heart-disease-classification.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Predicting heart disease\n",
    "\n",
    "This notebooks uses python and different data science libraries to build a machine learning classification model capable of predicting if a patient has a heart disease, based on their medical attributes.\n",
    "\n",
    "## 1. Problem Definition\n",
    "\n",
    "Given medical attributes about a patient, is it possible to predict the pressence or absence of a heart disease?\n",
    "\n",
    "## 2. Data\n",
    "\n",
    "The original data can be found in the [UCI Machine Learning Repository](https://archive.ics.uci.edu/ml/datasets/Heart+Disease). For this project, the dataset was taken in csv format from [kaggle](https://www.kaggle.com/ronitf/heart-disease-uci).\n",
    "\n",
    "## 3. Evaluation\n",
    "\n",
    "Given the nature of the prediction this model is going to make. A 95% accuracy is required in this proof of concept to be considered viable for further use.\n",
    "\n",
    "## 4. Features\n",
    "\n",
    "**Medical attributes (14):**\n",
    "\n",
    "- **age:** age in years\n",
    "- **sex:**: 0 = female, 1 = male\n",
    "- **cp:** chest pain type\n",
    "    - 0: Typical angina (chest pain related decreased blood supply to heart).\n",
    "    - 1: Atypical angina (chest pain not related to heart).\n",
    "    - 2: Non anginal pain (typically esophageal spasms - non heart related).\n",
    "    - 3: Asymptomatic (chest pain not showing signs of disease).\n",
    "- **trestbps:** resting blood pressure (in mmHg on admission to the hospital)\n",
    "- **chol:** serum cholestoral in mg/dl\n",
    "- **fbs:** fasting blood sugar > 120 mg/dl (0 = false, 1 = true)\n",
    "- **restecg:** resting electrocardiographic results\n",
    "    - 0: Nothing to note\n",
    "    - 1: ST-T Wave abnormally\n",
    "    - 2: Possible or definite left ventricular hypertrophy\n",
    "- **thalach:** maximum heart rate achieved\n",
    "- **exang:** exercise induced angina (0 = no, 1 = yes)\n",
    "- **oldpeak** = ST depression induced by exercise relative to rest\n",
    "- **slope:** the slope of the peak exercise ST segment\n",
    "    - 0: Upsloping: better heart rate with excercise.\n",
    "    - 1: Flatsloping: minimal change (typical healthy heart)\n",
    "    - 2: Downslopins: signs of unhealthy heart\n",
    "- **ca:** number of major vessels (0-3) colored by flourosopy\n",
    "- **thal:** thalium stress result\n",
    "    - 1, 3: normal\n",
    "    - 6: fixed defect\n",
    "    - 7: reversable defect: no proper blood movement when excercising\n",
    "- **target:** have disease (1) or not (0). Predicted attribute."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Tools to use\n",
    "\n",
    "For data analytics and manipulation, we're going to use NumPy, pandas and Matplotlip; and for modeling, scikit-learn."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Regular EDA and plotting libraries\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "# To output charts inside the notebook\n",
    "%matplotlib inline\n",
    "\n",
    "# Preprocessing\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "\n",
    "# Models from Scikit-Learn\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.svm import LinearSVC\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "\n",
    "# Model evaluation\n",
    "from sklearn.model_selection import train_test_split, cross_val_score\n",
    "from sklearn.model_selection import RandomizedSearchCV, GridSearchCV\n",
    "from sklearn.metrics import confusion_matrix, plot_confusion_matrix, classification_report\n",
    "from sklearn.metrics import precision_score, recall_score, f1_score\n",
    "from sklearn.metrics import plot_roc_curve\n",
    "\n",
    "# # Hide unwanted warnings\n",
    "# import warnings \n",
    "# warnings.simplefilter('ignore')\n",
    "\n",
    "# Export and import models\n",
    "from joblib import dump, load"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(303, 14)"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_csv('heart_disease.csv')\n",
    "df.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Exploratory data analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "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>age</th>\n",
       "      <th>sex</th>\n",
       "      <th>cp</th>\n",
       "      <th>trestbps</th>\n",
       "      <th>chol</th>\n",
       "      <th>fbs</th>\n",
       "      <th>restecg</th>\n",
       "      <th>thalach</th>\n",
       "      <th>exang</th>\n",
       "      <th>oldpeak</th>\n",
       "      <th>slope</th>\n",
       "      <th>ca</th>\n",
       "      <th>thal</th>\n",
       "      <th>target</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>63</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>145</td>\n",
       "      <td>233</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>150</td>\n",
       "      <td>0</td>\n",
       "      <td>2.3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>37</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>130</td>\n",
       "      <td>250</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>187</td>\n",
       "      <td>0</td>\n",
       "      <td>3.5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>41</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>130</td>\n",
       "      <td>204</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>172</td>\n",
       "      <td>0</td>\n",
       "      <td>1.4</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>56</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>120</td>\n",
       "      <td>236</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>178</td>\n",
       "      <td>0</td>\n",
       "      <td>0.8</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>57</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>120</td>\n",
       "      <td>354</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>163</td>\n",
       "      <td>1</td>\n",
       "      <td>0.6</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   age  sex  cp  trestbps  chol  fbs  restecg  thalach  exang  oldpeak  slope  \\\n",
       "0   63    1   3       145   233    1        0      150      0      2.3      0   \n",
       "1   37    1   2       130   250    0        1      187      0      3.5      0   \n",
       "2   41    0   1       130   204    0        0      172      0      1.4      2   \n",
       "3   56    1   1       120   236    0        1      178      0      0.8      2   \n",
       "4   57    0   0       120   354    0        1      163      1      0.6      2   \n",
       "\n",
       "   ca  thal  target  \n",
       "0   0     1       1  \n",
       "1   0     2       1  \n",
       "2   0     2       1  \n",
       "3   0     2       1  \n",
       "4   0     2       1  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "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>age</th>\n",
       "      <th>sex</th>\n",
       "      <th>cp</th>\n",
       "      <th>trestbps</th>\n",
       "      <th>chol</th>\n",
       "      <th>fbs</th>\n",
       "      <th>restecg</th>\n",
       "      <th>thalach</th>\n",
       "      <th>exang</th>\n",
       "      <th>oldpeak</th>\n",
       "      <th>slope</th>\n",
       "      <th>ca</th>\n",
       "      <th>thal</th>\n",
       "      <th>target</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>298</th>\n",
       "      <td>57</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>140</td>\n",
       "      <td>241</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>123</td>\n",
       "      <td>1</td>\n",
       "      <td>0.2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>299</th>\n",
       "      <td>45</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>110</td>\n",
       "      <td>264</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>132</td>\n",
       "      <td>0</td>\n",
       "      <td>1.2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>300</th>\n",
       "      <td>68</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>144</td>\n",
       "      <td>193</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>141</td>\n",
       "      <td>0</td>\n",
       "      <td>3.4</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>301</th>\n",
       "      <td>57</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>130</td>\n",
       "      <td>131</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>115</td>\n",
       "      <td>1</td>\n",
       "      <td>1.2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>302</th>\n",
       "      <td>57</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>130</td>\n",
       "      <td>236</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>174</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     age  sex  cp  trestbps  chol  fbs  restecg  thalach  exang  oldpeak  \\\n",
       "298   57    0   0       140   241    0        1      123      1      0.2   \n",
       "299   45    1   3       110   264    0        1      132      0      1.2   \n",
       "300   68    1   0       144   193    1        1      141      0      3.4   \n",
       "301   57    1   0       130   131    0        1      115      1      1.2   \n",
       "302   57    0   1       130   236    0        0      174      0      0.0   \n",
       "\n",
       "     slope  ca  thal  target  \n",
       "298      1   0     3       0  \n",
       "299      1   0     3       0  \n",
       "300      1   2     3       0  \n",
       "301      1   1     3       0  \n",
       "302      1   1     2       0  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.tail()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1    165\n",
       "0    138\n",
       "Name: target, dtype: int64"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# How many had heart disease and not.\n",
    "df['target'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df['target'].value_counts().plot(kind='bar',\n",
    "                                 color=['orangered', 'skyblue']);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 303 entries, 0 to 302\n",
      "Data columns (total 14 columns):\n",
      " #   Column    Non-Null Count  Dtype  \n",
      "---  ------    --------------  -----  \n",
      " 0   age       303 non-null    int64  \n",
      " 1   sex       303 non-null    int64  \n",
      " 2   cp        303 non-null    int64  \n",
      " 3   trestbps  303 non-null    int64  \n",
      " 4   chol      303 non-null    int64  \n",
      " 5   fbs       303 non-null    int64  \n",
      " 6   restecg   303 non-null    int64  \n",
      " 7   thalach   303 non-null    int64  \n",
      " 8   exang     303 non-null    int64  \n",
      " 9   oldpeak   303 non-null    float64\n",
      " 10  slope     303 non-null    int64  \n",
      " 11  ca        303 non-null    int64  \n",
      " 12  thal      303 non-null    int64  \n",
      " 13  target    303 non-null    int64  \n",
      "dtypes: float64(1), int64(13)\n",
      "memory usage: 33.3 KB\n"
     ]
    }
   ],
   "source": [
    "df.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "age         0\n",
       "sex         0\n",
       "cp          0\n",
       "trestbps    0\n",
       "chol        0\n",
       "fbs         0\n",
       "restecg     0\n",
       "thalach     0\n",
       "exang       0\n",
       "oldpeak     0\n",
       "slope       0\n",
       "ca          0\n",
       "thal        0\n",
       "target      0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Are there missing values?\n",
    "df.isna().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>age</th>\n",
       "      <th>sex</th>\n",
       "      <th>cp</th>\n",
       "      <th>trestbps</th>\n",
       "      <th>chol</th>\n",
       "      <th>fbs</th>\n",
       "      <th>restecg</th>\n",
       "      <th>thalach</th>\n",
       "      <th>exang</th>\n",
       "      <th>oldpeak</th>\n",
       "      <th>slope</th>\n",
       "      <th>ca</th>\n",
       "      <th>thal</th>\n",
       "      <th>target</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>303.000000</td>\n",
       "      <td>303.000000</td>\n",
       "      <td>303.000000</td>\n",
       "      <td>303.000000</td>\n",
       "      <td>303.000000</td>\n",
       "      <td>303.000000</td>\n",
       "      <td>303.000000</td>\n",
       "      <td>303.000000</td>\n",
       "      <td>303.000000</td>\n",
       "      <td>303.000000</td>\n",
       "      <td>303.000000</td>\n",
       "      <td>303.000000</td>\n",
       "      <td>303.000000</td>\n",
       "      <td>303.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>54.366337</td>\n",
       "      <td>0.683168</td>\n",
       "      <td>0.966997</td>\n",
       "      <td>131.623762</td>\n",
       "      <td>246.264026</td>\n",
       "      <td>0.148515</td>\n",
       "      <td>0.528053</td>\n",
       "      <td>149.646865</td>\n",
       "      <td>0.326733</td>\n",
       "      <td>1.039604</td>\n",
       "      <td>1.399340</td>\n",
       "      <td>0.729373</td>\n",
       "      <td>2.313531</td>\n",
       "      <td>0.544554</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>9.082101</td>\n",
       "      <td>0.466011</td>\n",
       "      <td>1.032052</td>\n",
       "      <td>17.538143</td>\n",
       "      <td>51.830751</td>\n",
       "      <td>0.356198</td>\n",
       "      <td>0.525860</td>\n",
       "      <td>22.905161</td>\n",
       "      <td>0.469794</td>\n",
       "      <td>1.161075</td>\n",
       "      <td>0.616226</td>\n",
       "      <td>1.022606</td>\n",
       "      <td>0.612277</td>\n",
       "      <td>0.498835</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>29.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>94.000000</td>\n",
       "      <td>126.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>71.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>47.500000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>120.000000</td>\n",
       "      <td>211.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>133.500000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>55.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>130.000000</td>\n",
       "      <td>240.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>153.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.800000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>61.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>140.000000</td>\n",
       "      <td>274.500000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>166.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.600000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>77.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>200.000000</td>\n",
       "      <td>564.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>202.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>6.200000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              age         sex          cp    trestbps        chol         fbs  \\\n",
       "count  303.000000  303.000000  303.000000  303.000000  303.000000  303.000000   \n",
       "mean    54.366337    0.683168    0.966997  131.623762  246.264026    0.148515   \n",
       "std      9.082101    0.466011    1.032052   17.538143   51.830751    0.356198   \n",
       "min     29.000000    0.000000    0.000000   94.000000  126.000000    0.000000   \n",
       "25%     47.500000    0.000000    0.000000  120.000000  211.000000    0.000000   \n",
       "50%     55.000000    1.000000    1.000000  130.000000  240.000000    0.000000   \n",
       "75%     61.000000    1.000000    2.000000  140.000000  274.500000    0.000000   \n",
       "max     77.000000    1.000000    3.000000  200.000000  564.000000    1.000000   \n",
       "\n",
       "          restecg     thalach       exang     oldpeak       slope          ca  \\\n",
       "count  303.000000  303.000000  303.000000  303.000000  303.000000  303.000000   \n",
       "mean     0.528053  149.646865    0.326733    1.039604    1.399340    0.729373   \n",
       "std      0.525860   22.905161    0.469794    1.161075    0.616226    1.022606   \n",
       "min      0.000000   71.000000    0.000000    0.000000    0.000000    0.000000   \n",
       "25%      0.000000  133.500000    0.000000    0.000000    1.000000    0.000000   \n",
       "50%      1.000000  153.000000    0.000000    0.800000    1.000000    0.000000   \n",
       "75%      1.000000  166.000000    1.000000    1.600000    2.000000    1.000000   \n",
       "max      2.000000  202.000000    1.000000    6.200000    2.000000    4.000000   \n",
       "\n",
       "             thal      target  \n",
       "count  303.000000  303.000000  \n",
       "mean     2.313531    0.544554  \n",
       "std      0.612277    0.498835  \n",
       "min      0.000000    0.000000  \n",
       "25%      2.000000    0.000000  \n",
       "50%      2.000000    1.000000  \n",
       "75%      3.000000    1.000000  \n",
       "max      3.000000    1.000000  "
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Heart Disease Frequency according to sex"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1    207\n",
       "0     96\n",
       "Name: sex, dtype: int64"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.sex.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th>sex</th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>target</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>24</td>\n",
       "      <td>114</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>72</td>\n",
       "      <td>93</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "sex      0    1\n",
       "target         \n",
       "0       24  114\n",
       "1       72   93"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Compare sex vs target column\n",
    "pd.crosstab(df['target'], df['sex'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "72 out of 96 women have heart disease. Almost a 75% for this data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "pd.crosstab(df['target'], df['sex']).plot(kind='bar',\n",
    "                                          figsize=(10,6),\n",
    "                                          color=['orangered', 'skyblue']);\n",
    "plt.title(\"Heart Disease Frequency for Sex\")\n",
    "plt.xlabel(\"0 = No Disease, 1 = Disease\")\n",
    "plt.ylabel(\"Amount\")\n",
    "plt.legend([\"Female\", \"Male\"])\n",
    "plt.xticks(rotation=0);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Age vs Max. Heart Reate for Heart Disease"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAmQAAAGDCAYAAACFuAwbAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAgAElEQVR4nOzde5wcVZ3//9cnN0IWBhJJkFwgwEaFgZDAgFwWgwZMAnzFJbqiQUJEgyIiqCtRWFGXaPyqBLOYxfjjohIRBXRxJYBEGVgBMZAAGYEvlyUhF0nCQIYYArmc3x9VnfT0dPecmunq09X9fj4e85jp09VVp6qrpk9/zqnPMeccIiIiIhJOn9AVEBEREWl0apCJiIiIBKYGmYiIiEhgapCJiIiIBKYGmYiIiEhgapCJiIiIBKYGmYhIyszsSjPbYGZ/C12XkMzsRDN7JnQ9RGqRGmQiCZnZfWb2qpntFrou5ZjZuWbmzOyqgvIPxuU3prDNG83syoKy0fH2+lV6e/H67zOzT5Z5Prf9TfHPi2Y2K8H6zzWz/+lF/UYBXwQOdc69vafrKVinM7N/LCj7upndVIn1F9let+9hvP2tZvZ6/PP/zOwaM9svt4xz7gHn3DvTqKNI1qlBJpKAmY0GTgQc8IGglfHzPPCRgg/Sc4D/F6g+FWORJP/D9nbO7QF8CPg3MzslpaoVOgB4xTm3LukL02rEpliHW5xzewJDgH8G3g48mt8oE5Hi1CATSeYc4GHgRmB6/hNm9jYz+62ZdZjZX+Juqv/Je/5dZvZ7M2s3s2fM7F+KbcDMzjKzJQVll5jZHfHfp5rZX+MoxGoz+1KZ+v4NeBKYFL92CHA8cEfB+n9lZn8zs41mdr+ZNcflA8xsmZl9Ln7c18z+ZGZf8zhWRZnZbmb2PTNbaWYvm9m1ZrZ7/NxgM/tvM1sfRyH/28xG5r32PjObbWZ/AjYDPyNqIF8TR7+u6W77zrklQBswLm+9s8zs+fiY/tXM/jkuPwS4FjguXv9r3e1Dwb6eDPweGB6//sa4/ANm1mZmr8X7dEjea140s0vN7Ang7z1tlJU738zsNDNbGp+rL5nZ1/Oey0XDzjOzlcAfgPvjp1+L9+O4ctt2zm11zrUBHwHWE0UIMbOTzGxV3rYujc/h1+M6TozL++S9J6+Y2S/jczf3uqLna/xcyevDzE6Pz+fXzOxBMxvbk2MrkgrnnH70ox/PH+A54ALgKGArsG/ec7+IfwYBhwIvAf8TP/cP8eMZQD/gSGAD0FxkG4OA14ExeWV/Ac6K/14LnBj/PRg4skRdzwX+B/gYUeSCuO4/Aq4Ebsxb9hPAnsBuwNXAsrznDgNeBQ4BLiNqkPYtsc0bgSsLykYTRRT7xY+vJmoQDom3+Vvg2/FzbwOmxsdgT+BXwG/y1nUfsBJojo9j/7jsk2Xes8LtH0vUmPvnvGU+DAwn+pL6EeDvwH75x7FgnSX3ocj2TwJW5T1+R7z+U+L6fzk+rwbEz78ILANGAbuXWKcD/rGg7OvATT7nW1ynw+P9HQu8DHyw4Hj9NF7P7oXHsESddm6/oPybwJ8LjwXwzriOw/O2e3D898VE59lIonPyR8DNnudr0esjPgbrgHcDfYm+UL0I7Bb6/4p+9OOcU4NMP/rx/QH+iagRtk/8+GngkvjvvvFz78xb/kp2Ncg+AjxQsL4fAVeU2NZNwNfiv8cQNdAGxY9XAucDTd3U91yiBtnu8QfuXvGH3AkUNMgKXrd3/OG7V17ZF+P9fZW8hmKR194IbAFey/vpyH2YA0bUGDk47zXHAf9bYn3jgFfzHt8HfLNgmfvwa5C9BrwR//09wMq8ZhlwRv5xzHsu6T6cROcG2b8Bv8x73AdYDZwUP34R+EQ3762Lj2v+cd7CrgZZ0vPtamBuwfE6qMgx7EmD7NPAs4XHAvhHogbSyUD/gtc8BUzMe7wf0fXVZfuF5yslrg/gP4F/Lyh7BphQ7ljrRz/V+lGXpYi/6cA9zrkN8eOfs6vbcihRg+OlvOXz/z4AeHfcVfJa3PU1jWiMTTE/Bz4a//0xoijR5vjxVOBUYIWZtXp0H70B/A64nKgx+af85+NuyDlx91AHUYMAYJ+8xX5C9KF8p3Pu2XLbA77nnNs790MUgckZShT9ejTvONwVl2Nmg8zsR2a2Iq7L/cDeZtY3bx35xzWJfYA9gC8RNQz6554ws3PyurJeI4oK7lN0Ld3sg4fhwIrcA+fcDqJ9GpG3jM8+HllwnOfkPVf2fDOzd5vZH+Ou4Y1EjabC/e3pcS40AmgvLHTOPUcUCfs6sM7MfmFmw/Pq/+u8uj8FbAf29ThfS10fBwBfLDgmo4jeD5Hg1CAT8RCPD/oXYEI8duVvwCXAEWZ2BNE4mW1EXSw5o/L+fglozf8Adc7t4Zz7TIlN3gPsY2bjiBpmP8894Zz7i3PuDGAY8Bvglx678FOiKNfPijz3MeAMokjFXkQNL4giQTnzgf8GJpnZP3lsr5QNRFGq5rzjsJeLBtsT1/GdwLudc03Ae4rUxRWss/BxSc657c657xNFky4AMLMDgB8DFwJvixs3y/O2Wbj+7vahO2uIGgfE2zeic2V1T/aphO7Ot58TdbmOcs7tRTROzgrW4Ur87c2imy7+D/BAseedcz93zv0T0fFwwHfy6j+loP4DnXOr6eZ8LXN9vATMLljnIOfczT3ZN5FKU4NMxM8Hib6hH0rUjTaOaEzVA8A5zrntwO3A1+Moz7uIbgDI+W/gHWb2cTPrH/8cnT+YO59zbhtwK/BdonFKv4edg+ynmdlezrmtRN1W2z3q30o0Zuk/ijy3J/Am8ApR5Odb+U+a2ceJxsydC1wE/MTMfBsfncTRoB8Dc81sWLz+EWY2Ka8ubxANHh8CXOGx2peBgxJWZQ7wZTMbSDROyhE1qjGzGUQRsvz1jzSzAZ770J1fAqeZ2UQz60/UCH0TeDDhPpTT3fm2J9DunNtiZscQNXLKWQ/swPM4x9s7BLiZKCp3VZFl3mlm77MofcwWovc9dy5fC8yOG8uY2VAzOyOv7kXP126ujx8Dn46jg2Zm/2DRzQ17+uyTSNrUIBPxMx24wTm30jn3t9wPcA0wzaI74S4k+sb+N6JI1M1EHxw4514H3g+cRRQh+RtRNKBcLrOfE0UBfhU30HI+DrwYd9d8Gji7u8q7yGLnXJeuI6Lo2QqiCM1ficaZAWBm+xONLzrHObfJOfdzYAkwt7ttlnEp0SD2h+N9uJcoKka8rd2JolAPE3UFducHwIcsuitznmcdfkc0Hu5Tzrm/At8HHiJqfB0O5Hfr/oHorsy/mVmuu7rcPpTlnHuG6D37D6L9/D/A/3HOveVZd59tdHe+XQB808xeB75GN1HWuLt8NvCnuLvv2BKLfsTMNhGNabuDqNF0lHNuTZFldyNqGG+I6zcM+Gr83A/i198T1/FhosH4UOZ8jRW9Plx0d+2niK7ZV4nev3PL7bdINZlzvY2Mi0gxZvYd4O3OuendLiwiIg1NETKRCrEo79PYuDvkGOA84Neh6yUiIrUveBZokTqyJ1E35XCi2/m/D/xX0BqJiEgmqMtSREREJDB1WYqIiIgEpgaZiIiISGCZHkO2zz77uNGjR4euhoiIiEi3Hn300Q3OuaKzemS6QTZ69GiWLFkSuhoiIiIi3TKzFaWeU5eliIiISGBqkImIiIgEpgaZiIiISGCZHkMmIiIifrZu3cqqVavYsmVL6KrUvYEDBzJy5Ej69+/v/Ro1yERERBrAqlWr2HPPPRk9ejRmFro6dcs5xyuvvMKqVas48MADvV+nLksREZEGsGXLFt72trepMZYyM+Ntb3tb4kikGmQiIiINQo2x6ujJcVaDTERERKqib9++jBs3jubmZo444giuuuoqduzYAcCSJUu46KKLAtcwHI0hExERkarYfffdWbZsGQDr1q3jYx/7GBs3buQb3/gGLS0ttLS0BK5hOIqQlbN4IZw9Gib1iX4vXhi6RiIiItWR8mfgsGHDWLBgAddccw3OOe677z5OP/10AFpbWxk3bhzjxo1j/PjxvP766wB897vf5eijj2bs2LFcccUVO9f1wQ9+kKOOOorm5mYWLFgAwPbt2zn33HM57LDDOPzww5k7dy4Azz//PJMnT+aoo47ixBNP5Omnn67ofvWUImSlLF4IV8+ENzdHj9etiB4DTJwWrl4iIiJpq9Jn4EEHHcSOHTtYt25dp/Lvfe97/PCHP+SEE05g06ZNDBw4kHvuuYdnn32WRx55BOccH/jAB7j//vt5z3vew/XXX8+QIUN44403OProo5k6dSovvvgiq1evZvny5QC89tprAMycOZNrr72WMWPG8Oc//5kLLriAP/zhDxXbp55ShKyUGy7bdSLmvLk5KhcREalnVfwMdM51KTvhhBP4whe+wLx583jttdfo168f99xzD/fccw/jx4/nyCOP5Omnn+bZZ58FYN68eRxxxBEce+yxvPTSSzz77LMcdNBBvPDCC3zuc5/jrrvuoqmpiU2bNvHggw/y4Q9/mHHjxnH++eezdu3aiu9TTyhCVsr6lcnKRURE6kWVPgNfeOEF+vbty7Bhw3jqqad2ls+aNYvTTjuNO++8k2OPPZZ7770X5xxf+cpXOP/88zut47777uPee+/loYceYtCgQZx00kls2bKFwYMH8/jjj3P33Xfzwx/+kF/+8pdcffXV7L333jvHsdUSRchKGbp/snIREZF6UYXPwPXr1/PpT3+aCy+8sEuaiOeff57DDz+cSy+9lJaWFp5++mkmTZrE9ddfz6ZNmwBYvXo169atY+PGjQwePJhBgwbx9NNP8/DDDwOwYcMGduzYwdSpU/n3f/93HnvsMZqamjjwwAP51a9+BUTRuccff7xi+9QbqUXIzGwU8FPg7cAOYIFz7gdmNgS4BRgNvAj8i3PuVYvejR8ApwKbgXOdc4+lVb9uzZjduf8cYLdBUbmIiEg9S+kz8I033mDcuHFs3bqVfv368fGPf5wvfOELXZa7+uqr+eMf/0jfvn059NBDmTJlCrvtthtPPfUUxx13HAB77LEHN910E5MnT+baa69l7NixvPOd7+TYY48FogbbjBkzdqbV+Pa3vw3AwoUL+cxnPsOVV17J1q1bOeusszjiiCN6tV+VYMX6biuyYrP9gP2cc4+Z2Z7Ao8AHgXOBdufcHDObBQx2zl1qZqcCnyNqkL0b+IFz7t3lttHS0uKWLFmSSv2BaFDjDZdFIdqh+0cnogb0i4hIBj311FMccsgh/i/QZ2CvFDveZvaoc65obo/UImTOubXA2vjv183sKWAEcAZwUrzYT4D7gEvj8p+6qIX4sJntbWb7xesJY+I0nXwiItKY9BlYVVUZQ2Zmo4HxwJ+BfXONrPj3sHixEcBLeS9bFZcVrmummS0xsyXr169Ps9oiIiIiVZF6g8zM9gBuAy52znWUW7RIWZf+VOfcAudci3OuZejQoZWqpoiIiEgwqTbIzKw/UWNsoXPu9rj45Xh8WW6cWS4b3CpgVN7LRwJr0qyfiIiISC1IrUEW3zV5HfCUc+6qvKfuAKbHf08H/iuv/ByLHAtsDDp+TERERKRK0kwMewLwceBJM8tlYPsqMAf4pZmdB6wEPhw/dyfRHZbPEaW9mJFi3URERERqRmoRMufc/zjnzDk31jk3Lv650zn3inNuonNuTPy7PV7eOec+65w72Dl3uHMuxXwWIiIiUm1mxhe/+MWdj7/3ve/x9a9/3fv1N954I0OHDmX8+PGMGTOGSZMm8eCDD+58/mtf+xr33ntvJatcNcrULyIiIlWx2267cfvtt7Nhw4Yer+MjH/kIS5cu5dlnn2XWrFmceeaZO6dd+uY3v8nJJ59cqepWlRpkIiIi0kVb+xbmL29nztINzF/eTlv7ll6vs1+/fsycOZO5c+d2eW7FihVMnDiRsWPHMnHiRFau7H7ezPe+973MnDmTBQsWAHDuuedy6623AtF8mIceeihjx47lS1/6EhBN1zR16lSOPvpojj76aP70pz8B8Mgjj3D88cczfvx4jj/+eJ555pnoGLS1ccwxxzBu3DjGjh27czLzm266aWf5+eefz/bt23t9bNQgExFJII0PKZFa09a+hUUrN9GxNZp2qGPrDhat3FSR8/2zn/0sCxcuZOPGjZ3KL7zwQs455xyeeOIJpk2bxkUXXeS1viOPPJKnn366U1l7ezu//vWvaWtr44knnuDyyy8H4POf/zyXXHIJf/nLX7jtttv45Cc/CcC73vUu7r//fpYuXco3v/lNvvrVrwJw7bXX8vnPf55ly5axZMkSRo4cyVNPPcUtt9zCn/70J5YtW0bfvn1ZuHBhbw9LqoP6RUTqSu5DalucITH3IQXQPGRgwJqJVFbrms07z/OcbS4q7+253tTUxDnnnMO8efPYfffdd5Y/9NBD3H57lCHr4x//OF/+8pe91ldsCsimpiYGDhzIJz/5SU477TROP/10AO69917++te/7lyuo6OD119/nY0bNzJ9+nSeffZZzIytW7cCcNxxxzF79mxWrVrFmWeeyZgxY1i8eDGPPvooRx99NBDNzzls2LAudUhKETKRnlq8EM4eDZP6RL8X9/4bktS2ch9SIvUkFxnzLU/q4osv5rrrruPvf/97yWWi7FndW7p0aZc5I/v168cjjzzC1KlT+c1vfsPkyZMB2LFjBw899BDLli1j2bJlrF69mj333JN/+7d/473vfS/Lly/nt7/9LVu2RJHAj33sY9xxxx3svvvuTJo0iT/84Q8455g+ffrOdTzzzDOJbkwoRQ0ykZ5YvBCungnrVoBz0e+rZ6pRVufS/pASqRVN/Ys3D0qVJzVkyBD+5V/+heuuu25n2fHHH88vfvELABYuXMg//dM/dbue1tZWFixYwKc+9alO5Zs2bWLjxo2ceuqpXH311SxbFmXfev/7388111yzc7lc+caNGxkxIpqt8cYbb9z5/AsvvMBBBx3ERRddxAc+8AGeeOIJJk6cyK233sq6dVFe+/b2dlasWNGDo9CZGmQiPXHDZfBmQVTkzc1RudSttD+kRGrFhOGD6FcQoOpnUXmlfPGLX+x0t+W8efO44YYbGDt2LD/72c/4wQ9+UPR1t9xyC+PGjeMd73gH3/rWt7jtttu6RMhef/11Tj/9dMaOHcuECRN23kQwb948lixZwtixYzn00EO59tprAfjyl7/MV77yFU444YROA/RvueUWDjvsMMaNG8fTTz/NOeecw6GHHsqVV17J+9//fsaOHcspp5zC2rW9z2Nvxfpes6KlpcUtWaJ0ZRLApD5RZKyQGdytaEm9KhxDBtGH1JT999AYMql5Tz31VJeGSzlt7VtoXbOZjq07aOrfhwnDB+k8T6DY8TazR51zLcWW16B+kZ4Yun/UTVmsXOpW7sNIH1LSCJqHDNS5XUVqkDWSxQujLrX1K6OGw4zZMHFa6Fpl04zZ0Zix/G7L3QZF5VLX9CElImlQg6xR5Aah5xoQuUHooEZZT+SOmRq4IiJSAWqQNYpyg9DViOiZidN07EQkU5xz3ukkpOd6Mj5ftwY1ivUlpqAoVS4iInVl4MCBvPLKKz1qLIg/5xyvvPIKAwcmG9qgCFmj0CB0aUQaNymy08iRI1m1ahXr168PXZW6N3DgQEaOHJnoNWqQNQoNQpdGo3GTIp3079+fAw88MHQ1pAR1WTaKidPg4gUw7IAoV9awA6LH+mCSeqXkvSKSIYqQNRINQpdGonGTIpIhipCJSH0qNT5S4yZFpAapQSZSSxYvhLNHR1MznT1ak5X3xozZ0TjJfBo3KSI1Sl2WIrVCg9ArS8l7RSRDNLm4VI9SEJR39ujiqUmGHQA3vVjt2oiISIVpcnEJT9Gf7mkQuohIw9IYMqkOpSDoXpYGoYce6xZ6+yIiFaYGmVSHoj/dy8og9Fy0c90KcG5XtLNajaLQ2xcRSYEaZFIdWYr+hJKV5L2ho52hty8ikgI1yKQ6shL9CW3itGgA/907ot/lGmOhuu1CRztDb19EJAVqkEl1ZCX6kxUhu+1CRztDb19EJAVqkEn1JIn+SHkhu+1CRzuTbF+D/0UkI5T2QiSLQnbbhU646rt9pVoRkQxRYliRLFIS2e7pGIlIjSmXGFZdliJZFLrbMAs0+F9EMkQNMiku4NibtvYtzF/ezpylG5i/vJ229i2lF06jnlkYd6SbJLqnwf/+snDOi9Q5jSGTrgKOvWlr38KilZvYFvekd2zdwaKVmwBoHjIw/XpmadzRxGm1V6daMmN25/cSFEUsJkvnvEgdU4RMugp4B1/rms07G2M521xU3kUa9VTS0fqhKKIfnfMiNUERMukq4Nibjq07/MsT1LOtfQutazbTsXUHTf37MGH4oK4Rt4TrTGTxwnB3JTYyRRG7p7F2IjUhtQiZmV1vZuvMbHle2Tgze9jMlpnZEjM7Ji43M5tnZs+Z2RNmdmRa9RIPAcfeNPUvfkoWLfesZ64bNNeoy3WDFh2blsa+a+5FqWUaaydSE9LssrwRmFxQ9n+BbzjnxgFfix8DTAHGxD8zgf9MsV7SnYB38E0YPoh+1rmsn0XlXXjWM1E3aBr7nqRLKCuDq7NST+me7tgVqQmpdVk65+43s9GFxUBT/PdewJr47zOAn7ooKdrDZra3me3nnFubVv2kjICJP3PdiF7di571TNQNmsa++3YJZWVwdVbqKX5CJ/oVEaD6Y8guBu42s+8RReeOj8tHAC/lLbcqLlODLJSAY2+ahwws3gDroab+fYo2vkp1j1bc0P2LJygt7BIqF0mr1nvhM9atFuoplZWVsXYaiyl1rNp3WX4GuMQ5Nwq4BLguLrciyxadQsDMZsbjz5asX78+pWpKJniOzUrUDZrGeC/fLqFijbZy5ZXmu+8aBC4haCym1LlqN8imA7fHf/8KOCb+exUwKm+5kezqzuzEObfAOdfinGsZOnRoahWVDPAcm9U8ZCBTXn+cpo1rwe2gaeNaprz+ePEoXJLxXvMugMn94P0W/Z53QfF6+qZf6NO3+OtLlVea775naRB4Vsa6ZaWeISk9h9S5andZrgEmAPcB7wOejcvvAC40s18A7wY2avyYdCvB2Kzmq2fSXJgg9K0ijSLfdc67AP47796THdt3Pb5oftfX+3QJ7dierLzSfPc9KwlXszLWLSv1DE2RWalzaaa9uBl4CHinma0ys/OATwHfN7PHgW8R3VEJcCfwAvAc8GOgRKhBJI9vpCbJN2vfdd65oPhypcp9DDsgWXml+e57VhKuZiWikpV6hpalyKxID6R5l+VHSzx1VJFlHfDZtOoidco3UpPkm7XvOtOIZs2YTdvi39F68r/SsfcIml5bzYR7v0vzxNN6vs6E2/eOfGVhEHhWIipZqWdoWYnMivSQMvVLdvneru97l2OSdfbpW7zx1YvxXm3jp7Jo8Mlss2gdHYNHsehDc2H0XjT3eK0J1Fv6gyTve0hZqWdo9XZ+ihSwKDiVTS0tLW7JkiWhqyG1rnCMDkTfrHvTzVY4hizn9M8UH0PmYf7y9pLpOS44bEiP1tnQ0njf05CVeopIr5nZo865lmLPaXJxqX9pjHm6aH7U+MpFxPr07VVjDBImsJXuZWWsW1bqKSKpUoRMpEYkjpCFTJKpBJ0iIokpQiaSARNeXUq/tzrfbdfvrc1MeHVp14VDJslUgk4RkYpThKyRJIlqKAJSfWePpm34UbROvnzXXZZ3XUnzmkfhphe7LFt0IPiwA7oum0I9g21baHtgMa3bh9HRtC9NHS8zoe86mk+cGLpamdXWvsVv7lyRCigXIdNdlo0iSfJJJaoMY/1KmtetoHnZ7Z3LrcjMYiFTJShNQzBtDyxm0YAxbBsQTcXVsdd+LHprL3hgsRplPdDWvoVFKzexLY5LdGzdwaKVmwDUKJOqU5dlo0iSfFKJKsNIkvgyZJJMJegMpnX7sJ2NsZxtAwbRun1YoBplW+uazTsbYznbXFQuUm1qkDWKJFGN0BEQ3zkiQ6v0/IMzZkO/AZ3L+g0onvjSd8LyNITcdqytfQvzl7czZ+kG5i9vp619S+9XmoH5JDua9vUuT+UY1Rnd2Sy1RA2yRpGV6Esuv1cu6Wpujshaa5SlNbC9cExnqTGeIVMlBE7TkOtmyn1o5rqZetXgyMiNCk0bi0/xW1ieyjGqQ039i38ElioXSZPOukaRJKoRMgKSxhyRSflE6NLo1r3hMti+tXPZ9q2l1zlxWjSI/u4d0e9yN2hUOvLju+0UJOpm8t33jHTTT/jjVcXvxP3jVZ3K6rYrrsLn8oThg+hXMESzn0XlvZKBaKvUHg3qbxRJph0JOUVJGnNEJlGYgT8XoYPOSV/T6NZNY511eIOGdzdTkn0P3U3vqfnPP4Mtm7reifv4r4Gf7Fyu463tRW8G6XirStdRGlI4l3MD9yt6l2UdXnNSHUp7IbVlcr/Sc0Teta12tp809YNPGpE00knUQoqKCqdQ8U6gm2Tfa+E4+fCs5/z7nqRjr/26LNa0cS0XnHR4z7cfMh1Onb1H0piUGFay49SZycorzTdCl6Rb13d8UhpdxaEjPymMzfLuZkqy7zVwo4IXz3oe3HZn0fGIB7fd2fNthx5nF/pc9pWVekrNUYNMqsdnXEUKc0Qmkttud+VJBrb7jk+aOA1Omd5530+Z3ruxYaFTVKQwNqt5yECm7L/HzoHXTf37MGX/Pbp2MyXZ9yTvZ8jxQZ71fL751K5dlmZReU/dcBlth0xm/qzHmDPnZebPeoy2QyZXb5xd6HPZV1bqKTVHXZZSHYXjKiD6Zl9rkygXjiHL6U2jcFKf4ndLmkWD4nOSHCPfZUMfd999T0Ma+x76eHqa89j64gmFnWPWkUN7tM62Sz/EojOv6pQHrd9bm5ly+xdo/s6tPa2qv4wc+8zUU4JQl6WEl5G72FKJ0Pl+Y04jeW/gFBVBowVp7HtGzuOmAcUjvaXKfbROuaJ4UtopV/R4nYmEPpd9ZaWeUnN0l6VUR5bGVVw0v7JdpDNmF//GXDg+Ka3kvROnhfsw8N33tLT9CTasiqJ0G1ZFj3tzLNI6jys8WH7C8EGdpnhs++QAACAASURBVASC3qdz6Gh6e6LyNOaIbBs/ldZ9p3ReZ6/WmJKQ15xkliJkUh2NPK7C9xvzHkOKv75YeVaOZ8hoQRpJhtM47ikMlvceZ5dAkqhbGolplexW6p3GkEl1ZGlcxbwLokS0O7ZHXZanzqzOTQUf2gc6Xula3vQ2uHVD57LAxzON6EfFTe5H29gzuubseuK/ep5CJY3jnkYKlRQUTsQNUdStWEPPOzVJAmmsU6Tayo0hU5elVEfIZLNJ+CaGTcPr7f7lAY9n4QdzLlIB1FSjrG3sGSyaOnfnuKeOwaNYNHUuQM+7udI47km6QQMmHU2SRDWNOSI176TUOzXIpHqyMK6i3NRNaTfIhu5fPFJSqjss0PEsNy1PLTXIWidfXnwQ+uTLa2vcUZL3vdxNBVU4F5qHDPR6j5v69ykZzeqpNNYpUkt0JovkCzl1U0aSk2YlUtGx98hE5V7SSI6a5H3PyM0xacwRmdq8kyI1Qg0ykXy+iWHTkJHb5UtFJKoaqfBIzppG6ofEaS98ksgmed8zcjNHGjcVpLFOkVqiLkuRfKfOLJ4YtlpTN2WgW/fgpv4sfeXNouVV4TmOKo3UD6mN9/J930OnEUnAt3sz9DpFaoUiZCL5Qk/dlAHPd2xNVF5xnlGqVCIqSSJUaSSRzUgUVUSSU4RMpFClE8NCsFQFaQg+hixBlKriEZUZs+H7n4Btb+0q6zeguuO9PKNpbQ8spnX7MDqa9qWp42Um9F1H84kTe7ftBpaJVC+SaYqQiaQtjYHgAQUfQxZ6HFVh7sZSuRwD1rPtgcUsGjCGjr32A+tDx177sWjAGNoeWJz6tuuRktJKNahBJpK2jMx/6Cv43W4h70a94TLYXtA1u31r8fcypXq2tW9h/vJ25izdwPzl7UUbBa3bhxVP+bF9WPGV+tx80MDKpXoRqRR1WYqkLSOpCnwlSRCaipBJhpPOIQoVradvUt6Opn2Lvr5oecBks1kRvJteGoIaZCKFKj3eK2nC1wxINDbL83gmGvMU6m7UtJL3eh4j36S8TR0vR92VBZo6XgYKomSBk81mYXxl0qS0Gm8mPaEuS5F8oRN/1hvP45mZMU/HnJqs3EeCc67jreIJigvLJ/RdR7+3Ojey+r21mQl913V9ccgIbkbGVybpptd4M+kpNchE8iUZ7+U77mbiNDhleudUGqdMr7koQCo8j2fiMU++Kj026pE7k5X7SHDONXX8regqCsubT5zIlBd+R9OrL4HbQdOrLzHlhd8VjzgO3Z+7z/g23/nWWubMWcd3vrWWu8/4dnUiuBkZX5kkhUqmxps18NhBn7GY1aYuS5F8vtGCJONuFi+E3/9k1/RLO7ZHj5tPqP9GWbHuvSLlicY8+UpjbFQa0STPYwQwYdE3WHTmVZ0ar/3e2syERd+Ak27dteDihTT//Es0FyaQHdbUZd/v/uxtLN1t/yivGeD69mPpcefBkacwqed75SdD4yt9u+kzM96sgccO+o7FrDZFyETy+aYqSPLNPiNRgFR4TkXV9NrqoouVKveSxnFPI5VFgum6mlcvYcptl3SOfN12Cc2rl3ReMMG+Lxt4wM7G2E5mUXnaQqcwSUHwtDC+Gvj/Uq1GMWvsDBEJzHe8V5Jv9hmKAlSc52TtE+66sviYp7uu7Pm216+kbdyZzJ/1GHPmvMz8WY/RNu7M3h33pOMBfbqEkkxoP2M2zU/dxQVzjmTWrH25YM6RND91V6/OzxJZ1EqW+/LqEqrD8ZXB08L4auD/S7UaxUytQWZm15vZOjNbXlD+OTN7xszazOz/5pV/xcyei59LPVIuUpTv1DRJvtnvMaT4sqXK68mwElGWgvLmNY8Wj/ysebTHm2478VMsmjqXjsGjohsFBo9i0dS5tJ34qR6vM9HURb4D1j2PUaLtJzg/zRX/ECpV7sN7YHsdTgWVmUnQ6zA66atWo5hpjiG7EbgG+GmuwMzeC5wBjHXOvWlmw+LyQ4GzgGZgOHCvmb3DOVfiq6NIinxSFSSZ5Nm6FpUtr7SQaQV8j9OM2TRfPZPmZbd3Xu7iBT3edOvky9nWd/dOZdsGDKJ18uU0F3tBpY+TbzqJpBOGe56fbYt/R+vJ/0rH3iNoem01E+79Ls0TT+uy6Lhlt7J03Ic7d1s6x7hlt8KRF3Ra1jc1iW96DoC28VNp3XdK5xQR5feu5jUvvY3mGk/lkaWJ6ittwvBBncaQQW1EMVNrkDnn7jez0QXFnwHmOOfejJfJ3YN9BvCLuPx/zew54BjgobTqJ9IrSZJ+vt5efB2lyisp9MBd3+OUQhLVjoLGWNly3+OU5Hj6dgmlkUB2/FQWDT6ZbRaNQ+sYPIpFH5oLo/fq0tiZdMuF8EYHy959Lq5PX2zHdsb9+UYm3fFV+MSuBlkuNUnuhoKOvfZj0Vt7wQOLuzTKfLuEanVwda+EvuZ8hUywHFjw5NYlVPsuy3cAJ5rZbGAL8CXn3F+AEcDDecutistEapdv0s+QiWFDJ/2EYElcjeLjoIoGJn2PU5LjmeR9T3KMPCJ5rWs272yM5WyzvkUjVAzdn5Er/sLzh0yiY+8R7NmxlpEr/tKlniVTk7wxrEsjzzeRapJIWmbUwjXnK1SC5RqQKLl1lVS7w7QfMBg4FvhX4JdmZhT/H1l0TKmZzTSzJWa2ZP369enVVKRSQg5czsrA3RQShCYarO57nBKkqEjlffc8TkkGLbedf13xsXbnX9f5tQlSk/gObK/VwdW9kpVrTmpOtSNkq4DbnXMOeMTMdgD7xOWj8pYbCawptgLn3AJgAUBLS0tvbwQS6aLi056E7BrIyrRNKUQVEk1343uc+vQtfvdjsdQVabzvnscpyb63Dh7PtoJltw0YROs/jO8U+UoyHZNvl1DSKYl8JbmGK369Z+Wak5pT7QjZb4D3AZjZO4ABwAbgDuAsM9vNzA4ExgCPVLluIulNezJxGtz0Ity9I/pdzUH1WUgrkCTy5GnCq0uLp9J4dWnXhX2PU5IUFVD5990z+pIk9YJvlCrRdExEjbILDhvCrPH7cMFhQ4o2ctJIEZHkGk7les/KNSc1J820FzcTDcp/p5mtMrPzgOuBg+JUGL8AprtIG/BL4K/AXcBndYelhFCrCQN7LCtpBRIkR/XV/KPziqfS+NF5XRf2PU4p1DMRz1QFSVIv+KYAaD5xIlPeepamjWuj47lxLVPeerb0BPAe0kgRkeQaTuV6z8o1JzUnzbssP1riqbNLLD8b0FcICaoux7RkYeBu0siTT4qK9StpXreicyoN6JqVPsfnOCWtZ6UlSFXgO2h5wvBBLHpxY6ebAPq57UwYvkfXhZtPgDWbYesO2GcEDB/To93otMoKp4hIcg2ndr1n4ZqTmqNM/SJ5ajVhYN1LkhzV9waANBJfJqlnGlKIvjQvvY0ptxZEEm+9hOalt3VaLpXuvRRu5khyDTdtf6P4siXKRdJk0fj6bGppaXFLlizpfkFJrOIDXTOiMC8SRGNaajLTtq8kCU9DJZEtzN0EuxLDFm7/7NHFx5YNOyAap5W3zpLJUXu6T4sXwvc/Adve2lXWbwB88fqai4h4X8Oex3P+8vaSA/AvOKyHs074vpcJJLmG2648n0WnXNZlsvYpv59N8+U/6tH2Rcoxs0edcy3Fnqv2XZaSAXWZrNFTrSYM7LEkSSpDJrRMckei58D2JMlREyn8EluDX2oTXcOexzOV7r0UUkQkuYabH/gxdGygdfLluxrtd11J8+O/BtQgk+pSg0y6qMtkjQnUYsLAHkuSTiJ0QssKJ9pNlBwVz4jSDZfB9q2dy7ZvLX2MAkUcE13DnsczlRQVKaWI8B6XNnR/mpfd3nWcYbW6oEXyaGCMdFGXA9sbVZIIRFYSWnqmFUiUHNV3fFSSY5TC+Chfia5hz+N5cFP/oussVe4lYPLc1LYv0kOKkEkXaSVrlACSRCCyktBy4jRo+xPcuSC6u7FPXzhlepcISKLkqL4RpSTHKGDEMdE17Nld/HzH1q6vLVPuZeI02ga83WvCckgQxfQ97g08n6PUHn3CShdpJGuUQJJEALISLVi8EH7/k12pJnZsjx4XREDSSI6a6BgFjDgmSooLXgls04ict7VvYdGeR0QzAFifaMLyPY/oXRLXpEmGQyVtFimgBpl0kUayxkxZvDC6+2tSn+h3FbqYctratzB/eTtzlm5g/vL2yswQ4JsmISsJLctFQPKkkRw10TFKknajwudcoqS4ntJICZNKEtfQyXtFekhpL0TyJUm/UGF1mXIjDZP6FL+z0SyKcvRAKsfe91xK45zLyDGas3RDyedmjd+nZ8u+v0TiX4B7svt5J/WhXNoLRchE8nlGX9JQd9M2pSWFhK+pRIV9o2lpnHMZOUaJkrj6Lhs6ea9ID3k1yMzsADM7Of57dzPbM91qiXiqdPdiwHE/icfo+O57gmNU8S7TNKQ01q156W1cMOdIZs0axgVzjuySqX6nJOecz/ikNM65GbNpa/ko82c9xpw5LzN/1mO0tXy098fIY8LwJJKM8/NeNq2xkAGHMkhj6PYuSzP7FDATGAIcDIwErgV6PqOsSCWkkcg04J2Gie6M8933BMcoMwmB07gzLoXj6S2Fcy61pLgVliiJq++yIc8PkV7odgyZmS0DjgH+7JwbH5c96Zw7vAr1K0tjyBpcCtOuZGYMme++JzhGqUyNkxUpHE9vKZxzDf1eJuWTvDeN99132z1ZVmpWb6dOetM595aZ5VbWD9DISAkvja6egHmJEk3b5LvvCY5RQycETuF4ekvhnGvo9zIJ38hX0lQaldx20mUls3waZK1m9lVgdzM7BbgA+G261RLxkFb3ou8UPil8Y/Wetsl33xMcoyRdpnU3+XwKxzOJtvFTad13SufjWWpZj2Nft8mdK33N+SaR7dN3V967fCVSafgmsG07ZHLXeTRrcVozqQqfq3MWsB54EjgfuNM5l/4tZyLdCZnINOC0OID/vic4Rr6Dpr0TdGZJCsfTV5Lj6btsXSZ3TuOa8414FmuMlSj3fY/aRrSwaOpcOgaPipLiDh7FoqlzaRtRpDcrK9OaSa/4NMg+55z7sXPuw865Dznnfmxmn0+9ZiLdCZnINK30GL53cvnue4Jj5JvWoCbSc1T6jrcUjqevNJKj1mVy55DpQRKk0vB9j1qnXMG2AZ0byNsGDKJ1yhU9r6dkmk+X5XTgBwVl5xYpE6k+3+7FSkvjG2vScSK++57gGPl0mQYfn5TWeJoUjqePJMczybLe3d9ZkVJ6kKI3VBSLjPosh/971NH09uLLFStPsH3JrpIRMjP7qJn9FjjQzO7I+/kj8Er1qihSg9L4xhowKW0SaUyhk0hGjpOvVJKj1qM0rrkUIqO+71HTgOLjz4qWZ2VaM+mVclfxg8D3gafj37mfLwKT06+aSA1LY/xa0ghAoESVwccn1dl4msTJUV3ncUv93PZsjw3zFXLMKHhPQu77fia+jjQJet0r2WXpnFsBrACOq151RDIijfQYSe7gC3gbfKL0HGkImLw3DYmSoy69DRb/jtaT/3XXnXn3fpfmiafV/wd0RhK++r6fwa8jqTk+iWGPBf4DOAQYAPQF/u6ca0q/euUpMazUlSQJQs8eTdvwo7reMr/m0S6JKusuRUWS45RGMs2QCTrTSlCaERU/lxv8eEr19TYx7DXAWcCvgBbgHOAfK1c9EQESRQDaRrSw6Myrdt6llbtlntu/0Cl/VWamQ0rC9zilEUUMnaCzzrprk0jlXG7g4ym1xydCtsQ512JmTzjnxsZlDzrnjq9KDctQhEwa1fz7nqRjr/26lDdtXMsFJ+2a1ayhp9BJI/oROqKSYPuhI6Pe2/eMOKZyLod+PxMI/X5KZZSLkPncmrPZzAYAy8zs/5rZJcA/VLSGIpKI7y3zwVNUhJRG9CN0RMVzYHvo5L3e20+Q7DWVczn0jQKeQr+fUh0+DbKPx8tdCPwdGAVMTbNSIlKe7y3zSpOQoDyFdba1b2H+8nbmLN3A/OXtvf8A9Ux/kDh5b4Xv2PXefjx90PxZjzFnzsvMn/UYbYdMLprCJJVzOSPpJGoiGbOkrtsz2Tm3wjm3xTnX4Zz7hnPuC8C+VaibiJSQ2q319SSN6EeCdaYW1fBIf5AompTClES+208yfVBq53IG0kk0dKS7gZRLDNs3Tg77JTM7LC473cweJBroLyKB+E6NU5dT6PhKI/qRYJ0hoxqJokkpJNo1z/Ik0wc18rnc0JHuBlLuLsvriLonHwHmmVkuJ9ks59xvqlE5kUaTZOBu3U2Nk4Y0ptbyXGeSqEalB2xPGD6IRS9uZJvt6sKOEsju0XXhBOPifOtZ6laxwvJE0wclETI1SQomDB/U6Q5TaKBIdwMp1yBrAcY653aY2UBgA/CPzrm/VadqIo0ljdv66zLtRUY09e9T8q7AfGm8R4kSyHom2k1ST999bxrQt/hyRcZIem8/dGqSFCiJbGMo1yB7yzm3A8A5t8XM/p8aYyI94xNZKNfF1dN/vInXmZXIQgYSvvpGNZK+R15Rqhsuo3ndCpqX3Ny5fOWDXfdpxmzaSjXeelhP331PEvnx3n65LthaPJc9eUfEs3INSxflGmTvMrMn4r8NODh+bIDL5SQTkfJ8v9mnMXC3R4O7az2ykJGEr75RjaRdm15RoiTdkOOnsmjwyTu7NzsGj2LRh+bC6L06JRlOUs80pg/y3n7o1CQhZeUalqLKNcgOqVotROqY7zd7326eJBKtMyuRhThVQpdpo3pTz5T23SeqkeQ98o4SJZjvs3XN5k5jzQC2Wd9en5++ER3f5by3XwtznYaKUmXlGpaiSv6nj9NdlPypZiVFssz3m30at/UnWmdGIgttw48qniph+FE9X2nAfT+4qb93uXeUKEF6Dt91JqlnGrzP5dDJXlNII+ItI9ewFKd7ZkVS5nvLehq39SdaZxqJVFPQOvny4qkSJl/e85UG3PfnO7Z6l3unP0iQnsM3RUWSeqbB+1wOnew1QRqRiicOzsg1LMX5TC4uIr2QZOByGqksvNc5Y3bn8SdQk9PIdOw9IlG5l4D7nmRs1sFN/Vn6yptFy7vwTM/hnaKiBpKTep/LaaQ78eUZpUrlDuiMXMNSXLcRMjP7vE9ZkWWuN7N1Zra8yHNfMjNnZvvEj83M5pnZc2b2hJkd6bsDIrUuMwktQ0cWPDV1vJyo3EvAfU+S9DONKJXv9n0jaQ3PM0qVSuLgjFzDUpxPhGw68IOCsnOLlBW6kSij/0/zC81sFHAKkP91YQowJv55N/Cf8W+RsiqdUDMtmUniGjKy4GlC33UsemuvTt2W/d7azIS+63q34kD7niSCmkay2QnDB3Hnyk1sz9t+3yLb942kJd1+8DQNld6+Z5QqtYhjBq5hKa5kg8zMPgp8DDjIzO7Ie2pP4JXuVuycu9/MRhd5ai7wZeC/8srOAH7qnHPAw2a2t5nt55xb2/0uSKNS0tPG1HziRHhgMa1vDKOjaV+aOl5mQt91UXkGJUn9kFay2R2u/OMk2060/dBpGtLYfu513TTy0rirWrKtXITsQWAtsA/w/bzy14Enir6iG2b2AWC1c+5xs06B7hHAS3mPV8VlapBJSWkkUhVPCaIKaUQxm0+cmJcja1iv1lULfCOoaSSbbV2zuUuUy9F12bpM4prW9j2iVJoOSQqVbJA551aY2Srg78651t5uyMwGAZcB7y/2dLEqlFjPTGAmwP77686RRlYLg4wbUoKogqKYlZVGslnfZesyiWvA7Ws6JClUdgyZc267mW02s72ccxt7ua2DgQOBXHRsJPCYmR1DFBEblbfsSGBNiTotABYAtLS0lBq+IA2g4UP+GUg+qShm5VU62WySZesuiWuS7adwvWVmbKlUhc8n1xbgSTO7Lr4Tcp6ZzUu6Iefck865Yc650c650USNsCPj+THvAM6J77Y8Ftio8WPSnTQSqWZGRpJPKooZRpJrI43ryDuJbOgkrr7bD3m9ScPwaZD9Dvg34H7g0byfsszsZuAh4J1mtsrMziuz+J3AC8BzwI+BCzzqJQ0uM+kk0pAg+WTFJUg+mSSlg1ROkmsjjevIOz1H6DQNvtsPeb1Jw7DoxsZsamlpcUuWLAldDZHqm9Qn+qZeyAzuTjn6VDiGDKKoQpEPssIxZBBFXxqm4dyg5jy2PjoXCznHrCOHVr9CvRXyepO6YmaPOudaij3nkxh2jJndamZ/NbMXcj+Vr6aIeAs5RUqCqEZDRzEbWFPH3xKV1zxNSSRV4JMY9gbgCqL8Ye8FZqDkzCKJVDz1Q+gpUhIkn9TA5cYzYdE3WHTmVV2T9y76Bpx0a4/XGywRdOjrTRqCz0CO3Z1zi4m6N1c4574OvC/daonUj1y3XW4gey71Q68mEg499kakjObVS5hy2yU0vfoSuB00vfoSU267hObVPR9iksp15EvXm1SBT4Rsi5n1AZ41swuB1dRDJkaRKkkt9YOmSJFaNWM2zVfPpHnZ7bvKcuMMeyh4ChVdb5IynwjZxcAg4CLgKOBsovktRcSDUj9Iw0khoqTrSOpdtxEy59xfAMzMOedmpF8lkfrS8AlspealMjarwhElXUdS73zusjzOzP4KPBU/PsLM5qdeM5E60dAJbKXmBR2blYCuI6l3Pl8trgYmAa8AOOceB96TZqVE6olSP0h32tq3MH95O3OWbmD+8vaqNobKjc2qJYmuo8UL4ezRUf6ws0cro75kgs+gfpxzL1nnJH/b06mOSH1S6gcpJfQE7Fkam+V1HRUmLs5NcwQalC81zSdC9pKZHQ84MxtgZl8i7r4UEZHeCR2hqrvprTTNkWSUT4Ts08APgBFEE4LfA3w2zUqJBLV4YfTPe/3KKBP3jNk1+c06WJJM8ebzHoWOUE0YPqjo9FbFxmaFPue8tr9+JW3jzqR18uV07D2CptdWM+GuK2l+/Nfpb1ukF3zustwA1N6nkUgaMtLdEbqbS7rn+x6FvnswV5fuGhuhzznf7bed+CkWnXLZzlkCOgaPYtHUudC0D80pb1ukN0o2yMzsP4CSM4875y5KpUYiIZXr7qihBlnwJJnSLd/3KEmEKom2BxbTun0YHU370tTxMhP6rqP5xIlFl/UZmxX6nPPdfuvky9nWd/fOyw0YROvky3vcIAu970BmIvfSc+UiZPlzXHyDaD5Lkfq2fmWy8kBCd3NJ93zfI98IVRJtDyxm0YAxu6JEe+3Horf2ggcWl2yUdSf0Oee7/Y6Cxlh35ZXcdmoyErmX3inZIHPO/ST3t5ldnP9Y6l/DjpcYun/0z65YeQ0J3c0l3UvyHlX6LtzW7cM6TewNcZTojWE9jhKFPud8t28U79qxImWV3nZq0orcK+pWU3zPppJdl1J/spIoMhUzZkdz7uXbbVBUXkOUJLP2HdzUP1F5JXU07Zuo3Efoc853+6U+rHrzIRZ631OJ3OeibutWgHO7om7K2RaMVx4yaSw1MV4ilNy3wxr/1phGN5dU1vMdWxOVV5K5HTjr+n3bXM+72EKfc77bTyOalXTfK97DkEbkPiPjZRtJuUH9r7PrS8UgM+vIPQU451xT2pWTMIKPlwitwnPwpUXJZmtbyOvIWd9E5b5Cn3M+20/rJgnffU/ljswZszuPIYPeR+4zMl62kZT8yuCc29M51xT/9Mv7e081xupb3SWKFAkg5HXUNKB4w6tUeT0JPVVZKol+J06DixfAsAPALPp98YLefXEsFV2rsfGyjURdltJFWt8wRcqptxtJQl5HjX4Nh4zkpRYZrXTkPo2om/SKGmTSReixItJ46jHxZsjrSNdwOMHvyPSVkfGyjcScy+4NlC0tLW7JkiXdLygiNW3+8vaSH2IXHDYkQI2kUVQ6Mlv45QKi6GQ1u02ldpnZo865lmLPKUImIsE1/I0kEkQakVlFJ6WnetQgM7MFzrmZla6MiNQfnwhEZrp5qIGxbkrmWTFppfgJfTeqZFNP/9v9qKK1EJG65JtkOGQS1SSCJ01WMs+KUmRWakm3ETIzG+icK/xvUyRDnUhjCR4pyQDfCETiJKqBokTBkyanlMwzjXM5C9dHWpHZkPueaNuKttYUn7PuL2Z2bO6BmU0FHkyvSiK1L3ikJCO8J4ROEqkIGCUKHlFJIZlnGudyVq6PNKZECrnvibataGvN8WmQfQz4DzP7rpktBD4FvC/daonUtlSSP9Yh3+SoiZKolosSpSx40uQUknmmcS5n5fpII4lsyH1PtO2A15EU122XpXPuSTObDfwMeB14j3NuVeo1E6lhwSMlGeGboDRRItOAU74ET7g6YzZti39H68n/SsfeI2h6bTUT7v0uzRNP6/Eq0ziXs3R9VHoAfsh9T7RtTZ1Uc3zGkF0HHAyMBd4B/NbMrnHO/TDtyonUqizdFRiSbwqARKkC0pho2VPolAZt46eyaPDJbIvnpOwYPIpFH5oLo/eiuYfrTONcbuTrI+S+J9p2wOtIivNJe7Ec+KSLMsj+bzye7Kp0qyVS24JHSjLENwLhHamYMRu+/wnY9tausn4DqjblS8iUBq1rNu9sjOVss769uqkgjXO5ka+PzEyZpamTao5Pl+XcgscbgfNSq5FIBoSOlDS8whlGMjzjSBJpdIelcS438vWRmSmzNHVSzel26iQzGwN8GzgU2PmuOucOSrdq3dPUSRKUbhkP4+zRtA0/itbJl+8aR3XXlTSveRRuejF07VL1naUbKPYf24BLx+/TubCRz89G3nepaeWmTvLp1L4B+E9gG/Be4KdEA/xFGpduGQ+mbUQLi6bOpWPwKLA+0TiqqXNpG1H0f1xdKfX1uUt5I5+fjbzvkmk+DbLdnXOLiaJpK5xzX0dpL6TR6ZbxYFqnXMG2AZ3HxGwbMIjWKVcEqlH1eKfdaOTzs5H3PUsWL4SzR8OkPtFvNZi9GmRbzKwP8KyZXWhmfSKTwAAAEmtJREFU/wwMS7leIrVNt4wH09H09kTl9cQ7kWkjn5+NvO9ZoShmUT4NsouBQcBFwFHAx4Hp3b3IzK43s3Vmtjyv7Ltm9rSZPWFmvzazvfOe+4qZPWdmz5jZpOS7IlJFKSToFD9NA/omKq8n3olMG/n8bOR9zwpFMYvyucvyL/Gfm4AZCdZ9I3AN0ZiznN8DX3HObTOz7wBfAS41s0OBs4BmYDhwr5m9wzm3PcH2RKon4S3jmZnbLgMaOaUCeKbdaOSUBinte71dR0EpillUyQaZmd1R7oXOuQ908/z9Zja6oOyevIcPAx+K/z4D+IVz7k2iXGfPAccAD5XbhkgwCW4Zz80vl2tA5OaXA1L/hx5y22lp5JQK3ho5pUEK+16P11FQSkpbVLkI2XHAS8DNwJ+J7qyupE8At8R/jyBqoOWsistEatfEaV7/5MvNL5f2P/OQ205TyOSsmeF5ftalCu97vV5HwTRyBLeMcg2ytwOnAB8lmmD8d8DNzrm23m7UzC4jSqORG8FXrLFX9A5vM5sJzATYf//Gbk1LNmRmbjsRKUrXUYU1cgS3jJINsnj81l3AXWa2G1HD7D4z+6Zz7j96ukEzmw6cDkx0u7LSrgJG5S02ElhTol4LgAUQJYbtaT1EqiUzc9uhcTIixTTy3JypaeQIbgllzyYz283MzgRuAj4LzANu7+nGzGwycCnwAedc/i0WdwBnxds7EBgDPNLT7YjUEu9UBYG3nRsnk/vgyY2TaWvfkno9RWpZyGtYGke5Qf0/AQ4DFgHfcM4tL7VsidffDJwE7GNmq4AriO6q3A34vZkBPOyc+7Rzrs3Mfgn8lagr87O6w1LqRVbmttM4mcpTxLE+6EYSqYaSc1ma2Q7g7/HD/IUMcM65ppTr1i3NZSlSOXOWbij53KzCeRKlW4V35kEUVSmaN0xEGkK5uSzLjSFT57hIA9E4mcpSxFFEktB/WhEBNE7GV1v7FuYvb2fO0g3MX95ecoyd7swTqbKMz4/ZbaZ+EWkMGifTvSQJQhVxFKmi3PyYudxmufkxITN3c6pBJiI7NXTC1cULu82LlKQbMq0pnnSjQB3xOOfEU7n5MTNyTNUgExHx/HadpBsyjYijpvCpI3UQ0akpdTA/phpk0hAUVZCyPL9dJ+2GrHTEUTcKVJ73/4ZKR7PqIKJTU+pgfkwNZpC6p4Sn0i3Pb9ehb3zQjQKV5f2/IRfNWrcCnNsVzerNoPE6iOjUlBmzo/kw82Vsfkw1yKTulYsqiAClv0UXlDcPGcjhQ3bbOfmuAYcP2a1q0alSkTjdKNAz3v8bykWzesrznBNPE6fBxQtg2AFgFv2+eEGmoo26iqXuKaog3fL8dt3WvoUn29/cmSnbAU+2v1m1aGvoCF298f7fkEY0qw4iOjVn4jS46UW4e0f0O0ONMVCDTBqAogrSLc9v16Gjrc1DBjJl/z12nrtN/fso838veP9vSCOaVQcRHaksDeqXupdW+gGpMxOndfthqGhrffH+3zBjduc7IqEy0SyPcy4p3cCUXWqQSd1TwlOplNDJXpX2orK8/zfkGk01njNM50e2qUEmDaGhE56KH4+0BkmjrZWOViRJe6FIiR/v/w0pRLMqTWlRsk0NMhERzySdSaKtaUQrfLtMFSlpTOpSzzY1yEREEiTp9I2opBGt8O0yVaSkMYXuUpfe0bskIpJCWoM0ohW+aS8UKWlMSouSbYqQiYgkmHbFd2xWGtEK3y5TRUoaU/OQgazatJVlr0S58qqduFh6Rw0yERHPtAZJxmallW7Fp8tUqV4aU6nExSP36K9GWQbo65KISAqJYUMmcVUC2cYUOnGx9I4iZCIikEpi2JDpVpTqJZxQKUc0djDbFCETEfE0sMR/zFLl0nhy3dq5RlCuW7sa851qmrhs07skIuLJzBKVS+MJ2W2ouyyzTV2WIiKe3tjuEpVL4wnZbahp4rJNDTIREU9KJ1F59TbFU+hzRGMHs0v/RUREPKlLqLJCjrdKi84R6SlFyEREPKXVJZRGlCgLkafUpnjymCg+Leo2lJ5Sg0xEJIFKdwmlMRF4ViYXT2W8ledE8WlSt6H0hLosRUQCSuOuvKwkCE0lTUO5ieJFapgaZCIiAaURJcpKgtBUxlulMFG8SDWoQSYiElAaUaKsJAhNZYqnIhPCly0XqRG1dXWKiDSYCcMHUZhW1uhdlKih7/SbMTuaGD5fkYniRWqNBvWLiATWxyA/t2yfXib+z8qdfqncfJAbuB/oLkuRnlKDTEQkoNY1mylM9L+9AqkfsnCnX2ppLzwmihepNeqyFBEJKCsD8NPQyPsuUkgNMhGRgLIyAD8NjbzvIoV01ouIBNTIA/Abed9FCqXWIDOz681snZktzysbYma/N7Nn49+D43Izs3lm9pyZPWFmR6ZVLxGRWpJK6oeMaOR9FymU5qD+G4FrgJ/mlc0CFjvn5pjZrPjxpcAUYEz8827gP+PfIiJ1LwsD8NPSyPsuki+1CJlz7n6gvaD4DOAn8d8/AT6YV/5TF3kY2NvM9kurbiIiIiK1pNppL/Z1zq0FcM6tNbNhcfkI4KW85VbFZWsLV2BmM4GZAPvvr8zLIiJZ1ta+pebzpYlUQ60M6i+WBtEVKcM5t8A51+Kcaxk6dGjK1RIRkbTkEsPm0lzkEsO2tW8JXDOR6qt2g+zlXFdk/HtdXL4KGJW33EhgTZXrJiIiVVQuMaxIo6l2g+wOYHr893Tgv/LKz4nvtjwW2Jjr2hQRkfqkxLAiu6Q2hszMbgZOAvYxs1XAFcAc4Jdmdh6wEvhwvPidwKnAc8BmYEZa9RIRkdrQ1L9P0caXEsNKI0qtQeac+2iJpyYWWdYBn02rLiIiUnsOburP0lfeLFou0mg0ubiIiPhZvBBuuAzWr4Sh+8OM2b2axPv5jq2Jyn3pzk3JIjXIRESke4sXwtUz4c14wP26FdFj6HGjLI0xZLk7N3M3C+Tu3ATUKJOapo56ERHp3g2X7WqM5by5OSrvoTQmF9edm5JVipCJiKBurm6tX5ms3MOE4YM6RbOg95OL685NySpFyESk4SlBqYehJWZGKVXuIY3JxdOIuolUgyJkItLwynVzKUoWmzG78xgygN0GReW9UOnJxdOIuolUgxpkItLw1M3lITdwv4J3WaYh17hT97NkjRpkItLwlKDU08RpNdcAK6bSUTeRatB/GxFpeBOGD6KfdS5TN5eIVJMiZCLS8NTNJSKhqUEmIoK6uUQkLHVZioiIiASmBpmIiIhIYGqQiYiIiASmMWQiIg1O00aJhKcGmYhIA8tNG5XLbJ+bNgpQo0ykitRlKSLSwMpNGyUi1aMGmYhIA9O0USK1QV2WIiIZUunxXpo2SqQ26IoTEcmI3HivXAMqN96rrX1Lj9epaaNEaoMaZCIiGZHGeK/mIQOZsv8eOyNiTf37MGX/PTSgX6TK1GUpIpIRaY330rRRIuEpQiYikhGlxnVpvJdI9ukqFhHJCI33Eqlf6rIUEcmIXLeisuqL1B81yEREMkTjvUTqk7osRURERAJTg0xEREQkMDXIRERERAJTg0xEREQkMDXIRERERAJTg0xEREQkMDXIRERERAJTg0xEREQkMDXIRERERAJTg0xEREQksCANMjO7xMzazGy5md1sZgPN7EAz+7OZPWtmt5jZgBB1ExEREam2qjfIzGwEcBHQ4pw7DOgLnAV8B5jrnBsDvAqcV+26iYiIiIQQqsuyH7C7mfUDBgFrgfcBt8bP/wT4YKC6iYiIiFRV1RtkzrnVwPeAlUQNsY3Ao8Brzrlt8WKrgBHFXm9mM81siZktWb9+fTWqLCIiIpKqEF2Wg4EzgAOB4cA/AFOKLOqKvd45t8A51+Kcaxk6dGh6FRURERGpkhBdlicD/+ucW++c2wrcDhwP7B13YQKMBNYEqJuIiIhI1YVokK0EjjWzQWZmwETgr8AfgQ/Fy0wH/itA3URERESqLsQYsj8TDd5/DHgyrsMC4FLgC2b2HPA24Lpq101EREQkhH7dL1J5zrkrgCsKil8AjglQHREREZGglKlfREREJDA1yEREREQCU4NMREREJDA1yEREREQCU4NMREREJDA1yEREREQCU4NMREREJDA1yEREREQCU4NMREREJDA1yEREREQCU4NMREREJDA1yEREREQCCzK5uIiISL1qa99C65rNdGzdQVP/PkwYPojmIQNDV0tqnBpkIiIiFdLWvoVFKzexzUWPO7buYNHKTQBqlElZ6rIUERGpkNY1m3c2xnK2uahcpBw1yERERCqkY+uOROUiOWqQiYiIVEhT/+Ifq6XKRXJ0hoiIiFTIhOGD6Gedy/pZVC5Sjgb1i4iIVEhu4L7uspSk1CATERGpoOYhA9UAk8TUZSkiIiISmBpkIiIiIoGpQSYiIiISmBpkIiIiIoGpQSYiIiISmBpkIiIiIoGpQSYiIiISmBpkIiIiIoGpQSYiIiISmBpkIiIiIoGpQSYiIiISmBpkIiIiIoGZcy50HXrMzNYDK0LXI88+wIbQlZCy9B7VPr1HtU/vUe3Te1SbDnDODS32RKYbZLXGzJY451pC10NK03tU+/Qe1T69R7VP71H2qMtSREREJDA1yEREREQCU4OsshaEroB0S+9R7dN7VPv0HtU+vUcZozFkIiIiIoEpQiYiIiISmBpkPWRmA83sETN73MzazOwbcfmBZvZnM3vWzG4xswGh69rIzKyvmS01s/+OH+v9qTFm9qKZPWlmy8xsSVw2xMx+H79PvzezwaHr2cjMbG8zu9XMnjazp8zsOL1HtcPM3hlfP7mfDjO7WO9RtqhB1nNvAu9zzh0BjAMmm9mxwHeAuc65McCrwHkB6yjweeCpvMd6f2rTe51z4/Ju058FLI7fp8XxYwnnB8Bdzrl3AUcQXVN6j2qEc+6Z+PoZBxwFbAZ+jd6jTFGDrIdcZFP8sH/844D3AbfG5T8BPhigegKY2UjgNOD/ix8ben+y4gyi9wf0PgVlZk3Ae4DrAJxzbznnXkPvUa2aCDzvnFuB3qNMUYOsF+LusGXAOuD3wPPAa865bfEiq4ARoeonXA18GdgRP34ben9qkQPuMbNHzWxmXLavc24tQPx7WLDayUHAeuCGuPv//zOzf0DvUa06C7g5/lvvUYaoQdYLzrntcYh4JHAMcEixxapbKwEws9OBdc79/+3dPYhcVRjG8f+jm4AaYVGCKFGXNBGMYjYgxgUJKkJAgh9rMSiIaGebxtaQwk4QFBsLC5FEESWIIhgFmwh+QSJooWKWuAkSCGihha/FPXGVCGIIe2b3/n/NzP0YeOFlhodzztxTn/399L/can/6W6iqeWAP8HSSu3oXpH+YAeaBl6pqB/ArTn1NpbYmdi9wqHct+v8MZBdBG77/CLgDmE0y0y5tAU72qmvkFoC9SX4AXmeYqnwe+zN1qupkez3NsO7lduBUkmsB2uvpfhWO3hKwVFVH2/EbDAHNHk2fPcDnVXWqHdujNcRAdoGSbE4y295fBtzLsND1CLDYbnsceLtPheNWVc9U1ZaqmmMYwv+wqh7F/kyVJFckufLce+A+4BjwDkN/wD51VVXLwIkk29qpe4CvsUfTaMLKdCXYozXFB8NeoCS3MiySvJQh2B6sqmeTbGUYkbkK+AJ4rKp+61epkuwG9lXV/fZnurR+vNUOZ4DXqupAkquBg8ANwI/AI1V1plOZo5fkNoY/x2wEvgOeoP3uYY+mQpLLgRPA1qo62875PVpDDGSSJEmdOWUpSZLUmYFMkiSpMwOZJElSZwYySZKkzgxkkiRJnRnIJI1SkgeTVJKbetciSQYySWM1AT5heHCwJHVlIJM0Okk2MWyv9SQtkCW5JMmLSY4nOZzk3SSL7drOJB+3DdDfP7cdjSRdLAYySWP0APBeVX0LnEkyDzwEzAG3AE8BuwCSbABeABaraifwCnCgR9GS1q+Z/75FktadCcNm8zBspTUBNgCHquoPYDnJkXZ9G7Ad+CAJDNul/bS65Upa7wxkkkal7e93N7A9STEErGJlT83zPgIcr6pdq1SipBFyylLS2CwCr1bVjVU1V1XXA98DPwMPt7Vk1wC72/3fAJuT/DWFmeTmHoVLWr8MZJLGZsL5o2FvAtcBS8Ax4GXgKHC2qn5nCHHPJfkK+BK4c/XKlTQGqareNUjSVEiyqap+adOanwILVbXcuy5J659ryCRpxeEks8BGYL9hTNJqcYRMkiSpM9eQSZIkdWYgkyRJ6sxAJkmS1JmBTJIkqTMDmSRJUmcGMkmSpM7+BI/LI4unD31rAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(10, 6))\n",
    "\n",
    "# Scatter with positive examples\n",
    "plt.scatter(df.age[df.target==1],\n",
    "            df.thalach[df.target==1],\n",
    "            c='orangered')\n",
    "\n",
    "# Scatter with negative examples\n",
    "plt.scatter(df.age[df.target==0],\n",
    "            df.thalach[df.target==0],\n",
    "            c='skyblue')\n",
    "\n",
    "plt.title(\"Age vs Max Heart Rate for Heart Disease\")\n",
    "plt.xlabel(\"Age\")\n",
    "plt.ylabel(\"Max. Heart Rate\")\n",
    "plt.legend([\"Disease\", \"No Disease\"]);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Check the distribution of the age column\n",
    "df['age'].plot(kind='hist',\n",
    "               color='lightblue');"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Heart Disease Frequency per Chest Pain Type"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- 0: Typical angina (chest pain related decreased blood supply to heart).\n",
    "- 1: Atypical angina (chest pain not related to heart).\n",
    "- 2: Non anginal pain (typically esophageal spasms - non heart related).\n",
    "- 3: Asymptomatic (chest pain not showing signs of disease)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "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>target</th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>cp</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>104</td>\n",
       "      <td>39</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>9</td>\n",
       "      <td>41</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>18</td>\n",
       "      <td>69</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>7</td>\n",
       "      <td>16</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "target    0   1\n",
       "cp             \n",
       "0       104  39\n",
       "1         9  41\n",
       "2        18  69\n",
       "3         7  16"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.crosstab(df['cp'], df['target'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Surprisingly, non aginal pain –which is supposed to not be related to heart problems– is significally more present on patients that do have heart disease than those who not (69 vs 18)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "pd.crosstab(df['cp'], df['target']).plot(kind='bar',\n",
    "                                         figsize=(10,6),\n",
    "                                         color=['lightblue', 'orangered'])\n",
    "\n",
    "plt.title(\"Heart Disease Frequency per Chest Pain Type\")\n",
    "plt.xlabel(\"Chest Pain Type\")\n",
    "plt.ylabel(\"Amount\")\n",
    "plt.legend([\"No Disease\", \"Disease\"])\n",
    "plt.xticks(rotation=0);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Make a correlation matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "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>age</th>\n",
       "      <th>sex</th>\n",
       "      <th>cp</th>\n",
       "      <th>trestbps</th>\n",
       "      <th>chol</th>\n",
       "      <th>fbs</th>\n",
       "      <th>restecg</th>\n",
       "      <th>thalach</th>\n",
       "      <th>exang</th>\n",
       "      <th>oldpeak</th>\n",
       "      <th>slope</th>\n",
       "      <th>ca</th>\n",
       "      <th>thal</th>\n",
       "      <th>target</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>age</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>-0.098447</td>\n",
       "      <td>-0.068653</td>\n",
       "      <td>0.279351</td>\n",
       "      <td>0.213678</td>\n",
       "      <td>0.121308</td>\n",
       "      <td>-0.116211</td>\n",
       "      <td>-0.398522</td>\n",
       "      <td>0.096801</td>\n",
       "      <td>0.210013</td>\n",
       "      <td>-0.168814</td>\n",
       "      <td>0.276326</td>\n",
       "      <td>0.068001</td>\n",
       "      <td>-0.225439</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sex</th>\n",
       "      <td>-0.098447</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>-0.049353</td>\n",
       "      <td>-0.056769</td>\n",
       "      <td>-0.197912</td>\n",
       "      <td>0.045032</td>\n",
       "      <td>-0.058196</td>\n",
       "      <td>-0.044020</td>\n",
       "      <td>0.141664</td>\n",
       "      <td>0.096093</td>\n",
       "      <td>-0.030711</td>\n",
       "      <td>0.118261</td>\n",
       "      <td>0.210041</td>\n",
       "      <td>-0.280937</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>cp</th>\n",
       "      <td>-0.068653</td>\n",
       "      <td>-0.049353</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.047608</td>\n",
       "      <td>-0.076904</td>\n",
       "      <td>0.094444</td>\n",
       "      <td>0.044421</td>\n",
       "      <td>0.295762</td>\n",
       "      <td>-0.394280</td>\n",
       "      <td>-0.149230</td>\n",
       "      <td>0.119717</td>\n",
       "      <td>-0.181053</td>\n",
       "      <td>-0.161736</td>\n",
       "      <td>0.433798</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>trestbps</th>\n",
       "      <td>0.279351</td>\n",
       "      <td>-0.056769</td>\n",
       "      <td>0.047608</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.123174</td>\n",
       "      <td>0.177531</td>\n",
       "      <td>-0.114103</td>\n",
       "      <td>-0.046698</td>\n",
       "      <td>0.067616</td>\n",
       "      <td>0.193216</td>\n",
       "      <td>-0.121475</td>\n",
       "      <td>0.101389</td>\n",
       "      <td>0.062210</td>\n",
       "      <td>-0.144931</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>chol</th>\n",
       "      <td>0.213678</td>\n",
       "      <td>-0.197912</td>\n",
       "      <td>-0.076904</td>\n",
       "      <td>0.123174</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.013294</td>\n",
       "      <td>-0.151040</td>\n",
       "      <td>-0.009940</td>\n",
       "      <td>0.067023</td>\n",
       "      <td>0.053952</td>\n",
       "      <td>-0.004038</td>\n",
       "      <td>0.070511</td>\n",
       "      <td>0.098803</td>\n",
       "      <td>-0.085239</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fbs</th>\n",
       "      <td>0.121308</td>\n",
       "      <td>0.045032</td>\n",
       "      <td>0.094444</td>\n",
       "      <td>0.177531</td>\n",
       "      <td>0.013294</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>-0.084189</td>\n",
       "      <td>-0.008567</td>\n",
       "      <td>0.025665</td>\n",
       "      <td>0.005747</td>\n",
       "      <td>-0.059894</td>\n",
       "      <td>0.137979</td>\n",
       "      <td>-0.032019</td>\n",
       "      <td>-0.028046</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>restecg</th>\n",
       "      <td>-0.116211</td>\n",
       "      <td>-0.058196</td>\n",
       "      <td>0.044421</td>\n",
       "      <td>-0.114103</td>\n",
       "      <td>-0.151040</td>\n",
       "      <td>-0.084189</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.044123</td>\n",
       "      <td>-0.070733</td>\n",
       "      <td>-0.058770</td>\n",
       "      <td>0.093045</td>\n",
       "      <td>-0.072042</td>\n",
       "      <td>-0.011981</td>\n",
       "      <td>0.137230</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>thalach</th>\n",
       "      <td>-0.398522</td>\n",
       "      <td>-0.044020</td>\n",
       "      <td>0.295762</td>\n",
       "      <td>-0.046698</td>\n",
       "      <td>-0.009940</td>\n",
       "      <td>-0.008567</td>\n",
       "      <td>0.044123</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>-0.378812</td>\n",
       "      <td>-0.344187</td>\n",
       "      <td>0.386784</td>\n",
       "      <td>-0.213177</td>\n",
       "      <td>-0.096439</td>\n",
       "      <td>0.421741</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>exang</th>\n",
       "      <td>0.096801</td>\n",
       "      <td>0.141664</td>\n",
       "      <td>-0.394280</td>\n",
       "      <td>0.067616</td>\n",
       "      <td>0.067023</td>\n",
       "      <td>0.025665</td>\n",
       "      <td>-0.070733</td>\n",
       "      <td>-0.378812</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.288223</td>\n",
       "      <td>-0.257748</td>\n",
       "      <td>0.115739</td>\n",
       "      <td>0.206754</td>\n",
       "      <td>-0.436757</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>oldpeak</th>\n",
       "      <td>0.210013</td>\n",
       "      <td>0.096093</td>\n",
       "      <td>-0.149230</td>\n",
       "      <td>0.193216</td>\n",
       "      <td>0.053952</td>\n",
       "      <td>0.005747</td>\n",
       "      <td>-0.058770</td>\n",
       "      <td>-0.344187</td>\n",
       "      <td>0.288223</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>-0.577537</td>\n",
       "      <td>0.222682</td>\n",
       "      <td>0.210244</td>\n",
       "      <td>-0.430696</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>slope</th>\n",
       "      <td>-0.168814</td>\n",
       "      <td>-0.030711</td>\n",
       "      <td>0.119717</td>\n",
       "      <td>-0.121475</td>\n",
       "      <td>-0.004038</td>\n",
       "      <td>-0.059894</td>\n",
       "      <td>0.093045</td>\n",
       "      <td>0.386784</td>\n",
       "      <td>-0.257748</td>\n",
       "      <td>-0.577537</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>-0.080155</td>\n",
       "      <td>-0.104764</td>\n",
       "      <td>0.345877</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ca</th>\n",
       "      <td>0.276326</td>\n",
       "      <td>0.118261</td>\n",
       "      <td>-0.181053</td>\n",
       "      <td>0.101389</td>\n",
       "      <td>0.070511</td>\n",
       "      <td>0.137979</td>\n",
       "      <td>-0.072042</td>\n",
       "      <td>-0.213177</td>\n",
       "      <td>0.115739</td>\n",
       "      <td>0.222682</td>\n",
       "      <td>-0.080155</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.151832</td>\n",
       "      <td>-0.391724</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>thal</th>\n",
       "      <td>0.068001</td>\n",
       "      <td>0.210041</td>\n",
       "      <td>-0.161736</td>\n",
       "      <td>0.062210</td>\n",
       "      <td>0.098803</td>\n",
       "      <td>-0.032019</td>\n",
       "      <td>-0.011981</td>\n",
       "      <td>-0.096439</td>\n",
       "      <td>0.206754</td>\n",
       "      <td>0.210244</td>\n",
       "      <td>-0.104764</td>\n",
       "      <td>0.151832</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>-0.344029</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>target</th>\n",
       "      <td>-0.225439</td>\n",
       "      <td>-0.280937</td>\n",
       "      <td>0.433798</td>\n",
       "      <td>-0.144931</td>\n",
       "      <td>-0.085239</td>\n",
       "      <td>-0.028046</td>\n",
       "      <td>0.137230</td>\n",
       "      <td>0.421741</td>\n",
       "      <td>-0.436757</td>\n",
       "      <td>-0.430696</td>\n",
       "      <td>0.345877</td>\n",
       "      <td>-0.391724</td>\n",
       "      <td>-0.344029</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               age       sex        cp  trestbps      chol       fbs  \\\n",
       "age       1.000000 -0.098447 -0.068653  0.279351  0.213678  0.121308   \n",
       "sex      -0.098447  1.000000 -0.049353 -0.056769 -0.197912  0.045032   \n",
       "cp       -0.068653 -0.049353  1.000000  0.047608 -0.076904  0.094444   \n",
       "trestbps  0.279351 -0.056769  0.047608  1.000000  0.123174  0.177531   \n",
       "chol      0.213678 -0.197912 -0.076904  0.123174  1.000000  0.013294   \n",
       "fbs       0.121308  0.045032  0.094444  0.177531  0.013294  1.000000   \n",
       "restecg  -0.116211 -0.058196  0.044421 -0.114103 -0.151040 -0.084189   \n",
       "thalach  -0.398522 -0.044020  0.295762 -0.046698 -0.009940 -0.008567   \n",
       "exang     0.096801  0.141664 -0.394280  0.067616  0.067023  0.025665   \n",
       "oldpeak   0.210013  0.096093 -0.149230  0.193216  0.053952  0.005747   \n",
       "slope    -0.168814 -0.030711  0.119717 -0.121475 -0.004038 -0.059894   \n",
       "ca        0.276326  0.118261 -0.181053  0.101389  0.070511  0.137979   \n",
       "thal      0.068001  0.210041 -0.161736  0.062210  0.098803 -0.032019   \n",
       "target   -0.225439 -0.280937  0.433798 -0.144931 -0.085239 -0.028046   \n",
       "\n",
       "           restecg   thalach     exang   oldpeak     slope        ca  \\\n",
       "age      -0.116211 -0.398522  0.096801  0.210013 -0.168814  0.276326   \n",
       "sex      -0.058196 -0.044020  0.141664  0.096093 -0.030711  0.118261   \n",
       "cp        0.044421  0.295762 -0.394280 -0.149230  0.119717 -0.181053   \n",
       "trestbps -0.114103 -0.046698  0.067616  0.193216 -0.121475  0.101389   \n",
       "chol     -0.151040 -0.009940  0.067023  0.053952 -0.004038  0.070511   \n",
       "fbs      -0.084189 -0.008567  0.025665  0.005747 -0.059894  0.137979   \n",
       "restecg   1.000000  0.044123 -0.070733 -0.058770  0.093045 -0.072042   \n",
       "thalach   0.044123  1.000000 -0.378812 -0.344187  0.386784 -0.213177   \n",
       "exang    -0.070733 -0.378812  1.000000  0.288223 -0.257748  0.115739   \n",
       "oldpeak  -0.058770 -0.344187  0.288223  1.000000 -0.577537  0.222682   \n",
       "slope     0.093045  0.386784 -0.257748 -0.577537  1.000000 -0.080155   \n",
       "ca       -0.072042 -0.213177  0.115739  0.222682 -0.080155  1.000000   \n",
       "thal     -0.011981 -0.096439  0.206754  0.210244 -0.104764  0.151832   \n",
       "target    0.137230  0.421741 -0.436757 -0.430696  0.345877 -0.391724   \n",
       "\n",
       "              thal    target  \n",
       "age       0.068001 -0.225439  \n",
       "sex       0.210041 -0.280937  \n",
       "cp       -0.161736  0.433798  \n",
       "trestbps  0.062210 -0.144931  \n",
       "chol      0.098803 -0.085239  \n",
       "fbs      -0.032019 -0.028046  \n",
       "restecg  -0.011981  0.137230  \n",
       "thalach  -0.096439  0.421741  \n",
       "exang     0.206754 -0.436757  \n",
       "oldpeak   0.210244 -0.430696  \n",
       "slope    -0.104764  0.345877  \n",
       "ca        0.151832 -0.391724  \n",
       "thal      1.000000 -0.344029  \n",
       "target   -0.344029  1.000000  "
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.corr()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA0AAAAJCCAYAAADp3Kf9AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAgAElEQVR4nOzdd3gU1f7H8fdJSG+bHgKE0ElCJ0GQ3qvlWkH8CVgQFVHEqyiCYlcsCFdBvBZUpCgXRUARAemQhJpGTyghkLrpPfP7Y0MKWRAjmQXzfT1PHsjO2ewnk++ZnTNnZlZpmoYQQgghhBBC1AdWlg4ghBBCCCGEEHqRAZAQQgghhBCi3pABkBBCCCGEEKLekAGQEEIIIYQQot6QAZAQQgghhBCi3pABkBBCCCGEEKLekAGQEEIIIYQQQndKqS+UUslKqejLLFdKqXlKqeNKqUNKqS7X4nVlACSEEEIIIYSwhK+AYVdYPhxoVf41EVhwLV5UBkBCCCGEEEII3WmathVIv0KT24CvNZPdgEEp1fDvvm6Dv/sDrnOapQMIIYQQQoh6QVk6wNVwCBij2/5xwZllj2KaublokaZpi/7Cj2gEnKny/dnyx5L+Tq5/+gAIh4Axlo5QK/mnlwKw4uSvFk5SO/c0N81mfnl0vYWT1M6E1kMJW7Hd0jFqLeKeXvRft8PSMWpt84iePLBli6Vj1MrXffsC8NOpXyycpHZuazocAGPROgsnqR2D7QgAJm7/w7JBamlRr34M/OXG7bsbh/dk2/m1lo5Ra739RtLt+xtz2x9+dy8Ant69ycJJamdu9wEAHEhbY+EktdPJc5SlI1yXygc7f2XAcylzg8q/PYCTU+CEEEIIIYQQ16OzQJMq3zcGzv3dH/qPnwESQgghhBBCmCh1Q81/rAYmK6WWATcBmZqm/a3T30AGQEIIIYQQQggLUEotBfoBXkqps8DLgA2ApmkLgXXACOA4kAdMuBavKwMgIYQQQggh6gl1HV0Bo2naFS/W1zRNA5641q97/awBIYQQQgghhKhjMgMkhBBCCCFEPXGDXQNUJ2QNCCGEEEIIIeoNmQESQgghhBCinpAZIJkBEkIIIYQQQtQjMgMkhBBCCCFEPaGUsnQEi5MZICGEEEIIIUS9ITNAQgghhBBC1Bsy/yFrQAghhBBCCFFvyABICCGEEEIIUW/IKXBmLJzzKMMHdiYlLYvQwc+ZbfP+7HEM7d+JvPwiJk5bwIHoBADG3tWH6U/eDsDb839kyQ9b9YpdjaZprFv4P45GxGJjZ8Md08bi37JJjXYbvlrDgY0RFOTkMXPVnIrHS4pKWPn+t5w7dgZHVyfueWEc7r6eumX/fdFKTuyNxcbOlpFPjcXPTPbzx0+zdu4SiouKadE1mEET70QpxY/vfEl6YjIABbn52Ds58OC853XJDtDDz8C0Ts2xUoqf4i+w+PDZasvva+3Pbc38KNU0jIXFvBpxjPN5hQA82SGQXg3dUUqx54KR9/ef1C33RWFeBiYHN8dawdozF1h6MrHa8rub+TOisS+lmkZmUTHvHjrOhQJT/nfCggk2uBCVkcWLkXG6ZwcwRkdzavlytLIyfHr1wn/48GrLs44e5dTy5eQlJtLykUfw7NoVgNwzZ0hYsoTS/HywsqLRiBF4hoXpnl/TNFZ/8j8OR8RhY2fDPc/eR+NWNev/1y/XsndDBPk5eby++t2Kx7f+sJnwX3djZW2Fs5szd08bg7uvh675P3h7FTu3xWFvb8PM18fQNrhm/oueffK/JJ5NY+kqUx/NzMzlpWe/5ty5dPz9PXjjvXG4ujnqkj09KprjS1egaWU07N2LgBHDqi03HjnKiWUryDmbSPCjD+Md2rVi2fkduzi9Zh0AAaNG4Nezhy6ZqwrzMvBEUHOsFKw7e4Fll/TduwL9GdHEl9IyDWNRMXOijpNc3nffCjX13eiMLGbstUzf1TSNpfNWEbUnDls7Wx58YQxNWzeu0e5/n61j1/pI8nLy+PjXtyseX/afHzmy/zgARQXFZBmzmb/2Td3yd/c1MK1z+bb/5AW+PnLJtr+VP7c29zOt/8JiXoussu1vH0jP8m1/+AUj7x/Qf9ufciiGuCUr0Mo0GvftSYtRQ6stLy0u5tCixWQlnMbG2YlOjz+Mo7cniTvDif9lQ0W77DOJ9Jz9Aq5NL9/vrzVN0/jqwx/ZvysOO3tbHntpNM3bVK+dwoIiPpzxNRcSU7GytqJrz2Due3wUABtW7WT9yh1YWVth72DLxOfvpnEzP93y601ugy0zQGZ98/0Wbnvg7csuH9q/Ey0C/WjXZyqTp3/GvDceAsDdzYkZT99Bn1tn0vvWmcx4+g4Mbk56xa7mWEQsaedSePrzl7htymh+/s/3Ztu1vakdkz56psbje3/bhYOzA1O/mEmP2/vx2xc/13XkCif3xpJxLoVHP53JsCfuZf2CFWbbrf9kBcMmj+bRT2eScS6Fk+Vv2rc/P4EH5z3Pg/Oep83NHWndo4Nu2a0UPNelBU9ti+Ge9fsYEuBNM1eHam2OZOTywO8HuO+3/Ww8m8qUDoEAdPB0oaOXK2N+28/o9fsIdnemi7ebbtnBtEF4KqQ50yNiGL91PwP9vWnqXD3/scxcJu04yMPbD7DlfBqPtg2sWLb8ZCJvHjyqa+aqtLIyEr77jjZTptBh9mzSIiLIO3euWhs7Dw9aTJiAV7du1R63srWlxYQJdJg9m7ZPPcWp5cspycvTMz4AhyPiSE1M4bkvZ3Dn0/eyap75vhvUPYQn50+t8bh/y8ZM+c80nvn0edr37sja/66u68jV7NwWx5lTKfyw9kWmv3wP777+w2Xbbv79EA4OdtUe+/rzjYTe1IqVa2cQelMrvv58Y11HBky1c2zJUtpPfZKw114heU8EuZfUjr2nB20eHI/vTdVrpzgnl1Or19B5xnQ6vzSdU6vXUJybq0vui6yAKSHNeSEyhge37WdAw5p993hWLo/tOMgjOw6w9UIaE6v03RXxibx9yHJ9FyBqTxzJZ1N5c8mLPPDs3Xz7gfna6XhzMDM+fbrG46Mn387Lnz/Ly58/y4A7etGlt47bfiq3/ff+uo+hAd40c7lk22/MZdzvBxi7YT+bzqbyZPm2v72nCx28XLnvt/2MWb+PYA/9t/1aWRkxXy8jdNpker81i6TdEWQnJlVrc3brTmycHOk751UChw7gyIpVADS6uRu9XptBr9dm0HHieBy8PHQd/AAc2HWY82dT+WjFCzzy/N18Pmel2Xaj7uvHh8um885Xz3AkKoH9u0z7DT2HdOG9b//Nu4uncevY/nw9T9/tptCfDIDM2BF+mHRjzmWXjxrSle9WbgMgfP9x3Fwd8fMxMLhvRzZuiyIjMxdjZi4bt0UxpG9HvWJXE7c7mk4Dw1BK0SQokPycfLLTM2u0axIUiItHzQ3t4V3RdBpkepMP6d2RkweOomlanecGOLY7inYDuqGUolHbZhTm5pNzSfac9EwK8wpo1LYZSinaDejGsd2HqrXRNI3D2/cT3LcregnxcOFMTgGJuYWUlGlsOJ1CX//qM2d7UzIpLC0DICotGx9H0w6gBthaWWFT/tXASpFeUKRbdoC2BhfO5RWQlF9IiaaxKSmFnpfMHhxIz6SwzJQ/1piNt71txbJ9aZnklZTqmrmqnPh47H18sPf2xqpBAzzCwsg4eLBaGzsvLxwbN4ZLbgPq4OuLva8vALYGAzaurpRkZ+uW/aLYnVF0GWzqu02DAsnPzScrrWbfbRoUiKtnzb7bslMrbMv/JgFBgWSm1HxuXdq6OZrht5ryt+8YSHZ2PqlmMuTlFfLd138w4dHBNZ4/8jbTzNvI28LYsjlKl9xZJ+Nx8PHBobx2fLqFkra/eu3Ye3nh3KRm7WTExOAeEoSNsxM2Tk64hwSRER2jS+6L2hpcSMyt7Lubk1K42efyfTfukr6738J9F+DA9mh6DA1FKUWLkEDycvIxpmXVaNciJBCDp+sVf1b4xv10G9i5rqLWEOLhwtmcAs7lmtb/b2dS6NPoCtv+9Gx8Lg7+NbC1Lt/2W1vRQOm/7TeeTMDJ1xtHH1P9N7wplOR91es/ed9BGvXqDoBfWBfSYg/X2C84tzsC/+76z5xHbIumz7CuKKVo3a4puTn5ZKRWrx07e1vadW0JQAObBjRr3Zj0ZNO2ydHJvqJdYX7RP/420UpZ6fZ1vbJoMqXUj0qpvUqpGKXUxPLHHlJKHVVK/aGU+kwp9Z/yx72VUiuVUhHlXz0tldvfz4OzSWkV3yeeT8ffzwN/P3fOnkuvfDwpHX8/d0tEJCvNiJuXoeJ7Ny83slKvfkfI9HxTdmtra+wc7cnL0ueIZnZaJi5Vsrt4Gsi+ZAewRhuvmm3OxJzAyeCCh79P3QauwtvBlgvlpzQAXMgvxNvB9rLtb2vmy86kDMA0GNqbYuSXW7rx6y3d2H3eSEJ2fp1nrsrL3pbkKm+8KflFeNnZXbb9iMa+7EnJ0CPaVSkyGrH1qNzpszUYKM746/ly4uMpKynBztv7Wsa7KplpmRi8K7cbBi8DmWYGQFcj4tfdtA0LulbRrkpKcia+fpV908fXQEpyzfyfzl/H2HH9sLev3j/S07LxKj/67eXtRkba5Q9GXUtFRiN2HpXr3c7dnUKj8aqeW5hhxM79kudmXN1zrxUve1tSqvbdgiK87C/fd4c39iX8Ouq7AMbULDx8KmvH3duAsRYD+LTz6aQmpRHUpdW1jHdFl277k/OuvO2/tZkvu86Xb/vTs9mbbGTdLd345ZZu7L6g/7a/IMOIfZX6t/dwp+CSGq7axsramgYODhTnVN8vSNqzl4bdQ+s+8CUyUjLx9K2sHU9vN9KvUDu52fns3RFDu9DKGlm/cjtT7nqTJZ+sYfzU2+s0r7A8Sw/NHtQ0rSsQCkxRSjUCZgLdgcFA2yptPwI+1DQtDLgT+K+5H6iUmqiUilRKRS5atKhOQitqHhnQNM3sEQOdJk1qMve6f+GIhrnc+h0RqfniNV7abMDq38Zt3UtQH/1mf8xEAMz/KQCGB3gT5OHMN+XniTd2tifQxZGRa8IZsSacUB83Ontd+SjntWY+v/nfYJC/N23cnFken2h2uUWYL9y/9COKjEZOfPEFzcePR1lZYBNptrT/et/b93skZ4+eoe/dA65BqKtnfqa4ev6jhxM5eyaVfgP1O0XpT/2tbfXfr7u6cLlZ+0H+3rR2c2bF9dR3uUzeWqzG8E376dq3I1bW+vVfs3/uy9TUsABvgtyrbPud7Al0dWTUmnBG/myZbb+5bWeN9/w/6drGE/FY29ni0rjRtc12Ff7Kpr+0pJR5L3/LsLt741tllm7onb2Y98OL3Pf4KP731e91lPT6oLDS7et6ZembIExRSv2r/P9NgP8Dtmialg6glPoeaF2+fBAQXKVDuiqlXDRNq3aOiqZpi4CLIx/tqdc3X/PQiefTaNywstM08vMg6UIGiUnp9O5RebS1UUMPtu3S72LSPT9vI/LXXabXbh1AZmrl0ZvM1Exc/+SUgarcvAxkpmbg5m2gtLSUwrwCHFzq7kLkvWu3cnC9KXvDVgFkV8menWbE+ZLT9Fy8DNXbpBqrncpXVlrKkV2HGP/hs3WW2Zzk/CJ8HSuPuvo62JGaX/NUhm4+bkwIbsKjm6MoLjNtufs18iQ6PZv8EtMpErvOZ9DO04X9qTVPAakrKQVF+FQ5Iu/tYEtaYc38XTzduL9lY57eHV2R/3pg6+5OUXrlLGyR0YiNwXCFZ1RXkp/PkfnzaXzbbbg0b14XEc3auXobe9aZ6r9JmwCMVY7MG1ONf6nvAhzbd4RNS39j0ntP0sC27jfz3y/dzk8rTfmD2wVw4Xxl30y+YMTbp3r+qIMJHI49y+1DX6WkpIyM9Bwem/AfFnw5GQ9PF1JTMvHydiM1JRN3T+c6zw9g626gML1yvRdmZGB3lbVj5+6O8Ujl9TOFGRkY2rS+wjOuvdSComqntHnbX77v3teiMc/suT767qZV29m2ZjcAgW2akJ5cWTsZKUYMXn/9WpjwjQcYO/WOa5bxaiTnVd/2+zjaVZuRuyjMx40JQU2Y9Mcl2/60bPLLT4/bmaT/tt/ew52CKvVfkJ6BncHtkjYGCtIzcPBwp6y0lJL8fGycKq9zTtodib+Osz/rV25n4+o9ALRo24S0C5W1k5aSiftlamfRO9/j19iLkff2Mbv85kGd+O9lriES/xwWG5oppfphGtT00DStI7AfOHKFp1iVt+1U/tXo0sGPXtZu2Md9d/YGoFvnlmRl53E+2ciGLQcZ1LsDBjcnDG5ODOrdgQ1bDv7JT7t2brqlN098/BxPfPwcQT3ac2BjBJqmcSYuAXsne7PX+lxO2+7tOPB7OAAx2w7SrGOrOp0B6jqyT8WNC1p170D0pnA0TSPxcDx2jvY1BkDOHm7YOtiTeDgeTdOI3hROq+7tK5YnHDiCZyMfXL30PQUxNj2bAGcH/J3saGClGBzgzdYqp0UCtDY48UJoS6ZtjyWjsLji8Qt5hXTxdsNagbVSdPF2IyFL39MgDmdm08jJAT8HOxooxYCG3uy8UD1/S1cnnmnXghmRcRiLii/zkyzDOTCQguRkClJTKSspIT0iAveOV3cdXllJCccWLMCrRw88Q/U9hePmW3szdeFzTF34HCE3t2ffBlPfPRWXgIOTg9lrfS4n8fhZVn60gnGvPoKzu0sdpq5095hefPvDv/n2h3/TZ0A7flltyh91MAFnZ4eKU9ouuvPenqzdNJsf189i0ddTCAj0ZsGXkwHo3a8da3+KAGDtTxH06d9Ol9/BtVkg+ReSyU8x1U5yeCSena6udtxDQsiIiaU4N5fi3FwyYmJxDwmp28CXuLTv9m/ozc7kmn13arsWzNx7/fTdAf/qVXHjgs6927NrfSSapnEiJgEHJ/s/vdbnUudPJ5OXk0eLkMC6CXwZsRnZNHF2wN/RtP6HNPFmm7ltf9eWPLuj+rb/vJltf7zO2363Zk3JvZBMXnn9J+2JxKdz9Rlan84dSNxuGqyej9iHZ1Cbiv0CrayMpIh9NLxJv23n0Dt78e7iaby7eBphfdqx9de9aJrG0ehTODrZ425mFm3Zp7+Ql1vAuKdvq/Z40pmUiv/v3xlHwyZedZ7fkuQaIMvOALkBGZqm5Sml2mI67e0zoK9Syh3IxnSq28UrYH8DJgNzAJRSnTRNO1AXwRbPf5LePYLwcnfh+J7/8NoHP2BjY1pV//32d37dtJ+h/TsRs20uefmFPPrspwBkZOby1rxVbP/5dQDe/Oh/ZGTqeyegi1qHBXM0IpYPH3wNG3tb7ph6X8Wyj594lyc+Nt3ee/3nP3Fo816KC4uZc/8sug7rwYD7h9NlaHdWzvmWDx98DQcXR+6ZPk637C1CgzkZGcOnE1/Fxs6WEU+NrVj2xZR3Km5pPfTxe1g7dwklRUU07xpM867BFe1it+7T9eYHF5Vq8O6+E8zr0w5rBavjL3AyK49HQwKIy8hh67l0nurYDIcG1rzdw3SG5/m8QqbtiGPj2VRCfdxYOrQLmmaaAdqWlP4nr3htlWkwL+Yk73YLwQr45WwyCTn5TGgVwJHMHHYmpzOpbSAODax5pUsbAC7kF/FS+R34PurejgAnRxwaWLGifyhzoo4TkarftRDK2prAMWM4MncuWlkZ3j174ujvz9mffsKpaVPcO3UiJyGBo598QmleHsZDh0hcvZoOs2eTHhlJ9tGjlOTkkLpzJwDNJ0zAqYm+dzNq2y2Yw+FxvDP+dWztbLn72TEVyz6c9C5TF5r67trPVnOgvO++cd/LhA3rzpAHhrP2s9UU5Rfy7WtfAmDwcWfCq4/olr9n72B2bo3jzhFvYG9vy8zXR1csu/+uOXz7w7+v+PxxDw3kxWcXs3rVHvwauvPm+/pse5S1NS3Hjibqw4/Qysrw69UTp0b+xP+4GpfApnh16khWfAIxHy+gJDePtIOHSPjpZ8JeewUbZycCRo1k3+tvAdD0lpHYOOt7B9AyDebHnuSdsBCslKnvnsrJZ3x5392VnM7ENoE4WFszq7Op7ybnFzFzn6nvzr2pHU2cHXGwtmJZ/1DeizpOpI59F6B99yCidsfx4n1vYmtnw4TplbU/+6H3ePlz04z+9wt+JnzjPooKivn3XbPpNfImbptgumX5no37CBvQWfeL2Es1mLPftO23UvBz+bZ/YkgAcek5bEtKZ0oH07b/rSrb/md3xLGpfNv/3ZAuaMDu8xls13nbb2VtTfD/jSZizny0sjIa97kZl8b+HP3fz7gFBuDbpSON+/Tk0KKv2PLvWdg4OdLp8Ycqnp9+5Dj2HgYcffS/bhKg881B7N8Vx1N3v4WtvQ2Pzajc7jw37n3eXTyNtGQjqxb/jn9TH6ZP+BCAoXf2ZOCt3Vn/ww6iIo9i3cAaJxcHHn9pzOVeSvxDKL3u7FXjhZWyA34EGmGa+fEGXsF0ytuzwDkgDkjXNG2GUsoL+BgIwjRw26pp2qQ/eRnNIeDGLOL800sBWHHyVwsnqZ17mpvejL48ut7CSWpnQuuhhK3YbukYtRZxTy/6r9th6Ri1tnlETx7YssXSMWrl6759Afjp1C8WTlI7tzU1fW6SsWidhZPUjsF2BAATt/9h2SC1tKhXPwb+cuP23Y3De7Lt/FpLx6i13n4j6fb9jbntD7+7FwBP795k4SS1M7e76XrFA2lrLJykdjp5joJaXbWmP8/WU3Tb+U87Ou+6XCcWmwHSNK0QGH7p40qpSE3TFimlGgCrMM38oGlaKnCvvimFEEIIIYQQ/ySWvgmCOa8opQYB9pgGPz9aOI8QQgghhBD/CNfztTl6ue4GQJqm6XvbLiGEEEIIIUS9cd0NgIQQQgghhBB1ozafLfdPI3NgQgghhBBCiHpDZoCEEEIIIYSoJ+QaIJkBEkIIIYQQQtQjMgASQgghhBBC1BtyCpwQQgghhBD1hJwCJzNAQgghhBBCiHpEZoCEEEIIIYSoJ2QGSGaAhBBCCCGEEPWIzAAJIYQQQghRb8j8h9I0zdIZ6tI/+pcTQgghhBDXDWXpAFfDL/gF3faPz8e+dV2uE5kBEkIIIYQQop6Qa4DqwQBoxclfLR2hVu5pPgwAh4AxFk5SO/mnlwKwMO43CyepnUlBQ1h0eL2lY9TaxLZDiUhZa+kYtRbmPZLnwjdZOkatvNttAMANWz8T2w4F4JPYG7PvPh48BIDRm7daOEntLOvfh4nb/7B0jFpb1KsfH8XcmLUD8FTIEMZv3WLpGLXyVZ++AAz8ZYeFk9TOxuE9ATiWucbCSWqnldsoS0cQf8E/fgAkhBBCCCGEMJEZILkKSgghhBBCCFGPyAyQEEIIIYQQ9YSS+Q9ZA0IIIYQQQoj6Q2aAhBBCCCGEqCfkGiCZARJCCCGEEELUIzIDJIQQQgghRD2h1HX52aS6khkgIYQQQgghRL0hAyAhhBBCCCFEvSGnwAkhhBBCCFFPyE0QZAZICCGEEEIIUY/IDJAQQgghhBD1hHwQqswACSGEEEIIIeoRmQG6DE3TWLfwfxyNiMXGzoY7po3Fv2WTGu02fLWGAxsjKMjJY+aqORWPlxSVsPL9bzl37AyOrk7c88I43H09dcm+cM6jDB/YmZS0LEIHP2e2zfuzxzG0fyfy8ouYOG0BB6ITABh7Vx+mP3k7AG/P/5ElP2zVJXNVmqbxx39XEr83Bhs7W4ZMuR/fFjXX/YXjp1k/71tKiopp1jWEfg/fiVKKXUvXEbVhJ46uzgD0vP8WmoWG6Jp/82crid8bSwM7W4Y9Nfay+X+dt4SSwmKadQ2m/yN3Vtyact+aLRxYuw0rayuahYbQd/xtuub/5qNVHNgVh529LRNfHEOzNo2rtSksKGLezMUkJ6ZhZaXo3DOE0Y+NAqC4qISFr39H/JEzuLg6MfnVB/Bu6KFb/guHYoj6ZgWUaQT060nrW4ZWW15aXMy+TxeTGX8aG2cnwiY/jKO3J3kpaWx8fjbODX0B8GjZjI4T7tMt90X/hPrZ8vlKEvbG0MDOliFP3o+PufwnTrOhvP8Gdg2h70N3Vrs1694fN7J98Y9MXPwWDuV9ua5lxURzdsUytLIyPHv2xm/Y8GrLc44d5eyK5eQnniXwoYm4d+1abXlpfj5xr8zCrVNnmozRv3bSo6I5vnQFmlZGw969CBgxrNpy45GjnFi2gpyziQQ/+jDeoZX5z+/Yxek16wAIGDUCv549dM0OptrZ/vlKTu0z1c7AyffjbaZ2kk+cZtN8U+007RJCr/La2fPdGuIjolBK4eDmwsAn78fJw023/BnR0SQsW45WVoZv7140Gl69frKOHiVh+XJyzybSeuIjeJbXT+7pM5xcsoTS/HyUlRWNRo7AKyxMt9wXhXkZeCKoOVYK1p29wLKTidWW3xXoz4gmvpSWaRiLipkTdZzkgkIA3goNJtjgQnRGFjP2xumeXdM0Fr3/I5E7Te9bT88aTcu21d+3CgqKePuFrzl/NhUrKyu69Q5m/GTT+1by+Qw+nL2U3Ox8yso0xj0xkrCeQbr/HnqRa4BkBuiyjkXEknYuhac/f4nbpozm5/98b7Zd25vaMemjZ2o8vve3XTg4OzD1i5n0uL0fv33xc11HrvDN91u47YG3L7t8aP9OtAj0o12fqUye/hnz3ngIAHc3J2Y8fQd9bp1J71tnMuPpOzC4OekVu0LC3liMSclMWDCLQY+PZtPC5Wbbbfx0OYMeH8OEBbMwJiWTsC+2YlmXW/tz/9zp3D93uq6DH4D4vbFkJKXw4MKZDH7iXn5fsMJsu98XrmDw46N5cOFMMpJSSNhnetM4fegoJ/ZE8cC85xn/nxcJu32AnvE5uDuO82dSeX/Zizz077v56r0fzLYbOaYfc76bzhtfTuNoVDwHd5ny/7FmD04uDnywfAbD7u3LsgVrdMuulZVxaPEyevx7MgPemUXirgiyEpOqtTm9ZSe2To4Mev9VWgwbQMzyVRXLnHy86P/GDPq/McMigx+48esnYV8sxnPJjPtkFgMfG82mT833380LlzPwsTGM+2QWxnPJnKrSf7NTMzh98DAu3u56xUYrK+PM0u9oMfkpgtgM4fAAACAASURBVF5+lYyIcPLPnavWxsbdg6bjJuAe1s3sz0ha/RPOrVvrEbcGrayMY0uW0n7qk4S99grJeyLIvSS/vacHbR4cj+9N1fMX5+RyavUaOs+YTueXpnNq9RqKc3P1jA/A6X2xZCYlM/bjWfSbNJoti8zXztZPl9PvsTGM/XgWmUnJnN5vqp3Otw9k9IcvcO8H0wkMDSFixS+6ZdfKyoj/7juCnppCp1dnkxoeQd4l69/Ww4MWEybg1a36+reytaXlgxPo9Opsgp5+ioTlyynJy9MtO5h2BqeENOeFyBge3LafAQ29aersUK3N8axcHttxkEd2HGDrhTQmtg2sWLYiPpG3Dx3VNXNVkTsPc+5MKotWvsDkF+7mk3dWmm13x9h+LPx+Oh99+wyxBxOI3Gnabi7/4nd6D+zEvG+n8dzr97PgXfPPF/8cMgC6jLjd0XQaGIZSiiZBgeTn5JOdnlmjXZOgQFzMHGE6vCuaToNMG7mQ3h05eeAomqbVeW6AHeGHSTfmXHb5qCFd+W7lNgDC9x/HzdURPx8Dg/t2ZOO2KDIyczFm5rJxWxRD+nbUJXNVJ8KjCOrXDaUUDds0ozA3n5xL1n1OeiZFeQX4t22GUoqgft04sSdK96zmnAiPIri/Kb//FfIXVskf3L8bx/ccAuDgr9vpdudgGtjYAOBocNE1/95t0fQaFopSipbtAsnNyScjNataGzt7W4K7tAKggU0DAls3Jj3FCMC+7dH0Hm46etmtXwdi9h7TrfYzTiTg5OuNk483Vg0a0Kh7KOf3HqzWJmnfQZr06g6Af7cupMYc1i3f1bjR6+dkeBRB/av339xL8uemZ1KUX0DDi/23fzdOhFf2361f/I9eD9wG6PdhfXkJ8dj5eGPnbaod97AwMg8dqNbGzssLh8aNzX6IYN6pUxRnZ+ESFKxX5GqyTsbj4OODQ3l+n26hpO2vXvv2Xl44N2kMl+TPiInBPSQIG2cnbJyccA8JIiM6Rs/4AMSHR9GmfNvv16YZRVeoHb82ptpp068b8eXbflvHyh324oIiXT/sMSc+HntvH+zL179XWBgZB2qufycz9ePg54uDr2nm2dZgwMbFleLsbN2yA7Q1uJCYW0BSfiElmsbmpBRu9qk+c38gPZPCsjIA4ozZeNvbVizbn5ZJXkmprpmr2rM1mgEjuqKUom37puRm55N+yfuWvb0tHUJbAmBj04AWbRuTmmyqL6UgL7cAgNycAjy8XPX9BXSmlJVuX9criydTSjkppdYqpQ4qpaKVUvcqpboqpbYopfYqpdYrpRoqpRoopSKUUv3Kn/eWUuqNusqVlWbEzctQ8b2blxtZqTUHQFd+vunopbW1NXaO9uRl6X9EzRx/Pw/OJqVVfJ94Ph1/Pw/8/dw5ey698vGkdPz99DsCe1FOuhEXr8rXdfY0mN0BdPY0XNLGWPH9wbVb+eapt/ht/hIKcvQ9kpaTlolLldpx8TKQk5ZZs02V/C6elW0yzqVwNvYES559n+UvfsT5Y6f0CV4uIzULT5/KbB4+BjKuUPu52fns3xFDSFfTke+MlEw8yp9v3cAaRyd7cjL1qf2CDCMOHpW14+DhTkGGsXqbdCMOnqY2VtbWNHB0oCjHlC8vJY0/XnqD7a9/QNqRY7pkvtSNXj85aUacPWvRf9NMf6eT4VE4e7jh3az66St1rSjDiK175Q6frcGd4ktq53K0sjISf1hBozvuqqt4f6rIaMSuSu3bubtTaLy6/IUZRuzcL3nuVf7u11JuuhHnKtt+J0+D2QFQ1doxtanMunvJzyx+ZCbHtkbSbfSIug9dzrT+q9SPu4FCY8Zf/jnZ8fFoJSXYe3tfy3h/ysvelpSCoorvUwqK8LK3u2z74Y19CU/5679fXUlLzsTLt7IuPH3cSEu+/PtWTnY+4dti6BRmOpB33yND2fzrXsaNepVXpv6XSc/+q84zC8uy+AAIGAac0zSto6Zp7YBfgfnAXZqmdQW+AN7QNK0EGA8sUEoNLn/e7Et/mFJqolIqUikVuWjRotqnMndA+C8cTTJ3QFnPo1FXoswcVdU0zWw+ixwYN/OiNZKZDWZq1WF4LyYsfJn7P3weJ3dXtn65ykzbumNuNqHmqjVXIKZ/ykrLKMzJ4745z9Bn/O38/O6Xus5QmM1/mbalJaV8/Mo3DL27Nz6NPC/7/L/Sd/6O2r62AuwMrgyZ+wb9Xp9Bu7F3EvnJlxTn51/7kH/iRq+fK2WraGH2d1QUFxYR/sN6uo8ZWUfZruTPc19O6pY/cG3XHlsP/a51q+Fv/Ykt12erpTDbf6+mTWWj7mNvYdxnr9GqTyhRv+h4DavZ7eZfW4dFRiPHP/+CFuPHo6wsv3t2ue3GIH9vWrs5syI+0exyS/gru2ylJaXMeelbbr23N37l71tb1u9n4KgwFq+ZxSsfPsz7ryylrHy2659IYaXb1/XqergJQhTwnlLqHWANkAG0AzaU75BbA0kAmqbFKKW+AX4GemiaVnTpD9M0bRFwceSjrTj561UH2fPzNiJ/3QVAo9YBZKZWHlXKTM3E1fPqp0TdvAxkpmbg5m2gtLSUwrwCHFwcr/r5dSnxfBqNG1bekKGRnwdJFzJITEqnd4/Ki/4aNfRg2y59LmY8sG4r0b/tBMC3VQDZqZVHlnLSjDUuZK16xPhiG+fyNk6Gyr9Tu8E389Mbn9ZldAD2r91K1AZT7fi1DCC7Su1kp5rPn10lf3aV/C6ebrTq0dF0ClHrpigrRX5WDo5udXcq04aV29n8824Amgc1IS25Mlt6shGDl/kLiT9/93v8mngx7J6+FY95+BhITzbi6WOgtKSUvNwCnF31qX0HD3fy0ytrJz89A3tD9ez2Hgby0zJw8HCnrLSUkrx8bJydUEphXX7amKFZU5x8vMhJSsa9edM6z32j18/BdVuJ3lDef1sGkJNWvf86u1fP72Km/zp5uJF5PpWsC2ksmfp2xePfTXuX0e8+i5N73Z6SYuvuTlFG5Qx4kTEDG4PhCs+olHvyBDnHj5O65Q9KCwvRSkuwsrej0b/urKu4Ndi6GyisUvuFGRnYXWV+O3d3jEcqr98ozMjA0Eafa5miftlKbHnt+LQMIKfKtj83zYiT+5W3/ebaALTuHcraNxbSbbQ+g2lbd3cK06vUT4YR26tc/wAl+fkcnj+fgNtvw6VF87qIeEWpBUXVTmnztrclrbDGLhZdPN24r0VjntkTTXGZZU8dXvP9dtb/uAeAVsFNSL1QWRdpyZl4eJt/35r/1vf4N/HitjF9Kh7bsHoPs+c9AkBQh0CKCovJMuZi8ND3FGKhH4sPgDRNO6qU6gqMAN4CNgAxmqZd7hY07QEj4Huts9x0S29uuqU3AEfCY9jz8zba9+3C2cOnsHeyN3utz+W07d6OA7+HExDUjJhtB2nWsdV1MwO0dsM+Jo0bworVO+nWuSVZ2XmcTzayYctBZj93b8WNDwb17sCst5fpkqnTiD50GmHaGJ2MjObguq206d2V80cTsHWyr9i5u8jZww1bB3uSjsTj1zqQuD/C6TTCtBOek55Z0f7EnoN4BjSs8/ydR/ah88iL+WPYv3YrbXt3IeloAnZXyH/uSDwNWwcSuzm84vktb+rA6UNHadK+FemJyZQWl9b5XbAG39mLwXf2AmD/zlg2rNxOj0GdORFzCkdne9zNnA/9/aJ15Ofm8/D0e6o93qVnCNt+iaBVu0DC/zhEcJeWutW+oXlTcs8nk5ucioOHgcTdkXR9/MFqbfw6d+DM9t14tGrOufB9eAW3QSlFYVY2ts5OKCsrcpNTyL2QjJOPly65b/T66TiiDx3L+298ef9t3cvUf+0c7WsM4Jw83LCp2n83h9NxZF+8mvozcfFbFe2+mPgyY977ty53gXNsGkhhcjKFqSnYGNzJiIgg8KGHr+q5gQ89UvH/tJ07yDt1StfBD4Brs0DyLySTn5KKnbuB5PBIgiY+dFXPdQ8JIf5/P1bc+CAjJpZmd+hzClD74X1oP9xUOwmR0UT/spWWvbpy4WgCtperHXt7zh+Jx7d1IEf+CKd9+bbfeC4Zg78PAPERURgaXfPdhMtyDgykIDmZgpRUbN0NpEZE0Orhq6ufspISjnyyAO8ePfAMDa3jpOYdzsymkZMDfg52pBYU0b+hN28cPFKtTUtXJ6a2a8H0iFiMRcUWyVnVqLt7Mepu0/tWxPZY1ny/gz5DOnMk+jSOzvZmr+P5ZsEv5OUUMGVG9fctbz93DkYcY9CobpyJv0BxUQlu7vrcfdIiruNrc/Ri8QGQUsofSNc07VulVA4wEfBWSvXQNG2XUsoGaF0++3MH4An0AdYopbppmlYnJyq3DgvmaEQsHz74Gjb2ttwxtfKOUB8/8S5PfGy6vfT6z3/i0Oa9FBcWM+f+WXQd1oMB9w+ny9DurJzzLR8++BoOLo7cM31cXcQ0a/H8J+ndIwgvdxeO7/kPr33wAzY2pj/1f7/9nV837Wdo/07EbJtLXn4hjz5rmiHJyMzlrXmr2P7z6wC8+dH/yNDp2o2qmnUNIWFvLF9OepUGdjYMmXJ/xbJvn36b++dOB2DApHv5bd63lBQWE9g1iMCupouPty3+iZT4syilcPXxYOBjo3XOH8zJyBg+n/QqNna2DH1ybMWyr59+hwfmPg/AoEn3mG5jXFREsy7BNCvP325Qd9bP/46vnnwL6wbWDH/6fl0Hz516BHFwVxzT7n0TW3sbJr44pmLZi+Pf482vniUt2chPX/+Of1MfXnrwA8A0iOp/S3f6jrqJha99xzP3voGzqyOTX3lAt+xW1tZ0eGA0u+bMRysrI6DPzbg29idu5c8YmgXQsEtHmvbtyb6FX/H7tFnYODsS+oRpJzHtyDEOr1yDsrJCWVnRcfx92DrrfxfEG71+Asv77+LHTP138JOV/XfJ1LcZ+2F5/3303orbYDftEkRgF8vcPOAiZW1N43vv48S8uWhlGp4398TBvxFJq3/CsWlT3Dp2IjchnviFn1Cal0dm1CHOr/mJoJdftWjui5S1NS3Hjibqw4/Qysrw69UTp0b+xP+4GpfApnh16khWfAIxHy+gJDePtIOHSPjpZ8JeewUbZycCRo1k3+umwWfTW0ZiY4Hab9o1hNP7YlnyuKl2BkyurJ3lz7zNvR+Yaqfvo/dW3AY7oEsQAeW1s/vb1RgTk8FK4eLtQd9H79Utu7K2ptl9Y4ibOxdNK8OnZ08cG/lz+qefcG7aFI9OnciJT+DIJ59QkpdHxqFDnPlpNZ1enU1aZCTZx45SkpND8g7TbFjLCRNwCqh5C/C6UqbB/NiTvBMWgpWCX84mcyonn/GtAjiSmcOu5HQmtgnEwdqaWZ3bAJCcX8TM8rtPzr2pHU2cHXGwtmJZ/1DeizpOZKp+15GF9gwicmccj9zxFnb2Njw9s/J9/8mx7zN/yTRSLxhZ/uXvNA704an/+xCAUXf3ZOjt3XnoqVuY/+b3/PjdVpRSPD1r9HVz0FrUDWXpux8ppYYCc4AyoBh4DCgB5gFumAZpc4FVwE5goKZpZ5RSU4CumqZdaWTxl06Bu57c09z0+Q0OAWP+pOX1Kf/0UgAWxv1m4SS1MyloCIsOr7d0jFqb2HYoESlrLR2j1sK8R/Jc+CZLx6iVd7uZbjt9o9bPxLamz036JPbG7LuPBw8BYPRm/T/D7FpY1r8PE7f/YekYtbaoVz8+irkxawfgqZAhjN+6xdIxauWrPqaZsIG/7LBwktrZOLwnAMcy9fvohGupldso0PPWlX9Di65zddv5P7H36etynVh8BkjTtPWAuT2FPmYeqzgpWdO0eXUWSgghhBBCCPGPZPEBkBBCCCGEEEIfcnrf9XEbbCGEEEIIIYTQhQyAhBBCCCGEELpTSg1TSh1RSh1XSk03szxAKbVZKbVfKXVIKXVNPuFYToETQgghhBCinrhePqBUKWUNfAwMBs4CEUqp1ZqmxVZp9hKwQtO0BUqpYGAdEPh3X/v6WANCCCGEEEKI+qQbcFzTtJOaphUBy4DbLmmjARc/1MkNOHctXlhmgIQQQgghhKgnlI4fhKqUmojpMz4vWqRp2qLy/zcCzlRZdha46ZIf8Qrwm1LqScAJGHQtcskASAghhBBCCHHNlQ92Fl1msbnb0V36GUVjgK80TXtfKdUD+EYp1U7TtLK/k0sGQEIIIYQQQtQX189tsM8CTap835iap7g9BAwD0DRtl1LKHvACkv/OC8s1QEIIIYQQQgi9RQCtlFLNlFK2wGhg9SVtTgMDAZRSQYA9kPJ3X1hmgIQQQgghhKgvrpPpD03TSpRSk4H1gDXwhaZpMUqpV4FITdNWA9OAz5RSUzGdHjde07RLT5P7y2QAJIQQQgghhNCdpmnrMN3auupjs6r8Pxboea1fVwZAQgghhBBC1BfXzzVAFnOdTIIJIYQQQgghRN1T1+A0uuvZP/qXE0IIIYQQ140bYmql9c0Ldds/Prpz0nW5TmQGSAghhBBCCFFv/OOvAfry6HpLR6iVCa2HArAw7jcLJ6mdSUFDAHAIGGPhJLWTf3opz4VvsnSMWnu32wAWH7sxax9gXKuhTNz+h6Vj1MqiXv0AeD5io2WD1NI7YQMB6Lp0m4WT1M7eMb0BMBat+5OW1yeD7Qh+T7wxswMMajSCB7ZssXSMWvu6b192Ja+1dIxa6eEzEoAtSTdm/fRtOAKApnNuzPfeU/8eYOkIV0+mP2QVCCGEEEIIIeqPf/wMkBBCCCGEEMJEk7vAyQyQEEIIIYQQov6QAZAQQgghhBCi3pBT4IQQQgghhKgv5Aw4mQESQgghhBBC1B8yAySEEEIIIUR9YSVTQDIDJIQQQgghhKg3ZAZICCGEEEKI+kJugy0zQEIIIYQQQoj6Q2aAhBBCCCGEqC9kAkhmgIQQQgghhBD1h8wACSGEEEIIUV/IXeBkAHQ5mqbx+6KVnNgbi42dLSOfGotfyyY12p0/fpq1c5dQXFRMi67BDJp4J0opfnznS9ITkwEoyM3H3smBB+c9r2v+P/67kvi9MdjY2TJkyv34tqiZ/8Lx06yf9y0lRcU06xpCv4dN+XctXUfUhp04ujoD0PP+W2gWGqJL9oVzHmX4wM6kpGUROvg5s23enz2Oof07kZdfxMRpCzgQnQDA2Lv6MP3J2wF4e/6PLPlhqy6Zq7pwKIaob1ZAmUZAv560vmVoteWlxcXs+3QxmfGnsXF2Imzywzh6e5KXksbG52fj3NAXAI+Wzeg44T7d82uaxoZFKzkRGUsDO1tuedp87ScdP82aD5dQUlRMi9BgBpfX/oWTZ/nl4+WUFJVgZW3FsMfuwb9NU93yp0dFc3zpCjStjIa9exEwYli15WXFxRz+/EuyT53GxsmJ4EmPYO/lRVlJCUe//pachFOgrGg55h4MbdvolvuiCwdjiPrme7Qyjab9bqb1rWbqZ+FijPFnsHVxInTyQzh5e1Ysz0tNZ+Pzr9H2jhG0GjlY7/j0aOjOs12aY60UP544z1dxZ6stv7OlH/e08qdU08gvKeX18OPEZ+UBMCG4Mbc196NU03hv7wl2nTfqml3TND54exU7t8Vhb2/DzNfH0Da4Zu0/NelTUlOyKC0tpVOX5vx7xl1YW1tx9Egi77z6Pfl5RTRs5M7st/8PZ2d7XfN//59VxOyJw9behv97bgwBrWvmX/35Wvb8Fkledh4frnun4vFdv4bz46ercfNyA6Dv7b3pObK7bvmN0dGcWr4crawMn1698B8+vNryrKNHObV8OXmJibR85BE8u3YFIPfMGRKWLKE0Px+srGg0YgSeYWG65b5I0zSWfLSKQ7vjsLWz5eEXxxDYpnGNdj8sWsfO9ZHkZufx6W9vVzyediGDz974jrycAspKy7h70kg69gjWNf/y+auI2m2qn/HTx9DUTP2s+u9adq831c/8X9+ptixy835+/mo9KGjSohEPz/w/veLTN9CDlwe2wloplh1KYkH4KbPtRrT2ZsFt7Rn1dQRRF7Lp6OfCW0PbAqYzw+bujGf9sVTdcgvLkAHQZZzcG0vGuRQe/XQm544ksH7BCsa9P61Gu/WfrGDY5NH4twnk+1cWcnJvHC1Cg7n9+QkVbTZ+vgo7R/3eBAES9sZiTEpmwoJZnD+awKaFyxkz59ka7TZ+upxBj4+hYZtAfnxtAQn7YmnW1TTQ6XJrf0JvH6hrboBvvt/CwsXr+e+Hj5tdPrR/J1oE+tGuz1S6dW7JvDceos9tM3F3c2LG03fQc+QMNGDn2jdYu2Evxsxc3bJrZWUcWryMm5+fgoOHO1tmvY1flw64NmpY0eb0lp3YOjky6P1XObsrgpjlqwib/DAATj5e9H9jhm55zTkRGUv6uRQmLTLV/q+frGD8BzVr/9ePVzB88mgatQ1keZXa3/TlT/QeM5wWocEcj4hh05c/cf/bU3TJrpWVcWzJUjpMexo7d3f2vfYWnp064OTvX9EmadsOGjg6cdNbr5O8J4KTP/yP4EkTSdq6DYDQV1+mKCuLqLnz6fLSCygr/c4U1srKOLh4OT2nT8HBw8Afs97Br2v1+jn1x05snBwZ/MFszu6KJHbZKsKefLhiedSSH/DtqN9OU1VWCqZ3bcHjm6O5kF/IN0M6sSUxvWKAA/BrQgorj58HoE8jD57p0own/4ihmasjQwK8uXvdXrwdbFnQvz3/WhtJmaZf/p3b4jhzKoUf1r5I9KFTvPv6D3zx3dQa7d54bxzOzvZomsb0Z75i428HGDK8C2++vJwp026lS1hLVq/aw7dfbmLSkyN0yx+zJ46UxBRe+eZFEuJOsWzuDzz3Sc387XuE0Pf2Xrzyf2/WWNalX2fufepOPeJWo5WVkfDdd7SdOhVbd3di3nwTQ8eOOFbpu3YeHrSYMIGk336r9lwrW1taTJiAva8vRUYj0a+/jltICA0cHXX9HQ7tjuPC2VTeWfoiJ2JP8fX7PzBr0dM12nXqGcygO3rx/H3V1//qxRvo1r8TA/7Vk8T483zw3Ge8/71+fTl6TxwXzqbw+pIXiY89xZIPf+DFBTXrp2OPEPr/qxczx1bPf+FsCr8s2chz/5mCk4sjWRnZekXHSsFrg9swdsV+zmcXsvr/Qvn9RArH0vKqtXOysWZ8lybsO5dZ8diR1Fxu+TqSUk3Dx8mWX8Z14/fjOyjVdNz46E3uAifXAF3Osd1RtBvQDaUUjdo2ozA3n5z0zGptctIzKcwroFHbZiilaDegG8d2H6rWRtM0Dm/fT3DfrnrG50R4FEH9TPkbtrl8/qK8AvzL8wf168aJPVG65jRnR/hh0o05l10+akhXvltp2lkN338cN1dH/HwMDO7bkY3bosjIzMWYmcvGbVEM6dtRr9gAZJxIwMnXGycfb6waNKBR91DO7z1YrU3SvoM06WU6qurfrQupMYfRrqMN7dE9UbSvUvsFl6v9/AIaB5lqp/2AbhypqH1FYV4BAIV5BTh7uumWPetkPA4+Pjh4m9a/T7dQ0vZXX/9pBw7ie7Np/XuHdiEjzrT+884l4R5kOgpo6+pKAwcHshPMH0GsKxknEnD29cbJxwurBg1o3L1rjfo5v+8QAb0v1k9nUmKOVNTPucgDOHl74VJlwKSnEA8XzuQUkJhbQEmZxm+nU+jX2KNam9yS0or/OzSw5mLp92vswW+nUygu0ziXW8iZnAJCPFz0jM/WzdEMvzXMVNMdA8nOzic1JbNGu4uzOqUlZZQUl6DKdyZOJSTTObQFADf1aM3m3w/VeG5dOrQzmpsGm/I3Cw4kPyefzLSa+ZsFB+KmY7+8Gjnx8dj7+GBf3nc9wsLIOFi99u28vHBs3LjGzpuDry/2vqaZc1uDARtXV0qy9dv5vmj/9mh6DgtFKUXLkEDycvIxpmbVaNcyJBCDl2uNx5WC/PJtZ35uAe5e+v6NDuyIpsdQU/00DzHVj9FM/TQPCcRgpn62rdlFv9t74eRiGni6uuvXfzs1dCUhI48zmQUUl2n8fDiZwS29a7Sb1qs5C8NPUVhSVvFYQUlZxWDHroEV18+7sahL1+0MkFLqAeBZQAMOAaVAARAC+ALPaJq2pq5ePzstExcvQ8X3Lp4GstMycfZwu3wbL1Obqs7EnMDJ4IKHv09dRTUrJ92Ii5d7xffOngZy0qvnz0nPxNnTcEmbylNODq7dStzmcHxbBtBnwr+wd9b3aNrl+Pt5cDYpreL7xPPp+Pt54O/nztlz6ZWPJ6Xj7+du7kfUmYIMIw4ela/p4OFOxon46m3SjTh4mtpYWVvTwNGBohzTLFVeShp/vPQGDewdCLr7FjzbtNIvfLmctExcr6L2XT2rt8kpr/3BE+9g2awFbPziR7QyjXHv1TyCWFeKjEbsqqx/O3d3suKrr//CDCP2HqadcmVtTQMHB0pycnFq0pjU/Qfx6RZGQXoG2adOU5ieAc2b6ZY//5L6sfdwJ+NEwmXbVK0faxsbjq3ZQM/pT3Js7e+6Za7Kx9GOC3mFFd9fyCuinWfNnaC7WzXk/jaNaGBlxaRNpkGCt4MdUWlZVZ5biI+jHaTptyObkpyJr19lXfv4GkhJzsTLu+bO3pRHFxIbdZoevYIYMNh0oKVFy4Zs3RxN3wHt2bj+IMk6n8KXmZqJwacyv8HbgDE18y8Ndg5sO8jxqBP4NPbmrsdvx91Hn21okdGIrUflYNnWYCD3kr57NXLi4ykrKcHOu+bOb13LSMnCo8r6d/c2kJGaaXawY87tE4bx3rSF/L5yO4X5RTw3d1JdRTXLmJKJu3f1/MaUTLODHXMunEkB4J3JH1FWqnHL+KG0uymoTrJeys/ZjqTsym1PUnYhnRtWX+8hPs74u9qx6WQaE8MCqi3r1NCVOcPa0sjVnqnrYv/Zsz8gd4HjOp0BUkqFADOAAZqmdQSeKl8UCPQFRgILlVI1zitTSk1USkUqpSIXLVr0tJcdBgAAIABJREFUN1LULP4aM4bmOsglbeK27iWoj76zP4DZbDXq3WwHN7XqMLwXExa+zP0fPo+Tuytbv1x1zSPWljLTczVNqzgKW/1xPRJVz1HDVUw1K8DO4MqQuW/Q7/UZtBt7J5GffElxfv61D/knzP8OV9Gm3L512xn08L948v/Zu+/wqKqtgcO/nZA6KZMKBEijBwg9gAQQFEFs9+qnggXhqlixgyiKIla8VxFsYMWGIBZAbKBIJ4UACQmElggJIX3S+5zvjxnSgRDJTDDrfZ48MHP2ZNacrLNPWfvs+fQFLr/736x/66sLHOFZNOnv3fjn6xg+EgdPD3YveJmjX6/CvVtXlK2Fu8imxH+GP8/B736k28RxtHO07HDb+nHU19hH+uZwGtf9GM2SfUnc1df/zK+18Abc+Ps1vv0uXnov6zfNp7yikuiIwwA888JkVn+9jak3/Y/i4lLa2dm2YLQNNRZ/Y/3imfQb0YcXvprH3A9n02tQDz571ZLbbvP6ztrKDQaOfvwxwdOmWXTo6mmNr/+mv37XxhhGXhnGm989x2Ov382yBV9hNBrP/cILRGv0uKfpH8BYZSQjJYvHFz3I3fNu57PXV1JcYPl92Gm1P48Cnh3bnRc3HWm07d60fMZ/Esm1n0dz/7BAHCzd9wuLa60VoHHAak3TsgA0Tcsxb4SrNE0zAoeVUseAXsDe2i/UNG0ZcPrMR/vk0K9NftPd67ew79edAHTs7k9BVs3Vu4JsQ50r4GCu+NRuk2XAtVYbY1UViTtjmfZmw3tvWsLen7aw/7cdALTv7k9BVm71ssJsA7p68bt46SnMNtRpc/oz6vQ1V076jr+ENS8tbcnQz0vqqWw6d6y56btTB0/S0nNJTcth1Iiaq02dOnqydecBi8bm5OlBSU7Nei/JycVRX3e9O3rqKcnOxcnTA2NVFZXFJdi56FBKYWtnB4A+KACdrzeFaRl4BLf8BALRP25hrzn3/br7k18v913r5Y6bt5787Hrbh/kqYdzvkYyfYbqHoHf4QH5avKKlw69m76E3VW3MynJzcdDr67Rx8PCgNCcHB08PtKoqKktKaKczrf9uk2+qbrfn5ddwam/Zyq2Tp75O/pTm5OLk4d5oGyevuvmTeySZ1Mg97P/6eyqKS6rzKfiKSy0Wf3pxGe2dHaoft3e2J6uk7Iztf/0rk6eGdAMgo6SMDnVe60BmSXnLBWv2zYptrPnWlPshff1Jr1W1yUg34ON75qv3Dg52jL60D1s27WfYJT0JDG7PkmX3AXA8OYPtW1q+/9n8wza2rzfFH9DTH0NGTfyGTAPuXk2rPgC4uOuq/z/yqhH88EGLDbJowN7Dg/Kcmgp+ucGAXb1t92wqS0pIXLKEztddh2twcEuE2KiN321j87pdAAT16kJOrfWfm2locvUEYMv6CB7/7wwAuvUNpKK8gsK8ohYdSrbp+21s/dGUP4G9/MnNrBu/exOrVwAePu4EhwTSrp0t3h296ODvS0ZqJoG9/M/94r/pVGEZHV1r+o+Org6kF9b0Hy72tvT01vH15IEA+Ojs+ej6UO78Lpa49Joq85GcYkoqqujhravz/D+OzALXOitAmE7WG7sUV/+5C3p5cPBVo/nP4if5z+In6T48lP1/RKJpGqkHk3BwdmxwAuTi6Y69kyOpB5PQNI39f0TSfXi/6uXJexPx6uSLm7dlhhAMmDSa2xbN4bZFc+g6LJQDf5riT0tMwl535vjTEk3xH/gzkq5hpvhr3/NxNGIfXv7WuaegMes3xHDLDaMACBvYjfyCYk5lGNiweR+XjwpF765D767j8lGhbNi87xy/7cLSBwdQdCqDoowsjJWVpO6KpsOg0DptOgwM5cQ20w7zZGQM3iE9UUpRll+AZr7aV5SRSVF6Bjpfb4vEPeTq0dy15EnuWvIkPUaEEteE3Heolftxf0TSY1i/6mXH40xX2ZL3HcLTz3JDUdyCAilJz6Ak07T+MyKj8RpQ9z4wrwGhpO8wrf/M6Bg8evVCKUVVWTlVZaaD9Zz4BJSNTZ3JEyxBHxxAYa38Sdm1u2H+DArl+NbT+bOnOn9GzXucCYteZMKiF+k6YSw9rp1g0ZMfgIScArq4OuKnc6CdjeIKfx82p+TUadOl1qxo4X6eHDdfId6cksMV/j7Y2Sj8dA50cXUkPqflD0BunBLOF6tn8cXqWYwe15ef10aZcnpfMi4uTg2GvxUXl1XfF1RZWcWOrQcIDDKdKOeYh+sZjUY+XraBf990SYvHP+Zf4Tz9wSye/mAW/cP7ErHBFH9SQjJOOqfzGv5W+36h2B376eDfviVCbpRLYCClGRmUZplyPycqCo/+TbuH01hZyeH33sN7xAi8hgxp4Ujruvz6cBZ88gQLPnmCQaP6sf2XaDRN40h8Mk4ujk0e/gbg1d6DhN2mauLJ5HQqyitx1bu0VOgAjP13OPM+msW8j2YxILwvO3815c+xeFP+nM8J3IDwfiTuNcVfYCgk/UQm3rUuVrakfWkFBHk408XdETsbxTW9fNlwpGYmt4LyKga+s43wZTsJX7aTPSfzq09+urg7YmuudHVycyTY05mU/FKLxC2sp7VWgH4HvldKvalpWrZS6vTA4BuVUsuBICAYSGypALoOCeFYdDxLZ7yAnYM9kx6+tXrZxw+9Vj2l9YT7b2L9oi+pLC8neHAIwYNrZmxJ2BJj8ckPTgsa3Ifk3Ql8cu8LtHOw44qHbqte9sUjr3LbojkAjLv3Zn5b/AWVZRUEDu5NoDn+rcvXkJmUglIKN19PLrtvssViX75kJqNG9Mbbw5UjEW+z4I3V2NmZUvXDLzbyyx97mDB2APFbF1FcUsY9T5iqU7l5Rbyy+Hu2rXsRgJff+o5cC84AB6Z7MkKnTmbn60vQjEb8R1+CW2c/Dny7Dn2QPx0H9SdgzEhi3v+UjY/Pw87FmSEP3AlAduJhDn77I8rGBmVjQ/9pt2DvojvHO154XYeEcCQ6nvfuNuX+1Y/U5P6HM1/jriWm3J94/02se9OU+10Hh9B1iCl3Js2czIZl32KsMtLO3o4rZ1oud5StLd1unUzcm2+hGY10CB+JrpMfST+sxTUwAO8B/ek4KpwDH3xMxFPPYKfT0fse0wxqFQX5xL6xGGWjsNfr6XXXfywW92k2traE3nEzOxa+jWY0EjBmhCl/Vq9DHxRAx8GhBIy5hN3vf8qGx57DzsWZoQ/eafE4z6RKg4XRR3n70r7YKsWaY+kcyy/m3n4BJOQUsCU1h5t7+BHWQU+lUaOgvJLndh0C4Fh+MRuOZ7F60mAqNY3Xoo9adAY4gJGjQtix5QA3THoJR0d7nn2xJndv+7/X+WL1LEqKy3li5kdUlFdSZTQyJKx79YnObz/HsPrr7QCMvawf1/wrzKLx9xkWQnzEAZ6/7SXsHe25bXZN/C/f/TpPfzALgO+XriX69xgqyiqYe9PzXDJpOFdNm8if320ldsd+bG1tcXZz5vYnp1gsdmVrS+CUKSQuWoRmNOIzciTOfn6krFmDLiAAjwEDKExO5tC771JVXIwhNpbUtWsJnT+fnOhoCg4dorKwkKwdplEQwdOno+vScArnltR/RG9idx1g9uSXcXC0486natbfs9P/y4JPTKNBVr67jl0bYygvreDR6+cz+uph/Ps/E5n8wLV8snAVv63aDEpx19NTzmsI2t/Vb3gI+yMOMPfWl7B3sGfakzX588KdrzPvI1P+rH5/LZEbYygvq2D2/z1P+FXDuXb6RPqE9SIhOpHn7ngVZWPDDfdeU6eq2JKqNI15Gw/x2f8NwNZGsSruJIezi3hsZBCxpwrYePTM01oP6aTn/uv9qTBqaBo8syGR3JIKi8RtNVIAQrWm2adqU0rdAczCNPnBHvPTucAQmj4JwnkNgWtNpvcwfffH+wd+O0fL1une3lcA4ORvuR3ohVRyfAWzI/+wdhjNtjBsHMsPX5y5D3BH9wnM2PantcNolmXhlwLwZNTv1g2kmV4bapr6fvCKrVaOpHl2TzFVhw3lP1k5kubR209iY+rFGTvA5Z0mMXXzZmuH0WyfjRnDzoz11g6jWUb4XgXA5rSLM3/GdDRNGR/w+sW57/1r1ji4SE4tul/5scUO/g///J9WuU5aawUITdOWA8tPP1ZKfQps1zTNclNKCSGEEEIIIf5RWu0JkBBCCCGEEOLC0uSLUC+eEyBN06ZZOwYhhBBCCCHExe2iOQESQgghhBBC/E0yDXarnQZbCCGEEEIIIS44qQAJIYQQQgjRVkgBSCpAQgghhBBCiLZDKkBCCCGEEEK0FTILnFSAhBBCCCGEEG2HVICEEEIIIYRoK2QWOKkACSGEEEIIIdoOqQAJIYQQQgjRVkgBSCpAQgghhBBCiLZDKkBCCCGEEEK0FTILHErTNGvH0JL+0R9OCCGEEEK0GhfFmUW36z+32PHxke9ub5Xr5B9fARq6apu1Q2iWqJvCAVh28FcrR9I8M3pNAGB25B9WjqR5FoaNw8l/irXDaLaS4yuYunmztcNots/GjOGaDVutHUazrBs/CoDvk3+2ciTN8+/AKwF4N+E3K0fSPPeHXAHAI7suzr5n0fBxTPj14txvAfw6IZw1f12cuQ9wXcCV3LXtT2uH0Swfhl8KwGMRF2fuvzFsHAC/pf5k5Uia54pOk6wdQtNJBUjuARJCCCGEEEK0HXICJIQQQgghhGgz/vFD4IQQQgghhBBmUv6QVSCEEEIIIYRoO6QCJIQQQgghRFshkyBIBUgIIYQQQgjRdkgFSAghhBBCiLZCCkBSARJCCCGEEEK0HVIBEkIIIYQQoo3QbKQEJBUgIYQQQgghRJshFSAhhBBCCCHaCpkFTipAQgghhBBCiLZDKkBCCCGEEEK0FVIAkgqQEEIIIYQQou2QCtAZjOig5/EBwdgoxZqkdJYfTKmz/JYeflwX1IEqTcNQVsELUYc5VVwGwMzQQMI7eqCUIiLdwP/2HLN4/JqmsemDb0nanUA7B3smPnwr7bt2adAu/chxfln8JZVlFQQNDmHs3TegzGNDY37czN71W7GxtSFoSB/GTLvOIrGnx8YT9/kqMGr4XzqSHtdMqLO8qqKCmKXLyUs6jp2LjqEP3oWzjxfFmdn8/uR8XDq2B8CzWxD9p99ikZhre//1e7jysoFkZuczZPzsRtv8b/4dTBg7gOKScmY8/h579ycDcOv/jWbOzH8B8OqSH/hy9RZLhV3NsH8/f61ciWY04hsejt+VV9ZZnn/oEH+tXElxaird7r4br8GDASg6cYLkL7+kqqQEbGzoNGkSXkOHWjz+QV4e3N3TtO1uSD3F6uS6224fvRt39+xKoIuOhXEH2ZGRVb1sWvdAhnp7olDszcllWaJ1tt11731HYuQB7BztuPHxW+jUveG2++sn64nZGEVJYTEvrFlY/fyxuKP8+P73nDp2kilPT6XfqAGWDB9N09j80bck746nnYM9V8y8Dd/G+p6jx9mw+AsqyysIHNyHMXea+p5dX//E/g07cHJzAeCS264haHAfi8SeGRvPgS9XoRk1Oo8ZSderG/Y9scuWk59s6nsG3G/qe1J3RJL084bqdgUnUhk5/yncAhp+7pY0xFvPvb2CsVWKn1PSWZVUN/f7erhxb69ggl10vBx7kG3p2dXL7uwRSJi3BwBfHTvB5lNZWJqmaax99zsORh3AzsGOm564hc6N5P4vn6xn9wZT7r+4tib3t6zeROQvu7CxtcHF3YUbH5+CR3tPi8WfE7efYytWoWlGOowKp8ukiXWW5yUe4ujXqyhKSaXXPXfhM2Rw9bL07Ts5/uNPAPhfPYn2I0dYLO7TMmLjSfjClP9dxoykWyP73n1Ll5OXfBx7Fx0DHzDlP0D+8RTiPvmKytJSlFKMfH4OtvZ2Fotd0zS+fft74iMOYO9ox22zp9ClR8PcWffReiJ/i6a4oJj//fRag+V7Nu/l4/nLmfXeo/j39LdE6NYhs8CdvQKklNIrpe6/EG+klJqmlPKr9ThZKeV9IX73hWajYPagrjy8NZ6bfo3hCn8fgtyc6rRJzC1i6sa93PLbHn5PyeKh0EAAQr1c6e/txpTf9jD51xhCPFwY5ONu8c+QtDuB3LRM/vP+s4x/4GY2vreq0XYb31/F+Psn85/3nyU3LZPkmAMAHI89xNGIOKYufpJpbz/N0H+Ns0jcmtFI7PKvGTHrQca9No/UnVHkp6bVaXN88w7sdc5c/r8X6DpxHPErv69epvP1ZuxLcxn70lyrnPwAfP7NZq6b+uoZl08YO4CugR3oO/pRHpzzAYtfuhMAD3cdcx+5ntHXPsuoa59l7iPXo3fXWSpswLT+k7/6ip4PPUTo/PlkR0VRfPJknTYOnp50nT4d77CwOs/b2NvTdfp0QufPp9fDD/PXypVUFhdbMnxsgHt7deX5PfE8sGM3ozv40EXnXKdNZmkZi+IT2Xwqo87zvdxd6a13Y+bOGB7cuZvubq709bD8tpsYdYCs1Eye+GQu1z98Mz8s+abRdr2H9+GBxY82eF7vo+fGx2+h/9hBLR1qo5JjEjCczOCOd+dx2X2T+WPpykbbbXp/JZfdN4U73p2H4WQGf8UkVC8beM1Ybn1zDre+OcdiJz+a0Uj8Z18z5PEHGfXKPNJ2RVFQr+9J2bIDO50zY15/gcAJ40hcZep7Ol0SRviCuYQvmEv/GdNw8va0+MmPDfBA7648szueu7fFMLajD/66uvutzJIy/hd3iE1pmXWeD/P2oJurjvt27uGhiH38X2AnnG1tLRi9yUFz7s/+ZC43PHIz3y8+c+7PXNIw9/26deahtx/nsaVP0m9Uf9Z/uLalQ66mGY0c/XIFfR6dyeAFz5MZEUVR/b7Ty5Oe/5mG77C6fWdFYRHH1/7IgLlzGPDMHI6v/ZGKoiKLxQ41+R/2xIOMeXUeJxvJ/xObTfk/9r8vEDRxHAfN+15jVRV7l35Kv+m3MOaVeQx/6lFs2lk2fxIiDpCRmsm8z59m8mM3sXLR6kbb9R3RhyfefaTRZaXFpWz+biuBvQNaMlTRSpxrCJweaHACpJRqTmZPA/zO1ag16OPpyonCUlKLyqg0amw4nskYP686bXZn5lFWZQQgLrsAX2cHADTA3sYGO/NPOxtFTmm5pT8CRyPjCBkbhlIKv55BlBWVUJiTV6dNYU4eZcWl+PUKQilFyNgwjkTEArDvl22E3TCednamKzjOeleLxJ17NBldex90vj7YtGtHp+FDOLV7X502aTH76BI+HAC/sEFkxR9E0zSLxNcU2yMPkmMoPOPyq68YzFffbgUgcs8R3N2c6eCrZ/yY/vy+NY7cvCIMeUX8vjWOK8b0t1TYABQmJeHo64ujj2n9ew4dSu6+uuvfwdsb586dG8wi49S+PY7tTdU3e70eOzc3KgsKLBY7QHd3V9KKS0kvKaVS09hyKpNhPnWvAGeUlpFcWEz9jDm97bYzb7u2NgpDueW33YSdcQy6fChKKfx7B1JSVEJ+dl6Ddv69A3HzaniC5tnBi47BfigrXeE7FhlHb3Pf09Hc9xTV63uKcvIoLymlo7nv6T02jKORcVaJ9zTDMVPf42zuezoOG0JGTN3cz4jZRydz39Nh6CCyExr2PSd3ReE33PKVz57urpwsLuVUSRmVmsafaZmM8K2730ovLSOpsBhjvez3d3EmNjcfowZlVUaOFRQxxFwNsqSEHXEMGm/K/YCz5H7AGXK/24Du2DvaA6btIy+z4WtbSsExU9/pZO47fcKGkLOnbv44enuj69Kw78yNj0ffpzd2LjrsdDr0fXqTuz/eYrEDGI4m4+xbk/9+w4eQXi//02P20blW/meZ8z9r/wFcu3TCzb8zAPauLigby95hEbdjP2Hm3AkKCaSksIS8RnInKCQQ90ZyB2D9xz9z+eRxtLNvA4OjlLLcTyt1rgx9FeiqlNqrlIpSSm1SSn0FxAEopW5TSkWaly9VStmafz5VSu1XSsUppR5VSv0fMAT40tz29GWpWebXRyqlupl/56dKqfeVUluVUoeUUlebn+9T671ilVLdW2aVgI+TPenm4WwA6SVl+DjZn7H9dUHt2ZGWC5hOhnZnGvj5mjB+uSaMXacMJBeUtFSoZ1SYnYert776sau3nsJ6nUFhdh6uXrXaeNW0yT2ZSUrCUb584n+sfPotTh3+yyJxl+YacPKs2fE6eXpQmmuo2ybHgJOXqY2NrS3tnJ0oLzRdLSvOzObPZ15i24tvkJ142CIxny+/Dp6kpNUMPUk9lYNfB0/8OniQcjKn5vm0HPw6WPYgpNxgwN6z5oTBXq+nIjf3vH9PYVISxspKHHx8LmR45+Tl4EBWWc22m11WjpeDQ5Nem5hXQFxOHstHD2P56GHsycolpcjy225+Vh56n5q/u7u3vtGDwNaqMNuAi1dN/C5e+kYvvrjU6ntcvPQUZtds5/t+2sIXj7zChiVfUlpomSpiaa4Bx1p9j2NjfU+tNja2trRzcqKisO6V+rSI3XQcPqTlA67Hy9GezNKa3M8qLcPb8cz7rdqOFRQx1NsDBxsb3Oza0d9Tj49j07abCykvu27u6731jR7ENkXUL7voNbT3hQrtnMoMBhxq5Y+9hwdlBsNZXlGjPNeAg0fNax08PCjPbdprL5TS3Jr9Kpwl/2vte+2cTflflJaOAiIWLmbrsy9zdP1vlgwdAENWHh6+NX2K3kdPXlbTc+fE4RRyMw30HWGZirOooZSaqJRKVEodUUrNOUObm5RSCUqpePN5yN92rtPcOUBfTdMGKKUuBdabHycppXoDNwMjNU2rUEq9C9wKxAOdNE3raw5ar2maQSn1IPCEpmnR5ucB8jVNC1NKTQUWAVeb3zcQGAN0BTaZT47uBd7SNO1LpZQ90GgVSik1A5gBsHTpUtCHnPdKaex89Uz1hSv9fejt6cI9m0xXLzu7OBLo6sxVP0YC8Pbovgz0dmNPVv55x/F3NFYRaXgi3sinMrcxVhkpKyzmltcf49Th46xb+Al3LXuu+v6gltJoJacJ76kAB70bVyx6CXtXFwxJfxGxaCnjXn0WOyenc77eklQjGaZpWqPr1uKFrWau/9rKDQaOfvwxwdOnW/wq4Plsu/V1dHKks86Z6VsjAFgwqB99snOJN1h4223syVZ8Fa2hM/cr1S0a7Z9MjfpNDCfsxokoBTu/Ws/WT75n/MxbWyLQ+kGdMaaaNo28rlYTw9EkbB3sce3c6cLG1gSN5n4Tkz8m20BPdxfeHBZKXnkFBwz5VFmjqt5o6px/7sdsjCbl0Anu/e/MCxBUE/2t1fX3+92/7xzJfRZGo5GcQ0cJnz8HW3t7dr26CPdAf7z79LqwIZ5Fc48dwBT/d+/+wG1PWmfYfFtmHlH2DjAeSAGilFJrNU1LqNWmO/AUpvONXKWU74V47/Ot80VqmpZk/v9lwGBzsABOQAawDghWSi3BdMJ0tksBK2r9+2at51dpmmYEDiuljgG9gJ3AXKVUZ+A7TdMavbyvadoyYNnphx+s2naeHxEySspp71xz9au9kwNZJQ2HwoT5ujM9pAv3bIqjwmja+C7t5MX+nAJKKk3D43aeyqWvl6tFToD2rN9C3IadAHTo5k9BVs3Vm4IsAzrPumVfFy89BbWuuhZkG3Axt3H1cqf7iP6mYSw9AlA2ipL8QpzdW3YonJOnByU5NRWHkpxcHPV143b01FOSnYuTpwfGqioqi0uwc9GhlMLWPGRPHxSAztebwrQMPIJb13je1FPZdO5YMzSlUwdP0tJzSU3LYdSImiuWnTp6snXnAYvGZu/hQXlOTRWq3GDATq8/yyvqqiwpIXHJEjpfdx2uwcEtEeJZZZWV4V2r4uPlYE9OrYrQ2Qz39SIxL59S89DW3dm59HR3s8gJ0M61W4n82bTtdu7hjyGzZhvIyzLg5unW4jH8Hft+2sL+DTsAaN/Nn8LsmvgLsw241LuXyrVexacwu6Z/0ulrPmvfKy5h7YtLWzL0ao6eHpTW6ntKc3JxaKTvKc2p1feUlGCnq7lPL21XNH5WqP4AZJWW16naeDs6kF3W9CGcK46lsOKYadKEOaE9SC22TPVzx9qtRPxkyv0uPevmviHLgJvX+eX+4ZhE/ljxG/f+d6ZFhzI5eOgpq5U/5bm5ODSx77T38CAv8VD147LcXNx79rjgMZ6No4cHJdl189+x3nbr6KGntNa+t8K873Xy1OPVqzv2rqaJS3z79yUv+XiLnwBt+WEbO9abcse/pz+5GTV9iiHTgHsTc6esuIy0pFMsfvRtAPJzClj6zEfc8+Kd/9yJEFrPNbUw4IimaccAlFJfA9cBCbXa3A28o2laLoCmaRkNfksznO/l2dq1fgUs1zRtgPmnp6Zpz5sD7A/8CTwAfHiW36c14f8AmqZpXwHXAiXAr0qpFrsrPyGnAH8XJ/x0DrSzUYz392FLraFJAD30Op4a0o3HtyWQW1ZR/Xx6cRmDfNyxVWCrFIN83EnOt8yOZOBVo5m66EmmLnqSbsNDSdgUiaZpnExMwkHnWH1yc5qLpzv2To6cTExC0zQSNkXSNawfAN2GhXI81tQh56RmUFVRVT0rU0vSBwdQdCqDoowsjJWVpO6KpsOg0DptOgwM5cS2XQCcjIzBO6QnSinK8gvQjKaD16KMTIrSM9D5tr55NtZviOGWG0YBEDawG/kFxZzKMLBh8z4uHxWK3l2H3l3H5aNC2bB53zl+24XlEhhIaUYGpVmm9Z8TFYVH/6bdh2SsrOTwe+/hPWIEXkOscxB4OL8AP2dH2js60E4pRnfwITIz59wvxDQ5Ql8Pd2zM225fvTsniiwz/GrEtaN4+L3ZPPzebPpc0o+YjVFomsbxA8k4Ojs1er9Da9J/0ujqSQu6DgvlgLnvSUtMwsHZscHFF52nO3ZOjqSZ+54DmyIJNvc9te8XOrJrH14BHS0b8bwVAAAgAElEQVTyGdyDAihKz6A405T7aRHR+A6s2/f4Dgwl1dz3nIqKwat3z+oqkWY0khYVQ8dh1sn9xPwCOjk70d7JlPuXdvRhV0bTct8GcLUznSwEuTgT5KJjd/b5D31tjkuuHcWj78/m0ffNub/BlPt/HUjGSXd+uZ96JIVv31rFHS/cjYuHZe5bPc01KJDS9AxKzfmTGRmN54Cm9Z0effqQG59ARVERFUVF5MYn4NHHskOx3IPr5v/JXdG0r5f/7QeFklIr/0/ve336hZB/IpWqsnKMVVVkHzyES6eW325H/yucOR/MYs4HswgN70ukOXeSEpJx1Dmd8V6f+pxcnHj1hxeZv2Ie81fMIzAk4J998mNhSqkZSqnoWj8zai3uBJyo9TjF/FxtPYAeSqntSqldSqmJXADnujxSAJypF/kdWKOUelPTtAyllKe5bRFQrmnat0qpo8CnZ/ldN2O6z+hmTBWe025USi0HgoBgIFEpFQwc0zRtsfn/ocAfTfmQ56tKg4UxR1k8ui+2CtYmpXMsv5h7+vhzILeQLSdzeLh/EE7tbHl1hOkKx6niMh7ffoDfU7IY4uvOigmD0DRTBWhrWtN2QhdS0OAQjkXH89G9L2DnYM+EWkNIPnvkNaYuehKAy++9yTQNdnk5QYNCCBpsGjLY9/Lh/LrkKz6d+Qq27Wy58pHbWnz4G5jGFYdOnczO15egGY34j74Et85+HPh2HfogfzoO6k/AmJHEvP8pGx+fh52LM0MeMM2ilp14mIPf/oiysUHZ2NB/2i3Yu1h2FjWA5UtmMmpEb7w9XDkS8TYL3liNnfng4sMvNvLLH3uYMHYA8VsXUVxSxj1PmK5w5+YV8cri79m27kUAXn7rO3LzLDsTkLK1JXDKFBIXLUIzGvEZORJnPz9S1qxBFxCAx4ABFCYnc+jdd6kqLsYQG0vq2rWEzp9PTnQ0BYcOUVlYSNYOUzUgePp0dF0sNxuWUYP3E48yf1BfbJRi48l0jhcVc2vXAA7nFxCZmUN3Nxee7h+Ci107hnp7cmtXfx7YGcOO9Cz6e+p5e/hgNCAmO4eoLMtvuz3DQjgYdYDXp7+InYM9Nz4+pXrZW/ct5OH3TFOr//ThWvZu2k1FWQUv3/ocQycOZ/ztV3Ii8Tifv/ARJQUlHNwVz4bPfuGxDxodVt0iAgf3IXl3Asvve4F2DnaMn3lb9bIvH32VW980xTLunpurp8EOGNSbwEGmvmfbZ2vITEoBpXDz9eSyeydbJG4bW1tCbp9MlLnv6Tz6Elw7+3Hou3W4B/rTflB/Oo8eSeyyT9k8ax52OmcG3H9n9etzEo/g6KnH2dey972dZtTgnQNHeXlwX2wU/Jaazl9FxUzt5s+hvEJ2ZebQw82FeQN749quHcN9PJnazZ8Z2/dga6P4X5jpYLe4spLX4g5htMIIuF5hIRyMPMBr017E3sGeG5+oyf03713Io++bcn/9BzW5/9Itpty/YuqVrP9gLeUlZXyx4BMA9L4eTH/hbovErmxt6XrrZPa/+Raa0Uj78JHoOvmR/MNaXAMD8BrQn4KkZBLeeY/KomJy9sVyfM06Bi94HjsXHf5XX8XeF18BwP+aq7Cz8L7LxtaWvlMnE7lwCZpWk/+J5n1v+0H96TJ6JHuXfsqmJ0z73kHm/LfT6QiaeBnbnjfNfurbvy/tB/SzaPx9hoWQEHGAF257CTtHe26bXdNvvHr368z5YBYAPyxdy+7fY6goq+DZm55nxKThTJp2QY6nLy4WnCSn3sis+poycr0d0B24FOgMbFVK9dU07W/dKKfONXuW+WajUEyVl3RN066utexmTOPybIAKTBWfEuATaqpLT2ma9rNS6gbgZfPyEcABc7tJ5rZTNE07opT6FMjFNGlCe+AxTdN+VEo9Bdxmfp9TwC2app3r6EQb2owhcK1B1E3hACw7+KuVI2meGb1M3x8wO7JFzlFb3MKwcTj5Tzl3w1aq5PgKpm7ebO0wmu2zMWO4ZsNWa4fRLOvGm6p73yf/bOVImuffgabvfXo3wfI3Ml8I94dcAcAjuy7OvmfR8HFM+PXi3G8B/DohnDV/XZy5D3BdwJXcte1Pa4fRLB+GXwrAYxEXZ+6/Mcw0sOe31J+sHEnzXNFpErSmwWVn0XX6Kotd4jj6yU1nXCdKqRHA85qmTTA/fgpA07RXarV5H9iladqn5se/A3M0TYv6O3Gdc4CspmlnvCtM07SVQGNf8tDgCyg0TfsW+LbWU4Hmf+c38vrtmqbVmeTfvDJeaaStEEIIIYQQoilazxehRgHdlVJBQCowGah/3vEDMAX41Pz9oT2Av/0t5ZadokkIIYQQQgjR5mmaVgk8CPyKaWTYKk3T4pVSLyilrjU3+xXIVkolAJuAWZqmZTf+G5uu1X3bk6Zp06wdgxBCCCGEEP9EWqspAIGmaT8BP9V7bl6t/2vAY+afC0YqQEIIIYQQQog2o9VVgIQQQgghhBAtpPXcA2Q1UgESQgghhBBCtBlSARJCCCGEEKKtsMD3OrZ2UgESQgghhBBCtBlSARJCCCGEEKKtkHuApAIkhBBCCCGEaDukAiSEEEIIIURbIeUPWQVCCCGEEEKItkNOgIQQQgghhBBthgyBE0IIIYQQoq2QabBRmqZZO4aW9I/+cEIIIYQQotW4KM4sgh/83mLHx8fe/nerXCf/+ArQ2J+2WzuEZtk0aSQAUZnrrRxJ8wz1uQqA5Yd/tXIkzXNH9wlM3bzZ2mE022djxuDkP8XaYTRbyfEVzN+z0dphNMtzAy8HYGPqT1aOpHku7zQJgFf3bbByJM0zp/94AB7Z9YeVI2meRcPH8UTExRk7wH+HjePlvRdn7gA8PWA8D+3cZO0wmmXxiLEAzNj2p3UDaaZl4ZcC8NXRX6wbSDPd0nWitUNoOpkGW+4BEkIIIYQQQrQd//gKkBBCCCGEEMJEk3uApAIkhBBCCCGEaDukAiSEEEIIIURbIeUPWQVCCCGEEEKItkMqQEIIIYQQQrQVMgucVICEEEIIIYQQbYdUgIQQQgghhGgrZBY4qQAJIYQQQggh2g6pAAkhhBBCCNFWyD1AUgESQgghhBBCtB1SARJCCCGEEKKtkAKQVICEEEIIIYQQbYecAAkhhBBCCCHaDBkCdwZDvfU8GBKMrYL1J9JZcSy1zvIbg/yY1Lk9VZpGXnkFC2OPkF5aBsBrQ0MI0bsSl5vP09EHrBE+mqbx+Vvfs3fnARwc7Znx9BSCenau06astJzFzy4nIzUbGxvFwJF9mHzf1QBUlFfy/otfkZR4Alc3HQ++MBWfjp4Wi33Dsm85Gp1AOwd7rnnkVjp069KgXdqR4/z45pdUllfQdUgI42fcgFKK9GMp/PzOSirLK7GxtWHifTfh1zPAIrEDGPbv56+VK9GMRnzDw/G78so6y/MPHeKvlSspTk2l29134zV4MABFJ06Q/OWXVJWUgI0NnSZNwmvoUIvFDfD+6/dw5WUDyczOZ8j42Y22+d/8O5gwdgDFJeXMePw99u5PBuDW/xvNnJn/AuDVJT/w5eotlgq7jpN749m9fDWa0UjXcSPpc90VdZZXVVSw853PyEk6joOLjpEP34mLrxdlBYVsffNDco7+RdCY4Qz9z81WiV/TNL55+3viIw5g72jH7bOn4N+jYf6v/Wg9Eb9FU1xQzJs/vVb9/M5fIvlh6Vrcvd0BGPOvUYy8arhF44/4ZDUpe+Jp52BP+P234x3cMP6sY8fZ+s7nVJVX0HlgH4ZN/z+UUiTtjGHvNz9hSE3nmpefwLur5bbdzNh4Dny5Cs2o0XnMSLpePaHO8qqKCmKXLSc/+Th2LjoG3H8Xzj5epO6IJOnnDdXtCk6kMnL+U7gFNPzcLSkjNp79X5ji9x8zku7XNIx/79LlGJKPY++iY/ADpviLM7PZNGc+Lh3bA+DRNYjQ6bdYNHYw5U7kp6tJNefOyPtux6uR3Mk+dpxt75pyp9PAPoRNM+VOWWERmxd9TGFmDi4+nox55E4cXJwtFn9mbDwHv1qFZjTSefRIgq+eWGe5saKCuA8+Jc+8/vvfdxdOPt6c3BFBcu38SUllxPNPWzx/cuL2c2TFKjTNSMdR4fhPqhu/IfEQR79eRWFKKiH33IXPkMHVy05t38nxH38CwP/qSXQYOcKisWuaxi9Lv+NwVAJ2Dnb867Fb6djIccPJwydY88aXVJRX0H1oCBPvuR6lFKeOpbL+7VWUl5Shb+/J9bOn4uDsaNHPYEmaTILQOipASqlPlVL/dx7tA5VS+1sqHhvg4T7BzImKZ9qWPVzm50OAi1OdNofzirh3+z7u2raXzaeyuadXYPWylcdSeXnfoZYKr0n27TrAqRNZ/O/rp7lz1o18+t/Vjba7asqlvP7VHF765HEOxSWxb6fphO3PHyPQuTrxxsq5TLx5DF+/96PFYj8anUDOyUzuXfYskx68mV/eXdVou1/eWcWVD07m3mXPknMyk2O7TbH/8ckaRk25kruWPMnoWyfxxydrLBa7ZjSS/NVX9HzoIULnzyc7KorikyfrtHHw9KTr9Ol4h4XVed7G3p6u06cTOn8+vR5+mL9WrqSyuNhisQN8/s1mrpv66hmXTxg7gK6BHeg7+lEenPMBi1+6EwAPdx1zH7me0dc+y6hrn2XuI9ejd9dZKuxqRqOR6I9XMXbOA1z1v2f5a3s0eSlpddoc3bQTexdnrn1rPj2vGsfer34AwNbOjtCbrmbgbddbPO7a4iMOkJmayfOfP80tj93E14sa33b7jejD7HcfaXTZoEsH8vQHs3j6g1kWPfkBSNmTQP6pTG5Y/ByXzJjCzg+/brTdzg9WMvKeKdyw+DnyT2WSujcBAI8ufox74m469O5qybDRjEbiP/uaIY8/yKhX5pG2K4qC1Lq5k7JlB3Y6Z8a8/gKBE8aRuOp7ADpdEkb4grmEL5hL/xnTcPL2tPjBq2Y0EvfZ1wx74kHGvjqPk43Ef2KzKf7L/vsCwRPHcWDl99XLdL7ejHlxLmNenGuVkx+A1L0JFJzK5N9vPceIu6ew66Mz5M6HKxkxYwr/fus5CmrlTtwPG+jYtyfXv/UcHfv2ZP+a3ywWu2Y0cuDzFQx+7EHCX36OtIgoClPr9v0pW7bTztmZ0QsXEHDFZRz6xrT+/S4ZxiULnuGSBc/Qb8Z0nLy9rJI/h79cQb9HZzJ0wfNkRERRVG/f5ejlSc//TKP9sLr7rorCIv5a+yMD585h4DNz+Gvtj1QUFVkyfI5EJ5CTmsnMD5/hmocms/7tbxptt/6dVVz90M3M/PAZclIzOWK+SL3urRVcNv0a7ntvDr0uCWX76t8tGb6wglZxAtTa9NK7crK4lLSSMio1jT/SMhnZvm71Y29OHmVGIwAJhgJ8HO2rl8Vk51FcWWXRmOvbvXU/4ROHoJSiW99AigpLyM3Kr9PGwdGekEHdAWhn147AHp3JyTQAELNtP6OuNFUfwi4NJX73YTRNs0jshyLi6DcuDKUUnXoFUVpUQmFOXp02hTl5lJWU0rl3EEop+o0LI3FXrHmpoqy4FICy4lJcvNwtEjdAYVISjr6+OPr4YNOuHZ5Dh5K7b1+dNg7e3jh37tzgi8ic2rfHsb3pCqy9Xo+dmxuVBQUWix1ge+RBcgyFZ1x+9RWD+erbrQBE7jmCu5szHXz1jB/Tn9+3xpGbV4Qhr4jft8ZxxZj+lgq7WvaRZFw6+ODS3hvbdu0IuGQwKdGxddqkRMcSNHoYAP7DBpIen4imabRzdMC3Vzds7axbGI/dsZ9h44eilCIoJJCSwhLysvMatAsKCcTdgrndVMejY+k22rT9+vYIoryohOLcuvEX5+ZRUVKKb49gUx81Ooy/okx/J33nDrj7tbd43IZjyeja++Dsa9p2Ow4bQkZM3W03I2YfncJNJ5Qdhg4iO+Fgg37x5K4o/IZbtnILkHs0GZ2vDzpz/H7Dh3CqXvynYvbR2Rx/x6GDyGwkfms6ERVLsDl3fJqYO8Gjwzhhzp0T0bF0HWPatruOGcbxqNgG79FS8o4l49zet1b+DCVjT933z9gTS6dwU2Wk/RnyJy0iio7Dhlgs7tPyjyXh5OuLk3nf5Rs2hOw9dfPH0dsbly4N91258fF49OmNnYsOO50Ojz69yd0fb8nwObhrP6GXmfrNzr0CKS0qoaDecUNBTh5lxaV0MR83hF42lIO74gDISskgoK/pokvwwJ4c2L6vwXv8o9goy/20UlY5AVJKTVVKxSql9imlPjc/PVoptUMpdex0NUiZvK6U2q+UilNKWWRMirejPRml5dWPM0vK8XZwOGP7SZ3bE5GZa4nQmiw3Kx8vX331Y09fPblZDQ+iTisqKGHP9nj6DO5hen1mHp7m19u2s8VZ50hhnmWu6BRm5+HmXRO7q5eegnoHgAXZebh51W1TaG4zfsb1/PHJGpZMm8fvH/3A2DuusUjcAOUGA/aeNSfL9no9FbnnnxuFSUkYKytx8PG5kOH9bX4dPElJy65+nHoqB78Onvh18CDlZE7N82k5+HXwsHh8JTkGdF417+vsqac4x3DGNja2ttg5OVFWYNmrlWeTl5WHvta2q/fRYzjLttuYvVv38dJdC/ng+U/IzbBs31ScY0DnXfM30Hk1/BsU5xhwrrX9OjfSxtJKcw04etbE7ejpQWmu4YxtbGxtaefkREVh3dxJi9hNx+GWP4AtzTXg5HXu+J1q576zE+Xm+Iszs9n8zEtsf+kNshMPWy7wWopz622/Z8gdnWdN7ug89RSbP2dJXgHOHqaLAs4e7pTmW+4CUmlubt388dBTWq/vL2tC/pyKiKaDFU6gyw0GHGrF7+DhQZmhadtkWa4BB496r8217PZckGXA3acmL9y83Smo128WZNU9tnDz1lOQZYrTN7AjibtMA4sStu4lP8u6/ZFoeRa/1KmU6gPMBUZqmpallPIE3gA6AuFAL2AtsBq4HhgA9Ae8gSil1FlvLFBKzQBmACxduhQ69zn/GBt5TqPxq2SX+/nQ092FRyLizvt9WlJjV/XOdB5eVVnFO89/zoQbR+HbyeuMr69/1aelNP7eTWhjFvPTNi6/69/0GjmAhK0xrH/rK2556cELHOUZXID1Vm4wcPTjjwmePh1l07qKtKqRLNI0DdXIZ2wtF5YbxtbIttGKLlI1uu2eR4D9RvRhyLhB2Nm3Y+va7Xz26lc8/MYDFzLEs2vs714v/sY3cSv/EZqy3hv9bDX/NRxNwtbBHtfOnS5sbE1yjuDO0sJB78blb76EvasLhqS/iHprKZe+8ix2Tk6NvKLlNN59NuFvYO3cgUbjqp/T59q3WTV//lZ/bb3jhbOq3+802veb2lz3yC38/P63bFnxCz2G9cW2na1FQrSa1vD3sTJrjPUYB6zWNC0LQNO0HHMC/qBpmhFIUEqdHv8QDqzQNK0KSFdKbQaGAmesa2uatgxYdvrhip+2n3eAmaXl+NYa0ubjZE92WXmDdoO83LmtW2ce2bWfCqP1j/Y2fLuNTet2ARDcuwvZGTVXMHIyDOi9Gx8u89HCb+jQxZuJN42pfs7TV09OhgEvXz1VlVUUF5Xi4tZyN5NG/7iFvb/uBMCvu3+dqy8F2QZcPevG7uatJz+7bpvTQ93ifo9k/IwbAOgdPpCfFq9osbjrs/fwoDynphJSbjBgp9ef5RV1VZaUkLhkCZ2vuw7X4OCWCPFvST2VTeeOXtWPO3XwJC09l9S0HEaN6F3zfEdPtu60/AQgTp56irJrrroW5xhw8nCv18aDouxcnL08MFZVUVFSgr2L5e9Xqm3zD9vYvt6U/wE9/THU2nYNmQbcvdya/Ltcat17NfKqEfzwQcvfv3fgl80c+n0HAN5dAyjKqvkbFGUbqq/Kn6bz0lNca/stzjbg5Gnd4XyOnh6U5tTEXZqTi4PevV4bPaU5uTh5mnKnsqQEO13N+k7bFY2fFao/AI4eHpRk143fsX7ue+gpya6Jv6K4BDsXHUopbO3sANAHBaDz9aYoLQN9cMtPQHHw13q5U3v7zW64/Tp76SmqVRUqyqnJLyd3V4pz83D2cKc4Nw9HN9cWj/+0BvmTa8DBQ99oG8cz5M+piCg6DrN89QfA3kNPWa34y3JzcWjivsvBwwNDYs19z2W5ueh79rjgMdYXuW4rMbWOG/Iya/IiPysP13r9ppu3vs6xRX5WzXGDd5f23P7S/QBkp2RwOCqhpcMXVmaNy8uKxq81lNVrU/tfizqYV0AnnRMdnBxopxTjOvqwIz2nTptubjoe69uVudEHMJRXWCPMBsbfEM7Lnz7By58+weBR/dj2SzSapnFkfzLOLo54eDc8iPpm2U+UFJVw20P/qvP8oJF92PpzFACRf8YSMqjbeV2FPl9Drh7NXUue5K4lT9JjRChxf0SiaRqpB5NwcHbEpd7BkYunOw5OjqQeTELTNOL+iKTHsH7Vy47HHQEged8hPP0sN4zMJTCQ0owMSrOyMFZWkhMVhUf/pt0LY6ys5PB77+E9YgReQ6xzEHUu6zfEcMsNowAIG9iN/IJiTmUY2LB5H5ePCkXvrkPvruPyUaFs2Gz5MdReXQMoOJVBYUYWVZWV/LVjN50G96vTpvPgfiRtiQDgeMQe2vfp0aK53RRj/hVePWlB//C+RGyIQtM0khKScdI5nde9PrXvF4rdsZ8O/i1/P03viWO47vWnuO71p/APC+XIFtP2m3EoCXtnpwYnQM4e7tg5OZBxyLT9HtkSif+Q0BaP82zcgwIoSs+gONO07aZFROM7sG5MvgNDSd1mush0KioGr949q3NHMxpJi4qxyv0bAPrguvGf3BVNh3rxtx8USoo5/rSoGLxDTPGX5Regme9pLcrIpCg9A2dfb4vE3WvCGK5d+BTXLnwK/6GhHDPnTuahJOzOlDuODmSac+fYlki6DDV9zi5D+nF0s2nbPro5gi4WzCm3oACK6+RPVMP8GRBK6jbTAXt6VAye9fLnVFQMHayUP25BgZSkZ1Bijj8jMhqvAU3bd3n06UNufAIVRUVUFBWRG5+AR5/zH31zvsKuGcW9b8/m3rdn02tEP2J/N/WbKQeTcdA5Nrhw6urpjoOTAykHk9E0jdjfo+g1vC8ARQbTcEnNaGTL178xZNLIFo/fqmws+NNKWaMC9DvwvVLqTU3Tss1D4M5kC3CPUmo54AmMBmYBLTo3oVGDxfHHWBjWBxvg55QMkgtLmN7dn8S8QnZk5HBvr0Cc2tny/KCeAKSXlPOMeRayt4b3xV/njFM7G1aNHcLrcUeIsvB40gEjerNv5wEev/ll7B3tmPH0lOplT0/7Ly9/+gTZGQbWfLYRvwBfnvnPG4DpJGrsNcMZc/Uw3l/wFY/d/BIubs48+PxUi8XedUgIR6Ljee/uF7BzsOfqR26tXvbhzNe4a8mTAEy8/ybWvfklleXldB0cQtchIQBMmjmZDcu+xVhlpJ29HVfOnGyx2JWtLYFTppC4aBGa0YjPyJE4+/mRsmYNuoAAPAYMoDA5mUPvvktVcTGG2FhS164ldP58cqKjKTh0iMrCQrJ2mK6IBk+fjq6L5WYDWr5kJqNG9Mbbw5UjEW+z4I3V2JknBfjwi4388sceJowdQPzWRRSXlHHPE0sByM0r4pXF37Nt3YsAvPzWd+Ra6J6x2mxsbRky/SY2vfwOmtFI8NgR6Lv4EbvqRzyD/ek8JJSuYy9hxzvLWfvwc9i76Ah/6D/Vr1/z4LNUlJRirKwkJTqWcU8/iHvnjhb9DH2GhRAfcYDnb3sJe0d7bptdk78v3/06T38wC4Dvl64l+vcYKsoqmHvT81wyaThXTZvIn99tJXbHfmxtbXF2c+b2J6ec6a1aROeBfUiJiefbh+Zja2/HqPtvq162ZtYrXPf6UwCMuOtmtr77hWkq4wEhdB5o2n7/itzHro+/oTS/kA2vvo9nYCcmzG35Iaw2traE3D6ZqNeXmKcxvgTXzn4c+m4d7oH+tB/Un86jRxK77FM2z5qHnc6ZAfffWf36nMQjOHrqcfa1zn17Nra29J06mV0Ll6BpRrqY4z/47Tr0Qf50GNQf/9Ej2bP0U35/Yh72Ls4MMsefnXiYxO9+xMbGBmxs6DftFqtURTsN7EPKnni+e3g+7eztGHlfTe6snf0K1y405c7wu25m+7tfUFlhyp1OA0y50/e68Wxe9DGHN+1E5+3BpY/e2ej7tAQbW1t633Yzu/+7GM1opNOoS3Dp5Mfh79biHhSA78D+dBo9krhln7Bl9rPY6Zzpf99d1a/PTTyMo4eH1fJH2drS7dbJxL35FprRSIfwkeg6+ZH0w1pcAwPwHtCf/KRk4t95j8qiYrL3xZK8Zh1DFzyPnYsO/6uvIubFVwAIuOYq7CycP92HhnA4KoEldy7AzsGe6x6tmcnw/QcXcu/bpq91uOqBm/jhzS+pLKug25AQupmPG+L+3E3Uj9sA6D0ylAHjh1k0fmF5yhozwCil7sB0IlMF7DE//aOmaavNyws1TXNRpksjC4ErMVWNXtQ0baVSKtDcvu853kob24whcK3BJvPVh6jM9VaOpHmG+lwFwPLDv1o5kua5o/sEpm7ebO0wmu2zMWNw8rfsge+FVHJ8BfP3bLR2GM3y3MDLAdiY+pOVI2meyztNAuDVfRvO0bJ1mtN/PACP7PrDypE0z6Lh43gi4uKMHeC/w8bx8t6LM3cAnh4wnod2brJ2GM2yeMRYAGZs+9O6gTTTsvBLAfjq6C/WDaSZbuk6EVrFDWnnFvjcLxY7+E+eP7FVrhOrzPeqadpyYPlZlruY/9UwnSjNqrc8GTjXyY8QQgghhBBC1GHdL7wQQgghhBBCWE4r/n4eS2nFtycJIYQQQgghxIUlFSAhhBBCCCHaCqkASQVICCGEEEII0XZIBUgIIYQQQog2QrPyd9+1BlIBEkIIIYQQQrQZcgIkhBucRAMAACAASURBVBBCCCGEaDNkCJwQQgghhBBthZQ/ZBUIIYQQQggh2g6pAAkhhBBCCNFWyCQIUgESQgghhBBCtB1SARJCCCGEEKKtkC9ClQqQEEIIIYQQou1QmqZZO4aW9I/+cEIIIYQQotW4KEorAa//YbHj479mjWuV6+QfPwRu6ubN1g6hWT4bMwaA2ZF/WDmS5lkYNg6AGdv+tG4gzbQs/FKu2bDV2mE027rxo5i/Z6O1w2i25wZejpP/FGuH0Swlx1cA8PmRX60cSfPc3m0CAK/u22DlSJpnTv/xADy7++LM/wWDL+fFi3jbfWbg5Sw7eHHmPsCMXhOYvGmLtcNolq/HjgbgrfjfrBxJ8zzc5woA3tx/cfY9j/Ydb+0QxHn4x58ACSGEEEIIIcxaZU3GsuQeICGEEEIIIUSbIRUgIYQQQggh2ghNZoGTCpAQQgghhBCi7ZAKkBBCCCGEEG2FkgqQVICEEEIIIYQQbYZUgIQQQvw/e/cdV2X5/3H8dbHhMA5TRBRRXGgu0HKi5ra0rbatbGrT0upnZTvb27Tya2XlaGhqw8yZAwEVxYmByFD23nD//jgIHDiUmtxIfJ6Ph486577OOe9z8bkv7vu+7vtGCCFESyHXAMkMkBBCCCGEEKLlkB0gIYQQQgghhO6UUmOVUkeUUrFKqTl/0+46pZSmlAq9EJ8rO0BCCCGEEEK0FErHf38XQylr4ENgHBAMTFVKBVto5wI8COw6369cl+wACSGEEEIIIfTWH4jVNO0vTdNKgW+BSRbavQDMB4ov1AfLDpAQQgghhBAthJWVfv+UUncrpSJq/bu7VpQ2wMlajxOrnqumlOoDtNU0bc2F7AO5C5wQQgghhBDigtM0bSGwsIHFlk6S06oXKmUFvA3cfqFzyQ6QEEIIIYQQLcRF9HdQE4G2tR77A8m1HrsAPYBNyhTaF1itlJqoaVrEv/lg2QFqQPaBA5xYtgytshKfwYPxGzfObHnu0aOcWLaMwqQkgqZPxzMkBICCkyeJX7qUiqIisLKizfjxePbrp3v+09Ex7P9yOVRqtBs2iM5XjjFbXlFWRtQnS8iJS8DW2UC/GXfh5O1JYVoGG2bPw7l1KwA8ggLpNe1GXbNn7j9A7DfL0bRKWg8ZTLvxY82WV5aVcfizxeSdSMDWYCD43uk4eHlRWV7O0S++Ij/+BCgrgqbegLFrF12zA/T1dGd6lw5YKcX6pFOsjE80W97d6Mr0Lh1p72xg/v7DbE9Nr152e6f29PPyQKHYm5nFwiN/6R2f5L0xRC5ZiVZZSccRg+g+abTZ8oqyMnZ8+AWZcQnYOxsY9NCdOPt4UpKXz9a3PyXz+AkCwy6j3x2Tdc++4PV7GHd5H9Iycgkd9YTFNm/Ou40xw3tTWFTK3Y99zN4D8QDcdN1Q5sy8CoBX3/+RpSu36BXbjKZp/PbJd8RGHMTW3o4rH7mJ1kFt67VLOZbA6reXUl5aRlBoMKPvuRalFKeOJ/Lzh8soLy3HytqKsfffQJsuAbrm37V4JYl7YrCxt2Pw/bfg1aF+/vS/Etj64ZdUlJbh36c7l067DqUUcTui2LtiHdlJp7ny5Vl4ddQve8q+GPZ+Yar9wOGD6Daxfu2Hf/wFWXEJ2DkbGPDgnRi8PcmIjSfys68B0DTofu14/Pv11i33GUl7Y4ioWneDRgyih4V198+qddfO2cDQWuvu5rc/JeP4CTqGXUb/Jlh3wVQ7Gxd9R1zkQWzs7Rj70E206li/dk7HJvDLe0spLykjMCSY4dNNtQ8QtWYze9duxcraisDQ7oTdbulygsaRG3OAxOXfolVW4jloCL5jzbcb8o8dJXH5MoqSEml/5924V203nFFRVMSh557BrXcf2k7V9/cumPp/22ffcSLKtO5ePuNmvC30f+rxBP54/yvKS8sI6NudwXea+n/X12uI270fpRSObi5cPvNmDB5uuub/8/OVJETFYGNnx/CZt+BtYexJO57Axg++pLy0jHZ9uzPoDtPYE/7NGuLDo1FWpvzDZ9yMwcOoW/4WajfQSSkVCCQBU4Dq4tc0LQfwOvNYKbUJmPVvd37gIroGSCn1oFLqkFJqqVJqVlNm0Sorif/6a7o8+CA9580jY/duCpOTzdrYe3jQcdo0vPr3N3veys6OjtOm0XPePLo+9BAnli2jvLBQz/holZVEL/mWAY/PYMRrz5C0Yze5SSlmbRI2b8fO4MTIN5+n49gRxCz7oXqZwceL4S89zfCXntZ950errOTY0m+45JGZ9HvhOVJ37aagTt+nbP0TGycDl77yIv6jRvLXyu9Nz2/ZCkDo88/S87GHOL7ctCGgJyvg3q4deW5PDA9sj2SorzdtDU5mbdKKS3gn5gibT6WaPd/VzYVuRldm7ohixo5IOrm60MNdv18eAJWVlUR8vpzhcx5gwptzOfFnBDmJ5rVzfOMO7JydmPjuPLpMGMHer38EwNrWlp43XEGfm6/RNXNtX67YzKRbX21w+ZjhvenY3pceQx9hxpxFvPfSnQC4uxl4+uFrGDpxLkMmzuXph6/B6GbQK7aZ4xEHyUxO4/5Fcxk/czI/f7jcYrufP1rOhJlTuH/RXDKT0zgeeQiADYtXMeTGcUz/YDZhN49nw+JVesYncc9Bck+lce17zzLw7qns+PRbi+12LFrGoHumcu17z5J7Ko2kvQcBcG/rx4hZ0/Ht1lHP2FRWVhK1eDlDnniAMa/PJWF7/dqP27QDW4MT49+eR+dxI4j+xlT7bm39GPnibEa/8hRDZz9A5GffUFlRoXv+8M+XM2LOA1z55lzi/4wgu07+2Kp196p359FtwgiiqtZdK1tbet9wBSFNuO4CxEUeJCsljTsWzGXUA5P5/WPLtf/7guWMun8KdyyYS1ZKGvFRptpPiD7K8V37ufW92dz+wVP0u2qEbtm1ykpOfvM1HWc8RLdnnydrdzhFdX532bp7EHDbNNz79bf4HimrV+HcubMecS1KiDpITkoqN334DMPuncLmhcssttvyyTKG3TeVmz58hpyUVBL2mNbdPlddzpS3n2TyW3NoH9qd3ct/1jN+Vf40pn7wLGH3TWXrQstjz5aFyxh671SmfvAsOSlpnKzK33vS5dzw9lNc/+aTBIT0IHKFvvn1pJR+//6OpmnlwAzgV+AQsFzTtBil1PNKqYmN2QcXzQ4QcD8wHjjW1EHy4+Jw8PHBwdsbKxsbPPr1I2vfPrM29l5eOPn71/vpOrZqhUMr0+yJndGIrasr5Xl5umUHyDoej6GVNwYfU/42l4VyKtI8f0rUPtoOvgwAv/59SY85jKZplt5OV7l/xeHo44NjVd/79A8lY4959oy9+2g10JTdO7QvWYdM2QuTU3Dv1hUAO1dXbBwdyYs/oWv+Tm4upBQWc7qomHJNY8upNC719jBrk1pcQnx+IXV7WwPsrKywsbLC1soKaytFdmmpbtkBMmLjcfb1xrmVF9Y2NgQMDCExItqsTWJENIFDLwWg3aV9OB1zBE3TsHGwx6drENa2TTex/Gf4YTKz8xtcfsXoEL7+zrSjHL4nFjdXJ3x9jIwK68WGrfvJyikgO6eADVv3Mzqsl16xzRzZuZ9LRvRHKYV/10CKC4rIy8wxa5OXmUNJYTH+3QJRSnHJiP4c2WH6OSmlKCk03SinuKAYFx2PwAIkREQTNNSU36dzIKUFRRRmmecvzMqhrKgYn84dUEoRNLQ/J3ab8hv9fXHza6VrZoDM2HicW9XUfrsBISRHmtd+UkQ07YeYat//0j6cPlBV+/Z2WFlbA6ZZln+892sjyIiNx8XXG5da6+7JOuvuyYhoOlatuwGX9uFU1bprexGsuwDHw/cTPNxUO35dAikpKCK/Tu3nV9W+X1dT7QcP70/sLtP33PfLNvpfOwobW1sAnIwuumUvjI/D3scb+6rfXe79+pETvdesjb2XF47+/tWzVWavP3GCsrxcXLrVuwOwbuLC99NlmKn/fbuY1t2COv1fkJlDaVExvl1M/d9lWH/idu0HwM7JsbpdWXGpxe/ZmOJ3R9M5zJS/VWdT/RTUGXsKsnIoKyzGt4tp7Okc1p+48Oj6+UtKaIr1uCXSNG2dpmmdNU3rqGnaS1XPPaNp2moLbYddiNkfuEhOgVNKLQA6AKuBdpjO7/sD03mB8zVNW6SUag0sA1wx5b5P07StjZGnNDsbO4+ajVY7o5GCuLhzfp/8uDgqy8ux9/a+kPH+UXFWNo4e7tWPHT3cyTpunr84MxtHT1MbK2trbJwcKc0vAKAwLYNN//cSNg6OdLv+Sjy7dNIte2l2Nva1stu7u5Nbp+9LsrJxqPr5KGtrbBwdKc8vwNDWn/Q9+/Dp34/izCzyTiRQkpkFHQJ1y+9pb096SUn144ySUjq7nt0v4SM5eezPzGHJ0EtRwNqTySQWFDVSUsuKMrMxeNb0v5OHkfTY+AbbWFlbY+voSEleAQ6uznpGPS9+vh4kpmRUP046lYmfrwd+vu4kJmfWPJ+SiZ+vu6W3aHR5GTm4etecduHqZSQvI8dsRyYvIwcXz/ptAEZPv4avn/mY3z/7ETSN2954RL/wQGFmNgavmr4zeBopzMzGqdZsZmFmNk618jtVtWlKRVnZOHnWHjeNZNat/VptrKytsXVypDSvAHtXZzJi49j9yVcUpmfS//7bqneI9FJYZ901WFh3Tf1+8a67+Rk5uHjV1IWLl5H8jByca9V+fp3ad/E0tQHISk4j8eBxtn21Bhs7G8KmXYVvJ31OoSzNysbOvfZ2g/tZbzdolZUkrVxOwLQ7yTt8qLEi/qOCzGyc66y7BZk5ZqexFWTm4Fyr/01tatbdnUt/4simcOydHJn0/Ex9gldnM8/v7GmkICMbQ62xpyAjG0Ot/M518u9aupqjm8Oxc3Jk4rwH9QneBPTeOb0YXRQzQJqm3YvpoqfhmO720BOYAAwAnlFK+WE6J/BXTdN6A72AvZbeq/bt9hYubOimE/8YyNIbn9NblGZnc/zzz+lw++0oK3272eJMzlnkV4C90ZXR77zEsBefpsdN1xLx0WLKinTcCD+rSShL3w9aDx6EvYc7kS+8zPFvl+MW1BFlrW/f/+3tTP5Ba0cH/A1OTNu6i9u37qKnh5HuRtcLGe+81B8o63+j5jKWKgs/IU3TLP4yaLIJUQsfXC+dpTZVjSLXbWPU9Kt5aMnzjJp+NWve+frCZ/w7lvqtTv9aHKKa+mjrWeRuIDgAnkGBjH19LiNfnM3hVb9RUVp2wSOes2a27lr63VU/X8M/g8qKSkryC7nx9UcZevtV/DR/sY5nNjSc65+kb96Ea49LzA68NgXL2w5n06am0WU3Xclti16g09BQ9v+s83WUFqOdzbZPTZtLb5rILQtfpNPQUA7onV/o6qKYAbJglaZpRUCRUmojpj+UtBv4XCllC/yoaZrFHaA6t9vTtm3efM4fbufuTmlmzdHg0uxsbI1nfyFceVERR95/H/9Jk3Dp0OGcP//fcvRwpygzq/pxUWYWDkbz02AcPIwUZWTh6OFOZUUF5YVF2DobUEphXXX6gDEwAIOPF/kpqbh30Ocomp270TRrU6UkKwv7On1v7+5OcWYm9h7uaBUVlBcVYWMwZQ+ackN1uz0vv4ZjKx9dcp+RXlKCl7199WNPezsya80I/Z3LfDw5kpNLcYXpuqXIjCy6uLkSk53bKFktcfQwUpBR0/+Fmdk41rkOydHDnYKMLJw8TbVTVlSEnXPTXC9zrpJOZeDf2rP6cRtfD1JOZ5GUksmQAd1qnm/twdYd+h2JjVizhT2/7ACgded25KbVHJHMTc/G2dP8Z+DiZSQvo06bqqO00RvCGX3PtQB0G9yHNe9+09jxOfTLZo5u2A6AV8cACtJraqggw3z2B6pmhWrlL8zIxlHnU/XqcvQwUphRe9y0XPuFtWu/sH7tu7bxxdrBjpzEZDx0GjfBNFtbe90tsJDfqSq/4SJad/es3cL+9aba9w1qR156TV3kpWfXu4je2dO89vMyamrfxdONTgN6oZSidecAlJWiKDcfJ7fGPxXOzt2d0qza2w1ZZ73dUPDXcfJjY0nfvImKkhK0inKsHOxpc/W1jRW32v6ft3BwvWnd9QlqR36dddfgXr//82v1v6U2AJ2HhLL2pQX0nzKhkZKbHPh5M4d+N+X3Dgowy5+fkY2TR/2xp6BWfkttADoN7se6lz+mXyPnbyoX04GPpnJRzABZUO/yCE3TtgBDMd0l4kul1K2N9eHO7dtTnJpKcXo6leXlZO7ejXuvs7seoLK8nGMff4zXgAF4hoY2VsS/ZewQQMGpVApSTfmTdkbg27enWRvfPj05uW0nAMnhUXgFdzFdO5CbV33jgILUNApOp2Lw8ar3GY3FNbA9RadTKUozZU8Nj8Czt3nfe/buyentpuxpEVG4d+2KUoqKklIqqnY2MmMOoqysMPj56ZYd4FhuHn5ODrRysMdGKYb6ehOelvnPL8R0c4Qe7m5YKbBWih5GN04W6HsDDc+OAeSdSiU/NZ2K8nJObI+kTcglZm38Qy4hbssuABJ27aFV987NZjp97foobrx2CAD9+wSRm1fIqdRs1m/ex8ghPTG6GTC6GRg5pCfrN+/7h3e7cEKvGMr0D2Yz/YPZdLmsJ/v/CEfTNBIPx+FgcKh3HY+Lhxt2jg4kHo5D0zT2/xFOl8tMPydnDzdO7I8FIH7fUTz8Gv8U3G5jw5j0+pNMev1J2vXvSewWU/7Uo3HYOTnW2wFycnfD1tGe1KOm/LFbwmkX2rOBd9eHR8cA8mvVfsKOSPzq1L5fyCXEbzXVfuKuPfhU1X5+anr1TQ8K0jLIS07F4OVZ7zMa05l1N6/Wutu2Tv62IZdwvGrdPbFrD74XwbrbZ8JQbn1nNre+M5ugy3pycKOpdpKPxGFvcDA7/Q1M9W3n6EDyEVPtHNwYTsf+pu8ZdGlPEqKPApCZlEpFWQWOOp3e5xTQnpLUVErS06gsLydr927cep7ddkP7O6fT45XX6P7yq7S59jo8Lh2gy84PwCXjhjL5rTlMfmsOgf17cmSTqf9PHYnDzsmh3g6owcMNWwcHTlX1/5FN4QRW9X92cs2NfeJ278fYpvGv5esxLozr33yS6998ksD+PTm62ZT/dNXYU3fnzFA19pyuGnuObg6nfb+e9fLHR0TjrkN+0XQu1hmgSUqpVwADMAyYo5QKAJKqrgcyAH2BLxrjw5W1Ne2nTuXIO++gVVbiPWgQTn5+JK5ahSEgAPfevcmPj+foRx9RUVhIdnQ0SatX03PePDIjIsg7epTy/HzSt5uOSnSYNg1D2/q3YmwsVtbW9Lx1Cjtefx+tspJ2Qwfi6u/Hoe9+whjYjtZ9exEQNoioBf/j98eewdbZidAHTHfDyjhyjMPfrUFZWaGsrOh1+426HiFU1tYE3TSF/W+/i1ZZie/gQRja+BH342pc2gfg1bsXrYcM5tCiz9n15P9hazDQ7Z67ACjLyyX6rfdQVgo7o5Gud92hW+4zKjVYcOQ48/r2wEopfk8+TUJBITd1DOBYbh7haZl0cnXmqV7BONva0M/Lg5s6tuOBHVFsP51OLw8jH1wWggZEZWSyO/3sdp4uFCtra0Kn3cDGlz9Eq6ykw/ABGNv6Eb18DR4d2uEf2pOOwwey/cMlrH7oWeycDQx+sKafV82YS1lRMZXl5SRGRDPiqRm4+bfWLf+S92cyZEA3vNxdiN31AS+8tRLbqgu7P/3qd375Yw9jhvcmZus7FBaVcM+sTwDIyinglfd+YNtPLwLw8rvfk5VToFvu2oL6BRMbEcOHdz1ffRvsMxbNeI3pH8wGYNwDN/DT20spKyklKDSYjqGmi6cnPDiF3z75jsrKSmxsbZkwc4qu+f37dCcxKobvHpyHtZ0tQ+6/uXrZqsdfYdLrTwIw4K7JbP3oKypKy2jTOxj/Pqb8J8L3sfPzFRTn5rP+1QV4tG/DmKdnNHpuK2tr+t5+A1teNdV+4LABuPn7cWDFGtw7tKNNSE86DBvIro+WsO6RZ7EzGLhspqn2048c5/Dq37CysQZlRci0ydjrfF2NlbU1/afdwIaqdTeoat3du3wNnh3a0Ta0J0HDB7LtwyX8WLXuDqm17n5fa909GRHN5U/NwKjjugsQGBLMXxExfHavqfbHzKyp/S8efo1b3zHV/sh7bzDdBru0lMC+wQSGmGqnx8jL+PX9r/nfzFewtrFm3MM367aDp6yt8Z98I8ffewetUsNz4CAc/dqQsnoVTgEBuPXqTUF8HHELTNsNOfujObVmFd2efV6XfGcjIKQ7CVEHWXr/89jY2zJiRs26u+zRV5n81hwAwu6ZXH0b7HZ9u9Gur6n/d361muykVLBSuHh7EHaPvrdTb9e3OwlRMXzzwDxs7G0Z9kBN/hWPvcL1b5rGniF3T2bjB6axp22f4Or8u75aRXZyKkqZ8g+5R9+xU0/N5Jhlo1IXw52/AJRS8UAoptvh+QEdMd0Q4cxNEG4DHgfKgHzgVk3T/ukKQ+3W8zgF7mLwRVgYAE+E/9HESc7P/P6m24/evW1T0wY5TwsHD+PK9Y1yjw1d/DRqCPP2/N7UMc7bs31G4thualPHOC9FCaZTzr6M/bWJk5yfW4JMfzPs1X3rmzjJ+ZnTaxQAcyObZ/2/EDKSF5vxuvt/fUay8HDzrH2Au7uOYcrG5nntx7fDhwLwbsxvTZzk/DzU3fR3q94+0DzHnkd6jIJmcuu4Tp9s0W3j/9g9Qy/KPrloZoA0TWtf9b/PNbB8CbBErzxCCCGEEEL816iL9QIYHUkXCCGEEEIIIVoM2QESQgghhBBCtBgXzSlwQgghhBBCiMYlN0GQGSAhhBBCCCFECyIzQEIIIYQQQrQQVjIDJDNAQgghhBBCiJZDZoCEEEIIIYRoIeQaIJkBEkIIIYQQQrQgMgMkhBBCCCFECyEzQDIDJIQQQgghhGhBZAZICCGEEEKIFkLJFJDMAAkhhBBCCCFaDqVpWlNnaEz/6S8nhBBCCCEuGs1iauWSL7bqtn28/9YhF2WfyAyQEEIIIYQQosX4z18DtOrEz00d4bxMChgHwMLDvzZxkvNzd9cxAMzevaGJk5yf1/pdzg/xzbN2AK5uP47fk9Y1dYzzNrLNeL6MbZ61f0uQqfYd201t4iTnpyjhG6B5r7sAi482z/qZ1nlMsx33wTT23/vnxqaOcd4WDBrerGsHYH70+iZOcn6e6DkKgBs3bW7iJOfn62FhTR3hrMklQDIDJIQQQgghhGhB/vMzQEIIIYQQQggTmQGSGSAhhBBCCCFECyI7QEIIIYQQQogWQ06BE0IIIYQQooWQU+BkBkgIIYQQQgjRgsgMkBBCCCGEEC2ElcwAyQyQEEIIIYQQouWQGSAhhBBCCCFaCLkGSGaAhBBCCCGEEC2IzAAJIYQQQgjRQsgMkMwACSGEEEIIIVoQmQESQgghhBCihVByGziZARJCCCGEEEK0HDID1ABN01j90fcc3n0IW3tbbph1I/6d2tZr98vitUSu301RfiEvrp5f/fyWlRsJ/2UnVtZWOLs5c/1jU3Fv5aFr/o2LviMu8iA29naMfegmWnWsn/90bAK/vLeU8pIyAkOCGT79WlTVyaFRazazd+1WrKytCAztTtjtk3TJfnpfDPu/XIFWqREwbCCdJ44xW15RVkbUgiVkx53EzsVA6Iw7MXh7Vi8vTM9kw+wX6HrNeDpNGKVL5to0TeOnj7/nSPghbB1suf6xG2ljoXZ+XbyWqN9NtfP8qpra+Wv/cdYs+IFTfyUz9albuWRIbz3jo2kaKz74gZhdh7BzsOWWJ6bSrnP9/Ks/W8uu3yIozCvk7XWvVT+/45dwfvxkNW5ebgCEXTWEQRMu0zX/b598R2zEQWzt7bjykZtoHVQ/f8qxBFa/vZTy0jKCQoMZfY+p9k8dT+TnD5dRXlqOlbUVY++/gTZdAnTJvuD1exh3eR/SMnIJHfWExTZvzruNMcN7U1hUyt2PfczeA/EA3HTdUObMvAqAV9//kaUrt+iSua7mvP5qmsbvC7/jeKSpdiY8dBO+FmrnVGwCa99ZSllpGR1Dghl5t6l2fnxtMZlJqQAUFxThYHDkjvdm65q/uY77ABn7Yzj29XLQKmk9ZBABE8aaLa8sK+PQp/8j70QCNgYD3e+7C0cvLyrLyzmyZCl58SdAKTrdeAPuXbvolvuM/0L97Fy8kpNRMdjY2zH0gVvw6lA/f/rxBLZ8+CXlpWW07dudy6Zdh1KKuB1RRC1fR3bSaSa+MgvvjvqMmwA5Bw5wcvkyqKzEa/BgfMeOM1ued/QoJ5cvoygpiQ53Tcc9JMRseUVRETHPPYuxd2/aTb1Rt9xNRa4B0nkGSCl1lVIqWM/PPF+Hdx8iPSmNJxY/zbUPT+aH91ZYbNftsu7MfP+Res/7Bfnz4AeP8egns7lkSC/Wfrq6sSObiYs8SFZKGncsmMuoBybz+8fLLbb7fcFyRt0/hTsWzCUrJY34qEMAJEQf5fiu/dz63mxu/+Ap+l01QpfcWmUl+5YsY8ATM7h8/lwSd0aQm5Ri1ubEpu3YGpwY9dY8Oo4dwcFvfzBbvn/pSlr1aroyO1JVO7MWP801D03mx/cbrp0H3qtfO0ZvI9c/diO9hvdt7KgWxew6RFpSGs99+RQ3PnoD376z0mK7SwZ054mPHra4rO+wPjy16HGeWvS4rjs/AMcjDpKZnMb9i+YyfuZkfv7Qcu3//NFyJsycwv2L5pKZnMbxSFPtb1i8iiE3jmP6B7MJu3k8Gxav0i37lys2M+nWVxtcPmZ4bzq296XH0EeYMWcR7710JwDubgaefvgahk6cy5CJc3n64Wswuhn0il2tua+/f0UeJCs5jXs+mcvYBybzawPj5q8fLWfsjCnc88lcspLT+Kuqdq6aPY073pvNHe/NpsvAXnQe0FPPva/ZcAAAIABJREFU+M123AdT7Rz96ht6PTKD/i8+y+lduylISjZrk7L1T2wMTlz26gu0HX05f60w1U7y5m0A9H/hGXrPeojYZd+hVVbqlv2M5l4/iXsOkpuSxvXvP8vge6ayfdG3Ftv9uWgZg+6ZyvXvP0tuShqJew8C4N7Wj8tnTce3W0c9Y6NVVpLwzdd0mvkgwc/NI3P3boqSzWvHzsOD9rdPw6N/f4vvkbx6Fc6dOusRV1wk/tUOkDI5l/e4CmgWO0AHt++n76h+KKUI6NaeooIicjNy6rUL6NYeV0+3es8H9e6EnYMdAO26tScnrf5rG9Px8P0ED++PUgq/LoGUFBSRn2meIT8zh5LCYvy6BqKUInh4f2J3RQOw75dt9L92FDa2tgA4GV10yZ11PB7nVt4YfLywsrHB/7IQTkXuM2tzKiqadkNMG9V+/fuQFnMETdMASI7Yi8HbC5c2rXXJa8nBHfvpO9JUO+3+pnbaNVA7Hr6etO7g12Tn6EZvP8ClVbUfGNyeovwicizkDwxuj5uF/E3tyM79XDLCVPv+XQMpLigir07t51XVvn83U+1fMqI/R3aYal8pRUlhMQDFBcW4eOj3Hf8MP0xmdn6Dy68YHcLX320FIHxPLG6uTvj6GBkV1osNW/eTlVNAdk4BG7buZ3RYL71iV2vu6++xnfvpUVU7bbr+/bjZpmrc7DGiP8d2Rpu10TSNw9v2EBxmfpS5sTXXcR8g9694HH18cPTxxsrGhlaX9iN9r3m/pu2JxnfgAAC8Q/uSdegwmqZRmJyCe3BXAOxcXbFxcjTNBumsudfPid3RBIWZ8vt0DqS0oIjCLPP8hVk5lBUV06pLB5RSBIX150S4Kb/R3xdjm1a6ZgYoiIvDwccHe29T7biH9iN7n/m4Y+/lhZO/f/VMp9nrT5ygLDcX1+BmsXl6QSil37+L1TnvACml2iulDimlPgKigFuUUjuUUlFKqRVKKeeqdq8qpQ4qpaKVUm8opQYCE4HXlVJ7lVIdq/79opSKVEptVUp1rXptK6XUD0qpfVX/BlY9P1cpdVgptV4p9Y1SataF6wpzORk5GL3dqx8bvYwWNwLPxu5fdtK1X7cLFe2s5Gfk4OJlrH7s4mUkv07+/IwcXDxrtfGsaZOVnEbiweMsnfUmy556l1PH9PllUpSVjaNHTb87eLhTVGcArt3GytoaGydHSvMLKC8u4dia9XS9ZrwuWRuSm25eO25eRos7QBernPQcjD41dWH0NpKdfm75927dx0t3zWfRc4vJSs260BH/Vl5GDq7eNfldvYzk1en/vDq1X7vN6OnXsOHzVbx72zNs+PxHht9+pT7Bz4KfrweJKRnVj5NOZeLn64GfrzuJyZk1z6dk4ufrbuktGlVzX3/z6o6bng3UTp2xtW6bkzHHMRhd8PDzadzAdTTXcR+gJDsLh1q1Y+9upCTLfOwozc7GvlbtWDs6UpZfgHNbf9L37KOyooKitHTy4xMoztR33IHmXz+FmdkYPGt+Bk6eRgoys83aFGRmY6hVPwZPI4V12uitLDsbW/eaSwzs3I2UZZ/dz1+rrCRx5Qr8r72useKJi9T5zgB1Ab4ARgF3AiM1TesLRACPKqU8gKuB7pqm9QRe1DRtO7AaeFzTtN6aph0HFgIzNU0LAWYBH1W9/3vAZk3TegF9gRilVChwLdAHuAYItRRMKXW3UipCKRWxcOHC8/x6gGbhvTn3Xdmo3yNIPHqSsOv1O5UAqD6iWlv9PXGLXxKAyopKSvILufH1Rxl6+1X8NH+xxfe84M7mIxqIffj7NQSNHYGNg8OFTnVOLH6Fi/kwSB2Wa+fs818yoDvPf/0MT3/6BF37duaLV7++kPH+maX8Z9OmqlHkum2Mmn41Dy15nlHTr2bNOzrn/xuWxiBN0yz+fPRYXet/6Pm1uXjW37MYNy11bJ02h7ZE0m2ovkfvoRmP+w3Eqhu+oe/nO2Qg9u5GIp9/hdhvluMa1AFl3RT3eGru9VP/uXpjzln8nPRnMdRZvTJt8ybcevTAzkO/a7QvBjIDdP43QTihadpOpdQVmE5p+7PqF7AdsAPIBYqBT5VSa4E1dd+gaqZoILCi1i9v+6r/jgBuBdA0rQLIUUoNBlZpmlZU9fqfLAXTNG0hph0rAG3ViZ/P+kttX72VXet2ANC2Szuy02qOIGSnZ+Pq6XrW7wVwLOoIf3zzG/e+MRMbu8a/38SetVvYv96U3zeoHXnpNUdl8tKzMdQ5lcfZ00heRq02Gdk4V7Vx8XSj04BeKKVo3TkAZaUoys3Hya1xT4lw9DBSVOvIXXFmFo7ubhbbOHq6U1lRQXlhEbbOBrJi40kK38OBb3+grLAIpRTWtrZ0GD2sUTMD7Fi9lfCfTX3v39m8dnLSs3H1OLfa0dvmH7fx51pT/oAu7chOramL7LRs3M6h9p1rXXsyaMIAflxUb/W/4CLWbGHPL6b8rTu3IzetJn9uejbOdU7VMx11rdOmqvajN4Qz+p5rAeg2uA9r3v2mseOftaRTGfi3rrlhQBtfD1JOZ5GUksmQATWzzG1ae7B1xyHd8zXH9Tdy7Rb2/VpVO53qjJu1xsQzXLyM9cbW2qdJVlZUcGRHNLe/3WgnKJj5L4z7APbu7mazNiVZ2dgbjfXalGSaZooqKyqoKCrCxmBAKUWnqTdUt4t8aT5OPvrMnjT3+jn4y2aO/L4dAK+gAAoyan4GhRnZONXJb/A0UlCrfgoysnFyb9pToW2N7pRl1cyAl2ZlY1undhpS8Ndf5B07RtrmzVQUF6NVVGBl74D/Ndc0VlxxkTjfrfKCqv8qYL2maVPrNlBK9QcuB6YAMzDt1NRmBWRrmna2t7hq9P3IgROHMHDiEAAO7Yph+6qt9B7Wl4TDJ3A0OFq8XqMhSbGJfPfucu58+V6c3fU5j7rPhKH0mTAUgL8iYtizdgtdh/Ql5Wg89gaHegOxs4cbdo4OJB+Jo3Xn9hzcGF79+qBLe5IQfZS2l3QiMymVirIKHF2dG/07GDsEkH8qlYLUdBw9jCTujCT0/mlmbXz79iRh6048OnUgOXwPXsFdUEox5JnHqtsc+m4NNg72uuz8AAyYOIQBVbVzeFcM21dvpdewvpw8fAIHp3OrnaYQdtVgwq4aDMCBnTFs/nEbISP6EH/IVPvncq1PTkZOdfvo7Qfwbdf454SHXjGU0CtMtXssPIaINVvoHtaXpCPxOBgc6l3H41JV+4mH42jTpT37/win35Wm1zt7uHFifyzte3Yift9RPPy8Gz3/2Vq7Pop7bxvN8tXb6d8niNy8Qk6lZrN+8z7mPTG5+sYHI4f05JlXLV/A3Jia4/obMmEoIVXjXuzuGKLWbKHb0L4kH4nH3qnhcTPpcBx+Xdpz4I9wQqpqByB+7xE82/jg6qXPKYj/hXEfwCUwgKLTqRSlpWPvbuT0rt10v+dOszZevXtyavsO3II6kBYRhbGrqXYqSkoBDWt7ezJjDqKsrTC08dMld3Ovn+CxYQSPDQMgIfIAh37ZQodBIaQdi8fWybHezo2Tuxu2jvakHo3Du1N7YjeHEzwuTJesDTG0b09xaiol6enYGo1kRewm8M67zuq1tdulb99O4Yl42flpIf7ttMRO4EOlVJCmabFKKSfAH0gGnDRNW6eU2gnEVrXPA1wANE3LVUrFKaWu1zRthTJNA/XUNG0fsAG4D3hHKWUNGIBtwCdKqVeqck8AFv3L/A3q2j+Yw+GHeO32F7Gzt+P6WTX7eG/fO59HFphuUbt20Wr2boykrKSMl258ln5jL2P0reNYu2g1pUUlfPXCYgCMPu5Me356Y8WtJzAkmL8iYvjs3uextbdjzMybqpd98fBr3PqO6daaI++9wXQ71NJSAvsGExhiugiwx8jL+PX9r/nfzFewtrFm3MM3n9NpUOfLytqanrdNZvv8D9AqKwkIG4Crvx+HVv6EMTCA1iE9CQgbSOSC/7H+0WexdXai34w7//mNddSlfzCHdx/i9WkvYmtvx/WP1dTOu/fN56GPTbWz7tOa2nn5JlPtjLplHCePJPDl859RlFfE4Z0xrP/iFx5dNEe3/N0vDSZm1yGeu/kl7BzsuPmJKdXLXp7+Ok8tehyAHz5ZTcSGKMpKynj6hucYOP4yJtw+lk3fbyV6+wGsra1xcnXiltn1jo80qqB+wcRGxPDhXc9X3wb7jEUzXmP6B6baH/fADfz09lLKSkoJCg2mY6ip9ic8OIXfPvmOyspKbGxtmTBzisXPaQxL3p/JkAHd8HJ3IXbXB7zw1kpsbU3D9Kdf/c4vf+xhzPDexGx9h8KiEu6Z9QkAWTkFvPLeD2z76UUAXn73e7JyChr8nMbS3NffjqGmcfOTu021M/6hmtr5/MHXqm9JPOb+G1j7jmnc7BASTIeQmounD26J0v3i9TOa67gPptrpfPNk9r31HlplJa0HD8TQxo+/fliNa/sAvPr0ovXQQRxatJidc+ZiY3Ci+z2mjdfSvFz2vfk+ykphbzQSfNe0f/i0xtHc66dt3+4k7olhxcx52NjZMuSBm6uX/TDrFa5+40kABk6fzJYPv6KitAz/3sH49zHlj9+1jx2fr6A4N5/fXlmAZ/s2jP2/GY2eW1lb027KVI69+w5aZSVegwbh6OdH8upVOAUEYOzVm4L4eI5//BEVhYVkR0eT/NNquj83r9GzXazk76CCOtfze5VS7YE1mqb1qHo8AniNmtPX/g/YDawCHDDN3LyhadoSpdQgTDstJcB1QCXwMdAasAW+1TTteaVUK0ynsXUAKoD7NE3boZR6DpgKnADSgE2apv3dTtA5nQJ3MZkUYLqH/cLDvzZxkvNzd1fT3/6YvXtDEyc5P6/1u5wf4ptn7QBc3X4cvyeta+oY521km/F8Gds8a/+WIFPtO7bTd8fvQilKMJ3y15zXXYDFR5tn/UzrPKbZjvtgGvvv/XNjU8c4bwsGDW/WtQMwP3p9Eyc5P0/0NP3drxs3bW7iJOfn62FhoMPZShfCwO+36XaV6PZrBl+UfXLOM0CapsUDPWo9/gPoZ6FpvZuta5r2J/Vvgz3WQrvTgKW/vvaGpmnPVc00bQHePPvkQgghhBBCtGwX880J9NL4V+ZfWAur/pCqA7BE07Sopg4khBBCCCGEaD6a1Q6Qpmk3NnUGIYQQQgghmivVFHeJv8hIFwghhBBCCCFajGY1AySEEEIIIYQ4f3INkMwACSGEEEIIIVoQmQESQgghhBCihdDr73tdzGQGSAghhBBCCNFiyAyQEEIIIYQQLYRMAMkMkBBCCCGEEKIFkRkgIYQQQgghWgiZAZIZICGEEEIIIUQLIjNAQgghhBBCtBAyAwRK07SmztCY/tNfTgghhBBCXDSaxa7F8HV/6rZ9vHH8oIuyT2QGSAghhBBCiBbC6qLcJdHXf34HKLt0XVNHOC9Gu/EAfHTwtyZOcn7uDx4NQMg3W5s4yfmJnDqk2fY9mPr/1X3rmzrGeZvTa1SzzT+n1ygAZu/e0MRJzs9r/S4HwLHd1CZOcn6KEr4BILNkdRMnOT8e9hM5VdQ8swP4Ok6kw0ebmzrGefvr/jCi0tc2dYzz0tdrAgDfHv+liZOcnykdxwLQ7q2NTZzk/CQ8OrypIzRLSqmxwLuANfCppmmv1ln+KHAXUA6kAXdomnbi336u3ARBCCGEEEIIoSullDXwITAOCAamKqWC6zTbA4RqmtYTWAnMvxCfLTtAQgghhBBCtBBWSr9//6A/EKtp2l+appUC3wKTajfQNG2jpmmFVQ93Av4XpA8uxJsIIYQQQgghRG1KqbuVUhG1/t1da3Eb4GStx4lVzzXkTuDnC5HrP38NkBBCCCGEEMLESul3k2RN0xYCCxtYbGmOyGI4pdTNQCgQdiFyyQ6QEEIIIYQQQm+JQNtaj/2B5LqNlFIjgaeBME3TSi7EB8sOkBBCCCGEEC3ERXQb7N1AJ6VUIJAETAFurN1AKdUH+AQYq2la6oX6YLkGSAghhBBCCKErTdPKgRnAr8AhYLmmaTFKqeeVUhOrmr0OOAMrlFJ7lVIX5O8EyAyQEEIIIYQQLcTFNPuhado6YF2d556p9f8jG+NzL6Y+EEIIIYQQQohGJTNAQgghhBBCtBB63gXuYiUzQEIIIYQQQogWQ2aAhBBCCCGEaCEuorvANRmZARJCCCGEEEK0GDID1ABN03jr1R/YvvUQDg62zH1xKl2D2zbYftbMT0lKzOCbH2YDkJNTwP/N+oLk5Ez8/Dx46Y3bcHVz0is+mqax+bPviI+MwcbejtEzb8anY/38p48nsP69rygvLaN9SHfC7rwWpWoODUT+uIFtS37k7iWv4OjqrEv2Aa3dmdW3A9ZK8ePxU/zvUKLZ8muDfLmhkx8VmkZReQUvhscSl1sIwLRgfyZ18KVC03gj8jg7TmXrkrm2f9v3O79dx4H126v7e+DNVxIY0l3X/LsWryRxjyn/4PtvwatD/fzpfyWw9cMvqSgtw79Pdy6ddh1KKeJ2RLF3xTqyk05z5cuz8OoYoFv2/0L+0/ti2P/lCrRKjYBhA+k8cYzZ8oqyMqIWLCE77iR2LgZCZ9yJwduzenlheiYbZr9A12vG02nCKF2zL3j9HsZd3oe0jFxCRz1hsc2b825jzPDeFBaVcvdjH7P3QDwAN103lDkzrwLg1fd/ZOnKLXrFrqZpGm+/tortWw+bxv0XJtMl2L9eu4fvXURGeh4VFZX06hvIrKeuxtraiqOHk5j/wveUlpZhbW3NrKevpvsl7XTN/978Vezadhh7B1uefH4ynbvVz//4/VX5yyvp2TeQh5805T/j2yWb+Pjttaza+BxGd4Nu+Ye2deeZwUFYWSmWH0xhwZ6TZstv7N6aW3r4UaFBYVkFT206SmxWIbZWipfCOnOJjzOVGjy/LZZdyTm65T5D0zSWvPMDe3ccws7BjvuenkpgF/P+Lyku5Z3/W0JqUgbKShEyuDtT77sCgEN7j/PFuz+ScDyFB+fdwqXDe+me/+dPvufY7oPY2tty1aM34RdUf+z8fcka9m3YTXF+IU9//3r18+Vl5Xz/xlekxJ7E0cXA9U/ehnsrz3qvbwxh7T14blgnrK3g2/0pfLQ7wWK78Z28WXBlD65YGkH06TyGtHNnzpCO2Foryio0XtoSy/aT+m836E1mP86xD5RSRqXU/VX/P0wpteYcX/8/pdR15/Ka8/2sf2v71kOcPJHGyrVPMefZG5j/4soG2278PRpHR3uz5774bAOhl3biu7VPE3ppJ774bENjRzYTH3WQ7ORUbvvoGS6/bwp/fLLMYruNC5Zx+X1Tue2jZ8hOTuVE1MHqZXnpWSTsO4yLt7tesbFSMCekIw9uiuG6dZGMCfAm0NV8x/GX+DQm/xzFjb/sYcmhRB7tGwhAoKsTo9t5c/26SGZuOsCc0KAmmea9EH3f58rh3PT2HG56e46uOz8AiXsOknsqjWvfe5aBd09lx6ffWmy3Y9EyBt0zlWvfe5bcU2kk7TXld2/rx4hZ0/Ht1lHP2NWac36tspJ9S5Yx4IkZXD5/Lok7I8hNSjFrc2LTdmwNTox6ax4dx47g4Lc/mC3fv3QlrXoF6xm72pcrNjPp1lcbXD5meG86tvelx9BHmDFnEe+9dCcA7m4Gnn74GoZOnMuQiXN5+uFrMLrpt+F9xo5thzl5Ip0Va2Yz55nrmP/i9xbbvfTGLXy58lGWfv8Y2Zn5/PFbNAAfvr2WO+8dxRcrHmX6A6P58O21esZn17bDJCaks3T1bGbNvY63XrKc/7n5t/D58kf533ePkZ2Vz6b10dXLUk9lE7HzGK1aG/WKDZjG/nlDOzFt7X7GfLObKzv5EORuPvavPprKuGWRXLE8kk/2nOTpQaZ1dEpwawDGLYvk1p+ieWpgR5riDJ+9Ow5xKjGdt5c9xfQnruezNyxvN1wxdRhvfjOHV//3GEei49i74xAAXq3cuffpqQwa1VfP2NWORRwkIymNBz/9P658cAprPlhhsV2XS3tw9zuP1ns+6tcdODo78tBncxlw9TDWf/5TY0cGTLXz4ojO3PbDPi7/XzgTu7aik0f9A84GW2um9fEnKqVm5zizqIw7foxm9Be7eeSXQ7wzrmnGTqG/c90JNAL3N0aQi82WjQcYN7EfSiku6dWevLwi0tPqH1EqLCzh6y82Me2eUfVeP2FSPwAmTOrH5o37dcl9xl/h++k2vD9KKVp3CaSkoIiCTPP8BZk5lBYV07prIEopug3vz/HwmpxbPv+ewbdOAh1/lXT3cOFkfjFJBcWUV2r8lpDGMH8P89zlFdX/72hjjVZ1M5Nh/h78lpBGWaVGckEJJ/OL6e7holv2My5E3zelhIhogoaa8vt0DqS0oIjCLPP8hVk5lBUV49O5A0opgob258Ru00aU0d8XN79WTREdaN75s47H49zKG4OPF1Y2NvhfFsKpyH1mbU5FRdNuyGUA+PXvQ1rMEbSqlSA5Yi8Gby9c2rTWPTvAn+GHyczOb3D5FaND+Pq7rQCE74nFzdUJXx8jo8J6sWHrfrJyCsjOKWDD1v2MDtP36DfAlo0xjLsyBKUUPXoFkJ9XTHpabr12BmcHACrKKykrq+DMpLlSioKCYgDy84rx8nbVLTvAtk0xjLnClL97T1P+jHPID/DBG6u59+EJKJ13IXr5uHIip4iTucWUVWqsiU1lVKD57EF+Wc3Y72RjVT32B7k78WdSFgAZRWXklZZziY/+Y3/ktgMMGRuKUopOPdpTmFdEVrp5/9s72NE9pBMANrY2BHbxJyPNNOPg3dqDgCA/s7Mw9HR45wF6X27a7mnbtT3FBUXkZdbf7mnbtT0uHm6WXz+yPwDBg3sRt+9o9djUmHr7uhKfXURCjql2fjp8mtEdveq1mzUokAW7Eygpr6x+LiYtn9MFpQAczSjA3toKO+v//gUyVkq/fxerc90BehXoqJTaS9VfZlVKrVRKHVZKLVVVa61S6hml1G6l1AGl1EJlYW1uqI1SKkgp9btSap9SKkopdeYwrMXPaixpqTm08q05AubTykhaav2B4JP313HTbcNwcLAzez4zIw8vb9MA4eXtRlZGwxsFjSE/Ixtnz5qZG2dPI/l1BrL8zBycPY3mbTJMA/Ff4ftx9nDDO7D+6RONycfJntOFJdWPTxeW4l1ndg3g+k6tWXVFKA/2CuT1yOMAeDvac8rstSX4ONV/bWP7t30PsG/dFr56+BXWv7+U4vzCxg9dS2FmNgavmvwGTyOFmdn12jjVyu9koU1Tac75i7KycfSoye7g4U5RnZ232m2srK2xcXKkNL+A8uISjq1ZT9drxuua+Vz4+XqQmJJR/TjpVCZ+vh74+bqTmJxZ83xKJn6++s08n5GWmms27nu3crM47oPpNLjxw+bhZLBn+KiepueemMgHb61l0qgXef+tNdz3kL4/i/TUXHzOMv+s+xYxacQ8nJzsCRtpyv/nphi8vN0I6uKnS97afA12pOTXjN8p+SW0MtQfv2/p4cfGm/oze2AHnt8WC8ChjAJGtffCWoG/iwM9vF3wc9Z/7M9My8XTp6b/PXyMZFo4cHpGQV4RUX/G0COksx7x/lFeejau3jX5Xb3cyE0/+1MJ8zKyca06Y8Ta2hp7JwcKcwsueM66fJ3tSc4rrn6ckl9CKxfzn393b2dau9izIS6j7surje/kTUxqHqUVcovoluBcd4DmAMc1TesNPA70AR4GgoEOwKCqdh9omtZP07QegCNwhYX3aqjNUuBDTdN6AQOBM+d/NPRZZpRSdyulIpRSEQsXLjzHr1fD8lEL832uo4eTSDyZzrDLe5735zQeC/nr7DJa+o5KKcpKSglf+SuXTZ3QSNkaZmmv1tJPYsWxFCatieD9fXHc1aNdw6/V4eiThU+t/9RZ9j3AJWMHc/vHz3LTW7MxuLuydfEP9do2Koulb/4FLHWr3keMG9Sc859NuTZQXoe/X0PQ2BHYODhc6FQXjKU+1jTN4hHvpll1G14v63pnwXR++mMuZaXlRIabNsS/X76Dhx6/klXr/4+HHp/Iy88ub9S4df3duFLXGx9P5/vf51JWVk5UeCzFRaV8+ekG7rh/dGPHtMxCTEs18OWBZIYvDWf+jjgeCDGN/SsOpXCqoIRV14cwd3BHok7lUF6pfwFZ7n/LbSvKK3j/uS8Zc90QWrXR5zqZf2J56Dz7cdHiuKrDbJbl3/3my58ZFsSLm483+B6dPZ14ckhHnvz9yAXPJy5O//YmCOGapiUCVM0KtQe2AcOVUk8AToAHEAPUPRm0Xhul1CagjaZpPwBomlZc9d5/91lmNE1bCJzZ89GyS9ed9ZdZ8c02Vn23A4DgHu04XesC+tTT2Xj7mJ/OsH9fPIcPJnLVmOcpL68kKzOf+6Z9wMeLZ+Dh6UJ6Wg5e3m6kp+Xg7tn4NxDYt24LB9ZvB6BVUDvyM7Kql+VnZOPsbj5l7VJn1iE/IxuDhxs5p9LJPZ3B0kderX7+68fmM2X+LAzujXtKx+nCElrVmrVp5WRHelFJg+1/PZHGk6FBAKQWleBr9lp70opKGy9sLReq7wEMxpo+7jF6IKtf/KQxowNw6JfNHN1gyu/VMYCC9Jr8BRnZONXJb/A0Ulgrf2FGNo4WTonQS3PPf4ajh5GizJrsxZlZONbJfqaNo6c7lRUVlBcWYetsICs2nqTwPRz49gfKCotQSmFta0uH0cN0/hYNSzqVgX/rmo29Nr4epJzOIiklkyEDutU839qDrVXXRTS2ld/+yervdgHQrXtbs3E/7XTO357GZm9vy+Bh3dmyMYb+AzqzbnUkj8yeBMDlo3vyynOWr6G4kH749k/WfG/K36V7W1LPMf+gsO78uSkGDy8XUpIyufOGt02vTc1h+tR3WPDVTDy9Gv9UvlP5pbSuNWvT2tme1MKGx/6fjqXywtBOPM6v4f/cAAAgAElEQVQRKjR48c+ajdsV1/QmPqeoUfOe8dt32/hj9U4AOnRrS0ZqTf9npmbj7mV5XFk0fwW+/l6MnxymS86G7PppK1G/mrZ7/Dq1IzetJn9ueg4unmf/s3f1MpKbloWbl5GKigpKCotxdGn8mz+l5Jfg51Jz4Ke1sz2ptWYTne2s6eJlYNn1vQHwNtjx2aRLuHPVfqJP5+HrbM/CiZfwyC+HOJFTXO/9/4uU/CHUf70DVHt0qgBslFIOwEdAqKZpJ5VSzwFmhyT/ps3fHSqo91n/Mns9108dzPVTBwOwbUsMK7/exuhxfTgQfQJnZ8fqU9rOuHbyIK6dbJqISk7K5LEZi/h48QwAhgzrwdpVu7ntrpGsXbWbocN7XOi49fQaP5Re44cCEBdxgH3rttB5cAinjsZj7+RQvYF9hsHDDVtHB1KOxOHbuT2HNobTa0IYXgF+3L3klep2n9/9LFPfeFyXu8AdzMyjrYsDfgZ7UotKGd3Om6e3mx+RaevswMl80yA12M+DhDzTL7rNiZm8NLALXx1OwtvRjrYuDsRk5jV6ZrhwfQ+m64POtI/duQ/PgMa/nqPb2DC6jTV9/smoAxz6ZQuBg0JIOxaPnZNjvR0IJ3c3bB3tST0ah3en9sRuCa9+fVNo7vnPMHYIIP9UKgWp6Th6GEncGUno/dPM2vj27fn/7N13dFTF38fx92x6I72QQBIgtIQSCL2KgEhR9AERsKEodlBEURRFKWIDFBXFCj+agAoIKr33hB4CGFogkN57snufPzamozHCLpjv6xwOm3tnN5/d3DJzZ+YuMTv34da4IVcOHMYjuClKKbq/+VJJmagf12Jpa3NTNX4A1m08xFOP3MHyNXvo0CaIjMwc4hLS2Lj9KG+/cn/JjQ/6dG/FmzOrvnnF9TZ0eFeGDjcex3fviGLl0t307R9K5LEYHJxsKzUgcnLyycnOx8OzDkVFevbuOkXr4huxeHjW4XD4Odq2b0T4/mjq+1eeh3C93Tu8K/cW59+7I4qffthN7ztDOXk8BgdHW9yryJ+bnY97cf59u07Rqm0DGjWuy+qtU0rK3d9/Bl8uGWeyu8AdS8gg0NmOek62xGfnMyjIixc2lm8EBzrblTRsegW4lzy2tdShgNwiA93quaI3aESnmmbo8B1DunHHEGO94dCek2z4cRdd+rQhOvIi9o62uFbRePxh/q/kZuUy5tVhJsn4Vzre1Z2Od3UH4MyBSPb/spMWPdty+fRFbB1sq5zrcy1NO7bgyKYD1G/egJO7jtKgVWOT9AAdjcukgYsd9evYEpeVz13NvBn7a2TJ+swCPaHzdpf8/MN9oUzfcZZj8ZnUsbHk+3tb8d6uc4Sb4c6Bwnz+aSMiE/i7mYV/NnaSlFKOwFCg4q1QqiyjaVqGUuqyUuoeTdNWKaVsAIt/mPG66No9mD07ohgyYDq2ttZMnja8ZN2DQz9g0cqX//L5j4zuzaQJC1jz83586roy46NHbnTkcgLDQrgQcZIFT7+DpY0VfZ9/sGTd4hdn8sDsVwG4/cn7S27FHNC2OYFtzXsHFL0G74ef5dPbWmChFKvPxXMuI4enWgZwMiWTHbEp3N/Elw4+LhQZNDILinhr3xkAzmXksDEmiZUDwijSNN4LP4sZRkH8689+18LVJJ6/DEpRx8uN3k8Nr/L33Cj12oRw+VAkP459GwtrK7o/U5p/9cvvMviD1wDo/Pj97Px8EfqCQvxCg6nXxpj/4oGj7Pt2BXkZWWyc+QVugX70e/05yV8NOgsLWj1yP3ve/xTNYCCgZ2fq1PMlauUvuDQIoG5YKwJ6diHii+/ZOP4trBztaf/caJNkq44Fc5+ne+fmeLg6Eb3/U6bOWomVlfE08/WiTfy+5TD9eoUSuXMOObn5PDnB2LuZmp7Nu5/8zK5fpgEw4+OfSE2/8XMHKurSvRl7dkZx38CZ2Nha88bU0grqw/fNYuGK8eTlFvDK2O8oKCjCYNAI6xDEvfcZb0rx2ltDmf3eavR6A9bWlrz61j++6em/0ql7M/btimLkXcb8r75dmn/0sFl8s9yY/7Vx31FYWIRBr9GmQxB3D+1k0pxV0WswZWc0C+5qiU4pVpyK44/UHF5oH8jxxEw2X0jmoZa+dK3nSpFBIz2/iAmbTwHgbmfFgkGtMKARn1XA+E2nzPIe2nRuzpG9UbwwbAY2tlY8OWlEybpXH/mQmQsmkJyQxqoFm/AN8GLSo7MAYyPq9rs7cTYqhlmvfVcyN2jF17/z4eKJJsvfuH0wZw6e5OPRU7GyseaeF0eWrJv33Ps8/anx1vYbvlnN8W0RFOYX8tFDb9K2X2d6Pdiftv068dOHi/h49FTsnOwZOtE09R69pjF56xn+N6Q1Fkrxw4mrnEnOYXyXBhyPy2DjuWvP+3kk1I9AFzvGdgxgbEfjVx48+ONRknMLTZLdXG7mmxOYivqncySUUkuAVkAuEK9p2qDi5Z8C4Zqmfa+UmgYMBy4Al4CLmqZNUUp9D6zVNG3lX5RpDHwJeACFwH2APzChqt/1N3H/0RC4m4mLtXHy7OcnN5g5Sc08E2wcRx62dKeZk9RMxIjut+xnD8bPf+bRjeaOUWOvtu57y+Z/tbXxjpATD5r21vfXy3vtewNg5z/ib0renHJjlgKQkr/GzElqxs3mbuJyb83sAD52d9Pw8+3mjlFj557pyaEk096+/Hpp62Gct7vs7O9mTlIzwxvdCYD/rK1mTlIzMeN7gSlvm/svDNu6w2SXh5f36nFTfib/eBiZpmkjr7H8uTKP3wDeqKLMqGqU+QO4vcLic8C2qn6XEEIIIYQQonrki1DlMxBCCCGEEELUItf9RgJCCCGEEEKIm5NO7gInPUBCCCGEEEKI2kN6gIQQQgghhKgl5C5w0gMkhBBCCCGEqEWkB0gIIYQQQohaQno/5DMQQgghhBBC1CLSAySEEEIIIUQtIXOApAdICCGEEEIIUYtID5AQQgghhBC1hHwPkPQACSGEEEIIIWoRaQAJIYQQQgghag0ZAieEEEIIIUQtITdBAKVp/+lxgP/pNyeEEEIIIW4at0TT4vFd20xWP/6622035WciPUBCCCGEEELUEjL/pRY0gMbs2mbuCDUyv9ttAAzfusO8QWpoWa8eAKQV/GrmJDXjYj2AF/ZtMXeMGpvT6fZbPv/kiE3mjlEjU8P6APDdmfVmTlIzjzbpB0BK/hozJ6kZN5u7AbDzH2HmJDWTG7OUdst2mjtGjYUP705U2lpzx6ix5i6DGLVju7lj1Mj3PXoC0Pu33WZOUjOb+3cFIKvw1jx3OVrdbu4I4h/4zzeAhBBCCCGEEEZyG2zpBRNCCCGEEELUItIDJIQQQgghRC0hd4GTHiAhhBBCCCFELSI9QEIIIYQQQtQS0gMkPUBCCCGEEEKIWkR6gIQQQgghhKglpPdDPgMhhBBCCCFELSI9QEIIIYQQQtQS8j1A0gMkhBBCCCGEqEWkB0gIIYQQQohaQu4CJz1AQgghhBBCiFpEGkBCCCGEEEKIWkOGwF1DyvETRC9djqYZqNu9G/4D7iy3Pu30Gc4uW07W5ViCn3wcz3ZhJevidu8lZu2vAPgPGoBP184mzQ6QEXmCy8uXoRkMuHftjs+d/cutz/rjDJeX/0Bu7GUCR4/BNSys3Hp9bi5RU97EObQN9UeMNGV0NE1j1syf2bMzCltbKyZPG0Gz4PqVyo176kuSEjPQ6/WEtm3Iy68PxcJCx5nTsbz3zgpycwqo6+fK2zMfwtHR1mT5E49FErV4OZpBo17PrjQa1K/cen1hIcfmLyDjQgxWjg6EPvM49p7uxO45wPnfNpaUy7wUS9e3X6NOQOX3Lvmv7erRSI4sXIlmMNCgV1ea331HpfwH5i0k9XwM1o4OdB47GgdPd5KjLxDxzRIANA1ChgygXvtQk2Y3/m6NTfN/5GzESaxsrBk47gF8gip/hnHRMaybs5jCgkIahQXTZ8wQlFKseu87UmITAMjLzsXWwY7HPplo0vyz31vNnp2njPvv1PtpGlyvUrkXnvqK5KRM9HoDrds2YMKke43776lY3p/6EwUFhVhYWDDh9XsJaelvkuxffPAk/Xu3ITE5g3Z9X6myzEdvP0K/XqHk5BYw5qV5HDlxAYAHhvbg1efvAWDm3FUsXrnDJJnL6uzjyoS2DdEpxapzcSyIulxu/QNN/Rjc0Ae9ppGaX8g7+88Ql5MPwPOtA+lW1w2AryNj2HgpyeT5NU3j61mriNgThY2tNWMnD6dRs/LbTn5eAe+/tpC42CR0Oh3tuwfz8LODStbv2nSEZV9tQCkIbOzLS1MfNFn+1BMnuLDsBzSDAe/u3fDrX/68m3HmDBd++IHsy7E0GfME7sXn3eyYS5xbvBh9bi5Kp8Nv4AA82rc3We4/tfdw4dnmDdEp+PVyPMvOxZZbPzTQlwH1vdEbNNIKCvngeDQJecbt5912wQS7OHEiNYPXI6JMnh2M288H7y5n985IbG2tmTL9YZoHX/vY8eJznxN7OYnlq94st3zhdxv5+KOf2LTzA1xdHW90bLOQ3g9pAFVJMxj4Y/FSWr30Ajaurhya+i7uoa1w8PUtKWPr7kbTx0Zxef3Gcs8tzMrm4pq1tJ08CRQcemcG7qGtsHJwMGn+S0uXEDTuRaxcXTn97nScW7XGrkx+K1c3Ah55lPiN66t8jatrVuPYpImpIpezZ2cUly4msnLdJE4cu8j701by7ZIXK5Wb/uEjODraomkar47/ns0bjnBH/7bMeOsHxr50N23bB7Hm5/0s+m4LTz0/wCTZNYOByIXL6PDKWGzdXNkzZSZebVrh5Fe3pMzlHXuwcrCn5wfvcGXfQU4v/5k2zz6OX5cO+HXpABgbDxEfzzN54+FWz28wGDj03XJ6vvY8du4ubHrjfXzbtsS5Xmn+89v2YuVgz4DZbxOzJ5xjS1fReexonOv70mfaRHQWFuSmprPhtRn4tm2JzsLCpO/hXMRJUq8k8uSXk7ly+gLr5y3nkY9eqlRu/efLufO54fg2DWTFlC84FxFFo3bB3DPx0ZIym7/5GRt70zX+AfbuOsWli0msWDuRyGMxvD/tJ75ZMrZSuekfPoRD8f47afxCtmw4Rt/+oXw2ex2jn+pL5+7N2LMzis9mr+Pzb582Sfb/rdjOFwvW8/XsZ6pc369XKI0CfWjR40U6tAnik+mj6TF4Mq7ODrz+wv/RdeDraMCeddNZtzGCtPRsk+QG45j+ie0a8ezWE8Tn5rOwbyg7YlM4n5FTUuZUahYrNxwmX29gSFBdxoY2YNKeU3St60ozV0dGrj+ElU7H/N6t2HM1lewivcnyA0TsOcXVS0nMW/kaZ07E8MX7P/LBt+Mqlbvngdto2S6IwsIi3nz2CyL2RBHWpTlXYhL5ccFmZn71HI517ElLyTRZds1g4PySJQS/+CLWrq4cnz4D19atsS9z3rV2c6PRo49yZf2Gcs/VWVsT9Nij2Hl7U5CWxrFp03AJCcHS3t5k+XXA2JCGvHIgksS8Aj7v0pq9CSlczMotKROdkc3Tu4+SbzBwl78PY5oFMu3IaQCWn4/F1kLHoPo+Jstc0e6dkVyKSWDVr29z4th53p26lIVLq774s2XjYezsbSotj7uawv69UfgUXwwQ/13SCKxCxrnz2Hl5Yefpic7SEq8O7Ug+fLRcGVsPDxzr1wNVfiZZamQkriHNsXJ0wMrBAdeQ5qSeiDRlfHIunMfGyxOb4vyu7duTfuxIuTI2Hh7Y1auHUpVnwuVcvEhhZgZOzYNNFbmcHVtP0P/u9iilaNk6kMzMXJIS0yuV+7NXR19koKiwqOS9XLyQQJt2jQDo2LkJWzcdM1n2tHMXcPD2xN7L+NnX7diOhEPlt52EQ0fx69YJAJ/2bUk+eQpNK39Lyiv7DuLbyfRXAG/1/CnRF3D09sTR2wMLS0v8O4dxJaL83z82/BiB3TsCUK9jG+JPnEbTNCxtrEsaO/rCQsA8s0T/2HecFrd3QCmFX7MG5GfnkpVSfvvPSkknPycPv2YNUErR4vYO/LGv/PvUNI1Tuw4T3LN87+6NtmNrJP3vCjPmah1AVmYeSYkZlco5lNl/Cwv1JYdSpRTZ2XkAZGXm4eFZx2TZdx84RUpa1jXXD7ojjCU/7gTgwOFonOvY4+PlQt+erdm88zip6dmkpWezeedx7ujZ2lSxAQhxc+JSZh6x2XkUGTQ2xCTS0698JS4iIZ18vQGAE0kZeNtZA9DQ2Z5DCenoNcjTG/gjLZvOdV1Nmh/gwI4T3NbfuO00bRlAdmYuKUnltx0bW2tatgsCwMrKkkZN65GcYNw/Nqzex4ChXXGsY2w4uLg5mSx71vnz2Hp6YVt83vVo357UI5XrDQ5VnHftfLyx8/YGwNrFBSunOhRmmq7xBtDMxYnY7Dyu5uZTpGlsvZpIF6/y28+RlHTyDcbtJyotE09b65J1h5PTyTFxg7mi7VuPMvDuTsV1h4ZkZeaQWEXdIScnj0ULN/P4k5UvjM56fyXjxv9fxardf45Ome7fzeq6NoCUUg8qpQ4opY4opb5USgUopf5QSnkopXRKqZ1KqTuKy65SSkUopSKVUmPKvEaWUmq6UuqoUmqfUsq7eHmj4p8PKqXeUUpd+yz1LxWkpWHjVnrwt3F1JT8trVrPzU9Nw8a1wnNTq/fc66UgNQ1r19IDl7WLK4XVzKAZDMSuXI7f/w29UfH+VmJCOt4+LiU/e3m7kJhQ+SAGMPbJL7iz52Ts7W25va+xwtEoqC47tp4AYPP6oyTEme7zz0tNw7bMtmPr5kpehc++bBmdhQWWdnYUZpW/Unx1fwR1O7W78YEruNXz56amYe9emt/OzYXclLRrltFZWGBlb0dBpjF/cvR5fn95KhsmTids9HCT9/4AZCan4+RRuv07ubuQmZz+12U8Kpe5FHkWBxcn3Hy9bmzgChITMsrtv57eztfcf1946isG3PY29g429Orbyrjslbv5dNY6BvedxtxZa3l6nGl6b6vD18eNy1eTS36OjUvB18cNXx9XLl9JKV1+NQVfH9M2ILzsbIgvHs4GkJBbgJdd5Svcfxrc0Ic9V1MBOJOWTZe6rthY6HC2tiTMyxnvKq6O32gpiel4eJduO+5ezqRUUYH9U1ZmLgd3RdKqfWMArsQkEhuTyKtPzOWVxz7m0N5TNzzzn4z1hjLnXVcX8tNS//HrZJ4/j1ZUhK2n5/WM97c8bK1JzCso+TkxrwAP22tvA/3reXMg8Z+/vxspIT4N7zL7nZe3K4nxlc//8+b+woOP9MG2TAMOjA0oTy8XmjSrPGRX/PdctwaQUqo5cD/QVdO0UEAP9ATeA74AXgJOapr2Z9/vY5qmhQHtgLFKKffi5Q7APk3TWgM7gCeKl38MfKxpWnvgyl/kGKOUCldKhc+fP79mb+ZffT9UFU82+aWEqjJU75lJ27dRp0VLrN3M1/1bsTfBqOo38MmXT7Fu69sUFBYRvv8PAN54Zzgrl+3i4WEfkZOTh6WVCSuxVWSv1Mv2N28v7ex5LGyscarnd32zVcctn7+KZZXyX3v/cA9qwJ0fTKbPtImcWr0BfUHhdY/496r6G1Qs8vf7eNSOCJr3MG3vD1C9bajYnC+e4JctkyksKCLiQDQAPy3fy7iX72L1xjcY9/LdzHhr+Q2N+0+oKo5DmqZV+f6qPIzdSFV8xNeK0D/Ak+Zujiw8ZZwjtD8ujd1XU/m2T2tmdGnG8aRM9CZ/A//sM9MX6Zk1eREDh3XHx89YfTDoDVy9lMS0ec/w0rQH+Wz6crIyc//mla6Tqrb7f9iLXJCWRvQ339Jo1CiUzvwDdKo+F0MfX0+aODuy/HxslevNpcrDYoU/welTl7gUk8jtfcrP78zNLeCb+b/z1HN33cCENw+lNJP9u1ldzzlAvYEw4GDxycAOSNA0bYpS6j7gKaDsFjdWKXVv8eP6QGMgGSgA1hYvjwD6Fj/uDNxT/HgJ8GFVITRNmw/82fLRwndt+8dvxNrVhfyU0isb+amp2Li4/MUzStm4upJ2+ky557o0Ne1cGmtXVwpSS69GFqSlYlXN/NnnzpIVHU3S9m3o8/PR9EXobG3wu3fIjYoLwIqlu1j9414Aglv4E1+m1yYhPg1Pr2sPg7GxsaLHbSHs2HqCjl2aEtjQm7nzjXMGYi4ksHuH6SZk2rq5kldm28lLScXGxblCGRfyUlKxc3PFoNdTlJtbbo7Y1X3h+Jqh98SY7dbOb+fmQk5yaf7clDTsXJ0rlHElJzkVe3dj/sKcXKwdy8/Rq+Png4WtNemXr+DWMOCG545Yt4Oj643bf93G/mQmlW7/mclpOLqVfw9OHi7lyySl4VSmjEGv5/TeY4yaPeEGJzdauWw3a37cD0DzkPrl9t/E+PS/HMZmY2NFt9tC2LE1kg6dm/DrmghenDgYgN53tOLdKStubPh/IDYumXp13Ut+9vNx42p8KrFXU+jeuXnp8rpu7Nxr2ongCTn55XptvOysSczNr1Sug7cLjwX7M2bLMQoNpZWTb09e4tuTlwCY1rkpl0zUcPh1xS42rDZuO42D65NU5op9ckI6bp7OVT7v83dXULe+B3eP6FGyzN3LhSYt/LG0tMDb1x3fAE+uXkqk8V9MhL9erF1dyU8pc95NTcO6muddgKLcXE7NnYv/PYNxatTwRkT8S0l5BeWGtHnaWpOcX1CpXFt3Z0Y2qsf4/SfKbT/msnzpNn5euRuA4BYBxMeVHv8T4lPx8Cr/Nzh25BxRJ2MYdMfr6PUGUpIzGTNqFi9Pup8rsUmMGDKt+LlpPHDfDBYum4iHR9XboLi1Xc9LDApYoGlaaPG/psWNH3vgz/5ERwCl1G1AH6BzcU/PYeDPmbqFWullBz1muFFDnQaB5MYnkJuYhKGoiIQD4biHVm88t2tICKmRJynMzqYwO5vUyJO4hoTc2MAV2AcEkp+QQH5SIoaiIlIPHsS5VfXyB45+ghbvvkfIjJn4DRmKW8fON7zxA3DfiG4sWvkyi1a+TI/bW/DbmoNomsbxoxdwdLTDo8JJMCcnv2ReUFGRnj07owhsYBzqk5JsHDttMBj4dv5G7h3W5Ybn/5NzgwCy4xPIKd52ru4Px6tNq3JlvNq0InbXPgDiDh7CvXnTkivImsHA1YOHqNvRPA2IWz2/W6MAsuISyEpIQl9URMzeCHzDWpYr4xvWkgs7jRWuy/sP4xXSBKUUWQlJGPTGMezZiclkXknAwcO90u+4EcIG9uCxTyby2CcTadypFSe2HEDTNGJPncfG3rZSA8jRzRlrO1tiT51H0zRObDlA406l7/PCkdO4+3lRx8M0w7CGDu/KwhXjWbhivHH//SXCmOvoRRycbCs1gIz7r3FuR1GRnr27ThFQvP96eNbhcPg5AML3R1Pf38Mk76E61m08xMgh3QHo0CaIjMwc4hLS2Lj9KH26t8LF2QEXZwf6dG/Fxu1H/+bVrq+TKZnUd7LF18EGS53iDn9PdsSmlCvT1MWBSe2DGL8zktT80t5NnQJna+OpNsjZnsbODuyLM83wpgH3dWPOopeYs+glOvZowbbfjNvO6eMXcXC0xc2jcuN58Re/kZ2Vx+gXB5db3rFnC05EnAUgIy2LKzGJePuZZh92DAwkLyGBvOJjZ9LBg7i2rt5511BUxOnP5+HZuTPu7cxz7DyVnomfgx0+djZYKkWvup7sSSi//QTVceDFFo2YHBFFmll6xysbNuI2lv74Okt/fJ3bbm/NujX7iusO53B0tMOzQt3hvuE9Wb91Jms3TOebhRMICPRi/vfjadzEj007PmDthums3TAdL28XFq+Y9J9t/MgcoOvbuNgMrFZKzdY0LUEp5QY4AROAxcBF4CtgEOAMpGqalqOUagZ0qsbr7wOGAD8Aw69j7kqUhQVBDwzn+OyP0QwGfLp1xcHPl/Or1uAUGIBHaGsyzl8g8rN5FGXnkHz0GBdW/0L7qVOwcnTAf9BADk17F4CAuwZi5Wi6O8D9mb/e/SM5+8kcNIOGe5eu2Pn6cXXNauwDAnBuHUr2hfOc/+Jz9Dk5pB8/Rtza1TR/6x2T5ryWrt2D2bMjiiEDpmNra83kaaV/7geHfsCilS+Tm1PAhOe/obCgCL3BQLsOjUsaOht+O8TKZcYrQr16t+SuezqYLLvOwoLgh4Zz8IO5aAYD9Xp0wameL2d++gXnQH+827amXo+uHJv/PdtffhMrB3tCnxld8vyU09HYurlg72Xa8d//pfxtRw1jx8zPjLfBvq0zzvV8ObFiLa4N/fELa0XD27qw//MF/PriW1g7ONDp+ccASDp9llNrNqCztAClI+zR+7GpY/pboDZqF8y58Ei+HPMOVjbWDBj3QMm6b8e+V3JL637PDGPdnMUUFRTQMCyYhmGlNy05ueOQyW9+8KcuxXdvu2/gTGxsrXlj6rCSdQ/fN4uFK8aTl1vAK2O/o6CgCINBI6xDEPfeZzwNvPbWUGa/txq93oC1tSWvvmW6+YgL5j5P987N8XB1Inr/p0ydtRIrK+Np8utFm/h9y2H69QolcucccnLzeXLClwCkpmfz7ic/s+sX49XjGR//RKoJ7wAHoNfgg4izzO3ZAgudYs25eM5l5PBkiwCiUjLZcSWFsaENsLO0YGZXY29VfE4+43eexFIpvuptrKxnFxYxed9p9Ga4uB/WtTkRe6J4asi72NhaMXZy6bH/hQc/Ys6il0iKT2PFd5uoF+jF+IdnAzDwvq70HdyJNp2acmT/aZ67/310FopRz99FHWfTnH+VhQUNRo4gas4cNM2AV9eu2Pv5ErN6NY4BAbiFhpJ1/gKnP/+copwcUo8d49LqNYS+8zbJ4eFk/nGGoqwsEnbvASDo0Udx8DfdXTQNGsw9eY732oegU/Db5QQuZuUyqrE/p9Oz2JuQwpimgdhZWPBmm6aAcW6mecwAACAASURBVJ7Z5EPGns45HVtQ39EeOwsdy3q148Pj0YQnmXb+c7ceLdi98wSD+7+JrZ01U6Y+XLJuxJDpLP3xdZPmETc3da0xnjV6MaXuB17D2LNUCIzHOAeoq6ZpeqXUT8AvGIewrQL8gNOAJzBF07RtSqksTdP+7CkaCgzSNG2UUqoxsAhjT9M6YIymaX83yUAbU4MhcDeD+d1uA2D4VtN/l8T1sKyXcVhCWsGvZk5SMy7WA3hh3xZzx6ixOZ1uv+XzT47YZO4YNTI1rA8A352p+hbzN7tHmxi/9yklf42Zk9SMm83dANj5jzBzkprJjVlKu2U7zR2jxsKHdycqbe3fF7xJNXcZxKgd280do0a+79ETgN6/7TZzkprZ3L8rAFmFt+a5y9HqdjDX7UP/odfDN5vsEsf0dr1vys/kug4v0zTtB4w9NGV1KrP+/8os708V/mz8FD9eCaws/jEW6KRpmqaUGg6EX5fQQgghhBBCiFrjVvoi1DDgU2WcbJAGPGbmPEIIIYQQQtxSdDfx3dlM5ZZpAGmathMw7TfLCSGEEEIIIf5TbpkGkBBCCCGEEOLfuZnvzmYq5v+mLSGEEEIIIYQwEekBEkIIIYQQopaQHiDpARJCCCGEEELUItIAEkIIIYQQQpicUupOpdRppVS0UurVKtbbKKV+KF6/XykVeD1+rzSAhBBCCCGEqCUsTPjvryilLIDPMH43aDAwQikVXKHYaCBV07QgYDbwXg3fdjnSABJCCCGEEEKYWgcgWtO0c5qmFQDLgMEVygwGFhQ/Xgn0Lv5O0H9FboIghBBCCCFELWHKL0JVSo0BxpRZNF/TtPnFj/2AS2XWXQY6VniJkjKaphUppdIBdyDp3+SSBpAQQgghhBDiuitu7My/xuqqenIqts6qU+YfkwaQEEIIIYQQtcRNdBvsy0D9Mj/XA65co8xlpZQl4Ayk/NtfLHOAhBBCCCGEEKZ2EGislGqglLIGhgNrKpRZAzxS/HgosEXTtH/dA6Suw2vczP7Tb04IIYQQQtw0bp6+lb/w/rGNJqsfv9Kq719+JkqpAcAcjDeN+1bTtOlKqXeAcE3T1iilbIH/AW0w9vwM1zTt3L/N9Z8fAtf7t93mjlAjm/t3BWDMrm3mDVJD87vdBsCm2F/NG6SG+vgNoN/6XeaOUWPr+3Vjwv4t5o5RYx92vJ1phzeZO0aNvNGmDwDzT603c5KaGdOsHwBxuRUvwt0afOzuBqDdsp1mTlIz4cO7Y+c/wtwxaiw3Zilwxtwx/oUm3Lbu1qw3bBtorDfc6vWezMLNZk5SM05Wvc0d4ZakadqvwK8Vlr1Z5nEecN/1/r3/+QaQEEIIIYQQwsjiluinurFkDpAQQgghhBCi1pAeICGEEEIIIWqJm+gucGYjPUBCCCGEEEKIWkN6gIQQQgghhKgldEpukiw9QEIIIYQQQohaQ3qAhBBCCCGEqCVkDpD0AAkhhBBCCCFqEWkACSGEEEIIIWoNGQInhBBCCCFELWFh7gA3AekBEkIIIYQQQtQa0gMkhBBCCCFELSE3QZAeICGEEEIIIUQtIj1AQgghhBBC1BLyRajSALqm9h4uPNu8IToFv16OZ9m52HLrhwb6MqC+N3qDRlpBIR8cjyYhLx+Ad9sFE+zixInUDF6PiDJHfFKOnyB66XI0zUDd7t3wH3BnufVpp89wdtlysi7HEvzk43i2CytZF7d7LzFrfwXAf9AAfLp2Nml2TdNY8enPRO6PwtrWiodeGYF/k/qVyq35Zh37N4STk5nD7F/fK1m+9/cDrPpyDc4ezgD0vKc7XQd2Mln+dh4uPNWsIRZK8dvleJafv1xufQvXOjzVrCENHR2YcewUu+KTS9aNbhJIBw9XAJacu8T2uCST5f5TwrFITixajmbQ8O/ZlcZ39Su3Xl9YyJEvF5B2IQZrRwfCnn0ce093chKT2frq2zjW9QbAtVEDWj060uT5Y49EEr5gJZrBQNDtXWkx+I5K+Xd/tpCU88b8PcaNxtHLnfzMLLbP/prksxdp1LMTHR673+TZwbj9b/3qR85HnMTSxpo7xz2Ad6PK2398dAy/f7KYovxCGoQF0+uJIShlHNdwaO12jqzbic5CR4N2IfQcNdik+T95fzX7d53CxtaK1965nybN61Uq9/IzX5GclIm+yECrtg144bV7sbAoHZSwbME25s1ex+qtU3BxdTBJ9s4+rkxo2xCdUqw6F8eCqPL77gNN/Rjc0Ae9ppGaX8g7+88Ql2M87j/fOpBudd0A+Doyho2XTL/vfvHBk/Tv3YbE5Aza9X2lyjIfvf0I/XqFkpNbwJiX5nHkxAUAHhjag1efvweAmXNXsXjlDlPFLqFpGtOnz2f79ghsbW2YOXMcISFBlco99NBrJCSkYmtrDcC3376Du7sLV64kMHHiHDIzs9HrDUyY8Ag9e7YzWf4Oni48F9wQCwXrLsWz5Gz5esN9DXwZWN8bvWasN7x/LJr43HyC6jjwYouG2FtaYtA0FkVfZutV028/t3q9R9M0Pnx3Bbt3RmJra8WU6Q/TLNj/muVffG4esZeTWL5qMgDz5v7C9i1H0el0uLo5MmX6w3h6uZgqvjCxf9UAUkplaZrmWMXy74G1mqat/DevX+E1RwHtNE177nq95rXogLEhDXnlQCSJeQV83qU1exNSuJiVW1ImOiObp3cfJd9g4C5/H8Y0C2TakdMALD8fi62FjkH1fW501CppBgN/LF5Kq5dewMbVlUNT38U9tBUOvr4lZWzd3Wj62Cgur99Y7rmFWdlcXLOWtpMngYJD78zAPbQVVg6mqYAARO6PIjE2kSn/m8SFqIssm7OSVz5/sVK5lp1D6HlPN6Y8NKPSura3teH+cUNMEbccHfBs80a8Fn6CpLwC5nYOZV9CMjHZpdtOYm4+Hx0/w9DA8pXCDh6uBDk58PTew1jpdHzYviUHE1PJ0etNll8zGDi+cBmdXhmLnZsrO9+aiU/bVjj51S0pc2n7Hqwc7On94TvE7jtI1A8/E/bc4wA4eHnQc9rrJstbkcFg4MC3y+nz+vPYu7vw26T3qRfWEpd6pfmjt+7F2tGeez5+m/N7wjm0ZBU9XhiNzsqK0GGDSLt0lbRLV8z2Hs5HnCT1aiKPfTGZq2cusGnech748KVK5TZ9sZy+zwynbtNAfnrnCy4ciqJBWDAxx85wdv9xHv5kIpZWVuSkZZo0//5dp7gck8TiNRM5eTyGWdN/4otFYyuVm/L+Qzg42qJpGm9OWMi2jcfofWcoAAlxaYTv+wPvuqarfOgUTGzXiGe3niA+N5+FfUPZEZvC+YyckjKnUrNYueEw+XoDQ4LqMja0AZP2nKJrXVeauToycv0hrHQ65vduxZ6rqWQXmW7fBfjfiu18sWA9X89+psr1/XqF0ijQhxY9XqRDmyA+mT6aHoMn4+rswOsv/B9dB76OBuxZN511GyNIS882af4dOyK4cOEKGzZ8ydGjp5kyZR4rVnxUZdkPP3yJli0bl1s2b95y+vfvxsiRA4iOjmHMmLfZsuUbU0RHB4wLaciE/cZ6wxfdWrM7vny94Y+MbJ7cZaw33O3vw5PNAnnn8Gny9HpmHPmD2Jw83G2smd+tNQcTU8ky4fZzq9d7AHbvjORSTAI//zqFE8cu8O7UZSxYWvWFgC0bD2Nvb1Nu2UOP9uHp5+8CYNmirXw171cmvWX6i3imYCFzgGQOUFWauTgRm53H1dx8ijSNrVcT6eLlVq7MkZR08g0GAKLSMvEsvhIFcDg5nRwTn/jKyjh3HjsvL+w8PdFZWuLVoR3Jh4+WK2Pr4YFj/Xqgyu8FqZGRuIY0x8rRASsHB1xDmpN6ItKU8Tm25wQd+7ZHKUWD4EBys3JJT06vVK5BcCDO7s4mzfZ3mjo7cSUnj7jibWfb1UQ6e7mXKxOfl8/5rBwMlO+C9ne051hqBgYN8vUGzmVm0664N8hUUs9ewMHLEwcv47bj26kdcYfKbztxh45Sr5uxR61u+7YknjyFpt0c3enJ0Rdw8vHEydsDC0tLArqEcSn8WLkyl8KP0ahHRwACOrYhLvI0mqZhZWuDV7MgLKzM2zF+9sBxgnt1QCmFb9MG5GfnkpVSfvvPSkknPycP32YNUEoR3KsD0fuN7/Po77voMKQvllZWANi7OJk0/65tkfQbFIZSipBWAWRl5pGcmFGpnIOjLQD6IgOFhfpyh6JPP1zDUy8MRGG6s3SImxOXMvOIzc6jyKCxISaRnn7lj/sRCenk643H/RNJGXjbGY/7DZ3tOZSQjl6DPL2BP9Ky6VzXtPsuwO4Dp0hJy7rm+kF3hLHkx50AHDgcjXMde3y8XOjbszWbdx4nNT2btPRsNu88zh09W5sqdonNm/dxzz23o5QiNLQZGRnZJCSkVPv5SkFWlrHBmpmZg1eF8/aN1MzFidic0nrDliuJdPWuUG9ILq03nCxTb7icnUdsTh4AyfkFpBYU4mxtZbLsJflv4XoPwPatxxhwd0eUUrRs3YDMzBySEivXHXJy8li8cAujn+xfbrmjo13J49zc/JIedfHfVO0GkFJqvFLqRPG/FyqsU0qpT5VSJ5VS6wCvMusuKKXeU0odKP4XVLzcUyn1o1LqYPG/rsXLOyil9iilDhf/37SKLAOVUnuVUh41fud/wcPWmsS8gpKfE/MK8LC1uWb5/vW8OZCYeiOi1EhBWho2bqUnXxtXV/LT0qr13PzUNGxcKzw3tXrPvV7Sk9JxKdPt7OLpQlpS5YPYXzmy8yjTH3+fr6Z8R2qC6f427rbWJBYPCQBIysvHo8xJ4q+cy8ymvYcrNjoddawsae3mgudfbHc3Ql5qGnbupX9/WzdX8ir8/cuW0VlYYGVvR0GW8UpxTmIy29+Yzu7ps0g+/YfpghfLSUnDoUx+BzcXclPSKpWxL5vfzo78TNNe6f4rWcnpOHmUbv9OHi5kVbgAkJWcjpN7mTLupWVSryRy+eRZFk/4iB8mfUzcHxdNE7xYUkIGXj6l2Ty9nUlMqHr/nfD0Vwy+/W3s7W3o2acVALu3ReLh6UxQU98qn3OjeNnZEJ9Tuu8m5BbgZXft/W9wQx/2XDUeW86kZdOlris2FjqcrS0J83LG2960+251+Pq4cflq6ZDb2LgUfH3c8PVx5fKV0oZG7NUUfH1M34CLj0/Gx6f0tO7j4058mSHCZU2a9DGDB4/ls8+WlVyAee65kfzyyzZ69BjFmDFTeOONJ02SG8DT1prE3PL1hr86fg+sX3W9oZmzI1Y6xZXiBpGp3Or1HoDE+DR8ymy33t6uJMRXrr/Mm7uWBx/pXTKEsqzPPl7NwN6T+G3dQZ56btANzWtOOmW6fzerajWAlFJhwKNAR6AT8IRSqk2ZIvcCTYGWwBNAlwovkaFpWgfgU2BO8bKPgdmaprUHhgBfFy8/BfTQNK0N8CZQbnyTUupe4FVggKZplQbJKqXGKKXClVLh8+fPr87bq5ZrXeHu4+tJE2dHlp+PrXK9Wfyri/FVPNnEV0Gq+qz/yZWYlp1DeGfJm7z+9Ss0a9uEhTOXXM94f6mqlNXtHDmUnMbBpBRmd2zFa62aEpWWgd7kPStV/T5VrRI2LnXoM3s6Pae9TsjIIRya9x2FublVlDaxSttOVduXaaJUR9Xbf6VSlZ9YXMagN5CflcPID8bTY9Q9/PL+dybtofsn+++H857gp02TKSws4tCBaPJyC/jf15t57Jk7qix/Q1UR8VqfWv8AT5q7ObLwlHGO0P64NHZfTeXbPq2Z0aUZx5MyzbDv/r2qetQ0Tavy72OO+FX9zqqyffjhBH755VMWL55JREQkq1dvBWDduh3ce29vduz4nvnzp/DKK7MwFPdYmIN2jS2or58nTZ0dK82xcbOxYlJoE947+se/O41fJ7dUvYfqHXtOn7rE5ZgEevUJrfI1nh03mHWbZ9B/YHuWL9l+Q3KKm0N1x3p0A37WNC0bQCn1E9C9zPoewFJN0/TAFaXUlgrPX1rm/9nFj/sAwWU2zjpKKSfAGViglGqM8fxTth+4F9AOuEPTtMpjKgBN0+YDf7Z8tB9+213Nt1gqKa+gXNeup601yfkFlcq1dXdmZKN6jN9/gkLDzXC4MrJ2dSE/pfTKTH5qKjYu1RtLb+PqStrpM+We69K0yXXPWNH2VbvYvW4vAAFN/UlLKL1qk5aYhrN7nWq/lqNz6XylrgM7s+qrtdcv6N9IqnDVz8PWpspt51qWnrvM0nPGStWrrZoQm2PaBoStqyu5yaXbTl5KKrau5YcZ2rm6kJucip2bKwa9nsKcXKwcHVBKYVE87MqlQQAOXh5kX03ApWGAyfLbu7mQXSZ/dkoadhXy27u5kpOcioN7cf7cXKwdTTfHrSqH1+3g+Ebj9u8T5E9mUun2n5mUhoNb+ffg6O5CZnKZMslpOBaXcXJ3pnHn1iilqNskAKVT5GZkYe9844bC/bxsN2t/2g9A05D6JMSVZkuMT8fD89r7r42NFV17hrB7WyRuHk5cjU1h9DDjaSIxIZ0nRszhi0XP4+5R/WNATSTk5JfrtfGysyYxN79SuQ7eLjwW7M+YLcfKHfe/PXmJb09eAmBa56ZcyrwJGv8VxMYlU69u6ZBcPx83rsanEns1he6dm5cur+vGzr2mmci+ePE6li9fD0DLlo2JK3Pjl7i45CqHsXl7G9+Do6M9gwb15NixM9xzz+2sXLmBr79+G4A2bZqRn19AamoG7u43fi5ZYl4Bnnbl6w1JeZWP/WHuzjwYVI9xe8vXG+wtLZjZPphvTl/k5F8MY7xRbtV6z/Kl21m10ljPC24RQFxc6fE/Pj4VT6/yx87jR84TdfISd93xBnq9gZTkTMaMms3878vPM75zYHvGPfM5T/5He4Fu5p4ZU6nuELjqfFR/tSdoVTzWAZ01TQst/uenaVomMBXYqmlaC+AuwLbMc88BTsANrZGfSs/Ez8EOHzsbLJWiV11P9lQYh2y8a0sjJkdEkVZQeCPj/GN1GgSSG59AbmIShqIiEg6E4x5avfHcriEhpEaepDA7m8LsbFIjT+IaEnJjAwM97+nGpK9eZtJXL9O6Wwv2bzyIpmmcP3kBOwe7fzTXp+x8oWN7TuDj730jIlfpdEYmfvZ2eBdvO7fV9WRfNcew6wCn4vknDRztaeDoQESyaYcYuDQMIDs+gZzibefKvnB82rQqV8a7bSsu79oHwNWDh/AIbopSivyMTLTiq63ZCYlkxydg73VDRqlek3ujADLjEshMSEJfVMTFPRHUD2tZrkz9sJac3WGsrF/cfxifkCZmH+vdZmAPHp4zkYfnTCSoUytObj2ApmlcOX0eGwfbksbNnxzdnLG2s+XK6fNomsbJrQdo1MH4PoM6tiLmmPEiRkpsAvpCPXZ1Kt2r5rq6d3hXvlk+nm+Wj6d7rxasXxuBpmlEHruIg6Mt7hUaQDk5+SXzgoqK9OzbdQr/Bl40alyX1Vun8MNvk/jht0l4ejnz1dIXbnjjB+BkSib1nWzxdbDBUqe4w9+THbHl992mLg5Mah/E+J2RpOaXHvd1CpytjftukLM9jZ0d2Bd3cw0PAli38RAjhxivXXZoE0RGZg5xCWls3H6UPt1b4eLsgIuzA326t2Lj9qN/82rXxwMPDGT16k9YvfoT+vTpxKpVW9A0jSNHTuHkZF+pAVRUpCeleE5cYWER27YdpHFj40WWunU92bvXmPvs2Uvk5xfi5maaeaKn0zOpV6becLuvJ3viK9cbxrdsxKSD5esNlkoxNawZGy4nsD2u6iF/N9qtWu8ZNqInS36cxJIfJ3Hb7a34dc1+NE3j+NHzODra4eFZ/u8/dHgPft/6Lr9smMbXC1/CP9CrpPETczGhpNz2rccIbGC+GzqIG6+6PUA7gO+VUjMxNobuBR6qsP5JpdRCjPN/egFlxx3dD8ws/n9v8bINwHPABwBKqVBN045g7AH6s191VIUcF4EJwM9Kqfs0Tbshs/MNGsw9eY732oegU/Db5QQuZuUyqrE/p9Oz2JuQwpimgdhZWPBmG+MUpYTcAiYfMl4xm9OxBfUd7bGz0LGsVzs+PB5NeJLp5tEoCwuCHhjO8dkfoxkM+HTrioOfL+dXrcEpMACP0NZknL9A5GfzKMrOIfnoMS6s/oX2U6dg5eiA/6CBHJr2LgABdw3EysRXx0M6BhO5P4opD07H2taaB18ZXrJuxhMfMOmrlwH4+cs1hG8+RGF+Ia8Pm0KXAZ0YOOpOtv20k2N7TmBhYYF9HXsemjjCZNkNGnwWdZYZYS3QKdgQG8/F7BweDvLnTHoW+xJTaFLHkTfbNMfJ0pJOnm48HOTPmN2HsdApPupgbGzkFBXx3vEzmPoCm87CghYPD2ff+3PRNAP1e3TBqZ4vp378BZcG/vi0bY1/j64c/vJ7Nk94E2tHe9o+MxqA5NN/cPqnteh0OtDpaDlqpMl7VnQWFnR4dBibZ3xmvA12r8641PflyPK1uDf0p367VgT16sKuzxawatxbWDs60H3sYyXP/+m5yRTm5mEoKuJS+DF6T3qu3B3kTKFBWDDnwiP55ql3sLKxpt/zD5SsW/jCezw8ZyIAfZ4aZrwNdkEBDdoG0yAsGIAWfTqxfu4Svn/+XSwsLej/woMmbeB16t6MfbuiGHnXTGxsrXn17WEl60YPm8U3y8eTl1vAa+O+o7CwCINeo02HIO4earpb1VdFr8EHEWeZ27MFFjrFmnPxnMvI4ckWAUSlZLLjSgpjQxtgZ2nBzK7G3pL4nHzG7zyJpVJ81dt4kSm7sIjJ+06jN8PF8QVzn6d75+Z4uDoRvf9Tps5aiVXxRZWvF23i9y2H6dcrlMidc8jJzefJCV8CkJqezbuf/MyuX6YBMOPjn0g18R3gAHr2bMf27eH07TsGOzsbZswYV7Ju8OCxrF79CQUFhTz++FsUFuoxGPR07hzKsGHGIZOvvjqaN974lO+/X41Sipkzx5ls29dr8PGJc3zQobTecCErl0eb+HM6LYs9CSk83TwQO0sL3m5rrDfE5xXwengUvXw9aO1WB2crS+6sZ5xCPfNYNNEZpvsb3Or1HoCuPVqwe2ck9/R/C1s7a96aWlpNHTlkBkt+nPSXz587exUXL8SjU4q6vm689uZ/8w5wID1AAKq6Y8OVUuOBP2sKX2uaNufP22Ar4xFmLnA78Of4qUWapq1USl0AvgMGYLzIPULTtOjiGxh8BjTH2BDboWnaU0qpzsACIBHYAjykaVpg2dtgF88/WgzcpWna2b+IrfWuwRC4m8Hm/l0BGLNrm3mD1ND8brcBsCn2V/MGqaE+fgPot36XuWPU2Pp+3Ziwv+JI1FvHhx1vZ9rhTeaOUSNvtOkDwPxT682cpGbGNDN+71Nc7hozJ6kZH7u7AWi3bKeZk9RM+PDu2Pmb7qLN9ZYbs5TSasCtqAm3rbs16w3bBhrrDbd6vSezcLOZk9SMk1VvqN6IKbP7X/R6k12ieSio3035mVT7fq+aps0CZlVY5lj8v4axN+daPtM07e0Kz03C2CNU8ffspfwQt8nFy78Hvi9+fBgIrm52IYQQQgghBFgo88/fMjf5HiAhhBBCCCFErXHDv/FP07TAG/07hBBCCCGEEKI6zPuV50IIIYQQQgiTkeFf8hkIIYQQQgghahHpARJCCCGEEKKWkNtgSw+QEEIIIYQQohaRHiAhhBBCCCFqCekBkh4gIYQQQgghRC0iPUBCCCGEEELUEvJFqNIDJIQQQgghhKhFpAdICCGEEEKIWkLmAEkPkBBCCCGEEKIWkR4gIYQQQgghagnpAQKlaf/piVD/6TcnhBBCCCFuGrdE0+KXmN9MVj++y7//TfmZ/Od7gHbGrTN3hBrp7jMQgI8jN5g5Sc2MC7kDgIe3bzdzkppZ2LMnqy/+Zu4YNTY4oD8zjmw0d4wamxTal/mn1ps7Ro2MadYPgKd2bzVzkpr5omsvABp+fmvuu+ee6QlAVNpaMyepmeYug4Az5o7xLzTBzn+EuUPUWG7MUv4XfWseex4KMh57Vp7/3cxJamZogzsBCP52h5mT1MzJx3qYO0K1SQ+QzAESQgghhBBC1CL/+R4gIYQQQgghhJGF9ABJD5AQQgghhBCi9pAGkBBCCCGEEKLWkCFwQgghhBBC1BI6JTdJlh4gIYQQQgghRK0hPUBCCCGEEELUEtL7IZ+BEEIIIYQQohaRHiAhhBBCCCFqCfkiVOkBEkIIIYQQQtQi0gMkhBBCCCFELSFfhCo9QEIIIYQQQohaRHqAhBBCCCGEqCXke4CkB0gIIYQQQghRi0gP0DVomsbST37m+P4orG2seey1EQQ0qVep3E9f/cre9eHkZOXw2e8zS5Yv+3QVpw9HA1CQV0hGWiZz180waf5d3/zIxUORWNpY0/u5B/FsVL9SuYSzMWyZu4iigkIC2obQbfQQlFLsX7KW8wePo5TCztmJ3s8/iIObs0myp504wcUffkAzGPDq1g3f/v3Lrc84c4aLP/xATmwsQU88gXtYGADZly5xYfFi9Lm5oNPhN2AA7u3bmyRzWZqmsebznzh1MAorGyuGTRhJvcaVP/vfv1tHxMaD5GblMG3N+yXLd6zcyoHf96Gz0OHo7Mh9L43A1dvNpPkPfL+S2MPGbafr0w/h3rBy/uRzMez6/H/oCwrxaxNCh1FDUUqRn5XN9jnfkpWYgqOnGz1fGI2No71J82/96kfOR5zE0saaO8c9gHcV2358dAy/f7KYovxCGoQF0+sJ47YPcGjtdo6s24nOQkeDdiH0HDXYZPmTj0fyx5LloBmo270rAQPvLLfeUFhI1Nffk3kxBksHB0Kefhw7Dw8MRUWcXrCYzAsXQSkajxyGa7OmJssN0KO+K292C0KnUyw/eZUvDl8qt35kSF0eauGLXoOcQj2Ttp0hOjUHK51ies8mtPRyxKDB+3IlpgAAIABJREFUO7ui2X8l3aTZwbjtfD1rFRF7orCxtWbs5OE0alb+uJ+fV8D7ry0kLjYJnU5H++7BPPzsoJL1uzYdYdlXG1AKAhv78tLUB02af/r0+WzfHoGtrQ0zZ44jJCSoUrmHHnqNhIRUbG2tAfj223dwd3fhypUEJk6cQ2ZmNnq9gQkTHqFnz3Ymyf7FB0/Sv3cbEpMzaNf3lSrLfPT2I/TrFUpObgFjXprHkRMXAHhgaA9eff4eAGbOXcXilTtMkrkiTdPY8OWPRIefxMrGmrtefIC6QZWPPVf/iGHN7MUUFRQS1C6YO540Hnvizl7mt89+oKigCJ2FjjufGYZf0wCT5l837ydOHzyJlY0VQ156AL8qzl0bvl/LkU3Gc9dbqz4oWV5UUMTKDxcR+8cl7Os4MPy1R3D1cTdJ9m5+rrzWqREWSrHyTBxfHyt/7Lm/aV1GNPfFoGlkF+mZsvsPzqbl4Otow9r/a8eF9FwAjiZm8PaeaJNkNie5C5yJe4CUUtuUUqY5mv5Lx/dHkXA5iRmLJ/HwhPtYNGtlleVad/l/9u48Lqrq/+P467DvzLBvIiouiAsqouZumZlWtqrtVmr502zxW5pZqS22fMts18r8pubW4la5lfsCuCLiDirIzgz7Pvf3xyAwMCqSXDTO8/Hgocw9M7xn+Nwzc+4599Ke6d+8UOv2URNH8OZ3U3jzuykMuq8PXft2aujIJs4fOEZ2chqPfPEGA54dxbb5y8222/7NcgY8N5pHvniD7OQ0zh88BkCXEbcy6pNpjPx4KkHhoUSt+EOV3IrBQMLSpbR9/nk6zZxJZlQUBRcvmrSxdXOj1ZgxeEREmNxuYWNDqzFj6DRzJu0mT+bc8uWUFRSokru641FxZCSl88rC6dz/wkh+nbfSbLuQnqFM+uzFWrf7BQfw/Ocv89I3r9Kxb2fWf7umoSObSDp0jNyUdO799E16jR3N3u+WmW2359vl9Bo3mns/fZPclHSSDhlrJ+a3Tfh2aMt9n76Jb4e2HF29Uc34xO8/hi45nae+nsHg/xvJ5q9WmG23+esVDJ4wiqe+noEuOZ2EA3EAnD9ykjP7Ynh83qs8+flrdB8xSLXsisHAycU/0fnFiUS8/Sap+6LITzKt/+Qdu7BydKDnnNk0u/1Wzq78FYCL23YCEDH7DcKmTOb08p9RDAbVslsImNmvNWPWxzDkpyjuau1FsNZ04LvmZBpDl+9n+Ir9fHPwAtN7twJgVHtfAIYu38/ja4/w2i2taIz35/27j5N8IYOvVk1jwtQH+fqDn822G/HIAL5YMZWPf3yJuMMJ7N9trJ2L59P5edEW5iyYyGfLXuHpF9UbOANs376fhISLbNz4DbNn/x9vvfXVZdt+9NHLrF49j9Wr5+HurgHgq69WMHRoH3777VM++eQ/zJx5+ftfbz+u3MY9j8+57PYhA8NoFeRDh34vMnHqAua98zQAWldHpr9wH/3unkHfu2cw/YX70Lg6qhXbxJnoY2RdTGfCghncOWkkf3xhvu/548sVDJs0igkLZpB1MZ0z+431s2Xhavo+PJSxn79K/0fvZMvC1WrG52TUMTIupvPS968zYvIo1nxu/r2rXY8OPPvpS7Vuj96wBzsne15eOIPe9w5gw/drGzoyYOx7Xu8VzPiNR7nrl2jubOlJK41p37PubBojftvPfasP8P2RC7wS0bJy24XcIu5bfYD7Vh9oEoMfyUgugbuMQzuP0mtIOEIIWoUGUZBXiD4zp1a7VqFBaNxdrvhYkVsOEnFrl4aKalZ8ZAxtB0QghMCnbQtK8gvJzzI9opqflU1JYRE+bVsghKDtgAji98UAYONgX9mutKik8sh4Q8uLj8fOyws7T08srKxw694d3eHDJm1sPTxwCAiAGpnsvb2x8/YGwEajwdrFhbLcXFVyV3dsdwxdB3dHCEHzkCAK8wvJyax9NLt5SBAu7rVn1YLDWmNTcWQ2MCSI7HR1j4RfiDpCy37G2vFsY6ydAp1phgJdNqWFRXi1aYkQgpb9IrgQdcR4/+gjtOrfA4BW/XtwvuJ2tZyJjKH9QGN+v7YtKM4vJK9G7edlZVNcUIRfO2Pttx8Ywel9xpyH/9xJxP2DsbK2BsBB46xa9pyzCdh7eWHvZax/7x7dyThk+vqlHzyCzy29APAM74ou7jiKolBwMRlt+3YA2Li4YOVgb5wNUklnLxfOZRdyIaeIUoPCutNpDG5hevQ3r7S88v8OVhYoFcvQg7UO7ErSAZBZWEpuSRkdvdR73S+J3H6UAUO7GfvDjs3Jzy0kK8O037e1s6FjuHFWxdrailZtA8hMM9bXxtV7ufOB3ji5GD98adzUfQ5btuxlxIhBCCEIC2tHTk4+aWlZdb6/EJCXZzxolJtbgJeXejPPuyKPk6XPu+z24bd3Y+nPOwCIPHgaVxcHfLw0DO7fmS07YtBl56PPzmfLjhhu799ZrdgmTuyNoeMgY98T0K4FRfmF5Nboe3Ir+p6AEGPf03FQBCf2GPdxIQTFBUUAFOUX4azSqotL4vYcpcutxveuwJAgivLMv3cFXua9K27PUbreZjwwGdq3M2cOnURRGv5ck44ezpzPKSQx19j3/HE2nUGBpn1PfrW+x97assEz3egshHpfN6oGWwInhHAEVgABgCUwu8b20cBrgADWK4ryasXtecA3wEBAB4xSFCVdCNEK+ALwBAqAsYqiHG+o/PqMHNy8NJXfaz016NOzrzrYqSkzJYuM5ExCura+3hGvKD9Lj5OHtvJ7R3cN+VnZJsvY8rOycXLX1Gijr/x+75K1nNgaia2DPffMmqRK7hK9Hhu3qjddG42G/Pj4a36cvPh4DGVl2Hp6Xs94dZKdmY3Gs+q113hoyM7MNvuGcTVRf+6lXfeQ6xnvqgp0ehzdq/I7uGsoyNLjoK3KX5Clx9GtWu24aSjQGWunMDu3sq2D1pWiHHUHoXmZ2Th7VGVz9tCQl5mNU7Xaz8vMxrla7Tu7G9sA6C6mk3jsDDsXr8PKxor+Y0bg01qdZSjFeh12blWvva1WQ85Z0/ov0euxrWhjYWmJpb09pXn5ODULIOPgYbwiwinO0pGXcJ6iLB0uLVuokt3H0YbkvOLK75Pzignzrt1fPtbBj6c6B2BtKXh0tfGDX1xmPoODPFh3Kg1fJzs6eDrj52TLkTR1aycrPRsP76q6cPdyJSs9GzcP8/1+Xm4hUTtjGT6qL2CcAQKYOvYzDOUGRo0dQtde7Ro+eIXU1Ex8fDwqv/fxcSc1NdPsQOa11z7FwsKC22+/hQkTRiKEYOLEh3n66TdYvHgdhYVFLFz4tmrZr8bPx43E5MzK75NSsvDzccPPR0vixapBXlJyFn4+WnMP0eByM7Nx8ayqHxcPDbmZ2SYDmdwafc+lNgC3j72PpW98xebvfgNF4YmPaq8QaEg5mXpcq+f3dCXnGt67jPc3vvaWlpbYOdpRkJOPo6tTg+S9xNvRlpT8qr4nJb+YTp61Dz6MDvHlidAArC0seOrPqgOr/k52/HxPV/JKyph3IIH9qbUPdkv/Pg05A3QHcFFRlM6KonQA/ry0QQjhB7wPDALCgO5CiBEVmx2BA4qidAW2AW9W3D4fmKQoSjdgCvCluR8qhBgnhIgWQkTPnz+/3uHNHrWox0g28q+DdOvfGQtLdSfb6pLffJuqRj0fuYsnFsymdb9wYv5QaU31VTLVRYlez5nvv6flk08iLBphktPsS3/txXNgczSJJy/Q/0H1lmDB5X4FNYvH3D1vjEM95uq6dgldfv8wlBsozivg4Q9fot+TI1j7wUJVjmJeLlbN8Jd7fj59b8FWq2H/rPc4/dMKXIJbItTsd8z8+s29bD8evcjAJZF8sCee/+sWCMDKuGRS8otZ/WA3ZvRpxYGUbMoM6l+l6Fp+zeVl5Xw8YzHDHuqLj7/xaLOh3EDyhQze/moCL7/9KF+8s4K83MIGSltbnfZd4KOPprB27ecsWTKH/ftjWb36bwDWr9/OvffeyvbtPzB//lu88srHGFRcRnkl5vpQRVHMPj+1dte6/ODaXc/l+6f9v+9k8Nh7mbxoFoPH3su6uUuvf8YrqGv9XP4B/uH966muP+GnuGTuWBXFx9FnGd/ZeFArvaCEW1fs4/7VB3g/8iwf9A/BsQnMEFmo+HWjasiLIMQAHwkh3gfWKYqyo9qO0B3YqihKOoAQYgnQD/gNMACXTlhZDPwihHACbgFWVnsMW3M/VFGU+RgHSwDKjpT1dQ7816872bFuLwBBbZuRlVY1G6JL16PxuPYj+JFbDvHIi/dd8/3qI+aP7RzbtBsAr+BA8jJ0ldvyM/U4ak3zO7lryMvUX7ENQJu+4ax/52siRg1roORVbLRaSrKqjuaV6PVYazRXuIepssJCTnz2GQH33INzy5ZXv8N1snvNDvb9vgeAZm0D0adXvfb6DD0u1zhzeOrACf76aSPPfjQJK5uGv1bJ8Q3bOLnFWDserZqTn1mVvyBTj32NunCoMVuYX22GyN7VmQJdNg5aVwp02di5NPwyoIPrtxOzyfj6+wQHkptRlS03Q1/rAh5O7hpyq9V+bqa+cobI2d2V1r06I4TAt01zhIWgMCcPB9eGfx62Wi1FWVWvfbFOj22N+rfVainOMs4UGcrLKS8sxMrRESEErUc/VNlu/zsf4ODl1eCZL0nJK8HXqapb9nWyJa2g+LLt155KY3a/1vyHE5Qr8PauM5XbVt4XVnlSckP7feVONq7eB0Dr9s3ISK2qi8y0bNw8zff7X763Et9mHtw9ul/lbe5eGtp0CMTKyhJvP3f8mnuSfCGd1u0DGyz/kiXrWbFiAwAdO7YmJSWjcltKivnZH29v44DNycmB4cP7c+TISUaMGMSqVRv59tuZAHTp0o7i4hJ0upzKc4QaU1JKJgG+Vcua/H3cSE7VkZScRd9eVbPk/r5u7NgTp1qu6HXbOfinse/xbRNITnpV/eRk6HGqMXvi7GHa9+RkVPU9R7ZEcvv4+wEI6dOFdZ/+1NDx2btmB1EV+QPaBJJdPX96Ns5udX/vcvHQkJ2uw9VTQ3l5OUX5Rdg7N/wFcFLyi/FxrOp7fBxtSSsouWz738+m88YtrWEHlBoUsovLADiWmceF3EKCXOyJzbz8ckzp36HBBmeKopwEumEcCL0nhHij2uZrOSSgYMypVxQlrNrXdV8XNOjePpUXLujStyN7NkSjKApnYhOwd7S75uVvKefTKMgroFVo0PWOalbHof0Y+fFURn48lRYRnTixNRJFUUg5EY+Ng12tD4GObq5Y29mRciIeRVE4sTWSFhEdAdBfTKtsFx8Vg8bfW5Xn4BQURFFaGkUZGRjKysiKikLbuW7ruQ1lZZz66is8evXCPVzda23ccndfXvz6FV78+hVCb+nIgU1RKIrCubgE7B3tr2n5W9LpRH7+dAVPzBqLk1adcwjaDenP3R9M4+4PphHYvRNntxtrJ/1kPNYO9ibL38C4tM3azpb0k8baObs9kmbdjRf6aBbekTPbjB8oz2zbR7Pwhr8ASJdh/Xh87qs8PvdVgnt24tjfxvwXT8Rj62hnsvwNwMnNFRt7Oy5W1P6xvyNpVVH7wT06cf7ISQCyktIoLy3H3qVhl3Bc4tyiOYWpaRSmG+s/dV8UHmGmr59HWCdSdhs/sKRHH0DTri1CCMqLSygvNg44smKPISwtcPT3UyU3wJG0HIJc7QlwtsPaQjA82IvN8ZkmbYJcq84tHNjcvXKQY2dlgb2V8e2oT4CWcoPCaZ06FzC588E+zF38MnMXv0yPfh3Y+sd+Y38Ycw5HJzuzy9+WfP0H+XlFtS5y0KN/B47uNw7kcvR5XDyfjrd/w14F65FHhlVezOC223ry229/oSgKhw4dx9nZodYAqKysnKyK81JKS8vYujWK1hVLPH19Pdmzx7g06MyZCxQXl+Km8nkol7N+0wEevt+41DCiSzA5uQWkpOnZtO0wt/XthMbVEY2rI7f17cSmbYev8mjXT/jwfoz9/FXGfv4qbXt2IuYvY9+TeDweO0e7WufxOFf0PYnHjX1PzF+RtO1p7Huc3Fw5F2M8CT/h8Enc/Bp+CXfPu/sy6ctXmPTlK4T06sjBLcb3rvNxCdg62l3Te1dIzw4c2BwJQOyOw7Ts3FqVGaCjGbk0d7XH38nY9wxt6cnf5037nuYudpX/79/MjXM5xr5Ha2ddeZ5KgLMdzV3sScwtavDMUuNryHOA/IAsRVEWV5zX82S1zfuAT4UQHhjP8xkNfFaxzQJ4AFgGPAzsVBQlRwgRL4R4UFGUlcK4R3VSFKXBermOPUOI2RvHaw+/i42tNWOmjq7cNvPpj3jzuykArPxqLZFbDlBSVMp/HphJn2E9uGeM8bK1+7YcoPugLqpdQKC65t1COX/gGEsmzMLK1ppBE6suxbr8pTmM/HgqAP3Hj6y8DHZg1xACu7YHYO/iNeiT0sBC4OzpRv/xI1XJLSwtCRo9mhNz56IYDHj27o2Dnx+Jq1fj2Lw52rAw8hISOPnll5QXFKA/coSkNWvoNHMmWdHR5J48SVleHhm7jbMZLceMwbFZ7ct4NqR2Ee05HhnH+0++jY2tDQ9OqaqdT579gBe/Nl7idf2CNRz6ez+lxaW88/CbdL+jJ7c/PpT1C9ZQUljM4tkLAdB4aRkza6xq+f27hJJ4MJZfJs/Eysaa3s9V1c6aV97j7g+mAdDzmZHs+nIxZaWl+Ie1xz/MWDsd7hnMtrnfc+rvPTh6aBnw4tOqZQdo0a09Z6Nj+e7ZWVjb2jBk0iOV2/73wvs8PvdVAG579iHjZbBLSmjRtT0tulXkv60nGz5byg+T3sPSypKhLzyq2j5sYWlJm0dHcvjjeSgGA759bsHR34+zv67BJag5Hl0649uvN3ELFrJ36gysHB0IHf8MACW5ORz+72cIC4GtRkP7Z8aokvmScgXe2nGaRXd1xEIIVh5P4ZSugBe6BxGTnsuWhEwe6+hH7wAtZRVHXadsMZ7G6W5vzaLhnTCgkJpXwkubG+z0zivq1juE/bvjePb+97C1s+b5GaMqt73w6H+Zu/hlMlL1rFy4mYAgL156/BMAhj3Ym8H39KRLz7Yc2neCiSM/wMJS8OSku3BR8Ypk/fuHs21bNIMHj8Pe3pZ3351cue2ee55n9ep5lJSU8swzb1JaWo7BUE6vXmE89NDtAEyd+jSvv/45P/ywGiEEc+ZMVq32F302ib69QvDQOnN63+fM/ngV1tbGjyjfLt7Mn38dZMjAMGJ3zKWgsJjxU74BQJedz3vzfmXnWuP5Su9++gu67HxVMtcU3L09p6Nj+eKZWZWXwb5kwcT3Gfu5se8Z+n8PsfaTJZQWlxAc3p5W4ca+Z9jzo9j4zc8YDAasrK0ZNmmU2Z/TUNpGtOdk1DE+fmo21rY23PfSw5XbPpvwAZO+NL53/fntag5vNb53vf/oG4QP6cWtjw2l2x09WfXBYv47Zjb2zg6MmvaEKrnLFXhnz2kWDOmAhRD8eiqF0/oCJnZpTmxGLn9fyOLhEH96+WmMfU9JGa9tPwFAuLcrk7o2p0xRMBgUZu4+RXZJmSq5G1MjfCytFyGEG8ZVYUFAAvCQoii6Gm3CgK8AF6AceEdRFPOXPq5+v4Za2y6EGAJ8iHFJWynwHPARMEVRlGghxMPANIyzQb8rivJKxf3ygE+AO4FsYGTFRRBaYHyCvoA1sExRlFlXiXFNS+BuJH19jMvNPo1V9xLC18vkUOMb6uPbtjVykvr5X//+rD6nzqW/G8I9zYfy7qFNjR2j3l4LG8z84xsaO0a9jGs3BIBnd/3dyEnq5+veAwFo+eXNue+endAfgDj9ukZOUj8hmuHAycaO8Q+0wT5w9NWb3aAKz//Ej6dvzr7nsWBj37Mq/s+rtLwxPdDCePC4/feN83ec/qljT/WDG+Vk2KuITF+v2plyEZ7D6v2aCCE+wDiZMkcIMRXQXrpoWrU2bQBFUZRTFZMv+4EQRVH0Zh6yUoPNACmKsgGo2YsMqLZ9KWD2DD9FUWYAM2rcFo/xwgqSJEmSJEmSJNXDTTFKM7qHqrHDImArYDIAqjjl5tL/Lwoh0jBeMfqKA6Ab+QINkiRJkiRJkiTdpKpfnbnia9w13N1bUZRkgIp/r3hlHyFEBGADnLlSO2jYq8DVi6Io6pxtLEmSJEmSJElNjJrnANW4OrOZLGIz4GNm0/Rr+TlCCF/gR+AJRVGuev3+G24AJEmSJEmSJEnSv5+iKLddbpsQIlUI4asoSnLFACftMu1cgPXA64qi7K3Lz5VL4CRJkiRJkiSpibiJ/hDqGuDS5QSfAFbXbCCEsAF+Bf6nKMrKuj6wHABJkiRJkiRJknSjmQMMFkKcAgZXfI8QIlwI8W1Fm4eAfsCTQohDFV9hV3tguQROkiRJkiRJkpoIIVS7CvY/oihKJnCrmdujgWcq/r8YWHytjy1ngCRJkiRJkiRJajLkDJAkSZIkSZIkNRE30d8BajByBkiSJEmSJEmSpCZDzgBJkiRJkiRJUhOh5t8BulHJGSBJkiRJkiRJkpoMOQMkSZIkSZIkSU2EnACSM0CSJEmSJEmSJDUhQlFujmuB19O/+slJkiRJkiRJN4ybYnLlSNY61T4fd3IbfkO+Jv/6JXARK3c2doR6iXywDwBPbt/WyEnq54d+/QHYk7a+kZPUTy+vYTyzc2tjx6i3b/sM4Pk9fzd2jHqb12sgo/7e3tgx6mXZwH4ALDy5oZGT1M+YNkMAOJBxc+67XT2GATd33zlg/a7GjlFvW4f15sfTN2ftAzwWPAT7wNGNHaNeCs//BMADf92cfeeqQca+U1e8rpGT1I/WdnhjR6gzixtySKIuuQROkiRJkiRJkqQm418/AyRJkiRJkiRJkpGcAJIzQJIkSZIkSZIkNSFyBkiSJEmSJEmSmgj5h1DlDJAkSZIkSZIkSU2InAGSJEmSJEmSpCZCTgDJGSBJkiRJkiRJkpoQOQMkSZIkSZIkSU2EnAGSM0CSJEmSJEmSJDUhcgZIkiRJkiRJkpoICzkFJGeAJEmSJEmSJElqOuQMkCRJkiRJkiQ1EXICSM4ASZIkSZIkSZLUhMgZIEmSJEmSJElqIoRQGjtCo5MDoMvo6a3h5S4tsRCC1WdT+d+JRJPtD7f24+6WPpQbFPTFpcyOPkVKQTEAkzoG0dtXixCCyFQ9/z10VtXsuqNHSVi2HMVgwLtvH/yHDjXZnnPyJAnLl5OfmESbcWNx79YNgPzzFzi7ZAnlhYUICwv8h92JR/fuqmYHUBSFJZ/+ypG9cdjY2vDMa6MJahtQq92q+b+ze0M0+bkFfLNxTuXtmak6FryzlIK8IgzlBh58dhide7VXLX9WzFHO/rQCRTHg07cPze68w2R79omTnFm2gvzEJNqNfwbP8G6V21J37eH8ut8BCBx+J969e6mW+5L0I7EcX7oCxWAgoF9vWg43zW8oLSVmwQ9kJ5zHxsmRzs89g72nBxd37yPhj02V7XITk+j11mu4NG+mav6c2KMkrliGYjDg3rsvPneY1n/eqZMkrlhOYVIiQU+PQ9utm8n28sJC4t56A9ewLjQb/bCa0QFj/W+e/zNn9h/D2taGYZMfwSe49muYcvo86+cuobSklFbd2nPbuPsRQvDb+wvJSkoDoCi/EDtHe56a96pq2RfN/ZVDe+KwsbPhuemjaVFj3y0uKmHu64tIS8pEWAi69Qll9HPDAYg7dIb/ffob588k8/zMx+gxsLMquS+52fvOCE8NE9u3xFLA+gupLD2TZLL9wRZ+DGvmTbmioC8p5YMjp0ktLCbYxZEXO7TEwcoKg6Kw+HQifydnqJ5fURQ2fvMzp6ONtX/Xi4/ga6b2k0+dZ80nSygrKSU4vD23jzfWfsqZRP74YjllJWVYWFpwx4SH8G/bXJXsX384nqG3diE9M4fwwa+YbfPfmU8wZGAYBYUljHv5Kw4dTQDgkQf6MXXSCADmfPYbS1ZtVyVzTbmxR7m4YhkoBrS9++I1xLT+80+d5OLK5RQlJRL49Dhcu1b1nTETxmHn7w+AtdadoAkTVc2uKAofv/8be3bEYWtnw4zZo2jXvvbnhheenU9GRg7l5QbCurZkymv3YWlpwakTF3l/9ioKC4rx8XNj1pxHcHSyU/U5SOqSAyAzLIBXurZi4vajpBWUsOi2MHZczCQ+t7CyzQl9Pk9sPkRxuYH7W/owqVMQ0/eeoKO7M508XHh440EAFgzqRFdPVw6kZ6uSXTEYiF+6lPYvvoiNVkvMO++i7dwZBz+/yjY2bm60GjOGixs2mj5vGxuCnxqDvbc3JXo9R95+G01oKFYODqpkv+TI3jhSEzN4/6fXOHPsHP/77yremP9CrXZhvdtz2319ePXhd01uX7NoExEDwxh0b2+S4lP4+JUF/HelOgMgxWDgzJKf6PDyC9hqtRya/R5uYZ1wrPb627q70fapJ0ncsMnkvqV5+Zxfs46wGa+BgEOz3sUtrBPWjo6qZL+UP+7Hnwj/z2Ts3LTsmfkeXl064eRflT9x+y6sHBzo98FskvdGcXLlr3SeMBa/W3rgd0sPAHIvJHFw3leqD34Ug4ELPy0lePKLWGu1nHjvHVw7dca+2utvrXWj+RNjSN20wexjJK9ZjVObNmpFruXs/mPoLqYz/psZXDyRwIavVvDEf1+u1W7Dlyu4Y+Io/NoGsfKtrzm7P45W4e0Z8eqYyjZbvvsVWwf13sQP7YkjJTGDT5a/xunYc3z30SreXlB73x0+egCh3VpTVlrG289/xaE9cYT1CsHDW8uz00ez/qetqmW+5GbvOy2AyaEtmbIvlvSiEr7u05ldqVmcy6t63zqVk8/4nYcpNhi4O9CH8e2CmHXwBEXl5bx76BRJBUW429owv09notJ15JWVq5Yf4Ez0MbIupjNhwQySTiTwxxcgS5qxAAAgAElEQVQreOqT2rX/x5crGDZpFP7tglj25tec2R9HcHh7tixcTd+HhxIc3p7TUbFsWbiax+c8r0r2H1du4+tFG/j2kwlmtw8ZGEarIB869HuRiC7BzHvnafrdMwOtqyPTX7iP3sOmowC717/D+k370Wfnq5L7EsVg4OKypbR4/kWstFrOzHkHl06dsfOt1ne6uRHw+BgyNtfuOy1sbGg9/U01I5vYs/M4F85lsHLdNGKPnOeDt3/m+6WTa7V756PHcXSyQ1EUpr20iL82Hmbw0C68+9YKJr18F13DW7H2130s/uFvxk8cauYn/TvIc4DkOUBmhbo5k5hXxMX8YsoUhY0X0unn727SZn96NsXlBgBisnLxsrc1blDAxtICawsLrC0tsBKCrKIS1bLnxcdj5+mFnacnFlZWeHTvju7QYZM2dh4eOAYEIITpLmDv4429tzcANhoN1s4ulObmqpb9koM7j9L7jnCEEASHBlGQV4g+I6dWu+DQIDQeLrVuFwIKC4oAKMwvQuvh2uCZL8k9G4+dlxf2Fa+/Z0Q4WQfNvP7NAoxBq9HFxqIJDcHayRFrR0c0oSHojsaqlh0g+2wCDt5eOHgZ8/v26E7awSMmbdIOHsG/j3Fmyrt7VzKPHUdRTKfTk/dF4dsjXLXclxQkxGPr5Yltxeuv7d6d7COHTNrYenhgb6b+AQrOnaM0NwfnEPVmDGs6tTeGDoMiEELg364FxfmF5GWZHkDJy8qmuKAI/3YtEELQYVAEp/aa/p4UReH4zoO07286w9WQ9u88St+Kfbd1hyAKcgvR1dh3be1sCO3WGgAraytatA0gM10PgKevG82D/cz+bhrazd53ttM4k1RQRHKh8X3rr4vp9PZ2M2lzKDObYoPxfeuYPhdPOxsAEvOLSKroMzOLS9CVlOJqY61qfoATe2PoWFH7Ae1aUJRfSG6N2s+tqP2AEGPtdxwUwYk9xtoXQlBc8TyK8otwdlOv798VeZwsfd5ltw+/vRtLf94BQOTB07i6OODjpWFw/85s2RGDLjsffXY+W3bEcHt/dWc+wdh32nh6YlNR/67h3ck5bNp32rgb+86a7103gu1/H+XOu7oZ+8POzcnLLSQjvfbnhkuzOuVlBkpLyytHAucS0ujSrSUAEb3a8PfmGNWyS43jhh0ACSEeF0IcEUIcFkL8KIS4SwixTwhxUAixWQjh3VA/29PehtSK5WwAaQXFeNrbXLb93S282ZOiA4yDof1pen6/K4I/7opgb6qehGozRw2tRK/H1q3qTc9Gq6FYr7vmx8mNj0cpK8PO0/N6xqsTXXoObl6ayu+1nhp0GXWfQRsx5g72bNzPi/fN5OP/LODRF+5tiJhmFev12LppK7+30Wop1uvrdN8SnR5bbdV9bbVaSnR1u+/1UqTTYVctv51WQ5HOtH6KdfrKNhaWlljZ21OaZ3q0MmVfND491V8CVKLTY6OtVv8aLaV1fA0Vg4GkVSvwv++BhopXJ7mZ2Th7VNW/s7uG3MzsK7fxqN3mQuwZHDXOuPl5NWzgarLSc3Cvtu+6eWnIusLsd35uIQd2xdKhW+PNuF1ys/ednnY2pBdWHWxLLyrB0872su2HNfMmMr3282vn6oS1heBixUBCTbmZ2bh4VtWPi5m6zs3MxtndfJvbx97Hlu9X8+kTb7Dl+98Y+ORd6gSvAz8fNxKTMyu/T0rJws/HDT8fLYkXs6puT87Cz0dr7iEaVJlej3W1vtNaq6W0ju9dYFwaffq9tzn9/rtkHzrYEBGvKD0tGy+fqrrw8nYlPc183zP52W8YOuBNHB1tGTTYONhsFezDjq3GA45bNh4hLUXd915JfTfkAEgIEQpMBwYpitIZmAzsBHoqitIFWAaYXWQrhBgnhIgWQkTPnz+/nj/fzI2XOV/sjkBPQrRO/FhxjlCAox1BLg4MXxfJsLWRhHu50sXMLEWDUWoHFdc42Vmi13P6u+9p9eSTCAv1S6TmbAJc2wGnvZsP0HtoBJ/88iYvfTiW+bOXYqg46tng/tF5hWburPaRNnMRatSPud9P9Sb6M/FY2trgHOB/ncPVxZWzXUnGtq24dOiIjZvb1Rs3qDrU/1V+BwBx2/cT0k+92R+4tn23vKycz976kSEP9MW7xgx7o/gX9J01KZfpkAb7e9LW1YllZ03PEXKztea1sDa8f/jUP+vK6svs76AObSoa7f99J4PH3svkRbMYPPZe1s1dev0z1pO5WlIUxexsp7ndu8H9w/fddu+8T/C01wl8aizJK5dTnJ52HcNdndku8TL5P/16POv+epOSkjKiI08BMH3WSFYt28UTIz+hIL8IK2vLBkzb+IRQ7+tGdaOeAzQIWKUoSgaAoihZQoiOwHIhhC9gA8Sbu6OiKPOBSyMf5duVO6/5h6cVlODtUHXkzMvBlnQzy9i6e7kyJqQZz26NodRg3PsG+LtzNDOXworlcbuTdXRwd+agmSVcDcFGq6U4q+poUolOj41Gc4V7mCorLOT4Z58ROOIenFu1bIiIZm3+ZSfb1u4FoEW7ZmSlVR190aXr0bjXfSnD9vX7ePmjcQAEdwiitKSUvOx8XLTO1ze0GbZaDcVZVUdVS3Q6bOv4+ttotWSfOFn5fbFOh2tbdY+M27lpKaqWv0inx1arMdvGzk2LobycssJCk/OUUvZF4dtD/dkfML6GJbpq9a/XYV3H1z//7BnyTp8mY9tWyouLUcrLsLCzxf/e+xsqbqX967dzeMMeAHxbB5KbUVX/uZl6nGos5XH20Ji2ydCbLPcxlJdzYs8RnvxkSgMnh40/7+SvNcZ9t2VIMzKr7btZafrLLkFd8MFKfAI8uHNk/wbPWBc3a995SXpRiclKBU87GzLMvG91c3fl0eAAJu85Wvm+BeBgZcmc7u357sQ5jl1hKdf1Fr1uOwf/rKj9NoHkpFfVT06GHid3M7WfWaNNRe0f2RLJ7eON+2tIny6s+/Snho5fZ0kpmQT4Vg30/X3cSE7VkZScRd9eIVW3+7qxY0+c6vmstFpKq/WdpTodVq51r/9L/ayNpyeObdpQdOECtp4NO/u8atlOVv+8D4CQ0GYmszZpqdl4eF7+c4OtrTV9B4Sy4+9YevRqS1ALb+Z9Mx6A8wnp7N6h/u9AUlfjH6IyT1D7MOhnwOeKonQExgMNdmbvMV0uzZzs8XOwxUoIbm/myY5qU9QAbTSOTOsWzJRdx9AVl1benlJQTFdPVywFWApBV09X4nPUWwLnFBREUVoaRekZGMrKyIiKQtu5buuJDWVlnPjyKzx79cI9XN3zN267rw+zF05h9sIpdO3bkV1/RqMoCqdjE7B3sjN7rs/luHtrObbfeFTnYkIqpSVlOGucGiq6CecWQRSlVr3+6ZHRuIXV7fXXhoaiiz1GaX4+pfn56GKPoQ0NbdjANbi0aE5BahoFFfmT90Xh1aWTSRuvsE4k7TR+YEmNOoBbSNvKo5iKwUBK1AF8GuH8HwCH5kEUp6VRnJGOoawMXVQUrp3q9voHPT2WDu+9T+i7c/C//wHcevRSZfAD0G1YP56a9ypPzXuV1j07cfSvSBRFIel4PLYOdrUGQE5urtjY25F0PB5FUTj6VySte3as3J5w6ATu/l64eDT8Uprb7+/DnEVTmLNoCuH9OrKjYt89dTQBByc7tGb23eXzf6cwr5DHJ49o8Hx1dbP2nZecyM4lwNEeH3vj+9YgP092p5q+bwW7OPJSx1a8FhWHvqTqfctKCGZ3a8fGxDS2pWTWfOgGFT68H2M/f5Wxn79K256diKmo/cTj8dg52tU6j8e5ovYTK2o/5q9I2lbUvpObK+diTgOQcPgkbn7qL+G+nPWbDvDw/X0BiOgSTE5uASlpejZtO8xtfTuhcXVE4+rIbX07sWnb4as82vV3qe8sqeg7s6OjcKlj31men4+h1FhPZXm5FJw5g62vb0PGBeCBUX34ceXL/LjyZfoP6sDva/cb+8PD53BytsPD07TvKSgorjwvqKysnN0742jewjhIy8o0nrNnMBhYOH8T9z6o/hVY1WSh4teN6kadAdoC/CqE+ERRlEwhhBvgClyar3+iIX94uQIfHjzDvH4dsBCwNj6VszkFjAsNJC4rjx3JWTzfqQX2Vpa816sdYBz4TNkVx1+JGYR7ubL09q4owN4UHTuTs678A68jYWlJi4dHEzd3LopiwKt3bxz8/Ti/ejVOzZvjFhZGXnwCJ778krKCAnRHjnBh9RrCZs0kMzqa3FMnKcvLI23XbgCCx4zBMVDdK3l17hXCkb1xvDLqXWztrHl62ujKbTPGfMTshcaj2su/XMvezQcoKSrlxftm0m94D+596g5G/d/dLPxgBRtXbAMheOa10aqdVC0sLWn1yCiOfvKp8VK6fXrj6O9Hwm9rcA5qjntYZ3LjEzj2xVeU5ReQdfgI51evpdvst7B2ciRw+DAOvf0eAIF3DcPaSb0rwIHxnJ6QR0ey/6N5KAYD/n1vwcnfj1O/rMG1RXO8unTGv19vYuYvZPsrM7B2dKDzc89U3l934hR2Wi0OXo3zwUNYWhIw8mHOzJuLYlBwv6U39n7+JK9ZjUPz5rh2DiM/IZ74r7+kvKCA7JgjpKxbTcibsxolrzmtwttzNjqWb8bNwtrWhjsnP1K57fvn36+8pPWQCQ+xfu4SykpKaNmtPS27VV244dj2A6pe/OCSLr1COLQnjhceMu6741+r2nenPvERcxZNITNNz2+LNuPX3IvXxnwMGAdRg+7uyZm483w8bWHluUErv/2Tj5aocwnvm73vLFfg06Nn+TAiFAsBfySmkZBXyJg2gZzQ57E7LYvnQoKwt7JkZte2AKQWlTA9Oo6Bfh50dnPB1dqKOwKMHwjnHDnN6Rx1r0QW3L09p6Nj+eKZWZWXwb5kwcT3Gfu5sRaG/t9DrP1kCaXFJQSHt6dVuLH2hz0/io3f/IzBYMDK2pphk0apln3RZ5Po2ysED60zp/d9zuyPV2FtbfyI9e3izfz510GGDAwjdsdcCgqLGT/lGwB02fm8N+9Xdq59G4B3P/0FncpXgANj/fuNepj4z+aCQUF7S2/s/PxJXbsa+8DmuHQOoyAhnnPfGPvO3JgjpK5bTZs3ZlGUkkzS0sUIIVAUBc8hd5hcPU4Nt/QNYfeOOB4Y9h52dta8Prvqd//Yg//lx5UvU1hYwn+e/56SkjIMBgPdIoIrBzqb/jjIquW7ABhwa0eGj4hQNb+kPmF2Pf8NQAjxBPAfoBw4CPwKfIJxELQX6K4oyoCrPIwSUY8lcDeCyAf7APDk9m2NnKR+fuhnXNayJ219Iyepn15ew3hm59bGjlFv3/YZwPN7/m7sGPU2r9dARv3dOH8L459aNrAfAAtPmr/M9o1uTJshABzIuDn33a4ew4Cbu+8csH5XY8eot63DevPj6Zuz9gEeCx6CfeDoqze8ARWeNy75e+Cvm7PvXDXI2Hfqitc1cpL60doOh5vkCtPn8taq9uG/udNdN+RrcqPOAKEoyiJgUY2bVzdGFkmSJEmSJEmS/h1u2AGQJEmSJEmSJEnX1w05JaOyG/n8JEmSJEmSJEmSpOtKzgBJkiRJkiRJUhNxI/99HrXIGSBJkiRJkiRJkpoMOQMkSZIkSZIkSU2EnACSM0CSJEmSJEmSJDUhcgZIkiRJkiRJkpoICzkFJGeAJEmSJEmSJElqOuQMkCRJkiRJkiQ1EXICSM4ASZIkSZIkSZLUhMgBkCRJkiRJkiRJTYZcAidJkiRJkiRJTYQQSmNHaHRCUf7VL8K/+slJkiRJkiRJN4yb4vSalMI1qn0+9rG/+4Z8TeQMkCRJkiRJkiQ1ETfkiERl//oB0At7/2rsCPUyt+cgAG79Y1cjJ6mfLUN7A7At+fdGTlI//X3v5KV9N2ftAHzcYxDjdm5t7Bj1Nr/PAD6N3djYMeplcujtAHxwZFMjJ6mfVzoNBmDZmT8bOUn9jGp1B3Bz9503a3Yw5l8Vf3PWDsADLe7ggb+2N3aMelk1qB8A9oGjGzlJ/RSe/wmAlMI1jZykfnzs727sCNI1+NcPgCRJkiRJkiRJMhJyCkheBU6SJEmSJEmSpKZDzgBJkiRJkiRJUhMhJ4DkDJAkSZIkSZIkSU2InAGSJEmSJEmSpCZCzn7I10CSJEmSJEmSpCZEzgBJkiRJkiRJUhMhrwInZ4AkSZIkSZIkSWpC5AyQJEmSJEmSJDUZcgpIzgBJkiRJkiRJktRkyBkgSZIkSZIkSWoihJwBkjNAkiRJkiRJkiQ1HXIAJEmSJEmSJEnSDUUI4SaE2CSEOFXxr/YKbV2EEElCiM/r8thyACRJkiRJkiRJTYQQFqp9/UNTgS2KorQGtlR8fzmzgW11fWB5DtBlpB+JJW7JChSDQkD/3rQaPsRke3lpKUfmLyIn4TzWTo6ETXgGB093knZHEv/Hpsp2uReS6D1zGi7Nm6mWvbuHhv8LaYmFgN8TU1l2Nslk+wNBftzZzJtyg4K+pJQPY06TVlQMwHvh7WmvceaoLofp++NUy1ydoigs/+xXYvbGYWNnzZNTR9O8Te3X79dv17N3QzQFuQV89uf7Jtui/z7I2h82gIBmrfx5ZsZjasUn7UgsxxYba6dZ/94E31W7dg5/s4jshPPYODnS5f+MtQOQcz6RmIVLKSsqQghB77emYmljrVp2gKyYo5z+aQWKYsC3bx8C77zDZLv+xEnOLFtBXmIS7cc/g2d4t8ptKbv2cH7d7wAEDr8Tn969VM0OxvrZ+d3PnDsQi5WtDbdOfBTPVrXrJ+3Mef76bDFlJaU07xpKn6fvRwjBvqXriI+KQQiBvaszt056FEc3V1Xz7124igsV+fv932N4tKydP+PMebZ/8SNlJaU06xpKzzEPIIQgfs8BDqz4HX1SKne/NwXPVs1Vy34p/x/f/MKpqGNY21oz4qVH8AuunX/zonUc3hJFUV4B03/5sPL2stIyfvloMcmnL2Dv7MiD055A6+2uSvabve+82fMrisL6r37hREXt3P/yI/i3rl07G39Yx6HNURTmFfDmb9Vqp6SMVR8tJunUBRxcHBk17Qm0PurUDkBu7FEurlgGigFt7754DRlqsj3/1EkurlxOUVIigU+Pw7VrVd8ZM2Ecdv7+AFhr3QmaMFG13ABffzieobd2IT0zh/DBr5ht89+ZTzBkYBgFhSWMe/krDh1NAOCRB/oxddIIAOZ89htLVm1XK7YJRVGY98Fq9u08jq2dNdNmjaRNSECtdv+ZsIDMjFzKywx06tqCF6bdi6Vl1Qf1ZYu28tUn61n991totI5qPgWptnuAARX/XwRsBV6t2UgI0Q3wBv4EwuvywI0yAySE0AghJlT8f4AQYt013v8HIcQDDZMOFIOB2P8tI/zlifR97w2S90aRm5Rs0iZx+26sHR3o/+EsgoYM4sSKXwHwvyWCPrOn02f2dDqPexJ7DzdVBz8WwPOhLZkWHctTOw4yyNeT5k72Jm1O5+Tz3K7DjN11iO2pmYxrF1S5bUV8EnOOnFQtrzlH98WRmpjO20te47GXH2LJJ6vMtuvcK5RpX79Q6/bUxHT+WLKFVz5/npk/TOWhiSMaOnKlS7UTMWUi/ee8wUUztXNhm7F2Bn40ixZ3DOL4cmPtGMrLOfTND3Qc8zD933uDntNexMLKUrXsl/KfWvITHV+cRPfZb5G2L4r8ixdN2ti5u9H2qSfx7hFhcntpXj7n1qyjy/SpdHl9KufWrKM0P1/N+ACcP3CM7OQ0HvniDQY8O4pt85ebbbf9m+UMeG40j3zxBtnJaZw/eAyALiNuZdQn0xj58VSCwkOJWvGHmvFJPHiMnOR0HvzsTfqMH83uBcvMttu1YDm9x4/mwc/eJCc5ncRDxvzaZn7cOmUsPiGt1Ixd6VT0MTKT0nn+29e56/lRrPt8pdl2bXt0YNzcl2rdfmDDHuyd7Jn83Qx63TuATd+vbejIwM3fd97s+QFORh0j42I6L33/OiMmj2LNZWqnXY8OPPtp7dqJ3rAHOyd7Xl44g973DmCDSrUDxr7z4rKltJg4mdZvzCI7KpKiZNO+09rNjYDHx6DpHlHr/hY2NrSe/iatp7+p+uAH4MeV27jn8TmX3T5kYBitgnzo0O9FJk5dwLx3ngZA6+rI9Bfuo9/dM+h79wymv3AfGtfGGTTs23mcxPMZLFnzKlNmPMDH7/xitt1bHzzG9yte4oefX0avy2PrpiOV29JS9ETvPYW3r0at2I1EqPYlhBgnhIiu9jXuGoJ6K4qSDFDxr1etZ2KcZvov8J9reQUaawmcBpjQSD/7qvRnE3D09sTByxMLKyt8e4STduCwSZu0A4fx79MTAJ/uXck8dhxFUUzaXNwbhV/P7qrlBmincSYpv4jkwmLKFIW/k9O5xcvNpM2hrGyKDQYA4vS5eNrZVG47mJlNQVm5qplrOrTrKL2GdEcIQcvQIArzCtFnZtdq1zI0CI177SPzO9btYcCIPjg6OwDgonVu8MyX6M8k4OBVVTt+PcNJrVE7qQcOE1CtdjIqaifjaBzOzfxxCTQesbJxdkJYqLuL5pyNx97LC3tPY36viHAyD5rmt/PwwKlZQK0/Ja2LjUUbGoK1kyPWjo5oQ0PQHY1VMz4A8ZExtB0QgRACn7YtKMkvJD/LtH7ys7IpKSzCp20LhBC0HRBB/L4YAGwcqj40lhaVIFT+k9nnoo4Q3N+Y36uNMX+BzjR/gS6b0sIivNu2RAhBcP8IzkUa38Q1AT5o/L1VzVzd8b1HCbvVuP82axdEUX4huVm1999m7YJwNjOzdnzvUcJuM35AbN+nM/GHT9bqWxvCzd533uz5AeL2HKVLRe0EhgRRlFdIjpm+PzAkCBczfX/cnqN0raid0L6dOXNIndoBKEiIx8bTE5uKvtM1vDs5hw+ZtLFx98A+oHbfeSPYFXmcLH3eZbcPv70bS3/eAUDkwdO4ujjg46VhcP/ObNkRgy47H312Plt2xHB7/85qxTaxc2ssQ4Z3QwhBaKfm5OUWkZmeU6udo5MdAOVlBkpLy01+HZ9/tIZnXxgmr5J2HSmKMl9RlPBqX/OrbxdCbBZCHDXzdU8df8QE4HdFUS5cS67GWgI3B2glhDgElAL5QohVQAdgP/CooiiKEOIN4C7AHtgNjFdU6M2KdHrs3KrOs7Jz06I/E3/ZNhaWlljZ21Oal4+Ns1Nlm+R9++n2wrMNHdeEh50N6UUlld+nF5UQorn8AGBogDeR6To1otWZPj0brWfV0RetpwZ9erbZwY45qRfSAXh/4qcYyhXuenIIHXqENEjWmop0euzd61A77lW1Y+1grJ385FQEsO+DeZTk5uHXM5xWw25XJfclJXo9ttVq31arJSc+/gr3qFKs02OrNb1vsU5/3TNeTX6WHiePqhyO7hrys7JNlrHlZ2Xj5K6p0aYq694lazmxNRJbB3vumTVJneAVCrL0OFarIYeKbA7a6vn1ONbIX5Cl/mttTm6GHpdq+6+Lhys5GdlmBztm75+px8XT+PwtLS2xdbCjICcfR1enq9zzn7nZ+86bPT9ATqYe1+q14+lKTma22cHO5e9fVTt2jurUDkCZXo+1tmrAaa3VUlDHvhPAUFrK6ffeBgsLPIcMxTWsS0PErDc/HzcSkzMrv09KycLPxw0/Hy2JF7Oqbk/Ows/nsuepN6iMtBy8fKrqx9PblfS0bNw9XWq1nfLcAuKOXqBH77b0v60TALu2xuLh6UpwWz/VMjeWG2mApyjKbZfbJoRIFUL4KoqSLITwBdLMNOsF9K1YWeYE2Agh8hRFudL5Qo02AzQVOKMoShjGKasuwAtAe6Al0Lui3eeKonRXFKUDxkHQ8Ks9cPWptvnz51+tuXlmxli1jgKbG4ZVa6I/E4+lrQ3OAf71y3AdXW7MeJufJ21cnVgRn2R2e2NRzLy413IU3lBuIC0xg5fnTmTsG4/xvw+XU5BbeD0jXsFVCuMKDAYDWSfP0OW5p7jl9SmkRB8iI/b49Y13Nf/o8IKZOzfCkU6z9V5z9zXbpqpRz0fu4okFs2ndL5yYP9Rdz24+fl36nxvjDc18tLpnM/+raZzndrP1nTXdbPn/8e++MWvH7OeGut+93TvvEzztdQKfGkvyyuUUp5v7nNd4zH1gVhTF7Our0qSbmZ9b988OH301ll82z6C0tIwDkacpKizhx2+38NQEdQ86Sle1Bnii4v9PAKtrNlAU5RFFUQIVRQkCpgD/u9rgB26ciyBEKoqSCFAxKxQE7AQGCiFeARwANyAWuOKi3oqptUsjH+WFvX9dcxg7Ny1FWVVHxoqydNhqXGu00VCUpcPeTYuhvJyywkKsHavWvSbvjcavZ53Ow7quMopKTJY1eNrZkFlcUqtdV3dXHm4VwEv7jlJqaKTeqpq/f93JjnV7AAhqF4guvepoti5dj6tH7SM4l6P1dKVl+yCsrCzx8HXHJ9CLtKR0gtoFXvfcNdlptRRmmtaOnbZG7Wg1FGVW1U5pQSHWTo7Yu2lwb9e6chbRq3MHshPO4xHarsFzX2Kj1VBcrfaLdTpsNXVbC22r1aI/UXUOQbFOh6Ztm+ue0ZyYP7ZzbNNuALyCA8nLqHoO+Zl6HGv8DpzcNeRl6q/YBqBN33DWv/M1EaOGNVByo2N/buPEZmN+j+Dm5FeroYJMPQ41Zk8c3TXk18jvYCa/Wvat3cGBDcb91691IDnV9t+cjGyc3eu+/7p4aMhJ1+HqoaG8vJzigiLsK5azNqSbte+85GbNv3fNDqL+NNZOQJtAsqvXTno2zm7XVjvZ6TpcPY21U5SvTu0AWGm1lOqqZkJKdTqsXOt+Hol1RT9r4+mJY5s2FF24gK1nrdMdGk1SSiYBvlUXlPD3cSM5VUdSchZ9e1WtsPD3dWPHHvUuovHrsl2s+2UfAG1Dm5GWUlU/6anZeJiZ/bnE1sNIbVEAACAASURBVNaa3v1D2bU1FjcPZ5KTsnj6oU+M903LZuzouXy9eBLu1/D54+ZxYxwwq4M5wAohxNPAeeBBACFEOPCsoijP1PeBb5TLYBdX+385YCWEsAO+BB5QFKUjsACwUyOMa4vm5KemUZCegaGsjOR90Xh16WTSxqtLJ5J27gUgJeoA7iFtK480KAYDyVEH8O2h/gDoeHYu/o72+NjbYiUEA3092Z2WZdIm2MWRFzu0Ysb+OPQlpapnNGfgvX1447v/8MZ3/yGsTwf2bIhCURTOxiZg72hf5+VvAGF9OnLi0CkAcvV5pF5Ix8NXnSsBubY0rZ2Le6PxrlE73l07kVitdjzaG2vHs2N7ci4kUV5cgqG8nMzjJ3Hy91Ul9yUuLYIoTE2jsCJ/WmQ07mF1W8+tDQ1FF3uM0vx8SvPz0cUeQxsa2rCBK3Qc2o+RH09l5MdTaRHRiRNbI1EUhZQT8dg42NW6ipujmyvWdnaknIhHURRObI2kRURHAPQXq468xkfFqHI+Tfs7+nPvR9O496NpNO/eidPbjPnTTsZj7WBfa3DjoHXF2t6WtJPG/Ke3RdK8e6fLPHrD63FXX577/BWe+/wVQnp15NAW4/574XgCdo52dV7+BsaLIxzaHAnAsZ2HadGptSpH8W/WvvOSmzV/z7v7MunLV5j0pbF2DlbUzvm4BGwd7eq8/A0gpGcHDlTUTuyOw7TsrE7tADg0D6I4LY2SjHQMZWVkR0fh0qlufWd5fj6GUuPvoywvl4IzZ7D1Vbfvv5r1mw7w8P19AYjoEkxObgEpaXo2bfv/9u47rqr6j+P46wOyBGUI7pl7ZDjL3JqZOSrTRNs/f5X1a6iNX5aZDbOystKybPyszL21qebeZu5ZiVsRFURAQPj+/jgHuMDFgXCvxOf5ePDg3nu+9973OZz1HeewhVtaNyQo0J+gQH9uad2Qhcu2XOLT8s9dES35atpgvpo2mNbtG/DLgt8xxrBj6wH8A3xzDH9LSEjKuC7owoVU1q7cTeVqpalesxxzlwxn6k8vMfWnlwgrHcgXkwf+Qys/hYcx5pQxpqMxpqb9+7T9+kZnlR9jzARjzGXdRcRdPUBxwKWuTE+v7ESLSADQC3B+O7B85uHpSb37I9gwagwmLY2KbW6mRMXy7J01n8CqlSnT+AYqtmnJ1vETWPb8MLz8ixP+RP+M95/e8ye+IUEULx3mirhZpBkYs/Nv3mlWHw+Bnw5HceBcIg/VrMye2HOsiTrNo7Wr4ufpybBGtQGISkzmlU1Wi82HNzagUkBx/Dw9mNK+Ke9t+5ON0a69tuD6m+qxfd0uXr53BN4+3jz034iMaa/3H8Wwr6wbfcz4bB7rF20iOSmFF3oNp1XXm+jx8G3Ub16HnRv38OqDbyMeHtw9oDsBLrorjYenJw0eiGD9u2MwJnPd2TNzPkHVrHWnUpuWbP58AkueG4ZXQHEa2+uOl78/1W7ryMrh1p14St/QgDLh17skdzrx9KTGvRFsG/0RJi2Nsq1a4l+hPPvnzKNE1SqEht/A2f2R7PhkHBfiEzi1ZSuRc+fT7I3heAX4U7lbVza9ORKAKt274hXg+rsBVWlSn4ObdvL9E69TzMeLDk/elzFt6uC36fOB1TPe9rE+GbfBrty4LpUb1wNg7cR5xByJAg+hRFgIbR/r49L8lRrX5/AfO5j+1GsU8/ai9X8y889+biR3vTcEgJsf6cPyTyaSmpxCxfB6VGxk5Y9ct4U1X0/n/Nlz/DryM0pVrcBtQ113V6mazeqxd8NOPur/Bl4+3tw5qF/GtHFPvsvjY61b7P761Vy2Lf2dlKQU3r9/GI07t6D9fV1o3PkmZr03kY/6v4FfieL0+u+DuX1Vvirs+87Cnh+gdnNr3fngX9a603Nw5roz5ol3eepTa935+cu5bLHXnXfuG0bTzi3oeH8Xmtx2EzPencj7D1vrTsQQ16w7YO07y0f0Y/+YDyHNEHxzS3zLV+DE/Ln4Va5CyRvCSYjcz4HPPyU1IYG4bVs5sWAutYa9zvnjxzgyaSIigjGGsM634VvOtdehfDPmKVq3qEtocAn+XDeWNz6YgZeXdYr45cRF/PzbH3RuH86OFR+SkJjEY899DsCZ2HhGfjyblfPfBOCtj2ZxJtb1d/8EuKl1Hdau3EW/7m/j4+vNi6/dkzGt/z0f8NW0wZxPTGbIM/8jJeUCaamGRs1r0KPXTW7J60758P95Cj1x1R1ScnyxyCSgIZAInDDGdLNfHwtsNMZMEJE3gQggEjgEHDDGDBeRCcACY8ylKkR5GgJ3Lfjwpg4AdPxplZuT5M3iLtZlXMuO/ejmJHnTttztDF5XONcdgA9u7MCjK5e6O0aejW/Vjo92/OruGHnyTH1rDPm7WxdeouS16YWGnQCY8tfPbk6SNxHVrf9bVZj3nYU1O1j5Z+wvnOsOQK9qt9HrN/f8H5urNaNDGwD8Kvd1c5K8STw4GYDjifPcnCRvyvr1gEIytuxsykKXnfyX9Op0TS4Tt10DZIzpl8vrTzo8HgoMdVLmoYJLppRSSiml1D/VNVkncSntA1NKKaWUUkoVGdfKXeCUUkoppZRSBexa+j9A7qI9QEoppZRSSqkiQ3uAlFJKKaWUKiK0B0h7gJRSSimllFJFiFaAlFJKKaWUUkWGDoFTSimllFKqyND+D10CSimllFJKqSJDe4CUUkoppZQqIkT0JgjaA6SUUkoppZQqMrQHSCmllFJKqSJDe4C0B0gppZRSSilVZIgxxt0ZCtI/euaUUkoppdQ1o1B0rSRcWOGy8+PixVpfk8tEe4CUUkoppZRSRcY//hqgzacWuDtCnoSX6gbAvtjCmb9moJW/yqjf3Jwkbw4834Ffj/zo7hh5dmuF25n018/ujpFn/arfxujtC90dI08GNegEQL+ly9ycJG8mtWsLQOUPlrg5Sd4cHNwegHMphXPfE+DVgbiUxe6OkWclvDpS7+vl7o6RZzv/1YYzSYXzuBvsYx13jyfOc3OSvCnr1wMAv8p93ZwkbxIPTnZ3hCug/R+6BJRSSimllFJFxj++B0gppZRSSillkcJxqVKB0h4gpZRSSimlVJGhPUBKKaWUUkoVESLaA6Q9QEoppZRSSqkiQ3uAlFJKKaWUKjK0B0h7gJRSSimllFJFhlaAlFJKKaWUUkWGDoFTSimllFKqiBDt/9AloJRSSimllCo6tAdIKaWUUkqpIkNvgqA9QEoppZRSSqkiQ3uAlFJKKaWUKiL0H6FqBShXxhgmjJ7DH2t24ePrzeNDI7iudsUsZZLOJzP65W85cSQaD08PmrSsR78nugGwcPZqfpm5Cg9PD3z9vHn0v72pWK2sS/OPf38OG1db+QcOi6BGnaz5z59P5u0h33L8cDQeHh40b12Ph5608kcdP8Po1yYTH5dIWprhwf90pVnLui7J3rZqCK92rImnCFO2HmPc+gNOy91eK4xxd1xPt283sO1EHDeULcHIznUAq3P3w9X7+WVftEsyOzLGMHPsbHas24W3rxf3vdCXSrUq5Sg3/6sfWP/rRhLiEnj/x3dyTP9j2Wa+fu0bnh83iMq1K7siOmDl//nzWezbsBMvHy/uHHwv5WrkzH903yHmfvA9Kckp1GxWj9se64mIcPzvI/wwdhrJiUkElQmh5wsP4FPc16X5V309g4ObdlDM25v2T91P2HU585/86yBLxn7HheQUKjeuT8t/9UJEWD95AZHrtyIegl9gCdo/eR/+IUEuyx+7fTuHpk2FtDRCW7Wi7G1dskyP27uXQ9OmknjkCNf9+xGCmzTJMj01MZEdw18lKDycyn37uSw3WNvu8HY18fSAKduO8emGg07L3V4zjM+6N6Db9xvZeiKO1pWDebF1dbw8hZRUw4jlf7L6UIxLs4O17owaOY1VK3bg6+vN8BEPULde7tveoCc/5cjhaKbNGZbl9W//t5CP3p/FohWjCA4OKOjYGYwxvDdyup3fi+EjHqDORfOPs/O/AsC4MfNZ9tsWPDw8CA4JYPiIBwgr7bp1v1WFYIbcVB1PEWbsPc6XWw9lmd6ndjn61i1PmjHEX0hl+Kp9/BWTQPkAHxb0bEpkbCIAW06e5bXVf7osdzpjDB+8M4c1K6zj7itvRFCnXsUc5QYOGE909FlSU9MIb3wdz73UE09PD/btOco7b8wgMSGJsuVDeP3te/EPcO2+8+N357Ju5W58fL0Y8nofatXNmf/5J77gVHQcqRfSaNi4GgOH3IWnZ+aAoinfLGXc6B+Yu2Q4QcH+Lsn+2ajH6NKxESdPnaVppxeclnn/tQfp3D6chMRkHn12HJu3RwJwb682vPjUnQC8PWYO389Y7pLMyr3ybQiciASJyBP59XkX+Z52InJzQX/P5jW7OX44mo+mDeGR//bmq1EznZbr1q8do6e8yDsTBrNnWyR/rNkFQMtbG/PexOd595tn6XFve779eF5BR85i4+rdHD0UzfiZQ3hySG8+fcd5/p73tuOz6S/y0cTB7NwSycbVVv6pXy+idcdwPp74LC+8eR/j3nX+/vzmIfBGp9o8OGMLt3y9jh51S1OzVPEc5fy9PHmocSU2HY3NeG1PdDzdv93I7d9s4MEZW3irUx083dDKsXPdLqKOnGTYdy8RMfgepn44w2m5Bi3q89ynA51OO59wnmWzVlC1bpWCjOrUnxt3cvrISZ76cijdn47gh7HTnZb74ZNpdHu6D099OZTTR07y50Zr3Zn/0WQ6Ptydx8e9SJ2bG7JqxmJXxufgpp3EHjtJ37Gv0vbxvqwYP8VpueXjp9JmQF/6jn2V2GMnOfTHTgDC7+jIPaNfovf7Q6jSpAG/T//JZdlNWhoHJ0+i5lNPU2/4a5zesIHEo0ezlPEOCaHqQw8T0ry50884Om8uATVruSJuFh4Cb3aoxYOzt9Bxwnp61ClDzRDn2+7DjSqy6Vjmtns6MYV/zdnKrd9uYNDPu/iwSz1XRs+wasUODh2MYs6PrzF0eD9GvjE517K/LfwDv+I+OV4/fuw069bsomy5kIKM6lR6/tk/Dufl4fcy8g3n6z5Y+Ytny3//w7cwZfZQJs18idZtr+eLcT8WdOQMHgJDW9TgsV+3033WRm6/LozqQVnXnwV/R3HnnN/pOXcTX289xAvNr8uYdijuPD3nbqLn3E1uqfwArFm5m0MHopm+YAhDhvXm3TedHzdHvPcAE2c8x6RZz3Pm9Dl++3ULAG8Nn8YTA7vy/aznadexARMnLHFlfNat3M3hg9F8P++/PPdKLz4YMctpueHv3s/X0wYzYeazxJw5x9KFWzOmRR2PYePafZQp57qKM8B305dxxwNv5zq9c/twqlctS4M2g3jyxS/4eER/AIID/Xl5YE/a9HiF1j1e4eWBPQkKdE2lzb3EhT/Xpvy8BigIuOwKkFjy8v3tgAKvAG1YsZ02tzVBRKjVoArx5xI5E302SxkfX28aNKkBQDGvYlSrVZHTUdZBvbh/ZqtNUmKyy7sb1y3fTofbrfx1rq9CfFwip7Pl9/X1pmFTK7+XVzGq16lItJ1fBBLizwMQf+48IaElXZI7vFxJIs8kcCj2PClphvm7o+hUIyxHuWdbXcdn6w+QdCEt47XzF9JINQYAn2IeGJckzmnb6u0079QMEaFavaoknksk9lRsjnLV6lUlsFSg08/44eufuCWiA8W8Xd9Ju3vtdhp2tPJXrFOV8/GJxJ3Omj/udCxJCeepVLcaIkLDjs3YvXYbANGHo6jSoDoA1zWqza5VW1yaP3LDVmq1bY6IUKZWNZLiE4k/kzV//JlYUhLOU7b2ddY23rY5+9dbB3Hv4n4Z5VKSknDlDjx+/358S5fGJywMj2LFCG7ajJgtWZefT2goxStWdLpPiT9wgJSzZylZz/UViPCyJYmMSeRgxrZ7glurh+Yo91zLany24WCWbXfHyXOciE8GYO+peHw8PfD2dP2Bc9mSLXTtcRMiwvU3XMe5uAROnsy57SYknGfit4v592O355j2wbszeGZwT9wxwmTZkq3c3uNGO3814uISiM4l//ff/kb/x7L2LgYEZK77iYlJLj1uXR9agoNnEzkcZ60/P/19kg6VS2UpE5+SmvHYz8vTZdku1/Il27m9u3XcbXBDFc7FJRJ98myOcum9OqkX0khJSc3YxRyIjKJRE6tS17xFLZYs2uay7AArl+6gczcrf/2GVTgXd55Tl5HfcTUZ+948Bgzsirj4xHfV+t2cjjmX6/RutzZh0swVAKz/408CSxanbOkgOrW9gcUrtnEmNp6Y2HgWr9jGrW1vcFVs5Ub5WQF6G6guIptFZLSILBaRTSKyTUTuABCRqiKyS0Q+BTYBlUSkv4jsFZGlIvKFiIy1y4aJyEwR2WD/tBSRqsAAYJD9Pa3zMX8WZ07GUqpMZgtGqbBATjs5kKSLj0vk91U7aNC0ZsZrv8xcydO93uL7Txfw0KA7CyqqU6eiYgl1zF86kFNRuec/F5fI+hU7CG9m5e/3SGeW/Pw7D3Z7neGDvmTAc3cVeGaAsgE+HItLynh+LC6JsgFZWynrlw6gfEkffvv7VI73h5crycKHm/PLQ815eeHujAqRK8VExxLsMGwkKCyI2Ojcl312h/Yd5szJGBq0qF8Q8S4pLjqGwLDM/CVDA4nLlj8uOpaSoY5lgoiLtoYsla5ajj1rtwOwc8Vmzka7dihT/OkYAkKDM54HlAoi/lTWDPGnYvAvFZS1zOnMMuu+n8d3jw5l3/KNNIvoWvChbSkxMXgFZ/YceAcHkRJz5rLea9LSODxjOhXv7lVQ8S6qbIAPR+POZzw/di6JMiWybbthAZQr4cPi/Tm33XS31wxjR1Qcyamu33ajTsRQpmzmulO6TDAnT+Rcf8eNmc99D96Cr693lteXLdlCWOkgatXJOWzIFU6eiKGsQ/4yZYKJcpp/Afc92DFHfoBPPppL144v8dMPGxhgD4l2hTL+PhyPz9z3H49PonTxnPn61i3Hz72a8WzT63hrbWZPT4UAX2be0ZhvujSkSRnXNNhldzIqltJlM/crpcsEcjKX4+4zAz6nS7tX8ff3oUMn64S7eo2yrFi6A4DFv24l6rhr953RUWez5A+7SP7nHv+COzq8RvHiPrS9pSEAq5buIDQskBq1y7sk75UoXzaEw8cy9ztHjp+mfNkQypcN5vDR05mvHztNeYdt6J9K8HDZz7UqP5O9CPxljAkHngfuMsY0BtoD70tmU1Jt4FtjTCMgBXgFuAnoBNRx+LyPgNHGmGbA3cCXxphI4DP79XBjzIrsIUTkURHZKCIbx48fn+eZcXbenFtjWOqFVD5+dSK39W5NmQqZLVad727FxzNeot8T3Zg1YVGes+SFs1OHi+UfNXQiPfq0pqydf9kvf9CxWzO+WTCM4aP/zfvDJ5OWlub8AwqYcZgbAV5pX5M3lzgf4rD52Fk6/W89Pb7byBM3VsXH0/Ubn7mSlSebtLQ0Zn06h7sevyOfU12lbPmNkzUsfRO/Y2A/NixYwfinR5GUeB7PYi5uqXW6+C+9/B1bLG+8twf3j3+Tmm2asv0nV44Hd7rlXtY7Ty5bSmCDBniHuH7oFThP6bgpCDCsXQ3eXPZXrp9Rq1RxhrSuzpBFe/I93+W4nE13z+5DHDp4kg63hGd5PTExma/G/8yAJ7sXYMKLc7bvyb7u79l9iMMHo2ifLX+6/zxzBz8sfosuXZsxbdKyAsnpzOX2F0zedYzbZmzgg41/89gN1hDhkwnJdJy2jrvnbuKd9X/zbtu6+Luhh+hKdv0fffYYC357leTkC2xcvw+Al1/vw4wpq3iwz2gS4s9TzMXzcDnrT7r3xj3CrEWvkJJygU3r/+R8YjLffbmYfz1xa0HHzBNnPVLGGKfz54Z2U+UGBTW+RoC3RKQNkAZUAMrY0w4YY9baj5sDy4wxpwFEZDqQPnj9FqCew8pZUkRKXOqLjTHjgfSaj9l8asFlh/5l5koWz1sHQPU6lTjl0HJ26mQswaHOhyuNf2c6ZSuG0rVPG6fTb74lnC9zuYYoPy2YvpJf5lj5a9arRLRj/qhYQsKc5x8zcjrlK4VyR9/M/AvnreO1jx8BoG7DqiQnpXA2Jp6gkEv+Ca7K8XNJlHNoNS5XwocT55Izngd4e1I71J8pEY0ACPP35queDek/ayvbTsRllPvzdAKJKanUCvXP8npBWT5nJat/WANA5dqVOROVuexjTsYQWOryWiSTEpI4tv84Hw8aC8DZ03F8PvQrHnuzf4HeCGH9/BVs+sXKX75mZWJPZuY/Gx1LiWz5S4YGZenZORsdQ4A9nC+0UhnuH2GNhj11OIp9G3YWWO50239axq5FqwEIq1GFc9GZvSbnTsVQPCTruu+frVfIWRmAmq2a8eNb41zWC+QVFEzKmczWyOQzMXgFXd5Y+vi//yZu3z5OLltG6vnzmNRUPHx8qdizZ0HFzeLYuSTKl8gc+lsuwIeoc5kt+unb7tTe1ol3mL83X91xPf3nbmPriTjKBvgwvsf1DPp5Fwdiz+f4/IIybfJSZs9YBUC9BlU4cTxz3Yk6cYbQbDcB2Lr5b3btPEi3W18mNTWN06fiePShD3j+pT4cPRJN37vftN8bw7293+LbKf8lNJdjR/7kX8Ych/zHHfKfOHGGsNJZv3vb5v3s2nmI7rcOdcg/mvETBmUpd1vXZjzzxKc85qJeoOPxSZT1z9z3l/X3ISohOdfyP/59kmE314QVkJJmiE26AMDOU+c4FJdI1ZJ+7DiV+5Co/DJjykrmzrSOu3XrV8rSaxN1IpbQXI67AD4+XrRuV58VS3ZwY4vaVK1Who8/fwyAg5EnWb1iV8GGB2ZPWcWCWVb+2tnynzwRS2hY7scuHx8vWratz6qlOwgJLcGxI6fpf89o671RsTzS90M+m/gUpVw0hP5ijhw/RcVymQ3UFcqGcOzEGY4cO03rFpk3eKpQLoQVawp+ubvftXttjqsUVAXoXiAMaGKMSRGRSCD9yBjvUO5ifwEPoIUxJtHxxYIck9z57lZ0vrsVAJtW7eSXmau4uVMj9u04SHF/X4KdbMRTPv+JhPjzPDbkniyvHzt0knKVrGtX/li9i3KVco6Fz2/dereiW28r/4aVO1kwfRVtbm3Enu0HKR7g6/Q6nu/G/UTCufM8/XLW/GFlg9myYR+3dGvOof0nSEm+QKAL7ma05Vgc1YKLUynQl+NxSXSvU5qnF2SeQMclp9Lok5UZz6f0acSIpX+y7UQclQJ9OXo2iVRjqFDSl+tCinP4rGtOpNrc2Yo2d1rLfvvaHSyfs5ImHRoRuesAvv5+uV7rk51fgB9vz3kz4/lHg8Zy14AeBX4XuObdW9O8uzWidO/6HWyYv4IGbRtzZM8BfPx9KZGtclAiJBAfPx8O746kQu0qbF28geY9rPfHx8ThH1QCk5bG8im/0vT2lgWaHaBBl7Y06NIWgAO/b2f7T8up0aoJUfsi8S7uh39wtgpQcCBefj6c2Luf0jWrsnfZ+oz3xxyNIqh8aQAiN24luEIZXMW/alXOR0WRFB2NV1AQZzZuoFr/f1/Wex3LRa9eTcKBSJdVfgC2HI+jWpAflUr6cvxcEt3rlOHpH3dkTI9LTiV83KqM51N7hzNi+V9sPRFHSZ9iTLirIe+s/JuNRy9/uGh+uKdvO+7p2w6AFcu2MW3yUjp3acr2rfsJCPAjLNsJbO+ItvSOsNaVo0dOMfA/nzB+wmAAFi0flVGu260v893UIQV+F7h7+rblnr5WnpXLtjFt8jI7fyQBAX45TsB7RbShV0Qbh/yfZlR+Dh6IonIVa91ftmQrVV1459Lt0XFUCfSjQoAvUQlJdLkujBeW7s5SpkpJXw7Y+/S2lUI4cNY6PQj29SI2KYU0AxVL+FKlpB+H41yz7+8V0YpeEda+f9XynUyfvIpOXRqxY+tBAkr45qhAJCQkkRCfRGhYSS5cSGX1yl2EN7au+zl9Ko6QUiVIS0vjf+MXclfvFgWe/66IltwVYe2j1yzfxaypq+h4Wzg7tx3EP8CXUk7yJ8YnUcrOv3blbho2rkb1muWYu2R4Rrk+Xd7i80nPuOwucJfyw8JNDHjwVqbNW03zRjU4G5fA8agYFi7bwmsv9Mm48cEtrRsy7O3cbx6i/jnyswIUB6R3DwQCUXblpz2Q262s1gOjRSTYfv/dQPpVf78CTwKjAEQk3Biz2S5X4M0JjW6uyx9rdvFM75F4+3rx+MsRGdNeePB93v3mWU5FxTD7m0WUr1KaFx+2Wj06392Sjj1u4pcZq9i2cS+exTzxL+HHE0P7FnTkLJq2rMvG1bt4pOdIfHy9GPhKZv6n7n2fMd8/S/SJGKb+bxEVq5bmmfut/N16t6TznTfR/5nujHlrOnMmLUdEGDgswiUXxKYaw7BFe/m2VzieHsK0bUfZdyqewS2rsfV4HIv+yv221k0rBPFEz8qkpBmMgaEL93AmMaXAM2dX/8Z67Fy3i9fvG4GXrzf3vZC57N9+ZBQvfvE8AHM+n8fvizeRkpTCK/cMp8XtN3H7Q7e5PG92NZvVY9+GnYzp/wZePt7cMSjzVsqfPfkuA8Zatxjt+p97mDP6ey4kpVCjaT1qNLUuvN+29Hc2LLAqqXVbNiS8040uzV+5cX0ObtrB5P+8RjEfL9r9576MadOfHUnv94cA0PrRPiwZO5HU5BQqNapH5cZW/nUT5xJzNAoRoURYCK0fi3D6PQVBPD2pHNGXfR99iElLI7RlS/zKl+fovLkUr1KFoBvCiY+M5K9xn5KakEDM1q0cnT+P+sNfc1nG3KQawytL9vLd3TfgKcLU7cfYeyqBwTdXY9vxsyx0cs1eugfDK1A1yI+nb6zC0zdah4v7Zm7hlIu331ZtGrBqxXbu6DIMXz9vhr/xQMa0vnePYPLMl12a50q1bNOAVSt2cGeXV/H1wOz5qQAAEFBJREFU8+bVN+7PmNbv7reYNPOli75/zOg5HIg8gYcI5cqHMGSY626jnmpgxJo/+aJzAzxEmL3vOH/GJPBkoyrsiI5jyaHT9KtbgRblg7iQZohNvsBLy62hkk3LBPJU4ypcMIa0NMNrq/cRm3zBZdnT3dy6LqtX7KJX15H4+nox9I3Mfcf9vd/nu+nPkpiYzPNPf01y8gXS0tJo0rxGRkVn4U9/MGOq1UjQruP1dLvT+Z0eC8pNreuwduUu+nV/Gx9fb158LbNhtP89H/DVtMGcT0xmyDP/IyXlAmmphkbNa9Cj100uzenMN2OeonWLuoQGl+DPdWN544MZeHlZp7hfTlzEz7/9Qef24exY8SEJiUk89tznAJyJjWfkx7NZOd9qeHzro1mciY3P9Xv+KVx9k4prkTi9XiGvHyYyCWgIbMC6nscL2Ay0BNJvN7PAGNPA4T2PAs8BR4FdwGljzMsiEgp8AtTFqqgtN8YMEJFawAysoXVPObsOyMEVDYG7loSXsoYd7IstnPlrBlr5q4z6zc1J8ubA8x349YjrbgGb326tcDuT/vrZ3THyrF/12xi9faG7Y+TJoAadAOi31HXXT+SnSe2s3oTKH7j2Frz55eDg9gCcSymc+54Arw7Epbj21vH5qYRXR+p9XXj/j8rOf7XhTFLhPO4G+1jH3eOJrv23G/mlrF8PAPwqu7bBOL8kHpwMhWRs2YW0LS670qmYxw3X5DLJ1yFwxpjLaS5qkO35JGPMeBEpBszG6vnBGBMN9HHyHXuxKllKKaWUUkqpK+Dqf81yLboW7k83XEQ2A9uB/cAcN+dRSimllFJK/UO5/r8sZmOMec7dGZRSSimllFJFg9srQEoppZRSSilXuRYGgLmXLgGllFJKKaVUkaE9QEoppZRSShURehts7QFSSimllFJKFSHaA6SUUkoppVSRoT1A2gOklFJKKaWUKjK0B0gppZRSSqkiQv8RqvYAKaWUUkoppYoQ7QFSSimllFKqyND+D10CSimllFJKqSJDe4CUUkoppZQqIvT/AIEYY9ydoSD9o2dOKaWUUkpdMwpJzWKvC8+Pa12Ty+SfXgEqUCLyqDFmvLtz5JXmd6/CnL8wZwfN726a370Kc/7CnB00v7sV9vwq/+g1QFfnUXcHuEqa370Kc/7CnB00v7tpfvcqzPkLc3bQ/O5W2POrfKIVIKWUUkoppVSRoRUgpZRSSimlVJGhFaCrU9jHkWp+9yrM+QtzdtD87qb53asw5y/M2UHzu1thz6/yid4EQSmllFJKKVVkaA+QUkoppZRSqsjQCpBSSimllFKqyNAKkFKXQUSCROSJfPqsh0SkvMPzSBEJzY/PLigiMkFEel1B+aoisr0gM11mjqdFZJeIfC8iz7k7z9UQkTtFpJ6bM2RsByLSTkQWXOH7r2g9cnjfFX9XUSAi53J5PU/L+RLf9ZCIjM3Pz7zIdy0Vkaau+C6VyV3b95XKz+PxJb6nnYjcXNDfo9xDK0BKXZ4gIMcOV0Q88/BZDwHlL1VI5YsngNuBfe4Okp1YrmQffCfg1goQuWwHSql/hMKyfV9Rzjzsa9O1A7QC9A+lFaCLEJE5IvK7iOwQkUft1/qLyF67heqL9BYxEQkTkZkissH+aene9CAi/iLyg4hsEZHtItJHRJqIyDJ7vn4RkXIiUszO3M5+30gRGeHm+FmIyAMistWel+/slqbPRGSF/ffoVsAR3gaqi8hme1ktEZFJwDY7330ist6e/rmIeNo/E+xlv01EBtmtY02B7+2yfvbnP2+/f72I1LA/0+k8ikh9h+/aKiI183tmsy9v++U2IrJaRP5Ob+WzDyyjHOaxT35nySsR+Qy4DpgHDAJuEJHfRGSfiDxilyknIsvtZbldRFoXcKaqYvVIfQpsAu4XkTUisklEpotIgF3ubRHZaf8N3rNbIXsAo+ys1e2fn+1teYWI1LHfW0ZEZtt/uy3pLZgi8oqI7BaRhSIyWfLWI5axHQCjgAARmWF/7vciIvZ3DbO3k+0iMj799WzLwmkZEakhIovs7JtEpLr9FqffdTWcbLdV7PUjVEQ87OV6q102x/HAfv2ciIyw864VkTL269Xt5xtE5HXJpbfmCrIOtpfVdhEZmG2aiMhYe535ASjtMC1SRN6RnPsXp8csEWlub+d/2L9rO8nS1V5vr7rnWpwcp7JN72vvW7aLyDsOr58TkfftdWSxiITZrzvdLtxFch67uovIOnv5LkpfX64R+bZ9uyqniIy2//6b7PXkDjtj9n1tJbmC8zcRqQoMAAbZ31OgxwblBsYY/cnlBwixf/sB24EKQCQQAngBK4CxdplJQCv7cWVg1zWQ/27gC4fngcBqIMx+3gf42n5cH9gFdAL+ALzdnd8hd31gDxCa/ncBJgA/Y1XiawKHAd8CzFAV2G4/bgfEA9Xs53WB+YCX/fxT4AGgCbDQ4TOC7N9LgaYOr0cCL9uPHwAW2I+dziMwBrjXLuMN+LloeU+3s9QD/nRYxxYCnkAZ4CBQznF5uXndiQRCgeHAFntbDgUOYfXCPeuw7D2BEgWcpyqQBtxk51gO+NvT/gsMs5f3HjLv0pm+3kwAejl81mKgpv34RuA3+/FUYKDDPAViVbo32/NfAqtH7Ll82A5igYr2erGGzH1giMN7vgO6Z5+Hi5RZB9xlP/YFil/su67ib5HbdvtvYAbwPPC5Q/nsx4NS9nPjkP1dYKj9eAHQ1348ADh3FVmbYDW2+AMBwA6gUfpnAj3J3A7LAzEOyzkS5/sXp8csoCRQzH58CzDTfvwQMBa4C+vYF5xP24Sz49RSe50tj7VPCQOKAb8Bdzos9/T94DAyj8VOtwt3/OB8XxpM5rb9b+B9d+Vzkrcq+bR9uzBnMaCk/TgU+BMQHPa19rTyXOH5G9Zx44r3k/pTOH6KoS7maRG5y35cCbgfWGaMOQ0gItOBWvb0W4B6Dg0hJUWkhDEmzpWBs9kGvGe3mi0AzgANgIV2Tk/gGIAxZodYLf3zgRbGmGT3RHaqAzDDGBMNYIw5beefZoxJA/aJyN9AHayTPFdYb4zZbz/uiHWCssHO5QdEYS3L60RkDPAD8OtFPm+yw+/RDq87m8c1wMsiUhGYZYzJ7+FduS3vOXaWnQ6tlq2AycaYVOCEiCwDmgFb8zlTfphrjEkEEkVkCdAc2AB8LSJeWPPnivXngDFmrVg9evWAVfby9cb6254FzgNf2q35Ocbhi9VTdDMw3WGf42P/7oB1oov9d4kVkVZkzj8iMj+f5mW9Meaw/ZmbsU46VgLtReQFrMpLCNYJe/bvzFFGRJYCFYwxs+385+3Pvth35ZXT7dYYM1xEemNVWsIdymc/HtQETgHJZP6NfsdqRAJogTVsEawTrPeuImsrYLYxJh5ARGYBji3SbcjcDo+KyG/Z3u9s/+L0mIVVAflGrJ5lg3WymK49VsXkVmPM2auYH0dZjlPGmBUOmZoBS40xJwFE5HuseZ2DdXI71S43EZh1ie3CHZztS68HpopIOaxtfv/FPsDNrmb7dhUB3hKRNljrRAWsxjiw97X24+Zc4fmbK8Ir99EKUC7EGg52C1ZlIME+MO/BajV0xsMum+iahJdmjNkrIk2wroEYidVCuMMY0yKXt1yP1XJ4LXXJg7WDc/YPq7K/5sp/ahXv8FiAb4wxQ7IXEpEbgM7Af4B7gH/l8nnmMh4DGGPMJBFZB3QFfhGRfxtjsp/wXI3clndStjKOvwsDZ8tyuX3g7Ap8JyKjjDHfFnCO9HVHsHoI+2YvICLNsU7QI4AnsU6kHHkAMcaY8OzvzUVB/Z0c14lUoJiI+GL1pjQ1xhwSkeFYPTmZYXIvc7GcOb7rKrM73W5FpDhWqzdYvS1xuRwP0ucpxRiTvm7lR67csl7KxfZ/zvYpTo9ZdoPNEmPMXfYwoKUOk//GGlZaC9h4GZkuKftxSkQcG4quZL01XPl2UdCc7UvHAB8YY+bZ69VwV4e6Annavl3sXqwewibGmBQRiXTIk/04nZvctoX8zKmuMXoNUO4CgTP2wa4O1pCV4kBbEQkWkWJYXffpfsU6UQFARNy+AxbrTmMJxpiJWK2PNwJhItLCnu4lIvXtxz2BUlitax+LSJCbYjuzGLhHREoBiEiI/XpvscbpV8c6KO8pwAxxWEOHcsvXS0RKp+cT61qCUMDDGDMTeAVofJHP6uPwe43D6znmUUSuA/42xnyMdX1Lw6ucN2fz42x5O7Mc6CPW9U5hWOvP+nzOk1/uEBFfe77aYbX8V8Fq9f8C+IrMv5ErrAVaSuY1GcVFpJbdih1ojPkRGEhmL0TGemO3vu+3eyrSrwG5wS63GHjcft1TREpitdp2t+c/AKvClxcX2w7SpZ98RNvf5eyuUE7L2PN1WETutPP72BWSguB0uwXeAb7HGlb1hV3W2fHgUtaSeYyIuMqsy4E77XXEn8xhaI7TI+y/dzmsnhpHzvYvuR2zAoEj9uOHsn3OAazhdt+mHzuulpPjlOM2uA7rmBsq1g1n+gLL7GkeZK5b/YCVl9gu3MHZvtRx+T7ormC5yK/tu6A55gzE2oeniEh7oEou71nPlZ+/Xc7yUIWU9gDl7mdggIhsxTqxXou103oLa6d8FNiJNUYW4GngE7t8MawD0gBXh87meqyLptOAFKyTogtYFZxArJwfisgJrIsKO9otOmOBj7hGds728LwRwDIRScW6Rgmsv8syrB6rAenDZQoowykRWSXWrZ0TgRMO03aKyFDgV7HuNJOC1eOTCPxPMu8+k97SPAH4TEQSsYbJAPjYvToeWAf5dDnmUayLhO8TkRTgOPB6Ps9rbsvbmdn2PGzBaul8wRhz3G45vtasxxqKWBl4wxhzVEQexLoBRQpwDnvomCsYY06KyEPAZBFJH6YzFOugO9duaRWsGzgATAG+EJGnsU467gXG2euelz19C/AMMF5E+mO12j5ujFkjIvPs6QewWu/T911XkjnX7cChTIyIfIE1tCkSa5jhlZS5H/hcRF7H2pZ6X2nOy5HLdjsYa9hVS2NMqojcLSIPYw1hy348uJSBwEQReRZrvbvi5e2QdZOITCCzceFLY8wfDi3Us7F6CbcBe8msJKRztn/J7Zj1LtYQuMFY19xkz7JHRO7FGmbW3RjzV17ny+bsOPWe/V3HRGQIsARrW/jRGDPXfl88UF9EfsdatumVvNy2C5fLZV86HGvZHcFaj6q5I5sz+bV9F7RsOTcAdURkI9YQ+N25vOeIiFzp+dt8YIZYN1Z4yhizIucnq8Iq/UI8dZlEJMAYc85uQZiNdROB2e7OVdTYJwMLjDEz3J2loBSFeVSu47DvKo51gH/UGLPJ3bn+qezlnGiMMSISgXVDhDvckCMSa7hStKu/uyCJyDljTIC7c6jCQ8/flCPtAbpyw0XkFqxu4F+xLsZUSqlr3Xix/pGqL9a1L1r5KVhNgLFiddPEkPv1f0op19DzN5VBe4CUUkoppZRSRYbeBEEppZRSSilVZGgFSCmllFJKKVVkaAVIKaWUUkopVWRoBUgppZRSSilVZGgFSCmllFJKKVVk/B/Oaa2qqRxeLQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 1080x720 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot the correlation matrix\n",
    "c_matrix = df.corr()\n",
    "fig, ax = plt.subplots(figsize=(15, 10))\n",
    "ax = sns.heatmap(c_matrix,\n",
    "                 annot=True,\n",
    "                 linewidths=0.5,\n",
    "                 fmt='.2f',\n",
    "                 cmap='YlGnBu');"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "cp, thalach and slope have the strongest positive correlation with target. exang, oldpeak and ca the strongest negative corr."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 5. Modeling"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "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>age</th>\n",
       "      <th>sex</th>\n",
       "      <th>cp</th>\n",
       "      <th>trestbps</th>\n",
       "      <th>chol</th>\n",
       "      <th>fbs</th>\n",
       "      <th>restecg</th>\n",
       "      <th>thalach</th>\n",
       "      <th>exang</th>\n",
       "      <th>oldpeak</th>\n",
       "      <th>slope</th>\n",
       "      <th>ca</th>\n",
       "      <th>thal</th>\n",
       "      <th>target</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>63</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>145</td>\n",
       "      <td>233</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>150</td>\n",
       "      <td>0</td>\n",
       "      <td>2.3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>37</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>130</td>\n",
       "      <td>250</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>187</td>\n",
       "      <td>0</td>\n",
       "      <td>3.5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>41</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>130</td>\n",
       "      <td>204</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>172</td>\n",
       "      <td>0</td>\n",
       "      <td>1.4</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>56</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>120</td>\n",
       "      <td>236</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>178</td>\n",
       "      <td>0</td>\n",
       "      <td>0.8</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>57</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>120</td>\n",
       "      <td>354</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>163</td>\n",
       "      <td>1</td>\n",
       "      <td>0.6</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   age  sex  cp  trestbps  chol  fbs  restecg  thalach  exang  oldpeak  slope  \\\n",
       "0   63    1   3       145   233    1        0      150      0      2.3      0   \n",
       "1   37    1   2       130   250    0        1      187      0      3.5      0   \n",
       "2   41    0   1       130   204    0        0      172      0      1.4      2   \n",
       "3   56    1   1       120   236    0        1      178      0      0.8      2   \n",
       "4   57    0   0       120   354    0        1      163      1      0.6      2   \n",
       "\n",
       "   ca  thal  target  \n",
       "0   0     1       1  \n",
       "1   0     2       1  \n",
       "2   0     2       1  \n",
       "3   0     2       1  \n",
       "4   0     2       1  "
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Split data into X (independent) & y (dependent)\n",
    "X = df.drop('target', axis=1)\n",
    "y = df['target']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "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>age</th>\n",
       "      <th>sex</th>\n",
       "      <th>cp</th>\n",
       "      <th>trestbps</th>\n",
       "      <th>chol</th>\n",
       "      <th>fbs</th>\n",
       "      <th>restecg</th>\n",
       "      <th>thalach</th>\n",
       "      <th>exang</th>\n",
       "      <th>oldpeak</th>\n",
       "      <th>slope</th>\n",
       "      <th>ca</th>\n",
       "      <th>thal</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>63</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>145</td>\n",
       "      <td>233</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>150</td>\n",
       "      <td>0</td>\n",
       "      <td>2.3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>37</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>130</td>\n",
       "      <td>250</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>187</td>\n",
       "      <td>0</td>\n",
       "      <td>3.5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>41</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>130</td>\n",
       "      <td>204</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>172</td>\n",
       "      <td>0</td>\n",
       "      <td>1.4</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>56</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>120</td>\n",
       "      <td>236</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>178</td>\n",
       "      <td>0</td>\n",
       "      <td>0.8</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>57</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>120</td>\n",
       "      <td>354</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>163</td>\n",
       "      <td>1</td>\n",
       "      <td>0.6</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>298</th>\n",
       "      <td>57</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>140</td>\n",
       "      <td>241</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>123</td>\n",
       "      <td>1</td>\n",
       "      <td>0.2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>299</th>\n",
       "      <td>45</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>110</td>\n",
       "      <td>264</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>132</td>\n",
       "      <td>0</td>\n",
       "      <td>1.2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>300</th>\n",
       "      <td>68</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>144</td>\n",
       "      <td>193</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>141</td>\n",
       "      <td>0</td>\n",
       "      <td>3.4</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>301</th>\n",
       "      <td>57</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>130</td>\n",
       "      <td>131</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>115</td>\n",
       "      <td>1</td>\n",
       "      <td>1.2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>302</th>\n",
       "      <td>57</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>130</td>\n",
       "      <td>236</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>174</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>303 rows × 13 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     age  sex  cp  trestbps  chol  fbs  restecg  thalach  exang  oldpeak  \\\n",
       "0     63    1   3       145   233    1        0      150      0      2.3   \n",
       "1     37    1   2       130   250    0        1      187      0      3.5   \n",
       "2     41    0   1       130   204    0        0      172      0      1.4   \n",
       "3     56    1   1       120   236    0        1      178      0      0.8   \n",
       "4     57    0   0       120   354    0        1      163      1      0.6   \n",
       "..   ...  ...  ..       ...   ...  ...      ...      ...    ...      ...   \n",
       "298   57    0   0       140   241    0        1      123      1      0.2   \n",
       "299   45    1   3       110   264    0        1      132      0      1.2   \n",
       "300   68    1   0       144   193    1        1      141      0      3.4   \n",
       "301   57    1   0       130   131    0        1      115      1      1.2   \n",
       "302   57    0   1       130   236    0        0      174      0      0.0   \n",
       "\n",
       "     slope  ca  thal  \n",
       "0        0   0     1  \n",
       "1        0   0     2  \n",
       "2        2   0     2  \n",
       "3        2   0     2  \n",
       "4        2   0     2  \n",
       "..     ...  ..   ...  \n",
       "298      1   0     3  \n",
       "299      1   0     3  \n",
       "300      1   2     3  \n",
       "301      1   1     3  \n",
       "302      1   1     2  \n",
       "\n",
       "[303 rows x 13 columns]"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0      1\n",
       "1      1\n",
       "2      1\n",
       "3      1\n",
       "4      1\n",
       "      ..\n",
       "298    0\n",
       "299    0\n",
       "300    0\n",
       "301    0\n",
       "302    0\n",
       "Name: target, Length: 303, dtype: int64"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Split into training and test\n",
    "np.random.seed(919)\n",
    "\n",
    "# Split into train & test set\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Since we have 303 rows, it's reasonable to use the standard 80-20% split."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((242, 13), (242,))"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.shape, y_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((61, 13), (61,))"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test.shape, y_test.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "What model should we use? We'll use scikit-learn machine learning map as a guide:\n",
    "\n",
    "[![Scikit-learn ml map](https://scikit-learn.org/stable/_static/ml_map.png)](https://scikit-learn.org/stable/tutorial/machine_learning_map/index.html)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Since we're trying to predict a category, have labeled data and <100k samples, we'll start with a **Linear Support Vector Classifier**, then go for **KNeighbors classifier** and **RandomForest** (Ensemble Classification). We're also gonna try **Logistic Regression**, which doesn't appear in the map. We'll create a function to fit and score all of this models to the data.\n",
    "\n",
    "We'll use a StandardScaler to speed up convergence for SVG and optimize the training."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 8.73903247,  0.46245928,  1.01424187, 17.55384401, 47.67897309,\n",
       "        0.34317429,  0.52257617, 22.3991799 ,  0.45716749,  1.18161596,\n",
       "        0.61999998,  1.01232079,  0.60596042])"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "scaler = StandardScaler().fit(X_train)\n",
    "scaler.scale_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Put models in a dictionary\n",
    "models = {'LinearSVC': LinearSVC(),\n",
    "          'KNeighbors': KNeighborsClassifier(),\n",
    "          'Random Forest': RandomForestClassifier(),\n",
    "          'Logistic Regression': LogisticRegression()}\n",
    "\n",
    "# We'll create a function to fit and score models\n",
    "def fit_and_score(models, X_train, X_test, y_train, y_test):\n",
    "    \"\"\"\n",
    "    Fits and evaluate a ml model, given its type, and train and test data.\n",
    "    \n",
    "    Parameters:\n",
    "        models: a dict of different Sklearn machine learning models\n",
    "        X_train: training data (independent variables)\n",
    "        X_test: testing data (independent variables)\n",
    "        y_train: training labels (dependent variable)\n",
    "        y_test: testing labels (dependent variable)\n",
    "    \"\"\"\n",
    "    \n",
    "    # Set random seed\n",
    "    np.random.seed(919)\n",
    "    \n",
    "    # Make a dict to keep model scores\n",
    "    model_scores = {}\n",
    "    \n",
    "    # Loop through models\n",
    "    for name, model in models.items():\n",
    "        # Adjust data if it's LinearSVC\n",
    "        X_train_adjusted = scaler.transform(X_train) if name == 'LinearSVC' else X_train\n",
    "        X_test_adjusted = scaler.transform(X_test) if name == 'LinearSVC' else X_test\n",
    "        \n",
    "        # Fit model\n",
    "        model.fit(X_train_adjusted, y_train)\n",
    "        \n",
    "        # Evaluate the model and add to model scores\n",
    "        model_scores[name] = model.score(X_test_adjusted, y_test)\n",
    "    \n",
    "    return model_scores"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/jorgepasco/Development/projects/ml-heart-disease-project/env/lib/python3.8/site-packages/sklearn/linear_model/_logistic.py:938: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  n_iter_i = _check_optimize_result(\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "{'LinearSVC': 0.8852459016393442,\n",
       " 'KNeighbors': 0.639344262295082,\n",
       " 'Random Forest': 0.8360655737704918,\n",
       " 'Logistic Regression': 0.8852459016393442}"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Run the function\n",
    "model_scores = fit_and_score(models=models,\n",
    "                             X_train=X_train,\n",
    "                             X_test=X_test,\n",
    "                             y_train=y_train,\n",
    "                             y_test=y_test)\n",
    "\n",
    "model_scores"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Model comparison"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "model_comparison = pd.DataFrame(model_scores, index=['accuracy'])\n",
    "model_comparison.T.plot(kind='bar', color='lightblue');"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Without tuning, the best models are, in order:\n",
    "\n",
    "1. Logistic Regression\n",
    "2. Random Forest\n",
    "3. LinearSVC\n",
    "4. KNeighbors"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Hyperparameter tuning (by hand)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Tuning KNeighbors\n",
    "# Let's create empty list to save the new train and test scores\n",
    "train_scores = []\n",
    "test_scores = []\n",
    "\n",
    "# Try tuning n_neighbors hyperparameter\n",
    "neighbors = range(1, 21)\n",
    "\n",
    "# Setup KNN instance\n",
    "knn = KNeighborsClassifier()\n",
    "\n",
    "# Loop through neightbors list\n",
    "for i in neighbors:\n",
    "    knn.set_params(n_neighbors=i)\n",
    "    \n",
    "    # Fit the model\n",
    "    knn.fit(X_train, y_train)\n",
    "    \n",
    "    # Update training and test score lists\n",
    "    train_scores.append(knn.score(X_train, y_train))\n",
    "    test_scores.append(knn.score(X_test, y_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[1.0,\n",
       " 0.8016528925619835,\n",
       " 0.7851239669421488,\n",
       " 0.7768595041322314,\n",
       " 0.7727272727272727,\n",
       " 0.7644628099173554,\n",
       " 0.7396694214876033,\n",
       " 0.743801652892562,\n",
       " 0.7355371900826446,\n",
       " 0.7272727272727273,\n",
       " 0.7396694214876033,\n",
       " 0.731404958677686,\n",
       " 0.7148760330578512,\n",
       " 0.7231404958677686,\n",
       " 0.7148760330578512,\n",
       " 0.6983471074380165,\n",
       " 0.7024793388429752,\n",
       " 0.7066115702479339,\n",
       " 0.7024793388429752,\n",
       " 0.6818181818181818]"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_scores"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[0.5737704918032787,\n",
       " 0.5901639344262295,\n",
       " 0.639344262295082,\n",
       " 0.639344262295082,\n",
       " 0.639344262295082,\n",
       " 0.6721311475409836,\n",
       " 0.639344262295082,\n",
       " 0.6229508196721312,\n",
       " 0.5901639344262295,\n",
       " 0.639344262295082,\n",
       " 0.6065573770491803,\n",
       " 0.6229508196721312,\n",
       " 0.6065573770491803,\n",
       " 0.6557377049180327,\n",
       " 0.639344262295082,\n",
       " 0.639344262295082,\n",
       " 0.639344262295082,\n",
       " 0.639344262295082,\n",
       " 0.6065573770491803,\n",
       " 0.6229508196721312]"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_scores"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYIAAAEGCAYAAABo25JHAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAgAElEQVR4nO3dd3hUVfrA8e+bRgoJKZQAARKp0oUQCwiILAsW7B3WtSyWddVd/e3adm1bXMu6urq6qKBiwa5YURQVRQihdwQJkNBCOiQhmeT8/jg3MYQkzISZTMK8n+fJkyn33Psmmdz33nPPfY8YY1BKKRW4gvwdgFJKKf/SRKCUUgFOE4FSSgU4TQRKKRXgNBEopVSAC/F3AJ5q3769SU5O9ncYSinVqixdunSfMaZDfe+1ukSQnJxMRkaGv8NQSqlWRUS2NfSedg0ppVSA00SglFIBThOBUkoFuFZ3jUAp1fpUVFSQlZVFWVmZv0M55oWHh5OUlERoaKjbbTQRKKV8Lisri+joaJKTkxERf4dzzDLGkJubS1ZWFikpKW6381nXkIjMEJG9IrKmgfdFRJ4Ukc0iskpEhvkqFqWUf5WVlZGQkKBJwMdEhISEBI/PvHx5jeBFYGIj708Cejtf04BnfBiLUsrPNAk0j6b8nn2WCIwx3wJ5jSxyDvCysRYBsSLS2VfxZGTm8c/PNqBlt5VS6lD+HDXUFdhR63mW89phRGSaiGSISEZOTk6TNrYmu5Bnvt5CTvHBJrVXSrVeubm5DB06lKFDh5KYmEjXrl1rnpeXl7u1jquuuoqNGzf6OFL/8OfF4vrOX+o9XDfGTAemA6SmpjbpkL5PYjQAG3YX0zEmvCmrUEq1UgkJCaxYsQKA++67j7Zt23L77bcfsowxBmMMQUH1Hx/PnDnT53E2xuVyERLim122P88IsoButZ4nATt9tbF+iTEAbNpT7KtNKKVamc2bNzNw4ECuv/56hg0bxq5du5g2bRqpqakMGDCABx54oGbZUaNGsWLFClwuF7Gxsdxxxx0MGTKEk08+mb179x627q+++oohQ4YwdOhQhg0bxoEDBwD4+9//zqBBgxgyZAh33303AMuWLePEE09k8ODBXHDBBRQWFtZs8+6772b06NE89dRT7Nmzh/PPP5/U1FTS0tJYtGiRV34P/jwjmAPcJCKzgROBQmPMLl9tLD4qjA7RbdiwWxOBUv50/4drWbezyKvr7N8lhnvPHtCktuvWrWPmzJk8++yzADz00EPEx8fjcrk47bTTuPDCC+nfv/8hbQoLCxkzZgwPPfQQf/jDH5gxYwZ33HHHIcs88sgjTJ8+nRNPPJH9+/cTHh7Ohx9+yKeffkp6ejoRERHk5dnLqFOmTGH69OmMGjWKu+66iwcffJBHH30UgKKiIr799lsALrnkEv74xz9y0kknkZmZyVlnncWaNfUOzPSIzxKBiLwOjAXai0gWcC8QCmCMeRb4BDgD2AyUAFf5KpZqfTtF6xmBUuoQPXv2ZMSIETXPX3/9dV544QVcLhc7d+5k3bp1hyWCiIgIJk2aBMDw4cNZsGDBYesdOXIkt956K5dffjkXXHABbdu2Zd68eVx99dVEREQAEB8fT25uLmVlZYwaNQqAK6+8kqlTp9as59JLL615PG/evEOuU+Tn51NaWlqzvqbyWSIwxlx2hPcN8Ftfbb8+fROjeXXxNiqrDMFBOpRNKX9o6pG7r0RFRdU8/vHHH3niiSdIT08nNjaWKVOm1DsmPywsrOZxcHAwLpfrsGXuueceJk+ezMcff8yIESP4+uuvMcYcNrzzSCMZa8dnjCE9Pf2Q7XtDQNUa6tspmrKKKnbklfg7FKVUC1RUVER0dDQxMTHs2rWLuXPnNnldW7ZsYfDgwdx5552ccMIJbNy4kQkTJvDCCy9QWloKQF5eHu3btyciIoKFCxcCMGvWLMaMGVPvOsePH8/TTz9d87z6AvjRCqgSE31rjRxKbh91hKWVUoFm2LBh9O/fn4EDB3LccccxcuTIJq/r0UcfZcGCBQQFBTF48GAmTJhAWFgYK1euJDU1ldDQUM4++2wefPBBZs2axQ033EBpaSm9evVqcITS008/zQ033MDMmTNrrmHUTgxNJa3tBqvU1FTT1IlpSspdDLh3Lree3odbxvf2cmRKqYasX7+e448/3t9hBIz6ft8istQYk1rf8gHVNRQZFkL3+Ei9YKyUUrUEVCIAe51gw27vDl1TSqnWLPASQWI0mbkllFVU+jsUpZRqEQIyEVRWGbbk7Pd3KEop1SIEXiLoZEcObdQ7jJVSCgjARJDcPoqw4CA26gVjpZQCAjARhAYH0bNjWz0jUCqAeKMMNcCMGTPYvXu3DyP1j4C6oaxa305tSd/a2Jw5SqljiTtlqN0xY8YMhg0bRmJiordDPExlZSXBwcE+3w4E4BkBQN/EGHYWllFYWuHvUJRSfvbSSy+RlpbG0KFDufHGG6mqqsLlcjF16lQGDRrEwIEDefLJJ3njjTdYsWIFl1xySb1nEo8//jj9+/dnyJAhTJkyBYDi4mKuvPJKBg0axODBg3n//fcBeOWVV2rWfddddwHUlLe+5557SEtLIz09nSVLljBmzBiGDx/OpEmT2LNnj09+B4F5RpDYFoAf9xSTmhzv52iUCjCf3gG7V3t3nYmDYNJDHjdbs2YN7733HgsXLiQkJIRp06Yxe/Zsevbsyb59+1i92sZZUFBAbGws//nPf3jqqacYOnToYet6+OGH2bZtG2FhYRQUFAD27KNDhw6sXr0aYwwFBQVkZWVxzz33kJGRQbt27Rg/fjwfffQREydOpLCwkGHDhvHXv/6VgwcPctpppzFnzhzat2/Pq6++yp///GemT59+dL+regRoIrCT1GzYrYlAqUA2b948lixZQmqqrbxQWlpKt27d+OUvf8nGjRu55ZZbOOOMM5gwYcIR1zVgwACmTJnCOeecw7nnnluz/uqzABEhLi6Or776inHjxtG+fXsALr/8cr799lsmTpxIWFgY5513HmDLRKxdu5bx48cDtqsoKSnJ678DCNBE0KVdONFtQvSCsVL+0IQjd18xxnD11Vfz4IMPHvbeqlWr+PTTT3nyySd55513jngkPnfuXL755hs++OAD/vrXv7JmzRqPy05HRETULG+MYfDgwfXOdeBtAXmNQETokxitQ0iVCnDjx4/nzTffZN++fYAdXbR9+3ZycnIwxnDRRRdx//33s2zZMgCio6MpLj58v1FZWUlWVhbjxo3jkUceIScnh5KSEiZMmMBTTz0F2B17fn4+J510EvPnzyc3NxeXy8Xs2bPrLTvdv39/srOzSU9PB6C8vJy1a9f65PcQkGcEYO8w/njVrnoztlIqMAwaNIh7772X8ePHU1VVRWhoKM8++yzBwcFcc801NfuHf/7znwBcddVVXHvttURERBwyQYzL5eLyyy+nuLiYqqoq/vSnPxEdHc29997LjTfeyMCBAwkODubBBx9k8uTJPPDAA4wdOxZjDGeffTZnnnnmYZPbtGnThrfffpubb76Z4uJiXC4Xt912GwMGeH9in4AqQ13bSwszuXfOWhbfdTqdYsK9EJlSqiFahrp5aRlqN9WepEYppQJZ4CYCp+bQJk0ESqkAF7CJIC4qjI7RbfSMQKlm0tq6oVurpvyeAzYRgO0e2rhHJ6lRytfCw8PJzc3VZOBjxhhyc3MJD/fsumfAjhoC2z00a9E2KqsMwUE6ckgpX0lKSiIrK4ucnBx/h3LMCw8P9/jGs8BOBInRHHRVsS33AMd1aOvvcJQ6ZoWGhpKSkuLvMFQDAr5rCNDJ7JVSAS2gE0HvjtGI6BBSpVRgC+hEEBEWTHJClNYcUkoFtIBOBAB9OrXVmkNKqYAW8Imgb2IMmfsOUFZR6e9QlFLKLzQRdIqmysDmvfv9HYpSSvmFJgJn5JBeJ1BKBaqATwTJCZGEhQTpEFKlVMAK+EQQEhxErw5tdQipUipgBXwiAOiXGK1dQ0qpgKWJAOiTGM3uojIKSyr8HYpSSjU7TQTUumCs1wmUUgFIEwE/T1KjiUApFYh8mghEZKKIbBSRzSJyRz3v9xCRL0VklYh8LSKe1U71ks7twokOD2Hjbp2bQCkVeHyWCEQkGHgamAT0By4Tkf51FnsUeNkYMxh4APiHr+JpjIjoBWOlVMDy5RlBGrDZGPOTMaYcmA2cU2eZ/sCXzuP59bzfbPp0solAZ1BSSgUaXyaCrsCOWs+znNdqWwlc4Dw+D4gWkQQfxtSgfonRFJW52F1U5o/NK6WU3/gyEdQ392Pdw+3bgTEishwYA2QDrsNWJDJNRDJEJMNXU9316aSlJpRSgcmXiSAL6FbreRKws/YCxpidxpjzjTEnAHc7rxXWXZExZroxJtUYk9qhQwefBKs1h5RSgcqXiWAJ0FtEUkQkDLgUmFN7ARFpLyLVMdwJzPBhPI2KjQyjU0wbHUKqlAo4PksExhgXcBMwF1gPvGmMWSsiD4jIZGexscBGEdkEdAL+5qt43NE3MUbPCJRSASfElys3xnwCfFLntb/Uevw28LYvY/BEv8RoXlyYi6uyipBgvddOKRUYdG9XS59O0ZS7qtiWV+LvUJRSqtloIqiln14wVkoFIE0EtfTq2JYg0USglAosmghqCQ8NJjkhShOBUiqgaCKoo29itA4hVUoFFE0EdfTpFE1m7gHKKir9HYpSSjULTQR19EuMxhj4cc9+f4eilFLNQhNBHX10tjKlVIDRRFBHckIUYSFBOkmNUipgaCKoIzhI6N2xLRu1a0gpFSA0EdSjb2K0nhEopQKGJoJ69EuMZk/RQQpKyv0dilJK+ZwmgnroJDVKqUCiiaAe/RJjAB05pJQKDJoI6tEppg0x4SF6RqCUCgiaCOohIvTTSWqUUgFCE0ED+iS2ZeOeYowx/g5FKaV8ShNBA/omxlBc5mJXYZm/Q1FKKZ/SRNAAnaRGKRUojpgIxJoiIn9xnncXkTTfh+ZffTpqzSGlVGBw54zgv8DJwGXO82LgaZ9F1EK0iwylc7twPSNQSh3zQtxY5kRjzDARWQ5gjMkXkTAfx9Ui9OkUrYlAKXXMc+eMoEJEggEDICIdgCqfRtVC9EuMZnPOflyVAfHjKqUClDuJ4EngPaCjiPwN+A74u0+jaiH6JkZT7qoiM/eAv0NRSimfOWLXkDHmVRFZCpwOCHCuMWa9zyNrAX6uObSfXs7FY6WUOtY0mghEJAhYZYwZCGxonpBajl4d2xIksHF3EWcO7uzvcJRSyica7RoyxlQBK0WkezPF06KEhwaT3D5Kh5AqpY5p7owa6gysFZF0oKaz3Bgz2WdRtSD9EqNZt1MnqVFKHbvcSQT3+zyKFqxPp2g+XbOb0vJKIsKC/R2OUkp53RFHDRljvsFeH4h2vtY7rwWEfonRGAM/7tXuIaXUscmdEhMXA+nARcDFwGIRudDXgbUUfZ1JajbojWVKqWOUO11DdwMjjDF7oeaGsnnA274MrKXoHh9JeGgQmzQRKKWOUe7cUBZUnQQcuW62OyYEBwm9O0bryCGl1DHLnTOCz0RkLvC68/wS4FPfhdTy9OkUzYIfc/wdhlJK+YQ7F4v/D/gfMBgYAkw3xvzR14G1JP0So9lbfJD8A+X+DkUppbzuiGcEIpICfGKMedd5HiEiycaYTF8H11L0dSap2bC7mJN7Jvg5GqWU8i53+vrf4tBqo5XOawGjOhFs0usESqljkDuJIMQYU9Mn4jwOiPkIqnWMbkNsZKgOIVVKHZPcSQQ5IlJTTkJEzgH2ubNyEZkoIhtFZLOI3FHP+91FZL6ILBeRVSJyhvuhNx8RoU+naD0jUEodk9xJBNcDd4nIdhHZAfwJuO5IjZzJbJ4GJgH9gctEpH+dxe4B3jTGnABcip0Ws0Xql2hnK8vTC8ZKqWOMO6OGthhjTsLuzPsbY04xxmx2Y91pwGZjzE9Od9Js4Jy6qwdinMftgJ3uh968TuvXkQPlLkY+9BUPfrSO3YVl/g5JKaW8wp0SE7eISAy28ujjIrJMRCa4se6uwI5az7Oc12q7D5giIlnAJ8DvGohhmohkiEhGTo5/xvOf1rcjX/x+NJMGJfLiwkxGPzyfO99dzfbcEr/Eo5RS3uJO19DVxpgiYALQEbgKeMiNdlLPa6bO88uAF40xScAZwCxnMpxDGxkz3RiTaoxJ7dChgxub9o1eHaP518VD+fr2sVw8Iol3lmUx9tH53Dp7uV4/UEq1Wu4kguod+hnATGPMSurfydeVBXSr9TyJw7t+rgHeBDDG/ACEA+3dWLdfdYuP5K/nDuK7P57Gtacex+fr9jDh8W+Z9nIGq7IK/B2eUkp5xJ1EsFREPscmgrkiEs2h9xU0ZAnQW0RSRCQMezF4Tp1ltmPnQkZEjscmglZTy6FjTDh3nXE83/9pHLec3pvFW/OY/NT3TH1hMYt+ysWYuidASinV8siRdlZOV81Q4CdjTIGIJABdjTGrjrhyOxz030AwMMMY8zcReQDIMMbMcUYRPQe0xXYb/dEY83lj60xNTTUZGRnu/GzNbv9BF68u2sZzC7ayb/9BUnvE8dvTejG2bwdE3DmJUkop3xCRpcaY1Hrfa21HrS05EVQrq6jkrYwdPPvNT2QXlNK/cwy/Pa0XEwcmEhykCUEp1fw0EfhJRWUV7y/P5pmvt/DTvgN0i49gVK8OpKXEkZaSQNfYCH+HqJQKEJoI/KyyyvDZmt28vXQHGdvyKS5zAdA1NoIRyXGMSIknLTmeXh3baheSUsonGksEDVYfFZH4xlZqjMk72sACRXCQcObgzpw5uDOVVYYNu4tYsjWPJZn5fL8ll/dX2MFU8VFhpPaIIy0lnhHJ8QzoEkNIcMDMAaSU8pMGzwhEZCv2Am699wMYY47zZWANaY1nBI0xxpCZW8KSrXmkZ+aRvjWP7Xn2JrXIsGCG94hjRLJNDCd0jyU8NNjPESulWiPtGmpl9hSVkb7VJoUlmXls3FOMMSBiK6F2jY2ga1wkXWLDSYqNoGtcBF1iI+gaG0F0eKi/w1dKtUBHlQjEdlpfAaQYYx4Uke5AojEm3fuhHlkgJIK6CksqyNiWx6qsQrILSsnOL2VnYSk7C0qpqDz07xcTHkKX2AiS4mxi6OIkiq6xEXSPjyShbRs//RRKKX862kTwDPYGsnHGmONFJA743BgzwvuhHlkgJoKGVFUZcvYfrEkO2QU2OVQ/zi4orbkwXS05IdJ2NaXEc2JKPN3jI/UCtVIBoEkXi2s50RgzTESWAxhj8p07hZWfBQUJnWLC6RQTzrDucfUuU1RWYc8gCkrZkrOfJZn5zFu/h7eWZgG2q6l61NKI5Hj6JUYTpPc6KBVQ3EkEFc7cAgZARDrgXokJ1QLEhIcS0zmU4zvHcPrxnZg22p5JbM7ZX3MNIn1rHh+v2uUsH0KqkxTSUuIY1DWWsBAduaTUscydRPAk8B7QUUT+BlyInVBGtVJBQXbGtT6doplyUg+MMWTll7Ik0yaGxVvz+GrDXgDCQ4MY2i2WtOR4RvZqT1pKvHYlKXWMcWvUkIj0wxaHE+BLY8x6XwfWEL1G0Dz27T9IhpMUlmTmsW5nEVUGJg/pwl/PG0iMjk5SqlVp0sXilnpDmSYC/yguq2Dm95k88eWPdG4XzhOXDmV4j0Y/IkqpFqSxRNBY5+9SIMP5ngNsAn50Hi/1dpCqZYsOD+Xm03vz5nUnIwIXPfsDT8z7EVelXi5SqrVrMBEYY1Kcu4fnAmcbY9obYxKAs4B3mytA1bIM7xHHJzefyjlDu/L4vE1cOn0RWfk6XadSrZk7w0FGGGM+qX5ijPkUGOO7kFRLFx0eyuOXDOXflwxlw+5iJj2xgDkr604+p5RqLdxJBPtE5B4RSRaRHiJyN5Dr68BUy3fuCV355OZT6dWxLTe/vpzb31rJ/oOuIzdUSrUo7iSCy4AO2CGk72MnsL/Ml0Gp1qN7QiRvXncyN4/rxbvLsjjzyQWs2KHzNivVmrhddE5EYoAqY8x+34bUOB011HKlb83j1tnL2Vt8kD9M6MN1o3vqjGxKtRBNHTVU3XiQU15iNbBWRJaKyEBvB6lav7SUeD69ZTS/HJDIw59tZMrzi9lVWOrvsJRSR+BO0bmFwN3GmPnO87HA340xp/g+vMPpGUHLZ4zhraVZ3DdnLaHBQfzzgkFMHNjZ7faVVYa9xWXsLCglK7+UnQVlZBeUUFjqYmCXGNJS4hnYtR2hOmmPUm472qJzUdVJAMAY87WIRHktOnXMEREuTu3GiOR4bpm9nOtfWcZlad3581nHExkWQllFpa2SWl1Su6CUrFrltXcVlOGqOvQAJTYylKiwED50RidFhAZzQvdYpyaSnbQnMsydj7NSqi53zgjeA5YBs5yXpgCpxphzfRxbvfSMoHUpd1Xxry828b9vt5AQ1QYw7NtffsgyQQKJMeGHTLBT/TjJmVMhqo3dyecUH6wplLckM491u4owBkKChIFd29VM8zkiOY7YyJZRJDe7oJQPVmTzw5Zc+nSKrolP54ZQzelo5yOIA+4HRmFrDX0L3GeMyfd2oO7QRNA6fb95H7N+2EZcVOjPE+Y4O/xOMeFN7uYpKqtg6bZ8Zw7oPFbuKKTcudu5b6doRqTE1Zw1dG4X4c0fqVGFJRV8smYX7y3PJn2rrcbSq2NbduSVcNBVVfO8usrriOR4kuIimy0+FXh0qkoVMMoqKlm5o6CmiuqybfkcKK8EoFt8hN3xOokhpX2UVyupHnRVMn9DDu8vz+arDXspr6ziuA5RnDe0K+cM7Ur3hEgOuipZnVVIemYeS7bmkbEtv2byoK6xEYxIjquZNKhnh7Za6VV5TVOLzs1pbKXGmMleiM1jmgiUJ1yVVazfVczirblOme188g7Yrqn2bcOcbhqbGI7vHOPxcNeqKkPGtnzeW57Nx6t2UlTmon3bNpw9pDPnndCVQV3bNbozr6wybNhd5JzR5JOemUdO8UEA4qPCSO0RR1qKja9/5xhC9AK5aqKmJoIcYAfwOrAY2y1UwxjzjZfjdIsmAnU0jDFsyTlwyKQ82QV2iGt0mxCG1drxDk5qR5uQ4HrX8+OeYt5fkc37y3eSXVBKRGgwEwcmcu4JXRnZM6HJO2xjDJm5JSzZmmfPGjLz2JZbUhPfpWnduPbU4+gUE960X4AKWE1NBMHAL7B3EQ8GPgZeN8as9VWg7tBEoLwtu6D05x3v1jx+3GvvmQwLCWJoUiwjUuJIS0mgR3wk89bv4b3l2azdWUSQwKm9O3DeCV35Rf9ONRe0vW1PURnpW/P4Yt0ePl69i2ARLkpN4voxPekWr9cVlHuO+hqBiLTBJoRHgAeMMf/xboju00SgfC3vQLntRnLOGtbsLKKy1nDWwUntOHdoV84a0pmO0c17ZL49t4RnvtnCO0uzqDSGc4Z04YaxPendKbpZ41CtT5MTgZMAzsQmgWRgDjDDGJPtgzjdoolANbcDB10s257PTzkHGNmrPb06tvV3SOwuLOO5BT/x2uLtlLkq+WX/RH57Wi8GJbXz2TYLSyuICQ/RC9itVFO7hl4CBgKfArONMWt8F6L7NBEo9bO8A+XM/H4rLy7MpLjMxeg+Hfjt2J6ceFzCUa3XGMPmvftJr75nY2seOwvL6BjdhhEpduTViOR4+iZGaz2pVqKpiaAKOOA8rb2QAMYYE+PVKN2kiUCpwxWXVTBr0TZeWLCV3APljEiO48bTejG2Twe3juBdlVWs3VlEunOtJCMzj/ySCgA6RLchLTme/l1i2LSnuCYpAESHhzgjmxJIS4ljUNdYwkJ0ZFNLpPcRKBUgSssreWPJdqZ/+xM7C8sY0CWG357Wi4kDEgmqdeReVlHJ8u0FNaOnlm3Pp8S536JHQuQh91v0SIg8LJlk5ZfUjLpK35rHlhx7zNgmJIih3WI5MSWeESnxDOse57OL6MozmgiUCjDlrireX57NM99sYeu+A/TsEMXUk3qwq6iMJVvzWJ1dSEWlQcTegV1dmiMtJb5JQ1P37T9IRmYe6VvzWZKZx9qdhVQZCA4SBnSJqUkqY/t21DMGP9FEoFSAqqwyfLJ6F0/P38yG3cWEBAmDk9rV9POn9oinXWSo17dbXFbBsu0FNcNyV+wooNxVRed24UwbfRyXjuhORFj992go39BEoFSAq774mxQX6Zcd8EFXJd9v3sezX/9EemYeCVFhXD0qhakn9yAm3PuJSB1OE4FSqsVI35rH0/M3882mHKLDQ7jy5GSuHpVCfFTLqBZ7rNJEoJRqcdZkF/L0/M18tnY34SHBXJbWnWmjjyOxXcson2GMIb+kgux8O3fG3uIy2kWEkhQXQdfYSDpEt2lVQ2f9lghEZCLwBBAMPG+MeajO+48DpzlPI4GOxpjYxtapiUCpY8vmvcX89+stfLBiJ0ECFw635TN6JPh2/itXZRW7i8pqZsCzO/wyZ8KkEnYWlFFaUdlg+9BgIbFdeE1Z9aRa82hUvxYe2nKug/glETi1ijZh6xVlAUuAy4wx6xpY/nfACcaYqxtbryYCpY5NO/JK+N+3W3gzIwtXZRVnD+nCjWN70TexaeUzSspdNUfz2QV2Jrzq5zsLythdVHZI6RCAhKgwuzNvZ3fqNTv5uAg6xrShoKSiZma9uuvcU1RGndXRvm1YzbwbSXGRXJzazW93pvsrEZyMncDml87zOwGMMf9oYPmFwL3GmC8aW68mAqWObXuLynj+u628smgbJeWV/KJ/J246rRdDuv3cWWCMIfdAec1Up9m1pj6t3kFX3xBXLSTo5yP46p1zzSRJzuOjOYKvqKxid2HZodOwFtp5t7MLSsnKK8VguHpUCjeP693s91f4KxFcCEw0xlzrPJ8KnGiMuameZXsAi4AkY8xh52IiMg2YBtC9e/fh27Zt80nMSqmWo6CknBcXZjLz+0wKSytIS46nTWhQzc6+eqa3alFhwfXu3Kt3+h2jw/3ap79v/0Ee/mwDb2ZkkRgTzt1nHs9Zgzs3W+0mfyWCi4Bf1kkEacaY39Wz7J+wSeCw9+rSMwKlAsv+gy5eW7yNtzKyiGwTQtfYn4/qq3f4SbGRxES0joJ4y6qPeXsAABsDSURBVLbn85cP1rAmu4hTeiZw/+QBzVI9tsV3DYnIcuC3xpiFR1qvJgKlVGtXWWV4LX07j87dyIGDLq4amcwt4/vQ1ofdRY0lAl/e670E6C0iKSISBlyKLWNdN7i+QBzwgw9jUUqpFiM4SJh6Ug/m3z6WC4cn8fx3Wxn36Nd8sCIbfwzp91kiMMa4gJuAucB64E1jzFoReUBEas93fBm2zHXruqFBKaWOUnxUGA9dMJj3bhxJYrtwbpm9gkunL2Lj7uJmjUNvKFNKqRagssowe8l2Hpm7keIyF78+JZlbx/cm2kslOPzVNaSUUspNwUHCFSf2YP5tY7k4tRszvt/KuMe+4b3lWT7vLtJEoJRSLUhcVBj/OH8Q7984ki7twvn9Gyu55H+LWL+ryGfb1ESglFIt0JBusbx340geOn8QP+4t5qz/fMfs9O0+2ZZOHaSUUi1UUJBwaVp3Jg5M5LHPNx31XNQN0USglFItXGxkGA+eO9Bn69euIaWUCnCaCJR7SvJg9dtQ6fJ3JEopL9NEoI6seDfMPAPeuQbevgpc5f6OSCnlRZoIVOMKtsPMSfb7iGth/RyYfTlUlPo7MqWUl2giUA3L3QIzJsGBXPjV+3DmY3D2k7B5HrxyIRxs3tvglVK+oYlA1W/PWpgxEVyl8OsPoVuafX34lXDB87D9B3j5HHvtQCnVqmkiUIfLXmqvCQQFw1WfQuchh74/6EK4ZBbsXg0vnQ379/onTqWUV2giUIfK/B5eOgfC29kk0KFv/cv1OxMufwPyfrLXEAqzmjdOpZTXaCJQP/txHrxyAcR0hqs/g/iUxpfvOQ6mvmfPCGZMstcUlFKtjiYCZa2bA69fCu17wa8/gZgu7rXrfhJc+SGU77fdSXvX+zZO5R95P9lkv+AxKCv0dzTKyzQRKFg5G976NXQZCld+BG07eNa+y1C46hP7eOYZsHO510NUflTpgnd+A9kZ8OUD8PhAmHefXhs6hmgiCHRLnof3roPkkTD1fYiIbdp6Oh4PV38KYW3hpcmwfZF341T+s+BRmwTO+x9c9y30Oh2++zf8exB88n/2HhPVqmkiCGTfPwEf3wZ9JsLlb0Gbtke3vvjjbDJo2xFmnQdbvvJOnMp/diyBbx6GwZfCwPPtCLKLXoSbMmDQRZAxE548Ad67HnI2+jta1UQ6VWUgMgbm/x2+fRgGnAfnPwfB3pkOD7BdBrPOg32b7E6j35neW7dqPgf3w7OjoKoSbvjOjiSrqzALFj4FS18EV5n9W5/6B+g6vNnDVY3TqSrVz4yBuXfZJHDCFLjgBe8mAbBnBFd+CImD4I2ptlidan3m3gn5mXD+/+pPAgDtkmDSQ/D7NTD6dshcAM+Nszcbbv3Wft5Ui6eJIJBUVcKHN8Oi/8KJ18PZ/7E3jflCZDz86gPocQq8c609YlStx4aPYdnLMOpW+zc8kqj2MO4euHUN/OIB2LPO3mz4/Hi7rqoq38esmky7hppLVZW9Y7fyoP9iyJgJa96GU2+DcX8GEd9vs6LUnhVs/gIm/BVOvql5tuttxkD+VohLaZ3xe6J4DzxzMsR0hWu/hJAwz9dRUQYrXrHXoQq2Q4fj4ZTfQVwP78fbXILbQNdhvjt48rHGuoY0ETSXj2+HJc/5Owo4/V7bh9ucXOW2hPX6OZA42G7/+Mmt6x9q4X/g83tsIpvw12M3GRgDr15ku3iu+7bhO8vdVemCNe/Ad49DzjFwj0m/s+DCGRDSxt+ReEwTgb9t+hxeuwiG/xoGXuC/OCIToNMA/2y7qhJWvAbf/xtyN0NCLxh5Kwy+pGlHnM1p92rb7x2ZAMW77N/xzH+1rkTmrvTn4JPb4YxHIe033ltvVRXsXAYVJd5bZ3PbkQ5fPWjvqL/kVQiL9HdEHtFE4E/7c+xpdlRHmDa/VR5JeFVVJaz/0N6hunuV7X445Xcw7FcQFuXv6A5XUQbTx0JpHtzwA/zwH3t0O+hiOPcZCD6Gpv3O2Qj/Gw3Jo+CKt4/ds56jsWyWvc7W7SRbays8xt8RuU1HDfmLMfZDU1YIFzynSQDsUfSAc223w5R3IC4ZPrvD3q36zcNQmu/vCA/15f22S+Oc/0JUAoy/D07/C6x+E966Elx+vObjTa5yePc3EBoJ5zytSaAhw6baMuxZ6fDy5GOmDLsmAl9a9jJs/MTuPPzVJdNSiUCv8bY0xdVzIWkEzP+bTQif/9lOj+lvW+bbEVZp06D3+J9fP/U2mPQwbPgIXrsEyg/4L0Zv+eYh2LUSJj8J0Yn+jqZlG3iB7Rrasw5ePNNeXG/ltGvIV3K32JtxkkbY0g1BmnOPaPdq2+2y9j0ICoUTroBTbj5yFVRfKMmDZ06BNjEw7ev6+4OXvwJzfgfdTnS6CRoYa9/SbVtoa0SdMAXOecrf0bQeP30Dr18G0Z3sUOnY7v6OqFHaNdTcKivsaXZwmO1H1iTgnsRBdkTGTRkw5FK7o/3PcFvwbM+65ovDGPjoVjiwD86f3vBFweob8rKW2PpKrbGboKwQ3r3OdtFNfMjf0bQux42xU7geyG31Zdh1D+UL3z5q7xk463Fo19Xf0bQ+CT1tF8UtK+GkG+wNSc+cDLOvsP90vrZyNqz7AE67y1ZWbczA8+HS12z57Zln+KZLyxjYugA+u9N+9+ZZ/Kd/gqJsW2bkaGtNBaJuafDrj2x5jRkT7RSvvrB7Dbx9tc8q+2oi8LYdS+DbR34u0qWaLqYL/PJvtnzBmDtg8zx48Qwo2uW7beZvsxU1u58CI29xr02fX8KUt+2NUzMmeq8apzGw8VN4YQK8dBYsesZ+f+EXsOGTo79bd+17sPJ1Wxqi2wjvxByIOg+2s/kFBdtrBtlLvbfu7Yvh1Yvh2ZGwaS7s+9F7665FrxF4kztFulTTbV1gJ8+Jag+/muP9u1SrKu0/8p61cP13nq9/xxJ49QJbivtXc+wkP01R6bI76e8eh71rbd/zKTfbuaLXvPPz3bod+8OoP9jCgZ4OYy3aCf892Z59XT3X+/WmAlHeVltjqSTPXjNKHtm09RgDW76EBf+Cbd9DRLw9M077DUTENTk8vY+gucz5nR1nfNUn7tVnUZ7LyoBXzofQKLhyDrTv7b11L3jMTrxy3nQYcknT1rFrla28KkF2Gs/Ege63rSiDla/ZHX1+JnToZ3f0A88/dEdd927duGR79jLkcggNP/J2qqrglfPsDVLXf2eTgfKOop02GRTsgEtfsSPj3FX3HpvoLjDyZq/dY6OJoDms/wjeuML+446/19/RHNt2r4aXz7VDUKe+Zy8yH62dy22BtOMn2wvWRzOOPmeT3RlUlMCUdyHpCCWZDxbbOlA/PA37d9sSzqfeBn0mNT7QoKoKNn1qdxzZS6FtJ1sCI/UqaBPdcLsf/msri579hL1LWnnXgX0w61zYu8F+lvpPbnx5V7m9L+W7x+1d9/E9YdTvvX7XvSYCX6su0tUuCa6Z1/JLJhwL9v1od7bl++GKd46uj7u8xN5RW34Ablx4VKffNfIzbXwH9jndBKMOX6YkDxY/C4v/B2UFkDLGJoCU0Z4lImNsyefv/gU/fQ3hsXDidbbCbGT8ocvuWQvTT7OzjF36mt445iulBbZmU/ZSOPe/dhRcXeUl9l6jhf+Boix7QHPqbT6rw6WJwJe8XaRLuS9/m727c3+O3dmmnNq09VQXBPzVB3DcWO/FV7TL6SbYBpe8Ar1/4by+8+fJXCoO2EJmo/5w5DMHd2QvtX3LGz6ydwkP/7U9S2jX1XY9PTcODuTAjT/Yay3Kdw7uh9mX2SR95r9gxDX29dIC+3lb9AyU5NqBCafeZpOzDxOzJgJf8lWRLuWeol32NDw/Ey6eBX0meNa+uiDgyTfZEUredmCfvWawd72dwGXXSljxOpgqO9XjqFvtfM/etneDLfC36k17vaL6iHT5LLj8TTvSSfleRZktRbLpMxh7l+0uXPIClBdD7wn2AKDHyc0SiiYCX9EiXS3DgVx78XPPOlsHZsC5brbbZ0fORHWA33zl3oXWpigtgNcuhh2LbU37YVPtKKDmqM2fv812PSyfZce6p14DZ/3L99tVP6usgHenwdp3bVLuf669BtB5cLOG4bdEICITgSeAYOB5Y8xhty6KyMXAfYABVhpjLm9snS0mEbjK4YXxds7WG36wt5kr/ykrtOOts9Jt0bShjX6MbJfe7CvshDnTvvZ9LajyA7DmXXsU6I/Pyv698OPntk5OaETzbz/QVVXas7NuaX4bpdVYIvBZDV0RCQaeBn4BZAFLRGSOMWZdrWV6A3cCI40x+SLS0VfxeN3X/7Cn+Ze8qkmgJQhvB1PfhdmXw/s32B1vY111y16GjR/DhL81T0HAsCh7JuAvbTvakhjKP4KCYehl/o6iQb68szgN2GyM+ckYUw7MBs6ps8xvgKeNMfkAxpi9PozHe7YttEO9TpgKx5/l72hUtbAouOwN6HuGvW6zoIEukNwttlxDymg46cbmjVGpFsiXiaArsKPW8yzntdr6AH1E5HsRWeR0JR1GRKaJSIaIZOTk5PgoXDdpka6WLTQcLn4ZBl5o5xL48oFDa/NUumx/bXAInPusFgRUCh92DQH1XTmte0EiBOgNjAWSgAUiMtAYU3BII2OmA9PBXiPwfqgeqC7SdfVcLdLVUgWH/lw1dMFjtpvol/+wO/0Fj0J2Blw4UwsCKuXwZSLIArrVep4E7KxnmUXGmApgq4hsxCaGJT6Mq+nWvGuLdI25Q4t0tXRBwXD2kxAWDYuetmO6h021s6BpQUClDuHLRLAE6C0iKUA2cClQdyjH+8BlwIsi0h7bVfSTD2NqupyN8NHv7e3/o2/3dzTKHSL23oA2beGbf8KqN+wcyWc87O/IlGpRfJYIjDEuEbkJmIsdPjrDGLNWRB4AMowxc5z3JojIOqAS+D9jTDMUnPfArlX21v11H9iqkuc/p5UaWxMRO69AWFs7T8R5z2pVWKXq0BvKGrJtoR11svkL272Qdq0dYdK29YxwVXVUVfqkhotSrYFf7iNolYyxk58seAy2/wCR7WHcn2HEtRAR6+/o1NHSJKBUvTQRgD1SXPe+vTdg92qISYJJD9v7BBqar1YppY4RgZ0IXAft/LTfPwF5WyChN5zzX1sMTEtJK6UCRGAmgvIDtgTwwqegeCd0HmpvQup3lnYfKKUCTmAlgpI8WzZ68bNQmgfJp8I5T0HPcVo5VCkVsAInESx72daXKd9vpwA89Q+2EqBSSgW4wEkEsT2g7yRbB7w5qk0qpVQrETiJ4Lgx9ksppdQhtPSiUkoFOE0ESikV4DQRKKVUgNNEoJRSAU4TgVJKBThNBEopFeA0ESilVIDTRKCUUgGu1U1MIyI5wLYmNm8P7DuKzWt7be/P9i0hBm3fetv3MMZ0qPcdY0zAfGGnyNT22r5Vtm8JMWj71t2+oS/tGlJKqQCniUAppQJcoCWC6dpe27fi9i0hBm3futvXq9VdLFZKKeVdgXZGoJRSqg5NBEopFeACIhGIyAwR2Ssia5rYvpuIzBeR9SKyVkRu8bB9uIiki8hKp/39TYwjWESWi8hHTWibKSKrRWSFiGQ0oX2siLwtIhuc38PJHrTt62y3+qtIRG71cPu/d353a0TkdREJ97D9LU7bte5su77PjIjEi8gXIvKj8z3Ow/YXOduvEpHUJmz/Eef3v0pE3hORWA/bP+i0XSEin4tIF0/a13rvdhExItLew+3fJyLZtT4HZ3i6fRH5nYhsdH6PD3u4/TdqbTtTRFZ42H6oiCyq/h8SkQbnum2g/RAR+cH5P/xQRGIaaV/vPseTz6BHfDEmtaV9AaOBYcCaJrbvDAxzHkcDm4D+HrQXoK3zOBRYDJzUhDj+ALwGfNSEtplA+6P4Hb4EXOs8DgNim7ieYGA39uYWd9t0BbYCEc7zN4Ffe9B+ILAGiMTOyjcP6O3pZwZ4GLjDeXwH8E8P2x8P9AW+BlKbsP0JQIjz+J9N2H5Mrcc3A8960t55vRswF3tTZ4Ofpwa2fx9wu5t/s/ran+b87do4zzt6Gn+t9x8D/uLh9j8HJjmPzwC+9rD9EmCM8/hq4MFG2te7z/HkM+jJV0CcERhjvgXyjqL9LmPMMudxMbAeu3Nyt70xxux3noY6Xx5dpReRJOBM4HlP2nmDc+QyGngBwBhTbowpaOLqTge2GGM8vTs8BIgQkRDsDn2nB22PBxYZY0qMMS7gG+C8xho08Jk5B5sQcb6f60l7Y8x6Y8xGdwJuoP3nTvwAi4AkD9sX1XoaRSOfwUb+Zx4H/thY2yO0d0sD7W8AHjLGHHSW2duU7YuIABcDr3vY3gDVR/HtaOQz2ED7vsC3zuMvgAsaad/QPsftz6AnAiIReJOIJAMnYI/qPWkX7JyK7gW+MMZ41B74N/YfsMrDdtUM8LmILBWRaR62PQ7IAWY6XVPPi0hUE+O4lEb+AetjjMkGHgW2A7uAQmPM5x6sYg0wWkQSRCQSezTXzZMYHJ2MMbucmHYBHZuwDm+5GvjU00Yi8jcR2QFcAfzFw7aTgWxjzEpPt1vLTU731IwmdGv0AU4VkcUi8o2IjGhiDKcCe4wxP3rY7lbgEef39yhwp4ft1wCTnccX4eZnsM4+xyefQU0EHhCRtsA7wK11jq6OyBhTaYwZij2KSxORgR5s9yxgrzFmqUcBH2qkMWYYMAn4rYiM9qBtCPY09xljzAnAAexpqUdEJAz7j/CWh+3isEdCKUAXIEpEprjb3hizHtuV8gXwGbAScDXaqAUTkbux8b/qaVtjzN3GmG5O25s82GYkcDceJo86ngF6AkOxCf0xD9uHAHHAScD/AW86R/eeugwPD0YcNwC/d35/v8c5Q/bA1dj/vaXY7p7yIzU4mn2OJzQRuElEQrF/kFeNMe82dT1Ol8rXwEQPmo0EJotIJjAbGCcir3i43Z3O973Ae0CDF7rqkQVk1TqLeRubGDw1CVhmjNnjYbvxwFZjTI4xpgJ4FzjFkxUYY14wxgwzxozGnrJ7ejQIsEdEOgM43xvsmvAVEbkSOAu4wjgdxU30Go10TdSjJzYRr3Q+h0nAMhFJdHcFxpg9zgFRFfAcnn0GwX4O33W6WtOxZ8cNXrCuj9O1eD7whofbBrgS+9kDezDjUfzGmA3GmAnGmOHYRLTlCLHWt8/xyWdQE4EbnKOOF4D1xph/NaF9h+oRHiISgd2xbXC3vTHmTmNMkjEmGdu18pUxxu0jYhGJEpHo6sfYi45uj6AyxuwGdohIX+el04F17ravpalHYtuBk0Qk0vlbnI7tM3WbiHR0vnfH7giaEscc7M4A5/sHTVhHk4nIROBPwGRjTEkT2veu9XQynn0GVxtjOhpjkp3PYRb2YuZuD7bfudbT8/DgM+h4HxjnrKsPdtCCp5U4xwMbjDFZHrYDe01gjPN4HB4eTNT6DAYB9wDPNrJsQ/sc33wGvXHFuaV/Yf/pdwEV2A/wNR62H4XtY18FrHC+zvCg/WBgudN+DY2MVnBjXWPxcNQQto9/pfO1Fri7CdsdCmQ4P8P7QJyH7SOBXKBdE3/u+7E7rjXALJyRIx60X4BNXiuB05vymQESgC+xO4AvgXgP25/nPD4I7AHmeth+M7Cj1mewsVE/9bV/x/n9rQI+BLo29X+GI4xCa2D7s4DVzvbnAJ09bB8GvOL8DMuAcZ7GD7wIXN/Ev/8oYKnzGVoMDPew/S3Y0T+bgIdwKjs00L7efY4nn0FPvrTEhFJKBTjtGlJKqQCniUAppQKcJgKllApwmgiUUirAaSJQSqkAp4lA+ZVTxfKxWs9vF5H7vLTuF0XkQm+s6wjbucipEjnfC+t6QETGH2GZ+0Tk9npeT65brVMpd2giUP52EDhfGilp7A8iEuzB4tcANxpjTjva7Rpj/mKMmXe062kKD39mdQzRRKD8zYWdh/X3dd+oe0QvIvud72OdomNvisgmEXlIRK4QO+fDahHpWWs140VkgbPcWU77YLG1/Zc4BdCuq7Xe+SLyGvbGp7rxXOasf42I/NN57S/Ym3+eFZFH6iw/VkS+lp/ncXi1ujaOiAx3foalIjK3VtmAmp9ZRM5w2n0nIk/KofNQ9HfW/ZOI3Fzr9RARecn5ud52agQhIqeLLRi4WmzBtzbO65ki8hcR+Q64SERuFpF1TvvZbvz91LHAG3el6Zd+NfUL2I8t7ZuJLe17O3Cf896LwIW1l3W+jwUKsDXb2wDZwP3Oe7cA/67V/jPsAU9v7B2e4cA04B5nmTbYO6ZTnPUeAFLqibMLttRFB2zxs6+Ac533vqae+QWc9RVi6/IEAT9gk0YosBDo4Cx3CTCj9s/sxLmjOhbsnaofOY/vc9q3wdbayXXWmYy9G3Wks9wM5/dZva4+zusvY4uY4fze/1gr5p38XO+/SXNO6Ffr+9IzAuV3xlZVfBk7WYq7lhhbs/0gtnhXdVnq1dgdYrU3jTFVxpYc/gnoh6219CuxZcEXY2/br67Dk26M2VrP9kZgJyLJMXZOgFexczQcSboxJsvYQmsrnNj6YifL+cKJ4R4On1ugH/BTrVjq1kb62Bhz0BizD1t4rJPz+g5jzPfO41ewiacvtmjfJuf1l+rEXrsA2yrgVbHVXVtthVblmRB/B6CU49/Y+jEza73mwum+dLpUwmq9d7DW46paz6s49HNdt4aKwc4Y9ztjzNzab4jIWOwZQX2aUu64bpyVTmwCrDXGNDbd55G2V996oeGftzG1f+YzsUliMvBnERlgfp4MRx2j9IxAtQjGmDzsFJTX1Ho5ExjuPD4H2/3hqYtEJMi5bnAcsBE71eINTplfRKSPHHmincXAGBFp71xUvQw701lTbAQ6iDPvs4iEisiAOstsAI4TOykJ2O4jd3SXn+eTvgz4zllXsoj0cl6fWl/sTlXMbsaY+dhJkGKBtm5uV7ViekagWpLHOHSylOeAD0QkHVtpsaGj9cZsxO70OmGrTpaJyPPYLpplzplGDkeY8s8Ys0tE7gTmY4+wPzHGNKkEsDGm3Lkg/KSItMP+H/4bWxm2eplSEbkR+ExE9gHpbq5+PXCliPwPW6HyGednvgp4S2w9/iXUXwI5GHjFiUmAx03TpyRVrYhWH1WqhRKRtsaY/U6yehr40RjzuL/jUsce7RpSquX6jXMxeS12RNX//ByPOkbpGYFSSgU4PSNQSqkAp4lAKaUCnCYCpZQKcJoIlFIqwGkiUEqpAPf/zPW8Asm9MnIAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(neighbors, train_scores, label=\"Train score\")\n",
    "plt.plot(neighbors, test_scores, label=\"Test score\")\n",
    "plt.xlabel(\"Number of neighbors\")\n",
    "plt.ylabel(\"Model score\")\n",
    "plt.xticks(np.arange(1, 21, 1))\n",
    "plt.legend();"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The max KNN score on the test data is 67.21%\n",
      "It improved by 3.28%\n"
     ]
    }
   ],
   "source": [
    "max_score = max(test_scores)\n",
    "improvement = (max_score - model_scores['KNeighbors'])\n",
    "print(f\"The max KNN score on the test data is {max_score * 100:.2f}%\")\n",
    "print(f\"It improved by {improvement * 100:.2f}%\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Tuning Hyperparameters with RandomizedSearchCV\n",
    "\n",
    "We're going to tune the rest of the models, with Randomized Search Cross Validation:\n",
    "\n",
    "- LinearSVC\n",
    "- LogisticRegression\n",
    "- RandomForestClassifier\n",
    "\n",
    "#### Creating Hyperparameter grids"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'C': 1.0,\n",
       " 'class_weight': None,\n",
       " 'dual': True,\n",
       " 'fit_intercept': True,\n",
       " 'intercept_scaling': 1,\n",
       " 'loss': 'squared_hinge',\n",
       " 'max_iter': 1000,\n",
       " 'multi_class': 'ovr',\n",
       " 'penalty': 'l2',\n",
       " 'random_state': None,\n",
       " 'tol': 0.0001,\n",
       " 'verbose': 0}"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "LinearSVC().get_params()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Linear SVC\n",
    "lin_svc_grid = {'loss': ['hinge', 'squared_hinge'],\n",
    "                'C': list(range(1, 100, 5)),\n",
    "                'max_iter': [1000000]}\n",
    "\n",
    "# Random Forest Classifier\n",
    "rf_grid = {'n_estimators': np.arange(10, 1000, 50),\n",
    "           'max_depth': [None, 3, 5, 10],\n",
    "           'min_samples_split': np.arange(2, 20, 2),\n",
    "           'min_samples_leaf': np.arange(1, 20, 2)}\n",
    "\n",
    "# Logistic regression\n",
    "log_reg_grid = {'C': np.logspace(-4, 4, 20),\n",
    "                'solver': ['liblinear']}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Tuning models\n",
    "\n",
    "First, we'll create a function that can be reused for tuning all models, and another one to output the improvement."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "def tune_model(model, model_grid, X_train, y_train):\n",
    "    \"\"\"\n",
    "    Tune a ml model with RandomizedSarchCV, given its type,\n",
    "    tuning model grid and train data; and return the new model.\n",
    "    \n",
    "    Parameters:\n",
    "        model: The model to be tuned.\n",
    "        model_grid: Tuning model grid for the specific model\n",
    "        X_train: training data (independent variables)\n",
    "        y_train: training labels (dependent variable)\n",
    "    \"\"\"\n",
    "    \n",
    "    # Random seed\n",
    "    np.random.seed(919)\n",
    "    \n",
    "    # Setup random hyperparameter search\n",
    "    rs_model = RandomizedSearchCV(model,\n",
    "                                  param_distributions=model_grid,\n",
    "                                  cv=5,\n",
    "                                  n_iter=20,\n",
    "                                  verbose=True)\n",
    "    # Fit hyperparameter search model\n",
    "    rs_model.fit(X_train, y_train)\n",
    "    \n",
    "    return rs_model\n",
    "\n",
    "\n",
    "def output_change(original_score, new_score):\n",
    "    print(f\"The model went from {original_score} to {new_score}\")\n",
    "    change = new_score - original_score\n",
    "    print(f\"Which makes for an {'improvement' if change >= 0 else 'downgrade'} of {change * 100:.2f}%\")\n",
    "    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Linear SVC"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 20 candidates, totalling 100 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.\n",
      "[Parallel(n_jobs=1)]: Done 100 out of 100 | elapsed:   13.5s finished\n"
     ]
    }
   ],
   "source": [
    "rs_lin_svc = tune_model(LinearSVC(),\n",
    "                        lin_svc_grid,\n",
    "                        scaler.transform(X_train),\n",
    "                        y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'max_iter': 1000000, 'loss': 'hinge', 'C': 1}"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rs_lin_svc.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9016393442622951"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rs_lin_svc_score = rs_lin_svc.score(scaler.transform(X_test), y_test)\n",
    "rs_lin_svc_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The model went from 0.8852459016393442 to 0.9016393442622951\n",
      "Which makes for an improvement of 1.64%\n"
     ]
    }
   ],
   "source": [
    "output_change(model_scores['LinearSVC'], rs_lin_svc_score)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Replace scores in the model_scores dict\n",
    "model_scores['LinearSVC'] = rs_lin_svc_score"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Random Forest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 20 candidates, totalling 100 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Done 100 out of 100 | elapsed:  1.7min finished\n"
     ]
    }
   ],
   "source": [
    "rs_rf = tune_model(RandomForestClassifier(),\n",
    "                   rf_grid,\n",
    "                   X_train,\n",
    "                   y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'n_estimators': 60,\n",
       " 'min_samples_split': 6,\n",
       " 'min_samples_leaf': 17,\n",
       " 'max_depth': 3}"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rs_rf.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.819672131147541"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rs_rf_score = rs_rf.score(X_test, y_test)\n",
    "rs_rf_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The model went from 0.8360655737704918 to 0.819672131147541\n",
      "Which makes for an downgrade of -1.64%\n"
     ]
    }
   ],
   "source": [
    "output_change(model_scores['Random Forest'], rs_rf_score)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Replace scores in the model_scores dict\n",
    "model_scores['Random Forest'] = rs_rf_score"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Logistic Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 20 candidates, totalling 100 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Done 100 out of 100 | elapsed:    0.5s finished\n"
     ]
    }
   ],
   "source": [
    "rs_log_reg = tune_model(LogisticRegression(),\n",
    "                        log_reg_grid,\n",
    "                        X_train,\n",
    "                        y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'solver': 'liblinear', 'C': 0.23357214690901212}"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rs_log_reg.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9016393442622951"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rs_log_reg_score = rs_log_reg.score(X_test, y_test)\n",
    "rs_log_reg_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The model went from 0.8852459016393442 to 0.9016393442622951\n",
      "Which makes for an improvement of 1.64%\n"
     ]
    }
   ],
   "source": [
    "output_change(model_scores['Logistic Regression'], rs_log_reg_score)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Replace scores in the model_scores dict\n",
    "model_scores['Logistic Regression'] = rs_log_reg_score"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Tuning Hyperparameters with GridSearchCV\n",
    "\n",
    "Since none of the models reached the desired 95% accuracy, we're gonna try with an exhaustive search Cross Validation with GridSearchCV.\n",
    "\n",
    "We'll use the same grid, and we'll create a functions similar to the ones previously used."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [],
   "source": [
    "def tune_model_exhaustive(model, model_grid, X_train, y_train):\n",
    "    \"\"\"\n",
    "    Tune a ml model with GridSearchCV, given its type,\n",
    "    tuning model grid and train data; and return the new model.\n",
    "    \n",
    "    Parameters:\n",
    "        model: The model to be tuned.\n",
    "        model_grid: Tuning model grid for the specific model\n",
    "        X_train: training data (independent variables)\n",
    "        y_train: training labels (dependent variable)\n",
    "    \"\"\"\n",
    "    \n",
    "    # Random seed\n",
    "    np.random.seed(919)\n",
    "    \n",
    "    # Setup random hyperparameter search\n",
    "    gs_model = GridSearchCV(model,\n",
    "                            param_grid=model_grid,\n",
    "                            cv=5,\n",
    "                            verbose=True)\n",
    "    # Fit hyperparameter search model\n",
    "    gs_model.fit(X_train, y_train)\n",
    "    \n",
    "    return gs_model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We'll create a dictionary to link the models to its correspondent grid, so we can iterate over all of them easily. We'll also save the model on each iteration."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [],
   "source": [
    "grid_dict = {'LinearSVC': lin_svc_grid,\n",
    "             'Random Forest': rf_grid,\n",
    "             'Logistic Regression': log_reg_grid}\n",
    "\n",
    "final_models = {}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Tuning LinearSVC model\n",
      "Fitting 5 folds for each of 40 candidates, totalling 200 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Done 200 out of 200 | elapsed:   31.4s finished\n",
      "[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best params: {'C': 1, 'loss': 'hinge', 'max_iter': 1000000}\n",
      "Tuned model score: 0.9016393442622951\n",
      "The model went from 0.9016393442622951 to 0.9016393442622951\n",
      "Which makes for an improvement of 0.00%\n",
      "\n",
      "Tuning Random Forest model\n",
      "Fitting 5 folds for each of 7200 candidates, totalling 36000 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Done 36000 out of 36000 | elapsed: 488.2min finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best params: {'max_depth': 5, 'min_samples_leaf': 15, 'min_samples_split': 18, 'n_estimators': 10}\n",
      "Tuned model score: 0.8524590163934426\n",
      "The model went from 0.819672131147541 to 0.8524590163934426\n",
      "Which makes for an improvement of 3.28%\n",
      "\n",
      "Tuning Logistic Regression model\n",
      "Fitting 5 folds for each of 20 candidates, totalling 100 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best params: {'C': 0.23357214690901212, 'solver': 'liblinear'}\n",
      "Tuned model score: 0.9016393442622951\n",
      "The model went from 0.9016393442622951 to 0.9016393442622951\n",
      "Which makes for an improvement of 0.00%\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Done 100 out of 100 | elapsed:    0.4s finished\n"
     ]
    }
   ],
   "source": [
    "# Calling the function for every model\n",
    "for model_name, model in models.items():\n",
    "    if model_name not in grid_dict.keys():\n",
    "        continue\n",
    "        \n",
    "    X_train_adjusted = scaler.transform(X_train) if model_name == 'LinearSVC' else X_train\n",
    "    X_test_adjusted = scaler.transform(X_test) if model_name == 'LinearSVC' else X_test\n",
    "    \n",
    "    print(f\"Tuning {model_name} model\")\n",
    "    tuned_model = tune_model_exhaustive(model,\n",
    "                                        grid_dict[model_name],\n",
    "                                        X_train_adjusted,\n",
    "                                        y_train)\n",
    "    \n",
    "    # Export model\n",
    "    dump(tuned_model, filename=f\"./models/{model_name}.joblib\")\n",
    "    \n",
    "    # Output best params\n",
    "    print(f\"Best params: {tuned_model.best_params_}\")\n",
    "    \n",
    "    # Calculate and output score\n",
    "    tuned_model_score = tuned_model.score(X_test_adjusted,\n",
    "                                          y_test)\n",
    "    print(f\"Tuned model score: {tuned_model_score}\")\n",
    "    \n",
    "    # Output change\n",
    "    output_change(model_scores[model_name], tuned_model_score)\n",
    "    print(\"\") # Blank line for visibility purposes\n",
    "    \n",
    "    # Add the model to tuned_model dict\n",
    "    final_models[model_name] = {'model': tuned_model,\n",
    "                                'score': tuned_model_score}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The best model(s) is/are ['LinearSVC', 'Logistic Regression'] with a score of 0.9016393442622951\n"
     ]
    }
   ],
   "source": [
    "max_score = max([model['score'] for model in final_models.values()])\n",
    "best_model = [el for el in final_models.keys() if final_models[el]['score'] == max_score]\n",
    "print(f\"The best model(s) is/are {best_model} with a score of {max_score}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Since LinearSVC and Logistic Regression had the same accuracy score, any of those could be used for further exploration. In this case, we're gonna keep the Logistic Regression, since it's a model that works better with identified factors, whereas SVM models work better with unstructured – or semi-unstructured – data.\n",
    "\n",
    "### Using more metrics to evaluate the tuned classifier\n",
    "\n",
    "We'll use:\n",
    "\n",
    "- ROC curve and AUC score\n",
    "- Confusion Matrix\n",
    "- Classification report\n",
    "- Precision, Recall, F1-score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([1, 0, 0, 1, 0, 0, 0, 0, 1, 1, 0, 1, 0, 1, 0, 1, 1, 1, 1, 1, 1, 0,\n",
       "       1, 1, 1, 0, 1, 1, 0, 1, 0, 1, 1, 0, 0, 1, 0, 1, 1, 0, 0, 0, 1, 1,\n",
       "       1, 1, 1, 1, 0, 0, 1, 1, 1, 1, 0, 0, 0, 1, 1, 1, 0])"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Save the model in its own variable\n",
    "gs_log_reg = final_models['Logistic Regression']['model']\n",
    "y_preds = gs_log_reg.predict(X_test)\n",
    "y_preds"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([1, 0, 0, 1, 0, 0, 1, 0, 1, 1, 0, 1, 0, 1, 0, 1, 1, 0, 1, 1, 1, 0,\n",
       "       1, 1, 1, 0, 1, 0, 1, 0, 0, 1, 1, 0, 0, 1, 0, 1, 0, 0, 0, 0, 1, 1,\n",
       "       1, 1, 1, 1, 0, 0, 1, 1, 1, 1, 0, 0, 0, 1, 1, 1, 0])"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.array(y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### ROC Curve"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot ROC curve and calculate the Area under the Curve\n",
    "plot_roc_curve(gs_log_reg, X_test, y_test);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Confusion Matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[23  4]\n",
      " [ 2 32]]\n"
     ]
    }
   ],
   "source": [
    "print(confusion_matrix(y_test, y_preds))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_confusion_matrix(gs_log_reg, X_test, y_test, cmap=plt.cm.Blues);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Classification report"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.92      0.85      0.88        27\n",
      "           1       0.89      0.94      0.91        34\n",
      "\n",
      "    accuracy                           0.90        61\n",
      "   macro avg       0.90      0.90      0.90        61\n",
      "weighted avg       0.90      0.90      0.90        61\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(classification_report(y_test, y_preds))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Accuracy, Precision, Recall and f1-score using cross-validation\n",
    "\n",
    "We're going to calculate recall and f1-score with cross_val_score()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'C': 0.23357214690901212, 'solver': 'liblinear'}"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Check best hyperparameters\n",
    "gs_log_reg.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create a new classifier with this paramas\n",
    "clf = LogisticRegression(C=gs_log_reg.best_params_['C'],\n",
    "                         solver=gs_log_reg.best_params_['solver'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8479781420765027"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Cross-validated accuracy\n",
    "cv_acc = cross_val_score(clf,\n",
    "                         X,\n",
    "                         y,\n",
    "                         cv=5,\n",
    "                         scoring='accuracy')\n",
    "\n",
    "cv_acc = np.mean(cv_acc)\n",
    "cv_acc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8215873015873015"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Cross-validated precision\n",
    "cv_precision = cross_val_score(clf,\n",
    "                               X,\n",
    "                               y,\n",
    "                               cv=5,\n",
    "                               scoring='precision')\n",
    "\n",
    "cv_precision = np.mean(cv_precision)\n",
    "cv_precision"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9272727272727274"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Cross-validated recall\n",
    "cv_recall = cross_val_score(clf,\n",
    "                         X,\n",
    "                         y,\n",
    "                         cv=5,\n",
    "                         scoring='recall')\n",
    "\n",
    "cv_recall = np.mean(cv_recall)\n",
    "cv_recall"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8705403543192143"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Cross-validated f1-score\n",
    "cv_f1 = cross_val_score(clf,\n",
    "                        X,\n",
    "                        y,\n",
    "                        cv=5,\n",
    "                        scoring='f1')\n",
    "\n",
    "cv_f1 = np.mean(cv_f1)\n",
    "cv_f1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualize cross-validated metrics\n",
    "cv_metrics = pd.DataFrame({'Accuracy': cv_acc,\n",
    "                           'Precision': cv_precision,\n",
    "                           'Recall': cv_recall,\n",
    "                           'F1': cv_f1},\n",
    "                          index=[0])\n",
    "\n",
    "cv_metrics.T.plot.bar(title=\"Cross-validated classification metrics\",\n",
    "                      legend=False);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Feature Importance\n",
    "\n",
    "Which features contribuited the most to the outcomes of the model?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['age',\n",
       " 'sex',\n",
       " 'cp',\n",
       " 'trestbps',\n",
       " 'chol',\n",
       " 'fbs',\n",
       " 'restecg',\n",
       " 'thalach',\n",
       " 'exang',\n",
       " 'oldpeak',\n",
       " 'slope',\n",
       " 'ca',\n",
       " 'thal',\n",
       " 'target']"
      ]
     },
     "execution_count": 70,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "list(df.columns.values)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "LogisticRegression(C=0.23357214690901212, class_weight=None, dual=False,\n",
       "                   fit_intercept=True, intercept_scaling=1, l1_ratio=None,\n",
       "                   max_iter=100, multi_class='auto', n_jobs=None, penalty='l2',\n",
       "                   random_state=None, solver='liblinear', tol=0.0001, verbose=0,\n",
       "                   warm_start=False)"
      ]
     },
     "execution_count": 71,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Use the previously created instance of LogisticRegression\n",
    "clf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [],
   "source": [
    "clf.fit(X_train, y_train);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[-7.70903683e-04, -9.38565663e-01,  7.00403259e-01,\n",
       "        -1.82229628e-02, -1.87460903e-03,  4.26805408e-03,\n",
       "         3.65445349e-01,  3.22296265e-02, -5.31479435e-01,\n",
       "        -5.14049165e-01,  3.06868229e-01, -5.64216904e-01,\n",
       "        -6.18643929e-01]])"
      ]
     },
     "execution_count": 73,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf.coef_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'age': -0.0007709036827013516,\n",
       " 'sex': -0.9385656633514624,\n",
       " 'cp': 0.700403258829611,\n",
       " 'trestbps': -0.018222962768096058,\n",
       " 'chol': -0.0018746090302703532,\n",
       " 'fbs': 0.004268054076135423,\n",
       " 'restecg': 0.36544534872178913,\n",
       " 'thalach': 0.032229626484217413,\n",
       " 'exang': -0.531479435136635,\n",
       " 'oldpeak': -0.5140491650312795,\n",
       " 'slope': 0.3068682291856542,\n",
       " 'ca': -0.564216903563187,\n",
       " 'thal': -0.6186439286561868}"
      ]
     },
     "execution_count": 74,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Match features with its coeficients\n",
    "features_dict = dict(zip(df.columns.values, list(clf.coef_[0])))\n",
    "features_dict"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualize feature importance\n",
    "feature_df = pd.DataFrame(features_dict, index=[0])\n",
    "feature_df.T.plot.bar(title=\"Feature Importance\", legend=False);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There's a clear negative relation between sex and the target (ratio). And prositive for cp and target."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "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>target</th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sex</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>24</td>\n",
       "      <td>72</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>114</td>\n",
       "      <td>93</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "target    0   1\n",
       "sex            \n",
       "0        24  72\n",
       "1       114  93"
      ]
     },
     "execution_count": 76,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.crosstab(df['sex'], df['target'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "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>target</th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>cp</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>104</td>\n",
       "      <td>39</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>9</td>\n",
       "      <td>41</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>18</td>\n",
       "      <td>69</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>7</td>\n",
       "      <td>16</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "target    0   1\n",
       "cp             \n",
       "0       104  39\n",
       "1         9  41\n",
       "2        18  69\n",
       "3         7  16"
      ]
     },
     "execution_count": 77,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.crosstab(df['cp'], df['target'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "cp: chest pain type\n",
    "- 0: Typical angina (chest pain related decreased blood supply to heart).\n",
    "- 1: Atypical angina (chest pain not related to heart).\n",
    "- 2: Non anginal pain (typically esophageal spasms - non heart related).\n",
    "- 3: Asymptomatic (chest pain not showing signs of disease)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "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>target</th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>slope</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>12</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>91</td>\n",
       "      <td>49</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>35</td>\n",
       "      <td>107</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "target   0    1\n",
       "slope          \n",
       "0       12    9\n",
       "1       91   49\n",
       "2       35  107"
      ]
     },
     "execution_count": 78,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.crosstab(df['slope'], df['target'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "slope: the slope of the peak exercise ST segment\n",
    "- 0: Upsloping: better heart rate with excercise.\n",
    "- 1: Flatsloping: minimal change (typical healthy heart)\n",
    "- 2: Downslopins: signs of unhealthy heart"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 6. Conclusions\n",
    "\n",
    "In this proof of concept, we didn't hit the metric of 95% accuracy as required, but we did come close (~90%), so our model would predict correctly 9 out of 10 times –given the 13 parameters taken into account– if a patient has a heart disease. Some recomendations for further investigation would be:\n",
    "\n",
    "- Collecting more data to check if the models can be improved.\n",
    "- Try some other models like CatBoost or XGBoost.\n",
    "- Improve model parameters beyond what have been covered here."
   ]
  }
 ],
 "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": "3.8.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}