DAKD2022_ex2_Peppi-Lotta_Saari.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "fRx75YvD3BPW"
   },
   "source": [
    "## <font color = red> *** FILL HERE *** </font>\n",
    "Peppi-Lotta Saari <br>\n",
    "517334 <br>\n",
    "plsaar@utu.fi  <br>\n",
    "28.11.2022  <br>\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Rpio-nMAoDwq"
   },
   "source": [
    "------\n",
    "\n",
    "# Data Analysis and Knowledge Discovery: Exercise 2, Supervised learning"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "90c9fF6woDwr"
   },
   "source": [
    "The previous exercise was about <i>data understanding</i> and <i>data preparation</i>, which formed the basis for the modeling phase of the data mining process. Many modeling techniques make assumptions about data so the exploration and preparation phases can't be ignored. Now, as we have checked the validity of data and familiarized ourselves with it, we can move on to the next stage of the Cross-Industry Standard Process for Data Mining (CRISP-DM) which is <font color = darkorchid><i>modeling</i></font>.\n",
    "\n",
    "The questions to be answered at this stage could be for example:\n",
    "\n",
    "- What kind of model architecture best fits our data?\n",
    "- How well does the model perform technically?\n",
    "- Could we improve that performance?\n",
    "- How the performance of the model is evaluated?\n",
    "\n",
    "<i>Machine learning</i> is a subfield of artificial intelligence which provides automatic, objective and data-driven techniques for modeling the data. The machine learning algorithms aim to learn from data to make predictions. The two main branches of it are <i>supervised learning</i> and <i>unsupervised learning</i>.  In this exercise, we are going to use the former -- <font color = darkorchid><i>supervised learning</i></font> -- for classification and regression tasks.\n",
    "\n",
    "For classification, data remains the same as in the previous exercise, but I've already cleaned it up for you. Some data pre-processing steps are still required to ensure that it's in an appropriate format, so that models can learn something from it. Even though we are not doing any major data exploration nor data preparation here this time, <i>you should <b>never</b> forget it in your future data analyses</i>.\n",
    "\n",
    "-----\n",
    "\n",
    "<b>General guidance for exercises</b>\n",
    "\n",
    "- Answer <b>all</b> questions below, even if you can't get your script to fully work.\n",
    "- Write clear and easily readable code, and include explanations what your code does\n",
    "- Make informative illustrations: include labels for x and y axes, legends and captions for your plots.\n",
    "- You can add more code and markup cells, as long as the flow of the notebook stays readable and logical.\n",
    "- Before saving the ipynb file (and possible printing) run: \"Restart & Run all\", to make sure you return a file that works as expected.\n",
    "- Grading: *Fail*/*Pass*/*Pass with honors* (+1)\n",
    "- +1 bonus point (grading *Pass with honors*) requires a <b>completely</b> correct solution and also thorough analysis.\n",
    "- If you encounter problems, Google first. If you can't find an answer to the problem, don't hesitate to ask in the Moodle discussion or directly via moodle chat or email from the course assistants.\n",
    "- Note! Don't leave it to the last moment! No feedback service during weekends.\n",
    "\n",
    "<font color = red size = 4><b>The deadline is 28 November at 18:00</b></font>. Late submissions will not be accepted unless there is a valid excuse for extending the deadline before the due date.\n",
    "\n",
    "------"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "2I2WLapM3BPc"
   },
   "source": [
    "### Gather **all** packages needed for this notebook here:\n",
    "\n",
    "You can use other packages as well, but this excercise can be completed with those below."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "id": "iypIAVquoDws"
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import scipy.stats as stats\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# Machine learning package - scikit-learn\n",
    "from sklearn import metrics\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.model_selection import train_test_split, cross_val_score, LeaveOneOut, GridSearchCV\n",
    "from sklearn.neighbors import KNeighborsClassifier, KNeighborsRegressor\n",
    "from sklearn.linear_model import Ridge\n",
    "\n",
    "# Show the plots inline in the notebook\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "______________\n",
    "## <font color = lightcoral>1. Classification using k-nearest neighbors </font> "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "7CqWZYx2oDw3"
   },
   "source": [
    "We start exploring the world of data modeling by using <font color = darkorchid><b>K-Nearest Neightbors (k-NN) algorithm</b></font>. The k-NN algorithm is one of the classic supervised machine learning algorithms. It assumes that similar points are close to each other. \n",
    "\n",
    "In our case, we'll use the k-NN algorithm to **predict the presence of cardiovascular disease** using all the other variables as <font color = darkorchid><b>features</b></font> in the given data set. So, the target variable is <font color = darkorchid><b>cardio</b></font>.\n",
    "\n",
    "But first, we need data for the task. The code for loading the data into the environment is provided for you. <font color = red>The code should work, but make sure, that you have the csv file of the data in the same directory where you have this notebook file.</font> \n",
    "\n",
    "***Exercise 1 A)*** \n",
    "\n",
    "Print the first 10 rows to check that everything is ok with the created dataframe.\n",
    "\n",
    "*note: as said, the data remains the same, but cholesterol has been one-hot-encoded for you already. There's also a new variable gluc (about glucose levels), which is one-hot-encoded for you. It has the similar values as cholesterol originally does [normal, at risk, elevated]. Also, binary variables have been changed to [0,1] values.*\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "### Loading code provided\n",
    "# ------------------------------------------------------\n",
    "# The data file should be at the same location than the \n",
    "# exercise file to make sure the following lines work!\n",
    "# Otherwise, fix the path.\n",
    "# ------------------------------------------------------\n",
    "\n",
    "# Path for the data\n",
    "data_path = 'ex2_cardio_data.csv'\n",
    "\n",
    "# Create a dataframe\n",
    "cardio_data = pd.read_csv(data_path, dtype='int')"
   ]
  },
  {
   "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>height</th>\n",
       "      <th>weight</th>\n",
       "      <th>ap_hi</th>\n",
       "      <th>ap_lo</th>\n",
       "      <th>smoke</th>\n",
       "      <th>alco</th>\n",
       "      <th>active</th>\n",
       "      <th>cardio</th>\n",
       "      <th>cholesterol_at_risk</th>\n",
       "      <th>cholesterol_elevated</th>\n",
       "      <th>cholesterol_normal</th>\n",
       "      <th>gluc_at_risk</th>\n",
       "      <th>gluc_elevated</th>\n",
       "      <th>gluc_normal</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>22767</td>\n",
       "      <td>0</td>\n",
       "      <td>162</td>\n",
       "      <td>75</td>\n",
       "      <td>120</td>\n",
       "      <td>80</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>19102</td>\n",
       "      <td>0</td>\n",
       "      <td>164</td>\n",
       "      <td>80</td>\n",
       "      <td>140</td>\n",
       "      <td>90</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>23251</td>\n",
       "      <td>0</td>\n",
       "      <td>155</td>\n",
       "      <td>76</td>\n",
       "      <td>130</td>\n",
       "      <td>80</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>22651</td>\n",
       "      <td>1</td>\n",
       "      <td>169</td>\n",
       "      <td>65</td>\n",
       "      <td>120</td>\n",
       "      <td>80</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>21770</td>\n",
       "      <td>1</td>\n",
       "      <td>171</td>\n",
       "      <td>67</td>\n",
       "      <td>160</td>\n",
       "      <td>100</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>14651</td>\n",
       "      <td>1</td>\n",
       "      <td>170</td>\n",
       "      <td>70</td>\n",
       "      <td>120</td>\n",
       "      <td>80</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>17391</td>\n",
       "      <td>0</td>\n",
       "      <td>152</td>\n",
       "      <td>65</td>\n",
       "      <td>110</td>\n",
       "      <td>70</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>14792</td>\n",
       "      <td>0</td>\n",
       "      <td>165</td>\n",
       "      <td>65</td>\n",
       "      <td>120</td>\n",
       "      <td>70</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>15982</td>\n",
       "      <td>1</td>\n",
       "      <td>180</td>\n",
       "      <td>81</td>\n",
       "      <td>110</td>\n",
       "      <td>70</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>19589</td>\n",
       "      <td>0</td>\n",
       "      <td>156</td>\n",
       "      <td>60</td>\n",
       "      <td>120</td>\n",
       "      <td>90</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     age  sex  height  weight  ap_hi  ap_lo  smoke  alco  active  cardio  \\\n",
       "0  22767    0     162      75    120     80      0     0       1       0   \n",
       "1  19102    0     164      80    140     90      0     0       1       0   \n",
       "2  23251    0     155      76    130     80      0     0       1       0   \n",
       "3  22651    1     169      65    120     80      0     0       1       0   \n",
       "4  21770    1     171      67    160    100      0     0       0       1   \n",
       "5  14651    1     170      70    120     80      0     0       1       0   \n",
       "6  17391    0     152      65    110     70      0     0       1       0   \n",
       "7  14792    0     165      65    120     70      0     0       1       0   \n",
       "8  15982    1     180      81    110     70      1     0       1       0   \n",
       "9  19589    0     156      60    120     90      0     0       1       0   \n",
       "\n",
       "   cholesterol_at_risk  cholesterol_elevated  cholesterol_normal  \\\n",
       "0                    0                     0                   1   \n",
       "1                    0                     0                   1   \n",
       "2                    0                     0                   1   \n",
       "3                    0                     0                   1   \n",
       "4                    0                     0                   1   \n",
       "5                    0                     0                   1   \n",
       "6                    0                     0                   1   \n",
       "7                    0                     0                   1   \n",
       "8                    0                     0                   1   \n",
       "9                    0                     0                   1   \n",
       "\n",
       "   gluc_at_risk  gluc_elevated  gluc_normal  \n",
       "0             0              0            1  \n",
       "1             0              0            1  \n",
       "2             0              1            0  \n",
       "3             0              0            1  \n",
       "4             0              0            1  \n",
       "5             0              0            1  \n",
       "6             0              0            1  \n",
       "7             0              0            1  \n",
       "8             0              0            1  \n",
       "9             0              0            1  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "### Code - 10 first row\n",
    "cardio_data.head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "age                     int32\n",
       "sex                     int32\n",
       "height                  int32\n",
       "weight                  int32\n",
       "ap_hi                   int32\n",
       "ap_lo                   int32\n",
       "smoke                   int32\n",
       "alco                    int32\n",
       "active                  int32\n",
       "cardio                  int32\n",
       "cholesterol_at_risk     int32\n",
       "cholesterol_elevated    int32\n",
       "cholesterol_normal      int32\n",
       "gluc_at_risk            int32\n",
       "gluc_elevated           int32\n",
       "gluc_normal             int32\n",
       "dtype: object"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cardio_data.dtypes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "age                       int32\n",
       "sex                     boolean\n",
       "height                    int32\n",
       "weight                    int32\n",
       "ap_hi                     int32\n",
       "ap_lo                     int32\n",
       "smoke                   boolean\n",
       "alco                    boolean\n",
       "active                  boolean\n",
       "cardio                  boolean\n",
       "cholesterol_at_risk     boolean\n",
       "cholesterol_elevated    boolean\n",
       "cholesterol_normal      boolean\n",
       "gluc_at_risk            boolean\n",
       "gluc_elevated           boolean\n",
       "gluc_normal             boolean\n",
       "dtype: object"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# change boolean columns to type boolean \n",
    "boolean = ['sex', 'smoke', 'alco', 'active','cardio', 'cholesterol_at_risk', 'cholesterol_elevated', 'cholesterol_normal', 'gluc_at_risk', 'gluc_elevated', 'gluc_normal']\n",
    "cardio_data[boolean] = cardio_data[boolean].astype('boolean')\n",
    "cardio_data.dtypes"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "mfq_3_WNoDw3"
   },
   "source": [
    "----\n",
    "\n",
    "We have the data so let's make use of it. \n",
    "\n",
    "To teach the k-NN algorithm (or any kind of machine learning algorithm) to recognize patterns, we are going to need <font color = darkorchid><b>training data</b></font>. But to test how well a model possibly learned those patterns, we'll need <font color = darkorchid><b>test data</b></font> which is new and unseen for the trained model. Keep in mind that the test set is not shown for the model before we are actually done with training. \n",
    "\n",
    "So, to **find the estimate for the performance of a model**, we are going to use this <font color = darkorchid><b>train-test split</b></font>. \"Split\" because we literally split the data into two sets.\n",
    "\n",
    "Sometimes <font color = darkorchid>stratification</font> needs to be considered. It can be used to ensure that train and test sets contain the same proportions of samples of each target class as the original data set.\n",
    "\n",
    "***Exercise 1 B)*** \n",
    "\n",
    "Gather the features as an array `features`, and the target variable as an array `labels`. Produce training and test data.  Divide the data **randomly** into training (80%) and test (20%) sets.\n",
    "\n",
    "- Would it be a good idea to use stratification? **Explain** your decision."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "### Code - Train-test split\n",
    "features = ['age','sex', 'height', 'weight', 'ap_hi', 'ap_lo', \n",
    "            'smoke', 'alco', 'active', 'cholesterol_at_risk', 'cholesterol_elevated', \n",
    "            'cholesterol_normal', 'gluc_at_risk', 'gluc_elevated', 'gluc_normal'\n",
    "           ]\n",
    "labels = ['cardio']\n",
    "cardio_features = cardio_data[features]\n",
    "cardio_labels = cardio_data[labels]\n",
    "\n",
    "# https://www.geeksforgeeks.org/how-to-do-train-test-split-using-sklearn-in-python/\n",
    "feature_train, feature_test, label_train, label_test = train_test_split(cardio_features, cardio_labels ,\n",
    "                                   random_state=104, \n",
    "                                   train_size=0.8, \n",
    "                                   shuffle=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>age</th>\n",
       "      <th>sex</th>\n",
       "      <th>height</th>\n",
       "      <th>weight</th>\n",
       "      <th>ap_hi</th>\n",
       "      <th>ap_lo</th>\n",
       "      <th>smoke</th>\n",
       "      <th>alco</th>\n",
       "      <th>active</th>\n",
       "      <th>cholesterol_at_risk</th>\n",
       "      <th>cholesterol_elevated</th>\n",
       "      <th>cholesterol_normal</th>\n",
       "      <th>gluc_at_risk</th>\n",
       "      <th>gluc_elevated</th>\n",
       "      <th>gluc_normal</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>109</th>\n",
       "      <td>22488</td>\n",
       "      <td>False</td>\n",
       "      <td>146</td>\n",
       "      <td>77</td>\n",
       "      <td>140</td>\n",
       "      <td>70</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>770</th>\n",
       "      <td>18195</td>\n",
       "      <td>False</td>\n",
       "      <td>150</td>\n",
       "      <td>72</td>\n",
       "      <td>120</td>\n",
       "      <td>90</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>797</th>\n",
       "      <td>23341</td>\n",
       "      <td>False</td>\n",
       "      <td>170</td>\n",
       "      <td>60</td>\n",
       "      <td>100</td>\n",
       "      <td>80</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>554</th>\n",
       "      <td>16587</td>\n",
       "      <td>False</td>\n",
       "      <td>163</td>\n",
       "      <td>63</td>\n",
       "      <td>120</td>\n",
       "      <td>90</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>951</th>\n",
       "      <td>21086</td>\n",
       "      <td>True</td>\n",
       "      <td>181</td>\n",
       "      <td>88</td>\n",
       "      <td>140</td>\n",
       "      <td>90</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       age    sex  height  weight  ap_hi  ap_lo  smoke   alco  active  \\\n",
       "109  22488  False     146      77    140     70  False  False    True   \n",
       "770  18195  False     150      72    120     90  False  False    True   \n",
       "797  23341  False     170      60    100     80  False  False   False   \n",
       "554  16587  False     163      63    120     90  False  False    True   \n",
       "951  21086   True     181      88    140     90   True   True    True   \n",
       "\n",
       "     cholesterol_at_risk  cholesterol_elevated  cholesterol_normal  \\\n",
       "109                False                 False                True   \n",
       "770                 True                 False               False   \n",
       "797                False                 False                True   \n",
       "554                False                 False                True   \n",
       "951                False                  True               False   \n",
       "\n",
       "     gluc_at_risk  gluc_elevated  gluc_normal  \n",
       "109         False          False         True  \n",
       "770         False          False         True  \n",
       "797         False          False         True  \n",
       "554         False          False         True  \n",
       "951          True          False        False  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>height</th>\n",
       "      <th>weight</th>\n",
       "      <th>ap_hi</th>\n",
       "      <th>ap_lo</th>\n",
       "      <th>smoke</th>\n",
       "      <th>alco</th>\n",
       "      <th>active</th>\n",
       "      <th>cholesterol_at_risk</th>\n",
       "      <th>cholesterol_elevated</th>\n",
       "      <th>cholesterol_normal</th>\n",
       "      <th>gluc_at_risk</th>\n",
       "      <th>gluc_elevated</th>\n",
       "      <th>gluc_normal</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>808</th>\n",
       "      <td>22453</td>\n",
       "      <td>False</td>\n",
       "      <td>160</td>\n",
       "      <td>111</td>\n",
       "      <td>130</td>\n",
       "      <td>80</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>83</th>\n",
       "      <td>21261</td>\n",
       "      <td>True</td>\n",
       "      <td>175</td>\n",
       "      <td>75</td>\n",
       "      <td>120</td>\n",
       "      <td>80</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>643</th>\n",
       "      <td>23593</td>\n",
       "      <td>False</td>\n",
       "      <td>167</td>\n",
       "      <td>85</td>\n",
       "      <td>120</td>\n",
       "      <td>80</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>895</th>\n",
       "      <td>22464</td>\n",
       "      <td>False</td>\n",
       "      <td>169</td>\n",
       "      <td>65</td>\n",
       "      <td>120</td>\n",
       "      <td>80</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>536</th>\n",
       "      <td>14716</td>\n",
       "      <td>True</td>\n",
       "      <td>170</td>\n",
       "      <td>73</td>\n",
       "      <td>130</td>\n",
       "      <td>90</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       age    sex  height  weight  ap_hi  ap_lo  smoke   alco  active  \\\n",
       "808  22453  False     160     111    130     80  False  False   False   \n",
       "83   21261   True     175      75    120     80  False  False    True   \n",
       "643  23593  False     167      85    120     80  False  False   False   \n",
       "895  22464  False     169      65    120     80  False  False    True   \n",
       "536  14716   True     170      73    130     90  False  False   False   \n",
       "\n",
       "     cholesterol_at_risk  cholesterol_elevated  cholesterol_normal  \\\n",
       "808                False                 False                True   \n",
       "83                 False                 False                True   \n",
       "643                False                 False                True   \n",
       "895                False                 False                True   \n",
       "536                False                 False                True   \n",
       "\n",
       "     gluc_at_risk  gluc_elevated  gluc_normal  \n",
       "808         False          False         True  \n",
       "83          False          False         True  \n",
       "643         False          False         True  \n",
       "895         False          False         True  \n",
       "536          True          False        False  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>cardio</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>109</th>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>770</th>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>797</th>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>554</th>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>951</th>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     cardio\n",
       "109   False\n",
       "770   False\n",
       "797   False\n",
       "554    True\n",
       "951    True"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>cardio</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>808</th>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>83</th>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>643</th>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>895</th>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>536</th>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     cardio\n",
       "808   False\n",
       "83    False\n",
       "643    True\n",
       "895    True\n",
       "536   False"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#display heads\n",
    "display(feature_train.head())\n",
    "display(feature_test.head())\n",
    "display(label_train.head())\n",
    "display(label_test.head())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>index</th>\n",
       "      <th>age</th>\n",
       "      <th>sex</th>\n",
       "      <th>height</th>\n",
       "      <th>weight</th>\n",
       "      <th>ap_hi</th>\n",
       "      <th>ap_lo</th>\n",
       "      <th>smoke</th>\n",
       "      <th>alco</th>\n",
       "      <th>active</th>\n",
       "      <th>cholesterol_at_risk</th>\n",
       "      <th>cholesterol_elevated</th>\n",
       "      <th>cholesterol_normal</th>\n",
       "      <th>gluc_at_risk</th>\n",
       "      <th>gluc_elevated</th>\n",
       "      <th>gluc_normal</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>109</td>\n",
       "      <td>22488</td>\n",
       "      <td>False</td>\n",
       "      <td>146</td>\n",
       "      <td>77</td>\n",
       "      <td>140</td>\n",
       "      <td>70</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>770</td>\n",
       "      <td>18195</td>\n",
       "      <td>False</td>\n",
       "      <td>150</td>\n",
       "      <td>72</td>\n",
       "      <td>120</td>\n",
       "      <td>90</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>797</td>\n",
       "      <td>23341</td>\n",
       "      <td>False</td>\n",
       "      <td>170</td>\n",
       "      <td>60</td>\n",
       "      <td>100</td>\n",
       "      <td>80</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>554</td>\n",
       "      <td>16587</td>\n",
       "      <td>False</td>\n",
       "      <td>163</td>\n",
       "      <td>63</td>\n",
       "      <td>120</td>\n",
       "      <td>90</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>951</td>\n",
       "      <td>21086</td>\n",
       "      <td>True</td>\n",
       "      <td>181</td>\n",
       "      <td>88</td>\n",
       "      <td>140</td>\n",
       "      <td>90</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   index    age    sex  height  weight  ap_hi  ap_lo  smoke   alco  active  \\\n",
       "0    109  22488  False     146      77    140     70  False  False    True   \n",
       "1    770  18195  False     150      72    120     90  False  False    True   \n",
       "2    797  23341  False     170      60    100     80  False  False   False   \n",
       "3    554  16587  False     163      63    120     90  False  False    True   \n",
       "4    951  21086   True     181      88    140     90   True   True    True   \n",
       "\n",
       "   cholesterol_at_risk  cholesterol_elevated  cholesterol_normal  \\\n",
       "0                False                 False                True   \n",
       "1                 True                 False               False   \n",
       "2                False                 False                True   \n",
       "3                False                 False                True   \n",
       "4                False                  True               False   \n",
       "\n",
       "   gluc_at_risk  gluc_elevated  gluc_normal  \n",
       "0         False          False         True  \n",
       "1         False          False         True  \n",
       "2         False          False         True  \n",
       "3         False          False         True  \n",
       "4          True          False        False  "
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#reset index\n",
    "feature_train = feature_train.reset_index()\n",
    "feature_test = feature_test.reset_index()\n",
    "label_train = label_train.reset_index()\n",
    "label_test = label_test.reset_index()\n",
    "\n",
    "feature_train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>age</th>\n",
       "      <th>sex</th>\n",
       "      <th>height</th>\n",
       "      <th>weight</th>\n",
       "      <th>ap_hi</th>\n",
       "      <th>ap_lo</th>\n",
       "      <th>smoke</th>\n",
       "      <th>alco</th>\n",
       "      <th>active</th>\n",
       "      <th>cholesterol_at_risk</th>\n",
       "      <th>cholesterol_elevated</th>\n",
       "      <th>cholesterol_normal</th>\n",
       "      <th>gluc_at_risk</th>\n",
       "      <th>gluc_elevated</th>\n",
       "      <th>gluc_normal</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>22488</td>\n",
       "      <td>False</td>\n",
       "      <td>146</td>\n",
       "      <td>77</td>\n",
       "      <td>140</td>\n",
       "      <td>70</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>18195</td>\n",
       "      <td>False</td>\n",
       "      <td>150</td>\n",
       "      <td>72</td>\n",
       "      <td>120</td>\n",
       "      <td>90</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>23341</td>\n",
       "      <td>False</td>\n",
       "      <td>170</td>\n",
       "      <td>60</td>\n",
       "      <td>100</td>\n",
       "      <td>80</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>16587</td>\n",
       "      <td>False</td>\n",
       "      <td>163</td>\n",
       "      <td>63</td>\n",
       "      <td>120</td>\n",
       "      <td>90</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>21086</td>\n",
       "      <td>True</td>\n",
       "      <td>181</td>\n",
       "      <td>88</td>\n",
       "      <td>140</td>\n",
       "      <td>90</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     age    sex  height  weight  ap_hi  ap_lo  smoke   alco  active  \\\n",
       "0  22488  False     146      77    140     70  False  False    True   \n",
       "1  18195  False     150      72    120     90  False  False    True   \n",
       "2  23341  False     170      60    100     80  False  False   False   \n",
       "3  16587  False     163      63    120     90  False  False    True   \n",
       "4  21086   True     181      88    140     90   True   True    True   \n",
       "\n",
       "   cholesterol_at_risk  cholesterol_elevated  cholesterol_normal  \\\n",
       "0                False                 False                True   \n",
       "1                 True                 False               False   \n",
       "2                False                 False                True   \n",
       "3                False                 False                True   \n",
       "4                False                  True               False   \n",
       "\n",
       "   gluc_at_risk  gluc_elevated  gluc_normal  \n",
       "0         False          False         True  \n",
       "1         False          False         True  \n",
       "2         False          False         True  \n",
       "3         False          False         True  \n",
       "4          True          False        False  "
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#drop level_0 and index\n",
    "feature_train = feature_train.drop(['index'], axis=1)\n",
    "feature_test = feature_test.drop(['index'], axis=1)\n",
    "label_train = label_train.drop(['index'], axis=1)\n",
    "label_test = label_test.drop(['index'], axis=1)\n",
    "\n",
    "feature_train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "</style>\n",
       "<table id=\"T_a6514\">\n",
       "  <caption>sex</caption>\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_a6514_level0_col0\" class=\"col_heading level0 col0\" >Feature_frequency_train</th>\n",
       "      <th id=\"T_a6514_level0_col1\" class=\"col_heading level0 col1\" >Feature_frequency_all</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_a6514_level0_row0\" class=\"row_heading level0 row0\" >False</th>\n",
       "      <td id=\"T_a6514_row0_col0\" class=\"data row0 col0\" >0.645000</td>\n",
       "      <td id=\"T_a6514_row0_col1\" class=\"data row0 col1\" >0.645000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_a6514_level0_row1\" class=\"row_heading level0 row1\" >True</th>\n",
       "      <td id=\"T_a6514_row1_col0\" class=\"data row1 col0\" >0.355000</td>\n",
       "      <td id=\"T_a6514_row1_col1\" class=\"data row1 col1\" >0.355000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x25bdd12add0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "</style>\n",
       "<table id=\"T_007d2\">\n",
       "  <caption>smoke</caption>\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_007d2_level0_col0\" class=\"col_heading level0 col0\" >Feature_frequency_train</th>\n",
       "      <th id=\"T_007d2_level0_col1\" class=\"col_heading level0 col1\" >Feature_frequency_all</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_007d2_level0_row0\" class=\"row_heading level0 row0\" >False</th>\n",
       "      <td id=\"T_007d2_row0_col0\" class=\"data row0 col0\" >0.893750</td>\n",
       "      <td id=\"T_007d2_row0_col1\" class=\"data row0 col1\" >0.902000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_007d2_level0_row1\" class=\"row_heading level0 row1\" >True</th>\n",
       "      <td id=\"T_007d2_row1_col0\" class=\"data row1 col0\" >0.106250</td>\n",
       "      <td id=\"T_007d2_row1_col1\" class=\"data row1 col1\" >0.098000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x25bdd12add0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "</style>\n",
       "<table id=\"T_90cdc\">\n",
       "  <caption>alco</caption>\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_90cdc_level0_col0\" class=\"col_heading level0 col0\" >Feature_frequency_train</th>\n",
       "      <th id=\"T_90cdc_level0_col1\" class=\"col_heading level0 col1\" >Feature_frequency_all</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_90cdc_level0_row0\" class=\"row_heading level0 row0\" >False</th>\n",
       "      <td id=\"T_90cdc_row0_col0\" class=\"data row0 col0\" >0.937500</td>\n",
       "      <td id=\"T_90cdc_row0_col1\" class=\"data row0 col1\" >0.940000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90cdc_level0_row1\" class=\"row_heading level0 row1\" >True</th>\n",
       "      <td id=\"T_90cdc_row1_col0\" class=\"data row1 col0\" >0.062500</td>\n",
       "      <td id=\"T_90cdc_row1_col1\" class=\"data row1 col1\" >0.060000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x25bdd12add0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "</style>\n",
       "<table id=\"T_e4e6b\">\n",
       "  <caption>active</caption>\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_e4e6b_level0_col0\" class=\"col_heading level0 col0\" >Feature_frequency_train</th>\n",
       "      <th id=\"T_e4e6b_level0_col1\" class=\"col_heading level0 col1\" >Feature_frequency_all</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_e4e6b_level0_row0\" class=\"row_heading level0 row0\" >True</th>\n",
       "      <td id=\"T_e4e6b_row0_col0\" class=\"data row0 col0\" >0.808750</td>\n",
       "      <td id=\"T_e4e6b_row0_col1\" class=\"data row0 col1\" >0.816000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e4e6b_level0_row1\" class=\"row_heading level0 row1\" >False</th>\n",
       "      <td id=\"T_e4e6b_row1_col0\" class=\"data row1 col0\" >0.191250</td>\n",
       "      <td id=\"T_e4e6b_row1_col1\" class=\"data row1 col1\" >0.184000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x25bdd12add0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "</style>\n",
       "<table id=\"T_ab43e\">\n",
       "  <caption>cholesterol_at_risk</caption>\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_ab43e_level0_col0\" class=\"col_heading level0 col0\" >Feature_frequency_train</th>\n",
       "      <th id=\"T_ab43e_level0_col1\" class=\"col_heading level0 col1\" >Feature_frequency_all</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_ab43e_level0_row0\" class=\"row_heading level0 row0\" >False</th>\n",
       "      <td id=\"T_ab43e_row0_col0\" class=\"data row0 col0\" >0.877500</td>\n",
       "      <td id=\"T_ab43e_row0_col1\" class=\"data row0 col1\" >0.883000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ab43e_level0_row1\" class=\"row_heading level0 row1\" >True</th>\n",
       "      <td id=\"T_ab43e_row1_col0\" class=\"data row1 col0\" >0.122500</td>\n",
       "      <td id=\"T_ab43e_row1_col1\" class=\"data row1 col1\" >0.117000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x25bdd12add0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "</style>\n",
       "<table id=\"T_1bd8c\">\n",
       "  <caption>cholesterol_elevated</caption>\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_1bd8c_level0_col0\" class=\"col_heading level0 col0\" >Feature_frequency_train</th>\n",
       "      <th id=\"T_1bd8c_level0_col1\" class=\"col_heading level0 col1\" >Feature_frequency_all</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_1bd8c_level0_row0\" class=\"row_heading level0 row0\" >False</th>\n",
       "      <td id=\"T_1bd8c_row0_col0\" class=\"data row0 col0\" >0.915000</td>\n",
       "      <td id=\"T_1bd8c_row0_col1\" class=\"data row0 col1\" >0.912000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_1bd8c_level0_row1\" class=\"row_heading level0 row1\" >True</th>\n",
       "      <td id=\"T_1bd8c_row1_col0\" class=\"data row1 col0\" >0.085000</td>\n",
       "      <td id=\"T_1bd8c_row1_col1\" class=\"data row1 col1\" >0.088000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x25bdd12add0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "</style>\n",
       "<table id=\"T_50fad\">\n",
       "  <caption>cholesterol_normal</caption>\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_50fad_level0_col0\" class=\"col_heading level0 col0\" >Feature_frequency_train</th>\n",
       "      <th id=\"T_50fad_level0_col1\" class=\"col_heading level0 col1\" >Feature_frequency_all</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_50fad_level0_row0\" class=\"row_heading level0 row0\" >True</th>\n",
       "      <td id=\"T_50fad_row0_col0\" class=\"data row0 col0\" >0.792500</td>\n",
       "      <td id=\"T_50fad_row0_col1\" class=\"data row0 col1\" >0.795000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_50fad_level0_row1\" class=\"row_heading level0 row1\" >False</th>\n",
       "      <td id=\"T_50fad_row1_col0\" class=\"data row1 col0\" >0.207500</td>\n",
       "      <td id=\"T_50fad_row1_col1\" class=\"data row1 col1\" >0.205000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x25bdd12add0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "</style>\n",
       "<table id=\"T_0cd65\">\n",
       "  <caption>gluc_at_risk</caption>\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_0cd65_level0_col0\" class=\"col_heading level0 col0\" >Feature_frequency_train</th>\n",
       "      <th id=\"T_0cd65_level0_col1\" class=\"col_heading level0 col1\" >Feature_frequency_all</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_0cd65_level0_row0\" class=\"row_heading level0 row0\" >False</th>\n",
       "      <td id=\"T_0cd65_row0_col0\" class=\"data row0 col0\" >0.943750</td>\n",
       "      <td id=\"T_0cd65_row0_col1\" class=\"data row0 col1\" >0.935000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_0cd65_level0_row1\" class=\"row_heading level0 row1\" >True</th>\n",
       "      <td id=\"T_0cd65_row1_col0\" class=\"data row1 col0\" >0.056250</td>\n",
       "      <td id=\"T_0cd65_row1_col1\" class=\"data row1 col1\" >0.065000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x25bdd12add0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "</style>\n",
       "<table id=\"T_e9a01\">\n",
       "  <caption>gluc_elevated</caption>\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_e9a01_level0_col0\" class=\"col_heading level0 col0\" >Feature_frequency_train</th>\n",
       "      <th id=\"T_e9a01_level0_col1\" class=\"col_heading level0 col1\" >Feature_frequency_all</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_e9a01_level0_row0\" class=\"row_heading level0 row0\" >False</th>\n",
       "      <td id=\"T_e9a01_row0_col0\" class=\"data row0 col0\" >0.940000</td>\n",
       "      <td id=\"T_e9a01_row0_col1\" class=\"data row0 col1\" >0.942000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e9a01_level0_row1\" class=\"row_heading level0 row1\" >True</th>\n",
       "      <td id=\"T_e9a01_row1_col0\" class=\"data row1 col0\" >0.060000</td>\n",
       "      <td id=\"T_e9a01_row1_col1\" class=\"data row1 col1\" >0.058000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x25bdd12add0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "</style>\n",
       "<table id=\"T_8496f\">\n",
       "  <caption>gluc_normal</caption>\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_8496f_level0_col0\" class=\"col_heading level0 col0\" >Feature_frequency_train</th>\n",
       "      <th id=\"T_8496f_level0_col1\" class=\"col_heading level0 col1\" >Feature_frequency_all</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_8496f_level0_row0\" class=\"row_heading level0 row0\" >True</th>\n",
       "      <td id=\"T_8496f_row0_col0\" class=\"data row0 col0\" >0.883750</td>\n",
       "      <td id=\"T_8496f_row0_col1\" class=\"data row0 col1\" >0.877000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_8496f_level0_row1\" class=\"row_heading level0 row1\" >False</th>\n",
       "      <td id=\"T_8496f_row1_col0\" class=\"data row1 col0\" >0.116250</td>\n",
       "      <td id=\"T_8496f_row1_col1\" class=\"data row1 col1\" >0.123000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x25bdd12add0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# checking if stratification is needed \n",
    "for b in feature_train.select_dtypes(exclude='number'):\n",
    "        relative = feature_train[b].value_counts(normalize=True).to_frame()\n",
    "        rel = cardio_data[b].value_counts(normalize=True).to_frame()\n",
    "        result = pd.concat([relative, rel], axis=1)\n",
    "        result.set_axis(['Feature_frequency_train', 'Feature_frequency_all'], axis=1, inplace=True)\n",
    "        result = result.style.set_caption(b)\n",
    "        display(result)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "daPc_o9boDw3"
   },
   "source": [
    "<a href=\"https://towardsdatascience.com/what-is-stratified-cross-validation-in-machine-learning-8844f3e7ae8e\">https://towardsdatascience.com/what-is-stratified-cross-validation-in-machine-learning-8844f3e7ae8e</a><br>\n",
    "<font color = skyblue> Stratification is not neccisary in my oppinion. Based on the above relative frequency check that original datas and training data does not have any signifficant diffenrence in any columns frequencies. This means that the training data is a good repsentation of the original. This is expectable as the sampling is done in random.  </font>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "----------\n",
    "\n",
    "***Exercise 1 C)*** \n",
    "\n",
    "Scale the numeric features. \n",
    "\n",
    "- Remind yourself briefly why you need this step. \n",
    "- What can you say about scaling especially from the perspective of the k-NN algorithm?\n",
    "\n",
    "*tip: You should now have **two** variables where you have the features divided into. Don't forget to scale the test data. Some good information about preprocessing and how to use it for train and test data can be found https://scikit-learn.org/stable/modules/preprocessing.html.*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "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>height</th>\n",
       "      <th>weight</th>\n",
       "      <th>ap_hi</th>\n",
       "      <th>ap_lo</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>1000.000000</td>\n",
       "      <td>1000.000000</td>\n",
       "      <td>1000.000000</td>\n",
       "      <td>1000.000000</td>\n",
       "      <td>1000.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>19138.613000</td>\n",
       "      <td>164.359000</td>\n",
       "      <td>73.114000</td>\n",
       "      <td>124.147000</td>\n",
       "      <td>80.400000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>2552.145587</td>\n",
       "      <td>7.746362</td>\n",
       "      <td>14.285625</td>\n",
       "      <td>15.332269</td>\n",
       "      <td>9.124761</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>14308.000000</td>\n",
       "      <td>130.000000</td>\n",
       "      <td>41.000000</td>\n",
       "      <td>80.000000</td>\n",
       "      <td>60.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>17263.000000</td>\n",
       "      <td>159.000000</td>\n",
       "      <td>64.000000</td>\n",
       "      <td>120.000000</td>\n",
       "      <td>80.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>19152.500000</td>\n",
       "      <td>165.000000</td>\n",
       "      <td>70.000000</td>\n",
       "      <td>120.000000</td>\n",
       "      <td>80.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>21178.500000</td>\n",
       "      <td>169.250000</td>\n",
       "      <td>80.000000</td>\n",
       "      <td>130.000000</td>\n",
       "      <td>80.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>23649.000000</td>\n",
       "      <td>195.000000</td>\n",
       "      <td>170.000000</td>\n",
       "      <td>200.000000</td>\n",
       "      <td>120.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                age       height       weight        ap_hi        ap_lo\n",
       "count   1000.000000  1000.000000  1000.000000  1000.000000  1000.000000\n",
       "mean   19138.613000   164.359000    73.114000   124.147000    80.400000\n",
       "std     2552.145587     7.746362    14.285625    15.332269     9.124761\n",
       "min    14308.000000   130.000000    41.000000    80.000000    60.000000\n",
       "25%    17263.000000   159.000000    64.000000   120.000000    80.000000\n",
       "50%    19152.500000   165.000000    70.000000   120.000000    80.000000\n",
       "75%    21178.500000   169.250000    80.000000   130.000000    80.000000\n",
       "max    23649.000000   195.000000   170.000000   200.000000   120.000000"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cardio_data.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "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>height</th>\n",
       "      <th>weight</th>\n",
       "      <th>ap_hi</th>\n",
       "      <th>ap_lo</th>\n",
       "      <th>smoke</th>\n",
       "      <th>alco</th>\n",
       "      <th>active</th>\n",
       "      <th>cholesterol_at_risk</th>\n",
       "      <th>cholesterol_elevated</th>\n",
       "      <th>cholesterol_normal</th>\n",
       "      <th>gluc_at_risk</th>\n",
       "      <th>gluc_elevated</th>\n",
       "      <th>gluc_normal</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.336831</td>\n",
       "      <td>False</td>\n",
       "      <td>-2.342353</td>\n",
       "      <td>0.275366</td>\n",
       "      <td>1.054367</td>\n",
       "      <td>-1.110335</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-0.357123</td>\n",
       "      <td>False</td>\n",
       "      <td>-1.834422</td>\n",
       "      <td>-0.074194</td>\n",
       "      <td>-0.255406</td>\n",
       "      <td>1.065459</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.673412</td>\n",
       "      <td>False</td>\n",
       "      <td>0.705230</td>\n",
       "      <td>-0.913140</td>\n",
       "      <td>-1.565179</td>\n",
       "      <td>-0.022438</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.991616</td>\n",
       "      <td>False</td>\n",
       "      <td>-0.183649</td>\n",
       "      <td>-0.703403</td>\n",
       "      <td>-0.255406</td>\n",
       "      <td>1.065459</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.783623</td>\n",
       "      <td>True</td>\n",
       "      <td>2.102038</td>\n",
       "      <td>1.044400</td>\n",
       "      <td>1.054367</td>\n",
       "      <td>1.065459</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>795</th>\n",
       "      <td>1.618170</td>\n",
       "      <td>True</td>\n",
       "      <td>-0.310631</td>\n",
       "      <td>-0.144106</td>\n",
       "      <td>-0.255406</td>\n",
       "      <td>-1.110335</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>796</th>\n",
       "      <td>0.572125</td>\n",
       "      <td>False</td>\n",
       "      <td>-0.564596</td>\n",
       "      <td>-0.563579</td>\n",
       "      <td>-0.910292</td>\n",
       "      <td>-1.110335</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>797</th>\n",
       "      <td>-0.346469</td>\n",
       "      <td>False</td>\n",
       "      <td>0.070317</td>\n",
       "      <td>-0.703403</td>\n",
       "      <td>-2.220065</td>\n",
       "      <td>-2.198232</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>798</th>\n",
       "      <td>-0.159042</td>\n",
       "      <td>False</td>\n",
       "      <td>0.705230</td>\n",
       "      <td>-0.563579</td>\n",
       "      <td>-0.255406</td>\n",
       "      <td>-0.022438</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>799</th>\n",
       "      <td>-0.724877</td>\n",
       "      <td>True</td>\n",
       "      <td>1.721090</td>\n",
       "      <td>0.135542</td>\n",
       "      <td>1.054367</td>\n",
       "      <td>1.065459</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>800 rows × 15 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          age    sex    height    weight     ap_hi     ap_lo  smoke   alco  \\\n",
       "0    1.336831  False -2.342353  0.275366  1.054367 -1.110335  False  False   \n",
       "1   -0.357123  False -1.834422 -0.074194 -0.255406  1.065459  False  False   \n",
       "2    1.673412  False  0.705230 -0.913140 -1.565179 -0.022438  False  False   \n",
       "3   -0.991616  False -0.183649 -0.703403 -0.255406  1.065459  False  False   \n",
       "4    0.783623   True  2.102038  1.044400  1.054367  1.065459   True   True   \n",
       "..        ...    ...       ...       ...       ...       ...    ...    ...   \n",
       "795  1.618170   True -0.310631 -0.144106 -0.255406 -1.110335  False  False   \n",
       "796  0.572125  False -0.564596 -0.563579 -0.910292 -1.110335  False  False   \n",
       "797 -0.346469  False  0.070317 -0.703403 -2.220065 -2.198232  False  False   \n",
       "798 -0.159042  False  0.705230 -0.563579 -0.255406 -0.022438  False  False   \n",
       "799 -0.724877   True  1.721090  0.135542  1.054367  1.065459  False  False   \n",
       "\n",
       "     active  cholesterol_at_risk  cholesterol_elevated  cholesterol_normal  \\\n",
       "0      True                False                 False                True   \n",
       "1      True                 True                 False               False   \n",
       "2     False                False                 False                True   \n",
       "3      True                False                 False                True   \n",
       "4      True                False                  True               False   \n",
       "..      ...                  ...                   ...                 ...   \n",
       "795    True                False                 False                True   \n",
       "796    True                False                 False                True   \n",
       "797    True                False                 False                True   \n",
       "798    True                False                 False                True   \n",
       "799   False                False                 False                True   \n",
       "\n",
       "     gluc_at_risk  gluc_elevated  gluc_normal  \n",
       "0           False          False         True  \n",
       "1           False          False         True  \n",
       "2           False          False         True  \n",
       "3           False          False         True  \n",
       "4            True          False        False  \n",
       "..            ...            ...          ...  \n",
       "795         False          False         True  \n",
       "796         False          False         True  \n",
       "797         False          False         True  \n",
       "798         False          False         True  \n",
       "799         False          False         True  \n",
       "\n",
       "[800 rows x 15 columns]"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "### Code - Standardizatio\n",
    "# standardize training data \n",
    "from sklearn import preprocessing\n",
    "\n",
    "#initialize \n",
    "numeric_features = ['age', 'height', 'weight', 'ap_hi', 'ap_lo' ]\n",
    "feature_train_numeric = feature_train[numeric_features]\n",
    "\n",
    "#scale\n",
    "scaler = preprocessing.StandardScaler().fit(feature_train_numeric)\n",
    "feature_train_numeric_scaled = scaler.transform(feature_train_numeric)\n",
    "\n",
    "# make dataframe again\n",
    "feature_train_numeric_scaled = pd.DataFrame(feature_train_numeric_scaled)\n",
    "# insert column names back\n",
    "feature_train_numeric_scaled.columns = numeric_features\n",
    "\n",
    "# change feature_train numeric values to standardized \n",
    "feature_train_copy = feature_train.copy()\n",
    "feature_train_copy[numeric_features] = feature_train_numeric_scaled\n",
    "feature_train_standardized = feature_train_copy\n",
    "# display\n",
    "feature_train_standardized"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "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>height</th>\n",
       "      <th>weight</th>\n",
       "      <th>ap_hi</th>\n",
       "      <th>ap_lo</th>\n",
       "      <th>smoke</th>\n",
       "      <th>alco</th>\n",
       "      <th>active</th>\n",
       "      <th>cholesterol_at_risk</th>\n",
       "      <th>cholesterol_elevated</th>\n",
       "      <th>cholesterol_normal</th>\n",
       "      <th>gluc_at_risk</th>\n",
       "      <th>gluc_elevated</th>\n",
       "      <th>gluc_normal</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.210617</td>\n",
       "      <td>False</td>\n",
       "      <td>-0.558726</td>\n",
       "      <td>2.657790</td>\n",
       "      <td>0.313821</td>\n",
       "      <td>-0.133761</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.753979</td>\n",
       "      <td>True</td>\n",
       "      <td>1.531272</td>\n",
       "      <td>0.118163</td>\n",
       "      <td>-0.331237</td>\n",
       "      <td>-0.133761</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.647335</td>\n",
       "      <td>False</td>\n",
       "      <td>0.416606</td>\n",
       "      <td>0.823615</td>\n",
       "      <td>-0.331237</td>\n",
       "      <td>-0.133761</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.214831</td>\n",
       "      <td>False</td>\n",
       "      <td>0.695273</td>\n",
       "      <td>-0.587289</td>\n",
       "      <td>-0.331237</td>\n",
       "      <td>-0.133761</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-1.753316</td>\n",
       "      <td>True</td>\n",
       "      <td>0.834606</td>\n",
       "      <td>-0.022927</td>\n",
       "      <td>0.313821</td>\n",
       "      <td>1.004630</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>195</th>\n",
       "      <td>0.372426</td>\n",
       "      <td>False</td>\n",
       "      <td>-0.280060</td>\n",
       "      <td>-0.305108</td>\n",
       "      <td>0.958879</td>\n",
       "      <td>1.004630</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>196</th>\n",
       "      <td>0.386600</td>\n",
       "      <td>True</td>\n",
       "      <td>2.506604</td>\n",
       "      <td>-0.305108</td>\n",
       "      <td>-0.331237</td>\n",
       "      <td>-0.133761</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>197</th>\n",
       "      <td>-0.189177</td>\n",
       "      <td>False</td>\n",
       "      <td>-0.419393</td>\n",
       "      <td>3.574878</td>\n",
       "      <td>2.248996</td>\n",
       "      <td>1.004630</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>198</th>\n",
       "      <td>0.343695</td>\n",
       "      <td>True</td>\n",
       "      <td>0.834606</td>\n",
       "      <td>0.118163</td>\n",
       "      <td>-0.331237</td>\n",
       "      <td>-0.133761</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>199</th>\n",
       "      <td>-0.143590</td>\n",
       "      <td>True</td>\n",
       "      <td>-1.116059</td>\n",
       "      <td>-0.657834</td>\n",
       "      <td>-0.976296</td>\n",
       "      <td>-0.133761</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>200 rows × 15 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          age    sex    height    weight     ap_hi     ap_lo  smoke   alco  \\\n",
       "0    1.210617  False -0.558726  2.657790  0.313821 -0.133761  False  False   \n",
       "1    0.753979   True  1.531272  0.118163 -0.331237 -0.133761  False  False   \n",
       "2    1.647335  False  0.416606  0.823615 -0.331237 -0.133761  False  False   \n",
       "3    1.214831  False  0.695273 -0.587289 -0.331237 -0.133761  False  False   \n",
       "4   -1.753316   True  0.834606 -0.022927  0.313821  1.004630  False  False   \n",
       "..        ...    ...       ...       ...       ...       ...    ...    ...   \n",
       "195  0.372426  False -0.280060 -0.305108  0.958879  1.004630  False  False   \n",
       "196  0.386600   True  2.506604 -0.305108 -0.331237 -0.133761  False  False   \n",
       "197 -0.189177  False -0.419393  3.574878  2.248996  1.004630  False  False   \n",
       "198  0.343695   True  0.834606  0.118163 -0.331237 -0.133761  False  False   \n",
       "199 -0.143590   True -1.116059 -0.657834 -0.976296 -0.133761  False  False   \n",
       "\n",
       "     active  cholesterol_at_risk  cholesterol_elevated  cholesterol_normal  \\\n",
       "0     False                False                 False                True   \n",
       "1      True                False                 False                True   \n",
       "2     False                False                 False                True   \n",
       "3      True                False                 False                True   \n",
       "4     False                False                 False                True   \n",
       "..      ...                  ...                   ...                 ...   \n",
       "195    True                False                 False                True   \n",
       "196    True                False                 False                True   \n",
       "197    True                 True                 False               False   \n",
       "198    True                False                  True               False   \n",
       "199    True                False                 False                True   \n",
       "\n",
       "     gluc_at_risk  gluc_elevated  gluc_normal  \n",
       "0           False          False         True  \n",
       "1           False          False         True  \n",
       "2           False          False         True  \n",
       "3           False          False         True  \n",
       "4            True          False        False  \n",
       "..            ...            ...          ...  \n",
       "195         False          False         True  \n",
       "196         False          False         True  \n",
       "197         False          False         True  \n",
       "198         False          False         True  \n",
       "199         False          False         True  \n",
       "\n",
       "[200 rows x 15 columns]"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# standardize test data\n",
    "\n",
    "#initialize \n",
    "numeric_features = ['age', 'height', 'weight', 'ap_hi', 'ap_lo' ]\n",
    "feature_test_numeric = feature_test[numeric_features]\n",
    "\n",
    "#scale\n",
    "scaler = preprocessing.StandardScaler().fit(feature_test_numeric)\n",
    "feature_test_numeric_scaled = scaler.transform(feature_test_numeric)\n",
    "\n",
    "# make dataframe again\n",
    "feature_test_numeric_scaled = pd.DataFrame(feature_test_numeric_scaled)\n",
    "# insert column names back\n",
    "feature_test_numeric_scaled.columns = numeric_features\n",
    "\n",
    "# change feature_test numeric values to standardized \n",
    "feature_test_copy = feature_test.copy()\n",
    "feature_test_copy[numeric_features] = feature_test_numeric_scaled\n",
    "feature_test_standardized = feature_test_copy\n",
    "# display\n",
    "feature_test_standardized"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a href=\"https://analyticsindiamag.com/why-data-scaling-is-important-in-machine-learning-how-to-effectively-do-it/\">https://analyticsindiamag.com/why-data-scaling-is-important-in-machine-learning-how-to-effectively-do-it/</a><br>\n",
    "<font color = skyblue> Scaling makes the algoritm less sensitive to the data. Scaling makes data more generalized and data point closer to eachother. </font><br>\n",
    "<a href=\"https://stats.stackexchange.com/questions/287425/why-do-you-need-to-scale-data-in-knn\">https://stats.stackexchange.com/questions/287425/why-do-you-need-to-scale-data-in-knn</a><br>\n",
    "<font color = skyblue> k-nn is based on classifying samples on their euclidean distance from other samples. The sample is classified based on it's k nearest samples. Standardizing all features makes sense here as it makes all the features influence the same size. </font>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "muqCazPIoDw4"
   },
   "source": [
    "-------\n",
    "\n",
    "It's time for us to train the model!\n",
    "\n",
    "***Exercise 1 D)*** \n",
    "sklearn.metrics.classification_report\n",
    "Train a k-NN model with $k=3$. Print out the confusion matrix.\n",
    "- What can you say about the confusion matrix?\n",
    "- How does the model perform with the different classes? Where do you think the differences come from?\n",
    "- What is the total classification accuracy? How would you interpret it?\n",
    "- Perform also the following:\n",
    "    - Make 1000 **different** train and test set splits.\n",
    "    - Run the k-NN model (with $k=3$) for each split and save its accuracy.  \n",
    "    - Plot accuracies in a histogram.\n",
    "    - Discuss your results\n",
    "\n",
    "\n",
    "*tip: `sklearn.metrics.classification_report` is a great way to build up a text report showing the main classification metrics*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "id": "NMR7Y2s6oDw4",
    "outputId": "33bd42f3-a25c-47de-cb12-908d698d08af"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\peppi\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\neighbors\\_classification.py:207: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples,), for example using ravel().\n",
      "  return self._fit(X, y)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([[124,  26],\n",
       "       [ 42,   8]], dtype=int64)"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "true neg: 124\n",
      "false pos: 26\n",
      "false neg: 42\n",
      "true pos: 8\n",
      "Total accuracy score: 0.66\n"
     ]
    }
   ],
   "source": [
    "### Code - kNN classifier\n",
    "from sklearn.metrics import accuracy_score\n",
    "from sklearn.metrics import confusion_matrix\n",
    "\n",
    "# initialize classifier\n",
    "neigh = KNeighborsClassifier(n_neighbors=3)\n",
    "\n",
    "#fit training data to classifier\n",
    "neigh.fit(feature_train_standardized, label_train)\n",
    "\n",
    "# make predictions from trained data \n",
    "predictions = neigh.predict(feature_test_standardized)\n",
    "predictions = predictions.astype(int)\n",
    "\n",
    "#labels as array\n",
    "labels = label_test['cardio']\n",
    "labels = list(map(int, labels))\n",
    "labels = np.asarray(labels)\n",
    "\n",
    "# print confusion matrix \n",
    "confusion = confusion_matrix(labels, predictions)\n",
    "display(confusion)\n",
    "tn, fp, fn, tp = confusion_matrix(labels, predictions).ravel()\n",
    "print(\"true neg:\", tn)\n",
    "print(\"false pos:\", fp)\n",
    "print(\"false neg:\", fn)\n",
    "print(\"true pos:\", tp)\n",
    "\n",
    "# total accuracy score \n",
    "print(\"Total accuracy score:\", accuracy_score(labels, predictions))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<font color = skyblue> Confusion matrix shows the 'best' case scenario of classifying (As many values are correctly classied). The threshold for classification could be changed so that all positives are correctly classied. This would lead to there being more false positives but it would be necissary if all positivies needed to be classified correctly."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "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>height</th>\n",
       "      <th>weight</th>\n",
       "      <th>ap_hi</th>\n",
       "      <th>ap_lo</th>\n",
       "      <th>smoke</th>\n",
       "      <th>alco</th>\n",
       "      <th>active</th>\n",
       "      <th>cholesterol_at_risk</th>\n",
       "      <th>cholesterol_elevated</th>\n",
       "      <th>cholesterol_normal</th>\n",
       "      <th>gluc_at_risk</th>\n",
       "      <th>gluc_elevated</th>\n",
       "      <th>gluc_normal</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.422412</td>\n",
       "      <td>False</td>\n",
       "      <td>-0.304682</td>\n",
       "      <td>0.132087</td>\n",
       "      <td>-0.270611</td>\n",
       "      <td>-0.043859</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-0.014353</td>\n",
       "      <td>False</td>\n",
       "      <td>-0.046368</td>\n",
       "      <td>0.482264</td>\n",
       "      <td>1.034480</td>\n",
       "      <td>1.052609</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.612151</td>\n",
       "      <td>False</td>\n",
       "      <td>-1.208785</td>\n",
       "      <td>0.202122</td>\n",
       "      <td>0.381935</td>\n",
       "      <td>-0.043859</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.376937</td>\n",
       "      <td>True</td>\n",
       "      <td>0.599420</td>\n",
       "      <td>-0.568268</td>\n",
       "      <td>-0.270611</td>\n",
       "      <td>-0.043859</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1.031565</td>\n",
       "      <td>True</td>\n",
       "      <td>0.857735</td>\n",
       "      <td>-0.428197</td>\n",
       "      <td>2.339572</td>\n",
       "      <td>2.149076</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>995</th>\n",
       "      <td>-0.630614</td>\n",
       "      <td>True</td>\n",
       "      <td>0.082790</td>\n",
       "      <td>-0.568268</td>\n",
       "      <td>-0.270611</td>\n",
       "      <td>-0.043859</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>996</th>\n",
       "      <td>-0.447147</td>\n",
       "      <td>True</td>\n",
       "      <td>-0.175525</td>\n",
       "      <td>-0.288126</td>\n",
       "      <td>-0.270611</td>\n",
       "      <td>1.052609</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>997</th>\n",
       "      <td>0.612100</td>\n",
       "      <td>False</td>\n",
       "      <td>0.082790</td>\n",
       "      <td>0.132087</td>\n",
       "      <td>0.381935</td>\n",
       "      <td>-0.043859</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>998</th>\n",
       "      <td>-0.116279</td>\n",
       "      <td>True</td>\n",
       "      <td>1.890994</td>\n",
       "      <td>-0.078020</td>\n",
       "      <td>-0.923156</td>\n",
       "      <td>-1.140326</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>999</th>\n",
       "      <td>-0.722739</td>\n",
       "      <td>False</td>\n",
       "      <td>-0.433840</td>\n",
       "      <td>-0.428197</td>\n",
       "      <td>1.034480</td>\n",
       "      <td>1.052609</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1000 rows × 15 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          age    sex    height    weight     ap_hi     ap_lo  smoke   alco  \\\n",
       "0    1.422412  False -0.304682  0.132087 -0.270611 -0.043859  False  False   \n",
       "1   -0.014353  False -0.046368  0.482264  1.034480  1.052609  False  False   \n",
       "2    1.612151  False -1.208785  0.202122  0.381935 -0.043859  False  False   \n",
       "3    1.376937   True  0.599420 -0.568268 -0.270611 -0.043859  False  False   \n",
       "4    1.031565   True  0.857735 -0.428197  2.339572  2.149076  False  False   \n",
       "..        ...    ...       ...       ...       ...       ...    ...    ...   \n",
       "995 -0.630614   True  0.082790 -0.568268 -0.270611 -0.043859  False  False   \n",
       "996 -0.447147   True -0.175525 -0.288126 -0.270611  1.052609  False  False   \n",
       "997  0.612100  False  0.082790  0.132087  0.381935 -0.043859  False  False   \n",
       "998 -0.116279   True  1.890994 -0.078020 -0.923156 -1.140326  False  False   \n",
       "999 -0.722739  False -0.433840 -0.428197  1.034480  1.052609  False  False   \n",
       "\n",
       "     active  cholesterol_at_risk  cholesterol_elevated  cholesterol_normal  \\\n",
       "0      True                False                 False                True   \n",
       "1      True                False                 False                True   \n",
       "2      True                False                 False                True   \n",
       "3      True                False                 False                True   \n",
       "4     False                False                 False                True   \n",
       "..      ...                  ...                   ...                 ...   \n",
       "995    True                False                 False                True   \n",
       "996   False                 True                 False               False   \n",
       "997    True                 True                 False               False   \n",
       "998    True                False                 False                True   \n",
       "999   False                False                 False                True   \n",
       "\n",
       "     gluc_at_risk  gluc_elevated  gluc_normal  \n",
       "0           False          False         True  \n",
       "1           False          False         True  \n",
       "2           False           True        False  \n",
       "3           False          False         True  \n",
       "4           False          False         True  \n",
       "..            ...            ...          ...  \n",
       "995         False          False         True  \n",
       "996         False          False         True  \n",
       "997         False          False         True  \n",
       "998         False          False         True  \n",
       "999         False          False         True  \n",
       "\n",
       "[1000 rows x 15 columns]"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#initialize \n",
    "numeric_features = ['age', 'height', 'weight', 'ap_hi', 'ap_lo' ]\n",
    "cardio_features_numeric = cardio_features[numeric_features]\n",
    "\n",
    "#scale\n",
    "scaler = preprocessing.StandardScaler().fit(cardio_features_numeric)\n",
    "cardio_features_numeric_scaled = scaler.transform(cardio_features_numeric)\n",
    "\n",
    "# make dataframe again\n",
    "cardio_features_numeric_scaled = pd.DataFrame(cardio_features_numeric_scaled)\n",
    "# insert column names back\n",
    "cardio_features_numeric_scaled.columns = numeric_features\n",
    "\n",
    "# change feature_train numeric values to standardized \n",
    "cardio_features_copy = cardio_features.copy()\n",
    "cardio_features_copy[numeric_features] = cardio_features_numeric_scaled\n",
    "cardio_features_standardized = cardio_features_copy\n",
    "\n",
    "# display\n",
    "cardio_features_standardized"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([  3.,  18.,  61., 169., 179., 291., 189.,  62.,  23.,   5.]),\n",
       " array([0.665, 0.683, 0.701, 0.719, 0.737, 0.755, 0.773, 0.791, 0.809,\n",
       "        0.827, 0.845]),\n",
       " <BarContainer object of 10 artists>)"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# initialize classifier\n",
    "neigh = KNeighborsClassifier(n_neighbors=3)\n",
    "acc = []\n",
    "for x in range (1000):\n",
    "    \n",
    "    # make splits\n",
    "    feature_train, feature_test, label_train, label_test = train_test_split(cardio_features_standardized, cardio_labels,\n",
    "                                   random_state=x, \n",
    "                                   train_size=0.8, \n",
    "                                   shuffle=True)\n",
    "    \n",
    "    #format labels\n",
    "    train_labels = label_train['cardio']\n",
    "    train_labels = list(map(int, train_labels))\n",
    "    \n",
    "    #fit training data to classifier\n",
    "    neigh.fit(feature_train, train_labels)\n",
    "    \n",
    "    #make predictions\n",
    "    pred = neigh.predict(feature_test)\n",
    "    \n",
    "    #format labels\n",
    "    test_labels = label_test['cardio']\n",
    "    test_labels = list(map(int, test_labels))\n",
    "    \n",
    "    #accuracy score\n",
    "    acc.append(accuracy_score(test_labels, pred))\n",
    "\n",
    "#plot histogram of accuracy values\n",
    "plt.hist(acc)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "1ceztBiPoDw4"
   },
   "source": [
    "<font color = skyblue> Values seem to have a normal distibution around a mean of about 0.76(This is the total accuracy). Accuracy score tells how many labels were classified correctly -> how many values mach in test_labels and prediction arrays. Accuracy seems to stay near the same values and don't vary much when no other parameters are changed. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "------\n",
    "\n",
    "One really common evaluation metric is <font color = darkorchid><b>the area under the receiver operating characteristic (AUROC or AUC-ROC)</b></font>.\n",
    "\n",
    "***Exercise 1 E)*** \n",
    "\n",
    "Explain what information you can learn about the one k-NN model you trained by using AUROC. Also, evaluate the performance by computing  the metric and plotting the related curve. Draw also the line for random guesses.\n",
    "\n",
    "- How would you interpret the results?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "### Code - AUROC and ROC Curves\n",
    "# https://www.statology.org/plot-roc-curve-python/\n",
    "\n",
    "from sklearn import metrics\n",
    "from sklearn import svm\n",
    "\n",
    "auc = metrics.roc_auc_score(labels, predictions)\n",
    "\n",
    "#initialize roc curve\n",
    "fpr, tpr, _ = metrics.roc_curve(labels,  predictions)\n",
    "\n",
    "#create ROC curve\n",
    "plt.plot(fpr,tpr,label=\"AUC=\"+str(auc))\n",
    "plt.ylabel('True Positive Rate')\n",
    "plt.xlabel('False Positive Rate')\n",
    "plt.legend(loc=4)\n",
    "\n",
    "#random guess \n",
    "ypoints = np.array([0,1])\n",
    "plt.plot(ypoints, linestyle = 'dashed')\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<font color = skyblue> k-nn is a better than just random guessing. The best results thats can be accieved with k-nn is classifying correctly litle over 50% of true values and only about 15% false positive values. AUC is the area under the curve. When it near 1 the trained model is good at classifying. </font>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "gQfehqAioDw4"
   },
   "source": [
    "__________\n",
    "## <font color = royalblue> 2. Classification accuracy using leave-one-out cross validation"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "DPZx-6JLoDw5"
   },
   "source": [
    "Even thought the train-test split gives us an unbiased estimate of the performance, we only evaluate the model one time. Especially with very small data sets, a test set will be very small. How can we be sure that the evaluation is accurate with this small test set and not just a good (or bad) luck? And what if we'd like to compare two models and the other seems to be better -- how can we be sure that it's not just a coincidence?\n",
    "\n",
    "Well, there's a great help available and it's called <font color = darkorchid><b>cross-validation</b></font>. This time, we'll take a look at <font color = darkorchid><i>leave-one-out cross-validation</i></font>.\n",
    "\n",
    "***Exercise 2 A)***\n",
    "\n",
    "Describe *in your own words*, how leave-one-out cross validation works? How does it actually differ from the basic train-test split?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "TXJmJ80DoDw5"
   },
   "source": [
    "<font color = skyblue> Data is divided into folds. In leave-one-out cross validations case the amount of folds id the count of the samples. All but one of the data is uded for training and one sample is used as the training set. Usually training/test split is bigger than in  leave-one-out cross-validation. </font>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---------------\n",
    "\n",
    "***Exercise 2 B)***\n",
    "\n",
    "Let's predict once again the **presence of cardiovascular disease** using all the other variables in the dataset as features. Now, use leave-one-out cross validation to estimate the model performance. Again, use a kNN classifier with $k=3$.\n",
    "\n",
    "- What is the classification accuracy? Compare the result with the one you got in the previous task.\n",
    "\n",
    "*tip: Sure this could be done manually, but `cross_val_score` is quite a handy function too.*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {
    "id": "znHjVwKDoDw5",
    "outputId": "3b73cc4a-eef7-4e9e-be17-19d79df71033"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.32\n"
     ]
    }
   ],
   "source": [
    "### Code - Leave-one-out cross validation\n",
    "# https://www.statology.org/leave-one-out-cross-validation-in-python/\n",
    "# https://medium.com/@svanillasun/how-to-deal-with-cross-validation-based-on-knn-algorithm-compute-auc-based-on-naive-bayes-ff4b8284cff4\n",
    "\n",
    "from sklearn.model_selection import LeaveOneOut\n",
    "from sklearn.model_selection import cross_val_score\n",
    "\n",
    "#define cross-validation method to use\n",
    "cv = LeaveOneOut()\n",
    "\n",
    "#build knn model and fit with training data\n",
    "model = KNeighborsClassifier(n_neighbors=3)\n",
    "model.fit(feature_train_standardized, label_train)\n",
    "\n",
    "#use LOOCV to evaluate model with testing data\n",
    "scores = cross_val_score(model, feature_test_standardized, label_test, scoring='neg_mean_absolute_error',\n",
    "                         cv=cv, n_jobs=-1)\n",
    "\n",
    "#view mean absolute error\n",
    "print(np.mean(np.absolute(scores)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<font color = skyblue> The lower the MAE score the better the model is able to to predict the labels. This model seems to be good at predicting the lables.</font>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "88BjCQL6oDw5"
   },
   "source": [
    "____________\n",
    "## <font color = forestgreen> 3. Model selection with leave-one-out cross validation"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "A8fcES_LoDw5"
   },
   "source": [
    "So far, we've trained one model at a time and I've given the value of k for you. Accuracy is what it is (no spoilers here), but could we still do a little better? Let's try that by <font color = darkorchid><b>hyperparameter tuning</b></font>.\n",
    "\n",
    "This time, we're going to train multiple models, let's say 30, and choose the best K-Nearest Neighbors model among the others. Almost all models have some parameters that need to be chosen. As does the k-NN, I just happened to choose the k value of 3 for you. Note, k-NN has many other hyperparameters too, but for the sake of simplicity, this time we'll focus only on the nearest neighbors. \n",
    "\n",
    "Now, you're supposed to find that *optimal value* for k.\n",
    "\n",
    "***Exercise 3***\n",
    "\n",
    "Repeat the model performance estimation with values $k=1...30$ using again leave-one-out cross validation.\n",
    "\n",
    "- Which value of k produces the best classification accuracy?\n",
    "- If the number of k is still increased, what is the limit that the classification accuracy approaches? Why? \n",
    "- Can you say something about the performance of this selected model with new, unseen data? Explain, how you could you estimate the performance of this selected model.\n",
    "\n",
    "*tip: `GridSearchCV` is a good class for this but will you get the other but the best k value out of it?*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {
    "id": "RJ7570_ZoDw6",
    "outputId": "91caaf74-2636-46f7-a4e3-4e5ba98e4c4e"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "### Code - Select best k\n",
    "x_range = range(1, 31)\n",
    "mae = []\n",
    "for x in x_range:\n",
    "    import logging\n",
    "\n",
    "    #build knn model\n",
    "    model = KNeighborsClassifier(n_neighbors=x)\n",
    "    model.fit(feature_train_standardized, label_train)\n",
    "\n",
    "    #use LOOCV to evaluate model\n",
    "    scores = cross_val_score(model, feature_test_standardized, label_test, scoring='neg_mean_absolute_error',\n",
    "                         cv=cv, n_jobs=-1)\n",
    "    #mean absolute error\n",
    "    mae.append(np.mean(np.absolute(scores)))\n",
    "\n",
    "#plot mean absolute error values\n",
    "plt.plot(x_range, mae)\n",
    "plt.xlabel('Value of k for KNN')\n",
    "plt.ylabel('Cross-Validated mean absolute error')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<font color = skyblue> The mea value setteled to little under 0.26 at around k=8-13. Havink k be over 13 doesn't gives us any better results.  </font>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "9r2v1LEDoDw6"
   },
   "source": [
    "________________\n",
    "## <font color = red>  4. Testing with training data <font color = red> (this should *never* be done out of this exercise!) </font>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "j6B6L5HWoDw6"
   },
   "source": [
    "Oh, but what if we just used the ***whole data*** in training? Wouldn't we like to use as much data as possible to discover the underlying pattern in the data so why **not** to use the whole data?\n",
    "\n",
    "*Never ever* do this out of this exercise. Why? Let's take a look.\n",
    "\n",
    "***Exercise 4***\n",
    "\n",
    "Repeat the previous task but use the whole data for training. Plot the resulting classification accuracy versus $k = 1...30$. Include the values from the previous task in the *same figure*.\n",
    "\n",
    "- Comment your result and answer, why you should not test with training data?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {
    "id": "ExWjmjQ5oDw6",
    "outputId": "f7bd8eac-a636-400b-a8a4-9ca269129628"
   },
   "outputs": [],
   "source": [
    "### Code - Train with whole data\n",
    "\n",
    "### Code - Select best k\n",
    "x_range = range(1, 31)\n",
    "mae_all = []\n",
    "for x in x_range:\n",
    "    import logging\n",
    "\n",
    "    #build knn model with all of the data\n",
    "    model = KNeighborsClassifier(n_neighbors=x)\n",
    "    model.fit(cardio_features_standardized, cardio_labels)\n",
    "\n",
    "    #use LOOCV to evaluate model with all of the data\n",
    "    scores = cross_val_score(model, cardio_features_standardized, cardio_labels, scoring='neg_mean_absolute_error',\n",
    "                         cv=cv, n_jobs=-1)\n",
    "    #mean absolute error\n",
    "    mae_all.append(np.mean(np.absolute(scores)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#plot mean absolute error values\n",
    "plt.plot(x_range, mae_all)\n",
    "plt.plot(x_range, mae)\n",
    "plt.xlabel('Value of k for KNN')\n",
    "plt.ylabel('Cross-Validated mean absolute error')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    " <font color = skyblue> When the whole data is used for training and testing the mea values are avereged lower. AKA better score -> means in this case the model is propably over fitted. The meas score setteled around 0.225 but not as clearly as with training/testing combo. We first get this low value et k=8 like with training/testing combo."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "_____________________\n",
    "\n",
    "## <font color = darkorange> 5. Comparison of ridge regression and kNN regression </font>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The previous exercises were about classification. Now, we are ready to see another kind of supervised learning - regression. We're going to use <font color = darkorchid><b>Ridge Regression</b></font> and <font color = darkorchid><b>K-Nearest Neighbors Regression</b></font>, and compare the performances of these models."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To make this exercise more intuitively, let's change the data to another somewhat famous dataset called *The Boston Housing Dataset*. The Boston Housing Dataset is a derived from information collected by the U.S. Census Service concerning housing in the area of Boston, Massachusetts. Let's have a quick peek at the variables:\n",
    "\n",
    "- **CRIM** - per capita crime rate by town\n",
    "- **ZN** - proportion of residential land zoned for lots over 25,000 sq.ft.\n",
    "- **INDUS** - proportion of non-retail business acres per town.\n",
    "- **CHAS** - Charles River dummy variable (1 if tract bounds river; 0 otherwise)\n",
    "- **NOX** - nitric oxides concentration (parts per 10 million)\n",
    "- **RM** - average number of rooms per dwelling\n",
    "- **AGE** - proportion of owner-occupied units built prior to 1940\n",
    "- **DIS** - weighted distances to five Boston employment centres\n",
    "- **RAD** - index of accessibility to radial highways\n",
    "- **TAX** - full-value property-tax rate per \\$10,000\n",
    "- **PTRATIO** - pupil-teacher ratio by town\n",
    "- **B** - 1000(Bk - 0.63)^2 where Bk is the proportion of blacks by town\n",
    "- **LSTAT** - % lower status of the population\n",
    "- **MEDV** - Median value of owner-occupied homes in \\$1000's\n",
    "\n",
    "\n",
    "No worries if the variables don't make a lot of sense. However, the object is to predict **the median house value**. So the target variable is now <font color = darkorchid><b>MEDV</b></font> and all the others are <font color = darkorchid><b>features</b></font>.\n",
    "\n",
    "There's the code chunk for loading data provided again. <font color = red>Again, the data file should be located in the same directory as this notebook file!</font>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "metadata": {},
   "outputs": [],
   "source": [
    "### Loading code provided\n",
    "# ------------------------------------------------------\n",
    "# The data file should be at the same location than the \n",
    "# exercise file to make sure the following lines work!\n",
    "# Otherwise, fix the path.\n",
    "# ------------------------------------------------------\n",
    "\n",
    "# Data path\n",
    "data_path = 'ex2_boston_housing_data.csv'\n",
    "\n",
    "# Load the data \n",
    "boston_data = pd.read_csv(data_path, dtype='float')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 122,
   "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>CRIM</th>\n",
       "      <th>ZN</th>\n",
       "      <th>INDUS</th>\n",
       "      <th>CHAS</th>\n",
       "      <th>NOX</th>\n",
       "      <th>RM</th>\n",
       "      <th>AGE</th>\n",
       "      <th>DIS</th>\n",
       "      <th>RAD</th>\n",
       "      <th>TAX</th>\n",
       "      <th>PTRATIO</th>\n",
       "      <th>B</th>\n",
       "      <th>LSTAT</th>\n",
       "      <th>MEDV</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.00632</td>\n",
       "      <td>18.0</td>\n",
       "      <td>2.31</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.538</td>\n",
       "      <td>6.575</td>\n",
       "      <td>65.2</td>\n",
       "      <td>4.0900</td>\n",
       "      <td>1.0</td>\n",
       "      <td>296.0</td>\n",
       "      <td>15.3</td>\n",
       "      <td>396.90</td>\n",
       "      <td>4.98</td>\n",
       "      <td>24.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.02731</td>\n",
       "      <td>0.0</td>\n",
       "      <td>7.07</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.469</td>\n",
       "      <td>6.421</td>\n",
       "      <td>78.9</td>\n",
       "      <td>4.9671</td>\n",
       "      <td>2.0</td>\n",
       "      <td>242.0</td>\n",
       "      <td>17.8</td>\n",
       "      <td>396.90</td>\n",
       "      <td>9.14</td>\n",
       "      <td>21.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.02729</td>\n",
       "      <td>0.0</td>\n",
       "      <td>7.07</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.469</td>\n",
       "      <td>7.185</td>\n",
       "      <td>61.1</td>\n",
       "      <td>4.9671</td>\n",
       "      <td>2.0</td>\n",
       "      <td>242.0</td>\n",
       "      <td>17.8</td>\n",
       "      <td>392.83</td>\n",
       "      <td>4.03</td>\n",
       "      <td>34.7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.03237</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.18</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.458</td>\n",
       "      <td>6.998</td>\n",
       "      <td>45.8</td>\n",
       "      <td>6.0622</td>\n",
       "      <td>3.0</td>\n",
       "      <td>222.0</td>\n",
       "      <td>18.7</td>\n",
       "      <td>394.63</td>\n",
       "      <td>2.94</td>\n",
       "      <td>33.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.06905</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.18</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.458</td>\n",
       "      <td>7.147</td>\n",
       "      <td>54.2</td>\n",
       "      <td>6.0622</td>\n",
       "      <td>3.0</td>\n",
       "      <td>222.0</td>\n",
       "      <td>18.7</td>\n",
       "      <td>396.90</td>\n",
       "      <td>5.33</td>\n",
       "      <td>36.2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      CRIM    ZN  INDUS  CHAS    NOX     RM   AGE     DIS  RAD    TAX  \\\n",
       "0  0.00632  18.0   2.31   0.0  0.538  6.575  65.2  4.0900  1.0  296.0   \n",
       "1  0.02731   0.0   7.07   0.0  0.469  6.421  78.9  4.9671  2.0  242.0   \n",
       "2  0.02729   0.0   7.07   0.0  0.469  7.185  61.1  4.9671  2.0  242.0   \n",
       "3  0.03237   0.0   2.18   0.0  0.458  6.998  45.8  6.0622  3.0  222.0   \n",
       "4  0.06905   0.0   2.18   0.0  0.458  7.147  54.2  6.0622  3.0  222.0   \n",
       "\n",
       "   PTRATIO       B  LSTAT  MEDV  \n",
       "0     15.3  396.90   4.98  24.0  \n",
       "1     17.8  396.90   9.14  21.6  \n",
       "2     17.8  392.83   4.03  34.7  \n",
       "3     18.7  394.63   2.94  33.4  \n",
       "4     18.7  396.90   5.33  36.2  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "(467, 14)"
      ]
     },
     "execution_count": 122,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "display(boston_data.head())\n",
    "boston_data.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "-----------\n",
    "\n",
    "With regression, we can examine **the relationship between two or more variables**. This relationship is represented by an *equation*, which itself represents how much y changes with any given change of x. So for example, we could use *age* as an independent variable to predict *height* (a dependent variable). \n",
    "\n",
    "We start by looking at the relationships between the variables in the given dataset.\n",
    "\n",
    "\n",
    "***Exercise 5 A)***\n",
    "\n",
    "Make scatter plots where you present each feature versus the target variable `MEDV`.\n",
    "\n",
    "- What can you say about the relationships?\n",
    "\n",
    "*tip: seaborn.pairplot is quite a handy function where you can set x_vars and y_vars to point out the variables you need*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 123,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<seaborn.axisgrid.PairGrid at 0x25bef589600>"
      ]
     },
     "execution_count": 123,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAADHcAAAD7CAYAAABaUZnuAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/NK7nSAAAACXBIWXMAAA9hAAAPYQGoP6dpAAEAAElEQVR4nOy9eZhkV3nm+Z5zl1gzcq2qrFVSVWkpSqWFRQhZSEDbFpiHMYtNd0O3+8HYzQjo6QZMewQez/TYRgaz9GIka2Q3DRiMgUbQGCQBNlIhSkJCa6pUUm2qLZfKNfa46znzx4kbGZEZy43MyMzIzO/3PIXIyJtxb9y455zvfMv7MSmlBEEQBEEQBEEQBEEQBEEQBEEQBEEQBEEQBEEQBEEQBEEQBEEQBLEm8LW+AIIgCIIgCIIgCIIgCIIgCIIgCIIgCIIgCIIgCIIgCIIgCIIgiM0MFXcQBEEQBEEQBEEQBEEQBEEQBEEQBEEQBEEQBEEQBEEQBEEQBEGsIVTcQRAEQRAEQRAEQRAEQRAEQRAEQRAEQRAEQRAEQRAEQRAEQRAEsYZQcQdBEARBEARBEARBEARBEARBEARBEARBEARBEARBEARBEARBEMQaQsUdBEEQBEEQBEEQBEEQBEEQBEEQBEEQBEEQBEEQBEEQBEEQBEEQawgVdxAEQRAEQRAEQRAEQRAEQRAEQRAEQRAEQRAEQRAEQRAEQRAEQawhVNxBEARBEARBEARBEARBEARBEARBEARBEARBEARBEARBEARBEASxhlBxB0EQBEEQBEEQBEEQBEEQBEEQBEEQBEEQBEEQBEEQBEEQBEEQxBqy4Ys7pJTIZrOQUq71pRAEsURoHBPE+ofGMUGsf2gcE8TGgMYyQax/aBwTxPqHxjFBrH9oHBPE+ofGMUGsf2gcE8TGgMYyQax/aBwTxPqHxjFBdBcbvrgjl8uht7cXuVxurS+FIIglQuOYINY/NI4JYv1D45ggNgY0lgli/UPjmCDWPzSOCWL9Q+OYINY/NI4JYv1D45ggNgY0lgli/UPjmCDWPzSOCaK72PDFHQRBEARBEARBEARBEARBEARBEARBEARBEARBEARBEARBEARBEN0MFXcQBEEQBEEQBEEQBEEQBEEQBEEQBEEQBEEQBEEQBEEQBEEQBEGsIVTcQRAEQRAEQRAEQRAEQRAEQRAEQRAEQRAEQRAEQRAEQRAEQRAEsYZQcQdBEARBEARBEARBEARBEARBEARBEARBEARBEARBEARBEARBEMQaQsUdBEEQBEEQBEEQBEEQBEEQBEEQBEEQBEEQBEEQBEEQBEEQBEEQawgVdxAEQRAEQRAEQRAEQRAEQRAEQRAEQRAEQRAEQRAEQRAEQRAEQawhVNxBEARBEARBEARBEARBEARBEARBEARBEARBEARBEARBEARBEASxhlBxB0EQBEEQBEEQBEEQBEEQBLHhsT1/rS+BIAiCIAiCIAiCIAiCIAiCIAiCIBpCxR0EQRAEQRAEQRAEQRAEQRDEhiZve5jIWGt9GQRBEARBEARBEARBEARBEARBEATREH2tL4AgCIIgCIIgCIIgCIIgCIIgVop00cFswQFjbK0vhSAIgiAIgiAIgiAIgiAIgiAIgiAaQsUdBEEQBEEQBEEQBEEQBEEQxIZkKmcjZ7lrfRkEQRAEQRAEQRAEQRAEQRAEQRAE0RIq7iAIgiAIgiAIgiAIgiAIgiA2FEJITOZsFB1vrS+FIAiCIAiCIAiCIAiCIAiCIAiCIELB1/oCAv78z/8cjDH8h//wHyqvWZaFD33oQxgcHEQymcS73vUuXLx4sePnFkJi5EIGDx+fwsiFDISQHT8HQRArC41jgtic0NgniO6CxiRBEKsJzTkEQRDdR7fMzZ4vMJYpUWEHQXSIbhnbRGvouyKI7oPGJUF0DzQeCYIgVg+acwmCIJpD82R70P0iNhtd0bnjiSeewD333INrrrmm5vWPfOQj+MEPfoBvfetb6O3txYc//GG8853vxM9//vOOnfvIyWnc/fApnJrMw/UlDI1h39Ykbr91H27aP9Sx8xAEsXLQOCaIzQmNfYLoLmhMEgSxmtCcQxAE0X10y9xsez4uZmx4QqzaOQliI9MtY5toDX1XBNF90LgkiO6BxiNBEMTqQXMuQRBEc2iebA+6X8RmZM07d+Tzebz3ve/Fvffei/7+/srrmUwGf/M3f4PPf/7zeNOb3oRXvepV+NKXvoQjR47gscce68i5j5ycxifuG8Gx8SwSER1beyJIRHQcG8/hE/eN4MjJ6Y6chyCIlYPGMUFsTmjsE0R3QWOSIIjVhOYcgiCI7qNb5uai42E8bVFhB0F0iG4Z20Rr6LsiiO6DxiVBdA80HgmCIFYPmnMJgiCaQ/Nke9D9IjYra17c8aEPfQhvfetb8au/+qs1rz/55JNwXbfm9auuugp79uzBo48+uuzzCiFx98OnkLc9DKeiiBoaOGeIGhqGUxHkbR93P3yK2vcQRBdD45ggNic09gmiu6AxSRDEakJzDkEQRPfRLXNzpuRiImNBSFoDCKITdMvYJlpD3xVBdB80Lgmie6DxSBAEsXrQnEsQBNEcmifbg+4XsZnR1/Lk3/jGN/DUU0/hiSeeWPS7iYkJmKaJvr6+mte3bduGiYmJhu9p2zZs2678nM1m6x53dCyLU5N59MdNMMZqfscYQ1/cwKnJPI6OZXFoV28bn4ogiOVC45gg1j9hx/FSoLFPEKsDrccEsTFYyTV5LaA5h9iMbLRxTGw8umFunsnbyJTcFXnvTkDjmFiPdMPY7ia6eRzTd0UQ4VjNcUzjkiBWhqWMYxqPBNF9dLNtTSwPmnM3DzSOCWJpdNM8uR7GcTfdL4JYbdasc8f58+fx7//9v8fXvvY1RKPRjr3vnXfeid7e3sq/3bt31z1utujA9SVMTd0CCYmS4yNnuSg5PkyNwRUSs0WnY9dGEEQ4ljKOF45hCYmIxmkcE8QaEXYcL4WFa3hAMA84no+S62Mmbzd4B4IgwkDrMUFsDFZyTV4qQkiMXMjg4eNTGLmQaUtNpZEdEEBzDrER6cZxTBDVLGduXs6aEPz9RMbq6sIOgMYx0R20O97I7qqlm8dxJ76r5c7HBLEeWKlxXG/80BxKECvDUsZxt4xHWmsJYp5utq2J5dFszpWQEEKi4Ph46uwczYPrHBrHBLE0uilneTXG8XJt4G6x5QliLVizzh1PPvkkJicn8cpXvrLymu/7OHz4MP7yL/8SDz74IBzHQTqdrunecfHiRQwPDzd83zvuuAMf/ehHKz9ns9m6E89A3IShMTi+gOdKTOVs2J4PKQHGAJ1zxE0NA3GzMx+YIIjQtDuO0yUXmZJbM4YjuobemAGDMxrHBLEGhB3HS6F6DY9yDQCQt73KWh5sBv7zP56AqXPctH+oI+cliM0GrccEsTFYyTV5KRw5OY27Hz6FU5N5uL6EoTHs25rE7bfuC7Vm17MDqrF9QXMOseHotnFMEAtZ6ty83DXB8wUmshYcT3Tss6wUNI6JtWYp443srlq6eRwv97ta7nxMEOuFlRjHjcbPbQeHaQ4liBVgKeO4G2waWmsJopZutq2J5dFozg1i+ZbrQUjgv/30BH7y4kWaB9cxNI4JYml0U87ySo/jTtjA3WDLE8RasWadO/7ZP/tnGBkZwTPPPFP59+pXvxrvfe97K//fMAz84z/+Y+VvXnrpJZw7dw6ve93rGr5vJBJBKpWq+VePgztS2Lc1icmchQuzRViuD84YdI2BAbBcHwXHQ6ZEVV0Esdq0M44HkybGMyWUHK8yhjljKDkexjMlDCZNHNxR/+8Jglg5wo7jpRCs4XNFF1JK5G0Po3MlWK4PBgDlhPLzs0V84r4RHDk53bFzE8RmgtZjgtgYrOSa3C5HTk7jE/eN4Nh4FomIjq09ESQiOo6N50Kv2QvtgGqklEgXXezbmqQ5h9hQdNM4Joh6LGVuXu6aYHs+xtLro7ADoHFMrC1LHW9kd9XSzeN4Od9VJ2x0glgvdHocNxs/9x4+hcGkSXMoQXSYpYzjtbZpaK0liMV0s21NLI96c24Qyy85HqQEYoaGvphB8+A6h8YxQSyNbspZXslx3CkbeK1teYJYS9asuKOnpwdXX311zb9EIoHBwUFcffXV6O3txfvf/3589KMfxU9/+lM8+eSTeN/73ofXve51uPHGG5d9fs4ZPnDLXtiegCskwFT7N19IeEJVipmahnsOn6ZWcASxDpAAhJTwfQkhJWjUEsTGhXOG22/dh2REw3imhLF0CZ4vAKbWcI1xbOuNYntvFHnbx90Pn6K1nCBWCVqPCYJohBASdz98Cnnbw3AqiqihgXOGqKFhOBUJvWbX2gEW5ooOsiUXc0UH4xkLyYiG22/dB87ZKn0ygiCI9cdyW6EvpHpunsjaKLmqo2LJ9TGRtRfNzctdE4qOh/G0BU+0V9ghpcQPnhvDx7757KJAEEFsVJYz3tod22Gvp5PzD6EIvquEyXFhroTpvI2C7aHoeE2/q07Z6ASxGWk1fgqOslMStHcliDVnLX1JtNYSBLHZWDzn2pVYvgSgcY6tqShipk7zIEEQm5LNkLPczAbe1mMiXXTxqR8ew7Pn023FhTvlnySI9YK+1hfQjC984QvgnONd73oXbNvGbbfdhrvuuqtj798bM2FqDI4HeD6AcvoZB9ATNdAXN3BqMo+jY1kc2tXbsfMSBNEZjo5lMZN3MBA3MFtw4frzCz4H0J8wMJN3aAwTxAbkpv1DeO9r9+C//uNJ2GWlVuEDGgMGEiaSEWXidHItF0Li6FgWs0UHA3HVhYA2CARB6zFBEOE4OpbFqck8+uMmGKtdPxljba3ZgR3wxYdOYTxdgkB5Hx8z8N7X7sGNewcxciFDazZBEEQdOtEKvR437R/Cp95xqPLeGSFhcIYD23sWvXewJvTFDFiugCcEdM4RNXjLNSFTcjGTt9u+vkzJxed/fBw/O6EUwV67dwDvfnXnWswTRLeyXBsszNgO6y9ZqfmHmCcVMzCRtZG1XACAzjmuHE7ijrccqLnHwXf21Nk5HBvPoi9mLNtGJ4jNRpj5dSbv4Larh3Hf06N1967dOveRH5zYiLTyJa3UeFyOLUZjkSCIbqfRPFU9546lLQR5uxpq4/i05yAIYrMRzJunJgvQGMBQm7Os8p0iSEb1dTc3Vq8Js3kHJy/mFtnAedvDVM6C5Qq8MJ7F7335l7iqTvxgIe3EHghiI9FVxR0PPfRQzc/RaBRf/OIX8cUvfnFFzvfIyWnkbR8AwBkqBqUAMFNwULA9RAwNs8WVb3NEEET7zBYdFGwfBcfFQq1GARW4T0RAY5hYUci5ujYcOTmNr/3iHBhTDniNq52PkBKzBQcxU0MyoiOicWSEbDkPtPoeKQGBIBpD6zFBbF7asYNmiw5cX8LU6jcQDbtmA/N2gM4ZdvTFwBggJVByffz1Iy/jfz07hpm8Q2s2QRDEAoJW6DnLRdzUYWgcQkq8MJbFJ+4bwafecWjZBR437h1suTYE9mOm5MLxBaQEGAMiOseWnijihlZ3TZjO28iW3Lav66mzc7jzgRcxk59/v//ykxN4+3U7Yepr1th6U0G+k7WjEzZYs7Ed1l8SzD9520N/3ISpcTi+wLHxXEfmn81O9f3dPRCDEIDl+ii6PrKWt+jY4DsrOD5ylouS42NrKlpJsgpox0ZfCjQ3EOuZMPPrlOPjhyPjdfeuX/vFORzc0VuZ+7plPJAfnNioNPMlLRyPYQkzbpdqi9FYJAii22k2TwGozLkDCROzeQdcY5AL4vjAyuw5usWuIgiCqKZ63syU3Eq3R17OeZJl+3QmbyOic7gr6I/pNAvXBF8IZC0XvgBkDIgaHAXHx+hcCb6U0DggfcDQWGjfYNjYA0FsJLqquGM1EULiwaMTkDKofVuM5Qm4vsD52eKqXhtBEOHoixkoOB68hZmkZTwBFGwPfTFjdS+M2DSQc3VtqG7hN5Q0UXR8gAGcMXAGeEJiKmchYSZg+wIGZxiImw3fr9X3SAkIBNEcWo8JYnPSrh00EDdhaAyOLxDl2qLfh1mzgVo7YFsqAtuVSvFd4+AcOD9bwmzBwSWDcUQ0jdZsgiA2NdXB7L6YgbseOoXZggNfSGQtq1JUYWocrq/m1xv3Di4rIMI5a6kmdn62iLzjAVJC17hKrAJQcgVG50oY6jFr1gQhJCZzNoqO1/R9F+L6Av/9kZfxzV9eqPH/3rx/CJ9797VU2LFKbGbfSTcklHTKBqs3tsP6S6rtt+FUtKLYF+UahlMcE1m7I/PPZqXR/Y1HNMRsH9N5G5958CV8+9IBPH5mtuY7ixoaCrYLy1UB9p39sZoCj7DPx1LYzHMDsTFoNb9ano+S44MB2Nkfrdm7pmI6Lmadytz32OmZrhgP5AcnNirNfEkLx2NYWyTsOrYUW4zGIkEQ3U6zeeqO+0aQiuqVOTdT8sA4A4MSa6yO4zPGOr7noH0GQRDdSPW82Rc3MFuYL9oQUglmBg5sISXGMxYGE8aK+GM6zcI1wfEEJrIuPAFM5m3MFB1EdQ5fSvhSddyQUHEJAOiN6shYbih7PEzsgSA2Epu2uOPoWBbnZwsNCzsCGGN44PkJvOeGPRRYIIguQ0gJXzQfxb6QELLVSCfC0A0B6W6CnKtrR9DGOqJrmMypxCQJgEGqTh6cwfYEiraH2aKLXf0xCCkhhFz0zLb6Hv/07VfjnsOnKQGBIJogpITXYj32aD0miA3FUuygA8M92JqK4vRUAVuSJmKmVllXpZRIF10c2N6DgztSTc9dbQecnSnB9nyo6UUimGaklIBk4JzRmk0QxKZECImvP34O33j8HCayFsobJmSK853WdM4qRRWWJ8CZwAtjmRVv9S6ExAPPT1S6KDMGMKggv8EBVwhM5Wy89rIBHNyRgucLTGQtOI0qiRtwbraIP/3BMZyczFde0znD779+Lz5+25W0FiyDdvxDm9l30i0JJQd3pLBvaxLHxnMYTvGK/QW0Z4MtpJ2CjcB+64+bNecHVPylL27g1GR+xeefjUq9+5u3PUzlbNieDyEkRi6k8a6/OgIhZc13JiERNXSUHA++FDVJVst5PlqxWnMD+bOJlaTV/DqTd8AYkIjU7l1VtzINvTE193398XP465+dDjUeVvKZpkI8YiPTyJe0cDyGtUXaWcfatcW6cSzSekoQK8t6G2Ot5qkL6RLG0xaGkmZlzhVCwkdZnb4cx7dcgajBO7rn2Mw+CIIguo9gfp8u2PgvPzlRmTctV8BvkTfh+ALxiN5xf0ynWbgmFBwf4xkLvhRgULEHISQs14cvAYMzCKk+HwBMZC1wBuic44WxLPkGCWIBm7a4Y7pgI11qrfaWimo4PUWBBYLoRp69kAl93PV7+lf4ajY23RKQ7ha60bm6mZgtOig4Poq2BwlA1xg8XxV4SAkIX4IBODtXAgCcnyvh9r99ctEzG+Z7/OyPjuNipkQJCATRhKfPp0MfR+sxQax/lmIHBbbk+dkiCraLvO0iomvY2hOBoasATjKi4fZb97W0nRbaARpnYBzwBSqFZgwMnhAAlCIirdkEQWwmjpycxp33H8ML4zkIKaExhojOEdEZ3PI8aWgMvDx/VxdVZC0P0wV7Ra/v6FgWp6fy2NoTxVTOhuerNuxBoQnKSmVvvno7XCFwMWOX5/RwSCnxg5FxfPGnp2BXFYTsGYjjk79xFa4Y7u4kiW6nHf/QZvaddFNCCecMt9+6D5+4bwQTWRt9cQMRjcP2RVs22ELaKdiYLTpwfQlTq98tJ6JxZITEbNGp+3uiOQvvb972MDpXUmsAZ+Ca6qh5crIAy/NrxiMDw5aeCEbnBHyhEqyKjg/O2bKej2as1txA/mxipWk1v0Z0DlcITGbtmr2rlIDl+rBdHzFTw989fi7UeFjp7h5UiEdsZBr5kqrHYzyih7JF2l3H2rXFum0s0npKECvLehxjreapuKEhU3IxmbUAxqBxVulgJMpxfM6AguMhXZId23NsZh8EQRDdR/X8XnR85CwXpq6h4PiQZWHaVhTs9rpYrwXVawIATOWscncODsEBxxM1wvsqPjH/ipDqnycEHM/BIyenaL9JEFVs2t7zs3mnpeI/AHCuAi8UWCCI7kNK2bL7jiwft9YohbQMHj4+hZELmVCGWrcQBKRfGMtA4wxxU4PGGV4Yy+IT943gyMnptb7EVacd5yrRefpiBizXh5ASusagcw5D56j2w6ikIImBuIm+mAGNMxwdzeBj33oWdz90EiMXMhgZzbT8Hs/PFlFyRdMEBJcSEIhNzni61NHjCILobtq1g6ptybipYTAZgaFxWK6P83NFzBVsHNjeEzrJcaEdwBkDq2i+K4JEtmo24pq9nvcYBEGsDEdOTuOO+0bw4kQOkBKmpoLolieQrRK5Ud0PZbkDm1IKY+XX0wV3Ra8xSELuixnY2R9D1NDUdfjqeqKGhmREx9aeCMbTVluFHZmiiz/+X0fx+R+fqCnseNu12/FX/+qVuHxbz0p8pE1DsKYfG88iUf6OEhG9UrCw0D80MprBi+M5mBqH5QpUe/E2su9kYUJJ1NBUNzFDw3Aqgrzt4+6HT63qun3T/iF86h2HcGB7D4q2h8m8jaLttWWDLSRMwUZgew3ETRgag+35KJUD2qVyMBsAbF/A4AwD5UAw0Z6dF9xfx1fjbCpn19jKAANnSshMSIl00akZj8mIXjMfzxScZT8fzVgNvyr5s4nVotn8+uE37a/YN9V7V84YdI1BSImS6+Ni1mo5Hr7++Dl84r4RHB1NwxfzdtvR0UzHnul25nWCWG808iVVj0fL9dEXM1q+VzvrWLCeu0Li916/F1cNJ1vaYt00Fmk9JdYb681Xul7HWMt5yuBKjFGiMudqnMGsiuULCbie39E9B+VvEATRLSz0ofbGDIAxOJ6P0TnV0ahlsiOAmYKDkdFwotdrRbAmGJoSCSm5Qs31DNAYg6nxSgePVggA337yQuj1e72t+wSxFDZt545MKdyGdzbvQNc4zs8WV/iKCIJol1R03sm20Bio/rn6uLVgtRQXVqJlZxCQni2ogrisZVVaFZsah+vLTalwQIqDa48Qsuz4EWBMKQ0aGocQAm45h8fUOdIlB3NFQEJCSkAWXXz+x8fx1UfPYiBhouD4lSryhUQ0DiGVeojjC0S5tugYSkAgCITbjbdzHEEQXU07dlAzWzJqcHAAuwcS+NK/eQ04Zxi5kFlkyy60cYVUazowX+wtJdCq7HujrdnrUdWNIDYjK7FPb3auux8+hUzRhZQSusYr3TkMDjh+rSKW7Yq6M+dYurAi1xdQnYScjOhIRDRYjoAnBHTOASaRtzwIoYr1wvLk2Tn8+f0vYqYwvw9PRXV8/LYr8Ss0Ly6bdhUwj5ycxqd+eAwzBRsMSt08omvY0hNBMqJCEivlO1ntcbfwXO0klATHr8a13rR/CDfuHezY+arHcit/ycEdKQwmTVV4BlTswYjOMZRUBS8Htvfg4I7Usj7jRqFdO+/gjhT2bkni+bEMYroqota4SlqVUsITEjGDI27q0JgD2/NhOQIxc/57S0Z0aH1RZIoePvSm/XjVnv66z0cnxlen/aoLr+nAcA/5s4lVpdH8OjKaqdm7+uW9LGOoJDZKqYo0mo4HX+Ibj5/DZNaC5QkIOV+wyxlguaKtZ7rROG5nXieI9Ug9X5Iy1WTN76vxPIHvPzeO0XQRO/vieNs120OvY4+cnMJnHnyxZj3fuyWJD75xP3YPxBuuo90yFik+TKw31puvdD2PsVbzlFvlf1o45xoag+urLrN3/MYr8I7rd4bec7Tai1D+BkEQ3UA9H6qUau/GOIPnC1zM2ggjZ+QLiWfOpXHt7r6VvuwlE8RtX54uwvF9+ALwoa5d1zgYU4V+21JRXMxaNWsEq0rsDF49P1vEyGim8pkbzf2t1v3V9A8TxEqyaYs7JrJ2qOMEVJLIvT87jb1Dia40/AliszKQNKFxVlacrCX4WeMMA8m1czYHFbl520PM0KBrynA7Nq4UFzqlRLBSDoujY1m8MJZByfWVugJnYEzdX8sT4EzghbHMpmvF3S3O1c3K3z1+ttyuD/ArXfvKEiBV2J5QBR9Sokq0FZ6vgmkX5orI2z7SJRcDicXfle0LxAwNW1NRXJgrYTjFaxIzpJRIF11KQCA2PTLkPjjscQRBdDft2EHNbEmn3H79/GwB3/jleTx4dGKRLXvL5UM4fGK65vX+hAmDMwjJYHu17XsDAhXTgI22Zq/WHoMgiOWx2okFQUJ53NSQtz1U55SzskqiqA6wN3ifL/zkJKYLLt5w5dYVCXwc3JHCvq1JHBvPVfZYKsFYg5QSY2kLl21JYP+2RKj3czyBv3nkZXzryQs1r7/qkn78n2++EoPJSMeufTPTTsFCznLxiftGkC464IyBcyXIYLlKnW5nfwzJiL4ivpPVHHeNznXTvsGQCX/TixL+Vjr5iHPWMf9dvbEcsND2euz0DCZzdkVtXuMMYEDJ8XFutogtPRHcfuu+ZSUlbxSWYuc9dnoGmZKDnOUhU/aT+0JCK/+txhi29EQRMzVEdI6i48P1fcQwb8ur70wpiP/OjZfUvaedGl+d9KvWu6atqShOT+XJn02sKvXm13TJRczQkLdFee9ai8agujux5uMBAF6ezqPoLk7/ERIouj6ePT8X6pluNo5v3DsYel4niPVGMB4LjqzrS9K52pOkS/NdDO89fApffOgUciUXAgAH8J/+4Sjecf3OluuYEBJfefQs3HKMKVjPX5zI4sJcEZ96x6GG47UdG2slofgwsZ5Yj77S9TzGWs1TuZKr9nyy8Zwbj+gY6oks2nc0slXqxQkW7kUof4MgiG6gng81anCYGkPRqS+2VI+KIECXu7wyJQcFx4PtClS7QgOhXs6AmKmjL24gZ7nIVHUXr1dc7QqJp8+rgpZGa8LN+4fwP46cQcH20Bsz0Bc34Pqy0l36va/d03LNIIj1Qv0IwyZguCd8YE9jQLrorHq7dIIgmjOUiCCiNbdkIhrDUGJtAvlBRe5c0UHJ8TGRtTCWtjCRtVByfMx1aF5Z2NJta08EiYheMVyW07JzumAja3kQUsLg5VbFTP1XJdVJZC0P04VwBXMbhcBpMVdWRK0mcK7u25qkQMcKcO/hU/jWk6OhjtU5U8mdC78jANN5B46nSjgncxaErA2OVX+Pf/DrVyAZ0TCRtZWTTUiUXB8TWRvJiBY6AYEgNirbQ9rVYY8jCKK7accOCmNLpkse/utPji+yZZ89n8GnH3gJz11I17w+mi6h5PlN1dyDhLaNuGav1h6DIIjlsZL79EYECoVRQ1MJAQumgbBOYE9IfOnnZ/Dvvv4UPvDVX+LffOnxjl4v5wy337pv0R6r6HgYTVuImRzvuWF3petIM87OFPDhrz9dU9hhaAy337oXn37XISrs6CBhFDBdITFdsCvqdDv7YogaHL5QhZe6ptb+qZwNIUXHfSerOe6anesrj56FkBKOX1+DTyX8CXzl0TOrOkd0mkZjeaHtBQB3P3wKvpDYMxBHzNQhAQihFPo0zrC1J4Ib9w62POeRk9P4N196HB/46i/xB998dkXmqLVkKXZe8CyOZywMpyKIGmqMSihxE0NjlYIqxhh64wY4Z8hYXls+rk6Or075VRtd06nJPLKWB1+QP5tYWwbiJiQk/AZbM18CDBK7B+JNx8OWlImC01zXNWf5mMxbTY9pNY4fOz0Tal5f73tqYnMyEDeha427vkpI6ByVhN97D5/Cpx94CZmiC84ZTI2p9bPo4quPnkXE4E3GrQNfShRsd0l+m7A21kqPRYoPE+uF9eorXc9jrNU8lYjoSJhcOQLqsHDODWhkqzx3IY1PP/ASnj2fbroXofwNgiC6gXo+1ILjw6sjWt0MUe54dH0Xd+0QQuKew6dhahyGxtQ1V/1eQonqb+kxAaniDo0I/k5K4OWpfIs14UVMZCwUbA8TWQvnZkvwhMRwKoK5ooPP/fg4XhjLrFufK0FUs2mLOxKRxZW69eAAPAEUbB8vjGVxdCy7shdGEERortyahNPIM17G8SWu3JpcpSuqRSkuZFGwfViuDwAV5UzL9Tsyryxs6RY1NHDOEDU0DKciyNv+shwW6YILIWTFoVBN4GAQQiJdcBu8w8ZkpZyrQkiMXMjg4eNTGLmQ6TpH02qz8H44jo8vPtTG88zKiZ11DmdQ84OU6pjRtNXwe7z58i341DsO4cD2HhRtD5N5G0VbqRl2o9oLQaw2mVK4NSDscQRBdD+3HRyGxoEL6RJKrtfQDqpnS0opIaSsOPl8IVF0/RpbNqJz+ELAFxKeLxExeMXG3dkbhSw37DI4ENE5TI2X/6uuT+MMnpAbcs2u3mPYngBnDLqm7HLbE+S7IIguYKX36Y0IFAo5ByK6Bs8X8KUo74kkwuuCKVxfIG5qKxL4uGn/UM0e62LOQs7ysHdLAh/9tStw/Z7+pn8vpcT3nhnDB/72KZycyldev2Qwjrve80r89qvDFYcQ4alWwARUMkbJ8ZGzVNKa7fswymt/oE7HOceWnig0xuAKtf/mDLBcVcjTSd/Jao67VudyPAFfSswVnboJJXMFB75UY2w154iVYOFYrmd7VSsW9kQNXDoUxyUDCezqj+HSwSQuGYxjJu+0tF3WomhutWnXzlv4LA4kIti7JYGYoYFDjTeNAYmykSylhONJvGJ7D67e0Rvax9Xp8VXPr+oLgXTRwfnZIkyN4QO37G06NzS7pt6Yro6RWJRQ1m3+bPIFb2yu3JpE0fGbHlNyBT7yzy5vGmd4xXBrtW4J4OiFxvNo2HF8495B8oMTG5IDwz3wpSouNbXFviQhVMHVgeEeOI6P//qPJ+EJVSSpcQbVh47B0Bh8oYqV4yavO25NjZeL18WS/TZhbKyVhuLDxHphPflKa2y/8xn4/vodY83mqU+/8xpwzmvmXENTc6jOynOuUHNuQCNbJWJweL6EL9S/iM4b7kW6pTiOIIjNzUIfat72MDpXgtsit7EeUgIHtvW0PnCNCHx+21JR7OyPI25q4Asy0RmU+MhE1kYqamChbhDD4lrAZ87N4a6HFq8JnpAo2F4l/ytY74Nu0QXHg+NJOJ5AX8xc1z5XggjQ1/oC1opHT8+GOo4xNRm4nkDWcjGT777KaILYrPzg+Ymmir0AIKTED56fwDteuXOVrmqembyNrKWcXxKAX2WsMQBMymXPK/VaulXOwRj64gZOTeaX3LKzv6zmJoSE5BKsyqySkJWNcn/cWPJnWK8ETougDVymrAZ3YHvPktq5NWuJvhmDJvXuRzKqI1tyoXFVeNkKp067+wDOVNKT66nM0p19McwVnIbf4037h3Dj3kEcHctituhgIG7i4I5UZXzUe50gNgtPnJ3r6HEEsRHZKGtF9fpsOQIl18d5p4SooSFhaovWz4W2pBCA5wssXMZNvbZ9e8lRwTiNA7YnUHJ8xE3lvrA9WVGkF1IJMnAe/MygcyAV1fEfb7sKA0lzXd/vegR7DCklDJ1X7HPGAKaBfBcE0QWs9D59IcEaM5O3sTUVxfnZIhIRDQXbg+9JoM2ijgDbEwAYhlMRTGTtSsJfo/m03bUu2GM9cz6NU1N59EQM7N+WaFmUkS46+OyPjuPIqZma13/zuh34wC17ETXCCfoQ7REoYB4bzyEZEZjOO7A9v9whRoIxhquGe9AfN2rU6ZIRHTv7Y5jKWbA9UVm/d/bF8MnfOFDj7wjzDDXyndx2cHjFx11wfU+dncOx8Sz6Ykbdc/UnTMwVbBiaSvjrixuIaBy2r7qVmDqH44lVmyNWmmb+EmCxYiEDQ8zUAKixKoRERniYLToNz7Ew0Se4b1GuYTjFQ81R64F27bx66w1nHMO9UYzOleALAduTKDq+KrwuukhGNNzxlgNNv7OF1DuPhITlCHhCIGZwnJzM43vPjIW2v6v9qi+MZSs+dM5VMt49h0+DM9bQJ9psrdWqMgmknBdbCq67W/zZ5Ave+ISNX80W3aZxhl+eCRfPbmbztWOftprXCWI9cmwiB42pQg0hVfFjtS9J44DGGL7xy/P4yqNnkLM9AEogrF77naLj41+/bgeeH80sGrev2zuIL/zk+LL9Nms9Fik+TKwX1ouv9MjJadz10Cm8NJGD4wslgAQAvoS2TsdYo3nq6Fi2Zs5lkIvEGAuOh2/88jz+1Y2XAGhsq1iOgFOOE1ie2lP3xdVevN7+udP5GwRBEO1S7UPd1sMwlbPgSwmNsZb7w3qsVb5jGKp9flGDIWEmYLkCOdtFtuTB9QWEVH6pHX1KuG8yW9txcuEdMXWO0XQJyFiIGRrytgedc0QNXu4GXf135c5XWrmAJGPD9XwwxuAvuNfr0edKEMAmLu7IWuEqnJW4D6tsXueK3VkZTRCbkdF0ERKAwdXCLGUwZtWGXSsv2KPp4ppc31zRhe/LRclrgLpOCQD+8uaVei3dqoloHBkhmwZImzGYjCAVNZApufB8CcYkWPnaVXCMIRU1MJiMLPkzdCthkgo65VwN1Afztof+uAlT43B8UVEfXKjEs1ESRBux8H4YnCFnezg7Xahp5Rc8i0uBsdr1/Xd/5TJcsa2n6T3lnC0y8ikQSxBAOmRHjrDHEcRGY6OsFcH6nLNcxE0dfXEDPUJDzvIQ1Tk++Mb9eM8NeyrrZ7B/jhoaSo4HxxN1u2kBQKbooTfmIRnRkbc9jGdKVe15JUbTJWzvjSEZ0eEJAQalRGzoHL6QkEKt7VFDw2DSRNHxMZA0cesVW1bl3qwmc8X5xDe2QM+GfBcE0R2s9D69moVrjCiLSHhivhBuIWH3UUICOdtFzIy2DHwsda2zPB8DCRN98eadOgKeODOLTz/wEmYL8/euL2bg47ddidftGwz1HkQtQkiMjGbwzLk0JAOu392HQzt76+6Hb791Hz7yzWdwbrYIBtUpi3HA91Vy6mTOxoV0qaJOF+UqeT8Z0SvBvYKjAnufeschXLu7r/L+YZ6her4T2/cxMprByGgGni/R1yDxZbnjrvr6ClUdS7amokhGakMsEY2Dc47fed2lOHJqelFCyU37BvHfHzmzKnPEalHPXxJQrVgYPBPV2L6AwRkG4mbD91/torm1ol07r9F6ExRVTWZVl9qZglO3EDvsvVp4nrztYSpnVwq8RNkv/8ffG0HU0GFoDHu3JPHmq4exeyDe1K8qpMTHv/0cYoaG3piBnogORwiMjGbwkW8+g3/3pstr9hiNrqkaQ+PQOCt3ARTQNV5ZE30hO+LPXq5vtl1fMLE+CeJXGqubGw6NKZtsNF3EO165s2Gc4ZdnwoqlNH4G27VPm83rBLEemS064IxhZ38MM0GR8gJfUrro4nM/eilU92khgemcjS+/74bKuO2LKTv0ByPj8IVUtvIy/TZrORbrxYchlZhhEKfriegbMj5MrC/Wg6/0yMlpfOSbz2C2oLo7VvtqBFAuXmBdl4MRxuatN09Vz7lqTzKfLcOZ+ucJif/2jyewdyiBm/YPVWwVX0jkLFcl8poceduDVxZSBSTGsyWkSw629Ki9eL3981oXxxEEsbkJfKifuG8Eo2kLlqsK1JZQ1wEpsWb5js0I1oeXpwsAANv3EeM6GFOCLjFTw5akRLrkomh7ePdrduOHz40hXXTr7o0DOAO290QwVXDgeAJpOACYEubnym+icQZRfhNZThBlUMXaqnhS7bX1hS1EsD59rgSxaYs7Du7oxT+9ONXyOCFVu2oh1YagL9G9ldEEsdnY2RcHBwAGmFo5sausxhUEcJhUx60Fqbhet7CjGlE+bql0IkDajIM7UnjFjhSeOjtXbl05/zvOgJih4RU7Uji4I7XUj9CVrGYSZrvqgxslQbQRC+9HwfExli7B9gT8cpJntcHPyp6udvZCQQFY9fo+kDTbdpJXB2L74gaEACxXJZbccd8I7qRALLFJuHxLAsfGc6GOI4jNxkZJ2gnW59mCA19IZEoWpFSJURGdo+QKPHh0Au+5YQ+AeVvq5MUcLNdv2XHLlxLjmRKGU1GMpS14ovYPHE/gwlwRu/pj0DmvFGnu7IuBgcETohLwsVwBg4sl27/dTl+irJwoZeU7CCDfBUF0Byu9Tw+om+ju+UiXnIr4RT3a2TtlSx629MimgY+lrnWZUnjlTMcTuPdnp/E/nxqtef2GS/vxH998FQYSG3POX2mOnJzGnfcfw/GLebi+Wnt1znHlcBJ3vOXAou/txr2D2NoTqRTXBEldMVPDUFK1tn/g+XHs3ZLEixM5DKfmO3MxxhA1ONIliQPbUzi0s7fmOlo9QzfuHVzkO8nbLiYyKsE8qAl9ebqA4XJBaDXLGXcLry9qaCjYLizXx+hcCTv7a88XnOvm/UP4wC1766qYfvXRs8haLjTOKjZMkIjUqTmiW6hWLKx+JgBlu6SLLg5s72nqW1zNorm1pF07r9l6k4zo0PqiyBQ9fOhN+/GqPf1LTmiqPo/nSozOlSCkSlqVQKUou+T6GEhE4AmJX7w8g0dPzyBp6khEtBrfZZAEMF2w8V9+cgKuL7BnIF4e16pwxHI9ZATw/37/BTzw/Dg++Ib9NXNSs88eNThMjcMSPgyNK1GocgJvROfQNRbKn90omW25vtnN0olms+F5At9/bhyj6SJ29sXxtmu2Y2dfHEwCfoO/8cudKIP4VaMk7qu2J0NdQ7PjVss+JYhuJRgDpsZx6VC80n0qsMNKjo+i7VWEBMPw7IU0RkYzSJdcnJ8t4t7nJ3B6Ko+85UFIlJPORE1iWaf9NispBNcsPgyofYAAkCmtb/uLWF/Ue+a73VcqhMSd9x/DVM4GA2oLj8tO8/k5Q9ENORjLsXmDOdfQVHGNKuhglRweKQEmJWxPVOze87NFZC0Xc0Ub84m8DI4va/xYnDGUXFHZi2uc1bVhqFCVIIi1JOgi9KkfHsML41lIPxCorl/43wgJYHtvbMWucylUrw+OJ5C1PKRLDnb2xdATrV1rLVfgwPYUHj01jam8g6LTaHesEBIYz1pwfeV01soFgRKA4/nwJaBp88WQ8/Lf8/lfEkr0I2os9iHSvpdYj2za4o7d/eEnP9sT0DWGVFTHUILUBwiiW3jbNdvxn/7hKDJFpRoCzC/dwc+9cQNvu2b7mlzfXD6cQyvscfXoRIC0GZwz3HL5EB49NQMpAZ2rRDoJCV+o+fGWy4c2VKCpncSUThRatKM+mLPcDZEg2ozq+1FwfJyfLVYpd9ciAbAltu6QUgWfOVva+l4diE1G9EpCiXL8SxRsD3fefwzf+9DNG2p8EEQ9LhkMV7QR9jiC2ChspKSdo2NZvDCWQcHxK8WWAAAp4Tk+NC7wwlimrr3iS2Ay1zh5N3DC2a7ARKYEX6rOfNWBmyDAdW62hF19VfdSV8rYgFa+nOXbv93OUCKCVFRH1vLgClm2z1FJrFuqbUMQROdY6X060HiNKZeyY2EJx1K7HtqeD8sRAEPdwMdS17rpvI1syK5uL08X8Gc/PIbTU4XKa4bG8IFb9uLt1+8EZ929hnYrgXJoJcGEM4ABvi9wdCyLj3zzGXzh3dfV+BiOjmUxk3dwyUAcqC6uNNRzrmscp6cK+OAb9+PCXBETWRt9cQMRjcP2BdJFF8mIhttv3VfT6SvMM5SI6DW+k6mcjYs5a1HyXckVuDBbxK6BeKXgYjnjrt71SUhEDR0lx4MvBaZyFhJmQv1uwbnqJZRkSg6Kro9syVXdyDhDRNewpSeChKltOFumWrEwzDNRj82SlNyundd6vfFwYHsPfufGS5a135g/TxYlx4eQErqm3s8pJ6QxKJv9Ys5SnazLg9MTAnHTrPgu3/vaPTh8QnW0KZa74Ji6hkI5wD9fOMLBuYQQEs+PZhf5PZt9dkAF8iUkkhEdMUOrJNCVXB/JiN7ymWvk873l8iF87RfnluWb3SydaDYT9x4+hS8+dAq5kgsBVbDxn/7hKD7w+staG2EMeOvVw03f/4UQgirBcb9+sH4sbDXsU4LoZmrHQAQxU0O1L2k6Z8OTEhyAoTPYXuvd08tTebz/y0/A9STyjgfOgK09UQz1RJB3VIGH60tICOiMddxvs9JCcNXx4XohOgnA9QT+6LvPb4jYJNH9NHrmbzs43NW+0pHRDI5fzINBdaFmVUmoOmeVGDhHufChC3IwlitYFcy5T5+bQ8FWdr6QspLDw5kSiRhMmjg1mcfXHz+Hew+fqnQE1DUAYLBcsaCwA9AYg8YAV0hMZi1EN6gQKUEQ65+b9g/hz95xCL/35V/C0Bjipg5X+Dg7U2rrfS4dWhsx63osXB/64xwRw8F4xsK52SK298bQFzMqPj+dAxezFk5PF+rak/VwyrmeTAJeOfeqeiV0fVkRAFbdy9Xv/bIfytA59DriMLTvJdYr9aWONgFj6fCTpaExxAwdr9jRSwOcILoIXed469XDqj0l5n3k1T+/9eph6PraTHUvXQzndA97XD2CAGkyomEia5fVUyRKro+JrB0qQNoMISQOn5hG3NQQN7VyEFsFmtRrOg6fmIYIa4l1OQuD9lFDA+cMUUPDcEqpUN798CkIISuG67HxLBIRHVt7IkhE9Ipj48jJ6VDnDKM+6AqJ6YId+trWM8H9MDjDeKZUcWoFKUoLqR77YVFtoyWiOl/y+h4EYiO6hrG0Bcv1wRmDrjFonEMIiRfGc/j64+favDqCWH9MhVRcDnscQWwU2kna6XamCzbSJa9S2MFQuzb7QiJd8jCVs2rslYjOUXS85m9efhOlvCKhcaX8V29994XE+bkSkhEdAwkTF3NOx+3fbkcpJ/YiZuiI6hxCSnhCLtu2IQiic6z0Ph1ovMa4QizaE7LyhL2UswkJ5CwX6aKLfVuTi+aWdtc6ISQuZq1QhR1SStz39Chu/9pTNYUdlw0lcNd7X4l3vnLXkgo7Nubq0B5CSNz1kOrIFSSYaJxDYxxGWUV0tuDgrodO1jxPwX49omuImRp6ogZiZV8RMO+/2D0Qx6fecQgHtvegaHuYzNso2irJfGESSNhn6Jlz6YrvJG+7lcKOer4CV0hMZCz4Qix73NW7PgaGLT0RaJwDUqnQFR0/1LmOnJzGH333eQihuh4EF19yPJyfLeL8XHFD2jKBYmGYZ6IeQXLQXNGFXFDREwRn681R64127bzVWG+qz2NoHCXXrwTRfTFfjG1oqiOG7QoIKSvBdMcXABiGUxHMFR187sfH8cJYBomIjt6YATAGx/MxOlfERKZUKRzhjIGXB0hvzFjk92z12QcSBj72a1fgwPYUfCFRclVH4APbUy2fuUY+3xfGsvjcj49jrugsyzcb1he83jvRbBbuPXwKn37gJWSKLjhnMMsK1Zmii7948HjLBBYhgRfGm/sEJrNWqGtpdtxqzRcE0a20GgNa+dlX/w03Dlyhckg8IQCpChKncjaElIibGoLh5Pud99t0Kj7ZjCA+HDM4qqcGBlRU+AEgb3sbIjZJdDfNnvl7D5/Cjr5Y1/pK1V5WQOOsUtgBABLzBdEAgLIy+VrnYLSTK9GIoDistEChPYjp+xJIRgxEdQ2uL/GNx8+h4PjY2ReDxjk8oeL49c4gAaCs5F5yfZg6JxuGIIiu5dDOXly1vQeOLxE1OAyuod3p6pnz6RW5tnZptD4MJCLYMxCHxhmm8zYmc8rnt703AglV3LGUpUwu+P/V+aCcMWxLRRE3tcq67wuJVEz5ggYSBu17iQ3Dpu3c0cpZVo2ajAwa4ATRZQghcW6uhLjBYXmixiDgDIjqGs7NlSCEXJOxW3KbtxRr97hGBAHSQK0iI1Ri/IHtPctWaAmC2NvKyXmWK2qUGS1PbCglsbBJBSOjmY4pcYdVH0wX3E2h6hbcj6zlwnbn1f8CfxeTS1OcXUhPWcEzFVva+h4EYouOWwk816itaEpx/BuPn8N7bthD9gOxoTl+Md/R4whioxAmaSezTpJ2ZvNOTWFHdZw7WJt9IfHcaKbGXik5vuqEyQKFFdS+R9kjp3HAF6qoQyn9ql9rnFUUu4I/Z4xhV38Mf/jmq3DP4dMdt3+7nWr162zJRSKiV9oXer5Ysm1DEERnWcl9OtB4jfH92uB3zUywxPYdMwUHu/pjdeeWdtY6zxeYyFoVlfdmzBUd/MWDL+Gx07M1r7/j+p34t6+/DBFj8d45DMmIjv7E+lb27wRHx7J4aSIHKSV0jdckmDDGoHMOX0i8OJGr8TG00z3h0K5e3Lh3EEfHspgtOhiIm5VOFtWEfYYkU8lztu9jImNXCjsql15+thnUntz2fIxnLMQMbVnjrtH1JSM6dvbHMJm1UHJ9zBQcJMzm56oOgu4ZiKPg+JjKWbA9AcaUzcMYw5++/equsWWEkC2/w7DctH8o1DNRj050/1gPVH/OnOWhP2FCYwy+lCjaHiKGhpv2qXsY3LuVXm8Cbto/hN953aX4/I9fUuqIsrqwg0HjrFLswVngI5PlzrkCYByOJ+F4Ajt7Y4gaGqRUPnzGGXwfcH0Bs0rNWCW3qcKRvjhf5PcM89nff/Petp65Zt2E+mIG5oqOur4Fb9GOb3azdKLZDHiewBcfOgVfSJg6A2dqrVBJzyKU8j8A3H90Atdd0t/w95bb2nYKc9xqzRcE0a00GwN7hxL46mNnVXJ1yD2TxgGNqUJGvVwg7fkS03kHQz0RjM1ZcH0BBqAnqsPQeEf8NqvVKTiIl/bFTVieBQ2qyJkxVDpiOb7AgGFuiNgk0b2EeeZTMaA/YSDXhb5SGZx2wemlrJ1vBhImkhF9zXMwOtFlLigOixocRWe++0awXwdUYVgyqmzhiayF/riJqKFhZz/DVM5CqcquYQAGEgZsT8L2fEgBAEow4XdedwnZMARBdC0L/VmG1v5adHS0OwQCm60PPVEDewYYMiUXH37jfly3pw9/8eCLGM9Y6I0ZyNlLy4k0uPKJVceIAcDUGJJRHQNxA1nbQ7bkIRHR8Be/dQ1uvnwLDu3sxd0Pn8LJizlMewIcwJ7BBP7g16+gNYNYd2za4o7pfPgEGsYY3vvaPTTACaLLCIyHnf1xmBpDpuTB9QUMjaM3psP25Zo6lK7f3Y+/fay1av/1uxs77sOynABpM6qD2IyxmlbFwPpKSgxD2KSCp8+nO1ZoEbYlen/c2DAJos0I7sdTZ+fmDfSqW6yCxIDGAE8A1+xK4cRkAYMJA0IAo5lwSmbpkotUzFjy+h4EWG3Ph8Zrk2GCi9aYckiRU5vY6BTsFqr8bR5HEBuFjZS0k7WqFNYXL3kVr9pUzq6xVzwhVPJlo4RiNh/IMjSGuKHB1Dnmii40rhR7ZTlxTAgJCYltqQhm8g56Yya+/L4bOm7/rgdu2j+E9752D7740CnM5G0IqLasPcuwbQiC6DwrtU8HGq8xDbaLAMInKVWjVFkZfu/1e+vOLWHXup6IjrG0pRJ8W/D4y7P49AMvYq44v/b0xw38xzdfiddeNtj+hwAQMTQMJlSiAKF8H45ftUYvIHjJ9Wt9DGH9F9VdBVrthcM+Q9fv7sO+rUmMjGZge37NhQbPNmeAzhmEkIhGdLz/5stw6xVblzXuml1fMqJD64siU/TwoTftx6v29Dc918IgaDKiI2EmKkIqnpDwfYHeWHfYhkdOTleSH11fwtAY9m1NLisBOMwz0YjNkpS88HPaQkAIAQHA8QT++yNn8NVHz9Z8Fyu53lRz8/4hfOXIy9A1Do0zeL7ERLZUUe4OhJfK+e2V4gydc1iOgOv7YOViFQCIGhwRnaPkCjBWG6yXZeXFmKGS26REXb9nq8/e7jPXLFnBLxdgub4PyxFlP/k8YX2z7c6lRPfy/efGkSu55W4ztUYYZxwMfpVIAQBZyTWtSSAvtOh2efWOFL7z9GjL67k6xDOzWvMFQXQrjcbAyGgGf/f4efi+gMYb18VrTCnOA4CpcWgcFZuagUHjKmakM46BhInJnFIqThddcNYZv00nEq/DEMRLDU2twzpntedjElKoz04dp4iVJMwzP5N3cNvVw7jv6dGu85Vev7tPCSj4AlyTlc9QI2YEoDdqIB6ZT99bq7h/JwSrgu9se28M45kSSq6AxlR3s2B+tVwP03mGnX0xXMxYlfMF++R00cV4tqSKyYRKHN4e1WE5av/sl1Xab96/ZQXuAkEQROeo9vMcG8+23cUibnaHP7vV+hDVNWSZh0u3JMAZw+mpAvrjJhxv6WLXjDGYnCmdQAkIKeALYNdAHHnLq/gGr93dW+MbvGn/EISU+OyPjuP8bBFCSkxmLdxz+DQ4YxvGh0hsDjZtcUc7wbyYzvC1X5zDwR29NMAJoouoLTxQ41rXlMIgYwwRja1psvtvXrsDn/zuSFPFpKjB8ZvX7ljFq2qPjZSUGIawn5dJVJ49CVlxJOicI2rythwuYdUHe6LGkr8LKeUih1e3EtyPD3/9KRSCdq1y/j8MgK5xSKmUPgxNg6lxmOXWulN5laTSDAZgS08EvpBLXt8P7khhayqC6bwNjVfCcuo6ywHoqK42NuTUJjY6AyHVj8MeRxAbhY2UtMMYq6jyVZSyy1QnqyRNvcZeUXZ5WdkP8+q+svoPy1yxtQe6xnBiMg8hJHRee798KRHROVIRHdNFF7NFZ1kJguuZIyen8bVfnIPOGXb0xSrfTcn1yXdBEF3GSs1TjdYYQ1Ot3UXVHmqprQ91DgwmI7BdH+mig5ELmUXJf2HWuiu2JdEXN1oWdjiewP93+PSi5MXXXjaAj9925ZJsSUPj6C+rXxLzDMTNii+tXoFHtRp/tY9hJbonNHuGhBSYztvY2RcDAHzglr342LeerSTDVhcsMagEcs4AnwEmZ7j1iq1NlUTDJLa2fsY9HNjeg9+58ZKWn7teELRaSEUIicm8vSwfQqc6bRw5OY1P3DeCvO2hv/y8OL7AsfEcPnHfCD71jkNrYmtslqTk6s/5yMlpfOXRM3B90fS7CLPeLPf5OLgjhf3besrjIQIwJZ5iuT7ApfI/AuUE9trijLztVTp16FyNAcYYtvREMTpXqnQJFEKAcQ5PSGjl3zPGYHn+qvigq8fpQp+vxhk41Bqr1rRa/2xYP/lm6USzGRhNFyEA6Ew985X9clnVnlclgQf76ErDqao1rJX416suHaix7+rBmTqOIIjW1FszD+3sxZXDSRwdy1aKEOsR/IoB2JaKVvxeQYQoSELO2S7mCq7qSstUPMrQeEf8Ns3Wqnbjk80I4qWqu1xVcVrVvQj2ExspTkx0H2Ge+SnHxw9HxrvSV1o9v7jl6y73uKscEzH4osLhtcrB6ERuSPCdRXQNW1MxZe9LiWAnLCHhCyCic/zLG/bgrp+eXHS+iM5haByOKyoF4wFSSuRsD1fv6F0XsRWCIIjAzzMymsHvf+WXmMzZof/2+j3LF4vuBK3XBx+QwMvTBbyMAhxPoD+uioQ1DrRI4QKgfNGuP78+SkgIWe4cV145dY3hD99yFbYmow39W0dOTuOPvvs88raHgUT3+DUJYils2sjW26/bgSOnZloel4rq2NEX71gLS4IgOkdgPKRLDtJFB7YnIMpBoojO0Rs319ShxDnD9t4oXp4uNjxme2+0I3PKQjU9ANiaiuBf3rAH77lhT9NzeJ7A958bx2i6iJ19cbztmu3Qy0npGykpMQxhP+91e/rKz56LTMlVLUDLjsyIrqE3ZrT17IVRHxRChv4uhJAouT6Kjo+i42FrT3SRU6ibuWn/EP7Dr12B//t7R1UVdtXvJFTST8CT5Q4fmZKLLT0mBhMmxrONu3dwBsRNHVt6IoBE6PW9OgDeG9FxarqA7b1RvDCWheNJmLqsqI0EAejeuAFIkFOb2PD85nU78Ojp2VDHEcRmojppZyxdgq7xrmoJ3w7X7+6DwTlcXyxamwEV3DU4x5sPDeOFiWzFXokaHKbGUHJFTWLLQlRBpMSFuRIs14cAYHsCnANCzJ/P8SXOzJYQN/mmXV+FkLj74VPI2x6GU9Eam7BXSvJdEMQmYXFiqA4hAMv1K12PgCXXdQBQc3u66ML1Bf7q4dP48pGzuHK4Bx98w7wKVqsE1ZjB8a5X7mx5rpenC/jTHxzDy9OFymuGxvCBW/bhHdfvaFusgJfVO3tjxroROlhNDu5I4crhHvziZQeeL2Do890opZRwfbVub+uJQEgJIWRlTVlK94RmCeWNnqF0ya0oHV+YLeH2v30S+7Ym8RuHtuNrj52r2CTAfLK4xhl8KSClxJ7BRENfVauOFAuv9wO37MUffff5ZSdhNwqCSilhuQIFx4OUEn0xo52vM/TnCksjWyPKNQyn+JrbGpuluJdzhoM7UvjMgy/C9cWyv4swz0er4o9643UwaWJ0rgTHk9A1BoMxuL6AzyQ0xivFGRxqL8AZULBdSCkRMzUkIzp29scwUVb09QXAmCoKGUpGoDGGbNn/evXO1KJx3ekOM/PxhsU+X1Pj4JzB95Xfr5p2/eSbpRPNRmdnX1w920Kq9bLK8OJscdPLenZZIqJVxL8axWkObk8hZmoo2I0VT2OmhoPbWz97R05O466HTuGliRwcX8DU+CL7jiA2I5wz3PGWA/jIN5/BbMEBGOoWeQSRqcGEgZ6oAVkWIim5AgYHfCEgBDCTdyqFi4mIjqGeCBgYUlJgNG3hz354DJ96xyEc2tnbtj1VvVap2LhfFRvX0Neh2HgQL31hLAtT47A89RkZY+XE7OCz+ziwffEaTRCdopl9FtE1pKI6So4PBmBXf6zrfKUL5xe/ymAIrmahj3c5ORidKOheTm6I5wk8ey6NkuthOg8MJU3s7I9hKmepXJ7y59c4wztfuQv/4tW78eDRicr5Co5fc6wAwCSQLtooZQRsz4cv1H3NlBw8dnqGbBiCINYFnDNcu7sP73zlTvzVw6dD/U3C1LpGLLrZ+pCzXIymS2Bg+Mt/PAHGGLKWh4jhqC5wIQo7dM4wlIjgYtkfDKCc+zi/bmpc5XFvTUabCvp0s1+TINpl0xZ3vPP6XfjD/znSMsi5ZyDe0RaWBEF0joM7UhhMmnh+NFszloUEPEeg6Fh1gz4rTbBpfvLcHPKWj8GEgbmCi2p7hQPoTxiwXbHseaVaTS+iayg6yrExnbfxf/+vo/j7J87hjrccqLuxvffwKXzxoVPIldxKi9L/9A9H8aE37MPv37Jv0ymJhf28h3b2YjBp4uhYttJJgpXbL5ccD0XHw8Ed7T17rdQHq69tPGMhZmrQGIMvJUqOj2REw+/ceAku5ixYrqgk86xX/tVrL8F/f+RlnJlpXBwFAKam7oEngMmcA47GLbODhI8tZUc6GEKt79WB4ozlouj4WHh7VfKLqhoPAtB5299QxU8E0YirtoV7xsMeRxAbiZv2D+FX9g3i7395AV5V4EbnDG+5enjdBB4O7ezFzv7GRcsSwGDSRLrk4raDwzg3U8BE1oapM/iiVmG03jpddAWOjucwlDCwuz+G83MlCDmv5KLsLbXOWq4PCYlMaXN2xgra2vfHzUUJy+S7IIjNRZAYeuf9x3D8Yh5uedLUGIPGAS9E0KQZroCqsIPa51quh1+87ODEZA5fePd1NW3O6yWo7tuSwLtfvaupupiUEvc9PYZ7Dp+qUeXaO5TAJ996AJcNJdq+7p6ogYGECW2D+ClWijdfPYznxzLIFF04roCuqT2y58mK/+pC2qoUVSxsbR+2e0KYxOuFz9CU4yNve+AMGE5F0BczK+pq52YKuGJbAmdmiiiVu31qnIGXk+lcXyJicPzBr1/R8HqadaR472v34PCJ6UXXW/36UpKwRTnxdyBh4sJcCTv7ouBcdTMIElc8odRo/+LBl9pOsu1kpw2yNbqHTn0XYZ4PAKGKJBaOV1dI9MYM+FJCY2rtcYUEZ8BQj4m4oWGmYONiWYjFFRLjWRsMNiIGx/beGBKmhripY2efgazlwfEEYqaGqZwN2xNK5ZcxZEpuTfLWSnSYCeIN9Xy+lquSZwOxKdsXNb7Znqjelp+8nbm0U115uomS40PjDKbOWx/cpbztmu34v743gnydootgD6y3sMnefu0O6DpvGqe5ce8QorqGkuPX7d7BGRDVNRybyLWcC4LEUillJTG2nn1HEJuRm/YP4Qvvvg5f/OlJPHluDr7bOMZWdFQMrroLle3NFx+LqsFqaKqfrbL7bFiuh2PjLt7/5SdwYHuq7aK+5rFxH0Wn1JHYeHVM0vUlOBNwykXgQqj5Q+NAMtLe+kcQ7dLMPis5Hgq2B11jGOrp3v1LML/c9dBJvDiRq9jbw6kIpvIO8rYPXePLzsHoROFzq3yEZjZvtT3jSyBv+5jK29jaE8WlgwnMFhzMFpQdrTPge0+P4vnRDG65fAjnZ4s4N1tCyfUgyvY/oLofAcB0wS3/zBA3OfriBsYzNimwEwSxrhBCLsozasaNewcqwsxrTTOBnvFMCQCwvVf5cW3fR7rkYCxtgTGgVZiCQdmUMwWnacdKIYEdfbGK4HE9Pwn5NYmNxqYt7hBChlKvCzbmnWphSRBEZ8nbXsOxLMu/X02qN80F20fOdhEzNOweiMEXKgHc0Dh6YzoAhsm8vax5pbrqNBnRMZa2IKSExjk4E/B8iRfGs/joN5/BZ3/7Wtx8+ZbK3957+BQ+/cBL8IVSVtPLrb0zRReffuAlAMDv37Jv0ymJhe2iUaHSS73q5yXWVbRSH7xp/xDe+9o9+OJDpzCeLinFOwDJqI63XL0L+7YmK8kNG4FWAquMAYwzcMkqSUdB8Auo/Ro0BsTKHTuSkXnzp9X6Xh0olhKLFNJ4edwICSQiHFuSUTAGZErehit+IohGTBfCtQ4NexxBbCTuPXwKf//LC8re4vNmgi8k/v6XF3DpUAK/f8u+tb7MUFSvn/WYytn4+DefhalzDCZNcO7h3GxJqX2Xl0Ipm5tJ0wUXuqZhd38MZ2dLldcjOq/cN50zRHSOew6fxk37hjbdOhu0tTe1+g5d8l0QxOYjW1IdMgYTJqKGBsZk0w6i7cIApbAF1XlqKmfjzvuP4Xsfurmmm0OQoDpTsAEJ7BqIVYLx9ZgtOPjMgy/h8ZdrO8C965U78fuv39t2smfM1DCQMBHR10/XyrWg2m8lpSqe9HxZKa6RUElaw6kY+mJGw0TpMN0T2km8Dp6hkdEMPnnfCC7MFbGzLwbO1XNQra6WijEMJiNIFx04noTr+/A95euPGBwf+7UravxfAa2U287NlvC5Hx9HMqKjL2ZAlJO5Ry5kcG6mgD97xyH0xsy2k6trfIXlwpXjk3mkojpylldRbtUZw1Ayghcn2ktM77QiHdka3UMnvoswz8ed9x9DtuSi4PihiiTqFSUcGO7BsYkcZosOzs8W8cDz4zg9VcCFdAk5y4WQtS5TCcByBc7NFpEwdQwkDPy/v3k1AODO+4/hhfEchJDQOBA3lAp5dfLWjXsHO67EKITEyGgGOUv5AOv5fBmAgYQJyxUYT5cqfsiemIH3vnZP237ysHNpJ7uTrCWW66NgeyjYPjwhsC0VXRfFHY2SRjhnSEb0usUdAYwx6EzCq7MRNjjDubkS7nn4FP7iwcZxmt96VQGcMeweiGM6Zy9S6R/qiaDo+C3ngjvvP4apnD2fGMvUHr2RfUcQmxUpJWx3PgWNQRUTq+JxCceTsDwfY+kSBpIRxA0NUYOhUWhurugCKKJg+zUJywZnGLmQwUf+/hn8u392Od5zw57Q42+1YuPV8dJnz88hZ/mV+YwD4Jwvaf0jiCXTICYvJbp+/9KosPex0zMtcxLCCit0qvB5YT5CGJt3Yd4Jl6qwW0hgImspO9Dx4PkSGmfY0ReDqXEcG8/h/GwR77lhN+56+DR8IdXXXI7rJyM6JsuF4gql4m7qGvrjGimwEwSxbjhychp33n8Mz49mQ//NY6dn4Xmiawo8FuXS+RJZy4XGGXb2xdATVd2IY1zHjt4ozs6WWhazaEx1ek2X3KbHMcybAUdOTeOew6fr+klcIcmvSWwoNm1xx3/96YlQx13MlrC9LwHbFx1pYUkQROcYGc1gdM6qJFcvhDNgdM7CyGgG1+7uW/HrWbhpjhoaCo4Hy/UxnrGxsz+G/sT8HFJy/WXPK0HVaV/cwETGhpBqwyyEUoQSACCAizkbH/67p/GX//J63Hz5FniewBcfOgVfSJi6MoNkOSBg6oDjSXzxoVN4302XQdd5W0piG4FWn/foWBYzeQfbe6PlFrCionQVMzSkYgZm8k7Hq30fOTGFrz52FhzA1p4IUA7A2J7A958bw+Xbkk1VUdcTX3/8HM626Nohy0UVgUJtwEDSRF/MhKkD52Yt2J6Pbako+uKG6thRRbP1vToAvq0nghcv5gEsCEZLQGeAJ4G85cPUHER0bcMWPxFEPX5y7GLo4371FcMrfDUE0T00s7c0LhfZW93M0bEsxtIlaJzVtI6vRkiJVMyAxhnG0hYKjoeowTGUMGFoGqIGx/GLOTh+c0/eZM7CpYPxmnMF6r9RQ8OWngg0znBqMo+R0Qw4Y5vCPg0YiJswNAbHF4jyxQnM5LsgiI1FsyC+5wl85sGXMFd0sSUZQSyigYGhaHs1PhJTY0rFtZnsVROC5uecMRg6h+sJHL+YX+Rr4eUEhIs5E7bbXHTgsdMz+MwDL9UEbfrjBv7wzVfhhssG2ro+Q1NFhXFz07q5Q7PQb9UfV0pq0zkHrLzOlhwfO/uj4GxxUUU7CRNLKTgI/juRtRA3ddieRNSQlb8N1NVm8g4++Mb9ePDoBE5N5lFyeSXp9Q9+vX5hB9C8CwKgfAuOJ2DEOCayVo2vp+B4+PQDL7ad9FrvnqdLLi5mS5gpuOVkQSBqzItRSCnbut+dVqRbS1tjI3YnWA7L/S6EkPjeM2N4YSyLuKlhgUus/HzoOH4xj5jBsas/HmqsNvqeqp+v99ywByOjGdzxnedw/KIHxiQMziGgEsmDJckv2/l/+varK8lrvTETPREdvTFd7SNMDgZWMzYSEb2jz31QQPHieA4zBbtSnO4JAQZW8flGdI6pvIOeiI4dfbFKcnzJ9fG1X5zDwR29HfUFrkR3ktXG8QQKtoe87S3y4a4HmhXXJCI65opuU60nt8n+1xUST5+dwTPn0k3jND94bhwJk8PUOC7bkoDlCHhCQOccUZPDcgUMLprOyyOjGRy/mFdFuzqv+MgZQ1P7jiA2Io3WsWDOnc47i7TcfCHBOYPGOHRNwBcSQz0R5C0Pc56PnKX2P2oYM3jl5OTgfeaKLjgA0+BK+EQKpEsePCGQsST+5B9ewAPPT4Tq3lYdGw+S5aprElmHY+M37R+CkBIf//ZzABhiploPV3L9I4hqWsXko4aGuYKDnOWhr85a2E2+0nqFvc1yEsIW+Xa64P/IyWl87RfnoJeLMFrZvHXjIAAMDfB8JcQQ+H/iBse23lhFSCq4vh8+fxExQ0NPRIfGGXTO4UuBC3OlGsV3zhksT2J0roSd/TFSYCcIYl1w5OQ07vjOcxhNl9rSCS44Pr737Bje9apdK3Zt7VK9bj11dg7/7acn0BczEKvyzUsp4bWIBQeEPEx1eO6NYixt4ePffg6uL+r6SX7v9XsphkpsKDZt1OsfnhsPdVym5GO4VyJddHFge8+yW1gSBNE5nj6fhidEJdhSnafAmfrnCYGnz6dX3CFdb9MspUTU4Ci5Pnyh1I8S5WQLKTszrwQKbkIAtqdamYtyh5CFXQuyJRcf//Zz+NxvX4uLWRu5kguNqwCDlLLi/Ata6eZKLr7/3Dje8cqdAMIpiW0kmn3e4L5v7YmgP67U2ioBlbJzdildWeo5lX0pUbR95G0X//knJ5CzPAwlzZoiBQmJ6byDrz9+Htfu7muqjroeEELi735xtmnLvQDXE4s2QJbrw4sIwOPY2mPiwpyFvO2hL2bUBLJbjcPqBIlMlZpm9fkk1IZD52oOeterduNt1+zY9AkIxObixMVcR48jiI3C958bb9ve6lamCzaylgcpJQwNgFQBmupEYSEBXwgkIib6YgbmiqrgsSemiitn8nbLwo7gfbIlDwxARGfwBDCUNNETMSqJXUJITDk+PnnfCGYLzrpXsG2HgztS2Lc1iWPjOQyneE0yW6f2GARBdAfNgvgA8JkHX8LIhTQAFWCP6KoAzqoqrFBrDoMQy0umDBTMGZRiresLPHOu1tdiez4uZmx4Tc5luz7uOXwa331mrOb1G/cO4D/edmXdRIxG8HLybm/MqJuoT9TSKNkjxnXs6tdwYa6E2YKD3QOxSmFHQHWi9PeeGcNA0myZeL+UgoMjJ6fxZz88htmCA84YGLMR0Tm29EQriR+ButrugTi+/L4b2ioEaNYFwXIFPF8AEpjK25AAdK4SuoOuNS+M5/D1x8/hX914ybLuueoww/DyTBGGxrC7P46YqVUl2raXmN7pThtrZWtspO4EnWI530VwP4+OZjBXcpEpqcTShR1tAz/yYCLcWA37PXHOwBnDZE518NS5un4NAC8nhAopIcrKvr0xNf8fHcvi9FQeW3oiiBpaw+t5+ny6Y899dQGFqXFVdKUFheYMPREdiYiOVFTD2XJnwt6YjlRsfs3qbbMoKwydTtJbTTxfVLqaO976K+gIqH42YoYGXVOFF8fGs/jEfSO49Yot5U7tar0QZeVwFsSmfNkyUSXvSHDmNfUbFB0P2/uSmC24GE5FEDM1AGp8hJ2XnzmXhusLtbYtqPRqZt8RxEaj0Tr2gVv24p7Dp5G3PcQNjnzQgJoBTM7bg1zn4AzwAbztmh14w5Vb8f3nxvDXPzsNLZgH6sSRAIBz9aLrq8R0Bz44Y9CYeu/nRzOhCveC2LiuqbVWVs09ai6S8PzOxcaFkLjn8Gm4vsCegXiNvbAS6x9BLKRVTF4IiUzJRbbkLvIPrBdfaXVOQpAr8MjJaXzl0TMNk1er54pOFvw3skGBxmO+URxkIQwA4/Wv79xMARLAjt4YOGeQkDgzbS0SmmJgMLgqkp3KWdjTHycFdoIguppgXs206EzRiGfOpbuquAOYX7dmiw4YWE0n7bztYSpnodiorV0bVBdLM8Zg6hyTORtRnWNrT7RiV0cNjuFUBBNZGw88P469W5J4cYJiqMTGYNMWd2RK4Yw7TygDNRnRcPut+2hTShBdBCs7rOq1tBZlRX9WPm6lqbdpZoxhS08Uo3Ml+FLAcj0UbR+cM6SLbkfmlUDBzXL9crKFhCdkjaoMoFrjSl+gaHu4++FTuOHS/kqwITguOLb63o2m5zsnkHrePDXKeYZWE1ABAMtrvytLtVPZ8QQ0zrBnMI5/8ZrduH5PP45P5HFupoBUdHH3CQaGnqiB8zMFnLxYwBXDyU591DXh6FgWF+aad+0IqDe887aPklMC5yoZQxV7MUxkLfTFTUQ0DtsXLcdhdYJE1mq82VLJrWrMpKL6piqCIggASJfCtXkPexxBbBRG08XF9lbZE9XI3upW0gUXfllly/NVYWk9AiFWXyqVbdf3YTkCUYO3FWARUlaSYzgDeiJG2d4qX0/JRd72cGGuiC090XWpYLtUOGe4/dZ9+MR9I5jI2uiLG6FtG4Ig1g/NlLo/8s1nAAAlxwcYqwhelBwP5+d8xKqCKRLl5Nll+kVqcgPK/19WvVZyfFzMWhBN+qyfmsrjz35wDGeqOjSaOsftt+7D/3bt9rYKNJIRHQMJE3qDpF5iMa2SPWKmhqzlolFtjusLTOVt/Mk/HIXGecvE+3YLDoJnPl10K4ItjDGUXFFR5ExG9Bp1tUaiHNX+q76YAUDZDrN5BzpHXeU2Twj4UkIAYFLC0OaDfwyArjE4nsTfPX4O77lhT6h1ttk9VzaOKlhlWJxo205ieqc7bayFrbERuhOsBEv9LqrvZyKiI2d5AFNiKNXjCUClIHBhIUWAyRmmXR8PH5/EsxfSuPdnp1FY8D29MJbFx771LH7ndZfg5v1bKj7j2aIDt+y4r11G2HwCqpBw/flnPezcwSQ68twvTF6zXKGSyaQqfJGQmCu5yNkuZgpaWVwJMLTGhSedUg3udFeelcYXEgXHQ97yagpN1yvBszFXdOD5Kmk0SJ42NQ7XF/jZyelKASxnHIunxXD3IfAPAI3jNNfu6sMTZ2YxkbUQM7QaBe1kRG85L1fstkaH1LHvCGKj0cze+Pi3n4Pj+RhIRNQ+q5qgwEPOd5NnAHb1x3FoVy8ePj5ZE6tuNIyEVHaoLP+9EICo8rCVXA+yKFsWSlRi3lKtB6x64gAqjshOxcaD9ShmaMjbXqVrkFrPu289IjYerWLyjhBIRXWYurZufaXVBR0PHp3AxUwJ0wUHvpCqm0VUImqwhkW+nSz4X4oNWi8OUg/GlLDCwj1JRFMd/jQ2b99bjqgIm0q/tviVMQadA7YnkLU9UmAnCKKrCebVuKkja7WfKxGL1PfXdAPVa3SEqTjwZNZuGiNoh2DuD2IcubIgr+tLnJsrVvbogThQX9zA6akCPvjG/bgwV1y3dgFBVLNpizv6Yzrmiq0nTQbgwPaeTa0QRRDdyjW7wzmJwh63HBptmpMRHTv7Y5jMWii5PmYKDhKm1rF5JVBwGxnNAJAQktVsnCXKAXGogGAqpjbcr7qkf1H3gYVIANt7YwBIPW8hnVYx/PmJKdxx3wjylodUzEDc1OD6Eicn8/j8j4/jo792BTwh4QqJlFbfyDQ1hpyUyFjzjhkhJU5eLODkVB7bU7F1U5AzW3TgLTMG6EsJ4SuFEADQmA/D4JgrOOCcweCs5Tis3owYIZKGqscMQWwmtiQMnJoKdxxBbCa298aqFEWqflG2zaRcP2vHeKa2/XkjguVS5xwcKnDtCYHZogfbDa/YGjM4bI+j6PiImxqi5vw6LKTAZM4CZ8DOvhg4V7+rDm598acnMJ6xMJ4pYWdfHG+7Zjt0feMkAN+0fwifesehin2eETKUbUMQxPqgmVL3th6G45N5AKqrUc7yatMGfYmc39mC2mDNUj0TJXxfQucc15dVYLOWi5m8A9kgaCOkxHeeGsW9Pztd2Z8BwL4tCXzyrQdw6WAi9LWYOsdQcrGaO9GaVskesfI9tVwfiUhtyCBvexidK0FIWVavN1om3rdTcFD9zO/sj+LsjITl+tA5YHBWUeSMG/GW/pZHTkzhsz86jvOzRdieD8dTBaNRQ0PC1FDyBApZG3sGYjW+HK2sehwo9i8uNmLQuMRk1g6duNbsni+0laqTkxben1YsxUfVSMAleN0VEr/3+r144PkJnJ5aWVtjPXcnWA3atfsW3k8AmCs6KLkCOlcJ+FM5W40Hx0fGcqFxBrbAOyylSv6azjnwfIF7D59CyZWQkNjZF6vMw56rxutc0cHnf3wcX330bMVnPBA3YegMcOYD8rXnmE+UD571sHPHdXv6OuKbXZi8FjU5dK4Ky6phjMH2fAiprrd6fxLQLGluKaJJne7KsxIEBR1F20fJ9RvaAuuRo2NZvDCWRcFWlpbGGRhXz63tCdgeADjQOIPvS3BNLnoOl1Jc2yhOc+PeQVyxLYkv/vQURgulyvhJRQ2897V7Ws7L1+/ug845fF/UvdaF9p3nCXz/uXGMposbck9NbD5a2RvnZ4souj629URhxnWMZ1TH9mCsAeWkMqEE9nrjBt52zXYAtf6/4Lh6BKrCticWFXQFAmIF28cLY9mKvVlvLF63pw+GxlUnES5rioQlJHyhCpWv29PXkXv3yMkp1dmuXM2mEui0SjewbliPiI1NmP3OK3b0VjrwrCdfqRASX3/8HP7u8XO4MFtCzlZCg4bGIaSyP2yvthiiXoFFJwv+27VBPU9gNF0KNQ8y1O7xE2ZC2dm+QEzn2NYbw4W5EoZTHJ5QxXC6xuCVfUmMze8pAl9VtuTh2t29pMBOEERXEfgApgs2Hj01g7zloSemgzO0vU/8tVdsXZmL7ADBGv3s+Qx8IVB0/EVC1MsleD8hgdmCUymYVsI86veBOND23ijccsdniqESG4VNW9yxsz+O0zNWy+MEgNfvH6SBTRBdCGcMusZqkgMWErSlXWmabZqTER28N4pMycW/e+PleOUl/R1Lsg8U3O64bwQF26soKgPzQTOdc/hCImpo6InqmMo7SEX1mhZm9WAA9m5JkHpeHTqhYiiERNH1kS+5+MJPTiBneRhKmhVHbERnGEqamM47+Prj5/H+my9TDg9fIqIvfl/HlzAYQ29UOWYeOz2DL/38DEbninB8pexx3Z6+dWGsDsTNhorgYQie7ep3EJDgUMlAC5UEG1HtMBSN5EsXnHfvlvCJSQSxUdjWGwOQDnkcQWwe9m5JzK9JsrbAo7oYt9vXDiEkfvTCxdr2t6hvR07mHHDOkTA16BqH7fmwPR9TOaetlV0C0DiHxgV0jcFyRcXWms7bEBIYTkUqhR0BjDFIKXDk1Cx+fmq2cq3/1/eex//xpn34t7fub/vzdys37R/CjXsHqbMeQWxAgkTTvpgByxXwhFDqqAaH7cmyWqxKzg0zt7arlBUoQQdIqIBJ9e9fMZzEoZ29mC04SDdJ4pnJ2/jMgy/hiTNzNa//9qt24f03XwYzZJIgZwz9CRO9MSoWXiqtkj0YU/6joutjQM4nfUpITGYt+EIibmrqO2CA9ICEyZEuubjrocWJ9+0UHFQnV3PGsaUngtG5EjxfQuMMGi+re6Yt9MWNhv6Wew+fwud+fByOJ1SntPLrHMqeSUZ0SClRdHycmy1hayoy78spuZWkkUUJ6OUkuUi5K07YxLVm9zxq8IqtpC3wW7YrGlLtG7yQLiFuaIgaGjgH0kVvkY+qkYDLLZcP4fCJ6ZrX925J4INv3I/dA/EVszXWW3eCtaAdu6/e/Qw6S3tCgjOg6Hg4MZmHW57bGQPOzBaxtSeKwYSJguNjPFOCVS5wYABsT8Iv+8bG0hZ29gO2KzCZU8meGmeV/wY+4z99+9W4argHv3h5Fp4QMDivmVtUUirDlcPzz3rYuePQzt6OdJhplbw2f3JUNkAy+J8Fb90oaW6pokmd7srTKYKCjoLtwXLFhiroqGYmbyNruZBSwtB5xWfPGMA0wPUEio6HnX1RnJ8rwS3ba/NJ2mIZ3u1aGICc7eKvH3kZedutSWrMl18/uKO36fN0aGcvrhxOlgv46l/rlWX77t7Dp/CX/3QSWcurPOr/z/96Hh9+0378/i37OvSpCGJ1aWVvpGIG8raHnO2hL25iS08UE1mVR1I9zXlCrXUfesO+SsFTtf+vGTGDYyhp4vycel9W/h9Z7lpraAyuJ8rF83bDsfihN+7DFdvK49kT0IPENgl4vhrPV2xT43m5HDk5ja88eha+COxyVRBd3Q1MK4upkWo+sVKEjcnftH8IN+0bWje+0iMnp3Hn/cfwwngOflWmr85V50whAa6xyj51KmcjEdHAwBYVWHRSlLIdGzSYpzIhlehFeTLTy0Urlqs6fgfX94Fb9uKPvvs8JrI2YgYHIOELNUf61bY4pBJ5lBIJUmAnCKLLCHwAL4xlkC13mhASyNlLE2Q6N1PCay7t7DV2Cs4Zbrl8CI+emoEvajO8VsJTECyXqutzeY8OwOCAKyQmczb6YjoG4iYO7eqlGCqxIdi0MhsxI3xdy5/f/xLuPXxqBa+GIIilkC65iwKgC9EYQ7rkrvi1BJvmuaK7KKAhpUSm5OHA9hT+9esuwaFdvR01GG7aP4Q733EIVw331ASVOFPFLUJKcMawpSeiCgA4m2/V2wTGgLmCU6Nmo4LDDFFDw3Aqgrzt4+6HT0EsQYZKCImRCxk8fHwKIxcyDd8j7HGrTaCcd2B7D4q2h8m8jaLt4cD2noYFL54vkCm5mMhYODtbxGTWwjPnMzg3U0AqatQo7AAAA0NP1MD5mQIAYPdgQgWVFqrpQSJbctAbN3H4xCR+9388gU/c9zxOTOZRdEW568d8Qc6Rk9Mrd2M6QKbkwA1RTNGI6rvDoFTEXV8gonOUHB8/PzkdynAPHIa+EJgttp5HGMOqzDcE0W0MJiMdPY4gNgpZy0O8Sllcyvl/AXFTW1Ib3tXk6FgWk1kLUUODxhoXdgCA46ngbt72YGgcps4xnXfatt9Kjo9rd/fiD998Ja7Z1Vdja+3siyFhaIjoGnKWi5IzrxA7lbMxU3AXORDztodP3f8SPvmd55ZyC7oWzhkO7erFrVds6fgegyCItWO26KBg+xjPWDg7W8CFuRLOzhZwZqaAnO2W25CHV/ny29xatcrRZADeds12TOftpoUdR05N4/e+8mRNYcdAwsSn33UIt79hX+jCjp6ogd0DcSrsaIN6fpQwfqsrh5PojRmYyNoouT6EkMgUXZRcH7rGsDUVRcHxcWa6iLOzBYymLeQtF0+cmcXXHz9X857BfjoZ0TCRtVEsJwJP521cmCshYfJKAsbC5OqgC27U0CDKauJCSuzqjzX0tzxyYgqf+/Fx2K6AFkggB58PqsgpU3Kxqz+GREQVPhQst2JfvGJHCv/6xkvAuUqcEVJ1KBBSwvOVX62vnOQSNnGt2T0HULGVMpZbud8l18dE1g6dmF5NKqqjaPuYyFo4M1PA+dkStvdGau5ZIOBybDyLRETH1p4IEhEdz11I49MPvIRnz6drXn9xIo+//tlpGGWbYyVsjTDKsC6pQYe2++rdz2BMxQxeWT8cT6kcmhoHJIMvgPGMhRcncjg/W6wUdgCqm43jC/hSBc89IXF2pojxjAVPSPhSJfxLqZI+A5/xPYdP43+/dR8GEiakVL45Xwr4QsD1VPLpYMLEB98w/6wvnDuajY16vtmCpcb5P3/NbvREDXieaOpXrk5eAwDLUX5UQ2PgNfOIRDyiIaIzuL5AyaltORwkze3bmqxJmms05sL4aFvN2/XOt1L4Qs2hY+kSzs4UMJ2za/ZhG5G5ogshZOXZFFKWk3LUZw66Hf3GNTuwpScCzhj8sqK/L8rrRjRcHLrVzMoY8KVHXsZUzoYQ6tyaxsrXoPbBd95/rOm+m3OGO95yoOG1bumJ4I63HMDfPHIad/7wRWTKyeSAWkczloc7f/gixcqJdUsre6MnqoNzhkzJg5QSW3oiGE5FlV1ZRW/MwB+++cqaQqd0yW0ZbwXUWJ5d6LMqd47TNVVEFswt33n6QsOx+Of3v4Rrdvaq8cxZZV31hSqa3Foez2Hstmbx16DbieP5iBma2n/KeeFFIVUReLrorNp6RGxewsbk14uv9MjJadzxnefw4kQOkBLVzUlVMYO67qBjherg4SNddJGzXGQtt6aoqh0buhVhbdAjp6Yr81RYOFOfSUolXFJwvJrru/nyLZXv2RdKdCIQMB1ORRE3dQgpy3OeRCpm4C9+65quF7UkCGLzEPgAnj2fRtbyyrl66ncLxWnDMpYudvISO4oQEodPTCMR0WBWCRSv1Oob3EtfoCZnjjElDmR7Prb1xip26XqxCwiiGZu2c4ft+a0PKiMAfPZHL+F9N11GbWcJootIRXVYXvMsBcsTSIV0oi+HVqoRCZPjtoPD+NnJ6RWpCL1p/xC+96Gb8be/OItP3/8iio5fMZiihmqPmzDVhv7A9p5yIQEatn3j5cS950dXRj0vrGLZUpXNOkHQKq9ZFW8r5TwpJSxXoOT6KDqeUo9cQMZy4AqJ1EIvcRlTY8hJiZzl4j037Mbnf3wc03kHiYgOISTytoeC48MXEulSHicm83Xfx3YF9g6amMyrgp2FqprdghAS9xw+jaiuwfWbr9U6V0FYAWAmX18RnDFACsAHcDFng4HhsdOz+OyDL+EPbruy5T244dIB5ENU0QdjJl2g4g5i8zGVa90Nr53jCGKjMBA30Rs3EPM45oouqpvNaQzKXtS1rle2CwLgqaiOybzf0PmosaBNumrFvqMvhmt2DeGBkYm2zscZ8L5f2YsPv2k/OGd4/817a2ytZ87P4U/+4RjOzSqHJmNAROcYTJiYbDHPfP3x83jzwe14/ZVb2rqmbiWMvUoQxPrj/GwReccDpISucYApda+i48Oqam0eJp9yKa3emxHRGCIGx0+OTeJNB7bV7ZRquT7+6uHT+F/PjtW8ftO+QfzBr1+BvpDrXsTQMJgwETUWK0USjWnmR2mldnrHWw4AQE3LelFW6d1R7sI3OleCkBKMzT9fji/w3/7xBPYOJWp8NUECzp33H8NLE3l4ZREHQ+PYPTDf1a+eMmgyoiMR0WA5Shnd9SX+7B2HcO3uvkWfWQiJz/5IdewwdAYGBs+b70ggy108bM+H7Ups6YmiaHv4+G1XYSBpVtZQAHjizCxenMiVOxQoYZSooWEoaSJv+6HVToHWvsKBhIH3vnZfpVtGRihBlgPbe9rye1V33N0zEIMoqykXHR/ZKgGKIEEvEHAJ/HwRzuH5KsFXdShRSq9RrmE4xTGRtVfUjzQQN6FzIGu50DirdCoKrm+tuhN0E+3YfI2UdpMRHXEjjpNTBQghMZgwkSm5ECh3vhZK5dAT1cFxNV41xuBDwPEl3DqLCkOw1kg4vqjxGffGTHzh3dfNzwP+/DxwxbYk7njLgUXP+o17B/F7r9+Lv3v8HCazdvn4+mOj2jf7yMlpPHh0ApNZC//9kTP465+dhi+V+BNnrK5feaHKsScEpAy6gavkM0Pn2NkbQyyiIWd5uDBbxFTewVbOmnYLaTTmwo6tTnSOXg6i0qHDR8ntbCFHpuTiqbNzGBnN4NrdfXjfr1zWsffuFH0Jo5w0LSuJldUNWxhTiZY3XDaA1+8fwl0PncSLE7nK2nvVcA/efPV2fOaBF5sKOkR11SV9fhzVEny9o+kSINU1VHeyZwCYBI5fzFfuZyNu2j+EL7z7urrX+sE37McNlw7gd7/8BBpF3ASAz/34OMXKiXVJKyV6x5dIRQ2YOq/MuYMJE4kIx8WcA50z/Msb9uBjv3rFouc/XXBDJbCVHIFIVdKbRCDSp9ZaWVah55zhhyPjTcfi/3x6FPf+q1fjnp+dxksTOTi+gKlxXDncgw++IZwd2Sr+GnQ7GUhE4AmJ0bkSXCGh8/Lcw4CS66M31rizHkF0ksDue+LMLCayFnb0xvDKPX3Qqoq21oOvNLARM+W9mq7xsn9HjXoJVVgbCCyJcrsLz5cYz1iAVOmsqZiBTGm+AD7Yf1fv55eyvwxjg/7+6y/Dv/3qkw3nqUb0J0xYroDlehDl4u8D21OV6xNCoidq4HdvvgzpgovxTBFffewcHE8gGdUxkDCQszxkSy4SER2f+a1rcPPlG8PPTxDE+ieY33OWW+nWwRmr7CeXipTdtY5VE9iLW3uiEFIJgQThAneZn7sanZcLOqT6/54AXE9C1+Z9v365G/NtB7e1tfavB9uB2Nxs2uKOmNneR7c9ib/40Uu44zcOrNAVEQTRLicu5kIfd/2e/hW+msab5u29EUgp8Z9/chyuJ2Ho8w7zThYocM7wO6+7FHuHEvj4t59D0faQihnoiepwfFmjfOB4oqIAY2qAhGqlyxjAghaXnAEcLdXzqtt+hqE66NwfN2FqKoARKJYFChthj1sJ2ikqCap9AzxfIG95KDk+So5fUfNqRG/UhMEZXF/WOHYDHF/CYAwRTYPt+7hquAePvTyLublSy8/By0kIPREdybLy0FILclaLYAMw3BtDwvJwMWvVTSLVOUPc1LE1FcVssX5hB1AbFAsCzZDAXx0+hcMnpvCJ31gcRK7mvmdGQ208pFRKhv1xUpMlNh9npgsdPY4gNgrViUJXbksia/lwfaWQm4pqmMyvntLqchiImxBSYq44H6xuVFDpC1UADKnW6p+fmIYrJDQGGOXEmEZJxjpX6ynnDNdUKahU21pHTk7jrx95GRJK4UsvS3OXXIELaatlArME8Cc/fAEPXH7LunfOHTk5PZ+Us4J7DIIgVhchJB54fqKSNC8l4AmlaCiBmsB5mPBI0MDA1DmiOkfB8dBIIyMIjNR7j2DK1DTVveDMdAEnLxZwxXCy5thTk3n86Q+P4ezMvKJYROe4/Q378LZrti8SjaiHxhn6EyZSUdpbtUsYP0qYZI9qEYvZvIPPPHAMhs4wkbHhC6W2L8X8M8igFDcbJSlnLQ9xU0PcNMsdYYHxjF25phv3DtYkVwfPCQND1OBIlyQObO/BoZ31fRhHx7I4N1MAYwwcTI2dqt8zlAs8JOAJgYSpIyMkBpImbr2iNhHkjrccwB33jSBTchE3tMr1povekhKpwyTYLCxkbSeI2Ch5PBHRMSBlTfJ44G9ZKOBiOQKOL6CXEx4tVyBmqqTHRsIunQx8ZkoOSp5ANmeDlzsNR3SOLT1RJEwN6aLbVlHNRiPwUZ68mEPJFeAM2DOYwB/8+hV1E5kWFivUfNeumO9s6/kQQKXDM9cYwESNDyyizf89ZwxsUS/hWhiAbMnFQMKs8RnfesUWfO9DN2NkNIOnz6fBJHDdnj4c2rlYNbHaJ+t4AmDAcCqKf3HDHrznhj0NiyBylou/f+JcZf5zPIGxjAPPVwVqO/tjMDW+yK+8MHktZnCg7BsXEtA4x/beGOIRFU80NI6+uIndA3FMZq2mSXONxhwQXjSpU0l6YZFSouj4KJTFhDpV0OELieMXc3j85dlKEV2wdzs7W+zK4o6hRAQxY3Gnzeoi26ShYSgRwaFdvXUFoADgwaMTeOLl2boiZVGdY9+WBI5P5pvGaQBWU3i18HokVDeeZ86lmxZ3AM3Fqr71y/M1XXvqYbkC9z0zit9+9e6mxxFEt9FsfQyU6F+xI4UP3LIX9xw+XTPnXr+7r+mcq2JCzfrcKlSxsawURjCo/RWH6g7klTvpmBpHwWkufGa5AhM5C1/53RuWZJOF2Te4Qlbiw1FDraVTOQu2JyodRzTO8K9fdyn5oYhVg3OGq7ansKVHdYk/P1dCPKIhGdHx9Nk53P3wqa73lQY2YtzUkbd9lQTL5meRYARrXK3/ni8hyvMLg4RkqnhZCOCPvvt8Tb5EK1HKsLSyQUfTpZY2Qz0Spo5tKQ2jaQs7+1RnzmBP0Cg343d/5dJFggjXtJiXCYIg1oLq+T1rWdB4uVsR0DS22oqrd3avP6y6Ox7jKjfNcpUwcaeIGqrDHQt8VoypmLOuYs5SKNva1DXETQ037w9f9LcSYtNULEJ0mk1b3LGjL9r23/yPI2dw6xVbyEgkiC7hpy9NhT7u3a/Zs8JXo1i4aT4/W8R//acTmCs486qaDvCLl2dxYvIZfOHd13V8Trn58i343G9fWzFCpvLOoqDPyIUMUlFDKbRJ5SBg5YQ6v2z8pKIGdvfFK2o2EcZhuQKeEBUVvXbV88Iqlt1w6cCylM2WQ7tFJb4v8NS5NCayFmKGhkuH4nXVSxuxf1sCuwcTOD2Vx1DSBMO8ulvR9TBbUMqFH/v2sy2N4ISpYfdAHGdnChhMmIgZGhhjMDReuU9LKchZTao3AFt6IogaHBez8w5jzlRA+V2v2oXHTs9gPGMhZ7XfLUNItcH6D3//NP7zP7++4Tj8pxcnQ79nKmpgMBlp+1oIYr2TKYUbg2GPI4iNQnWi0GTeRV/cQK9mwPYFJvMrr7TaKQ4M98CXSsnZ1FWiibOg8JEzYE9/HAKqc95kxsJUzkZvTKlpgZVVxhqYMqqOmFVs0HrraWBHFmwPO/tiGEtb8IWExlVCsu2FcxZemCt1bZFrWI6cnMZHvvkMZldxj0EQxOpwdCyL01NK7epi1oJTVjgPijTaDYtUJ/zZnqgUaQT7KkCinDvbsEBOln+nMfU+QgKulMhY83tKISX+51Oj+Oufna5JDN6/JYlPvvUqXDKYaHmtjDGkojr642bXr43dSFh/y5ffd0PLZI/qwkohJL7z9AWMjGZQdLzKc1L9DUkAJcfHC2PZRQUAwdq9qz9Wk0QXTanusnc9dAqJiI6b9g3h5GQeE1kLfXGzoTp9vQDZbNFRanjlawmS5SpPYtUPOm/uy7pp/xDurEpiydneshOpWyXYLBQNaYd2kser/S3VBJ0KtLIKnuqwMq9ovdCP1MnA55GT0/ij7z4PKVUyo5QSjDGUXB8X5oqIGUqddT3YzCtB4KOcLagiBdcXEACeu5DG733ll/jYr12B379lX83fNFPanco7YAD6EiZm8jb0cmHHQjQG+BLwpYQeFFsxphIT6uRwBYkKusZgewKWoxaW6nHGOcO1u/uaJp4v9Mn2x5VPdiJr469/dnpRd6CAhfMfAIxnShASMHUGTwAXsyqBbFvKxMVsbVfjhclrjKkEgZihoTdmQEqJkuMjYrBK8u+X/s1rcGwi1zRY32jMBYT10XYqSa8ZJcdHznZRtFsLFYVltuDgiTOzePzlWTx5dq5h94pnzqcxV3DQn+iu7jwHhnsq97g6Eaf6/3POcGC4p+F7cM5wy+VDePTUDAA1tgL8cgHva/cOYiJrN43TGBpDyW3+vUgAjftsLr6ueutOWD/4P704ScUdxLojbDekm/YP4aZ9Q23NuRfSpUridSuCZZRBJb85voBX7ooX1Tk0zmForGVxB6DG4rteuSvUeWuuIeS+4Q9+/cqa+LDGGAYTJnyh5ipfSvi+wM3kfyJWESEkXhjL4txsAb1RE/u3JZC3JH52fAqfuv9FZIrzvlLmsK70lQY2Yk9UR1kjqaY1a0VEgSnbQVbtwRljiFYVwtfLl1jO/rKaZjbo7X/75JLe0xMCF7M++mIGPvkbByp7g2a5Gedni/jTt1+N3phJibIEQXQFjZL3g/ndKHdkCsRmGACwmqk+NHFDzfndysLueFt6IrgwV1xSEUs9TJ1X8ucCz4oQEpwBO/viqnuzUDGP2YKL4V7VQUQI2XKdWAmx6ZUoFiGITVvckYy0r/5mewJ33n8M3/vQzWQsEkQXEHYUrvZoDTbNQkj85hcfwXTOBgsSJ8oRACElpnL2is0prYI+B3ek8IodKTx3IQ3Pl3B8Ualojegcusbwih0pvO2a7fjO0xfw7PkMfKGU/ALlKFNTjsZrd/eGVs8LG3T+/nPjy1Y2WwphnZqvvqQfli/ws5em8KUjZ3BupgC3rBaxezCB99ywO3S3GM4Y3nPDbnz2Ry9hLG2VW6uLUN0iEqaGQ7t6ce2uPly/pw/7tiRxarKAP/7eCDTO6wZo2y3IWW0WbgAYY+XEo3nHt5QS/QkT//w1u3HfU6OYKSytUEUCmMo5Tcdh2JGpcTVmNquSJLG5MbRwIyXscQSxkVhtpdWV4NhEDhpTyVqeL1HPRAkKPnrjOi5mbYABW5IRxEwNc0UHRcevqM7XI1AmDGzQeutptR0ZNTTs7GeYytmwPX/eURri80iJri1yDYMQEnfefwxTa7DHIAhi5QmCQFt6DMwWWKXAf1EXgpDvFyTiB9NBMIcLiUrypsaArakopvM2RNUkXx14kuW/ZRKwXR8GY+iNqj3lTN7Gpx94Cb88O1dz7ne/ehd+91cug6nXT2qtJm7qGEiYoY4l6tOuQnxYX0qQCPeRv3+mpgCoumuHrqmii6zlYiZvh76miM7xxJlZvP/LT4CBQUgBXwJzBQecs0U20yMnpvDZHx3HuZkChARiBsf+bT247eAwYgaH46kEPZ0pv5WUVUFUKD9WxGC4mHWadoJYiUTqTiXYLKSd5PGF/pYAnXMwpuYFxtTPEhKWowRe/LL9OhA3Oxr4rPbB7e6Po+D4FdsO5cICzoE/ffvVNe+5WRTwgvszW1C2tJCqA4XOlLK/7Qp87sfHcWB7alEHj0Z7kL1bEjg/W1SFUBIAk5UuUdVDlHEG5sty8qasBNJZgxWIATB0Ds5QLkLxUXBEWx1Xwvpk6wn9LJxrSo4P2xNgTHWkFlJ1jjg7U0S0XLCx0K9cPe4fOTmFe392GtmSi8mcP99FoZzUevut+6DrvOWYbjTmAtrx0a7EHGK5PvK2h4LtdURV0/UFXhjL4vEzs3ji5TmcnMo3PT4V1XFwRy9u3j/UcA5bS4J9sKGxsqjA/L7LFxI6V2qhxyZyyFlu3QSOD9yyF4dPTCNuavCFrImtRMuxleMXcziwPYWR0cZxmq09UZyYbH4/ASAVWV7Iv1vjbQTRCYSQ6Ika+Oev2Y0Hj17ExUwJGYm6Prp25lzVeXEcGgNCao6ov5Oq421/IgqNqUKJkuOjJ6pja08kVMxrJm/j33zp8baTx8LuGwBg39Zkx+LDBLFcgoTJExdzsD1Ricn/i9fswr2HT2M275R9pQgqJuALYDLbXb7SwEbk5bW+6Kj9z8IpxPMlTJ3D8QR0DgwmI+iJGErBvDx2VypfohVLvYu26+PAjt6aeSrMPuCew6fx5ffd0BXfH0EQm5tmyfvB/C7KhbuB3l1Q5LEUrt/T39W21sLueMmIjsFEBBNZa9nvzaHWdCllRXyKB0UyZd9P1NDg+gKTORtCAhdmS7j9b59saRMvxwfViJUoFiEIYI2LO+6++27cfffdOHPmDADg4MGD+OM//mO85S1vAQBYloWPfexj+MY3vgHbtnHbbbfhrrvuwrZt25Z97qUkgjIAL03kMTKaadnath02S0CCIDrNm67aih8+PxHquLVgZDSDlyaUA0wIQKC25RpjKzOnBDRzQFar1ORtDwOGWQl8l1wfyYheCVQF6lKBMjIDq7Ro17jALZcPhZ6zwgadR9PFjiibtUszp6YE0BPV8dJEDv/04hQKjovP//g4io6PVNRASmNwfYnTU3l8/sfH8dFfu6JpgUfR8fD8aBbPnE/j2QtpTGTslgk6iYiGQztrizm0Bfe+UScQYL69dDvB1dWmegOQjAilyi0ldM6hQcL1JASA/+/waSQjenlUKUM+cIK3G49sNg7DzjN98c2rJNkpyB5av8SMxUkKyzmOIDYaq6G0upLMFh1wxtEfNzGVs+seIwGMZUqYKahObzFDQ8xURZrJiIG83VhxsC+moydq1Nig9e7NQjsyGdGRiGiVxEPHV3ZDKxIRrSuLXMOug2u9xyAIYmUJgkA5y4PnS5U0L9UcEWxz2o0F6VXBpHowxir+APXzfCeP6llIQu27LFfgiuEe7N+WwM9PTuMvHnypRo17MGHi/3zLVXjVJa0FDwyNYzBpIm5uWv2hjtEphfh63LR/CO965S7c9fCpymtKNVQVAmicwYOAEBJzxfluffWuKSgayNsepvMWfKmSw/rjJhxfYLbgwNQ5fud1l+Dm/Vsq6+G9h0/hcz8+DscTSgQCgONxPHs+g3OzRQz1RFAo2wSeVP6BIPgXPPsDCQMXs06o7mkrVYyxHOrZCmGTx2fzDoSU2JqK4sJcsTZwaXKYmkoqipsaPCFwZtqC7fmVuScVMzBXtHHvz17uWOBzoQ9uoW3nC9U5rjc2b7dtJgW8o2NZHB3NIG/Pd8yRvkoQ0DUOQ1PdlD77o+O48bLBRV0kgj3IyGgGz5xLQzLg2l29+NyPjuP50Uy5q5MaIdXrA2eAFOo8GldJZWoIy0oBgMaAbb1R+EKqZwtSrRtCvV/G8tAXM3DbwWH87OR0qP1PuwVq1SycazwhIIQqEALmiyIZUwUNtusjHtEXzYfBuM9ZLgxNJc0tLHxpZw1emORQ/bnWykcbFHQUbb/cpWd5TGQtPPHyLB4/M4unz6VVcmIDNM5wcEcKu/pjeHm6gOmshRMXczg7U8BjL8903TgO9sE7emOYKdiqqzTU8xQ3NQwmIii6Ph45OY2/f+Jc3QSOj3/7OTiej22pKCL64q7olidweqqAD75xPy7MFRvGaW7cOxiquCO7hO7W1XR7vI3oLtaTL7+e/bCtN4bbDg7j5v1Dy7p21XmxgK09UYxnrLbWicFkRK1LQtlrr9iRwgdu2au6Hp2Za/n3Jydz4Jy3nTwWdt+QLrkdjQ8TnWc9jcPlUp0wmYoaSEZQicl/+v4XkS7vQ5WQRfmPqjpevDiewy/PzOKGvYNrcv3VVNuIiYje0G8uoewnXWPY0RtDKrZYvHi5+RLNnqFme6+wNoPGgywFiYiu4d+96Qr869ddUvOcLmcfsNpspjFHEMRiWiXv/+nbr8a+rUm8MJaFqXGUyvvj5UgpvOXQ9q6eZ+p1x4ubGvQ2C58XYmhKXMHxRF3hQAmJiayyvYNuxMOpKPpiZiibuN7aI6Ws7NmjBsfJi7nQa89KFIsQRMCaRs527dqFP//zP8fll18OKSW+/OUv4zd/8zfx9NNP4+DBg/jIRz6CH/zgB/jWt76F3t5efPjDH8Y73/lO/PznP1/2uWNG+0ownCnH8NPn0x1LkthMAQmC6DRvv24nPnHfCJwmHQ5MjeHt1+1cxaua5+nzabhCLGiXiYr6gpRK0eqHz43j0M7eVV/EF6q4BZ0nDmxPVeYgISQOn5hGROcouT48AQSmU6AocfjENN5/895Q1x826LyzL94xZbN2qHZqynKRgJBSKZpKFbx0hcBcycG3n7yAouPXFFBEdIahpInpvIOvP34e1+7uU2rKUMUcI6MZPHs+g2cvpPHSRK5lEQKDuse/9opteOs12+sWcywk6ATy+R8fx3TeQU/UgMaVol91e+luNRqDDcAd33kOo+kShFAJTUBZbQ+B6oqE5wvETPV8iPL3Y2gMjidDb5Ik1HfaaG1/+3U7ccd3noPbJN7JAHzht7unpe96hOyh9U0yGq4jXtjjCGIj0o0JgmEZiJvQOZApeWBMqRp6YnExpZBAqbxgJqM6GGMQQiBTchqqzDMABcdH3NRrbNBG19HMPuSMQeOspepsyfGRKXVX54521sFQe4wO+y0Iglg67QZ+gyD/U2fn4HZARRuoDaQwAIyr4jBABT6klJjOOZUkWA5A47yizFqNEj3Q8Fuv3In/8pMT+P5z4zW//5X9g/iDX78SvXWSDqrROENf3ESqvF4QS6P6+ZrNO9A5VsyP8pq9A9B/dqq892bg5SQNVi6iEFK1u+9LzH/3C9fuvO2pzgyuD09UFScyZStFuYbtvVFMZG0cOTWDD9yifBePnJjC5358HLYrYOgMHAyy/Jk8ITBXUArw/XEdc0Xla/N8UVO0lIhokBLrqntaNQttBQDYmorgX7xmN/ZuSeLFifrJ41M5C4wxfOaBY/AEIKRAwfFxfq6ILT1RRLRytxONlf1NEhfmSpDl7xNlIQ0hgD/8nyNwPIGBRGeSbuolFjKwsp9HgxASk3m7kqi02RTwHjk5hXTRrXSxDew9IQHXE6r4ACqh7bfueRSTWWuRHQlgkY05mDRRXaixCAlIpubprT0RpIsubM+HX+4kkDQ1aOW5hDGGuKljKmfB9gQ8ob7PXX0xMAbc9dOToX08yylQWzjXaJxVulMF941B7RcYV/fPcn301VmrgoC8LySu2JqE7clKMn5EZ7iYc0IH5OslOQRjrpGPttMJY0FyQt72UHSW36HDdn08N5rB4y/P4okzczg3W2x6/LZUBDdcOoBXXzqAV+7pw/GLuRqxpJihwZeyK8dx8FyZOselg4m6hRm65+PBoxMNEzjOzxZRdH1s64nWPUfwXO8eiDeN0zz04mSoa54IIXYANH7Ouj3eRnQP68mXX89+sH0fZ2YK+B9HXkZPVF9WkV2wfm3tiUBI4GI2fIHHr161FW+7bmdlLGZKDu45fBonJrIt/zZYGYZTan4J5qjeqI50yW26VoWN0/bFjI7Gh4nOsp7G4XKpTphMmDo8IRHRVCfIoaSJC3NW0wRSFQeWOHxyGtt6o0hEdCQjOqJrJEZWHQcfy1gN/eYRnWNXfwy5ktuwy+ly9vnNniEATfdef/K/XQ1TY01tBiCYKzTETA4hULdApdNCFStVgLGZxhxBEIsJ22XoA7fsxR9993mUHB/Ll1MAclWCSt3KorxDX0LTOKQvmgpO1SNYE02dI6ZrmCk4df9eSmBrTwSzBReOL8ABmLpW8S+3KqhYuPYon7XybVW6QDPgkZPTFR9ns/VlPRUqEuuPNS3ueNvb3lbz85/92Z/h7rvvxmOPPYZdu3bhb/7mb/D1r38db3rTmwAAX/rSl3DgwAE89thjuPHGG5d17rjZvrHul1v8sM7EVjddQIIgOg3nDImIDqfYWJEoEdHXzKkkpaxRvqx7DIC/e/wsXpjIrsnmr5WS9NGxLF4Yy8L2RDlhbt6gkhKwPdVuPawRElax7G3XbMd3nr6w6spmPREdGgcKjqccJQvme8eXMBhDtuTi/EwBqahR0xkDUIHonqiBs9N5fPepMUwXbDxzPo3jF1sXc3CmlO3jpoaYocHUGWYKLsYyFvZvTVYKRVpx/Z5+fPTXrsDXHz+P8zMFFB0PpsbXTRLDTfuH8Pu37MOf/MMLkEy1zQVEORCrVD0lAKscOA5whcRSdklSqueqHo+9PANN43CbqNldMhjDr1ze3fe0myF7aP1jhFxnwx5HEER3cXBHCtt6Y7iYS0PXVEJlI3ecWe5klrc8xE0XF7N2peCjGp2zytob1TV8/Lar8JvX7QiV8BzYhwXHVwmiVarSMUNDoYlaLABYno9P3jeCO995TVesL+2ugw33GGy+vXMz24YgiNVjKYFfzhluuXwIR07NrMg1SQAQqqhDVGUMe+Xk7lTUQM7y4AsBnQO+qJ3xd/XF8K5X7cJdD53C+blS5fWozvHBN+7HWw8NNy3WYIyhJ6qjP262FC4gmlPv+Sp5AoWsjT0DsY77UYYSEfTGDGQt1UUgSLsQUsITEpwxpKI6hhKRyt/U68wppOoIUC2mOpW1EdE1JCN6OeCl48XxHL7y2Flcv7sPn/3RS3A8AUNj0Jg6MwNgcOUHcH2B6ZyND73pcjx4dAInL+ZQ8lRgb/dAHL/1ql3YM5hYt8qa1bZCRNdQdFSy/XTexv/z/RewZyAGjWNR8vhUzkLBVt04klGjYmO4vrKf5goOOGcwOMM1u/pw8/4h3P3wKZRct+J/ihkatvREkTA1nJstoeR62JaK1L3OdpNuwiYWDsTNTaeAJ4TEg0cvLra4q+w9Vyg/Wd72cHoqj22paI0d+ZFvPgMA8IWssTHH0qWmXfUklK1+5bYezBYcxE0dcVPH1lQE//KGPbh0MI4/+u7z84qMhoZtqShm8g4iOsdbr9mOfzx2EQXHb8vH087zsJCF+4TqGxeYxJzNF3pUv76Q6oA85xyqccz89bQbkF+Y5JCpJO0v9tF2KmFMSomS66Ng+8su6JBS4vxcCU+cmcUTL8/+/+z9d5xdV30ujD9r7Xb6OdPUZ9Rly00yYLlgDCEYQ/JJApc0DIEQ4Me9kNwbICSXkPveH4SElxQISXACBBIIMQk1xBQbY7CNcZGxrWL10Uie3uf03dd6/1h77zm9TJFG0nn4GFuaPafutda3PM/zxaHRDCynfo1UkQj296fwkm3duHlbN/pLziPGOe49OFJmliT2ILom13H5faUFwjNg8Vzd0hXGdNaoS+BIhBVBEsmb0G13kSjikaITYSW4r6/fkqzbp3noRHNnbABwefPCeKP77JYdPdjcFca52fqinc1d4TXx/XRw8XAp1fJrxQ++2NiwHWQY8JH7juP+5yfw7lfsWtLrLj2/+uIaQgrF6HyxJafijclQcJY8PjiLP/nP55E3HYRkCokIXko9EAJ0RzWvNmaU7S8ypTg+nql7VrXapwWwov3hDlYOl9I6XAmUxmejCzp02wUhQFSVEFVlqBKgtzK8iguBc1a3kdXFpLaoNz1Qky+s0MPvg3/4vmNlMT8BglqJyzhmcyYGeqIYXdBXlC9Reg+FFQmyJCbWnZjI4oPfPBLEMHUJzD8ZwqZUCOfn9LrPIVOCvpiGjG5hLu9AosJ04JvPjZbFt8vJA2q9r9UQYFxpa66DDjqoRqvk/WRYxZtuHsBfP3i6QSe1dTxwbBLvumPtC2lLeYePDc7gG8+MYmi20PbjSJSgvzuMTckwDp6bDz4//937f2YcmC/aYJxBlQhcDszkDETVqDelvLGgovTscWyOsQUdLueQqZhw7noxw5eeOI993u82Ol9Wc6J2Bx2smZn3ruvia1/7GgqFAm699VY888wzsG0br3rVq4Jrrr76agwMDOCJJ55YtrhjLr+0BUMJwf6B1LKeG+iM5Omgg5XA4dF0MGazHtJFG4dH07hxoOsCvapFJEodyn07whrK1JAirdnkby5vImvY4JxDkWmZkIGDw3YYsoaNubzZ0uO16lgmyzS4biJjIKxKkAiByzl0y0U8JK/I9Am/2VW0XOiWi0RYxuauCIZm8mUTOfz3mzNs7OiLIRlSYDOOhLT4c5dxGLaLovd4psPw9w8PNnz+mCZjR18UQ9N5xEMKYppUlQzEQwpG5goYnCpgz4ZYy+/txoEu7OtPYXCqACoBGxPhS4rE0N8dQSKkIB6SBVnE5ZjKGcJljxAwJgJqzvmKJEZxrTokenxwFh/4+hFYDqtbTE+GZJg26xSwl4hOPHR5YL7Q2hnQ6nUddNDB2gKlBHddux5HR9NwXA5Kyqd2+E0nDmBjKozprCCujC4U6wpbiecGbDscRdtBpoWCVmkcObJQRMF0q1ylAY+4hfpNcMaAuULrjruriaWcgzVzjNI/e0h0piV10MFFxVIbv4xxPHJ6BoonrFgNcIi9UIwrD0OiBGndwpauCKayJjYmNczkLViOC+pPPiDApmQIr7luA/7uR4NlIvvd62L40C/uxUB3pOHzRlQZ3VG1ruNkB7XhuAxTORMRRUJYlRBSpLr3V8FyULRcDM/rWJfQmjrEN0OpI1kqrGDvxgSOjGbhMhZMdiFEiHskSnHNpmQZqSRwJP3W0bLJnP69TYCARDKdNUCTIeRNB5miDdNh+OQPTkORCTK6DXBUvXZCCGQqPiPdYejvjuCLbzuwKi6dFwulsUJMkwOBjEQpJCpqJaMLOroiCjYmBcHeJ4+LqQoSBrojZTHGQHcYExkD/d0R/P7P70ZPTMO1mxI4Np7Fv/z0HOIpsS/47vT+7ybCMgqmjZzhIFWDWNOua2yrxEL/tV1JDnjHxrOYyujQZDEdoBK+wAMQa7AvpgXOwyEqYX2C4PRUHgCwZ10MlIp917E5CqZbtodXwhdM/9nrrwclpOZa+vPXX497Hh7Eyclc0NS+bnMC//3lO/GZR4dQsNy2azzt3A+VKM0TJjJGTZMliYq9xmEcEhUTYtI1WICr0ZBvZq4ELJ8wtpKCjqLl4LnhNA6en8fT5xYwmW08DaK/K4ybtnXjwPZu3LAlWdcFe3CqUN8saQ2u48r7qlZ/4q5r1+MLj52ve7/EvQlls3kLlArCtS9w9HsS125KNCVkzhVau9+aXdfsPnvHy3bAtBl6o4pHlCn5PAjQFVE6dfArHJdaLb8yfsibDsYW9CCWopSDMY7nx7JL7s9Wnl8xTUZXVMVMCzyUhBcz+Z/rQtGC43Kki1bdmhYlQEyTkDfFnj9XsMqIaByA5bowi26Z03DZY7TYp10oWCvaH+5gZXCprcOVgB+fUYjzExCxcN50G4qWKxHTymultsuQLlpIFwFFEus3FpKh1DnXVxpbUmEAwjhBlohHRhUmkv4ay5kOXn3Nenz1ZyMrxpeo3HMyuh3k9qpEYdguJrNmmUjXhx+znZrMgYPXjBn89+QyjqmsAcbF+9uUDEOVaVV8u5w8oBSrJcC4EtdcBx10UI1WawWzBROPnplFTJOxOanA4Rwj8/V7pY2gSgTTWWPN5l++EZ2/91FKkDNs/MfTI8ibDjYmw0gXraaGfKUg4Lj7wABu2tqNX/vsEyBA0EswK+pjPoeLSgTEEx4bNvOMGRrXb/yz5/h4FobtwuWLtVQODs5Ebcx2GT72/RPI6nZDE5OVFCp20EElLrq44+jRo7j11lthGAZisRi+9a1v4ZprrsGhQ4egqipSqVTZ9evXr8fkZH2XEtM0YZqLSWQ2W3t0ZcFa2uiinqiK6zcvf9O80hoSHXTQDlpdx987OtGU0M296y6GuKM7JpwoXc9JmNQZORZWJcRDykVJ/po5GCwUbTAmSHO1JlRQKkj2C01ENqVo1bHstl29eNPNA/j0w2cxkdbBIIoB8bCCN908sGQRjGm7eHY4jcmsgYgiYee6aOBISAnB3Qf68YkHT2M2byEeUoKxojnDRkSVcPeBfkRVBRIBMroDh7FAzNEMMU3GDVuS2Nefwv4tSezoi+HZ4QX8v98/iWgNYQcggvYc58gY7YsSKSHYsyGGjclwEMReKLS6juvBD4AlShBVZOSMxeISh3DmBDwiJwXsJqNfm6FypKFfrCmYjnAElSkUDlje+EBANNMLlhD01CuUd9AYnXhobaPVdTzX4hnQ6nUddHA5YrXGgbeC5Z7JAHD7rj78w8NnkTOcKvdBiYrzmnEOhVKsS6h4YU6HwwCpzlt0XCHO9IZz4W9+dAYPnZpu6qJ1265efPR11+F3v/IcGOdVrtKMcQwvFKFQArdObMABFC13TbgLLuUcrJVj+PD/U6IE3bFOkfBywkqs4w4uHJbT+L334DCeGpoXUwlXERyCvJCKKDAchqgq440HBvDZR88iawhHbcYBy3ahOwxRVUJYlfHZn5wLHoMA+I2b+vG2l25rSH5QJIqemIqIetHL0BcVS13HHKKOYdouForic/+bh04ja9jYkAhB8kjbISqhvyuCkYUiAI50wYTNhIv61RvibbsR16oX9cRUhFUKxyXoiqotkUpu29WLd75sBz5y33HhfFYq7PAEBI4r6irnZwtlRLaIKoEQETdwCHKIXCnwAIJ6UXdEBaXkssod/VghpEiYzBpwGSsh1xHIEgfjHJbDkQwr+OjrrkdatzGft/AX959ALKTUjDG6oiqmswZ6Ylrwec0XLTgM6I4qNePUhCZjiopptsmwsmTSjY9WiYWUkjXjgHehzmP/u1gX1zCyoAcO2ZVOhQCgyRLCWnnNz7S5N8WNwHQ4wioCQqtTRzkoUwJKAcfhgtz/wgK2r4s1yF38WjEP7sfzc8Ul13jauR9qobSOnC3aZZ+RL1Rk4AgrYloCOGo21ZfSkG8l12u0Ny0nbtAtF3nTWZagg3OOoZmCEHOcn8fRsWzDxworEm4cSOHA9m7ctK0LG5Phlp4nY1hVZkmlWIvruFl/4pYdvfjXJ16oe79YLlsUG5VajnJUNYoa9Wmms60Rpxtd18p99pWDw7AchvWJMNbFNczmbVgugypR9MYUEEIxnTc7bqNXMNZKLb/VdVwaP3BwzORMMC7ExgQkOBuSYQV5011Sf7b0/Bqe12G7rOGEIx+ldZtj41kcH8+i4JHUSbBRlKMrrGBjUoPhcOSMQiDs8Ilo4ndFz8zhvMrpufK8+ujrrsNnHh2q26f91rNjK94f7mD5WCvrcCXQ6lr24zPDYYhrMvKm07bhHyVAMlLfCMd2GRaKFhaKFjRFQkwVEz3kFRJ61IoXfQ6GJJEgr/dRusY2JsMrypeo3HMkSkDo4kQexnlgClILmkdqBQe2dJXHDLbDoDtu0Ed3ufjse6KaiMGBmvHtcvIA//NdLQHG5bTmVgOdmnUHVwrq1Qo45zBsYbjDOcd83sLZ6TzCioSC5Yi9foll/ogqw15jOXIpXMYxltbRFVWRCCk19+KuqIKROR0Zo7V4UZUl/NtTwxhd0OEykRdTQgJuWCUYB1xXTIlmjMNhDP7EzUaCCv/sef/XDmOhaImpWURMHHW9CdXrEiFIFDg9lUdYodjSFal7vvzzW29aEaFiBx3UwkXvql111VU4dOgQMpkMvv71r+Otb30rHnnkkSU/3sc+9jF8+MMfbn7hEjfPN7xoy4qQcNZKQ6KDDtYiWl3Ho/P1x0Mv5bqVRm9UQyosI607AfmqEhIFFEm6KMlfKw4GKa+hy7g3IaEiCGGeW3Iq2p4zcKuOZf/21DBkSrApFRakfi5cOf7tqWFcuynZcsHCsF0UTAePnZnFl558ASNzBdhesbK/J4q7D/QHAqAbB7rwvjv34N6DIxiZKyDHORRCsLUnihdvTeGpc/M4NJLGbKF5ABrXZNzQn8S+LSns709he280GKfqIxlSoVAC2+XQ5OrzxXLF8ydDrRH0CCFQJAJNlqApFKpEoV0EZ9SWz+M6qHTqKHU2Y+7ielop0lPlJ+8Xa5JhBbrtwmU8IJT4EM114V7uj+RbS5N3lgLHYbjvyATG0kVsTkXwSzdshLyK908nHlrbaHUdVzZ3lntdBx1cblitceCtYrlnMgBkdAu67daMZx1vklZYlRFSKcyicHpXJFqXPFbJZYkoEk5MZPH+rx3GW27dhts9165auXcyrCIsU8RTYVBPxCFR0bymkvi31Uz0ydGyu+BqCnOWcg42yzEkSpAKy+iNaivyGjtYG1iJddzBhcNSG7+PnZnBJx48terCDh+a56ztNxd+8fqNCCs0yIVtLxfui2uYzBiYLCEM9sRUfPC1V+NFDYw0KCHoiqhIhOWaRgZXGpayjkfmi5jI6EiGlUBAc2oyj/MzBcQ1BY7L4TJBHKWEgBIgqsmYzVkgHtnbJ123g3r1oomMCYkCm1JiQoTJRJPsmk2JhnGNmMwpIxFWhFMyIZjM6jAcDttxA+Fo5fE9VzDRG9cgUwLby8klysvyCgYRhwz0RNdsg6yVWKLeNY8NzmAmb4rppd7nwx0GWaKQCAkmOIRVCUMzBVBC8PI9fXjk9AwchrZijGakdotxJEIyVFnySDcyGEMwxTYVVtp2jW3VAGatOOBdqPPYf7+aImF9IoSprFGLCw5CgHUJrSrXdtjiZB2HMXDQgNBKJYB5ponE+z/O4ZH5iff7QnytUFqVu9TbH05O5nBmOgfDYuiq8z00q/G0ej/48NfNXN7EU+fn8Y1nRgAObEqFMZMzYDtCEUOJEKfFNQWaTDCVs+o21dt1Dl6JXK/duEG3XBQsBwVz6YKOrG7jmRcWPEHHAuabTHzY0RfFgW1CzHHd5uSSHK2b1cDX4jpu1p/Yu3HxflmfIDBtQSaRKYWmiIkdBAQbkxqyhgPDdoO1GVZlJMMK5vIW7j04jH/6yZBnLiUcwxnnOD4upgn0xVv7THzSZC20cp9NZ02AAGndQkYXU7T811u0HSTCSsdt9ArHWqnlt7qOS+MH7gCm4woSs3fecYj7W4jeac0crZU4zheC/fWDp1sSdgDC1C1dEMTumZyBjG6BMUCWRROslrwjazqwFxgSYQWyRGAzBkUi5WcVRL9Kk6Uyp+d659W77tiBZFit+f5Wqz/cwfKwVtbhSqDVtVwan23ticBhHDndRt4Svf9GU+l8qBJFTnfKzILqwTdXmCsAIUVCLCQjqspV/f1WUW/9+UKDZmtsPKPjaz8bacqXaDXvfOb8PNK6Bc4ARSbB50EIQCTAskW92bBdRLVqOqHpCT85eFnM4LLFad8UwoRB8vr6c3kTHBxxTUFIpVXxbat5QL33uJoCjMtpza0GOjXrDq4U1KoV5E0HMzkDpsPgMLFP/O1DZzDt1V+WW+LPmQ5SVF5TObKPTNHGr/7j4/ilfRvx2us2IqvbmEgbVXsxAYFczwWwAhIBeqIK0rqNR0/PCCE252Auq3vWcyzyxQgQxOJ+/ebqDTEwLialV56Nt+3qxVtu3YpPPHganHM4rjehWpHQF9cQ02QUTAe2y9ATbXy+nJjMLVuo2EEH9XDRxR2qqmLXrl0AgBe/+MV4+umn8alPfQq/8Ru/AcuykE6ny6Z3TE1NYcOGDXUf74Mf/CDe9773BX/OZrPo7++vuk6325/cEdUkvO/OPW3/Xi2slYZEBx2sRbS6juuN+F7qdSuNazclcM2mJA6PZOAyBt12yxz/CRFOVyFFJIPLTf7aIZ616mDwB6++ComQjKzhwGYcMl0sKjqeYjURWhp5bCmOZQCQ5LypwwJj3ih6y4FuCWL+c8ML+MSDp1G0XCRCChKSaCYNzeTxiQdP43137ikTeOxcF8MPnp/C8+MZvDBXxJHRNA6NpBu+J995MqJKeMfLtuO1129sWiTatT6K/p4ohmby6I2pVaONc4aNHX0x7FofrX6+EiGHKgsRhybTNUGcaXUd10OlU0cyLEOVaNk6qoXankbNEQ3JODqaCdbPbMGE7XKkIgpm8wS6Xbso73IgJAu1+KU+dvVzj57Fpx8+i5xuB84vH/7OMbznFTvxzjt2rspzduKhtY1W13F3RMFECy6C3Q2ckTro4HLFao0DbwfLPZMZ4/j4/ScbNspcDiRCMjjnyOo2JEqwozeCnOliPK03PLslL27RLRfztoVPPHgKX3r8HHatF82bSkHwbMEUZEWFCGKps0iQUaVFMWg9EACUoiV3wdUW5izlHKzMMQThRjQANZlCohTXbEquWYJrB0vDctdxBxcWS2n8PnZmBr/7leeQuYCup+miDdtx0RvTcPdNA1goWrhxoAv7+lMYnCpgJm/ggWNTeGxwtuz37tjdi/feuQfJBgTCWEhGT1RbMvnhcsRS1vE3nh3F3/zwDDSZ4tpNCdywJQlNFrGE73zOOQ9ch4u2i5msAZsBPbKCrogKm3GcnGw97milXpQMq8GEiFaEj/55Z7tMEOoIQV88hJH5YtVEMB+KROByIFO0oMoUrlcLsF2vNuU5stkuh6ZQ/MGr96y5XJwxji8/9QK++NPzmMmboBAiDD/G8b+LevHGHbt78aUnXvAc4xAQ/RgHbIcBMi2rL+ZMJ9hXlhpjNCO1X7MpiXfdsQMfv/8kTk/lA+c8mVL0d7Xm3l+JVgxg2iXcrxYu1Hlc/n41hBSKqawZOORycEQ0GZpEa541vkGK/9+GxQJCq3DgFQuPL/5nUOv1YTkMm3pCsF0e5C6+w3e9/WE0rUO3XZiOi3CNaU2tTL24ZUdP0/sBWFw3x8ezyBp2MNGXEvE+uiIa5gsWXCZi5bzhIKbJmMo1bqq34xy8UrleK3FD2mU4P1dAIiwvSdDhMo7TUzkcPCemc5yczDXM0eIhGS/Z2oWXeIKO3tjyheOLNfAC1sVVSJTC/wrW4jpupT/xmUeH8K47duD9XzuM01N5MMaD3JRSgpgmI6xIUGSK6ixVTA8omC6+cnDYu185MrpRlt/ZLm/5O796fbzuz8omGHjusr4QJaRQaN79F9NkDM8XQQjKzJZ020XRcnHtpkQn17yCsVZq+a2u49LzNKpSsT69rZZzDoeJiU4hRfysMkd7fHAW9zw8iJOTOdgOhyJXT8RjjOPoWAbfPjSOkEyxKRGCCyHSbrR0dZvh//7X8/j8Y0NI6zZ8TYjt8Lo9LgIerMW+uIaZnCmciiUekL0XnYY1FC0X80Wr4Xn1J//5PP789dfj5Xv6qp6vN6qtWn+4g6VjrazDlUCra7kyPouHZIRVCZJEQSDyJ4BgrmCD8doLz3AY/vz7J/CPj57FTdu6cMuOHrx4axdiNcQLZb9nuzBsF3PEQliRENUkRFW55dyz0fobnBZO4LrN6q6xuCbhweNTTfkSjPNgEk+9GrYfPx8dywQTNUV+zYPajZgYIqZ25AwH3RVk1kWybBzpoonjEzkQ4tX3Sz56v6MuecJem3NMZU3MUTMQzOgWw7MvLASxfmleOFswkS7Y6IooiHtu8E8OzdWt09tebaA0ni6NdygRa2YpHJzLac2tBjo16w6uFFSeRapMRCzGOMCFqXNckzA8XyyL45bKXQJEXOdyYO+G+nneSmAp6/ieRwZxZjqPTzx4Bl/72Sje8bId0GQCw2ZIhkS/12GiDpxvcWqHy4GRBQOA6FsA8GLk1j5BDmA6Z8ByGSyHQ5aAdNHC//jyM3XPxtt39eFfn3gBEiWQvAnTIZUGnD3DFu4o9TinpX2el+/pa8uwpIMOWsVFF3dUgjEG0zTx4he/GIqi4KGHHsIb3vAGAMCpU6cwPDyMW2+9te7va5oGTWueRBp1SJqN8D9fuXvF3KvXSkOigw7WIlpdx9dtTuBbh8Zbuu5ioDTAy3kNpJm8CUpEI1amFH3xxUR8OclfK0XGUrTqYACgjDxmuYtuTaFVJI8txWHBcRmKtoui6Qpn6ZICDuMc9x4cQdFyywQUmkzQG1Mxm7fwr08OQ7ddHB7J4PBoGoPT+aZqak2mIqgucU0OqRLeeGAAv3jDppbeKyUEdx/oxycePI3ZvIV4SIEqCcdp4dYl4e4D/aCEQJEoNIWKqRxrSMhRC62u40aodOqQKGn6nSwlOaIE+JfHz2OhYAWB/bpECIyLe74ZLtTkndV0Dv/co2fx8ftPwWWisSh7+1SmaOPj958CgFUReHTiobWNVtdxT6y1c6vV6zro4HLBao4DbwfLPZOPjmVwajIPCoBKpMyFqxRTWQMzeTMQ0ToMnkjSbJh/u5xjImOIRhAVjSBZojgxkcN7v3oIfTEVk1kziHE3p8IwHRcLRQvcczDzjw/dctFKps+4N+GigbvghRDmLOUcrMwxuqIqJELgcg7dchEPyR0HmMsQKxFbd3Dh0G7j97EzM/j9/ziETNEGJdXTC1YTRYvhl27YiKs2LjaJKCEAEcSA0QU9+PuQTPG7r9yF1163oW4eqikSeqLqRTPZWMtYyjp+amgeAGA6DM8Op/HscDr4mW4zRFUJYUX8QygwnzfhMuF0pkgSHC5IGn0xFdO51uKOVuoxQzP5YEJEK8joFnSHIZszQQkJzvyGt7rXGLVcjr6YCttlsL1pXa43lYAD0BSK99+5B7fvbu21XCg8PjiLD/3nUZybXZwmTADoDsPhkUwQSwCoGW8cH8/i6fPzUCWCsCIFzUT/cTiEwINSQJUkFC0HnHOkPNHVcmOMRqR2QLj+hxUarHdKgImMueQYqZEBTDuvbbXjnwt1Htd6v9t7IsiaDrK6g6gm4eNvuB6f+8m5mt+xpiy6eGsyQcFyA0Jrq2XEouVicKaAzakwNiQ0TGZN/NUDpzGaLkKVKAyblTW8CRE11hFLx2zewpYuadWmXvhx+kLRQsF0wVhpDRgoWg4sh6E7qiJvOjBsB7rtIqPb2Lux8aQhoLUJIiuZ69WLGxjnYEwI9ygAhdK2hB3zBQtPn5/HwXPzeOaFBWSN+oZ3BMDVG+O4aWs3DmzvxlUb4isi0PQFCppMEVIk/P7P78b/+fbzmCvYwTo2HHdNruNW+xPn54qwHFY+8ZmLydMOYyAgGPPiKcnLXTkXPeqxBR1RVcLovI6CZ1AVgHM4lguJMjDeWj/72v76+6h/n6V1C+miBdNhYFzUxTWZIun9XA4UN/6bRfXozQ6uWKyVWn6r67j0PE3rNgAOxvzJVmKanN+jNRy3LEd7fHAW7/3qIeE278WdxASeHJrDmek8Pvnr+wEA//DIWZycyGGuIOJch4m6GCEEhFdPWy1F1nDw/Hi2bP5Wo+VGCAH39om4JsNxOYoecY6zcqdhiRIolCEVVvBXPzi1pPOq0lzkQvWHO2iMtbIOVwLtxNal8dmZqZyoqxCCneviuPuAIKB+4sHTyBsOQIRQynCqzQLnCxYeODaFB45NQaIE121K4OYdPbh5eze29UTq1jo45yhaDoqWg1liIapKiGoyIqpU93eax4sGKKUIyWJ6R601NtATxXTWaBiPHB/P4ANfPwLbZXVr2MBi3hlRJKQhSLPCQEHY/UmUCPMIiFw8rEl1c6///nIh+BcfDsAJatbka03XLpgu8qYLQoC/+/EZ/PDkVBBjU0qQM2x8/idDZZyX9XENE1kDjsuRDCtIhRXYbFGE/o6X7SiLp0vd9H26CCUEI/PFqtfTDJfTmlsNdGrWHVxJ8M+iex4exFND82UTIxhHbTO55ag7AFi2ixOTuVXjHwHtr+Oi5eArTw0Hfx5Z0PF//+sYtvdGYTkuzs0VSybKkrbqGMtJOwmEGGQ2b0H2JmNldQfJsIxkWAWlqOrvVpqrVO7xRduFXGJMUYnKPk8rBjYddNAuLqq444Mf/CBe+9rXYmBgALlcDvfeey8efvhhPPDAA0gmk3j729+O973vfeju7kYikcDv/d7v4dZbb8Utt9yy7OfW5NabjATCqea2nSunolorDYmVxGqSTjvooBYiamvruNXrVgO1yOku4wgrEtYlQoEjRKnTQb2xYPXgFxnnC1aQoMICnjo3jzPTh/DJX99f1bBq1Uk0rdsXhTzW6uubzBrYUgij6I1IrYfBqQJG5gqiqOqVSl1vuoduuchbDg6NNJ/MkQjJuGFLCvv6kwjJFF85OIyi5QaiA0rE4953ZBy718eCSSDNcONAF9535x7ce3AEI3MF5DiHQin2rI/jHbdvx8v29EGV6BW5p5YGwM+cn8dHvnt82SMMS+F/pC/MFhBSJchUuJKMzBdQsFwYjuG5iwK1dB6KROAw0RSwV3Hs6mo6hzsOw6cfPguXcagyAfXsqygBKBHK9k8/fBZvu237iolcfZTHQ4YgJZHFkcIxrUNQvRTQalzdTvzdQQeXA1ZzHPiFxHMjaeHoKRFIhEKiQthqVzRnGADKAUkiWBcLYTStI6FJLZ3bla7BEiWIaRJemCtiOmsuEossIKPbAXGGQJBm/P9uJUTgEKTXRu6CF0qYs9S6QGWOYTJRQLxmUwLvumMH4iGlrXyigw46EFipulY7jd/SiR0cF07Y4Yf1LgN+eHIGr3vRFlAimh5ffXoEX/jp+TIH9z3rY/jQL+xFf3ek5uNJlKA7qiIe6kxqWylwLiZSqDINRsqXwnTE9CZ4pAxVosEUBU2i0BQhmPSdS6OajFOTWTx2ZhYv2taFsCLVJO4uZfJMo7Xz+OAs/uQ/nxeCTEICIw7DLr/ZfSMUHzbjABN/r8gUvXENszkLiZAcOBL3d0fwB69em8KO3//qIUxXTDfkEN+b39S85+FBAKRmvJEKK1goWiCQsD6pYXxBh+uWEwQ5hKOqBRdTOReqRPGXD5zCu18h6gQrEWNUktpv2dGDt/7zQRQsF1u6yslPIUVaVfFyaRP95GQuqI00MrdZa2jnnKn3XezrTwa1IJnSut9xT1QFBzCVsxBWxNQAlwFOGweN5TCMLRSxuSsCTaY4Np6Bw7kQfVMxSbgvrgX1bU2SEFKEIU2rUy+EsY0MRaJgnOP4eBYf/NZRvPNlO9DfHan6nErjdP+9yB6BrRQuY8ibDrb1RFC0XMwVLPzez+3Gb926taV7s1lDfiVzvdK4YV2ciGlM3JvMBDEZsd5U51I4LsMxTxh28NwCBmfyDa/viii4aZsQc7x4a1fDiVytIpgCIUueQVF5HPSyS8jJsqXz0OX4p58MIaPb3qQLz7Xa6xHkDTeoc4qaqy+GAkA5LIfD9qYx1vMXchlH0azf9yhFtlhfwHPtpgR6YiqeH8uWnSWMA47FULQM7OiLomA62JgMI6PbYkqmRxgPqzKSYQVzeWvN1zA6WD1cirX8xfjhLJ4+Pw/LZZApEFaE+V5Mk6tyNMY4Pvb9E5jOmtV1Jg5MZ0186D+PgnEgb9hin+MAqBBbGLaYGqV4Jm7N0OrJzPniWiyYDvq7IxhdKCIZCsHlPHAaBgcmsyb2egL+pZ5XHXORtYlLcR2uFPz47Klz8xieLyAZUrFrfTQ4X0v77JRyRFWK3kQI23qiGJkv4vhEtizndBnH4dEMDo9m8NlHh7AuruHmHd24ZXsP9g+kEK5jWsE5R950kDcdSJQgosqIh+Qqk4vm8aKKhYKFcEiC5bCaa+yua9fjC4+drxuPqJQgazgIKxwD3ZGaNex7Hj4LgAd5p2Ezb6IfDwRljssAiOmZ/lSe//nK3Xjg2GTNmC0eEjGBHzPo9mIMQsnixA6gvGZfWtogAJJhuUqEUsl54SbHbN4KHlu3XSzIYg/3Rej3Pz+B7b1RHJvIQpUIFgo2XM/sgnh5ECjH534yhB290bZizit5zXXQwZWEVus1t+3qxdmZPB4/OwcCeH1TsXeXThT1z5s6A6VaRtFmmMkZANZO/hVRZdz/+3fgkw+exjeeHQ3e67nZQnCNTBFwIy9EuyOmSTBtBpeLKXhiqquoi88VbKR1G2FFRm9MRd50y2qXteunLmbzFjSJoqdbxULRRkhpbmICNDew6aCDdnFRxR3T09N4y1vegomJCSSTSdxwww144IEHcOeddwIAPvnJT4JSije84Q0wTRN33XUX7rnnnhV57pm80fK1W7vDKNhLG9PWCK04AF0qWE3SaQcd1MMDx6Zavu6NN29b3RfTAKXNoMcGZ/ClJ16A5Swmzn6DS6bCzbDRWLBK+EXGmZy5OCob/shOhpmciY99/wS+/Z7by4LPdpxEr9+SrEseW601Xu/1cc4DBzbCOVyXY6GFvTljWDBdBslxkdVtIQapQY6oRCIkY19/Cvu2JLG/P4VtvdGA8PJH3ziKjO6AcQ67xE1DkShs18G9B0ewrz8VFJXqgRACRSK4Y886/PzV63F2poCcaaMnqnXIeB78APhn5+dXTNghEUCVCExX3FN5y0XeEo0yAuH+KRHhhOkwoJKTTiAIJpSIxrhhu6s2dnW1ncPvOzKBnG5DlhaFHT4ooZAlhpxu474jE3j9izYv9+1U4bZdvXjTzQP49MNnMZ7WA4JuPKzgTTcPdOKISwAvzBeaX9TGdR10cLlgKeTItQjin73+Bg3UdlrhgCITFAwb500HRcsNRucCzcUXopHEg2LoWEZfNAslYry732wqfUmo8d8EgmBoOm7gsFj6M1WSGroLXkhhzlLrArUIZxndwmceHerk5R10sASsZF2rVeHWk0Nz+MDXjyCr25BoeZN7NSFTErhXMcIxkdYxOFVAV1TBx75/ssz0gAB444F+vPW2bVBqnGeEEMRDMrojaid3XWEQQvAvbzuAvGnjh8encHg0gyOjGRwbz9SciFVKbNYdhuF5PZjqEVYpVIkgZ3JMZHVMZ8MAAFWmCCsSIqqMkCIIuO1Onmm0dm7Z0ROQsPu7IihYLmZyJgy7nHzq18dqgXFgIq0jrEo4sL0LH7jraqR1e80KGBnjuOfhQcznF+O7IJTgfr2Ow3ZdnJjIghJaM95wOQchBLbrQiYUm7simMjotaehcUAmBL0xDScny+sEKxVj+J/10dHMGhAvE8+4hXv/Xlv3QD0s5ZxpJjBo9h0DCH5GCAnEEO0YRzqMYzKjw3E5bMYhESHmJiAwbBdjCzo2d4UR02SYrpgo9O6f21WXCFY69WK+YMFlHFnDCOqqEhWTBT5y33EkQjJUmZZ9Tn6cHlYkZHQ7EKmVLDNwT3BuOoJgSylBVJXwoq1dbe0ZjRryK5nrmQ7D3TcN4KPfO46JjNFwqnMlJrMGnj43j4Pn5/HccBpFq74IQKIE125K4MC2bty0rQs718Wa1q0bgRACVaYIyRSaIiEkU8h1Po9SXCpOlq2chxwcowu6qBNL5UIWiQgHbn9qpMMAmfKS3o0gAhFe21CoFK2GiFmjhlNsCfKmU3ftc+/nrsuxPqqgK6rAsJgwevAI45wB03lzzdcwOlhdXIq1fH/fuffgMP7uoTMwHYaemIqQLEG3q6cHHR5J48REruF6OTdbRMSLn02HgQFgLkAhSGQciyLrVkEgzsLSsNj/O1Wm2JwKI6xKwVq869oN+I+nh5ExnMVpSHZ5zpnW7WWdV43MRTq1pouHS3EdrhSod/+tT1Qb9tw40IV9/SkMThWQMawq8UdGt/Gz8wt46twcnj6/gIxefm5O50zcd3gC9x2egCIR7O9P4ebtPbh5Rzc2p8I1X4/LRLyWM2woEkVUkxHVJGiy1FK8SCnBW27disfPztVcY/GQgn994oW68UjWdMCYmGZRLz87NZkDBw9yuJAq6gBFywWB6JEziD0rJEvBVJ67Dwzg7gMDNWO2R07PwHY51sVVdEUVpAs2JrJGIOxgJU5O9XZCQgACig0J2ROhDCJdtMs4L4zzsloZ595ETG8C2uauMFIRBScmskhFVGR1u2wPZSUCO1WiWCjYSzIjuJLXXAcdXAlop17DGMd/PD0CDpSZppLSfc/bekrrFMvBkbEMXrl3/TIfZWWxKRXG//uGG/Ca6zbgn35yDk8MzZX93GEAIaKGdCHUHQVvIpQiCXFhpcCaMUC3HIynGfriWlntsjLenbFcGLbXW1YAh9koWA6G53WsS2iXvHF/B5ceLqq44/Of/3zDn4dCIXz605/Gpz/96RV/br1BkbMSeUs4X60GYfOWHT2IajKeG0mDcGD/QArXb05eUgt/pUmnnQkgHbSKsbS+otetJvxm0PVbkti3JVXVdNuY1DCdMzGRMdpaR0fHMjg1mRcNBLrYQPD/bLsMpybzODqWwb7+VPB77Y6QvNBNl2aOZZkWHMtyho0joxkcHk3jqaF5ZHQHGb2+exUARDUJLxrowj5vOsf23mjNJtfgVAFnp3MwvZGuEiXwjblMh4EShrPTOQxOFbBnQyz4PV/I4buXqVK1g9mLoyt/1lwuGM+0vpZ91wwfEkRxSpJEkb0rqkK3HDCnOpvgQOBcEpIpGGfBuGsfQgjhP4cYyXf95tZGULdzzl0I5/CxdBEMgFzn1ykBXO+61cDjg7P4t6eGQSEKkC7jXnOe49+eGsa1m5Kdwtgax0KhcfO43es66OByQSkZRKO0ihhRSY5cq9g/kIIiUTguA6U8iMsqIUmeG7XXZOmNqcgaTiCoDYQaWNSJVD4Mh2jccM5hlpAXCQgIIZ7jinDBQYPH4wASYRnzBeGOLRHiueWIhlVEk3DbTkEOq3UOr5Ywp14MsNRYu5Rw5jujr5YYtNl76GDtofNdtY7VEFO37H5vOiDwm998RcTs/tdc+VilbmK+OyOFyJMeOzuDbx8aR85YzJn7Yho++AtXY39JLaEUYVVCd1TtTGdbZWiyhBu2pHDDlhQAIXI8M53H945O4LHBOWR1u2aPzJ/skfYIK76o59REHtt6ouiLa7AcBsthnuM4QUih2JwKYXtvFKem8k3rRc3WzjtetqNMCBDziC5zeQvTWQOECEJrTdFoCSyXwzUcvHxPX1ltay3i2HgWJydzwYSSspISEc1eDvGeJIdDoqxmvCFT6pEDhXlLPKRgV18M80ULc3kriK+o56buT0/gnJfVCZrFGJVnxd4NcZyYzAV/ftmu3rKzo1mMpFKCWdvFI6enAWBFz57KSQ8RVQLjHCcmsisa66wGaq0V03FxdDSD9/7HIfzuK3fhhi2pmsKlZo5/9b5jQNyPv3P7dqQLNsbTBXzqobPBdNpWBYWEEBg2E2cIBRRJguUyKFScKY7LMZMzEVFpsD80IoL5r+v4eAa6LeqqMiUgBHA4DwRMnDAkwgokSsrOY9ubIChLol7se5SUkmGDnIADtuuiYLEqJ8Xlol0hXCUM20XedFAwHbiM46qN8eqpzoRgR18Mdx/oD6ZDm7aLI2MZHDw3j6fPL2B4vnGtbl1cw83bu/GSbd140UAKUW3pLWFKCEKKhJBCgwktlSTClh/rEnCybKV/kvAmWSgyqUmo9B1bEyEZtivOZV/IFFYoeqKa58K6fBCCht/H0bEMxhaMknp2mX8DCAHmC2JCll/DqMSlUsPoYHVxqdbyKSV48y1bsaM3GuRoWcOpKXx9ZnihbIph6dIq7TnpnoBQogBh3lS1kudkbYr3/TUpUxI8vy98TIUVT9S6uBZv39WLfVuSDYW8R0czbZ9XlfHhLTt6LglR3pWES3UdrhSiqgQS12BYLnTbLcsnKSFlfflSJMMKfn7vOvz83nVwGcepyRyeOjeHJ4fmcWa6fOKZ7XI8fX4BT59fwN//GNjSFcYtO7px8/YeXL85CVWuPidtlyFdtJAuCtGnTAhkiqbr7/ZdfXjXHTtrrjHGeMN4JKs7oFSYbtSC5k+44whyOAKCvriGsQUdLgdkImpRqYgKzlE1ladWzFbZ99BkKgi1DiuLNeQaru2+AIR7uS4hUiDQyBpOwHEBAdwKcq7/J4US2IxjJmegOyIc1W2XoyeqYiZfXa+XPbKvyxwcH8+0bUawEmuuU6ft4EKic7+1jnb7AsfGs5jI6KJm500F8sW4fo/Sj+kIAFkmXm2ifbMNH9PZlckZVxqUANduSuLPXn8dvndkEn/z0OmyGJpzoAYVq23U+8wqe8Lg4jlriWs4RP3IYQxzeRMxTS7r71aKwQk4emMaNFnUwBzGYDouFgomKKVrxri/s9avDFxUccfFRCsOMj7m8xau3ZxY0eIvcHlMu1hp0unl8Jl0cOEQb7EZ0ep1zbBSB2Nl0y0VVvCXD5zERMZoex09N5IWJEGpdgNBkggcl+G5kXRZA7xVJ9HS57tQTRfOOQzHxd0HBvDR77buWJbVbRwdy+DQSBqHPSfBZrGiRICQKoExju29MfzNb+6DXKNxUYkF3ULedEoakIuiGhEUivGmOdNGLCQLMYdcLeTooDbm8iY0RRIFuiV+XqVBOyFAb1xDwXREk5jUdgupLNK7TDis7VkXxfCCDtNmsL2ExHY5HFcka5Ikiuu1VNmV+0a7jtoXwjl8cyoSkEYIeOBy6CehzCPpbk5FlvT4jeDHETM5A7rNygholAhC0nLFKx2sPliLnapWr+ugg8sFPhnkyGgajivOFH9/VSUKWSK4YUtqxfPMlcb1m5PYsz6GY+NZ2A6reTYTiMkaFl88YyOqhERIwQse4chv5PhksnpxWiIsQ3cWp3P4Z3nZk3kQo31p8LkCHJYjnmehYHniYEEUW/wdAsYYPvvoWfzTT4bQ3x3BH7x6D27f3Rdcs1yyVi00y3WXE2tfCDFoK++hg7WDznfVOlZz/fi5/9GxDA4Np8EJcGO/MHXx4/xkWIFuuwAE8d4umU605PdU4wEoAFWhweQAzjlcxsVzOi6+/ORw2fUv39OH9925G/GQUvVYikTRE1MRUa/YsvJFhSxR7N2YwN6NCbz3To7TE3mcms5iLG3gh8enkKkj9nAYh8M4vvD4OXzh8XPYmAzhhi1JTziSxKZkCLrFocPFf3vRZnziwdMYTxtIRRSEZAqL8bJ6EYCma+crB4dhOQxdkcU6CwFBVJVBPTI347yMJFcLBEBIkfHomVm8/fYdFzw3bKceOF+0YDslstNS5mwJOOegXkxYK94IKcIB33RcSH7NiRD0RDWEZIpzc0UoEkF/VwRhVQrIfrXqBPVijMqzgnEOlwtXPUpozbOjUYyUNx1MZgxYjovPP3YO9z41vKSzp9bnDaDupAdVorBdvmZrB7XOmbzpYCZnwHQYMjrH//Nfx6BQgrAiBFDtfm6V33GtOGBdIoSQQiFR4k36ae20CYTUREzG641pGE8bsL0zhBLAsB2MpY2q2li92Ha2YCJriGnIvlkRBy9z1mUccBlDVFPLzuM/ePVVUDxhBzj3ppEskuh9+AKrjOHUrdktB+0aJwGLgo6i6cKpUR+p5Ta9c10EY2kD33x2FAfPL+DwSLrhNGpFIti3JYWbtnXhwPZuDHRHllxblSlFSPGmcij0ihNzttI/edmuXpyfLdTd631IlGBzKgrDLjF8UCgMh6HpIei/Ho8oWY/UolCKGxsIIEv7SJSQgPQiRCHiPHZcjlRYxXTOuKRrGB2sHi6HWn4r5hpT6UUCXeUWWmkqpvj9QZnDdspzuSXndSVP4BuszOQtzOQtKJRAU2iwFiklDd9Pu+dVp5aw9nE5rMPlQpYoEhJFwqtXWA6Dbrsw7WqxRz1I3nSMazYl8LaXbsdc3sTBc/N46vw8njm/gEKFUfDogo6vPzOGrz8zhrAi4UVbvake27vRF6+eImK7DOuTGjZ1RTA0U8D6hFbGJ6gXL1aiWTwS1aTApZwSXhVrmK4wE+DgZTlcTJOxuStcNVmz1ak8tfoeLhMiEdfhZeJRQbsQQhXFj0MAMPCAk6FJFKYjahaKx3nxawWVpF7fdFOm4n6fypkAhEFJrRibEtEPAAdsxpA1HMwWzIbvrxQrseY6e2sHFxKd+611LKUv8NjgLLK64Im5XqBGieiPlsaJYgI298S+i8IOP69rhzFRtBobGF8sEEKwIRmC6bjY0hNGTJMQUiTMF6yqyRlLfg6Ic6TWtEs/DZc8YTQHysUlFdc73tFuOAwutzFSwyzjgWOTcDnHlq5I2f3Q3xXBZNbAlq4I/terdqM3ql00IYVfN31scBYPHJvEdNborPXLHFdsF64npgHItXQtAzCVM3HvwWH0d0dWRO20Gq6AFwMrSTq9XD6TDi4cXtSfwrMjmZauWy5WOggubbodHc1gaKawpHVESqWotbYkXnFdCZo5iV7I9ea4DEXbhW6JfxjnuGpDY8eynX0xPHZmFodG0zgyksHZmeZiDt8NIqbJiHmin7zpIBKW8Pbbt7Uk7ACEkIQxEUiWfmeCfEhAiQjUZUqxLh5a4qdy5aJoucjoNighiGgS4pqCsCphU52xt/XgJ1IgQFxT0BtVMZY2sKUrjLe9dDv+9L5jKHqpU1WRHmL5uIxj1/o4Tk8XPCcOocL3Fd4cwNauMP7sddVnVDVRgqFgudBkcV+0cs6tlnN4KX7pho348HeOIV204ZSMyS1Vs6ciCn7pho1Lfo56ODaexeGRNApWdUbGOFCwGA6PpKv2v44KfW0hrCmA3nwqXlirJgd20MHlDEoJ7tjdiyfOzgVnCIEo3BUtF5L387W+f1FK8MHX7sV7v3pIEOpqFOWU4Jxa/JnLgHhIEIGKlhv8pB4XyS/SxTUFhZJipU82WbyuvKVDfTVi8Lfi/y0vZpdl4RDmP69EAN3msF0HnAMLRQvv+NLP8P479+Cddwiy6lLIWo1Q063ZdXF0LIP3fvUQfu+Vu3H3gYEl3wsXQgzaydcvHXS+q/aw2uvnyaG5spxApsD6ZBi718Wg2y42JkJYkCUYtgtZIlBA4TC2IhM8SsEgSLIEJMhzCIRovfSpQgrF/3zlbtx17fqqz0OiBKmIikRI7pgWXEDIlGBjMgzTcWHYwiXMJ6tQQnD1pjiu3hQHANy6oxt//YNTyJsizrBdBt1iZROvfExkDExkDDxwbAqAmLjlCz32bUniva/aja88PYqRuQIWPAL2zr4o/vvLd+DWnT14fqz52pnOmgCpdioVJGEaTNb2eBZV/FaJ+FNthPhzuWfZUtBuPbA7ogoHd2txSke9ct2O3ihiIQUnJ6vjDUDEVxwcGcMGoSQg8czmbVAAG5PhmiKrVuoElWeF5TLh3MrERIRNyTBUmVadHfVipLzpYGyhCNvlCCkSNiZDsF1e9+ypl9PX+7zvunZDzUkPHKIpSwlbkgPrhUDlOSM+Kx0u58F7ABcNaN12EA1Jy4oR68UBQzMFFEwHW7rDKJoMM/nWCE2+SIISgr54yCOCkUCc4pNON6fC+NAv7G0pxkgXbDDGPZGXT3BDFXnLb96XnscAsHNdDM++MC8cIP0PsAQEIufQZIrrNiXx7lesfJ27VeMk02EoWPUFHVWPSwi2dIcwM2zgu8+P4+lzC5hs4tLZ3xXGTdu6cWB7N27YkkRIWZoIQ5Fo2WQOpU4t8kpCs/5JVJPx70+PBFMuScmOz8EDYYTtJaNhVYKYL72YV0Y0GUWn+aTbsExRsGvfQxzA5q4Qrt9cf/8r7SMRj3hUdkB5a33nuhheOFkMzgOJijV+KdUwOlg9LLWWv9bQzFxjQ6q1nl7pxByJEEAWU2/9XC6qSjAcVkYyawbGAUWmIIzD9khqpUR1l3GYDsNAV7ilKV/tGP21UkvoTPC4+Lhc1uFKQpWpmKQRLhd7GPYi56AZemIaXnv9Rrz2+o1wXIbnx7N4amgOT52bx/m5cvKnbrv46eAcfjo4BwDY0RfFLdvFVI9rNiW8iQ4irvvNm7bgL+4/hdH5ImIhGfGQDIcJnoE/IaOyblSZczaKR951xw585tEhHB7JwGWsSpgpUYobtiQBcJycLJ/OGdNkRFSKsbSBzakw/vz11+P6zcmW1nRl30OWCBRJkGu9QSGQiNizwoqEZFjBTN4Uz03ERI6QIiGkinjTdFkwhdZPon1RK6lQd5CSfzMGOJwhpEgIaxLShZJJ4L7fgvcPJUJYwhhHutA89vKx3DXXqdN2cCHRud/aQ72+AIcQy6kSwcmJHI6OZbCvP4XHB2fxpSfOC2OUkomowmSOo3T7TEZkRFQZ01kThi3qn4SI6buqTDDfxj601qHJEnb2xhBWhLhja3fEE9JZLQkum6FRKaWm+QGprjFXPSbn+NxPhrCjNxqsieZ9IhXTWQO9Ue2ixVh+3fT4eAYLRXEPabKEdQkNqlRdx+3g8sAVK+7Y3NU6QbQ7omChYOEj9x1HIiRDlemySN0XylXzQmClSKeX02fSwYXD+mRrBbZWr6uH1Q6Cl7OO9g+koEi0bgPBZRyKRLF/IFXzsW/b1YsD27px35EJjKWL2JyK4Jdu2Ai5xjjRlYZhuyhaLoqWA6sOy6/UsWwiW8RUxsJ0zsA9D5/F0EyhqZgjGVawrz+J/VtS2NefwkLBDMgJecupOd6+FXRHVFCP5C95n3vQjPQaN5JE0BXpEJnbgWG7uPtzT2J/fwov3dWLge4wnj2fQ8aw0B3RsKM3CpnWJ4X66IrIyBkuGONg4AjJEjg4pnI2UhEFf/wLe2EzXk5wqcO4oITg7HQeEVUKCucEIjmTJQqZEiRCCkyX4ehopoyQULpvKJTg/FwRps3guhxOlCOkkKbn3Go4h1dClil+8boN+LeDI3WdpX7xug2rsi/M5AzkjHICb/D83gvIGcLREhBJUsdxYu2hN6ZgNN18JGhvrLMndnBlgTGOR8/MQpMJdJuXnV+UAJpMLpoDdbu4bVcvPvnr+3HPw2dxZCyNrC72buF8JRyAS0es+w40hs2gSBSM1xeAqZ67GIdoOqkSgVHycciUlsW4pWcFYwCjPCjWOSWO92FFuDg7jC+Kb7kQfVDmQpaocChlHKbN8NcPnsbejQncvrtvSVPu6qG+W7NwRssw4CP3Hcf9z0/g3a/YtaSzbLXFoJ18/dJB57tqH6u5fqrI0w7DTM7EVC6NIyNpwNsnk2EFlsPguBwSJZBI7ekby4XjAoRwEA4Qb98uxdUb4vjQL+ytqpkSQpAIyUhF1ICk0MGFQT0CvO0ymA6DYbswHQbLYeCc48aBLrz/1VcFJhkgFGGZYn0yjGs3JZDWbRwZzWAiUx27z+Yt/OjkNH50choAkAoruH5LEq+4aj02pjRctT6BPRtioETktoMzeZgOQzJc+57QvDW1IRHCZNYsI5EQQtAb0zA8XxQiTS/5K/WhUygR8QXn4F4ObjpsWcYG7WIp9cC9G+LYnApjoWgF8YlPWAcW82xNoviDu64CJaRuvNEdVfCmm3fi0TOzZSSeHX1RjMwX6+5bzeoElWcFCIJ7QpUJHAbMFUxs64liQ0KrOjsqYySVEkxmDE/AJlz7KCXgDkdUpUjrNu55ePH36+X0d+zuxb89NVzz8z41kUVadwAsTnoAfLf6pTmwXiiUnjOcc8zkDLhciP3sEtG0RMS+POV9FxnefoxY9d0CgXtvIiQhb9qYzppCVIHmjuIEQhxIAKxLaIFRTkyTEVXFFIKC5cB2Gf789ddjX3+qJTOOrogCSgWxint11YC8VfLcpWeOJlFkXI5Dw2nIFDXJVT5kStAdVfF7P788AXUz1CPa7Vkfw2/dshWbu8KYyOhNH4dzjsGZPO4/OoXDY2mcny00jAPCioQbB1I4sL0bN23rwsZke2Y4gNiHVZkiJPuCDqlzxtdBo/4JY7xsyqWfY5bmptt6ImCMYzytQ5ZocKM7LkMirGBnXwIPnZxp+jrIMsuyrfaRprNGWR3cz6cjqgSJ0kumhtHB6mAptfxLES8a6Aqc8IM4zicJl1zHOOBXwrgnOJQlCsdhoJTg/XddhZ19Ubz3Pw5hrg0Cn98vVWpMvvXP7288N4bXXr8xmELb6PxtxeivlVrCx75/AsmwiqGZTl/mYuJKWYfLgS/2SHpiD8MXetguTJs1FXvIEsX+/hT296fwrpfvxGTWwFND83jq3ByeG66eojY0U8DQTAH3HhxBTJNx07Yu3Ly9GyFFwn1HJmC7DIbjoph1MZs3EdMU7FwXw2/fuhVFy8Gffuc4CpbbMOdsNHXoxES2TGRRLsxkePmeXly7KVk370yFFXzoF/ZiXxtmqX7fo2bMoBAQQtEXV2HYLmyXI6xKUCUqJtd6HIq+uOrlAULwuqM3ihOTucVYxfdzqvi6uPc/l4vpk36uQiBqCEGeUyoS4QAnPBCYl/I3muUvy1lznTptBxcSnfutfdTqC/i9O9MRXCMO4EPfOor//dqr8ZlHh2C7DGFFgukI8Ydv0sQh9l5KgIgqQaHC9K4roiAWimAyoyMZVpAIyTg7U2jrdUZrTNZea7h2UwK71sdxYiKHdXEVXVEV8ZCMdNHGfNFacq+jlV8rFVJ3R5SW4u51MQ0F0y1bExfCdHc58OvUOcOGbjEQImqJtsswkTawuStcs47bwaWPK1bccWN/F+59aqSla7OGA+45KSXCCiRKlkXqvhCumhcKK0U6vZw+kw4uHJ5+YaHl6965xOe4EEHwctbR9ZuTNRsILuPCVZkAu9fF6ro21WqofvO50VUpxrmMQ7eFmEO3mo9FzXikg8MjaRweTbck5kiFFezrT2HfliT2D6SwtWL0/PbeKPYPdJWNt9+1PirIBHWgSMJRUpXF+HlVpiiYLpJhFRndhssAEd+JQq/rFS8SIcWbEtVBq3jszCyeHU7j2eE0vvDT89BkKpw6JYqQRDDQE8OGZAijC/VJ5KpEsTERRkR1MJ0zhKOjJBxBS4vVX37yBeglbmd+oaeScBFSJEzlTMQ0GfMFS4yBhTelBYIw+vxYBu/990OIqBJ2rosFTiml+4ZuubAZgyITuC7HZMbAuoQGxRuNW++cW2nn8FpgjGN4QUdUlQIXTB+UiKbx8IIeFL1WEs+PZReJwBUPHRBhvOteuXd9x3FijUKTW3OFbPW6Djq4XHBsPIvj41kYdrUDO+OCaHV8PHvJ5Dh+E+nbh8bx/3z7+WAUsHAc5ij1ficEmMlbcJggKzdC6Xhey2EYyxiIKBQxTULBdIXTPCknyQDijArJBJa76IiqSATM5VC9Ee9iwDsAzoPvgEMQY/zmCyUEiiye+69+cBq37RROpCs15a6eWzPjHBKloFQ0lp4fyy75LFttMWgnX7900Pmu2kfp+tEohWEJEqxMKUIqXfL6qawlFCwXExmP0CsROI7YhAzbheMy9MQ05E0Hhu00FbMvFamwDIdxFCvycQLg7psH8NZbtwrSYQkiqozuqCpcMDu4oGgmalckGhCtORdkCtNheNnuPrxkazdOTOTq1j2mswaOjGWCmsvIQjX5OK3b+MmZ2eDPEVXCvi3JYLqHSkW+XrRchBRxDgsnTLHfmC6DIhH85oEB/NNPhqpIJHnTRV9cQ19Mw5npfEBiI1icCGY5LCCWT2VNUEIwMl+seq2rgaXUA/3vbHShGEz89FH636pE8P5X7wnIeM3ijbffvqOMbLJ3Qxxv++LTVXUCDg7dFISh7X0x7N0Qr/neKs8K3XJhOsKRnRICmXJPPMQQVqWqs6MyRpq1XViOi5AiYYNnsHN+tgjTcT0iDcdT5+bwlz84hZfu7MGf/OfzNXL6LJ4+Pw9VohgoqeX5n/f5uaJHPCY1z7elOLBeKJSeM5wLkZLs3TPcIyMBEA623v6vSJ7Qr80YsfS7LVguprNiugbziiscgG4zSATiLKqY3lSKkEyhyhIiKkVvTMVk1vJ6ZIsirZBCkdY59m5M4PrNyZbNOHpiGhIhBRnd9oSNqJjEJ9yAZSruT4cxFEwXGd3C3/7odFOHSw7gL3/tBtyxZ13D61YCfo703HAa4xkdIVnC9r4IKKkWUZYiq9t45oUFPH1+AY+fnUXWcOpeCwhH6APbhJjjus3Jtqdq+N9XSBZCDk2mnSZ/i2jWPymbcskW3bIpJeiJqviz112P7x4Zx78/PYLS1FgiwGuv24D9A10tiTvyZuMAcXReDxxla6Gyj1Qq8vJJS1u6wpjKGoiHFKSLFnzaOvFWZjx0caZodbB20G4tf62iGZH3+s1JXL0hHrzfgCRcA4bNagomY6qEl2ztxvVbkvjUb96I//7lZ5A3m0++BhAISyqFHT5kSoIa1i3be/DvPxvBVw4Oi6l53u9Xnr+NiOFAeQwBIDh/Za9/pcoExydyiGsy+uJapy9zEXG5rMMLCV/ImvL+bDouDEsILgy7OV9hQyKEX9m/Cb+yfxMsh+HQSBoHz83jyXNzGK8wPcubDn58agY/PiXOdokSxDQZ6+MhOIwhb3p58k1bsHdTAn/0jaPIGg7WxTVIVNS/Q0rtnLPWlB5fZBHVJDguh+WKWjwhQuSiSASPnJ7B22/fsSJ1bh/+nlErZiBE9Lwth+F3X7kb9z8/gaOjmfL+M+OYzpqwIiL3jGkS/uCuq/Dx+0+WxSq1vhnGxVQ0DiE6VyQSEHEVScS4vgmVP7mMg3uGJ+X8jVbyl+WsuWZ7a6dO28FKotMXaB+VfbXy3h0BkYTB3OiCjg98/Qgsh6E7qiIe4sF1YuouwBiDy4F1cQ1/9Wv7kAyrVXW8wyMLmMoaZT3RVrB/S2pV3v9KotQMZjpnCTMYiSIWkmEz5k3VWrmmR2n87ZsNKhKFYbO6da6y3ycEqUh5fruUPmsrBicrgdI6dSqsImsUIVMqavHgcFyOmZyJaG+ks9YvQ1yx4o7dfbGWr3W5IIe4TIzmiSryskjda13t1Q5WinR6OX0mHVw4NBtJ3u51tbBaQXDpIZ8KK9jRF60ahwk0X0eUkrIGgu2WkwcpRGL55NBcVWJ+IUjSlsNQtBwULTcYN1cPmaKNw2NpHB7JBGKOZuiKKNi3JYV9/Uns668Wc9QCJQR7NtQ+A3whhy/iqNfounZTAtdsSuDIaDoolnAmPmtNppAlgms2JZZFuL8S8f3nJ8v+HLif2Aw6JciOZ6DJBHFNiBDcGsVtxjnGMgaiqoSbt/fgNddtQH93pCyYfnxwFp/7yRBqlcZL/0aiwNaeCIZm8pitSDYYBwyPkEWJmBKjymLUnUjuXHRHteB+dLzmHrj4Xd12MTqvg1Jxz/RENdg1zrmVdA6vB3+f25QKQ5OFq6btCqf1VFiB4bDVSwBKj31fNVP655LrOo4TaxetjpZvZwR9Bx1cDpjLm0II6t36lVucy4WYdS6/9hyGAfHaLIchpskIq6KQRSnBr+zfhG8+N1ozDqLEH0EMWK4LwqtPW/+aSlCIaykB3vPK3djWE8H7v3a4Jkkm4ZFZ86YwYvCFGjYTD+K4HC5jkCiFXCp89uC4ACAa8L67CSEEw3OFsvOuWfO7FZS5NUMU2JhH7vZdihk4kmEF+QqnmFax2mLQTr5+6aDzXbUPf/2U7WnefqNKIre7YUuq7fVT2cD1ndoVKvYbInM4jIESAodxZHQb23simMgYWNAFadWf4LFSEVRadyBLpIxYuC6u4YO/cDX2VTSJFImiJ6Yiol6xpeOLinbrNYK0Kwgr8NxJt3RHYDrCmVQIPxZJK+sSIbwqEcKrPOLBfMHCkdEMjo6Jesy5GuYaRcvFE0PzeGJoHoAwAVAkgqxpoDeqIqRIZQKS+YKFvRvieONN/djRG61LIrllRw/uPTiMv33oDOYLphBFcIbSFFymRBgtEOBzPxnCjt7oqpPH2q0Hln5n3VENMU3BZMaA6S6KVmSJYKA7gg//8rWBsANoHm/UIvFU1glsl2E6K5wFAWBkvoC3ffHpmkSdyrPCr1f4rvAEnqCWMQBSzbOj9DU/cnoan3/sHDYmQ9BtFjS4CRE1GsYB1+X4zCNn8aUnzkMipErAkQwpmC9YoDVGqhJCEA+JJrvLeDBB1wdHbQfWtYLSOC2qSsEZU7q/Uy9e9UFAQNF+jOh/t5bDMJbW6+bgLhdEp0YCj4gq4drNSfyPl+8EgKZ1qSeH5lret2rWVflic556ou3JrAHLZWCMw/Mwglvj9cqUBM6+jHNwzjGXX32hj+m4KJgucrqNmZyJouVAobVjMJdxnJ7K4eC5eTx9fh4nJ3MNnSsJgJBC8cv7NuENL96C3jYNhCTqnQuyBE0RNe5mNfMOqtHqeexPuTw1mYPlMqgSxVUb4nj3K3bi2HgGX3l6pOr7djnwladHoEi0pUk6zWAzjmdfWKgr7vD7SO+591mki3ZZfkwApCIKfvXF/fjHR86iaFrgQJBP+8K06ayBiCZ3cokrGW3U8tcqWiHyVvZdGeNl57ZECRRKYHrk4lJiGbw/Wy5DRhdr5fbdffjHN78Y//PfxePVguLxUDSZYl1Cw/Bcsepx/cd2mJhqd3Y6h7s+9SjOzxXBGA9y2J6YWvP8rRVT+iiNISYyunhvQV7scWQYRzIsi5wDnb7MRcNlsA4vNjRZgiZLSELkDqbjwrBZMOGjkdhDlSkObO/Gge3deA/fidEFHU+em8fBoTkcHs1Uxd+uV+/J6DYkQhDRJBQtF19+ahhRTcHIXAGJkLJoaBSYJhAkw81FlX7eui4egqZQzOctzBctrzYu/jl4bgH3HhzGm2/Zuuw6t4/5ooWC5aJoOg1jhpxhI120kTMdcC7ifB9F24WRNXDNxjg++Nq9uG1XLygheO9XD2E2b5ZNOfRB4U9El5AMK/j4f7sBn31sKKiJCzEzDYQkpUukkr/Rct1lGWuu0d7aiBvQQQdLQacv0D5K6zXrE6SsdwcuJqiFFYrNqRBGFgzotoP1CQ0hRcLmrnAw4cPvT8qU4K23bSur+fkY6AqXmei0g13rWuc3X0zUM8y7bnMSb7ypH//46BBOTuRWhC/CK/6bAeiKqpjNtdZvn86Z2CzTsj243T5rqwYnK4HSOrVd0r8C4E2uWhSvanJnrV9uuGI7dG3dxH7gRwDZK84uh9S92q6aFxIrRTq9nD6TDi4c5BZnUbd6XS2sRhBc65DviamQKJa0jvwGwp9/7wSOT2Q9ZSoJFP8TGbOqiLeaJGnDdlEwhaDDduurb8vEHCNpDM22I+ZIYX9/sqwZ3C4UiYrmliSaXKrUumNZ6d6XNx10K2rgzqDbLmKavGzC/ZWI33vlLuzoi+LTPx5E0SoXAzlMTMkwHZEYRTUZGV242lEAmkKRCCswbBchheLdP7cLdx8YqPoO/Hu/YDrY0hXB6ELtZjclQF9Mw5auEI6NZ6t+XpYwcPGafDeVkfkiiraL9fFQcI1MBaGz1IWXekU63WYYz+hIhOSa59xKOYfXQ+k+RwgJCGg+VjPZ709FFonAWHRRAcpJDv2pSMdxYg2jVcfIdp0lO+jgUsdcwQrOGBL8n4AvenAYx1ydxu7FxLHxjBC5EoKcYUOmFFFNQlQTTdx6cVBGt5A1XNHc5kCNHkxNEpNCCSSJwHYY8qaD+5+fxJd+50BdkszL9/Tinx47B5jiDPCPBe5N2OIcUGQSEP9EOFBb1Mm5IIFJRBQAawktl3OulLk1Owicsf3X5tcaFIkiFaFLOstWWwzaydcvHXS+q/ZBKcEdu3vxxNk5uEw0byQq9qqiJdbrHbt7214/pTG24ZHrZbroOC9IqKIWkNUdWI6L4YVikAfJ1C/Kr5w4lgNlzfFX7OnDe+/cjXjJaHfq5QOJsNwhgV4krFS9RqIEEVVG6XK3XU/oYbswHeGaxjhHd1TFK67qwyuu6sNzwwu458eDGJot1jg5F6HbLjwdEkY9t9KQV1thnCMekvHfXrQZIws6dvTF8Knf2I9zs0XkLaeKRPLmW7ZiR28UH/v+CeHS6eXNviCCc0EW2ZQKLVkI2S7aqQfW+s5CioR4SEbRdDCVs7A+oeETv7YPN/Snar7uduON0jrB8fEs0kVBxA3JknBT9swnaomBKs8KmS5OSBMT2cr7IPXOjtLXfO9Tw7BcFjTCKSGi2VhyPedAwXQhU6BgucHkGUCYaxGIe9SfGFKKsCKDEBMECCY9+K95rU/QLY3T0kWxaFjJzDsCsV79aXe++HgpMWJ3RIVMhdFRs2a5wzgkQiBLBDKlSIRlxDQZpu1Ctxn+5Bevwa/s3xTcr43qUrfs6MFb//lgy/tWvbqqbruYy1vgXJxXNoQIMhDLEyBnLk648AnxjHs1LRAQzmC7HGPp1ZnyYzkMBdNB3nRguwzPDS/g3oMjGJkrwPY+l/6eKO4+0I+tPVE8fX4eB8/N45kXFppO5wjJFBFV5FyaTDBXsDE4U0B3tHnc5te5fUFHZ+LW8tHOeVxPpMcYxzv/9Wd1hTyMA1/52QgiqoSCVd8Yy3fxb4aJFk3OKh/J/3NMk2DYrnCglWmQsxICEEm4ZBu2i1R47QnpOrgwaKeWvxbx+OAs/vc3jgTxqCZLdYWIpcKtk5NZIYriHBFFwrpECJxzvDBXftbIEvHqSwSqJOEzjw4F02Fv392Hv3/jjUEf198XSuNdmRJsSIZAScU6rVFPZBzImy4Ks4XFSbUcMByG8YyBdTGtrbi5O6KCcYbxjAXGF4WTHGLqF+O++LI8Ruv0ZS48LvV1uBYRiD28881yFqd6+NNda4EQgv7uCPq7I/i1F29B0XLw3HAaDxybwuNnZ2sIOzlyXjx4eCSDj373OHKmA02RFifkeXVqBpEfGQ7DyEIRezfGqyatAuV5a8F0MZu3xMRYb0IXA4ftMPzdj84EJgkrsU5TXj++UcxQtBx88fHzmCtYYkKnTAAQYXQCgq6oCsthSIYV3LKjB4DYe//61/bhd+99DhnDFrkRFnMjAhI851/86g24fXcfJIkENXFVFlM7Sj97mYp9inME/A0ALcd5y1lzzfbWRtyADjpoF52+QPsorU2MpQ0YtgNKiGd4ImK6vngI1KuZFEwbOcNBKqIipsmIalJwTviCutt3VQs7Hjszg288N7ak1xhSaNNawlpCIwObWEjGu/71mVWZWs65mJC6PqFhImME8XIl/DOFA5jOmuiKKMGaaKfPeiGMtEtRZiJYWr/13xcBOPOMelx01vplhitW3HF4NN3ytRzCESisSggpi0HzUsmOq+2qeaGxEqTTy+0z6eDCIKK01sRt9bpaWOkguN4hP5ExIVFgY1LDXN5qex3dsqMHqYiCREhGMqxCkYQ7ASEEnPOqJtpKkqQZ49BtFwXLgW7Vd7RIF4UT5OHR9sQc+/tTuGFLCjf2p9DfHV4SwUSVqTeJQ4ydb0fIUQ+Ve58dfGeJVVHjXgnY1hvFHbv78MXHz6ErrMByGXKmA6NiaoZwtvWEHQSIKDKSERkRRUIiJGM2b+M7R8bxqy/aglAFIaD03g8pEvq7iXDLqBjRt703it94yRb81Q/ONH3dpb9HCEEirCBvOsiZIrkDAE0ur8gTiASBEgKZcliOcCHcuyFe8zlWwjm8Hi5msv9LN2zEh79zDOmiXeZsVVqwS4QV/NING/HTobmO48QaRVRpjTTQ6nUddHC5IGMsusU2MnUqvW6t4I+/eRRHRjPYuzGB23f14KW7etHfHUFGt6F4AosP//K1+KfHzpXFQTv64hiaycNwXFgtEpL9M5GAQJYoXMZwajKHY+PZmuefP8LYZRx71sVgOjwYZ+4yF+fmdDBgsSmG6pHppX/nu+pwACFp5c+7crdmWuaMzbkQr4YVEbtzjiWfZaspBu3k65cOOt9V+2CM49Ezs4ioElwm3MNdbxpRRJUgUYpHz8zi7bfvaCv2Lo2xHVbupgQskqjjmoKeiIqxjI5UWAWBBd12IHnPRVDbUX05UCWK9965G6++Zn3ZPRIPKeiOqsFzd3BxsJqidkWiUCRaRqq3vKkepsPw+OAs/voHpzCdMz2iGQAQuK5wAg6rkiB3sdr3pXA7Ffl70XLx+cfO4+nzC9jXn8S1m5LoiiroiYlcPGc6CCuLBOTbdvXi2++5HX/5g1P4p58Mee7DHsFCEYKFmCZDlpYmhGwX7eTJ9b4zQgiiIQUbJYqi6UBagVpUKW7b1YsD27rxq595ArbL0BfTEFal4DXUEwNVnhUhVdTLDNsFKA+cCUVs0vzs8B/v6FgGhu1AorRK2EGJN3WBi+bqTM5AVI0Gr9UXmDAsTgwpBSGAQilUmXgT4i6tCbp+nHbPw4M4eG4BtsMgUSHqKJ1YAXhnBedwONqOEa/dlMD6RAgT2dacCmWJYEtXGGFFCmrIGd3BNZsSZcIO/z3Uq0sdHc20vW/VrqtSXLMxjvNzQuwYELm81yoRsSYXH9wjufLFc5V5n+fmFSQ0Vgo6fDw3vIBPPHgaRctFIqQgTgXR9thYBv/7m+mmZPxESA4mJSbDStX5Gw8JN+fBqULZFGpCSDBxWog5aE2yXwfLQ7vncS2R3reeG0PBdEt+b/Fnfh5q2gzhsIKeqBSQx31I3vPYLoPtNif0bEjUF7gxxvGx759ARrdBAUjS4otxXeEq/i+PnwdjJQlyjSIGX+nAtINLCu3U8tcaGOP42x+dQd500BNTQSAmKEqUoC+mYjpXHTP5Z9+3D43jT79zDFFvvyaEQLdcUAoxdc57Ds7FdLu+eAgSJTXPvv/63dtx78Fh/PNj53BurhB8jqXxbtGqv955xX/7azIwlfHOxOm8iU3JUMtx894NcbhciGhVmQQT+QgWJ0qK11l93nT6MhcWl/I6vFTgcwoSoUWxh267ngi6Pg8iosp46a5eqDLF82NpxEMKdMvjT9jVTNZxzyRhZEEXpoaeyDeiSKCUwHI5ZAJQEAzPFxFSJERVQSb2Yz8/bzUdt2piLCBiZYlymA5bcZOEICaoEzM4rhCzcM4hSzTYVxQqnNKLloP1iRCGZsonaSfDKsIKRTwUhkSFEF2TSdADcBiH6zIkw6KG7+cVH/v+CRyfyImpjt7rISBwOMd8wS6bENJO/rKcNddob22FG9BBB+2g0xdYGvw95M++dwInJkSvlkD07PrioaCGmtBkTFGCrG4H8SABQViVwLmov9X6fBnj+KsfnIblMMgUbQsbUpegAKyegU1PVHyeFC5MlzWcZtoqSutpLuNIRVSkdcFbrLQ18E1iKBHxrem4WJco/85a6bOuppF2PZTVqb3prLrNoFB49Tzx/iRCOmv9MsQVK+44NZVr63riKfJKD8Clkh1X21XzYmC5pNPL8TPpYPUxW2yNCNfqdbWwkkFwK4d8Mqzio6+7HmndbmsdHRvPYmimgL54KBiJ66NWs2G5E0ksz3GhaLneCMnqyCtdtHB4NINDI2kcGc3gXAtiju6oin1bkksWcxBCoEgEmudQpsmrO3Z+NQn3VyrEuFggEZIQJXIwWi5vOsibonBWCsaBvOUgbzmgRLhuaLKEs1M5PHxqBtf3JxHThPugREnVvR/TZOxaF4NuuijaLjjnMGyGP/7Fvfi7hwYbTp/xQQDkTcdz56SIh2RQSpDRnSC5Mx2+mDF4vySazhwuEw19iRCcmMzVLbQv1zm8Hi5msi/LFO95xU58/P5TcJnnvulR2FwmXEze84qdkGXacZxYwxjLtEYcafW6Djq4XCB5nlWlTd5KEO+6tYTJjIHDoxkAwPGJLI5PZPHZn5zDQHcEL93Vg9t39eKqDXH0d0fwZ6+7DiPzOgzHxbp4CHN5Ex/4+hFYDkOtSRn1wDkHJ4tHpeWyIA6tPP9Kmy+UUog+jjgXcgYLnLQcl0OWuEckrPFaKv6Kc2BDKrLi512ZW7NuA+BgHhmx1P2HEALDcZd1lq1WbNrJ1y8ddL6r9uET99YnQtBkMWXDF4yFFArDYUsikpfG2ImQ2KMctug27zIuCJmqeE5FoihaLtYnNUxmxERM37G1NI1YCfzyvo2469oNwZ81RUJPVK2qJXRwcdBqvWa2YOLoaGbZ+71PXIkyjq89M4qC54ovS4J4wAEQWRA/GQM2p0LQTQdvvmU75osmDo9mcGIiC7OiO8n4Yhzx70+PgBJgZ18MN2xJYt+WFK7fnEQyooi1plKEFQkRVcYtO3rw9Z+NIh6SwTgPfu47gV4o8lg7efJPBmcvmhHBickcprMG1idaqwcCtc+KnpiKsQUdliMmGPVENRhOa2eH/3jv/eohZBgAUu6UKogrFAAPhEKmUz6hI6RSKJIkJpxV1PB8wcFVG2LIGg4Kl+gEXT9Ou/fgMP7uR2dgOgxRVcJ01iybsiERAodjSTEipQT7+lM45MXyDa8lYlKK5TCEZAmG4zb9vuvVpZZaZ64VuzLO8T++/Ay6IkJsaDpiIowvfKCUgHmiiVJhDOcAA4PjciQjyyc0+rXvvOl4uU05GOe49+AIcoaDkEIxmzeh225DcoJECa7dlMCBbd24aVsX5osW/uL+U0hFlIDoVQpVIshxjqxpI6LKi2IOZfVq3R0sYiUmuj83shD8NyktDqA8vgupFI4LXL0hhowhRESKRJEMyZjKWdiQDCFr5Ju+Zp/kWAtHxzI4PZUXztnK4rkKCNKn7TAMzxehShQ2Y7AZh0wX15hPgg+rkpfXdnAlop1a/lrD82MZHDw3D0qIZ7IhIaxIUCRBgo5qMk5N5vD42Tm8eGtXsNdSStAdUyHRRaK3brnIGTbACRRJ9H0cl6MvrqEnKojCjPGaewSlBG++ZSt+8yX9+NXPPIFzM3n0xjSENSlYl47LmtYTq0AW/+0LH+fyFuIhpaUY9MRkDhIhkCiBywBQHuxTpWdbrQlrnb7MhcWlvA4vVfg5M2pM9jBtVtVHToaEuSclBN1RFd1RFS7jKFgOCpaLgulUxYwO48gYDjKGAwJ4ZGGOHX0x7FofBSDqRIbtYq4g6jhRVcKe9TEhtB/NVE2MFfcER0iR0RtTV9QkIa3bCCsSdJvXjBmIt4FoMoXpMM9hnAeibOrlhJwDdsVeOV+04DCgO6qU5SR+D4Axjum8WfY7t+zoQTKsIq7JSIZlKJIETSEwbQ7bZch4ZGx/Qkg7cd5y1lyjvbVVbkAHHbSKTl9g6fAJ/W//4tNQJYqIKlfl3RbjSIRkqLLU1ud7bDyLkfmiJwirSApbwGr0LC8W5osWKKHY3htC1nAwntGXJfAgAKQSwYzLBM+rL65hZL5YswcMiLXi94zvunZD1XfWrM+6msZM9VBep9bQFw9hbEGHzTgkKuqtqiwhY9iXRI20g/ZwxYo7aJsEmr6YVuZqtlyy42q6al4sLJd0ejl+Jh2sLloZSd7OdbWwUkEwYxzfPjSOY2MZRLXqrdc/5Idm8qCE4OV7qse1NUK7zYalkKQNWxQcipZbk/BeKuY4PJLG+YqRxLXgizn296ewrz+F/q7WxRy+S5nqjZ1XpdUVctTDahHur1SIcdhEjCf1xlgrEkVXREVXBChYLuYLFiyHVRE1GYc3llAUvz77k7N4/Y2b8ZKtXQipMsK+mx1F2b1PQBDRZEQ0GbrtghIH2aKD4blCUGRplGpxADM5E7MFE6pEvSk6ClSZBvuGv2Z8ByXqFfwJEa5MPTEVRcu9KO5GFzvZf+cdYgTupx8+i5xuB+OGkxEF73nFzuDnHceJtQuzBRFUO9d10MHlgv0DKagyhe2wYMyrD9/ZSZEp9g+kLsrrq4eRhSI2p8IYS+tlfz88X8TwwSK+cnAEPVEVt+3swW27enBjf1fQ5MoZNjjngaNuqZtVPXAAtstBPLKIcGYmmM9beOT0TFXhrFHcK1MqCIFcEBNdVu7+QiEcof3nJVgk2VAC3HXt+lU57xbdms/i6fPzsFzh0lPq/rNSZ9lqxaadfP3SQee7ag+lewohxCOqLObISyVll5Kdh+eLQZPBJ/BKBOiNqQAH0kUbGxIhTGVNaJJwbB1b0BfznRUSdvi72+71whFQogRdUTUgKXWwNtBKvYYxhk/98AymswZsl0ORCHauiy1rjR8dEyINiVJwzkA85y8Cr/lFAct1AQ64INiQ0vBL+wV52nbF1K0joxkcGU3j+fEsila1McOZ6TzOTOfxjWfHAADbeiLYtyWFG7YkccOWJHpiGhxHCDUJAWKqXFXjuVDksXby5ItpRLAU8jFjHPGQgt+4qR8PHJvCVEaHw4FkWIHLOSQCFG0XistaPjtu29WL33vlbnzkvuNw2WLeR4kXn3mNUwKAM4DTigkdHGIqByjSug1CSdXn/cHX7gWAS3qCrk/o3NEbDd5HRJNh2C4sL2cAAcLy0mPE7X2xuqLA0tpW0ps6W7RcGA5rO1ZgjAeN7vm8BZkK10OAlIkkCSEN10Bl7PrI6RnYLveE3ASy5WKWmHCZIKdVNv79PzLO4LribFsqodF2Fyd01BJ0AILMd3g0jR8cm8Kh4QW4HMga9R+zO6Litl09eMm2Lrx4oKusL3B6Mg+FEtguFxN/PRAiciLb4QhJFHs3JLAhGWr7/XSwPKzE3h5VS3vKFT8s+fNLBrrx/HgGUzkLqYiCZEiB6TJM5SzENAm/sn8zPvHg6bpO4YC497tj9V/LoeE0bNcjfVb0xwkE6dBhHJJCkIpqSBctmI7IpykBQjJFMiJi1w6B+8pGq7X8tYbjk1kwLsR5WcPx+khi7UQ84ZzpMowsFLApFQqMvEKyhOmMAZcxTOcMFC0HlsvBmHBbZ66YdidRgmhJ7Npsj5Blij+86yr88beOCjJ3SeyTMRxQz8Xff52l8M/zZvRAvzbXypoVZDuCzV1hzOUtmI4bTEkLKRJczj0Su4twSb7c6ctcHFyq6/ByQeVkD0EmFeaEhu1i1/oo+nuiGJrJo9ebFCRRIia9hWTM5IANyTBu2taFh0/NYGShvAbPgSCfPjaexW//89M4sL0bN2/vxr4tot9gepNEAOBXX7QFJyeyXl4I8EA8wEEJQV9cgyZJyDBnxXrQ3REVUU1CLCTVjBlCioSFoo2wKqZ2+rE85yUTCwGkdQshWSrbp2rFYL4xoz+5Qyble5swQc2jL66VmR6EVSAMCYpMyyaEtBvnLXXNNdtbLyY3oIPLE52+wNJx/eYk9m5M4MREDt1KbR7KNZuSeNcdO/CZR4da/nznixYY54H4rR0QrF7P8mLA33ttxpGKKEjrFopWY5OMRpBLRBoKJd5nzBAPKViX0DCRNqpiZQqvNy9JiKgUt9dZE436rCthBNEuatWpNyZDmM6ZMB0x+TaiSpdUjbSD1nHFijt+7up1+N7zky1dSyESYN12WyI7lha2Gzmn+aPT7zsygbF0EZtTEfzSDRvLCs+tPtblgo4Lfgft4HX7N+Pxs/MtXbccLDcIfnxwFv/wyFkcH89iQbeRMRzM5EzENBnRkAzFa3ZpEkXG5fjOkXE8cnq6bE9othe0m4S2QpK+ekMMW3simM4aNceMLhQtHGlTzBEPybh6fRwv3S2IgFtaFHOUTuTQvDFjPvmmXVxp++qlhms3JbBrfTxQHQMkKLozb0zr1p4IZnOGGEvrMBgOg2mzqpHUzw6n8exwGiGF4tYdPXjZ7j7ctL0Lm7siGJopYENCgyRRcHAYlihMzxds9HeHcXwiA8thLZFSAY8oygCdMegZE+sTGt79ip144NgkhmYKKNouwIU774akBonQxWa379ZLWcNC+2reu63uc6v1Gt55x0781oGt+NPvncD5+QK2dUfxf35hL0KhxVC1NGmZyBgIqxIkIpI23XIRD3VU6BcLN2xO4txs8zPghs0dIVwHVxau35zEnvUxPD+WrTpH/D/vWR/D9Wtsbdy0rRuP/dHP4amhOXzv+Un8dHAWZ2fKJ7DNFSzcd2QC9x2ZgCZT3LyjGy/b1YubtnUjFVUwWxBFK7/Z3Aj+zxkHmCvGptuM48++dzwokF21IY53v0KcR6Vxr0bEGWq7rud2Jc4L22teCXpZecs7pFDAcwUDAMI5JC+23JAIi9HtqyTwCNyaHzoDw3YRC8lgjGOhaF0SZ9lq5uudGH1l0amttI7VImUzxjE0WwicCCvhcmAsrUOlBImwgpfu6sFXDo5gLF0UDd6oIkaIM3dFtB0SFa6AEVVCXJMxtqDj1h09HTfPNYhm9ZrprAnLdTG6UERXRDiCWi7DiYkcPvjNI3jnHTvR3x1pa90/PjiLP/veCcwXLG/iFcAcQJHFhClCiJiM5S5OncoZDqayJrZ2h3F2ugDddvGSrd34zQP94BwYnM7jyKiY5Hp0LFOTmHZ+rojzc0V8+/A4ADEVZGtPFBwck1kDG5MhyJSCeAYJBGhIHqt1lgCC6DGXN7FQtJGKKuiNalWfTa3fbTVPvphGBO3uYX6NcnAqB90WQpreuIbX7d+Ml+3uw94NcZyYzC3p7Lj7wADuf34Ch0bSKJpi+hCl4t7h3tRSTRHCZ8Y5HI9Ir9sudMtFKqzgzbdsxaNnZht+3pfD+VZ5TqfCCg6PpvHpHw3CdBh6YipCsgTdrj1No/J+Lf3e4poMmRC4WBRC+EYjJcNkEVWFAcofvmYvUlEF6YKNroiCeEgB836x3np6bHAWDxybLBOYZU0H03nL+74FYUmTKXpjGnKGg/7uCObyYuJQo+/Mv6ezhi2cbL1/DLtcbFEpYGG8PUKj/xlO5wyEFQn93ZGaZkYuY3hscA4/PjmNF+aLGE/rsN3GJ3NIplBkAsaAD7x6D27e2VPzul3roxjoieLsTAHr42IqoRC3EXDOMWs4HbLsRcRK7O2vvXYDPvvoUMNYjgB4x+3bUbRd/P2PBnFoNB1M7ti/JYXffeUuxEMKvvDYENK6U1PgIVGCVFhGb1Srm1fxElf/Uuds4gkq/cS8L6ZhJm8BIOBc5NHi3wR5w8W+/mTnnuygpVr+WsNMtvZEaZdx5EwHOe/HH/3uCezvT+GGLSmEZIJHTs9idL6IjG7D3/5lIshkzOWBYYnq9Zk4eCDgL90j/LU5WzCRLthIhmUsFG28bFcvfnRqGtNZA5QQyJRgcyoMw3YxnTUR8lzvywxjvNiYM+5Nh180KCsF58C6hFb1GmrFUP75q0oU23ojMCxW1r+aL1iYyppIe59Dpy9z8XEprsPLFRIV0398Ea/LON798p34/993DDM5C5pCQSFyM9NiiGgS3n77Nrx4azfe9tLtePT0NP7l8RcwkTEWBd8lGF3QMbowhm8+O4aQTPGirV242RN7rEuEcP2WJN50y1Z8+seDgYhCiAdk9MWFgbFu154E2CivaJRr+XHS4ZEMfBurxRo8geEsir1lSqDbi0ZQfpWeA1go2Lhucxh7N8SDyaSpsIIdfVGcnMxjQ4KiYLmYzhqegETsu8mQgoxePu3DdsV7z+qW6BNIpJyDU0K0vXZTAjv6Ynh+PINkSIbLxb4uUwpNITXjvKWsuWZ7ayvcgA46aBcXuy9wqfZ5WjV5uW1XL27b2dvwPZbGfYdH0jCdpQkYQgrF7bvaM4VeS6h1xlROnxidLyxZ3MHBIXsmBRxiapSYHCxMpufzFowS4w4CQJYoeqIKLJfjmk2JJeW2F8vkp7JObTOOVFjG+mQcd127Hrfv6rtk1lsH7eGKje5fv38zPvSto7CaFGIBIBqSsSkVwlzeakrq9hskZ6fzTZ3Tal37zedGg2vbeazLCR0X/A5axev2bcYffuNoS9ctF0sNgh8fnMUff+so8qYDSj2nBC6SS8OxMFuwIBEg5JGUc6aDzz06BE6EsOzD3zmGX7xuA4YX9IZ7QbvNhlrBqUoJDC84jSgUr79xM2bziwXP+YKFI6NpHBrJ4PBoGi+0IOboiakY6I5gLm8hb9hgnOOFuQIYgP6uCPq7IzV/T5VpIORYyYkcq7GvtpqgXKqJzIWE/xndtrMXg9N5TGYNpCIqNInCYeI+ToRkvGggiW8dKkC3GCgASoGQTBDWZJgOQ8Esdwg1bIYfn5rBj0/NQJFEcdx2XYyldYQUCTnDhl4iDjk+kcOJidyySFRTWRN/9cApXLc5iXf/3C5s7grjUz88g9GFYomD06LjSOkarXWvPDk0t+oxQbN9bjXjks89erbE8QR48uwcvvv8RFVT/rZdvXjTzQP49MNnMZHWwSD2ynhYwZtuHris46O1jN+6ZSu+fXiipes66OBKAqUEv7xvE46PZ1Er7aQE+OV9m9ZkPEAIwTWbk1ifDOO3b9uGiYyOrz49godOnGtmugABAABJREFUziBvlpMzTYfh0dOzePT0LCRK0FMxNa/ZeVr5c8aBnGGXEcOeOmfhzHQOn/z1/bhlR0/QPHIZg+EIYUcldJtBkfyx56K5zgAkQjLWJUIwLIacaSOr20Fz6C8fOFmWky8H9WK/N9+yFbrl4NMPn8VkxrjkzrLVyNev1NrHaqNTW2kNyyXuMcZxdCyDQ8NpcALc2J9CzrDxj48M4enz8zDrOH8DggRkuxy6beIzj55b/IFeTYJfKgjEeePvk5QAn3jwNFSZdtbZGkWjZuJCwYLlutBkivVxDabDUbAcyJQiqlKMZwz86XeOIxFSWtpLGeOB6LBgucH94h2bsBwGVRZTsbhnujCdM0EJwT8+fBYcPCB3iQYZwdaeKN588wCu25LCVRvi+LWX9INxjvOzBXz36AQePjWDjG7XbNyNpQ2MpRct8AenCwjJFPGQDFkSgpKQIuGO3X2wHIaQuthAq3WW9MRUcM4xMq8jbzngTNQPkmEF12xKNqzBr0uEcNe1G3D7rl7881tvakhwuZjTMBvtYYwxzOZMbOkKg3GOx87M4E/+83nMFyw4LoftMjAAad3G3/7oDBRJnBtLPTsoFZOIf/bCAhjEfQSXg0CQi2RKsSERwkLRgss4prLmosOeRNHfHca1m5J4++07GtbPLpfzrfJ97OtPYVdfLLgXs4ZTsw9Ver9aDoPtMjicQyYEiiRchDWFltXGePB/AppMYTgMV29IIG/a+KfHhjDtkV79tQMAc3krWBNhVYLLOBYKNnKm7T2OhHUJDY7LoHtui8x7DBBAt1y8MFeEIhGMzBfxga8fabo3ZXQLBctBVhdTeSlFVaxPICZg+TnWuriGP3zN1fiVfZtaEi0+emoa9zxyFkMzBUGipwT9PVHcfaAfNw50QbdcPDu8gO8dncDT5xeaumtWGsMYDoPpiL//4pMvQFUo9vWnMDhVQNa0sS6m4YYtSUQ0Ge991W586D+fx2zBDvYOwykX9QAIyG6dmvKFQ7t7e63877rNScgSaSgIkiWC6zYn8a4v/wyPD80Ff2+7Lh4fmkNYpfjcW27CNZuSePLsHNwaj0EBXLMpiYxu4S1feAonJ3OwHQ5FJrh6QxzvfsUu3NifgkwpHJeBe8RIHwQI9uk79vTiS0+8UFHD4ChYLiTi4o7dvZ37r4OWa/lrCe/5uV34yZkZHBvPQpYIdMutyREpmC5+OjiHnw4urseQUn62OByBENKH5TK8MFeAIklQZYKuiBrsEYsGgELwXGsSlQ9KAN12wEEC535FWtxn/Je8Pq5hsmJ0VJUzMQHeeGCg7DXUq7uUx5RaMNGScw7ddJE3XKxPaCgYTqcvs0ZwKa7DKwUSJfj5a9ZjaDaPTz98FtNZs3x9EuDfnx4FJQQ3DnThjj3rcPvuPgxOFTBfNDGbszCW0XHw3DzOzZYbLhkOw+Nn5/D4WbFH7eiNBlM9rt+UxJmZPOKaBJlK0FSRy7te7FJZ23p8cBb3PHwWpyZzZVO4JQpQQqFIBDv6onjNdRurDCQoJbh9Vy8ePzsHxjgkKmIaxsXUEYkS9HeHkdEd8ArHk8q9Km86+O1/eRpDM+W5vESB4XkdBata3Jozbbz/a4fxyV/fj9t29WJkvoisYWOuYJbtr36dIqpKCCuLE0KeHJpDumghq9tIF+3gekoQ9Dcqc/ilrLl6eyvQmXzUweriYtVNLvU+T6smL40+39K4r544v1UYNisTsl1KqHcv3LFb7NmTWROKBK+OubTPyPUmA3Mu4vOQTBBSRdz+wlyxTNgB71ksl2EiayIRkvGuO3YsKbcVAsEonh/LIhlWoEiLk2tXe2+/2OKtDi4OCK+Mpi4zZLNZJJNJZDIZJBLlC+clf/oDzBbsOr+5iIhCceNAF/7bi7ZgIqPXnLABlJO4S53TFrxC35+//vqyInyja9908wD+7anhlh6rgw4ud9Rbx4dH0viVT/+06e9/+z0vxb7+1Cq+wtpgjOOt/3wQJyayiKoSxtI6GvA6AshUJI6MIyj8R1UJm1LhFvcVt2azoda+8eipafyD38zyxlj6zaytPdFFMcdIGi/MtybmuNFztdnfn8R01sAnf3gGRcsNyAW2y5E1bERUCe+7cw8ObO+BViHmWI3go509up3HbCVBudQTmZVAo/MYqP6MGGcBSYRSAoWKYk7edDA8r4OxxSaUf7tQAmiyjERYwp1712NkQceTQ/PQ7VptrwsDRSLojqr45K/vB4CmaxRATULKdM6E63JIkkgKCBFE1ZgmX5CYYDXWj4/PPXoWH7//FFzGIUskIBM5LodECf7oNVcFhbHgdRjORfssrmTUW8fPDS/g9fc83vT3v/Xu23DjQNdqvsQOOlhT8GPBI57zpunwYN/SZEG+umFLCl9824ELWnhpdiaXgjEO3Xbx6KkZfPR7x5HRbRRNN3DXWk3IFAFB0XWFBeG1mxL49ntux+cfG8LH7z/V9hhjQDR1NiQ1qJKE8YwenDebu0SsvRJnW6PYD0DnLCvBasYYlzPaWccdNMdScmn/9z72/RM4PZWH7QqxOAUAz6zBbiH/v1AgAMIKxeauSGedrRG0lSN7zcR1iRBG5osIKRQZT5zIuXAqY8x38QUGuqOQKKn6jkuJpyPzRXz/6AR+9sICLJcJ92Evz5YogeM5ERMAqkxgO1yQuAiCiRrNztEXb+uC6TBYDsPjg7P4+P0nUbRcyFQICpaSqkuUIKxQDHRH8fbbt+GOPetwZCSNj3znGAqWG5wlad3GeFqvGa9IFIiqCrqjSlUN3nIZprNipD0AdEXKhSCNjDtqfWc718Xwrjt2IB5SykRg129OLin+q2cGUbmHLegWZnKC2BLTZEQUCt1hsB1XOJp6zqQEAAOH7XBoCsU/veUluH330lwB/b10oWjBsF2YzuKnTwB0RxUQQiFTIQCwbIaIKiGkSKAUSBedlvaktWSestLnsS8YfG4kDcKB/QPl90pp3KTJFHMVLoQhhaInKmpnecMBB6rIo5SIycphRUJYlYIam0SFWCOkSFjwHG17oioIIUIQVPFAkjdhggAglMBxGRhHMHmCwJvCy8Wa294bhSZJDc+fxwdn8d6vHsJs3qwp3q6ERCDWvEzxmd96SUPyiu1Ninnk1Az+4oGTZbVqy2FY0G0QCKHI+bli0zxDkQiimoR0sb4gUyJiYpYmU6xLhLBQsOAwVNWF6+0dfu5wpdeUVxtLOY8rv4N6+d+d16zHn953PJgaWQsKJbhhSxLPDKfrXvPzV4t9+aGTM3WvubE/ifGMgfmCVTbVhhDhWvrXv7YP/+fbzzecvLutJ4KYJuHYeG3TIwLgus0iJ++QR65ctFPLv1Bo9Tx+fHAWf/SNI8ibDuIhBRKBmNphON5UMb6sWpe/KjSF4v137sE779gZnN3zBQu67S7pOSopbwRARKUwHd7wvNreG8FD73tFSazYuO5SmRfbDsN0bjEuBkS/LhaSoUjkiq5lXWysxXW4Erical2la79gOkGc7COqSUiEVfzBq/fghi2puo8zlTXw1Ll5HDw3j2dfWKgiqpYirEiA51zeFVERVigslyPncTL+5Bevwc9dvQ4hRQribhE3iHqC/8gSJdiSCsNhYg/wc8qoKgUxEOMcv/uV55Ap2mX7k0RETiJLFAPdEczkTEznzGBCaCkogFhIRsFyEdfElJHS/YkSjqzhVE3wCxh6Xq/gj15zNT70raMYTesNcwhKgD96zdW4fnOyYc4hUYLemBoIR4Dlrbml1hwvVVxO67iD9nA59Xma1Z3q/bx07897e/9y0RVR8Myf3Llme8i10Aof+t6Dww1z06WAEmBDMgTddLDQgnnWtRtjeOPN2wIBY6vTq/ye0PGJHBgXE641mSIZUWDawijpLbdu7UzS6GDF0Ja44+///u/x5je/GalUahVf0sqi3qZTLNq45iM/aOkxCETBNRFWQEBqFlBLSdwbEqEqp7/JrIm9G+P44tsOAEDDaycyBlwuxgc1e6zOJtDBlYB66/izj5zBn3//dNPf/+PX7sH/7+W7V/MlVoExjm8fGsdHv3schAALBaumY3MthORFVac/bpcS4JqNiWA/qLcXNGs2+AS9ouXCsF3h0sc5BqcKGE0XMJkxMZUzcGQk07aY48b+FDalFvcsxjn+6BtHMTSTR29MBYEo9vlb2kxOvP4v/c7NDfeylWjatrNHt/rYjYLSqCbhnS/bgf7uCEbmi/jco2fLSAaXaiKzHDRKAup9lvMFKwh+o5qMe348iJm8Cc5EY5jxxWKQ/63FwzL+9Feuww1bUmIN2S6++swI/v3gCAy7epztakMiovDuk1HLJnDUaNhWfg6mK1wOHe+NVhbxNYXiJVu78aXfae3eXcp6Wo3148NxGF7y5z9EpmhDlQkoWRTOMs5gORzJiIKf/fGrQCkJSNKOy2G5LHB0VyUKWSIXhSR9JaHeOv6bB4/hbx463/T3f//nt+H377x2FV9hBx1cHAjnPO4573n/ZsDRsQze/9VDIIQga1gwbQaAeGPRKZJh4SbdjIi00mi3MFd6DhRNB6bDQAlpSFBZSfguouCAIlP8xztvwV//8DQODs3BbDXABqBKRDTrmedkT0W8HVYkrEuEENNkcAhHwtm8ie19MXz9Xbe25P5bimYxYiIkY3i+2DnLsLoxxuWOTqNs5dEKca/y+vd+9RBmcmZAoFsr7jkEwEB3GIbtIl20QClFWBGGEZ11tnbQyjquzN/m8ib+178fQtGy4XhTKMBQ5uItEeGQGQ8pZd/xu+7Ygc88OoSz03kUTBd5ywnI1zIV9RrHFSYLlIjHcdiiwEOIPsQU1pgm4/xcAbrNIHuu+iFFwrbeCMBRdV/5+/3x8QyimoyxBb1ujUw4hZKWBJyEANt7osgaQny6KaVBkSSAAEMz+TJxQSVCMkVElcCAoAZfsFyMLehgnIuJN5xDkwQBPh6S8aabB/DomdmGJOvK7yyjW/j4/ScDERggnNGv2hDDH73maiTDas38vN2JnsAiAbxguWKCMAHWxUNIhRVkDRujC3pQY1Sl8ikfLmNwGMcNW1L45v+4bdl1NxAxAXi+YMF2BOlfkShu2taFrGFjImMs6exfa+YpK3keN3tvpZ9xTJMxntZrOo6rEsGmVBgLRQsEBIwxFGy3ZHKNik2pEKayBhaKNphHUAIEQdPlHK12DL0tqPrvieincQinYEKAbT0xz6m29nfNGMcv//1jOD6RDR7DZbVdyCVCAIKgiR5WZfztG2/Ey/cIAry/fmbyJsKKhC1d4aAG7tequyIKdMtFwfunXTdNAkCRCawG+4z/elsRuDQWbl1YcsxaElBdCLSyjh2H4b4jExhLF6vM//z8TxAnZVAi8s6i5YJDTLtpdJcs3ae0GkJcJc4Z/3EdJnLOazaKST3n5/S6v78pGcJs3oTtiolLpcvCn6ypyhRffdetF8VMrYOLj3Zq+e3WUpaDds7jx87M4J6Hz2JoOg+rwvBu9/o4nh/L4JFTM3jo5HTbhiJxTUIirKBgutizIYZ/uPvF+J///hxOTmZh2MzbF5YHShbPyMp1WoltPRF88jf240/+83mMLejY3BUq+85qncelbtMLRX9SFwXjHJZTbrhGvf2GUmDP+sSSamgdtI+1ug5XApdLrcuP2w+PLCBvulXnKYHIn8OqhL0bE/j8W14Ci4l+tuEwGJaLM1N5ZAwLyZCKXeujoEQIkg+PpvHU0DwOnp/H6EL9Mx0QebUmEQz0CGOEF2/t9l4Dwf/48s9wcjIv1rFEvVh58XcViXiCDDHWM6RI2JgKIV10IFEhnE4X7SDWLf1dITjXwDnHnddswJefeqEs1hZ1eRFX+HnHQHcYCa9XYthiOuFc3kTeauwIocoU12yIY2ShiJzh1MyPSiF7wg1ftFILmkQQCym4ZlMCX3zbATDGl73m2q05Xsq4XNZxB+3hSurz1Kvd+HXX4+MZFEy3oRivXXz9XbfgJdt7VuzxmmE567iVe2FDQsPgdB6Gw+rWldqFf7ZQtNe7pkSccYp3QMuUgpLavHCg0nRFQrpowXTcIDYPqxJCstTwMTrooF3I7Vz8oQ99CH/4h3+I173udXjHO96BV77ylav1ulYd/+trz7V8rRjNw6FKFF0RFabj4uhoBu/9j0P4vZ/fjd98ST/uOzKBY2OiQVUJQghSEQVnp/M4Ni6K02en8+iKqGUbmX9tWJUwkdarmq61HutyGEHeQQdLxdd/NtbydRdS3OEHdMfGMpgvNp8OVAmHMVBCwbFYKGNcuBp2eeMi6+0FtcZw7Vkfg+EwTGR0QW73MuX5goVDI2kcHk23LObojanYX0fMUYmz0wWMzhfRFVGhSFLg3OYjFVFwcjKHf33iBbxoa1fNRtFKNW2PjWcb7rvt7quMcfzDI2eRN52yoDREJcQ0hrG0jo/cdxyJkIycKVx/NqfCCClScN2GBMVk1sQ/PHIWt+zoueQTmaWi0We5MRnCZNbEd49MYHRBR8awg0IPq4jyOUQDW6EUO3pj2NodgW67yBsOjo5mPLKGBocBpuOumFq+GfyG3qnJPI6OZeqOygOE8NP/HEAAwxIF/3puThxiHOMTQ3O49+Aw3nzLVgCN3QKaradav7vS66cU9x2ZQE63PaeT8qIXJRSyxJDTbdx3ZAK71sVwfDyLgumCMS6Kf94HoVsuKCU4Pp7txEcXAZ9uQdjhX9cRd3Sw1sE9d1mHMTCGRdGGR3QqE3J4P6/nlzCWLqJguTAsJ3DBFoICsX+btoGIJmO+uLbH6vrnQFiRkNHtoOFzocCx2OixHIbvH5vEoeGFtoQdgMjpJSKaOLIkyGvrk2GkIsLEIW86mPEcCRnjODKaxl1/8yjedvt23H1goCUx8lzexN88dAY5w8bGZLgsrtmQoBhN6xj3SKOc44o/y1Yzxuigg3bRzkhrxjg+/eNBzOUtb2rC2oFPFOCcIBlRwTgwl7eQCClV13bW2doHpaTsezk8kkbBcoKpsLWOY8aFcySw+B0fH8/iA18/AttlSIUVZHQb4GJ6g7h/idd0osGDlmahvnPv+oSGiCYI0aYjpn1QQgDKYTouDIshrEpV91Xpfj+RNRZNGgiqSOQcgkjSG1cxkzUbGqVwDgzNFoI/D84UockUIYXWFHaUPp/pMMRDMjIFC5tSYYAIExLXm2BACIFEAJsxbAyHMFsw8dcPnkZEoYhqChRJkNyOj2fxx986WpNkfXg0jb996DRm8xYIRAwCArguw/PjWbztX55GIqRUNfyABhM9GS8jeZ+YyOGD3zqKd75sB37n9u2Yz1v4wk/PVRH4JEo8wYr3GVRsbZQSEMZxbraALz35Al48ULtOVw+l8WLedCBTiu6oiu6o6tU2HFgOw03bu/Glx8837KEMTuXw7UPj6I6pNR0QK4nuJyZydb+DSwXN3ttHX3cdZnIWnh9NQ5EoJjJGYAJS+g1xAC7jmM2bWJ8IQbdcfOCua5GKKkgXbHRFFHRFVfzlAycx4tWBZUk0rgEAFHCaiBVKUS8jYBzeGhTrCSBwGINuiTxLphSpiFy2T3z5qRdwbDxbVgsnWBRH+AgmgxAhLnEYh265SIXFOfeT0zO45+FBDM0UYLkMCiXo74niN2/agtmcjZMTWdguDwir9dCMcM+BpsIO/30AYmstWi7CUbluXbh0v29UL13NmvJaE1CtBTw+OIt7Hh7EyckcLJtBkQm+/sww3vNzu3HLjh78wyNnMV+w4DKOrGGUifd9kWQjrGQcyXm5eI8AUKjI309N5YJ9ox4ms0Zwz1aez/7fWy7DoeF0R9xxhaKdWv7rX7T5Ir3Kxrh9dx9u29lblvftXheDxRhMm6E7qkKiBD8dnINhO0Es2spazZkucqYLRSI4OprBn3//BE5NZqFIFPNNhF6twidREzQWdgDA+bki3vaFg8h6ot8X5jj64hpiXhxGCEEyLOPERDbo096yowcHtnXjVz/zBGw3j76YBt12MZExyl6//zocr1F3ZDSNX/3ME/jDu666Ys+LC4XLYR1e7jg2nsXx8SxyRrWgy1+3hiP2m7PTeZyayuP6LUnENDmIOwa9WEymBP3dEbzxQD9uHOjCTdu6cdM2IdIYXSjiqXPzeHJo3pscXv5sLuMoMo6Tkzn82XdP4qbt3bh5ezdSYQWDMwVhrOBNlfZflx8D214dXZGox5VjACdYn1BxeioP5hXDaoXDhs0wkdYRUiWoMgHhHBSLvRFh5ETgMhaIMWSJerV5I5hQ6ragOLcchtNTOXRHVcwVmnNyHCaIxX70XJlLAaKHEFJokKsMTueXveZKa46zBTPIzeIhxet3X5kckQ4uXVTySRjnF63PcyHNCRrVbj7w9SOwHAZCyIoKOwDg7358Bl+8gOKO5aBZzy8ZlnFiMheI/lbqkxL1MKBdKTXjolbjgxKgvzsi4vmxDN771UP4vVfuxt0HBgCgqk7TFVVgWAw5w8ZM3oTlMGxMhKDJ0mVTs+zg4qMtccfk5CS+9rWv4Z//+Z9x5513YmBgAL/zO7+D3/7t30Z/f/9qvcZVwXMjmbZ/J6xIKNpuEFRmDI4P/9cx/NUDp+Byjrw3unOhaKEvHgqSYwDQJIoM4wFxx/bEIrUgEQIGr+EEDsNiQdE7pNKqx+qggysVI+nWxnS1et1KoDSgi6gSFortF+xErFcdxvgkOsYYMrpoihYdF9N5A8BiEEwpwe71MRQtF0XLwVhaODfM5U0cHs3g8Egah0bSGGni6AAAfTEN+/qT2N+fwr7+FDYla4s5iNeI1mQJIYVCkyWMLOhgXOydlQF03nQwnTWg2y4+8cPTZaM8S8eZr1TTdr5oNdx3291XS4NSEEHGcxiD5TLM5iwwJtytNIUirYsixHjawOYuUlY47RBpmgf4qkxwfCIbNMYarSfL5UgXLTw2OINrNyUwOJ3HZx8dwnMjGXDGg8BckghCMkXRXn1iquO5CDqM4bkR0fiqbNgCwNHRTPA5FCy3jFzabA9xGcenfngKO3qjAKpJIDvXxXDH7l7821PDDddTvd+9bWfviq6fUoyli8Ip1Yt5fKcW4jnNUSJcaMfSRXRFhNuo630mVQIfxpE1bMzlzbZfRwfLQ6syxvbljh10sDJwGQ/EGlWiDe/fpf+9UoiHFLGXc0CWCaiYYyb2OAmwHQbDXiQirVX4cZQsETCOsobPaoN4/1fqin96Oouc2di1qx5cLqZqxTRZTAjzzra5gonprAnORTOFeWTXs7MF/N//Oob/eHoYH3zt3rru/f75qdsusroNTZZQsNyymgAhBBFFQrqESHaln2UrHaN30EEraNTwqRWn18K9B4dx8Px8206uqw2Zivfguhy642C+6EK3GRiAmbyJjGGjO6Ki25usCXTW2aWGI6Pppved/1O/TkEIkDVshBUJA90RGLaoXcje3ms6DDZjoFSQQWVPtBBVJeRN4Wav20LMMZk1sS4hyB9+3gZ4uTrzyV1S1X3l7/eMw5tkhjIyReU7Mm2GdMGuIpa2AtNhMOs0UUsfjwNwXB7U4OcLFgqmEOS6LkC8V0UJgeM5FZs2AwGQM8vJu7YrSNiM82A6iu2KM91yxO8oCg3WHacMrsthuxwF08GOvihsl+PERA7v/eohACgTcZiOi5Ne43OgO1JmHFJpMAIC5AwHvTGtjHgiERJ8zr5wtrQEIyY2iHvlkz84jahWXafzUbmP7t0Qx9efGcF0TsRSPvFek6WAQGi7DDN5E59/7FzDHortMMwWLHz0u8fLRC++A+KFJrpfCDQj8Q/P6/jdrzwHl3HkDKfq9yuXCQeg28yb2MzRHVODiRaAqD8NzRQQUWXkTRcg3CNormwu5MNlYj1NZIxA4Lw4tY5ivmjh8cFZ/M2Dp2u+l8p9wK+bC1IYAqV0wXTw7UNj+Oh3jqNouYiHZCRC4j0eGU3jueGFpjmMIonP/ubtPXhueB4Fq3ndUKZin2jmEgyIfaY7ujjdulFd+GKIoC9nAdVS4U9qm8ubizVLC3jq3DxOTT2Ht9++Hc+PpVG0HG9CCwnON8Nxq/K91QYHALI4rYMQ8ZokicB2mteYW9kCOAcu/GzsDtYK2qnlr2XUyvtCkICQ+O896+NBPU+VSGAC2ipsL877xrPCoFCmZMVXTauP53IOSoTI17BdTwAcRkyT6/Zp77p2A6azBtYnQnAYFzFek+dhHDg3k79iz4sLictlHV7OmMubSOtWw3XjejwCuyRvrheLnZ8r4lMPncGf/MI1uL4/GRh5bumKYEtXBG940RY8OTSLzz92HmMLOkxvUnQp0rqNB49P4cHjU4HpgRBRc7glZfbSXyOBaSdfzPdtKoyueGNCLoMgy37tZ6OLhg0VaVrpcxUtF3N5Cy4Xn4vfB24FRZuBFJZW06r1HAFBmIvvZqXWHKUEOcPGFx471xFSd3BJo5YhQFdURcF0A6PiSqxW/flCmhM4DsNfPHAK8wULfTENmizqqKW1m4K1tFpmMwzNFJpftEbQrOe3lMmpFxKMA2MLOkIKhWG7yDDgI/cdx/3PT+A1122sqtMQEIRUismsW3LMCTHj5VCzvFi40ibKNkNb4o5wOIy3vOUteMtb3oKhoSH8y7/8Cz7/+c/jwx/+MF71qlfh7W9/O173utdBUdY2OQUQwVe7KFhOEFRK1Hfv4sjotnD0AgTJ12ZlyTEAmJ5LULd3mCkSgeX+f+z9eZwkR3kmjj8RedVdfc709ByaGY2O0eiERQgZCxnbyGBjg/F6bfCBr7WNzO6CzfdrWNhdfICxd338bMB8wQYfHMYsGIONZHMIARISh6QZzYxGmrvvu+48I+L3R2RmZVVXdVd1V3VXz9TDB0ldnZ2VlZUR8cb7vs/zcMSosup9mM9eNl2G5bIL22NhgGioCrJxreZcffRxtaJegWCzx20W9cU4cx2ryLXQqLitKRQLRRvzRasm0f5fP/4E3vjSI3jdXQdRcTyYfkC0UTLH7QcGcNu+LG7b19yZQ1clgcPQKHSFhsFrFEMJveFcV7I9TK2YYJyDEmDYV8KJFooC1atOFW2bXUuA+jl6PQRBqcM4ZvJWOE8HgWiQrJVqWASqItUgFooWknoy/Dz9Rpq1A3whBPIVuQmS6rPrn08A+L/fmcS/HJ/BM7PF2sYX/z+ZJ7B2Wq1zEEISPAgByBpvWfNM5axQbVW0+MFLNse7v3AaBctDeVURtoBvXVyGrlAcGEo0HE/Rv41rClRFKsufnpGqJFzwjo2fKPYOJECBsKk6UD4OivSB8vDegQRWKm5TFxOg2pyznvpiH330cWWAcwHPnzs8LsCYdNSoJ3F4W91NUY+AlCBk8yDCRjrhv75N19UGAuWduYK95Uk3Ef6jisfOrWzqnIxL5TEAyJsu5ot22MwJACyyd9AUefzpmWLo3Bl18agvuqkKQcF0YTO2KicAYFURq9HnvZrWsk7H6N1GP6G489GJgs8jZxfx5196bl3l4+2AxwFVyHg5V3b9dcdXwRJSPXE6b2G54mBPVs5PvTbO+mgOzgU+8a2Jlo69sFiWDfaQMQgTwEBcAfGFB8LGflJ1swhe40I2o+XMahO55RMyyg7DxHIFoykDgMxzBFt5QqR1PbB6/g7me8tltY5nawwj12926ebW3eUcRAAVx8OKr6Yc5Pfh702ZkE3jtie132yPQ1NoKD5heRyUcDw1sYK3fOo4TIchoSvQVAKPyfZXAUnoVKhsBonGUx7ncDyBuK5gd4bg2bkSAOD6XSlQGuRJqgrsiyUHqZgauo5N56xQYCQT12C5DA7jmC9Y0FVa07gXDeMcJj+HQqSbWZA7VYh0CanP00WFWKLzKBcClsdQsavaeBRS5CJoIJQqqjKOTOgKTMdrWEMp2R6m82Z4XCam1SkgMgwl5bMXkJdUKp1adrJ4So2ADGo/m8c5TNeDxwRaDTm4z4iYyVvQFBo6dARYLNuoOAwxTYEQArYHdNuDqjp2SJhvslwGeByXl8r4t1NzsL328uhR4odCCb70zDw+8filkJBRbjEvr6sEMT9/FpDQnprIhfNeFEHqu34PR1qcrDwmQocjYO288FaToLfLKaSXwbnAu79wGgtFSbwP51B/Yl8sOfjAV8+haFWV/XkPxIf1z67rzx+dvLJUA/elPq4OtJPL38mghPjCHwKEEoggobdBbKcogMtE2KCtUrkWLRRtCAhMr1gN67TPzhZheRwDCS0kZ7aClE+qvNrWi63G1TIOdzKWyk5Dl816WG5139xKLPZ3j13C39x0JwiRfWSmw2C6DI+dX8J7v3IOFYchE9dQMB04nmhKvgiGNBcAXyNk5hzgtJo/CPYn7dQzeMQTVO45ZVyiEBLOLQTASsXxyWhYsw7cDOU2xRzXO7/HOXSFYiihd2zM9YnUfVwJaPYcT+VMlBwPOdMJ8zZRdCP/vJVj6pGzi/jDB8/gxGQOIATminQODgRLCCFIGxQFqyNvtwpGk7xAL2Ktmp+AwFKxSzepg/C4QNlm0FUKSgU4F3h6qoDn5kphjBzN3QlIR2vFFw4MxI+AvuDzRtB3lF2NDWdgDh8+jN/5nd/BO9/5Tnzxi1/ERz7yEbz+9a9HMpnE/Px8J6+xKzBW9w2si7lCVZkgmqRTKSB8BVYu5M/RJl4hBBaLNvYNxsGFwLE9GVy7K4XTM0WMZWobooUQqNgedJVisSgt2zWVQhD/d46HiuPh2HgGx8Yzm7sJ24h+U0QfnUCrvXpb1dNXr6gVKKK0YhtZj0Z/4TKOucJqFd+SzfCeL5zB5LKJA8OJkNDRLpnj9v0D2FPnzBF15JCEjsZEjkY4Nl6d63ZnCGxXwGUM80UHTGZDEdNUJAwFBKSmUJQ01I6qk0WvpdG8m6u4OLon3fK8GjQ5Tvn3OGjCD5K0LpeN+UHBHyBQiIDlMiyVHSR1FTGN9htpsHaAb7lScZP6GRLeQhJcoQTnfPb69pfTapE0FHz12YWG615wH+YLNhjnUBT5eVv5zABAwHFmtoSErmDfYLwm8ZeNaVguO6ANOkoD+8MzsyVoquzqyQWEGgIYKoGryDGzXHZWzREbGT9RvPLWPXjHZ09ItcgIgiI9F0DKUPDKW/fgK8/Nr/udCgCp+PpBXj8O6aOP3oXHeEPSRuCuwXxSh9hAfLXVKFrSwcF0PbgMCLtAfKiUIK4ryJm93cifNx2UHS9Unt7uO7+R2LoeC0UHhMrCUdC82QicI1QynivaeOfnTuILJ6Zx//dd15CMLIRU4RI+0Wg6Z2J3xoCuSHe7VomqqbhSs1YF7i45092SdWur1slOx+jdxJWeULwaYqP1Cj6/96qbkY3rNUr0p2eLNfcEqFpv9yo84asvQoCCwG0w71gux+RyBXsH4yjZrGfGWR9r4+R0IXRmXQ9cAEQIqJSEDR0rZRupmAqVVokJFAQalWPBYxyUknUFUjwuMFuwwrXT4wIEQEyTjtON5u9j4xkcHk3iyYl82AC71rsISHGEaL90uEQ0WLd1RV63QoN1u4WbBJlPowRYKlVddxsRSwumE55TpQQ0VGcDVCLgMIECY/J8kKImkhhSPYfHOSihq5wIokU/2xWhK0re9KAqBCqlcBkLCRym6yFXdpFNqFgo2uBCuqsxLuOVhK7K753XNe4JDoVUr0kI6ZIhFBKS1SiR32NCl0SgIE/3vofOImmoeOTcIv720UshycJhci4Jzhl8XVwAwnd983yiiQCQ0BWMpgyUbdmEpPjq8rN5C4eG45gvWPCYJHZkE5pUv6OS9HJxqeI3P1FM2dIRhfvznaFSjKRiNaq3OwWcC3zn8grKtlQHn7YcOEz4z4gI8yLBLqKdWFyK0HD82RfPYCpn4s5DQ5haMfGP355A0XKRN9t3m94smJ8vDUhUlACf+s4k5gsW4hptySmjEWyX4yPfuIBW+urr76HjCTgRYklCoxhJabi05NX8DYCG8w/j8j6vhyCsCsa7gHT34VxguSRdoBvlCbeKBL0dTiG9jhNTeZyZLYFANno3am6OEiF7GVy0N3+sh3IDF6E+rg60k8vfyciZLmKaggoX8HxyRLsgkLWq7XZ7tDwOXSVwGffjWMByPczlRdM67eSKCctlKFgeLNdbT6ckhKpQxHW6ar24GvINW4mrZRzuZKyYrTkylx0PzzswiGPjmbZjsYSuIqGr4FzgM09MwXKlg/SsT8hSFAIF0pWDECATUzGajuH8YrnleYlDxsoAEPcFDIRb7YlrBaUGMUOQa9cVQCUUjsfh+fv5jRA7uoGK7eLowWEcG8/g6Fh602MuSt7ZnTFguwJlx4NKKXandUzlLfz+v57Gu159C27Zm+3PkVcBduLauBYJbe9ADM/OlTBftDEQ1yJiJd2p82ylOEFQU1j2HYJUCgCkRrAEkMS+buG63TuHsLlWzW+pZKNNLt62QcB3aBIEHALZuIa86aLieDi34K+lfl5SUaRwDvUFoCu2FDfSFAUxnbYlztFsbtiJc8ZG0CdCNsam5TUIIVBV1W9kEHDd3m5MCbBQbv86g0ByVRKMEChEbtApkQldSmRSebZgIW+64AKYWDHx63//HVy7K4V7rhvBxHIFswUbAwkNhiKbfOcLNmzPCwNXATS0ku/lQvJ6uNKbIvq4crHeglmvqKVSCkqlFRfrwGa0EbEjvDYAf/fY5XXPsStt4Lb9A7h9/wBu35/FWCYGAeDsXBlTOROWy3HL3izihtLUkaNVUErw6y+5Fm/65JN4dq4kFat51aZTpQSjaQOBWXk0OfHk5VxH1cmCa3nbZ06smndzFRcpQ8Gvv+TalgOgo2PpsNFUV2Vhf5WStRDIxlXkTAUVxwuVHOYLFigl0BUKhVLctj97VTfSrBXgu4yF91igwT1uBF+pttdCWSGA3/38Keiq0nDdOzaewa5MDDN52fTAvPZmDNMVABFI6KsTf0zIZhuXcVhuVSEwem0u5/AchM+p/AXAHAFKOJKGirihYLZgIa4pobqr6ScM2xk/UVB/HijZze1qR9MGKCU4PVVs6Zynp4r4waNjTX/fj0P66GNrIQJihh8HRB01gtejxI0rCdmYDlUBeJNwhQsBlaKnSZ6cC3zg4fPQFQoefE/b/DUpHVjkOQDK6+k2q1H/WV0m8NiFZZxdeBJvfOl1q4puTMhGw0CdzfY4Li+bUChBTKUtF88efHoOf/vIJZybL4UNkIQAMU1BUlfCdeuuw8MdT+ht5TrZ6Ri9W7jSE4pXQ2y0XsFnYqWC+z/2XWiUwGVSRYoSAsVv4g7uyX3HxnBqOh8q+PcquAA0Wm3YVqncS0WvmXGBqZyJvQPxnhhnfayP5YrTlkKmgBSeUCkBIbJJfKFo45rhOAyVwnQ5NCoLVYqfozDd2maF0MUCtYQEAZ9EFCFqOB7HStmF7fFV8/c3zy8hb7pt5bMFhK+SvDYIgMGkjuXQbXtt9dF6rLc0qwQIbgsBws8kIEnIXmRscT8ZoRIifx/9PKLa/BF9y0C11HQYipYL5scnMwUrVAIlIGCRoGWmYGGpTOEwFrqlECJCFwtDpVLF1ZGNe83UVQWqbsfUV2IdTVfnSEIIdJXgsfPLeO2HvgnL4eAQSGgKXM6xVFod5AZ1k+i5A6RjGgghSBkqyo4Xxkumy/DcQtlv6CHYlYmFucKS7WGhaMNyGJgAZou1uVEuAM/hsHIVDMS1no6r6xGsv8/MFFGwXeQtWa/SKJHztqhV+iaA7+7a2vkFZEy6UHLx/ofO4f97+DyE7xK7UTGizULGydIdQyEUI2kdkysmbFc6lLSD6K1oVVCdQqqKF9ZoTM/EdaRiOnZnBGZ82c1Gp1ZotZmrFVBSVTwO3HRMVyo7/uEDp/F/vzuBH7p5D/YPJUKS6VaSoLfaKWQn4ImJnFTiVMiWiYd1E50a8XJt6seNVyvayeXvZAwldCR1BSlDRd50YXssXHeoT04MYrZmyMY1qD7xeHmb3VmD5myHibBuJoSMVBVKV9Vph1M6JpZNLJZsNGhRaQrLYUil1HC94FzgY49fxscfv4x5v759JeYbthpXyzjcKHqh+fFUi7VMAOG+OXDX0xQpDhTTauO/ZrHYyekCzi+UMZjUMZuXQsWqQvwxTUCUwG1R4PxiqW0X8eBwlRC4HgfxQ8V2zhPNHUThcSCmyuv1uIBgrcfW3YblCtxz3Uj47Gx2zAXkHUNVcGnJlHnFSG5ACIGC6eKX/uZbOLon058jr3Ds1Fz8WiQ0Sih2pWOYLViYylkYSRtdrfNslThBtKYwmjL8vKlfL6Ay7zqTN6VIYRcnsPOLve92EaB5zY9hvtga+bFXIOfpII7m4JyHLlRANS/leb44i58HnS/Jz6lQIK6pyMa1lsQ5ms0N91w3goefW9xxc0a76DvKNseGyR0TExP48Ic/jI985CO4fPky7rnnHnzwgx/Ea17zmk5eX9dgt9k0GUX9X3pcQPMD2aGkjrLNYLmyQLFcdqBQgrFMDAPxqpX45eUKXnrjLjx+YTnc0HLB4TAGjVKplobVdnnE///lZRMfe/wyfuauazb8ObYDV3pTRB9XLloJsusVtYKCasXZvsaPXWkDdxwYwG37BnDb/iz2ZOPh7zSF4vhkDn/3zUu4tFiWBRyVdjEQIFJ6cY2bESQnBEHH1cnuPjKCd736FrzvoXM4M1uEw6Sl5o1jabzh3vY+7+nZIhQSJBwAlTZoDSRyrUkZak0DQxBoVBwGhfKa5MDViGYB/orphOqOrteaOlKQVAcQkg96CUXbQ5wLDCT0VesepQQ3jqXx5ERuQ+cW/j8MbXURNlBm5ai1AQxgul5NEil6qwVkY2vFYXjFzWN44NQcpnNmSKBJxzW87oUHNjxfnJwuYDq39oZ0Omfh5HQBrZchmx/Xj0P66GNziBI1AjIG91WCA7cN7v9OiKpV9dWKa0eTsD3e3BVCSAW9o2PpLb2udhAkSXdnYvCCROU2S6tUOvT+Gz0L47I59kNfO1/TAFWyPUytWA1dZRgXqDisZbXHfz85C813zDNdL0wMV7iMLU/PFPGmTz6JXWkDSyWnYwm97Vgngxj9vV85i5MzBbgeh6ZSHNuTwf3fd2Tb1+UrPaF4tcRGaxV8yg5D0fL8RlPZkB3UYhQC7BtKQFcoTs8UcHI6j5zfmNNJ9eNugPnFBM1X/qdUwGO8Wij3G8Z/+XsPXxHf8dWAIOfUDgikY2vRdmE6Ml9tuwKj6RimVky4XHalxTQFg0kdUytmzXMd7DMbrV+Ba0L4M4DFko0XHByqybEE88xCofVCZECCoEQ2Y3Mhf2706SklSBsaErqKhaIFy+Xo5OiMlhDkdclYuJmiKBdSWEFTKCgRNQQY0eBeKpRgtmjB8XiNMAyB8F07+KomfMYFLB7kGmVEE9eUsAFoJGVgcqUCjwOWW5uTlASVau4keC2uKdiViSFlVMtF80UrFJpxIyqlZYfBXJKNLUodSaD+3NH3CFQOl8p2tbHQ/3dABBlK6eE1yNjKlMqzlITzWiMw3vtxdRTR9XcgrmGpXHWM97gIHeGi2AjBup7Ut+ETdRBcAHFNEokSmoJpx4TlsbYvq91PQSGfT8fjoFhN8pbuDPK5G0nrDfNrURgKBaVklYJvPWIahec3esQ0CpczTK1YYFy624xn4/A4x2MXlvHo+WWkDDUkczcTiOtGc8xWO4XsBBD/AeFcNGw07PVYsBEMlTYUEwwQ1yT5c00Q4LarxL2lj9VoJ5e/k11+ooJk1wzHYbsCJdvFUtkJc57NEPRxuIzDZYDptME67iJUWktIYQJI6qvjPwCI+QJl9cTvVhCsF5eXynjZ507i/GLZ32vLPUej2lgf7eFqGYcbQa80TMfXiSMD3HVoCHcfGcEjZxfxZ198DkXLRdH2QnfC0XR1fDaLxQKCLueA7bGQyB2AQu4DHD/epthYDFOwPRQXPKiUhMKgzeLg6Pkpae5gxAVgM4HnHcji25dy271NqYGmUDz83CJ+6cWHOzLmlisOyg5DxXf5VHx1d5fxyL2Cn3/sz5FXMnZyLn49QYCBuIaS7WHfYBzLZQd5LqBRgqN70h2fh7dKnCBaUzA0CkNVYLkMRJEkEoUK2C4Pc31BXqHT89lSaWeRIoKaX7Am5wNL2h0Gj/Hwu5zJr14HglxvMzAOVGwPFcfDsfHMmuIcjeYGmzE8cXkFj5xdRExTsGcgBkNRdsyc0S76jrLN0Ra5w3EcfPrTn8Zf//Vf48tf/jL27NmDn//5n8cv/uIv4vDhw926xq6AAujUdloIqXYAACtlFyNpHdmYgvmig5im4MBwHNSnMceogpTBMZUz8dFvXkbaUEAowe60Acvj8JgNy1tdtAHkQqCr8gF2mMAnHr+M1955YMc0EFzpTRF9bD1abcjabPtXq0F2IweClKGtW+TpJOKagnuuH8Gt+wZwx/4BjGVjAOQm0VAV6Gqg4qfgsfNL+KMHz3Rt8xCMecYFrhtNomAxVBwPedOTSopCNsYlfbtfoJqcuGP/QBfVyYRk2IZM2/YjyeWKA0ooxrNxLJVt2bgpqgkLjQICBC5jKNluqEpBUC3mJ3QFqkLC5MDVPO/VB/gLDkPJlnbPhkrhcQ6FAGydoaQQAtdfP3tF3SMKIWRjxULRxvhADCWbheseADwzW9j0e7hMPt+Ww6W6nK/cqSmKTOzVBcJCCBTNKvkoqswKVPlYHhf499NzUAgwPhAHJQRcyEbVjz52GcfGsxuaL2Zz5ppFRUCqns/mTDRu5WmExsf145A++liNeleNgLTBGzhqXK1EDcYFLJfBcqV7guVy2Rzp+T87DKbL/d8F/5c/LxRtv8GwORyX4+RMAbftH9iaD9QmoklSl3lwdopnbhcREBwnV0wMJKSAg0EoZvNWU2VsQMZ/hqq0VBxnQmBf2sCl5QqYkAUlEGlNnzddjKQ0XF42sVx2cM1QAoa6+YTedq6TJ6fzeHo6j6LpSVcVh+Hp6TxOTue3PTF5JScUr6bYqFnBR0CS1qp7h9pCAxPSeXEsG4PpsG0VbmgVAzEVedtDKqaiYsvCPiD3SlSloWL6WMaA7XHsH9o5tu5XO46NZ7BvMI7FBm4JzUAgBQBiegxTKxW4TKDiyBzQSFrHQtEGB6Aq0ukhcJ+tb75Yt4GNSJW+mErxlvtuCOOaYJ4pWu76zaI+RlM6ErqCpbIDl3FfLV2scrwAZME0rtGQ1JDQEri0bMJ2GXZnDaiUYqFkw3a5T0Ju7b6thfUEo0LVNn8/7kSYDlxw0Mh+VTbaIMx51DbiA0LwpgVhETkOAJK66jfYeFgs2eF3Fv1zSuR3rRACTjlcTyaqUoaKXWkDcT1C7ChYmFtDUS+4l9JntdooX3+PKeA/YwQely7n0XsVXNdwUsdSyUG+4iKuUqgKxXzBkk53CoHXiDGC2sahXo+rA9Svv5bL5Rrsf2n1TT4df/8unbdVjGWMMK4KnCs6sc9cr0mNAzAUBS5j0FQpgsKYgMsDMqR8zfYYTJthsWTXqAxTUs3ryjEOpDSCkaSG5bLbUKSNEmAkZaBgenAYQ0JXMJ2Tz3XCb6gFgIW8I++BkM0DCUM2vk4sV/C6Fx4IFRq72RyzlrNyN5xCdgJuPzAAhZKQfFb/jAn4JEex/eOqVRwdS+HEVKHh2qIQ4KdesB9//83L4bioFwACpHI32eF7gz42jnZy+TttfxxFVJBsruBgIKFhOGlAoRTzRWuVK2J0fgj+HQh8dHN+2DcYw+RKqwTqqptfANbElsjyGFwmENcVmA5rOYYO1ot0TMHvff4U7Mhkw4V0a3Mb1MZ2er5hq3G1jMN20UsN07vSsZaOG4jrNdetqwocj4FQAtPlmFoxsXcwjqSuNI3FAoKu5UonCFLX5ywg85fBHl+hJFzn24WAVKl3ncaxUfS4AOu5WKuU4PhkXvYAbKL3t148YLPwBMep6TxOThc6MuYG4hosl4ELAU2V7iwu52GdIWgQjusKhjTlisrJ9lHFTs/FtyIIkNQV/P6rbwElpKsOSp0WJ2jm+hStKRBI56ypFdN3nq3mB4iQ8yshpCYP2Cloqz9iz+PuIyO46/BweF+/dHoOf//NS9t9WW1hoySd6PrIUSsc0eg5azQ3SCdjK+w1LTsMMzkrJGbvhDmjXfQdZZujLXLH2NgYKpUKfuRHfgSf+9zncN9994HS1pjHvQZdBdwuuGA6jGM6Z0lLdUqwO2uExA5AKv9M5yxwLlXIswkdCiWYylkoWO6abYvRdJpCgNnCzmLdX8lNEX1cuYgupLvTBmxPoOx4UCnF7rSOuaJTs2BGHQiycRVFf1xvVQPIj966B79277W4uFTBcsWBSilu35+FXhfxbcXmoWoxSXF5RW58A6UrxqRSjOUyLJUcJHUVhkbC5MQte7NN7No2rk7WLLHzzGyp7cROsGHQVYqDw0lYrmykdxjHQsEGFwKESBVzy+WSnKdQjKTl+6qUIqZTWC7vz3s+ggD/xFQeb/vMCT9pFUPFkQksJgRU2lzhA1g/SdRpqFQ2UAgABctb9/ig2cFjHPMFC3sGYuH3D8gm0c1ipexguWRLa0BRVfqgRD6vOdMFoaRmPEXVX4NEUvTnAKbDcHg0VbOGZ+NiU/PF/31iquXj/uML9rV07LF9jQvN2xmH9IIVdB9XD6JkDI/zkLQR/W/GJHmjkcPAToQQAi4TMB0G06uSLKRCdZV4Yda9HrxmunXHOdXj3S5LRnlc4ImJXE82oXEusFxywDjHfNHCSmV109JWQacACAEXnV/vNxKrCyHHUiauYb5owXJYTdE6em417AQDhlOrVdEb/U2w77A9DtVPCgPSxtd2GeYKIrJeE1BKNh3Hb9c6+cGHz+E9D5wJ1Ys1v4muYHp4zwNnAAC/cs+1HXu/dnElJxSvphxNs4KPaTPYkYZzSgm8ugY+y+WYWJbK8Tth1ay4HhQi3Q+FEBCCVJ0CRFUxPaYpEAJXlQL3TgelBPcd24Pjk/l1G6yizfYKJUjoKkbSBhaLck6bL9nQKMELDw3hh27eg/1DCTx2bgkfePgc2mkp50LuiyklYJyjaHM8cm4pjGuCeUaltOUYQpIaFYxlY1gpuxhIaLi4VIEQInRQjWkKiv4ePPhvJgQqtgddIeCCQlcUJHQFKUOF6TBM5kw4aziqdQph8U5IAplKpPsHgSRicCKbs+E7dgufvFAvYifQXhFxvijzAGWbgQl5t3UFiIpFSxVX+d+UUCiKzNONZWNYLruIaQoIISja7prEjiiYqMshiNrGGgFJ8Kh3Oom6rUqXFuI74HJMrpgghMjYgErWS7PaeNBcTX0HkV6Nq6OoX389Xs0bepyvehauNEhCGAHnHCtlB3Fd2RDxKiAFFSwXtifnBxJpxG8ElRLYLqBQAQIKGlgDwX8mAXAO5C13VQNXldAkiSCGRmF7Ai5jIBTQ/BMICHi8SnYyHYbb9mfxq/ccxkLRwe9+/iSShopsXAMAXFwqgwkBjVIIyPoiBMFYxsBswcbDzy3iwz//ApyeLXY1p9TMWblbTiE7Acf2ZGCoFK6vNtToydIVAoUIlLtQa+4GBhI6fvvlN+IvvnwWBcsL88CZmIrfeOkRXD+Wwae/O42y48Hjq2NflRIkDBU5c4d84D46jnZy+T9481iXr6Z9tJOjDwTJ3vfQWTwzWwxdAO48OIQju1L4zBNTsFwGp4Gbm1j1H91B68SO2lxa4OJheQKTyxXsG0qE7gBCCCyVHBAC7MnGwLjA5aVKS3Gp5XHEdQWTftNjFML/h8c5FksOxrLGFZNv2Grs9HHYDfRaw/TN+1t7pr92dhHPzhfD6y47TNbDmVSB97jAdM5EUleQiWsNY7GAoHtiKg8piiD3mjKulTWgqOgfjbKXW0BAumiEtc7Sztu4TGB3OoaFkt123j+a+9AV4gsgbh5CyLr/Ytnu2JgLmsDrxcBE3Q9XWk62jyp2ei6+VUGAW/Zmuz7XdlKcYC3Xp/qaQspQsXcwjoWiDdtjocOsrlIMJHQsdslhY6duwykluGVfFpwL/Mm/Pwtg7XXlSkB9vZlA1oWncyZ+4gOPYr5grXrO0jGtZm4InIzridiWy0LiZ8pQe37OaBd9R9nmaIvc8fa3vx0/+7M/i9HR0W5dz5ZhHXLtpsEFIJiA43HENBEq/EnLTg5VIbKgIwSSmopsvJoQM1QC7jWxduey8BNTZWPBTmoguJKbIvq4chElKFxarlQdGoh0FMjEaxfMqAPB6ZkCTJeF1muqr4AoY0sBj0mmZicDmOv3pPE//vnkupajW7F5WK44KNsMpuv5n59AUWRx1uXCn4cF5vKWVIckBMNJPUxONLJr26g6WacTO9ENw+50NXhIaCrGBwim8xYoIShZUoUhrjW2N+7Pe7Wgvl3sStnBaFqSI1MG9TdJ1rrKGH4vwJa5duwbTOBdr74F3764jD/+4nMt/U1wbRWXYyZvgxKCS0tlJGPquuruayF4auvVhLmfMKcE+OkX7MflFXPVeDo8ksTfPnqpqi7V5P4lDLXj80XBbG2TWzBtjCSNdRtwCYCRpNHwd9sVh/SKFXQfOxONnDW48J01Iq9HiRu9TNiIumDUkytqXDEiRIzocY0IG8HvetGxqRUIAKIHLz6Yu87Ol5A33W1Ltqm++o2mKtidMZCrOFAIwUIHu2c28tGCRsIjo0lcXCw3ff4EAC9SRKOktpmxGQxVQSVQXIssvYHqo8O4r3Isxz+g+L/f+Lrc6jq5ULTwme+WMJWrYO9AAq+8dQ9UdWOiH57H8d6HpNOfrpJQGEM2aXI4nsB7HzqHX7j70IbfY7O4khOKOzFHs1HCbLOCT8Wtxs7BacLhSaoudvXKrL2MoJGbcfnZmCefUeoT5RVCMJIykDe9q06BG9j5pOsXHxnBh74mHWLX2h/XNwcIIeB4AnceGsRb7rsROdNdpRD2wYfP+yQN0Rbx0eMA4b5iHYAHT87iV+85XKNw185qu1JxUDBdGL57wzt++CZM5kx8/PHLmC/I/aOmEAwldcwXbayUHSyVHdkwDtl0bXscl5bKiGkUuzJx2B7bMLEjqSu4aTyDcwtlLJfbmw+jTeY/fMsYXnBwGIIAd+wfwPHJHH7vX06DcS5dOnwHDFXxlf7aLF4IACsVN3QLUChFQqdwIk6dLhM+mZJKRw0hQCnBq24fxye/PRkK1czlW28UbHYt0f9uFPvU/zxfskMlOxn/iFAZlgkR5lbhny8wWtIUImMiX0yG7IDJun79Val0jCBENgS4nnRsoZDKu10QfdwU5Dgj/n60/b+3XIai5WKl4oJCYLG0vlhKPah/HbbHMZDQsVS0wQWg+6Rot8l4L9ryvWxPQFcESOSo6ByxXHKaEtIkh0OAcw7bJ1JK14/qWqJQDpcJUELwtlccxavv2AtKCb767AIUSpGJadK5xGF1ZG4B4e/vCVHCuP70bBHHxjPh+nVyutCV9auTufgrAadni+s20WgqlbO32/w5DhyMegExTcGx8Sxu3pvFyZkCXI9DUymO7cng2HgW6ZiGpKFAU4Dliluzz6UEyMZVGKqyI/c+fXQG7eTyew0bz9ETED86lvVdgut2p5HUFT8eb74YKrT7/SjtQldlHCgiAnWzeQuHRhJwmGyGNFRJODQUBUQDhpI6FlpwDzw0kvRVhu3wsxMAINWmZgjAcj1wbsDtsXzDTsFOHofdQq81TLdSywQkCfjMbAkHhuIghCBlqHK8Fa1w7rA9Dl2heN0LDzScqwKC7ls/cwJl2wvj4Oh718YzrQfwwf5aEQIeEzA0inRMg6ZQTOXWFiskrSTAIcUQCSEwNIpMXMVymzn/6Ds4HSxgCCHzR7my25ExlzNdaJTUiC80AvPvWS/mZPvYPHZiLj6KXhIE2Mi1NMoJf/P80pquT7/3qptX1RRShoqETpGvuFgoO2CMY09Wx0LJhYCsaXY6/rPbT5v0FE5OFzBfsGCodFP9UDsBwUpEiOwDBQgoCBYrDlxWwu5MbNVz9p9esD+cGwSEdNoWAgqlYJGkIPWdbxeKNpKG0vNzRrvoO8o2R1vkjje/+c0AgOeeew6f/exncfHiRRBCcOjQIbzqVa/C4cOHu3KR3cB6gVMnIABMLJtQFWuVOhXzBBRKoPrOJ7VJegKVoqElnlQup8gmNEDsLIW9K7kpoo8rF7UEBVmUCZYQ05XF4biu1iyYgQPB3z16Cf/n389gMKlhsejAdBgIlUWa6PDu1F4zZSj420cvodyC5ehWbB4G4hpMl4Hx2iIXiaiiAUHDmvxd/a2ot2vbaPNFpxM7wYbhTZ98Es/Ol2ryE4QAgwkN/+X7r4fHBd775bPIJlTEtdVLbn/eW41Gz2bKUJHUk8hVXEzlzFBVTI04TnABQIgtbe6dXDFxcjqP/UPJDRXqKn4w8pZPHcet+7J+c+bGEG4U1vj98ak8PvPr37NK7e/EVB6f+NZEw6RfVHEzbTQOGzczXwi01qgpQMFbbFhvdtx2xCG9ZAXdR28gUEviEbIG4yJ8Peq6wbho+bnvJEIXjFWEiirRwnJr3S5Ml8FyakkajY5rt0mt10AJENcUxDQFcV1BTFUQ89XP47r/uiZfi2sKLi1V8NCzC+ueNx1ra1vedQRz10rFgeNt7doahVzrqSQFeRy2x6FQBTfsSmHh/NL2XFQElADnFsrrlsSEkGrVBEDRdNe8n0Hhr2C5iGlqqCAcrsd1izQBwpxCgI2uy62sk7bL8F/+4UmYtgcO2aj0zs+fxP33Xrshd43PHZ9B0XTDRlYuREhoIYRAVeQ9+9zxGbz6eXvbPn8nUEPqzhDYrnRCUimtcR7ciQnFnZaj2QxhtlnBx40k41SF+s3hImw6CcnP3fpQXQKBbwUPmVt0uYAiBGKagmxCEgOuRgXuK4F0fWw8g5vGszg1XcBggmCuYK9LfjddhrzlIWUoeMO9Rxq6GpycLuD8Qgm70jEsFG14nLfShxFCQBYuDZVgPuI0Hcwz7TSnCyGJkZ7DQIl03PiZu67Ba+88EOaGJpYr+ODD5xDXZIOp7TEsFh0wSEITIHNtZYfj0mI5vMZ2oSkEv/3yG/ETz98P22P4pY88ju9czm/gTMDDzy3gxj0Z/KcX7Ec6pmGl4iAT05COqeB+s8xcUQqFbAYB4aGZYiAXsoldIXJuzMRUfO91u3DrvgH87397FhcWSm0XWgPSq0rlfjcKVSGgWLvRJphzFV8URw/n5+rfpGMq8qYXHgvAJ7LI+xXkHm8/MNDWtW8H6tffmE4lsdfxqo2P8HNMPbh9CvJGhBBQVPcKrZLC8paHnOltyo0v+Mv9Qwm89eU34j0PPIOT07JRvdUx5DDuE4oDp50q1rrtXMh/2CJCXKp7S0IICGT978iuVLjW1n/3wVwbXHLw30F8H8T1Xz+7gD988JktWb86lYu/ErBYtlGy1y4kly2G9TjoHM3Vq9sUz940PCbw258+jrLtYTRlhHnKM3OlsHlpOKXj5HQBQlTFFgiRNeyViotj4/EduffpozNoJ5ffS9hIjr7Z3zwzW8TESgUJQ8VsYe2G314jaAIA4zwk346mDeQqHhyPYSZvIa4pOLonjZfdtBv/vy+dxWzBQsVhNfvmtfC6Fx7Aex44I2M0SkEi9a5gzfTLeLBc1lP5hp2EnToOu4lea5hutaZjaBRlx5OxqBBYLjvhHk5T5KBhQsYLH33sMo6NZxvGfncfGcG7X30L3vaZ47i4VCVdyNyQ3CsFaMelnPoKCpwDcV0KL5Vthv/5ypsQ1xQ8dmEJj11YxpMTuVXnZS0GOFwAqh//GqoCYGOCTpovaNKpsCrY7wwmtI6MuYG4BputHVcKIBSI6rWcbB+dwU7LxTdCLwkCtHMtjXLCh0dTyJvOmuLAH3j4PH71nsN4+z89HdYUXI9j3nfukDEPwcSKDQHZy8uiiasOYafvx4N1OpvQYOavfPIpQZDflutw3pJrWzqmwmUyDxTTaejY+uDJOagKQcFyQ5FMxSdyRFF1PWawHMmibmXO2CliV71EIOs1tN1F8u53vxvveMc7IITArl27IITAwsICfvu3fxvvete78Fu/9VvduM6OgwLoFL9DIXIQNSJjCDQPkhkXKFou4rqC6F6Dc5n4bRSA6irFWCaGks3WbSDotQHaCsvqxrEUuBD46rMLPXHNffQREBQ8LhuNuN/oETZYEenMM+BbqQNyMTUdhqShQAiBxaIDy+NdL8qZDkPZ8jCWXd+ZYqs2D+EwD1VQBby6ROCutIGkocFQCeaKzioHjcCubTPofmJHRIpxsjnt8EgSdx0expdOz+H0TBGxjNJnl7aAZs+mbPCrauppKoFCqt+ngIDjbq2irscF/uxLz+FPf+q2TQ1t02V47MJyR66JEGk9C5DIMykbRc7MlnBiOr+qCfSWvVlcvzuFE1OFVecL7zeVapiNsJn54oUHs3i0hQbdFx7M4qnJ/PrdAgR4ajKPOw4MrvrVVrO9o45Bu9MGbE+g7HhQKcXutN5wvutjZyEgXzA/Roi6anCBKknDf03GEp2bpbgQIeEi6nYRJVfUkzPMJi4YlldL0OhBI4m2oCmkloRRR7qo+bnm99XXY5G/j6vy35pCVpFE18KnvzvZErmj2EOyK8HctVJxULGZrzTTeK/bbQjI5sPgrWfyFjSFYr5L1sbtYnfawMSyCQpAVQlsb+17JADkfPVsTZGK2sH8QfzGMiaku99iycXhERWGSmG6HJovT8y48BseGRiD//xG4iEhkLdcMMal4jAXLa8x662TMzkTZZ+YqikEqt+IlK+4eM8DZwCgbYLHVK4im538RtOg0E8QKNvI2G8qV2nrvJ1EDal7rgRRR0CJOg/uNOwkJZxOEGYbFXyAqgI89f8dbTrZqdAUAiak2+hwTMVCyQYBkfOFwFWpwH2lkK6jRY6c6Yb5ZOn84B9DardNLhPrfudBzmRXWgfjAgsluyWVzXoQ//2C3Eowz5yazrfcwBoV+OQC+MTjl/DaOw+EuSHOBX7+w4+j7DCMZWIoOx6mcs1V9jeyV1f8az02nsHrXngNKCXQKMGzc+UNnE2iYDH87397Fn/xlbO4eTyLvQNxeIxDCIGkrgIQyJly3d90GxYhKNkuhO940ei+C0jS8k3jWeRNBx94+Dzm8uaGGu4JpMOJrlLkKtJ1QhIvmgtQNILHAYUI7MroUCnFxLIJh8lYsBAQO1D7rAMIhSqu353CLXs3lz/cCtSvv2WHhZ9zp8BlAqpCoKBabG718ttpKGsGTaHIxBS869W34Lb9A6CE4E2ffBJLJbvhviUqXBIF99dKN8jbtwgu5DxhqApsl8Fj0r0omL8YlwJRMU1Bzqw2qNV/94FrS3AHGZdEzJguZwGbcXAu8LePXoLL+JatX53IxV8JWPQdYdYCB+C08PDIZ61BvraeWdRlfO3sIgAgE1NhunJvp1KC0ZR0w3r/Q+fC/U5A3IS81Np/93HVop1cfq9gIzn66N80a/ZTyPpDYquGTBATtRJLMC77Ywik+FgmpqLsAj971zV46Y27UbJd/OVXzyFXccI1tZV4jhDg+FQBXAj4fWfSfVb4zmqRkwhIF81b9mZ7It+w07ATx2G30WsN0y3VMoGwVpQ3HcwVGCpO1UlO+P1iCoF0X7W8NeuJXAjkKt6qtxUgSOgEpiPj3Xb3HJ7vRrcrE0NcU1C0GQxNwQsPD+O2AwM4uieDD33tAs4vlGB6HF6b+xoB2fga0yg439hOOJhvWjQLaf28BBhM6rjzmtbG3J3XNB9zT06stKTkLyB6LifbR+ewk3Lxa6GXBAFauZZmOeGnp/MoWh7GMsaa4sDZuB7WFE5N57FSkXt8Q1WwK2PAYwLToTgt7wq5d3/W6PxJtxDBOg0ENb/tExXcCtSIh3ABm8v87Lwv0kT8fNJo2sBAQsPlpTJcJvcoEDLP0Cg/K908CAQHXMZQdvi6c8ZOE7vqJQJZL6EtcsdXvvIVvP3tb8c73vEO/Nf/+l8xOCgb55aXl/Gnf/qn+O3f/m3ceeeduOeee7pysR1Fq1JCLYCJWgZ2cOpW3mK+aPuNQkpYbOFCgIKsaqIhAEaSeksKe704QNdjWSkUyJsufv3vv9Mz19xHHwBqggtfpMB/3bdlFBxl28NTEyv45vllPHE5h6cmc5jJW+uemxDg6FgaZ2aLG3bwCAqmTABKg6bDRs4UW7F5yJkuYpqCChdhkUugNnGgEALDb5wE0DVr1E4ndoLkLuMC1+9OrVLwnStUE8J9dml7WOvZDBSCCABal1LmYusEDaNzQMlm+I2PPrHh83R636IqBDQgvYS3iEBROFzG8ZuffBIFy4PrCWgqwY1jabzh3iO4dW+2IbkjwHBKR870ENM6S1LaM5hs+TirlWop/GJBA2w12ztwDDJUikvLFdheVZHRUCky8a21gu6jMXhA0PCL19x3AKonaQghi+5RQkcrCFwwaggVdYSMetJFPfEi+NmqI23YV4ALRiNSRUC2iAUkC9X/na4gHiVn+M4ZQXN7XKv+XTMy2laj1Tm+l3JYwdzleJIwqSp+80kHla/aQc1Q88fm5eXta/SP4qbxLL70zLzvJNbmqi6kMrXqP6uMC7iR4pfDGJ5bKCMTU+F4PIyBKCEYSmqYL3JwITCSqiaeS7aH+YIF01d1+aMHn8Gnn5hseT+99jrphM0/hlqNNaTaMYfjCbz3oXP4hbsPQV1PujaCvQOJ0FkAqBbj5Hft50aIPK53QBDdbvXS+G0XO0UJp5XGmlYJs/UFn4G4hv/+mRM4NSPVvlWFhk3dOwlaJH9IifxuiZCq5ClDQyqmIl/xcP9Lj+D5BwavOiGVTj5DvYCgyPG+h87hWxeX4TAOhfrNznENaUODrgDTeRv7BuP4/Vffglv2Ztf8bEHOJGe6WC47IJAkISFqi0nNVjsCGTMExwa5leg8Y7my6LQeghR7cLVn5kr4m0cv4j9cM4Rj45kaZ9aywzCxXOn4mGVCNpf81stuCO/b547PoNwBQq7lcnz70gq+fWkFALBiujLG1WXMazkM6/BF14Xjcbjwlf2E3JMEp1QjhKCUoeKe60bw9n96Oixyx3UVl5crLZM8NEUSHf/fHzqKwYSG//PvZ3BiqgABeR9Zm8lOAWAmZ2MoqdfUWyiVYl01WyAh1acpldfw1pcf3RFjODouJlYqKNteS40+vQaPCSibvN0bbcByGcdQKoFjezLgXCAd0/D6uw/iM9+dwnPzJQB1Drf+dRJRdZqpxt1yDg0I/ctlp6U5RVUoDJUiG9eQN13YHoPg8jPFNAXZuAYhRE2uuT72ysZV6AqF5cfZCqUYTRsgIH6+zQETAo7HsCcb3/Hr107DmdnmedJ2EeyQQqdCBHspAraFO4ogj7RQqgpcqZRAVykUSvDERG7NQJhSgumc2c9lXsVoJ5ffK9hIjj4abwoI5MouXMahKRTZuApCBOaLzYXigj3lVo3ugEAmxNpuaVEIAHnTQx4yvv34tybwzfNLeG6+BMZ5zQLd6udI+vldy5V5LJUSuMwXTIychBIprNgL+YadiJ04DruNXmuYblajXHUckaJ6y2U3QviVcQIXAGdC7hUNBYSSpvXER84u4i2fOo6CJd2RFUpCt3hKgF3pGCZyJnibezMOOa5H0wZShiSGBr0Uukrx7YvL+KMHz6BkexhJG1ApgeUyrFTknGm26AhpebIxOmlszN2cUJm36GReQGb7SeSnVv9qNR45u4g/++JzLZ2hYHkoWFen2+7VgJ2Si28FvSQIsNa1rJUTzsZU5CsuchUXgwl9VY9dVBz4JdeP4s6DQ/iJDzwKl5UwmjLk3OyPey44ZvJ213I7xXUcJXsdwTp9YioPQgQUUo0Rr3SEpE0BUIWEeTDLZZhaMTGYkDklw9+Tr5VHdVlVRDRveevG0ztV7KqXCGS9grYipL/8y7/EL//yL+N//a//VfP60NAQfud3fgezs7N4//vfvyPIHZ107gAiwTap2tS2kpgWAOYKNg6PJKBQAiIEYqoCh/GQqSyE/L+qEHCxvsLedg7Q9dxCmrGs9mQNzBdtzOStHTWp9LG9aLWFajNT/ErZqRnLjd6PceD+j303ZOm2AgIgbSh492tuhUIp/p9/fAqWx8CEVEAIGGKswWY0+rm1YJPOpDJeM0XwemeKrdg8DCV0JHUFKUMNi1yBC2/YcALUuAh0yxq104mdaHKX1D2JBLVkmj67tD2s9WyWLJlorhdNl4XOrauEBwTOANYGOy92+WzsgulhoWRvSCFz3YvzwX1l8IuLFRCfEU9sgsfOL+PM7HfDJFv9vBr8XLI9DCW0js8X0yvm+gf5x33fDbtqrqsewXXftkYyYSvH43LFQdlmMF0PTMhiaTC9my6H49mI6+qWWUFfiRBChOtkUCDngXuGkM994KYRqGdK0kb1uABcCNg+kSLqYBESMeocMEKihcfqXDMizhgOg+XxlokgvYoaF4wIySL4OR4lYqxBuggdMlT5uq7SVcmyKw1Gi03ucW3T+swdw7LfxO8yBsWP0TrpOLMZEALsHYhjdhtdHAIkdYpzC756dxsF+7hGUXGleo+iCBCQkNgRPUc2pqLkMEmsVCkAETaKCQHcOJbGfNFGyWZQFak2PJ03fUI1wfhAHLpC295PN1snh5IGlksOtCiJ1AclFKrCUTRdfO74DF79vL0t3g3gh28ew1s+Vaf4XHczKSH44ZvHWj5np1FD6t6Vgu1FSN1NnAd3EnbCXqVm79WCkEEz1OervvfICCgleNsrjuJNn3wSy2UHjPNwz7qTEFyybLCn4R5VNuZz6QpAPBwaSfZM4W0r0alnqJcQFDk+9vhl/PmXnoPtcQyndMRUBTbjmC+5GEhoeNsrjuK2/QPrnu/YeAaHR1N47MISOBfQVPkcCV9JPixEAaucIAgAXZX31WHAroxRk1uJzjNPTeRQsNYnSARkPy5k4er//NuzyMRUXLsrhbuvHQlFgWbyVlcU8XSV4Dd/8Hq8+LrR8LVHzy22LSgxnjXgMNmgvVbKwnRbb3xpBcH3w5io2eMT/59yrgB+5q5r8PBzizVFbiEEYhpFxWnNzc9lAmOZOEbTOt7x2adxYXFzcZqmELiMY74gBXMUX8wm+ASUyL2eSgmShmyMv2EsjTfc2xtrVqu4+8gIfu9VN+M3Pv5EjdhPN0RAuomNCBWplCChK3A8FhLK24UAMJ0z8er3fwMAsFSS7kPMVzFM6gqKtief4ahCiw+FEFAi620KIdidiSFhKChZHpbL6+doCGRcv38ogckVE9cMxxsK/zTKNdfHXqpCAY+DEmA0bSChKTBdhlzFha5QCMEwlJQqnabDwveIaXRHrl87CdO59YW7WgUHGjo99sLuweMCntO4Wh7IGARq2IwL5E0XM3kTN+/NXPE5nT5Wo51cfq9gIzn6wNUub7lYKto1691krvH71IgCYuvW9IDo7zGxeq+D1vthFoo2FortO+VGP+d9x3bj1GwBxydzcDwBN+IuGL0XN46l8bZX3LSjYrdewk4ch91GrzVMr1WjjMJjAoZK4TG27oTRrH8iyFuWbQ+UyNg2EBQSVM4NswUb3gYC97Sh4sBwIkI8rvZSNGuY1lXZF9KuONNyxcVyG302UXCOtt0vU4aC0hrN0goliOvShe/iSmufpdFxwX2yvdY6Ex2P47b9Az2Tk+2j89gJufidgPX6UgOslRPWFAUKlQR4y+WhGHGAenHg07NFzBcs7M7EENNqjx1OGchVPFhud6j7O7zNIFyn3/qZE1Lg5CohdkQR7BGoLx5HFJlTXSjJ+HtPNg4WiBvW3RyFAoIHjh6y9+Dm8eyaudCdLnbVSwSyXkBb5I7HH38cf/d3f9f09z/7sz+Ln/u5n9v0RW0FdJXC7WDRJEDQ99IOecT2GCZzJjIxVTZnc4GBhAGVUjAhULE9GJqCn3vRNXjxkdE1GUnbOUBbdQtppJT4Rw+ewUze2pGTSh/bh1YX/M0EBisVF5yLNRNxwj+uEYaTOkq267Moq68TAhBKcGmpgj3ZOJKGgnQ8IEHwMPmX0BUYKkWu7IZzSlBMl6qipOYzNkuoN3Km6PbmIUqoCIpcFcfDfNEOnUZkI2h1y90ta9ROJ3aC5K7jcUznpNoPF/J7MVSKkVQMbiTJ0meXtodmz+ah0RTOzRdk0p1xSYqEgBtZcLcqad6J91iuOEgYKkbSBgRkgm2jyMQUFCwGjwtQKkJiaFB8CxpJOBBhnwowASyU5PwVkMXq/5ZxAdNh+OEXHcSJyRyemS2G63zg/LHR+eJ4s2pIg+NeenQ3VELg+qqj0dETfCKVEJA2FZu7NR4H4hpMl4Fx4au6+zbkRKqlekze14G41tH37XV4fgNzlIwRdcwImsjD3wGhg4bHOWyXo+J4qDiSVFHrhhG8xmteqyFnNHHFuFJcMGpIFo0cMdQIyWKt4yKv94oLxk7E159bbOm4B5+exU/feU2Xr6Y1DCV0UFJdC1wmFQ17IdFGiHSoS8d1WKXtI8YRAAMJAwXTCZXQlBYrSLbfuCUE4HoylvEauKKUHAZdoSAEuGY4id971c2gRCqqB+vWN88v4f1fPYezc0Uslh0wLpDQFezKxJDylc42sp9utE4+dGYOf/zFIpr9OSUy/zHVJvHmzHwJCZ2iYDXPniR0ijPzpW1L5EULAJRSxHUAqCbvr4Smul7fqwR7L73JQGtFIGC9fNWf/OTteN9D53BmtoiC5e64uID5OYugSRVAaPGtUtq1ffZOQSeeoV4EpQQ/c9c1ODySDJ/vguVtKK9DKcEP3TyGR88vAQi2jAKeWL1G1RcUpQsVgcekg8JP33lg1fwRnWcWihaOT+WxULBxYamMR84tQVMkWZDxKpkkmkPTFdnIf3qmiLPzJTDOsFRyYLpex3MAaUPFe193B+65viou8MjZRXz5zHzb53I8gYLlQYBApZJkPpjUUHY4hE82dzdq49siomcfSupIx1TZyOsxOIzjqYkcdJXCdBhiGoXtCSQNFbbLa8jwa2FXxsDbPn0cl1c23wgdCEWH7+zn2kZTMRgahUopBDjypoc3ft91eN41O9eNKBvXEVcp1KSO5bIjG7AEapzcrxToKgEFgeNxpAwFRcuDAKCpFIzzDalbapTi5LR0VtiTjWNXWkfBcpGrOChHGtUbPcYul64jKlVAKfFd8hSolK4isAGN808HhpP4rZddj7f/09OYKzgYSGhI6ipsxjFXsKErFHdfO4KT04WGAmjR2GtiuYIHnp7F+YUS5kt2OI/ffe0I/vrrF+B4HDN5c5Xa/HDSqMk/bxatNshcLRgfjHf9PbayARyQTeCGpsj8mv88rYWgAaX6HzLH/KdffA6PXVjCkdE0btyTxsHhJAyNQleoJCz1ccWinVx+r2AjOfqhhA7bYy2RkgME45n451cpEFekwEe3x3joBFS3gAXlYorq2kYA7MnGYHscRcsD43xDZM0A0bqMolD8+kuuxZs++SQKptfwc6djCn775Uf7zaubwE4ch1uBXmqYJpSEtcy1IF02KLJxuR8wI8VuShDmjS2HAwQN8zpB3jLrz3XR2ikBgUJRI47YTtyhUgLOBZwGvRQnJvNNG6Zl/lTdUqX5drcTqkKhUh7WJaOQuX8NukIxlNAx2SJRpdFxtc6ja++XFUrwP195DK++Y2/bMXg/jt9Z6PVcfK+j1b5UYO2ccEynMFQFFZfBZRzxSM2nkThws3MFfQ4DCQ0z+e7Me1dCnujuIyN496tvwbu/cFrmcnb+R2obrsdBNSmoJP8n43BDlb0Zlivrx7I/q/p3lBAQRe5bQIBffPEhvCXiNt0IV6LY1dWMtsgdc3NzOHjwYNPfHzp0CLOzs5u9pi3B4WEDJ2bbZ6wHjRhqA2v6KNrZCHMhB6FKKYQQsDwOs+ggritI6gqO7c22vOnYrgHarltIlGV1YjKP8wv9SaWP3oHlMtgux/nFEh5pU5lv70Act+8fwG37B3DnwUH8wQPP4NR0HgoRsD2pIAbIeSRvevijB89gMKGBQxYwrxlKrFKDnVixkDQUFO3GTF8uZBFM9ZvLhKhVaFnLmaKbm4cooSIocg0kNKxUXFgug6YQjKarhK5uW6N2MrEzlNDBBcdkzq4JrLgAPIfDylUwENdqkix9dml7aPRs5soO3vgPT4DZ8qZHmyACFbG4rsBxeU9vcoLkGeMCs3kL144mEa9j+Ld/ThIWDqzNkFfrb5uoNmwuFu3IOwn/35ubK/JWa0osectFznSR0FWUHa9hI6xKCRKGipy5/jm3ajxG1WajjTFhPHkV5mk+9tjlpkQLy4u6ZjQmZPTw0G4Jq1wwNAVxvTHJIlZ3XOCCUUu+kK9dDS4YOxGFFuYjoDlJeDtwbDyD/UMJqWDPqq5OvaAizLnAQtHC4DaT4gSA5ZKFhK5i70Acl5Yqa1rlRsGELMSl4lIdrJFLX0DqCJrLZ/MWVEpXrVtBrPTZJ6fxe/9yCgldQTahRSzjN76frl8nz86XwmaARtsELnOK2DuQaOn8AZYrDggoFMIa5lAUAhBCt7Xh+kptCq9HL+9VhhI6NIXAYRwxujpmXo+40Gq+KnBB+ON/PyOJ5JEk+nbPf80Q7IEg4CvuA8K3E/K4QFyrOsx0a5+9E7DZZ6jX0am8zv6hBFK6Co9LMQXOEeawGoEAoBQhcZtSSf5/7Z0HGh5fnWeyeOnR3QCApyZy+I9/+SgY5wCtdQmJomB5yCZ0jGUMXF6uoGx74KJxw9hm8Zrn7w2JHZwLnJjK4/f/9TQ8JuRa2Ma5irZsmKOEgAuChEGxZyAOCODychlCEMQUwPR4Z5w8myB4EmyPYUw3MLFigXGOP//y2bAJcLHkgEI2IslppbXrIZCNY51aBxvlcgiApbKDvYNxxHUFnFMUCMPB0Z3tRrTsO7pkYhrypusTmbFtkoyq/913I5/GmQAjAhxAzpSNshQyLyJd4drPY5kuC9fAvOliMKkhm9CwXHZQcdgq1996MAFUbA8KJShZLpbLDnZlDBiqssrFoP6OGBrFb71MuvvU55oDB1EhGP766xfwd49ebNhoUh97vfbOA6vm8ZPTBXzoa+cwnXdCx5qo2vx0XgrGdWL9aqdB5mrBD9+yBx/62oWuv89Wjvgbx1K4tFwBhT8GAcBvCgdkU2ErJOdTMwWcmimEP8c0ioPDSRwaSeLwSBJHdqdxw+4UxrNxGH7Oqi8ccmWgnVx+L6HdHP0Nu1KoNHG0WQvR8WyoCu44MADOBR45v7zBK18fwbrdiJRLfbK/gBRv4X5/i+u7Bdx0eAi//OJD2JWJ4aPfvIR//PYkONonHxPIeWCp7ODe60exK21gueyE4k0AoCkUYxkdZUfgAw+fx93XjvSbWDeInToOtwK90jBdX8tsBIUSvPSGXfjKmQUMxDUMJFScX6jA9hhUSsJr9piAyxjKDm+Y1wnylgNxDcsKheUyUCL/PhDwi84P7YzunOmiaHvIxDTcNJ6piQvXzZe24VTeiFzdbTgug6ESWC6gUfjFDymIGFMphACu3ZXCsfHMumTYAI2OC+7TSEpb93PGNYofu2287ee1H8fvTPRyLr6X0W5f6lo5YQKCgYQOq2Ahb7rQVBoRB3ZWCTYMJXSoFLA8Kcxmugxlx4PlyH6Gbub1BuJXBoH+7iMj+Oz9L8Yf/dsZfOhr5/3ext6tvXQKakTkz2U8FOcKHplsXAUhBB7nkK7FACdV8XDXd2Y2VAJDU1sS8rta6ppXC9oid1iWBV1vnijUNA2OszO+eIdvbPJTKQ2D4JGkhoWSsylFA0AG1ZpCMZSUi4/NGBZLDgyV4g3fdwQ/9R/24/RsEV99dmHVJqSehbtUsrd8gLbqFnLnwSGcni2u2kz1J5U+thuOx2G6Hs4vlPHNc0t4YiKHJydymG/RgpYAeP3dB/GqO/biwFACcd9t4+mpAp6bK6HiK/I1misEZAEqpimwPY6JFQu7Mkao9DWxYqHieEgaCoaTGhbLMhnDhbxuhSAs7vyn/7AP3zi31LYzRTc3D40IFQldgYCArkg1cM7Fllmjdiqxc3QsDcvjNcSOKBgHLI/j6Fi6A1d99SL6bD5ydhFv/+en4bKq4rVC5b0O1YEoxa50DJxzXFpem8BJAAzEFXBBkG9DgakTCK6XC1mMPrdQRtLYHLmjE5+B+cX2YBMVNIsFBfGv+Eql0c3yM7ONN8utYjjVWgF6OKVjKKEjaShIxRTkKr7LkU9mM1SKbEIDBHqmKStnulAVAttbndQLNmuqQtYlo1xpiiv/459PbvclrAtZkApIFDRCrogQLKLEC126YDQ6LlZHxugXs68uHNmVwjNzpXWPO7and5ptKSX4zR+4Hj/34cfD13olscYBVByGhK6GBIP18qUqxYaUgNeD6Ql43MOhXSlcWmrPrQIEePMP3oAHT87h+EQOBdsLXoaiEF+ZRSqxuB5HwXKxVGq8L6GUYCgl3VYysVpiR4BO7Kdfeese/M9/PoGCxcA5A6VENgRRKonmTCCb0PDKW/e0dd5AQROQiUnp4CRjD0p6w+XqSm8K3wmIOkKOZWjLQgZAbb4qG9NCN6KYTrE7o2MqZ+H3//U03vXqW3BsTwYPnpyVDpqaAsuTz2avzIH1UCnBSEo2zTDOISCT/K4nr1ihBJm4hrmi0/V9dq9jM8/QTkEn8jrBfiuh6xACmMqb4O7aLf6jKQOEEJgOQzau4q0vP9rWc3bL3ixuGEvh6alCU6JkQDBdKNowI4r/nUbwPt+6uALOReiQdXqmgOWyI5vjVFqjeroeggbZgCTDuEDZVy8t242JjZ2CQqriU8HbVGyGc4tlOL77bD04AHDZDKIQ6agAyHujKSR09ozeAgFsyoV0LRBIZwcK2bi4ULSQ1JOwPAYhBC4ulHf0HjmIMQIXYNPlULexXs+F2KR8R3N4AqsWVA5Z1FY38N0RSBKjSuUNsz0Gy+GI+y52l5fKLY0vTaUYSxuYLlioOAwTyyaGkzpMt7mwhK4Q/OYPSmIHUJtr/vrZBfzto5fgeAxDSWPdRpMoGs3jR8fS8LhsspX9cfJeUSLdgBxP1hs2m39ut0HmasEt41lofhN0M/SCCEI7uGYkidOzpdDBIGgydJl0s3nN8/fhX4/PoOCTsNDivttyZY74mdlizeuZmIpDI0kcHEniyGgK1/mkj5G0AU2h0JQ+6WOnoZ1cfq9gIzn6f3l6dtMNejFNwRvuPYKT0/mukjvWgscBQoRU+4Uc0tcMJfD2H74JwymjJob6Ty84gC8/I+s/i0ULThu5NAGgZDP87wfP4J+fnMLksomhhIq86cHl8gDGOZbKLjLxvqDnZtHpcXil1Z96oWG6US2TcyEdgxSKdEyFRgledO0Ivn52UeYcNQVj2RimVkwwIUBE4HIvkLc8DMS1hnmdYE+Rt1wwvxeFSWaVzDU32F8Q0piIEMVoSkdMU5A3Pegqxa/ec7jlhmkAsFsUIYw6C20lCJF9HPJ+VV9XKIGqKEjH1PB+a0pr46HRccF9Wiw7635OSoDTs8W2nt9+HN/H1YS18vxjGaOhg/16OWHb47hpTxrZuIbzC+UawQbOPfzlQ2fx/331LMaycdxz/QhAiBR7E6Jl4lcnkDO3zgmp26CU4C0vuwEnJvN4ejoPCBGKgFypiO4rGAeYX8dR/aY3Q5XraEDK9vyFKcg3BE6/HhdItliLbLeueaXFg1ca2iJ3AMCHPvQhpFKphr8rFosNX2+Gd7/73fj0pz+NZ555BvF4HHfffTfe85734IYbbgiPsSwLv/mbv4lPfOITsG0b9913H973vvdh9+7d7V56DTxvY5OD7zQPzgRylidVTQSgKnKTvJHgkwhgKKnB8O134lTFvgEFswUb//Cty3jg6RmcXyivYtoCWMXC3ZWJgQuxpY0HrbiFnJrO4yc+8CjmC9aqz9Fvluhjq+F4HBXHw4WFMh69sIQnL7dH5qiHphD82O3juG3/QM3rSyUbBcuFEGsXw4PAL2koIAQoWy7yvloMpfL1/YMJEELARKVG3ZlFiB2//+O31rDzt9NyNIpGhIq86eADD5/fluvsRGLn5EwBzjqJCcflODlTWPVc9NE+OBd49xdOYzpn+sksAgFRQ64hBBgfiCFlqDAd1rC4RohMFBFCkNAVXDOSwnzBQtn24G1jJc70nQm6oU5CID9zK1s+DtTcNAF/fhK+cgrnGB9INCVxtsISr0erCT7b5TUb8IPDq12Oek2ReCCurdv443h8zYbVvuLK2tAU0tDxQpIsa39ueFwN+aL6s9GmCwb1FVYJyCp1UOI3aQMRwpRfHV91bPgPrGrQbhZJNGrkrj+24XWF19f656z+TYP3J7W/qz9/o79t/j6N7w/QvOjQ6NhG11n7Wu01Bq81eovgiOjvRIOLibqRAcC7fuxmfP7E+s6Wb3/50XWP2UpcWqmAUjQksa63VnW7uYULoGLLYrDrcZTXUVLsBrEjgMsFvn1xBUB7n1tXKA4MJ/G3v3gn/p9PHcenvjsJ+H/PmAAnslFN8VXaOBdrurtE99MGobBcHq6PMY12ZD/94UcuwPETikwgbMClnPmEW4L7770W6gY6EsN5g5BQPbZKLhUtzRvdxNXQFN7riDpCtitkcHK6gFPTBVQchrzphuShoJHUZRynZ1z80t98C/sGE5hYrmAoacDjAhPLlZaV67uNYI4JCvOGSrErE0fKUBHXFcwXLJguk2rjvuJVXFcAgW3PB/QCNvMMXU2IznfZmAqvQSBA/WYUx+MQkG4Poykdt+xr3XW65nyU4EdvG8fpmWLTBjoB2XRetj2U7M4W+zRKQrcEIaRzyNSKiXd+7iT+7dQcHE+OK0pkAyxrpvDRAhQqm2enVkzwJgIwnQQTq+MTjlqnz2bxi8cEDI2ETiUCsoC4lU03lAR7p6pzmO1xLJUdLJZsUELwF185u6P3yNExN5IyMJ2z4PH2HWI6he1oqhLYmFOIrlEwJsIcIRfwFQ5lbanVJgvGOTRVwb7BRLiWLlccDCZ0WB6D41ZddVRKcM1wAv/rR4+FxI4AlBIcG8/gDx98Bi7j2JONdyR39olvT6DiE8Jc3/abEkAhBAIyBlcIabsJLIpWhds2kvPb6Tg5U1h3k0cJuj6fdxInJgtI6AoYlzVkxuVcm9QVKJTi8pKJm/dmcWIqD48J2J7c76m0umbuzsZx1+EhXFys4MJiGbMFq+n7FSwPT03m8dRkvub1kZSOwz7p49rRFK7fLYkf6ZgGXaF9p48eRju5/F7BRnL0l5fWF2pZCwTAnmwMdx4cwvseOrstqvQBXCagUoG4riIb11BxGIZTxqp149h4BsMpHU9PFTa0CxZY7eoD+PUpP+isOAy2y5AwtL6g5ybQyXHYrz91B63UMo/sTuOVt+7Bp5+YDHOOKUPF3sE4FooWbN/dUVcobh7P4g33Nv5OgrF7crogBYUpkU7RCPZx0hHCjhL8G+wVoxhMaNidicFyOQYTBHnTxV9+9VyN4866DdNua43I27ZaECKdy4jMN4S3RghcMxzHW19+NLzfa8U6UTTKVwTfz4mpQoO/qEXZ4Vgst96r1I/j+7ja0CzPb6gKRtNGQwf7VnLCb3350bCX7atn5vBX37gA0/ZkE76Qc+WKWcTp2fX7oVV/Du502Kdud5Gsw6CU4A33Xou3fuYEpnNmWBfcQVvrTSPIh8U0iorjIRvXYGi17EuBIBdPIISA40kCZysCH+3UNfvxYO+jLXLHgQMH8MEPfnDdY1rFV7/6Vdx///14wQteAM/z8La3vQ0ve9nLcOrUKSSTSQDAm970JvzLv/wL/vEf/xHZbBa/8Ru/gR//8R/HN77xjXYufRVmihuzIrQ9Li3nhVTRBGRRa+9gHAAwlzdhNVG/agYOYGLFDBedlKGGKtSnZopIGwpG07Eapu2bPvmkP3g5ErqKdEyql04sV3yLPx42gwfoVuPBes4brsexUnHhshJ2Z2KrGMO/96qb+80SfXQdubKD84tlfPPCEp64nMNTLZI5RlI6chUX3A/CFCobBIWvi0kBxHW1oWr+iv93hACcN98oUwo4jGMoGYPHBX76BQdAqfybj3/rcjgnlGwPZVs2gIfJQCGT7984t4RHzi72jOXo6s+4mlAR2Nj10nW2iicmcusq93hC4ImJXJ/c0QF87PHLODVTBISAqlC/8ULa1wXN0oyL0MKuaLuNxxoB4pqK4ZSOisPw377/OkzmTLz7X0/D24DNdq9AIUHDy+rfCbRG7IgiOgqDUxICDCWNpiTOjaotuS02x7iMr9qAx3UFCiGwPIaVCqtRUlkPW8E+50KEzHqg8X316qzgo7hSFVcOjyabkyz8n2NNSBcJXUVCl8cldAWaT8KIkiYCInagSBYlXoAETUK1f1PT8F/3mjyulpyByGt9XM1Y6xmQv7uUs6BRsmbDkkYJzi1XcEuiN9TqHjm7iD//8nPgvLHrheJ3r5C6tWcrR4SqUPzwzWP4xLcmtvBd10Y7SUdKJNHim+eX8PDZhVXnEQKh6hD3mQ65igPORcO1KkjSPTWRB+Mcjv+3hEgiiUIpbtuf3fB++oMPn8N7HjgD1qDZkAsgplH85g9ej1+559q2z53zHQxLjMN2ec19DBoAYpqyrstVNxGNP2byVhh/MCFdRdqJP/rYOBo5QrYiEPD1swvI+U0jqkIBIpu3TdcnKEF+x7pCcX6hjLLtIhVTZYG3R2Co0plnseSAUoJdaQNDyaq4SspQQbMx5E0Xb/y+63D7gQEAcnzttH12N7HRZ+hqQnS+WyhJVcn6eVlVKBRCENMkwSOuK/h/fugofuz28Q09Z4+cXcRHH7sMQ6VN9+QErTdqtwOVInTnlmrC8k3ylouPPnYZXAjENQWGpoSEbUUB2AaVITgPxKHaqxsAtflECiCbUFG05P1aKze13tsEpLH64wSkC5CqUGRiKpZaUBntNFKGiqLlwea1wehs3gIhwEBKQ1yjEAI4PVPYkXvkYMy99dPHZVNBTIHpMFjb1f25Q6BSgpSuYrnsQGbj5f2aK9ggRD4jrTaJcQEULRcpQ8XujIGC5cHxON72iqP4sdvGcXKmgCcmciACuP3AAG7Zm20617UigNZO7izYl0mHkmqDCBdy3oppFLszMVQctqkG2U5f95WEVvL/O2245k0HuzMxGOpqUQDL4zi/UMIbvu8IJldMlGwPSUNFsFpwIZAyVPz+q27GnYeG4DJZn86ZDp6dK+K5uRLOLZRwcbGM84vlNQUKFksOFksOHvfFEgC5Ho0PxHFoJIlDIwkcHk3hhrE0Do8kkTBUaJRCU4iM6fvYNrSTy+8VbCRHf3p2c+SOoaSG5bKDzx2fwTOzRVBCmtYAOgWVyny26+fsIKoNYWOZOAaSGgQH5kt203WjZHfeIU+glgTHBVA0XZiOB9NhMFTa36+2iU6Nwyu1/tQLaLWWqap0VdNxQlOwOxPDUsmBoVK88fuvw2vvPNDaOBEAVYjsMfHFeQSAXWkDKxUXZdtbtc+vuW4AIEDBrB4b4PELK/jY45fxM3dd0/Az1jdMJwwVJdvrWRIsAcH1u1KwPQGXcZgugxAcliuQiWm46/AwADlOJnOtES7yTZTnWxWpYFzgqYkcXnLdaEvfdz+O7+NqQ32en/h9u5bLMLViYs9ADG4DB/u1csK/8uJDuG53Cl96Zg6ffXIaD2zAvc3w+4XvO7Ybn/jWBAqW21A4bzNwenUy3QTuPjKCX/new/idz50CjWzxVIUgrirIW25N/TlwnSLYRmJgFxC4e7msgmxcW7VGKxHXDsUXA/zc8Zl1c/GUEvzqPYfxlk8dx+VlE5m4ioyhwuGiRuzqm+eX+vHgDkBb5I6LFy929M0feOCBmp8/8pGPYNeuXfjOd76De+65B/l8Hn/1V3+Fj33sY3jpS18KAPjwhz+Mo0eP4pvf/CbuuuuuDb/3RpW2BCIKpkIWWYdTOoSQ6pqHR1NYrjiYzbWeRAZkk2Gw6OwdjCNpKLKBQwhk49L2DpBM290Zgmdmi+BcNpqXbOYzEilGUgZcX1FltmBhIKF3XY1uLecNIUTYQD+aMmo+R8AY/sDD5/Gr9xzG2//p6b6CXh9dww/+ycNYKK2/+do/GMdt+wdwx4EBvODgEIqmh9/6xydBCEHB8kK1IkIIYqoiF1ghGirhDiQ10BbYuUFOz/E4lisOPvLIBSiUgnGOoi3tLg2VYqFogQlJpCJEEkw8JjCQ0FCyWQ37/pZ92bB5+WtnF3uyqaMXrFE3iqgbS3BHRd1/C9FY1buP9sC5wCcel40VukLCRitCAE2RiXeFSALDQslBymNYKtVu2oKEFAXBSFqHSik0yjGY1PFX37jQ9cR6q9iQ+xeAdFzDL959Df7q6xehUEkE28gmL2WoMF0WNrUAcsOQ0JVVibwoDIUi32Cz3ArSRnPXikbH3X1kBK974QH8xVfOYnrFAYf8ftNxFa974YGWNhdbxT4/PlGrStfsGzk+kccdBwZrXruSFVc+8Z/v8pVvSUi0oD4ZIyBiVAkZVSLGTvucffQBSBe3dVkPxD+uBxDMPbbLoVBAoRSqImMazmWhmxICz68QE0KgKbLwIJoUhrqhTvgDR3fh8oqJmKag4rBtUXOhAG7eJxVN2w0jBpM6jo6l8Qt/8y04rlTfDe5f0GApVVvk8UQIvPehs/jymfmGaxWlBPdcN4JHzy2BcQFVIVD8Il7FYVAoxz3XjWxoHvU8jvc+dC4k1Db6qIaq4BfuPtT2uQGZS1DXuS61B1w8OxF/9LF5tCtkwLnAgyfnpHW1QnziVO3enPvHgQCjKR0l28VMzgSlUg1pu9TTA2gKAeOyQGyoBLYnkDddDCWrY0IIgbzp4eieDH72Rdf0Y6Y10KtiGL2EoOD5hw+ewfGJXPh64NgRCCoI4b9GCYZS8nk8MZmvua8A1rzX0T1PXKNN3bi6tc579daVkTf0hIBKAcvjcAo2FIXAZXJOaBX166asKWzM/SL6JwlDwZ/91B14xz89jZLNEPf367lKY5GLds4dhaYS7MnGEVcplitOVwg2UQT3i/g/FK3GzYXBMbmyixW4IZnVZXzH7pEzcQ0zeStco4LnrDcyVduP6FiKaRQDcR1LJbuWfEZkXv3yslmT01oPXADzRTusXwU9UX/9jQvYk43h7iMjLYv2rCeA1k7ubPW+jISxTLAvUyiBSsmmHfo6ed1XGlrK/2/5VW0OAghrS3FdQeB2A1S/6/1DCbzuhQfw3ofOoWg64WdOx7WavY+qSKe4bELDNcNJ/OBNspHYZRyOxzFftPDMTBFn50s4v1jGxcUyLiyVUbabr/dTORNTORNfP1t9XaUEB4YSODiSxOGRJA6PJnH97jT2DyUQUxXoap/0sZVoN5ffC9hIjj6ub+55KjsMhBBM5Spwve75QUb3q1W3m9oEnSRoExAQWIw1XTdOTOUxuWxu6DpednQ3fvIF+/HYhSU88PQsJlbWPg8H8Gt//10cGU3h1v1ZPP/AIO48NIQ9A3HEVNofz+ugE+PwSq4/9QpazSU2azpu1R3z5HQBSyUHe7Jx5E037GMBpNtdOqai7DDoKkU6FkfedHyCpy964tfhmJA55uB1FqkxUyKbSv/8y8/h8Ehy3Ws/uieNl920G7/3L6chPN5yXk1p4iTeDexO66CUggkPS2Unct8Evn1JEllee+cBvP+r51pWm9AbOFq3O7f+9dcv4DuXVlr67vtxfB9XAloV46zP80f7hYjfLzRfsDGY0BrGOXcfGcELDw3hiYkcnp0vYWbFxFzBwh88cAbPzRfhttFXQwmwJ2OACaBsMyQNBf/rR2/Cl07PI2+6XREAKJpX5jjeP5RAJqYiE9fA/Z7rmEZRsj0UbU/2gnFgJKlDUynmChYIfGL2dl98hyCEJO943AuJuUowBIhP7PALyBwCJdvD737+JD79xOSaa8UjZxfxgYfPw/E4TNdD2XYxRwkyMRU3jcsY467Dw/j5Dz/eNB6cyVv4wwfP4L95HMMpo19H2Ua0Re7oNvJ5ucEeGhoCAHznO9+B67r4gR/4gfCYG2+8EQcOHMCjjz7akNxh2zZsu9ocUyg0tjijgSz+BhEklW1PYK5gh1k9XaFSz5/4NrWQpIf1wARAIcC5wHzBwlg2BttjUhGSCxQtN5zIVipVph8hJLRhNl2O6ZyFkbQO20VoJ91tNbq17HxMh8H2LezjRi3xI8oYzsb1lhX0tkJtu4/tRavjuB00I3bsH4zj9gMDuH2/JHPsH0z4iuHyWeZc4MjuNE7PFHHNcBy2G7HO1AjmCk5TZ5mRpIFMTEXe9NZsHg/G/1LZBhdA0lCRiWkoWC5ypoupFRO7MgZsj/vqK9ViumxwVzCQUGrY933rrO4iE6smxeqL9c2Ou9rQqXF8crqA2YLlN5HUKU9ANi96nCMd03BoOIFTM8VVjPqwZ0MILBRsxDQFN/lj9tx8CYaqwOyCVXjAIO8GqB9zpGMqNIVi/1AKSUNF0lAxPhBHxWEoWG6oOtgKEhrBweE0cqYLl3FoCsWAb9Vdtj1YLvMV22phM77hYnKrxJrguEfOLuJDX7+Asq90EjwRZdvDh75+AcfGs2vOcVE1orimhMXxbih9CiKfAcVvbI427koCg/96gxCmVxRXurEe70rHNn2OPvrYKQhc3Bo5YABSMZpzsaaiZSfQ6lgO5p6RtA7PFx8IEqWCAtwTsP0PIiBJBwQEukp9Uqtck5sRPToBCuDAUAqPnLvs29Vuj/NWOq7hx5+3D/N5E4pCsVRyYLW43mZiKr5yZh5n54pIxVRYHoPj8abK5LpKEVNp07WKc4GHn1tE0lDgetK5w+WBYxmFplI8/NwifunFh9veN3/u+AwKvmtGuIb5vwsutWi5+NzxGbz6eXvbOjcAHB1LgwnZ6Kqr8uzBHgcQcDzRss1wNxHEHxW/QYNCPvsVh7UUf3QC3ViTdyLaEQg4OV3AXN5ETFVgeWzNuHM2Z2E0bQDwlbC2WQ2LAqFjIaE+oZ0SKFTAdBlypotsTOuLomwA2ykysVPG8d1HRvCpg0N4xZ9/Dc/NlXzSoMxxA4CAdMDRFOmoN7Fcwc9/+PGa/NOwT/hYKjlNc1Inpws4NV1A2fbC+KJVNCMcAugYKYv5ThtcyNZXhTR3yVAjTTBrXedmC7zUF3caTBh494/fivd/9RzOzhVDlWUF7Tt3NkM2poIAmCuaW6ZMrypEqsuu837cF1PRFJkjtT0O2wNOTRd21B45yE8slx0olID5n4v75Cn5zHXiqncuMjEVr7pjLwiAB0/NwXGZbJZA7R5LCIBhY5uQmryuf++ncmbbOaK1BNCA9nJna+3LqCLV3x2PY6nk4JZ9G3fo6/R17xS0Oo5bzf8DzUUN1lqvtgMC63/XE8sVfPSxy1CIdNIIHAcqDsNHH7u85t5HUyg0hSKhAwMJHdfvls9mQPhwPIbJnIlnZoo4t1DC+YUyLiyWcWm50jR/7XGB874byJcjr8c0ioPDkvBxcCSJa33Sx1g2Bl1VoCt90kc30G4uv1toZz3eSI7+0HBqQ9cVjHnL5TAUjr0DCWgqAbqk6RKMmiCf5PFqzi4KlVIIIVV6m9Wzv3tpZU3n4bXw6IVFvP57DuKtLz+KR84tAeuQOwKcXSjh7EIJn/7uFADgmqEEbt2XxR0HBnHnoUEcGE6GJK4+qujEOLyS60+9gnZqmZsRogga/HeldQz6rkHLZQcekwKhK2UHhBDsHYijZDNcM5xA3vQwk7d80TWE7kbN9rvBy3nTXUX6aXbtXzu7iJimyEbVFh5ZTSHSUXOLIqeyzeAwgdm8JffSlID47YMO4/jzLz0HQPYQKJTUkF2aIdZgrmp3buVCtFyv7pU4vtvjuN8beOWinX62aJ7fZhyUinD9IiCgRAqg78pU4xyPcVQchlPTeXz70gqemszj6ak8ZvLWpq6bC4BQipGEhpEUMLFSwX/5+JMo2d0hdgBAqQt9TFFs13o8lNChqxQKJUhq1R4klVLfbVmuU+mYFtaeFcVvwLrC8mVSmIxDVwiycR3DSQ22J7BiOshVXOnOReXznjTUNZ01atzZkhoycRUly0PFZdBVBb96z2HcfWQEJybzTePBssNQcRhOTObwpk8+ibim9PtNtxFtkTte8YpX4OMf/ziyWRnE/8Ef/AF+7dd+DQMDAwCApaUlfO/3fi9OnTrV9oVwzvHf/tt/w/d8z/fg5ptvBgDMzs5C1/Xw/AF2796N2dnZhud597vfjXe+853rvt+ejIrzyxtjt9WHCqEakABMzsJj4qpsupzJt7ZrD05TcRgWizY8DlAiMFswAUgFY12lNUkuAhIqGmsUcLlAvuIiriv4rz9wHUaSRtcDnbVs9xZKDgiAXRkjLAJGEWUMv+T60XU3Lv2G9asDrY7jjeDAUAK37c/i9v0DuPPgEPYOJhCPkDnqEX2+5wq2zwKWqT9uSivqZk0Ux8YzuGk8i6cmcihYa1suBknFhC7dQAghyMblJrziSCeCaqNTtZge0xTEdArBEY6lvpVqZ7DWZnEopa+rRE0JQgXLqxGdGsfLFcd3ypL28BpF3VgVYBzYNxjH//zRY/iZDz2GZn2eHDIQNTSK+46NYaUsk16W052mWkoIWJcKGQolcs03XcR1FQNJLUK0NEKix1hGYDpvYrm8/mdcKLtwGJCNq8gkdSi+ByKlsjhXcRiGhKi5/+sVBdbDpaVyy8dxLvDuL5zGfMFeVVTlHJgv2Hj3F07js/e/uKmqw/u/eg4rFZnYzJtuOK92Q+nzjv0DUH0XJpUAglYbVokQYEJuTu9ooALZK4or3VyP++jjasBAUgNItdmvdvWSzUgKJfK4LqLVsRzMPYaiYDRtYGrFhMcECJHFoGDuDaZILmTxR/iOHjLRVrf2dXgZzCQ03Lw3g/LXGcrOxhSqOwGVynn+urEMTs8UkTSUlskdT08X8J///js19yb4T7WB65/LOOaKdtO1KijIpgwVOe4CjID4+xVCZaJvowXZiVylxvo4+hATvyGCC3ncRnB6tgiFyGZOj8v7Skl1fKgKgULkcdvVjB3EHwtFG4T4Dff+HfY4x0Jx7fijU+ivye1jueJIZaeU3pJy6FzRXtdsaavAIYvJQcOR4jeOj6QNLJUcVPxm+G4KufTReeykcayqFP/jR27CL//tt2G73HfS8xuofTcn3Xez/uDD51B2WJh/ypkuTk7LIuCebAy70kbDnNTXzy4g5ztCtLueNzt+rTEcXLPLWnPPEEDoiMW4wGjKQN50UHbkek8BEErCY5u9KdnA51vjdPC4bBA5OJrEb73sBpydL+GdnzsJSji8FhtnWsFc0QEtOltWKxWoNhW1erzDZIO/qhC4HkfBcrvuiNepcRzkJ5bLDiyX+fkBSaLiPsF1O0xmt9uxqh5Fy8MTl1fw2ftfjPuOjeEPHzyDE5M5AHL86Qo25FzbCGHGnxDszcYwV3Tw/q+ew50Hh3B6trhujW0tAbR2c2fN9mUKDYRkBDwuYKh00+TOTl73TkGr47jV/H9wxzRa3R8Fr/caQWs4oWGl4jb9rm8cS+OBp2dQsj3sycZrjsnGxYaV3APSR9JQMZg0cMveAQDSccdlHJbLcH6hjDNzBZydl4SPi4tlTKxUmt5/y+V4ZraIZ2aLNa9n4xoODidwaCSJQyNJHB5N4ciuJIaTBjSFQlep7/bRbxbfCNrJ5XcT7azHG8nR/+cXH8Iff/HZTV2jJwR++OYx/N/vTuDR88tdYXpFCWSB0Gg9qYwJoOJ4yJliTVGAqQ3mdQCgaDG8+wun8XuvugWTLRI7GuHScgWXliv43PEZAMBQUscNu9O457oRvPDwMA6NJBDXVRhq43r+1YJOjMN+/am72Egtk3OBs/MlTOUq2DuQwNGxdEtrbbTB33MFFotO6PJGiBQtYFygYLnQFIq5ggPdb4wllIT1CdGCy6Tlcjw1kVuVY24koiHdmluf+ggaC+F1C7ORHKBC4V+ofEWlUkDgE49f9tX8W/sULlvdnNDu3Fq2Wdjgu17M1StxfDfHcb838MpFu/1sQZ5/NG1I51NfVC+IezwucygvOjyMfzkxjScu53B8Mo/TM4WmTsGbwUze8gWXVJRtBiYESDdzOF3eV27XetxsHovpFLrfm5TQZS+n5cr43WN8Mzr+PYugJupxAbtkY9knZ7qMh6sQ50BMk32k2Tga7s+j7mwpQ8Vs3q5xp1r0ON7zwDO4+9qRpvFgyfYwtWKCCXmj0zEVhqr0+023EW2ROx588MEatta73vUu/ORP/mRIvvA8D2fOnNnQhdx///14+umn8fWvf31Dfx/grW99K9785jeHPxcKBezfv3/VcaqiAtjYhiRQdFjvmJQhH/B2k+ICQN5vBOdCqklQv6JrOaxGQDC6dyWEQKWA7XHEdQUjSWPLGiCa2e4dHk1iYrnSdHNYzxheS0Gv37B+9aDVcdwO/sePHMV/ODiEvQNxxHUFMVVpOQF995ERfM+1w/iHb0/WqBaolOC+m3Y3fe4CYshv/uNT65I7ALl5Hk0bvh2mdAcZTRuYyplwPC6b5iADQ8ZlE91o2qix0x2Ia/jf/3bmirJS3Q5G/nqbxaGEDsVXrWoGhVxZqmbtolPjOGBs6yqFU7ThMA6FEr+xlMBjApQS/PSdB7Dik6HkbyQafUOmw/C+r5zFrkwsVLfuBkgXW04VSqD4CR7TYRhK6E2Jlq1ucIQAcqZ0DFIoQcpQkDakZe91u5IoWt6qc29WKbjc4s0vuxwnpvI4PVNselcFgNMzRZyYyuO2BoSJqjKsfEYCJRYh0BWlz1v2ZnHDWApPTxX82K32ygmAm8ZSuGXv6vfqFcWVbqzHffRxNWEooa8idNSD+Md1E62O5ejckzJU7B2MY6Foo+yrQQfXO5aNwVAVTK5U4DJJsgxUtTTfL1YqfEeEEDoACuD+e6/FcMqA6TJwvn4DWn1Ru1OwPY5jezLh2rtQWj/Wj6JZCFmvkKYS2WC71lq1XHFQthlM16tpDBSQRTfXsxHX1Y0VZKM3t36Zj97cDcZSyxUHlFCMZ+NYKtuwPR42WcQ1iuGkgYrLttU+/sRUHmdmS76gRjXBHPzsMo4zs6Wm8Uen0F+T20cwp8lGzdYc9bo1Z2wEwZDjAhBM7v8pCEZTBt5y340YSul9xbodhp02jl983Sje9APX4//8+7NwPQ7CBSgATVWgqwSDfvxSdliYfxKQJPqAEJg3XQwmdMS02pzUnQeH8ODJOQhI0kininLDKalyL9Po0o1XVQhUKpW7S5YHh7Uu7sD9Pg4hpODSweEkLi5V4DKOhK5iJKWDCyBnOqGggkrl33GBsLBblYlpf45R/aYcCAGPAwXLxZ9/+VkQIhXJNYWgaHmh8nQnJ7GdUCv1/PwopWRLHPE6NY5PThdwdq4Ij8mmUi3i1KyAQFc4XNbcAbBb6MRbUcgGqU7k2giAZ+dKeGoyh3RMw/ffuAvnFkoYiKvQVQVzBRuMe75yIwHjfMPmW9Gan+0JZOMqjk/k8LI/fRgFS6ok6ipt2kxEKcGv3nMYb/nUcUwsV5CJa0jHVDhMtJ07a7Yvsz0G4dsT6wrFG7//uk3XwxoJt+n+vFIwXSQNFb96T/sOgL2MdvbHdJ38PyUE1wzFcX6xUvPs1RDkewj33TSGB07NYrZgIa4pYYxsugwpQ8UP3TyG933lbLjGmw4L61QxjXZcyT3I+ScNFcMpAy84NAQhBBzf6aNsMzw3V8SZ2SLOLZZ80kcFs4Xmart508VTk3k8NZmveX00ZeDQSAIHR6Tbx6HRFK4dSSIT1yT5RJXrqq5c3U3j66GdXH430c56vJEc/XOLZRgKgb0JAqFKCM7Ml/CGe4/g5PR3uxKjDCd1eH4Db4BGV7xUdvCCg0N4w73Nm2EXiptTsT41XcC3Ly6h4rSWH6OQDjzH9mZxfqGMpfLqvM9y2cGj55fw6PklAFK44dZ9A7htXxYvODiEG8bSbdf7rwR0Yhz260/dRbu1zA8+fA7vfegciqYLDjk+3vn5k7j/3mvxK/dcu+Z7VRtjCzAd6VqrKn5u2HcFjGsKCAh2pQ1k4zrOzBZDoklco8jENMwV7ZYSZ0XL8+eLteOAo2PpMM/bChgP7szGsZGePMBvpoXc/AeiENm4ipm8CQLiX9v6WDFXN5C3O7cS0nq9ei0B5q10+e3WON6q3sC+M8jWI9p83mo/W7Bu6SrF3sE45gsmLNefY/yvixDgLx8+19K8M5LScWw8i8GEhq+cmUfe9KApgNsiD4RA7qHKjgfi17O6WVPQ1J05jtfDWvNY4CStUCkArKsyv2u68k6rBPB6pZDTYQgBuKIxudBlAiXbQzqmNdyfB2KAhqpgOmdVCZ8UEILAYxynZor42OOXcdu+gVXxoBACC0ULTAgohEAQQFeUVbn9ndRveiWgLXKHqJsF63/eKH7jN34Dn//85/Hwww9j37594etjY2NwHAe5XK7GvWNubg5jY2MNz2UYBgzDWPc927V7r0crnzxvekjoyqYncZfLBlGFNpL6qqeaCDAhMJaJbbmaTiPbvaNjafzC33xr04zhjSzwfexctDqO28HP330IygafjQ8+fC4kdtQzgP/h25M4OJJsurm++8gIfu5FB/HH/35GFnfrlHijMFSCxVJtU5OhUgwmdCyXpYKA68m5IKapGE0bSBlqzVgC0BNWqp3CdjDyW9ksJg0VhBBQCNSHVST4/1VeBOjUOD42nsFwSsczs0VwyEQUZyJUDqWU4MaxNF575wH8xZfPVr8LEvlX3fKZ8l0tJpYrMB1vw4Xf9dDNAnzUySf4dzOi5c17s5jKVXBpqeI3fEgVFgG/WazB+RkXyJse8qYHSoDrdqXx4utG8cTlFVxeqiDvNx9sVil4b0bHdG795NbejI7vXF6paXqNDrEgJPW4wHcurzRsrlwq2ShYLoSf3ARk8lKqIUuV0E4qfVJK8KO3jePUdKHhM0YJ8KO3jXdd8XEz6MZ63EcfVxsUv9EMftNEsHsLzOA2Gp+2g1bHcv3ckzJUUAJcWmIApPJNTFMwlNBRdpj8XHVw/TU6phKkYhrmC51TT/6R2/bgV+65Fk9N5KTCIpFzLV9jIe9Wbs90OU7OFMK19w8feAZP1jWPdAqMiTXXqoG4BtNlYFxAU+rJB1Ui6EC8fYeYm/dmwr1PNPYAqmsv8Y+LotWiSDQpf3A4WUNyj2m+axvj20qYfmIiJ69JIQ33V4oiE6JPTOS6Su7or8ntI5jTnppofWz2UiNzlFQnADAhUHEYbtmXxY/d3jiG7KO3sdPG8SNnF/H1s4tIGwryXMjmVkKQ0BXcNJ7BfceqDaDB/Gg5HLbHoPoiP7bHYblShCiak/rc8RnM5U3EVAWW176CXiMOAyXSkfHoHpkzAlDNJ3EBygX2DsZRmGnP+cvjspjlcYG5ooPRtIHXvfAAHn5uMTy3Qih0v4HGE1JdX7DV+b924xKFyqZiETXk8P9jV9rASsXGpWW5nxZibXX5KxlSBETGhd12xOvUOF6uODBdqViv0tUxBqUEhImeU/1vhOh4lLk6+dwqBJvOt1Eqlf1/85NPwXIZTJehYntgTGAggXC+oYSAQWz6/TSFgguBou0ib7qwXI6C5UGhBIZKYWi0aTPRI2cX8YGHz8PxGCouQ8n2QClBJqbhpvFMW7mzRvuypKHAcjhcxpC3PNw8nsVr7zywuQ/sI5pPPDVdQMFywbkU01EVgg88fB6UkCtGWK2dcRzdz0fnWBrZz99z/SguLl1q/Pz1EnMYwK6sXMPe+9A5TOfMMD+Rjmt43QsPYP9QAi4TcDyOmby5qk41nDTgdlnJnRACQ1VgqArSMQ1j2Ri+9/pRcO6TPhjHStmRpI+5Is4vSKePC4vlNZvnF0o2Fko2Hr+4Er5GCTA+EJcuH8NJHBxJ4tBIAtcMJxHXlarTh+88shX5m15HO7n8bqKdcbyRHP1i2d7Q3jA4fUyVhKHlioOXXD+KN//g9XjHZ09u4Ixr4479WewbTuLD37i46ne6QjCSMiAg80Jvue+GNXMW86XNkU+YAP7y4fNtObG5XOD+e4/ge44M49n5Ej7x2GV86ruTsD3eMK5dLDn48jPz+PIz8wBkPuyWfVncti+L518ziJvGs0gaKmIqDfcjVyI6MQ779afuop1a5uMXlvCeB86ErpGqH3PkKy7e84AUVV6L4BEVGl12HblWCYBDhHvZXZkYFEqwVHLwe6+6BQDw3z9zApMrJvYOxFB2mIzjsX5eTAB4eqqAlx7dveZxJ2cKMNsg+sm8d8uHr8I1wwm89IZRfPiRSxs/CeTnsz2ByRULqkIwlNRa3mcPxlYTpdqdWwPhZy5aq1c36wvYSpffbozjreoN7DuDbA+C5vNW+9k4Fzg8ksS+oThOTxcR06SlpYA/Z60zRikBjuxK4di4HzMcHMSBoSR0heA//913QEB8R3kKlYpVAmzNzkkIge3J3J+C7tYVGtViO4ntXI+bzWO37hvAPdeNhPnXPBfQVQWm7zDdrf6uXgfjAhPLFewfSiCpq6uc1gI3jorj1hA+ARmPqAqBwwQ+8fhl/NR/2L8qHrRcDtuTeVbpFKIgplP/73dev+mVgrbIHZ2GEAJvfOMb8ZnPfAYPPfQQDh06VPP75z//+dA0DV/60pfwmte8BgBw5swZXL58GS960Ys29d4bbbptlovTwgEh4PrFG9tjcDqUgZeN5LWvUfgsZipChRXPV/P7qTsPbEuht5HzRicYw+0u8H30UY+NJl49j+O9D50D4wKGSkBJNSGkUA7HE3jvQ+fwC3cfgqo2Tha9+MgI/vaRC1AVCsfjWCrZ8ARCoggXMvgoOxyUwG/KkuPedDlsz0EmpuKnXnAN/u8TMrE1ktJhKApMl9WMpZzp9oSVaiewHW49rW4Wf+l7DiGuKTD9OV+l1YY3xmVQFNcV5MzuKgZeDfjm+SXMF20wX31TlWJ84L6V7FBKw1tfflQWviN+h9EGxPp1m0AGonuyMVgug9PAKrXXESihU0IQ06rPWiOi5bHxDL55fglv+uSTWC47voqBjFpUKu2t//vLj+Kxi8v43FPTq5yGuAAev7iMxy8uI2kouHlPFsf2ZnDPdaO46/AwYvrqhFWrUGhrSXaFUsxFEtb1YVxUjXmuSWJ7peKGBVnHE4i22gRkoU4qfXIu8PBzi0jFVHiMyw1+WBCVyrEPP7eIX3rxagXEXlFc6aOPPjaHnOkipingfkOkolTHbNAsE53DtxuN5h6XcSnqQORcvCsTAwgwkzdrkmcKEDa8CACWx2EWN0/sCKY5hRK8+nlSCCK4rxUuOuoM0g4YF/iuTya8+8gI3vHKm/Ca9z/a8t9nYyqYkPa2a8ELWBU+KJFkj/q1KlwX65eFOiJouxhNx5COqWFs0EjfIx1TMZqOhT+3UxSpLSYbiOsK/KdpS4vJayEML+t1NRB5PXpcHz2DaIG7aPXGPLsRRB+tuN6PAfvYGkTzMUNJA7vTMRRtSf7XVYpfvecwXC4JR5pCfQIohcf9JlCK0PHC4xzB3B7kpKZyFXgcGE0bmM6Za6qiB1D8PR+P/gw5NWfiKv7zPdfixUdGawiF9XvjZ+eKeMunngLQer8tF7LRhDFe0yDxSy8+HJ57IK7hjx58BscnCxC+kxZt0tjeTp9vszgnZ3pQFQeLmxQmSOpKSDg2XYZdaQO6qqDieHA8AdP14OyQaqkQ8jkYSe6M5jDpCCCf5yCjEgipCcgYd7t5Ha08q/XH7M4YWCjZHSuyB3uLuaKF8WwcAwkNlsthugxOkcs9lipdg1y/eLbRXnqFBrklgaWSEzZ0aD7B1/I4FooOxgdiKNmsppmolTmznRx2s5wQiKwfDMQ1vOHezsYDdx8Zgcc53vzJp6BRgnRCx1BCgyfQ1Vx8L6N+P6812c9/6+JK09yfxzcvNNhJfOfSCk5OF6D4pIbAmaTiMHz0scv45e89DC44pvMOuKi6RwV1qum8iUxM3RbyPaUEMSoVQzMxDdcMJ/EDN43B8wkfjscxl7dwZr6I5+aky8f5hTIuLZVRdhrn/rkAJldMTK6Y+Npzi+HrmkJwYCiBg8NJSfzw/79nQLqYaookfOgqhaoQaFdwE3k92snl9wo2kqPPld22SAoBCOTzM5TSAVF16j0wnEQ2riHf4fzf184tITNVAOA3GQq/dibk5zY0BQlNwXzJXjf3mI1vvmUoX2mPRB3UYjRVwdGxDC4uV5A0VBweMeD4qsQVh6HieHAbfB8508XXnlsMx2/SUHDL3ixu3ZvFHQcGcdu+LJKGBkOjiGkbr2H1GjoxDvv1p+6i1VrmzEoF//jdKTAuoEd6UCgBKGmtBwWoFRqFADwhQlfk0XQMKUMF5wJ5LpAzXbzk+lG87RVH8bbPnMBc0UFckxv4lsdvC4/gdy+vtJW3JySYxFr+kxADCQ2//6qb8Z4HnmlpL0D9e1N2msdoTAgwT2Am3/qeu9HZNjK3ulz4TbWt1avvPjKCOw8O4XPHZzCVq2DvQAKvvHXPms9Mr2MregO3ow+pD4mg+bxRP5sQsrHfZhxPTa7gxFQOxyfzODktHZE8LrBe2TFlqLhpPINj4xncvm8Atx8YwFBSR1xTasifJybzODdfwkhaugBbLoNCSUvkDhH+Q0JVCLhP9OgGv1/bweO5FTTrb6KU1ORfLyyW8YdfOA3T5b2kobBlqIqAyX3sroyxymkt+G/bY1AoDYkd0bMoBJgtWDg9W2zYE8AFAC6gUIrRtFFzjp3Ub3oloa1ogpDGSoUbxf3334+Pfexj+OxnP4t0Oo3Z2VkAQDabRTweRzabxS/90i/hzW9+M4aGhpDJZPDGN74RL3rRi3DXXXdt+H0BQKUbG+qNJmMZYBM/GSBZfS6XyYGKw2qC9PWgUUBVKIQQIcuv6bEqASBVIoGqcmmgYB5FI+VMAFtiMdYJxvBaCzzQmQmkb7nWRyN87vgMiqYLVakldgA+e1fhKJouPnd8Bq9+3t6G5zg2nsGR3Wmcmi7AchkEIYj5hSEBIdWGgwKeqLo+yMZrAccT4CB48w9ejxdfNxIZS96qsXRiMt8TVqqbxXa59bS6WVypuEgaCjRFNoyHNRpfUSEdU2GoSs/f515H8BwwLnBgKIHZggXbrbVxNV0eNoLsG0zWKIs0W3t1VY4NQgh2ZwycX6x070N0CVwIJHQV2bgGIUTNs9aIaHn3kRH8yU/ejvc9dBbPzBbDZssbx9J4w71HcPeREfzoHXvxuz92M7767Dy+eHoeJybzOD1bqEmcl22Gxy4u47GLy/jY45dx16FhfN+Nu/D9R3dhJGW0nSRvJxE9NhCrfTH6/UaG66rjfAwkNYCgoYWugHxdoeiY0mcwn+xKx2BoFJYTUSPXKSyXr5l86gXFlT766GNzGEroSOoKUoaKXMWB7TFwv9kvpikYSOir5vDtRv3cY/o+xIaiYHdWFoMqtge7TnmLATXzMuMyJiKbVOwNHKcysWqzXnBfk7qCuUJ71uadxNRKNX54erLQ1t/qCkUmrmEgrmKxZMNhoiUVMJ+jiJwpGzoTuoLlihOSXTwmIo1psim0ngjaDo6NZ3Db/gF899KyH3NVfxcUwm7bPxDmF9otitQWky3ENSW8dtNlSBnqtheTbz8wAE2h8BgHpTLPEzSCSGK3dEy5/cDAtl3jTkU3cjD157zr8DD+6Cduxa///XdRXIdMBaAjKuPdxEuuH9nSGLCfJ7s60SwfM5DQkY1rmC3YeM8Dz0ChBEXLRdGWbo+GSpGJaTV5cEIANbLnC3JSewcSoXPTvqEEZvImrDVUPWMaxZ5sHI7HsVC0fKdssmpPW4/6vfFSyZYq7A1cNZqBEuC1dx7ATzx/f80YqD/3G+49grd95gSWy4DLGCy39h0IAEOjgJBOE/XrfkylGErpWCja4EI03LcGYFxgvmhtyKmDQK7fmkpxaCSF//oD1+HyYgV/8ZWzGEoa0mnAzzMslGzMddCFrZsQQjYqbychtB0cG89g/1ACK5N5uP6AiRZpewUKJWs2ZUV/M5YxYHt8zWcXqKZv1vqYwb0I3np3uppvGsvGMLlcgec70jLGQ9VOAGGTtdPgQtZqsiAEYZ0tmCGoX28jINIRjwsslmzszsTCfM6x8cy6c+YHHj6Pu68daWsN3eqc0NefW8B/+cSTKJguKCFYrjgoOx5G0zGMZYyu5eJ7GdH9fN50YXss3AfE/Zys43HMFqymub/pFbOnyB1PXF6BwwT2ZOM1tY9sXGC2YOMLJ2bg+ftK2WQqj4nWqZgAjo6lt+kTrIaqSIX+hC7H3Q17MhBCCnI5jMNyPEysmHh2rohz8yVcWKrgwqIkfTRqFgekI+m5hTLOLZRrXo9rCg6OJEKXj8Mj8t9DST10+NBV6fIh/7+6p2KnYyeSOzaSo0/HlA01jOkqxZ6sJAFGhSqG/PwIJVVX307Acjko8WrOG7oCCCnMsncg3lJN+KU37sIDT89t6nrSMRVL5dZyTxxy/x3UYqK1WUopYhTh2i+EQMHykK84eN41g7i8XMH5hfKq+1i2Gb55fhnfPL8M4AJiKsWx8Qxu3TeA2/fL5s5sXENMU2CodMeOz06Nw379qXtotZa5UHY23YMSICo0qlASuiIHz3l9f0j99x/0rKwHSoD9A4l1j4vmzdcDgdx3OOun7Vb9na5S/NbLbkA6puHMbKmlv5MCDhQUvKOEerXBmNvo3NpOvbqR0NKnn5jc0eO4WW+ggIDlcDhMujqu52zSDNvVh9SHRODm7jAOg1CflMpheRwVh6Fse7Bchrf/U2uuZ4ZKceu+LF5y/Sied80gbtidRlxXEFOVNb+/4DkzFAWjaQNTK6YUm22hv1eK/1YP0hUKjzEw0bineLPIGr0TW3cLjfqb6l8fiGvweOCcIufKHkqhdR3Rz8qFwELRxgsPDdXkQo+NZ7ArY2CxZEOhwdPo/72QzjQxnywUuAxG44GKywAhXVLG/J6AKDrZb9oLtadeuIZW0Ba5QwiB17/+9aEdj2VZ+LVf+zUkk0kAgG23t3i+//3vBwDce++9Na9/+MMfxutf/3oAwJ/8yZ+AUorXvOY1sG0b9913H973vve19T6N4PHOfRkC0h46GEgEkqRB/IUIQEsb9phGMZTU4XgcedP1FR58Vf8Gk5ItZTyhKdIWmgAYTOr47R+6seZhaxTQDafkQFsqOVtiMbYW064VRBf4bjSs9y3X+miGqVwFHNItoBEokc1sU7nmm9SoYuhKxQldRDwuC16SIFZVI2O+db0QweaNQCEEp2eL646lXrFS3Sy2y62nVSLZQFLDcErHyekCCAE0WpWM5EIqKRwb3zlF5V5F9DnwuJCqaEQ+A8EtrzgMb/nUcfyf/3gbXnnrHrzz8yeRq7hNi8UKkUoZQghYrtwktov6DZlKq+5a3WDjN0JCV3HNcBxzBaflMd3KWkwpwffduBvfd6O01F0oWnjw5Bz+7eQcHruwVFOQtFyOh55dwEPPLuD3//U0XnBwEN93/S58/9Hd2J01/AbNtdf5AyNJfOP88rrXfmAkiecdGISmEOmQVn+T/Z81heB5BwYbnmMooa+7GRcRRa3NIjqfEJAaNXKgNWLqZuOnPvroY3sRxGVPTeT9VwhIRP6/aHm4bX+25+KF6NyzVLLxp196DhPLFSR9p6aKy1pa65KGipGUjskVSzqAbOBadIVCUxTcNF69T8F9PTGZl81WQsDjtY1ispmUgLHuqR4vRCSCBGkvBlgsO8hZLgyVYiwrFVNdxlG0vHVJGALA73z+FD7zxBS+59ph7B1IIKZSJDQDBcuFwziE7HFDTFMaEkFbRbCPedMni7A9O8xpEMj9SzqmheSLjRZF7j4ygte98ADe+9A5TOfM8PzpuIbXvfDAtu/Hb9mbxfW7U3h6qtCw8ZgCuH53Crfs7dye5GpAN3Iwa53zfa99Hn7hb761rvKW4s8bvYZgxDwzWwqVoruNfp7s6sV6+RhdJTg1U0TaUKGrChyPgVAC0+VwPEmecP2FOa4piGkyvxLNSb3y1j349BOToXPTkdEUlssOlitOONcSyALtUFLHUFKuYUXLw52HhvGW+25AznTb3h8NpwxkYhpyFaelBnpdkQ0xLz26e93cU7Q55tR03m+2k43mLuPSpRcyYFAVAtcTPolVftq9g3HENQX5ihvmKQKncO4X34Cqq1k7xA6FIowNdmdiGErqyJsuJpYryJVdn8hYm3eXeZf+vrOT4FzA8hhMh8HyOO46PISnJvPbTuZQqcyzrlRcmefmopqjbvFBS+lynAc1LUIaP6M1j9Qap47+ylAJEpFicspQsW8ogdm8BdNlcP3GJ0COGYUSCCHzh6pC4Hii5rzNYnbGAkcgErqoqBGVw0DUzfak6Izr53OicyYIULE9WQwHkNAVZOPqhnPYW5UTeuTsIt7yqeMomG5YhxCQgjpTKyb2Dsa7lovvZUTrLNcMx2G7ImwINzSCuYKD3RkDc3mrae4vGVOx3CGH4E5guexg/1ASIIDpsJoG94GEhmfnSlK8ihJJ1KLV5qb6OlUvPwdBvKKrFClDxUg6hjsODIJxAceTLh9l18XFhQqenS/iwkIZF5bKuLhYweRKpekaa7oMp2eKOD1TrHk9G9dwaCSBQyMpHBqRjh8HR5JIxzSolITuHhql0FQSkj92ItrJ5fcKNpKjPzVdbHCmtUEA7BuIIW+xVa4H0SYrlRK4HXShDWrJjc5ouRyzeQt3HBgAFwJffXah6Vpy3a7NkbYIJMEqV3Fbjm2itZi1arOEEKQNFabL8PrvOYS7Dg9hvmDjsQtL+NaFZTw1mcezc8VVY9fyOL5zOYfvXM4BgE8Mz+DWfVnctn8AdxwYwGBCh6FSGKoCfYeoYndyHPbrT91Bq7XMlK5sugclQCA0GuyxW+kPiX7/Xz+7gI984wLmi86a+e1MXMMrb92z7vUstOHonYypSBsqhpI6Fkt2GG83QkyjGE7qUAiwbHq4ZjCBW/Zm8Z3LK/A4b+qgWQ9NISC0ummp3yNIYqvcW1gtEnVdxjGTN/08iPz/ZubWVurVV6r7RKPewJLtYaFoSwE1/3v70y89B12lbX/G7epDutrBuYDtcewdiGHPQAxnZkswVJlPtFzWUp5LVylu2J3CeDaO8cE4nn9gEC++bgRxXYGhtidAGn3OUoaKvYNxLBRtmC4Da0G9PTo1lG1ZMw1chjsY6gEAUvHeEQjcbgT7U94k/r1awIU0vPqhm/es6vX66TsP4H/+80l4TEBVRLjGeVxAIQTZhFbjMhiNBxbLNv7si89hcqXaExCgk/2mvVB76oVraBVtkTt+7ud+rmZx+5mf+ZmGx7QK0cKEGIvF8N73vhfvfe97Wz5vKwgKSxsBJTIxW/In6FBpH74qAuREfmw8id991S140z88ickVE4AImXr159udiWEooWGu6GDvQBwF0/ULPFLJQaUICzn1C4HLBFzGQAlw03gcRT+wScdUfPfSyqqALme6ODktlUX3ZGPYlTa2JMhrxrRrBd1sWL9Sg94+OoO9AwnpBCCAQLMgVGpFVT137zoqBdIS8xr88b8/C8YFPFG1ZeP+ORV/D8mFgGDVhqzhlI6Kw8Lk4lpj6UqxUt0Kt55G2BCRTPgqbtQPJDsdrV/FCJ4DjRLM5E0wIRsjgjVAQBI+KrYXNgvef++1eM8DZ+BxAZVWHRkCjKZjqLhS7VMqCrb/fTXJxTX8XTsIYolW0kRCCMwVnLbHdLtr8Wg6hp+56xr8zF3XYKlk499PzeHfTs3h0XNLoZo7IEmu3zi7hG+cXcIfPPAMnn/NIF5y/Sh+4OhujA3EkNTVkNgWxY8/by8+/vjEutfx48/bi1v2ZrFvMI4Lazit7BuMN22uZA3UUevBha/62AHUKE/Q1apgrRJTNxM/9dFHH9sLSgnuuW4Ej55bAuMCqt9sxIUkJyqU457r2lNw3SpE5x5dpTWxZaN9PAn/UVW1qTgMcwUbTLRP7AiK0xolGEpqNWtdSDj4hyeRtyQpOkAQX2sKBQXgdjG950Syt3fsH2i7QE/8hprJFRO70gYGkxqKVmsSZUIAxyfzOD4piUNxTYEQAiNpHRQEHIBOKWI6wXxx88k2x1dhDj5dEF9FSZ8bLYo8cnYRH33sMhQiFbcpkU2sFYfho49dxrHx7Lbuxykl+NHbxnFqutCwOEgI8KO3jffkOO5VdCMH08o5/98fugF/9OAZOHVfZKSWu+p3vQIBqbQ/lze3pLDYz5Nd3VgrHyOEQL7ignOBbFwFpRRTKyY4F77zjQARJMyVSQcIwPJYTU5KVemq3NVgQkfcd+WqOAxxTcVY1kBMVWB51ZzWr73k8IZJB8fGM7hpPIPjkznYLlvTNZsCYJzDcoGJ5dZUR4Ni2N89egl//MVnMZzUwYXA5IpZ09ROQUKVOeEXFjwmYBEuFUSJLGoTIq8jeo0KJU2VxpuB8aqqKRfAxaUKLNcDF8Dvfv4kbhrPYDilYyZv1+Tdd9LKRgBM56yebr64vFwJnWefuLyCT357/VxIt2GoBOMDCSR1BZbLYLpckiNadLcJ6mEVl6PkyAYuWcuSzvSrYuMqz72mgaoR4SIIrcay8ZBgESBlqDg0ksDlZRMqJfCEgGl7fq5e+DlBipGkjoWSDS/yeRp9roCIpVACy2XQVQrXV+xc9XmFbPAO8jnBnOkwjomVinQcjhxvaLJhdKM57G7nhAKCdtm/f4rvVEKA0K1koWjhwGCiK7n4Xka0zjJXcDCQ0JDUVdiMhznZn77zAN73lbNNc38Q64sAKPCdMLcAnpCuNjN5q8aJxFBlDUo63kjC4VJJuo9GhQPq61Q7DQqVjf1xXUEWGsazCbzo2mHYHofDJOmjaLk4O1/ChcUyzi+UcXFJ/nt+jSbVvOniyYk8ngyFPSR2pQ0cHEni0HACh0aSODSSxIGhBAxfEElTiC9oQasEEH8u6lW0k8vvFWwoR7+BFpKEocB0eUPXg2iT1XrNgqpC4LUR61nu2jk302WYzpn49b//zppNSwXLQ1JTUHY3NiNpKkVclwQJcw1HviiitZh2arOGqmD/UAL7hxL4iefvh+0xLJccPH5xGd++uIwnJ/J4ps6RHpD9NCem8jgxlcdHH7sMSoDrd6cl2WPfAG7dl8VI2oChSoK6oSo9OR47PQ779afOo9Va5h0HhvDxxydCp+96tNqDAmy8PyT4/m/xx8F//6cTq6476F9RKMH998r9/Hpw2nAu+5FbxvCl05J85jGB2TWcujMxDQLATMEGF8BkzsQv/823oSgyD9HO9E3gq+3zWjIJATCS0iEA/PdXHMUnvnUJj13IrXu+80smfvZDj+OFh4d8JfUsLi1XkFApKhtwcluvXn0lu0/U9waWHSZzTz6JH0Q6zU8sVzaUp9yuPqSrCUJIIoftcViuh4llE09czuHkdB5P+3UkAWA985XhpCQd3rw3izsODODWfQNIx1QYqgICWZN6ajK/IXJi/XOWMlQkDQWWw7FctleR9AlknCaAmlgteEcBSS4LpshOtozpalut3VcscqYUzKsR3/c33DW50xaJfjsdMY1i/9DqGOG1dx7AP3zrMp6ZLcreU3/fH9coRlLGKpdBoDYeNBTaUjyxUdeLXqg99cI1tIO2ZoCPfOQjXbqMrcft+wfxzFx5/QMbIBPX8J9esB9fODGDS8smgNpEdJD0JYSgbHswXQYuJBtKU6UKYDCRUMjgXVcVzBVlQvCHbt6Di4sVmK4HlwuolPjnlbOSSgBFodg7EMf5/z97fx4myXWWieLvObHlnllr19J7V0vqVYuxJLfkls3iBWNsgz1gY2CM4efBhjvXMzD3msudOzOAPRgwA1gGD2Awg22wPTLGGEvetEuWZFnqXb0vte+5Z6zn/P44EVGZVblEZmVVV3Xn+zxSd1dFRkZGxjnnO9/3ve87t/QZGAeeu7yA5y4vYDgVwgO39OPMZBZZ3cJgIgRKKTi4r6AEIpJOXREVIUUEeZMZHR9/5Cz+T5uhJ6atGzu/0aBfq4b1Gzno7aA9KHcCsN2RXlaHAgeQigRTKbh/pA9/9eRlZEqmUGJwGf8OEzbRjrsxHkyEIEukwhZYoSyw2q6nFvipxy7i7FQOpsOgShS3DcTxwddtPJZhNay1W08tBCWSAcL5aDAZXrJlZ5W27PN5c0MXlTcDvOcgZ9gwbCbWw4rvRKwPifBSs+CvHN0DAHjwsYvIlayK9bk/oSGsShhfLAkFD9rYVrERYpqMorHUiClRgr6Yhtm80RRxJK4J0mjg9A7Butsj98Q0/Ozd2/Ezr96GxYKJR8/O4JFT03jm4jzyZffAZtyPR/7wm2dx57YUXntLH35k3xZs6wojosq+ClImoHqdd1xMk2sWRYn7+1p4+ORUoPd6+OQU7trZHejYevDmk+Njadhuod/bPKlu0e7w1tSGU+zvoIMO2gfGOJ44P4eoJsF2RDLTduOFiCpBlgieOD+H99+/e0PvNZbbtFdzL+Du/8o/hc04uOVAligo4YEaERVKXCc0Qaze3R/Hf3rjrSvWuiMjvfj1H9mL3/mX03AY89cGodBLIbkEgbVM4m0tS5odGk7itsE4ToxnA7/eVxHmHBMZHTM5HWqT6kIePMLl2KIOiQrlN02RsFB0ENNk/Pw9O2rG1fXAGMfHvnFG5BCI2L94myGHidzCx75xBl/90P0VRRHPIr28YaJaUaR8Pz6YDFfEeckw3xD78aVxLMN2mK8eJ5qgKGRJ2hTjeKNgLXIwQc/52ffdjX2DCfzhI+dwea4AhwsL6mRYwaX5wqr3BWuNVFiFxRhevLa4poqanTxZB/XyMbol5kGJAookmiM9ZTvDdgAO2A7Dnr4owqqE+byJmbwBhRLsG4zjA0d3Ix5SfNXg3337QXz6iUu4OJNHhglhhzu3d+Ho3l48cX4OF2fyyOq2//qje3v941tR1CrPL+cNG72KIJN4RdmlWELk4kVDOMdfPnkJu3ujODLSGyiHfdeOLkRV0QxGOfFV5bxljgOgFOiLa1gsWDBtBxndQkSRcPu2JEb6Y/jc966BMQ4G0WDuqf9RUmtHXB2C9CriK8tmmMrqoMRVX4dwWntlKg+JCvcDL+9uOaxuI+tGAwOQKZmYb9QZcB3BOMcrU1k8d2kBX35xDIUWnGTbBUEuIlAlCVFVNBn3xUMYXyzB4dUzUxQAiHiehNM890m55Wso40K1VpJWrhEcQlnQ+zsg4suoKqFg2mAMIBSIqTJ29UYxuliq2XBjOhyJkIwPvn4ED5+cwgtXFmA6DDIVBeu+eAhRTcJCwaxL+N6S0PAX730VKCF4aTSNB797AcmwjKmsAd1yQCT45BKOJYL2oa3CWfDURBaMc4wurFT75xDzpmmzwCSx9YZH0E6GFZQsp2KuKncryRr2muTiNzoa1Vnu3d2DR05N1c39RTQJRaO6+6Xbm7ZuIADGF0VNWyoTq9It0TAX1WQosiAc7OyNrNjTNVun2gwghPjq2gDQG9OwqzcG21kifJg2w2LRxPmZvHD5mPOcPgpYrJPbnskZmMkZeP7yksI+dUUNdvVGsct1+NjVG8HWrojfRC5R4hM+PPKH+K+yNnI90GwufyOglRz9tlSkQgSgvB5cDkFmlvHTd23FA7f2192jeU1WZyZzPtlzOSgBJEL8WnQQeETKWhEi48CluSI0d+8rUYLTE9kVTUvdERXJiIKwTTFfaF4qxVv2tSbIHcBSLWY1Ip+aLGEwFcbb7hjG2+4Yhm45yJRM/OBqGi9cWcCx0TROTWRXqO8zDrwylcMrUzl88ftjIAB290Vx2CV6HBpOYksi5Dt7aAqFKtHrvg/ejOPwZkSQWuZbDw3id75+GpmiBUoYKFmKexlnsB2OZMAeFGBlDt/bYwetJR8Z6cV3/sPr8DtfP40vvjAq3OjchtlkWMGHXr/Hr8E3wtYqzabVcM+uLvzcPTvxxLk5GI6DvGH7e99qrkTzBUPUIIjYM+iWg7y+5BgUdPYhROzTCREiCJx77ydOpCkSGOMY6Y/j4FAqELkDAC7M5nFhNo/PPXcNiZCMW7fEBYGTAU1MjT7q1atvZPeJyt5AHQXDgcPEHtNhgEQo+pMaJBDM5Q18/JGz+PLO7kDEI+D69SHdyDBdsrRuOSgYNk5PZHFyIoNT4+LPuXwwokxYkXDn9hRef2s/XrWjC7t6IwirMjS5MjZoVnG/PJeXCisABFHgjQcGcG2hWNGD6tW+KAG2xENg4FjIm35+zi5zB5UpcZvlhQsv45WOHu3Crg3kinc90R1RocrUz0UBS2L8lIicqed8HFYoDLux8Otmhmk5mMsZK9z5KCX4yJv34SMPHUemZEGWqHCUpAR5w0ZMk+v2VweJJ2qNwQ8c3Y1kWK2ZN98ItaeNcA3Noilyxy/90i81PIYQgr/+679u+YLWCz9y2xb8w/fHWnrtYtHC/3ziEigh6IkqKJrMVTQRE4YmUyTCCuZyBv7wm0Klf3t3xLeRI4RAIlyoZlGCnG4jrHB/IMRDCv7Xs1cQC0lu0/BSoiGiSkL1jHH81588AMY5/vnYBL55erpC5XM8rePzz18DICatnG4joskwLcGMlN2kuGBKMoRVCQXTQdF0cGIsjQ9/8WWEFWldLGeCLryr3ZBUw40c9HbQHsgyxVsODuBzz4/WVOd/y8GBQJuFfQNxOFyMfVUm/iZdovAVxDhfUj8EVutMI4rQ8DejmydyWUu3nnoISiRLlyxYDkd/XEVXVFlR6OAMmMkbHUb/KuE9B8dGMxXFRUA80w7jCCkS4iEZs3nTv9+/cnQP3ndkF752fBLj6SIKpoOvHxtHwRQ21J4Kh92GOvqf/ewdWCha+NPvnMdCwcT27rA/fuupi5QjGZLQbCnvv771AH7qrq3XJaAkhKA7puGnX7UN77hzK9JFE4+fm8O3zkzh6QvzyJSWksaMw7e+/pNvn8ehrUkc3duL19/Wj509UXzzVDDCxXfPzGBLIoyJdMlvjFl5XcBEuraa8Wi6FOi9gh7XCNUV+1Gm2E82rGJ/Bx100B54e42YpiBdNMArGp84Ypq6afYa5basT56fxR9+82zVpFg1xV1KSENnMwKx193ZE4FuMczmTezui+LLH3hNzTj7PXdvx8Mnp3ByIgNNIlgoWr6imVcAA0QjWrvzqQTAWw4uFdY8d4eg5I5adyMoMfSObUnsH0zg6QvzuLqsUcxhHBndBnQbEiHY1RfDWLqECzM5xEOKUCpVJERquGqV48R4Bmen8r4bSnlMLhEOy2E4O5XHifGMXxRJl6wl8nOZEmwyrKwoipTvxwHRKOfH1ArdEPtx7xrjIRnpouWLbogiIEEsJF33a9xMWIscTDPnvH9vH47s6a0o5nz84Vdwca418Zf1AgFEkVt38OB3LwDAmtlEd/JkHdTLx1gOg8M5IookFMmBCmU703GQ02389lv24+gtfRUkiEzJrErMqFX0ef/9u1e8/rf/6eSqFbWW55cjqoycbgnxCLdxHRBCLBKlGEqFkDcc/PnjF8E4D0QuKb+HWxJCXdhrEgcXBNiwQtEVVmDaDLcOxPHvf3QveqOan+c6P53DyfGsWD9dNdGJtA6rySoxJcTfc1plhU+HcURUsT4nw4LUMZjUkAgpODudw2LRBGOCeNuMM5kHmYi6B+ccjIlYzGtiyhnBnMqaBWOo22R7vfCNE5N4+NQUHjs7W5EvuV4IKRQ9Uc13qPDynxFFQm9cxWzOACfcdypWJIKYKiOjW+J5IgSEcNiOIB+pMoXtML+BwWticxxetaGtXDRJosA9u7rxn950GxaKJtIFC10RBT0xDfsG4njfZ19omBt+z93b8Z67t+Pzz1/Dn33nPAyboSemIiQLYodXt6sGmQK/fP8u3Lm9C4AgbH/nzDTOTObQG1MxkRbkL1FKE5+ZUoJkeKkQvm8gDpvVbxhgXDwH77l7+4bLAXkE7VREwWJRWkFoIRBzRrZk4/ZtyZtYoKR6naVR7o8SQeQrGrWT0GvQ+1MTMgF0h7u1Kfc7JgAoh2mLMXvrlhjOThcw4IokCW+Rta3JbETIEoUsUXjb1/5ECLdsiftOPZbbwDad0XFuJocrc0VcnhNOH5fnCijWIPAxDowtljC2WMKT5+f8nysSwfZu4fCxsyeK3X2C+LElri05ahHRkKNIFKos/vPcP9aL9PHdMzOBj/vR/QNrfDXB0EqOvlz0r5zYUf73kEzxO28/iHfcMRyoPuw1Wf3f//sYxtP6CiESAuF81UysV5HvIu71VWmGBkQ/yEzOgEQJQjKF5fCKpqXy+HVHdwhXFoLVtjyYDkfJcNBsncurxbRT5FMQtsJ486Ew3nRwAIbryvPyaBovXF7E8fE0ToxnUFg2N3MAF2cLuDhbwFdeGgcAbOsK4/ZtguxxeDiJ/kQIahnZIyRLvpDZemEzjsObDacmsoFqmWdn8vjQ6/bg9x8+C9PmkCXmE8tEDBrcKcNDeQ6/FWEQSgl+bN8WnJvK4dRE1ie2HhhK4MBQ8BzQWw4O4q+evFy3M4UA+Mib9vnzz4nxjN/H5t07xjgYF70AqbCMmZwJWRbK+Zmy3rhmJBAIAMt2oEqi8VeWRFzEwWExMb+XTAf7hxI4MJTA577X2h4uq9t44epiS6/1UK9efaO7T3i5m48/chYnxtIAIeBc9D7GNAVzOeEyxxjHibE03vnpZ6uKhFXD9epDulHgETmEIy7DbM7AyfGM78pxdipX4fheC1FVwm2DCQwmQhhMhXDXti68Zk8PwpoErY4IWrOK++X9qAXTge6KGoQVCVFNuBMmQjLm86bfg7qrL4bRhQJiIRkht5Y2m9NXOKalIgp6YypCCsW0KxCxFh15/QltDc66+XBgKIGBRAjzBROqLPK3Xg2SELF2UiLi40RIwWLREiKEPJg7rYfmZHWuHywGfOwbZ0BAVuSoj4z04r337sCDj13EYsH09zLxsIKfu2d7IMJnrXii1hg8NprBL//d9xFVZVCy8pqAjVF72gjX0Cyadu7YsWMH7rzzTmEZvomRX4Uykuxu+m3GkS1ZGO6KQKa0ogmBc2A8U8LoQhHdUeGMEdNk6BZDzrCQLdkwmAg2HMbRF9fwhv1bYLkB6u6+KF6ZymNHdwSGzf1zazLBdM7EvsE47nMf/ju2deGeXd342DdeQU63fSaah5LFcG2xhJAsro0xQKJuIpIDNmPIG7xCmcmzsqq1ALZqr7MczS68q92QLMeNHvTe6AjaqLWatApjHNcWS4iqEkpmpao+BRBWJVxbLIEx3vA5PDOVg+QqANoMkOlSkYsSoSrMudiMJkNKy840tcbVK1P5DWkhVQ1r5dYTBEGIZCfGMhWM/vJCBwDojtNh9LcB3nPwH790DHlDKG9ILlveG3N9cRWmw1fcb1mmeEeZ7fBrR3rx8UfO4vho2h/HBNXtE5vB7379DD7xM3fgZ+/ejr979gqmcwZSERU9UdF8EkSlyGJAIiQhbzqBmjq3JLTrRuxYDkoF0eMddw3jJ+8YQrpk4unzc/jOmRk8eWEOC4Wl9ZMDOD6WwfGxDD756EXsH4xXOH7Uw1S2hLmCgXTJrlkwZxxIl2zMFaordQ6nwoHeK+hxjeApfUdUCQ4ThT+nTLFfovSGUfpej/W4gw42IxaKJgpmpXoUAIADBZNBt0qIhZRNs9fwir0ff+QVhBWpptpwebHb4QAc1rApkEO4T2R0G7rF0B1V8J/eeGvd4hWlBB98nae+7aA/HkK6KJL6DhO/3z8Yx+HhJP7x+2Mr9smrwc7eCA5vS/n/9ud8haLYggwXIcBQMoTpXLBn4fBwCh98/Qh+5aiD0YUCnr4wj6cvzOHURLYi6ehwjpeupfHStTQUSaih3z/Sg9fs7kFPTIOmeEQPaYX6EQC8NJoWuYgyhVAO7quZEUpgM4aXRtP4hXt3oCem4tRE1o2xqK8EWzJtFE3bLQQuFUW8/bhpM0xmSitcMXqiGqzrvB9fKJooGA5Klu3aahMQiPhRtxmsnIGwKm+acXy9sRY5mGbPWW4xfWIsg7PTuSY/xfqDA1gsWJAlgmREhiZJa2YT3cmTdVAvH5MpiebuVET1G38B0QQcViXAAsIKR09Mqxhrz1yYq0nM+O1/OomPvuMQHrilb8V1eK9njOMX/+b5tilqLc8vf+/SPD7z1GU4zNuzieaRvriGmCZDlihOT2Twm18+DsthDXPY5fdwOmsiGVZgWA5MN0agVDSdTWUNxENy1QaED75uxI9xUhGKqCqjL65hOqtXJc0uV5QmEE38DhMN6YQs/Y5DqEL3xZfuZSqiYCJdgtQtVO4YE8QQz3G4mehGdlXibFcZu/x933r7IHZ0R/H7D7/S/oZmAqSiSrvPumr8y/FJfP3E5HW9BgLxXG6Ja+iKKpjOmtg/lMAHju5e4Z5zz65uHNqawj8+fw1RTbgT5w0badfJDRB5OY6luKjcUaa8+VVeRg6S3II7c3MjfTENH3r9CG4vi6vLUTkXyWBMOAwULQepsOLnhhnjuH1rCv/2vl145NQUpjMlzOcN5A3RXJ/QZGT0lTroHMDfP3cNB4aSODLSu8zdx0FfXFuxv7htII6PvHmfP2bPTOUCVf1Pjmc3VGHYg0fQthxRoxxfLPmEFkLEfoJxjuga5uI3Mrw6S063EFFlRFWhEOrN/b/79oN1c38AwXzeBHXn0uXwnt/1giJT2JytqE3ZDH7D+5sPDWE8fWndazKbAYQQqDIRjdxuf9WWRAiHtiZh2EtOH4blYGyxhIuzeVyZK+DSXAFX5oq4ulCo6WhqOdxvKi+HJ4KxqzeGXb0R1+kj6gskeNclUwJNdh0+XNKHQtvvMDCVDSaKFPS49UArOXpZpn7DtSD8ws0DCPE+iRK8/c5h3DbQXH/CkZFe/Pefvh0f+8YZ1/lJ/JxAEB4dxppqJFNkAsN3ha34o+790G0HlDCcnsj4a1P5GrhYMJtuanMYx2zeQFSTQUhwwm15LWYtRD7L3Xl+bP8AfnTfFhg2Q163cXIigxeuLOLEWBrHxzJIVyHhji6WMLpYwr8cF7HcYDLkEz0Ob01hKBWCLFGEFGnJ4UNeW3ePzTgObzY0U8v0nDAefPQisroF282LNuOUUa1vq9WYs7zHpC+u+fves9PN9Zgc3pbCzt4ILs/Vdq/b0SPy69788+EvvowMAyjlIFyMIQZAohT9iZBw12QGGFYKJDUzX8kSQViVUbIcGLZwuKRUxHhiD0wRDy2RuYM6Wt4+nMCRkV587/ICToxl2lKLqFevvhncJ46M9OLfOwwf/oeXVwhfMC4IUFQSseyl2ULgZ/R69iFtNjiMw7Ad18nXgWE5uDJX9IkcpyayuBbQpXIwGcLBYSEYcGg4iVsH4ohpcoWLXiM0q7hfPqdpsoSCbvlzc4EzRDWKyYyBqErxwdePYFt3BN0RdYXYREyTEVWjmMiUsFAQ8QIBkClZyBu27/y7Vju7oTb1rmx2UErws3dvx//3z6dgO4K0Td0apCB2EEQ1CTndRsF00BtTkSlZ0C1nBbG65nsAkNwcyWaAKlF0RdQVOWoA+Nxz1yAR8fx4bidF08HnyvJg9VCem/dQawzaFodu2bAcDgIHu3ojsBy+Im/ebO2pXb3p5diM9a+myB2/+qu/ii984Qu4fPky3ve+9+G9730vurur24BtdMRDwRaHaqCEghAO5ohN/FzewM6eKAgpa+y1HdHoxpceCEIIHM6xWLDAuJhoGBeKHKcnsvj/xk/5CppbEhokCkznTKQiCqKqDMNhmM6ZK4KJeEjGPx+bACHA7t4oDNtBpmQj46qOLV0T8y0nLZexBgDgwEyuBIdzSISAE0CVxAJaawFsxuKqFlq1uqk2gbSKmyHobRVrMUm2GxJBIButKi7wgeGx9lIRFSAmdNPxi1QhVUIqHFxxeaFoghKKoWQY8wWjonkprEqIaDIWCyaKhg3DZi0lrTajhVQtrEUir5n3rkck6zD61w9HRnrxB+88jF/7wkvIFC0/kPZUiGayRlXr7Grn+T8sB/+///UiwLnfbLCkRMdaCtIvzRXwS3/7PCihYJzB4cBiQRTtZIkG8lnd2RPBW28fwp9853wgcsev//DeDTl+JUrQE9Xwk3cM482HBpEtWfjepXl895UZPHl+bkUC7PRk8Ea6eEjBQt707w/x/+fCbRhxmLDGrIZbt8QCvVfQ4xrBWz88627dYhVEYN1mG4713UEHHbQXqbCComHXTBo5HCgatm8DvBngzW2DyRDG0iXoy9a5sr4uhGQKSjhKVrD1lXEgXTBx546uwLFeNfXtiCqjP6Hh3a6KL6UEO3uj+B/fOb9ClQ9oXoUlHpLxe28/VLEWl+8Z9IzedLMi58Bs3oRVR1m4HIeGExhIhuAwjp6Yir1b4viZV2/DQsHEsxfn8fTFObx4dbEitrEcjucvL+D5ywsAzmP/YBxH9vTi/pFebO+JgBLRmBtyyR6KREHKujE5EcIUQumm8nouz+Yrf1Ctw7TKTe6OqGCcYSJjgnnECeKRfRgmMiUkQvJ13Y+nwgpKliDglruXEAAKFSr2JdPZVOP4emItcjDeOQ1bqISWx1uEkLrnXCiaK+axjQpfFIIBRBbzRlSVkC5a+NRjF9q2x+/kyToAaudjDg4nkClZmMwIR7IguZB25KnWQlFreX75y98fQzwki5y968rq5QtUSpB1nbe3d0cCfYbl91CVKYpuTpEzkTeop5ZW6zu4Z1c3prI6ZnMmZCoa5zhWLrMyJRhIiuLvbM5AyVqKgQiAnpiKmLZUGrJshsWiBcspCHcNYkGSxDrHIMSuPFGaRmCcw7GrF7W/8PwoBpOqn1ttZ6mUAhtybnrglj6f3EFdAmsQIZB2QpEJBpNhKBLFdHapvnRkpLfC0crLfwLAyfEMzkzmkAwLJfElgpAQXCEQJCEAK0S6ve+VLXtgHPdhVaSVJIlq8MbBx75xBuem836sLFOKbV3i+V5eq3KYyO8xzv16V86wQYhwogFIGUlFkNbKx2/Q/YWHhaIZaFzkzdpiJNcTlTluDcNdYczmDKGC6zqNJ8IK/uCdhze8WFS74a1fCwWRj8zqul/LUSWhev+H3zyH6UypZu5vLm9gKmuDkpVzHgGaVhBdDRSJgIBiKKmtrE25jj5Fy8G27sh1q8lsVpQ3j3sY7orgrh1dQtnYJX4UTRtX512Hj7kCLs8L0sfYYrFmnbNoOjg9mVuRR0+FFZ/osas3gp094u9RrbLtQ6YUikwgUwpVEn9XJEECaQXxULB9b9Dj1gOt5uj9huvHLiJXsuDRLiRKEFEpvnNmBk+cm226P+LISC+++qH78fnnr+Evn7yEiXRJEGm5cEwdToVwdaGIIOkhTaYg4NDtxqSQ8nyd5ParZPXKtclbAz/6r2dwciKYM205bMZh2k6gur2H5bWYdot8Lkf5eH3drf144JY+6BZDybTxylQOL1xdxPHRNI6NpTFXpc40mdExmdHxyKlpAEBPVBVkj62C7LHDza8pEoWmLJE9qgmqtIrNOA5vNjRbyzwwlMSBobhwyrCFw9aBoXggp4x29W0B7e0xoZTg995+CB/6/A+Qdves5SBE5Ni/d2neVxj/9R/ei//2tdPCrQPcj1H64iHENKGazwFfhLr83jajS61KFL94ZCeeuTiP0xNZZHXLF5RMhGTsH0pW3L+gY2l7dwQfeGAP3vuanVjIG3jx2iKeuyRy8fOF1hpEBxIaMkVLOAUta36/WXpVeqMaIqpwSdIUiitzut9rSSD6LSnh6IupyOh24Gf0evYhbVQwxgVh2Rbr4omJLKYyJSwWLKR1E6cnsu6YaSzcqUgEe/tjODCUxMFhsU4OpcJQJYqLs3nkDRuLBQtDyXBTa3wz+cEDQwl/TtuS0HBxtgC7bK5gDJjOGtjaFUbBZHjk1BQ++767/etZTgCybIZMaemzC0E0+LWFtTTySoY3Xq7reuE9d2/HP75wDa9M5eAw4RNMiHjmOOfI6TYIITBtB3N5hrAioSsqCB+mzXzR7WqgBFBlCrNJ1+TrCiLWXI1SJEMy5vIGPv7IK4iqEvKGjcFkuGKsJMN8VT2j1cYg5xyzOeFMqEiiPmfaHGF1ZfzQTO2pnTFOOTZj/aspcseDDz6IT3ziE3jooYfwmc98Bh/5yEfwlre8Be9///vxhje8Yd1sP9uB0xOtK/MRNxHnfVrDZtAtJpIAFoPlKpnt7IlgJmf4DwQHx2zO8IMNoXbJkC7ZfoDv2UAtFA1EVRnbuyNYKJgVwcQHju5GPKTg8XOzbjMEx6XZAnqiQnlTkSjCqoyemILRhRLMKjMT43ADHSCrW9AtBkpFkTikSAipS4SU8gUwp1tNOW3Uw0awurlZgt5msVaTZLsRdKO2GqMhT3G5aAh1r3IrSMNimLF0RLRgSq3eIqHKFDt7olWTiBIhePert4NSYDgVwVsPDzZlt7kRxlU7sdaJvHqoRyTrMPrXF/fv7cOvPrAHH39EKCXJVCSgGWpbZ1dDpmSDu0oOEq0cVzIV3gM24zg0nERck/Hi1UWfFLn8zOXKk4QQ9Mc1mA7DQsGEKlM8cEsvvn1mBo2ijZBM8LGfOgwA+ONvnW94L/b0RvCeu7c3PO56Q5EoemIa3nJ4CD+6fwvyuo3vX13Ed89M44nzc5jMNGfrLVGCdKlsnl3+hZQVJ7J6dVWoXEDXtqDHNUI565u4DbPlDj8bkfXdKoJucTfRVriDDtoCxnmFQlN5aObFp7br3LhZ4M1tmixhMBnG+GIRDhPNnd4ybLnq0D0xFSFZwtWFoluUCQAC/MprdzW15wgSLx4YSqI3qoExA4Yl3Ph8AjiprqJaDcmwggffc+eK6/PuS0yTWm7OMWy2YnmrBW+tkihBIqQgEVJgOww9UQ2DqTDecngQJdPBC1cX8PSFeXzv0jxyy5LuXoPIXz11GVu7wrh/pBf3jfRg32DCL0Tv7I1ApgSmw0GqkDo8PHJqCnu3xDGfNzGYDCFTspY1C0lIhBXM582Kfci+gTgcV1VHlT3VZ9ddjXKYNofDxXHXE/7YrRZ/oHJsd1Afq83BVBOhODCUQE9MxStTOfc8S+4vvTENecOpec7uiLqpvj/L4RhdKEKWxLzljcnnLy/i889fw3vv3bHq9+jkyTrwUGt9/d6l+aZyIe3IU1VT1OKc+3k1SkRBptW9lZevkyhBVFlZLskaNhjjSIaVpj6Ddw8///w1/Nl3z0ORKGIhGdTNK5as+mppjb6DvOFgIOk5ly2tuxIl6HcbYAAgqgki2ES6JGoahCCmyv59LJkOJjOiabk3poIS6othyBLg2NxVeXcJHlXuIYXryMDrC/FwABOZtdkDr6P4fVN44NY+vPvubdg3kEBcU/Dfvn5q3cgdQmkcSIQUFE0HCmUrmlVq5T/L857JsAxVotAtx3dvo54qJeduTYnCsllFkdz7OwHQn9DgMPG8xUMy/q833RY45s+WLLf5XEVIkUAJMJkx8OEvvizeh3GhVGgL0rDtCpsRDsCteXlXIrnzE3cpRgoVwmtffXkCb7tjyCd4BM1Hd0dUlKcYK45wc1VeE1q6EEzFfD1RLce9vTuMnG4jW7IQ1WR8/J2Hcf/evsYnu8FwaiKL0xMZlCxnBRFdtxkoYbg0mwclBN3R6rk/73krV+evfC7W57N4z6UsoW5tSnEYuiMqDm1N4u6d3fja8UmMp4st1ak6gE+iiGpLP9vRHcW9u3v8xjnTYcjrlk/6KP+vnlJ4umTh5dE0Xh5NV/y8P665hI+l/7Z3R4TbSBk8tw9VppApgSK75A+J+s9tNQRVVQ563HpgNTn6Xzm6B+87sgtfOz6JZy7O4dGzM2CMoScWWlV/BKUE7713B372h7bhq8cm8NLoIqKqjDcfGAAAvOt/PgsnQHaJEIqemITZnOHntao1q5XnNMS8I3J4jK1cm46M9OJ33n4Q/+bTz7oq1M2AozuquTE7r2iArIVqtZh2inw2gvdMhFUJR0Y0vGZPD0qW6Am4OFfA968s4sRYBsfG0lVrWvMFE4+encWjZ2cBAImQjENbk7h9awqHtyaxpy8GiQo3XI/k4bl8yC2SrDbjOLzZUFGjbFDLrHTKCJU5ZTR2ISh/7Wr7toD295gcGenFg++5Cx/91zM4PZn1e9NCioSuiILJjFFxne+5ezsePjmJk+NZJMOyu2fiQnCAMeRd0Sbq7jvL721Q4QACkVs7sqcXHzi6B6cmspjN6Tg5ngUhHFu7oitinqBjKaLJSEVUpCLAlriG7T1RvGH/AEqmje+cmcXvP3wmsHK8h5m8gXmXhEddclpYkVzyGL0pelXK85TJkAzDFn0hBKKR22YcYYUirEoglDT9jLbah7QZhJLrgbluZobnyGGLPtfZnIGvH5/Aw6emsVAwAzvQdEUU7B9K4OCQcObYP5RAIqRAUySEXLJjO3oQm1HcL5/TFgtWVYEnhwMTaR19CW3Fs1NOALowncNcwQR3hS4IgV/L8mafgLptLWEjutReL1BK8JE378NHvnICmZKFiCKBgWMuZ8BxHSmHkmEoMsFc3oQmU/z6D+8F5xz/5WunAc6hyhScc59gDQAyATghiKqSOz4CxOKA32d9vTCT1cEYkDdsIdLBOI6PZUAJwZaE1vae0WpjUDj7MD9nYjscNmMApKqkqyC1p0zJrOkCvlon+81Y/2qK3AEAmqbh3e9+N9797nfj6tWr+Nu//Vt88IMfhG3bOHXqFGKx9qgdrzU4aS1rRgncxm5h70PcgGGhaLgLH4PjWrbZjKEnpmIyY2AgQX2rKomKDKCnNFRLwSir27g8V8D/9ebbsKMniu6IikzJ9K2yvQWvK6qiYDi+DSulBNRNVA8kw5hMF2ExICxT6K4dugfGgQXXGtNxxITVFVX9BDglSwvgfN7AXz99uW2OABvB6qbToL0S7d4IriVIwJ3aapo1UmEFuuWAcQ5FXlLtIwQgklC2061gSq3LlbDKk4gem5EQgi88fxU2E6zGh14aW7OAdrNgPRN5zaDD6F8/MMbx1IU5RMuts3l96+xq6Ioovt09p9wfz4AoBHvq1ldmC4hqUuXcsUxZpRyKJBofQlTCYDKEqayOh09Ng3C2lGSqgZAq48BgAqcms2IjiPoN8P/2vl2bbk3SZAlaTMIbDwzggVv6kNMtHB/N4NFzM/jy98d8Ak09nJ3MYrDMcpLzlQVRwJ2ba0z6kvuN11s2iHtcO7AZWd+tQgEQpEWhk3ro4GbDsbFM5Q+qTUBEHHfn9q51uabVonxui2kyhrsimM3pMGzmr3eKRLCtO4JMyYZMPWVDoYTTCHnDwf/xDy/jk+++s6kGonrxoqc6VjCFYiqlBKq0VACwglwYhJrYg++pfl3efQnqvFENXrNlow6fWmuVLFEkIxTJiALLYSgYNpKRLTi6tw+2w3BiPIOnL87j6QtzmM5WNomMLZbwDy+M4h9eGEVXRMGRPYLocce2FPoTGsYW9ZqXpboWxf/w/DVYDkd/XENXRF3RLMS5KEiV70POTOUgEZG7cBgAyn0ivcNEM5RECM5M5a7bfiBdshBSJBQZh+1w0STpXyP3C2vp0sZr1tuIWE0OplYB6OjeXsy4zSwEbhMdAUqmg2sLRfTFtarntByGnqiK4VQIr0zlV7zfRoRomhZ5SFUmLuGdw7IZ/uy757G7N7rqfWgnT9ZBOaqtr83mQtqRp1q+t8obth9/eOsTJQSjC8WWPmejok62ZINSgnioeiml0Wd45NQUHMaxtWuZWhpvrJYW5DuIajKiGjCQCOHfvHobvnlqGq9M5Xx3lYLhIF20lpqLOcdUtoR4SEXesKBbDrxt+UzORF9chSZL0C0HsiQIHrYjCp4SJSuK+yGFYiARwlzeQMl0RL7Gu3/1bnybwQDM5zeeM8KWRAgf+6nDuDJXwHOXF2A7vO2uJbXgMCES8omfuRPpktVU08ny5wxYahilLnHbcVUPZUox6DrFTGdKKFpLQilhhWJLMuyTjbj73H/6iUs4sqe+UMtSHO9ga1ekYvxoCsW5aXFdt/THQAjBZKYExgFVJrAdQeAoP7sgg4kx7jBRyF8omOAAfudfTlXk4YPmow8MJTCcCiNdKpN3KSf1u39KlKArsjGzIrXm9cPbUjd1jnuuYCCr26IuRKs4+DGGoukgrkk1c3/ltV8/j+n+SWqQ5VYDAjE2PBKwR/izbEGE3NYVwVhar1qbKm9gqBZ3N1un6qA6vBpCZaNoGLt6YzgywnynD8N2kClZuDpfwOW5Iq7MFXDJdfyot/ecyRmYyRl47vLC0nsSYDgVriB87OyNYjgVrprDoESQPRSJQKG04u/T6VKgzxn0uPXAanP0skzxtjuG8NBLYwCAoVQwF7dGqDbOTk9mcWRPDyKqjEygHINQ2B5KhTGXN0TTYrXkjdcEzb3cE+A4QqW+2tpECXH/E3Vx0aDVeMZKhBT/2Y6qckNyRztrMe0CIcR37eqNh/Cq7V0oWQ5KloNr80W8dG0Rx8czOD6awdUqe4+sbuPpC/N4+sI8AOG2eXA4iUOuavmtA3HfOcfLl2nyUqN2ECHdzTgObzYQQvzcYd1aJkfLThntdNnwsBY9Jvfu7kEqoiARkpEMq1CkJaddXmU//MHXjeBDn/8Bri2UKuKkcQARVRAS/fhqGcEjCLw5EBBrck638DfPXKkb87Qy5mSJIhmmSIaFINNP36Xik4+eD+R4UI6/fuoKXr6Wxj27e3DPrm4MpcIomuIchBDs6oviI2++DZ999iouz+aR4bjhelXK85RzeUO4rEgiR2oz4dTYFxdjoJVntJU+pM0ilOyhgsjhODBcwXKHcVycFQ3XpyayODmeqUsu9kAA7OyN4qArunRgKIkdPRGEVVk4w1chMLarB7GZuM6b0xRKKp6Jck4GB+BwjkzRQliVVjw7HgHoqy9P4D//80mYRAjFOq7LSbX2obXI92xEoYjriSMjvfhYWQ5jPm+CcdE31p9YErvZmpIwlTXwyKkp/M0vvhpf/P4oXpnKgbmOHxIl0Nw8p+XaFc8XrKaWloDl5QpIpLZ7SLOwGTCV1SFT8XmoJEiMNhPmA6osVTg3A6vrGa02Bm22JDjk/SmXqaCUv1+Q2tMHju7Gp5+41NYYpxybsf7VNLmjHJQuBV6O0x6V4/UC4a19CR4xw2OAJkIypnOGsJPjomklokhIRVRMZU1IVPxsLF0S9j6Mg1DxeoJK9aLyyR7uvwumg089egGf+Dd3IKdbVZlJ4+kS8qaNdMlEd5kECCUEybACy9EwlzMRVmVoYHAcwTIzqjRT2lwUQWy3WYdSMShlImze28nW3ihNj50G7SWsxUZwLUEpgABTD12loJCfC1u+SeTLft/wOmovErM5HQXDQUSVEAsp6xLQdrB6XE9nkZsJrVpnL0dPTEMipCBTsiqa88qJHRIBeuMaJEoqkyy86l9BINZ9/9+EIKxImEiX0BvTQEiNhDrgO5Ccmcr5jYOWw+ru+E5NZGr/chPAs7z+kf0hHBnpxbFrizgRwE3t1FQeUzmzgvzC/f8JeA0Fd25LVT3HHdtTkN0G1FqQJYI7tld/fbMobxDaEicwbO4/t5pMNiTru1VIMmAFyElKq9p5dNDB5gPxminIShVjj0jPuKsou0mwvPkxpsmIqtEKB8uDwwn8uwf24Lf/6STSrpBA0O23TIUy729++Tj+6F23t2Uv5sURYUVCpmT5yk4AXMcREkgRsWDYODOZrUru8O7LsdGMn8BqFoKA2vi4IGuVIlFXKUwoGBdNG3erMu7c3oUPvW4PLs0W8NSFOTx9cR4XZiqb2heLFr5+YhJfPzEpim0NrodxjlREwVRWB4G7D1GkFWqYuu2s2IcsFE1QQjDcFcZ83oRhO+BMjJuQIqEnpqJoOteVmN4dURFVRQJ0sSiaJbwEZUih6Ipo4Jx39ldNoJUcTK0C0OmJLF64sgBNptjeHcHcsueIug579+7u8c/lNa8tFEycn85jZ0/supA7ZEoQkgnyZmukMAKh/kk4IFGR42tXvqaTJ+ugEZpVtV9tnqo8/ohpDBNpHQ7nkF1pftvh4AT4yycvtURyalTUiWqSS+IUeYRmPsNaOezW+w5G+mL+Z9FkKlTcOF+KPQGULIaCqaP848gU0C0H44s6uqMihigXgJKIKHh6YjuKRNEbU9EdU/05aXShCPBmVZ7bh+PjGfzwvi3X6d3royem4tYtcWgKRdFsXyG3ERwOHBpOtrQ2LHefAVySoVv09vY48ZCEiCJBtx2hAi+J5vKuqIpURKkQV2nmua83fgyLg3ORrDdsoZbmKQVSQiBLrgscWyLTcLd5g7l5QAIx/sGBqCa3rL7+7nt24D9/9SSYe35vf+cTO4hoeO2JabVOc93RyXGvRLpgieYxV3G9HN4+kjGOvkQICwWrau4vZywVr8qfDe/fQHubfyiBm/MmIFQ882IeF4T0Nx4cwD++MFq3gWHJHWrji7/dSFBlKtw1yqYJ22HY0xdzyR6C9GEz4Rx+ZV64e1yZE44fV+YLKNZwo2YcGF0sYXSxhCfOz/k/VySCHd1R7OyNVJA+tsQ1MIvDqNJDNlsIRqLMGs01r64l2pGjb3c8V6/B8cJMHobduPBNIca4IlHkDQdbEiEwzjG+qMN0GCTirnllaxPHUj2MEFJzbVouchFUQDFa1jgWVqWG81s7azFrBVmiiEsU8ZCC/ngIt29LoWg60C0HE4slHBtP4/iYIHtcnM2v+LwF08Fzlxd8wpUmUxwYSuDw1iQOb01h30AcmltfXO7uEVKkqk46QcfXRhqHNxvu3JaCQqnf7Fhey/TIBTKliIeVlueWtdhnrkWPyamJLC7NFtAXD61wwKh2nacmMsiUrBVjiQMomgxhVYJpM3B3z7k87m4EiS6J9ARtNl/tmJMliomMLtyxyubmIHAYx/NXFvH8lUX8GYBtXWHcs7sb9+7qwcHhJDjn2Lsljv/2tgO4MleEItMbMo738pQff+QsToylYTOAEtGv2VfmWrpWfVDlLh2jC0X85ZOXUNhAsXL59aXCCvb2x2A5vILIAQB53capyYxP5jgzma3qZrEchAAhWYh3be2O4A/feRjdUQ0hlSLsusnUc6Oq14O4JUEwntbxe/96Bh99x6GGuYtmFPdPTWShSAQ5wxY5DFQ3U5IogWGL+aXas0MpwVS2hKI7xmWJQiIEFmMr6oCKJETMeJstGiscoToAUEm8+Z1/OYWoJq9wXC5fZ85M5VY4foQUCYYt3IwpAWKahKLJwDhvqcYbFAPJkFvHap/lB3VrYZ5LFoHIh83mdETVaMV9Wc1cWW0MypQKN2dwMNfZl4Mjp1suyUMEQ5fnCuiOqLh3d0/d2lM81Hp8FBSbrf7VdIuVYRh46KGH8JnPfAZPPfUUfuInfgKf/OQn8aY3vQl0tR3U64iDw6010lkOByMcEqWIajJmcobfgEEIoEgSkmEFikQQkmXMFQzB8rIZLFctxayycKyYF8p2vDndxqceuwiAV13whlMhnJvOYyZnIBVWKr4HzjksB7hndzd+4w23YiprQKEEO3ojODuVw+efu4bnryxUNDqWLIaSZYASAxFFAgdw20AcNhMJpFS4+mLaLLtrI1nddJLXAmtVcFwrqBKBEaASpkqtf4/pkoWwIqFkcViMQ6ZLw9NmIlEeVoMrtdZaJIQaiITt3atTfdlI4+pmwUZ1FrmRsBrr7HJ4NpDHx9KwHaFSwBl8y2pKRNI5ogqLuJ09EZyZyvkJcMBNvGHp35pr9VkOQpaKxgS12dceidJbdyKqhGyDueRfTkzhd992qMISdrMirEqIqsFD0flC/e+XcWCkP4pbt8Sr/v7AYAKaTGHVISRrMsWBwfbMj16D0Ie/+DLOzeQrNoGEAN1RdcOxvltFUIPvjhF4Bzcb7tiegiJRmDarWgh1mCjgb/RCZjlqNT+CiIJlKqLgg68b8WPeTz12Ac9fXgxkYUuJSKRyh6No2G1rUvbiCFkiYm1evoQGTNJxDjz42EW878iuFeuwd18+9PkXAxdnlsNhHDKlwmGszjmaXatEk4ggehi2g4Lh4DZZwp7+GH7xyE5MZXU8c2EeT12Yw/GxdMX1B0nu2wwwXKfDLXENMzkDAwktUMOEVzRUJYqdvRHoZhmBVxWEXoWy60qc8PZXL1xeqHAb41wUFRkz8Opd3Z39VZNoJgdTrQDEwcFtUTARKlkEsZCMWEiueI5AOObzpp/DyBs2FvImXrgyj79/7hpemcyiFOA5bxc8orJEgFu3xJA3HeTnW3Ma4Fy4EjuMI6TI6I2pbc3XdPJkHTRCM6r2q81TeevsRx46jvF0CYyJdR3wnJ4ohlIh5A2n5fihXlHHUwxr5TOspcNure9gKQ67iBeuLMB2iTCaTBHTFOQNC3m34ZhBFOktN8dJ3MbgvGFjuCuE2ZyJommLvAgX5BZForAchoGEhmTZGh3TZPTHNUxk9KY/S7swG0Bl8nohHlJweGsSAwkNc/n1I67mDXvVa4PnPrOrNwoAvuiKYQvBopzuwHKK0C3mOwXolgMUTCgSbVklsN74KVcGtN0CnfdvYInkSYhQ4gSWiB2e+rhQ/hXuIsmwgmQYLYlbvefu7fjHF67h9GQW5ZotBIAmE2iKhP2uqulGRifHXYl6DtAc3Cd+vP2OIXz22atVc39RVTTm8rLmEE8/zGvuBFoTB6iGZESBafMVpPlkWAHnHPeP9OH2ramaDQz37u7BL/7N85tG/O1GhywJslx0GeFjIBnCSP8S6cNyRJPrdM4QDh+zBZ/8cW2hWFPoyHI4LszmcWG2kugeUSXs7PFcPiLY2RvF7t4oUhG1QoG1HmrFPdcD7cjRtzOeaySyOL5YDJSLEUQNgl94zU48c3FOqIczjnhIRsG03cZyMR+U17eoSyCQJVJzbSoXuciUrEBkk5BMKupkYUVqKH7SzlrMekEpU+Pvj2vYN5TA224Xzh7zeQPHxzI4MZ7BsbE0zk3n/dqjB8Nm+MG1NH5wLQ3gKhSJ4NYtcRzemsTt21I4MJSocG5RJApNcckesiRyfAHH10YahzcbDg0ncetADCfHs1VJCuDArQMxdMfUlueWtdhnrkWPSTPXadsMDz52EZwDqgQQQsuUwBlMBzBthqgqmnF1m1fE3UD9NH9YoeiOCpGeVFjBH37zbKCYpx1jzhM46o1pmCuYLQd/o4sljL44ji+/OI6wIuGuHSncs0u4etw6EMe27khL590MODLSiy/v7MY7P/2sIAzFVEEkdL+7teqDKnfpMG2GrG6Dg2M4FfYJS9crVmaM4/FzM/j0E5dwebYgamAU2NYTxbtfvRX9iVCFK8eVgDloiRJEFAlhd/3RZApKKQzbQV4XrrDbe4I/a7V6EIUrrwHdsnFm0sL7P/sC9g0m6jZWN6O4Xy7IBiz19xDAd+3wxExMxjGQCK14dhjjODGewZe+PwYGUYvwIFMixGbKr48sCd21Wieshs0kELieoJSgO6ZCohSJUCWxw0P5OvPALX0Vjh853UZWF4KAQ8kQ5l1nV0/gp9F3yNGaC8d8wazqoFgLXv4AWBLBXQ6bcYA7sLnowQIXAj6GzVCyHBAQ2IxBIgQZ3cK+wdZyVNXGoCoJgoduOb7pwbWFIjiHT5SRKcEnv3Meqkx9p6PPvu/uqrWnx8/NrlkuvRybqf7VFLnjgx/8IP7hH/4B27Ztwy/90i/hC1/4Anp7NxZbJSj64iEolMBqYUZ1OKBRUSzgLutJchtFdMvBRMaBRAjgqx8Aw8kQbCas5TzQWoN82XOiSASvTGUBoG7T/XzexNXFEgYSGjRJqljAPvi6Edyxvct/TdG0EdVk7BtMIG/Y+NunL+Ofj01W2GoyDuRdpY8zk1k8dnYGFBwF00FY9Ww54V9Ps+yujWZ1Uy15Xc5y3cgDuV1Yy4LjWiDoetfEurgC3REVUU1CLCQhXbRguGoAhBCEZIpkRAE4mmo4Wr5ILORNfPzhM4hVCTiaJdVstHHVQQftQLvUQsrHR96w0a2osBzmF/8lSn3rTnE8xUAyhKmMvsJhy+FiczuYDK8Ytx6pwyN51Aro84aNJF1S0YiHFEy6TRDVCKAEQFG38bXjk3jHXcN1P+tmQdFqr/NbIiRhIlOCIlGfqCMKCcIhpdHURwlwZiq3RsVsXtZoEHwO3gyxiBkwnA56XAcd3Cg4NJzEUCqEy3PFFfl6b9s1lArh0PDmaqAJqmixXOk3b9jQLbaisAmIWdFrEqAUSITbRyr34ohyW1gQMScL9bRgk5NEgVzJWrN1WKIEibCM+QbNfq2sVdXWEosxFAwHikTxU3eF8FN3DSNbsvDc5QU8dWEOz19eCKwckzdsqBLFGw8O4H89ewVnp3P+M0/cprpqDROVRUOtgsC7UYjplBKoEqkgdpRDtxlUiWy4tXkzIGgD4fICkFf4MWwHjkuGKlkOFgomeqKVzxFjHBlmYzqnoyetQrccvHRtEb//8CuYc+261wuuRpGPYo35MCgY53AcoaTfFxd5wAyzV52v2QyxZwebC+3KUx0Z6cWvHN2D3/mX0y6xCSCEI6RI6ItriGkyZImuKn6oV9ShhLT0Ga6Xw+6RkV5ENRnv/+wLUCWKiCoLRy5CENUk6PNFeGXtoVQIMzkDJUtci1AtdCARit6YgmsLNiQKDCZCSIQV5HQb4+kSJrM6pGWN+6pM26pC3yxiTQhIXA88c3Fu3ckvtsNWtTaUr8MgqCBRxmIyopqE2ZxwsI8QCb1xFYwB1xaKrhNMCcNd4YrnJOhzX2/8eMqA3t+BpTwcwRLRIxlWkDdsP651mIj+ZeoRLklFLrAVcStKyZICZNGCLBHIlECWRHE7psmdfPwmRC0HaM4rVe8T4fLnuDL3RylFVOUoWcx1UFp6SL1/e+6HjdBobiUQeYYdPWEYVhnRXiGYzpr+vopSUnOtOzGW2VTibzcjPMJH+fTJmBCx6kuEcMuWOEzX5YNxQQQfWyziynwRl8tIH+PpUs29UNF0cHoyi9OT2Yqfd0UUGAFz+e3O+bcXzefo2xnPNRJZVGQJQDCVeEUiuH+kFx84urtiTGdKJj79xCWcnsjAsB1QAIpMkQrLUGUJpQZrU3muxptTxtMllOp8r5TSChJcEILK2tZi1h7EdUUKKRK6AAwkQtjTH8ObDg6gZDnIFC2cnszi+Jhw9zg9mV1BtrIcjpMTWZycyOLzz4+CEmBvf9x19kji0HASibCCvC6eCUoI0gHVu4OOw04OoP2gVLjI1lq3OYD+uIbeqObPLRqhPoFaphQhhdadW9Zin7kWPSbNXOfXjk8iVxKxtOSRCb23IhIUMDgOhyJTKDJFnAgHcNttwCVE5N6q1eQlSrAloSFvMOwbFAKBQWOeoGOp3nHdERWMcywWrbqikOWIKAQfeGAvxtNFPHd5AVeXNeaXLAdPX5jH0xfmAQAj/TG8Yf8WvOngAA5vTQW65s0GWab4T2+8Fb/1lRPI6DYIJWvaB7Xc3SWkcKRLFjjnmEjrGO4i/j53rWNlh3GX2Ov4BN/nL8/jE986h6LpIK7JkCWOguHg+FgaL11bDJT3linBSH8MB4cT0GQJ/3JsAn1xzc+DeYIJlIg9bsF0kGvSGapaD2LesDG+WFraI0EQpII4oAStT3pz2n/80jHkDQsSheuqIc5DIOYGz+nwZ+/eXvHseMSeM5NZzOdF07/puobWgmVzX/CCtdGuNR5W2nauGw3Nroflud8fXF3Enz16HqmwAuEKK9xghYhvAMFCiLzBbM5oKg8qHKiaeAFE3bxeuZZx0YcjESDq9kYZtgPb4RhdKIHxJTcxVaY4ure35bmy2hiMqBQOY7AcMVdJ7n305iEGjqgmQ5Ubj/P1zKVvFpGTprLdf/EXf4Ht27dj9+7dePzxx/H4449XPe6hhx5qy8WtJQ4MJbB3SwynJ3Mtvb5oMUgU6Itpgr3FUcHK4yhTYuGAxcRCSsnSw1ttLi9nCXrIlmy3oYCiP0ZQMh0/sLcZw1zehG7ZglxiOhhdKEFzLVwHEiH87N3bce/unor3iagyIqoMw3aQKVn49R+5Bffu7sYfffMcpnMrk/1Fi+HrJ6YgUWElPJDUoEoSQEQCnBC01Hixka1uylm4lsOhSMRnkG00C5524XoVHFtFPCShmG8cPMZDrWuFlzN6xQLu2jxCWL/ndQe3b0s23XBUvkg8fm4WNqvN6G+WVLORx1UHHbSCdqqFLB8fJYuBQygJ9SdCK9QEu8IqCoaDrV1hzGQNmA6DIhE/WRRd5trBOUfJchAPK8jrVt2GLYcJ9cB9A3FQSnDH1iTOTYu4ZPmrRNOriB3G062p+25EDKbCOD6ebXjcq3d24e4d3Xjw8Yt1j/ve5TR+/xuv4IFb+3DX9i7RWEIIwoqEqwvCnp4S+JsXDx4Zp2QxzAW0l28ET4XLYRy3bIlVLaw2UvDYLLFIRJaQqeOIUn5cBx10cGMgqKIFpQTvvXcHdvdG8eePX8TpiQwWCpXW6pSIhixKxZ46pEiIh2TM5s22kMqX4ogsVIm6Kv18xVrQCBIhsDivug4vzfmo2PM3A8ZEgafWS1tdqxqtJd1R0fCeN2xIlKA3pqJoOkhoErJc7AEbIVOy0RfXoJu2r0rr3V/CudsExcGWZSw3AzHdNB08ena27jGPnp2FaTpQ1c46txYoLwCVF34kSkAkr6gCzGQNaLJUEc/rbiOLbXPolgOHMfyP75zHTJW811qjfCRxDsxk9VUVZmzGQQkQDcmQCIHhOKvO12yW2LODzYd25am2dUeQCCmIh2Qwzn2nJ6+RrB2iNI3cMJr9DNfTYTddskBA0BVRK9ZSL08hUQKHibilLx7C+GLJdfAQcUnRtDGbFzHHtq4I4iGh/K7KFIpEYdgM05kSon0xn3w3kS5dN2IHAPzY/v7r+O718dT5WfzaF15q6NjaLnjfuMOBhbzpuww0C28dNh2GyYwuHAHc5ilNltATU1A0bYQV4UgNACXTgSwRmDaH7YqqRDUJBKSp577e+NEU4v9bk8XfNVnE+hRL7hyZkgVwDomIxlbFjScICEKKEHkpjx1anUeOjPRWKEBajIMy3lB5tIONi1oO0OLZF6r3+wYTePjkZJ3cnwFKKcIKEQ1ZDvMJSCFZgkQpVBmYywdT6W/0+2RIwXTWRCqiIKrKMByG6ay5Yl9Va63bbOJvHQhQShCikq8e7cFyBMmjN6Zh/1ASpuvyAQj3zasLHukjL/6cK2CmjgPWYjH4+jWYCrf2YdYA7cjRtzOeazTOArvLQrhBerm45WP6yJ5enJrI4qkLc3jk1BRmsjosJgg/zahiT2dNpMJyXWIHIEhBjDFQtyHbZqxmbmwtajEbAZQSv/8GAAaTHNt7Injg1j7oJkPRtPHKVBbHxzI4NpbBqYnMChIM48DZ6RzOTufwpRfHAAC7e6M4tDWJ27cmcXhrCr0xDUB++duvQJBx2MkBrA2C5hIf/Nm7/B4UhzERJ7hxtipRSJTW7EFZq31mu3tMmrnOx8/NgAGQa2xZKAEcAvzwrf2YzOq4OJNHIiRyYP0JDbol+t6imoSJRR2OmzukRNQcxtM6hlNh/OoDe5AuWYFjHjVgXbPemNs3EIfjki6DTPMEgOlw9Cc0vPX2Qfy7B/ZgKqPjucsLeO7yPF66ll4hynRhJo8LM3lkStamInc0SzBbrz6oak5bOZdcJ0siti/f5wLti5WrETmWq/zP5Q08+NhFzOVNEASP0xIhGQeGxLxyYDiB27bEoSkSFIni8mwB3zkzLT5Llee+1V7B5T2IHOLeMS5ceYVoKkdEldGtUExmdHz8kbP4P22GnphW9ZloVJ/0niuLcfzq6/bgE988h+wyciRxj6OU4LaBON5z93b/d+XEHlWiDd3IACEORwmBzTgkQhBWiBBRkQBwYDXm4V2RDrmjFlpZD73YeaFogoBAkyUUTNFzzcBXOLIsR7nIb063lkRGAl5zs7VpjvrEjnI4HJjJm76DPACYDvPXQ8UVC/jcc9dwYCjZ8py5fAymwgr+n6+cwFm3143xpfyzQgEGgvmCgZ09UQwktLpOR9czl75R0RS54xd+4Req2thsRlBK8J57duC3/+lky+cgEHbjCwXTZ9p7d2f5hjVdtMAYhyIJ+zjLTeItR7VmTkpF82fJcnBxtgBW1ojC2BLzT6bAloSGxaIF3XJAiEgefurRC3jk1JRvq7t8geuPh/Dk5Cz+9w8m/EWfuPeIABVuHo678b8yX0JIoegKK5DdhHhElfDee7bD4Ry0CUXqjWh1s5yFq0oUpsMCMUU3MzbbJBlE+aOZ46qBUoKje3vx7MV5OEwEmAoVY7xkMUgOXxWrEVgbUs1GHFcddNAq2t34V4uRHa6i8Gg4wuL1o+84BEpIhfrRb//TyRrXI+O+PT34hxdG616Ht4576kSv2dOLL784VkYUXdocyJSCg4NwjuHUjWOr+iO39eORU9MNj3vHHUOB7UIfOT2NR05PI6pKeM2eHhzd24dX7+wSjYBMxFUeW5zzpXjH4RyMcaQL7Wm0KFfhooRCCPktzfGNFDw2VywSeLu6plfRQQcbDSfGM5hI66CAq8y59DvqZnom0jpOjGdw+7bU9bnIVaAZRYvytffJ8zP4n09cQrpkQ3bnZAC+Qk5fXIPp8LYqb3hxRMl0VhAMgsLhHASoug57c34yrKBkOZC4u64seytKAcIJkhEZlBBkSxZMN5fAgaoZYtlV0G5lrQq6lnhKg+emcviT75xHXrcRDynQFAlj6cbq0hzATM7AXzxxGYBQh/FAiFCCKRgM/+Pb57GrN4qYpiCsSlBluuGJ6Q8+fqEhWYdxcdyHf+zW9bmomwzeftmwHUxldNiMQXGLKoQTIbwAEVd7RTVwkctaKFjY3RfFyJYoFosm/ss/n8boQqni/K0SssrRSHGPwHP8FW4DDEDBdFAwW1O09VxAGBfFu0xJ2IffNtB6vmZzxZ4dbEa0I0/lzQcSJYgq1ffvaylK08pnuJ5Exlr5Rs/xwFM3lalwvhzuCmM2p0N3RTCKlgMCgsGkhnhIcZ2TdBg2A3NTrUWLYSpTQjykYDytw3Q4KERRezWF61Zxeb6IH9rV0/jAdcYzF+bwm18+jmzJasu6EwTeW5RMBx9/+AweemmspdjKU5kdXxTrp0SF0JfDBQGoOG8L0ZSIhLxhYy5v+M+I9zmLpo2i4YBS0tRz32j89ERVcADTOdHM3hPVMJYWJCVAiKRQCjiOuB+JkIyfftU2/OPz1xDVZCTDK120VzOPdPLxNxaqOUB7TT2e6v2bDg7gU49eqJP7U7FYMBHWJJg2Q1dUhUSIqPeaDuIhGX1xDXP5xbZc86t2pDCbN1veV2028bcO6kORBKEtqi39zHP5MCyGnriGA0PJimbBvGHjypxw+Lg0W/BJH5kmiYk/ctvGIVuuNkcPtDeeazTOmmivwK0DtdcYL2d3aGtyhbNHkLWpPFfz0rV0oOu5PF/EUCoMTaIolu11FSrcTr2m9bWoxWxESJQgHo5OK9UAAPpnSURBVFIQD4mGTNNmGEyGce/uXuiWA9N2cH4mj2NjGZwYy+DEeAb5Kmrol+YKuDRXwFdfngAQvMHznp3dyBs2NJfcuhydHMDaIWgu8c+fvLiiB0Vye1CKpgOJspo9KGu5z2xnTNvMdQ6nIqIZlbu1kyr3jAB4zZ5evO2OoYrrY5zjV//+RdddQQLtpv7eVbh+Cre0X37tbhwZ6cWJsUygmGd0QayDQVBv7TszlQMLQOxQJOLHiowJ8klfXEPBcDCYCuNtdwzhbXcMwbAcvDyWxvOXF/G9S/OYLHOHfP2tG2cNboRWCWbrse86MZ7xxcJ0iyGk0DL3SDFWDduBbjLXRbq1WNlhHIbtwLAEicO0hYPP8mOuzBVwciKLUxMZnJrIVnzn9TCQCOGu7SkcGBaEjm1dYRBCoMoUYbcmFFIkSJRgOBXG3i1xnJnMIeQq/3tYTa/g8h5E3RLEFYkSP4cfVoRjUcF0UDQdnBhL48NffBlhRar5TNSqT1Z7rrZ2hXF1QeQwwJfa8Akh6I6q+Mib91UQQ8qJPSXLCeSyQEDQFxOuTZbD8M5XbcVfPnlZuD42dcdWIlNqzi3lZsJq1sPyuNwb3x6xox5ZQ7jZiL8nwgp0S5CVgxI81qNbpnwWoQB6ogoSIRUhRcSE9cgVQVE+Bk+MZbBQMLGjOwKAoGCKHDIlwo2LcQ7DZtAtMWfW24NRSvCBo7vxm18+jmsLJSTCMhKaDJPxDSMKuN5oitzxt3/7t2t0GdcH/+aurfh//+lkywPHZkDBtOGUzeTVzkUgJm8CAonAt9q1GRe2vrpVM8D3AniFEpgOFwGBJDaEps39xkQwjogqQZYoLEdY3Ug2w3AqBMvhODOZw4e/+DL64xrm82ZFcHR0by8+99w1pItCYVyhYhGzyiatap9NtxgmLQOUAFsSIfzSfTuxd0scowtFhBQJsZCMqCr7zTr1sJGsbqqxcAEgRCUMJGhbJrmNis2gnFoO3QxI7gh4XDUwxvHE+TlENclXaHJchSYx5gieOD+H99+/u+X7slakmo00rjroYLVod+OfNz4ODCXw7VemG25UDw0nV4zxj77jED712AW8MpXz19XbBuL4dw/swaefuARNpig26GYomDbmXSXMtx4exH/9l1NIFy0oEkBA/WQ3B4dlcyQjCt56eLCpz7qRsbMnGFFlpD+Gh16aaOrcBdPBt8/M4NtnZhBSKG7pjwNEOKYQqUxV3FU9dhzRdNcu9YPVqN1ttlhEQrB1NuhxHXRwo+Dla2lYDoMsCftib+/mze2Mc1gOw8vX0puS3NEsygvLh7em8GuffwmZkuWq48BXzo2qEqayRltJ5UdGevG7bz+IX/vCSyCWU+GWGTTh5jDUXIe9OT8VVrAoU5dEsvIcjAFhhWJLTMOVhZKIHyhBPCQjo9srHL+WxCPEWuUpCQVZq5pdS7zji6aDIVdpzA7g2lEN5U3ujIvcASUMF2ZyODGWxS0DMaCw1Ex6+7YU/uYXX40zU7kN1wjnWdsHOe7DP7bGF3OT4sBQAj0xFWcmc774iGkzgIjEtPeUEALolo28LppMsyULEVXCe+7ehhevLuK/f+OVFSpmnqLWamzSCUTjFGG8QhylHByiMNAOVGsKZm5ecCZn4HuX5pveG2222LODzYvV5qk2gihNK5/hehEZa92vkEqhuk13EVXyC3sxTUZEiWA8rWNrVxg//aqtePC7F5AKq75zksO5IJ5KAGGiCD9XsDBX1pjHAPBVuJmtBsfG0njXD21b3zdtAG+OLRi2u2Y1o+e3egynQlBlqeVGvXKVWVUmfqPC8u/WU3snrhKgJAGEw3e+nc7qSIaVpp/7auNHJsDWrjDeeGAL4iEFD5+cxKXZAkyHCZEUwF+vOAfCqoTemIa8IZpE9g8l8cpUDsll4rrtmEc6+fgbC8ufP8ufv4XqvcV4w9wfpQS/8JodePrCXEUOed9gHB983QieuziLF660h9xxbaGEL37gNS03um2EdbaDtUWly4fIK/iED5shHlLQHVVxaGsK3O194Fy4jF6eK+CZC3P4ysuNc/RBc/7rgXY50njzQbV60AdfNxJ4XWs0zmyHBY6hnrs0H8iZq9W1yWuefcufPokzU7mGx0uUoGjYyLhCoUtrsshrcbcxu9n81o0CVaZQZYokhBueYTP0xkO4Y3sXDEuI0VyeLeDYWAbHx9I4MZ6pqsQeVJ29O6pgJiuafyUqVKlDirgGhdJODmAN0UwuMaxKLfegrOU+s50xbdD506uRZ4oWKGGgZGneZpzBdpZq5Muv7/FzsxVzfUyTEVWj0C3RKE8JQVa3sM11GgwS89w2EMPDJ6dWNNrXQr21bz5voGA2bs7mnIOXzZE9UdUniTHGUbQcFA0bhBDcs6sH9+zqwa+9fg9GF0v4/pVFHB/P4MjIxhM7qIbVEszWct/1zIU5/N6/nsFCwRS9ncSAJlP0xjTfLVJ2HU8Lpg2bMUiEIF2ysH8oUTNWth3mk2y9P6s9XwXDxpnJrEvmyOLMZLaCNFkLBEBIkRBWqOhRsR18+Mf24tU7u6FIVPxOlRB2yRzLsVa9gsvPq0qiFkUk0XMrEYK+eAgF03HzTuKexEMytCZzGbWeq2sLJRQNGyAAIXypTlvlo5QTg4lnzxAAnHNkdQvbuyKYLZi4d1cPHjk1hctzwcRL6yEe6jjG10M73Ja3xFXIlMJyGo81xsV/EiUIKxIiqoSC6UCViC/mtxwyDe6+0QzKHTpqgQHIlmx0uQIplJBABPdm4O25NFkCpcSd24g/XxCI/Jz4uVR3D/bMhTl8+olLMG2GkmWjYFiYpgSJkIz9Q4K4Hg8pePzcbFtqyc06SF0PNEXuuNHw9ZNTqz7HVLaxXaQsEX8T7i0SIASUAj0xDamoirGFQlU1K29wl08AjAGglWsI56JJZDprwOEiILYZg2lzhFUJMY3h2kLRZ0ppsgTTYTg9kcULlxcgSQRRVYJgmQo2cLlyuCJ5A21lYYZxYDKj4/f+9RV855VZ/MThQbx6Zzd0y8E8MRFRJVEcUiubZjcqVizWZSBrMMltNGx05dRy2AEDqaDHVYP3PPTHQ9AUCt1kvmVvSBXM4tU+D5uNVNNBB9cLa6HEsPrxR1y7Te7+SXBlvoiLM3nENBlFq35RwGHAfEEcI8sUH3rdHvz+w2dhOxyyxF0VI1EMlyjBh163B7Jcw7J7E+Jz37sa+Dgt4OdOuarp5Za0usVwfDzj/7vCRdxdIwiAqCpDliiKpo2wsrq4ZTVqd5stFskHdJgNelwHHdwo4H7HsacOhWUKfLzyuJsIlAi1nKxuuWQ7kVA1bYacbq9J/JsMqwjLFPFUGJQSOA6HRIGi4WC20HiCIgQ112FvzrcYR0xTkDdqJwAVCZjMGTBtByFFQjKsYKEg7LNlSiqawzkqnTQpgERIQU9MW3He5Wh2Lal2vBkw2/izr96GH1xbxLnp/IrfeVfvcCBdsvHVY+N4t7YNW7sisBlDTmfI6RYIIeiJqdjWHRYKaxtk7xM0/gh6XAfN43uX5jGTM1YKq3DAG2mKJGylTZtjLm8iolDs7ovhXa8axtMX5/HQD8arntthgLPKBlsOYW3doilQQ6gygWULEhy42L8sJ6MQAFviGko2w+/96xl89B2HqpLTa2GzxZ4d3LzYzPmz66HoX+9+ya4DikQpdJtV3MdURMFv/fg+xEMKVJnCsB3M5nRRc6DEnycIWZqLlhcpfVJz2b/XA9EqjqzXG8sd3tg6Ml5UiSAeVkBAWm7UOzOVg0REjct2eMV6XA5e9hfPJVUhBBRCiGxLIoQ//pk7mlqfPJSPn6cuzOGRU1OYyer4zFNXoEgEu/ti+ODrR2Azjge/ewGJsAQCupTDV0TDlixRXJot4IOvH8HYYnHTzSMdXB/Um7+Dqj4L1faVOWQAmGmjar3lsFU1um3mdbaD1lFJ+BDgXDQZm7YgfYRVGb0xDQ+fCCa+9LnvXcXduzdGLbn9jjTVx3JQNBpnibCCWwfiePbSQsNzjS0W13yPRqlQtA6CQ0MJfOTH92OhaGIhb+J3XCEzo0qRvpn81o0IQoivlg6IprKS5SAVUXHbYAI/ddcwOOcYXSjh+LggexwbzWA237gvycOfP3YR/+nN+7CjJwIw10nNTX+em8rj7FQWcU123RCEm7J3bZ0cwOoQNEdoOQwTM6VV9aBsLue4+vNneY3ctDlkaYns1qhGXm2uJ4S4jgoSSpYDVaL+XB8k5nnTwUF86tELK8SYaqHe2jdfMBFEQ8l2BRSqzZGUEsQ0GTFN9okeed1GyXKwvTuCPX0x/MYbN4er9HqIzLTarPvMhTl85KHjmPdjdPG8liyGibSO7qgK0zZh2QwMSyIHnHOoMvXddoI4cnivm8zoOOUSOU5OZHB5thAoh0KJIHNEXcKGJi8RlQxXkGFnTxTbuyOQa5Bcl2OtegXLz/vKZM514RY9rp7g25X5AhzOIRECTgBVEutk0Gei1nOlEQqHifsRUSgGkmEUTcclw1Bkdbvi3MuJwQ7nvoNjPRBCYNgMWcOGQgnG0iWkAxIyG+H0RA4/un+gLee6UbFat+XpnImISlFym4gajUGPrODNC4bNYDEOiQgxWcY4GAMIBfrjGtJFKzBZsBkEPaNuM4wtlNAf16ApEhQq6tmNCO5B59Ll67DnhOL1yHMsuUcDtfdg5QSt7qiKLQkNOd1GtmRBlSXcP9KLTz9xqWnHp1po1UFqvbHxst3riPF0cc2FkihZUiEQLC8GUA7bXahCCkXesF121EolBgLB9ipv6BDki8rjOID5vAXmvUYiQuGBMTBOMJ013MQ69+1ubCaUjHSbAbZofHQ4h1Nlk205okgjEaF0NJgMIx6S8fJo2r9mxoFnLs7jmYvz2JLQ8JZDg3jzwQH0xDQUDBsSJYi6AWd5kigo1oottfy883mjLSoemxmbZSNIKQJthOgq+mzKgzeCpU2gh3Y9D5uJVNNBB9cTa6HE0Mr4q8X8f2Uqh/MzOegmgxZwqVssLiVkf+XoHgDAg49dRK5kwYFY15MRBR963R7/9zcKXh5LBz7ug68bwdMXGyve/PL9O/H6fVvwwuUFPHF+Dt+7NB9I0YIDiIUkDCQ1TGV0kfhTREIkrEpQm2zaXI3aXbsUxdYLPGA8fTM2sHdwc+PObSnIlMJxGKjEV8wDjsMhU4o7bwLXjnKUr6GDyZBb6GUoWg70rI79g3F85M372h7/LhRN2Ewo15Xva3QzmHLO62/tq7kOe3P+6YksdMupmWYQDedx/PC+fvzNU5cxmAjh2mIJjHPIkrgmp45FOwOQCMuBlFK9tUShBAsFAyXLASUEyZCCsLpSGWWuYKBoOlAkCs7dxLYRzO55KlPCT9+1Ff/9G6/4ake16l7fODmFb5ycwo6eCO7b04P7Rnpx60AcFIBuOdDd5KlESYWalBKwANFu/OTtg4Hij5+8/cZxVmuE9VSy8YoytsPQE1ExV4OI5TAOhwNRVcL7juzEgaEkJAn42L++gktzhTW5NqCyDL1WqUXTzdExV72RMZGb81WPCGDbHIslC7bDcGbSwvs/+4KvKB1kLt1ssWcHNzc2c/7seij617pfh7emcHRvL544P1fzPjLGsac/hhNjGRg2g1xG7ODgsB0hSCGR2upz62zcgR+7rX+d37Ex5vMiDvMaLWqRI9oN2RUH002GsCu61Uqj3kLRBCUUQ4kQxtKlhsd7zh6cC4U+h4uGE92NRVuNGSglyOkW/vGFa1VzcWOLRfzMq7cDAEKy7L5PZWLOW8+2dUc27TzSwfVBrfk7SO5vMKnhL5+4iILprHhuf+srJ9rqcPCqHalVn2Mzr7MdtA+ECJV/TZYQL/v5+ZnG7g1A8Jz/eqBdjjT16kHNOmM1GmenJzOByB2mzdZljxY0L/K2O4b8uZIxjj/97jks1GhqbCa/dTOAuj00UU20btmOyJPGQwp29UXxE4cHwTnHVFbHhz73ItKlxvWuszMFvP+z30ciJOPQcBKHXSflkf4YMrrIAUiUCLcYSkClpbHRyQGsDkHHzKt2pPC1Y1Or7kHZ6M5xzcyfrdbIW5nrG83FnkObYTUeb0D9ta+8DyAIGs2Ry4keBdOuEDzc6FhrkZlWm3UZ4/jYN85gIiNcjzgXIsKUCMKBwznyhoVURMZMzh2X7v5elSgUieCzz15Bb0zD7TVqfqbNcH4mt0TmqOHStByUAHv7xTO8fyiB/UNx/NE3z+PSbB6piCKcbgiHZXHYzEHWcHBwKIlX7+xelbhCO/P/3nlPjGfw/3zlBMYWSxhOhUApRckUop0SFbnvkCIhpIo8ddBnwnuuwoqEvGFXEOVMV+REtzkm0jpsxvymc5lSnHa/j0Nbk1Wb1CVKQNz6Q21wOAxYLJi4Y1sK3zgxgbxhg7oN7qvKBHV0xQJhtW7Lr0zm/Nqu6PMWDd/LSX6eYy0lgnyYN2wMd4UwkdbBOWDZQlgwrEpIRhToloNSNbX/NsPrTacEVc0FCqaDK/NFpCIKoqoMCiCuLdEGltf9MiUzMJFi+TocUik0WeQCy3vkNVk4/c3mTezui2LfQLzi/asRtFIRFcmwgtHFIv742+cQVSV0R7WmHZ+WY3l8olCCnGHj2GgG//FLx/AH7zyM+/f2NfUdrFXt9KYmdwynImvG7fC+G0oILIcjpEjoT6gYX9RdpjFBT1TDQtHElBsc9MRULObNCpeBpaJJObmDoj+mYiJdqmiY8A7zEuiUEpg2w3TW8NlljAGX54pVGy4kKthjte6HzYCQTHzWZEyTEQ/JsB0Oi/EKVc/prIHPPH0Fn332Ko7s6cFPHB7Eq3Z0IVuyXEYVFXZ4mhyoSWOt2FLVztufCIG5qiTtUfHYnNjoG0GgMTu22eOqof2qLrWxWUg1HXRwI6KZ8ddIUWIsXULJcmCzYGP39ERlUeRXju7B+47swteOT2I8XcRwKoK3Hh68oRw7PFjBekZh2cAtA8ES/ffs7sFwKozk/i04eksfTJvh+1cX8PjZOXzrzHTd146ndfzzyxM4eksfuqOqq2ZkAwWxuQ+rwloxHEBRfDVqd+Vrj0aobx3sKU5utFgkqNDAGggSdNDBhsah4SRuHYjh1EQWljuGvZ2dzRg4gFsHYjg0vLFj7nai2hraFVGhWwyWw5ApWUiGFdy7u/324bXi+moKgtWQDNeec705/z9+6RgWCmbNPbXhMIwuFLAtFUFYkZAzbOiWsFIXLrW19+MeJjO6b81eD90RFYbt4MxUruKcc3kTqkTRG1f9teSZC3P4k2+fR063kHMTzs04UTx/ZQEj/TE/+S7LxFdtdjivSvS4Ol/E1fkiPv/8KHqiKo6M9OD+kV7csS0FRaJwGEfBsFFwCSZBrMPXAkHjj6DHbXast5LNqYksTk9kUDIZrDqBhEQAiwOEErz18CC+dmISf/H4RVhlFZeQQkEBlCzWtjwgcZPl61E/9VTwvfwgc7hrH0/AAFg288eFKtGmktrrmffooIN2oJM/aw717tf779+94ucAcGIsg4WiiTceGMDZyazb6CII+9xVTvWKnc06JlcTuGoXLi8U8UNrEEe2imcuzOF3v346ULNGOyFTV2iMceQM8d4hlTbdqMcYx1zOgGkzWI4DzgFpmdhStfoaB2AxDjCxViVCciBFwEbX0kjd9ZFTU5ApAq1nh7YmO/NIB6tGkNwfIJooaj23L19Lt+169g22J7fQWWc7qAXbCfYMBM35rwfa4UizFgrj9cbZl18cC3QODqzLHq2VvAhjQom8HoLmt25GyBJFQqJIhBQAgGE70E2GiCZDpRKWfEwbI6vbePrivE82iKgSdvVEfffkeEiCaQI6cTZs/WmzIeiY2TeYxMMnp2/oXEwr82crNfJW5/ogDm1B+4zqrX3L+wCCIOgcSV2XuHjdozYWFoomTJshpHDkdKvCbRFYHcGsFpkoSJ70889fw+nJHMA5ZIlClgDLEbUNxjlk1wW+5Lo+9MVVKBKFRCg0RVz7XN7E3z93DYe2JkEJwWLR9IkDJ8czODudq8hX10JMk7F/KIGDQwkcGErgtsEEwmUC2pQQ/NKRnfjvD7+ChYKFkCIhXTRh2A4cJp6LTMnE9y7Nt5Szb2ev4PJm50PDSfzWj+/z3RJSEQWW6zQCxiFRir645jr8CAR5Jp66MIvZvAHORbaKEECTJUQ1yXeWdRiHwYXAGaEiv2XaDkzbwVMXZnFoa7Jmk3rJrB/gevWBkuVgKqtjLmeAc+GCSiDcXOoJu9XDQOLmdDpbLywnHo0uFNETU33nqMmMEDrxhq4mU//5lKiI0xhT0BfT8LY7h/HQD8aQN2wwxjGd0RuQgtoH7l4jIQQSqU5G4gAWixbSRQtDqTBCCsVMVseZySz+6qnLft2PcUEcVCWKLYlQw7m02jrcE1Mxvljye+QjqoxLc0UYtphDRxeKeN9nX/Bri/WIfyBCBM20GYaTYd9QoNX92PL4pGA6mMyUYNiC+JU3LPzaF17CJ999Z2CCx1rWTm9qcsdbDw/i//3qCeSN4BugoGBckOc0mcAiBLJEkC3ZSIYV10oKKJg2siXhaDGcCiMWkpHXHdhl7F+Hc8hlzx2BWFzAed1AUhR9OebyBuxlI3apILzsmhlv6GTCQDBfRkCJqjIY5yiZDuKahKFUGKcmsn6BxmEcT56fw5Pn5zCYDOEnDg/iTQcH0BVRsWCbWCiYCCkSYiEZMVWuOshWE4DVQ63zji4UUTBt2IxhW1ekZRWPDtYeQQuBqykYtkvVJSg2A6mmgw5uVAQdf40UJXpjKkbNUgXpsR4iVdysZJniHXcNB7vwTYyuiIyJbLDjsrqNkEyF41gNhGSKosWQiqhIuU2ted3Ga/f2Ia87DckdnAN/+t0L+LPvXsChrUkc3duL1+7tQ19cg80YcjpDTncbIxSX6KEKFbNqaFXtzlt7jo1m4DChKOEpSKgShUQpbt+W3DCxSNBldr1VWzvo4HqDUoKPvHkfPvzFl7FQMN0mOGHjTClBd0zFR96876YqYlZbQz2L9DAkKDLFpdlCy6pM9VArricB73+19bocR0Z68fP3bsfHHzlX8xiHAemShURExp7+GH5wZcFN+nIEnSV1i+Grxybw06/aWve4Zy/OIqtXTzibDsNkWsfB4QQyJRO//U8nkTdsqLIE03ZAqLAaD2o7zxyG5y8vIh5SkC1ZcLymc7jCF+5pFJmiN6ZhfJnq83zBxNeOTeJrxyYRUSXcs6sb94304u5d3Yi5yjGWIwhA3jqsee5ailRR/Gk3gsYfte71jYS1ys3Uw1MX5rBYtFwSRe3v2GaikZWA4z986RhOLgswX39rP06OLYIQoNgGlSRKgFRYQc6wAxXk1gqsLMFHKQFzi1dhVUK3IgVOaq933qODDtqBTv6sOdS6X8t/Xq0QlQgrWCxZYJyDOWJfKlECx2mtKL1WxA4AODaWxrt+aNvavUETeObCHD78xZcxk21OpbUdsBngcJFHmMuLGowmS0iGlcANYs9cmMPHvnEG56bzlSqw5cQOl+zj/bkchIga2XzBRDKsrKoxLYi660xWx5ZkGGOLpUDrWWce6aAdqJf7e+OBAXzq0Qt1n9srs+2ZIwiAvlj7mn4646ODamgml7+RsFpHmrVSGK81zhrlnzykyoi5a4lW8iJfPTYBvcHeO2h+qwP4TjpJKOiNKZip4Wpa8RqJwGIrBVeKpoNTk2IgF0wH07klV1KJivfaaPWnzYagY6Yvpt3wuZhW589WauStzvWNHNqmsvWJah7qrX1B5/Vy3Mhz5LX5AjK65TfqU0KgKRT98RBimtwysWk1ZEzGOL7w/DUwxqHKBJS4rdsuwQNYat6XiYi7U2Gl4hycc6gyxZnJLH7t8y9hvmBiNhcs1t/WFcaBoaTvzLGjJ1KRCyeEIKRQtxYi/tvZG0VXVMXHvnEGpydzYFw4jERUilREwWTGWLOcfVDUa3YuH69FS/TAqrKEgWTIrwl5aPRMPHNhDn/37FXYjnBM8G5dybShW44ruCe+R+G4sOSILUkEtsPxyKlpfODonppN6qPzjes/lAADiRDmcgayug1CAMcGhP9t6xhbKK7i1R0EAaUEt29L+cSjrO4gFVHAuOjfXnLzKPsHXKd1hyNbsnB4Wwq/+YZb0R1R8EffOgejjSJj9SBTQerw8nJ2gMQrBzCeLuHf/s0LeP1tfXjy3CyKbo0YALK6BcvhcByGgmnDkihkSrElrmI6Z1adS5evw5bDEdVk2A7z+9cBEWv2J7QVAmWeY1Y1d3ndFKIz1RyRW9mPlccnBdPB+GJJ9OdTQQxzOJAtWfjNLx/HH73r9oZz6FrXTjfW7nqdIcsUtw7E8eLVdFvOt9xOicG1RFMIGBeL/7buCP7Dj+1FV0TDi9cW8eB3LyAZkRFWxFeRDCu+ywYgih3lw526D9FERl8xCSyvJzMuAo2gk0W9AU4gGIWWG7EQwHfh8CyHbEahyhI+/yv34JFT0/j68UnMlAUrkxkdf/nkZfzN01dw/0gv3nr7IO7YloJuOdAtB/PERFQVRI+wIrnqpe1XwwDqB3aDyRCuLZRg2IJRmYqoTat4tIq1sui5UbEezaTtUHXpoIMObiwsFM2agSUAaJLYVDvMgWU0noHu2J5q8xVuHhh2MIKtYTtIhZWG8zkHKpIpmixBi0nojqq4Ml8IfF0cwPGxDI6PZfDJRy9i/2Acr93bh6O39GIwGQYAP35BQcRB5a4e5UnKVtTuKCU4urcXz16ch8MEm12iIrYrmg4kynB0b++GWXs65I4OOqiNIyO9+OX7d+GTj15ArmS744Ajosr45ft3Xbek6vVC4zW0dVWmRqgV1wcwkgQQbL3ekgg3PMZhQLpg4ejeXjzjquY1i5dG03ULO7bN8OBjF+ueg0PsP//i8UvIGza2xDUsyhZmsgyOw0XuIeD1qLKE2ZyOHT1RnJvKCmeEskmfQLgm7B9K4vd/+hCmMjqevjCHpy/O4+R4pqLQXDQdPHp2Fo+enYVMCe7YlsJ9I704sqcHffGlZiXDcmBYDtIQBaByVw+1jW5nrcQfNyLWKjfT6D0fOTUFAJAIQT01ElWmiKoSFotWBbEjFVHw3rt3oGjZeOHyPEynPcQOVaLI6BbacLqWsfxOLJFMOKYyOvoTocBJ7U7eo4MOOgCqF6IMx/GbERRJQn9chSxRjC2WGpytEmvp1lGOkLQx5inGOD712EUsFExQIoq96+HyVA6PdOGpzZZcZ9IDrvJnPXjEFO+7l8jK2pP3HgCgUAJz2QGKRCBRAs45TJvD4Rz7BlrXkw26j3jjgS34xxdGO+tZB+uKWrm/Jy/MNXxuSQPBPQ+NDpMlgq7o5lX27mBzoJlc/kbDahxp1juXdcf2FP7++WsNj/vJO4bWZU1rJS/y0mg60Lkb5bc6WImgeY0tCRWf/LkfwqmJDI6NilrX2elcVSEXP5uwQetPmw1Bx0xXVL3hczHrPX+2033Mq9M+eX4u0PH11r6g8/py3Ihz5F8+cRF/8MjZCrEcxjlsw8GoVcTWrjDyhtMSsalVMhHnHC+PpjGV0f2eT69HkxICVQYc18FDkQk0SUIiJIMxDt12ULIYSpaDkuX4e+RXpmq7tagyxa1bxOc7OJzAgcEkkpHK2oJH5gjJnshkdXGre3f3IBlWENckJMOq634ujuWcr0nOfjlq9TgGaXb+7PvuxqmJLOYKBv7k2+cxtlhEVK0kQzUiu3l1i4JhleVexBdBxAlQ/snLbyPnHA4T/STTmZL/bCxvUjcd4ZjNGiS2tnWFkQirCCsSMrrtPw/Ee1/eWq/E8YBxVQerx/Lv3uvfDskSEmEFCwUTFuOu0JgQ7OcAopqMX31gDwDgifNziGkyeqMUk1nhJrPWOdGgLlNAZW5hJmfgH18YK/tZ5VpmMWAyrUOinhsORSJcu+bkrcOff/4a/uH5a5jK6pAJRdYQZKf+mIaIJgsXGy7i1emsIIv8xhturekubzNRc6ZEOBUvR7PxhBefKJRgMiOIHQol/jwrETG3FA274Ry6HrXTm5rcwRjHYqF9ltjVBiMBR08s7C9UY4sl/OevnsJH33EIu3qjAEQTqIeYJvtJb+98nItGCNNmfhI96MBv5/ywXEtUooKxygEYNgMlDBdnckgXbPz8vTvwnru344UrC/jasUk8d3nev2abcTx2bhaPnZvF1q4w3np4EG84MIBkWEHesJE3hJtJTJNxdb64JmoYjQK7/oSGxYKBrV0RzGT1plU8WsFaWvTcqFivZtLVqrp00EEHNxa6I2rNwBIQ6gFRVcKPHxrGXz11peH5dvZE1uAqNwemAipYesc12jTX+j0hBMlQsKbLiCIJhYgynJ7M4fRkDp9+4hL29sdw9Bbh6LG9W3x3lsNglRiyJQvUVaCPqBIiqojrmlW7Y4zjifNziGoSbIfDdBgcJjb+EVWCLBE8cX4O779/94ZI7HbIHR10UBvPXJjD5567BpkSDHeFQYlQVi+aDj733DUcGEreVLFkkDV0Le3mq8X1TsCmgyDrda4ULL+QKZoiwadKyOp20/NjtIHa19eOTyJXaqwkdHGugHjOhCZTXF0o+pazjDfXgEkpgc2B2wbiODmeASCSX15CkQEwbYZX7+gCJQRDqTDe9UPb8K4f2oZM0cKzl+bx9IU5fP/qYoUytM04vn91Ed+/uog/+c553DoQx317enDfSC929iy5bIoxJRoWAbhFjCWyh7TKtbLV+ONGwloplTZ6z5msDs11lKlFxJKpyJktFCvH377BOBSJ4ovfvwbdZsib7Wsw8lzVNipKllAbGkyFYAVManfyHh10cHOjWiEqb9iYzRnQLRs2AyzHwURGR0SVKp0cGkCiIiZIhmTkDHtNC5qFNs71q8GpiSzOTuXAOYcsUVdVbv0XDomSpbcN2EBeTkzxCCL1vjOClUVkAuHYwbnrriURSAQ4M5VrOU4Iuo+4f6QPt29NddazDtYd1XJ/QZ7boFuVhsN3A8emHdw4aDaXv9HQqiPNeueygtaL3rBvS1veLwiazYs0yls1e1wHSwg6vubywjntnl09uGdXDwCRKzgzmcWx0TS+8tIE8kZl7pAQUR/baPWnzYigY+ZGz8Vcj1pAu9zHvDptUNQbm632Adxoc+RT52fxB4+cXSEM4MFmHKOLRWxNhVsiNgUhE6Udhqmsjh16BIbFYDoMps1wfkY4X6iy6MuUJIBA1PLsMkFty+awHRtX5p1A6viAaEY+OJzEfSO9ODiUwEh/DEqVa9QUCSGZIqLKgZ3KT01kcWm2gL54CKFlz8ta5ezLUavH8QNHd+PTT1wK1OzsXZcm0ZbIbqcmsjg9kUHJYiv2RHzZn9SvWXHfzUMiBP1xDUXLqcihl5PFfOH2sCyOK5gwbbYiVzKTM0GpEKEqT8FwAKtZTadz7RfD66A2yr/7+byB//Gd8xhdKKI3piKsCqG7pVoqRyKs4OPvPIwjI704MZbx62iWw0AgcnO15r12wEvRBkz7gQM4OJzA2amcT7Sr9zpvrMgSQcliMG0DYVWuWXP63qV5/NWTl3xil8O4IDsxYCpngOZNn/DkuQxfnMkDQE1HM8ntTfcIbMvRbDzhxSc5w4ZhM9exo9xBTcQT9YgsHtajdnpTkztOTWSR1dtH7qiGkKr4i+jyhcpjHQkWL4HNBNvPs+lTXIXmQdeOZjRdAlvDAV8PHIBd9t4KJZBcNhSBUC20GEPesFG0xEZQogT37u7Bvbt7MJPV8a8npvD1k5OYzy8N8LHFEv788Uv4q6cu44Fb+vAThwdxaDgJhwGZkoXzMznoFkM8JAgvywdCq2zuIIEdpRT//kf3ojeqrbmTxlpb9HSwerST6d9BBx1sbnhWrI2scmnAbdq3Ts/gh9zk6s2GoM2QgpBrrrDZWw6Hi+Oq4Y0HtuDTT15q+F6//1MHQSWKJ8/P4Ynzs5hIV1runp/J4/xMHn/91BXs6o3itXt78cAtfX6TKeMcBcNGwbABGNAUCRG3wXR5YqUWvE1AfzwETaHQTQabMciUIqRS6BZb02RMBx100B6UN8kNJsMV60UyvD6qORsNQdfQtbSbXx7Xf/WlcTz00njD1wVZr4Pu1iczOi7O5NGX0GA4DLrVnJTzGw/UL+CPp4sIckbdYpCIjUyJwWbcbyySKMBW5sLrQiZCkUqVCEqMV6g7ezmDY2Np/PxrdlQ0hCYjCt50cABvOjgA3XLw4tVFPHVhDs9enEdWrywyn53K4exUDp95+gqGUiHct6cX94/0Yv9QooLAYTkMlsOQc/M95WtxLZWrWlht/HGj4Hq47iwURZGkJ6ZiOqNXVZcERKI1U0ZmUiWKtxwewHOX5lGyGBIhBckwqXiegia6q4Fx8bw3pk+1F81es8MYZrIGuiJK4KR2J+/RQQc3L5YXovKGjfHFEhjnkCgFpaKhgXFeMecGAWNivf+V+3fiE9++0JycXJNohnSyllgomj4RkBDAtte/piNT4SRvMw5CgLAiVAbn82bDouTZqSVVZ8ZdlUlUX4c4AIuVqWICoFQ4fRBwhBWKnujKRolm0cw+grp1qc561sH1RpDnNqRIMJzVE9Nuhj1JB9cfzeTybySsdy7rW6dnAh+3HnWlVvIiQWsxjfJbHaxE4PHFgR09Ud+BvmQ5IITgru1diKkKHjk5he5IGIxz5A0HlsMwkAghrEmd+tMq0eyYuZFzMRuhFtAqvD1y0HxcvbEZdF5fjhtpjmSM4w8fOec3ONdyMWAMePOhQdy7u/n1bTmZiHPuu3BwLghuFMKFYy5XScZJhlQolEAJKVgsmLBtDkL4CgdLDvgN3bVAACTCMsKKhJBCkS5aUCSKn75rGLRsDHhkjrAqISRLLY3565Gz91Cvx/E3v3wcps3QHQ3e7Nwq2W2uYCCr23AYr1sTU6hwpLXYUq4mrFD0xUOQKIHisBU5dI8s5t0/TZYQVmVoMsW1hdKK/JZhC8Gl3rgKQip/vZpUWH9Ma3xQB21FOVFQlSuJR9u7IsgaNrIlG1FNwh+88zDu39sHoHJMes8ZsLp6VC0sP2e98xP3f95zeHm2gM/8wqvxX/7lFC7OFgK9n+MIgofDOEqmU+Ha56GaeFBOt9y5GO68z6FIFByAbjkwLAcRTUa6ZNV0NMvoFlSZQq4y17UST3jxybHRTMX3JC6Rw2EcIUVCPCRjNm/WnUPXYx6+qckdC0WzZnG4VSwfEDndRn98iZRQvlABQE9M9S25vAdGcumCNhOTxELBhFHm2gHAt3y6HukRiWLFgCFEuHgwxkEgVGkzRctn/PcnQvi39+3EweEEPv3EJYwuFCuYaZbD8e0zM/j2mRns6IngrYcH8WP7tyAZUiFTMaA5FwG/cAwR79kqmzsoS7w3qq35pnU9LHo6aA/axfTvoIMONjcoJYGscv/l+ESg8+WMtSWabmQI27zGBUyZUswXzIYKn4wD8zUKmZJEIRGsSMSUgwJQFRl7t8SwbzCBX3ntLlycLeCJ87N44twcri0UK46/PFfA5bkC/u7Zq9jWFcZr9/bi6C192Nsf89dzw92ULBZF/BZ2m0s9V49qKN8EEAgnEGApXlnLZEwHHXTQPlwPpfuNjqBr6Frve8rj+kdOTgZ6TZD1OlElmVUVROyBGQMcJhSNHBbMmlciYk2rh+FUcDUww2bCAhdlqsxNJhoMm2FnbwyjCwUYDlvxcg7AZAznpnNYKIgkW8FwkDMsmGVNmCFFwn0jvbhvpBcO4zg5kcEzF+bx1IU5TGYqyZYTaR1fenEMX3pxDKmwgtfs6cF9Iz141fYuaMvIlOVrMSWkwtVDlevfy9XGHzcKrofSXkimoFSo8mxJhrCQN6Hbzorvozy2G0yG8DtvO4A/f/wSShZDb0yFTCksR5CXfIdcLDWgtpJXuw49uoGE172ikVAWEgWl/kRzRfJO3qODDm5OVBQfwTGbM8C4KNoREDGvEKA3qmDKVQyU6ZI6XD10RRT86bvvxGzOBF9j26OwsjHKXd1uYwMhaHv9KShkStAdE9chU6FsxzkwkzcaFiVNZ0mBcnm9qx44xPrdHVVF0dV9X91mVRslmkGz+4jOetbBRkCQ51aOSMjoxcYna4CbYU/SwfVHM7n8GwnrncsKWi9ar7pSK3mRILWYIPmtDlai2XEYUoToWAqi+a1kOTg1kfX7kCghiIWUChX5Tv1pdWhlzNyosetGqQW0Am+PHBT11r5W5usbbY48NZHF5bllDcQEIGW5Wu/PLzw3ilMT2abcaxzGsbs3iu09EZydyqMvVrn35ODIlizs7othZEt0xev7EiriYQXjiyXhXll+UXVAiBAa8mocEgEGUmFEyuoTiTDB6HwB1+ZLuGN7CiGFtkzmWI7rkbMHGvc4XlsooWTZ2JKoTkqotc60QnZLFyxR31qWw8CyZ0uRBNlmMBKCw7mfrwCE8069xvDy+6xRitmcWTVJwrgQHpvLmb5jg+TWAleTGarmVNDB+qEW8ej2bckV81TFmFQoNJmiZDlr0tvdyjm9+lHRdPD3z1/FleXzcoP3sxwOikoyRDmq9UVIlPiEOM/9l0PUaokEWDaDbgmyyO3bUjVIXgkc3duLzz13rS3xhBef/McvHUPesOBwMX9zLtYTSgj64hpMhzecQ9djHt4Y2e7rhO6ICkoIKBBI1bIRFEkQHDjnMB3uKiMJFU7RlCfgLVTPXJzHTM6Aw7g/sYOIB5dDkCgsh8HEkjqSZ4PNGAelZN2LAxTwmVSkTJGcg/vX1BVRoMkS+hMSuhyGTMlCTrfxg6sL+JPvnEfRdDCYDAPgWCyK35V/iqvzRXzy0Yv4n09exutu6UNXVMX4YgmJsAKZUmiKsMMhABYKJva3wObeSCzxTtNXBx100MHmQxD1gK8FJHfcWDpWzYGQYJ+eEF7htlYtLPfOVMuVLV2yENMU5AyramKVEiAeUhBSJQwmw8gZFoqGg5H+GEb6Y/il+3bhynwBT54Tjh7LWeyjiyV8/vlRfP75UQwkQi7Roxf7BhO+EofDOPKG7ZJfDaiuvaqn3uHFAdcrGdMqgqoNbLz0bAcdrC2up2rORsZGs5sPug4HOa4nqkKmxLfkLp/3vNfLlGA4GYYiEVfEAJAlIWJgO7yunTclQCykIF2qXxB66+FBfOShY9ADdKB7xI7VQJEo3rC/H3/0rXNwaiRXHCbW4rmCAVlKIhmhSEYUmDZDwV0brbIXS5Tg9q0p3L41hX/3wG5cmS/iqQtzeObCPM5O5yrOnS5Z+MbJKXzj5BRCMsUP7ezGfSPCRTS5jHDDOEfRtFE0hRCF54rlrcfLiZerjT9uFKxnDsW0GeYLBnpiKrZ1R3FpNo/emIrhrhAMiyOvm1gs2Sue26N7e/FbP74PV+aKGJ0vIBFSoEgSKBHPSC1b9s0GqWyOWYGyopV3yBsPDGzIInkHHXSwsVC+B+W2IIdJlPj5fw5RlyCUQnJrEpwTKJJw+641p0oUUNymsa6IIgp6a+hM/rp9fWt27mZwYCiBWwfieO6yCec6ObF7DQXDXWG/NqXbTqCiZLmyYL0Nv7e8UCKIPgSAxRjm8kvv2844YaPtIzroIAgaPbd/9M2zAIKTO27mPUkH1x/N5PJvNKznGtTOPFU70EpeJEgtJkh+q4OVWM04JIQgosrY2RP1G0WrKSBvtPrTZkMnl1iJzRrDe3vkwPXPOmOz2fn6RpwjF4q1HW2W/1SRCc5M5vBbXzmBj77j0IpnhDEOw2YwbAemzWDYzK8rvOtVW/GJb53DbN5APKRAlQhMhyOnW4ioEt5z9zYAwLX5Ik5OZHBqIouT4xmMLpYCfQ5FEsJR4EK5PhaSYTORzw4pFD0xTRA7iNu4TIAolVA0bUgSQXe0vfPq9ep7bNTjmAjLKBiiFzRVZS2pt840S3briiighMDmfCmHUf6n+4BJlEBTJGR0228M1+1gjeEHhhLY3RfFyfEswgqFbtm++8xycACmw0AAhGWKnT1hZEq2cI4hBIbjIKc359zYFQsoKtfBmiEo8ahyTGroi4cwtlDE2tA7VqLhmlVWP/rOmZm6ROxaYBBz3kM/GMOh4WTFPajaF8FX/pV76mu87N8u6t3rA0PJtsUTR0Z68QfvPIxf+8JLyJYsv+c9pEjoi2uIqlJD4hewPvPwTU3uODCUwLbuCBYKJiSXNNHqcPIcNziHX2R1DThgM4ZytWXDYZAJ8MipKTiMY3t3BHN5E4btgDNxLkqIq6oEpCIaHMYxnRNKldzxVDVrX20jNYRWIBHB7i9ZDiybCQcN9zM7TLiTJEIKesosoRSJojemIRlS8JGHxlA0HfTGVL8wNJCQ0B9nmMwYoJQgpy/Zups2wzdPT/v/zpvCJk1TKOIhGZbDEVElvOPOrZjLG4hqMiKqtCJ4qIZ2scQZ46u2SOw0fXXQQQcdbE40CuIH4sEsEoMeFxTtWJvWC5QGo9hSSkFBltSIUb1hlhCA1qAQdEdURDUJsZCExaIJw1qy3dQUKpILXBwXVoWiN49xFEwHeV1sunf2RLHzNVH8/Gt2YHyxJBw9zs/h7FRlk+lUdklNvDem4rV7+3B0by8ODicrmkZNm8G0TaQhkh1hRaiI7+2PbRgSagcddNA6NhtRaz2xkezm27Fee2vv1fkioqrkCxiUb8kJRI4gGVZw544u7Dkbw4nxDAAOx3XwqNWvTQGE1Mq1qt61nJnKYf9QEj+4lm74udqRNnjb7YPYkghXEDuqrdMOAxbylftaVaZQZRVdURW65SCnC+JFuZAFIQS7eqPY1RvFz9+7A7M5A09fmMPTF+fx8mi64ljdZnjqwhyeujAHSoDDW5M4sqcX9430uCITlbAZQ15nyLu5CFWmCCvCYSukrD7+uFFQnkOZzOgIqxIkQuC4hax4SF610h5jHOmShUzJAudCHec9d2/DJ751DnN5E5pMYdgOMssEQhIhGb/147fh7l09AICMbsJmQMTNX83mdOgWWxFxtkvoZT3hib4sh8hH8ookuCJRxDQJ92/QInkHHXSwsVBeiIqqVLhzS3DXdkE+jagSIooEiiWyB3eVzZbzSRVJKJx1RRRMZ038+eMX8RtvuBWJkIJ00Wx73QIANIlgIL5yrb8eoJTgg6/bg/MzOczmjHVnFBII8rDDhAtLVJMAjkC5hANDCQynwkgXrRWxTzWoMhViYJy7zwSHwxlmsjpoMoRMyW6rIu9G2kd0cPNgtbnWes/t114ew4sB9m0ebuY9SQfXH83k8m9ErHYNCjqXXK+6Ui20khdpphbTQXNoxzisbEKTOvWnNqOTS1yJzRjDe+NkuZt0LdQbc0HnawIgHLAGsNnQHVERkinyhvg357WbkJMhBWG3qfZTj13A7VtTsBhbQeQoB+McF6YLsBnHu35oG548N4fRhQJynEMG0BfXsKcvhi+9OIb/+rXTyJb1JTaCTIFkWEUyLPsOLYxzzBdM/Ny9OwAAn/veVSTCMiKKDEqJm8MVz3fJcqBKdE2+z+vljtOoxzGhyZimBNmShWRYWdN1piemIVxWkyNlTePe8yVRAkki+IXX7MAzF+ebbgz/3qV5IWpu2MiUgomlcQAOY7g0V3R7hsVF1RJHqwfd2mxVjLXH9eiHCkI8qjYmu6IKZnLr0+9b89kkK3/vMB7Ylbna+/zds1fx/asL+H9+fL8/fqr1RThuvrB8TAKiHm4zDokShFUJaZdgUf69vnakt+J7bXc8cf/ePnzy3XfiN798HEXDRiKsIB6SYTocU1kj0By6HvPwTU3uoJTgzQcHcHwsveqCAiUEls0rCBcOBwjnMMtmZ2+h2toVxkxWR1dERUiREAvJ0E0GmzHIlEK3HUymSxhKhZEIK+J1JRMlyyk711IRpfJa0NDqrxlIlIAxLthXDvM3IszhIO5A02QKWSLYP5SougC/MpXDtfkieqIqFEly1b1cNVNC3M/v4ANHd2OxaOKRU9PIVGEiMwAli0G3TOztj+IDD+zBndtTvgq2RIXiQDwkC8ZsHayWJf7MhTn/tZbDoUgEe/pjTTPCOk1frSMsAaUApNZw/Uehgw466KBl1AviCQlWxAh6XBC0a21aL8RUCYulxgmUmCrhju0p39p0ecMsIPYkqkRxx/ZU1XOUJ6x39URh2NyPuzSZYDpnrkgkEEIQ02TENBkO4yiYNvK6Dd1yMNwVxrvv3o53370dU1kdT52fwxPnZnFyIlvxvnN5E195aRxfeWkcXREF94/04ugtfbh9a7JCFYkvUxL/qTuH8Udz5zCR0dEVURCSpQ1rVRwKuB6HOutxBzcZNpJb4EbERrGbX+16vXzt1W0GQgGFEtjMa3Qj0GQCRaLYP5TAoWFh1/uRr5xATrdg10lIKBLB9q4IQgqtulbVupby3MGKz4KlJvF2NFfesb0buWX791qnXX5cOUKKhJAigXMVRdNB3rBRNB0/d+ChL67h7XcO4+13DiOv23ju8gKevjCH568soGgufW7GgZdHM3h5NINPPXYRu/uiuG9PD+4b6cXe/lhVYQhBvBQOpIQQbO0OQ6FC4ava52oUf9xIODLSi5+7ZzsefOwiJtMlMAiCRDys4Ofu2b6qWDOnW1gsWG6hYwl3bu/CWw8P4e+fu4rF4spn58ieHvzGG26pUADrjmjQZIKsYWE2J1ThJAoQVvn9bbaSCIEYI5azcmwrkhjVjIkGbM0lKe0fSt60a0wHHdxMaEdBs7wQNZs34HDAsZd23gRATBMNFrJE4XAHfXEVOd2BbtkVE6wm04p11nOlBoD9QwkcH0sjr9ttJ3jIMsW+gXh7T7oKHBnpxR//mzvwW185gSvzwVX5VwtPcMxyRN1Gt2xkihZKFgtclHz33dvxn796Uoih1fmePHcXBg5KRYyWLdnQLSGQkSlZ2DeYaHtOaqPsIzq4OdCuXGut53YgGQn0eq9O20pOtIMO2oVmcvk3Klpdg5qZS65HXakeWqnLtFKL6SAY2jEOr1cz8M2C1dYyb1RsthjeGyfPXpwL1Phab8wFna+7IgoGk6Ebco48MJTAbYMJPHtpzm9ur7bV1GQCRaGwGUdUk3BuKocnz8/hloFYzXO/dG0Rn39+FKPzBViMgwJIhBUMdUUwlzcwkS5htmDh3HS+7jV2RRQcGEpi32AcuuXgqy+NI6KJ3kOyjIxlOhwyJdgSC6EvoWJ7dwTj6RLi2toSGarherjjNOpxNBlHIiRDlaU1X2cODCWwpy+GY6PpClcArwZGiZhzwzLF/SN9+MDRPU3l0J65MIff+soJ5A0bA4kQ5vMG9GWTwjKTEB+GI1YBAkCm1fdygbDOgiEbHRu9H8obk5967AJemcpBr1OrXQ1qPXfVzACq5fYoaeAS3wAcwNmpHH7zfx/DR99+CA/c2l+1L0J2XaA5577rL+MAAUdYoUiEFYADowtF/OLfPN/we213PHH/3j780btu95+p2bzZ9By61vPwTU3ueObCHL7w/DWEFQkFs7XBRAkQ12RENAmzORMUwr6QMQ7LTbTNZHUolEKRqb9QvfHAFnzmqSs+k5GAuPbYYuEzHWHKQ92FnxCCvngI44slgDB/INZis3pkxHoDMaiFnKeCSTiQCCnCjoZz//pSYQUcQEyrrdToMTc1WQKlxCWgEGR1Cwt5A6bNYHPgH18YxZ7+GP7vN92C//nkFYwuFKtePwcwltYxtljCbQMJ31rcYcJSLadbUCSKqNuQqcrVA+ZWWV3lC3hXRBUbNIfVtYerhU7TV+uIhxWU8o3tCOPhjk1ZBx10sP44sDXRcK0l7nHtQDvXpvVCrfW52nGHhpMYSoVweW6pKaL8/nIAQ6kQDg1XD+bLE9bTOROpiIKoKsNwGKZzZsNEgkSFQ1kipMByGAqGjZxuw3IYBhIhvPNVW/HOVwk3safOz+GJ83M4PpauINwuFi187fgkvnZ8EomQjCN7enH0ll7ctb1rxb24fVsKH/7RvX5CKg0LmkQ3pFVxZz3uoIPq6BTKNgdWs15XW3tVmWIiXYLhFhIEOCwHSISWvvMjI734vbcfxL/7+xeRN5yK9yq/FokQgKDhWrX8WlIRBUXThmEvNWX6e3E3V9EdUTBbWJ29u0wJeqIq8rod6D4GcdokhCCqyYhqMhjjyJeRK5cjFpLxI/v68SP7+mHaDC+PpvH0xTk8c2Ee84VKNZxLswVcmi3gf33vGvrjGo7s6cH9I704vIxw6YFzjh09EfQnNIwtLqm0NRN/3Eh45sIcPvfcNciUYCgV9sVGSpaDzz13DQeGkk3HJ7rlYL5gwqiR4H7p2iIeemmsgrTjIRmW8VN3DlUQO1SZ4oFberGnP4bnLi+AcQ6FijwLo6xqfqnhc0sqC0LtIoV4xaWgzc2UAsOpEBgHMiUTi0ULnItzWDYDpQSMi0ZeRZKQCCudNaaDDm4CtLOg6ZH4/uhb5yp+7s1XCwUTIYVCkSg4OCyHYyCpIa9LmMkJ+U2ZUneNWJp7PFfqdMnyY2ObcRSM9hY3ddPBqcksbt+Waut5VwsOIKpJbf+81aC4Tuu2w8D4UsxXNB3sHwpOsnjP3dvxjy9cw+nJLBhbuU56etHUdeqwmSjIdkdUdEdVFA2xvv/66/fi51+zo7MWdbBpsR651oPbgu0jeqMqpt25Frh59yQdXF80k8vvYAnNziXrXVdqhFbqMqutxXRQG+0ah9ejGfhmwWprmR1sHBwZ6UVfTMVktrHier0xF3S+1hTphp0jy90tp7NG1WMIgL54CMxNliqUIMs5Mnrt+//9Kwv4g0fOIm/YIISIvkPGMV+0gDoiCwTArt4oDrji1QeGkxhKhvxcBuMcpydzuDSbhychQYiorQDAYsGEJBH8+eMXYDkcjDMUTAeji0X0xUPrXgdcb3ecID2O+4eS+MDR3fj0E5fWdJ2hlOA33nALfvnvvg/DYpAl4jfMMs4hESG6NrIl7t+ToI3hjHH8+eMXfWIHIQRdEQXnZvIwXYIHhSu8REQupjzXX958bzGRQ/FqDM0QPfS1sL/dpNhc/VBCEEWiFBJlfu+1+E3w77/msTXYHY32EJQI13dxPcR172jtGWMMyBQt/Ol3L2BvfxzdMXVFX4QqC4KHbjlQKNCXCEGVaAXpezCp4S+fvITCdfpe2zGHruU8fNOSO8on4V29URRMG1MZw3Wm4BWspVqPMAHwC6/ZgUfPzmIiXQLnXCwUEMltmRLIVLAVJzIl9EZVf6GKhxT8r2ev1mQyeoXjcieQmCZjuCuM2ZyOkun4i4JEid80YrOlayekcgAGM2ms/VlliaAroiIRVjCbM6BbNhzGkSnZuHtXFz74upGag2k5c5MQgpJpYyZrwGZMMMIAhBSKS7N5fOJb52E5HMOpMAiAdMlCVrcqGiSLpoM//vZ5fPqJS/jRfVvw1sOD2NO/xNi1HIZ00US6aPruKFFVhrRs4DTL6qq2gANAiEoYSFBMZQ38+eMXce/unkCDtNP01Tq0gAmMoMd10EEHHbQT/bEQ4iEJWb12AT8ektAfC636vdq9Nq0XZCmYipd3XEyTKxzKymM0SsTv66FdCWtFEra4qYgK3RLK4gVDxEW9sSU18cWiiacvzOOJc7N4aTRdsWnL6jYePjWFh09NIapKeM2eHhzd24dX7+yC5rqP3bm9C7dvS+HCdAEMHNu6IhvSqrizHnfQQW10CmUbH62u17XWXlWmfsM2g1BJ8X63PLeQDKuIazLCioycYcF2nTLhHe86gTZSPa51LUOpCK7NF+C47puccTCRZ0YipODwti5855WZlu6bh4gqoSemIRVWAnTJA7c3qahCy8iVtsN8106zimSaKlPcvasbd+/qxr//EY6zUzk8dWEOT1+Yx7WFymLOTM7AP708gX96eQIxTca9u7tx30gvXr2zCxG1Mp6IKNKKIqx/fQHijxsBtZ4xAEhy3nSsaTsMC0UTeb222qTNGP742+dXWFZrMsVAQkNWt/GFF8Zwx/YuUJcQ1BfTQCnBmw4O4tlLC77VM2O8Ig4rR6O0da38YKMcW73h4DnnNIOwIqFg2LA5oMkSDgyFAQATaV3kyxh3x4uM/UPJzhrTQQc3AdpR0Cx3/UiFFTx+bg4xTUZXhGA2Z4JzuPUODtthGE+XMJQM4b337sET5+dwcSYvXMsAqLKEgWRoxbpY7kp9aGsSH33HIXz8kVfw8mimrffD4cCL1xY3DLnDWzsLho2dPRFcmi2gZK2db5REiU9WVSQKy1XHjigUv/2W/XjbHUOBcwmUEnzkzfvwkf8/e3ceJ1dV5o//c+5ee+9b0h2yEBJCEhBMILIpaARkQGfcQEXFnfmNyrgyo8yoA+6jzigqojKOuPEVF0QRGQkikbAnhEA2SDrpJb3WXnc9vz9uVXVVd3V3VXd119LPe14Z7O7q6ntv1Tnn1nPOc55f7kY4aWY3Mwsn3fkZIV0exOETiyVaAxP3B4LA4FNEvGRFY9XFLwgp1mLFWtv8GoKzfR5VBfg1CUMxfc4xUULKodRYPplbX7KY80rFmsu8DMVEF0Y52+FiLwZeSuY7l0mqR1ArLrljpjbX4lMhCshWrJiOaTlYX0JSfq3ZtqYF7zp3Jb507/PZKtm5Jnc9hs0hM4aQNrGxTzRl4tn+CJ45HsHevjB2HwujmDXJHlnE+s4ATusK4dSuIE7tDMKvTd8OBcbwlvTmFyNxE41eGVp6Dd+JSAoJ04bGBYiqhAavDNPmMG0dumVjLG5AENiij3mLWR2n2DWO29a0YNvqlgUfZ849uRX//Mq1+Mp9+2FYjpuIwwFFFCGLApp8c9sEaW9fBIdOxNDoVbL3cIwxtPgV9I27m4E5SPfxnE+s4U3PD0oiy1ZgtBwOWRSya4AdjmnnKyZbHlRLOu56VSvroQrFa0/EUhjKnedi7gb7s70DNMmNBYzHDTcGl/vDSb8ssXSdmPT6a7Cp7zFFFvCqU9vx6OERRFI2dNuBlB6f5pLe4QDQLQeHT0Sx+1gYazv8WNXqx79dvgG3/fWF7GcAryKCg0MRRXgVKdtfDEYN+BQ3lhmv8Otajj50ofrhJXvXOLkT9qsyVrdJSBkOLMfN5hyK6pAEATZ3oJtO9o3s7ngJnNoZxKdfswFr2o7iM799Foy5b3jG3J2KWgMafKqIcMJEwrDxsVevzwbRHYfPmMmYNG0EPDISho2Qh2d/7lcleGUvjo+nsKxBwzvPXYUmv4Imr4Knj43jtodewPGxJCyHZ29iGNyAvsDcRlVoQcJ0pHTlD8bcCXRNFtK7aIpIGQ4ShgXT5vjo9nXTTpw4DofDOZp8Co6NJbGswW2MQ9EUbO7uKGY7bmJHyCvDcTj6xnXoloVmvwKRMbQFVLT43d1Ax1MmUjkTIQnDxm+e7sNvnu7DqZ0BvGZTFy48pRWaPHEDnTJtd0dIZsCriO4uoIpY1K6hkxUawDMYY9lS83v7IkU3WgpwzI0oFvf6Ffs4Qggppw1dQWzubsTjL44hZdl5wQWBAZokYnN3Y1kqMy3E2LQYQp7iAtEhj4i9fRGMxAx0hjSMJ0333oy790iqLCDkkTESM2Y9x3IHrDVZhCaLaPYpSBhuokfCsME5R6NXwWs2deI1mzoRSZrYeXgED+4fxmNHRmHmBK7iho0/7TuBP+07AU0WsHVlMy5Y24KtK5vhUUSs7fCj2aci5K3Oyhc0HhMyM5ooq25zHa8Ljb0cbhyBMQZFdHctaQ2o8ClSdgeS3CDUaMKA5QBtARXtITUbj5AE97N3wihu1+Pp7gP8qoSeZh8GwkmkLAciY24FToHB4cCTveNzvm6ZheuCwLC+I4C9/RFIjMEsVN83TWIMbB7veyknuVK3bMRSFuK6DcuZOjMmMIb1nUGs7wzi3eetQu9oAn89NIK/HhzGs32RvHhITLey47AsMpzR04hz1zRj2+oWDEcNjCdMtAYURFMTSSWl3n/UunLda3LuLgwdT5h5m5lMNhLTceNvnsWxsWTe9xs9Mpr9CgTGENAYekfiODgYx0tXNqLZPzHJ0d3khV+VYNlOdsc2DrdNuwtUnWwJ6GKD1oKAdNCZg4GjLaDiRNQouKuRIrptTBbdmKJp8by/k948rOi/HdAkqJKAj2xfh5aAmh1HAKTvkXWMJUw0+GS0+FQaYwhZAsoxoTm56gcHRzRlocWvosmnQJMlDEXdxQmcu0lpAmN49/mr8ZazV+Dac1dhb18Ew3EdX//TARwbS8Cn5H/GL1SVetuaFtx50jZs/9qDODQcL+t1GRxPzf6gRZI7duqmO5GvSu7OeHPdEW8mnLuvIQMDY+4CAtN2sKIzWFJiR8a2NS24+XWbJt4jjKPBq8DmHCJzN6ywHQ6PLKItOJHUQ5XISb1YrFhr7ufRpGnn3R8yuAvRVrcFcCKSmndMlJD5KiWWT1xz6UsWc16p2HOY67wMxUTLr9ztcDEXAy8V5ZrLJNWhXG1OEtx45HSfBGWR4ZOXrsdrz1hWt32k43A8eGAYAU1CQJMQS1kIJ93PlWJ6LmU0ZsDT6CYvRpIGljV6cWg4it/u7sPevjBenKEaRy5JYNlNuD74irW4cF3rlE2gJxMFBo8sQlNEaJKIVa1+tAe17GfiSMqCbdtIGDYchyNl2hiIpDCWENEaUNHT5EF/OIXuJi8+dNHJaPbXd4x2ujWO6zr8ePVpnTAdjj3HwtjQFVyUvu7d56/G+s4gvvzH/Tg6EocDwCO5FTvmuuZyNGHAtDmUSVXfm3wKxuJGdgMPy+HZqhxieu6Nwa2w7VZuYGCMw3YcyKIIw7YBDvgUEfECFcsn82jVuT5ksdXCeqjp4rXtQQ2RhIFMUd8ZpsagSgKafAriuglBECAxuPOqNs/OKU2eW8p+nb4ssuT+D5G5iXJiOuHUcTgee3EsWw1GZIBl8zkldmSYNsdY0sItOw7iI9tPwbIGD3qavfjC6zahP5xCzLDQ5FUQThoFK/ls39CBb/35YFW/rpW2ZJM7CnXCDAweRQQgwnHcyQxNEmA5Alp8Aqx04N20HDR4ZXzykvUQBIbuJi+CmoSgR4bDeXYhRuZNF9RkpCwHTX4lO3DPnsko4eqtPfjxI0cL/rzBK+NfLjs1OwA9fHAY33/oBcR1CyuaPegdTWazXXOTOSY3yEzyRiGyyMDTPYowafelzLVSJQEnYjrGk2bB58idJIqnFzzuPxFDyCNDN92KHbbjPn9bUIMiinAEjqBHwmDY3QU7mB6oBMYQ9MgIemQMp/8mn5TJ9mx/FM/2P49vPXAIrzy1Ha/Z1ImVLb7szznniKd31xYYg1cVEVDl9OtenOkG8IxMqfnRhJt1l7sD2kxBi5kCHMU+x1ITSRZXzr7YxxFCSDllx/rRPYikdzrM3G1btoOgZ267BBRS6thULUYT0+/WPPlxmXNsC6ho9CpImfkLYDkHTsT0os5xIQLWLL1jtE+VYDscccNCLGUhZbpjUNAjY/uGDmzf0IG4buFvh0fx4IEh7HphFHrO7uMp08GO/UPYsX8IiiTgpSsacd7aVly2sbNqkztoPCZkdjRRVr3mOl4XGntThgPdstNBW8CBu4gv83lzchBqcoXLTDwi99iK2fV4tvsAgTGAA1b6871bgcrdWGG0xDWVkuCOeQ7nEBiDyIB9A1GMJ014FQkx3UShKtEiA7yqNG3soFSqJEL1i2j2A0nDRlQ3kdDtaRMGupu8eFOTF296aTdG4wZ2HhrBXw8N4/EjY3kJl6bNseuFUex6YRT/ed8BdDd5ENUttPpVhBplGJa7+cVc7j9qWTnuNeO6hdG4ATNnmzqHcxwcjCOcMhDSFKxp92HnoRF86d7nEcmp6iEKDB1BN1EqQxEZopyDM56X2AG41WN9igiv6t4z9odTbsXY9IQKBAbm8MIB8AIavTJ8qoSEbkEUHSQNB/0RHazAL8jpxA6BMTT6ZAyEUwUD7qWs642nLMQZ8IO/voAbLl2fN57Q2ELI0jTfCc1Cu8iNJQwY6V0oFUmAX5WyGzy5lbcZoikL3U1eAPn3t6oolFSVWhAY3nHuSnzmt3sL7tg5Vx1VtJth7tgZNyw3QUZwd250OC9pHChGZtdHUXAnie30pO/2De1zjjsVmq9Y3xHAvoEoHjo4hP/ZeQSG5WR3oqRK5KSeLFastZjPo9s3tOP7D71YlpgoIfNRSiyfuObSlyzmvFIp5zDXPohiouVF7bD6lXMuk1ReOdrceNKEJrtrEB1k194CSG+eBMCrSGgJqHX9GSoTx2jyqVAkAYooQlMkDIZTsB3382vKtNEfTkG3HNgOx3hfBHv7IrM+tyYJ0GQRHlmERxbczX04x0jCQNArFUzsYIylHy9CUwSo0tQ1g7mfiR86OIzbHjoMh1vZigycu8d8fCyJZY0eNPoUnIik0OxXl8TYNzlm0DuawB+eGcC3/nwQpu1u9L26zb9oG1qfe3JrWSuFTJ6/y2Bg6Ah50Dsah83dZA9JyFTA5ekED5bdKMX9Hff/Bz0SRuMObF5oBW9hIzF9Tsdfb2phPdR08VoGhq5GH46MxNPJP4AoADlLhCAyNwmkya+AgWXnoz66fR0GIin8z84XkTDc9c6O4/6+W1E3U6TAnZeS0v+1bCcbe8yEXm2HpxM6HKQsG4oogDOk34/Fy1SnyfX0sTDe/oNHcfmmTrz1nBVo9Cpo9MnoatTQ6FWgyWJe+2zwuOue/nJgCAnDzn49WTW8rpW2ZJM7puuEM3TbgU8R8YGXr8G9ewdw6EQMNgcUUcD6zvxSaE1eBYokQBQYfPLUS5pbejxXMdUaNnSFZq3mUCjzq8lnYyAy0cFbjoNC3ZvgzmkXnExwHGRvLtuCasHygNOdGzB1kqjRq2A8aeJENIXRdMkgSQA0WUJrYOL5BcbQ6JExFE0hrlsIaHLeDW7csDCWMCEKQEdIg2E5GE9aMHIWCMR0C3c9eRx3PXkcG5cF8ZpNXbhgbSsUaeIqOJwjlnIXXkqCAJ/qVvTIrfhRSDHvncw1mbwD2mw3L4UCHKU8x1JLAlGK3AG82McRQki5TR7rzfRYfkpHqKwfZEsZm6rJNJ+9Cj4u7xzlqQtgU5ZdNecoCgxBTUZQk2Hajnu/oVvZxYw+VcJF69tw0fo2JE0bj74wigcPDONvh0eQyNmhwbAcd5fxQyP48r3P49yTW3DDpeuxtj1QqVMriMZjQkitm8t4XWjstRwnu3AvE7iVhInBbnIQakNXcMaKnsXuejzdfUBMt9zKnra7K5iYXtxuWA6OjyXRFlDhUwTEjVlqwudwuPs8HllEs09FwrSznz8lkU274wxPBxUXYpz2KO59geNLJ1fqFpIz7HjU5FNw2aZOXLapE0nDxqMvjuKvh0bwt8MjiOYkFHAAR0fdyhFHx5KQRYaAKqPBK2c39Kim+4+FNJ97Td2yMRo3prwmTx4dwx27etE7EofpcLdsNGMYjuUHaT2yiK6QNmUSzrQ5VFHA8gbvlL+Z27Z8irsQhjGW3iDEge0AsiRkK7Fk5CZg5MbKHA4kDQu67SBp2NmNRia/3TOHqMnujm08Z/Gu7J5gtp/I/V05Hcgv1HwENhGQPzaWxA137cFNr91IFV4JWeLmM6E53S5yXkVy+5t0RXOfKuZthpU0bchi4b5+unmO5Y0ebN/QgYDmJpUKAsuLdQuslDpGswtW0XicO3ZKggDG3PshYQ6VxGfi3t8JMNOTttx2E1EVSYRXEXHumtZ5PX+h+YqNy0PYuDyEzcsbqBI5qVuLGWud7fNoQJPxo51HaiImSupbKbF84pprX7JY80rFqKV5maWA2mH1ozZTX8rR5jKb0PhVCeGkmVcdU5NEhDwyOOd1/56YHMewbAe2zaHKIhKGlV0oHNNn3igw5JGxoSuI9oCGPz8/iIAmw1NgjZ9hc8iMIaRNXFc1mwAi5m3YPRNBYNjQFcQX730OhuWkN7xi6aqZABPdne+Hojp6mjwzLkKux/V8mZjBwweH8b2/HM7bRMSwHezrjy5qPLucSaUzzd/5FHd9KWMMssCQtBwAHIokoiOkYihqIGXaYKK7sJ/DbfN+RULScOBwB4lZ3usZKbP4+bt6VgvroWaK1/pVCd1NXhwbS4CBZRMqGABVFtAZ8uSty87Edpv8Cl77kmXYvNxdP/5sXwSRlAnHcROJhPR40uRVMBI3wDlg2oUrRTkcOBHV3bkt7l6zuSzf4RyQGBDyyhhPmHnJI796qg/37h3ERevasGVlI9oCHqxp98GrSGjwytn+4st/fB6HTsSQNG1EUyZetBy0h7Qpa9Or4XWttCWb3FHsIoqrtvTgqi09Mw6w81mQMVs5ymLKVRbK/PKrMkTBgMM5eCZTK10GKl2tBwzu5LYkMpjWRMOWBDexQ5UYgh4FLX4FAxEjWy6wmHOzLAdfvPd5jMYNtPpVqLIABoYmn4IGr4SjI0mkTBvtQRUNPjfrDAA4OFKGg4RhwafI8KoiRuMGgh4ZEmPQbRuDEbe8entQg1eR4FWABq+CpGlhMGrkZZ8BwJ7jEew5HsE3/3wQr9rQjtO6GuBRhOyukAJzJ9fDSQfhpAlZzOxMJuUlg5T6eoeTBv71V8/M6+al0C5q0z1HqYkk9eDkNh8GorNn553c5pv1MYQQslAWo/R0uRaHLrb2gIrnB2ffMrw9oNbsOcqigEafgkafgpRpI5pKZ9SnP7B5ZBHnr23F+WtbYVgOHj8yhgcPDOGvB0cQ0ycWmFoOxwPPD+Hm122s1KlMi8ZjQkg9KHW8LjQuZRbuOQ6HzTkUUYRp24ABaLIwJQg1e0XP4nY9LnQsHO6Egu04AEO2woDpuKvKOQf6wymospAtyVvM0soGj4wmnwpNFpCyHMi2k93FWbdsTBdmduAu8l/fsXAJioLAENBkBDQZlu0gpruJHpMX8OfyKBPjsGU72H08jIcPjuChg8M4Ec3fEcm03YmZ0YSBJp+CrpBWtfcf5TaX+zDb4RhLGIimrGxV2Iwnj47hq/ftR8KwEdRkqI5bCSO3sqwqCWgPagind5J3LA6RCVBld/Ispls4tStY8Npn2taHf/4Ujo+nYHMOO2crIUlgaPTKGIro2fdsprqGwABJFCAwwLQc+DUZV5zehT/uHUAkZbm7KqUj3pld5USBZeMwyxu98CjugujRuPseYgAEMEAAJLDsbkocHLbjVniLpKyJnZOcieMRBQabAx5ZwLIGDYNRA7fsOISzVzXX/OQfIUtNOSfx5zOhOd0ucposQJUEJA0bKdNCynCy1ceK+cydv6PlEO7dO4jBcBLff+gF/Gjni1jd5sf5J7fgx48czca6VUnACyOJ7HPMJ9VDYECTv3om+nLHzvagAlUSkTJtcFb+qh0AhyQwtIc0SIK70CScMrG+s/A4WS6LEe8ipFIWOw45U3tyHF6TMVFSf0qJ5RPXQq4hWSy1Oi9Tr6gdVj9qM/WlHG0u9z2xotkD3eTZai6qzDAYMZbEe6LJq0ASgKOjCaQsO6+a9kx6mrzobvSis0HFxq4GbFvTBFFwK3McHUvg8FDMTdTI2TKagyOaMrGmzY8zehrg0yR4ZLFgBY9iZOIoIY+MpGln47aAu3BfFNy5j2jKmnFz7HpdzzfdJiKaIKIjKGAgotdkPHu2+btGr4LPXXkaQh4Fw3EdX//TARwbS7jVxwPuRk2m5bibfQBQRBHjSRMeRQDn7jpdvYh2sKGrujb9rJRaGF8LxWsz66Atx4HjcLQFNHzkVafg+cEofrarFwGPiAbP1MrMk2O7uZ8NRmI6xhImGnwyjo8lcetfDiOcNAHwghWSMxvOAOnEN5GlK9BzeGQJSdPOm5ubjSYzBDwKDMtBe1DFhae04f7nTmAwXYQgadq4e08/frenHz5VxNr2AK7e2oMzehrxzPEwvvzH55EwbDR6FTR4ZaRMB0nTxvGxBJY1erMJHpNf13pMkCtGRXO2H3zwQVx++eXo6uoCYwy/+tWv8n7OOcenP/1pdHZ2wuPx4OKLL8aBAwfK8rcznbBfFTEQ0ZE0bTgOR9J0K17kLqLIZPZdsLYVG5eHprwxSnmu6Y5ltuef6eeFMr80RXBvThiDIjKIDGgLqPCqOaWimPtPTCd4AO4NSCYRZG1HEF99w+m44dJTSzq3hw8O4x++sxN7jo0jrls4OpbAi8OJ7AJFgQloD7kl5WK6nZ2xiekWXhxO4MWRGPrDKSRNG80+BV0NGnTTxnjKREK3wcDQGlDcATGHR5bQEdTQ5JVx1ZYerGnz5/08krJw5+PH8W+/3Yt/uesZfOzOp/DRO3fjyaNjeY8zbQdjCQPHxhI4NpZAOGHCyqkKUszr/d7zV+E7Dx7O3rxosghBYNBkER1BFTHdxi07DsGZoXOcfAM003NkkkD29UfgUyV3B1ZVyiaBPHxweNq/U8tC3uICE8U+jhBCFspsY3k5nn8+9yKVEvQUt/Ai6FFq9hxzZXZxXtHsRVs6STX3g5oiCThndTM+/up1+OX7z8EX/n4jLtvYiVC6DOEZPQ3oDHkqdfjTovGYEFIvShmvC41LiuQmeBg2h+MAhm3j+HgKR0bjeGE4jhMRHavb/HnBxcxujOs7A0joFk7EdCR0C+s7A0XvZlToWBK6uzCTYyJxw04ndWQ+hXIgO4lS7C7SjT4lu9BzPGFmz2dvfwT6LDsI6aaDvf2zl1EvB0kU0OBVsLzRi2WNHoQ8cl4Vlel+5yU9jfjHV6zBT969Fd9965m45pwV6AppUx6riELN3H+UQ6n3YZGUiWNjCUSS5pTEDodz3LGrFwnDRrNPRtyw0DueygseexQRt7zlJbj0tA4kLRvHxpLoD6fQN57A0VH3fwc0qahrn9noZDKRMeS+7WXBvRdTJAEicxfRcQAnNXtxeCiOlGlny1xbNodp8+zXANyNU2wOw3LAHTeAHUtXgmHp0taG5cC0OSyHw3Q4LMe9tteeuxLrOgJgjOUndogsXS2HoTWgQRAENHhlHDoRw96+xWlLhJDyePjgMK75wS6890eP4SM/fxrv/dFjuOYHu+Ycs81MaI4lpvazmYmvyfccGdPtIsfSfY0oMNgOkDCskj9zCwJDNGXiZ4/24thYAn5Nzsapn+2L4Cv37cdYwsjGun2aBE2eOI755DwIrLp2ccsdOwcjBkIeGWyaSdY5PT9zE3IyCYaSKCCoSZBEAeGUBb9a3Dg57+NY4HgXIZVSiTjkdO2pHmKipD6UEssnroVeQ7IYqA+qLtQOqx+1mfpSjjY3+bMhGNy1bgwYjBhL5j2xoSuINe0BxA1r2sQOgQGnd4dw9dYe3PTa0/DZK05Fa0DFgcEI7t93At/88wF84pfP4MmjYxAYw1VbuuFVRAzHDKQst1qyYTsYiZsIahI+fPHa7E7wc03sACbiKAFVgioJsByeF4vJbPgVSRaOxdT7er7pNhEB3FhTLcezZ5u/O/dk9x7t5ae04WPbT4FfldA7lsBAWIftcNjc3cjJdgBZdKudtgVUOBzwa8Xtx09juqsWxtfJ8drMOugjo3H0jiZwfDyJZHoj/E+8eh02dYeQMqf2h9PFdjOfDS5c14bXvmQZXn5KG95y9grc/NqNOK0rlNf+WN7z5T9/Zo6Kc4Ymn5wXm52Nu1m+ApExrOsI4KPb1+Fd563C7e/Ygis2dyL38nO41Zie6h3H5363D4+9OILbdx5BNGWlE2EEiIKAjvRmNabNMRBOwbadKa/r3w6PlDW2XksqmtwRj8exefNmfPOb3yz48y9+8Yv4xje+gW9/+9t45JFH4PP5sH37dqRSqbL8/XIsoliI5ypVbuZXhpsAoUJgzN1JkDF4FMkt6ZZ+DOduh2A7DmyHQxKAZp+bNLFxeQPufO852LamZdZzO3tVM/YcC2PH/iH879+O4JN37cHhoRiQThoRGEPKtHF8LJlN8FBFEZosQpXcBRGjcQPHRhNIGhYcxy1bF9BEHBtLIpww8IGXr8GXX78Z/3jRyQh5ZDT71IIdsiIy2AA2d4fwnbe8BN+86gxccloHlEl1hCyHI6LbeOroOG78zV7c9+xAwWtrWA5G4jqOjibQH04imi5tNNs1CXmUed+8FHsDtOd4uOgkkHozEp99l/BSHkcIIbWskvcic1VqP16L51gIYwx+VUJHSENPkxfNPnVKtTBJFPDSk5rwz69aizvfdw5ufetZ+MirTqnQEc+MxmNCyFI1eVwaihmQxUxVSneRYWbdZMKwkTAsnH9yy5TPstvWtOD2d2zBd956Fr78+s34zlvPwu3v2DKvmMRI3Mgmc8xESEcIZsl7yHJ44WDpU0fHYXM3rjD5o7rA3AqhNud46uh40edULqokotmvoqfZi64GDwKaPOtkDmMMa9r8uGbbSfjfd23Fv1y6DssbPdnXF+A1d/8xX8Xch6VMG8fGEhiOupMXhRwcjKN3JA6vIqJvXMdwLP/+IKBJ8MsCdh8L47e7+6CIbsJFZjMSw3JgOg6u2tKNgCZjx/4h7DkWzot5ZDbLsB2OtW1+dDZkdjJ3kzg43KoimcQPSWBwMuFuDtjcTcJQJAH/cOZyHB6KQRQE2OmED2AiaYTD3e3Ir7mThAnDzl6bla1+BDURjLFsMsjkq8I5h1+T8bkrN+ItW3vQHlTTlXbdJ9ckAcsaJ0pyq6IA03GryBBCasNCTOLPZ0Kz0FxChl+V0BJQIYsCTJuX/Jl7ps2KGjwyDMuBYbnlizI71zV4ZIhlmKXi80oNWRi5YyfnHJI4tcrKXDEAzT4VmixBlQV4FRFDMaMmYySEVKtqikNW07GQpYtiwHNTD+23Hs6hXlA7rA3UZupHudocvScm4hiaPPG5WBIYvIoIjyyg1a/g86/biK++4XRce+5KqJKAb+84nK7MIcKXjo3uH4jiK398Hk8eHcNZJzXhhkvXY11HAKblbhhtWA42dAVx8+s25V1Xx+HZdY2TY8mzycRRTIe7m2Iw9387nKfXXLoRCV+BTRZK2dS5Vk23iUhGrcezi52/27amBVdv7YFuOdAtG4wxSAKDKgnwpt/D2zd0YCRmQJVEjCfMov5+Zp0tqf6+NDde2zuWQG96HTQAgLlV2h0H+NdfPYO/HR6ZNrbbH05BFhm2rXYrdczWP2xb04L/eecWXLVlRbY6/Uyx1sxcm83dTcg6QtqssVkBQFCV8JHtp+DLb3DbwY/euRUrW/x4pi+Cw8NxHBtLIahJaPTKecklDgfGEiY+/etncXAwgqDmbmxr2Q4My4FHFrG8yQtNFmFYNvoiqbzXFUBdJ8jNprg0sAVyySWX4JJLLin4M845vva1r+Ff//VfccUVVwAA/ud//gft7e341a9+hTe96U1lOYZylrSsVHnM6UoP+VUJXQ0ajo8nwcAQSZlQRAEbuoKI6RaOjiahW24HwAA4cAddRRJw2cYOSDmLDKc7t0xm1KETMRiWg0jKAgdHi19F0rQBuMkdTHR3NhyK6vCpInTbgU8R8YGXr8EfnhnAoy+OwuIcQnZynCGctMAYEDcs/OzRo/j1deeiyatASWfCarIILvB0tqM7Q27YHDJjCGluQsT6ziBO6QigP5zC8wMR2E7+7liZLLGbf/88/vjsIC7f3IVtq5ohFei1koaNpGFjmBnwqSLO6GnE7e/YUvD13rF/aNabl/AsNy/F3ACFHY4ne8eLTiTZuDw07d+rRdyZeWfYUh9HCCG1rlpKdRdrLv14rZ3jbESBIeSVEfK6C11iuoVYyoKVc86iwLBlZRNCXrmCRzo9Go8JIUvZ5FK4X7v/AA4PRWE77i5RtuMu0vYqIiSR4cEDw7j23FXTVuws17E8dmQUN92zD7bNIYmAaRf+HctJfwbnEwvVZzIaN+FTRKzvDOSVDOfZ02EFdhBHdqU6r/BwrcnuRhOcK0gYNmK6hYRhTznmyS5a346Xr2vDkZEEbIejM+Sp6fuPuZruPszhHCciqaImGsIpA3HDRjJuIDcmLQoMnUEVmixiJG7g7t19SBg2OtOVUyzbXbgrMoahmI5v7TiMHz58pGAZ+9zNMgRBQLNPhSqJGIrq0C0bcNwqGitbfIjqFpKGA9N2YNkOHLgxSVUW8M+vXIueZh9MmyOee25s4r8snbAR1y20+BR87NXr0eRX0ORVsL4jgLf/8FHsPJwfXGaZp0hX1P3qH/fDq4iI6zbihum2Qw73ySe1ysnluAkh1W3yJH4mdqsJIjqC7qZHt+w4hLNXNZc8pmQmNG/ZcQiHTsQQdjhkgU0Zoyebbi4BcPs/w+LYsrIRH92+DuNJs6TP3DNtVmRznk52szEaNxBJWtAtO52IWsxdyMxsBxiJ6fN6joWQGTv3HA/jn37yJI6MJsryvIy591ybu0N47/mrEPIodREjIaTaVFMcspqOhSxNFAOeu3pov/VwDvWA2mHtoDZTH8rZ5ug94V6Df37lWvzssV6MRHU4DJAZQ3ezD1dt6cYZPY0A8is/u5U5dBi2W5mDMSBpWvj+X1/Aa89YjhXNPlxyWueM1/Xhg8PZ2EmhWPJs8uMoKpY1ejAUTUFPVwtxOEfQI+OL/7BpyvOVUtWiVtfz5W4ioglTN7Woh3h2MfN3jsPx4IFh+BQRy0Ke9GZoQrYqwkBEx88e7YVpO+6cVBF/lzFAKFiXfOmq9r5025oWfO7K0/CPP3kSDucQ0u3eI4toDWjwKWI2Hnz7O7ZMie06jgM7vcnZ9x96ET/aeaSo/koQGP7hzOW4e/dxJA0Hhj3NpHAOBiCSNNHgkeGRJSQMC+BuRXnucFg5b1IOty1LAsMFa1vx8MFhvOP2R7P9qsMdRFMWGrwKGjwyGj0yRuIGwqmJebWU5SBlAaato9WvQJVFcM5hpddG9zR5cCKq49pzV+KCtW3ZqiXX/GDXgsTWa0VFkztm8sILL2BgYAAXX3xx9nuhUAhbt27Fzp07p03u0HUduj4RxI9EZi/rVI5FFAvxXKX8zfdfsBo33LUHAxEdDV4ZqihAtx3EdBvLGjx413mr0N3kzXZqAPDZ3z2LH//NnQgX0hPKsihCkRh+/MhRbOgK5XUMk88ts+tYTLfQ6FWgyRzjSbe00HBUhyi6OxPKgntTIgqAbtlI6jbCKTfD6qotPdi4LIRrb38UnHNEkhYcxiEJLDulY9kOnu2P4o5dR3HVlp4pk0+SyCByDstxO4pVrT6safdlj/PgYBx9Ywm0BTQoIkPKchBOmoimrLzB8vEj43j8yDiafAouOa0Dl23qREdQm3K9OeeIpdyFl5IgoKtBw9oOP1Rp4ialHDcvxT4H45h3Ikm1KbYdG0VmLxf7OEJI+cxlPCblUYl7kbmaaz9eS+dYCkUS0CQpaPIpSBo2orqJhG7DmW3b9QVC4zEh9YHG5IWXGZf2HAvjRCSF9qAHqiwgZTiwHMcN4CoCUqaz4IH6zLE4nIOBgTGezrwo3Adnlo5LDJgt1McAvOml3XjNpq4pwdIzuhuyO1ZlHpv5UM/hVs+UBYYzuhvmeYblwRiDT5XgUyXYDkfcsBBNWdCny4KBW4llXWcQyxo8i3ikrmpqx7n3YZxzhJMmxhNmUfcrSdPG3U/3I27kX2e/KqI9oEEU3JgNAIzEDAQ1Ob3DlQBNdt9vMd1C0rRhOxwBVUKjV4FhO9ldem567UaYDp8SJ3HLRYtIpQPb0ZSFf73sVCiSgFt2HMLBwSiSlgMBQE+zDx951Vqce3Ir9hwLAwBM28nGqTITiblMi6M95MEVp3fltY1Xn9aBnYdHIIBDFFk6aQOwHTewzwFEUiZkkSFpWtlqO5mWlLI4jo8lsazRA5/i7ui1vjOQV46b1IZqasdk8Sz0JP5cJjRnmksYT5jwqyI+cOEabJ7DmD3TZkWSIECAm4RxIuK2BVFgYILbJ84+5Ti7Z45H8Ir17WV4psLm2o4Fwd0AK25YkAQGa56fjRkAjyzgo9vXTRl3CCEzq/Qc8nxV07GQpadaYsC1el9dD+23Hs6h1lVLOyyHWm3LpaA2U/vK3ebq7T0xl3Z87Xmr8OYtPXhw/zDCKQMhTcGadl92ATQwUflZERlORPT0InnmxmMZg2U7eG4ghjt2HcVbzl4x43WdvK5REYUpseTZEjwKxVF6Gr2I6BYiSQs+VcSX/mETzj25dcrvFrupcy2t55tstk1Elko8OxMDbPKpeRVqMhq8MgYiKVgOh27ZbhUHe+a+Q2IMp/c0LNARu2pxPK72vjTkUeCRBAQaPBAFlk3yybSN3Hhwbmz3oYPD+J+dL8K0nTn1Vxu6gji1K4Rn+yLo9GgYSxoYjU+tEJOeooIkMuiWu9Y6oIpoD6o4OupucJcZ1hhz34cOd6sV3frgISQNCz9+5GhevzqWMDDqmBiN6VDS1WragxoavQ6GYnrenGDCtHFkLImAJqHFp0AWBXDOkbIcyCLDS09qwmnLgmCMYc+xcN0nyM2mDAWvF8bAwAAAoL09Pxjf3t6e/VkhN998M0KhUPZfd3f3gh5ntZit9NBbzl6BC9a2YuPyUDbYf/BEDEGPjFUtPnQ3eXFSsx+r23zobvTOWvqrUOmwzES+JLLsshGRIVuODOBwHI7hWH5p+PGkCXAgZbo7JMqC4Fb7YO6khyQy2A7H9x96AQ/uH8L2DR3wKUJeWaKU5WA4ZiKoiXjf+avyEi3CKQOm42bfMsbgkUV0BDWsavGh1e8mUOQajRv48SNHcfWtj+CTv9yDvx4chp2+Dg7n2D8Qw6MvjmL/QAyGbSOcNHF8LIljYwmEEyZsh2dvXsYSZoFdS92bl9Vt/hlvXop9jtN7GrJJIIXUYhZsse3YpxTXhRX7OEJI+SzV8ZiUhvrx6XkUEW0BDT1NXrQGVCjS4l8DGo8JqQ80Ji+e3EA9A4NHERHQZHgUEQxsUctPjydNeGQRYhELCAXmlsXNfRSb9N+MrSub8uIKGRs6g1DkiX4+swA+9zkVWcCGzuoL4IsCQ1CTsazBg+4mLxq9bjCxmlRjO47pFo6NJTEaN/ISOybHTTI/2z8YxXt/9Dj+klMimQFoD6joDLqJHRwc0ZSJFr8bv1BEAbIoQEy/3zjnGIqm0slL7mtXqIx9g0cuGCfJtEtFEuGRRTT71WxZ9e++7aX4+pvOwO3v3Ipfvn9bdjJuQ1cQbUEVDgckYaJNcJ7/HmcM2L6hY0rbWNbogUcSIUtugDpzPTyyAEmcyICK6hZsDsiikJ3ssx0OkQE2dzAQTqE/nMqLp5HaUo3tmCy8Yibx53tvkJnQnBz7n8lscwnF7FxZSO5mRZOpsttnO3D7N0lk2cUb1iyT2UVb4OF7Pu14NGEAHFAlYd6HyeHudtfgk2k8IKRENB4TMnfVEgOmdkyWsmpph+VAbZnUgnpqcwthru1YEBjWdvjx0pOasKbdh4OD8WwsWRYFWNyB7XBEUxYcuDFiURAgpNcSSiKD43D8ZNfRadc0AoXXNRaKJc/0HBmT4yhDcQOO41bT/MrrNxdM7ABmjpMAtbmeb7JM8otfFfPWcSZNGwMRvSzxbMfh2HMsjB37h7DnWLio12yxFRMDBICQJsN2igthqYswp0bjcfmNJgxYDhDU5Ik54pzEhMnxYEFg2NAVxMOHhmHazpz7q0xbDGgSwikLmiQWfJ8JDJAld97NcTgiSRNr2gP497/bgK6Qx12zDUARGWTRjeWKgoBlDR7EDRvffOAQoikz7zi9igRJYHA4MBzVkTBtpEwHssSwrMGD9oAKeVIfEE1ZeHEkgRNRPb2hv4nuJh+a/QqOjSURSZkYievZdsXB3Q1yUyaShg0Ovqjz7pVStZU75uqTn/wkrr/++uzXkUhkyXQ8udlcIzEdYwkTDT63o3AcnjdQ5u4aNiVjkGHWzKZCu45JgpDeudCt0mE7HK1+FZGUCd1yslldK1v9+Nj2U7KTRE1eBWCAbjnpbNv8xpypYPfiSBwf/vlT6Ql4BUGPu5NjoZLznHPEdAvhpImQpkAW3CoiqjTx3KLAsucfTZk4uT2Ip3rHYKYnkziAR14YxSMvjKLVr+KMngb0h1MYDCfdZBEhvzycYTkYsXSMxHV4FQlvP+ckfObuvdPugDbbzUsxu6i9/4LV2LgsVHdZsMW2Y8su7uav2McRQspnKY/HpHjUj89OEBgCmlyRv03jMSH1gcbkxVNN5aebvAp8qgi/JmIkpiNlFQ74KaIAu0DZeD7pv4C7eD2ctKY8FgD2DUShSQKSho1CsUWBAZokYN9AtKp3T5FFAY0+BY0+BSnTRky3ENet7IYPlVJN7Thl2hiNG0gVqHLy5NEx3LGrF70j8WzcZHmTF50hDffuHcxLNJJFd/MNVRbdks6WG7z1KiIu29SFOx45Ag6et2tbynSgW477PebGoTJyd+kBUFKcZKadngSB4U0v7caNv9kLmwOC4Map8hKXJAEBVcS5kxZDP3xwGF//0wEkTQuAG++SJbcP8Mgijo4l3Oodglt6OhMTExkgg8Gy0zskccCwbJzSEciLp5HaUk3tmCyearo3mGwuVT9mM91ujTHdwolICnq6OhMHoJtOOsltok915xRyEulK+NsCA7obvHM+9mLMpx03eRUokuBWdzPt0k4uR6aKlO1wjBfYfY8QMjMajwmZu2qJAVM7JktZtbTDcqC2TGpBPbW5hTDfdvzk0TH89NFeHB1JwHI4FJFhTXsA2zd0gAkMRk68NB+DKHCciOgz7tZe7mqqc4mjLJWqFpnkl1t2HMKhE7GC6zjn6uGDw9nnNW13Y+/Vbf55P2+5FRMDVEQBl23sxHf+chhmEZuyqZK44HNqNB6X31ziwXPtrxyH5/VJZ69qzrbF5/qjAANEuGukM/OcosjS1ZU5OACfKmXb07vPX4XP/PZZQHA3BcxUD24NaPCrEmyHo288ia4GT95xarKbLGLa7gb9/eMpMMahSAIavQoM28HG5SFsXh7Cj/52NDtXyOFuVjieNBHUJPzDS5ZBYAym7WA46iZ2iAIwljAQSVnQLRucu/PVqiQi5JFrPkFuNlWb3NHR0QEAGBwcRGdnZ/b7g4ODOP3006f9PVVVoarqQh9e1RIEhmjKxG1/fWHGgW2+pb8K/b4quaWEMuWjOHcnuE9q9iFp2BiKGVjV6sOd7z0HkiRkO5iRmI6QJmM4ZsB9uonGbzlOdkBjAAKaBFUS0R/W4VNEXHn6MnhVEcsavLh8Uyek9K7WjLkLIQOajAaPjJNafDhwIoYWvwKW8/wcbhLI6rYAbn7daYgkTdy7dxB37+7H8fFk9nFDMR1/fHYQAOCRRTR4JMiigMNDMXz1vv24/pVrcUZPY/bxCcPCylYfPnjRWvzssV4cGYkjzFHyzcvkG6DxdDZvR1DDm7b04OxVzUUngdTSLl7FtuP2UHFtvdjHEULKZ6mPx6Q41I9XNxqPCakPNCYvnmoK1OceS3tQQ+9YckrSBQNg84kqlbMRGEPII2HPsfCUyYvRhAHLnqgCMvV3ActBTe2eoskiNFlEs09B0rQRS1nZ67XYqqEdm7aDsbiBmF44wefJo2P46n37kTBsBDUZQZEhadrYfSyMJ46OZx/HALxpSzdO7w7h548dR+9IHFHOITOGVa1+XLWlG1tXNuOxF0fx3EAMmjyxq5HlOOmKGRweRYI2aVe8TCxrPGlOiZMoIkM0ZSGSNOFTJbz3/FVFxUkePjiMe/cOQGDpKjjpt4CQTi7xKQIMh6Mj6HFLU6c3V3n44DBuuGsPYroFRRJhWDaEdHB7OGag0Sdnz0WWBFg2z6uUk4mrtQVUSOlj/9BFJ08bS5ocxJ/vAm1SftXQjsniq6Z7g4yF7i+2b+jA/oEojo0l0exXYDkcx8eSbjUiYWIjJw7AtDkyf1oWGcABO93RioJbcQmMw7aBwntcTvAoIi7f1DnLo+ZnPu048154ujc8r2PI3IkIjKHRW5mNIAipZTQeEzJ31RIDpnZMlrJqaYflQG2Z1IJ6anMLYa7tWBQYDg/F8PX7DyCStOBVRPhUCYIA7OuP4uhoAn5VwkjcgLssOn+Nn+1wqJL7k8nzDbkxjxeH4tOui+TpOG7csPH40bGiYyMzbRI03ePrbT3fdBZiE5HcGHujV4EiCjBsB/v6o7jhrj3zqj5bbsXGAK9/5Vr85eAQ9vZFpn0uAUB3kwcJw1nwOTUaj8tvLvHguazjfujAEL78x/04OhKHw90kjDXt7prk29+xBXuOh/Evd+3BsbEkljVoSJgOhqLu5jtWusq8T5HwlrNXZAsHdDd5EdQkBD0yHM4hCul4LXerZjD3y7xN2QAgbth5lZkZc+fJdNNB33gKTX4FV2/twRk9jTi1K4j//vMh9I4m8uaTIykLX75vP6455yRccloHRIFhZYt7PAdOxMEASCKDKLpzZ0nDQsKwsKErWFJsvdbm0qo2uWPlypXo6OjA/fffn03miEQieOSRR/D+97+/7H+v1l646RQ7sM1317DJvx/TLQxFUzBsd5dOJ91gU5YNSRQQTllo8sn42PZTIEnClMzKTAky0+KQJQ7G3NckU0VDEtyEDUV0F1f4VQfHx5P43kMvIKhJUCQBv3zyWMGkCb8m40MXr8UnfrkbwzETAU2CIrpZvpldIa/a0g2BMTR4Fbzxpd14/VnL8dTRcfx2dz8eOjict0tn0rSRNG1IgruwJaZbuGNXLzZ3N0zpvE7vacCm7hAODsYR1U10hjw4s6cBHrX4ppe5Abpj11H8ZNdRnIjoGIzo+NafD+LevQPZc16oLNhqtqLZV9bHEUIIWVzUj8+uFu5R6XUkhBDXXAL1C9XP5x7LeMJMl9HlyImtgcMNJHJe3ObRHBzf+L8DOBHRp2wk0eCR0+W2J3aUzmBwF3ImDRsNntpbhMiYW1LYq0jgFUruqCQnnSwRTprTnr/DOe7Y1YuEYWc31YimLAxGU3nB2Ra/gk9esi67OcaZK5pwcDCOcMpASFOwpt2HoCajNaDiAxeumdKWLMcNOAuMoTWggoGBc46U6cByHFgOh8TcmNXG5aFsnOTZvggiKTObeCGJDN958DAExmaMl+TG2NqDKk5EUrAdd4GxwwHDdmAk3XjW84NRvPOHj+LUriDee/4qfOfBw4jpFjqCGuKGjeNjSTgOh8gAm7t9grvQmaHJp2AoqoMjZ7f69O5DXkUCGOCROZr9hSdZamX3MkKWomqbxF/I/iL3uVOWg5Rpo3c0CZu7iXleRUTIK2MoqiNd/BuWzSHnVBHLbPLE4SaFllLeoi2gVt1n5VyCwPDe81fhXf/zGByO9Hgw9+fzqdK04wIhpDbVQgyQLG0UAyak8srdDmnsIWRmNPYtDJExfP3+A+gPp8C5uyFzZhf2Fr+CmG4j5JHA4MYNJNFdQ8i5u9N8Zm0f5zxvTePkmAcHRzRlQZEENPkmHpdZ35gyHTic45v/dxD37xtcsFjqUlrPV2ryy0wch+OWHYeyMfbMAnlNENERFDAQ0XHLjkPZTbGLeb6FHPOKjQFKkoC/29w1Y3KHJLqbRNV7RYJ6NZd4cKnruG998BC+ct9+GJYDxtxKHIYl4OnecN768BsuXY8b7tqDwaiBBq+MnkYvIrqF0ZgBizuQRYYf/PVF/O/fjmB1mx/bN3RAkdwqHNwBBiN6XrWMzLro3A0DOecYiqbA4a7vdpeAM3e+mAGSwNAZVHHB2jZEdQtnrmjCbdc04tnjUfxhbz8e2D+EhGEDAEZiBr56337c+XgvLjmtEyc1eZEyHXfuLH25eLriyJTJ6CLU4lxaRZM7YrEYDh48mP36hRdewFNPPYWmpib09PTgQx/6ED73uc/h5JNPxsqVK/GpT30KXV1duPLKK8t6HLX4whVSysA2313Dcn/fr7pZVjbnkAQBAhyY6a20TkR0mF6OU7uC2etZKAFFt23EdBO24+4G6XYG7q6FYrozUyURmiIgplvoG0/BcdwbuKBHhiiwGTMzt61pwedftwm37DiEg4NRxHV3Qj2zK2Ru1Q3A7YxesqIRL1nRiEdfGMNn7t6LlOXkJXlYDsdIuvT5M8fG8Zun+vF3p3dOSfAQGMPaDn/26/5ICrIowK9K8GtuBZDZ/O3wCL73l8OzJu2UOwu22nU2aGV9HCGEkMVF/fjMauUelV5HQgiZUEqgfqH7+cyxfOuBg3jk8Cis9MfZzOdsxpDdUGE2DG5A7sXhBFoD6pTPpNeeu9Ld7arA73JMJJLUuqkl4OtbNGViLG7CcmbeL/3gYBy9I3EENdkN+MZSiKTyK3yoIsMnX70eZ/Q0ZL83OV7S4FWyE22F2pKUjgFxzuFTJjYa0S134bDDOYIeGeGkkX0Oh3N89M7d8MhuieaAKsF0+Ky7exWKsSmSiP5wEikz/3pIgvv+DidN7D42jo/euRuGZaPJp4IxBr8qYVmjB0NRNxAO7sa+fJoEkQGNHhmRpImk6UAWADB3olKTRagyw2DEmDZGV0u7lxGyVFXLJP5C9heTn7vRq0C3bQyGdcQNB80+Be1B9/Ngtr9jDLLI3MQ3QYBuzVabY3rNPhm66WBvX6RsiwgWQsijwKdIYLBh2nbJE5AZAgM2LgstasUXQsjCqpUYIFnaKAZMSOWVsx3S2EPI7OJG4QrGk9HYV5o7dh3Fs/1RgHNIopBN3EiZNvrGU2gNqEgaNla1+HBkNJHeEMKdz9DkiQSQ3HhpwXWIlo1w0kR/OAlZZAhoMmK65VYX5Q7A4caMvdKCx1KX4nq++drbF8GhEzE0epUp8zKMMTR4ZRw6ESsqFrRYY14xMUDH4XjwwDACqoRIyioYGjJsjoFwquSKBKR6lBoPLmUd90MHhvCV+/ZDNx3IEoMABg43AcRyHIzGkV0fXug4HMeBAw6PLKLZnz/fe3Qkjma/gqOjCSR0t76yKDAwwZ0zc5NJgPGEiZBHBmMMKdOBbjnpis0MPlVAe1CD7bjryME4RuMm+sMprOsMYCRmIGFYOG15EKctD+I956/CHbuO4q4nj2fnrI+OJvGdBw9DZO5Ga7LIsoUCMokmmiQg6JExEjOK7gdqcS6toskdjz32GF7+8pdnv77++usBANdccw1++MMf4mMf+xji8Tje8573YHx8HOeeey7+8Ic/QNPKd2NUqy9cIaUObHPZNSw3k3H7hg4cGY7h+Li7+6Akuo/lYJAFoCWgIq7b6G7y4gfXvBSSJEybgOIRJHQ3enF0NOHuwOhXwAEMxQxw7k7wtAbcnaiGojoc7v4923En732yNGtm5uSbpZAmY1mDu4OiM9MiE8ahSAI6gipSpoPxpIl4OmMsQ7fdHUzvfKIXr9nYie2ndaBxhuxJ03YwljAwljCgyiL8igSfKkIqkOhRajZqNU9gldtTR8eKftybXrpigY+GEEJIqagfn14t3aPS60gIIfmKCdQvVj+/bU0LtpzUhO1ffxCHh+MQ4O4aA156sgUDEPJI0GR315jcz6Q/3XU0L7EjsyM3cqqCOJhaJp1Up5RpYyRuQDft2R8MIJwyYDociuOgL6znJQ0xBrT6FRi2A4sXXrjLGEOzX0FQy6/sUqgthZMG/vVXz+DoaBJJ08pW8uA5//3XXz2Dm167EWevasZ3HjwM03bQ0+TNxlNEEbPGkArF2HyKCJHlbwikiAyiIICDw7LdfwndQsK00R6YiF/6VTfukzIcGLaNaMrCu89biZ8/dgyDUQNBjwzD0mGmK9sKjCHkkTEYMWaM0X3rgUPpQLrkBrWFue9eRghZOJWexC/3bofFPLdHkNDkcxDTLSQMG0nDhs05glq6v3O4O+nHOaYZHmYkpHeLY8xNmojqVtXfZ4wmDAiMYWWLF5Gkhb5wsugKarmCHhkfuHDxKr4QQhZWLcUAydJGMWBCKq9c7ZDGHkJm9/DBYfx0V29Rj6Wxr3iOw925BM6hiCy7eTJjABPdSh3jCQNeRcLL1rQgsncASd1GQHPnJQQBGE9YefHSaeMSioRlDR4cHU3g+HgSPU3uBtWZ6qGiIKAtqMEjS9CC4oLHUpfaer75Gk0YMG0OZZoNs1VRQNjhs8aCFnvMmy0GmJl3aA2o0G0nu5FUoTk1UttKiQcXW+0DAL78R7dihywyiMxtHwyALLhVkU3bwcHBaHZ9eO5xDMd1fP1PB3BsLDFtjDigAbrlwHI4ZIm5b0vuViCWRXezHtNx0B9OocErI65b7mb53K060xrQ3Ir0aY7DEXbcuK0sCugIaUgYFkZiBkzbQdAj430XrMZrz1iGr/xxPx47MnG/m5lqNG1387cmnwJZEiAyAarMwDnDWMLAUEyf8bVYyNj4QqtocseFF14442ICxhg+85nP4DOf+cyC/P1afuEKKXVgKzVLrFAmo1dxF3Uw5u7iyZi7s2BrQE1PXNs4EUlh30AUG5eHZkxA8WsSmv0KxhPu7l0AAM6hSCI6Qhr8qoSkYUO37Gw1D8bSWV7IT2D59VN9aPIrUzrGzM1SJknl2YEoQpqM7iYPYpnOZpKQpkAWGCzHLXfuUyWYtoNIykI4acLK+Z2+8RS++5cX8P2/vojzTm7B5Zu7sHl5KO9cHc5xcDCOcMpASFOwpt0H3bQxEgc8iuheN0WaMrCXIxu13jx1dLysjyOEELK4nu4Nl/Vx9aLW7lFpPCaEkKlmCtQvdj+/byCKpGGj2atgLB03KIVblcD93C+L+eWAM59Jj48nwTmHJLi7qPD0P4bM4sv07jLp6pekOlm2g9G4gZhe3K50GQFFhmE5GEvkv76qJKAzqLmJP9xCSJu6CYbAGNqDGjzK1FLTwNS25Dgc7zx3Jb7w++dg2RwCA8DcybrWgAqfMjER51FEPNcfhSIKSJkONEWAG4qePZ5SKMaWMh0YNs+WRXelnw8MogAYtgOfqiCmW4jqFhpyNv5gYO55moBH5njZmlYEPQp+susoTkR0eGQJCdNKXzsRnPMZd/a/Y9dRPPriKGzHXTzNmPt7mZjcUo4XEVKNKjmJv5Dx5ZmeO7ORUdyw8eJIHCydjCEKDALcRRvuGJGTFIjiJrEdDgiMw+EMKdOGLDA0zbDZUjVo8iqQRQbDcmBzPqfEDgB485YeWmxHSJ2otRggWdoolk9I5ZWjHdLYQ8jsMu0kWeTGNzT2FW9vXwQDkRRElg7q5sjEV5OmDcN28Ls9/TAtDt1ykIrpUEQBXkXCuo4APnDhRLx0prhEQJPREdIwHDUwEjORNG0IDNBkKRtDBWjtXTXKxlBsB5owde5At51ZY0GVGvNmigFm5h0c7lbvltJVbR2OvCBRo7f4igSkepUSDy5mHfeeY2H0jibcudec9ywHB4fbqxq2g6Rp5yU+ZY5jz7EwTkRSM8aIByIpqKIIAW4Sh8U5GAM8soDWgAZRYBiLG2j0yjg6moRpO9m3rligHRVqq15FgqdRRDhpYjxhwnLc+cXxpAGvLAIMSEzaeN/iwEjcQFtQA+BANwVwOBAZoJs2/rJ/CKbD0epXpyTR1PLa64omd1Ra7gsHBiQNG5bjQBIEaIpQ1S9cIXMZ2IrNEpsuk3EwosN2ONqCKjRJzF67zGT55ISS6RJQYrqFoaiOlGnB5m4CRHejBylLQThpwpee5Lccx53rYRyW43YcmjzxXKbtYCim47N374UoCAVLaU1Xbuu956/CxmUN6YSNie3C1rT70N3sw+GhGFr8ChgYZFFAs09Bo1dCf0SHJAiIpkxk8jwsh+PPzw/hz88PobvRg9ds7sL2U9txaCiGO3b1onckDjPdCXc3+3DVlm6c0dOIZHoXs2FmwJtO9BiJ62XJRq1HVpG7uhX7OEIIIYtLnmZsm+vj6kWt3aPSeEwIIaVZ7H5+NGEgbtiIpUyUmNcBYKL/9skiNGXqmKyKQnqBpVs1QZGEieQO5v4zLQeCwNDolaf8Pqm8TFnncNKcubJpAUNRHd/5y6EpFU6bvDKafe57fDhmYFWrH2vafXmPkQR3lx5FKu5eLxPP2dcfyU7EyZKAJp+CJp+SjUU1eGU82xfBR37+NEbiejbQrUoiWvwKJEGA5TgQmBtDKxRPKRRjy8SkciNmuXFgxgDuAJosQBAYwkkrW5o6I3OtO0MqvnTvczg8FIdhOTBtN0jukQQIjEEQGNqCGt57/qqCC3gfPjiM/7r/AIx0rE8QGDh3q64cH0tiWaMHXllcsvEiQki+cu12WMpzx3QL/eFkdnLP5gDjHGI6QU4AhypJaPErGIkbSOgWZMld3GE7PG9Dpem49yjucb/0pCZs6AqWfPyLaUNXEM1+Bc8NRGE7fE6JHbLIsGVlU9mPjRBSGbUWAyRLG8XyCam8crRDGnsImV2mnUgCK7hB8GQ09hVvNGEA3N0UKGU5kAXkb5acXuTOOEeDR4bpcOhhG7rFYdk2JGHqlhCzxTwaPQoMy8GrT+vEb5/uQ7NPgVcVs7HkjKW89q4abegKYnWbH/v6o+gICgVj7Os7AzPGgua7oDuzcXg5q/Bm5h1Spg2eXh9bqJsxLAeCIND7cYmZbR33aMKAk062yMzB2pzDsp28TWRiuo3e0cSU5y8mRmzabqpIe1BFKj1B7JVFeBQRjLnJSJbDEUlZ8CoiPLKM8aQJw3JgWE52fsqvSjO2VbcNKthzLIz//vNB7B+MYjxhQshsYuZXEE1Z2WMA3PhyfziV3ViQMaAjpOF7f3kBvaPuWmxFFLCm3Y/rLlyTnVtbyNj4QlvSyR2ZF86w3VIxumVn3/iqJKLZr8Cs0heukLkObLNlic2UydjqVxDTTYQTFlra1Ck3P5MTSgpNjsd0C8fHkunOh0ECEPLIGIwaEAU3qytTcsgtycZh2W6JtNbAxPHkPo9PldwS75NKaQGYttzWv/7qGdz02o04Z3UzorqFcMKEabuT/Vdt6cZX79uP4ZiBgCZDERkMmyOaMtHgkXH9K9eiq8GDe/b04549AxiJT7xneseSuOWBQ7j1wcOQRQZJYGj0KghKbod4eCiGr963H9e/ci3O6GnMvl5x3UJct6CbDgQGpCw7r2zRdNd4KVne6MHBoXhRjyOEEFJ9NnQFsft4pKjHLSW1do9K4zEhhJRmsfv5Bo+MhGFlEzvmWt5ZFNiUz/yA+5lUkdwYRFy3YNnc3ZE7XfHDst3P+kFNRrNfLcMZkXKK6RZGY0beJhfF2rF/CF+9bz+iqYlKH4wBrX4VQU2Cno6beBURV23pTsd0XKosoj2gZnd1n03upiOKmE6AYO77azhqQJXE7E5rpu1gPGEgZVrpRAl357eEYeHIiAVBmHgnC4wVDHIXirFJguAGzdMthyE/uSPTjjkHgpoMRRIKltCWBOBEVEd/2N0hSRYd9IWT2bazrNEDRRRwbCyZjVXlJnhk4nS65U5oMua2TcYAJrrXZCiqoyOkLtl4ESEkXzl2Oyzlud04fWJKtTAOwLY5RMFNzBC5g3edtxLfuP8gUqYFgLnJoizz6OKkTAc9jZ6q31n4b4dHcCKqw7LnltgBAEFNQouP7qcIqRe1FgMkSxvF8gmpvHK0Qxp7CJldpp3IggAds1fvoLGveE1eBYokQJUFDEUNmI5bDZwBcMBhple5t/k1dxHveAoOdzdttjmHafO8NYDb1rQUFfNQRAFn9jTi//YNzjjPQbHU6iEIDO+/YDVuuGtPwRi7XxXx/gtWzxgLms+C7uk2Dp+uynaxMvMOe46Fs1VdC4nqNmwOej8uQTOt427yKvDIIlKmu2EYZ8hWzshvCRy3/uUwVrX48t6vxfSXnHMkTRu9Y0kAmXtEd422X5Wg2zZSpg0GjuWNXjDGoCkSjo8lYXMHtuPgRCQFsUHDeMKasa0+fHAYn/r1M4jpFjRJgMDcpA3dsmHaDtqCKpKmg9F4fhvlcBM9wIHjYymMJ0w0+1QE05sKPdsXwcf+327cdOVpOP+UtgWNjS+0JZ0+2uRV4HCO42NJpEwbAmOQRHcCIbPTnePwqnzhCskMbH5VxEBER9K04ThugxuI6EUNbIXMlMnoUUSokgjdspHU829qMwklq9v82ZvZzCA1ljDBuZvpNRTV4XB3UsfhgCoLCHlkdARV2A7QFlCxriOAhG4hmrIgCu4OiF0NWnbinoPjRCQF2+HwyCJCHhmCwKDJIjqCKmK6jW89cBDfemAiSUWTxSmPuWXHoewkfHeTF21BdwfJM3oacf0r12JVqx8pw8JIwkDKsLCq1Z9NymgPanjHy1biJ+/ein//uw146UmNeR2n5XAkTQdR3X09wkkTksDQ4leQMGzcsau34A6Zq9t8WN7kxWjcgG7asGwn+7hC13gpufiUtrI+jhBCyOJ6zWmdZX1cvai1e1QajwkhpDSL3c87nMPKXVzJJv23SNGUBWdSAkDmM+m6jgA2LgvBp4pQJSH7Nx3OoUoCfKqIU7uCS/Jza7VKmTb6xpM4EUmVnNiRNGx88Q/P499/+2xeYsfm5SFsWhYCOC8YN8nwqRK6QlrRiR2TNx3xKlK6KozbdhzuxpZ4+v9ORHRwAO1BDZoswHbc58jsXOQ4bgwqE5e69S+H8fDB4by/WSjGpohugodlI7szUGZ1LgeHnd4VKGnaOLUriC/9wyas73TjWSdiOhK6hXUdAbQGVNgOR0dQgyoJGInrcDigSG6jHIkZUGUhL1bl5GzflYnTtfhVqJIAy+HgPJNwwiAKDCnTwnDMWLLxIkJIvskx+VzzjS9PF+/PrUzBMHHbwQHY6SpHPkXCpuUNOKUjAMaYu8scOMCKT33IPO/vnhmAVcXlIjNjmWU72f5+LtZ3hqhfJ6SO1FoMkCxtFMsnpPLK0Q5p7CFkdpl2kjRnT+wAaOwrRSaGoFscXQ0aPHJ6LsHhyISoVVFAk1/OriWURAZRECAJAkzHQYNHzouZFhvzuHxT54LFRsjC2LamBTe9duOUGPv6zsCUDZEKyV3QXch0C7ozG03t64/Ap0poC6jwqVI2sWjyXEIpMvMOAVWcNrEjI2HYOLnFN/ODyJKS6e8UiYGB5yV25MZhlzV4EC8wtzRbf3kiomfbC0+v5RYYQ9J0K3JEUyaGYwY4B1r8anYduV+VsKzRA4/sVvdImjbCiZnb6uR5P5/qrvVmjEFMz/uNxU00+2QoU/MxJo4bgG45kEX3WFVJQItfQVy38PX/O4iB8SRObvPXbP+/pCt3rO8IwObpyVeJZXcQZAyAwGFYHDbnWN8RqOyBliAzsGWyB8MOhywwrO8MFJ09OLms1EhMnzaTkTGGtoCK3rEEhmM6WgU2Y6bk5MxKjywgZVru5I0DiIzlVeNo8MoYiRn43JUbITCG0YSB3tEEbn3wEGK6DUkUoIoCIikTSdOGJDK05VQXyRxjg1fGcwNRMLCSym35VQl+VULSsOFdI2FzdwMODsYRThkIaQrWtPvydp4EAEkUcN7JLTjv5Bb0jSdxz55+/HZ3f96CB8N2MBQz0pVAJHgUEUeHYzg4GMfaDn/e8+VWDhmK6dnKIabDEU1ZCMwxaace3LdvsOjHveVlKxf4aAghhJTqe399sejHvWwJJQbU2j0qjceEEFKaxe7nd/eG876eLWA8mSK55edtBzg6mkR7SJvyuf8DF64BMFEps0lWshUMkqYNvyot2c+t1cayHYwmDMRyYhSl2NcfwU33PIfj48ns9zyyiP/vFWuwfUM7ODBj3CTkKb2Cy+RNRzRZgCoJSJru5IsoMOiWjZTh7iqkWzY0SYRXkdAaYDg2msju+ga4wV7LcSvCdjVo2cnAs1c1571HC8XYvIqQrdxh2m7wXBDcUtSZZJPM+33bmhZsW92SF2NzOMf7//fx7LkkDRu65UAS0n2BwLPn4lHEgrGqzI5jqiSiNaDh+Fgyb5c7N9HE3U2J2h0hBJhpt0MbwzEDqiRg+4aOsjz3RLwfAHf7JVkUshW9nPSuhB0ht/8dT5r4wIWrceBEFENRHWaJCRqKyOAAiCZN/HZ3P177kmVzOo+FlhnLvIqEcNKc03MwAK8+rYP6dULqSK3FAMnSRrF8QiqvHO2Qxh5CZpdtJ0XG0WnsK15uDCGm22gPatk5hGjSRNyw0RHSoJtujDS3ygaDG1ewOZ8SMy2mwoOUjpXOpxIEWXzb1rTg7FXNeTH2DV3Bol6nQtW5MzILutd3BvIWdFuWgy/e+zxG4wZa/SpUWQADgyaI6Ai6lboLzSWUek5bV7Wg94ljsz72uw+9gH+6+OQ5/R1Sf3L7UNvhMNMb8ecmdrQHNQQ9CmTJnjK3NFNFnLG4AcO2oUoCGr0K+sbdTfZFAZAEd27x+HgSjV4ZXHaLAeTyqxJ8ig8Jw8ZI3MB1r1iDt529Ytp2MtO8nwRk5/2GYwYs261gIYlu8kZMt/OqIps2x5GRJFoDKnyKm2AS0GT0jsTxVG8Yazv8eNNLu/HFPzxXc/3/kk7u2DcQhcjcF95yAEng2Uwmy3G/LzL3cdOVu6lG8xnYCpWVagtqcDiftjSNLAlo9MrobvLhRCQ1a0JJ7uT4s30ROBwQAXjkiRI+GZkSWONJExesbc1+f1WLL29y3d11kaEr5Mn7/dznMdP1eOZSbsujiPAoIgxLQcijIKZbUzK5Culq8OBd563ChmVB3PS759LZ3RMTVBxAJGUhkrIgCsAf9vZjWeNK+CadQ6ZyyB27etE7EkeUc8iMYWWLD1dt6caKFh/CSRN+VYJYhR3NQumL6GV9HCGEkMVVbFnnpVb+udbuUWk8JoSQ0ix2P8+ZO0ErMrdaZqZ6QSlPIAnuRhBNPgUJ3Zr2c3/uQngz+5jgvEtVk/lz0rGVcHLqzjTFsB2Onz56FD98+AjsnESJdR0B/Mul67Gs0QPADR5P3rQio9mvIuSRS/7bk8uns/TGIJmkBjH9/kwYFuK6m7TSGnB3DfKrElqDKvrHU3nve1lk6EzHkCRRmBLkzpgcY+sdTeCnu45i/2DMrQYCwLY5JIEh5JFxalf++31yCe0d+4fyzsVyHHCeXkQB97/cQbqailgwVpW741hmR6ShaAq65WSTt2RRwP/3ipOp3RFCsiYnrA0ZNlKm7d4XOMC3/nwQ9+4dmNOYXSjez5i7g5yUTsIDkK66BFg2h25O7I64cXkI//mG03Hz7/dh/2Bs2h0VC2IMAjhsAMfHEyUd92LKjGWyKMApfRgGAHhVEd1N3vIeGCGkomotBkiWNorlE1J55WiHNPYQMrtMOykWjX2lmRyfyMwjrGkPoHc0DkUS3N3oOcBylvdxuDEFSRCmxEyL3Qi7HBtmk8U3OcZeyu+VktDz8MFhfPHe57Hn2DgAN+nI3WBJhV+Vpt04fC54kbN0R0bjc/4bpD5l+rH/uGcf9vVH3AQ4BiiigPaQioDqzsFNtw56un6wu8mL3tEEmnwKNFnEskaGoagO3bKzc1gCY/j7M7vx6yePF1xHzhiDIDD4FBFn9jTOuFZ9pnk/y3Hc+WwAY4mJTXKCmowmnwLb4RiIpBA3JipsmQ5HXzgFjyxkq95HOUc45Z7/+s4g/umik/Gzx3rRO5JAmKMm+v8lndwxmjAgMAFdIQ9G4np2EpQxN9Gg2aciYdo1eSM2l4EtU1Yqplto9CpQRAGG7aB3NIG4YcFyHHQ3egtmMp7aFcIPrnkp9g1Ei0ooyUyO//qpPnz27r3wqRJCHnlKRY3pSmBNnlwfjRn44h/2QZEKJ27otgNZdDN6p0tSme5v5VIkAa0BFU0+BeGkiWjKzFvUMJ1mrwafIsKjSGAAwkkT4ZSZN5FjO8CvnurDH/YO4KJ17bh8cyfWtk/siHBGT+O0lUN004Zu2hiNG/DIIvyalM1Eq2fLGjQcOBEr6nGEEEKqz4bOIJ4+Fi7qcUtJrd2j0nhMCCGlWex+/ozuBkiCANtxP/OCsXSCB09vgjAzlk4KERjwTxedjLXtgWk/989nswmyMDjniKQsjCeMouIXhQxGUrj5989hd859m8CAN2/pwTXnrIA0zSYaE49laAuq8CpzC0PmJjNk4jm5SQ0p0y09bdocK1v92UnADEUUIAoMAnODwZy7ZakzxzPTZh/ARIzt4YPD+N5fDiOmW+hp8sDhQNKwENVteGUB/3TxWly1pWfG9/vkc5EEwU3owMTuc5lJSqBwrGryjmOZHZFSpgPTdhBOmjhtWRBXbemZ0/UmhNSvzDh9x66j+K/7D4CBpye7RBi2g339Udxw1x7c9NqNc0rwyI33exURYwkDKYuDc56NU7tJaBwJ08bGZaHs7ojb1rTg19ediz3Hw7hnTz9+tPPF7CZJM41emaRVBmBZQ/UmPmT6f2cOCZaAO9kYVKUZ5y4IIbWn1mKAZGmjWD4hlVeOdkhjDyGzy7QTvyIgZsy++QCNfaUrNI+wviOAd9z+KPb1RxHSpGxlcPe/HJbD4ZEFaLKAlDU1Zlrs3ATNYSwtxSb0ZNbMjsYNgLkbngEMKdPG8bEkljW6G0XNNpdQrBVNvrI+jiwtmff1tbc/CkUU4FUkaHJ+dZqZ1kEX6gdHYjo+eufubLKFX5XgU0WkDAeW40BgDNGUha0nNeGZ4+GSKuIUMt28X5NPwWBkYrM2kQF2ujp0OGlCU0R4ZRHNPhUpM5FNAslImg56x5LwyCI8soCQNnH+uWuvU7aNnkYvTu9uqOr+f0knd2TeJIok4KRmdxLUchxIQs7NgO0siYC543DcsuMQYrqFjqCWbXiaIKIzpOHoaBK6ZWMgkkKDV5m2hFkpCSWCwHDF6V345ZPH3JszT/7PZ2vwuQksjsOzzzNdx7GuIwCA4bmB+XUugFv6p8mnoMEjI5qyEE6a6V0VC1vT7kN3sw+Hh2Jo8StoDaho9rsVQMJJM6+aR8p08Ls9/fjdnn6sbffj8k1deMX6NnhkEQJjWNvhh8M5Dg7G8fiRsbwkD87dnSoThgWBMfhUCX5VgkeZmsxSD647fw0e2D9c1OMIIYRUn3+9ZD3ueLS3qMctJbV2j0rjMSGElGax+/mNy0I4pcOPvX0RWJxDYu4i92L3xHYcB5YDhLwyrtjcBWmaTRUy5rqLEik/w3IwGEnBLGUH9EkeeP4EvnrfAcTSFTEAoC2g4oZL12HT8oZZf18S3N2CJpdoLsV05dP9qgSv7MXx8RSWN3rwH6/diA2dwewkYOaxmQQKpKtiaLKYFycpZrOP6eJmPlVCs5+jP5zCnY8fw/IGD5r96rQTgpPPJa/UtMBhp49PU4RpY1XT7TgGBsQNGw1eGR+4cE1VB6QJIZV1794B2Jxjec5GTpogoiMoYCCi45Ydh3D2quaS+5HJ8f7WgIa+8RRMh6cnxTksm0MQGBo88pRy94LAsLm7ARuXhfDIC6PYfWwcouDu3GZYhZMiOBxYtnufcvmmzrlekgWX6f/3Hg9jLvkdiiRiTXtxcxeEkNpRazFAsrRRLJ+QyitHO6Sxh5DZZdpJMOhBbHj2XfNp7JubQvMImZhnOGVCEgQYlg1RZLAdQEzv6g5g2vV9xc5N0BzG0jJbQk9u7L/VryI5lgDAIDAGJroVaIeiOnyqWNRcQjHed/4qfOP/DsBKb8iVG4HLhI0kgeF956+a198h9WvjshDWdwaxrz+KJrn0ddCT+8E9x8JTki0YWHouTUTStCGLDM1+ddaKONs3dOAvB4dnTJ4rNO/HwRHTLTC4iX2KJKArpGEwoiNp2LA5x2hMh6fRA1V272eTpgNVYhAFAYmcSh5J00bStHH3nj5c41+BZr/qnnd67XXGiah7DppcnWurZ56Rr3OZN0mmfItHERHQ5OwE73jCxOo2/5IImO/ti+DQiRgavcqUag8svcOiTxGxvNGLhG7hRExHQrewvjMwp528MjIT0n5VxEBER9K04TgcSdPGQESfUgJrPs/zgQvX4AMXzv9vTf67Ia+M7iYPWgMq5Gl2qxQYw1VbuuFVRAzHDKQsd2GFIonQZBFtAQXnn9yCgJafb7V/MIav3Lcfb/j2Tnz9TwdwaCiGJ4+O4eP/bw8+/es9+Pzvn8Onf70HH/9/e/Dk0bG833U4RzRloj+cxNGRBEZibqmkeqKoImZ7tVj6cYQQQqrPodEE1Fnq2qoiw6HRxCIdUXWotXtUGo8JIaQ0i93PCwLDJy9Zj9aACoEx2I67y5Vd5MdD03E3OLjuwtWzJnaQ6mI5zpwTOxKGhS/84Tl85u59eYkdLz+lFd9721lFJXYokoCuBm1eiR3AzDGfwaiBBq+MGy5dj83dDZAkYcpjFdFN8DAtDgagNaC6paoxEeSerc3NFDeLGzYSho09x8bx4Z8/hff+6DFc84NdePjg1OTXyeeSshw0+1QIDNmFy81+BSnTmTFWldmZaX1noKxxOkJI/ZttHqDBK+PQiRj29kXm9Py5/VxMt9ESUKBJAmyHw7DdLTfXdczcVwkCw0detRaKJMCyAXB3QrsQ03arSVX7fUrmuqhzmCgUBYYW/9RkGEJI7au1GCBZ2iiWT0jllaMd0thDyOwy7WSkiJ35aewrr4mYZxDedL9k2RyKKKAzpEEU2JzX95GlLbOQ/YK1rdi4PJT33smNlXkUEaokwHLSlWjBIAoMumUjqdtlGycVRcQbz1qe/Zrn/Mt441nLodTpZt5k/sq15joj9x6RT9qZZvI82nTzU50hFQFNwrf+fBAf+fnTJc2VJU0bCd1GyrSy1e07Qx74VBltQQ2SKIAB0C0HuulAtzgEJkBkgCwKaPGr6App2cojGXfv7sdbb9uF2x56AfGcuc6MhGGhbzyJgXAKKbP61lUv6cod0+10N7kixVK4GRhNGDDTN0SFqKIAQRDwwYtPRotPLWtpsmJLYJXrecrxtyZjjCGgyQhoMuK6hfGkCX1Sgz+jpxHXv3It7tjVi96ROKKcQ2YMq1r9uGpLN87oaYRu2tixfwi/3d2fN4EWN2z8+uk+/PrpPsgigywKaPYpCEoCTJvj8FAMX71vP65/5Vqc0dM45fgsx0E46SCcNKFIAvzpih7SNK93rRhPmvDKIuIzdK5eRcR40lzEoyKEEFKs0YQBTZZg2CYKbVzJAGiKtOTKP9faPSqNx4QQUppK9PPb1rTgP99wOr71wEE8NxCFaXOAc0TTgSxnhh2kfaqID110Mt59/uqyHQ+pbvv6I/jc7/ahP5zKfs+riPini07GK9e3TVkMXIhXkdAWUMv2Pi4ldlTosV5FAAeHKgkQBQbH4SW1ueniZjHdwvGxJGzuJtEENAmqJGJffxQ33LWn4OLlycdnOhxBTYLN3R3oEoYNWXBmjVXNtuMYIYQUUsw8QNjh8/ocPrmf8yhuxaSOoIY3benBVVt6Zu2rzj25Ff/8yrX4yn37YVgOCgYN4MYNgh4ZG7qqf8fNbWta8LZzVuCr9+2HZfPpTimLwU1cObUzgE9esp4S9wipQ7UWAyRLG8XyCam8crRDGnsImV2mnXz450+BYdqPozT2LZDcmOdDB4dw795BDIaTSJg2ZHv2mCkhpcqNlbF0hZjjY8m8SrSOwzEc09HoU8o2Tl62qQu/froPMX3qGgu/KuKyTV3z/hukvpVrzTVQ+j3i5Pmp3tEEbn3wEOKGjUavAkUUYNhOSXNlccOGwwGPLKItqMGvuqkNflXCskYPTkRSSJo2wikLHlnAqV0BnNnTiEePjKF3JA6Tc4Q0EX6PB5GkiZG4Oz6nLAc/fuQo7t7dj7ec3YPLN3VBmbRJUMKwkDAseBQRjV6laip5LOnkDqC8b/Jalikrl1taJ1emrFSLT12Q0mTlmpAu5nkWevLbp0rwqRKSho3xpIFkTsmfM3oasbm7AQcH4winDIQ0BWvafRDSCyNUWcSrNnTgVRs6cHgoht/u7sefnh1EPOc5TJvDtG2kzCSCmoyQR0aLX8FwzMAdu3qxubsh+3yFGJaDUcvAaNyARxHhUyX4FakmP6A3eGSYDoeAqRmsLP3PtDkaPHJFjo8QQsjMGjwykqYNgbkLFhzu/sv9OmnYS7Ifr6V7VBqPCSGkdJXo5yd/Fh6O6vjE/9sNh3N31z/GwDkAxmHbHA53d4r+0Tu24CUnNZX9eEj1sR2OOx45itt3vpiX8HNqZxA3XLoOXQ2eop4n6JHRki5xXE6lxHMKPTacNPCdBw/Pqc0ViptxzjEUTcHmHCJj4AxQRLdCa0dQwEBExy07DuHsVc0FK29MPr71HQHsG4iWFKuaXDqbEEJmU+w8QJNXmdffKUcM/t3nr8b6ziC+/MfnsbcvAm5ziIK7Y6JXduPaQU3EiZg5bX9bbc5d04r/efhFjCZMmLbjfmZmDJy7916Z4VcSGE7pCODNRSbDEEJqVy3FAMnSRrF8QiqvXO2Qxh5CZrdtTQv+8eVr8Olf753yMwGAKNDYt5AyMc+Ny0N47/mraXMbsqAmx8oyC8mHoinolpOdK1nZ6sfHtp9SlnHScThu2XEImiyip9GD4bgJw3agiAJafHJNxbpIZZVzHXSp94iZvtpxOK75wS7EDRsdQS27QZwmlDZX9sSRMfzXnw+gwSPDo+SnNfhVCWKDhnDCwnWvWIMzexpxamcAUd3CG+IG9g/E8tZi2w7H73b34392HsluQhtOmvjmnw/hl08cxztfdhJevq5tyhrrpGEjaSSrJsljySd3ALTTHTBRWmdffxQdQSFvF8ZMaZ31nYEFLb9YrgnpYp5nMSa/3R3JPEiZNsJJM1vaR2AMazv8s/7+qlY/PnjRyXjP+avw5+dO4BePH8ORkYlyfg53d8oeT5rwyAK8ioSjwzEcHIwX9fxApkOyMcIM+NKJHl5FLGoXzmrB3DVIkEV3+Sjn7tcAh2Vz1NCpEELIkpTpp0VRgISc+w9wOJazpPvxWrpHpfGYEEJKV4l+Pvez8NO94xAEBsfmYAIDA3Mz8sAgSBym5bgLKGu84iMpzkA4hZt/vw97jk9UERUY8JazV+CtZ6+AWOT7stmnIuRduAnNUuI5hR67bXXLnNpcobhZynSgW447mesAmixCU9z2whhDg1fGoRMx7O2LFDzmQsdHiRqEkIW2mPMA5YjBn3tyKwKajGtvfxSK6MbANTn/uBu8mLG/rSYbuoJoD3kwGNUhiQwim7jPcrgDy+YQBQFBj4SbXrsRm7sbKnewhJBFU0sxQLK0USyfkMorVzuksYeQ2W1a3gBZZO7nNJFBYGwiQZ/GvkVDm9uQhVYoVuZXJfgUH5KGjaGYgVWtPtz53nMgSeWZL9vbF8GhEzE0ehWIooj2YP4C8lqKdZHKK2c/OZd7xNz38+Q1x6XMlW3oCuJPzw1iX38UmiwWiFtbWN8ZwNvOXpE9ngavAr8qwa9KiKXXZgOAIDJcecYyvGpDO37x2DH87LFepEwHANAfTuE/7nkOP3v0GN59/kqctaJxynHnJnk0eBR4lMokeVByR9pSvxmg8osLR5PdXRsNy8F40kBct8H5bEXXJ3hkEZdu7ERrQMXnfrcPDEBUt5D7FEnTQdI0wAD8eNcRvOvclehu8hb9NzjniOkWYroFUWBuNQ9Vqnj22WzGkyY0WUTC4bAdNzteEADOAdtxE2k0Wcxm4BFCCKkuuf24u4DBDUy7/Tinfhy1cY9K4zEhhMxdJft5GodJxv37TuBrf9qfVzW0I6jhhkvX4bRlxb0/GWNoC6jwqdUdapxrmysUNzPt9K5djrsQtzWguklSaaooIOxwjCaMMp4BIYTMTy3OA4wnTTAwNHqVgsdVS/2tIDBs39COPcfGYdscTOQ591+AKAjobNCQMGy6ByNkiamFGCBZ2iiGQEjllbsd0thDyMzGkyY8ioREZn2Wu78dHM5p7COkjswUKwunLDT5ZHxs+yllS+wAgNGEAdPmUKbZXK2WYl2k/pR6j1iu9/Nc49aSKKAtqCFg2BiO6TBtJ/szryLhmm0n4fLNXfjRziO4e08/7HQ5noNDMXz8/+3BS3oa8J7zV2Fte2DKMWWSPDRZRINXhldZ3DlQ2n6RZGVK66zvDCChWzgR05HQ3Yynm167kcovzpMiCWgLaOhu9CDkkUuujhHSFHgkAY1eBaubfWgLqFAn3ThwAH85MIxrfvAorv/503jg+RN5HVYxbIcjkjTRN55E72gCjlN8Ispia/Iq8Cki2oIaNFmEw91AhsM5NNn9vk8R0eRVKn2ohBBCCqB+vD7Q60gIIbWJ+m8S1y3cdM8+/Mc9+/ISOy5e34bvvu3MohM7RIGhM6RVfWLHfE2Om4VTJsA5FEnEskYP/JPOX7cdyAKjNkQIqTq1Ng/Q5FUgiwzGNHHuWutvz13TigavAkWaev+1rNEDWRRq6nwIIYQsDRRDIKTyqB0SsriozRGydCx2rKzeYl1kaSvn+3k+bdGjiFje6EGzT4UwaV12k0/BBy8+GT94+1m4cG1r3s+eODqO9/3vE/js3c/i+Hiy4HOnTBsD4RSOjyeRMKyCj1kI9T3rSkpG5RcXniQKaParaPAqiCRNRFJmNiNsJmvafehu9uHwUAwtfgUNHhkhTULKchBOmoik8juOp3rH8VTvOBq9MrZv6MBrNnWiq8FT0rGatoPqTe3IL422otkD3eSwHAeSIECVGQYjBtZ3BrChK1jpQyWEEFIA9eP1gV5HQgipTdR/L23PHA/j5t8/h/5wKvs9nyLigxefjIvXtxf9PLIooCOkQZ5mR556kxs3G47r+PqfDuDYWAK+SSWZ3RLRJrUhQkjVqqV5gNx7lo6gkLdpUi32txu6gji1K4h9/RGENA0255AEAZoiABwYiOg1dT6EEEKWBoohEFJ51A4JWVzU5ghZWhYzVlZvsS6ytJX7/TyftsgYQ8grw69JGInriE1aT7280YtPX34q3jAQwXcffAFP9Y5nf/bn54fw4IFhXL6pE289ZwUaCySj6KaNgbANJb1B/0Jverc0Zl5JSTKldS5Y24qNy0NVOaFTD0SBodGnoKfJi2a/OutCCIExXLWlG15FxHDMQMpyEy8YY+4umUEVrz2jCyc1e/N+byxh4qeP9uItt+3Cx+7cjQcPDMEqsZpHtcqUY/KrIgYjBsAAnyIBDBiMGNOWYyKEEFIdqB+vD/Q6EkJIbaL+e2myHY7bH34RH/rZU3mJHRu6gvju284sKbHDo4hY1uBZMokdGZm42ctPacPHtp8CvyphIKIjadpwHI6kaWMgolMbIoRUvVqZB8i9Z6mH/nbifCSEUxYkUYAqCUiZTk2eDyGEkKWBYgiEVB61Q0IWF7U5QpaexYqV1VusiyxtC/F+nm9bFAWGtoCGrgYPVFmc8vN1HUF85fWb8IW/34jVrb7s922H41dP9eEt39uF2x9+EUnDLvj8huVgMJJC72gCMX3hKnkwznk1b8w/b5FIBKFQCOFwGMEgZbOR6hbTLYwnDBjW9MkXTx4dwx27etE7EofJOWTG0N3sw1VbunFGTyM459jbF8Hdu/vx5+dPwLSnNvFmn4JLNnbg0o2d6AhqMx7TimYfxArfLMzWjh8+OIxbdhzCoRMxmA6HLDCsbvPj/ResLntpNELI3NB4TGZC/XhtoPGYkPpAYzKZjPrv2lNsO04YFgZyEjj6w0ncdM9z2NsXyX5PYMDbzlmBq7euKOmzf0CT0eJX8nbgWaqoDZG5oPGYkNJVW38733ZcbedDyFJE4zEhpau28YvaMVmKqq0dlgO1ZVLN6rHNLQRqx4SUrtr6F2rHZD6q7f2cK5IyMRY3YDtT11E7nONP+07gB399AYMRPe9njV4ZbztnBS7b2Alpho3uZFFAg1eGX5XKOm9KyR2EVKGkYWM8aUyb/eVwjoODcYRTBkKagjXtPggFOoZI0sQfnx3E3bv7cXQ0MeXnDMDWVU14zaZObF3ZXHAhRy0kdwCA4/BFKY1GCJkbGo/JbKgfr340HhNSH2hMJoVQ/11b5pLc8ad9g/j6nw4gnhNn6AxpuOHSddjQFSrp7zf5FDQUKEe8lFEbIqWi8ZiQuamm/rYc7biazoeQpYjGY0LmpprGL2rHZKmqpnZYDtSWSbWrtza3EKgdEzI31dS/UDsm81VN7+fJbIdjLGEgkjQL/tywHPzm6T7879+OIJLKr8axrMGDa89diQvWtsyYvNHV4IFWoFLIXElleyZCSNl4FBEexQPdshFOmFPK9wiMYW2Hf9bnCXpk/MOZy/H3L1mG3cfC+O3ufvzlwFC2mgcH8LfDo/jb4VG0+lVcmq7m0RpQF+K0FlSmHBMhhJDaRP14faDXkRBCahP13/Urplv4+p8O4P7nTuR9/5WntuOfXrEGPrX40CBjDK0BFf4SfmepoDZECCGLo97623o7H0IIIUsDjV+EVB61Q0IWF7U5QshCof6F1JNqfj+LAkOLX0VAkzASM5Ay8zfdVyQB/3Dmcrz6tA787NFe3Pn4MeiWAwA4Pp7EZ+5+Fqd0BPCe81bijJ7GRTlmmo0lpIqpkoi2oIhG20E4aSKasjCXYjuMMWzubsDm7gaMJ1bj3r2D+O3uPvSNp7KPGYrpuH3nEfzob0dwzqpmXL65C2edtDgdESGEEEIIIYQQQsrriSNj+OdfPJ1XRtinivjQRWtx0fq2kp5LFBjag1pZd5whhBBCCCGEEEIIIYQQQgghhJDFoEoiuho8iOkWRmMGLMfJ+7lflXDtuStxxelduP3hI/j9M/1w0su1nx+I4p9/sRtbVjbh3eetxOrW2Tfnnw9K7iCkBsiigBa/ikavgkjSRCRlwnZKT/IAgAavgje+tBuvP2s5njw6jt/u7sNfD45kn8/hwF8PjeCvh0bQEdRw9dk9eONLu9EW0Mp5SoQQQgghhBBCCFkAlu3gG/93EP/9fweQGzrYuCyIT166Hh3B0j7fy6KAjpAGWRTKfKSEEEIIIYQQQgghhBBCCCGEELJ4/KoEryxiPGkinDSnbLjf4lfxz69ai9efuRy3PnQYfz04kv3ZrhdG8egLo7j41Ha842UnlTzvWixK7iCkhogCQ6NPQcgjI5qyEE6aU7LHiiUwhjNXNOLMFY0YjRv4/TP9+N3uAQxEJqp5DERS+Mof9+P7D72AXf9yMS3kIIQQQgghhBBCqtxTveP4xv0Hsl8LDHj7tpPw5i09EAVW0nN5FBHtAQ1Cib9HCCGEEEIIIYQQQgghhBBCCCHVSBAYmnwK/KqEkbiOpGFPeUxPsxefveI07O0L47sPHsae4xEAAAdw37ODeOD5E7ji9C5cvXUFuho85T2+sj4bIWRRCAJDyCuju8mD1oAKRZpfU27yKbh66wr86Not+PzrNuJlq5uRu27j8s1dlNhBCCGEEEIIIYTUgLNOasK1564EAHQ1aPivN5+Bt5y9ouTEDr8moSNIiR2EEEIIIYQQQgghhBBCCCGEkPqjSAI6Qx60BTVIQuE10hu6QvjaG0/HZ6/YgBXN3uz3TZvjzseP4y3fewTf2XGoYILIXFHlDkJqGGMMAU1GQJORMCyMJ0ykzLl3EKLAsGVlE7asbMJQVMfvn+nHH/YO4qqtPWU8akIIIYQQQgghhCykj24/BYrIcNmmTniV0sN/TT4FDV5lAY6MEEIIIYQQQgghhBBCCCGEEEKqh1+V4JVFjCUMRFIWOOd5P2eM4WVrWnD2qmbcu3cAP3z4RQzHDABA3LDxn386gB8/chR3XfcyLCtDFQ9K7iCkTngVCV5FQsq0MZ4wkTCseT1fa0DF2845CTdceuq8K4MQQgghhBBCCCFk8WiyiP/vopMxEE6V9HuMMbQGVPhVChkSQgghhBBCCCGEEEIIIYQQQpYGQWBo9qsIaDKGY3rBjfZFgeHSjZ24aF0b7nryOO7Y1YuY7q7VXtboQVdIK8+xlOVZCCFVQ5NFdIQ0LGv0wK9JYIzN6/lEYX6/TwghhBBCCCGEkOonCgydIY0SOwghhBBCCCGEEEIIIYQQQgghS5IiCehq8KA1oE67flqVRbxpSw/+99oteONZy6FIAj7x6nXzXq+dQbO1hNQpVRLRFhBheR2MJ01EC5QKIoQQQgghhBBCCJFFAR0hDbJI+8AQQgghhBBCCCGEEEIIIYQQQpa2gCbDp0gYTRiIJM2Cjwl6ZLz3gtX44MVr0dXgKdvfpuQOQuqcJApo8ato9CqIJE1EUiZsh5I8CCGEEEIIIYQQAngUEe0BDQJV7iSEEEIIIYQQQgghhBBCCCGEEACAIDC0+FUENAkjMQMp0y74uCafUta/S8kdhCwRosDQ6FMQ8siI6hYiSROm7VT6sAghhBBCCCGEEFIhfk1Cq18tW4lgQgghhBBCCCGEEEIIIYQQQgipJ6okoqvBg5huYTRmwHIWdu01JXcQssQIAkPIIyOoSYgbNsYTBgyLkjwIIYQQQgghhJClpMmnoMFb3l1kCCGEEEIIIYQQQgghhBBCCCGkHvlVCV5ZxHjSRDhpgnO+IH9HWJBnLbNvfvObOOmkk6BpGrZu3Ypdu3ZV+pAIqXmMMfhVCcsbvegMeeBVKNeLEEIIIYQQQgipd4wxtAU1SuwghBBCCCGEEEIIIYQQQgghhJASCAJDk0/B8kYPfOrCrLuu+uSOn/3sZ7j++utx44034oknnsDmzZuxfft2nDhxotKHRkjd8CgiOkIaljV64NckMMYqfUiEEEIIIYQQQggpM1Fg6Axp8C9QoJEQQgghhBBCCCGEEEIIIYQQQuqdLApoD2roDHkgCuVdc131yR1f/epX8e53vxvveMc7cOqpp+Lb3/42vF4vvv/971f60AipO6okoi2gobvRg5BHhkBJHoQQQgghhBBCSF2QRQFdDR5osljpQyGEEEIIIYQQQgghhBBCCCGEkJrnUUTIYnnTMap6mz7DMPD444/jk5/8ZPZ7giDg4osvxs6dOwv+jq7r0HU9+3UkElnw4ySk3kiigGa/ikavgkrkd1A7JqT2UTsmpPZROyakPlBbJqT2laMda7KI9qBW9l1jCCHFofGYkNpH7ZiQ2kftmJDaR+2YkPpAbZmQ2kftmJDaR+2YkOpW1ZU7hoeHYds22tvb877f3t6OgYGBgr9z8803IxQKZf91d3cvxqESUpcEgYFVILuD2jEhtY/aMSG1j9oxIfWB2jIhtW++7divSegMUWIHIZVE4zEhtY/aMSG1j9oxIbWP2jEh9YHaMiG1j9oxIbWP2jEh1Y1xznmlD2I6fX19WLZsGR5++GGcc8452e9/7GMfw44dO/DII49M+Z1CGWXd3d0Ih8MIBoOLctyEkPmhdkxI7aN2TEjto3ZMSH2gtkxI7ZtrO04YFnTTQaNPWYzDJITMgMZjQmoftWNCah+1Y0JqH7VjQuoDtWVCah+1Y0JqH7VjQqqbVOkDmElLSwtEUcTg4GDe9wcHB9HR0VHwd1RVhaqqi3F4hJAFQu2YkNpH7ZiQ2kftmJD6QG2ZkNo313bskUV4laoO/RGyZNB4TEjto3ZMSO2jdkxI7aN2TEh9oLZMSO2jdkxI7aN2TEh1Eyp9ADNRFAVnnnkm7r///uz3HMfB/fffn1fJgxBCCCGEEEIIIYQQMoExVulDIIQQQgghhBBCCCGEEEIIIYQQUoKq377v+uuvxzXXXIOzzjoLW7Zswde+9jXE43G84x3vqPShEUIIIYQQQgghhBBCCCGEEEIIIYQQQgghhBBCCCGEzFvVJ3e88Y1vxNDQED796U9jYGAAp59+Ov7whz+gvb290odGCCGEEEIIIYQQQgghhBBCCCGEEEIIIYQQQgghhBAyb1Wf3AEA//iP/4h//Md/rPRhEEIIIYQQQgghhBBCCCGEEEIIIYQQQgghhBBCCCGElJ1Q6QMghBBCCCGEEEIIIYQQQgghhBBCCCGEEEIIIYQQQghZyii5gxBCCCGEEEIIIYQQQgghhBBCCCGEEEIIIYQQQgghpIIouYMQQgghhBBCCCGEEEIIIYQQQgghhBBCCCGEEEIIIaSCpEofwELjnAMAIpFIhY+EkPoXCATAGCv781I7JmTxUDsmpPZROyakPlBbJqT2UTsmpPZROyak9lE7JqT2UTsmpPZROyakPlBbJqT2UTsmpPZROyak9hXTjus+uSMajQIAuru7K3wkhNS/cDiMYDBY9ueldkzI4qF2TEjto3ZMSH2gtkxI7aN2TEjto3ZMSO2jdkxI7aN2TEjto3ZMSH2gtkxI7aN2TEjto3ZMSO0rph0znkm5qlOO46Cvr2/WTJdIJILu7m709vYuSOdXi+iaTEXXZKrca7Js2bIFyQxdau24Hs6jHs4BqI/zKPUcFirDu1A7rofru9DoGhWHrlO+xWzHhdTL61EP51EP5wAs3fOodFueTr28HpVE17A8auE6Vms7LlYtXONqQterNLVyvRarHdfK9SgGnUt1WsrnUuvjcTFq8fWlY14c9XLMC92OOefo6empqetUC2rx/VcLavW6VuN4XKvXslLoepWmXq9XJdpyPV5LOqfaUI/nBFR+TK6X60rnUV2W2nlUuh3Xmnp5fxRC51a7qHIHAEEQsHz58qIfHwwG6/LNMB90TaaiazJVMBhcsIF9qbbjejiPejgHoD7Oo9LnMFM7rvSx1QK6RsWh67SwaDyuXfVwDgCdR7mU2panU+nzqAd0DctjKV7HcrXjYi3FazwfdL1Ks1Sv13TtuJ6uB51LdaJzKZ/FHo+LUelrMhd0zIuDjrmwTDuORCKL9jeXIrquC4Ouq6sc4zFdy9LQ9SoNXa/iFNOW6/Fa0jnVhno8p4VA88i1jc6julTqPKox1lVO9fL+KITOrT4JlT4AQgghhBBCCCGEEEIIIYQQQgghhBBCCCGEEEIIIYSQpYySOwghhBBCCCGEEEIIIYQQQgghhBBCCCGEEEIIIYQQQiqIkjvSVFXFjTfeCFVVK30oVYOuyVR0TaaqpmtSTccyH/VwHvVwDkB9nEc1n0M1H1u1oGtUHLpO1aVeXo96OI96OAeAzqPa1Mt5VBJdw/Kg67jw6BqXhq5Xaeh65aun60HnUp3oXOpbLV4TOubFQcdcvX9zKaDrujDoupYPXcvS0PUqDV2v8qnHa0nnVBvq8ZyqQb1cVzqP6kLnQWZSz9eVzq2+Mc45r/RBEEIIIYQQQgghhBBCCCGEEEIIIYQQQgghhBBCCCGELFVUuYMQQgghhBBCCCGEEEIIIYQQQgghhBBCCCGEEEIIIaSCKLmDEEIIIYQQQgghhBBCCCGEEEIIIYQQQgghhBBCCCGkgii5gxBCCCGEEEIIIYQQQgghhBBCCCGEEEIIIYQQQgghpIIouYMQQgghhBBCCCGEEEIIIYQQQgghhBBCCCGEEEIIIaSCKLkDwDe/+U2cdNJJ0DQNW7duxa5duyp9SIvqwQcfxOWXX46uri4wxvCrX/0q7+ecc3z6059GZ2cnPB4PLr74Yhw4cKAyB7sIbr75Zrz0pS9FIBBAW1sbrrzySjz//PN5j0mlUrjuuuvQ3NwMv9+Pv//7v8fg4GCFjnhx3HLLLdi0aROCwSCCwSDOOecc/P73v8/+vNLXpNbacTHvswsvvBCMsbx/73vf+yp0xFP927/925TjW7duXfbnlX5PFOukk06ach6MMVx33XUAqvd1KEffPTo6iquvvhrBYBANDQ249tprEYvFFuX4a63NLjQai2dH43Nlldpmf/GLX2DdunXQNA0bN27EPffck/fzSrynSzmHW2+9Feeddx4aGxvR2NiIiy++eMrj3/72t08ZH1796lcv6DkApZ3HD3/4wynHqGla3mMq1b+Uch6FxmLGGC677LLsYxb79Zit3y7kgQcewEte8hKoqoo1a9bghz/84ZTHVMv4ONfj+OlPfwrGGK688sqFPcAaUOo1HB8fx3XXXYfOzk6oqoq1a9dO6TuXolKv49e+9jWccsop8Hg86O7uxoc//GGkUqlFOtr69fnPfx6MMXzoQx+q9KFUrePHj+Mtb3kLmpub4fF4sHHjRjz22GOVPqyqZNs2PvWpT2HlypXweDxYvXo1PvvZz4JzXulDq5hqGf9LUQ9xrYx6iW8BtRvjyqj1WNdCKaa9TVbMZ8GFNFu7KmS2OMJCm639TFaJa7xQ8cOFHIdmOmbTNPHxj38cGzduhM/nQ1dXF972trehr69vxuecy/trJrU4Dlcbim2XH8XC565c1+7o0aO47LLL4PV60dbWho9+9KOwLGsxT6UiCn3+p+s1YbbP/kv1frVUtXB/UIrZxsG5zl1U8h5ltnMqdN/MGMOXvvSlaZ+zkq/RQo2rdI9Tumq/966HuEQ93QuVY51iNZxHLrrXWlj1+tm03j8fVvua5Epa8skdP/vZz3D99dfjxhtvxBNPPIHNmzdj+/btOHHiRKUPbdHE43Fs3rwZ3/zmNwv+/Itf/CK+8Y1v4Nvf/jYeeeQR+Hw+bN++vW4XSOzYsQPXXXcd/va3v+G+++6DaZp41atehXg8nn3Mhz/8Yfz2t7/FL37xC+zYsQN9fX143eteV8GjXnjLly/H5z//eTz++ON47LHH8IpXvAJXXHEF9u7dC6Cy16QW23Ex7zMAePe7343+/v7svy9+8YsVOuLCNmzYkHd8Dz30UPZntdJOHn300bxzuO+++wAAr3/967OPqcbXoRx999VXX429e/fivvvuw913340HH3wQ73nPexb82GuxzS40GotnR+Nz5ZTaZh9++GG8+c1vxrXXXosnn3wSV155Ja688ko888wz2ccs9nu61HN44IEH8OY3vxl//vOfsXPnTnR3d+NVr3oVjh8/nve4V7/61Xnjw09+8pMFOf65ngcABIPBvGM8cuRI3s8r0b+Ueh6//OUv887hmWeegSiKeWM1sLivx2z99mQvvPACLrvsMrz85S/HU089hQ996EN417vehXvvvTf7mGoZH+d6HC+++CI+8pGP4LzzzlukI61epV5DwzDwyle+Ei+++CLuvPNOPP/887j11luxbNmyRT7y6lLqdbzjjjvwiU98AjfeeCP27duH2267DT/72c9www03LPKR15dHH30U3/nOd7Bp06ZKH0rVGhsbw8te9jLIsozf//73ePbZZ/GVr3wFjY2NlT60qvSFL3wBt9xyC/77v/8b+/btwxe+8AV88YtfxH/9139V+tAqolrG/1LVS1wrox7iW0DtxrgyajnWtZCKbW+TzfZZcKHN1K4mKyaOsNCKaT+TLfY1Xoj44UKPQzMdcyKRwBNPPIFPfepTeOKJJ/DLX/4Szz//PP7u7/5u1uct5f01k1odh6sNxbbLj2Lhc1eOa2fbNi677DIYhoGHH34Yt99+O374wx/i05/+dCVOadFM9/mfrpermM/+S/V+tVTVfn9QqmLmCkqdu6j0Pcps55R7Lv39/fj+978Pxhj+/u//fsbnrdRrtFDjKt3jlKbS7+ti1ENcop7uhea7TrFaziOD7rUWXr1+Nq33z4fVvCa54vgSt2XLFn7ddddlv7Ztm3d1dfGbb765gkdVOQD4XXfdlf3acRze0dHBv/SlL2W/Nz4+zlVV5T/5yU8qcISL78SJExwA37FjB+fcPX9ZlvkvfvGL7GP27dvHAfCdO3dW6jArorGxkX/ve9+r+DWph3Y8+X3GOecXXHAB/+AHP1i5g5rFjTfeyDdv3lzwZ5V+T8zHBz/4Qb569WruOA7nvPpfB87n1nc/++yzHAB/9NFHs4/5/e9/zxlj/Pjx4wt6vPXQZhcSjcXFofF58ZTaZt/whjfwyy67LO97W7du5e9973s555V5T8+337EsiwcCAX777bdnv3fNNdfwK664otyHOqNSz+MHP/gBD4VC0z5fpfqX+b4e//mf/8kDgQCPxWLZ71Xi9ciY3G8X8rGPfYxv2LAh73tvfOMb+fbt27NfV8v4OJfjsCyLb9u2jX/ve9+r6GtRLUq9hrfccgtftWoVNwxjsQ6xJpR6Ha+77jr+ile8Iu97119/PX/Zy162oMdZz6LRKD/55JP5fffdVxOfiyrl4x//OD/33HMrfRg147LLLuPvfOc78773ute9jl999dUVOqLKqpbxf75qMa6VUa/xLc5rM8aVUWuxrsVUqL1NNttnwYU2U7sqZLY4QiVMbj+TVfoalyt+uJjjUDGfnXft2sUB8CNHjkz7mFLfXzOpl3G4mlBse2FQLHzu5nLt7rnnHi4IAh8YGMg+5pZbbuHBYJDrur64J7BIpvv8T9drwmyf/el+dW6q8f5gPgqdz1zi5dV0j1LMa3TFFVdMictOVi2vEeflGVfpHqd01fS+Lka9xCXq7V6olHWK1XQedK+1+Or5s+lS+HxYLWuSK21JV+4wDAOPP/44Lr744uz3BEHAxRdfjJ07d1bwyKrHCy+8gIGBgbxrFAqFsHXr1iVzjcLhMACgqakJAPD444/DNM28a7Ju3Tr09PQsmWti2zZ++tOfIh6P45xzzqnoNamXdjz5fZbx4x//GC0tLTjttNPwyU9+EolEohKHN60DBw6gq6sLq1atwtVXX42jR48CqN12YhgG/vd//xfvfOc7wRjLfr/aX4fJium7d+7ciYaGBpx11lnZx1x88cUQBAGPPPLIgh1bvbTZxURjcWE0Pi+OubTZnTt35j0eALZv3559/GK/p8vR7yQSCZimOWWcfuCBB9DW1oZTTjkF73//+zEyMlLWY8811/OIxWJYsWIFuru783Y4ACrTv5Tj9bjtttvwpje9CT6fL+/7i/l6lGq2dlEt4+Ncj+Mzn/kM2tracO211y7GYVa1uVzD3/zmNzjnnHNw3XXXob29Haeddhpuuukm2La9WIdddeZyHbdt24bHH388W8r98OHDuOeee3DppZcuyjHXo+uuuw6XXXbZlP6L5PvNb36Ds846C69//evR1taGM844A7feemulD6tqbdu2Dffffz/2798PAHj66afx0EMP4ZJLLqnwkS2+ahn/y6FW41oZ9RbfAuonxpVRzbGuxTZde5tsps+Ci2G6dlXIbJ+XFtt07WeySl/jXHP5fF+N41A4HAZjDA0NDTM+rpT313Sq8fzrEcW2y4Ni4XM3l2u3c+dObNy4Ee3t7dnHbN++HZFIpKJ9/UKa7vM/Xa8Js332p/vVhbOY9wcLpZS5i1q7RxkcHMTvfve7ouYGquU1Kse4Svc4pam193UhtdrP18u90FzWKVbTedC9VuXVU79dz58Pq2lNcjWQKn0AlTQ8PAzbtvM6PwBob2/Hc889V6Gjqi4DAwMAUPAaZX5WzxzHwYc+9CG87GUvw2mnnQbAvSaKokz58LgUrsmePXtwzjnnIJVKwe/346677sKpp56Kp556qmLXpB7acaH3GQBcddVVWLFiBbq6urB79258/OMfx/PPP49f/vKXFTzaCVu3bsUPf/hDnHLKKejv78e///u/47zzzsMzzzxTs+3kV7/6FcbHx/H2t789+71qfx0KKabvHhgYQFtbW97PJUlCU1PTgr5G9dBmF9tSH4sLofF58cylzQ4MDMza/2S+N91jyqkc/c7HP/5xdHV15X1ofPWrX43Xve51WLlyJQ4dOoQbbrgBl1xyCXbu3AlRFMt6DsDczuOUU07B97//fWzatAnhcBhf/vKXsW3bNuzduxfLly+vSP8y39dj165deOaZZ3DbbbflfX+xX49STdcuIpEIkskkxsbGqmJ8nMvr89BDD+G2227DU089tQhHWP3mcg0PHz6M//u//8PVV1+Ne+65BwcPHsQHPvABmKaJG2+8cTEOu+rM5TpeddVVGB4exrnnngvOOSzLwvve9z7ccMMNi3HIdeenP/0pnnjiCTz66KOVPpSqd/jwYdxyyy24/vrrccMNN+DRRx/FP/3TP0FRFFxzzTWVPryq84lPfAKRSATr1q2DKIqwbRv/8R//gauvvrrSh7bo6uXzca3GtTLqMb4F1E+MK6OaY12Labr2NtlsnwUX2kztKhAITHn8bHGExVao/UxW6Ws82Vw+31fbOJRKpfDxj38cb37zmxEMBqd9XKnvr+lU2/nXK4ptzx/FwudurtduunEp87N6M9Pnf7peE2b77E/3qwtjse8PFkKpcxe1do9y++23IxAI4HWve92Mj6uW16hc4yrd45Sm1t7XhdRiP18P90LzWadYLedB91rVoV767Xr9fFiNa5KrwZJO7iBkNtdddx2eeeYZPPTQQ5U+lKpwyimn4KmnnkI4HMadd96Ja665Bjt27Kj0YdW86d5n73nPe7L/e+PGjejs7MRFF12EQ4cOYfXq1Yt9mFPk7qq5adMmbN26FStWrMDPf/5zeDyeCh7Z3N1222245JJL0NXVlf1etb8OhCxFND6TxfT5z38eP/3pT/HAAw9A07Ts99/0pjdl//fGjRuxadMmrF69Gg888AAuuuiiShzqFOeccw7OOeec7Nfbtm3D+vXr8Z3vfAef/exnK3hkc3fbbbdh48aN2LJlS973a+H1qEfRaBRvfetbceutt6KlpaXSh1OzHMdBW1sbvvvd70IURZx55pk4fvw4vvSlLy3Z5I65eOCBB3DTTTfhW9/6FrZu3YqDBw/igx/8ID772c/iU5/6VKUPr6b09vbigx/8IO677768sY8U5jgOzjrrLNx0000AgDPOOAPPPPMMvv3tb///7d15VBNX+wfwbyQGwQUUVNCqqIgoWHdxX6oo2KrltSoqSsGlrnVB3F5bLVal1ooVrUsbQFsr1Yp7XUDFfdcoLsUN9ecpihvu4sL9/eGblBAgBEkmwPdzDufozJ3kuZO5c+88mTvh5I4srFmzBqtWrcLvv/8ONzc3qFQqjB07FpUqVeL+KqAKal5LrTDmtwDmuAqr3OZDpL4WzKldFYRf+8uq/WQm9T4ubF6/fo3evXtDCIElS5bkWLagH19EhmIuPO+47/Tj9X/u8drf9ArL+KCwf3cRERGB/v376z2HmMtnxL6BipLCcLwX9PsUOdai/FYY2nVWCnpbN5ZiUgcgJXt7e1hYWODOnTtay+/cuQMHBweJojIv6v1QFPfRqFGjsGXLFuzZs0frSUsODg549eoVUlNTtcoXhX2iUCjg7OyMxo0bY86cOahfvz5+/PFHSfdJQW/H2R1nWfHw8AAAXLlyxRShGczW1hYuLi64cuVKgWwnN27cQFxcHAYPHpxjOXP/HIDcnbsdHByQkpKitf7Nmzd48OCBUT+jgt5mpVCU++KssH82rby0WQcHB73nH/Wy3L7m+3if8868efMQGhqKnTt34sMPP8yxbI0aNWBvb2+0/iE/zp/FixdHw4YNNTFKcX55n3o8e/YM0dHRuUr0G/vzMFR27aJMmTKwsrIym/7R0DiuXr2K69evo1u3bpDL5ZDL5Vi5ciU2bdoEuVyOq1evmip0s5GXz9LR0REuLi5aT2qrU6cObt++jVevXhk1XnOVl/341VdfYcCAARg8eDDq1asHHx8fzJ49G3PmzEF6eropwi40Tp48iZSUFDRq1EjTtvfu3YuFCxdCLpfj7du3UodoVhwdHVG3bl2tZXXq1MHNmzclisi8BQcHY/LkyfD19UW9evUwYMAAjBs3DnPmzJE6NJMzl/7/fRSmvJZaQc9vAYUrx6VmzrkuUzGkvWWW+VrQ1DK2q6zoyyOYUm7bT2ZS7+O8XN+bSz+kvnHzxo0biI2NzfGp3FnRd3xlx1zqX9gxt/1+mAvPu/fZd9n1S+p1hYm+6/+KFStyf/2Pvmt/jlfzl1TjA1PQ991FQRqj7N+/H4mJiQaPnQFpPqP87Fc5xjFMQTqus1PQzvOFZSz0PvcpmkM9ONYyH4XhvF2Yrw/N8Z5kc1CkJ3coFAo0btwYu3bt0ixLT0/Hrl27tJ64U5RVr14dDg4OWvvo8ePHOHr0aKHdR0IIjBo1CuvXr8fu3btRvXp1rfWNGzdG8eLFtfZJYmIibt68WWj3SXbS09ORlpYm6T4pqO1Y33GWFZVKBeBd8sgcPX36FFevXoWjo2OBbCeRkZGoUKECPv744xzLmfvnAOTu3N2iRQukpqbi5MmTmjK7d+9Genq65st9YyiobVZKRbEvzgr7Z2nkpc22aNFCqzwAxMbGasqb+pjO63ln7ty5mDlzJrZv344mTZrofZ9bt27h/v37Rusf8uP8+fbtWyQkJGhilOL88j71WLt2LdLS0uDn56f3fYz9eRhKX7swl/7R0DhcXV2RkJAAlUql+evevTs6dOgAlUqFKlWqmCx2c5GXz7JVq1a4cuWK1gSES5cuwdHREQqFwugxm6O87Mfnz5+jWDHtNJd6wowQwnjBFkIdO3bUadtNmjRB//79oVKptCYi0bs2nJiYqLXs0qVLqFatmkQRmbfs2mpRnIRlLv1/XhTGvJZaQc9vAYUrx6VmzrkuY8tLe8ss87WgqWVsV1nRd71kSrltP5lJvY/zcn1vDv2Q+sbNy5cvIy4uDnZ2dga/hr7jKzvmUP+igLntvGEuPO/yY9+1aNECCQkJWjdnqm8uz3xzf0Gn7/q/SZMm3F//o+/avyiPV/OblOMDU9D33UVBGqMolUo0btwY9evXN3hbU35GxuhXOcYxTEE6rrNTUM7zhX0sZMh9iuZQD461zEdBPm8XxetDc7gn2SyIIi46OlpYWlqKqKgoceHCBTF06FBha2srbt++LXVoJvPkyRNx+vRpcfr0aQFAzJ8/X5w+fVrcuHFDCCFEaGiosLW1FRs3bhRnz54VPXr0ENWrVxcvXryQOHLjGD58uLCxsRHx8fEiOTlZ8/f8+XNNmWHDhomqVauK3bt3ixMnTogWLVqIFi1aSBi18U2ePFns3btXJCUlibNnz4rJkycLmUwmdu7cKYSQdp8UxHas7zi7cuWKCAkJESdOnBBJSUli48aNokaNGqJt27YSR/6voKAgER8fL5KSksTBgwdFp06dhL29vUhJSRFCFKx28vbtW1G1alUxadIkreXm/Dnkx7nby8tLNGzYUBw9elQcOHBA1KpVS/Tt29fosRfENmts7Iv1Y/8sHX1tdsCAAWLy5Mma8gcPHhRyuVzMmzdPXLx4UUyfPl0UL15cJCQkaMqY+pg2tA6hoaFCoVCIP//8U+t4e/LkiRDiXZudMGGCOHz4sEhKShJxcXGiUaNGolatWuLly5dGqUNe6vHNN9+IHTt2iKtXr4qTJ08KX19fUaJECXH+/Hmtupr6/GJoPdRat24t+vTpo7Ncis9D33l78uTJYsCAAZry165dE9bW1iI4OFhcvHhRLF68WFhYWIjt27dryphL/5jXz0fN399f9OjRw0TRmidD9+HNmzdF6dKlxahRo0RiYqLYsmWLqFChgvj222+lqoJZMHQ/Tp8+XZQuXVqsXr1aXLt2TezcuVPUrFlT9O7dW6oqFCrt2rUTY8aMkToMs3Ts2DEhl8vFrFmzxOXLl8WqVauEtbW1+O2336QOzSz5+/uLypUriy1btoikpCQRExMj7O3txcSJE6UOTRLm0v8bqjDktdQKU35LiIKZ41IryLkuY8pNPiQv14LGpK9d5SWPYArZtR8hzGMf50cb+eijj0R4eLjm/8buh3KK+dWrV6J79+7igw8+ECqVSuv4TktLyzZmfceXIQpqP2xumNvOf8yF511+7Ls3b94Id3d30blzZ6FSqcT27dtF+fLlxZQpU6Soksllvv7n/nonN9f+RXW8aihzHx/kZ31y+92Fqcdo71MntUePHglra2uxZMmSLF/DnD6j/OpXa9euLWJiYjT/5xjHMFIf17lRGPIShWks9L73KZpLPTLjWMt4Cuu1aWG/PjTne5KlVuQndwghRHh4uKhatapQKBSiWbNm4siRI1KHZFJ79uwRAHT+/P39hRBCpKeni6+++kpUrFhRWFpaio4dO4rExERpgzairPYFABEZGakp8+LFCzFixAhRtmxZYW1tLXx8fERycrJ0QZtAYGCgqFatmlAoFKJ8+fKiY8eOmpOoENLvk4LWjvUdZzdv3hRt27YV5cqVE5aWlsLZ2VkEBweLR48eSRt4Bn369BGOjo5CoVCIypUriz59+ogrV65o1kt9TBhix44dAoDOuc2cP4f8OHffv39f9O3bV5QqVUqUKVNGBAQEaG5cNraC1maNjX2xfuyfpZVTm23Xrp3mWFVbs2aNcHFxEQqFQri5uYmtW7dqrZfimDakDtWqVcvyeJs+fboQQojnz5+Lzp07i/Lly4vixYuLatWqiSFDhpgkAWlIPcaOHaspW7FiRdG1a1dx6tQprdeT6vxi6DH1999/CwBa4081KT4Pfedtf39/0a5dO51tGjRoIBQKhahRo4bW+UvNXPpHQz+fjDi54x1D9+GhQ4eEh4eHsLS0FDVq1BCzZs0Sb968MXHU5seQ/fj69WsxY8YMUbNmTVGiRAlRpUoVMWLECPHw4UPTB14IcXJHzjZv3izc3d2FpaWlcHV1FcuXL5c6JLP1+PFjMWbMGFG1alVRokQJUaNGDfHf//5X6yaNosZc+n9DFIa8llphym8JUTBzXGoFPddlLLnJh+TlWtCY9LWrvOQRTCG79iOEeezj/Ggj1apV0+Q21IzZD+UUc1JSUrbH9549e7KNWd/xZaiC2A+bG+a28x9z4XmXX/vu+vXrwtvbW1hZWQl7e3sRFBQkXr9+beLaSCPz9T/317/0XfsX1fGqoQrC+CC/6pPb7y5MPUZ7nzqpLVu2TFhZWYnU1NQsX8OcPqP86hsyb8MxjuHMfexdGPIShWkslB/3KZpDPTLjWMt4Cuu1aWG/PjT3e5KlJBNCCBAREREREREREREREREREREREREREREREZEkikkdABERERERERERERERERERERERERERERERUVHGyR1EREREREREREREREREREREREREREREREQS4uQOIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiCXFyBxERERERERERERERERERERERERERERERkYQ4uYOIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiEhCnNxBREREREREREREREREREREREREREREREQkIU7uICIiIiIiIiIiIiIiIiIiIiIiIiIiIiIikhAndxAREREREREREREREREREREREREREREREUmIkzvIaG7fvo3Ro0ejRo0asLS0RJUqVdCtWzfs2rULAODk5ASZTAaZTAZra2vUq1cPv/zyi9ZrxMfHQyaTITU1Vev/ZcuWxcuXL7XKHj9+XPN6RGRc6raY3V+HDh1w/fp1yGQyVKhQAU+ePNHavkGDBpgxY4Y0wRMVYp9//jk+/fRTzb9lMhlCQ0O1ymzYsEGrr8zYnosVKwYbGxs0bNgQEydORHJycravn1Hm/vrt27cIDQ2Fq6srrKysUK5cOXh4eOj080SkX27G1AsWLNDZbsaMGWjQoIHO8lu3bkGhUMDd3T3L99u7dy8++ugjlCtXDtbW1qhVqxb8/f3x6tWr/KwWUZGV2/757du3CAsLQ7169VCiRAmULVsW3t7eOHjwoKbMkiVLYGtri//7v//Teq3Ro0fDxcUFz58/N25liAjAv+1aJpOhePHiqF69OiZOnKiVt1KvP3LkiNa2aWlpsLOzg0wmQ3x8vIkjJyIAOHz4MCwsLPDxxx/rrHv16hW+//57NGrUCCVLloSNjQ3q16+PadOm4Z9//tGUy3geyPjn5eVlyqoQFWmZ++OKFSvC09MTERERSE9P15TLfA195swZdO/eHRUqVECJEiXg5OSEPn36ICUlRYJaEBU9uRlLA/rzWRn735IlS6JWrVr4/PPPcfLkSVNUg6hIy+n7YplMpvV9sKurKywtLXH79m2t13j27Blq1qyJ8ePHay2/fv06ypQpg59//tkUVSEq8DL2qwqFAs7OzggJCYGfn1+O7dTJyQkA0L59e82yEiVKwMXFBXPmzIEQQue9srqWzu7aOKv3GTt2rNbrnT9/Hr1790b58uVhaWkJFxcXfP3118xxE+WzzO3Uzs4OXl5eOHv2rNShERUp2d17BeScq5oxY4be8bfa6tWrYWFhgZEjR2qWZezrs/pr3769kWtOhuDkDjKK69evo3Hjxti9eze+//57JCQkYPv27ejQoYPWCSMkJATJyck4d+4c/Pz8MGTIEGzbtk3v65cuXRrr16/XWqZUKlG1atV8rwsR6WrZsiWSk5N1/pYtWwaZTIYRI0Zoyj558gTz5s2TMFqioqtEiRL47rvv8PDhQ71lExMT8c8//+D48eOYNGkS4uLi4O7ujoSEBIPf95tvvkFYWBhmzpyJCxcuYM+ePRg6dKhm8gcR5U5ux9SGiIqKQu/evfH48WMcPXpUa92FCxfg5eWFJk2aYN++fUhISEB4eDgUCgXevn2bH1UiIujvn4UQ8PX1RUhICMaMGYOLFy8iPj4eVapUQfv27bFhwwYAwLBhw9CsWTMMGjRIs+2uXbuwZMkSREVFwdra2hTVISIAXl5eSE5OxrVr1xAWFoZly5Zh+vTpWmWqVKmCyMhIrWXr169HqVKlTBkqEWWiVCoxevRo7Nu3T2vCRlpaGjw9PTF79mx8/vnnmvHxwoULce/ePYSHh2u9jvo8kPFv9erVpq4OUZGmbofXr1/Htm3b0KFDB4wZMwaffPIJ3rx5o1P+7t276NixI8qVK4cdO3bg4sWLiIyMRKVKlfDs2TMJakBUNOVmLJ1TPkstMjISycnJOH/+PBYvXoynT5/Cw8MDK1euNEU1iIqsjOPfBQsWoEyZMlrLJkyYAAA4cOAAXrx4gc8++wwrVqzQeo2SJUsiMjIS4eHh2L9/P4B3+bGAgAC0atUKQ4YMMXm9iAoqdb96+fJlBAUFYcaMGahVq5ZWuwT+7TeTk5Nx/PhxzfZDhgxBcnIyEhMTMWXKFHz99ddYunSpzvtkdS39448/5vp9Mjpy5Ag8PDzw6tUrbN26FZcuXcKsWbMQFRUFT09PPnyMKJ9lzGHt2rULcrkcn3zyidRhERH056omTJig1dd+8MEHmvuvM/a/wLu+euLEiVi9erXmAQoxMTGacseOHQMAxMXFaZbFxMRIUm/KmlzqAKhwGjFiBGQyGY4dO4aSJUtqlru5uSEwMFDz/9KlS8PBwQEAMGnSJMydOxexsbHw9vbO8fX9/f0RERGBvn37AgBevHiB6OhofPnll5g5c6YRakREGSkUCk3bVbt48SImTJiAqVOnolevXrh+/TqAd08Onj9/PkaOHIkKFSpIEC1R0dWpUydcuXIFc+bMwdy5c3MsW6FCBdja2sLBwQEuLi7o0aMHGjZsiOHDh+PAgQMGve+mTZswYsQI9OrVS7Osfv36eaoDUVGW2zF1bgkhEBkZiZ9++gkffPABlEolPDw8NOt37twJBwcHrfNFzZo1+cRhonymr39es2YN/vzzT2zatAndunXTLF++fDnu37+PwYMHw9PTEyVLloRSqYS7uzuWLl2Kfv36ITAwEOPHj0fLli1NWSWiIs/S0lJzjVylShV06tQJsbGx+O677zRl/P39sXDhQixYsABWVlYAgIiICPj7+zOXRSSRp0+f4o8//sCJEydw+/ZtREVFYerUqQCAsLAwHDhwACdOnEDDhg0121StWhXt2rXTeXJpxvMAEUkjYzusXLkyGjVqhObNm6Njx46IiorC4MGDtcofPHgQjx49wi+//AK5/N3XpdWrV0eHDh1MHjtRUaZvLK0vn6Wmzm0D736lp3PnzvD398eoUaPQrVs3lC1b1nSVIipCMo6BbWxsIJPJshwXK5VK9OvXD+3atcOYMWMwadIkrfVt27bF6NGjERAQgDNnzuDnn3+GSqXCuXPnjF4HosIkY786fPhwrF+/Htu3b9eZOJmx38zI2tpaszwgIACLFi1CbGwshg8frimT3bW0jY0NbGxscvU+akIIDBo0CHXq1EFMTAyKFXv3jOpq1arBxcUFDRs2RFhYmM45g4jyLuN5wsHBAZMnT0abNm1w9+5dlC9fXuLoiIq23OSqMj4wzMLCQuv+a7WkpCQcOnQI69atw549exATE4N+/fqhXLlymjLqCR92dnbMa5sp/nIH5bsHDx5g+/btGDlypNZNaGq2trY6y9LT07Fu3To8fPgQCoVC73sMGDAA+/fvx82bNwEA69atg5OTExo1avTe8ROR4VJTU9GjRw+0b99e56aUvn37an7yk4hMy8LCArNnz0Z4eDhu3bpl0LZWVlYYNmwYDh48iJSUFIO2dXBwwO7du3H37l2DtiOif+VlTK3Pnj178Pz5c3Tq1Al+fn6Ijo7Wehqpg4MDkpOTsW/fvvcJnYj00Nc///7773BxcdGa2KEWFBSE+/fvIzY2FsC7G18WLFiA4OBg+Pn5oVSpUrxJnEhi586dw6FDh3TyW40bN4aTkxPWrVsHALh58yb27duHAQMGSBEmEeHdhEpXV1fUrl0bfn5+iIiI0EzaWL16NTw9PbUmdmSU8Sfuich8ffTRR6hfv36WTx50cHDAmzdvsH79ep0JW0QkjazG0vryWTkZN24cnjx5ormGJiJpPHnyBGvXroWfnx88PT3x6NEjzS90ZDRr1izI5XL4+flh6tSpCA8PR+XKlSWImKjwsLKyytMvXwghsH//fvz99986Oa6crqUNpVKpcOHCBYwfP14zsUOtfv366NSpE38Vk8iInj59it9++w3Ozs6ws7OTOhyiIi+/clWRkZH4+OOPYWNjAz8/PyiVynyMkkyFkzso3125cgVCCLi6uuotO2nSJJQqVQqWlpb47LPPULZsWZ2nJ2WlQoUK8Pb2RlRUFIB3TzrMy9OLiej9paeno1+/fpDL5Vi1apXOl9symQyhoaFYvnw5rl69KlGUREWXj48PGjRooPNEltxQ9+XqX+LJrfnz5+Pu3btwcHDAhx9+iGHDhmHbtm0Gvz9RUZaXMXXGv9mzZ+uUUyqV8PX1hYWFBdzd3VGjRg2sXbtWs75Xr17o27cv2rVrB0dHR/j4+GDRokV4/PhxvtaNiHLuny9duoQ6depkuZ16+aVLlzTLAgIC4O7ujs2bNyMyMhKWlpbGCZqIsrVlyxaUKlUKJUqUQL169ZCSkoLg4GCdcoGBgYiIiAAAREVFoWvXrnwaGpGElEol/Pz8AABeXl549OgR9u7dC+BdX1u7dm2t8j4+PprxduZfyVKfB/SNyYnI9FxdXbPMbTVv3hxTp05Fv379YG9vD29vb3z//fe4c+eO6YMkKsL0jaX15bNyktf8NhHlr+joaNSqVQtubm6wsLCAr69vljeYWVlZ4ccff8SGDRvQvn17zVidiAwnhEBcXBx27NiBjz76KNfb/fTTT5p7uNq2bYv09HR8+eWXWmVyupY2lDrPnVM+PGMunIjeX8YcVunSpbFp0yb88ccfOhOsiMj08iNXlZ6ejqioKE1f7evriwMHDiApKclYYZOR8KxM+c6QWWPBwcFQqVTYvXs3PDw8EBYWBmdn51xtGxgYiKioKFy7dg2HDx9G//798xoyEb2HqVOn4vDhw9i4cSNKly6dZZkuXbqgdevW+Oqrr0wcHREBwHfffYcVK1bg4sWLBm2n7tMNfSJp3bp1ce7cORw5cgSBgYFISUlBt27dcjWBk4jeycuYOuPfsGHDtMqkpqYiJiZG6wuxzE9psLCwQGRkJG7duoW5c+eicuXKmD17Ntzc3JCcnPz+lSIiLTn1z4acA86cOYNTp07B2to6y6ceEpHxdejQASqVCkePHoW/vz8CAgLQs2dPnXJ+fn44fPgwrl27hqioKD6ohEhCiYmJOHbsGPr27QsAkMvl6NOnT45PMfvpp5+gUqkQGBiI58+fa61TnwdyGpMTkTSEENnmtmbNmoXbt29j6dKlcHNzw9KlS+Hq6oqEhAQTR0lUdOU0ls5NPisnec1vE1H+ioiI0GnHa9euxZMnT3TKKpVKWFtbIyEhAY8ePTJlmESFQsZJk97e3ujTpw9mzJiR6+379+8PlUqFgwcPwtvbG//973+1Hm6Ql2vp3OAv6RGZTsYc1rFjx9ClSxd4e3vjxo0bUodGRHj/XFVsbCyePXuGrl27AgDs7e3h6empefAYFRyc3EH5rlatWpDJZPj777/1lrW3t4ezszPatGmDtWvX4ssvv8SFCxdy9T7e3t548eIFBg0ahG7duvHnwYgkEB0djXnz5mmeuJKT0NBQ/PHHHzh9+rSJoiMitbZt26JLly6YMmWKQdupbzZ1cnICAJQpUybLZHpqaiosLCxQsmRJzbJixYqhadOmGDt2LGJiYhAVFQWlUsnZ4ES5lJcxdca/cuXKaZX5/fff8fLlS3h4eEAul0Mul2PSpEk4cOCAzlOPKleujAEDBmDRokU4f/48Xr58iaVLl+Zr/Ygo+/7ZxcUl2wmZ6uUuLi4AgFevXmHgwIHo378/fvrpJ0ybNg2JiYnGDZyIdJQsWRLOzs6oX78+IiIicPTo0Sy/1Lazs8Mnn3yCQYMG4eXLl/D29pYgWiIC3t009ubNG1SqVEkzPl6yZAnWrVuHR48eoVatWjp9qqOjY5ZjbeDf80BOY3IiksbFixdRvXr1bNfb2dmhV69emDdvHi5evIhKlSph3rx5JoyQqGjLaSxtSD4rK+pr6JzOAURkXBcuXMCRI0cwceJETTtu3rw5nj9/jujoaK2yf/zxB7Zs2YJDhw6hdOnSGDdunERRExVc6pu2L1++jBcvXmDFihVa39/qY2NjA2dnZzRt2hRr1qzBokWLEBcXp1mv71raUOo8d075cHUZIsofGXNYTZs2xS+//IJnz57h559/ljo0Ivqf98lVKZVKPHjwAFZWVpq++q+//sKKFSuQnp5u5MgpP3FyB+W7cuXKoUuXLli8eDGePXumsz41NTXL7apUqYI+ffrk+sZTuVyOgQMHIj4+nk86JJKASqXCoEGDEBoaii5duugt36xZM/znP//B5MmTTRAdEWUWGhqKzZs34/Dhw7kq/+LFCyxfvhxt27ZF+fLlAQC1a9fG+fPnkZaWplX21KlTqF69OooXL57t69WtWxcAshwbEJGuvI6ps6NUKhEUFKT1JOEzZ86gTZs2OT6loWzZsnB0dGTbJTKSrPpnX19fXL58GZs3b9Yp/8MPP8DOzg6enp4AgJCQEDx48ABhYWHw9/eHp6cnAgICmJwjklCxYsUwdepUTJs2DS9evNBZHxgYiPj4eAwcOBAWFhYSREhEb968wcqVK/HDDz/ojI8rVaqE1atXo2/fvoiNjeVDSogKuN27dyMhISHLX9TKikKhQM2aNXkNTCSRzGPpvOaz1BYsWIAyZcqgU6dOJoieiLKiVCrRtm1bnDlzRqstjx8/XuuhCHfu3MHIkSPx7bffon79+oiKisLKlSuxbds2CaMnKnjUN21XrVoVcrn8vV6rVKlSGDNmDCZMmAAhRK6upQ3VoEEDuLq6IiwsTCenfebMGcTFxWl+JYSIjEMmk6FYsWJZ5rKJSHqG5Kru37+PjRs3Ijo6WquvPn36NB4+fIidO3eaIGLKL+83kiPKxuLFi9GqVSs0a9YMISEh+PDDD/HmzRvExsZiyZIl2c66HjNmDNzd3XHixAk0adJE7/vMnDkTwcHB/NUOIhO7d+8ePv30U7Rv3x5+fn64ffu21vrsblCZNWsW3Nzc3juRQESGq1evHvr374+FCxdmuT4lJQUvX77EkydPcPLkScydOxf37t1DTEyMpkz//v0REhKCgQMHYuLEibCxscG+ffuwYMECzJ07V1Pus88+Q6tWrdCyZUs4ODggKSkJU6ZMgYuLC1xdXY1eV6LCIq9j6sxUKhVOnTqFVatW6bTBvn37IiQkBN9++y2USiVUKhV8fHxQs2ZNvHz5EitXrsT58+cRHh5ujCoSFXlZ9c++vr5Yu3Yt/P398f3336Njx454/PgxFi9ejE2bNmHt2rUoWbIkjh8/ju+++w5bt26FjY0NAGDZsmVwd3dHWFgYgoKCpKoWUZHXq1cvBAcHY/HixZgwYYLWOi8vL9y9exdlypSRKDoi2rJlCx4+fIhBgwZp+lC1nj17QqlUYv/+/di6dSs6duyI6dOno02bNihbtiwuXbqEbdu26eS+0tLSdPJjcrkc9vb2Rq8PEb2jbodv377FnTt3sH37dsyZMweffPIJBg4cqFN+y5YtiI6Ohq+vL1xcXCCEwObNm/HXX38hMjJSghoQEaA9ls5NPkv9fVNqaipu376NtLQ0XLp0CcuWLcOGDRuwcuVK2NraSlATInr9+jV+/fVXhISEwN3dXWvd4MGDMX/+fJw/fx5ubm4YOnQo6tSpg7FjxwJ499DA4OBgDB06FOfOndMZtxORaXzxxReYOXMm1q1bB7lcrvdaetiwYQa9vkwmg1KphKenJ3r27IkpU6bAwcEBR48eRVBQEFq0aKE5LxBR/siYw3r48CEWLVqEp0+folu3bhJHRlS0PHr0CCqVSmtZQkICduzYkedc1a+//go7Ozv07t0bMplMa13Xrl2hVCrh5eWVn9UgI+Ivd5BR1KhRA6dOnUKHDh0QFBQEd3d3eHp6YteuXViyZEm229WtWxedO3fG119/nav3USgUsLe31zkZEZFxbd26FTdu3MBff/0FR0dHnb+mTZtmuZ2LiwsCAwPx8uVLE0dMRMC7p3tn9yTv2rVro1KlSmjcuDFCQ0PRqVMnnDt3TvOLGwBga2uL/fv34/Xr1+jevTsaNGiAhQsXYv78+fjiiy805bp06YLNmzejW7ducHFxgb+/P1xdXbFz505O7iIyQF7H1JkplUrUrVs3y8lVPj4+SElJwV9//YVmzZrh6dOnGDZsGNzc3NCuXTscOXIEGzZsQLt27fKzakSUQeb+WSaTYc2aNZg6dSrCwsJQu3ZttGnTBjdu3EB8fDw+/fRTpKWlwd/fHwEBAejcubNmW0dHR4SHh2PatGlITEyUojpEhHc3dI8aNQpz587VeZqSTCaDvb09FAqFRNERkVKpRKdOnbK8Qaxnz544ceIELl26hF27dmHSpEmIjIxE69atNTebtWrVChs2bNDabvv27Tr5sdatW5uoRkQE/NsOnZyc4OXlhT179mDhwoXYuHFjlg8jqlu3LqytrREUFIQGDRqgefPmWLNmDX755RcMGDBAghoQEfDvWHrKlClwcnLSm89SCwgIgKOjI1xdXTF8+HCUKlUKx44dQ79+/UwZPhFlsGnTJty/fx8+Pj466+rUqYM6depAqVRi5cqViIuLQ2RkJIoV+/cWpm+++Qa2trYYN26cKcMmogzKlSuHgQMHYsaMGbm6lj579qzB79GyZUscOXIEFhYW8Pb2hrOzM6ZMmQJ/f3/ExsbC0tIyP6pCRP+TMYfl4eGB48ePY+3atWjfvr3UoREVKfHx8WjYsKHWX2Rk5HvlqiIiIuDj45PlvdQ9e/bEpk2bcO/ePWNUh4xAJoQQUgdBRERERERERERERERERERERERERERERERUVPGXO4iIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiCTEyR1EREREREREREREREREREREREREREREREQS4uQOIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiCXFyBxERERERERERERERERERERERERERERERkYQ4uYOIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiEhCnNxBREREREREREREREREREREREREREREREQkIU7uICIiIiIiIiIiIiIiIiIiIiIiIiIiIiIikhAndxAREREREREREREREREREREREREREREREUmIkzuIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIgkxMkdREREREREREREREREREREREREREREREREEuLkDiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIglxcgcREREREREREREREREREREREREREREREZGE/h9aBXhFfHm7YQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 3250x250 with 13 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "### Code - Scatter plots\n",
    "sns.pairplot(data=boston_data, kind='reg',\n",
    "                  y_vars=['MEDV'],\n",
    "                  x_vars=['CRIM', 'ZN', 'INDUS', 'CHAS', 'NOX', 'RM', 'AGE', 'DIS', 'RAD', 'TAX', 'PTRATIO', 'B', 'LSTAT'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<font color = skyblue>I added linear regression lines to the figure. I can't see any clear dependecies I could be sure about. </font> "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "------\n",
    "\n",
    "We are about to compare two different regression models. We need to be sure that both are *the best models* among the other trained models. But to make sure that these selected models really are the best ones, we're going to use a combination of cross-validation and hyperparameter tuning right away. What's the other way to ensure that the chosen hyperparameters are appropriate for the given data?\n",
    "\n",
    "For the k-Nearest neighbors, we have previously get familiar with the optimization of the k value. Let's continue with this value to find the *optimal k* for the k-nearest neighbors regression too. With ridge regression, we have this hyperparameter called $\\lambda$ (read as 'lambda'). We'll use this parameter for model selection.\n",
    "\n",
    "\n",
    "To compare the models, let's use a loss function called <font color = darkorchid><b>mean absolute error (MAE)</b></font>. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "\n",
    "***Exercise 5 B)***\n",
    "\n",
    "Do you need to prepare the data a little? Explain your decision."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 124,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "CRIM       float64\n",
       "ZN         float64\n",
       "INDUS      float64\n",
       "CHAS       boolean\n",
       "NOX        float64\n",
       "RM         float64\n",
       "AGE        float64\n",
       "DIS        float64\n",
       "RAD        float64\n",
       "TAX        float64\n",
       "PTRATIO    float64\n",
       "B          float64\n",
       "LSTAT      float64\n",
       "MEDV       float64\n",
       "dtype: object"
      ]
     },
     "execution_count": 124,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "### Code - Preparation\n",
    "boston_data['CHAS'] = boston_data['CHAS'].astype('boolean')\n",
    "boston_data.dtypes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 125,
   "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>CRIM</th>\n",
       "      <th>ZN</th>\n",
       "      <th>INDUS</th>\n",
       "      <th>CHAS</th>\n",
       "      <th>NOX</th>\n",
       "      <th>RM</th>\n",
       "      <th>AGE</th>\n",
       "      <th>DIS</th>\n",
       "      <th>RAD</th>\n",
       "      <th>TAX</th>\n",
       "      <th>PTRATIO</th>\n",
       "      <th>B</th>\n",
       "      <th>LSTAT</th>\n",
       "      <th>MEDV</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-0.548749</td>\n",
       "      <td>0.320128</td>\n",
       "      <td>-1.303592</td>\n",
       "      <td>False</td>\n",
       "      <td>-0.137008</td>\n",
       "      <td>0.629691</td>\n",
       "      <td>-0.109981</td>\n",
       "      <td>0.108260</td>\n",
       "      <td>-0.971314</td>\n",
       "      <td>-0.668073</td>\n",
       "      <td>-1.593889</td>\n",
       "      <td>0.439740</td>\n",
       "      <td>-1.175095</td>\n",
       "      <td>24.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-0.544712</td>\n",
       "      <td>-0.483035</td>\n",
       "      <td>-0.603934</td>\n",
       "      <td>False</td>\n",
       "      <td>-0.726934</td>\n",
       "      <td>0.368889</td>\n",
       "      <td>0.378732</td>\n",
       "      <td>0.523617</td>\n",
       "      <td>-0.855688</td>\n",
       "      <td>-0.991540</td>\n",
       "      <td>-0.379700</td>\n",
       "      <td>0.439740</td>\n",
       "      <td>-0.569140</td>\n",
       "      <td>21.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-0.544716</td>\n",
       "      <td>-0.483035</td>\n",
       "      <td>-0.603934</td>\n",
       "      <td>False</td>\n",
       "      <td>-0.726934</td>\n",
       "      <td>1.662736</td>\n",
       "      <td>-0.256239</td>\n",
       "      <td>0.523617</td>\n",
       "      <td>-0.855688</td>\n",
       "      <td>-0.991540</td>\n",
       "      <td>-0.379700</td>\n",
       "      <td>0.394323</td>\n",
       "      <td>-1.313474</td>\n",
       "      <td>34.7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.543739</td>\n",
       "      <td>-0.483035</td>\n",
       "      <td>-1.322701</td>\n",
       "      <td>False</td>\n",
       "      <td>-0.820980</td>\n",
       "      <td>1.346048</td>\n",
       "      <td>-0.802028</td>\n",
       "      <td>1.042209</td>\n",
       "      <td>-0.740061</td>\n",
       "      <td>-1.111342</td>\n",
       "      <td>0.057407</td>\n",
       "      <td>0.414409</td>\n",
       "      <td>-1.472246</td>\n",
       "      <td>33.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-0.536684</td>\n",
       "      <td>-0.483035</td>\n",
       "      <td>-1.322701</td>\n",
       "      <td>False</td>\n",
       "      <td>-0.820980</td>\n",
       "      <td>1.598382</td>\n",
       "      <td>-0.502379</td>\n",
       "      <td>1.042209</td>\n",
       "      <td>-0.740061</td>\n",
       "      <td>-1.111342</td>\n",
       "      <td>0.057407</td>\n",
       "      <td>0.439740</td>\n",
       "      <td>-1.124113</td>\n",
       "      <td>36.2</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",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>462</th>\n",
       "      <td>-0.537919</td>\n",
       "      <td>-0.483035</td>\n",
       "      <td>0.110423</td>\n",
       "      <td>False</td>\n",
       "      <td>0.162229</td>\n",
       "      <td>0.660174</td>\n",
       "      <td>0.029141</td>\n",
       "      <td>-0.654829</td>\n",
       "      <td>-0.971314</td>\n",
       "      <td>-0.805846</td>\n",
       "      <td>1.174461</td>\n",
       "      <td>0.384950</td>\n",
       "      <td>-0.491939</td>\n",
       "      <td>22.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>463</th>\n",
       "      <td>-0.541258</td>\n",
       "      <td>-0.483035</td>\n",
       "      <td>0.110423</td>\n",
       "      <td>False</td>\n",
       "      <td>0.162229</td>\n",
       "      <td>-0.140859</td>\n",
       "      <td>0.300252</td>\n",
       "      <td>-0.745325</td>\n",
       "      <td>-0.971314</td>\n",
       "      <td>-0.805846</td>\n",
       "      <td>1.174461</td>\n",
       "      <td>0.439740</td>\n",
       "      <td>-0.577880</td>\n",
       "      <td>20.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>464</th>\n",
       "      <td>-0.538279</td>\n",
       "      <td>-0.483035</td>\n",
       "      <td>0.110423</td>\n",
       "      <td>False</td>\n",
       "      <td>0.162229</td>\n",
       "      <td>1.308791</td>\n",
       "      <td>0.810369</td>\n",
       "      <td>-0.802152</td>\n",
       "      <td>-0.971314</td>\n",
       "      <td>-0.805846</td>\n",
       "      <td>1.174461</td>\n",
       "      <td>0.439740</td>\n",
       "      <td>-1.078958</td>\n",
       "      <td>23.9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>465</th>\n",
       "      <td>-0.528887</td>\n",
       "      <td>-0.483035</td>\n",
       "      <td>0.110423</td>\n",
       "      <td>False</td>\n",
       "      <td>0.162229</td>\n",
       "      <td>1.000571</td>\n",
       "      <td>0.749726</td>\n",
       "      <td>-0.697307</td>\n",
       "      <td>-0.971314</td>\n",
       "      <td>-0.805846</td>\n",
       "      <td>1.174461</td>\n",
       "      <td>0.401242</td>\n",
       "      <td>-0.956601</td>\n",
       "      <td>22.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>466</th>\n",
       "      <td>-0.540846</td>\n",
       "      <td>-0.483035</td>\n",
       "      <td>0.110423</td>\n",
       "      <td>False</td>\n",
       "      <td>0.162229</td>\n",
       "      <td>-0.293276</td>\n",
       "      <td>0.446510</td>\n",
       "      <td>-0.642327</td>\n",
       "      <td>-0.971314</td>\n",
       "      <td>-0.805846</td>\n",
       "      <td>1.174461</td>\n",
       "      <td>0.439740</td>\n",
       "      <td>-0.752674</td>\n",
       "      <td>11.9</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>467 rows × 14 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "         CRIM        ZN     INDUS   CHAS       NOX        RM       AGE  \\\n",
       "0   -0.548749  0.320128 -1.303592  False -0.137008  0.629691 -0.109981   \n",
       "1   -0.544712 -0.483035 -0.603934  False -0.726934  0.368889  0.378732   \n",
       "2   -0.544716 -0.483035 -0.603934  False -0.726934  1.662736 -0.256239   \n",
       "3   -0.543739 -0.483035 -1.322701  False -0.820980  1.346048 -0.802028   \n",
       "4   -0.536684 -0.483035 -1.322701  False -0.820980  1.598382 -0.502379   \n",
       "..        ...       ...       ...    ...       ...       ...       ...   \n",
       "462 -0.537919 -0.483035  0.110423  False  0.162229  0.660174  0.029141   \n",
       "463 -0.541258 -0.483035  0.110423  False  0.162229 -0.140859  0.300252   \n",
       "464 -0.538279 -0.483035  0.110423  False  0.162229  1.308791  0.810369   \n",
       "465 -0.528887 -0.483035  0.110423  False  0.162229  1.000571  0.749726   \n",
       "466 -0.540846 -0.483035  0.110423  False  0.162229 -0.293276  0.446510   \n",
       "\n",
       "          DIS       RAD       TAX   PTRATIO         B     LSTAT  MEDV  \n",
       "0    0.108260 -0.971314 -0.668073 -1.593889  0.439740 -1.175095  24.0  \n",
       "1    0.523617 -0.855688 -0.991540 -0.379700  0.439740 -0.569140  21.6  \n",
       "2    0.523617 -0.855688 -0.991540 -0.379700  0.394323 -1.313474  34.7  \n",
       "3    1.042209 -0.740061 -1.111342  0.057407  0.414409 -1.472246  33.4  \n",
       "4    1.042209 -0.740061 -1.111342  0.057407  0.439740 -1.124113  36.2  \n",
       "..        ...       ...       ...       ...       ...       ...   ...  \n",
       "462 -0.654829 -0.971314 -0.805846  1.174461  0.384950 -0.491939  22.4  \n",
       "463 -0.745325 -0.971314 -0.805846  1.174461  0.439740 -0.577880  20.6  \n",
       "464 -0.802152 -0.971314 -0.805846  1.174461  0.439740 -1.078958  23.9  \n",
       "465 -0.697307 -0.971314 -0.805846  1.174461  0.401242 -0.956601  22.0  \n",
       "466 -0.642327 -0.971314 -0.805846  1.174461  0.439740 -0.752674  11.9  \n",
       "\n",
       "[467 rows x 14 columns]"
      ]
     },
     "execution_count": 125,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#initialize \n",
    "numeric_features = ['CRIM', 'ZN', 'INDUS', 'NOX', 'RM', 'AGE', 'DIS', 'RAD', 'TAX', 'PTRATIO', 'B', 'LSTAT']\n",
    "boston_data_numeric = boston_data[numeric_features]\n",
    "\n",
    "#scale\n",
    "scaler = preprocessing.StandardScaler().fit(boston_data_numeric)\n",
    "boston_data_numeric_scaled = scaler.transform(boston_data_numeric)\n",
    "\n",
    "# make dataframe again\n",
    "boston_data_numeric_scaled = pd.DataFrame(boston_data_numeric_scaled)\n",
    "# insert column names back\n",
    "boston_data_numeric_scaled.columns = numeric_features\n",
    "\n",
    "# change feature_train numeric values to standardized \n",
    "boston_data_copy = boston_data.copy()\n",
    "boston_data_copy[numeric_features] = boston_data_numeric_scaled\n",
    "boston_data_standardized = boston_data_copy\n",
    "\n",
    "# display\n",
    "boston_data_standardized"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 126,
   "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>CRIM</th>\n",
       "      <th>ZN</th>\n",
       "      <th>INDUS</th>\n",
       "      <th>CHAS</th>\n",
       "      <th>NOX</th>\n",
       "      <th>RM</th>\n",
       "      <th>AGE</th>\n",
       "      <th>DIS</th>\n",
       "      <th>RAD</th>\n",
       "      <th>TAX</th>\n",
       "      <th>PTRATIO</th>\n",
       "      <th>B</th>\n",
       "      <th>LSTAT</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-0.548749</td>\n",
       "      <td>0.320128</td>\n",
       "      <td>-1.303592</td>\n",
       "      <td>False</td>\n",
       "      <td>-0.137008</td>\n",
       "      <td>0.629691</td>\n",
       "      <td>-0.109981</td>\n",
       "      <td>0.108260</td>\n",
       "      <td>-0.971314</td>\n",
       "      <td>-0.668073</td>\n",
       "      <td>-1.593889</td>\n",
       "      <td>0.439740</td>\n",
       "      <td>-1.175095</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-0.544712</td>\n",
       "      <td>-0.483035</td>\n",
       "      <td>-0.603934</td>\n",
       "      <td>False</td>\n",
       "      <td>-0.726934</td>\n",
       "      <td>0.368889</td>\n",
       "      <td>0.378732</td>\n",
       "      <td>0.523617</td>\n",
       "      <td>-0.855688</td>\n",
       "      <td>-0.991540</td>\n",
       "      <td>-0.379700</td>\n",
       "      <td>0.439740</td>\n",
       "      <td>-0.569140</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-0.544716</td>\n",
       "      <td>-0.483035</td>\n",
       "      <td>-0.603934</td>\n",
       "      <td>False</td>\n",
       "      <td>-0.726934</td>\n",
       "      <td>1.662736</td>\n",
       "      <td>-0.256239</td>\n",
       "      <td>0.523617</td>\n",
       "      <td>-0.855688</td>\n",
       "      <td>-0.991540</td>\n",
       "      <td>-0.379700</td>\n",
       "      <td>0.394323</td>\n",
       "      <td>-1.313474</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.543739</td>\n",
       "      <td>-0.483035</td>\n",
       "      <td>-1.322701</td>\n",
       "      <td>False</td>\n",
       "      <td>-0.820980</td>\n",
       "      <td>1.346048</td>\n",
       "      <td>-0.802028</td>\n",
       "      <td>1.042209</td>\n",
       "      <td>-0.740061</td>\n",
       "      <td>-1.111342</td>\n",
       "      <td>0.057407</td>\n",
       "      <td>0.414409</td>\n",
       "      <td>-1.472246</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-0.536684</td>\n",
       "      <td>-0.483035</td>\n",
       "      <td>-1.322701</td>\n",
       "      <td>False</td>\n",
       "      <td>-0.820980</td>\n",
       "      <td>1.598382</td>\n",
       "      <td>-0.502379</td>\n",
       "      <td>1.042209</td>\n",
       "      <td>-0.740061</td>\n",
       "      <td>-1.111342</td>\n",
       "      <td>0.057407</td>\n",
       "      <td>0.439740</td>\n",
       "      <td>-1.124113</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>462</th>\n",
       "      <td>-0.537919</td>\n",
       "      <td>-0.483035</td>\n",
       "      <td>0.110423</td>\n",
       "      <td>False</td>\n",
       "      <td>0.162229</td>\n",
       "      <td>0.660174</td>\n",
       "      <td>0.029141</td>\n",
       "      <td>-0.654829</td>\n",
       "      <td>-0.971314</td>\n",
       "      <td>-0.805846</td>\n",
       "      <td>1.174461</td>\n",
       "      <td>0.384950</td>\n",
       "      <td>-0.491939</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>463</th>\n",
       "      <td>-0.541258</td>\n",
       "      <td>-0.483035</td>\n",
       "      <td>0.110423</td>\n",
       "      <td>False</td>\n",
       "      <td>0.162229</td>\n",
       "      <td>-0.140859</td>\n",
       "      <td>0.300252</td>\n",
       "      <td>-0.745325</td>\n",
       "      <td>-0.971314</td>\n",
       "      <td>-0.805846</td>\n",
       "      <td>1.174461</td>\n",
       "      <td>0.439740</td>\n",
       "      <td>-0.577880</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>464</th>\n",
       "      <td>-0.538279</td>\n",
       "      <td>-0.483035</td>\n",
       "      <td>0.110423</td>\n",
       "      <td>False</td>\n",
       "      <td>0.162229</td>\n",
       "      <td>1.308791</td>\n",
       "      <td>0.810369</td>\n",
       "      <td>-0.802152</td>\n",
       "      <td>-0.971314</td>\n",
       "      <td>-0.805846</td>\n",
       "      <td>1.174461</td>\n",
       "      <td>0.439740</td>\n",
       "      <td>-1.078958</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>465</th>\n",
       "      <td>-0.528887</td>\n",
       "      <td>-0.483035</td>\n",
       "      <td>0.110423</td>\n",
       "      <td>False</td>\n",
       "      <td>0.162229</td>\n",
       "      <td>1.000571</td>\n",
       "      <td>0.749726</td>\n",
       "      <td>-0.697307</td>\n",
       "      <td>-0.971314</td>\n",
       "      <td>-0.805846</td>\n",
       "      <td>1.174461</td>\n",
       "      <td>0.401242</td>\n",
       "      <td>-0.956601</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>466</th>\n",
       "      <td>-0.540846</td>\n",
       "      <td>-0.483035</td>\n",
       "      <td>0.110423</td>\n",
       "      <td>False</td>\n",
       "      <td>0.162229</td>\n",
       "      <td>-0.293276</td>\n",
       "      <td>0.446510</td>\n",
       "      <td>-0.642327</td>\n",
       "      <td>-0.971314</td>\n",
       "      <td>-0.805846</td>\n",
       "      <td>1.174461</td>\n",
       "      <td>0.439740</td>\n",
       "      <td>-0.752674</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>467 rows × 13 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "         CRIM        ZN     INDUS   CHAS       NOX        RM       AGE  \\\n",
       "0   -0.548749  0.320128 -1.303592  False -0.137008  0.629691 -0.109981   \n",
       "1   -0.544712 -0.483035 -0.603934  False -0.726934  0.368889  0.378732   \n",
       "2   -0.544716 -0.483035 -0.603934  False -0.726934  1.662736 -0.256239   \n",
       "3   -0.543739 -0.483035 -1.322701  False -0.820980  1.346048 -0.802028   \n",
       "4   -0.536684 -0.483035 -1.322701  False -0.820980  1.598382 -0.502379   \n",
       "..        ...       ...       ...    ...       ...       ...       ...   \n",
       "462 -0.537919 -0.483035  0.110423  False  0.162229  0.660174  0.029141   \n",
       "463 -0.541258 -0.483035  0.110423  False  0.162229 -0.140859  0.300252   \n",
       "464 -0.538279 -0.483035  0.110423  False  0.162229  1.308791  0.810369   \n",
       "465 -0.528887 -0.483035  0.110423  False  0.162229  1.000571  0.749726   \n",
       "466 -0.540846 -0.483035  0.110423  False  0.162229 -0.293276  0.446510   \n",
       "\n",
       "          DIS       RAD       TAX   PTRATIO         B     LSTAT  \n",
       "0    0.108260 -0.971314 -0.668073 -1.593889  0.439740 -1.175095  \n",
       "1    0.523617 -0.855688 -0.991540 -0.379700  0.439740 -0.569140  \n",
       "2    0.523617 -0.855688 -0.991540 -0.379700  0.394323 -1.313474  \n",
       "3    1.042209 -0.740061 -1.111342  0.057407  0.414409 -1.472246  \n",
       "4    1.042209 -0.740061 -1.111342  0.057407  0.439740 -1.124113  \n",
       "..        ...       ...       ...       ...       ...       ...  \n",
       "462 -0.654829 -0.971314 -0.805846  1.174461  0.384950 -0.491939  \n",
       "463 -0.745325 -0.971314 -0.805846  1.174461  0.439740 -0.577880  \n",
       "464 -0.802152 -0.971314 -0.805846  1.174461  0.439740 -1.078958  \n",
       "465 -0.697307 -0.971314 -0.805846  1.174461  0.401242 -0.956601  \n",
       "466 -0.642327 -0.971314 -0.805846  1.174461  0.439740 -0.752674  \n",
       "\n",
       "[467 rows x 13 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>MEDV</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>24.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>21.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>34.7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>33.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>36.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>462</th>\n",
       "      <td>22.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>463</th>\n",
       "      <td>20.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>464</th>\n",
       "      <td>23.9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>465</th>\n",
       "      <td>22.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>466</th>\n",
       "      <td>11.9</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>467 rows × 1 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     MEDV\n",
       "0    24.0\n",
       "1    21.6\n",
       "2    34.7\n",
       "3    33.4\n",
       "4    36.2\n",
       "..    ...\n",
       "462  22.4\n",
       "463  20.6\n",
       "464  23.9\n",
       "465  22.0\n",
       "466  11.9\n",
       "\n",
       "[467 rows x 1 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#split to feature/result\n",
    "features = ['CRIM', 'ZN', 'INDUS', 'CHAS', 'NOX', 'RM', 'AGE', 'DIS', 'RAD', 'TAX', 'PTRATIO', 'B', 'LSTAT']\n",
    "result = ['MEDV']\n",
    "boston_features = boston_data_standardized[features]\n",
    "boston_results = boston_data_standardized[result]\n",
    "display(boston_features)\n",
    "display(boston_results)\n",
    "\n",
    "# make splits\n",
    "boston_feature_train, boston_feature_test, boston_label_train, boston_label_test = train_test_split(boston_data_standardized, boston_results,\n",
    "                               random_state=x, \n",
    "                               train_size=0.8, \n",
    "                               shuffle=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "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>CRIM</th>\n",
       "      <th>ZN</th>\n",
       "      <th>INDUS</th>\n",
       "      <th>CHAS</th>\n",
       "      <th>NOX</th>\n",
       "      <th>RM</th>\n",
       "      <th>AGE</th>\n",
       "      <th>DIS</th>\n",
       "      <th>RAD</th>\n",
       "      <th>TAX</th>\n",
       "      <th>PTRATIO</th>\n",
       "      <th>B</th>\n",
       "      <th>LSTAT</th>\n",
       "      <th>MEDV</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-0.548749</td>\n",
       "      <td>0.320128</td>\n",
       "      <td>-1.303592</td>\n",
       "      <td>False</td>\n",
       "      <td>-0.137008</td>\n",
       "      <td>0.629691</td>\n",
       "      <td>-0.109981</td>\n",
       "      <td>0.108260</td>\n",
       "      <td>-0.971314</td>\n",
       "      <td>-0.668073</td>\n",
       "      <td>-1.593889</td>\n",
       "      <td>0.439740</td>\n",
       "      <td>-1.175095</td>\n",
       "      <td>24.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-0.544712</td>\n",
       "      <td>-0.483035</td>\n",
       "      <td>-0.603934</td>\n",
       "      <td>False</td>\n",
       "      <td>-0.726934</td>\n",
       "      <td>0.368889</td>\n",
       "      <td>0.378732</td>\n",
       "      <td>0.523617</td>\n",
       "      <td>-0.855688</td>\n",
       "      <td>-0.991540</td>\n",
       "      <td>-0.379700</td>\n",
       "      <td>0.439740</td>\n",
       "      <td>-0.569140</td>\n",
       "      <td>21.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-0.544716</td>\n",
       "      <td>-0.483035</td>\n",
       "      <td>-0.603934</td>\n",
       "      <td>False</td>\n",
       "      <td>-0.726934</td>\n",
       "      <td>1.662736</td>\n",
       "      <td>-0.256239</td>\n",
       "      <td>0.523617</td>\n",
       "      <td>-0.855688</td>\n",
       "      <td>-0.991540</td>\n",
       "      <td>-0.379700</td>\n",
       "      <td>0.394323</td>\n",
       "      <td>-1.313474</td>\n",
       "      <td>34.7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.543739</td>\n",
       "      <td>-0.483035</td>\n",
       "      <td>-1.322701</td>\n",
       "      <td>False</td>\n",
       "      <td>-0.820980</td>\n",
       "      <td>1.346048</td>\n",
       "      <td>-0.802028</td>\n",
       "      <td>1.042209</td>\n",
       "      <td>-0.740061</td>\n",
       "      <td>-1.111342</td>\n",
       "      <td>0.057407</td>\n",
       "      <td>0.414409</td>\n",
       "      <td>-1.472246</td>\n",
       "      <td>33.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-0.536684</td>\n",
       "      <td>-0.483035</td>\n",
       "      <td>-1.322701</td>\n",
       "      <td>False</td>\n",
       "      <td>-0.820980</td>\n",
       "      <td>1.598382</td>\n",
       "      <td>-0.502379</td>\n",
       "      <td>1.042209</td>\n",
       "      <td>-0.740061</td>\n",
       "      <td>-1.111342</td>\n",
       "      <td>0.057407</td>\n",
       "      <td>0.439740</td>\n",
       "      <td>-1.124113</td>\n",
       "      <td>36.2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       CRIM        ZN     INDUS   CHAS       NOX        RM       AGE  \\\n",
       "0 -0.548749  0.320128 -1.303592  False -0.137008  0.629691 -0.109981   \n",
       "1 -0.544712 -0.483035 -0.603934  False -0.726934  0.368889  0.378732   \n",
       "2 -0.544716 -0.483035 -0.603934  False -0.726934  1.662736 -0.256239   \n",
       "3 -0.543739 -0.483035 -1.322701  False -0.820980  1.346048 -0.802028   \n",
       "4 -0.536684 -0.483035 -1.322701  False -0.820980  1.598382 -0.502379   \n",
       "\n",
       "        DIS       RAD       TAX   PTRATIO         B     LSTAT  MEDV  \n",
       "0  0.108260 -0.971314 -0.668073 -1.593889  0.439740 -1.175095  24.0  \n",
       "1  0.523617 -0.855688 -0.991540 -0.379700  0.439740 -0.569140  21.6  \n",
       "2  0.523617 -0.855688 -0.991540 -0.379700  0.394323 -1.313474  34.7  \n",
       "3  1.042209 -0.740061 -1.111342  0.057407  0.414409 -1.472246  33.4  \n",
       "4  1.042209 -0.740061 -1.111342  0.057407  0.439740 -1.124113  36.2  "
      ]
     },
     "execution_count": 127,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#reset index\n",
    "boston_feature_train = boston_feature_train.reset_index()\n",
    "boston_feature_test = boston_feature_test.reset_index()\n",
    "boston_label_train = boston_label_train.reset_index()\n",
    "boston_label_test = boston_label_test.reset_index()\n",
    "\n",
    "feature_train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 128,
   "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>MEDV</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>32.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>29.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>14.3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>19.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>23.3</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   MEDV\n",
       "0  32.2\n",
       "1  29.0\n",
       "2  14.3\n",
       "3  19.8\n",
       "4  23.3"
      ]
     },
     "execution_count": 128,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#drop level_0 and index\n",
    "boston_feature_train = boston_feature_train.drop(['index'], axis=1)\n",
    "boston_feature_test = boston_feature_test.drop(['index'], axis=1)\n",
    "boston_label_train = boston_label_train.drop(['index'], axis=1)\n",
    "boston_label_test = boston_label_test.drop(['index'], axis=1)\n",
    "\n",
    "boston_feature_train.head()\n",
    "boston_feature_test.head()\n",
    "boston_label_train.head()\n",
    "boston_label_test.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<font color = skyblue>CHAS seemed to be boolean so I'll make it boolean. Rest of the features I standardized </font> "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "------\n",
    "\n",
    "***Exercise 5 C)***\n",
    "\n",
    "Then, train a ridge regression model  and a k-NN regression model. For hyperparameters, use $\\lambda=2^{-10}...2^{10}$ and $k=1...30$. Once again, use leave-one-out cross validation.\n",
    "\n",
    "- What are the total performances of each model and how they compare with each other? \n",
    "- What does the MAE tell you about the performances of your models and in general?\n",
    "\n",
    "*note: In a `sklearn.linear_model.Ridge` class, lambda is called alpha so don't get confused*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "metadata": {},
   "outputs": [],
   "source": [
    "### Code - Ridge regression and k-NN regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 130,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.824468085106383, 0.7244680851063826, 0.6684397163120568, 0.6372340425531914, 0.6344680851063829, 0.6255319148936171, 0.6232522796352581, 0.6789893617021274, 0.6722222222222223, 0.696595744680851, 0.7413926499032881, 0.7611702127659569, 0.8054828150572829, 0.8525835866261392, 0.8824822695035457, 0.8863031914893617, 0.919023779724656, 0.9324468085106381, 0.9722284434490481, 1.0218617021276597, 1.050709219858156, 1.1191972920696327, 1.1635522664199816, 1.227171985815603, 1.2744680851063832, 1.3287234042553195, 1.378250591016549, 1.4229483282674773, 1.4791636096845202, 1.5276241134751776]\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjcAAAGwCAYAAABVdURTAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/NK7nSAAAACXBIWXMAAA9hAAAPYQGoP6dpAABeIElEQVR4nO3dd3hUZcLG4d+kE1LoCSX0TkIIoQiIgoAURbCBFOkWRESwwUoR9BNXBdGVooIgFooIYkEQ6QICofcuCZCEmt5nzvcHa9ZIy4QkJ5k893XNteTMzJknw8g8e8573tdiGIaBiIiIiINwMjuAiIiISG5SuRERERGHonIjIiIiDkXlRkRERByKyo2IiIg4FJUbERERcSgqNyIiIuJQXMwOkN9sNhvnz5/H29sbi8VidhwRERHJBsMwiI+Pp0KFCjg53frYTJErN+fPnycgIMDsGCIiIpIDERERVKpU6ZaPKXLlxtvbG7j25vj4+JicRkRERLIjLi6OgICAzO/xWyly5eavU1E+Pj4qNyIiIoVMdoaUaECxiIiIOBSVGxEREXEoKjciIiLiUFRuRERExKGo3IiIiIhDUbkRERERh6JyIyIiIg5F5UZEREQcisqNiIiIOBSVGxEREXEoKjciIiLiUFRuRERExKGo3IiIiEiu2XDsIinpVlMzqNyIiIhIrlhzOJqBc7fTd/Y2ktIyTMuhciMiIiJ37MC5WIYv2I3NgBplvSjm6mxaFpUbERERuSPnY5IZNG8HSWlW7q5ZhrceDsRisZiWR+VGREREciw+JZ1B83ZwIT6V2n5ezOjbGFdnc+uFyo2IiIjkSLrVxnNf7+JIVDxlvd2ZO7AZPh6uZsdSuRERERH7GYbB+OUH2HT8EsVcnfm8f1MqlihmdixA5UZERERyYNaGUyzYHoHFAh/1CiGokq/ZkTKp3IiIiIhdftp3nn+vPALA+Afr06G+n8mJslK5ERERkWzbeeYKoxbvBWBAy6oMbFXN5ETXU7kRERGRbDlzOZGn5u8kLcNG+3p+jHuwvtmRbkjlRkRERG7ramIaA+fu4EpiGkEVffmoVyOcncyby+ZWVG5ERETkllIzrDzz5U5OXUqkYolizBnQBE83F7Nj3ZTKjYiIiNyUYRi8umQf2/+8gre7C3MHNqWct4fZsW5J5UZERERuaurqYyzfcx4XJwsz+4ZS28/b7Ei3pXIjIiIiN7Q4LIL/rD0BwNsPB3F3rTImJ8oelRsRERG5zuYTl/jX0v0ADGtbgx5NA0xOlH0qNyIiIpLF8eh4nv1qJxk2g67BFXipQx2zI9lF5UZEREQyXYhPYcDcHcSnZNC0aknee6whTgX0ku+bUbkRERERANIybDzz5U7OxSRTrUxxPn2yCR6uzmbHspvKjYiIiADw/q9H2R0eg4+HC3MHNKVkcTezI+WIyo2IiIiw9kg0n248BcB7jwdTtUxxkxPlnMqNiIhIEXc+JjlzMcyBrarSsYG/yYnujMqNiIhIEZZhtfHCgt3EJKUTVNGX0Z3rmh3pjqnciIiIFGFTVx8j7MxVvN1d+Lh3CO4uhW8A8T+p3IiIiBRRG45dZMb6kwC882hDqpQuvONs/k7lRkREpAiKjkth1KI9APS9qzIPNCxvbqBcpHIjIiJSxFhtBiMW7uZyYhr1yvsw9oH6ZkfKVSo3IiIiRcxHa47zx6krFHdzZnrvkEI5Ud+tqNyIiIgUIVtOXOKjtccBePuRIKqX9TI5Ue5TuRERESkiLsanMmLRHgwDejYJoFujimZHyhMqNyIiIkWAzWYwavEeLsanUtvPizceamB2pDyjciMiIlIEzNxwkk3HL1HM1ZnpvRtTzM2xxtn8ncqNiIiIg9t++gpTfj0KwKRuDajl521yorylciMiIuLAriSm8cKC3dgMeKRxRR5vEmB2pDynciMiIuKgbDaDlxbvISouhRpli/Nmt0CzI+ULlRsREREHNfv3U6w7ehF3Fyc+7t2Y4u4uZkfKFyo3IiIiDmhX+FXeXXltnM2Erg2oV97H5ET5R+VGRETEwcQmpTP8m91k2Ay6BlegVzPHH2fzdyo3IiIiDsQwDF5espdzMclULe3J2w8HYrFYzI6Vr1RuREREHMgHvx1n9aFo3JyvjbPx9nA1O1K+U7kRERFxEHN+P81Ha66tGzWxWwMCK/qanMgcKjciIiIOYHFYBG/+dAiAl++vTa9mlU1OZB6VGxERkULul/2RjP5uHwBP31OdYW1rmpzIXHaVG6vVysaNG4mJicmjOCIiImKPjccu8sLCazMQP9E0gDGd6xa5AcT/ZFe5cXZ25v777+fq1at5lUdERESyaeeZKzzz5U7SrQYPNCzP/z0cVOSLDeTgtFRgYCCnTp3KlRffuHEjXbt2pUKFClgsFr7//vtsP3fz5s24uLjQqFGjXMkiIiJSmBw6H8eAuTtITrdyb+2yfNCjEc5OKjaQg3Lz1ltv8fLLL/PTTz8RGRlJXFxclps9EhMTCQ4OZvr06XY9LyYmhn79+tGuXTu7niciIuIITl1MoN/n24hPyaBp1ZLM6huKm4uG0f7FYhiGYc8TnJz+9+b9/dCXYRhYLBasVmvOglgsLFu2jO7du9/2sU888QS1atXC2dmZ77//nj179tz0sampqaSmpmb+HBcXR0BAALGxsfj4FJ2pqEVExDGcj0nmsZlbOB+bQoMKPix4+i58isBcNnFxcfj6+mbr+9vuFbTWrVuX42C5Ye7cuZw6dYqvvvqKt95667aPnzx5MhMnTsyHZCIiInnrUkIqfeds43xsCtXLFueLQc2KRLGxl93l5t57782LHNly/PhxRo8ezaZNm3BxyV70MWPGMGrUqMyf/zpyIyIiUpjEJqfTb852Tl1MpGKJYnw1uDllvNzNjlUg5Wjt85iYGObMmcPhw4cBaNCgAYMGDcLXN+9mQrRarfTu3ZuJEydSu3btbD/P3d0dd3f95YuISOGVnGZlyBc7OBQZRxkvN74a0pwKJYqZHavAsnvMTVhYGB07dqRYsWI0a9YMgB07dpCcnMyvv/5K48aNcxbkNmNuYmJiKFmyJM7OzpnbbDYbhmHg7OzMr7/+yn333Xfb17HnnJ2IiIjZ0jJsPDU/jA3HLuLt4cKip1tQv0LR+/7K0zE3I0eO5KGHHuKzzz7LPDWUkZHBkCFDePHFF9m4cWPOUt+Gj48P+/fvz7JtxowZrF27liVLllCtWrU8eV0RERGzWG0GIxftYcOxixRzdWbewKZFstjYy+5yExYWlqXYALi4uPDqq6/SpEkTu/aVkJDAiRMnMn8+ffo0e/bsoVSpUlSuXJkxY8Zw7tw55s+fj5OTE4GBgVmeX65cOTw8PK7bLiIiUtgZhsG/lu7n5/2RuDpb+LRfKKFVSpkdq1Cw+6J4Hx8fwsPDr9seERGBt7e3XfsKCwsjJCSEkJAQAEaNGkVISAjjx48HIDIy8oavJSIi4sgMw+DtFYdZFBaBkwU+eiKE1rXKmh2r0LB7zM0LL7zAsmXLeP/992nZsiVwbbbgV155hUcffZRp06blRc5cozE3IiJS0E1fd4L3Vh0F4N3HGtKjia7yzdMxN++//z4Wi4V+/fqRkZEBgKurK0OHDuWdd97JWWIREREB4Pvd5zKLzbgH66vY5IBdR26sViubN28mKCgId3d3Tp48CUCNGjXw9PTMs5C5SUduRESkoNp55gq9Pt1GmtXGM/dUZ0yXemZHKjDy7MjNX6uCHz58mGrVqhEUFHRHQUVEROSaiCtJPD1/J2lWG/fX9+O1TnXNjlRomboquIiIiEBcSjqD5u3gcmIaDSr4MO2JRjhphe8cM3VVcBERkaIuw2pj+De7OX4hAT8fd+b0b4qnW44WEJD/svvd69KlCwAPPfRQrq4KLiIiUhS9+dOhzEn6Zvdrir+vh9mRCr1Ctyq4iIiIo/hiy598sfUMAB/0bERQpbxbo7EosavcpKenM2nSJGbNmkWtWrXyKpOIiIjDW3/0AhN/PAjAa53q0inQ3+REjsOuMTeurq7s27cvr7KIiIgUCcei4xn+zW5sBjweWoln761udiSHYveA4r59+zJnzpy8yCIiIuLwLiWkMmjeDuJTM2hWrRT/93BQljGscufsHnOTkZHB559/zm+//UZoaCjFixfPcv/UqVNzLZyIiIgjSUm38vT8MM5eTaZKaU8+6RuKm4vdxxnkNuwuNwcOHKBx48YAHDt2LMt9ap4iIiI3ZhgGr323j13hMfh4uPD5gKaULO5mdiyHpKulRERE8sF/1p5g+Z7zuDhZmNk3lBplvcyO5LByfCzsxIkTrFq1iuTkZOBaIxUREZHr/bj3PFNXXzvb8Wb3QFrVLGNyIsdmd7m5fPky7dq1o3bt2nTp0oXIyEgABg8ezEsvvZTrAUVERAqzXeFXeenbvQA81boavZpVNjmR47O73IwcORJXV1fCw8OzrATes2dPVq5cmavhRERECrOzV5N4en4YaRk22tcrx+jOWuU7P9g95ubXX39l1apVVKpUKcv2WrVqcebMmVwLJiIiUpjFp6Qz5IswLiWkUa+8Dx8+EYKzFsPMF3YfuUlMTMxyxOYvV65cwd3dPVdCiYiIFGZWm8ELC3ZzJCqest7uzOnfhOLuWgwzv9hdblq3bs38+fMzf7ZYLNhsNt59913atm2bq+FEREQKo3dXHWHd0Yu4uzgxu18TKpQoZnakIsXuGvnuu+/Srl07wsLCSEtL49VXX+XgwYNcuXKFzZs350VGERGRQmPVwSg+2XAKgCk9ggkOKGFuoCLI7iM3gYGBHDt2jLvvvptu3bqRmJjII488wu7du6lRo0ZeZBQRESkU/ryUyMuLr10ZNfjuajzYsILJiYqmHJ0A9PX15fXXX8/tLCIiIoVWcpqVZ7/aSXxqBk2qlGR057pmRyqytKCFiIjIHTIMg3HLD3AkKp4yXm5M79MYV2d9xZpF77yIiMgdWrQjgiU7z+JkgY96heDn42F2pCJN5UZEROQOHDgXy/gfDgLwcsc6tKyhpRXMpnIjIiKSQzFJaTz71c7MGYifvUcX1hQEOSo3GRkZ/Pbbb3zyySfEx8cDcP78eRISEnI1nIiISEFlsxmMWryXs1eTqVzKkymPN8JJMxAXCHZfLXXmzBk6depEeHg4qampdOjQAW9vb/7973+TmprKrFmz8iKniIhIgTJzw0nWHrmAm4sTM/o0xtfT1exI8l92H7kZMWIETZo04erVqxQr9r8ZFx9++GHWrFmTq+FEREQKos0nLjHl16MAvNUtkMCKviYnkr+z+8jNpk2b2LJlC25ublm2V61alXPnzuVaMBERkYIoKjaFFxbsxmZAjyaV6NE0wOxI8g92H7mx2WxYrdbrtp89exZvb+9cCSUiIlIQpWXYeO7rnVxOTKN+eR8mdQs0O5LcgN3l5v7772fatGmZP1ssFhISEpgwYQJdunTJzWwiIiIFyuRfDrMrPAZvDxdm9m2Mh6uz2ZHkBuw+LTVlyhQ6duxI/fr1SUlJoXfv3hw/fpwyZcqwYMGCvMgoIiJiup/2nWfu5j8BmNqjEVVKFzc3kNyU3eWmUqVK7N27l0WLFrF3714SEhIYPHgwffr0yTLAWERExFGcuJDAa0v2ATC0TQ061PczOZHcisUwDMOeJ2zcuJGWLVvi4pK1F2VkZLBlyxbuueeeXA2Y2+Li4vD19SU2NhYfHx+z44iISAGXmJpB9+mbOX4hgRbVS/Pl4Ga4aN2ofGfP97fdfztt27blypUr122PjY2lbdu29u5ORESkwDIMgzFL93P8QgLlvN35qFeIik0hYPffkGEYWCzXz8B4+fJlihfX+UcREXEcX/5xhh/2nsfZycL0Po0p6+1udiTJhmyPuXnkkUeAa1dHDRgwAHf3//0FW61W9u3bR8uWLXM/oYiIiAl2h1/lzZ8OATCmc12aVi1lciLJrmyXG1/fa7MvGoaBt7d3lsHDbm5u3HXXXTz11FO5n1BERCSfRcel8NzXu0i3GnQJ8mfw3dXMjiR2yHa5mTt3LnBtJuKXX35Zp6BERMQhxaek0//z7UTGplC9bHH+/WjDGw7HkILL7kvBJ0yYkBc5RERETJeWYWPoV7s4EhVPGS93vhjYDG8PLYhZ2NhdbqpVq3bLBnvq1Kk7CiQiImIGwzAY/d0+fj9xCU83Z+YOaEpAKU+zY0kO2F1uXnzxxSw/p6ens3v3blauXMkrr7ySW7lERETy1fu/HmXp7nM4O1mY0acxQZW00ndhZXe5GTFixA23T58+nbCwsDsOJCIikt+++uMM09edBGDyI0G0qVPO5ERyJ3JtJqLOnTvz3Xff5dbuRERE8sXqQ9GMX34AgJHta9OjSYDJieRO5Vq5WbJkCaVKaQ4AEREpPHaFX2X4gl3YDHiiaQAvtKtpdiTJBXaflgoJCckyoNgwDKKiorh48SIzZszI1XAiIiJ55fSlRIZ8EUZKuo22dcryVvdAXfLtIOwuN927d8/ys5OTE2XLlqVNmzbUrVs3t3KJiIjkmUsJqfT/fDtXEtNoWMmXj3s31ppRDkTz3IiISJGSlJbB4Hk7CL+SROVSnszp35Ti7nZ/HUoBlq2/zbi4uGzv8HbLkIuIiJglw2rj+W92s/dsLCU9XZk3sKkWw3RA2So3JUqUuO15yL9WC7darbkSTEREJDcZhsG45QdYe+QC7i5OzO7flOplvcyOJXkgW+Vm3bp1eZ1DREQkT3289gQLtkfgZIH/9AohtEpJsyNJHslWubn33nvzOoeIiEie+TYsgimrjwEw8aEG3N/A3+REkpdyNIIqJiaGOXPmcPjwYQAaNGjAoEGD8PXVVNUiIlKwbDh2kTFL9wMwtE0NnmxR1dxAkufsvu4tLCyMGjVq8MEHH3DlyhWuXLnC1KlTqVGjBrt27cqLjCIiIjly4Fwsz321kwybQfdGFXjl/jpmR5J8YDEMw7DnCa1bt6ZmzZp89tlnuLhcO/CTkZHBkCFDOHXqFBs3bsyToLklLi4OX19fYmNjdWWXiIgDO3s1iYdnbOFifCqtapZm7oBmuLloLpvCyp7vb7tPS4WFhWUpNgAuLi68+uqrNGnSxP60IiIiucxqMxixcA8X41Op6+/NzL6hKjZFiN1/0z4+PoSHh1+3PSIiAm9v71wJJSIicie+2PInO89cxcvdhdn9m+Dj4Wp2JMlHdpebnj17MnjwYBYtWkRERAQREREsXLiQIUOG0KtXr7zIKCIikm1nLify7qojAIzpUpdKJT1NTiT5ze7TUu+//z4Wi4V+/fqRkZEBgKurK0OHDuWdd97J9YAiIiLZZbMZvPbdPlLSbbSoXppeTSubHUlMYPeA4r8kJSVx8uRJAGrUqIGnZ+FoxhpQLCLiuL784wzjvj9AMVdnVr14D5VLF47vJrk9e76/czy6ytPTk6CgIKpUqcKvv/6aOeeNiIiIGc5eTeKdFde+i17tVEfFpgizu9z06NGDjz/+GIDk5GSaNGlCjx49aNiwId99912uBxQREbkdwzAYs3Q/iWlWmlQpSX9N1Fek2V1uNm7cSOvWrQFYtmwZhmEQExPDRx99xFtvvWX3vrp27UqFChWwWCx8//33t3z80qVL6dChA2XLlsXHx4cWLVqwatUqe38FERFxMN+GnWXT8Uu4uzjx7mMNcXK69WLP4tjsLjexsbGUKlUKgJUrV/Loo4/i6enJAw88wPHjx+3aV2JiIsHBwUyfPj1bj9+4cSMdOnRgxYoV7Ny5k7Zt29K1a1d2795t768hIiIOIio2hTd/PgTAqA61tdK32H+1VEBAAFu3bqVUqVKsXLmShQsXAnD16lU8PDzs2lfnzp3p3Llzth8/bdq0LD+//fbbLF++nB9//JGQkBC7XltERAo/wzB4fdl+4lMyCA4owZDW1c2OJAWA3eXmxRdfpE+fPnh5eVGlShXatGkDXDuqEhQUlNv5bslmsxEfH595JOlGUlNTSU1Nzfw5Li4uP6KJiEg+WL7nPGuOXMDN2Yn3HmuIs05HCTkoN8899xzNmjUjIiKCDh064OR07cxW9erV7R5zc6fef/99EhIS6NGjx00fM3nyZCZOnJiPqUREJD9ciE/hjR8PAvBCu5rU9tMs+XJNjue5gWuHAwEsljtvyhaLhWXLltG9e/dsPf6bb77hqaeeYvny5bRv3/6mj7vRkZuAgADNcyMiUsgN/WonvxyIon55H5Y/3wpXZ60d5cjyfJ6bOXPmEBgYiIeHBx4eHgQGBjJ79uwchc2Jv5Z7WLx48S2LDYC7uzs+Pj5ZbiIiUrit2B/JLweicHGy8N7jDVVsJAu7T0uNHz+eqVOnMnz4cFq0aAHA1q1bGTlyJOHh4UyaNCnXQ/7dggULGDRoEAsXLuSBBx7I09cSEZGC50piGuO+PwDAc21q0KCCr8mJpKCxu9zMnDmTzz77LMsimQ899BANGzZk+PDhdpWbhIQETpw4kfnz6dOn2bNnD6VKlaJy5cqMGTOGc+fOMX/+fODaqaj+/fvz4Ycf0rx5c6KiogAoVqwYvr76cIuIFAUTfzzI5cQ0avt5Mey+mmbHkQLI7uN46enpNGnS5LrtoaGhmQtpZldYWBghISGZl3GPGjWKkJAQxo8fD0BkZCTh4eGZj//000/JyMhg2LBhlC9fPvM2YsQIe38NEREphFYfimb5nvM4WeC9x4Jxd3E2O5IUQHYPKB4+fDiurq5MnTo1y/aXX36Z5OTkbE/IZxYtnCkiUjjFJqXT4YMNXIhP5Zl7qzOmcz2zI0k+suf7O1unpUaNGpX5Z4vFwuzZs/n111+56667ANi2bRvh4eH069fvDmKLiIjc3Js/H+JCfCrVyxRnZPvaZseRAixb5eafyxuEhoYCcPLkSQDKlClDmTJlOHjwYC7HExERgfVHL7Bk51ksFnj3sYZ4uOp0lNxctsrNunXr8jqHiIjIDcWnpDNm6X4ABrSsSpOqN5+VXgRyOM+NiIhIfpn8yxEiY1OoXMqTVzrWMTuOFAJ2XwoO165yWrx4MeHh4aSlpWW5b+nSpbkSTEREZMuJS3yz7dpVs+88GoSnW46+tqSIsfvIzcKFC2nZsiWHDx9m2bJlpKenc/DgQdauXau5ZkREJNdcTUzjtaX7AOjTvDIta5QxOZEUFnaXm7fffpsPPviAH3/8ETc3Nz788EOOHDlCjx49qFy5cl5kFBGRIiY+JZ3+c7cTcSWZiiWKMaaLLvuW7LO73Jw8eTJz2QM3NzcSExOxWCyMHDmSTz/9NNcDiohI0ZKcZmXwF2HsOxtLqeJuzBvYFC93nY6S7LO73JQsWZL4+HgAKlasyIED19b3iImJISkpKXfTiYhIkZKWYWPo1zvZfvoK3u4uzB/UjFp+3mbHkkLG7ip8zz33sHr1aoKCgnj88ccZMWIEa9euZfXq1bRr1y4vMoqISBFgtRmMXLSH9Ucv4uHqxOcDmxJYUWM5xX52l5uPP/6YlJQUAF5//XVcXV3ZsmULjz76KGPHjs31gCIi4vhsNoMxS/fx8/5IXJ0tfPJkE5pqPhvJIbvXlirstLaUiEjBYhgGk346xNzNf+Jkgem9G9M5qLzZsaSAsef7W5P4iYiIqT747ThzN/8JwLuPBavYyB1TuREREdN8tvEUH605DsDEhxrwWGglkxOJI1C5ERERUyzYHs7/rTgMwCsd69C/ZVVzA4nDULkREZF898Pe8/xr2bXFMJ+5tzrPtalhciJxJCo3IiKSr9YcjmbUoj0YxrVlFUZ3qovFYjE7ljgQuy8FT0xM5J133mHNmjVcuHABm82W5f5Tp07lWjgREXEsW05eYujXu8iwGXRvVIE3uwWq2Eius7vcDBkyhA0bNvDkk09Svnx5fShFRCRbdodfZcgXYaRl2OhQ34/3Hg/GyUnfIZL77C43v/zyCz///DOtWrXKizwiIuKADkfGMWDuDpLSrLSqWZr/9ArB1VkjIyRv5GhtqVKlNGukiIhkz+lLiTw5Zzuxyek0rlyCT59sgoers9mxxIHZXW7efPNNxo8fr0UyRUTkts7FJNN39jYuJaRSr7wPcwc2o7hW+JY8ZvcnbMqUKZw8eRI/Pz+qVq2Kq6trlvt37dqVa+FERKTwstkMhn+zi3MxyVQvW5wvBzfDt5jr7Z8ocofsLjfdu3fPgxgiIuJoFodFsCs8huJuzswf1IwyXu5mR5Iiwu5yM2HChLzIISIiDuRKYhrvrDwCwMgOtalU0tPkRFKUaKi6iIjkund+OUxMUjp1/b0ZoGUVJJ/ZfeTGarXywQcfsHjxYsLDw0lLS8ty/5UrV3ItnIiIFD5hf15hcdhZAP7v4UBcdMm35DO7P3ETJ05k6tSp9OzZk9jYWEaNGsUjjzyCk5MTb7zxRh5EFBGRwiLdamPs9wcA6NkkgNAqmjpE8p/d5ebrr7/ms88+46WXXsLFxYVevXoxe/Zsxo8fzx9//JEXGUVEpJD4YsufHImKp6SnK6M71zU7jhRRdpebqKgogoKCAPDy8iI2NhaABx98kJ9//jl304mISKERGZvMB6uPATC6c11KFnczOZEUVXaXm0qVKhEZGQlAjRo1+PXXXwHYsWMH7u66zE9EpKh686dDJKZZCa1SksdDA8yOI0WY3eXm4YcfZs2aNQAMHz6ccePGUatWLfr168egQYNyPaCIiBR8649eYMX+KJydLLzVPVALYoqp7L5a6p133sn8c8+ePalcuTJbt26lVq1adO3aNVfDiYhIwZeSbmXCDwcBGNCyKvXK+5icSIq6O17go0WLFrRo0SI3soiISCE0c/1JzlxOws/HnZEdapsdRyRnk/h9+eWXtGrVigoVKnDmzBkApk2bxvLly3M1nIiIFGynLyUyc/1JAMY/2AAvLYopBYDd5WbmzJmMGjWKLl26EBMTg9VqBaBEiRJMmzYtt/OJiEgBZRgG45cfIM1q457aZekS5G92JBEgB+XmP//5D5999hmvv/46zs7OmdubNGnC/v37czWciIgUXCv2R7Hp+CXcXJyY9FADLBYNIpaCwe5yc/r0aUJCQq7b7u7uTmJiYq6EEhGRgi0+JZ1JP10bRPxcmxpULVPc5EQi/2N3ualWrRp79uy5bvvKlSupV69ebmQSEZECbtpvx4mOS6VKaU+evbeG2XFEsrB75NeoUaMYNmwYKSkpGIbB9u3bWbBgAZMnT2b27Nl5kVFERAqQQ+fjmLflTwAmdQvEw9X51k8QyWd2l5shQ4ZQrFgxxo4dS1JSEr1796ZChQp8+OGHPPHEE3mRUURECgibzWDs9/ux2gweCCrPvbXLmh1J5Do5umavT58+9OnTh6SkJBISEihXrlxu5xIRkQLo250R7AqPobibM+MerG92HJEbuqMJCTw9PfH09MytLCIiUoBdSUxj8i9HABjZoTb+vh4mJxK5MbvLzeXLlxk/fjzr1q3jwoUL2Gy2LPdfuXIl18KJiEjB8e9fjhCTlE5df28GtKxqdhyRm7K73Dz55JOcOHGCwYMH4+fnp3kNRESKgJ1nrrAoLAKA/3s4EBfnHE1wL5Iv7C43mzZt4vfffyc4ODgv8oiISAGTYbXx+rIDAPRsEkBolVImJxK5Nburd926dUlOTs6LLCIiUgDN2/InR6LiKenpyujOdc2OI3JbdpebGTNm8Prrr7NhwwYuX75MXFxclpuIiDiOTccv8sHqYwCM7lyXksXdTE4kcnt2n5YqUaIEcXFx3HfffVm2G4aBxWLJXEhTREQKr+Q0K+/8cpgvtp4BoFm1UjweGmByKpHssbvc9OnTB1dXV7755hsNKBYRcUC7w6/y0uK9nLp0bb3Afi2qMLpzXZyc9O+9FA52l5sDBw6we/du6tSpkxd5RETEJOlWG/9Zc5zp609itRn4+3jw7mMNuUezEEshY3e5adKkCRERESo3IiIO5Hh0PCMX7+HAuWtjJ7s1qsCkhwLx9XQ1OZmI/ewuN8OHD2fEiBG88sorBAUF4eqa9YPfsGHDXAsnIiJ5y2Yz+Hzzad5ddZS0DBslPF15q3sgDzasYHY0kRyzGIZh2PMEJ6frL7CyWCyFZkBxXFwcvr6+xMbG4uPjY3YcERHTnL2axMvf7uWPU9dmlm9Tpyz/frQhfj5aVkEKHnu+v+0+cnP69OkcBxMREfMZhsGSnWeZ+OMhElIzKObqzNgH69G7WWVdJCIOwe5yU6VKlbzIISIi+eBSQir/WrqfXw9FAxBapSRTHg+mapniJicTyT13tCq4iIgUHr8ejOJfy/ZzKSENV2cLIzvU5pl7auCsS7zFwajciIg4uPiUdCb9eIhvd54FoI6fN1N7BtOggq/JyUTyhsqNiIgDS0rLoNdnf3DgXBwWCzx9T3VGdaiNu4uz2dFE8ozKjYiIg7LZDF5ceG3umlLF3ZjVN5Rm1bSitzg+lRsREQf17qqj/HooGjdnJz7rF0poFRUbKRqyVW5KliyZ7csDr1y5ckeBRETkzn0bFsGsDScBePexhio2UqRkq9xMmzYt88+XL1/mrbfeomPHjrRo0QKArVu3smrVKsaNG5cnIUVEJPu2nbrMv5btB2D4fTXpHlLR5EQi+cvuGYofffRR2rZty/PPP59l+8cff8xvv/3G999/n+19bdy4kffee4+dO3cSGRnJsmXL6N69+y2fs379ekaNGsXBgwcJCAhg7NixDBgwINuvqRmKRcSR/Xkpke4zNhOTlM4DQeX5T68QreYtDsGe7+/r11K4jVWrVtGpU6frtnfq1InffvvNrn0lJiYSHBzM9OnTs/X406dP88ADD9C2bVv27NnDiy++yJAhQ1i1apVdrysi4ohik9IZ9MUOYpLSCa7ky/uPB6vYSJFk94Di0qVLs3z5cl566aUs25cvX07p0qXt2lfnzp3p3Llzth8/a9YsqlWrxpQpUwCoV68ev//+Ox988AEdO3a067VFRBxJutXGsG92cepiIuV9PfisXxOKuelybyma7C43EydOZMiQIaxfv57mzZsDsG3bNlauXMlnn32W6wH/buvWrbRv3z7Lto4dO/Liiy/e9DmpqamkpqZm/hwXF5dX8URETGEYBhN+OMjvJy7h6ebM7P5NKKfFL6UIs/u01IABA9i8eTM+Pj4sXbqUpUuX4uPjw++//27X2JeciIqKws/PL8s2Pz8/4uLiSE5OvuFzJk+ejK+vb+YtICAgTzOKiOS3uZv/5Jtt4Vgs8OETIZp5WIq8HM1z07x5c77++uvczpInxowZw6hRozJ/jouLU8EREYex7sgF3vr5EABjOtelQ32/2zxDxPHZfeQG4OTJk4wdO5bevXtz4cIFAH755RcOHjyYq+H+yd/fn+jo6CzboqOj8fHxoVixYjd8jru7Oz4+PlluIiKO4EhUHMMX7MZmQM8mATzVurrZkUQKBLvLzYYNGwgKCmLbtm189913JCQkALB3714mTJiQ6wH/rkWLFqxZsybLttWrV2fOtyMiUlRcjE9l8LwwElIzuKt6Kd7sHpjtyVZFHJ3d5Wb06NG89dZbrF69Gjc3t8zt9913H3/88Ydd+0pISGDPnj3s2bMHuHap9549ewgPDweunVLq169f5uOfffZZTp06xauvvsqRI0eYMWMGixcvZuTIkfb+GiIihVZKupWnvwzjXEwy1coUZ1bfUNxccnQgXsQh2f1fw/79+3n44Yev216uXDkuXbpk177CwsIICQkhJCQEgFGjRhESEsL48eMBiIyMzCw6ANWqVePnn39m9erVBAcHM2XKFGbPnq3LwEWkyDAMg1eX7GN3eAy+xVyZ078JJTzdbv9EkSLE7gHFJUqUIDIykmrVqmXZvnv3bipWtG+K7zZt2nCrCZLnzZt3w+fs3r3brtcREXEUH605wQ97z+PiZGFmn8ZUL+tldiSRAsfuIzdPPPEEr732GlFRUVgsFmw2G5s3b+bll1/OcgpJRERy1w97z/PBb8cAeLN7IC1rljE5kUjBZHe5efvtt6lbty4BAQEkJCRQv3597rnnHlq2bMnYsWPzIqOISJG3K/wqL3+7F4Ahd1ejV7PKJicSKbjsXjjzLxEREezfv5+EhARCQkKoVatWbmfLE1o4U0QKmzOXE3l05lYuJaTSrm45Pu3XBGetGSVFTJ4unDlp0iSSkpIICAigS5cu9OjRg1q1apGcnMykSZNyHFpERK53LDqex2ddKzZ1/b35sFeIio3Ibdh95MbZ2ZnIyEjKlSuXZfvly5cpV64cVqs1VwPmNh25EZHCYv/ZWPp9vo2rSenU9fdm/uBmlPPWmlFSNNnz/W331VKGYdxwoqi9e/dSqlQpe3cnIiI3EPbnFQbO3UF8agbBlXz5YlAzXfItkk3ZLjclS5bEYrFgsVioXbt2loJjtVpJSEjg2WefzZOQIiJFye/HL/HU/DCS0600q1aKOf2b4O3hanYskUIj2+Vm2rRpGIbBoEGDmDhxIr6+/1t11s3NjapVq2oZBBGRO7T6UDTDvt5FmtXGPbXL8knfUIq5OZsdS6RQyXa56d+/P3BtluCWLVvi6qr/FyEikpt+2HuekYv2YLUZdGzgx0e9QnB3UbERsZfdY27uvffezD+npKSQlpaW5X4N0hURsd+iHeGMXrofw4CHQyry3mMNcXHWelEiOWH3fzlJSUk8//zzlCtXjuLFi1OyZMksNxERsc/nv5/mte+uFZs+zSsz5fFgFRuRO2D3fz2vvPIKa9euZebMmbi7uzN79mwmTpxIhQoVmD9/fl5kFBFxSIZh8PHa40z66RAAT99Tnbe6B+KkeWxE7ojdp6V+/PFH5s+fT5s2bRg4cCCtW7emZs2aVKlSha+//po+ffrkRU4REYdiGAb/XnmUWRtOAjCyfW1eaFfzhlNtiIh97D5yc+XKFapXrw5cG19z5coVAO6++242btyYu+lERByQzWYw4YeDmcXm9S71GNG+loqNSC6xu9xUr16d06dPA1C3bl0WL14MXDuiU6JEiVwNJyLiaDKsNl79bh/zt57BYoG3Hw7iqXuqmx1LxKHYXW4GDhzI3r3XVqYdPXo006dPx8PDg5EjR/LKK6/kekAREUeRlmFjxMI9LNl5FmcnCx/0aETv5lrdWyS35XhV8L+cOXOGnTt3UrNmTRo2bJhbufKM1pYSETOkpFsZ+tVO1h29iJuzEx/1CqFToL/ZsUQKjTxdW+qfqlSpQpUqVe50NyIiDscwDA6ci2PlwUh+3hfJn5eT8HB14pMnm3Bv7bJmxxNxWNkqNx999FG2d/jCCy/kOIyISGFntRnsPHOVlQeiWHUwinMxyZn3+Xi4MLt/U5pV0yLDInkpW6elqlWrluXnixcvkpSUlDmAOCYmBk9PT8qVK8epU6fyJGhu0WkpEcltaRk2/jh1mZUHo/j1YDSXElIz7yvm6kzbumXp2MCf++qW0wKYIjmU66el/ro6CuCbb75hxowZzJkzhzp16gBw9OhRnnrqKZ555pk7iC0iUngkp1nZePwiqw5E8dvhaOJSMjLv8/FwoX19Pzo18Oee2mXxcNX6UCL5ye4BxTVq1GDJkiWEhIRk2b5z504ee+yxLEWoINKRGxHJqbiUdNYducDKA1GsP3qR5HRr5n1lvNy5v4EfnQP9uat6aVy1fIJIrsrTAcWRkZFkZGRct91qtRIdHW3v7kRECjzDMJi35U/e+eUIqRm2zO0VSxSjU6A/nQL9aVy5JM5aNkGkQLC73LRr145nnnmG2bNn07hxY+DaUZuhQ4fSvn37XA8oImKmpLQMxizdz/I95wGoUbY4nQPL0ynQnwYVfDSrsEgBZHe5+fzzz+nfvz9NmjTB1fXawLiMjAw6duzI7Nmzcz2giIhZTl9K5Nkvd3I0Oh4XJwv/6lKPga2qqtCIFHB2l5uyZcuyYsUKjh07xpEjR4BryzDUrl0718OJiJjl14NRvLR4L/GpGZT1dmd678a6hFukkMjxJH61a9dWoRERh2O1GUxdfZTp664tatm0akmm925MOR8Pk5OJSHZlq9yMGjWKN998k+LFizNq1KhbPnbq1Km5EkxEJL9dSUxjxMLdbDp+CYCBraryry71dOWTSCGTrXKze/du0tPTM/98MzoPLSKF1d6IGJ77ehfnYpIp5urMO48G0a1RRbNjiUgO3PHCmYWN5rkRkX9auD2c8csPkma1Ua1McWb1DaWOv7fZsUTkb/J14UwRkcIqJd3KhOUHWRQWAUCH+n5M6RGMj5ZIECnUslVuHnnkkWzvcOnSpTkOIyKSXyKuJPHc17vYfy4WJwu83LEOz95TAydNxCdS6GWr3Pj6+uZ1DhGRfLPh2EVGLNxNTFI6JT1d+U+vxtxdq4zZsUQkl2Sr3MydOzevc4iI5DmbzWD6uhNM/e0YhgENK/kys28oFUsUMzuaiOQijbkREYdgsxnEJqdzMSGVS/GpXExI5WJ8KpcS0v77v6mcvZrEyYuJAPRqVpkJXetrxW4RB5SjcrNkyRIWL15MeHg4aWlpWe7btWtXrgQTEfmn2KR0fth7jvOxKZmF5dJ/S8zlhDQybLe/+NPNxYm3ugXSo2lAPiQWETPYXW4++ugjXn/9dQYMGMDy5csZOHAgJ0+eZMeOHQwbNiwvMoqIsOXkJUYt2ktUXMotH1fC05UyXu6U9XKnjLc7Zbzcrv3sfW1b/Qo++Gm2YRGHZne5mTFjBp9++im9evVi3rx5vPrqq1SvXp3x48dz5cqVvMgoIkVYWoaNqauP8cnGkxgGVC3tSdu65TILTFlvd8p4uVPG243Sxd1xc9FswiJFnd3lJjw8nJYtWwJQrFgx4uPjAXjyySe56667+Pjjj3M3oYgUWacuJjBi4R72n4sF4ImmAYzvWh9PNw0XFJGbs/v/4vj7+2ceoalcuTJ//PEHAKdPn6aITXYsInnEMAwW7QjngY9+Z/+5WHyLuTKzT2PeebShio2I3Jbd/0rcd999/PDDD4SEhDBw4EBGjhzJkiVLCAsLs2uyPxGRG4lJSmPM0v38ciAKgBbVSzO1ZzDlfXW5tohkT7bXlvrpp5/o0qULADabDReXa71o4cKFbNmyhVq1avHMM8/g5uaWd2lzgdaWEim4/j5o2MXJwkv31+Hpe6rjrFmDRYo8e76/s11uXFxc8PPzY8CAAQwaNIgaNWrkStj8pnIjUvCkZdj44LdjzNpwbdBwtTLF+fCJRjSsVMLsaCJSQNjz/Z3tMTenT5/mmWeeYeHChdSuXZt7772XL7/8kuTk5DsOLCJF16mLCTw2awsz118rNk80DeCn4Xer2IhIjmX7yM3frVu3jnnz5vHdd9/h4uLCE088weDBg2natGleZMxVOnIjUjAYhsHisAje+OEQyelWfIu58s4jQXQOKm92NBEpgPLktNSNxMfHs3DhQubNm8cff/xBYGAge/fuzenu8oXKjYj5NGhYROxlz/f3HV1T6e3tTbt27Thz5gxHjhzh0KFDd7I7ESkCtp68zKjFe4iM1aBhEckbOZrKMzk5mfnz59OmTRtq1arFwoULGTVqFH/++WcuxxMRR7Lp+EWenLONyNgUqpUpzndDWzK0TQ0VGxHJVXYdufnjjz/4/PPPWbx4MWlpaTzyyCP89ttvtG3bNq/yiYiDOHAulme/3EmGzaBTA3+m9AimuLsm5BOR3Jftf1nq16/P0aNHCQkJYfLkyfTu3RtfX9+8zCYiDiLiShID5+0gMc1Ki+ql+bBXI9xdnM2OJSIOKtvlpn379ixYsIDg4ODMbZs3b6ZJkya4u7vnSTgRKfyuJqbRf+52LsanUtffm0/6harYiEieuqOrpXx8fNizZw/Vq1fPzUx5SldLieSflHQrfWZvY+eZq1Tw9WDpc63w9/UwO5aIFEJ5MonfjWihTBG5GavN4IUFu9l55io+Hi7MG9RMxUZE8sUdlRsRkRsxDIM3fjjIr4eicXN24rN+Tajt5212LBEpIu6o3HzyySf4+fnlVhYRcRAzN5zkyz/OYLHABz0b0bx6abMjiUgRckflpnfv3litVr7//nsOHz6cW5lEpBBbuuss7648CsC4B+rzQEMtpyAi+cvuctOjRw8+/vhj4Npkfk2aNKFHjx40bNiQ7777LtcDFiaGYRCfkm52DBHTbDp+kVeX7APgqdbVGHR3NZMTiUhRZHe52bhxI61btwZg2bJlGIZBTEwMH330EW+99VauBywstp26zP0fbOSVb/eZHUXEFAfPxzL0q11k2Ay6BldgTOd6ZkcSkSLK7nITGxtLqVKlAFi5ciWPPvoonp6ePPDAAxw/fjzXAxYWJYu7cfxCAqsPRxMZm2x2HJF8FXEliQFzd5CQmsFd1Uvx/uMNcdKSCiJiErvLTUBAAFu3biUxMZGVK1dy//33A3D16lU8PIruZZ61/bxpXq0UVpvBN9vCzY4jkm9iktIY8N9J+ur4efPJk000SZ+ImMrucvPiiy/Sp08fKlWqRIUKFWjTpg1w7XRVUFBQbucrVPq1qArAgu0RpGXYzA0jkg9S0q0M+SKMkxcTKe/rwbxBTfEt5mp2LBEp4uxete65556jWbNmRERE0KFDB5ycrvWj6tWrF+kxNwD3N/DDz8ed6LhUfjkQSbdGFc2OJJJnrDaDEQt3E3bmKt4eLnwxqBnlfYuZHUtEJGeXgjdp0oSHH34YLy8vrFYre/bsoWXLlrRq1Sq38xUqrs5O9GpWGYAvt54xOY1I3jEMg4k/HmTVQU3SJyIFT45OS82ZMwcAq9XKvffeS+PGjQkICGD9+vW5na/Q6d2sMi5OFsLOXOXg+Viz44jkiVkbTjH/vwV+as9g7tIkfSJSgNhdbpYsWZK5MviPP/7I6dOnOXLkCCNHjuT111+3O8D06dOpWrUqHh4eNG/enO3bt9/y8dOmTaNOnToUK1aMgIAARo4cSUpKit2vm1fK+XjQKdAf0NEbcUxLd53l3yuPADDuwfo82LCCyYlERLKyu9xcunQJf/9rX94rVqzg8ccfp3bt2gwaNIj9+/fbta9FixYxatQoJkyYwK5duwgODqZjx45cuHDhho//5ptvGD16NBMmTODw4cPMmTOHRYsW8a9//cveXyNP/TWw+Ps954hN0qR+4jhWHojilf9O0jfk7moM1iR9IlIA2V1u/Pz8OHToEFarlZUrV9KhQwcAkpKScHa27/LPqVOn8tRTTzFw4EDq16/PrFmz8PT05PPPP7/h47ds2UKrVq3o3bs3VatW5f7776dXr163PNqTmppKXFxclltea1q1JHX9vUlJt/Htzog8fz2R/LDp+EVeWLAbq83g0caV+FcXTdInIgWT3eVm4MCB9OjRg8DAQCwWC+3btwdg27Zt1K1bN9v7SUtLY+fOnZnPB3BycqJ9+/Zs3br1hs9p2bIlO3fuzCwzp06dYsWKFXTp0uWmrzN58mR8fX0zbwEBAdnOmFMWi4UnW1QB4Ks/zmCzGXn+miJ5KezPKzw9fydpVhudA/3596NBmqRPRAosu8vNG2+8wezZs3n66afZvHkz7u7uADg7OzN69Ohs7+fSpUtYrdbrVhX38/MjKirqhs/p3bs3kyZN4u6778bV1ZUaNWrQpk2bW56WGjNmDLGxsZm3iIj8OZLSvVFFvN1d+PNyEptOXMqX1xTJCwfOxTJw7g6S063cW7ss055ohIvzHa25KyKSp+ye5wbgscceu25b//797zjM7axfv563336bGTNm0Lx5c06cOMGIESN48803GTdu3A2f4+7unlnA8lNxdxceDa3EvC1/8uXWP7m3dtl8zyByp05ciKff59uJT82gWdVSzOobqtmHRaTAy9H//dqwYQNdu3alZs2a1KxZk4ceeohNmzbZtY8yZcrg7OxMdHR0lu3R0dGZA5b/ady4cTz55JMMGTKEoKAgHn74Yd5++20mT56MzVbwZgT+69TUmiMXiLiSZHIaEftEXEmi7+ztXElMI6iiL3MGNKGYm4qNiBR8dpebr776ivbt2+Pp6ckLL7zACy+8QLFixWjXrh3ffPNNtvfj5uZGaGgoa9asydxms9lYs2YNLVq0uOFzkpKSMmdE/stfg5gNo+CNa6lR1ovWtcpgGPC11puSQiQ6LoU+s7cRFZdCrXJefDGoGd4eWlZBRAoHu09L/d///R/vvvsuI0eOzNz2wgsvMHXqVN5880169+6d7X2NGjWK/v3706RJE5o1a8a0adNITExk4MCBAPTr14+KFSsyefJkALp27crUqVMJCQnJPC01btw4unbtaveVWvnlybuqsOn4JRbtCOfF9rXwcC2YOUX+ciUxjb6ztxF+JYnKpTz5akhzShV3MzuWiEi22V1uTp06RdeuXa/b/tBDD9k930zPnj25ePEi48ePJyoqikaNGrFy5crMQcbh4eFZjtSMHTsWi8XC2LFjOXfuHGXLlqVr16783//9n72/Rr5pV8+PiiWKcS4mmZ/2RfJYaCWzI4ncVFxKOv0/387xCwn4+3jw9ZDm+Pl4mB1LRMQuFsPO8zk1a9bklVde4ZlnnsmyfdasWUyZMoXjx4/nasDcFhcXh6+vL7Gxsfj4+OTLa05fd4L3Vh0luJIvy5+/O19eU8ReyWlW+n2+jR1/XqVUcTcWP9OCmuW8zI4lIgLY9/1t95Gbl156iRdeeCFzsUyAzZs3M2/ePD788MOcJXZwTzQN4MPfjrP3bCx7I2IIDihhdiSRLFIzrDzz1U52/Hlthe/5g5qp2IhIoWV3uRk6dCj+/v5MmTKFxYsXA1CvXj0WLVpEt27dcj2gIyjt5c4DDcuzbPc55m89wxSVGylAMqw2Xly4h43HLlLM1Zm5A5oSWNHX7FgiIjlmV7nJyMjg7bffZtCgQfz+++95lckhPdmiCst2n+PHfed5/YF6GqApBYLNZvDad/v55UAUbs5OfNovlCZVS5kdS0Tkjth1KbiLiwvvvvsuGRkZeZXHYYUElCCwog9pGTYWh2m9KTGfYRhM/PEg3+06i7OThf/0DqF1LU02KSKFn93z3LRr144NGzbkRRaHZrFYMlcL/+qPM1i13pSYbMqvx/hi6xksFnj/8YZ0bHDjyTNFRAobu8fcdO7cmdGjR7N//35CQ0MpXrx4lvsfeuihXAvnaB4KrsDbKw5z9moy645coH19v9s/SSSX2WwGMzec5ON1JwB4s1sgD4doigIRcRx2Xwr+zxmCs+zMYsFqtd5xqLxkxqXgf/f2isN8uvEU99Quy/xBzfL99aXoyrDa+HHfeWauP8mx6AQARneuy7P31jA5mYjI7eXppeAFcQ2nwqRv8yp8tukUG49d5PSlRKqVKX77J4ncgZR0K9+GRfDJxlOcvZoMgJe7Cy+2r8WQ1tVNTicikvtytCq45Fzl0p60qV2WdUcv8tUfZxj3YH2zI4mDiktJ56s/zvD576e5lJAGQOnibgy6uxp976qCbzGtFSUijinbA4rXrl1L/fr1iYuLu+6+2NhYGjRowMaNG3M1nKP6a2Dxt2ERJKcV7NN4UvhcSkjl3ZVHaDV5Le+uPMqlhDQqlijGxIca8Ptr9zGsbU0VGxFxaNk+cjNt2jSeeuqpG57n8vX15ZlnnuGDDz7gnnvuydWAjuje2mWpXMqT8CtJLN9zjieaVTY7kjiAiCtJfLbpFIt2RJCace30cc1yXgy9twYPNaqAq7PdF0eKiBRK2f7Xbu/evXTq1Omm999///3s3LkzV0I5OicnC33vulZo5m89g51jukWyOBYdz6hFe2jz/nrmbz1DaoaN4IASfPJkKL++eA+PhlZSsRGRIiXbR26io6Nxdb35oWwXFxcuXryYK6GKgh5NApjy6zEORcaxK/wqoVU0K6zYZ3f4VWasP8nqQ9GZ2+6uWYbn2tSgRY3SWCwWE9OJiJgn2+WmYsWKHDhwgJo1a97w/n379lG+fPlcC+boSni60a1RBRaHnWX+1jMqN5JtNpvB2OUH+GZbOAAWC3Rq4M/QNjVoWKmEueFERAqAbB+r7tKlC+PGjSMlJeW6+5KTk5kwYQIPPvhgroZzdH8NLF6xP5KL8anmhpFCwTAM3vr5MN9sC8fJAo+HVmL1yHuZ2TdUxUZE5L+yPYlfdHQ0jRs3xtnZmeeff546deoAcOTIEaZPn47VamXXrl34+RXsWXfNnsTvnx6esZnd4TG81KE2w9vVMjuOFHAf/nacD347BsCUx4N5NFQzC4tI0ZAnk/j5+fmxZcsWhg4dypgxYzIHwVosFjp27Mj06dMLfLEpiPq1qMLu8Bi+2R7O0DY1cNHAT7mJuZtPZxabN7rWV7EREbkJuybxq1KlCitWrODq1aucOHECwzCoVasWJUuWzKt8Dq9LUHne+ukwkbEp/HY4mk6BGrck1/tu51km/ngIgJHtazOgVTWTE4mIFFw5OkxQsmRJmjZtSrNmzVRs7pC7izM9mwYA1y4LF/mnVQejePW7fQAMalWNF9rdeFC/iIhco3MgBUCfu6rgZIEtJy9zLDre7DhSgGw+cYnh3+zGajN4LLQSYx+op0u8RURuQ+WmAKhYohj31/cH4JUl+0jL0OKkcm0em6fmh5FmtdGpgT/vPBKEk5OKjYjI7ajcFBDjutbHt5greyNi+PfKI2bHEZMdjYpnwNwdJKVZaV2rDB/2aqTB5iIi2aR/LQuIiiWK8f7jwQDM+f10lllnpWgJv5zEk3O2EZucTkjlEszqG4q7i7PZsURECg2VmwKkQ30/Bt997SqYl7/dy9mrSSYnkvwWHZdCnzl/cCE+lbr+3swb0Izi7nZd1CgiUuSp3BQwr3WqS3BACWKT0xm+YDfpVo2/KSquJqbx5JxtRFxJpkppT+YPboav583XcxMRkRtTuSlg3Fyc+LhXCD4eLuwOj+H9VUfNjiT5ICE1gwHzdnAsOgE/H3e+Gtycct4eZscSESmUVG4KoIBSnrz33/E3n2w8xdojGn/jyFLSrTz1RRh7I2Io6enKV4ObE1DK0+xYIiKFlspNAdWxgT8DWlYFYNTivZyPSTY3kOSJDKuN4Qt2s/XUZbzcXfhiUDNq+XmbHUtEpFBTuSnAxnSpS1BFX2KS0nlB428cjs1m8OqSfaw+FI2bixOf9Wuilb1FRHKByk0B5u7izPTejfF2dyHszFWmrj5mdiTJJTabwaSfDrF09zmcnSzM6N2YFjVKmx1LRMQhqNwUcJVLe/LvxxoCMHP9SdYdvWByIrlTO89cpfuMzczb8icWC0x5PJj29f3MjiUi4jBUbgqBLkHl6deiCgAvLd5LVGyKyYkkJ6LjUhi5aA+PztzCvrOxeLu78N5jwXQPqWh2NBERh6LZwQqJf3Wpx84zVzl4Po4XFuzmm6eaazr+QiIl3cqc308zfd0JktKsWCzweGglXulYl7Le7mbHExFxOPp2LCQ8XK+Nv/Fyd2H7n1eY9ttxsyPJbRiGwepD0dz/wUbeW3WUpDQrjSuXYPmwVrz7WLCKjYhIHtGRm0KkapniTH4kiOELdjN9/QmaVSvFPbXLmh1LbuDEhXgm/niITccvAeDn487oznXp3qgiFotW9hYRyUsqN4VM1+AK/HHqMl9vC2fkoj2sGNEaPx/NZFtQxCanM+23Y8zfegarzcDN2YkhrasxrG1NrRElIpJP9K9tITTuwfrsCo/hcGQcIxbu5ushd+HspKMB9jp9KZHR3+3D082ZmuW8qFXOmxrlvKhZzgvfYvat6WS1GSwOi+C9VUe5kpgGXFsIdewD9ahSunhexBcRkZuwGIZhmB0iP8XFxeHr60tsbCw+Pj5mx8mxUxcT6Pqf30lMs/JCu1qM6lDb7EiFSnKalYdnbOZIVPwN7y/n7f7fwnOt7NT4b/kp4+V23WmlHX9e4Y0fDnLwfBwANct5MaFrfVrX0ilDEZHcYs/3t47cFFLVy3rx9iNBjFi4h/+sPU7zaqVoVbOM2bEKjYk/HuRIVDxlvNx5oV1NTl1M5MSFBE5cSCAqLoUL8alciE9ly8nLWZ7nW8w1S+nZdzaWH/aeB8Dbw4WR7WvzZIsquOpKNhER0+jITSE3+rt9LNwRQRkvd1aMuFsrSWfDst1nGbloLxYLfD24OS3/UQrjUtI5+d+ik3m7mED4lSRu9F+LxQJPNA3g5fvrUNpLV0CJiOQFHbkpQt54qAF7ImI4EhXPs1/u5JMnm+gS41s4cSGB15cdAGBEu1rXFRsAHw9XQiqXJKRyySzbU9Kt147wXEzgRHQ8Jy4mYLFYGHpvDQIr+uZLfhERuT0duXEAJy4k0H36ZhJSMyjr7c6HTzSiZQ2dovqn5DQr3adv5mh0PC1rlObLwc01EFtEpJCw5/tbAwMcQM1yXix7riW1/by4GJ9K39nbmPbbMay2ItVbb+uNHw5yNPraOJtpTzRSsRERcVAqNw6ilp83y4fdTY8mlbAZMO234/SdvY0LcVqHCmDprrMsCovAyQIfPdFIY5NERByYyo0DKebmzLuPBfNBz2A83ZzZeuoyXT7axKbjF82OZqoTF+L/Ns6m9g3H2YiIiONQuXFAD4dU4ofn76auvzeXEtLo9/l23l91lAyrzexo+S45zcqwr3eTnG6lVc3SPH9fTbMjiYhIHlO5cVA1y3nx/bBW9G5eGcOAj9edoPdn24iKLVqnqSb8cICj0fGU9XZnWs8QjbMRESkCVG4cmIerM28/HMRHvUIyVxPv8tEm1h29YHa0fPHdzrMsDjuLkwU+fKKRLpEXESkiVG6KgIeCK/Dj8LtpUMGHK4lpDJy7g8m/HCbdgU9TnbgQz9jv/zbORpfGi4gUGSo3RUS1MsX5bmhL+rWoAsAnG07xxKd/cC4m2eRkuS85zcpzX+8iOd3K3TXLaJyNiEgRo3JThHi4OjOpWyAz+zTG28OFnWeu8sBHm/jtULTZ0XLV+OUHOBadQFlvdz7oqflsRESKGi2/UAR1DipPgwq+DF+wi71nYxkyP4xBrarRvn45fIu5Zt683F2uWwG7oFuy8yzf7tQ4GxGRokzLLxRhaRk23vnlCJ9vPn3D+52dLPh4uGSWHZ//3nxvcKtYohj1yvvg5mLewcDj0fE89PFmktOtjOpQmxfa1TIti4iI5C4tnCnZ4ubixPiu9WlRozSzN53iUkIqsckZxCWnk2a1YbUZXE1K52pSevb25+xE/Qo+NAookXmrUtozX47+JKVlZBlnM6ytxtmIiBRVKjdCh/p+dKjvl2VbSrqV2OT0/92S0rP+nJxO3N/+fPJiAleT0tkTEcOeiJjM/ZTwdCW40v/KTnBACUoVd8v132H88oMcv5BAOW+tGyUiUtSp3MgNebg64+HqjJ9P9tZgMgyD8CtJmeVmT0QMB8/HEZOUzoZjF9lw7H9LQFQp7fm/wlO5BPXL++Dh6pzjrN+GRbAkc5xNCGW8NM5GRKQo05gbyTNpGTaORMVlKTynLiZe9zhnJ0vmAObi7i54u7vg5eGC13//19v9f3/2cnfB28MFL3dXvDxcSEzNYPAXO0hJt/FSh9oM1zgbERGHZM/3t8qN5KvYpHT2no1h798Kz+XEtDveb+taZZg3sJlOR4mIOCgNKJYCy9fTlXtql+We2mWBa6ezLsSnEpucTnxKBgmpGSSkZJCQmvXnxLSMf9z/v5/L+3poPhsREcmkciOmslgs+Pl4ZHtsj4iIyO1ohmIRERFxKCo3IiIi4lBUbkRERMShmF5upk+fTtWqVfHw8KB58+Zs3779lo+PiYlh2LBhlC9fHnd3d2rXrs2KFSvyKa2IiIgUdKYOKF60aBGjRo1i1qxZNG/enGnTptGxY0eOHj1KuXLlrnt8WloaHTp0oFy5cixZsoSKFSty5swZSpQokf/hRUREpEAydZ6b5s2b07RpUz7++GMAbDYbAQEBDB8+nNGjR1/3+FmzZvHee+9x5MgRXF1dc/SamudGRESk8LHn+9u001JpaWns3LmT9u3b/y+MkxPt27dn69atN3zODz/8QIsWLRg2bBh+fn4EBgby9ttvY7Vab/o6qampxMXFZbmJiIiI4zKt3Fy6dAmr1YqfX9YFG/38/IiKirrhc06dOsWSJUuwWq2sWLGCcePGMWXKFN56662bvs7kyZPx9fXNvAUEBOTq7yEiIiIFi+kDiu1hs9koV64cn376KaGhofTs2ZPXX3+dWbNm3fQ5Y8aMITY2NvMWERGRj4lFREQkv5k2oLhMmTI4OzsTHR2dZXt0dDT+/v43fE758uVxdXXF2fl/K0jXq1ePqKgo0tLScHNzu+457u7uuLtrlWgREZGiwrQjN25uboSGhrJmzZrMbTabjTVr1tCiRYsbPqdVq1acOHECm82Wue3YsWOUL1/+hsVGREREih5TT0uNGjWKzz77jC+++ILDhw8zdOhQEhMTGThwIAD9+vVjzJgxmY8fOnQoV65cYcSIERw7doyff/6Zt99+m2HDhpn1K4iIiEgBY+o8Nz179uTixYuMHz+eqKgoGjVqxMqVKzMHGYeHh+Pk9L/+FRAQwKpVqxg5ciQNGzakYsWKjBgxgtdee82sX0FEREQKGFPnuTGD5rkREREpfOz5/jb1yI0Z/upymu9GRESk8Pjrezs7x2SKXLmJj48H0Hw3IiIihVB8fDy+vr63fEyROy1ls9k4f/483t7eWCyWLPfFxcUREBBARESETlnZQe9bzuh9yxm9b/bTe5Yzet9yJq/eN8MwiI+Pp0KFClnG495IkTty4+TkRKVKlW75GB8fH32Qc0DvW87ofcsZvW/203uWM3rfciYv3rfbHbH5S6GaoVhERETkdlRuRERExKGo3PyNu7s7EyZM0HINdtL7ljN633JG75v99J7ljN63nCkI71uRG1AsIiIijk1HbkRERMShqNyIiIiIQ1G5EREREYeiciMiIiIOReXmb6ZPn07VqlXx8PCgefPmbN++3exIBdobb7yBxWLJcqtbt67ZsQqcjRs30rVrVypUqIDFYuH777/Pcr9hGIwfP57y5ctTrFgx2rdvz/Hjx80JW0Dc7j0bMGDAdZ+9Tp06mRO2gJg8eTJNmzbF29ubcuXK0b17d44ePZrlMSkpKQwbNozSpUvj5eXFo48+SnR0tEmJC4bsvG9t2rS57vP27LPPmpS4YJg5cyYNGzbMnKivRYsW/PLLL5n3m/1ZU7n5r0WLFjFq1CgmTJjArl27CA4OpmPHjly4cMHsaAVagwYNiIyMzLz9/vvvZkcqcBITEwkODmb69Ok3vP/dd9/lo48+YtasWWzbto3ixYvTsWNHUlJS8jlpwXG79wygU6dOWT57CxYsyMeEBc+GDRsYNmwYf/zxB6tXryY9PZ3777+fxMTEzMeMHDmSH3/8kW+//ZYNGzZw/vx5HnnkERNTmy877xvAU089leXz9u6775qUuGCoVKkS77zzDjt37iQsLIz77ruPbt26cfDgQaAAfNYMMQzDMJo1a2YMGzYs82er1WpUqFDBmDx5sompCrYJEyYYwcHBZscoVABj2bJlmT/bbDbD39/feO+99zK3xcTEGO7u7saCBQtMSFjw/PM9MwzD6N+/v9GtWzdT8hQWFy5cMABjw4YNhmFc+1y5uroa3377beZjDh8+bADG1q1bzYpZ4PzzfTMMw7j33nuNESNGmBeqkChZsqQxe/bsAvFZ05EbIC0tjZ07d9K+ffvMbU5OTrRv356tW7eamKzgO378OBUqVKB69er06dOH8PBwsyMVKqdPnyYqKirLZ8/X15fmzZvrs3cb69evp1y5ctSpU4ehQ4dy+fJlsyMVKLGxsQCUKlUKgJ07d5Kenp7ls1a3bl0qV66sz9rf/PN9+8vXX39NmTJlCAwMZMyYMSQlJZkRr0CyWq0sXLiQxMREWrRoUSA+a0Vu4cwbuXTpElarFT8/vyzb/fz8OHLkiEmpCr7mzZszb9486tSpQ2RkJBMnTqR169YcOHAAb29vs+MVClFRUQA3/Oz9dZ9cr1OnTjzyyCNUq1aNkydP8q9//YvOnTuzdetWnJ2dzY5nOpvNxosvvkirVq0IDAwErn3W3NzcKFGiRJbH6rP2Pzd63wB69+5NlSpVqFChAvv27eO1117j6NGjLF261MS05tu/fz8tWrQgJSUFLy8vli1bRv369dmzZ4/pnzWVG8mxzp07Z/65YcOGNG/enCpVqrB48WIGDx5sYjJxdE888UTmn4OCgmjYsCE1atRg/fr1tGvXzsRkBcOwYcM4cOCAxsDZ6Wbv29NPP53556CgIMqXL0+7du04efIkNWrUyO+YBUadOnXYs2cPsbGxLFmyhP79+7NhwwazYwEaUAxAmTJlcHZ2vm4kd3R0NP7+/ialKnxKlChB7dq1OXHihNlRCo2/Pl/67N2Z6tWrU6ZMGX32gOeff56ffvqJdevWUalSpczt/v7+pKWlERMTk+Xx+qxdc7P37UaaN28OUOQ/b25ubtSsWZPQ0FAmT55McHAwH374YYH4rKnccO0vKDQ0lDVr1mRus9lsrFmzhhYtWpiYrHBJSEjg5MmTlC9f3uwohUa1atXw9/fP8tmLi4tj27Zt+uzZ4ezZs1y+fLlIf/YMw+D5559n2bJlrF27lmrVqmW5PzQ0FFdX1yyftaNHjxIeHl6kP2u3e99uZM+ePQBF+vN2IzabjdTU1ILxWcuXYcuFwMKFCw13d3dj3rx5xqFDh4ynn37aKFGihBEVFWV2tALrpZdeMtavX2+cPn3a2Lx5s9G+fXujTJkyxoULF8yOVqDEx8cbu3fvNnbv3m0AxtSpU43du3cbZ86cMQzDMN555x2jRIkSxvLly419+/YZ3bp1M6pVq2YkJyebnNw8t3rP4uPjjZdfftnYunWrcfr0aeO3334zGjdubNSqVctISUkxO7pphg4davj6+hrr1683IiMjM29JSUmZj3n22WeNypUrG2vXrjXCwsKMFi1aGC1atDAxtflu976dOHHCmDRpkhEWFmacPn3aWL58uVG9enXjnnvuMTm5uUaPHm1s2LDBOH36tLFv3z5j9OjRhsViMX799VfDMMz/rKnc/M1//vMfo3Llyoabm5vRrFkz448//jA7UoHWs2dPo3z58oabm5tRsWJFo2fPnsaJEyfMjlXgrFu3zgCuu/Xv398wjGuXg48bN87w8/Mz3N3djXbt2hlHjx41N7TJbvWeJSUlGffff79RtmxZw9XV1ahSpYrx1FNPFfn/I3Kj9wsw5s6dm/mY5ORk47nnnjNKlixpeHp6Gg8//LARGRlpXugC4HbvW3h4uHHPPfcYpUqVMtzd3Y2aNWsar7zyihEbG2tucJMNGjTIqFKliuHm5maULVvWaNeuXWaxMQzzP2sWwzCM/DlGJCIiIpL3NOZGREREHIrKjYiIiDgUlRsRERFxKCo3IiIi4lBUbkRERMShqNyIiIiIQ1G5EREREYeiciMiIiIOReVGRG6qTZs2vPjii2bHwDAMnn76aUqVKoXFYslc2+fv5s2bR4kSJezab1JSEo8++ig+Pj5YLJbrFvoTkcJJ5UbEAXXt2pVOnTrd8L5NmzZhsVjYt29fPqfKuZUrVzJv3jx++uknIiMjCQwMzJX9fvHFF2zatIktW7YQGRmJr69vruwX4M8//7yuiMXHx9O2bVvq16/P2bNnMx9Trlw54uPjszy/UaNGvPHGG5k/t2nTBovFwsKFC7M8btq0aVStWjXXcos4ApUbEQc0ePBgVq9ezdmzZ6+7b+7cuTRp0oSGDRuakCxn/lptvmXLlvj7++Pi4pJr+61Xrx6BgYH4+/tjsVjs3ofVasVms932cRcvXqRt27YkJiayadMmKlWqlHlffHw877///m334eHhwdixY0lPT7c7p0hRonIj4oAefPBBypYty7x587JsT0hI4Ntvv2Xw4MFcvnyZXr16UbFiRTw9PQkKCmLBggW33K/FYuH777/Psq1EiRJZXiciIoIePXpQokQJSpUqRbdu3fjzzz9vud8NGzbQrFkz3N3dKV++PKNHjyYjIwOAAQMGMHz4cMLDw7FYLNk+SnHx4kWaNGnCww8/TGpq6nX3t2nThilTprBx40YsFgtt2rQB4OrVq/Tr14+SJUvi6elJ586dOX78eObz/jr99cMPP1C/fn3c3d0JDw+/ZZaIiAhat26Nr68va9eupXTp0lnuHz58OFOnTuXChQu33E+vXr2IiYnhs88+y9Z7IFJUqdyIOCAXFxf69evHvHnz+PvauN9++y1Wq5VevXqRkpJCaGgoP//8MwcOHODpp5/mySefZPv27Tl+3fT0dDp27Ii3tzebNm1i8+bNeHl50alTJ9LS0m74nHPnztGlSxeaNm3K3r17mTlzJnPmzOGtt94C4MMPP2TSpElUqlSJyMhIduzYcdscf5WJwMBAlixZgru7+3WPWbp0KU899RQtWrQgMjKSpUuXAtfKVFhYGD/88ANbt27FMAy6dOmS5WhJUlIS//73v5k9ezYHDx6kXLlyN81y9OhRWrVqRf369VmxYgVeXl7XPaZXr17UrFmTSZMm3fL38vHx4fXXX2fSpEkkJibe9n0QKapUbkQc1KBBgzh58iQbNmzI3DZ37lweffRRfH19qVixIi+//DKNGjWievXqDB8+nE6dOrF48eIcv+aiRYuw2WzMnj2boKAg6tWrx9y5cwkPD2f9+vU3fM6MGTMICAjg448/pm7dunTv3p2JEycyZcoUbDYbvr6+eHt74+zsjL+/P2XLlr1lhr/KRMeOHZk7dy7Ozs43fFypUqXw9PTEzc0Nf39/SpUqxfHjx/nhhx+YPXs2rVu3Jjg4mK+//ppz585lOWKVnp7OjBkzaNmyJXXq1MHT0/Omefr160fNmjX59ttvb1iy4NoRsXfeeYdPP/2UkydP3vL3e+655/Dw8GDq1Km3fJxIUaZyI+Kg6tatS8uWLfn8888BOHHiBJs2bWLw4MHAtbEib775JkFBQZQqVQovLy9WrVp121Mst7J3715OnDiBt7c3Xl5eeHl5UapUKVJSUm76pX348GFatGiRZbxLq1atSEhIuOGYoVtJTk6mdevWPPLII3z44Yd2j6E5fPgwLi4uNG/ePHNb6dKlqVOnDocPH87c5ubmlu0xSw899BCbNm3KPDJ0Mx07duTuu+9m3Lhxt3ycu7s7kyZN4v333+fSpUvZyiBS1KjciDiwwYMH89133xEfH8/cuXOpUaMG9957LwDvvfceH374Ia+99hrr1q1jz549dOzY8aanj+DaEYa/n+YCspyuSUhIIDQ0lD179mS5HTt2jN69e+fNL/k37u7utG/fnp9++olz587l2esUK1Ys28Xp9ddfZ/z48fTu3fu2R8XeeecdFi1axO7du2/5uL59+1KlSpXMU3cikpXKjYgD69GjB05OTnzzzTfMnz+fQYMGZX4pb968mW7dutG3b1+Cg4OpXr06x44du+X+ypYtS2RkZObPx48fJykpKfPnxo0bc/z4ccqVK0fNmjWz3G52mXW9evUyx7b8ZfPmzXh7e2e5oig7nJyc+PLLLwkNDaVt27acP3/erufXq1ePjIwMtm3blrnt8uXLHD16lPr169u1r78bN24cb7zxBn369GHRokU3fVyzZs145JFHGD169C335+TkxOTJk5k5c+ZtB2uLFEUqNyIOzMvLi549ezJmzBgiIyMZMGBA5n21atVi9erVbNmyhcOHD/PMM88QHR19y/3dd999fPzxx+zevZuwsDCeffZZXF1dM+/v06cPZcqUoVu3bmzatInTp0+zfv16XnjhhZueYnruueeIiIhg+PDhHDlyhOXLlzNhwgRGjRqFk5P9/0Q5Ozvz9ddfExwczH333UdUVFS2n1urVi26devGU089xe+//87evXvp27cvFStWpFu3bnZn+bvXX3+dN998kz59+tzyqrT/+7//Y+3atRw9evSW+3vggQdo3rw5n3zyyR3lEnFEKjciDm7w4MFcvXqVjh07UqFChcztY8eOpXHjxnTs2JE2bdrg7+9P9+7db7mvKVOmEBAQQOvWrenduzcvv/xylsG0np6ebNy4kcqVK/PII49Qr149Bg8eTEpKCj4+PjfcZ8WKFVmxYgXbt28nODiYZ599lsGDBzN27Ngc/84uLi4sWLCABg0acN999932Euu/mzt3LqGhoTz44IO0aNECwzBYsWJFlhKXU6NHj+btt9/mySef5JtvvrnhY2rXrs2gQYNISUm57f7+/e9/Z+txIkWNxfjnCXQRERGRQkxHbkRERMShqNyIiIiIQ1G5EREREYeiciMiIiIOReVGREREHIrKjYiIiDgUlRsRERFxKCo3IiIi4lBUbkRERMShqNyIiIiIQ1G5EREREYfy/5LW/pG/rDEeAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#knn\n",
    "x_range = range(1, 31)\n",
    "mae = []\n",
    "for x in x_range:\n",
    "    import logging\n",
    "\n",
    "    #build knn model\n",
    "    model = KNeighborsRegressor(n_neighbors=x)\n",
    "    model.fit(boston_feature_train, boston_label_train)\n",
    "\n",
    "    #use LOOCV to evaluate model\n",
    "    scores = cross_val_score(model, boston_feature_test, boston_label_test, scoring='neg_mean_absolute_error',\n",
    "                         cv=cv, n_jobs=-1)\n",
    "    #mean absolute error\n",
    "    mae.append(np.mean(np.absolute(scores)))\n",
    "\n",
    "#plot mean absolute error values\n",
    "print(mae)\n",
    "plt.plot(x_range, mae)\n",
    "plt.xlabel('Value of k for KNN')\n",
    "plt.ylabel('Cross-Validated mean absolute error')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<font color = skyblue> k should be 7 for best results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 132,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#predictions \n",
    "knn = KNeighborsRegressor(n_neighbors=2)\n",
    "\n",
    "#fit training data to classifier\n",
    "knn.fit(boston_feature_train, boston_label_train)\n",
    "\n",
    "#make predictions\n",
    "knn_prediction = knn.predict(boston_feature_test)\n",
    "\n",
    "#plot\n",
    "plt.hist(knn_prediction, bins=20)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 133,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[3.217111650051891e-06, 6.433756082869614e-06, 1.2865644306113392e-05, 2.5723825267655585e-05, 5.1417861760138833e-05, 0.00010271708185919593, 0.00020496365039871016, 0.00040807683387579773, 0.0008089918457607947, 0.001591435724408641, 0.0030916000124050587, 0.005894090389087099, 0.010943634217549708, 0.019810880774276597, 0.03481046055892684, 0.060938748266612754, 0.1066734715674272, 0.187167212954411, 0.3290877832515643, 0.578055722497608]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# ridge regression\n",
    "# https://machinelearningmastery.com/ridge-regression-with-python/\n",
    "x_range = range(-10, 10)\n",
    "ridge_mae = []\n",
    "for x in x_range:\n",
    "    # define model\n",
    "    model = Ridge(alpha=2**x)\n",
    "    model.fit(boston_feature_train, boston_label_train)\n",
    "\n",
    "    # evaluate model\n",
    "    scores = cross_val_score(model, boston_feature_test, boston_label_test, scoring='neg_mean_absolute_error', cv=cv, n_jobs=-1)\n",
    "    \n",
    "    #mean absolute error\n",
    "    ridge_mae.append(np.mean(np.absolute(scores)))\n",
    "\n",
    "#plot\n",
    "print(ridge_mae)\n",
    "plt.plot(x_range, ridge_mae)\n",
    "plt.xlabel('lambda')\n",
    "plt.ylabel('Cross-Validated mean absolute error')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 144,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#predictions \n",
    "# from array it seems 2^(-5) was the best so lets use it for predictions\n",
    "ridge_model = Ridge(alpha=2**(-5))\n",
    "\n",
    "    \n",
    "#fit training data to classifier\n",
    "ridge_model.fit(boston_feature_train, boston_label_train)\n",
    "\n",
    "#make predictions\n",
    "ridge_prediction = ridge_model.predict(boston_feature_test)\n",
    "\n",
    "#plot\n",
    "plt.hist(ridge_prediction, bins=20)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<font color = skyblue> Lambda is a regulairization term. When lambda is small very little bias is introduced to the model. Smaller mae is the better the result. We can grow the regularization tern extremely large and still get better or as good result as with knn. Ridge regrression is the better modle of these two for this type of modeling. </font>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "------\n",
    "\n",
    "We can say something about the performance now. One way to do the visualization is to plot actual labels vs predicted labels. Let's do that.\n",
    "\n",
    "\n",
    "***Exercise 5 D)***\n",
    "\n",
    "Using the best models found in the previous exercise, plot the actual labels vs. predicted labels. \n",
    "\n",
    "- How did the selected models perform?\n",
    "- In an ideal situation, where should the data points be located?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 145,
   "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>MEDV</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>32.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>29.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>14.3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>19.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>23.3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>89</th>\n",
       "      <td>25.3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>90</th>\n",
       "      <td>18.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>91</th>\n",
       "      <td>32.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>92</th>\n",
       "      <td>17.3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>93</th>\n",
       "      <td>20.4</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>94 rows × 1 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "    MEDV\n",
       "0   32.2\n",
       "1   29.0\n",
       "2   14.3\n",
       "3   19.8\n",
       "4   23.3\n",
       "..   ...\n",
       "89  25.3\n",
       "90  18.4\n",
       "91  32.0\n",
       "92  17.3\n",
       "93  20.4\n",
       "\n",
       "[94 rows x 1 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>0</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>32.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>29.70</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>14.30</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>19.30</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>23.45</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>89</th>\n",
       "      <td>24.95</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>90</th>\n",
       "      <td>17.75</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>91</th>\n",
       "      <td>32.55</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>92</th>\n",
       "      <td>17.25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>93</th>\n",
       "      <td>20.50</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>94 rows × 1 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        0\n",
       "0   32.00\n",
       "1   29.70\n",
       "2   14.30\n",
       "3   19.30\n",
       "4   23.45\n",
       "..    ...\n",
       "89  24.95\n",
       "90  17.75\n",
       "91  32.55\n",
       "92  17.25\n",
       "93  20.50\n",
       "\n",
       "[94 rows x 1 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>0</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>32.199980</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>29.000007</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>14.300011</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>19.800019</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>23.300029</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>89</th>\n",
       "      <td>25.300000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>90</th>\n",
       "      <td>18.399976</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>91</th>\n",
       "      <td>31.999988</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>92</th>\n",
       "      <td>17.299996</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>93</th>\n",
       "      <td>20.399996</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>94 rows × 1 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            0\n",
       "0   32.199980\n",
       "1   29.000007\n",
       "2   14.300011\n",
       "3   19.800019\n",
       "4   23.300029\n",
       "..        ...\n",
       "89  25.300000\n",
       "90  18.399976\n",
       "91  31.999988\n",
       "92  17.299996\n",
       "93  20.399996\n",
       "\n",
       "[94 rows x 1 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "display(boston_label_test)\n",
    "knn_prediction = pd.DataFrame(knn_prediction)\n",
    "display(knn_prediction)\n",
    "ridge_prediction = pd.DataFrame(ridge_prediction)\n",
    "display(ridge_prediction)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 146,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAh8AAAGdCAYAAACyzRGfAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/NK7nSAAAACXBIWXMAAA9hAAAPYQGoP6dpAAA26ElEQVR4nO3de1yUZf7/8fcEgqiIeQQSFJXwkHjW2DIxTUR0Nd3U1oqsTFNSs1xjXU+1LmaZhzStVcM289BByyx3XVPMs6J4WFs8fEkoUdoOICiocP/+6OfUyKCgM/c48Ho+HvfjwX3d19zXZy5u4M193zNjMQzDEAAAgEluc3UBAACgYiF8AAAAUxE+AACAqQgfAADAVIQPAABgKsIHAAAwFeEDAACYivABAABM5enqAq5WVFSk06dPy9fXVxaLxdXlAACAUjAMQ+fOnVNgYKBuu+3a5zZuufBx+vRpBQUFuboMAABwAzIyMlS/fv1r9rnlwoevr6+kX4qvXr26i6sBAAClkZOTo6CgIOvf8Wu55cLHlUst1atXJ3wAAOBmSnPLBDecAgAAUxE+AACAqQgfAADAVLfcPR+lYRiGLl++rMLCQleXAqAMPDw85OnpycvogQrO7cLHxYsXlZmZqfPnz7u6FAA3oEqVKgoICJCXl5erSwHgIm4VPoqKipSWliYPDw8FBgbKy8uL/6AAN2EYhi5evKjvv/9eaWlpCg0Nve4bEQEon9wqfFy8eFFFRUUKCgpSlSpVXF0OgDLy8fFRpUqVdOrUKV28eFGVK1d2dUkAXMAt/+3gvyXAffHzC4DfAgAAwFSEDwAAYKoy3/OxdetWvfrqq0pOTlZmZqbWrFmjfv362fT5+uuvNWHCBCUlJeny5ctq3ry5PvroIwUHBzuq7mIyE8298TTgccPU8ZzNYrHY/V5eERkZqdatW2vOnDml2t+WLVvUtWtX/fTTT6pRo8YN19WwYUONHTtWY8eOveF9AABuLWU+85GXl6dWrVppwYIFdrefPHlS9957r5o2baotW7bo0KFDmjRpEjeW/X87d+6Uh4eHYmJiyvzYhg0blvqPPwAAt6oyn/mIjo5WdHR0idsnTpyoXr16aebMmda2xo0b31h15dCSJUv07LPPasmSJTp9+rQCAwNdXRIAAKZy6D0fRUVFWr9+ve68805FRUWpbt266tSpk9auXVviYwoKCpSTk2OzlFe5ublatWqVnnnmGcXExCgxMbFYn3Xr1qlDhw6qXLmyateurQcffFDSL5c9Tp06peeee04Wi8X6/iZTp05V69atbfYxZ84cNWzY0Lq+d+9ePfDAA6pdu7b8/PzUpUsX7d+//6aeyz/+8Q+1b99evr6+8vf31x//+EdlZWUV67d9+3aFh4ercuXKuvvuu3XkyBGb7du2bVPnzp3l4+OjoKAgjR49Wnl5eXbHNAxDU6dOVXBwsLy9vRUYGKjRo0ff1PMAAJjPoeEjKytLubm5mjFjhnr27Kl//etfevDBB9W/f38lJSXZfUxCQoL8/PysS1BQkCNLuqWsXr1aTZs2VVhYmB555BEtXbpUhvHrvSPr16/Xgw8+qF69eunAgQPatGmTOnbsKEn6+OOPVb9+fb300kvKzMxUZmZmqcc9d+6cYmNjtW3bNu3atUuhoaHq1auXzp07d8PP5dKlS3r55Zd18OBBrV27Vt98840ef/zxYv3Gjx+vWbNmae/evapTp4769OmjS5cuSfrlEl3Pnj01YMAAHTp0SKtWrdK2bdsUFxdnd8yPPvpIs2fP1ltvvaXjx49r7dq1atmy5Q0/B9y6Bq3PsrsAKB8c+iZjRUVFkqS+ffvqueeekyS1bt1aO3bs0KJFi9SlS5dij4mPj9e4ceOs6zk5OeU2gCxZskSPPPKIJKlnz57Kzs5WUlKSIiMjJUnTp0/X4MGDNW3aNOtjWrVqJUmqWbOmPDw8rGcayuL++++3WX/77bdVo0YNJSUlqXfv3jf0XJ544gnr140aNdK8efPUoUMH5ebmqlq1atZtU6ZM0QMPPCBJWrZsmerXr681a9Zo4MCBSkhI0JAhQ6w3k4aGhmrevHnq0qWLFi5cWOw+ofT0dPn7+6t79+6qVKmSgoODreEMAOA+HHrmo3bt2vL09FTz5s1t2ps1a6b09HS7j/H29lb16tVtlvIoNTVVe/bs0cMPPyxJ8vT01KBBg7RkyRJrn5SUFHXr1s3hY589e1bDhg1TaGio/Pz8VL16deXm5pb4PSmN5ORk9enTR8HBwfL19bUGy6v3GRERYf26Zs2aCgsL09dffy1JOnjwoBITE1WtWjXrEhUVZX0b/as99NBDunDhgho1aqRhw4ZpzZo1unz58g0/BwCAazj0zIeXl5c6dOig1NRUm/Zjx46pQYMGjhzK7SxZskSXL1+2ucHUMAx5e3tr/vz58vPzk4+PT5n3e9ttt9lcupFkvaxxRWxsrH744QfNnTtXDRo0kLe3tyIiInTx4sUbei55eXmKiopSVFSUli9frjp16ig9PV1RUVFl2mdubq6GDx9u974Ney/LDgoKUmpqqv79739r48aNGjlypF599VUlJSWpUqVKN/RcAADmK3P4yM3N1YkTJ6zraWlpSklJUc2aNRUcHKzx48dr0KBBuu+++9S1a1dt2LBB69at05YtWxxZt1u5fPmy3n33Xc2aNUs9evSw2davXz+tWLFCI0aMUHh4uDZt2qShQ4fa3Y+Xl5cKCwtt2urUqaMzZ87IMAzrTagpKSk2fbZv364333xTvXr1kiRlZGTof//73w0/n//+97/64YcfNGPGDOslsn379tntu2vXLmuQ+Omnn3Ts2DE1a9ZMktS2bVsdPXpUTZo0KfXYPj4+6tOnj/r06aNRo0apadOmOnz4sNq2bXvDzwcAYK4yh499+/apa9eu1vUr92vExsYqMTFRDz74oBYtWqSEhASNHj1aYWFh+uijj3Tvvfc6rmo389lnn+mnn37Sk08+KT8/P5ttAwYM0JIlSzRixAhNmTJF3bp1U+PGjTV48GBdvnxZn3/+uSZMmCDpl/f52Lp1qwYPHixvb2/Vrl1bkZGR+v777zVz5kz94Q9/0IYNG/TFF1/YXL4KDQ21vjolJydH48ePv6GzLFcEBwfLy8tLb7zxhkaMGKEjR47o5Zdfttv3pZdeUq1atVSvXj1NnDhRtWvXtr6R2YQJE3T33XcrLi5OTz31lKpWraqjR49q48aNmj9/frF9JSYmqrCwUJ06dVKVKlX03nvvycfHp8KfVQMAt2PcYrKzsw1JRnZ2drFtFy5cMI4ePWpcuHDBBZXduN69exu9evWyu2337t2GJOPgwYOGYRjGRx99ZLRu3drw8vIyateubfTv39/ad+fOnUZ4eLjh7e1t/PZbt3DhQiMoKMioWrWq8dhjjxnTp083GjRoYN2+f/9+o3379kblypWN0NBQ44MPPjAaNGhgzJ4929pHkrFmzZoSn0OXLl2MMWPGWNfff/99o2HDhoa3t7cRERFhfPrpp4Yk48CBA4ZhGMbmzZsNSca6deuMFi1aGF5eXkbHjh2tz/OKPXv2GA888IBRrVo1o2rVqkZ4eLgxffp06/bf1rlmzRqjU6dORvXq1Y2qVasad999t/Hvf/+7xJpxayrNz/HAz87aXQDcuq719/tqFsMwbqn3Cc/JyZGfn5+ys7OL3Xyan5+vtLQ0hYSE8I6pgJsqzc9xSS+rXRVT15mlAbgJ1/r7fTU+WA4AAJiK8AEAAExF+AAAAKYifAAAAFMRPgAAgKkIHwAAwFSEDwAAYCrCBwAAMBXhwySRkZHWj46vKL755htZLJZinzUDAKjYHPqptq5U0jsiOgvvtGi+qVOnatq0acXaq1Spory8PBdUBAC4EZz5QDEXL150dQl2vfDCC8rMzLRZmjdvroceesjVpQEAyoDw4SLr16+Xn5+fli9fLkl6/PHH1a9fP7322msKCAhQrVq1NGrUKF26dMn6mIYNG+pvf/ubnnjiCfn6+io4OFhvv/32NceJjIxUXFyc4uLi5Ofnp9q1a2vSpEn67Uf6NGzYUC+//LIee+wxVa9eXU8//bQkadu2bercubN8fHwUFBSk0aNH39QZhsLCQj3xxBNq2rSp0tPTy/z4atWqyd/f37qcPXtWR48e1ZNPPnnDNQEAzEf4cIH3339fDz/8sJYvX64hQ4ZY2zdv3qyTJ09q8+bNWrZsmRITE5WYmGjz2FmzZql9+/Y6cOCARo4cqWeeeUapqanXHG/ZsmXy9PTUnj17NHfuXL3++utavHixTZ/XXntNrVq10oEDBzRp0iSdPHlSPXv21IABA3To0CGtWrVK27ZtU1xc3A0954KCAj300ENKSUnRV199peDgYElSdHS0qlWrVuLSokWLEve5ePFi3XnnnercufMN1QQAcI1yc8+Hu1iwYIEmTpyodevWqUuXLjbbbr/9ds2fP18eHh5q2rSpYmJitGnTJg0bNszap1evXho5cqQkacKECZo9e7Y2b96ssLCwEscMCgrS7NmzZbFYFBYWpsOHD2v27Nk2+73//vv1/PPPW9efeuopDRkyxHqTbGhoqObNm6cuXbpo4cKFZfpU4dzcXMXExKigoECbN2+Wn5+fddvixYt14cKFEh9bqVIlu+35+flavny5XnzxxVLXAQC4NRA+TPThhx8qKytL27dvV4cOHYptb9GihTw8PKzrAQEBOnz4sE2f8PBw69cWi0X+/v7Kyrr2zbZ33323LBaLdT0iIkKzZs1SYWGhdbz27dvbPObgwYM6dOiQ9bKQJBmGoaKiIqWlpalZs2aleMa/ePjhh1W/fn19+eWX8vHxsdl2xx13lHo/v7VmzRqdO3dOsbGxN/R4AIDrcNnFRG3atFGdOnW0dOlSm3surrj6v3yLxaKioqIy97kRVatWtVnPzc3V8OHDlZKSYl0OHjyo48ePq3HjxmXad69evXTo0CHt3Lmz2LYbveyyePFi9e7dW/Xq1StTLQAA1+PMh4kaN26sWbNmKTIyUh4eHpo/f74p4+7evdtmfdeuXQoNDbU5y3K1tm3b6ujRo2rSpMlNj//MM8/orrvu0u9//3utX7/e5nLTjVx2SUtL0+bNm/Xpp5/edG0AAPMRPkx25513avPmzYqMjJSnp6fmzJnj9DHT09M1btw4DR8+XPv379cbb7yhWbNmXfMxEyZM0N133624uDg99dRTqlq1qo4ePaqNGzdaQ1N8fLy+++47vfvuu9et4dlnn1VhYaF69+6tL774Qvfee6+kG7vssnTpUgUEBCg6OrrMjwUAuB7hwwXCwsL05ZdfWs+AXC8I3KzHHntMFy5cUMeOHeXh4aExY8ZYX05bkvDwcCUlJWnixInq3LmzDMNQ48aNNWjQIGufzMzMMr1kduzYsSoqKlKvXr20YcMG/e53vyvzcykqKlJiYqIef/zxa565AQDcuiyGvZsPXCgnJ0d+fn7Kzs5W9erVbbbl5+crLS1NISEhZXq1RUUWGRmp1q1bm3KGBSiN0vwcl/SOxbyzMHDrutbf76txwykAADAV4QMAAJiKez7KuS1btri6BAAAbHDmAwAAmIrwAQAATEX4AAAApiJ8AAAAUxE+AACAqQgfAADAVIQPF7NYLFq7dm2J27/55htZLBalpKSYVpMZHn/8cfXr169Mj7neXDnKjYzjjt+nEydOqF69eqpSpYq2b9/u6nIAVCDl5n0+sr/rZup4fndscsh+MjMzdfvttztkX6jYDh48qBkzZmjbtm363//+p4YNG2rEiBEaM2ZMsb6nT5/WAw88oHvvvVeBgYHq3bu3tm7dqpYtW9r0+/jjj7Vo0SIlJyfrxx9/1IEDB9S6dWubPvn5+Xr++ee1cuVKFRQUKCoqSm+++abq1avnzKcLwI2Vm/Dhbi5evCgvLy/5+/u7upQbduU54NaQnJysunXr6r333lNQUJB27Nihp59+Wh4eHoqLi7P2++mnnxQVFaXOnTvrnXfekYeHh6pVq6aoqCht375dISEh1r55eXm69957NXDgQA0bNszuuM8995zWr1+vDz74QH5+foqLi1P//v05mwKgRGW+7LJ161b16dNHgYGB1z09PWLECFksFj7UTL98wFtcXJzGjh2r2rVrKyoqSlLxU/x79uxRmzZtVLlyZbVv314HDhwotq9PP/1UoaGhqly5srp27aply5bJYrHo559/tvbZtm2bOnfuLB8fHwUFBWn06NHKy8srsb6pU6eqdevWeuuttxQUFKQqVapo4MCBys7Otva5cqlk+vTpCgwMVFhYmCQpIyNDAwcOVI0aNVSzZk317dtX33zzjfVxhYWFGjdunGrUqKFatWrpT3/6kxzxeYYTJkzQnXfeqSpVqqhRo0aaNGmSLl26VOw5LV26VMHBwapWrZpGjhypwsJCzZw5U/7+/qpbt66mT59ebN+ZmZmKjo6Wj4+PGjVqpA8//NBm+/W+T4WFhXryyScVEhIiHx8fhYWFae7cuTf9nK/liSee0Ny5c9WlSxc1atRIjzzyiIYOHaqPP/7Y2uf8+fOKiYnRPffco2XLllk/GTghIUGjRo1Sjx49lJX164e6Pfroo5o8ebK6d+9ud8zs7GwtWbJEr7/+uu6//361a9dO77zzjnbs2KFdu3Y59fkCcF9lDh95eXlq1aqVFixYcM1+a9as0a5duxQYGHjDxZU3y5Ytk5eXl7Zv365FixYV256bm6vevXurefPmSk5O1tSpU/XCCy/Y9ElLS9Mf/vAH9evXTwcPHtTw4cM1ceJEmz4nT55Uz549NWDAAB06dEirVq3Stm3bbP77tefEiRNavXq11q1bpw0bNujAgQMaOXKkTZ9NmzYpNTVVGzdu1GeffaZLly4pKipKvr6++uqrr7R9+3ZVq1ZNPXv21MWLFyVJs2bNUmJiopYuXapt27bpxx9/1Jo1a25kCm34+voqMTFRR48e1dy5c/X3v/9ds2fPLjYXX3zxhTZs2KAVK1ZoyZIliomJ0bfffqukpCS98sor+stf/qLdu3fbPG7SpEkaMGCADh48qCFDhmjw4MH6+uuvJZXu+1RUVKT69evrgw8+0NGjRzV58mT9+c9/1urVq6/5nKpVq3bNZcSIEWWao+zsbNWsWdO6XqVKFe3YsUOLFi2SxWKx6Ttx4kQdP35cdeuW/pNjk5OTdenSJZtw0rRpUwUHB2vnzp1lqhVAxVHmyy7R0dGKjo6+Zp/vvvtOzz77rP75z38qJibmhosrb0JDQzVz5swSt7///vsqKirSkiVLVLlyZbVo0ULffvutnnnmGWuft956S2FhYXr11VclSWFhYTpy5IjNf+8JCQkaMmSIxo4dax133rx56tKlixYuXFjix5jn5+fr3Xff1R133CFJeuONNxQTE6NZs2ZZLw9VrVpVixcvtl5uee+991RUVKTFixdb/5i98847qlGjhrZs2aIePXpozpw5io+PV//+/SVJixYt0j//+c8bmUIbf/nLX6xfN2zYUC+88IJWrlypP/3pT9b2oqIiLV26VL6+vmrevLm6du2q1NRUff7557rtttsUFhamV155RZs3b1anTp2sj3vooYf01FNPSZJefvllbdy4UW+88YbefPPNUn2fKlWqpGnTplnXQ0JCtHPnTq1evVoDBw4s8Tld74bV631M9W/t2LFDq1at0vr160v9mLI6c+aMvLy8VKNGDZv2evXq6cyZM04bF4B7c/g9H0VFRXr00Uc1fvx4tWjR4rr9CwoKVFBQYF3PyclxdEm3jHbt2l1z+9dff63w8HCbcBAREWHTJzU1VR06dLBp69ixo836wYMHdejQIS1fvtzaZhiGioqKlJaWpmbNmtkdPzg42Bo8roxdVFSk1NRUa/ho2bKlzX0eBw8e1IkTJ+Tr62uzr/z8fJ08eVLZ2dnKzMy0+cPu6emp9u3b3/Sll1WrVmnevHk6efKkcnNzdfny5WJ/nBs2bGhTW7169eTh4aHbbrvNpu23lxquPPer168Eg9J8nyRpwYIFWrp0qdLT03XhwgVdvHix2M2aV2vSpMk1t5fWkSNH1LdvX02ZMkU9evRwyD4BwFEcHj5eeeUVeXp6avTo0aXqn5CQYPMfYnlWtWpVU8bJzc3V8OHD7X4PgoODb2rfVz+H3NxctWvXziboXFGnTp2bGutadu7cqSFDhmjatGmKioqSn5+fVq5cqVmzZtn0q1Spks26xWKx21ZUVOTQ+lauXKkXXnhBs2bNUkREhHx9ffXqq68Wu7xztWrVql1z+yOPPGL3kt1vHT16VN26ddPTTz9tc3bIGfz9/XXx4kX9/PPPNmc/zp4969Y3UwNwLoeGj+TkZM2dO1f79+8vdj25JPHx8Ro3bpx1PScnR0FBQY4sy200a9ZM//jHP5Sfn2/9r/rqm/bCwsL0+eef27Tt3bvXZr1t27Y6evRomf+LTk9P1+nTp6336ezatct6aaIkbdu21apVq1S3bt0SLwkEBARo9+7duu+++yRJly9fVnJystq2bVum+n5rx44datCggc39LqdOnbrh/V1t165deuyxx2zW27RpI6l036ft27frd7/7nc09MydPnrzuuDd72eU///mP7r//fsXGxtq9kdbR2rVrp0qVKmnTpk0aMGCApF/OzqWnp9s9GwQAkoPfZOyrr75SVlaWgoOD5enpKU9PT506dUrPP/+8GjZsaPcx3t7eql69us1SUf3xj3+UxWLRsGHDdPToUX3++ed67bXXbPoMHz5c//3vfzVhwgQdO3ZMq1evVmJioiRZA9+ECRO0Y8cOxcXFKSUlRcePH9cnn3xy3RtOK1eurNjYWB08eFBfffWVRo8erYEDB17zP9ghQ4aodu3a6tu3r7766iulpaVpy5YtGj16tL799ltJ0pgxYzRjxgytXbtW//3vfzVy5EibV+ZI0vz589WtW+nfqyU0NFTp6elauXKlTp48qXnz5jnkJtYrPvjgAy1dulTHjh3TlClTtGfPHuv8leb7FBoaqn379umf//ynjh07pkmTJhULifY0adLkmsu1bgY9cuSIunbtqh49emjcuHE6c+aMzpw5o++///6G5+HHH39USkqKjh49KumXYJGSkmK9n8PPz09PPvmkxo0bp82bNys5OVlDhw5VRESE7r777hseF0D55tDw8eijj+rQoUNKSUmxLoGBgRo/frxDbjAs76pVq6Z169bp8OHDatOmjSZOnKhXXnnFpk9ISIg+/PBDffzxxwoPD9fChQut//17e3tLksLDw5WUlKRjx46pc+fOatOmjSZPnnzdVx41adJE/fv3V69evdSjRw+Fh4frzTffvOZjqlSpoq1btyo4OFj9+/dXs2bN9OSTTyo/P98aJJ9//nk9+uijio2NtV6CePDBB23287///a9UZwau+P3vf6/nnntOcXFxat26tXbs2KFJkyaV+vHXM23aNK1cuVLh4eF69913tWLFCjVv3lxS6b5Pw4cPV//+/TVo0CB16tRJP/zwQ7FXDjnahx9+qO+//17vvfeeAgICrMvV9wiVxaeffqo2bdpYbxwfPHiw2rRpY3PpZ/bs2erdu7cGDBig++67T/7+/jYv7wWAq1mMMt71l5ubqxMnTkiS2rRpo9dff11du3ZVzZo17d5P0LBhQ40dO9b6yovrycnJkZ+fn7Kzs4udBcnPz1daWppCQkJKfMVGRTR9+nQtWrRIGRkZN7yPqVOnau3atW719uBwT6X5OR60Pstu+6qY0r8MGIC5rvX3+2plvudj37596tq1q3X9yv0asbGx1tP/cK4333xTHTp0UK1atbR9+3a9+uqr172kAgDAraLM4SMyMrJML5H87TtdwjGOHz+uv/71r/rxxx8VHBys559/XvHx8a4uCwCAUinzZRdn47ILUL5x2QUon8py2cWhN5wCAABcD+EDAACYyi3Dxy12pQhAGfDzC8CtwseVt8U+f/68iysBcKOu/Pxe/Tb3ACoOh3+2izN5eHioRo0a1g8Bq1KlSqnfxh2AaxmGofPnzysrK0s1atSQh4eHq0sC4CJuFT4kWd/q++pPIQXgHmrUqMGHzgEVnNuFD4vFooCAANWtW1eXLl1ydTkAyqBSpUqc8QDgfuHjCg8PD36JAQDghtzqhlMAAOD+CB8AAMBUhA8AAGAqwgcAADAV4QMAAJiK8AEAAEzlti+1BeD+MhNLeIfiOmfNLQSAqTjzAQAATEX4AAAApiJ8AAAAUxE+AACAqQgfAADAVIQPAABgKsIHAAAwFeEDAACYivABAABMRfgAAACmInwAAABTET4AAICpCB8AAMBUhA8AAGAqwgcAADAV4QMAAJiK8AEAAExV5vCxdetW9enTR4GBgbJYLFq7dq1126VLlzRhwgS1bNlSVatWVWBgoB577DGdPn3akTUDAAA3VubwkZeXp1atWmnBggXFtp0/f1779+/XpEmTtH//fn388cdKTU3V73//e4cUCwAA3J9nWR8QHR2t6Ohou9v8/Py0ceNGm7b58+erY8eOSk9PV3Bw8I1VCQAAyo0yh4+yys7OlsViUY0aNexuLygoUEFBgXU9JyfH2SUBAAAXcmr4yM/P14QJE/Twww+revXqdvskJCRo2rRpziwDQDmR/V03u+1+d2wyuRIAN8Npr3a5dOmSBg4cKMMwtHDhwhL7xcfHKzs727pkZGQ4qyQAAHALcMqZjyvB49SpU/ryyy9LPOshSd7e3vL29nZGGQAA4Bbk8PBxJXgcP35cmzdvVq1atRw9BAAAcGNlDh+5ubk6ceKEdT0tLU0pKSmqWbOmAgIC9Ic//EH79+/XZ599psLCQp05c0aSVLNmTXl5eTmucgAA4JbKHD727dunrl27WtfHjRsnSYqNjdXUqVP16aefSpJat25t87jNmzcrMjLyxisFAADlQpnDR2RkpAzDKHH7tbYBAADw2S4AAMBUhA8AAGAqwgcAADAV4QMAAJiK8AEAAExF+AAAAKYifAAAAFMRPgAAgKkIHwAAwFSEDwAAYCrCBwAAMBXhAwAAmIrwAQAATEX4AAAApiJ8AAAAUxE+AACAqQgfAADAVIQPAABgKsIHAAAwFeEDAACYivABAABMRfgAAACmInwAAABTET4AAICpCB8AAMBUhA8AAGAqwgcAADAV4QMAAJiK8AEAAExF+AAAAKYifAAAAFMRPgAAgKkIHwAAwFRlDh9bt25Vnz59FBgYKIvForVr19psNwxDkydPVkBAgHx8fNS9e3cdP37cUfUCAAA3V+bwkZeXp1atWmnBggV2t8+cOVPz5s3TokWLtHv3blWtWlVRUVHKz8+/6WIBAID78yzrA6KjoxUdHW13m2EYmjNnjv7yl7+ob9++kqR3331X9erV09q1azV48OCbqxYAALg9h97zkZaWpjNnzqh79+7WNj8/P3Xq1Ek7d+60+5iCggLl5OTYLAAAoPwq85mPazlz5owkqV69ejbt9erVs267WkJCgqZNm+bIMgD8f4PWZ9ltXxVT1+RKAOBXLn+1S3x8vLKzs61LRkaGq0sCAABO5NDw4e/vL0k6e/asTfvZs2et267m7e2t6tWr2ywAAKD8cmj4CAkJkb+/vzZt2mRty8nJ0e7duxUREeHIoQAAgJsq8z0fubm5OnHihHU9LS1NKSkpqlmzpoKDgzV27Fj99a9/VWhoqEJCQjRp0iQFBgaqX79+jqwbAAC4qTKHj3379qlr167W9XHjxkmSYmNjlZiYqD/96U/Ky8vT008/rZ9//ln33nuvNmzYoMqVKzuuagAA4LbKHD4iIyNlGEaJ2y0Wi1566SW99NJLN1UYAAAon1z+ahcAAFCxED4AAICpCB8AAMBUhA8AAGAqwgcAADAV4QMAAJiK8AEAAExF+AAAAKYifAAAAFMRPgAAgKkIHwAAwFSEDwAAYCrCBwAAMBXhAwAAmIrwAQAATOXp6gIAlC/Z33Wz2+53xyaTKwFwq+LMBwAAMBXhAwAAmIrwAQAATEX4AAAApiJ8AAAAUxE+AACAqQgfAADAVIQPAABgKsIHAAAwFeEDAACYivABAABMRfgAAACmInwAAABTET4AAICpCB8AAMBUhA8AAGAqwgcAADCVw8NHYWGhJk2apJCQEPn4+Khx48Z6+eWXZRiGo4cCAABuyNPRO3zllVe0cOFCLVu2TC1atNC+ffs0dOhQ+fn5afTo0Y4eDgAAuBmHh48dO3aob9++iomJkSQ1bNhQK1as0J49exw9FAAAcEMOv+zyu9/9Tps2bdKxY8ckSQcPHtS2bdsUHR1tt39BQYFycnJsFgAAUH45/MzHiy++qJycHDVt2lQeHh4qLCzU9OnTNWTIELv9ExISNG3aNEeXgVtAZqLFbnvA4/bv/8n+rpvddr87NjmsptIoqe6xdc7abX+79cN226+uu6T9Sjc/JyXuu4SayzLXZf0+AsD1OPzMx+rVq7V8+XK9//772r9/v5YtW6bXXntNy5Yts9s/Pj5e2dnZ1iUjI8PRJQEAgFuIw898jB8/Xi+++KIGDx4sSWrZsqVOnTqlhIQExcbGFuvv7e0tb29vR5cBAABuUQ4/83H+/Hnddpvtbj08PFRUVOTooQAAgBty+JmPPn36aPr06QoODlaLFi104MABvf7663riiSccPRQAAHBDDg8fb7zxhiZNmqSRI0cqKytLgYGBGj58uCZPnuzooQAAgBtyePjw9fXVnDlzNGfOHEfvGgAAlAN8tgsAADAV4QMAAJiK8AEAAExF+AAAAKYifAAAAFMRPgAAgKkIHwAAwFSEDwAAYCrCBwAAMBXhAwAAmIrwAQAATEX4AAAApiJ8AAAAUxE+AACAqQgfAADAVIQPAABgKsIHAAAwFeEDAACYivABAABMRfgAAACmInwAAABTET4AAICpCB8AAMBUhA8AAGAqwgcAADAV4QMAAJiK8AEAAExF+AAAAKYifAAAAFMRPgAAgKkIHwAAwFSEDwAAYCrCBwAAMJVTwsd3332nRx55RLVq1ZKPj49atmypffv2OWMoAADgZjwdvcOffvpJ99xzj7p27aovvvhCderU0fHjx3X77bc7eigAAOCGHB4+XnnlFQUFBemdd96xtoWEhDh6GAAA4KYcftnl008/Vfv27fXQQw+pbt26atOmjf7+97+X2L+goEA5OTk2CwAAKL8cfubj//7v/7Rw4UKNGzdOf/7zn7V3716NHj1aXl5eio2NLdY/ISFB06ZNc3QZuIUNWp9lt/3t1s7b96qYuje/cydy5pw4izvW7K7c9bgGSuLwMx9FRUVq27at/va3v6lNmzZ6+umnNWzYMC1atMhu//j4eGVnZ1uXjIwMR5cEAABuIQ4PHwEBAWrevLlNW7NmzZSenm63v7e3t6pXr26zAACA8svh4eOee+5RamqqTduxY8fUoEEDRw8FAADckMPDx3PPPaddu3bpb3/7m06cOKH3339fb7/9tkaNGuXooQAAgBtyePjo0KGD1qxZoxUrVuiuu+7Syy+/rDlz5mjIkCGOHgoAALghh7/aRZJ69+6t3r17O2PXAADAzfHZLgAAwFSEDwAAYCrCBwAAMBXhAwAAmIrwAQAATEX4AAAApiJ8AAAAUxE+AACAqQgfAADAVIQPAABgKsIHAAAwFeEDAACYivABAABMRfgAAACmInwAAABTebq6AABwhsxEi932sXXO2m1/u/XDdtv97thUqv0GPG7Ybc/+rltJJRbbtzM5qu6y1DxofZbd9lUxdUu9D2dxxXzgV5z5AAAApiJ8AAAAUxE+AACAqQgfAADAVIQPAABgKsIHAAAwFeEDAACYivABAABMRfgAAACmInwAAABTET4AAICpCB8AAMBUhA8AAGAqwgcAADAV4QMAAJiK8AEAAExF+AAAAKZyeviYMWOGLBaLxo4d6+yhAACAG3Bq+Ni7d6/eeusthYeHO3MYAADgRpwWPnJzczVkyBD9/e9/1+233+6sYQAAgJtxWvgYNWqUYmJi1L1792v2KygoUE5Ojs0CAADKL09n7HTlypXav3+/9u7de92+CQkJmjZtmjPKQDmWmWixv6HOWbvN2d91s9vud8cmR5UE3DRXHNeD1mfZbX+7dal3Ua5+Hh0xH7g+h5/5yMjI0JgxY7R8+XJVrlz5uv3j4+OVnZ1tXTIyMhxdEgAAuIU4/MxHcnKysrKy1LZtW2tbYWGhtm7dqvnz56ugoEAeHh7Wbd7e3vL29nZ0GQAA4Bbl8PDRrVs3HT582KZt6NChatq0qSZMmGATPAAAQMXj8PDh6+uru+66y6atatWqqlWrVrF2AABQ8fAOpwAAwFROebXL1bZs2WLGMAAAwA1w5gMAAJiK8AEAAExF+AAAAKYifAAAAFMRPgAAgKkIHwAAwFSEDwAAYCrCBwAAMBXhAwAAmIrwAQAATEX4AAAApiJ8AAAAUxE+AACAqQgfAADAVIQPAABgKsIHAAAwFeEDAACYivABAABMRfgAAACmInwAAABTET4AAICpCB8AAMBUhA8AAGAqwgcAADAV4QMAAJiK8AEAAExF+AAAAKYifAAAAFMRPgAAgKkIHwAAwFSEDwAAYCrCBwAAMBXhAwAAmMrh4SMhIUEdOnSQr6+v6tatq379+ik1NdXRwwAAADfl8PCRlJSkUaNGadeuXdq4caMuXbqkHj16KC8vz9FDAQAAN+Tp6B1u2LDBZj0xMVF169ZVcnKy7rvvPkcPBwAA3IzDw8fVsrOzJUk1a9a0u72goEAFBQXW9ZycHGeXBAAAXMip4aOoqEhjx47VPffco7vuustun4SEBE2bNs2ZZdjITLTYbQ943LDbnv1dN7vtfndsclhNpVFS3WPrnLXb/nbrh+22l7buQeuz7LaviqlbqscDFU1JPzNvtza3Drie2b+v3ZFTX+0yatQoHTlyRCtXriyxT3x8vLKzs61LRkaGM0sCAAAu5rQzH3Fxcfrss8+0detW1a9fv8R+3t7e8vb2dlYZAADgFuPw8GEYhp599lmtWbNGW7ZsUUhIiKOHAAAAbszh4WPUqFF6//339cknn8jX11dnzpyRJPn5+cnHx8fRwwEAADfj8Hs+Fi5cqOzsbEVGRiogIMC6rFq1ytFDAQAAN+SUyy4AAAAl4bNdAACAqQgfAADAVIQPAABgKsIHAAAwFeEDAACYivABAABMRfgAAACmInwAAABTET4AAICpCB8AAMBUhA8AAGAqwgcAADAV4QMAAJiK8AEAAExF+AAAAKbydHUBt4pB67Pstr/d2nn7XhVT9+Z3fpMyEy32N9Q5a7c5+7tuJe7L745NjigJACSV/PtpbAm/nyTp7dYP220vL7+fyjont+p8cOYDAACYivABAABMRfgAAACmInwAAABTET4AAICpCB8AAMBUhA8AAGAqwgcAADAV4QMAAJiK8AEAAExF+AAAAKYifAAAAFMRPgAAgKkIHwAAwFSEDwAAYCrCBwAAMBXhAwAAmMpp4WPBggVq2LChKleurE6dOmnPnj3OGgoAALgRp4SPVatWady4cZoyZYr279+vVq1aKSoqSllZWc4YDgAAuBGnhI/XX39dw4YN09ChQ9W8eXMtWrRIVapU0dKlS50xHAAAcCOejt7hxYsXlZycrPj4eGvbbbfdpu7du2vnzp3F+hcUFKigoMC6np2dLUnKyclxdGmSpHMX7LdfOn/ObnvOuct22y1lqK/EfedULvU+nFW3o/brzH3bm2t33HdJ+3Xmvm/l+XDXffMzc+vOtTP3fSvMh7P3fbOu/N02DOP6nQ0H++677wxJxo4dO2zax48fb3Ts2LFY/ylTphiSWFhYWFhYWMrBkpGRcd2s4PAzH2UVHx+vcePGWdeLior0448/qlatWrJYLKXaR05OjoKCgpSRkaHq1as7q9RbHvPwK+biV8zFL5iHXzEXv2AefuWIuTAMQ+fOnVNgYOB1+zo8fNSuXVseHh46e/asTfvZs2fl7+9frL+3t7e8vb1t2mrUqHFDY1evXr3CH0AS8/BbzMWvmItfMA+/Yi5+wTz86mbnws/Pr1T9HH7DqZeXl9q1a6dNmzZZ24qKirRp0yZFREQ4ejgAAOBmnHLZZdy4cYqNjVX79u3VsWNHzZkzR3l5eRo6dKgzhgMAAG7EKeFj0KBB+v777zV58mSdOXNGrVu31oYNG1SvXj1nDCdvb29NmTKl2OWbioZ5+BVz8Svm4hfMw6+Yi18wD78yey4shlGa18QAAAA4Bp/tAgAATEX4AAAApiJ8AAAAUxE+AACAqdw2fEydOlUWi8Vmadq0qavLMsXWrVvVp08fBQYGymKxaO3atTbbDcPQ5MmTFRAQIB8fH3Xv3l3Hjx93TbFOdr25ePzxx4sdJz179nRNsU6UkJCgDh06yNfXV3Xr1lW/fv2Umppq0yc/P1+jRo1SrVq1VK1aNQ0YMKDYmwG6u9LMQ2RkZLFjYsSIES6q2HkWLlyo8PBw65tGRURE6IsvvrBurwjHwxXXm4uKckxcbcaMGbJYLBo7dqy1zazjwm3DhyS1aNFCmZmZ1mXbtm2uLskUeXl5atWqlRYsWGB3+8yZMzVv3jwtWrRIu3fvVtWqVRUVFaX8/HyTK3W+682FJPXs2dPmOFmxYoWJFZojKSlJo0aN0q5du7Rx40ZdunRJPXr0UF5enrXPc889p3Xr1umDDz5QUlKSTp8+rf79+7uwascrzTxI0rBhw2yOiZkzZ7qoYuepX7++ZsyYoeTkZO3bt0/333+/+vbtq//85z+SKsbxcMX15kKqGMfEb+3du1dvvfWWwsPDbdpNOy4c8mlyLjBlyhSjVatWri7D5SQZa9assa4XFRUZ/v7+xquvvmpt+/nnnw1vb29jxYoVLqjQPFfPhWEYRmxsrNG3b1+X1ONKWVlZhiQjKSnJMIxfjoFKlSoZH3zwgbXP119/bUgydu7c6aoyne7qeTAMw+jSpYsxZswY1xXlQrfffruxePHiCns8/NaVuTCMindMnDt3zggNDTU2btxo89zNPC7c+szH8ePHFRgYqEaNGmnIkCFKT093dUkul5aWpjNnzqh79+7WNj8/P3Xq1Ek7d+50YWWus2XLFtWtW1dhYWF65pln9MMPP7i6JKfLzs6WJNWsWVOSlJycrEuXLtkcF02bNlVwcHC5Pi6unocrli9frtq1a+uuu+5SfHy8zp8/74ryTFNYWKiVK1cqLy9PERERFfZ4kIrPxRUV6ZgYNWqUYmJibL7/krm/J1z+qbY3qlOnTkpMTFRYWJgyMzM1bdo0de7cWUeOHJGvr6+ry3OZM2fOSFKxd5OtV6+edVtF0rNnT/Xv318hISE6efKk/vznPys6Olo7d+6Uh4eHq8tziqKiIo0dO1b33HOP7rrrLkm/HBdeXl7FPrSxPB8X9uZBkv74xz+qQYMGCgwM1KFDhzRhwgSlpqbq448/dmG1znH48GFFREQoPz9f1apV05o1a9S8eXOlpKRUuOOhpLmQKtYxsXLlSu3fv1979+4tts3M3xNuGz6io6OtX4eHh6tTp05q0KCBVq9erSeffNKFleFWMnjwYOvXLVu2VHh4uBo3bqwtW7aoW7duLqzMeUaNGqUjR45UmHugSlLSPDz99NPWr1u2bKmAgAB169ZNJ0+eVOPGjc0u06nCwsKUkpKi7Oxsffjhh4qNjVVSUpKry3KJkuaiefPmFeaYyMjI0JgxY7Rx40ZVrlzZpbW49WWX36pRo4buvPNOnThxwtWluJS/v78kFbs7+ezZs9ZtFVmjRo1Uu3btcnucxMXF6bPPPtPmzZtVv359a7u/v78uXryon3/+2aZ/eT0uSpoHezp16iRJ5fKY8PLyUpMmTdSuXTslJCSoVatWmjt3boU7HqSS58Ke8npMJCcnKysrS23btpWnp6c8PT2VlJSkefPmydPTU/Xq1TPtuCg34SM3N1cnT55UQECAq0txqZCQEPn7+2vTpk3WtpycHO3evdvm+mZF9e233+qHH34od8eJYRiKi4vTmjVr9OWXXyokJMRme7t27VSpUiWb4yI1NVXp6enl6ri43jzYk5KSIknl7piwp6ioSAUFBRXmeLiWK3NhT3k9Jrp166bDhw8rJSXFurRv315Dhgyxfm3aceHQ21dN9Pzzzxtbtmwx0tLSjO3btxvdu3c3ateubWRlZbm6NKc7d+6cceDAAePAgQOGJOP11183Dhw4YJw6dcowDMOYMWOGUaNGDeOTTz4xDh06ZPTt29cICQkxLly44OLKHe9ac3Hu3DnjhRdeMHbu3GmkpaUZ//73v422bdsaoaGhRn5+vqtLd6hnnnnG8PPzM7Zs2WJkZmZal/Pnz1v7jBgxwggODja+/PJLY9++fUZERIQRERHhwqod73rzcOLECeOll14y9u3bZ6SlpRmffPKJ0ahRI+O+++5zceWO9+KLLxpJSUlGWlqacejQIePFF180LBaL8a9//cswjIpxPFxxrbmoSMeEPVe/0ses48Jtw8egQYOMgIAAw8vLy7jjjjuMQYMGGSdOnHB1WabYvHmzIanYEhsbaxjGLy+3nTRpklGvXj3D29vb6Natm5Gamuraop3kWnNx/vx5o0ePHkadOnWMSpUqGQ0aNDCGDRtmnDlzxtVlO5y9OZBkvPPOO9Y+Fy5cMEaOHGncfvvtRpUqVYwHH3zQyMzMdF3RTnC9eUhPTzfuu+8+o2bNmoa3t7fRpEkTY/z48UZ2drZrC3eCJ554wmjQoIHh5eVl1KlTx+jWrZs1eBhGxTgerrjWXFSkY8Keq8OHWceFxTAMw7HnUgAAAEpWbu75AAAA7oHwAQAATEX4AAAApiJ8AAAAUxE+AACAqQgfAADAVIQPAABgKsIHAAAwFeEDAACYivABAABMRfgAAACmInwAAABT/T9Sa1sXZ8KeKwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "### Code - Plot actual vs predicted labels\n",
    "colors = ['#E69F00', '#56B4E9', '#F0E442']\n",
    "names = ['Actual labels', 'knn pre. k=7', 'ridge pred. lambda = 2^-5']\n",
    "         \n",
    "# Make the histogram using a list of lists\n",
    "# Normalize the flights and assign colors and names\n",
    "plt.hist([boston_label_test['MEDV'], knn_prediction[0], ridge_prediction[0]], bins = 20,\n",
    "         color = colors, label=names)\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<font color = skyblue> Seems like the predicted values did well. Ridge regression could be over fitted as it is predictin so well. Also the regylarisation term is very small so we didn't introduce mush bias. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "________________\n",
    "## <font color = dimgrey> BONUS: Feature selection - most useful features in predicting the disease </font>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "You can stop here and get the \"pass\" grade! To get the \"pass with honors\" grade, you need to do the following exercise. This means you'll get one bonus point for the exam.\n",
    "\n",
    "The exercise is not as straightforward as the previous ones, and may require you to do some research of your own. You are also required to **explain** the steps you choose with your own words, and show that you tried to understand the idea behind the task. There's no single correct solution for this so just explain me what you did and why you did it.\n",
    "\n",
    "----------------\n",
    "\n",
    "\n",
    "Unfortunately, due to the lack of resources and time, doctors are not able to measure all the values represented in the given dataset. Luckily, keen students are ready to help: You should now find <font color = darkorchid><b>five [5] most useful features</b></font> in predicting the presence of the cardiovascular disease among the ones in the given cardio data set. \n",
    "\n",
    "Use the selected features to train and test a model of your choice. Evaluate the performance by computing the accuracy and drawing a ROC curve. Remember not to use any information from the test set when selecting the features!\n",
    "\n",
    "**Discuss** your results."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "### Code - BONUS!"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<font color = red> \\*** Answer here *** </font>"
   ]
  }
 ],
 "metadata": {
  "colab": {
   "name": "DADK2020_ex3_Valtteri_Nieminen.ipynb",
   "provenance": []
  },
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.6"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {
    "height": "180px",
    "width": "160px"
   },
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": true,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {},
   "toc_section_display": true,
   "toc_window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}