experiment8.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Experiment 8"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "from copy import deepcopy\n",
    "from glob import glob\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib.gridspec as gridspec\n",
    "from scipy.interpolate import griddata\n",
    "from matplotlib.lines import Line2D\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import seaborn as sns\n",
    "import statsmodels.api as sm\n",
    "from scipy.stats import mannwhitneyu, shapiro, ttest_ind\n",
    "from sklearn.decomposition import PCA\n",
    "\n",
    "from population.population import Population\n",
    "from utils.dictionary import *\n",
    "from utils.myutils import load_dict"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Hyperparameters\n",
    "use_backup = True\n",
    "# populations = ['default', 'biased', 'gru_nr']\n",
    "# populations = ['default', 'gru_nr']\n",
    "populations = ['default', 'biased']\n",
    "max_v = 30\n",
    "max_g = 100\n",
    "hops = 10"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Parameters\n",
    "FOLDER = f\"population{'_backup' if use_backup else ''}/storage/experiment8/\"\n",
    "COLORS = [sns.color_palette()[i] for i in range(10)]\n",
    "TITLES = {\n",
    "    'default': 'NEAT-GRU',\n",
    "    'biased': 'BIASED',\n",
    "    'gru_nr': 'NEAT-GRU-NR',\n",
    "}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Final result"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Fetch"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# The dataframe which only concerns the last generation\n",
    "df_sol = pd.DataFrame()\n",
    "for p in populations:\n",
    "    results = []\n",
    "    for v in range(1, max_v+1):\n",
    "        temp = load_dict(f'{FOLDER}{p}/v{v}/evaluation/{max_g:05d}/results')\n",
    "        results.append(max([t['Fitness'] for t in temp.values()]))\n",
    "    df_sol[p] = results"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Analyse"
   ]
  },
  {
   "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>default</th>\n",
       "      <th>biased</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>4.433534</td>\n",
       "      <td>3.759852</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>3.941045</td>\n",
       "      <td>4.547839</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>4.273178</td>\n",
       "      <td>4.733581</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4.202701</td>\n",
       "      <td>4.833064</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4.690362</td>\n",
       "      <td>4.247697</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>4.128300</td>\n",
       "      <td>4.609283</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>4.045464</td>\n",
       "      <td>4.668146</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>4.197117</td>\n",
       "      <td>5.015152</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>4.566661</td>\n",
       "      <td>4.629198</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>4.823381</td>\n",
       "      <td>4.027082</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>3.217225</td>\n",
       "      <td>4.832674</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>3.880164</td>\n",
       "      <td>4.732736</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>4.375982</td>\n",
       "      <td>3.494474</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>4.346900</td>\n",
       "      <td>4.733985</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>4.326136</td>\n",
       "      <td>4.284900</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>4.009214</td>\n",
       "      <td>5.433048</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>3.587879</td>\n",
       "      <td>4.884469</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>4.185684</td>\n",
       "      <td>4.629418</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>4.738969</td>\n",
       "      <td>4.489544</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>3.130933</td>\n",
       "      <td>4.887915</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>4.734927</td>\n",
       "      <td>4.293545</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>4.316961</td>\n",
       "      <td>4.074360</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>4.389420</td>\n",
       "      <td>5.388078</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>4.360852</td>\n",
       "      <td>4.588269</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>4.456401</td>\n",
       "      <td>4.746834</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>3.858603</td>\n",
       "      <td>4.428766</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>3.728361</td>\n",
       "      <td>4.448758</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>4.203232</td>\n",
       "      <td>4.758556</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>5.027099</td>\n",
       "      <td>4.651342</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>4.189648</td>\n",
       "      <td>4.490508</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     default    biased\n",
       "0   4.433534  3.759852\n",
       "1   3.941045  4.547839\n",
       "2   4.273178  4.733581\n",
       "3   4.202701  4.833064\n",
       "4   4.690362  4.247697\n",
       "5   4.128300  4.609283\n",
       "6   4.045464  4.668146\n",
       "7   4.197117  5.015152\n",
       "8   4.566661  4.629198\n",
       "9   4.823381  4.027082\n",
       "10  3.217225  4.832674\n",
       "11  3.880164  4.732736\n",
       "12  4.375982  3.494474\n",
       "13  4.346900  4.733985\n",
       "14  4.326136  4.284900\n",
       "15  4.009214  5.433048\n",
       "16  3.587879  4.884469\n",
       "17  4.185684  4.629418\n",
       "18  4.738969  4.489544\n",
       "19  3.130933  4.887915\n",
       "20  4.734927  4.293545\n",
       "21  4.316961  4.074360\n",
       "22  4.389420  5.388078\n",
       "23  4.360852  4.588269\n",
       "24  4.456401  4.746834\n",
       "25  3.858603  4.428766\n",
       "26  3.728361  4.448758\n",
       "27  4.203232  4.758556\n",
       "28  5.027099  4.651342\n",
       "29  4.189648  4.490508"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXQAAAD4CAYAAAD8Zh1EAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAARy0lEQVR4nO3df2zcd33H8dcLx1q6NG23pZiMAplYBVcda4FbodQgu0ZVfw3KxqZYIFZ0U5QOXLapI6UnwYJkqdkmjQJbQ4SrFiGObbBmqKmysnG39jrBlrRNl+Y6rRphVEHqyiDUbTLF5r0//E1w3Dv7vu7ZZ3/8fEgnf/39fH+8z/n0dd9+vt/vfR0RAgCsfq/odQEAgO4g0AEgEQQ6ACSCQAeARBDoAJCIdb3a8aZNm2LLli292n1yXnjhBW3YsKHXZQAvQd/sroMHDz4XERe2auso0G0flfS8pGlJUxFRmtM+JOnvJX03m/V3EfHp+ba5ZcsWHThwoJPdowP1el1DQ0O9LgN4Cfpmd9n+Xru2PEfowxHx3DztD0fEDTm2BwDoIsbQASARnR6hh6QHbYekL0TEnhbLXGH7kKRjkm6NiCfnLmB7m6RtkjQwMKB6vb64qvESk5OT/D2xItE3l487ufXf9i9HxDHbr5T0TUljEfHQrPbzJP00IiZtXyfpzoi4eL5tlkqlYAy9exinxEpF3+wu2wfnnsc8raMhl4g4lv18VtJ9ki6f0/6TiJjMph+Q1G9708uqGgCQy4KBbnuD7Y2npyVdLenwnGVeZdvZ9OXZdn/Y/XIBrBbValXFYlEjIyMqFouqVqu9Lil5nYyhD0i6L8vrdZK+EhH7bW+XpIjYLen9km62PSXphKStwdc4AmtWtVpVpVLRxMSEpqen1dfXp3K5LEkaHR3tcXXp6mgMfSkwht5djFNiJSkWi/rc5z6n4eHhM32zVqtpbGxMhw8fXngDaOtlj6EDQB7NZlODg4NnzRscHFSz2exRRWtDz279B5CuQqGgnTt3au/evWo2myoUCrrxxhtVKBR6XVrSCHQAXTc8PKxdu3Zp165duuSSS3TkyBHt2LFD27dv73VpSSPQAXRdrVbTjh07dPfdd585Qt+xY4f27t3b69KSxknRRHBSFCtJX1+fTp48qf7+/jN989SpU1q/fr2mp6d7Xd6qxklRAMuqUCio0WicNa/RaDCGvsQIdABdV6lUVC6XVavVNDU1pVqtpnK5rEql0uvSksYYOoCuO33z0NjY2Jkx9PHxcW4qWmIEOoAlMTo6qtHRUc7vLCOGXAAgEQQ6ACSCQAeARBDoAJAIAh0AEkGgA0AiCHQASASBDgCJINABIBEEOgAkgkAHgEQQ6ACQCAIdABJBoANAIgh0AEgEgQ4AiSDQASARBDoAJIJAB4BEEOgAkAgCHQASQaADQCIIdABIBIEOAIkg0AEgER0Fuu2jtv/d9uO2D7Rot+3P2n7a9hO239L9UgEA81mXY9nhiHiuTdu1ki7OXm+TdFf2E11mO/c6EbEElQBYafIE+nzeK+lLMZMc37Z9ge3NEfGDLm0fmXbhvOW2fTp6x/XLXA3wM4s52JA44OimTgM9JD1oOyR9ISL2zGl/taTvz/r9mWzeWYFue5ukbZI0MDCger2+mJrRBn9P9FKtVms5/6b9L+ieaza0XY9+2z2dBvqVEXHM9islfdP2UxHx0Kz2Vh/NL/nYzT4I9khSqVSKoaGhvPWinf37xN8TKxJ9c9l0dFI0Io5lP5+VdJ+ky+cs8oyk18z6/SJJx7pRIACgMwsGuu0NtjeenpZ0taTDcxb7hqQPZVe7vF3SccbPAWB5dTLkMiDpvuyExzpJX4mI/ba3S1JE7Jb0gKTrJD0t6UVJH16acgEA7SwY6BHxX5IubTF/96zpkPSR7pYGAMiDO0UBIBEEOgAkgkAHgEQQ6ACQCAIdABJBoANAIgh0AEgEgQ4AiSDQASARBDoAJIJAB4BEEOgAkAgCHQASQaADQCIIdABIRKfPFAUASdKlOx/U8ROncq2z5bZ9uZY//5x+HfrU1bnWAYEOIKfjJ07p6B3Xd7x8vV7P/ZDovB8AmMGQCwAkgkAHgEQQ6ACQCAIdABJBoANAIgh0AEgEgQ4AiSDQASARBDoAJII7RQHksrFwm9507235Vro37z4kqfO7UTGDQAeQy/PNO7j1f4ViyAUAEkGgA0AiCHQASASBDgCJINABIBEEOgAkouNAt91n+zHb97doG7J93Pbj2euT3S0TALCQPNehf0xSU9J5bdofjogbXn5JAIDF6OgI3fZFmrlt64tLWw4AYLE6PUL/jKSPS9o4zzJX2D4k6ZikWyPiybkL2N4maZskDQwMqF6v56sW8+LvieWSp69NTk4uqm/Sn/NbMNBt3yDp2Yg4aHuozWKPSnpdREzavk7SXkkXz10oIvZI2iNJpVIp8t4OjHns35f79mpgUXL2tcXc+k9/XpxOhlyulPQe20clfVXSVba/PHuBiPhJRExm0w9I6re9qdvFAgDaWzDQI+ITEXFRRGyRtFXStyLig7OXsf0q286mL8+2+8MlqBcA0Maiv23R9nZJiojdkt4v6WbbU5JOSNoaEdGdEgEAncgV6BFRl1TPpnfPmv95SZ/vZmEAgHz4PnQAueX+vvL9+ZY//5z+fNuHJMm9GhkplUpx4MCBnux7pbt054M6fuLUku7j/HP6dehTVy/pPgBpJvzzPBAD87N9MCJKrdo4Ql+Bjp84lfs/gLyXhvFEGCA9fDkXACSCQAeARBDoAJAIAh0AEkGgA0AiCHQASASBDgCJINABIBEEOgAkgkAHgEQQ6ACQCAIdABJBoANAIgh0AEgEX5+7Am0s3KY33Xtb/hXvzbMPSeI7qoGUEOgr0PPNO/g+dAC5MeQCAIkg0AEgEQQ6ACSCQAeARBDoAJAIAh0AEkGgA0AiCHQASASBDgCJINABIBEEOgAkgkAHgEQQ6ACQCAIdABJBoANAIjoOdNt9th+zfX+LNtv+rO2nbT9h+y3dLRMAsJA8D7j4mKSmpPNatF0r6eLs9TZJd2U/AawRttu37Wq/XkQsQTVrU0eBbvsizTyvbFzSH7VY5L2SvhQz/zLftn2B7c0R8YPulbq2LOqJQvs7X+f8c/rzbx+YR7tgzvs0LSxep0fon5H0cUkb27S/WtL3Z/3+TDbvrEC3vU3SNkkaGBhQvV7PU+uacc81G3Kvc9P+F3Kvx98fy2FycpK+tkwWDHTbN0h6NiIO2h5qt1iLeS/5uI6IPZL2SFKpVAo+tbto/z6OgrAicYS+fDo5KXqlpPfYPirpq5Kusv3lOcs8I+k1s36/SNKxrlQIAOjIgoEeEZ+IiIsiYoukrZK+FREfnLPYNyR9KLva5e2SjjN+DgDLK89VLmexvV2SImK3pAckXSfpaUkvSvpwV6oDAHQsV6BHRF1SPZvePWt+SPpINwsDAOTDnaIAkAgCHQASQaADQCIIdABIBIEOAIkg0AEgEQQ6ACSCQAeARBDoAJAIAh0AEkGgA0AiCHQASASBDmBJVKtVFYtFjYyMqFgsqlqt9rqk5C3663MBoJ1qtapKpaKJiQlNT0+rr69P5XJZkjQ6Otrj6tLFETqArhsfH9fExISGh4e1bt06DQ8Pa2JiQuPj470uLWkEOoCuazabGhwcPGve4OCgms1mjypaGwh0AF1XKBTUaDTOmtdoNFQoFHpU0dpAoAPoukqlonK5rFqtpqmpKdVqNZXLZVUqlV6XljROigLoutMnPsfGxtRsNlUoFDQ+Ps4J0SVGoANYEqOjoxodHVW9XtfQ0FCvy1kTGHIBgEQQ6ACQCAIdABJBoANAIgh0AEgEgQ4AiSDQASARBDoAJIJAB4BEEOgAkAhu/V9lbLdv29V6fkQsUTUAVhKO0FeZiGj5qtVqbdsArA0EOgAkgkAHgEQsGOi219v+V9uHbD9pe2eLZYZsH7f9ePb65NKUCwBop5Mj9P+TdFVEXCrpMknX2H57i+UejojLstenu1olgFWnWq2qWCxqZGRExWJR1Wq11yUlb8GrXGLmrNpk9mt/9uJMG4C2qtWqKpWKJiYmND09rb6+PpXLZUniqUVLqKPLFm33SToo6Vcl/WVEfKfFYlfYPiTpmKRbI+LJFtvZJmmbJA0MDKhery+2bswxOTnJ3xMrxu23365bbrlFtnXy5Emde+65Ghsb0+23367Nmzf3urxkOc9lbbYvkHSfpLGIODxr/nmSfhoRk7avk3RnRFw837ZKpVIcOHBgkWVjLh7zhZWkr69PJ0+eVH9//5m+eerUKa1fv17T09O9Lm9Vs30wIkqt2nJd5RIRP5ZUl3TNnPk/iYjJbPoBSf22Ny2uXACrXaFQUKPROGteo9FQoVDoUUVrQydXuVyYHZnL9jmS3i3pqTnLvMrZLYy2L8+2+8PulwtgNahUKiqXy6rVapqamlKtVlO5XFalUul1aUnrZAx9s6R7s3H0V0j6m4i43/Z2SYqI3ZLeL+lm21OSTkjaGtyiCKxZp098jo2NqdlsqlAoaHx8nBOiSyzXGHo3MYbeXYyhY6Wib3ZX18bQAQArF4EOAIkg0AEgEQQ6ACSCQAeARBDoAJAIAh0AEkGgA0AiCHQASASBDgCJINABIBEEOgAkgkAHgEQQ6ACQCAIdABJBoANAIgj0Va5arapYLGpkZETFYlHVarXXJQHokU4eQYcVqlqtqlKpaGJiQtPT0+rr61O5XJYkHvUFrEEcoa9i4+PjmpiY0PDwsNatW6fh4WFNTExofHy816UB6AECfRVrNpsaHBw8a97g4KCazWaPKgLQSwT6KlYoFNRoNM6a12g0VCgUelQRgF4i0FexSqWicrmsWq2mqakp1Wo1lctlVSqVXpcGoAc4KbqKnT7xOTY2pmazqUKhoPHxcU6IAmsUR+gAkAiO0FcxLlsEMBtH6KsYly0CmI1AX8W4bBHAbAT6KsZliwBmI9BXMS5bBDAbJ0VXMS5bBDAbgb7KjY6OanR0VPV6XUNDQ70uB0APMeQCAIkg0AEgEQQ6ACSCQAeARBDoAJAIR0Rvdmz/j6Tv9WTnadok6bleFwG0QN/srtdFxIWtGnoW6Ogu2wciotTrOoC56JvLhyEXAEgEgQ4AiSDQ07Gn1wUAbdA3lwlj6ACQCI7QASARBDoAJIJAX4Fs/4ntW+dpv9D2d2w/Zvudi9j+TbY/n03faPuSl1Mv0md7i+3DLeZ/cSn7j+2jtjct1fZTQ6CvTiOSnoqIN0fEwy9zWzdKItCxKBHxexFxpNd1YAaBvkLYrtj+D9v/KOkN2bzX295v+6Dth22/0fZlkv5U0nW2H7d9ju27bB+w/aTtnbO2eeboxnbJdn3OPt8h6T2S/izb1uuX6/1iVVpn+17bT9j+mu2ft123XZKkefrhHbaPZOv9eTbvQttft/1v2evKbP4v2X4w+7/PL0hyT97pKsUDLlYA22+VtFXSmzXzb/KopIOaudxre0T8p+23SfqriLjK9icllSLio9n6lYj4X9t9kv7J9q9FxBML7Tci/sX2NyTdHxFfW6K3h3S8QVI5Ih6xfbek35/T/pJ+KOkZSe+T9MaICNsXZMveKekvIqJh+7WS/kFSQdKnJDUi4tO2r5e0bTneWCoI9JXhnZLui4gXJSkL2fWS3iHpb+0zByk/12b937G9TTP/nps1M4SyYKADOX0/Ih7Jpr8s6ZY57a364RFJJyV90fY+Sfdny75b0iWz+vZ5tjdKepek35SkiNhn+0dL9WZSRKCvHHNvCHiFpB9HxGXzrWT7VyTdKunXI+JHtu/RzIeBJE3pZ8Nq61usDuQxt4+e+b1dP4yIKduXa+a8z1ZJH5V0lWb65RURcWL2BrOA5+aYRWIMfWV4SNL7svHwjZJ+Q9KLkr5r+7clyTMubbHueZJekHTc9oCka2e1HZX01mz6t9rs+3lJG1/+W8Aa8FrbV2TTo5Ias9pa9kPb50o6PyIekPQHkk4foDyomXBXttzp+Q9J+kA271pJv7A0byVNBPoKEBGPSvprSY9L+rqk01eufEBS2fYhSU9Kem+LdQ9Jeixrv1vSI7Oad0q60/bDkqbb7P6rkv44OwnFSVHMpynpd20/IekXJd11umGefrhR0v3ZOv8s6Q+z+bdIKmUnSo9I2p7N3ynpXbYflXS1pP9e2reUFm79B4BEcIQOAIkg0AEgEQQ6ACSCQAeARBDoAJAIAh0AEkGgA0Ai/h/Wc1U/n9eRYwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df_sol.boxplot()\n",
    "df_sol"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 180x216 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(2.5,3))\n",
    "df_sol.boxplot(medianprops={'linewidth':3}, widths=(0.6, 0.6))\n",
    "plt.ylabel('fitness')\n",
    "plt.yticks([round(3+i/2,1) for i in range(5)])\n",
    "plt.xticks([i for i in range(1,3)], ['NEAT-GRU', 'BIASED'])\n",
    "plt.savefig(\"delete_me/temp_boxplot.png\", bbox_inches='tight', pad_inches=0.02, dpi=500)\n",
    "plt.savefig(\"delete_me/temp_boxplot.eps\", format='eps', bbox_inches='tight', pad_inches=0.02)\n",
    "plt.show()\n",
    "plt.close()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "###  Statistics\n",
    "\n",
    "Compare the final results (the fitness at gen100) of the data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "default    0.404229\n",
       "biased     0.321862\n",
       "dtype: float64"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Do not consider the outliers\n",
    "Q1 = df_sol.quantile(0.25)\n",
    "Q3 = df_sol.quantile(0.75)\n",
    "IQR = Q3 - Q1\n",
    "df_sol = df_sol[(df_sol >= (Q1 - 1.5 * IQR)) & (df_sol <= (Q3 + 1.5 * IQR))]\n",
    "IQR"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "stat=0.986, p=0.965\n",
      "default is probably Gaussian (p=0.96455)\n",
      "stat=0.957, p=0.333\n",
      "biased is probably Gaussian (p=0.33331)\n"
     ]
    }
   ],
   "source": [
    "# Example of the Shapiro-Wilk Normality Test\n",
    "for pop in populations:\n",
    "    stat, p = shapiro(df_sol[pop].dropna())\n",
    "    print('stat=%.3f, p=%.3f' % (stat, p))\n",
    "    if p > 0.05:\n",
    "        print(f'{pop} is probably Gaussian (p={p:.5f})')\n",
    "    else:\n",
    "        print(f'{pop} is probably not Gaussian (p={p:.5f})')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "for p in populations:\n",
    "    sm.qqplot(df_sol[p].dropna(), line ='s', fit=True) \n",
    "    plt.title(p)\n",
    "    plt.show() "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "default vs biased -  p=0.0007836679598086043 - significant\n"
     ]
    }
   ],
   "source": [
    "for i in range(len(populations)):\n",
    "    pop1 = populations[i]\n",
    "    for j in range(i+1, len(populations)):\n",
    "        pop2 = populations[j]\n",
    "        stat, p = mannwhitneyu(df_sol[pop1].dropna(), \n",
    "                               df_sol[pop2].dropna(), \n",
    "                               use_continuity=False, \n",
    "                               alternative='two-sided')\n",
    "        print(f'{pop1} vs {pop2} -  p={p} - {\"significant\" if p < .05 else \"insignificant\"}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "default vs biased -  p=0.0004890754003067441 - significant\n"
     ]
    }
   ],
   "source": [
    "for i in range(len(populations)):\n",
    "    pop1 = populations[i]\n",
    "    for j in range(i+1, len(populations)):\n",
    "        pop2 = populations[j]\n",
    "        stat, p = ttest_ind(df_sol[pop1].dropna(), \n",
    "                            df_sol[pop2].dropna())\n",
    "        print(f'{pop1} vs {pop2} -  p={p} - {\"significant\" if p < .05 else \"insignificant\"}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data throughout its lifetime"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Fetch"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "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>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "      <th>10</th>\n",
       "      <th>...</th>\n",
       "      <th>21</th>\n",
       "      <th>22</th>\n",
       "      <th>23</th>\n",
       "      <th>24</th>\n",
       "      <th>25</th>\n",
       "      <th>26</th>\n",
       "      <th>27</th>\n",
       "      <th>28</th>\n",
       "      <th>29</th>\n",
       "      <th>30</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.126805</td>\n",
       "      <td>0.101633</td>\n",
       "      <td>0.050000</td>\n",
       "      <td>0.233934</td>\n",
       "      <td>0.327431</td>\n",
       "      <td>0.339325</td>\n",
       "      <td>0.328967</td>\n",
       "      <td>0.024798</td>\n",
       "      <td>0.150672</td>\n",
       "      <td>0.036195</td>\n",
       "      <td>...</td>\n",
       "      <td>0.177431</td>\n",
       "      <td>0.076589</td>\n",
       "      <td>0.274061</td>\n",
       "      <td>0.133763</td>\n",
       "      <td>0.067068</td>\n",
       "      <td>0.193684</td>\n",
       "      <td>0.297609</td>\n",
       "      <td>0.207275</td>\n",
       "      <td>0.069902</td>\n",
       "      <td>0.228359</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.989438</td>\n",
       "      <td>2.127396</td>\n",
       "      <td>1.310473</td>\n",
       "      <td>3.125223</td>\n",
       "      <td>0.909649</td>\n",
       "      <td>1.360482</td>\n",
       "      <td>3.211603</td>\n",
       "      <td>1.706146</td>\n",
       "      <td>2.567259</td>\n",
       "      <td>3.638776</td>\n",
       "      <td>...</td>\n",
       "      <td>0.882742</td>\n",
       "      <td>1.706798</td>\n",
       "      <td>2.642011</td>\n",
       "      <td>1.817979</td>\n",
       "      <td>1.874606</td>\n",
       "      <td>1.723449</td>\n",
       "      <td>2.416587</td>\n",
       "      <td>1.618331</td>\n",
       "      <td>2.199212</td>\n",
       "      <td>2.038805</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2.961487</td>\n",
       "      <td>2.534089</td>\n",
       "      <td>1.976462</td>\n",
       "      <td>3.139509</td>\n",
       "      <td>1.276539</td>\n",
       "      <td>2.572645</td>\n",
       "      <td>3.092016</td>\n",
       "      <td>2.443143</td>\n",
       "      <td>2.989573</td>\n",
       "      <td>4.775277</td>\n",
       "      <td>...</td>\n",
       "      <td>1.842822</td>\n",
       "      <td>3.383173</td>\n",
       "      <td>3.651890</td>\n",
       "      <td>3.066244</td>\n",
       "      <td>3.153999</td>\n",
       "      <td>2.725632</td>\n",
       "      <td>2.967159</td>\n",
       "      <td>2.560004</td>\n",
       "      <td>3.763708</td>\n",
       "      <td>3.841512</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3.441146</td>\n",
       "      <td>2.424617</td>\n",
       "      <td>2.714817</td>\n",
       "      <td>3.598022</td>\n",
       "      <td>1.665802</td>\n",
       "      <td>3.210748</td>\n",
       "      <td>3.239986</td>\n",
       "      <td>2.795792</td>\n",
       "      <td>3.208624</td>\n",
       "      <td>4.775277</td>\n",
       "      <td>...</td>\n",
       "      <td>2.908500</td>\n",
       "      <td>3.572875</td>\n",
       "      <td>3.868428</td>\n",
       "      <td>3.317505</td>\n",
       "      <td>3.449682</td>\n",
       "      <td>3.147576</td>\n",
       "      <td>3.137299</td>\n",
       "      <td>2.926190</td>\n",
       "      <td>4.217827</td>\n",
       "      <td>4.106410</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>3.990679</td>\n",
       "      <td>2.639615</td>\n",
       "      <td>3.089644</td>\n",
       "      <td>3.635355</td>\n",
       "      <td>1.996252</td>\n",
       "      <td>3.156435</td>\n",
       "      <td>3.361441</td>\n",
       "      <td>3.895206</td>\n",
       "      <td>4.148091</td>\n",
       "      <td>4.763007</td>\n",
       "      <td>...</td>\n",
       "      <td>3.625303</td>\n",
       "      <td>3.815851</td>\n",
       "      <td>3.964599</td>\n",
       "      <td>3.389667</td>\n",
       "      <td>3.609956</td>\n",
       "      <td>3.147576</td>\n",
       "      <td>3.159310</td>\n",
       "      <td>3.206322</td>\n",
       "      <td>4.217827</td>\n",
       "      <td>4.118630</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>4.106995</td>\n",
       "      <td>2.769737</td>\n",
       "      <td>3.263029</td>\n",
       "      <td>4.050302</td>\n",
       "      <td>2.147238</td>\n",
       "      <td>3.667577</td>\n",
       "      <td>3.724040</td>\n",
       "      <td>3.858519</td>\n",
       "      <td>4.606316</td>\n",
       "      <td>4.763007</td>\n",
       "      <td>...</td>\n",
       "      <td>4.172205</td>\n",
       "      <td>4.120605</td>\n",
       "      <td>4.196758</td>\n",
       "      <td>3.791838</td>\n",
       "      <td>3.659898</td>\n",
       "      <td>3.333232</td>\n",
       "      <td>3.185493</td>\n",
       "      <td>3.233318</td>\n",
       "      <td>4.253082</td>\n",
       "      <td>4.253365</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>4.162789</td>\n",
       "      <td>3.082459</td>\n",
       "      <td>3.449377</td>\n",
       "      <td>3.891735</td>\n",
       "      <td>2.306245</td>\n",
       "      <td>3.677926</td>\n",
       "      <td>3.783568</td>\n",
       "      <td>4.227342</td>\n",
       "      <td>4.668751</td>\n",
       "      <td>4.723657</td>\n",
       "      <td>...</td>\n",
       "      <td>4.226038</td>\n",
       "      <td>4.120644</td>\n",
       "      <td>4.188411</td>\n",
       "      <td>3.965832</td>\n",
       "      <td>4.262881</td>\n",
       "      <td>3.330664</td>\n",
       "      <td>3.244084</td>\n",
       "      <td>3.333827</td>\n",
       "      <td>4.253082</td>\n",
       "      <td>4.295890</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>4.278171</td>\n",
       "      <td>3.252926</td>\n",
       "      <td>3.626462</td>\n",
       "      <td>4.180924</td>\n",
       "      <td>3.098313</td>\n",
       "      <td>3.622870</td>\n",
       "      <td>3.987695</td>\n",
       "      <td>4.256615</td>\n",
       "      <td>4.668751</td>\n",
       "      <td>4.711361</td>\n",
       "      <td>...</td>\n",
       "      <td>4.367625</td>\n",
       "      <td>4.201678</td>\n",
       "      <td>4.396356</td>\n",
       "      <td>4.150417</td>\n",
       "      <td>4.339924</td>\n",
       "      <td>3.217996</td>\n",
       "      <td>3.398359</td>\n",
       "      <td>3.613967</td>\n",
       "      <td>4.406726</td>\n",
       "      <td>4.108170</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>4.357510</td>\n",
       "      <td>3.252674</td>\n",
       "      <td>3.683472</td>\n",
       "      <td>4.180924</td>\n",
       "      <td>3.153837</td>\n",
       "      <td>3.736347</td>\n",
       "      <td>4.120402</td>\n",
       "      <td>4.237638</td>\n",
       "      <td>4.682665</td>\n",
       "      <td>4.705055</td>\n",
       "      <td>...</td>\n",
       "      <td>4.453445</td>\n",
       "      <td>4.098451</td>\n",
       "      <td>4.360764</td>\n",
       "      <td>4.129152</td>\n",
       "      <td>4.339924</td>\n",
       "      <td>3.544088</td>\n",
       "      <td>3.549260</td>\n",
       "      <td>4.084654</td>\n",
       "      <td>4.623388</td>\n",
       "      <td>3.962428</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>4.345988</td>\n",
       "      <td>3.623530</td>\n",
       "      <td>4.054629</td>\n",
       "      <td>4.197221</td>\n",
       "      <td>4.606053</td>\n",
       "      <td>4.025181</td>\n",
       "      <td>4.163817</td>\n",
       "      <td>4.220711</td>\n",
       "      <td>4.682122</td>\n",
       "      <td>4.860543</td>\n",
       "      <td>...</td>\n",
       "      <td>4.587743</td>\n",
       "      <td>4.317112</td>\n",
       "      <td>4.370704</td>\n",
       "      <td>4.297323</td>\n",
       "      <td>4.358415</td>\n",
       "      <td>3.754612</td>\n",
       "      <td>3.525316</td>\n",
       "      <td>4.242246</td>\n",
       "      <td>5.027099</td>\n",
       "      <td>4.098954</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>4.433534</td>\n",
       "      <td>3.941045</td>\n",
       "      <td>4.273178</td>\n",
       "      <td>4.202701</td>\n",
       "      <td>4.690362</td>\n",
       "      <td>4.128300</td>\n",
       "      <td>4.045464</td>\n",
       "      <td>4.197117</td>\n",
       "      <td>4.566661</td>\n",
       "      <td>4.823381</td>\n",
       "      <td>...</td>\n",
       "      <td>4.734927</td>\n",
       "      <td>4.316961</td>\n",
       "      <td>4.389420</td>\n",
       "      <td>4.360852</td>\n",
       "      <td>4.456401</td>\n",
       "      <td>3.858603</td>\n",
       "      <td>3.728361</td>\n",
       "      <td>4.203232</td>\n",
       "      <td>5.027099</td>\n",
       "      <td>4.189648</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>11 rows × 30 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          1         2         3         4         5         6         7   \\\n",
       "0   0.126805  0.101633  0.050000  0.233934  0.327431  0.339325  0.328967   \n",
       "1   1.989438  2.127396  1.310473  3.125223  0.909649  1.360482  3.211603   \n",
       "2   2.961487  2.534089  1.976462  3.139509  1.276539  2.572645  3.092016   \n",
       "3   3.441146  2.424617  2.714817  3.598022  1.665802  3.210748  3.239986   \n",
       "4   3.990679  2.639615  3.089644  3.635355  1.996252  3.156435  3.361441   \n",
       "5   4.106995  2.769737  3.263029  4.050302  2.147238  3.667577  3.724040   \n",
       "6   4.162789  3.082459  3.449377  3.891735  2.306245  3.677926  3.783568   \n",
       "7   4.278171  3.252926  3.626462  4.180924  3.098313  3.622870  3.987695   \n",
       "8   4.357510  3.252674  3.683472  4.180924  3.153837  3.736347  4.120402   \n",
       "9   4.345988  3.623530  4.054629  4.197221  4.606053  4.025181  4.163817   \n",
       "10  4.433534  3.941045  4.273178  4.202701  4.690362  4.128300  4.045464   \n",
       "\n",
       "          8         9         10  ...        21        22        23        24  \\\n",
       "0   0.024798  0.150672  0.036195  ...  0.177431  0.076589  0.274061  0.133763   \n",
       "1   1.706146  2.567259  3.638776  ...  0.882742  1.706798  2.642011  1.817979   \n",
       "2   2.443143  2.989573  4.775277  ...  1.842822  3.383173  3.651890  3.066244   \n",
       "3   2.795792  3.208624  4.775277  ...  2.908500  3.572875  3.868428  3.317505   \n",
       "4   3.895206  4.148091  4.763007  ...  3.625303  3.815851  3.964599  3.389667   \n",
       "5   3.858519  4.606316  4.763007  ...  4.172205  4.120605  4.196758  3.791838   \n",
       "6   4.227342  4.668751  4.723657  ...  4.226038  4.120644  4.188411  3.965832   \n",
       "7   4.256615  4.668751  4.711361  ...  4.367625  4.201678  4.396356  4.150417   \n",
       "8   4.237638  4.682665  4.705055  ...  4.453445  4.098451  4.360764  4.129152   \n",
       "9   4.220711  4.682122  4.860543  ...  4.587743  4.317112  4.370704  4.297323   \n",
       "10  4.197117  4.566661  4.823381  ...  4.734927  4.316961  4.389420  4.360852   \n",
       "\n",
       "          25        26        27        28        29        30  \n",
       "0   0.067068  0.193684  0.297609  0.207275  0.069902  0.228359  \n",
       "1   1.874606  1.723449  2.416587  1.618331  2.199212  2.038805  \n",
       "2   3.153999  2.725632  2.967159  2.560004  3.763708  3.841512  \n",
       "3   3.449682  3.147576  3.137299  2.926190  4.217827  4.106410  \n",
       "4   3.609956  3.147576  3.159310  3.206322  4.217827  4.118630  \n",
       "5   3.659898  3.333232  3.185493  3.233318  4.253082  4.253365  \n",
       "6   4.262881  3.330664  3.244084  3.333827  4.253082  4.295890  \n",
       "7   4.339924  3.217996  3.398359  3.613967  4.406726  4.108170  \n",
       "8   4.339924  3.544088  3.549260  4.084654  4.623388  3.962428  \n",
       "9   4.358415  3.754612  3.525316  4.242246  5.027099  4.098954  \n",
       "10  4.456401  3.858603  3.728361  4.203232  5.027099  4.189648  \n",
       "\n",
       "[11 rows x 30 columns]"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 'default' population\n",
    "data = []\n",
    "for v in range(1, max_v+1):\n",
    "    row = []\n",
    "    for g in range(0,max_g+1,hops):\n",
    "        # Read in the generation's results\n",
    "        results = load_dict(f\"{FOLDER}default/v{v}/evaluation/{g:05d}/results\")\n",
    "\n",
    "        # Use the best performing genome\n",
    "        row.append(max([results[k][D_FITNESS] for k in results.keys()]))\n",
    "    data.append(row)\n",
    "\n",
    "# Create a dataframe containing the data\n",
    "df_default = pd.DataFrame(data).transpose()  # Rows indicate generations\n",
    "df_default.columns = [v for v in range(1, max_v+1)]\n",
    "df_default"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "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>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "      <th>10</th>\n",
       "      <th>...</th>\n",
       "      <th>21</th>\n",
       "      <th>22</th>\n",
       "      <th>23</th>\n",
       "      <th>24</th>\n",
       "      <th>25</th>\n",
       "      <th>26</th>\n",
       "      <th>27</th>\n",
       "      <th>28</th>\n",
       "      <th>29</th>\n",
       "      <th>30</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.156048</td>\n",
       "      <td>0.080691</td>\n",
       "      <td>0.069907</td>\n",
       "      <td>0.138917</td>\n",
       "      <td>0.053134</td>\n",
       "      <td>0.153652</td>\n",
       "      <td>0.178346</td>\n",
       "      <td>0.061173</td>\n",
       "      <td>0.136151</td>\n",
       "      <td>0.217417</td>\n",
       "      <td>...</td>\n",
       "      <td>0.158420</td>\n",
       "      <td>0.187629</td>\n",
       "      <td>0.179461</td>\n",
       "      <td>0.226030</td>\n",
       "      <td>0.024115</td>\n",
       "      <td>0.130385</td>\n",
       "      <td>0.024954</td>\n",
       "      <td>0.024514</td>\n",
       "      <td>0.061239</td>\n",
       "      <td>0.183872</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2.476002</td>\n",
       "      <td>3.126799</td>\n",
       "      <td>1.069502</td>\n",
       "      <td>1.743819</td>\n",
       "      <td>0.628329</td>\n",
       "      <td>1.027974</td>\n",
       "      <td>2.439807</td>\n",
       "      <td>3.376216</td>\n",
       "      <td>1.766950</td>\n",
       "      <td>2.021245</td>\n",
       "      <td>...</td>\n",
       "      <td>1.803602</td>\n",
       "      <td>3.047463</td>\n",
       "      <td>1.948476</td>\n",
       "      <td>2.364693</td>\n",
       "      <td>2.734676</td>\n",
       "      <td>1.002972</td>\n",
       "      <td>3.138642</td>\n",
       "      <td>1.292238</td>\n",
       "      <td>2.179443</td>\n",
       "      <td>2.577080</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3.137368</td>\n",
       "      <td>3.676572</td>\n",
       "      <td>3.011294</td>\n",
       "      <td>3.441800</td>\n",
       "      <td>1.310980</td>\n",
       "      <td>1.359591</td>\n",
       "      <td>3.761198</td>\n",
       "      <td>3.853109</td>\n",
       "      <td>3.293189</td>\n",
       "      <td>2.189106</td>\n",
       "      <td>...</td>\n",
       "      <td>2.908970</td>\n",
       "      <td>3.565233</td>\n",
       "      <td>2.915454</td>\n",
       "      <td>2.689415</td>\n",
       "      <td>3.177113</td>\n",
       "      <td>3.368175</td>\n",
       "      <td>3.193004</td>\n",
       "      <td>2.680134</td>\n",
       "      <td>2.353373</td>\n",
       "      <td>3.099926</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3.150766</td>\n",
       "      <td>3.751116</td>\n",
       "      <td>3.413886</td>\n",
       "      <td>3.635455</td>\n",
       "      <td>1.744564</td>\n",
       "      <td>1.512387</td>\n",
       "      <td>4.192663</td>\n",
       "      <td>4.077763</td>\n",
       "      <td>3.364808</td>\n",
       "      <td>3.240291</td>\n",
       "      <td>...</td>\n",
       "      <td>3.072776</td>\n",
       "      <td>3.746148</td>\n",
       "      <td>4.005686</td>\n",
       "      <td>3.559450</td>\n",
       "      <td>3.391891</td>\n",
       "      <td>3.501265</td>\n",
       "      <td>3.429637</td>\n",
       "      <td>3.046553</td>\n",
       "      <td>3.681894</td>\n",
       "      <td>3.105308</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>3.201373</td>\n",
       "      <td>3.986651</td>\n",
       "      <td>3.578233</td>\n",
       "      <td>3.800397</td>\n",
       "      <td>2.476164</td>\n",
       "      <td>2.281302</td>\n",
       "      <td>4.307949</td>\n",
       "      <td>4.077763</td>\n",
       "      <td>3.381364</td>\n",
       "      <td>3.396658</td>\n",
       "      <td>...</td>\n",
       "      <td>3.330538</td>\n",
       "      <td>4.008615</td>\n",
       "      <td>4.321141</td>\n",
       "      <td>4.222100</td>\n",
       "      <td>3.410010</td>\n",
       "      <td>3.647525</td>\n",
       "      <td>3.520989</td>\n",
       "      <td>2.870671</td>\n",
       "      <td>4.032315</td>\n",
       "      <td>3.222551</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>3.594190</td>\n",
       "      <td>4.135158</td>\n",
       "      <td>3.763934</td>\n",
       "      <td>3.731302</td>\n",
       "      <td>2.658765</td>\n",
       "      <td>2.749509</td>\n",
       "      <td>4.388189</td>\n",
       "      <td>3.911971</td>\n",
       "      <td>3.908636</td>\n",
       "      <td>3.395785</td>\n",
       "      <td>...</td>\n",
       "      <td>3.467426</td>\n",
       "      <td>4.005947</td>\n",
       "      <td>4.314480</td>\n",
       "      <td>4.331093</td>\n",
       "      <td>3.688292</td>\n",
       "      <td>3.757952</td>\n",
       "      <td>3.683453</td>\n",
       "      <td>2.934464</td>\n",
       "      <td>4.246455</td>\n",
       "      <td>3.294395</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>3.310188</td>\n",
       "      <td>4.135716</td>\n",
       "      <td>3.523174</td>\n",
       "      <td>3.747646</td>\n",
       "      <td>2.990278</td>\n",
       "      <td>3.448622</td>\n",
       "      <td>4.382606</td>\n",
       "      <td>3.922173</td>\n",
       "      <td>3.953072</td>\n",
       "      <td>3.379673</td>\n",
       "      <td>...</td>\n",
       "      <td>3.672620</td>\n",
       "      <td>4.096034</td>\n",
       "      <td>4.442276</td>\n",
       "      <td>4.331093</td>\n",
       "      <td>3.759900</td>\n",
       "      <td>3.957532</td>\n",
       "      <td>3.583483</td>\n",
       "      <td>3.159205</td>\n",
       "      <td>4.287527</td>\n",
       "      <td>3.051837</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>3.576015</td>\n",
       "      <td>4.206963</td>\n",
       "      <td>3.733330</td>\n",
       "      <td>3.703552</td>\n",
       "      <td>3.287396</td>\n",
       "      <td>3.783671</td>\n",
       "      <td>4.382606</td>\n",
       "      <td>3.922417</td>\n",
       "      <td>4.003380</td>\n",
       "      <td>3.427020</td>\n",
       "      <td>...</td>\n",
       "      <td>3.954723</td>\n",
       "      <td>4.357104</td>\n",
       "      <td>4.465127</td>\n",
       "      <td>4.354889</td>\n",
       "      <td>3.864615</td>\n",
       "      <td>3.932160</td>\n",
       "      <td>3.944563</td>\n",
       "      <td>3.201077</td>\n",
       "      <td>4.338000</td>\n",
       "      <td>3.318984</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>3.628713</td>\n",
       "      <td>4.213775</td>\n",
       "      <td>3.823057</td>\n",
       "      <td>3.553979</td>\n",
       "      <td>3.676105</td>\n",
       "      <td>3.783671</td>\n",
       "      <td>4.247758</td>\n",
       "      <td>4.004709</td>\n",
       "      <td>4.051243</td>\n",
       "      <td>3.486915</td>\n",
       "      <td>...</td>\n",
       "      <td>4.222928</td>\n",
       "      <td>4.360977</td>\n",
       "      <td>4.360001</td>\n",
       "      <td>4.164980</td>\n",
       "      <td>3.873307</td>\n",
       "      <td>4.151871</td>\n",
       "      <td>3.955310</td>\n",
       "      <td>3.414850</td>\n",
       "      <td>4.308147</td>\n",
       "      <td>3.248747</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>3.668481</td>\n",
       "      <td>4.128112</td>\n",
       "      <td>3.758078</td>\n",
       "      <td>3.648109</td>\n",
       "      <td>3.772470</td>\n",
       "      <td>3.707399</td>\n",
       "      <td>4.238658</td>\n",
       "      <td>4.045259</td>\n",
       "      <td>4.098325</td>\n",
       "      <td>3.596846</td>\n",
       "      <td>...</td>\n",
       "      <td>4.423864</td>\n",
       "      <td>4.338070</td>\n",
       "      <td>4.561890</td>\n",
       "      <td>4.215552</td>\n",
       "      <td>4.024527</td>\n",
       "      <td>4.208542</td>\n",
       "      <td>4.225947</td>\n",
       "      <td>3.741254</td>\n",
       "      <td>4.494074</td>\n",
       "      <td>3.463231</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>3.901886</td>\n",
       "      <td>4.217767</td>\n",
       "      <td>3.853896</td>\n",
       "      <td>3.864039</td>\n",
       "      <td>3.774737</td>\n",
       "      <td>3.801315</td>\n",
       "      <td>4.204621</td>\n",
       "      <td>4.058280</td>\n",
       "      <td>4.193806</td>\n",
       "      <td>3.556775</td>\n",
       "      <td>...</td>\n",
       "      <td>4.479409</td>\n",
       "      <td>4.250460</td>\n",
       "      <td>4.561890</td>\n",
       "      <td>4.151989</td>\n",
       "      <td>4.013142</td>\n",
       "      <td>4.279682</td>\n",
       "      <td>4.288351</td>\n",
       "      <td>3.926476</td>\n",
       "      <td>4.406835</td>\n",
       "      <td>3.525163</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>11 rows × 30 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          1         2         3         4         5         6         7   \\\n",
       "0   0.156048  0.080691  0.069907  0.138917  0.053134  0.153652  0.178346   \n",
       "1   2.476002  3.126799  1.069502  1.743819  0.628329  1.027974  2.439807   \n",
       "2   3.137368  3.676572  3.011294  3.441800  1.310980  1.359591  3.761198   \n",
       "3   3.150766  3.751116  3.413886  3.635455  1.744564  1.512387  4.192663   \n",
       "4   3.201373  3.986651  3.578233  3.800397  2.476164  2.281302  4.307949   \n",
       "5   3.594190  4.135158  3.763934  3.731302  2.658765  2.749509  4.388189   \n",
       "6   3.310188  4.135716  3.523174  3.747646  2.990278  3.448622  4.382606   \n",
       "7   3.576015  4.206963  3.733330  3.703552  3.287396  3.783671  4.382606   \n",
       "8   3.628713  4.213775  3.823057  3.553979  3.676105  3.783671  4.247758   \n",
       "9   3.668481  4.128112  3.758078  3.648109  3.772470  3.707399  4.238658   \n",
       "10  3.901886  4.217767  3.853896  3.864039  3.774737  3.801315  4.204621   \n",
       "\n",
       "          8         9         10  ...        21        22        23        24  \\\n",
       "0   0.061173  0.136151  0.217417  ...  0.158420  0.187629  0.179461  0.226030   \n",
       "1   3.376216  1.766950  2.021245  ...  1.803602  3.047463  1.948476  2.364693   \n",
       "2   3.853109  3.293189  2.189106  ...  2.908970  3.565233  2.915454  2.689415   \n",
       "3   4.077763  3.364808  3.240291  ...  3.072776  3.746148  4.005686  3.559450   \n",
       "4   4.077763  3.381364  3.396658  ...  3.330538  4.008615  4.321141  4.222100   \n",
       "5   3.911971  3.908636  3.395785  ...  3.467426  4.005947  4.314480  4.331093   \n",
       "6   3.922173  3.953072  3.379673  ...  3.672620  4.096034  4.442276  4.331093   \n",
       "7   3.922417  4.003380  3.427020  ...  3.954723  4.357104  4.465127  4.354889   \n",
       "8   4.004709  4.051243  3.486915  ...  4.222928  4.360977  4.360001  4.164980   \n",
       "9   4.045259  4.098325  3.596846  ...  4.423864  4.338070  4.561890  4.215552   \n",
       "10  4.058280  4.193806  3.556775  ...  4.479409  4.250460  4.561890  4.151989   \n",
       "\n",
       "          25        26        27        28        29        30  \n",
       "0   0.024115  0.130385  0.024954  0.024514  0.061239  0.183872  \n",
       "1   2.734676  1.002972  3.138642  1.292238  2.179443  2.577080  \n",
       "2   3.177113  3.368175  3.193004  2.680134  2.353373  3.099926  \n",
       "3   3.391891  3.501265  3.429637  3.046553  3.681894  3.105308  \n",
       "4   3.410010  3.647525  3.520989  2.870671  4.032315  3.222551  \n",
       "5   3.688292  3.757952  3.683453  2.934464  4.246455  3.294395  \n",
       "6   3.759900  3.957532  3.583483  3.159205  4.287527  3.051837  \n",
       "7   3.864615  3.932160  3.944563  3.201077  4.338000  3.318984  \n",
       "8   3.873307  4.151871  3.955310  3.414850  4.308147  3.248747  \n",
       "9   4.024527  4.208542  4.225947  3.741254  4.494074  3.463231  \n",
       "10  4.013142  4.279682  4.288351  3.926476  4.406835  3.525163  \n",
       "\n",
       "[11 rows x 30 columns]"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 'gru_nr' population\n",
    "data = []\n",
    "for v in range(1, max_v+1):\n",
    "    row = []\n",
    "    for g in range(0,max_g+1,hops):\n",
    "        # Read in the generation's results\n",
    "        results = load_dict(f\"{FOLDER}gru_nr/v{v}/evaluation/{g:05d}/results\")\n",
    "\n",
    "        # Use the best performing genome\n",
    "        row.append(max([results[k][D_FITNESS] for k in results.keys()]))\n",
    "    data.append(row)\n",
    "\n",
    "# Create a dataframe containing the data\n",
    "df_gru_nr = pd.DataFrame(data).transpose()  # Rows indicate generations\n",
    "df_gru_nr.columns = [v for v in range(1, max_v+1)]\n",
    "df_gru_nr"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "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>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "      <th>10</th>\n",
       "      <th>...</th>\n",
       "      <th>21</th>\n",
       "      <th>22</th>\n",
       "      <th>23</th>\n",
       "      <th>24</th>\n",
       "      <th>25</th>\n",
       "      <th>26</th>\n",
       "      <th>27</th>\n",
       "      <th>28</th>\n",
       "      <th>29</th>\n",
       "      <th>30</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.115264</td>\n",
       "      <td>0.215772</td>\n",
       "      <td>0.050018</td>\n",
       "      <td>0.174669</td>\n",
       "      <td>0.095919</td>\n",
       "      <td>0.135221</td>\n",
       "      <td>0.243556</td>\n",
       "      <td>0.118306</td>\n",
       "      <td>0.124377</td>\n",
       "      <td>0.105093</td>\n",
       "      <td>...</td>\n",
       "      <td>0.042237</td>\n",
       "      <td>0.179380</td>\n",
       "      <td>0.270821</td>\n",
       "      <td>0.217791</td>\n",
       "      <td>0.048462</td>\n",
       "      <td>0.192170</td>\n",
       "      <td>0.245334</td>\n",
       "      <td>0.610784</td>\n",
       "      <td>0.172314</td>\n",
       "      <td>0.024115</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.046696</td>\n",
       "      <td>3.676311</td>\n",
       "      <td>2.146224</td>\n",
       "      <td>2.269551</td>\n",
       "      <td>2.113905</td>\n",
       "      <td>2.114707</td>\n",
       "      <td>2.621355</td>\n",
       "      <td>1.729156</td>\n",
       "      <td>3.633864</td>\n",
       "      <td>1.752015</td>\n",
       "      <td>...</td>\n",
       "      <td>1.695648</td>\n",
       "      <td>0.654125</td>\n",
       "      <td>3.500848</td>\n",
       "      <td>2.114138</td>\n",
       "      <td>3.331517</td>\n",
       "      <td>1.240560</td>\n",
       "      <td>3.197277</td>\n",
       "      <td>0.777648</td>\n",
       "      <td>3.793369</td>\n",
       "      <td>2.901407</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.777647</td>\n",
       "      <td>3.714755</td>\n",
       "      <td>3.008915</td>\n",
       "      <td>3.477012</td>\n",
       "      <td>2.908226</td>\n",
       "      <td>3.322373</td>\n",
       "      <td>3.058498</td>\n",
       "      <td>2.618778</td>\n",
       "      <td>4.332502</td>\n",
       "      <td>2.431926</td>\n",
       "      <td>...</td>\n",
       "      <td>3.311496</td>\n",
       "      <td>1.921058</td>\n",
       "      <td>4.130446</td>\n",
       "      <td>2.833432</td>\n",
       "      <td>3.866569</td>\n",
       "      <td>3.680862</td>\n",
       "      <td>4.462375</td>\n",
       "      <td>2.071019</td>\n",
       "      <td>3.798973</td>\n",
       "      <td>3.521779</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2.545361</td>\n",
       "      <td>4.091895</td>\n",
       "      <td>3.361248</td>\n",
       "      <td>3.826317</td>\n",
       "      <td>3.546776</td>\n",
       "      <td>4.064863</td>\n",
       "      <td>4.080600</td>\n",
       "      <td>3.274158</td>\n",
       "      <td>4.596982</td>\n",
       "      <td>2.920237</td>\n",
       "      <td>...</td>\n",
       "      <td>3.396544</td>\n",
       "      <td>3.181549</td>\n",
       "      <td>4.313658</td>\n",
       "      <td>3.727081</td>\n",
       "      <td>4.076207</td>\n",
       "      <td>3.818209</td>\n",
       "      <td>4.765190</td>\n",
       "      <td>2.841428</td>\n",
       "      <td>4.089695</td>\n",
       "      <td>3.779437</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2.783155</td>\n",
       "      <td>4.088330</td>\n",
       "      <td>3.563896</td>\n",
       "      <td>3.872672</td>\n",
       "      <td>3.851611</td>\n",
       "      <td>4.176898</td>\n",
       "      <td>4.321965</td>\n",
       "      <td>3.885388</td>\n",
       "      <td>4.586705</td>\n",
       "      <td>3.006257</td>\n",
       "      <td>...</td>\n",
       "      <td>3.558225</td>\n",
       "      <td>3.895376</td>\n",
       "      <td>4.041809</td>\n",
       "      <td>3.820661</td>\n",
       "      <td>4.300999</td>\n",
       "      <td>3.743467</td>\n",
       "      <td>4.639857</td>\n",
       "      <td>3.084288</td>\n",
       "      <td>4.167292</td>\n",
       "      <td>3.771030</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>3.202661</td>\n",
       "      <td>4.149387</td>\n",
       "      <td>3.661168</td>\n",
       "      <td>3.895410</td>\n",
       "      <td>3.862550</td>\n",
       "      <td>4.382228</td>\n",
       "      <td>4.488198</td>\n",
       "      <td>4.046991</td>\n",
       "      <td>4.627837</td>\n",
       "      <td>3.427169</td>\n",
       "      <td>...</td>\n",
       "      <td>4.036633</td>\n",
       "      <td>3.901380</td>\n",
       "      <td>4.210385</td>\n",
       "      <td>4.260759</td>\n",
       "      <td>4.511173</td>\n",
       "      <td>3.928054</td>\n",
       "      <td>4.357225</td>\n",
       "      <td>3.346917</td>\n",
       "      <td>4.302208</td>\n",
       "      <td>4.037544</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>3.264155</td>\n",
       "      <td>4.387390</td>\n",
       "      <td>4.030953</td>\n",
       "      <td>4.405818</td>\n",
       "      <td>4.053993</td>\n",
       "      <td>4.452497</td>\n",
       "      <td>4.462647</td>\n",
       "      <td>4.691389</td>\n",
       "      <td>4.684200</td>\n",
       "      <td>3.438703</td>\n",
       "      <td>...</td>\n",
       "      <td>4.221724</td>\n",
       "      <td>3.900008</td>\n",
       "      <td>4.377848</td>\n",
       "      <td>4.298856</td>\n",
       "      <td>4.586302</td>\n",
       "      <td>3.755521</td>\n",
       "      <td>4.839274</td>\n",
       "      <td>4.199553</td>\n",
       "      <td>4.302208</td>\n",
       "      <td>4.294657</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>3.681930</td>\n",
       "      <td>4.490074</td>\n",
       "      <td>4.109378</td>\n",
       "      <td>4.384869</td>\n",
       "      <td>4.182746</td>\n",
       "      <td>4.507601</td>\n",
       "      <td>4.509110</td>\n",
       "      <td>5.063955</td>\n",
       "      <td>4.666387</td>\n",
       "      <td>3.738883</td>\n",
       "      <td>...</td>\n",
       "      <td>4.302450</td>\n",
       "      <td>3.917772</td>\n",
       "      <td>4.377646</td>\n",
       "      <td>4.541406</td>\n",
       "      <td>4.754683</td>\n",
       "      <td>4.100779</td>\n",
       "      <td>4.388978</td>\n",
       "      <td>4.176465</td>\n",
       "      <td>4.409489</td>\n",
       "      <td>4.409690</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>3.636116</td>\n",
       "      <td>4.500831</td>\n",
       "      <td>4.153034</td>\n",
       "      <td>4.600638</td>\n",
       "      <td>4.277730</td>\n",
       "      <td>4.578865</td>\n",
       "      <td>4.544028</td>\n",
       "      <td>5.046209</td>\n",
       "      <td>4.667625</td>\n",
       "      <td>3.884609</td>\n",
       "      <td>...</td>\n",
       "      <td>4.222818</td>\n",
       "      <td>3.933827</td>\n",
       "      <td>4.716496</td>\n",
       "      <td>4.568725</td>\n",
       "      <td>4.881343</td>\n",
       "      <td>4.161530</td>\n",
       "      <td>4.826509</td>\n",
       "      <td>4.354012</td>\n",
       "      <td>4.506054</td>\n",
       "      <td>4.533084</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>3.592435</td>\n",
       "      <td>4.547839</td>\n",
       "      <td>4.441216</td>\n",
       "      <td>4.555059</td>\n",
       "      <td>4.265624</td>\n",
       "      <td>4.624152</td>\n",
       "      <td>4.669095</td>\n",
       "      <td>4.989994</td>\n",
       "      <td>4.678259</td>\n",
       "      <td>4.025693</td>\n",
       "      <td>...</td>\n",
       "      <td>4.264852</td>\n",
       "      <td>4.058572</td>\n",
       "      <td>5.332158</td>\n",
       "      <td>4.556062</td>\n",
       "      <td>4.832375</td>\n",
       "      <td>4.260963</td>\n",
       "      <td>4.826864</td>\n",
       "      <td>4.629492</td>\n",
       "      <td>4.651342</td>\n",
       "      <td>4.479308</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>3.759852</td>\n",
       "      <td>4.547839</td>\n",
       "      <td>4.733581</td>\n",
       "      <td>4.833064</td>\n",
       "      <td>4.247697</td>\n",
       "      <td>4.609283</td>\n",
       "      <td>4.668146</td>\n",
       "      <td>5.015152</td>\n",
       "      <td>4.629198</td>\n",
       "      <td>4.027082</td>\n",
       "      <td>...</td>\n",
       "      <td>4.293545</td>\n",
       "      <td>4.074360</td>\n",
       "      <td>5.388078</td>\n",
       "      <td>4.588269</td>\n",
       "      <td>4.746834</td>\n",
       "      <td>4.428766</td>\n",
       "      <td>4.448758</td>\n",
       "      <td>4.758556</td>\n",
       "      <td>4.651342</td>\n",
       "      <td>4.490508</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>11 rows × 30 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          1         2         3         4         5         6         7   \\\n",
       "0   0.115264  0.215772  0.050018  0.174669  0.095919  0.135221  0.243556   \n",
       "1   1.046696  3.676311  2.146224  2.269551  2.113905  2.114707  2.621355   \n",
       "2   1.777647  3.714755  3.008915  3.477012  2.908226  3.322373  3.058498   \n",
       "3   2.545361  4.091895  3.361248  3.826317  3.546776  4.064863  4.080600   \n",
       "4   2.783155  4.088330  3.563896  3.872672  3.851611  4.176898  4.321965   \n",
       "5   3.202661  4.149387  3.661168  3.895410  3.862550  4.382228  4.488198   \n",
       "6   3.264155  4.387390  4.030953  4.405818  4.053993  4.452497  4.462647   \n",
       "7   3.681930  4.490074  4.109378  4.384869  4.182746  4.507601  4.509110   \n",
       "8   3.636116  4.500831  4.153034  4.600638  4.277730  4.578865  4.544028   \n",
       "9   3.592435  4.547839  4.441216  4.555059  4.265624  4.624152  4.669095   \n",
       "10  3.759852  4.547839  4.733581  4.833064  4.247697  4.609283  4.668146   \n",
       "\n",
       "          8         9         10  ...        21        22        23        24  \\\n",
       "0   0.118306  0.124377  0.105093  ...  0.042237  0.179380  0.270821  0.217791   \n",
       "1   1.729156  3.633864  1.752015  ...  1.695648  0.654125  3.500848  2.114138   \n",
       "2   2.618778  4.332502  2.431926  ...  3.311496  1.921058  4.130446  2.833432   \n",
       "3   3.274158  4.596982  2.920237  ...  3.396544  3.181549  4.313658  3.727081   \n",
       "4   3.885388  4.586705  3.006257  ...  3.558225  3.895376  4.041809  3.820661   \n",
       "5   4.046991  4.627837  3.427169  ...  4.036633  3.901380  4.210385  4.260759   \n",
       "6   4.691389  4.684200  3.438703  ...  4.221724  3.900008  4.377848  4.298856   \n",
       "7   5.063955  4.666387  3.738883  ...  4.302450  3.917772  4.377646  4.541406   \n",
       "8   5.046209  4.667625  3.884609  ...  4.222818  3.933827  4.716496  4.568725   \n",
       "9   4.989994  4.678259  4.025693  ...  4.264852  4.058572  5.332158  4.556062   \n",
       "10  5.015152  4.629198  4.027082  ...  4.293545  4.074360  5.388078  4.588269   \n",
       "\n",
       "          25        26        27        28        29        30  \n",
       "0   0.048462  0.192170  0.245334  0.610784  0.172314  0.024115  \n",
       "1   3.331517  1.240560  3.197277  0.777648  3.793369  2.901407  \n",
       "2   3.866569  3.680862  4.462375  2.071019  3.798973  3.521779  \n",
       "3   4.076207  3.818209  4.765190  2.841428  4.089695  3.779437  \n",
       "4   4.300999  3.743467  4.639857  3.084288  4.167292  3.771030  \n",
       "5   4.511173  3.928054  4.357225  3.346917  4.302208  4.037544  \n",
       "6   4.586302  3.755521  4.839274  4.199553  4.302208  4.294657  \n",
       "7   4.754683  4.100779  4.388978  4.176465  4.409489  4.409690  \n",
       "8   4.881343  4.161530  4.826509  4.354012  4.506054  4.533084  \n",
       "9   4.832375  4.260963  4.826864  4.629492  4.651342  4.479308  \n",
       "10  4.746834  4.428766  4.448758  4.758556  4.651342  4.490508  \n",
       "\n",
       "[11 rows x 30 columns]"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 'biased' population\n",
    "data = []\n",
    "for v in range(1, max_v+1):\n",
    "    row = []\n",
    "    for g in range(0,max_g+1,hops):\n",
    "        # Read in the generation's results\n",
    "        results = load_dict(f\"{FOLDER}biased/v{v}/evaluation/{g:05d}/results\")\n",
    "\n",
    "        # Use the best performing genome\n",
    "        row.append(max([results[k][D_FITNESS] for k in results.keys()]))\n",
    "    data.append(row)\n",
    "\n",
    "# Create a dataframe containing the data\n",
    "df_biased = pd.DataFrame(data).transpose()  # Rows indicate generations\n",
    "df_biased.columns = [v for v in range(1, max_v+1)]\n",
    "df_biased"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Plot"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_performance(plot_default, plot_biased, plot_gru_nr):\n",
    "    plt.figure(figsize=(4, 2.5))\n",
    "    max_fitness = 0\n",
    "\n",
    "\n",
    "    ### DEFAULT ###\n",
    "    if plot_default:\n",
    "        c = 1\n",
    "        # Get quantiles\n",
    "        q1 = df_default.quantile(0.25, axis=1).values\n",
    "        q2 = df_default.quantile(0.50, axis=1).values\n",
    "        q3 = df_default.quantile(0.75, axis=1).values\n",
    "\n",
    "        # Plot\n",
    "        plt.plot([i for i in range(len(q1))], \n",
    "                 q1, \n",
    "                 color=COLORS[c], \n",
    "                 linestyle=\":\", \n",
    "                 linewidth=.5)\n",
    "        plt.plot([i for i in range(len(q3))], \n",
    "                 q3, color=COLORS[c], \n",
    "                 linestyle=\":\", \n",
    "                 linewidth=.5)\n",
    "        plt.plot([i for i in range(len(q2))], \n",
    "                 q2, \n",
    "                 color=COLORS[c], \n",
    "                 linestyle=\"-\", \n",
    "                 linewidth=2, \n",
    "                 label=TITLES['default'])\n",
    "        plt.fill_between([i for i in range(len(q1))],\n",
    "                         q1, \n",
    "                         q3,\n",
    "                         color=COLORS[c], \n",
    "                         alpha=0.1)\n",
    "        max_fitness = max(q3.max(), max_fitness)\n",
    "\n",
    "\n",
    "    ### BIASED ###\n",
    "    if plot_biased:\n",
    "        c = 0\n",
    "        # Get quantiles\n",
    "        q1 = df_biased.quantile(0.25, axis=1).values\n",
    "        q2 = df_biased.quantile(0.50, axis=1).values\n",
    "        q3 = df_biased.quantile(0.75, axis=1).values\n",
    "\n",
    "        # Plot\n",
    "        plt.plot([i for i in range(len(q1))], \n",
    "                 q1, \n",
    "                 color=COLORS[c], \n",
    "                 linestyle=\":\", \n",
    "                 linewidth=.5)\n",
    "        plt.plot([i for i in range(len(q3))], \n",
    "                 q3, color=COLORS[c], \n",
    "                 linestyle=\":\", \n",
    "                 linewidth=.5)\n",
    "        plt.plot([i for i in range(len(q2))], \n",
    "                 q2, \n",
    "                 color=COLORS[c], \n",
    "                 linestyle=\"-\", \n",
    "                 linewidth=2, \n",
    "                 label=TITLES['biased'])\n",
    "        plt.fill_between([i for i in range(len(q1))],\n",
    "                         q1, \n",
    "                         q3,\n",
    "                         color=COLORS[c], \n",
    "                         alpha=0.1)\n",
    "        max_fitness = max(q3.max(), max_fitness)\n",
    "\n",
    "\n",
    "    ### GRU-NR ###\n",
    "    if plot_gru_nr:\n",
    "        c = 2\n",
    "        # Get quantiles\n",
    "        q1 = df_gru_nr.quantile(0.25, axis=1).values\n",
    "        q2 = df_gru_nr.quantile(0.50, axis=1).values\n",
    "        q3 = df_gru_nr.quantile(0.75, axis=1).values\n",
    "\n",
    "        # Plot\n",
    "        plt.plot([i for i in range(len(q1))], \n",
    "                 q1, \n",
    "                 color=COLORS[c], \n",
    "                 linestyle=\":\", \n",
    "                 linewidth=.5)\n",
    "        plt.plot([i for i in range(len(q3))], \n",
    "                 q3, color=COLORS[c], \n",
    "                 linestyle=\":\", \n",
    "                 linewidth=.5)\n",
    "        plt.plot([i for i in range(len(q2))], \n",
    "                 q2, \n",
    "                 color=COLORS[c], \n",
    "                 linestyle=\"-\", \n",
    "                 linewidth=2, \n",
    "                 label=TITLES['gru_nr'])\n",
    "        plt.fill_between([i for i in range(len(q1))],\n",
    "                         q1, \n",
    "                         q3,\n",
    "                         color=COLORS[c], \n",
    "                         alpha=0.1)\n",
    "        max_fitness = max(q3.max(), max_fitness)\n",
    "\n",
    "\n",
    "    # Show the figure\n",
    "    leg = plt.legend(loc='upper center',\n",
    "                     bbox_to_anchor=(0.5, 1.25),\n",
    "                     fancybox=True,\n",
    "                     fontsize=10,\n",
    "                     ncol=len(populations))\n",
    "    for line in leg.get_lines():\n",
    "        line.set_linewidth(4.0)\n",
    "    plt.xlabel(\"generation\")\n",
    "    plt.xticks([i for i in range(11)], [i*10 for i in range(11)])\n",
    "    plt.xlim(0,10)\n",
    "    plt.ylabel('fitness')\n",
    "    plt.yticks([i for i in range(int(max_fitness)+2)])\n",
    "    plt.ylim(0,int(max_fitness)+1)\n",
    "    plt.grid()\n",
    "    plt.tight_layout()\n",
    "    name = \"experiment8\"\n",
    "    if plot_default: name += \"_default\"\n",
    "    if plot_biased: name += \"_biased\"\n",
    "    if plot_gru_nr: name += \"_gru_nr\"\n",
    "    plt.savefig(f\"delete_me/{name}.png\", bbox_inches='tight', pad_inches=0.02, dpi=500)\n",
    "    plt.show()\n",
    "    plt.close()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 288x180 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_performance(plot_default=True, plot_biased=False, plot_gru_nr=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 288x180 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_performance(True, True, False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Genome space of 'default'"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Fetch"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "# More populations to compare with\n",
    "max_v = 54\n",
    "min_fit = 3.5"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "genomes = []\n",
    "for v in range(1, max_v+1):\n",
    "    pop = Population(\n",
    "        name=f'default/v{v}',\n",
    "        folder_name='experiment8',\n",
    "        log_print=False,\n",
    "        use_backup=True,\n",
    "    )\n",
    "    for gen in range(3*hops,max_g+1,hops):\n",
    "        pop.load(gen)\n",
    "        for g in pop.population.values():  # Go over all the current genomes\n",
    "            if g.fitness and g.fitness > min_fit: genomes.append(deepcopy(g))\n",
    "        for i in range(1, 6):  # Add the elites of the past five generations\n",
    "            g = pop.best_genome_hist[gen - i][1]\n",
    "            if g.fitness > min_fit: genomes.append(deepcopy(g))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Parse"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_param(g):\n",
    "    result = [v for v in g.nodes[2].bias_h]  # GRU biases\n",
    "    result += [v[0] for v in g.nodes[2].weight_xh_full]  # GRU input->output\n",
    "    result += [v[0] for v in g.nodes[2].weight_hh]  # GRU hidden->hidden\n",
    "    result += [g.nodes[1].bias]\n",
    "    result += [g.connections[(-1, 2)].weight, g.connections[(2, 1)].weight, g.connections[(-1, 1)].weight]\n",
    "    return result"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "params = []\n",
    "for g in genomes:\n",
    "    params.append(get_param(g))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "head = [\n",
    "    'bias_r', 'bias_z', 'bias_h',\n",
    "    'weight_xr', 'weight_xz', 'weight_xh',\n",
    "    'weight_hr', 'weight_hz', 'weight_hh',\n",
    "    'bias_rw', 'c1', 'c2', 'c0',    \n",
    "]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_g = pd.DataFrame(params, columns=head)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Final size: 1643\n"
     ]
    }
   ],
   "source": [
    "df_g = df_g.drop_duplicates()\n",
    "print(f\"Final size: {len(df_g)}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Visualise"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(20, 5))\n",
    "df_g.boxplot()\n",
    "plt.ylim(-6,6)\n",
    "plt.show()\n",
    "plt.close()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x360 with 13 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot as a whole\n",
    "plt.figure(figsize=(15,5))\n",
    "for i, col in enumerate(df_g.columns):\n",
    "    # Drop the outliers first\n",
    "    data_temp = df_g.copy()\n",
    "    q1 = data_temp[col].quantile(0.25)\n",
    "    q3 = data_temp[col].quantile(0.75)\n",
    "    iqr = q3 - q1\n",
    "    data_temp[col] = data_temp[col][((q1 - 1.5 * iqr) < data_temp[col]) & (data_temp[col] < (q3 + 1.5 * iqr))]\n",
    "    \n",
    "    # Plot\n",
    "    plt.subplot(int(len(data_temp.columns)/5+1),5,i+1)\n",
    "    sns.violinplot(data_temp[col], color=COLORS[1])\n",
    "    plt.title(col)\n",
    "    if 'bias' in col:\n",
    "        plt.xlim(-3,3)\n",
    "    else:\n",
    "        plt.xlim(-6,6)\n",
    "    plt.yticks([])\n",
    "plt.tight_layout()\n",
    "plt.show()\n",
    "plt.close()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Plot separate and save\n",
    "def violin_creator(h, t=None, height:float=.6):\n",
    "    plt.figure(figsize=(3.25,height))\n",
    "\n",
    "    # Format data\n",
    "    data_temp = df_g.copy()\n",
    "\n",
    "    # Drop the outliers first\n",
    "    q1 = data_temp[h].quantile(0.25)\n",
    "    q3 = data_temp[h].quantile(0.75)\n",
    "    iqr = q3 - q1\n",
    "    data_temp[h] = data_temp[h][((q1 - 1.5 * iqr) < data_temp[h]) & (data_temp[h] < (q3 + 1.5 * iqr))]\n",
    "\n",
    "    # Create the plot\n",
    "    sns.violinplot(data_temp[h].dropna(), color='#ff7f0e')\n",
    "    if 'bias' in h:\n",
    "        plt.xlim(-3,3)\n",
    "    else:\n",
    "        plt.xlim(-6,6)\n",
    "    if t: plt.title(t)\n",
    "    plt.yticks([])\n",
    "    plt.xlabel('')\n",
    "    plt.savefig(f\"delete_me/{h}.png\", bbox_inches='tight', pad_inches=0.02, dpi=500)\n",
    "    plt.savefig(f\"delete_me/{h}.eps\", format='eps', bbox_inches='tight', pad_inches=0.02)\n",
    "    plt.show()\n",
    "    plt.close()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAM4AAABQCAYAAABRc0r3AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAAQ2UlEQVR4nO2da2wdx3XHf2d37/vFl0hKJCXKkmyLUmLZokVJcZoASmI3aJO2QJoCKdAi/pCi6AsoirRNkbYI+ghcFGj7JWjRAkGbNl+SAC2S+hE5tFHElixZj8iibD1iRZQoUiJ5n7zP3dMP94omLVmk7iXvlcj5AYO7S87OmeHe/87MmbNDUVUMBsO9YbW6AgbDg4gRjsFQB0Y4BkMdGOEYDHVghGMw1IERjsFQB0Y4BkMdGOGsMURERWR7q+ux1jHCaREi4rS6Dob6McJpIiLyroh8WUROAzkReUpEfiwiSRE5JSIfX5D3N0XkkohkROSnIvKFBb/7ooiMicisiLwgIltqP3+1luWUiGRF5PNNbeB6QlVNalIC3gVOAgNAHzANfJrqA+yTtfMNQARIA4/UrtsI7Kod/xJwAdgJOMCfAT9eYEOB7a1u61pPpsdpPv+oqleAXwd+oKo/UFVPVV8CjlEVEoAH7BaRkKpOqOpbtZ9/CfgbVR1T1Qrw18CeW72OoTkY4TSfK7XPLcDnasO0pIgkgaeAjaqaAz4P/BYwISLfF5FHF1z3DwuumQGEag9maBJGOM3nVjj6FeDfVbVtQYqo6t8CqOoLqvpJqsO0c8C/LLjuS++7LqSqP256S9YxRjit4z+AXxSRp0XEFpGgiHxcRPpFpEdEPiMiEaAIZAG3dt03gD8RkV0AIpIQkc8tKHcSeKiZDVmPGOG0iNo857PAnwI3qPYkf0T1nljAHwLXqA7FPgb8du267wFfB74tImngDPDzC4r+C+CbtaHcrzalMesQqXliDAbDPWB6HIOhDoxwDIY6MMIxGOrACMdgqAMjHIOhDhqK0O3q6tLBwcEVqsraZ1Gsk2UhIq2ukuF9HD9+/KaqblgqX0PCGRwc5NixY40UsSZRVS5fvszZs2d55513uPzuT5mamiJfKC7KFwkF2dDdzZbBrQwODjI0NMRDDz2EbdstqrlBRC4vJ595J2SFUFXOnz/Pq6++ytEjr3NzegaAuN+jP5hnOFom0V7BZymCkndtshWbGzPTnLr+Lq+8Uh01h4IBPvzYHoaHh9m7dy9tbW2tbJbhAzDCaZBkMsnLL7/M4R++xMT1SRwLHo1mOTSQ49HYHG2+CssZkaXKNpeyId7Jhhk78RpHjhxBRBjauZMDBw+yf/9+Ojo6Vr9BhmXRUOTA8PCwrsehmqpy5swZnn/+eY4eOYLreWyL5nmyPc1jbVlCttdg+XA17+d0KsrpdJzreR8iMLRziIMf+QgHDhwwPdEqISLHVXV4yXxGOMsnmUwyOjrKSy++wMT1ScKOsq89yYHOFD3B8qrZvV7wc2I2ysl0gsm8U+2JhnZy8OBHGBkZMT3RCmKEs0IUi0WOHz/OK6OjvPnmm7iex9ZIgQOdSfa0ZfFbzYv1U62JKBnlVCrBZMFBBHbs2MG+fSM8+eST9Pf3G29dAxjhNEAqleLkyZMcPXqUN48fo1gqE/d77E2kGOlM0xsstbqKAEzk/fwkFeF0Os74nB+A7g1dPP7EXvbs2cOuXbuIRqMtruWDhRHOMlFVJicnuXTpEmNjY5x96wzvXr6MKsR8yofiaR5ry7Ajmse6jx/kyZLDW+kIY+kw53NRii6IwOCWLTy6c4iHH36Y7du309vba9zdd8EIp4aqUigUSCaTzM7OcvPmTaamprh+/TpXx69w5coV5vIFAHwWDIbzbI/OsTOeoz9UvK/F8kFUPLg8F+R8NszFbJjL+RCl2mtwAb+fzZsH6B/YzKZNm+jp6WHDhg10dnbS1taG46xvR+tyhfPA/JWKxSKpVIpMJkM2myWbzZLL5cjlcszNzS1KuVyWbCZDLpslnc1SLlduKy/uV3r8BR4PF+nvKtIXqib7ARTK+3Es2BYtsC1aAGZwa3Oj8bkgV/N+Jq4nefNn5/lRaXHElQjEohHi8QSJtjai0RjRaJRYLEYkEiEajS5KsViMWCxGMBhcd/OqlgnH8zxyuRzpdJp0Ok0ymSSZTJJKpZidna19zpCcnSWdTlMsfbDXyhIIOUrQVoKWS1DKxG2PXscj0uYScVzivgoJn0vCV6bDX2nqpL7V2AJ9oRJ9ocVzs6IrzJR8TJd8pMo26bJDppIkm50km7SZ8XzkXJu5ilC5i4fd59jEYlHi8QSxeGJeULfEFg6HCYfDBAIBAoEAjuPgOA6WVRWu53m4rkulUqFcLlMqlSgWixQKhUWpWCxSLBYpl8uUy2Vc18XzPEQEy7KwbZtAIIDf7yccDhOJRIjFYiQSCTo6OuZ71ZUYqq6ocPL5PNPT00xPT88PjW4JI51Ok8mkyaRTZDJZsrk57jRMFCDiU+K+CjG7xBbHJZpwiTrVFLarQgjbHiHbJWR7+C1d1iKjYTEBW9kYKrExtLSzo+QJeddirmIz51rkKjZzrk2uYpNzLXKVGbKzNtmbDpOujznXYq4ieCv0fPJZ4LcVv6U4Uk2WKBaKAorgqlBWi5InFCpC+Q5ity2Lrq5OejduZNOmPvr6+ujv72dgYOCe1sYaEs7MzAzPPfcck9cnmJqaIpubu92ABVGfErErhK0y3Y7LYMAlEvHmxRB1XGJOhahTFcVaGC6tNfyW4rdcEj536cw1VKHoCQXXouBZlDyLsidUtPolV711oxVbwBbFZyk+y6vZ8wjUPuuZa1Y8IVuxyFQc0mWbZNlhtuRjupTi5vlx3j5zmsKC5sSikWWX3ZBwkrOzXDzxf3T4CjwWLNMer9Dur5DwVYg7FWI+l6DlrYne4LvjXVzNB1a0zIJrkXctQrZHsMFog5WgL1TkV/pvrlh5IlSHz7bLe5v0NA/HUtr8Lm3+O9tWhXTFZrLgZ6LgZ+Ie7m9DwvFbHjG7SNkTrhf8XC/4GynuvuZqPkDBW1k3bjAY5NCnDnH48GEKucKKll0PV/OBFX84rFXMi2wt5NChQzz77LMcOnSo1VUx3CMN9TgbAiV+d8fVlarLfc0/ne/jYi68omUePnx40Wer6QsV1839/CD+bpn5GhLOeD7I185to9Mp0OEv0+Ev0+av0OarEHNcYr4KYbu+id39Rl+ouHSme6TgFnjtxe/RaXsEI/fHHGc94Smkys78NONaPkB1f8elaUg47e3tDD35MSYmrjE2NUVqJnNbHksg4mjVhWyViTgukZpLObLIq/bep3MfrrGs5KR5veAp5F2LgmtR9KppsVetmk8ELBTHAkc8fO/zqAUsxZZ7W3JQrbrQs5XqC4OpskOq7DBT8jFTcrhZDnCj6J+PqABoS8SXXf6KhtwUi0VmZ2eZnp5mdnaWZDJJOp2eX/HPZNKkUymymQyZbI6Ke2dvR8hR4j6XqFUi5ntPVLfc1WG7mkK2N++RWgu92v1K1a1s1dZmLObcBWs4FYvc/LFNznWY8xxyFZv87QEbdWMJ+G3eW8exFBtFpLaOo7fWcYSyJ+Q/YA3J59h0d2+gu2cjfX3vreNs3ryZWCzWmpCbQCBAb28vvb29S+ZVVfL5/KLIgVQqNR9BkEwmSc7OMjk7w9up1G3v699m265FD1guIatCyH5PYFHHJeZzSfiqw8gOf+W+cP+2mrmKNf8EvvVEztSe0DnXYW6ZC5kBv594PEasPU5HPMGWWuRANBolEoksK3LA8zwqlQqlUmlR5ECxWCSfz98xcuDWdQCWZeE4Dn6/fz5yIBwOk0gkSCQStLe309nZSTwen7fbCC0LuRGR+cYtR2jlcplMJkM6nSaXy5HJZO4Sq5Yjl81woxbXdifRRXxKt79IT6BIf7hIf6ia7sdhYqO4CtfyAa7MVd3NE4UAU6Ug2fLibtqyhEQsRqK9jVg8wab3CWBhjNrCWDWfz9eilrWOBybI0+fz0dHRUdfbjqVSaT4EaHp6msnJyVp09Dhv/ewyr89UIx4cCzaHCmyP5tgZn2NLuPBADgHLnvDTXJAL2RAXchGuzAXnw09CwQCDg1vZ398/Hx3d3d29ok/j9cADI5xG8Pv9dHd3093dfdvvVJXp6WkuXLjAuXPnOHv2LV66eIkXJ5WIT9kdq+4j8Ehs7r4OBZouOpxNRzibjnAhF6HsVXv1bQ9t5ZmhXTz88MNs27aNnp6edRfJvBqsC+HcDRGhq6uLrq4u9u/fD0Amk+HUqVMcO3aMN44e4chMgphPeaItyUhHmk3LCIpcbVThWsHP6WR1Q4+JfHW4tLG3h09+tPoG6NDQEOHwyq49Gaqs+RfZGqVcLnPixAlGR0c59sYbVFyXLZEiBzqSPN6WIWA3d8+Bq/kAJ5JRTqcT3CjYiAg7H32UfSMjDA8Ps2nTpqbVZy1i3gBdBTKZzPwuN+NXrxFylOG2FAc7U8sKza+HWz3LyWSMk6k4NwoOliV8aPeHOHDwIPv27TNbRa0gRjiriKoyNjbG888/z+uvvTbfC+1rT7GnLUPEaczV7SmM5wOcSkb5STrOVKG6JdTu3bt46qmPMjIyQjy+/MU6w/IxwmkS6XSa0dFRfvjSi4xfvYZjwY5Ijl2J6k6enf7ysla8Z0sOF7Mh3smGOJeNkS5ZWJawe9duDhw8yMjIiOlZmoARTpNRVS5dujS/d/Tk1A2g+hJff3CODn+FNl8Zx1KEaihKtuJws+hjohgiU1tTiYRDPLbncfbu3cvw8DCxWKyFrVp/GOG0EFVlfHx80X8ruDE1RWbBG7K3NsbYsKGHLYODbN26laGhITZv3my2b2oha26XmwcJEWFgYICBgQGefvrp+Z+Xy2UqlQqe5xEMBo1AHmAa6nFEJAO8vXLVWTZdQCvClY3dtW/7EVVdcnzcaI/z9nK6tZVGRI4Zu2vXbitti8iy5h4mMMlgqAMjHIOhDhoVzj+vSC2MXWP3/rG9LLsNOQcMhvWKGaoZDHXQkHBE5GsiclpETorIiyLStNBcEXlORM7V7H9PRJoSjyIinxORt0TEE5FV9/qIyDMi8raIXBCRP15tezWb/yYiUyKyvC1fVs7ugIj8SETGan/j32+S3aCIHBWRUzW7f7nkRapadwLiC45/D/hGI+Xdo+1PAU7t+OvA15tkdyfwCDAKDK+yLRu4CDwE+IFTwFAT2vhzwBPAmWbdz5rdjcATteMY8E6T2itAtHbsA44A++92TUM9jqqmF5xGgKZNmFT1RVW9tY/K60B/k+yOqWqzFn33ARdU9ZKqloBvA59dbaOq+iows9p27mB3QlXfrB1ngDGgrwl2VVWztVNfLd31u9zwHEdE/kpErgBfAL7aaHl18kXgf1tkezXpA64sOB+nCV+k+wERGQQep/r0b4Y9W0ROAlPAS6p6V7tLCkdEfigiZ+6QPgugql9R1QHgW8DvrEQjlmu7lucrQKVmv2l2m8SdXkhY825QEYkC3wH+4H2jmlVDVV1V3UN15LJPRHbfLf+SITeq+oll2v5P4PvAny8z/5IsZVtEfgP4BeCQ1gaozbDbRMaBgQXn/cC1FtWlKYiIj6povqWq3222fVVNisgo8Ax32Q+3Ua/ajgWnnwHONVLePdp+Bvgy8BlVvf0/Wq0N3gB2iMhWEfEDvwb8d4vrtGpIdfudfwXGVPXvm2h3wy2vrIiEgE+w1He5QW/Ed6iq8jTwP0BfEz0wF6iO/0/WUlM8esAvU+0JisAk8MIq2/s0Ve/SReArTWrjfwETQLnW1mebZPcpqkPR0wvu66ebYPfDwIma3TPAV5e6xkQOGAx1YCIHDIY6MMIxGOrACMdgqAMjHIOhDoxwDIY6MMIxGOrACMdgqAMjHIOhDv4f4FPr2Kn2/zcAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 234x43.2 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 234x43.2 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 234x43.2 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 234x43.2 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAM4AAABQCAYAAABRc0r3AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAAOYklEQVR4nO2deXAU153HP78ZjTQ6ESChY2QkQGAkIA7mWDsuErxUudgiTiq7ruyWs7v2QlLlrYorjvGdteMj6zMpQ+1uJZvC2N6yAR/Yu1sWYBaC7fUGH2ALMOYUlixxSaMDoWPu3/4xDZYVQKOWpkdI71M1pT5e9/fX0/r2+73Xr3tEVTEYDIPDleoADIbLEWMcg8EGxjgGgw2McQwGGxjjGAw2MMYxGGxgjDPCEZFbReT9VMdh+DrGOKMIEXlYRF5KdRxjAWMcg8EGxjhJQERURCr7zL8gIr+ypheLSJOIPCAifhGpF5Ef9Sk7UUT+W0Q6ReQjYFq/fa8WkUZr/W4RWWQtXwo8APy1iHSJyB5r+TgReU5ETorIcRH5lYi4HfgaRjXGOKmhGCgAfMAtwO9F5Epr3b8BAaAEWG59+vIx8E1gArAOeE1EvKq6BXgceEVVc1T1Kqv8i0AEqATmAjcAP07WgY0VjHFSx4OqGlTVd4Ea4IdWTfBXwEOq2q2qnxH/xz+Pqr6kqq2qGlHV3wAZwJV/sndARIqAvwDusPbXDDwL/E0Sj2tMkJbqAMYo7ara3We+ASgFComfk8Z+684jIiuJ1xilgAJ5xGuvC1EOeICTInJumavf/g02MDVOcugBsvrMF/dbP15EsvvMTwZOAC3E06or+q0DwGrP3Av8EBivqvnAGeCcK/oPdW8EgkCBquZbnzxVnWXvsAznMMZJDrXAzSLithrt37lAmUdEJN0yw3eB11Q1CrwBPCwiWSJSTbwNdI5c4sZqAdJE5CHiNc45TgMVIuICUNWTwFbgNyKSJyIuEZkmIheKxzAIjHGSw8+AG4EO4EfAf/ZbfwpoJ17LvAzcpqoHrXU/BXKsMi8Az/fZ7m1gM3CYeAoX4Otp12vW31YR+cSa/nsgHfjc0nydeMeDYQiIeZDNWURkMfCSqpalOhaDfUyNYzDYwBjHYLCBSdUMBhuYGsdgsIExjsFggyGNHCgoKNCKiophCsVgSD27d+/2q2rhQOWGZJyKigp27do1lF0YLkJTUxO1tbU0NzejquTk5JCbm4vH4zlfJi8vj+nTpzNhwoQURjq6EJGGgUuZsWojjgMHDrB+3To+278fgAw3iEAgcvFt5s+fx/LlKygu7j+yx5AsjHFGAKrKwYMH2bjxdT755FPy0mN8r7SVq/O7yE+POyaq0Bt1E44JQnxQ2pmwmwOd2bxXu4s7f76PlXfdzbx581J6LGOFIXVHz58/X02qNnhisRgNDQ0cPnyYuro69u2t5XSzn6w05frCVr5T2EG6K/Hz0h5K47l6HycDXu6+5x4WLlyYxOhHNyKyW1XnD1jOGMc5Ojo6eOutt9jxh+20d5wBICtNmZLVzZxx3czNP0uG2975CERd/PZYGSeCWTz8yKNUVVUNZ+hjBmOcEUQ0GmXTpk2sX/cywWCIWXldfCO/i6nZvUxMj/DVozJDozviYvXRcnrTxvHEk09RWlo6PDseQxjjjBCam5tZvepZDhw8RHVeN98vbaHIG06anj/oYdXRcnImFPH4E0+Sn5+fNK3RSKLGMTdAk4Sq8t5773Hnz+/giyMHuXnyKX4y5URSTQNQkBFmRUUTrS3NPPrII3R2diZVb6xijJME2traePrpp1m1ahVFrg7unlHPwglnhy0lG4gp2QGWVxynqbGBB+6/j4aGhG5NGAaBSdWGkVAoRE1NDa+/9irhUJClRX6un9SO2yHD9Keuy8sLDWX0xNwsXnw911xzDeXl5YwfPx6327wh6kKYNo6DRCIRduzYwWuvvoK/tY3qvG5+4GuhMCO5aVkinA272XJqAh+15xOOxZe53S4mFRRQMXUac+bMYcGCBUycODG1gY4QjHEcIBgMsmPHDt58YyMt/lbKs4MsK25hRm5vqkP7E0IxoaHby+lgOu2hNPxBD18GsmkPuhCBmTNncu213+Kqq67C5/Phco3NLD5R45iRAzbw+/1s3bqVrW9vofNsF+XZQX4yxU91Xo9j7ZjBku5Spuf2Mr2PqVXhdDCd2o4cauv3sfZA/LUHGekeiouLmVhQSGFhISUlJUyePJnKykpycnJSdQgjCmOcBGlububTTz9l584/sm/fPlCozutm8bR2KnN6B2WYN5oKON6bkVDZQNRFb9RFpjuG1x1LaBtfZpC/LPMPWE4Eir0hlha3sbS4DX8wjbquLI73ptPa0c5pfx0H93roiXx1cGW+UmbNnsOsWbOoqqoasymeSdUs1qxZQ11dHZFIhHA4TDgcJhQKEQwGCQR6iUSiAHhEyUmLkOeJkCb2vrvjvRkEYok1zr1eL0uWLGH79u0EAoHEtnFF8WUGbcV2IaIqhGKuuIljLgJR9/kXuGV6vcybP5/KykrKy8vx+XyXdeeDSdUsotEojz32GHv37rW1vccVw+uK4U2PkemOke5K7Ko/XCxZsoQVK1YAUFNT46j2OdyiZLqjZLqjjLeWBWPxmlBiIT776D3ef//iP+Fz0003cfPNNzsTrEOMeuMAtq9+AozPiFGa0Ut5VoAZuT2UekNDbsf8yxEfdd1ZAxcEtm/f/rW/ieDLDHL79OO2YkuESAwae7180e3li+5MGno9lyx/udY+l2LUG8ftdvPggw8OWC4Wi9HT00NHRwdtbW20tLRw6tQpmpqaOHb0CLUn2gCYkBFl7rgzXDOx03Z382DSqEA0wM6tbzLRHcObnXgbZ7hQhc6ImxO9GdR3eznWnUl9T9b5ru3ioknM/WYVlZWVVFRUUFpaSn5+PjJSe0mGCdPGSRC/309tbS07d+6ktrYWVWVmbg/XT2pjxiA7B0YiqtAS9FDXncnx3gz8QQ9nIum0hzwE4s07RKB88mRmzZ5DdXU1VVVVo24snLmPk0Ta2trYtm0bmzfVcKbzLFdkhbihyM/svO7LykCqcDJgdUefGUdzIJ6AZHozKCkpoaBwEgUFBZSUlFBeXs7UqVPJykosxbxcMcZxgHA4zDvvvMMbG1/ndHMLZVkhlhW3MDN35N3PCUaF+p5MTgc8tIc8tAQ9fBnIojPkQkSYVV3Ntd+K3wAtKSkZ9anWxTDGcZBoNMq7777LKxvW0+JvZWZuDz/wNSd9JHQinAm72XxyIrs6xhGx2iUeTxqTCguZOq2S2bNns2DBglGXctnFGCcFhMNhtmzZwisb1hMKBrhhkp8lRakb5HnkbCYvfukjEEvjz5csOT/Icyw03u1i7uOkAI/Hw4033siiRYt47rk1bPq/P7L/bC5/O/mk4wM+95/J4vkGH8Ulpdxz732UlZkfRxhOxuZIviSTn5/PypV3sXLlSvyaz68PV/BBax5Ovab7WJeXFxp8TC6v4J8ff8KYJgkY4ySR6667jmdXrabyyio2NBbxu2M+TgXSk6rZEvSwpr6MgklFPPjQL8nNzU2q3ljFtHEcIBaLsXnzZta9/BKBQJCqvG6+Ma6LKdm9FGaEcQ1Tc6Mr4mb10ckE0sbx1NPPmBcU2sC0cUYQLpeLZcuWsWjRImpqavjD9m1saGwHwJsGFZndzBnXxbzxXQmPgO5Pb9TFvx8royPq5dFf/pMxTZIxNU4KUFUaGxs5cuQIR48eZe+eWk6eOk1mmrK4oI3Fhe2Der9aazCNtQ0+TgW93HvvfcyfP+AF03ARTHf0ZYSqcuTIETZu3MjHH39Mrkf5dkErV+efZYL13rVwTOiJuojE4nmdAp3hND7vzOb91gmIJ4O77r6HuXPnpvZgLnOMcS5TDh06xIb169ljPQbhsbpvwhfJ4ERg4cI/49Zbb6WoqMihKEcvxjiXOSdOnGDPnj20tLSgqmRnZ3/tZz5EhNzcXCorK81d/2HEEeOIyFngkO0d2KcAGPjZYKN7ueqmUvtKVR2wD3+ovWqHEnHncCMiu4zu6NVNpbaIJJRCmRugBoMNjHEMBhsM1Ti/H5YojK7RHTnaCekOqXPAYBirmFTNYLDBkI0jIreLyCER2S8iTw9HUIPUv0tEVEQKHNJ7RkQOisheEXlTRJJ6E0VEllrf71ERuS+ZWn00rxCRHSJywDqvP3NCt4++W0Q+FZG3HNTMF5HXrXN7QESuveQGqmr7A1wPbAMyrPlJQ9mfDf0rgLeBBqDAIc0bgDRr+ingqSRquYE6YCqQDuwBqh04xhLgams6FzjshG4f/TuBdcBbDmq+CPzYmk4H8i9Vfqg1zj8CT6pq0DJh8xD3N1ieBe4BHGuoqepWVY1Ysx8AyXxKbCFwVFWPqWoI2AB8P4l6AKjqSVX9xJo+CxwAfMnWBRCRMmAZsMYJPUszD/g28ByAqoZUteNS2wzVODOARSLyoYi8KyILhri/hBGR7wHHVXWPU5oXYDmwOYn79wGNfeabcOgf+BwiUgHMBT50SHIV8Yuhk+8angq0AM9bKeIaEcm+1AYDjhwQkW3AhR7u+IW1/XjgGmAB8KqITFWrvhsqA2g/QDxtGnYupauq/2WV+QUQAV5ORgznQrnAMsdqVxHJATYCd6hq0n9MVES+CzSr6m4RWZxsvT6kAVcDt6vqhyKyGrgPuPgrYIeYF24BFveZrwMKHchH5wDNQL31iQBfAsUO5cO3ADuBrCTrXAu83Wf+fuB+h47RQ7z9eKcTepbmE8Rr1XrgFNADvOSAbjFQ32d+EVBzyW2GKHgb8Kg1PYN4WiFOfdF94qjHuc6BpcDnDl0g0oBjwBS+6hyY5YCuAP8BrHL6XPaJYTHOdg78L/EBngAPA89cqvxQB3muBdaKyGdACLhFLeVRzL8CGcD/WO8m+0BVb0uGkKpGROSnxK/8bmCtqu5PhlY/rgP+DtgnIrXWsgdUdZMD2qniduBlEUknfrH6h0sVNiMHDAYbmJEDBoMNjHEMBhsY4xgMNjDGMRhsYIxjMNjAGMdgsIExjsFgA2Mcg8EG/w9lusEBb3x98wAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 234x43.2 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAM4AAABQCAYAAABRc0r3AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAASpklEQVR4nO2deWxc13WHvzPbmzcbZ0iKokxSIilSskiJErXZSZ3CgdrUQIsWSIC0/aPI4hZIgbpLmrZZ0DYIirZpAsRtWqBo06RbithNkyaxHTs2a7iWrMWi1Viy6CEpURJFSuIqzr69uf1jhvRIkSlqhjOkpPsBBN/Mve+e+zjze3d55xyKUgqNRnNn2Na6AxrN3YgWjkZTBlo4Gk0ZaOFoNGWghaPRlIEWjkZTBlo4VUJELojIz9zi/feJSHiZ8/5ZRP5smXIlIl2r1U9NeTjWugP3G0qpV4Hta90PTWXoEUejKQMtnOqyR0TeFJEFEXlKRNwi8qiIXF6sICL9IvKGiERF5CnAXdqAiPyBiFwRkUkR+fhNZYaIfFlELonINRH5exExi2WPishlEfl9EZkqtvGxmlz1fYAWTnX5MPAY0AH0AR8tLRQRF/DfwL8B9cB/Ah8qKX8M+BTws0A3cPOa6YvANmAP0AW0AH9SUt4M1BXffxz4OxEJrcqV3edo4VSXv1FKTSql5oAfUPiCl/Iw4ASeVEpllVLfBl4vKf8w8A2l1BmlVBz4/GKBiAjwG8DvKaXmlFJR4M+BXyk5Pwt8odj2c0AMvb5aFfTmQHW5WnKcAB64qfwBYELd6Gl78abywXcp2wB4gMGChgAQwF5SZ1YplbupD74V917zrmjhrC1XgBYRkRLxbAbOlZS3ldTfXHI8AySBXqXURNV7qrkBPVVbW44COeC3RcQhIh8EDpaUPw18VER6RMQD/OligVIqD/wj8BURaQIQkRYR+bnadf/+RQtnDVFKZYAPUtg0mAd+GfhOSfkPgSeB/wFGi79L+aPi+8dEJAK8hF7D1ATRgWwazZ2jRxyNpgy0cDSaMtDC0WjKQAtHoykDLRyNpgwqegDa2Nio2tvbV6kr65t8Ps/iDqTNZqPkab3mHmJwcHBGKbXhdvUqEk57ezsnT56spIl1h1KKsbExTp8+TTgc5sLYOWZn58jmrBvq+X1eNjY309HRSWdnJ729vbS0tKxbQWUyGS5dusTY2BjXrl1jenqahYUFEok4mXQaAJtNcJseTNNDIBAgFArR0NDAxo0b2bRpE83Nzdjt9ttYursRkYu3r6VdbpaYmJhgYGCA144cZmp6BoAGw6LNTNBTnyXgsLBLYcSJW3YWsgvMXJ3hyIVRXnyxIJa6gJ89/XvZt28f/f39eL3eNbuefD7P8PAwp06d4s03f8zoyChWPg+ATSBk5PHZM5g2C9OmECCvID1rYybv4KLlIJK1kcu/06bDbqe1tYWtXd10dXWxbds2tmzZgs12/834K3oAun//fnU3jziWZXHixAmee/YZ3jo7hE1guy/O7mCMHYE4dU7rtm0oBTMZJ6Mxk5Goh7djPhI5wW6z0dvby0MPP8yBAwdobGysyfUMDQ1x5MgRjh19jYVIFAE2e9Ns9cbZ7EnTaqapd2WxrWBgVApiOTszGSfTaSdXUy4mkwbjKQ/xbKEBj+mmp3cnu3fvZvfu3et61F0JIjKolNp/23r3o3ASiQQDAwM884PvMz0zS71h8d76eQ7WRwisQCzLkVdwMeHmzIKX05EAU6nCoN61tZMDBx9i//79tLe3r9qXSynFyMgIr776Kq8dOcz89QVcdujxR9lVF2OHP4HHkb99Q3dkE2YzDsbiJufjJqNxH9OpwhSusaGe/r372Lt3L319fZimuaq2q40Wzi24evUqzz//PC/+6AWSqTSdvhSPNs6xsy6+ojtwOVxLOTm94ON0xM/FuAFAfSjInv7CF6unp+eOR6N0Os3Q0BAnT57k+LGjzM7N47DBDn+MvcEoPYE4hr22rlSzaQfhqIe3o16GYz5SFtjtNh588EH27Oln9+7ddHR0rPs1khZOkVwux+DgIC+99BJvvDGIALvrory/aZ7NnnRN+xLJ2jkb8fJ21EM45iOZK6i1IRSkvXMrW7ZsYcOGDTQ0NGCaJg6Hg1wuRzQaZW5ujvHxcS5cGGN0dBTLyuO0wXZfjL5gjF11cUz76o4s5ZLLw1jcZKh4nRMJFwCm22BHTy/btm2ju7ubjo4OgsHgGvf2RtatcJRS5BcXqVXa1s1kMpw5c4bjx49z/NhRItEYfmee99TP896GBYKuyqZjq0FewWTS4FzczcW4yZW0m2spJ/llPg7TodhkpGn3Juj2Jen0Jms+spRDNGtnOGZyLubhXMLLVNLBYq8Dfh+tbW00N2+iqamJhoYGQqEQfr+fQCCAx+NZuonUgjUXzszMDMPDw4yMjDA5OcnkxGUikQiJRHJpd0cETLcbr8dDXShEKFRPff07P6FQaOmP6PP5MAxjSWhKKTKZDAsLC8zNzTExMcH4+DjD4TCjo6PkLAvDDjv8UQ6EojwYiGNf52tWS0E06+B61k4mb8NSgk0UXruF32kRcFjcxevuJZKWjfGEwWTS4ErKxVTaYDZrEMm8+8U5HXZcLhduw8DpcuJyGbhcLgy3iWEYuN1uTNPE4/Hg8/moq6sjGAzS0NBAU1MTPp9vRTfplQpn1WRsWRZnz57l9ddfZ/Dk61y5eq1gwAaNRpYNrjTtRg6PN49DFIrCXTdl2UhYdqLXxrk84eJszrG0Y3MrnI7CHDlnWdyseYcNWsw076tP0O0v3JWdtvV/R17ELhB05Qi6crevfBdj2vNs8yfZ5k/e8H42L0SydiI5B/GcnXjOTipvI2XZSOeFbN5GJi9k0zayKSGbF1J5IaIcZJSNpGUnZQmZW0wovKab1rY2Wts209nZydatW+no6MDpdJZ1DRULJxwO88orr3D0tSMsRKI4bdDljfPQAwk6fUkecKdx3OE2fzYvLGTtRLIOojkHCctGPGcnmxeySgqB9aJw2hR+h4XfkaPJnaXelV33o4rm3XHaFA1GjgajshtHLi/EcnYiOTvXMw7mitvp1ybnOTY2zMBA4UvidDro6uqit3cnfX19bN++8hjAiqZqTU1N6pFHHsFpK2x/9odi9ATiuO6iu7zm/kIpuJ51cCnhZizu5nzcw3jCQAGGy8m3nnq6+lM1m5XmV9uusTsYw71OdnQ0muUQgZArR8gVY3cwBhTWXOdiJuGoh2+tsJ2KfCWa3Bkeaoho0Wjuakx7np11cT7UOr3ic7Sv2k1853IjE0ljrbuxYlKWjaRlw7Tn1/QG1mKm+WDrzJrZrzUVCWcq5eKrIy2r1Zd1wUTSIJVf30+3S3G73Rz6wCEGBgZIxVNr1o+JpHFX3XDenTMrqnX/ubXeYxw6dIjHH3+cQ4cOrXVX7isqGnGa3Bme6L63kkh+daSFc3HPWndjxQwMDNzwe61oMdP3xHfhyyusp9c4N9Fi1tZ/rVJSVoqjP/ouDfY8bu/arnHuJ7RwbuJ+WuBqyqfizYHXZgLsCcZWPeZDo6kV8ZyN0ZiHcHTlU/SKhJO3Gzx9eSPfmdzIg74Y/aEovYE47rvAY1dzf1IIwnNyKWEwFjcZS3iYSLhQgNtwrbidir2jn3rqKV555RWOHH6V+esLOGzQ6U3woD9OhzdFq5m+Y0fLtCVczzqKvmp2EpadRM5GJm8jqwp+Rg5RuGwKnyNHwGmxwcjS6MrcsV+c5t5CKcgUfdWiOQfzpb5qaYOraWMpDspwOdm2bTu9O3eya9cuuru7cTqdtQ0ryOfzDA0NMTg4yMnXT3B5YhIoJIZoNCw2uJL4HRYeh4Wz6B1tKVnyjo7l7EQsFwtZB6l38fGz22w4HA5EIJezyFk3usHaBFrMDB2eeNH7NqH95u4i0pYQyTlYyBa8oxPFh7sF72jbO97ReSGrbO/8VnbSRS/qZE6wbvGR+31e2to207a54B3d2dnJli1bfiLOZ83jcebn5wkXY2MmJye5MjlBZGGBWCJBNptDpBDI5jFNvF4PwWA9wVBoKRZnMaApGAwSCATwer0/4QKezWaJRCLMzc0xOTnJ+Pg44fDbDA8Pk83mcBbDifeHIvQE4ut+NMoVvcIXsg7SJfE4PoeF32FR58xVLcS7lsRzNsYTbiaLyT+m0gZzWRexZcJJDJcLt+HCZRgYRiEWx2UYuFyFWBzDMDBNE5/Ph8fjIRgMLsXjbNiwAY9nZeuXNRfOWpLNZjl79iwnTpzg6GtHuL4QwedUHAzN80jjAvXrIN7FUnA5YXA+bnIx4WYyZTKdeicy8lYYdmh2p2n3FGKNunzJu8JP8HrGzkjMw2jM5HzincQeAPWhOlpaN9Pc3MzGjRuXAhgDgQB+vx+v14thGDVLQXVfC6cUy7I4deoUAwMvceLE66AUu+pivL9png5vbV1UrmccvBUpJLQYKSa0AGja0EhH51Y2b968FD5smiZ2ux3LsohGo8zOznL58mUujI0xPFIYUR026PbG6QvG6KuL4V0nO5vZvHA+bnI2Usg5cDVZmCl4PSY9vTvZvn073d3dtLe34/f717i3N6KFcwump6d54YUXeOH5HxJPJNniTfNo4xx9wVhVAuCUgisp11KWm8vFpBUbGurZs3cffX197Nixg/r6+jtqN5PJEA6Hl7LcTE3PYBN40B9nbzDKzrpYTXc2F3PLvR3xMBT1Mhr3krEK0bo7enqWstzcDckLtXCWIZlM8vLLL/OD73+Pa1PThIw8D4fmeai+8kQelipkeFnMqzabtiMC3d3dHDz4EAcOHKC1tXVV86qdP3+ew4cPc/jV/2V2bn5pbbezLkZvIL7qI1FewUzayfmlvGpe5tKF6dfGpkb27jtAf38/O3fuxO12r6rtaqOFswIsy2JwcJDnnn2WN0+fxibQ5Uuwu64QyRpawVoor2A67WQk5mEkajIcL6R9ctjt7Orr4+DBgxw8eJBQKFT168nn84TDYQ4fPszxY0eZm7+OAC2eDF3eOG2ewuOBRmNlIeZ5BZGsg9mMg6m0aymT5+WUubSl6/d66Nm5aymT56ZNm6p7kVVGC+cOuXr16lLu6MVEI0FXns1mgpArS50zh10K05KEZWMh62A6bTCZci+tVepDQfr37mPfvsI0bKU7OdVAKcXo6CinTp3i9JtvEg6Hl7bvBQgaeXz2HG5bDpetMCLllZDO20irQq6HWFZuSFflcjppbWulq5g7evv27bS0tKz76dedoIVTJkopLl26xJkzZwiHw4ydP8fMzCzpTGapjogQDPhpam6ms7OQLaW3t5fm5uZ1mzc5l8sVExpe4Nq1a0xNTRXSdcXjpNMpRASRxf9WYFJXV0eo+HigubmZ5uZmmpqa7imR3AotnFVEKUUymVxKpLi446W596iJcEQkCoTLbqB8GoG1cGPWdu9929uVUrfdI680rCC8EnWuNiJyUtu9d+2upW0RWdEU6t6esGo0VUILR6Mpg0qF8w+r0gttV9tdP7ZXZLeizQGN5n5FT9U0mjKoWDgi8oSIhEXkLRH5q9Xo1B3a/5SIKBGp/n+nLdj7koi8LSJvish3RaSq/1JMRB4r/n1HReTT1bRVYrNNRF4WkaHi5/o7tbBbYt8uIqdE5Jka2gyKyLeLn+2QiLxn2ROUUmX/AO8HXgKM4uumStorw34b8AJwEWiskc0PAI7i8ReBL1bRlh04B3QCLuDHQE8NrnETsLd47AeGa2G3xP4ngf8AnqmhzX8Bfr147AKCy9WvdMT5TeAvlVLpoginKmzvTvkK8IewbPzXqqKU+pFSatH78xjQWkVzB4FRpdR5pVQG+BbwS1W0B4BS6opS6o3icRQYAmqS61hEWoGfB75WC3tFmwHgp4F/AlBKZZRS15c7p1LhbAPeJyLHReQVETlQYXsrRkR+EZhQSv24VjZvwceBH1ax/RZgvOT1ZWr0BV5ERNqBfuB4jUw+SeFmWMuovE5gGvhGcYr4NRHxLnfCbT0HROQloPkWRZ8rnh8CHgYOAE+LSKcqjneVchvbn6UwbVp1lrOrlPpesc7ngBzwzWr0YbErt3ivZqOriPiA/wJ+VykVqYG9XwCmlFKDIvJote2V4AD2Ak8opY6LyF8Dnwb++F3PqHBe+DzwaMnrc8CGGsxHdwFTwIXiTw64BDTXaD78EeAo4KmynfcAL5S8/gzwmRpdo5PC+vGTtbBXtPkXFEbVC8BVIAH8ew3sNgMXSl6/D3h22XMqNPgJ4AvF420UphVSqz90ST8uULvNgceAszW6QTiA80AH72wO9NbArgD/CjxZ68+ypA+PUtvNgVcpOHgCfB740nL1K3Xy/DrwdRE5A2SAj6ii5XuYvwUM4MVi7M0xpdQnqmFIKZUTkd+icOe3A19XSr1VDVs38VPArwGnReT/iu99Vin1XA1srxVPAN8UEReFm9XHlqusPQc0mjLQngMaTRlo4Wg0ZaCFo9GUgRaORlMGWjgaTRlo4Wg0ZaCFo9GUgRaORlMG/w8XLuxB6Dx3AQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 234x43.2 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 234x43.2 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 234x43.2 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 234x43.2 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 234x43.2 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 234x43.2 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAM4AAABPCAYAAACj8/q5AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAAMxklEQVR4nO3de3BU53nH8e+zV0loJe1qpV0jAQJZAsylYGyNweYmwBKOXcB4xlPbONM4ddxpbZO004mT6UzdaadJ80eSSSfJdFI7XBIcJgScmLp2ocR4nBgHWSIm5mJDuApJ6Iru0u6+/WNXjjCgy1l0VljPZ2aHvZxznvcg/fY95z0XiTEGpdToOFLdAKVuRRocpSzQ4ChlgQZHKQs0OEpZoMFRygINjlIWaHDGKRH5CxGpEpE2ETklIitS3Sb1J65UN0BdS0T+DvgCsAmoAeYA7SltlLqK6JkD44uI5AGngKXGmCOf+uxfgWVAPfCkMaYrBU1U6KbaeLQa+OA6oZkLFBtjlgL7iPdIKkU0OONPAGi9zvtLgdcTz18H7rOtReoaGpzxpxq4T0T+TOJKRGQ24AfaEtO0EQ+YShENzjhjjPkN8C/Aa8QHBHYD6UALkJ2YLBtoTkkDFaCDA7cMEZkHvGCMeUxEnga8xpjvpbpdE5UOR98ijDEfiMhZEXkbaACeTHWbJjLtcZSyQPdxlLJAg6OUBRocpSzQ4ChlQVKjasFg0BQVFd2kpiiVelVVVY3GmLzhpksqOEVFRRw+fHhE09bW1lJdXU1dXR0Oh4NwOMysWbMoKipCREZV1xjDpUuXqKurw+fzUVRUhNvttrIKSl1FRM6OZLoxP45z4cIFtmz5MVVV7wOQ5oSogf5Y/HN/TjZL7r2P8vJypk+fPuzyjhw5wrZtWzl9+o+fvJc5KYNVq9fw0EMPEQjomShq7CV1HOeuu+4yN+px+vv72bVrF7t2/RyPRFkRbObuwBUCngjGQEu/i4870jnalsmH7ZlEYjCztJR169dTVlaGw3H17ldHRwcvv/wyBw4cINcbZXmwiSkZvbT1u6hpzeRIqw+X20VFRSUbNmzA7/dbXi81cYlIlTHmrmGnG4vgHDt2jB/+4Pucv3CRRf4rbChoJNMVveFyuiIO3mvO4u2mAE29Tm4Lh1hzfwXz5s3D4XBQU1PDL1/dTXt7B6vym7k/1IzbcXW7G3vdvFkf4HBLFk6Xi8rKtWzcuJGsrCzL66cmnpQEp7a2lh07fso77/wGvzfGI5PrmJPdOeLlRQ0cac3kYGOAM53eqz4ryexm3eTLFGb0DrmMy71u3qyLBygtzctjjz9BZWUlTqdzdCunJiRbg9PU1MSOHTv49a8P4BLDimATq/Jb8DqtL7up18WFbi8GoTC9l6C3f1Tz1/W42XMxn+PtGcwsLeX5zZsJh8OW26MmBtuC8+KLL7J921aikX6WBFpYE2rB577xZpmdjIGqFh+7akMYp5envvhXlJeXj3oUT00ctgSnsLDQlJWVMTurk0cKGsj1Riwvayy19LnYfi7MqY505s+bx6Ynn6S4uDjVzVLjkC3B8fv95ruPz2V5Xivj/Us8ZuCdxmxer8+jKyJML5rGgoV3UlxczIwZMwiFQtoTKXuCc3so0/zf3wx/7GU8GRjBO9Lm42xXGrHE6vuzsyi7ZzHl5eWUlJSktpEqZWwJzvyCdPPal263PH+q9ceEuh4P57q8fNSRwYftPvqiMHv2LDZufISFCxdqLzTBaHAs6I46ONSUxVtNubT0Opg6pZCKyrUsWbKE7Ozs4RegbnkanCREYvB+q4+DjQEudHkQgeLiYubMmUtpaSklJSUEg8FUN1ONgZEGx5Z7DvziQpCL3d7hJ7yOnqiD7qiDdGeMNGdsVPMWpPfycGHjqGu6HFAWaOdufzu1PR5+35rJyUs97D19ilcHnWNXOnMWxcXFTJ06lVAoRE5ODpmZmXqwdQJIqscJ52abTQ8sHna6i91eemLWfpnS0tJYtWoV+/fvp6enZ3TzOqIUpA99psFoGIS+mNATdcQfMScRc+0+kIjgcDhwOZ243G7cbjcej4eCggIeffRRgsEg2dnZGrBxaFz1OMlYtWoVTz31FAB79+5NaVsEg9dh8DpiZCeuYoglwhSJCVETfxggaoRYVIhEeujpir/f0NBAdXX1DZe/bOlSNn/5y/asjEpKUsHJT+vj2ZKLw073vY8KONWZYanG/v37r/p3NArSe0fUvmR0RhzU93ho7nPTHnHSneiNuqNOOqNOrkTctPa76RjBGUMO7YFuGbb0OMlsLvVEe/jtm7vJdcZImzT6fZybqS8mXOjycrYrLf7onkRL79WXP4hAmtdLRkY6Pn8WeYFcZgeDhEIhwuEwoVCIvLw8fD6fDnXfwmwJjpUd9PHAGDjf7eWDtkxOdkzifJf3kwOmebkBZi+6enDA7/eTnp5+zbVE6rNn3O/jpEIkJvyuJT4cfanbjcMhlJSUsH7QcLReKDexaXAG6Yo4eLc5i7cac2nrc1A0bRrPrF3L4sWL8fl8qW6eGkcmdHD6YkJtt4fzXWmc7MjgeHsm/TGYN3cuD2/cyPz583U/RF1XUsHpitx6o0CdEQeHmrOoac3ifJeXgaNYuQE/qyvuYfXq1SO6aYia2JIKTlOfm/31fsrzW26JywoOXs7hjYYg3RHh9tuL2bhgIcXFxUyfPp28vDztXdSIJRWcSZMm8atLQU52ZPBIYQN5o7y82S5NvS62n7uNP3amsXDBAp7YtEl7FZWUpIITDod5+umn2brlx3zjeAaLc1tZE2omexxdOv1es4/dl0KIK43nnvsSy5cv155FJS3pwYHKykrKysrYuXMn+/ft41BzDsuCzawJNZOWxM06Lve6E/sgQmF6D6G00fVmtd0e9tTmc7I9nTtmz+K55zeTn59vuT1KDXZTbw9VV1fHK6+8wsGDB8n2xNg4uY75OaO7PVR1i4+3GgOc7/Jc9dmMzB7WT25g6jC3h6rvid9f7f2WLDIy0nns8SeoqKjQg5JqRFJ6Q8ITJ07wwx98n7PnzrMgp52HCy6TNcTmW8fADQkbA7T0OSmcPJk1FRXMnTsXp9NJTU0Ne3b/gra2K6zMb6Yy3IznUzckbEgEpqolC4/HzdoHPseGDRv0+IsalZQGByASibBnzx52/uxnuIiyLNhEWaCdXE8/hviI3MAtcI+3TyJqYM4ds1m3fgOLFi26Zj+ks7OTLVu2sG/fPgLeKEtzmxO3wHVS0+rjaFsmHo+byrUPsG7dOnJyciyvl5q4Uh6cAbW1tWzbupX3fvcexoDbER8ajibKBnMDLLn3PlauXMm0adOGrXn06FG2b9vKyY8+/uS9LF8mq9fcz4MPPqiBUUkZN8EZUFdXR3V1NfX19TidTsLhMDNnzmTKlCmWRrnq6+s/+TMf06ZN04vC1E1hS3BEpB04YXkB1gWBVJxyrXU/+7VnGmOG3TFOdjj6xEjSebOJyGGt+9mtm8raIjKiTSgdo1XKAg2OUhYkG5z/vCmt0Lpad/zUHlHdpAYHlJqodFNNKQuSDo6IPCsiJ0TkDyLy7zejUaOs//ciYkTElnvSisi3ROS4iPxeRHaLyJgecRWRysT/78ci8tWxrDWo5hQROSAixxI/1+ftqDuovlNEqkXkNRtr5ojIzxM/22MiMvSdNo0xlh/ASmAf4E28zk9meRbqTwHeAM4CQZtq3g+4Es+/CXxzDGs5gVPADMADHAHusGEdbwPuTDz3ASftqDuo/leAnwKv2VhzC/DFxHMPkDPU9Mn2OH8NfMMY05sIYUOSyxutbwP/ANi2o2aMedMYM/Cn594FCsewXBnwsTHmtDGmD3gFWDeG9QAwxlwyxryfeN4OHAMKxrougIgUAp8DfmRHvUTNLGAZ8F8Axpg+Y0zrUPMkG5xSYKmIHBKRt0Tk7iSXN2Ii8ufARWPMEbtqXscXgNfHcPkFwPlBry9g0y/wABEpAhYCh2wq+R3iX4aju/tkcmYAl4GXE5uIPxKRSUPNMOyZAyKyD7jen2v+emJ+P3APcDewU0RmmER/l6xhan+N+GbTTTdUXWPMq4lpvg5EgJ+MRRsGmnKd92zrXUUkE9gFbDbGXLGh3oNAgzGmSkRWjHW9QVzAncCzxphDIvJd4KvAP95wjiS3C/8HWDHo9Skgz4bt0XlAA3Am8YgA54CwTdvDnwd+C2SMcZ3FwBuDXr8AvGDTOrqJ7z9+xY56iZr/RrxXPQPUAV3AdhvqhoEzg14vBfYOOU+SBZ8B/jnxvJT4ZoXY9R89qB1nsG9woBL40KYvCBdwGpjOnwYH5thQV4CtwHfs/lkOasMK7B0ceJv4CZ4A/wR8a6jpkz3J8yXgJRE5CvQBnzeJyp9h/wF4gf9NXA7xrjHmmbEoZIyJiMjfEv/mdwIvGWP+MBa1PuVeYBPwgYjUJN77mjHmv22onSrPAj8REQ/xL6u/HGpiPXNAKQv0zAGlLNDgKGWBBkcpCzQ4SlmgwVHKAg2OUhZocJSyQIOjlAX/D5fqyl1piNv/AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 234x43.2 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 234x43.2 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "violin_creator('bias_r', \"reset\")\n",
    "violin_creator('bias_z', \"update\")\n",
    "violin_creator('bias_h', \"hidden\")\n",
    "\n",
    "violin_creator('weight_xr', \"reset\")\n",
    "violin_creator('weight_xz', \"update\")\n",
    "violin_creator('weight_xh', \"hidden\")\n",
    "\n",
    "violin_creator('weight_hr', \"reset\")\n",
    "violin_creator('weight_hz', \"update\")\n",
    "violin_creator('weight_hh', \"hidden\")\n",
    "\n",
    "violin_creator('c1', \"$c_1$\")\n",
    "violin_creator('c2', \"$c_2$\")\n",
    "violin_creator('c0', \"$c_0$\")\n",
    "violin_creator('bias_rw', \"right wheel bias\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "COLORS = ['#ffffff', sns.color_palette()[0], sns.color_palette()[1], sns.color_palette()[2], sns.color_palette()[3]]\n",
    "\n",
    "def compare(h, t):\n",
    "    plt.figure(figsize=(3,2.5))\n",
    "    gs1 = gridspec.GridSpec(5,1)\n",
    "    gs1.update(wspace=0, hspace=0)\n",
    "    for i in sorted([i for i in range(5)], reverse=True):\n",
    "        plt.subplot(gs1[i])\n",
    "        if i == 0: plt.title(t)\n",
    "\n",
    "        # Format data\n",
    "        data_temp = df_g.copy()\n",
    "        \n",
    "        # Drop the outliers first\n",
    "        q1 = data_temp[h].quantile(0.25)\n",
    "        q3 = data_temp[h].quantile(0.75)\n",
    "        iqr = q3 - q1\n",
    "        data_temp[h] = data_temp[h][((q1 - 1.5 * iqr) < data_temp[h]) & (data_temp[h] < (q3 + 1.5 * iqr))]\n",
    "        \n",
    "        if i == 1:\n",
    "            indices = (data_temp['c1'] >= 0) & (data_temp['c2'] >= 0)\n",
    "            data_temp = data_temp[indices.values]\n",
    "        elif i == 2:\n",
    "            indices = (data_temp['c1'] >= 0) & (data_temp['c2'] < 0)\n",
    "            data_temp = data_temp[indices.values]\n",
    "        elif i == 3:\n",
    "            indices = (data_temp['c1'] < 0) & (data_temp['c2'] >= 0)\n",
    "            data_temp = data_temp[indices.values]\n",
    "        elif i == 4:\n",
    "            indices = (data_temp['c1'] < 0) & (data_temp['c2'] < 0)\n",
    "            data_temp = data_temp[indices.values]        \n",
    "\n",
    "        # Create the plot\n",
    "        sns.violinplot(data_temp[h].dropna(), color=COLORS[i])\n",
    "        if 'bias' in h:\n",
    "            plt.xlim(-3,3)\n",
    "        else:\n",
    "            plt.xlim(-6,6)\n",
    "        plt.yticks([])\n",
    "        plt.xlabel('')\n",
    "        if i < 4: \n",
    "            plt.gca().set_xticklabels([])\n",
    "    plt.savefig(f\"delete_me/comp_{h}.png\", bbox_inches='tight', pad_inches=0.02, dpi=500)\n",
    "    plt.savefig(f\"delete_me/comp_{h}.eps\", format='eps', bbox_inches='tight', pad_inches=0.02)\n",
    "    plt.show()\n",
    "    plt.close()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 216x180 with 5 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "compare('weight_xh','input-to-candidate')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 216x180 with 5 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 216x180 with 5 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 216x180 with 5 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 216x180 with 5 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 216x180 with 5 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 216x180 with 5 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 216x180 with 5 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 216x180 with 5 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 216x180 with 5 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAALkAAAC3CAYAAABOp/8sAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAAgAElEQVR4nO29aZAk53nf+Xszs+6qvu9j+p6rpwczAOYAQPCCCUMGSXkZIUqWJVs2QckR1sr+tGvZqw+OXcfGamM3NkL7aRUiY9eQJZoiKIqESQIicQ0xAObAHH3NdM/0fZ/VdVce737IypzuxhzdPX1U99QvIqOqK7Mz36z655vP+7zP86SQUlKgwEFG2esGFCiw0xREXuDAUxB5gQNPQeQFDjwFkRc48BREXuDAUxB5gQNPQeQbQAgxLIT4B/f5/EUhxK29aFOBjaPtdQP2M1LKD4Aje92OAg+n0JPvE4QQhQ5pixREvnHOCCF6hRBLQojvCiH8QogvCiHGnQ2EEP9OCHFHCBHLbfvfrVrXLoR4TwgRFULMCyG+96gDCiGkEOJfCyEGgAEhxH8UQvxZbp1HCJEQQvxp7u+AECIthCjdgXPf1xREvnH+KfAPgTbgMPA/3WebO8CLQDHwH4HXhRC1uXX/M/AWUAo0AH+2weP+Y+AccBx4D/hi7vMzwDTwhdzfzwG3pJRLGz6jJ4SCyDfO/y2lHJNSLgL/Cfgn6zeQUn5fSjkppbSklN8DBoCzudU60ATUSSnTUsoLGzzu/yqlXJRSpoCLQIcQohz4PPAXQL0QIowt9vce6wwPKAWRb5yxVe9HgLr1Gwgh/pkQ4poQYlkIsQycACpyq/8HQACfCCF6hBD/crPHzQn9MragP48t6g+BFyiI/IEUBjMbp3HV+0PA5OqVQogm4M+Bl4CLUkpTCHENW9hIKaeBb+e2/Rzw90KI96WUg4847vpY6PeALwOngUu5v/8h9h3j/S2c14Gn0JNvnH8thGgQQpQB/x5YP3AMYQtyDkAI8S+we3Jyf/+GEKIh9+dSbltzC+14D/hnQK+UMgu8C7wGDEkp57awvwNPQeQb579gDxzv5pb/ZfVKKWUv8H9g280zQBfwq1WbnAE+FkLEgb8D/o2UcmgL7fgQCHCv1+4F0hR68QciCplBBQ46hZ68wIGnMPDcQ4QQLwI/vd86KWV4l5tzYCmYKwUOPAVzpcCBZ1PmSkVFhWxubt6hphTYLgzDwDRNTNNESsnqu7WiKO6iqiqqqu5hS7ePK1euzEspK++3blMib25u5vLly9vTqgLbxvLyMiMjI4yOjjI7O4tp3nO/ezweNM3+mQ3DQNf1Nf8rhCASiVBcXLxmKSoqIhwOoyj742YvhBh50LrCwHOfEovFGBwc5M6dOywuLgJQVlbGsWPHqK6uprS0lKKiIlfgDqZpkkqliMVirKyssLKywvLyMisrK0xNTWEYhrvt+gugrKyM8vJyysvL9434oSDyfUUmk2FoaIiBgQGmpqYAqK6u5vnnn6e5uZlw+NEOGVVVCYfDhMNhamtr16yTUpJMJl3xR6NR93VyctK9Q6iqSmVlJfX19TQ0NFBZWZnXot+Ud+XZZ5+VBXNld0mlUoyOjjI0NMT4+DiWZVFcXExHRwft7e0UFRXtSjssyyIWizE/P8/s7CxTU1PMz88DEAgEaGpqorW1lbq6uj0RvBDiipTy2futK/TkeYSUklgsxsLCAtPT00xPTzM3Z4ejhEIhOjs7aWtro7KyEiHErrZNURTXbGlrawMgnU4zPj7O8PAwd+7cob+/n0AgQGtrKx0dHXvSzvtR6Mk3STabdW/j8XicZDJJKpVC1/U19qzjvdA0DY/H4753ejnLsjBNk0wmQzqdJh6PE4vF3IGhqqpUVVVRX19PY2MjFRUVeSGYB2EYBqOjo9y5c4fR0VFM06SoqIj29nZaW1spKyvb0eM/rCcviPwBZLNZlpeXWVpaYmlpicXFRZaWlkgkEmu20zSNQCDgejEcITouPMMw3MU0TSzLAuxBnaqq+Hw+/H4/oVCISCSyZnC3X9172WyWoaEhBgcHmZycREpJSUkJTU1NNDU1UVVVte0mTUHk65BSkkqlSCaTJBIJEokE8Xjc7U1XVlZIpVLu9qqqUlpaSmlpKSUlJZSUlFBUVEQkEsHr9W7p+PncK28nyWSSoaEhhoeHXcF7vV5qa2upra2lurqa8vLyz3iBNssTZ5NblrXGSxCLxYjFYq6Qk8mk26M6CCEIh8NEIhEOHTpEcXExJSUllJaWEolEtrXneVIEDhAMBuns7KSzs5NMJsP4+DgTExNMTEwwMmK7toUQbudRXFxMJBIhFAoRDAbx+/14vV48Hs+Wv7ddF7lhGGQyGTKZDNls1rVlnVv56juLEMIV12qROfasrutks1nS6bTbMztCXr+fUChEOBympqaGcDhMMBgkFAq5SyAQyGs32EHA5/PR1tbmDlyTySQzMzPMz8+zsLDA4uIiw8PD3M+6EEK44xtn8Xq9eL1eAoHAQ4+7aZE79qWu665AHbFms1my2awr4nQ67b46y+rB2Xbh8Xhcu7aqqoq2tjYikQiRSGRPZu4+/PBDFhYWdu14q3F+A0cAO0V5eTnPP//8Y+0jGAzS0tJCS0uL+5lzF47H427n5ZyTozlncZwA09PTDz3OpkQ+NzfHd77znQ1t6/TCqxefz0cgEFizTgjhLs7/AZ+5mtf3zMCa/3dIJpNuD7FXLCwskM1m9+TYmqZx9OhR+vv7d6RDcVhYWNizC9lBURT8fj9+v/+h223JXHHcY+sDfVYvT5LdmU8cPXrU7WG7u7v3uDX5waZE7gjXsqzPDNw+s2NNc91j6xefz+feTr1eL5qmuRfN6ovDiaB7UCSdY5etdt3lAz/+8Y/daffdpr+/f83rTlFeXs7Xvva1HT3GdrHZUFtee+011yZfbZs7NpNjj6+2y9PpNLFYzP1su1EUhUAgQDAYdOMyHHvcsckf10W1GcrLy3ftWOvJZrOMjIxQXFy84zb5TmCaJvF4nEQiQSqVcsd12Wx2jd4czTkaexib/uUVRXmsQY1lWW7jVntXDMNASrlmsmT9snofzkTLau9KIpFgcXHRnXFbzeoLYLVXxXFVhUKhbZt8edwB2ZNCPB5f411ZXl4mHo/fd1tVVV2viuNl8fv9FBUV7YxN/jhsdLDwODiTPat95CsrKyQSCRYWFhgZGfnMRQD2heDcARw/ueO73a+zj/lEOp1mbGyM8fFxJicn3dljRVEoLS2lpqZmjZ88EAi45u3jfP8HcjJICEEwGCQYDFJTU/OZ9VJKstmsO9u5fsZzcnKSgYGBNftzJoacWU9nxtPn8214PLB6fLF+/HFQicfjDA0NMTQ0xMzMDFJKfD4f9fX11NTUUFNTQ2lp6Y52IgdS5I9CCIHP58Pn8z0wcEjXdaLR6JrYlbm5Oe7evbtmO+fOtH4A/KDYlfu1w+/3u+OIsrIyysrKqKys3NVxxHaSTqe5e/cug4ODrg+7rKyM06dPc+jQoV2PTtyf3+Iu4PF4qKiooKKiYs3nhmEQjUZdMyiVSq2JQnS8QI63yLEhV3uQwO7VnSjEVCpFPB5ndnaWvr4+wL4AKisraWhooLGxkaqqqrzu+XVdZ2RkhMHBQTfuvaSkhGeffZbW1lZKSkr2rG0FkW8STdPcKMHtxsnMmZ+fZ2ZmhqmpKT799FOuXr1KMBikubmZ9vZ2qqur80LwqVSKsbExhoeHGRsbwzRNgsEgJ06coL29nfLy8rxoZ0HkeYQTYxMKhWhqagLuJSbcvXuXW7du0dvbSyQSob29nY6Ojl3rIS3LIhqNMjc3x9zcHFNTU25uaTAY5MiRI7S1tVFTU5MXwl7NExlqu1/JZrMMDw8zMDDghq1WVlbS1tZGc3PzY6fCWZZFIpFYk9/pvI9Go657V9M0qqurqauro6GhIS8SOgrx5AeQRCLBnTt3GBwcdHMtS0pKqKurW5Otv34+Q9f1Ndn6joAdUa+eyVYUZY071UnoKCkpybuIzYLIDzgrKyuMjIwwNjbGzMzMmtoqzuAXuK+HZ3XuplNvxXkNhUJ73kNvlCcuaeJJo6ioiK6uLrq6urAsi+XlZbeWSiaTQdd1N93O7/e7s7/OpEu+9crbzaZ6ckVR5KMC1LcTwzD2ra/4Uez0ua0PbHNeV4dI7KS4d/u3SyaTUkp53xPaVCuefvrpXS0T9+yzzx7YsnTbfW7Ly8t89NFHfPLJJ1y7foPFhflH/o8Qgrr6ejqPH6ezs5OnnnqKpqambTFRdvu3E0JcfdC6g9lNPiFks1k++OADfvazn3PlymUsy0J4g2TD1ZgNLVjBUqQnhPT4kcLu5ISZRegplEwMJR1lOL7I5LsXeOuttwAoLinl7JlnOXPmDGfPnt3TSZztoiDyfcjCwgI//OEP+bsf/4SV6DL4wmSqTmCUtWAFy+AhPbHUfEhfBCtc5X6WlhKRiaHFptFXJvn7dy/w9ttvI4Tg8JEjPP/cczz33HN0dHTsm4HoavJa5L//+7+/103YMbZybmNjY/zVX/0VP3/rLUzDwChpJHv4LGZR/UOF/UiEQPqL0P1F6JWHSUuJkphHi47TNz7Ore9+l+9+97uUlJbx/HPnOX/+PM888wyhUGhbz2+nKLgQ9wEDAwO8/vrrvPf++wihkilvJ1tzAunfnTqIQk+hRsfRlsfxxiaRRgZFVTnZ1cW5c+c4d+4cLS0te9rLF/zk+xApJdevX+cv//K/cOnSJwjNS7ryKHp1J9Kzex6uz2BZqIlZ1OUxvCsTiKQ9tV9aVsa5s2c5c+YMzzzzzK7b8gWR7yMMw+D999/nr7/3PW7fuoXwBkhXHidbdQy0nUtn2yoim0CNTqBFJ/DGJ5G6nd7Y0trKs888w+nTpzlx4sSOV98tiHwfMD09zU9/+lN+9Hc/ZnlpEQLFpKs60SvaQcnrodM9pIWSWEBbmUCLTaHGZ8GyZ1gPNTVxorOT48eP097eTktLCz6fb9sOXRB5nrK8vMwHH3zAL995h08//RSkxCiuJ1t1HLO44fEGk/mAZaAm5lFj06jxWTzJeaRuJx0LIaiqrqG56dCamohOeEEgECAQCODz+dbE5D9oAqswrZ8nJJNJ+vv7uX79Oh9/8gm3+vvtmchAMZnaU+gV7UhfZK+buX0oGmakBjNipyCmcq5KNbmIklpkPBVlunsQ5eo118x55C6dkti5pGafz7+9ZeJmZ2f50z/9U1Kp1L2yE7kSXqZhYpgmlrTAuTsIgbaqRrfP63XTzvx+v3u13m9ZX59lddEiIYRb+8UJOlpdSsypBOAsTimD9fUWnUpeTvucYz2qXoyT5bMawzBIp9Mkk0k3qm92dpbp6WlGR0cZvHOXifEx+9hCYIUq0Wufwihtwgo83Ld9YMi5Kg1/EdC8dp2pI/QkQk8jzAzC1BGWYZs7lomQFuQW+33uc9OA6MMrhW1K5EvLy/z0F+8hFQ9SUbGEihQqCAWEghTCfo/zg0mQBkJmQZqIXGMVaeZOwEAaWddu208IIRBOKtu6QqXrNgR/BN1XglX7FGa4CjNUCdrj26O+0Y9Qct6Nh2JmEUYWqXlB3b7BqxUsI3Po/PbsTPUg1WKkv3iLO/j/HrhmUyK3guWsnPzNLTbiYTu2wNJzV6+eu0L13IWx+mo2c3cJee/CUrTcqwpCBSV34Smqu87+TLl3Aa66Bu0L0bL3bdkXIpbhXoTCNOypcCONtjiE0JN2G6TM7UYiVS2337VtundcgTCztm0ae3hxys2gJhfs7+kR+P1+vvr1r/KTn/yE9EYuig0ikwsbu8j2mM2VicvG8Y1+bP+IipZ7XSUgoeQEJLDFA66ILDP3auTEZLjCxjQQli1w1TIR0gAz19Obhn2benjLEIoKqi1uhIbMiV0KBWt1+xBIIRBu+3K3QctCkRYK94SOqSNN/Z759cCj3+9DBelcfO535cldAJv51h+fr371q/zhH/4hUkp+8IMf7O7B84BNiVyTBpHoHbKZzINvzxvE4/Hg89u2dzAcJBQsJhQKuba6Ywc7BdidkfXqbHfHJl9fQtoZL6wuX5fNZjFz2+MMGRQFVbVtcl/uOKvHDE4V3tV/O+1Z/TRjp/REOp1208eWlpaYm5tjYnKS6elprKxtkgnNhx6qwCiqxyyux/KXbNkeD/T/N7QN3Bl+8pOfIKXkzTff3NJxHoQZLCd19B9t6z5dpJW7o9p32Hu2uNMp5e7subvtw9iUyNvb2vj5z37mFudZvax+zLVlWe4A0RHD6sLpfr//wAfqr0bXdcbGxrh9+zY9PT18eu0a42OfwBi2Z6WkCaO8HSuwuVlCK1jGRoozx80s//XHP0dqQYhs30ykFdziw64sCyUTRUlFUdJRRDaOko2jGmkUIw1G1r6DbhMFP/keMTMzw0cffcS7773HtU8/tTuHcBWZisMY5a37ZwLoUUiJyKygxmdR43NoyXmU1NIaZ0OkqJjq6ioqKyooKSlZUybO5/N95gl6q71hPp/PKddRmAzKZxYXF3n77bf5yU/eZGxsFOHxk644jF51HOkN7nXzNocTtrsyiboyhTcxg8wmAQgEghw7dpTDhw/T1tZGU1MTDQ0NBIOPf46FGc99gpSSa9eu8cYbb3DhwgWkUMiWtZGt7XoM19ouYGTQVqZQV8bxxqYgHQPsoK1nnn6ap556iq6uLg4dOrRjZmphxnOfIITg9OnTnD59momJCb7//e/z5ptv4p2/jV7aTLb2KazQ3tU+d5ESJbmAFh3HszKBEp8FKfEHAjzzzNOcOXOGp59+msbGxrxIsij05HnO4uIiP/jBD3jjjR+SSiUxihvI1nTZU+W7KSAzixadRIvaIbaOCdLe0cH5c+c4e/Ysx48f37PE84K5cgCIxWL86Ec/4nv/9fvEVqJY4UoyVccxSpvtuYHtRkqUdBQ1Oo4nOoYanwHLIhgKcfbMGc6fP8/Zs2d3/HHiG6Ug8gNEJpPh5z//OX/9ve8xOTGB8AbIlLahl7c9Mr/zkRgZtNg0anQCb2wS0iuAHSb7wvPPc/78eTo7O/OyTEhB5AcQy7K4cuUKf/u3f8vFjz7CMk0IFJON1GEW1WEGy5He0INFb5ko6RWU1CJqfA5PYhaRsMtY+Hx+nn76ac6ft1Pb7vcgg3yjMPA8gCiKwpkzZzhz5gzRaJR3332XDz/8kE8/vUZ2NlfjXPMhvUEM1ZcLaQDFzKKaaWQm4YYreH0+jh49yulTr3L69Gk6Ozvd0nIHgUJPfsDIZrMMDAwwODjI0NAQi4uLLC0tuTUQI5EIJSUlVFVV0dTUREtLC01NTXlpgmyGbTNXCmXi7rH+2aLrnyj9KPL53LaDQpm4DZIPZeJM02R4eJj+/n5u3brF0N07TE1PY1n37xwi4RCtrW10nTzJqVOnHliqIR/ObScplInLYyzLYnh4mO7ubm7evElvTzeptJ2aVeK1aPAnOV6ZpdyrE9EMPIot9qSpEtU1plJRRgaWef3GDV5//XWqKiv43Iuf5wtf+AKNjY17eWpPLE+8yE3TZGhoiN7eXnp6eujt6SaRTAFQ6Tc4FUrQXp2kNZSm1LuRmD+bqK7StxLi2nKSv/3hG7zxxhu0t7XypS+/xIsvvrhTp1PgPuS1yHei1Jhpmty9e5ebN2/S3X2T/r4+0pksABV+kxPBOB0VKdrCqU2Jej3FHpPz5SucL18hpqtcWYrwyaTOn//5Xb77ne9w6NAhLl68yOnTp3f0wb17RaFM3C4Ti8W4fPkyly9f5sb1a25PXRPQaQsmaAunaA2lKPHubK6plDCe8nF5KcLV5RJiusDr0eg6+RSnTp3i+PHjNDY2PtaALZ1Os7S05Bbhj8fjJJNJMpmM62FxivGHw2FKSkqoqKigqqpqy4+SzweeSD95PB7n4sWLXLjwAT09PViWpMhrcSIc40hlko5wiohndxOohYDGYIbGYIav181zNx7gRjRMf88nXLlyBQCPR6O+rp66+nrKy8vdZ8d7vd7PPPvTfe7P8jKLi/az6Z3xw+bbJqiqrKC5pZW2tjY6Ojro6OjYljDYveZA9eSGYXD16lXeeecdLl++hGlatl1dvEJXcZzGQCZvKz8sZDRGkn5Gk35m0l7mdT9RXSX7gOtQAEGPJKhaRNQsEc2g2GNQ5DEp8hhENJOwZhJUTQKqhVeRKML+rS0pyFiCpKmyoqssZT3MZTxMp71MZoLMpe1YGCEELc1NdJ7o4sSJExw/fvyhlWz3kgM9rS+l5O7du7z77rt88P57rMTiRDySp0uWebY0RkMeC3sjZC1B1hIYlgJIFAFexRHtzhwzaSiMJv0MJfwMJoKMJAMYli361pZmOk90cfz4cY4ePbrjNQ43yoETuZSSsbExPvzwQ3514QMmJqfQFOiMxDhbFuNoUQJ1Hws739AtwXDCz2A8sEb0ANVVlXQcPkJTUxONjY3U1NRQWVn5yKpWDqZpkk6nSafTbvK5UwTKyQ92ktsf9hCvA2GTR6NR+vr6uHnzJlcuX2J2bh4BtIVTfLNhhVMlcYLao0pXFNgKHkXSEUnREUkBi+iWYDTpYygRYDQZp/fSDBcurBWfz+slEg4RCAbweLwoioIlJYZTTSGdJpW2K7BtFCEERZEwZWVlVFRWUVVVRU1NzSMDyPJK5Lqus7S0xMLCgvts+bGxMe7eGWR2zo6Q86rQHkrwhYY4J4oTFO/y4LGALfq2cJq2cNr9LGUqzKY9zGc9LGc9rBgqSUMlE1MwpMCSoAhQhcSrWHg0ib/YwqdY+FULr2LhUySaIlGFXbjJkmBIQdZSSJsKCVMlpi8TXZhmbGaQ61nPA8csq9m0yJ0aI+vrnaxe3DqJmYx7G1r9mkqlSKVSJJNJkok48bi9rPcMCOwJmVpfivN1aZqDaQ4F02h7XM3ijfEKJlIPL/OWNhVSpkJAtX/EfKM+kOEbDY9+QtxGCagWTaEMTaGteXe2gpQQN1TmMh5++JDtNiXyO3fu8Bu/8RuP2TQbBYkm7CtXERIvkoAHVMX+3JNbJ5DEDZWeaIieaH6M7CdSPtLWw7Nx/H4/L738Er/4xS9IJ9IP3XYvmEj5HnmhHhT2zFyxEGSlIGvaPbYq7NuUauVErlh4LIlXsd/vN1566SW+9a1vAWx75aoCm2NTIm9ra+ONN9545HZSSrds2mqzxfnbMVdSqRSJRIJEIkEsFiMajbK0uMDCwgIruVlJsO3whkCa5qA9idMaSuFTH69M3ePwZwP13Ek8fJLkF7/4xZrXfKM+kOG/75jYlWNZubKTArbF7WlYENU1FrMeFnI+/oexIz25EMItCxeJbK2ofCKRYGpqitHRUYaGhhi4fZv37t7hl7MWHgU6wgm6iuOcLI4T2mWvSn3g0XZn2kxz8a0fUq5a+EP5dyfayDlsBFPCTNrLZMrHTMbLfMbDsu4hZnpIGgpp0x50OigCPAr4VIlPsfApJj5h4FctPIptpirOwBMwpSBjKqQthYTlIW5oxLOC1V2cqj58kLav/OTpdJq+vj6uXr3KpU8+ZnZuHkXA0UiCZ0tXOFGcwKvsXQ//JJA0FO4mAtyJBxhOBhhP+dFz17AQgsqKMiqraigrKyMcDhMIBPB4PAgh3Dt8Npt17+zJZNK+qycTZHMPT3AesKCoCpqq4Q8ECASCFBUXU1RUZLsQKyqorq6mqqqKiooKNE07WJNBcG+m81e/+hUfvP8uC4vLBDTJqeIVzpSu0BJK7+uZznwhaSgMJQIMxAMMJkJMJL1IQFNV2tpaOXzkqFvyra6ubs9yQw/cjOd6TNOkt7eXd955h4sf/opMVqfCZ/JMyTJPl8ao9m9fhdTdwLAEC1mNFV0jbqhkLNvXDI6f2b7VhzSTkGoS8Zj4FeuxL2pLwlzGw1jSz3DSz1AyyGRO1B5N5fCRI3R1naSzs5OOjo68ilo88CJfTSqVsqvFvvMO3T3dSAm1AZ2TRSt0FidoCGR2LOZjK0gJMxkPt2NBhhMBxtJB5tMqmzW6PAoUeS2K1CxFmk7EY18AIc3MTbZIFCRC3LNzk6bCiq6xpGvMZ33MZrzu5Irf56Wj4zDHOzvp7Ozk8OHDeSXq9TxRIl/N4uIiFy9e5OKHH9LX34eUEPFIOkJxDkeStIeTlHuNXTdrpIThpJ9ry2FurBSxlLF97uVlpbR3HKapqYmamhoqKioeGGqbTCaJx+N2qG00ytLSEktLS3Z2/uI8y8tRN27+YXg8GhVlZdTU1VNfX09zczOtra00NjZ+5uFf+cwTK/LVRKNRrl27xtWrV7lx/RrRFbvyaonXojWXONEWTlHty+6Y6BcyGpeWiri0VMJCRsWjqZx86hRnz57l5MmTVFdXb+vxTNO0Z5JzT+tzHpKgaZqbNBEMBvOiKOfjUhD5Opwoxp6eHnp6eujr6WYpapdEC3sk7aE47eEUHeEkVT79sUSfNhVuRENcWixmIB5ACDhx4gRf/OKXOHfu3IFISsgHCiJ/BFJKpqam3GTm7hvXWVhaBmw7tz2YoDWcojmYpjaQeWgYr5SwpGvcigXpXQnRFwtjWHZI6pe+/BJf+tKXqKys3KUze3IoiHyTSCmZnp6mu7vbXm7eYGk5CoAqoMqvU+7JUOQx0YTt1UgYuZIUmQBx3b4KystKOP/cC7zwwgscOXLkQJgF+cqBiCffTYQQ1NbWUltby1e+8hWklMzNzXHr1i2Gh4cZHR1lfm6W0aVFdN1ASkk4HKKkpoxzTc20tLTQ1dVFQ0NDQdh5QKFM3B6xX89t9SPbH0ahTNwGOcil1PL53BYXF+nv72dgYIA7d+4wMjrC9PQ0qVUuSUVRKK8sp721nePHj3PmzBmOHj3qpqcVysQVyBuSySQDAwP09/fb6YXdN5mbnXPXKxEFM2wi6yT4sBUjwcpazCRmmO+d5+LFi/zFX/wFFZUV/Norv8ZXv/rVPTuf+1EQ+RPE8vIyt2/fZnBwkMHBQfpv9TMxPuGaIEpIwSg14CTIMgmlYGoPzy8zMCADYlowNzrHf379P/P6668zOTnJ7du3OXz48G6c2kPJa5HnU6mx7Wanzy2TydDf3093dze9vb309ibLBQ4AABN/SURBVPeyMLfgrldCCmaxiTwmkaUSysD0bzFf1geySSKbJCRBDArCs2Fee+01XnzxRV577TVaWlq26cw2T8GFeEAwDIO+vj6uXLnClStX6OntwdDtWo4iIjBLTCjFFnQJsNNhKFkQAwJ1UAUdXnnlFb71rW9RVVW1I4cr+MkPIFJKRkdH3RqPV65eIZ3K5ZKWglVpISskVGDb0ntFBkS/LXZN1fjmN7/Jb//2b285meZBFER+ADAMg+HhYW7evGnXnrl6haXFJQBEWGBWmchqCVXsfC+9FRIgugXKqEI4EuZ3f+d3+cY3voHPtz1XYEHk+wzTNBkdHWVgYIBbt27Rf8t+ykU2V2JaCSgYFQZUgaySEN7jBm+GJVC7VZiG4pJifuef/g5f+9rXHjuGpyDyPEZKyfj4OH19fe4yMDiAnrUTPYQmkMUSq9SCMpDlEkLYWcH7mTlQe1WYhUAwwK9//dd59dVXaWpq2tLuCiLPIzKZDLdu3XLNjpvdN4nlwn6FJpClEqvEsgeJJRIiwB4XU9pRFuwBqjKugIQjR47wuc99jnPnztHe3r7hWdOCyPcIXdcZGxtzJ1t6ensYuD3gFsMXRQKzzITyXA990AX9MNIgRgXKmAKL9kcer4eOjg6aDjVRW1tLeXk5xcXFhEIhAoEAPp8Pn8+H3++nsrJyewK0UqkUFy9eJB6Pk0gk3EzrTC7LWtd1LMtO3RZCoKoqmqbh8XjWNMipUuq89/v97nqv14vP58Pj8bj/+6BKpo+LU/JO13V3MQwD0zTdxckcl1K6r846PVe80snSiUajbh3H8Ylxpqem730fTi/dbtmC3mavh7gmEMvbbMPoucWTWzaBLJHIU5tI4vODPCwxD5u24GcFmcUMPbM99A/1YyW3XtZjUz15JBKRp0+f/uxONIFQhd0Lrf6eJWCCtCTSeEJKRQhABalIuwtRufe6k3b0Mgh9ew8QCAR49dVXefPNN0mlHp1KtxrpyfnjtwuJXYjFyr2Xqz6XcOGnF3Yu1FZ4BEIT9o+4XuQWoIIw7QtAZp8AoSsghbS/h9XLPuTVV1/lj/7ojwD4m7/5mz1uDbkSXNwT+QbltCmRNzQ08Cd/8ifu7Xm9ueLc6t2daxqqquL1evF6vQ80VRxzxTFVnO0dk8UxW1RVfez4bMuy1pgpq80Vp1qvs97Z1jFVVi/O/zll7+LxOMvLyywuLjI9M83ExATRXKIFgOJXMEtNZLm0zZWyzX77D0d5V4G5R2+3GZwajluq5VgC1he3aGLowDyIRdsEU+MqMrl1a6Aw8NxBYrGYXeIu5+/u7ulmfGzcXqlgC6HcsgeeZRKCbLnX3/c2uQ5iTCDGBcqcgrQkQggaDzXS3NRMbW0tZWVllJSUEAqFPjOWa2lpKWQG7QWRSISTJ09y8uRJ97NYLOam1d24cYPevl70Adsn7vb2pfJejEmADQlfnpLITVdryQNWQNwWqGMq0pDU1dfxhd/8AmfPnuXYsWPbkuhdEPkuE4lEeO6553juuecAe7p+cHCQ3t5eezKov4+xvrF74a9+BbMkJ/wyCeXsbSzKdrEMSq+CmLCLw77ya6/w6quvcuzYsW1PGSyIfI/RNI2jR49y9OhR97NUKsXg4CC3b99mYGCA3r5eRvpH7qWeFQnMChMqc9P6++mBzrFcDMu4QiAY4Dd/7zf5xje+QUnJdrpi1lIQeR4SCATo6uqiq6vL/SyVStl2fc7MuX7jOqm7tltPlOQCtGpy/vd8LHyVBtErUIYUfF4fv/XPf4tvfvOb2x6NeD8KIt8nBAIBTp06xalTpwDbzLl9+zZXrlzhk0uf0H2zG/O2iVAEslzaobaV2+/F2TRGzua+rSIswde//nV+7/d+j7Kysl1rQsG7ckBIpVJcv36dK1eucOnyJYbuDtnmjcD24pRa95Imitj53t4EcUeg3lKRacnnP/95/uAP/oDGxsYdOVyh7soTQCAQ4Pz585w/fx6wvTg3b960K4L1dNPf3++aNyg5u77YhJKc8EvZHjVkQNwVqHdUZEpy+unTfPvb36azs3Mbdr41CiI/oEQiEZ5//nmef/55wJ4Em5ycZHBwkIGBATto7FY/yyN2OTzEqoCxspzfvoiNBYwZwKwdYKVOqkhT8uyZZ/nd3/1d17zaSwoif0JQFIWGhgYaGhr44he/6H6+sLDA7du36evrs2tB9vaQGEoAIFSBLJJYEcuOYXdKUljYk0QJUKIKYkkgLUk4EublX3+Zr3/967S2tu7+ST6Agk1eYA1OEodTXGhwcJCRsREW5hbciEqHUDhES0sLJzpPcPbsWZ566qm8fJxKoScvsAYhBI2NjTQ2NvKVr3zF/dwwDBKJBOl0GlVVCQaDBAKBfVHrsVALcY/YzLk5v9H61/UIIdxlrynUQtwg+Vwv8H4sLS0xMjLC2NgYMzMzzM7OEl1eZiUaJZVKoes6Zu6W//bbb/Nrr7yCpmkoirLm0SVOhGM6kyGTzW6pLaXFxTS3tNDS2kp7eztHjhyhsrJy1y6AQi3EA4CU0i0R0dPTw62+PhaWltz1HgRFSIKGQUBKiqVEk46zQvK+YdI0v4AlnDwAR3wSBdAkaEh8UuKT4JMSv3T+lniQqDk3uAnoQpASgoQiiCoKi9ksE0vL3Lh2DSf4uaSoiKPHj3PkyBGOHDlCW1vbE/Gki4LIN0E6nbYflHvpEpc/+YTlFfsRLMVAXVanyzSpMk3KLYuglA8NHvyetHg5nd6+xjkmzJpKb2lMYE5RmNZUJrNZbn30ER999BFgXyB1dXV0HD5Ma2srra2tNDc3U1xcvH3tygPyWuT5UAsxk8lw+fJl3n//fS5fukRW1/ELQXMmy3nD4JBhULSJcY3D12pqd6C1n0UFaiyLmqzFKXRIpUkJwbSqMKWqzIyMcGVyinfffdf9n/LSUlrb22nNmTqHDx/e9DR8Pvx2DgUX4n0wTZPu7m7eeecdPrxwgVQmQwjoyGQ5rOvUm2ZexkA9DgkhmFMVZhWVOVVhzuNhQdx7hn1leTmdXV10dnZy8uRJamtr82KA61BwIW4Ax8Z+7733ePeXv2RhaQkvgsPZDMeyOo2meaCrRYSkJGSYNDv2TiqNDsyqKlOqyuT0NJcWFtwev6K8nNNPP83p06c5derUrkQTbpW86snT6TTDw8NMTU0xNzdHPB4nnU4jhMDr9RIOhykpKaG6upra2loqKysfq1yFZVncuXOHjz/+mF998AHjk5MoQLNucFzP0qYbm836OtBIYElRGNVURjSNUa+XjLTT1NpbWzn19NOcPHmSo0eP4vfvbpB73hYXSiQSXL9+nevXr9N98yZjY2NrErg8Qrgi0wF9XVu9Hg8N9fUcam6msbGR+vp6amtrqaioIBwOr7kATNNkYWGBqakphoaG7NrdN24QjcUQQINhcljPckQ3CG7Bxn4SsYApVWVY0xjxaEyrql2gQVFobW3l2PHjtOds+/r6+h31m+eVyCcnJ/nkk0/4+OOP6evtxbQsPEJQr+vUGibVpkmZZVFkWZ+xpQwgLgQrisKSorCoKiwoKgseD7F1+Y2KouDLZfvrhkEmm10ziVIENGSzNBkmLUZB2NtBBpjQNMZVlQmPxoyqYuTWqapKbXU1DYcOUVNTQ21tLdXV1VRXV1NVVYXX+3ilePfUJs9ms25x+Esff8z45CQAFVLyTCZLi2FQt8GBnAaUSEmJaXLINO3u3TkO9q10WVFIKAoJIdBFChOBhsQrIWJZFEmLStMiVBD1tuMDWg2DVsOATAYLWFAU5lWFOUVlcWSEgYlJLgtc8YPtyiwrLaWmtpbaujpqa2upq6tz78yPa/psq8illCwuLnL37l07si2XnJvVdVSgwTD4sm7QquuUbLPIvEC1ZVFtbb2c2MP4pd/HrLqzPpUMkBHCnvDZ0SNBlWny5XRmR4+hAJWWRaVlccyVdQqJfUeOKkpuESzPzLA8P89IXx/xddooLy2lobGRuvp66urq3DtBZWXlhiazNiXyTCZDd3e3W1xoZWWFpaUl5ufnmZmeZnx8nEQyCdhXZ6UlOaHbJkGjYeDFFsvPgrsX/7JdzKoq2R12mfn9fl5++WXeeust5rZzoug+zKrqjl+0m0VISYmuUwwYCHQBhhDI+XlmFxa4dfMm6XUXQDgYpKKiglAo1PGg/W5K5GNjY/zxH//xA9fXGwbP6Lb5UWWaefnAg3zm5Zdf5tvf/jZSSn784x/vdXP2DAF4kHgkIOWau05KCJZzoQvzisq4YTCeSKBpWtGD9rcpkauqSlFR0b0BnJRkMxkyum0cT2gaU5pGmZRU6zr1hkmjaVJiWe4U907fIneKvw4FGd/hqLq33noLKSVvv/32jh4HbHPltxLJHT/O45IGllSVHo/HHXMtaypRVSW1QZP3sb0rUkri8TgzMzNMTEwwOjrK3Tt3uNXfTyxhZ5gUAc3ZLC26QVPObNlvFGzynUFi987RXO/sCHlJVVnWNJKr9CmEoKqigrr6empqa6mqqqKyspKKigpOnjzZbZpm1/2O8fhVbYUgEokQiURob2+/1/hchkl3dzeffvop165e5UYmgwY06gZthk6bbhDZAS+Hif3F6UJgAmrOu/KooKmHkQ+C2K84A83FnNt3OSfkFU0jqihk12mgtLiYhsZGuurrqc8NNp0B54MyjyzLeuAPtGt+cl3X6e3ttX3kFy8yM2eXYK2UkkPZLI2GSZ1pbthfbQErimBeUVlQFRYUhSVVY0VTSTxgHyr2XaVS16nOmVLVBzAOZa9ZEYJJzQ4HmNE05jVtzYDR5/FQXV1NTV2d6ytf7TffisswryaD4F4vf+nSJa5cvkxfXx+6YbuYIkCpYVBkWfilxCslEoEhIC0EcSGIaR6W1/laS0tKaGhsdGc8i4uLCQQC9mSQrpNMJpmfn2dqaoq7g4NMz84CuBGFHbpOq1GYxt8KCSHcqf4xrxenYLVH0+wQ3rY2mpqaOHToEHV1dZSVlW17cFfeiXw92WzWrft39+5dpiYnmZ2ZIZ5IuOLXVJVIOExpaSlVNTXuZMGhQ4dobGwkFApt6pjLy8t0d3e7k1TRWAwPgvZslmO6TrNhHOiArMfBACZUW9TDXi+zii3YUDBI18mTdHV1cfz4cZqbm3ctBS7vRf4wLMva8bxF0zTp6enh/fff58IHH5BIJgkBRzIZjmd1qld5h55EssC0qjKhqYxpGpOahoEdo3Ls2DE3GrG1tXVNGt9usq9Fvtvous7ly5f55S9/yeVLlzBMkzIpOZLJckTXqdihGdV8IQvMqyozqsKsqjKjeZhXBBa2/7rp0CFOnjrFU089RVdXF7uZ2P4wCiLfIvF4nAsXLvD+e+/R3d2NBMqlpC2TpdUwqN3ng9akEMyoCjOqaidLeD0sc+9RPJFQiPaODg7nckKPHTtGOJyfj38uiHwbWFxc5OLFi3z44Yf0dHdjWhZeIWjMZqk3TOpzs7z5OnA1gDlVZTK3zOQE7VBdWUlrezstLS20tbXR3Ny8q9n9j0tB5NtMIpHg2rVrfPrpp9y4fp2p6WnADkgqlVBm6JTkwoWDliQgJd5cdr0zmHUeXmZBLmP/npgU7Mx+T86/75NyU3cMZ5ZwXrF76RmPxqyiuDnO5aWlHDl2jCNHjrjx3vnaQ2+UQvrbNhMKhXjhhRd44YUXALveiuMdGhkZYXR4mKG5OQzTfMSeNo5HCPyA37LwWBYeS6LmClmYORdrStNICLHGJ+33+Whvb+f84cMcOXKEo0ePUl5evm3t2g8URL4NlJaWcu7cOc6dO+d+ZlkW0WiU5eVlYrEY6XSabDaLaZqut0hVVRRFcQsMObjFhdJpUqkUiUSCeDzuPgk7kUiQSafRs1lMKfFoGiG/n8aiIkpLS93JlebmZqqrq3fsidb7hU2ZK0KIOWBk55rzGSqA+V083m5ykM8Ndv/8mqSUlfdbsSmR7zZCiMsPsrP2Owf53CC/zu/Jvo8VeCIoiLzAgSffRf7/7HUDdpCDfG6QR+eX1zZ5gQLbQb735AUKPDZ5LXIhxP8uhOgXQtwQQvxQCLFzz6beRYQQrwghbgkhBoUQ/26v27OdCCEahRDvCCH6hBA9Qoh/s+dtymdzRQjxMvBLKaUhhPjfAKSU/+MeN+uxEEKowG3gK8A4cAn4J1LK3j1t2DYhhKgFaqWUV4UQEeAK8I/38vzyuieXUr4lpXQSgD4CGvayPdvEWWBQSnlXSpkF/hr49T1u07YhpZySUl7NvY8BfUD9XrYpr0W+jn8J/HSvG7EN1ANjq/4eZ49FsFMIIZqB08DHe9mOPY9dEUL8PVBzn1X/QUr5o9w2/wE7WvQvd7NtO8T9Ylfz12bcIkKIMPAD4N9KKVf2si17LnIp5T942HohxD8Hvgq8JPN5ALFxxoHGVX83AJN71JYdQQjhwRb4X0op39jz9uSzboQQrwD/J/AFKeXcXrdnOxBCaNgDz5eACeyB529LKXv2tGHbhLCzLP5fYFFK+W/3uj2Q/yIfxK4IvJD76CMp5b/awyZtC0KIfwT8X9ilYL4jpfxPe9ykbUMI8TngA+Amdk4IwL+XUv63PWtTPou8QIHtYD95VwoU2BIFkRc48BREXuDAUxB5gQNPQeQFDjwFkRc48BREXuDAUxB5gQPP/w/gHGmqy8QThwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 216x180 with 5 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 216x180 with 5 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 216x180 with 5 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 216x180 with 5 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "for h in head:\n",
    "    compare(h,h)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "def classify_connections(r):\n",
    "    if r['c1'] >= 0 and r['c2'] >= 0:\n",
    "        return \"PP\"  # Positive Positive\n",
    "    elif r['c1'] >= 0 and r['c2'] < 0:\n",
    "        return \"PN\"  # Positive Negative\n",
    "    elif r['c1'] < 0 and r['c2'] >= 0:\n",
    "        return \"NP\"  # Negative Positive\n",
    "    else:\n",
    "        return \"NN\"  # Negative Negative"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "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>bias_r</th>\n",
       "      <th>bias_z</th>\n",
       "      <th>bias_h</th>\n",
       "      <th>weight_xr</th>\n",
       "      <th>weight_xz</th>\n",
       "      <th>weight_xh</th>\n",
       "      <th>weight_hr</th>\n",
       "      <th>weight_hz</th>\n",
       "      <th>weight_hh</th>\n",
       "      <th>bias_rw</th>\n",
       "      <th>c1</th>\n",
       "      <th>c2</th>\n",
       "      <th>c0</th>\n",
       "      <th>classes</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-0.655686</td>\n",
       "      <td>2.139471</td>\n",
       "      <td>-0.277780</td>\n",
       "      <td>1.787566</td>\n",
       "      <td>-0.728850</td>\n",
       "      <td>3.101659</td>\n",
       "      <td>-0.872249</td>\n",
       "      <td>0.580573</td>\n",
       "      <td>-3.192146</td>\n",
       "      <td>0.825515</td>\n",
       "      <td>1.448174</td>\n",
       "      <td>3.100648</td>\n",
       "      <td>-3.895533</td>\n",
       "      <td>PP</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.685464</td>\n",
       "      <td>2.163037</td>\n",
       "      <td>-0.327781</td>\n",
       "      <td>1.787566</td>\n",
       "      <td>-0.575491</td>\n",
       "      <td>2.941030</td>\n",
       "      <td>-0.872249</td>\n",
       "      <td>0.578717</td>\n",
       "      <td>-3.192146</td>\n",
       "      <td>0.825515</td>\n",
       "      <td>1.562524</td>\n",
       "      <td>3.100648</td>\n",
       "      <td>-3.514895</td>\n",
       "      <td>PP</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2.420199</td>\n",
       "      <td>2.467700</td>\n",
       "      <td>-0.926128</td>\n",
       "      <td>-0.807916</td>\n",
       "      <td>0.749830</td>\n",
       "      <td>2.346618</td>\n",
       "      <td>1.497829</td>\n",
       "      <td>-1.223468</td>\n",
       "      <td>-4.061283</td>\n",
       "      <td>-0.256585</td>\n",
       "      <td>-2.954488</td>\n",
       "      <td>-4.799525</td>\n",
       "      <td>-5.041083</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>2.591266</td>\n",
       "      <td>2.435434</td>\n",
       "      <td>-0.926128</td>\n",
       "      <td>-0.521316</td>\n",
       "      <td>-0.282335</td>\n",
       "      <td>2.346618</td>\n",
       "      <td>1.497829</td>\n",
       "      <td>0.015339</td>\n",
       "      <td>-4.061283</td>\n",
       "      <td>-0.256585</td>\n",
       "      <td>-2.954488</td>\n",
       "      <td>-4.799525</td>\n",
       "      <td>-4.596333</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>2.420199</td>\n",
       "      <td>2.467700</td>\n",
       "      <td>-0.926128</td>\n",
       "      <td>-0.794483</td>\n",
       "      <td>0.851053</td>\n",
       "      <td>2.346618</td>\n",
       "      <td>1.611991</td>\n",
       "      <td>-1.223468</td>\n",
       "      <td>-3.986925</td>\n",
       "      <td>-0.256585</td>\n",
       "      <td>-2.954488</td>\n",
       "      <td>-4.799525</td>\n",
       "      <td>-5.102641</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     bias_r    bias_z    bias_h  weight_xr  weight_xz  weight_xh  weight_hr  \\\n",
       "0 -0.655686  2.139471 -0.277780   1.787566  -0.728850   3.101659  -0.872249   \n",
       "3 -0.685464  2.163037 -0.327781   1.787566  -0.575491   2.941030  -0.872249   \n",
       "4  2.420199  2.467700 -0.926128  -0.807916   0.749830   2.346618   1.497829   \n",
       "5  2.591266  2.435434 -0.926128  -0.521316  -0.282335   2.346618   1.497829   \n",
       "8  2.420199  2.467700 -0.926128  -0.794483   0.851053   2.346618   1.611991   \n",
       "\n",
       "   weight_hz  weight_hh   bias_rw        c1        c2        c0 classes  \n",
       "0   0.580573  -3.192146  0.825515  1.448174  3.100648 -3.895533      PP  \n",
       "3   0.578717  -3.192146  0.825515  1.562524  3.100648 -3.514895      PP  \n",
       "4  -1.223468  -4.061283 -0.256585 -2.954488 -4.799525 -5.041083      NN  \n",
       "5   0.015339  -4.061283 -0.256585 -2.954488 -4.799525 -4.596333      NN  \n",
       "8  -1.223468  -3.986925 -0.256585 -2.954488 -4.799525 -5.102641      NN  "
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_g['classes'] = df_g.apply(lambda row: classify_connections(row), axis=1).values\n",
    "df_g.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "a,b = 2,75\n",
    "temp = df_g.iloc[a].copy()\n",
    "df_g.iloc[a] = df_g.iloc[b]\n",
    "df_g.iloc[b] = temp"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2\n",
      "bias_r       -0.481571\n",
      "bias_z        -1.24299\n",
      "bias_h       0.0709518\n",
      "weight_xr     -1.36397\n",
      "weight_xz     0.773489\n",
      "weight_xh     -1.21522\n",
      "weight_hr      1.13805\n",
      "weight_hz      5.69377\n",
      "weight_hh   -0.0203089\n",
      "bias_rw        0.44017\n",
      "c1            -2.41601\n",
      "c2              2.5955\n",
      "c0            -4.09624\n",
      "classes             NP\n",
      "Name: 4, dtype: object\n"
     ]
    }
   ],
   "source": [
    "i = 0\n",
    "while df_g.iloc[i][\"classes\"] != \"NP\":\n",
    "    i += 1\n",
    "print(i)\n",
    "print(df_g.iloc[i])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### PCA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Explained variation per principal component: [0.2988147  0.17083279]\n"
     ]
    }
   ],
   "source": [
    "pca = PCA(n_components=2)\n",
    "pca_result = pca.fit_transform(df_g[head].values)\n",
    "df_g['pca-one'] = pca_result[:,0]\n",
    "df_g['pca-two'] = pca_result[:,1] \n",
    "print('Explained variation per principal component: {}'.format(pca.explained_variance_ratio_))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "The PostScript backend does not support transparency; partially transparent artists will be rendered opaque.\n",
      "The PostScript backend does not support transparency; partially transparent artists will be rendered opaque.\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(5,5))\n",
    "sns.scatterplot(\n",
    "    x=\"pca-one\", y=\"pca-two\",\n",
    "    hue=\"classes\",\n",
    "    palette=\"tab10\",\n",
    "    data=df_g,\n",
    ")\n",
    "plt.tight_layout()\n",
    "legend_elements = [\n",
    "    Line2D([0], [0], marker='o', color=COLORS[0], label=\"$c_1 \\geq 0 , c_2 \\geq 0 , w_{xh} \\geq 0$\", markerfacecolor=COLORS[1], markersize=10),\n",
    "    Line2D([0], [0], marker='o', color=COLORS[0], label=\"$c_1 \\geq 0 , c_2 < 0 , w_{xh} < 0$\", markerfacecolor=COLORS[2], markersize=10),\n",
    "    Line2D([0], [0], marker='o', color=COLORS[0], label=\"$c_1 < 0 , c_2 \\geq 0 , w_{xh} < 0$\", markerfacecolor=COLORS[3], markersize=10),\n",
    "    Line2D([0], [0], marker='o', color=COLORS[0], label=\"$c_1 < 0 , c_2 < 0 , w_{xh} \\geq 0$\", markerfacecolor=COLORS[4], markersize=10),\n",
    "]\n",
    "plt.legend(handles=legend_elements,\n",
    "           loc='center left',\n",
    "           title=None,\n",
    "           bbox_to_anchor=(1, 0.5),\n",
    "           fancybox=True,\n",
    "           fontsize=10,\n",
    "           ncol=1)\n",
    "# plt.axhline(0, linewidth=0.5, color=(0.5,0.5,0.5,0.5))\n",
    "# plt.axvline(0, linewidth=0.5, color=(0.5,0.5,0.5,0.5))\n",
    "plt.xlim(-8,8)\n",
    "plt.ylim(-8,8)\n",
    "plt.xlabel(\"First principal component\")\n",
    "plt.ylabel(\"Second principal component\")\n",
    "plt.savefig(f\"delete_me/pca.png\", bbox_inches='tight', pad_inches=0.02, dpi=500)\n",
    "plt.savefig(f\"delete_me/pca.eps\", format='eps', bbox_inches='tight', pad_inches=0.02)\n",
    "plt.show()\n",
    "plt.close()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "hide_input": false,
  "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": "3.7.7"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {},
   "toc_section_display": true,
   "toc_window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}