global_insights.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "00tdaK8JgU81",
    "outputId": "2060c8b3-c88f-4e5a-cb3a-75d5c3383402"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "hi 240 205\n",
      "ta 236 235\n",
      "te 247 245\n",
      "kn 231 225\n",
      "bn 232 221\n"
     ]
    }
   ],
   "source": [
    "import warnings\n",
    "warnings.simplefilter(action='ignore', category=FutureWarning)\n",
    "import pandas as pd\n",
    "pd.options.mode.chained_assignment = None\n",
    "import numpy as np\n",
    "from scipy.stats import ttest_ind\n",
    "import matplotlib\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "from collections import Counter\n",
    "import re\n",
    "from statistics import mean\n",
    "from itertools import combinations\n",
    "\n",
    "sns.set_style('darkgrid')\n",
    "\n",
    "rp_data = pd.read_csv('muse_rp.csv')\n",
    "lng_data = pd.read_excel('language_stats.xlsx', sheet_name=['hi', 'ta', 'te', 'kn', 'bn'])\n",
    "\n",
    "lang_count = {}\n",
    "\n",
    "for lang in lng_data.keys():\n",
    "    d = lng_data.get(lang)\n",
    "    ct_col = 'Transcript word count'\n",
    "    print(lang, len(d), len(d[d[ct_col]>=100]))\n",
    "#!rm -r plots; mkdir plots\n",
    "\n",
    "rp_data['male_screentime'] = rp_data['masculine_faces']/(rp_data['feminine_faces']+rp_data['masculine_faces'])\n",
    "rp_data['female_screentime'] = rp_data['feminine_faces']/(rp_data['feminine_faces']+rp_data['masculine_faces'])\n",
    "rp_data['total_faces']=rp_data['feminine_faces']+rp_data['masculine_faces']\n",
    "\n",
    "male_cols=[col for col in rp_data.columns if 'male' in col and 'female' not in col]\n",
    "male_data=rp_data[male_cols]\n",
    "male_data['masculine_faces']=rp_data['masculine_faces']\n",
    "male_data['Programme Language']=rp_data['Programme Language']\n",
    "\n",
    "female_cols=[col for col in rp_data.columns if 'female' in col]\n",
    "female_data=rp_data[female_cols]\n",
    "female_data['feminine_faces']=rp_data['feminine_faces']\n",
    "female_data['Programme Language']=rp_data['Programme Language']\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "DEiJMc5Mm3sW"
   },
   "source": [
    "# Global Insights\n",
    "\n",
    "###I. Screentime x Gender"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 1000
    },
    "id": "4CnndGl1jPwg",
    "outputId": "3a7c4fc2-c439-468f-9202-e1136f280273"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Average screentimes: 0.5579562217877896 0.44204377821220975\n",
      "Median screentimes: 0.5654213881019831 0.434578611898017\n",
      "N=1199\n",
      "p-value for screentime x gender: 1.4580085871807877e-75\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "The PostScript backend does not support transparency; partially transparent artists will be rendered opaque.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "BENGALI-> N=230, 230\n",
      "p-value for gender for BENGALI language: 3.40441539237696e-36\n",
      "HINDI-> N=268, 268\n",
      "p-value for gender for HINDI language: 6.141317664090087e-06\n",
      "KANNADA-> N=232, 232\n",
      "p-value for gender for KANNADA language: 8.148864868427629e-07\n",
      "TAMIL-> N=240, 240\n",
      "p-value for gender for TAMIL language: 1.4315890722998364e-15\n",
      "TELUGU-> N=229, 229\n",
      "p-value for gender for TELUGU language: 2.857276189609455e-41\n",
      "corrected p-value for languages BENGALI and HINDI: 1.3657964463240296e-05\n",
      "corrected p-value for languages BENGALI and KANNADA: 2.2221544502835794e-05\n",
      "corrected p-value for languages BENGALI and TAMIL: 2.2193486011042367e-05\n",
      "corrected p-value for languages BENGALI and TELUGU: 0.07753813300550419\n",
      "corrected p-value for languages HINDI and KANNADA: 0.08478300163907115\n",
      "corrected p-value for languages HINDI and TAMIL: 0.0432326460555346\n",
      "corrected p-value for languages HINDI and TELUGU: 2.0280917496020265e-05\n",
      "corrected p-value for languages KANNADA and TAMIL: 0.05567836530039878\n",
      "corrected p-value for languages KANNADA and TELUGU: 2.9999552605916185e-05\n",
      "corrected p-value for languages TAMIL and TELUGU: 2.788155450580983e-05\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjkAAAHUCAYAAADY2TmsAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8qNh9FAAAACXBIWXMAAA9hAAAPYQGoP6dpAABTf0lEQVR4nO3deViN+f8/8OcpFUqUJTtZjpClbKMU2XfGMGNXmAwZWxjDYGgsU2MZO2NpyFg/9j1Li73sYjRaVFQoaUPl3L8//Lq/zpwT55yOyj3Px3W5Lr3v7XXu07nPs/f9vu9bJgiCACIiIiKJMSjsAoiIiIg+BYYcIiIikiSGHCIiIpIkhhwiIiKSJIYcIiIikiSGHCIiIpIkhhwiIiKSJIYcIiIikiSGHCL6T3r79m1hl0BEnxhDDuWpXr164j8bGxskJCRovOzYsWOVli9o+/btU7vtuLg4sX3Pnj0FXhcVDffu3cPXX3+t0n7lyhXx9+PixYuFUNnnKysrC4cPH4anpye6desGOzs7NGzYEA4ODujXrx9+/fVX3Lp1q7DLzJe8jitUdBUr7ALo8yAIAk6cOAFXV9ePzpuWlobz589/+qKIdBAcHIwxY8awJ0ePDhw4AB8fHzx//lxlWlJSEpKSkhAWFobNmzfD2dkZCxcuRPny5QuhUvqvYcghjWkack6dOoWsrKxPX5AOjIyMUL16dQCAmZlZIVdDheHZs2d5BpzixYuLvx/FixcvyLI+W15eXvDz8wMAGBsb48svv0SHDh1Qq1YtlCpVCqmpqbh9+zb27duHCxcuICgoCIMGDcKOHTsYdOiTY8ihj5LL5QgPD8fNmzeRkJCAihUrfnD+Y8eOAQDKly+PZ8+eFUSJGrOysoK/v39hl0FFVJMmTfj7oYU//vhDDDgNGzbEypUrUaVKFaV5ypQpg+rVq6Nnz544cOAAZs6cidjYWCxYsADLly8vhKrpv4RjcuijnJycUKpUKfGU1YckJyfj8uXLMDIyQocOHQqoQiIqaLGxsVi5ciUAoEaNGvjzzz9VAs6/9e3bV+wNPn78OB49evSpy6T/OPbk0EcZGRmhY8eO2L9//0dPWR0/fhw5OTlo27YtypQp89F1P3z4EH/++ScuXbqEp0+fwtjYGDVr1kTnzp0xdOhQlCxZMs9lHz9+jC1btuDChQt48uQJzM3N4ezsDA8PjzyXiYuLE8PXL7/8ggEDBqjMExQUhCNHjuDmzZt4/vw5srKyULp0adjY2KB79+7o06cPihVT/ujs27cPP/74I6ysrBAUFITbt29jy5YtCAkJQUpKCsqWLQsHBweMHj0atWvX/uh+Uef48eM4cOAA7ty5g5cvX8LMzAy1atVC+/btMWjQoA+efrtw4QJ2796Nu3fvIjExEaampqhfvz769++Pnj17Ks27cuVKrFq1Cvb29li7di28vLxw9uxZyGQy1KxZE4sWLVIaeHnjxg1s374doaGhSEpKQokSJSCXy9GjRw/0798fRkZGedb14sUL/Pnnnzh37hxiY2Px9u1bVKpUCU5OThg5ciQqVaqksoyu+/rKlSsYPny40rpyX8eiRYvQr18/pXm2bNkCBwcHlXm3bt2KunXrYv369Thz5gwSExNhYWGBL774AuPHjxdPdwUGBsLX1xdhYWF48+YNrK2tMXDgQAwcODDP/fHkyRP4+voiODgY8fHxkMlkqFatGlxcXODq6goLC4s8l1Vn7NixOHv2LABg2bJl6N69u8o88+fPx/bt2wEAy5cvR7du3TRa9+bNm/HmzRsAwJw5c1CqVCmNlhs9ejS2bt0KALh+/Tpq1KihMk92djb27t2LY8eOITw8HBkZGShbtiyaN2+OYcOGoWnTpirLvP/ZPnXqFEqUKIENGzYgMDAQCQkJMDU1RZMmTTBkyBA4OzvnWZ8ux5X36fJ5aN++PR4/foxffvkF1apVw6+//oqHDx+iVKlSaN26NZYsWaLRtkkVQw5ppFu3bti/fz9u3ryJ+Ph4tV8+wP+dqurRowciIyM/uM4tW7bAx8dHaXzEmzdvcOfOHdy5cwc7d+7Ehg0bUKdOHZVlg4KCMHHiRGRmZoptr1+/xt69e3HixAn06tVL69f46tUrTJ48GefOnVOZ9vz5c5w/fx7nz5/H4cOHsXHjRpWgk2vPnj2YO3eu0utKSEjAvn37cOTIEaxfv17py1MT06dPx8GDB5XaUlJScP36dVy/fh3bt2/H1q1bxS/YXG/evMHcuXOxf/9+lWUvXbqES5cu4eTJk1i6dKnKwTcrKwujR4/GnTt3xLaYmBjxS0mhUMDb2xtbtmxRWS4kJAQhISHYvXs31q1bBysrK5XXdPnyZUyYMAEvX75Uao+KikJUVBR2794Nb29vdOnSJc/98in29cdERkbC09NT6VRsYmIiDh48iODgYOzdu1d83e+7f/8+5s6di9jYWEybNk1lvUePHsWPP/4oBodcDx48wIMHD7Bz506sXr0azZs317hWLy8vXL9+HSkpKVi4cCHatGkDc3NzcfqlS5fw119/AQC+/PJLjQPO27dvcfToUQBAtWrV0KZNG41rsrS0hK+vL2xsbNQG8/j4eLi7uyM8PFypPSEhAUeOHMGRI0fw7bffwtPTEzKZTO02wsLCMG/ePKSkpIhtWVlZCAgIQEBAACZMmKA2tOTnuJLfzwMA3Lp1C/PmzUN2djaAd4O2TU1N89wmaUAgyoNcLhfkcrmwdOlSISsrS2jZsqUgl8uFLVu2qJ0/ISFBqFevntCoUSMhLS1NWLp0qbiOf9u9e7c4bejQocL58+eFpKQk4cmTJ8Lu3bsFR0dHQS6XC87OzkJSUpLSspGRkULDhg0FuVwuuLi4CEePHhWSkpKEmJgYYc2aNYKtra247n9vOzY2VmzfvXu30rR58+aJ0+bPny+EhYUJSUlJwqNHj4Tjx48LPXr0EKfv2bNHadn//e9/glwuFxo2bCjUr19f6Natm3Dy5Enh+fPnQmxsrLBy5Uqhfv36Ys1v377V+H04fPiwuN0FCxYIf//9t5CcnCxER0cL69evFxo0aCDI5XJh5MiRKsvOnj1bXPb7778Xbty4ISQnJwthYWHC5MmTxWnLli0Tl1mxYoXS/lu5cqWQmJgo3L9/Xzh27Jg437Jly8R5xo8fL4SGhgovXrwQYmJihC1btgjNmjUT5HK50KdPH+H169dKdT148EBo3LixIJfLhfbt2wsHDhwQEhIShOfPnwsBAQHCV199JcjlcqFBgwZCaGioXvZ1Tk6OkJ6eLuzYsUOsOz09XUhPTxeys7MFQRCEy5cvi9MuXLigtN3c9kaNGgmNGzcWNm/eLMTFxQnR0dHCL7/8Ik7v0qWLIJfLBXd3d+HWrVvCixcvhJCQEKFXr16CXC4X6tevL8THxyut+8KFC4KNjY0gl8uF3r17C/7+/sLz58+FxMRE4ejRo0Lnzp0FuVwu2NnZCVFRURr/7giCIBw9elSsbfbs2WJ7amqq0LZtW/E9SEtL03idN2/eFNfp5eWlVT0fkpGRIXTt2lWQy+VC06ZNhfXr1wtRUVHCixcvhJs3bwpTpkwRt7t+/XqlZd//bDdp0kRo3ry5sG3bNiEuLk54+vSpcPDgQfEYVr9+fSE6Olpp+fwcVwQhf58HFxcXcVknJychKChIePbsmXD69GkhPDxcb/v3v4ghh/L0fsgRBEGYNWuWIJfLha+//lrt/Js3bxa/TAVByDPkpKamCnZ2doJcLhfGjRun9gv/8ePH4kHh3wfR7777TpDL5ULLli2FJ0+eqCzr7++vdchJTU0VD3CzZs1S+/oSExPFL+YJEyYoTcv94v3QF8bixYvFee7cuaN2G+p8//33glwuF0aMGKF2+qpVqwS5XC7Uq1dPSE5OFtvv3r0r1KtXT5DL5cLcuXPVLjtu3DjxCyUjI0MQBOWQM3XqVLXLRUdHi1/KeX3J3blzRwwb/w7GQ4YMEffV+zXnevPmjTBgwABBLpcLvXr1UpqW3339/vL/pknIkcvlgr+/v8qy74dgNzc3QaFQKE1/+PChOH3//v1ie05OjtC+fXtBLpcL/fv3V/kCFARBSElJEef57rvvVKZ/zMSJE8XfkWvXrgmCIAgzZswQv/Bz2zT1/h8p//5jIT9+//13McDevHlT7TwLFiwQ5HK5YGtrKzx9+lRsf/+z3aBBAyEsLExl2bNnz4rzbNq0SWlafo4r+f08vB9yLl68qHZ50g0HHpPGcruyb926hSdPnqhMz+2+/vcYj387dOgQMjIyAAAzZsyAgYHqr2HlypUxdOhQAO/GYOTk5AB4dw+eoKAgAMDw4cPVnjbr2LGjVt3nuet1dXVFt27d4ObmpnaeChUqwNraGsC7sSR5GTx4sNpu+LZt24r/j4uL07i23MvxX758Ke6H9w0aNAgbNmzA0aNHlcZFHDlyBIIgoGTJkpg6daradbu7u6N27dpo3bq12nuc5HWqaNeuXVAoFChRogQmT56sdh5bW1v06NFDnD/Xw4cPERISAgAYN26c2nEmxsbG4nofPHiQ503k9L2vNWFtbY2OHTuqtNvb24v/HzVqlMqplNq1a4vvz9OnT8X28+fPizV6enrCxMREZd2lS5fGd999BwA4d+6c0vKamDt3LsqVKwdBEDBv3jycOXMG+/btAwCMGTNGqXZNvP/7/6HLwLOzs5GRkZHnv1evXonzCoIg/p706NEDTZo0UbvOCRMmoHjx4sjKylI5DZurTZs2aNCggUq7o6OjeJr5/d+L/B5X8vN5eF/p0qXRunVrtdNINww5pLEvvvgClpaWEAQBJ0+eVJoWExODO3fuwNTUVOkLRp2rV68CACwsLGBpaZnnAbBRo0YAgIyMDPz9998AgJCQEPGL/kODBzt16qTVa6tcuTKmTp2K5cuXqx0YnJWVhdDQUPGgrC5s5Mrr4Fy2bFnx/69fv9a4ttwxGLl36f3rr7+UDtCWlpZo27YtateurTRO6PLly+LyeQ1KbtKkCY4dO4Y1a9aojOcBgPr166tdLvc9rFWrFgDk+R42btwYwLtxLLlfjFeuXBHXI5fL81y2Xr16MDQ0BABcu3Ytz/rV0XVfayL3NX1om+q+YIH/uzfT++NuNN0ftra2AN6FgRs3bmhVs4WFBebNmwcA+PvvvzFx4kTxtWg6oPZ9CoVCo/n++OMP2Nvb5/kv90sfACIiIsSgXb9+/Tz3g0wmEweBX79+Xe1283qPjI2NxaD5fsDK73ElP5+H9/FOyvrHgcekMUNDQ3Tu3Bk7d+7E8ePHlXo8cntxOnXqpPYv0ffFxsYCePfXoKZ/QSYkJMDW1hbx8fFim7ov5Vy5BxtdxMTE4MqVK4iMjMSjR4/w6NEjREdHKwUbQRDyXD6vK2CMjY3F/2v6JQEAQ4YMwYkTJ3Dnzh2EhYUhLCwMwLsehTZt2qBDhw5o1aqVSo9YYmIiAKi9ekVTeV0hl/sehoWFafUeWlhYiMsCQP/+/TVa9v33/X363teayGub7+//vEKlul7L9/eHpn/F57U/PqRjx47o06cPDh48iOzsbJQsWRI+Pj55DqD/kNKlS4v/19e9sGJiYsT/L1q0CIsWLfroMtr+XgD/97vx/mc4v8eV/Hwe3qft1XP0cezJIa3kXoL671NW719V9THp6elabzd3mbS0NLGtRIkSec6v6eWs74uPj8eECRPQpUsX/PTTT9i8eTPOnDmDhw8fwszMDF26dEG1atU+uh5dvjQ+pESJEvjrr78wbdo0pYNsVFQUtm3bBldXV3Ts2BGBgYFKy+VetfSh/fQxeQXW/LyH+Vn23/S9rzWhyZ2Q87rqRx197o+Padmypfj/4sWLK11ppY33ezs/dDpw3Lhx4tVh7/9Td5WSPvfDh25boE5+jyv6qv39cE76wZ4c0kqLFi3EOxmfOHECI0eOxMOHDxEeHg4LCwuNLtfN/ZJo2rRpnuem8/L+X5CZmZl5HhS0faxEeno6hg4diri4OMhkMjg7O6NFixaQy+WoXbu2GG4GDRqk9Jd3QTE2Nsbo0aMxevRoREdH48KFC7h48SIuXbqEjIwMPH78GB4eHvjrr7/ELvHixYsjPT1dqVteX3LX3aNHDyxdulTrZXPdvn37oz1/Upe7P8qXL/9Jn/n29OlTeHt7iz8nJyfj559/xooVK7ReV9OmTVGyZElkZmbi9OnTeY5D0cb74WLjxo1wcnLK9zo1ld/jSn4+D/RpsSeHtGJgYICuXbsCgHj349xTVV26dNHoL+vKlSsD+PiAUHWnhHKXBfDB+/BoG0S2b98u1rNs2TL88ccfcHd3R7t27ZR6bz404Lig1KxZE0OGDMHq1atx+fJlzJw5EzKZDNnZ2dixY4c4X+6++ti+8PHxga+vr8p9ST5EX++hLstLTe7+SE5OVro/i77Nnj0bL1++hKmpqTiI+eTJk+LnVxvGxsbiceDhw4cIDQ3Nd33vD/Yt6N+L/B5X8vN5oE+LIYe0lnuV1e3btxEfH4/jx48D+PhVVbmaNWsG4N0N9vK6agYA1q9fj+bNm6N3797i+foWLVqIf/mePn06z2Vzr5TQVO5ATgsLizxviJaQkIDo6GgABXewevXqFUaNGoW2bduKd6V9n7GxMUaMGAG5XA7g/8bhAICdnR2Ad4Mq8+rNiYiIwMaNG7Fo0SLcv39f47py38OwsDAkJCTkOd+cOXPQqlUrfPXVV2L3/Ps3sztz5kyey16/fh1NmjRBly5dxN8xfdDmVFJByN2Xb9++RUBAQJ7zHT58GHZ2dujRo4fWoWLv3r3iuj09PTFp0iTx92P+/Pk6jasZM2aMeFrohx9+0OqUjbqB+zY2NuKN7z70e5GRkQFHR0e4uLjgt99+07Jq9fJ7XMnP54E+LYYc0pq9vT0qVaoEQRCwdu1aREVFwcrKSuM7sfbt21fsDvby8lL7BRwTE4MtW7YgLS0NWVlZYm9KyZIlxRCyfft2PHjwQGXZmzdvav3Xae6A0JcvX6o94L958wazZs0Sw01BPWW9RIkSSExMREJCAnbt2qVyN1zgXc2546PeHzT51VdfAXj3pZDXKYlVq1YBeLdf27dvr3FdX3/9NYB3X1bz5s1T+1TvW7duYf/+/UhJSUGZMmXEwbiNGzeGjY0NgHdX3+QGx/e9fv0av/76K968eYPHjx/nebWMLnKv2AIK7n38kA4dOqBcuXIAgN9++w3Jyckq8yQnJ2PFihXIzMzE8+fP87zqTZ34+HhxEK+9vT0GDx4MmUyG+fPnw8jICCkpKZg9e7bWddesWROenp4A3vVgDBo06KO9gS9fvsSiRYvUPgOvWLFi6NevHwAgODg4z2C7fPlyJCUl4cmTJ+LvUX7l97iSn88DfVoMOaQ1mUwmdlXv3r0bwLsByZr+hVy+fHlMmDABAHDnzh18/fXXOHXqFJ4/f44nT57gwIEDGDZsGFJSUiCTyTBr1iyldU+bNg1lypTB69evMXz4cOzatQtPnz5FYmIitm/fjlGjRmnd0+Lo6Ajg3ZU4Y8aMwaVLl5CUlITY2FgcPHgQ/fv3Vxov8SlPK/zbqFGjALy7X4ybmxuCg4ORmJiIxMREBAYGYuTIkUhLS4OhoSEGDRokLtekSRP07dsXwLvnDP3444+4d+8eXrx4gdu3b2PKlCnigPGxY8dqNVi7QYMG4rbOnj2L4cOH4/z580hOTkZMTAz8/Pzw7bffIjs7GyYmJiqPMZg7dy6KFSuG1NRUfPPNN/Dz80NcXBySkpJw/vx5uLq64ubNm+Lr/9iDH7Xx/hVjuV86ufdtKgwmJiaYNWsWgHfPTerfvz8OHDggvsf+/v4YNmyY2Jvp6emp1a3+Z82ahfT0dBgZGeGXX34RP0tyuVz83Tp37lye95z5EDc3N4wbNw4AEB4ejq+++goTJ07EyZMn8fDhQ6SkpCAmJgbnzp3DnDlz4OLiAl9fXwiCAFNTU3H7uTw8PMTTVp6envj1118RHh6OFy9e4O7du/jhhx/E5141a9ZM7bO4dJWf40p+Pw/06XDgMemke/fu2LJli/ih1+SqqveNHj0amZmZWLt2LcLDw/H999+rzGNkZIS5c+eqDEAsW7YstmzZAnd3dzx79gxz5sxRmm5iYoJJkyZp9VC7AQMG4Pjx4wgJCUFYWJjah5BaW1ujUaNGOHToEOLj45GVlVUgV0N8+eWXuHXrFnbs2IFr165h9OjRKvMYGRnBy8tL5T4b8+bNQ0ZGBvz9/bFv3z7xBnDvGzJkCL799lut65o1a5b4IMXQ0FCVLywAMDU1xdKlS1X+4ra3t8eKFSswdepUpKSkwMvLC15eXirLDxgwQAzE+mJraysOmp0zZw7mzJmD8ePHq/0dLCjdu3dHamoqfvnlFzx+/Bg//PCDyjwymQweHh5ir4Em/vrrL1y4cAEA8N1336ncA2rcuHE4ceIEoqOjsWDBArRu3RoVK1bUqvaJEyfCzs4OixYtQmRkJE6cOKG2pyZXiRIl0Lt3b3h4eKg8w8nCwgKbN2/G2LFjER0djc2bN2Pz5s0q62jcuDFWrVql9pJ8XeX3uJKfzwN9Ogw5pJPGjRujatWqiIuLQ/Xq1cUb92lKJpNh4sSJ6NKlC/z8/HD16lUkJiZCoVCgcuXK+OKLLzB8+PA8n9jdoEEDHDp0CH5+fvD390dsbCxKlCiBFi1awMPDQ+srioyNjbF582b8+eefOHbsGKKiopCdnQ1zc3PUqVMHXbt2Rb9+/XD//n0cOnQIr1+/RlBQkNo7334KP//8M1xcXLB3717cvn0bSUlJMDIygpWVFRwcHDB8+HDUrFlTZbnixYtj1apVOH36tLjsy5cvUapUKTRt2hRDhgzR+SoWIyMjLFiwAH379sXOnTtx48YNPHv2DAYGBqhWrRqcnJwwYsSIPL80O3ToAH9/f2zbtg2BgYGIjY3FmzdvYGFhATs7O3zzzTdiD5s+WVpaYt26dViyZAkePHiAYsWKqTwktDAMHDgQjo6O+PPPP3Hp0iU8efIE2dnZqFChApo3b46hQ4dqddouNjYWPj4+AIC6detizJgxKvOYmJhg3rx5GDFiBNLS0jBr1ixs2rRJ69qdnZ3h5OSE8+fP4/Tp07h9+zZiY2ORmZkJMzMzWFpaomHDhvjiiy/QuXNnpauZ/q1WrVo4dOgQ9uzZg5MnTyI8PBzp6ekwMzNDvXr10KtXL/Tr10/ptKO+5Oe4kt/PA30aMoHDvYmIiEiCOCaHiIiIJKnIhpybN29i2LBhaNq0KRwcHPDDDz8gKSlJnB4ZGQl3d3c0a9YMrVq1wsyZM5GamlqIFRMREVFRUiRPV929exeDBw9G69atMWzYMDx9+hRLly5F1apVsXPnTqSmpqJnz56oUKECxo4di6SkJPj4+KBRo0ZqB6kRERHRf0+RHHjs7e2N+vXrY82aNeLgMjMzMyxYsACxsbE4duwYUlNTceDAAVhaWgIArKys4O7ujtDQUI3v10JERETSVeROV7148QJXr17FoEGDlEbPd+7cGYGBgahWrRrOnz+PZs2aiQEHAJycnGBqaqr1nW6JiIhImopcyHnw4AEEQUDZsmXh6ekJOzs72NnZYerUqeJlnhEREbC2tlZazsDAAFWrVlV791QiIiL67ylyISf3luYzZ85E8eLFsWbNGkyfPh2BgYFwd3eHQqFAamqq2jt+mpqa8nkgREREBKAIjsnJzs4GADRs2BALFiwAALRu3Rrm5uaYMmWKePdOdY8QEAShyD18j4iIiApHkQs5uT00Li4uSu25d2W9f/8+zMzM1PbYZGZm6nQ3yeTkNBS9a8wKlkwGWFqW4r4gIiqCeIxWlrs/PqbIhZzcW9P/++nAOTk5AN7dpt7a2lp8WF0uhUKBuLg4dO7cWettKhT4z//S5HaAcV8QERU9PEYr0/SkTZEbk1O7dm1UqVJF5ZH2Z86cAQA0b94cjo6OCAkJEcfvAEBwcDAyMjI+ybNuiIiI6PNT5EKOTCbD9OnTcfPmTUyaNAkXLlzAtm3bsHDhQnTp0gUNGjTA4MGDYWJiAjc3N/j7+2PPnj2YNm0anJ2dYWdnV9gvgYiIiIqAInnHYwA4d+4cVq9ejQcPHqB06dLo1asXJk+eDGNjYwBAeHg4Fi5ciBs3bsDU1BQdO3bE9OnTYWZmpvW2nj/nOU6ZDChXrhT3BRFREcRjtLLc/fHR+YpqyClI/KXhB4iIqCjjMVqZpiGnyJ2uIiIiItIHhhwiIiKSJIYcIiIikiSGHCIiIpIkhhwiIiKSJIYcIiIikiSGHCIiIpIkhhwiIiKSJIYcIiIikiSGHCIiIpIkhhwiIiKSJIYcIiIikiSGHCIiIpKkYoVdgBR16tQWT58mFnYZWjMwkEGh+Dwfb1uhghX8/QMLuwwiIipCGHI+gadPExEf/6SwyyAiIvpPY8j5hAwMDFCxgkVhlyFpCU9fQKFQFHYZRERUBDHkfEIVK1jgn4u+hV2GpNV1cMWThKTCLoOIiIogDjwmIiIiSWLIISIiIkliyCEiIiJJYsghIiIiSWLIISIiIkliyCEiIiJJYsghIiIiSWLIISIiIkliyCEiIiJJYsghIiIiSWLIISIiIkliyCEiIiJJYsghIiIiSWLIISIiIkliyCEiIiJJYsghIiIiSWLIISIiIkliyCEiIiJJYsghIiIiSWLIISIiIkliyCEiIiJJYsghIiIiSWLIISIiIkliyCEiIiJJYsghIiIiSWLIISIiIkliyCEiIiJJYsghIiIiSWLIISIiIkliyCEiIiJJYsghIiIiSWLIISIiIkkqVtgFqPPq1SvY29tDoVAotRsbG+POnTsAgMjISCxevBjXrl1DsWLF0KFDB8yYMQPm5uaFUTIREREVMUUy5Dx48AAKhQJLly5FlSpVxHYDg3cdT6mpqXB1dUWFChXg7e2NpKQk+Pj4ICEhAZs3by6ssomIiKgIKZIh5/79+zAyMkLnzp1hZGSkMn3Hjh1ITU3FgQMHYGlpCQCwsrKCu7s7QkND0bx584IumYiIiIqYIjkm5/79+6hTp47agAMA58+fR7NmzcSAAwBOTk4wNTVFUFBQQZVJRERERViRDDl///03DAwM4ObmhqZNm6Jly5aYM2cO0tPTAQARERGwtrZWWsbAwABVq1ZFdHR0IVRMRERERU2RO12lUCgQHh4OAwMDTJ06FePGjcOdO3ewatUqPHz4EH5+fkhNTYWpqanKsqampmIQ0oZMpo/KqbDxfSQiqco9vvE4946m+6HIhRxBELB+/XqUK1cOtWvXBgC0aNEC5cqVw7Rp0xAcHAwAkKl5hYIgqG3/mLJlS+Wv6H8xMOBvYUEzMJChXDn9vo9EREWNvr+vpK7IhRxDQ0O0atVKpb1du3YA3l15ZWZmprbHJjMzExUrVtR6m0lJaRAErRfLk0Khx5WRRhQKAc+fpxV2GUREn4RM9i7g6Pv76nOVuz8+psiFnMTERAQGBsLZ2VkpsLx+/RoAYGFhAWtra8TExCgtp1AoEBcXh86dO2u9TUEAf2kkgO8hEUkdv6+0U+QGHmdlZWH27NnYtWuXUvuxY8dgYGCAZs2awdHRESEhIUhOThanBwcHIyMjA46OjgVdMhERERVBRa4np1q1aujTpw/++OMPGBsbo2nTprh27RrWrVuHwYMHo1atWhg8eDD8/Pzg5uaG8ePHIyUlBT4+PnB2doadnV1hvwQiIiIqAopcyAEALy8v1KhRAwcOHMCaNWtgZWWFCRMmYNSoUQAAS0tLbN26FQsXLsTUqVNhamqKrl27Yvr06YVcORERERUVRTLkmJiYwMPDAx4eHnnOI5fL4evrW3BFERER0WelyI3JISIiItIHhhwiIiKSJIYcIiIikiSGHCIiIpIkhhwiIiKSJIYcIiIikiSGHCIiIpIkhhwiIiKSJIYcIiIikiSGHCIiIpIkhhwiIiKSJIYcIiIikiSGHCIiIpIkhhwiIiKSJIYcIiIikiSGHCIiIpIkhhwiIiKSJIYcIiIikiSGHCIiIpIkhhwiIiKSJIYcIiIikiSGHCIiIpIkhhwiIiKSJIYcIiIikiSGHCIiIpIkhhwiIiKSJIYcIiIikiSGHCIiIpIkhhwiIiKSJIYcIiIikiSGHCIiIpIkhhwiIiKSpGKFXQAREVFB6tSpLZ4+TSzsMrRmYCCDQiEUdhlaq1DBCv7+gYWybYYcIiL6T3n6NBHx8U8KuwwqAAw5RET0n2RgYICKFSwKuwzJSnj6AgqFolBrYMghIqL/pIoVLPDPRd/CLkOy6jq44klCUqHWwIHHREREJEkMOURERCRJDDlEREQkSQw5REREJEkMOURERCRJDDlEREQkSQw5REREJEkMOURERCRJDDlEREQkSQw5REREJEkMOURERCRJDDlEREQkSfl6QGdOTg5OnTqFq1evIiEhAfb29nB3d8fevXtha2sLGxsbfdVJREREpBWdQ869e/cwadIkxMbGQhAEyGQymJubAwB27NiBe/fuYebMmRg2bFi+ixw/fjzu3buHs2fPim2RkZFYvHgxrl27hmLFiqFDhw6YMWOGWAMRERH9t+l0uurJkydwc3NDTEwM2rZtizlz5kAQBHG6g4MDihUrhoULFyI0NDRfBR48eBD+/v5KbampqXB1dUVycjK8vb3h6ekJf39/TJo0KV/bIiIiIunQqSdnzZo1ePnyJby8vDBgwAAAwPz588Xpnp6esLe3x9ixY7Fp0yY0b95cp+ISExOxYMECVKxYUal9x44dSE1NxYEDB2BpaQkAsLKygru7O0JDQ3XeHhEREUmHTj05wcHBqFevnhhw1HFxcYGtrS3+/vtvnYv76aef4OjoiNatWyu1nz9/Hs2aNRMDDgA4OTnB1NQUQUFBOm+PiIiIpEOnnpykpCQ0adLko/NVqlQJDx480GUT2LNnD8LCwnDkyBF4e3srTYuIiED37t2V2gwMDFC1alVER0drvS2ZTKcSqYjh+0hEVDTp+/is6fp0CjmWlpYahYnIyEil3hZNPX78GIsWLcKiRYvULp+amgpTU1OVdlNTU6Snp2u9vbJlS2m9zIcYGPDbtqAZGMhQrpx+30cikiYeowtWYR6fdQo5rVu3xqFDh3D8+HF069ZN7TzHjh1DREQE+vTpo9W6BUHAzJkz0bZtW3Tp0iXP+WRqYlzuVV7aSkpKw3vjpvNNodDjykgjCoWA58/TCrsMIvoM8BhdsD7F8Vkm06yDQqeQM3bsWJw6dQrTpk3DtWvX8MUXXwAA0tLScOnSJQQFBWHbtm0wNjbG6NGjtVr39u3b8eDBAxw+fBg5OTkAIF65lZOTAwMDA5iZmantscnMzFQZpKwJQYBeQw4VDr6HRERFU2Edn3UKOTVr1sTKlSsxZcoU+Pn5Yfv27ZDJZAgICEBAQAAEQUCJEiXg7e2NunXrarXukydP4sWLF2jTpo3KtIYNG2L8+PGwtrZGTEyM0jSFQoG4uDh07txZl5dEREREEqPzzQDbtGmDkydPYs+ePbhy5QoSEhLw9u1blC9fHi1btsTXX38NKysrrdc7b948ZGRkKLWtXr0ad+/exdq1a1GhQgXIZDJs2rQJycnJ4pid4OBgZGRkwNHRUdeXRERERBKSr8c6WFhYwN3dHe7u7vqqB7Vq1VJpK1OmDIyNjdGoUSMAwODBg+Hn5wc3NzeMHz8eKSkp8PHxgbOzM+zs7PRWCxEREX2+PssHdFpaWmLr1q2wsLDA1KlTsWzZMnTt2hXLli0r7NKIiIioiNC5J+fu3bvYu3cvYmJi8ObNmzznk8lk8PPz03UzAIDFixertMnlcvj6+uZrvURERCRdOoWcK1euYNSoUXj79q3SM6vU0eWSbiIiIqL80inkLFu2DDk5OejZsye6du2KUqVKMcwQERFRkaJTyPnnn3/QoEED/Pbbb/quh4iIiEgvdBp4XKJECZ0uDyciIiIqKDqFnA4dOuDGjRt4+fKlvushIiIi0gudQo6npyfKli0LNzc3XLhwAUlJSVAoFHn+IyIiIipoOo3JMTc3R7du3bBq1aqPPptKJpPh3r17OhVHREREpCudQo6vry9Wr14NAB+9hPxj04mIiIg+BZ1Czo4dOyCTyfDDDz+gd+/e4vOjiIiIiIoKnUJOfHw8WrVqBVdXVz2XQ0RERKQfOg08LleuHIyNjfVdCxEREZHe6BRyevbsicuXLyM2Nlbf9RARERHphU6nqzw8PBAaGorBgwdj1KhRaNKkCUqXLg0jIyO181erVi1fRRIRERFpS6eQ065dO7x9+xapqan49ddfPzgvLyEnIiKiwqBTyClRogQAwMzMTK/FEBEREemLTiHn7Nmz+q6DiIiISK90GnhMREREVNRp1JOTexVV5cqVYWhoqPVVVRx4TERERAVNo5DTqVMnGBgY4OjRo7C2tkanTp0gk8k02gAHHhMREVFh0CjkVK5c+d3MxYop/UxERERUVGkUcv490JgDj4mIiKio02ng8ZMnT5CSkvLR+eLi4hAcHKzLJoiIiIjyRaeQ06FDByxatOij83l7e8PT01OXTRARERHli1ZXV+USBAEZGRkfvMoqLS0N9+7dQ3Z2dv4qJCIiItKBRiHHy8tL6bSTTCbDmTNncObMmQ8uJwgCWrZsmb8KiYiIiHSgUcj58ccfERERAUEQAADx8fEoXrw4LCws1M4vk8lgYmKCGjVqYObMmfqrloiIiEhDGoUca2trpV4bGxsbdOrUCd7e3p+sMCIiIqL80OnZVVu3bkXZsmX1XQsRERGR3ugUcjjOhoiIiIo6nUIOANy+fRubNm3CP//8g9evX0OhUKidTyaT4dy5czoXSERERKQLnULO9evXMWLECOTk5IiDkfOi6TOuiIiIiPRJp5CzevVqZGdno2vXrhg6dCgqVKgAQ0NDfddGREREpDOdQs6tW7dQs2ZNLF++XM/lEBEREemHTo91UCgUkMvl+q6FiIiISG90Cjn169dHeHi4vmshIiIi0hudQo67uzuio6OxefNmfddDREREpBc6jcnJyclBp06d4OPjgyNHjqBp06YwNzdXeyWVTCbDhAkT8l0oERERkTZ0CjkeHh6QyWQQBAH37t3DvXv3VObJnc6QQ0RERIUhXyGHiIiIqKjSKeR8//33+q6DiIiISK90Gnj8b4Ig4MWLF0hJSdHH6oiIiIjyLV8h59KlS/j2229hb28PBwcHLFq0CAAwceJE/Prrr3j9+rVeiiQiIiLSls4P6Fy1ahVWr14NQRBgYGAAQRDE51g9ePAAp06dwu3bt7FlyxYYGxvrrWAiIiIiTejUk3Pu3DmsWrUKlSpVwooVKxASEqI0fcmSJZDL5bh+/Tp27dqll0KJiIiItKFTyPH19YWJiQl8fX3RuXNnmJqaKk1v2LAhNm/ejOLFi+PgwYN6KZSIiIhIGzqFnLCwMDRv3hzVq1fPc56yZcuiRYsWiI2N1bk4IiIiIl3pFHJycnJgZGT00fkEQUBWVpYumyAiIiLKF51CTo0aNXDnzh28evUqz3kyMjJw586dD/b2EBEREX0qOoWcHj16ICkpCT/99BPevHmjMv3NmzeYM2cOXr58iW7dumm9/rdv32LDhg3o1KkTGjdujN69e6uM7YmMjIS7uzuaNWuGVq1aYebMmUhNTdXl5RAREZEE6XQJ+YgRI3DixAkcPXoUV65cQZMmTQAA9+/fx/Tp0xESEoL4+HjUrVsXI0aM0Hr9S5cuxZ9//okJEyagUaNGCAwMxPTp02FgYIBevXohNTUVrq6uqFChAry9vZGUlAQfHx8kJCTwyehEREQEQMeQY2Jigq1bt2L+/Pk4evQozpw5AwD4559/8M8//wAAOnToAC8vL5QoUUKrdWdkZMDPzw8jRoyAu7s7AKB169YICwuDn58fevXqhR07diA1NRUHDhyApaUlAMDKygru7u4IDQ1F8+bNdXlZREREJCE63wzQzMwM3t7emDp1qthzo1AoUL58eTRv3hzVqlXTab0mJibYtWsXypUrp9RuZGSE9PR0AMD58+fRrFkzMeAAgJOTE0xNTREUFMSQQ0RERLqHnFwVKlRAjx499FELAKBYsWKwsbEB8O7qrOfPn2Pfvn24ePEivLy8AAARERHo3r270nIGBgaoWrUqoqOj9VYLERERfb7yFXJycnJw6tQpXL16FQkJCbC3t4e7uzv27t0LW1tbMazo6vDhw5g2bRoAoG3btmKwSU1NVbkBIQCYmpqKvT3akMnyVSYVEXwfiYiKJn0fnzVdn84h5969e5g0aRJiY2MhCAJkMhnMzc0BADt27MC9e/cwc+ZMDBs2TNdNoEmTJvDz80NUVBRWrFiBgQMHYu/evQAAmZpXmFuHtsqWLaVzjeoYGPDbtqAZGMhQrpx+30cikiYeowtWYR6fdQo5T548gZubG16+fIl27drB2dkZ8+fPF6c7ODggPDwcCxcuRP369XUeI1OjRg3UqFEDLVq0QLVq1eDq6oqTJ0/CzMxMbY9NZmYmKlasqPV2kpLS8P+fLaoXCoUeV0YaUSgEPH+eVthlENFngMfogvUpjs8ymWYdFDqFnDVr1uDly5fw8vLCgAEDAEAp5Hh6esLe3h5jx47Fpk2btAo5SUlJCAoKgrOzM8qWLSu2N2rUCACQkJAAa2trxMTEKC2nUCgQFxeHzp07a/16BAF6DTlUOPgeEhEVTYV1fNbpZoDBwcGoV6+eGHDUcXFxga2tLf7++2+t1p2ZmYkZM2Zgz549KtsEgHr16sHR0REhISFITk5Wmp6RkQFHR0ettkdERETSpFNPTlJSkngDwA+pVKkSHjx4oNW6q1Wrhr59+2L16tUwMDBAo0aNcPfuXaxduxZt2rSBs7MzGjVqBD8/P7i5uWH8+PFISUmBj48PnJ2dYWdnp8tLIiIiIonRKeRYWlpqdKl2ZGSk0r1sNOXl5YWaNWvif//7H1auXIny5ctj+PDhGDduHGQyGSwtLbF161YsXLgQU6dOhampKbp27Yrp06fr8GqIiIhIinQKOa1bt8ahQ4dw/PjxPJ9NdezYMURERKBPnz5ar9/Y2Bhjx47F2LFj85xHLpfD19dX63UTERHRf4NOIWfs2LE4deoUpk2bhmvXruGLL74AAKSlpeHSpUsICgrCtm3bYGxsjNGjR+u1YCIiIiJN6BRyatasiZUrV2LKlCnw8/PD9u3bIZPJEBAQgICAAAiCgBIlSsDb2xt169bVd81EREREH6XzzQDbtGmDkydPYs+ePbhy5QoSEhLw9u1blC9fHi1btsTXX38NKysrfdZKREREpDGdQo6/vz8aNmyIypUrw93dXXxaOBEREVFRoVPImT9/PoyMjHD27Fl910NERESkFzrdDPDly5ewtbXVdy1EREREeqNTyGnatCmuX7+O1NRUfddDREREpBc6na6aPXs2vv/+e/Tt2xd9+vSBXC5H6dKl83wCeOvWrfNVJBEREZG2dAo5vXr1gkwmgyAIWLdu3QfnlclkuHfvnk7FEREREelKp5DTokULfddBREREpFc6hZxt27bpuw4iIiIivdJp4HGut2/fIiIiQqktLi4OgYGByMrKyldhRERERPmhc8gJCgqCk5MTZs2apdQeGhqKMWPGoGPHjrh27Vq+CyQiIiLShU4h5/bt2xg7dixSUlJQpUoVpWl16tRB165d8fz5c7i6unLQMRERERUKnULO2rVroVAosHr1aixZskRpmq2tLZYvX47Vq1cjOzsba9as0UuhRERERNrQKeSEhYWhWbNmcHFxyXMeFxcXNG3aFCEhIToXR0RERKQrnUJOWloaLC0tPzqflZUVXr16pcsmiIiIiPJFp5BTtWpV3LhxA9nZ2XnO8/btW9y5cweVKlXSuTgiIiIiXekUcrp06YJnz55h1qxZai8Vz8nJgZeXF+Lj49G5c+d8F0lERESkLZ1uBujq6orDhw/j8OHDCA4OhoODg9hjk5CQgMuXLyMpKQnVq1eHu7u7XgsmIiIi0oROIcfMzAxbt27Fzz//jHPnzuHo0aMq8zg7O8PLywulSpXKd5FERERE2tIp5ADvBhWvXbsWiYmJuHr1Kp49e4acnBxUqFAB9vb2qF69uj7rJCIiItKKziEnl5WVFXr16qWPWoiIiIj0Jt8hJ1dKSgr8/f2RkpICuVyOtm3b6mvVRERERFrTKuRs374dO3fuhKOjI2bMmCG2BwQEYMqUKUr3xKlbty7Wrl2r8tgHIiIiooKgcciZOXMm9u/fD0EQIJfLxfYnT55g0qRJeP36NapVq4ZOnTohMjISAQEBGDlyJA4fPgxjY+NPUjwRERFRXjQKORcuXMC+fftQrlw5zJ07F23atBGnrV27Fq9fv0blypXxv//9D+bm5gCANWvWYMWKFdi5cyeGDx/+aaonIiIiyoNGNwPct28fZDIZ1q9fj06dOqFEiRIAAEEQcOrUKchkMri5uYkBBwC+/fZbmJub4+TJk5+mciIiIqIP0Cjk3LhxAzY2NmjYsKFSe1hYGF6+fAkAaNeundI0IyMjNG7cGJGRkfqplIiIiEgLGoWc5ORkVK1aVaU9NDQUAFCpUiVUq1ZNZXqpUqWQnp6ezxKJiIiItKdRyDE0NIRMJlNpv3r1KmQyGVq0aKF2uaSkJKVTWEREREQFRaOQU7lyZURERCi1ZWVl4fLlywAABwcHlWVevXqF27dvo3Llynook4iIiEg7GoUcBwcHREZG4ty5c2Lbvn37kJmZCSMjI7i4uKgss3HjRrx580ZtACIiIiL61DS6hHzEiBHYuXMnJkyYgO7du8PAwACHDx+GTCbDoEGDlE5JJScnY+vWrVi/fj2KFy+OgQMHfrLiiYiIiPKiUcipXLkyfv/9d0yZMgUHDx4U2x0cHDBlyhTx5ytXrsDV1RUAIJPJMHfuXFSqVEm/FRMRERFpQOM7Hrdr1w4BAQE4c+YMkpKSYGNjo3RTQAAoXbo0jIyM0LRpU3h4eKBVq1Z6L5iIiIhIE1o9u8rc3BxffvllntNtbGxw+/btfBdFRERElF8aDTwmIiIi+tww5BAREZEkMeQQERGRJDHkEBERkSQx5BAREZEkMeQQERGRJDHkEBERkSRpdZ+cj4mKikJUVBQMDQ1Rp04dVKlSRZ+rJyIiItKYXkJOcnIyZs6cicDAQKX2Pn364Oeff0bx4sX1sRkiIiIijekl5MyePRuXL1/GkCFDUKdOHbx+/RqBgYE4ePAgSpYsiTlz5uhjM0REREQay3fIycrKQkBAAKZNmyY+nBMAXF1d8dVXX+Ho0aMMOURERFTgNBp4PHnyZMTGxqqdlp2djbdv36JUqVIq00qWLImsrKz8VUhERESkA416cm7duoVu3brhm2++gYeHBywtLcVppqamaNCgAZYsWYK0tDTUrl0b2dnZCAgIQGhoKDp27Kh1UYIgYPfu3fDz80NcXBwsLS3Rvn17TJw4EWZmZgCAyMhILF68GNeuXUOxYsXQoUMHzJgxA+bm5lpvj4iIiKRHo5Bz8uRJbN++HevWrcP+/fsxcuRIjBw5EiVLlgQA/Prrrxg/fjwWL14MmUwG4F1QqVOnDmbPnq11URs3bsSyZcswatQotG7dGo8ePcLvv/+Of/75B1u2bEFaWhpcXV1RoUIFeHt7IykpCT4+PkhISMDmzZu13h4RERFJj0Yhx8jICK6urujfvz/Wr1+PjRs3YseOHRg3bhwGDhyIunXr4vjx4wgODkZkZCQUCgXkcjkcHBxgaGioVUEKhQIbNmzAN998A09PTwCAg4MDypQpg0mTJuHu3bu4ePEiUlNTceDAAbFXycrKCu7u7ggNDUXz5s213A1EREQkNVrdDNDMzAyenp44efIk2rZti4ULF6Jbt244fvw4DAwM0LZtW7i5uWHUqFFwcnLSOuAAQHp6Onr37o2ePXsqtVtbWwMAYmNjcf78eTRr1kzptJmTkxNMTU0RFBSk9TaJiIhIenS647GVlRUWLlyIgwcPolatWpg8eTL69++Py5cv57sgc3NzzJ49G82aNVNqP3XqFACgbt26iIiIEENPLgMDA1StWhXR0dH5roGIiIg+f1pfQp6amooSJUrAyMgIderUwbp16xASEgIfHx+4ubmhTZs28PT0hI2Njd6KvH79Ov744w907NgRdevWRWpqKkxNTVXmMzU1RXp6utbr///DiOgzx/eRiKho0vfxWdP1aRRyBEHAli1bsHHjRrx48QIGBgawtbXF7NmzYWtrixYtWmD37t04ceIEli1bhn79+qFnz56YNGkSKleunJ/XgdDQUHz33XeoXr06FixYILbL1LxCQRDUtn9M2bKql7/nh4EBv20LmoGBDOXK6fd9JCJp4jG6YBXm8VmjkLNt2zZ4e3ujcuXK6NKlCzIzM3H58mWMHDkShw8fhpWVFQCga9eu6NSpE3bu3Im1a9fixIkTGDRoEH788Uedijt69ChmzJgBa2trbNq0CWXKlAHwbmyQuh6bzMxMVKxYUevtJCWlQRB0KlEthUKPKyONKBQCnj9PK+wyiOgzwGN0wfoUx2eZTLMOCo1Czo4dO1C9enUcPnwYJiYmAICrV69i+PDhOHDgAMaMGSPOa2hoiCFDhuDLL7/Epk2b4Ovrq1PI2bhxI3777Te0aNECa9asUbrZoLW1NWJiYpTmVygUiIuLQ+fOnbXeliBAryGHCgffQyKioqmwjs8aDTx+/Pgx6tatKwYcALC1tQUAJCQkqF2mZMmS+P777+Hv7691UTt37oSPjw+6du2KTZs2qdxN2dHRESEhIUhOThbbgoODkZGRAUdHR623R0RERNKjUU9OzZo1cfHiRQQGBqJVq1Z48+YN1q5dC5lM9tEBxu9f5q2JZ8+eYdGiRahSpQqGDh2Ke/fuKU2vXr06Bg8eDD8/P7i5uWH8+PFISUmBj48PnJ2dYWdnp9X2iIiISJo0Cjmenp7w8PDAd999J7YJgoAGDRqgX79+ei0oMDAQr1+/xuPHjzFkyBCV6YsWLUK/fv2wdetWLFy4EFOnToWpqSm6du2K6dOn67UWIiIi+nxpFHLatm2L48ePw8/PD9HR0TA3N0eTJk3wzTffwMjISK8F9e/fH/379//ofHK5HL6+vnrdNhEREUmHxvfJqVatms5XSREREREVNJ3ueExERERU1DHkEBERkSQx5BAREZEkMeQQERGRJDHkEBERkSQx5BAREZEkMeQQERGRJDHkEBERkSQx5BAREZEkMeQQERGRJDHkEBERkSQx5BAREZEkMeQQERGRJDHkEBERkSQx5BAREZEkMeQQERGRJDHkEBERkSQx5BAREZEkMeQQERGRJDHkEBERkSQx5BAREZEkMeQQERGRJDHkEBERkSQx5BAREZEkMeQQERGRJDHkEBERkSQx5BAREZEkMeQQERGRJDHkEBERkSQx5BAREZEkMeQQERGRJDHkEBERkSQx5BAREZEkMeQQERGRJDHkEBERkSQx5BAREZEkMeQQERGRJDHkEBERkSQx5BAREZEkMeQQERGRJDHkEBERkSQx5BAREZEkMeQQERGRJDHkEBERkSQx5BAREZEkMeQQERGRJDHkEBERkSQV+ZATHx+P5s2b48qVK0rtkZGRcHd3R7NmzdCqVSvMnDkTqamphVQlERERFTXFCruAD3n8+DFGjRqFtLQ0pfbU1FS4urqiQoUK8Pb2RlJSEnx8fJCQkIDNmzcXUrVERERUlBTJkKNQKLB//354e3urnb5jxw6kpqbiwIEDsLS0BABYWVnB3d0doaGhaN68eUGWS0REREVQkTxd9eDBA/z888/o27ev2qBz/vx5NGvWTAw4AODk5ARTU1MEBQUVZKlERERURBXJnpxKlSrB398fFStWVBmLAwARERHo3r27UpuBgQGqVq2K6OjoAqqSiIiIirIiGXLKlCnzwempqakwNTVVaTc1NUV6errW25PJtF6EiiC+j0RERZO+j8+arq9IhhxNyNS8QkEQ1LZ/TNmypfRRksjAgN+2Bc3AQIZy5fT7PhKRNPEYXbAK8/j8WYYcMzMztT02mZmZqFixotbrS0pKgyDoo7J3FAo9row0olAIeP487eMzEtF/Ho/RBetTHJ9lMs06KD7LkGNtbY2YmBilNoVCgbi4OHTu3Fnr9QkC9BpyqHDwPSQiKpoK6/hcJK+u+hhHR0eEhIQgOTlZbAsODkZGRgYcHR0LsTIiIiIqKj7LkDN48GCYmJjAzc0N/v7+2LNnD6ZNmwZnZ2fY2dkVdnlERERUBHyWIcfS0hJbt26FhYUFpk6dimXLlqFr165YtmxZYZdGRERERUSRH5PTqlUrPHjwQKVdLpfD19e34AsiIiKiz8Jn2ZNDRERE9DEMOURERCRJDDlEREQkSQw5REREJEkMOURERCRJDDlEREQkSQw5REREJEkMOURERCRJDDlEREQkSQw5REREJEkMOURERCRJDDlEREQkSQw5REREJEkMOURERCRJDDlEREQkSQw5REREJEkMOURERCRJDDlEREQkSQw5REREJEkMOURERCRJDDlEREQkSQw5REREJEkMOURERCRJDDlEREQkSQw5REREJEkMOURERCRJDDlEREQkSQw5REREJEkMOURERCRJDDlEREQkSQw5REREJEkMOURERCRJDDlEREQkSQw5REREJEkMOURERCRJDDlEREQkSQw5REREJEkMOURERCRJDDlEREQkSQw5REREJEkMOURERCRJDDlEREQkSQw5REREJEkMOURERCRJDDlEREQkSQw5REREJEkMOURERCRJDDlEREQkSQw5REREJEmffcgJCgpCv3790KRJE7i4uGD9+vUQBKGwyyIiIqJC9lmHnOvXr2PcuHGoXbs2Vq5cid69e2PZsmVYt25dYZdGREREhaxYYReQH6tXr4aNjQ18fHwAAM7OzsjJycGGDRvg5uaG4sWLF3KFREREVFg+256crKwsXLlyBZ07d1Zq79KlCzIzMxEaGlpIlREREVFR8NmGnNjYWGRnZ6NmzZpK7TVq1AAAREdHF3xRREREVGR8tqerUlNTAQBmZmZK7aampgCA9PR0jddlYADoc6yymZkZSpUqhYxXObDrPF5/KyYVGa9yUKpUKZiZmcHgs43sRFSQeIwuGJ/y+CyTaTbfZxtyFAoFAECWxys10GKPWlqW0ktNucLDH+h1fUREpD88Rv93fLZ/+5qbmwNQ7bHJyMgAoNrDQ0RERP8tn23IqV69OgwNDfHo0SOl9tyf69SpUxhlERERURHx2YYcExMTNG/eHP7+/ko3/zt58iTMzc3RuHHjQqyOiIiICttnG3IAYOzYsbh16xYmTpyIwMBALF++HJs2bcKYMWN4jxwiIqL/OJnwmT8Dwd/fHytWrEBUVBSsrKwwZMgQjBw5srDLIiIiokL22YccIiIiInU+69NVRERERHlhyCEiIiJJYsghIiIiSWLI+QzMmDED9erVy/PfwYMHC7tEXLlyBfXq1cOVK1cKuxQioiJn2LBhqFevHgYOHJjnPJMnT0a9evUwY8YMrdY7bNgwfZQoSZ/tYx3+a8qXL49Vq1apnVa9evUCroaIiLQlk8lw8+ZNxMfHo1KlSkrTXr16hYCAgMIpTMIYcj4TxsbGaNq0aWGXQUREOmrYsCEePnyIEydOwM3NTWna2bNnYWJiglKl9Pssxf86nq6SkNOnT6Nfv35o1KgRHB0d8csvvyAzM1OcvnLlSnTt2hWnT59Gz5490ahRI/Tp0wc3btzAzZs3MWDAADRu3Bg9e/bEpUuXVNY9ePBg2NnZwdbWFl27doWfn98H6wkPD8eYMWNgb28Pe3t7eHh4IDY29pO8diKioq5kyZJo27Ytjh8/rjLt2LFj6Nq1K4oV+7++h+TkZMybNw8uLi6wtbVFy5Yt4eHhgbi4uDy3oVAosGHDBnTq1Am2trbo0qULtm3b9klez+eAIeczkpOTo/Iv9zZHhw8fhoeHB2rVqoXVq1dj/PjxOHToEMaNG6f02IuEhAQsWrQI3333HZYvX46XL19iwoQJmDJlCr7++mssXboUCoUCkydPxuvXrwEAAQEB8PDwQMOGDbFmzRqsXLkSVapUgZeXF65fv6621qioKAwcOBBJSUlYvHgxFixYgNjYWAwaNAhJSUmffmcRERVB3bt3x61bt/DkyROxLT09HUFBQejZs6fYJggCxowZgwsXLsDT0xObNm3CuHHjcPHiRcyZMyfP9f/8889YsWIFevfujXXr1qFr165YuHAhVq9e/UlfV1HF01WficePH6Nhw4Yq7RMnTsTYsWPx22+/wcnJCb/99ps4rWbNmnB1dUVgYCDatWsH4N1537lz58LZ2RkAEBERgSVLlmDBggXo378/AODt27eYMGECoqKiUL9+fTx8+BB9+/bFrFmzxHXb2dmhVatWCAkJgb29vUpdq1atQvHixeHr6ys+Eb5169bo2LEjNm7ciB9++EFv+4aI6HPRrl07lCxZEidOnBDvzu/v7w9LS0s0a9ZMnO/p06coUaIEfvjhBzRv3hwA0KpVK8TFxWHnzp1q1x0VFYXdu3djypQpcHd3BwC0adMGMpkM69evx+DBg2FhYfGJX2HRwpDzmShfvjzWrl2r0m5lZYXIyEgkJCRgzJgxyMnJEae1aNECZmZmuHDhghhyACiFknLlygGA0nifMmXKAABSU1MBAKNHjwYAZGZmIiYmBlFRUbhz5w4AIDs7W229ly9fRqtWrVC8eHGxJjMzMzRv3hwXL17U8tUTEUlD8eLF0b59exw/flwMOUePHkX37t0hk8nE+aysrLB161YAwJMnT/Do0SNERETg+vXrHzzuCoKA9u3bK30XtG/fHmvXrsW1a9fQsWPHT/jqih6GnM+EsbExGjVqpHZa7jiXefPmYd68eSrTnz59qvRzbs/K+z70QNPk5GTMnTsXp0+fhkwmQ40aNcS/OPJ6KkhKSgqOHTuGY8eOqUyztLTMc1tERFLXrVs3cWyNqakpLl26hEmTJqnMd+jQISxduhTx8fEoU6YMbGxsPnisTklJAQD06NFD7fTExER9lP9ZYciRAHNzcwDA9OnT0bJlS5XppUuXztf6p06dioiICGzZsgX29vYwNjbGq1evsGfPnjyXKVWqFBwcHFSuIACgNLCOiOi/xtnZGaVKlcLJkydRqlQpVK1aFba2tkrzhIaG4ocffsDQoUMxatQoVKxYEQDg7e2Na9euqV1v7nfBn3/+CVNTU5XplStX1vMrKfr4bSMBtWrVQtmyZREXF4dRo0aJ7c+ePcO0adMwcODAfN1L59q1a/jmm2/wxRdfiG1BQUEA3o3kV6dly5Z4+PAh6tevL4YaQRAwdepU1KhRA/Xr19e5HiKiz5mxsTE6dOiAU6dOoWTJkmp7Xm7cuAGFQoEJEyaIl5W/fftWPN2vUChgYKB87VCLFi0AAC9evFA6XgcHB8PX1xczZ85E2bJlP9XLKpIYciTA0NAQkydPxpw5c2BoaAgXFxekpqZizZo1SExMVDtgWRuNGzfG4cOH0bBhQ1SsWBE3btzA+vXrIZPJ8OrVK7XLjBs3DgMHDsSYMWMwaNAgmJiYYNeuXTh9+jRWrFiRr3qIiD533bt3x5gxY2BgYICffvpJZXrjxo0BAPPnz8dXX32F1NRU+Pn54e+//wbwbozkv4ceyOVy9O7dG7Nnz8bjx49ha2uLqKgoLFu2DFWrVkXNmjU/+esqahhyJGLAgAEwNTXFxo0bsWvXLpQsWRL29vb47bffUK1atXyte/HixfDy8oKXlxeAd1dtzZs3D4cOHUJoaKjaZWxsbLB9+3YsW7YM06dPhyAIkMvlWL16NTp06JCveoiIPncODg4wNzdHpUqVULt2bZXprVq1wpw5c7BlyxacOHEC5cqVQ6tWrbBq1Sp4eHjg2rVraNu2rcpyixYtwvr167Fz504kJCSgbNmy6N69OyZNmgRDQ8OCeGlFikzIa+QoERER0WeMNwMkIiIiSWLIISIiIkliyCEiIiJJYsghIiIiSWLIISIiIkliyCEiIiJJYsghIiIiSeLNAInok1AoFLh48SIOHjyIe/fuITExEdnZ2bC0tETjxo3RrVs3dOnSRenJy4Vt3759+PHHH9GrVy/89ttvhV0OEeUTQw4R6V1MTAwmTZqEsLAwAO9uN9+yZUsYGhri0aNHOHnyJE6cOIFmzZph9erVsLCwKOSKiUiKGHKISK8SExMxePBgPHv2DC4uLpg5c6bKA2IjIyMxe/Zs8UnLGzZsKKRqiUjKOCaHiPTq559/xrNnz9CtWzesXbtWJeAAQK1atfDHH3+gSpUqCAwMxI0bNwqhUiKSOvbkEJHexMTE4OzZszA1NcXPP//8wfE2JUuWxIQJE3D37l2ULFlSadrt27exceNGhISEIC0tDVZWVujYsSPGjBkDS0tLpXnr1asHGxsb+Pn5YcWKFfD398fz589RqVIl9OjRA2PGjEGJEiWUlklLS8Mff/yB48ePIzExEVWrVsWIESNgZGSUZ71RUVFYt24dLl68iBcvXsDS0hJOTk4YN24cqlSpojRv+/btkZqair/++gszZsxAeHg4ypcvjyVLlsDe3l7T3UlE+cQHdBKR3qxatQorV65Ez549sWTJEp3WsX//fsyaNQsKhQINGjRA1apVcf/+fcTExKBKlSrYunUrqlatKs5fr1491KhRAyYmJoiJiYGdnR2MjIxw6dIlZGdnw8XFBevWrRPnf/nyJYYNG4YHDx6gQoUKsLOzQ1xcHMLCwlCnTh08fPhQZeDx+fPnMX78eLx69QpyuRw1a9ZEdHQ0wsPDUaZMGWzatAm2trbi/O3bt0dSUhIsLCxgaGgIGxsb3L9/H0eOHFEJdET06bAnh4j05s6dOwCAVq1a6bR8REQEZs+eDRMTE6xZswatW7cG8O5Krd9//x3r1q3DtGnTsGPHDqXlHj16hNq1a+PIkSOoVq0aAODu3bsYOHAgzp07h/DwcMjlcgDAihUr8ODBA7i4uGD58uUoXrw4AGDPnj346aefVGpKTk7GlClTkJWVhaVLl6JHjx7itF27dmHOnDmYNGkSjh07BmNjY3Ha69evYWlpiR07dsDExAQKhQIGBhwhQFSQGHKISG8SEhIAABUqVFA7fc6cOcjMzFRpt7a2hoeHB7Zu3Yrs7GxMnjxZDDgAYGBggEmTJuHs2bO4fv06bt68iaZNmyqtY8qUKWLAAQBbW1vY29vjypUr+OeffyCXy5GVlYV9+/bByMgICxYsEAMOAAwYMADnzp3DmTNnlNa7d+9evHz5EkOGDFEKOADwzTff4OzZswgICIC/v7/K9EGDBsHExER8DURUsPipIyK9USgUH5x+7NgxHD58WOXf+fPnAQCXL18GAKWAk0smk8HJyQkAcPXqVZXpdnZ2Km25YevVq1cA3vU0ZWZmwtbWFmXLllWZv3PnziptV65cybMmAHB2dlaa7325vUdEVDjYk0NEemNlZYXw8HA8e/ZM7fTQ0FCln69cuYLhw4eLP+f2BH355Zcf3E58fLxKm7m5uUpbsWLvDnG5Qw+fPn0q1qnO+2N9/r2t8ePHf7Cm3No/VhMRFRyGHCLSm/r16yM4OBjXr1/HgAEDtF7+7du3AIAePXp88PSOjY2NSpsmd07+2DyGhoZ51uTi4gIzM7M8l61Tp45KG09RERUuhhwi0pvevXtjw4YN8Pf3x48//qh1T0aFChXw+PFjTJw4ETVq1NB7fRUrVgQAPH78WO303J6ef9cUHR2N4cOHw8HBQe81EdGnwz8ziEhv6tati06dOiEtLU28DPxD7t27p/RzixYtAACBgYFq558+fTq+/vprnD17Vqf6GjZsiNKlS+PevXt48uSJyvSAgACVto/VtGTJEvTr1w979uzRqSYi+nQYcohIrxYsWIAqVarg1KlTGD58OO7fv68yT0REBCZPnozFixcDgDgIeNiwYTA0NMTvv/+OS5cuKS2zc+dOHDx4EPfv30fjxo11qs3IyAiDBw/G27dvMX36dKSnp4vT/P39ceDAAZVlvvnmG5QsWRJ+fn44evSo0rRz585hy5YtCAsLU7pPDhEVDTxdRUR6Vbp0aezduxczZ87EuXPn0LdvX9SoUQPW1tYoVqwYoqOj8fDhQwCAmZkZRo0ahW+//RbAu8u+Z86ciV9++QWurq7izQCjoqLwzz//wNDQED4+PihXrpzO9Y0bNw7Xr1/HlStX0LFjR7Ro0QLPnz/H9evXYWdnp/KICSsrK/z666+YMmUKpkyZgtWrV6NWrVqIj4/H3bt3AQA//vgj6tevr3NNRPRpMOQQkd5ZWlpi3bp1uHHjBg4dOoRr167h5s2byMzMhKWlJdq1awcnJyf07dtXZTDv0KFDUb9+fWzZsgXXrl3DP//8gwoVKqBbt25wd3dHgwYN8lWbsbExNm7cCF9fX+zfvx+BgYEoX748PD09YWdnh6FDh6os07lzZ/zvf//Dxo0bcfnyZQQEBKBs2bJo164dRo4cqfPND4no0+JjHYiIiEiSOCaHiIiIJIkhh4iIiCSJIYeIiIgkiSGHiIiIJIkhh4iIiCSJIYeIiIgkiSGHiIiIJIkhh4iIiCSJIYeIiIgkiSGHiIiIJIkhh4iIiCSJIYeIiIgkiSGHiIiIJOn/ASNi10Ru7MwjAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "t, p = ttest_ind(list(rp_data['female_screentime']), list(rp_data['male_screentime']))\n",
    "rp_data['male_screentime'] = rp_data['masculine_faces']/(rp_data['feminine_faces']+rp_data['masculine_faces'])\n",
    "t, p = ttest_ind(list(rp_data['female_screentime']), list(rp_data['male_screentime']))\n",
    "\n",
    "average_female_screentime=rp_data['female_screentime'].mean()\n",
    "average_male_screentime=rp_data['male_screentime'].mean()\n",
    "\n",
    "median_female_screentime=rp_data['female_screentime'].median()\n",
    "median_male_screentime=rp_data['male_screentime'].median()\n",
    "print('Average screentimes:',average_female_screentime,average_male_screentime)\n",
    "print('Median screentimes:',median_female_screentime,median_male_screentime)\n",
    "\n",
    "fig, ax = plt.subplots()\n",
    "keys=['Female','Male']\n",
    "print(f\"N={len(rp_data['female_screentime'])}\")\n",
    "# create a bar plot of the dictionary\n",
    "mean_screentime_gender=[average_female_screentime*100,average_male_screentime*100]\n",
    "ax.bar(keys,mean_screentime_gender,\n",
    "       edgecolor='black',linewidth=2,width=0.25)\n",
    "for i, v in enumerate(mean_screentime_gender):\n",
    "    ax.annotate(str(round((v/sum(mean_screentime_gender))*100, 2)) + \"%\", xy=(i, v), ha='center', va='bottom', fontsize=15)\n",
    "\n",
    "# set the title and axis labels\n",
    "ax.set_title('Average screentime x Gender', fontsize=20)\n",
    "ax.set_xlabel('Gender', fontsize=16)\n",
    "ax.set_ylabel('% Screentime', fontsize=16)\n",
    "\n",
    "# get the x-tick labels\n",
    "labels = ax.get_xticklabels()\n",
    "yticks = ax.get_yticks()\n",
    "ax.set_yticks(yticks)\n",
    "ax.tick_params(axis='y', labelsize=16)\n",
    "ax.tick_params(axis='x', labelsize=16)\n",
    "\n",
    "#plt.show()\n",
    "#save the plots\n",
    "plt.savefig('plots/screentime_x_gender.eps', bbox_inches=\"tight\", dpi=300, format='eps')\n",
    "fig, ax = plt.subplots()\n",
    "\n",
    "keys=['Female','Male']\n",
    "ax.bar(keys, [median_female_screentime*100,median_male_screentime*100],\n",
    "           edgecolor='black',\n",
    "           linewidth=2, color='#F5DEB3', width=0.25)\n",
    "\n",
    "# set the title and axis labels\n",
    "ax.set_title('Median screentime x Gender', fontsize=20)\n",
    "ax.set_xlabel('Gender', fontsize=16)\n",
    "ax.set_ylabel('% Screentime', fontsize=16)\n",
    "\n",
    "# get the x-tick labels\n",
    "labels = ax.get_xticklabels()\n",
    "yticks = ax.get_yticks()\n",
    "ax.set_yticks(yticks)\n",
    "ax.tick_params(axis='y', labelsize=12)\n",
    "ax.tick_params(axis='x', labelsize=12)\n",
    "\n",
    "#plt.show()\n",
    "#plt.savefig('plots/median_gender_screentime.eps', bbox_inches=\"tight\", dpi=300, format='eps')\n",
    "# create a figure and axis object\n",
    "\n",
    "fig, ax = plt.subplots()\n",
    "\n",
    "fm_sc_time=[x*100 for x in list(rp_data['female_screentime'])]\n",
    "m_sc_time=[x*100 for x in list(rp_data['male_screentime'])]\n",
    "t, p = ttest_ind(fm_sc_time, m_sc_time)\n",
    "print(f'p-value for screentime x gender: {p}')\n",
    "\n",
    "# create a boxplot of the data with custom labels\n",
    "ax.boxplot([fm_sc_time,m_sc_time], labels=['Female','Male'],\n",
    "              patch_artist=True,\n",
    "              boxprops={'linewidth': 2},\n",
    "              whiskerprops={'linewidth': 2},\n",
    "              medianprops={'linewidth': 2, 'color': 'red'})\n",
    "\n",
    "# set the title and axis labels\n",
    "ax.set_title('Average screentime x Gender (Box Plot)',fontsize=20)\n",
    "ax.set_xlabel('Gender',fontsize=16)\n",
    "ax.set_ylabel('% Screentime', fontsize=16)\n",
    "\n",
    "# get the x-tick labels\n",
    "labels = ax.get_xticklabels()\n",
    "ax.set_xticklabels(labels, fontsize=16)\n",
    "yticks = ax.get_yticks()\n",
    "ax.set_yticks(yticks)\n",
    "ax.tick_params(axis='y', labelsize=16)\n",
    "ax.set_ylim(-2,110)\n",
    "\n",
    "# set the rotation angle for the labels\n",
    "#ax.set_xticklabels(labels, rotation=90, fontsize=16)\n",
    "#plt.tight_layout()\n",
    "plt.savefig('plots/screentime_x_gender_boxplot.eps', bbox_inches=\"tight\", dpi=300)\n",
    "\n",
    "\n",
    "#### BY LANGUAGE (STACKED)  ####\n",
    "gender=('Male', 'Female')\n",
    "width = 0.35  # the width of the bars\n",
    "lng_data = rp_data.groupby(\"Programme Language\")\n",
    "x = np.arange(len(lng_data))  # the label locations\n",
    "fig, ax = plt.subplots()\n",
    "#bottom = 0\n",
    "gxlng = {'Male':[], 'Female':[]}\n",
    "mul = 0\n",
    "\n",
    "for lang in lng_data.groups.keys():\n",
    "    msc = lng_data.get_group(lang)['male_screentime']\n",
    "    fsc = lng_data.get_group(lang)['female_screentime']\n",
    "    print(f'{lang}-> N={len(msc)}, {len(fsc)}')\n",
    "    gxlng['Male'].append(round(np.mean(msc)*100, 2))\n",
    "    gxlng['Female'].append(round(np.mean(fsc)*100, 2))\n",
    "    t, p = ttest_ind(msc, fsc)\n",
    "    print(f'p-value for gender for {lang} language: {p}')\n",
    "    #bar_value = [np.mean(msc)*100, np.mean(fsc)*100]\n",
    "    #offset = width * multiplier\n",
    "for gen in ['Female', 'Male']:\n",
    "    #print(offset)\n",
    "    #rects = ax.bar(x, gxlng[gen], width, label=gen, bottom=bottom, edgecolor='black',linewidth=2)\n",
    "    offset = width * mul\n",
    "    rects = ax.bar(x+offset, gxlng[gen], width, label=gen, edgecolor='black',linewidth=2)\n",
    "    ax.bar_label(rects, padding=3, fontsize=12, )\n",
    "    #bottom = gxlng[gen]\n",
    "    mul += 1\n",
    "\n",
    "ax.set_xlabel('Language', fontsize=16)\n",
    "ax.set_ylabel('% Screentime',fontsize=16)\n",
    "ax.set_title('Screentime x Gender (By Language)', fontsize=20)\n",
    "ax.set_xticks(x+0.5*width, list(lng_data.groups.keys()))\n",
    "\n",
    "ax.legend(bbox_to_anchor=(1.25, 0.5), ncol=3,fontsize=12, loc='upper center')\n",
    "ax.tick_params(axis='y', labelsize=16)\n",
    "ax.tick_params(axis='x', labelsize=16)\n",
    "\n",
    "filename=\"screentime_x_gender_x_language.eps\"\n",
    "plt.savefig('plots/'+filename, bbox_inches=\"tight\", dpi=300)\n",
    "\n",
    "## multiple comparisons t-test\n",
    "fsc = {}\n",
    "for lang in lng_data.groups.keys():\n",
    "    fsc[lang] = lng_data.get_group(lang)['female_screentime']\n",
    "combs = list(combinations(lng_data.groups.keys(), 2))\n",
    "num_comb = len(combs)\n",
    "\n",
    "for comb in combs:\n",
    "    #print(np.mean(fsc[comb[0]]), np.mean(fsc[comb[1]]))\n",
    "    t, p = ttest_ind(fsc[comb[0]], fsc[comb[1]])\n",
    "    print(f'corrected p-value for languages {comb[0]} and {comb[1]}: {p/num_comb}')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "ttest_ind(fsc['BENGALI'].values, fsc['HINDI'].values).pvalue*1e5"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "cHKhxvKPncmr"
   },
   "source": [
    "## II. Screentime x Age"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 1000
    },
    "id": "zOutaYhMna-f",
    "outputId": "d4053348-9e75-4a4b-f8da-a593258013c2"
   },
   "outputs": [],
   "source": [
    "dict_age_mapping={'age_1':\"0-18\",\n",
    "                 'age_2':\"18-33\",\n",
    "                 'age_3': \"33-60\",\n",
    "                  'age_4': \"60+\"}\n",
    "mean_age_wise_screentime=[]\n",
    "median_age_wise_screentime=[]\n",
    "values=[]\n",
    "age_screentime_total_list=[]\n",
    "\n",
    "\n",
    "for key in list(dict_age_mapping.keys()):\n",
    "    age_screentime_total_list.append(list([x*100 for x in rp_data[key + '_screen_time']]))\n",
    "    mean_age_wise_screentime.append(rp_data[key+\"_screen_time\"].mean()*100)\n",
    "\n",
    "\n",
    "fig, ax = plt.subplots()\n",
    "keys = dict_age_mapping.values()\n",
    "ax.bar(keys, mean_age_wise_screentime,\n",
    "           edgecolor='black',\n",
    "           linewidth=2, width=0.25)\n",
    "for i, v in enumerate(mean_age_wise_screentime):\n",
    "    ax.annotate(str(round(v, 2)) + \"%\", xy=(i, v+1), ha='center', va='bottom', fontsize=12)\n",
    "\n",
    "\n",
    "# set the title and axis labels\n",
    "ax.set_title('Screentime vs Age', fontsize=20)\n",
    "ax.set_xlabel('Age', fontsize=16)\n",
    "ax.set_ylabel('% Screentime', fontsize=16)\n",
    "\n",
    "# get the x-tick labels\n",
    "labels = ax.get_xticklabels()\n",
    "yticks = ax.get_yticks()\n",
    "ax.set_yticks(yticks)\n",
    "ax.tick_params(axis='y', labelsize=12)\n",
    "ax.tick_params(axis='x', labelsize=16)\n",
    "\n",
    "#plt.show()\n",
    "plt.savefig('plots/screentime_x_age.eps', bbox_inches=\"tight\", dpi=300)\n",
    "# create a figure and axis object\n",
    "\n",
    "\n",
    "fig, ax = plt.subplots()\n",
    "ax.boxplot(age_screentime_total_list, labels=[dict_age_mapping[k] for k in list(dict_age_mapping.keys())],\n",
    "          patch_artist=True,\n",
    "          boxprops={'linewidth': 2},\n",
    "          whiskerprops={'linewidth': 2},\n",
    "          medianprops={'linewidth': 2, 'color': 'red'})\n",
    "\n",
    "# set the title and axis labels\n",
    "ax.set_title('Screentime vs Age',fontsize=20)\n",
    "ax.set_xlabel('Age',fontsize=16)\n",
    "ax.set_ylabel('% Screentime', fontsize=16)\n",
    "\n",
    "# get the x-tick labels\n",
    "labels = ax.get_xticklabels()\n",
    "yticks = ax.get_yticks()\n",
    "ax.set_yticks(yticks)\n",
    "ax.tick_params(axis='y', labelsize=12)\n",
    "ax.set_ylim(-4,110)\n",
    "\n",
    "# set the rotation angle for the labels\n",
    "ax.set_xticklabels(labels, rotation=0, fontsize=16)\n",
    "#plt.tight_layout()\n",
    "plt.savefig('plots/screentime_x_age_boxplot.eps', bbox_inches=\"tight\", dpi=300)\n",
    "\n",
    "\n",
    "age_mean_ids=np.argsort(-np.array(mean_age_wise_screentime))\n",
    "age_keys=list(dict_age_mapping.keys())\n",
    "sorted_age_screentime=[mean_age_wise_screentime[i] for i in age_mean_ids]\n",
    "age_keys_sorted=[age_keys[i] for i in age_mean_ids]\n",
    "pvalues_list=[]\n",
    "confidence=0.05\n",
    "for i in range(len(age_keys_sorted)-1):\n",
    "\n",
    "        c_i=rp_data[age_keys_sorted[i]+\"_screen_time\"]\n",
    "        c_j=rp_data[age_keys_sorted[i+1]+\"_screen_time\"]\n",
    "\n",
    "        t, p = ttest_ind(list(c_i), list(c_j))\n",
    "        pvalues_list.append(p)\n",
    "        print(f\"t-test between {age_keys_sorted[i]} and {age_keys_sorted[i+1]}: t={t}, p={p}\")\n",
    "\n",
    "significant = np.array(pvalues_list)/(len(age_keys_sorted) - 1) < confidence\n",
    "for j in range(len(age_keys_sorted)):\n",
    "    if j == len(age_keys_sorted) - 1:\n",
    "        print(f\"{age_keys_sorted[j]}\")\n",
    "        break\n",
    "    cat = age_keys_sorted[j]\n",
    "    sign = \">\" if significant[j] else \"=\"\n",
    "    print(f\"{cat} {sign} \", end=\"\")\n",
    "\n",
    "### BY LANGUAGE\n",
    "age_groups = dict_age_mapping.keys()\n",
    "width = 0.2  # the width of the bars\n",
    "lng_data = rp_data.groupby(\"Programme Language\")\n",
    "x = np.arange(len(lng_data))  # the label locations\n",
    "fig, ax = plt.subplots()\n",
    "mul = 0\n",
    "#bottom = [0, 0, 0, 0, 0]\n",
    "gxage = {k:[] for k in age_groups}\n",
    "\n",
    "for lang in lng_data.groups.keys():\n",
    "    for age in age_groups:\n",
    "        d = lng_data.get_group(lang)\n",
    "        asc = d[age + '_screen_time']\n",
    "        gxage[age].append(round(np.mean(asc)*100, 2))\n",
    "\n",
    "for age in age_groups:\n",
    "    #print(age, dict_age_mapping[age])\n",
    "    #rects = ax.bar(x, gxage[age], width, label=dict_age_mapping[age], bottom=bottom, edgecolor='black',linewidth=2)\n",
    "    offset = width * mul\n",
    "    rects = ax.bar(x + offset, gxage[age], width, label=dict_age_mapping[age], edgecolor='black',linewidth=2)\n",
    "   # if age!='age_3':\n",
    "    ax.bar_label(rects, padding=3, fontsize=12, )\n",
    "    mul += 1\n",
    "    #bottom = [bottom[i] + gxage[age][i] for i in range(5)]\n",
    "\n",
    "combns = list(combinations(lng_data.groups.keys(), 2))\n",
    "age = 'age_3'\n",
    "for comb in combns:\n",
    "    l0 = comb[0]\n",
    "    l1 = comb[1]\n",
    "    pval = ttest_ind(lng_data.get_group(l0)[age + '_screen_time'], lng_data.get_group(l1)[age + '_screen_time']).pvalue\n",
    "    print(f'p-value for lang {l0} and {l1}: {pval/len(combns)}')\n",
    "\n",
    "ax.set_xlabel('Language', fontsize=16)\n",
    "ax.set_ylabel('% Screentime',fontsize=16)\n",
    "ax.set_title('Screentime x Age (By Language)', fontsize=20)\n",
    "ax.set_xticks(x+1.5*width, list(lng_data.groups.keys()))\n",
    "\n",
    "ax.legend(bbox_to_anchor=(0.86, 1.06), ncol=3,fontsize=12, loc='upper center')\n",
    "ax.tick_params(axis='y', labelsize=16)\n",
    "ax.tick_params(axis='x', labelsize=16)\n",
    "\n",
    "filename=\"screentime_x_age_x_language.eps\"\n",
    "plt.savefig('plots/'+filename, bbox_inches=\"tight\", dpi=300)\n",
    "\n",
    "#### SCREENTIME X GENDER X AGE X LANGUAGE\n",
    "age_groups = dict_age_mapping.keys()\n",
    "width = 0.2  # the width of the bars\n",
    "lng_data = rp_data.groupby(\"Programme Language\")\n",
    "x = np.arange(len(lng_data))  # the label locations\n",
    "fig, ax = plt.subplots()\n",
    "mul = 0\n",
    "#bottom = [0, 0, 0, 0, 0]\n",
    "gxage_male = {k:[] for k in age_groups}\n",
    "gxage_raw = {k:{} for k in lng_data.groups.keys()}\n",
    "\n",
    "dict_age_range = {'age_1': '[0, 18)', 'age_2': '[18, 33)', 'age_3': '[33, 60)', 'age_4': '[60, inf)'}\n",
    "for lang in lng_data.groups.keys():\n",
    "    d = lng_data.get_group(lang)\n",
    "    total_male_faces = d['masculine_faces']\n",
    "    for age in age_groups:\n",
    "        relevant_cols = [x for x in d.columns if '(male' in x and dict_age_range[age] in x]\n",
    "        asc = d[relevant_cols].sum(axis=1)/total_male_faces\n",
    "        gxage_raw[lang][age] = asc\n",
    "        gxage_male[age].append(round(np.mean(asc)*100, 2))\n",
    "\n",
    "combns = list(combinations(lng_data.groups.keys(), 2))\n",
    "num_c = len(combns)\n",
    "\n",
    "\n",
    "for age in age_groups:\n",
    "    print(age)\n",
    "    for comb in combns:\n",
    "        t, p = ttest_ind(gxage_raw[comb[0]][age], gxage_raw[comb[1]][age])\n",
    "        print(f\"{comb[0]} and {comb[1]}, pvalue: {p/num_c}\")\n",
    "    #print()\n",
    "        \n",
    "for age in age_groups:\n",
    "    #print(age, dict_age_mapping[age])\n",
    "    #rects = ax.bar(x, gxage[age], width, label=dict_age_mapping[age], bottom=bottom, edgecolor='black',linewidth=2)\n",
    "    offset = width * mul\n",
    "    rects = ax.bar(x + offset, gxage_male[age], width, label=dict_age_mapping[age], edgecolor='black',linewidth=2)\n",
    "   # if age!='age_3':\n",
    "    ax.bar_label(rects, padding=3, fontsize=12, )\n",
    "    mul += 1\n",
    "    #bottom = [bottom[i] + gxage[age][i] for i in range(5)]\n",
    "\n",
    "\n",
    "ax.set_xlabel('Language', fontsize=16)\n",
    "ax.set_ylabel('% Screentime',fontsize=16)\n",
    "ax.set_title('Screentime x Male x Age (By Language)', fontsize=20)\n",
    "ax.set_xticks(x+1.5*width, list(lng_data.groups.keys()))\n",
    "\n",
    "ax.legend(bbox_to_anchor=(0.86, 1.06), ncol=3,fontsize=12, loc='upper center')\n",
    "ax.tick_params(axis='y', labelsize=16)\n",
    "ax.tick_params(axis='x', labelsize=16)\n",
    "\n",
    "filename=\"screentime_x_male_age_x_language.eps\"\n",
    "plt.savefig('plots/'+filename, bbox_inches=\"tight\", dpi=300)\n",
    "\n",
    "\n",
    "#### SCREENTIME X GENDER X AGE X LANGUAGE\n",
    "age_groups = dict_age_mapping.keys()\n",
    "width = 0.2  # the width of the bars\n",
    "lng_data = rp_data.groupby(\"Programme Language\")\n",
    "x = np.arange(len(lng_data))  # the label locations\n",
    "fig, ax = plt.subplots()\n",
    "mul = 0\n",
    "#bottom = [0, 0, 0, 0, 0]\n",
    "gxage_female = {k:[] for k in age_groups}\n",
    "gxage_raw = {k:{} for k in lng_data.groups.keys()}\n",
    "\n",
    "dict_age_range = {'age_1': '[0, 18)', 'age_2': '[18, 33)', 'age_3': '[33, 60)', 'age_4': '[60, inf)'}\n",
    "for lang in lng_data.groups.keys():\n",
    "    d = lng_data.get_group(lang)\n",
    "    total_female_faces = d['feminine_faces']\n",
    "    for age in age_groups:\n",
    "        relevant_cols = [x for x in d.columns if '(female' in x and dict_age_range[age] in x]\n",
    "        asc = d[relevant_cols].sum(axis=1)/total_female_faces\n",
    "        gxage_raw[lang][age] = asc\n",
    "        gxage_female[age].append(round(np.mean(asc)*100, 2))\n",
    "\n",
    "for age in age_groups:\n",
    "    #print(age, dict_age_mapping[age])\n",
    "    #rects = ax.bar(x, gxage[age], width, label=dict_age_mapping[age], bottom=bottom, edgecolor='black',linewidth=2)\n",
    "    offset = width * mul\n",
    "    rects = ax.bar(x + offset, gxage_female[age], width, label=dict_age_mapping[age], edgecolor='black',linewidth=2)\n",
    "   # if age!='age_3':\n",
    "    ax.bar_label(rects, padding=3, fontsize=12, )\n",
    "    mul += 1\n",
    "    #bottom = [bottom[i] + gxage[age][i] for i in range(5)]\n",
    "\n",
    "    \n",
    "combns = list(combinations(lng_data.groups.keys(), 2))\n",
    "num_c = len(combns)\n",
    "\n",
    "\n",
    "for age in age_groups:\n",
    "    print(age)\n",
    "    for comb in combns:\n",
    "        t, p = ttest_ind(gxage_raw[comb[0]][age], gxage_raw[comb[1]][age])\n",
    "        print(f\"{comb[0]} and {comb[1]}, pvalue: {p/num_c}\")\n",
    "    #print()\n",
    "\n",
    "ax.set_xlabel('Language', fontsize=16)\n",
    "ax.set_ylabel('% Screentime',fontsize=16)\n",
    "ax.set_title('Screentime x Female x Age (By Language)', fontsize=20)\n",
    "ax.set_xticks(x+1.5*width, list(lng_data.groups.keys()))\n",
    "\n",
    "ax.legend(bbox_to_anchor=(0.86, 1.06), ncol=3,fontsize=12, loc='upper center')\n",
    "ax.tick_params(axis='y', labelsize=16)\n",
    "ax.tick_params(axis='x', labelsize=16)\n",
    "\n",
    "filename=\"screentime_x_female_x_age_x_language.eps\"\n",
    "plt.savefig('plots/'+filename, bbox_inches=\"tight\", dpi=300)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "5lVfDdvox0ge"
   },
   "source": [
    "## III. Screentime x Skintone"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 1000
    },
    "id": "kmL0jGnnwksB",
    "outputId": "79694832-704e-4d5d-b999-ebeedc3f5410"
   },
   "outputs": [],
   "source": [
    "skintone_strings=[\"[-inf, 1.1)\", \"[1.1, 2.1)\",\"[2.1, 3.1)\",\"[3.1, 4.1)\",\"[4.1, 5.1)\",\n",
    "                  \"[5.1, 6.1)\",\"[6.1, 7.1)\",\"[7.1, 8.1)\",\"[8.1, 9.1)\",\"[9.1, 10.1)\"]\n",
    "\n",
    "mst_skintone_dict={'light': [\"[-inf, 1.1)\",\"[1.1, 2.1)\",\"[2.1, 3.1)\"],\n",
    "                   'medium': [\"[3.1, 4.1)\",\"[4.1, 5.1)\",\"[5.1, 6.1)\",\"[6.1, 7.1)\"],\n",
    "                   'dark': [\"[7.1, 8.1)\",\"[8.1, 9.1)\",\"[9.1, 10.1)\"]\n",
    "                  }\n",
    "mst_interval_dict={\"[-inf, 1.1)\":\"light\", \"[1.1, 2.1)\":\"light\", \"[2.1, 3.1)\": \"light\",\n",
    "                    \"[3.1, 4.1)\": \"medium\", \"[4.1, 5.1)\":\"medium\", \"[5.1, 6.1)\": \"medium\", \"[6.1, 7.1)\": \"medium\",\n",
    "                     \"[7.1, 8.1)\": \"dark\", \"[8.1, 9.1)\":\"dark\", \"[9.1,10.1)\":\"dark\"}\n",
    "\n",
    "mst_skintone_group_dict={'light':[],'medium':[],'dark':[]}\n",
    "\n",
    "\n",
    "dict_skintone_screentime=dict()\n",
    "\n",
    "for string_example in skintone_strings:\n",
    "    if string_example in rp_data:\n",
    "        continue\n",
    "    col_sample=[col for col in rp_data.columns if string_example in col]\n",
    "    sample_df=rp_data[col_sample]\n",
    "    curr_sum=sample_df.sum(axis=1)\n",
    "    rp_data[string_example]=curr_sum\n",
    "\n",
    "#compute light, medium and dark screentime\n",
    "rp_data['light']=rp_data['[-inf, 1.1)']+rp_data['[1.1, 2.1)']+rp_data['[2.1, 3.1)']\n",
    "rp_data['medium']=rp_data['[3.1, 4.1)']+rp_data['[4.1, 5.1)']+rp_data['[5.1, 6.1)']+rp_data['[6.1, 7.1)']\n",
    "rp_data['dark']=rp_data['[7.1, 8.1)']+rp_data['[8.1, 9.1)']+rp_data['[9.1, 10.1)']\n",
    "\n",
    "#assert with light, medium and dark\n",
    "cols_to_sum=['light','medium','dark']\n",
    "\n",
    "check_equal=0\n",
    "#loop over and check the sums with total_faces\n",
    "for i in np.arange(rp_data.shape[0]):\n",
    "    if(rp_data['light'].iloc[i]+rp_data['medium'].iloc[i]+rp_data['dark'].iloc[i]==rp_data['total_faces'].iloc[i]):\n",
    "        check_equal=check_equal+1\n",
    "    else:\n",
    "        sum_mst_skintone=rp_data['light'].iloc[i]+rp_data['medium'].iloc[i]+rp_data['dark'].iloc[i]\n",
    "        print(\"++++++++++++++++++++++++++++++++++++++++++++++++++++++++\")\n",
    "        print('Sum mst skintone:%d' %(sum_mst_skintone))\n",
    "        print('Total faces:%d' %(rp_data['total_faces'].iloc[i]))\n",
    "        print('Feminine faces:%d' %(rp_data['feminine_faces'].iloc[i]))\n",
    "        print('Masculine faces:%d' %(rp_data['masculine_faces'].iloc[i]))\n",
    "        print(rp_data['Cat No.'].iloc[i])\n",
    "\n",
    "#plot the skintone unconditional distribution\n",
    "def plot_skintone_screentime(data,keys,fname,metric=\"median\"):\n",
    "    fig, ax = plt.subplots()\n",
    "    # create a bar plot of the dictionary\n",
    "    ax.bar(keys, data, edgecolor='black',linewidth=2, width=0.25, align='center')\n",
    "\n",
    "    for i, v in enumerate(data):\n",
    "        ax.annotate(str(round((v/sum(data))*100, 2)) + \"%\", xy=(i, v), ha='center', va='bottom', fontsize=15)\n",
    "\n",
    "    # set the title and axis labels\n",
    "    title=metric+\" screentime x Skintone\"\n",
    "    plt.xticks([0,1,2], keys)\n",
    "    ax.set_title(title, fontsize=16)\n",
    "    ax.set_xlabel('Skintone', fontsize=12)\n",
    "    ax.set_ylabel('% Screentime', fontsize=12)\n",
    "\n",
    "    # get the x-tick labels\n",
    "    labels = ax.get_xticklabels()\n",
    "    yticks = ax.get_yticks()\n",
    "    ax.set_yticks(yticks)\n",
    "    ax.tick_params(axis='y', labelsize=12)\n",
    "    ax.tick_params(axis='x', labelsize=12)\n",
    "\n",
    "    #plt.show()\n",
    "    #save the plots\n",
    "    #filename=metric+\"_skintone_screentime.png\"\n",
    "    plt.savefig('plots/'+fname, bbox_inches=\"tight\", dpi=300)\n",
    "\n",
    "\n",
    "rp_data['light_screentime']=rp_data['light']/rp_data['total_faces']\n",
    "rp_data['medium_screentime']=rp_data['medium']/rp_data['total_faces']\n",
    "rp_data['dark_screentime']=rp_data['dark']/rp_data['total_faces']\n",
    "\n",
    "\n",
    "#mean and median values of screentime\n",
    "light_screentime_mean=rp_data['light_screentime'].mean()*100\n",
    "medium_screentime_mean=rp_data['medium_screentime'].mean()*100\n",
    "dark_screentime_mean=rp_data['dark_screentime'].mean()*100\n",
    "\n",
    "light_screentime_median=rp_data['light_screentime'].median()*100\n",
    "medium_screentime_median=rp_data['medium_screentime'].median()*100\n",
    "dark_screentime_median=rp_data['dark_screentime'].median()*100\n",
    "\n",
    "keys=['light','medium','dark']\n",
    "\n",
    "#mean and median data\n",
    "mean_data=[light_screentime_mean,medium_screentime_mean,dark_screentime_mean]\n",
    "median_data=[light_screentime_median,medium_screentime_median,dark_screentime_median]\n",
    "print(median_data)\n",
    "print(mean_data)\n",
    "plot_skintone_screentime(mean_data,keys,metric=\"average\",fname='screentime_x_skintone_binned.eps')\n",
    "#plot_skintone_screentime(median_data,keys,metric=\"median\")\n",
    "fig, ax = plt.subplots()\n",
    "skintone_screentime_list=[[x*100 for x in list(rp_data['light_screentime'])],\n",
    "                          [x*100 for x in list(rp_data['medium_screentime'])],\n",
    "                          [x*100 for x in list(rp_data['dark_screentime'])]]\n",
    "\n",
    "ax.boxplot(skintone_screentime_list, labels=['light','medium','dark'],\n",
    "          patch_artist=True,\n",
    "          boxprops={'linewidth': 2},\n",
    "          whiskerprops={'linewidth': 2},\n",
    "          medianprops={'linewidth': 2, 'color': 'red'})\n",
    "\n",
    "# set the title and axis labels\n",
    "ax.set_title('Screen time vs Skintone',fontsize=16)\n",
    "ax.set_xlabel('Skintone',fontsize=12)\n",
    "ax.set_ylabel('% Screentime', fontsize=12)\n",
    "\n",
    "# get the x-tick labels\n",
    "labels = ax.get_xticklabels()\n",
    "yticks = ax.get_yticks()\n",
    "ax.set_yticks(yticks)\n",
    "ax.tick_params(axis='y', labelsize=12)\n",
    "ax.set_ylim(-2,110)\n",
    "\n",
    "# set the rotation angle for the labels\n",
    "ax.set_xticklabels(labels, rotation=0, fontsize=12)\n",
    "plt.tight_layout()\n",
    "plt.savefig('plots/screentime_x_skintone_binned_boxplot.eps', bbox_inches=\"tight\", dpi=300)\n",
    "\n",
    "\n",
    "skintone_data = np.array(rp_data[skintone_strings])\n",
    "y = skintone_data/skintone_data.sum(axis=1, keepdims=True)\n",
    "y = np.mean(y*100, axis=0)\n",
    "x = np.arange(1, 11)\n",
    "plt.figure(figsize=(22, 10))\n",
    "bc = plt.bar(x, y, width=0.6, color=[\"#f6ede4\",\"#f3e7db\",\"#f7ead0\",\"#eadaba\",\"#d7bd96\",\"#a07e56\",\"#825c43\",\"#604134\",\"#3a312a\",\"#292420\"])\n",
    "plt.bar_label(bc, [f\"{yy:.1f}%\" for yy in y], padding=10, fontsize=16)\n",
    "plt.title(\"Screentime distribution of MST skintone bins\", fontsize=20)\n",
    "plt.xlabel(\"MST skintone bins\", fontsize=20)\n",
    "plt.ylabel(\"% Screentime\", fontsize=20)\n",
    "plt.xticks(x, x, fontsize=16)\n",
    "plt.ylim(0, 100)\n",
    "plt.tick_params(axis='y', labelsize=16)\n",
    "#plt.legend([bc.patches[0], bc.patches[3], bc.patches[7]], [\"light\", \"medium\", \"dark\"], fontsize=16)\n",
    "plt.savefig(\"plots/screentime_x_skintone.eps\", bbox_inches=\"tight\", pad_inches=0, dpi=300)\n",
    "\n",
    "\n",
    "### BY LANGUAGE (stacked)\n",
    "skintone_groups = ['light', 'medium', 'dark']\n",
    "width = 0.25  # the width of the bars\n",
    "lng_data = rp_data.groupby(\"Programme Language\")\n",
    "x = np.arange(len(lng_data))  # the label locations\n",
    "fig, ax = plt.subplots()\n",
    "bottom = [0, 0, 0, 0, 0]\n",
    "gxst = {k:[] for k in skintone_groups}\n",
    "mul = 0\n",
    "\n",
    "for lang in lng_data.groups.keys():\n",
    "    for st in skintone_groups:\n",
    "        d = lng_data.get_group(lang)\n",
    "        asc = d[st + '_screentime']\n",
    "        gxst[st].append(round(np.mean(asc)*100,2))\n",
    "\n",
    "for st in skintone_groups:\n",
    "    #rint(age, dict_age_mapping[age])\n",
    "    #rects = ax.bar(x, gxst[st], width, label=st, bottom=bottom, edgecolor='black',linewidth=2)\n",
    "    offset = width * mul\n",
    "    rects = ax.bar(x + offset, gxst[st], width, label=st, edgecolor='black',linewidth=2)\n",
    "    #if st!='medium':\n",
    "    ax.bar_label(rects, padding=3, fontsize=15 )\n",
    "    #else:\n",
    "    #  ax.bar_label(rects, padding=-18, fontsize=15, )\n",
    "    #bottom = [bottom[i] + gxst[st][i] for i in range(5)]\n",
    "    mul += 1\n",
    "\n",
    "\n",
    "ax.set_xlabel('Language', fontsize=16)\n",
    "ax.set_ylabel('% Screentime',fontsize=16)\n",
    "ax.set_title('Screentime x Skintone (By Language)', fontsize=20)\n",
    "ax.set_xticks(x+width, list(lng_data.groups.keys()))\n",
    "\n",
    "ax.legend(bbox_to_anchor=(0.86, 1.06), ncol=3,fontsize=12, loc='upper center')\n",
    "ax.tick_params(axis='y', labelsize=16)\n",
    "ax.tick_params(axis='x', labelsize=16)\n",
    "\n",
    "filename=\"screentime_x_skintone_x_language.eps\"\n",
    "plt.savefig('plots/'+filename, bbox_inches=\"tight\", dpi=300)\n",
    "\n",
    "\n",
    "combns = list(combinations(lng_data.groups.keys(), 2))\n",
    "num_c = len(combns)\n",
    "\n",
    "st='medium'\n",
    "print(f't-tests for {st} skintone')\n",
    "for comb in combns:\n",
    "    t, p = ttest_ind(lng_data.get_group(comb[0])[st+'_screentime'], lng_data.get_group(comb[1])[st+'_screentime'])\n",
    "    print(f\"{comb[0]} and {comb[1]}, pvalue: {p/num_c}\")\n",
    "\n",
    "st='dark'\n",
    "print(f't-tests for {st} skintone')\n",
    "for comb in combns:\n",
    "    t, p = ttest_ind(lng_data.get_group(comb[0])[st+'_screentime'], lng_data.get_group(comb[1])[st+'_screentime'])\n",
    "    print(f\"{comb[0]} and {comb[1]}, pvalue: {p/num_c}\")\n",
    "\n",
    "### BY LANGUAGE and GENDER (stacked + grouped)\n",
    "\n",
    "for st in skintone_groups:\n",
    "    male_cols, female_cols = [], []\n",
    "    for age in age_bins:\n",
    "        for intr in mst_skintone_dict[st]:\n",
    "            male_cols.append(f\"(male, {age}, {intr})\")\n",
    "            female_cols.append(f\"(female, {age}, {intr})\")\n",
    "    male_data[st + '_screentime'] = rp_data[[col for col in male_cols]].sum(1)\n",
    "    female_data[st + '_screentime'] = rp_data[[col for col in female_cols]].sum(1)\n",
    "\n",
    "    \n",
    "gxst = {k:{'male':[], 'female':[]} for k in skintone_groups}\n",
    "\n",
    "\n",
    "for st in skintone_groups:\n",
    "    for lang in lng_data.groups.keys():\n",
    "        dm = male_data.groupby(\"Programme Language\").get_group(lang)\n",
    "        df = female_data.groupby(\"Programme Language\").get_group(lang)\n",
    "        msc = dm[st + '_screentime']\n",
    "        fsc = df[st + '_screentime']\n",
    "        mfrac = msc/(msc+fsc)\n",
    "        ffrac = fsc/(msc+fsc)\n",
    "        gxst[st]['male'].append(round(np.mean(mfrac)*100,2))\n",
    "        gxst[st]['female'].append(round(np.mean(ffrac)*100, 2))\n",
    "\n",
    "#fig, axs = plt.subplots(1,4, figsize = (32, 4))\n",
    "x = np.arange(3)\n",
    "width = 0.25\n",
    "\n",
    "plt.figure(figsize=(30,20))\n",
    "ax = []\n",
    "ax.append(plt.subplot2grid(shape=(2,6), loc=(0,0), colspan=2))\n",
    "ax.append(plt.subplot2grid((2,6), (0,2), colspan=2))\n",
    "ax.append(plt.subplot2grid((2,6), (0,4), colspan=2))\n",
    "ax.append(plt.subplot2grid((2,6), (1,1), colspan=2))\n",
    "ax.append(plt.subplot2grid((2,6), (1,3), colspan=2))\n",
    "\n",
    "for i, lang in enumerate(lng_data.groups.keys()):\n",
    "    mul=0\n",
    "    for gen in ['male', 'female']:\n",
    "        plot_data = [gxst['light'][gen][i], gxst['medium'][gen][i], gxst['dark'][gen][i]]\n",
    "        offset = width * mul\n",
    "        rects = ax[i].bar(x + offset, plot_data, width, label=gen, edgecolor='black',linewidth=2)\n",
    "        ax[i].bar_label(rects, padding=3, fontsize=15 )\n",
    "        mul += 1\n",
    " \n",
    "\n",
    "    #ax[i].set_xlabel('Skintone', fontsize=16)\n",
    "    ax[i].set_title(f'{lang}', fontsize=20)\n",
    "    ax[i].set_ylabel('% Screentime',fontsize=16)\n",
    "    ax[i].set_xticks(x+0.5*width, skintone_groups)\n",
    "    ax[i].tick_params(axis='y', labelsize=16)\n",
    "    ax[i].tick_params(axis='x', labelsize=16)\n",
    "plt.figlegend(['male', 'female'], loc=(0.85, 0.25), fontsize='xx-large')\n",
    "#ax.set_suptitle('Screentime x Skintone x Gender (By Language)', fontsize=20)\n",
    "filename=\"screentime_x_skintone_x_language_x_gender.eps\"\n",
    "filename=\"screentime_x_skintone_x_language_x_gender.png\"\n",
    "#plt.savefig('plots/'+filename, bbox_inches=\"tight\", dpi=300)\n",
    "\n",
    "                    \n",
    "# for st in skintone_groups:\n",
    "#     bottom = [0 for _ in range(len(lng_data.groups.keys()))]\n",
    "#     for i, gen in enumerate(['male', 'female']):\n",
    "#         print(x, offset, bottom)\n",
    "#         print(gxst[st][gen])\n",
    "#         #rint(age, dict_age_mapping[age])\n",
    "#         #rects = ax.bar(x, gxst[st], width, label=st, bottom=bottom, edgecolor='black',linewidth=2)\n",
    "#         offset = width * mul\n",
    "#         rects = ax.bar(x + offset, gxst[st][gen], width, label=st, bottom=bottom, edgecolor='black',linewidth=2)\n",
    "#         #if st!='medium':\n",
    "#         ax.bar_label(rects, padding=3, fontsize=15 )\n",
    "#         #else:\n",
    "#         #  ax.bar_label(rects, padding=-18, fontsize=15, )\n",
    "#         bottom = [bottom[i] + gxst[st][gen][i] for i in range(5)]\n",
    "#         #bottom = [bottom[i] + gxst[st][i] for i in range(5)]\n",
    "#         if gen == 'female':\n",
    "#             mul += 1\n",
    "    \n",
    "gxst_male = {k: [] for k in skintone_groups}\n",
    "gxst_female = {k: [] for k in skintone_groups}\n",
    "\n",
    "for lang in lng_data.groups.keys():\n",
    "    dm = male_data.groupby(\"Programme Language\").get_group(lang)\n",
    "    df = female_data.groupby(\"Programme Language\").get_group(lang)\n",
    "    \n",
    "    for st in skintone_groups:        \n",
    "        msc = dm[st + '_screentime']\n",
    "        fsc = df[st + '_screentime']\n",
    "        mfrac = msc/(dm['light_screentime'] + dm['medium_screentime'] + dm['dark_screentime'])\n",
    "        ffrac = fsc/(df['light_screentime'] + df['medium_screentime'] + df['dark_screentime'])\n",
    "        #mfrac = msc/(msc+fsc)\n",
    "        #ffrac = fsc/(msc+fsc)\n",
    "        gxst_male[st].append(round(np.mean(mfrac)*100,2))\n",
    "        gxst_female[st].append(round(np.mean(ffrac)*100, 2))\n",
    "print(gxst_male[st])\n",
    "\n",
    "    \n",
    "x = np.arange(2)\n",
    "width = 0.25\n",
    "\n",
    "plt.figure(figsize=(30,20))\n",
    "ax = []\n",
    "ax.append(plt.subplot2grid(shape=(2,6), loc=(0,0), colspan=2))\n",
    "ax.append(plt.subplot2grid((2,6), (0,2), colspan=2))\n",
    "ax.append(plt.subplot2grid((2,6), (0,4), colspan=2))\n",
    "ax.append(plt.subplot2grid((2,6), (1,1), colspan=2))\n",
    "ax.append(plt.subplot2grid((2,6), (1,3), colspan=2))\n",
    "\n",
    "for i, lang in enumerate(lng_data.groups.keys()):\n",
    "    mul=0\n",
    "    for st in skintone_groups:\n",
    "        plot_data = [gxst_male[st][i], gxst_female[st][i]]\n",
    "        offset = width * mul\n",
    "        rects = ax[i].bar(x + offset, plot_data, width, label=st, edgecolor='black',linewidth=2)\n",
    "        ax[i].bar_label(rects, padding=3, fontsize=15 )\n",
    "        mul += 1\n",
    " \n",
    "\n",
    "    #ax[i].set_xlabel('Skintone', fontsize=16)\n",
    "    ax[i].set_title(f'{lang}', fontsize=20)\n",
    "    ax[i].set_ylabel('% Screentime',fontsize=16)\n",
    "    ax[i].set_xticks(x+width, ['male', 'female'])\n",
    "    ax[i].tick_params(axis='y', labelsize=16)\n",
    "    ax[i].tick_params(axis='x', labelsize=16)\n",
    "plt.figlegend(skintone_groups, loc=(0.85, 0.25), fontsize='xx-large')\n",
    "#ax.set_suptitle('Screentime x Skintone x Gender (By Language)', fontsize=20)\n",
    "filename=\"screentime_x_gender_x_skintone_x_language.eps\"\n",
    "#filename=\"screentime_x_skintone_x_language_x_gender.png\"\n",
    "plt.savefig('plots/'+filename, bbox_inches=\"tight\", dpi=300)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "print(male_data['dark_screentime'].sum())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "sJCv9GWE4U9-"
   },
   "source": [
    "## IV. Screentime x Gender x Age"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "age =  '[0, 18)'\n",
    "male_cols = [col for col in male_data.columns if age in col]\n",
    "print(age, male_cols)\n",
    "print(male_data[male_cols].sum(1))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 576
    },
    "id": "NmiT232a4Qzg",
    "outputId": "78919904-7a1e-4117-b35e-a3801ff3a38a"
   },
   "outputs": [],
   "source": [
    "\n",
    "male_data_mean=[]\n",
    "female_data_mean=[]\n",
    "age_bins=['[0, 18)','[18, 33)','[33, 60)','[60, inf)']\n",
    "\n",
    "for age in age_bins:\n",
    "    age_col=[col for col in male_data.columns if age in col]\n",
    "    percent_val=(male_data[age_col].sum(1)/male_data['masculine_faces']).mean()\n",
    "    male_data_mean.append(percent_val*100)\n",
    "\n",
    "for age in age_bins:\n",
    "    age_col=[col for col in female_data.columns if age in col]\n",
    "    percent_val=(female_data[age_col].sum(1)/female_data['feminine_faces']).mean()\n",
    "    female_data_mean.append(percent_val*100)\n",
    "#plotting the male vs female grouped bar plot\n",
    "plt.rcParams[\"figure.figsize\"] = (15,10)\n",
    "index=['male','female']\n",
    "dict_age_sample={}\n",
    "\n",
    "for i,v in enumerate(age_bins):\n",
    "    dict_age_sample[v]=[round(male_data_mean[i],2),round(female_data_mean[i],2)]\n",
    "dict_age_sample\n",
    "gender=('Male', 'Female')\n",
    "x = np.arange(len(gender))  # the label locations\n",
    "width = 0.15  # the width of the bars\n",
    "multiplier = 0\n",
    "\n",
    "fig, ax = plt.subplots(layout='constrained')\n",
    "\n",
    "for attribute, measurement in dict_age_sample.items():\n",
    "    offset = width * multiplier\n",
    "    rects = ax.bar(x + offset, measurement, width, label=attribute, edgecolor='black',linewidth=2)\n",
    "    ax.bar_label(rects, padding=3, fontsize=15, )\n",
    "    multiplier += 1\n",
    "\n",
    "ax.set_xlabel('Gender', fontsize=20)\n",
    "ax.set_ylabel('% Screentime',fontsize=20)\n",
    "ax.set_title('Screentime x Gender x Age', fontsize=20)\n",
    "ax.set_xticks(x + 0.23, gender)\n",
    "\n",
    "ax.legend(loc='center', bbox_to_anchor=(0.5, 0.8), ncol=3,fontsize=16)\n",
    "ax.tick_params(axis='y', labelsize=16)\n",
    "ax.tick_params(axis='x', labelsize=16)\n",
    "\n",
    "filename=\"screentime_x_gender_x_age.eps\"\n",
    "plt.savefig('plots/'+filename, bbox_inches=\"tight\", dpi=300)\n",
    "filename=\"screentime_x_gender_x_age.png\"\n",
    "plt.savefig('plots/'+filename, bbox_inches=\"tight\", dpi=300)\n",
    "\n",
    "### Transposing the axes\n",
    "plot_data = {'male':[], 'female':[]}\n",
    "for age in age_bins:\n",
    "    male_cols = [col for col in male_data.columns if age in col]\n",
    "    female_cols = [col for col in female_data.columns if age in col]\n",
    "   # print(age, male_cols, female_cols)\n",
    "    male_faces = male_data[male_cols].sum(1)\n",
    "    female_faces = female_data[female_cols].sum(1)\n",
    "    frac_male = (male_faces/(male_faces+female_faces)).mean()*100\n",
    "    frac_female = (female_faces/(male_faces+female_faces)).mean()*100\n",
    "    plot_data['male'].append(round(frac_male,2))\n",
    "    plot_data['female'].append(round(frac_female,2))\n",
    "    \n",
    "    #age_data_mean[age] = [frac_male, frac_female]\n",
    "\n",
    "x = np.arange(len(age_bins))  # the label locations\n",
    "width = 0.25  # the width of the bars\n",
    "multiplier = 0\n",
    "\n",
    "fig, ax = plt.subplots(layout='constrained')\n",
    "\n",
    "for attribute, measurement in plot_data.items():\n",
    "    offset = width * multiplier\n",
    "    rects = ax.bar(x + offset, measurement, width, label=attribute, edgecolor='black',linewidth=2)\n",
    "    ax.bar_label(rects, padding=3, fontsize=15, )\n",
    "    multiplier += 1\n",
    "\n",
    "ax.set_xlabel('Age', fontsize=20)\n",
    "ax.set_ylabel('% Screentime',fontsize=20)\n",
    "ax.set_title('Screentime x Age x Gender', fontsize=20)\n",
    "ax.set_xticks(x + 0.5*width, age_bins)\n",
    "\n",
    "ax.legend(loc='center', bbox_to_anchor=(0.85, 0.95), ncol=3,fontsize=16)\n",
    "ax.tick_params(axis='y', labelsize=16)\n",
    "ax.tick_params(axis='x', labelsize=16)\n",
    "\n",
    "filename=\"screentime_x_age_x_gender.eps\"\n",
    "plt.savefig('plots/'+filename, bbox_inches=\"tight\", dpi=300)\n",
    "filename=\"screentime_x_age_x_gender.png\"\n",
    "plt.savefig('plots/'+filename, bbox_inches=\"tight\", dpi=300)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "jV22e1K6714s"
   },
   "source": [
    "## V. Screentime x Gender x Skintone"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 576
    },
    "id": "ijctiTGK5p8B",
    "outputId": "f25c050b-30da-4986-ea4a-2e4b23fc72b5"
   },
   "outputs": [],
   "source": [
    "for skintone in skintone_strings:\n",
    "    if skintone not in male_data:\n",
    "      sk_male_col=[col for col in male_data.columns if skintone in col]\n",
    "      sk_male_data=male_data[sk_male_col]\n",
    "      male_data[skintone]=sk_male_data.sum(axis=1)\n",
    "    if skintone not in female_data:\n",
    "      sk_female_col=[col for col in female_data.columns if skintone in col]\n",
    "      sk_female_data=female_data[sk_female_col]\n",
    "      female_data[skintone]=sk_female_data.sum(axis=1)\n",
    "\n",
    "male_data['light']=male_data['[-inf, 1.1)']+male_data['[1.1, 2.1)']+male_data['[2.1, 3.1)']\n",
    "male_data['medium']=male_data['[3.1, 4.1)']+male_data['[4.1, 5.1)']+male_data['[5.1, 6.1)']+male_data['[6.1, 7.1)']\n",
    "male_data['dark']=male_data['[7.1, 8.1)']+male_data['[8.1, 9.1)']+male_data['[9.1, 10.1)']\n",
    "\n",
    "female_data['light']=female_data['[-inf, 1.1)']+female_data['[1.1, 2.1)']+female_data['[2.1, 3.1)']\n",
    "female_data['medium']=female_data['[3.1, 4.1)']+female_data['[4.1, 5.1)']+female_data['[5.1, 6.1)']+female_data['[6.1, 7.1)']\n",
    "female_data['dark']=female_data['[7.1, 8.1)']+female_data['[8.1, 9.1)']+female_data['[9.1, 10.1)']\n",
    "skintone_bins=['light','medium','dark']\n",
    "male_data_skintone_mean=[]\n",
    "female_data_skintone_mean=[]\n",
    "for sk_bin in skintone_bins:\n",
    "    mt = male_data[sk_bin]/male_data['masculine_faces']\n",
    "    ft = female_data[sk_bin]/female_data['feminine_faces']\n",
    "    print(sk_bin, ttest_ind(mt, ft).pvalue)\n",
    "    male_sk_mean=((male_data[sk_bin]).sum())/(male_data['masculine_faces'].sum())\n",
    "    male_data_skintone_mean.append(male_sk_mean*100)\n",
    "\n",
    "    female_sk_mean=((female_data[sk_bin]).sum())/(female_data['feminine_faces'].sum())\n",
    "    female_data_skintone_mean.append(female_sk_mean*100)\n",
    "plt.rcParams[\"figure.figsize\"] = (15,10)\n",
    "sns.set_style('darkgrid')\n",
    "index=['male','female']\n",
    "x = np.arange(len(index))\n",
    "dict_sk_sample={}\n",
    "width=0.25\n",
    "multiplier=0\n",
    "\n",
    "for i,v in enumerate(skintone_bins):\n",
    "    dict_sk_sample[v]=[male_data_skintone_mean[i],female_data_skintone_mean[i]]\n",
    "\n",
    "fig, ax = plt.subplots(layout='constrained')\n",
    "\n",
    "for attribute, measurement in dict_sk_sample.items():\n",
    "    offset = width * multiplier\n",
    "    rects = ax.bar(x + offset, measurement, width, label=attribute, edgecolor='black',linewidth=2)\n",
    "    ax.bar_label(rects, padding=3, fontsize=15, )\n",
    "    multiplier += 1\n",
    "\n",
    "ax.set_xlabel('Gender', fontsize=20)\n",
    "ax.set_ylabel('Screentime',fontsize=20)\n",
    "ax.set_title('Skintone based screentime vs gender', fontsize=20)\n",
    "ax.set_xticks(x + width, index)\n",
    "ax.legend(loc='center', bbox_to_anchor=(0.38, 0.8), ncol=3,fontsize=16)\n",
    "ax.tick_params(axis='y', labelsize=20)\n",
    "ax.tick_params(axis='x', labelsize=20)\n",
    "\n",
    "filename=\"screentime_x_gender_x_skintone.eps\"\n",
    "plt.savefig('plots/'+filename,bbox_inches=\"tight\", dpi=300)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "age_st_data = {k: {} for k in age_bins}\n",
    "rp_data = rp_data.fillna(0)\n",
    "for age_grp in age_bins:\n",
    "    age_st_data[age_grp]['light'] = rp_data[f'(male, {age_grp}, [-inf, 1.1))'] + rp_data[f'(male, {age_grp}, [1.1, 2.1))'] + rp_data[f'(male, {age_grp}, [2.1, 3.1))'] + rp_data[f'(female, {age_grp}, [-inf, 1.1))'] + rp_data[f'(female, {age_grp}, [1.1, 2.1))'] + rp_data[f'(female, {age_grp}, [2.1, 3.1))']\n",
    "    age_st_data[age_grp]['medium'] = rp_data[f'(male, {age_grp}, [3.1, 4.1))'] + rp_data[f'(male, {age_grp}, [4.1, 5.1))'] + rp_data[f'(male, {age_grp}, [5.1, 6.1))'] + rp_data[f'(male, {age_grp}, [6.1, 7.1))'] + rp_data[f'(female, {age_grp}, [3.1, 4.1))'] + rp_data[f'(female, {age_grp}, [4.1, 5.1))'] + rp_data[f'(female, {age_grp}, [5.1, 6.1))'] + rp_data[f'(female, {age_grp}, [6.1, 7.1))']\n",
    "    age_st_data[age_grp]['dark'] = rp_data[f'(male, {age_grp}, [7.1, 8.1))'] + rp_data[f'(male, {age_grp}, [8.1, 9.1))'] + rp_data[f'(male, {age_grp}, [9.1, 10.1))'] + rp_data[f'(female, {age_grp}, [7.1, 8.1))'] + rp_data[f'(female, {age_grp}, [8.1, 9.1))'] + rp_data[f'(female, {age_grp}, [9.1, 10.1))']\n",
    "\n",
    "age_st_mean = {k:[] for k in skintone_bins}\n",
    "\n",
    "for st in skintone_bins:\n",
    "    for age_grp in age_bins:\n",
    "        grpmean = age_st_data[age_grp][st]/(age_st_data[age_grp]['light'] + age_st_data[age_grp]['medium'] + age_st_data[age_grp]['dark'])*100\n",
    "        age_st_mean[st].append(round(grpmean.mean(), 2))\n",
    "\n",
    "for age_grp in age_bins:\n",
    "    print(age_grp,sum([(age_st_data[age_grp][x]!=0).sum() for x in skintone_bins]))\n",
    "    \n",
    "plt.rcParams[\"figure.figsize\"] = (25,10)\n",
    "sns.set_style('darkgrid')\n",
    "x = np.arange(len(age_bins))\n",
    "width=0.25\n",
    "multiplier=0\n",
    "\n",
    "fig, ax = plt.subplots(layout='constrained')\n",
    "for attribute, measurement in age_st_mean.items():\n",
    "    print(x)\n",
    "    print(measurement)\n",
    "    offset = width * multiplier\n",
    "    rects = ax.bar(x + offset, measurement, width, label=attribute, edgecolor='black',linewidth=2)\n",
    "    ax.bar_label(rects, padding=3, fontsize=15, )\n",
    "    multiplier += 1\n",
    "\n",
    "ax.set_xlabel('Age', fontsize=20)\n",
    "ax.set_ylabel('Screentime',fontsize=20)\n",
    "ax.set_title('Skintone based screentime vs age', fontsize=20)\n",
    "ax.set_xticks(x + width, age_bins)\n",
    "ax.legend(loc='center', bbox_to_anchor=(0.8, 0.95), ncol=3,fontsize=16)\n",
    "ax.tick_params(axis='y', labelsize=20)\n",
    "ax.tick_params(axis='x', labelsize=20)\n",
    "\n",
    "filename=\"screentime_x_age_x_skintone.eps\"\n",
    "plt.savefig('plots/'+filename,bbox_inches=\"tight\", dpi=300)\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 974
    },
    "id": "lNolYKWVfDa-",
    "outputId": "72067861-b64b-4a49-f125-5eca02b16a62"
   },
   "outputs": [],
   "source": [
    "idm = male_data[[x for x in male_data.columns if '(male' in x]]\n",
    "age_bins=['[60, inf)','[33, 60)','[18, 33)','[0, 18)']\n",
    "\n",
    "### ALL skintones\n",
    "hmdata = np.zeros((len(age_bins), len(skintone_strings)), dtype=float)\n",
    "tot_faces = idm.sum(axis=1)\n",
    "\n",
    "for i, age in enumerate(age_bins):\n",
    "  for j, st in enumerate(skintone_strings):\n",
    "    fraction_faces = idm[f'(male, {age}, {st})']/tot_faces\n",
    "    hmdata[i,j] = np.mean(fraction_faces)*100\n",
    "\n",
    "cmap = matplotlib.colormaps[\"Reds\"]\n",
    "#cmap = matplotlib.colors.ListedColormap(cmap)\n",
    "#cmap.set_under(\"none\")\n",
    "height = len(age_bins) * 1.2\n",
    "width = len(skintone_strings)\n",
    "plt.figure(figsize=(width, height))\n",
    "ax = plt.gca()\n",
    "sns.heatmap(hmdata, xticklabels=skintone_strings, yticklabels=age_bins, annot=True, fmt=\".2f\",\n",
    "                    linewidth=1, cmap=cmap, annot_kws={\"fontsize\":\"small\"}, cbar=True, vmin=0, vmax=100,\n",
    "                    linecolor=\"black\", ax=ax)\n",
    "ax.set_xticklabels(skintone_strings, fontsize=\"x-small\", rotation=45)\n",
    "ax.set_yticklabels(age_bins, fontsize=\"small\")\n",
    "cbar = ax.collections[0].colorbar\n",
    "cbar.ax.tick_params(labelsize=\"small\")\n",
    "ax.set_xlabel('Skintone', fontsize=12)\n",
    "ax.set_ylabel('Age', fontsize=12)\n",
    "ax.set_title(f\"Male screentime x Age x Skintone\", fontsize=\"large\")\n",
    "figure = ax.get_figure()\n",
    "figure.savefig('plots/screentime_x_age_x_skintone_x_male.eps', dpi=300, bbox_inches=\"tight\")\n",
    "\n",
    "### BINNED skintone\n",
    "\n",
    "st_bins = ['light', 'medium', 'dark']\n",
    "hmdata_bin = np.zeros((len(age_bins), len(st_bins)), dtype=float)\n",
    "\n",
    "for age in age_bins:\n",
    "  for stkey, stval in mst_skintone_dict.items():\n",
    "    key = f'(male, {age}, {stkey})'\n",
    "    binkeys = [f'(male, {age}, {v})' for v in stval]\n",
    "    idm_sub = idm[binkeys]\n",
    "    d = idm_sub.sum(axis=1)\n",
    "    idm.loc[:,key] = d\n",
    "\n",
    "for i, age in enumerate(age_bins):\n",
    "  for j, st in enumerate(st_bins):\n",
    "    fraction_faces = idm[f'(male, {age}, {st})']/tot_faces\n",
    "    hmdata_bin[i,j] = np.mean(fraction_faces)*100\n",
    "\n",
    "height = len(age_bins) * 1.2\n",
    "width = len(st_bins)\n",
    "plt.figure(figsize=(width, height))\n",
    "ax = plt.gca()\n",
    "sns.heatmap(hmdata_bin, xticklabels=st_bins, yticklabels=age_bins, annot=True, fmt=\".2f\",\n",
    "                    linewidth=1, cmap=cmap, annot_kws={\"fontsize\":\"small\"}, cbar=True, vmin=0, vmax=100,\n",
    "                    linecolor=\"black\", ax=ax)\n",
    "ax.set_xticklabels(st_bins, fontsize=\"x-small\", rotation=45)\n",
    "ax.set_yticklabels(age_bins, fontsize=\"small\")\n",
    "cbar = ax.collections[0].colorbar\n",
    "cbar.ax.tick_params(labelsize=\"small\")\n",
    "ax.set_xlabel('Skintone', fontsize=12)\n",
    "ax.set_ylabel('Age', fontsize=12)\n",
    "ax.set_title(f\"Male screentime x Age x Skintone binned\", fontsize=\"large\")\n",
    "figure = ax.get_figure()\n",
    "figure.savefig('plots/screentime_x_age_x_skintone-binned_x_male.eps', dpi=300, bbox_inches=\"tight\")\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 974
    },
    "id": "Fo2ql9-yocP1",
    "outputId": "d2ebc279-b26e-4159-c82d-bb056a924033"
   },
   "outputs": [],
   "source": [
    "idm = female_data[[x for x in female_data.columns if '(female' in x]]\n",
    "\n",
    "### ALL skintones\n",
    "hmdata = np.zeros((len(age_bins), len(skintone_strings)), dtype=float)\n",
    "tot_faces = idm.sum(axis=1)\n",
    "\n",
    "for i, age in enumerate(age_bins):\n",
    "  for j, st in enumerate(skintone_strings):\n",
    "    fraction_faces = idm[f'(female, {age}, {st})']/tot_faces\n",
    "    hmdata[i,j] = np.mean(fraction_faces)*100\n",
    "\n",
    "cmap = matplotlib.colormaps[\"Reds\"]\n",
    "#cmap = matplotlib.colors.ListedColormap(cmap)\n",
    "#cmap.set_under(\"none\")\n",
    "height = len(age_bins) * 1.2\n",
    "width = len(skintone_strings)\n",
    "plt.figure(figsize=(width, height))\n",
    "ax = plt.gca()\n",
    "sns.heatmap(hmdata, xticklabels=skintone_strings, yticklabels=age_bins, annot=True, fmt=\".2f\",\n",
    "                    linewidth=1, cmap=cmap, annot_kws={\"fontsize\":\"small\"}, cbar=True, vmin=0, vmax=100,\n",
    "                    linecolor=\"black\", ax=ax)\n",
    "ax.set_xticklabels(skintone_strings, fontsize=\"x-small\", rotation=45)\n",
    "ax.set_yticklabels(age_bins, fontsize=\"small\")\n",
    "cbar = ax.collections[0].colorbar\n",
    "cbar.ax.tick_params(labelsize=\"small\")\n",
    "ax.set_xlabel('Skintone', fontsize=12)\n",
    "ax.set_ylabel('Age', fontsize=12)\n",
    "ax.set_title(f\"Female screentime x Age x Skintone\", fontsize=\"large\")\n",
    "figure = ax.get_figure()\n",
    "figure.savefig('plots/screentime_x_age_x_skintone_x_female.eps', dpi=300, bbox_inches=\"tight\")\n",
    "\n",
    "### BINNED skintone\n",
    "\n",
    "st_bins = ['light', 'medium', 'dark']\n",
    "hmdata_bin = np.zeros((len(age_bins), len(st_bins)), dtype=float)\n",
    "\n",
    "for age in age_bins:\n",
    "  for stkey, stval in mst_skintone_dict.items():\n",
    "    key = f'(female, {age}, {stkey})'\n",
    "    binkeys = [f'(female, {age}, {v})' for v in stval]\n",
    "    idm_sub = idm[binkeys]\n",
    "    d = idm_sub.sum(axis=1)\n",
    "    idm.loc[:,key] = d\n",
    "\n",
    "for i, age in enumerate(age_bins):\n",
    "  for j, st in enumerate(st_bins):\n",
    "    fraction_faces = idm[f'(female, {age}, {st})']/tot_faces\n",
    "    hmdata_bin[i,j] = np.mean(fraction_faces)*100\n",
    "\n",
    "height = len(age_bins) * 1.2\n",
    "width = len(st_bins)\n",
    "plt.figure(figsize=(width, height))\n",
    "ax = plt.gca()\n",
    "sns.heatmap(hmdata_bin, xticklabels=st_bins, yticklabels=age_bins, annot=True, fmt=\".2f\",\n",
    "                    linewidth=1, cmap=cmap, annot_kws={\"fontsize\":\"small\"}, cbar=True, vmin=0, vmax=100,\n",
    "                    linecolor=\"black\", ax=ax)\n",
    "ax.set_xticklabels(st_bins, fontsize=\"x-small\", rotation=45)\n",
    "ax.set_yticklabels(age_bins, fontsize=\"small\")\n",
    "cbar = ax.collections[0].colorbar\n",
    "cbar.ax.tick_params(labelsize=\"small\")\n",
    "ax.set_xlabel('Skintone', fontsize=12)\n",
    "ax.set_ylabel('Age', fontsize=12)\n",
    "ax.set_title(f\"Female screentime x Age x Skintone binned\", fontsize=\"large\")\n",
    "figure = ax.get_figure()\n",
    "figure.savefig('plots/screentime_x_age_x_skintone-binned_x_female.eps', dpi=300, bbox_inches=\"tight\")\n",
    "\n"
   ]
  }
 ],
 "metadata": {
  "colab": {
   "provenance": []
  },
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}