Will_version.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "id": "ijwJamy1VgHD"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\William\\Anaconda3\\lib\\site-packages\\statsmodels\\tools\\_testing.py:19: FutureWarning: pandas.util.testing is deprecated. Use the functions in the public API at pandas.testing instead.\n",
      "  import pandas.util.testing as tm\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "import seaborn as sb\n",
    "from sklearn import preprocessing\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.linear_model import LinearRegression\n",
    "from sklearn.tree import DecisionTreeRegressor\n",
    "from sklearn.ensemble import BaggingRegressor, AdaBoostRegressor, RandomForestRegressor\n",
    "from sklearn import tree\n",
    "from os import system\n",
    "\n",
    "# Just a start, any suggestions on different algorithms?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "id": "-R-ko-D-lOD1"
   },
   "outputs": [],
   "source": [
    "frequency = pd.read_csv('frequency_domain_features_train.csv')\n",
    "\n",
    "heart_rate = pd.read_csv('heart_rate_non_linear_features_train.csv')\n",
    "\n",
    "time_domain = pd.read_csv('time_domain_features_train.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 204
    },
    "id": "6GHTvkO3CqjV",
    "outputId": "664f8dd0-1b28-4115-9835-c992b58125ad"
   },
   "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>uuid</th>\n",
       "      <th>VLF</th>\n",
       "      <th>VLF_PCT</th>\n",
       "      <th>LF</th>\n",
       "      <th>LF_PCT</th>\n",
       "      <th>LF_NU</th>\n",
       "      <th>HF</th>\n",
       "      <th>HF_PCT</th>\n",
       "      <th>HF_NU</th>\n",
       "      <th>TP</th>\n",
       "      <th>LF_HF</th>\n",
       "      <th>HF_LF</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>89df2855-56eb-4706-a23b-b39363dd605a</td>\n",
       "      <td>2661.894136</td>\n",
       "      <td>72.203287</td>\n",
       "      <td>1009.249419</td>\n",
       "      <td>27.375666</td>\n",
       "      <td>98.485263</td>\n",
       "      <td>15.522603</td>\n",
       "      <td>0.421047</td>\n",
       "      <td>1.514737</td>\n",
       "      <td>3686.666157</td>\n",
       "      <td>65.018055</td>\n",
       "      <td>0.015380</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>80c795e4-aa56-4cc0-939c-19634b89cbb2</td>\n",
       "      <td>2314.265450</td>\n",
       "      <td>76.975728</td>\n",
       "      <td>690.113275</td>\n",
       "      <td>22.954139</td>\n",
       "      <td>99.695397</td>\n",
       "      <td>2.108525</td>\n",
       "      <td>0.070133</td>\n",
       "      <td>0.304603</td>\n",
       "      <td>3006.487251</td>\n",
       "      <td>327.296635</td>\n",
       "      <td>0.003055</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>c2d5d102-967c-487d-88f2-8b005a449f3e</td>\n",
       "      <td>1373.887112</td>\n",
       "      <td>51.152225</td>\n",
       "      <td>1298.222619</td>\n",
       "      <td>48.335104</td>\n",
       "      <td>98.950472</td>\n",
       "      <td>13.769729</td>\n",
       "      <td>0.512671</td>\n",
       "      <td>1.049528</td>\n",
       "      <td>2685.879461</td>\n",
       "      <td>94.280910</td>\n",
       "      <td>0.010607</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>37eabc44-1349-4040-8896-0d113ad4811f</td>\n",
       "      <td>2410.357408</td>\n",
       "      <td>70.180308</td>\n",
       "      <td>1005.981659</td>\n",
       "      <td>29.290305</td>\n",
       "      <td>98.224706</td>\n",
       "      <td>18.181913</td>\n",
       "      <td>0.529387</td>\n",
       "      <td>1.775294</td>\n",
       "      <td>3434.520980</td>\n",
       "      <td>55.328701</td>\n",
       "      <td>0.018074</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>aa777a6a-7aa3-4f6e-aced-70f8691dd2b7</td>\n",
       "      <td>1151.177330</td>\n",
       "      <td>43.918366</td>\n",
       "      <td>1421.782051</td>\n",
       "      <td>54.242160</td>\n",
       "      <td>96.720007</td>\n",
       "      <td>48.215822</td>\n",
       "      <td>1.839473</td>\n",
       "      <td>3.279993</td>\n",
       "      <td>2621.175204</td>\n",
       "      <td>29.487873</td>\n",
       "      <td>0.033912</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                   uuid          VLF    VLF_PCT           LF  \\\n",
       "0  89df2855-56eb-4706-a23b-b39363dd605a  2661.894136  72.203287  1009.249419   \n",
       "1  80c795e4-aa56-4cc0-939c-19634b89cbb2  2314.265450  76.975728   690.113275   \n",
       "2  c2d5d102-967c-487d-88f2-8b005a449f3e  1373.887112  51.152225  1298.222619   \n",
       "3  37eabc44-1349-4040-8896-0d113ad4811f  2410.357408  70.180308  1005.981659   \n",
       "4  aa777a6a-7aa3-4f6e-aced-70f8691dd2b7  1151.177330  43.918366  1421.782051   \n",
       "\n",
       "      LF_PCT      LF_NU         HF    HF_PCT     HF_NU           TP  \\\n",
       "0  27.375666  98.485263  15.522603  0.421047  1.514737  3686.666157   \n",
       "1  22.954139  99.695397   2.108525  0.070133  0.304603  3006.487251   \n",
       "2  48.335104  98.950472  13.769729  0.512671  1.049528  2685.879461   \n",
       "3  29.290305  98.224706  18.181913  0.529387  1.775294  3434.520980   \n",
       "4  54.242160  96.720007  48.215822  1.839473  3.279993  2621.175204   \n",
       "\n",
       "        LF_HF     HF_LF  \n",
       "0   65.018055  0.015380  \n",
       "1  327.296635  0.003055  \n",
       "2   94.280910  0.010607  \n",
       "3   55.328701  0.018074  \n",
       "4   29.487873  0.033912  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "frequency.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 204
    },
    "id": "wpWnJ_UpCud4",
    "outputId": "adea686d-2ac4-48a9-a76e-72fb1fee5cce"
   },
   "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>uuid</th>\n",
       "      <th>SD1</th>\n",
       "      <th>SD2</th>\n",
       "      <th>sampen</th>\n",
       "      <th>higuci</th>\n",
       "      <th>datasetId</th>\n",
       "      <th>condition</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>89df2855-56eb-4706-a23b-b39363dd605a</td>\n",
       "      <td>11.001565</td>\n",
       "      <td>199.061782</td>\n",
       "      <td>2.139754</td>\n",
       "      <td>1.163485</td>\n",
       "      <td>2</td>\n",
       "      <td>no stress</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>80c795e4-aa56-4cc0-939c-19634b89cbb2</td>\n",
       "      <td>9.170129</td>\n",
       "      <td>114.634458</td>\n",
       "      <td>2.174499</td>\n",
       "      <td>1.084711</td>\n",
       "      <td>2</td>\n",
       "      <td>interruption</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>c2d5d102-967c-487d-88f2-8b005a449f3e</td>\n",
       "      <td>11.533417</td>\n",
       "      <td>118.939253</td>\n",
       "      <td>2.135350</td>\n",
       "      <td>1.176315</td>\n",
       "      <td>2</td>\n",
       "      <td>interruption</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>37eabc44-1349-4040-8896-0d113ad4811f</td>\n",
       "      <td>11.119476</td>\n",
       "      <td>127.318597</td>\n",
       "      <td>2.178341</td>\n",
       "      <td>1.179688</td>\n",
       "      <td>2</td>\n",
       "      <td>no stress</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>aa777a6a-7aa3-4f6e-aced-70f8691dd2b7</td>\n",
       "      <td>13.590641</td>\n",
       "      <td>87.718281</td>\n",
       "      <td>2.221121</td>\n",
       "      <td>1.249612</td>\n",
       "      <td>2</td>\n",
       "      <td>no stress</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                   uuid        SD1         SD2    sampen  \\\n",
       "0  89df2855-56eb-4706-a23b-b39363dd605a  11.001565  199.061782  2.139754   \n",
       "1  80c795e4-aa56-4cc0-939c-19634b89cbb2   9.170129  114.634458  2.174499   \n",
       "2  c2d5d102-967c-487d-88f2-8b005a449f3e  11.533417  118.939253  2.135350   \n",
       "3  37eabc44-1349-4040-8896-0d113ad4811f  11.119476  127.318597  2.178341   \n",
       "4  aa777a6a-7aa3-4f6e-aced-70f8691dd2b7  13.590641   87.718281  2.221121   \n",
       "\n",
       "     higuci  datasetId     condition  \n",
       "0  1.163485          2     no stress  \n",
       "1  1.084711          2  interruption  \n",
       "2  1.176315          2  interruption  \n",
       "3  1.179688          2     no stress  \n",
       "4  1.249612          2     no stress  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "heart_rate.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 479
    },
    "id": "du7YWDgKC5gG",
    "outputId": "1ff37108-df9f-4225-eb48-691979a44a55"
   },
   "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>MEAN_RR</th>\n",
       "      <th>MEDIAN_RR</th>\n",
       "      <th>SDRR</th>\n",
       "      <th>RMSSD</th>\n",
       "      <th>SDSD</th>\n",
       "      <th>SDRR_RMSSD</th>\n",
       "      <th>HR</th>\n",
       "      <th>pNN25</th>\n",
       "      <th>pNN50</th>\n",
       "      <th>KURT</th>\n",
       "      <th>SKEW</th>\n",
       "      <th>MEAN_REL_RR</th>\n",
       "      <th>MEDIAN_REL_RR</th>\n",
       "      <th>SDRR_REL_RR</th>\n",
       "      <th>RMSSD_REL_RR</th>\n",
       "      <th>SDSD_REL_RR</th>\n",
       "      <th>SDRR_RMSSD_REL_RR</th>\n",
       "      <th>KURT_REL_RR</th>\n",
       "      <th>SKEW_REL_RR</th>\n",
       "      <th>uuid</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>885.157845</td>\n",
       "      <td>853.763730</td>\n",
       "      <td>140.972741</td>\n",
       "      <td>15.554505</td>\n",
       "      <td>15.553371</td>\n",
       "      <td>9.063146</td>\n",
       "      <td>69.499952</td>\n",
       "      <td>11.133333</td>\n",
       "      <td>0.533333</td>\n",
       "      <td>-0.856554</td>\n",
       "      <td>0.335218</td>\n",
       "      <td>-0.000203</td>\n",
       "      <td>-0.000179</td>\n",
       "      <td>0.017080</td>\n",
       "      <td>0.007969</td>\n",
       "      <td>0.007969</td>\n",
       "      <td>2.143342</td>\n",
       "      <td>-0.856554</td>\n",
       "      <td>0.335218</td>\n",
       "      <td>89df2855-56eb-4706-a23b-b39363dd605a</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>939.425371</td>\n",
       "      <td>948.357865</td>\n",
       "      <td>81.317742</td>\n",
       "      <td>12.964439</td>\n",
       "      <td>12.964195</td>\n",
       "      <td>6.272369</td>\n",
       "      <td>64.363150</td>\n",
       "      <td>5.600000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-0.408190</td>\n",
       "      <td>-0.155286</td>\n",
       "      <td>-0.000059</td>\n",
       "      <td>0.000611</td>\n",
       "      <td>0.013978</td>\n",
       "      <td>0.004769</td>\n",
       "      <td>0.004769</td>\n",
       "      <td>2.930855</td>\n",
       "      <td>-0.408190</td>\n",
       "      <td>-0.155286</td>\n",
       "      <td>80c795e4-aa56-4cc0-939c-19634b89cbb2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>898.186047</td>\n",
       "      <td>907.006860</td>\n",
       "      <td>84.497236</td>\n",
       "      <td>16.305279</td>\n",
       "      <td>16.305274</td>\n",
       "      <td>5.182201</td>\n",
       "      <td>67.450066</td>\n",
       "      <td>13.066667</td>\n",
       "      <td>0.200000</td>\n",
       "      <td>0.351789</td>\n",
       "      <td>-0.656813</td>\n",
       "      <td>-0.000011</td>\n",
       "      <td>-0.000263</td>\n",
       "      <td>0.018539</td>\n",
       "      <td>0.008716</td>\n",
       "      <td>0.008716</td>\n",
       "      <td>2.127053</td>\n",
       "      <td>0.351789</td>\n",
       "      <td>-0.656813</td>\n",
       "      <td>c2d5d102-967c-487d-88f2-8b005a449f3e</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>881.757865</td>\n",
       "      <td>893.460030</td>\n",
       "      <td>90.370537</td>\n",
       "      <td>15.720468</td>\n",
       "      <td>15.720068</td>\n",
       "      <td>5.748591</td>\n",
       "      <td>68.809562</td>\n",
       "      <td>11.800000</td>\n",
       "      <td>0.133333</td>\n",
       "      <td>-0.504947</td>\n",
       "      <td>-0.386138</td>\n",
       "      <td>0.000112</td>\n",
       "      <td>0.000494</td>\n",
       "      <td>0.017761</td>\n",
       "      <td>0.008660</td>\n",
       "      <td>0.008660</td>\n",
       "      <td>2.050988</td>\n",
       "      <td>-0.504947</td>\n",
       "      <td>-0.386138</td>\n",
       "      <td>37eabc44-1349-4040-8896-0d113ad4811f</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>809.625331</td>\n",
       "      <td>811.184865</td>\n",
       "      <td>62.766242</td>\n",
       "      <td>19.213819</td>\n",
       "      <td>19.213657</td>\n",
       "      <td>3.266724</td>\n",
       "      <td>74.565728</td>\n",
       "      <td>20.200000</td>\n",
       "      <td>0.200000</td>\n",
       "      <td>-0.548408</td>\n",
       "      <td>-0.154252</td>\n",
       "      <td>-0.000100</td>\n",
       "      <td>-0.002736</td>\n",
       "      <td>0.023715</td>\n",
       "      <td>0.013055</td>\n",
       "      <td>0.013055</td>\n",
       "      <td>1.816544</td>\n",
       "      <td>-0.548408</td>\n",
       "      <td>-0.154252</td>\n",
       "      <td>aa777a6a-7aa3-4f6e-aced-70f8691dd2b7</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      MEAN_RR   MEDIAN_RR        SDRR      RMSSD       SDSD  SDRR_RMSSD  \\\n",
       "0  885.157845  853.763730  140.972741  15.554505  15.553371    9.063146   \n",
       "1  939.425371  948.357865   81.317742  12.964439  12.964195    6.272369   \n",
       "2  898.186047  907.006860   84.497236  16.305279  16.305274    5.182201   \n",
       "3  881.757865  893.460030   90.370537  15.720468  15.720068    5.748591   \n",
       "4  809.625331  811.184865   62.766242  19.213819  19.213657    3.266724   \n",
       "\n",
       "          HR      pNN25     pNN50      KURT      SKEW  MEAN_REL_RR  \\\n",
       "0  69.499952  11.133333  0.533333 -0.856554  0.335218    -0.000203   \n",
       "1  64.363150   5.600000  0.000000 -0.408190 -0.155286    -0.000059   \n",
       "2  67.450066  13.066667  0.200000  0.351789 -0.656813    -0.000011   \n",
       "3  68.809562  11.800000  0.133333 -0.504947 -0.386138     0.000112   \n",
       "4  74.565728  20.200000  0.200000 -0.548408 -0.154252    -0.000100   \n",
       "\n",
       "   MEDIAN_REL_RR  SDRR_REL_RR  RMSSD_REL_RR  SDSD_REL_RR  SDRR_RMSSD_REL_RR  \\\n",
       "0      -0.000179     0.017080      0.007969     0.007969           2.143342   \n",
       "1       0.000611     0.013978      0.004769     0.004769           2.930855   \n",
       "2      -0.000263     0.018539      0.008716     0.008716           2.127053   \n",
       "3       0.000494     0.017761      0.008660     0.008660           2.050988   \n",
       "4      -0.002736     0.023715      0.013055     0.013055           1.816544   \n",
       "\n",
       "   KURT_REL_RR  SKEW_REL_RR                                  uuid  \n",
       "0    -0.856554     0.335218  89df2855-56eb-4706-a23b-b39363dd605a  \n",
       "1    -0.408190    -0.155286  80c795e4-aa56-4cc0-939c-19634b89cbb2  \n",
       "2     0.351789    -0.656813  c2d5d102-967c-487d-88f2-8b005a449f3e  \n",
       "3    -0.504947    -0.386138  37eabc44-1349-4040-8896-0d113ad4811f  \n",
       "4    -0.548408    -0.154252  aa777a6a-7aa3-4f6e-aced-70f8691dd2b7  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "time_domain.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "id": "z7mSyd8UDv72"
   },
   "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>uuid</th>\n",
       "      <th>SD1</th>\n",
       "      <th>SD2</th>\n",
       "      <th>sampen</th>\n",
       "      <th>higuci</th>\n",
       "      <th>datasetId</th>\n",
       "      <th>condition</th>\n",
       "      <th>VLF</th>\n",
       "      <th>VLF_PCT</th>\n",
       "      <th>LF</th>\n",
       "      <th>...</th>\n",
       "      <th>KURT</th>\n",
       "      <th>SKEW</th>\n",
       "      <th>MEAN_REL_RR</th>\n",
       "      <th>MEDIAN_REL_RR</th>\n",
       "      <th>SDRR_REL_RR</th>\n",
       "      <th>RMSSD_REL_RR</th>\n",
       "      <th>SDSD_REL_RR</th>\n",
       "      <th>SDRR_RMSSD_REL_RR</th>\n",
       "      <th>KURT_REL_RR</th>\n",
       "      <th>SKEW_REL_RR</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>237778</th>\n",
       "      <td>66403f0f-7574-4c2e-872d-13fe52e7571e</td>\n",
       "      <td>8.060510</td>\n",
       "      <td>134.886662</td>\n",
       "      <td>2.059638</td>\n",
       "      <td>1.176229</td>\n",
       "      <td>2</td>\n",
       "      <td>interruption</td>\n",
       "      <td>1602.213909</td>\n",
       "      <td>74.298023</td>\n",
       "      <td>536.938456</td>\n",
       "      <td>...</td>\n",
       "      <td>0.652782</td>\n",
       "      <td>0.257938</td>\n",
       "      <td>0.000015</td>\n",
       "      <td>-0.000086</td>\n",
       "      <td>0.014003</td>\n",
       "      <td>0.006935</td>\n",
       "      <td>0.006935</td>\n",
       "      <td>2.019195</td>\n",
       "      <td>0.652782</td>\n",
       "      <td>0.257938</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>140657</th>\n",
       "      <td>aef343fd-6161-4182-9bcd-1da1b063ce17</td>\n",
       "      <td>15.281364</td>\n",
       "      <td>135.879068</td>\n",
       "      <td>2.179748</td>\n",
       "      <td>1.232235</td>\n",
       "      <td>2</td>\n",
       "      <td>interruption</td>\n",
       "      <td>1778.240803</td>\n",
       "      <td>49.374174</td>\n",
       "      <td>1795.895058</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.288070</td>\n",
       "      <td>0.068838</td>\n",
       "      <td>-0.000074</td>\n",
       "      <td>-0.001199</td>\n",
       "      <td>0.023021</td>\n",
       "      <td>0.012191</td>\n",
       "      <td>0.012191</td>\n",
       "      <td>1.888379</td>\n",
       "      <td>-0.288070</td>\n",
       "      <td>0.068838</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>346606</th>\n",
       "      <td>63000685-346c-40f3-abb5-2ee594ecf7ba</td>\n",
       "      <td>7.705983</td>\n",
       "      <td>83.115353</td>\n",
       "      <td>2.203397</td>\n",
       "      <td>1.098727</td>\n",
       "      <td>2</td>\n",
       "      <td>no stress</td>\n",
       "      <td>1404.220057</td>\n",
       "      <td>74.000003</td>\n",
       "      <td>487.925499</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.347363</td>\n",
       "      <td>0.053569</td>\n",
       "      <td>-0.000158</td>\n",
       "      <td>-0.000193</td>\n",
       "      <td>0.012491</td>\n",
       "      <td>0.004608</td>\n",
       "      <td>0.004608</td>\n",
       "      <td>2.711036</td>\n",
       "      <td>-0.347363</td>\n",
       "      <td>0.053569</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3052</th>\n",
       "      <td>737afa4f-76cc-4425-9a92-773deb09a19f</td>\n",
       "      <td>11.847343</td>\n",
       "      <td>211.334861</td>\n",
       "      <td>2.147074</td>\n",
       "      <td>1.116217</td>\n",
       "      <td>2</td>\n",
       "      <td>no stress</td>\n",
       "      <td>4815.377254</td>\n",
       "      <td>81.582680</td>\n",
       "      <td>1061.517294</td>\n",
       "      <td>...</td>\n",
       "      <td>0.073401</td>\n",
       "      <td>0.282527</td>\n",
       "      <td>-0.000285</td>\n",
       "      <td>-0.000223</td>\n",
       "      <td>0.020096</td>\n",
       "      <td>0.009185</td>\n",
       "      <td>0.009185</td>\n",
       "      <td>2.187891</td>\n",
       "      <td>0.073401</td>\n",
       "      <td>0.282527</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10281</th>\n",
       "      <td>a0d276fd-a988-4858-a0b9-20912d234d8c</td>\n",
       "      <td>9.690084</td>\n",
       "      <td>94.280483</td>\n",
       "      <td>2.112930</td>\n",
       "      <td>1.285844</td>\n",
       "      <td>2</td>\n",
       "      <td>time pressure</td>\n",
       "      <td>1410.462465</td>\n",
       "      <td>72.041368</td>\n",
       "      <td>515.183968</td>\n",
       "      <td>...</td>\n",
       "      <td>1.719037</td>\n",
       "      <td>-0.389206</td>\n",
       "      <td>0.000022</td>\n",
       "      <td>-0.000245</td>\n",
       "      <td>0.016507</td>\n",
       "      <td>0.009748</td>\n",
       "      <td>0.009748</td>\n",
       "      <td>1.693408</td>\n",
       "      <td>1.719037</td>\n",
       "      <td>-0.389206</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 37 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                        uuid        SD1         SD2    sampen  \\\n",
       "237778  66403f0f-7574-4c2e-872d-13fe52e7571e   8.060510  134.886662  2.059638   \n",
       "140657  aef343fd-6161-4182-9bcd-1da1b063ce17  15.281364  135.879068  2.179748   \n",
       "346606  63000685-346c-40f3-abb5-2ee594ecf7ba   7.705983   83.115353  2.203397   \n",
       "3052    737afa4f-76cc-4425-9a92-773deb09a19f  11.847343  211.334861  2.147074   \n",
       "10281   a0d276fd-a988-4858-a0b9-20912d234d8c   9.690084   94.280483  2.112930   \n",
       "\n",
       "          higuci  datasetId      condition          VLF    VLF_PCT  \\\n",
       "237778  1.176229          2   interruption  1602.213909  74.298023   \n",
       "140657  1.232235          2   interruption  1778.240803  49.374174   \n",
       "346606  1.098727          2      no stress  1404.220057  74.000003   \n",
       "3052    1.116217          2      no stress  4815.377254  81.582680   \n",
       "10281   1.285844          2  time pressure  1410.462465  72.041368   \n",
       "\n",
       "                 LF  ...      KURT      SKEW  MEAN_REL_RR  MEDIAN_REL_RR  \\\n",
       "237778   536.938456  ...  0.652782  0.257938     0.000015      -0.000086   \n",
       "140657  1795.895058  ... -0.288070  0.068838    -0.000074      -0.001199   \n",
       "346606   487.925499  ... -0.347363  0.053569    -0.000158      -0.000193   \n",
       "3052    1061.517294  ...  0.073401  0.282527    -0.000285      -0.000223   \n",
       "10281    515.183968  ...  1.719037 -0.389206     0.000022      -0.000245   \n",
       "\n",
       "        SDRR_REL_RR  RMSSD_REL_RR  SDSD_REL_RR  SDRR_RMSSD_REL_RR  \\\n",
       "237778     0.014003      0.006935     0.006935           2.019195   \n",
       "140657     0.023021      0.012191     0.012191           1.888379   \n",
       "346606     0.012491      0.004608     0.004608           2.711036   \n",
       "3052       0.020096      0.009185     0.009185           2.187891   \n",
       "10281      0.016507      0.009748     0.009748           1.693408   \n",
       "\n",
       "        KURT_REL_RR  SKEW_REL_RR  \n",
       "237778     0.652782     0.257938  \n",
       "140657    -0.288070     0.068838  \n",
       "346606    -0.347363     0.053569  \n",
       "3052       0.073401     0.282527  \n",
       "10281      1.719037    -0.389206  \n",
       "\n",
       "[5 rows x 37 columns]"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_train = heart_rate.merge(frequency, how = 'inner', on = 'uuid')\n",
    "df_train = df_train.merge(time_domain, how = 'inner', on = 'uuid')\n",
    "\n",
    "train = df_train.sample(1000)\n",
    "\n",
    "train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 1000 entries, 237778 to 88647\n",
      "Data columns (total 38 columns):\n",
      " #   Column                   Non-Null Count  Dtype  \n",
      "---  ------                   --------------  -----  \n",
      " 0   SD1                      1000 non-null   float64\n",
      " 1   SD2                      1000 non-null   float64\n",
      " 2   sampen                   1000 non-null   float64\n",
      " 3   higuci                   1000 non-null   float64\n",
      " 4   datasetId                1000 non-null   int64  \n",
      " 5   VLF                      1000 non-null   float64\n",
      " 6   VLF_PCT                  1000 non-null   float64\n",
      " 7   LF                       1000 non-null   float64\n",
      " 8   LF_PCT                   1000 non-null   float64\n",
      " 9   LF_NU                    1000 non-null   float64\n",
      " 10  HF                       1000 non-null   float64\n",
      " 11  HF_PCT                   1000 non-null   float64\n",
      " 12  HF_NU                    1000 non-null   float64\n",
      " 13  TP                       1000 non-null   float64\n",
      " 14  LF_HF                    1000 non-null   float64\n",
      " 15  HF_LF                    1000 non-null   float64\n",
      " 16  MEAN_RR                  1000 non-null   float64\n",
      " 17  MEDIAN_RR                1000 non-null   float64\n",
      " 18  SDRR                     1000 non-null   float64\n",
      " 19  RMSSD                    1000 non-null   float64\n",
      " 20  SDSD                     1000 non-null   float64\n",
      " 21  SDRR_RMSSD               1000 non-null   float64\n",
      " 22  HR                       1000 non-null   float64\n",
      " 23  pNN25                    1000 non-null   float64\n",
      " 24  pNN50                    1000 non-null   float64\n",
      " 25  KURT                     1000 non-null   float64\n",
      " 26  SKEW                     1000 non-null   float64\n",
      " 27  MEAN_REL_RR              1000 non-null   float64\n",
      " 28  MEDIAN_REL_RR            1000 non-null   float64\n",
      " 29  SDRR_REL_RR              1000 non-null   float64\n",
      " 30  RMSSD_REL_RR             1000 non-null   float64\n",
      " 31  SDSD_REL_RR              1000 non-null   float64\n",
      " 32  SDRR_RMSSD_REL_RR        1000 non-null   float64\n",
      " 33  KURT_REL_RR              1000 non-null   float64\n",
      " 34  SKEW_REL_RR              1000 non-null   float64\n",
      " 35  condition_interruption   1000 non-null   uint8  \n",
      " 36  condition_no stress      1000 non-null   uint8  \n",
      " 37  condition_time pressure  1000 non-null   uint8  \n",
      "dtypes: float64(34), int64(1), uint8(3)\n",
      "memory usage: 284.2 KB\n"
     ]
    }
   ],
   "source": [
    "train.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>count</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>min</th>\n",
       "      <th>25%</th>\n",
       "      <th>50%</th>\n",
       "      <th>75%</th>\n",
       "      <th>max</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>SD1</th>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.068720e+01</td>\n",
       "      <td>2.988433</td>\n",
       "      <td>3.933001</td>\n",
       "      <td>8.347661</td>\n",
       "      <td>10.406724</td>\n",
       "      <td>12.816114</td>\n",
       "      <td>18.658011</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SD2</th>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.563554e+02</td>\n",
       "      <td>111.810301</td>\n",
       "      <td>42.541430</td>\n",
       "      <td>91.951437</td>\n",
       "      <td>119.869624</td>\n",
       "      <td>165.366767</td>\n",
       "      <td>787.039225</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sampen</th>\n",
       "      <td>1000.0</td>\n",
       "      <td>2.072104e+00</td>\n",
       "      <td>0.197839</td>\n",
       "      <td>0.784982</td>\n",
       "      <td>2.045724</td>\n",
       "      <td>2.143411</td>\n",
       "      <td>2.183874</td>\n",
       "      <td>2.229643</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>higuci</th>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.181310e+00</td>\n",
       "      <td>0.061355</td>\n",
       "      <td>1.035755</td>\n",
       "      <td>1.138727</td>\n",
       "      <td>1.174804</td>\n",
       "      <td>1.223818</td>\n",
       "      <td>1.340765</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>datasetId</th>\n",
       "      <td>1000.0</td>\n",
       "      <td>2.000000e+00</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>VLF</th>\n",
       "      <td>1000.0</td>\n",
       "      <td>2.264667e+03</td>\n",
       "      <td>1873.582552</td>\n",
       "      <td>172.406217</td>\n",
       "      <td>1024.918581</td>\n",
       "      <td>1714.988217</td>\n",
       "      <td>2688.005847</td>\n",
       "      <td>11858.788627</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>VLF_PCT</th>\n",
       "      <td>1000.0</td>\n",
       "      <td>6.460225e+01</td>\n",
       "      <td>16.640809</td>\n",
       "      <td>19.760189</td>\n",
       "      <td>53.424846</td>\n",
       "      <td>66.448456</td>\n",
       "      <td>76.913719</td>\n",
       "      <td>97.347732</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>LF</th>\n",
       "      <td>1000.0</td>\n",
       "      <td>9.613308e+02</td>\n",
       "      <td>585.662186</td>\n",
       "      <td>100.557446</td>\n",
       "      <td>550.160352</td>\n",
       "      <td>795.075949</td>\n",
       "      <td>1244.991182</td>\n",
       "      <td>3134.440114</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>LF_PCT</th>\n",
       "      <td>1000.0</td>\n",
       "      <td>3.378272e+01</td>\n",
       "      <td>15.857930</td>\n",
       "      <td>2.578741</td>\n",
       "      <td>22.353165</td>\n",
       "      <td>31.696977</td>\n",
       "      <td>44.207346</td>\n",
       "      <td>77.258691</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>LF_NU</th>\n",
       "      <td>1000.0</td>\n",
       "      <td>9.560251e+01</td>\n",
       "      <td>3.888404</td>\n",
       "      <td>76.755100</td>\n",
       "      <td>93.646927</td>\n",
       "      <td>96.555876</td>\n",
       "      <td>98.704841</td>\n",
       "      <td>99.977561</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>HF</th>\n",
       "      <td>1000.0</td>\n",
       "      <td>3.941687e+01</td>\n",
       "      <td>44.513494</td>\n",
       "      <td>0.105925</td>\n",
       "      <td>11.587324</td>\n",
       "      <td>25.128426</td>\n",
       "      <td>46.263759</td>\n",
       "      <td>314.677188</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>HF_PCT</th>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.615024e+00</td>\n",
       "      <td>1.729136</td>\n",
       "      <td>0.005989</td>\n",
       "      <td>0.342064</td>\n",
       "      <td>1.023458</td>\n",
       "      <td>2.291304</td>\n",
       "      <td>11.651651</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>HF_NU</th>\n",
       "      <td>1000.0</td>\n",
       "      <td>4.397490e+00</td>\n",
       "      <td>3.888404</td>\n",
       "      <td>0.022439</td>\n",
       "      <td>1.295159</td>\n",
       "      <td>3.444124</td>\n",
       "      <td>6.353073</td>\n",
       "      <td>23.244900</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>TP</th>\n",
       "      <td>1000.0</td>\n",
       "      <td>3.265415e+03</td>\n",
       "      <td>1973.977534</td>\n",
       "      <td>391.608310</td>\n",
       "      <td>1890.015550</td>\n",
       "      <td>2878.944316</td>\n",
       "      <td>4114.868867</td>\n",
       "      <td>12743.518295</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>LF_HF</th>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.133248e+02</td>\n",
       "      <td>316.316325</td>\n",
       "      <td>3.302019</td>\n",
       "      <td>14.740450</td>\n",
       "      <td>28.034969</td>\n",
       "      <td>76.210628</td>\n",
       "      <td>4455.461054</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>HF_LF</th>\n",
       "      <td>1000.0</td>\n",
       "      <td>4.783933e-02</td>\n",
       "      <td>0.045436</td>\n",
       "      <td>0.000224</td>\n",
       "      <td>0.013122</td>\n",
       "      <td>0.035670</td>\n",
       "      <td>0.067841</td>\n",
       "      <td>0.302845</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MEAN_RR</th>\n",
       "      <td>1000.0</td>\n",
       "      <td>8.483168e+02</td>\n",
       "      <td>125.975238</td>\n",
       "      <td>562.719701</td>\n",
       "      <td>761.960047</td>\n",
       "      <td>823.437353</td>\n",
       "      <td>927.483550</td>\n",
       "      <td>1318.322907</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MEDIAN_RR</th>\n",
       "      <td>1000.0</td>\n",
       "      <td>8.440596e+02</td>\n",
       "      <td>134.443566</td>\n",
       "      <td>539.226625</td>\n",
       "      <td>760.887516</td>\n",
       "      <td>818.320737</td>\n",
       "      <td>919.797356</td>\n",
       "      <td>1653.122250</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SDRR</th>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.108931e+02</td>\n",
       "      <td>78.984512</td>\n",
       "      <td>30.306770</td>\n",
       "      <td>65.578606</td>\n",
       "      <td>85.079604</td>\n",
       "      <td>117.147372</td>\n",
       "      <td>556.545076</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RMSSD</th>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.510962e+01</td>\n",
       "      <td>4.224835</td>\n",
       "      <td>5.560268</td>\n",
       "      <td>11.801485</td>\n",
       "      <td>14.712678</td>\n",
       "      <td>18.118744</td>\n",
       "      <td>26.377851</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SDSD</th>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.510894e+01</td>\n",
       "      <td>4.224872</td>\n",
       "      <td>5.560248</td>\n",
       "      <td>11.801436</td>\n",
       "      <td>14.712421</td>\n",
       "      <td>18.118675</td>\n",
       "      <td>26.377609</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SDRR_RMSSD</th>\n",
       "      <td>1000.0</td>\n",
       "      <td>7.462333e+00</td>\n",
       "      <td>5.375931</td>\n",
       "      <td>2.827594</td>\n",
       "      <td>4.588150</td>\n",
       "      <td>5.958895</td>\n",
       "      <td>7.996226</td>\n",
       "      <td>53.374678</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>HR</th>\n",
       "      <td>1000.0</td>\n",
       "      <td>7.382229e+01</td>\n",
       "      <td>10.222204</td>\n",
       "      <td>49.965219</td>\n",
       "      <td>66.648759</td>\n",
       "      <td>74.316171</td>\n",
       "      <td>80.140967</td>\n",
       "      <td>107.207111</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>pNN25</th>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.014827e+01</td>\n",
       "      <td>8.407705</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>3.866667</td>\n",
       "      <td>7.933333</td>\n",
       "      <td>13.466667</td>\n",
       "      <td>38.733333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>pNN50</th>\n",
       "      <td>1000.0</td>\n",
       "      <td>9.099333e-01</td>\n",
       "      <td>1.008891</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.066667</td>\n",
       "      <td>0.533333</td>\n",
       "      <td>1.533333</td>\n",
       "      <td>4.933333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>KURT</th>\n",
       "      <td>1000.0</td>\n",
       "      <td>4.248475e-01</td>\n",
       "      <td>1.578028</td>\n",
       "      <td>-1.863072</td>\n",
       "      <td>-0.375153</td>\n",
       "      <td>0.003436</td>\n",
       "      <td>0.625843</td>\n",
       "      <td>17.745183</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SKEW</th>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.577216e-02</td>\n",
       "      <td>0.656618</td>\n",
       "      <td>-2.069373</td>\n",
       "      <td>-0.347984</td>\n",
       "      <td>-0.068509</td>\n",
       "      <td>0.265797</td>\n",
       "      <td>3.031353</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MEAN_REL_RR</th>\n",
       "      <td>1000.0</td>\n",
       "      <td>9.726655e-07</td>\n",
       "      <td>0.000163</td>\n",
       "      <td>-0.001129</td>\n",
       "      <td>-0.000074</td>\n",
       "      <td>0.000004</td>\n",
       "      <td>0.000074</td>\n",
       "      <td>0.000874</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MEDIAN_REL_RR</th>\n",
       "      <td>1000.0</td>\n",
       "      <td>-4.562780e-04</td>\n",
       "      <td>0.000859</td>\n",
       "      <td>-0.003748</td>\n",
       "      <td>-0.000907</td>\n",
       "      <td>-0.000298</td>\n",
       "      <td>0.000154</td>\n",
       "      <td>0.001845</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SDRR_REL_RR</th>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.866805e-02</td>\n",
       "      <td>0.005448</td>\n",
       "      <td>0.009371</td>\n",
       "      <td>0.014408</td>\n",
       "      <td>0.017682</td>\n",
       "      <td>0.021936</td>\n",
       "      <td>0.034386</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RMSSD_REL_RR</th>\n",
       "      <td>1000.0</td>\n",
       "      <td>9.720930e-03</td>\n",
       "      <td>0.003798</td>\n",
       "      <td>0.003871</td>\n",
       "      <td>0.007114</td>\n",
       "      <td>0.008768</td>\n",
       "      <td>0.011490</td>\n",
       "      <td>0.024516</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SDSD_REL_RR</th>\n",
       "      <td>1000.0</td>\n",
       "      <td>9.720914e-03</td>\n",
       "      <td>0.003798</td>\n",
       "      <td>0.003871</td>\n",
       "      <td>0.007114</td>\n",
       "      <td>0.008768</td>\n",
       "      <td>0.011490</td>\n",
       "      <td>0.024516</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SDRR_RMSSD_REL_RR</th>\n",
       "      <td>1000.0</td>\n",
       "      <td>2.005170e+00</td>\n",
       "      <td>0.364793</td>\n",
       "      <td>1.231794</td>\n",
       "      <td>1.757003</td>\n",
       "      <td>1.927496</td>\n",
       "      <td>2.217348</td>\n",
       "      <td>3.667298</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>KURT_REL_RR</th>\n",
       "      <td>1000.0</td>\n",
       "      <td>4.248475e-01</td>\n",
       "      <td>1.578028</td>\n",
       "      <td>-1.863072</td>\n",
       "      <td>-0.375153</td>\n",
       "      <td>0.003436</td>\n",
       "      <td>0.625843</td>\n",
       "      <td>17.745183</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SKEW_REL_RR</th>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.577216e-02</td>\n",
       "      <td>0.656618</td>\n",
       "      <td>-2.069373</td>\n",
       "      <td>-0.347984</td>\n",
       "      <td>-0.068509</td>\n",
       "      <td>0.265797</td>\n",
       "      <td>3.031353</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                    count          mean          std         min          25%  \\\n",
       "SD1                1000.0  1.068720e+01     2.988433    3.933001     8.347661   \n",
       "SD2                1000.0  1.563554e+02   111.810301   42.541430    91.951437   \n",
       "sampen             1000.0  2.072104e+00     0.197839    0.784982     2.045724   \n",
       "higuci             1000.0  1.181310e+00     0.061355    1.035755     1.138727   \n",
       "datasetId          1000.0  2.000000e+00     0.000000    2.000000     2.000000   \n",
       "VLF                1000.0  2.264667e+03  1873.582552  172.406217  1024.918581   \n",
       "VLF_PCT            1000.0  6.460225e+01    16.640809   19.760189    53.424846   \n",
       "LF                 1000.0  9.613308e+02   585.662186  100.557446   550.160352   \n",
       "LF_PCT             1000.0  3.378272e+01    15.857930    2.578741    22.353165   \n",
       "LF_NU              1000.0  9.560251e+01     3.888404   76.755100    93.646927   \n",
       "HF                 1000.0  3.941687e+01    44.513494    0.105925    11.587324   \n",
       "HF_PCT             1000.0  1.615024e+00     1.729136    0.005989     0.342064   \n",
       "HF_NU              1000.0  4.397490e+00     3.888404    0.022439     1.295159   \n",
       "TP                 1000.0  3.265415e+03  1973.977534  391.608310  1890.015550   \n",
       "LF_HF              1000.0  1.133248e+02   316.316325    3.302019    14.740450   \n",
       "HF_LF              1000.0  4.783933e-02     0.045436    0.000224     0.013122   \n",
       "MEAN_RR            1000.0  8.483168e+02   125.975238  562.719701   761.960047   \n",
       "MEDIAN_RR          1000.0  8.440596e+02   134.443566  539.226625   760.887516   \n",
       "SDRR               1000.0  1.108931e+02    78.984512   30.306770    65.578606   \n",
       "RMSSD              1000.0  1.510962e+01     4.224835    5.560268    11.801485   \n",
       "SDSD               1000.0  1.510894e+01     4.224872    5.560248    11.801436   \n",
       "SDRR_RMSSD         1000.0  7.462333e+00     5.375931    2.827594     4.588150   \n",
       "HR                 1000.0  7.382229e+01    10.222204   49.965219    66.648759   \n",
       "pNN25              1000.0  1.014827e+01     8.407705    0.000000     3.866667   \n",
       "pNN50              1000.0  9.099333e-01     1.008891    0.000000     0.066667   \n",
       "KURT               1000.0  4.248475e-01     1.578028   -1.863072    -0.375153   \n",
       "SKEW               1000.0  1.577216e-02     0.656618   -2.069373    -0.347984   \n",
       "MEAN_REL_RR        1000.0  9.726655e-07     0.000163   -0.001129    -0.000074   \n",
       "MEDIAN_REL_RR      1000.0 -4.562780e-04     0.000859   -0.003748    -0.000907   \n",
       "SDRR_REL_RR        1000.0  1.866805e-02     0.005448    0.009371     0.014408   \n",
       "RMSSD_REL_RR       1000.0  9.720930e-03     0.003798    0.003871     0.007114   \n",
       "SDSD_REL_RR        1000.0  9.720914e-03     0.003798    0.003871     0.007114   \n",
       "SDRR_RMSSD_REL_RR  1000.0  2.005170e+00     0.364793    1.231794     1.757003   \n",
       "KURT_REL_RR        1000.0  4.248475e-01     1.578028   -1.863072    -0.375153   \n",
       "SKEW_REL_RR        1000.0  1.577216e-02     0.656618   -2.069373    -0.347984   \n",
       "\n",
       "                           50%          75%           max  \n",
       "SD1                  10.406724    12.816114     18.658011  \n",
       "SD2                 119.869624   165.366767    787.039225  \n",
       "sampen                2.143411     2.183874      2.229643  \n",
       "higuci                1.174804     1.223818      1.340765  \n",
       "datasetId             2.000000     2.000000      2.000000  \n",
       "VLF                1714.988217  2688.005847  11858.788627  \n",
       "VLF_PCT              66.448456    76.913719     97.347732  \n",
       "LF                  795.075949  1244.991182   3134.440114  \n",
       "LF_PCT               31.696977    44.207346     77.258691  \n",
       "LF_NU                96.555876    98.704841     99.977561  \n",
       "HF                   25.128426    46.263759    314.677188  \n",
       "HF_PCT                1.023458     2.291304     11.651651  \n",
       "HF_NU                 3.444124     6.353073     23.244900  \n",
       "TP                 2878.944316  4114.868867  12743.518295  \n",
       "LF_HF                28.034969    76.210628   4455.461054  \n",
       "HF_LF                 0.035670     0.067841      0.302845  \n",
       "MEAN_RR             823.437353   927.483550   1318.322907  \n",
       "MEDIAN_RR           818.320737   919.797356   1653.122250  \n",
       "SDRR                 85.079604   117.147372    556.545076  \n",
       "RMSSD                14.712678    18.118744     26.377851  \n",
       "SDSD                 14.712421    18.118675     26.377609  \n",
       "SDRR_RMSSD            5.958895     7.996226     53.374678  \n",
       "HR                   74.316171    80.140967    107.207111  \n",
       "pNN25                 7.933333    13.466667     38.733333  \n",
       "pNN50                 0.533333     1.533333      4.933333  \n",
       "KURT                  0.003436     0.625843     17.745183  \n",
       "SKEW                 -0.068509     0.265797      3.031353  \n",
       "MEAN_REL_RR           0.000004     0.000074      0.000874  \n",
       "MEDIAN_REL_RR        -0.000298     0.000154      0.001845  \n",
       "SDRR_REL_RR           0.017682     0.021936      0.034386  \n",
       "RMSSD_REL_RR          0.008768     0.011490      0.024516  \n",
       "SDSD_REL_RR           0.008768     0.011490      0.024516  \n",
       "SDRR_RMSSD_REL_RR     1.927496     2.217348      3.667298  \n",
       "KURT_REL_RR           0.003436     0.625843     17.745183  \n",
       "SKEW_REL_RR          -0.068509     0.265797      3.031353  "
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train.describe().T"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "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>uuid</th>\n",
       "      <th>SD1</th>\n",
       "      <th>SD2</th>\n",
       "      <th>sampen</th>\n",
       "      <th>higuci</th>\n",
       "      <th>datasetId</th>\n",
       "      <th>condition</th>\n",
       "      <th>VLF</th>\n",
       "      <th>VLF_PCT</th>\n",
       "      <th>LF</th>\n",
       "      <th>...</th>\n",
       "      <th>KURT</th>\n",
       "      <th>SKEW</th>\n",
       "      <th>MEAN_REL_RR</th>\n",
       "      <th>MEDIAN_REL_RR</th>\n",
       "      <th>SDRR_REL_RR</th>\n",
       "      <th>RMSSD_REL_RR</th>\n",
       "      <th>SDSD_REL_RR</th>\n",
       "      <th>SDRR_RMSSD_REL_RR</th>\n",
       "      <th>KURT_REL_RR</th>\n",
       "      <th>SKEW_REL_RR</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>0 rows × 37 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "Empty DataFrame\n",
       "Columns: [uuid, SD1, SD2, sampen, higuci, datasetId, condition, VLF, VLF_PCT, LF, LF_PCT, LF_NU, HF, HF_PCT, HF_NU, TP, LF_HF, HF_LF, MEAN_RR, MEDIAN_RR, SDRR, RMSSD, SDSD, SDRR_RMSSD, HR, pNN25, pNN50, KURT, SKEW, MEAN_REL_RR, MEDIAN_REL_RR, SDRR_REL_RR, RMSSD_REL_RR, SDSD_REL_RR, SDRR_RMSSD_REL_RR, KURT_REL_RR, SKEW_REL_RR]\n",
       "Index: []\n",
       "\n",
       "[0 rows x 37 columns]"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Checking for duplicates\n",
    "\n",
    "duplicate = df_train[df_train.duplicated()]\n",
    "\n",
    "duplicate"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### No duplicates detected."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "train = pd.get_dummies(train, columns=['condition'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "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>uuid</th>\n",
       "      <th>SD1</th>\n",
       "      <th>SD2</th>\n",
       "      <th>sampen</th>\n",
       "      <th>higuci</th>\n",
       "      <th>datasetId</th>\n",
       "      <th>VLF</th>\n",
       "      <th>VLF_PCT</th>\n",
       "      <th>LF</th>\n",
       "      <th>LF_PCT</th>\n",
       "      <th>...</th>\n",
       "      <th>MEDIAN_REL_RR</th>\n",
       "      <th>SDRR_REL_RR</th>\n",
       "      <th>RMSSD_REL_RR</th>\n",
       "      <th>SDSD_REL_RR</th>\n",
       "      <th>SDRR_RMSSD_REL_RR</th>\n",
       "      <th>KURT_REL_RR</th>\n",
       "      <th>SKEW_REL_RR</th>\n",
       "      <th>condition_interruption</th>\n",
       "      <th>condition_no stress</th>\n",
       "      <th>condition_time pressure</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>237778</th>\n",
       "      <td>66403f0f-7574-4c2e-872d-13fe52e7571e</td>\n",
       "      <td>8.060510</td>\n",
       "      <td>134.886662</td>\n",
       "      <td>2.059638</td>\n",
       "      <td>1.176229</td>\n",
       "      <td>2</td>\n",
       "      <td>1602.213909</td>\n",
       "      <td>74.298023</td>\n",
       "      <td>536.938456</td>\n",
       "      <td>24.898964</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.000086</td>\n",
       "      <td>0.014003</td>\n",
       "      <td>0.006935</td>\n",
       "      <td>0.006935</td>\n",
       "      <td>2.019195</td>\n",
       "      <td>0.652782</td>\n",
       "      <td>0.257938</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>140657</th>\n",
       "      <td>aef343fd-6161-4182-9bcd-1da1b063ce17</td>\n",
       "      <td>15.281364</td>\n",
       "      <td>135.879068</td>\n",
       "      <td>2.179748</td>\n",
       "      <td>1.232235</td>\n",
       "      <td>2</td>\n",
       "      <td>1778.240803</td>\n",
       "      <td>49.374174</td>\n",
       "      <td>1795.895058</td>\n",
       "      <td>49.864358</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.001199</td>\n",
       "      <td>0.023021</td>\n",
       "      <td>0.012191</td>\n",
       "      <td>0.012191</td>\n",
       "      <td>1.888379</td>\n",
       "      <td>-0.288070</td>\n",
       "      <td>0.068838</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>346606</th>\n",
       "      <td>63000685-346c-40f3-abb5-2ee594ecf7ba</td>\n",
       "      <td>7.705983</td>\n",
       "      <td>83.115353</td>\n",
       "      <td>2.203397</td>\n",
       "      <td>1.098727</td>\n",
       "      <td>2</td>\n",
       "      <td>1404.220057</td>\n",
       "      <td>74.000003</td>\n",
       "      <td>487.925499</td>\n",
       "      <td>25.712842</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.000193</td>\n",
       "      <td>0.012491</td>\n",
       "      <td>0.004608</td>\n",
       "      <td>0.004608</td>\n",
       "      <td>2.711036</td>\n",
       "      <td>-0.347363</td>\n",
       "      <td>0.053569</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3052</th>\n",
       "      <td>737afa4f-76cc-4425-9a92-773deb09a19f</td>\n",
       "      <td>11.847343</td>\n",
       "      <td>211.334861</td>\n",
       "      <td>2.147074</td>\n",
       "      <td>1.116217</td>\n",
       "      <td>2</td>\n",
       "      <td>4815.377254</td>\n",
       "      <td>81.582680</td>\n",
       "      <td>1061.517294</td>\n",
       "      <td>17.984349</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.000223</td>\n",
       "      <td>0.020096</td>\n",
       "      <td>0.009185</td>\n",
       "      <td>0.009185</td>\n",
       "      <td>2.187891</td>\n",
       "      <td>0.073401</td>\n",
       "      <td>0.282527</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10281</th>\n",
       "      <td>a0d276fd-a988-4858-a0b9-20912d234d8c</td>\n",
       "      <td>9.690084</td>\n",
       "      <td>94.280483</td>\n",
       "      <td>2.112930</td>\n",
       "      <td>1.285844</td>\n",
       "      <td>2</td>\n",
       "      <td>1410.462465</td>\n",
       "      <td>72.041368</td>\n",
       "      <td>515.183968</td>\n",
       "      <td>26.313751</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.000245</td>\n",
       "      <td>0.016507</td>\n",
       "      <td>0.009748</td>\n",
       "      <td>0.009748</td>\n",
       "      <td>1.693408</td>\n",
       "      <td>1.719037</td>\n",
       "      <td>-0.389206</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>323511</th>\n",
       "      <td>80a91466-048b-4078-866c-40526a4b7ede</td>\n",
       "      <td>10.275728</td>\n",
       "      <td>63.153487</td>\n",
       "      <td>2.208784</td>\n",
       "      <td>1.339211</td>\n",
       "      <td>2</td>\n",
       "      <td>456.342399</td>\n",
       "      <td>40.451545</td>\n",
       "      <td>570.349214</td>\n",
       "      <td>50.557448</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.000333</td>\n",
       "      <td>0.019839</td>\n",
       "      <td>0.013597</td>\n",
       "      <td>0.013597</td>\n",
       "      <td>1.459059</td>\n",
       "      <td>-0.403443</td>\n",
       "      <td>-0.297158</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>237201</th>\n",
       "      <td>d01a29e3-3f50-4c4e-8cbd-b964e2e3e72c</td>\n",
       "      <td>6.601993</td>\n",
       "      <td>61.527288</td>\n",
       "      <td>2.184708</td>\n",
       "      <td>1.293529</td>\n",
       "      <td>2</td>\n",
       "      <td>572.046455</td>\n",
       "      <td>64.300074</td>\n",
       "      <td>286.003628</td>\n",
       "      <td>32.147834</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000107</td>\n",
       "      <td>0.012314</td>\n",
       "      <td>0.007709</td>\n",
       "      <td>0.007709</td>\n",
       "      <td>1.597414</td>\n",
       "      <td>-0.063071</td>\n",
       "      <td>-0.294677</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>280915</th>\n",
       "      <td>4225b807-c9db-4dad-9308-8fa02231c265</td>\n",
       "      <td>13.073333</td>\n",
       "      <td>179.670969</td>\n",
       "      <td>2.077140</td>\n",
       "      <td>1.136337</td>\n",
       "      <td>2</td>\n",
       "      <td>2649.693691</td>\n",
       "      <td>61.246470</td>\n",
       "      <td>1659.417073</td>\n",
       "      <td>38.356674</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000153</td>\n",
       "      <td>0.021773</td>\n",
       "      <td>0.009343</td>\n",
       "      <td>0.009343</td>\n",
       "      <td>2.330404</td>\n",
       "      <td>-0.196876</td>\n",
       "      <td>-0.649162</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>264356</th>\n",
       "      <td>a3950d10-9408-470e-97e7-a40a112879bb</td>\n",
       "      <td>15.077603</td>\n",
       "      <td>200.249270</td>\n",
       "      <td>2.136449</td>\n",
       "      <td>1.115346</td>\n",
       "      <td>2</td>\n",
       "      <td>5520.274393</td>\n",
       "      <td>77.384440</td>\n",
       "      <td>1607.753509</td>\n",
       "      <td>22.537848</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000104</td>\n",
       "      <td>0.022381</td>\n",
       "      <td>0.008836</td>\n",
       "      <td>0.008836</td>\n",
       "      <td>2.532797</td>\n",
       "      <td>0.720213</td>\n",
       "      <td>-0.410352</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>88647</th>\n",
       "      <td>a6e1dd32-1bed-461b-b425-f85fc9399dd9</td>\n",
       "      <td>8.443651</td>\n",
       "      <td>175.565958</td>\n",
       "      <td>2.175297</td>\n",
       "      <td>1.148201</td>\n",
       "      <td>2</td>\n",
       "      <td>2267.538798</td>\n",
       "      <td>79.420356</td>\n",
       "      <td>572.056237</td>\n",
       "      <td>20.036222</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000178</td>\n",
       "      <td>0.014936</td>\n",
       "      <td>0.006850</td>\n",
       "      <td>0.006850</td>\n",
       "      <td>2.180475</td>\n",
       "      <td>-0.890769</td>\n",
       "      <td>-0.110774</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1000 rows × 39 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                        uuid        SD1         SD2    sampen  \\\n",
       "237778  66403f0f-7574-4c2e-872d-13fe52e7571e   8.060510  134.886662  2.059638   \n",
       "140657  aef343fd-6161-4182-9bcd-1da1b063ce17  15.281364  135.879068  2.179748   \n",
       "346606  63000685-346c-40f3-abb5-2ee594ecf7ba   7.705983   83.115353  2.203397   \n",
       "3052    737afa4f-76cc-4425-9a92-773deb09a19f  11.847343  211.334861  2.147074   \n",
       "10281   a0d276fd-a988-4858-a0b9-20912d234d8c   9.690084   94.280483  2.112930   \n",
       "...                                      ...        ...         ...       ...   \n",
       "323511  80a91466-048b-4078-866c-40526a4b7ede  10.275728   63.153487  2.208784   \n",
       "237201  d01a29e3-3f50-4c4e-8cbd-b964e2e3e72c   6.601993   61.527288  2.184708   \n",
       "280915  4225b807-c9db-4dad-9308-8fa02231c265  13.073333  179.670969  2.077140   \n",
       "264356  a3950d10-9408-470e-97e7-a40a112879bb  15.077603  200.249270  2.136449   \n",
       "88647   a6e1dd32-1bed-461b-b425-f85fc9399dd9   8.443651  175.565958  2.175297   \n",
       "\n",
       "          higuci  datasetId          VLF    VLF_PCT           LF     LF_PCT  \\\n",
       "237778  1.176229          2  1602.213909  74.298023   536.938456  24.898964   \n",
       "140657  1.232235          2  1778.240803  49.374174  1795.895058  49.864358   \n",
       "346606  1.098727          2  1404.220057  74.000003   487.925499  25.712842   \n",
       "3052    1.116217          2  4815.377254  81.582680  1061.517294  17.984349   \n",
       "10281   1.285844          2  1410.462465  72.041368   515.183968  26.313751   \n",
       "...          ...        ...          ...        ...          ...        ...   \n",
       "323511  1.339211          2   456.342399  40.451545   570.349214  50.557448   \n",
       "237201  1.293529          2   572.046455  64.300074   286.003628  32.147834   \n",
       "280915  1.136337          2  2649.693691  61.246470  1659.417073  38.356674   \n",
       "264356  1.115346          2  5520.274393  77.384440  1607.753509  22.537848   \n",
       "88647   1.148201          2  2267.538798  79.420356   572.056237  20.036222   \n",
       "\n",
       "        ...  MEDIAN_REL_RR  SDRR_REL_RR  RMSSD_REL_RR  SDSD_REL_RR  \\\n",
       "237778  ...      -0.000086     0.014003      0.006935     0.006935   \n",
       "140657  ...      -0.001199     0.023021      0.012191     0.012191   \n",
       "346606  ...      -0.000193     0.012491      0.004608     0.004608   \n",
       "3052    ...      -0.000223     0.020096      0.009185     0.009185   \n",
       "10281   ...      -0.000245     0.016507      0.009748     0.009748   \n",
       "...     ...            ...          ...           ...          ...   \n",
       "323511  ...      -0.000333     0.019839      0.013597     0.013597   \n",
       "237201  ...       0.000107     0.012314      0.007709     0.007709   \n",
       "280915  ...       0.000153     0.021773      0.009343     0.009343   \n",
       "264356  ...       0.000104     0.022381      0.008836     0.008836   \n",
       "88647   ...       0.000178     0.014936      0.006850     0.006850   \n",
       "\n",
       "        SDRR_RMSSD_REL_RR  KURT_REL_RR  SKEW_REL_RR  condition_interruption  \\\n",
       "237778           2.019195     0.652782     0.257938                       1   \n",
       "140657           1.888379    -0.288070     0.068838                       1   \n",
       "346606           2.711036    -0.347363     0.053569                       0   \n",
       "3052             2.187891     0.073401     0.282527                       0   \n",
       "10281            1.693408     1.719037    -0.389206                       0   \n",
       "...                   ...          ...          ...                     ...   \n",
       "323511           1.459059    -0.403443    -0.297158                       1   \n",
       "237201           1.597414    -0.063071    -0.294677                       1   \n",
       "280915           2.330404    -0.196876    -0.649162                       0   \n",
       "264356           2.532797     0.720213    -0.410352                       1   \n",
       "88647            2.180475    -0.890769    -0.110774                       0   \n",
       "\n",
       "        condition_no stress  condition_time pressure  \n",
       "237778                    0                        0  \n",
       "140657                    0                        0  \n",
       "346606                    1                        0  \n",
       "3052                      1                        0  \n",
       "10281                     0                        1  \n",
       "...                     ...                      ...  \n",
       "323511                    0                        0  \n",
       "237201                    0                        0  \n",
       "280915                    0                        1  \n",
       "264356                    0                        0  \n",
       "88647                     1                        0  \n",
       "\n",
       "[1000 rows x 39 columns]"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "train.drop('uuid', inplace=True, axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Preparing data model\n",
    "\n",
    "cols = [col for col in train.columns if col != 'HR']\n",
    "X = train[cols]\n",
    "y = train['HR']\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=.2, random_state=7)\n",
    "\n",
    "X_train, X_val, y_train, y_val = train_test_split(X, y, test_size=.25, random_state=7)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((750, 37), (200, 37))"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.shape, X_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((750,), (200,))"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_train.shape, y_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((750, 37), (250, 37))"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.shape, X_val.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((750,), (250,))"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_train.shape, y_val.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 2.22152272, -0.39403205,  0.7600835 , ...,  1.45773797,\n",
       "        -1.03046381, -0.4445276 ],\n",
       "       [-0.68641787, -0.30971775, -0.02177995, ..., -0.68599434,\n",
       "         0.97043679, -0.4445276 ],\n",
       "       [-1.31771957, -0.7531216 ,  0.51776063, ...,  1.45773797,\n",
       "        -1.03046381, -0.4445276 ],\n",
       "       ...,\n",
       "       [-0.06401577,  0.38079944, -1.22780669, ..., -0.68599434,\n",
       "         0.97043679, -0.4445276 ],\n",
       "       [ 0.17618855,  0.44725324,  0.45260171, ..., -0.68599434,\n",
       "         0.97043679, -0.4445276 ],\n",
       "       [ 0.4452803 , -0.2785555 ,  0.26528349, ...,  1.45773797,\n",
       "        -1.03046381, -0.4445276 ]])"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Scaling the training/validation/test data\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "scaler = StandardScaler()\n",
    "\n",
    "scaler.fit_transform(X_train)\n",
    "\n",
    "scaler.fit_transform(X_val)\n",
    "\n",
    "scaler.fit_transform(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:>"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sb.heatmap(train.corr())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import mean_absolute_error, mean_squared_error, r2_score\n",
    "\n",
    "def scoring(model, test_features, test_labels):\n",
    "    predictions = model.predict(test_features)\n",
    "    mae = mean_absolute_error(test_labels, predictions)\n",
    "    mse = mean_squared_error(test_labels, predictions)\n",
    "    r2 = r2_score(test_labels, predictions)\n",
    "    errors = abs(predictions - test_labels)\n",
    "    mape = 100 * np.mean(errors / test_labels)\n",
    "    accuracy = 100 - mape\n",
    "    print(\"Model Performance\")\n",
    "    print(\"Mean Absolute Error: {:0.4f}.\".format(mae))\n",
    "    print(\"Mean Squared Error: {:0.4f}.\".format(mse))\n",
    "    print(\"R^2 Score: {:0.4f}.\".format(mae))\n",
    "    print(\"Accuracy: {:0.2f}.\".format(accuracy))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model Performance\n",
      "Mean Absolute Error: 8.5390.\n",
      "Mean Squared Error: 113.2862.\n",
      "R^2 Score: -0.0023.\n",
      "Accuracy: 88.17.\n"
     ]
    }
   ],
   "source": [
    "# Creating baseline performance report\n",
    "\n",
    "baseline_y = [y_train.median()] * len(y_val)\n",
    "\n",
    "base_predictions = baseline_y\n",
    "base_mae = mean_absolute_error(y_val, base_predictions)\n",
    "base_mse = mean_squared_error(y_val, base_predictions)\n",
    "base_r2 = r2_score(y_val, base_predictions)\n",
    "base_errors = abs(base_predictions - y_val)\n",
    "base_mape = 100 * np.mean(base_errors / y_val)\n",
    "base_accuracy = 100 - base_mape\n",
    "\n",
    "print(\"Model Performance\")\n",
    "print(\"Mean Absolute Error: {:0.4f}.\".format(base_mae))\n",
    "print(\"Mean Squared Error: {:0.4f}.\".format(base_mse))\n",
    "print(\"R^2 Score: {:0.4f}.\".format(base_r2))\n",
    "print(\"Accuracy: {:0.2f}.\".format(base_accuracy))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\William\\Anaconda3\\lib\\site-packages\\sklearn\\linear_model\\_coordinate_descent.py:531: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 913.5792474050057, tolerance: 7.598068085316166\n",
      "  positive)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'Coefficients')"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABJcAAAHICAYAAADz6QttAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8vihELAAAACXBIWXMAAAsTAAALEwEAmpwYAADNCUlEQVR4nOzdd3gc1dXH8e9Rl5vkLrljTDMdTO8tlFACoYXeAwkJpJMCgfQGCZAGKdRACi8dQjMtgCmmd2zABtkyNka2sSUXSef948zaa1m2VbZI8u/zPHq8u7Oee2d3dubOmXPvNXdHRERERERERESkIwryXQEREREREREREem+FFwSEREREREREZEOU3BJREREREREREQ6TMElERERERERERHpMAWXRERERERERESkw4ryXYFMGzRokI8ZMybf1RARERERERER6TGef/75j919cGvLelxwacyYMUyePDnf1RARERERERER6THMbPrqlqlbnIiIiIiIiIiIdJiCSyIiIiIiIiIi0mEKLomIiIiIiIiISIcpuCQiIiIiIiIiIh2m4JKIiIiIiIiIiHSYgksiIiIiIiIiItJhCi6JiIiIiIiIiEiHKbgkIiIiIiIiIiIdpuCSiIiIiIiIiIh0mIJLIiIiIiIiIiLSYQouiYiIiIiIiIhIhym4JCIiIiIiIiIiHabgkoiIiIhIBzz93lyueuzdfFdDREQk74ryXQERERERke6msamZC/7vFabNrWfvjYewwdC++a6SiIhI3ihzSURERESknW5/aSbT5tYDcP2k6XmujYiISH4puCQiIiIi0g6NTc1c+fAUNh3WjyO2Gc6tL9Tw6eJl+a6WiIhI3ii4JCIiIiLSDre/NJPpc+s5f98NOXmnMSxa2sStL8zId7VERETyRsElEREREZE2Ss9a2neTIWw5spItR1Zy3aRpuHu+qyciIpIXCi6JiIiIiLTRbS/OWJ61ZGYAnLzTaN6bs4gnp87Nc+1ERETyQ8ElEREREZE2aGxq5vePTGWz4ZG1lHLQ5tUM6F3C9ZOm5a9yIiIieaTgkoiIiIhIGyzPWtpnRdYSQFlxIcduN5KH3vyImrr6PNZQREQkPxRcEhERERFZi/SspX3SspZSjt9xNAD/eOaDXFdNREQk7/IaXDKzA8zsbTObamYXtLJ8dzN7wcwazezIfNRRREREROTW1WQtpQyvLGffTYbyr+c+ZPGypjzUUEREJH/yFlwys0LgD8CBwHjgC2Y2vsXbPgBOAW7Kbe1ERERERMKypmZ+//BUNh9e0WrWUsrJO4/hk0VLueeV2hzWTkREJP/ymbm0PTDV3d9z96XAP4HD0t/g7tPc/RWgOR8VFBERERG57cUZfPBJPefvu0GrWUspO68/kPUH9+b6p6fnsHYiIiL5l8/g0nDgw7TnNclr7WZmZ5nZZDObPGfOnIxUTkREREQkPWtp741Xn7UEYGactNMYXv5wHi99OC83FRQREekCesSA3u5+tbtPcPcJgwcPznd1RERERKSHuO2FtmUtpRyxzXB6lxRy/aRp2a+ciIhIF5HP4NIMYGTa8xHJayIiIiIiebesqZkrH5nCFiPWnrWU0resmM9vO4K7X6ll7sIlWa6hiIhI15DP4NJzwAZmtp6ZlQDHAnfmsT4iIiIiIsvd9sIMPvykoc1ZSykn7jiapY3N/Gvyh2t/s4iISA+Qt+CSuzcC5wL3A28C/3b3183sR2Z2KICZbWdmNcBRwFVm9nq+6isiIiIi6470rKW9Nmpb1lLKBkP7svP6A/nH0x/Q1OxZqqGIiEjXkdcxl9z9Xnff0N3Xd/efJq9d5O53Jo+fc/cR7t7b3Qe6+6b5rK+IiIiIrBtufaGmQ1lLKSftNJoZ8xqY+OZHWaidiIhI19IjBvQWEREREcmUZU3NXPnwVLbsQNZSyr6bDKW6oozrJ03PcO1ERES6HgWXRERERETS3PpCDTV1DZy/74YdyloCKCos4IQdR/PE1I+ZOnthhmsoIiLStSi4JCIiIiKSSM9a2nOjwZ1a1zHbjaSksIAbn1b2koiI9GwKLomIiIiIJP7v+c5nLaUM6lPKZ7eo5pbna1i4pDFDNRQREel6FFwSEREREQGWNjbz+0emsuXIyk5nLaWctNNoFi5p5LYXajKyPhERka5IwSUREREREdLHWurYDHGt2WpkJZsPr+D6SdNx94ysU0REpKtRcElERERE1nlLG5OxlkZWsueGmclaAjAzTtppNFNmL2TSe3Mztl4REZGuRMElEREREVnn/d8LNcyYl9mspZRDthxG/17F3DBJA3uLiEjPpOCSiIiIiKzTljY28/uHp7JVhrOWUsqKCzl6u5E88MZHzJzXkPH1i4iI5JuCSyIiIiKyTstm1lLKCTuMptmdm575ICvrFxERyScFl0RERERknZWetbRHFrKWUkYO6MU+Gw/ln899wJLGpqyVIyIikg8KLomIiIjIOuuW57OftZRy0k6j+XjhUv776qysliMiIpJrCi6JiIiIyDppaWMzf3hkKluPym7WUsqu4wYxdlBvrps0LetliYiI5JKCSyIiIiKyTlqRtbRh1rOWAAoKjBN3Gs2LH8zj1Zr5WS9PREQkVxRcEhEREZF1TnrW0u4bDMpZuZ/fdgS9Sgq5XtlLIiLSgyi4JCIiIiLrnP88/2FOs5ZS+pUVc/jWw7nz5ZnULVqas3JFRESyScElEREREVmnLG1s5g8PT2WbHGctpZy00xiWNDbz78kf5rxsERGRbFBwSURERETWKf95/kNmzl+c86yllI2q+rLDegO44enpNDV7zssXERHJNAWXRERERGSdsaSxaXnW0m55yFpKOXnnMdTUNfDIW7PzVgcREZFMUXBJRERERNYZ/5lck9espZT9xg9laL9Srn96et7qICIikikKLomIiIjIOmFJYxN/fGQq247un9esJYDiwgKO32E0j78zh/fmLMxrXURERDpLwSURERERWSesyFraIK9ZSynHbj+S4kLjxqc/yHdVREREOkXBJRERERHp8ZY0NvGHJGtp13H5zVpKGdK3jAM3q+Y/z3/IoiWN+a6OiIhIhym4JCIiIiI93r8n11DbhbKWUk7eeTSfLm7k9pdm5LsqIiIiHabgkoiIiIj0aKmxliZ0oayllG1G9WfTYf24YdJ03D3f1REREekQBZdEREREcmzOp0v4xr9fZrMf3s+MeQ35rk6PtyJrKb8zxLXGzDhpp9G8NetTnn3/k3xXR0REpEMUXBIRERHJkcamZq558n32vvRR/u+FGhYuaeTtWQvyXa0eLT1raZdxA/NdnVYduuVwKsqLuX7S9HxXRUREpEMUXBIRERHJgeemfcLBVz7BJXe9wVYjK7npjB0AmDlvcZ5r1rP9+7kPqZ2/mK/t1/WyllLKSwo5ZruR3P/6LGbN1/4gIiLdj4JLIiIiIlk0+9PFfP1fL3HUnyfx6eJG/nzCNlx/2vbsMHYghQWmYEIWxQxx77LdmP7svH7XzFpKOWGH0TS5c9OzH+S7KiIiIu1WlO8KiIiIiPREjU3NXDdpOr978B2WNDbz5b3W58t7jaNXSTS/Cg2G9i1l5nyNuZQt/37uQ2YtWMylR2/ZZbOWUkYN7MVeGw3hpmc+4Ny9xlFSpHvAIiLSfeisJSIiIpJhz7w3l4OvfIIf3/0GW4/uz/1f251v7b/x8sBSSlVFmTKXsmTxsu6TtZRy0k6j+XjhEu57fVa+qyIiItIuylwSERERyZDZCxbzs3vf5PaXZjK8spw/n7At+286dLVZM9WV5bwxUwN6Z8O/J0fW0mXdIGspZfcNBjNmYC+uf2oah245LN/VERERaTMFl0REREQ6aVlTM9c9NY3fPTSFpY3NfGXvcXxpz3GUlxSu8f9V9yvjoTc+wt27TQCkO1i8rIk/PvIu248ZwE7dJGsJoKDAOGHH0fzknjd5feZ8Nh1Wke8qiYiItIm6xYmIiIh0wtPvzeXgK57gJ/e8yYQx0QXuG5/ZaK2BJYjMpSWNzcyrX5aDmq47/pWMtXT+vht0u6DdUduOpKy4gBsmTc93VURERNpMwSURERGRDvhowWLO++eLHHv10yxc0sjVJ27LNadsx3qDerd5HdUVZQAa1DuDFi9r4o+PTu12WUspFb2KOXzr4dz+0gzmK+goIiLdhLrFiYjIGi1tbKZhaROLljZSv7SJhqVN1CeP65PHDcuaWLSkiYbU68uaqF+SvH9ZvG/RkkYG9y3lnD3XZ+f1B+V7s0Q6bFlTM9c+OY3fPfQOy5qdr+49jnPa0AWuNang0qz5i9UFKkP+9dyHfLRgCb89Zqtul7WUcuKOY7j52Q/5z/MfcsZuY/NdHRERkbVScElEZB20cEkj/3ruQ16fOT8JFq0IGKWeL1raSMPSJhqbvV3r7lVSSK+SQspLCuldUkR58rx/r168NmM+x/3lGXYZN5BvfGYjthnVP0tbKJIdk96dy0V3vMaU2QvZa6PBXHzopowe2PZMpZaqK8oBmKkZ4zJiedbSegPYaWz3y1pKGT+sH9uN6c/1k6Zz2i7rUVDQPYNkIiKy7lBwSURkHVK3aCnXPDWN656axvyGZQyrKKN3adHyYFBVv7LlwaBeJUVpgaIVj1Ovp4JHqce9SgopKypc40XQ4mVN3Pj0dP706Lsc8cen2GfjIXz9MxsqY0O6vFnzF/PTe9/krpdnMqJ/OX85aQL7bjKk05kxg/uWUlhgzFK3uE5zd659alq3z1pKOWmnMXzl5hd57J057LXxkHxXR0REZI0UXBIRWQfMmr+Yv/zvPW5+9gPqlzbxmfFD+dJe49hqZGVO61FWXMgZu43lC9uP4tqnpnHVY+/y2Sue4LObV/O1/TZk3JA+Oa2PyNosa2rmmiff5/KHprCs2Tlvnw04Z8/1KStufxe41hQWGEP7llI7T5lLHeXuPDl1LldMnMKz0z5h13GDunXWUsr+m1YxpG8p10+apuCSiIh0eQouiYj0YO9/vIirHnuX/3uhhmaHw7Ycxtl7rs+GQ/vmtV69S4v48l7jOGHH0fz1f+/xtyfe57+v1XL41iM4f98NGDmgV17rJwLw1NSPuejO15k6eyH7bDyEiw4Z36kucKtTXVlOrbrFtZu7878pH3P5xCk8P72Oof1KufiQ8Ry7/ahun7UEUFJUwBe2H8UVD09h2seLGNOOgeJFRERyTcElEZEe6I2ZC/jjo1O599VaigoLOHa7UZy1+9guF7SpKC/mG5/ZiFN2HsOfHn2X65+ezh0vzeCY7Ubylb03oCoZ7Fgkl2rnN/DTe97k7ldqGTmgnL+dPIF9NhmatfKqKsp4fcb8rK2/p3F3Hn1nDldMnMKLH8yjuqKMHx+2KUdNGJmxjLKu4rgdRvGHR6Zy49PT+cHB4/NdHRERkdVScElEpAd5bton/PGRqTzy9hz6lBZx1u7rc9quYxjSt2sHaQb2KeUHB4/njN3G8vtHpvDPZz/kludrOHHH0Zyz5/oM7FOa7yrKOmBZUzN/f+J9Lp84haZm5/x9N+DsPTLXBW51hlWU8dAbH+HuPSLjJlvcnUfens3lE6fy8ofzGF5Zzk8P34wjtx1BaVHPCiqlDO1XxgGbVfHvyR/yjc9s1KEZCUVERHJBwSURkW7O3XnsnTn88ZF3eXbaJwzoXcI3P7MhJ+40hory4nxXr12qKsr4yec254u7r8/lE6fw9yff56ZnP+C0XdbjzN3Hdrvtke7liolTuPLhqey7yVB+eMj4nGX6VVeUs6Sxmbr6ZQzoXZKTMrsTd+ehN2dzxcQpvDpjPiP6l/PzIzbn89uMoKSoIN/Vy7qTdhrD3a/UcsdLMzh2+1H5ro6IiEirFFwSEemmmpqd/75Wy58efZfXZy6guqKMHx4ynmO2G0mvku59eB85oBe/OWpLzt5jfX770Dv8/pGpXD9pGmftPpZTd1mP3qXde/uk62lY2sQNT09n/02HctWJE3JadnXS/bN2foOCS2ncnQfe+IgrJk7h9ZkLGDWgF7/6/BYcvs1wigt7flApZbsx/dm4qi/XTZrOMduNVHabiIh0SWqdi4h0M0sbm7ntxRr+/Nh7vP/xIsYO7s2vjtyCz201vMfdxR83pA9/OG4bvrznAi578G1+88A7XPPkNM7Zc31O2HF0jxtfRfLnthdnMK9+GafvOjbnZVdXlgNQO28xmw6ryHn5XU1zs/PAG7O4fOJU3qxdwOiBvfj1kVvwua3XraBSiplx8s5j+O6tr/L89DomjBmQ7yqJiIisQsElEZFuon5pIzc/+yF//d971M5fzGbD+/HH47dh/02rKCzo2Xeyxw/rx19P3o4XPqjj0gfe5if3vMlf//c+5+49jqMnjOxxQTXJLXfnmiffZ7Ph/dhuTP+cl788c2nBuj1jXHOz89/XZnHlw1N4a9anjB3Um8uO3pJDtxxG0ToYVEp32FbD+Nm9b3LdpOkKLomISJek4JKISBc3v34Z102axjVPvk9d/TJ2WG8Av/j8Fuy+waB1rnvENqP6848zduSpdz/mN/e/zQ9uf42rHn+X8/bZkMO3Ht7jg2ySHU9M/Zgpsxdy2dFb5uU3NahPKUUFRu28hpyX3RU0NTv3vlrLlQ9P4Z2PFjJ2cG9+d8xWHLLlMP2mE71Kijh6wkiue2oasz+7CUP6de1JGkREZN2T1+CSmR0AXA4UAn9191+0WF4KXA9sC8wFjnH3abmup4hIPsxesJi/PfE+Nz49nUVLm9hn4yF8aa/12Xa07lrvvP4g/u+cgTz69hx+88DbfPM/L/OnR6fy9f024sDNqijQBam0w9+feJ9BfUr57BbVeSm/sMAY2q+MWfPXrcylpmbn7ldmcuXDU5k6eyHjhvTh8mO34uAtFFRqzYk7juZvT7zPzc9+yHn7bpDv6oiIiKwkb8ElMysE/gDsB9QAz5nZne7+RtrbTgfq3H2cmR0L/BI4Jve1FRHJnQ/m1nPV4+/yn+draGxq5uAthnHOnuuzSXW/fFetSzEz9tp4CHtsOJj7X5/FpQ++w5dveoHx1f345v4bstdGQ7pNZteypmYWLWnk08WNLFyS9rd4xb+fJv8uSpZ9uiQelxQW0KesiL6lRfQpK6JP8m/f5f8WL3+9b9rynjp1e3u9O2chj7w9h6/tu2FeP5PqijJmzl83Mpcam5q5KwkqvTdnERsO7cPvj9uagzarVmB4DcYM6s0eGw7muknTmLNwMb1Li+hTUkTv0iJ6lxYm/8ZvvHfJitf6lBZRWlTQbY6HIiLSPeUzc2l7YKq7vwdgZv8EDgPSg0uHARcnj28Bfm9m5u6ey4rmw/S5i5jz6RIKC2ylv6ICo8CMooICCgpY6d9CMwoLLf5N3l9gZKUx4e40OzQ2N9PcvPK/Te40NTuNTU6zO43NTnNz/NuU+vN4TTJraL+ynE2dnWvuTk1dA0ubmpPnAI47+PL3gKdeS15MPae196St2yHtfSv/n1xZ0tjMfyZ/yF2v1FJoxue3HcEXdx/LmEG9c1+ZbqSgwDhw82o+s2kVd7w0g989NIXTrp3MNqMq+eIe69O/18qzb7U8hbT2Vbf2/XvLd7bynmaHRUvTgkJJwGhR2uOFS5atHDha0sjiZc1t2tY+pSuCQ71Li+hdUkhjk/PhJ/UrBaQa23B8TQWl0tfZNxWAKiuiT2nximBU2vJtRvfvUQOpX/vkNEoKCzh+x/xO8V5VUcZrM+bntQ7Z1tjUzB0vzeT3j0zl/Y8XsXFVX/54/DYcsKmyDdvq3L3H8a3/vMy9r85i4ZJGlja27dhRWGD0KimMwNPyIFQhvUqSYFQqOFWyYlnqfb2KC5VJJiLSScP7l1NdUZ7vamSV5StOY2ZHAge4+xnJ8xOBHdz93LT3vJa8pyZ5/m7yno9Xt94Jffv65G23XfnFo4+GL30J6uvhoINW/U+nnBJ/H38MRx656vJzzoFjjoEPP4QTT1x1+Te+AYccAm+/DV/84qrLf/AD2HdfeOklOP/8VZf/7Gew887w1FPwve8B8P7Hi/goGdjzR/ucxRtDx7LLtJf4ylP/XOW/f2//c3lv4Aj2mfoMZz572yrLv3noN5ldOYSD33ycLzx/z/JgkwFmcOEJF7OwbyUHTr6f/Sfft8pF+5eP+zGLiko5+tm72P+Nx+N6Km2/Ofa46M145jO3ss+7z65U9uKiUk45+hIAvvLkzewy/eWVlteV9+Ocw2Obv/3YtWwz462Vltf2HcTXDvkmABc9dDXjZ7+30vL3Bgznewd8JT7G+65k7CczVlr+xpCx/GjfswD47V2/ofrTlXedF4ZvzK/2OAWAP932M/o3LFhp+ZOjt+TKXb4AwLX//iFljUtWWj5x/e35yw5HAPDPmy6gpbs33o0bt/ksZcsWc+1/Ll5l+S2b78stm+9L//r5/On2n6+y/MatD+LuTXanesEcfnv3pass/8v2hzNx3A6MnVvDz+7/PQB9yooZ3KeEgX1KKbrownbveyv53e9gq63goYfgJz9ZdflVV8FGG8Fdd8Glq9aPG26AkSPhX/+CP/1p1eW33AKDBsG118ZfS/feywcN8O4lv6L/PbezeFnTSot74r5XUGAM7VvGwMMOpM9Pov4ceCA0tMhoOPhg+GbUjz33ZBXd8Li3kk7se80Od37jF/zy9UVs+/QDnPDivatu3ue+S12vCo589SGOfPWhVZafctTFLC4u44QX7uHgt/63yvL27Hu7fvDKipsEZjT0reSq835Fn9IijrjlD6w35ZWVbgosqx7Ge7/9M31Ki1jvJ9+n/I3Xli8zgA03hKuvjsLOOgveeWel8n3LLVnym8v4dHEj5aedDDU1KwX2Z43fmsdP/TqfLm7kkB+dS+n8eXFDIFn+3Nit+eOuX+DTJY1cfdOFqxz3pu24F8f++4p40s33vcYLvssL0+sY2KeE9Qf3ieV5Ou5N/6SeWfMXs/0bk7DeveGPf4R//3vV///oo/Hvb34Dd9+98rLycvjvf+Pxj38MEyeuvHzgQPi//4vH3/0uTJq08vIRI+DGG+Px+efHZ5huLfseW20Vnx/ACSdATQ0Qv8mPFy7hgYr1+MH2xzO+uh9/v/sXDF22iJVCFvvsAxdeGI913Ft1eSv7XrNDc3JTb8YVVzFvUBXlt97C0H9cs/w3n/r9//PblzGntC+bP3Qb2z12F03Jzb/Ue047+mIWWElGjnvd5Zyb0h3be+mu3PlYnhyzFeM/eo+LJl69yvJf7X4yL4zYhG1q3uTbj1+3yvLOXmt87eBvUNtvMAe/+Xjez7na91amfa/r7HuHNXzA8Mq04FKWzrnL7bQT/Dz5zj//eZg7d+XlHTzn2mOPPe/uE2hFjxjQ28zOAs4C2KK0NM+1yYyqijL69y7BHS46eDzzN96UyqcWsv7UvqQHfhw4b98NmDdyLNX/q2XUu72X35VPveeEHUdRN7CKjZa9weB3SlfJ2NhmVCUL+lQy4t1y+pYWgaUCT3Ehc+hWw2gq68X2cwZQPaMsGoLJMjO44MCNKTRjy0+HMmp+75WWN5eVcfmxW1FYYGy48DEGLX4PM7CkkMbK/txw+vYAjKl7gL7NM1f6HMZWDVy+fOysO+n95soH7NFjhixfPu79IZRP+3Sl5SM3qWKDZPlGbw6kZNbSlZYP32IYmyfLN3mxP0XzWnwP245g+2T5pk9VULBk5fEwhuwwkt2T5Rs/smqXpUE7j2b/47enoKGejZ9YdfmZu4/lsCO2p+iTuWz8zKrLv7zXOI757PaU1M5go+dXXf71/TbklL23p/y9AYx7pR+LljTy8cIlvP/xIqbNrWfSg+8wvnoz9mpspniV/911NTY7cxcu4Wt/eZqnahs44a3ZHFdUQFVFGUVpd0+v+MLWGLD+smcYurDvSutoLi3jzydsAxjrL3qMgYvfI3UVY8Cyyv78/ZQJGMbYT+6nX9PMZN+N96w3NPf7ngG9S4tiG0t7xOE5LwoMPrf1cA44cBjT//QeVbNW/e38+YRtaRwwkCH/9x5DPl51+bWnbkdzeS+qbnydgfNXbiyYwT/P2hGAaptM/4b0hqph5eU89q096VNaRMWlkyl6pMXJfuAAdjhlu3j8ymCY2yLbsLKc4RsMjse9SqCdM+GZGWXFhZFdVFq0yr40eHgFm+81Lp78tQJKG1davsWu63H6hfsB0Dz5cpoWNSQXp83MXbSUiXMWMXnaJz1ixqrZny6m2Z2qivwPjlxSWIC788mipQzs3XOyFT9ZtJTpn9SzZFkTpQMKuPrEbdlv/FDs8TKYuyjf1ev2Cowko90YO7gPjOwP1f2gd8kq7/32ARtHYHPRi/B231WWv/LD/VlWWkbjle9RUP/aSpnmDsvPecN5jgH1K1+gN5eWLV8+cvH/qFz6/krL1d7LbHuvpQsO2Ih5u2xP7zfKGfvGqssvOmQTPt1me/q+AGOmrLr8J5/blEXjN6fyyQZGvr/q8l8duQUNY8cx4OE6hn+46vLfHbsVS6uHM+iemVS3cs794/HbxDn31vcY2so5928nT6C5vBfVZa8xqMU5F7Tvad/rGfvewNc/WmX9PU0+M5d2Ai529/2T598FcPefp73n/uQ9k8ysCJgFDF5Tt7gJEyb45MmTs1t5kS7M3XltxgJufbGGO1+aydxFS+nfq5hDthzG4VsPZ6uRlV1y3IUljU088tZsbntxBo+8NYelTc1sMKQPh28znMO2Gr5ypF9E8qJ+aSO7/fIRxg/rxw2n75Dv6nRKY1Mze/z6UUYOKOefZ+2U7+pw32u1nH3jC9z9lV3ZbHhFvquTEe7Opj+8n2GV5Xz3wI3Ze+PuMw6aiIiIrMrMumTm0nPABma2HjADOBY4rsV77gROBiYBRwIPrwvjLYl0hpmx+YgKNh9RwfcO2oT/TZnDrS/M4F/Pfcj1k6YzdlBvDt96OJ/benjex2dydyZPr+PWF2ZwzyszWbC4kUF9Sjlxp9EcvvVwNh3WTxciIl1Ir5IivrjHWH5271s8P/2Tbj1z4YNvfMSMeQ1cdMj4fFcFgKpkHIba+Yt7THBpzsIl1C9t4qSdRrPPJkPzXR0RERHJorwFl9y90czOBe4HCoG/u/vrZvYjYLK73wn8DbjBzKYCnxABKBFpo+LCAvbeeCh7bzyUBYuX8d9Xa7n1hRlc+uA7XPrgO2y/3gCO2Ho4B21RTb+y3HWce3fOQm5/cQa3vTiDmroGyosL2X/ToRy+zQh2WX8gRYXt6wYkIrlzwo6jueqx9/jdQ1O6dfbS3598n5EDytm3iwQ9hiVd82b1oBnjaupiW0b0V+apiIhIT5fXQT3c/V7g3havXZT2eDFwVK7rJdIT9Ssr5pjtRnHMdqP48JN67nhpBre+OIMLbn2Vi+58nf3GD+WIrYez+4aDKc5CcOfjhUu4++WZ3PbiDF6umU+BwS7jBvH1/TZk/02r6K0xhkS6hV4lRZy1+1h+/t+3eH56HduO7p/vKrXbqzXzeW5aHRcePL7LzII1sE8pRQXGzPmL1/7mbmJFcKlnzmIqIiIiK+hqTmQdNHJAL87dewO+vNc4XqmZz60v1HDXK7Xc80otA3uXcMiWwzhim+FsPryiU93SFi9r4oE3PuL2F2fw2DtzaGp2xlf34wef3YRDthzG0H75H0RXRNrvxJ1Gc9Xj73H5xClcf9r2+a5Ou13z5Pv0KS3i6Akj8l2V5QoLjKH9ypjVo4JL9QAaM09ERGQdoOCSyDrMzNhyZCVbjqzkBweP57G353DrizXc9MwHXPvUNMYN6bN8fKa2Xhw0NztPvzeX216cwX9fm8XCJY1UV5Rx5m5jOXzr4WxUteoMNSLSvfQqKeLM3cbyy/ve4sUP6th6VPfJXpq9YDF3vTKT43cYTd8cdgdui+qKMmbO6znd4mbUNTCgd4kyU0VERNYBOtuLCBDjM+07fij7jh/K/Ppl3PNqLbe9WMOv73+b3zzwNjuuN5DDtxnOgZtVtXpB9vasT5fPUFc7fzF9Sos4cLMqDt96ODuMHdhlup6ISGactNNorn78XS6fOIVrT+0+2Us3Pj2dxmbnlJ3H5Lsqq6iuLOeVmnn5rkbG1NQ1aLwlERGRdYSCSyKyiopexRy3wyiO22EUH8yt57YXZ3DbizV8+5ZXuOiO1/jM+CoO32Y4Gw3tyz2v1HLrizN4s3YBhQXGHhsO5nsHbcK+mwylvKQw35siIlnSu7SIM3cfy6/ue5uXPpzHViMr812ltVq8rIl/PPMB+2w8lDGDeue7OqsYVlHG/a8vxt17xEyZNXX1bDhU2aoiIiLrAgWXRGSNRg3sxXn7bsBX9xnHix/O49YXarj7lVrufHnm8vdsOaKCiw8Zz8FbDmNQn9I81lZEcumkncbwl8ff4/KH3uGabpC9dOfLM5m7aCmn7TIm31VpVVVFGUsbm/lk0VIGdvNjqbtTU9fA3hsPyXdVREREJAcUXBKRNjEzthnVn21G9eeigzflkbdn8/7Hi9hv/FDWH9wn39UTkTzoU1rEGbuN5df3d/3sJXfn70+8z8ZVfdlp/YH5rk6rqiuiC1nt/MXdPrj08cKlLGls1kxxIiIi64jMzzcuIj1eSVEB+29axdl7rK/Aksg67uSdx1DZq5grJk7Jd1XWaNJ7c3lr1qectst6XbbLWXVFzKBZ2wNmjEvNFKcxl0RERNYNCi6JiIhIh/UpjZnjHn5rdpcejPqaJ6cxoHcJh241LN9VWa3qylRwqfvPGFdTF9ugzCUREZF1g4JLIiIi0ikn7TSaivJiLn+oa2YvTZ+7iIfe/IjjdxhFWXHXnWhgUO9Siguth2QuRXBpuDKXRERE1gkKLomIiEin9C0r5szd1mPiW7N5tWZ+vquzimufmkZRgXHCjqPzXZU1KigwhvYro3ZeT8hcqqd/r2L6lGp4TxERkXWBgksiIiLSaSfvPCaylya+k++qrOTTxcv4z+QaPrt5NUP7leW7OmtVXVHWIzKXZsxrUJc4ERGRdYiCSyIiItJpfcuKOX3X9Xjozdm8NqPrZC/9Z3INC5c0ctqu6+W7Km1SXVHeI4JLNXUNDK9UlzgREZF1hYJLIiIikhGn7DKGfmVFXN5FZo5ranaufWoaE0b3Z4sRlfmuTptUV5Qxa/5i3D3fVekwd6emrl4zxYmIiKxDFFwSERGRjOhXVszpu47lwTc+6hLZSw+/NZsPPqnvNllLEMGlpU3NzF20NN9V6bC5i5ayeFmzgksiIiLrEAWXREREJGNS2UtXdIHspb8/8T7DK8v5zPih+a5Km1VVREBmVjfuGpeaKU5jLomIiKw7FFwSERGRjKkoL+a0XdfjgTc+4vWZ+cteerN2AZPem8tJO42mqLD7NHeGVcag4zO78YxxNXX1AIwYoMwlERGRdUX3aW2JiIhIt3DqLuvRN8/ZS9c8+T7lxYUcu92ovNWhI6oqIrjUnQf1TmUuaUBvERGRdYeCSyIiIpJRFeXFnLbLetz/+ke8MXNBzsv/eOESbn9pJp/fdjgVvYpzXn5nDOpdSnGhdfPgUj2VvYrpW9a9PnsRERHpOAWXREREJONO22U9+pbmJ3vppmc+YGljM6fs3H0G8k4pKDCG9iujdn737RY3o65Bg3mLiIisYxRcEhERkYyr6FXMqbuux32vz+LN2txlLy1tbOaGp6ez50aDGTekT87KzaRhFeXdPHOpQV3iRERE1jEKLomIiEhWnJ6H7KV7Xp3JnE+XcOou3S9rKaWqovtmLrk7NXUNmilORERkHaPgkoiIiGRFRa9iTtllDP99bRZvzcp+9pK787cn3mfckD7svsGgrJeXLdWVZXw0fwnNzZ7vqrTbJ4uW0rCsSd3iRERE1jEKLomIiEjWnL7revQpLeLKiVOzXtbk6XW8NmMBp+4yBjPLennZMqyinKVNzcxdtDTfVWm31ExxylwSERFZtyi4JCIiIllT2auEU3Yewz2v1vL2rE+zWtY1T75PRXkxR2w9IqvlZFtVRRkAs7rhuEsrgkvKXBIREVmXKLgkIiIiWZXKXrri4eyNvVRTV899r83iC9uPorykMGvl5MKwigjMzOyG4y7V1NUDMFzBJRERkXWKgksiIiKSVf17l3DyzqO599Va3vkoO9lL10+ajplx0k6js7L+XOrumUsV5cX0KyvOd1VEREQkhxRcEhERkaw7Y9ex9CouzMrMcYuWNPLPZz/gwM2qGFbZ/TNmBvYuoaSwoNtmLqlLnIiIyLpHwSURERHJusheirGXpmQ4e+nWF2pYsLiRU3dZL6PrzZeCAmNoRWm3zFyaMa9BwSUREZF1kIJLIiIikhNn7DaW8uJCrng4czPHNTc71zw5jS1HVrLNqMqMrTffqivKqZ3XvYJL7k5NXQPDKzVTnIiIyLpGwSURERHJiQG9SzhppzHc/cpMps7OTPbSY1Pm8N7HizhtlzGYWUbW2RVUV5RRu6B7dYurq19G/dImZS6JiIisgxRcEhERkZw5c7f1IntpYmayl/7+xPsM7VfKQZtXZ2R9XUV1RTmz5i+mudnzXZU2S80Up+CSiIjIukfBJREREcmZgX1KOXGn0dz1ykymzl7YqXVN+ehT/jflY07aaQzFhT2rSVNdUcayJmfuoqX5rkqb1dRFptWI/uoWJyIisq7pWS0xERER6fLO2m0sZUWF/P7hzs0cd81T0ygtKuAL24/KUM26juqKMgBqu9GMcanMpeHKXBIREVnnKLgkIiIiOTWwTykn7TSaO1+eybtzOpa9VLdoKbe+UMPhWw9nQO+SDNcw/6orIkBT241mjKupa6BfWREV5cX5roqIiIjkmIJLIiIiknNn7j6W0qJCft/BmeNufu4DFi9r5tRd1stwzbqG6sokc2led8pcalCXOBERkXWUgksiIiKSc4OSsZfueGkG77Uze2lZUzPXPzWdXccNYqOqvlmqYX4N6FVCSWEBtQu6T+bSjLoGDeYtIiKyjlJwSURERPLizN3GUlJU0O7spftem8WsBYs5bdcx2alYF1BQYFRVlFE7r3sEl9ydmrp6jbckIiKyjlJwSURERPJicN9STtxxNLe/NIP3P17U5v/39yffZ71BvdlzwyFZrF3+VVWUdZsBvefVL2PR0iZ1ixMREVlHKbgkIiIieXPW7utTUlTAlW2cOe7FD+p48YN5nLLzGAoKLMu1y69hFWXdZkDvmroIgqlbnIiIyLpJwSURERHJm8F9Szl+h9Hc8dJMprUhe+maJ6fRt6yII7cdkYPa5Vd1ZTkfLVhMc7PnuyprVVNXDyi4JCIisq5ScElERETy6ot7jKWowLhyLWMv1c5v4N5Xazlmwkh6lxblqHb5U11RxrIm5+NFS/JdlbVakbmkbnEiIiLrIgWXREREJK+G9C3j+B1i7KXpc1efvXTDpOk0u3PyzmNyV7k8qq6ILKDuMKh3TV09fcuKqCgvzndVREREJA8UXBIREZG8OzvJXlrdzHENS5u4+dkP+Mz4KkYOWDeyY6orygC6xbhLNXUNyloSERFZhym4JCIiInk3pF8Zx+0wiltfbD176faXZlBXv4xTdxmT+8rlyYrgUtefMS6CSxpvSUREZF2l4JKIiIh0CWfvsT6FBcYfHlk5e8nduebJ99l0WD+2X29AnmqXewN6l1BSVMCsLp655O7MmKfgkoiIyLpMwSURERHpEob2K+O47Ufxfy/M4IO59ctff3LqXN75aCGn7bIeZpbHGuaWmVFdUcbMLh5cmt+wjIVLGhleqeCSiIjIukrBJREREekyztlz1eylvz/5PoP6lHLwltV5rFl+VPUrY1YX7xanmeJEREREwSURERHpMlZkL9Xw4Sf1vDdnIQ+/NZsTdhxFaVFhvquXc8Mqy5nZxWeLq6mLLDN1ixMREVl3tSm4ZGa7mFnv5PEJZnaZmY3uaKFmNsDMHjSzKcm//VfzvvvMbJ6Z3d3RskRERKR7OXuP9SmwyF667qlplBQWcPwOHW52dGtVFWV8tGAxzc2e76qsVipzaaQyl0RERNZZbc1c+hNQb2ZbAt8A3gWu70S5FwAT3X0DYGLyvDW/Bk7sRDkiIiLSzVRVlHHs9iO55fka/j25hkO3GsbgvqX5rlZeDKsoo7HZ+XjhknxXZbVq6hroW1pEv/KifFdFRERE8qStwaVGd3fgMOD37v4HoG8nyj0MuC55fB3wudbe5O4TgU87UY6IiIh0Q+fsGdlLDcuaOHWXMfmuTt5UVURXs9ouPKh3TV09w/uXr1ODrYuIiMjK2nqL6VMz+y5wArC7mRUAxZ0od6i71yaPZwFDO7EuzOws4CyAUaNGdWZVIiIi0gVUV5Tzlb3HMf2TejYdVpHv6uRNdUUZALXzG9hyZGV+K7MaNXUNGsxbRERkHdfW4NIxwHHA6e4+y8xGEV3WVsvMHgKqWln0/fQn7u5m1qmBBNz9auBqgAkTJnTdQQlERESkzb6yzwb5rkLerQgudc3MJXdnRl0DO44dmO+qiIiISB61Nbj0NXf/TuqJu39gZpuu6T+4+76rW2ZmH5lZtbvXmlk1MLuN9RARERFZZwzoXUJJUUGXDS4taGjk0yWNmilORERkHdfWMZf2a+W1AztR7p3Aycnjk4E7OrEuERERkR7JzKiuKOuywaUP6+oBFFwSERFZx60xuGRm55jZq8BGZvZK2t/7wKudKPcXwH5mNgXYN3mOmU0ws7+mlf8/4D/APmZWY2b7d6JMERERkW6nuqKM2nkN+a5Gq2rqol4ac0lERGTdtrZucTcB/wV+DlyQ9vqn7v5JRwt197nAPq28Phk4I+35bh0tQ0RERKQnqK4o59n3O9zsyqoaZS6JiIgIawkuuft8YD7wBTMrJGZ1KwL6mFkfd/8gB3UUERERWWdVV5Tx0YLFNDU7hQWW7+qspKaugT6lRVSUd2YSYREREenu2jSgt5mdC1wMfAQ0Jy87sEV2qiUiIiIiANWV5TQ2Ox8vXMLQfmX5rs5KauoaGNG/HLOuFfQSERGR3GrrbHHnAxsl3dlEREREJEeqk4BS7fzFXTC4VK8ucSIiItLm2eI+JLrHiYiIiEgOVVcmwaUuOKj3jLoGDeYtIiIibc5ceg941MzuAZakXnT3y7JSKxEREREBYkBviMylrmR+wzI+XdLI8EplLomIiKzr2hpc+iD5K0n+RERERCQH+vcqprSogNr5XStzSTPFiYiISEqbgkvufgmAmfVy9/rsVklEREREUsyM6oqyLpe5VFMXwS51ixMREZE2jblkZjuZ2RvAW8nzLc3sj1mtmYiIiIgA0TWu6waXlLkkIiKyrmvrgN6/A/YH5gK4+8vA7lmqk4iIiIikqa4oY1aXCy7V07ukkMpexfmuioiIiORZW4NLuPuHLV5qynBdRERERKQV1ZVlzFqwmKZmz3dVlqtJZoozs3xXRURERPKsrcGlD81sZ8DNrNjMvgm8mcV6iYiIiEiiqqKcpmbn44VL1v7mHIngkrrEiYiISNuDS2cDXwaGAzOArZLnIiIiIpJlwyrKAJg5r+vMGFdTV6/gkoiIiABtny3uY+D4LNdFRERERFpRlQSXusq4S/MblvHp4kbNFCciIiLAWoJLZvZtd/+VmV0JrNLJ392/mrWaiYiIiAgAwyoiQ2hmFwkuzUhmihuuzCURERFh7ZlLqXGVJme7IiIiIiLSuspexZQWFTBrftfoFldTVw+gbnEiIiICrCW45O53Jf9el5vqiIiIiEhLZsawyvIuk7lUk2QuqVuciIiIQBsH9DazB82sMu15fzO7P2u1EhEREZGVVPUr6zJjLtXUNdCrpJD+vYrzXRURERHpAto6W9xgd5+XeuLudcCQrNRIRERERFZRXVlGbReZLS41U5yZ5bsqIiIi0gW0NbjUZGajUk/MbDStDPAtIiIiItkxrKKcjz5dQlNz/ptgNXUN6hInIiIiy61tQO+U7wNPmNljgAG7AWdlrVYiIiIispKqijKamp05ny6hqqIsr3Wpqatnwpj+ea2DiIiIdB1tCi65+31mtg2wY/LS+e7+cfaqJSIiIiLphlVGQGnm/Ia8BpcWLF7GgsWNmilORERElltjtzgz2zj5dxtgFDAz+RuVvCYiIiIiOVDVL4I5+R7Ue0YyU9zwSnWLExERkbC2zKWvE93fLm1lmQN7Z7xGIiIiIrKK5ZlLeR7UuyYJLilzSURERFLWFlx6MPn3dHd/L9uVEREREZHWVZQXU1ZckPfMpZq6ekDBJREREVlhbbPFfTf595ZsV0REREREVs/MGFZRTm3eg0sNlBcXMqB3SV7rISIiIl3H2jKXPjGzB4CxZnZny4Xufmh2qiUiIiIiLVVVlFE7P9/d4uoZ0b8cM8trPURERKTrWFtw6SBgG+AGWh93SURERERypLqinKfeze+EvTV1DeoSJyIiIitZW3Dpb+5+opn9xd0fy0mNRERERKRV1RVlzP50CY1NzRQVrm10g+yoqWtgm1H981K2iIiIdE1ra5Vsa2bDgOPNrL+ZDUj/y0UFRURERCRUV5bR1OzMWbgkL+V/ungZ8xuWKXNJREREVrK2zKU/AxOBscDzQHrnek9eFxEREZEcqK4oA6B2/mKqK3If4JkxL8Z7GtG/V87LFhERka5rjZlL7n6Fu28C/N3dx7r7eml/CiyJiIiI5FAqoFQ7Lz8zxtV8EsGl4cpcEhERkTRt6qzv7ueY2a5mdiqAmQ0ys/WyWzURERERSbcicyk/M8bV1NUDqFuciIiIrKRNwSUz+yHwHeC7yUslwI3ZqpSIiIiIrKqivJjy4kJq5+cpc6mugbLiAgb2LslL+SIiItI1tXWakcOBQ4FFAO4+E+ibrUqJiIiIyKrMjOqKMmblMbg0on8vzGztbxYREZF1RluDS0vd3YlBvDGz3tmrkoiIiIisTnVlGTPz1S1uXr26xImIiMgq2hpc+reZXQVUmtmZwEPAX7JXLRERERFpTVW/8jxnLim4JCIiIisrasub3P03ZrYfsADYCLjI3R/Mas1EREREZBXDKsv4aMFiGpuaKSps633Czvt08TLm1S9jRP9eOStTREREuoc2BZcSrwClyeOXs1AXEREREVmL6opymh3mLFxCdUXusohmzIuueMpcEhERkZbaOlvc0cCzwFHA0cAzZnZkNismIiIiIquqrigDYOa83HaNm1EXwaXhlQouiYiIyMramrn0fWA7d58NYGaDiXGXbslWxURERERkVdWVEVzK9bhLNXWpzCV1ixMREZGVtbWjfkEqsJSY247/KyIiIiIZUt0vModqczxjXE1dPaVFBQzqU5LTckVERKTra2vm0n1mdj9wc/L8GODe7FRJRERERFanX3kRvUoKc94tLjVTnJnltFwRERHp+tYYXDKzccBQd/+WmR0B7JosmgT8I9uVExEREZGVmRlVFWXMWpDrzKUGdYkTERGRVq2ta9vvgAUA7n6ru3/d3b8O3JYsExEREZEcG1ZRnofMpXrNFCciIiKtWltwaai7v9ryxeS1MVmpkYiIiIisUVVFWU4H9F64pJG6+mXKXBIREZFWrS24VLmGZbp1JSIiIpIHwyrKmP3pYhqbmnNS3ozlM8Wp+SciIiKrWltwabKZndnyRTM7A3g+O1USERERkTWpqiin2WH2p0tyUt6MefUADFdwSURERFqxttnizgduM7PjWRFMmgCUAId3tFAzGwD8i+haNw042t3rWrxnK+BPQD+gCfipu/+ro2WKiIiI9BTVlWUA1M5vYFhl9gM+NcpcEhERkTVYY+aSu3/k7jsDlxBBoGnAJe6+k7vP6kS5FwAT3X0DYGLyvKV64CR33xQ4APidmVV2okwRERGRHqG6IhVcys24SzV1DZQWFTC4T2lOyhMREZHuZW2ZSwC4+yPAIxks9zBgz+TxdcCjwHdalPlO2uOZZjYbGAzMy2A9RERERLqd6orIIKrN0YxxNXX1DO9fjpnlpDwRERHpXtY25lK2DHX32uTxLGDomt5sZtsTXfHeXc3ys8xssplNnjNnTmZrKiIiItLF9CsroldJYU4zlzRTnIiIiKxOmzKXOsLMHgKqWln0/fQn7u5m5mtYTzVwA3Cyu7c6JYq7Xw1cDTBhwoTVrktERESkJzAzqivKqJ3fkJPyauoa2Gx4RU7KEhERke4na8Eld993dcvM7CMzq3b32iR4NHs17+sH3AN8392fzlJVRURERLqd6orynGQuLVrSyCeLlmowbxEREVmtfHWLuxM4OXl8MnBHyzeYWQlwG3C9u9+Sw7qJiIiIdHm5ylyaMS81U5y6xYmIiEjr8hVc+gWwn5lNAfZNnmNmE8zsr8l7jgZ2B04xs5eSv63yUlsRERGRLqa6spzZny5hWVOrowZkTE1dPYAyl0RERGS1stYtbk3cfS6wTyuvTwbOSB7fCNyY46qJiIiIdAvVFWW4w+xPlzC8MnuBnxl1SeZSFssQERGR7i1fmUsiIiIi0gnVFWUAzMpy17iaugZKigoY1Kc0q+WIiIhI96XgkoiIiEg3VF0RmUQz52V3UO+augZGVJZTUGBZLUdERES6LwWXRERERLqh6spU5lK2g0v1DNd4SyIiIrIGCi6JiIiIdEN9S4voXVLIzBx0i9NMcSIiIrImCi6JiIiIdENmRnVlObVZ7BZXv7SRuYuWaqY4ERERWSMFl0RERES6qeqKMmoXZC+4tHymOAWXREREZA0UXBIRERHppqoryqidl71ucTXLg0vqFiciIiKrp+CSiIiISDdVVVHOnIVLWNbUnJX119TVA8pcEhERkTVTcElERESkmxpWUYY7fJSlrnE18xooKSxgcJ/SrKxfREREegYFl0RERES6qaqKMgBmzc9ScKmugeH9yykosKysX0RERHoGBZdEREREuqlhldFdbWYWg0vqEiciIiJro+CSiIiISDe1InMpO4N6z6irV3BJRERE1krBJREREZFuql9ZMX1Ki5g5L/OZSw1Lm/h44VLNFCciIiJrpeCSiIiISDdWVVGWlTGXZszTTHEiIiLSNgouiYiIiHRj1RVl1GahW9yHdbFOBZdERERkbRRcEhEREenGIriU+cylmuXBJXWLExERkTVTcElERESkG6uuKGfOwiUsbWzO6Hpn1DVQUljA4D6lGV2viIiI9DwKLomIiIh0Y8Mqy3CH2Z9mNnuppq6eYZVlFBRYRtcrIiIiPY+CSyIiIiLdWFVFjImU6a5xNXUN6hInIiIibaLgkoiIiEg3NqyiDMhWcEmDeYuIiMjaKbgkIiIi0o1VpYJL8zI3Y9ziZU18vHCJgksiIiLSJgouiYiIiHRjfcuK6VtalNHMJc0UJyIiIu2h4JKIiIhIN1dVUUbt/MxlLtXU1QMoc0lERETaRMElERERkW6uurKcWcpcEhERkTxRcElERESkm6vuV8bMDAeXiguNIX1LM7ZOERER6bkUXBIRERHp5qory/h44RKWNjZnZH0z5jUwrLKcggLLyPpERESkZ1NwSURERKSbq64owx0+WpCZ7KWaunqNtyQiIiJtpuCSiIiISDdXXRGBoEzNGFdT18CISo23JCIiIm2j4JKIiIhIN1ddUQaQkRnjFi9rYs6nS5S5JCIiIm2m4JKIiIhIN1ddmbnMpRnzkpniBii4JCIiIm2j4JKIiIhIN9entIi+pUXMykBwqaYuCS71V7c4ERERaRsFl0RERER6gOrKMmbO63y3uJq6egB1ixMREZE2U3BJREREpAeoqihnVgZmi6upa6CowBjStywDtRIREZF1gYJLIiIiIj3AsIoyZs7LwJhLdQ0MqyynsMAyUCsRERFZFyi4JCIiItIDVFeU8/HCJSxtbO7Uemrq6tUlTkRERNpFwSURERGRHqC6IrqxfdTJrnE1dQ0KLomIiEi7KLgkIiIi0gNUV0ZwqbYTM8YtXtbE7E+XaKY4ERERaRcFl0RERER6gFTmUu38js8Yl5ptTplLIiIi0h4KLomIiIj0AFUVERDqTOZSTV0quKTMJREREWk7BZdEREREeoA+pUX0LSuidl7HM5dWBJeUuSQiIiJtp+CSiIiISA8xrKK8k5lL9RQVGEP7lWWwViIiItLTKbgkIiIi0kNUVZR1ultcdWUZhQWWwVqJiIhIT6fgkoiIiEgPMayyc8GlGfMaGFGp8ZZERESkfRRcEhEREekhqvqV8/HCJSxpbOrQ/6+pq9d4SyIiItJuCi6JiIiI9BDVlTFW0uwFS9r9f5c0NvHRgiWaKU5ERETaTcElERERkR6iuiKCSzM7MGPczHnRnU6ZSyIiItJeCi6JiIiI9BDVFREY6si4SzV19YCCSyIiItJ+eQkumdkAM3vQzKYk//Zv5T2jzewFM3vJzF43s7PzUVcRERGR7iKVudSx4FJkO40YoG5xIiIi0j75yly6AJjo7hsAE5PnLdUCO7n7VsAOwAVmNix3VRQRERHpXnqXFtGvrIja+e3vFldTV09RgTG0b2kWaiYiIiI9Wb6CS4cB1yWPrwM+1/IN7r7U3VOjUZaiLnwiIiIia1VdUd7hzKWqijKKCtXkEhERkfbJV+thqLvXJo9nAUNbe5OZjTSzV4APgV+6+8zVvO8sM5tsZpPnzJmTnRqLiIiIdAPVlWUdylyaUdeg8ZZERESkQ7IWXDKzh8zstVb+Dkt/n7s74K2tw90/dPctgHHAyWbWahDK3a929wnuPmHw4MEZ3xYRERGR7qK6ooxZHcxcGtFf4y2JiIhI+xVla8Xuvu/qlpnZR2ZW7e61ZlYNzF7Lumaa2WvAbsAtGa6qiIiISI9RXVHOxwuXsqSxidKiwjb9nyWNTXz06WJlLomIiEiH5Ktb3J3Aycnjk4E7Wr7BzEaYWXnyuD+wK/B2zmooIiIi0g2lZoz7aP6Stbxzhdp5i3FHmUsiIiLSIfkKLv0C2M/MpgD7Js8xswlm9tfkPZsAz5jZy8BjwG/c/dW81FZERESkm6iuiOyjme0Yd6mmLt6rzCURERHpiKx1i1sTd58L7NPK65OBM5LHDwJb5LhqIiIiIt1adWVkLrVn3KWaunpAwSURERHpGM01KyIiItKDpLrFtTdzqbDAqOpXlq1qiYiISA+m4JKIiIhID9KrpIiK8uJ2Zy5VV5RRVKimoYiIiLSfWhAiIiIiPUx1RRkz57UnuNTA8Ep1iRMREZGOUXBJREREpIeprihj1oK2d4ubMa9BM8WJiIhIhym4JCIiItLDVFWUU9vGzKWljc3MWrBYg3mLiIhIhym4JCIiItLDDKsoY+6ipSxe1rTW99bOb8BdM8WJiIhIxym4JCIiItLDVCUzxn20YO3ZSzV10X1O3eJERESkoxRcEhEREelhhiWDc9e2Yca4mrp6QJlLIiIi0nEKLomIiIj0MKnMpdr5ax/Uu6augcICozr5PyIiIiLtpeCSiIiISA+TChTNbMOg3jV1DVT1K6OoUM1CERER6Ri1IkRERER6mF4lRVSUFzOrjd3ihqtLnIiIiHSCgksiIiIiPVB1RVmbusXNqGvQeEsiIiLSKQouiYiIiPRAEVxac+bS0sZmZi1YrJniREREpFMUXBIRERHpgaory9caXJo1fzHNrpniREREpHMUXBIRERHpgYZVlPHJoqUsXta02vfU1NUDCi6JiIhI5yi4JCIiItIDVVVEwGhNg3rX1MWYTCPVLU5EREQ6QcElERERkR5oWEUZwBq7xtXU1VNgUJW8V0RERKQjFFwSERER6YGqlgeXVj9jXE1dA9UV5RQXqkkoIiIiHaeWhIiIiEgPVJ10i1tz5lIDwzXekoiIiHSSgksiIiIiPVB5SSGVvYrXkrlUz4hKBZdERESkcxRcEhEREemhqivKVzug97KmZmYtWKyZ4kRERKTTFFwSERER6aGqK8qYOa/14NKs+YtpdhihmeJERESkkxRcEhEREemhqivKmLWg9eDSh3X1AMpcEhERkU5TcElERESkh6quKOOTRUtZvKxplWU1dTEWkzKXREREpLMUXBIRERHpoVIzxrU27lJNXQMFBlUVZbmuloiIiPQwCi6JiIiI9FDVSeBoZiszxtXU1VPVr4ySIjUHRUREpHPUmhARERHpoaor15y5pC5xIiIikgkKLomIiIj0UFX9InOptpXg0oy6BoZrMG8RERHJAAWXRERERHqo8pJC+vcqprZFt7jGpmZmLVismeJEREQkIxRcEhEREenBqirKqZ23cuZS7fzFNDW7gksiIiKSEQouiYiIiPRgwyrKmNmiW1xNXWQyacwlERERyQQFl0RERER6sKqKMma16BZXU1cPoMwlERERyQgFl0RERER6sGGV5dTVL6NhadPy12rqGjCD6goFl0RERKTzFFwSERER6cGqK2LGuFkLVnSNq6lroKpfGSVFagqKiIhI56lFISIiItKDVSXBpdp5K7rG1dTVq0uciIiIZIyCSyIiIiI92LCk61vt/JUzl4ZXKrgkIiIimaHgkoiIiEgPtjxzKRnUu7GpmVkLFmumOBEREckYBZdEREREerCy4kIG9C5Znrk0a8Fimppd3eJEREQkYxRcEhEREenhqvqVLQ8u1dRFBpMyl0RERCRTFFwSERER6eGGVbYWXFLmkoiIiGSGgksiIiIiPVxVRdnyMZdq6uoxg+rKsjzXSkRERHoKBZdEREREerjqinLm1S+jYWkTNXUNDO1bRmlRYb6rJSIiIj2EgksiIiIiPVx12oxxNXX16hInIiIiGaXgkoiIiEgPV10RwaRZ8xdTU9eg4JKIiIhklIJLIiIiIj1cKnPpw7p6aucvZriCSyIiIpJBCi6JiIiI9HBVSXDppQ/n0dTsjOjfK881EhERkZ5EwSURERGRHq6suJABvUt4blodgLrFiYiISEblJbhkZgPM7EEzm5L8238N7+1nZjVm9vtc1lFERESkJ6muKGPq7IUAylwSERGRjMpX5tIFwER33wCYmDxfnR8Dj+ekViIiIiI9VGrcJYBhlWVreKeIiIhI++QruHQYcF3y+Drgc629ycy2BYYCD+SmWiIiIiI9U2rGuKH9SiktKsxzbURERKQnyVdwaai71yaPZxEBpJWYWQFwKfDNta3MzM4ys8lmNnnOnDmZramIiIhID1CdZCupS5yIiIhkWlG2VmxmDwFVrSz6fvoTd3cz81be9yXgXnevMbM1luXuVwNXA0yYMKG1dYmIiIis01Ld4jSYt4iIiGRa1oJL7r7v6paZ2UdmVu3utWZWDcxu5W07AbuZ2ZeAPkCJmS109zWNzyQiIiIirUh1ixteqeCSiIiIZFbWgktrcSdwMvCL5N87Wr7B3Y9PPTazU4AJCiyJiIiIdMyoAdEdbr1BvfNcExEREelp8jXm0i+A/cxsCrBv8hwzm2Bmf81TnURERER6rGGV5dz2pZ353NbD810VERER6WHMvWcNUTRhwgSfPHlyvqshIiIiIiIiItJjmNnz7j6htWX5ylwSEREREREREZEeQMElERERERERERHpMAWXRERERERERESkwxRcEhERERERERGRDlNwSUREREREREREOkzBJRERERERERER6TAFl0REREREREREpMMUXBIRERERERERkQ5TcElERERERERERDpMwSUREREREREREekwBZdERERERERERKTDFFwSEREREREREZEOM3fPdx0yyszmANPzXY8MGQR83IPLy0eZKk/ldfUyVV73Li8fZao8ldfVy1R53bu8fJSp8lReVy9T5XXv8vJRZj62MRtGu/vg1hb0uOBST2Jmk919Qk8tLx9lqjyV19XLVHndu7x8lKnyVF5XL1Plde/y8lGmylN5Xb1Mlde9y8tHmfnYxlxTtzgREREREREREekwBZdERERERERERKTDFFzq2q7u4eXlo0yVp/K6epkqr3uXl48yVZ7K6+plqrzuXV4+ylR5Kq+rl6nyund5+SgzH9uYUxpzSUREREREREREOkyZSyIiIiIiIiIi0mEKLomIiIiIiIiISIcpuCQiWWFmlu86iIiIiIiISPYpuCTLKRiQW5n8vM2sKFPryhTP4IBuZtZljlVmVprvOkjXZGbl+a6DSEdZIt/1kK7LzIrzUOYOyb/aN0VEurguc8EmK6ROoGZWmMtyMxkMyJRsNCbMrLeZVWd6ve0of3MzG5f6vDO0jeea2YYZWE+nmdkxZvazDK5vE+CCLtSwvMbMNs1FQdkKqiW/gWO7StDOzL5qZmPyXY+OMrMvm9mlwE/M7Mh87atmdraZVeaq/FyXtza5rkeGbxCkzvu9M7XOdpbfzxNZWn+X2EfaorPHRTMry1RduqCfmdluuSrMzD4DfMnMhnXFNirkd9/ORUA43+2EbJef62uttkg7H2Rk23P9HXbl433qs8jG+Tt53CWuxfKpS1xYyCpKAdy9KTlvZPvAOsTMTjOzr5vZtslrXeLA4O6ehe3/PDDEzAaa2YAMr7stPgc8a2bfMrNenW0wmdlngSPd/Z3keb4btl8B7k3qkomT9q+A2cm+MCCfmUNm9gVggLu/noMG3WDgPDMbnYXfwC+ADd29OR93otOZ2fbAWcAHOSpvNzPrm8H1bQ2cCUwDaoDDgbGZWn876rEH8EV3n5el42Zey1tNHUrNbHszO9HM1s/1xWeGy7vAzK4DfmBme8FKjeCsfbZmtr+Z/Rz4j5ndYWabpy3L2DEu2UcKzWxTMyvpAuep5cysyMzWN7MJZjbA3ZuT19u9/cn/udDM+rd4PZvf4dZmdq6ZjclyOXsCO7v7/7JVRiteBZYCj5rZvjkst83yEfQys15mNjwVEM5meyRpJ2Q9iJWSHNf3MbNvm9nGqd9jFsoxiGut5HmXuSZOvtPiTGy7mZV25pjWEWk3z7tE4C79u019Fhn+3abO1V8Bjs/getsl9Xkn10pDzWykmZXkuh5d5ockK/mHmd1jZlsm543mpFGWrYPC34GtgZ2BS8xs+3zeITKzHczsS2b26+TkmTooZuog9Yy7vwx8Dfiqme2aywOgu/8YOAz4EjDFzE5NLevgd/x94GfJ//8icHIm6tkRZvYdoB/wNqw4aXdifUcDFe7+1+Sl3wBVnapkx+tiwI+A52Glk2e2jqMDgAOBS4EjMxUINbOtgB3c/UfJS182s9GZWHcH/Ry4Jb0Rla1jnZntAvwJ+J6ZbZGhk+6PgR+4+5XAX4kLoYMysN72+j7wHQAzOxw4pYeV15pfJn97A5eZ2Za5KjgJaH0jCc70TV7r0H6bBJMOBx4HmoFjzKwkrRGcrYurnYGfAHXAT4FZwMNm9tuk3Ex2bf46cC3wX+AB4ItmtkWyLJ+ZH5sBvwMeBL4JTDSzC6DD238xMMTd65L1FyXrysrFnZn9jjgvbQvcQ2QWbZXJMtL8BPhXUm7Wrx/MrI+717r7mcCvgW+a2SnWRbqmm9nRZvZPM7sil4EvM9ub+C39z8xutyTrMFmWkf3LzPqa2RFm9lMz2zE9qzEHv9dfAd8CNgBuNrP9M12AmVUA3zGzR8zs87AiiJYsz+cx6TdmdiVwqZlNSF5rd7aNmW1mZr8AbjWz78JK7dZstbG2NbMvmtmZSXmpwF0+P8+C5LsdYWZ/M7OLLRIqMnYTMEkGqSRurv/e4obFWWb2V8tRVr6ZWdo11x+By4C/Edf3OaXgUheTnKC2BJ4CrjOzy82swt2bkkh2v0wGQszsQCIT4yvufiTRwEoPdpTk4aBwAzAYKAceN7NLoPOBihR3fzt5+DjxG/g8cLbloKtT2nd3JPB74gDwbTN7ysx2bm9jNrmoqwCmmtkw4Gxi38lHGuwg4IvAC0Qa+/FmVpW2vCP70VHAuxZdb74J9HX36bnetsRFwDxiUy63SNdPv2jIaJ3c/W13/wxwO5EZ80sz2y4Dq74AeA/AzA4AjnX36RlYb7uZ2XHAxsDA5EJ9NCy/a5fx4467P0kE7MqJE++XLOki25HvL2mUHkRkmI1z90+BYmBx5mrd5nrsRRLUBb4HvJbF8o7IZXmrqcPmwHbAocQ+/QSwm5ltYGaHmFlVti6yzexs4AxgJ+B04GDoVDDmUuB77v43d/8+0Bu40sz+ZGb/NrPLzGxIJurewneBK939V+7+uLt/EdgF2MnMfrSW/9tmZrYPcTf3T8BGwNXABCJba1Seuzv9DngT2IO44fQ9YC8zuz65AG0zMxsFnAaclzw/H/iDmb1jZqdDxgN2/YFdgS+4+6nAAUAh8DszOyVT5SRlnQ5sRnw2pwHZ2B/Ty9sKuMGi620FcBtwIzCOCKTllZkNJAKJdwIOfMuiC38u/CApd3vgE+AsW5GBnKl2yKXEjcoy4mL54NSCbP5ezWw8cQw6Kgkq/hTY2iJD9QSLYSX2zEBRlwEjgbuBX5jZJcnFuUP+hglJzi1bEhl7s4nfNEAqy6o99fpV8v/+CmxvZkelFmRj+5LfxE3AKOD7ZnZ/6hiaz2N82s2ZHxO/1SXAhsR13zHWIsu0Ew4AXiLO35cQ55RPiDZKzpjZD4lM+l8S38WLFpmOObsxb/k9p0tLSXBpQ3f/o5ltDHwd2AH4q7tfaWa/Av7u7m9lqLxzgAJ3/0MS+BhM3FX8vLtPMbMLgWvd/cNMlNeG+uwJHObuX0ueb0mcwHcFduvMdqdFr4cDp7n7j5OAyP7AFkAf4F53v6dzW9Fq2ctPWhZ3Im50943Tll8PnAAc5+7/bMd6NyQCSmVEV5w6d/9CRivf9rrcDLxBXECcBIwmDqxPA4+7e0M711dIXLRtRwTQjgBOdfcHMlnvNtZlFHHH/bNEZtZniLtqs4D/c/dXslx+L+JY8Hngenf/bQfXUwScA6wPzAVOBL7j7rclyy1XjYAkePQMkXW3kPgd9iWCo7e4e32Wy9+FuAjsDVzs7s91YB2PERej+xHfz2Sg0N33SJaXuPvSzNV6tfUYRQQ7NiIuvt529+OSZQWZznxJjjvHAOOJRlpWy1tNHe4BPnX3Y5PnewJ3AM8BA4FlwE/d/Y4Ml1tK7Ld7uXudRdbRN4iMjiVEZuWL3sauQ0lAdSLwOvArd3/SzCYBLwMzgHeBrd39WxnejjOAc919q+R5GdDk7suSC8lTgaMzcVPHzB4iglh3pL3WB/gtsAmwv7sv6mw5HajXcUSX8iNavL4ecfPnD+5+bzvWdzjRoP8jcW44jzg/j0jW9wN3/0eG6m7Euegq4FHgH0lwOxXMuywp764MlFVK7I+fJS7Iv01ku90APJkqN5PM7ESi+3Yj8ft4ljhv7UYMHfFbd78+l+esFvW7AZjr7ucnz88jbpQMAKYTx+JveTJcQQbLPR04w913Sp4PJG4Kv0e0Y18APnL38zpRxrZEO25nd280s0OIYPATxDH/eaDS3e/u1Ma0XvbfgUoiuNSUtJcnAXcBY4CPgOva005upYxtgX8CGyXXBP2IIPM0Img6F1jk7r/q+JZ0qF69iO9vC3dfmlyrXEMEv0uIc8sk4Gp3n7eWdR1DdFvfO3m+K9FGeZkISk4CFrj7FRms/9XEvndh8vwXxPXkpsR5eTPic1+Uy3ZmcrNyW+DsJGCJxY3aHYmsnvvc/d+dWX/yuA9x/D+KuBn9F+Kmyv7unpOuckmQ+ddEj6SzgdeSeMLpRDvi3FzUQ5lLXYy7P0QcTHD3t9z9LCI9dDczm0PspBkJLCWuJu4IATS7+yySk7iZHQp8JoeBpT7AD4GNzGwTMyty95fd/XDg+M5ud9oFz6+J6DXu/nHS2LsamMKKO/GZ1ift8XvEmEvpd98uJtKcH2zPSt39HXf/OnFR9R6wwMw+b2bjksZnTljMkvUi8BN3/5i4aLiF+JwPILofTmjPOj2y9e4gGupPE+M47ZDcacjqXdNWjAWucvdpSSDpGuJ348A5ZvY9y9A4Pma2u5n90sw2Sr3m7vXu/hPiZLGLdXy8kp18RfctJ/aZajPbMSknl430A4Dn3P124BHiM32OCCb+OP1OaWdZZLOcmHZ3F3d/0t2PJi7MfmXt7CJnZt8jghtPuvvFxIXFLGCCmR2flJGLwNL+wAx3vwi4HHgFGGTRfSQjYza0lFwwXU5cML+Q7fJW48/EueIBi65NpxMNuoPcfWsimyOjgaXE4cRFzv4A7v4I0Ug9CjiOyKo4ra0r88ga3Ar4H5Gt9BTQ4O5nu/uP3f2mTAeWEgXAFmb2ZzOrdPfF7r4sWXY/0AsY1tlCLMapqwbuTz8nuftCIgNjFpFJmA/fB+anv5BcLLxPtIO2aOf6HiQuJoYS++dfkjbMPUR2Xcayoz3MJ9oumwM7WoyzUezuE4nAzM4ZKm4D4Mfu/q67P+ruBxHn5O8AP7TIKsnojLXufgORZXoD8BZx4X8/kdExlriBla/xjvoQ++w4MzspeXl94obaW0T2wttE5kmmVQEzzewMMxtJHI+M+L7PAa4jLnA70xXpq8BM4qYlwBziJsopxJih3yQy5LLhH8m6b7bo/vcd4gbYEe6+DXFjuMOBpcSewFTgCIvM5Wri5u7ipOz55GgMyBZ2IG7W/Ca5STODaFcUEjeungfq1xZYShwENKZdZwwiMm0XEueacuBhyEyXNYuuX7sBLydBMYhg0hVEW+/nRJJAYS5/s2llHQicbiu6Bz5HBH9uJm4cd+hzSAssnUgE+78ObOfufyY+828Sv82s9iZJq/tQ4D4iWLqNu/8xef1UOrGd7a5PHo7LsgZmNsTdZyeBlca013sRB/sD3P3pDJY30N3ntoi+fhb4AnECP9/dn81UeWuoRyqr6LPECWwJEWx51t0XpL2vU3epLO64X+/uOybPy9x9sZltmOk7TC3KvZg4afza3V+06OJ1OHHgnUUEEB9298vasU4DilMXsMndxZOIA/gc4m7ffbk4kJvZFcSdkGeB99L2pd5Els9ngJvd/fE2rCt1p6EEWJa2rs2IC7qRxP5xnbu/kZUNWrk+uxEn6ueJgFJzWp3GEJ/3xkT2S+Pq1tOO8vYEDiECki8C/3H3uWnLnwFObO/+mgT3vkM0fm9y9zeTsg4g7op9QGQM1XR2G9pQl6HEZ7rQ3f+T9no5kRK+K3GH9Jvu/kkGyjuKGOdsJnCPuz/WYvmdxF3mNgWXk/q/THRZvJfIVPkwWbYP8AfiM93Ms5iBlfxG/kWkPv/G3W9OGjCHERcBfYlM14zfYU6rwyDiIjbVRfe6LAV1Vlf+D4hzRl/i856T5fKGEo3oQ4gAbQkRZPyixcCpS9Leu9YsLjPr6ysyTsYQjdEDiAD9H7N5/La4Y/9n4rf4a3f/afL6+cA+7n5IBsrYmbhpNAP4P6I7+sK0Y+jTxAVj1o/lrdTtMGBf4pxys7v/K23ZN4Bx7n5OO9bX290XJe21CcCkVMDOzG4D7nf3P3e2HZOsb313fzcJ6pxNfIcvEll1C4hgwP3ufmlnyrOYWOLzRLDxFnf/IG3Z0KScLYFDPAMZTMnxazfgVXf/xMx2Is4HD3iMlUlyM2SWu0/rbHmdqOOQpJ6HEdlKG7v7KmO4tOUY0IHyDya62jiwF3Ceu/83g+vfjsjG7EXc9DkduNPdr06WD8jEeXktdTiXOKeMIK5HXgUaM3E8tMj0PYBoY9QRGfI3eSsZ4Zn4rbazboOJgPFuRFttsruf37Iea9uvkuuBbxJZ1U8n/3491fbJ9H6ZBFy/TLSFpxLH+28Au7r7/KQ+llxv5SS7uZU6Hg6cS/xuftSW65G1rK/QI7vuBCJLbLfk2DCGyJzeFNje3X+UzW1Ou37enmh7HkxkzQ4mxnDsRwT1js5G+a3WScGlrsHiTvfOxE7xPHHRMsndpyTLTyOyDs7MYnlPESn4o4gU/Ts96eqQbRapvb2BKnd/NmncfpbIrPh5phoRyV2KfxN34R5Ie+1fwOHpF/GZYtG9a0siMLIjEYS5lGgwnUsEl9zbka5oZhulLoRtxYChjcnzocQd4fs8C138WqnLCcRdn0PSGtIrHUjNrMojK64t6xvVogG7UteiJAC5FXExlIvMkAeAa9z95uR5y5N8ITHo+CedPYFYdNWaRPwWDiHuNBnwmLv/JwlKbuHuJ61hNatbdwVxZ2wToi/4A0R2WT0RiNiNCJB91NH6t6MuNxDHmsvcfUErn+lQoL+7v5WBgHIqgL4FcUxZn7gLfqu7v5HsT193933asc6/E3c5JxNp1bOTx4970r3HzPbyyGrJqqTRtjNxTHmL6C7yXHJMPRWo9cx1xakE5rf8PpIg1xDi7mCdu9+SifJWU4fNiUyKXu5+Y/JaBTG+xG7ArcAPPUNj9K2hHqOI8+fZRPbWD70dY5dZDCb6I2I/qgOuSAtQ7kqMhzQE2N3b2aV4LeWOJO6kXpBW3rZEo3QQkUV7NnB6W4OtayhrdOozsZhF5wgis+7GZB+9gLi7mrNGb1KXfsSg21PNbASRyXAo8V381uMG0PPASR4zg67tQu5Q4ni9jGg7XUWc01OD2X4e+Ia7ZySTKLnw3srdz0h7bTjRzXkIkYkx092/kYGy7iGC8psTWQ/nu/trLd4zwjN0U8LMjiTOR0OIi+K/AEcT+87f3P3vmSino9IuKEvdfYmZjSP2nSOITJB/t/x8MlRuMTDCI6sudcM5Nebf28Qx6Dl3r+1kOek3mk8ksjHXI7IxXwIWZ+MiOTkGbUdcl/4pea2CCJDskJR9uUc2T2fKWf5bNrNtgH2I9tDjxM3Rp909p2MmptVtaKr9lQRVf00E105qTyCkxQ2LcURG255EL4CJRKbzstWvoeMsbgKfSLSL5hIBpsXZDka2Uo9UwKWQyCYuSeozjzhvfwv4r3ei+2haWc8Q481OIc7pGxGB/otJbkbnIkhpMQj8C+5+jcX4SnsRgbRPiIms5ucssOfu+svzH7AN0eA6khgv41tEsOP3xAUWRB/z8hyWdyDRpzpXn8GdRJeGt4BTkteqidS+EZ1cd2GL52cm6z2VOOhcSwSwIMafyub3vDvRiL8LOLiD6zmEOEBeB4xKe71X8u9mRDfCXHxvluxLm7X8/IhU3m3a+5km388SIjsn9VoRUJI8/hZxNyAX23cScQJqbdkGwLAMllVOBD7/S2QoQtwBOYdI3X2dGNB0VMvPem3fUYt/9wfeJ2beuzo5DhTn6vdOzOw1eTV1HAQUZbCswuTYcjkwOHltP2KQ0NuTz3oikaWxyrFiNescQ5yo07fn4uSYcj4xTkXWP8dW6vVockx4gTiWD2ntM+7E+ickx639iYH1l3/Gyb9bAF/NVHmrqcNgYvyJK4gLgb8AfVrUcSIxbmGmy96SyPw7D9g7ea03Eei+mOja+fV2rO8/xMXDScm/xyWvp47j5cAmWdiOG4Ffpj0vT3t8OBFg+UsGytk42Sc3SnttCNH4nkiMtfY6aeewXP0RM+hcktp3kuPfWGJA738DHxKBb1jLcZboOvRq8j2eQgxou2va8lIiWLJF8nytx5g2lDcZ2Dx5vjmRfXU00C/tfUVtqf9ayjoYeCLt+ZeJQXHPJDI/vpXJ748YK20v4sJ09+R48wgxxtPtxHg753ZmmzJQx6uS7/hHqeMMkWW8LdH15wlg/SyV+xhxI+MLaa+PSY5L9xHjknamjD2S7/SitNeKk8//QSLLcXQWtm0gESD7KXHj50FgYNryTYgxZMZ1spyjiOPfVcD45LVCok1wMXH+PCwP+9TexLns38k+VJG27GxgEZE12JZ1fSvZjqdTx4jk9b2S9T9BBKYzWf/DiZtbG6S9tmuyv/yZCLz0yvFnWpD8+xti2JBJyeNziS7h6wHbJu/pUFuFuP4pJ9p+P0p+g98mAnm3ZuM4sIa6jCSCW+8TYyvldB9epT75roD+nGTHP6XFa+slr98H9M5hefcTjZdONYDaWZ+TiX7W1cTdidFJHbZIe09nL4xKiL77OxMpzGcnB8P/JQeFgkyU00q5Y4gA3mjigqiMGMfiyOQkdzewYzvX+dPkYP1VIrBzCWkX5ERGSK6CS6OSg+jg5IC90ucIfIkOND6JxsbdRFbddmmvfwGYmsN980ziTjZAWYtt25wYRLk4Q2UVJr+Bk5Pf/XXApmnL1ye5EKQdjeuWn3/yfZ2e/A5OIwJMv2nPOju5nV9nRTC3rMWy1OxfGQkwEQ3jCclv/VHi4siIC75xyf60XgfWO6DF877Jb/oXRCNx00zUfy11GM+Ki+PTgduSx2OSY81MMhycIBquDxJdtrZO//6IxtuJyeOMB5aS9V4FXJg8HpLUY6vke9yUtEZzpusAPElc4PyFuEEwsMXyA1LHqrWVTQQb0gOUOxKN/j8nv8/biEF7M/357dSi3G2Jbp23At9Ne73T53+iC9x30p5XJ8fMCiKQ8x+i22tWfyet1Gu39M8gee0zyfGgf/IZXcGKIN/avssfAX9Ke/41ouv98u3OcP2/BzyaPB5LBOpuIgZg/mcmywOuJ9qEO6V9dvOAe5Iy7yIJimaovLtJC5wkr22a/L5PIwIPP8z1PpNWl4uJdtupxDnlpy2W9yEy8TJd7qFp3/lxRLtoK5IgBHHzrVPfOzCcuDj9NtHl+3HSblAQGTS/IQvBYKIN8uO0578gAi7nETPHfY4VwdKOBgJGEAGsI5LvcDJp50fiGuFQWtyUydF+9QSRqXoYkV20eyvvSQWTV9tOS36fqQG7f030RhnW4j2fAUozWPefEjfwbgVqgLPSlhURAb1v5/ozTcofDbyZPC4nkiYeBE7o5HqtxfODiDH1jk+e70H0Nsj29rWsx4Dkt/Ma0Tbql+06rLZu+SpYf8t3hs8Sd71SWRmWtmwkcbcmYztIG8q7g7SoeY4+gy8SF5XfY8VF53ZEl51MNHLTI9gvEHfeUtkfxWnLM35xTTTeryPuhP4o7fViYgyms2jHnSAigHNish0nEFlKNyUnyn2JYMgDOf7+biZmHUo9T+1bA4hgYYcbI0QDampSRq/kZHloDrdtc6L7WEnaa6kg04Vk4A5/sq6W2UXrJ7+LB5PfRf9OrPtMYClxF3o/ortralk50SDJ2Z0O4kL61rTnBSSNHaLhfG2Wyv0McfF+LzExQqbWm34MHQMck4PPcDcisHwecVf3BVZkD6b2oS2yVPbQ5Fj6MNGgKkuOY5OyvM0bE+M4bJd2zH6YyGS6D3iHyELN6M2YpJyvE2NXpY5rlxEXl2cRF2Nn0o5sqeRzayYJ6BBBgunJ6/uTzJCahe34HHGTJTUT5y2smN3zIaJrOHTyXEic955Ke34UMRbQfUk5m2VzX1lL3f4DfCmtnr9K9qtXSM7RJDcM2vI5JJ/dkWnPhxE3rVIDBf+HDLapiAHDbyayVf4HfC15fQjR1tgng2VtSnRr+Qsx+PkzJOdfMniBmqxvf2LcydTzn9IigJvPPyI4MTPteR+ibXANkQn33+T3m9HPJSnredKyIokg/93EkAr/JoJCnbqhkRwXfpD2/CdEsONHRLb8Gay4uZWxwD1xznyTuDkzPHntbuIC+R4ic+3Gzv6Gks8pPWPz/GQbb0/27ZtT5ed4v/oG0U0Yoi10FBEwvowVNxyObeO6HiW6M6d/h79MPsPfEdcKIzP1HRIByTdYcZNrayIw/3kiGPx5YIe09+c045Bovz/Iyj089ieCmWWdWG+qjfUZ4gb6ASRBPCL7/lFiYhHIQaIGkWhwBTAoeb5JckyaQRayn9vyp9ni8q8vMebJj5M+tqWpBR7jIfQnsopyVV4lcbDPpZeIA//X3P27yWs/IO7UNFkHR9g3Wz4ifiGAu3+TOFmOAp40swuJRmRzsjwb/VBfJAIiBmxpMavYQI/+zhsSA5a3eZwOd2/2mEXlIuJuXoXHuFi/JjKYriLugOTSr4HjzeynyRgbqXGQLgYecfcPOvoduvtLRPezO4k7prj7nZ2ucRtYzLrzKnHR8bLF1K54DEg4iGjk/zh5b2ePpTeb2W+B35nZD4k7H4OI7f4K8LDFYL/t5u5/IS5yzibuQt+XtqzB3f/n7i92sv7t8SJQaWZ3mtmYZJ9ODYJ8HnG3vNOfqZmda2afM7OTLAZln080JO8DLjOz681scNpxokPcoz998niapw0KnC0e09yfQdyhvJ4Y3PY1i4kgPHnPK5ks00KBu3+UHEu/TTTebiEasd9K3petWYRmEN23fwKcb2YHEBmOu7v7AUTD9pseAypnrG2TjGl3AjEz3XoeY0e8TlxwfY44p55G3DBoE3f/BdGN8CAzqyECLie5+y/c/X53v9jdn8jUNqS5hzj//4q4mPofMc7a9cSF8k5J/Tp7LlwELDKzvSzGizyLOMefSTS8D4SVztG5NBEoTcakSGUTf4a46NjdYnKPZbD2zyE5l19PHFdS49XMJI5xmxEXsO94jHWRiRmZBnmMn/Yd4sbATCLTAXefTWRfbJCBcnqZ2SXAXHe/lNhXBiRlDjCz6rRjdqb8jhWznH2JyJZKn8hiWPJvPvYZiM/VzOxvSTtnIXHz600i6PZvYF6mP5dkbLcXiRk5z7IVM8Q9SHS/uRo4wt1f70QZ6xPHz6nJWHAQQfwfEje6zk6Wp84t3tGyWjGHuBD+LHCKmZ1CZJzs4u6fJS7cz0h+Qx29FhhEXOv0sZjcAyKwvBVxs30+Mb5rzscFInpTbGMxRmMz0dV6cFKf3xHdIJ9qw7rGE7//PhZjd0LccK4mMm7riMlMPoSMfYfHEp/Z1haTsbxLdDs7jMjuO4DkWJ+UmfWxfszsixZjQ5K0319kRXsBYr8u9g6Oq5W0gdxiJsMLie/qH6yYWbGUSJK4N6lDVsd+TDxIJIfcZWanA2+5++nETY83c1D+KjSgdx5ZDHA5jTg5nUqkf/6XyCx408zOIHaOA1a/lq5bXhvqkz6w3i5E9HUUkdY5JoPbfQrQRMycMjt57QLiQnZXd383E+WspQ7rE42zk4mDwIvEneM9k0ZKe9dXQDL2B3EX9sPkgDrB3R/KWMVXX/42RANnKLEt7xAN9D2JfWwhERT9rLs3ZGIwO4tBg/t4DgYGNLPvEFP+DiXGBlmf6O4wgNjegcTAjz/p7AB5FjOz3AM0ENkQGxG/0e2IC8FqYIm7H9PxLVpe1mbEhdALxOwWdZ1dZzvK3ovYJwqIIMHXiN/AJKJRsj1x0s/Edm5F3O39iAh+bAM8RwTt3iGyp+qJu73d7iSY/ntKGk3nE8GXfxCDunZ61qbWykqetxys/yjirt3lORq0chvibu/mxLaebmblnsFBr1spcwgRIDmKCMjsDHzG3T9u8b42bb+lTXBgZvsTY4LNAb6c6aBgUkYJkRlwM3FcGQk0pBrAyXueJjIXMnL+MLMziYDNLsRv/R53/9TMvkxcOOZkspAWdSohMuCuB2qJY8AZqeOgmU0munW80IZ1bUVkdVxLdJmcnfab/CyREfYeMT5Xc2d/G2nlXUeMA1ifBLfmJst3Jfaj7TpbnpmtRwSwqolMz78Rx+3PEd2MRxLdxTMyk7CZbUTsm08Q2fM/IrqYTEuWH0lkXPw2E+V1lMVEE18lAiFlRFfkVW7mZfo4aDHpzAFEJtkGRHbc9i3L6Gi5FoOFn0acF2uIgMYZRHtyWRLQK3b3pZ1t76yhDuOJrPxtyfBA0En9tyKOfZsTN+23c/eNWnlvzmcyS669vkmMOTsO2Ms7MOupxSyKxxBZ/hAZorukLU/NxJyRbUzOxQcQWZNTicz419z9+8ny9Gu8nHyuZnauu//ezL5NZH2VEOfusUQ78D3gbHf/uDN1MrN/EcHwYURX3uMtZmTu5+4PJ+/JRXuot6+YROYgIuPTiAkkJuaqHqvUqxu2q3sEi7tmjxD91V9MXtuD6AIzk7i7981k+avdrbw21uknRDDJiYvA64nU47nEuDp1lszM0YF1pw6ipcQd9SoihfgZd3/QzL5IpEVm5YLIVswosj9xMnOiC8cc4gR+APA/d7+ujetbL1nPSpk7yQH0YKKB3K6p6TvDzCYSQZbXiBPKw0TQYiQx+8YkYLq7z+jod5gvyQH6R8Rdu21Y0d1wGtG4biKmnn8veX8mAmfjia43fYj09Ekes88VuXujJTPmZeKzTBpaxxC/t1Pc/abOrK+NZW5EXDTcRHLniAiI9CMGLJ9BBIMecfd5mWiImNnJRGN1EhFget9jZrrUMWaIu09PfcadKSsfWjTciolg8+HEWBk/yPZvrrXvKJuNmFaCWvsT+9BHRNbA/TnY5k2Ii+5Dga94O2fhs5VnaX2ZGKfiFndfaJEpciHR8L06w/W+jDi2fNdbzIhqMXPat4hufZ0O7CbrHEgEy/sT04enZkAqJ/bPr7r7pEyU1Y46fZmY6TA1w+BooptTaobTbxNdSU9oxzrPIwbK/YDo/vZa0m4ZTHSF/7K7/y1T58C08mYSwZhXPDI6CpPyb3f36zN0nhhMHKuPJI7Tf3H3+81sU2Kw3n94B26MtVJOKdFuKCLaRScSF8fbunt98p6XiJtoa83gyAWLmUe/SwQqL/PIJs9FuZsSNxQ3I26SvA/c5Rma+Sv5TZxEZDDWERnyCz2LM8i2PGeY2c7EPuBEr4YbU/tBBspKDby+N9G2+y/wpLu/nIn1d6A+y3+nFhmy3yPGT/wr8Hvv4M0/MzuCCICWEsHoGe7+RmZqvUpZJcQ1wK7EzeV/EDNVT81GeW2sUxFxw3YoMQbRXUTAZRCxP3cosJS+r1rM2NlEZPUdlFzr3AhMc/cfZHBz1lSfDYkeLI8QEy80J8fTp4guxt/KRT1arZuCS/lhZn8mfvA/Tu5K7EWkGM4nUkKLiINeRnaOXJfXhvrsAFxJ9HWGmBmkP9HozFh028x2dfcnzGxrYiC/IcQBZwtiIO3ZWbjLlApsVRGze3ybaAge5u4PdnCdhxIHkb5EZstdRIr6s8SFzkx3/3Im6t+GupwF7OvuRycnlmOJu9Nf9cynyuecmT0JXOLuDyTP/wQ86O63Zqm89BPWF4gZh2YTqcwvJBcP2QiAlhKzfn281jd3vqz/ABPd/c9m1pfI0LrO3Z9MlmclKGGRcfI94o7vw0Tg7iNP7sbm+m5ONrQIMo0lZtR5IEPrPpG4eJm3hvcUJ5/nKGJ/6nD3jDbWKX17Uw3y3u7+nWyW26IO+xMZDGOIWT/fb8P/2YbIcvkRcdF0OHHxXkd05/vUIvu0yTObebYhMdDzDsn31IvIoHUiK6WY6MZwkyfZVJ0oKz149nSy/ifdfWpyPvwWMX7caZ0ppwP1GpfUZR/iXNlkZhXuPj9ZflRSt2Pc/f32XHiYWX/i89yX6Hb8mLu/YmaDvQPZB+0obz+ii2+qvH7uviDDZRUR46rsTbQZb3X3282sNFPn+iTwOd/dL0kCnRsTY7VsQgTMNiGyAc7NRHmZZJGhdjHwkrufmcVyWgZg9iT2t3Lge539LpIbTsu7SllkyZ1KdNF6BfibJ9kR2dLKzZLTiO/911koaxBxnNqeyGj5pkd31rxoEWQaS3Rb3pPIemxzJmuLz7AP0ZbchQhCXpLp9nmLtusAIlN1K2K/+a+735PJ8tZSl3KPXhK9iRs/v7C4Ufxjoj19qXciKzftuq6cuBH8HDF8yyvEDH/jiXGytvW4EZyrLO6TgWXEuGuPEzdqryLGwZ3bkSBaRurWA9rW3U5y4PwZ0cj5rZldQQRWPiTSgvdIli/M0N2unJbXxjr9nWgUXZc0docndfhLZy+MUj8mM9udSIt8EPiru79qZtsSAZpl7v5kpu4orqYeFxF31e8m7vLtaWbDiWDMlb5ibKK2ri+VDbU90U2rD9FNo4hosG+d2S1otQ5G3BGoJmaGes3MDgNOdvcjuvsFu0Va63+I4Mfd7j7NzB4mBoK8P1v7SysNq/OIYOhj5CALJZuShuq9xN3Ipz3GxLkNuMPdr23ZsM1QmUac31Kf6XbEZ1pJzGJ1g3ezbCVLstdWs2yl7U1eK/ZO3NG26IpykrsfmjxfY4ZX8js52ZMxHbKtZZAp199n0sg8wtuYvWRmdxDdaK5Ne209YlyNcqJL+oIs3Oz4LjDY3b9uZpsTXQR2BBYQd3IPILp0dTbTpbXg2QQiW/cioqv0dkS2TcaCZ22s263End3LkiDXXkSGQAWROfkesLG7P7W2xngSHD+HOP8+Tdysei5Z5/FEV7uJ7v7TTHyXbShvETGO1M8zUNYuRHDz6Rav9yFubu1KdKfu0Hglqylzd6Ir4glpr1UTY/wcR2QLr+9ZzJ7pjOT7qXD3mmwfh1pczFcAQz2DGest930zO46YGfX3mSqjPXXIwee5CZHB/FjLsnOtlTbLHql6dWBd6Z/heGKg8g7d2O5A2RsTv9n/c/cpubomSPbVLYluajPc/YK0ZV8jEhkOcPfJHVx/6rryPGC8u3/RzE4lbqQsI27W3Ozud2b5ujJ1HbgbMZh6ITGu1ZZED53NiLb1hXm9HvM8jCKuP4eImN9LjN/wBrBz2rIngT26c3lrqctIojvVLNKm3CS6CJzfyXWnZhHqT3TT+mPydxcRZc76NKNpdTiK6Gr4HMnsVMTd0VszXN4Aktkacvgdfo9okP+V6M6UmqmqOJf1yNK2bUfMTnJ7sn335bDsgrTHY4lxXfL+mWRgu84jgry/IroYPpOlclpOzVrQ4vlJwLfy/Xl0cNt+TjIbyBrek5queR9iTInOlPcUK6YgP4bWp0cuTP79AjEGS6a3eRBrmG2FSHVPbXMlMDbD5WdkphdyPCtsi7K3JrqF7kR0xbsc2CpZ9gcyNLsh0e31lBavrZe8nsqQyulsQUkdionxj8Ynz3+ffN7nJOfne9u5vjOImf4+IcaO+waRFfkTIvgzncjKzFT9c1Ze8pnMIbLKV2krEUGsjM0+lHw3A5P1rjIzHBGcHJ/rfaad25CxmdM6WiZQmeH1F7T2OJfbl3ZcH9jZc9maPr98fH9t/eyT5x1qU7e2XR1dV0fLy8Nn909gSXJsTP+OSzOxHxPXlc8T2YKp14Yk5/CiHGxfKiGoNzFDeKp9VkAkTWyZei3f30led4R1+S85eG5GpBFulPb6fsRYPN26vDbUp4wIULxOpHZ/i8juWV7fTq7/t0QaJMRo/gclP8ZbiW4MWf/hEdlYNxKz8GyRfNavZapxlqeTfmHa48FJI3QmETAoyXV9srB96Seko4gLo1uJfvqVGSxntZ9V8lvNSAOjK/0Rd9wvTU7O1yQnyIxP27yaslf5rXSFxlA76v99kgtI4oKsYnXbk+w/r9OJgDMxjsAbSYOlLzHA/Lg1lPd0Z8pbTR02JcZvOIA2TElOjLu0YYbK3ip9O9e2/wAHrmV9xxIzSv2SCPKUtVj+GLBlFvabVND/IiKg8idigNPUDZCJtHGa67WUs7bg2R1kKXjWxvqdBSwm2hqvsfL02BNp5zTuxOQVPwVuI5lyOm1fGcKKoGumgpNZL49og21HBHSvJ8b0+TpxYVZMdL/I2I0W4obnUcnx5Ynkd14A7JD8Xv5IjDuTl32mHduR+i39gBxOZc+K4MuRmfgNr6WMatJuSGdgvWvcT1scP24mSwHGtN/N+sS4pXnfn5L6pPapguQ32OG2ftp3uAvJTe4s1z2VgUXym876cT+tvPHA0ck+81pyHBlMdAsclYFyNifOpVOJa8yM3sxqRz0uBy5KHh8EvE10yUu/Pstr+zZj0/VK+3h4zWPa4bctjCGmdf9Zdy9vdcxsd4upifu7+8+IRulcYvrK5bNCePLraOe6e6c9XT4oubvP8ZgV50FiJox9spEuaMkU3GZ2vJld4O4zgN8Qdxa/SqSwX+oZmhrS85C+65GOaUlq5hx3/woxXscxwJQkzbjbco+ZNJLH/yHGfniaCKJ9xzI3zfolFv3+W62DR/ptEYCZ7UP0Y++Wkv0lNYX9N4hxAEqJu0zHZeozNbMTLZmCtiVfucshFtM5j89EudmWdOM5gciygOh2tHNrb03+vRD4jXdusN0PiEkevgrcSQwOOTWpT8tujN8hujN3enDfFqYSjaZzgG+Z2XZJF5Tl0o65+xHdeTrdRSTZR640s+fMbPe032NhattTb03e/yPi4mR16/sK0XXpeKCRyEA7L3WstJgpqMEzPKismX0L+G+Sxv8nd/+cu5/j7kuT7TmZ+Mz+mYHi+hLdwX5sZjsRv28APLpJVhJZTDllZp8xs908BkjfnugSe7i7P5MsP4IYcLxN44SZ2Y/MbB+Prvu/Jm48nGFm16TtK8u7GHrnuxrmsrwiYDd3/9jdTyKCSYcQ3fr/QIyB843ObE8LWxHHtK8TgewLiTEkf0ZcINYSE0B0Scl5rSj5LY0n2rIZG7dnTefF5HzamLQRfkAcozO2/mS5+YouaX8mMtoywleML9TqNWjq3JK0fQq8A4NRtzhWt7a8IO338msiUzcn1la3ND8GFq2urW9mBav7DFPL077Dy4nfV4cl5a2t7gVJO3pT4MtAVrtAJ9cibtHFvJC4xjuRaDd8g7iRebO7f7Cmz2oN6x+b/LsJkaH9K6K77mLgt2Z2YaqtngvJNtQDcyyGXtmf2M6xxBA3QGaHmeiQfEa29LdSJLKUuGNzdk8sLylzPHFH/W7igv1bwIhk2XbJ688A5R1YdwUxAKERWVHDiC6A1wK7EWNavE3cDb8X2DVL21hAMh5C8nwL4sJjWL73sQxs2ypdjtJfIxqE3SYbpC3fZdrjjHVRI8dZKF1pf2nxPGNd1Ii7t3emPV9jijLRrWRkvj+jNm7bN4gZ7zZMjmUvtty+tP1lJNFI7sydzt2I2ZqGEhlMdxBZE3vTyl1IYtadrGVRJtt0MXEBdSYx4G/L7X+EDHd5Br6SnDP+nb6vEBfhqTvew4lu5a3eiSfGQHgT2DrttT2IgZ+vIS7gXwU2z8Lndkxy7Ph98vdVoCpZtmtyHtwhA+V8hQhCjEv2k0eJgOMmyfIzyGHX4rR6lRBd758jMpc2bXFc3YG4u71r8trasinKk+16kLjZMCp5fQwRJHkY+H4G65+T8liRJXEw8ED6a8nj7YnuadUZ3LavEMGY1D7zEtH2+xxr6fqbo31neLK/FLZ4vdUMMaLbabuy31ZT7lDg6PTvJvWXPC9K3pPaj38EHN/JMte23x8M/D1Dn+uBwAtEEDN9G1MZj1sBE9KWPZo6ZrWznDa3l5Jj199ytF/tS4x/t3zbk39b68I2mriGaS1rtoy0HhBp+2XLrmCp/eRrdH7IkZK0x0Vpj1tt8xNjl2b8vLaG+j1HtBGuJWYBTF2DbZD2+bSrnUIEq45Ojk3TiImMIM4tg5PfxuVEskSutrOcaAM9TwzcPSR5/XmykP3c4XrmuwL6a/GF5PjiPJflAV8ETkse70ykLt4NnJq8VkRMBdyRdY8luhtsQFyMjUpe+xJxB/wO4g7ZCODlLG7jycAfksfnAG8RswkclI/vN0ffa8uGVtb7Hme4/jnrosaKi83ByfOf00qXmrRGx0Wp30dP+VtNY6mz3WDzPj5QFj+v7Yi7+r8mLoZ/sbrPjrhQ63D3BaJLzF+IYMQ+yWsbEXfHrySm387o+Bet1KGEGGz560QQ5HPJ8fw8YjaU24mgVyrd/0tk8KK+RV0GJtv8Hq0EQokA0S5r+P9/Bi5MHlcTdzzvJGYtfDzZb3+dpboXEReftxM3Xi4huhodlhzX2n3R1koZeQuetaFuhcSd8zeSz/tSIgNwA+LCa0/gO8l723T8SfbNTZLf4iPABanzAdGm2aw96+sK5QGj0x5PST0nLvr7EYGlNXb7zMA+k7q5eBWRdVCdqc+wg3X8RVKXQ0lugLZY/k+SYBKR3ZyR4ETyOTQQ3WQ3S3u9NPn3V8SkCRBBxmdp/wXzwUSXx0PSXitIK2NnVh635UliUO9MbF9von3+XHJ8GN5i+XskgSfi5tOFHShjf6IXxBktXk/9bg4Fzkl7/ZlMbd9a6jWIGMPzD8RkB8u7viX/bp7sd6nnN7Ca4D/RFWpWsj/0SXs99R1+KfWbJc5hz9PJoRWSdVzZ4rX0INPJrAh0fC5Tv4m11CnVpjue5DyaHDsOJbIeB2eonN8QE2DdR8zgl3r9GGC99O8xy9u5N/DdVpb/lcgehy5yjZn3Cuhv3fgj7lAdR4xB1C/t9dPpZPYCK19c7UCc+H9KZDOMIhrSvYmG2pXAUVnczh2IO4p3EuNrlBF3bn+b7++gE9t0BGsfUyT9Tudp2TzQZmH7cjZQMjnOQsnjZ3oiaxmjihWNvZF0YkwFusD4QDn4PMuJBuVlrBgUfaNW3te7k+UUERcXJxIZOz9kxUDIexJj9qwSuMvwtqYGMJ6f1OGS5DfzCyJQP5eYiSb1/oPJ0JhkxF3PnxEXlucQwYnDiMDWEuJGxfrJezcnplte7f5NBAW+ljy/grhg+FnyXf6SyLjNyLg86d9h2uOy5Du8HNg9+a38LfksO10ueQyetbF+JUSA/noiSHg9caFwCGkZo2s7xhJd+gpJLoaJu9bbJb+H24hZiDJZ75yVlxxLlhDtpYuIrKJfE9k4bxDBhXMzWF5r+8zdafvMZOD0PO4zfYmg2teSOl2SHPtSF84nAJPT3v9H2jAmXBvKTc9ouYQ4/l0D9E1e24yYfbh38vwk0gJQ7SinhGijPUF0Ex2ftqyImMp8h+T5GXQy4yVt3UexIvtuZHJMehX4YfLaucBdae/fk3Ye11kRmNmfOGc8TpLBklpOjL+TylY8GPhSDvetLYhj8J+J43/6xEoPAxcnjzcmZhZt9ftL/h1PBBVeSt8GIug4jyTTkLhJs3cG6r4h8BDR9fOkFsu+CDya9vxOchCwS8qqJIKlv2rx+k3AVzux3vRxryYk238aERi9gbhJ+XCu9p2kLq+kvkviWvPk5PEZqc+bLnK9kGpwi2SVmb1F3GWpJk6at7v7jS3e06FpQC2ZctvMDiUaZEOJ9NpCIpXxJeLAvYzIjHqh41vSpvrsT9wt+Ffy0uPESSMnU4FmksV0t08QU25PMbM+3srYKmnTdH4fWOruv855ZTsgqe+G7n5yMtZKL3ef3+I95u6e9DVPDQbbofFlzGw7kkEOiRTxu33lKVPTpxr+KXCPu+dsLIBMsBxPYW9mBxGNxBlE2vkb7v7lZNlKY6uZ2QXAHHf/W0fKypW0fa4YaPYVY1RUEd1JxhONnofc/Z70/9PJ8lL/9iPuVE5MynmVCNovdPeGTm/gmutSQUyzvjNxU+J37v54sqw3MTvmbHdfkuFyNwHuIS607yMarc8Rv9cXgB2JY1tqvx6QPF/tscDMtie69PUlzglnpH7PZvYkMetMh6abXkOZPye6Nd5GnGt7E+fDd9z9gWQck3p3n9TJcoqJQNlMd/+tmV1BzKbzIXHO2CNZvtCzNC3zGuo2iAiyzUqef4mYKe8lIpiyH5EB+L82rGsgsf/PIi5MBxP7wxbJOs8i9pnN3L3TY+7kury0Mq8lji31xGe0AJjimZ3qvq37zKI1nTOyJRnT7a/E+WixmW1IZGQMJzJcniW6+5zq7o8lY62UdbQ90KLsu4BfuvsTyfMBRJB7X6J75OeIbou/S5a3e7pzMzsdeMTd30u+8/OS9U5096+Z2XeAbdz9mOT92wPPd/b3a2bjiMDEVu4+L+317YkbbuOJbogbuPu7nSjnO8BL7n5/8vwcIrjyLBEY+C7RneyYpD03nNgXszp+qZmdSQTWLkzOreOJLJSxSd16ASe6+7Zp/6fY3Ze1sq6riQyiV5PnBwLnE9mY5xGf5zR3vyhZPsDdP2m5nnbUvTfwBXf/a/L8ECIw+AkxO+hrZvYaEYB+NNlv13P35ztaZjvrVwZ8m9j2u4jA9Ydm9hjwc3e/ryNto7RrmsuBJzzGYMXMRhDBtF2Ay9z97o5eu7azPocDJ7j7583sLCL7rRy4IdWe7UwbMOPyHd3SX8//I+5A/DF5PJxIYfwHcbdqFJ1L6U51L+pDNMT+RtwR+AGRvfRPkjsjWdq2VLriFkQW1rmsPNPQhWRwWuI8fHd/JAYHhjghXs6q3eBSEf7BxJ3qbtEtjjx1USNHWSh5/Fxz1kWNLjY+UAY/wwLSui0Rd5vTMzTHE4N9bpCl8r5DXGSVEJmLvyS6MQ/N0faXEt3xziTugv4+/biaxXK3S45xP0/2zREtlqeyCNqU9QM5nxV2QyLw0EDMUnpK8puYTQTKxmS4vO2JsZv+R2S5pN+JfxLYIxf7S4s6HQF8TIy7cTURHPwLkfVxTPKeNnfTIwKcdxE3Fk5M1r8d0fXipOTvjOS9mZjuOqfltSh7I6Ir/1VkeHr7rrzPpJV/OXHBCNE+2Ch5/BmiLfQccFPyWsa6nyS/0/tT3ynw+bRlOySf08y019qddUi0Ud+jxZimRLfLvxJZmc0k44NmePuuJ8l+S36Pf01bVkBkg321o9uW/L/NifZcefJ84+TfXkSPhRnEIMyp7ctJO4DIBpvEimypHYis1uFEYO+nQB1waLK8eHWfffL+Z9Kej03bxjOBd4lgfof3k1bK/BNweaqctNcvIM4rDcAtmd5n1lKnVGLMGJJzGjFG1Z+I4PjzdKKrfNr6NyMC+6nuhicTbdpCcjTTcVqdNiGySd8lMo83JHpSdMnry7xXQH89+4+4kH6AmOVkTPJaEXFxdGYn123EOAR/ISLXpyevb0l0afghcZGQaiBktPtBi7o8nBxsHyEatZcl21hMckHS3f6IwN+bwJHJ81tIUqRZuRtc6kD8ZzKQfpvD7ctJF7W0dRSz8lShVURA8lJiWtPPtvw/3e2PHHZRowuMD5TFz3ETotH2yxav90r+PRw4LgvlXcKKIHFqsoVCopG1b5a3uTT93+RxEXFH+4dEFtW3c/DZFxJ3ua8lurIdzIpxFTo7NpgRDeKXyOA4Ni3KGEkEB18mGZ+LmFRiy0yfi8hx8KyNdTqEyDy7jghifI0YlPUpIutntWPstbZ9aY+PJM7v3wa2X837O3XBmuvy1vCdHgcsJTIWsrH+LrXPJOVvnuwjqWPsP0kbLJs4f+9D0p2SDLUniUD6K6xoH/8auLGV9w1M/u3QzTsi4+rk5PFRRMZD+vLPknR3ytS2Jevalwj2pm4o3U9kYKSWD0g/LnX0GEt0oT4heXwELbqVJcfAAzK9fW2o1y+APyePRxABykFpy0fRxnZzsn9ulzy+kLTu4clrw1kxFlinb/Imv4nH034T/wD2TFvem+hmnJooIqc37ois4PeJrrypoOLuRLe1J0muXTqx/kuIm8oTiPELJ5N2ndPRfbUd5aefD4qJdtphaa89yIrrsy510zTvFdBfz/1jRcbHSUSW0t+IAVpLWizvTObSOOLifApwa4tl+xLTDmd7O09kxexfbxBjdDyYnAhWO9hrV/8jggPfJe5qXUNcrLQ6k1/yPeS0/3EGti8nAyWT4yyUPH+mBxF3eL9PXBj9obXPM3l+AZ0YW4MuMD5Qlj/L4qRhU0tyUZC8PoRISd8wC+X9kLh7vaaxhLLSoEp+g38nAoM/IgKFWyfLComLzy0zXYfkOLd56nGLz/nLRPfmH6/u2NfOsrI6Sysrz+gzgbgh8FQuzoOp74UsB8/aUIfC5Jh6VnLcvpQVF1ztni0o+Xdf4rx+aXIevJsIWm1CZi/Ec1reWupSSg5mbusK+0xSj3uAfyWP9yG6wqQvX+s4gh0s91ii++qBRPbFE6RliBJZPZ26aUcEjj5mRdv7ofRjAnGjq3/a80weXz9PnJ+/SzJmWIvld5A2e1wHy9ifOG8NStu+z6UtH8/KA8jnKsNmKDHj6OeT538HLkhb3peVx4Rabb2I66hZRDCqlAiebJy2fGcy3IYkAktXJo/3BJ5usXy9tMd5CW4QwaQHk7oelbxWTExicVMn171Nss++yIrA6++AS3KwXalzwXhiEPifETdOC5O/I1v+lrrSn8Zckqwxs1JPxsUws1Kiy9hBxEXEb7yDY6yspqzdiSwlS9Z9a/L6SmOJZKq8tHILiB/5i8m/fd39e2b2DaIhcmGmy8w1M9uKyJQYTzRyXyXGtVnc4n3lnuXxWDLNzMqJ8V32BRqJIOXj7v52i/f1dvdFHSxjEyK19gp3/07a673cvT7pS13u7jd1dDu6AjPbjUhH7kvsLwcSv/WJxACoC1q8fxtifIR29VXvSuMD5UIyNsZ1wDAiLf5rREPuvGz09Tez/sSdv6HAl9392Uyufw1l/p3Yxm8TKf6LiEDMMmJK6hfc/eYslP2HpLxbiN//IOJiaDqR9l9NZFjcm8EyM3o+SvaRC4jzXx1wlbt/bGapRuh5RHe5g7L5m0jO81sRF3J/zlY5qym7mNhnFyZ/5UQXrO2JGeLeIhrjU9q4vtTxpYq44L+X6Dq0CbGP9AFmuPsZGap/TsvrKvK5z7SoxyFEELmWOA6d5+6PJsv2AH7m7rtkodyhROb0QUSXsUfd/bhk2WAiiLCvu3/QiTIOJYI87xO/hRJ3Pypt+UTgInd/ssMbsmqZqf25jMhU253ImHqDGAtnavKZn+fu+3ayrCHEDa0diHbcLHc/Mm35g8Cl7n5fZ8rpQL2GEuezXkQm4H7ABHevT5b/nWinfLUN69o5+f+VxGf5irufmiwrJ4Kzn3P3NzNU9z7EZ7oxkVl3MNEeeDpZvhcxkPhRq19LdiTjZe0NPJm6DjGzbxLDTFzm7t9MXlvjeJ9tLGs3/r+9swyXpLq68LuYGRgGmwyaIWjwYB8kBAgWPFjQQJBAcAgWCBYkSNBAcHd3TQhuSXB3D67B3df3Y59mmpuRO7eru29f9vs8PNyurqlzqrrsrLP32vC27Uck/ZAYAy1g+61WeBxJuolIU14VeN325pKmIp7nH9t+vyf+a80mxaWkaRQjtLGIm/1NhI/AJES0yL62n6u4vX6ET8bGZdHqxMVYudGairG1pMWJmcV3iAfnwrY3LWZyR9s+b6Qb6sVIGtP253WflyVmU78gwkPP76ng0i7qXnhaYpRc1+4AYuZuM2Lm6rSyfBJi0DOvKzRObTUKA939iAfeJbavlzQjcU1MSlQYuc72XRW0NQZRuadm1rsj8cK8OfEC9FNidm8b26832l5vQdIcRMrtZIQY+VkzX24kzU6kF11de4ltJpJmIjxybnYYn65LiCLbEC+SJ1U5IVHX7nqEQef9RLTdpMTs5JWE4eq1trcu6/Yew8w6yiBlTGKSYwoiivTy2st1Ee9md8Xm4SPpT8uPk6QDCEPoj4CzbJ9clg8lxIvlgFvdpZDISLZXe1b8AfjE9lHFrHZmIm1sb4a93Dcs8ra6vd5Gb7m2JP2eeE5fTbynvirpasIn6IJmDeQkTUGkcy0MvET83rVzYY9GfnOF8fjMhCixKuG9dITt+yWtTqSTLV/FftS1OQYhhHxUnlXjEc/m+Yj0sMeIKlcb2r69iuNaJqx2JlKp97B9fpm8+61LIYZ2IGkJIl13ciK17DbiGBxKpJl90Z3zv7xnzUv4r/Uj0khPlfRH4p1omyqvoyJazUFEz61ARFKfYfsrSVcBp9o+t9X3ozLuWpk4h26zfXdZfhxRJOdvjR6HrvtUJjG3IgrCHNcKQUfSrMR4YW1JtxMG6o9L2pUo9nNfM9tvhBSXkqYgaS1C3DmImAX+O8MGnteVdZoVTTSE8LI51MOptlDB9qcA/o+ocrIXEYr6eXmJ/TthGvi07d9U3XYrkDST7cfL3yLKwdaLTDsC2D6gTV3sMcMRJsYEvqidh5JmIQTK093NGe7RbL+lUSitory8zkP4GiwPPAJcYPtRSYsQ94JzXKp+NdhWLRLsQMJv6TLgV7ZfKgLzzEQa4nWNttXbKNfjlLafr2JWrpvtTWv7mSber7+pilNmBrcjRKZDCK+Kpgv0ZYA1i+0/lc/PE4OwF4jqQjf1lsFvV8rM6gG25y+fa9WD/kUITR8TFYnO7Y39rwJF1anDiHSmlYgo6X0YFs34ORGB9mp3B3Jlu5MT95p7bf+ibvl5ZVmlz8BWt5eApNmIyYlBNeGxvEMeQqTOv0QIJCtV3O4cRKTSp8BDtm9QRItMT0ziLUtMkgwtg/meVLyqRfN9QIiuA4h02fmBaYhJn3WItK0HqxwwlwHw0kTK35kuUaeSpiHSjdYA3rW9UU/vrZIGEc/7ocDjjorGIoSQbYhUuZmIdMtK968bffspIcycVsS1fkQxo8WI474sMUY5eWT9Ukx8/hh4C/iP7dcVVQznJUTzoYRoOIvtdysSuickjt3Ytl8pkxMLEuflGGW1CWyv3Eg7o9mnWvW2KYnJoCcJEfErIsJ4AiIqbLkK2/zW7yJpIttvVrX9EbRZX516OiLybRZiYmR7ST8CLiL2teEqlc0ixaWkKUg6hcgRXYWYaT+NqDZ0HbBtvVjR5H40I21kEiI6ahtipntrYkbx3fL9tERlj09HtI3eiiIM+x6iIsFetm8sy+tTHLcgZobf6TRBRL0kRU0tjkJpFvUPwvL/lqSolZfmPxIeQVfXD8SG17+q2k2qRREdtbTtA8vLlAhT+L8AL9ieuVn3GEkT2H6v/D0Z4RVmIgXpDts7VN1mM5C0NlGNaVXbF5eBwWPEC+h9RMnk810iMfsiZVb3Lttbls9rEAOQaQhhbQHCCHW0IyclLUZMIo1JiNl3EufKZmUWueoUx5a2912mvO/8k/BsmZeIQv99bdAmaW4i3XTvIk5Udi+SdAvhx/MVIX7ubvutuu9/QTw3/9XTiQSNOprv10QkxrZVnldFsDuFmEBbnvAL2o14L+jniKacsrT9SU+Pq6SjiSp4zxEi1j62XynfDQC2BYbY3rGV140i+m054lkyDpH6V4uEn5QoMDCT7d91Y1v/IDJABhIRqbuV5WMSkbWrE5PZZ1Uhnimio84hPC0fIaKT7lakr05E/J6/I3wg7231GEDSsUR08zmSpifO72mIY3GI7Tt70qe6d9ghRPDA/bXrscWiZK0fOxKTW/8l0hOvJyIO1ySyAA5tZb9GlxSXksopD5ZpiYvhZOKl6K0y+3aq7Ss7TZToSplh2o3wlxGRF381EYnyuO3T29i9HlNeyn9CeD5sT4hMu7jk+0s6jYhkWrN9vWwM9ZIUtTKYblkUSjNQm1PU1AZ/oKQ6JF1BpJ0dKmlo3cBgaWJiYh9XnD5dtt+PqOj5NBFR92ZZvh+wsu0Za+v11pe3ejQsInJSIg19H9t/b2+vWoOkcYh0hbmA14nqTEcRRubXEgPr77l4hfSwjX5Easj+hAB5hO39Gux6r2nvu4oijeYl23uX5//OxHU0M+EzMxbFF7Bi8WVbYFbb65fB7K7E7/wEkUr2LnCjG4ieVvei+SYE3nJE81UpnN0M3Gd7m/J5LWJQPC0heP8EWMP2PQ20MQ+Rir844em0MeHndCVRAfheQpB5uQzWWzLmKGODfxPVf1+WtDIRWXQucdw/Je5Tr5R+jSxqaR3CR2kVRYrU/sDlxLn5BRFd9LTtD8r6VVg4HEf4DV5JRIBNTUygf02cM/0Iwe7lFh7TmuAyHfBnQlw5se77b1l49GD7/RzRgfMQ7wVvEOfoKbZ3b7T/o9GPWnTWYCIt9q+2ny3Xz/cJ0etKdzO1u52kuJRUSpkRORHYqNx8jiMuiIuI0tlztLWDFSJpsCMMdXXiJvwOEQa8oO1X29u7nlFurocQHlnHEVEEKxAi4bVEOduf2n6vDwiEfTJFrZWol6SoqcX+QEnjSFoNWMfFC0PSg4T/xp1llvRIoihC5YahJWJhReL8HIu4311IiDMXEgOgbQhfto55SSrXwcVEuetNa5FZfR1FesyMRNrG0kQ64FT1gn0R82nk9yxi9nrAusDtwJZuQup9u9r7LqHweLueuA/cUwZ1NxCDuOcJIeRWYtKiMm9JRQr5ncAnhNfRs5I2AHYkxO57CMFkQ9uPNNBOd6L5Vqx6QkbfNoJ+khgkn0xEXVxHVDod2qj4LekB4BmX1KwS6XUSkcb4KREV/jvb1zbSTg/6tQ0hDK9b+/0kfUI8Y6Ylokpfs71pN7b1MPF+tW4RV9YjzpNBRMTdMkTKXyW/YXn//xcRTf91WXYjISh9StxX/0UYebd8MlTSHsQkwpfE+OSh2oRURdu/FLjI9hmK9M0TiOIQq7ZyfyUdQUzQrmL70rKsqwdUr45gTXEpqRRJJxEX/KGSFiBufmsTosRZJYyyI2aDu6Jh5qgbEIJZP+JBcT7xojAV8Gkrol6aiaISwWGEYdwJitzxbQlz5t1t/7lTf8PhoT6SotYu1EtS1Mrgsan+QEl1FDHpaoePwMaEsemadd/PRfihPdQMsbecL1MRYf6zE54WJxFRAwcT6dtN9VdoBmW/1iCE83XdhAp7vZUSBVLzcxlEVP6svAJZiSJY1PbhVW+7N7T3XUBhML0n4dtyNRH1cgBRme2/JSJubEfVxUrvPxpmrbAaMVifH1iy6/2mp8+xVkTzjaL98YjjuhKRtvQD2z8Yzno9fk4rTJ2PIISG3xCR9tcS4sD7kqa1/Z+e7kNPkbQoUdVtAJFmuSAx1t5IkcrWD/i8RMqM8Lwq6y4BLEk8py4gBIdtbd9RBPVPKz4vf0QcUxGTzPcRItavy/68BUxs++ZWvWPVRfNsRBSEOpo4DjMRtgt3EVXjGrJaKefscUThkOvrlp8J7FCliNXN/uxPjLuOqwnEnUSKS0lllBe784gZ52mJlJi7iJzjozvxRb0rinzkG4nS1YcTpV1FvJxc2FdmihVpKdsROcz/KMuWtH1N+btPDd7LgKyjU9TajTJFLekmZfCzBSHojklER/7Mw9JvtwXmtr1WE9qeijBI/T5wu8NPYk4i6mUqYub7LNtPdHIEY4n+Gq8vPHdHh3IvH0oMalcGtnBJ202SrhQReztgNiLSZwNJY7tCj8CRtD0zEYmyAhGVdlaF225JNN8o+jAJUR1uDSLy5Wrb51aw3W/ePxX+RrsQhs5DXFLE2kmJjF0amJtISzycKHDy7mhuZwziObUQYbxe8yh8uG6dqj3f+hHFLLYgfLmOtL1zVdsfzb7UV3c+gUhTu7l8Nw+wPmG2v11F7W1GHOttCSFtRiJ4YK5W3A9KH6YmghXeAoYAxwA/B3a0fUIr+lAFKS4llaIwRN6AyM3dkMgjvwNYzE0oI90qJM1Voq42ImZEryRmhn9FKP3jEep2j3PI20V5gC1HzEzc7GH+I0sTpuy714cw9zVhKakWZYpaMhKKsDRREXLnJgY/CwKXAJfbfk5hdru57QeqFHjKPW0zwtfkfiKq9hbiJfUDIoppOtv7V9Fe0j6KuDaB7TfK544VCpPqGU6ayVJEKuzrxIDy6lZFZ5e2tyL8bZaz/WyF225JNN9I2h+D8F1cgkhB3Nj2yxVs95vo+SI+HE6IIufZ3qLR7fewT996Ny6pVb8gUhG/Js6pG3qw3TGJY7g8EeX2AbCd7Xcq6fjw2xxM2GKsTnhI7WX7v81qbxR9WYuwrXgI2K8+O0TS5O6h/1NdVFQ/YHwiankXwu/xAUKQPdv2Sc3M1qjrxzpE6upHwItERNYpkn5JGMB3TLXQFJeSSikX6VjEZMgnkk4E3impDx35clfCXNcmPA8+BK4gBLQPbR9fZtkHdNKFX09Ryo8AXiNm0C4jQo2fImZNfkKY3DbtQZb0LTJFLRkRks4mKpgdDJzhqDq5FDE7N7B8d7vtLZswK3sbsK/tv5XP4xH3vlkIr7DnVKrIderz6rtGd84RSQOcXkXJcKi/zhV+SH8ExnFdNdkW9WNs4j2rsuilum23PZqvRFJ9ryYsVXV/rY80V/honUr4u7Y8Ja6uT11FpvmJd+vzbN83qnvWiI6Nws9qNiJVbt9W3NMUJtp7A2+7G9XtmtSHcYFFifS8p4lUvUcaSVWri4qagogOepsY3+0HfEwIgv+1/Xyj/e9mP/oTGTDbEMEZUxL7e0ntfaWs3xHvJSkuJU2hPMymJswo97H9eacOMhUm5UsSIan9CVPCyQlVfx+iusDytu9rWycbpO4GNyVh9Pg5Ud1jLGLW62ftfFgnSdI3kLQb4cXxOBHxeZntE8tM6S+JKMoNHL4ZVUYtbQ/M7KjQNAZREvuL8t3ZwK22j6yiraT5jCp1ScMqAC1JFNnYrYXdSzqMriKTm5Aa38zoh2623/RovnbsY7mfq77d3jDe6JK+N9D2pxVttyVpm13aHGj70yKCfNWqY6tIWZ+JiCR6k0jX+yHhd3uI7Y8b3P5h5c9TgKWISLObCK+jV1t1HklaH9jE9k/LZxGFHFYg3oc6anJ/jHZ3IOmbOHgW+HMRlsZo942+J5SXjFeI/NeZiPzx5YkZ9ieIC3+bThaWYFjOve0XbJ9t+0JHFY61gFls/6dEpSVJkjTC/sClwDNE9NKPJZ0LzGH7NGC9JghL/YkU5pkkTWP7a0f57THLKhcBU5b1kl6MpF9KOh3YT9LakvqVF/HaILM2qKsNNHcmotOS7ziSJhrRe0xJS5GGFW4ZLGnaitqds7TxVWnjf8ZetXO4/D3cohij0Z6Gt9z2Z3XC0oAqhaUSBcqIhKXacZe0gKS9qmq3tPl1Obb1v+1EVbbRXSSNUTv+dcKSgM/K3z9WpIYP79+uLWmrUbVRE5ZKpFaj/Z23RJWNbJ1+NWHJ9pfNHsvVnSsrE+8LsxHG9+/Y/hPh6/ufngpLdc+L2QgvsL/Yvp94TmxDRA1tDs3zI+vSn36Ez9IQSddKWqC0+xJRMbejhCVIcSlpMrWZ4U4I4xsedbNXewG/t/0TwrR8HuL6uZPwCulT1D0c363d2No545YkSWcjaSB880w4BZiBiFQ6ipgp3F3SLi5lv6t8ZpT7+C+Ay4HLJB2mSH+rVZiZGxi3GdEKSXWUc2hPwsfxbWKmefpaWkHdOVMbPGwOXFsbUCffXRSVsA4DlpA04fDWKZOitXvA8USkeqPtDgCOkHSXpIVKG1/Xi6K1Vcv6exGRGT1pa2wY8YC4btC+JLB7T9oYzjbnkPQH4FRJt0lavmt7pU+198fdCHPmnrY1sBtiyPJE5b+mI6m/pB9KmlvSkCJ0uYuAWMsM+D6wByEkdN3OQELYuK58Hu74vO433JJ4pjXS9ymBLYnxTG3Z/7Rb99udIel/qv5VTV172xFeSx8ApxfRd1HgdZcqqCMSUkex/dr1sSKx/zuW5R8Xken3RFbKCH+HKin7eyHxPLsKOEDS/YSx+GalHwOa3Y8qybS4JBkFkiYlBkMHu5SoLA+JC4A/2v5nO/uXJEnSmykvxGcBjxDVNgcRnm5rADfZvk1h7v2uK/bpUhjM/8LFE0/SjETZ6p8QJrBXED5zy9t+o8qIqaRaJB0DfOlSmlnSNsC8wHtEQYoBRKrEvSWS4g5gTjdYpjrpfBQpYTsSQvJjRLTi466rLqZhqZRLAOvb/nWF7W9JpPQ8QJgxv1iW9yfGu19JmpwwE19odCbzFIa/qxCC693AOcA3IkctKqsukuZGYPVGRdcysL+bEO0fJKwwtgMeBtay/VZZr3Zc1wNm9GhWH5O0HLAssAhhjH0BcDPwr+FNCEi6rrT/es/2rNv9mpWwx1iGmGiekfBV2r/LejVx6USiEumNw9nWQcA0tlcZ0TOo7rccD7iBsKvo8b2tCBY1AeMw24cMZ536a2ID22v0tL3R6NcYRJXxbYFzCR+tpR1eiBcCF9XEpQramgc4iKhau5dLdex2ofCXmpmwlJkcuB44wRWlVLaKjFxKklFQHlAXAOtJWknSRMTM0ucpLCVJkoySHxJRSnsSht3TEYO7jYBdyozvPbafgcpD0fcjPOSQNNT2E7Y3JGYnfwm8AlyRwlLvRpHCOBRYRNKaZfG4wGREha8rgHeI3xPCI3GTFJYS+CYlbC9C4PmUiKBZQ9JMReCpj5j4I7B1xe0fQVT6ug+4WeEBR0kzqrX7Z6Lq8OgIS+2M5tsVuN/2nrYvKeLENMT1eLqkCco+flWu3+2JLIBuU/bvAEJMWYB4ZkxECA+L161Xi+jZHLih2cJS4VBCqFyYeJ78Efi5pG/2vYgzlrQAUSV1eMLSDMA6wGSSdgCm18htKPYkhJBGhKXxbX9RJl1+DSwuaQ9Jk5Tva9kLtXNxJ6KiYdOo22cRz+zvEWLsdUVYWgWYogphSSV90fadthcCjgWOlXRKo9vuYX9qx/tD23cRv/FxxLn1s3b0qREycilJukF5MK5LKMqLEQ/Pv9q+qq0dS5Ik6QDKAG5LYkCyse2LSoj9IOBZN6HyjaTVgHVsr1A+P0hEJNxdPo8NLGP7ovK57SawychR+HBsTUyOTm77f3xxyov6gBSWkvLutg4wARFdMhFwL1FUoBbtcxVwYkm72ZyoarZPBW2fSoidEwL3E+fsS0SEz4FEufGlSrTmbMCBtkcr1ald0Xwlpep+YMIinoxFjCk/LQLFhcBvaxMGRVwZz/aVo9nOgcBYtrfWt03XNyX81Jaz/VBZNl7p08zNvvaLwL2qw5u0fvk0hCfQUfVRMJIuIfxZ/6f6mKS/ET6EfyfOi4mJ43dFV5GsCFGnAfP39FlVIpZOBp4HTgJeJaLCFgAesn1eWa8WKbU9MND23j1pbzT6NRNxfRwGXGz7CknbEemC9xOi01G2r1RFxvH12yni1py272l0u6Nos3ZcBxKC40sj6hswsVtc2bEKMnIpSbqB7c9tn0CoycsS4aEpLCVJknSDMkN/COG1tJGkW4BJbD/ZDGGpsBtReAFJGwMP14Slwv8RL/UpLPViFPRTVEm6mHgGXwJ8LelgRTnpb3CQwlIC8BvC4+dPROTF/xG+md8HBgMLAkvUpVe9QAzwG0LSzMBChIj1MeFrY2ADIiLjRuDRmvgCvAysNppttCWar4i3Q4gx5HWSflwiwz5VVBV7gxAJpq/7Z/f0QFiqlWM/rSz65v5s+1jCq29IXZ8GEel+rbj2dyEEvPr+ylHI6E7iWNeWDwGOH4GwNAS40vZJtl+3vS5wDLAycJCk5UqqVI2BwMoNPqt+BMxBRIEdAWxMpIvOA/xJ0v4wzOSeECib6mFVMkKWJCxIFiAiwrB9MFFM6VDit72yLK/EA7ZOWOpv+6uasNTlmFdGOUdqqY0nA6covMqmrFunX61vnSgsQUYuJUmSJEnSYiTNAVwJXGV7/SZsfxwiBWYywk9hBcKj4oXy/VbE7G/TPSSSxijRGf2Br4AzbN9Slk9NpNqsDCxu+5G2dTLplZT0pJ8TKWnTAYe62BmUe8QQ4A3bnzWh7Z8AaxPi0oOER9BLdd+PZ/uDRqMwWh3Np6g090X5e1fgD0TUzRa231V4EV0A/F8RnHok3Cv8Tg8m/Hf+QXg7vQEhICuqjF5tu+WpTAqfq8WBKYBzatE+5bvtgOlsb9aN7axKCJ5nAE/VnwflGbUBEW1baTSNpMHAJsDYwEOEGLk4IcaOC8zqYZUFK4kS6kafJgTOJsTXs0u/brX9iaSVbDe1eJKG+Uv9lhB+72hCGzX/rYOJyMVbiKjJOYrA9lZfmORKcSlJkiRJkpZTBjzTunoT73GIcPPnFUbhvyQiFC4BLrf9XImc2tz2A0qvpV5LSfNZjDApngIYx/buksaqCQKSZrb9WBu7mfRiSsrW1EQk0fJEdNKJjspQzW67H2GpsBDwPnAN8IjtZxu555V75xjAmGXwPS4RgbI5UaDgUBfT8KqRdCZwTJ3IOz4RbbMcESE2L+GTc2JPhYkiLL1ZBvvLEabT7wNnlnSpeQmfnLncxbC8mZR9ncT204q07kWIiYtPifTD+yTdA/zG9iMje7Yo/JUWBb4kBJVjbZ/RZZ0JPcwYveF9LMftbdtPSpqMED8fdUnhk/RD4tl5RwtFpfp0xx8Q5/UmwJSE+PIr4InuiHUj2H5NNJqQmKAQ8F5dmyL0kK/LOv8k0uOaElFdBO/jiAIDhwFn2z5f0p6EAPzHZrTbSlJcSpIkSZKkzyDpbMIE82Ai0uUdSUsREQwDy3e3294y0+F6L2V2/T5CgHRJITmPmPHtT1TT+ScxqHu/bR1Neh018bGLCNmfqOi1KiH2XG274RS4Lu2OB0xt+6FaZFJZPgmR9rYQ8CSwr+3/KUk/Gu20JZpP0iaEV9VKtj/qIgzMREQsjTO86KnRbOcG4C/AjS6VsiRtCKxFVKhbEDjC9lmtEkFKH04iUv7+YvtDhX/RFMQExnzlvwtsbzsKYWkC4FYievY9SfMTkbaXEel/gwkvwmurekYpUjV3IvzH3gP2J1LOtgJuBw6w/W6j7Yxmn+qrGG5aFj9FGLjPT0wsQBzvTxqZCJJ0BeFndQchWt1u+7n6fkg6GrjE9rU93qnu9WVdQpT8xPbaZdl9wIa27+n095IUl5IkSZIk6TNI2o3wOnkcGA+4rMyiDyYGAcsRvnnvZ9RS70VhAvxP4Ewi5eZ9SS8QKSQPAJMCH9s+qY3dTHohkh4mvG8+InyHzgIGlciSfkTEyBslcrHKqMmjCC+bC4lB8kTA+YR58huE19OMbqDkebui+RQeT7cCq5XIq5mJNKpFgH8BBzmqbk5k+02Fj82XI9nkiNrZFljAxSy7RKJOC7wLvEVERw21vU4V+zUa/VqQ2Mef1i1bEvhP6ddMhEfUTrY/Htl5pfAA/AOwHeG59KWkT4C/EQbwzwJf9DRaZwRtLgR8QVwTGxPeRpcRUVNrABcDO9p+b4QbqZg6UWdPYBnCJ+z7pZ8n2b6vbt3RflZrmHn2WsTk0o5EJGHNPPxBonriR5J+DOzt0TTVH51+1P4GfkBELQ0CriUmvF62vUWnC0uQ4lKSJEmSJH2IMpu8M2FEej3x4jyYSOW4WdI4XWfdk96JoqLOUUS0xIvANba36/oC3hdeyJNqkPQ9wix3KLADIfZ8BPyUGLTeBNzrCkqaD6ft9YiUnvsJ0+5JCTPiKwmB5FrbW5d1R/ucbWc0n6TpCWPnnYFPiMiiD4B7iIid54rI1Uh0yUAiamU12y9LWgxYjzChfhM4y/ZpCnP/T1octXQBEUl1tCLdenUikuot4NKy7wNsfzGqY1DSr1YG5iREsxkJcWHr4dzbGn5OSdqR8IHaqG7ZxESp+wHEMR7b9kKNtDOafVLdObwNcT59RJiLz0+IljfZPqzBdgYSItqZLmmHkuYE1gS+tr1TWfanss4zI9pWoygq700G3G/7DEmrE/epZ4hU0o/7wntJVotLkiRJkqTjKS+ROLwSTiEq0/2SECduAnaXtIvtj8p6Hf0C19eRNLXtT21vQAw4/gOsJWnFroPyFJaSGrbfIQSQz4jUsLWJiJuviUiRj4B/N6ntUwkx6Q3b59s+goiQ2JOInLgEGhJDZwWmAk6TNL7ttwlh4lXCVPtS4KWqhSUA208B9wJbEtEWnwB/tX0MIQzMXYT7Ru6rA4iB9molkmTP8nlj4j4+f4mg+qz0qSXCUuF6YCyFV9HhRIr1ksBSwEKSZijPnpE+WyTtRAg9JxBeVe8D0wCfSJp+OPe2RoWlCYhI3sPL5z8oKvz91/aFts+xvRQROfRNtbJmU7efewC7AovY/tr2XcC5hC/R5aVPaqCpoYQwua+kbYoAeL/tHYBD6vqzZ5OFpZ2ISZLbgNUl3Qi8avsQ25fXRbt1/HtJRi4lSZIkSdLRlBfis4BHiIiBQURayhrE7OdtZbb5XVdsIJ5Ui6T5gGWJAdGHRGTG5Q4T3V8BxxMRBCu1sZtJL0TfrmT2Q0JMOp4YRB7ruqpeFbc7QS2dqIgPRxPeORMBd5SBbFVttTyar6QVXUUIKosQ0TZX1qX6HEcYRe/caLsl/WxTwhj8YuBA2/8tv+e5wGLNEM9G0acxiVSq0wkh72PCH+ed8v3dwMa27x3FdsYFtgV+TKRkHWH7dUkLE+LjDIQX2GkV9v004IOScjUDUXlv7i7n638JvaclwoakFYnr4tXyeX+i4uHJwJYVCGpdr4X+wNKEgDYhcLHt8+qip5oSAVe3/TGI6/VRF6N3wkx9R+Dftjepuu12kuJSkiRJkiQdTXlpfgAYi4ha+ADYkEiHe5So3vN22zqYdBtJ9wKnEZECsxGDgrEJU9e7yjoz2X68L6QQJNUgaXZgadsHlkgHESLFX4AXbM/cjPOlCNt/BZ4mytK/WZbvB6xse8baeo0OYEs033Pl7x+Wdn8KbGr70ka2PZI2lyfSCzew/WSX78YmRJEDgDncYOU2SUNtvyJpGkIQebPuu9OBx23v28rrXtLviOpiZ5bPUwGv1ImYOwCzlwi57mxvfEKoWpVI97scOAGYhLjX3W374Yr6Pi5xLx2LiKjbCLjC9unl+ymI3/YPLn5dzaYIK5sTAuy6wA2Oyq5DiUiuRYDdbB9eQVurAEsQqZtPA68RZvArEmLgS422MYr2a+LSZsAviAiqA4CnPKw63RDbT/WlZ1mKS0mSJEmSdDxldnJLIsR+Y9sXKUobDyIq7zSltHBSHZL+CExv+7d1y8YlZnjXBH5Z1cAr6VsoqkFda/vQmkhRli9NRA3sUxNmKm53YmKwOjMxiL+JMPSetPz/PsJT5usGRJe2RfNJuhnYy/b1kgYxrCra34EXiH1/vnzfIwGtbv9WJqJoLiZMwu8jijIsB/ze9k/K+i2JPJU0HRHpsxghKH3VJUptNaI63+oOk/NuCwTlvJmH2LdniyhaaQSNooLfx4Rh9K8J4WaxOpH+LEKw27vV0byKyornERF+FxNVDz9XVHadxfYhI93AiLdbM/Fek3gfOIPwtvqMiFi6sU7EbNo+135LSXMAZxNRb7X07n8BDzRb3GoXKS4lSZIkSdJnKLOBZxGDki1HlaqQ9A7K73YH8Gvbd3UdqEk6nzAqvq1tnUx6JWWQv47tFcrnB4m0pTsljQUcCQy2vVqT2hfhhbQ8MDth8nwSkT52MLBtfRROD7bflmi+kq61i+0lFYUSdiUEii+IFK71bV9XQTv1+zc74WPUn7je7y4D9M9tP9asFKYR9OtiIm3pryV97OfABsAEwGaEUDCT7VtHdtwVptWL2z6/y/IBhFC3PbC97ccr7PtSwHbltxsITEl4EM5HGM4/Q0SjLVrWb5Vg1zVl7ZeECf7bROTfFXXf9aRCXC1a6Byi4tx15frcBNgCWNT2G5XsTPf6cwhxDl0kaWpC5JuNSNvf3/YnrepLq0hxKUmSJEmSPkcZkFwJXGV7/Xb3Jxk5kuYiBuL/IUpy32j7vboZ4FOAW2yf2NaOJr2OIiZdbXt7RZn3RWyvWff9XERp94cqFl+mIvxzvg/cXoSQOYlB/FSEmfdZtp/oabvtjOZTeA2dSYj1qwJDgONsXy7p18BcwE6NiD0j2b8diP1bCvhPK6NqSh8GECbYR9h+VNKRRAn5q4FxCJFimW5u6yeE2fo9wJ627+ny/emEh1VlFQzLb3ctsIqHpWpOQKTkrUKkpm1o+9xWCnZ1/fumzZIqtxWwfunTnRVsf0+ieuKetl8syy4D9rN9e6Pb72YfpiQiGT8lfofHyvKfARO7Sams7SarxSVJkiRJ0uew/QDxcrkPNFxxJmkikuYFFicM2O8m0hn2ljRvEZZmJLxlLijr52+ZACBpHEL86C/pKGA3YKe677clIjgeguqqRJZ0u8OJQfH0wNWSLicEpf2JdKqPbT/R03ZLNN/6hD9NbRCO7Q9t70aIFeM1vjfDbXtW4EvgHCJKByK978ry98xENFgjwtLI9m93ojrdpK0WlkofviDS8u6VdBURsbWf7WNsH0RUjvvRqLZTUv7mtz05cBdwgaSjJU0iaWBJXZsbuKKs3/C9TdKYtj8nIuhmKMv62X7P9h1EJbaFbZ9b9rWlwlJ9m6VfX9s+lEjZ65GwJGlFSQfWHb9jiAqRK0raWOF7NHWzhSVJP5W0PYDtF4hIsWuB8yXtL2mg7VtqwlJffJaluJQkSZIkSZ/EwTO1v9vdn2SE7AV8Zvt1R2nzdYjUk10l/RY4Aji6LpIpf8ukJixNZPsAIsLmLcK4d8WSggIRpXFgWb/Kcc+fgBNtL2x7a2BqIrXnH8D3bV9IDHAbaXcqoiLcpooKW+OV7dXKxX9EmEJXSom++jdx3B6zvbTtdWw/BXwtaVriuO5S1m/m/s3S0/3oKZKWlLSg7eMJX6SLgJWKMIOklYEvbT/Sjc2NQQhT2N6bqII3DuFZdSRhtH2o7feruLdJGgysK2kiIgp0iXKdTClpeYXR/AHAy420UxVl8kAlsu+/EH5UdedAd7dzKfBn4E+SbiWihfYCxici7H5EGJozutseTR4GLpW0lqSDgDfL/WEdYrLrKUnfnNN98VmWaXFJkiRJkiRJW1AYEm9YvEH6AQsT3hhvEP4rHwLYXrJ9vUx6I5LOJsyKDyYMgd8pXjM/BwaW7263vWWVnjIlMmFm2+sXYaWfh1UPOxu41faRDbYxL7AQ4UW0MpGW9ghwtu3bSzTfJcB8RXStcv+mI8zIHyYqXL1HmCE/UFIBdyNEp4MbSPdr2/6Nol9jEkblLxJV3G4hSsi7RJnMQ/hpbWr73yNKKav1V1GR7QriXPywtg/lGE9Vtv1qhf2fn6iQ+BjwBJFa+CXwPvA5EY31uO3TW3VMR9JX1R2PmhH3RES08Waje14pvK3eI363lYB9bf+l6n6PpP1aCvcAIoXyFGBc4ADbtajbn9u+sVV9agcZuZQkSZIkSZK0izGBV8rf6xLl498hKuzcSaQdrQhNn3FOOo/HiGileYHTJW1o+2oiLe0+4DlKdA1QSfqJoirlr4CZJE1TUnq+KKIERJTLlGW9RmhbNJ/tp4nolkmJY/ge8DtJmxIm5dvZPri2eg+b6a3Ril8BexPRRQsSaXtrSZqeuFeNTQiZ/y7iyIhSyiTp+8Xv5ylg+TohZQLgE2AoId5VQhHs5iPEuvsIv6oJCZFjU9u/sP1H26dX1eZo9m9OSVNKmgT+J2qn9vfewGXdFZYkjS1ptvLxKGAuh8fifMAKkl6RtH5Zt2kpaGXbE5WPxwAT2l4EOAjYXdLfJc1TE5b6YjpcjYxcSpIkSZIkSdpCmW0+nDALngfYmiiz/oGkE4ky3fu0s49J76RECOwMDCAqja0BDAaOsX2zpHFsf9TT6JqRtDsRsCERFXIjsLuHlaffl/Ai2ryB7bctmq/+WEn6DeGFdh4htswHTEyIS6830EavjlYsQuFOwHSEgDlz6dvNwD/rfuuRVYibmfCMOgB4gLi3vUMYwE9L7ONrtteosN/XAFfYPqx8npgQa74PXEWYSz/j8GNqCXURSVsRlfbeAW4gJg7usv1W3bpzA3u7m0bp5d9MApwI/IR4Vszf5fsVgB/Z3q/xvRlpP6YA9gO+R5h1z9Pl+72ASWxv2sx+9AZSXEqSJEmSJEnaRnkxnxYY4FLaXNIg4FZgVdtPtzuFI+k9KExxPy1/T0FEB7wCnEykH60G3FC1KClpduAXxeOJkrq1PTGwPZxIf7qMiFJ5o4GUsbWJ0vXrlaiLpYEPGJZCdizwgu2PR5SW1cP925CIAruGSKkaRKRuDbZ9rKIC55S2/9ZgO23Zv270ayKgv+3XyufNCWHofsKvZwmifPy/urm9AYRItWdZtB1xnt5DCCH1ptYN7eMIBLutgNeAOQgPoqdtb9RIO6PZp1pq4CDgfOC3RDTcKsBkRPrhNS7V8yRdAWxdIue6s/0J6oS+WwlfpQuBHWy/VaLNJrX977JOpSLzcPrzPULAmxjYFTjH9ieSxiZ82P7Tin60mxSXkiRJkiRJkl6DwpT2aOBlR3n5Pv0ynnSfMnA+i/DnuZEQQJ4iopZusn1biYB41/YzVYqSZfB7ne1DJA21/UpZvihRSW05YA/bezVyzrYjmq9E69wJzEYcz8OIsvUrEv5Va9u+tk4w6PFx7Y3RigqT7uOJim4vAhcTIsgKwFa2z5M0m0vVwdHc9gREhbapgQ3cPSPw0W1jVILd0USk1Putvp+WdMoNgIVsf1KWLU2IwEfYvl9hwL96Tbjt5nYPI86jd4ApHJ5gZxLX4Z8J77W/l7TLpqJhfkvzE95WhxARlTsDvwOes/2HZvejN5DiUpIkSZIkSdIrUBgkTwcsCRzn8LPJqKUEAEkzENEkYxEDtw+IFLXBwKPAb2y/3YR2VwPWsb1C+fwgsL7tu8vnsYFlbF9UPjd0zrYjmk9RVWxFYBPCy2ojwqtqCkLofb2qNntbtKKk5QkR4HUi8uRaoqLaNsCpwKmNppMpqvD9A7ja9m8b2dZwtt3rBLvSr37A8gy7Vo90VHb7VgRiD7c9mDAq/weR9nei7Q8lzQXsQBiX79HQDjRAiQRcBXgS2LaIT33+WZbiUpIkSZIkSdKrqPPq6PMv48nooTDL3pJIPdnY9kWSfkBEMT3rUrmt4jYfJESB7SVtDCxie8267+cH7mjWALKV0XySJge2JQTe04C/NjvSpd3RikUEmRFYgEgFHAicbPuRKvsiScC0VUfVlW33KsGuS98mBxYlooo+AI61fXdPIuHK9b8ecKftByUtSURGjUWYrdcE3v62v2zHuVSf5ihpUEnx/E48y1JcSpIkSZIkSZKko5A0IZEiNx6wpe17m9TOOITh9GRExbAVgJ/ZfqF8vxUwvys0Z+7SflOj+ST9iRDmxiAG5w+W5fMQESA/Bpay/UQV7Q2n/bZFKxZfpEkJg+0PiWpw85T/pgceBy6x/VSz+1I17RLs6lLEfkZcK7MS1T+fAj4iqoK+aPuIHm5/dmB3opLhXcAFZSLiN8BaRHXDXdr9m3U93ioFBtrZp1aQ4lKSJEmSJEmSJB2Jwmj6SuAqRxnyKrc9DjCR7eeLl9MviRL1lxBpR89JugXYvHi+NDOiqPJoPknrEVEflxKphR/aPkDfrhq3iO2bqmhvFH1pebSipAMI8+uPgLNsn1yWDwXmJCJtbrV9Ziv6UxW9Ib1Y0uNEdOHkwDRAv/L5c+DzRqL8SvTSucBPiepzp9m+QdL4RNTd0bbfqGhXekyd0PZb4FHbd7S7T80mxaUkSZIkSZIkSTqWZqUbSTqbqEB3MBHV846kpQiz4IHlu9ttb9lpaS9lgH4PsEDx5pkOOIow9h6b8O+5mxBd3uu0/RsVJTLrMGAxYCUiOm0f4Hoiiutz4PvAq+0QZ6qg1YJdXZrbEoQR+vJl+RDC1Pw+2/s2sP2aWLMRsBBwCrAMkcr4LHBmXeRd0/a5rh8TAl8R3mTv1QmyInSWr8s6/wTmbEbKbm9jjHZ3IEmSJEmSJEmSpKc4eKb2d4Wbfowwt54XOF3ShravBvYH7iNSc3Yp66rCdlvBVMAMwHYAjhLwMxBVzV4G3gXGcyn33mnCSjc4HLjb9se2zyKEpu2I3/wU4HmilP0X0Jn7XxM7WiQsTVzXzpPAQElrSBriMNk/hkiR6zFF0OlHpNudZvsGRxW2MwgfprXr1m3aPtd5Kp0OXAPsBfxKUfWu1nat/b2Bbb4LwhJA/3Z3IEmSJEmSJEmSpBeyPyEaDSAiWtaQtDhwjO3TJF1o+6NWmwZXQYnyGh/YWdJLwP3AlbY3rzdahv/1j+l0SrrjZcBcko4kfudfE1XHriXS5L5n+6729bLjOFfSr4iUt+clnUKkHA6VNCawLHAo9Px8KuflV5JuBtaT9KLtJ2xfLukm4rdr6vlaFw22FvAq8BvCR2pRYAaF+f+15b7wY2Aa29c2oy+9kUyLS5IkSZIkSZIkKaiuTLqiCtdBwCvAyUQq3GrADW5DefdmUFJ3jgN+BKxi+9E2d6npKCqpzUj4aC0NTAFMZfvLunUEnRm11GpKRNE4RMTbXrb3UFRRXAD4HvC47dN6uO1vpbgVUXRfoOarNJhIO1u053swWv0ZSIiTZ9o+oyybE1gT+Nr2TmXZn8o6z7SiX72BFJeSJEmSJEmSJEn4ZpB8FvAIcCNRSe0pYA3gJtu3FXPvd6v2eGo3zTRH760UP6C5iN93EPBP28e2t1edSREpJwROAKYEtrB9RZd1eny9SFoGmJsQdmrX5FREdOFZtp+s+SE1sBvd6ce0RLrbQoQf21G1tDdJk9p+vZnt92ZSXEqSJEmSJEmSJAEkzQA8AIwF7Ax8AGxIREc8CvymeMj0SZpljt6bKfs8lEjjWpkQRV5rb686gxKd9AQh+qxje52yfGXgL4QNz4LAyz0RferS0FYE9iOEpU2Iio1b2/6gkh0ZdT+6Rk/1JyLeliEEtYttn1eXUtp0kas3kuJSkiRJkiRJkiRJoQwctyRKp29s+yJJPyAiW579rpjzfteQNBYwgUsZ+77mNdUMSmW4i4jooRVs3yipfy29sFR2O6OWZtpAO4cDl9m+vvxOxwOrAwfZ3rWFFfFWAZYgKi0+DbxGiGcrEveKl5rdh95MVotLkiRJkiRJkiQp2P7S9iFE9bSNJN0CTGL7yRSWOp+al1JXbH9WJywNSGFp5EiaophVb0VE9V0sadc6YWk34ELbn0rqse4gaWlgFmDhknb2me11iSqOT0JzfbFqfZe0JvAH4EHgJ0SU22S2jwc2tP3SiM6t7woZuZQkSZIkSZIkSTICvoteRH0RSWPb/mQk3/cr1ciWBBa0vVsLu9dRSFoQmB84DJjO9sOSZgJOBSYDzgMWtj1vBW3NRpjozwpcA9wFPGH7w7p1mha5VJfqdg5wku3rioi0CbAFsGhNlPyuk+JSkiRJkiRJkiTJSPguehH1FST9ElgFeBu4GziHqOrlOk+fb35TSTcCq6dgMGokbQasChxt+6KybHVgNuAc24/0xH+oTtAZCxhCVIabGfgdkYJ3J3B+vcDUbCTtCUwO7Gn7xbLsMmA/27e3qh+9mRSXkiRJkiRJkiRJkj5HKRt/O1HBbEJgemAf24938QaqiUybA4Nt79u+XvduJK0KjAtcb/tFSesC6wDPAcfbvrPB7dciyGYHDgQ+Iir67Wv7BEkbEGmq+zW0I6Pux4pEdNaOReiaDNgLeAj4DOgHbGp7jmb2o5NIcSlJkiRJkiRJkiTpc0g6BvjS9pbl8zaEV897wIDy3yG275U0HnAHMKftz9vU5V5N8R/ajBBdngRutP1PSeMCvycqqF1PRPc0VC1N0t+AfwDHESlxJwFXFwPvAba/aLbpuqTxgW2BJYnKcOMC6wJTAF8Cp9u+87taHa4r/dvdgSRJkiRJkiRJkiSpEkljAkOBaSWtaftsQhyYjBBGHgAWBl4p/2R2YJMUlkZMEXKOkvQlsAGwqaSjgEts7y3pYmCGCoSlicufV5Q2HywRU3tJ+h4hDtJMYanQH9gbmBJ4loie2qfrSiksBRm5lCRJkiRJkiRJkvRJJK0MbE1USp/c9rTDWUfAgBSWRo2kQcC/gfWAqYHFgR8BVwBn236trNeQN5mk/YFJbf+2fJ4RuACYz/ZHjezDKNodmzApf6iYeP/V9l3FsPwE4IfArrZPTv+1b9PjkoBJkiRJkiRJkiRJ0ttQ0K9UiLsYWBa4BPha0sGSpqhf30EKS91jMeBN2w/avtz2VsCDRCTTJLWVRld0KSl3SFqlCEkHAhNIelnSIcAxwAm2P5LUr6qdGQ7jAftIehWYyvZdALYft70gsCkwaVmWwlIdGbmUJEmSJEmSJEmS9BmK11J/4CvgDNu3lOVTA9sDKwOL236kbZ3sUCQNAU4FLiU8kF6WtDSwqu0Ne7jNmqH6UOByYN3abyNpXmBG4EHb91WxDyPpxwS23yt/30pEZF0I7GD7LUnTE9FU/67vdzP71EmkuJQkSZIkSZIkSZL0CYpp92LAOYTx8ji2d5c0lu3Pyjoz236sjd3sSGoV9iQtCawOvAYMAX5G+FXd1ojgIukw4C3be5Wos0/K8kG2P65qP0bR/uHAO8AUth+QdCawHPBn4OfA320f0+y+dCIpLiVJkiRJkiRJkiQdj6TBwH3AtKV8/BDgPOBFIpJpcuCfRIW499vW0Q6iVglN0o+JqmlDgZ2ACYEFgQmAF21fXoHP0rbAhLZ3qVu2HTDI9t4N7Uj32h8MvE9UqbsKONH2h5LmAnYAHre9R7P70amk51KSJEmSJEmSJEnSF5gVmAo4TdL4tt8mUqpeBf5OpHK9lMJS96mrhHYa8DkwmBDrVrV9pu2jbF9eUXPXALNJWk/SHEXs2YiIQqsZr1eOpP6SNgSmLFFXfwXmA86UtIrte22vQUQvfeMPlXybjFxKkiRJkiRJkiRJ+gSSBgJHAasQIsg1trfrGlWTlb5GTZ0X0sLAOjVPJUmzEgbb8wFzAo9W5T0kaXFgUaIK3UvA7bYPbKa/kaTZgd2B54C7gAvKfv8GWAt4D9jF9lPNaL+vkOJSkiRJkiRJkiRJ0vFImtr2c+XvHxIRKD8FNrV9aRu71rEUse54YBoicudW2x+U75a1fUUT2hwEjAn0t/1mWdZUMVBSf+Bc4ny5ATjN9g2Sxge2BY62/Uaz2u8LpLiUJEmSJEmSJEmSdCyS5gOWBVYCPiR8li63/bSkXxHiyI22V2pjNzuOusilRYEVgYHATcA9wFO1SKJOjgKr85TaCFgIOAVYhkivfBY40/aDZd2O3c9WkOJSkiRJkiRJkiRJ0rFIupfwBLoemA1YGhgb+Ivtu8o6M9l+PMvHj5o6wWU8YHzgB7bvkLQMUSXuIyKy5462drQiJPUj/LgOs31dWbYCcCJwqu0d2ti9jiGNqJIkSZIkSZIkSZKORNIfgQdsH2b7YdvnAL8DngDOLf5A2H68/D+FpVFQZ+J9ErA/cLikh4A3gC2B/wIvt6l7lVKikb4CbgbWkzQjQDEpvwm4tqyX2skoyMilJEmSJEmSJEmSpOOQNCFwB/Br23d1jUqSdD5wiO3b2tbJDkXSSsBWtn9ePq8NbAesZvvpsqxj08SGY/A+PrAvIaBBVMWb0/aibeheR5LqW5IkSZIkSZIkSdKJTEVUhNtU0orAePBNmhNE+taP2tO1jucLonIaksa0fSZwPuFHBECnCkswrO+SlpG0GzA1sD1xPo0BfABsWtbpN4LNJHWkuJQkSZIkSZIkSZJ0FJLmJcrVrwHcTaRr7S1p3uIXNCNR+euCsr7a1tkOoZb6VaKWBgI/l7SZ7c/LKnMTolNHU7efKwIHA+MQaXFHARfa3sP2n2w/Cd9KE0xGQqbFJUmSJEmSJEmSJB2FpGuAK2wfVj4PBfYDJgQuAn5NVIw7smZQ3b7e9n7qKsMNBS4HliMiw04CPiPSD2e0vVgbu1kpkg4HLrN9vaSxiKqCqwMH2d61k9P+2kH/dncgSZIkSZIkSZIkSbqLpF8B2D6spCwtDGxB+OUMIYQlbB9Z/p/C0iio86raEfi77deA14BZJa0MPA88CcOqybWnp9UgaWlgFuBdSQ/bfh1YV9IhwOzQ2Wl/7SDFpSRJkiRJkiRJkqSTGBN4pfy9LrA08A5wNvApcCzwAvQNIaTFPE9Ef9UzMTCL7Xugz4h1LwO3ArMCK0m6C3jC9v3A/dDZhuXtINPikiRJkiRJkiRJko5B0hDgcCJKaR5gayIF7gNJJwLP2t6nnX3sVCTNSlRNuxi4jxDpbgeWtf10Jwsutb6XFLghRKTbzMDvAAF3Aufb/rCN3exYUlxKkiRJkiRJkiRJOgpJUwDTAgNsX1eWDSKiUVbtdCGknUhaHFiUMEx/Cbjd9oE1X6b29q5n1CLYJM0OHEhUEpwL2Nf2CZI2ACaxvV9bO9rBpLiUJEmSJEmSJEmSdDSSBgNHAy/b3r6ThZDeQBHqxgT6236zLOt4sU7S34B/AMcRKXEnAVcXA+8Btr/Ic6dnjNHuDiRJkiRJkiRJkiRJTyml5Schopb+WBZ3tAjSbmx/bPvdmrBUlnX0MZU0cfnzCttf234QWBWYQtL3gK/gW+bmyWiQ4lKSJEmSJEmSJEnSsRSh4Eng6BJ50vERNkn12P4v8AiwZ93igcD/AZ+nqNQYKS4lSZIkSZIkSZIkHU9NHEhhKalRotqQtIqkGQm/pQkkvSzpEOAY4ATbH0nq186+djrpuZQkSZIkSZIkSZIkSZ+i5p0kaShwObCu7UfKd/MCMwIP2r6vnf3sK6S4lCRJkiRJkiRJkiRJn0TSYcBbtveSNLbtT8ryQbY/bnP3+gyZFpckSZIkSZIkSZIkSV/leWAsgDphaTtgu3Z2qq+R4lKSJEmSJEmSJEmSJH2Va4DZJK0naQ5Jg4GNgHMAJKmdnesrZFpckiRJkiRJkiRJkiR9FkmLA4sCiwMvAbfbPrDmy9Te3vUNUlxKkiRJkiRJkiRJkqRPI2kQMCbQ3/abZZmyumA1pLiUJEmSJEmSJEmSJEmS9Jj0XEqSJEmSJEmSJEmSJEl6TIpLSZIkSZIkSZIkSZIkSY9JcSlJkiRJkiRJkiRJkiTpMSkuJUmSJEmSJEmSJEmSJD0mxaUkSZIkSZIkSZIkSZKkx6S4lCRJkiRJkiRJkiRJkvSY/wcNOuinQZmjzQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 1440x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Building lasso regression model\n",
    "from sklearn.linear_model import Lasso\n",
    "\n",
    "lasso = Lasso(alpha=0.1)\n",
    "\n",
    "# Fitting model to training data\n",
    "\n",
    "lasso.fit(X_train, y_train)\n",
    "\n",
    "# Extracting coefficients\n",
    "\n",
    "lasso_coef = lasso.coef_\n",
    "\n",
    "# Plotting coefficients\n",
    "\n",
    "plt.figure(figsize=(20, 6))\n",
    "plt.plot(range(len(X_train.columns)), lasso_coef)\n",
    "plt.xticks(range(len(X_train.columns)), X_train.columns, rotation=60)\n",
    "plt.axhline(0.0, linestyle='--', color='r')\n",
    "plt.ylabel('Coefficients')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\William\\Anaconda3\\lib\\site-packages\\pandas\\core\\frame.py:4169: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  errors=errors,\n"
     ]
    }
   ],
   "source": [
    "cols = ['VLF_PCT', 'HF', 'HF_PCT', 'MEAN_RR', 'RMSSD', 'pNN25', 'KURT', 'SKEW']\n",
    "\n",
    "cols_to_drop = [col for col in X_train.columns if col not in cols]\n",
    "\n",
    "X_train.drop(cols_to_drop, inplace=True, axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['VLF_PCT', 'HF', 'HF_PCT', 'MEAN_RR', 'RMSSD', 'pNN25', 'KURT', 'SKEW'], dtype='object')"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_val.drop(columns=cols_to_drop, inplace=True)\n",
    "X_test.drop(columns=cols_to_drop, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Index(['VLF_PCT', 'HF', 'HF_PCT', 'MEAN_RR', 'RMSSD', 'pNN25', 'KURT', 'SKEW'], dtype='object')\n",
      "Index(['VLF_PCT', 'HF', 'HF_PCT', 'MEAN_RR', 'RMSSD', 'pNN25', 'KURT', 'SKEW'], dtype='object')\n"
     ]
    }
   ],
   "source": [
    "print(X_val.columns)\n",
    "print(X_test.columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Building linear regression model\n",
    "\n",
    "model = LinearRegression()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model Performance\n",
      "Mean Absolute Error: 1.7886.\n",
      "Mean Squared Error: 6.5897.\n",
      "R^2 Score: 1.7886.\n",
      "Accuracy: 97.66.\n"
     ]
    }
   ],
   "source": [
    "# Fitting lr model to training data\n",
    "\n",
    "mlr = model.fit(X_train, y_train)\n",
    "\n",
    "# Evaluating performance\n",
    "\n",
    "scoring(mlr, X_val, y_val)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model Performance\n",
      "Mean Absolute Error: 0.6471.\n",
      "Mean Squared Error: 1.2485.\n",
      "R^2 Score: 0.6471.\n",
      "Accuracy: 99.15.\n"
     ]
    }
   ],
   "source": [
    "# Building Random Forest model\n",
    "\n",
    "rf_model = RandomForestRegressor(random_state=7)\n",
    "\n",
    "# Fitting model to training data\n",
    "\n",
    "rf = rf_model.fit(X_train, y_train)\n",
    "\n",
    "# Evaluating performance\n",
    "\n",
    "scoring(rf, X_val, y_val)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model Performance\n",
      "Mean Absolute Error: 0.6380.\n",
      "Mean Squared Error: 1.2978.\n",
      "R^2 Score: 0.6380.\n",
      "Accuracy: 99.16.\n"
     ]
    }
   ],
   "source": [
    "# Building Bagging model\n",
    "\n",
    "bg_model = BaggingRegressor(random_state=7)\n",
    "\n",
    "# Fitting model to training data\n",
    "\n",
    "bg = bg_model.fit(X_train, y_train)\n",
    "\n",
    "# Evaluating performance\n",
    "\n",
    "scoring(bg, X_val, y_val)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model Performance\n",
      "Mean Absolute Error: 1.5979.\n",
      "Mean Squared Error: 3.8767.\n",
      "R^2 Score: 1.5979.\n",
      "Accuracy: 97.79.\n"
     ]
    }
   ],
   "source": [
    "# Building Boosting model\n",
    "\n",
    "ab_model = AdaBoostRegressor(random_state=7)\n",
    "\n",
    "# Fitting model to training data\n",
    "\n",
    "ab = ab_model.fit(X_train, y_train)\n",
    "\n",
    "# Evaluating performance\n",
    "\n",
    "scoring(ab, X_val, y_val)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Importing KFold and Cross-Validation\n",
    "from sklearn.model_selection import KFold\n",
    "from sklearn.model_selection import cross_val_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Defining a function that compares the cross-validation performance of the previously run models\n",
    "\n",
    "def cv_comparison(models, X, y, cv):\n",
    "    # Creating a dataframe to contain all averages and measures\n",
    "    cv_accuracies = pd.DataFrame()\n",
    "    maes = []\n",
    "    mses = []\n",
    "    r2s = []\n",
    "    accs = []\n",
    "    # Loop through the models, run a CV on each model, and add the average scores and CV scores to the dataframe\n",
    "    for model in models:\n",
    "        mae = -np.round(cross_val_score(model, X, y, scoring='neg_mean_absolute_error', cv=cv), 4)\n",
    "        maes.append(mae)\n",
    "        mae_avg = round(mae.mean(), 4)\n",
    "        mse = -np.round(cross_val_score(model, X, y, scoring='neg_mean_squared_error', cv=cv), 4)\n",
    "        mses.append(mse)\n",
    "        mse_avg = round(mse.mean(), 4)\n",
    "        r2 = np.round(cross_val_score(model, X, y, scoring='r2', cv=cv), 4)\n",
    "        r2s.append(r2)\n",
    "        r2_avg = round(r2.mean(), 4)\n",
    "        acc = np.round((100 - (100 * (mae * len(X))) / sum(y)), 4)\n",
    "        accs.append(acc)\n",
    "        acc_avg = round(acc.mean(), 4)\n",
    "        cv_accuracies[str(model)] = [mae_avg, mse_avg, r2_avg, acc_avg]\n",
    "    cv_accuracies.index = ['Mean Absolute Error', 'Mean Squared Error', 'R^2 Error', 'Accuracy']\n",
    "    return cv_accuracies, maes, mses, r2s, accs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "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>Linear Regression</th>\n",
       "      <th>Random Forest</th>\n",
       "      <th>Bagging</th>\n",
       "      <th>AdaBoost</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Mean Absolute Error</th>\n",
       "      <td>1.7726</td>\n",
       "      <td>0.7147</td>\n",
       "      <td>0.7696</td>\n",
       "      <td>1.7174</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Mean Squared Error</th>\n",
       "      <td>6.7247</td>\n",
       "      <td>1.8532</td>\n",
       "      <td>2.0926</td>\n",
       "      <td>4.6730</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>R^2 Error</th>\n",
       "      <td>0.9340</td>\n",
       "      <td>0.9816</td>\n",
       "      <td>0.9792</td>\n",
       "      <td>0.9536</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Accuracy</th>\n",
       "      <td>97.5915</td>\n",
       "      <td>99.0289</td>\n",
       "      <td>98.9543</td>\n",
       "      <td>97.6664</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                     Linear Regression  Random Forest  Bagging  AdaBoost\n",
       "Mean Absolute Error             1.7726         0.7147   0.7696    1.7174\n",
       "Mean Squared Error              6.7247         1.8532   2.0926    4.6730\n",
       "R^2 Error                       0.9340         0.9816   0.9792    0.9536\n",
       "Accuracy                       97.5915        99.0289  98.9543   97.6664"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Create models to be tested\n",
    "\n",
    "lr2 = LinearRegression()\n",
    "rf2 = RandomForestRegressor(random_state=7)\n",
    "bg2 = BaggingRegressor(random_state=7)\n",
    "ab2 = AdaBoostRegressor(random_state=7)\n",
    "\n",
    "# Putting models in a list to be used for Cross-Validation\n",
    "\n",
    "models = [lr2, rf2, bg2, ab2]\n",
    "\n",
    "# Running cross-validation on models\n",
    "\n",
    "comp, maes, mses, r2s, accs = cv_comparison(models, X_train, y_train, 4)\n",
    "comp.columns = ['Linear Regression', 'Random Forest', 'Bagging', 'AdaBoost']\n",
    "comp"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Creating hyperparameter grid for Random Forest\n",
    "\n",
    "# Number of trees in Random Forest\n",
    "\n",
    "rf_n_estimators = [int(x) for x in np.linspace(200, 1000, 5)]\n",
    "rf_n_estimators.append(1500)\n",
    "rf_n_estimators.append(2000)\n",
    "\n",
    "# Max number of levels in tree\n",
    "\n",
    "rf_max_depth = [int(x) for x in np.linspace(5, 55, 11)]\n",
    "\n",
    "# Adding default as possible value\n",
    "\n",
    "rf_max_depth.append(None)\n",
    "\n",
    "# Number of features to consider at every split\n",
    "\n",
    "rf_max_features = ['auto', 'sqrt', 'log2']\n",
    "\n",
    "# Criterion to split on\n",
    "\n",
    "rf_criterion = ['mse', 'mae']\n",
    "\n",
    "# Minimum number of samples required to split a node\n",
    "\n",
    "rf_min_samples_split = [int(x) for x in np.linspace(2, 10, 9)]\n",
    "\n",
    "# Minimum decrease in impurity required for split to happen\n",
    "\n",
    "rf_min_impurity_decrease = [0.0, 0.5, 0.1]\n",
    "\n",
    "# Method of selecting samples for training each tree\n",
    "\n",
    "rf_bootstrap = [True, False]\n",
    "\n",
    "# Creating the grid\n",
    "\n",
    "rf_grid = {'n_estimators': rf_n_estimators,\n",
    "           'max_depth': rf_max_depth,\n",
    "           'max_features': rf_max_features,\n",
    "           'criterion': rf_criterion,\n",
    "           'min_samples_split': rf_min_samples_split,\n",
    "           'min_impurity_decrease': rf_min_impurity_decrease,\n",
    "           'bootstrap': rf_bootstrap}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'n_estimators': [200, 400, 600, 800, 1000, 1500, 2000],\n",
       " 'max_depth': [5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 55, None],\n",
       " 'max_features': ['auto', 'sqrt', 'log2'],\n",
       " 'criterion': ['mse', 'mae'],\n",
       " 'min_samples_split': [2, 3, 4, 5, 6, 7, 8, 9, 10],\n",
       " 'min_impurity_decrease': [0.0, 0.5, 0.1],\n",
       " 'bootstrap': [True, False]}"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rf_grid"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 200 candidates, totalling 1000 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=-1)]: Done  33 tasks      | elapsed:   42.8s\n",
      "[Parallel(n_jobs=-1)]: Done 154 tasks      | elapsed:  7.4min\n",
      "[Parallel(n_jobs=-1)]: Done 357 tasks      | elapsed: 16.0min\n",
      "[Parallel(n_jobs=-1)]: Done 640 tasks      | elapsed: 26.2min\n",
      "[Parallel(n_jobs=-1)]: Done 1000 out of 1000 | elapsed: 41.5min finished\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "{'n_estimators': 800,\n",
       " 'min_samples_split': 3,\n",
       " 'min_impurity_decrease': 0.0,\n",
       " 'max_features': 'auto',\n",
       " 'max_depth': 20,\n",
       " 'criterion': 'mse',\n",
       " 'bootstrap': True}"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Hyperparameter Tuning for Random Forest using RandomizedSearchCV\n",
    "from sklearn.model_selection import RandomizedSearchCV\n",
    "\n",
    "# Creating model to be tuned\n",
    "\n",
    "rf_3 = RandomForestRegressor()\n",
    "\n",
    "# Creating randomized Random Forest\n",
    "\n",
    "rf_random = RandomizedSearchCV(estimator=rf_3, param_distributions=rf_grid, n_iter=200, cv=5, verbose=2, random_state=7, n_jobs=-1)\n",
    "\n",
    "# Fitting the random search model\n",
    "\n",
    "rf_random.fit(X_train, y_train)\n",
    "\n",
    "# View best parameters\n",
    "\n",
    "rf_random.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [],
   "source": [
    "import xgboost as xb"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create regressor\n",
    "xgb = xb.XGBRegressor(random_state=7)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Fit model to training data\n",
    "xbgr = xgb.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model Performance\n",
      "Mean Absolute Error: 0.5939.\n",
      "Mean Squared Error: 0.9341.\n",
      "R^2 Score: 0.5939.\n",
      "Accuracy: 99.21.\n"
     ]
    }
   ],
   "source": [
    "scoring(xgb, X_val, y_val)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'n_estimators': [200, 400, 600, 800, 1000, 1200, 1400, 1600, 1800, 2000],\n",
       " 'max_depth': [2, 4, 6, 8, 10, 12, 14, 16, 18, 20],\n",
       " 'min_child_weight': [1, 2, 3, 4, 5, 6, 7, 8, 9, 10],\n",
       " 'tree_method': ['auto', 'exact', 'approx', 'hist', 'gpu_hist'],\n",
       " 'eta': [0.1, 0.2, 0.30000000000000004, 0.4, 0.5, 0.6],\n",
       " 'gamma': [0, 0, 0, 0, 0, 0],\n",
       " 'objective': ['reg:squarederror', 'reg:squaredlogerror']}"
      ]
     },
     "execution_count": 89,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Number of trees to be used\n",
    "xgb_n_estimators = [int(x) for x in np.linspace(200, 2000, 10)]\n",
    "\n",
    "# Maximum number of levels in tree\n",
    "xgb_max_depth = [int(x) for x in np.linspace(2, 20, 10)]\n",
    "\n",
    "# Minimum number of instances needed in each node\n",
    "xgb_min_child_weight = [int(x) for x in np.linspace(1, 10, 10)]\n",
    "\n",
    "# Tree construction algorithm used in XGBoost\n",
    "xgb_tree_method = ['auto', 'exact', 'approx', 'hist', 'gpu_hist']\n",
    "\n",
    "# Learning rate\n",
    "xgb_eta = [x for x in np.linspace(0.1, 0.6, 6)]\n",
    "\n",
    "# Minimum loss reduction required to make further partition\n",
    "xgb_gamma = [int(x) for x in np.linspace(0, 0.5, 6)]\n",
    "\n",
    "# Learning objective used\n",
    "xgb_objective = ['reg:squarederror', 'reg:squaredlogerror']\n",
    "\n",
    "# Create the grid\n",
    "xgb_grid = {'n_estimators': xgb_n_estimators,\n",
    "            'max_depth': xgb_max_depth,\n",
    "            'min_child_weight': xgb_min_child_weight,\n",
    "            'tree_method': xgb_tree_method,\n",
    "            'eta': xgb_eta,\n",
    "            'gamma': xgb_gamma,\n",
    "            'objective': xgb_objective}\n",
    "\n",
    "xgb_grid"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 200 candidates, totalling 1000 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=-1)]: Done  33 tasks      | elapsed:    6.6s\n",
      "[Parallel(n_jobs=-1)]: Done 169 tasks      | elapsed:   32.7s\n",
      "[Parallel(n_jobs=-1)]: Done 399 tasks      | elapsed:  1.1min\n",
      "[Parallel(n_jobs=-1)]: Done 697 tasks      | elapsed:  1.8min\n",
      "[Parallel(n_jobs=-1)]: Done 1000 out of 1000 | elapsed:  2.6min finished\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "{'tree_method': 'approx',\n",
       " 'objective': 'reg:squarederror',\n",
       " 'n_estimators': 1200,\n",
       " 'min_child_weight': 8,\n",
       " 'max_depth': 12,\n",
       " 'gamma': 0,\n",
       " 'eta': 0.1}"
      ]
     },
     "execution_count": 90,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "xgb_base = xb.XGBRegressor()\n",
    "\n",
    "xgb_random = RandomizedSearchCV(estimator=xgb_base, param_distributions=xgb_grid,\n",
    "                                n_iter=200, cv=5, verbose=2, random_state=7, n_jobs=-1)\n",
    "\n",
    "xgb_random.fit(X_train, y_train)\n",
    "\n",
    "xgb_random.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "XGBRegressor(base_score=0.5, booster='gbtree', colsample_bylevel=1,\n",
       "             colsample_bynode=1, colsample_bytree=1, eta=0.1, gamma=0,\n",
       "             gpu_id=-1, importance_type='gain', interaction_constraints='',\n",
       "             learning_rate=0.100000001, max_delta_step=0, max_depth=12,\n",
       "             min_child_weight=8, missing=nan, monotone_constraints='()',\n",
       "             n_estimators=1200, n_jobs=0, num_parallel_tree=1, random_state=0,\n",
       "             reg_alpha=0, reg_lambda=1, scale_pos_weight=1, subsample=1,\n",
       "             tree_method='approx', validate_parameters=1, verbosity=None)"
      ]
     },
     "execution_count": 94,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Creating the final XGB\n",
    "\n",
    "xgb_final = xb.XGBRegressor(tree_method='approx',\n",
    "                            objective='reg:squarederror',\n",
    "                            n_estimators=1200,\n",
    "                            min_child_weight=8,\n",
    "                            max_depth=12,\n",
    "                            gamma=0,\n",
    "                            eta=0.1)\n",
    "\n",
    "xgb_final.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RandomForestRegressor(bootstrap=False, max_depth=25, max_features='log2',\n",
       "                      min_samples_split=4, n_estimators=200)"
      ]
     },
     "execution_count": 92,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Creating the final Random Forest\n",
    "rf_final = RandomForestRegressor(n_estimators=200,\n",
    "                                 min_samples_split=4,\n",
    "                                 min_impurity_decrease=0.0,\n",
    "                                 max_features='log2',\n",
    "                                 max_depth=25,\n",
    "                                 criterion='mse',\n",
    "                                 bootstrap=False)\n",
    "\n",
    "rf_final.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Defining function that compares all final models\n",
    "\n",
    "def final_comparison(models, test_features, test_labels):\n",
    "    scores = pd.DataFrame()\n",
    "    for model in models:\n",
    "        predictions = model.predict(test_features)\n",
    "        mae = round(mean_absolute_error(test_labels, predictions), 4)\n",
    "        mse = round(mean_squared_error(test_labels, predictions), 4)\n",
    "        r2 = round(r2_score(test_labels, predictions), 4)\n",
    "        errors = abs(predictions - test_labels)\n",
    "        mape = 100 * np.mean(errors / test_labels)\n",
    "        accuracy = round(100 - mape, 4)\n",
    "        scores[str(model)] = [mae, mse, r2, accuracy]\n",
    "    scores.index = ['Mean Absolute Error', 'Mean Squared Error', 'R^2 Error', 'Accuracy']\n",
    "    return scores"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {},
   "outputs": [],
   "source": [
    "final_scores =  final_comparison([rf_final, xgb_final], X_test, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [],
   "source": [
    "final_scores.columns = ['Random Forest', 'Extreme Gradient Boosting']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "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>Random Forest</th>\n",
       "      <th>Extreme Gradient Boosting</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Mean Absolute Error</th>\n",
       "      <td>0.7134</td>\n",
       "      <td>0.6467</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Mean Squared Error</th>\n",
       "      <td>1.4467</td>\n",
       "      <td>1.3379</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>R^2 Error</th>\n",
       "      <td>0.9877</td>\n",
       "      <td>0.9886</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Accuracy</th>\n",
       "      <td>99.0724</td>\n",
       "      <td>99.1441</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                     Random Forest  Extreme Gradient Boosting\n",
       "Mean Absolute Error         0.7134                     0.6467\n",
       "Mean Squared Error          1.4467                     1.3379\n",
       "R^2 Error                   0.9877                     0.9886\n",
       "Accuracy                   99.0724                    99.1441"
      ]
     },
     "execution_count": 97,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "final_scores"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "colab": {
   "collapsed_sections": [],
   "name": "Hackathon2020.ipynb",
   "provenance": []
  },
  "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": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}