tak-akashi/LinearRegression(2).ipynb

## LinearRegression(2).ipynb
{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# データ分析と可視化に必要なライブラリをインポート\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 統計解析で利用するstatsmodelをインポート\n",
    "import statsmodels.api as sm\n",
    "import statsmodels.formula.api as smf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 可視化の際に見やすくする\n",
    "plt.style.use('seaborn')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# csvファイルをpandasのDataframe形式で取り込む\n",
    "# 「RとPythonで学［実践的］データサイエンス＆機械学習」にて利用しているデータ\n",
    "df = pd.read_csv('Ch03/ESSurvey.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 124 entries, 0 to 123\n",
      "Data columns (total 3 columns):\n",
      "Meeting     124 non-null int64\n",
      "Reassign    124 non-null float64\n",
      "Score       124 non-null float64\n",
      "dtypes: float64(2), int64(1)\n",
      "memory usage: 3.0 KB\n"
     ]
    }
   ],
   "source": [
    "# データフレームの基本情報を表示\n",
    "df.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Meeting</th>\n",
       "      <th>Reassign</th>\n",
       "      <th>Score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>15</td>\n",
       "      <td>0.26</td>\n",
       "      <td>72.7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>11</td>\n",
       "      <td>0.45</td>\n",
       "      <td>30.1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>32</td>\n",
       "      <td>0.40</td>\n",
       "      <td>74.9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>9</td>\n",
       "      <td>0.44</td>\n",
       "      <td>55.9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>31</td>\n",
       "      <td>0.36</td>\n",
       "      <td>54.1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Meeting  Reassign  Score\n",
       "0       15      0.26   72.7\n",
       "1       11      0.45   30.1\n",
       "2       32      0.40   74.9\n",
       "3        9      0.44   55.9\n",
       "4       31      0.36   54.1"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Dataframeの最初の５行を表示\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Meeting</th>\n",
       "      <th>Reassign</th>\n",
       "      <th>Score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>124.000000</td>\n",
       "      <td>124.000000</td>\n",
       "      <td>124.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>22.306452</td>\n",
       "      <td>0.332903</td>\n",
       "      <td>66.154839</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>8.898869</td>\n",
       "      <td>0.112234</td>\n",
       "      <td>11.987387</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.110000</td>\n",
       "      <td>30.100000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>15.750000</td>\n",
       "      <td>0.250000</td>\n",
       "      <td>58.825000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>22.000000</td>\n",
       "      <td>0.340000</td>\n",
       "      <td>66.650000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>29.000000</td>\n",
       "      <td>0.410000</td>\n",
       "      <td>73.325000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>50.000000</td>\n",
       "      <td>0.630000</td>\n",
       "      <td>99.500000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          Meeting    Reassign       Score\n",
       "count  124.000000  124.000000  124.000000\n",
       "mean    22.306452    0.332903   66.154839\n",
       "std      8.898869    0.112234   11.987387\n",
       "min      2.000000    0.110000   30.100000\n",
       "25%     15.750000    0.250000   58.825000\n",
       "50%     22.000000    0.340000   66.650000\n",
       "75%     29.000000    0.410000   73.325000\n",
       "max     50.000000    0.630000   99.500000"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Dataframeの要約統計量の把握\n",
    "df.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x396 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 目的変数のヒストグラムを作成\n",
    "plt.hist(df['Score'],bins=30,rwidth=0.9)\n",
    "plt.xlabel('Score')\n",
    "plt.ylabel('Frequency');"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<seaborn.axisgrid.PairGrid at 0x1c2078a5f8>"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 540x540 with 12 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# ペアプロットを描く\n",
    "sns.pairplot(df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "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>Meeting</th>\n",
       "      <th>Reassign</th>\n",
       "      <th>Score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Meeting</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>-0.041355</td>\n",
       "      <td>0.371995</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Reassign</th>\n",
       "      <td>-0.041355</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.134221</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Score</th>\n",
       "      <td>0.371995</td>\n",
       "      <td>0.134221</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           Meeting  Reassign     Score\n",
       "Meeting   1.000000 -0.041355  0.371995\n",
       "Reassign -0.041355  1.000000  0.134221\n",
       "Score     0.371995  0.134221  1.000000"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 相関係数を計算\n",
    "df.corr()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 通常の線形回帰モデル"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = smf.ols('Score ~ Meeting + Reassign', data=df)\n",
    "LE1 = model.fit()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table class=\"simpletable\">\n",
       "<caption>OLS Regression Results</caption>\n",
       "<tr>\n",
       "  <th>Dep. Variable:</th>          <td>Score</td>      <th>  R-squared:         </th> <td>   0.161</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Model:</th>                   <td>OLS</td>       <th>  Adj. R-squared:    </th> <td>   0.147</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Method:</th>             <td>Least Squares</td>  <th>  F-statistic:       </th> <td>   11.59</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Date:</th>             <td>Mon, 14 Oct 2019</td> <th>  Prob (F-statistic):</th> <td>2.48e-05</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Time:</th>                 <td>16:42:39</td>     <th>  Log-Likelihood:    </th> <td> -472.58</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>No. Observations:</th>      <td>   124</td>      <th>  AIC:               </th> <td>   951.2</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Df Residuals:</th>          <td>   121</td>      <th>  BIC:               </th> <td>   959.6</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Df Model:</th>              <td>     2</td>      <th>                     </th>     <td> </td>   \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Covariance Type:</th>      <td>nonrobust</td>    <th>                     </th>     <td> </td>   \n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "      <td></td>         <th>coef</th>     <th>std err</th>      <th>t</th>      <th>P>|t|</th>  <th>[0.025</th>    <th>0.975]</th>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Intercept</th> <td>   49.4622</td> <td>    4.082</td> <td>   12.119</td> <td> 0.000</td> <td>   41.382</td> <td>   57.543</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Meeting</th>   <td>    0.5095</td> <td>    0.112</td> <td>    4.537</td> <td> 0.000</td> <td>    0.287</td> <td>    0.732</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Reassign</th>  <td>   16.0063</td> <td>    8.903</td> <td>    1.798</td> <td> 0.075</td> <td>   -1.619</td> <td>   33.631</td>\n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "  <th>Omnibus:</th>       <td> 6.619</td> <th>  Durbin-Watson:     </th> <td>   2.323</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Prob(Omnibus):</th> <td> 0.037</td> <th>  Jarque-Bera (JB):  </th> <td>   6.155</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Skew:</th>          <td>-0.515</td> <th>  Prob(JB):          </th> <td>  0.0461</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Kurtosis:</th>      <td> 3.363</td> <th>  Cond. No.          </th> <td>    229.</td>\n",
       "</tr>\n",
       "</table><br/><br/>Warnings:<br/>[1] Standard Errors assume that the covariance matrix of the errors is correctly specified."
      ],
      "text/plain": [
       "<class 'statsmodels.iolib.summary.Summary'>\n",
       "\"\"\"\n",
       "                            OLS Regression Results                            \n",
       "==============================================================================\n",
       "Dep. Variable:                  Score   R-squared:                       0.161\n",
       "Model:                            OLS   Adj. R-squared:                  0.147\n",
       "Method:                 Least Squares   F-statistic:                     11.59\n",
       "Date:                Mon, 14 Oct 2019   Prob (F-statistic):           2.48e-05\n",
       "Time:                        16:42:39   Log-Likelihood:                -472.58\n",
       "No. Observations:                 124   AIC:                             951.2\n",
       "Df Residuals:                     121   BIC:                             959.6\n",
       "Df Model:                           2                                         \n",
       "Covariance Type:            nonrobust                                         \n",
       "==============================================================================\n",
       "                 coef    std err          t      P>|t|      [0.025      0.975]\n",
       "------------------------------------------------------------------------------\n",
       "Intercept     49.4622      4.082     12.119      0.000      41.382      57.543\n",
       "Meeting        0.5095      0.112      4.537      0.000       0.287       0.732\n",
       "Reassign      16.0063      8.903      1.798      0.075      -1.619      33.631\n",
       "==============================================================================\n",
       "Omnibus:                        6.619   Durbin-Watson:                   2.323\n",
       "Prob(Omnibus):                  0.037   Jarque-Bera (JB):                6.155\n",
       "Skew:                          -0.515   Prob(JB):                       0.0461\n",
       "Kurtosis:                       3.363   Cond. No.                         229.\n",
       "==============================================================================\n",
       "\n",
       "Warnings:\n",
       "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
       "\"\"\""
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "LE1.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x396 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 残差のヒストグラムを描く\n",
    "plt.hist(LE1.resid, bins=25, color='cadetblue', rwidth=0.9);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'Residuals')"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x396 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 残差の分布を見る\n",
    "sns.scatterplot(x=LE1.fittedvalues, y=LE1.resid)\n",
    "plt.xlabel('Fitted Value')\n",
    "plt.ylabel('Residuals')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1c212ef5c0>"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x396 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 残差の密度関数分布を正規分布と比較\n",
    "from scipy.stats import norm\n",
    "x = np.linspace(LE1.resid.min() * 1.5, LE1.resid.max() * 1.5, 100) \n",
    "plt.plot(x, norm.pdf(x, loc=LE1.resid.mean(), scale=LE1.resid.std()))\n",
    "sns.distplot(LE1.resid)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x396 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# QQプロットの表示\n",
    "from scipy.stats import probplot\n",
    "probplot(LE1.resid, dist=\"norm\", plot=plt);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 残差とレバレッジの表示\n",
    "fig, ax = plt.subplots(figsize=(12,8))\n",
    "fig = sm.graphics.influence_plot(LE1, ax=ax, criterion=\"cooks\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "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>VIF</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Intercept</th>\n",
       "      <td>16.851164</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Meeting</th>\n",
       "      <td>1.001713</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Reassign</th>\n",
       "      <td>1.001713</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 VIF\n",
       "Intercept  16.851164\n",
       "Meeting     1.001713\n",
       "Reassign    1.001713"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# VIFを求める\n",
    "from statsmodels.stats.outliers_influence import variance_inflation_factor\n",
    "num_cols = model.exog.shape[1]\n",
    "vifs = [variance_inflation_factor(model.exog, i)\n",
    "        for i in range(0, num_cols)]\n",
    "df_vif= pd.DataFrame(vifs, index=model.exog_names, columns=[\"VIF\"])\n",
    "df_vif"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 交互作用項を含むモデル"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = smf.ols('Score ~ Meeting * Reassign', data=df)\n",
    "LE2 = model.fit()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table class=\"simpletable\">\n",
       "<caption>OLS Regression Results</caption>\n",
       "<tr>\n",
       "  <th>Dep. Variable:</th>          <td>Score</td>      <th>  R-squared:         </th> <td>   0.478</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Model:</th>                   <td>OLS</td>       <th>  Adj. R-squared:    </th> <td>   0.465</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Method:</th>             <td>Least Squares</td>  <th>  F-statistic:       </th> <td>   36.67</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Date:</th>             <td>Mon, 14 Oct 2019</td> <th>  Prob (F-statistic):</th> <td>6.80e-17</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Time:</th>                 <td>16:42:40</td>     <th>  Log-Likelihood:    </th> <td> -443.10</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>No. Observations:</th>      <td>   124</td>      <th>  AIC:               </th> <td>   894.2</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Df Residuals:</th>          <td>   120</td>      <th>  BIC:               </th> <td>   905.5</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Df Model:</th>              <td>     3</td>      <th>                     </th>     <td> </td>   \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Covariance Type:</th>      <td>nonrobust</td>    <th>                     </th>     <td> </td>   \n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "          <td></td>            <th>coef</th>     <th>std err</th>      <th>t</th>      <th>P>|t|</th>  <th>[0.025</th>    <th>0.975]</th>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Intercept</th>        <td>   97.7568</td> <td>    6.510</td> <td>   15.017</td> <td> 0.000</td> <td>   84.868</td> <td>  110.646</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Meeting</th>          <td>   -1.6738</td> <td>    0.270</td> <td>   -6.188</td> <td> 0.000</td> <td>   -2.209</td> <td>   -1.138</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Reassign</th>         <td> -129.7491</td> <td>   18.454</td> <td>   -7.031</td> <td> 0.000</td> <td> -166.287</td> <td>  -93.211</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Meeting:Reassign</th> <td>    6.6255</td> <td>    0.775</td> <td>    8.546</td> <td> 0.000</td> <td>    5.090</td> <td>    8.160</td>\n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "  <th>Omnibus:</th>       <td> 3.486</td> <th>  Durbin-Watson:     </th> <td>   2.362</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Prob(Omnibus):</th> <td> 0.175</td> <th>  Jarque-Bera (JB):  </th> <td>   3.183</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Skew:</th>          <td>-0.228</td> <th>  Prob(JB):          </th> <td>   0.204</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Kurtosis:</th>      <td> 3.639</td> <th>  Cond. No.          </th> <td>    626.</td>\n",
       "</tr>\n",
       "</table><br/><br/>Warnings:<br/>[1] Standard Errors assume that the covariance matrix of the errors is correctly specified."
      ],
      "text/plain": [
       "<class 'statsmodels.iolib.summary.Summary'>\n",
       "\"\"\"\n",
       "                            OLS Regression Results                            \n",
       "==============================================================================\n",
       "Dep. Variable:                  Score   R-squared:                       0.478\n",
       "Model:                            OLS   Adj. R-squared:                  0.465\n",
       "Method:                 Least Squares   F-statistic:                     36.67\n",
       "Date:                Mon, 14 Oct 2019   Prob (F-statistic):           6.80e-17\n",
       "Time:                        16:42:40   Log-Likelihood:                -443.10\n",
       "No. Observations:                 124   AIC:                             894.2\n",
       "Df Residuals:                     120   BIC:                             905.5\n",
       "Df Model:                           3                                         \n",
       "Covariance Type:            nonrobust                                         \n",
       "====================================================================================\n",
       "                       coef    std err          t      P>|t|      [0.025      0.975]\n",
       "------------------------------------------------------------------------------------\n",
       "Intercept           97.7568      6.510     15.017      0.000      84.868     110.646\n",
       "Meeting             -1.6738      0.270     -6.188      0.000      -2.209      -1.138\n",
       "Reassign          -129.7491     18.454     -7.031      0.000    -166.287     -93.211\n",
       "Meeting:Reassign     6.6255      0.775      8.546      0.000       5.090       8.160\n",
       "==============================================================================\n",
       "Omnibus:                        3.486   Durbin-Watson:                   2.362\n",
       "Prob(Omnibus):                  0.175   Jarque-Bera (JB):                3.183\n",
       "Skew:                          -0.228   Prob(JB):                        0.204\n",
       "Kurtosis:                       3.639   Cond. No.                         626.\n",
       "==============================================================================\n",
       "\n",
       "Warnings:\n",
       "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
       "\"\"\""
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "LE2.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x396 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 残差のヒストグラムを描く\n",
    "plt.hist(LE2.resid, bins=25, color='cornflowerblue', rwidth=0.9);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'Residuals')"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x396 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 残差の分布を見る\n",
    "sns.scatterplot(x=LE2.fittedvalues, y=LE2.resid)\n",
    "plt.xlabel('Fitted Value')\n",
    "plt.ylabel('Residuals')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1c21bc19b0>"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x396 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 残差の密度関数分布を正規分布と比較\n",
    "x = np.linspace(LE2.resid.min() * 1.5, LE2.resid.max() * 1.5, 100) \n",
    "plt.plot(x, norm.pdf(x, loc=LE2.resid.mean(), scale=LE2.resid.std()))\n",
    "sns.distplot(LE2.resid)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x396 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# QQプロットの表示\n",
    "probplot(LE2.resid, dist=\"norm\", plot=plt);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAtEAAAH2CAYAAAC2iNAUAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjAsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+17YcXAAAgAElEQVR4nOzdd3xb9b0//tc5WpYsy3slTmzHiTNJyCCLQIBABpQCBUoLpaX03tvyoxdaOiCQMtIAoZRSZrm93w56uZfVMksKZaUJBLLIHk484njbkmXJ2tI5n98fjk2Ml2Rr2byefeRRIslHb9vH0Usfv8/7IwkhBIiIiIiIKGxyogsgIiIiIhptGKKJiIiIiCLEEE1EREREFCGGaCIiIiKiCDFEExERERFFiCGaiIiSGodIEVEyYogmIgrTvffeC7PZHNHHBINBfOtb30JqaioyMzPx7LPPQpIk7Nq1K0ZVxtef//xnSJLU649Op0NRURG+973vobW1teexJSUl+OEPfxj2sf1+P2699Va8/vrrsSidiGhEtIkugAgAFEXBY489hr/97W/w+/0455xzcM899yAjIwOhUAgbN27E3//+dyiKgssuuww///nPodfrE1020ZDefvtt/O///i82btyIJUuWIBQKJbqkmHj77beRnp4OAAiFQti3bx9+/vOf48CBA9i+fTskSYr4mE1NTXj88cdxzjnnRLtcIqIRY4impPDoo4/itddew8MPP4yMjAysXbsW99xzDx577DH85je/wdatW/G73/0OwWAQd9xxB3Q6HW6//fZEl000pPb2dgDAjTfeiNzcXGzevDmxBcXI/PnzkZOT0/P3ZcuWoaOjA+vWrcP27duxePHiBFZHRBR9bOeghHO5XHj22Wdx7733YunSpZgxYwbWrl2Lo0ePwuVy4fnnn8ftt9+OuXPnYuHChVi3bh1eeOEF+Hy+RJdOX3I33HADrrrqKjz22GMoLi6G0WjE+eefjyNHjvTcf8MNNwAA8vLyev77i8eYNWtWr9tee+01SJKEEydO9Nz27rvvYtGiRTAajSgqKsLdd98NRVF67i8pKcGvfvUr3HTTTcjKyoLFYsF3vvMddHZ29jxGURQ88MADKCsrg8lkwplnnonXXnut13M//vjjmDJlCgwGA2bOnIkXX3xx2F+fefPmAQBqa2v7vd9qteLf//3fUVRUBJPJhAsuuKCnzeXEiRMoLS0FAFx99dU477zzhl0HEVEsMERTwu3atQuyLGP58uU9ty1evBjvvPMOKisr4fF4cNZZZ/Xct3DhQng8np6gQpRI7733Hp599lk89thjeO6553D8+PGesPyLX/wC69atA9DV7vCLX/xiWM/x/vvvY82aNSgtLcWrr76Kn/3sZ3jkkUdwyy239HrcAw88ALvdjhdeeAEbNmzA888/jw0bNvTc/+Mf/xj33Xcfvvvd7+LNN9/EokWLcNVVV+Gjjz4CANx33334yU9+gm984xt48803cdFFF+Gb3/wmXn755WHVffz4cQDoCcOnc7lcWLp0Kd577z1s3LgRL774IoQQOPfcc3HgwAEUFhbilVde6fm8nn766WHVQEQUK2znoISrra1FYWEhNm/ejCeffBLt7e0455xzsHbtWrS0tMBkMiEtLa3n8WazGUajEc3NzQmsmqhLZ2cn3nrrLRQWFgIAGhoacOutt8Jms6GsrAxlZWUAPm93qKuri/g51q1bh8WLF+OFF14AAKxevRpZWVm44YYb8LOf/QwlJSUAgKKiIjz//POQJAkrV67E5s2bsWnTJjz00ENob2/H008/jXvuuacn2K9YsQIVFRXYsmULZs2ahY0bN+L222/HL3/5SwDAypUr0dnZiTvuuANXX331oDUqitLT7+10OvHxxx/j/vvvx9y5c7FgwYI+j//Tn/6EqqoqHDhwADNmzAAArFq1CuXl5bj33nvxt7/9DXPnzgUATJkypecxRETJgivRlHButxutra146qmncMcdd+CRRx7B4cOHcdttt8Hr9fZ7AaFer0cgEEhAtUS9FRcX9wRooCvIAl3ndTR4PB7s2LEDX/nKVxAKhXr+rF69Gqqq4sMPP+x57MKFC3tdwFdUVNRTx/bt26EoCi699NJex9+8eTPuvPNOfPrpp/D5fLjkkkt6Pc+aNWtQXV2NmpqaQessKCiATqeDTqdDdnY2vvrVr6K8vBwvvvgiZLnvS82WLVswc+bMXuFYr9fjiiuuwL/+9a9hfa2IiOKJK9GUcFqtFm63Gxs3bsS0adMAAPfffz++9rWvYfHixf2G5UAgAKPRGO9SifowmUy9/t4dGFVVjcrx7XY7VFXF2rVrsXbt2j73NzU1DVpLdx3dFzjm5eX1+zw2mw0AsHTp0n7vb2pq6rcto9t7773XM53DYDCgqKgImZmZAz7ebrcjPz+/z+35+flwOp0DfhwRUbJgiKaE635Rnzx5cs9t3b8CDwQC8Hg8cLlcPfN5XS4XvF5vvy/ARKONJEl9ArfL5er5b4vFAqCrpeOyyy7r8/Hjxo0L63m6A25bW1uvj9m7dy+EED33v/rqqz2r6aebOnXqoMefM2dOr+kcQ8nKysLRo0f73N7c3Izs7Oywj0NElChs56CE676C//Dhwz23dV+QtHLlSphMJuzevbvnvh07dsBkMvWsWhONZhaLBS0tLb2C9NatW3v+Oy0tDXPmzEFVVRUWLFjQ80ev12Pt2rVh91gvXLgQWq0Wf//733vd/v3vfx8PP/wwFi1aBJ1Oh9bW1l7Pc/DgQaxfvz7quwYuW7YMhw4d6nWBcCAQwKuvvoqzzz4bAKDRaKL6nERE0cSVaEq4iRMnYtWqVbjzzjvxy1/+EjqdDnfffTeWL1+OsrIyXH311Vi/fj0eeughCCGwYcMGXHvttTAYDIkunWjE1qxZg8cffxw333wzrrnmGnzwwQd9xs6tX78el19+OdLT03HFFVfAarVi3bp1kGUZZ5xxRljPk5eXhx/84AfYsGEDdDod5s+fj5dffhl79+7FU089hdzcXNxyyy34yU9+ArvdjoULF2Lv3r246667cNlll/WsiEfLd7/7Xfz2t7/FxRdfjA0bNiA9PR2PPvooWlpacNdddwH4fPX8vffew5QpUzBnzpyo1kBENBIM0ZQUNm7ciIceegjf//73oSgKVqxY0TMO7Kc//Sl8Ph9+8IMfQKPR4NJLL8WPf/zjBFdMFB2rV6/G/fffjyeffBLPPvssVqxYgWeffRZr1qzpecxXv/pVvP7661i/fj3+9Kc/wWKx4KKLLsLGjRv79EEP5re//S1ycnLw5JNPwmq1YtasWdi0aVPP9Ixf/epXyMvLw+9//3vcfffdKCwsxI9+9CPcc889Uf+809LSsGXLFvz0pz/FzTffjFAohCVLlmDLli09UzksFgtuv/12PPHEE9i2bRv2798f9TqIiIZLEtH+HR0RERER0RjHnmgiIiIioggxRBMRERERRYghmoiIiIgoQgzRREREREQRYogmIiIiIorQqBxx19bWmegShiUz0wS73ZPoMmiU4vlDI8Hzh4aL5w6NxGg/f3Jz0wa8jyvRcaTVcvctGj6ePzQSPH9ouHju0EiM5fNnVK5EE4UrpKhotLpR1+qC0x1ASBFQhYAsSTDoNci2GFA6zoJMswGSJCW6XCIiIholGKJpzAkEFew82oqqBgdsTh+8/hA0stRvSBZCQBVAmkmHgqxUnDEpC+UTMhioiYiIaFAM0TRmWDu8+PRwC6oaHfD5Qz1BWKsZuGtJkiRoJMDjC6G60YHj9R3ISTdi2sQMLJlVMOjHEhER0ZcXQzSNekIIfHSgCbuOtiEYUgBg2CvJGlmCvdOHbQebUFFnx0ULJqC4wBLNcomIiGgMYIimUa2tw4N/bK9DY5sLshy9FgxJktDu9OPlzVWYVZqNCxcUcVWaiIiIejBE06hV3ejEm9tq4A8oUQ3Qp1NVgX2VbWi1e/D1CyYjRc8fGSIiIuKIOxqljtd34PWPugJ0rEmShCabG8+/dxy+QCjmz0dERETJjyGaRp26Fhfe+qS2p/85HiRJQqvdg5c+qERIUeP2vERERJScGKJpVAmGVPxjey0CwfgF6G7dK9Lv766P+3MTERFRcmGIplHl3V0nYe/0Jez5JUnCgSobapudCauBiIiIEo8hmkaN6kYHDtXYE74RiioE3t1Zx7YOIiKiLzGGaBo1PtrfBCFEossAANicPnxyqDnRZRAREVGCMETTqFDX2okmmyfRZfSQJAnH6jqSJtQTERFRfDFE06iw55gVCe7i6KOtw4vKBkeiyyAiIqIEYIimpOcPKqhqTL4L+WRJwr5KW6LLICIiogRgiKakV1nvgD9JNzlptXvY0kFERPQlxBBNSa/F7onZtt4j1ekJwukOJLoMIiIiijOGaEp67c7EzYUemkB1U/K1mhAREVFsMURT0mvv9Ce6hAFJkgRrhzfRZRAREVGcMURT0kvEFt+RCIa46QoREdGXDUM0JT1FTe4L97hxIRER0ZcPQzQlPTnZBkR/QZKXR0RERDHAEE1JT6tJ7tM02esjIiKi6OOrPyU9S6o+0SUMSAiBdHPy1kdERESxwRBNSS/bkpLoEgakqEBxflqiyyAiIqI40ya6AABQFAXr1q1DTU0NNBoNHnzwQUycODHRZVGSyMlIgRACUhI2H5tStCjIMiW6DCIiIoqzpFiJ/vDDDwEAL7zwAm655RY8+OCDCa6IkknZuHQAyRegga5V8mTdTZGIiIhiJylWoi+88EKcd955AIDGxkbk5OQktiBKKplpBhTlmdHQ5kp0Kb0IITBlQnqiyyAiIqIESIqVaADQarW4/fbb8ctf/hKrVq1KdDmUZGaUZEJNsnnRphQd5pfnJboMIiIiSgBJCJFUyaStrQ1f//rX8dZbb8Fk6r/XNBRSoNVq4lwZJZKiCjz83C44XYFEl9Jj/rQ8XHnBlESXQURERAmQFO0cr732GlpaWvD9738fRqMRkiRBoxk4JNvtnjhWFz25uWloa+tMdBmj1tTxFny0vykpepD1OhmzijPi+v3k+UMjwfOHhovnDo3EaD9/cnMHnsCVFO0cK1euxOHDh3Hdddfhe9/7Hu68804YDIZEl0VJ5uwzCjE+15zoMqCoAoum5yMriUfvERERUWwlxUq0yWTCY489lugyKMlJkoTViybguX8eRzCkJKyOiflmLJ5ZkLDnJyIiosRLipVoonDlZpiweGY+EtXKbzRosXphcVLOrCYiIqL4YYimUWfJzALMn5oHNc5BWq+TcdmyEmSns42DiIjoyy4p2jmIIrVifhEAYHdFa1xWhVMMWly6tAQT8y0xfy4iIiJKfgzRNCpJkoQLF0yA2ajDp4eaEQipMXkeIQSy041Ys2hiUlzUSERERMmBIZpGtcUzCzB5fDr+sb0WDW3uqI6/kyUJZ0zOwYr5RdBq2PlEREREn2OIplEvJ8OIb62cik8ONWNfpQ0Ol3/EYbowJxXnzi5EcQHbN4iIiKgvhmgaEyRJwtJZhVg8owAHq204eKIddS0uACKsnmlVFTAatCgbn455U3MxLjs19kUTERHRqMUQTWOKLEuYPTkHsyfnwOn2o7LBAavDB5vDB6c7iJCiQBWARpag18nISktBtiUFBdkmTBpngY7byRMREVEYGKJpzLKkGjCvPC/RZRAREdEYxKuliIiIiIgixBBNRERERBQhhmgiIiIioggxRBMRERERRYghmoiIiIgoQgzRREREREQRYogmIiIiIooQQzQRERERUYQYoomIiIiIIsQQTUREREQUIYZoIiIiIqIIMUQTEREREUWIIZqIiIiIKEIM0UREREREEWKIJiIiIiKKEEM0EREREVGEGKKJiIiIiCLEEE1EREREFCFtogug5HfgwD7cdNP3et1mNBrx7rtbe/4uhMBPf3orli49G1deeU28SyQiIiKKK4ZoGtKJEzUoK5uM3/zmyZ7bZPnzX2Koqorf/vZhbN++DUuXnp2IEomIiIjiiiGahlRTU4WSkknIzs7pc19DQz3uv/9etLa2wGxOS0B1w7Np05t44IH7+r3vySd/j4KCcfjNbx7Cvn2fIT09A9deez0uv/yqOFdJREREyYohmoZUU1ONM86Y0+99hw4dxKRJk7Fx4yO48cZvxbmy4Vux4iIsWrSk120PPLAeLlcnZs48A9dffw2ys7PxxBP/BafTeSpwS7j88isTUzARERElFYZoGlJNTTWMRhO+851vwul04Mwz5+GHP/wRsrNzsHLlaqxcuTrRJUbMYEiBwZDS8/ctWzZjz55d+J//eQnbtn2ElpYmPP30fyMrKxsAcNNN/4nf/e4JhmgiIiICwOkcNASXywWrtQ2hUAh33LEOv/jFejQ1NeInP7kFoVAo0eVFRSgUwjPPPIGrr/4mxo8vQkNDHcaNG98ToAFgypSpaG1tQXNzcwIrJSIiomTBlWgalNlsxjvvbEZKihEajQYA8MADD+Pyy9dgz57dOOusRQmucOQ2b34fra0tuO66bwMAMjOzYLfbEQqFoNV2/Yi0trYAADo62lFQUJCwWomIiCg5cCWahpSaau4J0ACQlZUNiyUdbW2tCawqel599a9Ys+ZSWCzpAIDFi8+Gqqp48snfwufzobW1BX/4w38BAILBYCJLJSIioiTBEE2DOnjwAC666NxebQzNzc1wODpQXFyawMqiw2ptw759e7BmzSU9t2VmZuL++3+FLVs+xMqV5+KGG67FJZd8FQCQmpqaqFKJiIgoibCdgwZVXj4VOTk5ePDB9fjP//wxAgE/fvvbX2Pu3PmYOXNWossbsU8++Rh5efmYPn1mr9vnzz8Lr7zyFqxWKzIyMlBXdxIajQYFBYUJqpSIiIiSCVeiaVB6vR6//vXjMJmM+OEP/wO33fZDTJgwERs2PJTo0qLi4MH9mDNnLiRJ6rmttvYEbr753+H3+5GTkwOtVott27Zi2rQZMJm4Ek1ERERciaYwjB9fhAcffGTIx/31r2/GoZroqqqqxPLl5/e6rbBwHJqaGvHEE4/i2muvx5Ejh/DnP/8/rF+/MUFVEhERUbLhSjR9qbW325CWZul1m16vx8aNj6Cy8hi+/e1r8Kc//TfuuONuLFnCLc2JiIioC1ei6UvtlVfe6vf28vJpeOaZP8a5GiIiIhotuBJNRERERBQhrkRT1PiDCirrHWhp98Dj79rN0KDXIDc9BVOK0pFq1Ce4QiIiIqLoYIimEWtu92DHkRZUNzrh84cgy1Kv+4UQeHeXjNJCC+aV52DSuPQEVUpEREQUHQzRNGyqEPjX3kZ8dqwNiqICQJ8ADQCSJEEIgepGB6qbHJhRnIlVCydCp9X0eSwRERHRaMAQTcOiqCr+urkKNU1OyFLf4DwgARyqaUeL3YtvrJiC1BRd7IqMEUVVUd/qRl1rJ9o7/fD5FQgIaGQJaSY9ctJTUDY+HRlmQ6JLJSIiohhhiKaICSHw2tYanIg0QJ8iSRKsHV68+EElvr1qKrSa0XF9q8Ptx47Draisd6DD5YNmgLqFEHhvVz3G55oxsyQTsydnQyOPjs+RiIiIwsMQTRHbVdGG4/UdwwrQ3SRJQpvdg/d21WH1ouIoVhd9iqpi895G7DlmhaIokCRpwAANoGf3w0arC/WtndhZ0YoL5xexF5yIiGgM4fIYRcQfUPDJwaYRBehukiThQHU7Gq3uKFQWGy3tHvzl7QrsPNICVVV7bQ8eDlmW0NHpx9/+VY23t9dCVUWMKiUiIqJ4YoimiOw40gLvqfF10SCEwO6K1qgdL5rq21x46cNKtHV4R/ymQQiBvZVW/HVzJRRVjVKFRERElCgM0RSRygZHxKuxQ6lqdCKkJFewbLF78NrW6qi+YZAlCTVNTry2tQZCcEWaiIhoNGOIprAFQwpsTl/Uj+vxBZOqpUNVBf7xaS08vugF6G6SJOF4vQM7jrRE/dhEREQUPwzRFLaTLS4EQ0rUj6uRJdS3uqJ+3OHaur8Rze2emB1floBPD7WgPQZvSIiIiCg+GKIpbJ2eYFQuKPwiSZIQCCVHO4fbF8TuiraYfJ6n8wcVbN7TENPnICIiothhiKawxXKcc2wja/i2H26JyWp7f2qaOuH2BuLyXERERBRdDNEUtoIcE9QYXBAnhEBaauJ3LhRC4Hh9R9QvnBxISFGw/UhyTiYhIiKiwTFEU9hyLEaYjfqoH1dAQlkSbETSavei3emP2/NJkpRUF1QSERFR+BK+Y2EwGMSdd96JhoYGBAIB3HTTTVixYkWiy6J+SJKEwmwTqhudUT1ubnoKLKnRD+eRqml2Qo5zX4m90w8hRNxWv4mIiCg6Eh6i33jjDWRkZODhhx+G3W7HFVdcwRCdxGaX5eB4vQOaCNOmABAMqvAFQujetE+nlWHQySifkJEUIbLd6Y97HZ3eAGxOH3LSjXF9XiIiIhqZhIfo1atXY9WqVT1/12g0CayGBuL1d02tcLoDCCkqOjqDkGUJaSY9tJqBg2dQUeF0B+D1Kz079UmnLiMUENBqZJxo6UR6lRVnTMpOaJgOBOJzQeHpJEhwuAIM0URERKNMwkN0amoqAMDlcuGWW27Bj370oyE/JjPTBK12dIbt3Ny0RJcQkZpGBz492ISKWjv8AQWSJEGn08Ib8EEIAbcvBKNBi7RUPVJTTr84UMDe6YfD5YeqCgASZKl3C74EIDfdBKvDj3d31aOq2YWvnT85YYHSaNJBp4vvj4SsCqSlG8M+L0bb+UPJhecPDRfPHRqJsXr+JDxEA0BTUxNuvvlmXHvttbj00kuHfLzdHruNMGIpNzcNbW2diS4jLEIIbNnXiO2H++6sJwPIMOvR7uzq53V7g3B5AzAbdciypAAArB1eePyhnlXnroaO044PAYtJD71OQjDYtTPg8dp2PP78Z7hsWSkm5sf/B87vD/XUEi+KosLn9od1Xoym84eSD88fGi6eOzQSo/38GewNQMKnc1itVtx444342c9+hquuuirR5dAp7++uxyeHmge8vyswG9DdfSFBgssbhLXDB5vD94UA3ZsEID1Vj8w0Q5/7vP4QXv+oBlaHNxqfRkTSTPG/uFGn0yA/yxT35yUiIqKRSXiIfuaZZ+B0OvH000/j+uuvx/XXXw+fj9shJ9Jnx9rw2bGhd+0zG3UoyDLCoJchIE4F6QAc7kC/AVoIAZ1WQm6mERnmvgG6m9cfwtvbT0LEYCb1YAqzTadaT+InPVUPg250tiYRERF9mSW8nWPdunVYt25dosugU4QQ2F0R/gYgOq0G+Zkm+AMKnN4AHC4FqipO7W4oQZIArUaGXisjNUULo0Eb1sWDda0ufHbcivnlucP/ZCJUWmCBTitDiWOQzklPidtzERERUfQkPERTcqk42QGbwwc5whF2Br0GRkULjzYEoQqkGXVIM+mh0UjQyJH/wkMjSzhU3R7XEG3Qa1BcYEF1oyMuz6eoAjNKs+LyXERERBRdCW/noOSyv8oWcYDu5j3VBy3LMgIhAb1OM6wA3a3R5obTHb8dBAFgzuTsuK1E56YbUV6UEZfnIiIiouhiiKZeWkYw+SQYUnv+OxBSR9xfLEGgqiG6uyMOZfL4dIzPNcf8eVQhcGZ5TlJsMkNERESRY4imHqoqEAiqQz+wHwLotYIrhIA6wgsDJUmC3RXflWhJkrDyrAnQamP7o1FSYIlrqwoRERFFF0M09TbMhVEhvjgJOjq+uJothIDV4UV1owO1LZ3w+qM/17kgy4SF0/IQq64Oo0GD1YsmcBWaiIhoFOOFhWOUEALH6jpworkTgaAKrUZCQZYJsydnD9inLMsS9FoZXiXy1WhJ6srfoufv0pAj8sKh12mgqCr2V9lwtNaOtg4f3L5g1zMJQKORkZFmQFGuGQum5iEvMzq7HZ59RiEc7gAOjKBHvD96nYyvLC1FhplTOYiIiEYzhugxRgiBHUdacaDKhjaHF5rTAuC+Sis+OdSM8gkZOH/e+H7DdGG2CdWNkfchSwC0WhnBU+0gep084vCpqAKKquIPbx1B+2kTQ7o+p8+P3dHpR0enH4eq2zGtOAOrF02EboTbwkuShIsXF0Mry9hbacVI3w8IIWA06PDVZSUoKbCM7GBERESUcAzRY4gQAps+rcWBahtkSeoVoIGulWaXN4jdFa1otnlwzYop0H2h93d2WQ6O1zv6fGw49KdCtBACqYaRn1oebxA7D7cCEsIK5AICh0+0o9HmwZXnliInY2Q7AUqShFWLJqIoz4zNexvg8gSG1YIhhEBJgQVrFk+EJXXgTWaIiIho9GBP9Bjy4Z6GrhF1QwQ9SZLQYHXh1S3Vfe6bUpSOvIzhtUSYjToAgEYjIc2kG9YxunW4/PAGQxH3aEuSBIfLj5c2V8HeGZ2dL2eWZuHGi6dh1qQcpOi1Ye+kqKoCuRlGXLRwIr5+wWQGaCIiojGEK9FjhD+gYH+VLewVZEmSUN3kRG1zJ4oL0nrdvmBaHv65sy7ibbcNOg30ehkpOs2ILprzBhQ43cER9Te7PEG89UktrruoPCoX8BkNOlyypBjBkIpdFa2ob3Oh3eFHh8uPkKJCkrourjSlaJGZloJsiwEzS7NRWpjGCwiJiIjGIIboMWLHkRb4A6GIApssAXuOt/UK0QAwZ3IOOlx+fHqoOaLjqarA0hkFqG9zIRAa/qi8dqcPqUYtUvQj62uua3Vh59FWLJyeP6LjnE6nlbFkZkHP3z2+ENy+IIIhFQZd10WOI9lghoiIiEYHvtqPESeanMNa8axt7ux3nvPyM8fjnDnjwj6mCoH50/Jw5XllWLGgaNirry5PALIEZKWNvPVBI0s4UG2LeEU9EqYULXIzjBiXk4rsdCMDNBER0ZcEV6LHCG9AGdbH+YMK/AEFxn4uBFw6qxCTx6dj59FWVDU4urb1Pi0cC9G1tfekcemYOyUHE/O7VrTPmJQDVQU+2FPfM60jHIoqoNPKyM8yRa0Foq3DixNNnSgdx4kYREREFD0M0WPEcDOnJEmDTr7IyzThkiUlCIZUfHasFQ5PEKGQAq1GRmqKDvPKc2A09L2IcM7kHIzPScU7O+tQ1+rCYK3aXePftFg4PR+7K7OGk2gAACAASURBVFrh9kVvAxVZklDbyhBNRERE0cUQPUaYjTrYOyPfIttk0EIfxhbXOq2MRTMKhnzc6XIyjLjuonLUNndiX6UVbR1e2F1+hBQBCYDZqEWWJQUlBWlYMC0P/qCKzXsbBw3cw9ExjK8LERER0WAYoseIaRMzUdvcGdEGJ0IITBpvifn0iOKCtJ6LF4MhFf6gAo0sIUXfe4pHhysARVEhhxHqI+GPoKWEiIiIKBy8CmqMmD05G5ZUfUQfI8sSFk7Li1FF/dNpZZiNOhgN2j7hfTgbvIRDw7OciIiIoozxYozQyDKWzS4MuzdaUQXmTslFdvrwZzFHmyVVD1NK9H85kmaK7M0FERER0VAYoseQ2WU5OG/ueMhDjFkToitAr5hfFKfKwqPVyMhMS4nqMYUQyE6P7jGJiIiI2BM9xpw1LR8FmSbsPtaGmkYnAiEFkiRBCAFZljAxPw2zJ2VjeklWokvt18R8M1ra3VHr0zbotDhjUnZUjkVERETUjSF6DJqQn4YJ+Wnw+II4erIDgaACjUZG2TgLsizJvSq7aHoe9hxvg6KMfIMUIQTKxltg0I1s50MiIiKiL2KIHsNMKTrMK89NdBkRSTXqceaUXOw80gJ5hKvRphQtlp85LkqVxZ7D5UdFXQfanT54Ts3KNug1yEwzYFJhOnJz04Y4AhEREcULQzQlnfPPHI+6FteI2joEgGVnjIMldeTbh8daxUk79hy3oq61E6oq+nzOQghs2duISQeaUFaYhnnluTEfS0hERESDY4impCPLEq46rwwvvH8cNoc34sAoAJw9qwBzk3wV3usP4h/bT+J4naNnqkp/n6skSZAkoKHNjar6DhyptWP1oonISaLJKkRERF82nM5BScls1OG6i6ZgclEGwu2OFkLAZNBi9cKJWDY7uds42p0+/OWdYzhe1xHRlu0aWUKj1Y3/e/c4qhudsSuQiIiIBsWVaBpUMKRid0Urqhqc8AaCAACTQYcpRemYW54LbQx3MjEadLhyeRkO1bRj7/E21FvdgOjb7qAKAZNBh7Lx6Vg+pxDmJJ8L7fYF8dKHlXC6A8Nuy/AFQvj7thp8bXkZinLNUa6QiIiIhsIQTf0SQmDr/ibsPW6F1x/sCXtCCDQE3NhzvA1//VcV8jKMKB1nweRx6ZhVlj3iiwH7M7M0CzNLs9DW4UFlgxN2pw++gAJZlpBm0iEvw4hpxVnQRXm78Fj5x6e1cLj8I+5r9gUUvLPjJG5YMw2aIWaDExERUXQxRFMfQgj8Y/tJ7K+yQZY+79N1e4NwegLw+kMAJEACnO4g2jp8qDjZgU8ONWPWpGwsnVUQkwvfcjNMyM0wRf248XSw2oaqBkfUvj7WDi/+tbcRF8xLro1ziIiIxjouX1Efnx5qwf4qK+TTcl6704vmdg9c3hBCiuj6E1IRDCmwOX2wdnhhdwWwdX8T/r6tFkKMfM7zWLT3uDWqbzAkScLhE+0IKWrUjklERERDY4imXlQhsK/K2qstw+bwwurwQ1EFurJx930SAAlCAE5PEC3tHgghcLDGhg/3NCSg+uTWZHOjweqO+nFd3iD2Hm+L+nGJiIhoYAzR1MvBahvsnf6ev3v9QbS7AugOzAMRQsAXCKHV3jWSbl+lFR5fMPYFDyKkqHD7gvAFQkmxMl7Z4IhoEke4ZElCo80T/QNT0tu06U0sW7ag3z97937W87hgMIhLL70UH3+8NYHVEhGNLeyJpl6qGpzQyJ9fRNjm8EFV+k7E6EuCqgKBkIIOlx+ZZj12Hm3F8jPHx77o0wRDCnYebUXFyQ7YnD4oigpJkqDXaVCcn4a5U3JQUmiJa03dbE7fqDw2Ja8VKy7CokVLet32wAPr4XJ1Ytas2QAAv9+P++5bh2PHjiWiRCKiMYsheowJhlTsOtqCY/UOdHoCkABkpKVg+sQMzJmSM+QUB39Q6flvrz+EQFANu4dXQECCDK8/hEyzAVUNzriFaCEEPtzTgAPVNvj8oZ6au/8/EFRwvL4DFXUdKMgy4oJ5RZiYH99ttD3eUAyPndhVf0oMgyEFBkNKz9+3bNmMPXt24X/+5yVotVpUVBzF/fffA1nWJLBKIqKxiSF6DKlvc+H1j2rg8vSeP+z2uVDX0onPjlvxtXMnIcuSMuAxTs/LLl8o7I1OTtfdRpHqi8/ppQqB17ZU41hdB2RZGjT0yxLQavfilS3VWLNoIqZOzIxLjd11xkoSdKtQgoVCITzzzBO4+upvYvz4rmktu3fvxNlnn4sbbvgeVqxYluAKiYjGFoboMcLq8OLVLdXwnrYKezpZltDu9OGvm6tw/apyGA26fo+Tavz89lBIRSQpumf1FxICIRVKHIKdEAKbPqnFsfquAB2uQFDB29tPwmjQxm1FWq+L3WqgLobHptFh8+b30draguuu+3bPbddee30CKyIiGtt4YeEY8dH+plPzmwdn7/Thk4MtA95/ZlkO1FPJWQARb0ndTQiBFH3sT6+aJicOVtuGtcmLP6jgw8/iN0UkM80Qs2NnxfDYNDq8+upfsWbNpbBY0hNdChHRlwJD9Bjg8QVR0+QM67GSJOFYfceArQVFeWaMy+raRlqWJMiSFNZkC1lGryArSxKK47DCu7fSGtEK9Bc1t3tQ29wZxYoGVphlgqpGf3leCIHs9IFbdGjss1rbsG/fHqxZc0miSyEi+tJgiB4Djtc7EDjtgsChtDv9sA8yzeGcOYXQamTotTI0soQhd5SWAK3m8wcJACaDFgum5YVd03C4vQHUNIb35mEgkgTsidOM5eklmbCk6qN+XFmWMHdKTtSPS6PHJ598jLy8fEyfPjPRpRARfWkwRI8BwVCku9UJ+IMDf0xpoQWrFk5ApsUASQI0kjzwarQE6DRyr1VonUbC9OJM5KQbI6wrMnsrbVHZqa+2pTMuc6Q1sozyCRlRf67i/DRkpnEl+svs4MH9mDNnblR3wyQiosGFHaKFEOjs/PzX3i+//DIeeeQRHD16NCaFUfgy0wwRTX7QauQhV0RnlmbjugunIj8rFbKmq9/58/AnAKmrhaN7tbqbKlSMy0nFZeeUDudTiYhvgIsoIxUMKsN4IzI8y88chwxz9PqXdVoNLpgX31nclHyqqioxaVJZossgIvpSCStEHzhwACUlJXjooYcAAPfeey+uueYarF27FnPnzsUHH3wQ0yJpcJPGWQYdW/dFRblmmI39T+c43YR8M267Zg5mlmQhy6KHyaiFRiNBI0sw6GTotZpeK9BCCEzIS8OPrp4NnTb20yLCb2AZnCoAJQa9yv3RaTW4YH5RVMK/EAKLZuQhJ8MUhcpoNGtvtyEtLTGbCBERfVmFFaJ//vOfIy8vD9deey3cbjcefvhh3HjjjfD5fPj617+Ou+66K9Z10iAkScK0iZlhrUYLITBrUlbYx05N0eE7q6dhRkk2xmWnojjfjCxLSlcLh9zVU6yRgVSjFhfML8Id182DXhefyYkGTXS6kbQaCQZ9/EbETSnKwPnzxg9rokg3VQjMmZyLJTMLoljZyKiqiOksbBrYK6+8hcsvv3LQx1RUVODss8+JU0VERGNfWGln27ZtePHFFzFjxgy88sor8Pl8+Ld/+zfIsowbbrgBX/nKV2JdJw3hnNnj0GTzoLbZOeAqp6IKzC7LxhmTsiM6ttmkx/WrpqKywYF9lVbUtboQDKmQJCA1RY+y8RYsnJ4X977cCflmfHK4eURhFACyLcYRHyNSC6bmwaDTYPOeBnh8wYhWprUaGQun5+HsMwoT2gPr9gaw40grmts9aHf64Qt0/W7AlKJFlsWAcdmpWDg9P65vUIiIiOIlrBAtyzK02q6Hbtq0CVlZWVi8eDEAoL29HWazOXYVUlhkWcLV55fhnR11OFbXAX/g835hVQikpugwuywb584ZN6zgJUkSphRlYEpRBlQh4A8o0MgSdFo5YUGutNCC/Awj2hwDTxoZiioEppfEb9fC050xKRulBWl4Z2cdapqcUJSht1gvyjXjgvlFKMhKXAtHMKTinztPouKk/dSbqd41d3oC6PQEcKLJid3H2jCrNAvnzxs/5JbzREREo0lYIXrJkiV4+OGHYbPZ8OKLL+Ib3/gGAGDv3r247777sGwZt5NNBhpZxsWLi7FifhF2HW2F0x0AAORkpGDulNxeY+hGQpYkGA2J3+xSkiRMLc5E677GYQf51BQd5pXnRrmy8JlNely5vAxOdwA7j7ai1d61qusNhAAB6PUaZJkNyE5PwQWLimFI8PCFRpsLb3x0Ag6XH5I0+BbrkiQhEFSwq6IVJ1s6cfk5g285T0RENJpIIox5W8eOHcNll12GiooKlJeX4/3338f48eMxbtw4pKenY9OmTSgtjf00hm5tbfHZHCPacnPTRm3tySoYUvDHt47C4fZH/LGqAJbOyse5c5JruoUQAooqIERXv3Z3UE30+dNs8+DlzZVh7YzZnzSTHt9YMZnj+BLk9POn0xNAVYMDDncAQgAajYTcDCNKCyxsv6E+Ev1vD41uo/38yc0deOO4sJYTy8vLceTIEVitVmRnZ/e8qL/zzjuYMWMGNBr+o0uJodNqcMW5pXjpwyp4fMGwP05VBaaXZOGc2eNGXIMQAhUnO3D4RDs8/hBCioBeKyMnPQVnDaNXXJIkaDXJNe83pKh4Y1vNsAM00BXc3txWi+tXlnOecQK4vUF8uKcelfUO2Jx+SBC9vg+qKqDRyCjIMmF6cQbOjOJvr4iIxqKwVqKTzWh9RzPa340lsza7B69srUFHp2/IgCYEMLssG6sXTRxRmBNC4JNDzThU0w6b09fn4kQhBGRZRklBGhbPKMCE/JFdO5DI8+ft7bXYV2kdcfhVVYFlswuxLApvXig8Qgh8eqgZe6ra0eka+ucD6LpWICPNgBXzijClKCMOVVIy42sXjcRoP3+GtRKdlpYW9gumJElwOByRV0YUJbmZJtx48XTsqmhFxUk7mm0eyHLvGdY6nQaTCi2YOyUXxQUD/1CEQ1FVvLa1BsfrOiDLUr/TPSSpa4OamiYn6q1urFxQhJmlkU1GSQb+gIIjtR1RWT2WZQn7q9qxZFYBLzSMA18ghFe2VKO2uRMpBl3Y30NZkuB0BfDq1mrMKcvByrMm8LcHRERfMGCI/slPfsJ/NGlU0WllLJlZgCUzC3Ci2YmTLS4EggpkWYLJoMXssmyYUobeZGYoQgi88fEJVNZ39ArqAz4egL3Th2ffPorxuakwp+ih12mQZTFgwdRc5GeljrimWNp5tAWBYHR2hwQAp9uPg9XtmDM5JyrHo/75AiE8/95xtNo9vXYVjYgA9hxrQyCo4CtLS/iaQER0mgFD9L333hvHMoiiq6TAgpKC2Ozgtq/ShoqTHUPOlhYAnK4AXL4gQooKCRKqGzoxLscESZLQZHPjYE07JuSasWhmPsrGpcek3pFqbvdENTzJsoT6NhdDdAwJIfD6RzVotY/8eyfLEg7WtCPDbMA5c9iGQ0TULew5ZTabDdu3b4ff70d3G7UQAh6PB5988gmefvrpmBVJlEwOnbBhqIU9AcDa4YPHH4R06n9A1wV6Lm8IaaauFXEJQH2bC81b3bhg/gScmYTBsr0z8sknQ7HH4Jj0uc+Ot+FE08AbL0VKI0vYWdGK8gkZyE/gjHIiomQSVoh+9dVXcd1118Hn+/yiFCE+v7K7vLw8dhUSJZGWdjfq29wYLJoIADbH5wH6dJIkwe0L9oTobiFF4L1ddUjRaTCtODGbvwzEf2onwmjyBoY/5YMGF1JUfHKwOeqtF6GQig/3NOAbK6ZE9bhERKNVWCF6/fr1mDdvHp588kk89dRT8Pv9WLt2Lf7xj3/grrvuwqOPPhrrOomSwsHq9kEDNAB4fSG4fX0DdDd/UIWiij59qqoq8P5n9ZgyIT2uF90JIVDZ4MD+Khva7F4EFBUyAINegwl5aQiGoh+i2VkbO7srWuHyBmOylX1diwtWhxc56caoH5uIaLQJK0QfPXoUL730Es4880xccMEFePDBBzF9+nRMnz4ddrsdGzZswOrVq2NdK1HC+YK9A6UquiZ1CNE10UCjkeDyDhyggVObqSgqNHLf+eqd7gD2Hbdi3tS8qNfen72VVuyuaEVbh7dP6PL4Q2h3+tDc7oUsS8gwG2DQRSfcm436qByH+jpW54hJgAYAAYHdFW1YtXBiTI5PRDSahPWKqNPpkJbWNRKsvLwcFRUVCAa7NrZYsWIFjhw5ErsKiZJI91T1QEiFzelDo9WNxjY3Gq1uNFhdaGhzweUNQmB449dlWcLhWnsUKx7Y5j0N+OeOOtgcfWdcd+ve+MUfUNDW4R3RZiuny0wzROU41JuiqrA5vDF9jlgfn4hotAgrRM+bNw+vvPIKAGD69OkQQmDr1q0AgJMnT0alkH379uH666+PyrGIYkWvk9Hu9KPZ5oHbG4Kqdl0bIEtdFw8GgipCigp/UEVIVfs9hgQMOnKsyeqBL8Y9w9sONmH74WYgjLCfotdCCAFVFbA5fX1W4yOlCoEpRck5iWS0a7Z54InSG52B2JyfX1xORPRlFlY7x9q1a3HJJZegtbUVL7zwAr7+9a/j2muvxerVq/HGG2/goosuGlER//3f/4033ngDRiP77Ch5CSHQZPOg0xMY8KKtrmghAQIIhQSgVaH9Qn+zXqcZdL50QFHg9StI0Yc9PCcirXYPPj7QFPaFZ2ajFk63DEUVUFXA7vSjINs07L7m/AwjSgtjM37wy67F7h3+TOgweQMKAkEVBn3fdiQiig8hBH7601uxdOnZuPLKa/rc/+CD65GRkYmbbvrPnttCoRD+/Of/h02b3oTH48GyZefi1lt/2tNpQJELayV61apV2LZtW0/f8+9//3tcfPHF2LFjBy699FI89dRTIypi4sSJeOKJJ0Z0DEqMYEjBJweb8PdtJ/D3bSewdV8jvP5gosuKiZ1HW9FodUGvCz88hEIC6mmrdkIImAzaQQOsBCCWe1rsqmhDJAuJkiTBbNL1rD4GQsqw2zqEABbNLOCmHTGixmOFWIj4PA8R9UtVVTz66K+wffu2fu//y1/+iLfeeqPP7X/4w3/hr399ET/84Y/xzDN/hM/nwx133Bbrcse0sJe6Fi5ciIULFwIATCYT/vjHP0atiFWrVqG+vj7sx2dmmqDVjs5VkMH2YB9NhBD4x7YT2H20BR5fqNfow/0n2jG7LAeXLZ8c81WxeOmaYHEcKQY90kwq2p2+foOgLElQIHqNn1BUAY226/2qRiMhK9046Eq03gBMGJeBFEPfH8+Rnj/BkIK6Njd0ushWubPTNQiEVHh9IciSDK9fQbo5JaJjCCEwe0oOlp9VHNHHUfgK7V5otdoBz69Iv+/90WokFBakQ6fltu1fJmPltWu0q6urw+23347m5mZYLBaYzSk93xubzYa77roLe/bsQUFBAUwmfc99Qgj87W8v4rbbbsM111wBAJg589c455xzcOLEUZx11lkxrXusnj9h/Yv6l7/8ZcjHfPvb3x5xMeGy2z1xe65oys1NQ1tbZ6LLGDEhBDZ9WosDVbZ+X6xDIQUf72tAs9WFq5aX9YRNlzeIdqcPep2MnHQjtJrR8yJ8vL4DJ1uckCUJqSkadHplBAJKnyAtS+gK0Kct1CmqgCq6+qMzjAYoigJlkLbi8bmp6HR68cUzJRrnz66KVnQ4vcNaCc4y69EaVBAIqXD7gvD7g2Ftew50nTNFeWacN7twTPwMJKtMoxahkNLvbzJ0Oi2CwZH3S6cZU9Bhd4/4ODR6jJXXrrFgy5ZPMWFCKX75y1/hxhu/BZfL1/O9+eyz/dBo9PjTn/4Pd9+9Fh5PoOc+q9UKj8eDkpKpvb6X48YVYcuWbSgpmRazmkf7+TPYG4CwQvQNN9ww4H0pKSkwm81xDdGUWEdPduBAdf8BupssSaiqd2Dn0VaYjTrsr7KhtrkTqqpCALCk6lE2Lh1LZhUgw5z8kxqOnzY2TJIk5KWnoLXDi0BQ7RVIuy4y7Bp91+PUSLssSwrSUgcf7aaqArNKs2PxKQAAXJ7gsFspNBoZeZlG2Jx+eHxBKKoIO0RPL87CxUuKR9Ubp9HIlKJDhtkAhzt2O0JmWZL/55VorFq5cjVWrux/pPC8eQswb96Cfu/LyMiARqOB1doKYDqArrYQm60Ndnt8JkKNRWGF6P6+wC6XC1u2bMGdd96J5557LuqFUfI6UGUNaw6tLEt4d2cdtFq5p89XcypEeXwhHKi2obrRicvOKUVRrjnGVY+M/wsreBqNjPwsE2wOP3yBz6d0AF2TN1RVBSBBCAFZ/jzcDCXdrMcZk2IXokNK/xNDwtUdpB0uGZZUPdy+gTf1EAIozE7Fwul5SbcL41hWlJcKR01sQrSqCkwq5GQVotFGq9XivPMuwO9//zRKS8uQm5uHZ5/9Azo6OnpGFlPkwgrR6el9/9FMT0/HN7/5TXg8Htx6663YtWvXiAopKirCSy+9NKJjUOwFQypOtrrCeqzTFUC7y4/xuanQDbAC6fYF8frWGnz34mkwpej6fQwANNrc2HPMCrc3CEhARqoeZ03PQ2ZaZH25wzVQ/3NuRgoURYXTE4QvoEBRVUiSDHFqExadRoYsS2FN2tBoJFw4f0LYq7vDEa0+VlOKDt9aWQ57px81TU60O33w+pVT92mRnZ6CKePTUZQ3NvvgktlZ0/JwsGbonTWHI91swOyy2L3JI6LY+fGPb8eGDXfjG9+4AhqNBhdeuAqLFi1BampqoksbtUZ8lUlxcTEOHToUjVpoFPAHFYQUgaFynhACnd4gJHT1BA820MLlDWD7kRacP7eoz30Otx9/31aL+lZXrz5PIQT2VVlROs6CS5eUwBCjcXDdUvu5yK+bRiP32TxEiK6Zym5vCJC6LsYajE4r48IFRZgyISMq9Q4kL8PY75bjkUoz6ZFm0iMzLQWTxnFlMpnkZZpQNs6CqgZHVKegqKrA7LLsmL7JI6LYycjIwK9//Thcrq6FMLPZjBtvvA5nn31ugisbvcJalmpvb+/zx2q1Yv/+/bj//vtRVlYW6zopSei0MsJZzPT4QggpXf3PmiFeyCVJwvF6R58NHJweP55/9zga2lx9LpSSJAlCAFX1Dvzfe8cRDI2sTWEoZ5bnDrqV9xdJkoRsSwoyzAYYDVoYDX3fRXR/vhPz0/C15ZNwxqScqNU7kGnFmci2jHz1vmx8Ovubk9jqhRMH/c3OcIzPNWPprIKoHpOI4mf9+l/go4/+BbPZDLPZjJaWZlRVVWL+/NhO5hjLwlq+y8nJGXBFw2Aw4OWXX45qUZS8DDoNCnPMaGgbvKUjEOq64E6nlaENI3U7XX6EFBW600YXvv1pHRxu/+AzlSUJrXYP3t9dh9WLYjc6LTfDiKK8VNSF2crSXZslVYczpxSipNCCQzXtcPtCUBQVeq0GWRYDzpqWh+z0yDcZUlQV+45bYXX4ur5uOg3KxllQWmgZ8us1ZUI6dhxuGfYqpQBw1rTcYX0sxYfZpMeK+UXY9GktVHXkM51TU3S4eHEx53sTjWIZGRl45pknkZubD1mWsHHjBlx44UoUFU1IdGmjVlgh+o9//GOffzwlSYLFYsH555/fb880jV0zSjJR19I55K91BU5tLBLGMQWkXhMt7J0+1LZ0hvWiLUkSjjc4cKGixnR1dOGMfDRYPacuGgxPqkmPpbMKkWbSY+qEkV9cFwyp+HBPPSrrHXC6A72+B59VtCIv04TZZdmYPzVvwGMsmp6P/VU2+AORb98thEBxgQV5maZh1U/xM6MkC4GQivd31UEZQZBONepwxbmlyE6Pz/UHRBQb//EfN8Pj8eBHP/r/oNHIuPDCVbjpplsSXdaoJokv/g59FBit8wbjPStRCIGapk4crLbB4w/BoNNgyoR0zCjJCmu6xmDH/evmKlQ3Dtxz6XQH4PYFkZ+VOmT/NACYjXrcdPnMnuN98Fk9dh4Jf7VUFQKrFk7E3CmxXSH97FgbPvisPqzVvRSDFpcvK8HE/OhscZ2aloKnXtqLlnb3oF8XIQRml+Vg9aKJAz7ueF0H3thWA0WJ7Mc/w2zAt1aWR71VgGKnpsmJd3fVweUNIRQK/42TKoCSgjSsWjghbhfwUnIa7XN+KbFG+/kzrDnRv/nNbyJ6kttu49aRyURRVbyypRpV9Y5eq5VHT9qx62gbrlxeijTT8Oa9SpKEry2fhE2f1KLiZMepiRSnzUqGhNll2Wi2e+ELY3toIQTKxvduQzh9F8RwyJKETk/sx/TMK8+FQSfj4wPNsDl9/V6g1zXazYSVCyeiICs6K7aqEHj2rcNDBmig6/uzr8oGo0GL8+aO7/cxUyZk4JIlJXh7+0kEgkMHKyEEcjJMuOq8SQzQo0xpoQU3Xjwduytt2H24uc9vML5IFQJ5GUbMmZyDeeW5bOEgIhrAgCvRstz71+JdF3IJaDQa5Obmwm63w+/3Q6/XIysrC42NjXEpGOBKdDdFVbG/yoaqBgdCikCaSY8FU3OQn5WKNz+uweET7f2+AAohUJCdim+vmorKBgf2Vdpgc/i6asxMwZmTc8KeuNDh8mHnkTbYXX5AAJZUHeaV5yEv04gPPqvHjiMtQ65667UyvrNmWq/Vrk2fnsDB6vYIvhrA0lkFWDZ7XEQfM1xCCBypteNgTTs6XH6EQip0Ohl5GSbMK8/BhCiPdttfZcW7uxugRLCSqNPK+MFlM2E0DBx6W9o92H6kBdWNTvgDfd+4KKpAlsWAKUUZOHtWQcynoFDs5OamobnFgf1VNjRa3Wh3+tDpCUIVAlpZRrpZj+z0FJQWWjB5fDrDM/UY7SuJlFij/fwZ1kr06X2f//znP/Gd73wHTz/9NC677LKegP3ee+/he9/7Hu6///4olkvh6PT48fKHVWi1e3utKh2ssWF2WTYq6gZutZAkCY1WN/7nnxVotLp7hVyHS1PDMgAAIABJREFU249jdQ7Mn5qLixYMfbFBhjkFF53V/+POmzse7U4fKr+wGn46nVbGqoXFfX5dnJthhCpE2G0nqkDUg+vp/EEFOw634ESTE56AAo0M5KQbsWh6PooLYj8L+WB1O2RJQiRdzIGggh1HWrH8zP5XowEgP8uEr55dCn9Qwc4jLWixexEIKpBlCQadBpOLRt7+Q8lDI8uYOyU35m1PRERfBmH1RJeXl2Pt2rX47ne/2+e+5557DnfddRdqa2tjUmB/Rus7mmi+G3vunxWnRr/1DTd2VwAaGbCYBt5i2uEOIBBUkJvR/2QIIQQuXDBh0AvUwiGEwEf7m3DkpB02hxcaWT61i5+M4oI0LJ6R12/PcEhR8czrB+HxDd0OAgD5mUZ8Z830EdU6kD2VVvxrTz38AaXP11sVQHG+GVecOymsDVWGw+UN4ulXD0Cn0yIYDO/r0S03w4jvXhybrwuNLqN9NYgSh+fO6NPh8uGzY1ZYHV7YOwMIhrpev9JMemRbDJhUaMG04sy4/MZptJ8/w1qJPl1TUxNyc/tfuTCZTOjo6BheZTQstS2daGgbpDdWCLh9yoAhWgBwe0PQDbIBiCRJ2F9l67cnstXuwc6jrTjZ4oI/qMBo0GBSYToWzshDeqqhz3HOmTMOy2YXorLBgXanH7IsYfJ4y6AXK2k1Ms6YlI1tB5uH3BhEADFbWdtTacW7O08CYqBdC4GTLZ144b3/n737Do+rOhM//j33TtfMqEu2mm25997ABpvimBIwIXWTDZsljQ0QEkiAUEIICxtKaEl+aYSwCwkQAoQQCM0YjAm2sXHFcpNkS1Zvo+kzt/z+GEm2kEYeWcXtfJ6HB2vKvWdGVzPvPfc977uXL6+YOGgdAY/U7Iug6wbWY0hFTvUkRJIkSTr5tfojvPlhNZV1foxPrFcCCIbj1DUH2V7ezNrttcyflCevTA1ASkH0mWeeyY9//GPmzJlDQcHhnNP9+/dz6623ct555w3ZAKWe9lf7ejQfOZLFohALxTAMs9c0irhmEItrOGzJZ6oBaltCtAdjpLsPB8Y7K5p5bWMV2hHNTWJxnc17Gth1sJVVS8ZQkt/zrE0Iwfii/nXjO2tmAe3BODsrW/qo8CFYPDWfGeMGv1FJXNNZu+VQIkrvgxCC+tYQ722rYfmcnl0Xj6+TrviOJEmSdAw272lk7dYaoh2LxfuaZVaEoM0f5bUNB9lX7ePTZ44esqupp7KUps0effRRamtrGTNmDPPmzWPlypXMmTOHyZMno+s6jz766FCPUzqCeZTAyO2wYulr9b1hdlzWOcrUpmkSjmkYpsn28mZeeHc/T76+F18g2qO7oBCCSFTjpXWVxPux+K0vQgguPmMU580tIj/LhWkm0kM6912c5+bTZ4zmrJlDs5hwY1kD4RSqi3SOdU91W4/3ZTBkee0I5dhmuPtaVChJkiSdGtZtr+XND6u7AuhUKUJQXpPo/JtKNS2pu5ROO8aPH8/u3bt5/PHHef/992ltbWXSpElcddVVfPWrX8VuP7ZSadKxKcn3sGFXQ9I0ByGgINeddBJSVRIB9JHdAXtjt6m0+qO8uLaCVn+UtkCUYFjDNE2sFoUsr73HmWswHGNjWQNnTBt5TK+t52sRzJuUx7xJeVQ3BqhvCaEIQVGeO2k+92CpqE2t2UunlvYI5TXtjC0c3OZDHpeNkjw39W2Rfj3PNE3GFAz9okdJkiTp+Pm4soV1O2qP+flCCBpbQ7ywtpwvnjteVubph5Tn7j0eD9deey3XXiu72xxv4wrTyct00ewL93q/YZicNasAh83Chl31tPqjqEqiRKHTbmX+pCwaWsMcqO870T/b4+Cf66uIazqqIojEDl8i0nSTxrYIeRlO7LbDwbgQggN1/kELoo9UlOumKNc96NtNpr9n5YoQiVJ/Q2Dy6ExqN/evjKTVorJwcv6QjEeSJEk6/sLROGs+Onra4dEIIThQH2BjWQML5PdGypIG0ZdccgkPPPAA48eP55JLLulzI0II/va3vw364KTeCSG4ePEo/vrOfvyhWLezRsM0mTw6i8VTRyCEYNb4HHYdaKXNH8VpV5lemoPVolDdGKB2TZBYvPcW1g6bBRO6pWZ8MlPBNMEXipFn6z4jHNdSb4t9IlMUCEbihKMapgFCEaQ5LDhsatIzddsQLCwEmDk2h7017VRUt6U0S2AYJjMnZqfUGCUYjrF1fzOxuIFFFYwvSic/K20whi1JkiQNofe21fWIA46VImDjrgbmTMjFog7Nd9mpJmkQ7ff70fVEANXe3i6n908w+Vku/uOCSazfVc/B+gAxTSc9zc600ZlMHp3V9ftShGDq6Kwezy/KdbNqaSlvflhFU1ukawGiYZrkZjg5a0YBL7xX0e05qiIwjERGtmEY6IZJTNPBNElz2nA5LAjAaR+8xQntwRhb9jUSixs4bCpzJuQOS8e8yrp2Kuv8NLVFuh37wUgcm0Uh2+vAZu2eDmO1qIwb5FSOTooiuOLCKfzy2Y9obAv3+fdomCZTS7M55yiLHJt8Yd7ZUsOBunbi2uFV3Ot21FGUm8aCyfn9XgwqSZIkDQ/dMNib4sRKqvzhGFv2NjJvkpyNTkXSaOftt9/u+veaNWuGYyxSP7kcVpbPPvZqEKNHeLnyoinsrmrrqjldnOdmXGE6NU1BNE3vdjbqtKtE4zpxzehYQJdIEQlFdcKxMPaQSk66g0mjMgf82uKazsvvH2B/jQ9dN7o6Zm7Y1cCE4gwuWFSCeoyL7Y7mYH07L7xbnsgZ/8SHk0AQ10wa2sLkZTqxHZFXPnqkZ0gDfLfLxldWTOStzdXsO+QjGI53a4KiG4l2zdNKs1kwOa/PD9bqxgAvrC0n3FECr3vbdjjUGORvzRWcO6eI2RNk+aP+aG/38dBD97N+/ftYrTYuvfQzXHHFlSiKgq7r/P73v+Yf/3iJWCzKwoWLuf76m/B6h+bkS5KkU9f+Q+34grGjloHtD0UIKmr9MohOUcpThm1tbQSDQQoLC9E0jQcffJCDBw/ymc98huXLlw/lGKUhJIRgUkkmk0q6B74uh6VHEOZx2Wj2RbsC6MTzE/8SCCIxjUhMY+qYnjPf/RGKajz28sc0+cI4bJauD4hELrbBzopmQpE4n1s+bkiukKzefIjWjkWUup6YcRciMROvKAKBwDCgzR8jLzORymK1qMOSR2azqlywcBSabrB5TyPNvgianljoWVrgYXxRxlHfk1hc56X3KroC6GQMw2T15kNkeR3D0pXxVPGjH/2AtrY27rnnASwWCz/72V1Eo1G+/e2r+e1vf8U///kyt912J+np6dx9953cd989/PSn/3O8hy1J0kmmtik4qAF0p5b2oVnbcypKaSrv3XffpaSkhEceeQSAa665hptuuol//OMfnHfeeTz77LNDOkhp+GV6HIzMdnW7LRrXUVXREaQlEqSFEF2zxHarit1moamt9wWPRxOKxPn7ugp+9tRmdh1oockXoaYpSFNbBE0/nGcthGB/TTsfV7Yc8+tLZl91G9v2N9PsixCN6aiqghCJgDKuGUfMwkMkpqNpOlaLysoFJcO66NGiKiyYnM8Fi0bx6TNHs3JhCROKU+s+tbGsAX8oltJ+dMNg056GgQ73tLFnTxlbtmzm9tvvZMaMWUyZMo0f/OBHPPvsnwgGA/zlL3/m+utvZv78hUyYMIlrr/0+e/fuQdNkaSlJkvrHl+LneH/5w7FTZm3TUEspiL711ls544wz+P73v09rayt/+MMfuOGGGygvL+f73/8+d99991CPUzoOZn6ii1E4oqEIgc2qYLEoKMLEaVNx2lVyMhyMyHJhVRW2lzf3e1+hSJyn3tjDx5UttIdiKEJBIDBNCEbj1LeGuwXSqiLYUTH4QfTz75YTjeuIzpl2ErPMFlVBiEROeFzvDKRN0t12vnjuOCaPHngKy3DZU9W/HLrK2naC4aH5sD7VVFdXY7fbmTBhUtdt48dPIBaL8cwzf0IIwRlnLOm6b86ceTz99PNYLLLJgSRJ/WPoQ9NMyzDMbt+3UnIpBdGbNm3iBz/4Afn5+bzyyitomsaXv/xlAC6++GJ27949pIOUhp9hmKiA02ahyRfBF4yhd8zACgQWRZCb4eqq15zmsHYFZnGt/3/Yb22qpqU90vH87sXiBYk0jhZ/90tMrYN8yam6IUCTL9IVQB/ef2Lm12ZVsVlVFEXgTbNRmOtm5rgcRmafPJUsdMPo96W6WNzgYENwiEZ0asnKyiIajeLztXXdVl9f3/Xv/PwRrFu3lv/4j39j1aoLuOeeOwkEAsdjqJIkneTUIaqgoaoK1iGqNHWqSWn6w+VyEY0mvnj/8Y9/MHLkSGbMmAFAVVUVWVkDy4GVBldc09lY1kBVQwDDMMlJd7JwSj7etL7bfHdKFF2voKU9UbXD47LiD8WIRDWEAJfdijet92YtpmkevRPiJ0TjOvsPtXelhfQmkXOtoxlmn90YB2LLvqauetq9zdQKEqkkSkedbCHA7Ty5OgIahonRz4KiQpCowiId1ZQp0ygoKOS+++7hxhtvxTQNfvGLB1HVRFnEpqYm/vjH33HNNd9DURQeeeQB7rjjR9x//yPHe+iSJJ1kMtypfaf3l9dllSXuUpRSEL18+XJuv/12tm/fznPPPcfVV18NwIsvvsgtt9zCihUrhnSQUuqafRH+smYfvkC0KxA8WO9nR0UzKxeMOmragaYbPL+2HF8g1lX2zm5Vsac7ibsN6ppDSQNoALvNwryJef0a88E6P+GYhqok8qttFrXX+tWmaRKOang6AtfsdEe/9nM07aEYHqcFfzh+1ML1umHgcVqZVpo9qGMYahZVwWFVU25nDmCYkDvI7/WpyDRNbr75BpYvP491697lwgvPweFwMn36TAzD4Ikn/oCmxbnuuh8wc+ZsAK677od85ztf54ILzkHXdRYvPpPvfvd6srJOruNKkqThV5zv5r3t5qAvLsz2ys/7VKV0qvHoo4+Sm5vLT3/6U5YtW8btt98OwHe/+13Gjx/PfffdN6SDlFL38vuVtAe7F14XQhDXDF7bcIBQJN7n8z/a20irv/fL/VZVweWw4Av2nh9rmCZzJ+Z262CYCsM0uyVQuByWpDPSncGtbpjMGDu4gYYiBBaLitOmJt9/5zBMk4klmYN2tu4Px1i9uZrn39nPi2vLeX9HbY+0lsEghGDMSG+/npOf6TypUlaOB8MwePDBe1m//n3y8vJ48sm/8NJLr/GNb1zFnj1lmKbJhRdeDMAzz/yp63nPP59YlP31r3+bX//6MdraWrn99puPy2uQJOnkMirfQ07G4Aa8umEysUT2B0hVShFAfn4+r776KoFAgNdff52MjMQbvHnzZt58801ycnKGdJBSag7U+altCSW9PxrXWb+rPun9ndtQ+lh0luV14EmzYT8i0DRME5fDwtIZBSydUdDvcRfnubs1LvE4rYmOiZ8IZAUCh13FMEwmj8pkQvHg/qHnZToxTZMsjx2bVUkaSBumwYistKM2M0mFaZq88WEVv/3bx2zcVc++Qz72VLWxdmsNv3pxB5t2D35ljDkTU6/7bJgmE0elVvXjdHXoUDVXX/1N3n//PdxuN8888yfq6mrJzMziueeeZvHiJWRn5/DlL18BwLvvvk1dXR2maRKLJU5YFy8+k9LScXzuc19k69aPuhpdSZIkJSOEYFJx5lEnffoj2+tgSi8N2qTe9Wsa7b333uMnP/kJ3/72tzl48CDr16+ntrZ2qMYm9VNVg5++ruoIIWjuWLyXjHGUv0UB5Ge6+K9V0zh3bhGLpo7gUwtK+Pal0zhz+sj+D5pE05gjZ0eFEORlOkhzWhHC7PqAsFsVPE4b8yblccmSMYMe2C2cnIfFoqCqCvlZLtwuK6qSCHQ7/1NVQabbzjc+Pbkr3WUg3vywik1lDeiG0ePqQTSm89bmarbsaxrwfo5UkJ3G3Al5GCnMtpfke1g8ZcSg7v9Us3PnDkpLx/GHPzyJ2+1B13UeeeTnbN++lZqaQ7z33hq+9rVvUFhYxLJl56AoCq+++hK7d5dRX1/H4sVnUlBQSGNjAy+88FdmzpyNqvbvao4kSaenM6aPICfdOSjbMoElM0bKSZN+SCknOhwO84UvfIGXX34Zr9eL3+/nm9/8Jg8++CDbtm1jzZo1TJ48eajHKh2FRVWSLorrekwvXf7ag1HWba+jss7PocYAgXCcNIcVb5rtkw37AMj02LFa1EHtaHTO3ELqWkJd9YuFEOSkOzAMO4FwHItFsGxWIWdOLxiyVcNpThsLJ+ezbnstihBkex1dedi6YaKqCnarwqKpI8jyDvxDyxeMsnV/c9/BuAkf7Kwb9NSVc+YWoqiweXcTmq73esyMLczg0iVjBuVk4VS2YsVKVqxY2fXzRRddwo4d27juuv8C4Mtf/iqrVl0OwC23/IRNmy7kySef4Omnn2LJkrP53vd+yAMP/IwXXvgLXm86v/jFb4/L65Ak6eSjKgrnzy/muTX7B1SWzjBNpo7OkrPQ/ZRSNHLjjTeyfv161q5dS1NTU9fM4JNPPklhYSE33yxz+E4E08dmY+kjwNQNk3FF3dsLN/siPPn6XraXN+MPxXDZVTTDpC0YpaEtzCcnK03TZOYgB3QA6Wl2vnTeeMaM9KJ0VMjonPmdPjabqz8zg2Wzi4a87M6SGQUsnVmIoyNdRQiBy2HF7bSS6bazZEYBy2YVDsq+Nu5qQE/hQ6/VH2XnINfEFkKwfHYR37x0CnMn5pGX4SQ9zU6W18GUUZl8ZcVEPrtsrCxzdAzS09P5+c9/wc9//gsAPvOZz3fd53Q6GTmygH//96/x2mvvcNttd+J2u/nsZ7/A7373BLNmzeF73/sObW1tyTYvSZLUTUm+h5ULS455jY5hmpQWeLlo8ejBHdhpIKWZ6Keffpp7772XM888s1uuXn5+Prfddhvf+MY3hmyAUurSHFamjc7io71NPWYPTdOkINvV4yzzrc1VBI5opGGxqGS6bbT6o0RiGv5QrKs0XmLhYB4TioemsUiG287nlo+jPRhjb7UPE5OxBV4yPcO7UviMaSOYPymPD3c30OyLYJqQ7bUzb1Jet9zthtYw+w8lgp2xhenkZbqSbbJXbUdUUOmLqggaj7EL5NF4nDbOm1c8JNs+3dntdgBisTiuIw6NWCyOw9H9mB41ajQAP/7xT/nMZy7irbde5/LLP48kSVIqpozOwmGz8MaHVbT5IymnZChCMHtCLufOKZJXHY9BSkF0MBgkL6/3smVOp5NIpO88W2n4rFhQgqoq7KhoIRpLlDFTFMHokel8+oxR3RYNBsIxDtb1bPTgcdmwWBT8oTgxzcBlt5CT4WR6aTZTxwz9pR5vmo25/Vj8NhSsFoXFU3vPBW71R3h1/UGqGw6/d+9uq6U4182nFpQMeuk96eSUm5v4zFy9fhemPQ/dMElzWmlsaiQ3N49IJMK6dWtZtGgxaWmJlvF2u4OCgkLa2lqP59AlSToJlRZ4+c8LJ/POlkPsPtiGPxTrNTA2TRNFEYk1L1NHUJLvOQ6jPTWkFEQvWrSIhx9+mBUrVnSd3XT+/w9/+AMLFiwYuhFK/SKE4Lx5xSydWcDO8hZ0w2BcUXqvs7nNvihxTe+165HTZsFps+B2Wvmvy6YPx9BPCoFQjGfe2kd7qHuZPwFUNwZ49u19fPn88XjT7EfdVm6Gk/2HfEedMTBMk+I890CGLR0HB5oM0tJzeWPN+4yZvhyA9uZqgoEA+1rdzInGueuu27n55ttZseICAILBAFVVB/niF//9eA5dkqSTlNWicN68YpbPKWRHeQt1LSHaAlFicQMhEpNUmW4700qzhv0q76kopSD63nvv5eyzz2bSpEldgfSvfvUrysrK+PDDD1m9evVQj1PqJ7tVPWopM2+aFaWXhYZHcthSOkROG+9tr8MXTJ6G4Q/FeG97HRcuGnXUbc2flMfmPY3Etb7zonMznIwrTO/zMdKJ5aM9jby5uZox085h98aXcHmysbvS2brmfykYO5e2qJ0X1lVz0UWr+PWvf0FOTi7p6Rn8+tePkpeXz9lnLz/eL0GSpJOYqijMHJfDzOM9kFNcShHS3Llz2bBhAz/96U954YUXUFWVF198kSVLlrBu3TrmzJkz1OOUhkCmx0FBbhq1TcFe7zc7FhtICaZpUl5z9Jnj8hofhmEeNb/M5bCyaMoI1m6rSfoYVVU4a1aBLDl0EtENk3U7asGEMTPOIRYJ8NHqP2IYOvmjZjDj7H9DCEFja4j5y76A0+ngJz+5lWAwwPz5C7nvvoexWOTJqyRJ0olOmINQpTsej2O1WgdjPClpbPQP274GU26u54Qb+4H6dl54p5xQVCOmGVgUgc2aqEyRne7gKysmytnoDtG4ziN/2YZ5tJ7gwNWfmYHLkdr7tmFXPRvLGvAHD+evGYZJdrqTZbMKGN/RVOZEPH6knj4sq+fNTdV9Ni3qlOVxcOXFk4flJEkeP9KxkseONBAn+/GTm5s8Z/yo3/JlZWUATJo0qdf7//rXv3LjjTeyb9++YxyedDzlpDvxpNmobgwSjmqYmLjsVmaMzeJzy8fJAPoIFlWgqgJN7zuIVlWlX6XhFkzOZ+7EXLbua6apPYICFOW5mVicIWegT0K1zSEEHLVmO0CLP0I4quFyDN8khCRJkjQ4kkZItbW1XHbZZWzcuBGA+fPn8/e//53c3ESe7datW7nuuut45513SE+X+ZonI003+PObe2j2RcjLdKLpJqGohqpAky9Kc3uUotzh+3KPawb1rSFsFoXcDOcJF0CqikJJvofyGl+fjyvOc/e7vrKqKMyZcHwrkkgDY5omW/Y1sWVvE/WtIQCsFhWnXSU9zdbr8awbJtrR2oRKkiRJJ6SkQfQPf/hDtm3bxs0334zH4+GBBx7ghhtu4IknnuCOO+7g7rvvBuCqq67iJz/5ybANWOq/YCTO+p311LeGEmVt8tzMm5TP5j0NNPkS9SRb/VF8wRi6YaIIQas/xrOr9/G9z88clGA2rhls3tOALxTHaVWZOzG3a/ZNNwze2FjFnipfR81qQX6mk7mT8pg1LmfA+x5McybkUlnXjpEk8BFCMGfCiTVmaeiZpslL6yr5uLKFUFTDBASCuGYQi+tEYjp5mc4eKR5pDisuu7zaI0mSdDJK+um9evVqbr/9dm666SYApk6dyle+8hW+//3v89BDD3H++efz8MMPJ03zkI4f0zTZVdnCgfoAvmCMitp2BGZXJY7K2nZ2VLTgtFtQhOhotx3v6EQp0DHRdIP9h3ys3lTNufOKqW7wU17rx6IIpo/LxuO0pTyenRXNvLX5EOFIHCES3Qg3ljWwaOoIFk/N56/vlFPRsWCvs+NSc3uE1zccRDdM5p5AM7SlBV7OnVvE6s3VaJrRdYJhmiYWVWHZ7ELGFWYc51FKw+2DnfXsqmxBVQQelw1/OE5n6rwQgmhMp6U9Ss4naoiXFniPucuYJEmSdHwlDaKbmppYtGhR189nnXUW7e3t/OY3v+Hxxx/niiuuGJYBSv0TCMV4bs1+alsSs841TUEiUQ1FEbgcFhw2Cx6XlZb2CNG4gWGYtIc6v/APz5KZJsQ0nU17GjnUFKC2OYwQiWDxXx/XM210JisWlBx1lrq2Ocg/N1Sh60a3GuOabrB2Ww2hSIzyGl/SRVgbPq5n9ricE6qT0uzxuYwrSmfDxw00dFy2z81wsmByfld3R+n0YZomOytauo5Riypw2S0EI3EEh4/5cCSO4bF3Pc5mVZg/qfcmVpIkSdKJL2kQHY/HcR3RqzYtLQ2An/3sZzKAPoG9/K8D1LeGUBVBeyhGMBLHNEAYJoGQRiSqEwzHyctwEI4m2nrTMQPdmwP1ftqDh1t/CyHQdYOP9jbhdFg4a2Zhn+PZtLsRXe+9DrIA1n9cj9Wi9no/gC8QZU91G5NKhqbV+LHyOG2cO7foeA9DOgG0+qM0tIW6zShneR3ohkkkpnUF0oaZSK1yO63YrSoXLBpFflb/WsVLkiRJJ45+J+MtXrx4KMYhDYKG1jAH6/1dM75tgSiYgs5JXt00sAgVTTdp9kdJT7PRFug9t7ezLWgsrvcaXyuKYEd5C+OLMti2v5lITMfjsrJgch7uI1I9Omdqk2kPxcnyqklCeBACQuH4UV+7JB0vMU3vcZsiIC/TSSAcJxiOJxrqCIHNojJjXA4LJ+eT5ZXdwiRJkk5m/Q6ij9bhTjp+Djb4u8pqxTW9oxTbEbPMZmLSWQiIxgxUjyAvy0FNYwjDMLvylYUAq6pgdjytt4VP4ZhGVYOfimd8eF127LZEbekte5s4f14R08cmFtcdLd3DZlX6LAWmCIWSEclrNErS8ZbpduCwWdA+ccVFAB6nFY/Timkm2rd/8dxxjBkpqxlJkiSdCvoMoh944AHy8/MB6OzJcu+993aVueskhODhhx8eoiFKqfK6bBgmqAKicQNFiI7AuOdjTdMk2+skP8tJKFxHezAGgEBBURInSzFNx2m3dLtMbZgmja1hghENTTfQDQhFQzhtFnIzHGi6wesbqynKc5PpcVCY66ahNZQ0SJ5ckkV9ayhp6+ui/DRy0p0pvwe+YDRRiaQtjCqgMNfNgsl5OO2yDq80NOw2lTEFHvZWJS99KASMzHQxeoTsACpJknSqSBpEl5SUsGHDhm63jRo1ig8++KDHY2UQfWIYV5ROlseOLxjDalEQAiyqQlw3wEykYChCYGKiqgrnzS/CYbNQUZvoJBSO6hxuYGmSm+7osVCu2RchEtMxOmaPhUhkfIajGs0d1Qc0XWfDrgY+taCERVPy2FPVRijSMyXDoiosmTESfyjGaxuqiGt6t2oXmR4HFy4alfLrr6ht56X3KhJ5qB3bqWoI8HFlC59fPp7sdHn5XBoaS6aNpLoh0bCoN4oQLJ4+4oRrH21TAAAgAElEQVSrfS5JkiQdu6RBdGVl5TAOQxoMihCcNbOQV9ZXYkfFZlGIayZCKBiGic2qoCoKDpvKxJLMrlJsly4ZwztbDlHdGCAc01EFjC/O4PKzxrL6o0PsKG8GwDQMwlG9a3bbooquXGYhBOGohq4bqKqCr2Nm25tm5/KzS3l9QxV1LSGESMxmZ6c7OHtmISX5iVSNgpw0NuxqoNkXQVEEo/ITtaxTbVpimiZvbKwiGte7BSpCCPyhOK9vrOJL540fpHdakrrLzXRx2VmlvL7hIA1t4a5qM4ZhkuGxc9bMAiYWn1iLYyVJkqSBkVX+TzGTR2ficqhsLGtEEYkSd6oqSHfbsVkS+cdup42LFpV0PWdsYTqlBV5qmoK0BWIU5LjI9DgIROLoukkgouEPxNB0A80wcDks2IXaVQe3k2GYRDUDl6pgtx4Ofkdmp3HFBZOoqg9Q1xoi021jbGF6t2A30+PgUwtKOFZlB1ppbo+gJimFV9UQoKU9TJY39dQQSeqPolw3X7twMvtrfByo9WMA+RlOppVmn1AlGiVJkqTBIYPoU9CoEV5GdeRetgdjrP+4ntrmIAgoyE5j0dT8bhU0IDFjW5jrprAj3T2u6fz5jb20tIfJSLPhddnwB2M0t0dQFAWXXaE9dLgObidFEeiGyZTRWT3GVZzvpjjfPSSvua8AGsAwDOpbIzKIloaUEIJxhRmy4Y4kSdJpQAbRpzhvmo3z5xf36znby5t59YMDHKz3o3Skf6S7bXjdNgKROLGYTprDgs2iEtP0rkDaalGwKoJJozIZVzi8FQi8aTYMw0w+4ycgJ90+rGOSJEmSJOnUJevVSd38a0cdr3Q0bAGBYZiEIhoNrWE03SA9zYZQEm2M87OceFw2VFWgKIKSER6WzS5k1ZIxw76AauqYLDx9dAsszEkjN0M2tpAkSZIkaXDImehTUHmNj+3lLTS0hmn1R7BaFFx2K8V5aZwxfSQZ7t5nZKNxnfUf1yeas3wi31nXTdoDMbLTHSiKQFUEVotCXoaDghw38yfnMbbg+NW/VRWFc2YX8cr6Az06JDrtFpbPkd0FJUmSJEkaPEmD6M2bN/drQ3PmzBnwYKSBe33DQTbvaSQc1WhpjyZiYUFH6bsolXV+vnBO7+Xetu9vIhpPlIezWlUise6VLiKxRGc2h01l+exC5k9K1BA/URZNTR6diddt5cOyRuo7KoEU5rpZNOXYu8OZpsnOyhaq6gMoHakqo/Jl8xdJkiRJOt0lDaLnzZuX9JJ8bx3mdL1n61tpeJUdaGXz3kaEAF8wdngy2YRWfxSn3UIgHGftthpWLS3t8fyYZnT9Xr0uK6FIHOOISd3O7WV67MyZkHfCBM9HKsxxU7hkcBYvBsJx/vL2PupaQl2LFrfsbWJcUTqXLS09IV+/JEmSJEnDI2lO9Ntvv83q1atZvXo1jz32GDabjauvvpq3336bsrIy3nvvPW6++Wa8Xi9PPvnkcI5ZSmJnZQuKSOQrf7IDoGGa+EOJhicH6v1HNFU5rHTk4W5qFlUhN92J1ZJozmKaJlZLooLH55ePS7l+88nsH+9X0tAa6lb1QwjYW93Gmi2HhmSfcc0gGpcnpJIkSZJ0oks6E3322Wd3/fuWW27hxhtv5I477uj2mMWLF+PxeLjvvvv40pe+NGSDlFITiiS6pcV0A8M0E22/O+4THYsEIRGoGaaJ+omrCSOy0xiV76Gyrh0hBHabysjsNKJxnbhucNHCUSycOmI4X9Jx09Ie5kC9v9erMYoQ7KlqY/nswkFbQFlVH2DdzloONQYxDIPcDCdzJ+YyvTRnULZ/IvvNb37JG2/8k+ee+zsA27dv5aqrruz2GKfTyRtvrD0ew5MkSZKkXqW0sHDz5s3cfvvtvd43c+ZM7rzzzkEdlHRshIDqpiChcBzDBIGBoghsHY1PVDUR8GV7HahK7zPJly4dw0vvVXCgPoBpmhimSYbHwaIpeV050KeDQ02hXtOWOrWH4sTiBnabOuB9VTcG+Ou7+4jFD189aGgN8+oHB4nGDOZNyhvwPk5UZWW7+NOf/pfc3MOvsbKygrFjx/Hzn/+i6zYlyfEqSZIkScdLSkH0pEmTeOKJJ1ixYkW32w3D4Je//CUzZ84cksFJqWv2Rfi4ooVguLMBiokJ6IZJNKbjslvwuKyYpsnMcclnNx02C58/ZzwNrWHKa304rCrTSrOxqKdXEJPlsWOaiROT3titKhbL4MxC/2t7XbcA+kgbyxqYPSEn6UnPySwej3P33XcwffpM6upqu26vqNjP6NGlZGef+rPwkiRJ0skrpSD6rrvu4tJLL2Xr1q2sXLmSnJwcGhoa+Pvf/05NTQ1vvPHGgAZhGAZ33HEHu3fvxmazcddddzFq1KgBbfN0886WatrDcayqgqYbHFmlzjDBZlVx2CzMmZDLnAm5R91eXqaTvMzTt7tfQU4a3jQrlXUBNN1AEQKXw0Ka0wqmSelIz6AEtoZpUtUUSHp/WyBCZa2fsUPYvCauGWi6gcOmDmt978cf/x2FhcXMmzefP//58LqKiopypk+XJ+aSJEnSiS2lIPrCCy9k7dq13HvvvTz55JO0traSnZ3N8uXLueWWW5gyZcqABvHmm28Si8V45pln2LJlC//zP//D//t//29A2zzd7KxsxTRMLKrS1Xob0+xKSfC4rFx16bRBST84HeyoaKbFHyUc1TCMxHsYjmkEI3EmlmRy7tzBqTttmiZm75PQQCKXXTP6eMAANPsirNlyiAN1fuKaQYbHxtTRWZw5feSQB9O7d5fx0ksv8MQTf2bNmre63VdRUY7T6eKKK75Ee7uPWbPmcPXV18mZaUmSJOmEknKzlUWLFvH8888PySA2bdrE0qVLAZg1axY7duwYkv2cyjoDPUgsfFPUw0GQiUma0zroAbRpmmi6iUUVw96hcCjFNZ23Nx/CZlEZmeVM5D93VDuxqiozSrNwOayDsi9VUcjPclGTZDba5bB2q5oyWNpDUZ5ZvY9AOAYk0lZ8gRhrt9USDMf51MKhuxLUmcbxne98t0dgHAgEaGpqRNM0brrpVsLhML/97a+4/vpr+f3v/xeLRfaHkiRJkk4MKX8j+Xw+Hn30Ud566y3q6up47rnnePnll5k5cyYrV64c0CACgQBu9+Havqqqomla0i/MzEwXFsvJOaOamzs0jTrGFHipbQqhd80+J8rUWdTEJfqpY3MGbd+RmMY/1lWw50ArwUgcb5qdaWOz+dSi0d3KwZ2s1m45hGaA1WrBarXgdHbv8FjbGh3U3+N5i0bxzOu7MT5RddAwDRbPKKBgZEbXbYO133Vr6onEdazW7n9jVqDskI+LHDYyPb13thyohx56iIKCkXz1q4mKPm63A1VVyM31kJvrYdOmTTidTlQ18Tc+d+50li5dSkXFLs4444whGdPpYqg+f6RTnzx2pIE4VY+flILoyspKzjrrLAKBAEuXLuXdd98lGo2ybds2br31Vl566SUuuOCCYx6E2+0mGAx2/WwYRp8zTq2toWPe1/GUm+uhsdE/6Ns1TRNdM4jp3Tuj6EYi1zXb62TuuOxB2bduGDz5+h7qmoNds8+NUY23mgMcrPFx+dljB7yP462uwY+mJa/V3NYeHtTf4wivnXNmF/Cvj+tpagsjBLidNqaV5jBv/OHfWyrHTzSmU90YIM1pIT/TlfQKQVllc9LXaMZM3tl4gDOnjxzYC0vixRf/RnNzE7NmzQJA0zQ0TWPWrFncf/8jzJw5m3D4yL9xO15vOnv3VjJ+/PQhGdPpYKg+f6RTnzx2pIE42Y+fvk4AUgqiv/vd7zJy5EjeeustHA4HNpsNgKeeeop4PM6dd945oCB6zpw5vP3221x44YVs2bKFCRMmHPO2Tkfb9zezvaIFm0Uhrhsc2UdFN2DB5FzyMl2Dsq/NexqpbQ6ifCI4U4Rgb1UblXXtjB4x+OkHwyk3w4FumEln1b1ptkHf5/SxOUwrzeZQU5C4plOU6+lXQxtNN3htQxV7qtqIxOKAID/LydLpBYwvzujxeKOXZjudhDhcU3woPProb9A0revn1157hZdf/huPPvob2tpaOf/8s/i//3uWESMSNcnr6urw+doYNWrMkI1JkiRJkvorpSB69erVPPnkk7jd7h7tvb/1rW+xatWqAQ3i/PPPZ926dXzxi1/ENE3uvvvuAW3vdLN5byPRuIFFVVDVxKJCsyNHWlFgf0070ZiOYZo47QPLKa2s8/cIoDspimDXgdaTPoiePDqL93fU0eqP9rxTwKw+SgQOhBCCotxja1n+0roK9la1IYToqhrS1Bbh7+9X8tllYynJ734mPTIrjdb2Xl4ficWME0syj2kcqRgxovsMd0ZGBqqqUlRUTF5ePjk5Odxzz51cc833iMWiPPTQ/cyePZepU6cN2ZgkSZIkqb9SiqhsNhvhcLjX+1paWrDbB5Y7qSiKbNgyAM2+SLfOhBZFdDV01w2Dg3V+HvnrNkzTJD/TyZxj7IS3u6qVsgOttPqjqIogzWnF9YmgfChnMIeLIgSXnDmGF98rp7U9iqIITDNR+eSMaSOHtNzcsWhoDbGv2tdr6oamG2zYVd8jiF40JY/9h3w9Woybpsn44ozjVt7QZrNx//2P8ItfPMjVV38TMDnzzLP47nevPy7jkSRJkqRkUgqiL7roIm699Vbmzp1LaWkpkJg1a2lp4Z577uFTn/rUkA5S6lu6245JO4ebfCfohkEsbqDYVMyOy/f1rWFeXX8QTTeZPf7o9aI7vbHxIJv2NBKKaERjOkIIQlEdr8vatQBNN8wewdrJKj/LxdcvnsL2/S00+sLYrQpzJuTidg5+KsdAlR1s6/P+2uaeawhyMlxcdtYY1nxUS01Tojul02FlQnEGn1pQPFRD7dXll3+Byy//QtfPhYVF3HPPA8M6BkmSJEnqr5SC6AceeIBly5YxdepUxo8fD8CVV15JRUUFWVlZ3HfffUM6SKlvFy4qYUd5M5refRZYN0yEAm7nJ8qxmbBhVz2zxuWkVJquoqadTXuaUITA67ISisTRDRCAPxTH5bBgsyjkZzqZOiZrEF/Z8aUqCrPGn/i1iY/6G0zygJJ8L19d6aXFFyEQiZOf5cJuHd6qN/UtITbuqqeuNYRAUJCTxsIp+WR5HcM6DkmSJEnqr5SC6NzcXDZt2sQTTzzBmjVrKCwsJD09nSuuuIL//M//xOM5NWYfT1YFOW6Wzylk9aZqND3R7juu6egGWJREqbtPtrBu9kWoaQpSmEIO7vbyZjrX2KmqQm6Gk5b2CNGOVtWBcJw5E3K5aNGopPnSR2OYJht31bOnykcoopHmtDJ5dCZzxqcW6B8PVQ1+dh9swzBhVL6bCcUZx2Wsk0dn8q+d9RzuUdldUU5an8/PSneQlT78QWvZwVZe+eAAmna4qkxze4S91T5WLR1NSf7JnVsvSZIkndpSXmXmcDj41re+xbe+9a1utwcCAd59913OOuusQR+clLpLl5QyeVQWT7+1l9rmYKKyg2Zgtai0BaNE4zq5Gc6uQFpAj5nrZMIxrdvPNqvKiOw0YnGduGYwriSdL5wz/pjHbpomL7xbzt6qNpSOaN0XjFLV4OdQQ4BPnzn6hAqkNd3g+XfLqahp73o/N+9uYER2GpefPQaPa2jqKyeTk+5kUkkGHx9o6XESY7MqLJicP6zjSYWmG7z5YXW3ALpTJKbx+sZqrrxoMoZh8Pvf/5p//OMlYrEoCxcu5vrrb8LrTefgwQM89ND9bN++FZfLyQUXfJqvf/3bsiGLJEmSNCxSqqGlqipXXnkl0WjP1fw7d+5k+fLlgz4wqf+KctNwu6yMGellVL4Ht9OGEAJBomV1eyjW9Vhvmo3C3L5nKCExQ9zSHqG2OURVQ4CapiAt/iimaWKzqrgcFkZkHH07fdm6r7lbAN1JVQQ7K1vYU9V3zu9gM02T+pYQB+r8PRbeAby6/iAVNb5uM/uKIqhvCfLi2ophHOlhF58xmrkTcnHaLcQ1A8OAkdlprFpamtLVhuG2bX9zV7fE3jS2hSmv8fHb3/6KV155idtuu5NHHvk1Bw4c4L777kHTNK6//lo8Hg+PPfZ//PjH/83rr7/KE088NoyvQpIkSTqdpTRlY5omTz31FFu2bOG5555jzBhZr/VEtHlPI/GOoC8Q1jAMg6imoZIofReOaqSn2TBMkxnjcrCofZ9DmabJi2vLafRFiOsGmInZa38wRiyuk5/pxKKqzJ+c+gLF3uyuau0RQHfqDKSHsuTakXaUN7N+Vz31LWEEicV244sSi+0sqkIoEmdvdVuvM+NCCGqaQhyo9zNqmBdYKorg/PklLJtdRKs/gsNmGZJ61oPFF4z2mfqjKoKquhb+8pc/c8cddzN//kIArr32+9x7793U1tYwefIUfvjDH5GW5qakZBTLl5/L5s0fcuWV30q6XUmSJEkaLCl3c3jqqadQFIW5c+fy8ssvD+WYpGMUjumYJCpwtLRH0I3E4jjNMIjGdWJxDZfDwhnTRrIkhW50e6rb2H2wDYdVJdNtBwEmifrTkZhGMKyxYn4RGe6B5dNGY8m7A6Zy/2DZUd7MP9cfpNkXwaIKVFUhFtfZUd7Ec2v2Y5omBxsCxPoYjxBQVX/8OjNZLQp5ma4TOoAGcNktXRVjeqMbJvVVuxFCcMYZS7punzNnHk8//TzFxSXceec9pKUlZtl37y7j3XfXMGfOvCEfuyRJkiRBP4Lo4uJi3nvvPS655BJWrVrFLbfcgmmaKErqXdWkoZXlddDii3SVoAOwKAp2m4rNopLmtPGtS6Zy1syClHKMyw60dnXt87isFGSn4XXZcDkspKfZGVvoZfrYgVevcNpVWv1R6ltC1LeEaPVH0Y+oN+1NG/ocY9M0Wb+rvtdOfkIIKuvaqahtx2mz9Nntr7OetNS3ORNycfTR+CfLa4dYK/n5I1i3bi3/8R//xqpVF3DPPXcSCAS6PfYrX/k8V175FbzedL70pX8f6qFLkiRJEtCPIBrAbrfzxz/+kZ///Ofcd999rFixgmAwOFRjk/ppemkmmm72CJAFAkUR2C0qNU09awYnE//Eoi+LKsj02MlJd5DpsSepBdE/tc1BKmvaaQ/FiMYNonGD9lCM2uYQkXjiZGDexKEvM9fsi1Dfmvy9UYSg7EAbxflusvuoZGG1qMwYmz0UQzzhbd++lSVL5nX77/zzlwKg6zq/+c0vueSST7Fy5TLu+umtzBnj7jWlw6IqnD2zkHAoRFNTE3/84++45prv8eMf38WePWXcccePuj3+ttvu5MEHf0k0GuGWW34wLK9VkiRJko5pyuzaa6/l9ddfZ9u2bVx++eWDPSbpGGmaSbrHnuiw1xHidv7fm2bF7bLS0NZ758neZHocfV5yz/R0DyZN0+zz8Z9kmiavfnAQUwjS0+x0lmgTCHTdoLU9wrJZI8nPGtjCxVRE4nqyCnFddMNEEYKFU0b0WnvZMGHmuGxcDmvPO08DlZUVjB07jr/97Z9d/z377N8Ael0g+PqLv+PSpWMYNcKD3WbBabcwtjCdzy4by6RRmVgsFkKhID/60R3MnTuf2bPnctNNt/PBB+9TXV3Vtd+JEycxf/5CfvSjH7Nx43rKy/cfr7dAkiRJOo2ktLDw7LPPxuvtXrN12bJlbNy4kcsuu4wtW7YMyeCk/rFZVTLdNuxWlUAoRlw3UITA7bRitSjohkmWJ/XUiAVT8ti2v5m41jMHWBGCuRMSM8QfV7bw0d5G6lvDKEJQmJ3Ggql5jDpKnd+KWj8NbSEUIchw20hzqPhDcXTDTLQVd1jI8AxP/eL8TCdpTmvS/GvTNMnJSIxl1rgcVAU27mqgrjWcuM/rYPrYHBZPPfHKyR0pEI7T5o/gdtnIcA9umkxFxX5Gjy4lO7v7lYNgMJB0geCYEW7GF2X0ur2cnMSC1TFjSrtuGz16NJCY9d6/fy9nn31O132lpeMAaGtrHbTXJEmSJEnJpBREv/32273eXlJSwgcffEBNTc2gDko6NhZVYcxILx9XtvS6sCwvw0lpQeoNLDxOGxctLuG1DVWEInGEEF2l7ZbNKqQoz8MHO+tYu62GIyegK+raqWr0c+Hi0Uzqo6pGfWuo2+V8q0Uly9u9Y15DazhpkDWYrBaVicUZbN3X1Gu+eJrTxryJeV0/Ty/NYdqYbNpDcQzDJN1tO+ZGM8OhujHAe9trOVjnR9MNhCIoynEzf3Jen7+j/qioKGf69Jk9bt+6dUvSBYJ96dzWnj1lTJkyrWsfALquceutN/LXv75MXl7ixGXXrp0IIRg9WlYPkiRJkoZe0iD62muv5YYbbqCkpIRrr722z40IIXj44YcHfXBS/503r4jGtjCNbeGuYNA0E6XaVswv6XfTkgnFmYwZmc7mPQ34gjFcdgtzJuTiciRmbT/4uJ7eMjg03eTdrTVM7KOLn9tpxTDMpOXtDNPs2bJ8CJ0/vxh/KM7+mvauDo2maeJ22vj0maMSDWyOIIQg/QSvggFQWdfOi2vLiXV0mOxc+FjbHOTldZWEYzqzxw0877yiohyn08UVV3yJ9nYfs2bN4eqrr6O6uqprgeDjj/+OtrZWFi5czDXXfB+3O3kN68LCIpYtO4d77rmTG2+8FYvFyr33/jeLF5/JypUX89e/Pstdd/2Y6667gba2Nu6997+55JLLyMo6PXPSJUmSpOElzCRJrGPGjOHFF19k5syZjB7dd8c4IQTl5eVDNshPamw8fiXEBiI31zMkY69u8LN5TxPN/gg2i0pJXhqKolDVEMAwTPIynSycmo/HObgB37921PLu1pqkx4ZuGHxu2ThGjfBQ1xzCalHIz3J1PV43DB58ZivN7REgUfbMbjs8E53msPLtVVNRh7kCTGVdO7sOtKJpJnmZDuZMyOsRQB8Px3L8mKbJ46+W0dRHLrzDZuGqVdMG9BoDgQArVy7jjDOW8rWvfZ1wOMxvf/srIpEIy5adw5/+9H8UFhZy9dXfQ1EUHnnkAbKzc7j//kf63G44HOaXv3yIt956A8PQWbLkbL73vR/idrtpaKjn4YfvZ9OmjVgsFs4//wKuuuoabLYT/8TmeBiqzx/p1CePHWkgTvbjJzc3ed+HpEH0iexk/WUMxYG0eU8jqzdVE9V0/ME4sY785bxMJ9d8Zjppgxw4H2n15mo+LGtIer9umOSkOwhFNYLhOKaZGNfs8blMKEnnxbUV7Klqoy0QAzOxCNJhU8nxOrBaVc6dVzwoM6SnimM5fspr2nn27b19ppqYpsnZswpZNHXEgMYXDAZwOJyoauJEqKWlmVWrLuCb3/wvfv3rX/D4439i/PgJQKKu85VXfoWnn36BoqLiAe1XSs3J/kUmHT/y2JEG4mQ/fvoKolOaejrnnHMoKyvr9b5t27Yxa9asYxuZNCChSJw1Hx0iGIlT1xzCH44RixvE4gYH6/3c//QWAuH4kO0/02PHMHo/BzNJtG6ubQoSjelYVAWrRaHVH+WtTVX88vnt1DWH8Lps5KY7sNtUFCGIxnQicYNLl4yRAfQgqGsOHjVXWwhBqz/a722bpknZwVZe23CQf64/wLYKP/oRVRGzsrLxetPRNA3ofYFgXV1tv/YZ1ww+2tPI+9tr2bynoddFr5IkSZI0HJLmRL/00ktdX35r1qzhpZde4uOPP+7xuDfffJP9+2VJqePhw90NxOI6Lf4oppkoDdfJNBM5r796YTsrF5QwrTQ7ae7xsZoxNpsPdtbjD8V63BeKaJim2WtDjWBEo6U9QlFeok6w054ob9YZjpsmJ0T6xKnAoiqYZs/a4Z+kqv07Nirr2nnzwyqa2iIoiqClrpx//f1BVv77Tzlz7kSWziigvr4en6+N+fMX8fvf/7rXBYIFBYUp7U/TDd74sIq9VT7C0cOLXNduq2VcYQYr5hfLY0aSJEkaVkmD6NWrV/PII4l8RSEEN910U9KN9HWfNHSCYY1QVEtUW+DwIsK4ZqAbJkLAwYYAL/+rkvW76lm1ZAy5ma5B27+qKJw/v5iX11USjWvdFjJquk62t/fydJGYjmkmyq15XYfTTTrDOCFgz8E2Ro9IvZKI1LtppVm8t70WTTeSPsY0YWJx6hU6Dtb7uxYqdp6YpecW40jLYP3rf8Af+BwHyj2sffV/mT17LlOnTku6QDCVIFrTDZ5+ay+HGgMIIbqOM9Fx5WJHeRPNvjD/dv4E2S1SkiRJGjZJc6JjsRi1tbWYpklpaSnPP/88s2fP7vYYVVVJT0/H40meLzIUTtbcmsHOC1q7tYZX/lWJP6x13RbX9K6uhUIkZnRHZruwW1UyvQ6+ftHkflfoOJpWf4T1uxpoaA2jAIV5aRxqDCTtjtjYFiEc1fCmWZPWKp4xNpuVC0cN6jhPdsd6/Lz8fiU7K5p7/b2bpklhrpuvrJiY8vaeemM3hxp7dioN+hrZ+f6zNNfsxTRNzlp6Njdc/wO83vQ+FwgezVubqviwrKHP49Y0TWZPyGXF/JKUX8fp5mTPS5SOH3nsSANxsh8/feVEJ52JttlsjBqVCGIqKiooKCjAaj09O7GdqOZOzOWtTdW0h+MIEpe3EzPQAjBRFAVFiK7Zuaa2MDsrW5g2pv8lwBpaQ2wsa6DZF0k0VMlLY+HkfFwOK5keBysXdA9eXvngQNIg2mZRCEWMpJffDcOkIHvouxSeLlYuLCEUiVNe294tP9owTPKzXKxaWtrHs7urbwlS3RjsrWEjaem5LLjgO0AiqJ07MQ+vNx0Ap9PJDTfczA033NyvsRuGyZ4q31FP/IQQ7K3yce5cY9iruUiSJEmnp5SarYwaNYqysjJeffVVgsEghtH90rAQgttuu21IBigl53JYueTM0Tzx2m503cQwzERutEjUXraoArtNRe245K4qgtrmUJ9BdEt7hPW76qlqCKDrJlleOw6byt5qX7dFhPKIwNUAACAASURBVIeaAnxc0cKqpaUU5vacTZw3MZed5S1dbceP5EmzEolruHrJlwbI9NqZVtp3oB+OasQ0A5fdInNhj8KiKnxu+Tj2HfKxo7yZUFTDZlWZWJTR71z5yrpArwH0JwkhaG7v/2LFTzpQ76fNH0FNIU2jPRRj/6F2JhQPfXMeSZIkSUopiH7sscf45je/CYDb7UZRejadkEH08TFnYh6BcJxX1h8gFNHQjMRMnKoKLIrSo823gKQLzTpzXSNHtL5ubo9Q1xzE7bJ125YQgmBE45UPDvD1i6f02F5epovF0/J5f0ddj/047RY+c3YpH5Y1Eol2z6V2O21cvHh00sBuT1Ubm3Y3UN0YRNN0nHYLYwrSWTJ9JNnpw9Mi/GQkhGB8UcawdH/spOk67249RCSqoyqCkhEexhWm9yudKBTRECkG+YqA4BBWo5EkSZKkI6UURN99991cdtllPPbYY6Snpw/1mKR+OmtWIWMKvGza3cj7O+qIawZ2q4o3zYZFTaR5BEJxAlGND3c3sLOyheJcNwsm51OUl5hFNk2T1zdWdQuggY7KGwJ/KIbTpuC0d0/pafJF2FXZwpReZreXzCggN9PJtr3NNLdHEAoU5yX2m5PuZPa4XNbvqqehNYwACnLSmD85H8MweGfLIWqbQxiGSYbHzvxJeVTWtbNm86GuuW1VVYhpBrsPtlJV7+ezy8YxInvwFk5KPY0Z6eGdLX0/RjdMmn1h2gJRqhsCXUHzxt0N5GY4WTQln6kpphRlemyJjpgpxNGGCVne3nPsJUmSJGmwpRRE19TUyAD6BFec56E4z8PokV5e33iQI7MoWvxRAqEYaU5bVy3mfYd8HKj3c/Hi0YwvzmDfIR+NbeGu1I9OcS2RuiMQBMJajyBaVQQ1zaFeg2hIVH1IVvnB5bCyfHZRt9v2VLXx6gcHiMQOz1BXNwbYsq+JYDhGelrvQVIoqvH6xoN8deWk5G+SNGB5mS4Kc9N6XVgIiQC6viVEXDfI8ti7zTorQtDsi/DqBweJxnTmTMw76v5GZqeRn+mkyRc56mNz052U5A/vImdJkiTp9JVSMuncuXPZunXrUI9FGgSzxuWwckEJWV4Hmm4SDMcJRTQ8afYe6Q5xzeDtjw5hmonA55MBNNAtCNL03hur9Pa8Y9HUFuaVDyqJxvUel/z9wRhtgVivNak71TaHOFB/8q4APlmcPasQu1Xt9b5mX4S4ZuB1/X/27js+rvLKH//nuXd6lTRqVrGtYlvuDRfAgG3AGCcGAiGVLCzJbpbvssmGxIGQhJBiQwIEAuRHNvmGhCR8d0kgLBACDsSY2Lj3KtmyJKu3GWl6ueX5/TG2sFCbkWY0M9J5v168Xng0c+8RHu6cee55ztFBoxn8OSrneO9wC5zuoUeRX8QYw7xyB9QRBquqHJhbnpPwzjOEEELIUGJaiX700Ufxuc99DrIsY8WKFTCZBt4yX7JkScKDI6OzoCIX88sd6HaH8Ld9jWjq8mOoPNfpCaGmqRc6rThorbReK0ZXhRFtmTeYOdNj7zE8nH3VnYhIg/czjshq32q4xagdNFliDGjq8GIarUYmVUmeBZ+4ugzvHGhGV2+wr+NHRFYQkRXYLTrYzcOPm5cVFfurO2NqY7isKh9OdwjHznUP+veucmBeWQ5WzikY3S9ECCGEjEJMSfTVV18NANi0adOAD7GLiZei0PjddMIYQ16WERqNMGQCDURXkXs8ISyZlY9dJ9oRkfr/PdrMWvhDEmRZhfEj3TQ45ygvsqMgJzHt6Jo6h15FvvgrhGUFkqxCN8RKKBkfUwtsuHvDbNS2uNHQ7gVXOZq7fRAFFlOLOcYY6ttiu2vAGMP6FVNR6DDhRJ0LLd3RDkECE1CcZ8ac6dlYMjOPVqEJIYSMq5iS6Pfeey/ZcZAkMeiG/ytWVQ6bRQ+9VsTCSgf2nuroW1lUVA5FUZFt1SMYlmEy9D9WeZEdt1xVNurYJFlFV2+0l3Relqmv/nowOq2AYEQG4xjy1j4HMKOE6vbHy0c7fvxlVwO6e0euXb4oEJKhqjymFnuMMSyekYfFM/LgCUQQCEkw6bWwjbDiTUiicM7xjW98FVdccSVuu+3TAACPx4Pnnnsau3btgCTJWLJkKb761W8gLy9/xNcSQjJfTEn0Nddc0/fvsiyju7sbubm50GhiejlJofnlOTh93gWGDzcJajVC36qd3aLH7KnRcozVi4qhFQXsPdWO1u4AwpIKzjkMehHzyx2YVmgF59HV69nTspE/yhHiobCMbYdbcK7F3VfjbDXp0OsLQyMKg9ZYW006eAMS+IX4B1OSZ0nYqjiJX7y18YKAIUuEhmMz6fqNiyck2VRVxVNPPYa9e3fhiiuu7Hv8kUe+j66uLjzyyBPQ6w34+c+fwv3334df/eoFiKI47GsJIZkv5ikVe/bswXXXXQeLxYKSkhIcO3YMd9xxB77zne8kMz4yRtMLrbAYtGjtDqDVGf2npdsfbTnHgFULpvStBDLGUFFsh0YjwmLSIduqR1GeGQXZJnT2BHGwugv5WUZcs6h41Al0MCzhD++cwfFz3QiGZWhEARpRQDAsIxiS0e7yQ1Y5JEWFpKi4uOgsCgw5Nj1MBs2AcgHOOWxmHTaspJHPqVToMPUbyDOSXLuRSjBI2mtpaca99/4rdu3aCYvlw/0Wvb292LHjfXzta9/EnDnzUFFRiW996yGcOVONurpzw76WEDIxxJREb9u2rW81evPmzeAXMpt58+bh0UcfxU9/+tPkRUjG5O8Hm+ELyTAbtdBdWIFWeXRV2mLUYW5ZTr/nv/L+OTR2+ODyhNDrC6O7NwSPPwKOaBnFP462QlGHLrsYyTsHmuF0BwdNnsxGDSKSivNtHrR2+dHa5UdLtw9tzgCcnhBUFbhyXiFmlNhh0IkQBQE2sw5LZubjCzfMQraVhq2k0oIKx5AtCD9KVXnCNqQSkkwnT55AeXklnn/+D7BYPpzOqtfr8dhjP8OMGTP7HmMs+pEaDoeHfS0hZGKIqR7j/vvvx6c//Wn87ne/gyzL2LRpEwDggQceQDAYxC9/+Uvcd999SQ2UxK/bHcThs04IDMiy6GC36KCqHIxFe/b2+sI4cc6JBZW5AIC/7jmP0w29/W6xS7KKHl8YoYiCvGwjPP4Ijp1zYvGMvLjjCUcU1LV6Bk2gZUVFlzuEvqVnFn0sIgOqKkMjMhTmmHDqfA9ybHqsXVKMOdPjG1lNkksUBCyZlYvth1tGXGHOzzZi0YzccYqMkNFbt2491q1bP+Bxo9GIyy/vX57x0ksvwmw2o7JyxrCvJelnpLr1Rx75AbKysnHPPf/R91g4HMKvfvULbNv2Dvx+H+bOnY///M9vYOrU6eMYOUmlmFaiT5w4gTvuuAMABnw4rlmzBufPn098ZGTMDtV04dKpKwzRsoiLGwdFgeF0Yw8AoLXbj10n2vo9/8PXMQQjMty+CASBwRXD4IvB1LW6EQrLg/7M5QlDklSIogiNRoBBKwIcEBmDTitAqxHhDcoQBQa3L4K39jbhZINr2POpKocnEIHHHxnT6jmJ3Yo5hVg5t3DIn3PO4bAb8Mk1FTF18SAkU7z77la89NKL+Ld/+w8YDHRXLJOoqoonn/wJ9u7dNejPf/e75/Hmm68PePznP/8Zdu58H9/5zvfxy1++AJvNjq997V4EAoFkh0zSREwr0fn5+Th16hTWrVs34GenT59Gfv7Ik8fI+AsMkbD2e04o+pwDNZ3QDLOqy8AQCMuwmbXQDLGxbyTKhVXwj5JkBaHIhwNWOOcIRhRoPzKsIyIpCEsK9Bd6Wu842oo507MHJGMefxh7TnagtsUNtz+6cdFi1KKi2I7ls/ORazeOKn4Sm2sWFWN6oQ2Hz3ahod2LcESBIAB5WUZUTc3G0ln5Q24OJSQT/fWvb+DHP/4Rbr/9s/jEJz6Z6nBIHFpamrF588Po7OwYULfe0+PCI4/8ECdPHkN+fv8+9Iqi4K233sS3v/09LFlyGQDgwQe/h/Xr1+DIkUO44opV4/Y7kNSJKYm+88478d3vfhd2ux033ngjgGiXjnfffRcPP/ww7r777qQGSUbHoBu5l/LF5zjdIRj1GmhEBmWIRVv5wsCT+eWDj/geydRCK0RRGLD5zB+S+w16UVVAYHzQ1g2BkNw3Lc/jj+BYrROLZ35YWtLS5cOrO+vhD0TAGOvrGBEMyzhR58TZpl58/MrpqCiiVnjJNK3QimmFViiqinBEhVbDBnwpImQieOmlF/HMM0/ijjvuwr/9272pDofE6WLd+qOPPoG7776j38/q6+tgMpnw29/+Nx566Fv9fsY5x+bNP8Hs2XP7Hot+hnGEw6O7W0syT0xJ9Pe+9z00NTXhi1/8Yl+ic8UVV4BzjltvvRU/+MEPkhokGZ2Flbk4ds455M9VlWNmabTHL1j0AmA2auH2RQataeWco6zIhhzb6G5V2kw6TC+0oq7V85Hj9i8TEi4pORkQ8yX5tyAwtDn9WIxoEh0Ky/jfHfUIhKQha3LDkoK/7GrAF9bNGvXvQWInCgJMBlp1JhPTX/7yGp555kn8y7/cgzvv/GKqwyGjMFzd+pIll/WtMn+URqPB8uUr+z326qsvQ1EUzJ07P+FxkvQUUxItiiJ+85vf4P7778f27dvhcrlgt9uxatUqLFy4MNkxklGa4jCjalo2TtW7BmzA45xjSq65b3NXQbYJXT1B2M3RzYe+YP/VYY5oLesnriofU0xrFpegzXUGna4gwpICgEOSORRFgSiKEAQGvSggPMj4bw4+oBfxpUn13uoO+IKDfwG4VCgsY9/pTqxfQS3xCCGj43R246mnHsN1192Aj3/8Zjid3X0/s1pt0Omol/lkcvDgfvziF8/gc5/7pwGlH2TiimtaSlVVFaqqqpIVC0mCjVdMh8Woxal6F7xBCQyAXidiisMEi1GLV7bXQRCAbKv+QvUEQ47NAKtJgTcgRbt5CAwmvRYfu3zamGtZW50+KIoKf0hC5EKirHIVksLBBI6CLCMCIblfjfRFAmOwmrR9f+acI8vy4QdVbZM7pr7DjDGca3VD5XzIFW9CCBnOrl07EQqF8O67W/Huu1v7/ezHP34SV155VYoiI+Nt9+6d+M537sdVV63Gv/zLPakOh4yjIZPom266Ka4Dvf76wJ2rJPUYY1i7pARXLyxCfZsXshJtM3eqvgf8kk4cKudQVQ6VAzpNtBtGji1aw6qoHAsrHVg6K/62dpc6VtuNv+1vgso5CnNMiMhqX7eOiKzCF4zAG4jAbtbBG5T6NQrh4DAbtP1WovVaDZbNjn7j5zzaiSNWvoCEUFgZMMqcEEKG8vLLb/T9+8aNt2DjxltG9Voycbzzztv40Y++h3XrbsQDD3wXAnUdmlSGzCA8nv79fHft2gVBEHD55ZejsLAQTqcTe/bsgSzL2Lhx47gES0ZPIwqYUWLHzmNtOF7nwkcbcQiMQRAZBIGhrNB2YbgJR47dgPnlDlRNzRrTdDlZUbHjWCtU/mFmrNMI0GmiK8kXH/UHJZgNGliMGnj90dpmDg6TXoOcS4apqJxj8czcvk2GwMD2i8NhbHQjpwkhhBAgugL9ox99DzfddCvuu++bNIF1Ehoyid6+fXvfv//kJz9Bb28v3nrrLRQWftgDtqenBxs3bkRJSUlSgySJoagqjp9zDkig+z1HUZFj0+O21RUJPffhs13wBqUhyycYgFy7ARajFhpRQFGuBW1OP3xBCSa9BkaDFgzRFWdRFLB0Zj6uXlj04esZg8NmQJvTH1M8WRZ9TN1LCCGEkI8KhUJ45JEfYuHCxbjzzi/C5fpwE7/FYoFeTxvXJ4OY7mU//vjj+PWvf90vgQaA7OxsPPDAA/jCF76Axx9/PCkBksQ50+SG2x8edsofYwwN7d6En7u1OxBT/bFBJ8Jm0uFfNkbbBvV4Q9hf3YUeb7RlkMNmwPLZBbCZB27aqZqWhZZu34jn4ZxjRqmdVg0IIYSMypEjh+ByOeFyOXHLLf27e3zjGw/glluoX/hkEFMSrSgKXK7Bp8M1NTXRLuQM4QtKMY3JjgzSGeNSgZCEvac70NDmRTAiQysKKMm3YHlVPhyDDDIJhCTUtbrR5gz0DVzRaURYTdpBV4MvnS6YbTVg3bLSGH47YPGMPBw/50T3CBMV7RY9Vsym3dOEkOTinCMYlqGoHKIgwKgX6ct7Ghuubv25537d788rV16BnTsPJDskkuZiSqJvvfVWbNq0CWazGevXr4fFYoHH48Gf//xnfOtb38KXvvSlZMdJEiDXZrhwMR/+Ij7cZrvaFjfe3N2AUFju92Hg8oRwss6FqxcVYfklCWp1Yw/e3nseHc4AItKHHTeCioxAWIbZoIHDZuhXn2wxftiBIx4aUcAn11Tg5ffOobMnOGhbvyyrAbdeVQaTYXTnIISQ4XT2BHH4bBec7iCcnjACIRkqOAQwmC5c7xx2A5bMzENeFk1PJSSTxZREP/XUU2hra8OnPvUpMMag1WohSRI457jjjjvw6KOPJjtOkgDTp1iRn2WE0zP0Si3nHJUlg0/z63AF8ObuBoQHaT/HGIPKObYfaYHZoMHcMgcaOzx4a895SLIKq1kLX0jq/xoA/lC07Z7Dbug7/4yLA2BGwWbS484bq3D0bDdOne9BjzcMALCbdZg1NQtLZtLIaUJI4jV2eLDzeDtaOv39Oh8xBoj4cHJqc5cPzV0+HKt1oiTfjFXzp6C0wDrUYQkhaSymJNpsNuMvf/kLjh07hg8++AC9vb1wOBxYvXo1Zs6cmewYSYIwxrB8Tj7e3tMIPsjPOefItRtx2az8QV+/93QHwhFl+JNw4GBNF+aWObD7RAckOVqaodWIMOg1CIb6r2AzMPhDMmxmFVqNAJtFj2VVg58/VqIgYMmsfCwZ4vcghJBEkRUV2w4241idE6o62JV1cBwcTZ0+vLStFgtn5GLtkmKI1B6NkIwSV5PcBQsWYMGCBcmKhYyD+eW5kFVg17HWAd0ySvOtuGlV2aArtZKsov4j47qH0uYM4HhdN853+Po9nmc3oFMNDhikwsHh9odRNsWOW68uh1ZDXTMIIekvFJHxx221aHP6R13rrHKOQzWdaHcG8Om1ldBp6fpHSKaIKYkuKysb8QJRV1eXkIBI8i2uzMX8shwcre2GyxuGyBiqpmWjKNc85Gt8QQm+kAStGMtKCceJehfwkfVuxhjys43wBWUEQhIiF1apNYKAPLsB/3xjFQx6Gn5CCEl/kqzipW21aB9DAn0RYwyt3T78z7ZafO66GdDEdJ0lhKRaTBnLzTffPOAi4fP5sGvXLnR2duLBBx9MSnAkeTSigKVxlDuIAotrRDbDEP2gL4zuvji+m3MOxhgKc82UQBNCMsbWfecTkkBfxBhDW7cPf9vXiA2XT0/IMUnqRCQFLm8I4Uh0sUivE5BjNdCdhgkm5o2Fg+Gc4zOf+QwaGxsTGhRJP1aTFg6bAb2+8IjP1WpELCjPQUO7Z8TE++IHkHWUHTnGSlFVRCQVOq1A9YiEkJicberFqfM9CW9XxxjDyXoXZk3LRkXR4Bu8SXoKhiUcrOlCV2+0K0uvLxztSHXh5xyATisi26KHw2ZAXnZ0/5Gehn5ltDEt/THG8OUvfxmf+tSn8OSTTyYqJpKGGGOYUWrHvlMdI35wTCu0Yvb0HOw51TFiz2YgWhO4oMKRqFBHJMkqDp3pRHVjL7p7Q1BUFaIowGHVY9bULFxWVUAdPAghg+KcY8ex1o9WqyXu+AB2HG1F+RQb9ZTOAE2dXhw60426Vne/Nq4ABpTlqCqH0xOC0xNCTVMP9ld3omyKDUtn5qEk3zLeoZMEGPP981OnTkGW5UTEQtLcqvlFaGjzorMnMOTF3WzQYPWiYjDGsKgyF+8caBr2g4BzjiKHGWVTbMkKu58zTb342/7G6OCZS+JSFBWdvUF09ASwv6YL1y8txezp2eMSEyEkc9S2uNHZG4yrvC1eHa4A6ts8KKfV6LQVish4e28japp6+1ab4/nSwxhDRFJQ09iDmsYeVE3NxrrlpTDoqKwxk8T0t/WVr3xlwGOqqqK1tRVvvvkmPve5zyU8MJJ+tBoBn712Bl7f1YCGNk9fPTMAcA4UOkz42MqpfT2fl8zKh8sbxsEzXYNWSHPOkWMz4JarR964mgjV53vw5p4GKAof8gOQMYZQWMabexogqyrml4/fCjkhJP0dr3MmNYEGotehY+eclESnqVMNLmw/3AJvIJKwz67T511o6fZjzeJiVE2jBZxMEVMS/frrrw86XMNms+G+++7Dt7/97aQER9KPQa/Bp9ZUwukO4kitE6ELY79nTcvGtEEGBlx3WSkKc0w4ds6Jpk4vVJVD5dHhKpXFdqyaP2VcNhT6gxG8vbcRihLbPVhV5XjnQBNK8szIthqSHB0hJFN09YxcopYInb3BcTkPic+Oo63YfbIdQHwrzyNhjMEbiOD1XfXodgexakFRwo5Nkiem7KWhoSHJYZBUkhUVx+uc8PgjEAWGGSV2FOQM3e4OABx2I65dWhLT8eeVOzCv3AFvIIIebxgGnYhcu3HAWO5k2nu6E2FJjuuiJ0kK9p/uxLrlU5MYGSEkUwRCEnp9YYxHqXKvL4JgWIaRuhaljfcON8e0L2hMOPDBiXYoKsc1i4qTdx6SEDHtnlq7di2qq6sH/dmxY8ewaNGihAZFxofKObYdasYvXjuBrfsasfdUB3adaMdv367Bi+/UoLnTN/JB4mA16TC1wIr8bNO4JtAq5zjT1Bv3hY8xhrMtbiiqmqTICCGZpKXLD5UnaUfhRyiKgtZu/7DP4Zzj61//Cl555aW+xyRJws9//jPccsuNWL9+Db71rW+go6M92eFOeHtPtWP/6c5xKT1kAPac6sD+6o6kn4uMzZBfcV9//fW+DYPbt2/H66+/jlOnTg143rvvvotz586NOZB33nkHb7/9Np544okxH4uMjHOO/91RhzONvRA+0gOaIfph8fL7tbjlqnJMLxyfTX/J0tkTgNMdGlXHDY8vjMYO37htfCSEpK9QRMZ4ff+/uPFsKKqq4qmnHsPevbtwxRVX9j3+wgu/xvvvb8P3v78FVqsNzzzzU3zrW9/A88//YTzCnpDaXQHsOjG+X0QYgA+OtaGs0IbcLOO4npvEbsgketu2bXj66acBRP9nfuCBB4Y8yHA/i8WPfvQj7Ny5E7Nnzx7TcUjs9ld39iXQQ4lIKt7e24h/3Th3XFeOE4VzjprGXuw83opudwiMAYwBeq0Im0kX0+8kCNE6NUIIEUUBnGNcyjk4j55vMC0tzdi8+WF0dnbAYum/F2Xnzn9g48ZbsHDhYgDA3Xf/K+6554twu3tht2clPe6JRuUcW/c1QpLH/45kRFbx1t7zuGPdLGp3mKaGXJr7yU9+gvr6etTV1YFzjj//+c+or6/v909jYyPcbje2bNkypiCWLFmChx9+eEzHILHjnONUgyumJLLXG8aJOuc4RJVY+6s78H/fPIX/3VmH1u4AIpKCiKQiHFHh9kXQ0u1Htzs04q1ZzkEthwghAIC8bAN4shpEDyI3a/BNzSdPnkB5eSWef/4PsFj69xe22Wz4xz/eg9PZjUgkgrfffhMlJaWwWulu2mjsOt6GthHKapKppcvft5GRpJ8hswOdTodp06YBAOrr61FUVAStdmxT5f70pz/hhRde6PfYli1bsGHDBuzdu3dMxyax6/WF0eYMQIxxJfZsixsLKnPHIbKx45zjb/ubcORsFxiLlqloNQIEgeFivsxY9N/9QQmSrCA/yzjkio9GI6DQYRrH34AQkq5ybUZYjDoEw8mfjWAzaZFt0Q/6s3Xr1mPduvWD/uzee/8T3/72/bj55vUQRRFWqxU///n/hUATWeOmqhzH62JbcEoWQWA4XufEyrmFSW+tSOI37BIb5xxbt25FSUlJXwJ9/vx5PPzwwzh9+jTmz5+PBx98EGVlZTGd7Pbbb8ftt98+5qCzs03QaDJzVGZe3sA2cOPNE1Gg0YgDpikNRaMVxxy3oqjYe7IdZxt7EIwoEAWGLKsel8+fgpL8xP032bqnAScbXNDpPvzCpwVgNmrhD/b/4GOMQZI5ur1hFOdaBr1dNmtqFiqmpU+v6HR4/5DMRe+fsSspsKKhzTsO57EhP3/k1WNRFGCxGPr+bg8dciE/PxcPP/wQsrKy8PTTT+P7338Qf/zjH2E0jr62djK+d/adbEcwokCrTe3dSF9IRmtPCItn5ac0jrGYqO+fId8ZPp8PN9xwA/bs2YPNmzdj3rx56O3txZVXXonu7m5s2LAB+/fvx/Lly3HgwIG+Vevx0NMTGLdzJVJenhVdXcm/+I5EDkmQJRlcjS2JliVlTHHvPtGOw2e74PFHBnyj33eiDcW5Zly/rBT52WNb8Q2EJGw/0ARZUQH0r18z6jTRxviDjH0JhWS4PEHYTLp+jysqR8WU9Pg7A9Ln/TNZ/PWvb2DLlu8P+rNnn/0lFi1a0vfnhx/+Nrq7u/Dss78cr/DiRu+fxCjKMaLmvCupq4Iq5yh2GGP6+1IUFT5fCF1dXvj9Ptx///3YvPkxzJ27FADw0EObcdttH8cf//gqNmzYOKp4Jut7Z/exFsjy0Js7x9POw80oycnMDYaZ/v4Z7gvAkEn0o48+ijNnzuCNN97AunXrAAA//elP0dbWhueffx533nknFEXBunXr8P3vfx/PP/984iMnSZFt1aPQYUJX78hDA1SVo7xo9LV07+xvxMEzXRAYG/KWWEu3Hy9tq8WtV5ejOM8y6HNisfd0ByRZGXRF2aAXYTZo4Q9JAxJpxhgCIblfEs05R9XUbMwspY04k9W1116PFSsu7/fYli0/gM/nxbx5C/oe27FjO959d2u/pJpMXEtm5mPf6c6klnRYjVosmpEX9+vOn29AMBhEZeWMvsfMZgtKqD3BHAAAIABJREFUS6eitbUlkSFOeGFp5BaD46ml2w9JVkfVZYokz5B/G6+88goefPBBbNiwARpNNNd++eWXkZ2djS984QsAAFEU8eUvfxlbt24dcyArVqzAk08+OebjkJExxlA1NRuqOvIGGbtFh4UVo6uHPlTTiUMXEuiRBMMyXttZD2mU3/o55zjb7B5yBzMD4LAZYDZoB90WFJYUhCPRc6sqR2VJFm5aNZ12RE9ier0BDkdu3z8nT57A4cMH8NBDP+y7Jno8bjzxxI8xf/7CFEdLxotWI2BBhSNp/aJVlWNBRW7M5XaXys2NJt61tWf7HguHw2htbUFxcWzDsUhUY7s3JR05hhKRZDR2ZO5q7kQ15P+lDQ0NWLx4cd+fOzo6UF1djdWrV/fboFBUVITu7u7kRkkSbuXcQlQU24f9INBqRNywfOqoNlVwznGktjuuJNQbiOBAdWfc5wKiUxfdvvCwz2EsOm48x6qHTiv0S6YZGEIRBVMcZly3rBS3Xl0OkTbikAtkWcYvfvEMbr/9s/2Skaeeehxr116HuXPnpzA6Mt6uWliEwpzkbDguyjVj1YIpo3ptfn4BrrrqGvzsZ4/j6NHDqK+vw5YtD8NoNGHNmusSHOnE1ur0x7T5fryIgoBWZ/qsjJOoIbMEo9GIQODD2uPt27cDAK6//vp+z2tubkZWFt3yzjSMMdx2TQWWzMyDQacBvySZVsExxWHGJ64uQ3mRfVTHP9fqRmdvMO6Yqht7+8USq4isxrSyzgBYTVoU5phQmGOEzayD1aSFzazDyrkF+MINs3DZrHxagSb9bN/+d3R2duDzn/+nvsd27nwfJ08ex7/+67+nMDKSCgJjWL9iGgwJHslt1GuwfsXUMV1/vvvdH+Cyy5bj4Ye/jXvu+SJCoRCeeea/YDAM3i6PDK7HO/yiTCq4POkX02Q35BVg+fLleO2117BhwwYAwIsvvghRFLFxY/+NCb/97W+xdOnS5EZJkkIQGNYtm4o1i0twpLYL3oAEQWCoLLKjJH/0tckAcK7FM6qNNx09AfhDMizG+Nop6jQiBFEAjyGRvkivFaHXftjlZawbG8nE9eqrL+PGGzfCZot+qfR4PHj88Ufx0EM/pORkkirMMeGmK6fj9Z0NCEXGXh9t0Glw86rpcV+HXn75jX5/NpnM+NrXvomvfe2bY45pMgtL6VPKcdFoyx1J8gyZRG/atAk33HADmpuboaoqtm7dirvvvhtFRUUAgP379+Ppp5/GO++8gzfeeGOow5AMoNUIWFZVkNBjjraWTFE4QpH4k2iNyJBj0cPpGXmz5GA4gLIpE7MFDxmb7u4uHD16GP/+71/te+xnP3sMq1ZdjSVLLkthZCTVphfacNs15Xh7byO63cFRrSBzzpGXZcSNK6dhisOchCjJaKhq+iXRShyLRGR8DJlEX3vttXj99dfx+OOPo7OzE1//+texefPmvp9v3LgRXq8XTzzxRN9qNSEXjbaWTBAYdKPoAc4Yw8zSLOw60TaqD7KSPDOtRJNB7d79AfLzCzB79ty+x7ZufQt6vR5bt/4VACBJElRVxfXXX4Xf//5PKCwsTFW4ZJwV51lw14YqbDvYjGN1TsiyGtM1iHMOrUbE/AoH1i4ppj0YaSYdh9OkU402iRq2oGvDhg1DJsivvfYaZs6ciezs7KQERjJbocOEo7XdcW9KzLLo416Fvmj5nAIcPNOFiBTfLS9V5ZhXlj4DVUh6OXHiGBYuXNwvMfqf/3m133P+8Iff4ty5Wnzvez9Cbm5mTPckiSMKAq5fNhVXzCvE3tOdONfqRo8nDFXl/a6BF/+cYzOgotiGFbMLYDKMbRIwSQ69Nv2SaJ02M4fMTWSj3hWxYsWKRMZBJpgFFQ7sPtEOX1CK+TWcc8wosY96xKpeK2LlnAK8f7R1kJEqQ59zWqEV8ysoiSaDO3euFtdcs6bfYyUlpf3+bLFYodfrBzxOJhezUYe1S0qwdkkJgmEZda0e9HhDUDkgMCDHZkDZFBuMCd6QSBIv2zr4yPVUGmoMPEkd+j+ZJIUoCKiamo191R0xbzDUakQsnzO22uyVcwsRlhTsPdUJDNoR+kOcc5TkWfDJ1RVJnT5GMpvL5YTVOvqBQ2RyMuo1mFuWk+owyCgVOcxQVJ42JRSKqqIol2rm0w0l0SRprllchPaeAJo6vCPWCDLGcN1lJQNGb4/qvIuKkWPT4/CZbrQ5A/joqVWVw27WY+ZUO1YvplpEMrw///nNEZ9z773/OQ6REELGy9RCK7QaIabWqeNBp9VgagFtfk83lESTpBEFAZ9aU4k3PmjAmebeQUssOOcwGrRYu7gY88oTV1IxvzwX88tzcb7di+N1TgTCMjjn0GlFTC+0YkGFg5JnQgghg9JrRRTlmtHc6Ut1KACA4lwzjfxOQ5REk6TSiAI+cXU5OnsC2F/dicYOH8IRBYLIYDfrMKs0C0tn5Sft4jCt0IpphfTtnRBCSHzmTc9BY7t31Pt0EkVROeaVU2lQOqIkmoyL/GwTPnb59FSHQQghJA6youJccy+O1XTAG5AgKxyKqkIQGESBwWSIToAtm2KFQTexUop55Q7sOtkOjz+S0jhybHrMmU5JdDqaWO94QgghhIxJc5cPR852o7s3CJc3DBWAqgw9fERVOURRQJZFB4fNgKpp2Zg9LXtM48vTgSAwzCvLwc5jbSlbjVZVjgUVubT5PU1REk36hCUF7c4AQhEZNrMOhTmmjL8IEkIIGZmiqjh8pgunGnrR6vJBuGQXi1arGTaJFgQGzjl6vGH0eMOoaerBB8eNqCyxZ3wv7CvnT0FdmxftTn9Kzl+Sb8HKMXatIslDSTRBa5cfB8904lyrB8GQBMYYVM7hsBswoyQLK+dk9kWQkHTAOUePJ4z6dg8kWQXnHKIgwGE3YGqBBdpRTOokJBEOn+3CruPt8AYjEBjrl0CPhigI6PGGse9UB46c6ca8DJ7KyBjDDctK8eK7ZyDL4zsKXKsVsH75VFrMSmOURE9yu0+2Y+exNnAebeMjitGLnAAGty+C/ac7cLqhBzevKkNJviWVoRKSccKSgn2nO9DU6YPLHYI3KEFg6PehqKgqtBoR2RY98rONWFCRS5thybjwBSL4697zqGv1RJPnBCdrjDFIiopDNZ1o6vDi+mWlKM3PvPd2QY4JV84rxPtHWsftnBzAqvlTkJtlHLdzkvhREj2J7a/uwD+OtA7oo3wpxhj8IQmv7qjDZ6+tRG6WafwCJCRDdbuD2H2iA+da3AhLcl/SPNjgBlGI9qJ1ekJwekI41dCDKQ4T5pTlYOnMPFqFIklxpLYbO462IhCSkl5vyxhDtzuEl7bVYlFlLtYuLcm4Gt8VcwoRDCvYe6p9XP6fvHxuAZZVURlHuqMkepIKSwp2HW8fNoG+VDAsY9uhFnxq7YzkBkZIBuOcY9eJduyv7kBEit76jfcDlzGg3RVAS7cfNY09uHHFNOTYDMkIl0xS7x9pwZ5THWCI//05FqrKcaCmEy5PCLdeUwGNmFnlHdHhXAy7T7WPNBB3TFYtmIIr5k1J3glIwmTWO5gkzIHqToQiclyvaezwodcXSlJEhGQ2tz+MF985gx1HW/sS6LEQBYaWLj9+t7UG+6s7EhAhIcDfDzZj98n2MVY9j57AGOrbPPjje7WQh9msmK6uWliEm64og9Wo6yuDTATOOWxmHW6+qpwS6AxCSfQkVdvcG/cKhKKqOFjTnaSICMlcTncI//3uWbR2+xPeCisiKdh2sAXvH2lJ6HHJ5LPjaCsOVHekvJSCMYamDi/+/P65tBmrHY+qadn4549VoWpaDhKVR8+ZnoO7N8zGrNKsxByQjAsq55ikfKH4VqGB6IUvMIrXETKRuf1h/Gl7bVIHMjAG7DnZDoExXLWwKGnnIRPXiTondp/sSJsae8YY6to8+Nv+RqxfMS3V4cTNoNPg5lVlWDozD4fPduFcqwfhiBzzf1/OOfRaDSqK7Vg6Kw9FueYkR0ySgZLoyWqU3555MgvBCMkwKud49R/14zLRjLFoLWZelhFV07KTfj4ycQRC0oU7Gel1/RYYw/E6F2aWZqO8yJbqcEalJN+CknwLwpKCA9Wd6OoNwuUJweUNIxxRcPHGlMoBg16DbIsOOTYD8rONWDorH3ottbbMZJRET1Imgwb+kBTXazjnMOnpLUPIRTuOtqLd5R+/2+MceO9wC6ZPwBHLJHne3tcEX1BKm1XoS3HO8e6BJvzzhtnQajK3wlSvFXHl/A9rmSVZQY8vjHBYAVh05TrLos/o35EMRH+bk1RFsS3uTRGCwLB4Rm6SIiIks7S7AjhY0zXu9aUefxhb9zWN6zlJ5jpR5xzVHpjx1OMN4d0DE+s9rdWIyM8yobTAitJ8K/KyjJRAT0C0lDFJLavKx8GabkiyEvNrSvOtcNgnV+N3pzuImqZeuDwh+IIyVJVDIwqwW3TItRswr9xBt+MmqZ3HWlPSXYAxhjONvejsCSA/m/q2k6FdbLmY7hhjOFHvxLLZ+cidZJ8xJLNREj1JGfVarJiTjx1HW2NaoTBoRVy9aPK03TnT1ItDZ7vQ2O4D5+qg/40459hxrA0VRTZcPreQJktNIh5/BA3t3pSdnyPab3fDyukpi4Gkv5MNLjg9oUGH/KQbVeU4UN2ZkZsMyeRFSfQkdsW8KZBkFXtPdwy534RzDpNBg49dXoYix8Qf+x0Ky3h7XyNqGnv7BtEM9SWDMYaIpOD0+R7UtrixfHYBrphXmNa3TUli7DvdAUUZ/MvVeKlt9kCSVbpFTIZ0ss6VEQk0EL2e1ja7IV+mZtwQFjJ50Tt1krtmUTFuu6YCFcX26PhhzsE5h6JymA1aLJ6Rhy/cMCtjd07Ho8cbwgtba1DT2BPzJMeLJFnFjqOteOX9c1DUzBsgQOLT0O5N+ZelYFjCyXpnSmMg6cvpDqKxM3V3S0bDF5Jw+GxXqsMgJGa0Ek1QUWRHRZEdvmAE59t9CEsKrEYtyottEIXJ8T3LF5Tw0nu18Pgio06OBIHhXIsb/7ujHrdeXZ7yJIskR1hS0OsLpzoMMMbQ2RtMdRgkTR060w1V5Rl1HRIYQ02jG8uqClIdCiExoSSa9LEYdZhblpPqMFLir3sa4PaGx/yBwxjDmaZe7D3dgZVzChMUHUknje1eSLKaFrfJnZ7UJ/MkPXW7gxmVQF/kdAehcp7yqYqExGJyLDMSMowjtd2oa/Uk7ANHFBj2nuxAjzeUkOOR9NLmCqRFAg0AvV5KoslAnHO4PJl5/fGHZHT2BFIdBiExoSSaTGqccxw+k/hev2FJiW7YJBNORIq9LWSypVMsJH30+sLwBOIbppUuRAEp7XxDSDwoiSaTWn2bBx1JWvWobXanpI8wSa50+htV4xyYRCaHulYP0uRmSdwYY3C6M3MVnUw+VBNNxoXHH0H1+R64vGGEJQUCY7CatSjJNaOi2J6y2r2zzb1Jq73zBiTUt3owozQrKccnqSGmUXIiTJKNvyQ+3kB6jviOlS+YmavoZPKhJJoMS+Ucpxpc6OwJQpJVaAQGm1mHRTPyYupP29Tpxe6THTjf7oWqDuyru1tVkWMzYM70HFw5bwqEcV4+cSVxY5YoMLQ4/ZRETzAGvQacp0fXA6OepmWSgTL9Dpii0B0WkhkoiSaD8oUk7DvVgTNNvejxhvttpOKcY8+pDlQU27CsqgB5g0zq45xj26FmHD7rhHqhb/JgSYcoCHD7IvjgeBtqm924ceU0FOaM3yhjtz+S1ON7knx8Mv6mFVjwvsqhSYMl6RybPtUhkDSU6WU+iprZ8ZPJg+4FkgEaOzz47V9PY//pDnj8kQGdCBhjCIZlnKhz4fdba3CktrvfzznneG1nPfaf7uxLoEciMIau3iD+9F4tmrt8CftdRpLsFQ8lw1eEyEBTHGaY9Omx/pBro1HzZKBMbw+XLt1vCBkJJdGkn+ZOH/78j3oEQnJMt6tlRcU7+xpx+JJEevuRFlSf7xlVaUYwLOONDxoQCstxv3Y0kn2xpprViUcUBDjsqU9eVc5RnGdOdRgkDYlpcJdkLDI9fjJ50Cc86SPJCl77oD7utlkcwLaDzWhz+tHW7cehmq4x1TZ7/GFs3d806tfHw2rWJfX49iQfn6RGZYkdPMW3zHNsBlQW21MaA0lPFqMu5e/PsbAYtakOgZCYUBJN+hyo7oQvMLoaXkVRcbCmC7tPtY+5nu3i1L/xGFbisBmSdmxV5Sh0jF99Nxk/l83Kh1Gfug96zjlmlWalxeZGkn4qimzI1Byac45sa/Kuy4QkEiXRBED0wlXd2DumD+UzTb2oa/EkJB5VVXGguishxxpO2RQr1CRtYjEaNKgoopXCiUirEVK6Gq3TarCsKj8l5ybpL9uqh8WUmau5igpML7SmOgxCYkJJNAEAnO/wjnnUao83BGeCVo8ZY+jqTf7o15mlWcgdpLtIIlQW22NqA0gy09rFxbCaxr9cR1U5llXlwWTIzCSJJB9jDDkZupprMmjGtUMTIWNBn/ATgMo5JFkd06pYpysY1yq0qnIEghLcvgh6fRH4ghIikprQVV2XN3k9nC9ijGFBhQOJXozWaASsmF2Q2IOStGLQa7B6cfG43zYvybPgyvlTxvekJOPk2jMzic61G8Z9XgAho5UefZpI3MKSgv2nO3CmyY0eXxgq59BpBEwtsGLJjDxMi/N2WCzN+Tnn8Idk+IMSwpLab+AE5xySokIUGbSiAKtJO+Z6zchHzpEsy6ryUdvsTlhrPVXlWDYrP2kr3CR9zJmeg9pmN06fd41LfbJeK2L9iqlUC01GtGhGHo7WOsGROcXRnHPMKKESOJI5KInOQGcae/DWviaEwhI4GHzBCGSFAxzocAVwusGFGaVZuGVVeczlBJoRnicrKlyeMIIRGQzRD/BLP8gZY2CMQVWBHm8YgZCMHLseOs3oJ6qJAhuXZIExhhtXTsV/v3t2zONmOeeYWmjFlQtopXCy+NgV0+APSWjs8Cb1/arRCPj4FdPpyxmJSX62ESX5ZjR1jl/f/bEyGbRYMpNq/UnmoHKODHO2qRd/2d2AQEiCyxNGa7cfPd4wfAEJvqAEtz+CVmcAe06244W3qmMuryjOMw95W1qSFXT0BBEMf5hAD4Zd/IcxRGQVHa4gQpH42uVdym4Zv3rTbKsBt15TAYtRO+qyGM45SvItuH11ZcYPOyCxEwUBt6+pRNkUW9Imxem1IjZeOR0V1NKOxGHu9JyMmv5XQftISIahd2sGUVWOdw42IRRR0OEKwBeSoKq8X2LLwAAOhCMqjpzrwv9+UBfTsYtzLSjKHTi4gXOOrt4QFGXksgqtRujXJJ9zoNsdjKlUZDDJbD83mMIcE+5YNxPTCm1xJ9KCwLBkVj4+c+0M+hCYhDSigE+ursSyqoKEDopQOUd+lhGfWluJGcVZCTsumRzmVTiQbc2M0fAMDJfNolVoklno0z6DHD3XjV5vBF29IcgKH3ZVGADAGXYda0dXjz+m41dNyxqwktbjDUOSY0uCTQYN9Nr+5RuKwuHyxL9BkHOOuWU5cb9urGxmPT69thI3LJ+KwhzTsJvGOOcQBIbyIjs+vbYS119WCpEmFE5agsBw7dISfHrNDORlGce0yZZzDlEUsGJOAf5pfRWmOGgyIYmfwBgun1uY9oNXOOeYPT0L+dlUqkQyC9VEZ5Dq8z0IhCVEZGXkBPqCYETGX/c04s4bZw/5nP/6r5/jnXfexv+89BoOn+lCry86cEVRVPhjHP/NGGAz6RCWFDilUL+66WBYRjiiQK+LvT66MMeMsim2mJ+fSIwxLJqRh0Uz8nC+w4v6Vg+cnhC8gQg4B0RRQJZFB4fNgPnlDthoKiG5REm+BXeur8LRc9043dCDpi5fjP+3RleerUYdKkvsWD47n4ZOkDFbWJmLM829qGtxp+2G1CyLHjcsn5rqMAiJGyXRGcQbkBAIDV+X/FEMDPVtHoQlZcAqMQBUV5/G//t/v0NeXj40ooBPXFWOl7bVIhCW4Q1eKBcZ6cLLoiOItRoBGo0Af0hG6CMbEH1BKeYkWiMyrFtWmhYX/GkFVkwroMb/JD6CwLB4Rh4Wz8hDh8uPE/U96PGG4PKG4fGFIasX7mQwBqNeRLZVjxybAUUOMxZUOuiOBkmoG1dMxW//Wo1AWE51KAMxYO3SEmjHsAmdkFShJDqDSLKCsBT7KvRFEUnFiTonln6k3kySJGzZ8jDmz1+I9vY2AEBetgmfWluJV/9Rj46ewLCJLEe0nMFhM8Ckj76VGKJ9PjtcAUiK2hdrJMaSEIDhyvlFmDJIfTYhmaggx4yCnA/fz2FJgSRHe6prRAaDXkMbUUlSWYw6XLWwCFv3NaY6lH4455hX5sCMEqr3J5mJljsyiCgKiLflJ+ccGo2A4CArEL/5za9QXFyKNWuu7fd4frYJd39sNhwXVpc/2meUg0MUGWwmHYoc5r4Eui9OgSE/2widRux7rSQrI9blaTQC1iwpwoo5NKSETFx6rQiLUQubWQeTQUsJNBkXCytzcVlVftrUR3POUVpgxfoVVMZBMhetRGeQ6YVWHK3thhjHh65GZDAbNBA/MgGqpqYar7/+Kl544b+xffvfB7xOqxFgN+uhEQWEItEV8OjFl0GrEWAyaIZdD9eIAgpzTOj1hS90EYl26xgsdI7oFLZrl5bQuFdCCEmSNYuLEZFUHKntSumXN845inKj7UCpdIlkMkqiM8hVC4rw1t5GKEpsKwmccxj0WnCOfrv7L5Zx/Pu/fxUOR+6Qr7/YqsugE2GIY1PgRYwB2VY97BYd/EEZlcV29PojiEgKBMZgNevgsOkxv8KB4lxL3McnhBASO8YYblheCq1GwIGazjgLAxODc46pBVZ8cnUltQMlGY+S6AxiNmpRVZqFkw0uYJjLnxzxo+XIq/B21EDUaDB36bUo/eyi6M9kGf/n/3wR58834Gc/ewI1NadRVFQ86HGyLHp4/JExxy0whqmFFnxyTeWYj0UIIWT0GIu2YrRbdPjgeBvCYxiINZpzL6hw4Ppl1A6UTAz0Ls4wt1xdDrtZN6BO+VL1u3+LoLsFl62/B5fdcA/a6g7hV796DgDwy1/+HGfO1AAAIpEwXnnlj3j22afQ0dGO66+/CkePHu47Tq7dkLD6uVw7teoihJB0cdmsfNy5fhamFliTNmnzIs45sqx6fHJ1BdavmEYJNJkwaCU6w0xxmHHDiqnYdqgZ/qCMUETpG+rAGSB72+DvPoerPvltZOVNRdkUG+5YNwNf/eo9uOuuL+HVV1/Gpk0PYtGiJQCAgwf348knf4KsrGw8++wvkZeX13eupbPycPhs17ADR2KhXtiBTQghJH1kWQz49NpK7K/uxN5THQiEpIS2Fr04NGhemQPXXVYKjUjJM5lYUp5Ee71ebNq0CT6fD5Ik4YEHHsDixYtTHVZau3phMRgY9p7qQESSISk82oVDFNDl9UDUaJGVNw0zS+24aVUZZCmCSCSC6upTCAaDWL36Wlgs0Rrk7OxsPPbYFqiqipKS0n7nybYaUFmShTONPWO6sBblWlBZbB/T70wIISTxGGNYPrsAi2fk4UB1B6obe9HhCkAQRn/NV1UOu0WPGSV2LJ9dQAOpyISV8iT6N7/5DVauXIm77roLdXV1+PrXv45XX3011WGlvasWFmF+RQ72ne7CuRY3fMEIOBistiwosoSbVuZjdkUJAKClowMA4HR2w2g09iXQAGA2W6DRaKGqg9fF3bCsFK1dPvhDo2vSrxGFtBmcQgghZHBajYDL503ByrmFqG/z4ti5bnT1BuH2haGow/f5VzkHA2Az65FrN2BmaRbmlzvGlIgTkglSnkTfdddd0Omi31IVRYFer09xRJkjy2LAumWl4JeVQFZUqCoAPgdndr2IF3/zDO6//zvgXMWzzz4JURQRiUSg1Q5cETAajfiP/7hv0HOYDFp87IrpeG1HPcJSfBtQREHA9cuobR0hhGQKxhjKi2woL7IBiA4HcodknDrXDa9fgqKokLkKAQyiGG13mp9lRHmRDVYTrTiTyWVck+g//elPeOGFF/o9tmXLFixYsABdXV3YtGkTHnzwwRGPk51tgiZDR4Tm5SV/hPRzz/1/uO+++7Bhw1oYjUbce++9qKk5DbvdBFmWBsQgyxLy87OHjC0vz4r8XCteea8WXSNMMQSidXB2qx4fv7IM8yqGbqFH4jce7x8ycdH7h4xGCYC55XQtJ6M3Ua89jKfB+KKamhrcd999+OY3v4lrrrlmxOd3dXnHIarEy8uzjmvsPT0umExmqKqK66+/Cs8992vcc88X8be/vQ+TKdo32u/34YYbVuO//uu3mDt33rDHkxUV/zjaiprGXvT6wgMGuKgqh9moRWWJHasXFcGo1ybtd5uMxvv9QyYWev+Q0aL3DhmLTH//DPcFIOXlHLW1tfjqV7+Kp556ClVVVakOJ+N5PB488MB9eOihH6KwcAoA4O9/fwcORy5mzpwFo9GIo0cP4/LLVwEADh8+BKPRiMrKGSMeWyMKWLukBKsXFeNcixutTj98QQmcAya9BgU5JswszaIG+oQQQgiZ8FKeRD/xxBOIRCLYvHkzAMBiseC5555LcVSZy2azIRKJ4Omnf4p77vkPtLW14sknf4wvfeke6PUGbNx4C37608fwne+YwTnHU089hk984vYha9F7fSEcOetEtycItzcCWeEQRYYsiw45NgOunF+ILAv1gCaEEELI5JIW5RzxytTbAuN1S6O1tQWPP/4Ijh8/iqysHHzmM5/Dbbd9GgAQiUTws589jr///W8QRRHXX38j7r33P6HR9P8+5fKE8PeDzWho90JV1UHroDnnEAQB0wutuHZpCXJslEwnU6bfEiOpRe8fMlr03iFjkenvn+HKOSiJHkeZ8kY6WNOJncfa4urGodeKWLVgCpbOyk+5NmTNAAAbqUlEQVRiZJNbprx/SHqi9w8ZLXrvkLHI9PdPWtdEk/Sy60QbdhxrQ7zdPcOSgncPNiMUUXDl/ClJiY0QQgghJF3QDjDSp/p8D3Yejz+BvogB+OBEG6rP9yQyLEIIIYSQtEMr0ZNYMCyjucsPWVGh0wr4+6FmYKzFPRx473ALpk+xwqCjtxchhBBCJibKciYZbzCCvSc7UN/mQY83DEWNjmvt8YURCErQakWYDBpYjdpRj+r2+MPYfbIdaxaXJDZ4QgghhJA0QUn0JME5x97THdhzsgPhiNyXIIsCAwcQjijgYIhIKsKRMPxBCdlWAwy6+CdDMsZQ2+zG6kXFo07ECSGEEELSGdVETwKcc7z+QT3eO9SCiKQMSGwlWYWkqH1/ZoxBkjk6e4PwBaVRnbPLHUS3JzimuAkhhBBC0hUl0ZPA23sbcbqhZ8CY7otCEXnwF3KgxxNGMBx/Ii0yhvrWzG1pQwghhBAyHEqiJ7gzzb04XueEMEQCDQCKwsGG6MnBAfR4I1DjbCfOGBv1KjYhhBBCSLqjJHqC23W8bczHkGQVHn8k7tel2xwfzjm+/vWv4JVXXkp1KIQQQgjJcJRET2AN7R60OwMjPk8QGPgwve0YYwiE5LiSYs45jGnU4k5VVTz55E+wd++uVIdCCCGEkAkgfbIcknC1ze5hyzgu0uvEaN3GME+VZBWyokKria1bh6JyTC20xBhpcrW0NGPz5ofR2dkBi2Xo8Z2EEEIIIbGilegJzOUJxfQ8vUaEKI78VghGlJjPbTPpMMVhjvn5yXTy5AmUl1fi+ef/AIslPRJ7QgghhGQ2WomewEIxJr2MAUa9CH9wiC4diJZ0KErs5RwVJXaIQnp8R1u3bj3WrVuf6jAIIYQQMoGkR5ZDkiOOQSc2s27Ep8d6OJ1WwIrZBTGfmxBCCCEk01ASPYFZjdqYn6sVBdjMuiE3GHLOoYmh5ENROZbPLkCOzRDzuQkhhBBCMg0l0ROYw66Pq6OGzayDSa8dPJFmDEb98NU/qsoxe1o2Lp9bGG+ohBBCCCEZhZLoCWxBhSOmDYMXMQC5WQZYDNoBabRBKww58RCINvdYUOHATavKBowVJ4QQQgiZaGhj4QSWZTGgrNCGc63umF/DADjsBhj0Mnp9YciKCgAwD1EaoqocWTY9rllYhNnTchIRNiGEEEJI2qMkeoK7blkJWt7yIxQZuvPGYMwGDUwGDXyBCHRaEXlZRvgCEagcEBhgNevhsOkxszQL88odEGj1mRBCCCGTCCXRE5zdrMeaJUV4e29T3GO4GYCSPAs+f/1MWEw6hCUFiqJCFAXotbENXUk3L7/8RqpDIIQQQsgEQEn0JDC/PBeyCmw/1AxJVmN6DeccDrsRt11TDotJBwDRxDlDk2dCCCGEkESiJHqSWFyZi9JcE7bub0Zzp2/Ins+cc2g1IuaW5eDapSUxtbUjhBBCCJlsKImeRHKzTPj89TPR1OnF8ToXnO4Q3P4wFBXQawRk2wwoyDZi+ex8mAyx95gmhBBCCJlsKImehErzrSjNt6Y6DEIIIYSQjEX36gkhhBBCCIkTJdGEEEIIIYTEiZJoQhJE5RxqnG0ECSGEEJKZqCaakDhJsoqzzb1odwXg8oTg8obhD0pQL3QPFATAbNAix6ZHjs2AgmwTZpZmQauh76yEEELIREFJdIbyeNx46qnHsXfvLmi1Otx88624884v4u2338SWLd8f9DXPPvtLLFq0ZJwjnTg6e4I4UNOJ2hY3AkEJgjB4n0BFBXp9YfT6wqhr9UBVOUwGLSqKbVhWlY/8bNM4R04IIYSQRKMkOkM9+OAm9Pb24pFHnoBGo8GPf/wjhMNh/PM/fwkrVlze77lbtvwAPp8X8+YtSFG0ma3dFcD2Qy1o6vSBI1quMVQCPRhBYAhFZJysd+FkvQul+RasXlyMKQ5zskImhBBCSJJREp2BzpypxpEjh/D883/AzJlVAIBNmx7EV77yb7jrri/B4cjte+4//rEdhw8fwO9//0doNPTXHQ+Vc+w81oYDNZ2QY5z0GIumTh/++92zWDorD1ctLIIw1OQbQgghhKQtKtLMQM3NzdDr9X0JNADMmDETkUgE1dWn+h6TZRm/+MUzuP32z6K4uCQVoWYspzuE32+twa7jbQlNoC+SFRW7T7bjd29Xw+kOJfz4hBBCCEkuSqIzUE5ODsLhMNzu3r7HOjo6AAA9Pa6+x7Zv/zs6Ozvw+c//07jHmMnaXQG8tO0sOlyBuMo24iUwhs6eIF7adhZtTn/SzkMIIYSQxKMkOgPNmTMPRUXFeOyxR+D1euHxuPHss09CFEVIktz3vFdffRk33rgRNps9hdFmlq6eAF5+/xx8QWnczukLSnjl/2/v/qOauu8+gL9vQhIgCb9/GRUoiBVx+AutW5Uftmql1XVzOvWpbfU8TNtp19Va1q5TZynUdlXrbE/dbK1ndmfC2GmP9bHOdiJTsbPWQrVSFXFYUYiA/AoJJPk+f/iYMx4EGwjcJLxf53CE++N7Pzd+uPnky/d+7+EK1NSbBuyYRERE1Dcsoj2QWq3Gyy+/hkuXLiIzczrmzZuDceMmIiAgEFrtzZvVrl83orT0FGbPflDmaD1Hq7kDhcUXYRrAAvoWk9mKwsMX0WJqH/BjExERkfN4p5mHSkgYid27C9DQUA9/fy3sdjveeusNGAxDAQAlJUcRERGJxMQkmSP1HPuPV6GxxQJJphv9mk0W7P+sCvMzRshyfCIiIvru2BPtgZqamvDkk/+Na9euIjg4BBqNBseOHUFoaBhiY+8CAJw+XYaxY8fLVhB6mtIL11Fx5Yasr5ckSaiobsSp80bZYiAiIqLvhkW0BwoICEB7ezu2bt2Ey5er8K9/HcfmzRuxdGmWowisqLiAuLh4mSP1DCZzB/5ZdtUtPnAoJAn/LLvKYR1ERERujkW0h9qwIQ9mcxuWLfsvvPZaHpYuzcLDD89zrK+vr4NeHyBjhJ7j2OlraG1zn6K1zdyBktPX5A6DiIiIesAx0R7KYBiKTZu2dbv+b3/bN4DReC67XeD8t41u0Qt9iyRJOHelEdPtdigV/JxLRETkjvgOTYNaWUUdGlsscofRRXNrO768UCd3GERERNQNFtE0qJVX1ffrA1V6S6GQUP7vBrnDICIiom6wiKZBSwiBmgb3feS28UYb7HYhdxhERER0GxwTPQgIIXC9sQ2V1c1oMXdA2AXUaiWiQrSIidRBrVLKHaIsjI1tMJk7oHTDnmjg5gNYahtMiArVyh0KERER/T8sor1YS1s7jp+pwYUrjWhotkAhodMNdDa7gEalREyUHuMSwhBvGFyPB6+sboab1s8AAKUCuHStmUU0ERGRG2IR7YWEEPj8m1oc++oazO1WSJJ0295WpUKC1WZHxZVGnP+2EQnDAjH7nmj4+6pkiHrgNTTL93TC70KSJDQ0u99Nj0RERMQx0V7Hbhf44J8X8enJb2HpsH3nIlEhARe+vYH39pejuq6ln6N0D+1Wm9wh3JHFA2IkIiIajFhEexEhbhbQ5y7fgKIXPaySJKGlrQN/O3wRNQ2mfojQvXjCTXt2m/vHSERENBixiPYiJWeu4ZvLN/o8RMFktuJ/jv/bI4rMvnDnoRy3uOP0e0RERMQi2mvUN5nxr7M1LptpoqbehMOlV1zSlrtSKd2/QPVR8leUiIjIHfEd2kt89nUN2jvsLmtPIUk4U1kPq811bbqbAK0aQrh3b3ugTi13CERERHQbLKK9gNVmx4UrjS5vt6WtA6fOG13errsYHqmHzY2HrNjsdhg4vR0REZFbYhHtBSqrm9DS1uHydhWShMu13jtTx9AwLXzV7jvLo1rlg+hIvdxhEBER0W2wiPYC1XWmfnvqXn2T985T7KNUIDRAI3cY3QrRa6Dy4a8oERGRO+I7tBdoNrX3W9stJtf3cLuTOEOgW46LFkIgzhAgdxhERETUDdn/lm0ymbB69Wo0NjbCz88Pr732GkJCQuQOy6P0Zw1oFwJCCI+YDq43Jo+OxMlvjG734BW1SonJiRFyh0FERETdkL0nOj8/H0lJSfjzn/+MBx98EG+99ZbcIXkclU//FbgqpeS1BTQAaFRKxA11vx7fOEMg/DSD4/HrREREnkj2nujHH38cNtvNXsDq6mqEhYXJHJHnCdJr+q23OEjvvmOGXWVyYiTOX77hNjN1KBUSJo1iLzQREZE7G9AiuqCgALt27eq0LDc3F8nJyXj00Udx7tw57Ny5847tBAf7w8dH2V9h9qvwcNfPtpCSJOHIV9egVLj+DwvRQ4L6JWZ3Eh6uR2qdCcWnrvTqcemuZBcCU5MNGJsYddv13v5/Qf2L+UO9xdyhvvDW/JGEG91VVVFRgeXLl+OTTz7pcTujsXmAInKt8HB9v8QuhMB7H5fD2NDm0nbtAlh0XwJiorwz+f+T3S6w6+NyGG+49jV0VligHx6bffdtPxD1V/7Q4MD8od5i7lBfeHr+9PQBQPYx0du3b8cHH3wAAPD394dS6Zk9zHKSJAmjooNhd/HnoagQ/0FRQAOAQiFh5uTh8JFxSjkfpQIzJw3vl78oEBERkWvJ/m49b9487N27F0uWLMHq1auRm5srd0geaXJiBCKC/F3WngRg2tghLmvPEwwN02GWTEWsQiFhxqThGBahG/BjExERkfNkv7EwLCwM77zzjtxheDyl4mYv5p5DF2Cz2fvUll0IJMeHId4Q6KLoPEfSXaGw2QQOnrwMm21gRjopFArclzIM34sLHZDjERERUd/J3hNNrjMsQoeZk4b1qSdVCIG4IQF4YHK0CyPzLMkjwvDglFj4aZT9/iAWX7UPHpwSg/EjOCsNERGRJ5G9J5pc63txYfBRKPHpF9+ita3duWnvJCApNhSzp0RD0U+PEfcUo2KCYQjzx/7PqnDpapPLpw+0C4HYyADM/n40ArXeP40gERGRt2ER7YUSY4MxPEqHA59V4WJ1I+z2nueQtguB0ABfTBtrwKjo4AGM1L0FaDVYkDECX5w34tjpa2ht6+jzFHhCCPj7qvD9pChMvDvcqx9kQ0RE5M1YRHspna8K89Li0dBsxomztahpaEN9kxlt7TYIIaDyUSBQp0ZYgB8SY4Jxd3QQC7rbkCQJE0dGIDkuDF+cq0V51Q1crTPB2Y56uwCiQm6+1hNGRkAl4ywgRERE1Hcsor1csN4XM/9vfLPdLmCyWGEXAr4qJdQqTif4Xal8FLhndBTuGR2Ff19rxrnLDahrsqC+yYwmUweEsEPCzcpaQEAhKaD3VyE00BfBAb4YOSwQsVHu93hxIiIi6h0W0YOIQiFB56eSOwyPFxOld8yfLYRAs6kddU0WtHfYIABoVEqEBmig91ezd5+IiMhLsYgm6gNJkhCg1SCANwcSERENKhyYSURERETkJBbRREREREROYhFNREREROQkFtFERERERE5iEU1ERERE5CQW0URERERETmIRTURERETkJEkIIeQOgoiIiIjIk7AnmoiIiIjISSyiiYiIiIicxCKaiIiIiMhJLKKJiIiIiJzEIpqIiIiIyEksoomIiIiInOQjdwDewm63Y/369fjmm2+gVquRk5ODmJgYx/r8/Hz85S9/gY+PD5544glkZGSgvr4ezz77LMxmMyIiIpCXlwc/Pz8Zz4Lk0JvcuXHjBmbNmoWRI0cCAO6//3489thjcp0CyehO+QMA9fX1WLhwIfbu3QuNRgOz2Yw1a9agrq4OWq0WGzduREhIiExnQHLqTf4IIZCamorY2FgAwLhx47B69WoZoie53Sl/3nvvPezbtw8AkJaWhpUrV3rX9UeQSxw4cEBkZ2cLIYQ4deqUWLFihWNdbW2teOihh4TFYhFNTU2O71966SVRWFgohBBi+/btYufOnXKETjLrTe4cPXpUbNiwQa6QyY30lD9CCFFcXCx++MMfivHjxwuz2SyEEOLdd98VW7duFUII8dFHH4mXXnppYIMmt9Gb/Ll06ZJYvnz5gMdK7qen/KmqqhI/+tGPhNVqFTabTfz0pz8VZ8+e9arrD4dzuMjJkycxbdo0ADc/lZ8+fdqxrqysDOPHj4darYZer0d0dDTKy8s77ZOamopjx47JEjvJqze5c/r0aZw5cwaPPPIInnrqKdTW1soVPsmsp/wBAIVCgZ07dyIoKOi2+6SmpqKkpGTgAia30pv8OXPmDGpqarBkyRJkZWXh4sWLAxozuY+e8icqKgo7duyAUqmEQqGA1WqFRqPxqusPi2gXaWlpgU6nc/ysVCphtVod6/R6vWOdVqtFS0tLp+VarRbNzc0DGzS5hd7kTlxcHJ566ins3r0b999/P3JycgY8bnIPPeUPANx7770IDg7usg+vPQT0Ln/Cw8Pxs5/9DH/605+wfPlyrFmzZsDiJffSU/6oVCqEhIRACIGNGzdi9OjRuOuuu7zq+sMx0S6i0+nQ2trq+Nlut8PHx+e261pbW6HX6x3LfX190draioCAgAGPm+TXm9xJTk52jJ+fMWMGtm7dOrBBk9voKX++yz689gxuvcmfMWPGQKlUAgBSUlJQU1MDIQQkSerXWMn93Cl/LBYLXnjhBWi1Wqxbt67LPp5+/WFPtItMmDABxcXFAIAvv/zSccMXACQnJ+PkyZOwWCxobm5GRUUFRo4ciQkTJuDw4cMAgOLiYkycOFGW2ElevcmdF198EQcOHAAAlJSUICkpSZbYSX495U9P+/DaQ0Dv8mfbtm3YtWsXAKC8vBwGg4EF9CDVU/4IIfDkk0/i7rvvxoYNGxwfvLzp+iMJIYTcQXiDW3eonjt3DkII5Obmori4GNHR0bjvvvuQn5+PPXv2QAiB5cuXY9asWbh+/Tqys7PR2tqK4OBgvP766/D395f7VGiA9SZ3Ll++jBdeeAEA4Ofnh5ycHERERMh8JiSHO+XPLdOnT8f+/fuh0WjQ1taG7OxsGI1GqFQqvP766wgPD5fxLEguvcmfxsZGrFmzBiaTCUqlEmvXrkV8fLyMZ0Fy6Sl/7HY7nnnmGYwbN86x/TPPPINRo0Z5zfWHRTQRERERkZM4nIOIiIiIyEksoomIiIiInMQimoiIiIjISSyiiYiIiIicxCKaiIiIiMhJLKKJiFwgPT0dDz300G3XFRUVQZIkfP75593uHxsbi5UrV/ZXeERE5GIsoomIiIiInMQimoiIiIjISSyiiYg8RGtrK1atWoXIyEj4+fkhPT0dp06dcqzT6XTIy8vrtM+ZM2cgSRL+8Y9/AABqa2vx6KOPIiQkBDqdDnPnzkVlZaVj+/Xr1yMlJQW//OUvERQUhHvvvRcAcPXqVSxbtgwGgwEqlQoGgwFPP/00LBaLY9/6+nosWbIEwcHBCA0NRXZ2NpYuXYr09HTHNlarFWvXrkV0dDR8fX2RkpKCTz/9tL9eMiKifuMjdwBERN5CCAGr1dpluc1mc0nbc+fORVlZGXJzczFkyBBs27YN6enp+OKLLxAfH4+5c+eioKAAzz//vGO/PXv2ICoqCunp6Whra0NGRgba2trw+9//Hv7+/sjLy0NqairKysoQHBwMACgtLYWvry8KCwthNptht9vxwAMPQJIkvPnmmwgMDMSBAwfw6quvIj4+HqtWrYIQAnPmzMHFixfxxhtvQK/XY+3atTh//jymTJniiCcrKwv5+fnYsGEDkpKSsHv3bsyePRtFRUX4wQ9+0OfXiYhowAgiIuqztLQ0AaDHrxMnTnS7f0xMjPj5z3/e7fqPP/5YABAHDx50LOvo6BAjR44US5cuFUIIsXfvXgFAXLhwwbHNqFGjxC9+8QshhBBvv/22UCqV4uzZs471jY2NIigoSPz2t78VQgixbt26LrFWVVWJtLQ0UVpa2imm5ORkMW/ePCGEEH//+98FAHHo0CHH+itXrgiNRiPS0tKEEEKcPXtWABB//OMfO7Uzffp0kZGR0e25ExG5Iw7nICJykalTp+LEiRNdvt5+++0+t33o0CH4+/sjLS0NVqvV0eM9c+ZMx3CIWbNmITQ0FAUFBQCAsrIylJeXY/HixY42EhISMGLECEcb/v7+mDZtWpchFYmJiY7vhw8fjqKiIowZMwbnz5/Hvn37kJubi5qaGsdwjqKiIgQGBnYaumEwGDr1LhcVFQEAMjMzHce3Wq3IzMzEkSNH0N7e3ufXiYhooHA4BxGRiwQGBiIlJaXL8paWlj63XVdXB5PJBLVa3WWdSqVy/PuTn/wEBQUF+NWvfoX8/HzExcVh8uTJjjbKy8sd2/+nhIQEx/darRZarbbT+nfeeQe//vWvUVNTgyFDhuCee+6Bn58fhBAAgOvXryM8PLxLu5GRkbh69arj+AAwdOjQ257j9evXYTAY7vhaEBG5AxbRREQeIDAwEBEREdi3b1+P2y1atAjbt2/HpUuXUFBQgEWLFnVqY+zYsdixY0eX/TQaTbdtHj58GFlZWfjNb36DlStXOorlW8U5cLPX2Wg0dtn3P5cFBgZCkiQcPXr0toV8WFhYj+dGROROOJyDiMgDTJ06FUajETqdDikpKY6v999/H7t373Zsl5qaimHDhuHVV1/FuXPnHEM5brVRWVmJ2NhYx/4TJ07E5s2b8dFHH3V77OPHj0OSJLz44ouOArq6uhpfffWVoyc6NTUVjY2NKC4uduxnNBpRUlLS6fhCCDQ3N3c6h08++QSbN2+Gjw/7dYjIc/CKRUTkJkpLS7Fly5YuyxcuXIg5c+Zg0qRJyMzMxLp16xAdHY3CwkK8+eabncZcS5KEhQsXYvPmzUhOTsbo0aMd65YtW4atW7dixowZeP755xESEoI//OEPKCwsxN69e7uNa9KkSbDb7Xj66acxf/58VFVV4eWXX4bFYoHJZAIAZGRkYNq0aVi8eDFeeeUV6PV65OTkwGw2Q6G42V8zbtw4zJs3D4888gjWr1+PxMREFBUVIScnB88995xjOyIiT8AimojITRw5cgRHjhzpsnzKlCmIiorCgQMHkJ2djeeeew5NTU1ISEjAzp078fjjj3fafvHixfjd737XaSgHAAQEBKC4uBhr1qzBihUrYLFYMGbMGHz44YfIzMzsNq7p06dj06ZN2LJlC959910MGzYMCxYsgEqlwpYtW2CxWKDRaPDXv/4Vq1atwhNPPAG1Wo0VK1bAz88POp3O0db777+PtWvXIi8vD7W1tYiJicErr7yCZ599tm8vHhHRAJPErb/FERER9VJlZSVOnDiBH//4x45hGTabDbGxsZg/fz42bdokc4RERK7FnmgiIuozIQSWLFmCgwcPYtGiRWhvb8eOHTtgNBqRlZUld3hERC7HAWhERNRncXFx+PDDD/H111/j4YcfxoIFC9DQ0ICioqJOc04TEXkLDucgIiIiInISe6KJiIiIiJzEIpqIiIiIyEksoomIiIiInMQimoiIiIjISSyiiYiIiIicxCKaiIiIiMhJ/wuIQdMfHkGnhwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 残差とレバレッジの表示\n",
    "fig, ax = plt.subplots(figsize=(12,8))\n",
    "fig = sm.graphics.influence_plot(LE2, ax=ax, criterion=\"cooks\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "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>VIF</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Intercept</th>\n",
       "      <td>68.386103</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Meeting</th>\n",
       "      <td>9.274764</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Reassign</th>\n",
       "      <td>6.866926</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Meeting:Reassign</th>\n",
       "      <td>14.562118</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                        VIF\n",
       "Intercept         68.386103\n",
       "Meeting            9.274764\n",
       "Reassign           6.866926\n",
       "Meeting:Reassign  14.562118"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# VIFを求める\n",
    "num_cols = model.exog.shape[1]\n",
    "vifs = [variance_inflation_factor(model.exog, i)\n",
    "        for i in range(0, num_cols)]\n",
    "df_vif= pd.DataFrame(vifs, index=model.exog_names, columns=[\"VIF\"])\n",
    "df_vif"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "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>Intercept</th>\n",
       "      <th>Meeting</th>\n",
       "      <th>Reassign</th>\n",
       "      <th>Meeting:Reassign</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Intercept</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Meeting</th>\n",
       "      <td>NaN</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>-0.041355</td>\n",
       "      <td>0.727493</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Reassign</th>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.041355</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.603472</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Meeting:Reassign</th>\n",
       "      <td>NaN</td>\n",
       "      <td>0.727493</td>\n",
       "      <td>0.603472</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  Intercept   Meeting  Reassign  Meeting:Reassign\n",
       "Intercept               NaN       NaN       NaN               NaN\n",
       "Meeting                 NaN  1.000000 -0.041355          0.727493\n",
       "Reassign                NaN -0.041355  1.000000          0.603472\n",
       "Meeting:Reassign        NaN  0.727493  0.603472          1.000000"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 交互作用項と各変数との相関を確認\n",
    "pd.DataFrame(model.exog, columns=model.exog_names).corr()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 中心化（Centering）処理を行って回帰分析"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 中心化(*)処理を行う。(*)元の値を平均値との差で置き換えるという手法.\n",
    "dfc = df\n",
    "dfc['Meeting'] = dfc['Meeting'] - df['Meeting'].mean()\n",
    "dfc['Reassign'] = dfc['Reassign'] - df['Reassign'].mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "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>Meeting</th>\n",
       "      <th>Reassign</th>\n",
       "      <th>Score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>1.240000e+02</td>\n",
       "      <td>1.240000e+02</td>\n",
       "      <td>124.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>1.375244e-15</td>\n",
       "      <td>-3.894734e-17</td>\n",
       "      <td>66.154839</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>8.898869e+00</td>\n",
       "      <td>1.122336e-01</td>\n",
       "      <td>11.987387</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>-2.030645e+01</td>\n",
       "      <td>-2.229032e-01</td>\n",
       "      <td>30.100000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>-6.556452e+00</td>\n",
       "      <td>-8.290323e-02</td>\n",
       "      <td>58.825000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>-3.064516e-01</td>\n",
       "      <td>7.096774e-03</td>\n",
       "      <td>66.650000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>6.693548e+00</td>\n",
       "      <td>7.709677e-02</td>\n",
       "      <td>73.325000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>2.769355e+01</td>\n",
       "      <td>2.970968e-01</td>\n",
       "      <td>99.500000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            Meeting      Reassign       Score\n",
       "count  1.240000e+02  1.240000e+02  124.000000\n",
       "mean   1.375244e-15 -3.894734e-17   66.154839\n",
       "std    8.898869e+00  1.122336e-01   11.987387\n",
       "min   -2.030645e+01 -2.229032e-01   30.100000\n",
       "25%   -6.556452e+00 -8.290323e-02   58.825000\n",
       "50%   -3.064516e-01  7.096774e-03   66.650000\n",
       "75%    6.693548e+00  7.709677e-02   73.325000\n",
       "max    2.769355e+01  2.970968e-01   99.500000"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 説明変数の平均がゼロになっていることを確認\n",
    "dfc.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "# モデル化\n",
    "model = smf.ols('Score ~ Meeting*Reassign', data=dfc)\n",
    "LE3 = model.fit()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table class=\"simpletable\">\n",
       "<caption>OLS Regression Results</caption>\n",
       "<tr>\n",
       "  <th>Dep. Variable:</th>          <td>Score</td>      <th>  R-squared:         </th> <td>   0.478</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Model:</th>                   <td>OLS</td>       <th>  Adj. R-squared:    </th> <td>   0.465</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Method:</th>             <td>Least Squares</td>  <th>  F-statistic:       </th> <td>   36.67</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Date:</th>             <td>Mon, 14 Oct 2019</td> <th>  Prob (F-statistic):</th> <td>6.80e-17</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Time:</th>                 <td>16:42:41</td>     <th>  Log-Likelihood:    </th> <td> -443.10</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>No. Observations:</th>      <td>   124</td>      <th>  AIC:               </th> <td>   894.2</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Df Residuals:</th>          <td>   120</td>      <th>  BIC:               </th> <td>   905.5</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Df Model:</th>              <td>     3</td>      <th>                     </th>     <td> </td>   \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Covariance Type:</th>      <td>nonrobust</td>    <th>                     </th>     <td> </td>   \n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "          <td></td>            <th>coef</th>     <th>std err</th>      <th>t</th>      <th>P>|t|</th>  <th>[0.025</th>    <th>0.975]</th>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Intercept</th>        <td>   66.4263</td> <td>    0.788</td> <td>   84.315</td> <td> 0.000</td> <td>   64.866</td> <td>   67.986</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Meeting</th>          <td>    0.5318</td> <td>    0.089</td> <td>    5.980</td> <td> 0.000</td> <td>    0.356</td> <td>    0.708</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Reassign</th>         <td>   18.0413</td> <td>    7.052</td> <td>    2.558</td> <td> 0.012</td> <td>    4.078</td> <td>   32.005</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Meeting:Reassign</th> <td>    6.6255</td> <td>    0.775</td> <td>    8.546</td> <td> 0.000</td> <td>    5.090</td> <td>    8.160</td>\n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "  <th>Omnibus:</th>       <td> 3.486</td> <th>  Durbin-Watson:     </th> <td>   2.362</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Prob(Omnibus):</th> <td> 0.175</td> <th>  Jarque-Bera (JB):  </th> <td>   3.183</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Skew:</th>          <td>-0.228</td> <th>  Prob(JB):          </th> <td>   0.204</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Kurtosis:</th>      <td> 3.639</td> <th>  Cond. No.          </th> <td>    79.4</td>\n",
       "</tr>\n",
       "</table><br/><br/>Warnings:<br/>[1] Standard Errors assume that the covariance matrix of the errors is correctly specified."
      ],
      "text/plain": [
       "<class 'statsmodels.iolib.summary.Summary'>\n",
       "\"\"\"\n",
       "                            OLS Regression Results                            \n",
       "==============================================================================\n",
       "Dep. Variable:                  Score   R-squared:                       0.478\n",
       "Model:                            OLS   Adj. R-squared:                  0.465\n",
       "Method:                 Least Squares   F-statistic:                     36.67\n",
       "Date:                Mon, 14 Oct 2019   Prob (F-statistic):           6.80e-17\n",
       "Time:                        16:42:41   Log-Likelihood:                -443.10\n",
       "No. Observations:                 124   AIC:                             894.2\n",
       "Df Residuals:                     120   BIC:                             905.5\n",
       "Df Model:                           3                                         \n",
       "Covariance Type:            nonrobust                                         \n",
       "====================================================================================\n",
       "                       coef    std err          t      P>|t|      [0.025      0.975]\n",
       "------------------------------------------------------------------------------------\n",
       "Intercept           66.4263      0.788     84.315      0.000      64.866      67.986\n",
       "Meeting              0.5318      0.089      5.980      0.000       0.356       0.708\n",
       "Reassign            18.0413      7.052      2.558      0.012       4.078      32.005\n",
       "Meeting:Reassign     6.6255      0.775      8.546      0.000       5.090       8.160\n",
       "==============================================================================\n",
       "Omnibus:                        3.486   Durbin-Watson:                   2.362\n",
       "Prob(Omnibus):                  0.175   Jarque-Bera (JB):                3.183\n",
       "Skew:                          -0.228   Prob(JB):                        0.204\n",
       "Kurtosis:                       3.639   Cond. No.                         79.4\n",
       "==============================================================================\n",
       "\n",
       "Warnings:\n",
       "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
       "\"\"\""
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "LE3.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "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>VIF</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Intercept</th>\n",
       "      <td>1.001628</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Meeting</th>\n",
       "      <td>1.002583</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Reassign</th>\n",
       "      <td>1.002857</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Meeting:Reassign</th>\n",
       "      <td>1.001931</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                       VIF\n",
       "Intercept         1.001628\n",
       "Meeting           1.002583\n",
       "Reassign          1.002857\n",
       "Meeting:Reassign  1.001931"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# VIFを求める\n",
    "num_cols = model.exog.shape[1]\n",
    "vifs = [variance_inflation_factor(model.exog, i)\n",
    "        for i in range(0, num_cols)]\n",
    "df_vif= pd.DataFrame(vifs, index=model.exog_names, columns=[\"VIF\"])\n",
    "df_vif"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Intercept</th>\n",
       "      <th>Meeting</th>\n",
       "      <th>Reassign</th>\n",
       "      <th>Meeting:Reassign</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Intercept</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Meeting</th>\n",
       "      <td>NaN</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>-0.041355</td>\n",
       "      <td>-0.028065</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Reassign</th>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.041355</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>-0.032562</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Meeting:Reassign</th>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.028065</td>\n",
       "      <td>-0.032562</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  Intercept   Meeting  Reassign  Meeting:Reassign\n",
       "Intercept               NaN       NaN       NaN               NaN\n",
       "Meeting                 NaN  1.000000 -0.041355         -0.028065\n",
       "Reassign                NaN -0.041355  1.000000         -0.032562\n",
       "Meeting:Reassign        NaN -0.028065 -0.032562          1.000000"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 交互作用項と各変数との相関を確認\n",
    "pd.DataFrame(model.exog, columns=model.exog_names).corr()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "LE1:AIC = 951.15\n",
      "LE2:AIC = 894.20\n",
      "LE3:AIC = 894.20\n"
     ]
    }
   ],
   "source": [
    "# 3つのモデルのAICを確認\n",
    "print('LE1:AIC = {:.2f}'.format(LE1.aic))\n",
    "print('LE2:AIC = {:.2f}'.format(LE2.aic))\n",
    "print('LE3:AIC = {:.2f}'.format(LE3.aic))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "回帰モデルが交互作用項を含む場合に、多重共線性を避けるために使われる手法が中心化（centering）。この方法で交互作用項のあるモデルLE3を作成すると、モデルの決定係数やAICはLE2の値とまったく変わらないことが分かる。しかし、VIFを算出すると値が非常に小さくなっている。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}