2results_across_SSRTscales.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "\n",
    "import json\n",
    "\n",
    "from scripts.utils import SimulateData\n",
    "from stopsignalmetrics.ssrtmodel import SSRTmodel\n",
    "\n",
    "from matplotlib.font_manager import FontProperties\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "import dataframe_image as dfi"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "SSRTscales = [85, 25, 5, 0]\n",
    "\n",
    "SSRT_method_map = {'standard': 'Weighted', 'fixed': 'Fixed', 'tracking': 'Tracking'}\n",
    "gen_map = {'gen-graded_both': 'Confusion',\n",
    "           'gen-graded_go': 'Slowed Go Processing',\n",
    "           'gen-guesses': 'Guessing',\n",
    "           'gen-standard': 'Independent Race'}\n",
    "\n",
    "gen_map_simple = {'graded_both': 'Confusion',\n",
    "           'graded_go': 'Slowed Go Processing',\n",
    "           'guesses': 'Guessing',\n",
    "           'standard': 'Independent Race'}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# looking at correlations across generating models x SSRT methods, SSRT scales"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# full_ssrt_df = pd.DataFrame()\n",
    "\n",
    "# for SSRTscale in SSRTscales:\n",
    "#     ssrt_df = pd.read_csv('ssrt_metrics/expected_ssrts_SSRTscale-%d.csv' % SSRTscale, index_col = 0)\n",
    "#     # reformatting - was multiindex\n",
    "#     ssrt_df.columns = [ f'gen-{gen}_SSRT-{ssrt}' for gen, ssrt in zip(ssrt_df.loc['underlying distribution', :].values, ssrt_df.columns)]\n",
    "#     ssrt_df = ssrt_df.drop(['underlying distribution', 'NARGUID'])\n",
    "\n",
    "#     for col in ssrt_df.columns:\n",
    "#         ssrt_df[col] = ssrt_df[col].astype(float)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "def read_in_multiidx_df(ssrtscale):\n",
    "    tst_df = pd.read_csv('ssrt_metrics/expected_ssrts_SSRTscale-%d.csv' % ssrtscale, index_col = 0, header=[0,1])\n",
    "    tst_df = tst_df.drop('guesses', level=0, axis=1)\n",
    "    tst_df = tst_df.drop('graded_go', level=0, axis=1)\n",
    "    tst_df = tst_df.drop('graded_both', level=0, axis=1)\n",
    "    tst_df.columns = tst_df.columns.set_names('SSRT Method', level=0)\n",
    "    tst_df.columns = tst_df.columns.set_names('Generating Model', level=1)\n",
    "    tst_df = tst_df.rename(columns=SSRT_method_map, level=0)\n",
    "    tst_df = tst_df.rename(columns=gen_map_simple, level=1)\n",
    "    return tst_df\n",
    "\n",
    "pd_df = {'SSRTscale=%d' % scale: read_in_multiidx_df(scale) for scale in  SSRTscales}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>SSRT Method</th>\n",
       "      <th>Weighted</th>\n",
       "      <th>Fixed</th>\n",
       "      <th>Tracking</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SSRT Method</th>\n",
       "      <th>Generating Model</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"3\" valign=\"top\">Weighted</th>\n",
       "      <th>Confusion</th>\n",
       "      <td>0.847651</td>\n",
       "      <td>0.823044</td>\n",
       "      <td>0.793582</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Slowed Go Processing</th>\n",
       "      <td>0.932054</td>\n",
       "      <td>0.967091</td>\n",
       "      <td>0.940639</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Guessing</th>\n",
       "      <td>0.882841</td>\n",
       "      <td>0.883268</td>\n",
       "      <td>0.831584</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"3\" valign=\"top\">Fixed</th>\n",
       "      <th>Confusion</th>\n",
       "      <td>0.945386</td>\n",
       "      <td>0.973496</td>\n",
       "      <td>0.927501</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Slowed Go Processing</th>\n",
       "      <td>0.948272</td>\n",
       "      <td>0.975553</td>\n",
       "      <td>0.930655</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Guessing</th>\n",
       "      <td>0.859395</td>\n",
       "      <td>0.906937</td>\n",
       "      <td>0.832101</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"3\" valign=\"top\">Tracking</th>\n",
       "      <th>Confusion</th>\n",
       "      <td>0.987024</td>\n",
       "      <td>0.984714</td>\n",
       "      <td>0.994806</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Slowed Go Processing</th>\n",
       "      <td>0.961292</td>\n",
       "      <td>0.983543</td>\n",
       "      <td>0.948930</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Guessing</th>\n",
       "      <td>0.964556</td>\n",
       "      <td>0.985161</td>\n",
       "      <td>0.950957</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "SSRT Method                       Weighted     Fixed  Tracking\n",
       "SSRT Method Generating Model                                  \n",
       "Weighted    Confusion             0.847651  0.823044  0.793582\n",
       "            Slowed Go Processing  0.932054  0.967091  0.940639\n",
       "            Guessing              0.882841  0.883268  0.831584\n",
       "Fixed       Confusion             0.945386  0.973496  0.927501\n",
       "            Slowed Go Processing  0.948272  0.975553  0.930655\n",
       "            Guessing              0.859395  0.906937  0.832101\n",
       "Tracking    Confusion             0.987024  0.984714  0.994806\n",
       "            Slowed Go Processing  0.961292  0.983543  0.948930\n",
       "            Guessing              0.964556  0.985161  0.950957"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "corr_df = pd_df['SSRTscale=85'].corr(method='spearman')\n",
    "corr_df = corr_df.loc[corr_df.index.get_level_values(1)!='Independent Race', corr_df.columns.get_level_values(1)=='Independent Race']\n",
    "corr_df.droplevel(1, axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>SSRT Method</th>\n",
       "      <th>Weighted</th>\n",
       "      <th>Fixed</th>\n",
       "      <th>Tracking</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SSRT Method</th>\n",
       "      <th>Generating Model</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"3\" valign=\"top\">Weighted</th>\n",
       "      <th>Confusion</th>\n",
       "      <td>0.729580</td>\n",
       "      <td>0.615414</td>\n",
       "      <td>0.475834</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Slowed Go Processing</th>\n",
       "      <td>0.661254</td>\n",
       "      <td>0.849783</td>\n",
       "      <td>0.698311</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Guessing</th>\n",
       "      <td>0.712721</td>\n",
       "      <td>0.675053</td>\n",
       "      <td>0.448605</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"3\" valign=\"top\">Fixed</th>\n",
       "      <th>Confusion</th>\n",
       "      <td>0.742904</td>\n",
       "      <td>0.890478</td>\n",
       "      <td>0.623447</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Slowed Go Processing</th>\n",
       "      <td>0.751064</td>\n",
       "      <td>0.895976</td>\n",
       "      <td>0.634714</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Guessing</th>\n",
       "      <td>0.587196</td>\n",
       "      <td>0.765179</td>\n",
       "      <td>0.429138</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"3\" valign=\"top\">Tracking</th>\n",
       "      <th>Confusion</th>\n",
       "      <td>0.876190</td>\n",
       "      <td>0.838457</td>\n",
       "      <td>0.965085</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Slowed Go Processing</th>\n",
       "      <td>0.779033</td>\n",
       "      <td>0.911575</td>\n",
       "      <td>0.677161</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Guessing</th>\n",
       "      <td>0.836705</td>\n",
       "      <td>0.951849</td>\n",
       "      <td>0.754173</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "SSRT Method                       Weighted     Fixed  Tracking\n",
       "SSRT Method Generating Model                                  \n",
       "Weighted    Confusion             0.729580  0.615414  0.475834\n",
       "            Slowed Go Processing  0.661254  0.849783  0.698311\n",
       "            Guessing              0.712721  0.675053  0.448605\n",
       "Fixed       Confusion             0.742904  0.890478  0.623447\n",
       "            Slowed Go Processing  0.751064  0.895976  0.634714\n",
       "            Guessing              0.587196  0.765179  0.429138\n",
       "Tracking    Confusion             0.876190  0.838457  0.965085\n",
       "            Slowed Go Processing  0.779033  0.911575  0.677161\n",
       "            Guessing              0.836705  0.951849  0.754173"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "corr_df = pd_df['SSRTscale=25'].corr(method='spearman')\n",
    "corr_df = corr_df.loc[corr_df.index.get_level_values(1)!='Independent Race', corr_df.columns.get_level_values(1)=='Independent Race']\n",
    "corr_df.droplevel(1, axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>SSRT Method</th>\n",
       "      <th>Weighted</th>\n",
       "      <th>Fixed</th>\n",
       "      <th>Tracking</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SSRT Method</th>\n",
       "      <th>Generating Model</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"3\" valign=\"top\">Weighted</th>\n",
       "      <th>Confusion</th>\n",
       "      <td>0.820041</td>\n",
       "      <td>0.540094</td>\n",
       "      <td>0.176639</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Slowed Go Processing</th>\n",
       "      <td>0.177755</td>\n",
       "      <td>0.681644</td>\n",
       "      <td>0.247106</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Guessing</th>\n",
       "      <td>0.812881</td>\n",
       "      <td>0.674141</td>\n",
       "      <td>0.194281</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"3\" valign=\"top\">Fixed</th>\n",
       "      <th>Confusion</th>\n",
       "      <td>0.564066</td>\n",
       "      <td>0.892707</td>\n",
       "      <td>0.255062</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Slowed Go Processing</th>\n",
       "      <td>0.563060</td>\n",
       "      <td>0.887390</td>\n",
       "      <td>0.255884</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Guessing</th>\n",
       "      <td>0.541527</td>\n",
       "      <td>0.879659</td>\n",
       "      <td>0.203974</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"3\" valign=\"top\">Tracking</th>\n",
       "      <th>Confusion</th>\n",
       "      <td>0.251913</td>\n",
       "      <td>0.176027</td>\n",
       "      <td>0.553595</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Slowed Go Processing</th>\n",
       "      <td>0.567366</td>\n",
       "      <td>0.886458</td>\n",
       "      <td>0.269588</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Guessing</th>\n",
       "      <td>0.594349</td>\n",
       "      <td>0.902361</td>\n",
       "      <td>0.342064</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "SSRT Method                       Weighted     Fixed  Tracking\n",
       "SSRT Method Generating Model                                  \n",
       "Weighted    Confusion             0.820041  0.540094  0.176639\n",
       "            Slowed Go Processing  0.177755  0.681644  0.247106\n",
       "            Guessing              0.812881  0.674141  0.194281\n",
       "Fixed       Confusion             0.564066  0.892707  0.255062\n",
       "            Slowed Go Processing  0.563060  0.887390  0.255884\n",
       "            Guessing              0.541527  0.879659  0.203974\n",
       "Tracking    Confusion             0.251913  0.176027  0.553595\n",
       "            Slowed Go Processing  0.567366  0.886458  0.269588\n",
       "            Guessing              0.594349  0.902361  0.342064"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "corr_df = pd_df['SSRTscale=5'].corr(method='spearman')\n",
    "corr_df = corr_df.loc[corr_df.index.get_level_values(1)!='Independent Race', corr_df.columns.get_level_values(1)=='Independent Race']\n",
    "corr_df.droplevel(1, axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>SSRT Method</th>\n",
       "      <th>Weighted</th>\n",
       "      <th>Fixed</th>\n",
       "      <th>Tracking</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SSRT Method</th>\n",
       "      <th>Generating Model</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"3\" valign=\"top\">Weighted</th>\n",
       "      <th>Confusion</th>\n",
       "      <td>0.846541</td>\n",
       "      <td>0.519747</td>\n",
       "      <td>0.129184</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Slowed Go Processing</th>\n",
       "      <td>0.092678</td>\n",
       "      <td>0.677320</td>\n",
       "      <td>0.146728</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Guessing</th>\n",
       "      <td>0.847930</td>\n",
       "      <td>0.662002</td>\n",
       "      <td>0.156904</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"3\" valign=\"top\">Fixed</th>\n",
       "      <th>Confusion</th>\n",
       "      <td>0.535177</td>\n",
       "      <td>0.903071</td>\n",
       "      <td>0.183349</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Slowed Go Processing</th>\n",
       "      <td>0.529635</td>\n",
       "      <td>0.895206</td>\n",
       "      <td>0.181814</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Guessing</th>\n",
       "      <td>0.537224</td>\n",
       "      <td>0.913442</td>\n",
       "      <td>0.185171</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"3\" valign=\"top\">Tracking</th>\n",
       "      <th>Confusion</th>\n",
       "      <td>-0.095310</td>\n",
       "      <td>-0.201330</td>\n",
       "      <td>0.021480</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Slowed Go Processing</th>\n",
       "      <td>0.529901</td>\n",
       "      <td>0.896101</td>\n",
       "      <td>0.182643</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Guessing</th>\n",
       "      <td>0.513343</td>\n",
       "      <td>0.879253</td>\n",
       "      <td>0.175802</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "SSRT Method                       Weighted     Fixed  Tracking\n",
       "SSRT Method Generating Model                                  \n",
       "Weighted    Confusion             0.846541  0.519747  0.129184\n",
       "            Slowed Go Processing  0.092678  0.677320  0.146728\n",
       "            Guessing              0.847930  0.662002  0.156904\n",
       "Fixed       Confusion             0.535177  0.903071  0.183349\n",
       "            Slowed Go Processing  0.529635  0.895206  0.181814\n",
       "            Guessing              0.537224  0.913442  0.185171\n",
       "Tracking    Confusion            -0.095310 -0.201330  0.021480\n",
       "            Slowed Go Processing  0.529901  0.896101  0.182643\n",
       "            Guessing              0.513343  0.879253  0.175802"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "corr_df = pd_df['SSRTscale=0'].corr(method='spearman')\n",
    "corr_df = corr_df.loc[corr_df.index.get_level_values(1)!='Independent Race', corr_df.columns.get_level_values(1)=='Independent Race']\n",
    "corr_df.droplevel(1, axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "SSRTscale=85\n",
      "min:  0.8476507208487455\n",
      "mean: 0.934802738158568\n",
      "SSRTscale=25\n",
      "min:  0.6612536988081903\n",
      "mean: 0.783511655554326\n",
      "SSRTscale=5\n",
      "min:  0.17775515842073933\n",
      "mean: 0.6261867288902175\n",
      "SSRTscale=0\n",
      "min:  0.021479529041817966\n",
      "mean: 0.5420881346425266\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th colspan=\"3\" halign=\"left\">SSRTscale=85 (range:0.85-0.99; mean:0.93)</th>\n",
       "      <th colspan=\"3\" halign=\"left\">SSRTscale=25 (range:0.66-0.97; mean:0.78)</th>\n",
       "      <th colspan=\"3\" halign=\"left\">SSRTscale=5 (range:0.18-0.89; mean:0.63)</th>\n",
       "      <th colspan=\"3\" halign=\"left\">SSRTscale=0 (range:0.02-0.91; mean:0.54)</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th>ABCD Weighted</th>\n",
       "      <th>Fixed</th>\n",
       "      <th>Tracking</th>\n",
       "      <th>ABCD Weighted</th>\n",
       "      <th>Fixed</th>\n",
       "      <th>Tracking</th>\n",
       "      <th>ABCD Weighted</th>\n",
       "      <th>Fixed</th>\n",
       "      <th>Tracking</th>\n",
       "      <th>ABCD Weighted</th>\n",
       "      <th>Fixed</th>\n",
       "      <th>Tracking</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Confusion</th>\n",
       "      <td>0.847651</td>\n",
       "      <td>0.973496</td>\n",
       "      <td>0.994806</td>\n",
       "      <td>0.729580</td>\n",
       "      <td>0.890478</td>\n",
       "      <td>0.965085</td>\n",
       "      <td>0.820041</td>\n",
       "      <td>0.892707</td>\n",
       "      <td>0.553595</td>\n",
       "      <td>0.846541</td>\n",
       "      <td>0.903071</td>\n",
       "      <td>0.021480</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Slowed Go Processing</th>\n",
       "      <td>0.932054</td>\n",
       "      <td>0.975553</td>\n",
       "      <td>0.948930</td>\n",
       "      <td>0.661254</td>\n",
       "      <td>0.895976</td>\n",
       "      <td>0.677161</td>\n",
       "      <td>0.177755</td>\n",
       "      <td>0.887390</td>\n",
       "      <td>0.269588</td>\n",
       "      <td>0.092678</td>\n",
       "      <td>0.895206</td>\n",
       "      <td>0.182643</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Guessing</th>\n",
       "      <td>0.882841</td>\n",
       "      <td>0.906937</td>\n",
       "      <td>0.950957</td>\n",
       "      <td>0.712721</td>\n",
       "      <td>0.765179</td>\n",
       "      <td>0.754173</td>\n",
       "      <td>0.812881</td>\n",
       "      <td>0.879659</td>\n",
       "      <td>0.342064</td>\n",
       "      <td>0.847930</td>\n",
       "      <td>0.913442</td>\n",
       "      <td>0.175802</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                     SSRTscale=85 (range:0.85-0.99; mean:0.93)            \\\n",
       "                                                 ABCD Weighted     Fixed   \n",
       "Confusion                                             0.847651  0.973496   \n",
       "Slowed Go Processing                                  0.932054  0.975553   \n",
       "Guessing                                              0.882841  0.906937   \n",
       "\n",
       "                               SSRTscale=25 (range:0.66-0.97; mean:0.78)  \\\n",
       "                      Tracking                             ABCD Weighted   \n",
       "Confusion             0.994806                                  0.729580   \n",
       "Slowed Go Processing  0.948930                                  0.661254   \n",
       "Guessing              0.950957                                  0.712721   \n",
       "\n",
       "                                          \\\n",
       "                         Fixed  Tracking   \n",
       "Confusion             0.890478  0.965085   \n",
       "Slowed Go Processing  0.895976  0.677161   \n",
       "Guessing              0.765179  0.754173   \n",
       "\n",
       "                     SSRTscale=5 (range:0.18-0.89; mean:0.63)            \\\n",
       "                                                ABCD Weighted     Fixed   \n",
       "Confusion                                            0.820041  0.892707   \n",
       "Slowed Go Processing                                 0.177755  0.887390   \n",
       "Guessing                                             0.812881  0.879659   \n",
       "\n",
       "                               SSRTscale=0 (range:0.02-0.91; mean:0.54)  \\\n",
       "                      Tracking                            ABCD Weighted   \n",
       "Confusion             0.553595                                 0.846541   \n",
       "Slowed Go Processing  0.269588                                 0.092678   \n",
       "Guessing              0.342064                                 0.847930   \n",
       "\n",
       "                                          \n",
       "                         Fixed  Tracking  \n",
       "Confusion             0.903071  0.021480  \n",
       "Slowed Go Processing  0.895206  0.182643  \n",
       "Guessing              0.913442  0.175802  "
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "corr_dict = {}\n",
    "simple_dict = {}\n",
    "SSRT_types = ['Weighted', 'Fixed', 'Tracking']\n",
    "for scale_key in pd_df.keys():\n",
    "    corr_array = []\n",
    "    simplified_corr_df = pd.DataFrame(index=['Confusion', 'Slowed Go Processing', 'Guessing', 'Independent Race'], columns=SSRT_types)\n",
    "    for ssrt_type in SSRT_types:\n",
    "        curr_corr_df = pd_df[scale_key].filter(regex=ssrt_type).corr(method='spearman').copy()\n",
    "        np.fill_diagonal(curr_corr_df.values, np.nan)\n",
    "        corr_vals = curr_corr_df[(ssrt_type, 'Independent Race')].values\n",
    "        assert np.isnan(corr_vals[-1])  # checking that the order is correct, aka independent condition is last\n",
    "        simplified_corr_df[ssrt_type] = curr_corr_df[(ssrt_type, 'Independent Race')].droplevel(0)\n",
    "        corr_array.append(corr_vals)\n",
    "        \n",
    "    corr_array = [r for rlist in corr_array for r in rlist]\n",
    "    print(scale_key)\n",
    "    print('min: ', np.nanmin(corr_array))\n",
    "    print('mean:', np.nanmean(corr_array))\n",
    "    corr_dict[scale_key] = corr_array\n",
    "    simple_dict[f'{scale_key} (range:{np.nanmin(corr_dict[scale_key]):.2f}-{np.nanmax(corr_dict[scale_key]):.2f}; mean:{np.nanmean(corr_dict[scale_key]):.2f})'] = simplified_corr_df\n",
    "    \n",
    "simple_table = pd.concat(simple_dict.values(), axis=1, keys=simple_dict.keys()).dropna()\n",
    "simple_table = simple_table.rename(columns={'Weighted': 'ABCD Weighted'}, level=1)\n",
    "dfi.export(simple_table, 'figures/correlations.png')\n",
    "simple_table"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Plot graded Mu function"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "plt.rc('xtick', labelsize=16) \n",
    "plt.rc('ytick', labelsize=16)\n",
    "plt.rc('axes', labelsize=22)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/henrymj/Documents/ABCD_simulations/scripts/utils.py:342: RuntimeWarning: divide by zero encountered in log\n",
      "  return self._at_least_0((np.log(SSD/max_SSD)/4+1) * mu_go)\n"
     ]
    }
   ],
   "source": [
    "SSDs = np.arange(0,600, 5)\n",
    "\n",
    "simulator = SimulateData()\n",
    "mu_gos = [[i, simulator._log_grade_mu(1, i)] for i in SSDs]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig,ax = plt.subplots(1,1, figsize=(14, 6))\n",
    "_ = sns.lineplot(x='SSD', y='proportion of \\u03BC', color='k', linewidth=5, data=pd.DataFrame(mu_gos, columns=['SSD', 'proportion of \\u03BC']))\n",
    "plt.savefig('figures/graded_mu_go_function.png', dpi=400)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}