YannAhlgrim
629 lines
131 KiB
Plaintext
629 lines
131 KiB
Plaintext
{
|
|
"cells": [
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "72dba059",
|
|
"metadata": {},
|
|
"source": [
|
|
"# Lernfall 4"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "4105df59",
|
|
"metadata": {},
|
|
"source": [
|
|
"## Datensatz laden"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 27,
|
|
"id": "d7a53682",
|
|
"metadata": {},
|
|
"outputs": [
|
|
{
|
|
"data": {
|
|
"text/html": [
|
|
"<div>\n",
|
|
"<style scoped>\n",
|
|
" .dataframe tbody tr th:only-of-type {\n",
|
|
" vertical-align: middle;\n",
|
|
" }\n",
|
|
"\n",
|
|
" .dataframe tbody tr th {\n",
|
|
" vertical-align: top;\n",
|
|
" }\n",
|
|
"\n",
|
|
" .dataframe thead th {\n",
|
|
" text-align: right;\n",
|
|
" }\n",
|
|
"</style>\n",
|
|
"<table border=\"1\" class=\"dataframe\">\n",
|
|
" <thead>\n",
|
|
" <tr style=\"text-align: right;\">\n",
|
|
" <th></th>\n",
|
|
" <th>Vertragsnummer</th>\n",
|
|
" <th>Jahr</th>\n",
|
|
" <th>Vertragsalter</th>\n",
|
|
" <th>Status</th>\n",
|
|
" <th>Marktzins</th>\n",
|
|
" <th>Tarifzins</th>\n",
|
|
" <th>Guthaben</th>\n",
|
|
" <th>Sparbeitrag</th>\n",
|
|
" <th>Beruf</th>\n",
|
|
" <th>Alter</th>\n",
|
|
" <th>p</th>\n",
|
|
" </tr>\n",
|
|
" </thead>\n",
|
|
" <tbody>\n",
|
|
" <tr>\n",
|
|
" <th>0</th>\n",
|
|
" <td>10000</td>\n",
|
|
" <td>1</td>\n",
|
|
" <td>1</td>\n",
|
|
" <td>gekuendigt</td>\n",
|
|
" <td>8.7</td>\n",
|
|
" <td>8.7</td>\n",
|
|
" <td>3933.85</td>\n",
|
|
" <td>3619.0</td>\n",
|
|
" <td>angestellt</td>\n",
|
|
" <td>48</td>\n",
|
|
" <td>0.154465</td>\n",
|
|
" </tr>\n",
|
|
" <tr>\n",
|
|
" <th>1</th>\n",
|
|
" <td>9999</td>\n",
|
|
" <td>1</td>\n",
|
|
" <td>1</td>\n",
|
|
" <td>aktiv</td>\n",
|
|
" <td>8.7</td>\n",
|
|
" <td>8.7</td>\n",
|
|
" <td>523.93</td>\n",
|
|
" <td>482.0</td>\n",
|
|
" <td>angestellt</td>\n",
|
|
" <td>28</td>\n",
|
|
" <td>0.154465</td>\n",
|
|
" </tr>\n",
|
|
" <tr>\n",
|
|
" <th>2</th>\n",
|
|
" <td>9998</td>\n",
|
|
" <td>1</td>\n",
|
|
" <td>1</td>\n",
|
|
" <td>aktiv</td>\n",
|
|
" <td>8.7</td>\n",
|
|
" <td>8.7</td>\n",
|
|
" <td>7281.81</td>\n",
|
|
" <td>6699.0</td>\n",
|
|
" <td>selbstaendig</td>\n",
|
|
" <td>22</td>\n",
|
|
" <td>0.231475</td>\n",
|
|
" </tr>\n",
|
|
" <tr>\n",
|
|
" <th>3</th>\n",
|
|
" <td>9997</td>\n",
|
|
" <td>1</td>\n",
|
|
" <td>1</td>\n",
|
|
" <td>aktiv</td>\n",
|
|
" <td>8.7</td>\n",
|
|
" <td>8.7</td>\n",
|
|
" <td>8905.79</td>\n",
|
|
" <td>8193.0</td>\n",
|
|
" <td>angestellt</td>\n",
|
|
" <td>26</td>\n",
|
|
" <td>0.154465</td>\n",
|
|
" </tr>\n",
|
|
" <tr>\n",
|
|
" <th>4</th>\n",
|
|
" <td>9996</td>\n",
|
|
" <td>1</td>\n",
|
|
" <td>1</td>\n",
|
|
" <td>aktiv</td>\n",
|
|
" <td>8.7</td>\n",
|
|
" <td>8.7</td>\n",
|
|
" <td>8566.65</td>\n",
|
|
" <td>7881.0</td>\n",
|
|
" <td>angestellt</td>\n",
|
|
" <td>30</td>\n",
|
|
" <td>0.154465</td>\n",
|
|
" </tr>\n",
|
|
" </tbody>\n",
|
|
"</table>\n",
|
|
"</div>"
|
|
],
|
|
"text/plain": [
|
|
" Vertragsnummer Jahr Vertragsalter Status Marktzins Tarifzins \\\n",
|
|
"0 10000 1 1 gekuendigt 8.7 8.7 \n",
|
|
"1 9999 1 1 aktiv 8.7 8.7 \n",
|
|
"2 9998 1 1 aktiv 8.7 8.7 \n",
|
|
"3 9997 1 1 aktiv 8.7 8.7 \n",
|
|
"4 9996 1 1 aktiv 8.7 8.7 \n",
|
|
"\n",
|
|
" Guthaben Sparbeitrag Beruf Alter p \n",
|
|
"0 3933.85 3619.0 angestellt 48 0.154465 \n",
|
|
"1 523.93 482.0 angestellt 28 0.154465 \n",
|
|
"2 7281.81 6699.0 selbstaendig 22 0.231475 \n",
|
|
"3 8905.79 8193.0 angestellt 26 0.154465 \n",
|
|
"4 8566.65 7881.0 angestellt 30 0.154465 "
|
|
]
|
|
},
|
|
"execution_count": 27,
|
|
"metadata": {},
|
|
"output_type": "execute_result"
|
|
}
|
|
],
|
|
"source": [
|
|
"import pandas as pd\n",
|
|
"\n",
|
|
"data = pd.read_csv('../3_lernfall/VertragsdatenABCAG.csv', sep=',', decimal='.') # Daten einlesen\n",
|
|
"\n",
|
|
"# Erste 5 Zeilen anzeigen\n",
|
|
"data.head()"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "a3ee9779",
|
|
"metadata": {},
|
|
"source": [
|
|
"## Dummies"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 28,
|
|
"id": "ef9e13fc",
|
|
"metadata": {},
|
|
"outputs": [
|
|
{
|
|
"data": {
|
|
"text/html": [
|
|
"<div>\n",
|
|
"<style scoped>\n",
|
|
" .dataframe tbody tr th:only-of-type {\n",
|
|
" vertical-align: middle;\n",
|
|
" }\n",
|
|
"\n",
|
|
" .dataframe tbody tr th {\n",
|
|
" vertical-align: top;\n",
|
|
" }\n",
|
|
"\n",
|
|
" .dataframe thead th {\n",
|
|
" text-align: right;\n",
|
|
" }\n",
|
|
"</style>\n",
|
|
"<table border=\"1\" class=\"dataframe\">\n",
|
|
" <thead>\n",
|
|
" <tr style=\"text-align: right;\">\n",
|
|
" <th></th>\n",
|
|
" <th>Vertragsnummer</th>\n",
|
|
" <th>Jahr</th>\n",
|
|
" <th>Vertragsalter</th>\n",
|
|
" <th>Marktzins</th>\n",
|
|
" <th>Tarifzins</th>\n",
|
|
" <th>Guthaben</th>\n",
|
|
" <th>Sparbeitrag</th>\n",
|
|
" <th>Alter</th>\n",
|
|
" <th>p</th>\n",
|
|
" <th>Beruf_arbeitslos</th>\n",
|
|
" <th>Beruf_selbstaendig</th>\n",
|
|
" <th>Status_aktiv</th>\n",
|
|
" <th>Status_gekuendigt</th>\n",
|
|
" <th>Status_verstorben</th>\n",
|
|
" </tr>\n",
|
|
" </thead>\n",
|
|
" <tbody>\n",
|
|
" <tr>\n",
|
|
" <th>0</th>\n",
|
|
" <td>10000</td>\n",
|
|
" <td>1</td>\n",
|
|
" <td>1</td>\n",
|
|
" <td>8.7</td>\n",
|
|
" <td>8.7</td>\n",
|
|
" <td>3933.85</td>\n",
|
|
" <td>3619.0</td>\n",
|
|
" <td>48</td>\n",
|
|
" <td>0.154465</td>\n",
|
|
" <td>0</td>\n",
|
|
" <td>0</td>\n",
|
|
" <td>0</td>\n",
|
|
" <td>1</td>\n",
|
|
" <td>0</td>\n",
|
|
" </tr>\n",
|
|
" <tr>\n",
|
|
" <th>1</th>\n",
|
|
" <td>9999</td>\n",
|
|
" <td>1</td>\n",
|
|
" <td>1</td>\n",
|
|
" <td>8.7</td>\n",
|
|
" <td>8.7</td>\n",
|
|
" <td>523.93</td>\n",
|
|
" <td>482.0</td>\n",
|
|
" <td>28</td>\n",
|
|
" <td>0.154465</td>\n",
|
|
" <td>0</td>\n",
|
|
" <td>0</td>\n",
|
|
" <td>1</td>\n",
|
|
" <td>0</td>\n",
|
|
" <td>0</td>\n",
|
|
" </tr>\n",
|
|
" <tr>\n",
|
|
" <th>2</th>\n",
|
|
" <td>9998</td>\n",
|
|
" <td>1</td>\n",
|
|
" <td>1</td>\n",
|
|
" <td>8.7</td>\n",
|
|
" <td>8.7</td>\n",
|
|
" <td>7281.81</td>\n",
|
|
" <td>6699.0</td>\n",
|
|
" <td>22</td>\n",
|
|
" <td>0.231475</td>\n",
|
|
" <td>0</td>\n",
|
|
" <td>1</td>\n",
|
|
" <td>1</td>\n",
|
|
" <td>0</td>\n",
|
|
" <td>0</td>\n",
|
|
" </tr>\n",
|
|
" <tr>\n",
|
|
" <th>3</th>\n",
|
|
" <td>9997</td>\n",
|
|
" <td>1</td>\n",
|
|
" <td>1</td>\n",
|
|
" <td>8.7</td>\n",
|
|
" <td>8.7</td>\n",
|
|
" <td>8905.79</td>\n",
|
|
" <td>8193.0</td>\n",
|
|
" <td>26</td>\n",
|
|
" <td>0.154465</td>\n",
|
|
" <td>0</td>\n",
|
|
" <td>0</td>\n",
|
|
" <td>1</td>\n",
|
|
" <td>0</td>\n",
|
|
" <td>0</td>\n",
|
|
" </tr>\n",
|
|
" <tr>\n",
|
|
" <th>4</th>\n",
|
|
" <td>9996</td>\n",
|
|
" <td>1</td>\n",
|
|
" <td>1</td>\n",
|
|
" <td>8.7</td>\n",
|
|
" <td>8.7</td>\n",
|
|
" <td>8566.65</td>\n",
|
|
" <td>7881.0</td>\n",
|
|
" <td>30</td>\n",
|
|
" <td>0.154465</td>\n",
|
|
" <td>0</td>\n",
|
|
" <td>0</td>\n",
|
|
" <td>1</td>\n",
|
|
" <td>0</td>\n",
|
|
" <td>0</td>\n",
|
|
" </tr>\n",
|
|
" </tbody>\n",
|
|
"</table>\n",
|
|
"</div>"
|
|
],
|
|
"text/plain": [
|
|
" Vertragsnummer Jahr Vertragsalter Marktzins Tarifzins Guthaben \\\n",
|
|
"0 10000 1 1 8.7 8.7 3933.85 \n",
|
|
"1 9999 1 1 8.7 8.7 523.93 \n",
|
|
"2 9998 1 1 8.7 8.7 7281.81 \n",
|
|
"3 9997 1 1 8.7 8.7 8905.79 \n",
|
|
"4 9996 1 1 8.7 8.7 8566.65 \n",
|
|
"\n",
|
|
" Sparbeitrag Alter p Beruf_arbeitslos Beruf_selbstaendig \\\n",
|
|
"0 3619.0 48 0.154465 0 0 \n",
|
|
"1 482.0 28 0.154465 0 0 \n",
|
|
"2 6699.0 22 0.231475 0 1 \n",
|
|
"3 8193.0 26 0.154465 0 0 \n",
|
|
"4 7881.0 30 0.154465 0 0 \n",
|
|
"\n",
|
|
" Status_aktiv Status_gekuendigt Status_verstorben \n",
|
|
"0 0 1 0 \n",
|
|
"1 1 0 0 \n",
|
|
"2 1 0 0 \n",
|
|
"3 1 0 0 \n",
|
|
"4 1 0 0 "
|
|
]
|
|
},
|
|
"execution_count": 28,
|
|
"metadata": {},
|
|
"output_type": "execute_result"
|
|
}
|
|
],
|
|
"source": [
|
|
"# One hot encoding für die Spalten \"Beruf\" und \"Status\"\n",
|
|
"data = pd.get_dummies(data, columns=['Beruf', 'Status'], drop_first=True)\n",
|
|
"\n",
|
|
"data.head()"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "a6a24941",
|
|
"metadata": {},
|
|
"source": [
|
|
"## Korrelationen darstellen"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 29,
|
|
"id": "81f1e895",
|
|
"metadata": {},
|
|
"outputs": [
|
|
{
|
|
"data": {
|
|
"image/png": "iVBORw0KGgoAAAANSUhEUgAAA90AAAJOCAYAAACqS2TfAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy88F64QAAAACXBIWXMAAA9hAAAPYQGoP6dpAAB9P0lEQVR4nOzdd1yV9eP+8euADAXEgVsUUXMU7pyVG7c5SnOvhlluy6wUtXJlWmmOMre5Cs2V28gcuVcq+jEVB7hwoqLC/fvDL+fnCVAkDzdwXs/Hg0ee+77POdcZHbjO/b7ft8UwDEMAAAAAAOCZczI7AAAAAAAA6RWlGwAAAAAAO6F0AwAAAABgJ5RuAAAAAADshNINAAAAAICdULoBAAAAALATSjcAAAAAAHZC6QYAAAAAwE4o3QAAAAAA2AmlG0CaMHPmTFksFu3atctm+eXLl1WhQgV5enpq3bp1JqWLz8/PT507d07WdUeMGKGlS5fGW/7777/LYrHo999//0/Z0ovOnTvLz8/PZlliz93TOHXqlCwWi2bOnPmfbicpatSoIYvF8tifGjVq2Gz/6OWnva/kXvdZsFgsGjp06FNfL+71GDt27BO3jfucOHXq1NMHfMx9x/0/l9D74q+//lLz5s1VoEABubm5KVeuXKpSpYr69+9vs92kSZP+83vqWby/UyuLxaL333/f7BhPbejQobJYLDbL/svn/08//aSvv/46wXWPvv/8/PyS9f8TAHNkMDsAACTX2bNnVbduXV24cEHr169X5cqVzY70TIwYMUKvvfaamjVrZrO8XLly2rZtm0qWLGlOsFRm8ODB6t27t82yxJ671GrSpEm6ceNGgutGjhypZcuWqXnz5jbbp1Xbtm1T/vz5zY7xTK1cuVJNmzZVjRo1NGbMGOXJk0fh4eHatWuXFixYoK+++sq67aRJk+Tj45PsMialvfe3o1qyZIkyZ86crOv+9NNPOnTokPr06fNsQwEwFaUbQJp0/Phx1alTR/fv31dISIgCAgL+823euXNH7u7u8fZaSNLt27eVKVOm/3wf/0XmzJnTzRcLz0LhwoXNjpAkj3tfJfYFSnBwsJYvX642bdrYfLGQ1r5wMQxDd+/eVcaMGdPle3fMmDEqVKiQ1qxZowwZ/v+fVG+88YbGjBljYjKYqWzZsmZHAJDKMLwcQJqzb98+vfTSS8qQIYP+/PPPeIX7zz//VO3ateXl5aVMmTKpatWqWrlypc02ccNQ165dq65duypHjhzKlCmToqOjVaNGDb3wwgv6448/VLVqVWXKlEldu3aVJN24cUMDBgxQoUKF5Orqqnz58qlPnz6Kiop6bOa7d++qf//+KlOmjLy9vZUtWzZVqVJFv/76q812FotFUVFRmjVrVrzhxYkNL1+2bJmqVKmiTJkyycvLS3Xr1tW2bdtstokbAvn333+rTZs28vb2Vq5cudS1a1ddv37dZlvDMDRp0iSVKVNGGTNmVNasWfXaa6/pn3/+sdku7nnauXOnXn75ZWXKlEn+/v4aNWqUYmNjH/t8xD3W999/XzNmzFCxYsWUMWNGVahQQdu3b5dhGPryyy9VqFAheXp6qlatWvrf//5nc/1/Dy9/3HOXmPPnz6tVq1by8vKSt7e3WrdurYiIiAS33bVrl5o2baps2bLJ3d1dZcuW1aJFi2y2edz7KqkOHz6sTp06KSAgQNOmTbNZl9AQ8Xv37unzzz9X8eLF5ebmphw5cqhLly66dOlSovdx//595cyZUx06dIi37tq1a8qYMaP69esnKenvXen/v6ZTpkxRiRIl5ObmplmzZlnXPToc9tKlS+rRo4dKliwpT09P5cyZU7Vq1dLmzZsTzBwbG6svvvhCBQoUkLu7uypUqKANGzYk+hgftX79etWuXVuZM2dWpkyZVK1atSRf93GuXLkiHx8fm8Idx8np//+J5efnp7///lshISHW92bce/dZfDYkNMRZSni4/caNG1WjRg1lz55dGTNmVIECBdSyZUvdvn07yY/bMAyNGDFCBQsWtL4W69atS/D9mdzPTMMw9PHHH8vFxUU//PCD9XlMaKTAf7nfuPfsnDlzVKJECWXKlEmlS5fWihUr4t3PypUrVaZMGbm5ualQoUKJHvKQUM6///5bgYGBypQpk3LkyKH33ntPK1eutPlMr1GjhlauXKnTp0/bHGICIO1jTzeANOXPP//U0KFD5evrq7Vr1ypPnjw260NCQlS3bl2VKlVKP/74o9zc3DRp0iQ1adJE8+fPV+vWrW2279q1qxo1aqQ5c+YoKipKLi4ukqTw8HC1b99eH374oUaMGCEnJyfdvn1b1atX19mzZ/Xxxx+rVKlS+vvvvzVkyBAdPHhQ69evT/QPpOjoaEVGRmrAgAHKly+f7t27p/Xr16tFixaaMWOGOnbsKOnhENxatWqpZs2aGjx4sCQ9dpjiTz/9pHbt2ikwMFDz589XdHS0xowZoxo1amjDhg166aWXbLZv2bKlWrdurW7duungwYMaNGiQJGn69OnWbd555x3NnDlTvXr10ujRoxUZGanhw4eratWq2r9/v3LlymXdNiIiQu3atVP//v0VFBSkJUuWaNCgQcqbN6/1MT3OihUrtHfvXo0aNUoWi0UDBw5Uo0aN1KlTJ/3zzz+aOHGirl+/rn79+qlly5bat29fos/x0z53d+7cUZ06dXT+/HmNHDlSzz33nFauXBnvPSJJmzZtUv369VWpUiVNmTJF3t7eWrBggVq3bq3bt2/H+wM7sffVk1y/fl3NmzdXhgwZFBwc/MTRFbGxsXr11Ve1efNmffjhh6patapOnz6toKAg1ahRQ7t27VLGjBnjXc/FxUXt27fXlClT9N1339k8T/Pnz9fdu3fVpUsXSUl/78ZZunSpNm/erCFDhih37tzKmTNngtkjIyMlSUFBQcqdO7du3bqlJUuWWN+7/y5REydOVMGCBfX1118rNjZWY8aMUYMGDRQSEqIqVaok+hzNnTtXHTt21KuvvqpZs2bJxcVFU6dOVb169bRmzRrVrl070ev6+fnJMAzr5Uf/LUlVqlTRtGnT1KtXL7Vr107lypVL8LVesmSJXnvtNXl7e1sPEXBzc5Nkv8+GhJw6dUqNGjXSyy+/rOnTpytLliw6d+6cVq9erXv37iV5NM8nn3yikSNH6u2331aLFi105swZvfnmm7p//76ee+4563bJ/cyMjo5W586dtXLlSi1fvlz169d/qsf5tPe7cuVK7dy5U8OHD5enp6fGjBmj5s2bKzQ0VP7+/pKkDRs26NVXX1WVKlW0YMECxcTEaMyYMbpw4cIT84SHh6t69ery8PDQ5MmTlTNnTs2fPz/eMeyTJk3S22+/rRMnTmjJkiXxbufR99+zmrcAQAoxACANmDFjhiHJkGR4e3sbFy9eTHC7ypUrGzlz5jRu3rxpXfbgwQPjhRdeMPLnz2/Exsba3F7Hjh3j3Ub16tUNScaGDRtslo8cOdJwcnIydu7cabP8559/NiQZq1atsi4rWLCg0alTp0Qfz4MHD4z79+8b3bp1M8qWLWuzzsPDI8Hrbtq0yZBkbNq0yTAMw4iJiTHy5s1rBAQEGDExMdbtbt68aeTMmdOoWrWqdVlQUJAhyRgzZozNbfbo0cNwd3e3Pi/btm0zJBlfffWVzXZnzpwxMmbMaHz44YfWZXHP019//WWzbcmSJY169eol+tjjSDJy585t3Lp1y7ps6dKlhiSjTJky1kyGYRhff/21Ick4cOCAdVmnTp2MggUL2txmYs9dQiZPnmxIMn799Veb5W+99ZYhyZgxY4Z1WfHixY2yZcsa9+/ft9m2cePGRp48eazP/+PeV08SGxtrNGnSxHBycjJWrlyZ4DbVq1c3qlevbr08f/58Q5Lxyy+/2Gy3c+dOQ5IxadKkRK974MABQ5Lx/fff21y3YsWKRvny5RPN+bj3btz/n5GRkfGuJ8kICgp64u3Wrl3baN68uXX5yZMnDUlG3rx5jTt37liX37hxw8iWLZtRp04d67K45//kyZOGYRhGVFSUkS1bNqNJkyY29xUTE2OULl3aqFixYqJ5kuLy5cvGSy+9ZP1scnFxMapWrWqMHDnS5jPIMAzj+eeft3n+E5Ocz4a4/7//7d/PR9xn1b59+5L8GP8tMjLScHNzM1q3bm2zPO6z49HH+DSfmZKM9957z7hy5Yrx0ksvGfny5YuXM7HP1X+/t5/2fnPlymXcuHHDuiwiIsJwcnIyRo4caV1WqVKlRN+D/37u/53zgw8+MCwWi/H333/bbFevXj2bz3TDMIxGjRrF+1wDkPYxvBxAmtK0aVNdv35dffr0UUxMjM26qKgo/fXXX3rttdfk6elpXe7s7KwOHTro7NmzCg0NtblOy5YtE7yfrFmzqlatWjbLVqxYoRdeeEFlypTRgwcPrD/16tVL0qziixcvVrVq1eTp6akMGTLIxcVFP/74o44cOfIUz8D/FxoaqvPnz6tDhw42Q1k9PT3VsmVLbd++Pd6Q0aZNm9pcLlWqlO7evauLFy9aH6PFYlH79u1tHmPu3LlVunTpeI8xd+7cqlixYrzbPH36dJIeQ82aNeXh4WG9XKJECUlSgwYNbPZExS1P6u0mxaZNm+Tl5RXvOWnbtq3N5f/97386evSo2rVrJ0k2z0vDhg0VHh6e5PfV4wwdOlTLly/X0KFD1bBhwyRdZ8WKFcqSJYuaNGlik6tMmTLKnTv3Y9+TAQEBKl++vGbMmGFdduTIEe3YscN6OEWcp3nv1qpVS1mzZk1S/ilTpqhcuXJyd3e33u6GDRsSvN0WLVrI3d3detnLy0tNmjTRH3/8Ee+zIM7WrVsVGRmpTp062Tw/sbGxql+/vnbu3PnEYc6Pkz17dm3evFk7d+7UqFGj9Oqrr+rYsWMaNGiQAgICdPny5STdzrP+bEhMmTJl5OrqqrfffluzZs2Kd8hIUmzfvl3R0dFq1aqVzfLKlSvHO5vA035mnjx5UlWqVNGNGze0fft2lS5d+qnzJed+a9asKS8vL+vlXLlyKWfOnNbPm6ioKO3cuTPR9+CThISE6IUXXog3J0ObNm2S9fgApD2UbgBpyuDBgzVkyBD99NNPat++vc0f21evXpVhGPGGnEtS3rx5JT08BvNRCW2b2PILFy7owIEDcnFxsfnx8vKSYRiP/QM7ODhYrVq1Ur58+TR37lxt27ZNO3fuVNeuXXX37t0kPfZ/i3ssiT3e2NhYXb161WZ59uzZbS7HDXG9c+eO9TEahqFcuXLFe5zbt2+P9xj/fXtxtxl3e0+SLVs2m8uurq6PXZ7c5yohV65csRkqHyd37tw2l+OGjw4YMCDec9KjRw9Jive8JPa+SsyyZcv02WefqUmTJvr000+TfL0LFy7o2rVrcnV1jZctIiLiiaWva9eu2rZtm44ePSpJmjFjhtzc3GzKwNO+d5P62MeNG6d3331XlSpV0i+//KLt27dr586dql+/foLvn3+/LnHL7t27p1u3biV4H3Gv3WuvvRbv+Rk9erQMw7AOc/8vKlSooIEDB2rx4sU6f/68+vbtq1OnTiVpMjV7fDYkpnDhwlq/fr1y5syp9957T4ULF1bhwoX1zTffJPk24j53Evp/59/LnvYzc8eOHTp27Jhat279n2a6f9r7fdLn2NWrVxUbG5voe/BJEvusSWgZgPSJY7oBpDnDhg2TxWLRsGHDFBsbq3nz5ilDhgzKmjWrnJycFB4eHu8658+flyT5+PjYLE/s+OCElvv4+Chjxow2xz//e31i5s6dq0KFCmnhwoU2t/00E2z9W9wfiok9XicnpyTvcYzj4+Mji8WizZs3Wwv5oxJallZlz55dO3bsiLf83xOpxb2ugwYNUosWLRK8rWLFitlcfprJj0JDQ9WhQwcVKVJEc+bMearr+vj4KHv27Fq9enWC6x/de5eQNm3aqF+/fpo5c6a++OILzZkzR82aNbN53zztezep+efOnasaNWpo8uTJNstv3ryZ4PYJTXAXEREhV1dXm5Etj4p77SZMmJDo7OnPuvi4uLgoKChI48eP16FDh564/bP4bIjb+xodHW3z/2hCX7q8/PLLevnllxUTE6Ndu3ZpwoQJ6tOnj3LlyqU33njjifcV97mT0LHMERERNnu7n/Yzs3Xr1sqdO7c++eQTxcbGxvsCyt3dPcHn5fLlyza39V8+qxOSNWtWWSyWRN+DT5I9e/ZEny8AjoHSDSBNGjp0qJycnBQUFCTDMPTTTz/Jw8NDlSpVUnBwsMaOHWudQCo2NlZz585V/vz5bSb5eVqNGzfWiBEjlD17dhUqVOiprmuxWOTq6mrzR3VERESCM0AndU9xsWLFlC9fPv30008aMGCA9bajoqL0yy+/WGc0fxqNGzfWqFGjdO7cuXjDR9OCp9nLXrNmTS1atEjLli2zGWL+008/2WxXrFgxFS1aVPv379eIESOead6bN2+qefPmio2N1ZIlS+Tt7f1U12/cuLF1UqdKlSo99f1nzZpVzZo10+zZs1WlShVFRETEG1r+NO/dp2GxWOJ9iXPgwAFt27ZNvr6+8bYPDg7Wl19+aS2YN2/e1PLly/Xyyy/L2dk5wfuoVq2asmTJosOHD8ebtOpZCA8PT3DPftyw8LgRNlLi781n8dkQV3QPHDigF1980bp8+fLliWZ3dnZWpUqVVLx4cc2bN0979uxJUumuVKmS3NzctHDhQpsvobZv367Tp0/blO7kfGZ++umn8vLyUt++fRUVFaWRI0faPM4DBw7YbH/s2DGFhobaFOn/8lmdEA8PD1WsWDHR9+CTVK9eXWPHjtXhw4dthpgvWLAg3rZP8xkGIO2gdANIs4YMGSInJycNHjxYhmFo/vz5GjlypOrWrauaNWtqwIABcnV11aRJk3To0CHNnz//P51+pU+fPvrll1/0yiuvqG/fvipVqpRiY2MVFhamtWvXqn///okWn8aNGys4OFg9evTQa6+9pjNnzuizzz5Tnjx5dPz4cZttAwIC9Pvvv2v58uXKkyePvLy84u1JlR6ekmjMmDFq166dGjdurHfeeUfR0dH68ssvde3aNY0aNeqpH2O1atX09ttvq0uXLtq1a5deeeUVeXh4KDw83Hp6tnffffepbzelJPW5k6SOHTtq/Pjx6tixo7744gsVLVpUq1at0po1a+JtO3XqVDVo0ED16tVT586dlS9fPkVGRurIkSPas2ePFi9enKy8HTt21JEjRzRgwADdvHlT27dvj7eNm5tbouf9feONNzRv3jw1bNhQvXv3VsWKFeXi4qKzZ89q06ZNevXVV9W8efPHZujatasWLlyo999/X/nz51edOnVs1j/Ne/dpNG7cWJ999pmCgoJUvXp1hYaGavjw4SpUqJAePHgQb3tnZ2fVrVtX/fr1U2xsrEaPHq0bN25o2LBhid6Hp6enJkyYoE6dOikyMlKvvfaacubMqUuXLmn//v26dOlSvD3tT6NevXrKnz+/mjRpouLFiys2Nlb79u3TV199JU9PT5tzrAcEBGjBggVauHCh/P395e7uroCAgGfy2dCwYUNly5ZN3bp10/Dhw5UhQwbNnDlTZ86csbn+lClTtHHjRjVq1EgFChTQ3bt3rXuD//26JyZbtmzq16+fRo4cqaxZs6p58+Y6e/ashg0bpjx58tjML5Hcz8zevXvL09NTb7/9tm7duqVvv/1WFotFHTp0UPv27dWjRw+1bNlSp0+f1pgxY5QjRw6b6/+Xz+rEfPbZZ6pfv77q1q2r/v37KyYmRqNHj5aHh8cTD1Ho06ePpk+frgYNGmj48OHKlSuXfvrpJ+thHY8+ZwEBAQoODtbkyZNVvnx5OTk5qUKFCk+VFUAqZN4cbgCQdHGz8P57NlrDMIwvvvjCkGS0aNHCuHfvnrF582ajVq1ahoeHh5ExY0ajcuXKxvLly5N8e9WrVzeef/75BHPcunXL+PTTT41ixYoZrq6uhre3txEQEGD07dvXiIiIsG6X0Cy7o0aNMvz8/Aw3NzejRIkSxg8//JDgrMP79u0zqlWrZmTKlMlmNuB/z14eZ+nSpUalSpUMd3d3w8PDw6hdu7axZcsWm23i7ufSpUsJPg9xsxvHmT59ulGpUiXrc1i4cGGjY8eOxq5du574PCU0q3hC9H+zFT8qbqbqL7/80mZ53GNfvHjxY+8nsecuMWfPnjVatmxpeHp6Gl5eXkbLli2NrVu3xpu93DAMY//+/UarVq2MnDlzGi4uLkbu3LmNWrVqGVOmTLFu87j3VWLPwZN+Hn2M/56l2TAM4/79+8bYsWON0qVLG+7u7oanp6dRvHhx45133jGOHz/+2OsaxsOZvH19fQ1JxieffJJgzqS+dxN6TR9d9+js5dHR0caAAQOMfPnyGe7u7ka5cuWMpUuXxntd494To0ePNoYNG2bkz5/fcHV1NcqWLWusWbPG5j4Sez+HhIQYjRo1MrJly2a4uLgY+fLlMxo1amTzfkqOhQsXGm3btjWKFi1qeHp6Gi4uLkaBAgWMDh06GIcPH7bZ9tSpU0ZgYKDh5eUV73X9r58NhmEYO3bsMKpWrWp4eHgY+fLlM4KCgoxp06bZPB/btm0zmjdvbhQsWNBwc3MzsmfPblSvXt1YtmzZUz3u2NhY4/PPP7e+FqVKlTJWrFhhlC5d2mbmecNI+mdmQu+d+fPnGxkyZDC6dOlixMTEGLGxscaYMWMMf39/w93d3ahQoYKxcePGBN/b/+V+DSPhz/Bly5YZpUqVMlxdXY0CBQoYo0aNSvB1Sui6hw4dMurUqWO4u7sb2bJlM7p162bMmjXLkGTs37/ful1kZKTx2muvGVmyZDEsFkuCs9IDSHsshvGvk04CAAAAT+HkyZMqXry4goKC9PHHH5sdJ014++23NX/+fF25csU6WSSA9Inh5QAAAEiy/fv3a/78+apataoyZ86s0NBQjRkzRpkzZ1a3bt3MjpcqDR8+XHnz5pW/v79u3bqlFStWaNq0afr0008p3IADoHQDAABAMTExetwASIvFImdnZ3l4eGjXrl368ccfde3aNXl7e6tGjRr64osvOA1WIlxcXPTll1/q7NmzevDggYoWLapx48bZHPcPIP1ieDkAAABUo0YNhYSEJLq+YMGCOnXqVMoFAoB0gtINAAAAhYaGJnqedOnhTPoBAQEpmAgA0gdKNwAAAAAAduL05E0AAAAAAEByMJHaE8TGxur8+fPy8vKSxWIxOw4AAAAAIBUwDEM3b95U3rx55eSU+P5sSvcTnD9/Xr6+vmbHAAAAAACkQmfOnFH+/PkTXU/pfgIvLy9JD5/IzJkzm5wGAAAAAJAa3LhxQ76+vtbOmBhK9xPEDSnPnDkzpRsAAAAAYONJhyEzkRoAAAAAAHZC6QYAAAAAwE4o3QAAAAAA2AmlGwAAAAAAO6F0AwAAAABgJ5RuAAAAAADshNINAAAAAICdULoBAAAAALATSjcAAAAAAHZC6QYAAAAAwE4o3QAAAAAA2AmlGwAAAAAAO6F0AwAAAABgJ5RuAAAAAADshNINAAAAAICdZDA7QHpS/oPZZkdI83Z/2dHsCAAAAADwzLCnGwAAAAAAO6F0AwAAAABgJ5RuAAAAAADshNINAAAAAICdULoBAAAAALATSjcAAAAAAHZC6QYAAAAAwE4o3QAAAAAA2AmlGwAAAAAAO6F0AwAAAABgJ5RuAAAAAADsJIPZAQB7ChseYHaENK3AkINmRwAAAADSNPZ0AwAAAABgJ5RuAAAAAADshNINAAAAAICdULoBAAAAALCTNFe6J02apEKFCsnd3V3ly5fX5s2bE902ODhYdevWVY4cOZQ5c2ZVqVJFa9asScG0AAAAAABHlqZK98KFC9WnTx998skn2rt3r15++WU1aNBAYWFhCW7/xx9/qG7dulq1apV2796tmjVrqkmTJtq7d28KJwcAAAAAOCKLYRiG2SGSqlKlSipXrpwmT55sXVaiRAk1a9ZMI0eOTNJtPP/882rdurWGDBmSpO1v3Lghb29vXb9+XZkzZ37stuU/mJ2k20Tidn/Z8ZneHqcM+284ZRgAAACQsKR2xTSzp/vevXvavXu3AgMDbZYHBgZq69atSbqN2NhY3bx5U9myZUt0m+joaN24ccPmBwAAAACA5Egzpfvy5cuKiYlRrly5bJbnypVLERERSbqNr776SlFRUWrVqlWi24wcOVLe3t7WH19f3/+UGwAAAADguNJM6Y5jsVhsLhuGEW9ZQubPn6+hQ4dq4cKFypkzZ6LbDRo0SNevX7f+nDlz5j9nBgAAAAA4pgxmB0gqHx8fOTs7x9urffHixXh7v/9t4cKF6tatmxYvXqw6deo8dls3Nze5ubn957wAAAAAAKSZPd2urq4qX7681q1bZ7N83bp1qlq1aqLXmz9/vjp37qyffvpJjRo1sndMAAAAAACs0syebknq16+fOnTooAoVKqhKlSr6/vvvFRYWpu7du0t6ODT83Llzmj374Szi8+fPV8eOHfXNN9+ocuXK1r3kGTNmlLe3t2mPAwAAAADgGNJU6W7durWuXLmi4cOHKzw8XC+88IJWrVqlggULSpLCw8Ntztk9depUPXjwQO+9957ee+896/JOnTpp5syZKR0fAAAAAOBg0lTplqQePXqoR48eCa77d5H+/fff7R8IAAAAAIBEpJljugEAAAAASGso3QAAAAAA2AmlGwAAAAAAO6F0AwAAAABgJ5RuAAAAAADshNINAAAAAICdULoBAAAAALATSjcAAAAAAHZC6QYAAAAAwE4o3QAAAAAA2AmlGwAAAAAAO6F0AwAAAABgJ5RuAAAAAADshNINAAAAAICdULoBAAAAALATSjcAAAAAAHZC6QYAAAAAwE4o3QAAAAAA2AmlGwAAAAAAO6F0AwAAAABgJ5RuAAAAAADshNINAAAAAICdULoBAAAAALATSjcAAAAAAHZC6QYAAAAAwE4o3QAAAAAA2AmlGwAAAAAAO6F0AwAAAABgJ5RuAAAAAADshNINAAAAAICdULoBAAAAALATSjcAAAAAAHZC6QYAAAAAwE4o3QAAAAAA2AmlGwAAAAAAO6F0AwAAAABgJ5RuAAAAAADshNINAAAAAICdULoBAAAAALATSjcAAAAAAHZC6QYAAAAAwE4o3QAAAAAA2AmlGwAAAAAAO6F0AwAAAABgJ5RuAAAAAADshNINAAAAAICdULoBAAAAALATSjcAAAAAAHZC6QYAAAAAwE4o3QAAAAAA2AmlGwAAAAAAO0lzpXvSpEkqVKiQ3N3dVb58eW3evDnRbcPDw9W2bVsVK1ZMTk5O6tOnT8oFBQAAAAA4vDRVuhcuXKg+ffrok08+0d69e/Xyyy+rQYMGCgsLS3D76Oho5ciRQ5988olKly6dwmkBAAAAAI4uTZXucePGqVu3bnrzzTdVokQJff311/L19dXkyZMT3N7Pz0/ffPONOnbsKG9v7xROCwAAAABwdGmmdN+7d0+7d+9WYGCgzfLAwEBt3brVpFQAAAAAACQug9kBkury5cuKiYlRrly5bJbnypVLERERz+x+oqOjFR0dbb1848aNZ3bbAAAAAADHkmb2dMexWCw2lw3DiLfsvxg5cqS8vb2tP76+vs/stgEAAAAAjiXNlG4fHx85OzvH26t98eLFeHu//4tBgwbp+vXr1p8zZ848s9sGAAAAADiWNFO6XV1dVb58ea1bt85m+bp161S1atVndj9ubm7KnDmzzQ8AAAAAAMmRZo7plqR+/fqpQ4cOqlChgqpUqaLvv/9eYWFh6t69u6SHe6nPnTun2bNnW6+zb98+SdKtW7d06dIl7du3T66uripZsqQZDwEAAAAA4EDSVOlu3bq1rly5ouHDhys8PFwvvPCCVq1apYIFC0qSwsPD452zu2zZstZ/7969Wz/99JMKFiyoU6dOpWR0AAAAAIADSlOlW5J69OihHj16JLhu5syZ8ZYZhmHnRAAAAAAAJCzNHNMNAAAAAEBaQ+kGAAAAAMBOKN0AAAAAANgJpRsAAAAAADuhdAMAAAAAYCeUbgAAAAAA7ITSDQAAAACAnVC6AQAAAACwE0o3AAAAAAB2QukGAAAAAMBOKN0AAAAAANgJpRsAAAAAADuhdAMAAAAAYCeUbgAAAAAA7ITSDQAAAACAnVC6AQAAAACwE0o3AAAAAAB2QukGAAAAAMBOMpgdAIBjqTahmtkR0rQtPbeYHQEAAABPgT3dAAAAAADYCaUbAAAAAAA7oXQDAAAAAGAnlG4AAAAAAOyE0g0AAAAAgJ1QugEAAAAAsBNKNwAAAAAAdkLpBgAAAADATijdAAAAAADYCaUbAAAAAAA7oXQDAAAAAGAnlG4AAAAAAOyE0g0AAAAAgJ1QugEAAAAAsBNKNwAAAAAAdkLpBgAAAADATijdAAAAAADYCaUbAAAAAAA7oXQDAAAAAGAnlG4AAAAAAOyE0g0AAAAAgJ1QugEAAAAAsBNKNwAAAAAAdkLpBgAAAADATijdAAAAAADYCaUbAAAAAAA7oXQDAAAAAGAnlG4AAAAAAOwkg9kBAADmCXmlutkR0rzqf4SYHQEAAKRi7OkGAAAAAMBOKN0AAAAAANgJpRsAAAAAADuhdAMAAAAAYCfJLt3Xrl3T2rVrNXfuXM2ePdvmx54mTZqkQoUKyd3dXeXLl9fmzZsfu31ISIjKly8vd3d3+fv7a8qUKXbNBwAAAABAnGTNXr58+XK1a9dOUVFR8vLyksVisa6zWCzq2LHjMwv4qIULF6pPnz6aNGmSqlWrpqlTp6pBgwY6fPiwChQoEG/7kydPqmHDhnrrrbc0d+5cbdmyRT169FCOHDnUsmVLu2QEAAAAACBOsvZ09+/fX127dtXNmzd17do1Xb161foTGRn5rDNajRs3Tt26ddObb76pEiVK6Ouvv5avr68mT56c4PZTpkxRgQIF9PXXX6tEiRJ688031bVrV40dO9ZuGQEAAAAAiJOs0n3u3Dn16tVLmTJletZ5EnXv3j3t3r1bgYGBNssDAwO1devWBK+zbdu2eNvXq1dPu3bt0v379xO8TnR0tG7cuGHzAwAAAABAciRreHlccfX393/WeRJ1+fJlxcTEKFeuXDbLc+XKpYiIiASvExERkeD2Dx480OXLl5UnT5541xk5cqSGDRuWrIy7v7TPsHokX4EhB82OgH/Z0nOL2RHwiOp/hJgdAf8ysf9ysyOkee9/1eSZ3t4X7V97prfniD6Z+/Mzvb0jX2x8prfnaEp8UuuZ3+bQoUOf+W06kmf9/C1aXPGZ3p4javX6jmd2W8kq3Y0aNdIHH3ygw4cPKyAgQC4uLjbrmzZt+kzCJeTR48clyTCMeMuetH1Cy+MMGjRI/fr1s16+ceOGfH19kxsXAAAAAODAklW633rrLUnS8OHD462zWCyKiYn5b6kS4OPjI2dn53h7tS9evBhvb3ac3LlzJ7h9hgwZlD179gSv4+bmJjc3t2cTGgCAp/Ss99ICAABzJeuY7tjY2ER/7FG4JcnV1VXly5fXunXrbJavW7dOVatWTfA6VapUibf92rVrVaFChXh75wEAAAAAeNaSfZ5uM/Tr10/Tpk3T9OnTdeTIEfXt21dhYWHq3r27pIdDwx89XVn37t11+vRp9evXT0eOHNH06dP1448/asCAAWY9BAAAAACAA0nW8HJJCgkJ0dixY3XkyBFZLBaVKFFCH3zwgV5++eVnmc9G69atdeXKFQ0fPlzh4eF64YUXtGrVKhUsWFCSFB4errCwMOv2hQoV0qpVq9S3b1999913yps3r7799lvO0Q0AAAAASBHJKt1z585Vly5d1KJFC/Xq1UuGYWjr1q2qXbu2Zs6cqbZt2z7rnFY9evRQjx49Elw3c+bMeMuqV6+uPXv22C0PAAAAAACJSVbp/uKLLzRmzBj17dvXuqx3794aN26cPvvsM7uWbgAAAAAA0opkHdP9zz//qEmT+LOrNm3aVCdPnvzPoQAAAAAASA+SVbp9fX21YcOGeMs3bNjAOa0BAAAAAPg/yRpe3r9/f/Xq1Uv79u1T1apVZbFY9Oeff2rmzJn65ptvnnVGAAAAAADSpGSV7nfffVe5c+fWV199pUWLFkmSSpQooYULF+rVV199pgEBAAAAAEirkn3KsObNm6t58+bPMgsAAAAAAOlKsks3AACAI/hk7s9mRwAApGFJLt3ZsmXTsWPH5OPjo6xZs8pisSS6bWRk5DMJBwAAAABAWpbk0j1+/Hh5eXlZ//240g0AAAAAAJ6idHfq1Mn6786dO9sjCwAAAAAA6UqyztPt7Oysixcvxlt+5coVOTs7/+dQAAAAAACkB8kq3YZhJLg8Ojparq6u/ykQAAAAAADpxVPNXv7tt99KkiwWi6ZNmyZPT0/rupiYGP3xxx8qXrz4s00IAAAAAEAa9VSle/z48ZIe7umeMmWKzVByV1dX+fn5acqUKc82IQAAAAAAadRTle6TJ09KkmrWrKng4GBlzZrVLqEAAAAAAEgPnqp0x9m0adOzzgEAAAAAQLqTrNItSWfPntWyZcsUFhame/fu2awbN27cfw4GAAAAAEBal6zSvWHDBjVt2lSFChVSaGioXnjhBZ06dUqGYahcuXLPOiMAAAAAAGlSsk4ZNmjQIPXv31+HDh2Su7u7fvnlF505c0bVq1fX66+//qwzAgAAAACQJiWrdB85ckSdOnWSJGXIkEF37tyRp6enhg8frtGjRz/TgAAAAAAApFXJKt0eHh6Kjo6WJOXNm1cnTpywrrt8+fKzSQYAAAAAQBqXrGO6K1eurC1btqhkyZJq1KiR+vfvr4MHDyo4OFiVK1d+1hkBAAAAAEiTklW6x40bp1u3bkmShg4dqlu3bmnhwoUqUqSIxo8f/0wDAgAAAACQViWrdPv7+1v/nSlTJk2aNOmZBQIAAAAAIL1I1jHdAAAAAADgyZK8pztr1qyyWCxJ2jYyMjLZgQAAAAAASC+SXLq//vprO8YAAAAAACD9SXLpjjsvNwAAAAAASJpkH9N94sQJffrpp2rTpo0uXrwoSVq9erX+/vvvZxYOAAAAAIC0LFmlOyQkRAEBAfrrr78UHBxsPX3YgQMHFBQU9EwDAgAAAACQViWrdH/00Uf6/PPPtW7dOrm6ulqX16xZU9u2bXtm4QAAAAAASMuSVboPHjyo5s2bx1ueI0cOXbly5T+HAgAAAAAgPUhW6c6SJYvCw8PjLd+7d6/y5cv3n0MBAAAAAJAeJKt0t23bVgMHDlRERIQsFotiY2O1ZcsWDRgwQB07dnzWGQEAAAAASJOSVbq/+OILFShQQPny5dOtW7dUsmRJvfLKK6patao+/fTTZ50RAAAAAIA0Kcnn6Y5jGIbOnz+vH374QZ999pn27Nmj2NhYlS1bVkWLFrVHRgAAAAAA0qRkle6iRYvq77//VtGiReXv72+PXAAAAAAApHlPPbzcyclJRYsWZZZyAAAAAACeIFnHdI8ZM0YffPCBDh069KzzAAAAAACQbjz18HJJat++vW7fvq3SpUvL1dVVGTNmtFkfGRn5TMIBAAAAAJCWJat0f/311884BgAAAAAA6c9Tl+779+/r999/1+DBg5lEDQAAAACAx3jqY7pdXFy0ZMkSe2QBAAAAACBdSdZEas2bN9fSpUufcRQAAAAAANKXZB3TXaRIEX322WfaunWrypcvLw8PD5v1vXr1eibhAAAAAABIy5JVuqdNm6YsWbJo9+7d2r17t806i8VC6QYAAAAAQMks3SdPnnzWOQAAAAAASHeSdUz3owzDkGEYzyILAAAAAADpSrJL9+zZsxUQEKCMGTMqY8aMKlWqlObMmfMsswEAAAAAkKYla3j5uHHjNHjwYL3//vuqVq2aDMPQli1b1L17d12+fFl9+/Z91jkBAAAAAEhzklW6J0yYoMmTJ6tjx47WZa+++qqef/55DR06lNINAAAAAICSObw8PDxcVatWjbe8atWqCg8P/8+hEnL16lV16NBB3t7e8vb2VocOHXTt2rXHXic4OFj16tWTj4+PLBaL9u3bZ5dsAAAAAAAkJFmlu0iRIlq0aFG85QsXLlSRIkX+c6iEtG3bVvv27dPq1au1evVq7du3Tx06dHjsdaKiolStWjWNGjXKLpkAAAAAAHicZA0vHzZsmFq3bq0//vhD1apVk8Vi0Z9//qn169dr8eLFzzqjjhw5otWrV2v79u2qVKmSJOmHH35QlSpVFBoaqmLFiiV4vbhSfurUqWeeCQAAAACAJ3mqPd1jx46VJLVs2VJ//fWXfHx8tHTpUgUHB8vHx0cbN27U6NGjn3nIbdu2ydvb21q4Jaly5cry9vbW1q1bn/n9AQAAAADwLDzVnu7Bgwcre/bs6tKli8qXL6+5c+da1928eVP16tXTjRs3nnnIiIgI5cyZM97ynDlzKiIi4pneV3R0tKKjo62X7fF4AAAAkHwlPqlldgQASLKn2tM9Z84c9ejRQ0uXLrVZHhUVpfr16ysyMlIbN25M8u0NHTpUFovlsT+7du2SJFkslnjXNwwjweX/xciRI62TtXl7e8vX1/eZ3j4AAAAAwHE81Z7u1157TdeuXVPbtm21cuVK1axZU7du3VL9+vV16dIl/f7778qdO3eSb+/999/XG2+88dht/Pz8dODAAV24cCHeukuXLilXrlxP8xCeaNCgQerXr5/18o0bNyjeAAAAAIBkeeqJ1N58801FRkaqWbNm+vXXXzV48GBFREQoJCREefPmfarb8vHxkY+PzxO3q1Kliq5fv64dO3aoYsWKkqS//vpL169fT/DUZf+Fm5ub3NzcnultAgAAAOnZ0KFDzY4ApFrJmr38ww8/1NWrV1W7dm35+fkpJCRE+fLle9bZrEqUKKH69evrrbfe0tSpUyVJb7/9tho3bmwzc3nx4sU1cuRINW/eXJIUGRmpsLAwnT9/XpIUGhoqScqdO/dT7ZEHAAAAACA5nqp0t2jRwuayi4uLfHx81KtXL5vlwcHB/z3Zv8ybN0+9evVSYGCgJKlp06aaOHGizTahoaG6fv269fKyZcvUpUsX6+W4oexBQUF8GwcAAAAAsLunKt3e3t42l9u0afNMwzxOtmzZbGZLT4hhGDaXO3furM6dO9sxFQAAAAAAiXuq0j1jxgx75QAAAAAAIN15qlOGAQAAAACApKN0AwAAAABgJ5RuAAAAAADshNINAAAAAICdULoBAAAAALATSjcAAAAAAHZC6QYAAAAAwE4o3QAAAAAA2AmlGwAAAAAAO6F0AwAAAABgJ5RuAAAAAADshNINAAAAAICdULoBAAAAALATSjcAAAAAAHZC6QYAAAAAwE4o3QAAAAAA2AmlGwAAAAAAO6F0AwAAAABgJ5RuAAAAAADshNINAAAAAICdULoBAAAAALATSjcAAAAAAHZC6QYAAAAAwE4o3QAAAAAA2AmlGwAAAAAAO6F0AwAAAABgJ5RuAAAAAADshNINAAAAAICdULoBAAAAALATSjcAAAAAAHZC6QYAAAAAwE4o3QAAAAAA2AmlGwAAAAAAO6F0AwAAAABgJ5RuAAAAAADshNINAAAAAICdULoBAAAAALATSjcAAAAAAHZC6QYAAAAAwE4o3QAAAAAA2AmlGwAAAAAAO6F0AwAAAABgJ5RuAAAAAADshNINAAAAAICdULoBAAAAALATSjcAAAAAAHZC6QYAAAAAwE4o3QAAAAAA2AmlGwAAAAAAO6F0AwAAAABgJ5RuAAAAAADsJM2U7qtXr6pDhw7y9vaWt7e3OnTooGvXriW6/f379zVw4EAFBATIw8NDefPmVceOHXX+/PmUCw0AAAAAcGhppnS3bdtW+/bt0+rVq7V69Wrt27dPHTp0SHT727dva8+ePRo8eLD27Nmj4OBgHTt2TE2bNk3B1AAAAAAAR5bB7ABJceTIEa1evVrbt29XpUqVJEk//PCDqlSpotDQUBUrVizedby9vbVu3TqbZRMmTFDFihUVFhamAgUKpEh2AAAAAIDjShN7urdt2yZvb29r4ZakypUry9vbW1u3bk3y7Vy/fl0Wi0VZsmSxQ0oAAAAAAGyliT3dERERypkzZ7zlOXPmVERERJJu4+7du/roo4/Utm1bZc6cOdHtoqOjFR0dbb1848aNpw8MAAAAAIBM3tM9dOhQWSyWx/7s2rVLkmSxWOJd3zCMBJf/2/379/XGG28oNjZWkyZNeuy2I0eOtE7W5u3tLV9f3+Q9OAAAAACAwzN1T/f777+vN95447Hb+Pn56cCBA7pw4UK8dZcuXVKuXLkee/379++rVatWOnnypDZu3PjYvdySNGjQIPXr1896+caNGxRvAAAAAECymFq6fXx85OPj88TtqlSpouvXr2vHjh2qWLGiJOmvv/7S9evXVbVq1USvF1e4jx8/rk2bNil79uxPvC83Nze5ubkl/UEAAAAAAJCINDGRWokSJVS/fn299dZb2r59u7Zv36633npLjRs3tpm5vHjx4lqyZIkk6cGDB3rttde0a9cuzZs3TzExMYqIiFBERITu3btn1kMBAAAAADiQNFG6JWnevHkKCAhQYGCgAgMDVapUKc2ZM8dmm9DQUF2/fl2SdPbsWS1btkxnz55VmTJllCdPHuvP08x4DgAAAABAcqWJ2cslKVu2bJo7d+5jtzEMw/pvPz8/m8sAAAAAAKS0NLOnGwAAAACAtIbSDQAAAACAnVC6AQAAAACwE0o3AAAAAAB2QukGAAAAAMBOKN0AAAAAANgJpRsAAAAAADuhdAMAAAAAYCeUbgAAAAAA7ITSDQAAAACAnVC6AQAAAACwE0o3AAAAAAB2QukGAAAAAMBOKN0AAAAAANgJpRsAAAAAADuhdAMAAAAAYCeUbgAAAAAA7ITSDQAAAACAnVC6AQAAAACwE0o3AAAAAAB2QukGAAAAAMBOKN0AAAAAANgJpRsAAAAAADuhdAMAAAAAYCeUbgAAAAAA7ITSDQAAAACAnVC6AQAAAACwE0o3AAAAAAB2QukGAAAAAMBOKN0AAAAAANgJpRsAAAAAADuhdAMAAAAAYCeUbgAAAAAA7ITSDQAAAACAnVC6AQAAAACwE0o3AAAAAAB2QukGAAAAAMBOKN0AAAAAANgJpRsAAAAAADuhdAMAAAAAYCeUbgAAAAAA7ITSDQAAAACAnVC6AQAAAACwE0o3AAAAAAB2QukGAAAAAMBOKN0AAAAAANgJpRsAAAAAADuhdAMAAAAAYCeUbgAAAAAA7ITSDQAAAACAnVC6AQAAAACwE0o3AAAAAAB2kmZK99WrV9WhQwd5e3vL29tbHTp00LVr1x57naFDh6p48eLy8PBQ1qxZVadOHf31118pExgAAAAA4PDSTOlu27at9u3bp9WrV2v16tXat2+fOnTo8NjrPPfcc5o4caIOHjyoP//8U35+fgoMDNSlS5dSKDUAAAAAwJFZDMMwzA7xJEeOHFHJkiW1fft2VapUSZK0fft2ValSRUePHlWxYsWSdDs3btyQt7e31q9fr9q1az/Vda5fv67MmTMn+zEAAAAAQEpYtLii2RHSvFav73jiNkntimliT/e2bdvk7e1tLdySVLlyZXl7e2vr1q1Juo179+7p+++/l7e3t0qXLp3odtHR0bpx44bNDwAAAAAAyZEmSndERIRy5swZb3nOnDkVERHx2OuuWLFCnp6ecnd31/jx47Vu3Tr5+Pgkuv3IkSOtx417e3vL19f3P+cHAAAAADgmU0v30KFDZbFYHvuza9cuSZLFYol3fcMwElz+qJo1a2rfvn3aunWr6tevr1atWunixYuJbj9o0CBdv37d+nPmzJn/9iABAAAAAA4rg5l3/v777+uNN9547DZ+fn46cOCALly4EG/dpUuXlCtXrsde38PDQ0WKFFGRIkVUuXJlFS1aVD/++KMGDRqU4PZubm5yc3NL+oMAAAAAACARppZuHx+fxw71jlOlShVdv35dO3bsUMWKDycF+Ouvv3T9+nVVrVr1qe7TMAxFR0cnKy8AAAAAAE8jTRzTXaJECdWvX19vvfWWtm/fru3bt+utt95S48aNbWYuL168uJYsWSJJioqK0scff6zt27fr9OnT2rNnj958802dPXtWr7/+ulkPBQAAAADgQNJE6ZakefPmKSAgQIGBgQoMDFSpUqU0Z84cm21CQ0N1/fp1SZKzs7OOHj2qli1b6rnnnlPjxo116dIlbd68Wc8//7wZDwEAAAAA4GBMHV7+NLJly6a5c+c+dptHTznu7u6u4OBge8cCAAAAACBRaWZPNwAAAAAAaQ2lGwAAAAAAO6F0AwAAAABgJ5RuAAAAAADshNINAAAAAICdULoBAAAAALATSjcAAAAAAHZC6QYAAAAAwE4o3QAAAAAA2AmlGwAAAAAAO6F0AwAAAABgJxnMDgAAAAAAeHZavb7D7Ah4BHu6AQAAAACwE0o3AAAAAAB2QukGAAAAAMBOKN0AAAAAANgJpRsAAAAAADuhdAMAAAAAYCeUbgAAAAAA7ITSDQAAAACAnVC6AQAAAACwE0o3AAAAAAB2QukGAAAAAMBOKN0AAAAAANgJpRsAAAAAADuhdAMAAAAAYCeUbgAAAAAA7ITSDQAAAACAnWQwO0BqZxiGJOnGjRsmJwEAAAAApBZxHTGuMyaG0v0EN2/elCT5+vqanAQAAAAAkNrcvHlT3t7eia63GE+q5Q4uNjZW58+fl5eXlywWi9lxku3GjRvy9fXVmTNnlDlzZrPjQLwmqRGvSerDa5L68JqkLrweqQ+vSerDa5L6pJfXxDAM3bx5U3nz5pWTU+JHbrOn+wmcnJyUP39+s2M8M5kzZ07Tb+z0iNck9eE1SX14TVIfXpPUhdcj9eE1SX14TVKf9PCaPG4PdxwmUgMAAAAAwE4o3QAAAAAA2Aml20G4ubkpKChIbm5uZkfB/+E1SX14TVIfXpPUh9ckdeH1SH14TVIfXpPUx9FeEyZSAwAAAADATtjTDQAAAACAnVC6AQAAAACwE0o3kAIMw9Dp06d1584ds6MAAAAASEGUbiAFGIahokWL6uzZs2ZHAQCkUXyBCyA9uXbtmtkRUgylG0gBTk5OKlq0qK5cuWJ2FABAGsUXuKnP/fv3VbNmTR07dszsKECqNnr0aC1cuNB6uVWrVsqePbvy5cun/fv3m5gsZWQwOwDs686dOzIMQ5kyZZIknT59WkuWLFHJkiUVGBhocjrHMmbMGH3wwQeaPHmyXnjhBbPjAECSNG/eXBaLJd5yi8Uid3d3FSlSRG3btlWxYsVMSOdYHv0Ct2jRombHgSQXFxcdOnQowf9HYL5r165px44dunjxomJjY23WdezY0aRUjmnq1KmaO3euJGndunVat26dfvvtNy1atEgffPCB1q5da3JC++KUYelcYGCgWrRooe7du+vatWsqXry4XFxcdPnyZY0bN07vvvuu2REdRtasWXX79m09ePBArq6uypgxo836yMhIk5I5tj179sjFxUUBAQGSpF9//VUzZsxQyZIlNXToULm6upqc0PGULVv2iSWvc+fOqlmzpgnpHE/nzp21dOlSZcmSReXLl5dhGNq7d6+uXbumwMBA7d+/X6dOndKGDRtUrVo1s+OmeytXrtSoUaP4AjcV6d+/v1xcXDRq1Cizo+ARy5cvV7t27RQVFSUvLy+b3ysWi4W/u1JYxowZdezYMfn6+qp37966e/eupk6dqmPHjqlSpUq6evWq2RHtij3d6dyePXs0fvx4SdLPP/+sXLlyae/evfrll180ZMgQSncK+vrrr82OgAS88847+uijjxQQEKB//vlHb7zxhpo3b67Fixfr9u3bvG4mqF+/viZPnqyAgABVrFhRhmFo165dOnDggDp37qzDhw+rTp06Cg4O1quvvmp23HQvd+7catu2rSZOnCgnp4dHpcXGxqp3797y8vLSggUL1L17dw0cOFB//vmnyWnTv/bt2+v27dsqXbo0X+CmEvfu3dO0adO0bt06VahQQR4eHjbrx40bZ1Iyx9a/f3917dpVI0aMsI74hHmyZs2qM2fOyNfXV6tXr9bnn38u6eFhMzExMSansz/2dKdzmTJl0tGjR1WgQAG1atVKzz//vIKCgnTmzBkVK1ZMt2/fNjsiYCpvb2/t2bNHhQsX1ujRo7Vx40atWbNGW7Zs0RtvvKEzZ86YHdHhvPXWWypQoIAGDx5ss/zzzz/X6dOn9cMPPygoKEgrV67Url27TErpOHLkyKEtW7boueees1l+7NgxVa1aVZcvX9bBgwf18ssvO9SkOGaZNWvWY9d36tQphZIgzuNG3VgsFm3cuDEF0yCOh4eHDh48KH9/f7OjQNL777+vFStWqGjRotq7d69OnTolT09PLVy4UKNHj9aePXvMjmhX7OlO54oUKaKlS5eqefPmWrNmjfr27StJunjxojJnzmxyOscTGxur//3vfwkeW/TKK6+YlMqxGYZhfS3Wr1+vxo0bS5J8fX11+fJlM6M5rEWLFmn37t3xlr/xxhsqX768fvjhB7Vp04a9RynkwYMHOnr0aLzSffToUeveCXd3d45pTSGU6tRn06ZNZkdAAurVq6ddu3ZRulOJ8ePHy8/PT2fOnNGYMWPk6ekpSQoPD1ePHj1MTmd/lO50bsiQIWrbtq369u2r2rVrq0qVKpKktWvXqmzZsiancyzbt29X27Ztdfr0af17gInFYnGIoTWpUYUKFfT555+rTp06CgkJ0eTJkyVJJ0+eVK5cuUxO55jc3d21detWFSlSxGb51q1b5e7uLunhF1hubm5mxHM4HTp0ULdu3fTxxx/rxRdflMVi0Y4dOzRixAjrREQhISF6/vnnTU7qOE6cOKEZM2boxIkT+uabb5QzZ06tXr1avr6+vA4m+t///qcTJ07olVdeUcaMGWUYBl9GmahRo0b64IMPdPjwYQUEBMjFxcVmfdOmTU1K5phcXFw0YMCAeMv79OmT8mFMwPByBxAREaHw8HCVLl3aejzejh07lDlzZhUvXtzkdI6jTJkyeu655zRs2DDlyZMn3i9ib29vk5I5tgMHDqhdu3YKCwtTv379FBQUJEnq2bOnrly5op9++snkhI7n888/14gRI/TWW2/ZlLxp06bp448/1ieffKLx48dr1apVWrdundlx072YmBiNGjVKEydO1IULFyRJuXLlUs+ePTVw4EA5OzsrLCxMTk5Oyp8/v8lp07+QkBA1aNBA1apV0x9//KEjR47I399fY8aM0Y4dO/Tzzz+bHdHhXLlyRa1atdKmTZtksVh0/Phx+fv7q1u3bsqSJYu++uorsyM6pLi/eRPCzg5zHDt2TL///nuCIz6HDBliUqqUQekGUoiHh4f2798fb+8dUqe7d+/K2dk53jfjSBnz5s3TxIkTFRoaKkkqVqyYevbsqbZt20p6eDrEuNnMkXJu3LghSRyeZKIqVaro9ddfV79+/eTl5aX9+/fL399fO3fuVLNmzXTu3DmzIzqcjh076uLFi5o2bZpKlChhfU3Wrl2rvn376u+//zY7ImC6H374Qe+++658fHyUO3fueLPJp/djuind6VxUVJRGjRqlDRs2JPit0j///GNSMsdTq1Ytffjhh6pfv77ZUZCA3bt368iRI7JYLCpRooTKlStndiQAiMfT01MHDx5UoUKFbEr3qVOnVLx4cd29e9fsiA4nd+7cWrNmjUqXLm3zmpw8eVIBAQG6deuW2REd3t27d/mS1mQFCxZUjx49NHDgQLOjmIJjutO5N998UyEhIerQoUOCQ5phXwcOHLD+u2fPnurfv78iIiISPLaoVKlSKR0PejipYOvWrRUSEqIsWbLIMAxdv35dNWvW1IIFC5QjRw6zIwKm+/nnn7Vo0SKFhYXp3r17NuvS+96J1CZLliwKDw9XoUKFbJbv3btX+fLlMymVY4uKikrwlFSXL19m7gkTxcTEaMSIEZoyZYouXLigY8eOyd/fX4MHD5afn5+6detmdkSHcvXqVb3++utmxzANpTud++2337Ry5UpVq1bN7CgOqUyZMrJYLDYTp3Xt2tX677h1HFtknp49e+rmzZv6+++/VaJECUnS4cOH1alTJ/Xq1Uvz5883OaFjyJYtm44dOyYfHx9lzZr1sV8Qch7ilPXtt9/qk08+UadOnfTrr7+qS5cuOnHihHbu3Kn33nvP7HgOp23btho4cKAWL14si8Wi2NhYbdmyRQMGDLBObIeU9corr2j27Nn67LPPJMn6unz55ZePPZ0Y7OuLL77QrFmzNGbMGL311lvW5QEBARo/fjylO4W9/vrrWrt2rbp37252FFMwvDydK1SokFatWmUtE0hZp0+fTvK2BQsWtGMSJMbb21vr16/Xiy++aLN8x44dCgwM5LzDKWTWrFl644035ObmxnmIU5nixYsrKChIbdq0sRk6O2TIEEVGRmrixIlmR3Qo9+/fV+fOnbVgwQIZhqEMGTIoJiZGbdu21cyZM+Xs7Gx2RIdz+PBh1ahRQ+XLl9fGjRvVtGlT/f3334qMjNSWLVtUuHBhsyM6pCJFimjq1KmqXbu2zWfX0aNHVaVKFV29etXsiA5l5MiRGjdunBo1apTgiM9evXqZlCxlULrTublz5+rXX3/VrFmzEhz6BDg6Ly8vbd68WWXKlLFZvnfvXlWvXt06cRTgqDJlyqQjR46oYMGCypkzp9atW6fSpUvr+PHjqly5sq5cuWJ2RIf0zz//aM+ePYqNjVXZsmVVtGhRsyM5tIiICE2ePFm7d+9WbGysypUrp/fee0958uQxO5rDypgxo44ePaqCBQvalO7Dhw+rYsWKHGufwv59SMyjLBZLup9niuHl6dxXX32lEydOKFeuXPLz84v3rRLH4qW8w4cPJ3hcJOeLNEetWrXUu3dvzZ8/X3nz5pUknTt3znpue6SMp/lyg5mzU1bu3Ll15coVFSxYUAULFtT27dtVunRpnTx5Unxvn/KGDx+uAQMGyN/fX/7+/tbld+7c0ZdffpnuT7uTGoWFhcnX11fDhg1LcF2BAgVMSIXnn39emzdvjjeScPHixSpbtqxJqRzXyZMnzY5gKkp3OtesWTOzI+D//PPPP2revLkOHjxoc5x33LGrHNNtjokTJ+rVV1+Vn5+ffH19ZbFYFBYWpoCAAM2dO9fseA4jS5YsSZ7okf9XUlatWrW0fPlylStXTt26dVPfvn31888/a9euXWrRooXZ8RzOsGHD1L1793ij127fvq1hw4ZRuk1QqFAhhYeHK2fOnDbLr1y5okKFCvGZZZKgoCB16NBB586dU2xsrIKDgxUaGqrZs2drxYoVZseDg2F4OZBCmjRpImdnZ/3www/y9/fXjh07dOXKFfXv319jx47Vyy+/bHZEh7Zu3TodPXpUhmGoZMmSqlOnjtmRHEpISIj136dOndJHH32kzp07q0qVKpKkbdu2adasWRo5ciTHdKew2NhYxcbGKkOGh9/TL1q0SH/++aeKFCmi7t27y9XV1eSEjsXJyUkXLlyId2aFjRs3qnXr1rp06ZJJyRxXYq/J6dOnVbJkSUVFRZmUDGvWrNGIESNshv0PGTJEgYGBZkdzCP369dNnn30mDw8P9evX77Hbjhs3LoVSmYPSDaQQHx8fbdy4UaVKlZK3t7d27NihYsWKaePGjerfv7/27t1rdkQgVahdu7befPNNtWnTxmb5Tz/9pO+//16///67OcEc0IMHD/TFF1+oa9eu8vX1NTuOQ4ub1f/69evKnDmzzciQmJgY3bp1S927d9d3331nYkrHElcivvnmG7311ls2ow9iYmL0119/ydnZWVu2bDErImCqmjVrasmSJcqSJcsTZ/LftGlTCqUyB6U7HeLUO6lT1qxZtXv3bvn7+6tw4cKaNm2aatasqRMnTiggIEC3b982O6LD+Pbbb5O8bXqfTTM1ypQpk/bv3x9vYqhjx46pTJky/L+Swjw9PXXo0CH5+fmZHcWhzZo1S4ZhqGvXrvr666/l7e1tXefq6io/Pz/ryBCkjLgSERISoipVqtiM+oh7TQYMGMAkdybbtWuXjhw5IovFohIlSqh8+fJmR4ID4pjudGj8+PHy8vKy/jupx0nCvl544QUdOHBA/v7+qlSpksaMGSNXV1d9//33NpPhwP7Gjx+fpO0sFgul2wS+vr6aMmWKvvrqK5vlU6dOZW+rCerUqaPff/9dnTt3NjuKQ4s7rKJQoUKqWrVqvIlRkfLi9sx16dJF33zzDZM8pjJnz55VmzZttGXLFmXJkkWSdO3aNVWtWlXz58/n90kKmz17tl588cV4pzG+e/euFi1apI4dO5qULGWwpxtIIWvWrFFUVJRatGihEydOqEmTJjp69KiyZ8+uBQsWMFM28H9WrVqlli1bqnDhwqpcubIkafv27Tpx4oR++eUXNWzY0OSEjmXq1KkaOnSo2rVrp/Lly8vDw8NmPWdesD9m9weeXmBgoG7cuKFZs2apWLFikqTQ0FB17dpVHh4eWrt2rckJHYuTk5M8PDw0c+ZMtWzZ0rr8woULyps3b7qfcJDSnc7VqlVL1atXV1BQkM3yq1evqmXLltq4caNJySA9HN7/pEMAkLJiYmJ08OBBFSxYUFmzZjU7jsM6e/asJk2aZDO5Xffu3dkzYQInJ6dE11kslnT/h1Jq4OTk9MTfE4Zh8HqkoKeZuT84ONiOSZCYjBkzauvWrfFOD7Znzx5Vq1ZNd+7cMSmZY3JyctLYsWP16aef6sMPP9TQoUMlOU7pZnh5Ovf777/r4MGD2rt3r+bNm2fdQ3Hv3j2b2YJhP0n5xZwhQwblzp1bdevWVZMmTVIgFeL06dNHAQEB6tatm2JiYvTKK69o27ZtypQpk1asWKEaNWqYHdEh5c+fXyNGjDA7BvRw9nKYK6kTDDEhZ8p59Jh6pE4FChTQ/fv34y1/8OCB8uXLZ0IitG/fXlWrVlXz5s116NAhzZkzx+xIKYY93emck5OT9u7dq3feeUdRUVFavny5/Pz8HOZbpdSgS5cuT9wmNjZWFy9eVEhIiAYMGKDhw4enQDJID8vd0qVLVaFCBS1dulTvvfeeNm3apNmzZ2vTpk3MOmuSa9euaceOHbp48WK80pfej/tKbWbPnq3WrVvLzc3NZvm9e/e0YMECXg+TXb9+XfPmzdO0adO0f/9+fq8D/+fXX3/ViBEj9N1336l8+fKyWCzatWuXevbsqYEDB6pZs2ZmR3Qozs7O1vPZh4WFqWnTprJYLJoyZYqqVq2a7j+7KN3pnJOTkyIiIuTt7a2uXbtq7dq1Wrx4sUqUKEHpToVWrlypd999V2FhYWZHcRju7u763//+p/z58+vtt99WpkyZ9PXXX+vkyZMqXbr0Ux1LiWdj+fLlateunaKiouTl5WUzrNZisXDWhRT26B9Kj7py5Ypy5szJ7xGTbNy4UdOnT1dwcLAKFiyoli1bqmXLlvGG0gKO5N+H7EVFRenBgwfKkOHh4N64f3t4ePC7JIXFdZK43yW3b99Wu3bttGHDBkVFRaX73yUML0/n4j543NzcNG/ePH3++eeqX7++Bg4caHIyJKRatWqqUKGC2TEcSq5cuXT48GHlyZNHq1ev1qRJkyQ9/GXg7OxscjrH1L9/f3Xt2lUjRoywOe8tzBF3rPC/nT17liG2Kezs2bOaOXOmpk+frqioKLVq1Ur379/XL7/8opIlS5odz6H9/PPPWrRokcLCwnTv3j2bdXv27DEpleP5+uuvzY6ARAQFBcnT09N6OVOmTFqyZImCgoL0xx9/mJgsZVC607l/D2T49NNPVaJECeupR5C6ZMmShQlXUliXLl3UqlUr5cmTRxaLRXXr1pUk/fXXXypevLjJ6RzTuXPn1KtXLwq3ycqWLSuLxSKLxaLatWtb9xRJDyccPHnypOrXr29iQsfSsGFD/fnnn2rcuLEmTJig+vXry9nZWVOmTDE7msP79ttv9cknn6hTp0769ddf1aVLF504cUI7d+7Ue++9Z3Y8h8Lft6lLuXLltGHDhsdOGjxs2LAUTmUOSnc6d/LkSeXIkcNmWcuWLVW8eHHt2rXLpFRA6jF06FC98MILOnPmjF5//XXrcavOzs766KOPTE7nmOrVq6ddu3Zx/nqTxR3vuG/fPtWrV89mD4Wrq6v8/PxsTvsC+1q7dq169eqld999V0WLFjU7Dh4xadIkff/992rTpo1mzZqlDz/8UP7+/hoyZAhDmFOJO3fuxJtUjdPr2d+RI0cUFRWlrFmzatiwYXr33Xcd9gt1Snc6N2zYMH3zzTfy8vKyWe7n56evvvqKbwTh8BKbJKpNmzZasGCBSakcW6NGjfTBBx/o8OHDCggIkIuLi816zgudMuJONenn56fWrVvL3d3d5ESObfPmzZo+fboqVKig4sWLq0OHDmrdurXZsSApLCxMVatWlfTwNFU3b96UJHXo0EGVK1fWxIkTzYznsKKiojRw4EAtWrRIV65cibc+vR9DnBqUKVNGXbp00UsvvSTDMPTll1/afIH7qCFDhqRwupTFRGrpXGIT4Fy+fFm5c+fWgwcPTEoGpA5MEpX6cF5oIHG3b9/WggULNH36dO3YsUMxMTEaN26cunbtGu8LdqQMf39//fzzzypXrpxefPFFvfnmm3rnnXe0du1avfHGG+ztNknc2UiGDx+ujh076rvvvtO5c+c0depUjRo1Su3atTM7YroXGhqqoKAgnThxQnv27FHJkiVtDlWKY7FY0v3cB5TudOrGjRsyDENZs2bV8ePHbYaYx8TEaPny5froo490/vx5E1MC5nNyctKFCxfiHYaxf/9+1axZkz+W4JCyZcumY8eOycfH57HH4kni/xEThYaG6scff9ScOXN07do11a1bV8uWLTM7lsN588035evrq6CgIE2ZMkX9+vVTtWrVtGvXLrVo0UI//vij2REdUoECBTR79mzVqFFDmTNn1p49e1SkSBHNmTNH8+fP16pVq8yO6FD+PXu5o2F4eTqVJUsW6wQ4zz33XLz1FovFYSYuABLCJFFpw927dxnWbILx48db95oyG3DqVaxYMY0ZM0YjR47U8uXLNX36dLMjOaTvv/9esbGxkqTu3bsrW7Zs+vPPP9WkSRN1797d5HSOKzIyUoUKFZL08PjtuC8IX3rpJb377rtmRnNIcf+POCr2dKdTISEhMgxDtWrV0i+//KJs2bJZ17m6uqpgwYLKmzeviQkBc8V96TRs2DD1798/0UmiXF1dzYrosGJiYjRixAhNmTJFFy5c0LFjx+Tv76/BgwfLz89P3bp1MzsiACCVK1WqlCZMmKDq1asrMDBQpUqV0tixY/Xtt99qzJgxOnv2rNkR071ly5apQYMGcnFxeeIonPQ+XwulO50LDQ1V0aJFEzxG8vLly/Lx8TEhFZB6zJo1i0miUpnhw4dr1qxZGj58uN566y0dOnRI/v7+WrRokcaPH69t27aZHdHhnDhxQjNmzNCJEyf0zTffKGfOnFq9erV8fX31/PPPmx0PMNXq1avl6empl156SZL03Xff6YcfflDJkiX13XffKWvWrCYndEzjx4+Xs7OzevXqpU2bNqlRo0aKiYnRgwcPNG7cOPXu3dvsiOneo0PKHX2+Fkp3OtesWTMFBwfHe6NfuHBBtWvX1qFDh0xKBgAJK1KkiKZOnaratWvLy8tL+/fvl7+/v44ePaoqVaro6tWrZkd0KCEhIWrQoIGqVaumP/74Q0eOHJG/v7/GjBmjHTt26OeffzY7ImCqgIAAjR49Wg0bNtTBgwdVoUIF9e/fXxs3blSJEiU0Y8YMsyNCD2eZ37VrlwoXLqzSpUubHQcOJvGvHJAuhIeHxxuKGR4erho1aqh48eImpQLMlS1bNl2+fFmSlDVrVmXLli3RH6S8c+fOqUiRIvGWx8bGxjvPKuzvo48+0ueff65169bZHG5Rs2ZNRh0Akk6ePKmSJUtKkn755Rc1adJEI0aM0KRJk/Tbb7+ZnA5xChQooBYtWlC4TXD//n3VrFlTx44dMzuKaZhILZ1btWqVXnnlFfXt21fjx4/XuXPnVKtWLZUuXZpzEMNhMUlU6vb8889r8+bNKliwoM3yxYsXq2zZsialclwHDx7UTz/9FG95jhw5Ejz3LeBoXF1ddfv2bUnS+vXr1bFjR0kPv+C9ceOGmdEc3oYNG7RhwwZdvHgx3kReTDyYclxcXHTo0KHHngkjvaN0p3PZs2fXmjVrrMcZrVy5UuXKldO8efMee2wFkJ516tQpwX8jdQgKClKHDh107tw5xcbGKjg4WKGhoZo9e7ZWrFhhdjyHkyVLFoWHh1tnAY6zd+9e5cuXz6RUQOrx0ksvWU8TtmPHDi1cuFCSdOzYMeXPn9/kdI5r2LBhGj58uCpUqKA8efI4dOFLDTp27Kgff/xRo0aNMjuKKTim20EcP35cL730kurWras5c+bwwQM8gkmiUp81a9ZoxIgR2r17t2JjY1WuXDkNGTJEgYGBZkdzOB9++KG2bdumxYsX67nnntOePXt04cIFdezYUR07dlRQUJDZEQFThYWFqUePHjpz5ox69eplPayvb9++iomJ0bfffmtyQseUJ08ejRkzRh06dDA7CiT17NlTs2fPVpEiRVShQgV5eHjYrB83bpxJyVIGpTsdypo1a4Kl+vbt23Jzc5Ozs7N1Wdw5CwFHxSRRwOPdv39fnTt31oIFC2QYhjJkyKCYmBi1bdtWM2fOtPmdAgCpRfbs2bVjxw4VLlzY7CjQw3lAEmOxWLRx48YUTJPyKN3p0KxZs5K8LUNr4eiqVKmi119/Xf369bOZKXvnzp1q1qyZzp07Z3ZEIFU4ceKE9u7dq9jYWJUtW1ZFixY1OxKQKiR23LbFYpGbm5vNBIRIOQMHDpSnp6cGDx5sdhSA0p2ePXjwQPPmzVO9evWUO3dus+MAqZKnp6cOHjyoQoUK2ZTuU6dOqXjx4rp7967ZER1OTEyMxo8fr0WLFiksLEz37t2zWc8IHfPE/cnAIUrA/+fk5PTY/yfy58+vzp07KygoiPl07Kxfv37Wf8fGxmrWrFkqVaqUSpUqJRcXF5tt0/twZqQuTKSWjmXIkEHvvvuujhw5YnYUINVikqjUZ9iwYZo2bZr69eunwYMH65NPPtGpU6e0dOlSDRkyxOx4DunHH3/U+PHjdfz4cUlS0aJF1adPH7355psmJwPMN3PmTH3yySfq3LmzKlasKMMwtHPnTs2aNUuffvqpLl26pLFjx8rNzU0ff/yx2XHTtb1799pcLlOmjCTp0KFDJqTBv+3cuVOLFy9O8Av14OBgk1KlDEp3OlepUiXt3bs33ql3ADzUtm1bDRw4UIsXL5bFYlFsbKy2bNmiAQMGWE/7gpQ1b948/fDDD2rUqJGGDRumNm3aqHDhwipVqpS2b9+uXr16mR3RoQwePFjjx49Xz549VaVKFUnStm3b1LdvX506dUqff/65yQkBc82aNUtfffWVWrVqZV3WtGlTBQQEaOrUqdqwYYMKFCigL774gtJtZ5s2bTI7AhKxYMECdezYUYGBgVq3bp0CAwN1/PhxRUREqHnz5mbHszuGl6dzixcv1kcffaS+ffuqfPny8WYKLFWqlEnJgNSBSaJSHw8PDx05ckQFChRQnjx5rKc6/Oeff1S2bFldv37d7IgOxcfHRxMmTFCbNm1sls+fP189e/bU5cuXTUoGpA6ZMmXS/v37481zcPz4cZUuXVq3b9/WyZMn9fzzz1vP5w3769q1q7755ht5eXnZLI+KilLPnj05T3cKK1WqlN555x2999571sP5ChUqpHfeeUd58uTRsGHDzI5oVxxYks61bt1aJ0+eVK9evVStWjWVKVNGZcuWtf4XcHQuLi6aN2+ejh07pkWLFmnu3Lk6evSo5syZQ+E2Sf78+RUeHi5JKlKkiNauXSvp4bA0Nzc3M6M5pJiYGFWoUCHe8vLly+vBgwcmJAJSl/z58+vHH3+Mt/zHH3+Ur6+vJOnKlSvKmjVrSkdzaLNmzdKdO3fiLb9z545mz55tQiLHduLECTVq1EiS5ObmpqioKFksFvXt21fff/+9yensj+Hl6dzJkyfNjgCkCYULF5a/v78kJokyW/PmzbVhwwZVqlRJvXv3Vps2bfTjjz8qLCxMffv2NTuew2nfvr0mT54cb9Kh77//Xu3atTMpFZB6jB07Vq+//rp+++03vfjii7JYLNq5c6eOHj1qPe3kzp071bp1a5OTOoYbN27IMAwZhqGbN2/K3d3dui4mJkarVq1Szpw5TUzomLJly6abN29KkvLly6dDhw4pICBA165dc4gRIAwvB+DwmCQqdfvrr7+0ZcsWFSlSRE2bNjU7jkN4dAbgBw8eaObMmSpQoIAqV64sSdq+fbvOnDmjjh07asKECWbFBFKN06dPa8qUKQoNDZVhGCpevLjeeecd+fn5mR3N4TxpNnmLxaJhw4bpk08+ScFUjitumP8777yjChUqqF+/fvriiy/0zTff6NVXX9W6detUrly5dD+RGqXbQRw+fDjBmQL5AxaOLrFJoiZOnKjevXszSZQJ/vjjD1WtWlUZMtgOxnrw4IG2bt2qV155xaRkjqNmzZpJ2s5isWjjxo12TgMASRcSEiLDMFSrVi398ssvypYtm3Wdq6urChYsqLx585qY0LE4OzsrPDxcGTJk0N27d5U3b17FxsZq7Nix+vPPP1WkSBENHjw43R9+QelO5/755x81b95cBw8elMViiXeO1ZiYGDPjAaZjkqjUJ+4X9L+H/125ckU5c+bkcwtAqnT79u0Ed3Awaa05Tp8+rQIFCnDImMmcnJwUERHh8EP6OaY7nevdu7cKFSqk9evXy9/fXzt27NCVK1fUv39/jR071ux4gOmYJCr1MQwjwT+Srly5Eu8MDLCvBw8eyN3dXfv27dMLL7xgdhwgVbp06ZK6dOmi3377LcH1fFFojtOnT+v06dOJrmfUVMrhiw9Kd7q3bds2bdy4UTly5JCTk5OcnJz00ksvaeTIkerVq5f27t1rdkTAVEwSlXq0aNFC0sNfzp07d7aZqTwmJkYHDhxQ1apVzYrnkDJkyKCCBQtSGoDH6NOnj65evart27erZs2aWrJkiS5cuKDPP/9cX331ldnxHFaNGjXiLXu0/PG5lnKee+65JxbvyMjIFEpjDkp3OhcTEyNPT09JD4fRnj9/XsWKFVPBggUVGhpqcjrAHI9OEmWxWDRt2jStXbs2wUmikHK8vb0lPdzT7eXlpYwZM1rXubq6qnLlynrrrbfMiuewPv30Uw0aNEhz5861OTYSwEMbN27Ur7/+qhdffFFOTk4qWLCg6tatq8yZM2vkyJHW0yQhZV29etXm8v3797V3714NHjxYX3zxhUmpHNOwYcOsv+MdFaU7nXvhhRd04MAB+fv7q1KlShozZoxcXV31/fffW0+PBDiaf4/wKF++vKSH55CUpBw5cihHjhz6+++/UzybI5sxY4Ykyc/PTwMGDGAoeSrx7bff6n//+5/y5s2rggULxntd9uzZY1IyIHWIioqyHq+aLVs2Xbp0Sc8995wCAgL4/8NECZW8unXrys3NTX379tXu3btNSOWY3njjDY7pNjsA7OvTTz9VVFSUJOnzzz9X48aN9fLLLyt79uxasGCByekAc2zatMnsCHiMDz/8UI/O8Xn69GktWbJEJUuWVGBgoInJHFOzZs3MjgCkasWKFVNoaKj8/PxUpkwZTZ06VX5+fpoyZYry5Mljdjz8S44cORjtmYI4nvshZi93QJGRkcqaNSv/E8DhMUlU6hQYGKgWLVqoe/fuunbtmooVKyZXV1ddvnxZ48aN07vvvmt2RACwmjdvnu7fv6/OnTtr7969qlevnq5cuSJXV1fNnDlTrVu3NjuiQzpw4IDNZcMwFB4erlGjRun+/fvasmWLSckcC7OXP0TpTqe6du2apO2mT59u5yRA6la4cGEFBwerdOnSZkfB//Hx8VFISIief/55TZs2TRMmTNDevXv1yy+/aMiQITpy5IjZEQEgUbdv39bRo0dVoEAB+fj4mB3HYTk5OdmcLjdO5cqVNX36dBUvXtykZHBEDC9Pp2bOnKmCBQuqbNmy8T5sAPx/TBKV+ty+fVteXl6SpLVr16pFixZycnJS5cqVH3v6F9hHTEyMxo8fr0WLFiV4DuL0PuMs8Dj3799XsWLFtGLFCpUsWVKSlClTJpUrV87kZDh58qTNZScnJ+XIkUPu7u4mJYIjo3SnU927d9eCBQv0zz//qGvXrmrfvj2FAkgAk0SlPkWKFNHSpUvVvHlzrVmzRn379pUkXbx4UZkzZzY5neMZNmyYpk2bpn79+mnw4MH65JNPdOrUKS1dulRDhgwxOx5gKhcXF0VHR3PIXipy584dbdiwQY0bN5YkDRo0SNHR0db1GTJk0PDhwynfSFEML0/HoqOjFRwcrOnTp2vr1q1q1KiRunXrpsDAQH45AP9n2LBhj10fFBSUQkkQ5+eff1bbtm0VExOjWrVqad26dZKkkSNH6o8//tBvv/1mckLHUrhwYX377bdq1KiRvLy8tG/fPuuy7du366effjI7ImCqUaNG6ejRo5o2bZoyZGB/ltmmTp2qFStWaPny5ZIkLy8vPf/889bTUB49elQffvih9QtdICVQuh3E6dOnNXPmTM2ePVv379/X4cOHrefvBoDUJiIiQuHh4SpdurScnJwkSTt27FDmzJk5Di+FeXh46MiRIypQoIDy5MmjlStXqly5cvrnn39UtmxZXb9+3eyIgKmaN2+uDRs2yNPTUwEBAfFGTAUHB5uUzDG98sor6tu3r5o3by7pYenev3+/9VS5c+fO1Xfffadt27aZGRMOhq/jHITFYrFOJhEbG2t2HAB4rNy5cyt37tw6c+aMLBaL8ufPr4oVK5odyyHlz59f4eHhKlCggIoUKaK1a9eqXLly2rlzp9zc3MyOB5guS5Ysatmypdkx8H+OHTum5557znrZ3d3d+uWtJFWsWFHvvfeeGdHgwCjd6dijw8v//PNPNW7cWBMnTlT9+vVtPnwAR8YkUanPgwcPNGzYMH377be6deuWJMnT01M9e/ZUUFCQXFxcTE7oWOL24lWqVEm9e/dWmzZt9OOPPyosLIzhmYCkGTNmmB0Bj7h+/brNMP9Lly7ZrI+NjbU5xhtICZTudKpHjx5asGCBChQooC5dumjBggXKnj272bGAVIdJolKf999/X0uWLNGYMWNUpUoVSdK2bds0dOhQXb58WVOmTDE5oWMZNWqU9d+vvfaa8ufPr61bt6pIkSJq2rSpicmA1KFWrVoKDg5WlixZbJbfuHFDzZo108aNG80J5qDy58+vQ4cOqVixYgmuP3DggPLnz5/CqeDoOKY7nXJyclKBAgVUtmzZx06axnFGcHRMEpX6eHt7a8GCBWrQoIHN8t9++01vvPEGxxADSFWcnJwUERGhnDlz2iy/ePGi8uXLp/v375uUzDH17t1b69ev1+7du+PNUH7nzh1VqFBBderU0TfffGNSQjgi9nSnUx07dmSGciAJIiIiFBAQIOnhEOa4Qte4cWMNHjzYzGgOy93dXX5+fvGW+/n5ydXVNeUDQaGhoZowYYKOHDkii8Wi4sWLq2fPnonuSQIcwYEDB6z/Pnz4sCIiIqyXY2JitHr1auXLl8+MaA7t448/1qJFi1SsWDG9//77eu6552SxWHT06FFNnDhRDx480Mcff2x2TDgYSnc6NXPmTLMjAGkCk0SlPu+9954+++wzzZgxw/oaREdH64svvtD7779vcjrH8/PPP6tNmzaqUKGCdbj/9u3b9cILL+inn37S66+/bnJCwBxlypSxTlRbq1ateOszZsyoCRMmmJDMseXKlUtbt27Vu+++q48++khxg3otFovq1q2rSZMmKVeuXCanhKNheDkAh/bRRx8pc+bM+vjjj63lws/PzzpJ1KPHs8J+WrRoYXN5/fr1cnNzU+nSpSVJ+/fv171791S7dm0Oi0lh/v7+at++vYYPH26zPCgoSHPmzNE///xjUjLAXKdPn5ZhGPL399eOHTuUI0cO6zpXV1flzJlTzs7OJiZEZGSk/ve//0mSihQpomzZspmcCI6K0g0Aj/jrr7+0ZcsWJolKYV26dEnytswUnLIyZcqkAwcOqEiRIjbLjx8/rtKlS+v27dsmJQPMd//+fb311lsaMmSI9TzQAPBvlG4ASIJGjRpp2rRpypMnj9lRgBTVsGFDvf766/G+GJkxY4YWLFigNWvWmJQMSB2yZs2q3bt3U7oBJIpjugEgCf744w/duXPH7BhAimvatKkGDhyo3bt3q3LlypIeHtO9ePFiDRs2TMuWLbPZFnA0zZo109KlS9WvXz+zowBIpdjTDQBJ4OXlpf3797Mnw06edHrDR+3Zs8fOafAoJyenJG1nsVgUExNj5zRA6vPFF19o7Nixql27tsqXLy8PDw+b9b169TIpGYDUgj3dAADTNWvWzOwISERsbKzZEYBUbdq0acqSJYt2796t3bt326yzWCyUbgDs6QaApGBPNxzNX3/9pcjISDVo0MC6bPbs2QoKClJUVJSaNWumCRMmcGo9AACeIGljxgAASEHXrl3TtGnTNGjQIEVGRkp6OKz83LlzJidzHEOHDtWBAweslw8ePKhu3bqpTp06+uijj7R8+XKNHDnSxIRA6nLv3j2FhobqwYMHZkcBkMpQugEAqcqBAwf03HPPafTo0Ro7dqyuXbsmSVqyZIkGDRpkbjgHsm/fPtWuXdt6ecGCBapUqZJ++OEH9evXT99++60WLVpkYkIgdbh9+7a6deumTJky6fnnn1dYWJikh8dyjxo1yuR0AFIDSjcAh1OuXDldvXpVkjR8+PAknWf4448/VrZs2ewdDZL69eunzp076/jx43J3d7cub9Cggf744w8TkzmWq1evKleuXNbLISEhql+/vvXyiy++qDNnzpgRDUhVBg0apP379+v333+3+cyqU6eOFi5caGIyAKkFpRuAwzly5IiioqIkScOGDdOtW7eeeJ1BgwYpS5Ysdk4GSdq5c6feeeedeMvz5cuniIgIExI5ply5cunkyZOSHg6b3bNnj6pUqWJdf/PmTbm4uJgVD0g1li5dqokTJ+qll16yOQtDyZIldeLECROTAUgtmL0cgMMpU6aMunTpopdeekmGYWjs2LHy9PRMcNshQ4akcDq4u7vrxo0b8ZaHhoYqR44cJiRyTPXr19dHH32k0aNHa+nSpcqUKZNefvll6/oDBw6ocOHCJiYEUodLly4pZ86c8ZZHRUUl+VSIANI3SjcAhzNz5kwFBQVpxYoVslgs+u2335QhQ/yPQ4vFQuk2wauvvqrhw4dbjxe2WCwKCwvTRx99pJYtW5qcznF8/vnnatGihapXry5PT0/NmjVLrq6u1vXTp09XYGCgiQmB1OHFF1/UypUr1bNnT0myFu0ffvjBZnQIAMfFKcMAODQnJydFREQkuJcC5rhx44YaNmyov//+Wzdv3lTevHkVERGhKlWqaNWqVfLw8DA7okO5fv26PD095ezsbLM8MjJSnp6eNkUccCT79u1TmTJltG3bNtWrV0/t2rXTzJkz9c477+jvv//Wtm3bFBISovLly5sdFYDJKN0AgFRp48aN2rNnj2JjY1WuXDnVqVPH7EgAYOXk5KSyZcvqzTffVEBAgH744Qft3r3b+pk1cOBABQQEmB0TQCpA6Qbg0GbPnv3Y9R07dkyhJHica9euMZEdgFRl27Ztmj59uhYtWqT79++rRYsW6tq1q2rVqmV2NACpDKUbgEPLmjWrzeX79+/r9u3bcnV1VaZMmRQZGWlSMsc1evRo+fn5qXXr1pKkVq1a6ZdfflHu3Lm1atUqlS5d2uSEAPD/3blzR4sWLdKMGTO0efNm+fn5qWvXrurUqZPy589vdjwAqQCnDAPg0K5evWrzc+vWLYWGhuqll17S/PnzzY7nkKZOnSpfX19J0rp167Ru3Tr99ttvatCggT744AOT0wGArYwZM6pTp076/fffdezYMbVp00ZTp05VoUKF1LBhQ7PjAUgF2NMNAAnYtWuX2rdvr6NHj5odxeFkzJhRx44dk6+vr3r37q27d+9q6tSpOnbsmCpVqqSrV6+aHREAEnXr1i3NmzdPH3/8sa5du6aYmBizIwEwGXu6ASABzs7OOn/+vNkxHFLWrFl15swZSdLq1autE6gZhsEfrwBSrZCQEHXq1Em5c+fWhx9+qBYtWmjLli1mxwKQCnCebgAObdmyZTaXDcNQeHi4Jk6cqGrVqpmUyrG1aNFCbdu2VdGiRXXlyhU1aNBA0sPT8xQpUsTkdADw/505c0YzZ87UzJkzdfLkSVWtWlUTJkxQq1atOL0hACtKNwCH1qxZM5vLFotFOXLkUK1atfTVV1+ZE8rBjR8/Xn5+fjpz5ozGjBkjT09PSVJ4eLh69OhhcjoAeKhu3bratGmTcuTIoY4dO6pr164qVqyY2bEApEIc0w0ASJMaNWqkadOmKU+ePGZHAeCAmjZtqm7duqlx48ZydnY2Ow6AVIzSDcBh3b9/X8WKFdOKFStUsmRJs+PgKXl5eWn//v3y9/c3OwoAAECimEgNgMNycXFRdHS0LBaL2VEAAACQTlG6ATi0nj17avTo0Xrw4IHZUQAAAJAOMZEaAIf2119/acOGDVq7dq0CAgLizTYbHBxsUjIAAACkB5RuAA4tS5YsatmypdkxAAAAkE4xkRoAIE1iIjUAAJAWcEw3AIf34MEDrV+/XlOnTtXNmzclSefPn9etW7dMTuY4ypUrp6tXr0qShg8frtu3bz/xOh9//LGyZctm72gAAAD/CXu6ATi006dPq379+goLC1N0dLSOHTsmf39/9enTR3fv3tWUKVPMjugQMmbMqOPHjyt//vxydnZWeHi4cubMaXYsAACA/4xjugE4tN69e6tChQrav3+/smfPbl3evHlzvfnmmyYmcyxlypRRly5d9NJLL8kwDI0dO1aenp4JbjtkyJAUTgcAAJB87OkG4NB8fHy0ZcsWFStWzOYY4VOnTqlkyZJJGuaM/y40NFRBQUE6ceKE9uzZo5IlSypDhvjfC1ssFu3Zs8eEhAAAAMnDnm4ADi02NlYxMTHxlp89e1ZeXl4mJHJMxYoV04IFCyRJTk5O2rBhA8PLAQBAusBEagAcWt26dfX1119bL1ssFt26dUtBQUFq2LChecEcWGxsLIUbAACkGwwvB+DQzp8/r5o1a8rZ2VnHjx9XhQoVdPz4cWXPnl2bN2+m/Jlg9uzZj13fsWPHFEoCAADw31G6ATi8O3fuaP78+dqzZ49iY2NVrlw5tWvXThkzZjQ7mkPKmjWrzeX79+/r9u3bcnV1VaZMmRQZGWlSMgAAgKfH8HIADu3KlSvKmDGjunbtqg8//FA+Pj4KDQ3Vrl27zI7msK5evWrzc+vWLYWGhuqll17S/PnzzY4HAADwVNjTDcAhHTx4UE2aNNGZM2dUtGhRLViwQPXr11dUVJScnJwUFRWln3/+Wc2aNTM7Kv7Prl271L59ex09etTsKAAAAEnGnm4ADunDDz9UQECAQkJCVKNGDTVu3FgNGzbU9evXdfXqVb3zzjsaNWqU2THxCGdnZ50/f97sGAAAAE+FPd0AHJKPj482btyoUqVK6datW8qcObN27NihChUqSJKOHj2qypUr69q1a+YGdUDLli2zuWwYhsLDwzVx4kT5+vrqt99+MykZAADA0+M83QAcUmRkpHLnzi1J8vT0lIeHh7Jly2ZdnzVrVt28edOseA7t30P6LRaLcuTIoVq1aumrr74yJxQAAEAyUboBOCyLxfLYyzBHbGys2REAAACeGUo3AIfVuXNnubm5SZLu3r2r7t27y8PDQ5IUHR1tZjSHdf/+fRUrVkwrVqxQyZIlzY4DAADwn1G6ATikTp062Vxu3759vG06duyYUnHwf1xcXBQdHc2oAwAAkG4wkRoAIFUZNWqUjh49qmnTpilDBr4bBgAAaRulGwCQqjRv3lwbNmyQp6enAgICrEP+4wQHB5uUDAAA4OmxCwEAkKpkyZJFLVu2NDsGAADAM8GebgAAAAAA7MTJ7AAAAPzbgwcPtH79ek2dOtV6vvTz58/r1q1bJicDAAB4OuzpBgCkKqdPn1b9+vUVFham6OhoHTt2TP7+/urTp4/u3r2rKVOmmB0RAAAgydjTDQBIVXr37q0KFSro6tWrypgxo3V53ARrAAAAaQkTqQEAUpU///xTW7Zskaurq83yggUL6ty5cyalAgAASB72dAMAUpXY2FjFxMTEW3727Fl5eXmZkAgAACD5KN0AgFSlbt26+vrrr62XLRaLbt26paCgIDVs2NC8YAAAAMnARGoAgFTl/PnzqlmzppydnXX8+HFVqFBBx48fV/bs2bV582blzJnT7IgAAABJRukGAKQ6d+7c0fz587Vnzx7FxsaqXLlyateunc3EagAAAGkBpRsAkKpcuXJF2bNnlySFhYVp2rRpunPnjpo2baqXX37Z5HQAAABPh9INAEgVDh48qCZNmujMmTMqWrSoFixYoPr16ysqKkpOTk6KiorSzz//rGbNmpkdFQAAIMmYSA0AkCp8+OGHCggIUEhIiGrUqKHGjRurYcOGun79uq5evap33nlHo0aNMjsmAADAU2FPNwAgVfDx8dHGjRtVqlQp3bp1S5kzZ9aOHTtUoUIFSdLRo0dVuXJlXbt2zdygAAAAT4E93QCAVCEyMlK5c+eWJHl6esrDw0PZsmWzrs+aNatu3rxpVjwAAIBkoXQDAFINi8Xy2MsAAABpTQazAwAAEKdz585yc3OTJN29e1fdu3eXh4eHJCk6OtrMaAAAAMnCMd0AgFShS5cuSdpuxowZdk4CAADw7FC6AQAAAACwE47pBgAAAADATijdAAAAAADYCaUbAAAAAAA7oXQDAAAAAGAnlG4AAGA1dOhQlSlTJsXv9/fff5fFYtG1a9dS/L4BALAnSjcAAGlE586dZbFY1L1793jrevToIYvFos6dO6d8MAAAkChKNwAAaYivr68WLFigO3fuWJfdvXtX8+fPV4ECBZJ9u4Zh6MGDB88iIgAAeASlGwCANKRcuXIqUKCAgoODrcuCg4Pl6+ursmXLWpcZhqExY8bI399fGTNmVOnSpfXzzz9b18cN516zZo0qVKggNzc3bd68Od79nTx5UkWKFNG7776r2NhYzZw5U1myZNGKFStUrFgxZcqUSa+99pqioqI0a9Ys+fn5KWvWrOrZs6diYmKstzN37lxVqFBBXl5eyp07t9q2bauLFy8+9rFu3bpVr7zyijJmzChfX1/16tVLUVFR/+XpAwAgxVG6AQBIY7p06aIZM2ZYL0+fPl1du3a12ebTTz/VjBkzNHnyZP3999/q27ev2rdvr5CQEJvtPvzwQ40cOVJHjhxRqVKlbNYdOnRI1apV0+uvv67JkyfLyenhnw23b9/Wt99+qwULFmj16tX6/fff1aJFC61atUqrVq3SnDlz9P3339uU/Hv37umzzz7T/v37tXTpUp08efKxQ+EPHjyoevXqqUWLFjpw4IAWLlyoP//8U++//35ynzYAAEyRwewAAADg6XTo0EGDBg3SqVOnZLFYtGXLFi1YsEC///67JCkqKkrjxo3Txo0bVaVKFUmSv7+//vzzT02dOlXVq1e33tbw4cNVt27dePexbds2NW7cWIMGDdKAAQNs1t2/f1+TJ09W4cKFJUmvvfaa5syZowsXLsjT01MlS5ZUzZo1tWnTJrVu3VqSbL4U8Pf317fffquKFSvq1q1b8vT0jHf/X375pdq2bas+ffpIkooWLapvv/1W1atX1+TJk+Xu7p78JxAAgBRE6QYAII3x8fFRo0aNNGvWLBmGoUaNGsnHx8e6/vDhw7p79268Mn3v3j2bIeiSVKFChXi3HxYWpjp16ujzzz9X3759463PlCmTtXBLUq5cueTn52dTnnPlymUzfHzv3r0aOnSo9u3bp8jISMXGxlrvq2TJkvHuY/fu3frf//6nefPmWZcZhqHY2FidPHlSJUqUSPT5AQAgNaF0AwCQBnXt2tU61Pq7776zWRdXaFeuXKl8+fLZrHNzc7O57OHhEe+2c+TIobx582rBggXq1q2bMmfObLPexcXF5rLFYklwWVyOqKgoBQYGKjAwUHPnzlWOHDkUFhamevXq6d69ewk+vtjYWL3zzjvq1atXvHX/ZcI4AABSGqUbAIA0qH79+tbCWq9ePZt1JUuWlJubm8LCwmyGkidVxowZtWLFCjVs2FD16tXT2rVr5eXlleysR48e1eXLlzVq1Cj5+vpKknbt2vXY65QrV05///23ihQpkuz7BQAgNWAiNQAA0iBnZ2cdOXJER44ckbOzs806Ly8vDRgwQH379tWsWbN04sQJ7d27V999951mzZqVpNv38PDQypUrlSFDBjVo0EC3bt1KdtYCBQrI1dVVEyZM0D///KNly5bps88+e+x1Bg4cqG3btum9997Tvn37dPz4cS1btkw9e/ZMdg4AAMxA6QYAII3KnDlzvKHfcT777DMNGTJEI0eOVIkSJVSvXj0tX75chQoVSvLte3p66rfffpNhGGrYsGGyT9eVI0cOzZw5U4sXL1bJkiU1atQojR079rHXKVWqlEJCQnT8+HG9/PLLKlu2rAYPHqw8efIkKwMAAGaxGIZhmB0CAAAAAID0iD3dAAAAAADYCaUbAAAAAAA7oXQDAAAAAGAnlG4AAAAAAOyE0g0AAAAAgJ1QugEAAAAAsBNKNwAAAAAA/6/9OhYAAAAAGORvPYp9ZdFEugEAAGAi3QAAADCRbgAAAJhINwAAAEykGwAAACYBpRoe5gtTgc4AAAAASUVORK5CYII=",
|
|
"text/plain": [
|
|
"<Figure size 1000x600 with 1 Axes>"
|
|
]
|
|
},
|
|
"metadata": {},
|
|
"output_type": "display_data"
|
|
}
|
|
],
|
|
"source": [
|
|
"# Pearson-Korrelation berechnen: Zielvariable ist \"Status_gekuendigt\"\n",
|
|
"# \"Vertragsnummer\", \"Status_aktiv\", \"Status_verstorben\" und \"p\" sind nicht relevant für die Korrelation\n",
|
|
"correlation_matrix = data.corr(method='pearson')\n",
|
|
"correlation_with_target = correlation_matrix['Status_gekuendigt'].drop(['Status_gekuendigt', 'Vertragsnummer', 'Status_aktiv', 'Status_verstorben', 'p'])\n",
|
|
"correlation_with_target = correlation_with_target.sort_values(ascending=False)\n",
|
|
"\n",
|
|
"# Korrelationen als Balkenplot darstellen\n",
|
|
"import matplotlib.pyplot as plt\n",
|
|
"import seaborn as sns\n",
|
|
"plt.figure(figsize=(10, 6))\n",
|
|
"sns.barplot(x=correlation_with_target.index, y=correlation_with_target.values)\n",
|
|
"plt.xticks(rotation=90)\n",
|
|
"plt.title('Korrelationen mit der Zielvariable \"Status_gekuendigt\"')\n",
|
|
"plt.ylabel('Korrelation')\n",
|
|
"plt.xlabel('Merkmale')\n",
|
|
"plt.tight_layout()\n",
|
|
"plt.show()"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "76057188",
|
|
"metadata": {},
|
|
"source": [
|
|
"## Train und Test Split"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 30,
|
|
"id": "15f15627",
|
|
"metadata": {},
|
|
"outputs": [],
|
|
"source": [
|
|
"from sklearn.model_selection import train_test_split\n",
|
|
"\n",
|
|
"data = data.drop(['Vertragsnummer', 'Status_aktiv', 'Status_verstorben', 'p'], axis=1)\n",
|
|
"X_train, X_test, y_train, y_test = train_test_split(data.drop(['Status_gekuendigt'], axis=1), data['Status_gekuendigt'], test_size=0.2, random_state=11)"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "081a8c31",
|
|
"metadata": {},
|
|
"source": [
|
|
"## Multikolinearität mit VIF prüfen"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 31,
|
|
"id": "59a84f11",
|
|
"metadata": {},
|
|
"outputs": [
|
|
{
|
|
"data": {
|
|
"image/png": "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",
|
|
"text/plain": [
|
|
"<Figure size 1000x600 with 1 Axes>"
|
|
]
|
|
},
|
|
"metadata": {},
|
|
"output_type": "display_data"
|
|
}
|
|
],
|
|
"source": [
|
|
"# Multikolinearität mit VIF prüfen\n",
|
|
"from statsmodels.stats.outliers_influence import variance_inflation_factor\n",
|
|
"\n",
|
|
"# VIF für alle Merkmale berechnen\n",
|
|
"vif_data = pd.DataFrame()\n",
|
|
"vif_data['feature'] = X_train.columns\n",
|
|
"vif_data['VIF'] = [variance_inflation_factor(X_train.values, i) for i in range(X_train.shape[1])]\n",
|
|
"vif_data = vif_data.sort_values(by='VIF', ascending=False)\n",
|
|
"\n",
|
|
"# VIF-Werte mit Balkenplot darstellen\n",
|
|
"plt.figure(figsize=(10, 6))\n",
|
|
"sns.barplot(x=vif_data['feature'], y=vif_data['VIF'])\n",
|
|
"plt.xticks(rotation=90)\n",
|
|
"plt.title('VIF-Werte der Merkmale')\n",
|
|
"plt.ylabel('VIF')\n",
|
|
"plt.xlabel('Merkmale')\n",
|
|
"plt.tight_layout()\n",
|
|
"plt.show()"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "c4b9dc3a",
|
|
"metadata": {},
|
|
"source": [
|
|
"## Löschen der Variablen mit hohem VIF"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 32,
|
|
"id": "8cf50a4e",
|
|
"metadata": {},
|
|
"outputs": [],
|
|
"source": [
|
|
"# Loeschen von Merkmalen mit hohem VIF: > 10\n",
|
|
"for feature in vif_data['feature'][vif_data['VIF'] > 10]:\n",
|
|
" X_train = X_train.drop(feature, axis=1)\n",
|
|
" X_test = X_test.drop(feature, axis=1)"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "a0360323",
|
|
"metadata": {},
|
|
"source": [
|
|
"## Multiples lineares Regressionsmodell"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 33,
|
|
"id": "71580cbd",
|
|
"metadata": {},
|
|
"outputs": [
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Mean Squared Error: 0.11601394197589698\n",
|
|
"R^2 Score: 0.27326238891781685\n"
|
|
]
|
|
},
|
|
{
|
|
"data": {
|
|
"image/png": "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",
|
|
"text/plain": [
|
|
"<Figure size 1000x600 with 1 Axes>"
|
|
]
|
|
},
|
|
"metadata": {},
|
|
"output_type": "display_data"
|
|
}
|
|
],
|
|
"source": [
|
|
"from sklearn.linear_model import LinearRegression\n",
|
|
"from sklearn.metrics import mean_squared_error, r2_score\n",
|
|
"\n",
|
|
"model = LinearRegression()\n",
|
|
"model.fit(X_train, y_train)\n",
|
|
"\n",
|
|
"# Vorhersagen auf dem Testdatensatz\n",
|
|
"y_pred = model.predict(X_test)\n",
|
|
"\n",
|
|
"# Modellbewertung\n",
|
|
"mse = mean_squared_error(y_test, y_pred)\n",
|
|
"r2 = r2_score(y_test, y_pred)\n",
|
|
"print(f'Mean Squared Error: {mse}')\n",
|
|
"print(f'R^2 Score: {r2}')\n",
|
|
"\n",
|
|
"# Feature Importances\n",
|
|
"importances = model.coef_\n",
|
|
"feature_importances = pd.DataFrame({'Feature': X_train.columns, 'Importance': importances})\n",
|
|
"feature_importances = feature_importances.sort_values(by='Importance', ascending=False)\n",
|
|
"\n",
|
|
"# Feature Importances als Balkenplot darstellen\n",
|
|
"plt.figure(figsize=(10, 6))\n",
|
|
"sns.barplot(x=feature_importances['Feature'], y=feature_importances['Importance'])\n",
|
|
"plt.xticks(rotation=90)\n",
|
|
"plt.title('Feature Importances')\n",
|
|
"plt.ylabel('Importance')\n",
|
|
"plt.xlabel('Merkmale')\n",
|
|
"plt.tight_layout()\n",
|
|
"plt.show()"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "80ac4986",
|
|
"metadata": {},
|
|
"source": [
|
|
"## Backwards Selection"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "df4f5ff2",
|
|
"metadata": {},
|
|
"outputs": [
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Feature \"Sparbeitrag\" mit p-value 0.2170 wird entfernt\n",
|
|
"Feature \"Guthaben\" mit p-value 0.4518 wird entfernt\n",
|
|
"Mean Squared Error (Selected Features): 0.11601088142214849\n",
|
|
"R^2 Score (Selected Features): 0.2732815609197494\n"
|
|
]
|
|
}
|
|
],
|
|
"source": [
|
|
"# Backwards Selection der Variablen\n",
|
|
"import statsmodels.api as sm\n",
|
|
"from sklearn.linear_model import LinearRegression\n",
|
|
"from sklearn.metrics import mean_squared_error, r2_score\n",
|
|
"\n",
|
|
"# Funktion zur Durchführung der Backward Selection\n",
|
|
"def backward_selection(X, y, significance_level=0.05):\n",
|
|
" num_vars = len(X.columns)\n",
|
|
" for i in range(num_vars):\n",
|
|
" X_with_const = sm.add_constant(X)\n",
|
|
" p_values = sm.OLS(y, X_with_const).fit().pvalues[1:]\n",
|
|
" max_p_value = p_values.max()\n",
|
|
" if max_p_value > significance_level:\n",
|
|
" feature_to_remove = p_values.idxmax()\n",
|
|
" print(f'Feature \"{feature_to_remove}\" mit p-value {max_p_value:.4f} wird entfernt')\n",
|
|
" X = X.drop(feature_to_remove, axis=1)\n",
|
|
" return X\n",
|
|
"\n",
|
|
"# Backward Selection durchführen\n",
|
|
"X_train_selected = backward_selection(X_train, y_train)\n",
|
|
"X_test_selected = X_test[X_train_selected.columns]\n",
|
|
"\n",
|
|
"# Modell mit den ausgewählten Merkmalen trainieren\n",
|
|
"model_selected = LinearRegression()\n",
|
|
"model_selected.fit(X_train_selected, y_train)\n",
|
|
"\n",
|
|
"# Vorhersagen auf dem Testdatensatz\n",
|
|
"y_pred_selected = model_selected.predict(X_test_selected)\n",
|
|
"\n",
|
|
"# Modellbewertung\n",
|
|
"mse_selected = mean_squared_error(y_test, y_pred_selected)\n",
|
|
"r2_selected = r2_score(y_test, y_pred_selected)\n",
|
|
"print(f'Mean Squared Error (Selected Features): {mse_selected}')\n",
|
|
"print(f'R^2 Score (Selected Features): {r2_selected}')"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "9dafd48c",
|
|
"metadata": {},
|
|
"source": []
|
|
}
|
|
],
|
|
"metadata": {
|
|
"kernelspec": {
|
|
"display_name": "base",
|
|
"language": "python",
|
|
"name": "python3"
|
|
},
|
|
"language_info": {
|
|
"codemirror_mode": {
|
|
"name": "ipython",
|
|
"version": 3
|
|
},
|
|
"file_extension": ".py",
|
|
"mimetype": "text/x-python",
|
|
"name": "python",
|
|
"nbconvert_exporter": "python",
|
|
"pygments_lexer": "ipython3",
|
|
"version": "3.10.9"
|
|
}
|
|
},
|
|
"nbformat": 4,
|
|
"nbformat_minor": 5
|
|
}
|