U3L5.3 - Gradient boosting - intermediate challenge.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Guided example and challenge\n",
    "\n",
    "### Gradient boost guided example\n",
    "\n",
    "Having walked through gradient boost by hand, now let's try it with SKlearn.  We'll still use the European Social Survey Data, but now with a categorical outcome: Whether or not someone lives with a partner.\n",
    "\n",
    "### DRILL: Improve this gradient boost model\n",
    "\n",
    "While this model is already doing alright, we've seen from the Type I and Type II error rates that there is definitely room for improvement.  Your task is to see how low you can get the error rates to go in the test set, based on your model in the training set.  Strategies you might use include:\n",
    "\n",
    "* Creating new features\n",
    "* Applying more overfitting-prevention strategies like subsampling\n",
    "* More iterations\n",
    "* Trying a different loss function\n",
    "* Changing the structure of the weak learner: Allowing more leaves in the tree, or other modifications\n",
    "\n",
    "Have fun!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 292,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import scipy\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "\n",
    "from sklearn import ensemble\n",
    "from sklearn import datasets\n",
    "from sklearn.utils import shuffle\n",
    "from sklearn.metrics import mean_squared_error\n",
    "\n",
    "import seaborn as sns\n",
    "\n",
    "from sklearn.model_selection import GridSearchCV  #Perforing grid search\n",
    "from sklearn.model_selection import cross_val_score\n",
    "from sklearn import metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 169,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Working with the European Social Survey data again.\n",
    "df = pd.read_csv((\n",
    "    \"https://raw.githubusercontent.com/Thinkful-Ed/data-201-resources/\"\n",
    "    \"master/ESS_practice_data/ESSdata_Thinkful.csv\")).dropna()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 170,
   "metadata": {
    "scrolled": true
   },
   "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>cntry</th>\n",
       "      <th>idno</th>\n",
       "      <th>year</th>\n",
       "      <th>tvtot</th>\n",
       "      <th>ppltrst</th>\n",
       "      <th>pplfair</th>\n",
       "      <th>pplhlp</th>\n",
       "      <th>happy</th>\n",
       "      <th>sclmeet</th>\n",
       "      <th>sclact</th>\n",
       "      <th>gndr</th>\n",
       "      <th>agea</th>\n",
       "      <th>partner</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>CH</td>\n",
       "      <td>5.0</td>\n",
       "      <td>6</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>60.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>CH</td>\n",
       "      <td>25.0</td>\n",
       "      <td>6</td>\n",
       "      <td>6.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>59.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>CH</td>\n",
       "      <td>26.0</td>\n",
       "      <td>6</td>\n",
       "      <td>1.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>24.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>CH</td>\n",
       "      <td>28.0</td>\n",
       "      <td>6</td>\n",
       "      <td>4.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>64.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>CH</td>\n",
       "      <td>29.0</td>\n",
       "      <td>6</td>\n",
       "      <td>5.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>55.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>CH</td>\n",
       "      <td>40.0</td>\n",
       "      <td>6</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>76.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>CH</td>\n",
       "      <td>41.0</td>\n",
       "      <td>6</td>\n",
       "      <td>2.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>30.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>CH</td>\n",
       "      <td>51.0</td>\n",
       "      <td>6</td>\n",
       "      <td>2.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>84.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>CH</td>\n",
       "      <td>53.0</td>\n",
       "      <td>6</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>62.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>CH</td>\n",
       "      <td>55.0</td>\n",
       "      <td>6</td>\n",
       "      <td>1.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>33.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>CH</td>\n",
       "      <td>56.0</td>\n",
       "      <td>6</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>40.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>CH</td>\n",
       "      <td>57.0</td>\n",
       "      <td>6</td>\n",
       "      <td>4.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>69.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>CH</td>\n",
       "      <td>58.0</td>\n",
       "      <td>6</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>59.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>CH</td>\n",
       "      <td>68.0</td>\n",
       "      <td>6</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>32.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>CH</td>\n",
       "      <td>73.0</td>\n",
       "      <td>6</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>70.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>CH</td>\n",
       "      <td>81.0</td>\n",
       "      <td>6</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>61.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>CH</td>\n",
       "      <td>82.0</td>\n",
       "      <td>6</td>\n",
       "      <td>2.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>30.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>CH</td>\n",
       "      <td>84.0</td>\n",
       "      <td>6</td>\n",
       "      <td>0.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>21.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>CH</td>\n",
       "      <td>86.0</td>\n",
       "      <td>6</td>\n",
       "      <td>2.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>36.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>CH</td>\n",
       "      <td>87.0</td>\n",
       "      <td>6</td>\n",
       "      <td>7.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>51.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>CH</td>\n",
       "      <td>96.0</td>\n",
       "      <td>6</td>\n",
       "      <td>1.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>25.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>CH</td>\n",
       "      <td>98.0</td>\n",
       "      <td>6</td>\n",
       "      <td>2.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>62.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>CH</td>\n",
       "      <td>104.0</td>\n",
       "      <td>6</td>\n",
       "      <td>2.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>20.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>CH</td>\n",
       "      <td>108.0</td>\n",
       "      <td>6</td>\n",
       "      <td>0.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>22.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>CH</td>\n",
       "      <td>117.0</td>\n",
       "      <td>6</td>\n",
       "      <td>0.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>32.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>CH</td>\n",
       "      <td>118.0</td>\n",
       "      <td>6</td>\n",
       "      <td>2.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>35.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>CH</td>\n",
       "      <td>119.0</td>\n",
       "      <td>6</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>26.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>CH</td>\n",
       "      <td>121.0</td>\n",
       "      <td>6</td>\n",
       "      <td>2.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>35.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>CH</td>\n",
       "      <td>133.0</td>\n",
       "      <td>6</td>\n",
       "      <td>7.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>54.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>CH</td>\n",
       "      <td>135.0</td>\n",
       "      <td>6</td>\n",
       "      <td>6.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>38.0</td>\n",
       "      <td>2.0</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8564</th>\n",
       "      <td>SE</td>\n",
       "      <td>3659.0</td>\n",
       "      <td>7</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>17.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8565</th>\n",
       "      <td>SE</td>\n",
       "      <td>3661.0</td>\n",
       "      <td>7</td>\n",
       "      <td>5.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>17.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8566</th>\n",
       "      <td>SE</td>\n",
       "      <td>3662.0</td>\n",
       "      <td>7</td>\n",
       "      <td>2.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>17.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8567</th>\n",
       "      <td>SE</td>\n",
       "      <td>3665.0</td>\n",
       "      <td>7</td>\n",
       "      <td>2.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>17.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8568</th>\n",
       "      <td>SE</td>\n",
       "      <td>3666.0</td>\n",
       "      <td>7</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>17.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8569</th>\n",
       "      <td>SE</td>\n",
       "      <td>3668.0</td>\n",
       "      <td>7</td>\n",
       "      <td>6.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>17.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8570</th>\n",
       "      <td>SE</td>\n",
       "      <td>3670.0</td>\n",
       "      <td>7</td>\n",
       "      <td>1.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8571</th>\n",
       "      <td>SE</td>\n",
       "      <td>3671.0</td>\n",
       "      <td>7</td>\n",
       "      <td>3.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8572</th>\n",
       "      <td>SE</td>\n",
       "      <td>3682.0</td>\n",
       "      <td>7</td>\n",
       "      <td>5.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8573</th>\n",
       "      <td>SE</td>\n",
       "      <td>3684.0</td>\n",
       "      <td>7</td>\n",
       "      <td>1.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8574</th>\n",
       "      <td>SE</td>\n",
       "      <td>3687.0</td>\n",
       "      <td>7</td>\n",
       "      <td>1.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8575</th>\n",
       "      <td>SE</td>\n",
       "      <td>3689.0</td>\n",
       "      <td>7</td>\n",
       "      <td>3.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8576</th>\n",
       "      <td>SE</td>\n",
       "      <td>3690.0</td>\n",
       "      <td>7</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8577</th>\n",
       "      <td>SE</td>\n",
       "      <td>3692.0</td>\n",
       "      <td>7</td>\n",
       "      <td>6.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8578</th>\n",
       "      <td>SE</td>\n",
       "      <td>3696.0</td>\n",
       "      <td>7</td>\n",
       "      <td>1.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8579</th>\n",
       "      <td>SE</td>\n",
       "      <td>3700.0</td>\n",
       "      <td>7</td>\n",
       "      <td>2.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8580</th>\n",
       "      <td>SE</td>\n",
       "      <td>3702.0</td>\n",
       "      <td>7</td>\n",
       "      <td>4.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8581</th>\n",
       "      <td>SE</td>\n",
       "      <td>3703.0</td>\n",
       "      <td>7</td>\n",
       "      <td>1.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8582</th>\n",
       "      <td>SE</td>\n",
       "      <td>3707.0</td>\n",
       "      <td>7</td>\n",
       "      <td>6.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8583</th>\n",
       "      <td>SE</td>\n",
       "      <td>3708.0</td>\n",
       "      <td>7</td>\n",
       "      <td>2.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8584</th>\n",
       "      <td>SE</td>\n",
       "      <td>3709.0</td>\n",
       "      <td>7</td>\n",
       "      <td>2.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8585</th>\n",
       "      <td>SE</td>\n",
       "      <td>3712.0</td>\n",
       "      <td>7</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8586</th>\n",
       "      <td>SE</td>\n",
       "      <td>3716.0</td>\n",
       "      <td>7</td>\n",
       "      <td>2.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8587</th>\n",
       "      <td>SE</td>\n",
       "      <td>3720.0</td>\n",
       "      <td>7</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8588</th>\n",
       "      <td>SE</td>\n",
       "      <td>3727.0</td>\n",
       "      <td>7</td>\n",
       "      <td>1.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8589</th>\n",
       "      <td>SE</td>\n",
       "      <td>3729.0</td>\n",
       "      <td>7</td>\n",
       "      <td>3.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>18.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8590</th>\n",
       "      <td>SE</td>\n",
       "      <td>3732.0</td>\n",
       "      <td>7</td>\n",
       "      <td>5.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8591</th>\n",
       "      <td>SE</td>\n",
       "      <td>3743.0</td>\n",
       "      <td>7</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>44.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8592</th>\n",
       "      <td>SE</td>\n",
       "      <td>3744.0</td>\n",
       "      <td>7</td>\n",
       "      <td>5.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8593</th>\n",
       "      <td>SE</td>\n",
       "      <td>3746.0</td>\n",
       "      <td>7</td>\n",
       "      <td>2.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>8147 rows × 13 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     cntry    idno  year  tvtot  ppltrst  pplfair  pplhlp  happy  sclmeet  \\\n",
       "0       CH     5.0     6    3.0      3.0     10.0     5.0    8.0      5.0   \n",
       "1       CH    25.0     6    6.0      5.0      7.0     5.0    9.0      3.0   \n",
       "2       CH    26.0     6    1.0      8.0      8.0     8.0    7.0      6.0   \n",
       "3       CH    28.0     6    4.0      6.0      6.0     7.0   10.0      6.0   \n",
       "4       CH    29.0     6    5.0      6.0      7.0     5.0    8.0      7.0   \n",
       "6       CH    40.0     6    3.0      0.0      5.0     2.0    0.0      2.0   \n",
       "7       CH    41.0     6    2.0      4.0      5.0     3.0   10.0      5.0   \n",
       "8       CH    51.0     6    2.0      8.0      8.0     8.0    9.0      6.0   \n",
       "9       CH    53.0     6    4.0      4.0      4.0     8.0    7.0      4.0   \n",
       "10      CH    55.0     6    1.0      6.0      7.0     7.0    9.0      5.0   \n",
       "11      CH    56.0     6    4.0      5.0      7.0     7.0    7.0      5.0   \n",
       "12      CH    57.0     6    4.0      7.0      7.0     4.0    9.0      6.0   \n",
       "13      CH    58.0     6    2.0      1.0      9.0     7.0    8.0      4.0   \n",
       "14      CH    68.0     6    4.0      4.0      5.0     3.0    8.0      1.0   \n",
       "15      CH    73.0     6    5.0      4.0      5.0     5.0    9.0      3.0   \n",
       "16      CH    81.0     6    4.0      5.0      7.0     7.0    8.0      3.0   \n",
       "17      CH    82.0     6    2.0      7.0      7.0     7.0    6.0      6.0   \n",
       "18      CH    84.0     6    0.0      9.0      7.0     8.0    9.0      5.0   \n",
       "19      CH    86.0     6    2.0      6.0      6.0     6.0    8.0      4.0   \n",
       "20      CH    87.0     6    7.0      3.0      4.0     4.0    7.0      6.0   \n",
       "21      CH    96.0     6    1.0      8.0      8.0     6.0    8.0      7.0   \n",
       "22      CH    98.0     6    2.0      7.0      7.0     8.0    9.0      6.0   \n",
       "23      CH   104.0     6    2.0      4.0      8.0     8.0    9.0      6.0   \n",
       "24      CH   108.0     6    0.0      7.0      7.0     4.0    9.0      4.0   \n",
       "25      CH   117.0     6    0.0      6.0      7.0     5.0    7.0      6.0   \n",
       "26      CH   118.0     6    2.0      7.0      6.0     5.0    8.0      4.0   \n",
       "27      CH   119.0     6    0.0      5.0      8.0     6.0    9.0      4.0   \n",
       "28      CH   121.0     6    2.0      8.0      8.0     8.0   10.0      6.0   \n",
       "29      CH   133.0     6    7.0      8.0      5.0     5.0    8.0      5.0   \n",
       "30      CH   135.0     6    6.0      8.0      8.0     3.0    5.0      6.0   \n",
       "...    ...     ...   ...    ...      ...      ...     ...    ...      ...   \n",
       "8564    SE  3659.0     7    2.0      1.0      5.0     4.0   10.0      7.0   \n",
       "8565    SE  3661.0     7    5.0      5.0      3.0     4.0    5.0      6.0   \n",
       "8566    SE  3662.0     7    2.0      7.0      8.0     6.0    8.0      7.0   \n",
       "8567    SE  3665.0     7    2.0      4.0      8.0     7.0    9.0      7.0   \n",
       "8568    SE  3666.0     7    1.0      2.0      8.0     6.0    9.0      7.0   \n",
       "8569    SE  3668.0     7    6.0      5.0      3.0     4.0    9.0      6.0   \n",
       "8570    SE  3670.0     7    1.0      5.0      7.0     7.0    8.0      7.0   \n",
       "8571    SE  3671.0     7    3.0      4.0      9.0     7.0    8.0      6.0   \n",
       "8572    SE  3682.0     7    5.0      7.0      8.0     5.0    8.0      3.0   \n",
       "8573    SE  3684.0     7    1.0      8.0      8.0     6.0    9.0      7.0   \n",
       "8574    SE  3687.0     7    1.0      4.0      5.0     6.0    7.0      7.0   \n",
       "8575    SE  3689.0     7    3.0      8.0      7.0     7.0    8.0      7.0   \n",
       "8576    SE  3690.0     7    4.0      5.0      4.0     4.0    7.0      6.0   \n",
       "8577    SE  3692.0     7    6.0      6.0      9.0     7.0    5.0      4.0   \n",
       "8578    SE  3696.0     7    1.0      5.0      5.0     5.0    7.0      5.0   \n",
       "8579    SE  3700.0     7    2.0      8.0      9.0     7.0    9.0      7.0   \n",
       "8580    SE  3702.0     7    4.0      7.0      8.0     5.0   10.0      4.0   \n",
       "8581    SE  3703.0     7    1.0      6.0      7.0     6.0    8.0      6.0   \n",
       "8582    SE  3707.0     7    6.0      7.0      7.0     6.0    8.0      7.0   \n",
       "8583    SE  3708.0     7    2.0      6.0      8.0     5.0   10.0      6.0   \n",
       "8584    SE  3709.0     7    2.0      8.0     10.0     6.0    9.0      6.0   \n",
       "8585    SE  3712.0     7    1.0      3.0      6.0     4.0    9.0      7.0   \n",
       "8586    SE  3716.0     7    2.0      4.0      6.0     3.0    7.0      7.0   \n",
       "8587    SE  3720.0     7    4.0      4.0      6.0     7.0    9.0      7.0   \n",
       "8588    SE  3727.0     7    1.0      6.0      5.0     5.0   10.0      7.0   \n",
       "8589    SE  3729.0     7    3.0      4.0      5.0     3.0    6.0      6.0   \n",
       "8590    SE  3732.0     7    5.0      6.0      4.0     4.0   10.0      6.0   \n",
       "8591    SE  3743.0     7    4.0      5.0      7.0     6.0    8.0      6.0   \n",
       "8592    SE  3744.0     7    5.0      8.0      8.0     6.0    9.0      7.0   \n",
       "8593    SE  3746.0     7    2.0      6.0      7.0     5.0    7.0      7.0   \n",
       "\n",
       "      sclact  gndr  agea  partner  \n",
       "0        4.0   2.0  60.0      1.0  \n",
       "1        2.0   2.0  59.0      1.0  \n",
       "2        3.0   1.0  24.0      2.0  \n",
       "3        2.0   2.0  64.0      1.0  \n",
       "4        2.0   2.0  55.0      1.0  \n",
       "6        2.0   1.0  76.0      1.0  \n",
       "7        2.0   2.0  30.0      1.0  \n",
       "8        4.0   2.0  84.0      2.0  \n",
       "9        2.0   2.0  62.0      1.0  \n",
       "10       2.0   2.0  33.0      1.0  \n",
       "11       3.0   1.0  40.0      1.0  \n",
       "12       2.0   1.0  69.0      1.0  \n",
       "13       3.0   1.0  59.0      2.0  \n",
       "14       2.0   2.0  32.0      1.0  \n",
       "15       3.0   2.0  70.0      1.0  \n",
       "16       3.0   1.0  61.0      1.0  \n",
       "17       3.0   1.0  30.0      1.0  \n",
       "18       4.0   1.0  21.0      2.0  \n",
       "19       3.0   2.0  36.0      1.0  \n",
       "20       4.0   1.0  51.0      1.0  \n",
       "21       4.0   1.0  25.0      2.0  \n",
       "22       3.0   1.0  62.0      1.0  \n",
       "23       2.0   2.0  20.0      2.0  \n",
       "24       2.0   2.0  22.0      1.0  \n",
       "25       2.0   2.0  32.0      1.0  \n",
       "26       3.0   2.0  35.0      1.0  \n",
       "27       2.0   2.0  26.0      1.0  \n",
       "28       3.0   2.0  35.0      1.0  \n",
       "29       3.0   2.0  54.0      1.0  \n",
       "30       1.0   2.0  38.0      2.0  \n",
       "...      ...   ...   ...      ...  \n",
       "8564     4.0   1.0  17.0      2.0  \n",
       "8565     1.0   2.0  17.0      2.0  \n",
       "8566     3.0   2.0  17.0      2.0  \n",
       "8567     4.0   2.0  17.0      2.0  \n",
       "8568     3.0   2.0  17.0      2.0  \n",
       "8569     4.0   1.0  17.0      2.0  \n",
       "8570     4.0   1.0  16.0      2.0  \n",
       "8571     3.0   1.0  16.0      2.0  \n",
       "8572     1.0   1.0  16.0      2.0  \n",
       "8573     3.0   2.0  16.0      2.0  \n",
       "8574     3.0   2.0  16.0      2.0  \n",
       "8575     3.0   1.0  16.0      2.0  \n",
       "8576     3.0   1.0  16.0      2.0  \n",
       "8577     2.0   2.0  16.0      2.0  \n",
       "8578     4.0   2.0  16.0      2.0  \n",
       "8579     3.0   2.0  16.0      2.0  \n",
       "8580     1.0   1.0  16.0      2.0  \n",
       "8581     3.0   1.0  16.0      2.0  \n",
       "8582     3.0   1.0  16.0      2.0  \n",
       "8583     5.0   1.0  16.0      2.0  \n",
       "8584     2.0   1.0  16.0      2.0  \n",
       "8585     3.0   1.0  16.0      2.0  \n",
       "8586     3.0   2.0  15.0      2.0  \n",
       "8587     3.0   1.0  15.0      2.0  \n",
       "8588     2.0   1.0  15.0      2.0  \n",
       "8589     2.0   1.0  18.0      2.0  \n",
       "8590     3.0   1.0  15.0      2.0  \n",
       "8591     3.0   1.0  44.0      2.0  \n",
       "8592     3.0   1.0  15.0      2.0  \n",
       "8593     4.0   2.0  15.0      2.0  \n",
       "\n",
       "[8147 rows x 13 columns]"
      ]
     },
     "execution_count": 170,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- idno >>  Respondent identification number\n",
    "- year\t>> 7 = 2014 / 6 = 2012\n",
    "- tvtot\t    >> TV WATCHING, TOTAL TIME ON AVERAGE WEEKDAY\n",
    "- ppltrst\t>> \"MOST PEOPLE CAN BE TRUSTED OR YOU CAN’T BE TOO CAREFUL \"\n",
    "- pplfair\t>> \"MOST PEOPLE TRY TO TAKE ADVANTAGE OF YOU, OR TRY TO BE FAIR\"\n",
    "- pplhlp\t>> MOST OF THE TIME PEOPLE ARE HELPFUL OR MOSTLY ARE LOOKING OUT FOR THEMSELVES \n",
    "- happy\t    >> HOW HAPPY ARE YOU ?\n",
    "- sclmeet\t>> HOW OFTEN SOCIALLY MEET WITH FRIENDS, RELATIVES OR COLLEAGUES\n",
    "- sclact\t>> TAKE PART IN SOCIAL ACTIVITIES COMPARED TO OTHERS OF SAME AGE \n",
    "- gndr\t    >> GENDER\n",
    "- agea\t    >> Estimation of age of respondent or household member who refuses, by interviewer\n",
    "- partner\t>> LIVES WITH HUSBAND/WIFE/PARTNER: 1=with partner / 2 = alone\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 217,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "# Definine outcome and predictors.\n",
    "# Set our outcome to 0=alone and 1=w/ partner.\n",
    "y = np.where(df['partner']==1,1,0)\n",
    "X = df.loc[:, ~df.columns.isin(['partner', 'cntry', 'idno'])]\n",
    "\n",
    "# Make the categorical variable 'country' into dummies.\n",
    "X = pd.concat([X, pd.get_dummies(df['cntry'])], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 215,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Shape:  (8147, 16)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead 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>year</th>\n",
       "      <th>tvtot</th>\n",
       "      <th>ppltrst</th>\n",
       "      <th>pplfair</th>\n",
       "      <th>pplhlp</th>\n",
       "      <th>happy</th>\n",
       "      <th>sclmeet</th>\n",
       "      <th>sclact</th>\n",
       "      <th>gndr</th>\n",
       "      <th>agea</th>\n",
       "      <th>CH</th>\n",
       "      <th>CZ</th>\n",
       "      <th>DE</th>\n",
       "      <th>ES</th>\n",
       "      <th>NO</th>\n",
       "      <th>SE</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>6</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>60.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>6</td>\n",
       "      <td>6.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>59.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>6</td>\n",
       "      <td>1.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>24.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>6</td>\n",
       "      <td>4.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>64.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>6</td>\n",
       "      <td>5.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>55.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   year  tvtot  ppltrst  pplfair  pplhlp  happy  sclmeet  sclact  gndr  agea  \\\n",
       "0     6    3.0      3.0     10.0     5.0    8.0      5.0     4.0   2.0  60.0   \n",
       "1     6    6.0      5.0      7.0     5.0    9.0      3.0     2.0   2.0  59.0   \n",
       "2     6    1.0      8.0      8.0     8.0    7.0      6.0     3.0   1.0  24.0   \n",
       "3     6    4.0      6.0      6.0     7.0   10.0      6.0     2.0   2.0  64.0   \n",
       "4     6    5.0      6.0      7.0     5.0    8.0      7.0     2.0   2.0  55.0   \n",
       "\n",
       "   CH  CZ  DE  ES  NO  SE  \n",
       "0   1   0   0   0   0   0  \n",
       "1   1   0   0   0   0   0  \n",
       "2   1   0   0   0   0   0  \n",
       "3   1   0   0   0   0   0  \n",
       "4   1   0   0   0   0   0  "
      ]
     },
     "execution_count": 215,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print('Shape: ',X.shape)\n",
    "X.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 218,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([1, 1, 0, ..., 0, 0, 0])"
      ]
     },
     "execution_count": 218,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 219,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create training and test sets.\n",
    "offset = int(X.shape[0] * 0.9)\n",
    "\n",
    "# Put 90% of the data in the training set.\n",
    "X_train, y_train = X[:offset], y[:offset]\n",
    "\n",
    "# And put 10% in the test set.\n",
    "X_test, y_test = X[offset:], y[offset:]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Since we're now working with a binary outcome, we've switched to a classifier. \n",
    "\n",
    "Now our loss function can't be the residuals. \n",
    "\n",
    "Our options are \"deviance\", or \"exponential\". \n",
    "\n",
    "Deviance is used for logistic regression, and we'll try that here."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 220,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "GradientBoostingClassifier(criterion='friedman_mse', init=None,\n",
       "              learning_rate=0.1, loss='deviance', max_depth=2,\n",
       "              max_features=None, max_leaf_nodes=None,\n",
       "              min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "              min_samples_leaf=1, min_samples_split=2,\n",
       "              min_weight_fraction_leaf=0.0, n_estimators=500,\n",
       "              presort='auto', random_state=None, subsample=1.0, verbose=0,\n",
       "              warm_start=False)"
      ]
     },
     "execution_count": 220,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# We'll make 500 iterations, use 2-deep trees, and set our loss function.\n",
    "params = {'n_estimators': 500,\n",
    "          'max_depth': 2,\n",
    "          'loss': 'deviance'\n",
    "         }\n",
    "\n",
    "# Initialize and fit the model.\n",
    "clf = ensemble.GradientBoostingClassifier(**params)\n",
    "clf.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 221,
   "metadata": {},
   "outputs": [],
   "source": [
    "predict_train = clf.predict(X_train)\n",
    "predict_test = clf.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 222,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([1, 1, 0, ..., 1, 0, 1])"
      ]
     },
     "execution_count": 222,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "predict_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 223,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7521472392638037"
      ]
     },
     "execution_count": 223,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.metrics import accuracy_score\n",
    "score_gdb = accuracy_score(y_test, predict_test)\n",
    "score_gdb"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 224,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Accuracy tables.\n",
    "table_train = pd.crosstab(y_train, predict_train, margins=True)\n",
    "table_test = pd.crosstab(y_test, predict_test, margins=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 225,
   "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>col_0</th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "      <th>All</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>row_0</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1533</td>\n",
       "      <td>1291</td>\n",
       "      <td>2824</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>341</td>\n",
       "      <td>4167</td>\n",
       "      <td>4508</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>All</th>\n",
       "      <td>1874</td>\n",
       "      <td>5458</td>\n",
       "      <td>7332</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "col_0     0     1   All\n",
       "row_0                  \n",
       "0      1533  1291  2824\n",
       "1       341  4167  4508\n",
       "All    1874  5458  7332"
      ]
     },
     "execution_count": 225,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "table_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 228,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_tI_errors = table_train.loc[0.0,1.0] / table_train.loc['All','All']\n",
    "train_tII_errors = table_train.loc[1.0,0.0] / table_train.loc['All','All']\n",
    "\n",
    "test_tI_errors = table_test.loc[0.0,1.0]/table_test.loc['All','All']\n",
    "test_tII_errors = table_test.loc[1.0,0.0]/table_test.loc['All','All']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 230,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training set accuracy:\n",
      "Percent Type I errors: 0.17607746863066012\n",
      "Percent Type II errors: 0.04650845608292417\n",
      "\n",
      "Test set accuracy:\n",
      "Percent Type I errors: 0.18527607361963191\n",
      "Percent Type II errors: 0.06257668711656442\n"
     ]
    }
   ],
   "source": [
    "print((\n",
    "    'Training set accuracy:\\n'\n",
    "    'Percent Type I errors: {}\\n'\n",
    "    'Percent Type II errors: {}\\n\\n'\n",
    "    'Test set accuracy:\\n'\n",
    "    'Percent Type I errors: {}\\n'\n",
    "    'Percent Type II errors: {}'\n",
    ").format(train_tI_errors, train_tII_errors, test_tI_errors, test_tII_errors))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Unlike decision trees, gradient boost solutions are not terribly easy to interpret on the surface. \n",
    "\n",
    "But they aren't quite a black box. \n",
    "\n",
    "We can get a measure of how important various features are by counting how many times a feature is used over the course of many decision trees."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 231,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.01497972, 0.08405402, 0.08240387, 0.07755603, 0.05077273,\n",
       "       0.15576888, 0.04939301, 0.03993169, 0.05960565, 0.27496302,\n",
       "       0.01505447, 0.01558719, 0.00600491, 0.03773901, 0.02324548,\n",
       "       0.01294032])"
      ]
     },
     "execution_count": 231,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "feature_importance = clf.feature_importances_\n",
    "feature_importance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 232,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([  5.44790521,  30.56920837,  29.96907409,  28.20598662,\n",
       "        18.46529383,  56.65084844,  17.96351113,  14.52256739,\n",
       "        21.67769557, 100.        ,   5.47508803,   5.66883278,\n",
       "         2.18389639,  13.72512388,   8.45403865,   4.70620465])"
      ]
     },
     "execution_count": 232,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "100.0 * (feature_importance / feature_importance.max())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 233,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([12, 15,  0, 10, 11, 14, 13,  7,  6,  4,  8,  3,  2,  1,  5,  9],\n",
       "      dtype=int64)"
      ]
     },
     "execution_count": 233,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.argsort(feature_importance)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 234,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x24cf4918e80>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sorted_idx = np.argsort(feature_importance)\n",
    "pos = np.arange(sorted_idx.shape[0]) + .5\n",
    "plt.subplot(1, 2, 2)\n",
    "plt.barh(pos, feature_importance[sorted_idx], align='center')\n",
    "plt.yticks(pos, X.columns[sorted_idx])\n",
    "plt.xlabel('Relative Importance')\n",
    "plt.title('Variable Importance')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "--------------"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### DRILL: Improve this gradient boost model\n",
    "\n",
    "While this model is already doing alright, we've seen from the Type I and Type II error rates that there is definitely room for improvement.  Your task is to see how low you can get the error rates to go in the test set, based on your model in the training set.  Strategies you might use include:\n",
    "\n",
    "* Creating new features\n",
    "* Applying more overfitting-prevention strategies like subsampling\n",
    "* More iterations\n",
    "* Trying a different loss function\n",
    "* Changing the structure of the weak learner: Allowing more leaves in the tree, or other modifications\n",
    "\n",
    "Have fun!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 235,
   "metadata": {
    "scrolled": true
   },
   "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>year</th>\n",
       "      <th>tvtot</th>\n",
       "      <th>ppltrst</th>\n",
       "      <th>pplfair</th>\n",
       "      <th>pplhlp</th>\n",
       "      <th>happy</th>\n",
       "      <th>sclmeet</th>\n",
       "      <th>sclact</th>\n",
       "      <th>gndr</th>\n",
       "      <th>agea</th>\n",
       "      <th>CH</th>\n",
       "      <th>CZ</th>\n",
       "      <th>DE</th>\n",
       "      <th>ES</th>\n",
       "      <th>NO</th>\n",
       "      <th>SE</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>6</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>60.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>6</td>\n",
       "      <td>6.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>59.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>6</td>\n",
       "      <td>1.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>24.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>6</td>\n",
       "      <td>4.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>64.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>6</td>\n",
       "      <td>5.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>55.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   year  tvtot  ppltrst  pplfair  pplhlp  happy  sclmeet  sclact  gndr  agea  \\\n",
       "0     6    3.0      3.0     10.0     5.0    8.0      5.0     4.0   2.0  60.0   \n",
       "1     6    6.0      5.0      7.0     5.0    9.0      3.0     2.0   2.0  59.0   \n",
       "2     6    1.0      8.0      8.0     8.0    7.0      6.0     3.0   1.0  24.0   \n",
       "3     6    4.0      6.0      6.0     7.0   10.0      6.0     2.0   2.0  64.0   \n",
       "4     6    5.0      6.0      7.0     5.0    8.0      7.0     2.0   2.0  55.0   \n",
       "\n",
       "   CH  CZ  DE  ES  NO  SE  \n",
       "0   1   0   0   0   0   0  \n",
       "1   1   0   0   0   0   0  \n",
       "2   1   0   0   0   0   0  \n",
       "3   1   0   0   0   0   0  \n",
       "4   1   0   0   0   0   0  "
      ]
     },
     "execution_count": 235,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- idno >>  Respondent identification number\n",
    "- year\t>> 7 = 2014 / 6 = 2012\n",
    "- tvtot\t    >> TV WATCHING, TOTAL TIME ON AVERAGE WEEKDAY\n",
    "- ppltrst\t>> \"MOST PEOPLE CAN BE TRUSTED OR YOU CAN’T BE TOO CAREFUL \"\n",
    "- pplfair\t>> \"MOST PEOPLE TRY TO TAKE ADVANTAGE OF YOU, OR TRY TO BE FAIR\"\n",
    "- pplhlp\t>> MOST OF THE TIME PEOPLE ARE HELPFUL OR MOSTLY ARE LOOKING OUT FOR THEMSELVES \n",
    "- happy\t    >> HOW HAPPY ARE YOU ?\n",
    "- sclmeet\t>> HOW OFTEN SOCIALLY MEET WITH FRIENDS, RELATIVES OR COLLEAGUES\n",
    "- sclact\t>> TAKE PART IN SOCIAL ACTIVITIES COMPARED TO OTHERS OF SAME AGE \n",
    "- gndr\t    >> GENDER\n",
    "- agea\t    >> Estimation of age of respondent or household member who refuses, by interviewer\n",
    "- partner\t>> LIVES WITH HUSBAND/WIFE/PARTNER\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "multiple interraction: ppltrst*happy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 236,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([6, 7], dtype=int64)"
      ]
     },
     "execution_count": 236,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X.year.unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 237,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['CH', 'CZ', 'DE', 'ES', 'NO', 'SE'], dtype=object)"
      ]
     },
     "execution_count": 237,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.cntry.unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 238,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create a new column 'Nordic' from NO and SE countries as a dummy.\n",
    "X['Nordic'] = np.where(( (X['NO']==1) | (X['SE']==1) ) , 1, 0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 239,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "cntry     CH    CZ  DE    ES    NO    SE   All\n",
      "Nordic                                        \n",
      "0       1475  1207  27  2292     0     0  5001\n",
      "1          0     0   0     0  1420  1726  3146\n",
      "All     1475  1207  27  2292  1420  1726  8147\n"
     ]
    }
   ],
   "source": [
    "print(pd.crosstab(X['Nordic'], df['cntry'] , margins = True))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 240,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 60.,  59.,  24.,  64.,  55.,  76.,  30.,  84.,  62.,  33.,  40.,\n",
       "        69.,  32.,  70.,  61.,  21.,  36.,  51.,  25.,  20.,  22.,  35.,\n",
       "        26.,  54.,  38.,  82.,  53.,  46.,  47.,  72.,  16.,  42.,  57.,\n",
       "        45.,  28.,  85.,  48.,  52.,  29.,  27.,  79.,  44.,  92.,  50.,\n",
       "        37.,  75.,  19.,  31.,  65.,  68.,  58.,  39.,  67.,  23.,  56.,\n",
       "        78.,  83.,  74.,  49.,  43.,  86.,  91.,  34.,  17.,  41.,  63.,\n",
       "        73.,  80.,  81.,  71.,  89.,  18.,  94.,  87.,  90.,  15., 103.,\n",
       "        95.,  97.,  93., 101.,  96., 114.])"
      ]
     },
     "execution_count": 240,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X.agea.unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 241,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x24cf4903908>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Analyze the AGE column:\n",
    "\n",
    "plt.hist(X.agea)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 242,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    8147.000000\n",
       "mean       47.262182\n",
       "std        18.316890\n",
       "min        15.000000\n",
       "25%        33.000000\n",
       "50%        47.000000\n",
       "75%        61.000000\n",
       "max       114.000000\n",
       "Name: agea, dtype: float64"
      ]
     },
     "execution_count": 242,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X.agea.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 243,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x24cf34c1550>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.boxplot(X.agea)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 244,
   "metadata": {},
   "outputs": [],
   "source": [
    "X['Adult_61plus'] = np.where(X['agea']>=61 , 1 , 0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 245,
   "metadata": {
    "scrolled": true
   },
   "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>year</th>\n",
       "      <th>tvtot</th>\n",
       "      <th>ppltrst</th>\n",
       "      <th>pplfair</th>\n",
       "      <th>pplhlp</th>\n",
       "      <th>happy</th>\n",
       "      <th>sclmeet</th>\n",
       "      <th>sclact</th>\n",
       "      <th>gndr</th>\n",
       "      <th>agea</th>\n",
       "      <th>CH</th>\n",
       "      <th>CZ</th>\n",
       "      <th>DE</th>\n",
       "      <th>ES</th>\n",
       "      <th>NO</th>\n",
       "      <th>SE</th>\n",
       "      <th>Nordic</th>\n",
       "      <th>Adult_61plus</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>6</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>60.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>6</td>\n",
       "      <td>6.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>59.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>6</td>\n",
       "      <td>1.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>24.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>6</td>\n",
       "      <td>4.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>64.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>6</td>\n",
       "      <td>5.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>55.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   year  tvtot  ppltrst  pplfair  pplhlp  happy  sclmeet  sclact  gndr  agea  \\\n",
       "0     6    3.0      3.0     10.0     5.0    8.0      5.0     4.0   2.0  60.0   \n",
       "1     6    6.0      5.0      7.0     5.0    9.0      3.0     2.0   2.0  59.0   \n",
       "2     6    1.0      8.0      8.0     8.0    7.0      6.0     3.0   1.0  24.0   \n",
       "3     6    4.0      6.0      6.0     7.0   10.0      6.0     2.0   2.0  64.0   \n",
       "4     6    5.0      6.0      7.0     5.0    8.0      7.0     2.0   2.0  55.0   \n",
       "\n",
       "   CH  CZ  DE  ES  NO  SE  Nordic  Adult_61plus  \n",
       "0   1   0   0   0   0   0       0             0  \n",
       "1   1   0   0   0   0   0       0             0  \n",
       "2   1   0   0   0   0   0       0             0  \n",
       "3   1   0   0   0   0   0       0             1  \n",
       "4   1   0   0   0   0   0       0             0  "
      ]
     },
     "execution_count": 245,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Which are the correlated columns? \n",
    "I'll try to combine some if possible."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 246,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x24cf33f3a20>"
      ]
     },
     "execution_count": 246,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x24cf33ee0f0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import seaborn as sns\n",
    "corrmat = X.corr()\n",
    "\n",
    "# Set up the matplotlib figure\n",
    "f,ax = plt.subplots(figsize=(12,9))\n",
    "\n",
    "# Draw the heatmap using seaborn\n",
    "sns.heatmap(corrmat , vmax=.8 , square = True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We may think sad people watch more TV, then recoding our \"happy\" variable will make the correlation between `'tvtot'` and `'happy'` a little simpler to parse."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 247,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x24cf48f26d8>"
      ]
     },
     "execution_count": 247,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAW4AAAD8CAYAAABXe05zAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvNQv5yAAAEwZJREFUeJzt3X9sXWd9x/HPZ04q3ADyqjodcdMZuiiAiEbQFU2XCWUUL4VWEFWbRNROaJqafxiDBQUltBKaRBekTBH8MSGlLT+mRkGsBA+RjhABEaNqA9cNYCCN0rDSxCm1UeRRLLMm5rs/7HiJr+N7Ljk/7nPv+yVV8X18cj/PIzufnpxznMcRIQBAOv6g6gkAAFpDcQNAYihuAEgMxQ0AiaG4ASAxFDcAJIbiBoDEUNwAkBiKGwASs6yIN73xxhtjcHCwiLcGgI40MjLyq4joz3JsIcU9ODioer1exFsDQEey/Yusx3KpBAASQ3EDQGIobgBIDMUNAImhuAEgMU2L2/Za2z+87L9f2/5IGZMDADRq+jhgRJyU9FZJst0jaUzSV/OeyBsfeEK/nfn/3Xhe1WM9+9B78o5pm1xJGtx5qGHs+U/d1bG5VWZ3W26V2ay5+NxWL5XcIel0RGR+3jCLheUpSb+dCb3xgSfyjGmbXGnxL/RS46nnVpndbblVZrPmcnJbLe73SzqQ9yQWlmez8dRzAeBaZC5u29dJeq+kf7/K57fZrtuuT0xM5DU/AMACrZxxv1vSMxHx0mKfjIh9EVGLiFp/f6YftwcA/B5aKe6tKuAyiTR7Q7CV8dRzAeBaZCpu29dLGpJ0sIhJPPvQexrKsoynO6rKla5+x7noO+BV5VaZ3W25VWaz5nJyHZH/jbharRb864AAkJ3tkYioZTmWn5wEgMRQ3ACQGIobABJDcQNAYihuAEgMxQ0AiaG4ASAxFDcAJIbiBoDEUNwAkBiKGwASQ3EDQGIobgBIDMUNAImhuAEgMRQ3ACSG4gaAxFDcAJAYihsAEpN1s+A+24/bftb2Cdu3Fz0xAMDilmU87jOSvhERf2X7OknX5z2R4eNj2nP4pM5NTmtVX692bF6rLesH8o5pm1xJGtx5qGGsjN2wq8qtMrvbcqvMZs3F5zY947b9WknvkPSoJEXEKxExmeckho+PadfBUY1NTiskjU1Oa9fBUQ0fH8szpm1ypcW/0EuNp55bZXa35VaZzZrLyc1yqeQNkiYkfd72cduP2F6R5yT2HD6p6QszV4xNX5jRnsMn84xpm1wAuBZZinuZpLdJ+mxErJc0JWnnwoNsb7Ndt12fmJhoaRLnJqdbGs9LVbkAcC2yFPdZSWcj4tjc68c1W+RXiIh9EVGLiFp/f39Lk1jV19vSeF6qygWAa9G0uCPil5LO2F47N3SHpJ/lOYkdm9eqd3nPFWO9y3u0Y/Paq/yOtHMB4FpkfY77Q5L22/6xpLdK+uc8J7Fl/YB237NOA329sqSBvl7tvmdd4U93VJUrXf2Oc9F3wKvKrTK723KrzGbN5eQ6InJ/01qtFvV6Pff3BYBOZXskImpZjuUnJwEgMRQ3ACSG4gaAxFDcAJAYihsAEkNxA0BiKG4ASAzFDQCJobgBIDEUNwAkhuIGgMRQ3ACQGIobABJDcQNAYihuAEgMxQ0AiaG4ASAxFDcAJIbiBoDELMtykO3nJb0saUbSxaz7ogEA8pepuOf8RUT8qqiJ3PbQEb308ivzr296zXU69sBQUXHzhvYe1anxqfnXa1au0JHtmwrPlaTBnYcaxsrYDbuq3Cqzuy23ymzWXHxuW1wqWVjakvTSy6/otoeOFJq7sLQl6dT4lIb2Hi00V1r8C73UeOq5VWZ3W26V2ay5nNysxR2Svml7xPa2vCexsLSbjedlYWk3GweAdpD1UsnGiDhne6WkI7afjYjvXn7AXKFvk6Rbbrkl52kCAC7JdMYdEefmfh2X9FVJb1/kmH0RUYuIWn9/f76zBADMa1rctlfYfs2ljyX9paSf5DmJm15zXUvjeVmzckVL4wDQDrKccd8k6Xu2fyTp+5IORcQ38pzEsQeGGkq6jKdKjmzf1FDSZT1VcrU7zkXfAa8qt8rsbsutMps1l5PriMj9TWu1WtTr9dzfFwA6le2RrD8j0xaPAwIAsqO4ASAxFDcAJIbiBoDEUNwAkBiKGwASQ3EDQGIobgBIDMUNAImhuAEgMRQ3ACSG4gaAxFDcAJAYihsAEkNxA0BiKG4ASAzFDQCJobgBIDEUNwAkJnNx2+6xfdz214ucEABgactaOPbDkk5Iem0RE3lweFQHjp3RTIR6bG29bbU+uWVdEVFtkStJQ3uP6tT41PzrsnaYH9x5qGGsjF24q8zuttwqs1lz8bmZzrht3yzpLkmPFDGJB4dH9djTL2hmbsf5mQg99vQLenB4tIi4ynOlxtKWpFPjUxrae7TQ3MW+wZYa74TsbsutMps1l5Ob9VLJpyV9TNLvipjEgWNnWhpPPVdSQ2k3GweAS5oWt+27JY1HxEiT47bZrtuuT0xMtDSJS2e8WcfzUlUuAFyLLGfcGyW91/bzkr4k6Z22H1t4UETsi4haRNT6+/tbmkSP3dJ4XqrKBYBr0bS4I2JXRNwcEYOS3i/p2xFxX56T2Hrb6pbGU8+VZm9EtjIOAJe0xXPcn9yyTvdtuGX+TLfH1n0bbin86Y6qciXpyPZNDSVdxlMlV7vTXcad96qyuy23ymzWXE6uo4DrubVaLer1eu7vCwCdyvZIRNSyHNsWZ9wAgOwobgBIDMUNAImhuAEgMRQ3ACSG4gaAxFDcAJAYihsAEkNxA0BiKG4ASAzFDQCJobgBIDEUNwAkhuIGgMRQ3ACQGIobABJDcQNAYihuAEgMxQ0AiWla3LZfZfv7tn9k+6e2/6mMiQEAFrcswzH/K+mdEfEb28slfc/2f0bE03lO5MHhUR04dkYzEeqxtfW21aXstn7vw0/pydPn519vvPUG7b//9sJzJWlo71GdGp+af13GLu+SNLjzUMNYGbtwV5ndbblVZrPm4nObnnHHrN/MvVw+91+uW8M/ODyqx55+QTNzO87PROixp1/Qg8OjecY0WFjakvTk6fO69+GnCs2VGktbkk6NT2lo79FCcxf7BltqvBOyuy23ymzWXE5upmvctnts/1DSuKQjEXEsz0kcOHampfG8LCztZuN5WljazcYB4JJMxR0RMxHxVkk3S3q77bcsPMb2Ntt12/WJiYmWJnHpTDvrOAB0s5aeKomISUlHJd25yOf2RUQtImr9/f0tTaLHbmkcALpZlqdK+m33zX3cK+ldkp7NcxJbb1vd0nheNt56Q0vjeVqzckVL4wBwSZYz7tdJ+o7tH0v6gWavcX89z0l8css63bfhlvkz7B5b9224pfCnSvbff3tDSZf1VMmR7ZsaSrqMp0qudqe7jDvvVWV3W26V2ay5nFxHAdeRa7Va1Ov13N8XADqV7ZGIqGU5lp+cBIDEUNwAkBiKGwASQ3EDQGIobgBIDMUNAImhuAEgMRQ3ACSG4gaAxFDcAJAYihsAEkNxA0BiKG4ASAzFDQCJobgBIDEUNwAkhuIGgMRQ3ACQGIobABJDcQNAYpY1O8D2akn/JumPJP1O0r6I+EzeExnae1SnxqfmX5ex47kk3fvwU3ry9Pn512Xt8i5Jf7LrkC5etlfzMkvP7S5+N+zBnYcaxsrYhbvK7G7LrTKbNRefm+WM+6Kkj0bEmyRtkPRB22/OcxILS1uSTo1PaWjv0TxjGiwsbUl68vR53fvwU4XmSo2lLUkXY3a8SIt9gy013gnZ3ZZbZTZrLie3aXFHxIsR8czcxy9LOiFpIM9JLCztZuN5WVjazcbztLC0m40DwCUtXeO2PShpvaRji3xum+267frExEQ+swMANMhc3LZfLekrkj4SEb9e+PmI2BcRtYio9ff35zlHAMBlMhW37eWaLe39EXEw70msWbmipfG8bLz1hpbG87TMrY0DwCVNi9u2JT0q6URE7C1iEke2b2oo6TKeKtl//+0NJV3WUyXP7b6roaTLeKrkane6y7jzXlV2t+VWmc2ay8l1xNJ3w2z/uaT/kjSq2ccBJenjEfHE1X5PrVaLer2e2yQBoNPZHomIWpZjmz7HHRHfk8Rf4AGgTfCTkwCQGIobABJDcQNAYihuAEgMxQ0AiaG4ASAxFDcAJIbiBoDEUNwAkBiKGwASQ3EDQGIobgBIDMUNAImhuAEgMRQ3ACSG4gaAxFDcAJAYihsAEkNxA0Bimu45aftzku6WNB4Rbyl+SuUaPj6mPYdP6tzktFb19WrH5rXasn6glOyhvUd1anxq/nUZO9tXmStJgzsPNYyVsQN4t+VWmc2ai8/Ncsb9BUl3FjaDCg0fH9Oug6Mam5xWSBqbnNaug6MaPj5WePbC8pSkU+NTGtp7tCNzpcW/uZcaJze9bNZcTm7T4o6I70o6X9gMKrTn8ElNX5i5Ymz6woz2HD5ZePbC8mw2nnougPzkdo3b9jbbddv1iYmJvN62UOcmp1saB4B2kFtxR8S+iKhFRK2/vz+vty3Uqr7elsYBoB109VMlOzavVe/ynivGepf3aMfmtYVnr1m5oqXx1HMB5Keri3vL+gHtvmedBvp6ZUkDfb3afc+6Up4qObJ9U0NZlvF0R1W50tXvshd917/bcqvMZs3l5Doilj7APiBpk6QbJb0k6RMR8ehSv6dWq0W9Xs9rjgDQ8WyPREQty7FNn+OOiK3XPiUAQF66+lIJAKSI4gaAxFDcAJAYihsAEkNxA0BiKG4ASAzFDQCJobgBIDEUNwAkhuIGgMRQ3ACQGIobABJDcQNAYihuAEgMxQ0AiaG4ASAxFDcAJIbiBoDEUNwAkJhMxW37TtsnbT9ne2fRkwIAXF3TzYJt90j6V0lDks5K+oHtr0XEz/KcyPDxMe05fFLnJqe1qq9XOzav1Zb1A3lGtJ2q1jy096hOjU/Nv16zcoWObN9UeK4k3fvwU3ry9Pn51xtvvUH777+98NzBnYcaxp7/1F0dm1tlNmsuPjfLGffbJT0XET+PiFckfUnS+/KcxPDxMe06OKqxyWmFpLHJae06OKrh42N5xrSVqta8sLQl6dT4lIb2Hi00V2osbUl68vR53fvwU4XmLvaHaqnx1HOrzGbN5eRmKe4BSWcue312biw3ew6f1PSFmSvGpi/MaM/hk3nGtJWq1rywtJuN52lhaTcbB7C4LMXtRcai4SB7m+267frExERLkzg3Od3SeCfoxjUDyEeW4j4rafVlr2+WdG7hQRGxLyJqEVHr7+9vaRKr+npbGu8E3bhmAPnIUtw/kLTG9uttXyfp/ZK+luckdmxeq97lPVeM9S7v0Y7Na/OMaStVrXnNyhUtjedp4603tDQOYHFNizsiLkr6e0mHJZ2Q9OWI+Gmek9iyfkC771mngb5eWdJAX69237Ouo58qqWrNR7Zvaijpsp4q2X//7Q0lXcZTJVe7u1/00wZV5VaZzZrLyXVEw+Xqa1ar1aJer+f+vgDQqWyPREQty7H85CQAJIbiBoDEUNwAkBiKGwASQ3EDQGIKearE9oSkX/yev/1GSb/KcTopYM2dr9vWK7HmVv1xRGT66cVCivta2K5nfSSmU7Dmztdt65VYc5G4VAIAiaG4ASAx7Vjc+6qeQAVYc+frtvVKrLkwbXeNGwCwtHY84wYALKGtirubNiW2vdr2d2yfsP1T2x+uek5lsd1j+7jtr1c9lzLY7rP9uO1n577exW+yWTHb/zj3ff0T2wdsv6rqOeXN9udsj9v+yWVjN9g+YvvU3K9/WER22xT3ZZsSv1vSmyVttf3mamdVqIuSPhoRb5K0QdIHO3y9l/uwZv+J4G7xGUnfiIg3SvpTdfjabQ9I+gdJtYh4i6Qezf47/p3mC5LuXDC2U9K3ImKNpG/Nvc5d2xS3StiUuJ1ExIsR8czcxy9r9g9z5/4D5HNs3yzpLkmPVD2XMth+raR3SHpUkiLilYiYrHZWpVgmqdf2MknXa5Fds1IXEd+VtHDD1PdJ+uLcx1+UtKWI7HYq7sI3JW5XtgclrZd0rNqZlOLTkj4m6XdVT6Qkb5A0Ienzc5eHHrFd/HZDFYqIMUn/IukFSS9K+p+I+Ga1syrNTRHxojR7ciZpZREh7VTcmTYl7jS2Xy3pK5I+EhG/rno+RbJ9t6TxiBipei4lWibpbZI+GxHrJU2poL8+t4u567rvk/R6SaskrbB9X7Wz6iztVNyZNiXuJLaXa7a090fEwarnU4KNkt5r+3nNXgp7p+3Hqp1S4c5KOhsRl/429bhmi7yTvUvSf0fERERckHRQ0p9VPKeyvGT7dZI09+t4ESHtVNyFb0rcTmxbs9c9T0TE3qrnU4aI2BURN0fEoGa/vt+OiI4+E4uIX0o6Y/vSLtB3SPpZhVMqwwuSNti+fu77/A51+A3Zy3xN0gfmPv6ApP8oImRZEW/6+4iIi7YvbUrcI+lzeW9K3GY2SvobSaO2fzg39vGIeKLCOaEYH5K0f+6E5OeS/rbi+RQqIo7ZflzSM5p9euq4OvCnKG0fkLRJ0o22z0r6hKRPSfqy7b/T7P/A/rqQbH5yEgDS0k6XSgAAGVDcAJAYihsAEkNxA0BiKG4ASAzFDQCJobgBIDEUNwAk5v8AujzxNtKLCukAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x24cefca31d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.scatter(X['happy'],X['tvtot'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 248,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([  22.,   21.,   55.,  118.,  166.,  560.,  580., 1453., 2424.,\n",
       "        2748.]),\n",
       " array([ 0.,  1.,  2.,  3.,  4.,  5.,  6.,  7.,  8.,  9., 10.]),\n",
       " <a list of 10 Patch objects>)"
      ]
     },
     "execution_count": 248,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x24cf0d4e630>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.hist(X['happy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 249,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "8.0     2424\n",
       "9.0     1671\n",
       "7.0     1453\n",
       "10.0    1077\n",
       "6.0      580\n",
       "5.0      560\n",
       "4.0      166\n",
       "3.0      118\n",
       "2.0       55\n",
       "0.0       22\n",
       "1.0       21\n",
       "Name: happy, dtype: int64"
      ]
     },
     "execution_count": 249,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X['happy'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "No visual help"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 250,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x24cf4884860>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.regplot(\n",
    "    X['tvtot'],\n",
    "    X['happy'],\n",
    "    x_jitter=.49,\n",
    "    y_jitter=.49,\n",
    "    scatter_kws={'alpha':0.3},\n",
    "    line_kws={'color':'black'},\n",
    "    ci=None\n",
    ")\n",
    "plt.xlabel('TV watching')\n",
    "plt.ylabel('Happyness')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## The more hours in front of TV the less happier people become!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 251,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYIAAAEKCAYAAAAfGVI8AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvNQv5yAAAIABJREFUeJzsvVlwHHme3/fJszLrrkIBIEgABK/mdA9nyOlDPT2jme3Z3bFm1xGWHBrZboUPSauYcDgsW/aDLYfDoQiHH2yHXxSyXjZ0rCOkbYd2/CBveNSr2d3hzmimt6cvspvT3SC7iZMsoFAo1J2ZlZcfEpWsAqqAKhAXyfq+gCgWMv/5z////7u/P8H3fUYYYYQRRnh2IZ70AEYYYYQRRjhZjATBCCOMMMIzjpEgGGGEEUZ4xjESBCOMMMIIzzhGgmCEEUYY4RnHSBCMMMIIIzzjGAmCEUYYYYRnHCNBMMIII4zwjGMkCEYYYYQRnnHIJz2AQZDL5fy5ubmTHsYII4wwwhOF999/v+j7/vh+33siBMHc3BzvvffeSQ9jhBFGGOGJgiAIS4N8b+QaGmGEEUZ4xjESBCOMMMIIzzhGgmCEEUYY4RnHSBCMMMIIIzzjGAmCEUYYYYRnHE9E1tAIIzxtyJcNbq+WKTVaZGMq16fTTKX1J+b6IzxdGFkEQyJfNnjrTp7ff2eJt+7kyZeNkx7SCE8Y8mWDH3+yjtFyycUjGC2XH3+yfmhr6aivP8LTh5FFMATaGyyhyeTiERqWw48/Wee7L0yOtK0RBsbt1TIJTSahKQDhz9ur5UNZR0d9/b0wskSeTIwEwRA4yQ02wtODUqNFLh7p+iwWkSnWrSfi+v3wpCpKgwqvp1nIjVxDQ6DUaBGLdMvOWESm1Gid0IhGeBKRjak0LKfrs4blkI2pR3b91a0mS5uNI3VpdipKoiCQ0BQSmszt1fKh3+uwMKgb7Wl3t40EwRA46g38pGEULzkYrk+nqZkONdPG831qpk3NdLg+nT6S6y+XGry/WOZMUj/SQ+xJVJQGFV63V8u4ns+9Qo2f3dvgXqGG6/mnWsgNg5EgGAJHvYGfJDztGtJRYiqt890XJtFViWLdQlelQ3Wf7Lz+etXk5bkMM9nokWrqT6KiNKjwWthocHe9huV4JDUFy/G4u15jYaNxnMM9MoxiBEOgvcFur5Yp1i2yMZWvXxx7avyEg6DtJ/3p3Q0issjzU8nwcIFRvGRQTKX1I52nzuv//jtLxxIzuD6d5sefrANgOR7za1U2Gy1eu5glXzZO5bpoC6/2+oXewqtstBBF0JXgyNQVGdN2KRun19oZBkdmEQiC8E8FQSgIgnCn47OsIAg/FgTh3vbPzFHd/6gwldb53rUp/vqr5/netalTubiPCp1WgCiAgMCtlQqlRnCgnHY3wLOK49LU24qSabv8/PMiAN+4NIamyKfWWhzUyk/pCp4HRsvF932MlovnBZ8/DThK19DvAd/b8dnfA/7E9/0rwJ9s/z7CE4JOf2pSVxBE0BWJxWITOP1ugDaetdjGcbo0p9I6mZjKr39pgm9eHicX10510HhQN93F8ThXJuNEZJGqaRORRa5Mxrk4Hj+hkR8ujsw15Pv+TwVBmNvx8V8GXt/+9/8F3AT++6MawwiPh53pcvc36lyeSAAwNxbn1koZTRGomm54uHz94tgJj3pvnNYUx6NMTTxul+ZJpa8eFIO46a5PpylULa5MxolFZBqWs68wfZLSTY87WDzp+34eYPvnxDHff4QB0SsYvFxqsroVaP/ZmMqNmTS+Dz7+oQc8jwqnMcXxOALvx+nSfBKDxvth2AD/k5ZMcWqDxYIg/AD4AcDs7OwJj+bZQ6/iuauTSebXa2SiKrGIjCIJXMjFj0UAHJZ2dRq11aetULEzaNypPZ92a3E/DBPgf9Le6XELgnVBEKZ8388LgjAFFPp90ff93wV+F+Dll1/2j2uAIwTodWCey+iYthtqRceVNXWY7pxBs0SOE6dROD0O9nNFPUkuk4Picd7pSczPcQuC/xf4z4D/dfvnvzrm+48wIPodmBfGY3zv2tSxjuUwtavTqK2eRuH0uOinPZ/WGM1h46Dv9KTm5yjTR98E3gauCoKwKgjC7xAIgO8KgnAP+O727yOcQpym4rnDrFg96mKug+A0zfVR4zTGaI4CB32nJzU/R5k19Eaf//qNo7rnCIeH9oF5c77Au4slBAGunU2eyFgOW2M+6mKuYfE0FSru59Z4Et1gB3HVHPSdntT8nNpg8QinA7br88pcNnSjnIQZfxrdOYeN0yacDoJB3BpPkhssXza4OV/gF19skourXD2TCLN/BtkDB3mnJzU/I66hEfritJjxp9GdM8JuDLJenhQ3WFuo3V2vMZ5QEQWRj1ar2K53pHvgpOZnZBGM0BenyYx/GjTmpx2DrJcnxQ3WFmqO55PUFARBAGCx2OTGbPrI9sBJzc9IEIzQF0+SGT/CyWPQ9fIkCPW2UEtoMqbjoisymhLQSxz1HjiJ+RkJghH64lnwzR8VnoZc+WGf4WlaL22h1qZSAQffA1kUHvuZTuPaGMUIRuhLwjbyzR8MTxq9QC8c5BmepvXS9tUrksBXp1N4nk+xYfHc5ONV0p/WtTGyCJ5x7Jfp8SSY8acNTxq9QC8c9BmelvXS6as3bJdXL44Npbn30/pP69oYCYJnHKd1YT7JOE1B9oPiaXiGx8VBhdpeytVe83qSLqORIHjGMdrwh4+nIcj+NDzDSWEv5SobU3mwZbBRN6mZDglNZjyuoSliKDxEAd65v8n/91Geb1wa4/WrE0cuEEYxgmccTyNl8EnjScmV3wtPwzOcFPaiRJlKary3uEXFsElEZCqGzXuLW5SbFglNxnY9PlqtIgoi4wmVu+u1Y4khjATBM47Rhj98nLag6UE6sp22Z3iSsJdyla+avDSXJqWp1CyHlKby0lya1bJJLCKzWGyiKxK6KqErQR3DE801NMKTgdNU4HMQH+lpTMWD0xM0fRw2y8N6htP6jo4Ke6XR/mS+wHQmymw2Fn7f830+L9SD71k2yW1Xkum4JDT5WFy1I4tghGPtXtUPB0mrO62peKcJJ00T8iy+o72sqX7WwlfOpaiZDrIoYLRcDNvBaHnMjcWPxVU7sghG6MJJaW8HyV46jRlPp037PelkgNP4jo4D/aypftbCd1+YBODmPLx9v8RYTOWr0ykU6fEL2AbBSBCMEOIkm4YcJK3upA+5nTiNTVdOOvvntL2j40S/dbuXK/aNV+d4/epk+He6Ko24hkY4Xpyk9tbvwBLw+x6uj3vIHbb2fhq135OmfTguQXTaLLFBCjX74STiS6MYwQghDrMT2LDol70EQl8f9+NkPB2F7/ok568fTjr75ziy0o4jDjFs5tVJx2aGxUgQjBDiJGsK+h1YPvTPyX6MQ+4oNupprck4yWSA4xBER33oHkTQnEalYC+MXEMjhDhpN0Ivk3g/18JBzeij8F2f9Pz1w0m7TY7a1XHUcYiDuPxOOjYzLEYWwQghptI616dTfLZW5UcfP+SztSrXp1OHtokPUth0VK6Fo9DeT9oN0wvPQvrmUVtiB9HuD7JuD7I/Dgsji+AZwKAaYfC9Cl86k+Sl81kalsPt1QoTSW3ow2znPaeSGrdXK0Nn1LSF04/u5FmvmkwmNX77ENwbR6W9n5ZCsjZOYwD7sC2Uo7bEDqrdK5LAu4slBAGunU3uudZPOuNsJAhOCMdlrg+zwA7r0Oh1zzffXea5ycTQ1z5M4dSJ01RRfZQY1m1yGOtyr2v0Whs//GCVXEzBRzjQPY/6XQ4raDqf8fWrE+H398JJC+yRIDgBHKf0H2aBHZavtdc9Xc+nULW6SusHufZRbpDTpr0fBYbRZg9jXfY86N9bIZeI4ANLmw3OJPVwPLbrs7DRoFSXee1S7sB74TDfZS9BNoxVepA1e9L1FiNBcAI4Tuk/zAI7rABXr3vm4irFerdPdZBrn/QGedIxjDZ7e7WM6/ncK9RCiuSILPJ7v1jg/FhsIG1959q2XY/FUpNSs8Vrl3L8/F6RXz2sMJHQOJPSqJsO6aiM7fphxk/7OidJdbJTkCEKA1ulw6zZttD51cMKqiTy/FQq3BPHGVweBYtPAMeZWjZMIO2wArO97jmR0JFEYehrt/nbP1gu8Wd3C3ywXOLBlnFqsy9OG4YJYC9sNLi7XsNyPJKawmajxU/ni6yWjAOnTS4Wm6Q0BcfzKTdtqqaD50Oz5WI5Hp+t1ahbDomIsv33FvNrNd66k98zYHpUgdWb8+ssFOt8sLzFreUytutRarYo1a2B01MH3XOdgfyvnEtRNR3+/P4mxbp17CzAI4vgBHCcqWXDaISH5WvtdU9JFHjjlRnyVXOoa08lNf7wVp50TCalKVQMm6WiwQ++fWGoMT0tOIgPf1C3SdloIYqgK8GxUGs6RFQRxx9cW9+5tmuWjSIFRYGLm3XOpiM8LJtUTZtLcpyILLK82eTl81lKDYtbKxUEwWcqpWG03C63Uvt5gSNxrebLBm/fL5GLRUhqMqbtcWulgmHbxNTdils/q7TX+l8tNcklIvz+O0vhc3RaTwkUXr2Q5dN8lY9Wy3z7ufFjjVmNBMEJ4DjzzYc93A/D17rXPa8Pea02f3ux1qJm2aQ0lUvjcfJVc+hrPek46thSSleoGg5Gy0VTRMpmC0kQiKmPHAf7ueV2rm3L9rhXaHA2pVOst5jLxTib0ikbNlXT5vxYjHzFQJFE5tdqCIKP7wtcyCV2uZXaz9sWLIftWr29WmYspiKIIAgCuioBkC83SevdSlrYX2AAPiEBH0QBTZHC/f7jT9apWzaXJxLhNbOxCK9dylGsW3zv2tSBn+MgOBFBIAjCfwP8bcAHPgb+pu/75kmM5SRw3BkrJxEUPax7lhotoqoMwvbhI/hE1dNboXmUOOrY0sXxOJoiUay1qJo2iYiCroqcST269n6Wa+fa/qJQx7BdMlGVtK5Sbrb45GGF6XSU33h+gmwsQs20Me0kuiqxVjWYSmlcyCXIxlQ+WNoipSnYntdlkby7WOL1qxNd9z2MuFGp0eLqmSQfrVYAB02W8PEQRYFsPBhrp+J2MRfrKZivT6fIV81QOGw1LKbT+q739qDcPDVFZ8cuCARBOAf8V8ALvu8bgiD8S+A/An7vuMdykngWMlYOAwLwy4USmZhCUlMwHZdfLpR4ZS5z0kM7dhx14Pz6dJpC1eLKZJxYRGZ1q8n7i2XG4xqe7w9subbX9lt38pxJaYFmX2ySjqo0Wy6KBOmoGvrBOy0ao+X2dCt1Pq8gcCQHaDamYrRcbsykWdysUzVtZFHgN5+f5PWrE7sUt16Cudy0efPdFV69kA2Fw9v3S3zj0hgJHo03FpFJ62qYVnrSlegn5RqSAV0QBBuIAg9PaBwjnHr4CALgC9u/CsHv+Cc4ppPBUceWdlqqZ9M6L307M3Rcp4224BIFgWwsAmQo1k0+flDpeb2dbiVZFCg3Hb5+MdX1vNfOJo/kAG3fP6HJ3JjJhNdtN4/f+dw/mS/sEsyFmoHr+V3CYSymMr9WI3dZ63qOC+OxMFZw0rUsxy4IfN9/IAjC/wEsAwbwb3zf/zfHPY4Rngz4CLwyl2F50wjdFa/MZfCePTlwLLGlXgfeQWIx+bLB0maDD5fLjMcjzOWiZGMRIrLEt66M9/SB7xREz03GKTYCq6DTImk3cTnsA3RYl20vwVystxiLdQuHq2eS/Pzz4i7XUvvap8EzcBKuoQzwl4ELQBn4A0EQ/mPf9//5ju/9APgBwOzs7HEPc4RtnDRhWdtcf/H8I1dQzbTDQN6zhCelGrod1D6T1KkaDhWzxYfLLa5MxpFFcU/BtfNgbK+/Xs97FM89zMHcLztuIhkIglKjxeJmnY2aRUqXMW0Xw3ZP5XsTfP94VStBEP4a8D3f939n+/f/FPi67/v/Rb+/efnll/333nvvuIY4wjY6s1R2ttU76kXcPgDub9RZLjW5OpnkXEY/1jGMcDC8dScf+vo7D8NsTOVvfOPCib+3YZSb/b7bj1PL9XzurtcQRfA8QiHYac0ch3IlCML7vu+/vN/3TiJGsAx8XRCEKIFr6DeAU3/Kn7RmfBI4Kf6TTgF0eSKBpkjMr9cwbZcL47FTp009aTjqtdwZ1M7GVLKxLJ7vU6xbA/BKDX7oHpQHadAU3EG+28uCmEhq/N4vFnA8n/HoI7dYzbS5Ob+O7XKq2pnCycQI3hEE4YfAB4ADfAj87nGPYxicNDPgSeEws1SG2cQ7BdBsNkYmqqKr0rHnV580DvvQPo61fJCg9n7j6vx/URB45/4mP/o4z2sXs7x+tffYe83dMMrNXt9t/+z1XqbSOufHYrx0PgiUtxGLyLy3VOLl89lTxQYLJ0Qx4fv+3/d9/0u+71/zff8/8X3/VBPHPGlt5w4LB+F571X6Pywn/pPS3emo+eOPopfAcazlg1CV7Deu9v/brs9HqxVEUSAXi3B3vd5zTvrN3cJGY+C11W8d3t+o7/te+u0d3+/fce8kMaosHgDHTXzWS5OB4/Erdt5bwKfYsJlOD0+/26nV9aoE3Wq2+pKZPQndnfZj2TyMd3RQ19xeVsRB1/IwlslBgtr7jav9/7dWttBVEV2R8X0/yCTbFhid1+83d8MUcfVbhxXDZiq1u0Cscwz9Mry+ci51Ktf2SBAMgOM8mPZiP5xO60fqmmrf2/E8ClWLzUYLy3aQBUhF1X03dL/Nt7MStNSwuLdex/G8kM2x83lOS8vHvQ6//Vg2D+MdHeTQ3s/FchRum17YK/um17zuN672/9dMh+T2d0zbIxFRiEVkPi8ERHXta97fqHfRN7TnLqUrA9cg9FuHaV3tqdV3vpd+whA4FWt7J0bsowOgbequlJq8v7TJH/0qzzsLm0wltf3/eEj0MpGHZT98nHs7nsfnhQa26zORiJDQFZZKTb5zdWLfxuf9TOl2JWgbi8UmogjjiUjP5xmGMfOo3DP7uWX2Ytk8rHd0ENfcfi6WYdw27bn9h396j4ViHdv1HvvZ+s3rVFLbc1ztccuigGEHfEiG7TKXi7K61WS51Oy65nKpyepWc9fcXRyP71pb16dT3F4t71pD/dbhhfHYQO9lKq3zvWtT/PVXz4d7Z5i1fZwYWQQDINBSU7z57gqu5zMWizCRjBxKp6yd6KUF2q7Pzkrao3BNlRotClULXZHCPP20rlKomfsGyCA4uFa3miFBXCKikEuouypBN+oWsigwNxbv+zyD5HPvpanuN9b9XB37uWX2Ytns90yd9745X+DjB5WwjeHOgGe+bLDVsHj7fomxmMrVM0kisriv9rifFTGo26Y7OAsCArdWKtyYSZGNRQ68/m7OF1go1nE8n4QmMzcWJ6HJ5KvmnuNqj/vmvMAvvtgkF1f56nQSRRK5u17nTFLj3no9XHdnkhp31+tkouqeRVz7WTv91uHjaPWnpYisEyNBMCDyVZNXL2S7TNeaaR96tL+XiaxIAgHrziMchWsqG1P5cLnMROLRQWI6Lrm4ysJGg0LV2juVbidltNliabPJD759gYmk1rHJFc4k9a7xH+R5+h3W+6Xo3V7e4s13l3E9n1xcxbI9CtXuZ7m/UadqODRaAVf+XC5KOqqGh9+gdAi9OOh/+MEqCxsN0lEZfIH3lsoUay2+//LMrgyZb1waY36txs8/L/KNS2P7ao+DuH4GOYg65zapK1iOh65ILBabZGORA72vfNngF19sMp5QQ96oWytlvjqdwrDdfcc1ldZ549XzIe9PqdFCVyUyusx61SKqysF1bY+tpkNGl0PNu5/A63zOUsNisdhko27xoNzcVfPQqTwoEvsWiD1JKecjQTAgjitg3MsvmY2qsN3UJRaRebBlML9eZTYb5a07+cdaYN3BYbBsh7IhkNZVTMfFaHlMT8RZqxqcSWl7BsjyVZOX5zJs1E1qpkNKV7g8nggoo2czu7SwXiX3w6DfO9krRQ/gzXdXkESBsVgE03G5V6hzZSIePku+bLBcaiKJ2/OwzUt/eSLG2R0aaicdwmKxyaf5CrbroUgi2XiE77843TW+26tlSnWLTEwJef8FQaDUbIX338lTn7ussVxqsLjZ4Cfzhb6HSmBFtEKN+eqZBBFZeuy5nRuLc2uljKYIVE03dNvsdc1+qZu5uIqAGNA8KzLg8P7iJtl4N1f/fgKh8/8/XN5CFMH1fBaKdZq2i+v6SGJ033Tj9nO2eyHoisR4XGW51OQf/MldZrNRLo7Hw0KxTuVir8LGJy3lfCQIBsRxBYx7me7ff3kGIKT2XSo1eW4ywXQm+lgLrNdinUzprFdNWo5HLq4yPRFHEgVSurJvgKzUaHEuozOTjYaftQuJ9nvGgxSJtbuXtQVPQpMZj2t9U/SKdStsxzgWi3QdRoWaQUQJQma3V8tcnUxyr1DHdFw0RcJ0HO6u1/mtjoOlM3C8sNEgXzFwXY9ay8N2PcbjFoWq2fVcpUYL2/VIqY/WjaaIVAw3TCHcKeDawfX1apPFYoONegtdEflb35jjN74cjKfzXX7zco75tSq/+GKT1y5m9zys+mmsnes9G1O5MZPm03wFHx9dlXa9r97ZZt3JDXXL5uqZBB+tVsPnrjVtPl2v81fPpg98YKZ1lbWKSaHWIBaRkAUR07VZq5jky8ZAFtRisYnrejxsWmzULBqWy9Uz8bA/w5vvrnB1MjFwBtdJFWMeFCNBMCCOu5lMr8UylX5E7XsYC6zXYn1hKsnFXJRMLLJLm9tPEPY7mKfSu4Pqh+En7de97CvnEn3HWmq0GIsFWn47DqLJEoWayfWZgM+oLdBiETmkI05qMklN3XX4tQ/fqtlCFAUWN03imowqC2w1bf7Jzxf4HzriSNmYiiKJmI4bWgTFmkWhZlK3bJY2G6xVTLIxleenkmRjERaLTTYbJkslk+m0zmQiQsW0+Ud/dp9cQuP6bGaHFQG5y+MhJ9NBNNad612RBC7k4vtW4IoC/NGddcpNh69Mp7h2LrnNPBqkbk6ldG7MpFgsNqmaNht1ixfOJELl4SDr+cJ4jIVinagq4Xg+UVlibiyGrkr7Xqf9nEubDaqmjSgINCyHZFRhs95CFODieIylzQYPy02unU2HlcKditBOodrOWmq7m2pW0OUsqcs9v3/SbqOnXhAc1oQflhb7uDhMF1W/axm229Ok3k8QHndbyX7dy3RF7JsieHu1jGm7fF5oAIFWWjZaSKIQZqi0tcQ2PQL0JrrrPHzrlstW06bRcgGB6YxOy/FYKTW5OV/gjVfPA8HBc69QZ2GjgR/1qRsO84U6GV3B96FuOfhAvmxSNR1emcuwUbdY3jQYT0SIbx+UaV2l6Fr86E6eiaTGT+9uIAqQ1BXmxuJkY+qe62I/jXWY9f6o2Mvjo9UqluuRjct8sVFjoVhnPKEymdSQBCFUEG7MpmlYDuWmzYvns13Xa4970L17fTrNjz7Ocz4bHP6m7WHYLlfPJPYt1Go/57+9t4HlBFlHaVclE1VpWC7Fuh26jEzbwXK8MGiuSGLYpWynUF0uNbEcl42aja5IJDWFstGibLS4vby1y8100m6jpzp99LArM3ulgx03DpJSeBjXGiTtrX0wpzSVmuWQ0lRemkuTrx5N87lSo8V0JsqL5zP82nMTvHg+w3Qmio/Qd6zXp9PIosjliRiKJFCoWbgevPHKbFcx0CAplqVGC8vx+GC5xNJmk3vrNRQhcIcJQtA3IRNV+PhBpWsev//iNK/MZTBtj7WaydXJOLNjUXKJCJlohPG4xkRSI6nJfPygQjamIEsiGf2RhWO7PpmowkIxqHKNyCKqJG0fVMHhude62Jn+Wmq0uLte7Woa356vtiV1e7Xcc++0r7VYbKIrEmlNxXJ8CjULURBotjyqpkOpaXN9OtX1Xr5xaYyI3H0MNSwHAQbeu1NpndcuZvEJCswissiNmRQRWRpoX0yldS5PxFFliYrhYNou+YpBoWawstmgUDWIR2QkSQTBR5NFPs1XwzXRK2X36mSS9xbLCIKPpgQWoO8LXJ1M8qM7+VPHVPBUWwRPmp9uEBymi2rYa+3nzmkfzLPZWPhZrxjBYWGvuM1e7rW2pqspEjdm07s0zUG14c7uaXO5GJ+tVSk1W4wnNVqOi+V4nElpCMLuMbzx6nneAH7/nSVy8Qg/u7eBJm+7qhSRluvy2qVxinWL71yd4M6DChXTJq2r2K6P5bj4vsfyZpNi3SYekdAUiZlMDE0R+DRf4UIu3vddds5dqdHi1kpwaJ1J6uGhG+TX76+5tq+1Vm3SbHlUmjZLpQZRVSKqSlRMh2wswnOTMfJVs8vabCtr0L0GFYmh9u7rVyfDTLFh90W+bFBq2ozHI5i2S920WS0bTKUi6KpMy/Uol01eu5jBBaqGjecTzkOvBjVt12JSk6maNgIgi/DFRo359RoXcnE6y92OkqlgEDzVguC4qSGOA4fpojpsd5eAz9tfFIMc8e2Uy7b5fBQ4qFAcJD4xWAzjUfe0RERiJqOzuNnEsF1kEcZSOrbn8bWZVN8rPDqQ5TBu0K6YfbBlsFY1+Ml8ga+eS/KzL0o425aA73t8XmiSicpMxlWatkd+yyARkdFVGR9/T1dD59wtFB81jb84HgsP3R/dyfOlM0ls1+dn9woslQxsx+PDpRJ/4WIWHyGkXv7p3Q0elk1USSIWkZBEAQSB9ZrJeFzjxkyqK/22c557rcGfzBf2TE7opCmvGIGATEcVTNsZmvP/9mqZ5ybjfF5obMczfCRJwGh5JHQJVZSYmtQpmzaJSGDtZGOPlI9+CsmFXJTnJoP5u7VSRldFfA/iqswvF0p8/eJYuDcCK8jnzXcWufOwiu/DV86lwu5oR42nWhD0e0ECdJWjn3SgZlgcZkHKYV3r9vIWb98vsbJpkInJ5GIR1msmc9lomPW0H4aN55x03GZn97Qvn00HvuWWy1hcQ5EEzkU1Xr862fca7QN5PK5xd5tq2/MARP7w9iZTKQ0BgYmkzmtzWQzHpW45rFVMvjqdQhIEbM8PYwf1lstfvDLeN0jcRufc5SsmZ5I6F8djYWA3FpFZr5pcyMV5+4sihZpFTJWEzD9hAAAgAElEQVRpOjZ/dq/I3UKdr82muDSepFC1UCSB2WyUxc0miixzaTxOy/WQBLGrUX0/t+POse5l7XVSoayUDEQRqoaDpiSQRIHvvjDc4dm2ZOPb7q3AOtCIRUS+fjHHrZUKjuPyWaHOc2cSyKLAmaQeWkf9FJLfvjbF7dUKC8U6miKAL2A6Lq9dGmN+vc6n+UpIR7JaavKwYrBQbCAKApoiUqq3KDZafP/F6SNf00+1IOj1glbLBng+miIdS6DmtGUHHAXyZYM3310mocl8dTrF5xs1Plgpk9RldGWwMNRB865PskqzV/e0lVKTtarRk0yvHxQJ5tdrGLZDPCKT0mU+flhldkznXDqKaQe0H+06hu9dm+Lv/t8fcDal02y53C8Gge+oIrJeswZ2iXTOndFysV2PD5a2qFlB0/Z4RGZ+rUrdcohHFDzP52HVRFNFJFEgX7Hw/KAGY3nL4DdfOEO5abO4WWe9YrFeM0lqclej+kFdmHtZe22X7731OlE1KBozbIeNusmVicTQrt9HyQGRQBAKPhUjSD7IxiLcmEnxrz/OI0siqY5gfLug9HvXpvoqJBNJjf/tX5fYbNoIApzPBIWJr8xluno3y5LAQrFBRJaJRSRs16di2iwX68fiyn6qBUEvjTEXU9CU44kbHFVRyWkTLp25+Q3LRRJFzqV1NEXCcvyBnrmTZvjWylbIK3NzXggzbuB4n71X96l81dzVjWqr2QpJ+iRR4I1XZrg+mxno+u318WvPTXT5xzcbLSYSWlDrsJ2tVKhaaIpEvmzQtFzeW9oiF4uQiEg8rFiUGi3Susz16dRQc3J9Os0P31thsRTwJSlSUCU9FldZLDYwWi4pXeRh1cR1fcbjEVzfw/VAV0UKNSPkk+rMtGoLxYNYa3tZe22ffM2yQwI6TZaobhcoDuv63Sl0xuMaS0WDS+NxPN9HkUTimsJfupzrcjXvpO7o92yaKjGryUGmV83iD2/nSWoy01md72yTMf6LP1+iULeIKg6lZkAoIyKwXDKOhaL6qRYEsPsF/f47S/sWRh0W9qJA2JmnP+gGOY0Vi0EsJqhELtQMIrKIIomUDZtL4/EwI2K/QLMoCHy0WkFXRZKagmE7/OKLzdBPepzPvvNeD7YM/vBWnpfm0mEh3+3VCudSEX78WSGkq5hI6ANzUPVbH+8tlcL5bNcaaIpIoWZxfixwSbxwNskff1rgs3qVStMmFQ0yT77zpYmhObCm0jq5RIRSs4XteSQ0ma9fTKFIAk3L5qHrUTFtHN8nF4/gIyALAroqhjUYL86md6XsSqLwWK0p+x2uodsoooT1IKbjktDkLvfRQWmzp9IaP/j2BfJVMxRC/bKb9mNt/b1fLFAzHKqmw5rU5GHZwnI9NiSRK5NxfvjBKng+juejiFCsB8pENqYGLMB1Zwe5zNHgqRcEO3GclNK9gtWW4/L2/RLfuTpxoMPstGVC5csGS5sNVreaVE0Hs+UyHt+2DASBuVyUWETmi0J9V1ymPe5So8XSZoOtRouE3kG9gEgurvakXtiPF+ZxsXOeN+om6ZhMsdZiNvsooPrLpS1evTB2IA6qfskMvg8TCZ17hTrgoMlSWOsAAglNQlNEzqY0PlurIgrguB5nUxpn0zqKJA69HnzgtUu5ro5axbpJsdaiWLeoWw6KLBJVJCpNh0xcZTweCcfVjoP00uAP24pra/C5hMq99Tqm4+B5MJ2OUjMdLuZiQ/eK6CV0rnf8u1920369OUqNFufHAobUdxa30GSBbCyCJAoUai1sJxCeiiRSrLVw/cDKqhk2kiSS1hV2Ek4eBZ7qOoJeOEj3pIOiV57+/FqNsZh64Bzi09S9q73YA4oEh2rT5kHZYHGzged7fOvKGNlYoE0v7aAJ/uF7K/zwg9Xws8mkxidrNWqmje/7Ic1wZ1FQ+9nbvDCW4zEeD3LcH7dz1050znOp0eLOgyqrpSa/elgJx9MOqB70ffSr4/jKuRSSKHBlIo4qCRRqJq7n88YrM/jb118sNpnJxrg4EeeVuSwXJ+LMZGMsFpsHWg87x1JqWPzZ3SIVy2FuLBaMo2KxuNnkbDrCC2cSGLYX1mBAb7bXx6nl6Ucx3tbgz6Z1ZrI68YjMbDbKVFrjuy9Mkq+aXXn67V4Rd9drQ42h8/63V8u7aiD2Ut7aioS+3W/7i2IDAZ+EppKLR5hIRNAVieVSk8XNJlNJjbguo8kiTculYTmMxRT+0pcn8Y/BJnjmLILjzDTpFfAq1lt883Ku63vDuKaO06LZD+3+BRs1m3MpnUpEBgGqpsOL59PM5eLUTJv59SrP7eBpKTWDg+r5M0kg6Ev8/GScjapFRJZIRBSunomjSGLoI+/khfE8n3ylScWwiWsKrucfqlXUvlc79U+RBGwHFFng1kqZGzNpFElgMqkd+H1MJbUOanOViWQEWRS7aLQjisj1mUyoqPzoTp4Pl8sUqiZzY1GiikSj5RJTZDRFpGraB1oPO9fqp/kqW02LibhG2bCZHYszk4Wq2QJB5FwmyoXxWDiufi67g1qwB6WHBnbl9bd7Rdjeo34K+42hzRJbqlu7SAQHWWNtV2fVsGm2XMyWh+14LGw2sByXGzNpNEVkq2kzmdQYT2pczMYpNW2aLYeoJvP9l2a61v9R4pmzCOD4KoR7VeMexNfYieO0aPZDZ/+C8aTG5YkEf/HyOC9MJTBtL3zm2WyU6Uy0629t18d2va7PXpobI64pvDib4cZsGkUSezYnWdpsBC0HW0FgOh1VuLteY2GjcWjP1r7Xp/kKmiIwHo/QbDnkYpGwYKudIniQ9xG4SypcnUyQiytsNizurtfCQO/ONQrBYXsmqSOLAj4+dwt1ZFGkYbokojJGy0UWhQOth51r1XI8ppIatusFVcuyhCoHLSLPpnUujMfCvbNXM5yDWrD7NdjZCzutm5plgxD0Pxh0DDfnCyxsNBBFgZSuIooCCxsNbs4X9r1/ewzza1XGEhGm0zqW44EgoIoC+LBetXhYNhAFAdN2+HB5C18ASRIYi6tczMV2rf+jxDNnERw3dmouw/oae13vNHAeQf/+BbPZKDPZGH99O9vnrTv5gXosRGSRb1wa68sh380L45HbbhAU1xS2mhZlY3B3SL+MoO4CJZmtRou4pqDKIufHdNaqJn4VsjGFH2xrpxNJjZvzBd5dLIWNZva75ycPK7Qcj4gikdBkXpkbQ5GEgLK7x992HoyO5/HFRo0vNhosFhvcmE7hez7FhsVrF7O7mtwMis61+tadPO/c3+Tueo2UHigptusjiyIRSeBn9zbCuVvYaHBpIk6p0WJxs07NdIhHJJKaGnbzGtZiepxi0IP0iti5Hn4yv47j+pSaFlFZZiIZIR0NKD/e6HPfnZTuK1tNZjMx6qbDubTGZqNFSg+4vEzbZbXUZDqrbwe9XapmkLabi6tEVakny+tRYSQIBsRpIq87zNz5x3mu69Npbs5vUDZau/oXdG6yQXos7Mfv3sZUWufGbJrP8jUalstCsYEsisQiEufSQdOZH93Js141mUxq/Pa1qV2pnP0ygi5PRilUW10FSpoqMRZX2KjZpKMaZ1JRykYL1+sO4Nmuzytz2fBZdiYAdDN0CvzqQY2IIvDcmUTID9Ru0NILj3jzW9x5UEESRb40EWej2WKt2kJRJH7nmxcGSlsdBFPJ4OB6WAnSF7NRFc8XiGsim80WZ5Ja6LJZKjVpuS4LxSY108b1fFzPJ7OdbXN7NeBa6qzlycWUPfsP9C8G9XnrTp6FjQZlo0VKV7g4Hu+6xs499txknGIj6CDn+f4u5avXevh4tUJMldEUGUm0eVBuENNkHM/v2QOk1zUqDZtb9RI10+VcRudGNsVW0ybu+jw/ldyuN4nywVIF1/dJagqaIqIpEv/ld64cq3I3EgQD4LDTFjsP8vZBvFfDkaPCzuda3Wpyc36D89lH/t/9DuU3XpnhzXeXKdTMrv4FneZsL+HX2WNhWIGYiQb+XgFACEJptudRM1r87k8XSMcCzfmTh1V+/sUm/+61Sf7K12bCa/fLCPosX2c6E+0qULo6meSPP11nLhfdQR6W4OZ8gUxM5Wf3NlAlkeenUn190J33vLdcIhtXcFwoVltcnIgDDvNrVV7dhxtocbMeFngBXI5GmMpoQcykjzXRfteDCvy22+prMxlkAT5cqfCwYnJjOoUiizgeXc/63GScf/1xHh9I6wqyKGI5Di3b45N8tevdC7Bd0Cn3FZrwSHkoN20KNYNivYVpu5xN6Zi2z3Kp2SGwJQpVq7tbXg9LvHMMikS457YaVtd6uF+sBW6ZlkMsotByHDbqLaKqxK89Nx4Gmzvv153R1uJeoc65rE6xFliUluMhSxKTSTmML+UrBoVqi3NpjUrTpmLa1C2Ba2eTx27hjwTBADiqlM2TrgnYmY75eaGBJAYBQaOlDTSW67OZsA3lXodMPytmmPqJR26VKvhwfjyKJgc55FsNm1/la8xkoyiixP1iHU2RyCLwy4UtYhE1fJadboea6ZDSFD4v1Lk8EfRRbhco3ZjpJg9LaDJXJ5N4vs/PPy/y61+aQCBgGm0HkXtRQHfes2Y6nEvrLBQblE0X3/fxvaCI7Pp0uueh3T4YN2oWjuujSD4t1+NcLoYmS1SMVl+f97DrrHNdzGSjfGkqxc35de48rOLhc/1cN3fSdCYaUjg7HuiqwHOZBJJI4Ep59XyXy0lTpH330lR6d5/wla0Gtx9UuLVSIa4FBHum6/LOQomkJu9KI+41j/AosG05Hu/c3+T95S2+ejbFV2fSZGMRlkoG09koS5sNinWTraaN5wWZbBu1FvfW6+QSateYO9/v4mYdXRXJRKOIgkBMlbnzsMp61eQvffkMivSIjrthuViOi+G6pHWFiCwdQ7LobowEwQA4KvK6QQTMsK6bYb7ftXi3KYQ1RSRfMRCEGhu1wXL0D+qqGnSsOzM4vthokIjI2I5HywkKoP7ChSxv/nKJa5rC4mYjDHDKYtAGsrOobafbIaHJYVygV4FSmzys003x8883yMWDNODktsanqz6Lm3WysWzPpj3te7YPobMpnbJhh77h1y4GFbn9Du3vvjDJg3KT1S0DRRS5mIsR3y6820nu1zm3S5sNziT1gRWZznVRalh88rCKj8B4IoKmSKxsGbifb/DNy+Phc6WjKnNjcaIdgeGm5exiXu23lz4v1HbVmXT2CS81LN5f3kJXRDaNFgld2u6YBrIkcCkXY6PeCueq3zwqkhBWsH+0WsHzfTzP588XSiyVmnzrSg7BD+oyFFHEdj0kEXw/yOb/vFjfrgCWuHYuyfeuTYW1NB8ubzGeiLBWMZlK6Zi2x5mkzovnM3z5XJKPHwT3a/v+1yoGnz6sEddlYorE5nbyxWG0oB0Wz2TW0LA4zB4Andgvo2LYHOxhv9/5XDXLRlNEig2Lje2skfF4hFLDPvQc/WHHujODI6pKYfrprz03wYuzWSKyyGRCo2LaNG0XRQqWdrPlkdbVrnndmXk1HtcoNxy+NBWn2XLYalo0LRdVEnlnoYQkCLyzUGK51AgzgzYbLa6eCYiE53JRDNvF9wKK4l6ZQ+17rpSa1Ayb95e2WCo1+NpsihdnM1zIxXn9ane65c5smam0zt/4xgW+cXGMVEzZ5vq32WrYZOOR8H4757bUsLm7XuuyGPbKmulcF4vFJnXLRpVF0rrKbDaKLIlsNVssFGvhs752MUvFtDFablgHUjHtrsB5+8D88SdrfLBcCu+/utVkeUedyY8/Wef+Rj3cHx+vVjBbLoWqRdW0MVtB85m6aZPSFSwnoL5oz1W/efz4QSXsPOf5Pg/LBumoigC4vs9P7xZJaBLr1RaKLJKLa2iKjCgKKJJIw7TZalg8rBjcnN/gT36V58efrDOZ1JBFkYphU6iZPCgHLLRzuSBbLiJLfOvKeFemouv7zIzpxBSZLaNFzXTIxII4weP2ThkWI0EwAI4qZbNzw5UaLT5YLvHjT9ZY2myEGt0wKXTDfr/zuWKqTNlo8bBskokGJfq3VstUt4N/va7Rr+Cn3+cHHevHDyqkozK6IiMIAtOZKLIocK9Q73of/+HL05QbDq7rYzkOddOhYdm8dD7dJbh3pkq2KQWen0qFBUppXWatanJ1MsGN2QxXJxPcXa/xeaGGrkq8djFLRG7XNwTEZD4+nk/PYqOptM65VIQ//nSNX+VrRCSBluPyoztrvLe4GaaN7qccTKV1vv/yDJfHo3y+UeP9pS2KdZOGaXNzfp0331niH/7pPRaKdWw3yJsfj0cQxcBl0cZeikznuqiaQV677xNmaD03EcdHIF8xw2f9K1+bYS4bxfM9KoaN53vMZaNhxXFbOHUemB8ub7FSanJ3vc7VbYurcy1UjKAmotSw+Gy9SlKX8XyfiCKRr5hUjBYt1ycZUcJDtz1X/eaxzYlUMx0qzRaR7ZTY6UyUlCaHSoSuClSbNpt1E3wfSRBQFQGj5WI6HpoiEZFF/ukvlnA9n9lsjK/NBk2ZUprK6paB7Th8tFrh559vsFpq7jovUttV9FNpnbSmBoynakA6d9zNak7ENSQIQhr4x8A1Aovrb/m+//ZJjKUfdrstUl3cI4eR1tUZELu7XkMUQRZFJpOBf75hOVza9lm3sZdLalgXVmcQN6kHgiAiCmxtF8NIgkBal0N65E708zsP2sxkd4P2QMPMV8xwbtrfb3P+t5HQFGYzUdZqu99HLqHxB++vcHu1Qi6m8psvTDCe0MIskZ3v9TsdfO+dgda37uS3GTkfkeDpikQmqobugO6evmLY0xfYlQAA8OPPNpjLxZBEkXvrNSKiyOWsjipLIT/QzoYxi5t1NmpWyJ/THmssovKdq5PhunlYNlkqNQN6Dt9Hk6WwpeJcLsqHyy02albPrJm91oWPT0SRQyEAIEsSl8djvHpxrKvJzPdfnunr6usU/G265426FWTOZKOcy3TvpVgkIGlbLTX5YHmLimGjiCJRVSKzTWm9XnFJxmTG4pGwj3An1XWvrKNrZ5MhoWHFaBFVFVquy4VcDFkUGYtrVAybr5xN8Wm+juW4ILiIIjQtD9v18HCISCLpqEplO5g9k42G7KWzYzp/8mmBuKZs18oIIO6uDr44HkdTJIq1FiWjha5IeK5HsW7zwXKJ2WysbxbZYeOkYgT/AHjL9/3vC4KgAtH9/uA40euQu71aOfQgbnvD/d4vFnA8n/Fo94J+UG4OlYN9kP4LOzOY/v6/uoPVchmLRYipMjXToVi32NzuIdsv82ZnM5P9/NGDdMhqz/e1s0neWyojbPO0m7aHC/zWtTNdzKQQBK+vz2a6Dvy2Txb6+993vtd2Zeiff1GkZjm4no8oBIyabRK8XmnA/e6hSIRBz4VinXhEBgRqTSeopG67f5Iab767TM20qRgBL76mSKFy0Fmte69QIxqR0BWZtXKQDXM2rbC6ZZARQVckFotNXjyf4cpknPUhFJn2urg+neaHH6yysNGg2bLBF6iYNnPZ6C4Nd+daageYfR82axbjiQiCKIRNi27MpsPxPNgy2KibYRB1PK6RjsoUGzam45HRFcqGTURRmMnqyGKUZV3h+myGuCqzUKzz7mJpm/11lont+QJ2pScD3JwX+PhBBdl1GY+rrJYMqpaNLMB0RueVC2O0XJ9CtYXRavFFsYXr+6iKsN2/2AN8HC/oEvfS+UdCdX6txnRG55uXx8PPevFPXZ9OU6haXJmMUzVbfLYWZCtdysWwHI9fLpR4Ze5w0oH3w7ELAkEQksC3gb8B4Pt+Czh+opw9cJzEblNpnfNjMV46H+ki/IpFZFK60rcJey8cpP/CTg350niMshHB9+FBuYkoCERkCUUSu/6un/WxXjV5qU8z8n5j7dchqz3fr1+dDLSmZouK4aJIQpfbod+87nxXb3X0ioW9g/O/elhhYaOO6QSHd0QVaFguTcfh5vw6b7w6N9Q93lsqMRYLgtFN2yW6TaxXNlpciMTDgGmhavHcZIJ3vijRcj2K9RbfupJjNhsLD5P23NdMJ6Rhdjxvu5+uREyVMFoemiJQNV1qpo0sigMT8+1cE9++nCMXU/n4QQVBgJfPp/csWsuXjS5a63rL5rP1OvmayY3pdNgAvt1fYSqp8Ye38qRjMilNoWLYLBUNvnIuwXRa52szGSzHw/V8VrYaLG02uZiL8RvPT/LCVLIrs2giGdlW2rQ963XeePU8L0wl+cf/doH7RYNMVOFCLsq99TqWG+TtfPNyjo9XK7x930ASgwZAruejykGuf8uF2TGd9ZrJzz/fwHZ9FElgpdTkN1840zUnvfZAOzPqR3fyfLxapml7PD+VIBEJutQFx8Hx5BANLQgEQcgAM77vf3TAe14ENoB/JgjCdeB94L/2ff/w+AEeE8fd4rKfJt8ulBk0176XlrpX/wXYrb1uGQ5nkhE+XavRtF18zwcBoqqM43l9M29KDYtP81U2Gy3e/qLI81PJsNtVLyumc6z9OmR1cr3v5XYYFPu919vLW+GhEpEFlktNIopMWpMBCR+f2UyUOw+rQ9+j7WP/vNBAFoL4gO0SMrQ2LIeKYTOVCrJ7FooNLk/GMR03rJhuj7VXe0vH8ahZDh+ubJHQFL50Jkm+YuDjD1Whupc1vNP66ofbq2VKzaDIUFel7SwajVLT4kHZ4EtnkpiOw931Or91bYrbq2VemktTrLWoWUFDmEvjcT4v1Lk4kWAuF+XWSgVdkbg6mWCjbm0H1ye4vVoOM4va6GwYs1/q8zcuVbm7Xgvaq2oymizh+n5oSX1lOsWdh1XOZ2OUGhZrNYuW45OKCEQUEV2WQrZc8GlYLhXD5mf3Cpwfi4VNbHrtgXzZ4Kd3N4IKc1UmFoF82UQWJc4kNV6Zy+AdUy7pQIJAEISbwL+3/f1bwIYgCH/m+/5/e8B7vgj8Hd/33xEE4R8Afw/4n3bc8wfADwBmZ2cPcJuD47iJ3fbqxjRsaubO7+/Vf6GX5fPcZJy76zVkEaKKhCiIeL5HWle4tVLm9kq7hN6n2LCZTgfU2u8ubuH78K3LOT5bq/POtlnbbLncXa9zvkdKXPvnh8tbbDYsRBHY7oO7c777zcMw6bL7tT98891lJFFgLBbBtL2A9tn3KNRazIzpnE3HkES23QK90e8eXzmXwnZ9Lk/EEAX41YMqEUXg9avjIadMO7sJCLn2NSWoZ+gc6872lpt1i4rlIArgeT4pXWF+vRa2CR1m/RyGNVxqtLBdn5S+nbnlOGSjAZ2y4/pUTZukJpPUVKbSQaOZ6UyU2WwsvIbn+3yxUQ87h92YSYVxhWzsUU1Ir8bxwyhtO+m3S40WHy5vsVEP4imf5qtEFJHxpEZElohrKht1g5JhkxYENhotptNRvnl5PHRxnsvoFOutMCD+3GTQQnOnJX9zfp3FUpO0rjKmB32nfVxiEZEXz2eomfaxEM7B4BZByvf9qiAIfxv4Z77v/31BEA5qEawCq77vv7P9+w8JBEEXfN//XeB3AV5++eVjrbE4aFP0g+IwaCf6Ya/Dr5f2Op2JYtout5bLWK7LWEwmpioUahZfFOskIzKvXRojIsvgBdWedx5WSGoyz0+lyMZU0tEIn+Yr/Pn9TTwfNFnkQblJoWZyr1APGRw7M0mqhrPvxtmJYQul9mt/2HYvtLuCTWdibNRNUlGFi7l4WLi2l9+23z06g8iaIvHls0nAx0cINfbbq+XwXbW1YNNxSGpyV6vHzvVi2i4fLm9xfkxnbNua8iHgrElE9l1DOwXpwkaDTEzl3nIp9NcPG7TMxlQUSQhrMqKyTGO7Dedzk8ldh1y/NdoO7AKkoypXJkXOpLSu93tQpa3TBdiuCg+6q6k8N5kIO6tZjsevXx1ncdMgEZUpNixsxycii1w7lwr6IdhuGNh/VEgGKU0NA+Jtt1znfL91Z40zCQ1dlZhMaXyar1Izbd7+wkQWH7GdHgcGFQSyIAhTwH8A/I+Pc0Pf99cEQVgRBOGq7/vzwG8AnzzONQ8bwx7Mh8FD1BmELTVaoevmsDKToPfh188lJRC4Rjzf42HZZLNuockCmirx0WqVGzMpprMBHcOXz6bIxR/FOLIxldcu5fjRR3lkUSChK2EFcJvB8Y1Xz++ZSbKXP7s93z+9u0FEFnl+KrnNO++zUKzzf/6kyreujPdsPrJ3+8PurmCXx2OUm0EOfcVoBZlB4zFe324v2Av7rZ393mf7XaWjKpcnYtxdr5PU1F3unbaFlC8brGw1EQWIa3LoivB8f1+tuJcg/TRfwfZ8zqZ1kpqC6bhDBy2vT6e5t1bjk7UaLdul3GxRbNhcHI8yO6bv6l88iPDstw/3Wt/99mU7hrGy1WRxs0GhFmjyv3XtDIokMb9eZTYbJRtT+cq5JJoik45GWNyss1hsoEoCST1CLhYhclbEww/J9pJaYMnF1GBPRWSR+naKeHu+Xc+nUDNY3TJYq5h8FT+kDAEQ2mSMx+UXYnBB8D8DfwT8W9/33xUE4SJw7zHu+3eAf7GdMXQf+JuPca0jwaAumcOiidjvOv3K5QehdtjrYNrL8tEUiT9f2MT1fQRRYCKukYyoYTbK7JjOe0slfJ/wQO6MC2w2La5MxHFdWCjXadouIvDOwiZvvHq+yyJpp961D7C9hMAj8rZg09xaqTA3FmVxs4mmCAgIPflg9nqv2ZiKZXtdXcEc3+fSeJTXLuX6drbqhc5DuhePVL8Daue7OpvW+a09/NztuYjIIgJCSF7X5rLZTytu95O4t16nZtkkthvUbzVapDSFh02DihXQK5Sbgyf2TaV1vv3cOHcLdcq2S1RTuJbWEAWRrYbNhfHerLIHEZ7DZm9994VJbs6v88lalXLTJqmpKJLIetXi//lglfPZGCld5mHZYKNmoUoi8YjMdDbKjZkMGzULORfna7MBHUWpYfHhcpmNmkUurlI2Wpn5ekwAACAASURBVNTNIFMvoO4W8H0hrGrealq8t7SF5/noikTVsPnoQYWsrlK3HEzHYyYb5fmp5IE6zR0UAwkC3/f/APiDjt/vA3/1oDf1ff8W8PJB//404bAyjPa6Duxe1O1ep9PZ6L4CqN/ht98GLFTXSWoKl3Ix5tdrNFsuE8nINg1Fk/WaSVKTuXomyS8XSmFcICJL1EyHsahK3XJZKjYwbBfX8/F9j41ai3zZOJBZ3zlPSX2b2kGReH95i+mMDr6AKHjcKwxOkQHtVL51rkzEKdQMCjUTSRT4nb94sSd76X5Cud3Yftg6i2FiQu25eH4qGQZT270SLuTi+7rW7m/UWSkZRFU51GTXaxa6IvCgbIaMmMmIzKdr9a704X7otNZ0ReTyeByEIOaRS6icTetdtQc7UW4GlBgLG42BiA9h+AyxOw+DPs8Ny6VmOqiyyGQyQr5iULUcZsYecVht1FtkokoXNfpkUgsVnmwsEqbmJjWVraZDRBYDDV/wMVt+KJj/7G4h7Emc0BUEAeqWTdWweVg2iEW213VE4dZKha9OJ09XHYEgCP878L8ABvAWQe3N3/V9/58f4dieCBxWhtFe1+klJEr1QEA8fzYVfgbDC6Bem6i9meuWTaXZwmgFJGlVw0aSAm17q2GTiUVC3+rXL47xab7Cxw8qfOvKOF+/OIYiwb98d4Wq5RJXZTzPpWQ4KKLDP/rJPf7aSzPspCjeLxbTOU9zY3FurZRxXIcvNmoUa1ZQ1eq5xCIqsggN2+XzQp3ffH4yzP/vNw9todjZFWxnWun9jTrLpSZXJ5Ocy+h9hfI/+fkCEUlAU+Uwbz6hyQPXWQyC9lyIghAGU6umi4/f1yLtFGK3lstEIxJj2/Opb3Pglxo2r15Ihz58w3aIDdABrtNaM1oOa1ULfIErk0FefNuf3u/vHM9jpWTsYBU9GAnjXvupYTnkKyaaLGK5HltNH893sVoBN1DbNagrMn7UZ3nL4D//zpWusXbSp3em5ubLBv/wT+/Rcl2SqsLVyWToqtust2g5HvHtNGvLdRG3rVddlpmMayR1mcr2tefXan3ZaA8bg7qG/h3f9/87QRD+fYJg718DfgI884LgsDKMhg3qtv2ujueHB006qj52imvnZr48kcByXH52bxMI6J9rRlDgk4oq/IUL2XB8i5t1Kk2bjYbF/Y2AyuCFqRSut4wqihgtJ2ipqEpMpSLcLza4vVoZuGK7fYDdeVDpckPNjUX508/W8TyomS0cJ8jT932HRsshGVVoWA4fLJWw3f4HJOydmdSek6rhIG3TW8QicvD8O4Sy7fqslJpkdJXnzwbkY20Nb9A6i0HQuWbaGuqn+SqW84gSpPN5drofVVlkpWQQ2ebUMR2XtK5SqFn4ePi+GPaY+Op0ak9209ur5S467qbtocoiqiSxUbe4NJ7AdIKEgJ0IC+TW60RVuYsG/MpEIhRAh5UhFo8EVBVl0yYiiSgiNFvQ8jxMx+V+oU7TcYjKMnFdCnmr2lAkoasJ0U76629vU1XvvPdYXA37d8dVCQkBy/XwgUvjMVRF2mYf9SjUDGIR5dg6Dw4qCNpP9NvAm77vl4SdtIJPOP5/7t4sxpIsve/7nXNiv3vuWXtVr9PTy8x0k8OZoYbi2DQJUoItQ5BA2DBgESAgC7bfDAM2INgPBvzoB/vBkG0ahEwYtiyRFAeCKJKjGc6+dE+vtW9ZlevNu8cecY4fTuStpZepWVoieYBCdqEr40bEjfi+833ff/lJB74/K4TRjzPUHcUFe7MMT8llSf8wQefHXY8rVW52A04UH4/mJedWwkZjxaCk5Le+cJ69WUZa1EvInDZWD98SalLyquYrV44otUYpSW0Ma21r3C0kFJXFbe/Nso9sFZyc30kAe/lM/5E21N405VQ/4nQ/ZHeacfc4xncUi7xCKUHo2jL9YFHw+YfE236c9XBFFhfV0oTHave0uXG0aMr7jJZvPYRDR5DV9RKBBJZx+uN6HJ98Nx9kxHLC+h0tJkySgr1pxqDl8UvPbixnJCeJ9nEV0lFcNAQpzVv3pnhqTl4bIk/RCxwWmfVrPpHddpX4QCjjw9yLg1nGVtfjjZ0Jkkax02iS3JCWFVpDP3w/lv5r144sOGGccGElIkQtZcBPkuQHzdD+3x/cY63lMk7KpXrsSTvpw8iVay2Xw3kGGKpKI7GSGe3ARRvN/UlGXtQYAcKAEZK/+fLmQ9d6l1rbZ3mjE/JwgXPyXb1xd8z374zRBlq+ous7SCURQFpUzJOCRS6QwDgpKCv7jK21fbqhY+dfUvIrL/ybcx580kTwh0KIy9jW0H8mhFgHso/vtP7Nrp9m4Puzgn7+OEPd9/amDCIfTwmyqiZw1SMEnR91rSdBX2CYJCXv7c9ZbXk8t9VlFBfM0mqJ4rH6Oi6zrOSXnt1gnpXWAKV50W4NFwSu4PL+gtEiZ9DyuDuMuX445/QgRCCYZRVxWrLagftTjTaGl051H3nJPyoJ//7r9/j27TFJYeWOT3V9bo9S/r8f3MN3FX/t6TV2xinPbHasRLLWVFqz1QsbXqZBGJ748x5fVm4Crt1ZcPc4wZGS04OASVwySSqyorYG5aUmKSob9IxAaU1aVpSV5t444WhR8ItPrfJnVw6JcyvdUdWafsvn733u/WSth1EmH2TE8sqZ3hJZchyXuEriK7kUbpskJb/33R0+e3GFtbbP63cnzNKKSmtuDRP6kctokXMwz616azeweBVjeG9vzrmViJbnsMgtM/nhzc3eJOWfvb7D7313h7zS9AKHWlsJjkHLpTZwbhCRFhVZWXFvnOIIwSQtlrOGk+vzlEQIqEvNt26OsH4OilP9YJkkH2+PlrXh1lHMzjEoKR+5N9cO7+EIw5WDBceLgtW2x/ObnaUhzqW1DgezDFdKpLKqov3AIVBwf5pzDLjSEsZcaQEJlmey84BnUtVcO1yw2fX4nW/coh963BklRJ7g+3cm3Bkl6GZ+VQIrocszG22OFgWBp1AC9mYZRQWesryJSVoghCUZnl8LPhKd9rNewpgngyg1jOKZMaYWQrSAjjFm/2M9u2a99tpr5nvf+97HdvwTgbHH2Ymhp55ot/rTQkefZD38Oe/sTnnpdA8p5Ps8Yv/+Lz/9kcc4SXgHMyujezDN2B6EnB2EKKlwpJXjVUJwe5QgjNV7P9UP+KVnNxkuMt66P+WTp3oIDN++OUZIq+2+1QnoRVYobJwWbPdCtLYm9beOY4w2nB6EOEoReZL1ts+g5bHZCTizEn2gZeUP7475b/7p26y0Xdqewyixg7WXz3TpRz6uEkxTe/2uI3n9zpgbwwVFg/X2HUHLtwYrf/3ZDQJXLne6H2aR+fA8YJqW3BoumKUVF1Zb+I7k6uGCqrZaM6HncPc4QQg43Q8JXHtvfVdxfiUi8hWX9+cEjmK97SGkZG+SsDNKcV2JrxRnVwJc6fAffnqb0vAQpn/O4bzg5lGMo05UVyW+I3lms83l/RnPb3Upa8MfvbkLwnoKn+6HvHi6z796b5/houAXLq1wYbXN7WHM3jTl1jCm5Ss8JbhxFHMcFyhhB5ivnO4yTivqWtNvefRCd6nfczI0P4Ff/umVQ+6NEgJXUtSGvLKyCBttH9+RzPIKVwr6kceZQUTgKp7d7LAoKtZaLm/vzvCU5FQ/5O37U3bGKdO0xFeCXuTRjzxe2Orwt187uySOnUCUf3B3RFbWvLs7p+1bhJcVSXSJc8vGfmG7B8KwO844jnM2OgHnV1v0I4evXTumNppe4HJqELIzSvj+nTG1th4XtbbV72bXY60dcKof8ubOGCMElQZXWnJcUtX0Ao8z/YDQc3lvb2p7/q5DWpZMsgpP2uvpBA5FZVDCME4spLTU9rMC11YMlTH0Q5dPnRnw85dWf+qYIoT4vjHmRwJznnRYHAH/ADiHZfueAp4D/vlPdHZ/wdZPOvD9uBzGPqodsN0PH0lcKy3bb34SFuLJrqqsNV+/fozvOHiuZJGV7E0l2z0fRyoOpynXjhJ8R1DVGm1guChIcrvz3eo98KsNPIXWhourLYQQCCEotLWPjPOK86sRaWFdmI4WjVSCp1BNAI88xe1RwmYv+ECbxy+/vcdK28V3HPLKcDjPyauaH96f8cvPbvCJ7S7fvmWDwt4kYW+aIoUkdDXztKJw7A5uELl87/aYF0932Op9uEnLBw0ux3FBXhnujBKe2WxzbjXiyv6cw1nG81tdtpr7tj/PECajbF7srNSsdwKe3exgjMCRgqLS3BtnVFrTc1xqDffHtqX0v3z1Jv/xZ88vbUP/6O0DXtjuNHacklvDmAurEUVdL3WdLq61efPeFEcJBBKE4Yc7E/ZnGTePYjCGr10d8nU5ZCVyubw3Z9bIU6RFRZxpPFfgu4o4r/j+3QntRmrhzIrDr35ye1kFbjw0C9gdp8ya709KRV5Zcx3PEUxSe/zNjs8srRBCMIwLvvjMGi3fegLsHAv2ZylVrbl+tMARWGVPrZmVhvOrLba6Po4SHzgb2p+m3Dm2EtadwGEQufQjlx8eJ2x2faQURL4l4k2zknFqVUnzSnP7OOWVMx3euDfjoLFYtc5vmqyobRUpoK41944zpnHJvUnCLKmotablK8YaqrK2xDHf4cpBjNbGqpWiCT2HrLLOc3VTeS/yknODFllZk5YVAkFR2+TZj1yqSjNNK5KsZrg4YHea8rmnVt9nw/lxrCdtDf0fWE2gzzd/v4eFk/6VSAQ/6cD3R0FHH68WtrvBsl/7YZn+R7UDfuWFzR85l/iwKuUk4b1xd0JtDJ1A0fJdksIqYNrArDmOS0JHELiKu4sCu1cp+NbNnF7o8ulzAyZJwe2hNVm5vD/jhe0u06xilpXM05K0rChqwwunuiSF5sygxVY35GiR4yhJ13cIfUXgObhScnuYNLjsR+Wobw0Tzq9EXDuMGSc5eVkTNCSdaVoAgp+7MOCt+1OmWUU/tFC/UVoRejVprtmfZriOYjVy+cbNY37jpVOUzYD74WoKLO3/1nBhd+FScmYlxHUULV+gpOTOsT2fXuAwSxVCQOi5VLVGCVBK8fSK3f0ezjPKYc1T620urnX41s1jjuYZeWX5FJPUYvTbgUNaCNsvbjwEhvOCQehyvChoedZRy3cU98cZz211iPOKzW7Alf0ZoSc5M4i4NYwbDkXJjcMabWpAYpoq4/rhnFoLpDAs0oq8su5bJ57PWVGTl5Kiqmn5LjvjhFFc0I9cbhwuOJxZb1/RDDkXhb13ZaWpjEHSdKoEvHphQKAU37gxJPQkjrRibOOgwFWCm8OYbuDgKcU8L3jvMOH5rQ6bXSt4OGj5bHYD3t2bsdULH5kNbXY8Xr875nBe4EiBNjBJK5LCVmnHi5yLq1a+/XCe0fIcpqpkuCiQ0m5AjhcOr57vk5WaeWbfL1dAgm3NGKDUBg1Msxq3qqlqy9qepDYBVBocAYuiRiGZFyVVI1g3TXIqbRACNKAEuErZdlBZkZcGyywHR0KS12RlTa3BdwW+ktw4ipmlJb/64tbHzid40kTwlDHm7wohfhPAGJOKv0LT4ocDa17VXNmfM1wUfP6p1Y/ETt86ipllBYvc2hpeWG3Tj9xHhluV1hzOcr5+bcjeNOOLz67xwqneh1YPHyQxnBY1w3nBM5vtpZjWh80THv/c1+9O+MqVI37z584uE948L+n5LmVtiDxJVgqMMcyyiryU5JVmsxuyO0txpLAPfW0ld6WEP3lvHw14UqAcRVFrLh/MWY1cZnnNSsvlYF5T1pqvXT3CYBPWi6d6aOCl0z2yUuM7dgeblTXz3LZgvnptSF7WDEKP3UlKnFcsPEXoClJHkUpNqTWbXZ+1js9b98c40gY7gNcurNCLfL53e8hwUVLpkklcsTtO2Z+mCAz/d1ax3vE5uxLRDVwmacE4qfjh3THfvDlireXbXbiAW8MYV0mKWvPpU132pxmztOTeJENJ667lSElR2XtTVJpe6CGl4LULA16/O+GNnSnv7M05mmV0I5fAVSyyiq4nqYWh0gZT17R8xZ9cPmSj43N1f7Zkd58ZRJS1nQWlRc1ax7bfBqHLl9/ZZy3yObNiVTx3pxlFbfBdWIlCRkmBxrDIKorK4DmCujaktQ1CLlBpQ13UIARaayqt8BxFreHt+xOb+NOCrZ4FEXQDF6FBaMPR3MohpIXd+bcDh1PdgKo2fPPukFlWURxZL+zrhwtWIo+krPCU4vxKyN2Rtd5c5CVv3Z+y3vH49LkBoSf59s1jLqxZNdoyzmn7Du/sTvnza0cEjiR0FZ6SJGWNEhAb8JpNwjMbNjylhaasKrKi5HCWkRb2ux8nBdcPEn77ixd5Z29GN7DVEqamaBLAyaoB3RgUypO/N8+bASZx2VTGje2AgLjUoMFxRMOfscPpaW6nyx3XvmeVscecptUyYRgDQghcpYizmsu7Cza7H+/Q+EkTQSGECGk0UYUQTwEfjxTnv4V1Mqj9ypUDvnnTygV/4ek1fEd+aKtnb2JhYEqyRJG8sTPhmY022/1gydq8fhhbwwltCDzJ9++MOdUPl3C/xzP9B0kMB658BEFxcs4flKBOPvfNnQmLrKY0NVrD//b1W/z7L2/zlVsj7hzHCCCr9LLcPpxl1v/Xdwg9iRHWZL2qNbWBwIF+6C6RQhdXWwzjgkoXnO4HJHnN7XHK6V5I4EomSUFcaYwEYwyjRcG3bx1zfiVid5JiDHRCRZzX7E5T1tset4YLkqImzS357M8uH3G673H9KMGRcKpndVl2xymXVtuMFzmv70zoRR4vneriKMnl/TkvnJKM4oLDeUZaaDAQFyVlbVtW03TG9UMrG3FhpY1Skmc3W3z57T1WWx5CQst1OVrkHM4zFlmJ60iSrGJRVBhAa9jqBUgpOJxnJGXJSuQThU6jt2QNy4/nOZ3QBc1SWqHlK2apbSMoIdAa2183musHc/Ki4nBeAobVlscsKxnHBZFrzW/ysuLa0QLfUfRDl3le8t5+yfObXf7mK9v809fvE7oSbQSne4rbo5Ta2MCfFhaueDIZLA34CFxHoLWhFtD2HU73QjxHcP1owdMbHXqhuxTE60cOs6LCYEgzjXI1WoPnCLqBy1bP442dMbWGwBHktaEuNMbYpO8rydqKz7WDOcdxyfHCAhdqrSm14fYw5uJ6m+Ok4Fe3thnF+ZIwt972GC4KXEfQkdhnvKpJNUS+pO05BK7gxjBhtMjZnaXMMgtbvrjWYpFXXN6f0w1dLqxGvLs3Y6XlcTBLqbRt03yQsoPABnljxXjtTl5Y0ySB9ZpwHIlpKqO6tp4ZVW3wFEghKGqD0xyj1JpW4JCWFVXFMvH4jsBz7BwrKXIcKbjfkC8/zvWkieAfYolkZ4UQ/xj4Ao2fwF+Vtd0PGbR8fvm5jUdaRGA9cwcNXv6k1fLDexOe3Wxz/TBeqkRmZc2Vgxm/9uIWf3blkMNZTuhakk5a1/QDl0lWLtsgHzSH+CCJ4azUdHz3idpVo7jg5tGCg3lO23dpKZei0lw/nPP7b8Knzw5wpOGd3Tl5VeFKK3kcupK//8VLfOfOmMhz2J1mJLk1RTe1LZ+LWtuHuNY2mQiJ69jGkZCCJK+4P05wXQUILm1Elq06y1lru0glqLTh+uGC1ZbLSrtNP1SkZc3uJGeeFQ2Uz6HWhr1ZynCe8/xWi7jQHCwK1ts+r53vM01rvnZ9SOgqXj5jmZtt3+VeHHN5b8Y4LilKG/SEsCgTre1/S2ND4a1hjNbwKy9scmYQ8ea9Cb9waY03700pdc2t4cK2TWqLLx8nBQ62bdENHSpj6LouZ/oR9yYJZ1ZarLVtYL4xXDCJc7b6IWcHEYeznJbvUNW22oo8h+GiQGCrJa01eSVpeQ6j2Fo9FrVmnFqM+SD0mOcllTH8wQ/3AbMc5OaV5tJ6a+mUtt72KLWhqDQCicQGolp/sLp9XttBq6Mk6x2P1y6sEBcW86+EWFafJ+3TSVJxcTXiYJoS+MoilTxLSHt+s8NhnNsBqCPxHY+iNuxN04Zx64CAqmGYp1VN5FvbRyEEoSOZZiVHs4yXT/fwHWnbQNOMCs3eJKXtK5K8YpSUCCFQUiCEwVeK1Y7PL1wY8LUbx+zOMhaZbblkZd38O0kncDndD0HAP/nBPX7+wgo/vDdls+tza5jQbOofSZgnz5E2D6qCyoA02M1OrRlEHpsdn0VhvZTTUjOILLz1JINEvqKsDV7zLgwin0lSUFX2k9LSkJY1jrR8hVIb4qJiuxs8cSz7SdaTSkz8sRDiB8AvYO/Rf2mMGX6sZ/ZvYX3Q0DivNN+4ccyXnt9YDvG+cuWIUVzw1HrEhdWISWK9XTu+Sze0HqQrLY/X707Y6NjjnSgw9gKHef6orPDD68ShapFVTNKStbZH6DqcHrSfiJ+w0vK4cRTT8hw8xxJhhLCB4Giec3Yl4uyKlfz908uHCCH47PkeG52Q+1PrchW4Fj0UeRbLrY2g7dl++DyzO+JRUhB5DtoIbgxLjLYPd+Tbzx0vcivlLCTbvZAzDQv33GpEVlg44Y3DGCPg/CBCIqi0JvAUi7TCVWrZtjlcFPxXv/r8Upqh5Tt888aQQejx8pneMnGfXQlpe5L39ueU2oAQCGNs26h5u5UUKGXNdlZaHt3GX3ZvmnG8KLiyP+PiWsS7u1NCV7IoNAYLMZRSkxWa2hjyytDD4CiYFrZtM0sKFlmJFHB/kpEUFZ97KqIduLQDl/WOz7t7Uza7Ab/+0hb/+uqQcZKja8NxoumGDv3I5f7ESndkZY2UkjivGc4zhouC+6OEuKw41Y+s2Xnbw1FySfb61Lk+v/WLl/jq9SFXdqe8uTtlHJe4SnykxYlBI6WD7wqOFjllralrw6fO9ZbAhTujhGc328yykkVe0QpcPnm6ReQ5GGOYpCVnVlo4juS5jQ7fuT0iq+pl4HOk4GwzP7lznFDWBqMN652ASWJ3+Yu8xpVwHBc8v9XhTy8fcPMoYavrWwnrvEI33guisVIdJ/Y7cR3BuUHINKtouYphrRFSsNn2GMYFd49THCVZiRyGc9uaiTzrMxE6krK0G52TBCAeuT82MEthKBrugNc8U0WznY/ziltFzXrHb5j19lm8P7HzLiUAJEWpaZTFifP6kTbUslLTtmXX8gRfen7dQrY/8s3/6daPY0wTAOPmd14QQmCM+erHc1o/m/Wkw9qTf/edW8dMk5JBy2WrG3FhLeLK/oy1trckWF0/jJsBm+2pT9OaT53ts9LyHkHuvHKmz1euHDFJrUFHJ3I4PMhZibzGArJ8X2C35zHl2c0Oh7McYxKO5jkvnbbwtSeBkb1ypk9Za4q6bkpMTV7pxmXswaM9SSpe2O5Rar202dsZJdw8sj3xUVKy1Q0pK0OhNXllyCpLClppueSlbWtM0oq8tC+U60iO5jmRpyi04WCa02+5nBmElLXBUYJO4LDILCP6hY0OVV1zf5xx69i2hVq+wndVo/tv2wj90GNvlj0yF8krzSdP23bQyQochVKSc4OIotIcLjJELWyZblu2GGOoa1hoq99f17A/S3luq8sXnl7j2uGC6d6corI9c88RzQBQEChFIQzTtGKe14wWBSvtnHFa4iDohQ7dyCNt+ARFVXN3lLDds4JtTuNrPGi5jGOLslmLPC4fLkiKCq01T290CD2HsjJcPZjRCRR705RJXFhXrMoG6HlWMoh80rxmJXKJi5pfe3H7Ebjz/XHKWsu3s4W8ZpZVVE3UEYCStsUFsNkJMMCd45Q7w4SVtsdqy+Or14+5M0o4txqx3fO5ejBnvCi4N7G8gONFjmwLpBS4SvCdW8ekZU3kqYaYJeiHHmWlKbR9Z86tRESew1U5a+6DBVpoY5hnBaVS9FuC82ttjuOCUtcMFzmlNpztR+yMLVzXkZLa2GFry3EIXYd745Ss0pRVbVtdteEgKygrjRHgKusncXeccE60+My5PoezgllekmvNCe5OY3f+Stq/VFi+gu8IRFGhlMR1BHlhljcxLe3wPU5L3rg35VTPZ3ecgLEVRH3SU3xoPZycm4965P+tRB7PbHY/lNX9s1pPCh/9H4G/C7zDg3M1wF/YRPA4tPP+OOUP39jj1Qt9zgyi94mA2YEOzPPKlmZCcjDPmKYF/+4nLLPwhGAVuHY4ZowtSW8N57iq+0hg3+6H/ObPneX3vnuXwwaidurZVfamGd3Q+UDXqIdRSOdWWrx2YeV9fIaPktY9wb4rYQfZoa841fU51Wtx5WBGnFf87rduYwwUVc351RarbVuRjOKCqwdzamNIMoteyXKrLeQ2MLeT1grYVtAkLZbMSteR1No0uyYbaJKy4pwfUNcwTayI2/4059rBgtBT1Fpz6zjBdxSrbY94lJIWdm4RFxVSCJ7ZaHOqHy1lK06u+6XTXdLSzmDAzlEmaUFeaYZxjiNB14a0fHQfLIR9uVUDlzmaZ7iOxFOC++MEgyErNPcnKY6UbPcCbmU1ukkmxhhqDVWlWRiIi4qWp4jrmlleoWYZK5HPILJKlFf353hKkhWaeV6y3vb49LnTgOHT5/pcP4y5MLDM2/15zlv3Jrx4usu1SUxaaCQls7wiLmxyr2rL68hKTVpUzYDatpweliPYm2U8u9lmd5rSCVyksJo2x0lp+92SZRTyHVgUFWmhmzafhb7ePU45t9qiNlDUhmlasN0LiPOalquYFxX705SDecZqo6eTFoaXT3cZxgVZrZklBfPMSnn3fIeyqrl9HNuNgZQ8tR6xP88bHL2F1271Ap7d6nJ3FOO7ik+d6fHW7pyNto9qzHeq2oCpSHJ7HCEERWWI85K00qS5rURCT7LISxDgGCgrTS1tv38U50v0k9aGJK/xPEle6uXQVmsQEjbarg3oGgLXwVcwLTQYbQlhzbvhSklWazpKcG+cUhmzrCwE72/NuZLlJuXhJCCxn1sawVcuH/JLz63zJm3x6wAAIABJREFUca4nrQj+A+A5Y8xfmgHx49DOo0VGv+UwnBecW2ktmZf/859dpxe5zLKKQeTx4qk+98Yx+4ucp9ZahK7Ed+w+YZ6XS432rV7IhdUWN48sSeeVs4P3a6WfG7DRDR4J3H/ntXMfqmfz1atHSEEzyGo3DMsHc4QP4y2cJLMT7Pt2PyTJayLfIS40R7OUaZyTVIZBqFlpKY7mlkH8tz5zGoDbx1YQrKgMpxvi0tv3Z+SVhZZGnkNR2V3JJClxlEAJSRAI8sryC7q+Q+AqKm24sBaCMShHMklylBRcXO/Q9hUGuyu7djAj9Kxzle8onttoszvLyCrDs+ttepGLlBLfFc3OOkQKwbdvHrMzTgFDoCT3ciuHsN7x+cRWm+NFzo3DhYV2upCX9iWzAVBajLmnyCtNXmsi3yEta6ZZRVrUxHlBXhpKXRPnFboZtNYaIs/CSOeZzYCBK+3w2EBdW9x40nA8BDYI3D1OKGqNKwX3qpp//K3bjJKS0JUMWj6dwKHShu2Oz1GcszvNQRsGbZedUWrvo7BVidEG6QjKuiYtLPu1qGu+9Pz6I8/VibKoJyRIi0I5PQiYZuWDACfsbtp1BPOsIvQUrlQgLNKlNjZ5ZaXfCLFVvLM7w5WS9Y5PPLLXuSgq7hwv8B2Hp9barLYDAldx83BBO3CJfGUZyllFUtYsiopPn+mTNuWJExeMk5Kq1rQ9h0vrreWQvOO7uFLiK/uc3T5OWGSF3cxoMFXNSmRtVBdZScsP2OoGXD+c47kSRyobiIWtyIpasxI6+K5kntfcOIhp+5L9WUGl7UxJNZBUgcHzJGcHIacHUZNYFONFzu3jGE9Brm1VCbba0EbjKDu7maYVjhBUHxGnpADTzB4eThQGG5y10Y1V6cfrTfCkieAmFmn2lyYRPCwLMM9L7owSzg/CZX/+ZAc8jAue3mhz6yjm7iihH7r0Q4duaM1VbhwuHhjIew6TtMAYsVQVdJXkU+f6H8pA/ighswdSDzBsaP4fpSv/gBBmeGNnzDyrcKTg6sGMV8+vLEW7Vts+vqtswJZw/TBmrRNYS8S84mBR0g8dtIHdScr51RZH87zRwA9Yb9s/WVlz43BOXJT4rsXnZ0VNqUFKC4lrKYetXovhPG965zVJUeFI2OyHnOoGPLfV4XBucdWBq3j1/IB5ZtEbm44i8hSrLWsE8rmnHf7ozT0OFzm3RwmhK7l6IPhrz6xR1pbBHHqSlZbLtYOFZaYOQlwpScuK13emFtKprOCZFtDyDLO0RDY7QbAm5GBbBydaNb5j5RkWucGRlvpf1A+ghAqrSyOFwJEWS17WBiltoqmM3eEZYJaWzPMahW1xBY5imhcUpeFI5AgBxxoWWcX5tRYC28Pf6ASstj0WnkIbgxSCe6Nk2dIJfIVonM3yStPyHNY7AZ9/6tEd47S53jMrITeHVn/JdxWesvj30FG0fHuPRnEB2iCEhWJ6jqSsDNpAklfkVcXXbwxZZCWzpOTCWosXT/c51Y+4M4q5N0qpa8Mgcum1XG4OFygJg8hhGBccLwpGokRikWdPrbUJPJfntyOuH87JS83T620urre4P065P8mWTOU4txpOUghGcU7LUyS5pG4ymRaGeaHp+oozK3Ye4wiBFJKqNuRlhVJ25iUFeK7DWse3lbarWO143BkuiHNrr6mahIExOMoKzq1EduC90fZp+YrRImOS2IRamwc8Ac+RFBVN0rbPmW7qgQ/D2ucPaRWd/NZJQig0HM8Ltrs+5kOP8LNZT5oIEuANIcSf8FAyMMb8Fx/LWf0MlsDw3dsT+qFHN3DxpODy/pznt7qA3QFLCettS2KyipX24UsKzSLPuT9OlyJWP7w3oRvaRLDV9bl5tOC7t4uGfn/2R57PBwX+E9nib96weOvnNtvcPk4f0ZVfafustVz+r2/f4e37JwYsKaFnBefSsuLN+1NePN1bViwAay2/MdVQ7E1zIt8h8hz6LZ+iIRKFrt0hnXARkrxirf0AndDyXFbaPiq1CSrN9TIgGQ1SCcraas0rKTicZWSlxnUaOeNFzjgpOI4LfuPl7eWxT3xh+5HLpfUWAkla1lxYi1jkFVu9gHleMmj6vvuznJuHVjcm9CSh67A7Sgk9h09sdyyhSUPXcRnOCqraImYiXyEM5JXd0fuOJY55jiTJK7QAKWq0scM7baAoNRo7rFNNG6nlSpLc8gR6gUULVbWVVSi1wRN2p26a+6KNZp5rnAZUnpQVi6JGCoNSgqzQOI4N/Vll2J3mnO4HSCX50nMb7M9SPPXApc5iygVKgSMkeVXbYXfb5bOXVtjohPzw3pSNbrDcdPRDzyqluoILaxE3DubsTzN6oc9626XUkJQagUYbDRKK0g7p7WakotQWMfXurj22J2077c4o4ewgZKsfEToOpwYh7UCRFDUCie8Y7oxipmnFOLGzGN+RZLUhLjLazTPqSGF1gLasqN3uKGNRlOyMUi7vz7m01qIMXILm+5pntQUkeA5CNP4W0g7y1zo+xgiuHczJG4mIfmj1j1xlYbNKCjY6PittH99x+NxTA64eLBo4p0FJGqc7y6a3hDxDWmmeW2uz2vI4jgt2xhlK2aqJugncBrLqBCnl0AtdIlexM05xZL18b55knSQEtyH7XT2I2TlePPkBfoL1pIngD5o/f4mWaNAi9raudQKO5gV5VaONsU5DUvLq+T5fuzak4zvMMov+aXm2Z30CBX14V/9BCoSPv4SPr8dbOieB/0RWodKGXgPLe1hXfpFVrEQegWuRMr4j+dMrR1xYbS010wWS1cjjyv58aXgeetZUoxM4HM1z1tselTYWveHYoDJJC7Z7HT57aZVfe3GbvUnK//QnV5fD7azUtANF1EgPlJVGSknLFxS15SastXxCVzRqjlDV4HuKtueglERJa5+4P8v4wZ0J/94nt5aS1fOspNaar18bYk4o9nW9hNydW42W1wgzRonVbXmxkXqeNkkvcBTXDxc8vdEmcBRpYfkXSWElJjzHKp/6joVJLvISndkXPS+t+qS9jw+w39AgR4xNDpbti52nCIFCNHBMm4CEsVwLsLj8rLSwX9dRxE1/PK/skFoI2/s1xiyBBXlRkZU1620fJQVxVnJrGDNOCnylaHmSpKipKhh0FBtewLObbU71Qz5z7oHEyMOclIvrLQJXcbSw/JB26PFqNyQpa3xHcvPIfge+I1lrBRzMM3uOprIzEN0MS4W9F6O4YLXl8+q5PlcOFryzO2WzFzLNSpQUnOm3OJhbYljkKqZNu0di21Bpae+5q2gYwPDe/ozNjs+g5S7RaBgsd0Xrhqxpv+e1tk+/5Vm+S0OGE8KS43JtW0YY08x7JEYYhnFBN3B55UwPR0m+e2fCPKvY6MKvvrjJxbU2+7OcC6ttfEdxMM+YNDLZUoCrJEo0xpHN9zucZ1S1ZrsbYAwczBuElQajDRttD40gyWtcR+I2iarS+gNnBAJoeZaQiIHioX9Qa7sRqbTm+3cmHxbofibrSeGj/+fHehYfwzLAz19c4e4oZpaVrLRcfuPlLW4fJ484DZ1baXF5f0ZSaCptqIxhvWMD/AkU9OG1N8v47MXVR7gGb90f8w9//20iX1lo4Ivbj7haPT6vOAn8t4cJIDic58zTAlcpLqy1lubel/dnnFmJlr934kR1NLPDuazUpGXNZy+t8s7ujE+fi6z5R4PuOdOPGMUFn9jucnsYc3ecWr3+qkYY+OSp3nLAaIfb5/hHf36rETmTbHd9nt/uMrs1QghBZWoCKYg8D4OxraDaSkqvdzy6ocvuOGWuS0pt6IYOLc+hqjQ3hwt2RglXD+ZNi8b26Rd5TT9ywRjujhNcKXHaHoHzQDfp9MAaeztKNhottuyfpwVv7IwZJyXnV1tgNFvdiFO9gN/7zg5gse3jOKfSwvb3ha0QTnxhhZLoWhM4FrN9gtw4maUKAQZB4IGnJJ4r2Wh5XCkqJFYqooGA4zTVe1WBcAxFVVkxNG0/TWOxibKJCIHr0A8dlJTMsorDecYP7oz4wc6EwJWc6QXszwsqLfAcRS9Q/PyldWZpSeg69EOPr1494s5xjMEwiB6AB04c19bbAcakvLObE3mSZ9bbvHV/jjaw3nY5nBdo/eC88sr+dCT0A8misANQgc0IRgjO9H3e3Vvwr68eMU8LWr6Fj/78hRVuDmNe35kwSkpKbXe1niOpKr0k4s2zArDzt3FqZyVr7YCzK4objdCeUjWhaxFnu5OE13cmtr2jFMIYpBTM08omlyaxljVoXdPvuUjpMc8snPf1nRm//tIW/+nnz3PtMCYpKu4ex1w9mLM7zfAdqyZb1jWhq8hLyxtp+YrawDi25lAX1iKKptIwCFxH0vIdam2TXOQpPnm6D8LqU90dpXiu4pW1iHd2bYypGx6CfaWtjSXY4b2nJEX+/tLBU5LdafpjRL8ff31kIhBCvMVHTCmMMS//zM/oZ7RWWh5pUS93TGB3TedWW/zai9tLHfX39ubEmSb0JWcGIb3QBrm8qhlE79/hP841uHk058+vHaMkPL2xwjQr+V+/eovf/iLLZPD473QajPj+zO5A+pHLrNHn+YM37tMLXTrNrOKEzWl30jGugp1JQi/y2OoGPLfVxlWSXugyaHlkZb3UZt/uB3zmXJ8f3psiBNyfpMyqGqPh/ErAwSzjd795m9pYctIgcjnVC1hre5S1tv3Rts/nBNwZWftGX0kCV7E/z8nKishz6QUOl9bbKAn70wzfUay3fWptuD9JWW17dDyHP7u8z84kI3QUnVDhSKtAKiWErmKaWgZtrTUrbX9ZEbhKcWE1wncVd48TkqJmltqebi90GYQO7+5OOdOP+PzTq6y0fN7dnXHnOGGaVxgDoSsIXYeksIHHdRx0Q/3PK/uQ18ZgGrTICXPUVXZw2AutcF1tBJc2OjiO4p17U5KywiBo+Q5Fqal0TVUbosDKDt88XHAQF3jKMo/KBj7oOjYbWCMS6AcOW72Q47jAaENVGWLg/ErINLND69WOz9mVkGlqUUL/6r0DhgvbN49cxU6W8j98+V0+sd3l0nqb0z2fP758RK0N6y2f0JdcHyYMWg4H+xmjOKOo7MXWBpR6MEBeaVqLlS4omoE6wnD7aMHevKDtWa8HJb1m0Cy50mDmz/RDtNYcL0p7XKPtsFbYzym15vYoZhC5rDRSIsdxQVJq7h5bfZ2O75BX1vh+f5ohBPiuA2jb/muqNSUbnoyEQFpSnJV8scSxvNJobfjEtq0kMXN+cGdEktcEnrLghtrwzGabStME9oqOEuTa0PYkaalxtOb2ccILWx3SvOLeOGO15bDV9dmfZg2SSXNlf8ZzW13W2lbG4mTjWdXaJkVpCZha242UkraldDJvOlmqQd55jlwymj/O9aMqgr/R/PwHzc/fbX7+R9i5wV/Y9VHCbCd4/ec2OxzOU5KitAPVtsdqywdhcfbDuHyf1tDjAnXfv2OtDfuhRbgMIhvwv/z23jIRPP47F1bbfOvmMeM45+J6C4FLJ3AYx/alK7Xmuc0OVw5m3BsntH2XN3YmhJ7kwkqLu2N768+tRnYA/JiM8uNroxvwO9+4xYW1Nj/X9ulHDreP7XW/eX/C+dUWs7Ti1tGC2oglBPb28YK7xzHTpOQz5/uMY+tdME0LXAk5lmPf8iWHi4xxXFhobVU3Ms227TVNCiJPMc9rLjUqpTeOYrTRbPcCjuOCrZ41v5mmJcdpgec4XFqPwAimWcmFlRYvbHf4/Tf3ltffi1wcaRVLk1nG3eMFo0a4raxrPrHdpeW7fOf2iCQvySorTCaE3Tm2fUXgK9ajAIFgELncm6SWHJUUGG3bD73Q6uIUteHsIOKXnt3g5tGCd+7PcJUVVItci0Lq+A4IYXXoa007cBgmGSA5vRJS1TZARp6Vn661bdWd7of0I59ZVjKIPEqt8ZTgOC6IfBcpHD59bsDf+vQZDmcZ/90fvsuiKGn7ilqLhkluk+ksteinr9wa8dxmh7Mr0VKm4dZRzFFdWznuyjTSxw0s2NhOat20VItaM2hbpc9ZUpKXJ6NPg+9YItsLp3pkRc0kLbk9WlDXmo2ub4O0sBWAMQJX2R27END2HF49O6AVKA6nGXfH9p6X2h7bdSSVNszzkrSo7WC4+eTa6ughDbRchacUjrJaWWlZU9QGKcyy1w/QDhVv359Qabg9jMmKilIbRNkowpY11w9jBpFjkV7CWL0mJexQGEE3gLKquTG0ZM2WJxHCkJSWENYJHBwpqI2dvdxqJMwDJRiX9XI77RrTGAKdvJ32O5DGSlCfrBMUUV1b2Y9L660nD34/wfrIRGCMuQMghPiCMeYLD/2v/1oI8XXgv/84T+6nWR9m9ALwO9+4xSguWW/7XFxr8+r5Vf7lO3sczXNKrekEDr9wqYfbSOACDw16DcO45EzfJpijZke20XkQhHuB+0gp93hScpVgve1xbxzz9u6MQejR9hRnB4OlrtDZFUtCunIwJ3QlgWtxZo6j+NLzG+xNM751c8hmN6AXuh9oTfjwvTi/2uLV81bP/Qd3xoSuYhTnVukx8kmLmrfvL2j5in/y/R08V3K6H7He9knymrd2Zqx1PFYil9vDGN8R9CKXvKgZJYLtnk/caBMlecWt45hASVq+4iitqbRZ6hQpKVhkJXlVM89KViKPUQORHbQ8AlcS5wWj2FZEr53v88J2j9/77l06gcN2L6CqrbtT4FqkSztwuDfOEFKQl1Y47P54ynbXxXME0wyy0izhG4En8V2JMlY4TUkLySzKmriwL27oW1GzlqdwHGv+M04KRnHOJC04vxZy/XBhSU3NYD7OK7Z7IYPI5WCeM89K2p6L5yhCR3Fxu8vzmx16kYtBLL0l3tmdEbh2YlHUNUeLHF9ZTPqaEPQCl61uyB+/e8DeJGGS5hzNCzxHstb2CDxlZcARxIXddNTacPNoztEiY55VKNl8B0mJFDbwSvFgHiKEJT6hYZqUrLU8ZqnVaKo1CGH73oPIQ2NRdIfzjPW2TzqviYvKwk5r2Oj65GUNlbZ6O45CYK0wn93q0g4dhvOcw0XOesfnYJoth7vT2qrKdj3F3XGKpxSDlmIQ+ezPbFIVaDyl8D3VtPFsgC1rKyWRlhVg79t2J+DOKOHpjTZ3RwlJwy3JTU1WWfZ2bQo+e2FAZewcYGecUpT1suo4mOX0IxeJbakt0pLDOKeq7KzwpdN99qZWmvvOccz+NCXyHYyx8PPIUSRliTGCti8pKluNOo3URFnbyswi1R4wmw2Grufw6y+d+ukC4o9YTzosbgkhftEY8+cAQojPAx9vivoZrMehmydD21FsNWssTHPKp8728BzJ6Wa3d7K0MY/I755g99FFoyles9Zg/dvBg1s5zUo2H9IGeTwpCayw1ydP9RDCStB+786Y1XZK2290UIDTg5CsrLk3SRAIuoHkua02Ky2fbujyjRvHPL/VXVY8H+WF8HBVcoIueljYrqo1x3GB7wbM84p1z2d3koKxQ9KjRcHRwrZtes0LW1Y2wAPsTyxiqG6u15FWqqBoyt7RwiKzMCClbSsUWlPUBldWTXkvaAcez212SEvNv/OJjSUs91+8vUetDaste17DRYYrLSBgf5ZR14bVtkvkuay0Agy2lL83ya2InqExpbEKoVpb1EpeG/qBx1rH48ZhQlZWeI7AYKUdlA+ea9tC/cDhcJ7zj752k6SocBvZBMuLKIlchzSv8ZV9hW1LyWO95eF7il7o8XMXBvzmZ88vUWTGWAtLARwvctLcyidUlWXZVjWkeUXRDRktMhxH8s/f2qPd+CVXWjNLq8ZQyMpDB67i5tGcg1nKzaOY57Y6nO6HFDXMMtsaKypLhHoYvlg/1J7OSotm0gYcZcXQXKUoy4rjRY5G4KsEKeDGYYwUlhsiMYySnMh1CFyHrLLJ+uUzdmM1Skp6oQU+9EKPo0XOWstlEhdWEj2v6EdWH2ulHXAwL+iGDpFnVVvLSltQprCotWlSELqSyLcbhOE8x3clLd9lELq0fIWSkoNZxrWDOeOkWrKRTzoxDpZ4OM8r/vMvPcP/8/0d7o4TkkrTCVxcKYjzmnFSMktLDuc5q22P9bat4Fbb1oN7PM9AWoRTVmm0KZHSSkoXdU1eGqQ0uMbBcwyudHBdSZzVBI6x4nzNlyBo2nRKEnoOL2x3f7zg92OuJ00EvwX870KIptHGBPh7H88pfXzrZGi73rFJIPTs5d8eJu8zqAarHfKw/C5YKvyZlYjQU7xyps80yfmjt22v9sJay1rOxRV/59VHIaUPJ6V/8fYegavY7IV84/oRB/McTwpmaYUUdnA4iq1u+8X1FhfXW+9zULvS2Es+7oXwlSsHDBpd/4eZx49UJQ0fQjbOSQD3JwmbXY+yglprBIbdacbVgzl+400Qei6+ssJoi6ImLQzdwLYj8sqifQzgKQsfFUIwXuQYIdBG4ytFXluryqI29EMfjKHSMM9rVlsuAmvUst3zH6HVj+KCwJFc3p9xv5Gn7vguVVEjEZRGkxQ1gyjAcyTGCNZannWaEoJBINif543Im4WVjhLbiz69EpFVVutHKVBCcn6txSwrGC0KW82Ummc3Wtw6TqxomjYUDerIdyTTrOJ4UdD2FdpYqQZHCYySjNKST/ZDOqE1Zfnrj3kwf+XyIQfzjNGiQCqYpQV1zVLTBgTaGP7o7QPWW65lYrd8tM5Z5AZtNMNFSSd06AUBrhT8y3cOqGpN5EmqGq7sz5ikVrQuz0posPYP81kf7ELtH9n09I2xzPG8rOw5NfDjGw3be73jc3G9xTQtraR4o5NlxQM1nmPbZq+cHjDLrQ8CwCK32ltxWRM2gnxKCkDy2oUe3dBjnORkZePZUGpcRxLUEkcpWp5ACYey0igh+I2Xtnh2o7Oci6y2PDa6PnvjlOEi5/7EemrrylY4UtJoXBkCRzVCh1kjE1/zrVsjtLE8EoMhLSp6oYsUsMhrAscyoW8cLpBS4AjBVsvjViMkVBuD0IYCC1+tDUSNQF5RgXE0Pc9DayuIWOkaI8HRDycBRTtQvLs3ewSA8rNeT4oa+j7wihCii7W3nH5sZ/QxrpOh7YXVNm/sTIAK35EcLXLOrUagrYZLy3e4N7buR6M4RyC4tN5aSke3fIfrh3MOZzlbvYi/8dIW37454o2dCa+c6fHbX7z4kV/ayXlIIVDSkorSqiYpNU+vd1hpWcu7i2vtZTvrcb+E790Z8/R6i69ePcRgiT87o5g7o5T1TmBx3r2Aa/tz/vZrZx+pSk74EK+d77M/zRknOdO05OJqqwnUNTeOEmptB1xZYUiLglpDpgRZXuG5FjcOgpanaPuWyDacZxzOMzswdQWLoiZ0reYRwvY87YBMs9ryODsIWRQVR/NiqckzinNGcca8EeV75UwfAeyMUvZmKdOG2XwwT1FCMmh5CAF705RxUlBUNBaJDs9vtLl6FBNXti/uO7Y11/IV98cpoacYhB7Xh3PA+tOeDHTbvlWOPN0LrEfAYcwgshXg9aM5VW2YpiXzzOL6hdVnIM5KRrGFJweuQoSGawdzosCyrn/nG7fY7AZLnf3AtfdvtMjJCk3Z6CO5AotXNzDLKvqR4PZxwnOnOmAE6x0f15HcGycYBFudgEvrLaZphedKJmnJK2cH7E4Sbg0TsqpmNXDYyyo6vkPbt7DG8qFKQGB3y/BYhdAYs5wgqgLH8iiq2vbmXSV5fqvL0SzlzjhFGysm95lzA57Z7LLW8bh6MOfZzQ7XDmz//HCW0WpmYwKaBGGhoUmpubDq8Z987iKH84x39+a8c3/aJHKPTqDY7oUUlaasNf/tb7ywfOdeON1/hM0fuopTg4gvv70HRnN3lFJpO7+QypYFg5Z1ODv5HUdJvnBphXebmZg2phEelESenaqP05Ju49U8TnLmpf7/yXuzUEuz7M7vt/c3n/lOEffGkBmRQ2WmqkqZVaqi1CWpVLLaxm4JY0yDEfhBYCMMxhiMMfSjMY1NQ2ODHmyLbtCDadFGbfxgZCGpS6WSVVKpssYcIyMjY7px53PP/M17bz+s75y4ERmZEZmV2VKXFmRF1I0z3XPOt9fea/3X79/whhSbnVA2SMaRNA1t5ZakV0cYaELfY15UOIQmYJwjTQ3dWOZdjHO0I5+1Vshr9yb8xkdf8p44nhg6p5T6NeCzQLz0pHHO/a3tETwqluWR9XbIK5cH3BrOOZ5J/+AffvESIKeGd49m3DlNeeF8j6OZzBf88O6EKxsJ47TmeF4wSUt+/lmRkXbjgOfO9VZcoMdl7uXrqIzlziijl/jSMFI17x7PKU3MZjdelXn2xxmBp/izd464N8q4stHi+a027x7NyY0l1IqjaUlpZCCoMoZ3DmYEvhjHf/PaIb/xlSsPnEqWpYnQEzvMC4OEduzzpZ0er9+bMMqGaCVSOrBUVhN4irysSWtLaeDiICavxVgl8eHOcM6sqGlHAdYaJnmFsRAmPp5SzYUB2kDgOaxzHM4K1pOAKNDNgiK7rKwyHExyvvPekOsHM9KyZpRJI7WoLbO8oqodlzYiXtju8q3rw4ZNLxTKeSmN3nFW8dxmm9f2JlQKjDF0wkD4+YHGWKln78+kjp5XBhp7St8TLLJ10gx853DGeiukMrIoBJ5M8Oa1wfMVvVjUVsOsEqRFQ3CdF8L4aQWa5893+e6tU1qhz9ee32Sc1mx0InqtgOGiJJ8XeHrJtNFSx3eOsjIYK4vFeitkkIQcTQt8X+rhF/sJv/7yRW4N59waimmNaQuXaJLVjcdFReHEy9nXGuVB24opilasGFGeVvieIi0sfjPU5M4kBQv4vqYVyWclNXCfVqS5c5oS+UsFXsQorRi0fI6mgqZ+ar1NJ/K5dZLSTyrGecVT64kowfKaRWl4ei1huxez2Y34+gvn+OM3D3lpG26fLEhLQ2EMaeHYHWW8tNPdMLnzAAAgAElEQVThq89uPnDNPVwS/hffuc3FtYSXtrvcOIJRVjNL5TMKtCL2PXqRzxt7U3ZHGUmgeXN/CmjWWvLd9bSmF3v0k4Ak8hsPBfmQNzs++5OcorY4BFWjlUIhi/zSy3mQSAKJA4+qMmAFwdJNZNgz9DS1sfhaUNyhr3mqHeLUfcbXpxVPCp3734AW8CvAPwP+IfDXn+Lr+lTibHlk0Ap43uuy3UseqKsvPYF3Gl/bduTz7XePuTfO+OubQ3b6MeutgNDXXD+cN7Xa+yeFx/kcn30dN0/mBJ7mcFrgac1TGy2slQv36y/03uc4lpUC0vr+nTGLvGJRCk+oMobQ98kLR+xDNw7Jqpq7o5wvXBrw+t70A19LvxVwdavNTi/mR7sTAq9Z+EKftJJdfC8RrIVrFA/d2GeW16uRfLQgFfLaEvrChckrR20UWjsZ3NK6UdO4VXPVUzIlenec8es/uwNK8b3bIypr2ekneFrKF7dOU949nNEOPSaFDIkJptrHohhldYM5prkYhXsfeTL3cWkt4VdfPMc7hzMOZgVWwXrLZ6Ptc3ec8+bBBOWslGSco+V5LEpRyVzdbBH6HludmMgXieSsWDZQRc3iaUUvklNJVVvKyuIrRWVcw2+yOOXIKtgZtFCTlP1Jzr98dZde7BP5iqNpReiLwYlyTanAgFYOi6O0sCgqzrUD3j1e0ApykkCTlTKEN89L/uq9oUiKz3cxznGuF3NzmFIZSyvw6MQBB+OM0BdlV6w8+rEvgD6nCD2Y5tLHWfKVGpr3alYCGkMWFEUlzU6QvpjFp9/yySuhc7YizXavwzitGTalve/fOWWWi9/GL31mi9NFwb1xxqCluLSUSSqZAh+lZVPW7PNP/uiazDlUNZ04IGrgj/fGhZjUf0go4C9vnLAoDNrz+MLlPjeOF8wbm9LQ14xzQzfyGM4LDib5yshnkpV4wEY3Yq0d4SlHVtTNiU9QHZX1iAKNwxIFGlCMU2l4X12LOE6lid4KA9qxJqtk8r00lqtbbT5zvsckK9mfZDLZreU72E9EcNCLPD534W9Hj+CrzrmfVUr92Dn33yul/inwf32aL+zTiA9SEj28S5ZEEHN1syt3VIqsMlgcoa9RWrMe+xjnViYz8GQ+x2dfx29/Y0rdTK1udQOSwKesDdNMrob9ccbvfvsmu6cZ07xmnBZklUwcTwtDJ/QorWVRWLYjReRrSitk0aw0HM8K1hIxK/nD1/cfwHEv2f5nPRbWWwH3xjJ9GzVTr/OiYlILDlr5WhKFFRZPXluKyqA0mNpxcRBjnQDMRI7pU1lHXlsCr2nQWbXakU7zmm4iHg5ZZfmF57Z4Y2/K5TWRmPqeDITNMrhzmq4cwRLfI9VGBnCsZZ7X+L6mF4cUdY3SYnaTlTW9RIamfvH5TYxTPHfe0Y0CzvUivndrzM/s9Lkzznj3aE4vCvC0lCWyQlDKvq/50tMDiqomLY2QTT1FHGhGCxlowsJpmuMpQXaUDX/IU7KVq3F4DrqtgINpxq1hSlEbnFMM5zlVLbffaEcrzXgUyO5wWRoKPMBBO5EkfzQpmZWGVuCx0wspa8fr98bcHgp2xBjHc+fFzUwrKS1t90LyKqA2Dmvl/TPOsd2LySpLEvksyrm4mBmIPEFtWLuycxBKpkMUQQq09tjuCoNn6Q/8H3/xIsOFzBZEnpRei6pmXohSaZKKedL374z4+vNb/NLzW+yNs5WbX2UMb+5Pmp5PCDgu9GN6kc96JyQtRZlmHawnPn/w+j5vNE5jD6Pa98dibjTNazxPYY3h5qlwvV54asAoE87UU2sx1jnuTTIELHHfjQwFWWWIy4ok9FHKEWj53ZUSL+YkUNRGNQTWWlRotWGUm9U1PsoK5oWH10hzK+vYG4t5TifWHM9yKmsoS0sUeASVXmE4vv7C+U9iCfzAeNJEsNRCpkqpC8ApcPUneWKllAe8Ctxzzv36427/ScWHQeCWDbxO5PP6vSmv3h7TDj22+wm9JOTywOfZ812ySpAL1sLxvMA69z4D+Sd5HV/7zBaVsby443Myy0kroRV+9mKPUVquFE61FUPzUVphrKAJpoWM5HejoHE3krr3LDcMF3IqCTzFjeOUsjb0k4AvPr1GVhp+77t3+Mz57vs8FmZ5he9pTuYl++NUNPc45qXBOdjuxXz2Qo+39mZM0hrfF2R05GmGacmiMDx7rouxjlvDBbNMXm8/9okCj+HCUVlLLw5kMrq2LApRofzRmwcsilrw3o3D2YWBCNN2JylKqdXrqZrF37mAX33pHIezguGioKgdnucR+4rT5rnakaO2hjfuTYSHNCuY5iX9JKQbBYSBz99/6TydyGNvLGUZX6uGEaNYSwK2+wnfuXnKM5sdXCMvLJ1rlFcVRkFWWJQS5MRyEdFakl5RKWpjiTzN6bykH4ecprLzBFY8nGleiaeCr+nHAWVtmBWyqPTjpoY9rzjXj4g8x9NbPvuTDBD0h7EV07zEOkXsK0JPM04r0kIS4uE0J62axKmlmZ7VhsD3+PoL5/jxvQlHoThoeVoQC7pqOpfqPv7ZOCiMzBS0A4/NXsQXLouv8XJaf4kSWZZeB0mLH9+bcnu4oB2KF/ckK/nu7REO+MbbR8SBx7lexF5j5PLsVod3DmcMFyW92GMvr+gnIf1E5gYOpznToubHu2MWhZQVlz0xkDLvn18/FjBhP+KHd6c4pbiy0WajE/D5S2ssiprrhzOyyvD92yNm2X2jGHF2k7+3Qp9+HHAwK2iHPp85L3ykrBaMdS/2UQi+GySRF7VBIafurKhZ5JYcSWC+lgpCN/bIjWE4lGT5mfNdFoVhuChZFDVfenrAy5fXP1XjenjyRPD/KKUGwD8Bvtf87J/9hM/9XwNvAZ/umecJ4z7Z0zZevQJl2xtnArEyjv5a4yHsi0HIZ853OZhmjzxdPC72xxmjRcG1wxlVZei3QjphQCfyeGarzeE0Z6efsNWVC6OfyHF42DCSkkAUOJG1DBJfHKMij8hzK2u+wPMIPRna2h2LC9dWN+TWyYK9USaSyeOUQKtV0+7LV9cJm/JQbSyB59GNVONhIHz5g6lAtyJP4SsZTOonPuNc/GcXpVgc5rVBK0VRO5JQsd2L2RunjYRR3M9AVEyVdfz1rRFVbRi0Ij53sUsnEk/X/UnBxUHENBfQmEYkebOiFlJnXsuYv62pnWKWNRaVTtQ9aWG4U6astUP6kbCPhvMFz261OV0U7I1Trh3OyEpBC6wlgXCb6po7p1lTBnIcTjM6kc/53gClFLeHcxxyEhQZrMDnFEvvWUVaGgJfURroJj5FZXHWMMvFjtCiZOIW2SVmlUHjmOaCTQg9j81OwGY3EilsoKhqx6Kq2AljFHA0zXhxZ8B6O+L2cMHOIGacVuyOMrbaEe/lFTeO5lT2fiPYD0XNtNEKGKclN44X+EoavjePxQegF/nsTjKUdSwqg0PmL4oGzLdsur98ccB/9AXBmf/xm4fM8up9pdc/vXbERttnmpUcTAqccnQDj4NJwc2TBUrD7ijlx7sTBknAM+faRA3SfKMdcjzLGx/qgm5jybl0NQs8j7ujBYHyOJjm/N8/uNs0lH0UCqUQo5i1pEFuOEFcxz7vHE7Zm+bgpKz4sDGMQ05F87xipx9zZbPNFy6vkVeW1+6NCLWoxoyVQVGrLHeHGbOiXokFZkXNNGuw5Z4iN67ZKMhm4YXNLt+7PSSrXcM6C/jsTp/aWQoj3KhPOx6HmPgycNc59z80/78DvAa8DfzPH/dJlVKXgF8D/jHw33zcx/kkY6nk+eGdMZudmEEr5HCa45Av/E4/QmupyTqELOlpxW9+9epHztZn6/7nuhE3jxccTnP0QBHUikVeM0hEnXJlo8Nr96Ysml3HNKswWDqRR2UaEiNwruujleLQpRg8OmFALwlIq5pFZZlkKU9vdGWhLgwHUwHwLcqaQEsjVyuaHZ8lLQxJ6KGVkB37sc8kq7g3LQk9heeJO5Q1hkHjnauKmjjUzAv5kveTkCQQLlBeWa5utqlqwzAtOZzldEJBI2eVIdKaXuQztpZZJruhwCsJPPl5OxYdflpZylpkm8ZZjmcFx4uCeS6QnLKUXfmSl1PXlllpSJqa8620opf4TNOSV2+PUIwbjbw0q5WC/UqkrJHv0Yp8bg3nbLRlALAVBI3/sTy3BTY7EaWVRDDLxW+4G4d4WqSJz6y3OJhltAKftCjISkPggbEepTG044C0FJCfMYK5ME7M3sEyzgSvIO+LZpqWVFZxND1pSJmWeV6QlpbTtMRTirSsME6z0Q6YZeVKGeSaQbkk9El8xaAlk/R3RxmLoqLfCtnqReSVZVZWaOeYlAZjbNMDkTJW7HlsdGK6saYd+/xod8zLlwYEHrx6+xTn4PMX+1zsR/zut2/yg7sjjqcFvThgZ5AQeIpbJymzouQ7N09lpqN2DaBNJMXXD2c8f65DO9D8ye6UXjOwl5dyinHGkltNt6uZZTVpVWKdZe805Td/8Rm6cUDvDGZiklYNXl08wNuRbKDWWgE3juf43vs5Dg2GiXkubnNrLfEO78QBvqeJPY+fe6rFm3tjXt8br+TK1hrW2hGVvY8sUUDeDC94Wmwqb52kTLKak1mJp2GSGsapeH+3I5+8towWxfsIB590PO5E8L8Dfx9AKfU14H8C/ivgFeB3kKbxx4n/BfjvgO7HvP9PFGfdvJZcnnEmQ2LLYSulZHR9qxtSGyEA/uylHtcOZpzMS7767AZff+Hcx/pwlqeP64dznl7vcKmx36uMY7svaom1Mwqnrz2/ybeuH+Oco9d49rZCn822z/Fc1AVPb7ToRD5/eUOu+No5Nrsh00zkhcY6ThYZ07QWHntTkjFITb8oDc+ea4v0s3KrwaTK1gznit1xhtfMybdjn2kuFMvQE2lo6eDLV9ZoxwHPbHakbFKUvHec0ot9Rqno3KO5x0YrZF6Ir0FdGsKG56+U4CJ8DaO05rMXBlzdanOuG/JHbx6x0Y7oJZrKOOZNzXua13zxqQHv7M9JAo1yhlLGYOk29pmzrKT0PdY6EYGnVj0Xhwy7ZeL9gq/FrzmrauZFzSDxWTMhr94+ZasTMS8MeWlQKPK6BqSBHocePU/AbM6JV0Av8bg0aIODIBAznF6jQX/vpKYV+NSmxDpNbcHXjkUpZRcF9BMPax15bRqUc043lmnciYE4UIJvQNANP7gzphMvp6hrJmlJ6GmR05qGXaNkd2uso6wNkxzCRYG1ll4igLdxVrF7mpHmJYVxUg6yInfESZL1lCz+Fwdig3k0Lcgrw/WjOafNwGToa17fHXMwLdgeSAP79knKNBdHt3YjaEiLmiSUcmzaJJwKxzQtxVipqHjt3gSNY5LXlLXD147L6wnvnaTEnuX2adrgQGQqu6wd37s95GcvrXNls8W33x0yycTXoK4dndjjq89tsihqWqFQBPbHGafLsfMmlmlBAWgaEq9tsO89Yl8G1SpTMyukTDovRPmkFGTjfMWuWkbgifWpEFiEnDovhMlUGyk7oiQBDRclO72QvHIfOiz6ScTjEoHnnDtt/v6fAL/jnPtXwL9SSv3w4zyhUurXgSPn3PeUUl//kNv9FvBbAE899dTHeapHxtnd+N3TDK1hmtVs96V52AqV0Ba1DIG8cnmNeVFxOM2xDr7yzMYTeQd/WLx3PGea1fzg7oi1llBQX9zusT/JiAPNd24O+crV9RXK4umNNv+O1lw7nOIruDPKmGQVk1zxc08NSEKfdpNYUDDLanxPcfskxTrH6VykkHsj1wDbZFc8L40gL7TCeorTRSnQNyfN4ElWCi1RCRZYrCg1O4NEpoitY1GLneV6S7jyd8cZZW04nNbsjjNunywIPZHIeVrUIEVlCH0Ph5XF1TnOdWOUklPWhUFMJw65utVuENkD3tyfMckqqkZed64XEviaQeRT1bDW9jmalWSVAa0ItcwDDGJp3pe1RTvH7ljsMD0FSou94bIcUFowhcHz5KKcNBO7RW0pSscvP7fB0bzk7f0pi9Kw008YNZC4imXfQpzFnJX373xPEB0/d2VdMAVt2f1X1rIoDWuR2E5O8qZc0LwW5yTJNIPbTPKa7W600vYrpWhFkoRrax/4+elCprxtIJTOpV+u1ySZunZURhrKrbAi0EJHPd+LuXE8p6pFolmdwVHjoGrKpTv9BLRCa8GQDBeCi05Ly1o7oB+GnMxzvnX9hHhJbO0KzHB3lPHG/pSrWx1qI2wgvynH9mIPnAzv5bXhZy/3eePejP1Jfv/0gmNRGhaFwVnH3NiVHWRR2hWE7vrhAus0VzdbOBzt0CcrLaU1WOcxb0o3n7/YJ/JluLOfhFw/mnI8k5mZ1WAdopTa6kaEocckrbg9TDnfjZjmNSfzptRprIgc9P3vk9+UgKxbbnbcqodEA/orq/uIatPA9KBBoFvFD++OeOXy2gOY8U86HpsIlFK+c64GfpVmYX7C+35Q/ALwHyql/gEQAz2l1P/hnPtPz97IOfc7yKmDL33pS58Ye+/sbrwVindwVsm045eurHHjeMbJomCjHa7G4n2tP1YJ6FGxP864c5riacV6IgqImycLeomUXpLQY7uXEAf+CmVxMMkbn4CAcVbxKy+c5+JawqKo+c7NIZ/pxZzMSlqhz4VBwmlYMpqXjPMKnKMVadLSsihrWoHI5Y4XBVvtiEESMnEV1joxfok8elFA6JlmcEtMwgNPY6wwfWZZzcW1hGleM80qPAWf3elxOCs4meW8sTfDWsfVrTYb7YCbQzEU32xHjQLLca4ru3mNpbQyF5A2ZvXTOGCjrVaTxTuDhP/8F5/h9757F2MFBlYay492x5zvJQznBbO8ZpAEOCdlKE/J6wYhm9a1oBucdStnKRH8PRgOad6Gnvx9llsurXu8dKHHtDBsdiO25jGTTIBoKOkRKC2QM6Ug8TUvXehzYZAwyQQ30ol8ThcVt4cLkWZax8uXutw+zVnkmdA5tUwTe4jd5XJxF1CpnBa0B8qK/HatFVFbi6dk+tn39IoxVNUli0LUWsrdV/wEGmoc2kpDM/Y1i9Lw3rF4P5dVjRjmyKLvKZn3qKwl9BV57TiY5/SigK0LEbWTAap5IZam++OM00XBNDcUVYPpbiabi1qAdLNc3N1mRY3vyXdL4Zjkjm4SsNFRXFxvk1fi5Zz4iryG2VjsSaNAMclK1lpiFONBgzKRAc2tZibjOeX47q1TtroRz0ZdfvWlFuOs5HhWcDjN+c2vit7l91+9y/E0Z5rXhJ5GKQgbBhBIX6QXBZymFVeSEK8F57oRn7s4YKMd8sdvHQnfqjI4xOmtbprp2hPkukUygUPha7fyQu40joOBbhRaD30Xp3nJvXHGhUHcSFM/nXjcYv57wJ8ppU4Q5dCfAyilngM+1nSxc+4fAf+oeZyvA//tw0ng04xlL+Csm1fsS+PylcsJUaD5lRfOrSYTH2Uy/2HxQebyy/jR7pgXzve4fjSn1w5YjHNqa3n3cM5LOz2ckynmJcoir2TXvt2P+d7tEfdGwo95cbvL5y4OMNZxNC2aHkLAuW7M3jjneF6gEdkbiNxRyi5iMu9sxSitZP6gFhqjVqCcQ2mx9Ou3HUXl6MYesS/H9GVSUEqx04+5NFh6E0s/QSlZXGalgNM6oc9nzrXJasfxouR0IRiGjW7EC9s9frw75r3jBUNTcXkt5uKghXGOYSqAv2Us/Z+/ee2Qv3zvFGsseWW4tj9lWlREnl4N6TknMwwKx3Y35nRR4nuao2mOlQLOqgm4jGVRwGsuyEjLguApxaCZG3lrf8pOL+bmyQILnMwLqtqQ1WL40wp8LvQTLq63+S9/5Xl2Bgn/4ju30Qq+/e4J89yAklLS/rjkhe0uvSTkG29nqzoyCPhtmQQ04mkbeYK6Hi40ftA4p0UyIe5pSYznuhFbHWkUdxOpjde1ID183UzSKtnZL60bs1ru52vNNBM6q0Mw2dZJWUgpqW+YpsnZ9gNiX/HXN2Vu4b/42jP8n6/u8sbeFE+LhabvKSxi9Rn60lsLtKa0jq1OxLluzLyomeUVtXXkxjXDVI7zm21+/uqG+P6mBa/tTlhUIndVSL2+qA0vbfcoGjCcdaKEQsmsxnIhfudgRl5Znl5vNY54HaxznMwLdgYJP7oz4vrxvAHm2QcG62SSGHa6Ca1INoyTrOTOacobe1O+f/uUThysdvxOaWwt2PPa3aevRpHGa5Dnxkn5MYkEQnhxLeHeOOPIFBT2/V4EpXHcPlmw3Yt5+fLfEGLCOfePG3vKHeCPnFtVuzTSK/i3LpZTvQrFtYOpGIcowQAva/IfJDH9oDjbc1hOJC937A/X9pbJxdewO8yZ5iVlZRlnFWlheGmn+8CA2ndvnfLlK+uNX++YopY5gt1Rylv7UzbaAbujTJyZlBi9n8wLrIVW5OGcMF26UUBZOzJXsZ74hL4iKy1744wo0IS+EhvF2nEh9JllOUngU9UVke/RjnzakceiqBgksjC+cmmNzW7IH76+z3hRMlwUnEwLtFac70YY5+i1AsrScDBLaYeelCG0NAJ3gwyHk4usNCxKS20MT291qGrH7jDlf/yDNzmc5ivDn7W2SBX/4LU9QOP7DpU7ZrnBILvpOJRFPK8tJ/OcvUnOZifCWofCrVzIzsayBGCaUojvKbqRz6KsGS9q3rw3pqgtP7g7xhhLN/GZZZVIVrVCo+gmwlraHaX89jeu87XPbDFJC/78+gkH04JOFLDREZWO1ook9Lm0lvAnbx0Reg8mpmVEvgym+Z54Ow9aPllh0ShO5ilF5VhUFqw0sN87mZGVwqy6vBZzd5TT9hxVZfECUX+1A7kO+q2oYWwJ/8A6RxTIgGM7VCxKKU3ZxrDHaTkpjbKCbtLm8poAEt88nHNvkpGVQv4cp2Xj6ywlrnlRNSUsaai3I821gymeUuKNsdmirC2n84J5abm60WZnEPPypQHfvz1c2chWTell2c85mOVs92PujcTgqTKOdvNdurLZIqssV7c6XN1sodArwKSIJTL+1z99l//v3RO2uiFfeWaD13cnjNJKqKHW4WmZCxllJbNCUdaGvUmOcxBqmJeK4aJCOUdWGzylGjaTJN6GPSh9pECz0XiSRL4mCsTE6nCSE/sa5x7x4QPOiFR3lFUrE6lPIx5b3nHO/dUjfvbOJ/HkzrlvAt/8JB7rSePlSwN+//u7HE1z5o0Wf7QoOJhkjLOK/+wXPtp4xNn5g2kmmu/rR3PaDR0SeKC2p4C/vnnKWjvgymaL64cWX1s2uxGbvZBbw4xBK2S9LaRTpSQh/Pn1I2Z5RVZJaaR2jqNZzt3TlO1BwuW1mDunOa/tzYl8zYVBhHWa7a7ijQNBP2x1YqzTLCpDUTqUglaoCXzNPK+praUTBaAU/cTnZF5iETZ/6Avi2KJ4erPFdi/h+fMddscZeWWJA7mvp0tmCymZBI0Jx3tDoVT2k4C0rDmcldRGDFy01lgnswadwOfWaYZTivUk4A/vnHJhrUVZy87/m28f8fRmi/PdWCSKSUBpNJO0wiLloNATL+eqNuQOjmYl272YojJktV0t9GcvOx8IfOHmL1Uds9yQFuJZkFc1u+OUjVbAyaJEK0daSrOZpoxQOSgrw72sksE3BfvjnFdvT7hzmjU6c9gb5xSV4eJazO1hyuFU5MB5I89c7twtktTakU8cCKkzCT2eP99lMi+5OVwwWlgCH1qBZr0VczQrGoCaLDRFDetJwKKZPm57gqzOS0teOzb0cmNUcTIrKc19OmxlxbHNa0pWjbcM7UhO0ZHvcTQrxbe3+WaLLaWUqGTqWvNcY5mZG1HaPbXR4vJ6i0Er472TeeOlLZDCjbZ873/h+S1evjTgm9cOeftQwHZaqSYJyPe2G0fEnmbQCrlzmtOJpJwWBR6lcZSVIwzgq89ucPMkxbqa42nOvzyY4FD80vMbTPOSyhqGc7Fn7ScBs6Kmavjc60nAJK8oKkOtFVlpUEphjaNU4FUV7cgjCMTJbKk6CxpYXN6cVrqNzeyiMFwcJHzmfJfL6wn7k1xOMS7h5snikeuLQTAYn7/Q+xtVDf3Uxc4gYbMdsDMQTPTtYUo3DunHAZGnHus//HCctaFclCL7zGvDreGc9fY6RW149fbpqlQ0TovGBERxPCkIPZng3eqEOCdkwveOF8yLmncO5+RVzV/eOOGdgxlaK8q6oqhElWEaRcezm202uzG1ddwczokCj0GD+N2fGdqhYlqIicZ6K2B3XKK0DANpJbpwrcAaueD3Jxn9WEig2jmcElvAJPS42JiO9BLpr2y2A17c7nJ3lDHNKmprKeqa+aTkhXNdabTWlkEsht7tMOCooYDmtaUbadphQBLJbEY39llvhby5P2G4EBeystkJVsYyLWreC+dM0prSiM/u0twj0HIa8JtBqrW2T1HLZ/6D22ORxyJllaUJy7KmXznL0t9acV9vX9aGUSYL4iSTBqvTCuXsyju3to7QEyerspa6eC8JuHkyp6yldjx2DueqB4iXngfjrJQeTiX+u+rM6wu0PPY0l/r7850Ol9Zi/uz6kHYUEAdC9qyaZrHo1jVPrSeAYm+cScmvMb7JyprjuUhyW4FHWhk2nOM0FdDimh9xOM1ZawWMG79hX0MUeOSl9IfSxusAhRBqpwX7s5yiMOwMEjytmRUls0wQJJO85qULfa6st9ibZPQS8ZmuKsP+OCP0NDed5XM7A3qtkJ+91OO9Y0G/v7Y7wlcCZytrMa6vjSIMFM9stnh2q8vhrOB8LwQUn+3HUtZq5NC/9pktmehd5Hz31oiyKRs+s9nm1kmKp2G9FZGVlqNpgUM18MQQpVTjxSA2lKrpLZXNZxR5Ctco7npJwPYgJisNz2512BvnzMqK2lkS36MdedQWamMa9Lnh2uGcK+stfutrzwpp4J9+k1snC6qHuqEOePF8hyPbsoQAACAASURBVFc+RfIo/B1MBCDLxt97dpMf3h1xcT0hCfwHhkw+Snf+rA3l0jw+boalThcF3701onfGy+Ctgzk/s9NlmhlOm4bX090EB3z+4oCbJzNuHM8YZyUvnO+RhB5/ffOUe+NMhpO0DFLllbBKYl9z4yTlfD/m55/ZYvc05b2TlHFaEXqyuHiephcrnt1sM2xqyr04pDTSBAy0DCkpT+R3dWEbUqJBac1a7FM1aob1bsTfe2adr78gqI7v3DylEwqv5mhagHJsdUKR+llHVdQ8v9Um0FoutERc1g5muVgo9iIUilvDxaq2vDtKOZmXtANPardOoZUjbgBrabFEHKgzO9Km0VpZvBjCQCwmjZFm+6KsG7mlROhBEohlqFMCHsMJCthXim7sE/jCQxql0mOIA4/IF5/dsnmsqDlJhL6w8pVWHM1LDicZN45TNtoh7diXad7GUjQrLbunOV+43KfyPPZURqDBVxqLTO8qR+OyZTBGLD+LynBzGDWsp4pA+7SjgMrUnMxLuo3u/GBacL4bcXk9JqsteSmsoVunKZ1Is9YKWGtF3B2l7E2lPNdJxC51lJZEvvyunpL+Uiv08VUzJV1bOkkz9ZyblcfDvDTsT4tVn+dGOZd+SOM93Yn8xmdDceNoyq3TjM222JnOMsN374y4shaxP05ZVDVbrZA3DmaiyKqF11QbR+CBdR7t0OdzFwe4e2Oe3WwzK+qVl4h1jm9eOyIta350d8zbB3M8rej4MjcwzQ13hxmdxGOnH7Mocg5mJdbB7mkqJkOhSLBrK1wgkd9Kf8kitpJaSy9gUdQ8tZZgjHiIfPZCn2sHoixrh5q39ue0Qo92pHj3eMa14xm9MFh5D+yPBVneTWQzuZzVWLaGJ7ll54y/yacRf6cSwf4445vXjvh/X9vHNM2vS+st9ouccV7SjQKKWhaOs2yeZW3urEsZzf/eHi7IK8NT622ubLb44d0JeS3DX2/tT3EOLgwSfnh3xCwXDMV7xwv+vc+K4UpRi5pnlJa8dm+MrxXnujE/9/T6ymPg55/Z4Id3TzmeFaAUnlINpVPuf2+U8uO7Hq/fm3A0ExpmLwrQKOZFSeBrtvsJeW0FGR36lEYolqdpSWXFmCPwldRHoTkSW0IfWlFCXtecLiq+8dYR3353yO9++xYvXxqgERe1dw9nWGT3uJYE/MKzG3z24oC/ujEk8IQrdL4ruvNbJwvKylBZx7tHc+paShUAR6bgROfiWWDlOJ6EHrUVxIZ1wiDKDTxsp60RpUsvDlgUFeOsxtaWawfzB0pBUkLyGs2/w/dkutk6aRTmzYRn4GlpxDYJKi8rJrl94FlLA1sdmTlZFJZW6LHZDnnncM7RTJr4ga/pJgFV7RinFRvtiI2NkH5LTl3FrqMdeKS1bfwg7p8GsEI+jQIpH72xN2Ut8aktxE2vaV4axmnVEFXFcvM0qyjKim4s/KJ7kxxPQVbZhm9TU1u3Ol3++O6YvBRYXF7V0qx20I41safoxhFpUeMpZPhJia4+K6W86izMs5LXFwWB7xF4igv9hF4iDK127DMva873Iv7yvSFaQSuWfkkcesS+5s6o4OKa5nCS8vrdMVnN6qS3lF4aB/04pNWUqLpRwN4kJassf/bOEQpFVgoU8F+/dUht3IqQCvJe1o3ktqgUnvZIQs3JzKKQzVVWVoyaSk1tG8heowTymtJd7UAbSQpF7bhxMqcf+7x6a8SlNVGyTbKSo5msF6EncyiTvCLQmiw3/Om1kuN5zucuDHhmo8XRTEB2tbpfHuzFinP9iP1pzss/6QL4IfF3JhH86M6If/4XN7lzmtLypRY8nBXsjlIur7cJPMUgCfizd07wtZLBrmYX//uv3gWtuDRI0Aq+e0ucpV7c7jBalPzFu0Ne2u7wc1fWee5cm3cO5/TikEmW8uJ2hzf2ppzMckZpRVrWvH1g2RnEPLWR8GfvnLA3ynhhp0PQ4GcnaSn+t8iJ4/V7Y/JKLPXq2mJwFM1qpAFPOaZZyeFMYFpPr7eaOYOS2kjppKwstRE+0rysKCuRFm61RcY5L2po9NizfOmwBHUF98YL8koImAB5XQs9sWGpZJUAtDwtO/X9yjDLKt7cn5IEgUy35iV/sjchrSrRTytNHGpOF/VqkY49sM6KaYdrdv1AnUn9XKrO7gFXrYdjltfcHM5lgW302Q+34SoDxkr9VgPWc0zzklZwfwrVOZFoZkVNbWF6hkFzNpyTOZSlMqufBJTWYmpp3t49zdDKcb4X0w41WQ2vXB7w2Yu9RkEVEAaab709ZD5N5blh9V63Q2lAjhbVyty8MNK8Hc4KxouyUbk4JnmJzqXkNysNp4uC88YS+D7jNEcpEczmpWBIfE8RqADrafHxNVA7+Y6AJMy8ssx0zblQkrmxlrSU78YJFYEH7dBDqcZuUYGzllrJyW5eVOxPc3qRz+W1mN3TnJNZgVKO40VOVlg6kcYYjUNzvhPx5t6YvL5vklM2/YlOrOnHEef6Arl7bXfE6bzg+7tTnlqL6YZy6qmM4z/43HmqXcfbhzOZ3vZ8ekkADo5mOVll6LdDznVDvnd7SFrUTLKKtLDUD33GUhoS/MTZ4WN75s95btEYSlOwe5pinSUOhICrtc/+rMRU0qBassMUjrcP5ozSipe2e2gcWXlf0hxpSEKfaVrxXmMC9GnF34lEsD/O+Od/cZPrhzNZYEoBrMmkoqEylmc22+yOU47mBS+c6zzg/HWaip79pe0eP3hnzCQtGacVbx1MefnSgGc2W7x1MOfa0ZzPX+jxG19+amUY/4evHzDKShSabtMYs9bxR68f0ol9jmcFoa9l4rcd8+JOl7f2p1w7mKJ3+vzw7pjDaUYrCpikMrwFrHbQgZbFKGuMX0JPca4TU1vZ7VRhzTAtOZ4XjS5cJktbgccsr8kqy2Yn5LwKKUojwzQPvX/T4sGdt6Mx6B7mUpdvSiilkelKYwzD1HC8EE9dXyvCRvsfeT4Wi6aZqG3KVzIVKo8vC77Uyusz2uqHm7wPh0VuHylRZhRGdtiPup118uX3vEbi6CnSspRdqGOFnUjP/M6PCoWcglpNc/BgktNPfGaFmK475yhqcSxrR5rtXsyirLh2MKOsDV+6ssGL2wO+f2uEmd43LV/+nvPSoVXNmQoYs9wI7dJC2fxD1PQVSgc3TlK8RgZ5vKjZams85VEaWexF3eLAwPGiYmClFIFtVFNKkkAcyi64rC3DRQXOkFYP/v6VkSS5VFop3Xw3a0mmlXV0QsNwXnBruOCVSwPK2pDXUkK1AVgnJjc7/ZD9MxJf1dRGliW90lh6sQzRvXs44+5Y2E8Xmr7VN985lhJQ5POX753y9HqLi0Us1wgKreQU302CpjekOJjmbHQijqYFWWU/8PtluS8xXn7uGlbXigFGuaFtnbDJHBhbE3ia0Jf3cJlIFOLn0Il9xmnJyUwx7su8yXAhC74HaE8zyQyDRAipn2b8nUgE37x2xN1h1si6NMeLspHYuaaMYHjnSHgytXXcOJ7zrXeOcQiyeJyWtCOf947nfPvGCc65Rv8M3zOntEOfXuLz4vkeDse33j0BK7vASV5TlBaU1POVkl3j7jhnqxdxvh8zy0p+cGfEjZMZnTAk9KQ08/aBGNdP8xqc1FnTStDPXjO6bqxMnQ5aIf3Yx7pGBI3o7IuGjZ8Ey4VRAGpY6Q9YZ8EFhJ7muCw/dKE9G8t1STARlqCB0NHsaqS6DNSWwkGqIPbB9z3mhcPXMnzlnLB1Hl7kZfhJoe19uWfwkMfuo0JpGQLLtSTGs4+5vJBd83elZddvLeSlpXZS86+tHPfTxnzkw6YZLXIiWBQ1unnMMPAIa8ukdKuEVlmYFZY4LXn15ohZUWOd4/XdMXHoc+1wQV67Rz6ffegHDvEUPvu7FfZ+8jRN6WLQ8ilrxziTE+Yy0Xo+OKsa0J3jaC6+wq3QsSgkKRt3/0QVNgY8zkmZpH7o9djmv9o4fMMDO+pFaVk0xxsfuHY0w/cUIQJOVMgA4LL5vD8VDIRtHngp6wUZTFuUhrbSjNJCGrGBx6I0HE5zSmMZtALaoScGOZ4m9DVeo147nuUsqpqwsSI914t5dqvL2/sz8loECfYRn/by+xJoRSfxGC+kHPaor2JW3n8E48DW9oHkubT+jHCczEvy2jEvS/7i3ROMcQ+cMqT/5WQiPXk83v4niZ/6RLA/zviD1/ZJq4q8NBhbY41rwFyNykM7wsCjGwdMm7rejeMZrzSUwf1JzkYr4PYwRWuYZzIav5yoHC9K1johf5WekOY1YejRjQK+cnWDTqTJymXt2XBlo827x3MCXxq0s7xuzEUM5czg2qIbFtxAzhDFNK0Jfbi41uLagdTidXNBukauNs0rkW8qWFQ12/2Y49mUaiUJhbJyqyGtrBb3qdrArWGGUjzQTH3SUMvXYpxgeQuzukAcsgBqmnH5Ejxt5XTQXOzG3l9IHg6N7Nhd84Cx72GdeaTefhnO3i8nPDyI6R7x93boU9WGSd0kH09jnX3gd3hceFpOYoWRid1ZXpE36pK62QE2M1mcLGp8VdOKPEJP8dbBfIUzeNLnW0bUNCuXJuwWVmUdMXExaA1pJQlpmWSWCSfw5D2uLHi1FbMg3ZjhNJsMpWCaf/BO+eF4xAHsgX87muQEgfQE5oXMYeDE+3p/WlDW5v57xv0NgkN2ybUV2mpWWy72I45mApqrjPw3WlQkgU+vmQGx1jbsqkpw3lHA+X7Mdi9htKh4x85IG6z8wwnubDikeaycox1qZuWj35HlT5ef+fLzcA/dZlY8eP+idg/cZnl/D0VeS4L7NOOnOhEsNf7GOTZbIQe2ZDjNwLpGcCiUx1B7pHnFPQvOCfN+f5LzymVAOdZaIUdzkX0GWpZSrQQglZY1de3QaUkvCsRJal5SVFKjDT2PyNO0Qh+0DKksiprNdkhRG9bbAW82emKnpP5b1oYk8NiblFwYxFwaxNybFBxNC1GmcP/iDz0Zq5+Vhp1+i5/Z6XBnlHF3JOP4rdDDIBTOJQJh2RSt7f0vnK8V2cNbzycId+a1ZOWjF4yzPxOlSbNAug8v9SwxzFpZlHJ4nqKlPCb5Bx8LlicRx4Oeu48K66Qha1GrckheuzOm8U8WqjGg8ZVA4BZ5/UCZa4l4oHldleOB3+FxJa9HvnbkFPCB/+5kcVnSMx+VbJefP0hZrra2kcPe34k/7jP6qFFYcJXA+zqRz1YnoNV4hHeiAIVjVtzfVeuHjkiVlYQV47g7KlanLeUETw6O41lBb6vFjaMZ1ilakbdiav3MhQGb3Yisqnn3aM7xLOdoWrzv1PWosA6GqUE//qZye/vgexco3icPXcbDJ0GRMMvwXOxrPvoW7aPFT3UiWGr8n9vqcP1ozloScDovmBY1SjvWkwClRNdbGAitZb0dCsyqNOxPMrb7Mb/8wjm+8fYR4DiZV6y3xZt2mstOwvcUSmnJ3InPLJdFaJrVPLvVYZ5XHM5yylp2LBiL1sKr78WhaOqbVWtmZeo09iXRGAezsibyHUezCqtkgKhsdjDKOXztMYg9nttqczgT/X47CrjRGKwPZ2UzOfr+MoOvBIql3vfuffR4kgXDIOWd5c79wyI3EGNoh+IGlxb1qm78YXF2EfmwC9w4RG3zwB0/3gVX1naFFoj0+5kxjysvfRpxdof78PM73r/IP6rkE3tgPmJifFwsm78WS1oalPLoRAFPrbc4mBYUtSUr7AqW52t5rXEo5FoxaBLFnNdUQvPmMVsNt+j2aUpRWy70E3xPTt93R+KF8dKFPq1QM14IpFBmAp48nuTzepRIwfcUyt4XXXzYbQ1SmvM1fPnKxmrj+mnFT3UiWGr8P3dxwMEk4/a8aOqmljDQtBtOSDozJJ4i0Ip5YbDOcnmtxXY/5otPrTPLK64207RZJWYr7dBnOC+4cSyj85GnxF7O84gjUelM8oqNbkDtmonXJCBoGqcns4K0NBwuNe7OsdkJSStDbR2jtCLwJFkUlWOaGy4OYtLSMM4qUKBdUxuuDS9vdFEa9idi4JFW9UrxkkQeVW1kUMm+/+IPgezDzsVPEMtd7eNq6sDKAlBpSUQfVvd3DlqR5nBWCkvf88TI/Qlek3UQKmmgPuq1wiezCJfmwSf4gKrB37pwNBgE3p8EoCGV2qXF6P37fBKhkb6MMaBVxdMbCXuTnFlWyecWSENbBuWESqu1YtJYP6allBiNdatSoQaMc1SlNGvX2gFb7YhpXjFpWEWTrKI28PrutPH0sFRW+iCV+eSS8qNOectr7Ozp8Oyf77t/c3LvJcGnnAZ+yhPB+hmm/3a/hXUyBp4EwvSf5WIq3Q7EYq4wlgAxxchKw8E4Z3auYpbX/IPP7fCj3QlfenqN794acTIXtc/PX11nd5TRjmUa0+E4340aDbbPjeMFke/xxRfW2OrEzPOKa4cz0qJinFbcHeermujRVBa7Jfhqq9da1Woj3+Pp9TanaUkv8bgzzMmt7IgGScA7RwuunyyoamHGBFpTOYcxMtk7LyxiAPn+ht9HLYU8KlbKnmaH9kF5Zdl0W17Aof7w5FFYOJiW4CD0Fdo92WLkK2H219ah6vv3UbDSgn9a8eke4j+5OFvW+6B/rx30I828sKvTlVaycNTufjI5+3k/bjOwRIHIdzwi8jXTwnIyzekmAbW1pI3gQAO5dXjN1PfCVSzymtDXbHVDdse59DWa05ip7jfcQ624PcqojCgDvWbTMcsrjqc5J574aXQjQVzM82o1oGisI6sbxtJjfp+HY4nkOBtn3xPH40/gFumJGAfvHi8YJOGnak7zU50IXr404I/fPGScVtwazhszcM2vvniOYVoTB9K0fGtvwu3TlEErpJf4eEoQyLuT9AH66JJ+udmOGKYFG62QF7Y77Ky1OF2UXOzH3Bml5JXl4iDhlctr/ODuiPPdiMjT3DiecXeYUVnR2Y+yclWLXUoaLTQ6ezgY56L11uKN+tr+BI2iMJZ+KyQoBXlQGEtW1GwPEmojvr/jSlQQUSPfdEizTtGMxzfes2fjYUncMj5KDfvDFhaQBXpZovKQ+u7jLrJGVEJdORL//ejoR0XtICslw559fMej4W4fNZ7k5PNJPMejygafRKw+hye4bd7ssJfhHGhfEeIwRk52qiGa4j54E7AMhfTiDDIx3gl9bp9mOGSDUJzpNZ39My0tSSBS1yT0ON9LOF1Uguxe9cwkyRsHo6ymF8upIgo95nlN7GvunqZUzuEZy9Yg5mgqmI/S08zqWkqyCFZk2WP5KPFwbwAeXZr7oPCaNylojmuehtOs4pvXDvmNr1z5iK/myeKnOhHsDBJevtTn9757l8o4ksBnqxcxTGuubrbYG8t4vdaKpzbabLTFwUrMuzXb/Yh//3M7DzxmZeCXX9iiHfkrw/qvPrvFm/sTXt+bcr4X0419Lg5a7Axi4mCd798ecf1oTicKKE0tRh65aLrP7jaWfy53Qj6W0iiRChpYINOgzjpSLRrlQcsX6qYnmu+8tign08eVFROPJPBWrz/QNMoQRfpQ50o3yhZgdREsd0Qe99UbP0nIcfe+JPRJ22DL26T1ky+Ljxg+fmR8lEVdnflzuRt+XPL7uPFRd6If5/EfF4pHN6WL2tEJFc9vtUgry71xAU6GsdyHPPBSnIBuvJoNeI0V7PFMTuy1g8iTk+ryqT3A6cZHQinaoYfnCVD87GJt7P3Fu6wsqZa5idCXifrQV1zZ6LA7SkmrmqKGc92Ak0XVqHwcRimUUySBYv4x6nw/SeJefr98LT0F8e229OOA1/emP8Ejf3j8VCcCgP1pzleurvPFpwb88O6EJBBnrL1xxtXNDv/uz5znt79xndoYZoUhLQ1JqHlqPcb37i+g++OM3/32TU4XJVvdiCsbnRVddH+a8xtfufI+L4KdXsxoUfDO4Zx5aaiM4XRRMctq0Rg/9FrPXj9iKQhaS6kq8QRMtzfJKS0kWoaeSuOYN6P/RW051w0ZpzVVVor5hRIq49nnqBqu/MPx8IK2RDrXxq3kkB81NIievyl/GaTx93DP99/EDvuD4qM876oRzYerQP62x8dtCS1PjVGgqCzsT0v6sY/nQV42i/xjQr6XjigSWNvVjQ7H84J+EvLu8ZysNO/bdFRNvyf0ZJPTTwLuDqUhfPZ2D3+Hy9qy1groJz6+Ehrt4TSnqC1rDdtnmlsiT+Osox37DX7dkJb1x1J0/SThkGsk0mKI5RzUxoJ69DX7ScWTKqH+rYv9ccYfvr7PH76+zzuHU0DxyuU+ka8payFaLn0CLq8l3B6K/WMcaHqxT20Vn7/YXz3WH795yOmiYqsTUdSWH96VBb8d+ZwuytVtstKw2YnYG2f8zrdukleOZ861GcQ+t05SjLNofZ9u+WFhAV+pVc19Xhi2e4n4CQTeyuy7G/ustUJQgryF5vje6NuVgthXBFp2/cY+uLN9VBgnpaR5aVdo5o/zPVyWvKBhtkDjAPWgpPFvYj39uCM67UCY+p/WSeDfVCzpovDB34NlPfv/Z+/NYiRLrzu/37fcLZZca+8qsru5qtUSRWqbkTCCYIvQgA+2YcgP8+BH69EwZMEy5tnPBmzAfhBs2DBgjF/GMOyxMJ7xaOARhzMSSZFNimSTvVRXV1VnVeUe212/7/PDuTciMiurKrOrsrpZqgNkRWZUxI0bdznnfOf8z/+/PIfgEKrqxGpCCGy1kOwu2D/O5o7OyirAKM0Hu1OMForsxjmZmHYLCGvX0NZGYL4eiKxhVjlSo+hF6kRH5pBm9yCx/N6XLzFpFcRmVUPdNLy/kzMuakCgyaULfO5iny9fGaK1QLmXA39XqjutPY1z1Upgr6qNrAezZu6PzsNeyBXBskbAlZWMUVHz/dsH/MqNNb722XXGhUhCXl3L2DrIaQJsDGLK2jEqag5mFb90fY3f/ZKwGXYw1C4IZLEFGj7YnfAFM2SjH/PWnQOcD/z1h3vc2svZGZcME8v7O2OurvQYzRyDNGJUNnx0kDNuFbieWB93YT4tWjSOC8OYyhkmZd3W/mVa+otXhjwYl9wblRQtCgmY0074IA0sZZAehBFH2NXrTwuJO6vvW35Plz3XHs53YP50Vn3M94UQOBWO9RO005SsFKdDA3UOmaXSYVV7jNZ4I32CTrPqSf2XzqGm1uKCarmVhCvrnfsFu+3Ubmddz0wrUD5Q+MBKq2vh8aSJFUima04EAGikV/DN93bpx5Z+LCXd2isiG6QU5drZj8bz7oMJm/2EUe6ItaIKkomrlu7kLPZxVxLdnJD3gcRq+rHltYv9uT86D/t0X80f05Y1Al6/2J/z/N/cGTMuBAW0zCh6fS3ja59ZZ1oKM+j7O1N+8OE+D0YFwDzzF9UjR16Jtu72uJxv6+b2lO99uM+7D6bEWoQ59vOK73ywz/1RzrvbY2Eu9YFeJCIbmVWnQg90k6dF5RkXIn6hlOLiIGU1jbm6mvKly0O8F3H6xBpeWUvJokUeY5SwbCbWoLVM6fYig24DRXwGfP5Z7NOcNH9cSN6sgVFxMgndJ2nL36fr7zzOztLz8Ryt/Td+MX/i/MMQ3cdtxwGjomajF7Hei5hWQs63dZg/8pzELeKn9hBZzWY/Yb0Xix6w0g/p+VotfQYU1I3nnftjbm5PubkjQ2RF1WCURiuNMaIWh5Lz+uHejMYLkaIL3YSvINxifd5AzhbZpgIb/YQ3r63x669t8Adfu/5SmOastqwRsNFP+JUbq7y/PWXrMOcrN9aPaBDvTUVW79s3d3kwrtjsxWgduDeq+G///F3+03/n8ygC/+a9HRETUVA7x6iQ4bOuvHSQVxzkLZuk1fSTiINZybgOHHbN6cOCvIYvXx5SNodsj5/MZROQkkqkQWnFpBCK6zcuD0niiEDg6mrCD+4ekkaGz272eHBYcH9UtqIt0E8MzmuySLExTERSsF1ydspRZQuVe5xzO/7/n2Rd/1nY8r4/71rweViXuYeln8dZfArupuOmaZXTMkOkNZOqnh88fYbGuVIy3/LjrREKmJQ1VdNe56qjGW+bxMDaIGGU14IySiISa7iykvGjjw4om0C11PRQyJu6JvdA0QociYSkZNtQe2G1dV6B9YR2nqBY4snQSFm227xVJ/NBPUtTCNmlCQ03+o5/97ph98573Hp7wng8ZjKZzH+e9Pdp7YUMBN38QMcgKrKPUh/cm1a8dedg/tpbu1N+cOeQu/sz0tiQJZbGBVZ7wkP0P/3rm3jg9t6M9V7EhWGK84FXN3r8wa/dmAcUERxxc577XmS4W3qySNMEz+XVjEnhubGR0U8sX7joadyIuvFMKieQtxO+i1YwSDSxtVwYJqRWM0jk9+1xxbXVlM1+zF99sM9aq5krdUURH/cInYEGNvuWzX7M5sWIH3w0xhJaLVhZ/j7JET4JEvdpCQzHB3ZOYz/vQaCzZZTNk3z8WYJAakXIvR8bjDZoLSXJ2BpGRcW49CfCkY8H286sVuxPKnZnFaqFkWqkL2UDc/ZUkHvrl15Z4+b2mN1ZiQ+ixNc4T9mIkE83zNjBbVtGbFKr8DCX3YyMwTVuTv0BQAjkjfTSjuf7J10XJ11XwTt8lROqfPFYF4u/l3+vCny9+D3UOb4q5PkqJ9Q5N6uc4B3/5Mmn5pnYCxkIuvkBEM3Xu/s53/lgn199dQ2t4F/8+B7//Z+/CxpurGdMi5q88jgfsEqEzy+tiBDHX98+4O994QK/dH2Vu/s5b2/JRKLzfh5Qrq5lvH5xwAc7U/YmNbO6YSWNuLSStOInsNmP+fybA/amJd96b5eNXswvXBlw56CEaUXVNOS1QDtVewc5FsNfv/O5Da6v9Ql4diY1v/+LV3h/e8x3bx3wzvZEJodVJNORQQZiloe8CEJ5MWu/q0GEXzb7ER7F4ayhOWsR9FNqn4Zg9EnbszyTq6nhxnqGMQK0ikblDwAAIABJREFUuDiMmbUiK+OipGr8fEW1TOvxqBVXagQEIYpplrqRSeGm3fGOL6oLLI3z/GRrxEbPkteGUS4N3qJ2eC9sqVkLvi+XpndleFDozw3MkzzXOFRdEFyOK3PKSpx0qApsKClmU5py4bg7h77szOWnINTivHHPp+ultWYwGDAcDhkMBvOf5b+Xf/+TP/mTU233hQwEV9cyvv6GSCnuTErefTCin2j+5u4ht/dmQtPrHcEr9qY1SoviV+OEXnq1F7PVjrvHVrGWyQrDu8DBtGJ/KnS4f/n+Lt+7tc+rF3oczITbP4oUG3HCzlREQyKjub4uMNTVLJYsymheuzAgiw3DdMp3b+2R10KBG9qRfqMVqVFYreTkJzGvXugxLR1393P+2Y+2uLkzJYsN19cyXllL+Otbh1ijGBX1XLwkMrIyoA0o7+7MuDhI6CUWVXuc1zS1P0Xb+sn20gG/mDYpRGO3nxiCD+zNSkCxnoiOxhHG0aWLoCvrLKPDQGZxjAn0jFBHTBpZlZ5U1lLB46qKvfEee1VOpmqaMqcqZuwejFpnnC89dtn1IhOnzucO21cFvi7P8Wgtm0LFKTpKUXGGjjNUlMrj/PcUFffkMUqJkh6b6yuYJOPK5hpJ1ufvvXGDfr/PN772Oq9dWZ8jiU5jf6sDAUgw6FBB//b9XS70E7YOBCI6KR2Nd6TWcmGgGSQRYRDYnVVMK0dARCoq51gh4l/9bJu8bhjnDXkjXEOJNexMSj7czdmf1fzeG5dJI8M3333A2/uiQbrei6AV0t6ZlNw7LGm859XNHkrDpGx4MCkwRmMNRK7lKwcgtFqwnl4UiJRkRe88mDCaVQKF84HYag7zhs9d7JG0QhtVS2nb0UOjlpAITWB3WrEz9RgEC26NQr904y+UPcuz6ZCa+ih3aAWVk77a/dHRTwkh0DRl64QLqirH1/mi5NFm19Q5tim5V81oSimF1MXCkVNLmcR1jvw5mYqSubNWcYaOMnHkcYaKMnSSta/ptf+ftE58+fl0sY0oQZ0RXWYVRInBB5gkltX1jGm8whevrvOz3ZLXr55Ps/qFDQSdvXXngNRqbu1Nubk9ZW8mIhzBQ+08N3dE1jCNhcwsrxxaWTZ6EWVtCMDWwQzTisZbHdDa4oG3742JtOJ7H+5zYZCw1rMc5jXb45KALKk/f3HIQdHwN3dHrGaWxge+/guX+WA359bOmDv7OfvTChcgtjCt2xF5D70IVhLRq/2zH9/nxlrGvVEhimS1n4uJH0wr3t8eMyuFrbDTGeiW1ioI4kHI74TI7iCHzl1ULUT153lA6qWdzUIIhKaaZ8qhOppVL9euu/+nludctXDuXdbdved5rQuVTU7MtjtHbNrnbJKh4xRvUkwizt2mGVGSsb66ysbaCpWKGDmDRzHOT0do+KzseC+nCULV3m+ps0d5zVt3Dnnj2gp5fX579sIHgpvbU8pG6G6bIPXEopYqpAuO2gkx22o/pvbwpStDVrJYiLRc4MGkYG9WkUUiYu40XBrGRFZxZz/nM+sZkVXsTkr+5U/v89FBThaLNOCocHz/zgErqSGgiI1md1rwrXd36CWWn92fMilr6qaV5VMGg1yIgmYIvLIeUzae7XHJT+9PKFvaa9cECg0BgaAao1rZRXH6yxdYQBAUCkFPlPnRCypwdoz0S3u+Fpp6KbuetQ3HoyWQI83KumhLJLP2dcVDr3uIGe28zESLskgkDvlo5rxcKmmderzIym0iP1m/RzAZjYmxRhz3ceZXeBgBZpe7yEqG4QiBUYBRLtd/L5bmdzhnxMPx+Y6TXLuQIgbK2rGaRQQf+Om9Mb/5+ua57dcLHwgO8opBalnvJexMSiaI+EjdyClw7bBKpBW1VuS14+LQsHWQ85nNHodFRdV4jBJxchc8ee2Y1a1Yuw/cWEu5Py64PyooatcKnwe0UpSNY1I1vLoxIDKGYWR5Z3tKFmk2BhGTspZJTQ/eLwlyIH2CnUnF3qxpERihJWoLLVWEiLkHQC3L3AWhhigbPx8YUgg07+eFIvnn2YJrFg3G1nEvnPFsqenYPnYOvm4dd4cgqReZOv5x2l/P0JRuHXFbt14qjTzkqOflkhPq4Em7jShFmYfdTGKQgTTvj4j4zHeDh/1x3j4aZPLWB04khjv+vma56RA44ol1iy46LALQnDvd82ngtd2K/96ooHaezUHEzqSazz6dh73wgWA1ixjlDYVz1LVjWjZz2JxRLWeMgYO8JlKwM6m4sR7weLYOcj7YnRFbybYTZZhVMC4risqTRprxrGKURuxOSmZFQ+3BeyeZB5JpR0Zw0iJO46gaz3bjSSJpp3Vj98uaAJ5WUlB1HQOxZQWu7nXL/w9S3gmdzBSiw+v84xWt/rZa8I5Ql4ssu4XvLZz0Muzv+N/F0Rp4LY4e9/yc9nLzUWrW2dLfS845TqWGfdxpx8vvz8DYMzUjP65ppWicf2xi8rjZjsqFudrecd/68w4b7qCv82E9B29cHb4cKHsae/3igKrx/Mu3H7A9FfUv44T9ci7j10AIjhoZOx8XNZHR3N6foRVcXeuxN6uZlY3o2bb89mtZzMYgYVLKtHLtW/1aa3BOBGscwrC4MymxrcZxB7GLrML5xeXe/WbbcXaRG3x48XiaC72TIQxA3Xz8KdpPk4Xgj5REjte3Oyd+okM//vo2Qw/N80KQ0JY/kjar7p2AIGkdcosgmTcfoyVkyZKDVzZ+Lk77PMy13ESPskcNxHXEhd0Lutn507z3NPZpaJEFWg4oBZ7AK2spa70XTLxeKXUD+F+AK4jf+9MQwn9zHp+1dZCzPy3585/c59bejMbJ9OBxYfKuzOIJxJHi3mHOpVbg2iOCHBv9iLJxWAxGKbJIU7rQKh55+rGmZaTA+0VIV0FO6moaibJYaNE7AQ6mzXzloJWMzhOkCTwqhJzuSbq7j7Pjq4jUHp2aPE8LIRDq8iFo33EEyYlZ9rITP5ZxPy9TUSKlj/hoPXuOIIlPhgTqecOyc+Kdo4/PjCB5Ua1bKWexpjljrbK7pi0Q2on7i8OYshEZ2acJAp8m62hlJnnNnf2CH390eK6f90msCBrgPw8h/LVSagh8Vyn1z0MIP36WH9IRz32wO+b2fs6ocHMmxZOsuxzz0rOWwTC2PDicMKscpZMeQWw01miquqXJ9YFJWc+ZPrNILwTRu+lIA0ZplFbUIaDCApkTkEARkFqlUYGgDaXzJFaC06P296zTswrhGSI4jmu/hxDA1UdRIOVSdn0cAricXT+UZS8c+vNDkMQLyN8JJZEjjjtaZOUq6R17bfveKEFp8+QPfmkfy7p7rXoKhSCtRYGu8rA9roms3BHH1dJ+Xs3TDpdqAbd8891d3vpwn698Zv1cPu+5B4IQwhaw1f4+Vkr9BHgFeKaB4K07B+xOC/7fHz+gatycB+Vx14gFjFVMCseP7h0yLRqsNvRiw8Gsom4gjdqxFwWDJKL2nkFimVWOkWuIrCY1kvnVjadyHhc846IROb0lBkdY0PzScgl9+dKAe4cF9w7LeSb/kLka6wrqMqcqHu+Mfd3hsnO26wJXSjPSHXv980OQ2AU+O1oqhRzPro+XSh76/0Wj8qXT/vm0x0GVn8T91FFVg4AkXJvcnAdn/ydFneI9eAV7k5Jr6yl/9jdbL04gWDal1KvAV4G/fNbb3ptWfO/WAXnliayh8c38YjlxX6AtxQSaEJjNZAXhvXAUdRS3znuGaUTjA4dFRWYNgyRinFeCYHCBWYv3DQiCRDcFrmqoZ1Mm0+kxnLY4ZNUUTHzFvisJdcF4PMZVM1xZtNOQnySCZIEEmSNI2nKJjpYGcI7VsI8gSLqs3ETPZ99f2s+1nSotacustYeOvDog8zdFfQxCujRfc9YVwye1wAi01PEW8spzc2d6bp/1iQUCpdQA+MfAfxZCeEiDTSn1h8AfAnzmM5858/Y3+jG3D3J6sWZWg7Oaxi+IsRRHZRI9UDWOvq7RVcHkcIKqC3RToF1JmU/nDjkPNVUxpcoFFvh2CwGkkizdlfk84w7PiYNEECTJMehfhokWztkkGUQZNpFySYgyTJwSlkonyxBBTPRz24x8aS+2eZgLJsFiZR04GgQU0n9LtCIoUfzT7vzRQs9qFeGCzP74c25+fCKBQCkVIUHgfw0h/O8nvSaE8KfAnwL82q/92kOHwHvPdDo9Qrm6TMN6+/4ud//iRzRVTpnPmE2n+GomTrqUDHxRImlLK88ZQaKPoEYSkqyHinuYOKPRCbRZ9hwlEmdEcUqc9Wh00jrwtg5ukxOddjdYFiuZnm7a7rNSkNenE4J/aS/tvO1pHefxUYEF6RxcGERMS8es9u0MgrCSVk1otT2Y64c/bv+6bZ92f56FdZoRPoh62nnZJ4EaUsD/CPwkhPBfn+Y9P/3pT/nqV796hGt7Oj2/ZdJx01EqGO0ko9cb4G1CsEIYZduGI3FKkvVpdEzWG9CYBG9TkjQjyQboOGU4HLBTKEKUAKJH2un5gqxQskgTGc24aB4i4JIhM+ZlqtNcnHOEVHvByzxNqyrFghr4ZUB4aZ+kfVzH+aheQmQgMhrnA5PSYZSUhnyQ5Egh8G25hxQh+Meqq32S/Wfn4XMXeufWH4BPZkXw28B/DPxQKfX99rl/GEL4s0e9YTKZ8P3vf/9R/33E0jRlMBiQZj2cSWlMTLCp/EQpG6tD+v0+Yx9RkdDr9+j1+oy9xZuUleGA2qQkWY+JtzTExIkhNQCW2Mq0cF55IqO5tpZycZjyk60RKOklZLElBKi9Jy8brFH02ueUr6CFlHaw1c4CMlHolhq3XWYTaFkc/dHXwwJL7Zb+9scetVaEIGykTR3mQytWH5u8fGkv7efIHuW7nWe++i3apkAH06bt5Xkv2slxJBPOnwb51EdZltgXa7I4hPBNzjjfdPXqVf74j//4Ib7tfr/PcDg8wr9trXylf/o3W+SVo3aeD3ZmjMuasnZUjePt+2Oua0OkFQd5Se3gcysJd/ZnjAuHC4GxC9gY+rEliwwHs5pBDNYKRLQ2EEWaJLJkkRX657whifRcY5gQsC2vSd140lhjFBRemlYPHZv2wDROMprGt8vWJ6TrnbPvzGhhE/Vesp8O6loHh0Md0al9STnx0j5tJorgT2cnUTl4eKjWPmsCtWs+1XlQbGEliV5OFl+7do0/+qM/OtN7OrlKrRQbfZGt9CHwZz/8iLUsJrGGcVELisDD/rSmcoHSebwHaxASKh8oa48PHh80wySitp4s1hxOG+7sTbm7P2NWNYQAF/oxRQOrmSHSEZUPHM5qnAvsTqp5k3q5mbV8EXYXa3N86u0JtuzPjQ5opakRwZCqZTNtAKVe9gVe2qfXFE8fBM5indTxp5V0MUKEtsI5YzZe2FHHTq5y2aZlw+6kEpnHSclHB7lQuwbPg0lJ7QJZpNtyjJrL3znvSazBtjKRqTU0DowRgZrKeRKr2BxETArPpWHMMLHkTWAljejFlnFR44PMCsR2cQE+bsjt4577soGyvbK10VzoWYx+iG/rpb20T509Ta+gq/2fxayWnpwGelYRtTDTZ2nxse2ddh8VsNaPuLbW4/JK+mx36pj9XKwIPo51cpX7s4oHo5LdaYXRoku6My7Z7CfktWN/Wgn1RBCRF9eSWXWZe9kIZ1AvNuSVZ2dSMkwsvUQk81KruLCSokOg8oH9Wcn4Xo3WimtrGcPUUjUOF2CQGpJIDvmkqHGVn3ML+XDUSS/3Bs5qgjIQ+l2Np5fEjMtP9/L3pb20p7Euaet4uk57rfvg2ejFKGqsMbimQaFoHkeEdEaLrSJGtf2/0+9bYuCV9YyqCXzjzavPbH9Oshd2RXB1LeMr11f52f0xu9OSC4OIL10ekleOg1nNYV4RG03SsjtprYi1ovSBxGqMVrggCJsk1mit6cWKSVFza2/GaFZjFGwMEvKqYXdWMW6JfBonFNSzsubte2NmlXBFlLXHOY8KgXqJRMiFoxORmo9/Yrr3dUHEaM3+rHqIVuJvo2mEguClPR97mkNt9dnfH5CV9lneJxodci9WjZNS8TNeNrsAWWSEigYRgDqNbQwS3nxllT/8ndfOFTEEL/CKAGBrVPCbr20yTCP2phUf7E7YmRRsjyuyyJBEhn5s0appM3NN6gK9WOGDoqg9WmkuDpO51OOsakiDohcb9mcVeemY1k4yeueou2ykbUynkaGoPCup5TBvMI0jsYYs0ozaC7CVFJ5bpEWk2vnH0/SeZH7pMdCO4tf+ExuT/zSZh0/kIHySx/6T/OyP87mpVTgv2t2ah5FwT7LZGRoMBoitlgTNB7LYkkUCN33ag7Z83AvhhQeWGtaPeG1nkYL/4CvX+C++8cbT7cgp7YUOBF3DeG9a8f3bB7jgKZtApBVJZEgjgwIuDRKS2HBlNePu/oxx2aCVYiWNuNCPGWQRP7xzwLXVlH5sqRrPYVHjvefuYSEOt53+62r+voXyNC4QW4H/pFZjlaJqg8F6TzMrG6oWJRSCOO6mlakM8Mirf44Gesz3rz1o5c/UfDvtDffSFqaQ8/Go43zSOVoWaj9vO8051UgW3iUe3eDV85Qu1QiX14VBzM3t/Ih6l1YLdS8VTlb2OutnARS1Zy2zbA6SeRl4o2+4P67OdG4eFXC758t2eM3yMMXFSe9rAvzj79/h99+8eu6rAXiBS0OwaBh/sDthVFT85ft77E1KURiraspGGr2jUhDEzokbvzxM+buvb/Iffu06l1dTnA98dqNHFhlmtYgR/MKVIUorEtsOhi1RVwQgiRSxUURGM6s8dSN6wlprjFK8eqHPhUHM2iAhi2QbPsgN6xDCOh8kM+hKGl1DLEJ4kR7lYJZ/7yCkp7WXQeD0ZtqfXiS9p7PYk4L4s7J+pFjvGVKrxLlz8k3vWQSBxCwSk/N0EMv70j1OSsedPdEiM0qOr1VgjWK9Z8msemIQSK0iRlbWqVXzORuNULH3IsWV1ZjLqylprBmXjlFRUTYiDTmrzhYEYJEMLP8dqaUpZyXBrOs/PsmMhp1xzX/1f/+Itz7cP+PenN1e6BVB1zB+e+uQW7sF00L0ADwiED1tRaJja1hJYl690GetHzHKay4PE35w54BZ2eBCYCWLGKQWhSavHW++sspfvPNAphdDIGjRNO4yqKoJksH4BtWtBqwmMoZZ1bA9KqhdaFcEnsZLU6lp04WgFCuJIa89Re3nFxJIVpRoRWSkfLXsvB/6/W97PegcTWnJTj/NVB2VCzgCoZU3bYlxjwwmdmZaZlwXnh+ccrmUGbc70CkIpkahlMIHT2wUe7OG2GgM4bHBoHYBY0Cj2OzHHOY1w9QSW83+tKJsAs7Des+Q14ZZU+O9Q2vNrWL6xLmdx32PzoxeEPp2aL1l5bGABDjFyauupk3gtscV/+jbH3JpJT3XOYIXekXQNYzfezDlMK+Y1Q5jNJE1REayuCZAXXsaPD/aOmRvUjErHR/s5fz6qxt845ev8dXPrLM3qfjOrX1+vHWIJjApG4ZZTGI1VuuHUD8+yPCZ0aqdZtRc6CVkkcE3nvvjiknVkNeOxouT9z7MswXnAqPCsdmLSIyal4w65TLng6g8PeEYeJhngo8zq2AQ6xf7gnjG1p2LT3OsrXy7umSh76vUyavE0CZJ5x0EFBAb6B/DVdYuUPlFyazxoR2O1OSNJEt143mSvo8QtQEhcFjU9GPDl6+usDlI0VqTRJpp1TAuGmIj37n20PjH00w86Tt1j7rdno3UQ9DRjiEgMYqk7WxHj/g+SsP+rObBKOetOwcfb8dOaS/0fb91kPPWnUOGaczFYYLRikBog4A0hMvaM3OOvXbYa1Y77h7kjPOaYRqhlWKQWCKrGcQSQN7bmfC9D/f5lWsrrRCN0EIkRpaDsZZpQKsVpROHTQhcWkmJjAjUGBXmkpVdxtDdgBoJUN4Hdqc1KEUvUiRm4XQaLzfOk6yNQURGPfKCM0pQDY2XktNLO709rxLP01gTxDHVQRwwKKxRD938z2tVo4B+YpkdQ0IcP47Ot7Tw3i9o4IM8/6TLVCFzPlZrXr3Q57MbfYraYbQi0gqNYlo5itoRG7lfn0bLoHtrbEQjHKCqF9QWy9aVqlKriY1mJTEn35utb/npvem5l4de6NLQW3cOGKaWV9YzDvOa7XE1P2NZZNBKhORdA7f3Zrz3YILWXS0vsDct2egn/PDOIfszmQ14bbPP3f2C97dnGOUxWpFZQ9N4Wa4qmUpWSrPai7G6Fu3iINl7XjlhPSRQVx6rVdubOMqeCHIDu1qmg5UW6girwnw5+aTaf9QS1HWBb1IdvdK75lvbGqHuvsMT7CUC6WH7tDfZu5JE3oBGkpDYAErNuXjO0zRt3wEpmQQv5ZtljZDj15WAMDwo5igiYJ5APcqskpmdtSwmiTQ705oHo5KNXsykaHgwLlAKYqXJa49VijQ2TMrmsV3oWEH1hEMVPFQswACd5OQybYanLdEpxWpmQSkGIbBfHL2CjFGspBat4Ad3R2wd5OdWHnqhA0GHGvrVz67xz350n41+xLhQ+BBoXGCQWPabihoYFdIETmOFUprdScW33t3htz5/gVv7M3zwjHPHt97bpRdbZlXFrA5UVUPeridjrebsnsGJKplViiRSzErHD27vSynIdTXO8MTG0XwpqSQ76ppNJkhweJz3ySJDWTuKOhBOcN3LGVBzyiDQ7dNLEzsevD+ttnypHKUxeT577hFFsYAkMIXzBPf4pMIqeV9qNb1IESnFuHb4Vgq2S3RM25TtNqa0YpQ7pmUuKl/A3d0pg9RKL8Jq2YdGdmiQ2bZC8Oj9jxX0M0M1e/gu6cpBkZHvKPrjMuHfqQx6ZH8jDcZo1nsJSaS5uiIsB3nl6EfhSL8p0oqVLEYrWMsi3rpz8DIQfBxTBP7NezvcGxUc5BU7Y6GRGMQGaxSNC0L5QJBSTJBySy/SOA8/vTdib1Zxe1caSMPU0osttXM8GFUoHagbIXfTBso2s298Oy3sPY2CpqWamJYNtTva6DrNbRhg3kBTtFA6/fharkI0kJXWWCU1Yt9uQ8+30RLieajb/sR5uYUOIfI8eWSep33aA8Hj9u95rWa6fXCAO6FD2j3Tyg+TxgbbIvPy2qG1YphYGh8oWkFvZQKmzexVaNF0be9suSFeeChmDYmBG+sZNzYG+ACTouT9vRwbBFbu2vtg+XgMYk0/iSjqRhymBq0kodQASqEIhCX4eEBhdZgTPHokmMTWYNtB1v/oV68zSCP+6v1dZtUBoxwa76hbdoPaBbwPfOnqkGtrPfam1bM5ESfYCxsItg5ydqY1H+zMeH97QhprLq+kcsKUYuugoGpkEGw+BIY0mcaF1OXHhWLYi4msoSwa6iZgjOfeuKBslpA8SmqRHukHaATuNkgjqsaTxoq1LKJoPJuJZntUnxkHfSTzDJDGmqZ89O3bi+DSSspHhwV5y68b6cVStQnSkFbAMNVUTcCdY4nA8+kunbzo9rhjHxtOPXl+PGg8C6bQZbMINXRXPm18wNeB1dTgUYyLBqXg2lpK0XhW0pi9aUntK8oG1AnqY11ZrFsJ3T3I2Z5UrPdiPnuhz0pUcFDIUFnnyLvvpZGS6aEvKZvFbE+kw3y1obTs36xyKC+lYa0CRbuxbqVhjEFpzcVBzLX1HluHBb+52QcFV1cSGu8xCia1p4sqqz3LII25tJKw0Y+f4ZF++Li/kPbWnQOur2V8J+wSR5q8khPdTww6QFk7IqupnX+oLglShjmY1bx+yfC5iwPujQoejAtGpbCRppGmaHzL8S9om9IH+rGm8YH1LObCIOHBpGBaNEzyhtI5Mm/Q7ZL2NG73pCzdI3QVj5ojiC28utnjg70C1TbHqyUn313s3U09KfwzH246LnjzLFcbL3sUZ7MnHa+yLWd0Ge3j7LiTPR4EnvbcOKCsA4PMSA09taylEVvjgmlZk7ZUDeO8oQmBfuxl3/0CjXR8MFi35VTfNpoBQu2YVQ3vPZgwKhq8D8Qt5rMKR8topQd97H5d5jOyXkAmCsUgDsyWEq9qaSXvnCd3jj0t+gLbk5IvXh4wSCzb05oLg4TIGCZFxdZhhcIzLRyfv9THav1i6RE8L+v6A4dFTWQ0/ThCq8CDcUFZO5wH5RdLt2XryjAhBD7YndCPItYHlqr2zGpHUfsjXCg+tMRSQRGQWuNmP6YJnnFe03hYzSJsrZhVzYmf+Sg76aZSyAVttZSInJMlcWwNjQ9EKvDBXiHlH6NRDmrlpJzULLbRDj/P+xDP0rrtJXpBdfGsnPeLHASOB9CnsUidTs8isAAOPK097SYCss+J0dQhMMobZlU7ib8ec2O9z939KduTihDg3qigcQsqlsqdkDi1T3RlVd2WaJRSTMuGxBjGTYPRkCWGunAEjq5+urmAbttdIuWQYBhqWMkERm60w/uHJ4gbL0lZ3nhGec3feX2DNLJksRUBKw1GCcdZZCCLLHXwZJHmd790+aUewcexOQ11kIEUa5TU532gqD2RVQtR6GM2r/MFeDAuWUkcdw/lkuhHGu8DZXuWBXjhqZ0mjQ0EhVWerXFBaGGjRiuCh0FimJbNx77Jpe4vTWLnpU4ZfGClJ9TYPrSfUTUc5g2EwLTwGCVTzt0+96zcHJWTElYI4VwEalpQCv0konKevHIvabCfYApxAs+CJNBoWEkjgSDzeFqGp1mhPmtzAcYthbzzgTS2GOXYOqi5vZcTCHgn95U1hthoIYx7xIpmnrkj16OUg+UGqF2gF7c6I94LcogW4qlAh8WcQdfnOqnMqbT0BeZMAwShgtfyuV1JVilF4zwHswqrFP/0R/e4uz+jcvL9un3d7MWsDxK+cGlwRMXwvOyFRY1/5foa46LhQj+mqj3joqJ2nsbJCZLantwsXWav1QLj22UmVmumpaMoXYtNNqz3Y9JIBkK0lgs30orXNvqtzYq2AAAgAElEQVS8eW3I9Y0emTVUTtg/I60ZZIYslnmEbsCra4o9yo7/t0Y0jb2THUytppdYEivlqBACr6xlTEtHCGCtafljApULRFZxsW/pp1JrjAzzZvF5mEMc2iiv51TcL+3RpoCNfkT0jChSCwe1c0Tmydw8p70EzvsUauSerF2Ywy+d8zwY10xKRwiBJDLUXiQo+4mQR1qt5o76UXoCpqXOAGQQ1CqMVhS1Y6VnudFm3B0VRMdC0Fl0DK/dVQU6n1E1vl3xe0rnWeslREYkYo1WGLWgty4bz//1w3v8+O4he9MKo2RjsVZksSFqKfG/cHnIMLXnPlD2wq4Irq5lfP2Ny9w9mBEU7Ewr6qYBlIjPaC1LylawVyN0AcsDQjK84glemrOxMaz3ImaVZ70fk1cOIoVRhhubGRcGKf1EJpfXewnb4xLnpXg5SCIAVrKGYSr46MbL0hcWF5/zj66rGy3zCf1E8MfOw0FeUXsZwbda8Vcf7FG7QGKhqmmnqIViN9KKv//mFb713p5csE7qoed9czeBp2cJ+1tgsZGg+bjSYaoFAXMaU8D4MYCCs9rTrgYMgrwrmoamhVYe316gI0sUiLTRMlgZkMSpn1h0S+zkgZ1JiVbMp/NBkqWOBLKfWPqxqBHWLqCVx6t2PkEKBlQusDMuuLKSkkWKaRnmKJ8jbAH+6GVs1aKh3HgI3mPNYjh0LdMUlSJoKd2mkaFomYrz0rOiNLPQkMaWJnjWejHTwrHei+nHli9cGbI/q/jsZp+dSfkUR/7J9sIGApBg8I03r/KPvv0hGy3nyCCp2Dos6EWGnWmNUQ7rPYPYUjSOWe3nPB/GyiBWIGCNJrKKzX7CpJwBMEwjrq9nDNKIV9ZSDmY1721PCKFTCAsc5DW9yDBVUqpxPkij2miurqTUbkZZe2KrSa1uFc8EPha1dUltRXDeGs2gHTCpvWeSN6wkFmM1Dw5LokgL7YSHWQVZFFABisZTO5Gw/N6Hh3z1xgp/8V7DXjtgdx7wwfOCJH7aB7eexk6qby+bVaCMxgb/UP35JAu0/StzskZ2Z8/jmCpE4/vaWtp+R8XW4YxR7o4AFbpHHwR2Hbcjt7EO9BKLDzBbKq8GwBqND5529ozKiaZIL7ZEWpHXjtgaPncp49buDK0ds8qTRoZ+bKh1w6Ty3DssMcpgTYMKR3srXT+g66mlVqOUR7tFcGiQfl2k5GdrVKGUMKpWLjCrGkEaha6xrRhXjtiKtrnwPMkRqBrfyuk2TMvmXBFD8IIHgo5i4ouXhzwYlfxka0xe1viWw+RLl/rszCoOZjWvbvYY5TUfHRQ4L8GAIKuENJbF31pP6oiD2FA6xWoWcXGQ8sa1Fd6+N+beoUhfaqVw3hNpyUwmvm4DidBUBB9II0MvsQzTCKNrNrKYyBi0rjic1VgjoLc4adOc4IlMS0znPGtZjPcyBDbOZepZt/2DuoXQTetAaiQzWskM19YyRmXDO9s5X72+wv/3sz1U8CRGcNjPEj26XNyItaJqYbVP63Be1CAAp0DsBObDi6c1YbxVaHc2YryTsv+Pc3l0jjOxgsffOiyIrUIrg/NweRi3zdNFbd6xKJv2Y41HUdQCJU2sIo40ZTsAGTyENggEYJgqageTwjEthT4iKMis5fb+TPiFrGK9H+HanmEThHfswiDitYsr/GTrkL1ZhSJgA2SxBIzKB7zzIjSFp6wfLt8GZPo4tTCrHMHLqka3TfuOzwngIK+JjEIrRWQ1u+MCoxX3DmTyeVI2vHlthXHR8Hde3/wYR//09kIHgo5iYphGDJKI7364T+0ccQjs5zX704osMSRGsXVYkFnNMDWMSvDO0Ys0aWwJBG6sZXz9jSv85N4YHwK28Xz+0pCvfXaNxBpK57m9P2UlsRzkNR5FbDXrPcvOpGaQWq4ME356f4I1imFmmVaOQWJY70mmk0aaYa/He24yh5harVEa1vsRK2nMej/m1s4U37a0GxcITmQ2q3ohjNPdyK5FNPViS1kHhqmF4Pn2rUNWMov3gdoHUgXeeSr3MNrh49hyAlq1ddFlR/RxygyG01eYzvLa52Gdw1g+N2e15eN3GnRRx1/zJE6q7n/TpSZ1B8VslpzXx9nvbv+KBhrfImN0RBwF0QtHFMJi3Q6a+cVKhrZU+oXLPW7uFNROnPhs1syRTomVeyBuabOLWga7uqnkDvtvadA6wofAuAzUjSeODMNE0ID9xOBQ3N2fMCokWQxtKSi2WlYtPkjShHvicejk0rs+x0lWe2mGj4pKVjp1IIsUmRX6+sOi5sP9nP/k+uq5IobgBQ8E729PGOUN06rhw70Zs7IhBEgiy+bA8u6DCZXzfO7SgJUk4mf3xxitubYi9fdJ5chizZVhwr/3K9f5B7/52XaVccD72xMO85r9ac1rF2P+4GvXuXeYkxjDzZ0pe9MS15aUJPO37OUV1irKmSOvGl7dGPDZC312JgUf7MyIreFrn1nnH/zadf7i3V3u7OfcPczpx5ZhbDBGM85rFIFx3rSMp57SBbR380njTrcVWrKxKmBUTWw1G72MuwdTauf58uVVbu3NiIPcjIVyoDwpMDlnRZLTOpVuClqxoPE9zfvCJzBs0PHPn6RLa7QwTpZNmL8GloAJPHl3rVoEaaMhixSj8tHv6seKWSkH7XEBIyBBwBrDaiRlCa1lVRuC9K5UWHD8POoTn/QdXBtUppVjWtUkVqOQ1XntpYwJ4uCHiZXJYa0Y5Q2vXcy4u5dzWNRSOkV6AWlkmZY1WWTIa4dSsJrFREazOy2JtGJWOZmVcH5+b0gZTurxq72IfmIY5zV7U0dZ+yNDpuOZoHliqwjWstqzTCtPhZ8HyO44csLvcHL5rYNwT0s/537SLcb1wjBhrR/hgygtfuUxx/VZ2AsbCLYOcj7cm2G0Yi2LuXeYQ1CsphEuSBROIk0/sfz6q5u8fW/EWj9GKUESNE6mbDTw5rU1fvdLl9g6yPnnP76P84FxUbM3rTjMa37rc5tcXct489oK37l1QOk819YyGg+39ia8dqHHqxf6vL015muf2eBn9w45zBvyxrMzzrl9UHJjs8dXb6yxdVjw//zkAQr4/KUBzntcEDjTa5sZ25MaheLeKKfxgV5sqJowJw5TnIwbL11oMx3PtHRoJRC9vHLC7hiC0HF7nkjz+yTTCELjcbC3s6BUTDuYs7wMf5LTeYba40+0LntOI8O0cvTbXk3ZBOJoQSoIQgDofBDkWFvGO6lp+iiL25Vi0pZXrGrm9NJ1WEzRGi2omhYLceLxWnZOSaTpxUYoqlvn65e21wVikJXGoypUbRWV+oR+R6AdWvOeJBL69r1ZM59hCHTiMQptYDWNcd6zl9dopRiXDqUU61nEqBAyx7pxND4I3DS07Lkh0E8Ns8riggg8+fY4WG2ItSRNWWT4hasrjPOaw7zmIK9xgTlLcQfcqNrj0NSB2NXo1MzneE4Luz6+mlPHnlvrRUJA6aGoG+6Nira3obm5PT3dhzyFvbCB4K07B3zp8grvPJhQNA7TagBPKsdvvLbBdz7YY7OvUUoawi4ELg5j9iY1VgtyINaKNDKstI2at+4c4Hzg+7f3mZSNlGUI/A/fvMk//EbK737pMjvjiq2DnPd3JgDS6OnF/OX7e21D2PBLr6zz0/sjtIb3tmf80vVVvnB5yM2dGVks8NDae6wxfOXGOpHVEASj3Eukrmjuwp29nHHZIGNsPHY6OAThLXnn/pjDQobc3r43xvmANQrVwkgjK43Fpxkw0houDARVdZgvhnNMm9GedbOdkziecZ11VdFl089qsZC0jLBF08mUSumg9gGNIosko63r0JY9ApERpxuCIFdOW75SMHfqkZLs/c1XVri9O2NvVlE1gSiEudNTS99UBqgUZb3gueqOQRcMIqMZJBH7s4qiahZEbu1ndqsBTdtzaJE8y8fRGtkOCOQzcPS86fY1tYdQe1S7BOj2tzunLrRKfAZmpUAx740LEqsoGs+slpW9Ue1Er5Iaf2JlxmdaOZjWbA4i7o8L0QbREoRK57FGE1t5/cVhQj/WfLA3pVgifOv2BY5m/KWHJndHVt0f51pavn6FNyzMB1WlIS7f65W1lFt7s3NlHoUXOBDsTSteWc/Ymxb8q3d3uTcq8N5zaSVlkFhqL5QTl4Yp72+P2drPmVQiOP/L19d44+oqpiW86seG//lbN7k/KtidlEwrTz+y3DmY8mBUULnAX72/w+//4hWurWVksaFXR6jg2Z9VvPNgglbw2c0+7+9MuLaa8bXPbPArn1njz374Eb/3xhW+f3ufLNZkkcWFgNaQxZq68eSVJ40Uo8IRGcVH+wVlE7i6luIPckR4TbL66gTPEhu56GaVI6e9iJ0EB3FIARc6egqD9xI4u3mLs1rj4cHhgiDLArblj1l2CrBwFI9yzoHFDaeQ8ogLC5hv1wvosnKjFQqF1bIKalz7HY9t/Fk0rgMCT4wjT/CBWeWJrcI3ARsp0siSVzWRlWMQAmil8UHhfUua9ojvfdJndVYHmOQN3765R2YNmTX40FA3i1kYa9QcfZbFQsuQxoFpm8KGpe0qYG8qiU0kGGVqLxBkF2S/FcxLN94tVp/d5/VTQ6S1JDBaS+IRFv2Jrs/g2xWQc5CaMF9dDBIZiNQtlj6NrAxFakWmZUKgFxvipmF/WtOPNXnj5jrHIGi8NJJAXDvP5iBhf1ahcO35FpCGKAE68sazPy0pm0BeuRMd//FjD22gOvbk41ZJx617WbcyiK2oG4bgqZ2ar8p6saiqffHy4FyZR+EFDgQb/Zgff3TIv3lvj7U0Yv2VFd57MGVnUvHNnz2gboQqYhw18xplUTuC8xRVw0/vj7g8THnj2pB37k9ovOfKSsb3P9wnryQryVutAO9he1rzf3zvDlfWM/pxzJVhzDsPpjIm3jg8sDupOFAVHx3k/NbnNrmzP+PySsq0bOZkWlv7E+7sT8nrwEeHBXXjubKSUjnPILH83dc32J1WxEbxYFTgXKBqAllkKBqP9X6ufQxt9qg1caSZFg2xFYSQbjMko9tp5XYkvvEeo2VQzuiTJQ1PY8vxyAM+hPn2uqnNuYPhFPVx5IVBQWw0oUV2DRJhks0rUZ/zHYtkEPZXpRbzGbC4+Y4PDCkWK5ZTm1JMq4bEWq6sJdzam8rnKOHD70UG52WiWpTBFFmiKWtpyoew4MY5qwVkalicoKJwjmBbzH1oJSq9OCilZPWRRAaNP/Kdu20FYFw6IuPmqLOyaVeGMKcsj9uhqnY8RuiXrZLEwbX3Q+jCi3+or+NZBOHGLbiKpqUkKB5QkSYQ0O0QVuUCiaHN7A1p7LnYj7k3KvFKlotGa9b7llfWMu6NCmaV4/6o4OowZTOzbI0bUJ5YSwWg308YRJq3bh/OaV+saWeJulLSY8xxNJE4C21MZx65rkXWVgAfKKhqTy+2vHqhx2Y/4vp679znCD6RyWKl1N9XSv1UKfWuUuq/PI/P+Mr1Nf71uzuMypqdacn+tBbd4ciQO89vf+4i672IvGq4PypIreHyMCaNLe9uz5gUDVoHDmYNWsPFobD/TUpP0QYRhWSrnfRfFRR39nJ6seLuQYEn8PrFPl+6ukJmDYdFTVE7NvvSp/juBwf8xmfXGRcNZe342b0xu9OSvJ2E3trP2RmXvHN/wtZBjjWaW3sFVd2wP6vZmZSEEBhmFtsyb3XTzla14t9WKDaKUviVfJupdXqqHTqjHxshzqsDeeOZVv6pJAuXs6tuBdI5j27M37P4eWIgsO1AXZAhvybIjXthmLA5iFkfxFwaJjLijzi+rkzR7UOk22Zuu02ps8vg0KOmUR9lBqllN17qu01bDlLI99yf1Nw9KJiUbo68cSEwKRx55VtOGuYTsac1jTR2B6mQF94fV+xMKqFN0RqjNEYprBIG2iSSVWVAAuPxWrVZ+nAfJFPvnKBHVh9dEAhI7X8lswwT6ScMUsPrF/vERr799fWMNDIMU0NqDVYtSlCZVUQWVlLDWmrmvSjV/tM1XmeVZ1w4rq6mrGYRjQvszSr2ZgXTStgC9mc1Say5PEhII0MWR6xlEfuzRggZA1waJLx+acDVtT4Ez2oSEVvNSmrpR5qygaJ2JJEmjURLRPmHIaEn2XHHf9ZbpUs8Iqvpx7ZtogsMNokMl4YpV1czrqxmL+YcgVLKAP8d8HXgDvBtpdT/GUL48bP+rHFRkxghgZMJQs+NzR6ND/zuly8zqxvuHeS8v5szTC0hBK6tpRzkNeu9mNt7OWUd2vFwzY/u7rCSGe6PZPpz+SYyRknTzgtVbu5k2vfm9pQstgStuLGWsj2p2MtrdiYVb1wdUgf4+huX+dn9EaO8YVLWlJVrOUna3kYI4kido6gb/vr2iIuDiNVezOGspsobGU83mhtrKXvTmlHRUiB6KftksaIXS4OucvV8aC4Euagb3ENi96dNjk/K6JeX1cuQ1naXzlRXNcg+xlamNLtaboQ0aAeJ5dJQMS4bqsaxN5MpUqvDfB8Grc7nrHR41QYlBcNEiPq892fidNHtsmajZ0is4qPDAqMhLxd89s4dLXvpAI6lOrQHp8NDTcfHrZC6oDktRV2vyyQnBUy9Y5gY6qDQIXBpGKOVYndWC/PlsdqFQ1g1jzx3wrnpmtFdPXtSei4NYtHfBsaF47ULPa6uZqz1Eu4fzvjunQNc5bk4FARP2Xg2etLkPcybI9+xo2qYHysl33QYW2ZlwWpq2BkLbUrTOCal0EK8eiGj8YosllLQ/VHFtPb0Y8MgMTyYVNzez7kwiNkYJOR1Q6wNQSkO85oQNJ4OcirHsUBWNY9r3nQlsY6t9aw9gnk/L0gvMGkTiK5vpJVmb1ryzr0xv/OlCy/sHMFvAO+GEN4HUEr9b8C/DzzTQPDWnQOurmbMKkftQ8sD4ngwKnj94hCAfhzhKVnNrJQbgmJcei4OYiGl85pRXjFo4WiowNXVjMO8oa7reVYHwhoIgX5iGZeOpKV+nlZOuFEax7ujgiwyfHElY60X8f6DSQs1rRgXDdc3Mj7c9RzmFVqLUlrtA1oFZs7xvduH9GLLaFYyLWTozLcSmK4drNmd1qz3IyHksoZLg5iPDguurqSUruHmzvSI/itIll0tsZIev7A/TnN1+T3HH09TDuoGkRSQRsLREmndqrQFTJCm5e64YHsCqRUaAavU3HF5FEZJIJQ+i55PbDuksbmSRoTgW677mqLynAY5W3toCkfZIjsuDhK2pxVJJN6h8Uc1qbty2HLJwSFB7ngZ4nEfb9SilNQ0IqKklCKKZOhqWjl6sXDxbE8qbmz0ubqi2JkIIqY5oUfQHWutFk3V5cAUkACUGinBBBVQRvOFC32GSYTWit/+/AW0gn/xk/vc3J2SGU1hpCey1osxCg4LmajvuIGUapiU/qGkyiqIjGFrXEgPJI6ondCp0B7DL18dcnUt46ODQhyzVtyuHL3ItCsMIZKblg1393OGmWWcNygctZNVWj/WbA4T9qe1IHSMEhqJx5yA7tqN9OK8dSJXp0UQdde+RxKSul0qr/WM/O08OxOZ+v/OzT3++Pe//ELOEbwC3F76+w7wm8/6Q/amFb9wdcg/eWuLLLH02one7XHF7/9iD4BZ1bA9KQgBDvOKYUvfoHXE5ZWMX76+yr99bxetFahAL7LMcFxdSbEaDlricYVkejrIsjmvHHtlTekDmZV69sGsRmvY6EVE1vDB9pS8dmz2RbdglNeMy4arqz3285qqrSc4H2hUaFcmjrx2OCdSf7pefF8FEGQVZKyUBxrnqV3g8mrCILWUE4dH1JQ600gfwUSKgRFxDR/EeXYX+vJ9sTzIZLUEoFYP/Uh9/XHObDkgPOq13T3VISrWM4O2lqoO+CDlOgjMGt+iphRJJDdS1pYZZuUC521UEOUnrbAGev9/e2caJNd13fffuW/vfXYMMAABUBRIiCIoERJtybIp0VosK46r4siiK05VXBUl5TixKuVyZfng5Gs+xClXJU4U2SknUWg7XhInkhdVJIZSREtczH0nAIIABpjBzPTer9928+G+7mkOBhsXzZB4vyoUenp6+t1+7/U9957lf1wTlA9cC1vZzNY81joR55p9NgbXtjXQmEn59MaQsgO9eNPdIvkiwbcUvcjEKkYTzCiLyWTLCEppPDZX3uFEC0c7f6DU9n7oYWLUZT3LFBSOGr6LmPNmqs5NbYDR3DfxsEl3j2Zid3iFxi5JCok2AfFhlPDChQ5V36HmWfzJE0PeM1fhfCvEVgplC/NVH1cpbp6rUPEsLnYjXlntstd3aIUJSgQlKUqZHY6ljKHxXUXJsRlGKSvtIXvqPhVfEWUmMO7aZpf81Jk2rmXu1cVKmeXWgKpnsdFPaPZMZXB/mORBaFOEFqUpeYdKhonmfCtEiVD2TN/xqm/RHabjxdLWczHe2QqUHBNvS7NrDxRvfc9RlpBlmRJoLSYOUnaFuapPN8p48MVV5mv+uy5YvJ0L7pK5QES+CHwR4MCBA9d9kOmyy0sXOhzb3+DExZ7JGtLGp7nSibjYDTlxsU/Nc6iXHE6v9+kOE/bUfGyluHN/A8cSSp7FHUsNTq/3KLk23Sjl6L4acVplpdnnyXNtRDQl20ZZimY/ZqHmGhmLLEVhGth4tkXVs1CWcfmcaw4YJim+q2j2IyNiF6dEuTBemqsv6gwinY2rLkd+bnh9oHMkWpcBg2GKbwupFpqDGM9RiBYT19DGxTJSTEVGaXgZJUdM1kWqkdSsjEfHGgX4tgbIFOA6Yvo/9+NtM3QuubZcmkd9udeV8sT0YaaoizBdd8m0R6sf048SojRFiaIfpcxWHNa6Ee1hxqRQpCbvyJYH97UWLGWyW8I4ZU/dp2QrWpbJ2roWQzBaRY8+am/kicP4mTVgZYwcbqYgK//9qEBOZKQfZBHYikwwRjpMEDFB5fGqU8CyyZsp6bwPdW5Q0ERxNk757cUZKo+hrPeG1AJ3HHS1cimSrbuQyU+cr3tMcF6b2geVuydVZibrfmQCnLGdsZGk1DScvNijPUjYNxUQpSkl16YTJpy42GWh5vHxIwt4jqKSC8eZxU/M02c7QIJrmcLLQd40ptmPEWAQpfSGCf1hQt/K8G2L/dMlLFGcWusxUzZBc88S1noxrmUxTBJag8hUtWeabqTz+gsL0WZHkGpNmpjzlGaaSuBw96EZHj55kbPNaNsFysi3HydgSWZctmzXEfzqKBhrDDmWSYX1bYVYQpxmnGn2mSq7PHZ6g9mqy313H3wDR7k2dsIQnAH2T/y8BJzb+iKt9ZeBLwMcP378us/zsaUGX3tymcCzmKl4zFY9tIZ6YHPiYo8ozUzhijK5lQemS7TDhLJrikwcy/j6D0wFvHC+TZKZlfWti1XONQdEacbPf/QwizWf5XbIei/iT588xzA1BVu+kzHjuYiYUvGlKZ9BnGKjOLMxINVGQtdRiv/1xDmGaYqthCQVZko2G0CrH70u/x02rehk+qOw2aDGyYPDg0QouzYippuZo0x6qWsL3bEipck2yrTGFmgNYpOTTa7+OOHD0eblr2O0I4hTkzpZ900/2eZVxPSvz6+qODjtc2p9wMYgoxXGeJZirR+PV2E2KbZtJoGRzO92I9D58zrV2GmKaNjopQyjhPWSR8lRaOSaUksnff/kj0uOohdnY5dPmq+w4dJOXmP3g5iV88HZEqcu9uiGMfWSR5pkOJYx3kk+gaeZxlaK+arHuWafKDVdteIoI80Hk5EH1MnvB0zMojM0WWlZpq9oqCW/3/y8JiBwLPpxhk5NEeJI8hwxSbuNwKEfpdR9B5TgOsLZZg/bsljrxaSZ2ZVWfIeji7Wxam/Vd1jvRZxa66K15vllTaI1SQaLdY+VtsmSSTO40AmJcn+do81k+eKFHo4lDOOUjZ5G6x7tMCFKzb1c9S1a/STvSw6OCBkZlqVI401JaS2jDDbBUYrBMKYVJibILZcqjo4WJ1GS0YvfmNt09D6Bo7AtUxuRYep8OsMUS4y7ycFCNLTDhIdOrL+tzWl2whA8DNwiIoeAs8AXgJ97qw+y2Aj4yM0z/MWzF0jzlMr2IOZss0/g2Kz3IuaqPqudCDvXAtda0xrE1AKbwLU4PFvmwRdXaYcJdd8hjFOeW+5waK7Mz3xwaXxRRuXfD7xwgRIOG/0uVd9lmCQErpCmGYFj0+wnzE87BI7NudaAMM5lIXRKd2hWPZ6juH2xRgY89Mpa3kpT6IZxLimRV6dq/brUznEVZGJuqEbZxbOt3D2Usa/hE6VpXjltYhpKmfxlEdhb9+kOUzrDmFz/C0dtujAmjc5oNT/ZrKPiKhOsTTWugLKMvsybQcT0iz29HuY7IaHVT0hJX7etTIAs0ZDLbVzty2liKkKj6tAOE5phwkI9YK7mUy85XOyGV02b3WrMBGMQR7EAwZy3UdrjpGTBZGHbIM443+ozV3VJMs0wg/fNl2mFMWvdmDgdolNTJZxqiPoJNd8kCsS5/v1ICnnsTpPN4yiBTmhWt0YI7ernnNw9FaYZrqM4NFMaB3k9SxHpDJ27LY3rxfT4GAXuX2mFzFQVUZwxTDKGacpGN+Rf/fnz/MJHDrKS3xiNksMtVpU9tYBfvvcIX/nOCS60Q841Q7rD2Oy6lOmtPdrtRKlxhzmWaSKj0az1EzpRYorqlNCNMqZtm6myR6pTsjw1zlKCLYKfu5ZUnsPsuzZl12YQJzz48hphvsIwq3V9yc3Ui7Lx/Vf1Td+AfqwvuR+2xsQmbymN2ZEMowxHYJDrwVv5LmyYapSkrHSHNAcxi3Xvba0l+IGnj2qtE+CXgD8HngN+X2v9zNtxrHuOzFP1bWbKLiutkCQzLpqSq3jmXJthmvHePRUc2wQjq77N+/fV+MWP38Jnbl9kuR2yNF3i7kPT+I5FnGrqgc1s2bnkgiw3B/SijM4wpuIZ+zpyWVfJPLkAAByBSURBVFQ8m2GaMl9zWWkPOb3Rw1FiBODI6AxTojRjoeoxFTi8fNHko89VTPxgruri22ocyMu0Hmuvw+sDfaaKVNHwbeI0ozNMSDPNSnuIZysOz1Wo+w4aI32rRAgci5mqT9W32dMIWKy5eDZj90q6ZdYbBRg9y6yOUg0l38Z37LGf+0qyx9dKpmGQB/mjXFhvMqio2BRWM3npl1a7bof5YmpWOhGOUlR9h31TAfunyoSRsajb+S+vhDmu8c2DSdsUNovfLoeZ3DRrvci4BbKMk2s9Tq31aQ+G5oKrvNtb/jlfXR+ANiq40yWXWmBRL9nj6zJaMIBp+RjGJqhsW/b4fF2OSQOogFY/4fkLHc53QlwL+kkyTj2OE825ttmZWJZwcKZEcxAjStEbxGz0h2z0o1xYzWTEfeP5VY4t1bnQGvBf/vIU93//VV443+LF8y1Or/c5udqn2Y9QebXw6DrY1mabyTQzBma0incsRZaZq5pk5txXSzafOjqPa9kcWahyeK5C2bNN/xFthBYDx6JRMtlPnWFsquvTzCw6tHEXTTKa0EfxM7MIMnGmrXE0NTH2y91LUWKE7YZ5ajUCJc9GKcmDxsYNl2SaV9f6PPna29ecZkcKyrTWXwe+/nYfZ7ER8MOHp/nq915DWYJjWei8D6oCXrrQYWkqoOTanFlv0w5jwiTlidMbHDswNe57rESYLnuAmYS3K+544kyTDx+c4tsvruHaimGSmFTBDD5xZJaT6wN8ywTFTOMYi8ARWrkRqHoOR/fWKXs2p9a61AObMM3IUtNHdbHuc7bZHwt3OXmLydFOQGO+9BmawFGcb4eICL5rGZeNNjferYslPEvx9Lm2+QKLMQgbg5jZqkecaqZnPR5/bZ3WIBkH1sbXjterPo5WT2udKNfR4ZI01DdLlGl0rImTiCQzaaMpjFt2jkjItW64ugaMiPmy9eOUmmdzthmy3os52xoQeBbtbaq8tsYF4NId0ohRBa/Sl0+XHcd6MuiFMUqMP3yYZigRehPpS468fkLZ6CUs1C0aJQcn72MRODIuJBsZhVFjGssSLKB3leuSsRm0llxGQWnzf09ylc5MEyZQyh/blpFbD/OiR61hudVHoUyjJmXco75j0xnEfPeVVZ4+2+XAdIm67/DU2Q2+9uR5lDLutThTJIlxZUp+Xkc9r33bVN36jqkHONsMsSVjkFdFZ3kG0vlmSCuMWay77Jsq0Q0T1rpDrJLNMBbIs6vuPjTDI6fWzDVIzQKr7No4qaa/JZ9YY86JZOZ/T4nJTFMK3xLC3PqO3HOT2UHb4eZZFsPR32WYeSNPFogzjatgvhZQDRxeXule+eK9CXakoOwHyT1HFsYFYcMkoZ0LTNmW0B3EnNsY8PCpdVIN+6ZK+LbFlx88yROnNzb7Hk9wueIOk6VU59O3L9AouUBGHKdYSji9MeCmGZ962efY0hSLdR/bFla6EVprFms+x5bqVHyHi70hrUHMMNH80MEZZiouojX92GgPjYKMcarHKzPXMtvYUVWt71hkCDXfYTpwWe8lqDw+8uL5HiBMlxwCR1HxTWpsGKUMopTuMOHFC22ag4Stu+LRDmEk95tmm7uF9jBlrZ8Q60uDym8Fo0k2A2I2YxZb57Ust4rbrcImQh4MYk2ms7yrlHBqrcer6/08c8Qeu1de996Yld5s2WGubFP3TcGUa196nBHJyG9/hc+WYlI9l1uRmbCVyqtzX39syIsElZkMB8OUk2t9emFMN4xzt+DmanS0U6wFRpiwG6XX1tBGb9ZujI6tMQ2O2sMYnWkagY1liilYqLk4lrDcHNAouRjJNkGUqWoPk4w0he4wphPGPHRinUbZZqrk0YtSTqwO8Bw1TvkdxhlhYhrSR/nuzBaT7mmJwrU25RfiNGOQZzVEcWZ2C1FGqjNWuxGfum2es62Q51c6ZJkpMvQ9m/fMlan6Ns8ut1jrx7iWcR3NlF0z3izbNpV0rMaK8eOnmR630LTZrFwHs0O35fLXPkyNC2hkLFJgGJsFjcZcg2Gi8wD1KIL/9vCuNwSLjYD3762x0g5pDxIspZgpu2Y7p+CF8x3KrkWmTYrnfNWnUbb5+tPL477HnTA2OuZhTCdMOLbUuOQ4I6NxaLbCvbcucOf+GX7kvXPce9sC9ZLDi+d7ZFnGbNXj5rkqdd/BUoowzpipuGRa89xyi0dPbRAnGt9W3LG/gWcL59tmi621xrZMFXDJUXnDbMaTv50LnmUZLFQ9ZqsuUWYaevSGxhfe7Ed0h7HZKQSmGne26uM5ik5kAtbNfoRkm9vfUZOQ0fpoNNFbuTtqNFG8HbfpaPIGXqejZHyslxqqDKMW6SioupvpsaMvqD2S1MgfDxPzRS47ikGU0ItTktwCTgbmJ91kgzhloR5wZKHKbM0DFCVbcJWRmvYmmlGPjnvVc5NHn4eJphY4LNZL+JPWSBsDoPMTneQ9NbIsM92vYj0W9BudF8+Cmm+zNBWg1Ga/3Ku5vS7XyyHVkCQmK6nk2cxWXGYrLoLFnprHIE5ZrAWk2tR1hFHGRt/0qy67RhBurRfR7EcmuAysdEISnae1phlhbHbIk/IjIlANbHzHwrKEiudwz3tn6Q7Nqj7JNivkR66xmZLLnorL42e6zFVcsiRjpRux2gmZCRzqJZ8k06z3IpNCbCuUGOl5lS+0LAU1T1F1FTVPjSf1kVs205o00QS2UPJtLMs04Kn7Fp5tXFSj+8i1zD25tYJ967UY/X70fcrQrHQjSq7FodnKVa7cG+ddbwgAPn/8AMNY0yg5zFRcjKqz4r2zZS50QrQ2NQLTJYeVzhAlYnRK8r7HgWtxsTskcC0+eXT7yP2k0Th5sYOIRmvh8FyZuapHlKW08rqDim+zdyrgY7fM8Ikjc6SZ5qmzLdDQCIyPsJNXBluWxXzVw7FML+TZakAt8FBKYSnJfb7G12/K/KGfZMyU3Xz1pLh5LiDLndcjqeQsg8WGz537p/i5u29iqREQJxlTJddUnE5M8JOrSD+fEW0xrgGRidxqNlVGR2zp933dbPWzOhY08pX4CIVxVbm2MsqsMnIfOCzWPZamS+yt+wSuoh64eLYQuIJlGdeK51hoMTGYuu9iiyJwN6stbGUqs0dGseRaxGnGej/m1vkyYGQ7RsdP8oC1WTXmVchXwbZkU45BTJFR2TM+/ZEh0bm7xrFM3UDVt3Fty0xejNIR894HChbrPuU8k6vsGRmSUaxp65BsMZNeYMvmjmsbjIvdFGjOlD32T5dANM8ud6h5DmXfohG4eLbCs809anrymr7dFc+mUXJp5ff3IDJNaZphjGdbeSaUjFNrq67Cs4VBnLFQ8/nUbQscWaxhWUaEbrEemHhWrkvkOIqZssNdB2d4ebVHo2yzNFUGpWgELo2SRzNMqPg2I7N41/4GcZ6lpUSRJGa3u1jz2T9d4f376uypB0aE0lHcNFNiuuyilDmRCTBf9Zkq+8xVzTn3HWOAbWvz/rVyd+QkW89zmi9iHGUytKYChzjNONsK+ezti1e/kd4g71rRuUmOHZji6N4aZ5t9OsOYsmdS2da7EbXAYaFWouKbUxElKacu9ji6twaYHcW1ROpHRuOJM02WWyF7agGH58p5bEF48rUWa/2Iw1oTJkZvZmm+wmLD58BMmRcvdEgyjW0rGiWHiutw6mKfME5ZbPimh8IgNsVKmFWPYwmpmInGsVQusGZT9ixuXaxx8mLPTCaWotlLcJRJKU0yzZHFKrNlj3YYM112ma/57J8uUwtsXrrQoZtp1JZgmQZ81yaLjPZ7lPtlfTvXSdGMFVujJB2Ldzli8t9d20hcJDqlF5qV18gfmpC7PDBSv/B6n3zZFdCCUlD2HKYrPuu9YS6vDbWSTbMfkyQ674egCFybO/bXOLs+wLUt6iUHETjf1EyVXZSSPICYUfYtTl3sEbhGObMeuPSjoTFygmknKBrPNc1QlCgaFQtRFjfPlTh1cUDgWPiOxYVOOPYRO5bC1hotmlF5wnY7BCWCsjQ2JjWx5FpYyhj3RqDohAlRZnSRpsom28mzbQb5tbDy3aFSQtUxDeIHiea2PRVSLShC+sOEwLUYRqnpjZ1tjqdesvN7I4NeTJhcPjc+cMwEn2rN4ZkyYZxytjmgHjh4tuLmuRJK4ExzgKDxbZt+lKKU8PFb52kPYl5dGwCMtaHSFOYqDqttU+CllFmRe67NrG+jdcYnbl1goz/kUNXjppkySvL+xI7FamdImosMHpyp4DsWUZqOdx4j8UEl5PU6KVlqdt5Rqjk0W2atF5ENYsQS/DxYvVDzCFzTTbDqm06CRsZaMV1y6Q4TwjijO4wJHMVGPyZwLWq+jWvZ1EvGU2By/wQHbZrkTNzb5vqbydiyjXsqTTWdKCHM74XFmsexA1NXnYfeKDeEIQC458gsj7zapBG4+I5xyby80uWOfTXWumZ1UnJNEc/GIH5D1nfSaIxypcG4jY4fnOLhVzdY6YTMVlyW5itYSji21OBbL6zwwzfPokRY70U8/loTTUY7NCXzrUHC3kYJJQMGUcogSan4NoKpNZguezQChzBJcWwLP+9BPFtxafUTU8cwU2Kx5jNT8VhuDbBE0RxEVD3HVCMrs3s5MF3mQivkgRdXKbkmsJ5B3jbTfM6lekArjLEtC98W+lFCP0+Dq+TaPbZlMnwybXoTJPnefabsorXmtY2BSc2zIdEpVmp60VY9x0gD5ymDjm2yQkp5L0LHVqbLV6apehb9KCORjJrnkGVGRMx1zEpqTz2gEbjYs4q7D83w9LkWrmUC6d0wMbGYPI7jKsVc2aNWcmj2Y+aqxoAKxjeeac10xaLi2izUApamAs5s9FntDfncHXt5brnD8+eNMS85FqAJXAfQtIcptjIZKaMA7KQxcARmKm4+gVica0amslzg/Xtr2I7CtSxOXexScu28GMxUw4ZJhq2M8mqSJyH4rkWYJsRpxlwtAJ3RCWMjwWwp9s37vLYe0h5ERofKMZPavinTt1tjEhFW2tElsR7fAkQ4MF3iwHQJrU1w9UdvmeWum6ap+g7/98UVbl+q0x8mtIcJVd/h4FyZmbLDQs3n4GyZT79vD19/epnVTojn2Hxgf0Av1qx1hyiEI3uqaJ2ZYLdA2XXZ6A9p9hI+f9d+ltshwzjjpZUuUyWHkqvY6MaEacrh2TKH5sqsdYe0wpipksnG60YJg2GKJWbnNlMxWkytMKYeuMxVfXrDlH5k7otWaCrYh3FKZ2h2EZYIh2crnM+Vf93MGKJ+nOQLH5ONtFD1cpeOzVRgc7Y9JMtMIykVm7225GnrSWJcYb6nsBC0FkRplhoB87WAPVUP52rpXm+SG8gQmKYx6/2I1sDo+i9NBXzgwBRJmvHoq03W+xElx+Inb194U9b32FKDbzx7ATB69SbA7PGrnzoyLj6bLrscW2qw2AjG8YWq7zBddrlzf4PnlltoNB8+NMUjr7ZM6XtoE8ZGM+f2vTW6YcILF4zEQpJlzFZ8Zisuh2bLDOKUTig0Si63LZrdzeOvtWgOIhZqHvPVgBcutMc1E/d96ABPnGnRCWM+dGia5863WetFRpoiyXAtReBZ1D2bmYoPrQGOElOboISpwGaYZXn6nTBb88aT+b5GwJlmOHZHjCS1D86UEOCVi/1ci10TZxmNksueukeqhSPzFfpxylNnWri2+TxTZYf+0Ij6nWkNEG1yy5VoaoHLTMVh/1SJj75nDscSAtcy6cB5hzmAnm8CfKc3BuzPDbithOmKx0/fuY/psseJ1S7ffP4CSpnKaWOMoOo7ZGgOzJT48aPz3Hf3QZabAx544QJPn2uz2glZbg65eb7M3nrAKxd7PHO2xf4pxdnmkDjLSBI9dqvdNF3iw4dmmK16nG+H3HWTs5kKCjQCl0NzZRZrPk+caZFkGX95Yo3nznVQOstTDI3UROAoojRlKnB4395anl1m8fnj+5mvuPzPJ5dZ7QyZqbjsrXtkKG7ba3aHzUFEzbcJXMXJtQFe3kMizccSOOSd9xQ/9t559k0FYwn1Y0t1njjTAsxioB0mHFmsIcBMxUdjJruRgNpiIxh/x+7/3qvjHfHNsyVeWu3huxZVz2MOzUsXejQCl4pn8/m79nPswBTzzQEr7QvcMl/BEs1T59pUA5u/dmSRhVpAJ0z42eNL/NFfLQNwcLbEI6c2QIQPH5wyO6NEc3imRHMQmdjFIMUS4ZNHF+hFCd99eY3e0MwVe2oebp6d1Rsm1H2b0xsDbBFuWSiz3otoD1NmSg6BZ6PzhIwD0wFhrKn6A1Y6w3xOMEknpnBMEdvme/L+fTUu9iLOd4a4ymKhHrCn6pNozQdyD8XbhWh9uQ3g7uH48eP6kUceedPvM+o3PJqIR1+sqm+PJ+xOmFw2DvBmjjWa9C/32m88e+Gy43ji9AZff3qZkxd7hHnzjbmqz+17a8xXPL7x/Cppppkpu8zXPGyl+OTRBYDXve+ZjT4vXuhy03SJQ3PlS8Y0OeZWf8hz57tkmeZ8OyRKzUSDNqmI02WXum8TJpqji1UE+PZLq6z2IqLYFCLtnyrxcx/az9nWkI3+kMde3eC15gDRwqeOzvO3P3Joy+frghbeM19h/3SJZn/ImabRgjowFVAvOWz0TdbX5OT47HKL753Y4Fy7TxhlHFmocNfBGTxbXXI9RxP2QyfWmSm7LNZ9llsha72I2/ZUSVLN0nRpfB3OrPeZrXo0+zGvrvd570KFpanSVe+V0We60A5ZqPl8+KYpYg2Pn97gxGoPRDg0Wxo/fy33yeQ1OrHa5aXzHU6s9VhuDojTjKrv4LsWZcfi+E3T/MLHDm9b7zL6+9PrfRbrPmGccrEbYSnhvg8ZOZff+OZLPHuuxXovzmMLwnzNJ/Ac7ju+j2rJu2TMo/c+udobn6uSa/HC+Q4XuxEfuXmGe47Mbzumyfv02XMtHj61TtmzOTRb5rO3L267MJu8X03IyGQrTY5p8joIebBdYKHm89nbF5mv+TzwwgrffWWN2YrLkT1VPNsaG7fRwk2Ai50hqdY8cmqDKEvReVKGKMVHbp7h6GKNZ5fb277XJ48usNIOuf/h1+gMjIT8Sick05oPHpjic+9fJNZwcrXHc8stE2NxbBxLmC65/Mzx/W9oThKRR7XWx6/6uhvJEGzH9UzYbydvZhxX+tu34n1PrvY4s9GjNzQO/Ypns7cRcHiuck3H+kGe42s91hsZ6265V7ZijNsKT51tIQK3761dkxzBtXzWJ05v8MpqD4QrTsrX895vxevfaq7l+JPfh+Ygoh44l3wHrvZe13Oct+JcFIagoKCg4AbnWg3BDZE+WlBQUFBweQpDUFBQUHCDUxiCgoKCghucwhAUFBQU3OAUhqCgoKDgBucdkTUkIqvAqzs9DmAWuLjTg9iGYlzXRzGu62O3jgt279h2y7hu0lrPXe1F7whDsFsQkUeuJRXrB00xruujGNf1sVvHBbt3bLt1XJejcA0VFBQU3OAUhqCgoKDgBqcwBNfHl3d6AJehGNf1UYzr+tit44LdO7bdOq5tKWIEBQUFBTc4xY6goKCg4AanMATXgIh8RkReEJGXReSf7PR4RojIb4vIiog8vdNjmURE9ovIt0TkORF5RkR+eafHBCAivoh8X0SeyMf1L3d6TJOIiCUifyUi/3unxzJCRE6JyFMi8riI7BrlRxFpiMgfiMjz+X32w7tgTEfy8zT61xaRL+30uK6FwjV0FUTEAl4EPgmcAR4G7tNaP7ujAwNE5EeBLvCftda37/R4RojIIrCotX5MRKrAo8BP7/Q5ExEBylrrrog4wHeAX9Za/+VOjmuEiPxj4DhQ01p/bqfHA8YQAMe11rshJ36MiPwO8G2t9VdExAVKWuvmTo9rRD5vnAXu1lrvhhqoK1LsCK7Oh4GXtdYntNYR8LvAX9/hMQGgtX4QWN/pcWxFa72stX4sf9wBngP27eyoQBu6+Y9O/m9XrIREZAn4SeArOz2W3Y6I1IAfBX4LQGsd7SYjkHMv8Mo7wQhAYQiuhX3AaxM/n2EXTGrvFETkIPAB4Hs7OxJD7n55HFgBvqG13hXjAv4N8KtwSZvgnUYDfyEij4rIF3d6MDmHgVXgP+WutK+ISHmnB7WFLwD37/QgrpXCEFwd2ea5XbGK3O2ISAX4Q+BLWuv2To8HQGudaq3vBJaAD4vIjrvURORzwIrW+tGdHss2fFRr/UHgJ4B/kLsjdxob+CDwm1rrDwA9YDfF7lzgp4D/vtNjuVYKQ3B1zgD7J35eAs7t0FjeMeQ++D8Evqq1/qOdHs9WclfCA8BndngoAB8Ffir3x/8u8AkR+a87OySD1vpc/v8K8McYV+lOcwY4M7Gb+wOMYdgt/ATwmNb6wk4P5FopDMHVeRi4RUQO5Zb+C8Cf7PCYdjV5UPa3gOe01v96p8czQkTmRKSRPw6AHwee39lRgdb6n2qtl7TWBzH31ze11n9rh4eFiJTzYD+56+VTwI5nqGmtzwOviciR/Kl7gR1P3pjgPt5BbiEwW6yCK6C1TkTkl4A/Byzgt7XWz+zwsAAQkfuBe4BZETkD/JrW+rd2dlSAWeH+PPBU7o8H+Gda66/v4JgAFoHfyTM6FPD7Wutdk6q5C1kA/tjYdWzgv2mt/2xnhzTmHwJfzRdnJ4C/s8PjAUBESpgMw7+302O5Hor00YKCgoIbnMI1VFBQUHCDUxiCgoKCghucwhAUFBQU3OAUhqCgoKDgBqcwBAUFBQU3OIUhKHjHIyIzE4qP50Xk7MTPn97y2i+JyL97C4/9pTxl8Eqv+Rci8iuX+d1336qxFBS8UQpDUPCOR2u9prW+M5eO+PfAr+ePfxNToDXJW60B8yXgiobgSmitP/IWjqWg4A1RGIKCdzN/AHxORDwYC+DtxchPjxGRXxWRf5Q//nUR+Wb++N6R1IOI/KaIPDLZxyD/m73At0TkW/lznxGRx/KeB/9n4jBHReQBETkxOlb++m7+/z3570ca+1/NK7QRkc/mz31HRH5jN/UrKHh3UBiCgnctWus14Pts6gl9Afg9fWkV5YPAx/LHx4FKrpX0I8C38+f/udb6OHAH8GMicofW+jcwulMf11p/XETmgP8I/A2t9THgb04c41bg0xitnl/L338rH8DsMI5iFDY/KiI+8B+An9Ba/wgw90bORUHBlSgMQcG7nfvZdA9dzi30KHBXrqszBB7CGISPsWkIPi8ijwF/BbwPM1lv5YeAB7XWJwG01pO9Ir6mtR7mDV5WMPINW/m+1vqM1joDHgcOYgzIidF7Xmb8BQVvisIQFLzb+R/AvSLyQSAYNcyZRGsdA6cwejXfxUz+HwduBp4TkUPArwD3aq3vAL4G+NscS7i8RPlw4nHK9jpf271mOxn0goK3lMIQFLyryTuSPQD8NldeTT+ImewfxBiCvw88nruRahjN+5aILGBkhkd0gGr++CGM2+gQgIhMvwUf4XngcB7fAPjZt+A9CwpeR2EICm4E7geOYbT+L8e3MeqkD+U68mH+HFrrJzAuoWcwBuX/Tfzdl4E/FZFvaa1XgS8CfyQiTwC/92YHrrUeAL8I/JmIfAe4ALTe7PsWFExSqI8WFOxyRKSite7mWUT/FnhJa/3rOz2ugncPxY6goGD383fzvg7PAHVMFlFBwVtGsSMoKCgouMEpdgQFBQUFNziFISgoKCi4wSkMQUFBQcENTmEICgoKCm5wCkNQUFBQcINTGIKCgoKCG5z/D53lXNGbXp4GAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x24cf4a27550>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Reverse-coding the 'happy' variable so that higher scores mean less\n",
    "# happiness (and thus possibly more sadness).\n",
    "\n",
    "X['Sadness'] = max(X['happy']) - X['happy']\n",
    "\n",
    "sns.regplot(\n",
    "    X['tvtot'],\n",
    "    X['Sadness'],\n",
    "    x_jitter=.49,\n",
    "    y_jitter=.49,\n",
    "    scatter_kws={'alpha':0.3},\n",
    "    line_kws={'color':'black'},\n",
    "    ci=None\n",
    ")\n",
    "plt.xlabel('TV watching')\n",
    "plt.ylabel('Sadness')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Perhaps watching more TV only makes people sad if they do it alone. \n",
    "People who live with a partner are probably less likely to watch TV alone. \n",
    "\n",
    "To represent the concept that people who live with a romantic partner don't get sad watching TV but people who don't live with a romantic partner do get sad, we can <b>multiply</b> our 'partner' indicator by the 'sadness' feature to create the interaction."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 252,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x24cf4a86208>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Create a better indicator for 'partner.' Currently it is coded 1 for\n",
    "# living with a partner and 2 for not living with a partner.\n",
    "X['LiveWithPartner'] = np.where(y == 1, 1, 0)\n",
    "\n",
    "# Create the interaction.\n",
    "X['Sad_Partner'] = X['Sadness'] * X['LiveWithPartner']\n",
    "\n",
    "# A plot of an interaction.\n",
    "sns.lmplot(\n",
    "    x='Sadness',\n",
    "    y='tvtot',\n",
    "    hue='LiveWithPartner',\n",
    "    data=X,\n",
    "    scatter=False\n",
    ")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 253,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x24cf4aa9f98>"
      ]
     },
     "execution_count": 253,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x24cf48dbfd0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import seaborn as sns\n",
    "corrmat = X.corr()\n",
    "\n",
    "# Set up the matplotlib figure\n",
    "f,ax = plt.subplots(figsize=(12,9))\n",
    "\n",
    "# Draw the heatmap using seaborn\n",
    "sns.heatmap(corrmat , vmax=.8 , square = True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 266,
   "metadata": {
    "scrolled": true
   },
   "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>year</th>\n",
       "      <th>tvtot</th>\n",
       "      <th>ppltrst</th>\n",
       "      <th>pplfair</th>\n",
       "      <th>pplhlp</th>\n",
       "      <th>happy</th>\n",
       "      <th>sclmeet</th>\n",
       "      <th>sclact</th>\n",
       "      <th>gndr</th>\n",
       "      <th>agea</th>\n",
       "      <th>...</th>\n",
       "      <th>CZ</th>\n",
       "      <th>DE</th>\n",
       "      <th>ES</th>\n",
       "      <th>NO</th>\n",
       "      <th>SE</th>\n",
       "      <th>Nordic</th>\n",
       "      <th>Adult_61plus</th>\n",
       "      <th>Sadness</th>\n",
       "      <th>LiveWithPartner</th>\n",
       "      <th>Sad_Partner</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>6</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>60.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>6</td>\n",
       "      <td>6.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>59.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>6</td>\n",
       "      <td>1.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>24.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>6</td>\n",
       "      <td>4.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>64.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>6</td>\n",
       "      <td>5.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>55.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>6</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>76.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>10.0</td>\n",
       "      <td>1</td>\n",
       "      <td>10.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>6</td>\n",
       "      <td>2.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>30.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>6</td>\n",
       "      <td>2.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>84.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>6</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>62.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>6</td>\n",
       "      <td>1.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>33.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>6</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>40.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>6</td>\n",
       "      <td>4.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>69.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>6</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>59.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>6</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>32.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>6</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>70.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>6</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>61.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>6</td>\n",
       "      <td>2.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>30.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>1</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>6</td>\n",
       "      <td>0.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>21.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>6</td>\n",
       "      <td>2.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>36.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>6</td>\n",
       "      <td>7.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>51.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>6</td>\n",
       "      <td>1.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>25.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>6</td>\n",
       "      <td>2.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>62.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>6</td>\n",
       "      <td>2.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>20.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>6</td>\n",
       "      <td>0.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>22.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>6</td>\n",
       "      <td>0.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>32.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>6</td>\n",
       "      <td>2.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>35.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>6</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>26.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>6</td>\n",
       "      <td>2.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>35.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>6</td>\n",
       "      <td>7.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>54.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>6</td>\n",
       "      <td>6.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>38.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</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>8564</th>\n",
       "      <td>7</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>17.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8565</th>\n",
       "      <td>7</td>\n",
       "      <td>5.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>17.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8566</th>\n",
       "      <td>7</td>\n",
       "      <td>2.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>17.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8567</th>\n",
       "      <td>7</td>\n",
       "      <td>2.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>17.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8568</th>\n",
       "      <td>7</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>17.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8569</th>\n",
       "      <td>7</td>\n",
       "      <td>6.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>17.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8570</th>\n",
       "      <td>7</td>\n",
       "      <td>1.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8571</th>\n",
       "      <td>7</td>\n",
       "      <td>3.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8572</th>\n",
       "      <td>7</td>\n",
       "      <td>5.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8573</th>\n",
       "      <td>7</td>\n",
       "      <td>1.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8574</th>\n",
       "      <td>7</td>\n",
       "      <td>1.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8575</th>\n",
       "      <td>7</td>\n",
       "      <td>3.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8576</th>\n",
       "      <td>7</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8577</th>\n",
       "      <td>7</td>\n",
       "      <td>6.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8578</th>\n",
       "      <td>7</td>\n",
       "      <td>1.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8579</th>\n",
       "      <td>7</td>\n",
       "      <td>2.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8580</th>\n",
       "      <td>7</td>\n",
       "      <td>4.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8581</th>\n",
       "      <td>7</td>\n",
       "      <td>1.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8582</th>\n",
       "      <td>7</td>\n",
       "      <td>6.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8583</th>\n",
       "      <td>7</td>\n",
       "      <td>2.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8584</th>\n",
       "      <td>7</td>\n",
       "      <td>2.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8585</th>\n",
       "      <td>7</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8586</th>\n",
       "      <td>7</td>\n",
       "      <td>2.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8587</th>\n",
       "      <td>7</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8588</th>\n",
       "      <td>7</td>\n",
       "      <td>1.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8589</th>\n",
       "      <td>7</td>\n",
       "      <td>3.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>18.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8590</th>\n",
       "      <td>7</td>\n",
       "      <td>5.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8591</th>\n",
       "      <td>7</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>44.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8592</th>\n",
       "      <td>7</td>\n",
       "      <td>5.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8593</th>\n",
       "      <td>7</td>\n",
       "      <td>2.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>8147 rows × 21 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      year  tvtot  ppltrst  pplfair  pplhlp  happy  sclmeet  sclact  gndr  \\\n",
       "0        6    3.0      3.0     10.0     5.0    8.0      5.0     4.0   2.0   \n",
       "1        6    6.0      5.0      7.0     5.0    9.0      3.0     2.0   2.0   \n",
       "2        6    1.0      8.0      8.0     8.0    7.0      6.0     3.0   1.0   \n",
       "3        6    4.0      6.0      6.0     7.0   10.0      6.0     2.0   2.0   \n",
       "4        6    5.0      6.0      7.0     5.0    8.0      7.0     2.0   2.0   \n",
       "6        6    3.0      0.0      5.0     2.0    0.0      2.0     2.0   1.0   \n",
       "7        6    2.0      4.0      5.0     3.0   10.0      5.0     2.0   2.0   \n",
       "8        6    2.0      8.0      8.0     8.0    9.0      6.0     4.0   2.0   \n",
       "9        6    4.0      4.0      4.0     8.0    7.0      4.0     2.0   2.0   \n",
       "10       6    1.0      6.0      7.0     7.0    9.0      5.0     2.0   2.0   \n",
       "11       6    4.0      5.0      7.0     7.0    7.0      5.0     3.0   1.0   \n",
       "12       6    4.0      7.0      7.0     4.0    9.0      6.0     2.0   1.0   \n",
       "13       6    2.0      1.0      9.0     7.0    8.0      4.0     3.0   1.0   \n",
       "14       6    4.0      4.0      5.0     3.0    8.0      1.0     2.0   2.0   \n",
       "15       6    5.0      4.0      5.0     5.0    9.0      3.0     3.0   2.0   \n",
       "16       6    4.0      5.0      7.0     7.0    8.0      3.0     3.0   1.0   \n",
       "17       6    2.0      7.0      7.0     7.0    6.0      6.0     3.0   1.0   \n",
       "18       6    0.0      9.0      7.0     8.0    9.0      5.0     4.0   1.0   \n",
       "19       6    2.0      6.0      6.0     6.0    8.0      4.0     3.0   2.0   \n",
       "20       6    7.0      3.0      4.0     4.0    7.0      6.0     4.0   1.0   \n",
       "21       6    1.0      8.0      8.0     6.0    8.0      7.0     4.0   1.0   \n",
       "22       6    2.0      7.0      7.0     8.0    9.0      6.0     3.0   1.0   \n",
       "23       6    2.0      4.0      8.0     8.0    9.0      6.0     2.0   2.0   \n",
       "24       6    0.0      7.0      7.0     4.0    9.0      4.0     2.0   2.0   \n",
       "25       6    0.0      6.0      7.0     5.0    7.0      6.0     2.0   2.0   \n",
       "26       6    2.0      7.0      6.0     5.0    8.0      4.0     3.0   2.0   \n",
       "27       6    0.0      5.0      8.0     6.0    9.0      4.0     2.0   2.0   \n",
       "28       6    2.0      8.0      8.0     8.0   10.0      6.0     3.0   2.0   \n",
       "29       6    7.0      8.0      5.0     5.0    8.0      5.0     3.0   2.0   \n",
       "30       6    6.0      8.0      8.0     3.0    5.0      6.0     1.0   2.0   \n",
       "...    ...    ...      ...      ...     ...    ...      ...     ...   ...   \n",
       "8564     7    2.0      1.0      5.0     4.0   10.0      7.0     4.0   1.0   \n",
       "8565     7    5.0      5.0      3.0     4.0    5.0      6.0     1.0   2.0   \n",
       "8566     7    2.0      7.0      8.0     6.0    8.0      7.0     3.0   2.0   \n",
       "8567     7    2.0      4.0      8.0     7.0    9.0      7.0     4.0   2.0   \n",
       "8568     7    1.0      2.0      8.0     6.0    9.0      7.0     3.0   2.0   \n",
       "8569     7    6.0      5.0      3.0     4.0    9.0      6.0     4.0   1.0   \n",
       "8570     7    1.0      5.0      7.0     7.0    8.0      7.0     4.0   1.0   \n",
       "8571     7    3.0      4.0      9.0     7.0    8.0      6.0     3.0   1.0   \n",
       "8572     7    5.0      7.0      8.0     5.0    8.0      3.0     1.0   1.0   \n",
       "8573     7    1.0      8.0      8.0     6.0    9.0      7.0     3.0   2.0   \n",
       "8574     7    1.0      4.0      5.0     6.0    7.0      7.0     3.0   2.0   \n",
       "8575     7    3.0      8.0      7.0     7.0    8.0      7.0     3.0   1.0   \n",
       "8576     7    4.0      5.0      4.0     4.0    7.0      6.0     3.0   1.0   \n",
       "8577     7    6.0      6.0      9.0     7.0    5.0      4.0     2.0   2.0   \n",
       "8578     7    1.0      5.0      5.0     5.0    7.0      5.0     4.0   2.0   \n",
       "8579     7    2.0      8.0      9.0     7.0    9.0      7.0     3.0   2.0   \n",
       "8580     7    4.0      7.0      8.0     5.0   10.0      4.0     1.0   1.0   \n",
       "8581     7    1.0      6.0      7.0     6.0    8.0      6.0     3.0   1.0   \n",
       "8582     7    6.0      7.0      7.0     6.0    8.0      7.0     3.0   1.0   \n",
       "8583     7    2.0      6.0      8.0     5.0   10.0      6.0     5.0   1.0   \n",
       "8584     7    2.0      8.0     10.0     6.0    9.0      6.0     2.0   1.0   \n",
       "8585     7    1.0      3.0      6.0     4.0    9.0      7.0     3.0   1.0   \n",
       "8586     7    2.0      4.0      6.0     3.0    7.0      7.0     3.0   2.0   \n",
       "8587     7    4.0      4.0      6.0     7.0    9.0      7.0     3.0   1.0   \n",
       "8588     7    1.0      6.0      5.0     5.0   10.0      7.0     2.0   1.0   \n",
       "8589     7    3.0      4.0      5.0     3.0    6.0      6.0     2.0   1.0   \n",
       "8590     7    5.0      6.0      4.0     4.0   10.0      6.0     3.0   1.0   \n",
       "8591     7    4.0      5.0      7.0     6.0    8.0      6.0     3.0   1.0   \n",
       "8592     7    5.0      8.0      8.0     6.0    9.0      7.0     3.0   1.0   \n",
       "8593     7    2.0      6.0      7.0     5.0    7.0      7.0     4.0   2.0   \n",
       "\n",
       "      agea     ...       CZ  DE  ES  NO  SE  Nordic  Adult_61plus  Sadness  \\\n",
       "0     60.0     ...        0   0   0   0   0       0             0      2.0   \n",
       "1     59.0     ...        0   0   0   0   0       0             0      1.0   \n",
       "2     24.0     ...        0   0   0   0   0       0             0      3.0   \n",
       "3     64.0     ...        0   0   0   0   0       0             1      0.0   \n",
       "4     55.0     ...        0   0   0   0   0       0             0      2.0   \n",
       "6     76.0     ...        0   0   0   0   0       0             1     10.0   \n",
       "7     30.0     ...        0   0   0   0   0       0             0      0.0   \n",
       "8     84.0     ...        0   0   0   0   0       0             1      1.0   \n",
       "9     62.0     ...        0   0   0   0   0       0             1      3.0   \n",
       "10    33.0     ...        0   0   0   0   0       0             0      1.0   \n",
       "11    40.0     ...        0   0   0   0   0       0             0      3.0   \n",
       "12    69.0     ...        0   0   0   0   0       0             1      1.0   \n",
       "13    59.0     ...        0   0   0   0   0       0             0      2.0   \n",
       "14    32.0     ...        0   0   0   0   0       0             0      2.0   \n",
       "15    70.0     ...        0   0   0   0   0       0             1      1.0   \n",
       "16    61.0     ...        0   0   0   0   0       0             1      2.0   \n",
       "17    30.0     ...        0   0   0   0   0       0             0      4.0   \n",
       "18    21.0     ...        0   0   0   0   0       0             0      1.0   \n",
       "19    36.0     ...        0   0   0   0   0       0             0      2.0   \n",
       "20    51.0     ...        0   0   0   0   0       0             0      3.0   \n",
       "21    25.0     ...        0   0   0   0   0       0             0      2.0   \n",
       "22    62.0     ...        0   0   0   0   0       0             1      1.0   \n",
       "23    20.0     ...        0   0   0   0   0       0             0      1.0   \n",
       "24    22.0     ...        0   0   0   0   0       0             0      1.0   \n",
       "25    32.0     ...        0   0   0   0   0       0             0      3.0   \n",
       "26    35.0     ...        0   0   0   0   0       0             0      2.0   \n",
       "27    26.0     ...        0   0   0   0   0       0             0      1.0   \n",
       "28    35.0     ...        0   0   0   0   0       0             0      0.0   \n",
       "29    54.0     ...        0   0   0   0   0       0             0      2.0   \n",
       "30    38.0     ...        0   0   0   0   0       0             0      5.0   \n",
       "...    ...     ...       ..  ..  ..  ..  ..     ...           ...      ...   \n",
       "8564  17.0     ...        0   0   0   0   1       1             0      0.0   \n",
       "8565  17.0     ...        0   0   0   0   1       1             0      5.0   \n",
       "8566  17.0     ...        0   0   0   0   1       1             0      2.0   \n",
       "8567  17.0     ...        0   0   0   0   1       1             0      1.0   \n",
       "8568  17.0     ...        0   0   0   0   1       1             0      1.0   \n",
       "8569  17.0     ...        0   0   0   0   1       1             0      1.0   \n",
       "8570  16.0     ...        0   0   0   0   1       1             0      2.0   \n",
       "8571  16.0     ...        0   0   0   0   1       1             0      2.0   \n",
       "8572  16.0     ...        0   0   0   0   1       1             0      2.0   \n",
       "8573  16.0     ...        0   0   0   0   1       1             0      1.0   \n",
       "8574  16.0     ...        0   0   0   0   1       1             0      3.0   \n",
       "8575  16.0     ...        0   0   0   0   1       1             0      2.0   \n",
       "8576  16.0     ...        0   0   0   0   1       1             0      3.0   \n",
       "8577  16.0     ...        0   0   0   0   1       1             0      5.0   \n",
       "8578  16.0     ...        0   0   0   0   1       1             0      3.0   \n",
       "8579  16.0     ...        0   0   0   0   1       1             0      1.0   \n",
       "8580  16.0     ...        0   0   0   0   1       1             0      0.0   \n",
       "8581  16.0     ...        0   0   0   0   1       1             0      2.0   \n",
       "8582  16.0     ...        0   0   0   0   1       1             0      2.0   \n",
       "8583  16.0     ...        0   0   0   0   1       1             0      0.0   \n",
       "8584  16.0     ...        0   0   0   0   1       1             0      1.0   \n",
       "8585  16.0     ...        0   0   0   0   1       1             0      1.0   \n",
       "8586  15.0     ...        0   0   0   0   1       1             0      3.0   \n",
       "8587  15.0     ...        0   0   0   0   1       1             0      1.0   \n",
       "8588  15.0     ...        0   0   0   0   1       1             0      0.0   \n",
       "8589  18.0     ...        0   0   0   0   1       1             0      4.0   \n",
       "8590  15.0     ...        0   0   0   0   1       1             0      0.0   \n",
       "8591  44.0     ...        0   0   0   0   1       1             0      2.0   \n",
       "8592  15.0     ...        0   0   0   0   1       1             0      1.0   \n",
       "8593  15.0     ...        0   0   0   0   1       1             0      3.0   \n",
       "\n",
       "      LiveWithPartner  Sad_Partner  \n",
       "0                   1          2.0  \n",
       "1                   1          1.0  \n",
       "2                   0          0.0  \n",
       "3                   1          0.0  \n",
       "4                   1          2.0  \n",
       "6                   1         10.0  \n",
       "7                   1          0.0  \n",
       "8                   0          0.0  \n",
       "9                   1          3.0  \n",
       "10                  1          1.0  \n",
       "11                  1          3.0  \n",
       "12                  1          1.0  \n",
       "13                  0          0.0  \n",
       "14                  1          2.0  \n",
       "15                  1          1.0  \n",
       "16                  1          2.0  \n",
       "17                  1          4.0  \n",
       "18                  0          0.0  \n",
       "19                  1          2.0  \n",
       "20                  1          3.0  \n",
       "21                  0          0.0  \n",
       "22                  1          1.0  \n",
       "23                  0          0.0  \n",
       "24                  1          1.0  \n",
       "25                  1          3.0  \n",
       "26                  1          2.0  \n",
       "27                  1          1.0  \n",
       "28                  1          0.0  \n",
       "29                  1          2.0  \n",
       "30                  0          0.0  \n",
       "...               ...          ...  \n",
       "8564                0          0.0  \n",
       "8565                0          0.0  \n",
       "8566                0          0.0  \n",
       "8567                0          0.0  \n",
       "8568                0          0.0  \n",
       "8569                0          0.0  \n",
       "8570                0          0.0  \n",
       "8571                0          0.0  \n",
       "8572                0          0.0  \n",
       "8573                0          0.0  \n",
       "8574                0          0.0  \n",
       "8575                0          0.0  \n",
       "8576                0          0.0  \n",
       "8577                0          0.0  \n",
       "8578                0          0.0  \n",
       "8579                0          0.0  \n",
       "8580                0          0.0  \n",
       "8581                0          0.0  \n",
       "8582                0          0.0  \n",
       "8583                0          0.0  \n",
       "8584                0          0.0  \n",
       "8585                0          0.0  \n",
       "8586                0          0.0  \n",
       "8587                0          0.0  \n",
       "8588                0          0.0  \n",
       "8589                0          0.0  \n",
       "8590                0          0.0  \n",
       "8591                0          0.0  \n",
       "8592                0          0.0  \n",
       "8593                0          0.0  \n",
       "\n",
       "[8147 rows x 21 columns]"
      ]
     },
     "execution_count": 266,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 281,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "--- 0.004011869430541992 seconds ---\n"
     ]
    }
   ],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.svm import SVC\n",
    "\n",
    "import time\n",
    "start_time = time.time()\n",
    "\n",
    "X2 = X.drop(['LiveWithPartner','Sadness','Sad_Partner'] , axis = 1)\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X2, y, random_state=42)\n",
    "\n",
    "print(\"\\n--- %s seconds ---\" % (time.time() - start_time))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 282,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "GradientBoostingClassifier(criterion='friedman_mse', init=None,\n",
       "              learning_rate=0.1, loss='deviance', max_depth=2,\n",
       "              max_features=None, max_leaf_nodes=None,\n",
       "              min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "              min_samples_leaf=1, min_samples_split=2,\n",
       "              min_weight_fraction_leaf=0.0, n_estimators=500,\n",
       "              presort='auto', random_state=None, subsample=1.0, verbose=0,\n",
       "              warm_start=False)"
      ]
     },
     "execution_count": 282,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# We'll make 500 iterations, use 2-deep trees, and set our loss function.\n",
    "params = {'n_estimators': 500,\n",
    "          'max_depth': 2,\n",
    "          'loss': 'deviance'\n",
    "         }\n",
    "\n",
    "# Initialize and fit the model.\n",
    "clf = ensemble.GradientBoostingClassifier(**params)\n",
    "clf.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 283,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7628865979381443"
      ]
     },
     "execution_count": 283,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "predict_train = clf.predict(X_train)\n",
    "predict_test = clf.predict(X_test)\n",
    "\n",
    "predict_train\n",
    "\n",
    "from sklearn.metrics import accuracy_score\n",
    "score_gdb = accuracy_score(y_test, predict_test)\n",
    "score_gdb"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 274,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Accuracy tables.\n",
    "table_train = pd.crosstab(y_train, predict_train, margins=True)\n",
    "table_test = pd.crosstab(y_test, predict_test, margins=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 275,
   "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>col_0</th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "      <th>All</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>row_0</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2251</td>\n",
       "      <td>80</td>\n",
       "      <td>2331</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>15</td>\n",
       "      <td>3764</td>\n",
       "      <td>3779</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>All</th>\n",
       "      <td>2266</td>\n",
       "      <td>3844</td>\n",
       "      <td>6110</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "col_0     0     1   All\n",
       "row_0                  \n",
       "0      2251    80  2331\n",
       "1        15  3764  3779\n",
       "All    2266  3844  6110"
      ]
     },
     "execution_count": 275,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "table_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 276,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_tI_errors = table_train.loc[0.0,1.0] / table_train.loc['All','All']\n",
    "train_tII_errors = table_train.loc[1.0,0.0] / table_train.loc['All','All']\n",
    "\n",
    "test_tI_errors = table_test.loc[0.0,1.0]/table_test.loc['All','All']\n",
    "test_tII_errors = table_test.loc[1.0,0.0]/table_test.loc['All','All']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 277,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training set accuracy:\n",
      "Percent Type I errors: 0.01309328968903437\n",
      "Percent Type II errors: 0.0024549918166939444\n",
      "\n",
      "Test set accuracy:\n",
      "Percent Type I errors: 0.017673048600883652\n",
      "Percent Type II errors: 0.004418262150220913\n"
     ]
    }
   ],
   "source": [
    "print((\n",
    "    'Training set accuracy:\\n'\n",
    "    'Percent Type I errors: {}\\n'\n",
    "    'Percent Type II errors: {}\\n\\n'\n",
    "    'Test set accuracy:\\n'\n",
    "    'Percent Type I errors: {}\\n'\n",
    "    'Percent Type II errors: {}'\n",
    ").format(train_tI_errors, train_tII_errors, test_tI_errors, test_tII_errors))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 285,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x24cf4fd7748>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "feature_importance = clf.feature_importances_\n",
    "100.0 * (feature_importance / feature_importance.max())\n",
    "np.argsort(feature_importance)\n",
    "\n",
    "\n",
    "sorted_idx = np.argsort(feature_importance)\n",
    "pos = np.arange(sorted_idx.shape[0]) + .5\n",
    "plt.subplot(1, 2, 2)\n",
    "plt.barh(pos, feature_importance[sorted_idx], align='center')\n",
    "plt.yticks(pos, X2.columns[sorted_idx])\n",
    "plt.xlabel('Relative Importance')\n",
    "plt.title('Variable Importance')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Tuning:"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![title](tree-infographic for greadient boosting.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "1. min_samples_split\n",
    "    - Defines the minimum number of samples (or observations) which are required in a node to be considered for splitting.\n",
    "    - Used to control over-fitting. Higher values prevent a model from learning relations which might be highly specific to the particular sample selected for a tree.\n",
    "    - Too high values can lead to under-fitting hence, it should be tuned using CV.\n",
    "2. min_samples_leaf\n",
    "    - Defines the minimum samples (or observations) required in a terminal node or leaf.\n",
    "    - Used to control over-fitting similar to min_samples_split.\n",
    "    - Generally lower values should be chosen for imbalanced class problems because the regions in which the minority class will be in majority will be very small.\n",
    "3. min_weight_fraction_leaf\n",
    "    - Similar to min_samples_leaf but defined as a fraction of the total number of observations instead of an integer.\n",
    "    - Only one of #2 and #3 should be defined.\n",
    "4. max_depth\n",
    "    - The maximum depth of a tree.\n",
    "    - Used to control over-fitting as higher depth will allow model to learn relations very specific to a particular sample.\n",
    "    - Should be tuned using CV.\n",
    "5. max_leaf_nodes\n",
    "    - The maximum number of terminal nodes or leaves in a tree.\n",
    "    - Can be defined in place of max_depth. Since binary trees are created, a depth of ‘n’ would produce a maximum of 2^n leaves.\n",
    "    - If this is defined, GBM will ignore max_depth.\n",
    "6. max_features\n",
    "    - The number of features to consider while searching for a best split. These will be randomly selected.\n",
    "    - As a thumb-rule, square root of the total number of features works great but we should check upto 30-40% of the total number of features.\n",
    "    - Higher values can lead to over-fitting but depends on case to case.\n",
    "    \n",
    "    [source](https://www.analyticsvidhya.com/blog/2016/02/complete-guide-parameter-tuning-gradient-boosting-gbm-python/)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Create a function for GBM with CV:\n",
    "Lets define a function which will help us create GBM models and perform cross-validation."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 388,
   "metadata": {},
   "outputs": [],
   "source": [
    "def modelfit(alg, dtrain, predictors, target, performCV=True, printFeatureImportance=True, cv_folds=5):\n",
    "    #Fit the algorithm on the data\n",
    "    alg.fit(dtrain[predictors], target)\n",
    "        \n",
    "    #Predict training set:\n",
    "    dtrain_predictions = alg.predict(dtrain[predictors])\n",
    "    dtrain_predprob = alg.predict_proba(dtrain[predictors])[:,1]\n",
    "    \n",
    "    #Perform cross-validation:\n",
    "    if performCV:\n",
    "        cv_score = cross_validation.cross_val_score(alg, dtrain[predictors], target , cv=cv_folds, scoring='roc_auc')\n",
    "    \n",
    "    #Print model report:\n",
    "    print (\"\\nModel Report\")\n",
    "    print (\"Accuracy : %.4g\" % metrics.accuracy_score(target, dtrain_predictions))\n",
    "    print (\"AUC Score (Train): %f\" % metrics.roc_auc_score(target, dtrain_predprob))\n",
    "    \n",
    "    if performCV:\n",
    "        print (\"CV Score : Mean_ %.7g | Std_ %.7g | Min_ %.7g | Max_ %.7g\" \n",
    "                   % (np.mean(cv_score),np.std(cv_score),np.min(cv_score),np.max(cv_score)\n",
    "                     )\n",
    "              )\n",
    "        \n",
    "    #Print Feature Importance:\n",
    "    if printFeatureImportance:\n",
    "        feat_imp = pd.Series(alg.feature_importances_, predictors).sort_values(ascending=False)\n",
    "        feat_imp.plot(kind='bar', title='Feature Importances')\n",
    "        plt.ylabel('Feature Importance Score')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Lets start by creating a <b>baseline model</b>. In this case, the evaluation metric is AUC so using any constant value will give 0.5 as result. Typically, a good baseline can be a GBM model with default parameters, i.e. without any tuning. \n",
    "\n",
    "Lets find out what it gives:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 304,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Model Report\n",
      "Accuracy : 0.7744\n",
      "AUC Score (Train): 0.820911\n",
      "CV Score : Mean - 0.7918901 | Std - 0.01463337 | Min - 0.7700551 | Max - 0.8120606\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x24cf0ab5550>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Choose all predictors except target & IDcols\n",
    "train = X2\n",
    "target = y    \n",
    "predictors = X2.columns\n",
    "\n",
    "gbm0 = ensemble.GradientBoostingClassifier(random_state=10)\n",
    "modelfit(gbm0, train, predictors , target)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "So, the mean CV score is 0.79 and we should expect our model to do better than this."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Parameter Tuning \n",
    "Though, GBM is robust enough to not overfit with increasing trees, but a high number for pa particular learning rate can lead to overfitting. But as we reduce the learning rate and increase trees, the computation becomes expensive and would take a long time to run on standard personal computers."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 338,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=5, error_score='raise',\n",
       "       estimator=GradientBoostingClassifier(criterion='friedman_mse', init=None,\n",
       "              learning_rate=0.1, loss='deviance', max_depth=8,\n",
       "              max_features='sqrt', max_leaf_nodes=None,\n",
       "              min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "              min_samples_leaf=50, min_samples_split=500,\n",
       "              min_weight_fraction_leaf=0.0, n_estimators=100,\n",
       "              presort='auto', random_state=10, subsample=0.8, verbose=0,\n",
       "              warm_start=False),\n",
       "       fit_params=None, iid=False, n_jobs=4,\n",
       "       param_grid={'n_estimators': range(20, 500, 20)},\n",
       "       pre_dispatch='2*n_jobs', refit=True, return_train_score='warn',\n",
       "       scoring='roc_auc', verbose=0)"
      ]
     },
     "execution_count": 338,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Choose all predictors except target\n",
    "predictors = X2.columns\n",
    "\n",
    "param_test1 = {'n_estimators':range(20,500,20)}\n",
    "\n",
    "gsearch1 = GridSearchCV(estimator = ensemble.GradientBoostingClassifier(\n",
    "                            learning_rate=0.1\n",
    "                            ,min_samples_split=500\n",
    "                            ,min_samples_leaf=50\n",
    "                            ,max_depth=8\n",
    "                            ,max_features='sqrt'\n",
    "                            ,subsample=0.8\n",
    "                            ,random_state=10)\n",
    "                            , \n",
    "                            param_grid = param_test1, scoring='roc_auc',n_jobs=4,iid=False, cv=5\n",
    "                       )\n",
    "gsearch1.fit(train[predictors] , target)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Check the output:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 339,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean of: 0.78074 and std: (+/-0.03969) for {'n_estimators': 20}\n",
      "Mean of: 0.78448 and std: (+/-0.03954) for {'n_estimators': 40}\n",
      "Mean of: 0.78788 and std: (+/-0.03737) for {'n_estimators': 60}\n",
      "Mean of: 0.78732 and std: (+/-0.03800) for {'n_estimators': 80}\n",
      "Mean of: 0.78728 and std: (+/-0.03771) for {'n_estimators': 100}\n",
      "Mean of: 0.78656 and std: (+/-0.03561) for {'n_estimators': 120}\n",
      "Mean of: 0.78638 and std: (+/-0.03506) for {'n_estimators': 140}\n",
      "Mean of: 0.78611 and std: (+/-0.03577) for {'n_estimators': 160}\n",
      "Mean of: 0.78551 and std: (+/-0.03614) for {'n_estimators': 180}\n",
      "Mean of: 0.78458 and std: (+/-0.03646) for {'n_estimators': 200}\n",
      "Mean of: 0.78448 and std: (+/-0.03716) for {'n_estimators': 220}\n",
      "Mean of: 0.78268 and std: (+/-0.03619) for {'n_estimators': 240}\n",
      "Mean of: 0.78268 and std: (+/-0.03623) for {'n_estimators': 260}\n",
      "Mean of: 0.78205 and std: (+/-0.03605) for {'n_estimators': 280}\n",
      "Mean of: 0.78130 and std: (+/-0.03712) for {'n_estimators': 300}\n",
      "Mean of: 0.78052 and std: (+/-0.03674) for {'n_estimators': 320}\n",
      "Mean of: 0.77988 and std: (+/-0.03586) for {'n_estimators': 340}\n",
      "Mean of: 0.77933 and std: (+/-0.03627) for {'n_estimators': 360}\n",
      "Mean of: 0.77849 and std: (+/-0.03709) for {'n_estimators': 380}\n",
      "Mean of: 0.77737 and std: (+/-0.03730) for {'n_estimators': 400}\n",
      "Mean of: 0.77655 and std: (+/-0.03736) for {'n_estimators': 420}\n",
      "Mean of: 0.77579 and std: (+/-0.03640) for {'n_estimators': 440}\n",
      "Mean of: 0.77581 and std: (+/-0.03631) for {'n_estimators': 460}\n",
      "Mean of: 0.77544 and std: (+/-0.03755) for {'n_estimators': 480}\n"
     ]
    }
   ],
   "source": [
    "means = gsearch1.cv_results_['mean_test_score']\n",
    "stds = gsearch1.cv_results_['std_test_score']\n",
    "\n",
    "for mean, std, params in zip(means, stds, gsearch1.cv_results_['params']):\n",
    "        print(\"Mean of: %0.5f and std: (+/-%0.05f) for %r\"\n",
    "              % (mean, std * 2, params))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 342,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "best param:  {'n_estimators': 60}\n",
      "best score AOC:  0.7878762846454682\n"
     ]
    }
   ],
   "source": [
    "\n",
    "print('best param: ',gsearch1.best_params_)\n",
    "print('best score AOC: ',gsearch1.best_score_)\n",
    "\n",
    "\n",
    "#gsearch1.cv_results_['...']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As you can see that here we got 60 as the optimal estimators for 0.1 learning rate. Note that 60 is a reasonable value and can be used as it is. But it might not be the same in all cases. Other situations:\n",
    "\n",
    "1. If the value is around 20, you might want to try lowering the learning rate to 0.05 and re-run grid search\n",
    "2. If the values are too high ~100, tuning the other parameters will take long time and you can try a higher learning rate"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Tuning tree-specific parameters\n",
    "Now lets move onto tuning the tree parameters. I plan to do this in following stages:\n",
    "\n",
    "   1. Tune max_depth and num_samples_split\n",
    "   2. Tune min_samples_leaf\n",
    "   3. Tune max_features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 343,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=5, error_score='raise',\n",
       "       estimator=GradientBoostingClassifier(criterion='friedman_mse', init=None,\n",
       "              learning_rate=0.1, loss='deviance', max_depth=3,\n",
       "              max_features='sqrt', max_leaf_nodes=None,\n",
       "              min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "              min_samples_leaf=1, min_samples_split=2,\n",
       "              min_weight_fraction_leaf=0.0, n_estimators=60,\n",
       "              presort='auto', random_state=10, subsample=0.8, verbose=0,\n",
       "              warm_start=False),\n",
       "       fit_params=None, iid=False, n_jobs=4,\n",
       "       param_grid={'min_samples_split': range(200, 1001, 200), 'max_depth': range(5, 16, 2)},\n",
       "       pre_dispatch='2*n_jobs', refit=True, return_train_score='warn',\n",
       "       scoring='roc_auc', verbose=0)"
      ]
     },
     "execution_count": 343,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Choose all predictors except target\n",
    "predictors = X2.columns\n",
    "\n",
    "param_test2 = {'max_depth':range(5,16,2), 'min_samples_split':range(200,1001,200)}\n",
    "\n",
    "gsearch2 = GridSearchCV(estimator = ensemble.GradientBoostingClassifier(\n",
    "                            learning_rate=0.1\n",
    "                            ,n_estimators=60\n",
    "                            #,min_samples_leaf=50\n",
    "                            \n",
    "                            ,max_features='sqrt'\n",
    "                            ,subsample=0.8\n",
    "                            ,random_state=10)\n",
    "                            , \n",
    "                            param_grid = param_test2, scoring='roc_auc',n_jobs=4,iid=False, cv=5\n",
    "                       )\n",
    "gsearch2.fit(train[predictors] , target)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 346,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean of: 0.78901 and std: (+/-0.03635) for {'min_samples_split': 200, 'max_depth': 5}\n",
      "Mean of: 0.78806 and std: (+/-0.03312) for {'min_samples_split': 400, 'max_depth': 5}\n",
      "Mean of: 0.78622 and std: (+/-0.03453) for {'min_samples_split': 600, 'max_depth': 5}\n",
      "Mean of: 0.78579 and std: (+/-0.03881) for {'min_samples_split': 800, 'max_depth': 5}\n",
      "Mean of: 0.78523 and std: (+/-0.03879) for {'min_samples_split': 1000, 'max_depth': 5}\n",
      "Mean of: 0.78785 and std: (+/-0.03852) for {'min_samples_split': 200, 'max_depth': 7}\n",
      "Mean of: 0.78784 and std: (+/-0.03587) for {'min_samples_split': 400, 'max_depth': 7}\n",
      "Mean of: 0.78778 and std: (+/-0.03750) for {'min_samples_split': 600, 'max_depth': 7}\n",
      "Mean of: 0.78641 and std: (+/-0.03407) for {'min_samples_split': 800, 'max_depth': 7}\n",
      "Mean of: 0.78657 and std: (+/-0.03909) for {'min_samples_split': 1000, 'max_depth': 7}\n",
      "Mean of: 0.78713 and std: (+/-0.03346) for {'min_samples_split': 200, 'max_depth': 9}\n",
      "Mean of: 0.78831 and std: (+/-0.03692) for {'min_samples_split': 400, 'max_depth': 9}\n",
      "Mean of: 0.78661 and std: (+/-0.03510) for {'min_samples_split': 600, 'max_depth': 9}\n",
      "Mean of: 0.78759 and std: (+/-0.03537) for {'min_samples_split': 800, 'max_depth': 9}\n",
      "Mean of: 0.78640 and std: (+/-0.03722) for {'min_samples_split': 1000, 'max_depth': 9}\n",
      "Mean of: 0.78639 and std: (+/-0.03430) for {'min_samples_split': 200, 'max_depth': 11}\n",
      "Mean of: 0.78756 and std: (+/-0.03577) for {'min_samples_split': 400, 'max_depth': 11}\n",
      "Mean of: 0.78553 and std: (+/-0.03293) for {'min_samples_split': 600, 'max_depth': 11}\n",
      "Mean of: 0.78772 and std: (+/-0.03660) for {'min_samples_split': 800, 'max_depth': 11}\n",
      "Mean of: 0.78532 and std: (+/-0.03773) for {'min_samples_split': 1000, 'max_depth': 11}\n",
      "Mean of: 0.78375 and std: (+/-0.03741) for {'min_samples_split': 200, 'max_depth': 13}\n",
      "Mean of: 0.78757 and std: (+/-0.03505) for {'min_samples_split': 400, 'max_depth': 13}\n",
      "Mean of: 0.78727 and std: (+/-0.03524) for {'min_samples_split': 600, 'max_depth': 13}\n",
      "Mean of: 0.78784 and std: (+/-0.03640) for {'min_samples_split': 800, 'max_depth': 13}\n",
      "Mean of: 0.78533 and std: (+/-0.03736) for {'min_samples_split': 1000, 'max_depth': 13}\n",
      "Mean of: 0.78368 and std: (+/-0.04005) for {'min_samples_split': 200, 'max_depth': 15}\n",
      "Mean of: 0.78480 and std: (+/-0.03337) for {'min_samples_split': 400, 'max_depth': 15}\n",
      "Mean of: 0.78590 and std: (+/-0.03467) for {'min_samples_split': 600, 'max_depth': 15}\n",
      "Mean of: 0.78740 and std: (+/-0.03549) for {'min_samples_split': 800, 'max_depth': 15}\n",
      "Mean of: 0.78528 and std: (+/-0.03768) for {'min_samples_split': 1000, 'max_depth': 15}\n"
     ]
    }
   ],
   "source": [
    "means2 = gsearch2.cv_results_['mean_test_score']\n",
    "stds2 = gsearch2.cv_results_['std_test_score']\n",
    "\n",
    "for mean, std, params in zip(means2, stds2, gsearch2.cv_results_['params']):\n",
    "        print(\"Mean of: %0.5f and std: (+/-%0.05f) for %r\"\n",
    "              % (mean, std * 2, params))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 345,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "best param:  {'min_samples_split': 200, 'max_depth': 5}\n",
      "best score AOC:  0.7890058694185427\n"
     ]
    }
   ],
   "source": [
    "print('best param: ',gsearch2.best_params_)\n",
    "print('best score AOC: ',gsearch2.best_score_)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Second param tuning:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 347,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=5, error_score='raise',\n",
       "       estimator=GradientBoostingClassifier(criterion='friedman_mse', init=None,\n",
       "              learning_rate=0.1, loss='deviance', max_depth=5,\n",
       "              max_features='sqrt', max_leaf_nodes=None,\n",
       "              min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "              min_samples_leaf=1, min_samples_split=2,\n",
       "              min_weight_fraction_leaf=0.0, n_estimators=60,\n",
       "              presort='auto', random_state=10, subsample=0.8, verbose=0,\n",
       "              warm_start=False),\n",
       "       fit_params=None, iid=False, n_jobs=4,\n",
       "       param_grid={'min_samples_leaf': range(30, 71, 10), 'min_samples_split': range(100, 400, 20)},\n",
       "       pre_dispatch='2*n_jobs', refit=True, return_train_score='warn',\n",
       "       scoring='roc_auc', verbose=0)"
      ]
     },
     "execution_count": 347,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# use again 'min_samples_split' but with 200 in between the range:\n",
    "\n",
    "param_test3 = {'min_samples_split':range(100,400,20) \n",
    "               ,'min_samples_leaf':range(30,71,10) \n",
    "              }\n",
    "\n",
    "gsearch3 = GridSearchCV(estimator = ensemble.GradientBoostingClassifier(\n",
    "                            learning_rate=0.1\n",
    "                            ,n_estimators=60\n",
    "                            ,max_depth=5\n",
    "                                                        \n",
    "                            ,max_features='sqrt'\n",
    "                            ,subsample=0.8\n",
    "                            ,random_state=10)\n",
    "                            , \n",
    "                            param_grid = param_test3, scoring='roc_auc',n_jobs=4,iid=False, cv=5\n",
    "                       )\n",
    "gsearch3.fit(train[predictors] , target)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 348,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean of: 0.78723 and std: (+/-0.03746) for {'min_samples_leaf': 30, 'min_samples_split': 100}\n",
      "Mean of: 0.78927 and std: (+/-0.04008) for {'min_samples_leaf': 30, 'min_samples_split': 120}\n",
      "Mean of: 0.78804 and std: (+/-0.03817) for {'min_samples_leaf': 30, 'min_samples_split': 140}\n",
      "Mean of: 0.78879 and std: (+/-0.04027) for {'min_samples_leaf': 30, 'min_samples_split': 160}\n",
      "Mean of: 0.78891 and std: (+/-0.03789) for {'min_samples_leaf': 30, 'min_samples_split': 180}\n",
      "Mean of: 0.78800 and std: (+/-0.03990) for {'min_samples_leaf': 30, 'min_samples_split': 200}\n",
      "Mean of: 0.78803 and std: (+/-0.03866) for {'min_samples_leaf': 30, 'min_samples_split': 220}\n",
      "Mean of: 0.78879 and std: (+/-0.03617) for {'min_samples_leaf': 30, 'min_samples_split': 240}\n",
      "Mean of: 0.78816 and std: (+/-0.03946) for {'min_samples_leaf': 30, 'min_samples_split': 260}\n",
      "Mean of: 0.78709 and std: (+/-0.03838) for {'min_samples_leaf': 30, 'min_samples_split': 280}\n",
      "Mean of: 0.78764 and std: (+/-0.03730) for {'min_samples_leaf': 30, 'min_samples_split': 300}\n",
      "Mean of: 0.78717 and std: (+/-0.03690) for {'min_samples_leaf': 30, 'min_samples_split': 320}\n",
      "Mean of: 0.78868 and std: (+/-0.03602) for {'min_samples_leaf': 30, 'min_samples_split': 340}\n",
      "Mean of: 0.78697 and std: (+/-0.03639) for {'min_samples_leaf': 30, 'min_samples_split': 360}\n",
      "Mean of: 0.78691 and std: (+/-0.03624) for {'min_samples_leaf': 30, 'min_samples_split': 380}\n",
      "Mean of: 0.78744 and std: (+/-0.04144) for {'min_samples_leaf': 40, 'min_samples_split': 100}\n",
      "Mean of: 0.78862 and std: (+/-0.03496) for {'min_samples_leaf': 40, 'min_samples_split': 120}\n",
      "Mean of: 0.78804 and std: (+/-0.03621) for {'min_samples_leaf': 40, 'min_samples_split': 140}\n",
      "Mean of: 0.78909 and std: (+/-0.03977) for {'min_samples_leaf': 40, 'min_samples_split': 160}\n",
      "Mean of: 0.78999 and std: (+/-0.03983) for {'min_samples_leaf': 40, 'min_samples_split': 180}\n",
      "Mean of: 0.78819 and std: (+/-0.03628) for {'min_samples_leaf': 40, 'min_samples_split': 200}\n",
      "Mean of: 0.78924 and std: (+/-0.03963) for {'min_samples_leaf': 40, 'min_samples_split': 220}\n",
      "Mean of: 0.78794 and std: (+/-0.03868) for {'min_samples_leaf': 40, 'min_samples_split': 240}\n",
      "Mean of: 0.78700 and std: (+/-0.03940) for {'min_samples_leaf': 40, 'min_samples_split': 260}\n",
      "Mean of: 0.78727 and std: (+/-0.04040) for {'min_samples_leaf': 40, 'min_samples_split': 280}\n",
      "Mean of: 0.78709 and std: (+/-0.03840) for {'min_samples_leaf': 40, 'min_samples_split': 300}\n",
      "Mean of: 0.78705 and std: (+/-0.03862) for {'min_samples_leaf': 40, 'min_samples_split': 320}\n",
      "Mean of: 0.78661 and std: (+/-0.03923) for {'min_samples_leaf': 40, 'min_samples_split': 340}\n",
      "Mean of: 0.78733 and std: (+/-0.03792) for {'min_samples_leaf': 40, 'min_samples_split': 360}\n",
      "Mean of: 0.78664 and std: (+/-0.04079) for {'min_samples_leaf': 40, 'min_samples_split': 380}\n",
      "Mean of: 0.78669 and std: (+/-0.03829) for {'min_samples_leaf': 50, 'min_samples_split': 100}\n",
      "Mean of: 0.78804 and std: (+/-0.03839) for {'min_samples_leaf': 50, 'min_samples_split': 120}\n",
      "Mean of: 0.78624 and std: (+/-0.03816) for {'min_samples_leaf': 50, 'min_samples_split': 140}\n",
      "Mean of: 0.78831 and std: (+/-0.03453) for {'min_samples_leaf': 50, 'min_samples_split': 160}\n",
      "Mean of: 0.78963 and std: (+/-0.03625) for {'min_samples_leaf': 50, 'min_samples_split': 180}\n",
      "Mean of: 0.78782 and std: (+/-0.03771) for {'min_samples_leaf': 50, 'min_samples_split': 200}\n",
      "Mean of: 0.78877 and std: (+/-0.03760) for {'min_samples_leaf': 50, 'min_samples_split': 220}\n",
      "Mean of: 0.78625 and std: (+/-0.03643) for {'min_samples_leaf': 50, 'min_samples_split': 240}\n",
      "Mean of: 0.78965 and std: (+/-0.03745) for {'min_samples_leaf': 50, 'min_samples_split': 260}\n",
      "Mean of: 0.78694 and std: (+/-0.03940) for {'min_samples_leaf': 50, 'min_samples_split': 280}\n",
      "Mean of: 0.78794 and std: (+/-0.03690) for {'min_samples_leaf': 50, 'min_samples_split': 300}\n",
      "Mean of: 0.78810 and std: (+/-0.03389) for {'min_samples_leaf': 50, 'min_samples_split': 320}\n",
      "Mean of: 0.78739 and std: (+/-0.03946) for {'min_samples_leaf': 50, 'min_samples_split': 340}\n",
      "Mean of: 0.78695 and std: (+/-0.03931) for {'min_samples_leaf': 50, 'min_samples_split': 360}\n",
      "Mean of: 0.78596 and std: (+/-0.03792) for {'min_samples_leaf': 50, 'min_samples_split': 380}\n",
      "Mean of: 0.78862 and std: (+/-0.03877) for {'min_samples_leaf': 60, 'min_samples_split': 100}\n",
      "Mean of: 0.78862 and std: (+/-0.03877) for {'min_samples_leaf': 60, 'min_samples_split': 120}\n",
      "Mean of: 0.78878 and std: (+/-0.03720) for {'min_samples_leaf': 60, 'min_samples_split': 140}\n",
      "Mean of: 0.78762 and std: (+/-0.03540) for {'min_samples_leaf': 60, 'min_samples_split': 160}\n",
      "Mean of: 0.78695 and std: (+/-0.03735) for {'min_samples_leaf': 60, 'min_samples_split': 180}\n",
      "Mean of: 0.78874 and std: (+/-0.03695) for {'min_samples_leaf': 60, 'min_samples_split': 200}\n",
      "Mean of: 0.78768 and std: (+/-0.03814) for {'min_samples_leaf': 60, 'min_samples_split': 220}\n",
      "Mean of: 0.78647 and std: (+/-0.03527) for {'min_samples_leaf': 60, 'min_samples_split': 240}\n",
      "Mean of: 0.78736 and std: (+/-0.03601) for {'min_samples_leaf': 60, 'min_samples_split': 260}\n",
      "Mean of: 0.78790 and std: (+/-0.03727) for {'min_samples_leaf': 60, 'min_samples_split': 280}\n",
      "Mean of: 0.78789 and std: (+/-0.03741) for {'min_samples_leaf': 60, 'min_samples_split': 300}\n",
      "Mean of: 0.78720 and std: (+/-0.03764) for {'min_samples_leaf': 60, 'min_samples_split': 320}\n",
      "Mean of: 0.78772 and std: (+/-0.03734) for {'min_samples_leaf': 60, 'min_samples_split': 340}\n",
      "Mean of: 0.78774 and std: (+/-0.03360) for {'min_samples_leaf': 60, 'min_samples_split': 360}\n",
      "Mean of: 0.78783 and std: (+/-0.03755) for {'min_samples_leaf': 60, 'min_samples_split': 380}\n",
      "Mean of: 0.78880 and std: (+/-0.04147) for {'min_samples_leaf': 70, 'min_samples_split': 100}\n",
      "Mean of: 0.78880 and std: (+/-0.04147) for {'min_samples_leaf': 70, 'min_samples_split': 120}\n",
      "Mean of: 0.78880 and std: (+/-0.04147) for {'min_samples_leaf': 70, 'min_samples_split': 140}\n",
      "Mean of: 0.78613 and std: (+/-0.03797) for {'min_samples_leaf': 70, 'min_samples_split': 160}\n",
      "Mean of: 0.78711 and std: (+/-0.04043) for {'min_samples_leaf': 70, 'min_samples_split': 180}\n",
      "Mean of: 0.78826 and std: (+/-0.03485) for {'min_samples_leaf': 70, 'min_samples_split': 200}\n",
      "Mean of: 0.78803 and std: (+/-0.03636) for {'min_samples_leaf': 70, 'min_samples_split': 220}\n",
      "Mean of: 0.78717 and std: (+/-0.03403) for {'min_samples_leaf': 70, 'min_samples_split': 240}\n",
      "Mean of: 0.78749 and std: (+/-0.03810) for {'min_samples_leaf': 70, 'min_samples_split': 260}\n",
      "Mean of: 0.78709 and std: (+/-0.03986) for {'min_samples_leaf': 70, 'min_samples_split': 280}\n",
      "Mean of: 0.78703 and std: (+/-0.03539) for {'min_samples_leaf': 70, 'min_samples_split': 300}\n",
      "Mean of: 0.78619 and std: (+/-0.03622) for {'min_samples_leaf': 70, 'min_samples_split': 320}\n",
      "Mean of: 0.78755 and std: (+/-0.03826) for {'min_samples_leaf': 70, 'min_samples_split': 340}\n",
      "Mean of: 0.78751 and std: (+/-0.03498) for {'min_samples_leaf': 70, 'min_samples_split': 360}\n",
      "Mean of: 0.78866 and std: (+/-0.03752) for {'min_samples_leaf': 70, 'min_samples_split': 380}\n"
     ]
    }
   ],
   "source": [
    "means3 = gsearch3.cv_results_['mean_test_score']\n",
    "stds3 = gsearch3.cv_results_['std_test_score']\n",
    "\n",
    "for mean, std, params in zip(means3, stds3, gsearch3.cv_results_['params']):\n",
    "        print(\"Mean of: %0.5f and std: (+/-%0.05f) for %r\"\n",
    "              % (mean, std * 2, params))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 350,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "best param:  {'min_samples_leaf': 40, 'min_samples_split': 180}\n",
      "best score AOC w/CV:  0.7899933512661959\n"
     ]
    }
   ],
   "source": [
    "print('best param: ',gsearch3.best_params_)\n",
    "print('best score AOC w/ CV: ',gsearch3.best_score_)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The AOC score is rising very slowly"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let’s fit the model again on this and have a look at the feature importance."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 351,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Model Report\n",
      "Accuracy : 0.7745\n",
      "AUC Score (Train): 0.827417\n",
      "CV Score : Mean - 0.7899934 | Std - 0.01991735 | Min - 0.7550125 | Max - 0.8124049\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x24cf63b4e10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "modelfit(gsearch3.best_estimator_, train, predictors , target)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can see a slight improvement over the base model.\n",
    "\n",
    "The importance of 'age' rose and 'tvtot' was lowered 1 place to make room for 'sclmeet'."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Third param tuning:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 355,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=5, error_score='raise',\n",
       "       estimator=GradientBoostingClassifier(criterion='friedman_mse', init=None,\n",
       "              learning_rate=0.1, loss='deviance', max_depth=5,\n",
       "              max_features=None, max_leaf_nodes=None,\n",
       "              min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "              min_samples_leaf=40, min_samples_split=180,\n",
       "              min_weight_fraction_leaf=0.0, n_estimators=60,\n",
       "              presort='auto', random_state=10, subsample=0.8, verbose=0,\n",
       "              warm_start=False),\n",
       "       fit_params=None, iid=False, n_jobs=4,\n",
       "       param_grid={'max_features': range(6, 18, 2)},\n",
       "       pre_dispatch='2*n_jobs', refit=True, return_train_score='warn',\n",
       "       scoring='roc_auc', verbose=0)"
      ]
     },
     "execution_count": 355,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "param_test4 = {'max_features':range(6,18,2)\n",
    "              }\n",
    "\n",
    "gsearch4 = GridSearchCV(estimator = ensemble.GradientBoostingClassifier(\n",
    "                            learning_rate=0.1\n",
    "                            ,n_estimators=60\n",
    "                            ,max_depth=5\n",
    "                            ,min_samples_split=180                     \n",
    "                            ,min_samples_leaf=40 \n",
    "                            ,subsample=0.8\n",
    "                            ,random_state=10)\n",
    "                            , \n",
    "                            param_grid = param_test4, scoring='roc_auc',n_jobs=4,iid=False, cv=5\n",
    "                       )\n",
    "gsearch4.fit(train[predictors] , target)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 356,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean of: 0.78936 and std: (+/-0.03620) for {'max_features': 6}\n",
      "Mean of: 0.79190 and std: (+/-0.03571) for {'max_features': 8}\n",
      "Mean of: 0.79023 and std: (+/-0.04107) for {'max_features': 10}\n",
      "Mean of: 0.79148 and std: (+/-0.03374) for {'max_features': 12}\n",
      "Mean of: 0.79065 and std: (+/-0.04029) for {'max_features': 14}\n",
      "Mean of: 0.79060 and std: (+/-0.03963) for {'max_features': 16}\n"
     ]
    }
   ],
   "source": [
    "means4 = gsearch4.cv_results_['mean_test_score']\n",
    "stds4 = gsearch4.cv_results_['std_test_score']\n",
    "\n",
    "for mean, std, params in zip(means4, stds4, gsearch4.cv_results_['params']):\n",
    "        print(\"Mean of: %0.5f and std: (+/-%0.05f) for %r\"\n",
    "              % (mean, std * 2, params))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 357,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "best param:  {'max_features': 8}\n",
      "best score AOC w/ CV:  0.7918986586371386\n"
     ]
    }
   ],
   "source": [
    "print('best param: ',gsearch4.best_params_)\n",
    "print('best score AOC w/ CV: ',gsearch4.best_score_)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The AOC increased above 0.79"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Tuning subsample and making models with lower learning rate\n",
    "\n",
    "The next step would be try different subsample values. Lets take values 0.6,0.7,0.75,0.8,0.85,0.9."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 368,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=5, error_score='raise',\n",
       "       estimator=GradientBoostingClassifier(criterion='friedman_mse', init=None,\n",
       "              learning_rate=0.1, loss='deviance', max_depth=5,\n",
       "              max_features=8, max_leaf_nodes=None,\n",
       "              min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "              min_samples_leaf=40, min_samples_split=180,\n",
       "              min_weight_fraction_leaf=0.0, n_estimators=60,\n",
       "              presort='auto', random_state=10, subsample=1.0, verbose=0,\n",
       "              warm_start=False),\n",
       "       fit_params=None, iid=False, n_jobs=4,\n",
       "       param_grid={'subsample': [0.6, 0.7, 0.75, 0.8, 0.85, 0.9]},\n",
       "       pre_dispatch='2*n_jobs', refit=True, return_train_score='warn',\n",
       "       scoring='roc_auc', verbose=0)"
      ]
     },
     "execution_count": 368,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "param_test5 = { 'subsample':[0.6,0.7,0.75,0.8,0.85,0.9]\n",
    "              }\n",
    "\n",
    "gsearch5 = GridSearchCV(estimator = ensemble.GradientBoostingClassifier\n",
    "                        (\n",
    "                            learning_rate=0.1\n",
    "                            ,n_estimators=60\n",
    "                            ,max_depth=5\n",
    "                            ,min_samples_split=180                     \n",
    "                            ,min_samples_leaf=40 \n",
    "                            ,random_state=10\n",
    "                            ,max_features = 8\n",
    "                        )\n",
    "                            , \n",
    "                            param_grid = param_test5, scoring='roc_auc',n_jobs=4,iid=False, cv=5\n",
    "                       )\n",
    "gsearch5.fit(train[predictors] , target)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 369,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean of: 0.79015 and std: (+/-0.03313) for {'subsample': 0.6}\n",
      "Mean of: 0.79302 and std: (+/-0.03474) for {'subsample': 0.7}\n",
      "Mean of: 0.79145 and std: (+/-0.03784) for {'subsample': 0.75}\n",
      "Mean of: 0.79190 and std: (+/-0.03571) for {'subsample': 0.8}\n",
      "Mean of: 0.79064 and std: (+/-0.03795) for {'subsample': 0.85}\n",
      "Mean of: 0.78972 and std: (+/-0.03608) for {'subsample': 0.9}\n"
     ]
    }
   ],
   "source": [
    "means5 = gsearch5.cv_results_['mean_test_score']\n",
    "stds5  = gsearch5.cv_results_['std_test_score']\n",
    "\n",
    "for mean, std, params in zip(means5, stds5, gsearch5.cv_results_['params']):\n",
    "        print(\"Mean of: %0.5f and std: (+/-%0.05f) for %r\"\n",
    "              % (mean, std * 2, params))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 370,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "best param:  {'subsample': 0.7}\n",
      "best score AOC w/ CV:  0.793020218934444\n"
     ]
    }
   ],
   "source": [
    "print('best param: ',gsearch5.best_params_)\n",
    "print('best score AOC w/ CV: ',gsearch5.best_score_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 371,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "GradientBoostingClassifier(criterion='friedman_mse', init=None,\n",
       "              learning_rate=0.1, loss='deviance', max_depth=5,\n",
       "              max_features=8, max_leaf_nodes=None,\n",
       "              min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "              min_samples_leaf=40, min_samples_split=180,\n",
       "              min_weight_fraction_leaf=0.0, n_estimators=60,\n",
       "              presort='auto', random_state=10, subsample=0.7, verbose=0,\n",
       "              warm_start=False)"
      ]
     },
     "execution_count": 371,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gsearch5.best_estimator_"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Finally, we have all the parameters needed. \n",
    "\n",
    "> Now, we need to lower the learning rate and increase the number of estimators proportionally. \n",
    "\n",
    "Note that these trees might not be the most optimum values but a good benchmark.(!?)\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As trees increase, it will become increasingly computationally expensive to perform CV and find the optimum values.\n",
    "\n",
    "Lets decrease the learning rate to <b>half, i.e. 0.05</b> with twice (120) the number of trees."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 376,
   "metadata": {},
   "outputs": [],
   "source": [
    "predictors = X2.columns\n",
    "\n",
    "gbm_tuned_1 = ensemble.GradientBoostingClassifier(\n",
    "                            learning_rate=0.05\n",
    "                            ,n_estimators=120\n",
    "                            ,max_depth=5\n",
    "                            ,min_samples_split=180                     \n",
    "                            ,min_samples_leaf=40 \n",
    "                            ,subsample = 0.7\n",
    "                            ,random_state=10\n",
    "                            ,max_features = 8\n",
    "                        )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 377,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Model Report\n",
      "Accuracy : 0.7681\n",
      "AUC Score (Train): 0.815651\n",
      "CV Score : Mean - 0.7879847 | Std - 0.01873715 | Min - 0.7549648 | Max - 0.8105826\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x24cf4fe2940>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "modelfit(gbm_tuned_1, train, predictors , target)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 378,
   "metadata": {},
   "outputs": [],
   "source": [
    "predictors = X2.columns\n",
    "\n",
    "gbm_tuned_2 = ensemble.GradientBoostingClassifier(\n",
    "                            learning_rate=0.01\n",
    "                            ,n_estimators=600\n",
    "                            ,max_depth=5\n",
    "                            ,min_samples_split=180                     \n",
    "                            ,min_samples_leaf=40 \n",
    "                            ,subsample = 0.7\n",
    "                            ,random_state=10\n",
    "                            ,max_features = 8\n",
    "                        )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 379,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Model Report\n",
      "Accuracy : 0.7761\n",
      "AUC Score (Train): 0.830992\n",
      "CV Score : Mean - 0.7925294 | Std - 0.01807977 | Min - 0.7620313 | Max - 0.8150461\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x24cf6426ef0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "modelfit(gbm_tuned_2, train, predictors , target)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 380,
   "metadata": {},
   "outputs": [],
   "source": [
    "predictors = X2.columns\n",
    "\n",
    "gbm_tuned_3 = ensemble.GradientBoostingClassifier(\n",
    "                            learning_rate=0.005\n",
    "                            ,n_estimators=1200\n",
    "                            ,max_depth=5\n",
    "                            ,min_samples_split=180                     \n",
    "                            ,min_samples_leaf=40 \n",
    "                            ,subsample = 0.7\n",
    "                            ,random_state=10\n",
    "                            ,max_features = 8\n",
    "                        )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 381,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Model Report\n",
      "Accuracy : 0.7761\n",
      "AUC Score (Train): 0.831353\n",
      "CV Score : Mean - 0.7922256 | Std - 0.0183006 | Min - 0.7607958 | Max - 0.8147694\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYsAAAE/CAYAAAC3u+/oAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvNQv5yAAAIABJREFUeJzt3XmYnFWd9vHvTZRFdiSKLCEBA4iKIAHXQZFdNhUUUBQBRRwYZNR5RUXEuCFu4ygqqCDKIKtixCAiIqiIJuyyXYaAEkFZJRkEJHC/f5ynSKXS3fVUdVV3J31/rquurmc7daq6u37P2WWbiIiIoSw32hmIiIixL8EiIiLaSrCIiIi2EiwiIqKtBIuIiGgrwSIiItpKsIiIiLYSLGJESbpT0qOS/q/pse4w03ytpHm9ymPN1/yupE+N5GsORtLxks4Y7XzEsi3BIkbDnrZXaXrcPZqZkfSM0Xz94Via8x5LlwSLGDMkvVzSlZL+Iel6Sa9tOnawpFskLZA0V9J7qv0rAxcB6zaXVFrv/FtLH1UJ50OSbgAekfSM6rrzJd0n6Q5JR9XM92RJrvJ4l6SHJB0uaRtJN1Tv52tN579T0m8lfVXSw5JulbRD0/F1Jc2Q9KCkOZLe3XTseEnnSTpD0nzgcOAjwH7Ve79+qM+r+bOQ9AFJ90q6R9LBTcdXkvRFSX+u8vcbSSvV+B29s3qtBdXn97Y6n18sHXJXEmOCpPWAnwJvB34G7ACcL2kz2/cB9wJ7AHOB7YCLJM2yfY2k3YAzbK/flF6dlz0A2B24H3gK+Anw42r/+sAvJN1m++Kab+NlwNQqfzOq97Ej8EzgWknn2r686dzzgLWBNwE/lDTF9oPAD4CbgHWBzYBLJM21fWl17d7Am4F3ACtUaTzf9oFNeRn086qOrwOsDqwH7AScJ+kC2w8BXwBeCLwS+FuV16eG+h0B/wT+B9jG9m2SngesVfNzi6VAShYxGi6o7kz/IemCat+BwEzbM20/ZfsSYDbwegDbP7V9u4vLgZ8D/zbMfPyP7btsPwpsA0y0Pd32v2zPBb4F7N9Bep+0/ZjtnwOPAD+wfa/tvwK/BrZqOvde4L9tP2H7bOA2YHdJGwCvBj5UpXUd8G3KF3TD72xfUH1Ojw6UkRqf1xPA9Or1ZwL/B2wqaTngEOB9tv9q+0nbV9p+nDa/I0rAfZGklWzfY/umDj67GOMSLGI0vMH2GtXjDdW+DYE3NwWRf1C+NJ8HIGk3SVdVVTP/oHxBrT3MfNzV9HxDSlVW8+t/BHhuB+n9ven5owNsr9K0/VcvPovnnykliXWBB20vaDm23iD5HlCNz+sB2wubtv9Z5W9tYEXg9gGSHfR3ZPsRYD9Ktdg9kn5alThiGZFgEWPFXcD3m4LIGrZXtn2CpBWA8ynVI8+1vQYwE2jUNQ00dfIjwLOattcZ4Jzm6+4C7mh5/VVtv36A63phPS1eVzYJuLt6rCVp1ZZjfx0k30ts1/i8hnI/8Biw8QDHBv0dAdi+2PZOlAB/K6VkFsuIBIsYK84A9pS0i6QJklasGmLXB5an1M3fByys2ih2brr278CzJa3etO864PWS1pK0DnB0m9f/AzC/avReqcrDiyRt07N3uLjnAEdJeqakNwMvoFTx3AVcCXy2+gy2AA4F/neItP4OTK6qkKD95zUo208BpwJfqhraJ0h6RRWABv0dSXqupL1UOhw8TqnWerLDzyTGsASLGBOqL8m9KVU/91HuYv8LWK6qkjkKOAd4CHgrpQG5ce2tlEbhuVX1yLrA94HrgTsp9fVnt3n9J4E9gS2BOyh32N+mNAL3w+8pjeH3A58G9rX9QHXsAGAypZTxI+DjVfvAYM6tfj4g6Zp2n1cNHwRuBGYBDwKfo/weBv0dVY8PVHl+EHgN8O8dvGaMccriRxEjS9I7gXfZfvVo5yWirpQsIiKirQSLiIhoK9VQERHRVkoWERHR1jIz3cfaa6/tyZMnj3Y2IiKWKldfffX9tie2O2+ZCRaTJ09m9uzZo52NiIiliqQ/1zkv1VAREdFWgkVERLSVYBEREW0lWERERFsJFhER0VaCRUREtJVgERERbfU1WEjaVdJt1aLzxwxx3r4qC95Pa9r34eq62yTt0s98RkTE0Po2KE/SBOAkymLw84BZkmbYvrnlvFUpc+//vmnf5pS1j19IWWbyF5I2qdYciIiIEdbPEdzbAnOqhe+RdBZl4ZSbW877JHAiZcGVhr2Bs6pF4u+QNKdK73edZGDyMT9te86dJ+zeSZIREeNSP6uh1mPxheXnsfii80jaCtjA9oWdXltdf5ik2ZJm33fffb3JdURELKGfwWKgxeGfng+9Wi/4y5SlGDu69ukd9im2p9meNnFi23mwIiKiS/2shpoHbNC0vT5lfd6GVYEXAb+SBLAOMEPSXjWujYiIEdTPksUsYKqkKZKWpzRYP71ovO2Hba9te7LtycBVwF62Z1fn7S9pBUlTKAvb/6GPeY2IiCH0rWRhe6GkI4GLgQnAqbZvkjQdmG17xhDX3iTpHEpj+ELgiPSEiogYPX1dz8L2TGBmy77jBjn3tS3bnwY+3bfMRUREbRnBHRERbSVYREREWwkWERHRVoJFRES0lWARERFtJVhERERbCRYREdFWgkVERLSVYBEREW0lWERERFsJFhER0VaCRUREtJVgERERbSVYREREWwkWERHRVoJFRES01ddgIWlXSbdJmiPpmAGOHy7pRknXSfqNpM2r/ZMlPVrtv07SN/uZz4iIGFrfVsqTNAE4CdgJmAfMkjTD9s1Np51p+5vV+XsBXwJ2rY7dbnvLfuUvIiLq62fJYltgju25tv8FnAXs3XyC7flNmysD7mN+IiKiS/0MFusBdzVtz6v2LUbSEZJuB04Ejmo6NEXStZIul/RvfcxnRES00c9goQH2LVFysH2S7Y2BDwHHVrvvASbZ3gp4P3CmpNWWeAHpMEmzJc2+7777epj1iIho1s9gMQ/YoGl7feDuIc4/C3gDgO3HbT9QPb8auB3YpPUC26fYnmZ72sSJE3uW8YiIWFw/g8UsYKqkKZKWB/YHZjSfIGlq0+buwJ+q/ROrBnIkbQRMBeb2Ma8RETGEvvWGsr1Q0pHAxcAE4FTbN0maDsy2PQM4UtKOwBPAQ8BB1eXbAdMlLQSeBA63/WC/8hoREUPrW7AAsD0TmNmy77im5+8b5LrzgfP7mbeIiKgvI7gjIqKtBIuIiGgrwSIiItqqHSwkrdzPjERExNjVNlhIeqWkm4Fbqu2XSPp633MWERFjRp2SxZeBXYDGILnrKV1bIyJinKhVDWX7rpZdT/YhLxERMUbVGWdxl6RXAq5GYh9FVSUVERHjQ52SxeHAEZQZY+cBW1bbERExTgxZsqjmZ3q77beNUH4iImIMGrJkYftJWhYsioiI8adOm8VvJX0NOBt4pLHT9jV9y1VERIwpdYLFK6uf05v2GXhd77MTERFjUdtgYXv7kchIRESMXXVGcK8u6UuN5UslfVHS6iORuYiIGBvqdJ09FVgAvKV6zAdO62emIiJibKnTZrGx7X2atj8h6bp+ZSgiIsaeOiWLRyW9urEh6VXAo3USl7SrpNskzZF0zADHD5d0o6TrJP1G0uZNxz5cXXebpF3qvF5ERPRHnZLFe4HTm9opHgLe2e6iakDfScBOlJHfsyTNsH1z02ln2v5mdf5ewJeAXaugsT/wQmBd4BeSNqnGfURExAir0xvqOuAlklartufXTHtbYI7tuQCSzqIM8Hs6WLSktTKlSy7VeWfZfhy4Q9KcKr3f1XztiIjooTq9oT4jaQ3b823Pl7SmpE/VSHs9oHm22nnVvtb0j5B0O3AiZZLCTq49rNFL67777quRpYiI6EadNovdbP+jsWH7IeD1Na7TAPu8xA77JNsbAx8Cju3w2lNsT7M9beLEiTWyFBER3agTLCZIWqGxIWklYIUhzm+YB2zQtL0+cPcQ558FvKHLayMioo/qBIszgEslHSrpEOAS4PQa180CpkqaUq2DsT8wo/kESVObNncH/lQ9nwHsL2kFSVOAqcAfarxmRET0QZ0G7hMl3QDsWO36pO2La1y3UNKRwMXABOBU2zdJmg7Mtj0DOFLSjsATlF5WB1XX3iTpHEpj+ELgiPSEiogYPXW6zmL7Z5JmUdbevr9u4rZnAjNb9h3X9Px9Q1z7aeDTdV8rIiL6Z9BqKEkXSnpR9fx5wB+BQ4DvSzp6hPIXERFjwFBtFlNs/7F6fjBwie09gZdRgkZERIwTQwWLJ5qe70BVnWR7AfBUPzMVERFjy1BtFndJ+g9KN9aXAj+Dp7vOPnME8hYREWPEUCWLQylzM70T2K9pYN7LyRTlERHjyqAlC9v3AocPsP8y4LJ+ZioiIsaWOoPyIiJinEuwiIiIthIsIiKirTpTlG8i6VJJf6y2t5B0bLvrIiJi2VGnZPEt4MNU4y5s30CZFDAiIsaJOsHiWbZbZ3xd2I/MRETE2FQnWNwvaWOqxYck7Qvc09dcRUTEmFJn1tkjgFOAzST9FbgDOLCvuYqIiDGlznoWc4EdJa0MLFfNDRUREeNInd5Qn5G0hu1HbC+QtKakT41E5iIiYmyo02axW9O8UNh+CHh9/7IUERFjTZ1gMUHSCo2NatbZFYY4/2mSdpV0m6Q5ko4Z4Pj7Jd0s6YZqLMeGTceelHRd9ZjRem1ERIycOg3cZwCXSjqN0iPqEOD0dhdJmgCcBOxEmeZ8lqQZtm9uOu1aYJrtf0p6L3AisF917FHbW9Z/KxER0S91GrhPlHQjZQEkAZ+0fXGNtLcF5lQN5Eg6C9gbeDpYVDPYNlxFellFRIxJdUoW2L4IuKjDtNcD7mrankdZknUwh7a8xoqSZlMGAJ5g+4LWCyQdBhwGMGnSpA6zFxERddXpDfUmSX+S9LCk+ZIWSJpfI20NsM+DvMaBwDTg8027J9meBrwV+O9qYODiidmn2J5me9rEiRNrZCkiIrpRp2RxIrCn7Vs6THsesEHT9vrA3a0nSdoR+CjwGtuPN/bbvrv6OVfSr4CtgNs7zENERPRAnd5Qf+8iUADMAqZKmiJpecrkg4v1apK0FXAysFe1Ml9j/5qNHliS1gZeRVNbR0REjKw6JYvZks4GLgCa7/x/ONRFthdKOhK4GJgAnGr7JknTgdm2Z1CqnVYBzpUE8BfbewEvAE6W9BQloJ3Q0osqIiJGUJ1gsRrwT2Dnpn0GhgwWALZnAjNb9h3X9HzHQa67EnhxjbxFRMQIqNN19uCRyEhERIxdbYOFpBUp3VpfCKzY2G/7kD7mKyIixpA6DdzfB9YBdgEup/RqysyzERHjSJ1g8XzbHwMesX06sDtpT4iIGFfqBIsnqp//kPQiYHVgct9yFBERY06d3lCnSFoTOJYyTmIV4GN9zVVERIwpdYLFpdUaFlcAGwFImtLXXEVExJhSpxrq/AH2ndfrjERExNg1aMlC0maU7rKrS3pT06HVaOpCGxERy76hqqE2BfYA1gD2bNq/AHh3PzMVERFjy6DBwvaPJV0IfMj2Z0YwTxERMcYM2WZh+0nKsqgRETGO1ekNdaWkrwFnA480dtq+pm+5ioiIMaVOsHhl9XN60z4Dr+t9diIiYiyqM+vs9iORkYiIGLvqrMG9uqQvSZpdPb4oafWRyFxERIwNdQblnUrpLvuW6jEfOK2fmYqIiLGlTrDY2PbHbc+tHp+gmvajHUm7SrpN0hxJxwxw/P2SbpZ0g6RLJW3YdOwgSX+qHgfVf0sREdFrdYLFo5Je3diQ9Crg0XYXSZoAnATsBmwOHCBp85bTrgWm2d6CMoXIidW1awEfB14GbAt8vJrMMCIiRkGd3lDvBU6v2ikEPAjUudPfFphjey6ApLOAvYGbGyfYvqzp/KuAA6vnuwCX2H6wuvYSYFfgBzVeNyIieqxOb6jrgJdIWq3anl8z7fWAu5q251FKCoM5FLhoiGvXa71A0mHAYQCTJk2qma2IiOhUnd5Qz5b0P8CvgMskfUXSs2ukrQH2eZDXOBCYBny+k2ttn2J7mu1pEydOrJGliIjoRp02i7OA+4B9gH2r52fXuG4esEHT9vrA3a0nSdoR+Ciwl+3HO7k2IiJGRp1gsZbtT9q+o3p8ijITbTuzgKmSpkhaHtifstLe0yRtBZxMCRT3Nh26GNhZ0ppVw/bO1b6IiBgFdYLFZZL2l7Rc9XgL8NN2F9leCBxJ+ZK/BTjH9k2Spkvaqzrt85RlWs+VdJ2kGdW1DwKfpAScWcD0RmN3RESMPNkDNiMsOkFaAKwMPFXtWo5FEwra9mr9y15906ZN8+zZsxfbN/mYtjGNO0/YvV9ZiogY8yRdbXtau/Pq9IZatTdZioiIpVWdcRZI2gKY3Hy+7R/2KU8RETHGtA0Wkk4FtgBuYlFVlIEEi4iIcaJOyeLltlun6YiIiHGkTm+o3w0wp1NERIwjdUoWp1MCxt+Axymjq11N/hcREeNAnWBxKvB24EYWtVlERMQ4UidY/MX2jPanRUTEsqpOsLhV0pnATyjVUEC6zkZEjCd1gsVKlCCxc9O+dJ2NiBhH6ozgPngkMhIREWPXoMFC0lcZZP0JANtH9SVHEREx5gxVspg9xLGIiBhHBg0Wtk8fyYxERMTYVWcEd0REjHMJFhER0VZfg4WkXSXdJmmOpGMGOL6dpGskLZS0b8uxJ6vV855eQS8iIkZH22AhaRNJl0r6Y7W9haRja1w3ATgJ2A3YHDhggAkJ/wK8EzhzgCQetb1l9dhrgOMRETFC6pQsvgV8GHgCwPYNwP41rtsWmGN7ru1/AWcBezefYPvOKr3MORURMYbVCRbPsv2Hln0La1y3HnBX0/a8al9dK0qaLekqSW/o4LqIiOixOtN93C9pY6oBelXbwj01rtMA+wYd5DeASbbvlrQR8EtJN9q+fbEXkA4DDgOYNGlSB0lHREQn6pQsjgBOBjaT9FfgaODwGtfNAzZo2l4fuLtuxmzfXf2cC/wK2GqAc06xPc32tIkTJ9ZNOiIiOjRkyULScsA02ztKWhlYzvaCmmnPAqZKmgL8ldLO8dY6F0paE/in7cclrQ28Cjix5utGRESPDVmysP0UcGT1/JEOAgW2F1bXXgzcApxj+yZJ0yXtBSBpG0nzgDcDJ0u6qbr8BcBsSdcDlwEn2L65w/cWERE9UqfN4hJJHwTOBh5p7LT9YLsLbc8EZrbsO67p+SxK9VTrdVcCL66Rt4iIGAF1gsUh1c8jmvYZ2Kj32YmIiLGoznoWU0YiIxERMXa1DRaS3jHQftvf6312IiJiLKpTDbVN0/MVgR2Aa4AEi4iIcaJONdR/NG9LWh34ft9yFBERY043s87+E5ja64xERMTYVafN4icsmqZjOcoMsuf2M1MRETG21Gmz+ELT84XAn23P61N+IiJiDKpTDfV625dXj9/anifpc33PWUREjBl1gsVOA+zbrdcZiYiIsWvQaihJ7wX+HdhI0g1Nh1YFftvvjEVExNgxVJvFmcBFwGeB5vWzF9SZFyoiIpYdgwYL2w8DDwMHAEh6DmVQ3iqSVrH9l5HJYkREjLa2bRaS9pT0J+AO4HLgTkqJIyIixok6XWc/Bbwc+IXtrSRtT1XaGFeOX73GOQ/3Px8REaOgTm+oJ2w/ACwnaTnblwFb9jlfERExhtQpWfxD0irAr4H/lXQvZXBeRESME3VKFntT5oM6GvgZcDuwZ53EJe0q6TZJcyQdM8Dx7SRdI2mhpH1bjh0k6U/V46A6rxcREf1RZ9bZRyRtCEy1fbqkZwET2l0naQJwEmVQ3zxglqQZLWtp/wV4J/DBlmvXAj4OTKPMS3V1de1D9d5WRET0Up3eUO8GzgNOrnatB1xQI+1tgTm259r+F3AWpZTyNNt32r4BeKrl2l2AS2w/WAWIS4Bda7xmRET0QZ1qqCOAVwHzAWz/CXhOjevWA+5q2p5X7auj1rWSDpM0W9Ls++67r2bSERHRqTrB4vGqZACApGewaMryoWiAfXWuq32t7VNsT7M9beLEiTWTjoiITtUJFpdL+giwkqSdKGtZ/KTGdfOADZq21wfurpmv4VwbERE9VidYHAPcB9wIvAeYCRxb47pZwFRJUyQtD+wPzKiZr4uBnSWtKWlNYOdqX0REjIKhZp2dZPsvtp8CvlU9arO9UNKRlC/5CcCptm+SNB2YbXuGpG2AHwFrAntK+oTtF9p+UNInKQEHYHomL4yIGD1DdZ29AHgpgKTzbe/TaeK2Z1JKIs37jmt6PotSxTTQtacCp3b6mhER0XtDVUM1NzJv1O+MRETE2DVUsPAgzyMiYpwZqhrqJZLmU0oYK1XPqbZte7W+5y4iIsaEoRY/ajulR0REjA91us5GRMQ4l2ARERFtJVhERERbCRYREdFWgkVERLRVZ1nV6KEXn/7iWufdeNCNfc5JRER9KVlERERbKVksxW7Z7AVtz3nBrbeMQE4iYlmXkkVERLSVYBEREW0lWERERFsJFhER0VaCRUREtNXX3lCSdgW+QllW9du2T2g5vgLwPWBr4AFgP9t3SpoM3ALcVp16le3D+5nX8e6kw3/Z9pwjvvm6EchJRIxFfQsWkiYAJwE7AfOAWZJm2L656bRDgYdsP1/S/sDngP2qY7fb3rJf+YuIiPr6WQ21LTDH9lzb/wLOAvZuOWdv4PTq+XnADpJERESMKf2shloPuKtpex7wssHOsb1Q0sPAs6tjUyRdC8wHjrX969YXkHQYcBjApEmTepv76MoX99uj7TkfOPvCEchJRPRSP4PFQCWE1rW8BzvnHmCS7QckbQ1cIOmFtucvdqJ9CnAKwLRp07JO+DJm3jFL3B8sYf0T/q3tOccff3yt16t7XsR41M9qqHnABk3b6wN3D3aOpGcAqwMP2n7c9gMAtq8Gbgc26WNeIyJiCP0MFrOAqZKmSFoe2B+Y0XLODOCg6vm+wC9tW9LEqoEcSRsBU4G5fcxrREQMoW/VUFUbxJHAxZSus6favknSdGC27RnAd4DvS5oDPEgJKADbAdMlLQSeBA63/WC/8hpR16W/3LjtOTu87vYRyEnEyOrrOAvbM4GZLfuOa3r+GPDmAa47Hzi/n3mLiIj6MoI7IiLaSrCIiIi2EiwiIqKtBIuIiGgrwSIiItpKsIiIiLYSLCIioq0Ei4iIaKuvg/IiYmDrXHZdrfP+tn2WdImxISWLiIhoKyWLiKXc5GN+2vacO0/YfQRyEsuylCwiIqKtlCwi4mkppcRgEiwioveOX73meQ/3Nx/RMwkWETGmvfj0F7c958aDbhyBnIxvabOIiIi2UrKIiHHjls1e0PacF9x6S9tzTjr8l23POeKbr6uVp6VFX4OFpF2Br1CWVf227RNajq8AfA/YGngA2M/2ndWxDwOHUpZVPcr2xf3Ma0TEaPjifnu0PecDZ184AjkZWt+qoSRNAE4CdgM2Bw6QtHnLaYcCD9l+PvBl4HPVtZtT1uN+IbAr8PUqvYiIGAX9bLPYFphje67tfwFnAXu3nLM3cHr1/DxgB0mq9p9l+3HbdwBzqvQiImIUyHZ/Epb2BXa1/a5q++3Ay2wf2XTOH6tz5lXbtwMvA44HrrJ9RrX/O8BFts9reY3DgMOqzU2B22pkbW3g/mG8tV6nM1bTSp5GPq3kaeTTSp5gQ9sT2yXUzzYLDbCvNTINdk6da7F9CnBKR5mSZtue1sk1/UxnrKaVPI18WsnTyKeVPNXXz2qoecAGTdvrA3cPdo6kZwCrAw/WvDYiIkZIP4PFLGCqpCmSlqc0WM9oOWcGcFD1fF/gly71YjOA/SWtIGkKMBX4Qx/zGhERQ+hbNZTthZKOBC6mdJ091fZNkqYDs23PAL4DfF/SHEqJYv/q2psknQPcDCwEjrD9ZI+y1lG11QikM1bTSp5GPq3kaeTTSp5q6lsDd0RELDsy3UdERLSVYBEREW0lWERERFsJFqOk6uXVdl+NdF5VZ99IUrFB+zNrpTWm3p+k5SS9crRev58kbTrEsXzm49wyHywkTZV0nqSbJc1tPLpM60hJa/Yoa+cPsO+8Afa189Wa+2qRtI6kvSTtKWmdbtKouj9f0G0eWvTs/Um6tM6+odh+CvhiN6/f8rrvljS1ei5Jp0maL+kGSS/tMs3PSFqjaXtNSZ/qIIlbJJ0uaZUBjnX8mUtaWdJy1fNNqr+rZ3aaTg8/83Oann+u5djPO0jn/UM9usjXcyR9ovqeOrd6/twu0nld0/MpLcfe1Gl6rZb5YAGcBnyD0gV3e8ost9/vMq11gFmSzpG0azWPVUckbSZpH2B1SW9qerwTWLGDdF4h6QPAxJY/1uMpXZU7JuldlPEsb6KMe7lK0iHdpFVdu02X1/b0/UlaUdJawNrVF+ha1WMysG4X2fu5pH26+f03eR9wZ/X8AGALYArwfspMzd3YzfY/Ghu2HwJe38H1N1EGxF4j6eUtx7p5r1cAK0paD7gUOBj4bhfpQG8+86lNz3dqOdZ2uosmqzY9PtiyvWonGapKbLOqze8BZ1TPf99Fae4LTc9bb0aP7TCtJYyH9SxWsn2pJNn+M3C8pF8DH+80IdvHSvoYsDPlD/9r1d3Kd2zfXjOZTYE9gDWAPZv2LwDe3UF2lgdWofwOm/9A51O+6LvxX8BWth8AkPRs4Erg1C7S2h54j6Q/A49Qvmxse4ua1/fy/b0HOJoSGK5m0RfffMrMyJ16P7AysFDSYyx6b6t1kMZC209Uz/cAvld97r+QdGIXeQKYIGkF248DSFoJWKGD65+w/VFJFwNnSPoe8Knqzr6bPvay/U9JhwJftX2ipGu7SAd685kP9R5qvz/bn2g8l/SG5u0ufBF4g+3mz+XHkn4EnEyZK68uDfJ8oO2OjYdg8VhVFP6TyiDBvwLP6TYx25b0N+BvlNLKmsB5ki6x/f9qXP9jyh/DK2z/bhj5uBy4XNJ3bf9Z0qpV9v6v2zQpd5ULmrYXAHd1mdZuw8jHEu8PSt01sIrt+R2m9RXgK5L+w3bXVXRN6XV09ziIpyQ9D3gI2AH4dNOxlbpM8wzgUkmnUb78DmHRrM612b5C0jTg68CvJb2ty/xI0iuAt1GWI4BB4rJwAAAS3ElEQVQuv3N69Jk/S9JWlBqVlZqq+0T3n/lwB6qt1hIoSqL2ddX/dLd5ac3XsAfUjYdgcTTwLOAo4JOUO96DhrxiEJKOqq69H/g28F+2n2gEI6BtsGjyQFVf/lzbL5K0BbCX7U7qmAFWre7W1qryeD9wkO0/dpgOlED6e0k/pvxx7Q38oVEPa/tL7RKQtFr1Zb6g3bk1fVbS4ZRFsK6mVN99yfbnu0jrKUlrNKpqVNqfDrD99ToXq01bgu1rOsjLccBsSpXaDNs3Va/xGqCrNrXqzv1GSvAR8MkOFw17+u6z+ozeKukg4NeU/6FOHQ18GPhRNSvDRsBlXaRTMld+X1Npqq61fUUHSdxDuZMX5Wavudrmb93ma5gkac2qyrB551p03kywkaQZlPfXeE613XHnmSUyOl5GcEta2fYjw0xjOqXK6c8DHHuB7fbrMS46/3JKtc/Jtreq9v3R9os6zNOVwEdtX1Ztvxb4jO2Oe49IGrJqrk5xW9KFtveQdAdLziBs2xt1mKfrbG9Z3d1uDXwIuLqD6qwl0mrZd23j869xfeOLbkVgGnA95f1tAfze9qs7zM8zgFWbvygkrUz5vxxOCbEr1c3Q1bZ/27L/NcCxtlvr+UdM1Z72PsqkotcBLwd+Z7v22qWStgXusn1PtX0QsA+l7eh42w/WTOdGFt2pP5+y3s7TOvnbVFlm4d2Uto/GzcbWlIXgTrV9cgdpvWao41VpvXu2l+kH8ArKHFN/qbZfAnx9GOm9lFJK+Q/gpcNIZ1b189qmfdd1kc71dfYtrQ9Ko+szgXOB1wzn/QE3UN0gVdsTgJu6SOcs4MVN2y8CvtthGv+v6fmbW459psO0flP9XEBph2k8FgDzO0jnQmCLAfZPA37Sxed0GfDL1keXv7sbKUH6ump7M+DsDtO4Blirer4dZSbrfSg1Dud1kM5U4NXAhi2P7YDnd/He9qB0BnigelwB7NnN59SU5kRg4nDSaH2Mh2qo/wZ2oZrx1vb1krbrJqGqcfstwA+rXadJOtedVx0B3C9pY6o7FJXFou7pIp25Vb4aPbwOBO7oJAFJP2GIOk3be3WRr15UGwB8k3Lndz1whaQNKV+E3bgYOEfSNynv93DgZ12ks5ntGxsbtv8oacuhLhjA/kCjIfvDlGDYsCvwkboJuSrRePj1+pNt3zBA+rOrnmOd+mDT8xUpX8wLu8saj9l+TBJVI/6tGmJcyCAmeFHpYT/gFNvnA+dLuq6DdL4MfMQtNQySJlbH9hzwqkHYvpASqIel6il2HOVGVsBykhZSOhdMH2764yFYYPuulh533c5g+1ZKb6HHACSdQLlb6SZYHEGZEXIzSX+lfMEf2EU6hwCfYFEAuwJ4Z4dpfKH9KZ0ZrNoA6KTaYDng77bXa9r3F0q7Uzc+ROkZ9V7KP9PPKW1PnbpF0rcpDcqm/N5qV0FWetZzparfHpRrVq8wdNftjhuAbV/dsuu3VfVrN+apjCG5ALhE0kN0vsbNBEnPsL2Q0q5zWNOxTr4LexZUJR03xGHb/mQHyR1NKfFs47IcNVU70Tck/aftL3eSt1bjIVjcpTL60yrrahxF5//YDXdS/qEeq7ZXAOp2mV2M7bnAjlUd9XK2u20Q3tH2Uc07JL2Zxe9U2+VleHWZA3sfsA1ledztJW1GCWq12X6q6sF2TtM+0+XdqUsX0G9Uj+E4mBJw3ldtX9FFmr3suXI1i9qHJlF6WInSPfsv1G/cnCXp3ba/1byz6vra+sXfVksQW45SF9/tQM83Vk+Pr9qOVqfzUuEPKD3s7gcepTTcI+n5wMMdpNPLoDpQO+rKlN5jz6ZUkdX1DmAn208vo2p7rqQDKTdGwwoWy3wDt6S1KYOcdmTR3eT7XI0l6DCtCyhfgJdQ/jl3An4D3AvQ+qXdJq3nAp8B1rW9m6TNgVfY/k6HebrG9kvb7auZ1qso659vSLmRaPRl76hRukprlu1tquL9y2w/PlADc410Pkb5xz6bpn+sDu6Wm9OaCnwW2JzFq8Y6fn/DJelJFo0/WQn4Z+MQsKLtjkc6V9VrM2zPrLZ3o9xMfKDm9c8FfgT8i0XBYRplzMsbbXfUY6ilk8NCSul5uu3fdJJOU3qvBqbaPq2q8lmlcQfdQRovB54H/NxVhxdJm1Rp1erNJukHlLaXgYLqzrb36yRPTdevSrkBOZRyg/RF2/d2cP2gHWS66TyzRBrLerDopar3xKBs1+7TLukiyujyj9p+SdUz5lrbL655/W6U0blvoXyRNqwGbG5727p5aUrzVuA/KV8UT1fVdRlYf0S5Az+aUvX0EPBM252MKG584bTqNoD9hjIYs1GvfDDlf6CjAZoDBNVGpkY86DSTdLXtrVv2dbwGs6TtKY32UDoA/LJXeexW1VNvGrCp7U0krQuca3vE56zqQ1BdizLo8G2UcTFfcUtX2prpDHqT2O0N5GJpLOvBQtL/DLD7YcpqfT/uIr3lKT0xDNxm+19d5qtx5/10181O7rwlvQTYEphOadRqWABc1uUf2+9tdzJidKA0prTe7VVd+lYHftbp5yVpxUYb0VD7aqZ1te2tJd3YCMqSfm373zpMp2dBtZdURl7/msXbUrazvcsI52PIeYhs/3Co44OkeR2wFXBN0//LDe6iC3Wv9CKoSvo8ZXqdU4CTPIwu002l1SUO0WVptdl4aLNYkfLl3qjD34fSHfNQSdvbPrpuQpJeTxmCfzvVQBdJ77F9URf5ekRlOo1Gb6iX00G9qe3rgeslPQr82NUUD93QosFml1V/vD8Enk6vbvG8ch6wtaRLbe9QXT+cNpErKd2V2+2ro1ej+R/u8nfebwdQSk4/ovxdXVHtG2lD9QYyizpjdOJfti2p8f+yclc56yGXsU1dDzKsfIDyv3Ys8NGmjjgdT2diu6s54eoaDyWLX1LqERdW28+gtFvsBNxoe/MO0roV2MP2nGp7Y+CntjfrIl8vpczk+SLgj5R+0fsO1MuiTTqnUap5rqD0/7+48V47SGOoP3i7s4FP11J6rLyLARrUXGMUeJXOOsB6lLvkt7Koh9BqwDe7/My3oXRuWIPScLg6cKLtqzpM5wTKGI3hBNWekjQBOMH2f41WHvpJ0gcp3bB3orQ7HQKc6R5M3xL1jIeSxXqU3gWNu/aVKY3KT0rq9G783kagqMylatzulO1rquqZTSlfhLd50cRynaRzsMq0z7tRvlS/rjJP1bs6SKPbrqgD2R94A0tOANipXShdgNcHmgPMAjoYg9DMdmN2z/+jtFd0q1FV12gfEOWOuXZQ7bXq73nr9meOnKrk/HFKd05TOoNM77K67nHgF5QxNpsCx9m+pFd5jfbGQ7A4EbhO0q8o/9TbAZ+pirG/6DCtmyTNpPRUMPBmSnfDN0FndbHVneDrgcmU38POkmrfeTdzmZ/qoipPK1G+rGsHC7WZg7+TPNm+DfhcVZ/cdVVN1VngdEn7uAyc6pp6P+jwQhafysTAfElb2u5kcFevXasyH9C5LN5zrJtqn144i1Li3afafhulM8aOXaT1XEpPoWsosyB3+r8bw7TMV0MBVD0n3g7cSilZzHPnI4kbVT6Dse3aaz9UQecxyjQGTzUl0tFYBEm7Uu7mtwd+Rfln/HknVVHqwZxQTWn1JPD0MoCpx3PmSDqT0vulMWnb7pQ1CTaj9NDpdorxYRnk77Ojv8te6lXvrKZrxaLlAaZRbto6WR4ghmGZL1moByOJG2wPp+qi1fo96slxMGWw0Xu8aB2Dz1FGK9fSaYBqoxdTSfcynaeDQVWafNRlcF6jdNfJeg8Nz6bMC/Z/VTofpzTsb0fpITUqwaLHf5+9cJmk/Vk0qHJf4KfdJlY1cHe9PEAMzzJfslCZIbIxknhLVSOJuxk4I2lFyoCZF7L4oK6O79yqL/RLbddeznGQdAYalNdVl0KVqQG+QgmopgTV/3QZbb7Uk3QVZZBa40t+FUoprKMZeiXdAryk0Q1Y0gqUCe5eoA5mse01SetTOk28ikVtBO+zPW+U8rOAUpJvdC+ewKLqsY56+mjJ5QEucNPyALY37l3OYyDjYVnVx7xoLqcVbN9KaSDrxvcp0xXsAlxOKa10O03HVcCPJD2qsvbyAkm1J8iT9N4qEG6qsm5z43EHZXbVbpxJuQt8HmVVuXMppZaOSdpI0k8k3SfpXkk/roLRqKRTWbG5H3v1vJt1Gs6kLBv78apU8VvgB1XJ5eYu89YLp1GqxtaldOz4SbVvVNhe1fZytp9ZPZar9q3aSaCorA28yfYuts9tdAapSol79DzzsST3cArbsfig9DlfgzLi9grgx8DMLtO6tvp5Q/XzmXQ/5fJcyjoI6vL61SmN4z9g8WmS1xrGZ/X7AfZd1WVaV1HaiZ5RPQ4cKP2RSqdK67c0TStPqff+XZdpbU2p3jwamDbcv9NePBhgivuB9o1yHjcb7Tzk0d1jma+GajackcTV9X+wva2kK4B/p9Sd/sHdTT1xMbCbq/rzsaAaP/AwJQCZMo3zClTrVLuD+Zg0wGhwSVfZfnmHeepJOtV10ygdAO6mvL91gf285OyoSyVJvwC+y6LS4AHAwa4GR44Fkv5ie9Jo5yM6t8w3cDfz8GdXPUVljYZjKcX9VYCPdZnWPcCvqi6vzQO7Ou4620ONdpx3Vz8bXUMPoXy5dhIUL5P0YRYPPD9VNRNpB4GnV+kATKFMGTEJeCOL2maWFYcAX6MMhjRlpPuI94TSwFPsAE/PhBtLoXFVshiuqiFzH0r1T2OeFbuLhUUG667q3vZM6oiklSglpsYgql8D33B38zA15odq/IF1tbxqr9Kp0rrB9hYqs5d+hrIe80daSy4xPFXDdmMai1ZftL32CGcpeiDBogOSfkappmmdQO6Lo5apHpJ0DmWE7P9Wuw4A1rD9li7S6kng6XEAu9b2VpI+S5nq5czR7L3UK+rtAjrDpjLFzrG2rxzg2B22666vEWNIgkUH1Is54fu0hGkvSLre9kva7auZVk8CT48D2IWUyQN3pDRQP0ppc+r4/Y0lkgZar+LpBXRsrzLC+VmL0gvxn21PjqXGuGqz6IErJb3YTesvd6HnS5j20LWSXu5qYj1JL6P0IOrGpi1fwpdJun4U04Gy9seuwBds/0PS84ClfuK95pKtFi2gczBluo0RL/UO1I4k6dke5SncY3gSLGqoxjOY8nkdLGkupT62MY1w7QFw7v1o4l56GfAOlXWuoTQE39J4/528T3oXeHoWwKo73R82bd9D6Wiw1NOSC+i81F2sadKjvJxACcj3Vz3QzgGeUpnw8h096GgSoyDVUDVI2nCo47b/3EWaPRlN3Eu9fJ/VKOdNKWtAQxV4KPNg1Q48vUpnWaYeLqDTo/w0Ly51GfD/bM9SWb70THc5N1SMrgSLUaIBVsUbaN/SqleBpx+Belkj6SlKSXchi7eHdbyATo/ycyvwItsLW8fENAeSWLqkGmr0PCLppa4WzKmK64+Ocp56pldf4gkG7dkea9P2nATMrKqjfibpvynVfztQJvOMpVBKFqNkWR9NHOObpNcC7wU2odyU3kVZQfE0d7HIV4y+lCxGz7I+mjjGMdu/oqyvMihJB7kschVLgbFWfB1PPmZ7PmX6g50ojZPfGN0sRYyo9412BqK+BIvR0xgBvjvwTds/BpYfxfxEjDS1PyXGigSL0fNXSSdTBorNrOadyu8jxpNUuy5F8uU0et4CXAzsavsfwFosA6OJIzqQksVSJA3co2RZHk0cASBpiu07htjX7VQyMQrSdTYi+kIDrw9/te2tRytP0b2ULCKipyRtBrwQWF3Sm5oOrQasODq5iuFKsIiIXtsU2IPSLXzPpv0LWLQKYyxlUg0VEX0h6RW2fzfa+YjeSLCIiJ6S9FWGXuDrqBHMTvRIqqEiotdmj3YGovdSsoiIiLZSsoiIvqgWPlribtT260YhOzFMCRYR0S8fbHq+IrAPZYGmWAqlGioiRoyky22/ZrTzEZ1LySIi+kLSWk2bywFbA+uMUnZimBIsIqJfrqa0WYhS/XQHcOio5ii6lmqoiIhoKyWLiOiplvmglmD7h0Mdj7EpwSIieq0xH9RzgFcCv6y2t6esy51gsRRKsIiInrJ9MICkC4HNq7VakPQ84KTRzFt0LyvlRUS/TG4EisrfKTPSxlIoJYuI6JdfSboY+AGlV9T+wKWjm6XoVnpDRUTfSHojsF21+RDwXNtHjGKWokuphoqIfroDeAJ4I6WB+5bRzU50K9VQEdFTkjahVDkdADwAnE2pxdh+VDMWw5JqqIjoKUlPAb8GDrU9p9o31/ZGo5uzGI5UQ0VEr+0D/A24TNK3JO1AmfIjlmIpWUREX0haGXgDpTrqdcDpwI9s/3xUMxZdSbCIiL6rZqB9M7BfFj9aOiVYREREW2mziIiIthIsIiKirQSLiIhoK8EiIiLa+v90v/MePfC2TwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x24cf64cbfd0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "modelfit(gbm_tuned_3, train, predictors , target)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here we see that the score reduced very slightly. So lets run for 1500 trees."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 382,
   "metadata": {},
   "outputs": [],
   "source": [
    "predictors = X2.columns\n",
    "\n",
    "gbm_tuned_4 = ensemble.GradientBoostingClassifier(\n",
    "                            learning_rate=0.005\n",
    "                            ,n_estimators=1500\n",
    "                            ,max_depth=5\n",
    "                            ,min_samples_split=180                     \n",
    "                            ,min_samples_leaf=40 \n",
    "                            ,subsample = 0.7\n",
    "                            ,random_state=10\n",
    "                            ,max_features = 8\n",
    "                        )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 385,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Model Report\n",
      "Accuracy : 0.7804\n",
      "AUC Score (Train): 0.837651\n",
      "CV Score : Mean> 0.7925545 | Std> 0.0180046 | Min> 0.7617991 | Max> 0.8148918\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x24cf6546e80>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "modelfit(gbm_tuned_4, train, predictors , target)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A microscopic improvement:\n",
    "\n",
    "    gbm_tuned_1 CV Score = 0.7879847\n",
    "    gbm_tuned_2 CV Score = 0.7925294\n",
    "    gbm_tuned_3 CV Score = 0.7922256\n",
    "    gbm_tuned_4 CV Score = 0.7925545"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 393,
   "metadata": {},
   "outputs": [],
   "source": [
    "predictors = X2.columns\n",
    "\n",
    "gbm_tuned_5 = ensemble.GradientBoostingClassifier(\n",
    "                            loss = 'exponential'\n",
    "                            ,learning_rate=0.005\n",
    "                            ,n_estimators=1500\n",
    "                            ,max_depth=5\n",
    "                            ,min_samples_split=180                     \n",
    "                            ,min_samples_leaf=40 \n",
    "                            ,subsample = 0.7\n",
    "                            ,random_state=10\n",
    "                            ,max_features = 8\n",
    "                        )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 394,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Model Report\n",
      "Accuracy : 0.7794\n",
      "AUC Score (Train): 0.838994\n",
      "CV Score : Mean_ 0.7926079 | Std_ 0.01820108 | Min_ 0.7617069 | Max_ 0.8158141\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x24cf664a978>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "modelfit(gbm_tuned_5, train, predictors , target)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The `loss`=exponential parameter performed a bit better than 'deviance' :)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "### The age is the most important predictor for whether the person is single or not.\n",
    "\n",
    "Then <b>happiness</b> is anothe factor but without a huge impact, but still on second place close to `tvtot` and `sclmeet` (social interraction)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python [conda root]",
   "language": "python",
   "name": "conda-root-py"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.5.4"
  },
  "latex_envs": {
   "LaTeX_envs_menu_present": true,
   "autocomplete": true,
   "bibliofile": "biblio.bib",
   "cite_by": "apalike",
   "current_citInitial": 1,
   "eqLabelWithNumbers": true,
   "eqNumInitial": 1,
   "hotkeys": {
    "equation": "Ctrl-E",
    "itemize": "Ctrl-I"
   },
   "labels_anchors": false,
   "latex_user_defs": false,
   "report_style_numbering": false,
   "user_envs_cfg": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}