model_simulation.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import pickle\n",
    "import numpy as np\n",
    "from scipy.special import expit\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "import pandas as pd\n",
    "\n",
    "from utils_models import *\n",
    "from funcs_model_simulation import *\n",
    "from funcs_learning_curves import *"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "datasetName = 'takahashi2016roesch2009burton2018Valid'\n",
    "dataValid = pd.read_csv('data/' + datasetName + '.csv')\n",
    "ratList = dataValid['rat'].unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "models = ['sixState_full', 'fourState_full', 'hybridValue_full', 'hybridLearning_full']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "dir_simu = 'model_simulation/'"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Run simulation / Load simulated data\n",
    "\n",
    "Running the simulation takes a long time. For easy reproduction of figures in the paper, we have saved the aggregated results, and you can load them directly by setting `run_simulation = False`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_simulated_data(varName, modelName, w4=None, NSessions=10000, run_simulation=False):\n",
    "    \n",
    "    fileName = dir_simu + modelName + '_' + datasetName + '_groupMean' + (('_w4' + str(w4)) if w4 is not None else '') + '_NSessions' + str(NSessions) + '_NTrials' + str(NTrials) + '_' + varName + '.pickle'\n",
    "    \n",
    "    if varName == 'curveData':\n",
    "        curveData = None\n",
    "    \n",
    "    if run_simulation:\n",
    "        # set parameter values\n",
    "        params = get_params(datasetName, modelName)\n",
    "        if w4 is not None:\n",
    "            params[1] = w4\n",
    "        # run simulation\n",
    "        dataSimu = model_simulate(modelName, params, NSessions, NTrials)\n",
    "        if varName == 'avgR':\n",
    "            # calculate average reward\n",
    "            value = dataSimu['rewardAmount'].mean()\n",
    "        elif varName == 'curveData':\n",
    "            # prepare data for plotting learning curve\n",
    "            value = plotLearningCurve(dataSimu, N=10, ifReturnCurveData=True)\n",
    "        # save results\n",
    "        pickle.dump(value, open(fileName, 'wb'))\n",
    "    else:\n",
    "        # load results\n",
    "        value = pickle.load(open(fileName, 'rb'))\n",
    "        \n",
    "    return value"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Average reward"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "# simulation parameters\n",
    "NSessions = 100000\n",
    "NTrials = 57\n",
    "w4List = [0, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "run_simulation = False\n",
    "\n",
    "avgR = dict.fromkeys(['rat'+str(rat) for rat in ratList] + models + ['w4'+str(w4) for w4 in w4List])\n",
    "\n",
    "# all models\n",
    "for modelName in models:\n",
    "    avgR[modelName] = get_simulated_data('avgR', modelName, None, NSessions, run_simulation)\n",
    "            \n",
    "# hybrid-value model: w4 from 0 to 1\n",
    "modelName = 'hybridValue_full'\n",
    "for w4 in w4List:\n",
    "    avgR['w4'+str(w4)] = get_simulated_data('avgR', modelName, w4, NSessions, run_simulation)\n",
    "\n",
    "# animals\n",
    "for rat in ratList:\n",
    "    avgR['rat'+str(rat)] = dataValid.loc[dataValid['rat'] == rat, 'rewardAmount'].mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Load fitted $w_4$ value for hybrid-value model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "allSamples = pd.read_csv('model_fits/' + datasetName + '_hybridValue_full_allSamples.csv')\n",
    "w4_fit = phi_approx(allSamples.loc[allSamples['warmup']==0, 'mu_pr[2]'].values.mean())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Plot"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/pdf": "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\n",
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 288x216 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from IPython.display import set_matplotlib_formats\n",
    "set_matplotlib_formats('png', 'pdf')\n",
    "plt.rcParams.update({'font.family': 'arial'})\n",
    "lineWidth = 2\n",
    "fontsize = 24\n",
    "labelsize = 20\n",
    "ylim=[1.05, 1.14]\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(4,3))\n",
    "\n",
    "# simulation\n",
    "ax.plot(1, avgR['fourState_full'], 'o', color='C1', linewidth=lineWidth, markersize=8, label='Four-state')\n",
    "ax.plot(0, avgR['sixState_full'], 'o', color='C2', linewidth=lineWidth, markersize=8, label='Six-state')\n",
    "ax.plot(w4_fit, avgR['hybridValue_full'], 'o', color='k', linewidth=lineWidth, markersize=8, markeredgewidth=2, label='Hybrid-value')\n",
    "ax.plot(1.2, avgR['hybridLearning_full'], 'o', color='k', linewidth=lineWidth, markersize=8, fillstyle='none', markeredgewidth=2, label='Hybrid-learning')\n",
    "\n",
    "from matplotlib.colors import ListedColormap, BoundaryNorm\n",
    "from matplotlib.collections import LineCollection\n",
    "N = 256\n",
    "vals = np.ones((N, 4))\n",
    "vals[:, 0] = np.linspace(44/256, 255/256, N)\n",
    "vals[:, 1] = np.linspace(160/256, 127/256, N)\n",
    "vals[:, 2] = np.linspace(44/256, 14/256, N)\n",
    "cmap = ListedColormap(vals)\n",
    "x = np.array(w4List)\n",
    "y = np.array([avgR['w4'+str(w4)] for w4 in w4List])\n",
    "points = np.array([x, y]).T.reshape(-1, 1, 2)\n",
    "segments = np.concatenate([points[:-1], points[1:]], axis=1)\n",
    "norm = plt.Normalize(np.min(y), np.max(y))\n",
    "lc = LineCollection(segments, cmap=cmap, norm=norm, color='w', label='Hybrid-value (varying $w_4$)')\n",
    "lc.set_array(y)\n",
    "lc.set_linewidth(lineWidth)\n",
    "line = ax.add_collection(lc)\n",
    "\n",
    "# animals\n",
    "avgR_rat = np.array([avgR['rat'+str(rat)] for rat in ratList])\n",
    "ax.errorbar(x=1.5, y=np.mean(avgR_rat), yerr=np.std(avgR_rat)/np.sqrt(len(ratList)), linestyle='', color='k', capsize=7, linewidth=lineWidth, label='Rats')\n",
    "ax.plot(1.5, np.mean(avgR_rat), '.', color='k', markersize=8)\n",
    "\n",
    "ax.set_xlabel('$w_4$         ', fontsize=fontsize)\n",
    "ax.set_ylabel('Average reward', fontsize=fontsize, labelpad=10)\n",
    "ax.tick_params(axis='both', width=1.5, pad=5, direction='out', labelsize=labelsize)\n",
    "ax.spines['top'].set_visible(False)\n",
    "ax.spines['right'].set_visible(False)\n",
    "ax.spines['left'].set_linewidth(1.5)\n",
    "ax.spines['bottom'].set_linewidth(1.5)\n",
    "ax.set(xlim=[-0.15,1.7], ylim=ylim)\n",
    "ax.set(xticks=[0,0.5,1])\n",
    "\n",
    "ax.legend(frameon=False, fontsize=labelsize, handletextpad=0.5, bbox_to_anchor=(1, 1.1))\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Learning curves"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "# simulation parameters\n",
    "NSessions = 10000\n",
    "NTrials = 57"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "run_simulation = False\n",
    "\n",
    "curveData = dict.fromkeys(['w4'+str(w4) for w4 in w4List])\n",
    "            \n",
    "# hybrid-value model: w4 = 0 and 1\n",
    "modelName = 'hybridValue_full'\n",
    "for w4 in [0,1]:\n",
    "    curveData['w4'+str(w4)] = get_simulated_data('curveData', modelName, w4, NSessions, run_simulation)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/pdf": "JVBERi0xLjQKJazcIKu6CjEgMCBvYmoKPDwgL1BhZ2VzIDIgMCBSIC9UeXBlIC9DYXRhbG9nID4+CmVuZG9iago4IDAgb2JqCjw8IC9FeHRHU3RhdGUgNCAwIFIgL0ZvbnQgMyAwIFIgL1BhdHRlcm4gNSAwIFIKL1Byb2NTZXQgWyAvUERGIC9UZXh0IC9JbWFnZUIgL0ltYWdlQyAvSW1hZ2VJIF0gL1NoYWRpbmcgNiAwIFIKL1hPYmplY3QgNyAwIFIgPj4KZW5kb2JqCjEwIDAgb2JqCjw8IC9Bbm5vdHMgWyBdIC9Db250ZW50cyA5IDAgUgovR3JvdXAgPDwgL0NTIC9EZXZpY2VSR0IgL1MgL1RyYW5zcGFyZW5jeSAvVHlwZSAvR3JvdXAgPj4KL01lZGlhQm94IFsgMCAwIDY0My44Njg3NSAyMjIuOTQxMjUgXSAvUGFyZW50IDIgMCBSIC9SZXNvdXJjZXMgOCAwIFIKL1R5cGUgL1BhZ2UgPj4KZW5kb2JqCjkgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAxMSAwIFIgPj4Kc3RyZWFtCnicvZhNrx1JDYb3/St6CQsqVeVyfSwnAiKxI1yJBWKB7mQyRMlIzAgi/j2P3R+nupMQZsEsbnKOT5XL9fq1X3en9d3y4pu0vv1pjes7/j6uf1n/yv/frml9tb747Zt//f35zetXL9fnn5aI/cNSi4Ree1O+vZ++5ZzDKCkr1nj59v2y/LBwBhte4fbtsrQaqm9KPQhr8Crlano/mXLsQbvZzo0PE96/4w55u8NbDni3pqDTTexYfvnMqU1Cuh66vCTSj8vLp/XF7xOnrE/fLdpCLGKr0r786dvlV/HX69O79XdPix/2cK41SGuz893yeeclBtHK2WWEnOPhPOQvuB8llDxm97vlK+57C7nl0335gvskOZSeLtjspq8cQK7BsJwn1C+d0GJQkcsJu+lrJ1QByHqe0L9wwsGN6YQHXf5rgnNMYHlkOB3+j8VqiyMM+02CUNw3jmiB1cx+3/XBi8Q2f/P8/M8f//b878PHP9Y7/VT7mgaZruuPb9Y/rz+sef0Dfx85gASNHmXkpI0vUcuIadTY19evlmtVLh1UUqqxwc8cksbROnfvIwwtpRY395hiMoRHDTqa5ISZa9eaOf/9kqKEXGJqA7uEFnOv4vYRepcoFftgfc/VqMCBQbP2ISSeO2Qd3TiY8pnUXPg03EnGyR4iqFWRqNnsUkPZY8w92GWH22kqcQ9SUogt1Wr1k6yS9mBaCrmPUbip8SLuscCk0ap0v1MdB+K18ympR0geDrzMCXhVP7TnEzC8xCzRl4PjAVhVel2mL5kdIA/Aagul6MavHPMJGG7Yl2tx+zgBqzUMQPA7ZYBMZ5BpSO6+nP65Q1ZLkKGWvQ+Yx4EYwcTeh980A+QOGE4okeLWkkPbQqQQRlFicfM4UGwjJFARDxxwdxRBAvqpWu4y5N5ANELlahGZdRwgDtg8hib30fTEEEIZuaJHgseTdTlQjWlbP/QECxBFC82GawognmCRRNpiZzkZObGCRollZk16sgsncUj3QwUID3bhJHdJzdeD4cEusDVko0Euogdc7qekLaMCjAfrtmM1bvZx0o6Mdq2tul31pB0lT/6rBw+QO2SVhqWtDGvjQsJ2yAgSM6S21SC5I9aN3bVZ1Qk47qQjG9RcTu669xPG2KxbpOI3Bd4DRmCXAQ7mvADvUaTQDOp6NgpN8ixTslpkixwgTiDBiEKp3hsKmw8gS6GPafPeUFh0ANZTqL1WtY5UAPIArFtTq8l7Rin97IwsZ2rwgiSss1Cxc4Pk/C3Ad5DM7DRKp3sByYNkELtK5sJmB8qjUjmWnZDb7Gw+K5VUltbLZu8naqwvKZbh8IDmMS1EO4kNmOHPgVppgcXwiMsqYO6gKXMM6S52V5A+MFMrIfBwc05TqZaReypu7geSHScpdbGbKgDvQI4Wmm4wkvUDRmvaWpm/3Nwf+tADOjCc7kpYB4zZ6G53VKPm0XlraC324otbeWAFPI2eUe2WYHhgBYYdtjuGCoaHtpPcJuKEVGaVg2HNrkYCLcZqunv0XstiVYcFwTsZBuYw1SGvqZyV2k23hnp5VFCUR5S0E4vA7P1BPPSEjd5P5qGTLhd78/JY/rT8cf05eh0DnWWXaL7U3AqqAi2Xyy+v7wP1RbpjaHn0clNuah8AWrkpN4yGft7+LsqNEFFrm/rNyk0fiHFri7Ny8xHlbj6cTsqNuVcwSjftxj12dWrO2m20Qom82mbtxl6jbKozazcky5l6kJt42xwFsdJdvIHG6HwTb7UOmPdoJvHGyxiDQfGu3tQpvWOLclZvhIqE+amzeqMNvTXvgBfxthWd4rqJNz1S+HdcxVu7jQxjtJt4Q9SofZsMJvGm9gvtW6/qjZkWLc7xSb1RlwTFfTSf1BvtahBn0/pTvWmOPISo5qt6szg1US+HSb1pQp0Gt61+iHfz8h3lrt3FBS1645u1Gy5k23eR7sK+CK5y024kRJUf5KbdrG9cXu/ajR0JTd6bZ+2m36k02bR+0m66M4Imo960W6PdevR+1W6lmyaSMa7aDUadeTSnq3ZblyvqqZikm5bYhOLXm3Zbs2GOKFft7mpzt9Z6lW7MsaBB7aLctA6E0jv5pNs85WnJIukq29612jZZTqo9nG7VNWUWbZjPUCQ6bqJtPOSJZxsWHqINOCyQbbSYRHtbbhe+irYVHMtjv4k2x6J79SbZNlwzoDW5STYbo6Dx5SbZzPQ0fxrCVbIxi5WWXCVbbbgeNi5dJZspmnFR2lWy/RIUf79KtnUWNvtUOEm2Pa5m+CxXyab4GH3Ur//Q7AbDJG/PaJNkE/ZgHkrjqtjUp/Qi/SrZzYqZFllvik0DtWf/XG6KbXbNGy0mxSafrZbsojMrNgeVOJIvnxWb1kfRN8/QrNjWckqUTeEnySajlcL3gr5INmo5eJAaN8mu/ty58WuSbBBDjHJuP0+y7b0QlFppTcf7IXsG4OGa+U/W++N15qm3jH3W2F8bTbbpxcL/LQSBjeMWwmT7JUKgPX6CwmT7JASbjOInL+KuPj/zOu7Tt3v4+59eFtq5y38A1g8vWgplbmRzdHJlYW0KZW5kb2JqCjExIDAgb2JqCjE5MTEKZW5kb2JqCjE2IDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMTEyID4+CnN0cmVhbQp4nDWMyw3EQAhD71OFS+DPUE9We8r2f10GJULCD2w5IkEzERvJgQ+vc/6WpM7jXpr+0AlRqwePWiaEs4lzP1Q1Fs0WckhZ15ntoVOnsJJDTnAXqCqCWkUQxriWssLLodQJK8g2OFk7b+O9vn9DCiPoCmVuZHN0cmVhbQplbmRvYmoKMTcgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCA0MzcgPj4Kc3RyZWFtCnicNZJJciUxCET3dQou4AgxaDrP7/DKff9tv6TciyqQEJCZMOeyYbnsy92ml80c9sefGmWr7O+T2xXOGuYYHzbt88QZ9jUtandieLb9PD7fUu5hYSeMcp9nUvuozE39PTeXclZY8fhM+lqMxfWibGyjQVTZvRZTr923xbrkHot9OzWOEpxTXJ4PMm+Rhk8eHQCSY9OStuTn5YuOyKPza89rZ/yPHG7OBRp5O63iKDKHld6iQ02hiLaoAcm+GbIIBfjKxrbwAoaFdnc285z3lUTS/jycCowqRhAetSSZj2OFfp5pc0j8zXyEMAZqxstT04pWTzniJi/RUC8SuTXTzyMMk5EWnOvyTfytSRZ1Sxg2rKhYh6gLex3eiBEqtwqy2dXwAj1/2vOcxJhYdekljAVXwRsMaTSG0SgCSxY8HHpDotjvqr3SB+RZuvXuloYWCBnajtAaqG1I0KutQ1Bgx7gtNTcjzTUAiHupSe8koH0dgBKBeNvsHHmqwlJdmLC8p8XfDUnznEJQ1jsYFNH5HkUThff9XQz2LreKYY+a9zwuECAkYs4+NcF3l36p/jzf/wAZT6ZTCmVuZHN0cmVhbQplbmRvYmoKMTggMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAyNDYgPj4Kc3RyZWFtCnicNVFJcgQxCLv7FfpAqgzCxn5Pp3JK/n+NYGYOXcLGaKHXnJjgxZcZ4hoWJ75txAzYnfgbdWl58TviqLKLWIl0hJ0afAbVXSCtSTxX4zN8siuLDXOYjuvofk/VUbOWre0iFM0FjQixbsdOlIdnnIUogmplwlb5LBo7kpUdl+NyXcKFYuIrB/V+meabtjQiyr6+t/84LrXd6gkuKibFS71mRk9yCWnNKx4xcGpSUTxvo6IyujI5i/AOFlZeTCemOq45f6Hv7qiqFdcLq2VsfbOT3YXcHWzXQqNA8mqqILV+gUiu1/Vhb4fSrlV/ft0zfv4BzmpYzQplbmRzdHJlYW0KZW5kb2JqCjE5IDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggNDE2ID4+CnN0cmVhbQp4nD1SS24FMQjbzym4QKXwT84zVXe9/7Y2M6/Sy4MJBGxDZsmSUPlSlVST1iPfemm3+N7ye2kgvlRUXcKWnJSIlPtCRtQR/OKMSQ9c09klu6XsiFpKFdO1XBp1DHXYxDrH3pd7j8d+vdDNCm9YK/BftiW2o2g81o0ReHEQ6RgUkf3CCj+DM4gX/fxlgojxC/kZ4ql4i8ggSHQ1IKYAFuexi9XoabAXmBtaMIm1lgsQR41w1o+9L76ip7ERV3xNetm85n3Q2GoWgZNghGaIooYbWUNNzxR1B9wS/SegSZGbQ6EHCNogCiGV1ZOPTdHASQM3BssxGSVzwKLnvifDa71vfNtU8QMwlOx8ZB4PQ/CN7TiIoG9B2Gdo5XizcAMZKAEtEKDz3AAJM4itlH2INvE16KvlwwcRfzT5HU/RTZ1xHIxDZ7G0uIQLnDfFtAXJODybFYOSS8CIfGFI99BxCNw+BHStmdzGVj6iaL0irdk2egDODNt4yrMovCdlw3wUJ9kkxqI5hYSn2EVmaOtIAriYtA0RUMPafCje188fiKGkNgplbmRzdHJlYW0KZW5kb2JqCjIwIDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggOTAgPj4Kc3RyZWFtCnicTY1BEsAgCAPvvoInGChS/tPpyf7/WpFx9EJ2EiCqjSpBxtB6k6HRgyIcxjcVBuoFB7DyABGf671cwEGZxrNNeRrppho/Zk9qbGejmg7PfRXxqnx/MdkhKQplbmRzdHJlYW0KZW5kb2JqCjIxIDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMTE0ID4+CnN0cmVhbQp4nDVOyw1DMQy6ZwpG8N/xPK/qKd3/WsdqLwZhQLgHCEzVV1ORXHjxupTwWbK98Qx6DAuFG0G0lTYLMawKz+JIWBZYAxY2peZ2P81cq9Psu3tkUl63ZSNE2yNpCHcoEWInlGPGPOs/6/xWnfX+Ai2WIl4KZW5kc3RyZWFtCmVuZG9iagoyMiAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDQyID4+CnN0cmVhbQp4nDMyt1AwULA0BBKGQNLQwEAhxZALzM/lggrkcBmisEA0lEoDAH7MDBIKZW5kc3RyZWFtCmVuZG9iagoyMyAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDE0NCA+PgpzdHJlYW0KeJxNjzGyAyEMQ/s9hY9gyTaw58mfX23u30aGFClAAs2TxxFhbqN0RU6rcPvDtd/vlsJtz4XMH5fp7YbcHIZFK3ejToH2uggVMY0xdiHrqJJ1XEDK70hvpn+S3ctbvUHjDOktmltFx3FESWNPTDaNJQZuEBPiMabFzE5KLmkoJb62cmInvflzdn2u/w/DNDGPCmVuZHN0cmVhbQplbmRvYmoKMjQgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAzNTMgPj4Kc3RyZWFtCnicPVI7ckQxCOvfKbhAZswfzrOZVNn7txE4SSUGI0DC7kGHtOiDmZwPJTd98mNd5Mb0fgxJV6Pvx46QR5E2SltJMylO0+tRVwpFRoAWJNWLr0eCb0YDFUXCToGq18M1bCWOIE+gzxyfF0yzDmJxMuwwqDFzGJ3sIJNCJkYCvtmZOZhu0CIBTuCle3F2y43sAA08GwXTzbClgjOy9ZxFUcHLRLOfI8uGjoWZ4IMzeooMi2rfXdaASDgI/b1GSsUi9mLZiH0UbpNVEiNLaB3hAV3pCCzHy2sJG8IZy17goTHXXkjaFkfevZl2Uo4dMCKSd1UcYZZsSIOdfycdg3uFvm+Es45I3iPUxqhJSOeEDUUgj46SNRoHTBJHc9hVPoA1hCeYj5CwpkA+08YO7giKCTJ5UWSdnCjnnkHKl6N2L2b4WzBHxhnMqosYbr+ZU1vB/xx22S4rKIv+JL6erx9TMIkxCmVuZHN0cmVhbQplbmRvYmoKMjUgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAyNjcgPj4Kc3RyZWFtCnicNVFJcsMwDLv7FXiCuIvvSaen5P/Xgsx0xjJocwOgiMSBNl9xDCWNH3kmvI7PBgfvx84C61RgBpd5UvB6wtBsVIg6+kI48fXMh9yCpEGVR3gymNHmBL0wN1gRa2ZNj7HbY4Yr/N7FUGVmI861bsQllnP5mR5vpDlMCukJrS9yj8X3zyFRUpZK5LnDLSj2krAVwuuLZ/cw4Kr3iIpqTAulC+llriySKA5UkfVKp4g4RGoj56AkTXfiGuRclbRgVgQnLuqQn8hph3fCaZTTc7dYIsycWjvGJFeK9rsm9Yq1cG4PjJ26JqnRfNbqofWslduUNhmhXaNBkqIP9YTO9b7+7/n9/P4BKx1g2AplbmRzdHJlYW0KZW5kb2JqCjI2IDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMTg5ID4+CnN0cmVhbQp4nE1QwW0AMQj7ZwovUClAIDBPq76u+39rkqvUBzIyYAPugYlY+BDBygWXwpeMNZv/ObgDz7BwSMJknlZdcvBz6MtIkmGPO4K0aEE2ZBZMUQGfZNMQib2QhWCDOMlWZg+XEK8DtZu+yzxDXP9lavJmkm1njJ6kl2Ubhx1doQt1pBQ86+65J8ePk14n0hRhYtyTVTO73ca/cNq2HjFLolUXiqeGwniSMBr7GV2Jd82/P95sNp53PuP7F/bNRIwKZW5kc3RyZWFtCmVuZG9iagoyNyAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDIwOCA+PgpzdHJlYW0KeJw1kMtxAzEMQ++qgg1ohuDqQ9bjTE5O/9cAkn0iVlg9Apru5oZlPbxsFGyi7AdthXVU2V+bUvzj3TaoApapOQz+XPFqiPicPVuCzFG6d8yli74N261XcpKQW1YO6/xCETWXhU9GAZ1wgXRCTlngYVCxTsJ3A/BRAdkp9TyGdIsxuZ+TRK4jbC0mS4t9l0cFryw6gx0u5luemGKow9tstcbhdLUVmO2mkKF+kybUbnNjSWQppZ4qPJhi36cFyUfQdFyZoTnt+9av9vsP1sVKIAplbmRzdHJlYW0KZW5kb2JqCjI4IDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMjcyID4+CnN0cmVhbQp4nDVRS24FMQjbzyl8gUr8Sc4zVXe9/7Ym6ZNmBAnYGCezIAjDlypSFlo3vvXhjafjd5LwZolfohYyG++zmMuGElXBGDz3FLQ3mp1mfris88T3cb03Hs2o8C7UTlZCguhNGQtF+mBHMWelCrlZWVRXgdiGNGMlhcKWHM3BWRNH8VQWrIghzkTZZdS3D1tik943kiiqdBlhreC61seEETzxfUxupsnfuatRFe/JoqZjmukM/5+r/vFsMvM8rij30R70OpQCLmrOheWMqqNAT+KxpjrG3PYxZgiCpyGceIdNFtQ9HRkXk2swQ1JWcOWMKA65kcPd7w1NKeOj8cHyPIDS8Dxb0tQWfFZ5n58/9JZlSwplbmRzdHJlYW0KZW5kb2JqCjE0IDAgb2JqCjw8IC9CYXNlRm9udCAvQXJpYWxNVCAvQ2hhclByb2NzIDE1IDAgUgovRW5jb2RpbmcgPDwKL0RpZmZlcmVuY2VzIFsgNDYgL3BlcmlvZCA0OCAvemVybyAvb25lIC90d28gNTIgL2ZvdXIgNTQgL3NpeCA1NiAvZWlnaHQgNjUgL0EgOTcgL2EgOTkKL2MgMTE0IC9yIDExNyAvdSAxMjEgL3kgXQovVHlwZSAvRW5jb2RpbmcgPj4KL0ZpcnN0Q2hhciAwIC9Gb250QkJveCBbIC02NjUgLTMyNSAyMDAwIDEwMDYgXSAvRm9udERlc2NyaXB0b3IgMTMgMCBSCi9Gb250TWF0cml4IFsgMC4wMDEgMCAwIDAuMDAxIDAgMCBdIC9MYXN0Q2hhciAyNTUgL05hbWUgL0FyaWFsTVQKL1N1YnR5cGUgL1R5cGUzIC9UeXBlIC9Gb250IC9XaWR0aHMgMTIgMCBSID4+CmVuZG9iagoxMyAwIG9iago8PCAvQXNjZW50IDkwNiAvQ2FwSGVpZ2h0IDcxNiAvRGVzY2VudCAtMjEyIC9GbGFncyAzMgovRm9udEJCb3ggWyAtNjY1IC0zMjUgMjAwMCAxMDA2IF0gL0ZvbnROYW1lIC9BcmlhbE1UIC9JdGFsaWNBbmdsZSAwCi9NYXhXaWR0aCAxMDE1IC9TdGVtViAwIC9UeXBlIC9Gb250RGVzY3JpcHRvciAvWEhlaWdodCA1MTkgPj4KZW5kb2JqCjEyIDAgb2JqClsgNzUwIDc1MCA3NTAgNzUwIDc1MCA3NTAgNzUwIDc1MCA3NTAgNzUwIDc1MCA3NTAgNzUwIDc1MCA3NTAgNzUwIDc1MCA3NTAKNzUwIDc1MCA3NTAgNzUwIDc1MCA3NTAgNzUwIDc1MCA3NTAgNzUwIDc1MCA3NTAgNzUwIDc1MCAyNzggMjc4IDM1NSA1NTYgNTU2Cjg4OSA2NjcgMTkxIDMzMyAzMzMgMzg5IDU4NCAyNzggMzMzIDI3OCAyNzggNTU2IDU1NiA1NTYgNTU2IDU1NiA1NTYgNTU2IDU1Ngo1NTYgNTU2IDI3OCAyNzggNTg0IDU4NCA1ODQgNTU2IDEwMTUgNjY3IDY2NyA3MjIgNzIyIDY2NyA2MTEgNzc4IDcyMiAyNzgKNTAwIDY2NyA1NTYgODMzIDcyMiA3NzggNjY3IDc3OCA3MjIgNjY3IDYxMSA3MjIgNjY3IDk0NCA2NjcgNjY3IDYxMSAyNzggMjc4CjI3OCA0NjkgNTU2IDMzMyA1NTYgNTU2IDUwMCA1NTYgNTU2IDI3OCA1NTYgNTU2IDIyMiAyMjIgNTAwIDIyMiA4MzMgNTU2IDU1Ngo1NTYgNTU2IDMzMyA1MDAgMjc4IDU1NiA1MDAgNzIyIDUwMCA1MDAgNTAwIDMzNCAyNjAgMzM0IDU4NCA3NTAgNTU2IDc1MCAyMjIKNTU2IDMzMyAxMDAwIDU1NiA1NTYgMzMzIDEwMDAgNjY3IDMzMyAxMDAwIDc1MCA2MTEgNzUwIDc1MCAyMjIgMjIyIDMzMyAzMzMKMzUwIDU1NiAxMDAwIDMzMyAxMDAwIDUwMCAzMzMgOTQ0IDc1MCA1MDAgNjY3IDI3OCAzMzMgNTU2IDU1NiA1NTYgNTU2IDI2MAo1NTYgMzMzIDczNyAzNzAgNTU2IDU4NCAzMzMgNzM3IDU1MiA0MDAgNTQ5IDMzMyAzMzMgMzMzIDU3NiA1MzcgMzMzIDMzMyAzMzMKMzY1IDU1NiA4MzQgODM0IDgzNCA2MTEgNjY3IDY2NyA2NjcgNjY3IDY2NyA2NjcgMTAwMCA3MjIgNjY3IDY2NyA2NjcgNjY3CjI3OCAyNzggMjc4IDI3OCA3MjIgNzIyIDc3OCA3NzggNzc4IDc3OCA3NzggNTg0IDc3OCA3MjIgNzIyIDcyMiA3MjIgNjY3IDY2Nwo2MTEgNTU2IDU1NiA1NTYgNTU2IDU1NiA1NTYgODg5IDUwMCA1NTYgNTU2IDU1NiA1NTYgMjc4IDI3OCAyNzggMjc4IDU1NiA1NTYKNTU2IDU1NiA1NTYgNTU2IDU1NiA1NDkgNjExIDU1NiA1NTYgNTU2IDU1NiA1MDAgNTU2IDUwMCBdCmVuZG9iagoxNSAwIG9iago8PCAvQSAxNiAwIFIgL2EgMTcgMCBSIC9jIDE4IDAgUiAvZWlnaHQgMTkgMCBSIC9mb3VyIDIwIDAgUiAvb25lIDIxIDAgUgovcGVyaW9kIDIyIDAgUiAvciAyMyAwIFIgL3NpeCAyNCAwIFIgL3R3byAyNSAwIFIgL3UgMjYgMCBSIC95IDI3IDAgUgovemVybyAyOCAwIFIgPj4KZW5kb2JqCjMgMCBvYmoKPDwgL0YxIDE0IDAgUiA+PgplbmRvYmoKNCAwIG9iago8PCAvQTEgPDwgL0NBIDAgL1R5cGUgL0V4dEdTdGF0ZSAvY2EgMSA+PgovQTIgPDwgL0NBIDEgL1R5cGUgL0V4dEdTdGF0ZSAvY2EgMSA+PiA+PgplbmRvYmoKNSAwIG9iago8PCA+PgplbmRvYmoKNiAwIG9iago8PCA+PgplbmRvYmoKNyAwIG9iago8PCA+PgplbmRvYmoKMiAwIG9iago8PCAvQ291bnQgMSAvS2lkcyBbIDEwIDAgUiBdIC9UeXBlIC9QYWdlcyA+PgplbmRvYmoKMjkgMCBvYmoKPDwgL0NyZWF0aW9uRGF0ZSAoRDoyMDIxMTAyMjEwNDcwMC0wNCcwMCcpCi9DcmVhdG9yIChtYXRwbG90bGliIDMuMC4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcpCi9Qcm9kdWNlciAobWF0cGxvdGxpYiBwZGYgYmFja2VuZCAzLjAuMikgPj4KZW5kb2JqCnhyZWYKMCAzMAowMDAwMDAwMDAwIDY1NTM1IGYgCjAwMDAwMDAwMTYgMDAwMDAgbiAKMDAwMDAwODI0MSAwMDAwMCBuIAowMDAwMDA4MDQ3IDAwMDAwIG4gCjAwMDAwMDgwNzkgMDAwMDAgbiAKMDAwMDAwODE3OCAwMDAwMCBuIAowMDAwMDA4MTk5IDAwMDAwIG4gCjAwMDAwMDgyMjAgMDAwMDAgbiAKMDAwMDAwMDA2NSAwMDAwMCBuIAowMDAwMDAwMzk3IDAwMDAwIG4gCjAwMDAwMDAyMDggMDAwMDAgbiAKMDAwMDAwMjM4MyAwMDAwMCBuIAowMDAwMDA2ODIzIDAwMDAwIG4gCjAwMDAwMDY2MjMgMDAwMDAgbiAKMDAwMDAwNjI0MSAwMDAwMCBuIAowMDAwMDA3ODc0IDAwMDAwIG4gCjAwMDAwMDI0MDQgMDAwMDAgbiAKMDAwMDAwMjU4OSAwMDAwMCBuIAowMDAwMDAzMDk5IDAwMDAwIG4gCjAwMDAwMDM0MTggMDAwMDAgbiAKMDAwMDAwMzkwNyAwMDAwMCBuIAowMDAwMDA0MDY5IDAwMDAwIG4gCjAwMDAwMDQyNTYgMDAwMDAgbiAKMDAwMDAwNDM3MCAwMDAwMCBuIAowMDAwMDA0NTg3IDAwMDAwIG4gCjAwMDAwMDUwMTMgMDAwMDAgbiAKMDAwMDAwNTM1MyAwMDAwMCBuIAowMDAwMDA1NjE1IDAwMDAwIG4gCjAwMDAwMDU4OTYgMDAwMDAgbiAKMDAwMDAwODMwMSAwMDAwMCBuIAp0cmFpbGVyCjw8IC9JbmZvIDI5IDAgUiAvUm9vdCAxIDAgUiAvU2l6ZSAzMCA+PgpzdGFydHhyZWYKODQ1NQolJUVPRgo=\n",
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x252 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "application/pdf": "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\n",
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x252 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from utils_dataset import n_blocks\n",
    "import pickle\n",
    "\n",
    "model = 'hybridValue_full'\n",
    "modelLabels = ['Six-state', 'Four-state']\n",
    "\n",
    "# plot setting\n",
    "N = 10\n",
    "from IPython.display import set_matplotlib_formats\n",
    "set_matplotlib_formats('png', 'pdf')\n",
    "plt.rcParams.update({'font.family': 'arial'})\n",
    "lineWidth = 2\n",
    "    \n",
    "for trialType in ['free', 'forced']:\n",
    "    fig, ax = plt.subplots(figsize=(10,3.5))\n",
    "\n",
    "    for w4 in [1, 0]:\n",
    "        trialIndices = curveData['w4'+str(w4)][trialType, 'x']\n",
    "        y = curveData['w4'+str(w4)][trialType, 'y']\n",
    "        err = curveData['w4'+str(w4)][trialType, 'err']\n",
    "        ax.plot(trialIndices, y, linewidth=lineWidth, color='C'+str(2-w4), label=modelLabels[w4], lineWidth=lineWidth)\n",
    "\n",
    "    for blockChange in np.array([N*2+1, N*4+2, N*6+3])+0.5:\n",
    "        ax.axvline(x=blockChange, linestyle='--', color='gray', linewidth=lineWidth)\n",
    "    \n",
    "    ax.set_ylim([0, 1])\n",
    "    ax.set_xlim([0, N*8+4])\n",
    "    ax.set_xlabel('Trial')\n",
    "    ax.set_ylabel('Accuracy')\n",
    "    ax.spines['top'].set_visible(False)\n",
    "    ax.spines['right'].set_visible(False)\n",
    "    ax.set_xticklabels('')\n",
    "    ax.tick_params(axis='x', length=0)\n",
    "    ax.set_xlabel('')\n",
    "    ax.tick_params(axis='y', width=1.5, pad=5, direction='out')\n",
    "    ax.set_yticks([0, 0.2, 0.4, 0.6, 0.8, 1])\n",
    "    ax.set_yticklabels([0, 0.2, 0.4, 0.6, 0.8, 1], fontsize=20)\n",
    "    ax.set_ylabel('Accuracy', fontsize=25)\n",
    "    ax.yaxis.labelpad = 10\n",
    "    ax.spines['left'].set_linewidth(1.5)\n",
    "    ax.spines['bottom'].set_linewidth(1.5)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}