SI_4_OOS_predicted_behavioural_metrics.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "import os\n",
    "\n",
    "from bambi import Model\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib.patches as mpatches\n",
    "%matplotlib inline\n",
    "\n",
    "from analysis_functions import make_sure_path_exists\n",
    "from plotting_functions import plot_correlation, cm2inch, add_regression_line"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "make_sure_path_exists('results/descriptives')\n",
    "make_sure_path_exists('results/figures')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "np.random.seed(16) # from random.org"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Load subject summary"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr:last-of-type th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th colspan=\"8\" halign=\"left\">rt</th>\n",
       "      <th>best_chosen</th>\n",
       "      <th>gaze_influence</th>\n",
       "      <th>dataset</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>min</th>\n",
       "      <th>max</th>\n",
       "      <th>se</th>\n",
       "      <th>q1</th>\n",
       "      <th>q3</th>\n",
       "      <th>iqr</th>\n",
       "      <th>mean</th>\n",
       "      <th>Unnamed: 10_level_1</th>\n",
       "      <th>Unnamed: 11_level_1</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subject</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1434.192000</td>\n",
       "      <td>538.289384</td>\n",
       "      <td>419.0</td>\n",
       "      <td>4090.0</td>\n",
       "      <td>11.350659</td>\n",
       "      <td>1058.00</td>\n",
       "      <td>1681.00</td>\n",
       "      <td>623.00</td>\n",
       "      <td>0.904889</td>\n",
       "      <td>0.077317</td>\n",
       "      <td>krajbich2010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1443.249796</td>\n",
       "      <td>759.370108</td>\n",
       "      <td>259.0</td>\n",
       "      <td>5570.0</td>\n",
       "      <td>15.344725</td>\n",
       "      <td>915.00</td>\n",
       "      <td>1754.75</td>\n",
       "      <td>839.75</td>\n",
       "      <td>0.640000</td>\n",
       "      <td>0.216067</td>\n",
       "      <td>krajbich2010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>4486.572000</td>\n",
       "      <td>2976.300751</td>\n",
       "      <td>539.0</td>\n",
       "      <td>33182.0</td>\n",
       "      <td>59.537924</td>\n",
       "      <td>2416.75</td>\n",
       "      <td>5643.50</td>\n",
       "      <td>3226.75</td>\n",
       "      <td>0.836000</td>\n",
       "      <td>0.007773</td>\n",
       "      <td>krajbich2010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1966.197273</td>\n",
       "      <td>505.283550</td>\n",
       "      <td>903.0</td>\n",
       "      <td>4722.0</td>\n",
       "      <td>10.775131</td>\n",
       "      <td>1648.00</td>\n",
       "      <td>2185.00</td>\n",
       "      <td>537.00</td>\n",
       "      <td>0.656818</td>\n",
       "      <td>0.128273</td>\n",
       "      <td>krajbich2010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1650.926400</td>\n",
       "      <td>719.954680</td>\n",
       "      <td>514.0</td>\n",
       "      <td>5851.0</td>\n",
       "      <td>14.401974</td>\n",
       "      <td>1190.00</td>\n",
       "      <td>1913.00</td>\n",
       "      <td>723.00</td>\n",
       "      <td>0.744800</td>\n",
       "      <td>0.387967</td>\n",
       "      <td>krajbich2010</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  rt                                                   \\\n",
       "                mean          std    min      max         se       q1   \n",
       "subject                                                                 \n",
       "0        1434.192000   538.289384  419.0   4090.0  11.350659  1058.00   \n",
       "1        1443.249796   759.370108  259.0   5570.0  15.344725   915.00   \n",
       "2        4486.572000  2976.300751  539.0  33182.0  59.537924  2416.75   \n",
       "3        1966.197273   505.283550  903.0   4722.0  10.775131  1648.00   \n",
       "4        1650.926400   719.954680  514.0   5851.0  14.401974  1190.00   \n",
       "\n",
       "                          best_chosen      gaze_influence             dataset  \n",
       "              q3      iqr        mean Unnamed: 10_level_1 Unnamed: 11_level_1  \n",
       "subject                                                                        \n",
       "0        1681.00   623.00    0.904889            0.077317        krajbich2010  \n",
       "1        1754.75   839.75    0.640000            0.216067        krajbich2010  \n",
       "2        5643.50  3226.75    0.836000            0.007773        krajbich2010  \n",
       "3        2185.00   537.00    0.656818            0.128273        krajbich2010  \n",
       "4        1913.00   723.00    0.744800            0.387967        krajbich2010  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "multiplicative_odd_summary = pd.read_csv(os.path.join('results', 'predictions', 'out_of_sample', 'descriptives',\n",
    "                                                      'subject_summary_multiplicative_oos.csv'),\n",
    "                                         index_col=[0], header=[0, 1])\n",
    "multiplicative_odd_summary.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Run mixed effects regression, accounting for datasets\n",
    "## Does GLAM qualitatively recover empirically observed behavioral associations?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "n_samples = 20000"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# create dataframe for mfx models\n",
    "df = pd.DataFrame(dict(subject=multiplicative_odd_summary.index, # index equals subject ID\n",
    "                       dataset=multiplicative_odd_summary['dataset']['Unnamed: 11_level_1'],\n",
    "                       mean_rt=multiplicative_odd_summary['rt']['mean'],\n",
    "                       best_chosen=100 * multiplicative_odd_summary['best_chosen']['mean'],\n",
    "                       gaze_influence=100 * multiplicative_odd_summary['gaze_influence']['Unnamed: 10_level_1']))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Previous results loaded from \"results/mixed_effects_models/si3_predicted_individual_differences_a_best_chosen.csv\"...\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>mean</th>\n",
       "      <th>sd</th>\n",
       "      <th>hpd0.95_lower</th>\n",
       "      <th>hpd0.95_upper</th>\n",
       "      <th>effective_n</th>\n",
       "      <th>gelman_rubin</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1|dataset[folke2016]</th>\n",
       "      <td>-3.850017</td>\n",
       "      <td>6.037433</td>\n",
       "      <td>-1.691205e+01</td>\n",
       "      <td>7.211606</td>\n",
       "      <td>13357</td>\n",
       "      <td>1.000146</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1|dataset[krajbich2010]</th>\n",
       "      <td>2.607832</td>\n",
       "      <td>5.068189</td>\n",
       "      <td>-7.289657e+00</td>\n",
       "      <td>13.346603</td>\n",
       "      <td>9021</td>\n",
       "      <td>1.000533</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1|dataset[krajbich2011]</th>\n",
       "      <td>-3.795367</td>\n",
       "      <td>4.978577</td>\n",
       "      <td>-1.458692e+01</td>\n",
       "      <td>5.549114</td>\n",
       "      <td>8718</td>\n",
       "      <td>1.000163</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1|dataset[tavares2017]</th>\n",
       "      <td>5.405715</td>\n",
       "      <td>5.500875</td>\n",
       "      <td>-4.156288e+00</td>\n",
       "      <td>16.980059</td>\n",
       "      <td>9761</td>\n",
       "      <td>1.000823</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1|dataset_sd</th>\n",
       "      <td>7.412221</td>\n",
       "      <td>4.592130</td>\n",
       "      <td>7.640197e-04</td>\n",
       "      <td>16.027238</td>\n",
       "      <td>10633</td>\n",
       "      <td>1.000276</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Intercept</th>\n",
       "      <td>78.914418</td>\n",
       "      <td>4.894836</td>\n",
       "      <td>6.890519e+01</td>\n",
       "      <td>88.872204</td>\n",
       "      <td>6775</td>\n",
       "      <td>1.000607</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>best_chosen_sd</th>\n",
       "      <td>10.158341</td>\n",
       "      <td>0.690398</td>\n",
       "      <td>8.840166e+00</td>\n",
       "      <td>11.530310</td>\n",
       "      <td>31408</td>\n",
       "      <td>1.000099</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean_rt</th>\n",
       "      <td>-0.001133</td>\n",
       "      <td>0.001590</td>\n",
       "      <td>-4.126580e-03</td>\n",
       "      <td>0.002291</td>\n",
       "      <td>9556</td>\n",
       "      <td>0.999978</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean_rt|dataset[folke2016]</th>\n",
       "      <td>-0.001203</td>\n",
       "      <td>0.001663</td>\n",
       "      <td>-4.966469e-03</td>\n",
       "      <td>0.001598</td>\n",
       "      <td>9537</td>\n",
       "      <td>0.999980</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean_rt|dataset[krajbich2010]</th>\n",
       "      <td>0.000909</td>\n",
       "      <td>0.001677</td>\n",
       "      <td>-2.068267e-03</td>\n",
       "      <td>0.004774</td>\n",
       "      <td>13639</td>\n",
       "      <td>0.999978</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean_rt|dataset[krajbich2011]</th>\n",
       "      <td>-0.000833</td>\n",
       "      <td>0.001691</td>\n",
       "      <td>-4.717872e-03</td>\n",
       "      <td>0.002174</td>\n",
       "      <td>10919</td>\n",
       "      <td>1.000062</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean_rt|dataset[tavares2017]</th>\n",
       "      <td>0.000959</td>\n",
       "      <td>0.001947</td>\n",
       "      <td>-2.660421e-03</td>\n",
       "      <td>0.005320</td>\n",
       "      <td>15528</td>\n",
       "      <td>1.000048</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean_rt|dataset_sd</th>\n",
       "      <td>0.002260</td>\n",
       "      <td>0.001741</td>\n",
       "      <td>1.198212e-07</td>\n",
       "      <td>0.005646</td>\n",
       "      <td>11060</td>\n",
       "      <td>0.999977</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                    mean        sd  hpd0.95_lower  \\\n",
       "1|dataset[folke2016]           -3.850017  6.037433  -1.691205e+01   \n",
       "1|dataset[krajbich2010]         2.607832  5.068189  -7.289657e+00   \n",
       "1|dataset[krajbich2011]        -3.795367  4.978577  -1.458692e+01   \n",
       "1|dataset[tavares2017]          5.405715  5.500875  -4.156288e+00   \n",
       "1|dataset_sd                    7.412221  4.592130   7.640197e-04   \n",
       "Intercept                      78.914418  4.894836   6.890519e+01   \n",
       "best_chosen_sd                 10.158341  0.690398   8.840166e+00   \n",
       "mean_rt                        -0.001133  0.001590  -4.126580e-03   \n",
       "mean_rt|dataset[folke2016]     -0.001203  0.001663  -4.966469e-03   \n",
       "mean_rt|dataset[krajbich2010]   0.000909  0.001677  -2.068267e-03   \n",
       "mean_rt|dataset[krajbich2011]  -0.000833  0.001691  -4.717872e-03   \n",
       "mean_rt|dataset[tavares2017]    0.000959  0.001947  -2.660421e-03   \n",
       "mean_rt|dataset_sd              0.002260  0.001741   1.198212e-07   \n",
       "\n",
       "                               hpd0.95_upper  effective_n  gelman_rubin  \n",
       "1|dataset[folke2016]                7.211606        13357      1.000146  \n",
       "1|dataset[krajbich2010]            13.346603         9021      1.000533  \n",
       "1|dataset[krajbich2011]             5.549114         8718      1.000163  \n",
       "1|dataset[tavares2017]             16.980059         9761      1.000823  \n",
       "1|dataset_sd                       16.027238        10633      1.000276  \n",
       "Intercept                          88.872204         6775      1.000607  \n",
       "best_chosen_sd                     11.530310        31408      1.000099  \n",
       "mean_rt                             0.002291         9556      0.999978  \n",
       "mean_rt|dataset[folke2016]          0.001598         9537      0.999980  \n",
       "mean_rt|dataset[krajbich2010]       0.004774        13639      0.999978  \n",
       "mean_rt|dataset[krajbich2011]       0.002174        10919      1.000062  \n",
       "mean_rt|dataset[tavares2017]        0.005320        15528      1.000048  \n",
       "mean_rt|dataset_sd                  0.005646        11060      0.999977  "
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "result_path = os.path.join('results', 'mixed_effects_models', 'si3_predicted_individual_differences_a_best_chosen.csv')\n",
    "if not os.path.isfile(result_path):\n",
    "    m = Model(df)\n",
    "    a_best_chosen_fit = m.fit('best_chosen ~ mean_rt',\n",
    "                              random=['mean_rt|dataset'],\n",
    "                              categorical=['dataset'],\n",
    "                              samples=n_samples,\n",
    "                              chains=2)\n",
    "    a_summary = a_best_chosen_fit.summary(ranefs=True)\n",
    "    a_summary.to_csv(result_path)\n",
    "else:\n",
    "    print('Previous results loaded from \"{}\"...'.format(result_path))\n",
    "    a_summary = pd.read_csv(result_path, index_col=0)\n",
    "a_summary"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Previous results loaded from \"results/mixed_effects_models/si3_predicted_individual_differences_b_mean_rt.csv\"...\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>mean</th>\n",
       "      <th>sd</th>\n",
       "      <th>hpd0.95_lower</th>\n",
       "      <th>hpd0.95_upper</th>\n",
       "      <th>effective_n</th>\n",
       "      <th>gelman_rubin</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1|dataset[folke2016]</th>\n",
       "      <td>2282.926492</td>\n",
       "      <td>911.790639</td>\n",
       "      <td>507.065636</td>\n",
       "      <td>4183.234992</td>\n",
       "      <td>151</td>\n",
       "      <td>1.006803</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1|dataset[krajbich2010]</th>\n",
       "      <td>-563.426890</td>\n",
       "      <td>854.904519</td>\n",
       "      <td>-2281.350344</td>\n",
       "      <td>1339.570610</td>\n",
       "      <td>118</td>\n",
       "      <td>1.012702</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1|dataset[krajbich2011]</th>\n",
       "      <td>-535.514348</td>\n",
       "      <td>855.098489</td>\n",
       "      <td>-2179.560165</td>\n",
       "      <td>1396.168825</td>\n",
       "      <td>123</td>\n",
       "      <td>1.011203</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1|dataset[tavares2017]</th>\n",
       "      <td>-915.237116</td>\n",
       "      <td>864.180156</td>\n",
       "      <td>-2714.940752</td>\n",
       "      <td>894.326109</td>\n",
       "      <td>129</td>\n",
       "      <td>1.012524</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1|dataset_sd</th>\n",
       "      <td>1639.233157</td>\n",
       "      <td>592.504242</td>\n",
       "      <td>668.523260</td>\n",
       "      <td>2845.189521</td>\n",
       "      <td>874</td>\n",
       "      <td>1.000054</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Intercept</th>\n",
       "      <td>2890.695623</td>\n",
       "      <td>841.690873</td>\n",
       "      <td>842.609265</td>\n",
       "      <td>4429.605635</td>\n",
       "      <td>110</td>\n",
       "      <td>1.013987</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>gaze_influence</th>\n",
       "      <td>-1.270321</td>\n",
       "      <td>16.280947</td>\n",
       "      <td>-34.109216</td>\n",
       "      <td>32.460239</td>\n",
       "      <td>272</td>\n",
       "      <td>1.005470</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>gaze_influence|dataset[folke2016]</th>\n",
       "      <td>4.429943</td>\n",
       "      <td>17.044439</td>\n",
       "      <td>-35.506268</td>\n",
       "      <td>40.701000</td>\n",
       "      <td>650</td>\n",
       "      <td>1.000895</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>gaze_influence|dataset[krajbich2010]</th>\n",
       "      <td>-10.576384</td>\n",
       "      <td>17.145366</td>\n",
       "      <td>-49.311211</td>\n",
       "      <td>19.545357</td>\n",
       "      <td>345</td>\n",
       "      <td>1.004976</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>gaze_influence|dataset[krajbich2011]</th>\n",
       "      <td>3.129911</td>\n",
       "      <td>16.432574</td>\n",
       "      <td>-30.801500</td>\n",
       "      <td>41.049461</td>\n",
       "      <td>338</td>\n",
       "      <td>1.002944</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>gaze_influence|dataset[tavares2017]</th>\n",
       "      <td>-0.586848</td>\n",
       "      <td>19.150089</td>\n",
       "      <td>-47.039035</td>\n",
       "      <td>36.899101</td>\n",
       "      <td>358</td>\n",
       "      <td>1.002708</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>gaze_influence|dataset_sd</th>\n",
       "      <td>23.279494</td>\n",
       "      <td>33.764545</td>\n",
       "      <td>0.000152</td>\n",
       "      <td>61.679723</td>\n",
       "      <td>91</td>\n",
       "      <td>1.013257</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean_rt_sd</th>\n",
       "      <td>1040.489770</td>\n",
       "      <td>70.532950</td>\n",
       "      <td>908.052308</td>\n",
       "      <td>1183.661591</td>\n",
       "      <td>31382</td>\n",
       "      <td>0.999985</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                             mean          sd  hpd0.95_lower  \\\n",
       "1|dataset[folke2016]                  2282.926492  911.790639     507.065636   \n",
       "1|dataset[krajbich2010]               -563.426890  854.904519   -2281.350344   \n",
       "1|dataset[krajbich2011]               -535.514348  855.098489   -2179.560165   \n",
       "1|dataset[tavares2017]                -915.237116  864.180156   -2714.940752   \n",
       "1|dataset_sd                          1639.233157  592.504242     668.523260   \n",
       "Intercept                             2890.695623  841.690873     842.609265   \n",
       "gaze_influence                          -1.270321   16.280947     -34.109216   \n",
       "gaze_influence|dataset[folke2016]        4.429943   17.044439     -35.506268   \n",
       "gaze_influence|dataset[krajbich2010]   -10.576384   17.145366     -49.311211   \n",
       "gaze_influence|dataset[krajbich2011]     3.129911   16.432574     -30.801500   \n",
       "gaze_influence|dataset[tavares2017]     -0.586848   19.150089     -47.039035   \n",
       "gaze_influence|dataset_sd               23.279494   33.764545       0.000152   \n",
       "mean_rt_sd                            1040.489770   70.532950     908.052308   \n",
       "\n",
       "                                      hpd0.95_upper  effective_n  gelman_rubin  \n",
       "1|dataset[folke2016]                    4183.234992          151      1.006803  \n",
       "1|dataset[krajbich2010]                 1339.570610          118      1.012702  \n",
       "1|dataset[krajbich2011]                 1396.168825          123      1.011203  \n",
       "1|dataset[tavares2017]                   894.326109          129      1.012524  \n",
       "1|dataset_sd                            2845.189521          874      1.000054  \n",
       "Intercept                               4429.605635          110      1.013987  \n",
       "gaze_influence                            32.460239          272      1.005470  \n",
       "gaze_influence|dataset[folke2016]         40.701000          650      1.000895  \n",
       "gaze_influence|dataset[krajbich2010]      19.545357          345      1.004976  \n",
       "gaze_influence|dataset[krajbich2011]      41.049461          338      1.002944  \n",
       "gaze_influence|dataset[tavares2017]       36.899101          358      1.002708  \n",
       "gaze_influence|dataset_sd                 61.679723           91      1.013257  \n",
       "mean_rt_sd                              1183.661591        31382      0.999985  "
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "result_path = os.path.join('results', 'mixed_effects_models', 'si3_predicted_individual_differences_b_mean_rt.csv')\n",
    "if not os.path.isfile(result_path):\n",
    "    m = Model(df)\n",
    "    b_mean_rt_fit = m.fit('mean_rt ~ gaze_influence',\n",
    "                          random=['gaze_influence|dataset'],\n",
    "                          categorical=['dataset'],\n",
    "                          samples=n_samples,\n",
    "                          chains=2)\n",
    "    b_summary = b_mean_rt_fit.summary(ranefs=True)\n",
    "    b_summary.to_csv(result_path)\n",
    "else:\n",
    "    print('Previous results loaded from \"{}\"...'.format(result_path))\n",
    "    b_summary = pd.read_csv(result_path, index_col=0)\n",
    "b_summary"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Previous results loaded from \"results/mixed_effects_models/si3_predicted_individual_differences_c_gaze_influence.csv\"...\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>mean</th>\n",
       "      <th>sd</th>\n",
       "      <th>hpd0.95_lower</th>\n",
       "      <th>hpd0.95_upper</th>\n",
       "      <th>effective_n</th>\n",
       "      <th>gelman_rubin</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1|dataset[folke2016]</th>\n",
       "      <td>-2.395006</td>\n",
       "      <td>4.096390</td>\n",
       "      <td>-11.389149</td>\n",
       "      <td>5.063947</td>\n",
       "      <td>10266</td>\n",
       "      <td>0.999993</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1|dataset[krajbich2010]</th>\n",
       "      <td>1.641229</td>\n",
       "      <td>3.510146</td>\n",
       "      <td>-4.876838</td>\n",
       "      <td>9.567186</td>\n",
       "      <td>7989</td>\n",
       "      <td>0.999978</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1|dataset[krajbich2011]</th>\n",
       "      <td>-1.747444</td>\n",
       "      <td>3.519306</td>\n",
       "      <td>-9.414014</td>\n",
       "      <td>5.017206</td>\n",
       "      <td>9191</td>\n",
       "      <td>0.999977</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1|dataset[tavares2017]</th>\n",
       "      <td>3.687209</td>\n",
       "      <td>4.364017</td>\n",
       "      <td>-3.281590</td>\n",
       "      <td>13.150224</td>\n",
       "      <td>8226</td>\n",
       "      <td>0.999995</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1|dataset_sd</th>\n",
       "      <td>4.970288</td>\n",
       "      <td>3.613081</td>\n",
       "      <td>0.000475</td>\n",
       "      <td>11.978790</td>\n",
       "      <td>7168</td>\n",
       "      <td>1.000008</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Intercept</th>\n",
       "      <td>84.844101</td>\n",
       "      <td>3.356356</td>\n",
       "      <td>77.708689</td>\n",
       "      <td>91.528262</td>\n",
       "      <td>7986</td>\n",
       "      <td>0.999983</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>best_chosen_sd</th>\n",
       "      <td>7.906527</td>\n",
       "      <td>0.533079</td>\n",
       "      <td>6.904450</td>\n",
       "      <td>8.979606</td>\n",
       "      <td>32225</td>\n",
       "      <td>0.999998</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>gaze_influence</th>\n",
       "      <td>-0.419106</td>\n",
       "      <td>0.158849</td>\n",
       "      <td>-0.738456</td>\n",
       "      <td>-0.078450</td>\n",
       "      <td>5927</td>\n",
       "      <td>1.000756</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>gaze_influence|dataset[folke2016]</th>\n",
       "      <td>-0.215301</td>\n",
       "      <td>0.184679</td>\n",
       "      <td>-0.606061</td>\n",
       "      <td>0.110609</td>\n",
       "      <td>7279</td>\n",
       "      <td>1.000420</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>gaze_influence|dataset[krajbich2010]</th>\n",
       "      <td>0.080187</td>\n",
       "      <td>0.162480</td>\n",
       "      <td>-0.259966</td>\n",
       "      <td>0.417209</td>\n",
       "      <td>6196</td>\n",
       "      <td>1.000744</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>gaze_influence|dataset[krajbich2011]</th>\n",
       "      <td>-0.154894</td>\n",
       "      <td>0.171505</td>\n",
       "      <td>-0.528219</td>\n",
       "      <td>0.163660</td>\n",
       "      <td>6509</td>\n",
       "      <td>1.000425</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>gaze_influence|dataset[tavares2017]</th>\n",
       "      <td>0.176770</td>\n",
       "      <td>0.195697</td>\n",
       "      <td>-0.180548</td>\n",
       "      <td>0.592000</td>\n",
       "      <td>7108</td>\n",
       "      <td>1.000687</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>gaze_influence|dataset_sd</th>\n",
       "      <td>0.289509</td>\n",
       "      <td>0.171813</td>\n",
       "      <td>0.000023</td>\n",
       "      <td>0.620943</td>\n",
       "      <td>6551</td>\n",
       "      <td>0.999984</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                           mean        sd  hpd0.95_lower  \\\n",
       "1|dataset[folke2016]                  -2.395006  4.096390     -11.389149   \n",
       "1|dataset[krajbich2010]                1.641229  3.510146      -4.876838   \n",
       "1|dataset[krajbich2011]               -1.747444  3.519306      -9.414014   \n",
       "1|dataset[tavares2017]                 3.687209  4.364017      -3.281590   \n",
       "1|dataset_sd                           4.970288  3.613081       0.000475   \n",
       "Intercept                             84.844101  3.356356      77.708689   \n",
       "best_chosen_sd                         7.906527  0.533079       6.904450   \n",
       "gaze_influence                        -0.419106  0.158849      -0.738456   \n",
       "gaze_influence|dataset[folke2016]     -0.215301  0.184679      -0.606061   \n",
       "gaze_influence|dataset[krajbich2010]   0.080187  0.162480      -0.259966   \n",
       "gaze_influence|dataset[krajbich2011]  -0.154894  0.171505      -0.528219   \n",
       "gaze_influence|dataset[tavares2017]    0.176770  0.195697      -0.180548   \n",
       "gaze_influence|dataset_sd              0.289509  0.171813       0.000023   \n",
       "\n",
       "                                      hpd0.95_upper  effective_n  gelman_rubin  \n",
       "1|dataset[folke2016]                       5.063947        10266      0.999993  \n",
       "1|dataset[krajbich2010]                    9.567186         7989      0.999978  \n",
       "1|dataset[krajbich2011]                    5.017206         9191      0.999977  \n",
       "1|dataset[tavares2017]                    13.150224         8226      0.999995  \n",
       "1|dataset_sd                              11.978790         7168      1.000008  \n",
       "Intercept                                 91.528262         7986      0.999983  \n",
       "best_chosen_sd                             8.979606        32225      0.999998  \n",
       "gaze_influence                            -0.078450         5927      1.000756  \n",
       "gaze_influence|dataset[folke2016]          0.110609         7279      1.000420  \n",
       "gaze_influence|dataset[krajbich2010]       0.417209         6196      1.000744  \n",
       "gaze_influence|dataset[krajbich2011]       0.163660         6509      1.000425  \n",
       "gaze_influence|dataset[tavares2017]        0.592000         7108      1.000687  \n",
       "gaze_influence|dataset_sd                  0.620943         6551      0.999984  "
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "result_path = os.path.join('results', 'mixed_effects_models', 'si3_predicted_individual_differences_c_gaze_influence.csv')\n",
    "if not os.path.isfile(result_path):\n",
    "    m = Model(df)\n",
    "    c_gaze_influence_fit = m.fit('best_chosen ~ gaze_influence', \n",
    "                                 random=['gaze_influence|dataset'],\n",
    "                                 categorical=['dataset'],\n",
    "                                 samples=n_samples,\n",
    "                                 chains=2)\n",
    "    c_summary = c_gaze_influence_fit.summary(ranefs=True)\n",
    "    c_summary.to_csv(result_path)\n",
    "else:\n",
    "    print('Previous results loaded from \"{}\"...'.format(result_path))\n",
    "    c_summary = pd.read_csv(result_path, index_col=0)\n",
    "c_summary"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Previous results loaded from \"results/mixed_effects_models/si3_predicted_individual_differences_c_gaze_influence_plotting.csv\"...\n"
     ]
    }
   ],
   "source": [
    "# we also have to estimate gaze-influence ~ P(best chosen) for regression line in Figure 2:\n",
    "result_path = os.path.join('results', 'mixed_effects_models', 'si3_predicted_individual_differences_c_gaze_influence_plotting.csv')\n",
    "if not os.path.isfile(result_path):\n",
    "    m = Model(df)\n",
    "    c_gaze_influence_plotting_fit = m.fit('gaze_influence ~ best_chosen', \n",
    "                                 random=['best_chosen|dataset'],\n",
    "                                 categorical=['dataset'],\n",
    "                                 samples=n_samples,\n",
    "                                 chains=2)\n",
    "    c_summary_plotting = c_gaze_influence_plotting_fit.summary(ranefs=True)\n",
    "    c_summary_plotting.to_csv(result_path)\n",
    "else:\n",
    "    print('Previous results loaded from \"{}\"...'.format(result_path))\n",
    "    c_summary_plotting = pd.read_csv(result_path, index_col=0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Supplementary Figrue 3 (reproduces Fig. 2 of the main text)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "def figure_2(subject_summary, n_bins=20, figsize=cm2inch(18, 10)):\n",
    "    \n",
    "    fig = plt.figure(figsize=figsize)\n",
    "\n",
    "    ax00 = plt.subplot2grid((11, 3), (0, 0))\n",
    "    ax10 = plt.subplot2grid((11, 3), (1, 0))\n",
    "    ax20 = plt.subplot2grid((11, 3), (2, 0))\n",
    "    ax30 = plt.subplot2grid((11, 3), (3, 0))\n",
    "    ax40 = plt.subplot2grid((11, 3), (5, 0), rowspan=6)\n",
    "    axs0 = np.array([ax00, ax10, ax20, ax30, ax40])\n",
    "\n",
    "    ax01 = plt.subplot2grid((11, 3), (0, 1))\n",
    "    ax11 = plt.subplot2grid((11, 3), (1, 1))\n",
    "    ax21 = plt.subplot2grid((11, 3), (2, 1))\n",
    "    ax31 = plt.subplot2grid((11, 3), (3, 1))\n",
    "    ax41 = plt.subplot2grid((11, 3), (5, 1), rowspan=6)\n",
    "    axs1 = np.array([ax01, ax11, ax21, ax31, ax41])\n",
    "\n",
    "    ax02 = plt.subplot2grid((11, 3), (0, 2))\n",
    "    ax12 = plt.subplot2grid((11, 3), (1, 2))\n",
    "    ax22 = plt.subplot2grid((11, 3), (2, 2))\n",
    "    ax32 = plt.subplot2grid((11, 3), (3, 2))\n",
    "    ax42 = plt.subplot2grid((11, 3), (5, 2), rowspan=6)\n",
    "    axs2 = np.array([ax02, ax12, ax22, ax32, ax42])\n",
    "\n",
    "    datasets = ['krajbich2010', 'krajbich2011', 'folke2016', 'tavares2017']\n",
    "\n",
    "    # Correlation plots\n",
    "    for d, (dataset, marker, markercolor) in enumerate(zip(datasets, ['+', '+', '+', '+'], ['C0', 'C1', 'C2', 'C3'])):\n",
    "        dataset_summary = subject_summary[subject_summary['dataset']['Unnamed: 11_level_1'] == dataset].reset_index(drop=True)\n",
    "        plot_correlation(dataset_summary['rt']['mean'],\n",
    "                         dataset_summary['best_chosen']['mean'],\n",
    "                         marker=marker,\n",
    "                         markercolor=markercolor,\n",
    "                         regression=False,\n",
    "                         xlabel='Mean RT (ms)',\n",
    "                         ylabel='P(choose best)',\n",
    "                         xlim=(0, 8500),\n",
    "                         xticks=np.arange(0, 9000, 2000),\n",
    "                         ylim=(0.20, 1.0),\n",
    "                         yticks=np.arange(0.4, 1.1, 0.2),\n",
    "                         ax=ax40)\n",
    "\n",
    "        plot_correlation(dataset_summary['gaze_influence']['Unnamed: 10_level_1'],\n",
    "                         dataset_summary['rt']['mean'],\n",
    "                         marker=marker,\n",
    "                         markercolor=markercolor,\n",
    "                         regression=False,\n",
    "                         ylabel='Mean RT (ms)',\n",
    "                         xlabel='Gaze influence\\non P(choice | value)',\n",
    "                         ylim=(0, 8500),\n",
    "                         yticks=np.arange(0, 9000, 2000),\n",
    "                         xlim=(-0.25, 0.85),\n",
    "                         xticks=np.arange(-0.2, 0.9, 0.2),\n",
    "                         ax=ax41)\n",
    "        # fix xticks\n",
    "        ax41.set_xticks(np.round(np.arange(-0.2, 0.9, 0.2), 1))\n",
    "        ax41.set_xticklabels(np.round(np.arange(-0.2, 0.9, 0.2), 1))\n",
    "        \n",
    "        plot_correlation(dataset_summary['best_chosen']['mean'],\n",
    "                         dataset_summary['gaze_influence'],\n",
    "                         marker=marker,\n",
    "                         markercolor=markercolor,\n",
    "                         regression=False,\n",
    "                             ylabel='Gaze influence\\non P(choice | value)',\n",
    "                         xlabel='P(choose best)',\n",
    "                         ylim=(-0.25, 0.85),\n",
    "                         yticks=np.round(np.arange(-0.2, 0.9, 0.2), 1),\n",
    "                         xlim=(0.2, 1),\n",
    "                         xticks=np.round(np.arange(0.4, 1.1, 0.2), 1),\n",
    "                         ax=ax42)\n",
    "        # fix yticks\n",
    "        ax42.set_yticks(np.round(np.arange(-0.2, 0.9, 0.2), 1))\n",
    "        ax42.set_yticklabels(np.round(np.arange(-0.2, 0.9, 0.2), 1))\n",
    "\n",
    "        # Marginal histograms\n",
    "        axs0[d].hist(dataset_summary['rt']['mean'],\n",
    "                     bins=np.linspace(0, 8500, n_bins + 1),\n",
    "                     color=markercolor)\n",
    "\n",
    "        axs1[d].hist(dataset_summary['gaze_influence']['Unnamed: 10_level_1'],\n",
    "                     bins=np.linspace(-0.2, 0.8, n_bins + 1),\n",
    "                     color=markercolor)\n",
    "\n",
    "        axs2[d].hist(dataset_summary['best_chosen']['mean'],\n",
    "                     bins=np.linspace(0.3, 1, n_bins + 1),\n",
    "                     color=markercolor)\n",
    "\n",
    "        for axs in [axs0, axs1, axs2]:\n",
    "            axs[d].set_xlim(axs[-1].get_xlim())\n",
    "\n",
    "    patches = [mpatches.Rectangle((0, 0), 0.5, 1, fc=\"C{}\".format(i))\n",
    "               for i in range(len(datasets))]\n",
    "\n",
    "    ax40.legend(patches,\n",
    "                ['Krajbich 2010', 'Krajbich 2011', 'Folke 2016', 'Tavares 2017'],\n",
    "                ncol=2,\n",
    "                loc='lower left',\n",
    "                handlelength=1, handleheight=1,\n",
    "                fontsize=6, columnspacing=1,\n",
    "                frameon=False)\n",
    "\n",
    "    # Labels\n",
    "    for label, ax in zip(list('abc'), [ax00, ax01, ax02]):\n",
    "        ax.text(-0.35, 1.15, label, transform=ax.transAxes,\n",
    "                fontsize=7, fontweight='bold', va='top')\n",
    "    for label, ax in zip(list('def'), [ax40, ax41, ax42]):\n",
    "        ax.text(-0.35, 1.05, label, transform=ax.transAxes,\n",
    "                fontsize=7, fontweight='bold', va='top') \n",
    "        \n",
    "    # Fine-tune marginal histograms\n",
    "    for ax in np.array([ax00, ax10, ax20, ax30,\n",
    "                        ax01, ax11, ax21, ax31,\n",
    "                        ax02, ax12, ax22, ax32]):\n",
    "        ax.set_xticks([])\n",
    "        ax.set_ylim([0,10])\n",
    "        ax.set_yticks([0,10])\n",
    "        ax.set_yticklabels([0,10], fontsize=6)\n",
    "    ax20.set_ylabel('          Frequency', fontsize=6)\n",
    "    ax21.set_ylabel('          Frequency', fontsize=6)\n",
    "    ax22.set_ylabel('          Frequency', fontsize=6)\n",
    "    for ax in [ax00, ax10, ax20, ax30]:\n",
    "        ax.set_xticks(np.arange(0, 9000, 2000))\n",
    "        ax.set_xticklabels([])\n",
    "    ax30.set_xticklabels(np.arange(0, 9000, 2000).astype(np.int), fontsize=6)\n",
    "    ax30.set_xlabel('Mean RT (ms)', fontsize=6)\n",
    "    for ax in [ax01, ax11, ax21, ax31]:\n",
    "        ax.set_xticks(np.arange(-0.2, 0.9, 0.2))\n",
    "        ax.set_xticklabels([])\n",
    "    ax31.set_xticklabels(np.round(np.arange(-0.2, 0.9, 0.2), 1), fontsize=6)\n",
    "    ax31.set_xlabel('Gaze Influence\\non P(choice | value)', fontsize=6)\n",
    "    for ax in [ax02, ax12, ax22, ax32]:\n",
    "        ax.set_xticks(np.arange(0.4, 1.1, 0.2))\n",
    "        ax.set_xticklabels([])\n",
    "    ax32.set_xticklabels(np.round(np.arange(0.4, 1.1, 0.2), 1), fontsize=6)\n",
    "    ax32.set_xlabel('P(choose best)', fontsize=6)\n",
    "    \n",
    "    fig.subplots_adjust(wspace=0.5, hspace=1.05)\n",
    "    \n",
    "    return fig, ax40, ax41, ax42"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Make Figure SI 3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAcIAAAEFCAYAAABw9edJAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nVYQWi8ViadP8f0COHIW6PAy/AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x133c25438>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axa, axb, axc = figure_2(subject_summary=multiplicative_odd_summary)\n",
    "\n",
    "add_regression_line(axa, a_summary.loc['Intercept', 'mean'] / 100, a_summary.loc['mean_rt', 'mean'] / 100, color='darkgray', linewidth=1)\n",
    "add_regression_line(axb, b_summary.loc['Intercept', 'mean'], b_summary.loc['gaze_influence', 'mean'] * 100, color='darkgray', linewidth=1)\n",
    "add_regression_line(axc, c_summary.loc['Intercept', 'mean'] / 100, c_summary_plotting.loc['best_chosen', 'mean'], color='darkgray', linewidth=1)\n",
    "\n",
    "plt.savefig('results/figures/si_figure_3_predicted_individual_differences.png', dpi=330, bbox_inches='tight')\n",
    "plt.savefig('results/figures/si_figure_3_predicted_individual_differences.pdf', dpi=330, bbox_inches='tight')\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python [Root]",
   "language": "python",
   "name": "Python [Root]"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}