SI_3_additive-vs-multiplictive_group-averaged.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "import os\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "\n",
    "import glam"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Data I/O"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "datasets = ['krajbich2010', 'krajbich2011', 'folke2016', 'tavares2017']\n",
    "dataset_labels = ['Krajbich 2010', 'Krajbich 2011', 'Folke 2016', 'Tavares 2017']\n",
    "N = [2,3,3,2]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "data_all = pd.read_csv('data/data_all.csv')\n",
    "data_odd = data_all[(data_all['trial']%2)==1].copy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def dataset_iterator(datasets=['krajbich2010', 'krajbich2011', 'folke2016', 'tavares2017'],\n",
    "                     dataset_labels=['Krajbich 2010', 'Krajbich 2011', 'Folke 2016', 'Tavares 2017'],\n",
    "                     N=[2,3,3,2],\n",
    "                     path_data='data/',\n",
    "                     path_pred='results/predictions/out_of_sample/'):\n",
    "    \n",
    "    subjects = np.unique([np.int(s.split('_')[1]) for s in os.listdir(path_pred+'multiplicative/')])\n",
    "    \n",
    "    data_odd = pd.read_csv(path_data+'data_aggregate.csv')\n",
    "    data_odd = data_odd[(data_odd['trial']%2) == 1]\n",
    "\n",
    "    pred_multiplicative = []\n",
    "    pred_additive = []\n",
    "    for subject in subjects: \n",
    "        pred_multiplicative.append(\n",
    "            pd.read_csv(path_pred+'multiplicative/prediction_{}_multiplicative_oos.csv'.format(subject)))\n",
    "        pred_additive.append(\n",
    "            pd.read_csv(path_pred+'additive/prediction_{}_additive_oos.csv'.format(subject)))\n",
    "    pred_multiplicative = pd.concat(pred_multiplicative, sort=False)\n",
    "    pred_additive = pd.concat(pred_additive, sort=False)\n",
    "    \n",
    "    for di, (dataset, n, dataset_label) in enumerate(zip(datasets, N, dataset_labels)):\n",
    "        \n",
    "        data_odd_di = data_odd[data_odd['dataset']==dataset].copy()\n",
    "        pred_multiplicative_di = pred_multiplicative[pred_multiplicative['dataset']==dataset].copy()\n",
    "        pred_additive_di = pred_additive[pred_additive['dataset']==dataset].copy()\n",
    "        \n",
    "        if n == 2:\n",
    "            data_odd_di = data_odd_di.drop(['item_value_{}'.format(n),\n",
    "                                      'gaze_{}'.format(n)], axis=1)\n",
    "            pred_multiplicative_di = pred_multiplicative_di.drop(['item_value_{}'.format(n),\n",
    "                                      'gaze_{}'.format(n)], axis=1)\n",
    "            pred_additive_di = pred_additive_di.drop(['item_value_{}'.format(n),\n",
    "                                      'gaze_{}'.format(n)], axis=1)\n",
    "        \n",
    "        yield di, dataset, n, dataset_label, data_odd_di, pred_multiplicative_di, pred_additive_di"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Supplementary Figure 3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false,
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfkAAAH1CAYAAAD1dIefAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\npZRD/T8IiFVo9T5SyAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x12d419b00>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axs = plt.subplots(3, 4,\n",
    "                        figsize=glam.plots.cm2inch(18, 18),\n",
    "                        dpi=330, sharex='row')\n",
    "\n",
    "# plotting settings for predicted data\n",
    "predalphas = [1, 1]\n",
    "predlws = [2, 2]\n",
    "predls = ['-', '--']\n",
    "predmarkers = [None, None]\n",
    "predcolors = ['blue', 'red']\n",
    "\n",
    "# iterate datasets\n",
    "for dataset_bundle in dataset_iterator():\n",
    "    (di,\n",
    "     dataset,\n",
    "     n,\n",
    "     dataset_label,\n",
    "     data_odd,\n",
    "     pred_multiplicative,\n",
    "     pred_additive) = dataset_bundle \n",
    "    \n",
    "    # Dataset labeling\n",
    "    axs[0,di].set_title(dataset_label, fontsize=7)\n",
    "    \n",
    "    # Response time by difficulty\n",
    "    if di == 0:\n",
    "        _ = glam.plots.plot_rt_by_difficulty(data_odd,\n",
    "                                             [pred_multiplicative, pred_additive],\n",
    "                                             prediction_labels=['multiplicative', 'additive'],\n",
    "                                             xlims=(-1, 10),\n",
    "                                             ax=axs[0, di],\n",
    "                                             prediction_alphas=predalphas,\n",
    "                                             prediction_lws=predlws,\n",
    "                                             prediction_ls=predls,\n",
    "                                             prediction_markers=predmarkers,\n",
    "                                             prediction_colors=predcolors)\n",
    "    else:\n",
    "        _ = glam.plots.plot_rt_by_difficulty(data_odd,\n",
    "                                             [pred_multiplicative, pred_additive],\n",
    "                                             prediction_labels=None,\n",
    "                                             xlims=(-1, 10),\n",
    "                                             ax=axs[0, di],\n",
    "                                             prediction_alphas=predalphas,\n",
    "                                             prediction_lws=predlws,\n",
    "                                             prediction_ls=predls,\n",
    "                                             prediction_markers=predmarkers,\n",
    "                                             prediction_colors=predcolors)\n",
    "    axs[0,di].set_ylim(0, 8500)\n",
    "    axs[0,di].set_yticks([0, 2500, 5000, 7500])\n",
    "    axs[0,di].set_yticklabels([0, 2500, 5000, 7500])\n",
    "    axs[0,di].set_xticks([0, 1, 2, 3, 4, 5, 6, 7, 8, 9])\n",
    "    axs[0,di].set_xticklabels([0,'', 2,'', 4, '',6, '', 8, ''])\n",
    "        \n",
    "    # P(choose item) by difficulty\n",
    "    _ = glam.plots.plot_pchoose_by_value_minus_mean_others(data_odd,\n",
    "                                                           [pred_multiplicative, pred_additive],\n",
    "                                                           prediction_labels=None,\n",
    "                                                           xlims=(-7, 7),\n",
    "                                                           ax=axs[1, di], \n",
    "                                                           prediction_alphas=predalphas,\n",
    "                                                           prediction_lws=predlws,\n",
    "                                                           prediction_ls=predls,\n",
    "                                                           prediction_markers=predmarkers,\n",
    "                                                           prediction_colors=predcolors)\n",
    "    axs[1,di].set_yticks([0, .25, .5, .75, 1])\n",
    "    axs[1,di].set_yticklabels([0, '', .5, '', 1])\n",
    "    axs[1,di].set_xticks([-6, -5, -4, -3, -2, -1, 0, 1, 2, 3, 4, 5, 6])\n",
    "    axs[1,di].set_xticklabels([-6, '', '', -3, '', '', 0, '', '', 3, '', '', 6])\n",
    "    \n",
    "    # P(choose item) by gaze advantage\n",
    "    _ = glam.plots.plot_corp_by_gaze_advantage(data_odd,\n",
    "                                               [pred_multiplicative, pred_additive],\n",
    "                                               prediction_labels=None,\n",
    "                                               xlims=(-1.25, 1.25),\n",
    "                                               ax=axs[2, di], \n",
    "                                               prediction_alphas=predalphas,\n",
    "                                               prediction_lws=predlws,\n",
    "                                               prediction_ls=predls,\n",
    "                                               prediction_markers=predmarkers,\n",
    "                                               prediction_colors=predcolors)\n",
    "    axs[2,di].set_yticks([-1, -.5, 0, .5, 1])\n",
    "    axs[2,di].set_yticklabels([-1, '', 0, '', 1])\n",
    "    axs[2,di].set_xticks([-1, -.75, -.5, -.25, 0, .25, .5, .75,1])\n",
    "    axs[2,di].set_xticklabels([-1, '', -.5, '', 0, '', .5, '', 1])\n",
    "    \n",
    "# add panel labeling\n",
    "for label, ax in zip(list('abcdefghijkl'), axs.ravel()):\n",
    "    ax.text(-0.4, 1.1, label, transform=ax.transAxes,\n",
    "            fontsize=7, fontweight='bold', va='top')\n",
    "    ax.spines['top'].set_visible(False)\n",
    "    ax.spines['right'].set_visible(False)\n",
    "\n",
    "    # Font sizes\n",
    "    ax.tick_params(axis='both', labelsize=7)\n",
    "    \n",
    "fig.tight_layout()\n",
    "fig.savefig('results/figures/si_figure_3_additive-vs-multiplicative_group-average.pdf', dpi=330)\n",
    "fig.savefig('results/figures/si_figure_3_additive-vs-multiplicative_group-average.png', dpi=330)"
   ]
  },
  {
   "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
}