.04,hard,33 1.0,0.937000000002,easy,34 1.0,1.923,easy,34 0.0,0.445999999998,easy,34 0.0,0.550999999998,easy,34 1.0,2.893,easy,34 1.0,0.599000000002,easy,34 0.0,1.549,easy,34 1.0,1.462,easy,34 1.0,0.749000000002,easy,34 1.0,1.31,easy,34 1.0,0.660000000002,easy,34 1.0,0.610000000002,easy,34 1.0,0.784000000002,easy,34 1.0,1.922,easy,34 1.0,0.688000000002,easy,34 1.0,0.729000000002,hard,34 1.0,1.281,hard,34 1.0,0.686000000002,hard,34 1.0,1.053,hard,34 1.0,1.104,hard,34 1.0,0.522000000002,hard,34 1.0,0.670000000002,hard,34 1.0,0.518000000002,hard,34 1.0,0.663000000002,hard,34 1.0,1.811,hard,34 1.0,1.556,hard,34 1.0,1.081,hard,34 1.0,0.584000000002,hard,34 0.0,2.581,hard,34 1.0,2.193,hard,34 1.0,2.309,easy,35 1.0,0.694000000002,easy,35 1.0,1.995,easy,35 1.0,1.155,easy,35 1.0,0.480000000002,easy,35 1.0,0.902000000002,easy,35 0.0,0.601999999998,easy,35 1.0,1.816,easy,35 0.0,0.989999999998,easy,35 1.0,1.728,easy,35 0.0,1.052,easy,35 1.0,1.166,easy,35 0.0,0.570999999998,easy,35 0.0,0.695999999998,easy,35 1.0,1.778,easy,35 1.0,0.458000000002,hard,35 1.0,0.612000000002,hard,35 1.0,1.42,hard,35 1.0,0.681000000002,hard,35 1.0,1.015,hard,35 1.0,0.713000000002,hard,35 1.0,0.431000000002,hard,35 1.0,2.549,hard,35 0.0,0.688999999998,hard,35 1.0,2.378,hard,35 1.0,0.896000000002,hard,35 1.0,0.710000000002,hard,35 1.0,1.165,hard,35 1.0,1.626,hard,35 1.0,2.431,hard,35 0.0,0.475999999998,easy,36 1.0,0.894000000002,easy,36 1.0,1.025,easy,36 0.0,0.716999999998,easy,36 0.0,1.239,easy,36 0.0,0.565999999998,easy,36 0.0,3.411,easy,36 1.0,0.861000000002,easy,36 1.0,1.418,easy,36 1.0,2.059,easy,36 0.0,0.712999999998,easy,36 0.0,1.726,easy,36 0.0,0.719999999998,easy,36 1.0,1.651,easy,36 1.0,1.244,easy,36 0.0,0.380999999998,hard,36 0.0,1.161,hard,36 1.0,1.17,hard,36 1.0,0.464000000002,hard,36 1.0,1.077,hard,36 0.0,0.486999999998,hard,36 1.0,0.510000000002,hard,36 1.0,0.467000000002,hard,36 1.0,1.575,hard,36 1.0,1.18,hard,36 1.0,0.480000000002,hard,36 1.0,1.918,hard,36 1.0,1.53,hard,36 1.0,2.96,hard,36 1.0,1.05,hard,36 1.0,0.818000000002,easy,37 0.0,2.086,easy,37 1.0,0.785000000002,easy,37 0.0,4.274,easy,37 1.0,1.237,easy,37 1.0,1.755,easy,37 1.0,0.599000000002,easy,37 1.0,0.927000000002,easy,37 0.0,1.352,easy,37 1.0,0.576000000002,easy,37 1.0,0.538000000002,easy,37 1.0,0.464000000002,easy,37 0.0,1.294,easy,37 1.0,0.541000000002,easy,37 0.0,1.14,easy,37 0.0,1.377,hard,37 1.0,1.12,hard,37 0.0,0.861999999998,hard,37 0.0,1.426,hard,37 0.0,1.204,hard,37 0.0,0.738999999998,hard,37 1.0,2.285,hard,37 0.0,0.792999999998,hard,37 1.0,2.096,hard,37 1.0,1.569,hard,37 1.0,1.261,hard,37 1.0,0.903000000002,hard,37 1.0,0.603000000002,hard,37 1.0,1.011,hard,37 1.0,1.187,hard,37 1.0,1.211,easy,38 1.0,0.397000000002,easy,38 0.0,0.591999999998,easy,38 1.0,0.650000000002,easy,38 1.0,0.529000000002,easy,38 0.0,1.808,easy,38 1.0,1.123,easy,38 0.0,1.088,easy,38 1.0,0.721000000002,easy,38 1.0,0.913000000002,easy,38 1.0,1.278,easy,38 1.0,0.798000000002,easy,38 0.0,2.492,easy,38 1.0,0.646000000002,easy,38 1.0,0.596000000002,easy,38 1.0,0.997000000002,hard,38 1.0,0.471000000002,hard,38 0.0,0.860999999998,hard,38 1.0,0.536000000002,hard,38 1.0,1.775,hard,38 1.0,0.503000000002,hard,38 1.0,0.738000000002,hard,38 1.0,0.517000000002,hard,38 1.0,0.670000000002,hard,38 0.0,1.101,hard,38 1.0,4.019,hard,38 0.0,1.132,hard,38 1.0,1.863,hard,38 1.0,1.825,hard,38 1.0,0.753000000002,hard,38 1.0,1.492,easy,39 1.0,1.761,easy,39 1.0,0.922000000002,easy,39 1.0,0.672000000002,easy,39 1.0,1.003,easy,39 0.0,1.363,easy,39 0.0,1.287,easy,39 1.0,0.825000000002,easy,39 1.0,0.715000000002,easy,39 1.0,0.680000000002,easy,39 1.0,0.765000000002,easy,39 1.0,1.705,easy,39 0.0,1.43,easy,39 1.0,1.17,easy,39 0.0,0.514999999998,easy,39 1.0,0.713000000002,hard,39 1.0,0.657000000002,hard,39 0.0,0.542999999998,hard,39 1.0,0.524000000002,hard,39 1.0,1.332,hard,39 1.0,0.535000000002,hard,39 1.0,0.503000000002,hard,39 1.0,0.756000000002,hard,39 0.0,0.959999999998,hard,39 0.0,1.392,hard,39 0.0,0.738999999998,hard,39 1.0,1.202,hard,39 1.0,0.653000000002,hard,39 1.0,0.690000000002,hard,39 1.0,0.461000000002,hard,39 HDDM-0.2/hddm/examples/simple_difficulty.conf0000664000102500010250000000010111713025571021735 0ustar wieckiwiecki00000000000000[depends] v = difficulty [mcmc] samples=20000 burn=3000 thin=10 HDDM-0.2/hddm/examples/v('easy',).png0000664000102500010250000012725411726446356017562 0ustar wieckiwiecki00000000000000‰PNG  IHDRèXñÌ%sBIT|dˆ pHYsaa¨?§i IDATxœìÝy\TåþðÏ€€€ã‚àŠ .¸R*š¨intÝKÔ›”æ½V¦iÀL24Ã4¯%êÍ«–KvSÃL+»$z¯ ie&Z¡bi.”‚Ãëüþ˜f˜aöá çÌ™ÏûõêÕÌœ3gžxd>çûœçQ¨Õj5ˆˆˆˆˆˆˆHTb7€ˆˆˆˆˆˆˆЉˆˆˆˆˆˆ$ˆˆˆˆˆˆHЉˆˆˆˆˆˆ$€ˆˆˆˆˆˆHЉˆˆˆˆˆˆ$€ˆˆˆˆˆˆHЉˆˆˆˆˆˆ$€ˆˆˆˆˆˆHЉˆˆˆˆˆˆ$€ˆˆˆˆˆˆHЉˆˆˆˆˆˆ$€ˆˆˆˆˆˆHЉˆˆˆˆˆˆ$€ˆˆˆˆˆˆHЉˆˆˆˆˆˆ$€ˆˆˆˆˆˆHЉˆˆˆˆˆˆ$€ˆˆˆˆˆˆHЉˆˆˆˆˆˆ$€ˆˆˆˆˆˆHЉˆˆˆˆˆˆ$€ˆˆˆˆˆˆHЉˆˆˆˆˆˆ$€ˆˆˆˆˆˆHЉˆˆˆˆˆˆ$€ˆˆˆˆˆˆHЉˆˆˆˆˆˆ$€ˆˆˆˆˆˆHЉˆˆˆˆˆˆ$€ˆˆˆˆˆˆHЉˆˆˆˆˆˆ$€ˆˆˆˆˆˆHЉˆˆˆˆˆˆ$€ˆˆˆˆˆˆHЉˆˆˆˆˆˆ$€ˆˆˆˆˆˆHЉˆˆˆˆˆˆ$€ˆˆˆˆˆˆHЉˆˆˆˆˆˆ$€ˆˆˆˆˆˆHЉˆˆˆˆˆˆ$€ˆˆˆˆˆˆHЉˆˆˆˆˆˆ$€ˆˆˆˆˆˆHЉˆˆˆˆˆˆ$€ˆˆˆˆˆˆHЉˆˆˆˆˆˆ$€ˆˆˆˆˆˆHЉˆˆˆˆˆˆ$€ˆˆˆˆˆˆHЉˆˆˆˆˆˆ$€ˆˆˆˆˆˆHЉˆˆˆˆˆˆ$€]"ÒÒÒзo_dgg›ÝçöíÛ˜7oæÌ™c´­°°ððð€‡‡>ÿüsg6—ˆˆÜŒ©¿SÄðáÃѵkWüóŸÿ„Z­Öm+//ÇË/¿ŒE‹aþüù¨¨¨£ÙDDD.…]rssQVV†'N@¡P˜Ýï·ß~Ã/¿ü‚¢¢"£mï¿ÿ>6n܈àÀ7nœ3›LDDnÄÔß©’’lÛ¶ ,À´iÓ0þ|¤¦¦êž“””„Ž;")) ­ZµÂâÅ‹Åj>‘ËP¨õOw“hÊÊÊàããƒ+W® uëÖf÷[°`®]»†-[¶è«¨¨À¸qã’’‚úõë×Fs‰ˆÈÍTÿ;•‘‘Î;£^½z€§žz رcîܹƒàà`ܸq 4ÀåË—Ž«W¯ÂÏÏOäwBDD$]¬ K„———MûÕ©SÇè±´´4ìÛ·!!!xã7ŸŸ/tóˆˆÈÍUÿ;Õ»wo]8€Î;£mÛ¶€3gÎ 00 4#??gÏž­µö¹"ã´G’fjü˜1c R©päÈLŸ>?ýôvíÚeòùyyyØ¿?Ú¶m ___g7—ˆˆþT\\Œ+W®à/ù ÅnŽà¾øâ |ôÑG€¬¬,è¶yyy¡I“&ÈÎÎFŸ>}ŒžË¿MDDâûß&WÄ€.~~~1b222н{wäåå™ìdû÷ïÇ”)SDh!À¶mÛðÔSO‰Ý AŠvíÚ!44«V­BûöíëÖ­Ãðæ›oâþýû˜3gŽÈˆHúXA—ÄÄÄ &&ÆàñŒŒ ƒû}úôÁÑ£GžŸ`p­‘LýZ´h-Zdñ9îp*‘Љþ-v$‹Ÿiü\ÌãgCdšœú†J¥‚Z­†B¡€R©4Ú.§÷j ß«üˆù>­õ-’7t¢?¹ËGð³1Ÿ‹yülˆL“Sßظq#T*”J%âãã¶Ëé½ZÃ÷*?b¾Ok}‹äˆˆˆˆì…òòrԩï“DBbßroü©‘ÝÚ¶m+vˆd‰}˽qw""""""" `@'""""""’q'""""»¥§§£¤¤>>>›7WôÊÊÚi oåÊ•P©TP*•ˆ»9D²Á¾åÞЉˆl $yèÝ»æǼq£ê¶\ºþ{zé%ÓûxxTÝÖ?ñAD®%22eeeðòò»)D²Â¾åÞЉˆl $==…9¦BQu[.•dm@ ¼½Mï£ÿù1 ¹®°°0±›@$Kì[îÍÃú.DD¤$=ú—S? Ë¡‚®VWôæÍÍïÇ€NDDDd:9݉'Än‚ÝnÞ¼‰›7oŠÝ ’ý ·Ptý /‡€~÷.PZª¹Í€NDDDd?q'§Q©TX°`¦ÿ¹6ÕÈ‘#ŠþóŸ"·LcæÌ™¸yó&>ÿüs£mÍš5Crr2zôèarZ šÆ!îæUTË—_}Uõ:‘ü}111:t(°hÑ"ƒãýðØ2e zôèݨQ£FéÚÙ¾}{>úè#øûû£^½zÈÌÌD×®]m:a0lØ0¼þúë(//7¹}Æ 8tè`òäÉ2d®^½Šüãèܹ3>ŒY³f!88Glß¾S§NE·nÝ0tèP¤§§³¬¬ ‰‰‰xøᇎùóç#//O·]­VcÅŠøË_þ‚¶mÛbìر¸~ýºÕ÷BŽsF].CÜõ»FPùåÕ´8ĤäæÍ›1búõë‡^½záþýû ABB‚ØM³Ill,:w‚³û<óÌ38tè/^lÓ1ù大§cË–-«/??7oÞÄåË—mÚŸˆˆLc@'›edd`È!P(8|ø°îñ·Þz ýúõCff&`Û¶m&¯ÃîÖ­›.WVVbÈ!xì±ÇpàÀìß¿~ø!–/_@@'OžŒîÝ»ãСCˆ‰‰Áo¼ÔÔTÝñbbb0oÞ<üøã8zô(<þL;_~ù%† …Býû÷ÃÏÏðä“O"22Ÿ}öxàtëÖÍà„9:u‚J¥Â¹sçLnŸ9s&^yåÀnjÇ£¬¬ 8þ<6oÞŒ1cÆ ÿþðôôÄwß}‡çŸkÖ¬ÑUÝ'MšdpÌqãÆ!//ÀáDZvíZÌ;W·=>>÷îÝÃÞ½{qöìY\½zS§Nµú^Èq\fÍ<ý z.œb@'©(--Å„ ðöÛo#""ð믿";;ûöí¹u¶9tèΟ? .XÜ/!!—/_ÆæÍ›­³S§NˆÇ<`s;zôè¼¼<ìرÃæçè{ñÅz‘Ü0 “Í<==±hÑ"¨Õjû,|||àëë‹ÿû߀wÞyõêÕCTT”Ññ|ðA<øàƒÝ»w³gÏÆرcѪU+ôíÛÿùÏ ö¹ÿ>Ö®]‹'žx={öDß¾}±`Á£QV?ýôæÌ™ƒÁƒ£nݺðððÀßÿþw@TT”î}xxxèFšëå;tè4mÚyyyèÑ£‡]'F…Ï?ÿ\7wŠ-Ôj5þõ¯!::ÁÁÁxñÅ¡V«Ñ¯_? <½{÷ÆÇŒaÆá™gž1xnNN-Z„Q£F¡qãÆððð@HHM8íµ×PRR‚7nèN:Ÿ?Þæ¶É‘¶oõë×Oì¦Ð% 77eee8qâúßØ«ùí·ßðË/¿è&=Óºyó&æ΋+V`ùòå˜9s&îÝ»ç´öN™2 4ÀÚµk±cÇ<©·žR`` ‘››kñM›6ÅáÇÖE¿~ý:îÞ½‹7n`àÀhݺ5¶nÝjò(ooo¬X±_ý5‚‚‚0bÄüüóϺí À€ðü™™™hÕª•Í³Ñê»sçè&Á«©Ý»wã‰'žÀÒ¥K‘””¤«Ðk5oÞׯ_7x/€æäGAAš6mŠüü|ìÞ½Û`{yy¹ÑsH8\fÍ<{+èœÅ„Ö«W/\¾|Y7Ìû½÷ÞÃc=¦Û¾iÓ&<òÈ#HIIÁ´iÓPPP€­[·¢M›6ºÑNú|ðAƒÉPoÞ¼‰ˆˆ<õÔSøÏþƒË—/£E‹xì±Ç°sçNÝ~¯¾ú*’’’°iÓ&üðÃ5jÞ|óMƒ^jµO>ù$^xᤧ§ãÎ;xâ‰'tÛ?ýôSÌž=€æKúÿû_ݶfÍša÷îÝÀÙ³gˆ|Ð`ˆ¾5ÚŠÜ×_mÓþjµóæÍÃäÉ“‘’’‚>øk×®ÅúõëuûìÚµ Ÿ}ö>lt2"11ÁÁÁØ»w/rss±jÕ*ݶþýûãÀhÚ´)š7o®;éìÌuµ‰ˆ¤Ž]‚‚‚t×b[Ò³gO„……ýñ[¾|9"##¡P(àããƒððp¬]»ÖYÍ…¿¿?¦M›†­[·bïÞ½g÷à0˜qÜ”þýû£²²Ó¦MÓU¾oܸ  Q£FX·n®\¹‚˜˜ƒçUþYnËËËÃk¯½ˆˆˆÀþýûáááa4zà•W^ÁÙ³g1uêT̘1Ãb›ÌÍT«½VÏÔÄwZÚ êþø㋯qqq=z´Ñí{0`à‹/¾0ؾk×.øûû#""X´h>üðC”””   sçÎ…R©´úúägTÐå2Ä× “4hЯ¿þ:­Œ‘››‹ÊÊJÝ\%Ú“ÝÚÑ_Õ-[¶ ­ZµÒÝŸ;w.F…þýûÐ\¤TûÿââblÚ´ }ûöÕý[¬­Îÿøãºc¥§§ãÂ… hÒ¤ ÀÏÏ«V­Býúõuû¼õÖ[hÞ¼9~úé'´íàÁƒ˜1c†n~”N:aáÂ…6~Jн®­KÊ) ,]ºT÷zLJR©4˜ .::ÚdŠŠŠðÑGéND{xx 66V7¿ c@—kCµLéÏÌÌ4X.¬uëÖNæþ /   Àde;,, 999Ÿÿ÷¿ÿ]»vÅáÇñ裢qãÆhÙ²%†hÙ²% ðÃ? ¬¬L7$±°°P7Q\rr2¾ÿþ{š™xýýý®c=z4ºwŠ ]ð5åÊ•+hÖ¬z÷î­»f]Kе_ÌLiÑ¢MˆNIIÁ;wtÃÔ«-[¶ÄñãÇQ^^Žëׯë&²Ó¾·iÓ¦¡k×®X¸p!V­Z…#GŽ`öìÙhÓ¦  :uê„gžy—.]Âßþö74iÒ7¶øe“jÎÙtWêÍkÐI*&Mš„Ö­[ãƒ>ÐM¼ h*êú×]_ºt lú7³¬¬ ;vìÀ¡C‡ æM™;w®îßåââbøúúbÓ¦MXºt©Ásœ´îر#***ððÃëVDiÖ¬™nrSðõõÅÿýßÿ¡¨¨ï¾û®A{6oÞltݼ=´aÙž5ß½«­X¯^=äååáüùó8{ö,Ο?orB9???ãïÿ;Þ{ï=݉蔔‡ÛOäôû¹tcj|VV–nu@ó…ÃÒµìB Åرcu×Ìé7n.\¸•JeöùuêÔÁ‘#Gðä“O"(( 6ÄÊ•+uÃå'OžŒ'žx#GŽÄôéÓѲeK<üðøzõª.ø#"";wFLL ¶nÝŠ† ½Ö´iÓðÜsÏY|?jµ~~~øᇌ†ýýôÓOèÝ»·ÑõˆúˆÈÈHlÚ´ ¸pá–/_…BåË—cÿþýº}ß|óMTVV⡇ÂÊ•+1uêT4kÖ GEƒ àé鉯¾ú ãÇ× |ôÑG Ö•_¿~=.\ˆÔ­[ûÛß ¾Ü‘ð¸Ìšy¬ “Txzz"66………xï½÷hBrZZƯÛOû7²ú%F¦\¼x•••HLL4˜7åèÑ£¸|ù2.^¼___ÀÔ©SŠ÷Þ{“&MÒ«Ô;ŒeË–áòåË6l €£G½î¬Y³Ð¨Q#$''£¤¤pøða´k×­[·vðÒ6lˆk×®áîÝ»£¤¤©©©ØµkRSSÍN(·nÝ:Ô©S³fÍB‡ðþûï;Üv"w¡ß·ÈýØðUŠ¤N¡PüaÌÏÏ·:›k\\œQ˜ŽŽ¶k–óÏ?ÿÜäãC† Á„ pðàAÝpBS°mÛ6“Û6lht†][]ת´±ä˜žžŽ­[·ZÜ'$$·nÝ2Ö~ø¡Áä¦øøø„pKÑé:t¨nFv­ëׯÜoܸ±Åµd=<<””„¤¤$‹í"á8c™5¹ qgÝXJJŠÑ¿aÎœ„ª<ûì³HJJš5kO>ù'N48Á­æmËÏD{éRff&&Mšdqß~øQQQ˜5k–nµíx}óæÍÄ °páB|üñÇ2d þM÷÷÷Ç‹/¾ˆÅ‹cÓ¦M˜={6Þ}÷]Ìš5Ëú‡`…ö;ƒµeÔ,Q«ÕºkåÍej¦ >/^ÄòåËñÎ;ïàÙgŸÅ§Ÿ~ŠO>ùDwbƒˆ é÷-r?¬ Ë@DD~ýõWÝýœœtëÖÍâsV¯^/¾øÂà?G– 3gÕªUºê…˜¾þúk´hÑÂde½:•J…Û·o,‹–‘‘üíosb+É8»‚îÊCÜYA7môoìêÕ«Ån–[P*•xî¹çpíÚ5|üñÇزe ž}öYƒ}zõêÀø$S´ë¡¯[··oß6Ú¾lÙ2Ý$ÄÈ‘#1oÞ<«Çmß¾=>úè#?~Í›7Ç’%KŒF¿ÅÆÆÂßßo¿ý6®^½Šü£F²zlKJJJpûömÕ8 ûøøèþ³¤~ýúXºt)~ýõW 4{÷îÕÀ "c¶ö-’'tˆˆˆ0¸F%''Gôu›7oŽ'Ÿ|Ò`–×ÚòÓO?¡]»v9r$bccñÖ[oY}Ε+WššŠ·ß~[÷Ø;wðÖ[oáÃ?tfsÉE8£‚.—!î5© »ò‰ ’®ØØXÔ©S/¿ü2ºwïn4fçÎáíí[·nY=–¯¯/F»wïbØ°aº ß ±aÃxyy¡N:8xð ŠŠŠt“©U#½ôG|íÞ½Û`tEïÞ½ñÊ+¯@­Vãܹs¯Ý¨Q#Ìœ9¿ýöÆgò²2SòóóÍnÓŽØÒ?í,yyy´6mÚëÖ­ ¹>>ÈÊÊ2ZcÜ”={ö 00Ðì„­×®] ùî`Mnn.ÔjµÁ„{€æ»Fõ“Úת¾Ìê¦M›ðÁè.8sæ ||| .è3ðô IDAThß¾=òóó±jÕ*,]ºÔìpy""wÀ€.………X¿~= ¶lÙ¢›íuÇŽ8sæŒn¿'NàÈ‘#ÈÈÈ0XB¦K—.ˆÅ‚ °dɬZµÊê5èµå¡‡˜1cjõ5[·n .`ÿþýwèõêÕCBB‚C릓'%%¡C‡غu+ÂÂÂj¥ºO$e)))ؼy3W}LÎô 'NÄĉÖÆš0µ4œÔ¹b›É¹´AR¨áí€|*èúCÜy :IEÏž=±{÷n³ÛçÌ™ƒO>ù©©©6M¼Ö¼ys«ŽÆÇÇ#>>Þà±>øÀàþ„ lžäTëÞ½{6O5fÌÜ¿}ô‘ÑPwµZ 6`êÔ©hܸ±Ùc´hÑÉÉÉHNN6Ú¦]ÿûßÿtKÌ-_¾Üä±ly¯}úôárRDzu}‹ÜS‘ ´ß1…ÞÈçtVÐÉ) ¼ÿþû˜4ižy棵¾¥âÚµk8þ<hósÔj5víÚ…ÿûß:tuêÔÁŠ+i[dd¤ Ç!"Cì[îC܉ˆlàŒ º\†¸³‚N®ªC‡xæ™g°hÑ"±›b ??_·öùË/¿Œ7ÞxÃæçž>}Ë–-Ã[o½e09Þ½{÷ðÖ[oaÏž=¨_¿¾àm&""a0 Ù@$…¬ Ëeˆ;+èäÊæÌ™ƒþýû›Î-†«W¯"((õë×GûöíѨQ#Œ=Úæç7oÞ¯¾ú*:tè {ìÚµkˆ‹‹ÃÛo¿ÐÐPg4›ˆˆÂ!îDD6ÐV¸UAwå€^“ º+ ù;v,nݺ…sçΡK—.¢¶¥iÓ¦4h._¾Œéӧ㥗^²ëùú˼iýüóÏxÿý÷á!äFh&“«¨¨€§§'Z´h!豉Üû–{c@'"²‡¸›ÇeÖHš4i‚&MšˆÝ xyyáÿûŸ ÇÔ.'´;v@¥RA©TMŽGDŽcßro èDD6pÆ$qrâ®­ {xØöùpˆ;‘<<ýôÓ¨¬¬¼2OäîØ·Ü:‘ XA7O[A·uuBt"y05œžˆjŽ}˽ñ´ ‘ ¸ÌšyÚ º-Ä èDDDDæ0 ÙÀt¹ qgˆˆˆHâNDdg,³&—!î5© »òû&rw©©©(..†¯¯/ÆŒ#vsˆdƒ}˽1 ـˬ™go³¸ÉCii)JJJà)ä?ŒDľåæЉˆlÀ!îæÙ[Ag@'’‡¨¨(±›@$Kì[îˆ,úñG ) ¸}»ê1…=˜7O¼vÕ6gOçÊC½í­ +šÏ±²’ˆˆˆH:Y´p!šjüø‘#Àc:Õ~›ÄàìeÖÜ©‚h>G[zE•tèøú:ÖF""""WÀY܉Ȣ[·ÌoËÍ­½vˆÍt¹ q··‚Tè°Ð++‡xv¯ß9"©S©TÈÏχJ¥»)D²Â¾åÞXA'Y+)¾ü¸sGì–jÓ6LØ°ç,ú•ãâb`þ| 9Yó˜+˶—³+è®üY:ZA¬ô‹S§4·ïÜžxÂþ6‘ð6nÜ•J¥R‰øøx±›C$ì[îÝM=ú¨í_¦ûöRR\shil,ðÞ{b·Â´uë€Y³Än…uúÁÔËË°JêÊ¡Ò^Î^fÍ]+èÖ~‡ŠŠ ïkO‘ø¢¢¢P^^Ž:öt~"²Š}˽ñ§î¦ô'ü²fÏ - pÅ %¿ÿ^ì˜÷ý÷®ÐÀêNÝˬÉaˆ»Z]õ;bO]ûÞ­UÐÿøÃð>'•#’Ž¶mÛŠÝ"Ybßro ènªY3ÀÇÇò>*UÕÐðü|ç·ÉôƒÃúõâ¶rr€ÄDÍmW · èή »êg©­žι½z@×=""""¹a@wS_~ ôêeyŸ €çŸ×ÜvÕð ýòïí <óŒ¸m€_u½€^½rì®tÓô³3®AgˆˆˆÜ :™%‡ 按½jBÿ3µ5h¤¥gÎ3gJ¥sÚeIõʱ~/ì¥ÿ>¹Ìš!ýkÂk£‚΀N$ééé()),vsˆdƒ}˽1 “YrbΨzÖ„~;lùL³³Q£ªnÿë_Îi—%ânø>]}ˆ{e%pàpþ¼0ÇÓŸÄt"÷’ÂÂBøûû‹Ý"Yaßro èd–‚˜3®® {?ÓÇ«n¯]Ë€.ýPèêCÜwírÞ2eΨ Þç5èDÒ1mÚ4±›@$Kì[îM"±…¤HAÌÕ‡¸Kaس¥€.…öÕ†Ú¨ ×Ögyò¤óŽÝ¿¿íûÚºÌ+èDDDäNXA'³Ð…gïgjkh+.JK·Õ© Ô«gÛ±ªã$qΫ ‹1Ä]ÿ½,\tê$Ìq[´yÄöý¹Ì‘1t‰HKKÃ믿Ž;w¢M›6&÷9xð öîÝ‹ŠŠ Lš4  Ðm+,,D«V­pïÞ=À§Ÿ~ŠÇ¼Fm’CsõkÐm èC†Ç›ß œ;çXõ·úerXÌ^ú¡PÈ º£ôß˨Q@DDí¼nu¼ˆˆˆÈºäæ梬¬ 'Nœ€B?ýè¹yó&æ΋S§N¡´´½{÷ÆÑ£GÑ°aCÀûï¿7¢Q£F ÈŒrèR¾]¨!î%%–Ã9 ™ìæMM•Ó^¼½vfq£‚.æ‰+t"×·råJ¨T*(•JÄÇÇ‹Ý"Ù`ßro è„‘#GZÜgùò刌Œ„B¡€ÂÃñvíZ$&&¢¢¢ûöíCJJ êׯ/X»äÄÜaˆ»þqÚµ«Z»¶mNŸÖÜv4Ø0 ×Îw1*è®Ð9I‘tDFF¢¬¬ ^ö,á@DV±o¹7t‰°Ö3331eÊÝýÖ­[ã›o¾ ¿oß>„„„ ..±±±‚u9L&µ€îŒ!îúû´m Ì›Wuÿ‡Ð…à¬IâÄâî ½ú,IGXX˜ØM ’%ö-÷&‘¿dMVVt÷[¶l‰ììlÀ˜1c R©°}ûv¬_¿Ó§Oä5åĤv º3†¸ëÿlª_!¡ÿ¾­õ×¥®i]{ìêïaãF ,L3xŸ>ÀÞ½Ž½Ž”Éi™5©tÎâNDDDdŒCÜ]DDD~ýõWÝýœœtëÖÍh¿-Z`èС8vìÆŽköxqqqº æ´¢££­»/§€.• º3†¸;»‚niˆ{i)U=öê«šÙÁåÄY¡–CÜ5ï[­6þÝÕrÕkÐSRR’’bð˜v• "¹8yò$JKKáííððp±›C$ì[îÝEDDDàüùóºû999èÙ³§É}=<<ÌnÓZ½z5zõêeqŸê½¸¸sÇpOO Y3Ëm“Ôº½'=ì ñά › è%%†ánßvìu¤ÌYˬqˆ»FE…áïª>W­ W?á §Nâ-’•cÇŽéfšæï6‘pطܺD”••ÊõÊC+V¬À¸q㊱cÇbÆŒ€¢¢"\ºt ï¾û.`Ïž=èÔ©:wîŒo¿ý­ZµBpppÛ¤DÒÓ·Þòó÷=HM­ñË.Ô ‘®i51, X³†× ;»‚nê˜þ*ËJm,³æ®t@ž]ŽÒÒÒðúë¯cçÎhÓ¦ Í߬ùóçÃÏÏ¥¥¥X²d‰îr,KÛHæÌ™#vˆd‰}˽1 K@aa!6oÞ …B-[¶ 66رcÚµk‡ÐÐPtéÒ±±±X°`*++±jÕ*Ý5è?ÿü3f̘ oß¾HNN¤]úAìÀóû}ù%píвeÍ^ïÔ)`É’šC뇀¥w º³gq¯þ>«‡ {éÛ:‡¸ ÏÖßMtiÈÍÍEYYNœ8…Þ™¥¤¤$tìØ3fÌÀºuë°xñb$%%YÝFDDD¦1 K€¿¿?bbbcðxFF†Áý‰'bâĉFÏOHH@BB‚àí2j‡4¾ÿøýwÍc%%5½ ªn{z:**ª¾ÀçæJ¯‚îŒ!Í–*è5ânjh·µ€^ZjÿëH]mLçÎCÜÿø0·ÒdõK(\åt¹ ÂÈ‘# »sç’““qãÆ ÀÈ‘#Ž„„üñÇf·ùùùÕzû‰ˆˆ\…DêŠ$Eæ‚È?ÿ ìÚ \õ˜_š¯]«ºýïkB¿½ÿ}þyÕ1ôƒ¢T:`ûìÕ€°tG~F¦Â+è—tÆooÓÿU;OÉ ºˆ¼ªE9sæ Ñ A@pp0òóóqöìY‹ÛˆˆˆÈI:9Âßß111ˆ‰‰1z|õêÕfŸcnÉCjj*T*”J%âããÅn‘l°o¹7t2Ëž!îöVgµû×ø7ЮAzˆ»¥ zM~F€ë q//ú÷NœpþkÉi™5),±¤¦¿ÇDä³gÏ» D²Ä¾åÞÐÉ,gqwV@w…kÐ]©‚îè$qµÐOªpÁÁÂKÌ!îRêÖ°‚N$M\^È9Ø·Ü:™åÌIâj; K©RèÌkÐ-MW›týϾ6è/××¹3Э›s^§S'`Ú4áŽ'æwt""""éa@'³j£‚îåe»,q÷kÐ-M'çYÜ‹ŠªnO™$&Öîë;JÌ!îRêÖ0 ‘»`@'³\±‚îî×  ½Ìš©Iâô_C*³¸ët?¿Ú}íšàwÛÔt5"rŽ””ÁÏÏÑÑÑb7‡H6Ø·Ü:™å¬ ºZ]Üùt¡†¸ÛZA—ó$q®Ð9ÄÝ6¬ IS`` Š‹‹áëë+vSˆd…}˽1 “YΪ ë‡:wè®PA·6Iœ+èþþµûÚ5!æw¡ûž3yxh>+ý“{D$¾ÈÈH±›@$Kì[îMBSg‘Ô8«‚^[]ˆ»”&‰«ÍeÖĪ ×ö$q®ZAçwÛiÛË€NDDDræB5ªm®XAw…kÐ9ÄݙˬIyˆ{aaÕmW èB qýu`Ïãß…xüq`Ñ"ÃÇ]9 ——ót"""’7t2Ë\@׆ ©WÐÝqˆ{õŸYM—Y35IœÔ‡¸»j@w¤‚þË/@R’ùí?þDG¡¡U¹j@¯SGsÒt"é¸~ý:***àéé‰-ZˆÝ"Ù`ßro èd–©€®½fݺÐCÜ¥RA÷ð¨j ºí´×V;ÐoݪºíéYµlaYYÕÏ.7WC܉¤gÇŽP©TP*•ˆ»9D²Á¾åÞÐÉ,s]ËÑê¬þ¾µµº”®A·§‚noˆº‚nm’8ýcÖ­[”9Iœí´Ý‘!îüQuû•W€%K4·_}X¶Ls»ú|ÚŸ:ÕÔÓO?ÊÊJxHé,‘ °o¹7t2ËZ@—âwý6é‡D)…{®A·7Ä‹YA÷ñ‘F@wµ ºvä#tý“PuëVÝöö®º]= ³‚NDB » D²Ä¾åÞxZ†Ì2Ðõ¿ÔKqˆ»Ba½Ýb³gˆ»½É9³‚n- ëDÎân;íIGº~Ý\@×ñ€ëtíï2'‰#"""9c@'³\±‚nî˜R #BqwfÝžIâô"gq·ögVÓ!î>>¦o³‚NDDDä:8ÄÌrV]?¼9# › Rº„G? «ÕÆ¡ZŸ+èúí­>Ä]KÌ!î¾¾µûÚ5¥ý™qˆ»e èDÒ“ššŠââbøúúb̘1b7‡H6Ø·Ü:™Å ºsè†ÖºØt{'‰Ó+ ûúJëdŒ-8ÄÝ6 èDÒSZZŠ’’xºÚ?(DǾåÞÐÉ,W¼ÝÜ1¥ôï›~[*+-ÊšÎâîìIâôƒ¸3+è%%ÀgŸׯ›Þ®}ÜÕ†·Îâ.ç :¯A'’Ž¨¨(±›@$Kì[î]"ÒÒÒðúë¯cçÎhÓ¦É}<ˆ½{÷¢¢¢“&M€lÚæ(VÐCÿ3¬¨°üØ;Ó{õ zmNgéºçšZ¾X´Èú~®Ð1ÄÝÜÏB97)õ [h—YA'"""9c@—€ÜÜ\”••áĉP˜ï|óæMÌ;§NBii)z÷î£G¢aÆ·ÕDmTÐ…^pkÐëUS±+èÖ&‰«­!îÇ۶ߠA¾nmª‚nË5è¦~ž®‚C܉ˆˆÈ0 K@PPFŽiqŸåË—#22 …>>>ÇÚµk‘˜˜hq[M°‚îÕ‡¸["õ º¹Yܯ\±ÿµ,¹{·êöŽ¦7ë׆öukCM®Aׯ ›⮿©Ÿ§«`@'’•JµZ …B¥R)vsˆdƒ}˽1 K„—•Rrff&¦L™¢»ßºuk|óÍ7€Ÿ~ú S§N5¹­&x ºsTân‰Øt{&‰ÓˆÛ·kû¦Mö¿¦)Ú€^¯ðÄÂS*j2ÄÝÞ º:¯A'’Ž7B¥RA©T">>^ìæÉû–{c@wYYYÐÝoÙ²%vïÞ 8wîœÙm5a­‚îhu–½ê¶«WÐõÙ¶­f›¶=￬_/Ìe Ú€Þ¨QÍ%5ÎânîtWè¼Hz¢¢¢P^^Ž:ÎøcNäÆØ·Üê.B¡P,µŸŸzõê™ÝæïïoñxqqqFרGGG#::ZwßU‡¸Kýt{†¸K©‚n- ·i¤¤TU¸ÕjàÖ- eKû_WŸZíCÜ-Ó¶W­¶¾<¡¤¤¤ %%Åà±{÷î‰Ô"çhÛ¶­ØM ’%ö-÷Æ€î""""ð믿êîçää k×®f·uëÖÍâñV¯^^½zY܇CÜ£6+莞D1ulS]»‡ð׿‡ﾫyìÆšôââª*°:‡¸Û¦úÉ&©ªŸð€S§N!<<\¤‘+øWÒŠˆˆÀùóçu÷srrгgO«ÛjÂYtý‰ÅÜ= Ûs º¹Š¡¥ º£'QL=ÇÔ5èú´7o^õØö¿fuúÄÉ1 ×æ,îr è›69gˆ‡:uþ¸DDDD¶b@—ˆ²?Sk¹^Ò]±bƇÐÐPŒ;3fÌáÒ¥Kx÷Ï2¥¥m5a*²‚^sŽÎân. ;³‚nmˆ»©×j֬걛7íÍêÜ%  9Ä]ÿ¶Ü†¸À¬YÎ{Ë—V­œs|"¹IOOGII |||0xð`±›C$ì[î] ±yóf( lÙ²±±± ÄŽ;Ю];„††¢K—.ˆÅ‚ PYY‰U«Vé®3·´­& Íæ*´B\ƒÎuÐ-ï+¥ ºös5×VÐÃ!î¶éÛWsù„3UT èD¶ÊÎÎFaa¡ ß9ˆ¨ û–{c@—ÄÄÄ &&ÆàñŒŒ ƒû'NÄĉMÃÒ¶šðð0ÿ¥žtÇØ3ÄÝ•*è¦zr2ðÁö¿®¾¢¢ªÛr èB q77Iœ~@×ÿRŸ°ÅâÅ@ÿþ@NŽðÇ>tH3Á! ™ó€ˆl3mÚ4±›@$Kì[î,ªÐ]awW èR¯ [›$NŸv{ëÖUåçkþŠ~ø—‹šTеÃ×½¼ .r\fÍË ;Ö9ÇV««ºþ !"""¢ÚÆ€NYZ¶‹tÇ8ºÌZM+èµ5I\“&ÀÂ…ÀæÍŽUíÍéј4I¸ãIEM®A×VÐõ‡·¶-³&õYÐk“Ÿ_ÕmVЉˆˆHLüŠFUc®PAw¥kÐ…â®ìªï£ÿùÖÖ$q€f8òâÅö¿ž;ªÉ÷ŸÖü_¿bÈótgòõ­ºÍ€NDDDbb@'‹XAžÐCÜ«¯E®Oÿ}W¯ÚÖF­;wŒe­‚NöqtˆûÎUϱTAg@·ŽÈ1+W®„J¥‚R©D||¼ØÍ!’ ö-÷Æ€NYª ;Zu÷€.ô2k–*èú?¯S§j6;5ºs8:Ä}ïÞªÛJ¥á6OOÍòYfÍ™ô‡¸ót"ÛEFF¢¬¬ ^ÎX’…ȱo¹7t²È‡¸K= Û3Ľ¦t@3iÛo¿ÙÞ>s‚ƒÍ¿ ­ÏØ•8:Ä]?x¿÷žñvooM5˜tëXA'rLXX˜ØM ’%ö-÷Æ€N9cˆ{YYÕm®ƒnyßšVÐ`ûvM€Ó_’Ë^C†½zin³‚.,G‡¸ë÷£v팷3 ÛŽˆˆˆ¤‚,b]xBÏâni™5xøaÍBa@–£týà­ÍyõÇ €3g4·/\¨Ú.¥>!6ý€Î!îDDD$&t²ˆ“Ä Oè º-!^Hâ.,G¯A·е3»_»˜)ÇŸW.³F䘓'O¢´´ÞÞÞ»9D²Á¾åÞÐÉ"KtýÛRª WŸÑ0^†JLB_ƒn­‚.4sí`Ý1Žq×è¦.éÐÁò¬ý!!ö½žœqˆ;‘cŽ;¦›iš!‚H8ì[î,²TA4»¼¸u ضͶcff>_hS§»wwïjî÷ì  üë8JèYÜ¥RAg@wŒ³†¸oؼû.Ÿo¼­cG`útû^OÎЉ3gα›@$Kì[î,²TA4a³¼¸rEŒí匀>hpã†æÚ[¨àj+G×A7Wa­í :‡¸ «¦CÜ ÓŸ}ÇŽ@rrÍÚæ.||4Ÿ£ZÍkЉˆˆH\ èd‘µ zx8ðí·ŽÛÏèÔɱçZãã#­aíúân.̳‚îÚj:‹»··´N@¹"…BSE/*bHÎ.\¸•J%øq•J%:vì(øq‰È=1 “EÖ*èÿùðùçö©U(€ÈH aÚµÏ9:IœÔ+è 莩éwSÃÛÉ~ èDòvᄆ†:íøçÏŸgH'"A0 “EÖ*è¼–Õ^Ž.³f. ×vÝÜkpˆ»cj:Ä]ÚëÐЉl·aà^½z˜1c†Øͱ¨ªr¾ @|À§TæÉ}¹Rß"á1 “EÖ*èd?{†¸ëo7æŘÅ]{½®>þn8¦¦³¸›šÁì§]j× Ù.,, %%%ð‘ê5e&uÐKìFYäš}‹„€N1  OèIâj»‚hÞCõ“ üÝp ‡¸Kƒ¶‚~ïP¯’•@# IDATžs^ÃÚ 9"WÓ¯_?±›@$Kì[î,²6Äìçè2kR© k_§zØàï†cjZAg@FÓ¦U· Åk¹7t²ˆtá9:‹»Ô*è¶ù––†}ûö!$$o¼ñòóók+èâs• :OÞ8Æ‘kÐYA'"""’'Æ-‘eee! @wßËË Mš4Avv¶Á~“'OÆž={››‹ÒÒRœ={eÎ5f̨T*lß¾ëׯÇôéÓkºøXA—7q'""""-q™B¡0šN¥RÁßßßà±W_}˜9s&þú׿âûï¿ÇòåËuÛýüü0bÄddd {÷îÈËËC`` Ù׋‹CÆ ‹ŽŽ6šˆ‚CÜÅç*ttÇ82ij¸K_JJ RRR ÓNâI$ׯ_GEE<==Ñ¢E ±›C$ì[î]dHHHÐÝ/((@aa!ºvíj°ŸR©ÄË/¿ øßÿþ‡ÜÜ\ 2Äèx-Z´ÀСCqìØ1Œ;Öìë®^½½zõ²Ú>VÐÅgOCÜ]‡¸Ë“©ž§NBxx¸H-"ÞŽ; R© T*uóäQͱo¹7t‘uïÞ•••ÈÍÍEPPrrr T*ѦM³ÏY³f æÏŸovôìÙSö±‚.>kN?´×Ö S'xòÆ1âND®ê駟Fee%<ø€HPì[î]džžž?~<Ž?Ž1cÆ 55K—.¬X±ãÆChh¨nÿ¤¤$4mÚT·|ñÅ EçÎñí·ß¢U«V¤}¬ ‹ÏZ€“J¿ŽÑÿ™mÝ è-Þ`VffÕmÎâNDb±t)9Ž}˽1 KÀºu똘ˆÌÌL”——ãµ×^ ÞÒ®];„††"-- yyyhݺµn»Ö¹sçðÜsÏaÀ€èÛ·/’““k[õ/ÿ¬ÖÕ>kCÜŨ sˆ»pô¾sæØÿ|öI""""ù`@—¬^½ÚèñŒŒ Ýí‘#Gš}~BB‚ÁuìBš:8pøã 0xüq§¼ Y Å z×®À… U÷[·¶­òKÆúôöísüù=$\[ˆˆˆˆH\ èdÑ_ÿ Œ Ü¿>>b·ÈýHq™µ €Aƒ€{÷4܉YAwÔÂ…@ïÞÀÍ›ö?7, ˆˆ¾MDD¶HMMEqq1|}}1f̱›C$ì[î¬R*5ÿ‘8¤¸ÌZ“&'†—0nœØ­ "²_ii)JJJŒ–‹%¢šaßro èD'Å :QTT”ØM ’%ö-÷ƯóDgÏ2kµUA'"""""᱂N$qöLÇ :Õ¶K—.a÷îÝðööFAA&L˜€Î;£¼¼óçχŸŸJKK±dÉ×$""²‚_ç‰$ΞeÖXA'¢ÚöꫯbüøñˆÅc=†¸¸8@RR:v숤¤$´jÕ ‹/¹¥$4•J…üü|¨T*±›B$+ì[ît"‰cˆ¤ìСCº/‘5•+Wp÷î]$''ãÆ4K…†‡‡#!!~\“Q66nÜ•J¥R‰xÎJ$ö-÷Æ€N$q¬ ‘”M˜0O=õþûßÿbåÊ•xçwpúôi¢Aƒ€àà`äççãìÙ³èÓ§È-&¡DEE¡¼¼uêðë$‘Ø·ÜëmDÇ :IÙš5kо}{tëÖ ={öÄ£>Š¬¬,èöñòòB“&M-bKIhmÛ¶E‡жm[±›B$+ì[î§eˆ$Žt"’²ââb„††¢N:ˆ‹‹Chh(<<<Œ&„S©Tð÷÷·x¬¸¸84lØÐà±èèhDGG Þn""w“’’‚””ƒÇîÝ»'RkÈt"‰³¶Ì+èD$¦É“'cõêÕ Áĉ1yòdìÙ³óæÍÓíSPP€ÂÂBtíÚÕâ±V¯^^½z9»ÉDDnÉÔ ÏS§N!<<\¤‘) èDgmˆ;+èD$–ÂÂB|÷Ýwèر#à“O>App0êÖ­‹ÊÊJäææ"((999P*•hÓ¦È-&!¥§§£¤¤>>>zè!deeÁÓÓ7oÞÄܹsqêÔ)”––¢wïÞ8zô(6l(Ò;"!±‚NDDDDäXoÙ;wœœŒ'žx0räH¬Y³EEEûíܹýúõ4hÐ8xðàÿ³wïqQ•ùÀ?ÃU.£ˆã50P,/¦’b²¢»˜¦„eæ%SS£ŒE-Õ²Ä2Ö0õej`Ûš¦þ𶘋ÀTDÃdiˆ2\ÇùýÁ:5(:Ì3œù¼_/_yÎœ9ç3óßóœóÀ‡~ˆÀÀ@Èd2ØÚÚB©TâÓO?5î¡&Ãt"""""óÀ_çËÊÊ‚B¡@«V­...())Á¹sçô¶»qãNž<©[îر#þõ¯NŸ>­w¹»««+Ž=j„ôd A'""S”€ØØX$$$ˆŽB$)ì[æº`ÙÙÙpvvÖ-[[[£]»vÈËËÓÛnâĉعs'  V«qîÜ9ܹspþüy½}tîܹÎû©ùâcÖˆˆÈùøøÀ××>>>¢£I û–yã=è‚Éd²:©T*888è­[¸p!œ1sæL<ÿüó8qâ>üðÃz÷QRRRçýÔ|ñ1kDDdŠîÞzGD†Å¾eÞX  æç燈ˆÝrii)ÊÊÊàíí­·\.Ç‚ )))(((@@@€n—.]Òm›ŸŸž={6xÜùóç×™D.44¡¡¡ôyÈðs‰;GЉLCbb"õÖ JCDDDÍ tÁzõê…šš mÛ¶ÈÏχ\.‡››Û=ßóÉ'Ÿ 22R·ŸŸrrrt¯ççç£OŸ> 7..¾¾¾†ùÔ¤8‚NÔüÔwÂ333J¥RP""""j8Þ&˜¥¥%BBBpüøq@rr2¢££+W®Ô+¼ÚG²µoß^· Œ=gÏž”——ãÊ•+xõÕWô ¨©qˆˆLQNNÎ;Wçw"z4ì[æ#è& >>‹-™3g Ñhðî»ï’’’Э[7xzzbÏž=¸uë\]]u¯ßååå…yóæañâŨ©©ÁªU«xº„pˆˆLQrr2T*är9ÂÃÃEÇ!’ ö-óÆÝ888 ..®ÎúôôtÝ߃‚‚ÜǸqã0nÜ8ƒg#ñ8‚NDD¦höìÙ¢#Iû–ycNdâî÷˜5Ž ‘¶¶¶¢#Iû–ycNdâþ8*~þ<кµþ뜚#èDDDDDÍ t"÷ÇQñû݆Ät""""¢æ‹:‘‰óô|°í:uìí›6 Ñ]‰‰‰(//‡½½}Ç ÑÃcß2o,ЉLÜرÀºu@VÖ½·±±BCKKãå"""ó¦P(PQQ;;;ÑQˆ$…}˼±@'2qÖÖÀ´i¢Sé H’ط̧”"""""""2,ЉˆˆˆˆˆˆL/q'"""¢F»~ý:ª««aii‰N:‰ŽC$ì[æ:5ZRRT*är9Âï÷P"z`ì[æ:5ÚK/½„ššXXðŽI"Cbß2o,Љˆˆˆ¨Ñ …èD’ľeÞxZ†ˆˆˆˆˆˆÈ°@'""""""2¼Äˆˆˆˆ-99°³³Cpp°è8D’Á¾eÞX Q£©ÕjTUUÁÒÒRt"Iaß2o,Љˆˆˆ¨ÑÆŽ+:‘$±o™7ÞƒNDDDDDDdX ™^âNDDDD¦R© Õj!“É —ËEÇ!’ ö-óƈˆˆˆmݺuP©TËå7Ø~/^¼•Je°ýÀùóç º?¢¦ÔT}‹šèDDDUVVÑ1H¢ÆŽ F++Ãý:yñâExzzluq4’L_Sô-j>ø¯NDD$Q¯¼ò Ö¬Yƒ¶mÛŠŽBäîînð}þ>r¾€—÷.àaà}^Sô-j>X I”\.Ç’%K ÑhðÊ+¯ÀÏÏOt$¢äÀWt""£cNDD$Qk×®…••Š‹‹±qãF|üñÇ>|8^xá´hÑBt<"""ú>fèEG0Yl›ú±]îmcîÞ¿èää???âÕW_Åøñãñ÷¿ÿ·nÝœš³Ã‡cß¾}8|ø°è(D’¾eÞX ›Fƒ `É’%ˆŒŒDuuu½ÛmܸqqqˆE||¼Þkeeepvv†……,,,°cÇcD—÷ƶ©ÛåÞØ6¦aÛ¶mØ¿?†Ž¡C‡¢K—.¸pá¾ýö[Ìš5 ,À÷ß/:&5Syyy¸|ù2òòòDG!’ö-óÆKÜM@TT<<<0cÆ ÄÇÇcéÒ¥ˆŠŠÒÛæüùóØ¿?¶nÝ ?~<”J% Ø°aÖ­[‡Ö­[†jÔÏ@DD¦'44¶¶¶xå•W°aøººê^suuň#0oÞnß¾9sæ¨}7++ @í,îûÛß°zõjlÙ²-Z´Ðè.\À!C0vìX9r111Â> ™___|ûí·Ðjµ8räFÌÌLüå/ˆˆˆêÁKÜM€ƒƒâââê¬OOO×ý]&“ÝóÒôˆˆ½ûØïE£Ñ 22öööP«ÕX¶lY ê¤lÏž=xï½÷ðÕW_én!h¨MÌ¡½8€>øׯ_Ç«¯¾Š¹sçB&“™}»@ff&"""——‡¹sçêNœ±mj©Õjôë×ü1† ÂvùŸ÷ß‹/<ñÄ8uê”ð¶quuÅ’%K°dÉݺ .àñÇ7èqÈüddd@­VÃÆƆ“Bû–y㺉ŠŠ‚‡‡¢¢¢Ð¥K,]ºTt$£)((À;wpòäIÈd2Ýú†ÚDêíUUU…-[¶`ñâŘúyyyºþÄv©ýÞ\»v û÷ïÇþýûu»Ù6wîÜÁ—_~‰åË—céÒ¥Xºt)ÞyçL:Õ Ç!ó”––†Ã‡#--Mt"Iaß2sZ2 ………Z;;;mqq±V«Õj¯\¹¢mݺµ¶¬¬Lp2ãQ«ÕZ™L¦ÍËËÓjµ ·‰9´×É“'µ*•J·ü /h'L˜ -**2ëvÑjµÚÊÊJmMMnÙËËK{âÄ ³ÿÎÜõý÷ßk7nܨuww×þç?ÿa»üϺuë´ëÖ­Ó['ºmF¥uppÐöë×O;tèPíСCµƒÖ¶iÓÆ`Ç0„ŒŒ -mFF†è($ØÝï¡´Íài}w;wíܹ³QŽ×tÿ^¦Õ®RÆŸ¿¦‡#èf"++ …­Zµ¸¸¸ ¤¤çÎœÌxþü¨Š†ÚÄÚ«oß¾pttÔ-?þøãpwwÇÙ³gͺ]ÀÖÖV72|ûöm<ýôÓèׯŸÙg€ÚGAnß¾]ï l—Z ˜1c† †#GŽß6§OŸFNNNœ8C‡áСCøÏþƒ/¾øÂ`Ç """Ãan&²³³áìì¬[¶¶¶F»ví——'0•X÷j“ÜÜ\³l¯]»vaêÔ©l—?Ø¿?†h4¶ €+V`áÂ…zëØ.µNœ8ÜÜ\tïÞC† AVV–ð¶yñÅuOùø#ƒƒˆˆˆ ‡“Ä™ ™LVgâ!•JA‰Ä»W›8::â·ß~3«öÊÈÈÀ¸qãàáჲ]þÇÇdzgÏÆ‚ àää„.]º˜uÛìÙ³ýû÷×{†¶V«e_úWWW$$$ÀÞÞëׯ‡———жñðð@ll,~þùgÝ:Fƒäädlß¾Ý`Ç!"""ÃນðóóÃåË—uË¥¥¥(++ƒ···ÀTb5Ô& 0›öR«Õغu«î¹Èþþþl—ÿiß¾=^~ùeÄÄÄ %%Åì¿3±±±˜4iZ·nÖ­[㧟~Bpp0òóóͺ]ê3mÚ4¤¦¦ ÿÎ|öÙgHMMÅ¢E‹t3¹/Z´;vì0Ø1È|%$$ 66 ¢£I û–y㺙èÕ«jjjPPP€¶mÛ"??r¹\÷¸1sÔP›¸¸¸˜M{­\¹ .„…Eíùºž={²]þ¤gÏžèׯŸÙg¶lÙ‚ªª*µ#çƒ BLL †Ž•+Wšm»Ü‹¯¯¯ðïÌ믿ŽñãÇÃÆÆFoýæÍ› v 2_>>>¨ªª‚­­­è(D’¾eÞ8‚n&,--‚ãÇ’““-8•qݹs@íå@Ãmb.íõÎ;ï ÿþ(--Å¥K—°~ýzäææš}»TUU!33S·¼mÛ6„……Á¬ۦ}ûöpuu…««+ÜÜÜ`ee…:ÀÙÙÙ¬Û ±iÓ&h4¨T*|òÉ'X¼x±ðïLXX´Z-Nž< 8sæ ´Z-^|ñEƒ‡Ì“¿¿?† ÑQˆ$…}˼ɴZ­Vt2Ž²²2,Z´íÚµƒF£Á»ï¾+:’Ñ”••aýúõxã7°páBÌ›7 …¢Á6‘z{EEE!**Jo··7²²²Ìº]€ÚIÏÆŒ777x{{côèÑ1b€†?¿9´ÍuíÚ›6mÂàÁƒÍ¾]òóó„òòrøùù!""^^^Ä~göîÝ‹ÐÐPøûûã»ï¾Caa!-Z„7Þx=zô0è±Eff&”J%222àëë+: t÷»dhß…L¦óÝuqqÁ/¿ü‚Î;ëÍ=ÑTšîßË´ÚUÊøó×ô°@'""’¨‰'âoû._¾ŒE‹¨}bÃŽ;°aÃÁé~Ç_é.è†:5þš^âNDD$QJ¥'N„£££n]FF8 0IENNÎ;‡œœÑQˆ$…}˼q’8"""‰òôôDYY™nùرcˆÇk¯½fðc]¸p»wïF÷îÝ1hÐ 899!22öööP«ÕX¶lYÇÊQ󖜜 •J¹\ŽððpÑqˆ$ƒ}˼±@'""’¨ÁƒãÍ7ßÄéÓ§±jÕ*ܸq#FŒÀìÙ³ zœ­[·âÒ¥KX²d‰nÝ;ï¼̘1ñññXºtiy/¨y3ô÷ˆˆj±o™7èDDDÕºuk¬]»eeeHMMEÿþýѦMƒãÇDbb"’““u늊Šƒ7n‚‚‚ T*{{{ƒŸÄá# ˆšû–yã=èDDDçàà€   ´iÓåååØ´i“Áö=mÚ4ôë×QQQ˜1c®]»†¬¬,( ´jÕ @íÄU%%%8wîœÁŽKDD$E,Љˆˆ$ªk×®z\]]Ñ©S'¬]»Ö û/..FZZÆŒƒ¿ÿýïÐjµP*•ÈÊÊ‚³³³n;kkk´k×yyy9.‘Tñw"""‰òðð@hh(d2™nÝîÝ»1jÔ(ƒì?==r¹}úô,_¾ëׯ‡••U áT*Üßüùóáä䤷.44¡¡¡ÉK†•˜˜ˆòòrØÛÛó߈Ȁšªo%&&"11Qo]qq±ÁöO†ÁˆˆH¢Ö¬YOOO½uO=õ6nÜhýWUUé-·iÓnnn¨¬¬ÄåË—uëKKKQVVooï÷Ççð6# …°³³…HRšªoÕwÂeit IDATó÷gÙ“©`NDD$Q.ÎÀÆÆŸþ9>øàƒGÞïÞ½qûömܺu …@íäFÇÇâÅ‹QPP€¶mÛ"??r¹nnn|L2¢#Iû–ycNDD$QzË•••8uêTõ«K—. úuë‰_~ùVVVðööFHHŽ?Žàà`$''#::Ú Ç$""’2èDDDUVV†   Ý=è666xíµ×ðüóÏìŸ|ò –/_Ž•+WâÒ¥Køî»ïñññX´hÎœ9Fƒwß}×`Ç$""’*èDDDõ駟¢_¿~ nóË/¿ sçÎ} ''',_¾¼ÎzÄÅÅ=ô~Éô]¿~ÕÕÕ°´´D§NDÇ!’ ö-óƈˆH¢ÊËË‘ššzÏ×5 ’’’`ÄT$IIIP©TËå‡H2Ø·Ì t"""‰Z¶l8Ðà62™Œ:=”—^z 555°°°…HRØ·Ì t"""‰²··ÇÅ‹ѽ{wݺ/¿üJ¥=zô@uu5Ö¬Y#0!5gwgî'"Ãbß2o<-CDD$QžžžzÅ9øûûcÊ”)KKKÌœ9SD4"""ª t"""‰ºví*++õÖmÚ´ 7oÞÔ-ÛÚÚ;Ý/q'""’¨ÈÈHôîÝÆ C×®]±{÷n=zŸ~ú©èh$ÉÉɨ¨¨€‚ƒƒEÇ!’ ö-óÆt"""‰zâ‰'°{÷n”––"11nnnصk¦OŸ.:I€Z­FUUÔjµè(D’¾eÞ8‚NDD$aøì³Ï`ooòòrØÛÛ‹ŽD1vìXш$‰}˼qˆˆH¢ÎŸ?‘#Gb„ €¬¬,,[¶ ÅÅÅ‚“Q}X IÔŠ+àîîOOO@ÿþýáêꊘ˜ÁɈˆˆ¨>,Љˆˆ$J¡P`íÚµpuuÕ­S«ÕØ´i“ÀT$*• %%%P©T¢£I û–yã=èDDDåïﯷ\UU…ØØX >\P"’’uëÖA¥RA.—#<<\t"É`ß2o,Љˆˆ$ÊÇÇ/½ôJKK‘‘‘]»vÁÊÊ o¼ñ†èh$cÇŽ…F£•$2$ö-óÆu"""‰zì±Ç°iÓ&>|{÷îÅæÍ›1räHX[[‹ŽFàîî.:‘$±o™7èDDD5yòdŒ7£GF@@€è8DDDt,ÐÍÌ­[·°oß>¸»»ÃÎÎNt""³QQQÜÜ\<ûì³P(F9æ?ü€yóæÕYÿ믿¢C‡FÉ@DDDŽº™Ù·o&Mš$:‘ÙÚ²e ÂÂÂŒr¬U«VaïÞ½pvv†…Eíƒ[*++‡øøx£d é:|ø0ªªª`kk‹¡C‡ŠŽC$ì[溙éÚµ+€Ú_½¼¼§‘¦ùóç#..Nt Ébû6-¶oÓ9þ<&Mš¤û9l Ó§OÇÕ«W±xñb½õ2™Œ:=²¼¼<”••ÁÁÁAt"Iaß2o,ÐÍL‹-^^^ðõõœFšœœœØ¶Mˆí۴ؾMïîÏac˜:u* „Î;ÃÒÒ Õj±víZ£e éš666õ¾^VV†O?ýÔÈ©ˆ¨1.^¼•JÕè÷ݹsG÷ßÌÌLCǪãüùóM~ "sȈHB"##ïû,{…Ba¤4DÔX/^„§§ç#íãæÍ›P*•Jô ød"CaNDD$!÷+ÎÚG±=ªŒŒ ¨ÕjØØع”¶ßGηðj仃ÜÐÀCÆj€€‡‘ŽeØ·Ì t""""j´´´4ÝLÓ,"š‚€ûŸpÓgý‡ÿ6ö½d*Ø·Ì ôf`Ïž=xï½÷ðÕW_ÁÍÍ­Þm8€ï¾ûÕÕÕ?~<hä”DDDdNæÌ™#:‘$±o™7è&®  wîÜÁÉ“'!“ÉêÝæ×_Å[o½…ÌÌL¨ÕjôíÛGŽ“““‘ÓÑò€Ö¶m[5¸Í‡~ˆÀÀ@Èd2ØÚÚB©Tr†^"""""¢f†z3`mmÝàëgÎœ—×¸ººâèÑ£M‹ˆˆˆˆˆˆ ˆ—¸K@vv6œuË;wÆ¿ÿýo‰ˆˆˆHêPZZ GGG̘1Ct"É`ß2o,Ð%@&“ÁÒÒR·\RR‡ß3þü:÷¨‡††"44´I2=ªC‡á™gžPûœç÷ß_p"¢{KLLDbb¢ÞºââbAiˆš†ªªª`kk+: ‘¤°o™7èàçç‡K—.é–óóóѳgÏß÷@ÏÊ%"¢Æ«ï„gff&—C’âïï/:‘$±o™7Þƒ.~~~ÈÉÉÑ-çççÃÇÇG`"""""""j,Ž 7wîÜh4ݺ•+Wb̘1ðôôÄèÑ£u÷§”——ãÊ•+X»v­¬DDDDDæäâÅ‹P©T߯\.‡‡‡‡Á÷K¦º‰+++Ãúõë!“É°qãFÌ›7 …IIIèÖ­<==áåå…yóæañâŨ©©ÁªU«î{º1}:lllðôÓO1‘é›={¶èD&Ä €a&`^³¦'d² ÕFXX˜AöIÍ tj±±±°¶¶Öçàææ†îÝ» L¥ÏÍÍ ÖÖÖ nóÁ $$‡Ö{”‘¹ã# ˆš†Zm À€Zt€³¸“Á•––"** “'OÖ[ïáá//Ã\úc’G¡P S§Nøø㔊ˆˆˆˆˆÌ tz`}úô……,,,ðä“OêÖ¿ýöÛ°²²‚‹‹ 11666Ð{¿···ÞˆºJ¥ÂÌ™31|øp¸ººbîܹP«?SxöìYÌ;Ï<ó |||°dɽý:u “&M“O> ÝùË_t9»wïŽòòrÀ—_~ 8::âÌ™3ðöö~ Æ Ã{ｧ7‹þŸ•——cÅŠxþùçÑ­[7„„„àúõëzÛäçç#,, Ä Aƒ°zõjhµZÝëZ­+V¬À„ еkWŒ1'OžÔ½ž––†©S§¢wïÞøñÇÑ¿ 8YYY˜?>Ú¶m‹›7obøðáðòòBqqñ}?™èôÀÒÓÓ™L†C‡éÖôÑGð÷÷Ç™3g[¶l››[÷÷ìÙSW×ÔÔ ýë_±ÿ~ìÛ·_|ñ>üðCµÅêĉÑ«W/;vìÐeuvv†R©Ôe€%K– ¦¦F7£}@@”J¥nŸiiiøàƒðî»ïÚ´iƒ¥K—¢²² €~ýúÁÇǘ5k¦OŸŽ¯¿þ>>>èß¿?€Ú“'NÄÎ;Ѻuëû~6""¢‡¥P(жm[Ýÿ'‰È0nÝR  @®û=—Ì tj”Áƒ£W¯^ˆ×­[¿~=^}õU@ee% бcÇ÷óý÷ß#//º?¹¹¹())A~~>¼¼¼pòäIôîÝ¥¥¥º%%%j?7lØ0Œ5 ÑÑÑ(++Ó¾µ£õ£FªU«P]] HJJz¨3‘mÛ¶Ðp¾uëV¼óÎ;jg‡/..Öe€ýû÷ëNFµ'Nž<‰ØØXÝëVVVxì±ÇtÛtíÚ=övìØ¡»¼ÞÊÊ ÎÎÎhÕª•Þñ­¬jç{twwoôç#""z3fŒîÄ7FJJ víz)))¢£,ЩÑfϞÇãüùó(//GQQ:wî  v‚8­V{ßÙѯ]»OOO:tH÷çìÙ³¸|ù2Ú·oxì±ÇðÖ[oáµ×^Ó]ÿÇ˶“““1{öl,Y²Ý»wÇþýûõŽÜÜ\$%%¨-¢_zé¥FÞ»#Ñ¿üòË=·éÞ½;öîÝ‹Ñ£GãÂ… xüñÇQSS£{ý§Ÿ~Ò»¿þÏN:FƒÂÂB½õ#FŒÀíÛ·qáÂÝ:^ºNDDDD$M,ЩÑ&Mš„V­ZáÓO?ERR^xáÝk … … î£}ûö8tèPç¢_¿~¿ýönܸ§Ÿ~®®®Ø¼y3üýýëìÃÆÆ+W®Ä÷߶mÛbäÈ‘z…ìÀ1pà@,_¾gÎœA—.]ꌽÁÏ£ÕjaooS§NÕÅ>}ú4úöí‹'žx¢Þ÷:99aÅŠøæ›oˆƒâå—_FË–-qðàA´oßݺuÃÑ£GѵkWÌž=ß|ó Þ~ûm´iÓF·Ÿýë_˜>}:&L˜€áÇcݺuضm:uêX»v-öíÛ§›ÅýÚµk€mÛ¶!)) 2™ sæÌAvvvƒŸ•ˆˆÈ^zé%¼öÚk5¿ ÝÛ¦M/áÓO‡ê=Ž—̇•èÔ|íر£Þõxî¹çpàÀÝ£ÏêãììŒ-[¶Ôûš““SçªÞ]¿ë—7äðáÃؼysƒÛtíÚ7oÞÄ|€.]ºè½öÅ_à­·ÞjðýsæÌÁœ9sôÖ½ñÆzËO>ù$RSSï¹+++¬X±+V¬¨÷õ™3gbæÌ™uÖO˜0&Lh0‘¡ññjDM£°P@ ð~›’qšÄªU«ðÙgŸ‰Žï¿ÿ:uªwdýÏT* Ñ­[7ݺôôttíÚ/¿ür¦$"""""â:5‘Ž;â…^À?ÿùO¼öÚkF=öéÓ§‚=zàæÍ›8xðà}ß“››‹´´4½Ñ뢢"|ôÑGøâ‹/š2.èÔ„^|ñEœ8q;vìhðRwCkݺ5,--QSSƒ„„„z´š»»;ÜÝÝuËÅÅÅX·nÖ­[÷Pf#""’ºäädTTTÀÎÎÁÁÁ¢ãIFpp2ì쮣¢‚ýʱ@§&Õ¿h4£ÓÕÕ/^|¤}8::"""Â@‰ˆˆˆ¤G­V£ªªÊ¬JÔllÔ°µÕ ºÚFt€:59+«æ÷5kŽ™‰ˆˆŒiìر¢#IÒöícd˜E‹‰ŽCFÆI∈ˆˆˆˆˆL t"""""""Àëx‰ˆˆˆ¨ÑT*´Z-d2är¹è8D’á訂…EjjEG!X Q£­[·*• r¹ááá¢ãIÆŒëв¥ %%3DG!X ‘Ézæ™gpìر{¾KKKäççÃÑ‘g™‰ˆŒiìرÐh4œX•ÈÀ¶o KË ¨®~ aaa¢ã‘ñ'*™¬ÊÊJTTTÜóõªª*#¦!"¢?rwwH’rsÝÈ„„à$qDDDDDDD&€#è&N£Ñ 22öööP«ÕX¶l,--ël·qãFܾ}555hÑ¢fÍš% -=,è&.** ˜1câãã±téRDEEémsþüyìß¿[·nŒ?J¥ ™ˆˆˆÌÀáÇQUU[[[ :Tt"É2ä0Z´¸†ÊÊ!¢£,ÐMXQQbbbpãÆ @PP”J%"""`oo¯ÛîàÁƒ(--Õ-·nÝW¯^eNDDDM&//eeeppp…HRÜÝóààp ee,ÐMXVV ZµjpqqAII Î;‡~ýúé¶9r$"##ƒ   TVVbìر¢b‘˜úyyyÉdl"""ÒÇÝ„ùùù!""B·\ZZŠ²²2x{{ëm—––†;v`óæÍ6l‚ƒƒ€—_~ùžûŽ‹‹ƒ¯¯oSE'"2kõðÌÌÌ„R©”¨i=z...hݺ5€ŸC…ˆˆˆôñ:hÖ«W/ÔÔÔ   ŸŸ¹\777½ívíÚ¥û¥oèСˆŽŽFJJŠÑó‘ù)--ÅöíÛõN 74‡ IGll,¢¢¢+: ‘¤„‡Çâ½÷¾Exx¸è($ tfii‰?~œœŒèèhÀÊ•+‘““>>¢#IÒÙ³>î8kÐýšãü(Í tçàà€¸¸¸:ëÓÓÓu—Éd 3f,¢&õÁà£>zàí½¼¼ðä“OâÔ©SM˜Šˆþ,66VïÿG%%%Æ‚ pùòeÝú{Í¡ògœ…ˆ¨é˜Ûü(Í/q'""¢‡²eËœ>}§OŸÆ?ü€N:aýúõ˜3gÎÍ¡BDDDú8‚NDDD¥}ûözËVVVèСœus¨ëÍ¡BÒ‘‘‘_~ùÕÕÕ;ùrþüyƒì‡¨9S*3`csj5G¶Í t"""2¸{Í¡BÒ‘ššŠ‚‚¨T*¬Y³ÆÀ{—xD͇¿äòÛP©üEG!X ‘A\½zU÷÷{Í¡BÒ1dÈÿÝ»ºÀîYÀÀû#j^Ö¬™ €S¦²oQsÀˆˆˆˆNîGDdœ$ŽˆˆˆˆˆˆÈ°@'""""""2¼Äˆˆˆˆ-55ááá(-MEB/q'2”3àèøJKgˆŽB°@'"""¢Fëܹ3vî܉ª*?ÑQˆ$åìYØÚ梪j§è($ t""""j´îÝ»ãðáÃbDG!’”´4¶ÒDG!x:™Œ´´4 6 III~ï¥K—0lØ°&x/‘qpˆLFaa!<øPï---ÅÁƒѳgO§"""""2èDDDDÔhùùùðöö†F“œÑiˆ¤ÃÓ3VVסÑxŠŽB°@'"""¢F;sæ žþy””œAllè8D’œŒ–-U() …`NDDDD€ÁƒC«M…HRÖ¬™ ™ìhµƒ&: t""""j4+++TUU¿N–Zm À€Zt€³¸™èDDDDDDD&€×$‘pwîÜÁÿýßÿáäÉ“¼¯ÜÜ\ìÞ½¾¾¾èر£ÒQ}Nœ8©S§¢¼ü}EÇ!’ŒÐÐDØÛ ¼3f :tè€qãÆ¡  sçΛˆˆˆÌÄéÓ§1~üxTTœFr²¯è8D’œ ;»ë¨¨…ຉ’Édu&„S©TpppÐ[—žžŽÂÚÚĸqãðïÿÛ˜Q‰ˆˆÈ UWWÃÖÖ66¼R‰ÈllÔ°µÕÀÆÆFt€#è&ÊÏϺåÒÒR”••ÁÛÛ[o»ªQ~yß IDATª*½e¥R‰ÄÄÄûîþüùprrÒ[ŠÐP>ΈèQ%&&ÖùY\\\,( QÓðõõÅôéÓ¼!: ‘¤lß>@&€ÙX´h‘è8dd,ÐMT¯^½PSSƒ‚‚´mÛùùùËåpssÓÛ®wïÞÈÉÉÑ-ÛÚÚ"00ð¾û‹‹ƒ¯//G#"j õðÌÌÌl’ ‰ˆˆH:x‰»‰²´´DHHŽ?HNNFtt4`åÊ•º¢<<<;vì€J¥>|Ï=÷œ˜ÐDDDDDDôÐ8‚nÂâãã±hÑ"œ9sï¾û. )) ݺuƒ§§'¼½½eË–¡eË–ðòò¸qã''"""©«¬¬DË–-QSS‰ÒRÑiˆ¤ÃÑQ ‹ ÔÔ8ŠŽB°@7aˆ‹‹«³>==]oy̘13fŒ±bÄÑ£G‘¢¢¢&=Ζ-[Ð¥KL˜0¡ICDdnŽ9‚ððp””AlìS¢ãIÆŒëв¥ %%3DG!X ‘_~ù%âãã›ü8 ,Àc=ƈÈÀúôéƒY³f¡ººé–™“íÛÇÂÒòª«_CXX˜è8dd,Љˆˆˆ¨Ñ ._¾ @!: ‘¤äæº(+6 ÁI∈ˆˆˆˆˆL t"""""""ÀK܉ˆˆˆ¨Ñ~üñG<û쳨¬üÿù¯è8D’1dÈa´hq ••CDG!X Q£¢{÷î(++Äþ#: ‘t¸»çÁÁáÊÊÜEG!X ‘Qeggc×®]ÈÌÌ4Ú1ûí7,_¾>>>5j”ÑŽKD$eO=õ^ýu¢£IʦM“dx sç·ŒŒ÷ ‘Q9s‘‘‘8vì˜ÑŽYXXˆÈÈHüûßÿ6Ú1‰ˆˆˆˆ‹:‘ `NDDDDDDdx:5ZJJ Þ{ï=””¤ 6–³¸Jxx,Z¶T¡¤$\t€#èDDDDÔh^^^صkRR¼DG!’”””@ìÚÕ)))¢£A'"""¢Fsqqùß9\DG!’”³g}ÜpVt€#èDdÅÅŘ={66lØ ,CZZfÏž´´4aˆˆˆˆˆî…#èDdeeeˆšáÂ… ¸páúöí ¡YˆˆˆˆˆþŒ:5Z^^üýý¡Vç!#ƒ“ÄŠR™›+P«•¢£,Љˆˆˆ¨Ñ®\¹‚¡C‡B¥º‚Œ Ñiˆ¤Ãß? rùm¨T¼Úϱ@'"""¢F €R©ÀêœÈÖ¬™ €S¦L‡ŒŒ“Ä™èDÔäNœ8cÇŽ‰Ž¡óã?"55¢£é°@'¢&7qâDŒ7Nt ?üC† Á/¿ü": ‘ïA'"""¢FKMMExx8JKS‘ÀY܉ eÆŒ8:þ†ÒÒ¢£,Љˆˆˆ¨Ñ:wîŒ;w¢ªÊOt"I9{Ö¶¶¹¨ªÚ): Àˆˆˆˆ­{÷î8|ø0€ÑQˆ$%-Í€-€4ÑQHèDÔdŽ=Š+W® ´´Tt”zíرîîî&u<™/èDÔdðÅ_ˆŽqO ,@ÇŽY ‘I`nÂ4 "##aooµZeË–ÁÒÒòžÛ'%%á³Ï>áC‡Œ˜’ˆˆˆÌQ~~>¼½½¡Ñä#'Gt"éðôÌ•Õuh4ž¢£|Ìš ‹ŠŠ‚‡‡¢¢¢Ð¥K,]ºôžÛ^¿~ÑÑÑÉdFLHDDDæêÌ™3xþùç|Ft"I NÆóÏg 88Xt€#è&ª¨¨111¸qã ((J¥°··¯³ýG}„Y³f᫯¾2vT"""2C]tŒR\\Œ‰'âùçŸgNSß ÏÌÌL(•JA‰ˆ ïĉ˜:u*ÊËO 1‘¿SJhh"ìí P^Î3sÄÝDõêÕ 555(((@Û¶m‘ŸŸ¹\777Ý6ýû÷ÇùóçuËÿú׿ðõ×_cÛ¶m""‘qttDAA**EG!’”[·°³«DEÅ-ÑQHÞƒn¢,--‚ãÇ’““ X¹r%rrr`kk WWWÝŸ6mÚ E‹puuˆˆˆÌ€··7víÚ…”ïûoLD,%%»v=””ÑQHè&,>>û÷ïÇûï¿Û·ocÎœ9€¤¤$deeÕÙ^&“ñ9è$Tnn.¼¼¼ðÎ;Òh{öì——¶nÝ*: ™)^ânÂWg}zzz½ÛOž<“'OnêXD÷¤V«qáÂÑ1ŠJ¥Â… ðÛo¿‰ŽBDDDDfŠ#èDDDôÐ8€áÇÃÛÛ«W¯†V«h4,X°K–,Add$ª««'%C+..†‹‹ :vä ˆ ©cÇëpq)BÇŽEG!X ÑC©ªªÂ–-[°xñbLž<‘‘‘HNNDEEÁÃÃQQQèÒ¥ –.]*8-ÚÉ“'1mÚ4„†ž…HRBC“0mÚQ>þØLñw"""z(gÏžÅ'Ÿ|GGG :gÎœÁÖ­[1hÐ ÄÄÄàÆ€   (•JDDDÀÞÞ^pj2Œ755_‹ŽB$)›6½ ‹s¨©yaaa¢ã‘qˆ âàÁƒHMMã‘eggcïÞ½())…ÈäõíÛŽŽ¿?bëñLJ»»;Ξ= …BV­Z\\\PRR‚sçΉŠJMàîcÖ ù˜5"C*,T  @ŽÂÂBÑQHèDd¡¡¡˜>}ºèìŸÿü'‚‚‚põêUÑQˆš]»vaêÔ©ÈÎΆ³³³n½µµ5Úµk‡¼¼<鈈ˆL/q'""¢G–‘‘qãÆÁÃÄ¥¥¥Þë*•  îcþüùprrÒ[Êû0‰ˆ 11‰‰‰z늋9É£©aNDDDD­VcëÖ­X¹r%€Ú{“#""t¯—––¢¬¬ ÞÞÞ î'..¾¾¾Mš• çôéÓ?~<**N#9™ÿnD†œ ;»ë¨¨6è~ë;á™™™ ¥RiÐãУá%îDDDôHV®\‰… ¢ö׊ž={¢¦¦€üü|Èår¸¹¹‰ŒIV]] [[[ØØðzD†dc£†­­666¢£,Љˆˆè¡½óÎ;èß¿?JKKqéÒ%¬_¿¹¹¹ ÁñãÇÉÉɈŽŽœ” Í××[¶lÁöí='2¤íÛÇbË?lß¾]t€:=’7Þx½zõ’ÜL£Ï=÷  :‘I‹ŠŠBtt4FŒnݺÁÓÓ«V­B÷îÝýû÷ãý÷ßÇíÛ·1gÎÑq‰ˆˆLïA'¢GríÚ5I>:éòå˸yó¦èD&mÉ’%X²dI½¯988 ..Îȉˆˆˆš7èDDDDÔh•••hÙ²%jj*QZ*: ‘t8:ª`aQšGÑQHèDDDDÔhGŽAxx8JJŽ 6ö)Ñqˆ$cÆŒuhÙR…’’¢£,Љè¡üúë¯ø駟PTT$:J“©®®Æ‰'ЪU+ôèÑCt""“Ò§OÌš5 ÕÕñ¢£IÊöícaiyÕÕ¯!,,Lt22èDôP¶mÛ†ùó狎ѤÊËË1`Àüå/ÁîÝ»EÇ!"2) …—/_ …HRrsÝÈ„„à,îDôÿíÝypuþÿñ×äB&\ ¢DA!° ì.D´Ä Q@A<JÅÕèú]Ž¢._nW«¢QTXŠÅxì²€,ÊñÀƒ•#r##Y(0B˜!ÉÌôïÌH ™äÓ3ó|TQÐ×çóéžÌ{º§€ P `\â V¶oß®n¸AEEE¦£4˜?þX­ZµÒ”)S4vìXÓqÀöîÝ«¨¤d¯>ÿ¼‡é8@Èè×ï3]pÁA•”ô3P ¨Ç£¼¼<Ó1Tii©òòòÂêC 8—cÇŽ©}ûö*,<¦Ï?7íÚå(.î¨ Û™Ž(ÐPkW_}µÆŒ#i‹é(@HY¼ø~I™’&rå^â;èØ:6@€ ðt56eÊüðC?~\úÓŸÔ®];ÓqÀ¨µk×jÆŒr¹ÖjÞ<îâÊ„ óï–Ë5Át@ ÆæÏŸÖw2_¿~½Ö¯_¯!C†P {;wÖÌ™3UV6ÝtãöìÙcËm!8­]{£¢£÷«¬lºFŽi:ºÍy<Mš4I±±±*--ÕìÙ³Ya“'OjÎœ9Zºt©úöí«¿üå/êÚµ«¡Ä $%%)33SR’é(9%I£Fª·m#üìØÑMR™¤¦£À t››9s¦’““•ššª hÖ¬Yš9sf…uÞ~ûmµmÛVo¼ñ†fÏž­x@›7o6” \$KÊ’äðv?o@¸¡@·±üü|Í;WGŽ‘$ 8P)))zꩧë_¯C‡ºîºë$ImÛ¶U§N´wï^uêÔÉHn€ðA! p(ÐmlçÎJHHPÓ¦M%¾”Ìåri×®]êÙ³§½òâ\’.½ôR9N%%…óåf´}ûöÉårÉçó™Žb ß}÷.¸à¥¤¤˜ŽÆäää¨OŸ>*-ÍÑ–-Ü$”””-Š‰ÉVi)ï3ºíÞ½[-Z´ðOGGG«U«VÊÉÉ©P ŸiåÊ•>|¸âââ*&ÂÀ˜1c´fÍÓ1l㡇’Ãáà a-;;[×^{­ÜîlmÙb: :úôÙ §ó„Üî>¦£À ts8•nçv»«-¾?þøcÍ›7ïœÛ?~¼š5kVaÞˆ#4bĈó ðËÈÈPFFF…y†Òõ£ÿþ?_IDuÒK/ýYR¦¤=Út40 tëÝ»·žzê)ÿôÉ“'UXX¨.]ºœuý•+Wê¶Ûnó__´´4õèÁåhPÎögff&_‹ÕŠ0UëÚµ«|>Ÿòòò$I¹¹¹r:jÛ¶m¥uwîÜ©ÜÜ\ÝrË- è6©¡C‡ê›o¾‘$­ZµJsæÌ‘$=ÿüóÊÊÊ’$íÝ»W‹-ÒM7ݤèÛo¿Õ«¯¾j,7 ö¸ÄÝæ,X É“'kûöíòx<š6mš$éwÞÑå—_.§Ó©ë¯¿^‡Vzzº¤Óß]çwLÆFˆxùå—5oÞ<ÿ£þð ˲Ծ}{]uÕU•¾k á`ýúõš0a‚Nž\¯… ùÚ(©© Õ¤Éq<™j:  @·¹¸¸8¥¥¥Uš¿yófÿ¿:Ô‘F ”m:†meggóHCa«M›6Z±b…Nêm: Rvìè¦FèÔ©¦£À t€ó°oß>¹Ýî€o×ét*999àÛ ´öíÛë³Ï>“4×t ¤lØÐGR#ILGèµ´oß>uìر޶Ÿ••E: °(ÐT²ÿ~¥§§kÓ¦M¦£ØÞþýû5nÜ8ÝtÓMºõÖ[MÇÐ@~9sþwI¸å=’FÕË™y€ýQ ¨äðáÃzñÅMÇ å}åt:)аÔYRxÞ -77W]ºt‘Ç“«Ÿ, :vÌRTÔay<õw•ì‹µ¶}ûvÝu×]r¹¶kÞ¼¦ã!cРUŠwËåd:  @@­õïß_}ûö•e­7)/½ô?r8¶Ê²újäÈ‘¦ã Q  Ö¢¢¢têÔ)ñv¬ÒÒF’¢%•šŽ"L`–eéæ›oÖĉMG :ºù曵eËÓQ¤(ÐÔeY¦#À¦Ö¬YÃÝÛÏCvv¶Ö¬Y£cÇŽ™Ž²,‹×]pN\“¦–/_®””Ó1 ,Œ‘¡ØØ<0p=L­_¿^³fÍâlÔ³^xAiii¦cפIåååéèÑ&¦£!åèÑåå9uôèQÓQ`gÐÃÔúõëµ~ýz:uJmÛ¶UjjªéHR>úè#}÷Ýwš5k–Ž?n:p]ºtÑÊ•+%M7)k×Þ()SÒƒzöÙgMÇA£@sÏ<óŒš7o®«®ºJM›6Õe—]f: 9qâ„ÊÊÊLÇz.—KÇŽS‹-äp8LÇGÕ¡C‡”žž®Õ«W›Ž‚—¸CÇW÷îÝ5~üx¹Ýn•–òH‡ptë­·*11ÑtŒ wçw*!!A%%%¦£ y½^¹Ýn½ýöÛêÞ½;Å9¨5 tø­\¹Rñññzå•WLG€ “••¥øøx7Ît A())I]T`: R.ºè°’’òuÑE™Ž¸Ä•|ûí·zï½÷tã7ªyóæ¦ã€í-_¾\ÙÙÙ¦c jÓ¦Mz衇ärmÒ¼y×™Ž„Œ#ÞQ|¼[.wqGè¨dñâÅZ¼x±¶lÙBÜn·ŠŠŠøþy€åææÊétªeË–¦£ <ôÐC<a§OŸ>>|¸|¾˜Ž„”Å‹ïSDÄ.ù|whäÈ‘¦ã q‰;ª4cÆ =òÈ#¦c žýõ¯UëÖ­µqãFÓQBÊe—]¦˜Žz–žž®Ñ£GëäÉ“¦£ ®ü1kÇŽñ˜5 Ž;ý˜5>ø OœAG•>øàÅÆÆêµ×^3lé“O>ÑòåËMÇ!‚3èØgÐQ­ââbµoß^}ûöÕ›o¾i:¨¤¤Dùùù\š[JKKuøðaÅÅÅ©iÓ¦¦ã €^ýu=óÌ3úñÇMGŒÙ¶m›î¼óNoÓªU=LÇAÚ³gOÀ·ét:•œœðíÒ A«Ô¸ña2P £Z–e);;Ûö/d¨½O?ýT·Ür‹é!mÇŽjӦƯùó盎ƒ:qâwmGØóz½jÔ¨‘¼^¯é(9NIÒ¨Q£êeëYYY¶~oSªF<òzcLG訵=zèᇖôÿLGAÈI–”%Éàíî‘4Jnw ·XË– “”)é4yòdÓqÐÀ(ÐQ#Ÿ|ò‰Z¶l©É“'k„ ¦ã€™™™ºñÆU\\l: „8ûžáê7‰C”••)??Ÿ7¥!ÀëõjëÖ­úþûïMG yyyÚºu« LGAy½^^ @½á ºÍy<Mš4I±±±*--ÕìÙ³Yi½uëÖiõêÕòz½ºóÎ;uÍ5×ÔKž•+WêСCš0a‚­¿»ƒª«{÷î¦c„•¥K—jéÒ¥úç?ÿ©aÆ™Žƒó4fÌ>|Øt À6JJJ/Ÿ¯DÜoœ&MÜŠˆ(–Ï×Ät@ns3gÎTrr²RSSµ`ÁÍš5K3gά°Î?þ¨‰'*33S¥¥¥ºòÊ+õÅ_¨Y³fϳqãFmܸQ#FŒ @V-Z¤S§N™ŽØÆ_|¡ &ÈåúBóæ]m:2RS)>Þ-—+ÕtÀ%î6–ŸŸ¯¹sçêî»ï–$ 8Pééé***ª°ÞsÏ=§o¼Q‡C5RJJŠ^~ùåzÍöí·ßê³Ï>“eYõúsÀ´mÛ¶é³Ï>“Ïç3°•îÝ»kÉ’%Z¶Œ«²€@Z¶l˜–,é¥eË–™Ž8ƒnc;wîTBB‚ÿùÉIIIr¹\Úµk—zöìé_oûöíCq饗ꫯ¾ª×lãÇ—$Þ°™ 6觟~2#lmݺUMš4ѵ×^«˜,&Nœ¨?þØt Àv~¾ŸI‚é(@H9p ¤|IÌœA·±Ý»w«E‹þéèèhµjÕJ999ծצM›JëÔ—!C†èÉ'ŸlŸ…º{ì±Çtûí·›Ž¶ž~úi 0@'Nœ05‘‘¡ÁƒkëÖ­¦£€0Áts8•nçv»Wíz.—«Ò:¿±F–u…"""åóyá¿\ÝápÈçó)22RW‘‘•—EDDÊëõ裢µn]”–,ñÈçóÉë=½-¯×«ÈÈHù|>9I’eY•–•ÿ\Ë’""òz}ŠŒŒðÿíóY:ýß²,_­¶}þË"$g¦_ÆçlÛvÈíþ?EEyýãZu[¢äóy«ø¹Uï3ÕíOuÝ×Î7Sdd¤¿ TÞ_2mì"ÏèûÊÿßáp¨K§,«8¤÷µúÉ[1Ó™¯!ǧn}%‡Ã¡S§ëäÉë~^¿îûš×»UÒÀª^–(Ðí¬wïÞzꩧüÓ'OžTaa¡ºtéRi½ýû÷û§sssuÅWT»mŸo¶¤fòzOOŸþ{ÄÏNóxÎ\VQù<çôž8,ýÿ*Ã3o#Pq¨~YuûEm—Õd_;ßLµÙöù.«m¦£G+σEK:}ÝÚïk?ÿ9sÙChÙ»w¯  ’’½úüó¦ã!£_¿ÏtÁURÒÏt@nc]»v•ÏçS^^ž•››+§Ó©¶mÛVX¯wïÞÊÊÊòOçææžó1Z‰‰iŠ‰ Ü/S·Û%—Ë°í!𢣣¥¦M›*2’Cß”ãÇËç󪤤ÄtT#&&F‰‰­ê°…ŠxJRii¦òòRê” 8_ûöí“Ûíè6<¨öíÛ«°ð˜>ÿ< ›ÂZ»v9Š‹;ªÂÂv¦£ÀÞ¥ÛXdd¤†ªo¾ùFƒ ÒªU«4gÎIÒóÏ?¯!C†¨cÇŽ}zà6ˆ€{öÙ¹š0a‚éPsíܹSݺu3ÕèÙó}ùå—Ýff¦”B}öíÛ§Ž;ÖãOÈ:÷*jlñâû%eJš¨±cÇšŽƒFns ,ÐäÉ“µ}ûvy<M›6M’ôÎ;ïèòË/WÇŽÕ¹sg7NS¦L‘ÏçÓüùóÏùô@0`€š5k¦ hïÞ½ ú³ ÒÒÒtñÅ›ŽÌ/gÎÿ.©s€·î””àm@ø¢@·¹¸¸8¥¥¥Uš¿yóæ ÓÇ×ðáÃ*V%½zõR¯^½ôÁP jcÇŽõß<-%ñ]q°3 tÔèÑ£Õ¿MŸ>]žò;&@xøá‡Õ®];Ó1@£@G@Ý{ï½’N_òþã?jÙ²e:|ø°¶nÝ*—Ë¥¢¢"à ÃÏþð=ûì³6rùå—뫯¾Ò¿þõ/=ÿüó¦ã„‡Ã¡6mÚ¨Y³fúÃþ nݺ©gÏžêÔ©“œN§éx@И0ažâãÝr¹œš7{œò˱ÅqŽ(ÐQ/R~¾Ò­·ÞêŸ7tèP-_¾ÜT¤°Õ²eK]}õÕ¦cà ±±±ºúꫵmÛ6ÓQÂRãÆuðàAÓ1€ ·v튎.SYY´é(@H9}líWYÙt9Òt40 t4˜©S§*55U÷ÜsdkMš4Ñ¢E‹Ô¦MÓQP…n¸AJOO××_m:NXxñÅÕ©S'Ó1€°cO£êÃéc«LÒÓQ`:LŸŸëÍ'í ¡Q£Fºçž{LÇ@5’““•œœ¬•+WR 7k®¹ÆÿZ`7èhpíÚµSãÆuèÐ!ÓQ„‰–-[ªiÓ¦Š‰‰1 Jèhp›7oÖž={Ô¥KÓQ„‰©S§jܸq¦c!%%e‹bbJUZ£-[RLÇBÆéc+[¥¥Wá(Ât÷á‡jãƦcÔ‹ŒŒ ÓnÞ¼yÚ·oŸ.¼ðBÓQÔ»Ðy ëÓgƒ®½ö3õ鳡Š5B§­çF[CÉ÷§­,õéÓÇX˜ÃtѼys7N;vìÐ'Ÿ|b:NÈHIIQûöíÕµkW%%%™ŽS/2224bÄÓ1ªuëÖ’N?éà¿ÿý¯V¯^m8QèèСƒn½õVýþ÷¿7%,y<Mš4I±±±*--ÕìÙ³RkµIDATi:–a’Bã5쥗þ|Ž5B§­çF[CÉ÷§­LI)=z´‘ 0‡3è0¢uëÖJKKÓÝwßm:JHIMMջﾲÅy¨{å•WôÒK/™ŽR~÷»ß)--Mýúõ3%,Íœ9SÉÉÉš9s¦.¹äÍš5Ët$lÆ Ó‘#G4jÔ(ÓQPG—\r‰Ž9¢_|Ñt Nòóó5wî\ÿ±TzzºŠŠŠ 'À¾(ÐaTÛ¶muÛm·qÆ·Ž7n¬Ö­[+66ÖtÔQTT”Z·n­¦M›šŽÔbbbtÛm·©gÏž¦£„­;w*!!Á¿/'%%Éåri×®]†“`_|=Ì”””H’öìÙc8Éi‰‰‰š6mš¦L™Âc×ê ??_™™™¦c4ˆ‚‚‚°hëLGjqqqš6mš$Ùf)Ý-..6œ¤aìÞ½[-Z´ðOGGG«U«VÊÉÉ©ôÁ‰Ý~7ÕÄ/Yk›¹@§¿[Zå–%I«W¯Xüðö(Æ-SãÆE*.ŽÕ²eÃβƹÚJìÒÖÒ3þ®¯]“&M’×ë5)è=óÌ3þ}¶{÷î’èçºøðÃÕ«W/åääøçUןáÐ×5Rtt´N:¥âââ*Ûøæ›o*--MóæÍÓ‚ *,+,,T‹-üûêòåË"z­Ôd,½^¯þò—¿èÉ'ŸÔìÙ³+Ý™yݺuzüñÇ5~üx}õÕW ½ÖÑÖ`SéìÇô¯U7nÁ2¦RÝÛ,c*Õ¬­ÇŽÓ“O>©?ÿùÏ•–…Ú¸V×Ö`×b!$åääX .´G¥e£F²Ö¬YcY–e=ñÄÖo¼Q£e¨Ú?þh7ÎZ·nµnÝ:ëСCþeôi`L™2Åzíµ×,˲¬—_~Ùš6mšáDÁ­¤¤ÄzôÑGýûìþýû-Ë¢ŸÏ×O?ýd­X±Âr8VNNŽ~uý}]“6îÞ½Ûº÷Þ{ýÓÇ·þóŸÿø§_|ñEëÿø‡_õz½õ¼–jÒÎùóç[&L°,˲ ¬Ž;ZDz,Ë:räˆõûßÿÞòù|VII‰ÕµkWëøñã ×€Z¨k[-+8Æ´ªcúLÕ[0i]ÛjYÁ1¦–U³¶Z–eeffZC† ±F]a~¨«eUÝVË žq 5è!,++«R¾uëV«uëÖ–Ïç³,˲>ýôS+99ùœËP½É“'[k×®­4Ÿ> ŒcÇŽY7¶ ,˲¬ììl«yóæVaa¡ádÁkÑ¢EÖ¢E‹*Ì£Ÿë¦´´´Â¡êú3úº¦m|饗¬!C†ø§~øa+##ò,Ëòx<Ö-·Üb8q¢á‚×RMÛÙ§Oëý÷ß÷Oßwß}þoÇo=ñÄþe÷ß¿5{öìH_;hk0Œi¹_Ó¿Vݸ˘–«K[ƒiL-ëÜm-7yòdë¨0/ÔƵÜÙÚlãJ¸Ä=„EGGWš·mÛ6%''ËápH’.¹äíß¿_?ýôS•Ëòòò4w°ñxø`ƒ¶¡&jÚÎ{î¹G+V¬P^^žJKKµk×.•••I’öìÙSamÚ´±å¾ˆ¶ØÖÔ¯ûãÌq –1­©êÚJcz¦ò“+gª®‚ÙÙÚªã *\‚ 0uêTíܹó¬Ë\åA¡ÈÈHÿtù§aMš4©rY\\\ b­šôw‡´bÅ ÝqÇZ²d‰ÆGŸˆÃá¨Ð’äv»éÇ:ºôÒKµpáBÅÆÆêõ×_WçÎéçªj¿mÒ¤‰Ž?ò}]›ã6??_?þ¸æÏŸ¯Ñ£Gë7ÞðÿßØØXÝ|óÍÚ¼y³ºvíª£G*!!¡AÚP5mçÿþïÿªE‹zôÑGu×]wiãÆzî¹çκ —ËeË}!m•ì?¦5u¶qkÒ¤I•Ëì8¦5u®ö„ʘžK¨ë¹„Ë¸Ú zzúé§ÏëÿõîÝ[«W¯öOçææ***J;vT¯^½ª\îjÚ߇C>ø Þ|óM7®ÚþFÍõîÝ[O=õ”úäÉ“*,,T—.] ¦ =ôî¿ÿ~=ZO>ù¤>ý\7Õí·_|qÈ÷uMÛ 6hùòåZ²d‰®¿þz 4Hýû÷×ÿÿ«nÎîСCzì±Çôæ›oj×®]š3gŽýËéÓÀ¸é¦›ô·¿ýM™™™*..ö_>‰Ús»Ý3fŒ^xáíرC'NT‡$ÑÏç«°°P .Ôš5kät:õÛßþV±±±Õög8ôuUm;v¬:uê¤Î;«C‡*++Ó§Ÿ~ªýû÷+""B÷Üs"""´bÅ =òÈ#Ú°aƒ¼^¯fÍše¸EgW“v~øá‡úÏþ#§Ó©©S§V¸)Sbb¢âããõöÛoëóÏ?×øñãuÙe—™jNµêÚÖåË—ŘVuLŸÙÎêÆ-˜Æ´®m –ãTªY[%iÓ¦MZ¸p¡¾ÿþ{]qÅj×®¤ÐW©ê¶˱Š–eY¦Cî¸Ä @À(а tl€ @À(а tl€ @À(а tl€ @À(а tl€ @À(а tl€ @À(а tl€ @À(а tl€ @À(а tl€ @À(а tl€ @À(а tl€ @À(а tlàÿŒðwn$!0ÅIEND®B`‚HDDM-0.2/hddm/examples/wfpt('easy',).png0000664000102500010250000004110111726446356020257 0ustar wieckiwiecki00000000000000‰PNG  IHDR Xšv‚psBIT|dˆ pHYsaa¨?§i IDATxœìý{x–Õ/þ¿NŠDЂ JS¬H©:JÁm­ZùdæÍ›—K/½4kÖ¬IuuuV®\Ùeï‘eË–åÅ_l4€œ{î¹ùÏÿüÏ|÷»ßm…ÊØ[À>âÞ{ïMeee’dÈ!ù¯ÿú¯ôíÛw—cÞxãTUU5h[¹re&Ož\ß~ÔQG%Iú÷ïßUŽ:*½{÷N×®]íÿô§?¹sçæøCÁ•°·€}ijÏ>›:ìññ7n̘1c²fÍší×]w]þûÿï)/ûŸˆŠŠŠ}ôÑ­@Ž;î¸{ì±»ô¡åÉ'ŸÌ¤I“rÄGä3ŸùLª««“¼½^䥗^ÊoûÛ\uÕUyøá‡S[[›¹sçfĈ êùÛ¿ýÛú@’$ßøÆ72iÒ¤|ìcË—¾ô¥ú³mÛ¶\wÝu™9sf.¼ðÂ\tÑEY·n]ý¸ï}ï{¹ñÆsÑE¥sçÎyøá‡ó /ä/ÿò/S^^žÉ“'góæÍIÞ¾ ¬k×®¹ï¾ûrä‘Gæ´ÓNÛåÿ³³Ï>;wÝuW^zé¥wÿÐrJ´IgžyféŸøDý÷ÿôOÿTêÑ£GióæÍõm#GŽ,mÛ¶­T*•Jýúõ+]ýõ¥R©TZ»vméž{î)••••.¿üòÒOúèúÅëÕÕÕY¸paî¸ãŽ$ɤI“êר{ì±9ï¼órçwÖŸgîܹ¹üòË÷êÿY¯^½òÌ3ÏìÕŠ%€´QÇO§NòÀ$IæÏŸßà‡ôx £Gnrüö`òç ·Ï9çœÔÖÖfÁ‚ŽyõÕWs衇69ç믿žG}4&LÈŒ32cÆŒ|÷»ßͼyó’$ÿøÇ3räÈÌš5+3fÌÈ–-[êw®êß¿.»ì²Œ7.'Ÿ|r^xá…œp õsOš4)>úh–.]š$yì±Çò±}l·ÿŸþ\Ïž=óì³ÏîÕŠ%€´Qx`FŒ‘9sædùòåéÙ³gÆŸû·˲e˲qãÆ|ðÁ{5ç AƒêçnLÏž=S[[ÛäøÚÚÚlÞ¼9<òHƒöí»O½ôÒKùë¿þëœ}öÙùò—¿¼Ó•”Ûo¿=<ð@^}õÕüÕ_ýUî¿ÿþú¾³Î:+Çw\þéŸþ)Ï>ûlŽ=öØ”••íÕãK’Î;ïõŠ#€´a—^zi~üãçïÿþï3~üøüõ_ÿuzöì™±cÇæ‚ .Ø£9¶ßn•$¿øÅ/Ò¥K—wÜqIÞ¾MêÏoÁ:þøãóÖ[oeãÆÎUYY™“N:)sçÎm0nûU™ñãÇçCúPŽ?þø$ oñzî¹ç²hÑ¢\tÑEYºtiN<ñÄÜ|óÍ æ¿úê«sÏ=÷äë_ÿú^ß~•$+V¬H¿~ýözÅ@Ú°3Î8#•••©««KïÞ½Ó¡C‡|ò“ŸÌï~÷»œsÎ9 Ž]·n]£,ðW¿úUý?òÈ#™4iRýmV½zõÊóÏ?Ÿ5kÖ䩧žª¿%jÇ­yÿܨQ£òè£æ¬³ÎʬY³rÑEåƒü`’¤S§Nù·û·,\¸03gÎ̲e˲råÊ<þøãY¿~}¦OŸž$9øàƒë¯”ü¹±cǦ[·nyî¹çêçÜÑúõësâ‰'æ+_ùJƒöR©”Õ«W7¸­ €¶Ghã.¹ä’Lš4©þû+®¸"—^ziýíIkÖ¬Éõ×_Ÿ7Þx#ÿò/ÿ’»ï¾»~ázòöv½“&Mʈ#r 'ä†n¨ï3fL:v옱cÇæ¸ãŽËûßÿþ 0 .l²žÏ|æ3¹öÚkóÛßþ6?øÁ2nܸœþùI’©S§¦S§N=ztþò/ÿ2Ÿþô§óÒK/åàƒNYYYþå_þ%Çw\¾øÅ/fÓ¦M W’pÀ7nÜ.¯~¬_¿>¯¼òJn¸á†WjžþùtíÚu§-„h[ÊJ~ €}ƲeËòþ÷¿?O>ùdN?ýô=÷ÄOäÈÿù?ÿ§«kÚøñãsÓM7¥G»<îþáꯨ$o‡Ÿ-[¶äÿñ[ºDÞW@hଳÎÊÖ­[óûßÿ¾ðs¯\¹2uuu» ‹/ÎÀë¿ß²eKžyæ™Wwh›:¶v´Œí·'mØ°a¯Ç~õ«_Í7¾ñ\sÍ5éÞ½{s—¶“Ï}îs©««ËË/¿¼Û+/«W¯Nyyy.¾øâ$ɦM›2uêÔüÏÿù?Ó¥K—¯€wÇ€}ÐË/¿œY³f¥¬¬,ßúÖ·ò›ßüf¯Æz衹îºëòôÓO·P… ½ð Y°`A®ºêª~øá»<ö°Ã˱Ç[ÿý¯~õ«Üpà éß¿K— @3°(Œ+ @a 0P(ŒF #€…@€Â @a 0P(ŒF #€…@€Â @a 0P(ŒF #€…@€Â »ð“Ÿü$\pAªªªÞÕ<_ûÚ×R^^žW_}µ™*€ö©ckнúê«ùÖ·¾•{î¹'¿ÿýïsâ‰'¾ã¹yä‘L:5eeeÍX!´O®€4¢²²2_þò—3sæÌw5ÏóÏ?Ÿë¯¿>Æ K©Tj¦ê ý@vá=ïyÏ;»nݺ\qŹ÷Þ{Ó½{÷f¬ Ú/d:v|gw¨mÛ¶-—]vYn¸á†uÔQÍ\´_H ˜2eJÎ?ÿü :´µK€6Å"ôfv÷ÝwgÛ¶mù»¿û»½ûÚk¯åÑGM¿~ýrà¶@u¼o½õV–-[–sÏ=7={ölírÚ%¤ýâ¿È~ðƒÌŸ?¿ÑþÝ-DôÑG3vìØ–( €f4gÎœ|ò“Ÿlí2Ú%¤7.ëׯÏ1ÇÓ }åÊ•I’3Ï<3;wÎO<‘Ã?|§ñGyd’·_Ðüà[¾`öʤI“2{öìÖ.ƒ&x~Ú.ÏMÛå¹iÛŠ'€4¡²²2_þò—3sæ̽;{öìÜyçY³fMV­Z•9sæäÈ#Ì[o½•Ï~ö³-P-MéÓ­O–_³<˯Yž>Ýú´v9û=k@vã=ïyÏ^ÿÛßþ6ýèGsÙe—Õ·3&7n̸qã²xñâæ.Ú W@v£cǽËhýû÷ÏUW]µSû˜1cÒµk×|èCj®Ò Ý@šYyyãÿK7oÞœ7føðáWm‡RÇ{,gŸ}v>ó™Ï´v)Ðj¬)ÀÆsóÍ7çhíR U ˜1cFnºé¦zè¡{tü¤I“Ò£Gm£GÎèÑ£[¢<1wîÜÌ;·AÛºuëZ©š}‡ÒÂfÏžÓN;-'tÒ^:‡rH^|ñÅtìØ1[·nÍøñã3þüôîÝ»þغººÔÖÖfÕªU¹õÖ[ «Ú¤ 555¹å–[2þü”••å¾ûîËÁœ1cÆdÀ€I’®]»¦²²2‡vXý,¼üòËsß}÷¥¬¬,ëׯßiÞN:eäÈ‘…>h+&TTTdÊ”)™2eJ“ÇtëÖ-‹/nÐ6gΜ̙3§¥Ë€vÉ 0P(ŒF #€…@€Â @a 0P(ŒF #€…@€Â @a 0P(ŒF #€…@€Â @a 0P(ŒF #€…@€Â @a 0P(ŒF #€…@€Â @a 0P(ŒF #€…@€Â @a 0P(ŒF #€…@vã'?ùI.¸à‚TUUíõØÛn»-çž{n>úÑæ /ÌÒ¥K[ Bh?:¶vmÕ«¯¾šo}ë[¹çž{òûßÿ>'žxâ^¿ùæ›óÕ¯~5?ûÙÏÒ·oßÜyç9ùä“óÈ#ä#ùH U m›+ M¨¬¬Ì—¿üåÌœ9s¯ÇþøÇ?ÎäÉ“3~üøôíÛ7IrÅWä¸ãŽËUW]•R©ÔÜå@» €ìÆ{Þóž½óÙÏ~6ååå¹òÊ+´ô£ͯýëÜ{ï½ÍU´+Èntì¸ww©mÚ´)K–,I¿~ýÒ«W¯}GuT’äÁl¶ú =@šÙ³Ï>›mÛ¶í>’¤ÿþI’—^z©è² M@šÙÓO?$#Ž8"‰ÀþË.XÍlݺuI’Î;ïÔ·qãÆ$I÷îÝw9ǤI“Ò£Gm£GÎèÑ£›©Jvgîܹ™;wnƒ¶í?ëñΠͬ²²2Iý›µµµI’N8a—sÌž=;ƒnþâØcýxÑ¢E2dH+U´op V3;ú裓$Ï?ÿ|êêêô­\¹2Éîì«f6hР{ì±Ù´iS^|ñÅ}Ï<óLý1°?@vcûUŒ¦þxà† ²yóæm“&MJò§À±}žþð‡9ãŒ32|øðªÚ6d7^ýõ$Éo¼±SßêÕ«Ó§OŸ 0 [·n­oÿÔ§>•qãÆåºë®Ë›o¾™$™5kVþð‡?ä{ßû^ÊËýo`ÿdzjjjrË-·dþüù)++Ë}÷Ý—ƒ>8cƌɀ’$]»vMeee;ì°tèСÁøo~ó›¹é¦›òñõí]*R5´ªµÊ ûŒ>Ýúdù5Ë[» vÁ"t 0P(ŒF #€…@€Â @a 0P(ŒF #€…@€Â @a 0P(ŒF #€…@€Â @a 0P(ŒF #€…@€Â @a 0P(ŒF Ó±µ €¢­¨]‘¾³úÖ_Ñ¥"UC«2☭XÀþÁö]*’$u¥ºT×T×-}mi¦-˜ÖÊÕìöUC«2°çÀQqDýWyÙÛÿÖlªiåêönÁ`¿1â˜;ÝfÕwVßT×T·REûW@€Â @a 0Ö€@vÞš7±=/@K@vaÁ‚¹õÖ[³|ùòpÀ¹úê«sá…îÑØ•+Wæûßÿ~~øálܸ1ýúõË\óÎ;/:ujáÊØS]*’š?mÍÛ@M2mÁ4  MX¸pa† –|0C‡ÍÒ¥K3xðàÌœ93'NÜåØ_|1ÿí¿ý·|êSŸÊw¿ûÝtëÖ-¿ûÝïrþùçgÎœ9™7o^A€Ý©Z•i ¦í´ ïŠÚ©+ÕÙž ™ XµjU†žSN9%C‡M’ 80“'OΔ)Sò‰O|"GqD“ã¿þõ¯gëÖ­©ªªÊ˜$yßûÞ— &dâĉyùå—óþ÷¿¿ÇÀ®5¶5ob{^€–bz#n¼ñÆTWWg„ ÚO?ýô¼õÖ[™1cÆ.ǯ_¿>o¾ùfª«þÃÕ±cÇxàéÕ«W³× íÒˆE‹¥¬¬,'Ÿ|rƒöþýû'Izè¡]Žß¾NäŸøDV­Z•$©­­Í·¿ýíL˜0!-P5´}ÈJ¥Ržyæ™$IÏž=ôUVV¦¼¼;ëׯoÞ‚ ± Ö*++“$kÖ¬ÉÚµksÈ!‡4语­MEEEúõë×ä6lȨQ£rÚi§eÊ”)ùÜç>—‹/¾8O>ùd† –ÿøÇ»¬aÒ¤IéÑ£Gƒ¶Ñ£GgôèÑïìA°×æΛ¹sç6hÛþËjÞ9d½zõJEEEjkk³dÉ’œvÚiõ}k׮͖-[òáx—s|ó›ßÌã?žùóç'Iúôé“Ç{,ÿøÇóïÿþïY¸pa† ÒäøÙ³ggðàÁÍó€xGûð¢E‹vùs»ç¬F\rÉ%)•Jyî¹ç´o_œ>hР]ŽŸ?~ Ðà6®8 ·ß~{’äh抠}@1qâÄtìر>pl7þüôîÝ;×^{m}Û† ²yóæÇuÖYyöÙg³lÙ²í|p<ðÀ]Þ¾û2¤ÌwÜ‘»îº+K–,I’<õÔS¹í¶Ûrÿý÷çðÃO’¬^½:}úôÉ€²uëÖúñ&LÈA”)S¦Ô‡µk×æK_úRŽ9æ˜\~ùåE?$h¬iÂرcÓ¥K—|á _Hmmm*++óðÃçŒ3Ψ?¦k×®©¬¬Ìa‡–:Ô·÷îÝ;‹/ÎÔ©SsöÙgç½ï}oŽ>úè\pÁ™={v“;lÀ¾NÙ…‘#GfäÈ‘MöwëÖ-‹/n´ï}ï{_î»ï¾–* Ú%·`…@€Â @a 0P(ŒF #€…@€Â @a 0[»h/æ-™—éONOͦší]*R5´*#ŽÑJ•´ì¡éONÏÒ×–îÜQ“L[0MØì¡íW>ÊËÊÓ§[Ÿ$ÉŠÚ©+ÕítU€ÆY{©O·>Y~Íò,¿fy}`Ï @a 0P(ŒF #€…@€Â @a 0P(ŒF #€…@€Â @a 0P(ŒF #€…@€Â @a 0P(ŒF #€…@€Â @a 0P(ŒF Ó±µ hË,X[o½5Ë—/Ï«¯¾:^xá^ϳqãÆÌž=;¿ÿýïÓ£GœtÒIïhhï&,\¸0Æ ˃>˜¡C‡féÒ¥¹A{ÿþý“$=ôÐ.Çÿîw¿Ë'?ùÉ 2$#FŒh±: ½@vP*•òÌ3Ï$IzöìÙ ¯²²2åååY³fMjjjšœãë_ÿz^ýõŒ1"÷Þ{o®ºêª <8Ÿþô§³jÕª­Ú2d/¿ürjjjÒ½{÷tìØp~§NÒ»wïúãšòÓŸþ4:tH]]].¼ðÂÜrË-©ªªÊ=÷Ü“N8!+V¬hÑÇm•]°v°nݺ$IçÎí߸qc’¤{÷îö¿ùæ›yæ™gÒ§OŸL™2¥¾ýüóÏÏÕW_™3gfÊ”)¹ûî»›¬aÒ¤IéÑ£Gƒ¶Ñ£GgôèÑ{õXxçæΛ¹sç6hÛþ³"ƒÊÊÊ$Éš5k²víÚrÈ! úkkkSQQ‘~ýú5:þùçŸO]]]þâ/þb§¾3Ï<33gÎÌÿý¿ÿw—5Ìž=ÛÂu€VÖØ/€-Z”!C†´REû·`í W¯^©¨¨H’,Y²¤AßÚµk³eË–üñMŽ0`@’äøÃN}§vZ:tè°Ëõ#°/@î|\Ç IDATqÉ%—¤T*å¹çžkо}qú AƒšÛ­[·|àȲeËòÆo4è+•JéرcŽ;î¸æ/Ú¤'NLÇŽëÇvóçÏOïÞ½síµ×Ö·mØ°!›7onpÜç>÷¹$Ùé/Ÿ?òÈ#Ù´iS®¾úêªÚ6¤ÌwÜ‘»îº«þ6¬§žz*·Ýv[î¿ÿþ~øáI’Õ«W§OŸ>0`@¶nÝZ?~ܸq3fLªªªò«_ý*ÉÛkCfÏžqãÆåoþæoŠPÐX„Þ„±cǦK—.ù¾ÚÚÚTVVæá‡ÎgœQL×®]SYY™Ã;,:th0þŽ;îÈ°aÃRUU•Õ«W§k×®¹ì²Ërå•WýP Í@vaäÈ‘9rd“ýݺuËâÅ‹íëܹsF•Q£FµTyÐî¸ (ŒF #€…@€ÂØ €veÞ’y™þäôÔlª©o[Q»¢Åη¢vEúÎêÛâçØ_ ´+ÓŸœž¥¯-m´¯¢KE³§¢KER“Ô•êR]SÝbçØß ´+Û¯|”—•§O·>õí]*R5´ªÙÎS5´*ÓLkp¥¥%ΰ¿@h—útë“å×,o±ùG3"#ŽÑbóì¯,B #€…@€Â @a 0P(ŒF ã/¡@3XQ»"}gõ­ÿ¾¢KEª†Vùkê;pÞ…Š.I’ºR]ªkªë¿–¾¶4ÓLkåêÚÞ…ª¡UØs`Ž¨8¢þ«¼ìí^k6Õ´rum[°à]q̈n³ê;«oªkª[©"€¶Í 0P(ŒF #€…@€Â @a 0P(ŒF #€…@€Â @a 0[»ØW­¨]‘¾³úÖ_Ñ¥"UC«2☭X@ëršYE—Š$I]©.Õ5Õõ_K_[ši ¦µru­K€fV5´*{ÌGÔ•—½ýOnͦšV® u¹ šÙˆcFìt›UßY}S]SÝJ´®€…@€Â @a 0P(ŒF #€…@€Â »°`Á‚Œ5*§žzjÎ:ë¬üèG?zÇs}ík_Kyyy^}õÕf¬Ú—Ž­]@[µpá 6,>ø`†š¥K—fðàÁ™9sf&Nœ¸Ws=òÈ#™:ujÊÊÊZ¨Zh\iĪU«2|øðœrÊ):th’dàÀ™üá79¾¶¶6Ë—/ß©}û•W^y%eeeÙºuk“sÌž=;ƒ~€w«±_/Z´(C† i¥Šö Ö€4â’K.I©TÊsÏ=× }ûâôAƒ59ö•W^ɶmÛvúºôÒK“$Ë–-˶mÛ꯴ÀþDiÄĉÓ±cÇúÀ±ÝüùóÓ»wï\{íµõm6lÈæÍ›‹.Ú%¤ÌwÜ‘»îº+K–,I’<õÔS¹í¶Ûrÿý÷ׯÝX½zuúôé“ìò–*àmÖ€4aìرéÒ¥K¾ð…/¤¶¶6•••yøá‡sÆgÔÓµk×TVVæ°ÃK‡v9_YYYÊÊÊZºlhÓ]9rdFŽÙd·nݲxñâ=šëÎ;ïÌwÞÙ\¥@»ä, 0P(ŒF #€…@€Â @a 0P(ŒF #€…@€Â @a 0P(ŒF #€…@€Â @a 0P˜Ž­]ìOVÔ®HßY}뿯èR‘ª¡Uq̈]Ž›·d^¦?9=5›jv{Ž= 5 P€Š.IMRWªKuMõŸ:j’i ¦í6,Lrz–¾¶tÏN¶‡s´ P5´*ÓLkpcEíŠÔ•êöèªÆöcÊËÊÓ§[Ÿ&Û›9Zƒq̈®HôÕ·áÕ=Ч[Ÿ,¿fy“ýïdN€"Y„F #€…±ژƶÜ]Q»¢+h>´1»Úr·¢KEÁÕ4/Ú˜¦¶ÜÝþÚ3Ú¨Ým¹ ÐY„F #€…@€Â @a 0P(ŒF #€…@€Â @a 0P(ŒFÙ… dÔ¨Q9õÔSsÖYgåG?úÑ}â‰'rú駧Gyï{ß›Q£Få…^hÁj í@š°pá 6,ãÇÏOúÓÜ|óÍ=ztþ÷ÿþß»ûË_þ2çœsN~þ󟧬¬,ÕÕÕ™7o^ ”§Ÿ~º€ê m@±jÕª ><§œrJ†š$8p`&Ožœ)S¦¤ººz—ãgÏž;ï¼3k֬ɪU«2gÎœyä‘yë­·òÙÏ~¶ˆ‡m’Òˆo¼1ÕÕÕ™0aBƒöÓO?=o½õVf̘ÑäØßþö·ùèG?šË.»,Ý»wOçÎ3f̘|éK_J’,^¼¸Ek€¶LiÄ¢E‹RVV–“O>¹A{ÿþý“$=ôP“cû÷®ºj§ö1cƤk×®ùЇ>Ô¼Å@;"€ì T*å™gžI’ôìÙ³A_eeeÊË˳fÍšÔÔÔ4:¾¼¼ñÿ¥›7oÎÆ3|øðæ-Úd/¿ürjjjÒ½{÷tìرA_§NÒ»wïúãöÆc=–³Ï>;ŸùÌgš­Vho:îþý˺uë’$;wn´ãÆI’îÝ»ïñœ7nÌÍ7ßœx`ŽŸ4iRzôèÑ môèÑ=zôŸ€wgîܹ™;wnƒ¶í?+òÎ ;¨¬¬L’¬Y³&k×®Í!‡Ò ¿¶¶6éׯßÏ9cÆŒÜtÓM9ôÐC÷èøÙ³ggðàÁ{ö±í¯¼òJ>ÿùÏ·LáÐÆYÒˆæŽ;îÈøñã3a„{ì±yꩧrÛm·å_ÿõ_søá‡'IV¯^£>:‡rH^|ñÅtìØ1[·nÍøñãëÃÊvuuu©­­ÍªU«rë­·¶ÖC€V%€4aìرéÒ¥K¾ð…/¤¶¶6•••yøá‡sÆgÔÓµk×TVVæ°Ã«ß²÷òË/Ï}÷Ý—²²²¬_¿~§y;uꔑ#Gö8hûVÔ®HßY}|ßÜsVt©HÕЪŒ8fÄ»žàÝ@vaäÈ‘» ݺuËâÅ‹´Í™3'sæÌiéÒØTt©Hj’ºR]ªkªïo®9k’i ¦ @«@ •T ­Ê´ÓR³©f§¾íW,šcε+RWªkô<E@ •Œ8fD³_‘hlξ³ú6z… 5Ø (ŒF #€…@€Â @a 0P(Œ¦ck@ó™·d^¦?9=5›j´Wt©HÕЪŒ8fD+Uo@ö!ÓŸœž¥¯-ݹ£&™¶`š@«@ö!Û¯|”—•§O·>I’µ+RWªÛ骴k@öA}ºõÉòk–gù5Ë냴P(ŒÆ"t`—lï 4'Ø%Û;ÍIvÉöÎ@s²Ø#¶wšƒF #€…± ™µÄ–¥Í=§mUiê5x-Khf-±eisÏi[Uš| $^@‹@šYKlYÚÜsÚV•Æ^‰×Ðò€²}ËÒ$é;«oªkªÛÜœ-Q#íËŸ¿¯ åY„F #€…± •ìjÔ­¨]±Ë¾¾³ú6Ûqì»Zâ5ðçs&¶ðvOh%»Üµ ]*þwMRWªÛiÑð;9Ž}WK¼šœÓ¾Àn ­¤©mP›²ý7ËÛU ­Ê´Óšüc‚{{û®–x 46§-|=!€´²·AÝS#Ž±G¿eÞÓãØwµÄk ±9má ì ‹Ð€Â @a 0P(ŒF #€…@€ÂøKè@‹™·d^¦?9=5›j´Wt©HÕЪfûËÌM§1;ž{oÆ66ØÙŠÚé;«oý÷{ú¾ÙÓÏŒ–øl)êó @€4ýÉéYúÚÒ;j’i ¦5Û?èMž§1;œ{¯Æ62ø“Š.IMRWªKuMõŸ:öð}³§Ÿ-ñÙRÔç €-hûoËËÊÓ§[Ÿ$oÿf´®T·ÇWÞéyÓع÷tlSã?©Z•i ¦5xìÍûfO?3Zâ³¥¨Ï+@ Ч[Ÿ,¿fy’¤ï¬¾ 3ÚBçi̮ν»±»$#Ž±Ó•‚wò¾ÙÓÏŒ–øl)êó ög¡…@€Â @a¬a¿Öž·]|7µ·µÇÝÞ¶ÂÝq‹ÑÆú¡½ûó×ys¼¦w÷¾i©óìé¾mí}ÛÖ>§¡9 »°`Á‚Üzë­Y¾|y8à€\}õÕ¹ð ÷xüm·Ý–yóæeÆ 9ôÐCóÕ¯~5lÁŠÙ[íyÛÅwS{[{Üíe+Ü&·ÝÕñÐÎìêuþN^Ó{û¾iöóìåÖÛmå}ÛÖ>§¡9 MX¸pa† –|0C‡ÍÒ¥K3xðàÌœ93'NÜíø›o¾9_ýêWó³Ÿý,}ûöÍwÞ™“O>9<òH>ò‘ðØ{³íâܹs3zôèÂklÊ»Ù2²­m7Ù,[á.NrJ øGm1Ú”í¿¥¤í½wø“Æž›¦^çïô5½7ï›æ>ÏÞn½Ý–Þ·5›j’ÅIùñmãsš“ÒˆU«Veøðá9å”S2tèÐ$ÉÀ3yòäL™2%ŸøÄ'rÄG49þÇ?þq&Ožœë®».}û¾}øŠ+®Èí·ßž«®º*ÿùŸÿ™²²²B {fO¶]l«?D½›-#ÛÚv“ïj+ÜgÓ⤱-FÙ½¶úÞ¡ñ禹_çE½oöv à=ù¼iuÏ&}Ni[ŸÓÐ,BoÄ7Þ˜êêêL˜0¡Aû駟ž·Þz+3fÌØåøÏ~ö³)//Ï•W^Ù ý£ýh~ýë_çÞ{ïmöš =@±hÑ¢”••åä“OnÐÞ¿ÿ$ÉC=ÔäØM›6eÉ’%éׯ_zõêՠ﨣ŽJ’<øàƒÍ\1´nÁÚA©TÊ3Ï<“$éÙ³gƒ¾ÊÊÊ”——gÍš5©©©IEÅÎ Õž}öÙlÛ¶m§ð‘ü)À¼ôÒK-PyÛò£%¯¾Út©´û9šã˜Ýõ×þx\²i}jºtϬ?¶ÕüøÊdÓ ÚJ¥ä¥—’¯½åjÝ“cþ¼¿æ'Wý±öƒóÕ-ÞV“šÎç¦ÍMÏSó“ÿQ?öÆMlû?µ}e㻯5IÖÿÇÕÉæõYßùà|yÃÛ~ú§¶ª7ÿØö³Ï&›Ög}—?µ5užõ?›T?þj··MN^ûVÖÿÛßåúFnÞ>Çú§®©{ÝúÚ:œo¼»Ç¼«þõ?ÿÿ'›k²¾óÁùû×ÿØö‹Ïÿ±­¢¾íÝžçÝóƯ¦&›kòFçŠ\»ºùÎóÜsÉç?ÿîæøócÞX8-Ù\›7:wËçªßÙïô˜}í<¿þurÕU-žw3ÇÚÅ3“-ofm§ƒ2î¹?¶=;3Ù²!k;u­oklŽµÏ~­þ¸O-Ù>ß?&[7dmÇ®ùÿ=»ëZÖ.ùzýø+7ÏãÙÓc^nv²â+yý¾/å’ÿücÛo¾‘lÙ×;u­o{'çÙÝ÷{[ëžs É¿¸ûyØ÷••J{ò’Ú¼ôÒK9úè£Ó£G¼þúë;õqÄY¹re-Z”N8a§þÛo¿=W^ye†žïÿû ú^xá…|àH÷îݳvíÚÆþìg?Ë©§žš9sæäƒü`ó=¨Vð?þGòóŸ·vÍmR’Ù­]Mòü´]ž›¶ËsÓ¶í[ÏÏAGþ:ÿ÷ã[»Œwí7¿ùMÆŽ›ÿøÿÈ©§žÚÚå´K®€ì`ݺuI’Î;7Ú¿qãÛ¿îÞ½û^ßÝØW^y%I2vìؽ¨˜b iíØ%ÏOÛå¹i»<7mÛ¾óü¼ùJ2dßy8Y¶l™ò ;¨¬¬L’¬Y³&k×®Í!‡Ò ¿¶¶6éׯß.Ç/]ºóÞݵµoß/ÒØ•“$9÷Üs3gÎœôë×/xà;}´7æ•W^ɹçžÛÚ¥´[Èzõê•ŠŠŠÔÖÖfÉ’%9í´ÓêûÖ®]›-[¶äÃþp“ã>úè$ÉóÏ?Ÿººº”—ÿiÿÊ•+“4@zöì™O~ò“Íñ0h!§œÒÂû¾ïãì‚ÕˆK.¹$¥R)Ï=÷\ƒöí‹Ó ÔäØAƒåØcͦM›òâ‹/îõxØ— ˜8qb:vìX¶›?~z÷îk¯½¶¾mÆ Ù¼¹ávC“&MJ’ãëêêòÃþ0gœqF†Þ‚Õ@Û%€4bàÀ¹ãŽ;r×]weÉ’·÷í{ꩧrÛm·åþûïÏᇞ$Y½zuúôé“dëÖ­õã?õ©Oeܸq¹îºëòæ›oï):kÖ¬üáÈ÷¾÷½·eí‰ÿú¯ÿÊ5×\“k®¹&3fÌÈÓO?ÝL”æPWW— .¸ Gydk—²ß+•J¹å–[r 'ä ƒÊ>ð|îsŸËoìf]ZÄ‚ 2jÔ¨œzê©9묳ò£ý¨µK"É‹/¾˜#F¤OŸ>yÏ{Þ“¡C‡æñÇoí²hÄOúÓtéÒ%wÝuWk—ÂæΛñãÇgêÔ©™={ßÙ©¬0%šô½ï}¯tÞyç•N?ýôÒرcKO<ñDƒþšššÒqÇW:óÌ3Kuuu úêêêJ_ùÊWJõWU:ýôÓK“'O.=ÿüó{uþ-[¶”fÍšU4hPéñÇׇ–1uêÔRYYYéÈ#líRö{3gÎ,••••***Jp@©¬¬¬TVVV8p`iýúõ­]Þ~åÿý¿ÿW:è ƒJÿþïÿ^*•J¥ßüæ7¥<°ô¿þ×ÿjåÊöoo¾ùf©gÏž¥òòòRÏž=ëß#eee¥o~ó›­]æÕW_-õîÝ»TVVVºë®»Z»þhÉ’%¥3Ï<³4a„RMMMk—Ón mÔÖ­[KçœsNiàÀ¥uëÖµv94aîܹ¥¿ýÛ¿-½ï}ï@ZÙÆKùÈGJ?þxiË–-¥×^{­ôÿøõAdúôé­]â~cåÊ•¥¾}û–Î9çœí_üâÿ¿öî-4Êkâøúbb¢! õ2>”!TƒműZ5ñ…* ©¯y¢VñJª´%/*ñ†iAEK¼¤¨EŒZ/ÑQ‰ Qu´MƦ&Nw49N“œs,§{OÎü—½‡°`¾Y|ûbºwïnêêê%æM›Ìâŋͽ{÷Œ1Æœ}ºž?®ââb×qþ/P@bЊ+” œœmß¾]Ó¦MÓ¨Q£T\\¬ææf×ñâ^SS“fÍš¥-[¶Èç󹎃×::Üá×_UBB‚¦L™b9Q|2Æ´žØ«W¯¨9¿ß¯„„=~üX .âÅ=ÏóÚ¯¯¯×Ǭ¤¤$ˉð¦C‡éÆZ¹r¥ë(xÃáÇuîÜ9M™2EgÎœÑêÕ«•••¥¼¼<]¼xÑu¼N‰‹ÿa«V­REEÅõYÏótúôiõíÛW *((ÐìÙ³URR¢Õ«Wëĉ:yòd‡¼·ý~Ξ=«µk×jþüùÜçò{Ûïæܹs­'Ô½éرc*..æû²¤¶¶V êÑ£‡£1IIIòù|zøð¡jkk;¼”vƒA=yòDUUU®£Äµ@  ÒÒRÞÆ êêjIRrr²>øàMžùä}ôÑG:tè·¢¿-×›P­ªªÊxžgfÎœÙfnĈÆó<óÝwß9HcŒÙ°aƒY·n]›q6¡Ç¦@ `¦M›æ:F\¹xñ¢ñ<Ïø|¾vçßyç“`îܹc7:4{ölS]]í:F\ …B&77׃Á¨ñòòrãyž©¨¨p” Æüë÷×íÛ·£Æ[~³ùý~GÉ:/–`Ř–Û×ûöíÛfnüøñ’¤S§NYÍ„W‚Á /^¬]»v)333êïÁƒª««Sff¦&Nœè:*$ýþûïZ·n‡XÖ²¹ùñãÇzöìY›ùp8¬´´4¥§§[N†öTVVjÈ!5j”ë(qmÍš5ºvíšÆŽõlY¾|¹$iùòå8p *++'Oׯ_—çym~›3F]ºtQ]]îܹã(]çĬ“™™)I …BmærssõÍ7ß°vÚ‘ÆÆFE"‘vO‰1¯7 ÞºuKMMM¶£á/^¾|©¢¢"}õÕWêÑ£‡ë8q¥wïÞJKKS8V ÐèÑ£[çž={¦ææf1ÂaB´ø駟tåÊN%‹ápX¡PHOŸ>oy¶<|øPÁ`Põõõ.⎨¦¦F¡PH©©©­ã©©©ÊÎÎÖÏ?ÿ¬p8ì0açÃ3lØ0Ij÷¶ó–õÔƒ¶š ¯¤§§ë?þP$ióç÷û•žž®H$¢ÚÚZ×QãZ$ÑÚµkµdÉõë×Ïuœ¸4kÖ,cZßè¶hÙœÎ~÷Ο?¯K—.Q>bDyyy»Ï—–7¸ŠD"úüóÏ'OÙÙÙ2ÆèòåËmæ’’’”””¤8HÖyQ@bŒÏçS~~¾®\¹Òºé©Å?ü nݺµ+ Úo¿ý¦/¿üRsçÎm³ÄçĉÚºu«›`q¦°°P‰‰‰­…£Åäóù´lÙ2GÉ IUUUºpá‚ŠŠŠ¢ÆëëëµpáB½xñÂQ2ü•áÞ˜°`Á¥¤¤hË–-Qã=RuuµæÌ™£®]»:J×9Q@bÐöíÛ•‘‘¡/¾øBOž<‘$?~\ÔæÍ›õî»ï:NÄžG)77W{÷îÕ¤I“Z×P÷ïß_½{÷ÖĉÙŸcIff¦¾ÿþ{íܹS@@’töìY•••iïÞ½ížV;ÊÊÊ”——§ÒÒÒ¨½~¿_}úôÑ/¿ü¢ääd×1˜òþûïkãÆúñÇUZZ*Izúô©V®\©Ä ëËIDATáÇs7ÈßÀ”ššªÓ§OkÏž=š:uª<ÏSÏž=µgÏ3Æu<´ƒc‘Ýzñâ…FŽ©»wïvø™aÆ)##Ã^¨87sæL%''kéÒ¥ ‡Ãòûý:zô¨rrr\G‹[eeeš3gŽ<ÏëðÆóüü|Ë©ðïxžÇó%F(##C»wïVVV–ÒÒÒ4räH:uŠ·ƒgx¿À–`°†À k( ¬¡€°†À k( ¬¡€°†À k( ¬¡€°†À k( ¬¡€°†À k( ¬¡€°†À k( ¬¡€°†À k( ¬¡€°†À k( ¬¡€°†À k( ¬¡€°†À k( ¬ùBŸLJ‰®]IEND®B`‚HDDM-0.2/hddm/examples/RatcliffFrank.conf0000664000102500010250000000014511713025571020746 0ustar wieckiwiecki00000000000000[model] data = RatcliffFrank.csv [depends] a = da,conf v = conf [mcmc] samples = 10000 burn = 2000 HDDM-0.2/hddm/examples/a.png0000664000102500010250000011430111726446354016323 0ustar wieckiwiecki00000000000000‰PNG  IHDRèXñÌ%sBIT|dˆ pHYsaa¨?§i IDATxœìÝ}XÔu¾ÿñ× ˆ “wèvÌÐ(Q(Ï9 V[¸…e+µ¶Q–µœ´ËJù‘éb¸,Éuïb5ËUϹ’ÀÔn\¶N›eÖšÂ9!êz›´E‘w$7Êp÷ûƒe‚™ïwfžëâ’ïÍÌ÷=£~f^ó¹K}}}½€¡|Œ.Ð0:&@@Àè˜  `tL€€€ Ð0:&@@Àè˜  `tL€€€ Ð0:&@@Àè˜  `tL€€€ Ð0:&@@Àè˜  `tL€€€ Ð0:&@@Àè˜  `tL€€€ Ð0:&@@Àè˜  `tL€€€ Ð0:&@@Àè˜Ý`Û·o×wÜ¡°°0­X±BõõõÏ;}ú´æΫ§žzªÕûÊÎÎÖ¸qãœU*À‰èªªªÒ† ”œœ¬iÓ¦)))I999ÏýòË/uèÐ!UVV:<^\\¬´´4Y,g– ðRï¾û®n¼ñFµ8f³ÙôoÿöoúðÃíûjjjôì³Ï*%%EIIIª­­ue¹¸%ºöíÛ§ŒŒ EEEiÞ¼yš4i’233ž;räH1¢Õößÿþ÷š9sf«Çh¯“'OªººZ{öìqøAðïÿ{5;–ššª¡C‡*55UW_}µžþyW– €[" (""BöíáÇkÈ!­žßµkW‡ûÿô§?iêÔ©êÑ£Gg—€ú÷ï¯ &8<¶k×. 4H}úô±ï;s挖,Y¢_þò—’¤ &(##£ÕQ` ÝD¶nݪøøøV;êµ8zô¨ÊÊÊáÌÒ^Î×׷žòòr½ñÆzôÑG›í/,,T`` zõê%I4hÎ;§ýû÷»¢TÜ–ã.Y¸\^^žbcc5tèÐ6ߦ¶¶V«V­Rzzz›osêÔ)½÷Þ{2d=îàBçϟ׉'tçw*00Ðèr:ÅþðÍŸ?¿Åþ¨oß¾öm___ 0@EEE=zt‹óymcxâk“»# ›€ÍfSffæemIúä“OôòË/kýúõöû±ÙlêÛ·¯Îœ9ãð6ï½÷ž¦NÚáší³aÃ=ôÐCF—Ñaï¾û®ÆŒ£~ýúÙ÷5®ƒb±XÔ¥K—fç—••Éßßßá}ñÚÆò”×&O@@7ôôtÍŸ?_>>—7ã`̘1:xð }{Ó¦MÚ¼y³6nÜØêm®¹æI ÿ CCCÛW°‡JHHÐòåË.ÔxnãyiÏMKÔÔ©Síí°»[ºt©öîÝkß>wîœ&Nœ¨ è®»îÒ¼yóìÇÊËËUQQ¡°°0‡÷ÕY¯M;wîÔ… Ô½{wýô§?m÷ý´…»ý§ÞÎÓøYúOIÿŸ—HzæGg>-鬤>’2:áÊ_HZÐ)ïáÌüü:âmõº²-ó´×&O@@7Ø‚ tÛm·©¼¼\¥¥¥úðÃ¥7ß|S÷ÜsBBBìçÚl¶f_Sãç秠  ûv¿~ýÔ½{÷fû~¬{÷î’¤ÐÐP5Ê È}õîÝ›ç¤<7Žñ¼´Žç¦uí°»Û°aƒªªª$5ôœßrË-Z²d‰îºë.ùûû«®®N'OžTÿþýURR"«ÕªÁƒ;¼¯Îzm:qâ„*++Õ³gO§ÿûs·ãÔë wKj¬q£¤÷>ÎSC@ïéàX{äKZÐ)ïáÜãùý·Õëʶ¬‘§¼6yºRSS•––¦´´4û¾°°0ÅÇÇ+;;[ÁÁÁö€¾gÏ}ôÑG:}ú´rssÕâþ, ߃pŠêêjI ßo.I?ùÉOšïÚµ«®¼òJ]qÅ’¤I“&i÷îÝš8q¢rrrš½Ö9ËäÉ“~ p6Ú2ïÆ*îJIIQ]]]³ŸÂÂBIÒÞ½{›ýç=z´vîÜ©8 ç’4mÚ4íرÃ¥¼HEE…^zé%Y,­_¿^§NºämV­Z¥mÛ¶é…^Ð÷߯§žzÊ•àÞèAåïï¯Y³fiÖ¬Y­žóÅ_´¸;ÍÀ èÀ?ÅÅÅ]‚iñÜ8ÆóÒ:ž¡¬¬Lõõõ²X,²Z­N½–»ý§^gs¯zÝíùõ¶z]Ù–Á|èÀ?¹[ãïJ<7Žñ¼´ŽçFX³fÊÊÊdµZ•˜˜èÔk¹Û¿qêu6÷ª×Ýž_o«×•ṁ€<ÂäÉ“USS£®]y{À}Ñ–y7þÖ€G2dˆÑ%@‡Ñ–y7VqÀè˜CÜ€GÈÍÍUUU•üüüet9Ð.´eÞ€}ZsçÎÕSO=Õlyy¹’’’¤©S§ª°°Ðe^ÉÈ€n±4üI@ð\tUUUiÆ JNNÖ´iÓ”””¤œœ‡ç~ùå—:tè*++›ííµ×4xð`­[·N_}õ•}ôQTx'zÐsËËËÓ§Ÿ~ª¼¼<£K€v£-ónÌA7о}û”‘‘¡€€EEE©  @™™™Š‰‰iqîÈ‘#5bÄ}ý£ïXºîºëô³ŸýL’4xð` 6L‡Ò°aÃ\òÚjáÂ…JNN6º  C耹}úé§ö•ÃÃÃ.Ú…¶Ì»Ð Ñl{øð᪨¨hõü®][þu5†sI ’ÕjÕ Aƒ:¯ÈNðÙgŸiÇŽt¸=3t‰Z÷ãi`àŽh˼CÜMdëÖ­Šoõ¸¥qêEn+ÿÎ.­Ý¾ùæ=ðÀF—t 3ÌA¯¯'¤í믿ÖóÏ?¯!C†èÃ?¼äùõõõúéOª[o½µÕµchzÐM"//O±±±:th»ïcÛ¶mZºti›ÎMHHPïÞ½›í‹‹‹S\\ÜEowâÄ ­ZµJ………:v옢££µbÅ uqVª««• ³gÏêìÙ³7nœn¹åÝ}÷ÝúÓŸþ¤={öhÓ¦Mzøá‡åëë«>úHgÎœÑÊ•+õùçŸëóÏ?×M7ݤիW+ À~¿GUbb¢Nž<©ž={ê‘GÑ´iÓìÇKJJ” o¾ùFEEEzì±Ç´`Á‚K~À\ŠzÐ¥†€Î?gsËÊÊRVVV³}¥¥¥Uƒ‹9þ¼¶lÙ¢ 6èË/¿lÓkEuuµŠŠŠd±XT[[ëp„íÁ+Š Øl6eff*==½Ý÷±uëVÝ{ï½êÕ«W›Î_¾|¹FuÙ×yâ‰'4jÔ(½óÎ;zÿý÷uçwjøðá‡âøúújãÆ7nœ,‹vìØ!IÚ³göîÝ«“'Oê7ÞÐã?®÷Þ{O’´`Á}ÿý÷zóÍ7uäÈ 6LW^y¥ý¹9räˆ"""´yófEGGkݺuzì±Ç4dÈÝvÛm*//×­·Þª¬¬,5J¹¹¹?~¼®¹æ=üð×ýx¦štW¿oÐëêšoÃ|}à™ŸŸÏ\BêÑ£‡fÍš¥¯¾úªÍ¯ÃݺuÓ±cÇ$9ž~v)K–,Qll¬|Ù·x6Þâ™@zzºæÏŸ/Ÿv¾ã.,,TII‰î¾ûîN®¬¹ŠŠ ½ÿþûŠŒŒ”$ÝvÛmòóó³¿Ii͇ÿ=Z#FŒÐùóç5sæL=þøãÚ¼y³ºt颷ÞzKcÇŽ•$ :TC† ivÿ‰‰‰6l˜¢££%I÷ÝwŸ®½öZýßÿýŸ$iÙ²eºñÆí>DEEiØ°azã7:çI€W3K: ÅŽ½òÊ+Zºt©^yå•Ë¾­ŸŸßeïëë+__ß˾Î7ß|£ôôtFuhUGÚ2¸?zÐ ¶`Á{Ïoii©>üðCEEEéÍ7ßÔ=÷Ü£û¹6›MµM‚¤C‡iÍš5JLLÔ‰'töìYíÞ½[O<ñD§×êïﯼ¼<1BUUUzûí·åçç§sçÎ]ö}uíÚU}ûömÑã¿cÇ]sÍ5ª­­Õ_ÿúWÕÖÖÚ￾¾^|ðæÎk?¿W¯^:räˆ}ûã?Ö‰'4nÜ8û¾šš9rD555 CD‡Ðsk|}ºÜ°-ÉáT­ÎV\\¬»ï¾[ß}÷s×´ª#mÜ=èJMMUZZšÆ¯àà`…„„hÙ²eºöÚk•­ÂÂBû¹{öìÑG}¤½{÷*77WRà ýí·ß®ŒŒ ]sÍ5 VDD„úõëç´šCCCõ /衇ÒW\¡>}ú´ûM†£Û…††jíÚµšøàûÏÁƒUXXH8G‡™a‘8‰Eâ€Öüä'?ÑG}¤?üáºîºëtÇwèÕW_½¬û¨¯¯×_ÿúWÅÇÇkÈ!º÷Þ{URRÒìœ>úH?ü°ÂÂÂší¯©©ÑªU«4yòdËÇÇG>>>úòË/uêÔ)Íœ9S%%%ª¯¯×< qãÆéõ×_ïðãàYÆŽ«¨¨(û¨RxºRRRTWW×ì§1”ïÝ»W“'O¶Ÿ;zôhíܹSPTT”$iàÀúꫯšÝ¾¶¶VS¦LqJ½eeeš0a‚JKKµyófÝu×]zÿ555zðÁõñÇkË–-úå/ÙlØ¿~ýäëë«mÛ¶µ¸mqq±$iÀ€zýõ×uþüùfÇÿþ÷¿«¦¦¦Së…÷¡0¯ºº:ýâ¿ÐîÝ»õÆo¨°°P¾¾¾zôÑGõÉ'Ÿ´ù~222Ô¯_?­]»VŸ}ö™víÚ¥§Ÿ~Ú~üã?Ö{ｧÌÌL]¸p¡Ùm³³³õ駟ê7ÞÐñãÇõÉ'ŸØ‡Áê­·ÞÒwÞ)IÚ¸q£>øàÝÿýðèž‚€Ž6{ýõו››«¤¤$û¾úúzÕ]"-øøø´ÌŽ|øá‡Ú¸q£æ̙Ӭ·»ñþ»té¢o¼Qyyyúæ›oìÇ+**ôñÇK’nºé&;wN?ü°vïÞ-©aa¹?þñô £Ãè€yíرC………úÙÏ~&IêÞ½»&L˜ úúz´ù~¦OŸn_ÌoÀ€ºé¦›ôæ›oªººZ’tË-·(--M}ûömqÛ5kÖ4Å©„„†³ÚŒ€Ž6»êª«$I;wîT}}½Þzë-={VúÇ?þaóâèv………Ú½{·233%5̧?yòd³óh¿IúàƒtüøqUTT¨´´TgÎœÑ /¼ I?~¼þüç?kÓ¦Mš>}º&Mš$IzöÙg5`À½ñÆ;v¬5lØ°K~}ÐtÀ¼nºé&%''+22R‡Vmm­=Ÿ:uªÍ÷Ó½{÷fÛV«Uµµµ:{öl³ý=zôhqÛýñÔÿûÿϾ~ÊâÅ‹Y­Àe9|ø°öï߯Ç] @@G›?^?þ¸f̘¡ûï¿_555Š‰‰ÑéÓ§õÝwßµºšíÌ™3Õ«W/ÍŸ?_QQQZ½zµÞ{ï=û*îÿøÇ?$5Ì?Ÿ?¾/^¬˜˜;vLqqq²Ùl*,,TŸ>}tË-·èþçÔ½{w=þøãúä“O”ššj¿vïÞ½õÙgŸiâĉêÓ§  ÌÌLÝ|óÍ.{žà¹è€yõìÙS½{÷Ö[o½¥'Ÿ|RS§NUQQ‘$]r¤×Å4¾¾TUU5ÛïhQ¹äädÝpà Z±b…‚‚‚”””Ô¦dÐTNNŽ6oÞ¬œœ£Kó‹6óññÑË/¿¬—_~Ù¾/66ö’·;v¬¾úê+ûöO<Ñê*ó .ÔÂ… ›í[¼xq³íÛo¿]{öìiõzAAAÚºuë%ë.‹ÄæÖ§O-]ºT/½ô’bbb”››«åË—wÊ}·e˜úàÁƒ•ŸŸ¯õë×ëùçŸ×âÅ‹õÖ[oéí·ßnö­,p1O>ù¤Ñ%À@ô @у˜×«¯¾ªøøx-]ºT111†Õáãã£øøx9rDO?ý´:¤ßýîw†ÕÀýøùùÙà}èÐFtÀ¼GNÚ÷5mïÈ÷ËñàƒÚ×céÖ­›–,Y¢+¯¼R°ŸÓø†»q†Àš" @ÐóºýöÛ%IÏ=÷œÞ{ï=Í;WóçÏ—$}þùçZ·n}ÍG󃫨¨¤ß…þÝwßµX|®  @ :sæŒ$é믿Vii©zè!û9×^{­$éÅ_Ôý×9üêP€÷" @Ðóz衇4~üx8p@?þ¸†®ÜÜ\ >\'NœÐu×]§«¯¾Úám-Z¤M›6Éb±háÂ…Z·nJJJôŸÿùŸúË_þ"‹Å¢ßþö·ÊÍÍUaa¡žxâ ]¸pAš9s¦ %I‹E/½ô’  1cÆèÙgŸÕÿ÷7ûõÿøÿЭ·ÞªÜÜ\:€9eeeiíÚµÊÊÊ2º€Eâ X$0/«ÕªgžyFçÏŸW=-I͆—·&))IIII-ö/X°@ ,h±õêÕZ½zu‹ýûöí»äµúöí«ÜÜÜKžÀ{ÚÛ2x:´=耹5†rpg´eÞ!îÐFt8Úˆ€gbˆ;´0·ââbÕÖÖªK—.8p Ñå@»Ð–y7:´‹Äæ–­²²2Y­V%&&]´ m™w# @у˜Û#<¢ºº:ùø0ƒ€û¢-ónth#:`nF—F[æÝøXÆ`Û·o×wÜ¡°°0­X±Bõ­Œ]=}ú´æΫ§žzÊá}<óÌ3JHHЮ]»œ]2àµèp&zÐ TUU¥ 6(99Y»wïVRR’‚ƒƒÓâÜ/¿üR‡R¿~ýšíÿöÛo5gÎåççËf³)""B;wîTïÞ½]õ0¯A@€3Ð ´oß>edd( @QQQ*((Pff¦Ã€>räH1B_ýu³ý‹/Vtt´,‹üüü®•+Wê¹çžsÕü‹Äæ–““£óçÏ«Gš8q¢Ñå@»Ð–y7†¸(""BöíáÇkÈ!­žßµkËÏS jß b˜;à$ô æf³ÙTUU%›Íft)Ðn´eÞtÙºu«^{íµV[šv¡ýÓÔ·o_ûöUW]¥-[¶8¥>ÀÛÐs›­ÜÜ\EEEI’BCC5{öl%''«®®NË–-c:à$tÀÜ&Ož¬šš‡ßz̻ñ·n ””¥¤¤8<¶wïÞfÛ£GÖÎ;ž«ØØØN¯@stÀÜ.öU¥à.h˼«¸@ÐàLth#³,G@ðL q€6jÐ]=-¬i:«¸Žåæ檪ªJ~~~öµZÀÝЖy7zРâoöî»ïêÆoTQQ‘}ßöíÛuÇw(,,L+V¬P}“Ojjjôì³Ï*%%EIII-¾…ÄŠŠŠtìرf5€»¡-ónô @Ðá­Nž<©êêjíÙ³G–稪ªÒ† ”œœ¬Ý»w+))IÁÁÁŠ‰‰‘ÔðM%C‡ÕôéÓµjÕ*=ÿüóJMMujÓ¦Msêý€+Жy7:´Þªÿþš0aB³}ûöíSFF†¥‚‚eff*&&FgΜђ%KôÍ7ßH’&L˜ ððpÍ›7O={ö4â!àâmÄ"qðf¾¾¾Í¶#""`ß>|¸ý« ¨^½zI’ ¤sçÎiÿþý.«wD@€62K:‹ÄÁŒ¶nݪøøxIÒÔ·o_û1___ 0€ù”\CÜ ÌÐéA‡Ùäåå)66VC‡•$Y,uùÑ’²²2ùûû;µŽ¥K—ª¬¬LV«U‰‰‰N½8 m™w# @Жl6›233•žžnß©yóæÙ·ËËËUQQ¡°°°‹ÞWBB‚z÷îÝl_\\œâââÚTKtt´ª««[ Çw⬶,++KYYYÍö•––vê5Ðqth£¦ÝÇÅ„˜ƒ³JOO×üùóåÓä?Å 7Ü ºº:ºöíÛ§ŒŒ (**JÊÌÌlÐ/^¬èèhY,ùùù)<<\+W®ÔsÏ=§;v¨¼¼Ü~nŸ>}ôÅ_ÐÑ&M{bï¼SZ´È¸Zpqô x>º"""šm>\«P5QPP ©S§Ú·ƒ‚‚´k×.IÒ]wÝ¥¤¤$-Y²D&LÐ… 4yòdçÑ4è1¯mG@.í•W^Qyy¹4}út£Ë€v¡-óntÙºu«^{íµû8 ¾}ûÚ·¯ºê*mÙ²E’tíµ×êí·ßÖÏþsmܸQûÛßäãÃ×Û£mš=þÙ˜‹Ä—6bÄUUUÉÏÏÏèR Ýh˼Ý$òòò«¡C‡¶8f±XÔ¥I÷æ¹sçäïïoß>s挞yæ-[¶L=ö˜Ö­[×ì| 5ßë-ÐÍŽtàÒÆŽkt Ða´eÞ€n6›M™™™JOOwx<22RGµo—””èú믗Ôð=ãhìS IDAT‰o½õ–^}õUÝ~ûíš8q¢Æ§G}ô¢×LHHPïÞ½›í‹‹‹S\\\Ç Ü CÜ݋Ĺ—¬¬,eee5ÛWZZjP5À]ÐM ==]óçÏouhzdd¤>lß.))ÑÈ‘#%5 ‹—$EEE)--Mï¿ÿþ%úòåË5jÔ¨Îyp[ qwô »GxæççÛÛkGxKn° h̘1*//×Ñ£GµvíZ;vLéééöP£}ûöI’*++uüøq͘1C’të­·ÚŒ“¤nݺiÒ¤I® pKt÷A@.íðáÃÚ¿³µÀÝЖy7zÐ ”ššª´´4¥¥¥Ù÷………)>>^ÙÙÙ VHHˆBCC5{öl%''«®®NË–-³ÏA¿ë®»tæÌ­X±BýúõS÷îÝ èh3æ »‰.-''Geee²Z­JLL4ºhÚ2ïF@7PJJŠRRRÛ»wo³íØØXÅÆƶ8Ïb±è¡‡rJ}ð|ÌAwô —öä“O]tm™w# ^Œ!îîƒEâ€Kã+‰xÚ2ïÆ[rÀ‹1ÄÝ}ЃàùxKx1zÐÝÀó1ÄðbÌAw,\ZVV–*++Õ³gÏ_s̻Ð/Fºû ¸´ÀÀ@?^=zô0ºh7Ú2ïF@¼sÐÝ‹Ä—mt Ða´e޷䀣Ý}ЃàùxK·ñå—Ò 74Ì•núÓµ«t×]RU•Ñºæ »:€ç# ÃmdfJû÷7„“¦?µµÒ{ïIo¿mt…î‡t÷Á"qÀ¥ëÿø‡Š‹‹.ڶ̻1n£¤ä‡ßÃÂ$éÂiß¾†}3fHÏ=gLmíqË-Òš5 #ŒÂt÷A:piÙÙÙ*++“ÕjUbb¢Ñå@»Ð–y7:ÜÆÙ³?ü¾e‹4|¸TS#ýË¿H§NI¥¥ ?îâèQéÁ%#סÝ}ÐK{ä‘GTWW'4nŒ¶Ì»Ðá6Îœùá÷¾}þìÚUzñE))I*/7¦®Ëuþ¼TYÙð{ÓÇdæ »Vq.-00Ðè Ãh˼n£i˜íÓç‡ßãâ~ÜÅÿ(=ýtÃïÕÕÆÖBºû Àóñ–n£qˆ{@€äëkl-Ñ´v£:sÐÝ‹Äx>zÐá6{Ї·»+3tzÐÝ=èÀ¥åääèüùóêÑ£‡&Nœht9Ð.´eÞ€·P_O@wæ »:pi6›MUUUêBƒÀÑ–y7:ÜBeåa–€ÞyèAw,\ÚäÉ“.:Œ¶Ì»Ð ¶}ûv-Z´HÅÅÅš1c†fÍš%KÓɦMÎ{çwT[[«)S¦èæ›onv¼®®N¯¼òŠz÷î­Aƒé–[n¹èu|PêÙ³SŠS5 ²MˆsGf èÌAwô x>ºªªª´aÃ%''k÷îÝJJJRpp°bbbš÷í·ßjÎœ9ÊÏÏ—ÍfSDD„vîÜ©Þ½{K’Î;§øøx-^¼XÁÁÁmºö¡Cþp\æ'?1º‚ŽéÚäÑ!îîƒEâ<Ý@ûöíSFF†¥‚‚eff¶è‹/Vtt´,‹üüü®•+Wê¹çž“$%$$è駟ns8—B¢;ö˜,MŸntc¦t†¸»zÐK+++S}}½,‹¬V«Ñå@»Ð–y7º"""šm>\-Î+((ÐÔ©SíÛAAAÚµk—$iÛ¶múøãõóŸÿ\“&MÒý÷߯xÀá0ù¦vï–Fê„Ëf¦€Îw÷A@.mÍš5*++“ÕjUbb¢Ñå@»Ð–y7º‰lݺU¯½öZ‹ýPß&+£]uÕUÚ²e‹$é/ù‹FŽ©{î¹GW^y¥bccuòäIÍš5Ëeuãò˜) Óƒî>X$¸´É“'«¦¦F]»òö€û¢-ónü­›D^^žbcc5tèÐÇ,K³¯Y8wîœ$I{öìÑý÷ß/___Ý|óÍŠÕ–-[.ÐìsØÅÅÅ)..® .Ƭ9èæFº{ÉÊÊRVVV³}¥¥¥Uã=† bt €Ó|˜8ÝÒÓÓ5þ|ù´’"##uøðaûvII‰FŽ)Iú×ý×fÇüüüíÜ‚Ñ!fêAgºû ¼ÝPI‡%•p탒¦ª¬ÌˆkÞ…€n° è¶ÛnSyy¹JKKõá‡***Jo¾ù¦î¹ç…„„(&&FÓÿ¹tyee¥Ž?®Õ«WK’5~üx-Z´HV«U¹¹¹ºï¾ûŒ|H¸3tzÐݸ´ÜÜ\UUUÉÏÏOQQQF—8½×Þ€¶Ì»Ð ”ššª´´4¥¥¥Ù÷………)>>^ÙÙÙ VHHˆBCC5{öl%''«®®NË–-“¿¿¿ýüU«ViáÂ…ºâŠ+ªØØX£ÚÀ¬9èæÆ"qÀ¥©¢¢Âþ ̻Ð ”’’¢””‡ÇöîÝÛl;66¶Õà}Ï=÷èž{îéôúàf èô ›=èÀ¥M›6Íè Ãh˼p13tæ »¦‹Ä½û®tûíŽÏ5JZ´Hâ›YÜoá3S@§Ý}tïþÃï%% ?ŽìØ!EFJ,Eà~xK¸˜Y:sÐÍmð`éþûÛvîñãέÎA:àbf èô ›ßÆÒÿ·ã9è~(Ý}wÃï§N¹¶.À,–.]ª²²2Y­V%&&]´ m™w# .f¦€Ît÷Ót¨{SW]õÃï§O»¦Àl¢££U]]-ߦ -¸Ú2ïF@\¬éâ]FtzÐ=G¿~?üN:¼Õˆ#Œ.:Œ¶Ì»ñ–p1³tæ »·¦t÷D@\Ìbù!¤›) ÓƒîÞºw—~§À=1Ä0€¯¯TSc|@gºgé×O*/§Þ+//O6›MݺuSxx¸Ñå@»Ð–y7:`__éüyã:CÜ=K` TTÔÐ.4ºc„†J¿øÿž½Õ§Ÿ~j_ù˜7µÜm™w# h\”ÓLt÷×8½®NZ°ÀØZŒ´qcÛ¿3žå©§ž2ºè0Ú2ïF@ @@‡3L˜ ýõ¯FWa¼ÂB:à Ž9¢²²2—_÷àÁƒ.¿&ïA@ `–€ÎtÏ2{¶ôÓŸJß~kt%®÷ÿ'%'7üÞôƒ'žéÈ‘# 1¸ «Á×à‰è€ÌЙƒîY,É[§ªõìùÃïtÀóýÐs¾AR¨X% 5àº<0€:=èpgMÿý6ïòÊ+¯¨¼¼\š>}ºÑåÀ%B%2º SÑ–y7:`³t†¸ÃS4Bº÷1b„ªªªäççgt)Ðn´eÞ€n°íÛ·kÑ¢E*..ÖŒ34kÖ,Y,‡ç½óÎ;ª­­Õ”)StóÍ7·8';;[/¿ü²>øàW”Ž0K@§ž¢é¿_º÷;v¬Ñ%@‡Ñ–y7ºªªª´aÃ%''k÷îÝJJJRpp°bbbš÷í·ßjÎœ9ÊÏÏ—ÍfSDD„vîÜ©Þ½{ÛÏ)..VZZšú÷ïïê‡vh è55Ò‹/J½zI“&IW\áÚ:˜ƒOA@w¾wß}W¿ýíoõúë¯kðàÁ’¤šš%%%©gÏž²ÙlZ¸p¡ºü³1¹Ø1à}fÚ·oŸ222¥yóæiÒ¤IÊÌÌlqÞâÅ‹-‹Å"???…‡‡kåÊ•ÍÎùýﯙ3gª¾¾ÞU壺԰òö£J?îú:èA‡§`ºsé´ûW¡-ónt°ÙlÊÌÌl5hGFFêèÑ£öí’’………é“O>ÑË/¿¬õë×ÛïÇf³©oß¾:sæÌE¯¹|ùrÅ÷†šA÷î ÐéA‡;3[@wôg~~¾ÂÃà ª¨óýøÃcI*++S@@€Îž=ë𘿿ÿEï³£ó•D<³Ú2oøðØÐM ==]óçÏ—O+ )22²Ù”’’9R£GÖÁƒíû7mڤ͛7kãÆN¯§i@¯¯—|ËžÓ4ƒîÊëEâ\/22Róæͳo———«¢¢Baaa8p æÎëðØÅðá187|xì è33Ø‚ 4fÌ•——ëèÑ£Z»v­Ž;¦ôôt{(‰‰Ñ¾}û$I•••:~ü¸f̘!???Ùúõë§îÝ»+((Èȇ„ËÔ¸¶_]뿽±§‘Þs¸;æ »Þ 7Ü ºº:>^ÙÙÙ VHHˆBCC5{öl%''«®®NË–-s8LÐb±4ûú¸‡Æt©¡½[7×]»1È0ÿîÎlCÜ=Qõ??A¬©©‘$uéÒE“&MÒîÝ»5qâDåääØ_Ï.vÌ™²²²TYY©ž={²¦ ·E[æÝèJIIQJJŠÃc{÷îm¶«ØØØ‹Þß´iÓ4mÚ´N«®ñã€~Å®»6=èðt窨¨ÐÚµke±X´~ýzÍž=[Zµj•ž{î9¨¦¦F¿ùÍoì·¹Ø1g ÔùóçùÚQn¶Ì»Ðƒý8 »Rã\]:ÜsÐËßß_³fÍÒ¬Y³Zì_¾|y«·i혳DGG»ôzà ´eÞ·å€Áš~8êêïBgˆ;<=èÀЃكÎwx ‰ž€!î€Áè@ÇуI*..Vmm­ºté¢]´ m™w# c:ÐqÌA‡$egg«¬¬LV«U‰‰‰F—íB[æÝè€Á˜ƒt=è¤GyDuuuòáSGnŒ¶Ì»Ðƒ1Äè8æ Cjøj"pw´eÞ·å€Áât=èÀð¶0=è@ÇЀ'`ˆ;`0æ Ç"q¤œœ?^=zôÐĉ.Ú…¶Ì»ÐƒÑƒtsÐ!I6›MUUU꧎Üm™w# c:Ðq q‡$Mž<Ùè Ãh˼0XÓ€¾y³tâ„ë®]VÖð'<0XÏž?üžŸßðãj]i à昃<o˃-]sôÅÆÕpÿýÆ]è ô C’ÊÊÊT__/‹Å"«Õjt9Ð.´eÞ€¬{wéÐ!éï—êë]ý^½¤Áƒ]] 3Ð!IkÖ¬QYY™¬V«.Ú…¶Ì»Ð ¶}ûv-Z´HÅÅÅš1c†fÍš%‹Åâð¼wÞyGµµµš2eŠn¾ùfIRyy¹ÒÒÒ”™™©[o½U¿þõ¯uà 7¸úa ƒ|}¥#Œ®p_KÃO}=Ý›Mžª½{÷ôˆÀ8>> óÏ èÞkÈ!F—F[æÝèÚ·oŸ222 ¨¨((33³E@_¼x±¢££e±Xäçç§ððp­\¹RÏ=÷œ®»î:ýìg?“$ |¸ÃOÌ jß Ò®]»$ÉÎ÷[­V 4ÈyE€IuéÒð'=èÀ]уn"[·nÕk¯½ÖbÿÔ·o_ûöUW]¥-[¶8¼}ll¬üýýZ'˜QãBqtï•››«ªª*ùùù)**Êèr ]h˼Ý$òòò«¡C‡¶8f±XÔ¥±kHÒ¹sç†ðmÛ¶iéÒ¥N­ÌŠ€Ž¢¢"UTTðA5·F[æÝè&`³Ù”™™©ôôt‡Ç###uôèQûvII‰®¿þúfçlݺU÷Þ{¯zõêÕ¦k&$$¨wïÞÍöÅÅÅ)..î2«sh èf˜ƒž••¥¬¬¬fûJKK ªÆ{L›6Íè Ãh˼ÝÒÓÓ5þ|ùø8^ 22R‡¶o—””häÈ‘öíÂÂB•””èñÇoó5—/_®Q£Fµ¿h03õ ;úÀ3??_áááUÜ‹ÄlÁ‚3fŒÊËËuôèQ­]»VÇŽSzzº=”ÇÄÄhß¾}’¤ÊÊJ?~\3fÌ$:tHkÖ¬ÑøñãuâÄ ýïÿþ¯V¯^mØ㣰Hpwô (55UiiiJKK³ï S||¼²³³¬…††jöìÙJNNV]]–-[&ëöÛoWqq±222$5ÌWÏÎÎ6ê!€aÌÔƒÐt¥¤¤(%%Åá±½{÷6ÛŽUlll³}ÔW_}å´úÀ˜i:Œ±téR•••Éjµ*11Ñèr ]h˼àèAGtt´ª««åëëkt)Ðn´eÞ€ðÌALj#Œ.:Œ¶Ì»±HÀ#ЃÜàèÀÝ1ÄàX$yyy²ÙlêÖ­ß9ÀmÑ–y7:À#0Ÿ~ú©}åcÞÔpW´eÞ€ð qÇSO=et Ða´eÞ9è@@ðÌAðô wÇt€G`‘8¼òÊ+*//W@@€¦OŸnt9Ð.´eÞ€ðô cĈªªª’ŸŸŸÑ¥@»Ñ–y7:À#Ð1vìX£K€£-ónÌAx‰ðsÐ%©¾Þ¸:Ú‹!îàÓä#纺æÞáðáê®®–¯¯¯BBBŒ.Ú…¶Ì»Ð ¶}ûv-Z´HÅÅÅš1c†fÍš%‹Åâð¼wÞyGµµµš2eŠn¾ùæ6oA@GNNŽÊÊÊdµZ•˜˜ht9Ð.´eÞ€n ªª*mØ°AÉÉÉÚ½{·’’’¬˜˜˜fç}ûí·š3gŽòóóe³Ù¡;wªwïÞ=Þ¤i@¯­•|}«ÆxòÉ'.:Œ¶Ì»1Ý@ûöíSFF†¢¢¢4oÞÿüóV€7i:¤€ÜÝD¶nݪøøøû8 ¾}ûÚ·¯ºê*I’<Øê1ð&ô wÇt“ÈËËSll¬†Úâ˜ÅbQ—&]Cçγ÷¼;:æïïïü‚ÀdèÈÊÊRee¥zö쩸¸8£Ë€v¡-ónt°ÙlÊÌÌTzzºÃã‘‘‘:zô¨}»¤¤Daaa­»þúë/yÍ„„„ ÉÅÅÅÑp[?^$ÎHYYYÊÊÊj¶¯´´Ô j¼G`` ÎŸ?¯=z] ´m™w# ›@zzºæÏŸ/Ç3"##uøðaûvII‰FŽyÉc³|ùr5ªƒ•€y˜iº£<óóónPEÞ!::Úè Ãh˼sÐ ¶`Á3Fååå:zô¨Ö®]«cÇŽ)==ݼcbb´oß>IRee¥Ž?®3f\òx†¸wGºRSS•––¦´´4û¾°°0ÅÇÇ+;;[ÁÁÁ Qhh¨fÏž­äädÕÕÕiÙ²eöyæ;Þ„€ÜÝ@)))JIIqxlï޽ͶcccëðÜ‹oa¦9è0Fqq±jkkÕ¥K 8Ðèr ]h˼àÌ4ÆÈÎÎVYY™¬V«.Ú…¶Ì»Ð!îxä‘GTWW×ꢫàh˼àè 4ºè0Ú2ïÆÇ2ÀtàîèAxzÐqüøqmÙ²EݺuSyy¹î»ï> >\555JJJRÏž=e³Ù´páBuiºPht€G`‘8cÌŸ?_“&MÒìÙ³uï½÷*!!ARÃW‰:T©©©ºúê«õüóÏ;½–œœmÚ´I999N¿8 m™w£àèA7ÆŽ;TVV&IêÓ§Nœ8¡³gÏjÉ’%úæ›o$I&LPxx¸æÍ›§ž={:­›Í¦ªª*zê¸5Ú2ïF@xº1î»ï>=ôÐCúë_ÿª¥K—êÅ_Ô¾}û¨^½zI’ ¤sçÎiÿþý=z´Ój™ €ÉÄÅÅÉÏÏO¿úÕ¯´nÝ:Ùiüøñš={¶òòò ¬²óL›6Íè Ãh˼=è&пM˜0á¢ç¼þúë;v¬$©W¯^ŠŒŒÔ¶mÛ$I;vìPy“åûôé£/¾øÂyÜÂرcuäÈ­\¹²Y8od³Ùt…Y—½À ÐMÂ××÷¢Ç¿ùæíٳǾý/ÿò/Ú´i“$é®»îÒ| %K–èÀºpá‚&»úËv¦“­Aƒµz|Ú´iúàƒ\X¸†¸»‰x@óS" IDATo¾ù¦æÌ™£^½ziÿþýêׯŸ$éÚk¯ÕÛo¿­ŸÿüçÚ¸q£þö·¿ÉÇÇñg/.\¤sØ!•––*??ßè2L‰çÆ1ž—ÖñÜ´ÔØîž?Þ¥×1c†~÷»ß©OŸ>ö0>nÜ8—ÖÚ†€î&æÏŸ¯¾}ûê‰'žÐý÷߯Ï>ûL‹/¶?s挞yæ-[¶L=ö˜Ö­[§.]º´¸ŸÆ¡ïS§NuYíî$<<ÜèL‹çÆ1ž—ÖñÜ8vâÄ Ý|óÍ.¹Ö³Ï>«ï¾ûÎþ•jãÆSZZš<¨™3gº¤WZºt©ÊÊÊdµZ~7;¸Ú2ïF@wV«UÏ>û¬$éý÷ß×É“'í= Ÿ~ú©Þzë-½ú꫺ýöÛ5qâD7N>úh‹û¹óÎ;µaà 2D=zôpåC¯váÂ}ñźóÎ;]vÍë®»NÙÙÙÍö…‡‡kîܹУ££U]]}Éic`f´eÞ€î†222”””¤ÁƒK’¶nÝjï©ŠŠŠRZZšÞÿ}‡=00P=ô+ËüÓM7ÝäÒë 8°Å¾ßýîwö)RžfĈF—F[æÝX$Î$ª««%I555ö}ééé:|øp³óRSSõ“ŸüDiiiö}·Þz«víÚeßîÖ­›&MšääŠf©û·ÓóÏ?¯9sæèúë¯×G}d‘Ì…€nz饗d±X´~ýz:uJRÃê»………’¤wß}W¯¾úª‚‚‚ôÊ+¯4»ý]wÝ¥_üâZ±b…6lØ îÝ»Ð5j”þüç?«¾¾^;wîTLLŒòóóu÷Ýw]pÀR___otpšš%%%©çÿgïÎÚ:óößaUAÀ­"àªm%ÕJk[—2b§8jeÔÚŽKß:.”NË VŠJ§ XÜx]j~ê`k•¢­Qk]@ÜEÁŠIXòûƒ1óR×F’'äÜŸëòº<‡“œ›Çð•ož'ç89A£Ñ`Á‚÷¼œµúá‡ðñÇcãƆ^jµëÖ­Ãœ9s0aÂÄÆÆ"##€´Çh›]»v!55³fÍÂŒ3pìØ1›;>ÿüs”””~ž8. ¯›Ë—/#++ YYYØ´i±cSWW‡ýë_øôÓO1oÞ<Ì›7}ô&NœØ¤ç±@vv68 : ‘ÑXË$NO’P^^®oÙ²¥¾²²R¯×ëõ.\зnÝZ_UU%8™ùh4½L&Ó—””èõú‰ÆëðáÃz¥RiØþóŸÿ¬ã7ô’½^¯¯­­Õ××׶õ‡’ükæŽ}ûöé׬Y£÷óóÓïÙ³‡ãò«V­Ò¯ZµªÑ>Ñc3lØ0½³³³¾wïÞúÐÐP}hh¨¾ÿþzOOÏ&;GSÈÍÍÕÐçæ抎BÍÄ× «ôú§ã2v´€,û‡?£ÖŠõ×òp]" áåå777€··7 Nœ8!8™ùüöV)Œ×sÏ=Ãö“O> ???Hz\ÀÑÑÑ03|ûöm¼ôÒKèÝ»·ä_3 R©°yóæFw‰à¸4X¹r%¦L™‚W^y{÷î ~lŽ?Ž³gÏâСCؽ{7vïÞ={öàÿý¿ÿ×dç ""¢¦Ã]"Nž< ö½½=Ú¶m‹’’©Äºß˜Kr¼ÒÓÓ1qâDŽËÿ‘••…AƒÁÁÁZ­–cà‹/¾À¬Y³íã¸48t芋‹ÑµkW 0………ÂÇfüøñ8}úô]ûýýý›ìDDDÔtxt‰Édw]xH©TÂÙÙYP"ñî7&...øõ×_%5^¹¹¹9r$üýý±k×.ŽËôìÙýë_ñÁÀÝÝ:u’ôØüðÃèÓ§O£{hëõzþ,ýwî´áää„Õ«W#00PèØøûûcáÂ…¸r忨ÒjµÈÈÈÀæÍ››ì ÈÏχ^¯Çøñã›ô<–"$$¡¡¡ …ˆÈh¬eÒ&ÓëõzÑ!È<ªªª0{öl´mÛZ­sçÎÉlªªª°zõj¼÷Þ{˜5kfÎœ //¯Ž‰µW||<âããí Baa¡¤Çh¸èÙk¯½___aøðá2d€ÿR›ÿ«sçÎHMMEÿþý%?.¥¥¥ Cuu5úö틘˜ûšùñljlß¾ååå˜={6Þ{ï=tïÞ½IÏõ8òòò —Ë‘››‹àà`Ñq¨¸óšrXêkÆÀU\yȱ–.F­ë¯åaƒNDDd¥ÆŒƒ×_EEE˜={6€†;6lÙ²_ýµàtÿÅ_é÷bƒnnlЭë¯åáw"""+%—Ë1f̸¸¸öåææbçÎS™ÎÙ³gqâÄ œ={Vt""£±–I/GDDd¥PUUeØþå—_’’‚wß}W`*ÓÉÈÈ€R©„««+¢££EÇ!"2 k™´±A'""²Rýû÷Çûï¿ãÇãË/¿Äõë×1dÈüõ¯Í$¬õû""ia-“66èDDDVªuëÖX¾|9ªªª““ƒ>}úÀÓÓSt,“á-‰ˆÈ°–I?ƒNDDdåœOOOTWW#55Ut$"""ºΠY©Î;7ÚÖétP( Ä„ ¥"""¢ûaƒNDDd¥üýý ™LfØ·mÛ6 6L`*ÓIKKCuu5œœœ):‘QXˤ :‘•Zºt)í{á…°fÍA‰LËËË 555hÙ²¥è(DDFc-“66èDDDVê·Í9888૯¾Â?þñ‰LkðàÁ¢#=6Ö2icƒNDDd¥Øh»¶¶G½k?Y6èDDDVªªª aaa†Ï ;88àÝwßÅèÑ£'#""¢{aƒNDDd¥–-[†Þ½{?ð˜«W¯¢cÇŽfJdZ×®]ƒN§ƒ­­-žxâ ÑqˆˆŒÂZ&mlЉˆˆ¬Tuu5rrrîûu­V‹ 6`åÊ•fLe:6l€R©„««+¢££EÇ!"2 k™´±A'""²R ,ÀÎ;xŒL&³šýÍ7ßD}}=lllDG!"2k™´±A'""²RNNN8wîºvíjØ÷¯ý r¹Ý»w‡N§ÃÒ¥K&lZ^^^¢#=6Ö2iãÛ2DDDV*  Qs!!!xûí·¶¶¶øŸÿùшˆˆèØ Y©Ë—/£¶¶¶Ñ¾ÔÔTܼyÓ°íèèhîXDDDt\âNDDd¥bccñôÓOã•W^AçαmÛ6ìß¿Ë–-Í$222PSSƒ–-["<<\t""£°–IgЉˆˆ¬Ô3Ï<ƒmÛ¶A¥R!-- ¾¾¾HOOÇäÉ“EG3 FµZ F#: ‘ÑXˤ3èDDDVÌßß+V¬€““ª««áää$:’ÉŒ1Bt"¢ÇÆZ&mœA'""²R§N«¯¾Š7ÞxPXXˆ  ²²Rp2"""º6èDDDVê‹/¾€ŸŸ}úô’’’'#""¢{aƒNDDd¥¼¼¼°|ùrøøøöi4¤¦¦ Le:J¥ …J¥Rt""£±–I?ƒNDDd¥BBBm«Õj,\¸ƒ ”È´V­Z¥R WWWDGG‹ŽCDdÖ2icƒNDDd¥zöì‰7ß|*• ¹¹¹HOO‡Þ{ï=ÑÑLbĈÐjµ°³ã¯7DÔ|±–IÿÕ‰ˆˆ¬T·nÝššŠììlüøãX»v-^}õUØÛÛ‹Žf~~~¢#=6Ö2icƒNDDd¥&L˜€‘#Gbøðá8p è8DDDôlÐ%æÖ­[رcüüüвeKÑqˆˆ$£¦¦ÅÅÅ:t(¼¼¼ÌrÎcÇŽaæÌ™wí¿qãÚ·oo– DDDôèØ KÌŽ;0nÜ8Ñ1ˆˆ$kݺu;v¬YÎõå—_âÇ„‡‡llnÜR[[‹ääd¤¤¤˜%ƒ9eggC­VÃÑÑ¡¡¡¢ã…µLÚØ KLçÎ4ü‚(8uŠŠŠBrr²èV‹ãkZ_Ó9uêÆg¨Ãæ0yòd\¼xsæÌi´_&“Yeƒ^RR‚ªª*8;;‹ŽBDd4Ö2icƒ.1-Z´"88Xpëäîîα5!Ž¯iq|MïN6‡‰'¢_¿~èر#lmmz½Ë—/7[sš0a‚èDDµLÚØ YLŸ>³gÏÆ´iÓàììlhÎïøíŒ:Y6èDDDVäùçŸG||<à³Ï>ÃÑ£GÑ«W/Œ?ÞpëWWW ‰ˆˆè~Ø Y‘îÝ»þ>gÎ<óÌ3Ø´i“á"qDDDd¹Ø 5±ÈÈHѬÇ×´8¾ÍŸ½½½áï2™ r¹ü®æüرcxöÙgÍÍä.\¥R WWWDGG‹ŽCDdÖ2icƒNÔÄØà˜Ç×´8¾Í_EEΞ=  á‚p·oß6l ÷c_´hÖ¬YÓ¤ç=}ú4¶mÛ†®]»¢_¿~pwwGll,œœœ Ñh°`Á‚»> ßÔŒºººFoR57¬eÒƈˆÈŠ¬Y³æ®æ{Ë–-¶e2Y“6èëׯÇùóçgØ÷ÑGÁßßS¦LAJJ æÍ›gøl¼©ôìÙÓ¤ÏODd¬eÒƈˆÈŠüéOÂäÉ“áààpϯWUUaÙ²eMv¾3gÎ -- †}HJJÂõë×aaaË刉‰““S“›ˆˆÈÚ°A'""²"±±±½—½——W“oÒ¤I4hâããqõêU|ôÑG¸xñ"¼¼¼àææðöö†B¡À‰'лwï&;7‘µá%]‰ˆˆ¬ÈÚs áVlM¡²²Àk¯½†¿ÿýïÐëõËå(,,„‡‡‡á8{{{´mÛ%%%MrÞûÉÍÍÅ››kÒó™k™´qˆˆˆŒräȸºº¢W¯^€O?ý«W¯†Ý]„S*•pvv~àóEEEÁÝݽѾÈÈÈG¾xá W>–Ëå¿ã;!"²¦ªeiiiHKKk´¯²²²ÉžŸšôfà‡~ÀÇŒ7Â××÷žÇìܹÛ·o‡N§Ã¨Q£ðâ‹/š9%IZ­n´íéé ___ÔÖÖ¢¨¨È°_¥R¡ªª AAA|¾äääGZp?Ó¦M3ú±DD–ÂTµì^oxæååñ M ÃÝ•••¡®®‡†L&»ç17nÜÀßþö7äååA£Ñà¹çžÃÞ½{ïš… ""jJO?ý4nß¾[·n>×îèèˆAƒaÎœ9(++C›6mPZZ WW×û¾ÉLDDD øt צM„……=ð˜Ï>û ƒ†L&ƒ££#äry“^¡—ˆˆè^:uꄱcÇbÕªU€«W¯ÂÎÎAAAˆˆˆÀÁƒHHH•ˆˆ¨Y`ƒÞ ØÛÛ?ðëùùù 4lûøø`ÿþý¦ŽEDD„%K–àöíÛHLLÄüùó±}ûv@JJ ²²²ðÉ'ŸàöíÛ\~NDDô¸ÄÝ œ‹£Gš0YºÿÎœ¯h泟0NØì=Y6èô»üÏÿü‚ƒƒ±}ûvdffbèСxòÉ'1mÚ´{¯×ë1fÌÌœ9‹/ƪU«ðÎ;ï wïÞ¤¥¥aÎœ9رcºuë†7ß|¯¿þ:ŠŠŠàîîŽÊÊJ„‡‡#11C† Á0nÜ8¼ôÒKÈÍÍ…J¥B«V­°oß>ØÛÛcÒ¤Iøãÿˆ–-[ÂÓÓ;wîD¯^½Ð»woŒ3æo&QóuèÐ!Lœ8ÕÕJ¤¥‹ŽCDd”´´4TWWÃÉÉ ‘‘‘¢ã™±A§GVUU…ÌÌLL:0`À8::¢¨¨è¾9}ú4N:…¾}û† †ÇTWW#** &L@·nÝ“'OFVVÎœ9ƒçŸ111pqq1<6$$ãÆÃüùó‘““ƒ`øðá€N:!22²Q1kÓ¦ 1eÊ”&jj·oßÆ… PZZú»[]]£G¢M›6ðöö6A:"²t...(++CM /IDÍ———jjjвeKÑQH6èôÈœ‘››‹ž={B­VcëÖ­ptt„B¡¸ïcqøða<ýôÓP©Tعs'SPP€²²2Ãl:¼ôÒK¸víša;++ ¡¡¡žwèС˜?>¾ùæ 0vv /å.]ºÜ•ÁÎÎîžûÉòìÝ»×ðfËïuöìYcúôéX¼xq'#¢æ ((éééâDG!"2ÚàÁƒEG xã`ú]ñÉ'Ÿ`ìرhÕªZ·n ½^ÿÀÇtëÖ ûÛßðî»ï"0°á3wSRRÐh4÷|¬B¡ÀÅ‹qóæÍFûûôé777:tèq¿%""""""‹À™R©DXX*++±iÓ&¼úê«}Ìõë×ñÒK/ÁÇÇk×®EHHH£¯wèÐ@Ã,ùo]»v ­ZµB=°ÿþF_³³³Cûöí!—Ëã;"K¡Õj‘““ƒÂÂÂÇ~®«W¯"''eeeMŒˆˆˆˆÈ|Ø Ó#Û¸q#²³³kا×ëQ__ßǤ¤¤ ¸¸3fÌh´ÿÎczöì lß¾½Ñó=zÔð9ä?ýéO¨¬¬lÔ¼ÕÕÕáêÕ«4hP“|o$VUU Ðèµe¬Í›7cÀ€ÈÎÎ~ü`DÔ¬TVVÂÛÛ:TŠŽBDd´k×®áòåË>òIÒÁYÇŽ4|NX¯×cË–-øõ×_QUU…Ë—/£®®îžQ©T8zô(êêêšš  ¡!»s•öØØXbÔ¨QÈÎÎÆ¢E‹°qãFôêÕ 0~üxxxxàý÷ßG]]4 æÏŸ#F`Ô¨Qþ»Dþ^S«ÕF]tŒˆˆš—ÇcÒ¤IˆŒ<,: ‘Ñ6lØ€¯¿þ6l…`ƒNlÈ!˜¿‹7n ;;Ÿ}ö´Z­©‡‰~§¢¢"œ>}ºÉŸ÷Ê•+8yò$ÿ͉$$$$Ë–-CjjÈÃ&"²Po¾ù&Þ}÷]¼ù曢£2ýîðEV%//r¹¹¹¹æ=bI¼§Ÿ~&{þëׯ£}ûö&{~¢GÅú{M56wžÈ`î1ÎÀ_sûoý¨¼\ÐÀÁY_ßÖŠÿ7YΠYÞˆ„P(¨©©1ùô²²2ØÙÙÁËËˤç!""""z\œA'"!bccѾ}{œ:uʤçyúé§ïº½Y§ãÇcÔ¨Q?.: ‘Ñ222ðïÿ¢£œA'"""« ÓéàèèN': ‘Ñ4 Ôj5lmmEG!Ø ‘YÕ××C¯×ÜקÔëõÐétÉd°±áÂ!"kŒÉ“'xOt""£1Btˆ¿©‘Ymܸvvvøßÿý_³³¨¨vvv˜4i’ÙÎIDDDDô{±A'""""""²\âNDDDV¡¶¶­ZµB}}-T*ÑiˆˆŒ£T*¡×ë!“Éàêê*:™gЉÈ,®]»777üå/–aݺupssúuë„e "ÓÙ»w/¢££1eÊ^ÑQˆˆŒ¶jÕ*|ùå—Xµj•è($gЉÈ,ôz= …Ð uuu†?Dd}zõê…©S§B§K…ˆÈh#FŒ€V«…[5)â¿:Y///ð…ˆÈh~~~¢#@\âND&7`À„„„ˆŽaðᇢK—.(..…ˆˆˆˆÈ€3èN«Õ"66NNNÐh4X°`lmmï:nÍš5¸}û6êëëÑ¢E L:U@Z¢{»|ù2._¾,:†Á­[·pëÖ-hµZÑQˆˆˆˆˆ 8ƒnáâããáïïøøxtêÔ óæÍ»ë˜S§N!++ QQQˆŽŽÆîÝ»qðàAi‰ˆˆÄ9sæ †ŠΈŽBDd´ììlìرÙÙÙ¢£lÐ-XEE’’’ðÆo°dÉTWW7:n×®]PýŸûÉ´nÝ/^4kV¢{‰‹‹ÃsÏ=‡«W¯ŠŽrOxõÕWEÇ ¢&R^^Ž®]»Âϯ\t""£•”” ¨¨%%%¢£\ânÁ áåå777€··7 Nœ8Þ½{Ž{õÕW‹¤¤$„……¡¶¶#FŒ›È ¸¸¹¹¹¢cÜWaa!ÊËù‹<‘µxá…0}út–[wˆˆf„ ¢#@lÐ-ØÉ“'áááaض··GÛ¶mQRRÒ¨AïÚµ+¶nÝŠaÆá›o¾Á/¿ü.Ž """""jNØ [0™Lv×á”J%œï:¶¢¢ï¿ÿ>¾üòK¼ýöÛøúë¯ïy1¹;¢¢¢àîîÞh_dd$"##›&}¹¹¹¢cY”Ï?ÿ%%%Éd€øøxøûû#>>:u¼yó'$""²||[†ˆjúôéرc‡ècÒ¤IÉd|Âè?öïßooo´nÝ@ÃÝ’’’ ×k ƒ\.GLL œœœDF%""²hœA'"""£©T*lÞ¼o½õ–a_aa!¼¼¼àææðöö†B¡0ù­ :„‰'"2òIÏCDdJ‘‘ µìÐ!Ö2)bƒNDDDFûâ‹/0kÖ¬FûNž< ö½½=Ú¶m‹’’“fqqqAYYnÝr1éyˆˆLéÖ­†ZæâÂZ&ElЉè¾>ùäØÙÙqyû=èõzØÙÙáå—_…H˜~ø}úô§§§aŸ^¯‡L&ƒ­­m£c•J%œMš'((éééÈÌ 2éyˆˆL)3³¡–±–I?ƒND÷¥×ëÍrååæJ§Óq|HÒ.\ˆ#GŽ¶ ÂÃÃñÁ ¨¨È°_¥R¡ªªê¡¿lFEEÁÝݽѾÈÈHDFF6mp"" JKKCZZZ£}•••‚ÒÐý°A'"""£¬[·jµ@ÃzýúõCRR „ÄÄD”••¡M›6(--…««+|}}ø|ÉÉÉ6Gt""ɹמyyyËå‚ѽp‰;ÝåرcèׯV®\):ŠÅãX‘”µk×>>>ðññ¯¯/ìììо}{xxx ""ddd !!Áäy*++áíí8#DDÍW‡ µŒ³ÛÒĈî¢P(°ÿ~\ºtIt‹Ç±"º·””deeá“O>ÁíÛ·1mÚ4“ŸóðáØ4i"#›ü\DD¦ÙPËf-“".q·`Z­±±±prr‚F£Á‚ îºèÎõõõX¹r%ÜÝÝáíí~ýú™9-IÝÅ‹ wvvFrr²YÏ‚‘#G¢¾~“YÏKDÔ”RSC`c3›6±–IgÐ-X||<üýýN:aÞ¼y÷]ºt•¬ÌÂ… EGh–<ˆƒbèСèÞ½»è8DDDDÔ qÝBÂËË nnnooo( œ8q¢ÑqYYYØ·oÊÊÊ´´4èõz‘‰ˆˆˆˆˆè1pÝB>>8räˆè(V% ááá¢cQ3Ât –’’‚Ù³g#??Z­sçÎlØ°]ºtA@@‚‚‚’’‚  U«V ÄÈ‘#''""2¿ÚÚZ´jÕ õõµP©D§!"2Ž‹K-llZ¡¶¶Vt€ ºsvvFrrò]û;ÓùÚk¯áµ×^3W,"""‹´wï^DGGC¡Ø‹… _‡ˆÈ(S¦ìE«VÑØ»w/^xµLj¸ÄH¢víÚ…'žx‹/Åjåççã‰'ž@\\œè(D’ЫW/¬]»›7÷…ˆÈh›77Ô²^½Xˤˆ3èD¥V«qýúuÑ1¬Z]]®_¿…B!: ‘$xyy¡¨¨€—è(DDF+.öP//Ö2)â :‘à :‘ÄÔÔÔ`ìر¸qã†è(’ñý÷ߣ¤¤~ø!úöí+:Y(6èD£ÓéðÝw߉Ž!)EEE(**¸qãDG!²jgÎœÁСCQ[{{ö‹ŽCDd”ΠE‹¡8sæ ‚ƒYˤ†K܉ˆˆÈ*”——£k×®ðó+…ˆÈh~~ µ¬¼œµLŠ8ƒN$!ÿüç?qùòeÑ1$kË–-8þ<¦M›'''Ñqˆ¬Î /¼€éӧȅˆÈh©©/˜ŽÜ\Ö2)bƒN$!‹-±cÇDǬµk×Þ~ûm6èDDDDt.q'""""""²œA'’€óçÏãêÕ«P©T¢£€Ÿþ:t@Ÿ>}DG!"""" ÂH/^Œ%K–ˆŽAÿñú믣{÷î8}ú´è(DV%33ü1ŠL,\È+Qó‰V­>Fff&¯â.A\âNDDDV!00éééÈÌ …ˆÈh™™ µ,0µLŠ8ƒNdÅt:´Z-t:è(ôõõõP«Õ°±±½½½è8DVÁÛÛyyy¼EG!"2ZA7€úˆŸYj†þùÏâçŸÆ‚ àåå%: Àt"+³aì]»Vt úvïÞ+V@©TŠŽBDDDD‚°A'"""«““ƒèèhL™’#: ‘ѦLi¨e99¬eRÄ%îDVbß¾}8þ< …è(ô6mÚ___Œ=Zt¢f§cǎغu+Ôê¾¢£­  #·¢o_Ö2)â :‘•øúë¯ñöÛoãæÍ›¢£ÐcøðÃ-:Q³ÔµkWdggãÀ®¢£íÀ†ZÖµ+k™±A'""""""²\âNÔÌUVVâƸ}û¶è(ÔDêêêpúôi´jÕ O<ñ„è8DDDDd&œA·`Z­|ðâââ N÷Àã7lØ€š)YŠï¾ûؼy³è(ÔDnÞ¼‰ÀÀ@üío…¨Y)--EPPJEG!"2Z@@C-+-e-“"6è,>>þþþˆG§N0oÞ¼û{íÚ5$$$@&“™1!‘åÈÏÏÇèÑ£ž/: ‘ÑÂÃjY~>k™q‰»…ª¨¨@RR®_¿ ƒ\.GLL œœœî:þóÏ?ÇÔ©S±qãFsG%""²Dÿþý¡×óÖDDÔ|-]:2YÞfM¢8ƒn¡ áåå777€··7 Nœ8q×±_}õƇ–-[š;& tḹ¹aêÔ©¢£‰lÚ´ nnnX½zµè(DÍ‚Ôj54Î?Qó¥Ñ4Ô2;;Ö2)bƒn¡Nž< ö½½=Ú¶m‹’’’FÇ?J¥Ï=÷œ¹#’`õõõP(¨­­…L¤®® …Ft""""2¾-c¡d2lmmíS*•pvv6lët:¤¤¤ 11ñw?TTÜÝÝí‹ŒŒDdd¤q‰ˆÈ -- iiiöUVV JCDDDÍt Õ·o_ÄÄĶU*ªªªdØ÷óÏ?cÅŠX³f @£Ñ@£ÑÀÃÃ|þääd›&<™\dd$®\¹":™É’%K°eˬX±~~~¢ãÐ#¸×žyyyËå‚IáC‡0qâDTWBZÿ#¢æ)2òœœ&âСCü}]‚Ø [¨=z ¾¾eeehÓ¦ JKKáêê ___Ã1}úôÁ©S§ Ûÿþ÷¿±iÓ&|óÍ7""“íÙ³ÇpA²~§N©S§ R©DG!²h...(++CM‹è(DDF»uË-[–ÁÅ…µLŠøt ekk‹ˆˆÌZ&EœA'j2220}útÑ1ÈTUUaØ°a>|8ÒÓÓEÇ!²(!!!9r$êë7‰ŽBDd´ÔÔØ،ĦM¬eRĈˆˆ¬ÂÛ¬¼55_åå.x›5©âw" ¦ÕjQQQêêjÑQÈÂh4TTT ¦¦Ft""""j"lЉ,؉'àé鉘˜ÑQÈÂìرžžžXºt©è(DDDDÔDØ ‘U8~ü8F…ðð㢣-<¼¡–?ÎZ&Eü :YNGGGèt:ÑQˆˆŒæàÀZ&eœA'²PÏ>û,†.:Y¸Ï>û þþþ8vì˜è(DÂcݺuؼ9Xt""£mÞÜPË‚ƒYˤˆ3èDêÂ… P*•¢c…+//Gyy9jkkEG!"""¢ÇÄt""""""" ÀÈÂ$&&büøñœ¥ß%>>o½õ–è$A;wîÄ Aƒ„E‹A¯×h¸Mä|€¸¸8ÄÆÆšå³”µµµhÕª\\X?‰¨ùrqi¨eü]PšØ Y˜Ý»wcݺu¨««…š‘üëׯƒ$F­Vcݺu˜3g&L˜€ØØXdddhxÓÈßßñññèÔ©æÍ›gò<{÷îEtt4¦LÙkòs™Ê”) µlï^Ö2)âgЉˆˆÈ(X²d \\\Šüü|¬_¿ýúõCRR®_¿ ƒ\.GLL œœœL–§W¯^˜:u*tº“ƒˆÈÔ6oî[Û©HIa-“"6èD"77¿üò JJJDG¡fª¾¾Ë–-CÇŽñú믋ŽCðÜsÏ5Ú~òÉ'QUU…‚‚xyyÁÍÍ àíí …B'N wïÞ&Ëãåå…¢¢"^&;‘©{(‚—k™q‰;‘…ÈÌÌÄ´iÓpâÄ ÑQ¨™ª¯¯Ç´iÓ°xñbÑQH¢ÒÓÓ1qâDœúè# 0*• •••Ø³gBCCƒ"<<HHH0y–^xÓ§Okòs™Jjê ¦#7—µLŠØ ‘Qâãã‘Шù Âĉ‘’’‚Ù³g#??Z­sçΘ”ˆˆ¨y`ƒN$½½=êëëEÇ +5þ|$$$`ÇŽxå•WDÇ!+‡¸¸¸{~ÍÙÙÉÉÉfNDDDÔ¼±A'H§ÓA¯×‹ŽAVJ¯×ó5FDDDÔŒð"qDDDd233ññÇ#::St""£EG7Ô²ÌLÖ2)bƒN$À´iÓЯ_?Îl’YDEE¡_¿~¨««…Ȥ‘žžŽÌÌ@ÑQˆˆŒ–™ÙPËYˤˆKÜ-œV«Ell,œœœ Ñh°`ÁØÚÚ6:F¥R!!!ëׯGÿþýñ÷¿ÿ=zô”˜EAAöïß/:Iĉ'€o‘ÕóööF^^oÑQˆˆŒVPà ÞÞ¬eRÄt ÄÇÇ£S§N˜7oÞ]Çüë_ÿ‚¯¯/¾þúk\¹ro½õ–ùƒÒ#Y½z5BCCqüøqÑQH‚Œwß}Wt """"ºΠ[°ŠŠ $%%áúõ뀰°0ÈårÄÄÄÀÉÉÉp\·nÝðòË/|}}ѽ{wœ9sÝ»w’›îïâŋسgè$Q999¨­­ƒˆˆˆˆîƒ3謰°^^^pssаtO¡P–«Þq§9¸ººrIŒ…9rä „õë׋ŽBwúôi 4+V¬…¨É•”” $$ry‰è(DDF“ËjYI k™±A·`'Ož„‡‡‡aÛÞÞmÛ¶}àkzz:FŽ gggsD¤GTQQ;w¢¸¸Xt’8…B;wâܹs¢£5¹ . 44!!DG!"2ZHHC-»pµLŠ¸ÄÝ‚Éd²».§T*Ø|geeaáÂ…}(¸»»7Ú‰ÈÈHãÂÒ=©ÕjüñÄ­[·DG!jäÛo¿Eaa!âââ":ŽÕIKKCZZZ£}•••‚ÒHÇÀ!—Ë䊎BDd´¥K#7—µLŠØ [°¾}û"&&Æ°­R©PUU…   {ŸžžŽ?þñ†%ñ’œœŒààà&ËJ÷V__;vˆŽAt—ââbó¢q&r¯7<óòòþÓ<Ý—¸[°=z ¾¾eee€ÒÒR¸ººÂ××÷®c QZZŠ?üáæŽI÷…#FˆŽAô@ ,@xx8”J¥è(DDDD’ÇÝ‚ÙÚÚ"""ddd !!˜˜ˆ³gÏÎœ9ƒU«VaÈ!(..ÆÑ£G±|ùra¹©Á¾}ûðã?ŠŽAô@GŽÁ¶mÛPWW': ‘äq‰»…KIIÁìÙ³‘ŸŸ­V‹¹sç6lØ€.]ºÀÕÕ¯¼ò ®]»†%K–høìú† DÆ–´Õ«Wãûï¿GQQ‘è(DlüøñèÒ¥‹¡Ž5G999ˆŽŽ†J•ƒ•+ù1."jž¦LÉ‹K4rrrø‘T bƒnᜑœœ|×þ#GŽþ~åÊsF¢û¨ªªBYY~þùglݺUt¢ßeûöíèÖ­Š‹‹áêê OOOÑ‘ˆ~·Ž;bëÖ­P«ûŠŽBDd´‚‚ŽptÜŠ¾}YˤˆK܉šÈŽ;йsg|ýõ×¢£åüùóèܹ3fÏž-: ‘QºvíŠììl8ÐUt""£8ÐP˺ve-“"6èDI¥R!**Š9YŸþQQQVê‘éq‰;ÑcÐëõ¨®®Æ¢E‹DG!j2(((ÀsÏ=g¸-˜L&œŠˆˆˆÈúqè1¢]»v¢c™ÄøñãaccƒË—/‹ŽBôHJKK„€€RÑQˆˆŒÐPËJKYˤˆ :‘233ñå—_â×_…Èä¾úê+|õÕW¢c=T~~>Fðð|ÑQˆˆŒÞPËòóYˤˆK܉Œ°qãF6,$óçχ&Mš$: Ñ 8ýû÷‡^Ÿ#: ‘Ñ–.™¬?rrXˤˆ3èD¿Ã·ß~ ___¤¥¥‰ŽBdVW¯^…¯¯/fÎœ): Ñ}ÙÙÙA­VC£áü5_MC-³³c-“"6èD ªª ß}÷rrrpéÒ%TUU‰ŽDdV:—.]±cÇðÝwßáâÅ‹¢#Y6èDàæÍ›1b/^,: ‘P9991b233EG!"""²:\7Aô×®]Ãùóç±gÏ( lÛ¶ -[¶‹È¢$&&âóÏ?Çþðh4üùφ§§'žzê)ÑÑHÂ:„‰'¢ºúÒÒ‚EÇ!"2Jdä!89MÄ¡C‡ÌZ&5lÐI’®^½ŠK—.açΨ©©Áwß}9r@Ãý͉èþÎ;X²d `ÅŠî—Þ­[7ØÚÚbÈ!Ðh4xóÍ7áæ憠  ayI\\\PVV†šÑQˆˆŒvë– Z¶,ƒ‹ k™±A'«uãÆ Ü¼y?ýôêêê°víZ´jÕÊЄët:Á ‰¬^¯74ï§OŸ,_¾666°±±AçÎáää„W^y2™ ãƃ‹‹ ºuë&26Y‰   ¤§§ˆ…ˆÈÕ‘7ƒ3IDATh™™AÒÇZ&ElЩYS*•¨®®FVVªªª°fÍ8;;ã—_~^¯GuuµèˆD ¾¾õõõ†æýøñ〤¤$ØÙÙÁÉÉ ¾¾¾ðòòBhh(ÜÝÝ'''xzzŠŒNDDDd6lÐÉâ©T*ÔÖÖbß¾}P«Õؼy3ììì°oß>ÔÔÔ ¬¬LtD"z Z­ …€Ý»wfÎœ‰-ZÀÛÛ;v„ŸŸ‚ƒƒáéé‰þýûÃÙÙ"£5)6è$”V«E]]Ο?Û·o#33-Z´@FFpâÄ ¨Õj( ÑQ‰H€ÚÚZœ?ÞpáÆÔÔTÃ×ìììЮ];¸¹¹á‰'ž@=ТE „††ÂÙÙr¹666pttø9UVVÂÛÛ:]%®_†ˆÈ8:TÂÖÖ•••¢£lÐÉd4 ÊËËqéÒ%\¾|G…^¯Gff&<<‰N:ñjºÍ@HHFŽ‰úúM¢£-5566#±ik™±A§ûÒét8yò$JJJ3gΠ¨¨W®\AMM ´Zí_[[k¦¤DDÆ«ªªÐ0û Ë–-¬Y³ÆpŒt:zöì ™L†?üáðññÁ‹/¾¸¹¹™?8ÝåÎmÖ¾™BDÍWy¹ ÞfMªØ €†[’8q{öìÁ•+W°k×.ÃòQ""©»ó†d~~>€ÿ^…Ú´iFƒˆˆèõzŒ?:tà}߉ˆˆèwcƒ.QÛ¶mCll,lmmñÓO?ÁÆÆuuu¢c5;wî$ñÏþšš [[[ØØØ 88vvv3f nß¾-0%5lÐ%jîܹ¶u: $DDÖG§ÓA§ÓáàÁƒ€ýû÷ N$ ÇǨQ£PSsÁ¢ã%<ü8Z¶…ãÇ#8˜µLjØ Óÿoï:ï6Ž_»ˆÆô EÓVD§+j;ŒbÒ©Ò4µv¢-‘&}"S5i¬Ô’ %¨«iÑhª±¾4jj'N`Œ}2„±©iSµÚ@«(Ód)††ìîýüì°¼‰°»çÀ~?3ŒsÎ}öìåsßË]vèÜn· Ä/ôi²–…2tÐ/$$$èÉ'Ÿ”ôK«£@9’ éIýò—¬e¡Èiu@ƒ€-ðw›s¹\ÊÊÊÒ!CÔÔÔ¤uëÖ),,¬Ýq………:vì˜Ün·~üãëþûï· -ÖillÔðáÃåñ4ª¡Áê4Ð3Æ5Êé®ÆÆF«£À<ƒns999ŠUNNŽÆŒ£µk׶;æêÕ«zöÙgµqãFmØ°AK—.Umm­i°Î‰'”‘‘¡%KNXzlÉ’æµìÄ Ö²PDƒncׯ_צM›ôè£J’æÌ™£­[·êæÍ›>ÇmØ°A>ø ‡ ¤ÄÄD½üòËVDÀ2ñññÚ¿¿Ž‰·: ôØ‘#ÍkY|Ç+..λÃgîBNTT”ÊËËuùr”ÕQ Ç._n^Ë¢¢XËB º•––*22Ò»®Q£F©²²²ËãFÝñ&q6æp8Ú½!\}}½†ÚåquuuíŽiËé|[ÆL–Ó&Ç-§Ó)cŒ÷|Gaaar¹Ü k?æt†Éív),l€<·§$#c$§Ó!·»ùö]Ûíîx¬«L·kÎ&Ç#‡Ã!I2ÆÜsw™©§÷{»ÿKóý:åñ5GrȘö5¼Ós÷¦†grÞëÙ÷µ75äZãZëïךÛý¤9è º%%%iÅŠÞ톆ݸqC“&Mjw\YY™w»ªªJ“'OîòÜÏ:Iÿ#·»y»ùßÿ½õÕÌåj=æË÷vÒ­ŸE;ëèöwrÝi¦žŽYU'Õ°«Lý½†\kÝ;7×ZOj˜{ë«õ>Þ¼3Ð.\¸ Ù³g«±ñ‚Þ}7Áê8!ãüùó!qŸ@°ÌœyAwÝ5[.\PBkY¨¡A·±)S¦Èãñ¨ººZ#GŽTUU•"""4vìXŸã’’’tñâEïvUUÕmßTbäÈ-8 áû OIjj:«êêDk℈šš?^7nÔèÝw­N "$Iiii–gú“qãj4tèxÕÔÔX A·±°°0¥¤¤èäÉ“z衇tôèQ­_¿^’´qãFÍ›7O&LÐܹsµdÉIÒÍ›7UQQ¡;vtyîÿû?‰_È@ðœ=+%ÒŸÔ}÷ݧeË–I:cu”+颤z‹î?âV yõÕû$-Ó™3¬e¡ˆÝæ¶oß®ììlËåriõêÕ’¤C‡é+_ùŠ&L˜ ¸¸8¥§§kåÊ•òx<Ú¼yómÿ ÷hÀŸhÐmnèСڲeK»ýÿüç?}¶SSS•šš¬X?ãcÖ°tÐ/hÍš5ÊÈ(°: ôXFFóZVPÀZŠhÐ@¿§üü|ÄYz¬  y-‹‹c- Eü :袣£uöìYIÑVG€ûðÃhIgÍZŠx AÀhÐ@¿PYY©3f(1±Òê(Ðc‰‰ÍkYe%kY(¢Aü,77×êýõ ,ꋾ¬¢¢B³fÍÒŒA¸·¾6WÈXä ¤¾÷ØÔ»¼3f4¯eÁXË`74耟õ½‘¾…úõ…?¹\.effêùçŸWVV–Ünw@ï/99Y¿ùÍo´m[r@ï§Y_›+ä ,òRß{lê]Þmۚײää`¬e°t999ŠUNNŽÆŒ£µk×Z [ãcÖ€ß]¿~]›6mÒ•+W$IsæÌQbb¢V¬X¡!C†Xœ{âtàw%%%ŠŠŠÒˆ#$5Fy]]Î;gq2ì‹gÐCLcc£$éüùó'é¿jkkuöìY«cô[Ô7°¨oà´¬»Ÿ}ö™ÅI‚£´´T‘‘‘Þíððp5J•••º÷Þ{}Žõ×cÓÁƒµ`Á}öÙA9Ò«SuC­¤Ös¥9û±cÇ‚þ{éÒ%Ÿ k›×îúCÞ¦VÿÚíÿr§õµîú–¤>úHê}vo^u¦w×ïÃÔàÁ tðàÁŸ£»Bí±©/pcŒÕ!<PZZšÕ1 d½öÚkZ°`Õ1nçÎÚµk—Îœ9ãÝ¡×_]sæÌñ9–Ç&°V¨<6õ4è!æ¿ÿý¯Þ~ûm7Nƒ¶:„ŒÆÆF]ºtI³gÏVTT”Õq®¨¨H3gÎTmm­$©¡¡AÇ×¥K—4vìXŸcylk„ÚcS_@ƒüÎívëî»ïVyy¹FŽ©òòr%$$èÓO?µ:¶Å›Ä¿ SJJŠNž<)I:zô¨Ö¯_oq*ìgÐ@@ܸqCÙÙÙ5j”\.—V¯^mu$lgÐû)cŒrss×n¬®®NO=õ”Ö¬YÓîÙŒ®Æе%K–ÈétÊétêG?ú‘w?5õ—Ë¥ÌÌL=ÿüóÊÊÊ’Ûí¶:RŸ÷ /x¯ÙøøxIÔ¹7þüç?kúô骬¬ôî몞¡Pë¿þõ¯zÿý÷µ`Á‚›s·Û­_ýêWZ¾|¹Ö­[×îÝ¡÷îÝ«-[¶èÅ_ÔöíÛšµ°°Pßýîw5iÒ$½ôÒKêìù‹ÂÂB=óÌ3zúé§õ·¿ý­ÛcvÌÜÐР¬¬,ÅÄÄ(--M%%%¶ÍÚÚ¡C‡”œœ°¬-9ü‘×ãñhÇŽ:tèÞ{ï=Ûg¶ãœ«©©ÑòåËõ‹_ü¢ÃskÎõ6o0ç›?ò¶Œ9‡V ú¥ÊÊJ³k×.ãp8Ú¥¥¥™·ß~ÛcLff¦Ù³gO·Æй«W¯šôôtSXXh ÍG}䣦þ±råJ³sçNcŒ1/¿ü²Y½zµÅ‰ú¶ÆÆF³téRï5[VVfŒ¡Î=uíÚ5“——g‡©¬¬ôî爵ý½ÖÕ¤µÍ›7›ŒŒ cŒ1µµµf„ ÆívcŒ)--5=ö˜÷ØÔÔTóü# YÍ¢E‹ÌñãÇÍoû[3xð`“ŸŸßî¸+W®˜o~ó›Æãñ˜ÆÆF3eÊóÉ'ŸÜvÌ®™wîÜi~ÿûß›‚‚3sæL“˜˜hÛ¬->þøc3eÊ“œœ¬þÌû駟šÔÔTS^^°¬þÌlÇ9gŒ1gÏž5óæÍ3‹/îöÿÅŽyƒ5ßü•·E0æ|Ñ ÷c/^l× ðÁæ‹_ü¢ñx<ÆcŽ?nbcco;†®egg›‚‚‚vû©©ÔÔÔ˜Áƒ›ÚÚZcŒ1æî»ï67nÜ°8Yßµ{÷n³{÷nŸ}Ô¹wššš|šÑ®ê*µn[“¶f̘a>ìÝþéOêý…æ¶mÛ̼yó¼cO>ù¤ÉÍÍ HÎÓ§O›úúzïöc=f}ôÑvÇ=ýôÓ&33Ó»½páB³nݺێÙ5saa¡wËÏ ÿþ÷¿m™µEzzºÙ¾}»™5k–ßsú;ïâŋͻﾰœ­ù#³ç\‹ììl³hÑ"Ÿ}ÁœsþȬùfŒò¶Æœƒ/^âÞ…‡‡·ÛWTT¤ØØX9IÒ˜1cTVV¦k×®u:V]]ÔÜ}ËåÒ¾}û4{ölÍŸ?_ÅÅÅÞ1jê%%%ŠŠŠÒˆ#$IÑÑѪ««Ó¹sç,NÖwíÚµKK–,Ñ< 'NH¢Î½ÕvÍíªž¡R뎇Z»råŠNŸ>íÝþÒ—¾¤Ã‡K’¾ÿýïëøñãÚ´i“JKKÕØب‡~8 9§NªaÆy·'Nœ¨qãƵ;®¸¸ØçOÇbbb¼/«-**êtÌ®™¿óïø숈Ptt´-³JÒþ𥥥ü£øü‘÷wÞÑ{ｧêêj¥¤¤(77·Ó—Û%³ç\‹´Ûw»ëÅŸü‘7XóMòO^)xs¾hÐCLii©"##½Û- Ceee—cèÜ€ôŸÿüG.\Ð矮iÓ¦éúõë’º®7º¯mÃÃÃ5jÔ(êØ §NÒåË—5~üxÍœ9S%%%ÔÙÏ:«çåË—©õ-?ùÉO”——§êêj555éܹsúüóÏ%IãÇW^^žV­Z¥E‹iß¾}8p`Pråççë‰'žh·¿í÷môèÑÞïÙùóç; †ždn{ûÔÔT :4 9[îëN³–••©¾¾^S§N x¾¶z’÷­·ÞR||¼æÍ›§gŸ}VÚºu«­3Ûqεhy¢£µî^ÛГ¼mo¬ùÖrwš×Ê9ê:þu loÕªU¾¹Äܹsõøãw8æt:æÝ®««“$ 6¬Ó±`-vÖzõ«_U^^žæÏŸ¯ýû÷+==šú‰Ãáð©£$Õ××SÇ^Š‰‰Ñ®]»4dȽòÊ+Š‹‹£Î~ÔÙu;lØ0}òÉ'ÔZÒsÏ=§ÈÈH-]ºT<òˆN:¥ 6xǯ_¿®gžyF›7oÖâÅ‹µgÏžvuó·3gÎ(55U±±±íÆÚ~Oëêê¼ÏRu4¬ïçfî(×;ï¼£_|1 9¥žÕ×ãñhûöíÚ¸qcÀóµÕÓëáôéÓzä‘G®ûï¿_©©©zã7ôÔSOÙ.sëëÁns®+V͹žæm-XóMêY^·ÛmÙœ zŸõë_ÿºG·KJJÒ±cǼÛUUU0`€&L˜ éÓ§w:êº[o‡Ã¡Ç\{÷îUzzz—õF÷%%%iÅŠÞ톆ݸqC“&M²0Uÿñ³ŸýL .ÔâÅ‹µ|ùrï~êÜ;]]·_þò—©µ¤ˆˆeffJ’ T]]í}§à÷ß_o¾ù¦öï߯x@=ô’““µhÑ¢€åijjÒ:ý¡4))IeeeÞíªª*ï÷¬£±É“',k‹ždn›+??_?üá½ra§¬“&MÒßÿþwíܹS{÷îõž§©©I‘‘‘ÞW¬Ù)oËm[KLLTnnnÀ²¶èÍõ`Ç9×+æ\oò¶Ö|“zž×ª9‡f¼Ä=ÄLŸ>½Ýb6qâD………u¸ÐµŒáÎ$$$HêºÞè¾)S¦Èãñxÿv¿ªªJ;v¬ÅÉú„„êìg]Õ“Z··uëVeeeykŸŸ¯ÄÄDIÒ¬Y³´~ýz4ÃÆõÜsÏÉéìøÇ£¤¤$]¼xÑ»]UU¥¯ýë· ¤Þd–šß+¡ªªJ?øÁl›õÞ{ïÕùóçUTT¤¢¢"­]»VS§NÕ|`˼’ôo|ÃglРAzðÁšWê]f;ι®X1çz“W î|“zžwÚ´i–Ì94£AïÇZþŽÏãñx÷ÝsÏ=JLLô¾\û­·ÞÒ /¼pÛ1tîÒ¥K:|ø°Œ1ª®®ÖŸþô'-[¶L5õ—°°0¥¤¤èäÉ“’¤£Gò™ò½PSS£W_}U.—KõõõÚºu«V®\)§ÓI{¡eÍu¹\’º¾nCåšn[©ùÆÖ?TKRNNŽî¹çŸ|ûÛßöyç*%%%`YW­Z¥iÓ¦©¡¡Aeeezå•WT^^î“wîܹúðÃ%I7oÞTEE…~þóŸßvÌ®™/\¸ Ý»wë{ßûž._¾¬ýë_Ú±c‡í²4H111Þ¯/|á ºë®»¬½Í+IzóÍ7U__/IúË_þ¢ùóç,¯?2Ûqεhjj’ÛíöÙì9×Û¼Áœo½ÍkÅœC+V¿<ãÚµk&;;Û8Nó»ßýÎç£{®\¹b–-[fÖ­[g6oÞìs»®Æбââb3qâD“`žxâ ŸÏ@7†šúKCCƒIOO7ëׯ7999VÇéÓ®^½jâããÍ×¾ö5³páBSZZê£Î=ÓÐÐ`^zé%ãt:ÍÊ•+Muuµwgõìïµî¬&‰‰‰æ7Þ0ÆsìØ1óÇ?þÑìÙ³Çûq”-<yíµ×Ì–-[ÌþýûÍž={ŒËå HÖ5kÖ‡Ãáó5yòävy1æðáÃ&;;Ûdee™“'Oúœ§«1»eþøãÍèÑ£}nït:Í믿n»¬míÛ·/ ŸÉ쯼yyyfùòåAyü÷Gf»Î¹S§N™o}ë[&..Î?~Üç<Ášs½ÍÌùæ¼mzÎÁ—Ø~æè^†Р`4èØ :6@ƒ€ Р`4èØ :6@ƒ€ Р`4èØ :6@ƒ€ Р`4èØ :6@ƒ€ Р`4èØ :6@ƒ€ Р`4èØ :6@ƒ€ Р`4èØ :6@ƒ€ Р`4èØ :6@ƒ€ Р`4èØ :6@ƒ€ Р`4èØ :6@ƒ€ Р`4èØ :6ðÿ#£½¹ÖÂRƒIEND®B`‚HDDM-0.2/hddm/examples/wfpt('hard',).png0000664000102500010250000004720011726446357020243 0ustar wieckiwiecki00000000000000‰PNG  IHDR Xšv‚psBIT|dˆ pHYsaa¨?§i IDATxœìÝ{TÜÕ½ÿÿ× ·H!–'Á)‰Ö ´5MÒ¤/G-I´ ’‚Úu´”Z’C¬ß€–ãI5¥µžeŠÍEIÒ6M¨·Zh­'½Á)")^‚Q¡ˆBœ>¿?òcê„@›!y>ÖÊZ²÷goÞÆ(¯ÙŸ=ÛfY–%0À>Ù8}@C` €1Æ@C` €1Æ@C` N!Û¶mÓ† ”””¤ï}ï{'5×ßÿþwmݺU---:çœsT[[«7ÞxCßþö·5oÞ¼qªØ¿êêj}õ«_Õí·ß.IúÕ¯~¥„„õ÷÷¹öùçŸ×_þò— ­0~ pŠøÅ/~¡—^zIwß}·~úÓŸêg?û™š››;®¾¾~H[uuµÊËËuóÍ7«££C---:pà€¦M›¦––¹Ý¯˜˜ýíoÓ‘#G$Iû÷ï×[o½å7€\}õÕú¿ÿû?ýüç?ŸÐšãƒ§ˆ§žzJ.—K’”’’¢þóŸŠqÌ¡C‡T\\ìÓvàÀ­^½ÚÛ~ÁH’âããår¹tÑEɲ¬ ¸ƒ™3gŽÎ=÷\ï×\pbbbî÷úo~ó›*++Óûï¿?¡uNN¯½öš‚‚‚F}}oo¯V®\©ŽŽŸö{ï½WW]u•ìö£ÿ‹p8ºð /Icú'cðûKÒE]¤‹.ºhÄëÓÓÓu×]wMtY€“D€TXX(»Ý®E‹éàÁƒúõ¯­³Î:K‡CO?ý´$éÕW_U||¼®½öZÝ~ûíjmmÕ¯ýkÝ~ûízùå—õÄOèÒK/Uuuµòóó5{öl}ûÛßVKK‹¤£ûEöíÛ§7ÞxCwÜq‡*++åñxTVV¦ôôtŸznºé&Ÿ@ ])¹òÊ+uöÙgë±Çó¶9rD÷Þ{¯6nܨåË—ë†nPWW—¤£veddèßø†žxâ sÎ9úÕ¯~%IzóÍ7uà 7hÍš5ºûî»uàÀïœçŸ¾-Z4âÏìŠ+®Ð–-[´oß¾ü©Œ°éòË/·®¿þzï×?úѬ¨¨(«¯¯ÏÛ–‘‘a9rIJ,ËŠ‹‹³î»ï>˲,«³³ÓÚ¶m›e³Ù¬[o½Õz饗¬ÿøÇÖ§>õ)ëÆoôŽ¿õÖ[­ÔÔTï×ååå–Íf³ÚÚÚ†­«¨¨È:묳¬‡zÈ:pà€µ~ýzëŒ3ΰ:dY–e=ñÄÖç>÷9˲,«¿¿ß · -˲¬¿ÿýïVll¬u饗ZeeeÖš5k¬?üáÖ?ÿùO+::Úzíµ×,˲¬wÞyÇ µ¾þõ¯égö©O}ÊÚ±cǘÆÌbÔm·Ý¦çž{N~ø¡$)))I‡ÒÎ;%IúÓŸ´téÒ!+’¥ H’233uùå—ëÎ;ïÔµ×^«gŸ}V}}}’$˲|ösìÛ·OAAAš9s戵EDDhÍš5Š‰‰ÑòåËÕÛÛëÝð~öÙgëŽ;î$Ùl6]xá…ÞÕŒyóæé¼óΓËåÒM7ݤ‡zH_úÒ—ôýï_gu–÷1+—Ë¥äää1ÿÌf̘¡ººº1˜C€µbÅ …„„è™gž‘$UTTèÚk¯Õ“O>)Izæ™g”••5ìøÁ`òÉÛW^y¥<ªªªüŽy÷ÝwuÖYg·¶O†žéÓ§K’:;;%I_ùÊW”‘‘¡M›6©¨¨Hýýý>Ÿ^4ä{<ÿüóZºt©OÛ´iÓŽ[DZ¢££õÚk¯yÀ¨3Î8CéééÚ¾}»š››­U«Vé…^ÐþýûÕÛÛ«3Ïn%Iþ󟦋/¾XÒÑG¤>ùÖ%—\¢žžõööžpÍ«V­Ò¥—^ªK.¹DÒÐǼü¹þúëõÊ+¯è£>ò¶Y–å 5£ÕÚÚª¸¸¸1× 0‡lÉ’%r¹\PLLŒ‚‚‚ôµ¯}Mï¼óŽ®¼òJŸk»ºº|>9jÐ_ÿúWï??ûì³ÊÏÏ÷>5cÆ ½þúëêèèО={¼'œûѼŸÔ××çóHÕ`¸ -!!!zá…TSS£7jÿþý:pà€~ûÛßJ’úûû½ûZeffª¿¿_7ß|³šššô›ßüFo½õ–Þxã!Ÿjõá‡êsŸûœþë¿þ˧ݲ,µ··Oø)퀓C€—““£üü|ï×_ÿú×uóÍ7Ëf³I:î»ï>:tH¿ùÍo´uëVŸ_ð_}õUåçç+==]óæÍÓý÷ßïí[¹r¥‚ƒƒ•­‹/¾X矾TSSã·–?þñª¬¬T{{»yä½÷Þ{z衇d³Ùôä“OêÝwßÕÝwß­eee鳟ý¬¾ùÍojß¾}:óÌ3UZZªW_}U¿ûÝï´uëVï¼_þò—õð붶VK–,Q]].ºè"}æ3Ÿñ~„ï ?üPo¿ý¶î¿ÿ~Ÿ•š×_]áááC>BXlÖñÖÅSÒþýûuþù竺ºZ‹/õ¸—^zIÏ<óŒ~ò“ŸL`u'ïûßÿ¾ ½_ß}÷Ýêïï×øÃI¬ p<¬€|,]ºT‡Ö{ï½7Ù¥ «¾¾^‰‰‰Þ¯ûûûUWW糺L8E >žÔÝÝ=æ±<ð€6oÞ¬C‡wY'­½½]v»]7Þx£$éã?VAA~øa………Mru€ãá,8555é?ø~úÓŸjùòåºÿþû5gÎœ1ÍaY–þð‡?hÉ’%Tåøøãÿ¨”””a?ZX Œá,Æ@C` €1Æ@C` €1Æ@C` €1Æ@C` €1Æ@C` €1Æ@C` €1TUU)33S .ÔÒ¥KµsçÎQÝ¿¿V®\©E‹é3ŸùŒ®ºê*ýéOšÀj€ÀGFMMÒÒÒ´jÕ*½òÊ+zôÑG•••¥Gyä¸c_~ùe]zé¥úò—¿¬êêj544èÊ+¯ÔâÅ‹U^^n z 0Ù,˲&»ˆ@ÓÖÖ¦Ï}îsš3gŽ^xáoû=÷Ü£’’½ñÆš={¶ß±ü±æΫ   ½ñÆ>}×\sþö·¿©±±QŸþô§'ô€@Ä ˆ6lPKK‹rss}Ú/^¬žž ;¶¦¦Fo¿ý¶®ºêª!}+W®Ôû￯ÿþïÿ÷š€©€âGmm­l6›æÏŸïÓ/IÚ½{÷°cÛÚÚ$IÕÕÕCú.¹äIR]]Ý8U L-cX–å ÑÑÑ>}.—Kv»]r»Ý~Ç'%%I’öîÝ«úúzŸ¾C‡I’Þzë­ñ.˜‚'»€@ÓÔÔ$·Û­¨¨(ûþxBBB£¨©©IóæÍ2þ¼óÎÓš5k´iÓ&Ýu×]zà4oÞ<½ûî»zâ‰'$IgŸ}¶ßï}ðàA=ÿüóŠ‹‹Ógœ1þ7€“ÒÓÓ£ýû÷ëꫯòf5F‡rŒ®®.IRhh¨ßþÞÞ^IRddä°süð‡?TNNŽ¶mÛ¦Ûo¿]×^{­>ÿùÏëÍ7ß”$%$$ø÷üóÏ+;;ûdÊ€Û·o××¾öµÉ.cJ"€ÃårI’:::ÔÙÙ©éÓ§ûô{<9ÅÅÅ8ϼyó|VHjjjô—¿üE6›M7ß|³ß1çwž¤£ÿBÏ™3ç$î!??_%%%“]†Áë¸xm¯M`ãõ LÿøÇ?”íý½ cG9ÆŒ3äp8äñxÔÐРE‹yû:;;Õß߯Ë.»lÌó>ðÀ’Ž~’Öpã§M›&Iš3gŽ’““O zL¤¨¨(^—Æë¸xm¯M`ãõ lƒ¿·aìØ„îGNNŽ,ËÒÞ½{}Ú7§n4­?þñª¨¨PJJŠ~ñ‹_Œ[ÀTCñ#//OÁÁÁC>.·¢¢B111Z·n·­»»[}}}ÃÎõÜsÏéꫯÖ¿øEýîw¿òHp:!€ø‘˜˜¨ÒÒRmÙ²E ’¤={öhóæÍÚ±c‡fÍš%Ijoo—ÓéTBB‚>ì3ÇþýûõðÃkùòåš?¾^xá9ã÷ö€ #;;[aaaZ»v­<\.—*++µdÉï5ááár¹\š9s¦‚‚‚$I|ðRRR$IŸýìgU^^®eË–MÊ=`|eeeMv ¯Oàâµ \¼6ק*›eYÖd£jkk•’’¢šš6 ~_;y<‚Àc Œ!€0†Àc Œ!€0†Àc Œ!€0&x² àd•7”«°ºPîÝ>íŽ0‡ŠS‹•>7}’*‹˜ò « Õx°qh‡[Z_µž„˜òW>ì6»œNIR«§UÖÀUÀäbà”áŒpªyM³š×4{ƒ °@C` €1Æ@C` €1Æ@C` €1Æ@C` €1Æ@C` €1Æ@C`LðdȪªªôø㫹¹YÓ¦MÓwÞ©åË—jìô‹_üB•••êííU\\œ–-[¦k¯½V!!!\9˜ è©©QZZšvíÚ¥ÔÔT566*99Y7nT^^Þˆcßzë-}þóŸ×¿ÿû¿ëç?ÿ¹"""ôÎ;ïèºë®ÓöíÛU^^nè.€ÀÂ#X~´µµiÅŠZ°`RSS%I‰‰‰Z½zµ ÔÒÒ2âø‡zH‡Vqq±"""$Içž{®rssUQQ¡¦¦¦ ¿ @üØ°aƒZZZ”››ëÓ¾xñbõôô¨¨¨hÄñ~ø¡>úè£!A%88Xgœq†f̘1î5SÄÚÚZÙl6ÍŸ?ߧ=>>^’´{÷îÇî¹þúëÕÖÖ&IòxݧßãñÈáp(..nØ9º»»•™™©E‹©  @ßýîwuã7ªººZiiiúýï?b %%%JNN>é{À‰ó÷pmm­RRR&©¢S`cÆŒÞMâ >}êïï×%—\2âÿó?ÿ£ßþö·Z½zµ$ÉétêÅ_Ôå—_®—_~Y555S<à ~äääȲ,íݻק}pszRRÒˆã+**”àó×´iÓôÓŸþT’ôÌ3ÏŒsÅÀÔ@ñ#//OÁÁÁÞÀ1¨¢¢B111Z·n·­»»[}}}>×-]ºT¯½öšöïßïÓ~æ™gêŒ3Îññ-àTFñ#11Q¥¥¥Ú²e‹÷1¬={öhóæÍÚ±c‡fÍš%Ijoo—ÓéTBB‚>쟛›«O}êS*((ð†ÎÎNÝsÏ=š;w®n½õVÓ·6¡#;;[aaaZ»v­<\.—*++µdÉï5ááár¹\š9s¦‚‚‚¼í111ª¯¯×Ýwß­+®¸BçœsŽ.¼ðB-[¶L%%%Ã~Âpª#€Œ ##CÃöGDD¨¾¾Þo߹瞫§Ÿ~z¢J¦$Á` €1Æ@C` €1Æ@C` €1Æ@C` €1Æ@C` €1Æ@C` €1Æ@C` €1Æ@C` €1Æ@C` €1Æ@C` €1TUU)33S .ÔÒ¥KµsçÎQÛ³gìvû°’’’&¸r 0Ovª¦¦FiiiÚµk—RSSÕØبäädmܸQyyy#ŽýÉO~"›Í¦¨¨(ÅÄÄøôíÛ·OY:° ~´µµiÅŠZ°`RSS%I‰‰‰Z½zµ týõ×köìÙ~Ǿÿþûª¯¯×›o¾©óÏ?ȼçœsŽ²³³'ü€@Ä#X~lØ°A---ÊÍÍõi_¼x±zzzTTT4ìØööv½ð C‡$=õÔSš?¾Î=÷Üq¯˜ Xñ£¶¶V6›MóçÏ÷i—$íÞ½{رsæ̶oÛ¶mCB p:aä–e©®®N’íÓçr¹d·ÛÕÑÑ!·Û=¦yÔØبo¼qÜj¦È1šššäv»©à`ߢï¦ò¦¦¦1Í»mÛ6]wÝuŠŒŒ·Z€©†rŒ®®.IRhh¨ßþÞÞ^ISÐÓO?­œœœ“/˜ÂØr —Ë%IêèèPgg§¦OŸîÓïñxäp87ê9«««õÑGéºë®ÕõùùùŠŠŠòiËÊÊRVVÖ¨¿'NNYY™ÊÊÊ|Ú߬Ɖ#€cÆŒr8òx¸9},'™÷ôôè—¿ü%_"€ø•——§àà`oàTQQ¡˜˜­[·ÎÛÖÝÝ­¾¾¾açÚ¹s§¢££õÅ/~qÂê¦ ˆ‰‰‰*--Õ–-[ÔÐÐ IÚ³g6oÞ¬;vhÖ¬Y’Ž:èt:• Ã‡ûkÛ¶mœ|üÿØ2Œììl………iíÚµòxžœnx €1Æð~”7”«°ºPîÝ>íŽ0‡ŠS‹•>7}’*€©€…Õ…j<Ø8´Ã-­¯ZO€DÀÁ•»Í.g„S’ÔêiÕ€50dU0zì`Χš×4«yM³7ˆN€1Æ@C` €1Æ@C` €1Æ@C` €1Æ@C` €1ÆOvL¤VO«b7Åz¿v„9TœZ¬ô¹é’¤ò†rVÊý±{È8Àø#€NIŽ0‡ä–¬µ¸[þÕá–ÖW­÷ÂêB5lyÀ¸!€NIÅ©ÅZ_µÞge£ÕÓªkÀ§mðŸí6»œNŸ9WKã‡8%¥ÏM÷®r ŠÝë»ò Χš×4›( NklB` €1Æ@C` €1| /à´óÉÓÑ9ñÌ"€NÞŽ.N<S €Ó†¿ÓÑ%N<“ €Ó†¿ÓÑf@FPUU¥Ç\ÍÍÍš6mšî¼óN-_¾|Ìóôööª¤¤Dï½÷ž¢¢¢ô…/|á„æ¦:È0jjj”––¦]»v)55UJNNÖÆ•——7êy*++uÏ=÷è¶ÛnÓ£>:áõ£­­M+V¬Ð‚ ”šš*IJLLÔêÕ«UPP –––ãÌpTqq±n¼ñF•––Ž)´§*ˆ6lPKK‹rss}Ú/^¬žžwŽçŸ^÷ÝwŸ¶lÙ¢äää‰*˜R ~ÔÖÖÊf³iþüù>íñññ’¤Ý»w8þwÞÑ×¾ö5¥¤¤(=ÍŽÀ È1,ËR]]$)::Ú§ÏårÉn·«££Cn·ÛßpIÒC=¤>ø@ééézꩧtÇw(99Yßüæ7ÕÖÖ6¡õŒrŒ¦¦&¹ÝnEFF*8Øw~HHˆbbb¼× ç•W^QPP´|ùr=öØc*..Ö¶mÛ4oÞ<µ¶rê.NOctuuI’BCCýö÷ööJ’"##ýöôÑGª««ÓÙgŸ­‚‚EDDÈf³éºë®ÓwÞ©öövLLñ@€ãcxár¹$IêììÔôéÓ}ú=‡âââüŽýõ×500 ³Ï>{Hßå—_®7êøÈ5äçç+**ʧ-++KYYYc¸œŒ²²2•••ù´ ¾YG9ÆŒ3äp8äñxÔÐРE‹yû:;;Õß߯Ë.»lØñ ’¤÷ßHߢE‹4âþI*))ᓳ&™¿7€kkk•’’2IxËœœY–¥½{÷ú´nNOJJvlDD„>ó™Ïhÿþý:tèOŸeY ÖÅ_<þESļ¼<{Ç ŠŠ ÅÄÄhݺuÞ¶îînõõõù\÷Ýï~W’†œ|þì³Ïêã?ÖwÞ9A•âGbb¢JKKµeË544H’öì٣͛7kÇŽš5k–$©½½]N§S :|ø°wüm·Ý¦•+Wª¸¸Xýë_%ÝRRR¢Ûn»M_ýêWÍßØ2Œììl………iíÚµòx}.—Kv»]r»ÝÃÎñꫯ*55UkËÁD IDAT.—Kï¿ÿ¾¶nݪ‹/¾X¥¥¥Z;è Çhjj’ÛíVdd¤‚ƒ}·È„„„(&&Æ{Ýp²²²ôÛßþVû÷ïWss³®ºê*y<Ý~ûízñÅ'´~ ± ý]]]’¤ÐÐP¿ý½½½’¤ÈÈÈQÍwöÙgë¹çžÓƵnÝ:}÷»ßÕ«¯¾:â˜üü|EEEù´eee)++kTß'¯¬¬Leee>mƒ¿+âÄ@Žár¹$IêììÔôéÓ}ú=‡âââÆ4ïÚµkµuëVýãÿPww·ÂÃǽ¶¤¤DÉÉÉc®ãÇßÀµµµJII™¤ŠN <‚uŒ3fÈápH:º™ü“:;;Õß߯K.¹ä„æÎÎÎVXX˜ÂÂÂNºN`*"€ø‘““#˲´wï^ŸöÁÍéIII'4oDD„¾øÅ/úLœN ~äåå)88Ø8UTT(&&FëÖ­ó¶uww«¯¯ï¸sZ–¥'žxB<ðÀ¸× L?UZZª-[¶xÃÚ³g6oÞ¬;vhÖ¬Y’¤ööv9N%$$èðáÃ’ŽîY¶l™î½÷^8p@’´oß>ÝsÏ=úÞ÷¾ÇÞœÖØ„>ŒÁýk×®•Çã‘ËåRee¥–,Yâ½&<<\.—K3gÎô>V¡ .¸@;vìÐÃ?¬›nºIQQQúÖ·¾¥Ù³gOÖí2‚ŒŒ edd Û¡úúú!í?úÑô£ýh"K¦$Á` €1Æ@C` €1Æ@C` €1Æ@C`LðdÀX”7”«°ºPîÝÞ¶VOë$V `J)¬.TãÁF¿}Ž0‡ájcEL)ƒ+v›]Χ·ÝæPqjñd•%`JrF8Õ¼¦y²ËŒ›ÐC` €1Æ@C` à ÀµzZ»)Öûõà$ésÓ'±*˜X`”OZ°Ôânñþi<بõUë'¹:˜ ŒRqj±£5Û1ÛûÇn;ú¿ÒÁÚ#ã,F)}núǬb7ŪÅÝ2IÀÔà c Œ!€0†Àc Œ!€0†Àc Œ!€0†2‚ªª*effjáÂ…Zºt©vîÜyÂs=øàƒ²Ûíz÷ÝwDZB`j žìUMMÒÒÒ´k×.¥¦¦ª±±QÉÉÉÚ¸q£òòòÆ4׳Ï>«»ï¾[6›m‚ª¦V@ühkkÓŠ+´`Á¥¦¦J’µzõj¨¥¥eÔs½þúëºï¾û”––&˲&ªd`J €ø±aõ´´(77ק}ñâÅêééQQQѨæéêêÒ׿þu=õÔSŠŒŒœˆR€)…âGmm­l6›æÏŸïÓ/IÚ½{÷qç8räˆn¹åÝÿýºà‚ &¤N`ª!€ò,ÕÕÕI’¢££}ú\.—ìv»:::äv»Gœ§  @×]w÷.!šššäv»©à`ß=ú!!!Š‰‰ñ^7œ­[·êÈ‘#úÆ7¾1¡µS Ÿ‚uŒ®®.IRhh¨ßþÞÞ^IvOÇŸÿügýò—¿TEE…ß~6¢àtF9†Ëå’$utt¨³³SÓ§O÷é÷xmƒoVãÄ@Ž1cÆ 9y<544hÑ¢EÞ¾ÎÎNõ÷÷ë²Ë.v¼ÇãQssóöÁ•·ß~[6›M‡vŽ’’%''ŸÄ]àdù{¸¶¶V)))“TÑ©= ~äääȲ,íݻק}pszRRÒ°cß~ûm9rdÈŸ›o¾Y’´ÿ~9rĻҜN ~äåå)88Ø8UTT(&&FëÖ­ó¶uww«¯¯Ït‰À”Dñ#11Q¥¥¥Ú²e‹$I{öìÑæÍ›µcÇïÞööv9N%$$ŒøH€£Ø2Œììl………iíÚµòx›Í&›Í6Ñe2‚ŒŒ edd Û¡úúúQÍõä“OêÉ'Ÿ¯Ò€)‰G°C` €1Æ@C` €1Æ@C` €1Æ@<Ùp:(o(Wau¡Ü»}Úa§+}nú$Uf@0 °ºP‡v¸¥õUë N \ù°ÛìrF8%I­žV XCVEàTÆ rF8Õ¼¦YÍkš½AN'Æ@C` €1Æ@C` À)¤¼¡\…Õ…C¶s„9TœZÌiÛ¨ÕÓªØM±Þ¯ù™€8…Vªñ`ãз´¾j=¿ OG˜CrKÖ€ZÜ-ÿêàg~ñœBW>ì6»f;fk¶c¶ì6»OÆWqj±£½?o~æ02V@àäŒpªyM³$)vS¬ï;óWésÓ‡¬rð3€á±Àc Œ!€0†Àcø^8 œ<Ž‘|òtôVOë$W'“ÇáÏ°§£öÀiŒ'á“';#œ’Ž¾Ó=` p öi¬8µXë«Ö»2§3ŒNÇ'ù;p›ÐC9Žªª*effjáÂ…Zºt©vîÜ9ê±ÿû¿ÿ«U«VÉårÉåréÖ[oÕûï¿?Õ2‚šš¥¥¥iÕªUzå•Wôè£*++K<òÈqǾúê«úÉO~¢k®¹F?þñ5wî\mݺU7Ýt“Ê€ÀÄa´µµiÅŠZ°`RSS%I‰‰‰Z½zµ týõ×köìÙÃŽùå—µmÛ6ï×Ë–-“ËåÒK/½¤wÞyGçž{î„ßhXƆ ÔÒÒ¢ÜÜ\ŸöÅ‹«§§GEEE#ŽÿÖ·¾åóuPP–/_®3Ï—Ë%»Ý®ŽŽ¹Ý£ÿŒÿ|PIIICVF€Ó {@ühjj’ÛíVTT”‚ƒ}D!!!Š‰‰ÑÔÔÔ¤yóæ8׶mÛ´yóf½üòË Ull¬îºë.Ùl¶‰¼ @üèêê’¤a•êíí•$EFFw®ôôt]pÁ*++Óã?®uëÖiÚ´iºóÎ;‡“ŸŸ¯¨¨(Ÿ¶¬¬,eeeöp’ÊÊÊTVVæÓ6ø{"NÄ—Ë%IêèèPgg§¦OŸîÓïñxäp8wܹÎ8ã -X°ÀûiZéé骨¨1€”””(99ù¤î'ÇßÀµµµJII™¤ŠN ìñcÆŒr8’¤††Ÿ¾ÎÎNõ÷÷ë’K.ó¼7ÜpƒâââôÚk¯KÀTCFNNŽ,ËÒÞ½{}Ú7§'%%мsçÎÕE]tÒõSdyyy öŽAŠ‰‰Ñºuë¼mÝÝÝêëë;îœ|𪪪ôÿñã^/0@†‘˜˜¨ÒÒRmÙ²ÅûÖž={´yófíرC³fÍ’$µ··Ëét*!!A‡ö¶ÝrË-zôÑGõÑGI’Þ{ï=­]»VßûÞ÷´lٲɹ)`’± }ÙÙÙ ÓÚµkåñxär¹TYY©%K–x¯ —ËåÒÌ™3$I Ó| x@6lÐW¾òõ÷÷+??_Ÿýìg'ëv€IG9ŽŒŒ edd Û¡úúzŸ¶ÈÈHíÚµk¢K¦Á` €1Æ@C` €1Æ@C` €1Æ@<Ù€SÞP®ÂêB¹?vû´;Â*N-VúÜôIª †G`Š*¬.TãÁÆ¡ni}Õz€€D`Š\ù°ÛìrF8%I­žV XCVE P°€)ÎáTóšf5¯iöTÆ@C` €1Æ@C` `ÊÊUX]8ä0AG˜CÅ©Åœp`J#€` « Õx°qh‡[Z_µž`J#€`W>ì6»÷dóVO«¬!«"0Õ°€åŒpªyM³š×4{ƒLuÆ@C` €1Æ@C` €1Æ@C` dUUUÊÌÌÔÂ… µtéRíܹsÔc_zé%-^¼XQQQ:çœs”™™©7ß|s«d555JKKÓªU«ôÊ+¯èÑGUVV–yä‘ãŽýË_þ¢+¯¼RúÓŸd³ÙÔÒÒ¢òòr%%%éïÿ»ê€ÀDñ£­­M+V¬Ð‚ ”šš*IJLLÔêÕ«UPP –––Ç—””èÉ'ŸTGG‡ÚÚÚ´}ûvwÞyêééÑw¾ó·$ˆ6lPKK‹rss}Ú/^¬žž ;ö7ÞЗ¾ô%ÝrË-ŠŒŒThh¨V®\©{î¹G’T__?¡µŒâGmm­l6›æÏŸïÓ/IÚ½{÷°cãããuÇw i_¹r¥ÂÃÃu饗Žo±ÀB9†eYª««“$EGGûô¹\.ÙívuttÈívûo·ûÿ‘öõõ©··W+V¬ß‚€)$x² 4MMMr»ÝŠŠŠRp°ï'$$D111:pà€ššš4oÞ¼QÏûâ‹/êŠ+®Ð·¿ýíñ.Àiª¼¡\…Õ…rü¯7DZ=­“XNT«§U±›b}¾€Sä]]]’¤ÐÐP¿ý½½½’¤ÈÈÈQÏÙÛÛ«G}TÏ<ǫ́®ÏÏÏWTT”O[VV–²²²Fý=œú « Õx°ÑoŸ#Ìa¸œG˜CrKÖ€ZÜC?à„ט\eee*++óiü]'Žr —Ë%IêèèPgg§¦OŸîÓïñxäp87ê9‹ŠŠôƒü@guÖ¨®/))Qrrò¨çpz\ù°ÛìrF8½íŽ0‡ŠS‹'«,ŒAqj±ÖW­÷YÅÄëL>o×ÖÖ*%%e’*:5@Ž1cÆ 9y<544hÑ¢EÞ¾ÎÎNõ÷÷ë²Ë.õ|%%%Z´h‘¾ð…/LD¹ g„SÍkš'» œ€ô¹éJŸ›>Ùe€QlB÷#''G–eiïÞ½>탛ӓ’’F5ÏÏ~ö3%&&*--mÜk¦"ˆyyy öŽAŠ‰‰Ñºuë¼mÝÝÝêëë2ǦM›är¹ôoÿöo>ío¿ý¶îºë®‰)p<‚åGbb¢JKKµjÕ*åææꢋ.Òž={´yóf=÷Üsš5k–$©½½]^x¡¦OŸ®·ÞzKÁÁÁ:|ø°V­Zå +ƒäñxÔÖÖ¦Ç|²n ˜Tadgg+,,Lk×®•Çã‘ËåRee¥–,Yâ½&<<\.—K3gÎô~dï­·Þª§Ÿ~Z6›M~øáyCBB”‘‘aì>€@BAFFƈa!""Bõõõ>mÛ·o×öíÛ'º4`Jbc Œá,¥ò†rVú×êi×ù¤@Çù€SF©°ºPýö9Âã7Ÿ[Z_µž8%@`”W*ì6»œNoûàŠÅxÌ×êiÕ€50dU€SÆÈáTóšæ ™/vS¬ZÜ-ã67†MèŒ!€0†Àc Œ!€0†À΀ÓD«§U±›b½_ È‰ëSË'_ÇVO먮“x½œâaÉ- X¾‡º¥õUëù…tŠöuì;Þu¼Þ`À)®8µX‰Ñ‰ší˜íýc·ýÏ¿ûc÷$W‡Ñò÷:ÎvÌVbt¢ŠS‹G¼Ž×@ aNqésÓ‡¼ë»)vÈ»èlþ^ÇÑ^Çë °Àc Œ!€0†Àcø^˜ ãyuyC¹ « }Îqéìɨ€Ñ €À8›ˆ“¨ « Õx°qøï50<‚ãl"N¢g·ÙG<{2k`4X€q6‘'Q;#œj^Ó|ÒópZ6`²°Àc Œ!€0†Àc Œá@­žVÅnŠõùtå å*¬.r˜¥#Ì¡âÔâ!gÏ8=@ €8Â’[°ü èsLBUÀèVªñ`ãз´¾j=€$”âÔb­¯Z?ädé_ï"jðß[»Í.g„SÒÑÕ»kÀï¿ÓNO ésÓy—Sž3©æ5Í’¤ØM±~Wóœ¾Ø„> UUUÊÌÌÔÂ… µtéRíܹsLã>¬§Ÿ~Z—^z©~ÿûßOP•@àcä8jjj”––¦]»v)55UJNNÖÆ•——wÜñ¯¼òŠÊÊÊô³ŸýLÝÝݲÙlª+ #hkkÓŠ+´`Á¥¦¦J’µzõj¨¥åøKÊ .Ô#<¢œœœ‰.xlØ°A---ÊÍÍõi_¼x±zzzTTT4ê¹>ýéOwyÀ”CAmm­l6›æÏŸïÓ/IÚ½{÷¨ç æi7€2 ˲TWW'IŠŽŽöés¹\²ÛíêèèÛÍÇ £EFSS“Ün·"##‡¬^„„„(&&Æ{€Ñá¹ atuuI’BCCýö÷ööJ’"##Õ`ôÊÊUX]8äð³ÁÃü&ë¬VO«b7Åzÿ4úwÀ©ƒ2 —Ë%IêèèPgg§¦OŸîÓïñxäp87îß;??_QQQ>mYYYÊÊÊ÷ « Õx°qh‡[Z_µÞø/QŽ0‡ä–¬!‡²9ÂFkFhw€ÉTVV¦²²2Ÿ¶Á7©qâ Ø1c†‡<´hÑ"o_gg§úûûuÙe—MÈ÷.))Qrrò„Ì œ.ß½µÛìrF8%]q°†¼³kBqj±ÖW­ö]e PÚß`2ù{¸¶¶V)))“TÑ©2‚œœ=öØcÚ»w¯OÜœž””4Y¥%g„SÍkš%I±›b‡¬>˜’>7wŽ1¥Êߧ6¡ //OÁÁÁÞÀ1¨¢¢B111Z·n·­»»[}}}ÃÎ500 éè§k§+ÈUZZª-[¶¨¡¡A’´gÏmÞ¼Y;vìЬY³$Iííír:JHHÐáÇýÎõÁH’:d¦x ñÖqdgg+,,Lk×®•Çã‘ËåRee¥–,Yâ½&<<\.—K3gÎTPPÏøüã*//×3Ï<#›Í¦ûî»O úÎw¾£ððpÓ·L*È(ddd(##cØþˆˆÕ××ûí›3gŽ UXX8QåS`0†ÀÁ€cüíÝpõýÇñ× Æ´(Ð6“Æ%V*4àð#ÈÐo[43ůÁêÔ(Rf*`«´´ÕJ‡±6`©‘ßQ ú¥¢_Æ „ª)ÖB*ê T´%@B¹—ÜçûGLÈ%w1—»—Üó1“IòÙýì¾w÷vïÞ·?ÞÁ*AS¹}ÝYÏY%­KúÜq:ÛŸê躂ÚZ ú3T.G_/UI>ãët=Ö¯ó ý©Ž  H@ @• ›Q¹}Ñc3ÓÊü•®dÞöu¨?ÕÑt Ѻ4ЗÍûƼn¥ÔŸê躊›Ð؆€mH@؆€mH@؆€mH@؆: ú¼—Ý/+ç`N»"iÂS }¶YsÄPë–›fÛéõƼ„†@Ÿ—s0GÇËŽos4>OGû¬ª¤•ù+CN‚N³ÍôzcÞBC Ïkþ&Óe¹”xubK{ó·™ÂK°}ö¬ç¬|ÆðìDW¦hz½1o¡!Ðo$^¨ÓËN;€Nj»Ï&­KRqUqM³£éõƼt7¡° Û€° Û€° Û€° u@€ö²ûeåÌiW̪¹(vÛ e÷¬ç¬m±ðwÖsVIë’Zþîi=qL %ÆÖãvv>ýí¸ßß–á‡èa9st¼ìxûUÒÊü•¸eWMoŽì/UI>ãkWÀ¯'÷ÅîSC‰1踘O;î÷·åAø!zXó7F.˥ī%5}£æ3¾vß&¡I(ë,иҕoæØã±ieþÊ ß’÷”îSC‰1иO;î÷·åAø!zIâÕ‰:½ì´$)i]R»oßÐ^(ë¬õ¸ì7ïólý&¼+ÇÔPb 4n¨ÇîþvÜïo˃ðÁMèlCÀ6$ lCÀ6$ lCÀ6<†ŽënÅU;*¶›GOϧ78YѶ£õ֚ݕÌÛV:4<’Z?N½Î½†"}ûD‚ÎV.ïí*ìæ#u½ {°¾=}œezÝÙǨ˜Ž®"ãº[qÕŽŠ­Uß÷Ê°NV´íp½ÐÛ•Ì;ªŠìD<á¦ÃõãÐë¼£×P¤mŸHÐÙÊåvUaïNuôPúöôq:”éug£b:ºŠŽënÅU;*¶«¾Ý*Ã:YÑ6Øz ÄŽJæÁª";O¸ ¶~œ|{ Eâö‰­\nWöîTG¥oO§C™^wö1*¦£«H@6º[qÕŽŠ­m«o÷¥Ê°NV´ —ªåvWnîk‚­Ÿpx‡Ëk½«³û¨]ûrwª£w¥oO§C™^wö1*¦#TÜ„À6$ ÈÏÏ×í·ß®ôôtÍš5K»wï©^^žæÌ™£iÓ¦éÖ[oÕñã¿èH@‚8räˆæΫE‹©  @ëׯ×~ðåæævªÿúõëµfÍ=ÿüózë­·”••¥I“&éСC½9zÍN€Žlß¾ÝéÛ&|±mÂï;è§H@(--UVV–¦L™¢3fH’RSSµtéR=üðÃ*.îøÚÆ¿üå/Zºt©-Z¤¤¤¦GïÝsÏ=3fŒxàcz}Ð >t:t„Rá‹m¾Ø6aŽ÷ôS$ <ñÄ*..ÖâÅ‹ýÚ322TSS£U«VuØÿ'?ù‰\.—²³³ýÚ§M›¦÷ß_/¾øbÇ ô<+€ÂÂBY–¥I“&ùµ9R’ôꫯí[WW'·Û­#FhØ°a~ÃFŒ!IÚ³g,XÐÃQ£¯j>!ÖÕߟ7Ž¯î*©îjù¢ãtéRS»¯&^ªý‚|Q_Peeקýy¿/—Eª¾JÞÚ8q¥Ýã‘ŽëåèèwoN;\â¯yÿ¿¤ÚrÕÄÑΡϣ¸XÚ¶­kñU¿ûßRm…ªc¯Õ¦MỎÂuÍ?mçáó5ýv»¥eË‚÷mûws¿`ñ·í'IýJòV«"êjÝóAS{Åñ'¥†Ëº§»Þ ܯ£øƒÅìwsÜÍÚ.Ggú6ÿ_v&OòÕªÌ5H³^kj/+Ù$5Öéü€XMos»hëyŸ?÷?’¯Nç]±Joµ//{QòÕ霫I/5µŸ»°]:ó°ÎmxJ7mï8ƶËqîâNÉçU©5P7nij/½¸K2õ*µb4¦Õ¾TZµGòÕ«%ü4 IDAT﬑٠>í’êÿ“LƒJ¬hÜðYÛå×%ãU‰jÄú¦qKjÞ| *±¢”òÌ•iDEI³fIøƒY†ëüc4xð`y<ÕÕÕ)*êJŽæõz5hÐ cTYY©øøözŽ9¢›nºIééézë­·ü†Ü®oaa¡Æ¯#GŽ(--­çÎF?ü¡”Ÿ/y½Mÿwtмà-–Ïøä²蚨/K’*¼MÏ·äÒà¨Ä€}›ÿ®j,“‘O– « ‘1Rµï‚Œ1²äÒU®k‚ö4½@×›©až4àm jÖ`ê>ÃÒ ”1’O^I–$É¥¨¦7G5¶´Épâ±wÜ"駃@@l›ðŶ oýkûÌž-ýùÏNGÑ}ýéóšS8ÒFQQ‘ªªª4xð`¿äC’¢££• ’’iìرíú¿÷Þ{’Ôîì‡$}å+_‘$ýë_ÿ 8ïÚÚZIÒ±æ¯qû°'Ôòwgù$•«Ä¯ÍHªÐÙNõ7’ªôi»¶j - ª¤†còÚ Í  ½¥RR¡ÓA ¶MøbÛ„7;·Om¯Ì·¬+5H–iù¯i„+ÿGY±’¤Õ¶Ǻ2®Œ}rL……ã{ŸO.וә%%M—K@†ª;ï¼³'½dÊ”)N‡Ð§q7l .”1F}ô‘_ûÑ£G%IãÆ Úwܸqºá†TWW§“Íî ¡?П‘€°dÉEEEµ$ ÍvîÜ©„„=òÈ#-m—/_V}}½ßx>ø $ùõ÷ù|Úµk—¦OŸ®¬¬¬^Œ_$ ¤¦¦jÓ¦MÚºu«Ün·$éСCÊËËÓŽ;ôå/7=*öܹsJLLÔ׿þu54\¹ýêG?ú‘î½÷^­^½ZÕÕÕ’¤uëÖ©¼¼\/½ô’ßeYqæÌ-[¶LË–-ÓªU«Zž´…ðàóùtË-·èºë®s:”ˆgŒÑ† 4vìXÅÅÅiôèÑZ¾|¹.^¼èth)??_·ß~»ÒÓÓ5kÖ,íÞ½ûó;¡×ø &L˜àt(øLqq±î¸ãeggkÑ¢EN‡Óç‘€„¡={öè׿þµî¿ÿ~}ík_s:PZZª%K–èü£bcc’æÏŸ¯´´´víÙÙÙ’¤o}ë[v‡± eY–&Mšä×>räHIÒ«¯¾êDXoÀ€Z»vm»ö#FhúôéŠÕõ×_ï@dhí—¿ü¥n¸á}ÿûß—iW Nðz½ºí¶ÛT]]­5kÖ8N¿@†~ñ‹_ÈåriúôéÚ¸q£²²²4eÊ­Y³F^¯×éð"^}}½.\¨ 6(!!Áépð™`w¸xñ¢\.—æÎksD‘ÉÓòÀ¡C‡ú KNN–ËåÒùóçUUUåDxϲ¬€í—.]Ò·¿ýmEGGÛZ{å•Wtüøq=úè£N‡‚VöìÙ£wÞyGsçÎÕ_ÿúWåää(--MóæÍÓáǯO¢a/[¹r¥¶lÙÒ©q-ËÒÛo¿-·Û­áÇ«ªªJÙÙÙºûvíZåääèÀÊÏÏú&‚Єº}:¤Õ«Wë žK/ uÛ¼óÎ;-O¨kíõ×_ך5kØ^6)**RUU•¬¨(ÿ·˜èèh%$$¨¤¤DEEEA‹²Â^¥¥¥*++ÓÞ½{%¢¹Ýnåæær†0 H’bbbtã7ê{ßûžî¼óN}÷»ßUzzº^y媢‡Êé›Pàoï޽Ʋ,³`Á‚vÃ&Nœh,Ë2Ï?ÿ¼‘Ácž~úióøã·kç&ôðäv»MVV–ÓaD”Ç˲LBBBÀá×^{­q¹\æÔ©Sö† î¾ûnSPPàt­¼¼ÜdffšÒÒR¿öÍ›7˲̖-[Š Æ\ùüuòäI¿öæÏlÉÉÉEÖwq V˜i®¾>|øðvÃfÎœ)IzóÍ7m MJKKõÐCé…^PjjªßÏ™3gtúôi¥¦¦êæ›ov:THª­­Õã?ÎClÖ|sóùóçUQQÑn¸ÇãQ||¼RRRlŽ ìÚµKcÇŽÕ”)Sœ%¢­ZµJ~ø¡222üÞ[V¬X!IZ±b…®¿þzíÚµËáH#ÓG}$˲Ú}6›6mš  Ó§OëÔ©SE×7q V˜IMM•$•——·–™™©'Ÿ|’k§RSS£ÆÆÆ€O‰1ŸÝ(øñÇ«¾¾ÞîÐÐFCCƒ–/_®Ç{Lƒv:œˆ2lØ0ÅÇÇËãñÈívkêÔ©-Ã***äõz5qâD#D³ƒêèÑ£<•, x<•——ëÂ… ~íÍï-%%%*--Õ¥K—œ/â=Z………*//W\\\K{\\œ&L˜ ¿ýíoòx<FØ÷p$ÌŒ?^’V;o¾žz̘1¶Æ„&)))òù|jlll÷“œœ¬””566ª¨¨ÈéP#Zcc£V¯^­Ÿþô§1b„ÓáD¤… ÊÓrF·YóÍéÜã¼wß}WGŽ!ù›7oøþÒ|wË–-jllÔ]wÝåp¤‘i„ 2ÆèÿøG»aÑÑÑŠŽŽÖèÑ£ˆ¬ï" 3 š?¾Ž=ÚrÓS³?ýéO4hPËcEø»|ù²~øaÝÿýí.ñ9pà€ž}öYg‹0K–,QTTTKÂÑlçÎJHHÐ#<âPd¤½{÷êïÿ»–/_î×~éÒ%ýøÇ?V]]C‘¡-ÃcxÃÂâÅ‹« 6øµŸ;wNºï¾û4pà@‡¢ë›H@ÂÐÆ5jÔ(Ý{ï½*++“$íß¿_»wïÖúõë•””äp„@ø9wîœ233µcÇÍž=»åê‘#GjØ°aºù曹?Ç&©©©Ú´i“¶nÝ*·Û-I:tèòòò´cÇŽ€O+ƒ=òòò4oÞ<åææúÝkœœ¬/}éKª¬¬TLLŒÓaaå›ßü¦~ûÛßê7ÞPnn®$éÂ… zôÑGuÓM7Q¤ ¸$ ÅÅÅéí·ßÖ¶mÛt뭷ʲ, 2DÛ¶mÓ´iÓœðXdgÕÕÕiòäÉúä“O‚Ž3~üx5ʾ "Ü‚ £Ÿýìgòx÷õ¹>·¥¢¢¢BÀP~F7 ` è˜:&@€€  `è˜:&@€€  `è˜:&@€€  `è˜:&@€€  `è˜:&@€€  `è˜:&@€€  `è˜:&@€€  `è˜:&@€€  `è˜:&@€€  `è˜:&@€€  `è˜:&@€€  `è˜:&@€€  `è˜:&@€€  `è&PZZªÇ{LÓ§O×Ô©SUVVVm›?þX~~~?ýû÷·¯_¿~½þøÇ?*11QŸ|òIS6àÖ¬Y£(77×¾lýúõºþúëÕ«W/½ø⋪¨¨°¯«ËgpÔÌè@JMMUTT”&Mš¤ŒŒ ͘1C©©©Ûüýï×믿®Ë.»L‹E~ø¡ýËÎ?þ¨G}TÛ·oWII‰úõë§M›6©uëÖF<€—ÉÏÏ×Ù³gõÅ_Èb±H’Š‹‹õÖ[o驧žÒ–-[4uêTuëÖM£F’T·Ï6àÈRQõr7šÜ‰'©#GŽ¨U«Vúæ›o§ï¿ÿ^ÁÁÁöí¾þúk]zé¥öÇ“&MRbb¢zõê¥GyDzþùç%I¿ÿý凜””&>ïtöìYéСCêÔ©“¶nݪ=z¨eË–’¤Ûo¿]eeeÊÊʪógpDŠ»ÁvîÜ©ððpµjÕJ’©‚‚íÚµËa»ªÁyyy¹öìÙ£^½zI’vìØ¡ž={Ú×wêÔ‰4w€K8<îׯŸ=8—¤=z¨K—.’êþÙ‘ân°Ý»w+,,Ìþ8 @ÊÍÍÕUW]åtŸœœ]}õÕ5£C‡Z±b…Ó};¦üãêÒ¥‹Z´há¢g¸Ó§OëСCúÍo~£ððp£›ãr«V­Ò_ÿúWIõÿlã³ ŒáíŸMžˆÝ`‹EþþþˬV«BBBjÜgåÊ•;vlÇ(((¨qÿüã?~|#[ h¨·ÞzK·ß~»ÑÍp©mÛ¶iܸqŠŠŠ’TÿÏ6>›ÀXÞøÙä©Ð 6pà@%''Ûª¨¨Èž¾îÌ–-[ô /8ãÀöÇyyyêÝ»·Ó}»ví*©²VM‹‡”˜˜¨ôôt£›aJ¼6ÎñºÔŒ×¦º={öhüøñö¿ÃÞ¢¤¤DË–-ÓܹsíËêûÙæKŸMîì›6mÒ™3gÔ¼ysýêW¿rË9꣩þ >\ùùùjÛ¶­Ö®]ëöó9ãKó|å¹ÞyçºãŽ;LÓŸÜÅ[?›<ºÁbbbT^^nÿ`ÉËËShh¨:wîìtû;wVû‚3pà@íÛ·Ïþ8//O}ûöuºóæÍ%I={öTll¬‹ž…whݺ5¯I xmœãu©¯MÍl‡½Åܹsõä“OÊÏï\Y›ú~¶ùÒg“;ûÆ¡C‡têÔ)›âulª¿özÞ¾ô7ÏWžk‹-Ô¡CÓô'wó¶Ï&OF‘8ƒùûû+>>^[¶l‘$eggkÖ¬Y’*¿ôT ¼¥Êôö1cÆ8,5j”þóŸÿH’N:¥ƒêÞ{ïm‚Ö|ÅÙ³g%UÞßÜæ駟VÿþýUXX¨èÕW_Õ×_]ëgÜg̘1?~|µï ê¯]»vô'‚tÈÈÈPJJŠvìØ¡ÒÒRM›6M’”••¥nݺ)::Ú¾í† ôÄO8ìß³gOM™2EO=õ”ÊËË5þüZç°PEEEzõÕWe±X´téRM™2E .Ô¬Y³ï^½zé®»î’Tóg¨º „„„8˳uëÖjËÖ­[çôãÆÓ¸qã\Þ6BBBôðÃëᇶ/›>}º¦OŸ^ë>¾0OW"@þ'!!Áè&˜¯s¼.5㵜sgß°Z­ª¨¨ÅbQhh¨ÛÎSW¾ôw€çê}n¾ùf˜¦?Áw ÿã+8 Ákã¯KÍxmçÜÙ7–,Y"«ÕªÐÐP%%%¹íF7ðô'…9è˜:Üî‹/¾¨¶ìË/¿Ô­·Þª»ï¾»IÚP\\¬ŒŒ uíÚUß~ûí·?uê”öíÛ×-€Jèp«Õª)S¦(00P’4bÄM™2EGÕÚµkµ|ùr•——7I[Ö¯_¯·ß~[¹¹¹öe³gÏV¿~ýTQQQmûàà`åää(++«IÚ@cmÛ¶M999Ú¶m›ÑM<ý F!@‡ÛÜwß}º÷Þ{uÅWH’6nܨ?üPš:uª"""š¬-7Þx£–mÞ¼YÛ·o׉'œî3aÂíÛ·OË–-«Ó96mÚ¤?þ¸Ñm !rrr´aÃåääÝÀãÑŸ`ŠÄÁ-æÍ›§€€õêÕ˾¬sçκôÒKí›7oÞ¤m²äÛtêÔI­ZµÒÅ_\ã>O<ñ„¬Q£F颋.ªõøsæÌÑã?î’¶P_“'O6º €× ?Á(Œ Ãå •ššª &8,ŠŠRÏž=í-K“¶ËÏÏñí­=zÔºO`` ~õ«_é駟®u»9sæhõêÕNÓåæVZZªùóçkûöíÕÖ?~\óæÍSÏž=õÆoкJ§NÒ«¯¾ªhÆŒõÚ7--Mû÷ï¯×>7nÔí·ßîð¹]›œœuèÐA .¬×yŽÐQg}ûö•ŸŸŸüüütå•WÚ—?þøãjÖ¬™"##%I™™™ Ô!CöïÕ«—Ⱥ$UTThË–-úÃþ Î;kæÌ™ë:¤Ç\7Þx£.»ì2Mžÿüs}üñÇÊÌÌTqqqö9vì˜òòòôÝwßÕi{€s訳­[·jÈ!²X,úè£ìËŸþy]}õÕÚ±c‡$é­·ÞRçΫíß»wo‡€¸¢¢BŸþ¹N:¥¥K—꥗^ÒôéÓõïÿÛ¾Í}÷ݧfÍšiõêÕú¿ÿû?edd襗^’$•””èðáÃ:|ø°-Z¤;î¸CW^y¥,‹~úé' 4HÝ»wWNNŽþô§?éí·ßvørÕ»wïj œ¹ì²ËTVVVc‘›o¾Yééé’¤_|Q}ô‘~ñ‹_Èb±èßÿþ·V¬X¡¾}ûêúë¯W`` ¾ûî;Ý|óÍzâ‰'´mÛ6uëÖMwÞy§Ã\øûï¿_}ô‘>øà}þùçúàƒ4qâDûú_|Q6lЪU«´wï^µhÑB#GŽ¼àsœSRR¢±cÇê…^ÐÀ%IPnn®Ö®]+©òïñwÞid3%IãÇ׃>XmùúõëõÙgŸéäÉ“5îëç秴´4eeeiýúõ¿`:êO€»P$õòë_ÿZ111ÊÈÈиqã$I¯¾úªþøÇ?J’Îœ9£üü|µk×®ÞÇ–Ty{6I ѶmÛÔ§Oëý÷ßWPP ìû4kVù¶}¡²yã7Ô¯_?‡«ÿÎFõë"$$D-Z´¨5@w¦¦¶ýþ÷¿×µ×^+I:|ø°¶nÝ*Iöçµ~ýzéºë®³ïóÈ#è‘G‘TyÛÂÂB‡{ÈŸ9sF]ºtÑ'Ÿ|b¯š¨Ùž={ôÊ+¯(33Óay›6mÔ®];]~ùåöeþþþMÝ<§œµ#&&FmÛ¶•$ûçeMºwï®ððp¥¤¤hþüù :Ÿ«ýíoÓ}÷ݧk®¹ÆíçêãBý pFÐM ´´T=ö˜¦OŸ®©S§ÚçX;S^^®—_~YYYYÚ¼y³}yQQ‘ÂÂÂìsÄW®\é¶ö>øàƒÚ°aƒöìÙ£S§NéĉêСƒ¤Êq ¨÷qmé祥¥öe={öÔsÏ=§Ûo¿]]t‘Ú´isÁBl'OžÔÑ£Gë\ئ.Z·n­Ã‡»äX-Z´¿¿¿î¸ã¥¥¥)..NÒ¹QøÜÜ\•••Õø>øî»ïTVV¦·ÞzK}ô‘>úè#åääèàÁƒzà\ÒFð$Û·owø ×ûï¿o_÷Ýw+((H-Z´°zûóŸÿ¬–-[j̘1ÕŽwÅW8½ØYQQ¡/¾øB÷ÝwŸ¢££5lØ°jõ?JJJôÚk¯éöÛo×ÀÕ·o_Mž‡)`ÎÚzõÕWkðàÁºúê«k|­n¼ñF-X° ^EãJJJ4{ölÅÇÇ«K—.zöÙgÖé•W^qZÀîàÁƒzì±Ç4lØ0…††ÊÏÏÏž9÷þûïkÁ‚*--Õ—_~©!C†hÈ!â£Ô¥?î@€n©©©ŠŠŠRjjª:vìXcuÖ‚‚ýö·¿Õ 7Ü Ûn»Íájók¯½¦%K–hݺuZ·nFí¶öŽ?^­ZµÒÂ… •••¥ßýîwöuááá W~~~£ÏcµZ5bÄÓ”)S”‘‘aŸc]XX¨°°0mܸÑ^!>..NÇWqq±ÊËËUQQ¡;î¸C/¿ü²ÒÒÒ´iÓ&=òÈ#zî¹çêÜÖšôéÓGåååúôÓOë´ýÉ“'•ššª)S¦èoû›žzê)MŸ>]«W¯¶o³lÙ2½ýöÛÕ ØIÒ¤I“tà 7è_ÿú—Ž?nϸ“¤›nºI~ø¡¤ÊB´¶‹Î M•o@€n°'N(--M¿ýío%I#FŒÐ‚ œÎKLLÔC=T-eº¬¬Lk×®Õ°aÃ4tèP :´Ú-Å\)$$D&LÐ_þò­^½Ú>×ÚæòË/×±cÇ}žwÞyG6lÐÔ©SíË***T^^^ë~~~~;v¬¾ùæ:t¨ÚúšF¦Ïœ9ã0zoóóÏ?ëÌ™3µ¦ÈÛ^oÛIm¦Nª+¯¼²ÚYÛó²Í«_µj•ÃúåË—+$$D—_~¹BBB´`Á-\¸P:sæŒfÍšUçy…àmZµj¥gžyF’´aÇuùùù*//×Í7ß,©²VÊ©S§ìÙ_ç›={¶:vìXmù-·Ü¢aÆIªœŸ=|øp}þùç:xð }›_|QQQQêÔ©“$)>>^~~~öÑïÖ­[«ÿþöì©~ýúéÆoÔ+¯¼¢µk×Êßß_o¾ù¦–-[¦… Úk·\vÙeN ©ýæ7¿ÑwÜQ×—I’¤/¿ü²NÛ‡„„hÖ¬Yöih¶ŒÛÝK¤Ê ÜYÛ<¨?üО‚¨3fT»0 8‡Ý`;wîTxx¸Zµj%©²¸XAAvíÚå°Ýºuë´yófåçç+>>^™™™ö4¸5kÖhíÚµêÚµ«ž}öÙ&I {àTXXè4í§OŸ>ÊËË»à1 TXXh\RRb_.IíÛ·—$mÚ´IZ¹r¥~úé'é»ï¾ÓÙ³gíûœ¾„„I•5>¬“'OêŸÿü§¤Ê/ gÏžuؾ°°P]ºtQ=tìØq}ðôÜs•?Ï<ó£$éðá_Ú—=÷œ”‘!Ù ¯ÛÚºråJ}ðÁ:pà@mëСƒ¾üòK:uJ?ýô“Þ{ï=I•)‚_ýµ†ªaÆéå—_ÖÓO?­Í›7ë‰'žPYY™BBBÔºuk=ùä“:zô¨zè!µk×NaaaúøãÕ¿ÿ ¾îà­n¹åuêÔI¯¿þº~þùgûòE‹9L²Ô5è59¿hZhh¨$9\”ÎÈÈPFF†ýq~~¾ZµjUíÂu`` $Ùk’TõôÓO«uëÖúõ¯í°ÜV4µ&ûöíÓ®]»´oß¾·±èÎîùîÌùSÖœ=gÉyA¹:¨eË–úÿïÿiùòå’*§y-]º´NçŒT—þ¸ºÁvïÞm¯~.U~FDD(77×a»>ø@}ûöÕèÑ£õè£*))I ,$9RV«UË–-ÓK/½äp;.w‰ŽŽÖ¨Q£ª¥öIÒèÑ£µÿ~{±·ó驧žÒO?ý¤ 6hñâÅúþûïíEÐþþ÷¿kÍš5úÍo~£{î¹G÷Þ{¯n½õV•––jÔ¨Q:~ü¸òòòôÑGéå—_–ÅbÑ“O>©Ï>ûÌ~Ž_þò—ÊÌÌÔþýûÕ¿M™2Å€úé§:räˆC›,‹š7o®o¾ùF/¼ðw)¥¤Tþ¤§¥ŠŠK´lÙ(û²”éÁ¥É“+÷ïÞ½»Æ¯U«Vé믿Vii©RRRd±X´hÑ"ûýÏ%é‰'žPTT”bcc•’’¢qãÆ©]»vúꫯì…HV®\©{ï½W¯¼òŠRRRÔ»wo{–…T9 ÿç?ÿY—]v™š7o®Q£FU+t¾Æßß_S¦LQQQ‘-Z$I:{ö¬Ö¬Y£øøxûv¶ÏXÛÈnCقת÷ =z´~ùË_*++K·Ýv›þò—¿(  ZöWM÷T?r䈾ÿþ{uíÚµÚº eÇeggkùòåÊÎήq[€ÞЪéΞ³ä¼ \PP222ôóÏ?ëÖ[oÕå—_îP0³ºô'À¨ân0‹ÅRíCÍjµ*$$ÄaÙÖ­[uË-·( @ƒ Ò¸qã´bÅ =üðÃ’*+ >\[·nULLŒŽ;f¿=˜3‰‰‰jݺµÃ²„„ûÈs]ÔTˆnÈ!;v¬Ö¯_oO'¬*$$D3gÎÔÌ™3–/\¸P .tX¶hÑ"û—,IöÊñ6ç§×WõÛßþÖ!¨•ätþž­M‡Rff¦>úèç­}SÒ$9»žõŸÿTÙêÍ7Ö½ûî»NÏuå•WVK¿<¿]‹-ôÒK/ÙïùîÌäÉ“5Ùv…€©dffV»hVÛ}ªá:wß}·RSSµ`Á%%%éÝwßÕ¸qãb[êªÿ“ª…Ý:¤Ñ£GkèСzë­·Ô¬Y3{aººøþûï%U^Ì®/g÷I?Ÿí;G}o¡v¾ lµ?~¼®¿þz͘1C¯¾úªâããõ‡?üA¯¾új£Î¸[]úàè8p ’““í UTT¤^½z9lwþ•ê¸¸8§#¦íÛ·×СC•““£Q£FÕxÞôôtÅÆÆ6²õ5›?¾&Nœè4@7«’’mÙ²E­[§Û—=ùäz÷Ý£š3'EUëÝz«tæŒTÇï'|Œ³ žÛ·o·Ï;†û„††êž{îQZZšÞ~ûm½þúëö©D6¶Ï¿ºLǪãÇkÀ€Š‹‹Szzú…wp¢{÷î’¤o¿ýV¥¥¥N‹ØÕ¤.·ƒ²]v6Bï.—\r‰222”””¤øøx-]ºT&L¨–˜ ·WƒQHq7XLLŒÊËËíUÏóòòZ­ Ùå—_î0&((È^¨æ|~~~êÛ·o­çýþ{é믥ó2½]¦]»vúÝï~Wë°™?~\ýë_5gÎ?n[zF;v<¦µk³ Q£dÿ±MÉ#@ó™2eŠš5k¦Ç{L111öyÓ6=zôP`` Ž=êÒó~öÙgÊÏϯ–:_^^^cÓó—·iÓF¿üå/uæ̽öÚkN÷9¿ŽÊùªÎ¿?Ÿ-@?¿à¬;lÛ¶M³g϶?îÞ½»æÌ™#©rŠŸM@@€}N{EEEµA ð%èó÷÷W||¼¶lÙ"©r¾Ë¬Y³$IsçεåIIIZ¹r¥}^÷† 4vìXI•÷Ý»w¯$éÓO?UÇŽ/xK°›n’ºw—Ú·—žxÂ-OMwÜq‡âââÜzOvW¹øâ‹õûßÿ^AAAªüŽpVÒ‹š1cŽÓ/1¶LIt0ŸÈÈHÝzë­:zô¨}*XUÍš5Óå—_®~ø¡Ndz]Dÿ駟–ÛÒÐm#ñ×\s‚‚‚´bÅ ½öÚkzýõ×5lØ0éèÑ£Z·n}ú“mßT;ßK/½¤M:Uü±¤ÊÔ÷%K–Hª n‹ŠŠœý .TXX˜ d/VZ•-@¯-ËÎæرc|Î6¶ÇUo³ZQQ¡ÔÔT­^½Ú~—”;wªM›6ºñÆíÛ]zé¥Ú½{·Þ|óMM›6;’ðiè&‘‘¡uëÖé¹çžÓÏ?ÿlŸWœ••e¿¯u¯^½”‘‘¡™3gjÖ¬YêÙ³§}>öÞ½{uíµ×j̘1Ú´i“ÒÒÒêuþzL«·þýûkäÈ‘î;Tèþ’’Õ»w{§Û €¹ýñÔM7ÝTc*÷wÜ¡C‡é»ï¾«õ8/½ô’-Z$‹Å¢… jÞ¼y***ÒóÏ?¯7ÞxC‹EóçÏתU«ÔªU+-^¼XQQQzꩧ´e˽ýöÛzðÁÕ¦MmÚ´I×]wRRR”-‹Å¢)S¦8ÔZ‘*}ûv{Í™¡C‡êÉ'ŸÔ Aƒ¡6mÚhýúõÕægffjË–-jÖ¬™>ûì3§Å­rrrÔ®];Ýzë­5>ç½{÷êž{îÑ©S§TPP {ï½W[·nÕǬ &Hª¬p=mÚ4=zTYYYzá…d±Xôî»ïjÞ¼yöQþ’’9RmÛ¶Õ!Côã?jóæÍöÛÐIÒóÏ?¯öíÛkñâÅ>|¸Ú´iSëÿ Ð233õꫯR„MÎRQ×*ð ¶9#FlÓ¦M±*,”.¾Ø”B’zôþû_)4TªéŽumÚH'OJÑÑ•/2ïW IDATÛÀ…Øþþn۶ͭ5@<‘¯MEE…®½öZ%$$èþûïo’sºÛ¿þõ/>}Z-Z´Ð_|¡ØØX >ܾ¾¨¨HíÛ·Wrr²ž|òI[êZ‘‘‘:|ø°:tè`/²4VÕþTÓ´RoÀg“ùP$ÎGÍœ)Ýv›´¿Tô8Ÿe»XQK|FÐÀÃY,½öÚkºå–[t×]wÙïKîÉlADii©æÍ›W-ÏÊÊÒW\¡Ç܈æÅ›ƒr˜º³ÝÝÍô’iûöêϽ¢B²M·«-@·ÝŠ–}Z-Z´ÐÈ‘#nàÑèO0 º« —•9>öfèoTRR¢ââbù74= €ý F!@÷aUÿÞøRª6EâÞh̘1F7ðô'ÅGÆLáLÕs_ ÐA`FŒ û°óSÜ}Eâ€/Ø¿¿¬Vk½ö UTT”›ZàBÐ}#茠¼‡ÕjUEE…,‹BCCn ¶ÿ~EGG7hß}ûöù|N‚QÐ}˜¯ÎA'@x£%K–Èjµ*44TIIIF7;7rþ–¤žuÜk¤ñõu÷Fô'…݇ùjŠ;EâÞh̘1*--U³†~¸ÁKõ”kt#<ý FáçÃHqgà=ºtébt¯A‚Q¨âîÃHq§Hó @÷aŒ 3‚ÀŒwFÐÞc„ F7ðô'…9è>Œ÷Æè¾ta€{ û0FÐ_Åt®BŠ» ”––jêÔ© VII‰fÎœ)ÿ"Çòòr-^¼X­[·Vdd¤®¹æIÒúõëµzõj•••é–[nÑ Aƒ.x^_ƒNŠ;3"@7ÔÔTEEEiÒ¤IÊÈÈÐŒ3”ššZm»‚‚Ýu×]š3gŽºuëf_þã?êÑGÕöíÛURR¢~ýúiÓ¦Mjݺu­çõÕtŠÄ¼Ñ¼yódµZª¤¤$£›x4úŒB€n°'N(--MGŽ‘$1BqqqJNNVpp°Ã¶‰‰‰z衇‚sIš3gŽ† &‹Å¢   ÅÅÅiáÂ…JII©õÜÌAgà=† ¦³gÏ* Àè¦þ£0Ý`;wîTxx¸Zµj%IŠŒŒTAAvíÚå°Ýºuë´yófåçç+>>^™™™ªø_t¸cÇõìÙÓ¾m§NôÉ'Ÿ\ðܤ¸S$à=úôé£ØØXõéÓÇè¦þ£ l÷îÝ ³?PDD„rss¶ûàƒÔ·o_=Z>ú¨’’’´`Á§ÇèСCµýñÕwŠÄ0#tƒY,–já¬V«BBB–mݺUƒ R@@€ ¤qãÆiÅŠNQPPPmgHq'Å€y0Ý`Trr²ýqaa¡ŠŠŠÔ«W/‡íŠ‹‹ÇÅÅ)33Ó~ŒØ×ååå©wïÞµž711Qß}w®ˆÜ#H÷ߟ „„„?OáŠ"qj“™™iÿmsòäIƒZ_±mÛ6•””(00PqqqF7ðhô'…Ý`111*//W~~¾Ú¶m«¼¼<…††ªsçÎÛ]~ùåÚ·oŸýqPP† &©2@¯º.//O}ûö­õ¼ééézóÍX½øbåã^ pÑ“29WTq· K•AzÕÇPý‚çöíÛù’·ÊÉɱWæ½4ý F!@7˜¿¿¿âããµeË9RÙÙÙš5k–$iîܹ=z´¢££•””¤n¸A³gÏVhh¨6lØ ±cÇJ’F¥I“&I’N:¥ƒêå—_¾à¹Iq'@€ºZ³fžyæ½óÎ;ö‹È¥¥¥š:uª‚ƒƒURR¢™3gÚ§\Õ¶î1yòd£›x úŒB€nJIIÑŽ;TZZªiÓ¦I’²²²Ô­[7EGG«W¯^ÊÈÈÐÌ™3uÑE©gÏž7nœ$©gÏžš2eŠžzê)•——kþüùõžƒN‘¸ú9?@‡÷ËÍ•&N”ªÌ&q™¤7ß”š7wý±WÈÏÏ×Ù³gõÅ_ÈRå`jjª¢¢¢4iÒ$eddhÆŒJMM½à:ຠ„„„(==½Úò­[·:<=z´FíôãƳìuÅmÖ\7‚ï·t©ôá‡î9ö·ßJ R|¼{Ž4VÛ¶m5bće'NœPZZšŽ9"I1b„ââ├œ¬3gÎÔ¸.88¸ÉÛ€§ @÷aŒ 7¾HœD€î+~þùÜï_,6þ˜EERAAõãfàðxçÎ W«V­$I‘‘‘*((Ю]»túôé×]uÕUMÞv<ºóÕ9è®.çK7|YÕÿçÕ«¥þýÌÅ‹¥{ï­~|ÀìÞ½[aaaöÇŠˆˆÐ¡C‡tüøq§ërss ÐÝhñâÅ*,,TË–-íµi4 ý F!@÷a¤¸“⎺«za§êÅ­ÆðÕ,x‹ÅR­è›ÕjUË–-õÓO?9]w¡ú(‰‰‰jݺµÃ2gñá\Ÿ>}T\\¬   £›x)͞횶4Ä5×H7¤ûŠ³gÏJª¼¿¹$ùûû+>>^[¶lÑÈ‘#•­Y³f]p¨™‹¾fÂùêtw‰{ÿ}÷e#äçK›6¹çظ03Ž /]êšv4ÔæÍÒ¾}ƶM£¨¨H/½ô’,‹–.]ªcÇŽI’222´nÝ:=÷ÜsúùçŸ5yòdû>µ­ƒ{ìÛ·O»víÒ>:&Ðhô'…tFŠ»ëSÜ«¾Ž R›6 ;~U_~)åäTþþ¿,ÀŒ#è¶÷CHˆ4r¤kÚT_|!Œw׉«ú:¦¦JQQ ;~Uóæ Ы¶MËŒEâlûtè ee¹¦MuqÿýÒË/WþîK÷³{ðÁnà5èO0 ºóÕw[ëç×ð¹³u Ð]ÄUå'@7ŽSÜmÁ±«ÞkuÅ{0'oº`4úŒÂtæë#è Mo—j.WõwtïbÆwÛ>®º`PWUÏÇ:€ë û0_ƒn{® Ðk*WõwWU¶&@73§¸7õzÕ÷$€ëâîÃAoø1Hq÷=fLq7*@¯úžô¥‹{€ÙeffêÔ©S VBB‚ÑÍ<ý F!@÷až:½¢BZ¼XÚ²¥aû>\ùoc‚,tßcæwFÐHRxx¸NŸ>­-ZÝ”jöïß/«ÕZ¯}BCCåŠj«@˜¹?Á» û0OMqÿøcé¾ûœÀÀ†ï[S€ÎtïeÆw3‰ó¤¿€·6l˜ÑMpjÿþýŠŽŽnоûöí#H‡!ÌÚŸàýÐ}˜§¦¸:äšãÜ~{Ã÷­©H#èÞËÌ)î‰`fçFÎß’Ô³Ž{í‘4¾Þ£îàéÐ}˜§¦¸Wmë3ÏH·ÞZÿc„„H:5¼ ‰ó=fKq¯¨8wqˆwž¡§¤X£¦F€îÃÌRÜ›:0±¡Hœï1[Š»;ÚSWŒ ætçwª¼¼\~F}8^„þ£ û0oHq7êofM¯sн—ÙRÜAp¾ððp£›x úŒÂ%!FŠ{Ñâ\ÊÏ—¼¹Ø®™SÜ›ºŠ;#èîA€îÃAo8ôs¬V©wo)"B “æÎ5ºEîa¶w³Œ  ¸)î>Ìæ ›mÝç ¯_/íÝ[ù{i©´h‘ôè£Æ¶ÉHq?‡wÀœ²³³uúôiµhÑB#GŽ4º9€G£?Á(è>ÌRÜA7^Q‘ãã3gŒi‡»¹ã§褸æTRR¢ââbù7õ¼À ÑŸ`tFŠ{Ã5e‘¸€€s¿›1@/)q|lÆ6º‚-õó3G€ndwFÐs3fŒÑM¼ý Faº ”––ê±ÇÓôéÓ5uêT•Õ0$UTT¤°°0ùùùÉÏÏO+W®¬ÓºšâÞpŒ ŸS\ìøØŒmtÛÿ­+ƒaŠÄ *FÐM 55UQQQš4i’2224cÆ ¥¦¦VÛîµ×^Ó’%KÔ¦MIÒàÁƒë´®&ž:‚NŠ»¹œ?‚~ö¬1íp·ª#è®â©)î‰ptƒ8qBiii:räˆ$iĈŠ‹‹Srr²‚ƒƒíÛ•••iíÚµÊÌÌÔE]äpŒÚÖÕÆSç ›!Å"qçøÚº+G«=5@7û{ðUV«U²X, 5º9€G£?Á(¤¸lçÎ W«V­$I‘‘‘*((Ю]»¶[³fÖ®]«®]»êÙgŸUAAAÖÕÆFÐÍœâ™=ò•9èfNqg€$-Y²DóçÏ×’%KŒn àñèO0 #èÛ½{·ÂÂÂì¡ÜÜ\]uÕUöå#GŽ”ÕjÕÆ5qâD}õÕWZ¾|ù×ÕÆæ ›µHœ+/˜=@÷•t³¥¸S$®aÊË¥={Œ¹°°ÓŸ¾e̘1*--U³f|½‹þ£ðŽ3˜Åb©vû«ÕªjÛkøðáÚºu«bbbtìØ1…‡‡_p3‰‰‰*+kmüÎ;RLL‚\ôÌÜÇÌ)îîe5{€~þzYYåkbTvƒ»˜9Å"quSQ!]s”“ÓgËüßOU'›âÄða]ºt1º €× ?Á(¤¸làÀúúë¯í UTT¤^½zÕ¸Oûöí5tèPå8ù–YÛºªÒÓÓ5{ö*I•?ññ«<"8—<#ÅÝ—ôóGÐ%s¶³±Hq?ÇSGпý¶©‚sIJíï빟ô¦:9ðPŒ ,&&FåååÊÏÏWÛ¶m•——§ÐÐPuîܹÖýüüüÔ·oßz¯«Š÷†»P‘8_ ÐÏA—*ÛYõþíÞÀl)îf)çI#è§OŸû=:ZªÃÍ.\êØ1é½÷šöœÀ³ Ìßß_ñññÚ²e‹FŽ©ììlÍš5K’4wî\=ZÑÑÑzÿý÷uÙe—©GúôÓOÕ±cGEFFJ’V­Z¥èèh§ëjã©EâŒ÷†kÊ9èÒ¹K3C¾2‚næôº©z1‰t`F¤¸û°ª_ꕾü²éÎݹ³Ö°}}-Å]ª ˆJJÌøúÊtRÜÏñ†tt`Fè>¬ê—ì¿ý­ò§©I7Jýû×_3¤¸7e‘8‰t3`ýOA'ÅÞnÞ¼y²Z­ URR’ÑÍ<ý F!@÷aíÛKÍ›KgÎ4ý¹‹‹¥U« {BŠ»/èÌAo8o¨âîI:)îðvÆ ÓÙ³g`tSG‚QÐ}X›6RvvåÈyS‰;zTZ±¢ò÷†ÃÌ)îî(' Ð͘:îlÝŒíl,w¤¸7æN ‰«?RÜáíúôéct¯A‚QÐ}Üu×Uþ4•; ›!Å"qç0‚Þp¤¸7-RÜ€ÙQÅMªeËs¿»bÝ )îÌA¯¾ÌŒíl,[ J€î¹#褸³cMÊÕºŠÄ1½ú23¶³±lÿ·Tq÷Žttx£mÛ¶©¤¤DŠ‹‹3º9€G£?Á(èhR®ÐÍœâîî9èf |}eºÙRÜ)Wèðv999öªÓ@ãП`t4©  Ê/÷eeRQQÃŽa¶wFЫ/3c;ËÌ÷A§H\Ý0ÞnòäÉF7ðô'…9èhR‹Rù»'§¸S$î_˜ƒ^QážìRÜ›sЀÙ1‚Ž&ײ¥TPà=)îMY$®¤DZ³ÆµÇn,_ Ðmнc˜:šœmº'Wq7ªH\Y™tã®=¶;xÛôª£ÄfLqg½nHqfG€Ž&W5@¯¨¨m¶ô¦(#}þ¹kéNž4ªZî †½¡Hœ'ý_“âo·xñbªeË–š4i’ÑÍ<ý F!@G“³襥•_˜ëûEÙ #èM==-MŠ•Žwíq]¥mÛʶ=ýtåcO ÚêÂ]zMS%êÂ,)îž:‚N€oÔ§O+ˆ78Ðhô'…Mîü[­5&@7Ûº»RÜ[·–|еÇtµ—^:÷»·èîJq—*ß+ååžUÅÝb9×nO ÐIq‡7ºúê«nà5èO0 :š\ÕýÄ )4ÔùvÍš9tÍ–âÞEâFÀìª΂£S{%ŠÄ9ãÍsÐÝ9Zí©)î}ú4ý9]åòËnàAAAöC‚QA‡é\èËfHq§H\u¤¸7Œ§è™™ÒÛoWÞ*Ñ“´h!Ýz«Ñ­pŽ¦S5Øð¤÷ªº/¦Ïzs€N‘¸ê.¹Dš2¥éÏ àÍÐM ´´TS§NUpp°JJJ4sæLù;‰ŠŠŠÔ±cGiÚ'âBžâî¬HœFöT5a£°ÐõÅþÿ^zï=iÂ×·.Ì>‚î‹ï7oD€n°Ý»w+,,Ìþ8 @ÊÍÍuÐGŽ)«Õª7jâĉúꫯ´|ùr§ÇèСƒV¬XÑtOÂż!ÅÝ‹Äõì)=ôô·¿9þ5ÖéÓÒ‰•¿çç»î¸õaöû  xtƒY,–já¬V«BBBªm¬áÇkëÖ­Š‰‰ÑñãÇuñÅW;FAAÓý«JLL¬6G=!!ÁE0<-Å9è•,éÏ®üq¥5k¤o¬ü½¨ÈµÇ®+3Þ"qæ–™™©ÌÌL‡e¶Ÿ€»üðÃ*++“¿¿¿Ú·ootsF‚QÐ 6pà@%''Ûª¨¨H½zõªqŸöíÛkèСÊÉÉÑÈ‘#5pà@8pÀ¾>//O½{÷®õ¼éééŠmüp‹¥ò§¢Â¼)îÌAo:UJ1èÔ)cÚ@Š;êËÙÏíÛ·+..ΠÁdeeÉjµ*44TIIIF7ðhô'…Ý`111*//W~~¾Ú¶m«¼¼<…††ªsçεîççç§+¯¼RRe¿oß>ûº¼¼<õíÛ×­ív7ÿÊJàž”ân†vy£ªÉ F ›=Å*î$éÎ;ïTyy¹üøþ£ðŽ3˜¿¿¿âããµeËIRvv¶fÍš%Iš;w®=ð~ÿý÷µwï^IÒ§Ÿ~ªŽ;*22R’4jÔ(ýç?ÿ‘$:uJÔ½÷ÞÛÔOÅ¥l Íšâî¬H#šîQ5@7jÝŒ)î¼ßœ/<<\ 7º)€Ç£?Á(Œ ›@FF†RRR´cÇ•––jÚ´i’*Skºuë¦èèhíÝ»W“&MÒ Aƒ4`À¥¥¥Ù÷ïÙ³§¦L™¢§žzJåååš?þç ›mDðBº™FÐÍpáÀUMq7Ã::Ü…ÝBBB”žž^mùÖ­[í¿''';ÌU?߸qã4nÜ8·´Ï¶ÝYÐb†TrŠÄ5RÜãýà}ÐaJfOq§H\Ó1C‘83¦¸SÅÀù²³³uúôiµhÑB#GŽ4º9€G£?Á(è0%OLq7ÃȾ7jÑâ\UÿóGÐÏž•¤O>©Û±¥nÆŽ•šÕã¯ßÿJ©™3gª{÷îÚ¹s§µvíZ¥¦¦***J“&MRFF†f̘¡ÔÔT£›ëÕ¬V«***d±XjtsF‚QÐaJuA7rž÷…æ 0¹–-u½¨¨òõ¶ýßW ÐÛ·oúv¹Eâàé>üðCY­VIR›6mtèÐ!ýôÓOJKKÓ‘ÿÍ+1b„âââ”êA"… IDATœœ¬àª·f€K-Y²DV«U¡¡¡JJJ2º9€G£?Á(è0¥Úæ Ûb#ƒà ¥¸S$εlzE…tæLeewé\€,]t‘1mk,Ûû¸¢ÂñâÃ…pAf1vìXÝ~ûíúç?ÿ©yóæéÏþ³þóŸÿ(<<\­Zµ’$EFFª  @»víÒUW]ep‹½×˜1cTZZªfõ¹M§èO0 _ë`Jµ¥¸›a9èM«ê€[ûöÒÅWþìÝ{n™§^q–Q¼ß` ,Ð¥—^ªÞ½{«oß¾ºá†´{÷n………Ù· PDD„rss l©÷ëÒ¥‹ºwï®.]ºÝÀãÑŸ`. Á”ê’ânæt&×úÅ/Îý~òdõõ]»6][\­ê{¥¼¼îïÞo0‹Ó§O+::ZÍš5Sbb¢¢££åççW­ œÕjUHm·Z”˜˜¨Ö­[;,KHHPBB‚ËÛ ¾&33S™™™ËN:ûbC Ã”<)Å"qî—”$}ótøpõuááÒ3Ï4y“\æü½®(³¸í¶Û”žž®®]»jܸqºí¶ÛôþûïëñÇ·oSXX¨¢¢"õêÕ«Öc¥§§+66ÖÝMŸäì‚çöíÛgP‹à :LÉì)î*ç©éÖf#}ô‘Ñ­p†茠à ŠŠŠôÙgŸ)**J’ôî»ï*22RÍ›7Wyy¹òóóÕ¶m[ååå)44T;w6¸ÅÞmÆ *..VPPltsF‚QøZS"žÂ:UÜa”uïÞ]»ví’$Y,]yå•ŠŽŽV||¼¶lÙ"IÊÎÎÖ¬Y³ŒlªOÈÍÍÕ×_Í\ÀèO0 #è0%OJq'@Gc0‚O·lÙ2­ZµJŸ~ú©5{ölIRFF†RRR´cÇ•––jÚ´i·ÔûM˜0Áè&^ƒþ£ Ã”ÌžâN€W©ú^ùë_ÏÝBîBþû_çÇšZ÷îÝÞ#8$$Dééé´ÏE€Sò¤wŠÄ¡1ª¾Wî½·ñÇ€çâkL©¶´_3¤¸S$®ÒØ©mÚHÜ¢À;0‚SªZôª¬ÌyÀNŠ;¼Á´iRl¬ôãõß×ÏO>\ v}»xžyóæÉjµ*44Ôé´uG‚QÐaJUôóGÐÍ–âN€ŽÆ ”ÆŒ1º¼Á°aÃtöìYÝÀãÑŸ`t˜RÕ÷üyèfHqg:Àlúôéct¯A‚Q#`J秸Wå )îÌAP_è0%³§¸_¨H#èê‹w˜’'¥¸ Ì`Û¶m*))Q`` â{‹ÀÇÑŸ`t˜’§§¸ šZNNŽ½ê4Ð8ô'…¦döwŠÄÌfòäÉF7ðô'…0ÂJKKõØciúôéš:uªÊÎ2>OVV–† â°¬¨¨Haaaòóó“ŸŸŸV®\éÎ&»]m#èžâN‘8õE€n©©©ŠŠŠRjjª:vì¨3fÔ¸í?ü Y³fÉr^øÚk¯iÉ’%Z·nÖ­[§Ñ£G»»ÙnUÛt3¤¸S$€«‘ân°'N(--MGŽ‘$1BqqqJNNVpppµíŸþy=ðÀzçwìËÊÊÊ´víZeffꢋ.j²¶»“'¥¸ pƒíܹSááájÕª•$)22RÚµkWµm_yå?^-Z´pX¾fÍ­]»V]»vÕ³Ï>«‚‚‚&i»;ybŠ;sÐFZ¼x±æÍ›§Å‹ÝÀãÑŸ`FÐ ¶{÷n………Ù(""B¹¹¹ºêª«ìË8 «Õª~ýúiçÎÇ9r¤¬V«6nܨ‰'ꫯ¾ÒòåË›ì9¸ƒÙS܉c`¤>}ú¨¸¸XAAAF7ðxô'…Ý`‹EþU‡‹%Y­V…„„Ø—••)##Csçέñ8ÁÁÁ>|¸¶nݪ˜˜;vLááánk·»yzŠ;EâMíꫯ6º €× ?Á(è8p ’““í UTT¤^½zÙ—}úé§Z´h‘–.]*I*))QII‰ÂÂÂtâÄ ‡ãµoß^C‡UNNŽFUãyÕºuk‡e JHHpÅÓj4³§¸S$@m233•™™é°ìäÉ“µx tƒÅÄĨ¼¼\ùùùjÛ¶­òòòªÎ;Û·éß¿¿öìÙcüî»ïjùòåzûí·ÓÏÏO}ûö­õ¼éééŠuÍ“pOJq'@p>g<·oß®¸¸8ƒZ<a„Áüýý¯-[¶H’²³³5kÖ,IÒܹsµoß>©S§NöŸ‹/¾XÍ›7W§N$I«V­ÒÞ½{%UŽ¶wìØQ‘‘‘Æ|¸é€Ï`<ÁJx$Nqào(qà‘ŽvŠû/¿çw¾„SÜá‘Ž6ƒÎ)î4oÏž=:räˆtÆg˜Žx5ÆL¡@‡G:Ú5èœâ@ó²³³e³ÙdµZ•ššj:àÕO0…‰SÜ81wÜq‡Ô†O°SÆx‚)èðHœâÀ‰ 7ðŒ'˜B‰Ä)îü 3èðH¿œAŸ5Kúç?¿¯«û¹SÜø tx¤ví~þþ믿~-(È}yðt999:|ø°BBB”””d:àÕO0…“„á‘’’¤3Ï<úãV«t÷ÝîË€§³Ûíª­­•Ýn7ðzŒ'˜Â :<Òé§K[·J?þØüã¡¡Rp°{3àÉFe:B‹+,,üÍ}êþwý[]]òóóeµZÕÚÑàã|q<Á;P Ãcµi#uîl:p?«$iìرÇýŒýû÷+>>^’TTTD‘À+Q ÀÃDI*’d;Ž}%í—Ô]Rº¤±²ÙŽçyày(Ð|€Íf“ÃáÅb‘Õj5§ï xà/þÓJYào|o<Á[P ø€ Èf³Éjµ*55ÕtÀ«1ž` :>ªªªJaaa¦cÀMF¥úúzµmËŸwÀ©b<ÁÞqø¨»ï¾[sçÎU·nÝLGôéÓÇtÀg0ž` ÷AÀGY­VM›6MãÆÓgŸ}f:ø Ì à£æÍ›§¶mÛª¼¼\‹-Ò?þñ]uÕUºõÖ[Õ®];ÓñÀ¯0ƒüOVV–鋾iýrtôgøéÚÉN:iРA:xð î¿ÿ~ÝtÓMúË_þ¢Nˆ–”››«>ø@¹¹¹¦£^ñS(Ð=@}}½žxâ M›6M“&MÒ‘#GŽ¹vv¶† æÒ¶zõj=öØcš0a‚>ýôÓÖŒë³((ŽŽ¾iýrtôgXºt©V­Z¥«®ºJC‡U¯^½ôÍ7ßèí·ßÖ< 'žxBŸ|ò‰é˜h!%%%Ú¶m›JJJLG¼ã ¦pŠ»HKKSTT”ƧÌÌLMŸ>]iiiÍî»gÏÍœ9ÓeÁŸ}ûöéñÇW~~¾ìv»ú÷ﯵkתS§Nîú(99YÁÁÁºûpáBõîÝÛùXïÞ½uõÕWë‘GQ^^žÁ”h))))¦#>ƒñS˜A7ìСCš3gŽn¹åIRbb¢^xáUWW7»ÿ³Ï>«x@‡ÃÙöÌ3Ïhøðá²X, V||¼^|ñE·äx®„„ëÅ_t)Îb·ÛÕ¡CÉ@s(Ð +((Pxx¸:vì(IŠˆˆPEE…6oÞÜdßW^yEcÇŽUHHˆKûÆãÜîÝ»7§¹”­ˆˆˆ£>ž’’¢>úȉÀ±pŠ»a[¶lQ—.]œÛêÞ½»JJJ4`ÀgûÖ­[e³ÙÔ¿ó5zöì©7ß|³ÙãÕÔÔH’ [òÇð åååÊÏÏ7Ã#Ñ7Í£_ŽŽ¾iê§ß»‡vëqï¿ÿ~Íš5K;wvã¿^Çx tÃ,‹\Úl6›ÂÂÂœÛGŽQff¦fÏž}\¯QQQáòü_Ú±c‡$iìر§Ý'ÅÇÇ›Žà±è›æÑ/GGß4oçÎ8VŸøzÕÖÖjÉ’%š:uªRRR4iÒ$çi¯þÜ/Rcß|øá‡Z¼x±&Ož¬‡~X_}õ•$úæ'Ï>û¬JJJœã‰~i|ßìÚµK«V­ÒªU«´lÙ2Ifû¦®®No¼ñ†fÍš¥éÓ§kúôézòÉ'uÏ=÷´èqàÖ­[§ÜÜ\­[·ÎtÀë1ž`Œ~áàÁƒŽGyy¹Ãáp8¶oßîèܹ³£ªªÊp2÷±Ûí‹Åâ())q8Çîè¯/¾øÂa³ÙœÛ·Þz«ã–[nq:tȯûÅáp8jjj Î혘ÇçŸî÷|òÉ'ŽE‹9úôéãøøãé—ÿY°`cÁ‚.m¦ûæ÷¿ÿ½#,,Ì1`ÀÇСCC‡u\~ù厮]»¶Ø1ZB^^žC’#//Ït´‚Ÿþ}¥<‡äh…¯žÿ{ýžÿ;ï%‡ãdûþó7üþõ<Ì û‰‚‚…‡‡«cÇŽ’¤ˆˆUTThóæ͆“¹Ï¯o“q¬>ñ‡þêß¿¿Ú·oïÜ>÷ÜsÕ§OmڴɯûE’‚ƒƒ3Ã?þø£.»ì2 0Àïß3Rã­ ßzë-ÝyçÎ6ú¥Ñüùó5nÜ8]yå•Z»v­$ó}óõ×_«¨¨HŸþ¹>úè#}ôÑGúøãõÏþ³ÅŽZºŸØ²e‹ºtéâÜ T÷îÝURRb0•YGë“;wúe­X±B÷Üsýò «V­ÒUW]¥   Õ××Ó7’þþ÷¿kòäÉ.môK£Ï?ÿ\;wîÔÙgŸ­+®¸BÆûæöÛo×7ß|Ó¤=**ªÅŽZ‹Äù ‹ÅÒdá!›Íæ׫½­OÚ·o¯~øÁ¯ú+//O£GVTT”>üðCúåâââôàƒê‰'žP§NÔ«W/¿î›÷Þ{Ot¹‡¶Ãá`,ýBïÞ½5þ|…††êÕW_ULLŒÑ¾‰ŠŠRzzº¾ÿþ{g[}}½rrrôÖ[oµØq@Ë`ÝO 4HÛ¶msnWVVªªªJ±±±S™u¬>¹ä’Kü¦¿ìv»^ýu=ñÄ’¤„„úåN;í4Ýyçš3gŽV®\é÷ï™ôôt;V;wVçÎõÝwß)))I¥¥¥~Ý/͹÷Þ{µfÍãï™—_~Yk֬є)Sœ+¹O™2EË—/o±cÀsÌŸ?_éééš?¾é(€×c<ÁfÐýÄù矯††•••©[·n*--•ÕjuÞnÌ«O"""ü¦¿fÏž­É“'«M›ÆÏëÎ;ï<úåWÎ;ï< 0Àïß3K–,Qmm­¤Æ™ó!C†hÎœ9ºêª«4{öl¿í—£éׯŸñ÷ÌC=¤›nºIAAA.í¯½öZ‹ž#..Nµµµ 6ðzŒ'˜Â ºŸÐÈ‘#µ~ýzIRNNŽfÎœi8•{ÕÕÕIj<½S:vŸøK=ùä“8p *++µuëV½úê«Ú¹s§ß÷Kmm­òóóÛK—.Õm·Ý¦6mÚøußœvÚiêÝ»·z÷î­ÈÈHµmÛV=zôP—.]üº_$éàÁƒZ¼x±êëëe³Ùô /hêÔ©Æß3·Ýv›‡¾øâ IÒÆåp8tûí··èqà4tèP%$$˜Žx=ÆL±8‡épªª*M™2EÝ»wW}}½žzê)ӑܦªªJ¯¾úª}ôQMžªsÎ9§Eu*òó󯼼<õë×Ït´°Ÿþ}¥[C† Q§N4iÒ$…††Ên·kÆŒMn+‡–•““#›Í&«ÕªÔÔTÓq¯Æx‚)èø¨Ë/¿\=ö˜¾þúk=÷ÜsÚ»w¯®¾új=øàƒ-zœ×_][·nÕ´iÓœmO>ù¤¢¢¢4nÜ8effjúôéMÖ½@ËjéWÀŸ1ž` :>ªsçΚ7ožªªª´fÍ 8P]»vmÑc|ûí·ÊÊÊRNNŽ³íСCš3gŽöîÝ+IJLLT||¼&Nœ¨ÐÐÐ=>~Æí €–Ãx‚)\ƒ€ Sbb¢ºvíªêêj-^¼¸Å^ûÞ{ïÕ€”––¦qãÆi×®]*((Pxx¸:vì(IŠˆˆPEE…6oÞÜbÇÀQ à£Î<óL—¯Þ½{ëŒ3ÎмyóZäõËË˵nÝ:]ýõúË_þ"‡Ã¡øøx¨K—.ÎýÕ½{w•””´ÈqðUœâ€ŠŠŠRrr²,‹³íwÞÑïÿûyý 6Èjµêâ‹/–$Íš5K¯¾úªÚ¶mÛdA8›Í¦°°°c¾Þ„ Ô©S'—¶ääd%''·H^_—••¥êêj…††ÒgÀ)òÅñ”••¥¬¬,—¶òòrCip4èø¨¹sç*::Ú¥íÒK/Õ¢E‹Zäõkkk]¶»víªÈÈHÕÔÔhÛ¶mÎöÊÊJUUU)66ö˜¯—‘‘Á}xOAxx¸>¬ÓQ¯ç‹ã©¹<º:<:>ê׏$é•W^Ñßþö·S~ý .¸@?þø£8 ððpI +]uÕUš:uªÊÊÊÔ­[7•––Êjµ*22ò”‰£>|¸é€Ï`<Á t|Ô°aÃ\¶kjjôå—_6i?Y½zõÒm·Ý¦ hÒ¤IÚ½{·Ú¶m«ØØX9RëׯWRR’rrr4sæÌ9&¾ŒUUU¥ÄÄDç5èAAAúÓŸþ¤›o¾¹ÅŽñ /hÖ¬Yš={¶¶nݪwß}W’”™™©)S¦hãƪ¯¯×SO=ÕbÇÀWQ à£^|ñE 0à˜ûìÞ½[={ö<éctêÔI³fÍjÒ¦ŒŒŒ“~]œ¸={öèÈ‘# Ðgœa:àÕO0…U]]­5kÖõñúúzeggkþüùnL…Ö’-›Í&«ÕªÔÔTÓq¯Æx‚)èø¨3fhõêÕÇÜÇb±P ûˆ;î¸C jÓ¦é(€×c<Á t|Thh¨Š‹‹uöÙg;ÛÞxã ÅÇÇëœsÎÑ‘#G4wî\ƒ Ñ’~ZIÀ©c<Á>ÀGEGG»ç’” »îºK’ ?þñ&¢€fP à£víÚ¥šš—¶Å‹kÿþýÎíàà`wÇGÁ)îø¨I“&é‚ .ЕW^©3ÏýôS½ø⋦£¡äääèðáà QRR’é8€Wc<ÁfÐðQ^x¡ÞyçUVV*++K‘‘‘Z±b…î»ï>ÓÑÐ ìv»jkke·ÛMG¼ã ¦0ƒ€‹ŠŠÒË/¿¬ÐÐPUWW+44Ôt$´’Q£F™Žø ÆLaUXX¨k¯½V·Ür‹$©  @3fÌPyy¹ád 9èø¨¿ÿýïêÓ§¢££%ITïÞ½5gÎÃÉ@s(ÐðQáááš7ožz÷îíl³ÛíZ¼x±ÁTh-6›M²Ùl¦£^ñS¸•à²][[«ôôt]uÕU†¡5-X°@6›MV«U©©©¦ã^ñS(ÐðQqqqºãŽ;TYY©¼¼<­X±BmÛ¶Õ£>j:ZÁ¨Q£T__¯¶mùó8UŒ'˜Â;Õ·o_-^¼X¹¹¹zÿý÷õÚk¯éÚk¯U`` éhh}úô1ðŒ'˜B€JIIÑèÑ£5bÄ 6Ìtð(ÐýÌôÁ¨OŸ> 1üÆáǵsçN]sÍ5 wË1¿úê+=òÈ#MÚ÷íÛ§=z¸%8~è~æƒ>ÐرcMÇ¿µdÉÝvÛmn9ÖsÏ=§÷ß_]ºtQ›67n©©©QFF†233Ý’î“››«ÚÚZkèС¦ã^ñS(ÐýÌ™gž)©ñĘ˜Ãi|Ó„ ”‘‘a:†Ï¢[ýÛz 5vìXçïaw¸ï¾û´cÇM:Õ¥Ýb±P û ’’UUU),,ÌtÀë1ž` ºŸi×®$)&&Fýúõ3œÆ7uêÔ‰¾mEôoë¢[ßO¿‡Ýáž{îÑ!CÔ³gOH’‡æÍ›ç¶ pŸ””ÓŸÁx‚)èø.]ºè¡‡Ò”)S4~üx………9‹óŸüzFx t|È%—\¢´´4IÒ3Ï<£/¿üR_|±n¿ývçmƒ¬V«Á„àh(Ðð!çœsŽóû©S§ê /Ô²eËœ‹ÄÏE´°äädÓ|ýÛºè_ïèüÞb±(>>¾IqþÕW_颋.rw4´²ôôtÙl6Y­V¥¦¦šŽx5ÆL¡@ZNë¢[ýëý:¤¢¢"I ÂýøãÎm©ñ~ìÏ?ÿ¼-Zd*"ZÉðáÃUWWçò! €“Ãx‚)èøE‹5)¾—/_î²m±X(Ð}P\\œé€Ï`<Á t|¾=@2 IDATÈ7Þ¨ûî»OAAAÍ>^UU¥_|ÑÍ©Àñ @À‡Lš4é7ïeî¦4àDP àC~«8—oÅß“——'»Ý®   ÅÇÇ›Žx5ÆL¡@ðëÖ­s®:MAœÆL¡@÷ï½÷žþú׿ê_ÿú—"##›ÝgõêÕz÷ÝwuäÈÝtÓMÏ<óŒ†.‹Å¢àà`ÅÇdzB/x t/xÌÇ7nܨ˜˜çvïÞ½õ駟¶v,@ âw°eËuéÒŹݳgO½ùæ›w›?¾*++Õ¾}{7ÎtÀ«1ž` º°X, pnWTT(,,ì˜Ï™0aB“kÔ“““•œœÜ*SõÑGéw¿û¤Æû`РAÚºu«s»´´TçwÞ1Ÿ“‘‘q\÷Êœ¸æ>ðÌÏÏçV=hU ¦#>ƒñS¸Ý 4HEEEÎíÒÒRÅÅÅL8Qè^ ®®N’T__ïl›={¶³(1b„6mÚ$Iª®®ÖöíÛuÿý÷»?(à¤Q {¸ªª*½ôÒK²X,Z´h‘8 IÊÎÎVAA$)&&F<òˆ¦Nª3fè¹çžûÍkÐÝé‹/¾hÒöå—_êæ›oÖ½÷Þë– µµµÊÌÌÔ™gž©ï¾ûî7÷¯®®v9+OWTT¤Í›7óÿ/ 0ž` × {¸°°0=üðÃzøá‡]Ú7lØà²=zôh=ÚÑ~“ÍfÓÔ©Su÷ÝwK’­)S¦èý÷ßײeËtçwº%ËêÕ«µtéR•””8Ûþö·¿éÍ7ßÔ_|!‹Åâ²hh¨Ö­[§üü|3Æ-8999²Ùl²Z­JMM5ðjŒ'˜BŽVóÇ?þQS¦LQll¬$iÍš5úþûïõüóÏkÒ¤IzþùçÝ–åºë®Ówß}§µk×:Û>ùäåççëСCêÚµk“礤¤húôézýõ×uÛm·¹-+'ãÁ4ðŒ'˜Â)îhééé tç’©³Ï>۹ݮ];·f rÙîÝ»·:vìØlqþ“¿üå/zñÅUQQÑÚñ8%ÁÁÁÎ/§†ñS(ÐÑâ*++•––¦””—ö¨¨(ÅÄÄ8·}ZykkÓÆõí­sÏ=÷˜Ï Òe—]¦'Ÿ|²5£:ŽßÅ_¬6mÚ¨M›6ºè¢‹œíþóŸÕ¶m[EDDH’²²²¤aƹ}ú¨¸¸X’´ÿ~3FÔÝwß­-[¶4ÉóË ŽæÊ+¯Tff¦öìÙsÔ}‡233uë­·***J‰‰‰ÎEü~b³ÙôÀè²Ë.Ó€4cÆ UWW»ì³páB3FçœsŽ®¸â ½÷Þ{ÎÇ 4aÂuëÖMû÷ï×UW]¥˜˜åçç³x t· 6hØ°a²X,ú補íÏ>û¬´qãFIÒ’%KÙäùçwžKAìp8ôù矫ººZ‹-ÒK/½¤iӦ髯¾rîóÇ?þQm۶ջﾫ¹sç*33S/½ô’$Én·k÷îÝÚ½{·^~ùeÝ~ûíºè¢‹d±XôÃ?hðàÁêÛ·¯Ö­[§Y³fiéÒ¥.³öçw^“ šsÎ9çèÈ‘#ÊËË;ê>K—.Õ¬Y³ôÚk¯é믿ÖöíÛ]À³ÛíëÜgÁ‚úç?ÿ©7ÞxCùùù8p FŒ¡7ß|ÓÙ_………:xð ^yåÝ{ｊŽŽVMMÍQûà?²²²ôꫯ*++ËtÀë1ž` :Ž[@@€¦M›&‡Ã¡Õ«W;Û7oÞ¬aÆ©K—.’¤o¿ýV=zôhòü±cǪÿþÎm‹Å¢K.¹Ä9Ó~ÅWÈáp8r«ªªÒÊ•+5hÐ çãÁÁÁÚ¶m›¤ÆÂyðàÁ’¤n¸Aøô|ùrõíÛW³fÍÒöíÛõè£* @Ý»w×Ýwß-‡Ãá<~DD„üñßü¹ÃÃÃ%ÉùDs–/_®þýû+ @¡¡¡ºôÒKµuëVçã/¼ð‚¾ùægž¾}ûê†np¾æ®]»ôàƒjòäÉjÓ¦ÂÂÂ4}útjîܹ’¤¸¸8 0@’të­·j̘1úÏþ£K/½ô¨ýðáááêÖ­›óÿ[Nã ¦°Š;NÈå—_®óÏ?_™™™ÎÛº½úê«zì±Ç$I555*++Ó駟~¯*©ñTp©ñsyyyŠ‹‹Smm­þóŸÿ(88ØeÁ¶¶mßÂgu–Ëk-^¼Xýû÷wY®¹Yýã¦cèéééÎï¿üòK}÷ÝwΟC’V­Z¥!C†¨C‡ζŸfÆ%)77Wõõõ.g„„„è²Ë.Ó‡~¨}ûö©GΟ·OŸ>.Ç?Z?üÇðáÃMG|ã ¦0ƒŽöàƒ*77W………ª®®Ö¡C‡Ô³gOI Ä9žðëþtJv}}½³-&&FO?ý´n»í6uèÐA;wv™oNyy¹öïß\×—¯N:i÷îÝG}üŒ3ÎЦM›tà 7hõêպ袋\r–””¨¶¶ö¨ÏÿòË/%I¥¥¥.íW_}µšÜ÷ÞŸüt)Daa¡³mÏž=úüóÏU^^n0в(ÐqÂÆŽ«Ž;êÅ_Tvv¶n½õVçcááá WYYÙ)Çf³)11QåååZ¶l™®½öÚãz^‡Ô£G—{žŸª~øÁ¹^s^xá=öØcš7ožüqY­V—ÇO?ýt}ñŪªªri/--UCCƒ† "©ñÞì¿Ô«W/IR|||Kü^éÈ‘#ºä’K\îGºxñb]rÉ%Z·nÁd@Ë¢@Ç SJJŠ^{í5½ûúê«]¿à‚ tàÀS>οþõ/åææjÒ¤IÎ6‡Ã¡†††c>¯M›6ºñƵcÇíܹ³Éã?­ÿk555.³÷?ùñÇUSSsÔSä>¬Ç\wÞyg³×ÞKÒàÁƒU__¯wÞyÇ¥=''Gmڴѵ×^«ÐÐÐ&úŽ;sR— üËž={´k×®cÞuÀña<Á tœ”x@•••JHHhòX\\\“Sµ›SQQ¡ÊÊJç¶Ýnw¶K§KÒÚµkåp8´|ùrýðꪪҮ]»TWWç|ί—œœ,Iš0a‚vïÞ­òòrýßÿýŸ$iûöíª««sÙ¿²²R}úôѹ瞫ƒº<¶oß>IÒ¥—^ÚìϤððp}ú駒¤ââb­Y³ÆùºÛ·o×c=¦îÝ»kܸqZ¸p¡V¯^­;ï¼Ó¹¸[hh¨î»ï>­X±B+W®”$mÙ²EÙÙÙÊÎÎnÒG¿>CáhýðÙÙÙZ¸p¡Ëÿ7œÆL¡@ÇI‰ŽŽÖˆ#t×]w5yìúë¯Wqq±Ë"i¿TUU¥©S§ê‡~Pnn®æÏŸ¯ï¿ÿ^>ú¨$éí·ßÖ{ｧk®¹F÷ÝwŸî¿ÿ~Ý|óͪ¯¯×ˆ#tðàA•––ê£>Ò¼yód±X4yòd}öÙgÎc\zé¥ÊÊÊRqq±¨GyÄY€ÿ÷¿ÿÕÞ½{]2Y,µk×N;vìÐÛo¿íòØ×_­=zhĈÍþ<JOOWaa¡† ¢%K–è®»îRÏž=•““£îÝ»«cÇŽúì³Ï4xð`ýùÏÖ‚ t×]wéÜsÏu¾ÎsÏ=§éÓ§ëñÇ×å—_®§Ÿ~Z/¾ø¢âââ$5ÞÊ-;;[‹EãÇwÞ×ýÿþïÿŽÚÿqÇwèOú“î¸ãÓQ¯Çx‚)Ço­¸Ÿ’ŸŸ¯øøxååå©_¿~­vœ1cÆh̘1ºá†Zí­!++K§vš~÷»ß9Û’’’Ô¿ýõ¯5ÌÕ××uÑÁwß}W‰‰‰nNœwýþõFôoûéßWÊ“Ôÿ¾’vKê)i…$ÞKÒÉö{¾è?ÿÂï_Ïà :ZÅsÏ=§—_~ÙtŒb·Ûµ~ýz—â|ß¾}Ú¿¿¦L™b0@ŽVqúé§ëÖ[oÕK/½d:Êq9xð Þxã =óÌ3ζšš=ñÄÊÎÎ>©ÛÆÀ‰hk:|×í·ß®Ï?ÿ\Ë—/÷øSÝ»víª;ï¼Ó¹]WW§çŸ^Ï<óŒs±:¬%%%™Žx5ÆL¡@G«8p`³·.ótš8q¢é7»Ý®ÚÚZ˜Žx=ÆL¡@G«kÛÖûÞfmÚpõÀ»Œ5Êt¿s¬»ÖÕjUTTT+%BKa<Á﫜Ê‹‹}RÏ-**¢HÐ, tàý«PÒØžuà?(Ð|€Íf“ÃáÅb‘Õj5ÇĨuîï“O0…€ÇzóÍ7µ{÷î£>þöÛokûöíºï¾ûäÆdày,X ›Í&«ÕªÔÔTÓq¯Æx‚)è<Öœ9s´nݺ£>þÒK/I’RRR(Ðø½Q£F©¾¾Þ+g< ã ¦ðŽð}úô1ðŒ'˜Â½¤ðÌ {¸úúzMš4I¡¡¡²Ûíš1c†šì·hÑ"ýøãjhhP»víôÀH `^aaá íϽÉx t—––¦¨¨(7N™™™š>}ºÒÒÒ\ö),,ÔªU«ôúë¯K’nºé&ÅÇÇë’K.1››«ÚÚZkèС¦ãÒ¸ÚöرcOø™Ü›¿Äx‚)èìСCš3gŽöîÝ+IJLLT||¼&Nœ¨ÐÐPç~~ø¡*++Û;wÖŽ;(Ðð#%%%ªªªRXX˜é(EI*’t"÷çÞähŠñS(Ð=XAAÂÃÃÕ±cGIRDD„***´yóf 0À¹ßµ×^«I“&iÎœ9JLLTMMFe*60 %%ÅtÁ,8Nã ¦P {°-[¶¨K—.ÎíÀÀ@uïÞ]%%%.úÙgŸ­ÿüç?úýﯥK—ê³Ï>S›6¬ÿp»Ý®èÿø‡®¸âŠã^CüŒ݃Y,–&ÌØl¶fOµ9tè{ì1=÷Üsºë®»´páÂcþ!4aÂuêÔÉ¥-99YÉÉÉ-üXVV–²²²\ÚÊËË ¥qgŸ}V%%%²X,’Žo àŠ݃ 4H'NtnWVVªªªJ±±±.û­[·NË—/×k¯½¦+¯¼RIII6l˜î¼óΣ¾vFF†úõë×ZÑÀ¯5÷g~~¾âãã %j]Ÿ~ú©"""Ô¹sgIÇ¿† pÅyÐìüóÏWCCƒÊÊÊ$I¥¥¥²Z­ŠŒŒtÙoÅŠÎ?ú†ª™3gjåÊ•nÏ ð?•••zë­·\>>Ö*h=éééJKKSzzºé(€×c<Á t ‘#Gjýúõ’¤œœÍœ9S’4{ölI’.¿ür}úé§ÎçiäÈ‘î ð;ÿûß5yòd—¶c­¡‚Ö3|øp1BÇ7ðzŒ'˜Â)î.33SS¦LÑÆU__¯§žzJ’”­³Î:KÑÑѺöÚkuèÐ!=ÿüóêÚµ«ÚµkGhuï½÷ž¨®]»:ÛÇ ­¡òK¬rjâââLG|†/Ž'\ÅQ {¸°°0edd4iß°aƒó{‹Å¢Ûn»Í±€VµtéR½ûî»Úºuëqíÿý÷묳ÎÒÿûÿ¯•“ø¥ôôt—ÿUTT())IO<ñ„¶mÛæl?Ú*¿Æú(Ðzüm}oÅ)¾úJááኈˆh…tàÙŠŠŠTWW§ÀÀÀSZ¡ã æP ðk×®Õˆ#Nê¹ß~û­.¾øb=ôÐCúÇ?þÑÂÉÀóåääÈf³Éjµ*55ÕtÀ«1ž` :€xðÁMG|ã ¦P øn´ÆLa‘8<:€SÜ|@VV–ª««ªäädÓq¯Æx‚)è> <<\‡VHHˆé(€×c<Á t0|øpÓŸÁx‚)\ƒ€ @ÀpŠ;€سgŽ9¢€€qƦã^ñS(Ð|@vv¶l6›¬V«RSSMǼ㠦P {°úúzMš4I¡¡¡²Ûíš1c†šÝ·¡¡AóçÏW§N¡!C†¸9-0éŽ;îPCCƒÚ´á FàT1ž` ï8–––¦¨¨(¥¥¥©W¯^š>}z³ûUTTè–[nÑÕW_­1cÆPœÃëTVVjØ°aš`Ô¨Qª¯¯WÛ¶üyœ*ÆLáà劋‹ëÑHÒ¡C‡Ž¹aaakG¼ZŸ>}LG€Ÿ¢@ðbÅÅÅŠŽŽ>Égsm-x tFåææjß¾}-þº6lP»ví4|øpµk×®Å_<ÅÏ3çK$ÅœÀ3­’¢Z>à¤Q 0êá‡Ö¦M›ZüuŸ}öY=ûì³Ú»w¯zôèÑ⯞æŠ+Ú¨]»2ÕÔë㇚ŽxµÜÜ\ÕÖÖ*88XC‡5~„ÀôésPaa?¨ª*Ll: àÝJJJTUU¥°°0ÓQàg(Ð|ÀâÅ—Jêg:àRRRLG€Ÿjc: @À#P à¸À¤¦®T‡o«¢ÂªôôTÓq¯–žž.›Í&«ÕªÔTÆ܇À¬\£ÀÀÕÕšŽx½áÇ«®®NŒ'¸:€Ø´)B¬â´Œ¸¸8Ó৸@€àwF,X°@ï¿ÿ¾JJJZõ8÷ÜsÎ<óLÍ;·U¦ÅÇ—((¨Vv{òòâMǼZ^^žìv»‚‚‚Ïx‚ûP 0⫯¾Ò[o½ÕêÇy÷ÝwÕ·oßV?˜–°]Vë7²Ù¬èÀ)Z·nsw t¸:€˜;w˜X$hãÇ7~ŠkÐðèx t<× ø€qãÖ¨}û\UV¶×üùãLǼÚüùóUYY©öíÛkÜ8Æ܇ÀlÚÔSÁÁÝT[l: àõâââT[[«à`ÆÜ‹À¬[w¶XÅh ¦#ÀOq :€À)¾^“&MRhh¨ìv»f̘¡€€€£îŸ­—_~Y}ô‘S'¦®®N‡–ÝnwÛ1TQQ¡ÀÀ@…„„¸í¸àNÑÑ¥jÛv³êëUTm:àÕŠŠŠTWW§ÀÀ@EG3žà>Ì {°´´4EEE)--M½zõÒôéÓºïž={4sæLY,7&NÜ›o¾©Ž;ê•W^qÛ1·oß®Ž;êÁtÛ1ÀÝ’’6êæ›—)))ÇtÀëåäähÙ²eÊÉa<Á½˜A÷P‡Òœ9s´wï^IRbb¢âãã5qâD…††6ÙÿÙgŸÕ< ýë_îŽ <ÀܹÃd±\(‡ÃtÀûñ¡>L¡@÷P WÇŽ%Iª¨¨ÐæÍ›5`À—}_yå;V&¢`··•Ä-¡|Qqq±l6Û =§°°°•Òøn¯S(Ð=Ô–-[Ô¥Kçv`` ºwï®’’—}ëÖ­²Ùlêß¿?:€)..>Åk ­-–@ë£@÷P‹¥É‚p6›MaaaÎí#GŽ(33S³gÏ>áן0a‚:uêäÒ–œœ¬äää“ pÊÊÊRVV–K[yy¹¡4¼ÙÏ3çK$Ŝ೭’¢Z6€VEî¡ ¤‰':·+++UUU¥ØØXgÛÿû_½üòËZ´h‘$Én·Ën·«K—.:tèÐ1_?##CýúõkðàçšûÀ3??_ñññ†Á$'®ÐÐ/U]ª¬,>p÷=1’øÛÍ]²²²T]]­ÐÐP&°àVèêüóÏWCCƒÊÊÊÔ­[7•––Êjµ*22Ò¹ÏÀ]®/ú÷¿ÿ­eË–iéÒ¥&"ƒh¯:|˜ÛI§*<<\‡æö¬p; t ‘#GjýúõJJJRNNŽfÎœ)Iš={¶®¿þzEGG«wïÞÎçtíÚUíÚµsi<ÅÞ½{uÁ¨¶¶ÖX†7ÞxCo¿ý¶žþyÝzë­Ær@kX¹2VÌ°-cøðá¦#ÀOQ {°ÌÌLM™2E7nT}}½žzê)IRvv¶Î:ë¬& †X,îƒÕÐРÍP[[ëü< º SFFF“ö 64»JJŠRRRZ; P €“¶zõjýíoÓž={tÿý÷ëᇖÅbQ}}½&Mš¤ÐÐPÙív͘1£ÉÝIвN?½\»täH€öî=ÃtÀ«íÙ³GGŽQ@@€Î8ƒñ÷ic:ðNµµµZ²d‰¦Nª””Mš4I999’¤´´4EEE)--M½zõÒôéÓ §õ}ÉÉ_èÞ{*99ÛtÀëeggkáÂ…ÊÎf<Á½˜A'eÓ¦Mzá…Ô¾}{ :T7nÔ믿®!C†hÎœ9Ú»w¯$)11Qñññš8q¢BCC §ö]‹'¨M›sÕÐÀü pªî¸ã544¨MÆÜ‹œ”þýû»lŸ{ªªÒ¦M›®Ž;J’"""TQQ¡Í›7kÀ€&¢ú…ƒÛKên:àÂÃÃMG€Ÿâ#!­îºë®ó¨Û•L›6M^x¡¾ûî;ÓQŸ²bÅ ÝsÏ=Ú²e‹ºtéâl T÷îÝURRb0-AüB IDATžt­î›o¾ÑŽ;LÇpÚµk—víÚ%»Ýn: à3òòò4zôhEEEéÃ?l² œÍfSXXØ1_c„ êÔ©“K[rr²’““[|X!!!JJòñÔÜžùùùŠ7”ÍáwpJfÏž­É“';S:ï¼óÔÐР²²2IRii©¬V«"##MÆôyAAG\«  ÎN•ÝnWmm-gÛÁí˜A'íÉ'ŸÔW\¡ÊÊJ•——ëã?ÖСC5räH­_¿^IIIÊÉÉÑÌ™3MGõyo½ÕOg Í)..–Íf;îýûôé#«Õª¨¨¨VL4E Õ<úè£Z¶l™óVKžæ²Ë.SÏž=µaÃÓQ¯”––¦™3gºß±±±ºçž{”™™©)S¦hãƪ¯¯×SO=e0)V\\¬èèè“znQQE:ÜŠ@«9tè¾ÿþ{Ó1Žjß¾}²X,¦c^kÚ´iš6mZ³………)##Ã͉ ©ŸgΗHŠ9ÎgJ{B³î@K @ðíÛרM› 54XTYi5ð@1:ÞË@ÇSÕÔÔ´n$àW(Ð|À¸qkÕ¡ÃJUTX•žžj:àÕÇSªÖ®]«K/½Ôtø tðÖ[+ à,9ŸwÀ©jO(33Ótøn³ Å½ù曊‹‹ÓŠ+LGùMeeeŠ‹‹Óã?n: œ’;õm[_íÜÙÇtÀë5Ž§m 7~†X´¸~øA¦c—úúz(66Ötø9fÐðÌ ø€+®øVíÚ•©¦&X<ÔtÀ«5Ž§kôí·ßª_¿ã[ùh è> OŸƒ ûAUUaúøcÓiïÖ8žÎÖÁƒMGŸ¡@Ðb¾ûî;Ýzë­Ú·oŸé('ìÃ?Ô!Côè£êÆo4NØâÅ—êxïñ àØÇÓCÊËË3~†@‹©©©Ñ§Ÿ~j:ÆI9pà€8 1cƘŽ?Å"qx t<§¸ø€ÔÔ•êÐámUTX•žžj:àÕÇÓ_µråJVq‡[1ƒ Eüîw¿ÓþðÓ1NÙ¬Y³tÑEéÛo¿5NÈÊ•1Z±b„V®n: àõÇÓ ÅÄĘŽ?à º‡«¯¯×¤I“*»Ý®3f( ÀeŸÊÊJÍœ9S¯¿þº.¿ürýå/ÑùçŸo(1üÕæÍ›µÿ~Ó1NÙîÝ»µ{÷nÕÔÔ˜Ž'dÓ¦±Š;Ð2ÇS¾"""LGŸaÝÃ¥¥¥)**JiiiêÕ«—¦OŸÞdŸ7ÞxC‘‘‘Z¸p¡¾ÿþ{Ýyçî 8%Ì {°C‡iÎœ9Ú»w¯$)11Qñññš8q¢BCCûõíÛW¿ûÝï$I‘‘‘:çœsôí·ßêœsÎ1’pâ˜A÷` WÇŽ%Iª¨¨ÐæÍ›]öû©8—¤Þ½{Ëjµr:~&>¾D ëŸg: àõÇS‚JJJLGŸ¡@÷`[¶lQ—.]œÛêÞ½û1Q¬X±B£GVXX˜;"?~¼úöí«ƒšŽÒ¢FŒ¡‹.ºÈt 8n Û5th®Ö™Žx½Æñ4TÛ·o7~†SÜ=˜Åbi² œÍf;fñ½jÕ*¥§§ÿækO˜0A:uriKNNVrròÉ……ßÚ·oŸ¶mÛf:F‹Ûµk—ÊËËMÇ€—ÊÊÊRVV–Kï'´¶¹s‡‰E‖Ñ8žâ•—Ç)p/ t6hÐ Mœ8ѹ]YY©ªª*ÅÆÆ6»ÿŠ+ô‡?üÁyJü±dddpOGh%Í}à™ŸŸ¯øøxC‰€7àwvþùç«¡¡Aeee’¤ÒÒRY­VEFF6Ù·  @¥¥¥ºîºëÜÐ(Ð=X@@€FŽ©õë×K’rrr4sæLIÒìÙ³UTT$IúöÛoµ`Á]}õÕÚ¹s§¾üòKÍ›7ÏXnÀ‰ãw—™™©)S¦hãƪ¯¯×SO=%IÊÎÎÖYg%«Õª+¯¼R{öìÑ /¼ ©ñÚõììl“±á,X 'Ÿ|Ò§¯«­¬¬T=tå•Wêõ×_7Žiܸ5jß>W••í5þ8Óq¯Ö8žRµfÍ. …[Q {¸°°0edd4iß°aƒóûï¿ÿÞ‘IRuuµJKKMÇhU‡C¥¥¥>ý!ß±iSOwSmm°é(€×kOÿÑ AƒLGŸ¡@ðëÖ-VqZFãxÊÕœ9sLGŸát<:€À ùꫯ¦'žxÂt·ùàƒÆinWE…Ué驦㭢¸¸X6›í„žSXXxÂÇiO7kãÆJLL<áç'‹ÀÌ;LË…r8L'ZGqq±¢££Oá¬Ç½gãxº\kÖ¬9…ã'ŽÀØím%q‹5ø®ŸgΗHŠ9Ág[%E÷Þã©VmÛR.Á½xÇ8n‡C?žšùåÏo±X §À_ň[ ÂWQ 8níÛ·÷»ëÏiòäÉšùä <Øt'ád®_•$«Õª¨¨ãŸ}àdP ¿pª×¯yt‘žœü¹BC¿Tuu¨²²’MǼZãxºGŸþ¹úõc¶îCüÂÉ_¿Z(iìIͼ»ÓíÒA‡‡˜Žx½ÆñT¦öíÛ›Ž?CþÿÛ»÷ð˜îüàïÉä"—‰\‰HÔ%!I…µM…uí®®Ö¶,úÓ¥ôâѤˆÒÒj£­j)*íZªêV¥¨V\V]âA‘ˆlh BLî1™óû#2I&3ÉLÎ93ï×óäaæÌœy¾ß9“ùdæœCDDD6Æ:÷_Ý»·3¬±."1TnOÛ1wî\±£±;I_QQ4Ø1$ƒãADDDD–ÀOЉŒPZZŠÔÔT€üq‘5­Q£Fa÷îÝbÇŒÁƒC¡P@§Ó‰¥I©ÕjdddÚ·o–-[ŠœˆˆˆˆÈºðt¢z”••!;;ÑÑшŽŽFbb¢Ø‘ˆD±sçNývðÝwß¡¬¬ÌæþHA$e­Zå# ­Zý.v"Ù«ÜžŸŸ/v²1lЉꕕ…“'O"00:u;‘¤¼üòËhÖ¬¶lÙ‚3gΈ‡ˆŒ{'®ÁرÄŽB${•ÛÓD?~\ì(dcøw¢:ÄÇÇcýúõbÇ ’´Ñ£GÃÛÛ·oß; ‘Å]¸pÁ¤Û7õ¹Ó““#ag× :?!j¬Êíi46mÚ$v²1lЉLtüøq¼ôÒK;v,úõë'v‹*((@AAÊÊÊÄŽ"9‚ à÷߇££#|||ÄŽcQ7oÞÄܹsñÛo¿‰…H$*À¸qãL¾gSž;=/Ï @‹&y,"kW¹=ñ4kÔôØ ™(##èÖ­›Õ7èü1ÄŽ!YþþþèׯRRRÄŽbQŸþ¹Ø1ˆD àS΃.s§‘´°A'zDzz:8€Ë—/‹…HJKK±|ùr<öØcˆ;‘…4ÝWÕ‰ˆÈv±A'zÄÁƒ1uêT±cÉFQQ¦NŠádzA'Q\Ü8;g¢¤Ä;vĉ‡HÖ*·§18sæ BCCÅŽC6„G!j 3f@¥RáüùóbG!²˜àà`tïÞ]ìDdGÇ 89•Áѱ\ì(D²W¹=9¡¢¢Bì(dcø :Q•••Yíy ‹‹‹‘““Ã#s¡¸¸/^„——|}}ÅŽcvEEE(..;aóæPü¤È*·§bÆŒbG!ÃOЉ¸vížþy¬]»Vì(¢;qâ:uê„¥K—ŠEòªÆê£>;ŠèRSSñüóÏc×®]bG!"""’%6èDäççã믿ƱcÇLºß°{÷n ¥"ÇÉ“'ñÃ?˜tŠ½œœ|ýõ×&Ÿ+šˆˆˆˆ*ñ+î§Õj1{öl¸¸¸ ¼¼ ,€R©¬q»}ûöaçΨ¨¨À˜1c%BZÛ4uêTxxxàîÝ»bGi4A““µZ-vÙ¹wïþ÷¿ÿÁ××ÎÎÎbÇi´Å‹cãÆbÇ "¸¹•ÂÎNNÂB•Øqˆd­r{rGii©ØQÈÆðt‰KLLDpp0ѦMÌŸ?¿Æmnܸ×^{ K–,ÁâÅ‹1eÊäçç‹–¬AÛ¶mñÌ3ψCv–-[†¶mÛâàÁƒbG!"5iÒAÌœ™„I“¾; ‘ìUnO3ù{š?A—°;wîàƒ>@nn. 66aaa˜5k\\\ô·[¼x1 …B'''„……áÓO?E||¼XÑegÏž=ÈÈÈ;‘U¸pávî܉¾}ûÂÃÃCì8D6cóæžP*Û¡¢‚or{ú?,_¾\ì(dcø º„¥§§ÃÇÇÍ›7@£ÑàܹsÕn—––†ýåÀÀ@>|¸I³ÊÝ_ÿúW›?÷ùíÛ·õ ¢†ËËËCnn.A;ŠhV­Z…!C†àòåËbG!²)W¯ú 3³®^ ; ‘ìUnO™ðññ; Ù6èvþüyxyyé/;88 E‹ÈÎÎ6x;ÿ·!ªÏðáÃáïï/v Ù{öÙgѺukî³FDDDD&ãw $L¡PÔ8 \AA\]] ÞN£ÑԸͣ,‡RY;;;TTè TÚA§ P€‚ {°¬J¥:ŠÊ…ÁlËììì °³Sè³4u¦¢¢ß`o/ ¢¢BŸ©jl+s*¡ÓÕ¾L©TâÞ½ 4kv öööú„™sœ ;õÓyk[·á1,,Ü {{-´Z-”J{ètuܪ]jèÞ¹cæX²²þÁÏÃxðN²¬˜˜‹hÖìJK’Ò_ì8D²V¹= ÆÅ‹*v²!lÐ%¬k×®Ðét¸uë|}}¡V«¡R©Ð¶mÛj·‹ˆˆÀ¥K—ô—Õj5zöìYÏÚgx¢–ë¯5:·ÜxzzÁÙ¹™þSæ†*))Á;yæ &___8::‰Cö~ÿýº¬÷Awuu…‡‡g£Ö!:‚`ÃÇ5xòÁÏÃÎàÑ€Ér‚‚òàêzEE®HI; ‘¼UnOí‘—'ï÷v$?lÐ%L©TbĈ8zô(âââ°cÇ,\¸°dÉ 6 ;vÄСC1iÒ$@qq1®\¹‚Ï?ÿ¼žµ?máô–׺uk4oÞ.\hÔzV­ú#GŽltžÍ›waÔ¨Q^1gαcXüú믈ŒŒ;HƒŒ7Ùˆ×úØ¡°°*U›F­ÅÃÃ:tÀ‰'™‡Èú%'÷ÀOúˆÌ¡r{úRSSÅŽB6† ºÄ-_¾ñññHKKƒV«Õ7Q6l@»víбcG„„„`úôéHHH€N§CRRR½û [ƒ=z`„ Ø·o6mÚ„›7ošt///øùùA¥RY(¡ô988@¥Ré÷›'ó±··‡——JJJPRR"vQ( „„„ °°999&ßâĉ ‚››t"""² lÐ%ÎÕÕK—.­qý£oVGÑ£G7U,Q¹¸¸àÀpwwGpp0FŽ‰'N˜Ü ?ÿüóHJJ2[®Ž;bÖ¬YØ·oŸlš‰˜˜ìÝ»WìV©W¯^ÈËËÃûï¿7ÞxCì8FñööÆĉѧO³¬ÏÕÕçÏŸÇ®]»ðôÓ¦kgÉ’%hÞ¼9Ôj5¢££ñÆoàçŸ6K6"""")bƒN²Ò³gOøúú",,¬Úõk׮Ž{÷DýJq×®]±hÑ"̘1C z@@|}}ÅŽaõT*pûömÉŸzÍÏÏ‹-;’’’ðä“OÂÍ­ò`u~~~ðóóCdd$ÊËËqôèQÜ¿_ä”DD gÊîyÝ•ˆä… :ÉÊ—_~‰îݻ׸>$$Z­Ï=÷rssñÓO?Õ¹oooÄÅÅ¡wïÞ–Œ*iNNN úÊ1™nÊ”)˜2e † †ï¿ÿ^ì8¢ð÷÷ÇøñãqòäIœ={¶ÎÛuéÒ¡¡¡ˆ‰‰©õ@— ,PÙ°›ú"[0sæ^¸»F…?œ)vªUånuãÆkð}©iTnOó°wï^Åšt’… & 66¶Æìfoo¯¾ú ?ýô“Á=((k×®µ@Jy „»»»Ø1lŽ¿¿?:v숌Œ èt:±ã4©nݺaíÚµxóÍ7 6èC† ÁâÅ‹ë]ßêÕ«‘••…iÓ¦™3&‘ìí݇ܿï vªS0€K L¼ŸêÁ}©©TnO‰˜;w®ØQÈÆ°A'Isuu…¯¯/ž|òI£ƒ;wî`Ñ¢Exï½÷ª- ¯qy[³iÓ&›þö€X>ûì3€§§'òómó|ØmÚ´ADDÎ;‡‚‚?ÞœöîÝ»w“q§ø‹‹‹Ã•+Wðᇢ  €§À!zàìÙð(îrÀF[*·§“; Ù;±8àæ憱£4dÈdeeaüøñFßÇÁÁžžžxꩧ0wî\é—íß¿_~ù¥’þÁÉÉ *• vvÒÚ¼üýýѳgO¸¸¸ˆŦuëÖ :u;F nnn?ûÃK/½„#GŽ [·núëÞzë-Lž<žžž&=7Ûµk‡¬¬,Iì3oHPPT*Úµk‡:ˆ‡ˆˆˆ$NZ5™mÛ¶¡  @²GðvwwDzeËðâ‹/6x À¼yóðÄOÀÃÃfLX·E‹A£ÑÔº­˜fÎœ‰“'O¢K—.bG±i)))øöÛoÅŽQCvv6Ž;Ö$¥R©ôÛäœ9s0a„¯«oß¾X¶lÂÃÃ͘Ð|æÌ™FƒÌÌL|óÍ7bÇ!"""‰ãWÜmœ‹‹ ^xá\¹r)))bÇPù ´··7¦Nj–õmß¾Ý,ëúuëðí·ßâÆøõ×_´///„„„ uëÖfNG åâ₨¨(äææâÊ•+ ZGÕ×Å£¢¢°bÅ Yåw×®]f[WçÎѹsgœ:u iii(++ƒ f[C#::ÁÁü*+5°°l8:–¡¼Ü©©aõ߈êT¹=E";;›‰£&ÅOÐmœ§§'Ö¬Yƒ)S¦ˆE/==½ÁM‹¬_¿¾ÑëFxx8† ‚-[¶`ûöí8uê&Mš„ÈÈH8::½®~ýúáСCøÛßþÖè\R`Žñ[»vípèÐ!¼öÚk&Ý/44QQQ8uêöïß7búôéèÝ»7zõêe–lrßU«V¡¤¤¤Ú.-bŠŽŽÆš5k-v’Knc‘‘Wпÿ~DF±Øc˜B®¯' c;µÚʼz{‡þýûËú=)É?A'•çOJJ¦M›pøðaQ2 :Ý»w‡§§§(o.ëׯÇرcµŽ÷ßùË_ô—}}}áëë‹+VvìØëׯã“O>Ayy9._¾\cèß¿­§¥“3sŒ¯T`РA8{ö,nܸQc¹¯¯/üýý1hÐ „‡‡cРAhÞ¼yÛ%''£¤¤Ä,Çûø¾üòËÈÎÎÆÇ,Êãûøø >>ÞæFYE«ÕböìÙpqqAyy9,X¥R)v,QYr[¶l¤t8¹¿ž˜f=Û¨ÕVæuÏžìÙ³©©©bG!Ã?þ8üqdggãÔ©S())iò¯‰Ž1/¼ðB“>¦ÔÄÆÆbìرõ~•*..0yòdœ>}ßÿ=nÞ¼‰Î;cáÂ…¸~ý:ÜÝݱgÏž¦ˆM 4tèP :Ï>û¬þ‰Ï?ÿÇGÛ¶mÑ¥KŒ9Ò¨u988૯¾ÂÙ³gñþûï[2¶¤Íœ9ׯ_¥AwvvF«V­ðÊ+¯4ùcKUbb"‚ƒƒ1iÒ$,_¾óçÏGbb¢Ø±ˆˆˆ$‹ :U“””„¤¤$¤¦¦B­VcÈ!{,•J…¸¸8DFFâÿøš5kf±ÇÃÛo¿7nu ;'''´iÓýúõÃsÏ=gÒãôèÑ=zôÐ_NKKCnn.T*•É™Iaaa(.. >“'O6yöööxî¹çpèÐ!lÙ²yyy¸{÷n½÷›1cÂÂÂ,~÷¦ÔªU+h4;v «W¯Fjj*.]ºd±Ç›7o €ˆˆ“v?±vwîÜÁ|€ÜÜ\•€ ìY³x6 ""¢:°A§ZU5 ãÇGNN~þùg³¬÷±Çƒ££#–.]Š–-[Vk,­Mll,JJJŒjлwGšåq«Î·MòñꫯâÕW_5˺¢££qùòe$$$`áÂ…õÞ>&&¦ÚîÖÀÎÎ*• ÄÀ‘••µZ„„¨Õj¤§§›åq:uê„>}ú 66V²G‘Szz:|||ô»e@£ÑàܹsèÝ»·È鈈ˆ¤‰ º)--£ø>>bDZ¸3gÎ &&ùùù€ÂÂB¸»»#++ mÛ¶­v[þn""‡­ýn’6èDDDdvðôôDff&|}}‘™™‰ÐÐPÜ»wOìhDDD’e'v"""²>J¥#FŒÐsÇŽF¸ˆˆÈ–ñt"""²ˆ¢¢"ÄÇÇ£E‹Ðjµ˜3gŽØ‘ˆˆˆ$Gq·R‚ `Æ ˜?~£pj4$$$ÀËË ˆ7j6iÒ$¬Zµ 0lØ0lݺÇÔ\´Z-fÏž ”——cÁ‚P*•bÇ’µwÞy *H~êÔ)Žs#ìÚµ óæÍÃÆõûXO[k'''888 ¬¬ åå娨¨0Xㆠ°bÅ üòË/úëöíÛ‡;w¢¢¢cÆŒATTTSD7‰©sY[EEEhÓ¦~ŸýÍ›7cøðáÏn*ck5Tæ0O­r˜WSž¿:+W®„‡‡ Àú樻V9Ì)`\­)))0`@µëzõê…cÇŽϼZ¬Rvv¶°råJA¡PÔX6nÜ8a÷îÝ‚ Â믿.¬Y³Æ¨eT·7nÓ§OöíÛ'ìÛ·O¸víš~ÇÔ<„+V‚ Ÿ~ú©0gΑÉ[ii©0eÊýs6##CŽsCݼySضm› P(„ììlýõ†ÆÓÆÚ”¯_¿.tíÚU0`€þºÜÜ\¡G‚N§JKK…®]» wïÞµxnS5¶NA„?þXØ´i“~›¬¨¨°hæ†2¶Öºê‘Ëœ Bãk­o™T[ç½{÷„Ñ£G ™™™Õ®·Æ9­«VAÇœ ‚qµ¾úê«Brr²ð믿 GÞ}÷]aÁ‚‚ Èk^­ t+véÒ¥ úéÓ§…–-[ :NAøå—_„àààz—‘añññÂÞ½{k\Ï15¼¼<ÁÙÙYÈÏÏA®\¹"xzz EEE"'“¯/¾øBøâ‹/ª]Çqnœòòòj º¡ñ´…±6µÆéÓ§ Ë—/ú÷ﯿî•W^|^°õÇIDAT^ýuýåñãÇëßójØ Û˜óçÏÃËËK9 mpÕÍÞÞ999¸xñ"îß¿ððpܹs€áñ&ã=:ŽhÑ¢DZŽ;†«W¯¢}ûöˆ‰‰Azz:ÇÙÌêÏ«W¯ÚÄXFF ЫW¯z×áïï/¹12Gqqq(((ÀºuëðÙgŸáÅ_´xî†0¶VCõÈaNóÔ*‡y5¶Î~ø={öÄ°aÃðÚk¯aæÌ™øä“Oj]‡ÜçÔP­r˜S a9‚ÈÈÈ:×!ÕyµF>>Ëݦ¼?ZO^^¼½½e1§@ãj}xî¤>¯ÆÖyâÄ Œ3ˆŠŠÂèÑ£ñÝwßaÚ´iV7§†j¤?§@ÃÞ7mݺ£FªsRWkÄOÐeêí·ßÆ–-[jý1ô×¼ˆˆdffê/«ÕjØÛÛ£cÇŽèÓ§OËl±ã­P(ðâ‹/âÀ 7ïÑq,,,DQQ‘þkXÔ8'NÄ8Îffh<}½µÆ±6æùôßÿþ+V¬€··7<==1uêT:tHÿ©MDD222ô·W«ÕèÒ¥KÓasÔù°Ö­[ãOúŽ9Ò$ùMÑ׈Gë‘Ãœæ©ÕØeb2¶Î²²²j—ÃÂÂô_—¶¶95Tëä:§@Þ¿GEDDDµuÈa^­tÓ§OŸ[§N T*kÝ«–‘iBCCo2^×®]¡Óéôûî«Õj¨T*ý©¬¨ñBCC9Îffhùä$$$ÀÎÎŽãÜU¯¹Z­€áç­-<§ Õ¸dÉ\ºt NNN Ôÿx{{£Y³f P¹ ÑÙ³gÅÅŸrå &Ož,NAu0GÛ·oÇo¿ý òÓö6mÚèY"%ÆÔ Û¶m«³9Ì)`žZå0¯ÆÖ9sæLlݺ€ýû÷ë¿mmsj¨VCó-%ÆÖZeëÖ­9rdµëä2¯ÖH!T–¬Ê­[·ðÑGáÝwßÅÒ¥K1a¸¸¸nܸwÞy~~~puuÅ+¯¼¢¿Ÿ¡eT»³gÏâ™gž‹‹ zöì‰ÄÄDøûûë—sLÍ£¨¨ñññhÑ¢´Z-æÌ™#v$ÙR«ÕˆEqq1"""0kÖ,ý‘Z9Î STT„Õ«WcÆŒxóÍ71}útøøøO[ëºjìÕ«Þ|óÍo“““‘œœŒŸþYݦM›púôièt: >áááMZƒ1[çâÅ‹ñá‡"** }úôÁ¬Y³š¼cSk}õÈaNÆ×*—y5öù»}ûv>|îîî5ÞÏXÓœu×*—9L{]úóŸÿŒüööÕO&—yµ6lЉˆˆˆˆˆˆ$€_q'""""""’6èDDDDDDDÀˆˆˆˆˆˆHØ It""""""" `ƒNDDDDDD$lЉˆˆˆˆˆˆ$€ :‘°A'""""""’6èDDDDDDDÀˆˆˆˆˆˆHØ It""""""" `ƒNDDDDDD$lЉˆˆˆˆˆˆ$€ :‘°A'""""""’6èDDDDDDDÀˆˆˆˆˆˆHØ It""""""" `ƒNDDDDDD$lЉˆˆˆˆˆˆ$€ :‘°A'""""""’6èDDDDDDDÀˆˆˆˆˆˆHØ It""""""" `ƒNDDDDDD$lЉˆˆˆˆˆˆ$€ :‘°A'""""""’6èDDDDDDDÀˆˆˆˆˆˆHØ It""""""" `ƒNDDDDDD$lЉˆˆˆˆˆˆ$€ :‘°A'""""""’6èDDDDDDDÀˆˆˆˆˆˆHØ IÀÿi˜PD<1LIEND®B`‚HDDM-0.2/hddm/examples/t.png0000664000102500010250000011767411726446355016367 0ustar wieckiwiecki00000000000000‰PNG  IHDRèXñÌ%sBIT|dˆ pHYsaa¨?§i IDATxœìÝ{|“õÝÿñwZJ¡%¡0ÅÒ"Rª T¨n¨ÓŠ¹©8í-·0èyp%¹rå“øµÉ;ßÃ娭­­0U˜Ù:–@@ÀèX  `t,€€€Ð°:@@ÀèX  `t,€€€Ð°:@@ÀèX  `t,€€€Ð°:@@ÀèX  `t,€€€Ð°:@@ÀèX  `t,€€€Ð°:@@ÀèX  `t,€€€Ð°:@@ÀèX  `t ¨ªªÒÌ™35þ|edd¨ºººÙýæΫ„„¥¦¦j÷îÝîÛKKK•‘‘¡øøxM™2EyyyU: D¬_¿^?ýéOUPPà¾íïÿ»®¼òJ >^N§Sqqq®J"""\¿ð Ýá\’Î9çõïß_’ïŸm ¡NfêvîÜ©ž={º¯GDD¨OŸ>*((ÐE]Ô`ßââb-[¶LÏ>û¬FÝìñÖ®]«É“'+::ºÙû>¬·ÞzKýû÷W×®]ý÷B-*//×¾}ûtÕUW)66ÖìrüníÚµzþùç%µî³Mâ³ ÌìŸMvD@7™ÃáPxxxƒÛ\.W³Ûétj„ ÊÏÏ×e—]¦ƒ6轤 6héÒ¥^Ÿï­·ÞÒ”)SüS< Õþò—¿è¦›n2» ¿Ú¶m›&Ož¬ÄÄDI­ûl“øl³ãg“]ÐM6bÄ¥§§»¯—––ª¬¬Lƒn²oDD„’““õ /())I›6mÒUW]å¾íÚµš8q¢{HasÎ:ë,IÆÿ„çž{®_‰ý͘1CYYYf—aI¼7Íã}ñŽ÷¦©]»viÊ”)î¿ÃÁ¢²²RkÖ¬ÑâÅ‹Ý·µæ³M ­Ï¦PúÃóµ~øá‡:q℺té¢Ë.»ÌäÊü¯­ÿ]¯¾új:tH½{÷Ö›o¾€ÊüÏNm¸=íÎN¯³=‚õ³ÉÎè&2dˆjjjÜœ‹ŠŠät:•àõ1aaa2dˆ† æ¾-//OEEEºí¶ÛZ|¾.]ºH’Î=÷\ >Ü?/"Hôèу÷Ä Þ›æñ¾xÇ{ã]Ýßá`±xñb=øàƒ «_Ö¦µŸm¡ôÙJÿox¾Ö}ûö¹¼ Æ×ßÖÿ®;wvÿk—÷ÅNm¸=íÎN¯Ó‚í³ÉÎè& פI“´eË?^¹¹¹Z´h‘$ãKϵ×^«AƒiçÎ:ãŒ3tÚi§©¨¨HaaaêÓ§$i÷îÝZµj•ÒÒÒ´oß>=zT[¶lÑwÜaæK‘“'OJ2Îo^çÿý¿ÿ§Ÿÿüç*--Õ±cÇôþûïkäÈ‘:ûì³½~¶!4¥¤¤˜]B‡Û»w¯\.W‹ûÔýuòäImß¾Ýë~N§Ó=…¾ Åvû# [ÀŠ+4gÎíرCUUUš7ož$)''G РAƒ´råJ½úê«JIIQçÎõÜsÏI’ 5zôhjùòå’Œ¹999¦½@p)++ÓÓO?-‡Ã¡gžyF÷ÝwŸþð‡?hÑ¢E ‚÷àÁƒuë­·JòþÙ„‚½{÷jРA>ïðàA%''·¸Ïž={é@  [@ttt³s\¶nÝêÞ^¶l™–-[ÖdŸ¾}ûjÿþý­Ú¢££uï½÷êÞ{ïuß6þ|ÍŸ?¿ÅÇ„ÂüM 9õ=ç‘ÔÒº ã$”ÔGÒz/ûì’4唽ñ‚øAjjªÙ%XïMóx_¼ã½šJÿox¾V—Ë¥ÚÚZ99N« ïÿ]Ï•ÔÒ<æí1ßÙNm¸=íÎN¯Á…€ü€?ÄÞñÞ4÷Å;Þ y¡ôÿ†çk]µj•\.—œN§ÒÒÒL¬*0Bõ¿«Õµ§ÝÙéu"¸ÐP)))ªªªR§N|õDÇ¡ÝÁŽh­¨þýû›]Bívvê]@ Ð°†¸  6nܨŠŠ EFFjäÈ‘f—ƒA»ƒÐP*++Stt´Ù¥ „Ðî`GtÔÔ©SÍ.!ˆv;b:@@ÀèXsÐPK—.•Ëå’ÓéTZZšÙå DÐî`GtÔ˜1ctòäIEDD˜] BívD@@@ :Ôì‚hw°#æ `t,€!î¨mÛ¶©²²R;wVrr²Ùå DÐî`Gô £Õ.\xÊ}jkkõðÃw@5Àê6oÞ¬7jóæÍf—‚B»ƒуŽV)((PNNŽæÎÛâ~/¿ü²öîÝÛAU+»ûî»Í.!ˆv;¢>+))Ñ 7Ü ãÇ·¸ß®]»tûí·wPU´ÎÉ“'•““£+®¸B·Þz«OY¾|¹Î<óL}üñÇ®ÊèðÙ¬Y³TPP hÔ¨Qºí¶ÛšìsäÈÍš5KåååzóÍ75jÔ(-_¾\ï¼óŽn¸á]}õÕúÇ?þ¡Áƒë—¿ü¥$iß¾}š5k–®¹æ%%%éî»ïVuuuƒãnݺUW^y¥.¾øbM˜0AëÖ­kpÿÞ½{5qâD]zé¥4hV®\¸7`k¹¹¹Z¿~½6nܨÚÚZŸ³ÿ~½zõÒºuëtúé§kܸqzï½÷tÇwÈétêý÷ß×þýûµeËýêW¿R=$IwÜq‡:uê¤7ÞxCO<ñ„V¬X¡ÿýßÿuóƒ>иqãôè£jóæÍ6l˜®»î:}þùç’¤o¾ùF'NÔ’%K´iÓ&ÍŸ?_wÞy§>üðÃŽyc–qÏ=÷œrŸI“&)==½UÇ}ä‘GtäÈýâ¿huM………zä‘GZý8@è! £Õ|ím¨¡#FhàÀª¨¨Ð½÷Þ«™3gꩧžRYY™Þyç1B’ôóŸÿ\‘‘‘úâ‹/Ü¿ë®»4jÔ( >\’ô_ÿõ_:ýôÓµcÇIÒC=¤É“'kàÀ’¤›nºIN§Sýë_ýñr6‘ŸŸ¯§Ÿ~Ú§}###[}üîÝ»·ú1’4{ölUTT´é±Á⩧žÒÒ¥KõÔSO™] BívÄ"qè0:uR¿~ýî¾-::ZÛ¶mÓСCUQQ¡×_]‘‘‘*))‘$)??_3gÎt?&11Q_ýµûúG}¤.]º4è1ïÓ§þõ¯uÀ«XÁgŸ}¦«¯¾Úç ìùY(ÕÕÕš3gŽþò—¿hþüù>+«ûœoË#@[Ñî`Gô [DUU•fÎœ©ùóç+##£Éì:sçÎUBB‚RSSµ{÷î÷ÕÍÿ¶ÛŠ•çž{®~øaÝtÓMêÞ½»N;í4w/}AA$©²²Òëã÷ï߯n¸Aï½÷žû²wï^½ûî»R?´äÄ ©´Ôì*‚ÛÞ½{uÿý÷«´´Tµµµ5j”F¥>øÀ§Ç÷ÝwJKKÓ¥—^ª .¸@¯¾újƒû¿ýö[ýÏÿüÎ:ë,½ÿþû îûè£ôë_ÿZ—\r‰"""¦ÌÌLIÆiIßzë-IÒ³Ï>«Q£Fé–[nñÃ+¶Ÿ‹/¾X#GŽÔÅ_lv)!´;ØÝ"233•˜˜¨ÌÌLõë×O ,h²Ïºuë4zôhmØ°A………š5kVƒû¿þúkíÞ½û”«¬[‰ËåÒ¸qãtìØ1½üò˺úê«ÜÆgH’6lØÐä±………’ŒÞògŸ}¶Éýyyyþ/ZaáBÉé4.>.Ž6HLLÔúõëuþùçËáp¸¬½üòËOùØ-[¶è¯ý«–.]ªM›6iÈ!ºå–[tàÀIÆL²³³õÌ3Ϩ  @‡ÃýØÒÒRÝ|óÍzøá‡õüC‡Òe—]æÞgþüùÊÊÊ’$ýêW¿Ò{ｧիWà ºkÉ’%ºá†$IãÆÓòåË›íÑ£GkÔ¨QJLLÔ]wÝ¥ªªª÷6LC‡mõñÖ Ó‰'N¹ŸÃáði¿_|Q7nTFF†û¶ÚÚZÕÔÔH’âââ¯wß}·Áûñí·ßêßÿþ·$é’K.Ñ矮;î¸C;wî”d¬úž““Óª×þöøãRÝŸêÿû?skAóÎ?ÿ|Ýyçîë)))úþûïõÆoH2æ?ðÀºîºëš<ö¥—^Rii©zöì)IêÑ£‡žxâ‰Ãçù™ >t ÈËËSll¬bbb$¡´¤¤Dùùù ö‹ŠŠroïÛ·O .lr¬N»¬À™gž©¢¢"mØ°¡Å^€3Ï|X×\sÞyç­^½Z™™™ºêª«$I ,P×®]õÔSOiÈ!êÕ«—.¹ä’fO©¬Ìì p*]ºtipÝétJR“Ó©5ÞO’ ¤Ã‡ëÒK/Õ§Ÿ~*ɘó:gÎœUk_{öìQ~~¾öìÙcv)!´;ØÝvîÜéþõ]2V=ïÓ§{þµ§ââb=ôÐCZ±b…»—Ù“çл@˜:uª¨9sæ袋.òºßý÷߯òòr-Z´HcÇŽÕï~÷;mÛ¶M»wïÖœ9stìØ1IÒرcuÛm·éöÛo×/ùKUUUi„ :räˆ<¨ˆˆ¥¦¦jÍš5:räˆn½õV}õÕWîù}’töÙgkóæÍ9r¤ºwﮄ„­_¿^ }/àTNž4»´VDD„$5Yl®¹Eå.½ôRÝsÏ=Úºu«’““uÍ5×0½Ê‹ÜÜ\½üòËÊÍÍ5»„ÚìˆUÜ-Àáp4ùàw¹\ŠŽŽn²¯ÓéÔ„ ”ŸŸ¯Ë.»LT·nÝZýœ3fÌpŸ‡¼NjjªRSS[|ÜÀ}ú2%%E)))îë³gÏÖìÙ³›ì¦'Ÿ|RO>ù¤û¶É“'7ÙïTµwÞy, À2²³³•­†3‘Ž™UÚÀסé?þ¸¦M›¦|Po¾ù¦Þ{ï==þøãš>}z€+´—»îºËì‚hw°#ºŒ1Béééî륥¥*++ÓàÁƒ›ì¡ääd½ð JJJÒ¦M›ÜC½[#++Ë}^q€¥¦¦êúëSõCgì¶KJ6©"Ò°aô~ýz½ýöÛºþúëõÀhÊ”) ¦¦…:Ns3Ðî`G q·€!C†¨¦¦F‡’dœûÛét¶8D;,,LC† Ñ°aÃ:ªL@+0¼½ãEFFª¶¶Ö=¼¼¼< Ï—••¥M›6¹¯;VÓ§OWii©¾þúkwM’tøðaIÆÐyŽxC@·€ððpMš4I[¶l‘dÌ—Y´h‘$iñâÅî!å;wîÔÑ£G%!>,,L}úôip¬ÊÊJ¯çPtzÇ;ûì³%I<òˆ–-[¦;vxÝ·îGñºÏÕ:e?¬ì×xQÓºëu“¤ªª*M›6ͽ¨kUU•vïÞ­äädsÎ9’¤H2>Û_yå=úè£_/`_t‹X±b…6lØ ‡~Xßÿ½î¾ûnIRNNŽ{Á™•+Wê¼óÎÓ}÷ݧŋë¹çžkpŒO>ùD|ð¶nݪ7vôKx  w¼™3gêüóÏ×k¯½¦èèhýô§?mv¿·ß~[³gÏ–ÃáІ ”™™©ï¾ûNÏ=÷œæÍ›'‡Ã¡W_}UË–-ÓÉ“'µxñb½üòËr8zôÑGõ /H2Öùâ‹/4tèPõïß_×\s.¹ä½ùæ›îçúÑ~¤ùóç«´´TÙÙÙ!{†‘ììl=ýôÓÊÎÎ6»„ÚìÈQË8«²}ûv%''kÛ¶mÌA€úî;éŒ3›‚ß;ï¼£òòruíÚUcÆŒ1»œ€«kÓÒ6I-µé8IßJ:SÒ~oG;Ú&ÔÚ][ð÷×zX$€ ¨G8‚hw°#: ¤œ8!ÕÔ´¼O§NRçÎí{:h-:ØØÑ£ÒHŸ}VÛi§I=$]x¡ieYRm­”š*ý0}¸EaaÒ/)Ýr‹q}Ø0)6¶uÏG@­E@{î9iõꦷWVJo¿ÝñõXÙÞ½¾…sÉèaÏÉ1.’Ô¥‹ñø¸8ߟ¯ªªõ5Áª°°PÕÕÕ Wß¾}Í.!‚v;" €8кÛ}UZ*íÞÝôöðpiðàöÿ6Ã_Ôo õïßü~_-}þyÃÛNœÞ_ºé&ߟt ^NNŽ\.—œN§ÒÒÒÌ.!‚v;" €8Q¿½i“4a‚T\,UT´ý˜ß~k„ð’’æïON6ž«qH·ú©¿ü²~ûÁ¥[oõ¾ïßþ&mÙ"}ú©”›kÜÖÚ÷”€Ô»ùæ›USS£°0Îð‹ŽC»ƒÐÀÆ<zT”1»ñí­õöÛÞù$mÛVÿùÄØv¹Úþœ4}º±À[ˆß{ó èô €@# €y èíéA÷ èÑÑ ï»å–úSÕíÛ­›±]YÙöç ¤o¿­ß—fÎ4þõ…çP~zжËÍÍUyy¹ºvíªñãÇ›]BívD@ó žCÜ++S…µe]œ–zcž ÅY5 ×ÕuþùÒæÍR×®¾?–!î€TVVª¢¢Bá¾þ:øívD@óÖƒ.Á´ñbn¾8~¼~ûT½“ǧˆzmmýpü®][Î¥†ï_kG%pt ^JJŠÙ% Ñî`GœslÌ34FF¶/PÖiMºÃQß‹nÅ€îÙ‹Ý–s·Óƒ:l¬.„GFa¹=‹šÕñ èQQ§ÞßsX½ÕxÖÔ–€Þž<è µâ6Vë‚d{5«Óšt©>ø¶gåø@io@§ð—Ë¥ÚÚZ99N³ËAˆ ÝÁŽè`c-ôŽâ.Y{ˆ»?:=è@Û­ZµJ.—KN§Siiif—ƒA»ƒÐÀÆtqˆ0.§ÌÓ¬þ‘’’¢ªª*uêÄWOtÚìˆÖ 6È!î¾ôžKÁÐéAü£ÿþf—€D»ƒ±HØX ‡¸ÐéA‹€6U[ÛpwÏ%ºÔþ€Þ©“±:¾D@Çw°)ÏðéÏ!îÇÿúrŠ5©áiÖjkë­´7 ;Æ{Z^Þú<ªªZÿ|@°Ú¸q£***©‘#Gš]BívD@›ò Œþâ¾o_}Ïok{Ð%ã±m ÂÒÞ€.?@”—Óƒ´GAAÊÊÊíëÀhw°#:ØTs½½CÜó›úínÝ|{Œg𭬠΀.±HÐS§N5»„ Úìˆ9è`S§êAoË÷íÛë·'Oöí1º•ø# ×½§ô €@# €Mù{ˆû‘#ÒÁƒÆöÙgK¿þµo ö€N:è( q·€ªª*edd(**J•••Z¸p¡ÂÃÛì7wî\ýùÏÖ%—\¢‡zHIIIîûþþ÷¿ë7ÞPuuµ®¿þz]zé¥ùÐAöì‘x@úöÛ†·wê$Ýr‹tûíæÔ…ŽõàƒÒâÅ `sCÜ?ûLÚ´É÷ãîÜY¿=v¬ï³r@÷|""Úv zÐ@G! [@ff¦5}út­X±B ,Pfffƒ}Ö­[§Ñ£GkêÔ©úõ¯­Y³féõ×_—$}÷Ýwzà´}ûvUVVê /Ô‡~¨=z˜ñr@K–HëÖ5ß'ŸHãÆIññ[:Ve¥ôè£RuuÃÛO?Ýø׳ý0.m1x°ïûzþ(ÐÖ•ãÅŸ=èRi©ï«Ô——·íù€`´téR¹\.9N¥¥¥™]BívD@7Yqq±–,Y¢H’Ƨääd¥§§+ÊãG£Gv_¿ë®»ô§?ýÉ}ß#<¢1cÆÈáp(22RÉÉÉúÃþ 9sætì‹AÀÕ ?–¤ðp#(ÔÔ—ÚZé‚ ¤®]>¦S'é¶Û¤¹s;¶VƉõá<&F:÷\iÀéŽ;ŒÛ<Ö´ËÏ~æû¾VîA÷çôÚZÉélM@(3fŒNž<©ˆ¶eÚ€v;" ›,//O±±±Š‰‰‘$ÅÅÅ©¤¤Dùùùºè¢‹Üûy†õ}ûöiáÂ…îë;vìД)SÜ×ãã㵩5ãZaža£¨HêÕË8-ÖYg·=j\›?_ºûn‰AöçÙ.»¬éˆŠÄDéo“Þ}·íÏqÙeÆ=¾ ö€~æ™í«¡S'£¾¬¬}ÇìlèСf—€D»ƒÐM¶sçNõìÙÓ}=""B}úôQAAAƒ€.½íË–-Ó³Ï>«Ñ£G{=Æ™gž©W^y%ðÅ£Ã56ú÷7†<¯\ÙtÎëáÃÆ0Ûšch.Ýþ<‡{ œ×\c\:J°ô‡2F«xŽ`ñUx¸tãÒ¢EÒ_´íù@è  ›Ìáp4YÎår)::ºÉ¾N§S&LP~~¾.»ì2:tHÑÑÑMŽQRRÒìãaÞÂÆ̙ƥ±ÔT)'ÇØ®ª lmèžmÀsî·™‚= '%IÏ?ß¾:y¤}¡€n²#F(==Ý}½´´TeeeÜÌ MJNNÖ /¼ ¤¤$}ôÑGºêª«4bÄ}þùçîýŠŠŠôãÿ¸Åç1cF“EäRSS•ššÚÎW„@jíŠÔ<þ' _zÐ;Z°ô¶ÈÎÎVvv¶ûza¡$븋ٶm›*++Õ¹sg%''›]BívD@7Ù!CTSS£C‡©wïÞ***’ÓéTBB‚×Ç„……iÈ!6l˜$#äïÙ³Ç}QQ‘û>o²²²4|øpÿ¼t˜º°Ñ©“vêý=:§| VìA÷¬ƒ€nhüƒç°aÒÿýßvI|ADhÚ¼y³{5m‚: ívD@7Yxx¸&Mš¤-[¶hüøñÊÍÍÕ¢E‹$I‹/Öµ×^«AƒiçÎ:ãŒ3tÚi§©¨¨HaaaêÓ§$i„ š>}º$éøñãúòË/µråJÓ^§.lø4<{ÙéAVïAÆÓ¬ùÃ?ÿ)ýë_R£¥Eleýúõz衇ôâ‹/ºD®ªªRFF†¢¢¢TYY©… º§\µtBÏÝwßmv A´;Ø‘}p´+VhÆ zøá‡õý÷ß»ÿ˜äää(//O’´råJwÞyºï¾û´xñb=÷ÜsîÇŸ{ï¾û4wî\-\¸P¿ÿý¤ZÐâ|¬8=1ÄýÔ""|õbU‡ÒÉ“'õÉ'ŸÈáq"øÌÌL%&&*33SýúõÓ‚ |º4t ˆŽŽVVVV“Û·nÝêÞ^¶l™–-[æõ“'OÖäÉ“R¬£.løz:Ozð±âwÏà»p¡R‰Ä IDATôôÓRt´tÿýÆ)ÛÌd•€nw½{÷Ö¸qãÜV\\¬%K–èÀ’¤qãÆ)99Yééé:qâ„×û“\.ãú%—Hùù·`Ü’%ÒÌ™§ÞÏ*ïW°p8M}s¹\êÖ­›Ž=Úì}§Z…S€¯¡C‡ª¢¢B‘Vùe!v×PãS€JÒ±cœÔjè€Ø) >\¿Ù|OwUU}]ß|ã= ×ÔHï¿ßôöM›B' ó4ujÃÛ¬8©{wc»kWãߎZ0nÆSïÓ³§1êþ3bÄ¥§§»¯—––ª¬¬LƒVß¾}5kÖ¬fïk § ^_|±Ù% ÑîjîÏíÛ·s :‹! 6b§9èÕÕõÛ……F@j쩧¤Ûo7¶õ~¬cÇŒaÌ’tÆÒëN5xŽ`×xø¸d!îÕÕU]m\}f­º¶ãp4?U"<܆_÷ÃücÈ!ª©©Ñ¡C‡Ô»woÉét*!!Aqqq^ïÞЛ¨©©ïmöuw3ç {†goÍ3´{?–ç}‰‰¡Лûon¥tOÏ‹èE»ëÚGLŒäÑi ?;ùï|U?ü1 פI“´eË?^¹¹¹Z´hÑ)ïÞÐCÔãK?úQýõÎ¥o”† ñÿs½÷žñ¥Ù³‡41QúÓŸ¤>}üÿ|ÁʳÜCÜýЩßöl3¡Эރ.uL@¯ûÛÒÜ( øGYY™ž~úi9=óÌ3ºï¾û«+VhÎœ9Ú±c‡ªªª4oÞ<÷cZº¡gÏž=:yò¤"""4hÐ ³ËAˆ ÝÁŽè!ê¹çš¿í«¯ü?uÁÉã”î’¤/¾V®”ø¾æ»¶œ.ÊÌ€îù|ÞÚÔi§ÕoûÚƒîÐCéÜîÍýaÕôÆ=jjêºg{‚EGGëÞ{ïÕ½÷ÞÛäö¬¬,¯ñvBOnn®\.—œN§ÒÒÒÌ.!‚v;" Ãí›o¤ï¿÷/”gïgÏžõa+?ß¿ÏìÚÛƒnæt_zÐ[šƒl=è_|!?Þð¶ž=¥3Ïôþ˜æ‚®]zÐÉå2BºD:`ew…ÊŠž°Úì(Ìì`Ž?þQúàã2fLýí-…¤¶ª †Ý»ç)® kÍ-zïÚÒƒˆ9èµµÆ íu‹¶yãž;yù)0‡¸ÿæ7ÒÀÒyç5¼ÄÅI-u66·"ºzн’»g»¡°®ÈÈH÷è(´;Ø=D &]v™q8°þö@ôˆãrÖYÆõ½{àU\|ê §h ì÷—^’úö5Öøøã–k¸ðB©woiĈ–éžÃ¼ü¥éÖ­¾ÆPâþ Þï{é%ï÷Ù) 7^$.<ÿfуìŽ!îhÐëè€.IƒIŸ.••I±±¾gÔ(ã|ÇÞÂ^°ó h¾®âîk@Ÿ3§~eôÿý_#|7gófiûvcûŸÿ4Öð¶o]@oiM‡ÃU?ØÜxcóû}úiý¶Ý{Ðëþ;ôè!]½±½j•ñoK¯‡!îÍ£:|© DOuã€>l˜ôÆ­?Î{ïáð ýW›´·½¥9èß~[¿íryßïĉ†×[ _¾tIêÕË襥-÷.×ñ<û€z]ÍqqÆyà%éé§yÔus©›c§tôêjéÚk¥·Þjy?Ï÷ŒtÀº²³³uüøqEEE)55Õìr"hw°#:Ô£GývGô ß¿ñ<ûöùöø/¿”>ûÌØ.,ô{y¶È9èž÷ùºŸ¯ûž* ÿú×ÒÌ™-‡Ó:cÆ4œ’aÇ!îu¯Óó} 3noi^s=˜{Ð7mjýyññm{.«òòruíÚÕìRBhw°#::|ˆ{¯^ÒþàûãŸ~ÚqRhô@žݳ'º¥ý÷ÂûrÌSô´4iêT£½%ááÆ*ç‡q©­µgz]@÷œªáp4¼¯9Í]«ö {ÖÕÖEâªßŽ‹“N?½åý““¥ÿø¶=€Àã¹"-ÐAhw°#::< ·Vß¾õۡйHœgÐmi(|kzÐëŽémwO½z_…‡Ï,½n›!îõ¾ÿ¾~{î\éöÛÛW€„èr[ðÔÑsÐ[ëŒ3ê·ë2 Eœƒî©5=è-d_{ÐÛ¢îuÙyˆ»?z0q÷ü[ä9  ˜ÑƒŽ_~Ÿ|Rúñ¥{îñϱkkýÐéA7øú^ú2½q(ôwz zÝ1íÖƒ^[Ûr@oiº†¸û»=&¦}õ0_aa¡ª««®¾žC〢ÝÁŽèA‡bcëçÀJÒŒƪÚþààÚÐ{÷®do¼al7¾të&-^Üþz­,PCܨŽžƒÞvèuüу基ütåäähõêÕÊÉÉ1»„ÚìˆtÈé4VÑ~ôQãzMqnaÏóM·•g kk@ “Î>[Ú³§¾¾ÆÊʤyóŒ|™ólG èOæ¯t_Wqo »q÷l»žï‹/‹Ä5èóæIV]”Ö‹Äу—›o¾Y555 £o‡v; Ò(ƒÖzä£Çªî¼ÌmýRÝX[VoÎOH¿ýmó+}ïÛg¬ø\^nœŽmȶ?O[<ö˜´fMà{s={Ðýuš5zЛò à­íA÷ì‰Þ¼Y1¿µù=è‹5»„ Ú숀7ôz5ætÉ8ÿµ·3eüîwRF†±ýôÓÒÅ·ýy|.m„ôô–çB÷î¾íçË"qTK=èVY$.z[FP˜…9èmC@‡›•zK†¯ßÎÊ2.!!AÊÍ­çááR—.ÞaÃ|?ß³/‹Äº=Sì:ÄÝó…Ö.ª="ºCùü€7Ï/ý¾ž’ëT:" ÿô§Æ©âq÷–H_}Uý–[ê§X…¿‡¸³H\ûxëA÷eºgеêéÕêxÖw豸ckL©;¥bLLÃE,ØSnn®ÊËËÕµkW?Þìr"hw°#:ÜìÚƒ#ýãÒ[oµpüeÍiÛ6c»¼¼þv+.N×–UÜ9ÍZàx[$.؆¸{Ö÷ÄÆ¥­˜‡ÊÊJUTT(< €´;Ø‘#Ìâ íÐ%éœsŒKGر£> ?^{ __[ù2=PCÜYŽ©PœƒÞ^ÉÉþ;󤤤˜]BívD@·€ªª*edd(**J•••Z¸pa“_úJKKµhÑ"­Y³F—_~¹fÏž­!?,W^]]­9s樦¦FÝ»w×u×]§sÏ=·Õuص½£y¾«ô¶ÌAg‘¸À9U@oiºçH;ô[n‘lÝqzôn¸Á?5ØÝ233•˜˜¨éÓ§kÅŠZ°`233ìóüóÏ+!!A«W¯ÖÂ… 5mÚ4mݺU’´|ùrùäMžø zöì©;î¸C¿üå/õÏþS<òH‹Ï;cÆ õhtþ¢„„TI©’ìuôŽlsÐq¯­5¡go_‰ó|®@®â.µÚåì)§:ú‰ÆÙ¾ùÆû1¬>¼†ììlegg7¸íرc&U˜¯ÿþf—€D»ƒÐM6bÄ¥§§»¯—––ª¬¬Lƒnvÿµk×jâĉî!ñ’ät:5sæLIÒ;ï¼£C‡iÔ¨Q->oVV–†Þà¶W^©?_1=èÞÛôÆ=è’ñß­¹…Ȭփ^÷\v èÞ‰«Û.)1.-ñò§“ššªÔÔÔ·mß¾]Éœ7´À‚‘"´ 2D555:tèz÷î­¢¢"9N%$$4Ù7//OEEEºí¶Û¼oùòåÊÈÈhöñ§ÂwßصýË/¥F'$}ðAÓÛªªšè¾.×ÑÝ.N5ÝÓðáÒUW5¼-*ÊXp Á‰€n²ððpMš4I[¶lÑøñã•››«E‹I’/^¬k¯½Vƒ ÒîÝ»µjÕ*¥¥¥iß¾}:zô¨¶lÙ¢;î¸Ã}¬ÌÌLýèG?r?¾µè¾±Óô°0ãRS#íß/=ôoó¼ÛÒƒÈUÜ[ªÁŠZÐ/½TzøáÀ×aãƪ¨¨Pdd¤FŽiv9´;ØÝV¬X¡9sæhÇŽªªªÒ¼yó$I9990`€œN§F­ÂÂB-_¾\’1w=''G’´~ýz>|XñññîǶÝ7vêA—¤1c¤·ÞjÝc¼­A`Õ!îvám‘¸æ}k.´€]¨¬¬¬Å5v£ÝÁŽè­¬¬¬&·oݺս½ÿ~¯7nœ_ê  ûÆNsÐ%iÝ:iÓ¦æçš×q8¤ßýNڸѸî-x[e‘¸`è§êA' &S§N5»„ ÚìÈ¢‘fð žtïì4Ä]2êòeT×ÿX¿mõt»q?Õ"qž衇¯€póìAç4kÞy†C; q÷•/¡—EâÚ‡t´„¯€p ÄwÏãØ=ÀÖ±[º¯<ë·zz°ôææ ÛåÔqð†¸Ã9è¾±Ût_ùÒƒn•€n×!îô UK—.•Ëå’ÓéTZZšÙå DÐî`G6ð'ôêê†ÖóøvÊ=è¾.èÓ¬Cºçk  vcÆŒÑÉ“'a÷KØ ívD@‡›?úGI“&I‡×ß,=_‡gï­Ý_Ÿ?{нŠ»g­v è, T :Ôì‚hw°#¾ÂÍŸ«¸?ùdÃp.IgœÑ¾cZ…· n÷!îÞ~xðdÅEâ‚aˆ;sÐ уþìA/(¨ß¾â iÌé‚ ÚwL«ðЃ©=˜‰kK€Ô-ÌA@KèpóçiÖ¾þÚø·W/éïoß±¬&XºzÐ}â^Yiü0ôÁ­?þ¤IÒ+¯4ß³Ýt¡jÛ¶mª¬¬TçΕœœlv9´;Ønþâ^]-íßol÷ë×¾š¬(Xz[zÐ}Y$ά!îï¿ß¶p.I¯¾*åçKC†´íñÞ°H€Pµyóf÷jÚ%tÚ숀·°0#¤UUµ/ ÷]}@‹÷OmV¬sÐuš5³Vq/.®ß0@êÛ÷ÔÇýê+éÛoí’’¶×ç sЄª»ï¾Ûì‚hw°#›G ø[çÎFèúôS©wï¶Ã3´Ñƒnm9Íš•Wq÷ Ø>(Ýzë©;gŽôðÃƶç)ôü…UÜÐ:èÕË&55MWao‹Û « Ö€¨t³†¸{ôîÝ};nTTývyyëë:æ  %t4ðØcÒ¢Eþé=}ºÙå DÐî`G6@Ç ÖôsÎiýc¬¼Š»zÐY$@°:t¨***iv)!´;Øh¥` èW^)=ÿ¼´}{Ëûõè!Íkl[yw+t‰ª.¾øb³K@¢ÝÁŽè@+y œv臔šj\NeÙ2éàAû¬ânÅ!ît4ÆW@ •‚uzkÔ½V+,WXØü>ž½[7ߎË"q0_Vj. ;¡5gØJýÎ;›?o{]@ïÖÍ÷Ì:Ísл={ö(??_{öì1»„Úì(„úüÿh. Û}x{kÕôâbé?þ£éýžŸƒš^ØðúK/I?ùIÃÛŠ‹}Þ.5âètVqJrssår¹ät:•––fv9´;Øh¥ºÞrÏ^âP è]ºÿVVJ¯¿Þò¾è—]&Ý{¯1^’nºÉû¾­ è,q×]w™]BívÄW@  ¢£^÷ìy ·Ýæûœûº0ïO‡´p¡osË}?®YCÜ è‚]dd¤ûtÚìˆt  î¿_Z´È˜ƒ!ÝwŸÙu¬3¤[om¾—yñbiéR#Œ(˜œNiíZé¹çšŸƒ.I§fÔê«ŽâÎt4F@Úࡇ¤Ù³€Þ©S`z‰­Îé4.=ö˜”™)UTH11í 5ʸø =è0ݪªª”‘‘¡¨¨(UVVjáÂ… oÔ}VZZªE‹iÍš5ºüòË5{öl 2Ä}ÿ3Ï<£ï¿ÿ^555êÒ¥‹î¼óÎŽ~!'C¹¯¢¢†]»èÜÙÆ55,þ”­ãÇ+**J©©©f—ƒA»ƒÐ- 33S‰‰‰š>}ºV¬X¡ (33³Á>Ï?ÿ¼´zõj-\¸PÓ¦MÓÖ­[%I»víÒ† ´fÍIÒõ×_¯äädýô§?íð×Ø™Ãa s/+“”Þx£õÇèÛWºà‚æïc‘8¡*66Våååêj‹¶ÀT´;ØÝdÅÅÅZ²d‰8 I7nœ’““•žž®(.ÈêŠ+®$%$$())I»wïVRR’Þ}÷]•––º÷=í´ÓôÕW_Ð6ˆŠ2zAô‹_´íË–I÷ÜÓôöÖÌA' &cÆŒ1»„ Ú숯€&ËËËSll¬bbb$Iqqq*))Q~~~ƒýê¹$ÅÇÇËét*..N’tõÕWë½÷ÞÓ’%K´sçN8qB)))÷"€ rþùí?ÆÛo7{kæ ³H@è¡Ýd;wîTÏž=Ý×#""Ô§O袋.jö1k×®ÕäÉ“ýù¾Î>ûl½þúëúÅ/~¡^xAü±Âè~ÚäÏ–rr$—«u«¬4Ný& ä5‡EâкÉG“á\.—;|7gÆ ZºtiƒÛŠ‹‹õßÿýßúýï¯_ýêWZ½zu“ãzš1c†zôèÑà¶ÔÔTÐ@È;ýôÖš­Žg@?q¢ù}è¡#;;[ÙÙÙ n;vì˜IÕæ+,,TuuµÂÃÃÕ·o_³ËAˆ ÝÁŽè&1b„ÒÓÓÝ×KKKUVV¦Áƒ7»ÿÚµk5qâD÷xIÚ¼y³^{í5ýùÏÖèÑ£5~üx5JÓ¦Móú¼YYY>|¸ß^ê""Œ¹äµµÞºç"qž¿Ÿ1=ø4÷ƒçöíÛ•œœlRE€¹rrrär¹ät:•––fv9´;ØÝdC† QMM:¤Þ½{«¨¨HN§S MöÍËËSQQ‘n»í¶·¯]»Öý¥oäÈ‘Z´h‘Þyç:ÿr8ŒSï•—û§9è‚ÉÍ7߬šš¦à¡CÑî`G´V“…‡‡kÒ¤IÚ²e‹$)77W‹-’$-^¼X{öì‘$íÞ½[«V­Òرcµoß>}úé§Z¹r¥$éòË/צM›ÜÇìܹ³&MšÔÁ¯@—.Æ¿ÌA€†bccÕ§OÅÆÆš] BívDº¬X±BsæÌÑŽ;TUU¥yóæI2†å 0@N§S£GVaa¡–/_.ɘ»ž““#ÉXŽ¸¸X?þ¸zõê¥.]ºÐÔtæ  -è­¬¬¬&·oݺս½ÿ~¯w8ºé¦›Rßµ5 3üæTÝÛ"qÌAìrssU^^®®]»jüøñf—ƒA»ƒÐÀOâÍ«¬¬TEEE‹§€üv;" €ŸDFÿVVa¼qÈ& U)))f—€D»ƒñü¤®]j~%wæ  %ô €Ÿxô'¤®]ÞO:‚Ñ—_~©W^yE;wVii©®»î:sÎ9ªªªRFF†¢¢¢TYY©… 2Ì€Sà+ øÉ©zÐ=‰;U@'ÇÀ.|ðAMš4I÷ÝwŸ&Nœ¨3fH’233•˜˜¨ÌÌLõë×O ,0¹R˜Éår©¤¤D.—ËìRBhw°#zÐÀO<zI‰ôÖ[Ò¡CÒ7Jqq {ÐOµŠ;=è°‹wß}×ýå÷´ÓNÓ¾}ûtôèQ-Y²D$7NÉÉÉJOOWTT”™åÂ$«V­’Ëå’ÓéTZZšÙå DÐî`GtðÏ€þøãÒŠÆöÚµÒ0Áéºë®ÓM7ݤ·ß~[K—.Õ²eËôïÿ[±±±Š‰‰‘$ÅÅÅ©¤¤Dùùùºè¢‹L®fHIIQUU•:uâ«':ívÄW@ðÏ€þþûõÛ~hüËt£åË—ëì³ÏÖüc 6LcÇŽÕÎ;Õ³gO÷>êÓ§ L¬fêß¿¿¨þýû›] BívÄÏIà'žý«¯ÞWUÕº€ÎtØEyy¹ ¤N:iÆŒ4hš,çr¹Ýâ±f̘¡=z4¸-55U©©©~¯BMvv¶²³³ÜvìØ1“ª7tð“ºó KÒñã ï+)iÝ"qô Ã.n¼ñFeee鬳ÎÒäÉ“uã7êõ×_׬Y³Üû”––ª¬¬LƒnñXYYY>|x K€ÔÜžÛ·oWrr²I¡9tðÏôƾÿžEâ|ÊÊÊôñÇ+11Q’ôÒK/)..N]ºtQMM:¤Þ½{«¨¨HN§S &W ³lܸQŠŒŒÔÈ‘#Í.!‚v;â+ øIKýØ1‰Cð‰ŽŽÖÀ•ŸŸ/Ir8ºà‚ 4hÐ Mš4I[¶l‘$åææjÑ¢Ef– “è‹/¾`t(Úìˆtð“Öô 3ÁbÍš5Z»v­þñ¨sçÎúío+IZ±b…æÌ™£;v¨ªªJóæÍ3¹R˜iêÔ©f—€D»ƒÐÀOüÐéA‡] 8°Ùó GGG+++Ë„Š°/¾€Ÿ\x¡÷ûè8zÐÀO.¼PÚ¼Yڶ͸þÉ'ÒŸþdl;æ}wæ @" €_a\$)7·> ·vwæ &K—.•Ëå’ÓélvJ´;Ø$&¦~û™g¤'ê¯3Ä@(3fŒNž<©ˆˆ³KA¡ÝÁŽè ½zÕoõUÃû"#ë· è‚ÝСCÍ.!ˆv;â+ ȹçJãÆ5½}Ò$)>¾þ:sÐ уãpHo¼!•”ÔÏ? “ºwo¸=èèpyc, ØmÛ¶M•••êܹ³’““Í.!‚v;" €ÉèAì6oÞì^M› „ŽB»ƒÐÀdÌAìî¾ûn³K@¢ÝÁŽø hUUUš9s¦æÏŸ¯ŒŒ UWW7Ù§´´TŠ×”)S”——ç¾ïý÷ßWXXXƒËO~ò“Ž| ÚtHtKÈÌÌTbb¢233Õ¯_?-X° É>Ï?ÿ¼´zõjíß¿_Ó¦Msß·nÝ:=ûì³Ú¼y³>þøc=üðÚ8qb¾íÁtH q7]qq±–,Y¢H’Ƨääd¥§§+**ʽßÀuÅWH’”””¤Ý»w+))I¿ùÍotöúîÁi IDATÙg»÷ýãÿ¨3ftì ÐfÍô憽 ¸ÐM–——§ØØXÅÄÄH’âââTRR¢üü|]tÑEîýê¹$ÅÇÇËét*..N’„óššíÚµKƒî W ½‡q†·6O=õ”JKKÕ­[7MŸ>Ýìr"hw°#ºÉvîÜ©ž={º¯GDD¨OŸ>*((hÐ=­]»V“'OVttt“û6oÞ¬K.¹$`õð¿Æœ€ Ø :TŠŒŒ4»„Ú숀n2‡Ã¡ðF“M].W³á»Î† ´téÒfï{íµ×tÝu×ùµFÕ83ÿ@°¹øâ‹Í.!ˆv;" ›lĈJOOw_/--UYY™×!êk×®ÕĉÝCâÛ²e‹{ì±S>ïŒ3Ô£G·¥¦¦*55µÕðzЃOvv¶²³³ÜvìØ1“ªvA@7Ù!CTSS£C‡©wïÞ***’ÓéTBBB“}óòòTTT¤Ûn»­Ùcåååù<÷<++KÇoWíüƒ9èÁ§¹<·oß®ääd“*`w»víòÛ±œN§ýv<þC@7Yxx¸&Mš¤-[¶hüøñÊÍÍÕ¢E‹$I‹/Öµ×^«Aƒi÷îÝZµj•ÒÒÒ´oß>=zT[¶lÑwÜá>Ök¯½¦””³^ €6¢@°Û³gNž<©ˆˆ 4ÈìrlÆ)Iš2eŠ_ºgÏž é´;ØÝV¬X¡9sæhÇŽªªªÒ¼yó$I9990`€œN§F­ÂÂB-_¾\’1w=''§Áq6nܨٳgwxýÚ‡€ ØåææÊårÉét*--Íìrl&QÒI.?o—¤)r¹üu<ë¢ÝÁŽè­¬¬¬&·oݺս½ÿþSgÆ ~­ @Ç`‘8Áî®»î2»› îžî@¡ÝÁŽè`2æ vœæ f ÝÁŽø&cˆ;$:˜Ž€‰!î`ºÆsΙƒ Ødggëøñ㊊Šjr B Phw°#:˜ì’K¤=¤cÇŒë&˜[ø[ll¬ÊËËÕµkW³KA¡ÝÁŽè`²ž=¥¯¿–þõ/Éé”Î;ÏìŠÀ¿ÆŒcv A´;Ø,Àé”~ö3³«€™XŠ  UXX¨êêj…‡‡«oß¾f—ƒA»ƒÐP999r¹\r:JKK3»„Ú숀€€ºùæ›USS£°0fW¢ãÐî`GtTll¬Ù% Ñî`Güœ€Ð°†¸  rssU^^®®]»jüøñf—ƒA»ƒÐP•••ª¨¨Pxx¸Ù¥ „Ðî`GtTJJŠÙ% ‘]»vùå8N§S‰‰‰~9–¿Ñî`Gt d8%IS¦LñÛ÷ìÙcÙØ ‰’öHrùáX»$M‘Ëåcè0—Ë¥ÚÚZ99N³ËB£·›v;"   V­Z%—Ë%§Ó©´´4³ËAˆ ÝÁŽè©²²2EGG›] ””UUU©S'¾z¢ãÐî`G´V‚Ô-·Ü¢'žxB½{÷6»„¸þýû›]Bívfv 0œN§æÏŸ¯éÓ§ëã?6»p ô ¤V®\©N:éرczæ™g´lÙ2]yå•úÏÿüOuéÒÅìò@#ô ?ÈÎÎ6»Ëâ½iï‹w¼7ÖP7ï²G1b„Ž9¢Ûo¿]×_½fÏž­Ã‡›\!BÅÆõÖ[oiãÆf—‚B»ƒÐMVUU¥™3gjþüùÊÈÈPuuu“}JKK•‘‘¡øøxM™2EyyyMö©©©ÑÊ•+•““£>ú¨#J: ïxošÇûâï5¼ð ڰaƒ®¼òJ9RýúõÓgŸ}¦uëÖéÎ;ïÔÌ™3ùÌ@‡(((Ð_|¡‚‚³KA¡ÝÁŽân²ÌÌL%&&júôéZ±b…,X ÌÌÌû<ÿüóJHHÐêÕ«µpáBM›6M[·nuß_RR¢[o½U<òˆ ÐÑ/`Q©©©ŠŒŒÔ-·Ü¢Õ«W+>>Þ}_||¼ÆŽ«ûî»OÛ¶m3±J„‚©S§š]BívD@7Qqq±–,Y¢H’Ƨääd¥§§+**ʽßÀuÅWH’”””¤Ý»w+))I’4cÆ ÝsÏ=„s@_|±^xáÅÅÅ5{ee¥ºwïÞÁUoân¢¼¼<ÅÆÆ*&&F’§ÿÏÞ‡GUÞm¿‡ ƒU‹h@"ÐJ ‚ú²TP© }óJ¥ÊbDLµA4i¨¶BŒ¢)K•‹ &ÚB) ¨ ‚”‚$²°FÖÀÉ:Yæýƒ:mÊb€É<39ßÏuq•sròœ;OÄùÍsæœÂÂBíÚµ«Æqß6çÒù»Ýî~±µfÍ}þùçÊÏÏ×°aÔ––&—Ëå½à³ÒÓÓ/ÙœKçW—Ö®]ëÅDàrXA7h÷îÝjÑ¢…{;00P­[·Vnn®¢££/ú=+V¬Ðˆ#"IZ¹r¥n»í6 2D×_½FŒ¡üü|Mš4é¢ß_VV&IÚ³g‡ÿWPP ¬¬,Ó1|ssqÌË¥17úö÷nii©WÏûÄOèÕW_UóæÍÝÍø€¼šÔ ºA6›M5ö9wó}1kÖ¬Qrr²{{ëÖ­9r¤u×]wiĈZºté%ôC‡I’F埠þ‰ŠŠ2Ág17Ǽ\ssq‡Ö]wÝå•s=ÿüó:uê”û‘j ÐÌ™3µgÏ=õÔS^ÉHRrr²‡ìv»âââLÇEPwðG4èõîÝ[S¦Lqo©¸¸X‘‘‘=~ÅŠ:t¨û’xI*//¯qLTTÔeïž|ß}÷iñâÅêر£š4ir? ¶ÊÊÊtèÐ!Ýwß}^;çÍ7߬ôôôû¢¢¢ôË_þ’^5hÐ UTT(00ÐtXuDƒnP·nÝT]]­üü|µjÕJyyy²ÛíêСÃÇfgg+//OãÆ«±¿GÊÉÉqoiРA—¬/¿üRsçΕ$ÅÅÅiùòår8’¤uëÖé'?ù‰Ÿàkz÷î­ïÿûJLLÔsÏ=§[o½Uëׯ×óÏ?o:¸VÐ KMMU||¼vìØ¡ÊÊJ͘1CÒù;ïvêÔIv»]÷ÜsŽ?®9sæH:ÿÙõo/YŒŒŒTjjª’’’Ô¬Y3uíÚU#FŒ0öó|GÏž=õ׿þU .Ô† ôàƒjÉ’%úÁ~`:¸tÃBBB”’’rÁþ­[·ºÿ~ôèÑËŽ1dÈ 2ä;ÏUYY©©S§*88XN§SIIIܤ®>ûè£ôòË/ëÃ?tŒàrsb…ùúôÓOõÊ+¯èøñãzâ‰'4iÒ$Ùl6ËÏ‹$eeeiÊ”)ÊÍÍÕ¤I“4qâDIÔÌ·œN§¢££õæ›oª_¿~ÌË¿üú׿ÖôéÓ%Ißÿþ÷õå—_Ÿ›öíÛ륗^ÒK/½äÞ÷ÕW_é–[nñèy€ËÉÌÌ”ÓéT£F¸$¼†ºƒ?âw IHHPDD„Ô®];%&&šŽä5ùùùª¨¨Ð_|!›ÍæÞ¹9©ïóU^^®Å‹kúôé=z´¦NªŒŒ IÖžéüÜüýï×¢E‹4mÚ4Mš4IÛ¶m“ÄÜ|ë·¿ý­rssÝÿž˜—óusäÈ­Y³Fk֬џþô'If禢¢Bï¿ÿ¾^}õU%&&*11Q/¾ø¢ÆŒãÑóßeÓ¦MZ·n6mÚd: ,„ºƒ_rÁÎœ9ãjÒ¤‰«  Àår¹\t5oÞÜU\\l8™÷8N—Ífsåææº\®Ëωæë‹/¾p9÷öÿýßÿ¹~øa×Ù³g-=/.—ËUVV檮®vowíÚÕµeËË×Ì·>ÿüs×Â… ];vt}öÙgÌË¿,X°Àµ`Á‚ûLÏÍ<à qEGG»ú÷ïïêß¿¿«oß¾®–-[zìž™™é’äÊÌÌ4ðˆokZÊtI®Ëüiû¯ãÚ~Çq¾ú‡»þŽß¿¾‡t‹ÈÎÎVXX˜ûmááá*,,Ô®]» 'óžÿ~ÄÆåæÄ óuûí·«iÓ¦îí[n¹E;vÔÎ;-=/Òù§!|»2|îÜ9ýÏÿü¢££-_3ÒùÇA.[¶L?ûÙÏÜû˜—óæÏŸ¯ñãÇëž{îц $™Ÿ›íÛ·+''G[¶lÑÚµkµvíZ}öÙgúÃþà±sÏ¡A·ˆÝ»w«E‹îíÀÀ@µnÝZ¹¹¹S™u©99|ø°%çkÅŠ3f óòÖ¬Y£ªQ£Fª¬¬dn$½öÚkš6mZ}ÌËy[¶lÑáÇuÓM7©_¿~ÊÎÎ6>7?ýéOõÕW_]°?""ÂcçžÃMâ,Âf³]pã!‡Ã¡C‰Ì»Ôœ4mÚTß|ó¥æ+33S#FŒPDD„þþ÷¿3/ÿÒ½{wM˜0AÏ?ÿ¼BCCÕ®];KÏÍG}¤^½zÕx†¶ËåâßÒhß¾½æÏŸ¯àà`½óÎ;êÚµ«Ñ¹‰ˆˆPrrr›VVV*##CË–-óØy€g°‚n½{÷ÖÜÛEEE*..Vdd¤ÁTf]nNî¸ãËÌ—ÓéÔ’%KÜÏEîÓ§óò/ßûÞ÷ô³ŸýL³gÏÖêÕ«-_3ÉÉÉ5j”š7o®æÍ›ë믿VLLŒòòò,=/3vìX­_¿ÞxÍÌ›7OëׯW||¼ûNîñññZ¾|¹ÇÎÔÆüù󕜜¬ùó盎 ¡îàXA·ˆnݺ©ººZùùùjÕª•òòòd·Ûݳ¢ËÍIxx¸eækÖ¬Yš6mš48ÿ~Ý­·Þʼü—[o½UÑÑÑ–¯™Å‹«¼¼\Òù•ó»ï¾[³gÏÖÀ5kÖ,ËÎË¥ôìÙÓxÍ<ýôÓ9r¤5jTcÿ{ï½ç±sµÑ½{w•——+((ÈtXuDƒn6l˜6oÞ¬˜˜eddhæÌ™¦cyUEE…¤ó—wJ—Ÿ«Ì׋/¾¨~ýú©¨¨Húì³ÏÔ¿ËÏKyy¹víÚ¥ž={J’>øà=òÈ#jР¥çæ{ßû^í† êúë¯W‹-,=/’tæÌeddè‘GQii©æÌ™£éÓ§¯™GyDåååúâ‹/­;v¨{÷îúéOêÑóߥOŸ>u2î¾}ûäp8<2–Ýnçþ õL]ÕP—l.—Ëe:¼£¸¸Xñññjݺµ*++5cÆ Ó‘¼¦¸¸Xï¼óŽž}öYM›6MÏ<óŒÂÂÂ.;'õ}¾”Pc_dd¤²³³-=/Òù›ž 2D:tPdd¤|ðAÝ{ï½’.ÿó[anþÓ7Þ¨E‹©oß¾–Ÿ—¼¼< ù¤ÇÏõÕW_iåÊ•ºé¦›t÷Ýw+44TS§NUpp°œN§’’’.x¬¬###C‡Cv»]qqq¦ãÀ"¨;ø#tꩾ}ûê¿ø…¶oß®×_]'NœÐ½÷Þ« &xô¢A žºñÆküiß¾½Ú´i£¹sçzdü‚‚mÚ´IC† Ñ /¼ —Ë¥¨¨(egg«E‹îãÕºukåææzä¼ÔW\â@=¡ØØXÙl6÷¾•+WêðÈø[·n•Ýn×m·Ý&IzõÕWõÎ;ï¨aÆÜÎáp($$ä²ãMžªêêj5hÐÀtXuDƒ@=¬}ûö馛nrï{ÿý÷¥.]º¨ªªJo½õ–Á„°ŠoïòxuÄÛIÔS;w®ÑœKRŸ>}ôØcI’ôóŸÿÜD4p4èÔSGŽQYYY}‹-Ò©S§ÜÛAAAÞŽ.Kܨ§¦Nª=zèž{îÑ7Þ¨•+WjãÆzûí·MGƒÅddd¨´´TMš4QLLŒé8°êþˆtê©ïÿûZ¹r¥ŠŠŠ”––¦:hÅŠ7nœéh°§Ó©òòr9NÓQ`!Ôü+èÔcš7ož‚ƒƒURR¢àà`Ó‘`AÇ7DÝÁ±‚@=µgÏÝÿýzøá‡%IÙÙÙJJJRAAádàbhШ§^{í5uìØQ;w–$õêÕKíÛ·×ìÙ³ 'Cƒ@=¦¹sçª}ûöî}N§S‹-2˜ Väp8TXX(‡Ãa: ,„ºƒ?â3èÔS}úô©±]^^®ääd 8ÐP"XÕ‚ äp8d·Ûg:,‚ºƒ?¢A žêÞ½»}ôQ)33S+V¬PÆ õì³ÏšŽ‹>|¸*++Õ°!/=á=ÔüÕ @=uóÍ7kÑ¢EZ·n>þøc½÷Þ{ºÿþûh:,¦cÇŽ¦#À‚¨;ø#tê©Ñ£GkĈzðÁ5`ÀÓqÀw A·˜Ó§Oë“O>QÇŽÕ¤IÓqÀ2JKKuøðaÝwß} óÊ9·mÛ¦gžyæ‚ý'OžÔõ×_ï•  öhÐ-æ“O>ѨQ£LÇËZ¼x±y䯜ëõ×_×Ǭ-Z¨Aƒón)++SJJŠRSS½’¤uëÖ©¼¼\AAAêß¿¿é8°êþˆÝbn¼ñFIç_ víÚÕpšúiòäÉJII1£Þb~ëó[wöìÙ£Q£F¹{øqãtèÐ!MŸ>½Æ~›ÍFƒ¯ÊÍÍUqq±BBBLG…PwðG4èÓ¸qcIR×®]Õ³gOÃiê§ÐÐPæ¶1¿u‹ù­{ßþö†1cÆèî»ïVÛ¶m Ir¹\š;w®×2Òùû!ÞFÝÁÑ P´hÑBO?ý´âãã5qâD…„„¸›óoý÷Š:ð 4èÔ#wÜq‡$I¿ùÍoôå—_ê¶ÛnÓOúS÷#‡ìv»Á„àRhШGºtéâþûôéÓõýï_úÓŸÜ7‰¾‹ð°ØØXÓê5æ·n1¿þ/00Ðýw›Í¦¨¨¨ šómÛ¶é?ø·£ÁÂ’““åp8d·Ûg:,‚ºƒ?¢A<Œ§n1¿u‹ùõgÏžUNNŽ¤ó7„;wîœ{[:ÿ<ö7ÞxC .44hÐ UTTÔx ¨kÔü :õÈÂ… /h¾—/_^cÛf³Ñ Ã«ºwïn:,ˆºƒ?¢A ùÉO~¢qãÆ©Q£Fýzqq±Þ~ûm/§µAƒ@=2uêÔï|–}XX˜—Ò€+Aƒ@=ò]͹tþQl€7effÊétªQ£FŠŠŠ2AÝÁÑ  NmÚ´É}7m%x uDƒî>úè#½üòËúðÃÕ¡C‡‹ó駟jÕªUªªªÒÈ‘#u×]wy9%ÀÅMœ8ÑtXuDƒîãòóóUQQ¡/¾øB6›í¢Çœ}ú˜Ž ¢îàø z=лwoåä丷óòòÔ½{wƒ‰WŠÝTTTH’*++ÝûfÍšånÊ|ðAíܹS’TRR¢ƒê‰'žð~PÀUãwW\\¬wÞyG6›M .Ô3Ï<£°°0¥§§«S§Nêܹ³ºvíªgžyFÓ§OWuuµ^ýõïü :€·ää䨢¢Bêܹ³é8°êþˆÝÇ…„„hÒ¤Iš4iRý[·n­±=bÄ1›Ñj%##C‡Cv»]qqq¦ãÀ"¨;ø#t\±¤¤$MŸ>Ýt à'&L˜`:,ˆºƒ?â3è¸"¹¹¹JOO7ø‘   ÷À[¨;ø#tÔZaa¡~øa•””˜Žõ :jí—¿ü¥rssuâÄ 0@ãÆ»ä±7nÔÏþsÝy犊ŠÒÛo¿}Á1ï½÷žîºë.EGGëñÇ׎;j|}çÎ;v¬úöí«[o½UÓ¦MSUU•$éܹszûí·Õ³gO-Z´H/¿ü²Ú¶m«?üáZ´h‘úöí«ÄÄDÍ›7Oááázýõ×=;àa|µ6wî\•••iýúõZ»ví%+..ÖСCõþûïkîܹŠ×¤I“Ô·o_÷óÙ_}õUýùÏÖòåËuà 7¨oß¾1b„vïÞ­† ê믿ְaôlÙ2õèÑC«V­ÒOúSíÞ½[Ë—/—ÃáÍfÓ¶mÛ´téR3FTpp°ôùçŸ+00PcÇŽÕСCÕ¸qcoMø/iii*))Qpp°bccMÇEPwðG¬ ãŠ¹\®Ë~}ýúõ*((P¯^½$I÷Þ{¯\.—<(I:zô¨5zôhÝpà ’¤‰'ÊápèäÉ“’¤qãÆ©W¯^êÑ£‡$éG?ú‘bbb´bÅ >|Xáááº÷Þ{%I={öÔСCµhÑ"1BC† ‘$µk×N±±±zûí·õä“Oz~"àÕÕÕúðõnÝ:÷¾={öèÃ?Ô‰'ÌÔZXX˜Zµj¥°°0ÓQ`!Ôü+èð¸{ï½W[·nUhh¨Îœ9£Ï?ÿ\ÒùÏ°Kç//++SLLŒû{zè!=ôÐC’$§Ó©Ï?ÿ\/¼ðBqï»ï>½÷Þ{úðÃõË_þR ž/ßN:Õ8. à¢û៪««õðÃ×Ø·|ùr-_¾\«V­r¿Éð]UK€½ IDATƒ 2DÝÁÑ ÃãÔºuk?^6›M>ø ¤¯¼çææJ:߈_ÌW_}¥ÒÒR:uªÆþAƒÉf³iË–-u˜ÌàwxÜW_}¥Þ½{ëмyóÔ­[·_ÿvÅsÍš5|ïñãÇ©æÍ›kÆ 5¾ÖªU+)**ªî€!4è¸" 4PYYÙe™9s¦š5k¦¡C‡ÖØ_]]-IêÝ»·4h +VÔøúêÕ«U]]­† jÈ!ÊÎÎv_/I'OžTYY™î¹çý4ÀŽ?®#GŽèøñ㦣ÀB¨;ø#t\‘¶mÛ*//Ok֬ѻï¾{Écöï߯“'Oª¤¤Dï¿ÿ¾¤óww?pà€"""4fÌ}üñÇúùÏ® 6())I»víRxx¸$é‰'žP£F4yòd¹\.iúôéš1c†ûæsß^"Ÿ——Wãü—ÚÌHOO׻ﾫôôtÓQ`!ÔüŸAÇ=z´>øàÅÇÇë÷¿ÿýEyòÉ'µk×.õêÕKƒ Ò˜1ctóÍ7kß¾}ºÿþû%I©©© Ó{ｧ-[¶hìرzüñÇÝcôîÝ[Ÿ~ú©^|ñEÝzë­êÒ¥‹úöí«gŸ}VÒùËè_|ñEÙl6Í›7O:tÐC=¤cÇŽ)11Q’ôᇪGzâ‰'êxVÀå<ú裪®®Vƒ¬ Á{¨;ø#t\‘›o¾Y999—=¦C‡úë_ÿZcßO@@€fÎœ©™3g^rœ>}ú\ôsê’tË-·èüãûÛ¶m«yóæiÞ¼y—ͼ‡Ç\Áêþˆ·“ð4èø.q@ÊÈÈPii©š4i¢˜˜Óq`Ôü :ê”ÓéTyy¹LG…PwðG4è|VzzºŽ=zɯÿùÏÖW_}¥ &¨Q£F^L¸Ç7DÝÁÑ ðYo¾ù¦6mÚtɯ/X°@’4nÜ8tø=n€`uÊápÈårÉf³Én·›Ž‹ îàhÐP§,X ‡Ã!»Ý®¸¸8Óq`Ôü :êÔðáÃUYY©† yé ï¡îà¨VÔ©Ž;šŽ ¢îà¸I>€tWYY©©S§*88XN§SIII ¸à¸… êܹsª®®VãÆõÔSOH ¸Z¬ û¸„„EDD(!!AíÚµSbbâÇìÙ³GkÖ¬ÑäÉ“§µk×jóæÍÒ\hݺuúä“O´nÝ:ÓQ`!Ôü+è>ììÙ³š={¶Nœ8!IÌf³]ðbÆáp\ô2³gÏê¿ø…^ýu=öØcz÷Ýw/ûBhòäÉ ­±/66V±±±ž –––¦´´´û ¥ñŽßþö·ÊÍÍ•Íf“ôï{¨Œ?^©©©JLLTBB‚á”ø6tÖ»woM™2Ž]TT¤ââbEFFÖ8nÓ¦MZ¾|¹Þ{ï=ÝsÏ=Š‰‰Ñ€ô³Ÿýì’c§¤¤¨gÏžu,íboxfee)**ÊP¢ºµqãF…‡‡«yóæ’jP×Aû°nݺ©ººZùùù’¤¼¼<ÙívuèСÆq+V¬p¿èëß¿¿fÎœ©Õ«W{=/ÀzŠŠŠ´lÙ²o _î*°¦ääd%$$(99ÙtXuDƒîÃ4lØ0÷#Ó2224sæLIÒ¬Y³”““#IêÛ·¯6nÜèþ¾FiØ°aÞ °œ×^{MÓ¦M«±ïr÷P5 4H>ø  d: ,„ºƒ?âw—ššªøøxíرC•••š1c†$)==]:uRçÎuÿý÷ëìÙ³zã7Ô²eK5nܘPç>úè#õêÕK-[¶tïs¹\Wt•ÿÄýQê¯îÝ»›Ž ¢îj²âýQü º QJJÊû·nÝêþ»ÍfÓ#<âÍX@X¹r¥†ªêêê+þÞmÛ¶©aÆš8qâEÿÍð¼äääÿ=*,,TLLŒžþy8pÀ½ÿR÷PùoÜêŽÕî⯸Ä€Ïp¹\ªªª’Ë庪ﯪªºªæÀÕY¼x±¶oß®íÛ·kÛ¶mjÓ¦ÞyçMœ8±V÷P5±‚®Ê÷¾÷½Û 6Ôõ×_¯-Z¸ï¡Sã*°¦ÌÌL9N5jÔˆÕ:x uDƒ<îR÷P5mÚ´I‡Cv»F ^CÝÁÑ 8tèûï—º‡ ¬iâĉ¦#À‚¨;ø#>ƒ€ A`ܹsçd·Û5räÈkkîܹ²ÛíúóŸÿìd€÷p‰;ŸPTTä‘q***TQQ¡ÊÊJŒx :êÔüùóUTT¤¦M›jüøñ¦ãÀ"¨;ø#tÔ©îÝ»«¼¼\AAA¦£ÀB¨;ø#tÔ©>}ú˜Ž€:´gÏe·Ûᑱ¨;ø#tWÁ.I5j”GGÍÉÉñX“øtFÅÄÄh÷îÝwÒ¤Iš6mšþñ¨U«VDHÊ‘äðÐx{$’Ãá©ñÿCƒÀ¨¯¿þZ‡òø¸'Ož”$UUUy|lÀ•ÉÉÉQEE…Õ¹sgÓqàQ¾»ÒMÝÁÑ  NeddÈápÈn·+..ÎtXuDƒ€:5aÂÓ`AÔü :깂 ÔüQÓXSee¥œN§\.WžÇétÊétÖé9O A`Ä3Ï<£   egg×éy:tè [o½µNÏx—¸ N¥¥¥©¤¤DÁÁÁŠ5AÝÁÑ  N………©´´TMš41BÝÁÑ  N 4ÈtXuÄgÐð4è¼*77WéééÚ·oŸ×Îép8”žž®Í›7{íœÀ•âw^µiÓ&¯ß¨%//O±±±zì±ÇtÇwxõÜéøñ㪪ªR@@€Ú´ic:,‚ºƒ?¢A@JOO—ÃáÝnW\\œé8°êþˆ݇UVVjêÔ© –ÓéTRR’.zluuµæÏŸ¯ÐÐP…‡‡ëî»ïörZàòÎœ9£Ç\ÇŽ3–aÍš5:t¨ž~úi 8ÐX°šG}TÕÕÕjЀOWÂ{¨;ø#t– ˆˆ?^©©©JLLTBBÂÇj̘1úÍo~£N:H |·²²2­X±Âh†#GŽèÈ‘#úñl4XMXX˜é° êþˆ·“|ÔÙ³g5{öl=üðÃ’¤ÁƒkÎœ9*))¹àØÉ“'ë駟¦9‡Ï:r䈾þúkÓ1ÜòóóuèÐ!UTT˜Ž¸Ñ û¨ììl………éºë®“$…‡‡«°°P»víªqÜš5kôùçŸ+??_Æ SZZš\.—‰ÈÀ%õë×OwÞy§énS¦LQ§N”››k: àÆ%î>j÷îÝjÑ¢…{;00P­[·Vnn®¢££ÝûW®\©Ûn»MC† Ñõ×_¯#F(??_“&M2à*--U“&Mc:|Üž={<2ÎöíÛår¹Ôºukê~ƒÝGÙl¶ nçp8RcßÖ­[5räHê®»îÒˆ#´téRtà3œN§ÊËË/y³[à<»$iÔ¨Qmøðá Ö=÷Üã‘ño A÷Q½{÷Ö”)SÜÛEEE*..VdddãÊËËklGEE)--í;ÇŸ|¸éð ’r$9<2Ú²e{$Ò³Ï>ë‘ño A÷QݺuSuuµòóóÕªU+åååÉn·«C‡5ŽëÑ£‡rrrÜÛAAA4hÐwŽŸ’’¢ž={z<7ðŸþö·¿éøƒé—´hÑ"Ýpà 4è𸋽ᙕ•¥¨¨(C‰À_D˜ÅMâ|T@@€† ¦Í›7K:ÿÙ­™3gJ’fÍšånÊãââ´|ùr9çßi\·n~ò“Ÿ˜ ¸j¬ û°ÔÔTÅÇÇkÇŽª¬¬ÔŒ3$IéééêÔ©“:wî¬ÈÈH¥¦¦*))IÍš5S×®]5bÄÃÉþÍápÈårÉf³Én·›Ž‹hÚ´L 4SYY™é(@­Ñ û°¥¤¤\°ëÖ­5¶‡ ¢!C†x+ÀY°`‡ìv»âââLÇEŒ¿AÍšÅiÆ >õ¸WàrhÐP§†®ÊÊJ5lÈKOxϲe·) à)¥¦¦šŽÔ¿%x\zzº¦Nªüü|ÓQ¾Ó©S§tã7êÐ[o½e:ÔK;v”$íÛ·OYYYÓSÏÊFýuøp˜¤ 3¨5tWTT¤Ã‡›ŽQ+UUU:|ø°_¼™þlß¾}êܹsŒÌgÚÔ4è¨sß>qFZ,©«‡Fµ‹Çr¨OhÐP§Ö­[§#GŽ¨_¿~úì³®’zšŽ è×o¯7¾O{÷îUÏžÔüÏAà1N§Sû÷ïשS§LG¹b‡Cû÷ïWAAé(PïäææêÔ©SîÏ¢ÞбãÝtÓM:sæŒé(@­±‚Àc>¬.]º˜ŽqU>úè#EDDhÎœ9š8q¢é8P¯Œ=ZYYYzî¹ç$M2±hÑ’žVff¦é(@­±‚À#Î;§ÂÂBÓ1®YII‰ TUUe: ,†€GtîÜYÑÑѦc\³)S¦¨yóæÊÎÎ6Cƒ€à3è®Éš5ktøða•––šŽâQË–-Ó¶mÛ4zôhÓQÀï%''Ëáp(..NÉɦÓÀ*ââV«Y³—µzõjîâ¿Á :€k2wî\7î?žo[?$&&ê駟6ê…Aƒ©GZ½zµé(°Õ«»jÅŠêÚµ«é(@­±‚€:Õ½{wUTThçΦ£ÀBvî —”¥ððpÓQ€Zc@ƒ€ A@ÊÌÌÔe: ,$**W}úôQnn®é(@­Ñ ¸*K—.Õ€´~ýzÓQêLII‰  ^xÁtðk›6mRNNŽúôéc: ,¤OŸƒêß¿¿ Ò¦M›LG…lÚt“¤uš={¶é(@­Ñ ¨µ7ÞxC%%%¦c³aÃmØ°AÆ £A€Çq‰;>€tVYY©©S§*88XN§SIII ¸äñéééš7ožÖ®]ëÅ”°‚}ûö©°°PÕÕÕ¦£ø„¯¾úJ7VTT”é(àrrrtüøquîÜY99¦ÓÀ*:wÎSÆ‘ÊËË3¨5VÐ}XBB‚"""” víÚ)11ñ’Ç?~\3gΔÍfóbBXÅÓO?­Ûo¿]eee¦£ø„±cÇ*::Út ðÊÌÌTLLŒé(°˜˜z衇´cÇÓQ€ZcÝG={V³gÏÖ‰'$IƒVTT”¦L™¢ààà Žÿío«§žzJ~ø¡·£\Ö„ ´mÛ6õíÛWÒ#¦ãÀ"Þzk€l¶¾Êf³]pC8‡Ã¡÷vUU•RSS5kÖ¬+òäÉ ­±/66V±±±WõÒ—_~©yó晎á³æÍ›§¾}ûÒ ãiiiJKK«±¯  ÀPß-BRŽ$‡Ç´ÿk\¨=tÕ»woM™2Ž]TT¤ââbEFFº÷ýãÿмyó´páBI’Óé”ÓéT‹-töìÙËŽŸ’’¢ž={ÖMx°¸‹½á™••ÅÿaYiiiÊÏÏWll¬þë½+B3]ßÄÆnQpðmÙ²…×½ð4è>ª[·nª®®V~~¾Zµj¥¼¼<ÙívuèÐÁ}L¯^½j|Æéü£þô§?éƒ>0ÔCûöí“Ãqm+ËUUUªªªÒéÓ§=” øn§O7U“&ùjÚ´©é(@­Ñ û¨€€ 6L›7oVLLŒ2224sæLIÒ¬Y³4dÈuîÜYíÛ·wOË–-Õ¸qãû€«ár¹jü/.Íår¹ç‰Ç¨oöíÛ§Î;{xÔ·=<pq«WGJZ¡—^zÉt ÖhÐ}XjjªâããµcÇUVVjÆŒ’¤ôôtuêÔé‚ÿ`Úl6xĆ Ô¯_?Ó1ü† Ô AÅÇÇ+))Ét𨯜/–ÔÕ#ò¹l¸t¢””” öoݺõ¢Ç=Z£G®ëXÀrºJâ3¼P×hÐÀUûôÓOõÊ+¯èøñãzâ‰'4iÒ$Ùl6UVVjêÔ© –ÓéTRRÒO'uÜpÃqT©ª*@'N´1qà  ç)ð+ Lþ©¼¼\‹/ÖôéÓ5zôhM:U’¤„„EDD(!!AíÚµSbb¢á´0)66]cǾ«ØØtÓQ`!±±_hìرúâ‹/LGjtn.—Kÿó?ÿ£sçΙŽâw-Z¤uëÖ)99Y½zõ2ðŠ;wjÎœ9jÚ´©ú÷ï¯;vhÉ’%ºûî»5{öl8qB’4xð`EEEiÊ”) 6œ&,Zô¨4¨Vu5kCðžE‹ú¨AƒúÓŸþd: Pk4èjظq£é~éèÑ£:zô(—ÑÁRn¿ýöÛ·Ür‹Š‹‹µsçN………éºë®“$…‡‡«°°P»víRtt´‰¨0ìÌ™0Ó`AgÎ4•ÄcÖà_xxÄŠ+4fÌíÞ½[-Z´pï TëÖ­•››k0¾tpÍ2335bÄEDDèïÿû7„s8 ¹ì“'OVhhh}±±±Šõx^°š´´4¥¥¥ÕØÇ•¾‡€$ièСڼy³é~/66V5ÒÁƒÕ¤IÓq¯p:Z²d‰fÍš%IêÓ§¦L™âþzQQ‘Š‹‹yÙqRRRÔ³'òªbb2Ô¤I©JK›(##ÆtXDLÌv5i2RÛ·oçw‹.þ†gVV–¢¢¢ %ÂÅР$9sFyyy¦cø½³gÏšŽxݬY³4mÚ45hpþ“s·Þz«ª««•ŸŸ¯V­Z)//Ov»]:t0œ¦4jäTPP¹ªªxÔ¼§Q£*©ªªÊt ÖhЋ۵k—öîÝ«3gΘŽR¯üå/QëÖ­õÃþÐt N½øâ‹êׯŸŠŠŠTPP Ï>ûLýû÷×°aôyófÅÄÄ(##C3gÎ4-[6ÜtXвe=%Ó³Ï>k: Pk4è€Å-Y²D¯¼òŠéõNll¬¢¢¢´uëVÓQ€:“ ™3gÖh¾###5fÌ¥¦¦*>>^;vìPee¥f̘a0)þ\•—^zI/½ôÒE¿¢””/'À¿ñ˜5À¢öìÙ£¸¸8­Y³Æt”zëÈ‘#Š‹‹ÓòåËMG£š6u¨Y³B5mê0Ò´i™š5k¦²²2ÓQ€Zc°¨Ã‡ëõ×_7£^;uê”^ýu¹\.ýøÇ?6Œ?~š5s¨°Ð®ää8Óq`ãÇoP³fqÚ°aƒî¼óNÓq€Z¡A@Z¶l¸*UUÅKOxϲe·) à)¥¦¦šŽÔ¿%‹)//W||¼:d:Še¬_¿^Ï=÷œF­îÝ»›Ž^wøpGÓ`A‡‡I: °°0ÓQ€Z£A,¦¢¢B³gÏ6ÃR²²²”••¥;３—Dƒ ÞÚ³gGƱÛ튈ˆðÈXÀ¥Ð òÖ[o)77×t Ëzÿý÷•™™©iÓ¦)$$Ätðš~ýÖ©qãr••é³Ïú›Ž‹è×ï°7¾O ,ÐgŸ}æ‘1srrhÒQ§hÐ yçw´mÛ6Ó1,kéÒ¥’¤É“'Ó °”ŽsR¬âây¨O¾SÇŽ…„ܯââ†úì³äkm¤Qr8xT ê :êÔ¢E£MG€QwðG L¬ –°nÝ:mß¾]ùùù¦£@ÒüùóÕ®];=ú裦£À‡Ð °lÙ2Í™3Çt üËôéÓÕ¥KtÔ@ƒ€:—¬fÍ*,´+99ÎtXuDƒÔc_|ñ…~÷»ßéŸÿü§é(ø/'NœÐã?®¾}ûêg?û™é8P§V¯¤ÀÀ UTšŽ ¡îàhÐ}\ee¥¦Nªàà`9N%%%)  Æ1EEEš9s¦–,Y¢¾}ûê…^P·nÝ %†/9|ø°.\h:.¢°°P .T@@ :€zoçÎî¦#À‚¨;ø#îâîã¡„„µk×N‰‰‰óþûï«C‡z÷ÝwuôèQ^ì€b݇={V³gÏÖ‰'$IƒVTT”¦L™¢àà`÷q7ß|³~øÃJ’:tè .]ºhïÞ½êÒ¥‹‘Ü0¯ªªJÇ×éÓ§MGÁw(**Ò‘#GÔ¢E …„„˜ŽƒXA÷aÙÙÙ Óu×]'I Waa¡víÚUã¸o›sIjß¾½ìv»ÂÃýš¾%??_íÛ·×SO=e: ¾CzzºÚ·o¯¥K—šŽu&**S}úlRTT¦é(°êþˆt¶{÷nµhѽ¨Ö­[+77WÑÑÑýž+VhĈ¬ÄŸÑ§Ï&Ùí9vefF™Ž‹ îàhÐ}˜Íf»à†p‡ã²Í÷š5k”œœücOž.$$D)))ìߺu«ûïGõf$€+²sgw•«¼<ÈtXuDƒnQçÎ3vôèQp‰´ŸÛ»w¯ÊËËuË-·˜Žr:Ú¿¿é€1›6õ1DÝÁñt‹Z½zµ***är¹LG‡Ì˜1CÝ»w××_m: ®ÁàÁƒ5pà@Ó1àA:räˆ~øaÓQ€£A·¨W^yE5Ò¾}ûLG€zíæ›oÖÍ7ßl:ð\â€:Õ¹sŽ6¬Pee rr:›Ž‹ îàhЯÖÌ0.IDATP§bb2Ô¬™C……v%'Ç™Ž‹ îàhÐ-®gÏžjݺµ1>nûöíúàƒtæÌÓQ,åw¿ûZ¶lIsÔÒ™3aÊÏo­3gÂLG…PwðG4訕}ûöiÆŒúì³ÏLG©÷F¥n¸»´ãŠÄÄĨ]»v¦cÔ{ÇŽÓŒ3ôñÇ›Žê!.qG­}ºvïÞMs¯xùå—•ššª>ø@¦ã\³üà:}ú´Ú·o¯üããÆúÉO~¢¢¢" ü·¯œ/–ÔÕd”Ë°KŠ¸¦Æ_ fÍ*,´+99Î3±€ï@ÝÁÑ û°³gÏjöìÙ:qℤóø‰ŠŠÒ”)Sì>î7¿ù $›Í¦   EEEéí·ßV||¼WrVTT(//OyyyúÇ?þ!›Í¦µkתmÛ¶ ÕàÁƒÕ AƒZ?£¹´´T6lÐŽ;´råJ:tHÇŽ“$8qB-[¶Ô÷¾÷½ºü‘ê܆ ´~ýzÓ1`7n”$¿Cè›o¾Qyy¹Ž;¦Ó§Oë›o¾Ñ½÷Þ«N:iøðኌŒTxxx­Æª®®Vee¥¶mÛ¦½{÷êäÉ“êÑ£‡l6›‡òòòêø§þ[WI=M‡¨3Ë– W@@¥ªªxé ï¡îà¨V–­°°0]wÝu’¤ððpj×®]ŠŽŽv·cÇ5ʽݾ}{÷ rozóÍ7õæ›o^ôk×]wÕ£G8qB=zôÐÁƒÕ²eKmß¾]•••j×®ìv»žzê)=üðÃgàÀ²Ûí*,,¬Ë¥ÎüêW¿Òk¯½ÆãÔ`DXX˜¢¢¢´víZÓQ®Ê¸qã´ôÿÛ»÷à¨êóãŸÝMR“pÄÊ%Œ¦Tl‡¤`¨€X’€·†b[À™*PhbBÇ–k±”Âp±ýUpðµh "UJEnÁH`a¹æ4ÙÝóû#²C–${6{vó~ÍdÈÙïž“çyö{Nx²{ÎyûmïòÕ«Wµ}ûvIÒ_ÿúWÍš5Kû÷ï×¥K—ôù矫E‹ºóÎ;¯¨¨(>}ZíÚµÓ®]»ëýã'€ÀËÏïìÐ1ïŠhÐ-ìÈ‘#jÙ²¥w922Rqqq*((¨Ò ÷ymÛ¶­òŸX+¸råŠ$éƒ>të‹Î8NIÒ®]».°òÁhøðáºvíW‚FД””è£>R‹-ôüóÏë÷¿ÿ}°C2ÕÌ™3«,«  @ð>öÅ_xǬ†ÝÂl6[µ •””(66ÖçóŠ‹‹«=ç»ìö÷e÷ÊnwÈãqËn·{?þj³Ùäñxäp8är¹åpT³Ûr»]r8"äñ¸e³Ù%2 Én·Éí®\ß׶Ýî[ùŠÉnw¨´Ô­¨¨"EDDxó4 Cv»]n·û›ŸëñÞŸ½îc·ÊÅîý×ã1T¹ºM†á©Õ¶ ã~•–ý&‡ëU­a]ëäO +ãõ•Ký^×ÛýÜP›k¾kèªá5wÜ4̯“¯\j~]«Ö°´Ô¡E‹îÐ_þr­ÞóÿÛŸ{«ù_û}£6cºzõª<ž?ÉáX²sÍíÞ/i`µã1À 4è–’’¢iÓ¦y—KKKUVVVí"i)))ÊÍÍõ.éÞ{ïõ¹mg¶¤;åvW.WþûÌ7_•\®›Çªªºžtó©­ß»ÕúuÙö­Æ***¿BëîGUÿhâoü­am^»º¾®õ³ò\ DLÁ®“Ë%•–V~…†hIͪ<ìÞ~Ûk¿ùºù±ËÕW‰üPwÜñ?]¿þ=íÚÕ/Øá ‘`Þ!Ñ [X·nÝäñxät:u÷Ýw«¨¨HM›6UBBB•ç¥¤¤èøñãÞ墢"uïÞÝç¶ï¾{‘¢¢Bÿb4†á‘Çcè…󪨨v85Š—Íf“Ý^ýy@C«Üo<Þ‹®YUëÖ­e·ÛCt¿©úOI*/ß'§398áAÖ±cbcËTV«0<“ żC(¢A·0‡Ã¡ÔÔT}òÉ'­/¿üR{÷îõk{?ü°RRR4räHõîÝ[ååå>ï;ž’’¢Ý»wûõ3†ôÕW_©K—.&oµ©ÉÛ³–I“¨Y³7Õ‚“‚ æB :ÂŽÍfó~ù2nÜ8-\¸P’4xð`9sFwÜq‡<¨?ü°Æõ"""d³ÙÔ¦M«mÛ¶º~ýºÆ§N:)--ÍÌt€Ñ¡CýíoÓ¾}û´cÇmÛ¶MN§S'OžT›6m”››+‡Ã¡òòò·qÏ=÷hðàÁ6l˜zè!I’ÓéÔ¶mÛ4hР׫¼ÃÝôœ€@ùöóÿ“ÔÕ„-6•”hÂv¬kûöGY¡ŠŠÈ`‡‚F„y‡PDƒŽ°•œœ¬1cÆhëÖ­*,,”$½üòËúþ÷¿¯Q£FU»:t›6m´xñbmÚ´I6›M999:wîœw<""B]ºtÑ AƒôÇ?þ±AsJRR’’’’4eÊ”jcï¾û®f̘¡ÂÂB]¾üí=½#""Ô§OõîÝ[³fͪ¶Þ€tíÚ5ýûßÿÖÞ½{µbÅ H’~ýë_«S§NK¨®’Ââ–(wèÐ}Áó¡ˆakèС:t¨öìÙ£ .H’xÛÛ6ÝXïìÙ³*))ÑñãǧèèhuëÖ­!B,)55U©©©*((ÐÅ‹UXX¨„„EGGûl²‡‡  ¨_¿~ºté’$iРA¼{ð t„½^½zÕk½øøxÅÇÇ+11¼?vÔUBB‚Ô½{÷z­ÿÀ˜@x A@@%'®¨¨r•—GéóÏ“ƒ æBŸ+L¶víÚ`‡Ö¨o`Q_ ÐÓ>öm®½zíQ¿~ªW¯=AŒ'çëjuþÍ»ÐÉá…wГ­]»VÏ<óL°Ã[Ô7°¨/Ìär¹”žž®˜˜•——köìÙ·½Hø[+©±ìcßæúÚkÏ7”€kœ¯«Õù7ïB'O„ÞA‘••¥ÄÄDeee©}ûö·¼Ê?øï Ó]¼xQóçÏ×™3g$UÞE#99YÓ¦MSLLL£Àšx˜îðáÃjÕª•š7o.Ij×®Š‹‹•““äÈ°.ÞAod®_¿.I:zôh# _—/_Ö¾}û‚FØ¢¾E}çÆq÷ÚµkAŽ¤a9rD-[¶ô.GFF*..NêÑ£G•ç†Úï¦oã¬O¼—%5–}ìÛ\ÓÒÞQtôU]»£wÞI nXQß×µü¦Ce^„ÎöoÞ}7ÏÊý=TŽSµÕØ~7…ôF&//O’4räÈ GÞ’“¹•G QßÀ¢¾•ŸŸß(îo³Ùª]®¤¤D±±±Õžº¿›êocÚÇ*s}ç››”HÏŸ×Õéçú -4bõÞUÏ3ôŽSµÓX~7…›aF°ƒ@Ã9þ¼Þÿ}uìØQÑÑÑÁëׯ+//O>ú¨ZµjìpîÀzðÁuùòeIRii©š5k¦¼¼<%$$Ty.¿› 8Ûï¦P@ƒLçv»Õ¢E }ýõ׺ûî»õõ×_+))IW®\ vhX‰¦s8JMMÕ'Ÿ|"IÚ²e‹æÌ ×6`ÞAQVV¦ŒŒ ÅÅÅÉåriæÌ™Á Kãô0e†Ö®]«®]»V+..Öøñã•™™YíÝ _cðmìر²Ûí²Ûí6l˜÷qjj—Ë¥)S¦èå—_VzzºÜnw°C ysçÎõÎÙîÝ»K¢Îþضm›î¿ÿ~xóUÏp¯µËåRff¦š7o®’’eddT{Nii©ÒÓÓÕ¡C9R‡®öÇ£eË–iݺuúøã"ô:«Íké+×]»vy÷Å_={ölÈjÍß\%iõêÕZ´h‘,X ¥K—6Tèuæo®n·[¿ûÝï4uêTÍž=Û²Wÿ®í±hÆŒJHHÐ3Ï<£cÇŽUÛ±c‡&Ož¬‰'j÷îÝ v½™‘ï… 4uêT=ÿüó r½ø›gmŽÏ ô0uêÔ)•––V; HÒo~ó ­Ñ£GW/..ÖSO=¥è駟VïÞ½(úºñ7×Í›7kÍš5Ú³gþûßÿjîܹúùÏÞ€Ôž¿¹=zTÙÙÙš8q¢&Mš¤;wzOƒ°s]²d‰***ô‡?üA/¼ð‚† fÉ?ÄÕ&ÏÍ›7«ÿþÊÎÎVaa¡¦Nê;{ö¬^|ñEÍ›7O¯¾úªž{î9ïÅ!­Èß|%éäÉ“:v옮^½ÚPa×™¿yÞîøŒ2¶Ž?nØl¶*í߿߈7<a†±sçN#11ñ¶cð-##Ãؾ}{µÇ©©9.\¸`DGG—/_6 Ã0Nœ8a´hÑÂ(++ rd¡kÅŠÆŠ+ªöì³Ï»víj˜ ëÉŒ\sss«ªáÇëàÁƒÞ1jjŽÃ‡«U«VjÞ¼¹$©]»v*..VNNN# ]Ë—/×رcÕ¿}ôÑG’¨³¿¾{ÌõUÏp¯umóûÙÏ~æý¾C‡jÚ´©Úµk'IÊÎÎÖÇ,§Ó©ÔÔT­]»V†/›cF®:uòŽy<=zT?üá úº1#×Ç{L;wîÔüùóuäÈ]¿~]iii —D-™‘ë™3g´wï^ïx›6m´~ýúˆ¾öj›gLLŒ÷ûüü|Íž=Û»|ðàÁ*§TvèÐÁ²s7#ß"""¬ÌÈÓ×ÜF`Ñ 72GŽQË–-½Ë7v´‚‚Ÿc¨YDD„N:¥cÇŽ©¢¢B={öÔÅ‹%ù®7jï»uŒŒŒT\\uôçŸ~ªüü|uêÔI>ø >LMVS=óóóþÖõÉoÓ¦M1b„bcc%I[·nU÷îÝ5tèP½øâ‹š4i’–,YðØëÊŒ\o¶gÏýô§? H¬þ2#×N:iãÆz饗4zôh­Y³FQQQ½®ÌÈõ駟ÖÆåt:U^^®œœUTT<öº¨Kž/^Tff¦–.]*ÇSã6Ú¶mkÙc™ùÞpóéLVcfž’ïcÌgÝ?ýÀ§—^z©Æ‹5 2D¿üå/o9f·Ûåp8¼ËÅÅÅ’¤&MšÔ8ÆÎX»zwîÜY7nÔðáÃõ÷¿ÿ]&L ¦&±ÙlUê(I%%%ÔÑO:tÐòåË£•+Wªk×®ÔÙD5ÍÛ&MšèÒ¥Ka]ëúì³ÙÙÙZ°`wù³Ï>ÓO<¡ÈÈH=ðÀ1b„Þ~ûm?>`qׇ¹ÞlÆ >|¸©1šÅ¬\/^¼¨É“'káÂ…zöÙgµjÕªjÛ 63r>}ºZ¶l©çž{NO>ù¤>ýôS½úê«‹¹>ê’gÓ¦M5dÈåää¨OŸ>r:Š­¶ââbËËüÉ÷ܹsjÒ¤IC…ê³óôuÌBû3öœ¼¼¼jç oܸÑøñì]>tèi¸\.cÆ 5Ž¡ö6oÞl¤¥¥†á»Þ¨½ýû÷Í›7÷.—””6›ÍÈÏÏbTáãСCFRR’qàÀê짛ÏA÷5oÃ}N×5¿7[·n­òX=ŒÅ‹{—W¯^mÜu×] Øfäz³>}ú˜£YÌÈõ?ÿù1räHÃ0*¯Ëk¬^½:`1×—Ù¯ë¿þõ/Kîãõ9¹Ýn£sçÎÆ{ï½g†a¤¦¦ .ôŽOž<Ù²çf›‘ï ™™™"ÏÛÍm˜¸72÷ß¿rss½ËEEEúÁ~ ‡Ã¡”””ÇP7III’|×µ×­[7y<ï¹ûEEEjÚ´©‚YøHJJ¢Î&óUÏp¯u]ò;|ø°ŠŠŠ4hР*ÿèG?ÒñãǽËßûÞ÷ôÈ#6ðz0#כǭxîù fäºiÓ&%''K’úõë§9sæhûöí¾ŽÌ|]¥Ê+º§§§[n¯Ï±Èn·«[·nÞÛs¦¤¤TÙW‹ŠŠtß}÷6ðz2#ßP`Vžµ™Û0 z»qžÓÍç“´nÝZÉÉÉÞkoݺUsçνíj–——§õë×Ë0 9N½ùæ›zá…$QS³8¥¦¦zoųeËî)ï‡ .èõ×_—ËåRII‰–,Y¢3fÈn·Sg?Ü8æº\.I¾çm¸Ïi_ùÍ›7ÏûŸùcÇŽiÅŠ0`€òóóõÅ_hÙ²e’¤I“&iÆ *))‘$}øᇖüè·¹Þ°aÃK^0í3ríÛ·o• ˆEEE)55µ3¹=3_׬¬,µnÝÚ’ûxmó.\¸P}úôQ•*U­»ï¾[?ÿüs¾Ç[­V½ñÆêر£Ê—/¯FiØ°a:zô¨Ó±íÚµSÓ¦MõÚk¯¹äu(Ùü‹»|ËáǵnÝ:3Fo¿ý¶KÚ¼Ò—ÂØØX7.ßú‰'*44T>ø KúSÒöKt@@€}ôQ½ôÒKÚ¾}»6lèPߦMµiÓFÔ;ï¼ãŽ®æÉ4Í"¿|ùòb»ööíÛµxñb7N.ëǵŠU¿~ýœÊÿú;¶yíµ×4vìXU­ZU}úôÑÅ‹õÍ7ßhñâÅš6mš äp|FF†î½÷^-Z´H 4ÐàÁƒõÇhÖ¬YZºt©¾ùæ5iÒÄáœQ£FièСZ²d‰zõêåº  Ä!€ð¨U«VÉ4M5nܸ¸»")g:ØÂ… õøã«lÙ²ÅÝçÑGÕ¤I“4iÒ$}úé§ySÔ/åžv¥2wzõÕWxM£|îp¥P~¹ßÿ]cÇŽU5´fÍEFFJ’âããÕ¾}{ >\}úôQhh¨ýœÙ³gkÑ¢EºõÖ[µxñbùùùÙˇ ¢ÇÜidqÀ€zöÙg5iÒ$à嘂 PRSS¨¶mÛ:”§¥¥)00P‹EŸ|ò‰CÝûï¿/‹Å¢3fا~ RçÎU¾|yÅÆÆêùçŸWVV–½í•+WÚçãÿuZIQÖ"L™2EV«U?üp‘ÞãÇë‘GQÓ¦M¢zõêéÁTRR’Ó±¿ÿþ»þùϪE‹ WHHˆš4i¢Ñ£GëìÙ³y¶¿xñb3F7Vpp°ªU«¦[n¹Å>G>;;[QQQ*W®œÒÒÒòlã±Ç“ÅbÑ‚ ŠôÚ.¡>}úè‹/¾Ð±cÇ®º+±Z­š:uªºté¢ÈÈH…„„¨~ýú=z´Nœ8át¼išúôÓOÕ­[7………©N:zâ‰'”žžîtla¦é]nãÆêÙ³§ÂÃÃuýõ×kèС:räH‘_Ó‘#GôÅ_¨OŸ>ª\¹r¾ý²Z­š0a‚Z·n­5mÚTï¿ÿ~‘¯ç.¶Ñ­§žzÊ>$©Q£F:t¨RSSõá‡:œ3eʆ¡7ÞxÃ>$éþûïWll¬~þùgmÚ´ÉáœÀÀ@ÝÿýZ³fM¾ëŽx€B)[¶¬Z·n­M›69|á]½zµ=,[¶ÌáœeË–É0 uéÒEåË—×øñãÕ·o_IR¿~ý4aÂM˜0A±±±ç½úê«zúé§U­Z5 2D—.]Ò¤I“Ô¹sgûµbbb4~üx•+WNåÊ•³·5aÂÝ~ûí…~]ëׯW@@€n¸á†<ëóš®´bÅ 5hÐ@}ô‘Ο?¯Áƒ+ @3fÌPÆ µsçN‡ã,X >ø@5kÖÔ Aƒ4pà@¥¤¤è½÷ÞS£F”œœìpü+¯¼¢¾}ûjþüùª[·®ÆŒ£îÝ»ëøñãš?¾$ÉÏÏOÿ÷ÿ§óçÏkîܹN}¼pá‚>ùäU­ZÕþž_­† *;;Ûé £«¤¥¥©mÛ¶zì±ÇtøðaõìÙSÇ× 7Ü ¸¸8íÞ½Ûéœ7ß|S÷ß¿Nž<©»îºK'NœÐ;ï¼£»ï¾;ÏkvÚÙ×_­víÚiÅŠêܹ³z÷î­;v¨S§Ny®](ÈæÍ›eµZÕ¨Q£|ûtéÒ%Ýu×]zå•WT©R%ÝrË-Ú¾}»FŽé²)Šy9zô¨¦Nª×_]_|ñEžï±Í÷ß/Ã0òüÿÊVöã?:´½sçNÕªU+Ï)]¶s~øá§:Ûñ6l(Ú ð÷b@!7Î4 Ãüæ›oìeO=õ”`vîÜÙŒŠŠ²—ggg›*T0k×®íÐF\\œi†9sæL§ö‡ b†a†……™{öì±—gdd˜=zô0 Ã0§NêpNÍš5͘˜˜<û›`†a:4Ïúììl388جS§Î•_üŸ.\¸`FGG›AAAæÚµkêV­ZeZ,³sçÎåG5333ÚúÏþc†aþãÿp(ŠŠ2Ë–-k&&&:sêÔ)ûvbb¢h¶hÑÂé¸iÓ¦™†a˜/¼ðB¡_[~>ùäÓ0 sÒ¤IyÖ_é}¾’1cƘ†a˜ƒvzŸRSSÍsçÎÙ÷mŸ‘ððpsãÆöòóçÏ›76 Ã0ýõW‡6òúŒäõ9LOO7kÖ¬iZ,ó§Ÿ~r8þ¡‡2 Ã0-‹yðàÁB½®W_}Õ4 ÃüôÓOó¬¯Y³¦i†Ù¾}{óСCöò}ûö™þþþf¥J•Ì¬¬,‡s&OžlŽ?¾Ð?_}õ•Ãù¶ßU^?ýû÷7=êÔÏàà`3 Ï×pàÀÓ0 ³Aƒö²uëÖ™†a˜]ºtÉóœ3f˜†a˜#GŽtªûùçŸMÃ0ÌGy$ÏsxÖ€(´.]ºèå—_Ö²eËÔ£GI9£­[·V¿~ý4zôhíÝ»WuêÔÑÖ­[uöìÙ|ÿ"]®]»ªnݺöýÀÀ@ýßÿýŸ–.]ªwß}×eó锞ž^¤ÛïÎœ9SÔ°aÃtã7:Ôµoß^mÛ¶ÕŠ+tðàAÕ¬YS’¦­\îá‡Ö«¯¾ª+V8”—+WN/^THHˆÓ9+V´oW©REýúõÓ¼yóô믿:ÜÎøƒ>°’\«:uêH’ÓÈŽ+ØF‚ ÃÐرcj—)S&Ïóî¼óNµhѾ_¶lYõèÑCñññŠwU+Œÿýï:tèš4i¢víÚ9Ô9RÓ¦M+R{¶÷ëJŸ¯#F(**ʾ_«V-5oÞ\7nÔ±cÇêÞyç¨Ó§Oëúë¯×èÑ£õÝwßå»ÎcĈ’r‡ÍÖ­[µqãFuïÞÝáËëÕ²})ܱcÇ5·õW»ví’ÕjU½zõtýõ×ú¼þýû;•µoß^RÞ_l Ã6 )¯Ï_ll¬"""ŠÔÞ®]»dFŸ¯àà`õéÓÇ©¼}ûö2MÓéµ$$$Èjµê';;[Ó§Ow8?<<Ü>í144T×]wž}öY­^½Z¡¡¡Ú¸q£âââŠô:]©råÊ qËg @ÉÁ€B ÐÍ7߬eË–éôéÓZ¹r¥¬V«ºté¢úõë«J•*Z¶l™yä-[¶L‹åªHË–-ÊBBBtà 7hÇŽ:qKV­zͯǶ.À,Â]œ8 I>TÎ0 ?~ܾÿÒK/i„ W—.]TµjU…††Ê0 ÅÅÅ9=Øn̘1ªS§ŽÞ}÷]½÷Þ{zï½÷¤Ûn»Mo½õ–C¨èСƒêÕ«§¹sçêí·ßV™2eìaäÑG-ôë*ˆÕjµ¿.W³ý5ß +¯/õ¥K—–”¯†m‘}^Ÿ?[ù×_]èö óùŠˆˆ°÷ûr×úZŠªL™22dˆÞ{ï=mܸÑ~S† *¨T©R:sæLžçÕjÅŠjÛ¶­}¾x—.]4gÎM:U.\°[Tß}÷Ómq—-[¦¬¬,Ýpà  TjjêU]§fÍš*Uª”öîÝ[èsl_Ò~þùçB½{÷ÊjµªÿþN£6K–,QJJJ_jÃÃÃu÷Ýwëî»ïV“&M´téR9rÄáËÛ!CôÜsÏÙG>RSSõÔSO¹lÄÂöþØ›+Õ«WO‹E»wïÖž={Š4 Ë}‘r>¯¿þºCÝÖ­[•””T¤÷Ôö~íÝ»×i½ÐÕºÖ5 ùøã%Éé®]Ý»w×Ô©SõÕW_éñÇw¨[¸p¡$©[·nö²jÕª©AƒÚ±c‡âããÚ[¸p¡ Ãp8ÇÆŸ5%cœŠ¤iÓ¦*W®œ,X ={ö8„ Ûök¯½&éêÖH9acÏž=öýŒŒ ûsFåpldd¤Ž?žçó7®Äb±èúë¯×áÇž1RÁƒ+::Z/½ô’¾ùæeff:Ô[­V­\¹Ò¾ß¨Q#EEEé»ï¾s˜N³gÏž}ú¨|ùòúñǵ{÷nµk×Îé‹N·nÝôÓO?©{÷îöçÄÆƪwïÞ…ºÖ7Þ¨mÛ¶iûöí…ºsR™2e4{ölÝqÇêÝ»·BCCÕ«W/U®\Y‡ÖÚµkuöìY]¸pARNÈyøá‡5nÜ8Õ¬YS·Þz«´jÕ*uèÐAÍš5Ó–-[ìí_¸pA·Ür‹¢££ÕºukEEEéðáÃZ³f’’’4hÐ ûûË >\~ø¡Õ·oß|５Û¿ÞÊ&>>^þþþjÞ¼y¡Û,Š—_~YëÖ­Ó¬Y³´f͵k×N+VTBB‚¾ÿþ{-^¼¸ÈkD®F`` Þ{ï=Ý}÷ÝêÖ­›n»í6U¯^]«W¯Vrr²ºté’g8ÌO‹-äçç§mÛ¶¹±×Eóì³ÏjçÎêر£ªU«¦S§Né믿Vrr²üüüôöÛo;ýP»vm½úê«;v¬Z´h¡ž={*==]K–,QFF†¦M›ætǶûï¿_K–,ÑW_}¥¦M›ªS§NÚ·oŸ~øáEFFê½÷Þ˳¶›<ftÀß# ŠÌ:*T¨à4WÛV×¼yó*ô97ß|³öìÙ£üãª_¿¾¾ýö[MŸ>];wîT×®]õÙgŸ9ÿôÓOëóÏ?W›6môÕW_)##CÏ<óŒ¾ýö[…„„8¼eË–Õ믿®n¸AkÖ¬ÑþóíÙ³GwÝu—/^¬3fäÙ§&MšØ¿8uñ¹íKß½÷ÞëTwüøq-^¼Xýû÷/R¨)Š2eÊè—_~Ñ”)ST­Z5-^¼Xÿïÿý?íÚµKÆ ³$H…ûüüU^Çç×NïÞ½µzõjuéÒE+W®Ô’%KÔ°aC­\¹RÕ«W/Òµ«U«¦þýûë›o¾Qbbb¡úu¥þ]«Áƒ«uëÖÚ¸q£>þøc-^¼X5kÖÔàÁƒ¯‘#GæyÞ³Ï>«ùóç«yóæZ²d‰~þùguíÚU?üðƒèt|`` æÏŸ¯×^{M+VÔ¬Y³tèÐ! ë°^¢¸Ì˜1ÃþÀÆ‚åïA¸T§N®ø`AÛ3*þz›X€÷#€ð¡ 55UÏ<óŒºuë¦ððpY,½øâ‹…:wþüùêß¿¿bbb¢îÝ»ëý÷ßÏóVŒ€¯ €àÔ©Súè£téÒ%Ý~ûí’ ¾oûå^ýu={VO?ý´.\¨¶mÛê­·ÞRÆ 57ðFLÁ*¤Ó§O+<<\&Lиqã®x|RR’"""ÊÖ­[§›nºI£GÖäÉ“ÝÕU Äb¤ŠšÓþ>$éÆoTxx¸\Õ-ào…âA{öìÑÉ“'Õ¸qãâî P,x¡‡\ºtI÷ß¿ÂÃÃ5zôè<9uꔾûî;EGGçûÄfŸ‹/êÀêÞ½»*UªTÜÝù["€x€ÕjÕ°aômÛ6-X°@+VÌó¸ï¾ûN÷ÝwŸ‡{€¢úä“O4hРâîÆßÄͬV«zè!Í™3G³gÏV¯^½ò=6&&FRκ$<ÍØ“ÆŒ£)S¦w7à!ü¾} ¿oßÂïÛ·øâï{×®]ºï¾ûìßÛPt72MS?ü°fÎœ©éÓ§kÀ€_ªT)IR½zõÔ¬Y3Ot±Ä ó¹×ìËø}û~ß¾…ß·oñåß·í{ŠŽEènbš¦yä͘1C}ô‘† RÜ]Š# Wðí·ß*--MçÏŸ—$íرC_~ù¥$©W¯^ Ö°aÃ4kÖ,íß¿_QQQ’¤Q£Féã?Öƒ>¨ hݺuö6ƒ‚‚Ô´iSÏ¿ ˜@®`Ĉ:x𠤜§ Ï›7OóæÍ“aJHHP5dµZeµZžòõ×_Ë0 MŸ>]Ó§Owh3::Zû÷ï÷èëJÈæ¡qqqŠ‹‹+òyÈu¥õ1ð.ü¾} ¿oßÂïÛ·ðûÆÕ0Ì¢>ân³eË5oÞ\›7oöÙ]%ß×®‹Ðx S°@‰³wï^û `O Q:uŠ»^J”½{÷ªnݺÅÝ ø¨ßÿâfP¢ØF>>ùäÕ«W¯˜{_a{Â9#oîG%R½zõXä x!¡ð!€ð!€ðÀßHtt´bbbŠ»ÀU#€ü͆qMç[,uêÔÉE½Š†àƒ®5ÄW‹Àc %ÐìÙ³Õ¶m[………©eË–zî¹ç”™™ét\JJŠÞ|óMuîÜYÕ«WWpp°êÔ©£þýûk÷îÝÇΘ1CKÎ׿•+WÊb±Ø^|ñE‡ãî¼óNÕªUKeÊ”QTT”ºvíªo¾ùƽ/>Á¿¸;GÏ?ÿ¼&Mš¤ˆˆÝu×] ÐgŸ}¦;v(++KöcwîÜ©^xA:tPŸ>}¬•+WjþüùZ´h‘–/_®›o¾Y’Ô´iS?^/¾ø¢¢££õÀØÛéر£}{ĈjØ°¡:vì¨Ê•+kãÆúå—_´|ùr½þúëzúé§=õVÀ @J;wêõ×_W5´víZU­ZURÎHG§NtìØ1EGGÛ¯_¿¾U¡B‡vâããÕ´iS=Z›6m’$5iÒDMš4±qãÆåÙ‡;v8ÝiëÂ… ª_¿¾^|ñE ªêÕ«käÈ‘1b„gÞx5@ óÊ+¯hÖ¬Yª]»¶æÏŸ¯M›6éÞ{ïÕ—_~©€€‡iUmÛ¶ÕŠ+4hÐ ­]»Vñññ8p ~ùåÕ«W/Ï)X“&MÒÃ?¬£GjòäÉ?~¼V¬X!)çv»ëׯ×رc•””¤¥K—ªcÇŽúöÛou÷Ýw{ì=€÷2̼b1ŠÅ–-[Ô¼ysmÞ¼YÍš5+îîP,ø÷Å¡°Ÿ;>Ÿ×ŽCà1Cà1Cà1Cà1C îh IDATà1Cà1C€&L˜ ‹Å¢U«VwWàå >`ÆŒ²X,š9sfžõ†aØJ‹Å¢N:w7àB’_Àxì±Ç´k×.µlÙÒÃ=º²’Špmü‹»ðÓ4ó,¯X±¢*V¬èáÞÀ1PíÝ»W÷ÝwŸ6l¨5iÒDcÆŒÑùóçŽ;tèÞyçõíÛWåË—W… T¿~}=úè£:sæŒ$©cÇŽzðÁ%IC‡•Åb±ÿ:tHRîŸ~úÉ¡}Û¨´´45JMš4QåÊ•5hÐ mØ°A’téÒ%½ôÒKj׮ʕ+§Ö­[ë‡~pzM)))zóÍ7Õ¹sgU¯^]ÁÁÁªS§Žú÷ï¯Ý»w;k›2&I+W®tèó‹/¾xUïJF@J˜Ï>ûL>ø ÒÓÓÕ°aC 2DK—.ջﾫE‹iýúõª\¹²’’’Ô¦M8qB­ZµÒý÷ß/???íß¿_Ÿ~ú©F¥ *hèС*_¾¼-Z¤~ýú)66Ö~­råÊ]±?)))êÚµ«¶mÛ¦»îºKš;w®–/_®_ýU£GÖ‚ tÛm·©U«VZ¶l™zôè¡]»v©N:övvîÜ©^xA:tPŸ>}¬•+WjþüùZ´h‘–/_®›o¾Y’Ô´iS?^/¾ø¢¢££õÀØÛéرc‘ß+” &JŒÍ›7›’ÌÍ›7wW(6¾þïabb¢j†‡‡›{öìq¨›={¶i†9tèPÓ4MsÚ´i¦aæO<áÔÎ… Ì‹/Ú÷ãââLÃ0Ì™3gæyÝñãÇ›†a˜«V­r(7 Ã4 ü÷Þ{Í3gÎØËßzë-Ó0 ³bÅŠfÇŽÍ£GÚë.\h†a¾óÎ;m%''›§OŸvºöo¿ýfúùù™Í›7wª3 ÃìÔ©Sž}.Ê{u%…ýÜùúçÓ˜‚¼Æ¿×þ[Õÿ]=ߟÎ3;_±Î3;ØÆ¿×þÛ­¯aÊ”):þ¼F­ºuë:Ô 4H5kÖÔ§Ÿ~ªììl…††J’"""œÚ V©R¥\Ò'Ã0ôä“Oª|ùòö²~ýúI’Ξ=«#F(22Ò^×½{wê»ï¾sh'44T*Tpj¿Q£FêׯŸ¶lÙR¤iS…}¯¬Vk¡Û„û1 x”Œ=4ßúr¥®<Ý()-©À6R2R®ªo…/I:yò¤&L˜àT_±bE!-3M{Oï•$5¬ÜÐi –Õ*µn-EEIÓ§GÀ7@>!þD¼LåÜw%¯éW{÷J¶Ç#|ù¥'{¾…ð ¶éW’Ô$¢‰Sý¡C¹Ûy<× à"ÜઇVWßëûjëñ­N# i™iÚ0@R $éøñbè œìÚµ«¸»ÂçÍs ŸÐ»noõ®ÛÛ¡lú¯Ó5éçIÚwfŸî³þ(©³¤œ«U²0O X„„„H’î»ï¾bî |‘íó÷!€|–išúãÌ’¤=ç~“-€dgKgÎH•*cç|X:uôûï¿ëüùóÅÝø˜Õ©S§¸»áõ ŸÕ8¢±}ûpæouÇ@Š_ïÅà2Àg5¨Ü@#çŸÂ3ÛêXîAÉGjjªžyæuëÖMááá²X,zñÅ }~ZZšFŒ¡ë¯¿^áááêÙ³§6oÞìÆŠªt@iÕ©ó—öŒr;$K–½ŽîAÉÇ©S§ôÑGéÒ¥KºýöÛ%I†aêܬ¬,uêÔIóçÏ×SO=¥iÓ¦);;[:tкuëÜÙm@Ù§aùgHöÚË à¬ÉGtt´Îž=+I:}ú´>þøãBŸ;þ|mÚ´IË—/WÇŽ%IÝ»wWãÆõòË/kÉ’%îè2à*4Žh¬y;çåìTÙ&ª'‰g€»0R¦iéø™3g*::Ú>$)((H ÐÒ¥KuòäI÷Ÿ´Ì4m<ºQ/]̳Þá¡„¹ Ñ÷ €¸ÁþýûÕ¸qc§òFÉ4M8pÀóµñØFµú¸•ÊN*«—W½ìTߪZ+õ+5Eš±\Z󔽜îAqƒ£GªR÷n ·×<ãdZΨ³Õ´ªt@i§úˆ²ŠI-è$]¬`/'€€{°Ä »X=?cÆŒQXX˜CÙ€4`À€kj|QJFŠ};4(4ÏcÊÝ ”23Y@š;w®æÎëPvîܹbê÷ €¸AµjÕtêÔ)§rÛÚjÕªxþ”)SÔ¬Y3·ô |MrF²}»\©ryc †!5l(mÙ":%]º$x¢—J¢¼þ¼eË5oÞ¼˜z䘂å×]w~ûí7§òøøx†¡èèhÏw |ÔåäJ# ‘‘RTTζiJÜ3\âC† Ñ´jÕ*{YFF†æΫž={ÚׂÜïò)Xå‚œG@ÒÓs§[Õ¨!U©’[Ç:p=¦`àÛo¿UZZšÎŸ?/IÚ±c‡¾üòKIR¯^½¬aÆiÖ¬YÚ¿¿¢þü³ÙwÜ¡–-[ªÿþš8q¢ÂÃÃõþûïëرcš5kV±½ðEÉé€9’»M÷#€`Ĉ:x𠤜…åóæÍÓ¼yód†T£F Y­VY­V‡g…øùùiÅŠzꩧôÆoèÌ™3jÕª•V­ZÅÚð°+-B¿|zT”‘»ÏBtp=H®xL\\œâââœÊK—.­©S§º£[€"xã–7ôÄO(9#Y•ËTvª·èÚ¾^¿e%HoK™e7 €¼Z­òµT«|­|ëí$v†_œ%]”ö˜t¢Ü€EèŸf çbr Ë猀@Àõ Ÿf g/% Ë ¬×#€|š-€]Ì ¨Ì¸ à³L37€T/“@#öK"€€;@>ëôiéâÅœíZá‘ ô ”$™NÁJNÎyP!Àu ŸuøpîvÍE‡EK’2‚$å<߉u àZ܆ൎ?¦¸_ãT®T95¯Ú\m¢Ú8Ô_þÂ5¤Ø*±*PZ)c´ß?]Ê ÖñãRÍšî8x1Àkí;³O/¬xA’ôÔMO]1€üë®Ï%I'J/d唳\‹)X¯•œ‘lß uªÿk±©R%w›®Ex­”Œûv¹Råœê ày€×JN¿l$¨àêÕs·#"r·Y„®Ex­ËG@ UªHAA¹åŒ€€û@^+%ó²)XAŽS°23¥ÄÄœí¨(Çó*WÎÝ&€€k@^« )XGæ< ]r\ÿ!IRÅŠ9Ûp-Àk•+UN1a1ª\Aa¥Âêò[€nS©ZŠä—©¤¤Ü ¸v€×šØy¢öÞ¯ÓÏœVŠuêò SÖMQ¥7*iÏå¤?ëÂ)5ÕC@ø¤üH _ N_<³– ‰iXàJ€OÊ/€Ä„Åäî”ß/‰®Dø¤|HùË# àj€O²  )<<·<:,:wçÏ)X<Œ\‡ð9¦™@jÔ #·®”)E†Dæì0.GøœääÜ;[åu ^û:²IR@\ˆð9WzˆÃ:°p!À+%œMÐ ÿ¹A­?n­×~~Í¡îJ䡦iVßOdLÿE:[‹5 àBþÅÝÜáôÅÓÚsz$©Ed ‡º+Ñ$IÏdJdzX®ÄÀ+¥d¤Ø·CƒBê®@lªTÉùoR’dµº²wà» ¯ty)TΡ®¨äÒ%éìYWö|à•’Ó“íÛ€T¯ž¹Û¬× €¼RaF@*U’J—ο ÛˆÄ:pÀ+%gä=’•%=š³]Ðô+‰î@x¥ü¡;–» œžÇmx^é–Z·(8 XÉéÉŠ‹¶—vú¡äCúÕ\.uLþè¡ãÇot_gÀ‡@^é–ënÑ-×ÝâT^زîÈ:½ñûP©£¤¬`%%@À˜‚ð)… 1a1¹;a LÁ!€|JaH­òµrwÊ@ÀU ŸrøpîvA¤Bp…†äì”ßO!€|Šm$ ÀñAƒe†bÊÿ9 «Ü!<­ìl÷÷¼àSl$*J²\á_Aû4,¿K2ËÕÉ“îíøÀg¤¤HçÎål_é  ÑÀ¸ /Àë\ʾ¤­Ç·ª\©rª\QKW”Tøõ6µÊ×RY³ŠRä\;ÀëœH;¡V·’$ÝQïÍï?_Ráï€e3²åHÓc/æì@àÚ1 àu’3’íÛ¡A¡öí¢Ã0T¥Jî~R’+z¾ð:))öírAåìÛ—¨¨Âµuya®àu’Ó]3"@ÀÕ ¯ãÊËŸB€kGxËHHPˆ}Û@Ê——BBþzVÞÊ–•Ê”ÉÙf \;Àë¤g¥Û·ƒýƒ%IÙÙÒ‘#9e…~ec›†Å\;ÀëddgØ·ƒüƒ$儇¬¬œ²« ¦Î¦d*#ãJG Âs@^çÁ¦ªW^JÏJWr9iãj KÒ¶ãÛô[»R‡ÒÆ‘JJz£È‹ð:a¥ÂV*Ì¡ìjHÙÀ²:_jçŸ '()©è#(€\LÁø„« 5ÊÕaûç2ìë@à@>áðáÜí¢¿•÷«ž³S>׈ð W;"IUƒ£s6JŸÖÁÄó.ëø"À'؈ŸŸTµjÑÎ.cßÞ{*Á…½ßCø[©^='„EÝðÜrø<®àõÒÒ¤Ó§s¶¯æVõ«EÛ·§pIŸÀWq^€×‰û5NûÏíW_þÑæ:|8Ø^w5¤CÌM þém]<#k@ ö|àuælŸ£÷ÿ(IÝz´Ãô¨¨¢·W§bÕ<ö¤vï–N•qQ'ÀG1 àuÒ³ÒíÛAþA×t,›*Urþ›–&¥¦^CçÀÇ@^'#+þ` pi‘¤¤¤«ì€ð>Ù9¤”)†áòÂÃàê@^Ç6ä$éÚBh‘»M€«GxÛH¿c •Ê•»º6× €¼Žmz_¬Véðáœò«ýX®Âmx^§~x}U*]IUÊVщRffNùµ°ð‹RíURX‚Ö«"év—ô| àu– ^fßÞ°!·üZHé°é¾’¤mÉ=D€«Ã¬deeiìرjÔ¨‘Ê—/¯.]ºè»ï¾+Ô¹ëׯ×?ÿùOÅÆÆ***JÆ Ó—_~)Ó4ÝÜkÀålÓ¯¤k õjT–.å}ºxàIR=”˜˜¨ýë_ºýöü‡Þ?ùäY­VÍ™3GµjÕ’$ÝyçÚ¶m›âââ4qâDO¼€\@$©tFŒÎk—Lÿt%ž?®ÈЪ×Ö ø F@ò1sæL•.]Z÷ÜsCù!C´sçNmÚ´)ßs«T©¢ÀÀ@………9”W¬XQU«òx’+H9EÛ··9pm€"€äcÿþýª]»¶‚ƒƒÊ5j$IJHÈþïàÁƒ¢Ç\›7oÖôÖ[oiݺu5j”[û pd ‹ymmUˆ±oo;Ä:¸LÁÊÇÑ£Gqùcoÿn¯ÏO:u´}ûvµiÓF-[¶”$è›o¾Q×®]ÝÓa@žl$2R ¸¶¶¢ÊÆhËŸÛ{’ p5 n°yófÝsÏ=ªV­šž~úiU©RE .Tÿþý5qâD >¼ÀóÇŒã4}kÀ€0`€;» ^çâEéĉœí¨¨ko/¦B´”ê'%×Pzxе7 D›;w®æÎëPvîܹbê÷ €ä£ZµjJLLt*?yò¤½>?o¼ñ†NŸ>­øøxû®~ýúiÀ€zî¹çôàƒ*((ÿ¸¦L™¢fÍš]ã+ß´ëä.õšÓKAþAê9PÒ¿$]ûúIj+ N—¬þj1ùÚÛP²åõà-[¶¨yóæÅÔ#ïÀ|Ô®][ûöíSzzºCy||¼$Ùïn•—ï¿ÿ^]ºtqZ?Ò§O¥¤¤híÚµ®ï0@’t>ó¼Î%h÷©Ý:øç$תUü$kÎßööÀ@ò1dÈ]¸pAŸþ¹½Ì4MÍœ9S 6,0ùÖ¬YS?þø£.\¸àP¾`Á†¡(WÌä)#+þžš;ÚìŠR¥Jî6®S°òѲeKõë×O#GŽÔÙ³gU·n]}úé§úå—_ôÅ_Ø6l˜fÍš¥ýû÷ÛƒÅ3Ï<£ûî»O]»vÕÀU¹re}õÕWZ´h‘î¸ã]wÝuÅõ²Àëedç çÝ@’’®½=ðEÌ›7OãÆÓ´iÓtäÈ5kÖL_ýµºuëf?ÆjµÊjµ:<wàÀª]»¶¾þúkÅÅÅ)))I={öÔ—_~©[o½µ8^ øŒËG@RÏ•²o»"€”/Ÿs'­K—€«E)€ŸŸŸ&NœXà“ËãââçTÞªU+µjÕJ/½ô’;»ø‹ô¬Üµ{)g];b±H•+KG@àj±àU.Ÿ‚•|:'€”)“3zá ¶iX'NHÙÙ®i| àU.Ÿ‚uödά5$ÃpMû¶bµJ§N™ pBx•æ‘Í5©Ë$=Ýr¼²µäšéW6•«\’w•F]§¾óoq]Ãà#Xð*#«qDcmÙ"½ùçód]@ªFH—¶IeNiò%×5 >‚€W:t(wÛ•¤JIçb$I§2(3;Óu€ €¼’{H´$É”©Cɇ <àˆðJn gcìûÎp]ãà ¯ä®!û,IJ8›àºÆÀ@^)11w;2Òuí^>K’Î@ ( ¯”ššó߀©T)×µ"]d \-nà ð*œùC)):ë_J²ÔQéÒ.mß0¤*¥jêà/O«Ì¥h=þ`S—¶ÞŽð*c—Õ¼ó¤Þ’öT™2.\ò§ÈÊ¥tð‡7”&©y„Ë›¯Æ,€WÉÈÎÈÝÉ R™2®¿FÄe¡ãÄ ×·ÞŒð*éYé¹;ÙA*]ÚõרR%wûøq×·ÞŒð*Yî¹<€$%¹¾}ðf€Wq˜‚•íþ àUì# Ù’iqˬË×€@ h ¯bÉ ’$F@ „!€¼Š}z¶HÄ6mN[¨ÙÛf»þ"à¥xÀ«üôÀOúeÃÝ3 [’Ü;k@_­ ;¨ÝKËëþ&÷»þBà…x•j¡ÕTAu¤S7HrÏHp°*élŒ$élúY%§'»þBà… ¯sáBî¶;F@¤?§a‹¶ï8wÀ=/Cx´´ÜmwŒ€H¶cßO8—àž €—!€¼ŽÇÈÙËÈYàu<1+"B# ’¸çBàe ¯ã¹)XÑö}F@ p¸ /Àëx,€œ”²èï/?‹Ÿ{.^†€×8sñŒþñý?ôCöóRý/%¹ù.X¦Ezû˜žµ¦já= Ýs!ð2Œ€¼Æé §õïµÿ–JIºa´ó.÷Ž€HÒÅŠ:‘äžk€7bà52²3rw²JIró"ô??îžk€7"€¼FFÖå$H’ûÖ€T®œ»M€Â#€¼†ÃH¶{H@€T©RÎ6 ðŽ# î‚%å®IJ’LÓ}×oBxô¬ôܬ ùûKmHzº”’â¾ë€7!€¼Æ_§`¹kú•ýNXbà5BƒBÕ2²¥N7–ÎWuëô+é²Òå9ÝùÍMŠy'FVÓêÞ‹ÀßÏxÎ1µáá ªPA:{V*S۽׫^ýψxíHY+I:žz\‘!‘î½0ü1ð:.äü×ÝS°îºK ’t6Æ^–p6Á½€¿9À«dgK.q÷¬êÕ¥‘#%­e/ÛwvŸ{/ s€WIKËÝv÷ˆ$ýóŸR©´ºöýõüîþ‹ÀßàUlÓ¯$÷€HRx¸ôйäëµ(àU<="IFGKÙ9÷u9”ö»âã=s]ø;"€¼JqŠåîÿç: {õü <òCxOOÁ²Ý~˜B¶üKúú-þ:[k×zîÚðwBxg|VýVFK]/UÜã±Iz¾ã3z»÷KÒo÷IV+™ „€Àk$¥&))ã TéwÉ0=@$é¤Ú>üpåJé‡<{}ø; €¼FFvFîNVG§`IR@€ôò˹ûŒ‚€3Àkdd]@²ƒ<>"IýûK±±9Û›7K x¾P’@^#=+=w'«xˆÅ"Mœ˜»ÿ RV–çû%à5¦`e{~ –MÒÍ7çlïÞ-}òIñôJ"Àk8LÁÊ*U,#¢o)l IDAT ’dÒ¤I¹ûãÇKù¾„ðö«E²ú{|ä|Æy­;²N³¶ÍRåz{Ô£GNù¡CÒx¶/PR@^cLë1jòº´"çVTžY°kÚLk£!_ Ñ·|ë°dâD)5Õ³ý€’ˆðƒRÏH«ÇJò|©[±®}û÷Ó¿«iÓœ»bIÒ‰Ò;ïx¶?P@^åÂ…ÜmOOÁº<€ì=³WRÎsAüürÊÞ|S:sƳ}€’†ð*ii¹Ûž©Xº¢*W”3"Iuëæú(Çç¹çžsŸïÖ­›£)».®{1vpç ȨHô%ÖÅâp8Xº{)›m¢\X9\.‡Ã}½^=X±ÂlD¸sgÞ¶,‘@§ H!&OžLTT7ÜpCžódíÚµ,[¶¬Ðç¾òÊ+ñØcù:L9!<$œ¨ìZp¸¤V±|zÃ؆$g&³'yOžkZ KDÊ2% …HLL$>>žÈÈÈ<ç›7oÀ–"~c,_¾œ:uê0þ|:vìHtt4Mš4áå—_&''ǧq‹ˆ”e))ž±Õs@Unäo<¸1Ï5MD‘²L H!víÚE•*Uò¯Zµªûza6oÞLRR·Ür íÛ·gÚ´i4mÚ”#Fpÿý÷û,f‘²Î. ˆ6#‘²Ls@| ;;›ääd^yåžxâ ®¾újúöíËøñã>|8µk×.ôùÆ #&&&Ϲ0`ÀŸÆ-"bw©©ž±Õ-XÞ È?‡þÉsM-X"ö0mÚ4¦M›–çÜ‘#G,Š&p()DÍš5IJJÊw~ÿþýîëE=wãÆôéÓ'Ïùk¯½–Ù³gó×_™€Œ=š6mÚœaä""e—]* uêxƪ€ˆø¯‚>^±bmÛ¶µ(¢À ¬BÄÇdzyófÒÓÓóœ_³f õë×/ô¹ 6ÄårzÝ{%)9Þ«ZjäO@Ê•ƒ½J@D¤ÌQRˆ’ššÊçŸî>çr¹˜}ú0tèP>L£FøôÓOY¸p!Ó§Ow?nРAL™2…ÄÄDw[Õ5×\CçÎyñÅÙ²e Ý»wç“O>á›o¾á©§ž"..ΪKD$`}½ák~‰˜ÝCaù]DG7°:$^èöB¡×êÖ…¥KÁå‚; õኈ” % E˜1c#FŒà£>bçδiÓ†9sæУG÷cœN'N§3OË•ÃáàÇä_ÿúßÿ=S§N¥Q£FŒ7ŽÁƒ[ñ£ˆˆ¼ß¶ÿƺ˜· 3°±·_$ E9y%,% "RV()Bpp0#GŽdäÈ‘…>fÒ¤ILš4)ßù°°0FåËðDDÄKfŽW“3Ôò¬SÑJX"RViˆˆˆ„,g–ç 'ÔòI觢ÍE¤¬R"""!+Ç“€8\¡„…YL1h3B)«”€ˆˆH@ð®€D†…áï+ž{ï¢,)K”€ˆˆH@ð®€DE„ZIñDFBª€ˆHY¢DDD‚w$*Ü£éG™·eï,~‡5{×ä¹–Û†µ{7ddXœˆˆ”€ˆˆH@ˆÇ±«#$µ&:,ÒêpܾÚð—L¹„¿{ï6}—çš÷DômÛJ7.«(‘€ðâÅ#qMø>XAÅЪV‡ã–P-Á=þkÿ_y®i%,)‹”€ˆˆH@HMõŒýiÆUä0¿nÿÚ—7ÑJX"R)‘€’âûÓ ‘¡‘ÄÇÆ°vÿZrœ9îkÚŒPDÊ"% ""¼+ þ”€4¯Ö€ôìt'ºÏ«KDÊ"% ""¼+ þÔ‚'ÍñjÃòÞ D ˆˆ”J@DD$ øk äM@Öìó,Å5j˜±Z°D¤¬P"""Á_'¡ƒI@8hP©!y®å¶aíÝ ii¥›ˆHi ±:‘’àÏF•qü_ljËX½z°h‘oÛ—rp""¥L ÿÝ >†tð»$ÈT`òZ KDÊU@DD$ ÌÚw‚#ÇïZ°Š¢•°D¤¬QDDDB¶3Ó rÂü®RmF("e Ù®,3p†Ú*Q –ˆ”5J@DD$ ää& 9¡¶jÁª]3VDDÊ% ""r°g$, jÕ2c% "R(‘€ƒgˆ¿W@R³RÉvf»sÛ°öï‡ädkb)-J@DDÄörœ9àp8ðß ÈûKß'þíxʽ\Žå»—»Ï{ÏÙ¶­ôã)MJ@DD$ ´Ü=¾–Þç· H¶3›Í‡7ãÂÅ_ûþrŸ×JX"R–h±½à `jìxˆÕ'v÷׬„j îñ_û= ˆV‘²D ))ž±¿V@ò$ û N@T‘@§DDDBjªù3(Âí¥0Õ¢«Q5ª*€Z°D¤ÌR"""!·åÙWÃåVAö$ïá@êÀ,Ãl®«KD ¹ ˆ¿¶_å*¨ +$D{ˆHÙ¡DDDBn –ð´a:ÇŽ•vT""¥G«`‰ˆH@ðnÁògÝêvãíËß&¡Z­Îiå>òDô-J=4‘R¡DDDlïhÚqÒª®†œPB*×j[R¡VnHÃÊ óW""e…±½;ÿ†;»p`çCÀhk:Z KDÊ ÍÛ;ž’凅„ZÉ™Óf„"RV(Û;žêI@ QDD™±½ä´L÷8<$ÌÂHÎ\Íšf9^P""M ˆˆˆØ^rš§jÏ Hp0œ{®«KD™±½ï$Ìž xÚ°Ž…#G, EDÄg”€ˆˆˆí¥¤Û¯²t×R^[ø×ϸž”L³‰‰æˆHY DDDl/%Ý3$2Ìs@ƯÏ“?>ÉŒµ3Xµgw)^µa‰H R"""¶×"ø:µFo¡KÅ[¬§XÚUoç/Û½ PDDÊmD(""¶—• ÇkP¹œÅÁS»^ H’);TÛKIñŒ££­‹ãt$TK ,Ø´‹åV@Ô‚%"e±½ÔTÏ8*ʺ8NGxH8-âZ°áÀŽg§zuÈC¯ ˆˆ*% ""b{v¬€€§ Ë…‹•{Vuê˜k[·‚Ëe]l""¾¢DDDlÏ»b§¤mõ¶îñÉmXÇáCVD%"â[š„.""¶ç]±K @ûí9¯òy´«ÑŽ&Ušù'¢W®lIh"">£DDDlÏ®-X-ÏiÉúû×ç9wrÒ¶-""E ˆˆˆØ^bøL¸üp†r0ç~Î¥®Õ!1­„%"N ˆˆˆØÞÞÈùÐé’]׃í""N“ÐEDÄö2³³ÜãèˆP #9{J@D$Ð)ÛËrz°{' çœf¬, DJ@DDÄö¼Ð {' ‡ö‘À¦DDDl/˙采YIÉÈmÃJM…, ED¤Ä)ÛËqyU@‚íY9œv˜$ñp¢V‘€¦DDDlÍé<)±a ÖÌu3‰}-–Ë>¹ŒÏþúLÑE$ )[KOž;Χü±vD„DXÒikVµ™{¼l÷2% "дˆˆˆØZJ ðÃkt½*Ú/ÿ a冔+ÏñÌã,Û½Œ§.ô\S –ˆU@DDÄÖRR<ãèhëâ8AŽ ÚToÀŽc;ˆ®¶×}M 4J@DDÄÖRS=c»& íj´s·f.'*êÄx«5ñˆˆøŠ±5ï Hî›v;òN@–'yæh/ 4J@DDÄÖ¡ ò& ÞÑÓÓaïÞ‚Ÿ#"bGJ@DDÄÖ¼[°ì\iP©Ã+âÀÁñÌãZ KD–VÁ[ ” ˆÃáà×;~¥^L=ʇ—çýžk[¶@§NÖÅ&"R’”€ˆˆˆ­Ê$t€q-ÜcU@D$P©KDDl-%¸§%^±bmÛ¶µ(¢À ¬BÄÇdzyófÒÓÓóœ_³f õë×/ðy‰‰‰¤§§óàƒëþZ´hëÖ­£R¥J >Üçñ‹ˆ”))æϨ(kã(IQ¡Q´«Ñgåõ½íÛ!'ÇâÀDDJ€B 8ÔÔT>ÿüs÷9—ËÅäÉ“IHH(4ómݺ5óçÏÏó5oÞC‡-­CD$àå& ¶ÞG—s½Ú°Î]@v6ìÞm]<""%E-X…hß¾=}úôaèС>|˜Fñ駟²páB¦OŸî~Ü Aƒ˜2e ‰‰‰Ô®]›Š+Òµk×|÷«X±"ÙÙÙ^‘3—Û‚h H×:]y}ÑëæàÜ°îZ¶lÚµ­KDäl©R„3fðÐCñÑGqóÍ7³gÏæÌ™CŸ>}Üq:8ÎSîóáp8´ºˆH s¹òÎ $ÏíÌkÝ_ãárÀ¯Z KDƒ* EfäÈ‘Œ9²ÐÇLš4‰I“&ò^óæÍ+ÉÐDDHK3I^$&"†Ç/|œÙà-§9§DD±­ÔT âÄ­&³Z;ŽÖ¢vÅÀêQªWÏ3޴ɺ8DDJŠZ°DDĶRR€sVÁ³¨é¼·ô=«C*qAȉ —/·6‘’ DDDl+5Êr‡‡ZŒDD@Ë–f¼n;fm<""gK ˆˆˆØVJ ì•€^СƒùÓåRDDìO ˆˆˆØ–I@2ÝÇaÁaÖãC¹ À’%ÖÅ!"R”€ˆˆˆmåkÁ ð (ûÓ*X""b[ùZ°pHVNs!ø–_ÉI+Ç’ÅS­IDä¬(Û* -X!A!Œ^ü&9ñIQž³²IJ ¡zu«#93jÁÛ* -X‡ƒ.uº˜ƒðã·š¥K­IDäl(ÛJIþ¼ÞHâý†Û¹¾ÙõV‡ä]Îíâ9¨ó›æˆˆ­)ÛJI²#!ùêÄÔ¦|xy«Cò‰®uºzÎU""ö¦DDDl+5Õ3ŽŠ²._K¨–@LDŒ9¨óK–ºp:­IDäL)ÛJIñŒ££­‹ÃׂAt>·³9ˆÞÏÑð5lÚdmL""gJ ˆˆˆØ–w$€Ëê_æ9ˆÿNmX"b[J@DDĶ¼+ Ü‚pEÃ+èR¹Ìþþ¼E ˆˆØ–öÛ*+-Xñ±ñ̾ù b0ÇJ@DÄ®TÛ*K-X•*A£Ff¼r%dfýx¤DDDl+%Hø z=ÀÓ¿>ÂÎc;­ÉçÚ·7ffš5ÖÆ""r&”€ˆˆˆm¥¤õ~†Žï2zñ[J;duH>סƒg¬6,±#% ""b[©©@P–û84(Ôº`J‰±;% ""b[))@°Wø H«Vrb % "bGJ@DDĶÊb$"Z¶4ãuëàØ1kã9]J@DDÄ–\®Ü ˆg)¨°à0ë*Eu/X —% ""bKî, :Úº8¬Ò*ÚÓ†5yáÿ,ŒDDäô([rW@€¨(ëâ°ÊÕM{¹ÇóvhˆˆØ‡±%ï¤,V@úuiÇj°•ù¤f¥žâ""þ!ÄêDDDÎDYoÁª]ÛA¹ƒIÎ9HdÒåý[¿ÒEÄT[*ë-X\ôüïRV]Éö-g· ãúõðÈ#°tiÉÄ'"R% ""bKe½Сƒg¼dÉ™ßÇé„>}à­·àúëOlð("â#J@DDÄ–Êz }{Ïøl*_ 6˜ñÖ­”tVa‰ˆI ˆˆˆØ’;y !wÿs}|‘¥ñX¡];Ïøl* o¼‘÷xõê3¿—ˆÈ©([r·`•ÛÑì½H=`iïðaøðC3ŽŒ„{ï…–-ͱËý廘E¤lS"""¶”šŠ»ý Ên VDĉÄáïëÝçfü=ã”Ï?ÞSå¸ã¨\Ù$3¹4DD|E ˆˆˆØRJ ä•€”Ñ,8±ÈÚþîãék§ùøÌLxûm3v8àá‡ÍØ;Ñ<ñ% ""bK))@V¡ß¿Ç螣ÒfˆÕ!Y¦CàH]Øi&„¬Ú³Š7úøÏ>ƒÝ»Íøšk >ÞŒ[´ðV‡d÷Dôb´a¹\y7|ì1ϸbE¨WÏŒÿüÓ3A]D¤$)[Ê¿emþ qc(WX{û\amX?þèÙ/¤cG¸à‚¼×s'¢§¤ÀæÍ>VDÊ<% ""bK¹ Ht´µqøƒà`h×8ZvtàϽ²ùPþ Â{ãÁÇ3s@¼iˆˆøš±—ëD J@r¹Û°–ÜÏ€sžgí}kiÛ ÏcÖ¬¹s͸^=èÛ7ÿ}r+  y "â!V ""rº22<óÔ‚e¸57S4©šÿ1o¾é?ü°©œœLñ5U@DDÄv¼wéVÄp' À’%ù¯ïÞ Ÿ~jÆ11fï‚Ô©c&£ƒ* "âJ@DDÄvrÛ¯@ H®Zµ .ÎŒ—.5mjÞÞ}²Nl›rï½'&­Àáð´aíÜ ú&^)»”€ˆˆˆí¸+ ‘‡8;Û°ãèKc²šÃá©‚9›6y®%'Ãرf ÷ß_ô½¼ç¨ KDJš±w¤æ¾¯q ]&uᣕY“?(¬ kÒ$“”Ü|3Ô¨Qô}¼ç¨ KDJš±w$(Ë}.4(Ôš`üHA HN¼õ–çü#œú>šˆ."¾¤DDDlÇ€{% ÁJ@ÚµóŒsY³`Ëá­”EϞмù©ïÓ´©g…,U@D¤¤)Ûq·`gºÏ…‡YŒ‰…† ÍxåJømË žß ªæðè£Å»OD4ibÆëÖAffÑ9J@DDÄvÔ‚U¸Ü6¬Œ ˜:ó(G«~å/O÷îÅ¿OîDô¬,X»¶ä㑲K ˆˆˆØŽZ° ç=dòˆËàH’ϙ˶£[‹}Í_Q"""¶£¬Âµo痢Áò!¸pÖJaÞKñjˆˆ”$% ""b;jÁ*\«VâubÕaf”O\9‘lgv±î£½@DÄW”€ˆˆˆí¸C˜Ùi»ÙEŸÆ},É_DFB‹žãhg zÕ¿ €ÝÇwóí?ßë>ÕªAõêf¼jUþÕEDΔ±w Vv5+U¥FùD‡E[“?ñž2x0 ít—ûxüòñžOî<ÇaçÎ’ŠNDÊ:% ""b;î e]þêæ›MVÕªðè£Ð£Aέx.¡A¡”/OŽ3§X÷Ñ<ñ…S?DDDÄ¿¸+ @´ ùtî »v™v¬òå‚™Öoñ±ñT‹®Vìûœ¼ÖUW•x¨"R)ÛQäÔª”g\Pû‚Ó¾‡* "â jÁÛñN@Tñ† M”€ˆHÉQ"""¶ãÝ‚¥ ˆï{VÔÚ¼Ž·6 J@DDÄvr+ ¤ßd>å݆õçŸÖÅ!"C/ÛEÈÎÎføðá4oÞœJ•*q饗2wîÜS>ïË/¿äú믧^½z”/_žž={2vìX’’’J!j‘À—›€·ÿ{¿¹—G¿”ý)û­ *@<]Däl))Bÿþý3f ƒ bêÔ©ÄÅÅÑ»wofÍšUäó^}õU>Ìã?άY³èܹ3o¼ñ lݺµt‚ `¹-X®z?0nÙ8ÞüýMR³R‹~’œMD‘’¦U° ±xñbfÏžÍĉ¹ýöÛèÕ«III<óÌ3ôíÛ·Ðç~ýõ×ÄÅŹ»wïÎe—]Æ\À˜1cxë­·|¾ˆH@Ë­€…{’Ž¨PM9näµ…¯Ñ±fG†´Rèãš7‡Ãì„® ˆˆ”U@ 1yòd¢¢¢¸á†òœ8p k×®eÙ²e…>×;ùÈÕ©S'ªV­Ê–-[J¶|yhÐÀŒ×¬œâía("R(% …HLL$>>žÈȼ¿Ðš7opډĆ Ø¿?-r—‘3’™éyìóT@"C”€Wݘº\Vÿ2À$#Ÿÿõy‘Ï’–ÿüãëèD$Ð))Ä®]»¨R¥J¾óU«Vu_/®¬¬,n½õVªV­ÊC=Tb1Šˆ”EÞ{€¸BL$<8œà `‹"²§uþ—{üÊÂWpºœ…>Vó@D¤$iˆ9N ÄêÕ«™9s&•+W>ås† FLLLžs `À€¾ SDÄ6¼÷q…˜µ_¾®uºr~­óù}çïüµï/¾ýç[z7ê]àcO^ ëÆK)H‹M›6iÓ¦å9wäÈ‹¢ J@ Q³fÍ—ÍÝ¿¿ûú©8NÌÔ©Sùä“O¸òÊ+‹õ½GM›6mN/`‘2»â 2M@?}‡ƒ§:?Å5Ÿ]Àü‡+^‰ÃáÈ÷XU@¤¬*èà+Vжm[‹" jÁ*D||<›7o&===Ïù5kÖP¿~ý"Ÿïr¹2d“'Oæ£>RõBD¤„x' Õ³/ k®t¬ÙѺ€l¬w£Þ4«Ú €E;ñÛöß |\­ZkÆZ KDΖB 8ÔÔT>ÿÜ31Ïår1yòdŠÌ|].wß}7ü1&L`àÀ¥²ˆH™àÝ‚Õ'û3~¹ýfÞ0Óº€l,ÈÄ“>é>þ!ñ‡çpxª II°woiD'"J-X…hß¾=}úôaèС>|˜Fñ駟²páB¦OŸî~Ü Aƒ˜2e ‰‰‰Ô®]€|?ü;３fÍšñǸNëÖ­Kýç Þ(u^µndÑÎE n=˜¶5 ÿp­U+˜7ÏŒW¯†=J)@ 8J@Š0cÆ FŒÁG}ÄÎ;iÓ¦ sæÌ¡‡×«®ÓéÄétâr¹ÜçæÌ™ƒÃá`âĉLœ81Ï=ëÖ­Kbbb©ý ""Æ;‰Ž¶.Ž@ÊØ+Çžòq'ODW""gJ H‚ƒƒ9r$#GŽ,ô1“&MbÒ¤IyÎi³AßñnÁRRz4]DJŠæ€ˆˆˆ­¨ËMš@h¨k"ºˆœ % ""b+ª€X#, š65ãõëá¤E"EDŠM ˆˆˆØŠ* ÖÉ’“ÿmm,"b_J@DDÄV4 ½ô8]Î<Çš""%A ˆˆˆØŠ»«î|®ZTžj¯Wãµ…¯YS 9žqœgæ=C·ÉÝò$!'¯„%"r&´ –ˆˆØŠ»–LjN2©©ÉddgXS é7½Ÿ{S©k¦rK‹[U@D¤d¨"""¶âN@BÒÜç¢B5¤$=~Áãîñ🆓–eþ®ccáÄž»¬^ ^[`‰ˆ›±w V¨g9¬ÈÐHk‚ P—5¸Œ^ñ½Øqloýñ–ûZnäØ1غՂàDÄö”€ˆˆˆ­¸+ ¡ª€øÒë—½NüMøÏ‚ÿ°7y/ y "rö”€ˆˆˆ­X Q¤¤5«ÖŒ!m†œ™Ìs¿<hˆˆœ=% ""b+¹U@|íù‹Ÿ§\X9Æ/ÏÚýkU‘³¦DDDlÅ€Diˆ¯Å•‹ã© Ÿ õ9­ÉÈΠ~}(grU@DäŒh^±•Ü¬Š»®åíçë’–Fã*-)=rþ#4ˆmÀõÍ®wÏ iÑ-2“Иkc{Q"""¶’[‰Ik˶m­ ¦ ˆ äÆ„óœkÕÊ$ þ ]»Z˜ˆØ–Z°DDÄ6\.Omm,e™&¢‹ÈÙPDDÄæÒ³ÓY¶{Ëv/#¡ZÝëw·:$ŸÉÊ‚œ3VbMD‘³¡DDĦ~Ýö+#æà÷¿“™“é>wÛ»y³ç›¹2”{ *ð~<ÛHH€  p:U‘Ó§,ùëHºMîÆ/Û~É“||°ünÿïíÖæcÞ ˆ* Ö‰Š‚FÌøï¿MeJD¤¸”€ˆˆØÐñÌã8]NêÅÔãŽVwðL×gˆ ""$‚g/zÖâ}#Õ³ò®‹5k•ÝŸ$£Î6l°:±µ`‰ˆØЋÝ^dÑŽE\Vÿ2†wNpP075¿‰¿öýE³jÍ,ŽÐ7Ô‚åö¥ìã×&]ÁµŽý‹VüMB‚Öâ‘âQ""bC¡Á¡ü<ðgB‚ò¾Œ7®Ò8 ÷Ä𮀬ð?lvR!¼ku´.¨2¨jTUêT¨Ïþ£ ÂnÞZûw1Ñê°DÄ&Ô‚%"bS''ewdnÔ@zü_úMïg]@e”Ãá`Òµ@FyÖGNbʲ/,ŽJDìB ˆˆˆØ†w’ãHÈÕ¾ì áÜÚtM{Ó}|×7w°nÿ: #»P""⧲ٸ\.«Ãð+Þ-XY˜ƒÈÐH‹¢‘qw ‚?o ƒd®ýüZŽg·8*ñwJ@DDüP¶3›¿¸‘as‡¹W»¯ ˆÃIéD†(±J“&zå|{›°þàzžû°ÅQ‰ˆ¿S""âgœ.'ƒ¾Ä—ë¾äíÅoóàÿ<«ûË8Æ»KÞe×±]%¡uÜ HHºûœZ°¬õäÃÑðùLH¯Häá¶<ÝåßV‡$"~®ìÍ`ñc.—‹û¿½Ÿ)«§ÆU®:ãûÍÙ8‡_ 93†vZR¡ZÂÝ‚êéÅR –µºv…¶õâY>ùgÒö7eë5ÔëfuT"âÏTñ.—‹'|’±ËÆìæ³~ŸÑ3¾'GŽÀ„ °tiñïY-ºÉ™É,Ù½¤Äc.mž Hšûœ* Ör8à±Ç€¤6ÁoX‘ˆø;% ""~â¥__âõE¯àÀÁä>“éÛ¤/.̘MšÀ]wA‡æÏÇO}Ï–q- `É.û' V@4Är×]çžkÆß~ k×Zˆø7% ""~àíÅo3bþ÷ñ¸Þ㸹ÅÍlßW_ ×_{öx?a‚IH¦O‡¢Ê  §Õ9­X`=GÓúêG(î ÈÁF,¹&•ƒOäí^o[“@H æ9~óÍÂ+"¢DDÄ4«ÚŒèÐhFõÅ Vw1f 4m sæx×­”7{¿±w/ÜpƒIP¶o/üÞjvp—í^æ‹ðKg±å#‰Œ%&"ÆÊä„Aƒ B3þäóïSD¤ J@DDüÀ¥õ/åû[¿ç•K_áÒ¨G8ÿ|ó‰rîîêÕá‹/য়L{ËÕW{ž;gŽITÆŒœœü÷öN@ìÞ†å½Ht´uqH~*ÀÝw›qf&¼÷žçÚ”ÕSx~þógtßàÃ!)©‚¿ DDÄO´ª|‡ç¬é1¼É.® /4 Ë /˜LBÓ£‡IpêF¶¤\X¹ú)¬“[Qõõj—\bÆ›6Á×_ó=Ëýî Ç•Ãõ3®ç§ÄŸ ½ÇW_Á‹/šqPœsŽÏŸË—û0x)J@DDJÉ‹Þ éûMÙth³g›ö’ƒ͵˜³´îüùиñ™Ý?<žyþüÓìNë“O`¨½7@wËM@4ÿÿ=ö˜güÆàp8sùnj~9\1õ ¾\ûe¾çnÜ·Þê9~õUxî9ϱª "ö§DD¤Œüu$ÿð8»ï¦ÛÇ—pïÃGÜ×ú÷7s96Ÿöž­ó΃yóLBSîDÑcúôâm\èïr[°œç}ÉÀÿäÞoîeí~ízço.¿Ü¬Ì°p!üñ‡iüøš¹æ¼kÈÌÉ$#'#Ïó’“¡o_8vÌ÷ïoZºn» ªT1ç¦O/zÙiñJ@DD|(Û™ÍÐo‡òïyÿvŸ«¹ç.’¶˜Yå½zÁçŸ{ZLJJPIh6ÇYYfÕ,»Ë­€äÄ-gÊê)Œ[6Ž¤ãZŸÕß8yç‚äV-BƒCùâú/¸½ÕíŒî9Ú]³¡æwzvQoÚ&N4÷ŠŒôTñrrÌ’Ó"b_J@DD|äHú®œz%ï/}ß}îþF¯³ä “ŒDE™ ä‡ïb¸ÉóþŽ©S}÷}JCVdg›qpxšû|T¨&„ø£›o†¸83ž9Ó³ÈBHP¯žÈCÊóøQ£`Æ 3®PfÍòTðî»""ÌxÂ8z±)% "">°ùÐfÎÿè|¾ßü=¡A¡|Ôûcæ¿ü.—yÌ /@ݺ¾£];ˆ7ãùóa×.ß~?_ò^‚7(³#¡ÿ<`ÆN'Œí¹æ8)ëþùg³tt®O>FòÞ¯Z5ÓŠpü¸IBDÄž”€ˆˆøÀ»KßeýõTŽ¬ÌO·ýľò×_æzëÖðÐCEÜ „8ž*ˆËeúçÍØåûo^¼G¨§iA4R÷ÜcÚ§À´S4iûv¸áÏ&ƒÏ<W_]ðý~Ø33ÆTÅDÄ~”€ˆˆøÀ«Ý_å¢:ѤJ^Lõ¬.<ÿ¼¹d>½ÍÝ1Ú× 81?Æóÿ\IüÛñ\;ýÚÒùæ%(ÕSôÀ¦ ˆT® wÜaÆ))ðÁy¯§§C¿~pà€9îÕ ž}6ÜÈ/[Éw¿Æ᪫ÌxçNOB-"ö¢DDÄ‚ØyÃL~ô;õ+5àž{Ì›-0•¶mK/–Æ¡M ³G+ÍgóáÍü½ïïÒ  „xW@\!šb?ì™çôöÛ™iÆ.—™X¾l™9®_þïÿ ÙÊuÓ¯ãÒ)—òÚÂ×òUëNžÜnÃbžH™§DDÄGb#c©Q‘ÿû?øéÄžkçžkæ~”¶W4õ‰‡ÉÌÉ,ý@ÎBÞÄS‰ Q –?‹‡>}Ì8) ¦M3ã L[˜6­™3!6Þ_ú>kö­!ǕÓ?>IßÏûr$ݳlu×®fn˜Í7çÏ/½ŸEDJ†:p oßúûïç]Ù§´Üxã‰O¡ž˜Ý¨'–~ gÁ»ëÜ NôŠïÅEu.""$º ¤X¼7&5 ~ÿî¿ßsnÂhÙÒŒîô0ÿîêY¶zö†Ù´ߎU{Vù—ø}ã _F."¾ DDä ìMÞ˰‘Qäã}Ô³Ûùõ×ÕW–Bp¨UëÄîèÎsŸÛp`ƒ5Áœ!ï HïˆÿðíÍß2ÿöùùVTÿsÁЩ“¯Y=zx&?ôY²7WpP0/v{‘onú†J•Ø|x3&tàùùÏ“™“Éu×™j"À·ßzö{P""r’ÌLøâ X½:ÿ5—ËŬu³h1®cá韟.ô>?ýS¦˜qÅŠÖožvÓM¸+ €{•.»ð®€DG[‡œï*Hr²ù³Kxýõ‚EÃ+Xq÷ ÚV7¦²œY<÷ËsLZ9‰6ÌóØ7ßôQÐ"âJ@DD¼?—_ýûC«Vp～ Ï'Ò{Zo®~-ûRö0uÍÔ<ýé¹ÒÒàî»=ǯ½Vò»Ÿ®~ý äHc÷ñzW@”€ØOŸ>f¢y®5Ì*V¡¡…?§nL]Þ¹á]†ì¦Mõ6 j3€ÁƒMbfß={|¼ˆ”(% ""'> —]óæyÎÒ0îEš½ßŒoÿùÖ}­w£Þ¬ºg11ùîõÒK°y³wîlÞ,Y­reèÞÚ³»Û²-öM@¢´ð•í›%vÁL:ÿâ‹â%åá!ጼd$ËîZÆä>“ 2ëW—/wÝe“™ ï½ç£ÀE¤Ä)öíƒnÝ`ñbssâSö ;ÙÓ·ŸíAz¶YG·fùšÌè?ƒ¯nüŠjÑÕòÝkÍSñóéîøñfïpÛåà×áðÍ»ÔK|ÅêpN‹Z°ìï¶Û`éR3gãüóOï¹­ÎiEBµ„<ç|гŸÎرyÿˆˆÿò“_‰""ÖÙµ .ºÈ3ç£Z5øåó&éŠ.5 -Ö\pºô1†…®£ïy×8ùÙé4ŸÊfg›ãý š4)¥¤®¾¢þ K‡²àÓ.î=ì@-X¡];¨[·dîU«–Yá Ìb/|¸Œ½É{Kææ"â3J@D¤LKL4aן˜]«üö´haVÙ™óu¯tK讋aÜJ²¾yÇ,OçΦÒq²qãà?Ìø¼óLâO¢£={2> sçZÏéP –Ľ$op£¶ß@ü;ñ¼ôëK¤f©"⯔€ˆH™µnI>¶l1Ç ˜ä£‘gš‘äÌdž™÷ ñoÇóê‚W9œvØêEä$J@D¤LZµÊ싱۹š|IÓ¦ð믅·†Tªdærüò‹©l€i³zùeS-ùùgxೊÀ A¦­Ëõèa&¤|õ•gIT简TÜA›Ù!”ÿOyå³ûÅr=üu#,½œÁ$%'ñÔOOQë­Z<ð¿Øth“µAŠˆ›±Lv¶y\Ð~¾ôË .¼çS\sÜÓ†à>w3÷§tjÔ8õs»v5ñ>û¬gùÐM›Ì'°³f™ãjÕ<“ÐýQh(\w§¦šÿvàN@BÒpºœ$g&“í̶4&ñ={B³ºqðÍXxÿ/.Ž»fŽVjV*ï.y—+§^‰Ë¥Ê™ˆ?P""–HO‡¾}áškL ÅÝwÑüÛi”¨íG·sóÄá\<»6©½nÚ¿~Ÿö|^ìû„‡ÃsÏ™*Ê…æ¿>f ÄÆ–PÐ>bÇ6,w V¨§·?24Òš`į8ðÈ#'4¦ÊO_²áþ m?”¨P3aèÁ¸p„ˆ”>% "Rê’“áÊ+aÎϹñãÍjQ3f@IHy4ý(}>ëC½Ñõ˜ºã?½ß}­i•Æ^9–~Mûö}s۶ƃ ̹>}à†J*rßéÜÙL¸3ýÀkã)w$4Í}.÷Í¥ÈÍ7C\œÏœ !Çòîï²ãá¼ÚýUnou»¥ñ‰ˆ‡)UGŽ˜9?ÿlŽ££¡\93Þ³®¿Þ,»cGÉ}Ï áX±uNœæDN5ÝÀ7ýÊ_÷ýÉ=íî¡\X¹3ºwP©Þ$&šŸé‹/̧±þnwòNÚßò\ð:Ù‘»ùâ «#:µÜ HH¤W$D1ÂÃÍ<,0ËamƱ‘±ŽTDB¬@DÊŽýûMò±j•9Ž‰ÿýjÖ„¡Cáë¯Íù9s`þ|9Òœ>ó':Ø9s(ty –ßÍ5µ†0cRu÷Ž’P¹²ÙÈÐ.Æ/Ϭˆ¡pð<¦M»š{î±:ª¢åV@ÂË¥’;óCñvÏ=faˆÔTøè#³¢]q?xùЛìÊYÍðŸ‡Ó±fG$ àúf×S½|uß]„Œ øî;Sö^OÄˆH©Øµ ºw÷ì·Qµ*üð´liŽgÏ6m÷ßo*!ÉÉðÐCðÿ&xç-Ç™ÃoÛc溙8»¤&œWù<ÏAå üú«©:Õ®m]L§’›€„E§‘Û¥ ˆx«\î¸Þ{Ïü{¹ÿþb>1f 󬆱x×bïZÌÃsæâºÓ·q_ú5íGòÅX©¢„üö›ÙÔtýz³pÄ矛ys"@-X"âs[¶äÝì¯fM3wÂ;©p8 _?³7ÇÝw{Î/] mÛšý5RS]l<¸‘ñËÇ3àËTUn“»ñÎ’w˜¼d&&ÀW˜äfÀ˜>Ýk‰Ùœpz ”ñã•|4©êµ={å|ö™EÁSn VX´&¡KáyÄ´vž–#õ`t"üø2ìiá>íÂż­óxð»Yœ¸®d-,”#æ5°kWÏkfVôïo> —Ö¤ó+V¬ mÛ¶,_¾œ6mÚXŽH‰X¿ÞT>ví2ÇõëÃ?B½zE?oÁóéߺÜßù=&¸åtr¢vü„ÄK`ÊOùNÇÆÂUW™„¤G{ÌÏ( ÙÎlª¼V…£G!µ2¼¾—V-ƒY¹ÒêÈ fÞˆ5½0‘§ÇþAjV*Ö¾0o2%üó,^|z Zdd˜y\ß|ÇÂ×BÂg0 *o‚¬‚^?L·.\{­Y½¯fͼÏw¹\gµÊ–Ëeæ=ø ©犋ƒ½{ÍØေcó~H#¥Oï×ΞZ°DÊ€ìl3¯¢BS‰(ɹEYµÊ¼éßbÑ©&MLÛÕÉ¿¸ Ò¹3¬\ ¯¾jæ‚d–Û“?ùÈ(›{Àº~°±·ût­Zf5ª¾}ͧˆ!z¥Ë'$(„Kë_ÊÌu3!ê T_ÁªUíY·Îüwò7YYæ  õ¹©y}k¿Ö°¡ù:]ƒ›DdÞ¼¦Ìšõ³>žýŽ?¡Êœ™üôüô“™›Ö±£yéÛ×ÌÏèûy_'Ò®F;ÚVoKÛmiײXUº;à¾ûò® X®œyí»÷^6 Þß$)÷Üc6<}ì±ÓÿùDü…~-‹¸+`Èó'˜½{÷6¿4{ö„(ÍáýãèÕ˳·GëÖf¹×ªUÍ'…{’÷°jÏ*VîYɪ=«x÷Šw©]-Ï=ÂÃaij2ÖUÏucSæ×°½3l½¶tƒ¤6à4ÙTãÆž7íÚ©ÒQ=ô4 @üw°»=Ó¦Á /XWAR=]W§ß^#rÂÃáòËÍ×ûï;øý÷–ÌšÕ’Y©¦4×âÅæ멧 ISÛûÿAŠc/kö­aÒªI;‚iV­m«·e`Ë\T÷¢<ß+'ÇÌWyúi¯vQLÕö½÷3o¢Jß|Ÿþ1s%ξ÷~ï^ó‹iÚ4ϹðpS yì13÷#w¢å¬YféÛ}' 99äëo¾ê*Ó2u0{»³ÿæ¸sÇœ{9š“ÄœDä$²?;‘zay´ò<öï7¿Xs{õ¯ºÊ¬DIã*ù{ÿß”ßXžN—w¢Õ9­Ü_*o‘ûªUM‚#%cù]ˉ db( šhÎMZ² HIüÿíÞ% þÎ_Ï}¡AóºúØcfÒøìÙ0kV$?ÿÜ‘¬]È“”œ³šîo&Ы‡™Û¶y³ç>MšÀøñð}Ö?Ìûu»“w1ëü|߯õ9­Yq÷ † 3íXC†˜ªÈøñ& ùøc 9»‰ðg¢8ÿ½].Sñٴɼ¶œ{n)'~K Hú÷ïÏ÷ßÏÈ‘#9ï¼óøä“OèÝ»7Ó§O§o‹qgggÓ­[7¶mÛÆK/½D\\ï½÷]t?þø#:u*ÅŸÂ÷\.X»+Ëm›IDATÖóIPî\0+åN¾à‚Â?¡ñ—_Xæ xî›óÜÉÓAAf2sß¾æç9›O—Ë$3gšïó÷ßžkççi)jß¾øeu— &M2¿öœ¿ä7Î3Óår‘åH¡Qǽܞ°—žíå¿ö‘²ô:¾›Y¹Àþf.úº=Wè÷߸/‘»‡ç=wà ðÉ'ž ï/]ò¡A¡´:§÷Þr/_ÝúUñ~8ñ©Ü ²×^kªf™™f¿Q£JîÕ’N@|5oIJ†¿¼ž—¦sÎ1+SÝ}7=j*ͳf™jtJJ$ììD f•«\aafîÇ“OšŠ¾ŸU¢ªp õ@ßÃ{ÓÍAƒL"~ë­f‘‘©SÍÿ#º7#%+™êå«S½\uÏŸåªsN¹sh]½µ{¿¤’RØïœX´ÈóÞ`ëVϵ6mÌï¹k¯5 ˜ZÈÊ% …X¼x1³gÏfâĉÜ~ûíôêÕ‹¤¤$žyæ™"/¿ü’eË–ñóÏ?sñÅгgOZ´hÁ‹/¾È7ß|S ?o9°d‰ç…åŸ ~Ü–-fÂ[oAµjpõÕæç’KÌ ®?8~<ï¦u¹íIÞœN³3÷üùfs¼¶m=‰Bq^݉ùó=÷‰5¥ùÛnƒÃé‡hö~Žgç@êÒ²Óò}ß&ðÞëò矞Ähͯ¤Äpv8®‡â`þ"|ûmôiܧè¿$±TL \y¥ùo¿gù·~é¥VGåánÁ ;ΑrkŽV†#R¨ŠᦛÌWZšYv|Ö,øê+8xÐ<¦Køàƒ¼UﺽÀ Ý^àHú6ÚÄ?ÿáŸCÿ˜ñ¡hW£]žïsã&ïßß|p0{¶‹ [q§±ãØŽc›pÕ·\hì+’V0ò·‘ÄDĘ¯ðÏ8"†Š¹ ö„üöÑûC¼¯¾òTØó}Ÿæë™gN¬ æõÁ[v¨+”€bòäÉDEEqà 7ä9?pà@î¼óN–-[F»ví }nݺuÝÉ@xx8 àÅ_dÿþýT­ZÕ—áûDVüò‹yƒ:{¶™Z6m̾+V˜71ÙÙæü¾}ðá‡æ«|yóf§o_³RRi۷ϼ8Κe~9dfæLd”‹n½÷R£V?üœÉ¶Yœ AY,ß›Éò Yüûå¦4ªU¥ÀÏŒ øèë¿™´pk×g‘š‘iž^4Í„à,ÀÅù? kWóIÑ‚žuëwí2« ¼w¬',ü¡ÐŸ%hõ8ç{ªj7ßl’j'” gíþµEÿ}¤ìËÓßüüó¦E`áBóßo[fGV¦ BP‚«Q!8Ž*Áõˆ ®AP½?pwóGÆf ]}¢e?7Ýä™Ü=uª% ))@øQ¸år> ù“;¶ý®uºZ–H‘"#M+êUW™×ÒE‹LRrÙe…¿ÑŽ‰ˆ¡]vùŽ‚\}µ©¶\s ¤f¦ãÜ߈ÐJId‡ïÇEþ5†b"bŠ¼ß–Ã[<+ã"ýéô< Hv¶©šº?Äë2|ý"!+²#qdGR%Æ|%¯½€3‡ºŸ¿q£YÝëÕW¡F óÁ[õ‹æкe(á¡Á„…äûªSŠ 1Û™M¶3› ‚Á¥Þ–&§¦¤‰‰‰ÄÇÇ™wýîæÍ›°eË–BÄÄDZ´h‘ï|óæÍq¹\lݺµÈä©©ŸóË|̬.\8O¼¸hчꡅ/÷³3k ËÒ>Çåò<×sóbÔ¿âëEþ쿦Œg[ær÷÷uº\ìÝëbÇ™YNØÕv{^<‚‚̧9¹íIµÏuÒFÊ·qÑó6I{\ìNr±o¿§ÓüÇÎgŸ]ÈgŸ™JHÅŠðöÛžÖŠõ?óõ±ç½"÷ú9Nlöô¯ª¿ùs|~äaÖf|Óë¹99.’S\¤gd›}#þ÷NžçTªd~Q\{-\tI•Þ¬n.öáýg³Ø¸¾ûųfMó áÐ!óK!¹Þjè÷ÒuìfшÜÇ&F™E7¼;CLɦn]Ónuò¥¨Ð(*„W :4šØÈXªEW#®\qÑqfGëê­óÝ·Aóe´>ñ%ìÊ+͇ÇÃ60uã<²F‡˜H†;ÊÑ!êÆ"ïµ2í¿$;M+ÉÊÄí óažëµB[P/¬ð‰&éÎã,MûÜ}¼e pÛPsYÀm³ncã +öÏ'b¥ÓÊ[Òºw‡ï¿‡+®ˆäظUdmÚeqãà}ÍIâhNGœ»9ž³õ¿´àÃE…ßkAÊ‘"¿W(|ò±i_HM5ß÷»ïL«™[…fóF/.`ÿ‰¯7¹øÏ›CùïÍš xVjܽÞ— Õ®‚"6ž¿/v­# ¯ªÿ‘ú9¾Å}D0A„ì!ˆBŒª¾§ÐçÌ8ú8›3„yN¹ ÊÜSyPxƒŸBìÚµ‹¸¸üm'¹‰Ã®Üm ynç΋|nûöíó]O?±lÐ[Þ„ä|—˜ù½¶¤| þ÷pÙȯßpŠÞÜîÓ¡Aþ¥É]¢5y7!!çÓ©“i¥êÚÕ¼qS^>pÀÅÌŸ øÅë}tDù‹HÇ$w°oßzÈkòHý%pÙ¯E†9døŠ"¯Ó} 4(â“ÿ¬õÀ ªV…nÝÌW›6žMë6otARÑߢnÍõlÛ°"OÕbìX¯$ï,ò9ä°lù2‚žÂÚ´1_O>i>-›°²[W´Ã™YÑi’’¬(ÈŠÂq´>·ÞöÿíÝHT]ÇñÏUw³Ò¥X#Êj]3 Ó Bú%ÔZ< E?@Ø ¥§@©þ ²$(ÿjû#ÍÄ£RD‚„‰ ö-»I»Ä®Ù¦YXij¥=éì¦e£V:žë÷ Ñ™¹ù½NÇs?3×sîëçŸ{Þm»?À·¥lsÙçñºöµ^ëõÐ'ëE---º?P¡°Ý† RÉ‹Ô±4Cÿ=È“Zçèü¥o,°ýˆ4ëÓD»”ÿ¯}ýÿû_¥_‡˜‚¤„¯Žñ‘Ô(Mµf(cM†jþQ3t °ãÛ¼©S{VÂJNîù”û÷zZÏ  þÔ¤«j—4Äkã³Tò/•~ß&MiûâÏviJ›~³º´Ïãµméû7¥ÀßuªùÅÔmuª³«ÓãK¼ |«W¾¯uëzÞ¼lnî¹»¢¬¬gá—z÷͹÷ožHÿâ<ýGÚôù¯ÝêR·ºôQ=õ¼ÿí½ö9¿ñô/UÒŸ¾x£óíµÏÕ{LÏ϶÷ï=oiÆwrc@‘‘‘þ'Ož¸-Ërgffzl`` ;))É£¿¼¼ÜmY–ûÊ•+çr¹z> Ñh4F£éær¹†¡9Éñ È ‚ƒƒÕØèÁ›>-‹<Ä& ÁÁÁzùÒs‹o»e˹\.…„„xÜúûuttèñãÇßµ-FDXX˜îܹ£ŽŽùûû÷õ?ø´4Phhè Ç.\¸Pÿì·Eêçc-ËRHHÈ€Ç)!!ad…`T­]»ÖîÆ5;„ÃáлwïôË/Ÿ Òív+//O‘‘‘Cn(èp8TWW§òòò¾¾ÎÎN]ºtI[·n—+`ÞÀ' ƒˆŽŽV\\œ’““ÕÜܬððp¨²²R………}ÏKJJR~~¾jkk5þ|IÒöíÛ­]»v)==]³fÍÒ¹sçôìÙ3åççÛuJ€íødEEE:tè²³³• çÏŸ«¤¤DqqŸ—~ëîîVww·Ü½Ë IòõõUYY™vìØ¡S§N)))I>>>*//תUlžõ#²²²äãã£ÀÀ@»KÁ(¸uë–‡-Z¤€€­[·N§OŸÖã/·dǸóñãG9rDQQQš9s¦bbbtãÆ »ËÂ(aOnÌÓËýå•30†444héÒ¥ Pkk«Z[[í. ^¶sçN=}úT{öìQDD„jjj”››«G©¢¢bÀåª1öÅÇÇëæÍ›JOO×âÅ‹uñâE©°°Pñññv—/cO^ÌÓ.ƬŸ~úI~~~š1c†Š‹‹ÕÖÖfwIð²ÆÆFÍ™3§__}}½BCCµmÛ6]éÝ–ãÆÝ»wµfÍ]¸pA{÷îíë߸q£šššTSÃÞ ãxòbžÆpq Æ$—Ë¥ŠŠ ={Vdä‰ëë‹Iš7ož–/_ÎíãT^^ž¦M›¦Ý»w÷ëw8zøð¡îÝ»gSe-Œãɉy#AÁ˜óâÅ >|Xš;w®ÝåÀ°ææf=xð@Ë–-³» Cmm­ÂÂÂ<ö2ŠŠŠ’$.H' ÆñÄÆ<‘"€`ÌINNVDD„8`w)°Abb¢$)55ÕæJ0 òèï]~¼¡¡ÁtI°ãxbcžÆH@0jnß¾-Ÿïj½7«¤¤DçÏŸ·¹zü¨á¼Þ_KMMÕÕ«W•™™©ððpÃgÀÇó4¼}@0j–,Y¢¬¬¬ïzî‚ ÔÞÞ®””æix cFgg§ªªªdYV_ŸÛíVFF†ÊËËuýúu)""ÂÆ*ámÇ—Óé”ÓéTZZšÝå`„ª««µjÕ*åääÈápHêÇ›6mÒ«W¯½ýããxr`ž†·p ÆŒ)S¦hÆ ý999òõõÕúõëm¨ £éÌ™3r:ŠUll¬ªªªú=¾zõj›*ÃpEGG+..NÉÉÉjnnVxx¸ TYY©ÂÂB»ËÃ(`OÌÓðÆ<˲ú½Û‚‰ãÚµk²,K¥¥¥*--í÷˜eYêê격2ŒDQQ‘œN§²³³U__¯+V¨¤¤D›7o¶»4ŒÆ1˜§ñ£¸ €1,à Àc Œ!€0†Àc Œ!€0†Àc Œ!€0†Àc Œ!€0†Àc Œ!€0†Àc Œ!€0†Àc Œ!€0†Àc Œ!€0†Àc Œ!€0†Àc Œ!€0†Àcþ¸Ÿ«n÷IEND®B`‚HDDM-0.2/hddm/examples/plots/a.png0000664000102500010250000013031611726176127017466 0ustar wieckiwiecki00000000000000‰PNG  IHDRèXñÌ%sBIT|dˆ pHYsaa¨?§i IDATxœìÝy˜Õ½ÿñÏ ¨€‚ø3ê$hŒ ^TB@ŒAcDÅ-.É ŠŠ‰D£Çå&ê Šq‰·+^ãEeÔ$B4F7ŒFA6 ²o3óû£rèêšSkWwWϼ_ÏãÓ3Õµœî®ÁþÔ÷œSUÍÍÍÍeU]î:™@@ èd€  t2€€@ÐÈ:@@ èd€  t2€€@ÐÈ:@@ èd€  t2€€@ÐÈ:@@ èd€  t2€€@ÐÈ:@@ èd€  t2€€@ÐÈ:@@ èd€  t2€€@ÐÈ:@@ èd€  ô¹ùæ›5`Àí¼óΪ©©ÑرcµlÙ²Àmžxâ }ôÑÚ}÷Ýõå/Y£GÖœ9sZ¬÷ÜsÏiàÀÚyç5hÐ =ÿüóÅzhöìÙ:ñÄÕ£GUWWkÊ”)ÛžÛºu«~ö³ŸéàƒVçÎuØa‡éÖ[oÕ'Ÿ|RÆPè%2{öl]|ñÅš3gŽn¹åÍœ9S\pAà6/¼ð‚Ž9æMŸ>]O>ù¤6oÞ¬N8A6lضμyó4lØ0uëÖMO?ý´öØcwÜqzë­·Šý’mÔúõëUSS£»ï¾[’TUUµí¹uëÖiÞ¼yºæšk4wî\]qÅzøá‡5f̘r5€ŠQÕÜÜÜ\îF´Eï¼óŽúôé£ùóçkÿý÷´ÍÖ­[Õ«W/]sÍ5:ÿüó%I#GŽÔ‚ ôòË/o[oРAêÓ§~ûÛߥíÕÕÕšø $'œŸ}öÙºãŽ;ôµ¯}mÛ6:uR—.]$I—\r‰¨ÓN;Mµµµ7nœêëë™ÁP4ëÖ­ÓÂ… ·ý¾xñbÕ××k×]wU÷îÝuúé§kÞ¼yzú駵eË–mÿÿÚu×]µÝvÛ•«Ùd³¸—È„ ´zõj 2DÝ»wßößc=¶m††}ðÁÛ~ÿŸÿù555颋.ÊÛ¦¶¶vÛ:}ûöÕ´iÓ´|ùrp jhhÐôéÓÕ§OŸ’¾>@ÛñÚk¯©_¿~êׯŸªªªtÙe—©_¿~ºöÚkµ|ùrM:U+V¬Ð!‡²íÿ]=zôÐ+¯¼Rî¦iTÐKÄÜV-Èĉó~Ÿ1cF¤}uÔQzõÕWµ €¸Ž8âˆÀÿ¯Eùh‰ :ð/Ç/w2‹÷ÆŽ÷Åï ÚªÖrî·–×!µž×ÂëÈžÖôZUÍÍÍÍånJgîܹêß¿¿^ýõV3©TþýõÇ{åÁ¿¿ÙC€  t2€€@ÐÈ:@@ èd€  t2€€@ÐÈ:@@ èd€  t2€€@ÐÈ:@@ èd€  t2€€@ÐÈ:@@ èd€  t2€€@ÐÈ:@@ èd€  t2 }¹´f .Ôš5k ÚGçεï¾û¦Ô"YE@ŠdáÂ…Úo¿ýRÙׂ é@+G@Š$W9HRï„{™/iDÁUxÙG@Š®·¤~ån€Œc’82€€@ÐÈ:@@ èd€  ô¹ùæ›5`Àí¼óΪ©©ÑرcµlÙ²ÀmtÆghÿý÷W»vít饗¶XçPuuuÞíÚµÓæÍ›‹õREÐ¾Ü h+fÏž­‹/¾X Ðû￯›o¾Y¯¾úª¦OŸî»Í¦M›´ûî»k̘17nœªªª¬ëuéÒE ,È[¶ýöÛ§Ú~€¶háÂ…Z³fMâíçÏŸŸbk´vôqñÞ½{«gÏžêÓ§Þ}÷]í¿ÿþÖmzõê¥ñãÇK’î¿ÿ~ß}WUUi÷ÝwO·Á@+³v­Ô¹³ôÿ'~z¹[¨ .Ô~ûíWîfhCèeÒÔÔ$Ijll,x_k×®Õ^{í¥]vÙEgœq††®=÷ܳàý­É§Ÿ:t@T¹ÊùC’z'ÜË4IcÒi€V€^MMMª­­ÕСCuà´¯Þ½{ëÁÔA¤úúzMœ8Q·Ür‹V®\©vØ!¥­Gss¹[¨<½%õK¸-]ÜDG@/ƒÑ£GkÉ’%zñÅ Þ×À5pà@IÒA¤ÓO?]Ý»w×O<¡áǼ µð™ÂÈ z‰=ZÓ¦MÓ¬Y³Ô­[·Ô÷ß±cG :T?þx`@¯­­U×®]ó– >œPV :J¡®®NuuuyËV­ZU¦Ö€JA@/¡Ñ£GkÊ”)š5k–zõêU”clܸQÏ=÷œÎ<óÌÀõƯ~ý’vÕ±]ðœ;w®ú÷ï_¦€J@@/‘ /¼PuuuzꩧԩS'544H’ºvíª:H’®ºê*­\¹R>øà¶íêëë%9“”|üñǪ¯¯×öÛo¿mìúu×]§C=Tûì³þþ÷¿ëþûï׆ ôãÿ¸Ä¯Pz‰L˜0AUUU2dHÞòx@gŸ}¶$©¡¡A|ðAÞó¦Ê]UU¥¹sçjÒ¤IÚk¯½´xñbIÒ_|¡ýèGúè£tÐAé;ßùŽ~ýë_sKÀÃŒA§‹;²Š€^"æ¶jA&Nœ{»qãÆiܸq‰Û´L€¬«.wóê«ÒG•»ÅE@dÞ AÒa‡•»¤Ù³gëÄOT=T]]­)S¦´Xç7¿ù<ð@íºë®:þøã[ ßvt”ÔÂ…Ò矗»h˃^¹-*w IëׯWMMî¾ûnIÎ)nwÝu—®¼òJx≚4i’Þÿ}~øáÚ²eK9š @Ea :Jj¿ýœÿÞ}·Ü-A[Ãt Ç{¬Ž=öXësÍÍÍ7nœ.¿ürÝtÓM’¤Ã?\]ºtÑ£>ª#F”²©T*è(¹ ÊÝ@1|ôÑGZºt©jjj¶-ëÔ©“öÝw_½üòËel•€R±bÅ IRß¾}ó–÷íÛwÛsÀ]Ü´)ŒAJ¯¹¹¹ÅXu›ÚÚZuíÚ5oÙðáÃ5|øðb5 ÚŒºº:ÕÕÕå-[µjU™Z?tmcÐâëÑ£‡$©¾¾^ûî»ï¶åõõõ:ꨣB·?~¼úõëW´ö@[f»à9wî\õïß¿L-‚ ]Ü´)TÐâÙc=´÷Þ{kÞ¼yÛ–­_¿^ï½÷ž\Æ–P¨ €ÈÖ­[§… nû}ñâŪ¯¯×®»îªž={ê²Ë.Óå—_®ææfqÄúéOªž={ê?þã?ÊØj*DöÚk¯é[ßú–$çè—]v™$iäÈ‘úßÿý_]tÑEÚ¼y³î»ï>Ý{ï½éÚ :ÐJÐÅYð«_åfª—¤‰¥Ë/—~8ù>éâÚ :¡ ™]|6v¼/Åå÷þnÝê<66&ß·ßE§?ýIš2%ù~²†û ­td‘ î…\ ñÛö¸ã ß7@–ÐK1¾d#¥ƒ¾y³ôç?KÇ_šã¥¥ÏÃ_üBª®–~þóÒï?pº¦Çy¯üºŸ›}rá¨?3(†ùóç´}çεï¾û¦ÔÅ@@G"I¾0ó%»4Jõ>ÿâÒ7Jï½'í³OiŽ™†J<¯½Öy,U@Ÿ81þ6~ïktz…€3ûæˆ# ÞÓ‚ é@†ÐQ2•Œ*I©»¸¯\é$©wÂ}Ì—4BkÖ¬I§IŠ‚€ŽD²vÚªRwqO{fíÉ“¥pþkkšš¤Ûo' û¡‚FoIýÊÝEÄ,î%róÍ7kÀ€ÚyçUSS£±cÇjÙ²eÛ444èŒ3ÎÐþûï¯víÚéÒK/µ®÷ÜsÏiàÀÚyç5hÐ =ÿüóÅx #Ô—F©zZN9E:ðÀt÷iÃyXÌ_Ž€^"³gÏÖÅ_¬9sæè–[nÑÌ™3uÁn³iÓ&í¾ûî3fŒ9äUYJ•óæÍÓ°aÃÔ­[7=ýôÓÚc=tÜqÇé­·Þ*ÖKIŒ/ÙÅUê.îŸ7u)ÎÎÃÒ¢‚]ÜKdúôéÛ~îÝ»·zöì©>}úèÝwßÕþûïoݦW¯^ÿ¯>¯÷ß¿uÛo¿]_ûÚ×4yòdIÒ7¾ñ 4H·ß~»~ûÛߦü*r˜$.{*½‹{©p–V1o³ÐÚPA/“¦•„ ÚÏœ9sTSS“·¬¦¦F/¿ürAûEåªä z1òʃ€½ šššT[[«¡C‡êÀÝ®\¹R}ûöÍ[vÈ!‡hÅŠí×_²³‹ z4œ‡Å±v­}9]Ü¢£‹{Œ=ZK–,Ñ‹/¾X¶6ÔÖÖªk×®yˆ®áÇGÚž.î…[±BúÇ?¤£ŽJgôh8‹ãì³íËÓ® 77WÆ9WWW§ººº¼e«V­*Sk@¥  —ØèÑ£5mÚ4Íš5Kݺu+x=zôмyóò–Õ××kÏ=÷ Ünüøñêׯ´·é å;ì0iÙ²ôß—RW+!,¹qGX½óÒ½mS“sßø¬³]ðœ;w®ú÷ï_¦€J@@/¡Ñ£GkÊ”)š5k–zõê•Ê> ¤úúú¼eõõõ}ºúôéS²×æµzµôøãÒ‡æ/'å+V°¥‹{0ÎÃÒJ{ :´ftq/‘¦©iâĉ‰¶;ꨣôꫯ&jW‡–èÁJÙÃ`ëViûíKw¼,¢‹;@tÔvHÐm¿.­a_Îý&MëÜYj+غßç7Þ(^ŸҺŠ—£½Aaò´Ó$×­V1&‰h­è$mÙ"}ë[Ò»ï¾/¿ eЗóÓNËýl*ðƺu’çNr­B±'‰›?_êÛWúŸÿI÷8F¥VÐKÔƒÂäOHË—ÇÛ_¥]‘¨ ÄA@‡$gì÷ŒÒM7¾¯$ô'žÈýü駅·!mÏ>+yæïKdÒ¤ÜÏÅîâþÉ'ÎãÂ…éǨİ(e' KñÛR‰ï9t€èè¿ôE;IÝ­gOé•W¢­[ k×JÇ#}ÿû…ïë®»r?{’¸bW¶+1,J•Ð+1œÐ¢# CRº¡.Û*ýá…·#-[¶8s离ß,Þfí¨£¤SO·M¥õr´sëÖàçÛR@oj¢‹;@fq‡¤è_œ£¬Wh=îºÅf^OXØzçiŸ}¤vð_ÇýºŠ==‰çŸ¾n–>£8ŠÝn÷d‡iWÐK}û4PAˆŽ :$§‹{œ1èÛm'ÝpCîèY ¦-A}Ñ"éßÿ]3&x_î€Uªû {÷,}VQ»½ûïŸû™.îéœ'TÐ@[A@Gž4»¸G].9_À¿ô%©]»ÂŸ¶(U˵kÇùó£ï«Ø]Ü+q z)f„on–î¹Gzï½âìß}ë´bTÐ7ovîlP)¨ DG@‡¤tg“ö« ilL¿ËwZ¢ô¨¡Ø.JUA/–bTЋÐÝí¼ðBéØc‹w,#,LÆí²ÞØ( $í´Sò6•‹7 /[}[*è ­Èh$B¹”£‚n–·k— éYê6%DE"àÞW©Æ »‹{š =ÿþö·èUkw¥»XŠ1IܼyÉÛSJÞ 8Þ€~Ë-ñ÷%Ð@ëF@‡¤tÃVÜ1èfýêêÒtqŽ«X½-tqohpÚñÖ[ÑöYH›—,‘ Æ ^¯’o³Öš&‰ëÐ!þ>¤Ê|¢" CRüI⮼Rúüóà}E]nBLV»¸Ç™¹>ìýs‡’ÖRAÚ¿¹5Ý“OFÛg!ý‹/œÇE‹‚׫ä€^‰Õc¿€糦‹{e?~¼†ª/}éK:þøãuï½÷–»Id^F#²îµ×¤Ù³íÏŃnÖo×®ô㨣H³‚nº<ï½wñÆ —ªÂ8gNø:¥lSš· LK[® ‡uqóZèâ^y¦Nª«¯¾ZgŸ}¶^|ñE|òɺ袋ôÒK/•»id’âWÐ%ÿñµqïƒî® ·æ.îÍÍÒÇçöY¬‹Å ¶P´ë®áÇ3½ÂÞÃ$ç_Ü}”ã¶pIúÌ™Îlíq÷—¶-[ ߇_@ó9PA¯lÚ”ßÕÚ½^1ÐmºÊ1I\¡ãÐ ™$nëVé/É¿ØRì€nÂ9 óë_ÿZ#GŽÔ¥—^*Iúú׿®¥K—jüøñ}üøñêgëú$°páBí·ß~ån¶ žsçÎUÿþýËÔ"ØÐ!©8ô¨Ë³~›µ´zUU~@/v÷B™*â;ØŸ/Æ$qëÖEo_X;î¸Cºä’ÜxôrTзn•Ž;Nºýöüçmmö^Ôð«›ó©Øl“ÕEa»@·‚~ûíÒWä/‹2là”S¤Ï>“&OÎMfè¶aƒsÑÆS¨EŠÚ·o¯í<]/Ú·o¯öíùÚÒÉUÎ’Ô;á^¦I“Nƒ þO IÅ™Å=É$qYìâžÖ$qî z)»¸ºÿG¿€$j@ÿüsç±6ûÃ=æ<šïi¥ èæodÅÏ»ˆ IDAT é¹ç¤ë®Ë¾.î[¶$ûLâ*gýÓO[. ûüV¯–žzÊùùÝw¥Áƒ[®sè¡Òodëß™ÖæüóÏט1c´çž{êÐCÕ¬Y³ô»ßýN'N,wÓÐ&õ–”´g]Ü”VFk–¨Åœ$.KJYAon–.¼Pš1#~;“VÐÜ ,k×ÚŸ7tow# ú8i\Tð~æws¾f} ºa»Ðå~ ©l¿ù¦´xqËŠþŠÒ]wå/+´‚^ÈtÛˆ°÷ÔÝûÂo ùo„…9çœs4vìX=ùä“:âˆ#ôÜsÏiܸq:ùä“ËÝ42 :$eg’¸,*F@ª ßsôì³Ò{ïÅkgÒpûë_KsæHË—ÛŸ÷« §9IÜ_ä¯×õ×K¿ù}æ¼2•`¿Å$ {ÛçWAOœ÷;iäÈÜï—\’ûùÎ;¥_þ2ýrVГt÷…&&y+Ÿ:è /Ô…^Xî¦PQ2‰PjI&ÎJ»‚î¾z–ºžF q»¸ûÝfÍËtù.´½¶ß½ÚµsM÷J£‚÷óŒ»þ5×ø¿gæ}6ÝÍÍy»b…ôÀñŽ—ßßÈn»9Þ{ƒ›eRø$qIƒ³÷ÂÏgŸå~^¿¾åú•\A·½€,# CRô z”ÛùUÐÃ&‰«„.î×_o_'IwÛň¸½ܼ?ê{i‚·™­Ý+l’¸ q'Âó»@ó8~%Ù‡iƒ­‹û¹çFÛgRæXî׳ãŽÒ©§æ–ûýØ*ýiTнç…{5ÛyõBÀå—Û—UÐ_|Ñiߤ€:´\F´ftHJ6ówØ$q^q*èYbfÿ–¤±cí•fo´ñNç7Ýý‡_¸ Û—™hÙ¯‚n–r›µ¸ÝëÀ™¹“ìï‚^ Þ¿‘ª*çøf"k[@ëâîí²WØE$¯¨¼÷J·UнÏO˜àüüá‡ö}Ú. …KiVгԓ´ tHJ·‹»ùR;s¦töÙá_rm“Äeé‹ñ¨QÎãOê<N«‚¾qcü÷ é{Ð ¹ÍZ܉ë’^djW–&‰knè—]&}õ«¹çÜ’3¹™ èÞà¼b…ôê«…µ×ö9E½v{Eowþ¦¦\¯ 3Ì«ЀNT:$%»5WØôO?•~ÿûÜ—ð°.îY½ÍšaB„­miwq߸17éY\q{C˜°VAO2I\Üϳ˜³¸›jvÚ÷‰bÎmw›sE¼}üøÜ϶.î£Gç^‡7 t4hPaí-¤‚îwQ%h ºù·Áﶀvn˜.î»ìB@•‡€I-¿˜§нûªÔ.îFPØtWÐÿùOiÑ"{×eÛ}Ðm]Ü%iêÔxíó†Û¨ïeÔ z’Iâ’Þú-Í€n>·rtq÷ŽAonvÞ ¿ ºmB¸º:çñ¿Èß÷¨QÒ‚¹ßÃ&\¾Ü™.è¼(d zX½¹9B:w½{˜¶tÛ½ÑÝ>øÀyìÒ…Iâ@å! CRËP%P…uqºÜÝÅ=ËÜ¡ÊË,Û°AêÙÓé¦|ÿýùëÄ© Kñ/VøMu ºß$q&¸òùD}-Ÿ@ã­ ûÍP ¶.î’ݶ®¹KÕ—¾”ÿük¯I¯¿½-={JûíW¼1èaÚ›¥ŸÿÜùy§œß;׭‹þZ.»,øy3K}§NTÐ@åá>è”ÿÅü½÷¢Ý“<ê$qa±5uqß°!t½ÍÆÆÜ8û°Iâ wQo³wò¿$m)Fwo½”]ܽ#æ\7{l·Y3¼íìØ1÷ó™gJ?ÿbCXõ¹˜cЛ›¥5k¤#tÎ9w]jyèÎ;¥'ŸŒvì;îpöÉ%NÕ½¦ÆYnªé•‚ :$åW]÷Ýש´ymÚ$-]šû=j½µtq7¢¡ÁiçóÏçž3¯Áp¼aÇï>èiWÐã†[Óΰ1èI†AÄUŠ€^JÞ¿o@RA7ÜÝTà“t×ÛÅ=Í1è«VI½{çz‘¸÷ím×O~"͘íØ—\"ÕÖ:?759Ã1Ö¬‘.Œ¶=@VÐ!ÉÆk73‹¹u º÷^îû g™iŸéBë#n^›û=ñƒîíâîWA/´‹{T¦Íq+èÅ èi𶯜ÝÛÍ>I@wßÜÌök*v}Õ*©kWçï`ýúüníiõlhnvöß¿¿3ë}){LŠ€IѾÄ~ñEþﶀÞÐÐrb¦$ô,vqê~o èAt÷$qiUнûˆú^†t¿ º÷xAL[Ö­“¦Lñ_¯˜¡?jµ9ÍÐë$μ‡Qº÷ýn×.W9o×Îù/I[K=ÝЫª¤Å‹ƒ·OÊümMžìüþ‡?¤³_€R  CR´/ÇMMÒ>ûä~·…ƒnݤgž‰·oÛ}г(h’8[÷° zÚcГ^ÜHZAÂÛ–+®¾óg¦{›4>ÿ°Û¬…0 ð6…tq÷¾ßÕÕ¹YÍ««}¤ÝŽ˜ôÆFç"Ÿ© {ÛžvÝŒo7·]¨tH²ßsÙËÝ-[Ê}Á®«“N:IºòÊàc„uq¯”1è¶ §‚^ªYÜ£2m¾åûó& ÇßæãǨUÙB$íâ^_Ÿûù† kƒ÷˜…tq¯ªÊus¯®N^ARHÝË[AŸ1ÃùÙTнmO+ »{§*‹=y@ëF@‡¤Ü—ã Šœ÷‹¯Y÷á‡ñØ·Þjß®µtq2‹{ÔIâŠÑŽ ú{HC†ØŸ/d’8[ÈŒÓƸÝõÝëzOfóÿç6o–öÞ[š3Ç¿ z”Û¬yß着–ôJƒ¾zµóxÐAÎkñ»]¡Ü½S* _c É^ö[Çp¯5Uá í¢LW©·YKÒÅ=l’¸¸Lè{ôQç1ÎôÓ9Äþ|!·Y3¼¡9êët_¼ jCб èAíüøc箿ú•ÿ:a]܃^÷ض.îiŽAÿûߥÇk¹<­1è’4t¨tÀNûmôgŸ•-Šv§4o³æíâžöô—^²/O«‚î>¶­‚ÞÔ$sŒtê©ÑŽÔ*è Rñ5’¢Uн•)S èQ'‰ó~©öë2_.&Tî×·‚.¥›5cÕªxÛ76ßâ.ÍÛ¬Å'oŽaÎ$ÔH»‚^Ž€î® §=Ýï3)´‚îfÞ+¿ º$½ùf´ãùñþ;•ÅÞ8~è­‚î7Iœ- Çéâ.åfaΊ4'‰so“VÝlë½^ГVУ„X#¬Û½_õ?N@/Uw¿ã¹½öZþ~ ÅÝ[AO³‹»_;ŠUA÷¶=í1èts•ˆ€I…Mgff§‹»ù=K¼=¬‹»_ÝûºÂªÉË—K½{Kû[ðzÅ èŨ Çƒž¤‹»·Â]Ž úN;å¯[H@wWЫªÒïâVA_ºÔé~îؽ¯ÃŒÍ·twÝ ÍH{w:¨D‹@(Û,ä~ëÞIâ~üãàí¢TÐÝ‚º]—CÜIâÂ*èFXýí·ñøüc´vš.î¶}Ú”²‚g²7÷ö…tq2|ÃO!cЫ«[žû…tq¯®–öÜÓùù“OJ×ÅÝ´ùúë¥'Ÿ”,°¯çmï7¶\nƒ>ujðñã´Õ¯T÷>üЙ/ -tH ® »Ã§_=¨bg z%TЃn³æ`I*è~aÆ︶6”ª‚‡÷µøñÇ\È$q¦ÝIÂlZÝ(ô6kgœ‘Û.íÛ¬E½õ‚‚m¹­‚î~/Œ‘#›øºÓ¬ ¿újðógœ!UøqŒŒE ”KP•ÑÜçÙ;Ý=Iœ¨ãvq¯¤ º·ÂvÛÙ+èî l¶ihh¹)y@_¿^ª¯×Üï½Þ´IZ¶,·žíxQÚdăî@Ðœsï¼ãT.ëëý÷aŽ±xqþ¶qÒÅ=J½©)Þ$qîí’tsÁÅ+, Ç½¸b[n« Ûn9÷àƒÁÇ o*èiܲñ꫃ŸOã‒‚'‰3Ò¯‚.W¬¶n•®¸BúôSûó~“ÄUR@÷ö@Øn»ðYÜÍÄzîŠwÐÌ×Q˜{™/Y}›ÆFÿIµÌŒÚ;ìPØ}дº¸_q…S¹ôVáöÖ[Îc±*è~ŠÑÅÝÐ ™ÕÞ«Ôts¼ Éý”ª‚Æ|Q'ÒC@‡¤à.îŸîþ¶a¼Aâ7¿É…ko}ûíÃïƒÞ¹sËcÄ© ¯^-ýõ¯ùÛþ¿ÿ—û9lì¿ÔÅݼ†ÃK·‚t;:ÛöÞ€nfø·UƒýºÓg¡‚n^wÒIâܽTcУVÐýÓRþ{fÒ$ç1J½LÛÓü@ÛF@‡¤\·}Iw?çþb½Ë.Îc¡cÐMèò~©Nò%ûÓOvÞuWümÃØ*è×]—¿Ì]A7ïåSO9aùÝfÍ-, »ƒÉÈ‘ÒÀùÛÆ%Ý´Ù/ ›ý´k¯‚¾nÔ¯Ÿ´hQþkH+ ›ímAÍï=H{’8ïñ¼ ­ ‡ô¤]Üýþ®üÞŸ¨çRÜ ºaëâîÇ\Ô è娠Ð@Zè$MŸî<Ú¾Ôº+è¶.èiMçíÒЯ¾:×ýÚ¶M1Æ„Úúo~“<Ü]ÜO>Y:é¤ø÷A÷zöY§2ÿ‡?8¿ðóhfYw>IºM”€n3w®4ožtÏ=ùËm“½í×Ö ÁÝ®8Á¨ÔtÛòRtqO:æ:j@/¤‹»a›k"ÊùˆoºIzúéôÆ Ú€¸è”»¿²ÍM7IãƵ¬ {fX@2Iœ[P@¿ùfé”SZ.·UhÓ$¼Ño’¸°àTA7ãÊ—/w»tqÝcØÝûö{¿.”n¿]zï=ç÷bTÐýØ>Ÿ(ã¼mãù¥ð®Înîán{ï-uë&ÕÕ…ïÃÖ†8ts~D èo¿Ëü0÷\” ºßxïýÔ ¿/½~³¾»a›»P~ÝýûÖ­¹*¹_=i@÷®c« »÷˜Ü¡1ÉtïþƒÂŠm÷O>‘z(8¬û}žiWÐÝÝÑ£JÒÅ=Î$qîõÍþÒèâž4 ‡] JZA÷ë¶nÓ£‡ó8fLþçiû{/æpƒ€ÒF@‡¤\u6ˆ ˜Rðøo¯¤_ý¾ôšåAôbLç7CüâÅùmkßÞø¼]܃&›“’ô¤to@÷vç2Iœ-hyå®ú%uVËsÏ ›poïWA·#~çhP@wW»£ò~_|‘¡Ë¶Ï¸ÝÛÆ(]Ü‹ÐÃÖ Zn« ›s-J½{÷Üd~îõ››¥aä .ÈíŸ :¨D1¿ª¡µŠÐÝZï—ß  zÒ°VAÓÅ= ~¹¯~5÷³ èUUát#¬‚îWUëâîTÎv ³½7 {»`G© ›ctwÝÌà>÷6nÌ¿5›÷óô>ƹº»7ˆW’IżÁµkW©oß–íp¯7 {Ïw@ßa‡d]Üýn©¥ ¶ßã,šÅ=jȵ}¾Ò®»JýûçÖ)e@g’8*èÐÛ·® 2Ý»#I÷bVУŽA7ôB'‰óãƒîWA7½…ÃvtÛïQÇ »?7ÛybÚl*îÙÇ™gÚ÷m˜€ºukô¤ º›wŽƒ?ýIzí5ÿIââtÛ}ГTÐã¾Î¨K¥ƒîÞ‡w~Œªªüýf¡‚ÞÔDxñÐKäæ›oÖ€´óÎ;«¦¦FcÇŽÕ²eËB·«¯¯×!CÔµkWõë×O<òHÞó<ð€ª««óþk×®6• -Âú;$¯ —£‹{)+ènÍÍN˜ êâw zXÝ/ ¿þº½Þ÷(J÷ Iâ¼¢vqß~{çgw@7Çðß/ KùC $iåJéê«¥+¢µW*l’8¿mÍûpÁÉ+èînùUUùïK»vÒóÏ;“6ÚÚå'îÅ«Bÿ–Â*èqgBº• ûüÏÂ,î眓;Ï¢  —ÈìÙ³uñÅkÎœ9ºå–[4sæL]`LúøðÃuøá‡K’žzê)õíÛW#FŒÐôéÓóÖëÒ¥‹¶ý÷á‡jû˜ß ƒf¸–¤wL^A÷ Y6Åš$. ¶Èm¢vq« ‡½ŸæùBgq·u½÷î'Jwóû³ÏúÛVAw‡PóZüºûó5}àÀücL*Ý|³üÚ•4ˆ†}NŸ|ÒrßîJxÐqGŒÈ?Žû¢Ì÷¾çüìùç ôüLÐý^çÌ™ÎsK–oïÞ‡ûoל{Qºm‡  zØçúÏJ/¾íØ^aô‡J¶_Ðv1½DÜ¡ºwïÞêÙ³§úôé£wß}Wûï¿¿u›{ï½W]ºtѬY³$Ißüæ7µ~ýzÝvÛm:î¸ã¶­WUU¥Ýwß½ ö…UÐwÙ%¸‚î·LŠÐݺM)&‰ {;Æ›$.l~¤…Íâî·}±&‰‹óù†uq7Ǻ͚¹õé§ùÏ™×óØcÒi§EkO!cÐÃØþ.Üﯷ»ßûì=wN¢VÐÃ>WÓ“#‰]wu««¥_þR:öØäûèe1zôh-Y²D'N,x_Ôgœ¡ƒ:Hgu–ž~úi555é‰'žˆ¼°ñçRË º·jg z” УÂ¥K¥E‹$Ïœz-¸ÃmXp‰ú…߬u÷Bz!]Ü››£WÐÃ*©AïËï/}ë[N¸t·Y’nº)÷sX@·A÷²}Ža#@Š1Iœßg`¹¹ˆã~.¨Òo;Ž- ‡ýé§YAûíÜ϶ VA7ÿnD© {·µ]„J2‹{¡ÝÒ;v,Í„t m`Ä\‰=ZÓ¦MÓ¬Y³ÔͯÄõ/=zôмyóò–Õ×׫C‡êÚµ«u›Ž;jèСzüñÇ5Ü”-jkk·íã‹/ÌÒáÿú¯¥°.î~Ë$éóÏ}›á»êêð€n7|¸5 D èq*ènQöí #¦Ýî.îQ*è&œKéŒAò:Ìûíîâîæ èAcЃº÷‚Ñ°aÒwØחһ͚íyï¾Ík‹[A·½_ÛoŸ»¿»ÙßEÖn¿õm=H̽ì%{Û½¯Ï0ç©Ù&èïÝ èIÇ › ¶Û †qñª««“T'I:é$g™éùà‡€^B£GÖ”)S4kÖ,õêÕ+týAƒéÞ{ïÍ[V__¯Ã;Ìw›7ê¹çžÓ™æFÒ>Ư~ýúI’fÏ–† nK!]Ü/¾¸åºnûíçL°ä·‹»÷¹¨l}éRçV]ƒçžÛu×è“ÄIÉ*èwÞ)=úhþ>lÇó« ûmÔÅÝ}Û¯¸=NÝËÝÅݶ(cÐãô(cÐ¥x÷¬« K—æ8o@啖ϵÑÍ6=LZ“Ä57çf’÷ÛoXÝ}Á/, {/¦Ð͹±y³S òùçÒ²e’{ä‘ßq†®3Îp.x>õ”³lîܹêß¿´†€6‰.î%rá…êá‡Ö¤I“Ô©S§m·DÛèê_~ÕUWéœsÎÙöûù矯µk×êˆ#ŽÐ¬Y³tÞyçéøƒ.73DIºîºëôì³ÏjÑ¢Eš2eŠ¾÷½ïiÆ úñ¹mîÛ\IÒÞ{·\§Iâ´o/y;$íâžF@ÿêW%÷5=öjk£A—’UÐò饗r¿'™$.n@7¯ß`“L¥’õ=t߆̻¿§Ÿ–>û,|’8·°YÚMXŽ1uCÞ¶6î×¾`Aîgw@ojÊïa·‚^Ê€îݾº:}ÛßkØt÷c!]ܽq.™ó(ÊPŸo[ª©‰¾o€¸è%2a­^½ZC† Q÷îÝ·ý÷Øcm[§¡¡A|ðÁ¶ßÿíßþM³gÏ–$|òÉš;w®~øa뚉è‹/¾Ð~ô#õéÓG7Þx£ ¤úúzõîÝ;rÛ¼ýä“[®³ukò ºá÷¥Ù¶ß¤ô¸ÝwÝ¡Àüì ÞIÇ¢vq:‹»ßþâNç×¥:¨ÒmÞco÷$“ÄEiû8~]ÜÍik‚êäÎ=ã„œa~ÝûyIÎ{Õ¡ƒ}}ÓÞææxáÕ´É}QÅö¼÷ç .î…ŒA/U@ollyk;³Ü¶m” z”€§‚î^?ìïÕœGQ.ÎÌ›ÿ»mß[¶8ûŠûo€D÷’iŠð­Ø6i\ß¾}5sæLßmƧqãÆÒ´ÝVmL£‚нǬ®–Ö¯ÞO’ ºw›(cÐÃúñÇ;U]·8]ÜãòÞFÌÛÅÝУŒAO»‹»ßkÛ}wéãíïu§N¹í®$»ƒ“{¿;Ûaëânº_•Ô6\ ȼyÒ‘G:?×ׇ¯ïÞoÒ1è~]Ü׬ ?¾[¡cí/¿\úÛßZ®' {+èQº¸k º*¥‚Ô.ãŸÿ”zô<£“"¡‚ŽmÝŒ¿ô [¶øWÐý–Eᮦº+è‹Ù×O2Iœ_Û è_ùŠtÜq-·±uq7"Ò® {»¸G èRþ=¬Ý¿sœ zÐ1ŸüÄÿy÷1ÜaÔ/øûÝ¡pýú–a«º:x|qÜ úâÅÎ-Æ‚ø½Þ€n{ί^îIâ¢*t ú’%Ñ÷›f=î,îQ™ ú† éT½?úÈyœ>½ð}€¶‡€mØàÌdlº×ú}Éݼ91èÞõlÕ® /ÚiNå6k¦}¶*«_ ²UÐ%çÞÔA·YóîÃö¼w"µBº¸¿û®óè}¿½°™Û¬÷sÞ6æ8“&ÙŸ3ï™»«wPWñ/Ù¿=îsj·Ý‚»¸ÇíÕ&½û²3¶Û¬•{’¸_ü¢å2[w73Ý…_Ý-h zÚ“ÄEýLÝôB»ÿK¹ 7I/XÛ[o½Uî&€thÆüê¢_HØ´©øcÐÃÚ´Ÿ°çl¢ÞfÍýúÜ_âÝq7[}î\§;|œ±ì6ÞŠ¦·‹»÷Ø~ôY³rã§8 ÿux¶ 1·Ýæß7¿÷Ò„¦éÓ[¶Ë]Euþ~Ý=\ ÈyçI—\’n=ÊzQ+èŃn ½+V8hÿ IDAT—^š¿|§ìÂÎ7w› ƒžvw÷z7Ü ÝÝÌГüMzß÷°žåvì±Çê ƒÒM7ݤ%~Ý!@ÙСðâýÒìÕÐвòe¤ÐKUAߺ5wì .î’·±Uпò•üý$­ {»ó&íâþþûÒÕWçöÐO=ÕytŸªÌû½—~ÁÛ´Á{tïzÞ€t!ÇtÓ<Øymaá7 =íYÜûÛ–Ë^xÁyüå/ýÛëæ}­ÞÏÖ´9ì® î6¦=Iœw؇ûß”?”¦ñ(8 {™9<³ZAÿàƒtÇwhÙ²eúú׿®AƒéŽ;îÐG¦o>(+:ZTЃÂq±&‰KRAO+ ›/èIÆ Ç© wé’{λ¿÷ÓF éâînoP@ßi'çÞÏQ÷kš_gÛsî€e º·bêuØaÎ…‘#[îÓ˯:ì'Jw7ï9cŽé½(ô7”¤‹»m¹öí¥aÃZîÏob5ïçä âÞ¡Þmm=íÛ¬Ù*ènAûvßf-n@÷®oæXp·3kª««uä‘GêÞ{ïÕ‡~¨k¯½V÷ÝwŸzôè¡£>ZÓ¦M+whÓèØÐÃ*èRq&‰sƒ ýA!*Éô¸Ý}ŒÆFÿ÷ÂÛ¿€‘Ö$qf›8“ÄÞÏÜú½û ÚŸ_@w3Ô¹Û6Ý;¼ÀÝî±c±æFU•´×^ÑÎksA¥tw@ojŠ mçÙvÛIï½—Ÿõ°m6lp.ÙÞ €nžóôBº¸»{ăî?Â;=èbá¾@—F÷‹.²/Ï¢·ÞzK×\s.ºè"µk×N7ß|³~øÃêúë¯×HsU ”‘Ç K¥›$.‹t[åû ƒìÛ¸+è;:·]r?çÝO~=nwo{ƒ*èAݶ_¿÷Ò} ïmô¢ŒA÷®ï^ïÔSózбmû :–WÜ º_@÷ °6¶çÌÞ~;z[6lpÎùª*©{÷ܹÔÃÅͯ‹{¡ô¸cÐÝï©÷oÂû³{]÷{uÕU¹¿ÿ4fqo×.ûÁüÆoTŸ>}t '¨¹¹YO=õ”êëëõŸÿùŸúþ÷¿¯3fèÉ'Ÿ,w3h³¸:´qcþ ×I*èQúܹöåqÇ Ç™$.¨[¬ä„ÓÅ5í.î&ôî- {n»íœÇ—_=¶ç½ÝÀÝaÚ¤Óèân›|λ?[÷ ÞɴܽÜÁ»_?{Û½ô°s0ì¹bVÐÝç–ûœôØ ¶¿‰ÿ÷\{¢öjqÿ½Ÿ~ºôüóÁëÍÁ % èIÆ tóúM[ü*è RŸ>¹å/¼;¿lôsÎ n“m˜†·½Y³lÙ2ý÷ÿ·?üpUYÙ¡CÍš5« -r” Rñ+è»ì’ÿ»­‹{Z“Ä…µ)Éô(c†ÝaÓüîßm;®ûÑ»ÜoÜt1+趪¥M” ³·‚îЗ-³Ó6I\Ü1ÜÞí‚wS“t÷ÝÑÎu¿‹$îPg ºß… ɨ…ÍƹsÞ}®D­ {™.î~oÒª {·ñt÷s^æyï˜|÷)l=èÖ|†û}¯„€~Ë-·hûí·Ï çk×®Õ–-[´Ë¿þqîÛ·o¹š@›GwhëVçKv!cУô«¯–vÜ1Yšô¸½±1WA»zÜ º;øx¿Ä»»¼û½ž ö˜ý¸×+Ww[À󻺛7 »ÃªßE¢ 1èaç {]3aŸ{Û(ô?ÿٙ쮻Âoñæ÷UÐã^`0ï“_ÝÆ}QÊÝΨ“ÄùµÁ&j½ª* ºßt³®÷"À¦M¹öÙ˻û½Èj@6l˜Þ{ï½¼e ,ÐñÇ_¦7:´ukþË$t¿eQÖ)wÝô¯ݾNÐô(·Y³Ÿ(XÙ‚¯;${»»º­ºt¼$“ÄuÒÅÝÐÝÇ : w—fwƒÖ7î¿?ÿ¹¨“ÄmØà<®_/ »ÃgÐô Aô8]ʯ ‡#¬‹{Ðíëü^_’Yܽÿöto{MlÝìÇVA;üÄoH–üãÿÐÁœ·ìàƒÖÛq&2ECw¨±1~=É}ÐÍz¶/ÁÞm“N÷˱·šèw¼  º-ÔUÐÍóal¯¥]»Üñ¢vqOk º_Ý&Ê,îÞ1èRx=h ºi§÷s§Ÿžÿ¹D$Î{?ó  IÇ }´³ï?ýÉ¿ý†i÷™gJ;·|Þ½ÍN;Ik×:?§ÙÅ=èoÕ{kí 1èiN×ÜlŸ$Î{ç÷ûo›$.Êùî~+¡‚~ÔQGiêÔ©:ùä“·-›>}º† RÆV(¥ùóç´}çεï¾û¦Ô^tlëânë>è~’vqRAû‚í¾ÍZXûlÁ ºZúüó–Û„UÐýîómk»ûùvíZ†‘{î‘9$zw¿ î6zvÜ.îI&‰“Â+èa]Ü£\\ò{.JÝý™F©òî÷Ã{›5o÷÷gŸµ·7¨‹»¿ñÑflµ7 '© µaÒ¤hcЫª èa÷A÷« »ÿN“TнlC.²VQ?ÿüó5jÔ(½þúë:ôÐCõòË/ë÷¿ÿ½&L˜Pî¦(º÷%I#FŒ(xO , ¤EBwlë⥂^h@ZAwßöÍ+N÷°/ÇI*èÞ@õñÇ-·I£‚n{¶€þê«RMMñÇ {»ï…™ 1è·Ýæ-_íu¸E™ÅÝ»?¿.Ý^£FI—_îØùóßï»Ïy®.î…ô8ts|[»–zõrÞƒwßu–Eíâ·ý…t÷òbtqwOÂæn£i·Ù6Jwï>l¡üîƒîWAwo›´‚¥‹û3ÏHǾ¯R4h Tîf(›Þ’ú…® <èmÔ_þ"í½·sË3o÷,L·Ë.Ò¡‡Jû[Ëõƒ*èQ–}ü±tÝuÒùçWÐM»ÌcœIâ ­ ·oïßÅÝïõ{+èGáô=÷l¹ž—»…Ô2`u×}ýuiØ0ék_Ë-ó [UU¹Þ68³Ù›±Ó¤‚nöŸ„ù¼¦N ^ϯ‹»÷}ñ¾W¶ º9fÔ€n;gÜëûm»p¡ôË_æ_\°M—´‹{Ð$qÞÐt/Ê­ùÜÛzÇ £‚žV÷cÍVW÷•+Wê/ù‹>üðCIRss³ªªªôÓŸþ´Ì-ô6ê§?•^yEzâ‰x]Ü ­ ûuq º^qÆ ÛÖ{é%é¿ÿÛyÍîÛ¬IÒòåùÛ¹ºßt›BÇ ûUæƒ*è HC†8Çš8Ñ f¶ý'éân^“mÛ÷Þsf5á…àíöíÿÖ­s‚¢í¸å¨ ?ûlËÙݽüzØyç†Þ zXÕÞ»]œç«ªœ Q3fä/O³‹{P@÷« ›Û öìé<Úº¸&Í›'ý+C¶Ø‡mrD¿Éþš›¥ .h9_DØtóÜï~'õ?ÛßC%Lwçwj̘1:ôÐCõoÿöoånð  ·aŸ}æ<š.îF’ zœ.î^A·ˆŠ¢t³NccË.î·ßÞr»°€îx èacн6nÌmk¹mÿIÇ {‰ùÙöž›@úæ›ö¶vìè„úì•N¿sÐÛU¿ý%ýŒJ©k×®Ú>lâ P6TÐÛ0óåØtq7Š}›µ €î]'¨‹{”}{—¹»77· èwÞ™ûùØc±úÞ1¸Þ*ª_P.¤‚î®”G­ ›ã†}aÁÞtýw¯4ÝvQ$lvØ ¿ ºí6kAm âfÝݺo¿Ý©ï¼³}›Rt7[wo¯ïßaXwÛ}нÜçôªUÎÄ”nA½ l‹•ÐÅý‚ .ШQ£tî¹çêàƒÎ{îë_ÿz™Z zf¾À&© ÛÖ‰R=ò Ŭ {º;øÝýùçóÛç7I\Øò°€n¥‹{PUÐ-J=J@÷{­¶÷ü½÷r?wíÚ²û¸éÚ\UåL(9Ý®ÃÆ »¥]AÐÍ6†í‚D’ z¤Ý­º:ÎÝmð¶×½¾7ÛŽáwñƯ‹»í8Þ‹a]ÜmïyØ}Ðmšš¤o}K:òHi̘–Ï…íϯ‹ûÔ©Ñ'¾+¥n¸AUUUºñÆ[<·dÉ’2´¸ÐÛ0ó%3Êtó¼T÷Aw¯ï×î8]ÜÝÕo³Ý„ Îë)ä>è~÷o.´‚n‚ÑŒ¹ R² z”€n¶•œsÁtõ÷VУŒñ—¤Gño£”›¹}»íœ[®í¹§ôÇ?J&D èiN·‚în£ûˆ:=î,îaíûôÓàçýötüª*é‹/¤ýö“<ÅÖmÌ…[/s™ NA綀ž´‹»m º³®ûó¼ûn颋ÂÇå{¹ÏG3ÇGÖ,]º´ÜM*`ÄŠÅ|ùŒ3‹{1&‰K³‹»wÆgïúî.î’tÕUÒYgEk_šcУTЛ›ÊÞ]wå/O:}óf'Gé²ÜØèT?¯¼2¿¶1èQg!w»ççÉ8“ôu–%ÒOÒ ú=÷Äo¯áWA·…E·4&‰ Óµkø:¶¿Û°€þæ›Î¤ƒ?î¿O¿ÏËTЃnÏ(ùwq;Iœ_ó s´©É9Žû"å^{IGíÿÙ„õb©¯¼òŠn»í6IÒ| ÷ß¿Ì-½Msß8JÝöè~¾Tô €j»oÐ$q7Ü }å+Ám6í°t÷­—¼ëUÐmckÝü*åîû£µÕæÇ?v^ëÛoo×¾½Óö¿ÿ=7ÁXP@÷¶§{÷ð x¿~Îí¿¶?_Ì1èßþvËe[·F«x£‹{Ô1èaþøÇðuüö_ȤgAÝŒAOÐÓª õ1ëºÏ9³¯¸]Ü+Á›o¾©îÝ»ëÊ+¯Ôu×]'É©ª_pÁenèmšûvcQÇ —j’8÷~ö—¤Þ½Ç° ä DRp˜ôvñöîï{ß+þto[ƒªwAôyóœÇõëíÇtA—¤-[Zî7Ê°‚üCgÔ.ÜQº¸{ƒaÔ ºm¿þÕ¾ÍÏ~&}õ«¹ßýº¸Ûføëân« ûµÄûr·$s@xÛ÷<5Û„UЃÞ{³ÜÛEß;‹{œIâ¼ÇŠ[A:ßÍ6}äô0ñû{ʪ[o½Uwß}·fÍš¥vÿ:‰¨×^{-õc­ZµJ£FÒW¿úUí´ÓNª©©ÑܹsS?­ cÐhnnÖòåËÕìúööå/¹Œ-JÆÛÅÝHRA÷[eÛø .î¶ zØ$Pîõ¼ë…ï$iÞýŽ4Æ G]î>®_@7÷IÚVÊ}þ›7·ÜoÔqÿaÇ{Î}ô϶\?i÷¸O÷kö›ÅÝÀ}T:ê(ÿ€î­ Gé&}öÙáëu÷ÖÅ=j@÷» â­ ûÙ°¡å²¸t³äÓ=^XÝï\ ª O˜àßö,{õÕWu§ûv’>ûì3í¾ûî©gݺu0`€zõê¥_üâ:ì°Ã´xñbu2€6Œ€ÃâÅ‹5fÌ=óÌ3úÜuï¢vØAë+­Œ¢ü€î®³‚žFwÛ—í ãÞ!ÝVA÷;vœ zœ Ÿ¤Ý} )z@·½ŸîeîÛ|Eå~vÜÑ? 'éâç^ÕÞל´‹{ÐmÖ¢TУ„ø¨½¼¿]$ˆú7íí`|ð´ûîá]ÜÏ=7×ÃÈ;=IÝ|Þ.îæØacÐ+ÑøC?^×\s$ç‚ó„ tÎ9ç¤zœñãÇkýúõzî¹ç¶-ëÕ«WªÇ 5¢‹{ ãÇW—.]ôÆo¨K—.zÿý÷uî¹çê¾ûî+wÓñvq7‚¾ûU!ãtq÷Js º·‚§âЃÆÔFƒnã7=Iw#¬ÒÔÅ]jXü&‰Kz¸·sŸ;îh_?n=Iû¼Ó¯×‰ àæ–q^QÇ "¨çKÐïivq·­¿fM®m~ï½-·UWK§œb_.%ƒî¾kÃ?þá\¸òëâî÷Ù¤õ™•Ã¨Q£ôúë¯kŸ}öѺuëô•¯|EsçÎչ瞛êq&Ož¬o~ó›9r¤öÞ{o9R/½ôRªÇ 5" ÇðØcéÆoT=Ô¾}{í²Ë.ºá†týõ×—»i‰¸'‰ ›ÅÝÛå¼Xô(]Ümµ z[@÷vq÷ Ê…tq÷ …Üf-n@wWЃÚämOÒ€îæ›;íd_Çv›µ @™¤‹»w]÷ߌmw÷ýÝÃ*è悇û6z…HRA{>êlúA=Ê$qÛmgßï7Jÿõ_ÎïæBÙDzeömlÜçm§NùÏ}üqËnúæ\òû[{ã ûòJ°ë®»êü£Þ|óMMž|¸†^´¶@[QWW§ººº¼e«V­*Skà‡€ÃøCÍš5K§žzªFŽ©ïÿûÚ°aƒ®­Ðojqfq« û-“¤¯}MúÛßü× êâ¾ÿþÒÏ.ýë;^^»£tïzI+è¶/ÿA]ó/v&’Z¾<Ùt¿ zІªª–Éì?ì:’÷ó« ûuq÷ÝàãD% Û*æitwH›=[:è gÆ÷-[ì]Üýºß$qA7âv£ŽRA÷zäçñÙg[ßö»»úì7©cÒ ºäüÛtôÑ-Ûä¾èa–»÷ï÷^¶o/ýàÒÿþ¯óûºu-Ç ›s«’»²ûéhºyxTUU¥:—Êž{î©ššš¼e‡~¸þ?{wEyÿü³!D!á– EA®$„Kð‚ P´AQ@ªˆ%‚­G¹ÔR*‡U+DVP[)räÐ$(‚+V‚âOÎòüþXg3;™swfg6ûy¿^y%;;óÌ33›d¾ó}Ž… ên7oÞ­Zù¿ß{¯zÌôAWÈÒr£>èÊ@Üê€VR b6ƒn¥ü[nÑ~/” º•iÖÔöcf]yK“>}üº²‰»€Ëc3tµÁåÔ¶7ÛjÈòØOŸö—ê„ ;ÑÄ]ëƒT†òA”ü=½>èòßk ´ z4Û³gOÐ뢢"<ÿüó¶×£G|øá‡AË6oÞŒ~ýúÙº""¢ª†}ÐÔ”µÁ9à4|Ðÿó®]Ë¥ãÖ­+o#Vtµ¬ÖäÉ@§Nþ/µò”·ÚHÚzt­²äMÜå}›;v¬ø¹W/ß׋. }8åàV¡dÐÕúŒõA€£G+/33’ºÕ&îZt-_~ 4nln]yp–‘áŸjMÉé&îwݼîùó@B‚ÿgyVXz¨h>­Ü‡V] ‡Ó?œ>èf®im+oâ®d&@×j⮵? r€®uý•­Nâãƒ_Kt+Ó¬E³ôôô ¯ÁƒãùçŸÇ¬Y³lÝO^^JJJ0nÜ8|òÉ'˜1cþùÏâöÛo·u?DDDU ôV^^ž?¯ŸA—œ8áÿN€.pìЮ¿ù»|Ÿf£>ØÊúkeÐÐÞ^jÞj¦ºZ=”ý\C ÐÕúŒ›$NIàðaím”Ûê5q'ƒn@jeÐ>ø òúV3èVtåçìüùàLo¨MÜ•Ó¬íÐ àškŒëJôp×—ÖÑË ‡ÛÄ]«ŽjMÜ¥@?33øó¬ ÐåÎSÏ ë WÕ”””àœZ“„0´mÛk×®ÅgŸ}†~ýúáË/¿Ä|€+¥?úDDD¤ŠMÜcXyyÅM©üfW/ %ƒ®–Ѿp¨[·òÈÌj´2·’³gýý½õö'ÿ¹¼ÜÈÇ!’ïßj€®FÊ´jÑ«£ô¾V]oš9­ÝL2èZ‹Ù,£Z–ÔhÝoôC œ#[ªò³jw€®lâ.Õ Ï Öš¸›íƒ.m?w®z«Jt£saÇ4k™™@ÏžÀ_ÿjo€®O&$øÝKO ¨(G~-ªWWŸy@9ÕdUÍ ÷ïß?èõ‘#G°wï^<û쳶ï«gÏžµˆˆÈ"è1¬¼¼â¦Þl]/ȱÒÿWÚ·tcJ÷£G-[€ãÇý¯SS~0×]op1y€®Uw#RFN ÌŒ2èÊ€¡ZµÐ2èj¬Ž\ouš5³e[É §¤óæUYjœîƒJ€.eÐõš¸ËƒY3º™ YÚVù»¢|P¤uìÊYŒÖ—“?(Q«ë¨QÆGìjâTŒÿ õ DÙ]y Ò6U5ƒ~ë­·½nܸ1zöì‰DåüsDDDä è1LˆŠYÞ$ÛÌHívdÐãâ*+ÙMùMsÀ[où;Ößw\mòì·Ú tÊføjté»™Aâÿ9Õ ÐõHþýz#ð{Ù]wÝß/'Kô³Äçóá%i:"ãäÉ“!o¯œµ€ˆ(0@aRÚ»7pñÅËõ²]Z7âzº2;-eÐÍ6qW¾V¾'èÒ¨ìqqÀ?þlÚäï÷*_ÿ»ïÔë(g5@× PåÍ‹2“jMÜÕȃ³±Õ>èRÝÔ p‰³ÒÝèó„žA·rÞäßåtiÿJRýüùàýìÜ\o©>z£¸›Í _w°vmp ,1›,ÕjânÇ(îòrÍèMšèïÛl€.ÿ<óMÅrùïµ|]«ƒÄ)©‹iiiX°`²³³‘™™‰ÂÂBlÛ¶ ¿ýíoqÑEíDn*..Æe—]æv5ˆˆ"Žz “nT­Œ Ê<èjÁ–V€®×¬^¾½ÜÙ³þæäò)®n¹Åß:௠΂ieĜȠ˃£ ºZw5FAC8t‰Ù&î'ú [íû,mcµ‹„žòry‡û!TèjÇ#è?ÿ¬½ùgÉÌ<èVš¸+×U4hÔÕ"Ô]o}3×Eëí[zßl€äçW¼V><¹ê*ÿXV§Y‹Ö&î;wîÄÂ… qóÍ7–½ýöÛxñűfÍkF¬"s¾@›KY `Š=""Šè1lï^ÿüçÊ)œ¤à¼<¿n¹œœàuÔ¶Ó[®lâ.ݘ‡ÛĽfÍŠŒµr¤y½ì»VµtµY«L½ÝLt­zê5-×:^ÅÝlt³¶Ótù699þ팚¸Ët-zMÜCÉ Ë?£FëZMŒší2på•Àž=ÀÿýŸ~9f2èft+MÜå”ú† þïÁe= “_»ÛH­í IDATöíÏ?×^×KvïÞL©?Ó/:vìˆ]»v¹T#"#md…¸-›¸Qô±Ð»‘ªš‹.ò?uÊÿ]Ü}wåm.½´òºÒk+ti-3º’Z÷5*nò¥ÌštcnwÝl@j¶‰»• º• ¾¼|+}ÐÍŽânu8£&îF™R¨Ó¬Yiâ.msì˜ÿ»Ù&îf3èzº$œ ºÖ¾Í¾o6ƒ>i’ÿgéo‰V¹ft³ÎYiâ.wá‚úµ±:Íš¼Œ?¾ÿ^{]/ÉËËÃ]wÝ…eË–¡¸¸Ë–-ÃèÑ£ñ»ßýÎíª˜Aiû0t¨z fÄJ€®5H\BBxÓ¬I¤ º)8èm«vèf›¸[íƒ.Êlh¤ú ›ý<ªeŒÍ÷jÇ0j ´ZžA—èzÞÖ­ýßOŸ?ƒnGt­}›e6ƒ®œo]ù¾Ñ~Cmâïk¢¤Ä|]+@W냮÷p+Zû ×¨Q'NÄĉÝ® ©`÷f¶Iº™u¬6q·Ò]IÔœ=ëÐ¥QÜ%Òy¨t rv+te¿vµºIõ3;ÍZ(MÜ¥}‘ê(M¦œÙFëAI(- ìZOÙl^ïsôKâ0hÆÉ¢E@½zÁËRS+7§6 ÐëÖ­øY뼘fMzÏÉ>èò-zï[)3”u”Á1`>@7ú¢\®5¸œÖ4kU-@€/¾ø3gÎÄ Aƒ;vìÀ‡~èr­ˆˆˆ`€ÓB¹×[×Ì ²< 0;ÍšÙ ºÖ~Céƒ. ÊÀE­Yr(t½ã€~ý€nÝ€.]*o§7ÍšÏ<ôPðÀVû '&ú¿Ÿ8Qùýp2èfû?kÍ¥­¶¾òº«› ¹þz9ÒhÿzûxõU`úôÊu6jâ.¯O(£¸«eЭ ®j]º†¡öA€_¶l±¶_½uZäœ?ï?7Vf¶³ÒÄ]ë:Zi•â%«V­Â AƒPVV†õë×’’’0yòdw+FDDD Ç4­a³ýN•ÛØ‘Aµ‰»ZÐj]è(ôÓ§mÛÔËÐcµúСþ`çºë*—c4ÍÚ“Oú›l땯WGé\––êïÛê qf™Í Á±Ïç°a‡#‚3è£ ºüs(i6+å£]bgw-jŸ–-^½ÌíÆŒñ?TR£ö»hTžÑߣ ºV9Vš¸——W|¾•¿‹ÑšAŸ={6Þ|óM<þøãˆûåÂ\~ùåØ»w¯Ë5#"""€zL³3ƒ® « –Î(îÊ€ð­·ÔûÀªeÐÍZúE7cÖ+Ólwµ2´‚ £>èjû³šA—ZHý®Õö­W÷PfjûUk9á$£iÖÔti=©e‚Ù&îÊñ̬ggý«¯Ô»•9rdåuœ ÐÆÒw8i]åCÀhÍ ÿüóϨ§èϱgÏžJS¯‘; Ç°HèÊ@UÊ ÇÅÙ×Äýßÿ®¼_+ôǃ·Ó Ð[·ÖŸãZ‹2Sj”A—¨g3¹Ùòµ$$¨â®vþìT¤>ðf¨èyy@r²=uÑË «5­–v¦,C Ð娺|?f»(Étùhä¡´¦yùeíuìÐCiâ®5Íšwnżå¡ì_o™Qwy°¨q½ òý?Z÷sçÔë¡œ‹]ú¼k­ôaÆ¡G¸ûî»qüøqÜ~ûíèÑ£~ó›ß¸]5"""ô˜¦¼½ê*ëÛ-S.—~¶2Š»Qýè¼<(0 "Ξ ÞŸ´íöíÁõS ЧMV­Ò/_9r´Q†[/@7;Uœ^ùjäûÒÊ ;Ù]Ú¿üZè‰Dwµ]/€T[Ö¶­ÿ»”™ÖË Ë÷wÛmþc´ÒýСŠ×fZ,ççû¿Ë3èÉÉÙd+ƒÄé½o&@ïÖ È퉻Ùúµhan?­©%V2èééæ×uÒ¹sç0sæLŒ?[¶lÁ‰'°cÇLœ85´FO$""¢ˆb€Ô7œÒý™te†Y¹½V€¾s§ñ~ÍòþÖj}yõô„`Àýòš¸kjz}ÐÕ¨mgµºþcR>TÐz8`wSßîÝÍ­§ ЭtÑ•Qwåu€GŠŠ€´´àuŒš¸[¥<þûï7^GÔMkw»2è¡tŸ1ÛÄÝJ=%Å?B¿yvŸøüsíéî¬~îKK»î²¶ÝªW¯Ž—^z g~ùƒ—¤œ›’ˆˆˆ\Ç=†)ohÕúÏšÝÖLÿWùvR€.XV¦.+/÷÷çèÐA{=+ý^庼º’Z€n†^}õjàw‚ß7“A×kântµdšÓºQÓfµýz9ƒ.}¦«U:v¬\¦Q=ܺšiå"(c¦ÌPÖ‘·pc8ùòq「 ízÈË’~T/ßÊç^ ¤øì3óÛ8eܸq˜6m~úé'·«BDDD*,ôn¤ªFy#,õ‘µ;ƒ®Å]ÊèV« â¿q½ñFí}«eЭ4½•ô¤E«‰»²,+ºüXôú oÙRyþh½ÝJ6\bu­Qñìƒ.1;: ×©co=ë}Ðõ>—ÊÀÞçÓo9bU(ºÖ`ˆze•©&º•ú™ Я½˜=»âÜ(¯­•Ï}(ÉœòÒK/á믿ÆsÏ=‡-Z þ—&>Ÿ»wïv¹vDDDÄ=†)o\¥ ÌjeéµÕ&îqqþ¾®S¦ïK9»Õ ÀJ€®–A%@—SfzÍæåtýÛo+²ûFºVóz»û ë]Û{ïõÿÛßüu‘F|OOw&›® «†Ê®jïY݈Vi«±ÔÊ°#@—?ð²òY1 [$Ðþ}?|Qž›pš¸{!@¿ôÒKñõ×_ã…^<ú裘9s&„Û#""¢ Ðc˜V€nf1µ›ÿPš¸kÑ»g43è›|_vgЭ8°"°”˜ ÐåëýêWÚYlåvfû 7kVy[iwµ²#1͚ѹ¹ñF€Þ´iÅHó¡L}§d&ãÊ qʲäÛ;ÙÄÝÌï¨MÜÍ<°’A7š^NZ~þ|ð4kfZð(÷¡µ®ÚT‹áènúþÃ?úôéÀ?÷yïÞ½]¬©aô¦ ¤àÓL°cGÝ()_Jw+Aº´Í¸q@ýúêu3ÚV~>Š‹ƒ³Øfui½PæAW+3}Ðõ‚BŸÏ  ¤¦VdÍÍŽü.+ƒÄµküžZ]´I]=Ba&@×ú\;=H\(ºÑ4kÊ º•úèò¿gZôhmâNDDDÞÆ z ÓÊ ›iÂj€n”A¿ï>ÿˆ×=zhïÛlw+ô_ôg¹×®õ~vfÐÕêNw½À@+»+móÌ3Àï~g\O¯õA×"è Ø·ÿgŸ­øY-è5›AóMÿC%ùµ•Ÿ;©ÛG¨ôô›nòO»võÕêëh]C7t£¿'?ÿ¬ŸA×jM¢W©,ùuÖêŸmMÜËË˱ý—y#…¸páB൤K—.nTˆˆˆd˜A™3g¢sçÎHNNFff&¦OŸŽÿýï†Û¡wïÞHIIAVVÞxãJë¬]»]»vErr2²³³±nÝ:Su §‰{8º^½aCÿœâ‰‰•·‘„’A7º9þÕ¯üó©/[üþ÷Îèfúíª­jwiµÀ[¹­^w·3èJÉÉÀ¼yÀ’%öÔrä¯ÍW§•UñZ:O‡W”%ŒC Îåç¿eËàºJn¹Å?ÖCrrð¶‘îƒng€øg_°2Š»Ñþ«r=55·Þz+n½õV 6 _|qàµôEDDDîc€!7nÄý÷ß­[·âÉ'ŸÄúõëqÏ=÷ènsøðaôêÕ °råJddd`ĈX³fM`ÂÂBÜpà hÔ¨Þÿ}4hÐýû÷ÇçŸnX'e@‰ÅÝ e`èD]2t¨?·3@7ÓÔXy.ÌèzǤ5Hœ^sàP3èvVÕ €+¯´oÿj->ä¯Í6qWúúkÿ÷üüŠ²ìž]«îF}̽ØÄÝÌß“¬,íÏ´Þ+t­sm}пùæìß¿?ð¥|½ÿ~w+HDDDØÄ=bäAu›6mЬY3´k×{÷îÅå—_®ºÍ‚ P·n]lØ°pÕUW¡´´O?ý4ú÷ï˜?>®¼òJ¼ûž={";;óçÏÇÂ… 5ëÓ¹såN)»êM¹Õ&îád µ¶íСb®a+t¥H7q‹SŸrK­Ùr8ƒÄõו¸9Š»•A½ìd¶¼Ù&îZåË_Û9ÍšrðDµ9Ùµêää<èfö#g%@‹ÓΠ ¼û®¦­}(éâíM܉ˆˆ(:0ƒî’ò_îú.èÜ¡oݺAË222°E6iöÖ­[‘™™´NfffÐ:jþö·ÊËî¹Çß\8'Ǩö¡eÐå7éF£¸+·7›Aÿä ¬,xV2è·3èÊýÊ_KC(Ó¬™Í k à‰QÜ­4qw‚Ù º×t½ý¤VçAŸ>ÝxµrÌüjÄjû‹‹ÓžfmÀ€à©øäB$NY~´5q'""¢èÀÝåååÈËËCß¾}qÅWh®wèÐ!ÕýÌ™38vì˜æ:;vÄwß}§[µ€:!Áß\ØÊüÉzå)×U¼j€®7ÍZõꇅ“A×jrÊ(îfZX ÐC™§ÞL]D:¥þ¾[ôH2 Ð¥eZ1>üÐ?Ø ’Z« ;ÎÝœ9þ`´F û›¸k]‹† × •ÙQÜ¥uŒþŽ(VÉ¿ïܼnUnâNDDDÑMÜ]0~üxìß¿ÿùÏ\«C^^ÊËSûö¿þ5››‹ÜÜ\SÛ‡šA%@W2;Hœ•¬’—3èz}ÐíÈ  <þ8Чðàƒêû–³;3hÔÄÜiáfЯ»Î¸\éµA[f¦`Eå>ÌèFïkïfÍŒ×Q+ÇŽ&îòó.ÿÙLÆ]¹å yjt­VÆÇ’ÿËpü¸ÿá'pÌh#"""Šq Ð#lüøñX½z56lØ€Fé®Û¤I-+**BÍš5‘’’¢»NÓ¦MuËž7oÊʲг'к5°re##\ÔÞ‚ƒC§š¸«í×í&îfû ›ÙÖÉ>èòm23ý_fôHfÐ#Ù]ù³ô:ÔL¿Sº3}Еõ0ÛĽV-àšk€uëÜïƒntôÆ4Ðj1`&ƒn|írùòè?w.•U “цDDDÃØÄ=‚Æ+V`ݺuhÞ¼¹áúÙÙÙ(** ZVTT„²IµÖéÞ½»aùád#õ²·ZË«W½‰»rßvO³¦µ­²NèvõAW+ßl݈2èNÒë’ _Gow½ÏG$ôP2è¡èf2×ò÷ ЦY³2.€•>è$ŽˆˆˆœÀ=BƇ×^{ ¯¿þ:QRR‚’’œ9s&°Î#<‚‘#G^ÿö·¿Å©S§Ð§OlØ°cÆŒÁ²eË0iÒ¤À:&LÀ§Ÿ~Š¡C‡bÓ¦M¸é¦›PTT„ þö”ºÌ$Z”7›fš¸+3èV2ádÐÿò@6ˆ¾)vèz™Öpû ÿðpâ„öúáôA×{ßí º—›¸ií3í•]¯/¼^ÐkåºÖ qjëíC¯ºVf°:™Å=B^xáœ8q½{÷FãÆ_ÿøÇ?ë”””àÀ× 6Äƃ BAA^{í5äȆYÏÈÈÀêÕ«qðàA 0%%%X³f ÚµkgX';3èftéçóç­7qýuàŽ;üS'™©{ýúÀÍ7û¤¦šÛ—ÄËMÜ¿ú HK–.­¼¾4‡}¸}ÐÕ8=Íš|?J‘È>ªmj×.Ô&îj¿3NŽâ®68š§3èòAÌ kÑ$Ή&îjz™A'"""³Ø=BÊMD1‹/®´,##ëׯ×ÝîÚk¯Å¶mÛ,×ÉÍ]ëƒÄÍ <´jåÔÎHõêÀ[oU¼NNVÏ:«Ñ ¾ê×7·½QYZMÜ»v~õ«Šåf3íås€õ>èz†©}Œí <ÂiÑa'£ ¾^w=‘Я¹س§¢|­}+ëN€nµ»‰«Ó¬5q×Ë k=(“ÿ=Óš‚NëX>ùè×Ø°!x]èDDDd3è1ÌÎ&îzåH%eph5ƒ~ú40l°q#0c†ùºÕOVý¦›ü£Þ‡»ßáÃ_Kçâ/ ÎŒ[ ÐÕ§û G"ƒɠݨº]t»¦Y“›7ÏßÊ>Ž¤$ím´®«² 9+MÜÕúsë±:ÍšQw+ƒÄY™fMëXjÔ¨|¾ÙĈˆˆÌb€ÃìÌ k•·{70q¢ÿgù q€µýÜ9f°vmëuÕ«Ÿ­Ýç Îp‡R4nüZèmk`kNöAD=’ÙG£]oš5=Ê͉>èññs”Ëëwë­ÚÛÄÅig‰õ‚^£¦èjåØÑÄ]Y£¿#jåHçÝL=”AâÔÖeˆˆˆÌ`€ÃÌÜoÝ ¼ñ†ÿgéÆ0 ·iSqÓ;dHð{ æê)çv€ÊþÔ¶Uöe—zû5 FÔyp`e”qµ÷Õ®ûÿýŸþvVY |í¢×ß\þÚÎ&îNfU•™f½õœìƒ._Ç®]ÞŠ!”QÜ¥ãU^Cµ úرúõÔ«Ÿ|]èDDDdû Ç037Ö]»ú¿.ºÈ?_ºÄ(@ïÚw‹?uÊ?wò«¯V,3 ËËT€^£FèÛ*Yéƒn€®ÜÈ›×êmo¦º“A†Wú ë §‰{$2èÊòÍ®§u]µÆluš53ÌèÒt‹VFq—ÊÖʠ˯­ôÞ?¨—o4·¼t"""2‹ôf% êÛ¸äsåÀ{ï÷Õ®]»òÍ°—ôK/,°¶í;ïjãù™ÙÖî]~®¿ýØ´ÉÜözœÐ%nO³¦·¿pš¸KçnÙ²Š²Nž ¯~zû± keÐVßÆê4kʺ™YÇj]Ëõ×W.[jÒ¯ÜV­‰»Q=Õ(?ìƒNDDDf1ƒÃììƒ.ŸJ êÔQñ\¾N(MÜ­o#±:Gõu׿VûYnð`í²ÌÖÍ(@7ꃮhú¿Œ‚£zJýŽ}>ÿœßv 53I¡Œ¡ ‘®4†AA=uÒbö<êè?þh\¶Úçææ›+DÈ×q"@7Ê  \~9ЫWÅ2+MÜ%jA·Qýäë2ƒNDDDfxøV˜œN€.e²›7W/ÏLsSe3r-ndÐà`8œ>èf%»2èiŸ]ºóçW”e&;mfšµ%KÌÕÊhiâ. 5ƒ.mwú´=uÒbö<ê&ÿÇ?÷ߤ§WÞFígÉ[o/¿\¹.Vt3ŸÕ¸8ëž»ïöOmf´¾DÊ ›$N« èDDDäè1,œ€¨U+àÃ?ü¡ryfËÔ›úIέ]> Z8A£•&îFƒÓYíƒîóU<‘÷« ¥ž^hâî$3ƒÄÉYHPx:}œv4qoÛxæ™ÊÛ˜éƒ~ãþ)¶ ̓./.ÎßåïZ¶4.[¹åïÓsϹ¹þåFçÏjt6q'"""3 Ç°p3•×]ܼÔj½E ëûŒÆÝL æÔ(îB„ß@¾­Õ½cGkåË¿Ëy%ûèóyyþŒíµ×¿gTGåµ g<=vöA×b&@¯_øüs Ag›¸[aÔ½Càõ×éƒî•Ï0yû Ç0»›[-ÏÊ<è’+®°¶­rÌ7@·ÒBÁl÷PæA—Ê–F½Öbw½¤ÄßJÂlK ‰×û ¬=Þ€eÖÖ©]b%@·¯^ù ‘·1@aNfÐØg¸õ çÙ«ƒÄù|þÑí¥>ÞF}Ðõú‚Çz}Æ ÿÜÝrVš¸½>ènö­Dô„`Õ* E ëõSn€n†ÖµWèjuðòC&"""ò6è1ÌÎ>è>ŸsA†2ènõAWn«ìƒ®Ì˜ë̓©©À‹/×G›ôHì{Ô(à‡œ+_¯ºÛºQð¨—Aw#ð43y¸”×DùCÙ"B./¸ÿþàe¡Ì€@DDD±‡z ‹¶ z¸" [yj]+@W›]2fŒ~}õ¦Y“332»zt¯díìƒn× „JáL³f%ƒîæïriiE=¬°rn”ǧö»¥U‡^½€gž ^Æ>èDDDdôæF€.±2Å”]™þP‚¼^½Âß·“}е‚­ º«MÜ/ºøñÇàevô¥Ž•i֜Π[ ¶ÍŽâþôÓÀ“Oú¯½›¤Y ¬çD€nv6 èDDDdôfgw«ƒÄYÉ|Ù¼„ÒÄ411ü:8™AW®§6Š»ÙÂj€^»6pñÅÁ˺t1·/½òÝÂå¬ôA7¢×ÝÍ‘ëåt-ÒïNv6ðÐCî_£ŽýßCÅÝ ­²Í4qWÃ&îDDDdôfWßóPÖ·’A·ËéÓÖ·±#hÔÚvÿþŠŸ¥ÁÉŒ¦YÓ ¶µš¸öke]¤‡’Y³€»ï?^šfÍJ€n”U>šœV`¢×ÄÝjt£]êg.mvóR])œÏÁ„ @ݺþ)Ç”jÕ ½ÜpÉû ‘艉Àu×9S'#¡è¡öAo×N»ºÙ߯âbóû&""¢Øåá[arZ´dÐC)ßÈ-·XÛg8£¸‡pZ ÐÕ‚»‰kÚÔ\yVEC=sç?ýTÑ¢A:Î *Ï¿'š¸K”ÝÜbµ‰{8¿¿ÊŸC ЧO·^"""Š=lâÃÜ„Ü$Nº¡^·¸újkÛ:A7»­2¸“ØÑÄ]Í–-þ R^fBðÔSöä­×§×­iÖ¬\;3ëª[V–ù}Xag÷+€¯¾òÎÓH7q‹«üùܵËù:Qìa€ÃÜÈ »ÙÄÝjÆK¾o§§Y3Ú¿Äh´÷p‰S;ÆnÝ*¿_£ðàƒæÊ´BmŠ8%¯ˆvrû˜”㨹þzÿ—WD:@—Ÿ#å>Ý$ˆˆˆª|ÏÆ¥ IDAT.>ûanôA—¸ÙÄ=”¬h8MdíÌ &jA„]ƒÄÕ¬éÿîÔ `fFÅv+“î¯4é7“A÷šHâ®7 t"""²3è1,V3è‘ Ð•eر­V€®×ÄÝ®>è7ßìäÌ©Á¬N[I;;vØ[¦SÁ°Õ‡òAâ¢%@ô qjMÜ•u!"""²ƒo…)Rܼwã¦6”]-MÜíÅ]Kb"0x°ót/5q—öwçÎeï:¦Pú G «ºƒÄi½&""" o-b˜Ml…°–Ae87(HõôJ]k8eótùk«ƒÄ¹ÕŒÜ‹ôH|æÜÎZGcw)@wòAŸÙ&îá ÂHDDD¤ä¡[aŠ4;nÆõn\õÖ¶&îá°3ƒ.e¯»vU߇݃ÄER8-ÜT½zhÛy¥‰{4è‘×ÄÝK”ˆˆˆ(úñÙ s£º$Zš¸»Ý]yó_¿¾z§Zµ€–-+ïC­‰»]ƒÄ9Í‹t#›7çϽ{‡^†ÛMÜå}У…VW;™mâÎ>èDDDd§(º&»¹ÑtÜÍyÐÃNV;3èжmå~àZÓ¬ÙÙÝizPܽ]kÿÝ»©©á•íöyæ ºÕ ¡ÇiÖˆˆˆ(R Ç0;›oû|ÎeнÐÄ=œÑ‰ì»r¹²‰»üýhé#ëÅ º•nV¯³SȬf—£9@¿pÁÜú¡üke͵~÷ˆˆˆˆìÀ[‹fÇÍø€À¨QÀÀÎõAWn*·t»3èje*›¸Ëëmôh x.¿˜<˜5ËÚvNïéÓiÓ€NÌ×Ãí VIŸi«ºf›¸GË07=ù䓈‹‹Ã<àvUˆˆˆ<·1ÌŽ ^j*°h‘ÿç£GÍï3” º]A„ƒÄÕ©ã0îþúôñϱmu[åò:uüß{ô¨¼®Ú¹nÜ8tH½L·û {iš53|>à‰'ÂÛÞNÉÉþ ÝÊþ£-ƒ.}¦ì;o¶‰{´=PŠ´;vàïÿ;:tè_´|Àˆˆˆ\Ä=†¹y¯d%@ÿÕ¯üß÷ì oŸážÊmjÖÔßføp`Û6ýóüÑGúe˜Í ''%%@ZZðr)ƒ>n\Åy€ââÊÙÇP?+V]Ú¶r^Π;ùÐÂí˜%.Î?Ð]4±ÚÄ]ú]µò9U6kïÒÅÿsv¶özF¢±µB8N:…#F`áÂ…x"œ§XDDD1„z sc÷PNëØxðA`ðàÐê% %ÔÊ*תe¼íý÷û¿Âa6ƒ T^&ÍOÿÜsÁËë—ܯüºzuàÜ9ëåèeÐ;uFŒzÈz¹áˆD«·ôhΠ› ÐÛ¶^{ :Ôü>”ÍÚÛ·÷”3¬¶Š¥ý¾ûîÀpÍ5×àñÇw»:DDDQz ³ûfÜLàÛ¤‰ÿ»ro#³g[¯’[}ÐÃa%@Wc¶¾¼Ìpõ‹/€½{­o§÷%!xõÕðêå5^ †c!@€Ûn³¶³}УmFŠHyã7PTT„;v›·™Äl»i4SNÇŽþ›j7š1O™üü³¿fyõžÒÌù³hKƒ]5lh½>r­Zù¿¬Š¶ Ñ.no4fuCfÍ ­=œQÜݾ֑ràÀL˜0k×®EBB@a⃖——‡””” e¹¹¹ÈÍÍu¤®DD±$??ùùùAËŽ;æRmH tŠ8·ú§¥ „¶­×'nô[´. ¿+A¨¼Ø=•ÛA[\\E&Úíº˜Õ¸±ÿ»"Ž³•ÙiÖܘ2Òë>ýôSüðÃÈÊÊ ,»pá6mÚ„çž{gÏžǪ̃ϛ7/h;""²ÚÏ‚‚t2;õ EôfwÐYo>¥&ù¿$<É ÖçF¶¿\³¼ Kœx@ãÔ<è¡Ô#ÚZ/Üv›Ì…ë®snzYs9fÐ+»îºë°k×®Àk!îºë.´iÓ=ô›»é`€Nmâmþô'àšk*2v^QÏu(c8Íɺxå8£±‰»Ïçlp.íCíg%è•%%%áŠ+®Z–˜˜ˆ‹/¾¸Òr""" æÁ\E+/f>ÃU³&pà n×¢²ªx®½œAw’ÛA[4fÐ#Ao`896q7Ççó1sNDDd3èd[óí¤${Ê!cUñ>׋t §Y‹=Ì Ûë£>r» DDDQ!ÆrU$—’⟾ìÅÝ® YUoô½˜A•A✠=ZÙ çåi—IDDD¤ÆC·Âä†ô¶DÑÅKA¬]b-ƒ‰ãlßxî9ãz0ƒ^™MÜçΞ}Ö\9DDDD›¸yšVÀT)/fЉQÜ?ûÌ\=t¾.ÑÆ®&îÒvDDDDfÄÈ­0QtŠÅ½*›··Nw÷ïUvèòîn_k"""ò>èDä çÏû¿W¯în=äbašµ?ÿ¹âg¯ÔÉ Ìž +tž_"""2ˆ<áÜ9ÿw»f°SUÅ]`º]/1{.Œ¦Y“è±Ò}ƒˆˆˆBÇ>èDä ?ÿìÿÉýèQý÷¥l~ÎÕÁí ØíýG»ôtý÷9B>YÁˆ<Á z½zúï÷í Lž äæÚ¿ïø_þúº 3ƒºÿèÝ[K.©ø™ç—ˆˆˆŒ0@'ò°‹/v»‘ãFÝHíÚÀO8Söþ”–º?Í¡ÙÁШÂÛoû¯Ý5ׯ››ëȳ?›¸‘1èDÖ¡ƒÛ5ˆ/èNºòJà½÷Ü®ƒÆPÜt“ùu}> sg€Î DDDd„·fDK7ô^$®*cÝy<¯DDDdôÚ¸q#ˆ&Mš ..+V¬0ÜfåÊ•¸í¶ÛP¿~}téÒ+W® zÿå—_F\\\ÐWµjÕ𳔎$ŠÒGÖÉÙ¨2öAwžt^y~‰ˆˆÈô*--Eff&ž{î9€ÏànmË–-6lºté‚õë×cØ°aÈÍÍźuë‚Ö«[·.JJJ_‡FÓe˜AwƒFçIç˜Ý ˆˆˆÈû GPNNrrrL¯ÿüóÏcÈ!ÈËË´k×[·nÅSO=…kd£ù|>¤¥¥Ù^_¢H’2èÒÔfÌ ;O:Ç<¿DDDd„Ïó=ìøñã¨]»v⤤$lÞ¼9hÙ©S§žžŽÌÌLÌž=Œd5‰l!eЙeŒ,Îã9&"""³x+ìa£F›o¾‰÷Þ{§OŸÆÊ•+±lÙ2”••áرc€6mÚ`É’%Xµj&Nœˆ5kÖ cÇŽ8{ö¬Ëµ'²&=ÝíÄ&fÐÇ>èDDDdt8p &Nœˆ‡~uëÖŤI“0aÂ!ïïеkWÜvÛmhß¾=n¿ýv¼ÿþû(//ÇÛo¿ír퉬yã `Ó&·k{8Š»ó¤‡ lBDDDFØÝÃâãã1uêTL:ÇŽCJJ žy椥¥!))Iu›Zµj¡oß¾X¾|9rss5ËÎËËCJJJвÜÜ\Ýmˆœ”’ôìév-bƒFgäçç#??PTä_öãÇ\¬QdãäÉ“!o¿gÏkCD= G‰””!ðÊ+¯`È!šë9sk×®ÅðáÃuË›7o²²²ì®&EfÐ!à9j°x1P¿~ìär͈œW\\ŒË.»ÌíjE%ètúôi^ýõ×(**B½zõЬY3<òÈ#8tè–,YÀÿnÛ¶mèÚµ+¶lÙ‚Y³f᧟~´iÓe<öØcèÖ­Z¶l‰]»vaÑ¢E(++ý÷Þñã#¢èÃ>èÎã4kk*2çK´ ±”Õ¦ØS!"¢(Â=‚vìؘÍçóaâĉ€;ï¼/½ôJJJpàÀÀú.\Àœ9s°wï^Ô«W½zõÂœ9sРAƒÀ:ÇÇرcqäÈ´o߃Æ_þò>¹&"S”;O ÌËËÝ­Qäµjk=6q'¢ØÄ=‚úôéƒr;´Å‹½nݺ5 tËœ3gæÌ™cKýˆ(ö0ƒî<é¼–•¹["""ò>6¸#"Šaìƒî<é¼vîìn=ˆˆˆÈû Å0ö‹vžtŽÓÒÜ­yo͈ˆb3èÎã qDDDdoˆˆbû ;:™ÅÛ"¢Æ Üy ЉˆˆÈ,Þ.Å0fÐ'ãjÕÜ­y§Y#ŠR_ tèàv-(Ú±ºó˜A'"""³ E©½{ZµÜ®E;Îc€NDDDf1@'ŠRõë»]ª ˜Awž˜3@'"""#¼] "Šaìƒî!"""# ЉˆH§Hç•t"""2ˆˆÈAìƒNDDDfñvˆˆ0ƒî!üß ‘‘x·+@DDÞÀ=ØÈ‘@­Zá—ȈˆÌb€N¶úøcàÐ!·kAD¾—_¶§œòrÿwöA'"""# ÐÉVÙÙn×€ˆBÅ º3¤ :Ï/aƒ;""ÀÒ)R€Î :a€NDDä ©‰;û ‘Þ.fР:™Åˆˆ0@w û ‘Y$ŽˆˆÈAňªš={ö„]F:uЪU+jCTµ0@'ò¸ÚµÓ§Ý®ÅfxÁyЉ¨êø0bÄ[JÛ·oƒt"èDwàðóÏnׂbtg0@'¢ªCÊ,Ð&Œrö“'O†_%¢*†:‘Ç]t‘Û5 ¢ppw"ªzÚÈr»DUoˆˆ3èNa€NDDDfñvˆˆ0@w ›¸‘Y¼] ""rt"""2‹· DD€t§pš5"""2‹:`€î)ƒÎóKDDDF 9H ЙA'"""# Љˆ3¼Ná(îDDDdoˆˆt§p8"""2‹· DDDbˆˆˆÌâí`Ý)ìƒNDDDf1@'""rGq'"""³ 9¨vmÿwfЉˆˆÈH¼Û ""ªÊzȤ7jävMˆˆˆÈë˜A'""rPË–Àܹ@<‰‘èDDDDDDDÀˆˆˆˆˆˆÈ yt"""""""`€NDDDDDDä ЉˆˆÈV3gÎDçΑœœŒÌÌLLŸ>ÿûßÿÜ®‘çqÒ"""²ÕÆqÿý÷£sçÎøöÛo1sæLlÛ¶ kÖ¬q»jDDDžÆˆˆˆl%ÄÛ´iƒfÍš¡]»vØ»w/.¿ürkFDDämlâNDDDŽ*//\¸pÁåšyt"""rLyy9òòòзo_\qÅnW‡ˆˆÈÓØĈˆˆ3~üxìß¿ÿùÏ ×ÍËËCJJJвÜÜ\äææ:U="¢˜‘ŸŸüüü eÇŽs©6¤…:9büøñX½z56lØ€F®?oÞŸOwý-[¶`Ø°aèÒ¥ Ö¯_aÆ!777(/,,Ä 7Ü€Fáý÷ßGƒ п|þùçŽ U Љˆˆˆ¼MÜ#(''999¦×þùç1dÈäååÚµk‡­[·â©§žÂ5×\˜?>®¼òJ¼ûž={";;óçÏÇÂ… í?"ªr žQ„0oâaÇGíÚµƒ–%%%aóæÍ×[·nEfffÐ:™™™Ø²eKDêHDÑt""""oàm™‡5 o¾ù&Þ{ï=œ>}+W®Ä²eËPVV†cÇŽ:„ŒŒŒ í:vìˆï¾ûÎ*Qb€NDDDä lâîaÄĉñðÃcÏž=¸ôÒK1aÂ̘1ññá]º¼¼<¤¤¤-ËÍÍEnnnXåQôa€n¿üü|äçç-“¬ia€îañññ˜:u*¦NŠcÇŽ!%%Ï<ó RSS‘””hÒ¤ ƒ¶+**BÓ¦MuËž7o²²²«;EèöS{àYPP€N:¹T#"""Š¼-‹)))Bà•W^ÁСC˳³³QTT´nQQºwïé*Q”b€NDDDä ¼-‹ Ó§O£¨¨(Pýõ×(**Â<òFŽX¿¸¸K—.Eqq1–,Y‚víÚáСC˜6mZ` &àÓO?ÅСC±iÓ&ÜtÓM(**„ "{pDµ yoË"hÇŽÈÊÊBVV|>&Nœˆ¬¬¬@À]RRÖàÂ… ˜3g2220eÊddd °° 4¬“‘‘Õ«WãàÁƒ0`JJJ°fÍ´k×.âÇGDÑé­·€;îp»DDDDÄ>èÔ§O”——k¾¿xñâ ×­[·FAAa¹×^{-¶mÛvýˆ(6]{­ÿ‹ˆˆˆˆÜÅ :‘0@'""""""òèDDDDDDDÀˆˆˆˆˆˆÈ yt"""""""`€NDDDDDDä Љˆˆˆˆˆˆ<€:‘0@'""""""òèDDDDDDDÀˆˆˆˆˆˆÈ yt"""""""`€NDDDDDDä Љˆˆˆˆˆˆ<€:‘0@'""""""òèDDDDDDDÀˆˆˆˆˆˆÈ yt"""""""ˆw»DDDDäÅÅÅ8yòdÈÛïÙ³ÇÆÚÅèDDDDÀœ_vÙenWƒˆ(f1@'""""eΗhb)«L±§BT¥…ÛÚ¢N:hÕª•Mµ!òèDDDD¤Ð@VˆÛ²‰;ù0bĈ°KÚ·oƒtªR Qþå{8-5öÖx D^Ĉˆˆˆˆ\NK ¢ª‰Ó¬yt"""""""`€NDDDDDDä Љˆˆˆˆˆˆ<€:‘0@'""""""òèDDDDDDDÀˆˆˆˆˆˆÈ yt"""""""`€NDDDDDDä Љˆˆˆˆˆˆ<€:‘0@'""""""òèDDDDDDDÀˆˆˆˆˆˆÈ yt"""""""`€NDDDDDDä Љˆˆˆˆˆˆ< Þí ‘=Š‹‹qòäÉ·ß³gµ!""« UÅÅŸì²ËÜ®…:QP‘9_  Mˆ¥¬0Åž ‘e Љˆˆˆª”6²BÜ–M܉ˆÜÄA∈ˆˆˆˆˆ<€:‘0@'""""""òöA'""""¢¨îÔ€uêÔA«V­lª Qø Q”ù0bĈ°KÚ·oƒtò èD¿ÈÏÏGnn®ÛÕð$žuâââÀ믿þEEE¨W¯š5k†Gy‡Â’%KcÇŽÅW\ǃ ÂÞ½{ñÄOàŽ;î@bb"à±ÇC·nÝвeKìÚµ ‹-BYYî½÷^WŽ‘ˆˆ&Mš„Áƒ#)) ÇÇÔ©SqæÌŒ;ÖíªyGq ;v ++ YYYðù|˜8q"²²²0mÚ4@II 8X¿uëÖX¾|9Š‹‹Ñ¯_?<û쳸ýöÛ1uêÔÀ:ÇÇرcÑ®];̘1ÙÙÙ(**B›6áÌ£JDDžà•W^ÁöíÛ1hÐ ÄÅÅaãÆUjt"""Û¹Ý ž"géÒ¥¢cÇŽ"99YôéÓGìܹÓí*9jÚ´iÂçó}5jÔ(h9sæˆ6mÚˆ‹/¾XÜpà âÛo¿ zÿÂ… â‘G—^z©HKKÆ ?ýôS$#l6l 7>ŸO¼ûֱã<”––Š±cÇŠ¦M›Š¦M›Š{î¹Gœ9sÆÑc ‡Ñy9rd¥ÏO·nÝ‚Ö©ŠçE!þüç?‹+¯¼RÔ­[WdddˆiÓ¦‰o¾ù¦Òz±ö¹1s^bùs#„¹¿7J+V¬·Þz«HKKݺuÓ§O'Nœ¼¿xñâJç4..Nœ={ÖsÇ‘››+êÕ«':wî¬:ì‡~(ºté"êÖ­+ºví*þýï;Qý N‹×Äìß%¥ÂÂBqÕUW‰ääd‘™™)òóó+­ÉëâÔqDúš„r‡¹¹¹â²Ë.qqq"//Ou½Hÿž8u,ÑpM–/_.®»î:‘šš*š5k&î»ï>ññÇWZÏ¿]±ŒzŒX¹r¥¨V­š5j”X¿~½èÙ³§HNNßÿ½ÛUsÌ´iÓDûöíÅ‘#G_G ¼ÿ׿þU$$$ˆ?üáâƒ>íڵ͛7?ÿüs`I“&‰¤¤$ñÔSO‰eË–‰æÍ›‹ììl7'dkÖ¬S¦Lï¼óŽêìv‡¡C‡ŠÔÔT±páBñâ‹/Šúõë‹aÆEäCat^î¼óNqà 7}~”ATUøàÑ»wo‘““´N,~nÌœ—XþÜaü{¥TRR"jÕª%&Ož,víÚ%Þyç‘žž.þô§?ÖY¼x±HNN:§GŽñÔqlÞ¼YÔªUKÌ;W|þùçâ駟‰‰‰A7±¢zõêbРAbÓ¦Mâ׿þµHHHŸ}öYÔ‹×È>ÐG4IDATÄÌïŸÒ¡C‡DRR’¸êª«Ä† Ä]wÝ%ªU«&V¯^X'Ò×Å©ãˆô5 å8¾ùæ1aÂñꫯŠÌÌLñÀTZÇߧŽ%®I^^ž˜={¶øä“OÄ'Ÿ|"î¾ûnQ¯^=QZZXÇ­¿]±ŒzŒèÝ»·ÈÍÍ ZÖ°aCñÄO¸T#çM›6Mddd¨¾W^^.Z´h!yä‘À²Ó§O‹jÕª‰W_}U!ĉ'DݺuÅ‚ ëìÙ³Gø|>±iÓ&g+ïåÍ™]çá¿ÿý¯¨V­šøç?ÿXgÍš5¢Zµj¦2nS»i9r¤¬_¿ 6D¿~ýðÊ+¯àÔ©S÷cé¼”——.\¸€Ÿ‰ò¼üÜX•““ƒÒÒR¼üòË(--Å®]»°|ùr 2$h½S§N!==™™™˜={6<èRÕ?~µk×Z–””„Í›7^oݺµÒÿáÌÌLÏý6s,€û×Dí÷OiëÖ­ÈÈÈZ¦¼÷qûºØu€»×ÄÌq˜áöõì; ú®I\\.\¸´®I¬a€~üñGœ9s¦Ò÷Ž;â»ï¾s©VÎËÎÎÆ«¯¾Š>ø·ß~;^zé%ôîÝÇ­öOzï»ï¾C\\Ú·ox_z]UΛ]çá»ï¾CÆ ƒjØ°!4hµçªÿþxýõ×ñ¯ý }ûöÅŒ3ð›ßü&ð~¬œ—òòräåå¡oß¾¸âŠ+ðs¨Ÿ€Ÿ«.¹älÛ¶ yyy¨[·.:tè€þýûãñǬӦM,Y²«V­Âĉ±fÍtìØgÏžu±æÁF…7ß|ï½÷NŸ>•+WbÙ²e(++ñcLJŠŠÿÃfŽÅík¢õû§¤vÎ322pæÌO\;ÃÍkbö8Ìpû÷ÄÎc‰Ækò§?ý qqq¸í¶ÛËܾ&±ˆÓ¬Q••““ø¹}ûö0`Z´h?þ ªÛ!àóùtËsÐW%< ª:tè€îÝ»£W¯^8xð š6mª¹]U;/ãÇÇþýûñŸÿüÇpÝXúÜh~n¬ùì³Ïpýõ×cìرøàÑ¡C‘žž.Î;(ãÁuêÔ³fÍË—/ééé•F²ôºS§N‰ÂÂBQXX(|>Ÿ˜;w®(,, ÌYm×y¸ùæ›EZZšX´h‘X¸p¡HMM­4s€—è—S§N‰I“&‰?þX|õÕWâïÿ»èСƒhݺµ(++ ”QÏ‹BÜ{ï½"%%ElܸQ>|8ð%?öXüÜ—XÿÜaü÷æá‡wÜqG`}iûÇ{L‰7ß|S´k×NÜÿýu¦OŸ.þùÏŠ¯¾úJ¼ûî»bàÀ¢V­Zb÷îÝž9Ž}ûö‰W_}UìÛ·O¼üòËâŠ+®5%%%u EBB‚2dˆØ¸q£,êÔ©#z÷î-Ö¯_/F-ªU«&Ö¬YX'Ò×Å©ãˆô5 å8„Ïb§NÄðáÃEaa¡øâ‹/‚Þôï‰SÇ ×äõ×_ñññâùçŸÚæøñãAÇéÆß®XÆ=†¼öÚk¢cÇŽ"99YôéÓ§ÊÏ_8lØ0ѸqcQ³fMÑ»wo1wî\qàÀ uæÌ™#Ú´i#.ºè"qã7VzÿÂ… â‘G—^z©HMMÆ ÇŽ‹äa„í£>>ŸOø|>øù®»î ¬cÇy(++cÇŽMš4Mš4÷Üs8{ölDŽ1z祬¬Lôë×O¤¥¥‰ÄÄDÑ¿±`ÁñÃ?•QÏ‹¢Ò9‘¾–,Y´^¬}nŒÎK¬n„0þ{sçwŠ«¯¾:h›·ß~[Œ1B4hÐ@ôêÕK<õÔSAçìÍ›75kÖ;w3fÌ{÷îõÔqìÙ³GdffŠÄÄDѬY3qÛm·´’µk׊.]ºˆºuëŠììl±nÝ:Gécq㚘ù»¤öù*,,½{÷ÉÉÉ"33S¼ñÆ•ÊŽäuqê8"}MB=µÏb‹-‚Ö‰ôï‰SÇ פOŸ>ªÛÈï…pçoW,ó £#Óyû yt"""""""`€NDDDDDDä Љˆˆˆˆˆˆ<€:‘0@'""""""òèDT%lÚ´ W^y¥ÛÕ "¢òÄO == 6Äï~÷;|ýõ×nW‰ˆ¢t"²Uzz:¶mÛ†éÓ§ãÞ{ïut?Û·o¼îÕ«>ùäÇöGDD¤Ô¶m[|úé§Ø½{7Î;‡)S¦¸]%"ŠrñnW€ˆªŸÏv?ÿü3 ÷#„{_DDD¡2dHàçÉ“'£cÇŽBØò¿ˆb3èDd»õë×cæÌ™X´hêÔ©ƒAƒ~üñG<ùä“hÕªFŒÏ>û,°Mzz:^xátêÔ :tŒ7Mš4AóæÍñè£âûï¿Œ3ß~û-®¹æÔ©So½õÖ¯_6mÚÊÛ½{7ÆŒƒ-ZàñÇÇÑ£Gß|ó jÖ¬‰üü|´jÕ Ýºuÿþõ¯H""ªB–-[† 55íÚµÃO?ý„ƒº]-"Šb ЉÈvW_}5}ôQŒ3'OžÄŠ+#GŽDYY¶oߎ¡C‡"'''°ÏçÃÂ… ñâ‹/bçÎüÍÖ¿üòKlÞ¼_~ù%æÏŸX¸p!.¹ä|ôÑGøÿöî&Ħ>€ãøïšÑ”šÅAc¸²¢”4¹e&R%)YÌBÚPJwa!7¥dg¥(;KK ”•—ka²e%C3 Ë;JÍ”yVÏ}LÑó‚Ç?>ŸÝé¼ßÅé|ïÿtÎììlFGGíaa!{÷îÍ–-[Òn·óúõëœ9s¦;~~><ÈãÇsüøñœ>}úGÿ$üb>|ø“'OæìÙ³™œœÌÄÄD’xº ø&øîþ¼9ùü&¥Óédbb".\HUU9|øp6lØv»Ý]æèÑ£i4éëëK’9r$ýýýY»vmšÍfîܹóö?>>žþþþœ:u*ƒƒƒ¹téRîÞ½›OŸ>uëܹsȱcÇ233cÄ€evv6?~̪U«ÒétÒjµ~ö!¿ü/=z”ééé¬\¹2UU¥ªª<þ}š‘‘‘Ôëõ\½zµ»®ÏßP¢Ú‚WMÀOg Р Р Р Р Р Р Р Р Р Р Р Р Р ðXÐkŠ ÅžIEND®B`‚HDDM-0.2/hddm/examples/plots/simple_difficulty_hard.png0000664000102500010250000012540111726176127023756 0ustar wieckiwiecki00000000000000‰PNG  IHDR Xšv‚psBIT|dˆ pHYsaa¨?§i IDATxœìÝy\”åþÿñ÷ ; ‚¸&.`˜$ifžÜñhæ†ÓÔ —ÓùššZϯ,E+ÊN‹Ö)[LÍLËÌÓrÁ-ÓÜWÔ\qß•‘m~ã‚ ¯çãÁÃ{®{ûÌ0ż纮û6˜Íf³ÀŒŽ.@ùA`7vC`7vC`7vC`7vC`7vC`7vC`7vC`7vC`7vC`7vC`7vC`7vC`7vC`7vC`7vC`7‘””¤qãÆ©I“&2ºçž{Jå<‡–Ñh´üxyy¸m—.]T¿~ý"×h4ª]»v%Uf‰0`€ŒF£Ž=zËÇxå•Wäáá¡cÇŽå»~øðáV¯í„ nù\EÕ¶m[ÅûÓ•ßïjüøñ2ZµjU©ž[’¾üòËbŸ«¼™5k–ŒF£Ö¯_ïèRØ€C úè£ Ök¯½V"Ï@Ùæêè”?üñ‡~ýõW5Jo¿ýv‰óf ###5nܸ×ÇÅÅÉ××Wƒ *‘zʳٜo{Q?DéÑGÕ§Ÿ~ª±cÇ*00Ðj}çÎÕ¹sg%&&*!!á¶ë--IIIòöövȹøámÛ¶Mñññ9AÚ¶m«¶mÛÚ´·lÙ2ß퇪>úHuëÖÕ£>ª“'Oê›o¾Ñ?ü ùóç+**ÊjûóçÏ«k×®Z·nî»ï> úè#ÝsÏ=Z¹r¥å¿‘e˖顇Ò“O>©ßÿÝjxÚÛo¿­uëÖiðàÁš:uª¥ý•W^Qll¬^zé%Íœ9Óê’›››Íð½_|Qš;w®RRR¬Ö«[·nš7ožŽ?^ŠÏ€£@YÅŠÕ¼ysmܸQW¯^µ´¯^½Ú –-[fµÏ²eËd0,èbccÕ£GIRtt´Æ¯ñãÇ+22Òj¿×^{M£GV5£ôôt½þúëjß¾½å\uêÔQll¬*Uª¤J•*YŽ5~üxõìÙ³ÈÏkýúõrssÓwÞ™ïú‚†,]¼xQ:tД)SÔ A5kÖL¿ýö›zõꥯ¾úÊjÛ³gÏjÔ¨Qºzõª:uꤧŸ~ZÕ«W×’%KôÀhîܹùž7--M=zôÐo¼!___ :Ô2ßåûï¿W«V­´bÅ µoß^]»vÕ®]»Ô®];%''ùùßLDD„.^¼¨¤¤¤;f^ÉÉÉjܸ±ÆŽ«K—.©W¯^|¸åƒ\«V­,ëž|òIM›6ÍæCwHHH¾=7“sÞÝ»w—Ê°¹áÇkïÞ½zþùçm&!Ÿ;wN*T°ÙgÙ²eÚ½{·%4¾ûî» ×'Ÿ|¢qãÆ)((¨Øuœ>}Z£GVÅŠõ믿ZŽ÷½WœIè9´ëÕ«Wà6ÅyIÙWï*ŽvíÚ©M›66íŸþ¹>ÿüsËcýóŸÿÔĉU©R%Kû©S§”žž®ZµjåÛ‹“3#o¯[ÎríÚµó­)gŸüBrÞ÷Z÷îÝoúüü5@KTT”^yå-[¶Ì*€4oÞ\ÑÑÑ9r¤öï߯ÐÐPmݺUçÏŸ×#ÉùP¸k×®;fŽ;wJÊþÝúøøy¿Þ½{Û´å\æöVçÀöšæ¼÷ŠcÏž=ªP¡‚ªT©Rà6Å}eee)33SYYY7ýÉÌÌ´élØ°¡F­† ÊÛÛ[š0a‚~úé'F}óÍ7úù矋õ>>ºóÎ;•‘‘¡S§NÝös‘rçååÕ­[7ßöœK¾¦§§[ÚvíÚ¥&Mšè“O>Qpp°{ì1½øâ‹?~¼eNÌ¥K—lŽU³fMU«Vͦ=g’n~¯SNû­<§üä Û*û|9rD’Šu ƒÁï}_ò{Ý‹£°×4ç½WEé©(Î{¨4Õ®][]ºt‘$mØ°ÁÒ^£F I²™‡“ãôéÓ’²ß§·³OŽÒ|¯(;‚ Ø¢¢¢´téR­X±BË—/W… Ô¢E IÙ½‹/VZZšV¯^­»îº«À¡…Ù°aƒ7nlÕvéÒ%%%%ÉÕյľáoØ°¡¤ìÞ‰ÒôꫯêÊ•+š4i’ÍDé!C†¸Ÿ‹‹K¾í9ù6lØ üãVëÌf³~ûí·û—óÚÜu×]%r¼¼BBB$e_šù?ÿùO‰¿8ò¾¦M›6µZ—óÞ+Žððp­Y³F'NœPÕªUK¤Æ’š’ŸÈÈH}÷ÝwVÁ¡J•*ruuÕ‰'”––&www«}rdÞ0‘3ß« ‰íùí“£4ßkÊ€bË;ÔjÅŠjÙ²¥exJTT”>ÿüsM™2E×®]³¹AYQ-Y²Ä沸˖-SFF†î¼óN« ãîîîºråÊ-§víÚòôôÔþýûoiÿ¢JJJ’———ÍÜ•ôôô[ºÚQxx¸¤ì×é7Þ°Z·uëVjÞ¼¹ºuë&-]ºTIIIjÕª•MÏHÇŽµjÕ*uêÔÉr¿‘ÈÈHuíÚµHçºÿþûµmÛ6íܹÓæ~$%å_ÿú—&Nœ¨^½z©M›6ª^½º¾üòK¹ººjРAš>}z±Žçîî®ÿýïzä‘GÔ±cGuïÞ]5kÖÔêÕ«uñâEEEEÙÜ“¥ 9ãîó»Ìª$íرC~~~ÅžQTï¿ÿ¾’’’ôÚk¯éûï¿·Üpqÿþýúù矵wï^Ë¥[¥[›¯SzóÍ75tèPÝÿýêÞ½»å½çëë«ÈÈHmݺµÈÇ»ÿþû%IÛ¶m+•zoEÿþýe6›õ·¿ýM5jÔÐÁƒµdÉ¥¦¦ÊÓÓSñññ6Wk‹ŠŠÒSO=¥?üP‘‘‘êÔ©“N:¥E‹©B… š:uªMÏÐsÏ=§+VhÚ´iÚ±c‡î¿ÿ~mß¾]‰‰‰ ׫¯¾šo};vì«««Í8Î…·$'tèž{îÉw]Ó¦M󽲑Á`¸éP–1cÆè­·ÞRrr²fÎœ)www3FË—/·™ñÌ3ÏhôèѺzõª>üðCÅÆÆëNÒ£F’ÑhÔÔ©S‹¼Oaò{~wÜq‡6lØ aÆiÏž=Z³fºwïn¹‡G~¯ÇÍ^£®]»jõêÕŠŠŠRbb¢~øá5jÔH‰‰‰ªY³f‘‡ íرC’ôè£Ú¬Û¼y³¶lÙ¢¡C‡PnWõêÕµeË?^+VÔܹs5}út;vLÿþ÷¿­.©[Ø{§(ï«›mû¯ýKßÿ½î½÷^ýðÃJLLTûöíõÓO?©R¥JÅ‚Õ¼ysµhÑB³gÏ–Éd*ò~…Õw»†ª híÚµš2eŠV®\©† ꩧžÒ¾}û,D¸Ñ| O>ùDõêÕÓܹsµ}ûvõìÙSk׮ͷ—ÓÏÏOË—/×óÏ?/ƒÁ øøx]¼xQÇ׺uë,óDò:qℾûî;õîÝ;ßKVpsi}•·`À€š9s¦>lõ­÷­:|ø°êÖ­«˜˜Ë÷üôîÝ[?þø£’““:Á~øamÚ´ÉÒk•WLLŒæÏŸ¯#GŽXÝ«"¯œéãÇ¿¥›A:³E‹©k×®š:uªìèrÊ´ØØX½úê«Ú¾};“Ð'G€r!!!AF£Q^^^ù®3fŒ®^½Z¢7ðû+0›ÍZ³fž{î9›ðqêÔ)Í;Wÿüç?ó Ç¿åË,—=ô"##õ¿ÿýÏÑ¥”iiiiúøãÕ­[7ÂP0€Só÷÷Wll¬e8KA7“‹ŒŒÔC=¤)S¦X&Ý–ƒ¡À{ª|üñÇ’¤Ñ£G绾K—.V7ÙËïªFÈ·}úôѪU«Šu¿“òdîܹ:}ú´ÆŒãèRØC°”)ÔÌ™3uèС‚û›}º `io×®NŸ>­;w¸ï“O>©éÓ§kÿþýª[·®¥ý®»îÒùóçuüøñÒ,(³è)@BB‚¼½½Õ§O«ö˜˜íÞ½[7n,pߪU«ÊÝÝ]~~~Ví•+WVµjÕJ¥^ீR€ƒª~ýú67-‹ˆˆ$:t¨À}Ÿxâ ùøøè駟֦M›tøða½õÖ[úõ×_5bĈR­(˸a’““lÓdY_ÐÐPíܹS-Z´P³fÍ$InnnZ´h‘:tèPà~gΜђ%KRàÝšà8©©©:|ø°:uê¤ÀÀ@G—ó—D)›6mRŸ>}T£F =ZU«VÕ‚ Ô»woÅÅÅ驧žÊw¿%K–è±Ç³sµ(®Ï>ûLýû÷wtIÔ¨QC)))6í§OŸ¶¬/Èÿû_={V;vì°ôdDGG«oß¾zá…4hÐ yxxØìW§NIÙoèððð’x£FÒäÉ“]ì„ßwùÂï»|á÷]¾”Çß÷ž={ôØcY>·¡ø ¨_¿¾Ö¬Y#“É$OOOKûŽ;$ÉêêV7ú駟e3Œª[·nš;w®Ö­[§¶mÛÚì—sžððp5iÒ¤žÅ_‡ŸŸ_¹{Îå¿ïò…ßwùÂï»|)Ͽ？Q>Þ¦ý¾ûîÓ}÷ݧ—_~¹4Kt}ûöut °#~ßå ¿ïò…ßwùÂï·Â`ÎûÕ=jóæÍjÚ´©6mÚTn't”e|^»}LB`7 ÁeÎþýûuùòeG—rÆÇÇG¡¡¡Ž.Ãé@@™²ÿ~………9º ”Sûöí#„”2(Srz>>ûì3…‡‡;¸”9w8§ç­ô@@™Î$_À 1 €Ý@Ø €Ý@Ø €Ý@þBBBBT§NG—Ü2À_ŒÁ`¸­ýF£ÚµkWBÕÅC(‡n7Ä·ŠÀn eЬY³Ô²eKùùù©Y³fzá…”––f³Ý¥K—ôæ›oª}ûöªY³¦¼¼¼ªÞ½{+))ÉjÛ3fÈhÌþø—˜˜(£Ñhù™0a‚ÕvÿøÇ?T·n]U¨PAµjÕR‡´hÑ¢Ò}Ò(\]¬½øâ‹zýõ׬^½zÉÍÍM_|ñ…víÚ¥ŒŒ ¹»»[¶Ý½{·^zé%µiÓFݺu“———5þ|-\¸PË—/×< Iºçž{« &($$D °§mÛ¶–å¡C‡ªQ£FjÛ¶­ªT©¢ß~ûMk×®ÕòåËõÆohôèÑöz)à„ eÈîÝ»õÆoèŽ;îкuëT­Z5IÙ=íÚµÓñãÇbÙ¾aÆJIIQ@@€ÕqvìØ¡{î¹G#GŽÔÆ%I7VãÆ-dܸqùÖ°k×.›+m]»vM 6Ô„ ôÄO(88¸Ÿ5Êp ÷Þ+8aÿóV­*ýùù¾D|ðÁÊÊÊRÏž=-áC’|}}õøãkË–-VÛûúúæ{œˆˆEGGë믿ÖåË—åããSäò»Ì¯···þýïkĈZ¿~½ºwï^äãy@€S8qBJNvt·/gÞF§NlÖ=øàƒzöÙgmÚ·nݪI“&iÆ JNNÖ•+W,ë ƒ8 ÈÈÈ"×’’¢÷Þ{O?þø£Ž;¦sçÎY­?xð`‘܈œBÕªÎqÞãÇË`0¨Y³f6ëî¼óNU¬XѪmùòåêÔ©“ÜÜÜÔ¾}{uìØQ~~~rqqÑŠ+´råJ]ºt©ÈçOIIQ³fÍ”’’¢æÍ›«OŸ>òó󓇇‡:¤„„„b¸8…’åHÕ«W×Þ½{µaÃuîÜÙjÝž={tåÊZÚ^|ñEeeeiÎœ9êÑ£‡Õö¿üòK±ÏÿÞ{ïéøñã9r¤&MšdµnâĉÅ>p#.à P†„‡‡K’–,Yb³nñâÅ6mIII ± §OŸÖ²eËò½á «««L&S¾çÏöÌ3Ïج›3gÎÍŸp€2dØ°arqqÑ‚ tüøqKûÅ‹5kÖ,›í»té¢C‡içΖ¶3gÎ(66V™™™ùž£zõêÚ¾}»²²²ò=ž$}÷Ýw–¶ëׯkÚ´iÚ±cÇ-?/   ×ÿûÿOüñ‡š4i¢Áƒ멧žRdd¤jÕª¥êÕ«Ël6[¶2dˆ$éoû›~øaõë×O5kÖÔwß}§Þ½{[m›£cÇŽºvíšZ·n­çŸ^¯¾úªV¯^-IêׯŸ|}}5jÔ(µiÓFC† QÍš55lØ0 :Ô>/œ ŒyõÕW5sæLÕ¯__óçÏ×Æõè£ê«¯¾’›››Õ°ª–-[jÅŠêß¿¿Ö­[§;v¨_¿~Z»v­ÂÃÃó‚õúë¯ëÉ'ŸTrr²&Mš¤ØØX­X±BRöåvׯ_¯1cÆèäÉ“Z¼x±Ú¶m«üQ<òˆÝ^8/ƒ9¿X ‡Ø¼y³š6mªM›6©I“&Ž.‡àï!¡¨ï;ÞŸ·vC`7vC`7vC`7vC”;k®U«øV:uõ”£K€r‡(7Ìf³þóóÔ*¾•Ö]£?ŽptIPî@åÆž3{ôæ/oÊ,³$éÐ…Cº’vÅÁU@ùB”ûÎî³,Œ¨_ý¢ŠîX”?@¹qøÂaËr»vr1º8®(§ €r#o ñ qXPž@å(Øøñãe4µråJG—'G”9ÄÕèªê>Õ-íÉ—’õéæOõòÊ—TPúf̘!£Ñ¨„„„|× ËOYb4Õ®];G—äêè°—ïú~§CéÔÕS–ùf³YL@G.‘»‹»žkñœ*¸Wpp¥@é)(` >\}ûöU­Zµì\ÑÍ•µP„ÛC Ü¨U©–Z×n­^ {YÚ ƒ:‡v–$¥e¦iù¡åŽ*° ³Ùœo{åÊ•&//¡;‘ IDAT/;W„ò†(÷ªÿeyÑï‹X kÿþýzì±ÇÔ¨Q#ùøø¨qãÆ5j”._¾lµÝÑ£Gõî»ïªGò÷÷W@@€6l¨!C†èܹs’¤¶mÛjРA’¤Êh4Z~Ž=*)wȪU«¬ŽŸ3êêÕ«1b„7n¬*Uª¨ÿþÚ°aƒ$)==]/¿ü²Zµj¥J•*©yóæúùçŸmžÓ¥K—ôæ›oª}ûöªY³¦¼¼¼ªÞ½{+))ÉjÛœ!c’”˜˜hUó„ néµBÙÀ,@¹×¾N{¹Ý”ž•®UGVÝ| ”}ñÅ4hL&“5j¤˜˜-^¼Xï½÷ž.\¨õë׫J•*:yò¤Z´h¡S§Né¾ûîÓã?.‡É“'ëòåË9r¤Â¬Öõïß_µk×ÖìÙ³•™™)___IRpp°Íq¼¼¼äééY"5 =ûì³ò÷÷·´EGGK’Ο?¯¡C‡ªzõÜ+ËuêÔIîîî6Ÿ›|}}`süˆˆEGGkóæÍÅ6UÔ×*++«ÈÇDécV!yäýôÓOŠ‹‹Sƒ 4kÖ,uíÚUóæÍSÏž= ÜïÃ?´ùçêÕ«zðÁuï½÷Ò ePD•Ëòö“ÛÕ¸jcVƒ[uéú%%_N.p}%Ï›7:yõd¡Ç¸týÒ-ÕVT;vì$>}ZãÇ·Y_¹re9rD‡R=¨—^zI?ÿü³zõê¥víÚ©Aƒ%ZS­ZµÔ¬Y3«¶zõê)00PW¯^U×®]­Öyyyéî»ïVbb¢Í±¶nݪI“&iÆ JNNÖ•+W,ë ƒ8`5D¬0E}­8`5 ŽE)ÀúõëµpáBMŸ>Ý2æ°sçÎJIIÑرc ááá6m JOO×?ÿùÏÒ*pî¾[Õ}ªëîà»àeû -þ|=|UçFëƒ+Øöä·ÍEÓÅBÏQš>,Izï½÷ ÜÆ`0(%%Eõë×׶mÛ4eÊ}òÉ'Z¾<û*nuêÔÑ‹/¾h™x~»êÖ­›o»···*V¬˜ï•³¼½½e2™¬Ú–/_®N:ÉÍÍMíÛ·WÇŽåçç'­X±B+W®Ô¥KExE}­Nœ8A)C HHH···úôécÕ£Aƒiãƺ÷Þ{‹|¼iÓ¦ÉÇÇÇæx€Òg6›5lÑ0KÀèÞ »Í6êuRò³뿆g[<«g[<{[ÇXãØK1׬YS{öìQJJJ¾C«nT­Z5½òÊ+?~¼¶lÙ¢ŸþYÿûßÿôÏþSwß}w±>¯”¶_|QYYYš3gŽzôèaµî—_~)öñŠûZ¡l`H<¨úõëÛ$úˆˆì.úC‡ùXûöíÓš5kôè£ÊÛÛ»DëÜÜÓ}¸ñC]1Vï­Ïÿ›Rnt†²"gøÑš5kŠµŸ‹‹‹î½÷^½ð úàƒ$Iü±e½»»»$ÙôJØSRR’BBBlÂÇéÓ§µlÙ²|ÿ;tuu-°æ[}­àX$''+00Ц=((Ȳ¾¨¦OŸ.I£F’ŸŸŸU[ß¾}Õ·oßÒ(œ%Õ«WקŸ~ª¨uëÖ R×®]U¡B=zT«W¯Vpp°vïÞ­Ï?ÿ\S¦LQ›6mT¯^=yyyiÓ¦MÚ°aƒ|||4lØ0Ëq›6mª€€}þùç:wîœ"""d04bÄËÕ´Jšù†»®2DŸþ¹þö·¿©C‡òôôÔ×_­   õîÝ[óæͳ9FÇŽõ駟ªuëÖjÙ²¥*V¬¨6mÚ¨U«VÅz­nÅœ9s4gΫ¶ .ÜÒ±‹R€5j(%%ŦýôéÓ–õE±hÑ"¥‡(;‚U€fÍš)::ZÆ Óùóç¦Ù³gkíÚµV㬙3gêàÁƒªU«–Õ1”™™É½?ÀÁrˆ›ÑMÕ*V»éöi™iJ:“¤FUÉhà»:(IB|ùå—7nœ¦M›¦cÇŽ©I“&úþûïÕ±cGË6YYYÊÊʲ™d%IñññªS§Ž¢¢¢ìY6à÷V¿W.2 r1ººíÈÅ#5å·)ÊÈÊÐþ§÷«~@};U å¤...Š‹‹S\\\ÛÄÇÇ+>>>ßuIII¥U ¦uŸVäm+yTRFV†$iÇÉ(aô+ÇÝÁw[–wœÚáÀJÀ9@È#¢J„eyûÉí¬œC°È£~@}yºzÊ”a¢ÄÁöìÙãèPŽð~³y¸]Ô rm;¹MÎPzfºÜ\Ü]V¹âãã#Izì±Ç\ Ê£œ÷J€„UÓ¶“Û”iÎÔá ‡Z9ÔÑ%•+¡¡¡Ú·oŸ._¾ìèRPÎøøø(4”ÿÞK€ä ûÏí'€8çEàßý¸ÚÔn£°Êaªå[ëæ;ŠŒ«`œÚ°EÃäÿ†¿î˜t‡’ÎíþLwÞ©Žõ:*Ä/ä¦7.= §v>õ¼.˜.è‚é‚ÜŒL&G£àÔ.§åNdöñàê6àh€S»’vŲìãNG#€œÚåëÙ= FƒQž®ž®@8µœ!X>î>2 ®@8µœ!XÌÿ€²«`œZάâÎÿ¸hº¨©›§jßÙ}jP¹žûÛs¥Q”;€SJK“~üQz½å‡2x^TE÷ŠÅ>ÆèŸGK’Z×n]`Ù¸QªPA ¿­r Ü €œÒ3ÏHS¦H5jô×Ö­R``ñö¯äYIU*TÑ©«§´ÿìþ|·ùäéÿþOru•’’¤zõJ pprÌ8¥-[²ÿMN–ž|R2›‹ŒÐ€PIRÊ•«ËùJÙcÔ¨ìåŒ iÍšÛ©ÊÀ)¥¦æ.ó_üc„U³,çíIK“ú÷·>Ç™3·R%”?€S2™¬!8P¼cäô€HÒþs¹dÂiófëm P4€Sº1€\½*=öXöp©¢Ê¯dõjéõ×m·%€@Ñ@N)'€T©’;9ü×_¥×^+ú1B+çö€ì;·O/J?ž;ŸdØ°Üm P4€SÊ™Ÿáï/}ö™äâ’ýøå—¥õë‹vŒúõQ%B‡?¬5[hÄéÈ‘ìu-[Jÿýoî¶(À)™L’ü(­öº^u¥ž›23³‡b]¹Røþ’äíæ­íOm×üÞóppˆfÎÌn÷ñ‘fÍ’¼½¥ŠÞ^„EC8³Yº~]Ò uèo]Õ6¡­÷\¦æͳ×ÿþ»ôì³E?^r²4dHîã>BB²—sî/B€¢!€œŽeº{n7‡Ÿ—fÍʾk¹$M*}ûíÍ••%  ?Ÿý¸wïì”9äܹìÞ@á §“@.[Ú|Ü}*Mžœ»ÝàÁÒ‰…ë½÷¤¥K³—kÔ>üP2r×ç¬,éÂ…Û¯œàt,Ä#7€TtÏž¬1x°Ô£GvÛ™3Ù ºKúŽÒóÏç>NH¬·É 9ÇŽp:–;”ç‚åãá#)»÷bêT)88»}Ñ"é£la2eßíüúõìÇÏ>+EEÙnG€â!€œNAC°rIññ¹Û?÷œ””d}Œ—^Êˆ)..ÿs@ x §Sج;çÞH055»·#--ûñ²eÒÛog/»»K³gKé†ËÚ’²EæÆk@ x §sã,£Á(o7o›íþû_éÎ;³—7o–&LȾÚULLî6'Jq{•ïD_5ù¤‰N]=eu àt,= ŸþªçÒ.éØ3ÇdÈ{éª?y{g÷n¸ºf?~ýu©K—ìû~HÙs>FŽ”‚*YöÙwvŸÕ1 P<€Ó±³Q~^>ªæS­Àm›4‘^yåÏÍÍÒºuÙËþþÒŒ’Ñ(…„Y¶ßn¿Õþ(ÀéXˆ$OÏ›o?z´Ôª•uÛÇK5kf/‡Vµ´@àö@NÇ2DE ..Ò̙ٽ’4p ôÈ#¹ëÃ*çö€Ü8+ï}A ps®Ž.€’VÜI ‘¶n;ôîƒZ¯»£Òr3º)=+]ûÏZ÷€¸¹I~~ÙwA'€ÀÍÑp:yˆ—WÑ÷»ãŽìIè..Öí®FWÕ ¨'IúýÜïÊ2gY­Ï†E€›#€œNq‡`Eh@ö<ÔŒT%_J¶Z—@.\ÒÓKæ|ଠ§s+C°nf\›qZ;h­N>­š¾5­Öåˆ~î\Éœœs@NÇd’´Kjó²>øÃG™{»«{ƒî·uÌ{«ß[ຯ„|[§§FÀ餦Jò;"5š§ONÓ–”-¥z>.Å EG8“I’ûeËãŠîKõ|(:Àé˜L’ÿüs«k×®Z°`ÁM÷ýàƒÔ¡CU®\Yï¾û®¾ûî; :TéÜ¡ J]vHÉÁ:xþ Æ'ŽWÿ¯ûkÞ®y–v÷)ÀúõëµpáBMŸ>] $uîÜY)));v¬zöìY྇ÖsÏ=§1cÆhüøñ–övíÚ•rÕéÏ9 Wî–’z¨ýƒ—èxÓ}Š"årŠ&¬œ Iò÷ôWï»zK’üü$£QÊÊ"€ÀÍÐR€„„y{{«OŸ>Ví111Ú½{·6nÜXà¾'N”ÑhÔ¿ÿýïÒ.“IÒÆ!òXð–Å,Sm¿Ú%rÜÐÊ¡–åýçö[–]\¤€€ìeŽR€ƒª~ýúòòò²jˆˆ$:t¨À}7mÚ¤Úµk+11QÍ›7W… ®×^{M™™™¥Z7àÏ"é†ÿ…߶ ï Uò¨$IÚwvŸÕºœaX(¤ÉÉÉ ´í² ²¬/È”’’¢Ç{LÍš5Óœ9sÔ°aC7NÇ/µšÙRS³ÿõô,Ùã K/È‘ Gt=ãºe]ΟŒ+WrÀ¤dddèÒ¥K3fŒÞÿ}uïÞ]óçÏW·nÝôÉ'Ÿè?þpt‰àÔr@II «&I2ˬçXÚó~guölÉŸœ“Ð P£F ¥¤¤Ø´Ÿ>}Ú²¾°}÷íÛ§èèh«ö‡~X .ÔÎ;U«V­÷5j”üüü¬Úúöí«¾}ûç)@¹UZC°$)4 wȾ³ûÔ0¨¡$Û+aògÀ_Äœ9s4gΫ¶ .8¨çA)@ýúõµfÍ™L&yæù mÇŽ’¤ºuë¸ohh¨öîÝ[àzƒÁPè¹'Ož¬&Mš³b@ŽÒ‚%åö€HÒþ³¹ѹ/à|òûxóæÍjÚ´©ƒ*r Á*@LLŒ®]»¦¹sçZÚÌf³Ô¨Q£Bßxƒ ’$›û…ÌŸ?_nnnºçž{J§h€23¥œ[.•FiÔP÷×¼_ßý¸ÂƒÂ-í(z@ ЬY3EGGkØ°a:þ¼ÂÂÂ4{öl­]»VóæåÞ|jðàÁš9s¦î>v9/Š†p*Ùä²å±ÊÀ©¤¦Ê2K²_ˆ¯¯äúçÀfŒp*7ö€Tt¯h—ó ¹½ (àTì5Äl6ëÄ•Úzb«¥-o1Ûgî;üåp,€SIM•äj²<öró*•óÜ5å.í9³G•½*ë̲»зŸÜ®·~yK’Ô©^'ÜàTL&Iïý®j52uðpz©Ÿ/´r¨eyßÙ}jÁ,(s@N%55û_/yºz–úùÂ*‡Y–÷ŸÛϸ À©˜LÙÿz–~ö$U«XMÜ*HÊî!€@á §bïb0,ð?¤J¹Ã¾ `‹pÙ?’äåe¿óæ ÃÊ4gê²Ë!K;l@N#§÷C²_ˆ$…äND?ze¿*VÌ^&€€-®‚pŽ M«5UT(…VU5Ÿj ”®\!€@~ §‘7€ØsVÏðžêÞÓò80P:|X:{VÊÊ’ŒŒ7 þ—p©©’¼OKFë·Ú}ôþ†÷RGΕ°²²¤ R”Y€Ó0™$y^”î\¨#>óôë±_R— @NÃd’ärÝòØÃÕÃ!u@ `€ÓHM•äš'€¸@ ¬!€œ= Pö@NÃd= PÆq^€ÓHM•C{@2²2täÂ]öJ•ÔHnD8 Gö€\M»*ÿ7ü•ž•®&-%­–D€1 à4L&I—«I;û¨‰WOÝx§ÝÎ]Á½‚*{W–$ýqu¿¥Öè8 “IÒñfÒW_hô?¤‡Ãí{þЀP¸rB§SOJ—¤ë¾¸= §‘šš»ìéiÿó‡U³,W¬Ý Bk€Ó0™r—@*ÔÚ'‰7"€œFÞâåeÿó‡„Z–ݪf÷€œ?/edØ¿(« §Q–†`™öY–ϳ-PV@NÃÑC°êÔ“AIRZ…–v†a@.Ài8z–§«§~~ügíz¿ËXei'€@..à p©©’Œé’Ù(OO‡ÔU7J’˜ÛF€\ô€"##CcÆŒQDD„üýý¥%K–Üt¿3fÈh4æûsêÔ);Tå“É$éÁg¤XWÕáªMÇ79¬–@ä‹B<òÈ#ú駟§ hÖ¬YêÚµ«æÍ›§ž={Þtÿ3fèÎ;­ïÂPZå@¹g2IòL“$eš3åæâæ°Z ?HÖ¯_¯… júôé0`€$©sçÎJIIÑرc‹@5j¤&Mš”r¥€©©’*\·<öpñpX-ÈC° oooõéÓǪ=&&F»wïÖÆoz ³Ù\Zåòa2IrÍ@\ PÖ@ pðàAÕ¯__^7\F%""B’tèС›£k×®rssSHHˆúõë§Ã‡—F©€?™L’\耲Œ!XHNNVpp°M{PPe}AªU«¦—^zI÷ß¿ŒF£¾þúk-X°@?üðƒ¶mÛ¦Ò*ʵ{@Ü]ÜV‹ŸŸd4JYYÈ‹R :uê¤N:Y?øàƒŠ‰‰Q«V­ô /hÎœ9…î?jÔ(ùùùYµõíÛW}ûö-•zÀY¤¦ÊºÄAC°–Z®ù{æËeÐne}ûžÎœ‰pHnÏœ9sl>·]¸pÁAÕ8HjÔ¨¡””›öÓ§O[ÖÇ< »îº«HsG&OžÌäu¸ÙC°Ò,5ksÊfMùmŠTSRÕ­:sˆüå÷ðæ͛մiSUä˜R€úõëëÀ2å½­®¤;vH’êÖ­{KÇ5 ·] &“¤e¯©Öº/õYÏÏäjtÌ÷lwÝ•û h·._–®_/x{(O ˆ‰‰Ñµk×4wî\K›ÙlVBB‚5jTìä»zõjíÚµKÍš5+éR’Ìæ?‡`m¥à³½ÔÿîþûÒ§aPÃÜA»%IgÏ:¤(s‚U€fÍš)::ZÆ Óùóç¦Ù³gkíÚµš7ože»ÁƒkæÌ™:xð jÕª%Iúûßÿ®öíÛë®»î’Á`Ђ ôÍ7ßÈßß_'NtÔS§–‘‘=á[’<=[K­JµTÁ­‚®¦_•ªì’”=½zuÇÖe¤_~ù¥Æ§iÓ¦éرcjÒ¤‰¾ÿþ{uìØѲMVV–²²²¬îù¡Ù³gë?þPzzºBBBÔ§O;VÕùë¥"ïˆÙ® nwFƒQáAáÚx|£äPrMÕ™3. ÊH!\\\§¸¸¸·‰W||¼UÛ;ï¼SÚ¥nšš»ìè){ÖÆã%ƒY Ü«3g"]” Ì8…¼= e!€XODߎ@àOô€œBY ͪ7Ó}•7~È IDATºjà ¥3áøàòÁrôIjW§ÞkÑN÷?“ý˜Ù‚p &“$C–Ôx¦ŽúÎÕª#«]’s— €SȾ úu©gŒK2%¶ÕŠ˜­‰¶è8“I’kîíÆÝ]ÜWÌŸ|}%×?¿ê#€@6À)¤¦*»äO.Ž+æOCn/²@N!{Všå±‡«ãˆD€@NáÆ!Xe¡DÊ ©©ÒµkŽ­ÊÀ)Ø Á*#= þA&)x›Ôà[zA@\ à$ÊjÈê°¿Iwm‘²\”rêšî¸Ãñ“ãÀ‘è8“I’1CÊÈþ€_®‚%IU]³Œ™Úvl¿c‹€2€€SHM•”|Ÿôêu-^lV‡¿g9º$IÒÞ µÓ”½¼óä.Iw9´p4z@NÁdÊ]öò2ÈÅèâ¸bòókhYÞa·+€²p yˆ§§ãê¸Ñ]Á¹äÈ5@N!55wÙËËquÜ(¢F=˼”Y»\ 8àÊjHÕ*®ÒÙ’¤ .û”ž™îàŠÀ± §PVH` ¤ÓÙð\2**ùr²c #€œ‚õ$tÇÕq#ooÉcõDé­ãºóÛs ñ qtIàP€SÈ;¤,õ€ R[ˆt¥šÎž18ºp8À)˜L’~)=ÚC~xDottIÙÿž9#™ÍŽ­p &“¤ =Òßj~ÒW:uõ”£K²È ééÒåËŽ­p ©©’\®[{¸x8®˜ä)»Ê3À)˜L’\óW”E€S0™Dü@N!5Uô€À_àÊ|ˆç©Û“zéH³ÿÏÞ}‡GY¦mþM&“FK(Fé`@BGETTQA`¬‹«è®û©+î®Êêꮂb dÅ‚]¤é Ò$ô ¤ÍÌ÷Ç›)LH™™wÊu‡ÏLf&·a®¹ïç}ýñh³K1M¬ÙˆˆˆÔ”Ó{@ŠjAç·Ùá(dñN] KD¢—ˆˆˆ„½¢¢Òó5¶_Ài ­ô¹¨ºñuÍ.Ë­aCÀaƒ½¡ñr6Øı¢cÔŠ«evi""A§, {¥‹¥wÓ%û]>þõë›Z“7÷œ.8q²vïZó 1‘ˆˆˆ„=w̫× J9Ý÷­ÊYeN1""&S‘°ç@ͫׄ¨U w`Õ‰N ""öŽ÷¬C±¥cX{:¹o¯ÎYm^1""&R‘°ê#XP@ ëÁÁÖ¬Ù³»ÃnnQ""&ÐU°DD$ì…úxmD_~;ýÛqzµê‚ÃéÀŠÕÔºDD‚MDDDÂ^ØŒ`,ü?2΄¶mM-GDÄ4Á‘°çî€Øò±%ãt:M­§<îìßo^""fS‘°ç ÷5ãqâ8kòY¦ÖSƒˆˆˆ„=÷Vl!6«Í¼b|P1(€ˆˆHØsw@¬F‰·Æ›WŒ "" {€ÅÖâc@DDB•ˆˆˆ„½ãÇqw?@‘P¦ËðŠˆHØ+(À½ÿB¿²o¿%;—±zÏjl16Ætc^a""A¦"""a¯ €ï€4hàYï;`§Ï´>;ŠéШƒˆˆD`‰ˆHØ ‡ˆÍõêëƒûâhߨ=öoàxññ ž)"Y@DD$ì?ä7„w?åéîqÿy÷›]R¹\cXû÷C—Ó»`wÚùu߯&V%"\ ""ö €â$Ø<ˆ-‡Ò«Y/³K*—+€:Nëâ¾UÎ*“* > ”””ððÓžžNJJ ýû÷gþüùU~Gy„˜˜ÒÓÓP¥ˆˆ¸ÏÍ«ãT¼7¢·Lìì^¯Þ³Ú„jDDÌ¡RaÆ1iÒ$222˜1c©©© <˜9sæTú5V­ZÅÿû_RSS±X,¬VD$z÷ÚB‘`^§â@ΰxˆ: "Mt,–,YÂܹs™:u*£G`àÀìÞ½›ñãÇ3tèÐS¾FII cÆŒáÎ;ïdÕªU8p ÀU‹ˆD'ïH¸â¼ú4¯×œí‡·³:g5§ƒ‹>‘ȧ?é|˜>}:IIIŒ1¢Ìý£Fbݺu,[¶ì”¯ñÔSO‘››ËO<Óé T©""Q/ˆk#zã:éÓ¢y…yæ&"Dê€øMZZ‰' »öqlݺ•=zø|þºuë˜0asæÌ¡V­Z­UD$Ú…ãÖþý0û–Ù!yÉ`‘@Rć]»vÑÐûoŠR5rÝ»ÝÎ-·Üµ×^Ëå—_°EDÄàê€Øl`µš[KEN "Ô €çž{Ž¬¬,æÍ›W­ç7Žäää2÷9’‘#Gú£<‘ˆSPÔߌµíBffÆsN“s8³þ™f—u’ˆˆ„¶™3g2sæÌ2÷åææšTMäPñ¡I“&ìÞ½û¤û÷íÛçþzy¶oßΣ>ÊÓO?Mll¬û‡´¤¤»ÝÎálj'¡‚‰'Ò­[7?ü[ˆˆD‡ãÇ?Rpyׯ ~MDDj¬¼€W¬XA÷îÝMª(2hˇ´´4²²²(ðÞÙdffкuërŸ—MAA÷Üsõë×wÿZ´hëׯ'%%…‡~8àõ‹ˆD“‚ÀZè¾g3¯˜ (€ˆˆ¨âÓ¨Q£x饗˜5k£FÀét2}út:vìè3ùvíÚ•ï¿ÿ¾Ì}N§“qãÆ‘——Ç´iÓ|vODD¤z €XO Õ½))`±€Ó©""ÑKćž={2dÈÆŽË¡C‡hÛ¶-ï¾û. .döìÙîÇeddðÖ[o‘M³fͨW¯}ûö=éõêÕ«GIII¹_‘š9±o ÍbµBýúpà€ˆˆD/`Uàý÷ßçÞ{ïeÊ”)Üpà äää0oÞ<† â~ŒÃáÀápœòœ‹Å¢“ÐEDÀé<¹ª#XàÃÚ³ìvÏý…%…ì8¼Ãœ¢DD‚H¤V«• &™™É¡C‡øæ›o¸ì²ËÊÝCŒ#«ÅŠ-&tÈ€pfé‚¿þÖ¯‡s›œ Ý»W˜XˆHà)€ˆˆHXsw@¾ùÃ7§äÑz7ïmjM‰‰±c=·_z ÎirŽûö’KL¨JD$x@DD$¬y×TÁ¯!eôhHJ2Öo¾ íëyÈÒß—šR“ˆH°(€ˆˆHXs`‰‰æÕQ))pãÆúèQøå³³©W€¥»@D$²)€ˆˆHX Çœ0†õ¢•gô`ûáíäÍ1©*‘ÀS‘°®¤S'pM»a¤–xÆ°Vå¬2©*‘ÀÓIè""ÖÂqËåî»aÁc½ï››ùß?.àœ&çZ;ÕÜÂDDH káÚ276ÖßÏî@§„+>D$â)€ˆˆHX çb³Áwk‡^yÅÜzDD‚ADDDš;€\ò oäf謡–šZSUÜv›D^½l ‰D ""ÖÜ{@š-bMÁ§|¼ác,‹©5UÅé§Ã°aÆúÀ˜5ËÜzDDMDDDš»c`-rßg‹±™SL5Ý}·gý àtšW‹ˆH )€ˆˆHXs˜bbcbêpÞyЭ›±^¾–ê,B‰` ""ÖÜ#XV#€„[÷Àb)Û™<Ù¼ZDDMDDDÂÚ‰›5üÀþõë-~`Æ‘ÛhÿB'–ì\bvY""~§"""aíÄ= qÖ8óŠ©ÄDÈÈ­ÃÑõ ÖÌdÉ.‰< ""Ö"aËå®»€ßÏqß^¼CDD"O¬ÙˆˆˆÔ„»²î:†\¿& ë™ZOM´jWôHçÓ’xˆ-äûÍÚ."‘GDDDš;€Ì–§BãƦ–Sc÷ŒãÓÙÝ ÙÏì.ÚÂÁã©ŸXßì²DDüF#X""ÖÜ#Xû(ÂÝ%—@ò1ÏÖû‹Ô‘È¢"""aÍÝÌ«Ã_bbàªîçºoOùBDD"‹ˆˆˆ„5ïo^þtß0OdÅž¥>lb1""~¦"""aÍ@âãîA$èܼ5i#àÛÇ°/x€éÓÍ®HDÄ"äj‰V®= ‘0~åb±XøøÆ÷`ÁxØÖ_‡ÃìªDDüCDDDš«I C¸è"c½i|ýµ¹õˆˆø‹ˆˆˆ„µ‚Àâ ¾n%Øv³Kò›»ïö¬'O6¯R‘°vü8´í7Ô#qB"×̾Æì’ü檫 iSc=olÝjn=""þ """a­ °»oÛblæãg±±p×]ÆÚé„7Þ0·P‘°åp@Qã@¬‘@Fö¬þÙ´2DDüFDDD–û k‘û¾8kœ9ÅHãÆšj¬33NˆˆH8S‘°å ‘9‚åt:Y³šäË^€ ždÿ~سÇìªDDj&ÖìDDDªË@"xkÀ;ØsæhRþ•ÌL+§ŸnvU""Õ§ˆˆˆ„-×!„Þ#X‘Ô±X,ômÑ׸‘©kÈÌ4·&‘šR‘°UÞV¤íq€–?(€ˆHØÓ–ˆˆ„-wÙÛ+r?ãö;‹iÒÚÔšü­Li±€ÌÌqæ#"â ""¶Ü¤ …vÖ\u–©åDÇÓ:’œLnA.´XÀÚ¹Nìv V«Ù•‰ˆTF°DD$l¹÷€ æÕH1–ú4ïcÜH:@aõde™[“ˆHM(€ˆˆHØrw@ˆÜåa™W‹ˆHM)€ˆˆHØò ‰‰æÕh}[ô%96~‡Z)€ˆHXÓ [Ñ0‚гqO_·›vX@D$¬©"""a+ZF°, iiw—GDD™ˆˆˆ„­h V+´oo¬·lü|së©. [î¬F¿²èøTÞYó¿åþffI•žnüÓé„uëÌ­ED¤º@DD$l¹; g~ÉK»2¸iÎM,Ù¹ÄÔšÉ@@cX"¾@DD$l¹ˆµØ}ŸÍj3§˜ P‘H """aË=‚ã@b@DDB™HJJJxøá‡IOO'%%…þýû3þüS>ï믿æÒK/¥I“&$&&Ò½{wzè!-Z„ªED¢G´u@RS¡AêqhñËfv9""Õ¢RaÆ1iÒ$222˜1c©©© <˜9sæTø¼ƒ’žžÎĉùâ‹/5j , OŸ>|üñÇAª^D$òyH‘û¾8kœ9ÅA¡½€C·7„1’Ûó¯ìÛgvE""U§ƒ}X²d sçÎeêÔ©Œ=€²{÷nÆÏСC}>wøðá >Ü}»_¿~Œ9’–-[òꫯ2dÈ@—/"Ü$JF°bhäìÀ~Ó~å§û: ¡Ùe‰ˆT‰: >LŸ>¤¤$FŒQæþQ£F±nÝ:–-[V¥×«W¯uëÖ¥~ýúþ,SD$ª¹÷€DÉ@çä~îõ§™?™X‰ˆHõ(€øMZZ‰®cgK¥—îܺuë)_ÃápP\\Ì/¿ü¸qã8zô(·ß~{@ê‰Fîˆ#–¤Ø$l1¶ˆÁ¸´m_÷zIÎ+©`ù°k×.RSSOº¿Q£FʠAƒøòË/£²xñb:tèàßBED¢˜;€|ùÌýOÄŸ†0¢×<°Ü'Ùv ? 4yòd:ÄâÅ‹™;w.^x!S§NåÊ+¯¬ðyãÆ#99¹Ì}#GŽdäÈ‘,WD$ì¸G°€øxóê¦f Sˆ;”NQý5ä×]Iîñ<’ëš]–HDš9s&3gÎ,s_nn®IÕDš4iÂîÝ»Oº_é%Gš4irÊ×HKK gÏžŒ;–´´4þùÏž2€Lœ8‘nݺU£j‘èâê€$$€Åbn-ÁÔÄÞ—­¬³–|Ï^evI"©¼€W¬XA÷îÝMª(2hˆiiideeQàîï2KO~jݺu•^/&&†Áƒ³bÅ Ž9â·:ED¢™w‰&=ë_j,öMöoÅ?XD$Ä(€ø0jÔ(òóó™5k–û>§ÓÉôéÓéرc•“oAA³fÍ¢qãÆÔ©SÇß劈D%×Ö × ‰xW¶¿žÛ/®£öŽkÍ.GD¤J4‚åCÏž=2dcÇŽåСC´mÛ–wß}—… 2{öl÷ã222xë­·ÈÎΦY³f\}õÕtéÒ…Î;S»vm~þùg>ûì3öïßÏ»ï¾kÖ¿’ˆHĉÖH÷Ήp¸9¥y‘°¡R÷ߟG}”)S¦°sçNºuëƼyó¸ì²ËÜq88œN§û¾ .¸€>ø€É“'sìØ17nLçÎùþûïéÓ§ÿ*"")ZH›6ƦûÂB ?§÷;g1•kSÓòå˵ ]D¤l6()nÝ`ùr³« ®®]aÕ*°ZáèÑè a"fÑûµšÓ K%%Æ/€ÝÇ1èÝA\ýÞÕ8œs ’Òsq±Ûaýzsk© `‰ˆHX*,ô¬¦,âó-¿`ÁBŒ%:>[s0Æ°ºv5¯‘ªˆŽ?¥ED$âx_%Ý[@œ5Τj‚ïÄ"".@DD$,yŸ‚îŒ1ΰYm&U|Þdùú}|¶ù3óŠ© KeΉu˜è CJ 0ô&¾ë‘ÊàƒÙwlŸÙe‰ˆœ’ˆˆˆ„%ïâŒ1F°¢©b±”vAŽ4‹'N¾ÊþÊì²DDNIDDDÂR™,‹Ñ‰¦= P@²¸oÏÏšo^1""•¤"""aÉ»â°Dß”í½¡( €/³¾DÇ{‰H¨ÓexED$,yt®çìι4Ljh^A&HOìñðÛEÐöSrŽæ¹7“N©Ì.MDÄ'‘0Q\ _m¼élÚÔìjÌç=‚uuüDb^-féرt±e´ý€ù[æ+€ˆHHÓ–ˆH˜xòI4Î;ï„+@E)ïÿ æÕa¦ºu¡E ´DDŠˆˆH˜˜_ú¾r×.X³ÆÜZBˆ!=8еÄcÃfµiˆˆ„4‘0±q£g­“¯ËŽ`%&šW‡ÙŒ -0ó¼~€Ïoø‹ÅbvY"">)€ˆˆ„Œ_. ꀸ¸ODß›NÖº:¦Ö""R ""a`Ó¦²·@@\Üý\ˆHxP  'ó Ñ<‚uÖY`+=þD?"@DD€÷þ€}û`Ïsj î= %±¹+:F±½ØÔšÌ`³A»vÆzãF(*2·‘SQ 'v@@Ÿv»; uwpåÂj?Y›çÜhjMfqa•”À† æÖ""r* ""aàÄ(€¸ˆÕÓõ°ÅØÌ)ÆdÚ""áDDD$ÄÙí°y³±ŽõÜío4Ý#X1^Ī’™ N§“Õ9«Y³Ú¼¢DD|P q;v@a¡±¾è"pñíÄÓñlzˆ³Æ™SŒÉ¼È’¿qæógÒåÕ.<¶à1óŠñADD$Äy_uí gži¬×®/æýµ²fOt‹®,fÍ ^=c½eE3Žà‹-_p¼øxÏ >‘ç½½mÛÒO»-v ®º†á^Çw[¿3­63iËÃbŽõÎíV.ouùÅù|™õ¥‰•‰ˆœLDD$Äyw@Î:«4€\úW8k=÷0§0“i«,ï1¬tÛP÷zΆ9&T#"⛈ˆHˆ;±²§éëpþ³îû í…&Te>`•å@~ïOí¸Ú|²éJ%&U%"r2‘çê€$'ïǾgÊî?º¿Öuç«\Üêb“*3—k+1·+_ÜðŸŒü„;Eç9 P6€lX›À 6ƒ8xü ¶-0©*‘“)€ˆˆ„°ãÇaûvcvV1·}r+%ÎÒO³¾c n7¯8“¹: ‰4`@Ú·L»†íÌ-ÊD®= `\!mh;a‰HhR a®ó?¬Ý§‘u( €Úûû—ϰe‹×fì(ã æÖ*RR iSc™ Óg£aRCêÆ×5·8/±§~ˆˆˆ˜Å{úÔÀi¬û=ÍgN+'¬[Ý»—}Þš=kh^¯9É ÉÁ+6ÈÜDsë%éé°s'äæ‘ýuYvÛ2Ú7j5Æjvi""nꀈˆ„0ï èO´ûœ7¯~“?öü#·=×}¿÷„ ¶- Ë+]èüJgfdÎb¥Áçêü¨âqâ‰èé©é "r@DDB˜wäìvVFuÅ‹ƒ^<é¦KJB «÷¬à­Õo©Jshëd¾~.DDB‰ˆˆHó¥yÖ¾Þh¦§¦Óåô.,Ùµ„ͼ6‘D’°Ûµˆ‡ˆˆ„‘åtz: Í›CR’çk§Ÿ ëßhk?̽^ºki€«4‡÷Æ{íñh׬¥W "ª@DDNPR_|¿ýfnû÷›‰Á8€Ð›Åâù´;'Çx¬K‡FÜëM7‰Ü‡…)«x}ùëL_5퇷›WTˆ‡³Î2Öë×CqqÅ1ƒˆˆÈ ž}„ž=!/ϼ:¼÷¸ÞTzó5nÓ¶'­Dê–w9ÔàsnŸw;£çŽfåî•æ"\?EEe/ã,"*@DDNðñÇÆ?÷ï‡4¯ïrb|Ö)­‰±¼o>™ï@½G°¬qžùmV› Õ„_?…%…|¾ùs¾Ýúmð‹ñ¢""⥸V¬ðÜþé'ój™žý/h÷1XìUê€ÄÇÆÓ¼^sÀè€8ÎW|Þ˜X¯£RÞÏÅ®¼]4z¦ƒf büwãÍ)LD¤”"ñ’™ ……žÛfy;ùÉö(üÁ;Ï¡mÛÅ€¥Ìc:x¶zœ´áxúéÔ¯KZý4,–²Ï‹ÞÄb+r¯Õ)?€4®Ó˜É-X»w-‹v,"ë`gÖ?ÓœE$ê©""âeéÒ“o{¿Ù –W–½‚Ób\gÖúÛš7?9DÔ©­ZëµkÁáð|­o‹¾t9½ µãj£Ü óÁòî€ÄYãL¨&´´hµKÛ]Äb±pS§›ÜygÍ;&T&"bPñrb)*‚å˃[CAI¯-͸a%-÷N÷¥UOäú´ûØ1ó¯ÚLe: Á*#&:v4Ö[·Â‘#Æúúôë±”vÑÞÉ|'"GóD$<(€ˆˆx91€@ð7¢Ïþu6ûò÷7Ö]G‡æ}>6Zž+@¬Ú„~"K.¹s¡qí¦\Ôê"¶ÜÂ’]KLªND¢ˆˆH©#G`Ý:cÝ°¡çþ`îq:<¿äyÏKÿTîtˆ£N\b4‚UªÏzéR2:u‚´cž1¬·×¼mBe"" ""nË—§\s 4jd¬.,»¿"ï\ÌòÝ¥3_¿wƒ½Ê½¯K´ï= ק^AD$p@DDJy_{.\p±ÎÍõtFí¥e/yn,¹°TØiÓâJ?ô¦âÝIH0¯ŽPe±À°a°z5|ò œ~éŠêà\?öMá÷æéÿ›zئˆD'‘RÞäœs<‚7†õÜ€çx²ÿ“Ô9Ò~” º‹ÍgŸm¬7m*{ áH¦R9 lüüþð üïuxñWŽ~þã¬E‹0~¼qð¦ˆH0(€ˆˆ”rZµŒ7õf†I yð‚©ûÞ2(I A¨_¿âç¸ÆlìvX¿ÞsÿÏ;~æ¯`È{CXöû²Àmï¬ÄDóê ôí _|Ë~®Åu×Yp“› O|8­Zµ¢N: 0€—_~™Ý»wè»Å± IDAT¡j©Ž_~ñ¬Ï9Çø§Íçg¬·mó”@Û¼Ù³®hº‹¯Ò¦AÏkôzÑ RsgS§Bv6Ü{¯'l/_99æÖ&"‘M¤Æ cÒ¤Iddd0cÆ RSS“&Mâ¹çž tù"RE• ?ýøCàkٴɳ®L¤IHN6fù½Hóz͉³ÆQd/bÓM¾_ ©â'þ¬_{­yµˆHdSćéÓ§“””Ĉ#ÊÜ?jÔ(Ö­[Dze¾7tz‡—óÎ;F±uëV¿×*"5ãtzF°RSO„]Î;bJÿ¤ Ö>ªv@,Od×.8tÈX[c¬´Ni @Ö¡,Î fÞä©Ì»ðΆÎòýÁœšûR½wÏ“ˆD²³³IKK#ñ„Þ~zéßòU 7ndß¾}têÔÉo5ŠˆlÙbtÀè~¸®P·.tîl¬×¬Ã‡ýû½÷ÛÇš=kÊÜç  ¤¥UîuN5†UPRÀμ5)5¤x`­:°ˆ/³¾dþ–SïÑ“²r rycÅô{³ÙÇ—Ó±£qÿÊ•ÆÅDDAć]»vÑ°aÓîoTz4ò.×%j*¡¸¸˜›nº‰Fq¯÷­ˆ„_ãW.®Ñ§~þÙ¿ßûÅ_^¤ó+¹ðÍ ÉÜ“‰ÓéÁjÙâã+÷:¾HÛž®HÃò8Ó¼mV›IÕ„¯Ö}ÀmŸÜÆ‚m x{ÍÛîŸu»–,1·6‰\Ú`‡ƒŒŒ V¯^ÍG}Dƒ NùœqãÆ‘œœ\澑#G2räÈ@•)Õ*@^xÁXÿô\~¹¾oaI!¯,{ÅxÝí?‘’˜BN9b|½2ãW.¾Hïf½Ùx`# “R'®Žª ÞÄN1qÖ8“ª _×µ¿Ž»?»›B{!o¯y›ÿœÿ$¯¼btþú ú÷7¹@“Íœ9“™3g–¹/×Õ2—jSñ¡I“&å^6wß¾}ŠÃáàÖ[oeÆŒ¼ýöÛ\qÅ•úÞ'N¤[·nU+XDªÍ;€ôèqò×{÷ö¬ý9ÿþº÷Ùslל} Më6凕ž¯Wfº‹ktÊ¡geèÙ‘·7¢LÄa[Œ: U•œÌ°ÃxgÍ;<~ƒg£íò?^±bÝ»w7©¢È ,ÒÒÒÈÊÊ¢Àûo9 ³ôoöÖ­[Wø|§ÓÉm·ÝÆôéÓ™2eŠº"!ª¨È˜whÓ¦üSÇ›4V­Œõ’%ÆsjÊét2iÉ$÷í{ν¨út—zõ ysc½v­1.É\{@¬V(rh«&þØãîõì­/Ñ´©±þùg()1©(‰h >Œ5Šüü|fÍšå¾Ïét2}út:vìXaòu:ÜqǼù曼þúëŒ5*%‹H5dfBa¡±.oüÊÅ5_P+VÔüû.Ùµ„e¿WÓërzz73Ú,U½¯7×V^lß^óC™ë³¡„(¶«Rç5=.§w`鮥´ï¿€cÇ`õj3+‘H¥âCÏž=2dcÇŽeâĉ|öÙgÜxã,\¸Ç{Ìý¸ŒŒ l6;¼ŽH¾çž{xã73f :t`ñâÅî_+W®,ïÛ‰ˆINµÿÃåÄ3jê…¥/¸×÷œs–ÒKoU·¾÷D¢2Ä5‚¥HµX,–2]#í^v¯5†%" R÷ߟ{ï½—)S¦pà 7““üyó2dˆû1‡‡ÃÓkÞaÞ¼yX,¦NJ¯^½8ÿüóÝ¿®ÕÉN"!¥²ÄŸû@~?ò;³ @䆌L÷Œhº: IIÆèWUDSq`%&¨ΣÓe CÛEÞ^—`¹>ýzêÆ×`3Ÿ‚ÕkS‘@Ð&ô X­V&L˜À„ |>fÚ´iL›6­Ì}:lP$|¸Hl,téâûqgŸ ))Æ!?ýdì±ð>/¤*NcºŒá5ïp[·ÛHˆ5Žò..†ìlã1mÚx@¬¬h Þç>on1 V\-þÞïïÔŽ«Íˆö×ÓüÉ8òòjþ³."Ru@D$jååÁúõƺsgãͬ/11ž.ÈeG¥ªªiݦ¼våkì¸oîõg÷ýÙÙžM¿U¿r='¶ôc¥h â÷÷ºŸÛ»ßN½ÄÚîSÑsr<¡XDÄ_@D$j-_î¹ZTEãW.þÞÒ © ’Ö}À¦F{£_‹~ônÞ»ÜǪ’™ ë4æ½kßã뛾手Ÿ¨Ù‹‡€²ijï#ÎgB5‘ëæÎ7“lmlÜ8ûc>úq¹‰HÄP‘¨SÝñ+—ʇÓÁ„'¸o?Ò÷Ÿuu@¬Vhݺê5y;1€$Ä&0¢ãú·îOZý´š½xðÁ*ÓÑ–_ÅÇÆsWgO䃜›XˆD‰:5 Í›CÓ¦ÆzñbãdîòÌÛ4̽ƥ¨Îmr.ý[õ/÷qN§§ÒªÄÕðƒü”ÏXYf¦çÔðHQ¦§¬@zðÒÛ±¤ð{ýdØfrE" @D$êx=ªþ|‹ÅÓ9v V¯>ù1N§“'xÆéû³^»váÞà[Óý.®.Hn®ñú‘Ä;€ÄØ4‚HuêÐ.ïOÆ k |ö_s ‘ˆ ""Q¥°V­2ÖgUýÿN5†u´è(§×>€Î©¹¢Í>_Ë{ÿ‡¿xΉÞ#XVu@îÆ´{àÈðÓÿ‘žû°ÙåˆHˆ5»‘`Z³ŠJ?4¯Îø•Ë‰dܸ²_¯_‡ÿü«sVs¬ø˜ÏVM7 »œ@ôÏë†ïHûºçòÕe_Qd/¢Mý6æÁômÀßÎÙk2»" w "UjºÿÃ¥cGãÄñ¼<#€8å_M«óéOùZþ¼¯K$w@¼ÈiµrIëKÌ+& tî µm;?þèûg]D¤²4‚%"QÅ_ÄjõœŠ¾gdeUÿµüy ^—³Ï6j„È Þ#X‰‰æÕ-bc¡W/c½klßnn="þ@D$ª¸ˆÍf|²[U9¤"®HíÚÆÁˆþï 3ë×æ}[™¶rÏ,|†å¿/÷Ï71‰w$!Á¼:¢‰¿~ÖED@DD¢Èáðaƒ±îÒÅx“^þxSVX¿ýf¬Û¶õïh‹k «¨>Y±”[þw ýú¯ü°íÿ}(€Ÿˆˆø“ˆˆDeË<뚌_¹ôìitR úoʲ²Àá0ÖþÚÿáâ½$wWC÷z_þ>ÿ~£ S ¾sÏõŒô)€ˆHM)€ˆHÔð×þ—¤$èÞÝXoÜÙ»òªü¸¯‹wÉÉöýùûýû‚L{@‚¯vmèÚÕX¯] ‡Áš=kø|óçæ&"aIDD¢†¿x¦4ý™ŽS›òØq´èh¥ŸˆKðºtìèYoÛ9Ds¸ÖcJ¸îÝQty¥ c掩ÒÏ»ˆ(€ˆHqºuý÷fßxSæ„Ëà¸ãÿþïÌþuv¥ŸˆKðº´jåélý52Ȇ¼e¼ºìU¦®œÊμæÜÄËνGqâdϱ=<÷ós¦Ö%"áGDD¢Â®]ðûïƺgOˆñÓŸ~矴› ÍЮa;nî|s¥Ÿï=‚ÕÆÏçèÅÄx^ó·-ñÔ‰«„ñÁún×<îüôN2þ—ÁÚ½kÍ+* ôîíY×^ò/¬cSÈÓ‹žfß±ðÞW$"Á¥""Q!ãW) Jˆô ûö?/xŠØ˜ÊŸñêê€4n uêø¯.WW¥¤êÙŒ.H¸¿Yôî€8,Åîµ-ÆfB5ÑãôÓ!-ÍX¯ýþ,Æt¾€£EGy|Áã&V&"áFDD¢B È”S(ª[š"¶]@£ƒWUú¹ÂþÒf„¿÷¸xuÕ¡1gÔ>ƒ¦u›ât:ó ƒÀ@bc¡ÄYä¾?ÎgREÑÃ5†UTƒëü$[¯,{…¬ƒ58SD¢ŠˆˆD…_~ñ¬ý@Žå?üÃsÇWO³paåòä°\¼ƒM†å'~ÿóךּs8d®¬ÄD(¶{u@¬ê€š÷y ~9ƒû{Ý@±£˜G¾{ĤªD$Ü(€ˆHÄs8<¤IcÜÉžýùYrŽæ7Ö] ;{ñÔSð—¿xö›T$Äûu½¿_8su@Œ7¾.Á ¼$|àüh˜dŒö}°îvÞaRe"N@D$âmÚy¥Gtøsüªw³ÞtNíLlL,­·þ €cÇà¿ÿ5®@uÇÆAƒ¾ò¼å½®÷÷ gÞ¤ Ä³!$>¶†GÛË)µm K/¨¶p!Ô¶ÕåѾ2¤Ý2ïʤY½fæ("aADD"^ öôoÝŸw¬`á- ùin[ÆŽ…øÒ÷ÀEEðÚkƶ믇5kN~~0: ÉÉpÚiÆ:Rˆ÷ÖñbÏ%±jÙj™TQô°X<]C‡`ýz¸ûœ»™3bí¶3·8  "ñ@b,1œÓäÎ8&O†mÛàÁ=W´r8`æLè܆E‹<Ïu› Z¶ôo]Þ\]œO'(œyw@bHIH!ÞO¢MÇ¢ÉcX἟HDÌ¡""Ï@,èÞ=°ß+5ž|¶o‡  Q#Ï×>ýÔ8K¡_?øâ ؼٸÿÌ3+:J$íq:Ë©WOåàÿ¤à‘N¯}º¹ÅE‰ˆˆHU)€ˆHD+,„U«Œu»vP¯^p¾or2<ü0üö<ÿ<4ó_°ô¼‘Ôþ—H Ež«î’`^ѬkWcü @D¤z@D$¢­^ Å¥Jò÷øUe$%ÁŸþ[¶À´iF:Q ö¸DÒFtïSÐ5qeŠ¸88÷\cýÛo°s§©åˆHR‘ˆ¶|¹gÝ£‡yuÄÅÁèÑðë¯ðá‡eGÁ.¾8°ßÛpjíá傾t~¥3÷|~O`¿i€xŸ‚®ˆy¼Ç°.ôý¸Â’ÂÀ#"aGDD"ÚÊ•žu×®ÕU9«˜°`Eö¢S?¸11pÍ5ƹ$‹Á?Âå—×è%O©uk°Z‡}I?²fÏ6ÜØo ¡áTû@æOŸÿ‰ž¯÷¬ñÿ3"y@D$¢¹ˆÅb\‰ª:ì;wÌ»ƒG¾{„.¯taë¡­5®Ëb^½Ê¾‘ ”¸8ã\ ëºï;\p8ðß84‚zõ2Â4”@nšs“—N&so&Ïþülp‹‘§""«¸23u›6P»võ^çµå¯±t—q)-'N×ñÓQêAÔ¶-àˆ…Bã?BnA®¹U“: ¡¡n]èÔÉX¯Y‡Oȳÿ¸ðÄXŒ·ýð˜_B»ˆD‰X6WÁèÖ­z¯‘s4‡‡¾yÈ}û•+^ Ë·Ýû@ ’8\žÐáêÞ9°xqÙ¯u;£:çO/9ÎÝŸßÓé r…"ª@D$bùcÿÇ}óïs¿YÝe4ýZöóCeÁç …ÆuˆÃµâ=‚eK, ÷ÔÞ\úö¥üíÛ¿™WT”òœ3çä¯?~Ñã4©Ó€Ï6ÆGë? Re"ê@D$bÕ4€|™õ%ï­}€‰ xæÒgüTYð¹/Å[ÚÉ/ΧØ^l^AÕä݉‰ÏgÑŽE|ý5+w¯ôý$ ˆ‹.‚øÒf૯Â7ß”ýzø:Lº|’ûö=_ÜÑÂ#A¬PDB•ˆˆD¬š#…G¸sÞîÛÏ\ú “ú©²àóŒ`yNbÌ+Ì3§˜ð Ö„|÷:É–dB5Ñí´ÓàÉ'=·G‚C‡Ê>æš³¯áŠ6Wðû‘ßÿÝø V("¡JDD"’Ãá M›BÃ*f‡ƒÇ’œ`t ú¶èËè.£ý[`qFé&üõ×¼ö¯L¸x6«Í첪Ì{+&^Äl÷Þ ýûë]»àÎ;Á{«‡Åbaò É$Ù’èÓ¼ãÎgN¡"RbÍ.@D$¶n…¼Òø«3~Õ"¹‹o]Ì?ø']3°X,þ-0È,c kÅŠ òVßßñŒÏ„ïˆ%ΓF@Ìo¾i\ëÐ!˜=®¼n¼Ñó˜–É-ùiÌOtJí„5ÆjZ­":Ô‘ˆä èqÖ8&\<Ö)­ýS”É\û@ÈÎ6·–êò Ø<D›1KӦƗ±cá·ßÊ>¦ë]>DÄMDD"’¿N@$î} ÀÆæÕQ¾ˆ: æ6 n¾ÙXçåk»ÝÜšD$t)€T ¤¤„‡~˜ôôtRRRèß¿?óçÏ?åóvîÜɸqãèׯÉÉÉÄÄÄ0}úô T,". 'ó ›6™WGMxï!Ök+VÄl/¼-[ë„gÂ÷¢q"` 6l“&M"##ƒ3fššÊàÁƒ™SÞϽlÙ²…3fÀWWÿ÷ùq‘pã ))м¹¹µ„ ÷¥x‰ŒHÓ:͹ûœ»¹¥ë-t=C)ÓluëÂ[oûBƇ+Ì­IDB“6¡û°dÉæÎËÔ©S=z4d÷îÝŒ?ž¡C‡ú|n¿~ýØ»w/Ë—/gæÌ™Á(YDJåä¿Àè~œ*ÿ;ΨøÀ;€„kÄ;€´oЉÛ/xÁ¼bä$}úÀƒ¿þ%%pà °|9$UРÚrp iõÓ‚W¤ˆ˜N¦OŸNRR#FŒ(sÿ¨Q£X·nË–-óù\ï72NïëŠHPTuüê¾ù÷ñ÷ïÿN‰£$pE…€:uŒËñBøv@¼G°µï<$ýýïн»±Þ°þú×òw¼ø8÷~q/gM>‹o·~¼EÄt >dgg“––Fâ Ã¥§§°uëV3Ê‘Jð ݺUüØ÷־Ǥ%“xì‡Ç¸ìíË"þCƒ¶gÙ¡Þ6öŬaÙÖÍf—SeÞ„óêßââàw<ñÅá³ÏN~Ü»™ïòü’çq8Œüp$;ïn¡"bvíÚEÃrN.kÔ¨‘ûë"š*ÛY¹{%·Ì½Å}{x‡á?ŠÕ´]Ü×îêÌó4»œ*S íÚÁÿë¹}Ë-°o_ÙÇÜÒõ.m}){íe謡/>ŽˆD>í AãÆ#99¹Ì}#GŽdäÈ‘&U$^\$)©ì¾oûŽícȬ!/1ÞðÜÒõîè~G*4O‡´zpÔXï=œkn1Õ ¬ðqç0ožÑýسn¿>úȳ'+ÆÃÌkgÒóõžlÍÝÊòÝ˹}Þí¼5ä­ˆÿ @ÂÇÌ™3OÚË››~v†š4iÂîÝ»Oº_éG8Mš4 Ø÷ž8q"ÝN57""å:|²²Œu§N`-çì³b{1×½Ûo༦çñÒ —¢âMOǶµ`™bìÌ?lv9U¦Hø°X`êTHO7ºlÜÎÈð<¦AR>þÃÇôšÒ‹üâ|ÞYó]OïÊý½î7¯p/å}¼bÅ º»6:Iµhˇ´´4²²²((sêdffкudœŒ,iV­ò¬}_Ý7ÿ>l[ÀµÏàÃá„êÌ×® êp¤8ü>ÅS /©©ðÆžÛ÷Þ [¶”}L§ÔNLâ9+ë¯૬¯‚T¡ˆ˜AćQ£F‘ŸŸÏ¬Y³Ü÷9N¦OŸNÇŽ•|EBÔ©ölËÝÆ›«Þ ÎÇœsh\§qpŠ ­Z…ƈgá×ñÁ:bßÇž£{È+Ì‹ø‹„³«®‚Ûn3ÖÇŽÁ7—èõv]ûëø[Ÿ¿Æ©ö…ö W)"Á¤,zöìÉ!C;v,‡¢mÛ¶¼ûî»,\¸Ù³g»—‘‘Á[o½Evv6Íš5sßÿÁÆÕ´~ùå’J/„~Ýu×ñßD$ºœ*€´HnÁÏ?sõ{W3¾ïxÎmznðŠ ±±ï¬G!`ÍÅnwbµ†Ï虫w~v;oø€]÷ nž}¾ûÎè~,Y&—ëõöØEq¸ð0wõ¸‹öÚ›S¨ˆ…HÞÿ}}ôQ¦L™ÂÎ;éÖ­óæÍã²Ë.s?Æápàp8NúômøðáîµÅbáÅ_äÅ_Äb±`·Ûƒöï m\Äj…ŽËLzj:kÿ¸–$[§£E°Ú6#€`-a˶ãœÕ:|þ;¸HBäç»ïÖßËpQ»¶qiÞÞ½Án‡Ç‡à¼ó<‰±ÄðÂ@,) 4‚U«ÕÊ„ ÈÌÌäСC|óÍ7eÂÀ´iÓ°Ûí4oÞ¼Ìý®`âp8°ÛíeÖ"°n±nß¾â=Ñü†5%Ñs•½•ëÃkˆwñ¾dk4ÿ~†‹sÏ…G5Öv»1Šuô¨¹5‰ˆ9@D$b¬]k¼±Ê€­F7y^Ê„g·³7;ÕìrªÄµ$1ÑÓ±Z¬Øbl&V%•õðÞ®GVÜwŸ¹õˆˆ9@D$bTöÂhק}ØÛòš±es9×)aå`%Ú£âÊ‘ 6ÖŪU˸ýÆÆåyE$º(€ˆHÄ𠺔°åàߎbgåYoÜh^ÕQf«ôI_…—3Ï„çŸ÷ܾõV(çØ­“¬ß·ž™3W˜ˆˆˆ˜Æá€Ï?‡Í›ýózîbqðÆÞ z¾Þ“E;ùçÅ#Èi§A=ã(6m2·–ªp8 °ôê¬Þ#X ágÌ:ÔX8·Ü]Iy鮥\0ínšssÖÏ N‘"0 "bš‰aÐ èÑrrjöZv;¬Yà¤Þ°?3sý[ää2xÆ`„ßyd±@Û¶ÆzÛ¶²gk„²B¯£!¼G°@Âů½§ŸnÜþâ xé%ßÿhýG<~‡ÓÁ>üßný68…ŠH@(€ˆˆ)œNxùec—3gÖìõ6m‚ü|'ôÿ‡ÛOŒËz¾qÕÔK¨WÃj#k Ëé46‡ƒOAÿöæoùnÔw¼:øUóŠ’jkØÞ|Ósû/õëËì¿úÿ‹›:Ý@‘½ˆ«f^Åw[¿ |‘" "bŠeËŒCÉ\fÔp´{ù  ¼ú<é¾ïµÁ¯qÍÙ×Ôì…#”«á³Ä»S“˜=›ôä–rAó Ì+JjdÀøÓŸŒuAÜpü¸K S®šÂ•m¯àXñ1Íħ›> bµ"â/ "bŠDzeÕßPâ(á©õ·À¹“Ý÷M8™Œn5¨0²…ãFô; þýo8ûlc½råÉ'¤»Ø¬6f›í!% ™5„Ykg©Rñ :»Þ{ïäû«;†õΚwøÕ6ݸáˆaÒ…o2öœ±Õ/0Â.8ÌÎ'àÒ ë”°Ùˆ®™áÝwÁVz”Ë¿ÿ ”ÿØ„Ø>þ!#:ŒŒ¾Èú"H•Šˆ¿(€ˆHÐ}ÿ½gÓùyçRÁèŠTt%_nêt3q¿Ž»ä¯fsO¿Q~«5Ù‹xiÃxèý8{NØt@¼G°@"K×®ðÄÆÚé„›n‚Ã>®a³Úx÷šw†Þö† IDATÉèšÁ5g_ÃëW¾¼BEÄ/@D$è¼Ç¯î¿.¼ÐXoÚ+VTýõvèƒ×áŸé•|­_jŒde6åÇËHb¢yuH`üùÏЯŸ±Þ¾î¾Û÷c­1V^¿òuÞ»ö=bcbƒS ˆøˆˆUA|ø¡±®]†‘#=_¯Îfô•+§vw× è•g#!¶´…ËÁƒ°¿¹5U†F°"›Õ o½å9£æwÊÕt±X,ج¶à'"~¥""AõùçžÑŠk®1>ɾöZÏü÷{ï{DªÂûtÊINH6 ÆoF8tA@"_óæeϹë.رüzD$0@D$¨¼;×_oü³~}8ÐXÿþ»ï ¨ÅöârïW©ºzñ¥3Ç$öxï9¿™ÿ.ú//ÿò2«sV›W”øÝõ×{º¢¹¹0j8U"{÷‡Á¶HR‘ É˃O>1ÖAÿþž¯¹Âœ<†Uâ(aÜã¸fö5Ø'·G\¤n]hÕÊÏEG(O$,ö° ÞýÖUüå«¿ðÇÏþÈ×Ù_›W”Ä‹/B³fÆú»ïà¹çªö|§ÓÉØÏÆÒãõüoãÿü_ ˆÔˆvn‰HÐÌ™……ÆzøpˆõúèÊ+¡V-8v >ø&O†øxÈ-ÈeÄ#ø2ëKþöíßxê’§ÜÏÛ¿vî4Ö]»BŒ>V©”²Ñ°iS²yÅT’wqÆæ»×I¶$ª‘@JIéÓ)œNxøacL3©’¿Õ‹óg0åÐ yoÿêÿ/þ¯÷ÿaq]rODL¥""ASÞø•KR jl<ÍÍ…/¾€3z,eä‡#É>” @lL,g¦œYæy¿ªž¶ Ú²÷è^V/­‡ÓZï,§Í@mº$V$ºè"øË_à™gŒÓÑï½· O¶ …«G@ÇY8qòÐ7±xçb¦\5…I V³ˆTŽ>+‘ Ø³¾.”iÙzõ:ù1îPbq0~þSôžÚÛ>&5ä››¿á¶î·•yŽHõ¼0ðVÞ¹’öK¿‡c§±eKÕ7ÿ[™ˆÕ“FÔ‰\?^Íÿ¯‹“àƒ™ðíãî»ænœKçW:óÃo?ø¯@©u@D$(fÏöl$9Òsø ·K.úÍörð¢ëÉLýJn“syïº÷h™Üò¤ç(€ÔLÛ¶ðë¯Æ'ÌÛ¶AëÖfWä›w±Çh+ÄÇÃ7ßÀ§Ÿ–ýý?•àÑG--xr:c6†bÛvÙÅEÓ/bêÕSÝetÀê‘Š)€ˆHPÌœéYŸ8~åb³Á5W%ò†õ7,Xx¨ÏCü£ß?|^ïß@âã¡];?%Î:˳޴)´ˆ÷–Ãê5‚«¬H–’7ÞXõçuëC†@þ¦+)~~5õÆÜÌáúß’œLÿVýOý"0Á‘€ËΆŸ6Öééб£ïÇŽYÇ8Ô’ë¾aÂÅ|†£G=çW¤§{ΑÊkÛÖ³õ} ÞŸ€—X4‚%»ôRøòKãêxiÂᾤÉú'yî¢)4«×ÌìòD¢š: "pÞ§ûê~¸ôê-l=ÙöÂ&VXlìÝ §VþcW¯6®¿ª®; ¡Ì;€Ô¯M“:MÈ/ΧV\-óŠ’Ö»·qßË.ƒ¬ìšõ OgÂe_Ãg¿žM›|Ÿsä‹Å}ûB›6©IÄ "PN'¼û®çöþPñãcbŒ="O=eüÿ>Œ[þcµÿ£æ©â=‚u_—Çy½ýã¾,Rª[7ãMÿ%—ÀîÝ°nôéc\£eËàÕ1e Üv›çC“ª¨SÇØ«ÕL‰Á‘€ÊÌ4þÂãÓÈÓ›ðü’ç}žjJèM¤æ64N¢‡Ð ÞDmû*hß~úÉ8²²ŒRÞÏü=ŸßÃÜ sýúý'M‚[o­^ø8rþùO¿–$b*u@D$ ¼D—ëæÓñ¥±dÊ¢Ø^ÌŸÏÿs¹ÏqíY»-‚­[Ë?áÜ@bb S§%Ú´µ³d±•;ƒ k…èD“wIH0¯ O­[Ã?Lûö5ö‰tîl<æ³ÍŸñÂÒxaé \ÙöJžø|¹Wß«,§&L€ñã=÷edÀyçUþù<‡ôiÆÚ{lR$\)€ˆHÀ8¥W¿ª³ ËÀûxñðûî¯Møqwö¸Óçüþõקƒ±‡ä¡‡Ê~½¨È(`ü…\Ù’Å_œO墳ٙ·[¿.°ø+6o†.]L.Ω©¦Mq¬Ë.3öíÝ ^Ÿn„‚9æ¸ûɦOø:ûkí÷(÷÷ºŸ8k\•¾—Ó >O?í¹ïÑGáÿ(ÿ2ä¾ìÝ Þ¼:$ü¥¤ÀW_ÁÅ·ƒAƒà“O, ï0œ woàžsï!Æb¼MÚrp £>Å3 Ÿ9åk—”ÀèÑðòËÆm‹^yþò—êÕZ»¶Ñqqu†E™ˆˆD›äö8r7¸ØmÜÞe,ïÞÈÍoÆRÉ+ÚŒ® è5ç: ¡Ð™ ¹@hw@\$>¾êŸ"‹œ¨vmã„õ+®0nÀÕW¡dÉ‚ºL0‰_nû…‹[)¥–­·u¿­Â×,,„#àí·ÛV+¼õÜqGÍj½ývhÑÂXÏŸ?üP³×1›ˆˆÄ·ßX±ýür'ƒ³·ðêÕ“©ŸX¿J¯1l˜ñ8{IÏ×¼H¨îYu®õŒ1¬PF°4~%þ’}Ç{îsrÎ9ðÛÏÝøêÆoøaô¼0ð&5ôùZùùF€ùè#ãv\œ12USÜO_ö*X=Tý+j‰„ ˆ3€u×Á§/sÛðæÕzÓN3N3ضÍsšºÃ«Vë-<—‘•ªiV×s¨Àii;#€„êW$!úLëC墳Éø_†¹EIØ‹‹3>à˜2¥ìaË–Áµ×WäÛú}_nì8ÆçkäåÁ€ËÌÿº0Bò'ŸÀСþ«óÆË ƒñgá¼yþ{m‘`S‘*q:ü¸íG®ÿðz^[þZ¹9v >þØX§¤Àå—WÿûéY»Æ°²²à¨±­DãW5à@´6: yyÆUwB‘wY¼s1‹v,bUÎ*s‹’ˆ·Üë×Ãÿ·wçñ1Ý{À?“É0Y„H4H”¦$ÛU"ZE­EÑ ªª½èØQzKàÿ-à”€ì6Átܽì]—w!/?Ïbõ üýåñù󒢜¨2bBDEºغ8pá,Fí…º êbìpçBákn§ÝÆ­”[&ûN7œH^..@Ÿ>òøî]àÀN@·”æÍÔ9kú­AºC Ë+êDô‚*Nú&h TT*é¹ýÏdõ‚‰ê ™¨ž}þø88¨`¯’åÕ(Øuez¯ïçCžÇÜ𹸓^ö®E•  ÒÿyYDFÊ°®Š:ü’ž< @ˆÈĵk2¡ÿä]趹%¾?ý=ÒsÒ Ÿ¹ÞËXÛo-b&ÅàÙêÆó;¥å<=[KëálX @,ÃÓÅSÚOÁ»ÍßE¿ Ë+bH^žþFKã¨+,w°çŒt*_íÚÉ|‹³ge}"»wN))À­í/Ï>«Â¹Y¡¸5éfwšm4¼ñÆý˜úËT…qô’ëÚUŸ>øÚ5™»R‹MšHÚó 8¬‹¬ƒ‰Œ” !:Àõ.€ÎMžÈrF£û£±¹ëï86ü†4­½érÔ›7K|@.ÖY¬Ê¢gO zuy¼uëƒáÜÝ%È¡²óõÕ?®ˆ= †k€ ì!kiÑB&«_º¼ÿ> ÑH¹ô’4lÔv®Ï:|†ë®cÛÛÛУa¨ 9£G·m±ºö‚Ìš%¸Jcî\VVà›o€#$à'*O @ˆžbÙyÙF?Ÿ9#ã™ããåçúžZT9>Øù-° W-Å€Wš#08q¢ècZrøUªUe. sK’’äqË–\ÂR¼¼ôém+bª(6¾ÿ^i DêÕ3~ÚN¾¾}±{Èn\K__ŠÆî‹=n|jØñv\Ú”¬”b_?{¶>hîÝ{¼zççãÇË>‚ƒ%ÉGA¯NXп¿iÆ/"K±·uˆ¨|Ý˼‡Ãчq êDÀ¥DãæíC×Au½ÞxCßÿÊ+2æ¹`ØÓŒÀ¤IÀ²eÀ‚@l¬”ÿò‹lþþÒšvåŠ>yçËöNtìÔ©£ï€Xšpîœ ¡‹Ž6^ÁÖ êjLn7º:Öïh»JYÈ]Ý]Œ:ˆ<%ËÎ,ò3Ë V©à€®Þ]Ñõ¹®ð @µ>¥`“&À{ïÉJëÉÉÀüùÆiz‹’—Œ)½–,þñyìä$Gf¦~•ømÛd•{@¨RÊË3͆D¦îêîÂ}ž;úmè‡%§–˜^.^ØžŒ×_×Ý»{öèƒÎÎÀĉÒ:¾|¹ C(pê”\´>ýT_f¸~‡%¨Õ2§ÄЋ/Zö=žv†ó@*Ú0,ÖXwM}w FHÏ ð`»JYHjV*6'SaYž’‡ã·ŽcÖ‘YèÚÛ/m7ÙïóÏõ½_}üõ—ù÷ÈΖïå‚àÃÎX½Z|2ßn÷nù¾dUønÝ$ÀyR]½ lÚ$IÈz€P¥’•%­ðË—â‹/Jë ÍÝÑ>n>…?ÛÛÙ£]½vøïŽÿãÃ#Øã|«p¬o¿~ÀöíÒ fNÕª2IñâEÓ» z?üý˧õÜp–³³qDeW‘'¢ö€hMsUjϹ>‡õÖãÎ?ï`ÛÛÛ0®Í8£9#*¨Ð©A'“ý4F’ÇééÀœ9E?#C¾ßÃÂägF†\ýý羚}õUiàsu•ŸOœ²â‚›ÊèÌ`à@éù8PÒ*öYÅ]ˆõIÃ!XT)¤¥É¤¿ €¸8ãçÆŽRS[ߟTùJ>®$^AD\"â#“ƒ°aÅî3ìoׇ®Þ]ѱ~GT«*}éË–É…« hx÷]`Õ*Àþ1¿ ì8PzLæÎÂÃå¹>(í',^«VÜœ:%-uvlB)³ôìtœˆ=[÷oAç^À«*^z8jÑ×·/úúö söFÄ•¤+pst+rŸiÓ€•+]ÇñXœ|ªŸýÑõ…Öh]·5êT«ƒÔT o_àðay½V+“ç{ô0_€y}·nrC~îСƒ&O¸¯LE®SAAÀÞ½ÆÏÝ¿/å‹É0µO>‘ „ʪВ’$-`Hˆ>óQ¦M ÖÄ›2E‚ÃIyO‚¤Œ$쿶ñˆˆ‹Àé¸ÓHÍN5zMBzj9Õ2{Œµû—IÙÂ…ÀÇë=Zz’JsC¯RI0г§LZOJ’‹VyP©dxÀ±c2žÊ.ú^4ºü ¿ÌAßCARÑz@ ‡`9pézJx¹xaè߆ûࣀ ”½PÜ/#ä÷Ýù]ž«íT™Q­q/ߨý:ª¥·ÄÎLœ)ŠÌi™;8~Üø¹gž‘ϵ}»$ßÈÈûŽ¥K¥anòduA–ÅöCªbcåùÙg™3ƒ‚°çÏçB ’/á'i¥Kw/áíÍocþñù8}Ø$ø€ówÎ?öñEÆ Ÿ||û­ezZ·–9$庹Ik^qÃÄèñÕ«®o쓃ZbYö€Uã&fBå˜dR~;=÷<¶§Ã¡ÉA<øxÁG B Ž7äf]V¯ørseí–¿ý èÓÇ8øhÐ@Þ¢£eȵkráè¨ß74ðó“4ð§OÛà<ÁØBÊÕ«À¼y21Î0¹Z "-~~úò)S$;GÁBJ!!Ò²l™eÀ³¤´ì4\N¼ŒË‰—qéî%\NºŒ—¼^ÂØ6cÍîÓ¢v ¨Ujä)²*T=—zh]·uáÖªN+³ÝòSàŸÿ”alfÍ’îû'©×ˆJÆ¥ª \ªº %+·RnÁ×HHlc))€‹‹­k(€™WÛ]‹ ×;˜òe4P7õ_ŠÀM2jDZ™]½xjkøû›?ƾkû0û?³áëî _7_ø¸ùÀ×ÍÞ^Þ× [7 <âã%+áÞ½2,¶"ÊÊ’ûˆyó$°0äç'÷ƒé'ð2´lÑ"™"½ ÷îɵsófÙºu“}+Ê÷beƤKIþ÷e~DI¸¸½zé³\T¿ÿ®ÏCn؃¡ÕÊ„çO>‘ÖŠ¢Œ+ŸeÄÙwÕ*ù¬] T©Rô>å%'Ø·OVènÓøæd~¾ø3.'^Flj¬ÉëÓ²ÓŠ @5ŽøºÇ×hP£ZÕm…ÚεKU¯ìlÉûþÝwú²E‹¤µ‡ÈËÅ ‘ ‘ˆI‰A{GJDzåJŹÁ0‚Å„ÈÔøñ*„„<‡øÈçp#r ªòQÛï>Yx Û2Ÿ½}á7Ã~3ܨ\­RÃÛÕ~Sþ‡¯6""HL:w–aKË)¶¢ÈºRW®”l¿Û·eøÔÃóEÛ´‘…ûô)¾ÇßÝ]ç>ùD®™ Ê1`ÿ~Ùš5+YÈJèÎà믥ëðþýÒ£fMé57N†´Øʱc2&s×.ãrI øÑG2¾õQ† “ äw$ “´²aa–/ž‘“¸Ô8ĦÆ"6%±©±ˆKìN³à\ÅøüûßÒ /¿ Ôr‡™=î•ÄG³Ži3¦ÔõNO—”¹óç11R¦RÉ„þ‘#K}XzÂÔs©‡È„Hdæf¢žO"w²ÆÀ”)’þØ°¥Ðšrr€Ÿ~2n™]å/\OÖÁAãwGwØÛñRFäèLŸnœV·áóv8°«ê×tZ¸Ը"Ëó”<\Iº­½G½{Ë°¬”ÉŽõúëò=Ñ®0óðL(PàUÍ ^.²yºxÂUë ÕcvµçåIoÃܹÀÙ³µK±ºv•úuêT²Þ~50nœ Åš7¸~]ž;ÿø#ŸÉ…ÌÊÉÉQ¦L™¢4mÚT©Q£†Ò¹sgeÏž=µoZZšòá‡*>>>Š»»»Ò³gO%""¢Ø}NŸ>­PNŸ>]äó7n(ʸqŠâà (Ò6PöÍÉIQ&MR”˜˜ÿzJ-?_QvïV”öí ëò£(J­ZŠ2gŽ¢$'—îØ»v)ŠV«?n§NŠ’’RTò•œ¼œby'RiúmSÅ5ØUÁL¹ŒŠT‚ƒÅÃÃÌï¸í"3¡8ÏrS^Zþ²2ìçaÊÜð¹Ê–È-ÊwþP2s2K÷A!)IQ¾øBQÜÝë£V+Ê?–Ë[–È¡Thäö‘…Ó¡{Ϙüׯ¯(‹+ŠNWºã—æÿ[§“÷¬_߸.*•¢ô_9º°¾¿ÅÿVºJQ¹áùm;ÙÙŠâç'çJÓ¦ŠW²ýïeÜSNÆœTÖü¾F™öË4eàÆJ‹¥-‡ÙÊ›ßTEQÒÓ¥{wÓëw‡ŠR}¶{‘×J‡ÙJ£Fʺsë̾wV–¢,]–©<ÿBŠä—ùþ¦_?E9y²,¿Mc99Š²nü^âï×èÑØlTŒbß¾}˜3g|}}±fÍôîÝ7nD¿~ýÌî—››‹N:áƘ={6<<<°dÉtìØ@Û¶mKT?ÿ¾üX·N&EÐh¤…ò¥—Jö¹ÂÃåXyyÒB¾p!°x10t(ð¯•ßÚ ­ÁÁ¦ :8¬Ç¼yƒ1|¸~Xiôê%Y’úôÒ|¿Ã¡ª‘ðþ8/¶KBZ^2’3’‘”‘„äÌdL˜€yÝæ™=–£Æî\(öý:õCú…ŒÊºu“®ß?þpvðû{HËpÃ_ÞÀ°ÉÀв–Fy¸}[†V-]*sa õé#+š·n]>ï]ëׯÇ`K¯TH¥æUÍ«ð±ký[Ø¿¿%fΔJ@&žŽ+Ã&N>üÐt¡Êâ”äÿûþ}ùû]´È4ÿ+¯HRŠRuÀM)sÔ”á ƒÊÏoÛÑh$}îñãr-*éõ´º¶:ü=ýáïiÍ~¬Ó~à>@žйA‹šðtuC5ûšp²sƒ³]fÃ^e~|µôÆ®kd öö2Êâí·eʤI–=þÓ†ˆ'NœÀ¶mÛ°råJ 6 гgOÄÇÇcúôéÅ ›7oFDD~ùå¼úê«€×^{ Í›7Ç_|]72#"Bº ·nÕ¯ÕÈÉÿÁ’ÉÈËËüþæŒ)™æÏV¬ÉÙÙ2q{Å Y×aÊ E‹’Û¢(ÈÉÏÁ½tÖü¤Ã7ÿ£Ã8 ÑZ_ ÓËú›6É N£7"ôl(t9:däf@—£3Ú5Ž8ýAÑ)) QzeùzäzÁ]û¢M_—”aš1ÄPju µ×¢¶smxV󄧋'\à‰‹'=qâ€'r=‘þ—ü’T*É’ñ駲8b~>°cT'OÊñ¢¢dÝ™3å‹kÔ(™@o QQ2ükÕ*™|WÀÎN¾,?ý”cVɼ‚LXNÈÌÍDß®2l!<\¾ƒvï–×ݹ#ß ÁÁÀ˜1À„ ’ÂÒŠZÚ³§¼oûö@v^6>[¥OÑÅ„ÈX­ZÀoXö˜v*»Â5¤ tÆŽ•{‘€¹¸xI,= ¸Ä.1¨Ù ~mc }&±©1ˆI‰—‹þ¦%)I?CBdN  Åƒê Úmdâ6®åxÐøªV©±{Ò—Å¥zgó;˜²ã œ÷:ÃIãç*Îpªòà_:Öïˆ÷[½ovEQðÛíßà¤q2ÚO£–q¨vvå7ïåiÂČիWÃÑу 2*:t(†Žˆˆ´6Ó”¼zõj4hР0ø€ªU«bðàÁøâ‹/€ZµÌ¯Ûpê”Ü0îßoP¨ÊCõš¹û{|4^ w÷¢÷MÊHµ¤kÈÍÏEN~Žü›—Sø8_ÉGÿú£A9ñ§O¾úJÒ°¦¤Èó†?~†‰‘ðn”ƒ€¶¹x¦NNáñròrÐƳ F·m¶þiÙi¨» .t9ºÂìMƒ/Ć'váË‘½('óæÍÆǸ–t +~[aö=ªª«BQ³cJ€¿»â—xãrgM5Ôtt…«Öž.žfUÔU ›ªƒJ¥ÂÅ‹Ò º¶ˆ^¨áÒsdØÚbg'€¾}C‡ä&îÀy.>^Æ•ÉØÒñãK?ç¹üé'éa*¬{à¿þKÞ§²åk'ë{§Ù;ÒlªÚw͵o/Ûo¿ÉßYX˜4†XrÁ®›7¥1dùrã‰æ*•4†|ú)в¥”ért°qNÄž Ó3NŠ€ˆ¨ÄªT‘ìQ?ÿ üü³AA-pú´4Ì%Žn–†Ò?Þ8:*ˆ‹“‘ß}gšD' ¹òjvFn•$$ꑘ‘]Ž®ðùš59—$A—€;éwp'½è%Í«ÚW-6IÏIG«ïM'ì»;º#០ž7=> fDEE¡aÆpxhs³ÍÈׯ_7€DEE¡yóæ&åÍš5ƒ¢(ˆŽŽ.6}¼3Ð €`—+-*÷´zk ÜÝÍ÷¾ì¹ºB^èý gIDATC¶ 1û¼Z¥Fî ý´‡‡ÜOž,AÈW_ M6/lE€¨›(êPàÂ¥ ØŸ5€ä)Ú"׫0ôoú71ÿ¼ƒ¦è™ã; 5ŽpÔ8"+/ Z{óép¾zcþ¼6F¹âvTM ³Üêi0uêƒ WåW‹­&E…½{-[Šî…š4©øUaU*ÉÒ¹³–=Zœ,CZæÏÞ_V|\ùù2qûvãrgg3q"P§ÎãžnÅGîvÝ&_Dóa²êýÿýŸ¼¾9,ÙWïµí…W^)zÿ›7!k«d•_¹"kúäçpjd¼F?´‘Á¼ÙkI×Ð{}o\¼{±°Î«ÞX…*j+§º#"vv@ÿþ@¿~ÒØ6w®4¾’eâDY(¸S'¶o7M³?x°Ü‡4mÚ@o£cgæf")# IIFÁˆ9Ï8=ƒúÕë#-; é9éÈÌÍ4zÞISü"RiÙE§ÕØÙ(ÇŠˆ±±±ð("ýRAàkšRÕpßWŠ¸îë_D2îÌ‚D÷i÷ä¢}éÂ%œÉ8cö½o^» Ä›}yÈÃéÓ§‹lAèÑC†/ ÿ1—Š9Æ©¨;8µß|Š=«rµp«®…·µÝµÐÚk‘›…3YúcÜ»wgÎèöÈòÀ†v µ×ê7µöjýŸlä¹Èâë ¡¶ –§côètÄÄÜÂ2ô©´œeHÓàÁ@²^BÂc6ˆ¨Õ²æÆ!’Qc÷n¹‰ÓédèIYT¯.u4H2wÄÇËVQ=üÿMÛ¢ðEØòç}Á‹ÆÏçÇøcõêÚX½ÚÜîá÷3cm ís¦â­îðî»Ò0’š þ‰œˆ9‹ç%øpÒ8áë_Ã#ÕƒGÏï§ËÃÿßnnÒ°vþ<°r¥¤Òd˜Õ¦Múý4)ð÷¿žž”<êÏÆö8óWñ/ú¸ÁÇø¸~µÝ¼üãíímvßçŸçÎ3)?þ>>X·nŽ;†7¾nĈøᇅzÒ!õïßþþþxë­·0gÎÔªU K—.E\\~øá[}$"""""›cH1ÂÂÂ0a¬X±C† ÁíÛ·±sçN¾&??ùùùF«{ªÕj:t À¼yó0bÄØÙÙáÈ‘#°ÅG©´–/_;;;T³ÔŠ}T¡Bpp0êÖ­këê•%''ãüùóhÞ¼¹­«B¥…† š¬eÔ¬Y3à éS‚çñ“×i*+ TáŒ3~~~=z´­«B60|øpÀ´iÓl\*ØØX¸»»›”¤µv•Èx?Ùx¦²bBåæðáð³³{¬­`áÆM›6açÎX¶l™kO%Ušÿï‡M›6 Û¶mâE‹àããcåO@D–ÀóøÉÆë4YסrÓ¸qc,_¾ü±^ûì³Ï"-- cÇŽÅøñãááá{÷î²³³÷ï߇½½=œœœÊ­ÎTz%ùÿ.X3ÇÐÌ™3„¹sçbÔ¨Q–®Y‰§§'âããMÊ Ÿ§'Ïã'¯Ód)Ì‚EFtt4¼½½‹}M`` ¶lÙb¥‘µ|þùç…ÛôéÓm]*ƒ±cÇ"44wïÞ…V«-, ÅðáÃqêÔ)´jÕʆ5¤òÂóøÉÇë4Y ª0²²²ð믿B¥R–)Š‚àà`9r{öì»»;üüülXK²´Ù³gcÆŒ˜1cfÎœiëêP:u Xµj† @ÎãÎ;#11Ñìð;ªÜx?x&Ká,ª0ªV­ŠŽ;š”¯Zµ jµ:t°A­¨<-\¸3fÌ@Ïž=ѳgOüúë¯FÏ·mÛÖF5£Òò÷÷G`` ÆŒƒäädøøø`ݺu8vì6nÜhëêQ9àyüôàuš,…Ux*•Ê¨µ…ž;vì€J¥ÂîÝ»±{÷n£çT*òòòlT3*‹°°0̘1+V¬@LL ^|ñEìܹÝ»w·uÕ¨ð<&^§©¤8‹ˆˆˆˆˆ¬†ix‰ˆˆˆˆÈj€‘Õ0!"""""«aBDDDDDVĈˆˆˆˆ¬†Y """""² DDDDDd5 @ˆˆˆˆˆÈj€‘Õ0!"""""«aBDDDDDVĈˆˆˆˆ¬†Y """""² DDDDDd5 @ˆˆˆˆˆÈj€‘Õ0!"""""«aBDDDDDVĈˆˆˆˆ¬†Y """""² DDDDDd5 @ˆˆˆˆˆÈj€‘Õ0!"""""«aBDDDDDVĈˆˆˆˆ¬†Y """""² DDDDDd5 @ˆˆˆˆˆÈj€‘Õ0!"""""«aBDDDDDVĈˆˆˆˆ¬†Y """""² DDDDDd5 @ˆˆˆˆˆÈj€‘Õü?,Z°ˆ õ‘IEND®B`‚HDDM-0.2/hddm/examples/plots/simple_difficulty_trace_a.png0000664000102500010250000013300111726176127024431 0ustar wieckiwiecki00000000000000‰PNG  IHDRèXñÌ%sBIT|dˆ pHYsaa¨?§i IDATxœìÝ{œeýÿñ÷. ¢((ˆâS9xB !ûF¨Yú +ÒL󄈅Ú/µ¾”þÊC¥†–…™ÇR´RCM2O@æYØÔoˆ (,Š~9îþþ‡{öš™kæž¹gî½_ÏÇÃÇîÞ;‡ë>,Þïû󹮩kiiiÈU}Þt €@Ð(:@@ è€  Pt €€@Ð(:@@ è€  Pt €€@Ð(:@@ è€  Pt €€@Ð(:@@ è€  Pt €€@Ð(:@@ è€  Pt €€@Ð(:@@ è€  Pt €€@Ð(:@@ èrå•WêàƒÖ6Ûl£¡C‡êÒK/Õ¢E‹B÷ijjÒ׿þuí½÷ÞêСƒÎ?ÿü6ÛÜvÛmª¯¯/ù¯C‡Ú°aCVwPÃfÏž­ãŽ;NýúõS}}½¦OŸ¸íøñãU__¯ë®»®‚# zÐ+döìÙ:ï¼óôì³Ïꪫ®ÒÌ™35~üøÐ}Ö¯_¯Þ½{kòäÉ&I477ëßø†¾÷½ïiРAy€ªB=ÍÍÍš4i’F­}÷Ý·¬c 4H·ß~»öß566êÖ[oÕUW]¥¥K—j«­¶j³ýŠ+ôÈ#hÀ€êÒ¥KYçØûøãõÖ[oéóŸÿ¼¶ß~û¼‡“™Ÿþô§êÔ©“Î;ï<ë}øä£VþßTMè9˜8q¢Þ|óM=õÔSeëÐCÕ¡‡*IÚÿý5vìXõíÛW÷ÝwŸÆ×fûGyD'Ÿ|rÙç$ó‡?üA'tRÞÃÈÄK/½¤ë¯¿^sæÌ)¹½¥¥%t?þßùjÏÿoª6ô ›8q¢~øaÍš5K;í´SêÇïÒ¥‹F­{ï½×ÐwÛm7IÎ!­‡¥&Mš´eÎ?Jñؘñ¸ã±ikÞ¼y:ùä“·ü;Üýãÿлﾫ]vÙeËm›7oÖ…^¨ë®»N .4îÇÿ›‚ñ·ŒÇÆŒÇ%M[µðÿ¦jC@¯ ‰'júôéš5k–vÝu×LαnÝ:=öØcŸ€uîÜY’Ó?lØ°LÆP­zöìÉc€ÇÆŒÇ%M0÷ßáöè”SNÑèÑ£·üÜÒÒ¢Ïþó:å”StÚi§îÇÿ›‚ñ·ŒÇÆŒÇ%M°öüÿ¦jC@¯ &hÚ´izàÔµkW555Irþ¡pÿ .¹ä-]ºT·ß~û–ý%I«W¯Ö»ï¾«ÆÆFuêÔiËÜõË.»L‡v˜öØc½ú꫺ùæ›õñÇëœsΩð=Ô‚5kÖhÁ‚[~^¸p¡Õ«W/õïß_Ûm·]Éö;vÔŽ;î¨Vz¨Tz…L:Uuuu9rdÉí·Ýv›N9åIRSS“–,YRò{÷S¾ºº:Í™3GwÝu— °¥MðÃ?ÔYg¥åË—kÿý÷×ñǯ«¯¾Z{íµWî Ö¼ð :ꨣ$9ÿoºà‚ $I§žzªn¹å–<‡@Õ# Wˆ{Yµ0·Þzkìý®½öZ]{íµ‰Ç@£F²úšëÍ7ßÌp4´/\ø„iQ=8xlÌx\‚ñØéào)K0T:ð þÑÆccÆãŒÇHKÁxlÌx\‚ñØ Ð(:@@ è€  Pt¹›3GZ¶,ïQùjÈ{pàRïÞÒòåyÈt…ðî»yÈ€  Pt €€@Ð(:@@ èÈcIýúIÍÍy }  ¹üriéRiƼGÐ>Ð(:@@GæV­’>ü0ïQH[]óµ¥%ßq´ yí_ïÞÒúõ¼‰Ú7  TБ¹õëó€,ñá@:è-î vi@{±`Á­^½:µãuïÞ]LíxŠ‹€ èÕ‡Eâ ­ h¯½öJý¸¯¿þ:!¨t¹rýzРUkåü’¥pÄy’NNµ" ¸èrÕÜì|% W*èf¤ay@•a‘8¹rz=ÿU-:@:xK WTЫ t¹" :€\Ñâ^½˜ƒ.ÞÈôêG@H@®èÕ‹ :@ºèrE@¯~t€tÐäŠ9èÕ€ÞÈtÀA@+zõb:@ºèrE@¯~t€tÐäŠ9èÕ‹ :@ºxK WTЫ t¹" W/*èé" È•Ð€ZG@+·úJ½zQAH@®èÕ‹w€tÐäŠpWýxÒA@+‰«^TÐÒE@+Â]õã9H@®˜ƒ8èrEõµúñ¤ƒ€ÐÇKwß÷(*ƒpW½˜ƒ.:P@“'KãÆIÿùOÞ#Éá®úñ¤ƒ€Ðûï;_ׯÏw•ÀôêE ]t¹â2kÕ€:P`µZ w@õ™={¶Ž;î8õë×Oõõõš>}ú–ßmÚ´I]t‘8àuïÞ]Gq„~ö³ŸiÅŠ9Ž€ê@@+ZÜ«²Ôžµk×jèСºá†$Iuž?à5kÖhîܹúá¨9sæè»ßý®î¼óNMž<9¯áP5ò€ÚF¸«^ÌA¯]cƌј1cŒ¿Ûf›mô÷¿ÿ}ËÏÔV[m¥Où¤¶Ýv[«ý¦L™¢aÆe<:¨M¦<çÌ™£<0§Á„€ WEª¾"™¢<‡Ë—;_/®Ìùj9 ¯Y³F ,ØòóÂ… ÕØب^½z©oß¾;v¬æΫ¿þõ¯Ú¸q£ššš$I½zõRÇŽó6…G@+*èÕ«Ö‰«å€þ /訣Ž’ä\ý‚ .$zê©úÑ~¤|PuuuXÛl³†ªK/½T‹- ݧ©©I_ÿú×µ÷Þ{«C‡:ÿüóÛ=öØc:ôÐCµÍ6Ûhøðázâ‰'²¸ 5mÍšÖï‹F²P”9èíù1nx¾ÒA@¯Ù³gë¼óÎÓ³Ï>««®ºJ3gÎÔøñãC÷Y¿~½z÷î­É“'kðàÁª3”‡æΫcŽ9F;í´“þú׿ªOŸ>:úè£õÊ+¯duW€T±H\õ# ¤ƒ9è2cÆŒ-ß4Hýû÷×~ûí§ùóçkï½÷6î³ë®»jÊ”)’¤›o¾Ù¸Íu×]§ƒ:HùË_$IŸþô§5|øp]wÝuúÝï~—ò½ÒÇeÖªWÑZÜ]E€-*è9iþde¬Í›7—uœgŸ}VC‡-¹mèСzúé§Ë:.j sÐäx €´ÐsÐÜܬI“&iôèÑÚwß}Ë:ÖÒ¥K5dÈ’Û¬wÞy§¬ã¢6e:ªCQ+èÕŠ÷Lœ8Qo¾ù¦žzê©ÜÆ0iÒ$õìÙ³ä¶qãÆiܸq9¨øòµZ¹ÌZõ+ÚóU„ñL›6MÓ¦M+¹måÊ•9T z…Mœ8Q?ü°fÍš¥vÚ©ìãõë×OsçÎ-¹­±±Q;ï¼sè~S¦LÑ°aÃÊ>-©ä›þZªLÖÂ}lïŠòš>ìÉ‹éÏ9sæèÀÌiD Ð+hâĉš>}ºfÍš¥]wÝ5•c>\%·566êðÃOåøÈG¥Ožp™µêW”ç«(ãHŠ9è2aÂÝyçºë®»ÔµkW555©©©IëÖ­Û²Í%—\¢o~ó›%û566ª±±Q«W¯Ö»ï¾«ÆÆFýûßÿÞòûï|ç;z饗ôå/YÿøÇ?t '¨±±QßùÎw*vßJŽ",W íüíMÑ:=Š2€¤è2uêT­ZµJ#GŽTß¾}·ü÷Ç?þqË6MMMZ²dIÉ~Æ Ó°aÃ4wî\Ýu×]6l˜¾ð…/lùý!CôðÃëí·ßÖ¾ð555iÆŒÚo¿ý*vߊ쩧¤{ïÍ{C¨BZŠÔâ-îÒlñÎñÖ[oM´ßg?ûY=÷Üs‰ÆÕÞá|ݼYª/óã¨<*¼µ^¹zõ*j½Òã)ÊýÕ :jBµ½Î+øäñ8Qõ¬~Eùû*Ê8’" £&Ãm:A«ºåùâïT;:jBµ¾q/JðÉR’û¸r¥ôÓŸÖÆãS Šò<eIÐQª5 ×‚$ôï_ºøbiÞ¼ìÆ…hÌAH5!i@Ï{³Z˜ƒžä2kë×—î H|ª5!é÷¼ßð×ÒuгÞé+j½VÎ Ú:jÂæÍéí—g˜mOžxB:ê(é[ßjû;ÛÀÃsQ E ¨îj•¯=:jBÒJøw:_;vl½­=·¸§íÙg¥5kÌ¿{ôQéùçË;~µ?>ÕŽ :@ºè¨ IúòåÎ×cMo,y˜7OÚeéž{·K;àv˜tæ™ÁçêÓ§2ã¨e¿ø…Ôؘí9Šò|eIÐQÊ]$n«­ÒK’ó—kþ|iÉiæLóï³l×}í5»í’ÜWÚŒ£]p4|x¶ç(J0Î{Í€r5ä= ªu‘¸´lÚä|íСòçz [Zâ]ZÍ¿/ìeõxeÕâþÖ[ÒŸþä¼^O9EÚ~{»ýx]€jG@GM(·‚^í«C1 ‡a‘¸ÚöÛßJW]强u“Î:Ën¿Jü½64´þ=¤wÔ„4*èy„ÁJô,‚MtÄ“õc›öñ7m’öÜÓYœ1Î*±Šû¶Û¶½×.H 5¡Z+èi‰ èY~ø`ûØ{ãj¼ÓðÖ[ÒvÛIPþ±ª- »êê’M}ÈòþÒµ²D‹;jB¹=/iß­BæÑâT¥‚nñbéÿ×ù¯\Õ6Ý}m$ èíeíP{¨ £ÝX²$øšÛi´¸Ws€ÌsÎla©šŸ [ivTs=Îë§-îTÐ@–èh7vÙE:æó蝹Å=í9è•:ŸW’9è,—î‡CTЪ¹JûôìÙéž'¯EâÒ>n@:^–@$™ƒæäKùÏA_¼XzûítŽ`ƒ€ŽÜ¼ü²3'úų?WµÎAOKT@wU[½=KóëjkqOZAw³´»¥ý÷OçX6èÈÍܹÎ×—^Êþ\µÞâu©ª,ï_œËd¡U5}`Q”€žv‹û† ÒÊ•¥·™:iŠþü€êA@GMH£Å½šÙVгE½Q5,P˜Õtïñ¹Ì¨%tÔ„jmq¯æ9èq«™IÆОÃRÞ¯½ --ÒÒ¥moKûEhq¨4:Ú• À–´ÍÚT‘«DpJûÕ¶Š;Š[A¿öZ©_?çúìTÐÒE@Gî*>ªµÅ½š+è®$a‘i^Š:ýÙg¯«WgwŽj© ÷ìY™ó€ÚA@G»TÝJ³Å½´´+“nAA/¨{ z¸¢ôúOþ¯‘åøÜ×F}}ñ®ƒî}Í~÷»ÙÔ&:r—fàM; »û) %Õâ^„ :‹Ä•Ê»{#ˆéãZmqϺÅÔ:r—æ›ÛJTÐ+©R-îiŸÏ{,Û9è(UÔ ºËûüÕê"q¼~@ÚèhW*Ћ–¢TÃôj|\³TÔEâ¼-î•X$.Îßn^t€´Б;Z܃Uº‚ž…,®ƒîjÏ©¨ôJ´¸'­ WbºW{~ý€|ÐkX]ôíoç=Šb¯âžw0J; g}ÛKÜåýXMQç »²ž¢PN‹;tP­è5î—¿Ì{•Q­—YK‹í*îy^=Î8j!Ì×zÝ=WÑ*ètiöìÙ:î¸ãÔ¯_?Õ××kúôém¶ùÅ/~¡}÷ÝW½zõұǫ%K–ä0RªíJVsÐË ï½'-[–üüåÊczÜ°T¤ZEƒN‹;Ö®]«¡C‡ê†n$ÕùþáýÕ¯~¥‹/¾XÇwœîºë.-^¼X#FŒÐÆó.U¥!ïiÊr‘¸rªe½{;_?ø@ÚvÛäÇI*ªÅ=Ka ”;½=+êcà>g—_.Í™ã|_”ˬUz͈Z½ÌÚ˜1c4fÌãïZZZtíµ×ê /ÔW\!I1b„zôè¡{î¹G'Ÿ|r%‡ @Õ¡‚Žv%(ðe‘¸5kâm_Íô¸’Œ¡=·½‚~ÇÒ«¯:ß×j½=¿þ’Z¾|¹Þzë- :tËm]»vÕÀõôÓOç82ªíJÚ=àQ„Êp×AÂuÐÃáucbzΊRA¯ÄeÖÒêªi¯ÞyçIÒ!CJn2dȖ߀`´¸#Sao”³xs›U½ÞcÄ=^ZA#j%õ,ç 'Ù>«>ø@š:U1Âù¯Èª) §-iÝýûZ¹2›qùÐíµ´´´™«n2iÒ$õìÙ³ä¶qãÆiܸqY jÆ´iÓ4mÚ´’ÛVVêš°F@G¦ÂÞ\g:Ê Æ®4*rå´*§ñÍ+=ôPzÇKKØô¬üíoÒ~à„óÙ³³?Ÿ­çž“/–N<±õ¶´Û´»u“N?]ºî:iíZéÅ¥#ŒœJTÐËmqè!é·¿•Î:+ÝqùÐÛêׯŸ$©±±QÜr{cc£>ûÙÏFî?eÊ 6,³ñ@-3}à9gÎxà9&´¸#S•ªz/Ëd:_-·¸ßzkôù³׺uÑÛTâqr‘.Ú*ßÇK_ùJémi?kÖH¯¼â|ÅÒÈ‘É®,P -î’´tiºcY¾\ú׿Šõ!WõéÓG»í¶›æλ嶵k×ê7ÞÐᇞãȨTБ)ÈÈêÂnà èQ-ÞQÇ-GÞôæfið`©Áâ¯=«ðñ m[ÊÓ˜ƒww.~5T>³X$Î=μyÎ×õë㣒ôúúø-î»ïîܯ´Çt Ò3ÏHýû·ÞV ¯£,¬Y³F ,ØòóÂ… ÕØب^½z©ÿþºà‚ tá…ª¥¥E£FÒ÷¾÷=õïß__ýêWs5Õ€ŽvÁTA/b‹{•Û–'è¸ßm“%›Ë¼U¢2™ôƒšÿmµà…^ÐQG%ɹú\ I:õÔSuË-·èÜsÏÕ† tÓM7éÆoÔᇮ©S§ªÁæSBjÿ·D¦ª½ÅÝ´_ÜûQ„ z’ù¼åŠzü˹ºÍýhnv*©]º´Þæô¢µ¸›d1FÿNE èå,çî‡ìŒ5JÍ/ÐóÏ?_çŸ~…F@ûÁtdªRÝÔâžf½Izš•9›ªb%ì ’ö.¹DêÚµô¶¢ôŽ¯Þ 5UÐÓVÎ"q•¬j×óP2Þ^ S•ªØšÎ“æ"qåܼ‰s«Š6²_Ðó4®4Æqÿýmo³½|¥m½µóõå—[o+·óbÂéÞ{ÛÞösREZ$Îý0*+\d‰€ŽLÙ„Ó4«Ô\Ý|ŒöZA H¦ßµ‚~ÐAÎ×W_m½­œçdãFé7¿‘¾ñÒÛÓ¨ G}–†r[ܳ@%ÐÑ.d½HœWܪY«qÇ=%ç ô‘ôýï—. 6ý“µ¦*ÒiPÔ€þÉ¥£?TJºvÁÇ›oOz\©uNÿÎ;—wI[ܳ|­SAY" #S6¡+7¹YWÐË y/gÓö›f˜™2EºòJéïo½-ìñÿÛߤɓãÃf¼¦ûë~hP´€nªú¦õš »=Éóþ©O9_Ï:«¼ã„)gzÖ-î^Të@ÚXÅ™ªT‹»éiVÐ~8ÙþÞcø¿÷Û´Ijjj­¦Fmo+NÛošÏEÔãŽ9Øi‡œjjq7µœv‚î_ZsÐëê¤NÊ?NÔ9ŠØâNd‰ :2U©ÊRVtï1’†tÛ Î:Kêß_úéO“'H¥ç »+[Çù€Ä?¶4ÆSMÝ•Öce@w÷éÐ!þ¾qÏQÄ÷$ž~ZÚcéØcó (::*&Ë÷¬æ 'ÙoófiÝ:ó1ÂBÃ{ï9_W¬°ÛÞV¥¯ m è6«¸gµHœ÷¸Emqw¥Õâ´}Z‹ÄÕÕI ¥·¥)IÐ~òIé†òy­GyåiáB:ˆF@G¦*½Š{Ðq+yôÏÞYHkéÒ¶ç;žû»,ÂŽ$*YA÷^ÓÛ¦‚žöxL½¨ôJ·¸—ûZ««+­ a‘8÷unJ‘ZÜÝ¿Á ²h?èÈT¥ÚLMô¼úã;_ß¿í¹ãô´+蕘ƒî>¶sÐÝ}²šƒî=wQº«‰3µ¸û³y³ôÏJ3g¶]IÞöq+èÍÍR÷îÎdE›îïB@G¦*qé,É<ÝF.ºHêÓ'þjáå9Sà;^Öô"ÎA/·Å=L5µ¸›Æêã¦M¥—® “õô¨÷Ç“>ýié3Ÿq®ÇžT]ýsÕÜ\Úr^´ :Ø  #S• èaôÓN“.¾X2D«FÇ IDATúÉO¤ â·œ1ÙVÐÝíÒ¾nºweëJ0µÓGMAˆ»HœÍãVA/ÒbAüÇ;H»ï_/Óe“ˆjq_³ÆùÚÐ ­]ÿøn ­¯WAw_{Y.ç}=y?XÙaé?0ïC@¶èHlÕ*çÖÿùOÞ# ¯ õ«Îçï|Çùù¹çâ·œ1¥Ùâž´âYô zÒðw‘¸jjq÷¿nV®”–,±;N%Vq÷VÐ_}Õ¼­í"j¦s$iq÷ô¬gÅ éŠ+‚÷! t$víµÒÑGKcÇo“vØŒ:©‚îÞ¶ï¾Î×õëã·œ1¥¹H\’ñTzzœUÜ]•ª WS‹{9ÏEÖsÐýô­¶2·ÜPš4 {Ç6ï⇶@Ѐ-:ôï'{3붮†-U©€ÖBí¾‰v+~¶óx½ÇH"9èiVÐ+Åv‘8ï¶Reæ ½‚îôÇøñÒé§;—î²Ù×+­¿Á¨÷rzôJ¶¸Û   Ñ› ½ûÔ§¤yó¤}öIÿØiWƒ£ŽejqwoKÐMAÇLlÆT ô4Ÿ‹8«¸{÷‰S=¶ ;íu÷ot^ƒýûKûï}¯4Wqoù¿‡÷ﱜ×VÜ€^î” ÛóÄEØ" ×¸‘#¥Y³¤Õ«ãï7Pú·K3$™æ ûC»(6n´?®é¾U2 §ñ˜ÙÌA·Ÿ-ÛEâ¤ä*N‹»wÛ4[ÜßzKzôQçû¤ã—¶Û.Ù±¢º-ü÷·wïÖÛ¦M“î»OúÓŸÌûšÎô³íX‹XA÷wŠ©‚N@¶è5¬S'iï½€žÕ›Y¯JÏAwßH»oÚ;vt¾–[ASÑóÃ& ‡m_n=êÜipoï\ݸô4dYA_·N:è çZ÷îØW®”.¸ ¼ãz=[oí„c÷÷_ÿzÛm¢æ Û|¸Æ™µ°í’*ê"qt%æ ×0ï›ÆrZ]“nc ¤I…ÀrZÜÓ^$®Üë —;½ÄØ.'ÏAÏj‘¸´ú¤IN8?âçX;í$}ôQyÇ”ì*è=zD×J¬âžuÝÝ¿h‹Ä%ùÐŽ€lÐkX¹Á-­÷4ÞHÛTÐÓjq{mƒ§éñH£Å½ÒsÐm/³–F=lŸ,/³vãÎW÷5Õ¹³SU/—Íô"ô"¶¸y‘8:°E‹{ «De5,l–Ûfëe3=­w“+¤wß•öܳíùó^$Î;=ÈÛo'?¾Ÿ) ¿öšôóŸ;ßwï.y¦ó}Ðô¨q$]$Î}î˹ŸÞ žv@÷ cιªÌAû¦±Hœ{Î".Wéë s™56zK£‚žt›4[ÜM;÷6÷>º!µÜwÓmŸþ´4¾t÷Ým·KЃ¦”3=lÿsΉÜ ¦sM™â<þ[oí,HxØaåUÐã´¸§]Aß°¡õ{7¤vé~¹Á(¦gÂZÜß{¯ôõíïïß×tž ŸãŒ5¬‚îªT}ãFéÞ{Y$T?z ‹ª ¯_/u–tÆÒˆæýƒö :_9û‡1SU½¡!› zS“óõý÷Ûî›dzЛùr+èyÍA—¤CqZÃn ¹•¸zÚsн}Âçkž-î:Øô´.³§‚žTœ€~íµÒ믗î[É9èQè@ú,X ÕI.A {÷î8p`jÇ€¤è5,* /^,Ýq‡´h‘4sfòs˜¾—²Ÿƒnº­¡¡ü9è&¦ùíqç ÛlŸÕô4ôZ«®›6…WУÆis?L¯í4.³æ>Ç=${¬ó}-îþ z·nÎBtçŸ/zjÛ‘ }½ªiº»¿ÍsµbEéÏ• ÃTÐ|,X°@{íµWêÇ}ýõ× érG@¯aa­ÔÞÛÖ¯ÞßæQ¿Ks÷°Eâ$gz-î.ou5i‹{sskàO+PÅ™+[®°€î_I?éuÐ]aãÍjw÷9v×4Êoqw…}H1x°³ÆÁˆÎcÐ+1Ý& '=GÜ÷8Ašl¯‚A@ÒÕZ9ÿƒ¤A)qž¤“S­È@RôõðÃÎ×°àf[Å´ ¦íÒlqSAÏ¢ÅÝeª ç½H\¥+è¦`,µ è儺8‹Ä¥=Ý}Ž;uj½­sçì/³æ yþw¯·ßîIsºÍuÐm? ÛßfŒþÇ ha˜€de¤ayRE@¯Q“';_ÃZÜË sgŸí¼!uÏc:^¹-î×]ç¬.I/¿ì| [$NJ§Å½œ zX0Ìê2k•žƒîòÏ‹Žª Ç›Í¶a«¸§]Aïܹm›uQ¯3Û€þüóÒ¡‡JGm>ÿ5Vä÷¬+诽&í±Géó-î m\½Æ•ÐÃ~ÿÛ߶^+:jÿ¤a颋œyÀÏ?/=ýtÛß›ÚÞ㶸—SA7-—g½R‚ž×° zÜˬÅÑÜ,}ðtðÁÒ‹/šÇ‡ûû[ÜÓ^$.ª‚nšÂñÁÎ×… £oúÙV%‰s÷OУöÛ¸Q4Húö·í:‚ŽE@i# ×¸4z^-îÍÍÒ%—H/½ä\æL²« g1ݽÍTA·mݶپœ9è6a'`Ô‘W½¥ÅYèðÅ¥±c«¤QA÷·¸g=ݦ‚Õ•’Ö*îþó­‚u^÷q˜:UêÞÝn<&t6z³Y$®6íÚIÃ’÷M¯ûÝ´H\9sÐmw»«¯n{[9ô4ZÜ+=Ý=‡{é9Wœ9èqÏeâmqw·»à§’žv½sgiþüäÇ4±ƒî]”/ª+%­ÇÜßâvžrªôYµ¸Ç ËtP)ÌA¯qYµ¸ÛlSn=* ›ªxqç ›‚NÔxwÙŹD]Ü9èYµ¸{+èYð~àžO’~ó›Òí¼ÝŸ´Š{ÔýŒ»Hœ÷ÃÛ…Ç‚˜*èëÖ9W;ؼ9<¸F ¶¶½t÷üÞ÷ ã$]$Îû·›E=éxüLçyäçZõ‡Vº?(¾yóæê8j½Æ•SAOÚâ~ÇÒ¿þ%õécwž°cÛTÐÓ¾ÌZ”G•öÞ;ùô°í˃žv}Ó&©W/iäHéÂÏÑÐ~ô8â´¸{+Ëõõ¥J% M¦ úA9_׬qÂY\¶ÓÂZÜ7o¶« _|±ôàƒæsÄk‘[ÜMc²ý]’í½¿3¦ímt èK’N>ùäœÇôš·í¶Î׬ZÜMÇûæ7¯W^é|-§Åݧ‚žå"qRkp³ ]^6xd½Š{Üç|ÃiÕªÖÀv ›ë »A&×^PÝÜ“T»M—Ys¹Í6éýÝ$iqª ·´H¿úUéÏq¹ã¨æEâÒ è´¸íÅšO¾¦u]õ‡%MŽÜ Lè5ìùç¥vÎ??< ½±,b‹»—ÛŠ\‰Ë¬yosƒ‹MEÜËæ2kåTг˜ƒn:^ЇQsÐÝãÅ 26óŒ½t·Åݽ=I@7]f-­ð•¤ÅÝ[A·iq÷^þ®œw› zRåVУž¸ã»â óíq:€jÖuÕiq‹ÄÕ°=öHg‘¸¸-îþŸËyóê¯Àzß0»-×ÞUš³Z$Î+, ÇƒžFðq+è¶ÛÆ=¶í1:t(mq—Ìô´˜*èÞ÷¤¦ zÒ¹Ö~Þûo[A·mqwçÞ{_ÿå´¸wë½m9 Å è¦ZÒ|-M™b¾=, û?¡‚lÐk˜ia/¯4Ú¡mÂfÒ¶m)|zçÎNˆØyçÖÛÒ¸zÔýö?¦i^½šVq÷ëСµòšÖuÐ{,z›° zYVнL/Þ×{Ü÷œßÇyý«®ÎùÛzï=gUü,æ »ûÛË" ©Å=I@¿çžÒóЀ æ ×°rºkÕ*'75I={–þ.¬¶œ÷ €îÕÜÜöö¬æ {ùÓ°.ÓqÿùOç?ÛñDñ®âž¶rzؾQ÷Ó&\{?(ñ_fÍö&î‡Þ¶ê¬æ »59èõõm_ûåüí¹L÷=­@jЃÆ¤Ýô\D¡‚^ÃÒlq_¿Þ ê~a ReÝân èi´¸Gñ_[>îô°Û²® g=ÝýttÛóÇ©ª]f­œ z§N¥¯µ,—÷õ»v­ôòË¥—ÿ2‚ÞûiºoiUÐMa3‹ÎŒ"UÐË=Ø" ×°JÌA Õhq7UÐËmqº­Ü9èaÖ®•Þ~;z;¯JÏA[Å]2Wг/A—Yó€׆ ¥óϽç*—¿êêŽõ´ÓJÏ“¤‚¾zuÛç¦Üwÿ÷Þ㦱ðŸw®}˜¬‰£‚*…€^ÃÒjqÛ¾R-î¦7¿i´¸'©²¦Ðý·+õïïL'°•e8Hkºß4Õ¿‚î_# Éñü0‹€î}ýΟßö|qç ›nO£Å=è¸å<&iTÐóhqºÏt`‹€^ÃÒ® ›¶/Z‹{‹ÄÙÎA·Y1Þ j+W¶Þ¶hQë¼è ãq‘¸°®ƒ$çõ‹ª ' è¦s§µHœÿ¶ ÖmS=jwÓíi´¸‡UЋ´HÜ9ç$¯ê' èTÐ@ôУؼÙϪÅݵHœÿMqÜ9èi·¸ÇƒĽ³fIHgŸ~\Û°“F€w±Ë.¥·ÇYÅ=qz?(ñ~ Sn@÷;è礂*èþóUÐÝûôÞ{ñÏGÔô´>JÐMc˜:5øwIŽ/ÑâÒG@¯ai·¸›ØTÐÓœƒîÃåVÐmCœ÷>„-Wn‹»Ë½n ky÷VÐÓ4æÎ¥®]Koohhýš×ô4‰Ë¢‚n:¶mÝÝæÎ;KÿÞâž/é>A!º=µ¸›æ¸»ç ú™€’  WÈ•W^©ƒ>XÛl³†ªK/½T‹-ŠÜ¯±±Q#GŽTÏž=5lØ0Ý}÷Ý%¿¿í¶ÛT___ò_‡´!¬çùQ-¨i´¸ÛÌAO³Å݆˙ƒÞÒ"-Y|î Û²\$Îå>®î}i¹`¢7¤UÙüðCéÇ?nË7} `jqO:=Ï ºiÜY}ø4×Ý@wÛÍùzé¥ÒþÿIÆU©Eâ²ZÅÝäÃEmŽïgóz% ô ™={¶Î;ï<=û쳺ꪫ4sæL?>tŸeË–iĈ’¤x@C† ÑÉ'Ÿ¬3f”l×£G555mùoÙ²eêÔ©Sä˜*1=¬ZíVûÒ¬ {+ˆå^fíããË;¦JTÐmzZsÐçÌ‘FŒÆŒ‘~øCçÒ_~A×]÷^fÍÿ¤ý‚ÿ˜iVнÇú9©$-îýû·þ>Nwˆÿ|qö±iqÏsºiLQ¿ëÙS<ؼ}Ðý°½t`+äm=Òä Õƒ Rÿþýµß~ûiþüùÚ{ï½ûÜxãêÑ£‡fÍš%I:òÈ#µvíZ]sÍ5:ú裷lWWW§Þ½{ÇSÞ-în¸L# ›Væ.·Å=h;Û ºiŽ}¹×A÷Í} ƒZpÝs¦5ýù祧žŠ>†)œy+èîŠãAsУä]A÷‹ª°Ú3I‹»WÜû”Æô  z¹Ò^$Î+è˜o¼awü óÐâÊE='ÍŸ¼“Þ2iôÙgŸÕ!CJn2dˆž~úé’Û>úè# 0@C‡ÕÏþs½my¡ìJWЃzš-îQô8-îþí¾õ­èñxÇRÉ÷0¦ —”ŒA•Ë¨€TA:OÔ8L¼=í9èyWн×÷î“Çô ÛʃîÃæyJ£‚÷ø&tP.zš››5iÒ$=Zûî»oàvK—.5ôuëÖiå'×Ù4hn¿ýv=øàƒºà‚ 4cÆ ·_Òw›Ë¬ùÇfóÆ© GI# wè ýæ7ÒŠñƒy--¥/­¹Ña?§ÁvºW¥º¿‚ÔâžT¹-îi¼ÖÃŽu»Ø¢Å='NÔ›o¾©§¢z…-zè¡:ôÐC%Iûï¿¿ÆŽ«¾}ûê¾ûîÓ¸qãBöœ¤Nè¹%à]u•´aø’}â¾ÁMZAO£ÅýŠ+¤_ýJòÎ(wzT•×4ïv•¬ G7­9èþýà èG)½ø¢ôÏJÇ[zSUØ뤓¤ÝwOÞâ^M«¸ÛÎA÷*R‹{%‰3í&­ zø5ÓôÅ/N“$½úªsÅ…I“VÆ;1¨9ô ›8q¢~øaÍš5K;í´Sè¶ýúõÓܹsKnkllTçÎÕ³gOã>]ºtÑèÑ£uï½÷Fô)š>}˜V­rGúÞ÷¤O,Ý"6coÈ]½ZZµªõç4[Ü»w—N8AZ·®u›¼ç {æ4Òšƒî>vq*è6â>çAáÌhÜ€þƒH×^[º¿íøîºËùú¹ÏEokZ$®è«¸GÍA÷žÏ4þJt›×iЇ4qÎQn×Gš-îÉŒÓ8ÿç;ÒãKS¦ÌÑX‰“@¨ hõêÕ©¯{÷î8p`jÇj½‚&Nœ¨éÓ§kÖ¬YÚu×]#·>|¸n¼ñÆ’ÛuÄGî³nÝ:=öØc:餓"4_Ò[9áÎr‡ +ý]štÿ÷’9˜–3=nÝ KîîÂh6LjR” º)@z[ê½Ü€nz+µH\QWqª†‡ô¼*è¦÷,Ä è¦Ž*Ðý¯Û Ç‚wE²`Áíµ×^©÷õ×_'¤)  WÈ„ 4mÚ4=ðÀêÚµ«ššš$I={öTçÎ%I—\r‰–.]ªÛo¿]’töÙgëšk®Ñ¨Q£tÙe—é÷¿ÿ½þüç?롇ÚrÜË.»L‡v˜öØc½ú꫺ùæ›õñÇëœsαW¥æ _z©äN·äéæ›Óèîí«VIK—š+å´¸Ûò/wýõÎã$4}zX{³ÿ±7ô 9èqÆ~‚*è®… K§DÄ9_йÊå{öliúti—]ÌÛå9Ý=ØqÒ˜ƒîž§-îIÏã=@Q´VÎÿ iP Gœ'éäT+ò@-# WÈÔ©SUWW§‘#G–Ü~Ûm·é”SN‘$555iÉ’%[~·ãŽ;jöìÙš4i’¾ô¥/i÷Ýw×wÞ©1cÆlÙæÃ?ÔYg¥åË—kÿý÷×ñǯ«¯¾Úú“ÑJôc•:Èù¾gÏÒ€UõÛ´É S^Ûnë|5-V5v¬ôè£Ò×¾fnq_±"ü|ÞóJÒßþVZýÛâ¾q£´çžÒ›oæÐÃZ¥ã²­fV²‚ÞÜl¾Ìœ©‚^WçÌa—¤ ìÎe’u‹ûO~â|ýøcóy¼ÁÕÿX$9_9ûDÍAOÂû!O’ zÐm~;ï,í°Cüñ¹l+èQ¿€| ’4,r+•E@¯f‹wηÞzk›Û† ¢™3gîsíµ×êZwBo þ¹¿IßàFqƒ¤÷ÍiÃ'¯:Û ú%—HW_]zÛ™g¶=®ûýÿüOëñýÁÔ eï¿/õêe7ö^½¢ßćôúzé¸ã¤ë®‹?=jliWÐã>çA+Ï›ÂHX½œÕ܃¶5UÐë륭¶r¾/â"qþã·H«¸›ŽãݦœG²® ›>Ø1Ùi'g¡· ¯¿}>:°ÁeÖjTݦ‚žF@ÿ}i¿ýœ¶ßÙ³Vù÷Þk{\÷XÛoï|ÿÞ{mú”ž;Œ»Mƒïc,Û ºm¼Êê2kYÎAz,Ltÿãhs|Izê)éí·ƒ·‹ èþ º B\¦°'|=÷œ³ý¢Eæc{7 çµŠ{Ðë¢R‹ÄÅý÷+n ýÎ;›ow_K'Ÿ=>:°A@¯Q•è^nÕ*N‹ûvÛI#F8ÿm·]k%ÜÜ€¾|yÛ€'œ¹c÷KÛ7Ùþd›°QN=ì1L:}Ú4gŸ5k‚Ç´’¸û€Ôú|¸ë"†UÐMmÛ#FHžYm~·‚ž÷"q>ê|õVdý÷Û=^Q‰óÅ$@jóš]¸ÐYs"Î~q?@ˆzýìÝŸ€lÐkT-îa-·aô3»iSiHîØ18 K­pþüàëHÛ7wìXz»Íce èQói‹6ýÏv¾®ô\ºÙ& û+èS§:ã=ê(çç¸sÐ÷Ø£ôCÓùL‚*èq^¶çK¾Âö‰z¾’´¸Í7yã ɳe›}ÒZ$¤ÿü'ø6}¿ý¤ÿûKo³mq·}Þl_+ôRS¦LÑèÑ£µýöÛëØcmsUнƥYA7± èþ…°ülº;¦>}Z èiWÐmz]]z×AÏkz’€^WW:××fºWçÎñ§xÙÜl® §Ùâžö"|QUé {Â^_qâ¾û:ë&Ø+(øÛü­w\ÛË/zkóš ú÷ÃæùM»‚´]-ô|Pßÿþ÷uÊ)§è©§žÒ—¾ô%{î¹úç?ÿ™÷Ð(4‰«QEšƒµP“) ¯ZÕö¸î›ù®][o+' Ç™ƒî—u‹»;¶JÌAk'›ƒîŠÓnjq7…_›wïïMsÐójq·y ¸c´™ƒî=^X½¾¾í}KØ 6ÿVÄ ¤îß²éxI_³qZÜmŽõ!PÔvµЧM›¦cŽ9F'2AŸ}öÑÝwß­»ï¾[GqDΣ ¸¨ ×(÷ xÖ-îr·µ5jåé¸-îÞ7ÉAAªô¸‹Äe]A˦%Ú¦‚v\ÿEK—:+öðAé6a Ø´¸{+èÕÐâÕYdzÔ¶LÀØV˜ãž#ªë$H^-îôV'œp‚ž|òI=ûì³Z¿~½žxâ =ÿüóú⿘÷Ð(4*è5.N`õ‹SA÷Úc餓¤?”,h»B·é]º´þÕâV}MÐýsÐMüÇó¶ —3=jl•^ÅÝ/( {ƒ° wÛÛo—~ñ éôÓ[—´‚î}®Mô4[ÜË­ û«á6sЃè ÛÇf,6â´¸—£œ×¬m½œcùŸ'z«O¯Þ×DÒ©¦ zZ-îAç*wp ú€)ï zÔmåRÛ®“ ¶ç.·‚N‹{°;î¸C]t‘®¹æ=ýôÓºüòËuÆgè/ùKÞC Ð¨ ×(7tÚ̶eÚÞÐýã° èÞyêåôrç ÇyãwzœÀX‰Uܽû›¾—Êoq:WÔô¨1xÏ“öôr+è6lZÜM㚃î—V‹{Ø6åTé½3wÝå¬Ø~ÀÁûüîw¥c’¤‚nšÃt\ÓíµЯ¾újzê©:ÿüó%I‡rˆÞzë-M™2EÇ|à~“&MRÏž=Kn7nœÆ—éx L›6MÓ¦M+¹m¥÷r=(zòtÿË7ÞÞ?ü6ÕÔ×_/=_‡æ6x/½S'û9èi\f-ª‚4'9n@/§‚žÅuÐMUT@_·Núë_ÛVÀMÇõ~ïúz¹sн¬Ü9èÞcý&Np,âtÓ¸Òž:á}ZZœ¿÷“Nr¦Ç¼ñFð~§œRºŸíñ£­ã'à×Z@ohhPGßü ††5øÿAõ™2eŠ†-í(‹éÏ9sæèÀÌiD0! ×¨¨€>p`ô1l‚ª{-í°€îV8ƒ¶ kq7)­ ºiºí<^oµ­œ9èi´¸—;=, û¾ÿ~éå—¥Ã>^XÐq³°ç0jLþcº¯ ÷8iÌA:Wa¯¥8×A÷ŽÇûÁH—.¥— ˪ÅÝôºŠê¢ðïo:‡÷C%÷~™®™n{Lÿñm¶ón__߶C€ìì³ÏÖäÉ“µóÎ;ë°ÃÓ¬Y³tÇwèÖ[oÍ{hsÐkTT@/W]ôƒHÝ»Û#¬5wóæôæ ' èÞöú¥ï|Gzî¹ÖÛâTÐãÎAòûß·Ž'jß8«¸{“¤‚îÂ'ŸL6Ýû¡†i{Ó8l+èA犫Üw›b’VнGAë/øÏ»n]xk¶é¶¨ zœ@¶éoÚöXA÷iüx§un¯ææèËA†©Å€þÍo~S—^z©î¿ÿ~5J=ö˜®½öZ}éK_Ê{hô•dá3¿¨¶ã]w•zõ’V¯¯ Km«ä^q/³îâÎA¯¯ol6m’~üc顇‚÷M³ÅÝt»M}Þ<çrenõ¸-în@ìÔ)Y 1ÝŸ¤«ß{Çï­ »¿Ë«Å=lŸ¨ º7¸Gµ¸Ût÷ W_-]x¡ùxþ`jšƒžv@÷wÄ}®ÂÎ}ãvÛ™Æç<6û·g;wÖ„ 4a„¼‡@U¡‚^£ü Ÿ¥QAªú¿÷rßü‡UÐmº;›KMÙVЃ>4ð¶•E IDAT¶Øg9ÝäÈ#%w= €þÁÒ§>%}úÓÎãѳ§Ýóú¢Zܽ.¨ lz]lµ•óõ¶ÛÚ×,½c²]$Î{Þ¨Ð&­wɹÄ`]ôüó¥ÇŽ:^š]r®¨`´>DÔ߶M‡CØþîíYTзۮt{›î·ÅÝÏvñÅZ è zŠЃÞXÚÌAjÕuǶPœ) {/›å=—ÿͼÿ½íüã–éo‹^a>èX¦€žtºÿöÎ[ÇÐ׬qn»é&çZó_ùJ:=êwÞJµM‹»ëÈ#¥Ï|Æ|\÷qóŽ-êïyLt›¹ƒ¤ÙâþÊ+ÎWÿÜê¨ã½–¼÷É_ù›ƒôšþ.mZô%ûÇ$l\þ9èQ¢: $©ÿøcl/‹Ä½â¾à@aÐkT-î~YVÐÃÎVñµ­ /\è¬FÞ¥KøvAL‹ÄEµU]²Ì¯cÇÖÀÐÝÇsÀiÏ=“ÍA7˦Å=N@w¿ïØQú?ÿÇ|ÓóêÞï 1{÷»ì2iÆŒ¶çM³Å=éeì‚ærg1}ǯ{îÙö|A¡ÿïò½÷¤Ûo/}ìÂZÔm^wiVУŽ)Iûîk·WPÝVQú˜1c´ÿþûëŠ+®Ð›o¾™÷p€½Fy«ki·¸G-ðeGT@÷Ž·S§àãûßÌô¨7üÿû¿Î×Ç7ÿ>ª…7Íwÿíîýoii­pçržg›<¼??ñ„¹Å=è¸~ÞdšƒîÇË/Éž>fÏn[A¯t‹ûÿþ¯ôÇ?JMMÁûEÍA÷îgú{ « ï¹§´~}k;½w{Û úå—K«VIÿþ·yÜÞŸ£SŒÿxY´¸G]•Á´¿¿Ãfÿ±ŠЗ,Y¢ë¯¿^‹-Ò!‡¢áÇëúë¯×òåËóø½F¥1=ªÅÝ[AK»ÅÝ{ï÷Q-î¶ôµk¯={†ot¬´ºÿ¶í·o]Ú{}oS«¹Ôva´´ç {÷ÙÏ:­õQt/ÿ¼pÓq£>T û`ãC¤Þ½Š»ÿ\AÇûò—¥¾}Kÿ»å–Òã&iqŸ:UúêW¥ßþ6z?÷9>ã óyüîí/¾Øz›)PvêTÜÝÆlº÷Ãï¹Ë è&þ®“$t›®ÓŸ|Ò¹<¤Íôj èõõõúÌg>£o¼QË–-Ó~ô#ÝtÓMêׯŸ>÷¹Ïéá‡Î{ˆÔ<zÊ2 ›*è•lq«nÛô5kœ¯[o¾]бҚƒîuë­N‹ñ8?oÞÜ¢ì>žqÛrËmq_¿>øÒv&AÝÿ «¢F- ØÐÐ6 ‡µ¸ÿýïÒ!Ò9ç8ÿµ´”ߨûdݺ¶ûUÐûôqþ6N9%ø|¦¿3ïß‘ÍeÁ¢‚¯? {§øÏÄ;εkÛVí£^I*èaÇôÞ'Ó¿}G%x¢´hQëmAt[E è®W^yE?üáuî¹çªC‡ºòÊ+uúé§ëÇ?þ±N=õÔ¼‡@Mã2k5*‹UÜýlæ Û\Ý6 Kö-k×ðí‚xøíô  ºÿv7(x+èiµ¸ÿãmo‹³Š»w:‚Mðò èþq„Á{‡m+ Aô¥>rÚi§9·MŸÞö1HRA·i÷?Ÿ¶sÐm03…׸-îA.Atÿ¶ÞZºàéškÂç¿û»N’¬´‚îò.DØ^ZÜ/¿ürM›6M«W¯Ö׿þu=ðÀÚo¿ý¶üþøãWŸ>}t›{9…2oÞ¼B@6è5ªhô8-îÞÐÕâž4 »-îåTÐM‹ÄÅ}œM¥û˜Ùô¸-îÛoßö|q*èîµã½û›¸c±© ûç  ½õV𘼷»tÿx‚ºÜË×m»mé¶þÒè¦kû{›çÒ½t]˜¸ÝôAZÐò ±ßxci@RN@{\mºM‹»­¢ôE‹é׿þµFŒ¡:À:wî¬Y³få02áK’N>ùäœÇ è5*, _~¹Ý1¢zT‘’µ¸ŸpBë÷¦*}X ´½ÌÚš5ÎؼÕz¯°6zw,qæ Çi}·© 'mq7=_qVžß´É> û…tï¹vßÝ. »çq_c6twq@o@ïÐ!ü5åž'‰¨ zØ~QôÑ£¥ÿ÷ÿ¤eˤ / nmº©ÅÝ´X u»Sl^óqzÐcéçoqƒÔ~.³vÕUW©S§N%áü£>ÒƵí'/ú!C†ä5<>ù‡S4(…ã=,ir Çæ ×¨°€þßÿì˜aÁ%ªÅ=N½[7éK_27Í úÖ[—׺ìè›7Kù‹ù6íÛîXüÕ4†¤-îåÎA÷¯¸ï·é¶  º4ïój{‰@÷vS=hº[A÷.Ø¡CéH•hqwÖ¦Ð;w–Æs½óïëŠ;Ýôwš4xFôr+è¶-îAÛù_÷í¡‚~Ì1Çè7Þ(¹íõ×_ױǛӈÄ3HÒ°þÛ­ÒôÐãT£‚‘Íô$ô ±¹÷£¥¥5T%]Å}ÍšðùçQo¶M½[·à…»‚ÆsÍ5­U]Wœ9è6-îþJ¡_œ9è¶-î&¶sÐýaÔ"qAtÓcqÓMÎ×¼[ÜãÌA 6cŠª ûïwÜ9èAÇ^±¢õï+ª"I=N@âÝÍû×sÏž­ôxU„€þÚk¯éw•ÉOpÀúŸÿùŸœFü¨ ×¨ €¾zuÛmmÛ²“TÐmúæÍÁÁ6hºÛšTAzÃïVÐm˜î··…Ú}sþ_ÿï’ÓžìŸrPN‹{X@ßn»ð`lÚ?¬‚nÓânSA÷ÏAO£‚njqß°AúÝïœïûôi½=­÷4*èÞßEUÐÝÛvÜÑ ßþvÛmÓX$Ά½pïm®•+³­ ûÏçeнç|áû ºi»¢ôÏ~ö³zðÁKn›1c†FŽ™Óˆ€ôЗ/O~LÓtï9ÂÆV™Û¼98”ÍAïØѹ¬Uœ÷ œëTŸqFt=Éô°Ë¬EUô½c7ô ë û[ÜÃÎ4¾¸=ë9èqzPÝßâþÁ­‹ªýᥠ¬ù+èiµ¸›xšsÐÝÛ¶Úªm†+( »ÇOÚânzŒü¯#›û˜EÝ{ÂB·›ËÖmW”€~öÙgëÌ3ÏÔK/½¤Ã;LO?ý´~ÿûßkêÔ©y |‚ z è?ú‘ý1¢*èa@WTݽ=¨ÅÝ4¦æf©S'çç8-îË–9sÄÿû¿ÀæmsŽ+( Ç½T˜‰*®ƒž¤Å=I@OÚâ^‰9èŠ¾ÿ2x½zI“'›Çí¯ ›¶I«‚îÿ]ÐïÃ>ì‰3&÷ïË{¬+Z÷-÷2k^AÝæ5éÝ7É‚q^I*èAÛuϘ¶+B@=z´î¹ç­]»V?úÑ´nÝ:ÝsÏ=3fLÞCŸ ‚^£üÝå½öo›jp¹-în ðW¥‚*èQ-îa«¸»·mÜèTmzœ z9ݽ¯6—Y3µ¸gÐmZÜê»^AÕLïRv-î>j›i‘8Û±G3ª‚ôA€ÿ¹ŒþM-î®÷ß—vØ¡õç4‰‹SA÷·¸{ÿÍ— ›.àåÿ·&h;ÿcÓZÜ%iøðá>|xÞÃè5jÀ€Öïýa2©° zÒ÷ …Á¼çðo3=,¨ÖÕ9ãéÖÍ|Πý¼üÚ,*èýúïõ¸™Îí¯'™ƒžV=l.¼í"q¶-îA¯ï¸üA’ º_ÐÜm›cú[ôî‚Æÿwï_Ü,ˬ™¦aÄi[÷o»aCt@{ì’\f-h[ÓmEnq—¤¥K—êñÇײeË$I---ª««Ó÷¾÷½œG$zÍÚqÇÖï£ú{ïI¿ú•ó&yÌi¯½ÌÛ†#iÝÔªv\·2çÿ$Ý=¯mÅ,èþú‰KÐ;urB‰zØþYUУÎk;Ý$«÷  zÐë7ª‚tž(Q¯9›±tïÞz»÷ï-ieÛÿ!…Ÿ4‰ ª oØà,NéÞ7ÿþÍÍ¥ûþæ7’M– zl/.{› zµµ¸ÿò—¿ÔäÉ“uØa‡iGïÿ@aÐÐ_]:ïÿ8Ÿ¦¦Öïã,×Zܯ»î:Íœ9SC† É{( V-î{î)õíÞJk:FÔ›R7ü0^=I‹{}½¹ý7i‹»ÿ8Q·¥}™5÷<¶sÐÃ*èëׇ?èuþê«N@ôQéˆ#Úî4ÓX¾ñÒ}“t¯¨Ë¬½ùfé϶-îaçr·÷îóÚkæã™z¹‹Äù·³ý›´ý÷ ¨ÅÝvÿ,õìÙSÜ4@!Ðaзۮm «þ¹Çª Û®ânSA—¤·ßvZò·ß>xL¦Û¥ÖûæÞÏ°7ÔÞóÆ èIZÜ»t)ý9«÷Ì‹ÆyÎã^f-Éô¸‹ÄÙÎAú0Èt™5¿ ûºjUë±[Zœ×åŠÑc;¦û;ï¾»îêÜöÖ[Ñû‡UÐMSF¼Šp™µ ý¢ÆWÒ÷"ÏA?~¼Î<óLvÚi:à€J~wÈ!‡ä4*àE@‡u@7]WÜ+ìQ-îi-÷Þ{Î÷;íd‹) oØ %½óNë6qæ ›¤¹H\çÎmítô¢·¸{…toH‹»H\sГ.çônÝœýü×¥SA÷ŸÏ;þnÝœj½©£$LT5Úôüš˜KwŒþ@ÜßÛü[§‚nÛâ´OÐl¶õ*r‹ûO~òÕÕÕéòË/oó»7ýí tH² èË–•þ>ªr§‚~ë­Ò‹/:a´KçkçÎÒ.»8¿·ƒîŠs™µ>žy¦ô8qæ } ‘Ö"qn½-îåô¬®ƒî½-éuГÎA7-gÐW¯v¾º=jìþ&lzÐ6QL×A÷ èA—it•[Aw÷SA÷*'L»ÚãeÖÞò¶X€B" Ãº‚ÞÔL“´ëöè!í,5w®S¥ÿøcçëÆ­—³]ÅÝg÷ ùèYÌAOºŠ»—?t›‚u«¸»UX{wÒw“°EâÂ*è¶-îþslç —Óâît· "ªú§‚þÎ;ÒgH3fHÛnk?¦8-î6Ý»˜¸-îîxÖ­“–. “wÛ¸cŠSAoÝõÌ3Ïè駟օ^¨%K–¨¥¥E»¸Ÿ†€\`ÙäÍ& ¾È™w_S=HCƒôÈ#ÒüùÒ¢EÒòåN[ð† ÎÊñîŠÍq+èn¨ Z$Î{?LÕÑ4.³–V@÷LÓé ÊvP@ol”þþwiöìÒ0o_XtX‹»÷¼~qæ ûWq3ÝÛâû_¯Y´¸»ÝÝ/ªÅÝæ˜îïV¬n¾Ù ¯IÃ_Ô"qAÝû¡‘éƒÓc¶Š»Ÿ·‚¾j•ô•¯”þ.JœÇ#êží¶³?–Tì€þòË/«oß¾ºøâ‹uÙe—IrªêãÇÏydÀE@‡U@?㌶¦‚Õâ¦C‡Ö@gwï÷Þk+{o« Û´¸G1ÍAkK j¦Š¹ÿ÷¶-î--Òš5ÒAIŸÿ¼4r¤ôÄmÇ4¶´[ܽlç »Ó‚Æä?©‚gw=è<&ît7FUÿm+è6Å© íãýýÆNw‹TúDµ¸'™ƒî×ô\Tª‚îžg«­¢·õ²Y/#/?ûÙÏtà 7hÖ¬YêðÉ@=ôP½ð 9 ¸è ´´´hÉ’%Z¼xñ–ÿª™M@÷o¶mœ9èa¼×ŸŽ[AïÕËùzá…¥ûùç3û¿w·I³‚îþœ´‚î_ä̦‚Ôâ.9!kóféÊ+Ÿ×¬ >ŽT¼UÜã,'™‰ó¯àýjú$iý•WZ÷ñÏAŸ7/øC÷˜6!;ˆÍ6Q×A÷þûä=´²$ è~Aó½ýßÛêÐ!8 Gwë­ñŒiÞ>ªÅ=jzTHp 7h”Ûâî*·Åݽ¿}ú´Ý6,›Æiª §±Š»‘8ï¸ÓZ$.(ŒšB¡¿=ÛTÙ5yÿýÖ}üôñã£?ä ’U@ú½TÚ®ïoq÷óv¸n¼±ô÷6-îþǵ¡!øo/É¿76‹Ä…UÐM÷¡È—Y;ýôÓ5eÊmþäÝÒÒ¢©S§ê›ßüfÎ#.z S¦LQ=ô¯ýK=zôÐâÅ‹uÚi§é¦›nÊ{heI«‚n:F¹ô ë ûïÿ>j{›÷°7ï7JÓ¦I/½|½!$ º+NÝ?çÙtÝß¹ûú+Å®¸—YK»‚w‘¸ Ë¬Ù®â^N‹»÷õfjqÏ£‚njqºïÞñyCyœi&6-î¦ñ44w¿$mqzý¸÷Ûö>9 Ÿyæ™z饗´Ç{hÍš5Ú}÷Ý5gÎvÚiy |‚UÜcøãÿ¨ÿûßÚn»íÔÐРm·ÝV?ùÉO4jÔ(tÒIy/1Û€î‚QíÎaÆ÷®íuÐ]AáÚt™µ P6ögž‘žzʹÞúœ9æq¥ЃZÜM!ÃTAwùW⪠ûe§ÍuГT3½Çõ¯âw‘¸¨Ë¬ùcª 'mqwk÷µ"mç ÝÞØèLëxüñô+è^þˬٴ¸ûoSA7ô4[Üøû»Óezô°Û¯È½W¯^z衇´jÕ*=õÔS>|¸¶‹» Èô:wOÞµzè¡zöÙgõþûïkýúõ9¬&ÓùãŽ/É"qþßÛ´ñÛ|Xb‡í¿7aÛ¹ç9ýtçþ³Ý1‹<ÝÕ£GsÌ1„s ˆ€Ã駟®Y³fI’N=õT}ík_ÓðáÃuî¹çæ<²òÄ èqÂZsÐ]A‹Ä/N@Oë2kƒK3g:ß_t‘4v¬´paë1Ã.S§‚îoÜwÛ z’9èq[Üý÷ÃtwÜI‰KczØc`ÚÇUAºz¸~íÕºt¿£ZÜÓ¨ ûÇ*µ­ ‡}Àaóa¢Í: ·¶Y¶È½K—.Æÿºvíš÷ÐÀ'hqaòäÉ[¾;v¬ÆŒ£Õ«Wk§vÊqTå+òtÓ÷þ1™¾· èÍÍÒ„ Ò®»¶ÝÆö2kÞûzÅÒ‘G:ߟx¢ó߬YÎjÝn50(PĹ̚ú\a‹Ä™îƒ±5oð5²¤×AO«‚îÿ  éuÐmç ›æøG-gÐMc ú9éô8kI[Ü£*èQ¿ z<žyFzï=iñââTÐmŽ5~¼tÛmÒºuÅèóæÍ+ù¹±±Q¿þõ¯Y$€! —¡[·nêÖ­[ÞÃ(›÷cœ º_ô¸ÆÐ?ü°teiï6¶tÿJíAçKsº? ]Ýp‚*èA-§ÅÝ{^¿4+èQ¯YSÝôZ d¦ÅóÒšƒõz(' ÛìЃ>‹êhIRAï=éˆ#‚C·d?=êwÞã'éj1môút·+B@0`@›Ÿ÷Ûo?}ùË_®êuThOhqGjtÓíå¼)Ͳ‚î Ö&QsнۥПxÂùzÎ9¥·÷铬‚¾ysÛýüsУZÜmzV«¸û‰óÞ§‚.™“÷¾FÍA7]fÍ/i@ª )' {…ô UýýÇŽšƒnb èo¼áÜæ^µ²Ü ºMÀ§‚î=_‘+è&MMMÚè^ÏäŽ :¬ºí*îÞ SN‹{Ü ºM@÷®â>l+èQsimúºuÎ×_þÒ9ÞÔ©ÒH?,yfVHrž5ªu¾»Û‚½`ô›ß8çøë_ƒÇo;ݽ=NݶÅÝ$íˬ¹c8Ðù~ï½Ki»Š{š-îqç gQA;¿—ÿ’faç« µöûoŸ=Ûùº×^Òk¯™zØô$;6sÐm…]9¡(ýè£.ùyùòåš?¾~õ«_å4"àG@G¬ zÜE™Òšƒns™µ }MÛÛô¸sУBST‹{}}éýÜ}w©_¿¶-î]»JO>)m¿½ôþû­Áú믗vÜQzçsˆ4Uн•ærº·ó ­ º¿ÅÝ_Ajÿò—ÍÓ/‚ªÇ6-î~Iúã‡Ó6 ›Âi¹-îþ€·Z6~S@¿øb©wïÒ˙ũ ›ÎÄíŠ{F=çAÛ9 õ«_-ù¹oß¾úô§?Í"q‘ý–[ÚngÚ6¬úW” ºm@ϪÅ=¬;(XE}8áwÓ&çºìS¦8+È›‚gT‹»÷1± è¦Ç0‹UܽçŠZAÜ{{Ø8A¯ß¬*è¦íÜ¿-ÿ¢*èa¯ÍJ´¸›ÎçU}ê<™9SºêªÖãú¦LǺz˜Žç4¬ÆÔâöaDØ”Š¢ôSO=5ï!€tDôÓNk»]œc{¿Æ‘õ*îaoî“´¸GRï¹M-ÂA‹«™V!÷þìŽ×wÞ¥Køxm[Ü“ÌA—Ìsßý¢*èþwï*q[ÜMâ¬ânªÚÚt7ØUÐýòžƒî=¿í|oÿïÜŸÃÎô÷·Ï>¥ÿqÇI/¼ |°yLI/³æï±Ý?èß@ÿãfú}úi§¦ºOÒÒÒRò½«®®N·ø?9Ã"qˆÕâ·‚^ΛҰ ºwLInO³ÅÝ& ºtÉ|^ïùüçÝv[çëV[™íب€înÔân3ÝÏtßÓ® —»H\Ðùã¬âu½ð¨ùþQ¯¿,ç ‡µ¸{Çh;=¨‚ž$ {ÿNÜý½¯û° ºÿ„rzP=J‘[Ü{÷î­ûî»OK—.ÕŽ;K—êþûïWŸ>}´Ï>ûhï½÷ÖÞÞÅ@ÅQAGêÝ»}Vô,[Ü%û ºéØA¿óMS8¨*ûÝïJÇKÿõ_æcûzÐtR÷ùKs‘¸r[ÜãTÐÝl[AÛâг\$Î4fï1“ô8¯Û¨9èaØëÅ¥?nýpÈ{_ƒ*ÎAÏ›üÞdÂæÅÛvø$脨mŠ\Aÿ׿þ¥ßýîw;vì–Ûî»ï>ÝtÓMš1cFŽ#.*èdÐÃ9‹:FÑZÜ››Ó™ƒ.…W;m+èÞ7ðþãôè!sŒÔ©“ùØq*èÞ ëÎåö¶Ž§1=UÜM«sÇ]$.l ¦÷ }²¼ºŸíô"UÐ_yÅùú“ŸogZ­Þÿ¸»Ï¿ÿtïÏþöî<<ªòüÿø' »A‚²£‚VP$ ²¸²X-T«ÕÅâ¾¢ ÚV­Šöûõ¢Öjµ­~]pÑþ¨¬ PQ!®¨©"F«e•%ç÷Çx&gNÎ:sf¿®+W23ÏœóÌÉ™dîsßÏóü+vË*ªÝéþ\Π¿óÎ;ªªªJ¸oÀ€zë­·²Ô#`G€Ž&ô>}¤}÷õng¶õÛn*ô|XfÍÚ·m™ßƒèöçúmÛœ}Ý/ƒn²gÐÝJÜÍŸoºI:ê(çµã½Jܽ^G˜ ºyÞEYâžê2kv~ºù¼0ïƒ0CœÞaÆ ›üt§m›¯ñÛo·cmc"à ›ïw{€ný}»Mdhï»õoÕâÅÒË/'îËzî»eóìƒîô:¬r%@Ÿ]?ýéOuñÅg»kàŒAG“Ý+(𠆼Ơ'ȹent¯àÛ|<%îÖmZ³÷NÛ›A7E9ÝÚ—†ég?“N:)P™ËRY%[âžÌôd2èaÆ Í ;µqÚõbHsÄÚÎl›©ÝiÿA3èæï;L€þï7í‹=@w*q·^X±÷×í÷é {qÊ Ÿ|rìûµ×º?ÏzðT];zË–-5eÊM™2%Û].Бt€$ƒîÖ6ˆTÆ oÛæÝ>%î~t·ñ¸æ}öà'h‰{Ð ºÉmwkÑe IDAT efKKƒS€îUâ¦DÛú<ûv“Í ûèöì±[%×rcö>ZÙËâÃiA†ø=–j½¬Lª¯—ž}V5ªñ¢ÕÎκÛL÷Ö öýZ™x˜ º[¹»_=È9”Ìß,³ö}çJ€.Io¿ý¶æΫåË—kÎœ9zå•W´~ýzuÔQÙî X]]6lØÙöV¯^Ù¶ × #p€nÿ$ˆ1"6Ôi\¦ŸT–YûéO½Û{èRnŒAʼ `èöåج킔¸[Ï…ÒÒÄÀ=ª ºµO¦dÇ '3Iœ=kݧS‰ûŽÒÊ•Ò€Á«ìãσ|neÚNý÷{,™Ýºó÷xôÑMMÐ ºÙΞA÷:o’Í ÛÍ »ýnÂþ¾LÖ¿[Ö Š\П|òI]zé¥:å”S´hÑ"IRyy¹&NœH€Ž´©««SïÞ½³Ý È舼ÄÝÚæšk¤ .VRj—J}·Ý¼Ûçâô°t“=ƒî÷û3ƒ%3ð1ƒ=@75ƒng¾–7Þ<0q;Öý:õÅéu˜ì%îÖ}E5Iœý|u Ð+*b (=ü°óö¼ts_ÉŒAwc›ˆÛ±/©è^ïY§ +^%îA2èn%îÖ~xUñ8]ÔñÛ—×ñ3IœÓ ë6r%@¿é¦›ôØciàÀúË_þ"IÚo¿ýôÞ{ïe¹g(d™ó’úD´Õy’®‰h[[Б¶w3¸ëÔ)¹~™ÅÝÞ?“Û>­åÒ¹ºÌZXö]’V­’ZµrÞ}=q· ºYE$ƒn-«bt{¹r:Æ ¿þºs6Öþœ“O–ú÷]€Ø°!|Ý\Ì­?#GJ 6½ßëøáÞ6Ìyd#oåtœ~·öóhófé»ï¼Ç {íϯÄ=ì:èNû²_œJW‰{®eзmÛ¦Ýwß=á¾Õ«W7Yz H>’ª#Ú%î ³¸#ò=*ÉÎâ~ÑERûöÞÛÌT‰»)H€n~’Í KRe¥´ÿþMû ¹Og´Ì>ƒî4¡_Øm· *Ù ºWZ´Ö­“–,ñÏ —”H}û6.g4@·®7ï ïµWÓ×âwNÉn§#ƒ¤ÄýÇ?–žx"|Ý>ÝZâ>b„tÜqMÿy•»{eЃŒAOv™5§ Mö ÙrÑEéÒK/ÕÛo¿-Iz÷Ýwuíµ×êÌ3ÏÌrÏ€‰‘•¸›¼2Êa„ƒnþ앱2=l‰{Ø º[¶.Ê ºët§þ¤’AwÊ„§’A·…¶þ¾‚TSxÝ/I7Þû¾qcÓöÉœÿ^t¿Ýú; :Ýë÷f¹>¯IâzôH¼í÷ÚÍ |Ù²¦ÏKf ºuû Æ‚~¯eÖœª.Üöä"O˜ ŽnçñÎÒŸÿ,}ô‘sEK¦7N‡z¨Î9ç}ûí·:í´Ót衇êdszzu”¸#T€$c¯ ºS0eþdýí zÐ º}ÛN÷Y'ÌŠjw“ÀúèRc°d P¼JÜícнt³ü8Š1èævÿøGi—]ï³fôím¸Cs"=ëEû… §m…É Ût·vNûó{/9ÇSÍ $=òˆ÷>œ¶emcÞÝÚJ±*Éûƒ×ß ·‹^Ùzó÷é$Ìß3û%ë¾×®ýü›ßHkÖßfÔ¶oß®iÓ¦é׿þµ.¿ürmܸQåååÙëpDMJ‹½Æ톃ž ¯uÐM~I·ö^ÌÓ5Ý+@ÏT]j:™\*cÐíó HÑÎâ~Þy‰ýt Г °Ü&6óê·58L&@wûÛg÷ÚfQèNÏ÷:Fnk¦KþôóÏ—î¿?ö³W€îµÌZ˜÷믗~ùK逜Ÿk}ž×qtzžõü¼ðBiûöØp“Ü·“ Í›7×}÷ݧ­[·JÁ99Š‘•¸ÛK5@Ï— zTcÐÃfЭ¯7L‰»™A· s;Öþ™)¬ë\»eЭÏrtkw±?×+@÷ºØd¶ zaÉšAÂ)@wâTâno¦š#™Ý‰Ó{.ÙÝ/ƒ>xpã¼~ïß Ë¬ùeÐ÷ÜSzðÁ¦³¹[EQâþþûÒ—_&·Ìd:\xá…š:uª¾ùæ›lw¸ ÄIèQ”†zIv zÐ ºß$qA/0xeڜƠûM–YZ¾s§ÿäaN¼ßt¯~ÛKÜ­Û‰j z²t¯>$›A÷Ê„›¶o—þñé‡D¥o=ȹ}È!îû³·M%@·Ë°¨¼ÞO~t§‹qnÇʾ¾¼ÓÏ~ý²Wy$›Awbÿû³eKrËL¦Ã}÷ݧ?üP·ß~»öÞ{o5û¡c%%%zçw²Ü;  C©eÐÏ:+V†6ñšW‰»[=l‰»×ëp "›5‹]ì¹ë®pt·ã¤Ä=ô/¾hü¹¢"6É×óüfq5Ê}_ný Ê|ý_}%Ù¨NçûöÒ1ÇHW^é¾-k?̾D™Awƒnß·W›¨gqwÛ·$mÚ”ý úþóIÒðáÃ%ÅÖ>6lX{Ü0t³”Úd¶õ hRáUâžìt³_‰»}Û^¼2è¦ ¥âa3çÖmK± È뵟tRl&é3Ï”n¾9qßRãñ¶ž a–Y³èA*ì÷Ýy§tàΙúdJÜÝxeÐÝúm‡~º5[lfüÖ¸§r‘+Ìydíë9ç4î·Gçã\Z*ÝwŸÔ«—÷¶ìû7;î¸Ø÷tgЭ¥côÿý_©[·Ä~¹Ù¸1ût?È #¥wk€žot¿=l=ªwûsýXt¯×Þ£‡4mZâ>­ý±÷/È2kÉ–¸›ÜŽ™õ˜Ú3èaJÜÃdнúdÞäܶf‹ÍÙò½²óɬƒîåG?Š}ßgÿ¶æë=Zª¯o¼Æ é½÷š¶wê“[ lmk“ßüFš3'º zª%î^ÇØëwýÛßƾ^yÅ¿í† RÛ¶îgBCCƒV¬X!I2 C;wîŒß6 <8]6èH©ÄÝ*ê1èÖÀÅ-( ;Ý|ŽW€î—Iµ :Ý>æÛ¾Ï(2èa2uAK܃fÐËËc%ôæºåaÇ [sËh:eÐoº)|?Nt¿ç˜ü2èö²¿ zYÜÃ3Æÿùwß[jÌìkëÖÒ¶mÏkÞ<üyèõ~2ß#æúã^Á|²ë »m/Ù÷d—Y³Ë… zÇŽõ‹_ü"~{·ÝvK¸-I}ôQäû]¿~½~õ«_éùçŸW}}½zõê¥{ï½WÕÕÕ‘ï €BA€IÉè™Ê »IGÝžQöc/íuê_ z±ÛN¬“¸%ø­ƒt úÂ…ÒÚµÒÁ'¶ :Ýú˜[ݺ~»é㥿ÿ½i€t’8·¾9Ýï7N]ò.q÷êKÐ~Dáœsb¥êk×Æn·l›}Þ©"Âä÷¾ö*q73ènËš-q:ݺd'‰ó«°±rú;dÚ°!ûcÐ?þøãŒïsÓ¦M4hzôè¡ßýîw:ôÐCõᇪ¢¢"ã} Ÿ #²wS &ˆ Áf*t¿€#hÿƒdÚ”¸ݯ}ÌýöíÑèAÆ [×½z%ŽMN%ƒîÄ\Ýúû½é&é‰'‚O¼gJ6ƒä¼±–¸g*ƒžÌ{Íz!¤E‹XÝÊé}´rÀlk¶÷Ë -q*ƒd z²‡ I¼ ³¸gí·ÞªÍ›7kÁ‚ñûzôè‘Ř$¡t¯Y¯£Êœ›¼m§‹öÉμ¶e‰{Ø1èAƒÉ¨Ç ;±þî­¸ùÝ>Ý.ÈkwÛ¯×óœÆb»A÷îu.›Û 󜰓ÄYÇ ›ÏûéOÛG==(ëû9Š z1èf&9ÙIâ‚ŽA·>ÖÐÐ4ƒí ŸtRbÖ¸+}óMâ}ÙΠgÃO<¡#Ž8B§Ÿ~ºöÞ{o~úéZºti¶»@Î#@GàÝéñµ­=û—é zÐÀÚ L2==³¸'“©ûaiä”Æ ;I%ƒî9eÐMa3 fv×é5y•œ§Zâ^R"=ð@b{§ßYԻܘûiÙ2q ºS€nÞï¶ ¯sÁžA·_ìI6ƒîµÌšµ_æ~MöIâ̶?.Íœ™x_%%ÒªUÒ[oÅn›ó0Ø÷WL>øà=ñÄ2 C<ð€ºvíª#ø $iØ°azçwôè£ê°Ãs}ÞäÉ“›ŒS¯©©QMMMZû Å`Ö¬Yš5kVÂ}ëׯÏRoà¦?6À.ÙÝ>Iœõ~³}*ÂfÐ9DêÓGê×Ïÿ9^VºJܽʪýʦ½¤šA·o'Ù1èv\[¾Ëi¹/¿×ê4Iœ™A·>×|<™1Äf€žÌô0Ãì™wûs­çY²cГa­6nÑ"vþxõ3ì$qæv¬… Ѓn}Ì)ƒîµý0ƒ­(Æ ú{졪ªª„û?üpMŸ>Ýóy·Þz+³¼@š8]ð\¹r¥˜¥Á :" Ѓ:A…ƒþãKï¼ã¿Ý ô(të>ƒÎâ6ƒžêt“Ûtë,ñaJÜKK×ävÛo˜ º[‰»W€eÝž½ Âi ºÓv¤;ƒnºí¶Øw¯²s“WŸìÇÒz;™1è©fЭz‹îý¶n'l€nÝŸ×íbÌ zè¡ú׿þ•pßÒ¥KuôÑGg©GäÆ #T€dY£¨²;J­ZI?THºö)¬¨ô¨KÜíÏõ“® zª%î^ÂŒA÷Ê {lbŸ¿ý6¶Æ÷üùþûI&ƒdtû>ìö €d3è3fH#Fx÷9ˆK.‰ö z2“ÄYÛXßéÊ {A·¶qË Ûƒq§J—(.”c}òäɪ¯¯×…^¨W_}U7ÜpƒžyævÚiÙî9)gЭ·ííSѱ£´q£ôË_6},•2útA÷+q:‹{Ø×d~øO&@·ƒ¨KܽøeЬ_ß4ƒÞ§OâyùøãÒyçIǼÄ=èþS)q—Ü/¸½‡üöqê©ÒQG뇟fÍ/À8èn}²^бå©fÐ$;‹»_‰{ªt?ŘA?à€´`Á½ñÆ:úè£õî»ïê駟ÖA”í®ÓŠðcì‚è^³¸{ê©2á[Xæë0û¹ï¾RïÞÒ¼yMÛt ºù!}ܸØøßîÝ·‘j=™uЭ¿« t;¯Š /~t§@õÞ{cqyM÷ý÷±ïæ’aQgЃœÛ(½ñFã>ì}Tt¯¾†UVæ|,¬‚dЭÍ™ç†a4ž;at§ýƒ4ƒnnÃéBS²Ë¬9)Æ º$vØaÌÚ@HdÐá /Y+vjçµ ë}é–lݺÌÚŠãp¥äKÜý‚Î^½¤«®’>ù$öe•LÝlU‰»ßt¯eÖ’ý]»]ÔpÚnU•ô8_È0ûeþ.‚döÝ2è~ˬùþýkãÏNcнtûþüDõ3g3Àv{¯{íoçΦºÙÞ<7íåã¦0ºõþ CnÂŽAwÚ¿ßC1fÐ@rÐÑ$ûmÞ>üpéœsÛ)qÏÄ ÔQ–¸Ûƒ¦°Ag˜1èæðKûs¬“ÄÙŸëÇ|Þ¶m™ƒž‰wóv§NRÏžÞ“ÄYkÐ÷d2è~c¬ûð ìÝJܳ‘A—ì°ïc{Ý>½Y³ÔtûE• “ăn½?ì2k^Š5ƒÂ#@G¨1ènkç[Ý ;Í nÞDÐw)qFt«TÆ ·oû¾m›´Ûnážke¾Þ;îˆmgݺXP$ƒVÐ×ìº=×)@÷ÛOºÖAw ÐÍÇ‚”¸ûõ=L› œ2èöc“l‰»¹]kݾ-§Ýíwd ºý‚[Ð vØ zЋdÐ@Pè2mÚ4 4HíÚµSUU•®»î:}b¯svP[[«aÆ©¢¢BÕÕÕzôÑG›´Y°`† ¢víÚièСZ¸pa¨¾… ÐÝ>;eÓ)Ê º}fð°%îQèöçú9RúàéÝw¥ / ö'æë}í5©¼\ºé&éôÓÓ;‹{˜ ÔìGY™tæ™Ò76ÝF&2è~%îÖÇÌß·WÝé< ›AO5P7ûùÒKÛ zñš©vË ïØáAwºã–Az’8§1èQ—¸“AA gÈ’%KtñÅkùòåúýï¯E‹éüóÏ÷|Î矮Ã?\’4wî\UVVjüøñšo®!%iÕªU:öØcÕµkW=õÔSêܹ³Ž9æ½ùæ›û4@÷›$Î^ž›Î ú$¿{&Ô [÷ ¼JÜ­û”Ò“A/)‘öÙGÚo¿à}vbömÛ6i¯½¤Ë/Md—ÎYÜÃ÷Þ+ýú×·Í~¸v’Ì:èa+ü–Ys»8äÕ } ËÌðþö·±ŸÛ·.ƒ¾j•të­ÞcÐÛ¶müÙ/ƒžÌ2kÉŒAr™52è (>6dˆ5¨îÓ§öÜsOõë×Oï½÷žöÛo?ÇçÜu×]Úu×]µxñbIÒG¡Í›7ëæ›oÖ1Ç#Iºí¶ÛtÐAé‰'ž›5wèСºí¶Û4}úô@}K6ƒnÞgÊDæÜtÇÒÅ'÷\óBƒ_=ìôT2è©ÌâUfËî0ýáÒÊ•‰ÏIGÝ^nÏ ;µw*q '“A÷[«Ûº{æ=H‰{¦Ç [ÈÏ?—:tH}’8sÝúåËc·;vtÐÿõ/©ÿÆÛéÈ wèàÞÎÚ6]%îdÐ@PèYÒðÃ'Ôé×åË—«²²2á¾ÊÊJýîw¿Khsä‘G&´©ªªŠõADUâžÉ z«V±Y½“a/q·Ia3èÙƒžŽÝZlöÙœKÉA÷â—Aw’Ê$qN÷»í?l‰{IârawkiÎiöwûÑG±á&s’Aë>œôþýƒ—¸Û3èA'‰+/—>XZ¶Lºá†ÄvÖí$›A÷kCù¦®®N6lˆd[«W¯Žd;P,øØ šï¢‹Û†ôøã‰ÏKæwí… :òI]]z÷îín@Ñ"@Ï‚‰'ê£>Ò‹/¾˜µ>Lžû6£È [³gòíû[»VºôÒÄÇ¢ Ðgýð%=úhìÂÅúõëSÝ(v™ó’úD°Åy’®‰`;PÐ3lâĉš7ož/^¬®]»z¶íÞ½»V­Z•p_mm­Zµj®ÝÚì±ÇžÛ¾õÖ[U]]-IúÉObYMMl\wª%î¹Î^âî´ÌÚ×_7–úúñƒnݧ”žYܳ™AOu™5¿ ºõvØ z.Ìâî´Ìš½m%î©žn%ØÍ›Ïêûãæ>‚èA2èAKÜÃNž¶ÄÝYÍ_Ò„ Ò•WJ+W®ÔÀ“Ý a}$UG°JÜ fqÏ ‰'jÎœ9Z¸p¡zôèáÛ~èСª­­M¸¯¶¶V‡z¨o›C9$p¿‚–¸çÒ,î©pƒnïkÐà\Jßô\ ÐŽ)ݳ¸»ýloï4‹{@:™ zA2èÙ*qï=©®.ö³[•„½Ê ÈïÍ ÄwîLlŸÉ ºý˜½’®wÆ € Ð3ä /ÔÌ™3õÈ#¨M›6ª¯¯W}}½¶nÝoså•Wj„ ñÛçwž6nܨáÇkñâÅ:ûì³5{öl]vÙeñ6“&MÒk¯½¦O߯ÄݯQ¼¿ÊËSß(èrçwê»ï¾Ó°aÃÔ­[·ø×ßþö·x›úúz}úé§ñÛ]ºtÑ’%K$IÇwœV®\©™3gjôèÑñ6•••š7ožÖ®]«1cƨ¾¾^óçÏW¿~ý÷-ÙÝþ:lö/[΄G=I\¶3èö1èÖÇì29‹{º&‰ ºkûdJܪ ÜJÜ3=IœWÝï}m]nºµ]²%îÉdÐÔ¸[/EµÌÚŠÒYgÛ?…wÒ ’¹ÿþû›ÜWYY©E‹y>ïÈ#ÔË/¿œl×>ô¦2‹{ÐLd¶]wtÚiÙC)üU«0A[˜1è™æVâ.¹âé^Ý©^í3µzPÉ–¸{õ#H_ÃrË 'Sânž;öa"ÖÝú{J%ƒžÌ$q^ܲüAŸçdРàÛ ƒŽ„àÚëì׸L{>×µnx;Š÷BƒnMtëØn«tŽAz<Ü2èA'‰ Ú·dJÜ­eÖÜ2èö‹wAúÒ«—tË-Ò=÷$>n=W¼.D„É {-³¤ÄÝk z˜‹Oùð7ät.q÷›$Îi»¹ªM›ÄÛ™ƒnfL£Ì G KMô!CœŸ“‰YÜMA.Y/$dkt¿1èö>•YÜ“iëÄ+ƒ4«o½u饱eÚœª‚è©dÐS-qú»ˆ:Rƒîô<>Øf+ƒ.Ř(2èa3í~¼ôåËcë[Ûebtë¸`7nô zоÍ Û÷a}ß8½¶TÞCé.q“AÊí÷kŠb ºõ13@Óÿ°ô\¾ ò cÐ:@ß°A²‹w*qÏå­-Z$ÞÎÔtÉ9@Ï…YÜ­eÎöÝm?©þ®ƒ^ÔÂAZâþúëþ}2ï¢ÄÝ^b­YÜ­RYf-è9àu‘ÂÊ<÷**œû4ƒžê,îùp¡t„ Ф /”f̈}€Î×´~ABO=å¾0ô\™Å]  çBÝzî%3IÜš5±ÛÒúõÉÍÁàÖ¿w{€nÊ• z˜Ì³__ÜÎûíAƒ¤ÆŒqÞGÐ1èÖ6æ…?aÿžåÓß<Ð:ƒ¾n]Ó¶ncMóE²}m×NzõUÿíZ·¿yslŒî·ßJ×^Ûx\³A÷ Фc º½Ûmëý©L·m[ìvyy,@÷»@•l‰»SÜJÜÍÇü¤;ƒ¦Ä=Ê ú„ ÎÛ(-•¾ù&6×@³f©—¸»õ%Š÷_”vÛÍÿ¹VŒAG“Á/r Êò9›”L‰{Û¶ÒÿÛ¸\[˜rm)¶ÜÛ† ±ŸsiwÉ9£e‰{”tk?’É ÿ}c{¯¾™A^!Îâå$qn÷ н´n» 2n\â6í¾üRúì³ØúãÙ*q?ôP©OŸÔ¶Š:gÐÍëf c €ìÛ0ïËÉôÕ,K÷ ¨íeëvf@鼑Ž,j¦JÜý‚:«tÍânfÐýÎÛdJÜÃènÛñÛWªÁd1èÉìcÅŠÆŸ£ÐÏ=WÚsÏXð-¹_,üôÓØ÷GM}ô|ú;ò:B—¸[KG£Ì8eSº>„Ož, .uéâü¸S€ž­€ %î^Ù¿§k’83ƒ4@Sâî4ݾM§,sØ zªï=3ƒn?Æö÷u*UQèååÒ1Ç4^ˆq{Ý۷Ǿ7oÎ2k ¿0¡ô†çŒZ1ŽA·€wÚÆСÒóÏ»oà (‚ˆl–¸çÂ,îaîOe™5{‰»W_ƒ¼^· ºÓóÜJÜýöaoUÝ~z­IîÖ7QèRìb‚_€nVE8èÖýyA'@Ù@€Ž¤3è;w:æÖö¹ì¥—bNI©-³fJæõF‘AÏf‰{:Ç ‡É '3}ËéÁc㚥`cËÝ&{sê¿Ùgk´Ä=Óô6m¤Nbc¶­¬k’»í#“t)vn† ЧàÕÉ IDATwìÿþ¦Äd :"/qÏ—ÌÓÁ7þœ® ºŸTÆ G]¿Ï>ÒWH7JGí¾?«tÌâ6ƒ.%èf†Õï<üüóØ÷-[bßýJ¡“)q2ÝzŸßšãN}’¢)qÿâ ÷í V³‘Aw;ÿÌ÷fÍbÁzÐåâÂfÐóåoÈè û•¸{Ý—«²Õ×\Ê 7k&M›î9¹’A7Yô t« t?ÖmØÇ Û•–æÆ,în¬%în} Ú‡(t3÷Ë 7knwkS]f L‡Àº}wk‰{>fЭò1ƒžÊ~“ÙN¦2èn‚ÎânNxæwÚt¿€Þ Ú¢^fÍm˜H&3è~Û÷ V3]âf º5@²s»ŒAÙB€IÁû‡u{À‘ocЭ‚öõ¹ç¤K.ImV¹”A ·oŸÞYÜí·ƒŽA7ÇÏûÜ[·&Þ²,[:JÜ­¯+ì’féЃŒA·÷ÅÍ AÎíÒ]ânèNý·ï›d:š”àúèÖ‰¸ìÛòóƒmÐ aäHéC¢Ûo.Íâ4@<8ö;îÖ-\°tŸn*ÂŒA7ï Ó·L”¸ûè¹–A·Ïâžlý•Wbòùµ "Ì$qæLýAKÜ­:“Ä€l` :TR"­Y#ýñ·ÝÚIÎY9eÒ õƒ­[P”jÝ "r5ƒîÔÞ/ëœÌ¾RÍ ›³Ñ‡ Zƒ”¸É »•¸Û3o[ï˵ ºÓ{Þ¯­“ƒ Ö.ˆ0cÐÍ!9Agq7·Kd thà@iófé׿ŽÝ“Awƒ^èìÇ'• ^%À¶ÄÝšeLv_a2èA&‰{ûíÆ zØ‹Ö‹$nûIw‰{.gУÚGº_Ý ´’ЃdЋåoÈtè’K¤?l¼íöÔ:Õ‰½Ä½P3èö²_S2¯×m<˜íek z¦2è&¿ úÿ»Ðôý÷±ûÃö-hÝu»îûžî1èébÏ'ó{K¶› cÐ/º¨ñq¯w{_’Í êß:yè$µhÑø3to©Nre•Ê$qQgÚÃèRòcЃdЃ´5ï7ŒÆIßN:©±oa¤£ÄýW¿J|žS[§wûvüö•‰÷T'‰ ÛÎM úõ×Ç*‚ÂfЙ$d:$5Î|-ùèæ‡X§Ç¤ÂÏ Ûƒ¢¨KÜíû ÚŸTÝΦMíÓ™AZâ.%ÀÖ1è^û9þøÄÛÖeÝú¦¤¿yóØ…¯dJÜse zT“ÄY¥z¾Z'‰óú™Ä… Ð÷Ø£±AªÚ´‰}ßm·`ÛðC€IÁf9vš0Ê)0/t~Ç' §YÜÃn'Óô=o§ú;“%öƒnž—A'‰ûÿ/v±Á0¤Cý.ºuóîG˜ º5p [ânÝŽ—Lè©N—L;7æÒi ‰“Ú™º}n§ýï¿¿ôÁ+4˜Ïõ3p ôØcÒå—‡NÐÑDª%î…žAOu ú›oJO=û9Ÿ2èæãf¶0•û(3èæ¹gÐ7mJ¬ ±kÖ¬1úüóÒ×_Ko¼áݯ°%înÏëׯñ1§‹\¹’ASâv›É2'ܱ#Ö/¯Ý|ù]ÜiÖLÚgŸÄûƒ.³vòÉÞç@èˆ3?d† ЭE1iZ6-Y¬ýu…½ ѯŸÔ«WìçTÆ »õ'YAt{¦?ÝcÐMAtk»‹/ÖŸæÍc¬s18í?™ ÕúÞ0_Û^{5n32èAÎñLfÐ¥X€¾s§ûyQZÚX ï÷»³ùŒAÙB€Ž8óƒ¯_€ä–A7åëÛÃÖ.Šw3 ȧ º½]*û “A÷ Ý2èJ]»&ßG·ý¤³Ä=Ÿ3èÙ ÐÝ2è%%‰“¾yíŸä tÄùèf†ý‹/ïó Ìó-ƒ”[PU€ž«t·öéΠ;eÆí¬“Ä%;–ßO”%îÖïÖöÉfÐÓ%È2kÙ˜$NŠ-‰æ7ÝšA÷Ú¯};vHÿúW4ýƒq~%î£G»?×i z¡Šâ»S€n@3øíψ¥{ ºÉ/@·fy¥Æc›®=H;§ç¹-³ÖªUãí\Ë »Í·àÕ— íöÞ[ºõÖð} šA·Awz }ûn¶ã5Y @: #Î/ƒ¾Ë.ÒÏžxŸ5Hc’¸àÛ°èÖåÊ2Yjî´=7éÈP{eÐíÁgØIâ‚.­TÐ z˜ç•”ÄÕþSš0!÷Æ ›ÛÿÕ¯¤—_Žv›½zI“&…~ª“ÄYëÿûþ‡?$>לÍÝÞ ÝÐç KÒ~û%ÞŽ:HÌ…:ݽÌ9•ß½9IZ÷îMsË ‡ƒžŽÝ‹u‰ûO~[¾Î| ¹’A?àiäHiÁéµ×¼÷“© KÉNçdÛ¶Øwû\æ>2qAôK/•Î=·ñ¶ÓôBÏ Ûƒ¢d^o1è™Ð3Yâ~ãÒæÍÒ/~áß¿ ºÙ&ScЃ´szžßE{v.dлu“ž{Núíoc·ÛµkÚ&Ù1è©èaÇ ;1tûìýÖmêß1›Ðç7]’vß=VŠkjÙ²ñ9…>öÜ”© zªýI×v¢¨š()‘Z·Ö°ôlOgr \½.n˜zësƒô)Lûd]r‰ôæ›Òœ9þ}ñ“êùcþ 2‹»_‰ûöí±ïö :ÈtÄÉ Û7?ØîØ!Ý|sìçBÏ §c z±eÐÃô#W&‰ ºº[–<ôt++“úõ“**RïKTô'žÞxÃ;ƒn0À¹ ›A¿Ém›éFžqAtë_3@ïÛWZ³&=ýʤšÎèlg?>Q•¸';f:Š ­o_ÿ6QŽA÷bßn=—&‰s ­:µ--ÍÝ z”R=_ºvUî\yeìöðáÒÇ7mgþþGŽ”~ô#çc$ƒ^¨@n"@G\wûãæÛ1c¤… c?çs}Ý:ÿ6nAQ*ºÓ,î™Ì ¿ý¶›leÐóuwûñrš¡Ý^âžkcÐÃz˜}ݺ5¹ýôè!mØÐøÞiÑÂ9ãmþþíYûé6w-| A\*%îee‰H¾²—º:éÕËùþ|ƒî'Ý“°™Ü2è^ûËõIâ$ç ºõ"B¾fн.È]w]Ó™ú÷ØCzÿýÆ zÉhÞÜÿ½jÐú¤Ä=/4€üE€Ž¸TJÜ­kçs={v-•wk@–îÌtªr9ƒ.åÎ$qÉ–¸b}êÔ¦÷=ýtÓŒu:ûc¾×¼JÜí:t-L‡¸TJÜÍ ®ÂùÈ/P ¢¤$q–éd2è™äÝJú35Ým?™,q²ºÓó$ÿa ùšA7=&AªT¢`þþíÁ¶S‰»½ žIâ@¶ #®C‡Ø÷V­¼Û¹•¸KMƒôBdü’­(+sž$.cЃpË gz¿vnz®•¸ûÍâžJM¹A·2t;&‰ÙB€Ž¸¿üEš;76ë±·w)qôBý`EÝÜN!ŒAÏT‰{Ð1èéГ-q²ÌZ1dÐ3%H€n–¸Û‘AÙB€Ž¸îÝ¥±cýKP­Ä[¶Œ}·Ž©Î…`!¢ ЭôlÏâD¶–YK6ƒu‰»Ó0/N³¸{;³„ÞzÛO.è¹:çDýê«¥Ö­›ÞOd :Bsƒ.ÅÎ\ýÀ•¨?{‰»WvÕK¦Žs±ïG™¸ß¨gï›Aw›$.]ˬyõÅé1¯ãdÏ ›Û'ƒž:³?^çÁÅK›77½Ÿ :ÈæªEh^%î…°Ôš·ï©ŽAO6#©À¨[7iÇŽÆ×_^û¾~}´ûI5ƒž¯³¸[Kܶãµ/ô¦ìçA¯^Ò«¯{.³¸€l!ƒŽÐ¼&‰³–¸çÚö¨DUâÞ¢…tãÒ!‡H|>ƒ^Q‘Ü~Sa½8qï½éÙG2ô\š$ÎéyR°÷|Πç{€~Ï=ÒÒ¥ÁžË:è [К_‰{¡‹j÷Î;OêÝ;¤}t¸íüõ¯Ò¯~%uîn¿QÙ}÷ôl7—'‰ ²ÌZªô†飂÷)WäR_¤¦CvÙ%ö> ‚ z´~ÿûß«´´T—^zi¶»@ÎãcBs*q'ƒþõ{lìËßö>8öUh’)q—27=ŠYÜÖv7·ûíÒ²eÞç†Ó¶³%Wßï© u`’¸è¼òÊ+ºûî»u઄ƒ €/2è zz·ÏgØD…Xân^X±žKfý¿ÿݾÿþàÛÎ…÷\;oSYe€I⢱qãF?^Ó§OWûöí³Ýò:Bó›$.W3jQq+qZ¡??¹ºÌš_ý¿ÿ•¾úʽŸßßøXÛ¶Û´o¿¡AÚsO©S§`}’ÐD•AGò.ºè"3F#GŽ”‘ ')y€!Íú×ü k-q/téDr-ÐÉ´d'‰K÷të:èNÛ6„_xÿ A±rõíÛ¥#b÷™ú¶m‰Û—bU(A/.äB€~ã±¹rM*:“Ä¥îÑGUmm­^yåI¢¼€€Ðš}b+ë÷bX=SŠõø¹è^Ak&—Y kï½¥§žJ¼Ï Ð7nl¼Ïú Ú÷\Ð8Bzùåìíß ôìùôÓO5iÒ$-X°@-~(³2 ƒ,:ð1¡9JNôb 0£þÌY,ÇÍùúÍìr¾OçÄ)@7Ÿ—LM¥rAOÍk¯½¦ÿüç?ª®®Žß·sçN½ð ºýöÛõý÷ß»fÔ'Ož¬ s ÉÔÔÔ¨¦¦&­}€b0kÖ,Íš5+á¾õë×g©7pC€ŽÐœ>W9eБšb ÌÀ|ëÖØ÷\›$Îïü oÚÔxŸùÚ±#¿JÜsU*“Ä‘AOÍüc½õÖ[ñÛ†aèŒ3ÎPŸ>}ô›ßüƳÜýÖ[oMìÑqºà¹råJ 80K=‚>† 4¯½3èQK%°(ßû~×]±ï¹ ;-–ŒòòØ÷¨2èèME5á•——«oß¾ ÷µiÓF»í¶[“ûŽÕ«WG¶­¶mÛªW¯^‘mÈ'èͫĽ3èù^â¾ï¾™ÝŸs©>³Ê5ÈœN“Äår‰{×®±ïÇÓxŸ5ƒžOcÐsë ç–’’&Š ÖIÒøñã#Ýêûï¿OŽ¢D€ŽÐü2èL—?¾üRjÝ:Û½H´cGì»$…Í §{ zv~Ú·•ð›ËZŸG=dÐsËóÏ?Ÿí.Hs¼Ö I}"ØÞjIãµaƶät„æô·˜–YK·L^ØèØ1sû Ê Ð›7}Ov’¸tŽAbÛ-[&ÞÎ×eÖrtÈ´>’˜CHUÄ9&§àeÖ³Ä=Ù z¶ÖA·ö'Ìâ­¨fqÈ$t„t’8$§Øƒ®dtûôl­ƒžì~™Å=Zæ±I&@ÏÅÊPКW‰;t¤Ê ™9=“cУ*q·³V¡0I\êR)qßu×ÆŸù;2‰1èͫĽ3èù>‹{®±O7a‚ôä“ÒðáÎíícÐÓµL]IIlb·íÛÓ[âN=¼yóšÞ—®¹Ò‰ :Bó›$®Ø2è“'g»…ÅÌ ›“Äuë&½ôRbVÓÊ>Ýu½¼\ª«“Ö®•vÙŽ]ª%îLÞ1Ç$.Y'E Ëß1Кß$qÅæW¿Š6@*ö€À^âîÇ-@ú8^u•´dIìkÊïþ$ƒeÖ¢•®¡éD‰;BcôÌ(Öãg/qÂ:Iœ)êÀl—]¤Ãv›VÖIâÂŽA‡;ŽÈ'äšß$qHM±©dÐKKƒM*—NdÐsCT ‹ýý2‹¡ùMG©03èæt?ÖIâ¬K¡å[€Îôheû<H:Bó›$ΫüûqK5ƒž¯z´È €|D€ŽÐ¼Æ [×A’‘l€n‰%îÙš,ŠeÖX=uùz¡7t„F‰{zûq  KÎcÐó-@O&ƒw;ƾp@vû³¸#4&‰C:¥2=ŸKÜ­“ĵhî9hªé“O¤=÷Lm;cIèeÖÒ«Ø[*cЙ$V{í•í„C!%Bs È #*a×Aw›$.ßJÜ­cÐ ÐsG±_0™E¡ùeнÚÁ_±·T'‰3KãwÝ5=ýKó=Ä$qÅ‹¡ùèLp…T˜t³*#k}Ð éž{¤ñãÓÓ??©fÐ)qÏ-Å~Á d:Bó+q7?ÐòÁ69Å~ÜÌ zØ,² —”HgŸž¾…éOX©ŒA@a ׉Ђ–¸É;}’¸|E€tDšAÏö,Úù®Ø[²º9=Û¢˜$.lõÒ‡c 2)>΢AGT’ ˆ¬cгw$‹1舄P\}ãÏÉ)öãöÜsÒ‚ÁÛÛ—YËW©L‡ôá€L"@G$š5“n¸AúðÃØíššü[æ ¹¡oßØWPöIâòU1eÐKK³;‘@®"@Gd®º*Û=( ùte‹™Aë­ðã×sI1AχßSYY~ô–<Î7@c€^V&mܘíÞ$¯˜2èù€¡: ÐC@ž9ýàƒ³Ý“ä™á7ß çse €L"@׬“«åsPÕ»wãÏèÙGd:cÂ1KÜó=@oÝZ:òÈØÏèÙ—ÏçÈ_èHÚ~ûe»@áèRl5©ð'‰Ëù~.€üÄ,îHÊcIýúe»Ù3sfúfx&è ¯PôæÍcßÉ g%î È gÐ’%K4vìXuïÞ]¥¥¥š3gŽïsæΫSN9E:tÐàÁƒ5wîÜ„Çxà•––&|•••iÛ¶méz’¤“O·Vu¡9åé´Ó²Ý H…™A'@Ͼ ¿€(ñ$ƒ6oÞ¬ªª*Ý~ûí’¤ŸO×/½ô’ƧÁƒkÑ¢E7nœjjj´pá„v»îº«êëëã_Ÿþ¹Z´h‘¶×ô"è §P&‰“ÂgÐÛ¶]¯©I_ŸŠt ”¸gÐèÑ£5zôèÀíï¸ãp š(Ä zР°ukiÓ¦ôõ§˜åû¹òôöí·ßj—]vI¸¯¼¼\K—.M¸oãÆêÙ³§ªªªtÓM7iíÚµ™ì&"FÆ.3@ß±#ÿƒª°%îRìõsÎD/ßÏ%ŸÐsØ™gž©Ç{Lÿüç?µiÓ&Í;W³gÏÖ–-[´~ýzIRŸ>}ôàƒêÉ'ŸÔ”)S4þ| 0@ßÿ}–{dl…W(ô°%îHJÜ@6PâžÃÆŽ«)S¦èŠ+®ÐêÕ«µÏ>ûhÒ¤Iºá†Ôì‡TÛ!C4dÈIRÿþýõóŸÿ\ݺuÓã?®©“'OVEEEÂ}555žÏr‘u z³<ÿ‹–Lé‘êÅžY³fiÖ¬Y ÷™VÜäùÇÙÂÖ¬Y3]{íµºöÚkµ~ýzUTTèÏþ³:uê¤òòrÇç´nÝZ£FÒßÿþwÏ`ûÖ[oUuuuººŽäSÆî¹ç¤O?Ín i º™Aϧs P¥šAwºà¹råJ 80Åž€BFž&OTTTÈ0 =ôÐC:á„\ÛmݺU ,P·nÝ2Ø;«‘#¥ ²Û‡B ÐÉ çŽ|¯Æù‰´iÓ&ÕÖÖª¶¶V’ôᇪ¶¶VŸþ‚¼òÊ+5ÁíÔÕÕiÆŒª««Óƒ>¨~ýúiݺuš:uj¼Íõ×_¯gŸ}V|ðæÌ™£“O>Y[¶lÑ\Ùd:¢vùåÒ~ûI»ïžíž€bBŽ ƒ^yå•øòh%%%š2eŠ$éôÓO×}÷ݧúúúx°.I;wîÔ-·Ü¢÷Þ{O»ï¾»?üpÝrË-êܹs¼Í·ß~«sÏ=W_|ñ…ú÷ï¯ã?^üãÕ»wï̾8 K )ƒÎ$q¹ãì³c_™D€žAÇWCCƒëã÷ßÂíý÷ß_+W®ôÜæ-·Ü¢[n¹%’þùÈ:I\¾èdЊÅDaÁRÝ ÐùÝ'ty­tJÜŠci»ì’Ý~ä“B Ð)q(nŒArÌÈ‘ÒCIÇ—ížä‡BƒN ¸ 9¦U+é´Ó²Ý‹üQH:cЊ:€¼Vˆ:t@±[½zudÛjÛ¶­zõêÙö€t"@×Ì}ÇŽüÐÍw2è€âµF’4~üøH·úþûï¤#/ (… ›ô††ìö€ìÙôÃ÷’úD°½Õ’ÆkÆ l H?ty­3èèô‘TíNÇHGy­Æ ïºkâw2èòZ!eÐ=VZ¾\0 Û=@6 (ù —•IC†d»ÈJÜä5ëŒçù  ¸ Èkè(èò: :€¼F€€BA€ ` ¹aÚ´i4hÚµk§ªª*]wÝuúä“O²Ý-r³¸ÈkdÐܳdÉ]|ñÅ4hÖ¬Y£iÓ¦éå—_Öüùó³Ý5r:€¼Ö¼yãÏ-Zd¯Yñ>}úhÏ=÷T¿~ýôÞ{ïi¿ýöËbÏÈmèòÚ1ÇHÿ÷±LúœíÞpÒÐÐ IÚ¹sg–{R˜êêê´aÆH¶µzõêH¶H:€¼V^.~¶{ÀMCCƒ&Ož¬Q£F©oß¾ÙîNÁ©««SïÞ½³Ý @DÐ@ÚLœ8Q}ô‘^|ñE߶“'OVEEEÂ}555ª©©IW÷ò^cæ|†¤>lqž¤k"Ø€\3kÖ,Íš5+á¾õë×g©7pC€Òbâĉš7ož/^¬®]»ú¶¿õÖ[U]]ž¢>’¢8v”¸…Êé‚çÊ•+5pàÀ,õNÐ@ä&Nœ¨9sæhñâÅêÑ£G¶»@^ @‘ºð 5kÖ,Í;WmÚ´Q}}½$©¢¢B­ZµÊrïÈ]¥Ùî(,wÞy§¾ûî; 6Lݺu‹ýíoËv×ÈidÐ@¤ÌeÕ@8dÐÈèätr:9€€@€@ @  ÐÈè£Fe»È7ͲÝ(46H-[f»È7è±òòl÷ùˆwr:9€€@€@ @  ÐÈèätr:9€€@€@ @  ÐÈèätr:9€€@€@ @ 4ËvŠI]]6lØɶV¯^Év¹ CêêêÔ»wïlw£Ð2¤1s>CRŸ¶8OÒ5l Ð2®¤ê¶C‰;&‰  ÐÈèätr:9€eÖ<ÔÕÕYÖ/OÍêÕ,‹pG€ࢮ®N½{÷Îv7E‚ÀEcæ|†¤>lqž¤k"Ø  øê#©:‚íPâdCÔÃKÚ¶m«^½zEºM@"@P°ÖH’Æù–ßÿ}‚tDŽ@ÚôÃ÷¨†©H±J˜ñ‘M X (pQ SÒ‹BŠr\;cÚa"@€ÀÒ3®1íÐ „¨Çµ3¦Ð 4Ƶ#z¥Ùî @ 'PâYƬðиY³f©¦¦&ÛÝÈIgw *³$ñ^rö´ÿë„sÆ]®f…G#JÜ3hÉ’%;v¬ºwï®ÒÒRÍ™3Ç÷9sçÎÕ)§œ¢:hðàÁš;wn“6 ,Ð!CÔ®]; :T .LG÷ Þ¬Y³²Ý…œÅ±qÆqqDZ$Íœ9S•••ª¨¨Ðˆ#ôÆod»Kyˆ÷’»g²ÝÅ9ã.WuVø×"øš!IÌ Ÿ§Ð3hóæͪªªÒí·ß.I*))ñlÿÒK/iܸq}zô/€n¾ùf|òɺ÷Þ{%I/¼ð‚ºvíª»îºKW_}u–{@n"ƒžÃ¾ýö[í²Ë. ÷•——kéÒ¥ñÛË—/WUUUB›ªª*½ôÒKé#NV¬XÑäÿSee%ÿŸð@=‡yæ™:ãŒ34vìX1BÏ=÷œfÏž­-[¶hýúõª¨¨ÐºuëTYY™ð¼è‘GqÜæÖ­[%E;Kd¡X¿~½V®\™ínä$Ž3Ž‹;ŽMSæßÝ-[¶d¹'é÷õ×_këÖ­ŽÿŸæϟߤ}.ÿoZµjÕ?ýCRý[r{k%ÍŒp{~r}{Öm~.ïcv{¹úšÃnsÆ}{~Ç&™mæâkþHR°¿©Åô¿)oÈŠ’’cÎœ9žm¶oßn\ýõÆ`”––ûqõÕW%%%Ɔ  Ã0Œ¶mÛwß}wÂóþïÿþÏØu×]·9cÆ C_|ñÅ_Yúš1cF4ÿHrØW_}e”””Ï>ûlÂý¿ùÍoŒ<°I{þ7ñÅ_|e÷«þ7å 2è9¬Y³fºöÚkuíµ×Æ3æþóŸÕ±cG•——K’ºwïn¹ºS[[«=öØÃq›G}´f̘¡ž={ªuëÖi €˜­[·ê£>ÒÑGí®¤Ýî»ï®V­Z©¶¶VGuTü~·ÿOüo€ì(¦ÿMù‚=OTTTÈ0 =ôÐC:ñÄã÷:Tµµµ mkkkuÈ!‡8n§C‡:õÔSÓÚW€3·¿Í…hÈ!M. ¿þúëš8qb“¶üo€ì)¦ÿMù€Iâ2hÓ¦Mª­­Ô~ø¡jkkõ駟J’®¼òJM˜0!Þ¾®®N3fÌP]]|ðAõë×OëÖ­ÓÔ©Sãm&Mš¤×^{M'žx¢^xáýìg?Smm­&Mš”Ù€Åe—]¦¿ýío:÷Üsµxñb 6L[·nչ瞛í®³Ð3è•W^Quuµª««URR¢)S¦¨ºº:p×××ǃuIÚ¹s§n¹åUVVêšk®Qee¥V­Z¥Î;ÇÛTVVjÞ¼yZ»v­ÆŒ£úúzÍŸ?_ýúõËøëÀ4fÌ=ôÐCZ±b…Ž;î8•––jÉ’%êرc¶»@Î"@Ï áÇ«¡¡A Ú¹sgüçûî»O’tÿý÷káÂ…ñöûï¿¿V®\©M›6iÍš5š9sfBpn:òÈ#õòË/ëÛo¿Õ²eË4bÄÇýÏœ9S•••ª¨¨Ðˆ#ôÆo¤ç…æˆë®»N¥¥¥ _ݺuKhó§?ýI}ûöÕî»ï®Ÿüä' H$©¡¡AW]u•~ô£©sçΪ©©Ñúõë3ù2R¶dÉ;VÝ»wWii©æ̙ӤMÇaË–-:ï¼ó´çž{jÏ=÷Ô\ ï¿ÿ>­¯-~ÇåôÓOorþØKÀ ñ¸HÒ´iÓ4hÐ µk×NUUUºîºëôÉ'Ÿ4iWlçMãRÌçÔô}µË.»¨¶¶Vëׯ×óÏ?¯þýû7yÎܹsuÊ)§¨C‡oÒu\Šõœ©¯¯×)§œ¢ýöÛOeeeºôÒKÛÛ9ä¸Â9“w²=K2cîܹFYY™qæ™g‹-2;ì0£]»vÆ—_~™í®¥ÍÔ©Sþýû_|ñEü뫯¾Š?þ—¿üÅhÑ¢…ñë_ÿÚxúé§~ýú=zô0¶mÛosÙe—åååÆ7ÞhÌž=ÛèÑ£‡1tèÐl¼œ¤ÍŸ?߸æškŒü㎫DuN<ñD£cÇŽÆôéÓ{î¹ÇèСƒ1nܸŒ¼Ædø—ÓO?Ý8öØcΟo¾ù&¡M!Ã0ŒÑ£G>ø ñÎ;ïO?ý´1lØ0côèÑ mŠñ¼ r\Šù¼1 ÿ÷•ÝÒ¥KÖ­[úÓŸŒ7ß|Ó¸ùæ›6mÚÏ=÷\¼ÍÊ•+æÍ›Çwœñ /?ýéO-Zo¼ñFº_N¤Òqlî¿ÿ~£]»v çÛ_|‘î—© ï+»uëÖåååÆGa,^¼Ø8ãŒ3Œ²²2cÞ¼yñ6ù~Þ¤ë¸ë9óñÇ“&M2~øa£ªªÊ¸ôÒK›´)Æs&Èq)„s&ß ‰aÆ555 ÷uéÒÅøŸÿùŸ,õ(ý¦NjTVV:>ÖÐÐ`ì½÷ÞÆ•W^¿oÓ¦MFYY™ñðÆaß}÷±ë®»wÝuW¼ÍêÕ«’’ã…^HoçÓÄþ¡0ªãðÁeeeÆ3Ï<o3þ|£¬¬ÌøøãÓý²Ræôay„ ÆñÇïúœb8.¦·ß~Û())1Þ}÷]Ã08oLöãbœ7VA‚ÐSO=Õ8õÔSî;餓ŒQ£FÅoO˜0Á8øàƒÚ 2Ä8묳¢ël†Eulî¿ÿ~£¢¢"-}̧÷•ÝÔ©Snݺ%Ü÷‹_üÂ8òÈ#ã· í¼‰ê¸ë9c5|øpÇ@´Ï+·ãRˆçL®£Ä½H¬X±BUUU ÷UVV꥗^ÊR2£®®NÝ»wס‡ª;î¸C_}õ•$é‹/¾ÐÇœpLÚ´i£^½zÅÉÛo¿­ 6$´éÝ»·Ú´i£eË–eö…¤ITÇaÅŠjhhHhSYY©††­X±"C¯&Z%%%Z´h‘ºt颣>Z=ô6nܼ˜ŽKCCƒ¤Ø¼çÉ~\$Λ°¾ýö[í²Ë. ÷•——kéÒ¥ñÛË—/oòÿ«ªªªàÿ96’´qãFõìÙSUUUºé¦›´víÚLv3rNï+»åË—«²²2á>ûgšB;o¢:.Rqž3Aã9T¡3¹Ž½|ýõ×Úºuk“?Ú ÐgŸ}–¥^¥ßСCõðÃë駟Öi§¦ûî»OÆ “¤øëvúGf>öÙgŸ©´´4a̤y»PŽ[TÇá³Ï>S—.]&êÒ¥‹:wÇê˜cŽÑ#<¢gŸ}V£FÒ 7Ü “O>9þx±—††Mž>„@ŠÍ úÅ_̱:øàƒÕ¡C‡øÌËÅp\&Nœ¨yóæiáÂ…êÚµküþb?oÜŽ‹“b7é8.vÅrÎQŒçL2òõœÉ+Ùš™õä“OeeeÆ9çœc,Z´È8âˆ#ŒŠŠŠ‚^fí²Ë.3/^l|ðÁÆÌ™3Ã?ÜèÔ©S|iˆ¿þõ¯FË–-+®¸Âxúé§<ÐèÙ³§±}ûöø6.¿ür£mÛ¶Æþðãïÿ»Ñ³gÏ&3|æº7«V­2V­Ze”””úÓŸŒU«VkÖ¬1 #ºãðóŸÿÜèÔ©“qï½÷Ó§O7:vìØdå€\âu\6nÜh\vÙeƲeËŒÿûßÆÝwßmxàÆþûïolÙ²%¾B<.†a\pQQQa,Y²ÄøüóÏã_Ö×^Œçßq)öóÆ0üÿÞ\qÅÆ/ùËxû÷ßßxøá‡÷ßßxàŒ¾}û]»v5êëëãmV­Ze´hÑÂ8á„Œ%K–ǼѲeKãÍ7ßÌøëKE:ŽÍu×]g<óÌ3Æ¿ÿýoã‰'ž0ÆŽk´nÝÚxçw2þú’äïýØ|þùçFÛ¶maÆ‹-2Î:ë,£¬¬Ì˜?~¼M¾Ÿ7é:.ÅzΆÿ 8Ð8õÔSU«Vo¿ývÂãÅvΆÿq)„s&ß ‘™3g 0Úµkg >ýôÓ„6·Ür‹Ñ§O£}ûöÆO~ò“&ïܹӸòÊ+}öÙÇèر£1nÜ8cýúõ™|){þùç’’£¤¤Ä(--ÿ|ÆgÄÛDq¶lÙbœ{î¹F÷îÝîݻ矾ñý÷ßgä5&Ãë¸lÙ²Å8úè£N:mÚ´1Ž9æã®»î2þóŸÿ$l£‹aMŽ‰ùõàƒ&´+¶óÆï¸ûycþoN?ýtcĈñö«W¯6ªªªŒ6mÚ{qÊ)§$ ¦ ƒ6vÝuWcèСÆÂ… 3öš¢’Žcs饗=zô0Zµje 4ȸᆌ÷Þ{/£¯+UAþÞØaÄ‚ŠaÆíÚµ3ªªªŒG}´É¶óù¼I×q)æsÆéý·÷Þ{'´)ÆsÆï¸Â9“oJ £Hg¦ ‡0€@€@ @  ÐÈèätá…^ÐA”ínŠÈÿüÏÿ¨gÏžêÒ¥‹.¹ä}øá‡Ùî€[kÖ¬ÑÈ‘#Õ¶m[ý¿ÿ÷ÿ´hÑ"õéÓ'¾½wÞyGgŸ}¶öÞ{oýîw¿ÓW_}%IúøãÕªU+Íš5K½zõÒÁ¬gŸ}6S‡P@fÏž­1cƨcÇŽêׯŸ¾ùæ­]»6ÛÝÇÐDnĈºêª«töÙgkÆ š3gŽ$i„ Ú²e‹V¬X¡O]{íµ—žþymØ°A'tRÂþ ÃÐücUUUiÙ²eúøã5iÒ¤øãÛ·o×sÏ=§¥K—ê¼óÎÓĉÓ}Hfãƺð 5eÊ}òÉ'zë­·$‰ê.)!@9óÉõCÊwß}§·ÞzK×^{­Ú·o¯ŸýìgêÕ«—–-[o3~üxUWW«eË–’¤ššµmÛV{챇.¿ürýóŸÿ ´ÿ×^{MmÛ¶ÕE]¤.]ºèÿ÷õôÓO«¡¡!Þ¯+®¸B:uÒi§¦úúz2€P6lØ -[¶¨sçÎúî»ï4uêÔlw @ @/¾ø¢Ö­[§:¨}ûöjß¾½V­Z¥—^z)ÞfðàÁ Ϲ뮻tÔQGi·ÝvÓ±ÇÏNøY¶l™ª««ã·»uë¦æÍ›ëÝwß•$µlÙRû$©¬¬L»ï¾»Ö­[—êK‘®]»jêÔ©3fŒ†®#FÄ/0@²˜$@Z”••%dÐ>ø`uïÞ]~ø¡ëä9Íš5þIúä“OtõÕWëÙgŸÕ W_}U‡v˜ëö­9äÝqÇñÛŸ}ö™¶oß®ý÷ß_kÖ¬Iõ¥ IºüòËuùå—ÇoŸzê©Yì €B@@äJJJ4`À½ýöÛÚ¶m›$©}ûöêß¿¿®¿þzÕ××kÛ¶mzá…ô駟:nãË/¿T³fÍÔ©S'}öÙgºñÆ0`€^{í5ÇçVWWë»ï¾ÓÝwß­/¾øBS§NÕèÑ£UZÊŸ<ä.>­H‹aÆ©K—.Úk¯½âËÐÜ{ï½jÓ¦Ž<òHuïÞ]Ó¦MsÍ‚4H§žzª¬ãŽ;N'tRBæýüóÏ׌3Ô¾}{Íž=[%%%ñÇKJJôÌ3Ïhùòå2dˆºwï®[o½5þ\–¿@.*1˜j€¬#ƒ@ @€ÿß~ ò·žÆŽ²Ì2Ýc(;}ýIEND®B`‚HDDM-0.2/hddm/MPLTraits.py0000664000102500010250000000556411716202371015746 0ustar wieckiwiecki00000000000000import wx import matplotlib # We want matplotlib to use a wxPython backend if __name__ == "__main__": matplotlib.use('WXAgg') import matplotlib.pyplot import pylab as pl pl.ion() from matplotlib.backends.backend_wxagg import FigureCanvasWxAgg as FigureCanvas from matplotlib.figure import Figure from matplotlib.backends.backend_wx import NavigationToolbar2Wx from traits.api import HasTraits, Instance, Range,\ Array, on_trait_change, Property,\ cached_property, Bool from traits.api import Any, Instance from traitsui.wx.editor import Editor from traitsui.basic_editor_factory import BasicEditorFactory try: from IPython.Debugger import Tracer; debug_here = Tracer() except: pass class _MPLFigureEditor(Editor): scrollable = True def init(self, parent): self.control = self._create_canvas(parent) self.set_tooltip() def update_editor(self): pass def _create_canvas(self, parent): """ Create the MPL canvas. """ # The panel lets us add additional controls. panel = wx.Panel(parent, -1, style=wx.CLIP_CHILDREN) sizer = wx.BoxSizer(wx.VERTICAL) panel.SetSizer(sizer) # matplotlib commands to create a canvas mpl_control = FigureCanvas(panel, -1, self.value) sizer.Add(mpl_control, 1, wx.LEFT | wx.TOP | wx.GROW) toolbar = NavigationToolbar2Wx(mpl_control) sizer.Add(toolbar, 0, wx.EXPAND) self.value.canvas.SetMinSize((10,10)) return panel class MPLFigureEditor(BasicEditorFactory): klass = _MPLFigureEditor def main(): # Create a window to demo the editor from enthought.traits.api import HasTraits from enthought.traits.ui.api import View, Item from numpy import sin, cos, linspace, pi class Test(HasTraits): figure = Instance(Figure, ()) test = Range(0,5,1) view = View(Item('figure', editor=MPLFigureEditor(), show_label=False), Item('test'), width=400, height=300, resizable=True) def __init__(self): super(Test, self).__init__() self.axes = self.figure.add_subplot(111) self.t = linspace(0, 2*pi, 200) param = self.test self.line, = self.axes.plot(sin(self.t)*(1+0.5*cos(self.test*self.t)), cos(self.t)*(1+0.5*cos(11*self.t))) def update_plot(self): self.figure.axes[0].clear() self.figure.axes[0].plot(sin(self.t)*(1+0.5*cos(self.test*self.t)), cos(self.t)*(1+0.5*cos(11*self.t))) #self.axes.plot(sin(self.t)*(1+0.5*cos(self.test*self.t)), cos(self.t)*(1+0.5*cos(11*self.t))) #self.axes.redraw_in_frame() wx.CallAfter(self.figure.canvas.draw) def _test_changed(self): self.update_plot() #wx.EVT_IDLE(wx.GetApp(), callback) Test().configure_traits() #if __name__ == "__main__": # main() HDDM-0.2/hddm/likelihoods.py0000664000102500010250000001104611726176127016437 0ustar wieckiwiecki00000000000000from __future__ import division import pymc as pm import numpy as np import scipy as sp from scipy import stats from kabuki.distributions import scipy_stochastic np.seterr(divide='ignore') try: import pycuda.driver as cuda import pycuda.autoinit import pycuda.gpuarray as gpuarray import pycuda.cumath as cumath import wfpt_gpu gpu_imported = True except: gpu_imported = False import hddm def wiener_like_contaminant(value, cont_x, v, V, a, z, Z, t, T, t_min, t_max, err, nT, nZ, use_adaptive, simps_err): """Log-likelihood for the simple DDM including contaminants""" return hddm.wfpt.wiener_like_contaminant(value, cont_x.astype(np.int32), v, V, a, z, Z, t, T, t_min, t_max, err, nT, nZ, use_adaptive, simps_err) WienerContaminant = pm.stochastic_from_dist(name="Wiener Simple Diffusion Process", logp=wiener_like_contaminant, dtype=np.float, mv=True) def general_WienerCont(err=1e-4, nT=2, nZ=2, use_adaptive=1, simps_err=1e-3): _like = lambda value, cont_x, v, V, a, z, Z, t, T, t_min, t_max, err=err, nT=nT, nZ=nZ, \ use_adaptive=use_adaptive, simps_err=simps_err: \ wiener_like_contaminant(value, cont_x, v, V, a, z, Z, t, T, t_min, t_max,\ err=err, nT=nT, nZ=nZ, use_adaptive=use_adaptive, simps_err=simps_err) _like.__doc__ = wiener_like_contaminant.__doc__ return pm.stochastic_from_dist(name="Wiener Diffusion Contaminant Process", logp=_like, dtype=np.float, mv=False) class wfpt_gen(stats.distributions.rv_continuous): wiener_params = {'err': 1e-4, 'nT':2, 'nZ':2, 'use_adaptive':1, 'simps_err':1e-3} sampling_method = 'cdf' dt=1e-4 def _argcheck(self, *args): return True def _logp(self, x, v, V, a, z, Z, t, T): """Log-likelihood for the full DDM using the interpolation method""" return hddm.wfpt.wiener_like(x, v, V, a, z, Z, t, T, self.wiener_params['err'], self.wiener_params['nT'], self.wiener_params['nZ'], self.wiener_params['use_adaptive'], self.wiener_params['simps_err']) def _pdf(self, x, v, V, a, z, Z, t, T): if np.isscalar(x): out = hddm.wfpt.full_pdf(x, v, V, a, z, Z, t, T, self.dt) else: out = hddm.wfpt.pdf_array(x, v[0], V[0], a[0], z[0], Z[0], t[0], T[0], self.dt, logp=False) #out = np.empty_like(x) #for i in xrange(len(x)): # out[i] = hddm.wfpt.full_pdf(x[i], v[i], V[i], a[i], z[i], Z[i], t[i], T[i], self.dt) return out def _rvs(self, v, V, a, z, Z, t, T): param_dict = {'v':v, 'z':z, 't':t, 'a':a, 'Z':Z, 'V':V, 'T':T} sampled_rts = hddm.generate.gen_rts(param_dict, method=self.sampling_method, samples=self._size, dt=self.dt) return sampled_rts def random(self, v=1., V=0., a=2, z=.5, Z=.1, t=.3, T=.1, size=100): self._size = size return self._rvs(v, V, a, z, Z, t, T) wfpt_like = scipy_stochastic(wfpt_gen, name='wfpt', longname="""Wiener first passage time likelihood function""", extradoc="""Wiener first passage time (WFPT) likelihood function of the Ratcliff Drift Diffusion Model (DDM). Models two choice decision making tasks as a drift process that accumulates evidence across time until it hits one of two boundaries and executes the corresponding response. Implemented using the Navarro & Fuss (2009) method. Parameters: *********** v: drift-rate a: threshold z: bias [0,1] t: non-decision time References: *********** Fast and accurate calculations for first-passage times in Wiener diffusion models Navarro & Fuss - Journal of Mathematical Psychology, 2009 - Elsevier """) def wiener_like_gpu(value, v, V, a, z, t, out, err=1e-4): """Log-likelihood for the simple DDM including contaminants""" # Check if parameters are in allowed range if z<0 or z>1 or t<0 or a <= 0 or V<=0: return -np.inf wfpt_gpu.pdf_gpu(value, float(v), float(V), float(a), float(z), float(t), err, out) logp = gpuarray.sum(out).get() #cumath.log(out)).get() return np.asscalar(logp) WienerGPU = pm.stochastic_from_dist(name="Wiener Simple Diffusion Process", logp=wiener_like_gpu, dtype=np.float32, mv=False) HDDM-0.2/hddm/generate.py0000664000102500010250000005477511726662171015742 0ustar wieckiwiecki00000000000000 from __future__ import division import numpy as np import numpy.lib.recfunctions as rec from scipy.stats import uniform, norm from copy import copy import random import hddm def gen_rand_params(include=()): """Returns a dict of DDM parameters with random values. :Optional: include : tuple Which optional parameters include. Can be any combination of: * 'z' (bias, default=0.5) * 'V' (inter-trial drift variability) * 'Z' (inter-trial bias variability) * 'T' (inter-trial non-decision time variability) Special arguments are: * 'all': include all of the above * 'all_inter': include all of the above except 'z' """ params = {} if include == 'all': include = ['z','V','Z','T'] elif include == 'all_inter': include = ['V','Z','T'] from numpy.random import rand params['V'] = rand() if 'V' in include else 0 params['Z'] = rand()* 0.3 if 'Z' in include else 0 params['T'] = rand()* 0.2 if 'T' in include else 0 params['z'] = .4+rand()*0.2 if 'z' in include else 0.5 # Simple parameters params['v'] = (rand()-.5)*4 params['t'] = 0.2+rand()*0.3+(params['T']/2) params['a'] = 1.0+rand() if 'pi' in include or 'gamma' in include: params['pi'] = max(rand()*0.1,0.01) # params['gamma'] = rand() return params def gen_antisaccade_rts(params=None, samples_pro=500, samples_anti=500, dt=1e-4, subj_idx=0): if params is None: params = {'v':-2., 'v_switch': 2., 'V_switch': .1, 'a': 2.5, 't': .3, 't_switch': .3, 'z':.5, 'T': 0, 'Z':0, 'V':0} # Generate prosaccade trials pro_params = copy(params) del pro_params['t_switch'] del pro_params['v_switch'] rts = np.empty(samples_pro+samples_anti, dtype=[('response', np.float), ('rt', np.float), ('instruct', int), ('subj_idx', int)]) pro_rts = gen_rts(pro_params, samples=samples_pro, dt=dt, subj_idx=subj_idx) anti_rts = gen_rts(params, samples=samples_anti, dt=dt, subj_idx=subj_idx, method='drift') rts['instruct'][:samples_pro] = 0 rts['instruct'][samples_pro:] = 1 rts['response'][:samples_pro] = np.array((pro_rts > 0), float) rts['response'][samples_pro:] = np.array((anti_rts > 0), float) rts['rt'][:samples_pro] = np.abs(pro_rts) rts['rt'][samples_pro:] = np.abs(anti_rts) rts['subj_idx'] = subj_idx return rts, params #################################################################### # Functions to generate RT distributions with specified parameters # #################################################################### def gen_rts(params, samples=1000, range_=(-6, 6), dt=1e-3, intra_sv=1., structured=False, subj_idx=None, method='cdf'): """ Returns a numpy.array of randomly simulated RTs from the DDM. :Arguments: params : dict Parameter names and values to use for simulation. :Optional: samples : int Number of RTs to simulate. range_ : tuple Minimum (negative) and maximum (positve) RTs. dt : float Number of steps/sec. intra_sv : float Intra-trial variability. structured : bool Return a structured array with fields 'RT' and 'response'. subj_idx : int If set, append column 'subj_idx' with value subj_idx. method : str Which method to use to simulate the RTs: * 'cdf': fast, uses the inverse of cumulative density function to sample, dt can be 1e-2. * 'drift': slow, simulates each complete drift process, dt should be 1e-4. """ if params.has_key('v_switch') and method != 'drift': print "Warning: Only drift method supports changes in drift-rate. v_switch will be ignored." # Set optional default values if they are not provided for var_param in ('V', 'Z', 'T'): if var_param not in params: params[var_param] = 0 if 'z' not in params: params['z'] = .5 if 'V' not in params: params['V'] = 0 if 'Z' not in params: params['Z'] = 0 if method=='cdf_py': rts = _gen_rts_from_cdf(params, samples, range_, dt) elif method=='drift': rts = _gen_rts_from_simulated_drift(params, samples, dt, intra_sv)[0] elif method=='cdf': rts = hddm.wfpt.gen_rts_from_cdf(params['v'],params['V'],params['a'],params['z'], params['Z'],params['t'],params['T'], samples, range_[0], range_[1], dt) else: raise TypeError, "Sampling method %s not found." % method if not structured: return rts else: if subj_idx is None: data = np.empty(rts.shape, dtype = ([('response', np.float), ('rt', np.float)])) else: data = np.empty(rts.shape, dtype = ([('response', np.float), ('rt', np.float), ('subj_idx', np.float)])) data['subj_idx'] = subj_idx data['response'][rts>0] = 1. data['response'][rts<0] = 0. data['rt'] = np.abs(rts) return data def _gen_rts_from_simulated_drift(params, samples=1000, dt = 1e-4, intra_sv=1.): """Returns simulated RTs from simulating the whole drift-process. :Arguments: params : dict Parameter names and values. :Optional: samlpes : int How many samples to generate. dt : float How many steps/sec. intra_sv : float Intra-trial variability. :SeeAlso: gen_rts """ from numpy.random import rand if samples is None: samples = 1 nn = 1000 a = params['a'] v = params['v'] if params.has_key('v_switch'): switch = True t_switch = params['t_switch']/dt # Hack so that we will always step into a switch nn = int(round(t_switch)) else: switch = False #create delay if params.has_key('T'): start_delay = (uniform.rvs(loc=params['t'], scale=params['T'], size=samples) \ - params['T']/2.) else: start_delay = np.ones(samples)*params['t'] #create starting_points if params.has_key('Z'): starting_points = (uniform.rvs(loc=params['z'], scale=params['Z'], size=samples) \ - params['Z']/2.)*a else: starting_points = np.ones(samples)*params['z']*a rts = np.empty(samples) step_size = np.sqrt(dt)*intra_sv drifts = [] for i_sample in xrange(samples): drift = np.array([]) crossed = False iter = 0 y_0 = starting_points[i_sample] # drifting... if params.has_key('V') and params['V'] != 0: drift_rate = norm.rvs(v, params['V']) else: drift_rate = v if params.has_key('v_switch'): if params.has_key('V_switch') and params['V_switch'] != 0: drift_rate_switch = norm.rvs(params['v_switch'], params['V_switch']) else: drift_rate_switch = params['v_switch'] prob_up = 0.5*(1+np.sqrt(dt)/intra_sv*drift_rate) while (not crossed): # Generate nn steps iter += 1 if iter == 2 and switch: prob_up = 0.5*(1+np.sqrt(dt)/intra_sv*drift_rate_switch) position = ((rand(nn) < prob_up)*2 - 1) * step_size position[0] += y_0 position = np.cumsum(position) # Find boundary crossings cross_idx = np.where((position < 0) | (position > a))[0] drift = np.concatenate((drift, position)) if cross_idx.shape[0]>0: crossed = True else: # If not crossed, set last position as starting point # for next nn steps to continue drift y_0 = position[-1] #find the boundary interception y2 = position[cross_idx[0]] if cross_idx[0]!=0: y1 = position[cross_idx[0]-1] else: y1 = y_0 m = (y2 - y1)/dt # slope # y = m*x + b b = y2 - m*((iter-1)*nn+cross_idx[0])*dt # intercept if y2 < 0: rt = ((0 - b) / m) else: rt = ((a - b) / m) rts[i_sample] = (rt + start_delay[i_sample])*np.sign(y2) delay = start_delay[i_sample]/dt drifts.append(np.concatenate((np.ones(int(delay))*starting_points[i_sample], drift[:int(abs(rt)/dt)]))) return rts, drifts def pdf_with_params(rt, params): """Helper function that calls full_pdf and gets the parameters from the dict params. """ v = params['v']; V= params['V']; z = params['z']; Z = params['Z']; t = params['t']; T = params['T']; a = params['a'] return hddm.wfpt.full_pdf(rt,v=v,V=V,a=a,z=z,Z=Z,t=t, T=T,err=1e-4, nT=2, nZ=2, use_adaptive=1, simps_err=1e-3) def _gen_rts_from_cdf(params, samples=1000): """Returns simulated RTs sampled from the inverse of the CDF. :Arguments: params : dict Parameter names and values. :Optional: samlpes : int How many samples to generate. :SeeAlso: gen_rts """ v = params['v']; V = params['V']; z = params['z']; Z = params['Z']; t = params['t']; T = params['T']; a = params['a'] return hddm.likelihoods.wfpt.ppf(np.random.rand(samples), args=(v, V, a, z, Z, t, T)) def add_contaminate_data(data, params): """ add contaminated data""" if not params.has_key('gamma'): gamma = 1 else: gamma = params['gamma'] t_min = max(np.abs(data['rt']))+0.5 t_max = t_min+3; pi = params['pi'] n_cont = max(int(len(data)*pi),2) n_unif = max(int(n_cont*gamma),2) n_other = n_cont - n_unif l_data = range(len(data)) cont_idx = random.sample(l_data,n_cont) unif_idx = cont_idx[:n_unif] other_idx = cont_idx[n_unif:] # create guesses response = np.round(uniform.rvs(0,1,size=n_unif)) data[unif_idx]['rt'] = uniform.rvs(0,t_max,size=n_unif) * response data[unif_idx]['rt'] = response data[unif_idx[0]]['rt'] = min(abs(data['rt']))/2. data[unif_idx[1]]['rt'] = max(abs(data['rt'])) + 0.8 #create late responses response = (np.sign(gen_rts(params, n_other))+1) / 2 data[other_idx]['rt'] = uniform.rvs(t_min,t_max,size=n_other) * response return data def gen_rand_data(samples=500, params=None, include=(), method='cdf'): """Generate simulated RTs with random parameters. :Optional: params : dict Parameter names and values. If not supplied, takes random values. samlpes : int How many samples to generate. include : tuple Which inter-trial variability parameters to include ('V', 'Z', 'T') :Returns: data array with RTs parameter values """ if params is None: params = gen_rand_params(include=include) # Create RT data data = gen_rts(params, samples=samples, structured=True, method=method) if params.has_key('pi'): add_contaminate_data(data, params) return (data, params) def gen_rand_cond_subj_data(cond_params=None, samples_per_cond=100, n_conds=None, num_subjs=10, include=(), noise=.05): """Generate simulated RTs with multiple conditions. :Optional: cond_params : a dictionary of params if cond_params[key] is a list then a new condition is created for each item in the list for instance, if cond_params = {'a':2, 't':0.3, 'v':[1,2], ...} then two set of params will be created: 1) - {'a':2, 't':0.3, 'v': 1, ...} 2) - {'a':2, 't':0.3, 'v': 2, ...} if cond_params is None then it is genreated randomly samlpes_per_cond : int How many samples to generate for each condition. n_conds : int number of conditions num_subjs : int How many subjects to generate data for include: tuple which extra parameters to include noise : float Amount of noise to add to each parameter :Returns: data : array RTs params_subj: list parameter values for each condition combined_params: dictionary a dictionary used by check_model """ # Create RT data if cond_params == None: cond_params, original_combined_params = gen_cond_params_v(n_conds, include=include) data_out = [] params_subj = [] combined_params = {} #loop over subjects for subj_idx in range(num_subjs): i_cond_params = _add_noise(cond_params, noise, include=include) i_combined_params = cond_params_to_combined_params(i_cond_params) #create combined params for (key, value) in i_combined_params.iteritems(): combined_params[key + str(subj_idx)] = value #create data for subject data_subj, dummy1, dummy2 = gen_rand_cond_data(cond_params=i_cond_params, samples_per_cond=samples_per_cond, n_conds=n_conds, subj_idx=subj_idx) data_out.append(data_subj) params_subj.append(i_cond_params) return rec.stack_arrays(data_out, usemask=False), params_subj, combined_params def gen_rand_cond_data(cond_params=None, samples_per_cond=100, n_conds=None, include = (), subj_idx=None): """Generate simulated RTs with multiple conditions. :Input: cond_params : a dictionary of params if cond_params[key] is a list then a new condition is created for each item in the list for instance, if cond_params = {'a':2, 't':0.3, 'v':[1,2], ...} then two set of params will be created: 1) - {'a':2, 't':0.3, 'v': 1, ...} 2) - {'a':2, 't':0.3, 'v': 2, ...} if cond_params is None then it is genreated randomly samlpes_per_cond : int How many samples to generate for each condition. n_conds : int number of conditions include: tuple which extra parameters to include noise : float Amount of noise to add to each parameter :Returns: data : array RTs cond_params: dictionary see Input combined_params: dictionary a dictionary used by check_model """ if cond_params is None: cond_params, combined_params = gen_cond_params_v(n_conds) else: combined_params = cond_params_to_combined_params(cond_params) params_set = cond_params_to_params_set(cond_params) conds = range(n_conds) if type(conds[0]) is str: cond_type = 'S12' else: cond_type = type(conds[0]) arrays = [] for cond, params in zip(conds, params_set): i_data = gen_rts(params, samples=samples_per_cond, structured=True) if subj_idx is None: data = np.empty(len(i_data), dtype = ([('response', np.float), ('rt', np.float), ('cond', cond_type)])) else: data = np.empty(len(i_data), dtype = ([('response', np.float), ('rt', np.float), ('cond', cond_type), ('subj_idx', np.int)])) data['subj_idx'] = subj_idx data['response'] = np.sign(i_data['response']) data['rt'] = np.abs(i_data['rt']) data['cond'] = cond arrays.append(data) data_out = rec.stack_arrays(arrays, usemask=False) if cond_params.has_key('pi'): add_contaminate_data(data_out, cond_params) return data_out, cond_params, combined_params def gen_cond_params_v(n_conds = 3, params = None, include = ()): """ generate params set with multiple conditions for v Input: n_conds - number of conditions params - the basic params. if not supplied than it is randomly generated include : tuple Which inter-trial variability parameters to include ('V', 'Z', 'T') output: cond_params : list of params combined_params: all the params combined together in one dictionary (used to check the model) """ if params == None: params = hddm.generate.gen_rand_params(include) cond_params = copy(params) combined_params = copy(params) del combined_params['v'] cond_params['v'] = np.linspace(min(0.5,params['v']/2) , max(params['v']*2, 3), n_conds) for i in range(n_conds): combined_params['v(%d,)'%i] = cond_params['v'][i] return cond_params, combined_params def cond_params_to_combined_params(cond_params): """ transform cond_params to combined_params (combine_params are used by check_model) """ cond_params = copy(cond_params) combined_params = copy(cond_params) n_conds = max([len(x) for x in cond_params.values() if not np.isscalar(x)] + [1]) dep_keys = [x for x in cond_params.keys() if not np.isscalar(cond_params[x])] for key in dep_keys: del combined_params[key] for i in range(n_conds): combined_params['%s(%d,)'%(key, i)] = cond_params[key][i] return combined_params def cond_params_to_params_set(cond_params): """ transform cond_params to params_set """ n_conds = max([len(x) for x in cond_params.values() if not np.isscalar(x)] + [1]) params_set = [None]*(n_conds) for i_cond in range(n_conds): params_set[i_cond] = {} for key in cond_params.iterkeys(): if np.isscalar(cond_params[key]): params_set[i_cond][key] = cond_params[key] else: params_set[i_cond][key] = cond_params[key][i_cond] return params_set def _add_noise(params, noise=.1, include=()): """Add individual noise to each parameter. :Arguments: params : dict Parameter names and values :Optional: noise : float Standard deviation of random gaussian variable to add to each parameter. include : tuple Which inter-trial variability parameters to include. Can be any combination of ('V', 'Z', 'T'). :Returns: dict with parameters with added noise. """ params = copy(params) for param, value in params.iteritems(): if param in include or param in ('v','a','t'): params[param] = np.random.normal(loc=value, scale=noise) return params def gen_rand_subj_data(num_subjs=10, params=None, samples=100, noise=0.1,include=()): """Generate simulated RTs of multiple subjects. :Optional: num_subjs : int How many subjects to generate data for. params : dict Mapping of parameter names to values. If not provided, gets set randomly. samples : int How many samples to generate for each subject. noise : float Inter-subject variability. include : tuple Which inter-trial variability parameters to include. Can be any combination of ('V', 'Z', 'T'). :Returns: numpy.recarray: with fields 'RT', 'response' and 'subj_idx' and samples*num_subjs entries. dict: Mapping of parameter names (with subject ids) to parameter values. """ if params is None: params = gen_rand_params(include=include) #{'v': .5, 'V': 0.1, 'z': .5, 'Z': 0.1, 't': .3, 'T': 0.1, 'a': 2} params_orig = copy(params) data_gens = [] # Derive individual parameters for i in range(num_subjs): params_subj = copy(params_orig) params_subj = _add_noise(params_subj, noise=noise, include=include) for name, value in params_subj.iteritems(): params['%s%i'%(name,i)] = value if name in include or name in ('v','a','t'): params['%stau'%name] = noise # Create RT data data_gen = gen_rts(params_subj, samples=samples, structured=True, subj_idx=i) if params.has_key('pi'): add_contaminate_data(data_gen, params_subj) data_gens.append(data_gen) return (rec.stack_arrays(data_gens, usemask=False), params) def gen_correlated_rts(num_subjs=10, params=None, samples=100, correlation=.1, cor_param=None, subj_noise=.1): """Generate RT data where cor_param is linearly influenced by another variable. """ if params is None: params = {'v': .5, 'V': 0, 'z': .5, 'Z': 0., 't': .3, 'T': 0., 'a': 2, 'e': correlation} if cor_param is None: cor_param = 'a' data = np.empty(samples*num_subjs, dtype = ([('response', np.float), ('rt', np.float), ('subj_idx', np.float), ('cov', np.float)])) params_subjs = [] # Generate RTs with given parameters (influenced by covariate) trial = 0 for subj_idx in range(num_subjs): params_subj = copy(params) params_subj = _add_noise(params_subj, subj_noise) params_subjs.append(params_subj) for i in range(samples): params_trial = copy(params_subj) data['subj_idx'][trial] = subj_idx # Calculate resulting cor_param values data['cov'][trial] = np.random.rand()-.5 # Calculate new param values based on covariate params_trial[cor_param] = data['cov'][trial] * params_subj['e'] + params_subj[cor_param] # Generate RT (resample if it didn't reach threshold) rt = np.array([]) while rt.shape == (0,): rt = gen_rts(params_trial, samples=1, structured=False) data['rt'][trial] = rt trial+=1 data['response'][data['rt']>0] = 1. data['response'][data['rt']<0] = 0. data['rt'] = np.abs(data['rt']) return data, params_subjs HDDM-0.2/hddm/tests/0000775000102500010250000000000011727455647014727 5ustar wieckiwiecki00000000000000HDDM-0.2/hddm/tests/matlab_values.py0000664000102500010250000000656611713025572020116 0ustar wieckiwiecki00000000000000# Values obtained from running against the Fuss & Navarro 2009 reference implementation vals = [(0.3325402105490861, 0.18224585277734096, 2.0210322268188046, 0.37178992456396914, 0.7513994191503139, 1.6883221884854474, 1.0, 0.082198565245068272), (-0.13074510229340497, 0.44696631528174735, 2.4890334572448456, 0.3816245478330931, 0.9498762676625047, 1.0790903665954314, 0.1, 0.41591722288506577), (0.6056010862462016, 0.16871208306308166, 1.5524386791394562, 0.1286318302691049, 0.19969302867825994, 0.3836652143729468, 1e-07, 0.62063048896871487), (0.3895703269526853, 0.3435094918413777, 2.1750677359226023, 0.25344975464683345, 0.5512703840098271, 3.688305473258689, 0.01, 0.0090709779903460266), (0.42003699302964476, 0.10014569263424461, 2.3012075231555444, 0.0019673685361806803, 0.004527323276278492, 3.009144940928847, 0.0001, 0.0001884560320134888), (-0.5255928511420052, 0.41071488550017876, 2.2389065101119003, 0.2301372155046244, 0.515255710012329, 0.5621315133209579, 1e-05, 1.8640810217155657), (0.008338807811151538, 0.30523807964872274, 2.4560203575759623, 0.032544407891560656, 0.07992972830692877, 3.802538736807515, 1e-05, 0.0030406047220150243), (-0.709463423093855, 0.07788052784603428, 2.0241086070651075, 0.1491208977241244, 0.3018368925766758, 1.213966089059249, 1.0, 0.082017502354950353), (0.460913299169564, 0.2168584116589385, 2.298380160881188, 0.4461371311255309, 1.0253927312113693, 0.25156507383357796, 0.0001, 1.0372891338943479e-05), (-0.6117806447011472, 0.49463268264815274, 1.9686542627132415, 0.25832570907838626, 0.5085540083455858, 1.94831451612509, 0.001, 0.097069006460126561), (0.32861624103817627, 0.12335537773273014, 2.002692863678406, 0.47249065415592595, 0.9462536612328146, 2.796439295540539, 0.001, 0.018459337529972406), (0.021183546657572494, 0.47799029047254493, 2.3569808966748047, 0.4855005752397232, 1.1443155811646561, 2.610216277868382, 0.1, 0.082974887901830302), (0.7175299762398688, 0.3566056886339434, 2.220795104898356, 0.45773365470000554, 1.0165326597050066, 2.1966769477884225, 1e-07, 0.030138660738276948), (-0.7178918281791459, 0.057776773410309845, 1.9358335812071767, 0.10703342161705387, 0.20719889187779905, 1.9963028315012206, 0.01, 0.015193622821588188), (0.38713456753496744, 0.07157449665799648, 2.2096131432869957, 0.017240717413939843, 0.038095315797538463, 3.426246718806154, 1e-07, 0.00089937286951603833), (0.7331877690347837, 0.46552429958519015, 1.8562721149635684, 0.34330970798729826, 0.6372762377331072, 1.6101500341565407, 1.0, 0.071864719581762077), (-0.006430595681853335, 0.1207541730285373, 1.6676102104060564, 0.2604537143639433, 0.4343352734114944, 0.7713437644783603, 1.0, 0.26065820692218827), (-0.5187897244595112, 0.21532637320553233, 2.481938472673327, 0.18482168074390404, 0.4587160400224425, 1.0697490678038166, 1e-09, 0.23165988979463398), (-0.24925121543920914, 0.05753410930092823, 2.1709492103847565, 0.477487812061141, 1.036601788562479, 2.211403925199262, 1e-08, 0.084455290527924637), (-0.5719411096293123, 0.28837526790403306, 2.1929805341020145, 0.2243768947148022, 0.49205416241181843, 1.0791694385505024, 1e-10, 0.27868915917309367)] HDDM-0.2/hddm/tests/test_models.py0000664000102500010250000001071611726176127017620 0ustar wieckiwiecki00000000000000from __future__ import division import nose import sys import unittest from hddm.diag import * import numpy as np import numpy.testing import numpy.lib.recfunctions as rec import matplotlib.pyplot as plt import platform from copy import copy import subprocess import pymc as pm import hddm from hddm.likelihoods import * from hddm.generate import * from scipy.stats import scoreatpercentile from numpy.random import rand from nose import SkipTest def diff_model(param, subj=True, num_subjs=10, change=.5, samples=500): params = {'v':.5, 'a':2., 'z':.5, 't': .3, 'T':0., 'V':0., 'Z':0.} params[param] = [params[param], params[param]+change] data, tmp, tmp2 = hddm.generate.gen_rand_cond_data(cond_params=params, n_conds=2, samples_per_cond=samples) model = hddm.model.HDDM(data, depends_on={param:['cond']}, is_group_model=subj) return model class TestMulti(unittest.TestCase): def runTest(self): pass def test_diff_v(self, samples=1000): m = diff_model('v', subj=False, change=.5, samples=samples) return m def test_diff_a(self, samples=1000): m = diff_model('a', subj=False, change=-.5, samples=samples) return m def test_diff_a_subj(self, samples=1000): raise SkipTest("Disabled.") m = diff_model('a', subj=True, change=-.5, samples=samples) return m class TestSingle(unittest.TestCase): def __init__(self, *args, **kwargs): super(TestSingle, self).__init__(*args, **kwargs) self.samples = 10000 self.burn = 5000 def runTest(self): return def test_HDDM(self, assert_=True): includes = [[], ['z'],['z', 'V'],['z', 'T'],['z', 'Z'], ['z', 'Z','T'], ['z', 'Z','T','V']] for include in includes: data, params_true = hddm.generate.gen_rand_data(samples=500, include=include, method='cdf') model = hddm.model.HDDM(data, include=include, bias='z' in include, is_group_model=False) mc = model.mcmc() mc.sample(self.samples, burn=self.burn) check_model(mc, params_true, assert_=assert_) return mc def test_HDDM_group(self, assert_=True): raise SkipTest("Disabled.") includes = [[], ['z'],['z', 'V'],['z', 'T'],['z', 'Z'], ['z', 'Z','T'], ['z', 'Z','T','V']] for include in includes: data, params_true = hddm.generate.gen_rand_subj_data(samples=500, num_subjs=5) model = hddm.model.HDDM(data, include=include, bias='z' in include, is_group_model=True) mc = model.mcmc() mc.sample(self.samples, burn=self.burn) check_model(mc, params_true, assert_=assert_) return mc def test_cont(self, assert_=False): raise SkipTest("Disabled.") params_true = gen_rand_params(include = ()) data, temp = hddm.generate.gen_rand_data(samples=300, params=params_true) data[0]['rt'] = min(abs(data['rt']))/2. data[1]['rt'] = max(abs(data['rt'])) + 0.8 hm = hddm.HDDMContUnif(data, bias=True, is_group_model=False) hm.sample(self.samples, burn=self.burn) check_model(hm.mc, params_true, assert_=assert_) cont_res = hm.cont_report(plot=False) cont_idx = cont_res['cont_idx'] self.assertTrue((0 in cont_idx) and (1 in cont_idx), "did not find the right outliers") self.assertTrue(len(cont_idx)<15, "found too many outliers (%d)" % len(cont_idx)) return hm def test_cont_subj(self, assert_=False): raise SkipTest("Disabled.") data_samples = 200 num_subjs = 2 data, params_true = hddm.generate.gen_rand_subj_data(num_subjs=num_subjs, params=None, samples=data_samples, noise=0.0001,include=()) for i in range(num_subjs): data[data_samples*i]['rt'] = min(abs(data['rt']))/2. data[data_samples*i + 1]['rt'] = max(abs(data['rt'])) + 0.8 hm = hddm.model.HDDMContUnif(data, bias=True, is_group_model=True) hm.sample(self.samples, burn=self.burn) check_model(hm.mc, params_true, assert_=assert_) cont_res = hm.cont_report(plot=False) for i in range(num_subjs): cont_idx = cont_res[i]['cont_idx'] self.assertTrue((0 in cont_idx) and (1 in cont_idx), "did not found the right outliers") self.assertTrue(len(cont_idx)<15, "found too many outliers (%d)" % len(cont_idx)) return hm if __name__=='__main__': print "Run nosetest.py" HDDM-0.2/hddm/tests/test_generate.py0000664000102500010250000000226311726176127020125 0ustar wieckiwiecki00000000000000 import unittest import pymc as pm import hddm from hddm.likelihoods import * from hddm.generate import * from scipy.integrate import * from scipy.stats import kstest from nose import SkipTest class TestGenerate(unittest.TestCase): def test_compare_sampled_data_to_analytic(self): sampler = hddm.likelihoods.wfpt_like sampler.sample_method = 'cdf' sampler.dt = 1e-6 self._test_compare_samples_to_analytic(sampler.rv) def test_compare_drift_simulated_data_to_analytic(self): sampler = hddm.likelihoods.wfpt_like sampler.sample_method = 'drifts' self._test_compare_samples_to_analytic(sampler.rv) def _test_compare_samples_to_analytic(self, sampler): includes = [[],['z', 'V'],['z', 'T'],['z', 'Z'], ['z', 'Z','T'], ['z', 'Z','T','V']] for include in includes: params = hddm.generate.gen_rand_params(include=include) [D, p_value] = kstest(sampler.rvs, sampler.cdf, args=(params['v'], params['V'], params['a'], params['z'], params['Z'], params['t'], params['T']), N=1000) print 'p_value: %f' % p_value self.assertTrue(p_value > 0.05) HDDM-0.2/hddm/tests/test_likelihoods.py0000664000102500010250000003242311726176127020642 0ustar wieckiwiecki00000000000000import numpy as np from numpy.random import rand import unittest import pymc as pm import hddm from hddm.likelihoods import * from hddm.generate import * from scipy.integrate import * from scipy.stats import kstest from nose import SkipTest np.random.seed(3123) def gen_like_from_matlab(samples=20, seed=3123): # Generate samples from the reference implementation by Navarro & Fuss 2009 # This was used to generate matlab_values.py for comparison with our implementation. import mlabwrap np.random.seed(seed) vals = [] for i in range(samples): v = (rand()-.5)*1.5 t = rand()*.5 a = 1.5+rand() z = .5*rand() z_nonorm = a*z rt = rand()*4 + t err = 10**(round(rand()*-10)) # Test if equal up to the 9th decimal. matlab_wfpt = mlabwrap.mlab.wfpt(rt-t, v, a, z_nonorm, err)[0][0] vals.append((v, t, a, z, z_nonorm, rt, err, matlab_wfpt)) return vals class TestWfpt(unittest.TestCase): def __init__(self, *args, **kwargs): super(TestWfpt, self).__init__(*args, **kwargs) self.samples=5000 self.range_ = (-10,10) def test_pdf_no_matlab(self): # Test if our wfpt pdf implementation yields the same results as the reference implementation by Navarro & Fuss 2009 from matlab_values import vals for v, t, a, z, z_nonorm, rt, err, matlab_wfpt in vals: python_wfpt = hddm.wfpt.full_pdf(-rt, v, 0, a, z, 0, t, 0, err, 0) print v,t,a,z,z_nonorm,rt,err, matlab_wfpt, python_wfpt np.testing.assert_array_almost_equal(matlab_wfpt, python_wfpt, 9) def test_pdf(self): # Test if our wfpt pdf implementation yields the same results as the reference implementation by Navarro & Fuss 2009 try: import mlabwrap except ImportError: raise SkipTest("Could not import mlabwrap, not performing pdf comparison test.") for i in range(500): v = (rand()-.5)*1.5 # TODO! t = rand()*.5 a = 1.5+rand() z = .5*rand() z_nonorm = a*z rt = rand()*4 + t err = 1e-4 # Test if equal up to the 9th decimal. matlab_wfpt = mlabwrap.mlab.wfpt(rt-t, v, a, z_nonorm, err)[0][0] python_wfpt = hddm.wfpt.pdf_array(np.asarray([-rt]), v, 0, a, z, 0, t, 0, err, 0)[0] print v,t,a,z,z_nonorm,rt,err, matlab_wfpt, python_wfpt np.testing.assert_array_almost_equal(matlab_wfpt, python_wfpt, 9) def test_summed_logp(self): params = hddm.generate.gen_rand_params(include=('V','Z','T')) # Generate random valid RTs rts = params['t'] + params['T'] + rand(50)*2 p = hddm.wfpt.pdf_array(rts, params['v'], params['V'], params['a'], params['z'], params['Z'], params['t'], params['T'], 1e-4, logp=True) summed_logp = np.sum(p) summed_logp_like = hddm.wfpt.wiener_like(rts, params['v'], params['V'], params['a'], params['z'], params['Z'], params['t'], params['T'], 1e-4) np.testing.assert_almost_equal(summed_logp, summed_logp_like, 2) self.assertTrue(-np.Inf == hddm.wfpt.wiener_like(np.array([1.,2.,3.,0.]), 1, 0, 2, .5, 0, 0, 0, 1e-4)), "wiener_like_simple should have returned -np.Inf" def test_pdf_V(self, samples=50): """Test if our wfpt pdf_V implementation produces the same value as numerical integration over v""" func = lambda v_i,value,err,v,V,z,a: hddm.wfpt.full_pdf(value, v_i, 0, a, z, 0, 0, 0, err) * norm.pdf(v_i,v,V) for i in range(50): V = rand()*0.4+0.1 v = (rand()-.5)*4 t = rand()*.5 a = 1.5+rand() z = .5*rand() rt = rand()*4 + t err = 10**(-3- np.ceil(rand()*12)) # Test if equal up to the 9th decimal. sp_result = quad(func, -np.inf, np.inf, args=(rt,err,v,V,z,a), epsrel=1e-10, epsabs=1e-10)[0] hddm_result = hddm.wfpt.full_pdf(rt, v, V, a, z, 0, 0, 0, err) np.testing.assert_array_almost_equal(hddm_result, sp_result) class TestWfptFull(unittest.TestCase): def test_adaptive(self): for i in range(20): V = rand()*0.4+0.1 v = (rand()-.5)*4 T = rand()*0.3 t = rand()*.5+(T/2) a = 1.5+rand() rt = (rand()*4 + t) * np.sign(rand()) err = 10**-9 Z = rand()*0.3 z = .5*rand()+Z/2 logp = 0#np.floor(rand()*2) nZ = 60 nT = 60 my_res = hddm.wfpt.full_pdf(rt,v=v,V=0,a=a,z=z,Z=0,t=t, T=T,err=err, nT=5, nZ=5, use_adaptive=1) res = hddm.wfpt.full_pdf(rt,v=v,V=0,a=a,z=z,Z=0,t=t, T=T,err=err, nT=nT, nZ=nZ, use_adaptive=0) print "(%d) rt %f, v: %f, V: %f, z: %f, Z: %f, t: %f, T: %f a: %f" % (i,rt,v,V,z,Z,t,T,a) print my_res print res if np.isinf(my_res): my_res = 100 if np.isinf(res): res = 100 self.assertTrue(not np.isnan(my_res)), "Found NaN in the results" self.assertTrue(not np.isnan(res)), "Found NaN in the simulated results" np.testing.assert_array_almost_equal(my_res, res, 3) def test_pdf_integrate_to_one(self): for tests in range(2): V = rand()*0.4+0.1 v = (rand()-.5)*4 T = rand()*0.3 t = rand()*.5+(T/2) a = 1.5+rand() err = 10**-8 Z = rand()*0.3 z = .5*rand()+Z/2 func = lambda x: np.exp(hddm.wfpt.wiener_like(np.array([x]), v, V, a, z, Z, t, T, err)) integ, error = sp.integrate.quad(func, a=-5, b=5) np.testing.assert_almost_equal(integ, 1, 2) def test_wiener_like_full_single(self): for i in range(20): V = rand()*0.4+0.1 v = (rand()-.5)*4 T = rand()*0.3 t = rand()*.5+(T/2) a = 1.5+rand() rt = (rand()*4 + t) * np.sign(rand()) err = 10**-8 Z = rand()*0.3 z = .5*rand()+Z/2 nZ = 60 nT = 60 my_res = np.zeros(8) res = np.zeros(8) y_z = np.zeros(nZ+1); y_t = np.zeros(nT+1) for vvv in range(2): #test pdf my_res[0+vvv*4] = hddm.wfpt.full_pdf(rt,v=v,V=V*vvv,a=a,z=z,Z=0,t=t, T=0,err=err, nT=nT, nZ=nZ) res[0+vvv*4] = hddm.wfpt.full_pdf(rt,v=v,V=V*vvv,a=a,z=z,Z=0,t=t, T=0,err=err, nT=0, nZ=0) #test pdf + Z my_res[1+vvv*4] = hddm.wfpt.full_pdf(rt,v=v,V=V*vvv,a=a,z=z,Z=Z,t=t, T=0,err=err, nT=nT, nZ=nZ) hZ = Z/nZ for j in range(nZ+1): z_tag = z-Z/2. + hZ*j y_z[j] = hddm.wfpt.full_pdf(rt,v=v,V=V*vvv,a=a,z=z_tag,Z=0,t=t, T=0,err=err, nT=0, nZ=0)/Z res[1+vvv*4] = simps(y_z, x=None, dx=hZ) #test pdf + T my_res[2+vvv*4] = hddm.wfpt.full_pdf(rt,v=v,V=V*vvv,a=a,z=z,Z=0,t=t, T=T,err=err, nT=nT, nZ=nZ) hT = T/nT for j in range(nT+1): t_tag = t-T/2. + hT*j y_t[j] = hddm.wfpt.full_pdf(rt,v=v,V=V*vvv,a=a,z=z,Z=0,t=t_tag, T=0,err=err, nT=0, nZ=0)/T res[2+vvv*4] = simps(y_t, x=None, dx=hT) #test pdf + Z + T my_res[3+vvv*4] = hddm.wfpt.full_pdf(rt,v=v,V=V*vvv,a=a,z=z,Z=Z,t=t, T=T,err=err, nT=nT, nZ=nZ) hT = T/nT hZ = Z/nZ for j_t in range(nT+1): t_tag = t-T/2. + hT*j_t for j_z in range(nZ+1): z_tag = z-Z/2. + hZ*j_z y_z[j_z] = hddm.wfpt.full_pdf(rt,v=v,V=V*vvv,a=a,z=z_tag,Z=0,t=t_tag, T=0,err=err, nT=0, nZ=0)/Z/T y_t[j_t] = simps(y_z, x=None, dx=hZ) res[3+vvv*4] = simps(y_t, x=None, dx=hT) print "(%d) rt %f, v: %f, V: %f, z: %f, Z: %f, t: %f, T: %f a: %f" % (i,rt,v,V,z,Z,t,T,a) print my_res print res my_res[np.isinf(my_res)] = 100 res[np.isinf(res)] = 100 self.assertTrue(not any(np.isnan(my_res))), "Found NaN in the results" self.assertTrue(not any(np.isnan(res))), "Found NaN in the simulated results" np.testing.assert_array_almost_equal(my_res, res, 2) def test_failure_mode(self): rt = 0.6 for i in range(2): rt = rt * -1 v = 1 V = 1 a = 1.5 z = 0.5 Z = 0.2 t = 0.2 T = 0.1 nT = 10; nZ =10 z = 1.1 self.assertTrue(hddm.wfpt.full_pdf(rt,v=v,V=V,a=a,z=z,Z=Z,t=t, T=T,err=1e-10, nT=10, nZ=10)==0) z = -0.1 self.assertTrue(hddm.wfpt.full_pdf(rt,v=v,V=V,a=a,z=z,Z=Z,t=t, T=T,err=1e-10, nT=10, nZ=10)==0) z = 0.5 z = 0.1 Z = 0.25 self.assertTrue(hddm.wfpt.full_pdf(rt,v=v,V=V,a=a,z=z,Z=Z,t=t, T=T,err=1e-10, nT=10, nZ=10)==0) z = 0.5 a = -0.1 self.assertTrue(hddm.wfpt.full_pdf(rt,v=v,V=V,a=a,z=z,Z=Z,t=t, T=T,err=1e-10, nT=10, nZ=10)==0) a = 1.5 t = 0.7 T = 0 self.assertTrue(hddm.wfpt.full_pdf(rt,v=v,V=V,a=a,z=z,Z=Z,t=t, T=T,err=1e-10, nT=10, nZ=10)==0) t = -0.3 self.assertTrue(hddm.wfpt.full_pdf(rt,v=v,V=V,a=a,z=z,Z=Z,t=t, T=T,err=1e-10, nT=10, nZ=10)==0) t = 0.1 T = 0.3 self.assertTrue(hddm.wfpt.full_pdf(rt,v=v,V=V,a=a,z=z,Z=Z,t=t, T=T,err=1e-10, nT=10, nZ=10)==0) t = 0.2 T = 0.1 class TestWfptSwitch(unittest.TestCase): def __init__(self, *args, **kwargs): super(TestWfptSwitch, self).__init__(*args, **kwargs) self.tests = 2 self.skip = False try: import wfpt_switch except ImportError: self.skip = True def gen_rand_params(self): vpp = (rand()-.5)*4 vcc = (rand()-.5)*4 tcc = rand()*0.3 t = rand()*.5 a = 1.+rand() return vpp, vcc, tcc, t, a def testDriftDensIntegrateToOne(self): if self.skip: raise SkipTest("Could not import wfpt_switch.") # Test for small tcc where no drifts should have crossed the threshold integ, error = sp.integrate.quad(hddm.wfpt_switch.calc_drift_dens, args=(.05, 0, 2, 1, False), a=0, b=2) # Not sure why this returns 2, but the resulting likelihood seems to work np.testing.assert_almost_equal(integ, 2, 2) def test_pdf_integrate_to_one_precomp(self): if self.skip: raise SkipTest("Could not import wfpt_switch.") for tests in range(self.tests): vpp, vcc, tcc, t, a = self.gen_rand_params() func = lambda x: np.exp(hddm.wfpt_switch.wiener_like_antisaccade_precomp(np.array([x]), vpp, vcc, 0, a, .5, t, tcc, 0, 1e-4)) integ, error = sp.integrate.quad(func, a=-5, b=5) np.testing.assert_almost_equal(integ, 1, 2) def test_pdf_integrate_to_one(self): if self.skip: raise SkipTest("Could not import wfpt_switch.") for tests in range(self.tests): vpp, vcc, tcc, t, a = self.gen_rand_params() func = lambda x: np.exp(hddm.wfpt_switch.wiener_like_antisaccade(np.array([x]), vpp, vcc, 0, a, .5, t, tcc, 0, 1e-6)) integ, error = sp.integrate.quad(func, a=-5, b=5) np.testing.assert_almost_equal(integ, 1, 2) def test_pdf_precomp_integrate_to_one(self): if self.skip: raise SkipTest("Could not import wfpt_switch.") for tests in range(self.tests): vpp, vcc, tcc, t, a = self.gen_rand_params() func = lambda x: np.exp(hddm.wfpt_switch.wiener_like_antisaccade_precomp(np.array([x]), vpp, vcc, 0, a, .5, t, tcc, 0, 1e-6)) integ, error = sp.integrate.quad(func, a=-5, b=5) np.testing.assert_almost_equal(integ, 1, 2) def test_ks(self): if self.skip: raise SkipTest("Could not import wfpt_switch.") for tests in range(self.tests): vpp, vcc, tcc, t, a = self.gen_rand_params() tcc += .1 # Test for bigger tcc sampler = hddm.sandbox.model.wfpt_switch_like.rv [D, p_value] = kstest(sampler.rvs, sampler.cdf, args=(vpp, vcc, 0, a, .5, t, tcc, 0), N=100) print 'p_value: %f' % p_value self.assertTrue(p_value > 0.05) @unittest.expectedFailure def test_ks_small_tcc(self): if self.skip: raise SkipTest("Could not import wfpt_switch.") for tests in range(self.tests): vpp, vcc, tcc, t, a = self.gen_rand_params() tcc = 0.02 sampler = hddm.sandbox.model.wfpt_switch_like.rv [D, p_value] = kstest(sampler.rvs, sampler.cdf, args=(vpp, vcc, 0, a, .5, t, tcc, 0), N=100) print 'p_value: %f' % p_value self.assertTrue(p_value > 0.05) if __name__=='__main__': print "Run nosetest." HDDM-0.2/hddm/tests/__init__.py0000664000102500010250000000011211713025572017013 0ustar wieckiwiecki00000000000000import unittest from test_models import * from test_likelihoods import * HDDM-0.2/hddm/model.py0000664000102500010250000002001211726447331015220 0ustar wieckiwiecki00000000000000""" .. module:: HDDM :platform: Agnostic :synopsis: Definition of HDDM models. .. moduleauthor:: Thomas Wiecki Imri Sofer """ from __future__ import division import numpy as np import pymc as pm import matplotlib.pyplot as plt import hddm import kabuki from kabuki.hierarchical import Parameter from copy import copy, deepcopy class HDDM(kabuki.Hierarchical): """Implements the hierarchical Ratcliff drift-diffusion model using the Navarro & Fuss likelihood and numerical integration over the variability parameters. :Arguments: data : numpy.recarray Input data with a row for each trial. Must contain the following columns: * 'rt': Reaction time of trial in seconds. * 'response': Binary response (e.g. 0->error, 1->correct) May contain: * 'subj_idx': A unique ID (int) of the subject. * Other user-defined columns that can be used in depends on keyword. :Example: >>> data, params = hddm.generate.gen_rand_data() # gen data >>> model = hddm.HDDM(data) # create object >>> mcmc = model.mcmc() # Create pymc.MCMC object >>> mcmc.sample() # Sample from posterior :Optional: include : iterable Optional inter-trial variability parameters to include. Can be any combination of 'V', 'Z' and 'T'. Passing the string 'all' will include all three. Note: Including 'Z' and/or 'T' will increase run time significantly! is_group_model : bool If True, this results in a hierarchical model with separate parameter distributions for each subject. The subject parameter distributions are themselves distributed according to a group parameter distribution. If not provided, this parameter is set to True if data provides a column 'subj_idx' and False otherwise. depends_on : dict Specifies which parameter depends on data of a column in data. For each unique element in that column, a separate set of parameter distributions will be created and applied. Multiple columns can be specified in a sequential container (e.g. list) :Example: >>> hddm.HDDM(data, depends_on={'v':'difficulty'}) Separate drift-rate parameters will be estimated for each difficulty. Requires 'data' to have a column difficulty. bias : bool Whether to allow a bias to be estimated. This is normally used when the responses represent left/right and subjects could develop a bias towards responding right. This is normally never done, however, when the 'response' column codes correct/error. plot_var : bool Plot group variability parameters when calling pymc.Matplot.plot() (i.e. variance of Normal distribution.) wiener_params : dict Parameters for wfpt evaluation and numerical integration. :Parameters: * err: Error bound for wfpt (default 1e-4) * nT: Maximum depth for numerical integration for T (default 2) * nZ: Maximum depth for numerical integration for Z (default 2) * use_adaptive: Whether to use adaptive numerical integration (default True) * simps_err: Error bound for Simpson integration (default 1e-3) """ def __init__(self, data, bias=False, include=(), wiener_params=None, **kwargs): # Flip sign for lower boundary RTs data = hddm.utils.flip_errors(data) include = set(include) if include is not None: if include == 'all': [include.add(param) for param in ('T','V','Z')] else: [include.add(param) for param in include] if bias: include.add('z') if wiener_params is None: self.wiener_params = {'err': 1e-4, 'nT':2, 'nZ':2, 'use_adaptive':1, 'simps_err':1e-3} else: self.wiener_params = wiener_params wp = self.wiener_params self.wfpt = deepcopy(hddm.likelihoods.wfpt_like) self.wfpt.rv.wiener_params = wp self.kwargs = kwargs super(hddm.HDDM, self).__init__(data, include=include, **kwargs) def get_params(self): """Returns list of model parameters. """ # These boundaries are largely based on a meta-analysis of # reported fit values. # See: Matzke & Wagenmakers 2009 params = [Parameter('a', lower=.3, upper=4), Parameter('v', lower=-15., upper=15., init=0.), Parameter('t', lower=.1, upper=.9, init=.1), # Change lower to .2 as in MW09? Parameter('z', lower=.2, upper=0.8, init=.5, default=.5, optional=True), Parameter('V', lower=0., upper=3.5, default=0, optional=True), Parameter('Z', lower=0., upper=1.0, init=.1, default=0, optional=True), Parameter('T', lower=0., upper=0.8, init=.1, default=0, optional=True), Parameter('wfpt', is_bottom_node=True)] return params def get_subj_node(self, param): """Create and return a Normal (in case of an effect or drift-parameter) or Truncated Normal (otherwise) distribution for 'param' centered around param.group with standard deviation param.var and initialization value param.init. This is used for the individual subject distributions. """ if param.name.startswith('e') or param.name.startswith('v'): return pm.Normal(param.full_name, mu=param.group, tau=param.var**-2, plot=self.plot_subjs, trace=self.trace_subjs, value=param.init) else: return pm.TruncatedNormal(param.full_name, a=param.lower, b=param.upper, mu=param.group, tau=param.var**-2, plot=self.plot_subjs, trace=self.trace_subjs, value=param.init) def get_bottom_node(self, param, params): """Create and return the wiener likelihood distribution supplied in 'param'. 'params' is a dictionary of all parameters on which the data depends on (i.e. condition and subject). """ if param.name == 'wfpt': return self.wfpt(param.full_name, value=param.data['rt'].flatten(), v=params['v'], a=params['a'], z=self.get_node('z',params), t=params['t'], Z=self.get_node('Z',params), T=self.get_node('T',params), V=self.get_node('V',params), observed=True) else: raise KeyError, "Groupless parameter named %s not found." % param.name def plot_posterior_predictive(self, *args, **kwargs): if 'value_range' not in kwargs: kwargs['value_range'] = np.linspace(-5, 5, 100) kabuki.analyze.plot_posterior_predictive(self, *args, **kwargs) if __name__ == "__main__": import doctest doctest.testmod() HDDM-0.2/hddm/sandbox/0000775000102500010250000000000011727455647015223 5ustar wieckiwiecki00000000000000HDDM-0.2/hddm/sandbox/model_extended.py0000664000102500010250000000364611713025572020547 0ustar wieckiwiecki00000000000000import hddm from hddm.model import Base import pymc as pm from kabuki import Parameter class HDDMFullExtended(Base): def get_params(self): [self.include.add(param) for param in ['V','Z','T']] params = [Parameter('z_trls', False), Parameter('v_trls', False), Parameter('t_trls', False)] params += list(super(self.__class__, self).get_params()) return params def get_rootless_child(self, param, params): trials = len(param.data) if param.name == 'z_trls': return [hddm.likelihoods.CenterUniform(param.full_name+str(i), center=params['z'], width=params['Z']) for i in range(trials)] elif param.name == 'v_trls': return [pm.Normal(param.full_name+str(i), mu=params['v'], tau=params['V']**-2) for i in range(trials)] elif param.name == 't_trls': return [hddm.likelihoods.CenterUniform(param.full_name+str(i), center=params['t'], width=params['T']) for i in range(trials)] elif param.name == 'wfpt': return hddm.likelihoods.WienerSingleTrial(param.full_name, value=param.data['rt'], v=params['v_trls'], t=params['t_trls'], a=[params['a'] for i in range(trials)], z=params['z_trls'], observed=True) else: raise KeyError, "Rootless child node named %s not found." % param.name HDDM-0.2/hddm/sandbox/mpi.py0000664000102500010250000001426311713025572016351 0ustar wieckiwiecki00000000000000from __future__ import division import numpy as np import pymc as pm import matplotlib.pyplot as plt from copy import copy import matplotlib.pyplot as plt import numpy.lib.recfunctions as rec import os.path import os from ordereddict import OrderedDict import hddm import kabuki from mpi4py import MPI def create_tag_names(tag, chains=None): import multiprocessing if chains is None: chains = multiprocessing.cpu_count() tag_names = [] # Create copies of data and the corresponding db names for chain in range(chains): tag_names.append("db/mcmc%s%i.pickle"% (tag,chain)) return tag_names def controller(models, samples=200, burn=15, reps=5): process_list = range(1, MPI.COMM_WORLD.Get_size()) rank = MPI.COMM_WORLD.Get_rank() proc_name = MPI.Get_processor_name() status = MPI.Status() print "Controller %i on %s: ready!" % (rank, proc_name) task_iter = iter(models) idx = 0 results = [] while(True): status = MPI.Status() recv = MPI.COMM_WORLD.recv(source=MPI.ANY_SOURCE, tag=MPI.ANY_TAG, status=status) #print "Controller: received tag %i from %s" % (status.tag, status.source) if status.tag == 15: results.append(recv) if status.tag == 10 or status.tag == 15: try: model = task_iter.next() name = model['name'] del model['name'] #print "Controller: Sending task" MPI.COMM_WORLD.send(('%s_%i'%(name, idx), model), dest=status.source, tag=10) idx += 1 except StopIteration: # Task queue is empty print "Controller: Task queue is empty" print "Controller: Sending kill signal" MPI.COMM_WORLD.send((None, None), dest=status.source, tag=2) elif status.tag == 2: # Exit process_list.remove(status.source) print 'Process %i exited' % status.source print 'Processes left: ' + str(process_list) else: print 'Unkown tag %i with msg %s' % (status.tag, str(data)) if len(process_list) == 0: print "No processes left" break return results def worker(): from mpi4py import MPI rank = MPI.COMM_WORLD.Get_rank() proc_name = MPI.Get_processor_name() status = MPI.Status() print "Worker %i on %s: ready!" % (rank, proc_name) # Send ready MPI.COMM_WORLD.send([{'rank':rank, 'name':proc_name}], dest=0, tag=10) # Start main data loop while True: # Get some data #print "Worker %i on %s: waiting for data" % (rank, proc_name) name, model = MPI.COMM_WORLD.recv(source=0, tag=MPI.ANY_TAG, status=status) print "Worker %i on %s: received data, tag: %i" % (rank, proc_name, status.tag) if status.tag == 2: print "Worker %i on %s: received kill signal" % (rank, proc_name) MPI.COMM_WORLD.send([], dest=0, tag=2) return if status.tag == 10: # Run emergent #print "Worker %i on %s: Running %s" % (rank, proc_name, recv) retry = 0 while retry < 5: try: print "Running %s: %s\n" % (name, model) result = run_model(name, model) break except (pm.ZeroProbability, ValueError): retry +=1 if retry == 5: result = None print "Job %s failed" % recv #print("Worker %i on %s: finished one job" % (rank, proc_name)) MPI.COMM_WORLD.send(result, dest=0, tag=15) MPI.COMM_WORLD.send([], dest=0, tag=2) def run_model(name, model, load=False): import sys dbname = os.path.join('/','users', 'wiecki', 'scratch', 'theta', name+'.db') dbname = name+'.db' if model.has_key('effects_on'): mc = pm.MCMC(hddm.model.HDDMRegressor(**model).create(), db='pickle', dbname=dbname) else: mc = pm.MCMC(hddm.model.HDDM(**model).create(), db='pickle', dbname=dbname) if not load: try: os.remove(dbname) except OSError: pass mc.sample(30000, burn=10000, thin=2) mc.db.close() print "*************************************\n%s"%name kabuki.group.print_group_stats(mc.stats()) print "DIC: %f" % mc.dic print "logp: %f" % mc.logp sys.stdout.flush() return mc.stats() # else: # #print "Loading %s" %name # db = pm.database.hdf5.load(dbname) # return db def load_parallel_chains(model_class, data, tag, kwargs, chains=None, test_convergance=True, combine=True): tag_names = create_tag_names(tag, chains=chains) models = [] for tag_name in tag_names: model = model_class(data, **kwargs) model.mcmc_load_from_db(tag_name) models.append(model) if test_convergance: Rhat = test_chain_convergance(models) print Rhat if combine: m = combine_chains(models, model_class, data, kwargs) return m return models def combine_chains(models, model_class, data, kwargs): """Combine multiple model runs into one final model (make sure that chains converged).""" # Create model that will contain the other traces m = copy(models[0]) # Loop through models and combine chains for model in models[1:]: m._set_traces(m.group_params, mcmc_model=model.mcmc_model, add=True) m._set_traces(m.group_params_tau, mcmc_model=model.mcmc_model, add=True) m._set_traces(m.subj_params, mcmc_model=model.mcmc_model, add=True) return m def run_parallel_chains(model_class, data, tag, load=False, cpus=None, chains=None, **kwargs): import multiprocessing if cpus is None: cpus = multiprocessing.cpu_count() tag_names = create_tag_names(tag, chains=chains) # Parallel call if not load: rnds = np.random.rand(len(tag_names))*10000 pool = multiprocessing.Pool(processes=cpus) pool.map(call_mcmc, [(model_class, data, tag_name, rnd, kwargs) for tag_name,rnd in zip(tag_names, rnds)]) models = load_parallel_chains(model_class, data, tag_names, kwargs) return models HDDM-0.2/hddm/sandbox/model_gpu.py0000664000102500010250000000223111713025572017527 0ustar wieckiwiecki00000000000000import hddm from hddm.model import Base import pymc as pm from kabuki import Parameter class HDDMGPU(Base): """Experimental.""" def get_rootless_child(self, param, params): import pycuda.driver as cuda import pycuda.autoinit import pycuda.gpuarray as gpuarray if param.name == 'wfpt': data = param.data['rt'].flatten().astype(np.float32) data_gpu = gpuarray.to_gpu(data) out_gpu = gpuarray.empty_like(data_gpu) return hddm.likelihoods.WienerGPU(param.full_name, value=data_gpu, v = params['v'], V = self.get_node('V', params), a = params['a'], z = self.get_node('z',params), t = params['t'], out = out_gpu, observed=True) else: raise KeyError, "Rootless parameter named %s not found." % param.name HDDM-0.2/hddm/sandbox/model.py0000664000102500010250000004651011726176127016673 0ustar wieckiwiecki00000000000000import hddm from hddm.model import HDDM import pymc as pm from kabuki import Parameter from kabuki.distributions import scipy_stochastic import numpy as np from scipy import stats try: import wfpt_switch except: pass def wiener_like_multi(value, v, V, a, z, Z, t, T, multi=None): """Log-likelihood for the simple DDM""" return hddm.wfpt.wiener_like_multi(value, v, V, a, z, Z, t, T, .001, multi=multi) WienerMulti = pm.stochastic_from_dist(name="Wiener Simple Diffusion Process", logp=wiener_like_multi, dtype=np.float) class wfpt_switch_gen(stats.distributions.rv_continuous): err = 1e-4 evals = 100 def _argcheck(self, *args): return True def _logp(self, x, v, v_switch, V_switch, a, z, t, t_switch, T): """Log-likelihood for the simple DDM switch model""" # if t < T/2 or t_switch < T/2 or t<0 or t_switch<0 or T<0 or a<=0 or z<=0 or z>=1 or T>.5: # print "Condition not met" logp = wfpt_switch.wiener_like_antisaccade_precomp(x, v, v_switch, V_switch, a, z, t, t_switch, T, self.err, evals=self.evals) return logp def _pdf(self, x, v, v_switch, V_switch, a, z, t, t_switch, T): if np.isscalar(x): out = wfpt_switch.pdf_switch(np.array([x]), v, v_switch, V_switch, a, z, t, t_switch, T, 1e-4) else: out = np.empty_like(x) for i in xrange(len(x)): out[i] = wfpt_switch.pdf_switch(np.array([x[i]]), v[i], v_switch[i], V_switch[i], a[i], z[i], t[i], t_switch[i], T[i], 1e-4) return out def _rvs(self, v, v_switch, V_switch, a, z, t, t_switch, T): all_rts_generated=False while(not all_rts_generated): out = hddm.generate.gen_antisaccade_rts({'v':v, 'z':z, 't':t, 'a':a, 'v_switch':v_switch, 'V_switch':V_switch, 't_switch':t_switch, 'Z':0, 'V':0, 'T':T}, samples_anti=self._size, samples_pro=0)[0] if (len(out) == self._size): all_rts_generated=True return hddm.utils.flip_errors(out)['rt'] wfpt_switch_like = scipy_stochastic(wfpt_switch_gen, name='wfpt switch', longname="""Wiener first passage time likelihood function""", extradoc="""Wiener first passage time (WFPT) likelihood function of the Ratcliff Drift Diffusion Model (DDM). Models two choice decision making tasks as a drift process that accumulates evidence across time until it hits one of two boundaries and executes the corresponding response. Implemented using the Navarro & Fuss (2009) method. Parameters: *********** v: drift-rate a: threshold z: bias [0,1] t: non-decision time References: *********** Fast and accurate calculations for first-passage times in Wiener diffusion models Navarro & Fuss - Journal of Mathematical Psychology, 2009 - Elsevier """) class HDDMSwitch(HDDM): def __init__(self, data, init=True, **kwargs): super(self.__class__, self).__init__(data, **kwargs) if 'instruct' not in self.data.dtype.names: raise AttributeError, 'data has to contain a field name instruct.' def get_params(self): params = [Parameter('vpp', lower=-20, upper=20.), Parameter('vcc', lower=-20, upper=20.), Parameter('a', lower=.5, upper=4.5), Parameter('t', lower=0., upper=.5, init=0.05), Parameter('tcc', lower=0.01, upper=1.0), Parameter('T', lower=0, upper=.5, init=.1, default=0, optional=True), Parameter('Vcc', lower=0, upper=2., default=0, optional=True), Parameter('wfpt_anti', is_bottom_node=True), Parameter('wfpt_pro', is_bottom_node=True)] return params def get_subj_node(self, param): """Create and return a Normal (in case of an effect or drift-parameter) or Truncated Normal (otherwise) distribution for 'param' centered around param.group with standard deviation param.var and initialization value param.init. This is used for the individual subject distributions. """ return pm.TruncatedNormal(param.full_name, a=param.lower, b=param.upper, mu=param.group, tau=param.var**-2, plot=self.plot_subjs, trace=self.trace_subjs, value=param.init) def get_var_node(self, param): #return pm.Gamma(param.full_name, alpha=.3, beta=.3) return pm.Uniform(param.full_name, lower=1e-7, upper=10) def get_bottom_node(self, param, params): if param.name == 'wfpt_anti': data = param.data[param.data['instruct'] == 1] if len(data) == 0: return None return wfpt_switch_like(param.full_name, value=data['rt'], v=params['vpp'], v_switch=params['vcc'], V_switch=self.get_node('Vcc',params), a=params['a'], z=.5, t=params['t'], t_switch=params['tcc'], T=self.get_node('T',params), observed=True) elif param.name == 'wfpt_pro': data = param.data[param.data['instruct'] == 0] if len(data) == 0: return None return hddm.likelihoods.wfpt_like(param.full_name, value=data['rt'], v=params['vpp'], V=0, a=params['a'], z=.5, Z=0., t=params['t'], T=0., observed=True) else: raise TypeError, "Parameter named %s not found." % param.name class HDDMRegressor(HDDM): def __init__(self, data, effects_on=None, use_root_for_effects=False, **kwargs): """Hierarchical Drift Diffusion Model analyses for Cavenagh et al, IP. :Arguments: data : numpy.recarray structured numpy array containing columns: subj_idx, response, RT, theta, dbs :Optional: effects_on : dict theta and dbs effect these DDM parameters. depends_on : dict separate stimulus distributions for these parameters. :Example: >>> import hddm >>> data, params = hddm.generate.gen_correlated_rts() >>> model = hddm.sandbox.HDDMRegressor(data, effects_on={'a':'cov'}) >>> model.sample(5000) """ self.effects_on = effects_on self.use_root_for_effects = use_root_for_effects super(self.__class__, self).__init__(data, **kwargs) def get_params(self): params = [] # Add rootless nodes for effects for effect_on, col_names in self.effects_on.iteritems(): if type(col_names) is str or (type(col_names) is list and len(col_names) == 1): if type(col_names) is list: col_names = col_names[0] params.append(Parameter('e_%s_%s'%(col_names, effect_on), lower=-3., upper=3., init=0, create_subj_nodes=not self.use_root_for_effects)) params.append(Parameter('e_inst_%s_%s'%(col_names, effect_on), is_bottom_node=True, vars={'col_name':col_names, 'effect_on':effect_on, 'e':'e_%s_%s'%(col_names, effect_on)})) elif len(col_names) == 2: for col_name in col_names: params.append(Parameter('e_%s_%s'%(col_name, effect_on), lower=-3., upper=3., init=0, create_subj_nodes=not self.use_root_for_effects)) params.append(Parameter('e_inter_%s_%s_%s'%(col_names[0], col_names[1], effect_on), lower=-3., upper=3., init=0, create_subj_nodes=not self.use_root_for_effects)) params.append(Parameter('e_inst_%s_%s_%s'%(col_names[0], col_names[1], effect_on), is_bottom_node=True, vars={'col_name0': col_names[0], 'col_name1': col_names[1], 'effect_on': effect_on, 'e1':'e_%s_%s'%(col_names[0], effect_on), 'e2':'e_%s_%s'%(col_names[1], effect_on), 'inter':'e_inter_%s_%s_%s'%(col_names[0], col_names[1], effect_on)})) else: raise NotImplementedError, "Only 1 or 2 regressors allowed per variable." params += super(self.__class__, self).get_params() return params def get_bottom_node(self, param, params): """Generate the HDDM.""" if param.name.startswith('e_inst'): if not param.vars.has_key('inter'): # No interaction if param.vars['effect_on'] == 't': func = effect1_nozero else: func = effect1 return pm.Deterministic(func, param.full_name, param.full_name, parents={'base': params[param.vars['effect_on']], 'e1': params[param.vars['e']], 'data': param.data[param.vars['col_name']]}, trace=False, plot=self.plot_subjs) else: return pm.Deterministic(effect2, param.full_name, param.full_name, parents={'base': params[param.vars['effect_on']], 'e1': params[param.vars['e1']], 'e2': params[param.vars['e2']], 'e_inter': params[param.vars['inter']], 'data_e1': param.data[param.vars['col_name0']], 'data_e2': param.data[param.vars['col_name1']]}, trace=False) for effect_on, col_name in self.effects_on.iteritems(): if type(col_name) is str: params[effect_on] = params['e_inst_%s_%s'%(col_name, effect_on)] else: params[effect_on] = params['e_inst_%s_%s_%s'%(col_name[0], col_name[1], effect_on)] else: model = WienerMulti(param.full_name, value=param.data['rt'], v=params['v'], V=self.get_node('V', params), a=params['a'], z=self.get_node('z', params), Z=self.get_node('Z', params), t=params['t'], T=self.get_node('T', params), multi=self.effects_on.keys(), observed=True) return model def effect1(base, e1, data): """Effect distribution. """ return base + e1 * data def effect1_nozero(base, e1, data): """Effect distribution where values <0 will be set to 0. """ value = base + e1 * data value[value < 0] = 0. value[value > .4] = .4 return value def effect2(base, e1, e2, e_inter, data_e1, data_e2): """2-regressor effect distribution """ return base + data_e1*e1 + data_e2*e2 + data_e1*data_e2*e_inter class HDDMContUnif(HDDM): """Contaminant HDDM Uniform class Outliers are modeled using a uniform distribution over responses and reaction times. :Optional: init : bool Use EZ to initialize parameters (default: True) """ def __init__(self, *args, **kwargs): super(hddm.model.HDDMContUnif, self).__init__(*args, **kwargs) self.cont_res = None self.t_min = 0 self.t_max = max(abs(self.data['rt'])) wp = self.wiener_params self.wfpt = hddm.likelihoods.general_WienerCont(err=wp['err'], nT=wp['nT'], nZ=wp['nZ'], use_adaptive=wp['use_adaptive'], simps_err=wp['simps_err']) def get_params(self): params = super(hddm.model.HDDMContUnif, self).get_params() params = params[:-1] + \ [Parameter('pi', lower=0.01, upper=0.1), Parameter('x', is_bottom_node=True), Parameter('wfpt', is_bottom_node=True)] return params def get_bottom_node(self, param, params): if param.name == 'wfpt': return self.wfpt(param.full_name, value=param.data['rt'].flatten(), cont_x=params['x'], v=params['v'], a=params['a'], z=self.get_node('z',params), t=params['t'], Z=self.get_node('Z',params), T=self.get_node('T',params), V=self.get_node('V',params), t_min=self.t_min, t_max=self.t_max, observed=True) elif param.name == 'x': rts = param.data['rt'] outlier = np.empty(rts.shape, dtype=np.bool) outlier[np.abs(rts) < params['t'].value] = True outlier[np.abs(rts) >= params['t'].value] = False return pm.Bernoulli(param.full_name, p=params['pi'], size=len(param.data['rt']), plot=False, value=outlier) else: raise KeyError, "Groupless subj parameter %s not found" % param.name def cont_report(self, cont_threshold=0.5, plot=True): """Create conaminate report :Arguments: cont_threshold : float the threshold tthat define an outlier (default: 0.5) plot : bool should the result be plotted (default: True) :Returns: report : dict * cont_idx : index of outlier above threshold * probs : Probability RT is an outlier * rts : RTs of outliers """ hm = self data_dep = hm._get_data_depend() self.cont_res = {} if self.is_group_model: subj_list = self._subjs else: subj_list = [0] conds = hm.params_dict['x'].subj_nodes.keys() #loop over subjects for subj_idx, subj in enumerate(subj_list): n_cont = 0 rts = np.empty(0) probs = np.empty(0) cont_idx = np.empty(0, np.int) print "#$#$#$# outliers for subject %s #$#$#$#" % subj #loop over conds for cond in conds: print "*********************" print "looking at %s" % cond nodes = hm.params_dict['x'].subj_nodes[cond] if self.is_group_model: node = nodes[subj_idx] else: node = nodes m = np.mean(node.trace(), 0) #look for outliers with high probabilty idx = np.where(m > cont_threshold)[0] n_cont += len(idx) if idx.size > 0: print "found %d probable outliers in %s" % (len(idx), cond) wfpt = list(node.children)[0] data_idx = [x for x in data_dep if x[3]==cond][0][0]['data_idx'] for i_cont in range(len(idx)): print "rt: %8.5f prob: %.2f" % (wfpt.value[idx[i_cont]], m[idx[i_cont]]) cont_idx = np.r_[cont_idx, data_idx[idx]] rts = np.r_[rts, wfpt.value[idx]] probs = np.r_[probs, m[idx]] #plot outliers if plot: plt.figure() mask = np.ones(len(wfpt.value), dtype=bool) mask[idx] = False plt.plot(wfpt.value[mask], np.zeros(len(mask) - len(idx)), 'b.') plt.plot(wfpt.value[~mask], np.zeros(len(idx)), 'ro') plt.title(wfpt.__name__) plt.xlabel('RTs for lower (negative) and upper (positive) boundary responses.') #report the next higest probability outlier next_outlier = max(m[m < cont_threshold]) print "probability of the next most probable outlier: %.2f" % next_outlier print "!!!!!**** %d probable outliers were found in the data ****!!!!!" % n_cont single_cont_res = {} single_cont_res['cont_idx'] = cont_idx single_cont_res['rts'] = rts single_cont_res['probs'] = probs if self.is_group_model: self.cont_res[subj] = single_cont_res else: self.cont_res = single_cont_res if plot: plt.show() return self.cont_res def remove_conts(self, cont_threshold=.5): """ Return the data without the contaminants. :Arguments: cutoff : float the probablity that defines an outlier (deafult 0.5) :Returns: cleaned_data : ndarray """ # Generate cont_report if not self.cont_res: self.cont_report(cont_threshold = cont_threshold) new_data = [] if self.is_group_model: subj_list = self._subjs else: subj_list = [0] # Loop over subjects and append cleaned data for subj in subj_list: if self.is_group_model: cont_res = self.cont_res[subj] data_subj = self.data[self.data['subj_idx'] == subj] else: cont_res = self.cont_res data_subj = self.data idx = np.ones(len(data_subj), bool) idx[cont_res['cont_idx'][cont_res['probs'] >= cont_threshold]] = 0 new_data.append(data_subj[idx]) data_all = np.concatenate(new_data) data_all['rt'] = np.abs(data_all['rt']) return data_all HDDM-0.2/hddm/sandbox/__init__.py0000664000102500010250000000002411713025572017311 0ustar wieckiwiecki00000000000000from model import * HDDM-0.2/hddm/sandbox/model_lba.py0000664000102500010250000000461611713025572017503 0ustar wieckiwiecki00000000000000import hddm from hddm.model import Base import pymc as pm from kabuki import Parameter class HLBA(Base): param_names = (('a',True), ('z',True), ('t',True), ('V',True), ('v0',True), ('v1',True), ('lba',False)) def __init__(self, data, model_type=None, trace_subjs=True, normalize_v=True, no_bias=True, fix_sv=None, init=False, exclude=None, **kwargs): super(self.__class__, self).__init__(data, **kwargs) # LBA model self.normalize_v = normalize_v self.init_params = {} self.param_ranges = {'a_lower': .2, 'a_upper': 4., 'v_lower': 0.1, 'v_upper': 3., 'z_lower': .0, 'z_upper': 2., 't_lower': .05, 't_upper': 2., 'V_lower': .2, 'V_upper': 2.} if self.normalize_v: self.param_ranges['v_lower'] = 0. self.param_ranges['v_upper'] = 1. def get_rootless_child(self, param, params): return hddm.likelihoods.LBA(param.full_name, value=param.data['rt'], a=params['a'], z=params['z'], t=params['t'], v0=params['v0'], v1=params['v1'], V=params['V'], normalize_v=self.normalize_v, observed=True) def get_root_node(self, param): """Create and return a prior distribution for [param]. [tag] is used in case of dependent parameters. """ if param.name == 'V' and self.fix_sv is not None: # drift rate variability return pm.Lambda(param.full_name, lambda x=self.fix_sv: x) else: return super(self.__class__, self).get_root_param(self, param) def get_child_node(self, param, plot=False): if param.name.startswith('V') and self.fix_sv is not None: return pm.Lambda(param.full_name, lambda x=param.root: x, plot=plot, trace=self.trace_subjs) else: return super(self.__class__, self).get_child_node(param, plot=plot) HDDM-0.2/hddm/diag.py0000664000102500010250000002166411720307731015033 0ustar wieckiwiecki00000000000000from __future__ import division from copy import copy import numpy as np import pymc as pm np.seterr(divide='ignore') from sys import stdout import hddm from scipy.stats import scoreatpercentile from time import time from operator import attrgetter def check_model(model, params_true, assert_=False, conf_interval = 95): """calculate the posterior estimate error if hidden parameters are known (e.g. when simulating data).""" print "checking estimation with %d confidence interval" % conf_interval fail = False nodes = sorted(model.stochastics, key=attrgetter('__name__')) nodes = filter(lambda x:x.shape == (), nodes) for node in nodes: if node.__name__ not in params_true: continue # Skip non-existent params name = node.__name__ trace = model.db.trace(name)[:] est = np.mean(trace) truth = params_true[name] lb = (50 - conf_interval/2.) lb_score = scoreatpercentile(trace, lb) ub = (50 + conf_interval/2.) ub_score = scoreatpercentile(trace, ub) fell = np.sum(truth > trace)*1./len(trace) * 100 if lb_score==ub_score: fell = 50 print "%s: Truth: %6.3f, Estimated: %6.3f, lb: %6.3f, ub: %6.3f, fell in: %.2f" % \ (name, truth, est, lb_score, ub_score, fell) if (fell < lb) or (fell > ub): fail = True print "the true value of %s is outside of the confidence interval !*!*!*!*!*!*!" % name if assert_: assert (fail==False) ok = not fail return ok def check_rejection(model, assert_=True): """check if the rejection ratio is not too high""" for node in model.stochastics: name = node.__name__ trace = node.trace()[:] rej = np.sum(np.diff(trace)==0) rej_ratio = rej*1.0/len(trace) print "rejection ratio for %s: %.3f" %(name, rej_ratio) if (rej_ratio < 0.5) or (rej_ratio > 0.8): msg = "%s still need to be tuned" % name if assert_: assert 1==0, msg else: print msg def test_params_on_data(params, data, include=(), depends_on = None, conf_interval = 95): thin = 1 samples = 10000 burn = 10000 n_iter = burn + samples*thin stdout.flush() if depends_on is None: depends_on = {} if 'pi' in include or 'gamma' in include: m_hddm = hddm.HDDMContaminant(data, bias=True, depends_on=depends_on) else: m_hddm = hddm.HDDM(data, bias=True, include=include, depends_on=depends_on) model = m_hddm.mcmc() #[model.use_step_method(pm.Metropolis, x,proposal_sd=0.1) for x in model.stochastics] i_t = time() model.sample(n_iter, burn=burn, thin=thin) print "sampling took: %.2f seconds" % (time() - i_t) ok = True if check_model(model, params, assert_=False, conf_interval = conf_interval)==False: print "model checking failed. running again" stdout.flush() model.sample(n_iter, burn=burn, thin=thin) if check_model(model, params, assert_=False, conf_interval = conf_interval)==False: print "model checking failed again !!!!!!!!!!!!!!!!!!!!!!!" ok = False res = {} res['params'] = params res['data'] = data res['mc'] = model check_rejection(model, assert_ = False) check_correl(model) stdout.flush() return ok, res def run_accuracy_test(nTimes=20, include=(), stop_when_fail = True): """ run accuracy test nTime times""" n_data = 300 for i_time in range(nTimes): params = hddm.generate.gen_rand_params(include) data,temp = hddm.generate.gen_rand_data(n_data, params) positive = sum(data['response']) print "generated %d data_points (%d positive %d negative)" % (len(data), positive, len(data) - positive) print "testing params: a:%.3f, t:%.3f, v:%.3f, z: %.3f, T: %.3f, V: %.3f Z: %.3f" \ % (params['a'], params['t'], params['v'], params['z'], params['T'], params['V'], params['Z']) ok, res = test_params_on_data(params, data, include=include) if stop_when_fail and not ok: return res return {} def str_params(params): s = '' keys = params.keys() keys.sort(reverse=True) for name in keys: s = s + "%s: %.3f, " % (name, params[name]) s = s[:-2] + "\n" return s def gen_cond_data_and_params(n_data, n_conds = 3, include = ()): params = hddm.generate.gen_rand_params(include) params_set = [None]*n_conds params_true = copy(params) all_v = np.linspace(min(0,params['v']/2) , max(params['v']*2, 3), n_conds) del params_true['v'] for i in range(n_conds): params_set[i] = copy(params) params_set[i]['v'] = all_v[i] params_true['v(%d,)'%i] = all_v[i] cond_data, temp = hddm.generate.gen_rand_cond_data(params_set, samples_per_cond=int(n_data/n_conds)) positive = sum(cond_data['response']) print "generated %d data_points (%d positive %d negative)" % \ (len(cond_data), positive, len(cond_data) - positive) print "used params: %s" % str_params(params_true) stdout.flush() return cond_data, params_true def run_simple_test(nTimes=20, stop_when_fail = False): return run_accuracy_test(nTimes) def check_correl(model): nodes = filter(lambda x:x.shape == (), model.stochastics) threshold = 0.05 fail = False for node in nodes: t_lag = np.inf for lag in range(1,min(101, len(node.trace()[:])//2)): corr = pm.diagnostics.autocorr(node.trace()[:], lag) if corr <= threshold: t_lag= lag break if t_lag < np.inf: fail = True print "%s: correlation drop under %f after %d steps" %(node.__name__ , threshold, t_lag) else: print "%s: correlation don't drop under %f!!!!" %(node.__name__ , threshold) ok = not fail return ok def test_acc_full_intrp(include = (), n_conds = 6, use_db=False): from numpy.random import rand burn = 10000 thin = 1 n_samples = 10000 n_iter = n_samples*thin all_wp = [] all_wp = all_wp + [{'err': 1e-5, 'nT':3, 'nZ':3, 'use_adaptive':1, 'simps_err':1e-5}] all_wp = all_wp + [{'err': 1e-5, 'nT':2, 'nZ':2, 'use_adaptive':1, 'simps_err':1e-4}] all_wp = all_wp + [{'err': 1e-4, 'nT':2, 'nZ':2, 'use_adaptive':1, 'simps_err':1e-3}] initial_params = hddm.generate.gen_rand_params(include=include) full_params = copy(initial_params) params_set = [None]*n_conds v_0 = rand() all_v = np.linspace(v_0, max(4,v_0*n_conds), n_conds) for j in range(n_conds): params_set[j] = copy(initial_params) params_set[j]['v'] = all_v[j] full_params['v(%d,)'%j] = params_set[j]['v'] data = hddm.generate.gen_rand_cond_data(params_set, samples_per_cond=150) print "Using the following params: \n %s" % str_params(full_params) i_res={} i_res['params'] = copy(full_params) i_res['all_wp'] = all_wp i_res['data'] = data i_res['sample_time'] = [None]*len(all_wp) i_res['init_time'] = [None]*len(all_wp) i_res['burn_time'] = [None]*len(all_wp) i_res['stats'] = [None]*len(all_wp) i_res['logp'] = [None]*len(all_wp) i_res['dbname'] = [None]*len(all_wp) i_res['mc'] = [None]*len(all_wp) for i_params in range(len(all_wp)): print "working on model %d" % i_params model = hddm.model.HDDM(data, bias=True, wiener_params=all_wp[i_params], include = include, depends_on = {'v':['cond']})#, init_value=params) i_t = time() if use_db: dbname = 'speed.' + str(time()) + '.db' i_res['dbname'][i_params] = dbname[:] else: dbname = None nodes = model.create() mc = pm.MCMC(nodes) i_res['mc'][i_params] = mc [mc.use_step_method(pm.Metropolis, x,proposal_sd=0.5) for x in mc.stochastics] i_t = time() mc.sample(burn+1, burn) d_time = time() - i_t; i_res['burn_time'][i_params] = d_time print "burn phase took %f secs" % d_time i_t = time() mc.sample(n_iter, 0, thin=thin) d_time = time() - i_t; i_res['sample_time'][i_params] = d_time print "sampling took in %f secs" % d_time stdout.flush() check_model(mc, full_params, assert_=False, conf_interval = 95) check_rejection(mc, assert_=False) check_correl(mc) if dbname is not None: mc.db.commit() mc.db.close() return i_res def check_geweke(model, assert_=True): # Test for convergence using geweke method for param in model.group_params.itervalues(): geweke = np.array(pm.geweke(param)) if assert_: assert (np.any(np.abs(geweke[:,1]) < 2)), 'Chain of %s not properly converged'%param return False else: if np.any(np.abs(geweke[:,1]) > 2): print "Chain of %s not properly converged" % param return False return True HDDM-0.2/hddm/__init__.py0000664000102500010250000000053711726176127015673 0ustar wieckiwiecki00000000000000#!/usr/bin/python __docformat__ = 'restructuredtext' import likelihoods import generate import utils import sandbox from model import * import wfpt try: import wfpt_switch except: pass from kabuki.utils import load_csv, save_csv try: from IPython.core.debugger import Tracer; debug_here = Tracer() except: def debug_here(): pass HDDM-0.2/hddm/utils.py0000664000102500010250000006776411727151500015300 0ustar wieckiwiecki00000000000000from __future__ import division import numpy as np import matplotlib.pyplot as plt import pymc as pm import hddm import sys import table_print from scipy.stats import scoreatpercentile from numpy import array, zeros, ones, empty from copy import deepcopy from time import time def flip_errors(data): """Flip sign for lower boundary responses. :Arguments: data : numpy.recarray Input array with at least one column named 'RT' and one named 'response' :Returns: data : numpy.recarray Input array with RTs sign flipped where 'response' == 0 """ # Check if data is already flipped if np.any(data['rt'] < 0): return data # Copy data data = np.array(data) # Flip sign for lower boundary response idx = data['response'] == 0 data['rt'][idx] = -data['rt'][idx] return data def histogram(a, bins=10, range=None, normed=False, weights=None, density=None): """ Compute the histogram of a set of data. Parameters ---------- a : array_like Input data. The histogram is computed over the flattened array. bins : int or sequence of scalars, optional If `bins` is an int, it defines the number of equal-width bins in the given range (10, by default). If `bins` is a sequence, it defines the bin edges, including the rightmost edge, allowing for non-uniform bin widths. range : (float, float), optional The lower and upper range of the bins. If not provided, range is simply ``(a.min(), a.max())``. Values outside the range are ignored. normed : bool, optional This keyword is deprecated in Numpy 1.6 due to confusing/buggy behavior. It will be removed in Numpy 2.0. Use the density keyword instead. If False, the result will contain the number of samples in each bin. If True, the result is the value of the probability *density* function at the bin, normalized such that the *integral* over the range is 1. Note that this latter behavior is known to be buggy with unequal bin widths; use `density` instead. weights : array_like, optional An array of weights, of the same shape as `a`. Each value in `a` only contributes its associated weight towards the bin count (instead of 1). If `normed` is True, the weights are normalized, so that the integral of the density over the range remains 1 density : bool, optional If False, the result will contain the number of samples in each bin. If True, the result is the value of the probability *density* function at the bin, normalized such that the *integral* over the range is 1. Note that the sum of the histogram values will not be equal to 1 unless bins of unity width are chosen; it is not a probability *mass* function. Overrides the `normed` keyword if given. Returns ------- hist : array The values of the histogram. See `normed` and `weights` for a description of the possible semantics. bin_edges : array of dtype float Return the bin edges ``(length(hist)+1)``. See Also -------- histogramdd, bincount, searchsorted, digitize Notes ----- All but the last (righthand-most) bin is half-open. In other words, if `bins` is:: [1, 2, 3, 4] then the first bin is ``[1, 2)`` (including 1, but excluding 2) and the second ``[2, 3)``. The last bin, however, is ``[3, 4]``, which *includes* 4. Examples -------- >>> np.histogram([1, 2, 1], bins=[0, 1, 2, 3]) (array([0, 2, 1]), array([0, 1, 2, 3])) >>> np.histogram(np.arange(4), bins=np.arange(5), density=True) (array([ 0.25, 0.25, 0.25, 0.25]), array([0, 1, 2, 3, 4])) >>> np.histogram([[1, 2, 1], [1, 0, 1]], bins=[0,1,2,3]) (array([1, 4, 1]), array([0, 1, 2, 3])) >>> a = np.arange(5) >>> hist, bin_edges = np.histogram(a, density=True) >>> hist array([ 0.5, 0. , 0.5, 0. , 0. , 0.5, 0. , 0.5, 0. , 0.5]) >>> hist.sum() 2.4999999999999996 >>> np.sum(hist*np.diff(bin_edges)) 1.0 """ a = np.asarray(a) if weights is not None: weights = np.asarray(weights) if np.any(weights.shape != a.shape): raise ValueError( 'weights should have the same shape as a.') weights = weights.ravel() a = a.ravel() if (range is not None): mn, mx = range if (mn > mx): raise AttributeError( 'max must be larger than min in range parameter.') if not np.iterable(bins): if np.isscalar(bins) and bins < 1: raise ValueError("`bins` should be a positive integer.") if range is None: if a.size == 0: # handle empty arrays. Can't determine range, so use 0-1. range = (0, 1) else: range = (a.min(), a.max()) mn, mx = [mi + 0.0 for mi in range] if mn == mx: mn -= 0.5 mx += 0.5 bins = np.linspace(mn, mx, bins + 1, endpoint=True) else: bins = np.asarray(bins) if (np.diff(bins) < 0).any(): raise AttributeError( 'bins must increase monotonically.') # Histogram is an integer or a float array depending on the weights. if weights is None: ntype = int else: ntype = weights.dtype n = np.zeros(bins.shape, ntype) block = 65536 if weights is None: for i in np.arange(0, len(a), block): sa = np.sort(a[i:i + block]) n += np.r_[sa.searchsorted(bins[:-1], 'left'), \ sa.searchsorted(bins[-1], 'right')] else: zero = np.array(0, dtype=ntype) for i in np.arange(0, len(a), block): tmp_a = a[i:i + block] tmp_w = weights[i:i + block] sorting_index = np.argsort(tmp_a) sa = tmp_a[sorting_index] sw = tmp_w[sorting_index] cw = np.concatenate(([zero, ], sw.cumsum())) bin_index = np.r_[sa.searchsorted(bins[:-1], 'left'), \ sa.searchsorted(bins[-1], 'right')] n += cw[bin_index] n = np.diff(n) if density is not None: if density: db = np.array(np.diff(bins), float) return n / db / n.sum(), bins else: return n, bins else: # deprecated, buggy behavior. Remove for Numpy 2.0 if normed: db = np.array(np.diff(bins), float) return n / (n * db).sum(), bins else: return n, bins def parse_config_file(fname, map=True, mcmc=False, data=None, samples=None, burn=None, thin=None, only_group_stats=False, plot=True, verbose=False): import kabuki import os.path if not os.path.isfile(fname): raise ValueError("%s could not be found." % fname) import ConfigParser config = ConfigParser.ConfigParser() config.optionxform = str config.read(fname) ##################################################### # Parse config file if data is not None: data_fname = data else: try: data_fname = config.get('model', 'data') except (ConfigParser.NoOptionError, ConfigParser.NoSectionError): print "ERROR: No data file specified. Either provide data file as an argument to hddmfit or in the model specification" sys.exit(-1) if not os.path.exists(data_fname): raise IOError("Data file %s not found." % data_fname) data = np.recfromcsv(data_fname) model_name = os.path.splitext(data_fname)[0] try: include = config.get('model', 'include') except (ConfigParser.NoOptionError, ConfigParser.NoSectionError): include = () try: is_group_model = config.getboolean('model', 'is_group_model') except (ConfigParser.NoOptionError, ConfigParser.NoSectionError): is_group_model = None try: bias = config.getboolean('model', 'bias') except (ConfigParser.NoOptionError, ConfigParser.NoSectionError): bias = False try: db = config.get('mcmc', 'db') except (ConfigParser.NoOptionError, ConfigParser.NoSectionError): db = 'ram' try: dbname = config.get('mcmc', 'dbname') except (ConfigParser.NoOptionError, ConfigParser.NoSectionError): dbname = None # MCMC values if samples is None: try: samples = config.getint('mcmc', 'samples') except (ConfigParser.NoOptionError, ConfigParser.NoSectionError): samples = 10000 if burn is None: try: burn = config.getint('mcmc', 'burn') except (ConfigParser.NoOptionError, ConfigParser.NoSectionError): burn = 5000 if thin is None: try: thin = config.getint('mcmc', 'thin') except (ConfigParser.NoOptionError, ConfigParser.NoSectionError): thin = 2 group_params = ['v', 'V', 'a', 'z', 'Z', 't', 'T'] # Get depends depends = {} for param_name in group_params: try: # Multiple depends can be listed (separated by a comma) depends[param_name] = config.get('depends', param_name).split(',') except (ConfigParser.NoOptionError, ConfigParser.NoSectionError): pass print "Creating model..." m = hddm.HDDM(data, include=include, bias=bias, is_group_model=is_group_model, depends_on=depends) if map: print "Finding good initial values..." m.map() m.mcmc().sample(samples, burn=burn, thin=thin, verbose=verbose) if only_group_stats: print kabuki.analyze.print_group_stats(m.mc.stats()) else: print kabuki.analyze.print_stats(m.mc.stats()) print "logp: %f" % m.mc.logp print "DIC: %f" % m.mc.dic if plot: m.plot_posteriors() print "Plotting posterior predictive..." kabuki.analyze.plot_posterior_predictive(m, value_range=np.linspace(-3, 3, 100), savefig=True) return m def EZ_subjs(data): params = {} # Estimate EZ group parameters v, a, t = EZ_data(data) params['v'] = v params['a'] = a params['t'] = t params['z'] = .5 # Estimate EZ parameters for each subject try: for subj in np.unique(data['subj_idx']): try: v, a, t = EZ_data(data[data['subj_idx'] == subj]) params['v_%i' % subj] = v params['a_%i' % subj] = a params['t_%i' % subj] = t # t-.2 if t-.2>0 else .1 params['z_%i' % subj] = .5 except ValueError: # Subject either had 0%, 50%, or 100% correct, which does not work # with easy. But we can deal with that by just not initializing the # parameters for that one model. params['v_%i' % subj] = None params['a_%i' % subj] = None params['t_%i' % subj] = None params['z_%i' % subj] = None except ValueError: # Data array has no subj_idx -> ignore. pass return params def EZ_param_ranges(data, range_=1.): v, a, t = EZ_data(data) z = .5 param_ranges = {'a_lower': a - range_, 'a_upper': a + range_, 'z_lower': np.min(z - range_, 0), 'z_upper': np.max(z + range_, 1), 't_lower': t - range_ if (t - range_) > 0 else 0., 't_upper': t + range_, 'v_lower': v - range_, 'v_upper': v + range_} return param_ranges def EZ_data(data, s=1): """ Calculate Wagenmaker's EZ-diffusion statistics on data. :Arguments: data : numpy.array Data array with reaction time data. Correct RTs are positive, incorrect RTs are negative. s : float Scaling parameter (default=1) :Returns: (v, a, ter) : tuple drift-rate, threshold and non-decision time :See Also: EZ """ try: rt = data['rt'] except ValueError: rt = data # Compute statistics over data idx_correct = rt > 0 mrt = np.mean(rt[idx_correct]) vrt = np.var(rt[idx_correct]) pc = np.sum(idx_correct) / np.float(rt.shape[0]) # Calculate EZ estimates. return EZ(pc, vrt, mrt, s) def EZ(pc, vrt, mrt, s=1): """ Calculate Wagenmaker's EZ-diffusion statistics. :Parameters: pc : float probability correct. vrt : float variance of response time for correct decisions (only!). mrt : float mean response time for correct decisions (only!). s : float scaling parameter. Default s=1. :Returns: (v, a, ter) : tuple drift-rate, threshold and non-decision time The error RT distribution is assumed identical to the correct RT distrib. Edge corrections are required for cases with Pc=0 or Pc=1. (Pc=.5 is OK) :Assumptions: * The error RT distribution is identical to the correct RT distrib. * z=.5 -- starting point is equidistant from the response boundaries * sv=0 -- across-trial variability in drift rate is negligible * sz=0 -- across-trial variability in starting point is negligible * st=0 -- across-trial range in nondecision time is negligible :Reference: Wagenmakers, E.-J., van der Maas, H. Li. J., & Grasman, R. (2007). An EZ-diffusion model for response time and accuracy. Psychonomic Bulletin & Review, 14 (1), 3-22. :Example: >>> EZ(.802, .112, .723, s=.1) (0.099938526231301769, 0.13997020267583737, 0.30002997230248141) :See Also: EZ_data """ if (pc == 0 or pc == .5 or pc == 1): raise ValueError('Probability correct is either 0%, 50% or 100%') s2 = s ** 2 logit_p = np.log(pc / (1 - pc)) # Eq. 7 x = ((logit_p * (pc**2 * logit_p - pc * logit_p + pc - .5)) / vrt) v = np.sign(pc - .5) * s * x**.25 # Eq 5 a = (s2 * logit_p) / v y = (-v * a) / s2 # Eq 9 mdt = (a / (2 * v)) * ((1 - np.exp(y)) / (1 + np.exp(y))) # Eq 8 ter = mrt - mdt return (v, a, ter) def hddm_parents_trace(model, obs_node, idx): """Return the parents' value of an wfpt node in index 'idx' (the function is used by ppd_test) """ model.params_include.keys() params = {'a': 0, 'v': 0, 't':0, 'z': 0.5, 'Z': 0, 'T': 0, 'V': 0} if not np.isscalar(idx): for (key, value) in params.iteritems(): params[key] = np.ones(len(idx)) * value #example for local_name: a,v,t,z.... #example for parent_full_name: v(['cond1'])3 for local_name in model.params_include.keys(): if local_name == 'wfpt': continue parent_full_name = obs_node.parents[local_name].__name__ params[local_name] = model.mc.db.trace(parent_full_name)[idx] return params def _gen_statistics(): """generate different statistical tests from ppd_test.""" statistics = [] ##accuracy test test = {} test['name'] = 'acc' test['func'] = lambda rts: sum(rts > 0) / len(rts) statistics.append(test) ##quantile statistics of absolute response time quantiles = [10, 30, 50, 70, 90] for q in quantiles: test = {} test['name'] = 'q%d' % q test['func'] = lambda rts, q=q: scoreatpercentile(np.abs(rts), q) statistics.append(test) return statistics def plot_posteriors(model, **kwargs): """Generate posterior plots for each parameter. This is a wrapper for pymc.Matplot.plot() """ pm.Matplot.plot(model.mc, **kwargs) def data_plot(data, nbins=50): data = hddm.utils.flip_errors(data) plt.figure() plt.hist(data['rt'], nbins) plt.show() def quantiles_summary(hm, method, n_samples, quantiles = (10, 30, 50, 70, 90), sorted_idx = None, cdf_range = (-5, 5)): """ generate quantiles summary of data or samples Input: hm - hddm model method - which parameters to use for the computation of quantiles 'observed' - quantiles from the observed data (not using any parameters) 'deviance' - quantiles of data sampled using the parameters that have maximum deviance 'samples' - quantiles of data generated from samples from the posterior distrbution qunaitles - the qunatiles to be generated sorted_idx - the condtions will be sort according to sorted_idx. if sorted_idx == None then they will be sorted according to the accuracy achieved in each condition. cdf_range (advanced) - the range of the cdf used to generate the estimated quantiles. Ouput: group_dict - a dictionary with group statistics subj_dict - a list of dictionaries wih subjects statistcs conds - the list of conditions sorted associated with the statistics sorted_idx - the indices of the conditions such that: if we set: keys = hm.params_dict['wfpt'].subj_nodes.keys() than: keys[sorted_idx] == conds """ #check if group model if hm.is_group_model: is_group = True n_subj = hm._num_subjs else: n_subj = 1 is_group = False #check method if method == 'observed': n_samples = 1 len_trace = 1 elif method == 'samples': db =hm.mc.db name = list(hm.mc.stochastics)[0].__name__ len_trace = db.trace(name).length() params = hddm.generate.gen_rand_params() elif method == 'deviance': db =hm.mc.db params = hddm.generate.gen_rand_params() else: raise ValueError, "unkown method" #Init n_q = len(quantiles) wfpt_dict = hm.params_dict['wfpt'].subj_nodes conds = wfpt_dict.keys() n_conds = len(conds) temp_dict = {} group_dict = {}; subj_dict = [None]*n_subj; count = zeros((n_conds * 2, n_samples)) temp_dict['q'] = zeros((n_conds * 2, n_samples, n_q)) temp_dict['prob'] = zeros((n_conds * 2, n_samples)) temp_dict['n'] = zeros((n_conds*2, n_samples)) group_dict = deepcopy(temp_dict) for i_subj in range(n_subj): subj_dict[i_subj] = deepcopy(temp_dict) #loop over subjs for i_subj in range(n_subj): #get wfpt nodes if is_group: wfpt = [x[i_subj] for x in hm.params_dict['wfpt'].subj_nodes.values()] else: wfpt = hm.params_dict['wfpt'].subj_nodes.values() #loop over conds for i_cond in range(n_conds): wfpt_params = dict(wfpt[i_cond].parents) data_len = len(wfpt[i_cond].value) for i_sample in xrange(n_samples): #get rt if method == 'observed': rt = wfpt[i_cond].value else: #get sample_idx if method == 'deviance': sample_idx = np.argmin(db.trace('deviance')[:]) else: sample_idx = ((len_trace - 1) // (n_samples - 1)) * i_sample #get params for key in params.iterkeys(): if isinstance(wfpt_params[key], pm.Node): node = wfpt_params[key] name = node.__name__ params[key] = db.trace(name)[sample_idx] #generate rt rt = hddm.generate.gen_rts(params, samples=data_len, range_=cdf_range, dt=1e-3, intra_sv=1., structured=False, subj_idx=None, method='cdf') #loop over responses for i_resp in range(2): cond_ind = i_cond * 2 + i_resp if i_resp==0: t_rt = rt[rt > 0] else: t_rt = rt[rt < 0] #get n_q if len(t_rt)>=n_q: t_quan = [scoreatpercentile(abs(t_rt),x) for x in quantiles] subj_dict[i_subj]['q'][cond_ind, i_sample, :] = t_quan group_dict['q'][cond_ind, i_sample, :] += t_quan count[cond_ind, i_sample] += 1 else: subj_dict[i_subj]['q'][cond_ind, i_sample, :] = np.NaN t_prob = 1.*len(t_rt) / len(rt) subj_dict[i_subj]['prob'][cond_ind, i_sample] = t_prob group_dict['prob'][cond_ind, i_sample] += t_prob subj_dict[i_subj]['n'][cond_ind, i_sample] = len(t_rt) group_dict['n'][cond_ind, i_sample] = len(t_rt) #compute group values group_dict['prob'] /= n_subj for i_c in range(n_conds * 2): for i_sample in xrange(n_samples): if count[i_c, i_sample] > 0: group_dict['q'][i_c, i_sample,:] /= count[i_c, i_sample] else: group_dict['q'][i_c, i_sample,:] = np.NaN #sort if sorted_idx == None: sorted_idx = np.argsort(group_dict['prob'][:,0]).flatten() group_dict['prob'] = group_dict['prob'][sorted_idx,:] group_dict['q'] = group_dict['q'][sorted_idx,:,:] group_dict['n'] = group_dict['n'][sorted_idx,:] for i_subj in range(n_subj): subj_dict[i_subj]['prob'] = subj_dict[i_subj]['prob'][sorted_idx,:] subj_dict[i_subj]['q'] = subj_dict[i_subj]['q'][sorted_idx,:, :] subj_dict[i_subj]['n'] = subj_dict[i_subj]['n'][sorted_idx, :] conds = array(conds)[(sorted_idx // 2)] return group_dict, subj_dict, conds, sorted_idx def _draw_qp_plot(dict, conds, conds_to_plot, title_str, samples_summary, marker, handle = None): """ draw QP plot Input: dict - subject/group dictionary from qunatiles_summary conds - conditions from quantiles_summary conds_to_plot - a list of conds to plot title_str - the title of the plot samples_summary - sould we plot only samples sammury and not the samples themselves maarker - marker used in plot handle - axes's handle """ colors = ['b', 'g', 'c', 'm', 'r', 'y', 'k'] n_q = dict['q'].shape[2] n_conds = len(conds)//2 n_samples = dict['prob'].shape[1] #group plot if handle == None: f = plt.figure() handle = f.add_subplot(111) color_counter = -1 for i_c in range(n_conds): #check if we need to plot this condition if i_c not in conds_to_plot: continue color_counter += 1 #loop over correct and incorrect reponses for i_resp in range(2): if i_resp == 0: idx = i_c else: idx = (n_conds * 2) - 1 - i_c #get color of marker cc = colors[color_counter % len(colors)] #plot it if samples_summary: ok_idx = ~np.isnan(dict['prob'][idx,:]) if sum(ok_idx) == 0: continue mean_prob = np.mean(dict['prob'][idx,ok_idx]) xerr = np.empty((2,n_q)) xerr[0,:] = mean_prob - scoreatpercentile(dict['prob'][idx,ok_idx], 2.5) xerr[1,:] = scoreatpercentile(dict['prob'][idx,ok_idx], 97.5) - mean_prob mean_q = np.mean(dict['q'][idx,ok_idx,:],0) yerr = np.empty((2,n_q)) yerr[0,:] = array([scoreatpercentile(x, 2.5) for x in dict['q'][idx,ok_idx,:].T]) - mean_q yerr[1,:] = mean_q - array([scoreatpercentile(x, 97.5) for x in dict['q'][idx,ok_idx,:].T]) handle.errorbar(ones(n_q)*mean_prob, mean_q, yerr=yerr, xerr=xerr, marker=marker, fmt=cc) else: for i_sample in range(n_samples): if np.isnan(dict['prob'][idx, i_sample]): continue handle.plot(ones(n_q)*dict['prob'][idx,i_sample], dict['q'][idx,i_sample,:],'%s%s' % (cc, marker)) #add title and labels if title_str != None: plt.gcf().canvas.set_window_title(title_str) plt.title(title_str) plt.xlabel('probability') plt.ylabel('RT') plt.xlim([-0.05, 1.05]) handle.get_figure().canvas.draw() return handle def qp_plot(hm, quantiles = (10, 30, 50, 70, 90), plot_subj=True, split_func=lambda x:0, method = None, samples_summary=True, n_samples=50, cdf_range=(-5, 5)): """ generate a quantile-probability plot Input: hm - hddm model qunatiles - the quantiles that are going to be displyed plot_subj - generate a QP plot for each subject as well split_func - split the plot into several plots according to the output of the function method - which values will be used to compute the estimated quantiles. may be one of the followings: -'none' : display only observed quantiles -'deviance': use the point of minimum deviance -'samples': samples n_samples and plot them all (not recommended) -'samples_summary': samples n_samples and plot the mean and 95 credible set of them n_samples - see method cdf_range (advanced) - the range of the cdf used to generate the estimated quantiles. TODO: there should be an option to use the average value of the parameters to create samples. it make much more sense than 'deviance' """ #### compute empirical quantiles #Init n_q = len(quantiles) if hm.is_group_model: is_group = True n_subj = hm._num_subjs else: n_subj = 1 is_group = False #get group and subj dicts obs_group_d, obs_subj_d, conds, sorted_idx = \ quantiles_summary(hm, 'observed', n_samples = 1, quantiles = quantiles) #get splits keys = [split_func(x) for x in conds] dic = {} for i_s in range(len(keys)): if dic.has_key(keys[i_s]): dic[keys[i_s]].append(i_s) else: dic[keys[i_s]] = [i_s] splits = dic.values() splits_keys = dic.keys() #plot g_handles = [None]*len(splits) s_handles = [[None]*n_subj for x in range(len(splits))] for i_s in range(len(splits)): g_handles[i_s] = _draw_qp_plot(obs_group_d, conds, splits[i_s], title_str = "QP group (%s)" % splits_keys[i_s], samples_summary=False, marker = 'd') if not plot_subj: continue for i_subj in range(n_subj): s_handles[i_s][i_subj] = _draw_qp_plot(obs_subj_d[i_subj], conds, splits[i_s], title_str = "QP %d (%s)" % (i_subj, splits_keys[i_s]), samples_summary=False, marker = 'd') if (method == None) or (method == 'none'): return elif (method == 'deviance'): quantiles_method = 'deviance' elif method == 'samples': quantiles_method = 'samples' else: raise ValueError, "unknown method" #### compute estimated quantiles #get group and subj dicts print "getting quantiles summary of samples" i_t = time(); sim_group_d, sim_subj_d, conds, sorted_idx = \ quantiles_summary(hm, method=method, n_samples=n_samples, quantiles=quantiles, sorted_idx=sorted_idx) print "took %d seconds to prepare quantiles" % (time() - i_t) sys.stdout.flush() #plot for i_s in range(len(splits)): _draw_qp_plot(sim_group_d, conds, splits[i_s], title_str=None, samples_summary=samples_summary, marker='o', handle=g_handles[i_s]) if not plot_subj: continue for i_subj in range(n_subj): _draw_qp_plot(sim_subj_d[i_subj], conds, splits[i_s], title_str = None, samples_summary=samples_summary, marker = 'o', handle = s_handles[i_s][i_subj]) if __name__ == "__main__": import doctest doctest.testmod() HDDM-0.2/README.rst0000664000102500010250000001135511727455235014336 0ustar wieckiwiecki00000000000000************ Introduction ************ :Date: March 8, 2012 :Author: Thomas V. Wiecki, Imri Sofer, Michael J. Frank :Contact: thomas_wiecki@brown.edu, imri_sofer@brown.edu, michael_frank@brown.edu :Web site: http://github.com/hddm-devs/hddm :Mailing list: https://groups.google.com/group/hddm-users/ :Copyright: This document has been placed in the public domain. :License: HDDM is released under the GPLv3. :Version: 0.2 Purpose ======= HDDM is a python toolbox for hierarchical Bayesian parameter estimation of the Drift Diffusion Model (via PyMC). Drift Diffusion Models are used widely in psychology and cognitive neuroscience to study decision making. Features ======== * Uses hierarchical Bayesian estimation (via PyMC) of DDM parameters to allow simultaneous estimation of subject and group parameters, where individual subjects are assumed to be drawn from a group distribution. HDDM should thus produce better estimates when less RT values are measured compared to other methods using maximum likelihood for individual subjects (i.e. `DMAT`_ or `fast-dm`_). * Heavily optimized likelihood functions for speed (Navarro & Fuss, 2009). * Flexible creation of complex models tailored to specific hypotheses (e.g. estimation of separate drift-rates for different task conditions; or predicted changes in model parameters as a function of other indicators like brain activity). * Easy specification of models via configuration file fosters exchange of models and research results. * Built-in Bayesian hypothesis testing and several convergence and goodness-of-fit diagnostics. Usage ===== Command line ------------ The easiest way to use HDDM is by creating a configuration file for your model: example.conf :: [depends] v = difficulty Then call hddm_fit.py: :: hddm_fit.py --samples 10000 --burn 5000 example.conf mydata.csv Python ------ Of course, you can also use HDDM directly from within Python: :: import hddm # Load data from csv file into a NumPy structured array data = hddm.load_csv('simple_difficulty.csv') # Create a HDDM model multi object model = hddm.HDDM(data, depends_on={'v':'difficulty'}) # Create model and start MCMC sampling model.sample(10000, burn=5000) # Print fitted parameters and other model statistics model.print_stats() # Plot posterior distributions and theoretical RT distributions model.plot_posteriors() model.plot_posterior_predictive(savefig=True) Installing ========== See further below on how to install most of the necessary dependencies for the most common platforms. Assuming you have the dependencies, the easiest way to install HDDM is via pip_ :: pip install kabuki pip install hddm If you already installed a previous version of HDDM, you should upgrade instead :: pip install --upgrade --no-deps kabuki pip install --upgrade --no-deps hddm Dependencies ============ HDDM relies on the following packages: * Python * NumPy * PyMC_ (installation instructions: http://pymc.googlecode.com/svn/doc/installation.html) * kabuki_ * SciPy (optional) * Matplotlib (optional) * Cython_ (optional) Windows ------- The easiest way is to download and install the `Enthought Python Distribution`_ (EPD) which is free for academic use. An untested alternative is the Python(x,y) distribution. Please let us know your experiences if you test this option. We recommend using pip to download and install HDDM. The easiest way to install pip is via easy_install. Start the windows command shell (cmd.exe) and type :: easy_install pip Linux (Debian based, such as Ubuntu) ------------------------------------ Most of HDDM's dependencies are available from your repository, you can install them by typing :: aptitude install python python-dev python-numpy python-scipy python-matplotlib cython python-pip gfortran liblapack-dev which requires sudo rights. OSX --- We recommend installing the `SciPy Superpack`_ maintained by Chris Fonnesbeck. Getting started =============== Check out the documentation_ or the tutorial_ on how to use HDDM. Join our low-traffic `mailing list`_. .. _HDDM: http://code.google.com/p/hddm/ .. _Python: http://www.python.org/ .. _PyMC: http://code.google.com/p/pymc/ .. _Cython: http://www.cython.org/ .. _DMAT: http://ppw.kuleuven.be/okp/software/dmat/ .. _fast-dm: http://seehuhn.de/pages/fast-dm .. _documentation: http://ski.cog.brown.edu/hddm_docs .. _tutorial: http://ski.cog.brown.edu/hddm_docs/tutorial.html .. _manual: http://ski.cog.brown.edu/hddm_docs/manual.html .. _kabuki: https://github.com/hddm-devs/kabuki .. _Enthought Python Distribution: http://www.enthought.com/products/edudownload.php .. _mailing list: https://groups.google.com/group/hddm-users/ .. _SciPy Superpack: http://fonnesbeck.github.com/ScipySuperpack/ HDDM-0.2/PKG-INFO0000664000102500010250000000127711727455647013755 0ustar wieckiwiecki00000000000000Metadata-Version: 1.0 Name: HDDM Version: 0.2 Summary: HDDM is a python module that implements Hierarchical Bayesian parameter estimation of Drift Diffusion Models (via PyMC). Home-page: http://github.com/hddm-devs/hddm Author: Thomas V. Wiecki, Imri Sofer, Michael J. Frank Author-email: thomas_wiecki@brown.edu License: UNKNOWN Description: UNKNOWN Platform: UNKNOWN Classifier: Development Status :: 4 - Beta Classifier: Environment :: Console Classifier: Operating System :: OS Independent Classifier: Intended Audience :: Science/Research Classifier: License :: OSI Approved :: GNU General Public License (GPL) Classifier: Programming Language :: Python Classifier: Topic :: Scientific/Engineering HDDM-0.2/INSTALL.rst0000664000102500010250000001366211713025571014501 0ustar wieckiwiecki00000000000000************************* Installation Instructions ************************* :Date: September 11, 2010 :Author: Thomas Wiecki :Contact: thomas_wiecki@brown.edu :Web site: http://code.google.com/p/hddm/ :Copyright: This document has been placed in the public domain. Dependencies ============ HDDM requires some prerequisite packages to be present on the system. Fortunately, there are currently only a few dependencies, and all are freely available online. * `Python`_ version 2.5 or 2.6. * `NumPy`_ (1.4 or newer): The fundamental scientific programming package, it provides a multidimensional array type and many useful functions for numerical analysis. * `PyMC`_ (2.0 or newer): The Markov-Chain Monte Carlo sampler which allows construction of hierarchical models. It is used to estimate the posterior distributions over the parameters. * `Matplotlib (optional)`_ : 2D plotting library which produces publication quality figures in a variety of image formats and interactive environments * `Traits and TraitsUI (optional)`_ : Allows, among other things, automatic creation of GUIs and is used for the DDM demo applet (brownian.py). * `SciPy (optional)`_ : Library of algorithms for mathematics, science and engineering. * `IPython (optional)`_ : An enhanced interactive Python shell and an architecture for interactive parallel computing. There are prebuilt distributions that include all required dependencies. For Mac OS X users, we recommend the `MacPython`_ distribution or the `Enthought Python Distribution`_ on OS X 10.5 (Leopard) and Python 2.6.1 that ships with OS X 10.6 (Snow Leopard). Windows users should download and install the `Enthought Python Distribution`_. The Enthought Python Distribution comes bundled with these prerequisites. Note that depending on the currency of these distributions, some packages may need to be updated manually. If instead of installing the prebuilt binaries you prefer (or have) to build ``HDDM`` yourself, make sure you have a Cython and a C compiler. There are free compilers (gcc) available on all platforms. .. _`Python`: http://www.python.org/. .. _`NumPy`: http://www.scipy.org/NumPy .. _`PyMC`: http://code.google.com/p/PyMC .. _`Matplotlib (optional)`: http://matplotlib.sourceforge.net/ .. _`MacPython`: http://www.activestate.com/Products/ActivePython/ .. _`Enthought Python Distribution`: http://www.enthought.com/products/epddownload.php .. _`SciPy (optional)`: http://www.scipy.org/ .. _`IPython (optional)`: http://ipython.scipy.org/ .. _Cython: http://www.cython.org/ Installation using EasyInstall ============================== The easiest way to install PyMC is to type in a terminal:: easy_install hddm Provided `EasyInstall`_ (part of the `setuptools`_ module) is installed and in your path, this should fetch and install the package from the `Python Package Index`_. Make sure you have the appropriate administrative privileges to install software on your computer. .. _`Python Package Index`: http://pypi.python.org/pypi .. _`setuptools`: http://peak.telecommunity.com/DevCenter/setuptools Installing from pre-built binaries ================================== Pre-built binaries are available for Windows XP and Mac OS X. There are at least two ways to install these: 1. Download the installer for your platform from `PyPI`_. 2. Double-click the executable installation package, then follow the on-screen instructions. For other platforms, you will need to build the package yourself from source. Fortunately, this should be relatively straightforward. .. _`PyMC site`: pymc.googlecode.com Compiling the source code ========================= TODO First download the source code tarball from `PyPI`_ and unpack it. Then move into the unpacked directory and follow the platform specific instructions. Windows ------- One way to compile PyMC on Windows is to install `MinGW`_ and `MSYS`_. MinGW is the GNU Compiler Collection (GCC) augmented with Windows specific headers and libraries. MSYS is a POSIX-like console (bash) with UNIX command line tools. Download the `Automated MinGW Installer`_ and double-click on it to launch the installation process. You will be asked to select which components are to be installed: make sure the g77 compiler is selected and proceed with the instructions. Then download and install `MSYS-1.0.exe`_, launch it and again follow the on-screen instructions. Once this is done, launch the MSYS console, change into the PyMC directory and type:: python setup.py install This will build the C and Fortran extension and copy the libraries and python modules in the C:/Python26/Lib/site-packages/pymc directory. .. _`MinGW`: http://www.mingw.org/ .. _`MSYS`: http://www.mingw.org/wiki/MSYS .. _`Automated MinGW Installer`: http://sourceforge.net/projects/mingw/files/ .. _`MSYS-1.0.exe`: http://downloads.sourceforge.net/mingw/MSYS-1.0.11.exe Mac OS X or Linux ----------------- In a terminal, type:: python setup.py install The above syntax also assumes that you have gFortran installed and available. The `sudo` command may be required to install PyMC into the Python ``site-packages`` directory if it has restricted privileges. .. _`EasyInstall`: http://peak.telecommunity.com/DevCenter/EasyInstall .. _`PyPI`: http://pypi.python.org/pypi/pymc/ Development version =================== Get the code from the GIT mirror:: git clone git://github.com/hddm-devs/hddm.git hddm Running the test suite ====================== ``pymc`` comes with a set of tests that verify that the critical components of the code work as expected. To run these tests, users must have `nose`_ installed. The tests are launched from a python shell:: import pymc pymc.test() Bugs and feature requests ========================= Report problems with the installation, bugs in the code or feature request at the `issue tracker`_. Comments and questions are welcome and should be addressed to PyMC's `mailing list`_. .. _`issue tracker`: http://code.google.com/p/hddm/issues/list