Boston (1).ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Linear Regression Models on Boston Housing Dataset: Predicting Median Home Value and Nitric Oxides Concentration"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Frame the problem and look at the big picture:\n",
    "\n",
    "#### \"The name for this dataset is simply boston. It has two prototasks: nox, in which the nitrous oxide level is to be predicted; and price, in which the median value of a home is to be predicted.\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Get the Data.\n",
    "\n",
    "#### For this dataset, we are given a messy information collected by the U.S Census Service concerning housing in the area of Boston Mass. It was obtained from the StatLib archive (http://lib.stat.cmu.edu/datasets/boston). When downloaded to a TextEdit file, the data saves as an RTF, and when read into jupyter notebooks the data comes out noisey in a nested dictionary of sorts. In order to turn this disaster of messy data into a clean, useful dataframe, we must parse our data with a regex delimiter to get all the numeric instances from the data. The code in the next few cells shows how this was accomplished. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "result = []\n",
    "\n",
    "\n",
    "with open('/Users/ryanmartin/Desktop/Boston.rtf') as f:\n",
    "    for line in f:\n",
    "        line = line.split(\" (\")  # remove '\\n'\n",
    "        result.append(line)\n",
    "\n",
    "import re\n",
    "\n",
    "inilist = result\n",
    "regex = re.findall(r'[-+]?\\d*\\.\\d+|\\d+', str(result))\n",
    "data = regex[33:]\n",
    "#print(data)"
   ]
  },
  {
   "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>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>18.00</td>\n",
       "      <td>2.310</td>\n",
       "      <td>0</td>\n",
       "      <td>0.5380</td>\n",
       "      <td>6.5750</td>\n",
       "      <td>65.20</td>\n",
       "      <td>4.0900</td>\n",
       "      <td>1</td>\n",
       "      <td>296.0</td>\n",
       "      <td>15.30</td>\n",
       "      <td>396.90</td>\n",
       "      <td>4.98</td>\n",
       "      <td>24.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.02731</td>\n",
       "      <td>0.00</td>\n",
       "      <td>7.070</td>\n",
       "      <td>0</td>\n",
       "      <td>0.4690</td>\n",
       "      <td>6.4210</td>\n",
       "      <td>78.90</td>\n",
       "      <td>4.9671</td>\n",
       "      <td>2</td>\n",
       "      <td>242.0</td>\n",
       "      <td>17.80</td>\n",
       "      <td>396.90</td>\n",
       "      <td>9.14</td>\n",
       "      <td>21.60</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.02729</td>\n",
       "      <td>0.00</td>\n",
       "      <td>7.070</td>\n",
       "      <td>0</td>\n",
       "      <td>0.4690</td>\n",
       "      <td>7.1850</td>\n",
       "      <td>61.10</td>\n",
       "      <td>4.9671</td>\n",
       "      <td>2</td>\n",
       "      <td>242.0</td>\n",
       "      <td>17.80</td>\n",
       "      <td>392.83</td>\n",
       "      <td>4.03</td>\n",
       "      <td>34.70</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.03237</td>\n",
       "      <td>0.00</td>\n",
       "      <td>2.180</td>\n",
       "      <td>0</td>\n",
       "      <td>0.4580</td>\n",
       "      <td>6.9980</td>\n",
       "      <td>45.80</td>\n",
       "      <td>6.0622</td>\n",
       "      <td>3</td>\n",
       "      <td>222.0</td>\n",
       "      <td>18.70</td>\n",
       "      <td>394.63</td>\n",
       "      <td>2.94</td>\n",
       "      <td>33.40</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.06905</td>\n",
       "      <td>0.00</td>\n",
       "      <td>2.180</td>\n",
       "      <td>0</td>\n",
       "      <td>0.4580</td>\n",
       "      <td>7.1470</td>\n",
       "      <td>54.20</td>\n",
       "      <td>6.0622</td>\n",
       "      <td>3</td>\n",
       "      <td>222.0</td>\n",
       "      <td>18.70</td>\n",
       "      <td>396.90</td>\n",
       "      <td>5.33</td>\n",
       "      <td>36.20</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  18.00  2.310    0  0.5380  6.5750  65.20  4.0900   1  296.0   \n",
       "1  0.02731   0.00  7.070    0  0.4690  6.4210  78.90  4.9671   2  242.0   \n",
       "2  0.02729   0.00  7.070    0  0.4690  7.1850  61.10  4.9671   2  242.0   \n",
       "3  0.03237   0.00  2.180    0  0.4580  6.9980  45.80  6.0622   3  222.0   \n",
       "4  0.06905   0.00  2.180    0  0.4580  7.1470  54.20  6.0622   3  222.0   \n",
       "\n",
       "  PTRATIO       B LSTAT   MEDV  \n",
       "0   15.30  396.90  4.98  24.00  \n",
       "1   17.80  396.90  9.14  21.60  \n",
       "2   17.80  392.83  4.03  34.70  \n",
       "3   18.70  394.63  2.94  33.40  \n",
       "4   18.70  396.90  5.33  36.20  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "i=0\n",
    "data_list=[]\n",
    "while i<len(data):\n",
    "  data_list.append(data[i:i+14])\n",
    "  i+=14\n",
    "\n",
    "data = pd.DataFrame(data_list, columns = ['CRIM', 'ZN', 'INDUS', 'CHAS',\n",
    "                              'NOX', 'RM', 'AGE', 'DIS', 'RAD',\n",
    "                              'TAX', 'PTRATIO', 'B', 'LSTAT', 'MEDV' ])\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Explore the data to get insights:\n",
    "\n",
    "#### Luckily this dataset will not need much feature engineering because there are no missing values (see cell 5). But in our data retrieval processes we have accidentally converted these features to objects, when they are actually numeric values. This will be easily handled with a simple apply() function. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "CRIM       0\n",
       "ZN         0\n",
       "INDUS      0\n",
       "CHAS       0\n",
       "NOX        0\n",
       "RM         0\n",
       "AGE        0\n",
       "DIS        0\n",
       "RAD        0\n",
       "TAX        0\n",
       "PTRATIO    0\n",
       "B          0\n",
       "LSTAT      0\n",
       "MEDV       0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.isnull().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### We will then explore the data by studying its attributes and characteristics. By calling the pandas info function we can see all the datatypes present in each feature in the data. This information will be useful during the feature engineering stage, where we must prepare the data for the machine learning models. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 506 entries, 0 to 505\n",
      "Data columns (total 14 columns):\n",
      "CRIM       506 non-null object\n",
      "ZN         506 non-null object\n",
      "INDUS      506 non-null object\n",
      "CHAS       506 non-null object\n",
      "NOX        506 non-null object\n",
      "RM         506 non-null object\n",
      "AGE        506 non-null object\n",
      "DIS        506 non-null object\n",
      "RAD        506 non-null object\n",
      "TAX        506 non-null object\n",
      "PTRATIO    506 non-null object\n",
      "B          506 non-null object\n",
      "LSTAT      506 non-null object\n",
      "MEDV       506 non-null object\n",
      "dtypes: object(14)\n",
      "memory usage: 55.4+ KB\n"
     ]
    }
   ],
   "source": [
    "data.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>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>count</th>\n",
       "      <td>506</td>\n",
       "      <td>506</td>\n",
       "      <td>506</td>\n",
       "      <td>506</td>\n",
       "      <td>506</td>\n",
       "      <td>506</td>\n",
       "      <td>506</td>\n",
       "      <td>506</td>\n",
       "      <td>506</td>\n",
       "      <td>506</td>\n",
       "      <td>506</td>\n",
       "      <td>506</td>\n",
       "      <td>506</td>\n",
       "      <td>506</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>unique</th>\n",
       "      <td>504</td>\n",
       "      <td>26</td>\n",
       "      <td>76</td>\n",
       "      <td>2</td>\n",
       "      <td>81</td>\n",
       "      <td>446</td>\n",
       "      <td>356</td>\n",
       "      <td>412</td>\n",
       "      <td>9</td>\n",
       "      <td>66</td>\n",
       "      <td>46</td>\n",
       "      <td>357</td>\n",
       "      <td>455</td>\n",
       "      <td>229</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>top</th>\n",
       "      <td>14.33370</td>\n",
       "      <td>0.00</td>\n",
       "      <td>18.100</td>\n",
       "      <td>0</td>\n",
       "      <td>0.5380</td>\n",
       "      <td>6.2290</td>\n",
       "      <td>100.00</td>\n",
       "      <td>3.4952</td>\n",
       "      <td>24</td>\n",
       "      <td>666.0</td>\n",
       "      <td>20.20</td>\n",
       "      <td>396.90</td>\n",
       "      <td>8.05</td>\n",
       "      <td>50.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>freq</th>\n",
       "      <td>2</td>\n",
       "      <td>372</td>\n",
       "      <td>132</td>\n",
       "      <td>471</td>\n",
       "      <td>23</td>\n",
       "      <td>3</td>\n",
       "      <td>43</td>\n",
       "      <td>5</td>\n",
       "      <td>132</td>\n",
       "      <td>132</td>\n",
       "      <td>140</td>\n",
       "      <td>121</td>\n",
       "      <td>3</td>\n",
       "      <td>16</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            CRIM    ZN   INDUS CHAS     NOX      RM     AGE     DIS  RAD  \\\n",
       "count        506   506     506  506     506     506     506     506  506   \n",
       "unique       504    26      76    2      81     446     356     412    9   \n",
       "top     14.33370  0.00  18.100    0  0.5380  6.2290  100.00  3.4952   24   \n",
       "freq           2   372     132  471      23       3      43       5  132   \n",
       "\n",
       "          TAX PTRATIO       B LSTAT   MEDV  \n",
       "count     506     506     506   506    506  \n",
       "unique     66      46     357   455    229  \n",
       "top     666.0   20.20  396.90  8.05  50.00  \n",
       "freq      132     140     121     3     16  "
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = data.apply(pd.to_numeric)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### In our data description, we see that we have a almost a perfect bell curve distribution of \"MEDV\" which are the median home values; however, the \"NOX\" values are much more sparse and sporatically distributed. "
   ]
  },
  {
   "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>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>count</th>\n",
       "      <td>506.000000</td>\n",
       "      <td>506.000000</td>\n",
       "      <td>506.000000</td>\n",
       "      <td>506.000000</td>\n",
       "      <td>506.000000</td>\n",
       "      <td>506.000000</td>\n",
       "      <td>506.000000</td>\n",
       "      <td>506.000000</td>\n",
       "      <td>506.000000</td>\n",
       "      <td>506.000000</td>\n",
       "      <td>506.000000</td>\n",
       "      <td>506.000000</td>\n",
       "      <td>506.000000</td>\n",
       "      <td>506.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>3.613524</td>\n",
       "      <td>11.363636</td>\n",
       "      <td>11.136779</td>\n",
       "      <td>0.069170</td>\n",
       "      <td>0.554695</td>\n",
       "      <td>6.284634</td>\n",
       "      <td>68.574901</td>\n",
       "      <td>3.795043</td>\n",
       "      <td>9.549407</td>\n",
       "      <td>408.237154</td>\n",
       "      <td>18.455534</td>\n",
       "      <td>356.674032</td>\n",
       "      <td>12.653063</td>\n",
       "      <td>22.532806</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>8.601545</td>\n",
       "      <td>23.322453</td>\n",
       "      <td>6.860353</td>\n",
       "      <td>0.253994</td>\n",
       "      <td>0.115878</td>\n",
       "      <td>0.702617</td>\n",
       "      <td>28.148861</td>\n",
       "      <td>2.105710</td>\n",
       "      <td>8.707259</td>\n",
       "      <td>168.537116</td>\n",
       "      <td>2.164946</td>\n",
       "      <td>91.294864</td>\n",
       "      <td>7.141062</td>\n",
       "      <td>9.197104</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.006320</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.460000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.385000</td>\n",
       "      <td>3.561000</td>\n",
       "      <td>2.900000</td>\n",
       "      <td>1.129600</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>187.000000</td>\n",
       "      <td>12.600000</td>\n",
       "      <td>0.320000</td>\n",
       "      <td>1.730000</td>\n",
       "      <td>5.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>0.082045</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>5.190000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.449000</td>\n",
       "      <td>5.885500</td>\n",
       "      <td>45.025000</td>\n",
       "      <td>2.100175</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>279.000000</td>\n",
       "      <td>17.400000</td>\n",
       "      <td>375.377500</td>\n",
       "      <td>6.950000</td>\n",
       "      <td>17.025000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>0.256510</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>9.690000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.538000</td>\n",
       "      <td>6.208500</td>\n",
       "      <td>77.500000</td>\n",
       "      <td>3.207450</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>330.000000</td>\n",
       "      <td>19.050000</td>\n",
       "      <td>391.440000</td>\n",
       "      <td>11.360000</td>\n",
       "      <td>21.200000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>3.677082</td>\n",
       "      <td>12.500000</td>\n",
       "      <td>18.100000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.624000</td>\n",
       "      <td>6.623500</td>\n",
       "      <td>94.075000</td>\n",
       "      <td>5.188425</td>\n",
       "      <td>24.000000</td>\n",
       "      <td>666.000000</td>\n",
       "      <td>20.200000</td>\n",
       "      <td>396.225000</td>\n",
       "      <td>16.955000</td>\n",
       "      <td>25.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>88.976200</td>\n",
       "      <td>100.000000</td>\n",
       "      <td>27.740000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.871000</td>\n",
       "      <td>8.780000</td>\n",
       "      <td>100.000000</td>\n",
       "      <td>12.126500</td>\n",
       "      <td>24.000000</td>\n",
       "      <td>711.000000</td>\n",
       "      <td>22.000000</td>\n",
       "      <td>396.900000</td>\n",
       "      <td>37.970000</td>\n",
       "      <td>50.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             CRIM          ZN       INDUS        CHAS         NOX          RM  \\\n",
       "count  506.000000  506.000000  506.000000  506.000000  506.000000  506.000000   \n",
       "mean     3.613524   11.363636   11.136779    0.069170    0.554695    6.284634   \n",
       "std      8.601545   23.322453    6.860353    0.253994    0.115878    0.702617   \n",
       "min      0.006320    0.000000    0.460000    0.000000    0.385000    3.561000   \n",
       "25%      0.082045    0.000000    5.190000    0.000000    0.449000    5.885500   \n",
       "50%      0.256510    0.000000    9.690000    0.000000    0.538000    6.208500   \n",
       "75%      3.677082   12.500000   18.100000    0.000000    0.624000    6.623500   \n",
       "max     88.976200  100.000000   27.740000    1.000000    0.871000    8.780000   \n",
       "\n",
       "              AGE         DIS         RAD         TAX     PTRATIO           B  \\\n",
       "count  506.000000  506.000000  506.000000  506.000000  506.000000  506.000000   \n",
       "mean    68.574901    3.795043    9.549407  408.237154   18.455534  356.674032   \n",
       "std     28.148861    2.105710    8.707259  168.537116    2.164946   91.294864   \n",
       "min      2.900000    1.129600    1.000000  187.000000   12.600000    0.320000   \n",
       "25%     45.025000    2.100175    4.000000  279.000000   17.400000  375.377500   \n",
       "50%     77.500000    3.207450    5.000000  330.000000   19.050000  391.440000   \n",
       "75%     94.075000    5.188425   24.000000  666.000000   20.200000  396.225000   \n",
       "max    100.000000   12.126500   24.000000  711.000000   22.000000  396.900000   \n",
       "\n",
       "            LSTAT        MEDV  \n",
       "count  506.000000  506.000000  \n",
       "mean    12.653063   22.532806  \n",
       "std      7.141062    9.197104  \n",
       "min      1.730000    5.000000  \n",
       "25%      6.950000   17.025000  \n",
       "50%     11.360000   21.200000  \n",
       "75%     16.955000   25.000000  \n",
       "max     37.970000   50.000000  "
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>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</td>\n",
       "      <td>0.538</td>\n",
       "      <td>6.575</td>\n",
       "      <td>65.2</td>\n",
       "      <td>4.0900</td>\n",
       "      <td>1</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</td>\n",
       "      <td>0.469</td>\n",
       "      <td>6.421</td>\n",
       "      <td>78.9</td>\n",
       "      <td>4.9671</td>\n",
       "      <td>2</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</td>\n",
       "      <td>0.469</td>\n",
       "      <td>7.185</td>\n",
       "      <td>61.1</td>\n",
       "      <td>4.9671</td>\n",
       "      <td>2</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</td>\n",
       "      <td>0.458</td>\n",
       "      <td>6.998</td>\n",
       "      <td>45.8</td>\n",
       "      <td>6.0622</td>\n",
       "      <td>3</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</td>\n",
       "      <td>0.458</td>\n",
       "      <td>7.147</td>\n",
       "      <td>54.2</td>\n",
       "      <td>6.0622</td>\n",
       "      <td>3</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  0.538  6.575  65.2  4.0900    1  296.0   \n",
       "1  0.02731   0.0   7.07     0  0.469  6.421  78.9  4.9671    2  242.0   \n",
       "2  0.02729   0.0   7.07     0  0.469  7.185  61.1  4.9671    2  242.0   \n",
       "3  0.03237   0.0   2.18     0  0.458  6.998  45.8  6.0622    3  222.0   \n",
       "4  0.06905   0.0   2.18     0  0.458  7.147  54.2  6.0622    3  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  "
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### We now have all numeric features. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 506 entries, 0 to 505\n",
      "Data columns (total 14 columns):\n",
      "CRIM       506 non-null float64\n",
      "ZN         506 non-null float64\n",
      "INDUS      506 non-null float64\n",
      "CHAS       506 non-null int64\n",
      "NOX        506 non-null float64\n",
      "RM         506 non-null float64\n",
      "AGE        506 non-null float64\n",
      "DIS        506 non-null float64\n",
      "RAD        506 non-null int64\n",
      "TAX        506 non-null float64\n",
      "PTRATIO    506 non-null float64\n",
      "B          506 non-null float64\n",
      "LSTAT      506 non-null float64\n",
      "MEDV       506 non-null float64\n",
      "dtypes: float64(12), int64(2)\n",
      "memory usage: 55.4 KB\n"
     ]
    }
   ],
   "source": [
    "data.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### This histogram shows the distribution of each of the features."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x1080 with 16 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "import matplotlib.pyplot as plt\n",
    "data.hist(bins=50, figsize=(20,15))\n",
    "plt.show();"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "housing = data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "corr_matrix = housing.corr()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### First, we will build a model to make predictions of the \"MEDV\" label, or median home value. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### By calculating the correlation between the features and MEDV, we can see that RM, ZN, and B are the most positively correlated features. The CHAS of the housing set is less correlated, and the LSTAT is very negatively correlated to the MEDV, so it is the most important variable feature. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "MEDV       1.000000\n",
       "RM         0.695360\n",
       "ZN         0.360445\n",
       "B          0.333461\n",
       "DIS        0.249929\n",
       "CHAS       0.175260\n",
       "AGE       -0.376955\n",
       "RAD       -0.381626\n",
       "CRIM      -0.388305\n",
       "NOX       -0.427321\n",
       "TAX       -0.468536\n",
       "INDUS     -0.483725\n",
       "PTRATIO   -0.507787\n",
       "LSTAT     -0.737663\n",
       "Name: MEDV, dtype: float64"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "corr_matrix[\"MEDV\"].sort_values(ascending=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAtMAAAHoCAYAAACLoWPCAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvnQurowAAIABJREFUeJzs3XeUZNd92PnvfZVzVXd1zjM9OSeEwSAMAJKgSBAiRUkURYqyJEuWd9da6ayPtWfPri37nD2yfc7aK8urs5J8JEtakhKTSEogQQIEQGRgcu4JnVPlnOu9u39UTWMG0z3TOczczz9d3V1V71bVq6rfu+93fz8hpURRFEVRFEVRlIXT1noAiqIoiqIoirJRqWBaURRFURRFURZJBdOKoiiKoiiKskgqmFYURVEURVGURVLBtKIoiqIoiqIskgqmFUVRFEVRFGWRVDCtKIqiKIqiKIukgmlFURRFURRFWSQVTCuKoiiKoijKIplXc2NCiF7gPeAyUJZSflwI8S+BF4AR4FellJW5bh8MBmVvb+8qjFRRFm54eBi1fyrrkdo3lfVK7ZvKenby5MmolLLpXtdb1WC67sdSyi8BCCGagONSymNCiH8F/Czwjblu2Nvby4kTJ1ZpmIqyMIcPH95w++eZsSRvXY/SF3Txyd2tCCFWbFu6IfnumQmmUkWe3dHCtlbPnNe9OJnitYEInQEHz+9tR9PmN65IpsR3z0xg1gSfPdiJz2GZ9XqFss63To2TK1X59L52OvyOeT+OD4bjvDcYY1url4/tbJnz/v/m3RHeHYxiNZvw2E0MRfI0eW38/ie38+5gnGi2xCd2tbK5yX3bbau6wXfPTDKdLtLd4GQ0nqe30cXP7Jn79bk4keJPXr+BlPBbT25ib6f/tv/f3Dd//5tn+PqJCQD+yaPd/OsX9sz7cS+HkyMJ/vgn1/DYLfyb53fR4Lau6vYfdG9fj3JqNMGuDh/HtzUv+PavDoS5OJHiYHeAo/3BO/7/yuUQl6fSHOlt4OFNjfO6z5v7ppSSF89P8b2zkzR5bPzaY31sanLzxz+5xt+8O0K2WKXZY6XBbcfrMPM/Hu9nOJbnBxemafXZ6Gt0UagYPL29mR1t3jm39+a1KGfGEuzt9PPE1g9jpLF4nn88P4XXbmFXu5c3r0fp8Dv4zL75f/4o9x8hxMh8rrcWaR7HhRBvCCF+F3gIeK3+95eBR9ZgPIrywDo3nqRcNRiYzpAv6yu6rVi2xEgsT7lqcH4iddfrXphIUa4aDEZyJAtznqy6w7VwhkyxSiJfYSiam/N6Y4k8kUyJfFnnylR63vcPcHYsSUWXM2OczXgiz7VwhlShyo1wlouTGcKZErFsmR9fDDGRKFCqGFyY5XmIZsuMxmvP02sDYcpVg6uhDNlSdc4xvT8cJ5IpEc2WeG8wPuf1vntmauby352cWMCjXh6vD4TJl3VC6SInRuYep7IyzozX9t1zYymklAu6rZRyZt8/M5684/+6ITk3nqKiS87O8v97KVYMTgwnSOYrjMXzXJhMoxuSn1wJky1WyVd0wpkyw9EcuZLOjy+FOD2aJF2oMBjOcXq09ll2fvzuny1nxhK1MY7dPsZLU2kK9X3zjWsRylWDoWiORL684MeiPHhWO5ieArYCx4FngcPAzW+yFBD46A2EEL8phDghhDgRiURWbaCK8iDY3eHDrAm2tLhxWk0ruq0Gl5WuBicWk2B3x9wzRwC72mvj6g068c8xuzyb/mY3LpsJn8NCX6Nrzut1BhwE3VbsFtNdZ8hns6f+nO1o82I1z/4R2hlwsrnJhcdupjfoYkerh0a3lQaXlWd2NNPut2M1a+xq991x26DbSmfAgcUk6jNnkiaPDdddXp/DPYGZ+z/ce8fH6IxP7flwJv1n97fN/0Evkye2NmHSBAGnhYM9/nvfQFlWezv8mDTBnk7vgs9CCSHY2+nDpAn2dtz52pk0we4O38zPhbJbNPZ3+fHaLXT4Hexs82LSBE9tbcZlM+OwmAi6rXQ3OnFYTDyzo5l9XX7cttp7bG+XH4tJsKv+2RLJlCjMMkGwp/Pmc3D7GHe0erFZNJo8No72N2IxCXoanfid6uyJcm9ioUeny7ZhIX6bWiDdIaX8D0KIg8CXpJS/N9dtDh8+LDfaaXTlwbER0zyUlZUrVTk3nqLdb6fnLsH9rSq6gVkTCCGo6gZ//e4IyXyFfV0+nt4+e1rJvdy6b+pG7TPftAanri9Npnnp4jRWs8YvPdRNg0sFKveTULrIjUiWHa1eAvN8bRf6uRnLlrg8nWZT0EW73znrdU4Mx3njWhSH1cSXH+nBZVuLjFblfiCEOCmlPHyv6632AkSPlDJT//Ux4L8AXwT+A7WZ6ndXczxrrff3/3FJtx/+w08t00gURVkJL18OMRjJoQnBrx3rxWO/+yz7+fEUr1wJEXTb+MUjXRQrOsl8Lc1lOlVa8njC6SLfPDWOQPD5Q500eWxLvs+FmE4XAChXDeK5kgqm7yOGIfnWqXFKFYPr4Sy/8mjvimznmyfHeet6FAn86+d30t9855mlqVQRqK1dSBUqKphWVtxq72GPCyH+HVAC3pRSvieE+KkQ4k1gFPjPqzweRVGUFWPWamkgmgBtHqfVr4YySFk7RZ3IlWn22nlia5DhaJ6HNzUseTxD0RylSi3PezSeW/Vg+nBvA5liFbfNTF/Qfe8bKBuGEGDWBCVW9qxHrlSlVDWwmjQGI7lZg+mjmxupGgZBt402n33FxqIoN61qMC2lfBF48SN/+/fAv1/NcSiKoqyGZ3c20xFw0Oq1z2t27GBPgGShQpvPTtBdC3QP9TRwqGfpgTTA9jYv1yNZBIKtLQvLFV8OXruFF/Z3rPp2lZUnhODnD3UxHMvR37xyB0pfOdqLEGC3mNjXNXvefaPbxmcPdK7YGBTlo9S5D0VRlGVQrOgMRnJ0BBwzJflsZhP75/jCn01f0MWvH+tbqSHic1h4bHMQIbhnyomiLFTAZZ13rvRiNbptfO5gJ4aEFq+adVbWBxVMK4qiLIN/ODfFWDyPy2bi149tWpMFfvdyZTrND85PA/D8vrZZT5Erynp2PZzh+2drJR6f291615rSirJaVDCtKIqyRO8PxXnrehS/w4JZExhSYmL9BdO3lgorlGevkb2SUoUKr14J47KZeXp787o84FDWl2JF55XLYSSSZ7a33LbfFiorWxtfUeZLBdOKoihLkClWeOt6lKDbSlk3+Mz+diwmjYpucC2UpcVro9G9ugv95rK300+5aiCEYFf76s/onRiOc3IkjsWk0dvoZMsa5G0rG8ulqTRXQ7UiYC1eO70NTjoCDtr9dvZ1qlrlyvqggmlFUZQlcFhMBJy1/OOj3f6ZetI/vhRiYDqD1azxa4/14VhAUxwp5Yq0djdpYt5tnldCIl/maiiLSRMz9a4V5W5avHZMmkBK8Dss/N3JccpVA7fNrM5sKOuGCqYVRVGWwGzS+OLDPaSLFRpvWXxVrJ+CruqSqmEA9w6mdUPy96cnGE8UeHp78x1d2pYqmi3x7Xqd6Z871LnqdZ5bvHb2d9U60NktK9txU7k/dPgd/JPHepGAoPZ+yper/ODCFBOJwprsx4ryUSqYVhRFmSfDkCQLFfwOC9ots2JWszZTyu6mZ3e2cHo0SYffMe/KGalChdF4HoCLk6llD6ZvhLPkSrUgfyiapcG1PCX35uuhvgZMQuCqt4BWNhbdkKRm2f9XmtNqntnup/a28ZMrYUyaIFuqrsl+rCgfpYJpRVGUefr+uUluhLP0Bp187mDXXa9brNS6r3ns8/+Y9Tss9De7GU8U5qyhuxQ9QSffOT2BJgTdDasfzNrMJh7d3LgiKSzKyvv70xOMxvP0N7t5fl/7qm3326fGGYvn8TktBN02jvQGODOWRBOC/iaVd6+sPRVMK4qizNPAdIZz4ynOjCV5uC9Iu98+Z2D42kCEiUSBG+Esm4NufM57z05rmljRIGU8XpipzTueyK96B8SxeJ5vnxrHbTPzhYe6VZvnDURKyXii1g7+5s/VYBiSUyNJhqJZEvkyn9jVymSyyG8/tXlR91es6FhN2qrOrCv3P22tB6AoirJR9Aad2MwaUsJ/fvkq/3Buas7rNtXTPtw2M3br/D9qR2P5elvx5V+g1+i2IUSt9XOja/UrjLx+NcI7gzFeHQgzEMowHM1xrV6pQVnfhBAc395Em8/OU9uaVm27miZocFlw2cy4bGbGE3n8jsUdhJ0cSfAnr93g/3t/lIq++qUhlfuXmhZQFEWZp4/vbKVQNvjp1QgNLiuDkRzZYoV0sUqb7/ZZ6qe2NbG11UPAacFmnt9iu7F4nm+dGgfgia1NHOoJLOv4+4IuvvxID0KINVm0ZTNrWEwaVpNGOF3i9YEIAM/s0NmrypytuopuEEoXafbYsZrvfcC3t9O/Jq/Tx3e3UtYNJpMaXocF2yIXrw5GsgBEMyWS+cqqn5lR7l8qmFYURZknv9PKV472sq/Lz5nRBFuaPfzNe6MUyjqHewM8vuXDGTshBB1+x4Luv3zLbFmpujINKday5vXx7c2UdQOX1UxnwMHlqTQApaqaJVwL3zk9wUSiQJvPzhce6l7r4czp/HgKi0lDN8DnsFJdZFnFh/oayJcjtPsdBN2qAoiyfFQwrSiKskD7u/zs7/ITzZZ4fzgO1MrOLdXmJjdPb2+mUNGRUvIXbw2xr8vPwe7lmaFOFyu8eG4KIeBTe9txr3LOctBt45cf7gFqObgV3aCiSw6swGJL5d5i2TKwPPvucrsayvDW9Sg9DU7iuXKt0U/QyeHewKJnx3saXXzlqKoioyw/FUwrivLACKWLfP/sJHaLic8d7MBpXdpHYNBt4/EtQaZSRR7dvDzNUG5W8fijV66hG5J3bsSWLZj+YCjODy7U8rzbvA6eWMXc148SQnBgmR6Xsjif2NXC+YnUmnTDvJf3BmMk8xWS+RTHtzcxEsuzr9N/W0nFWLbEd89MYtIEP3ugA59jfiUoFWW5qWBaUZQHxqWpNJlilUyxynA0z85lCCIO965MjdvNTW6uhjL0N7uX7T7LVYOyLhFA2ViZNBJl49jU5GZT0/LtX8tpc7ObaDZOu9/O3g4/+7vuPPC6GsqSKlSAWj60OjhT1sqaBNNCiN8DPielPCaE+E/AYeCUlPJ31mI8iqI8GLa2eLg8lcZhMdHd6Fzr4dzVz+xp5ZkdzcvaKfBIXwPDsRwCwcFu1ehCWb+Obg5ysDuAzazNWX5yc7OLc+NJTJqgTzUBUtbQqgfTQggbsK9++SDgklI+LoT4EyHEESnlB6s9JkVRHgwdfge//eTmFW0aIqXkwkQaiWRPh2/R2xJi+VtuN7qsPNYfRCAIzKPutaKspXvt/80eO7/15MLqTUspuTiZRjdq709Vb1pZDmsxM/0bwH8H/i3wKPBy/e8vA48AKphWFGXFrHT3vYuTaV6+HKptC7HsLcGX4sJEmlcuhwHQNNjVvn7GpiirYSCU4ceXau9PSW0xsaIs1ao2bRFCWIAnpZQ/qf/JD6Trl1PAHQlPQojfFEKcEEKciEQiqzRSRVGUxdFuCda1ddYW69bjCJOakVMeQLe9P9VbQFkmqz0z/WXgq7f8ngRurgDy1n+/jZTyT4E/BTh8+PDytwRTFEVZRjvbvWgaSAnbWz2zXkc3JN8/O8lEssDT25vZ0bY61RTafHYKFR0BtNbbiivKRrOU98/WFg9yT+0+drTN/v5UlIVa7XmTbcBvCyF+COwCgsAz9f89C7y7yuNRFEUBal+umWJlWe5re6uXHW3eOVNKEvkyQ9Ec5arB+YnUsmxzPq6Hs1g0gVkT3Kh3g1OUjea29894imJFp1iZf3Waba0edrbP/f5UlIVa1WBaSvmvpJSfkFI+B1yUUv4BUBRCvAEYUsr3V3M8iqIoAIYh+bsTY/z5G0O8fnXl08mMegc3KeWq1vh128y8NxTn/aE47iXW2FaU1ZIpVjg9miCRqzWZCTit9DQ6sZgETV4rf/bTQf7bm0OE0sU1HqnyoFqzT1Mp5bH6T1UOT1GUVaPrBufqs8H7Ov1omiBfrvLOjRilqo7TauLJrSvXzEQ3JN86NQFAR8CxqosAz0+kKNRn8C5Npdm2SuklG0lFNzBr4r6ctbweznB2LMWONu+y1FhfLX9/ZpJopkSmOMnOdi8HuwN87mAnUGvuUtENpuIFXr0S5hePdN2Xr52yvqmpCUVRHhjZUpX/++VrXJhIsaXFjRCC/V1+wpkSZpMgUzRwWJe3HN1HSSkxZG1mumqs7jIQQ0oK5VowXVnlbW8E58dTvHIlRNBt4xePdGExrbMVpEv08uUwhbLORLLAjjbPhgk6dd3AkJKBUBaP3UwyX5lpNrOrw8drA2HOT6QZjuXxO608t7t1jUesPGjur08KRVE2tKpuEEoXqerGitz/VLJAoVLFkJJEroy5vpy/0WWj0+9AE4K3r8f463eGyZerKzIGs0njcwc7eHhTA5/e074i25jLnk4f21s97GzzsLdj9cvihdNF/vKtIb5xYmxBOa6r5dRogqvTGc6NJ2dSCu4nNxedtnrtGyaQvjKdxmExs7XFw+P9jYCg1ffh4lm3zcyB7gA+hwUp4cr0/NcgvDYQ5s/fGOTc+O21D4oVnVC6iJTqgFOZHzUzrSjKuvHt0xNMJAp0Nzj5uUOdy37/3Y1ODnQF8NktHO1vnMlX1qUkmisxEs+hIRiL55lMFvnSoz28PxTD77Ty1NamZQtA2nwO2nyOZbmvhcgUqkwkCwCkl2mx5UK8OxTn3cEYFpPG7g7fslQx+WA4zjs3YmxpdvPJPW1Luq9S1SBbqsLGiDMX7Pl97cSyJRpc1rUeyryUqjovXQhhSEmDy8qvP76JRK5M0G277XrHtzVxaSrNu4MxUvkqHwzHOdL7YYfPqm7w7dMThNNFPrazlb6gi5cuTvHi+Wn6gi4+GE6wt7NWb7qiG3zt/VGS+Qp7O308s6NlVR+zsjGpYFpRlHUjXF9ANL1CC4lsZhPP72vnRiR72+xcPFfCbjbhsZmJZEr4nVai2SIvnZ8mVawAeTYH3eu+Bfm9XI9kZ7rK3YjkeKivcVW3ny2WmUoVsZk1ytXlmZk+P55CNyRXpjMc37609uvbWj3kSlXsFhNu+/339WjSBM1LKImYLVUZjeXpDTpxrsICVoum4bGbSRUqBFxWLCbtjvEPRXOMJ/I83h/k5kTyRxcixnJlJhK1g8iLkylKVZ3r4RyGlITSRR7rD85ct1jRSeYr9fspreCjU+4n99+nhaIoG9bHdrZycTLFnmVKQciXq7xxLYrDYuJYfxBNE/zo0jRnx5LYzCZ+/nAnXQEnfUE3x7YE6W50YdPgxQvTxLIVynot4LNbTPhdd2+/PRTN8crlEM1eO5/a07Yum6J8Zl87Y/E8mhB8eu/SZnEXI5WvzYxbTRrLdQJ9X5ePd27E6G/2LLn9+hNbgnjtZjr9qxMsLjfdkEQytZlnq3n5szi/cWKMZL5Cs9fGLz/cc8f/y1WDeK5Mk8e26P2/ohvEsmWCbitmk8YvPdRNJFOi3X/nQcClyTT/709vcGkyTYffzrM7WtGlwZHeD/u/3YhkeeVSiHiuRIvXzu4OH167BbMm2Nnm42f2tLKl5cN60x67hSe2NjESy/HwptU92FQ2ro33aaEoyn1rW6uHbXM0OlmME8MJLk3Wmqy2+uxsbfFwPZzl7FiSSKZEtlTh0c1BPrGrlaObg7xy+RLjiTxOqwm/00bVgC893I3bbrnnwsQzYwkyxSqZYpZIpnRbXud6YbeY2NflRyDWZHHd1VAGiyaQ1Nqu7++6o+ntgh3qaeBQT8O9rzgPb9+I8f5QHJfNxK882rvk4Hy1/cO5SQYjuTmD3Wi2xMB0hv5mNy0LnKGWUs5Ugrm5iPWjvnFyjHC6xKYmFy/s71j4AwC+c2qCieSHqV4Oq+mOM0KXp9JcmEgxnihwfjxFJFvCrMHJ0TjdDS7evBabSRM7ORznWjiL2ST4hSNdNHtqj/tXH+vFMMDnvPMg+VBPgEM9S983lQeHWoCoKMp962ZuqEkT+B21L83+JjfNHjseu4WqLglnaqdyxxN5roezFCsG6WKVUlUnX64ylS7Oq8LH1hYPmhA0eWzrNif1/ESKa6EsV0MZLk2lV337z+xsJuCy0uq133Zqfb0IZ2rpAbmSTqa4MgtQV1Kkvi9HM2X0Waq1fPfMJO8PxfnO6YkFL64TQvDC/g72d/t5ft+dC2d1QxLNlG8bx2LcfA1CmblTvX5yJcx4osBQNMu2Vg8NLgs9jU5cNvMdt63oktF4nlC6RPKWRaUeu2XWQFpRFkPNTCuKct/a3eEj6LZhM2sE6gHuk9uaQMB0qkjAZeHhet5wV4OTvqCLy9NpdrV5CbptCCEwa/Obc9jV7mN7q3ddpnfc1Oq1owmBECx4ZnI5HN/WQipfy3/tDqy//PNj/U1oIkqL106Tx3bvG6wzz+xo4cxYgm0ts++HVlPtb4uto93hd9Dhn33hrEkTfGxnCwOhpZ1x+NjOVi5N3T3Vq81nZySW56HeRrxOC5/e28bB7gDpYvWO225r9fBQugFNCBzzTN0pVnTeuBbFrAmObQnedyUSleWngmlFUe5rzR4b5VtK7XnsFra2eEgXKmxu8tBZD+psZhNPbmvCYzfjsJhw2UxYNI22errGe4MxTgzHOb69mZ1zNFpZz4E0QNBtw2HV0ISgcQ1mz98fjhPOlAlnymxtydHf7F7yfQ5GspwYSbAp6GJPpw+bee6zCFJKroYy9bMHdwbLTR7botMT1oO+oIu+oGvO/x/b0sQbVyM8skK5wDvbl94M5qOpXlOpAm9ei9Lhd3C0fjbj2R0tvDYQYXOTi10dPnRDUjUM2vyOmdvmSlXOjSeYSBSwmDQe62+kq2F+B3DnxlNcqDd2anRbZyp9KMpcVDCtKMp9q6ob/N2JcULpIse2BDnS20AyX+bf//AK+ZLOtVCWoNvKDy9ME8mW2NTowmk1MxLNMhovYLOY0CX8yqM9/PGr1ylXDa6Hs/ynLxxY64e2KD+8OMUrl8NALYf803tXts51LFvipYshXDYTz+1uJVgPYM2awL9Mp9hfvxohkinxD2cn2dfl54mtQfZ2+mcNqv/qnRFePD+F22bmP3x+L43ujTf7vBQ/uRImXajwypXwTNOi5TQczXEtnGV3h3fZSj++eS3KeKLAeKLAtlYPjW4bL56f4vvnJskWqzy3q4WyLgHBp/a20upzgIT/+uo13hmMkylWOLo5yHiiwKFb0sirusFLF0Mk8mU+trPltjM1DS4rQoBAEHCuz5QtZX1RwbSiKBvayZEEk8kCj2xqvO3U/Ggsz2A0y2Qyj0nTuB7OcqS3gYlkAbtZI1usIoFz40lODseJ5ysUyzpPbmtiYDpDNFeriiCRVA2J3WKiXDU23KK0W5mEIJEvI2CmYc1KOjeemilTNhTNsafTR7PXht1sWrZ81Xa/g5FYDrOmEUoX+Yu3htnf5edLj/TcUZHjWihDplihWNGZTBUeuGD6ZoUPi2nx7dKllLPe1jAk3z87SdWQjMXz/NqxviWN9aYOv4PxRAGvw0K6WGE4liOeK1Oq1Kp+vDcYx1SvW/7OjRgTyQKj8TzpYnVmHy9V9TvSUyaSBQam0wghODWSuK1GeX+zm19+uAeTJtbt+gdlfVlwMC2EOCylPLESg1EURVmIeK7MT69GgFpZrpsr+BO5Mt8+Nc54Io8BNLlsmE2CYkVnc5Ob49uaieXKvLCvnW+cGmc4lsdm0fA5LJg0QV/QiZSSzU1uvvxILw6rid95ZguXp9K1nOsNqtlrZ2uzG4S4o/HFSuhpdHJ+IoXNrNHmrQUzy52r/fGdLezv9PH2YJx3bkRp89nJlXTiufIdwXSH386b13UCThPts8yc5kpVTo8mafXZ6G9evqoy68VnD3QwFMktul76jy+FuDiZ4kB3gCe33v4+EAJctlpN6OWs0X24t4FcuVb7+/tnp9ANSYffwaObGzk/kaK/yU3VkAQ9NgIOC2/diJHIlWlwWdnW4mZnm5cmby3Vq1TVsZlN6IbkzetRzown6fA7Zm0/vhFz5pW1s5g9/s+EEG7ga8DXpZSXlnlMiqIo8+K0mihUqpQqBgc/UspqOl1kNJ6nohskcmWsFo23b0R5ensLnz/cBUAyX6aqS45vb6ZcNegNumhwWtne5mV7m5eP76x9yZ4dS3ItlKFY0anoG7fFsN2scWk6gwAcqzDDvqnJzW8+sQmTtnKl+IQQtPgcfPZAB4/1N/LaQISg2zrrQrlkoUqL144QkMxX7piZ/smVMNfDWYSAXz1qw3+fneJ328zs6VxcDXcpJRcnU0hZa3xyZzAt+MUjXUylCnPmJt+IZLkRzrKvyz/vg6oPhuNcmEhT1g2KZR2zSXBmPMlTW4N8+ZEewpkiVV3icVh4bSBMRTcIOK3savfx+UMdvHY1wovnpmnx2ilVDY5vayZVqBBOl9jXWRvHcnTiVB5sCw6mpZQHhBDbgC8A3xRClPkwsB5Z7gEqiqLM5Vooi1nTqGqSjluaOgRcVj62o4X/niwQz5XIl3UKFR2PzczT2z9sD+xz1BYjjiXyZIoVRmJ5xuIFfuvJTdgtJgxD8qdv3GAyUWAwmudgt5+TI4k5Kxqsd1/7YJRQqpZ28bcfjPFvl6k5zt18NC0mnC5iW8Y0j1s1e+z8Qv1AaTYHuv0k8mXcNvOsM/O2ehqESYhFLSYNpYu8Oxiju8HJge47K1rkSlUuTaXpDKxNO/mlEEJwsDvAxck0B7pnX5DnspnnnNGv6Ab/eK42szyVKvKVo73z2u7N18Rq0ji2M8jX3h9lOlUkmSujCY03r0eJZIrkSjodfju723081BdgX1eAV6+EuR7KMhTN4baZZ+7L77DQ3+xmIlngcO/CK4/kSlVShQptPvuy550rG9OizsVIKQeAPwD+QAixj1pg/RMhxLSU8rHlHKCiKMpcyrqOxSSYTJb5x3NT/MIR80yQ8mh/I69cCZEsVMgWqwRdVrKlWge+dp+d02NJKlWDT+xqwWzS+N7ZSW6Eszitpplcyx9dmuY63vt+AAAgAElEQVS9wTgV3cBsEiBg012qJax3AYcFrf7YbtbdXk2nRhL8zbsjOCwmfvfjW1cl1eRWz+5ooavBSaPLNmswf3x7M41uK+0+Bx77wp+f1wbCTCaLDEZybG524/3IffzwwjSj8TwWk+A3Ht+07Pn3V0MZTo0k2N7mZX/X8legeGJrE09sXVyak0kIXDYz6UIFzwLSQA71BPDYLfVGShZev+ogmi2RK1d56eI0V6bTlKsGVUNSqFTxOCzs6fTjtpnxOiy4bGb2dfn4+M5WDtYPcDRNzForez7y5Sp//e4IhbLOkd4Gjm1ZWL30YkXn1EgCr8PC7lU4mFVWx5ISm4QQGtAMtAAuIHKP6+8G/hTQgevArwH/F3AYOCWl/J2ljEdRlAfL/q4AsWyZqVSRfLnKq1fCfPHhHt68FuH1qxHSxSqHegKE0yV6gk7sVhNeu5mroSyvD9Q+riq6QUWXBJwWPr2njbaAA3M9JSGSLbOl2U0kW+a3n9xMd6NzQy9A/F8+vh1NgKZp/Itntqz69t++EWUklkMIwcWJFE9ua17V7ZtNGrvmKGsIcGEixU+vRml0W/nCke4Ft+QOum1MJot47GbsdynRt1JeH4iQLVWZThfZ0+G7Y3Z9PJHnwkSa7a0eehdxUKgbkkS+TIPTOnNQNl+aJvjCkS6mUkW651miDmoz4reWyvvCkS72dvoYi+UJpQsk8hV0o5aidbArwGcPdswcxDzeH6Qr4CDgtGK3mMgUq/icFooVnUJZn6k9H82WODmSoKfRyfbWu6d85Er6TAfIaHbhzWneGYxxZjQJ1M6Mzbdcn7K+LSqYFkI8DvwS8LPABeDrwO9KKVP3uOmAlPJo/T7+AngIcEkpHxdC/IkQ4oiU8oPFjElRlPuDYch5f1GbNEGL144h4cxYimShyuNbcnz9gzEimRKN9eYgiVyZq6EM/8Pxftw2M8PRLKF0kWaPjcFojli21hntU3vbcNs+/Fh8ZnszJ0YS9DW62LqMbc7XylAsy/mJWs70WCJPX9PS6zwvhN9Zq8hgMZlo8tiQUiIlCw7MVsrAdIbJZIFUoUIyX6Z5gYsln97ezI42LwGnddZA/LndrVyeStMZWJmDss6AgyvTGdp9jlnTVF48P0WupHMjkuWfP7V5wSkKtUW9BTY3u/nMImZ2a2kgS9vnehtdbGpyc3o0wZ+8doNSRac94GBXm48XDnTclmaiaYJNTW7i2RL/5nsXSBer/NJDXVyeypAv6zy1rYkD3QFevhRiKlXk8lSarsCHnRRn0+SxcWxLkOlUkaDbypmxJHs7fHfsw7ohZ30NbqaaCPHhZWXjW0w1jzFglFoA/QdSytB8byulrNzyawl4Fni5/vvLwCOACqYV5QF1ZTrNSxdCNLqt/MLhrnvODF6YSPKXbw+RKlRo9dpp9doZiRdoqKd0NHvtBFwWpup5wq8NRHh9IMLbN2I0uq0c2xKkp8HFy5dDCMEdp5/b/Q4+s0Hzo2fzV++McHI0Dgi++t4o/9und67q9tt8dh7qa8Rq0ihUdP78jSEqhsFnD3Ssixzism4QyZYoVQ2si1gwKYSg/S77i8tm5nBvw1KGeFfP7W7l4U2N+OZI4fE5LORKOh67ecGBtJRy5n00kSgseawLpRuSb50cZzJV4Pi2Zg50B/DazTS4rKTyFY5tCaIbkjeuRTjUE7itksul6QzDsTwAL10MzaQXTSaLHOgGr6P2GeG0mua1UPZIbwPXw1m+f3ayPjaDQz0fvq7vD8V5+0aU3kYXL+xvv+25fqSvkQaXFY/dsuCDNWX9WszM9LGlLDQUQnwG+D+Bq8AUkK7/KwXsmuX6vwn8JkB3d/diN6soygYwMJ3BkJJIpkQkW5pzod9UqsCPLoY4P55kIlEgU6xScBr0Nbk43BNgJJoD4OcPdxJ0W3l1IIKU0BVwcKp+ijVf1gmnizS4rPzcwQ7sVhPNnvv7y61QvlmNRM6cql5Nh3saSReruKxmNCHIlqoADEZyqxJMj8XzvHI5RLPXznO7Wu+YTWxy29jX6cekCcQKzJYPTGd463qUvqCL49sXnuIylSpwdixJf7N71oV+Qty9LvIL+zuYSBYWtYBWCMHT25u5NJlm3wrkY99LMl9mIlkL4q9Mp9nS4gYBupS0eO1kSxVOX6i9t3Ml/bZyd7vaPLT77YzG8uhGLaXLbbdgs2j8+FKIo5sb2d7qocljm3dqz+3HIrfvK5en0khZq62eL+u3zXRrmrhnKomy8SymmseIEOIrwO8A2+p/vgz8kZTyr+Zx++8B3xNC/BegCtzcq7xAcpbr/ym1PGsOHz68cWtSKYpyT/s6/UQyJYJuGy2z1Hmt6AZ/f3qCt65H8TosVAxJm89OWS/gsZt4+eI0E4kCFd2gu8HJ9XCWXe0d/O6zW7k0leLopkacVjMOi4bdYiKULvLTq1G2tXr4jcc3rcEjXl0eh4lqvbSf17H6Pbt8TgufPVCrBZ6tV7aoVA22LyGFpqobTKWKNHls90ydODmSIJGvkMhX2N/lv2MW+fj2Zlp8dpo9tjlnd5fi/eE4qUKFM2NJDvcGFrzI8aUL0yTyFa6Gsvz2U64FlxvMl3WimRJeu2VRaSa7O3xrtmiuXDVqXQkFHOgOcDWUpc3nYCiax+cw89OrUcbieaxmjalUgUS+zKagi6lUkf5mN//sqc1844NxXDYzuoRHNjXwjRPjABhS8oldrcRzZcLpIs1eOz+6OM21cJZHNjVyqOfOih+bm9x8am8b5arBzo+U1jvUE+Ct61E2NbnvmjKi3D8Wk+bxK8D/DPwecIraIdlB4D8KIbhbQC2EsEkpb2bspwEJPAP8HbWUj79c6HgURVk/ylUD3ZA4rIvLB+0NuuYMasPpIq9fjXBpMo3LZiaSKXGoJ8Dh3gA/uhDiq++PUNElhYqO12FFUAvOM8UK//GlAaq6QSRT4itH+/jknjbOjSf56nujFCs6VUNyLZS57xs1nBtLc3NG4vRYYk3H4raZ+eWHe277m5SSwWgOj80871Pgf/HWMO8MxuhpcPK/f3rnXfOv+5vdDMdyNLisNLrvnMG1W0wzFR9WQn+Tm2imdsbFZV14kOVzWkjkK7htZkyLKMn296fHCWdKnB5N8M+e6l/w7deKbki+c2YCKWuLPLe2eIhlSzitZtp8dgJOC/Fcmc6Ag5FYAa/dzGgsz6XJFGatVqlnT4eXRreNZL5Mvlzl7RsxBLUgxOewMJEs8M0T4+iGQUfAwftDcdr9Ds6OJWcNpgG2ttQOAstVA71qzHzureVBh7I2FnPI9M+Bz0oph2/520+EED9HLY/6brPTzwkhfq9++Rq19I3/JIR4AzgrpXx/EeN5YPX+/j8u6fbDf/ipZRqJcivDkERzJQJO64o1qliPUvkKX/tglFLF4NP72ti8zIvb/uHcFPFc7VTvng4vX3qkm20tXjRNcD2cxWExUapWKFYMgm6BBMqGweWpNNFsiWK5ykg9bxJgb6cfi6bxo8sh7GZtyQujNoJ2v42To7XLXYH1V0Xgg+EEb12PognBFx/untfBzcnRBLlSlctTabKlCl7H3GkOuzt8bGlxY9G0NVn0+OjmRg50+7GZtUXVJ/703nbG4nlaffZFjX9gOstwLEfrBszVFfVUCq3+vDW6bfzTx/v41qlxLk6mafU6uDiVwmoWlHVJZ8BGQLcwGMnhspoAwc42D4l8maFonolEgaObG2nzOehqcHBxMo0hJdFsidFYnmypiqbBng7fnIsJAdLFCl97b5RixeBn9rSypWXtFirfPMBY7GSGsniLCaa9HwmkAZBSDgsh7poIJKX8LvDdj/xZlcNT7iu12qe1Wc5ffrj7gSnqP50uzuThjsbyiw6m8+UqsWyZDr8DQ9a+xIQQOKwmTAXBQ30Nt81eTyYL7O3w0Rt0oRuS/V0+Lk5mmEoVOTuW5Gf3d1Ao60ynS0QzJULp4kz3tR3tXna0Pzj5i4YUM9mdy9XIcTpVxGrW7pqrO1+5eg61Ieef0318WzOvDYTZ2uKZV9qEbQ1K1t1qKVU8LCaNTUs4SG3x2ZFIGlxWpJQb5rPJpAl+/nAnI7E8W1s+fPyaEEyligScVjLFCg/3NWJIybEtQR7ua6RY0YlkSlwLZUjkK5weSxJKFylWdLY0e9jS4pnZb7e3eohmS9wIZ3n5cghDQsBt5eJkmmLVmLN6SShVJF/fV0di+VUPpiOZEhLJeKLA6wMRHFYTX3qk57aqRMrKW8yzfbdlvKu/xHcNLXVmWLk/3VzxHs2WqOgSq3ljfGEtVV/QxeZmN/lSddELlCq6wVffGyVTrOKymciXdZo8Nn7hcBc/u7+DoWiOroYP81zPjCV59UoYsyb4pYe6CddTP3LlcTLFCgGnFQn0NTmx1mcDT40k2NTkYmuLZ8MEE8ul9sVbE0ot/eP60mSaly5Oo9VbSbf6ljbj+ejmRoSoVVfobpzfzPkvHO7kuV2tuOdZoSJbqmIzaw/UWaObPrajhTNjCXa2+zbcvh902+5o8qNpgmP9Qc5NpHi8P0g4W+L1qxFODCdo9drp8DvoDDjwOS2Mx/MMRrKMxvP4HVb6m923HQCaTRpPbWumv9nNWKJAsVJlIlmkyW1jKjn3e6U36KK/2U2uVJ2zM+RKGYnl+M7pCYCZ2tqFsk4iV1bB9CpbzLO9Qwhxbpa/C+D+X8GjKPdwfHszJ0cSbGl2L7jpw0Z2s7Sc125eUIezW5WqxkyFh2uhLO1+B+F0iUSuVvN35y2zyCeG47w2EKFqGBgmjRuRWgWPQlnn6e3N/PDiNK8PRLgyleZYf5ALE2lsFsFfvDWEzWLit57YxLEti+vmtlE5LR/ujy7b0mdoE/lafW5DSpKF8oKD6Uyxgm5I/M5aUGO3mHhqgY1chBD4nBbeH4qTyJc5urlxzhnq06MJXhuI4HNY+OLD3Ru6Ac9i7On0safz/srl7Q26eG8oztuDMXoanbR67QxGsvzXV6+zp8OH22bm4mStUc3mZjc3IjksJsE/nJvC67DwyKbaDPab16LYLSYe3dTA8e3NfPf0BE1uK9PpIp872DHn9i0m7Z7dFM+MJZlOFXi4r3GmUcxySOQryPrRcU+jk3SxQoPLRmdg7ctMPmgWFUwv+ygU5T7SF3TRt4FbTi/WufHUbZ29DixiIZfbZubZHS2MxPIc6w9ydjxJm89xx4zUZLLAG9eiGPWmH3s6fVwYTyGE4MxYkqohuTqdqV03JelpdPPzh7v5b28OIkSt/XAoXVz6g95gMmWdm+F0urj00niHegLkyzo2s8bWWUq13U0oXeTvPhhDl5Ln97UvKcd+LJ7nretRAKTktrJotxqN13LmU4UKiXx5XdS2VpZmKlmkXDWAWgpP0GNjLJ6n0WUlmi3zzo0YsVyZwWiWP/uVI/QFXfzFW8O0eG28PxTnkU2NnBxJcH6i1nMu6LHy5NYmLk2mKVZ0nFbTrGUI5yuWLfHqlTAAhYo+U81mOexq95LIl5FScqy/6YGavFlvFhNMO6SUV+CO6hwIIR4BFl2DWlGUjSvgtMzUXr0507gY21o99DQ68dgtHOmbvcGFy2ae+eJ4fEsjj25uxGExMRLLE3BZuTqdYXOTi5F4nk1N7pnFhZ/a004kU8JmNvHJ3W2LHuNG9RvH+vhfw+cBwa8f613y/dktJj62s2VRt41kSlSN2rRaKF1cUjDtsZuxmMRMW/i5HOltIFeqpQ5txEV4yp22tLgZjGYpVQ0e3VxrWHN0UwOvDkTY0uKhVNUpVHSaPXZKVZ0D3QGi2TIXJlIz+df++j6jCTFTEvEz+9sZmE4vuSa0w2rCbjFRrOgElvC5OBuLSeP4As/kKCtjMcH0V6mVwgN455bLAP/PR35XFOUBsanJzRcfqjVWWkxnr9FYnlCmyImhOFfDGawmE+1+O09ta57JwR6J5fhgKM6h3gCfP9TJmbEkvcFabm2Dy0q6UGVvhw+TgFafjZ871MneTv/Maf9Wn53f/+SHJ9dS+Qo3olk2B9347hKE3S/6m5w4LSaEgO2tq3+6P5wu8uL5KYQQbG6qncExmwT766+vbkguTabx2M301s/u5MtVBqYzdAQcczbV8TutfOmRHjLFKl0Nc+dat/sdfPFh1fzrfmK3mHhhfy0NY2A6w3g8z6mxJIlcmRPDcQJOKx67mSaPjctTGbobnHxsZwsdfjsgkFKyq91Hg6tWfSnotnF+PMnX3x+lu9FJ9132p/lwWs388iPdJHOV29Z7KPeXxQTTYo7Ls/2uKMoDZL5B9GAky6nRBDaziWd2NJPKV/jmyTFyZZ0zowkmkwVSxSr9zS6uhjI4LCZ2tfv4/rlJxuJ5TJrgYzta8DktfOfUOL1BF7FsGafVRDRb5AsPdfPnbwwxnUownSqxvc3DTy6H6Qg4eGF/x0yZq2+dGq810RhN8mvH+lbyqVkX/qevnWUyXTuZ+M//5gTf+xdPrOr2z0+kiGbLnBlNcDHgIJqpvTYP9TbgtJr53pkJvnFyHJtZ418/v4veoIsXz0/PNOP4p49vmvNUtt9pXdIZkcW4MJHi1SthOhscvLCvY8XL7b0/FOfdwRhbWzxzprLcTamqM5Us0uqzz5ovfmU6zfnxFHs6feuyS9+7gzHeH4qzvdXDx3fd/vi/f3aS169G8DvMXAvnGJhOkSlWsVtMdAacXJxMMxLL0ei2EXBamUwWaPbYqBoGezv9Myk/hiH5na+fZjJVRBNwbjzJ7o4Az+9rozPgRDckUkrM81jAWq4aTCYLtHjtMwtqpZT1hekqJeN+sphgWs5xebbfFUVRbnNiOM63To5zfjKFz25hJJrl7ESac+NJOv0ODKBYNbCYBOF0iWJZJ18xeP1qBLNJEMuV0YTg++emaPbYSBfKRHMlUvna4puqUcujlvWVObqUXJxIUzUkI7E8yXyZxnoOtlG/zs2f97vReG7m8vAtNbdXWqpQwWU1YTNrnBpNEM6UQBqcGktxeiyJxaTxux/bxnC93XO+rDOZKtAbdGHUU0GklMh19hVzYSJF1ZAMR/OkCpV7Li4rVw1uRLK0+eyLCvzPjSfRDcnlqTTHtzctuMzfd09PMpEsEPTY+PIjPeRKtdrr3Y1O3DYzr1wOU64ahDOldRlMnx9PoRuSi5Npnt7ePBPQjsRyvDYQZiyeZ9IkiGVLRLO1xa2lqkGxnMFhMxNKF4lkSghRL2PX7OFof3Dm/osVnR+cnyKULlHRa3nY4/ECDS47P70a5eO7Wvjrd0Ywa4JP7W3jzFgSo56jP1v1jO+fnWQ0nifgtPCVo71UdMnfnhgjli3x9PZm9nauflt2ZWUsJpjuFEL8EbVZ6JuXqf8+95JXRVHuG4ORLLmSzs5275zNDGZzbjzJjy+FQNRSLExC8P5wglCmSLpQ5kqpSrPbRldDrUOc3SwYjRdrpZ7sZgqlMoYEgcRirp2irUq4MJ7GYhJ4bGZShQp/9fYwPocFq1njuV0tjCeKRLK1znNVXfKXbw1RNSSPbwkSz1XY0nL/N2wBqFQ+DEYr1aUvQJyPt29EeW8wTqPbSovXzuGeAPlylcFIFpNJQzckY/Fa6bFHNzfw1vUoAZeFXfXKLc/taeXSZJquBifSkHzzxBjtfsdtQdBa2dPpI1bvvOdzWBiYzvDTqxG6Gpx8YlfLHeXnfnBhisFIDrvFxK8f61vw7OS+Lj/vDcbY0uJZVL3sZKFWfSVVX7T27dMTRDMlAk4Lv/pYH61eO6Px/LrNJ9/X5ef9oRjbWr23zQyfGUsyHM0RzZawW0xMp0pYTLWa6tb654RuGFwLZylXDUxC4rRZcFpNtPnsfDAc5/RoAofFxIWJND6HGSMn8busbG/30uCy0ua3870zE7x+NUxFl+TK1fprIPnbD8bo8Ns52h+cKVEHkCxUAEgXqxgSkvky0UztzNC1UPaBDaYrusHFyTSNLutd07I2ksUE0//ylssnPvK/j/6uKMocylUDsybWpBPbUown8nz3zCRQq9n76OZGAKr1mRyzSbvt8k2ZYoUXz08xmSxg0gSf2NXKcCyHSUC5YmDWNDQhKFUNdvgc9Le4OT2apKzrWEwCl9VMqVwFakGgy6ohqX1JWswaFd1ACIFuSIZjeaZSBfZ2+hlLFNjV7pspq/faQJiB6QyD0Ryj8Ty/++zWZS1XtZ7dGj6XjdXZ5mh9BjyWrZWtC6WLJHKCA90BQpkyFk3wpUd66tep0NvoxGYxEcmU8TmseO218mUA//nlq7x0YRqLWeMPf24PO9vWtszb9lYvLquZRrcVTROcHEmQrXdjfGRTwx2zz8VK7RWo6Aa6sfBZ9iO9DRzpnX1R7nw0um2cGw/z6KZGhBAU681GCpXazvDC/nam0gXafeszwHmor4GHblmUbBiS756d4MVzU/XntFbn32bW8DktdAYcxLIlplNF3FYToUwZAeQNicUs0aWkWNb5y7eG0UStadBQLIchaw1uHu5r5NFNDXQ1ONjZ6uVGOEuhrFOqGiTr1WCyJZ10MU+6UEGI2uca1M6k9DU6OTeR4tkdzZg0Qblq0OqzkSlW52xRPhvDkFSN21ND0sUK8WyZ7gbnhvsOeX0gwvmJFJoQfPnRnmVp+LTWFhxMSyn/+0oMRFEeJAPTGX54YRqP3cwvPdS9odq/3poRcTOVYiJR4M/euIHbZuFn9rTy2tUIUsLnD3XOdBu0mU0MR3OEMyWKVR2zBulChdF4HrNJsKWl1vClWJVcDmUYTRSIZstUdB23zUw8WyJZL+dm0iCRr2LWBHazhiHBMGvoUrKnw8vFyVqe9UfHC7C5yU2xamA1azgtJuL58gMTTN8aPy8ilrtDqarz3mAcu8XEkd7ArI1Ajm4O8taNKJ0BB5ub3BiG5G9PjOO2mfnk7lZ+4/FNM/m78VyJC5Np7GZtZt+KZkucGU3SG3QxFMmRLtZm+8Lp4oKD6UyxwgfDcZo9dnZ3LD0Q/8mVMBcmUjisJn71aC/bWj2EM0Xa/Y5Za11/fGcrZ8eTdDc41+Q9f7NqSjRbm5l+fl87l6fTbKt37Xv5cojLUxl2tHl4bgNUu4nlSnwwlCBQL4PXE3QyHM1hM2vsavcxlSqSLeloQpAq1GaHq4ZEUFvs6rKa+foHY+RKVQoVHa/DwpZmD7FsqX4mxcbp0SR//c4IqWIFn92K1aRhrX/mfHpvK5oQ/OP5aaLZMoORLFOpAm0+B3/1zjAvnp+mzWdnIllE0wQ/OD/FlekM7T4HY4n8zCLbuymUdb7+wSjpQpVP7G5he6uXYkXnq++NUijr7Onw8ewiK+qslZtpdRJ536TYLTiYFkJ8727/l1J+ZvHDUZQHw41IFkNKUoUK4UyRnsaNU5e6q8HJz+xpI1uqsq/eAOKbJ8e5PJXBpAkaXBZK9ZmukVhtxuaNa1GGohkuT6dJ5Ss0uqzYLRrDsRyZQhUhamWesqUq0pDYLSbilTKaBppRO02bLlUxUZtdtZk1KlUDm92M32ljZ7uHG+EsdouJZKFCX6OTaK7MznbPTLrAreP/P57fyRtXI3gdFvo20HO/nJbjK+zkSIKTIwmgVl5s6yytlLsbnXQ31ipofDAc581rUUZjeboaHOzr8t+2EC5drOKxm9EEZIq15j0/uhgilC5ycTLNx3e2kCqU8Tqs7Gpf+Cnyn16NcjWUAVK0eO00eWz3vM3dJHIlssUqumFQqhoc6gmwr9M35+K0gMu64KY0txqL5zkzlqS/2c2OtoXnNO/v8nN6NMmudi9CCFp99tsa7VwLZWd+Prd70cNcNWfGksTzZWLZEo9sCmAzm7CZTVSqVWK5MrFciUJFx2o2oRsGZlMtqL65psJpNVGqGhQqOh0BOx6bBZvFxOcPd3IjlGEolufqdIYbkRzlqs60uYjXbqbZY+fyZJp/+c3zbGl2808e6+U7pye5Ec7xL752miO9DaQKFQQQy5WxmASpfIWqIUnmKzS4rFwLZXn8lqZRE8kC06kiu9q9t70nIpkSyXztAPKHF6a5Ec6xp8NLoX5W4WYqyUbyxNYmAi4rjS7rHT0ENqrFpHk8CowBXwPeQ1XwUJQFO9DtJ5Ip4Xda6PCvXbmkfLnKVKpId4NzQe2Vt7XeHjQ1uK0zXQ+f3NbEieEEVUPSF3Twp68PcXY8yXAsT7lSpWKA226m1eekL+hiNJ4nW9KJpEuUdAOTJtjb5cdmMuF1WhiJ5Tg9mpyZSRXUOiU6LGZavHYOdPvJlXQyJZ0Wr4NC2cDhqv3v/2fvPaPkSs87v99NdSvnrs4ZOacJmAEmkByRnBkmiUGJklaiKEs+xz72Oetd7x77HPuLd/esrT0r7a6s9XplrbQUSYkiRVIkxRlyZshJSIMBMIgNdI5VXTnf5A+3utANNIDuRg+ABu7vUwEV8Hah+r1PPe//+f9dksRbV+fZ3RWiWje4PFtkS5ufRMDNp/c9nCMeM7kqFtY9CSTxKBKTmQqKLOBdQad1QeawrT3A84ssDxfoCrup6SY+l9wsdBeSGj0uka6oB5csEvYqN6VsXpzJky7VOdATuWWy4cJrKZKAqty9m4JbkZjJL7hjiOSrGmcncnSGPSvqOq6WVy7Mki1rXEuW2JTwrzoS/WBvhNZGzPZyHB6McWYix54NkpJY1y22tgY4qxvoJsxmyvhcEuPFGqmSLemI+VzUNANRlIkHVAoVndl8lZaAimGanB7PUdVtKVlrpxtVkQi4ZP7uzDSqLFLVTaI+hXTJHmx2KxLDySI1wyLglon5XGQqGq1BleMjaUZSJWbzNZ7eFOOJgRj9cR/7usNMZSsE3QqbEn76Yl6eGLguVylUNf7m5ASGaTGTq/LSnuunAh1hN4MJPyOpEtlyncuzBSwsXtjRykSmskT2slGwT7I23rpvx1qK6TbgBeBXgF8Fvg98zbKsD9ZzYQ4ODzPtIQ+/+VTffV2DaVp87bArz20AACAASURBVNg4+YpGX9x7V8lcn9rTTk/US3fEw0CLXaz+1bFx/ugnVznXsEMLe2RGKxoWdgzu8eE0yUKN1oCbFr/FtVQRw7QQsJgv1HhmS5zNrUHGGse2hmmCZUsVTAu6Ih5e3t3Oof4Y/+qHFwm4ZfyqzFefGWAiU8GyLI6PpDEtOzExXapTrhtcmM7zu88MrNv7+CAxnCrx7fcmAfjU3va7Sm5bCalijcmsbVuXK2t03UEG+nh/FEkQ8KnyTYU0gKpI7OsO45JEao1Uu0/uamd0vkRryM2fvDZErqKTqxQ4M5Hl8X5bSz2Tq/KDszOAfSz+0e3LH3s/s7mFroiXqM+1ZFBsrVQ1sxkIVK2bvHJhlrGGdePvHOnHt4zDw90Q86tkyxphr4K0jKTmTnz7vUmmslXifhdfPtx30/2H+qIc2kBFznNbWwi4ZYJuhfFMmVShTlUz7CREn4uOsIfZQo2ZegW3LPHM5hZqmslb1+bpCrv5+dA82XIdQRBIBFRSxTpxv4tvn56kXDOYy9fY0REk7PFxZa5Ii1+lM+zhjStJhLpJVTPxqjI/v5KiI+zhQHeY2XyVSl3j1FiGX9zXyRP9Ub5+fJz3xjLUdNuGb09XmIG4/blJl+p8+71JLk7nGWjxLZE9jKfLvHZpjkTQzZcOdfHvX79KVTOI+VR2dYZWLFXSDZMffjBDtqzxwo7WpvTOYf1Yi2baAH4I/FAQBBW7qH5NEIT/3bKsP1rvBTo4OHw4GJZFqWYfpecr+l29VthrR/AuMJOrUqzpzBdrtIfc9MV8DLR4+fO3RjEtk3xF4/2JLKWaTsijkAi6cUkiFUw0w+4yXpor4pJEgm4JUYDuqIdcRaeqmYCFqkg8MRgn7nextS1ApW7wuf0dSKIdAiII8N54lnpDH+2SRcp146H2d80vOvLN3eX/6UoYS1cwTIuqZjKdr7LzDhd3WRTxqXKzQ3wjuzpDjKXLBN328BiASxbZ3JCPDLb4OTORx6OIS1wAZElAEGx9/O26taIoNIvf9eDZrS28c23edvPwKs3PliQKq3K5WSkv7mpjOmd3VdcydJaraLZkqqpjWdayGveNhE+VeWZLC0c2xfnppTl8LolCTadY06nUDXqiXvJVDa/Llg4d7I1yeCDKk1fn+ebxcYzGYJ8qQ1/MT7Fa51S6jG6aBBr7zq6OIG9cSSEJItO5Kv/sxe1M5qqcn8rTE/PS4lc5N5nj9FiWz+zrZGS+xNvX5smnSvzlsTH8boWLMwUmsxWKVZ2WgMr3zkxzYjTDFw91c24yR66ikQi66Yv7+NiiL4InRtOkinVSxTqTmUojcEngsb6VDy8CTGQqTQnPqdEMn9z94OvhNxpr+trcKKJfwi6k+4B/C3xr/Zbl4ODwYaNIIi/ubmNortRMoLsT15JFToxkGEz4bzmN/s61ea4liwTdMjs6QghYdEV9vDmUpKIZlGsGXtV237AQKNUNJrMVXJKASxapaia6CZZpYpoWpmVHle/piuB1SZydyNEV8fLlwz3NdX/lyABlTWcqU+Hrx8cJuGW+fLiXLz3WzXS2yuZWP3XDZDRVbiYmPozs7AiSr2pYFvfkqH53R4DvnZnCLYtsWUGR+uZQiu+cnkSRRL5ypJ/euN2JW7B564p4+eozg0ueoxkmY+kyiYDKZ/d3sb09RMirLJGxxP0qXzjUTbZcv6f+yK1BdzN9D+AXdrbSH/fRGlw+FOVukSXxrqzEAm6Zt4bmeXZLy4YvpBcjigLPbU3gViRyZY3BhI+OkO1ZP52t8P2z08T9Lja3+hFFkaObWzg+nGYiW8GtSHx2fwfvXktzaSaPZlqoskhX2EswbIe7VOo6qaJtgXh6PMvHd7YR96uIgn0adHmmgAW8fnmOoaQdbW5a9pfbyWyZgCojCgIdYQ8DcR8VzaSum8zmq/TGvJydzNEecvPC9rYlg6kDcT8jKdunGsGOO1dlCcOyVlW8tQRUAm6ZYk2nv+XRnBH5sFnLAOL/B+wCfgD8b5ZlnVvFc58A/hB7huiEZVn/gyAI/xj4DDAK/JZlWRtPTe/gsEHZlAisSgrw+uUk2bLGZLZCT9TDbL5GX9zXDCzIVzXevjoPQMzv4nMHbOnIydEMV2aLlOoGJlCqm7YHrASiADXNpFyzL0CCAEGPRKVuIQr2/aoskq9o/IvP7WYyW8Hnlmnxq5yfyhPzu5rpYmfGc4A9vFbVTOJ+tTng4lYkdm8QLehakSVxyVDTh01Ft3hqMIYoCOSrd+6EX5jOM5GpkCrWKFQ1VFlioMXHZ/Z13rJI/OG5GYbmivhVmX/0dB+dYc+yeufOsOe+zh+A7VizHi4hHxavXUqSr2q8fjnJbz3d91AV1JIosL8nzJ+9NcKl2QJ7u0N8ZFsrQbfMTy8lSZc0vvv+NL9zpJ+huYLt3uFW8LslXr+cZDJTaQYGAcwWqjw+EOXchK3Fr+smlgWj6TKf2NnOa5fm7MHCikZr0M18qU6+ohH2quQqOjXNIOCWOT+dxzLt+ZS6YTIQ9zNXrBJQFVKFGiPzJbYk/FhAsa4T8l6XH+3tDrOlNYBLFnnlwiynx7Jsbg0gi6s7XfOpMr/1VB96Y7jbYf1ZS2f6y0AJ2AL8d4t+GQXAsizrdm2BUeAjlmVVBUH4S0EQjgLPW5Z1RBCEfwJ8FvjmGtbk4ODwIVOoalyZKzI+X2J3V4i/PTVJqW4009QAvIpE1OciWahxebbAf3htiHeuzlOoaWTLWnOI0LRABGq6hUsRGsNbZtPGTrQE/KpIRTMQgPaQm0N9EdwuiU2tAXIVjf/n58OkijWCboXfPNxHyKs0j927o15CnrvXxDrcnu3tAYbmiqiySP8KBu72dIW5lixSrunUNJN0uU531MtYutwspk3TWiJhWLDCK9cNTo9n+dmVFB6XxK890bOs/ZzDrSnVdJKFGuJDVEQvpqqZTSehBemaICyceF2XeL19LY1HkfC4JCxLYjpbxSWJ6KZFQLW97vvjPgzDolzXKdUNROw5je6IFwS7OJ7KVvEqtiWnZVkkG7I2t+RjplBHEgUm0hUCqsxouoJgWfybVy8R9qoMxv0EvTKSIPD6pTkSIQ+vXpzj6OY4+7rDTYenhU51slCjN+ajrpuU6vqqNP/TuQoBt7JsSqNlWfzogxlG5ssc2RR/oL8MPsisRTO9ZsGhZVkzi/6oA3uA1xp/fgV7oNEpph0cHkCuJUvEfS6uzBT40QezSILAR7e3Ums4NOiGycnRDNvbAxzsDfEnr1/jvVSJa8kSummhGba/60Knudy46C14vka8CpmSZmtfRYFizcAwLVS3wLVkiaBHYTZfpS3k4dvvTXJ6LNMMP9BM+7VuPHZ3+HBJBNz8zpH+FT/+yYEoEZ/Cn705zESmil9VaAu52dVhX8BHUiW+d2YKvyrzpcds//UXdrQ2faavztm6z0rdIF2q31RMV+oGp8YyRH2uNVnHPez0t9jF2Eq++GxEoj4XL+xoZTpX5fFFg5SfP9jFePq6r/Ngi48LU3nifhe6YeKSJbojXjyKxInRNIYpEPO5QLB97L0uif6YjwM9EX5hZxtXZgtEvSqT2Qr5mo5l0jxtqesWMb+LXEWzA6UkkUJVQzdMJAEmMlUCbhfD80U+mmjlymyB2XyNiWyFYs1gOFnkx2EP/+vLO0gsGhR8ciDKz6+kbGmUafE3JyfwqTIf2564pRUj2LK7t6/Ooyoiv3G476aC2g4ZKgD2jMmuzhDZcp35Up3+mG/DBcLcL9Z31HiFCIKwB4gDWa6HcuWAm0SYgiB8FfgqQE9Pz7quo++ffn9dX8/B4WGmL+Yj4JZJlWpU6gaiKBD0yHx6XwcAp8ayvHV1HtOySBVrnJ3IkSlrGKbF4q3etGyzflkEw7S7SSGPTMSrUtdNEOxuiU+VqOumnfyliKSKNY4Npzm6uYU3h1JkyxotAZUXd7c/NF6lGw3TNPnx+TkCbnlF8d6CILC1NUBn2Etr0IPXJfErj1/f1y/NFtAMi0xDSrQpYTvD/EIjVS7mc1GqG0S8it0hvIE3riQ5P5VvPjbxiLkWlGo615IleqLeJXKBBTyyRMCj4Fbuy6X/nrCcy0XIoxBa9HebEn4iXgVJFGgJqMQDbiTRPiEbz1aQRYGqZjIyXyJZrFHXLSZzFXZ0BGkJqFhYbGkLMJWrUNMMJjJVZElAM+wOdV03EQTbuq+maaiKbAdMKRLPbokzlq4gizI+l4QiCc1h8KpmNkKsRN4ZTiMAmXKdT+xsoyXgZl9PhPfGMlyczjcGsWFTwndbqV6qWGMuX2UqV6E74uFTe5c2G3wumf64j5H5Ejvag5RqOn/57hh13WRfd5jnt63dF/1R4p7/RgmCEAX+GPgicBBY+J8NYhfXS7As60+BPwU4dOjQwxGV4+CwAQl5FX73mUEMy+Lb700R8So8s7mFRMAuWFySwFyhim5aSNiFk2VZBFSJvriXsfkyyZJ99FrRlv4ql+sGsiQQ9tphLkcGY1ycLZIpaxzsDTGVrSFLIiGPwl++O0rU58LrkjjQG2Gb04G8b3z9xETTik8UBJ5sRMvfDkEQ+PiuNi5M5/G6ZN4fz7K7M4QoCsT9Ls5N5gh7FRKBm1MpY36Vzx+8tYWj2jjGFwVh1R7M94KpbIXjI2l6Y74VD/2uhu+cnmI2X8WvynzlaP9NmujumA+PSyboUR4qvfSqsWAiW0E3LHZ1hnh5Twdm4wv8ZKbCxZk8F2fyZMsaVc3+4l+tG/zfr18l4nOxpyvMV4/2U9EMjg3Ps7XNT76qE/G6CLpl5go1xEYEh0+V8boU9naFeHF3O5taA/yHnw6RKWu8O5wm6FYIexQ8Lol0qYZugEcRuTKb5zunpzAt+PEHsxzsi1Co6vhVicuzRcIema6o946NhKcH47x6YY64X2VoroRumEs62aIo8Nn9nU13l/lizW5qcF1i5XBn7mkxLQiCDPwF8I8ty5oRBOE48AfAvwI+BrxzL9fj4PCwohsmmmF9KJHFv/VUP1vbgvhcMjsWpQuqioQsChimRc0wEQU79jvocVHWLPpbAqTLGYwbvhJLAsR8Cl5FpFi12NoW4MjmBJM5O70s4HbxK0+0IgLvDqepagYTmTJel33he/faPE8M3LmIc1jKepRSCylsAKX6yq34trQGEAX47vvTgB3xfLA3QqpYb3YV5wp20uFqOLq5hdagm4jX9UBGxP/00hxz+RrXkiU2J/zr7kOtGXYRVDfs+YMb6+Ud7QF+mquyo/3D9R9/0JnJV4n5XFxNFjkxmuFAb6RZlG5pCxL0KPzDB7N0RxV00yJTqpGr6FxLlfj3r13lt5/u48pckXzFTjPUdKvR8bV47VISC0gEVUK6wq7OEJ/Y2cqZCdsppFjVEUW78eCSRX5xf5RKXadYM+iP+chVdTyKxPHhNKWqnQ5bqtupjT5VYi5fAyDkdfH0YJyw9/af84jPxaf3dnB6PEt/3HdLScjCl6uYX+WFHa3M5Ko8tsECYWq6vR8tuAPdS+51Z/oLwGPAv2z8x/3PwBuCIPwcGAP+zT1ej4PDQ0eppvO1Y2MUazof39m2Ju2oYVr84Nw088U6H92eoGvRkbpbkXh+mUjkhVCUbFlr6vJ8qkJdN4n4XAiCQHfUw0yuhqabRH0yHWEPm1uDHOyL8N33p0kEBAKqzPfOTjGdq6AZFqWqwbFraUzL4uxkDs2wCLpl3IqMW5GYL9XX/matkEszBd4cStEf923oY0+B6zHi69G3/dXHu5FEAZ9LumVQyu1WkynXMU2rWdhva7MHGgNueU3OHJIoPNBa6Ra/yly+RsijNLvo68nLe9q5MF1goGV5reux4QySKHBsOMOTA7FHtjsd86tYli0xq9QNjg+nm97LOzuCpIt1vnioi6Db9g7/+7MzHB+ZRzctfC6JN4dsOdtMvsqujhBRn4svHeriR+dn6Y3ZoUDzpTq7OoL8wfODvHohiUsWmS/W8asSbkXiQHeE+XKNN64kqWgmOzuC9kleVWe+UGMiU8WrykR8CtvbA4xnyvzeM4O8emGW46cmmC/VeGFH24p+3ue3JTg8GFvxZ241gTB34qeX5hhOlnh6U/ym5Nz1ZCpb4VunJhAEgS8c7LrnEq97WkxblvU17BjyxbwN/Mt7uQ4Hh4eZZKFGoWFTNjpfWlNxMZ27bvJ/cjSzpJhejqG5Iv/+p0Ncni3gUyU+uq2VbW32xaEv5kUQRWZyVbwu21faMu3QFVWRmclXyZY1Dg9EOTORQ5UlhlNlO07agi8+1sUbV1IcG07jdUmkijV6Y0FURWJLa4CnB69rda/MFshXNfZ0hdf1mP/YSJpcReP0eJZDfZEN6yKx3sV0vqqjGSYVzR5kWs4t4FaIgv2lzTStZge1N+bjD54bvKsibzJbaabiPWh8bHsru7tCRLyu2w6NrZWYX+XI5lsf+3dHPVyZLdId9TyyhTTYNopfOdrPN06MA8ISW8YDPRH2d4eb78/VZJFzUznG0x5aAiq//kQvF2cLzOSqvLirnb64l82tAX5yaY6fX0kykSnTHvaypzPER7YncCsye7pCTGTKnJvM8Tcnx/EoEqIoIAoC7UE3lgURr4uXGzHi//xvz2JaVjO1UWpY4V2YyXM1WcStSOiGhWXZJxGmaTGZrRDzu/C6bv4dvDRToFjT2dsVQpbW/v8+nCpxejzD1tYgOzqCGKbFmYksqiwtOaVcoFjTOT1mq3dtOcyHV0yPp8tohgVYTGQrD3cx7eDg8OHTFfGwpTVAtlLnQM/qkrIWiPtVIl6FbEVbUWLcTK5KulSn2tDaiYLAplY/dc3khZ2tGCb89clxZvO2XysWxAMKqWINRRI5O5njn31yG1O5KpfninRFPbSHPGxK+NndFaYj7ME0LU6OZqjWTU6NZtjTHaYj5G4OWk1lK3zvjC0bKNUMntmyfn7Lm1r8pAo1OsMefMtcrDYKiy+ja/dlus6FmQKFqk4B24ljNd0szbCaR+uGeV37czdF3oJzgUsW+fLh3geuoBZFYUnYzL3mxV3t5AY1xzYS6Ix4+YPnN1HTzZs+J4s/g4Mtfl7c3YFbkZBFkXxN5wsHuyhUdcLe69rz1y8neftaGlEA3bQL9vfHs+zpCtMd9fLxnW28fnmOVLGOgIVPVWgNqYDAP395x5KTmP09kYZbjcyW1gCpYp1CVWdrq59Msc50rkrIozQKaot/OD/Lhek8AbfMbz7Vt6SRMJ4u8/dn7X2xphkrGhS+Fa9emKVQ1Rmbr7C1LcDJ0QxvDqUAUBWRwZal1wqvItEZ9jSGiT9cadGOjiCj6TKiILDtQyzab8XGvSo4ODgsiyyJvLTn7uJis2WNfd1hBhP+FXVh9/WE+cj2BH97ahJVFpnIVGgJuHDJEuen8vz0UpLxTJlK3SDsUUgEVfrjfqazVYaSBSbSZV69OEci4Cbmc/H0pjj7eyLNY8mYX+X3nh3k9/7iBLIoMF+qMzRX5E/euMq2tiCH+iJsbmzW2XKdEyMZemPeplfr3XJ4MMb+njCqLG7ojt6CgwqAax2K6Ra/i/PTeTyKRNsqO0FbWv1UtQSaYbJ/jV/6biTdkPzUdZNidXVevI8Coig8kFry+4UqS6iyxEyuyo8+mCHkUXhpT/tNp1p7u0KkijWqmsGBnjCyJOJTZVLFOnG/C8uyGxBhr4IsCrSHbXeQnYu+XF6cyVOo6FhYGIZln+RUZXrj3mYhrRsmb16dZyDuQ9rZhmaY9Ma9vLS7Hb2RFLqtLcjurhB/+sY1/vCVK7y4u9hcb7Gmc2W2ANgFpSgKS3Tzd7t3tQRUClWdmN+FJAosVhIt510uigJfONRFTTeXDYt5ayjFhZkCB3sjdz2QG3ArfPFQ9129xt3gFNMODo8wl2YKvHJhlo6wm0/v7UQSBQpVjW+eGEc3bU3gJ3bduTD3qzL/+OPb6Ap7ODmaIVeu8+bVeVRZZGd7P8WaTtijcKgnwhcf60aWRKayFV6/PEdFM2zdrSojSyKmabGtLYhbkahqBt86NUm+qvHS7nZ2dQSZzlYRBQi5Fco1nUyjeH68P8pn93fyn98cxsLi+2en+YPnNq3be/UwJIdJosBCco4s3v3PkyzW2dGQEc3kq8QDK7coFASBvTdcQHMVO50v6JZ5ZnPLqj1uFyQ/MZ+LjvuchuiwcTg9niFdqpMu1RlPlxm4ocMqSyIf33ldn2yYFl87Nka6VCfsUShrBj6XxO6OAPmawef2d/LslqWzFQG3QnfUy2S2iiSCIomEfcqSIvLcVJ5ToxkAjm6O8/qVJP/yB5fYnAjwv7y8HWi4JHHdaeNqssh/+/wmjo+kccsSP/pgFrCHFh/ri9IV8fLpfR2Ua8ayUozV8PKeDuYKVWI++/f8QE8EtyItCW260S1EEIRl907DtHh3OA3Au9fmPxR3m3uJU0w7ODzCnJnIUtdNhmaLTOcqdEW8mCYYjShC7UbrjUXMF2u8cmGWoFvhhR2tGKbFsZE0J0ezKJJAxKugGRbHRzMc6AmjyBKf2tNOwK3wF++MMjpfoitiSzkGWvx8Ylc7uYrGG5eTnB7P8tzWFiazFWbzVQA+mMrz5ECc0fkyflXG65LxNtxK+uJeVFmiP+5je3uQmVyV8CqdIB4F3IpEWbf19J51aE0PxH2cnciiSOKyvs+r5dhwuhnM0hP13lTU3ImQV+HF3Xd3KuPw6LEp4efSTBG/W6YtdOcTlppuNE9B3p/M0RV2U9dNvKpCPOBhJFXm2S1Ln7OtLUDU5+K5rXGShRqCIHCgN8JTi2Y+prMVTo9nqesGiiRwZizLTK5KplRnMltpnrTt6w7zCzvaODuR5WPbE7QG3by8p4PhVImzkzngurML2FKVmVyVkfkSA3HfmjvU0g0ypapucGXODnzpjHi4PFvktUtztAXdfP5g123nAiRRYDDh5+pckS2tG99dximmH2HuNrRm5F+8tE4rcbhf7OoMcWE6z0SmwrdPT/Krj/cSbVgpzeSrt+0W/Pnbo5way9AZ9rClLcAHkzneH89hmCbtIQ9Bt8JEtkK2rOF1yfw3z/YT8ip85/Qk5bpOslgjEXTzu88MNLV2f/TqFc5P5+kMe9jZEaQz7CHsVZjLVwl5ZHTTYktrgGTBtocSRYHP7OugP+6nqhn83ftTlOs6Tw3G6Ii4m96pa2U8XSZb1tjeHvhQBsbuNTXtun1dRbt7D9muiIdntrTgUaRlQ0JWS2tQ5dwkuGSRqCNHcLhHbEoE+P3nfMiicNNpyFS2QqpYY1tbsBlH7nXJPLMlznCqzEDcxzdPTeBTJB7rj1KuG+zqDGGaFhdmbAnUQIsfv1umL+6jUNXpifmwLLAsGJkvN+dSzk/nUWWBUs2ippnUDZNyXSfk9vCnr1/FtGAw4ee5rS0MzRWZK9R451qax/pieFx2M+GFHa3kKxoeRWIqW6Ej7GEuX+Wvjo9hWXbH+1Df+ljenZ/KM5IqA3azYzhZwrJgOlclV9GI3cED+9N7O6jpxn2xsltvnGLaweERZSRVIlfReHIgxrnJHDXNZDpXIepzMdDiv21XsFjTSBaq1HWTZKFGS0BFFkUSQZWpbIXBFj9Rn4tsWePCTJ6KZjCZLeFTg/TGfFxLljiyKc4vHuhiOFXi2+9N0BHyki7VqNYNrswWmS/VCXoUDNPiarLEG5dTJIIqj/VH+ej2Vk6OZrAs0Bvd82vJEpOZCoZp8Z3Tk7QE3OzvCfPcMjZ+KyFVrPE3pyawLJgv1db8Og8SVWPR7XXIYzg1luGNy/YA0i8eEO9ao76nMWzqVqRVOYM4ONwtrmVs43Jljb8+OYFhWszma7yw47r948HeKAd7o5yfynOwofnf3Brg2S0tZEp1fnR+houNmO5fOtBFT8zLrz/ZS76ikS7X+eG5GdyKZLsWYUuchlMlkoU6PlUiU64TdCso0kL6q0a2ojGds92Pjg3PU9FMXLJI3TDxIJGr2MOlE5ky7w6nkUSB33yqj3LdoHHYSLG2cj/4O9ER9qBIApZlD1xGvAr5qta4fesvw6WazpmJHJ1hDz2x6ydalmWhm9aanZhM08K0rPvS+HB2KweHR5BcRWuka1lEfQrXUiVM00IR77wJnZ3I8erFWeqGxb7uME8MRAm6FZ7d2oIkwfvjObojHuZLdSI+FxGfQlUz+d6ZGXZ2VHhpTzuDLT7cisTQXJFvnhxnOltBIMOerhAZr0ZVM/jb9yb50iF7an48U7Yt/2p2t0OVRdpCbqJexbZ7syw6Ix78qkyuUm92OmZyVSp1g9F0ia6Id1UFmmlazQvQYreJjYywyBtvPeYo9UUyoNtJglaDEw3v8KBgWnZxBrYWeDkGEz4G5nzUNLM5qPhf3x1jdL6ELAq0hTzopv1ct2J7TCeCbroi3mYh+sblJNlyHcOwCHpkPrGzjcf6ovxfr1wmEVBJFWqkSzUqusm5qRyVuo7XZYdkHegJE/IoTGbK/MefDeNTbdeRhfVblkVf3MezW1soVnUeX2EQS1UzMC1rWau9BTrCHn7nyABAMyBsJa4dPz4/y3CqhCQK/PaRfvyqjGFafPPEONO5Ks9ubVm1E1WppvNXx8cp1fTGNWZ1ErG7xSmmHRweUI6PpJkv1jg8GF93K6t0qcbQXAGfKhNwy82NZzxTZssdbIUuzxbIVTTKNZ3P7GnnaKNjO5Ep8/rlJKlCHVEQeKI/yke3J5jN1Xj9ShJFEkmXbZ1hwK0w3+j81jRbf7i9zR6OmcpWmMlVGZ0v82tPdPPkQIyhuQIRr4IqS7gkgXxFJ1/Rmc1XOT9d4OjmOH1xH4os0BH20BP1MZuvcngwxt++N8ls3raS+u0j/St+jxJBNy/vaSddqrOvZ2MPxyywuB7QjFs/bqUc6osiSwKqLK3IQtHBYSOxkB44V6jdUvKmyhKf2dfZ/PO1ZBHDtOgIeQh4ZJ7eFF/2/vpuGAAAIABJREFUlC9VqPHBVJ5cpc5svkaqUEUzTYo1nXeH0+zsDPHbT/fzw3PTXEuV0AyTkyNp8uU6w/NwZFOcZ7ck+NiOVnTD5K+Oj3NhOk/Io/CJXW10hD0kAiqqLDFXqN62OM1XNb77/hSWBZ/a00HdMPnGiXEM0+Iz+zpue+K0XMquYVq8OZSirpsc2Ry/aQBRakhpRIGmI0i+0XUHuDxTWHUxPZ2rkq9oVDSDbxwf5+O72lb1GoWqxptD80S8ypoSdZ1i2sHhAWQqW+HnV+zjc8uimc61XhwfzpAIqpRqBp/c1cZbV+cp1w1238InWDNMJMHWE25u9fNnb4/gkkROjGZ4clOc8XSZP/rJEJPZCgK2lrYv7qUj7GV7u0BryM3FmTwuSeL8VI53h9O8fsnuxkR8Ll7a3cbrl+c5M5klW66jGaAqJsOpMs9tTbC1LcC71+aJ+V0Mp0pMZatEfQrpkq1VKNcNvnVyghOjmabPtqqIvHphlvPTecIehYomrlpDvfkhGIxZzOLe8fJ9ttUhiQIHezdW5LCDw2q4k+TtRvrjPp4YiFKs6jy9KX7LyPgfnJuhqhnM5qu0Bt0EPS58qsz7E3bX+T/9fJjBFj97usIEPS5kUeDMRA7NNChUdZ7f2sKnGkV8XTdxSSK9MS+iIFCs6pwczXBkc5wfvz3C0FwRtyLyO0f76Yvd/LNcmMrx3dNTlBod766wl2vJIh5FYiJTuaN8K1mo8bMrSVoCKkc2xbk4necnF2fxKBJel3STt/ULO1rpjnppD7mbne+wV2F7e5CJTJmDvau3yuyJeumJenlzKEXE5+L1S0m6I96mjOZOvHV1ngvTecDuuC8O8lkJTjHt4PAA4lNlWwunm4Rvoz1bKxGfC7+q0Bp0Ew+ofOE2/pxXk0W+f2Yar0vilx/vYS5fozPkIVOuo5sWkiAwm68R9bmYK1RpD7p5YXsrf/ueLSP58pO9bG8PcjVZ5PJsgTcuJ5nNV7iWLKEqIru7QmxKBHntcqqh7RNQZYGBuJ++xiYe9bloDbn5+zPTJIs1gqqMaVr43DI7O4JsSQT46aU5ijWduXyNfEXjwlSek2MZWgIqXpfMZ/Z1bGiPaAcHhwcfQRCWOHTciqhPYSprNzCeGIgxk6vw7vA8UZ+CIooYps6x4Xn+7v1Jjm5qwe2S2N4e5PJMAa8qIYkihmlxaSbPydEMMb+LnR1B2kIefvTBDADvjWVIFmocG0kjiwLFmsG//sLeZme4rpukilXOTuYYz5QxLfjzt0bojfvIVzRkSVyRu8lbV1OMzpcZbQxTnp3McXG6gCwJfGKZRpBbkW7q9AuCwCd2rSwefTlcssgvHeyiJaBycjSDW5HwqSsfbFwYeFYkgYB79aWxU0w7ODyAhDwKv/5kL4Wqdsco77Xw0W0JtrYGiPldd5ykvjpnH1sWqjozuQptITebE36qusGXD/eimxbtITeP90VRZRFVFvnWe5PM5O0BxajXxW881YcsCozOl5jKVshXdSzszXywxc/B3gjfPq1imZAIqvTGfGxrC/DKhVlaAiov7+lgaLbIlbkik5kybkXiQG8Et0vi6OYWarpBV8RLQJU52BuhN+bjvfEskijQ4lfZ0xX6UN7HjYbI9Y70xp+fd3DYuHxufxczuSqtIVuKcX46z1i6zMWZIqWwzraOEOlSnRa/SqZc52hnC08PxviPb1yjP+5jaK5I/sQ4Pz4/S6Gm0R708E8+uY1EQOX8tJfZfJWnB+NcDRR5+5rt+S+LQjMFtVTT+ct3Rvn5UIqIV8HnkpFEgVhAtRNq/SqdEQ/6QsDMbeZN2kMeriVLeF0SIY+CKArs7gphWTT9p+8VRzfH6Y/7CHuV2+q9b+SxvijtITcBVVmTM5FTTDusGcda78PFsixMk7u2d1sOURRIBNVmwuDt2NsdZjxdZr5Up1w3mo4LqizidUn85btjJAs1NiX87OgIMpKyfaCxwO+WsRrigo9sa+X0eJaY30VNs11AEkGVUt1AEAT++FcPUKzpZMsaumlyYSpvR1VXbW30gd4w//mtYQo1HUUSiftdPDlgSwxUWeLXnuihUNWbx3pfOTrAfLFmx/Deh3jZB5G4T2GuIY1pDTnWcw6PLpW6wUy+SmfYs6yTx4eNSxaXOFns7w7zNycnEAWo6iZbW/1sSvgp13QOD8bYlAggiQKqIvH+eJYdHUHevZbGwmI2V8OjyExkyiQCKiGPncSYCLrZ3BpgIOHnzESOZ7bEm9Z/88U6hZqOZlhopsVLe9rZ3OojVbT3h4G4j1JN5+/PTlOu6zzWH+WJ/tiy8zuP90cZaPHhV2XcisRHtiUIumW6It5lU0hruoFL+nDSZAVBWLVEY4G7abg4xbSDwwNIrqLxF++MohkWTwxEV3RsuBpev5zk1GiG/riPz+7vvO1jW4NuBhJ+8mNZXr0whyqL6KZFf9yHYVqkijWKjU33he2t7O0O0x3xMJuvMZYuc6Shl3PJIl881M0719L0xe1N68Rwhrl8la8dG+NTezvYlPAT96scH0lzZa5IrqKxrT1I3K8iCirtIQ+qJFLTTQSEZhIXXJ+UXyDkUdZ9cHOj41VkwL5YBlbRtXFweJiwLIuvHx8jU9bojnr5/MGue76G98eznJnIsqszxP6eCN1RL//0k9v4o59cweuSebw/ysWZoh0G0+gag91BfazhE61IIpIIiYCbREClNehmOFXiXCO4xTWUojfmpS/muykYpSviYXdnCK9Loj3kZjJT5dJMibaQim7CidEMbllCN0zOTebRDJNsWbtlZPdiF564X71lcu5bV1O8ey1NZ8TD5w90rTrl9EHF2U0dHB5ASo2OAdhep+vNlVnb/3Q4VbKHV+7QmXE1fDsFwR6csSy7yP7VJ3o41Bvh3756BUkUGEoW+eW+Hnpi3mWH924c5vEoEq9dSgL2FwiwJ8FfPT+LhUV31MOXn+xtPv75rS2cGMlQ0w1002I4VaI35nW00CukZFy38CjU1sHOw8FhA2JaUKjafssL+869YuGk8ecNt4ufXUmxv+E6sb09yL/71QOYlp0wu1AUHx9JL9ttXSis5wpVBARaAirpUh2XLKIZJucmc1yaKRD1KfzK473Ml2okAm6kRjjNLzQi0k3T4t/9dIjZfJWTo2miPhf9cT9+VaYlqDJXqBHyKCjS3e+zQ42E08lMhYpm3HJAc6PxcPwUDg4PGR1hD89siTNfrHN4cPU2PWDrkUWBZQ3snxyIcWw4zda2wIqOOJ8ciBH2KgTdMn/3/jR13aTa8FYzLdtb9GrS7qLEA8vLBy7PFjg9lmGgxd9M4NrdGSLfuKjt6bKdRIZTRX42lCJX1vjcgaVd819+vIfntyV45+o8Fd3E4xL5458MkQiq/OKBrjWb/T8qbIr7SBZse0JH+uLwqCKJAi/uaefKbIHdXffO9vK9RshRb8xLb9TLlbniTZpiQRCQBAh5FSJehUxZo+8OuuNE4PqQYNTn4h893UddN/n68XHOT+c5NVrnarJExOtioMXHS7vbsaC5X4qiwEt72vkPr10l5FYYSZVoD3k4uiVOb8zHs1tamMpW1hT7bZgWumk2Z3Me74/y1tA8/S2+h6aQhntcTAuC0AF8D9gB+C3L0gVB+EPgEHDKsqz//l6ux8HhQeZuLMfG5st85/QkLlnkS491L3EEsSwLlyzyCztbV6wRmy/WOD2eJeRReHlvO8PJEjvabV/owYSfzoiHroiHLz3evezQx4/Pz/KffnYNRRI50BtmR0cQr0tGlkSe3dKy5LHXkiV8LhEsmfwNXSNFEumKePn8IXvd3zo1gW5aTGWrpEt1WoN3njy/mizy5lCKnqj3oUg1XA2lRelnhdq97cg5ODxIDLb473mwxwdTeUzLPlH77SN9PLctgVsWuTiTJ+xxLXHOUGWJHR0h3h/PIq7y5M3rkvG64NP7OrgyW2RTws/FmQIHesO8c3Wea8kSLlnkc/s76Qh7APvU8NP7OvjjnwzREfYQ97ualnhxv3rLMKV8VeOHZ2cQRYEXd7ct2f/LdZ2vHRunWNX5xK42trYF2NYWZFsjU2AxharG6HyZ/vjGLLLvdRsnDXwUeAdAEIQDgM+yrKOASxCEx+7xehwcHkqG50vopkW5bjCZrSy578Rohu+fmeavT04w1bjPNC2mspVmt/lGTo1lmcvXuDJbxDAsntuaINEoXDvDHn7/2UF+/7nBJRrmxZwcTZMu17mWKlKq6bd1EDmyKU5LwINh2al64+nyso+zLIs9XSHcikRf3Lvi5Lx3rs0zX6zz3lj2nh/x3m/mS/Xrtwu1+7gSB4dHj33dYVRFZGtbgKBbwa/KvHV1nh+cneHrx8dJL/r9BHuvKtZ03hxKNf/OslaeNNoe8vCLBzqJB1Re3tvOfKlOrqJxcjRNTTMYmS8tefzRzS18el8HW9uCdN6m0TKTqza/mJ+fyjOZrTCeLnNxprDkcQs2paZlcS1ZvO1av3ligh+fn+Vb702u+Od7kLin5b9lWVWgukjfeBh4pXH7FeBJ4Pi9XJODw8PI7s4QE5kyblm6qftS02xzNMuCmm7ffuXCLB9M2elZv3G49yZpSH/cx6WZAj5VahbRi7nTEMnuzhDnp/IE3QpHt7Q0h2mWw6vK/NqTPbx7LW2vt7HGyWyFN4dStPhVJjJlchWNl/Z08PvPDd7h3VjKQNzPXL5Ga9C9qnjxh4GuiIfJnF1Ed0fvrWWVg8Ojzq7OELtuCMZa2N9My6KuL41SGmjxcWW2yEDcR1Uz+OaJcXIVjRd3t684SOapTfFmaMp/eWeU6WyFq8kiiaCbne03h3R96VA3F2cKnJvK8eZQiqdvCFx5ayjFu8NpPC6J3zjcS0/Uy4mRNIIg0BXxLHlsV8TDYMJPrlxv6sJvxcL7UFuPaNb7wP2+koSBq43bOWDnjQ8QBOGrwFcBenp6ltx3t9ZsDo82G9nar6YbVOrGLQNdoj4Xv/ZE77L3Pd4fRRTB55Kber1k0S6wchWNmm7eVExvbQvQG/Mii8KyGuw78dKeDroiXkp1nUMN+Uq2XL/JgWOBx/qiCAh4XNdjqt8cSjGZqXBuModbkfAoEpdm8qv2MT08GGNvdwi3LD00k+QrZVtbiDOTdsrXtvabj1odHBzuLUc3x/G6JCJe100BKS/tbqe81cDrkhiZL5Mq2p3rSzOFOxbThmmRq2iEG77PAC/uauPcVJ5febznlvZxsiRyYTrPdLbKdLbKtrYAsUWnfgvXikrdoFjV6Qh7+N1nBhAQbpq/kSWRT+/tWNH78Om97bw3nuXgKmPEHxTudzGdBRZ29GDjz0uwLOtPgT8FOHTo0MrPNxwcHlIqdYO/eGeUYk3nmS3xVWurXbJ4k9Xe81sTHB9J0xe7tV5toeg1TWtJEVrTDc5N5pdo7BaYzlWYzFTY3h5kbyPxyrIs3ric5I3LSRJBlV9/srepsxuaK/LqhVnaQm5e3tOxpIPdFfYwmanQFfbgd8uUagY7O5aPP19MqaaTq2i0h9xN14/VmPk/THSE3FQ10458D69MFuPg4PDh4Vakm7q/CwiC0NyPO8MeOiMesuX6Td3tBWZyFcYzFba2BXjl/Cyj82UGWnx8phE7HvOrN82ozOSq+FSJwCI/6M6Ih4lMhZBHvilF8MimOIIg0BpQCXoU/urYGMWazkt72mkPeRhJlZgv1dndGVqVf/fJsSxX54rUNJNfWqVVYU03SBZqtAXda2r2rAf3+4ryNvB7wDeAjwF/dl9X4+CwAchW6hQberWJTIWDyzegV0VH2MNn9nUyNl/mjctJdneGiPiWdr2rmsE3T06QLdX55O42NiXsye7XLyX5YCqPIMBvHO5rxrJWNYO/OTmBZliMzJf5/MEuprIV/ss7o7x2aQ6vS2Jra5BMWWsWt++PZynXDa4lS6SKtSUDhU9tirOtPYi/EbW+Eip1g//yziiVusFjfVGObF5fv+6NxjdOjGMBFvC1YxP8xtOrk8g4ODjcHxZ8+m/Fucks/8cPLuKWJZ7fliDT0F9PZCq3fM7J0TRvXE7hkkV+/cnepi//U4NxOsMefnhuhv/3zRE+t7+T9pAt4Yj51Wa3+cpsgelcFYAPJvO4JJFvn57EsiBdqvPCjtZb/tvDqRLj6TJ7u8OEPAoTGXs2ZjJbWVVQmWVZfOP4OKlifckXh3vNPS3hBUFQBEF4BdgL/AhQsDXUPwNMy7KO3cv1ODhsRNqCbvb1hOmMeDg8sDbbvOWo6QbfOT3JydEMPzg3c9P9c/kaqUIN3bRuGjS5FTduiFfmiuQrGm5FQhZF2kJuOhYdbW5vDyKJAu0hd7MoX0zU51pVt2MmX2kOGaaKzsCdtmh4STPM2zzSwcFhI3FxpoBlQbGmU6zpPL8tQUfYzfO3cSxasMms6+ZNzkmFqk65blDTzKY39AJVzSBTqtMR9hDxKrhk8SbbvNuVwuW6znffn+LkaIZ/+MC+1jy/1V7vR7YlVpUbYFqQbqS6Ju/jUPW9HkDUsDvQi3n3Xq7BwWGjIwjCshvkjfKLlfLutXlOjmXY0upHkUX0uoFbuV6wWpbFB1N5dMOkO+ohU9LY0xlu/nvPbm0h5ldp8atLCmC3IvFLB7qYyJTZ3tDnbm8PcGW2gCwJ7O+O8MKO1iUb546OINvbAzdtprmyxqXZAn1xL7phMZGpsLMjeFsLpfF0me++P026WGdnZ5Cjj3hXGuALBzr4P1+xx1R++dDKtIwODg4PPvt7Ioyly2DB431Rarp5R+/9w4MxDNMi4lVuGh7sjXpJBFXqurnEyq5c1/mLd0Yp1Qye2RLnt57uX9JJ/tz+TlJFW+ZxI7mKxqWZAu0hFVkSMEwLtSEf3N4eJOCWmc5VKdf15mnlW0MpTk9k2dMZXvZkURIFPr6rlUszBfZ3r05vvdZr5nLcb5mHg4PDOjCSKvG9M1P4VJkvPba81/OteG88S02zI2O//GQv07lqc+gP4Px0nh+fnwXghR2tbEr4+caJcXJljU/ubmdTws/B3uU3sbaQe8lQTSLg5itHB267nuW6Et89M0WyUOPtayAioJu2Zd7ttHXTuSqGadEWcrOrM7RkiOZR5fXL6ebtn17J8JXn7t9aHBwc1o/BFj//4wtbmc5V+Ktj44A9L/LMDRrpxYQ8Ci/tWRr7XdMNvnFigkypzsd3tt0U7pQta5Qa6amT2SoHe5fu2b0x302zMwt89317H1cVkc8f6GKuUGNzq7+51m+dmsRo2LQuyDVOjWXQDItTY5lbyvRu5V19O354boYL03kO9EZu0pGvBScuzMHhIeDybAHNsMiWNaay1VU9d1dHCFEQ2NEeJOZX2dUZWuKwISw6sBMEmM1XmS/W0U2Ly7Mrk3uAvVmemciSLdfv/OAbWNirF69FvMPutbszxGDCz+ZWfzNg5lFHlu13UIB1iQZ2cHB4sLhxv14tyYIt5zNMi0vL7O/tITf7e8J0R708ObC64ffF+3jcr7Ip4efSTIH5hgRPaD7u+sJ3dtrXp12d67eHm6bFxRnb1ejCdH5dXtPpTDtsWBxrxOvs6gwxnqkQcMt0Rz13fsIijmyO8/Sm2C11atvbA0tua4ZFd9RLtrz8Ud6t+O77U0znqnhdEr97dGBVx2uf2tvBldkCvTEfumExmb0uHbkVHpe0YlumR4XffqqXK7MFBEHgNw+vw+Sqg4PDA0VbyM1n9nWQr+rs6lh9AdoWdNMX95Iq1NnbdfP+LgjCmpNjF/bxnqgPURT4/nvTjKXLqIrI7x4d4JcOdjGdq7Jz0bqf35rguS0tq9JR3wlRFDjYG+H8lN2ZXg+cYtrhvuEUw+tHR9jD7xzpX/Pzb7dRCYLAjkWbm0sW+PwqrYvg+sCbblpY2Jrmnw+l6Ip4OLr59sdsQbeyxALwRj9Wh5UR9Kgc6rPfx4BneY9yBweHjc1KA12WQ5ZEPrd/dfv7+ak8p8ezbG8P3Dac5cZ9fOGaYBgWpmXREfY0480Xs56F9AJHN7fc8bqzGpxi2sHBYU2kijUmMxW2tAbwuJZ6kaZLdcbTZTa3+pv67Zf2dHBh2g5ZkUSBt6/OM5OrMpOrsrMjtKx7h8P6U6zqt520d3BweHQwTYsLM3n8qnxLrfOdeONKkkrdYK5QZW9XeMWnjp/c1c65qRw9US/5is5UtsDWtsCyQV4POo5m2sHBYdVohsk3Tozzk4tzfP/s9JL79EX3fe/96/dFfS6e3hRvdh4WEriiPtcjF+t9v7g0k2d0vszIfJkrs8U7P8HBweGh5thImn/4YJZvnZpkKntrT+rb0dPYy7sj3lXJ90Jehac3xWkJqM1rxo8+uNmWdSPgXMEcHBxWjWlZmKbtWazf4FdsYUfZAmjmrb2MDw/G2NEexKtKt7Vvclg/fKrclMh41Y3X/XFwcFhfdMNa9vZq+OSuNp4ajBFclKK4Gizr+jWjrm9M/3unmHZwcFg1qizx2f12YuLOG4YQFUnks/s7GU2VlmitlyPkXdvm67A2PruvE820kESBT91gieXg4PDo8cRAFJcs4lMlemLeNb2GIAiEvWuX6XlcEp/b38l4usyuZYYeNwJOMe3g4LAmuiJeuiLLb76dYQ+dywySONxfZFnkVx7vud/LcHBweEBQJJHH+1dncfdh0B31NqV/GxHBstbW1r8fxONxq6+v734v475gNBwQ5HVK63FYf0ZGRnhUP58ODzaLP5t64zjV2UscHgTWc990rpMO683Jkycty7LuqEPcUJ3pvr4+Tpw4cb+Xcc+Zylb45okJTMvi5T3tbG4N3PlJDh8almWHowTcMvIire+hQ4ceyc+nw4PPwmdzOFXir0+MgwBfOtSz5mNdB4f14m73zapmUDdMClWdvz4xgYV9ndyUcK6TDnePIAinVvK4DVVMP6qkS3XMxglCslhziun7zNeP21PHAy0+/qePb1vV9LKDw/3kwnSOk2MZBGB/d/ieF9P5qsZrl5L4VYlntySQnN8dh7sgV9b4r8fGqOkGlmVxcaZAd8RLqlhn09pyRRwc1oRTTG8AtrUFmCtUqesWB25jiO5wb3jjSpJcReO9sSz5qnZXgxcODveSkMdFi18FwL/Gyfu74eRIhqtztiVfT9TrdA8d7opksUZVM6jUDSazZVRZQjNM9nWH7/fSHB4xnGJ6AyBLIh/Z1nq/l+HQ4In+GG9cTtIb863ZCsjB4X6wvydMqlgDuC8FRyJoF/IuWSTqU+/5v+/wcNEX87K9PUimVMPvljFMi0N9kQ0Z+uGwsXGKaQeHVfLlJ3v5xK42Il6XI/G4BXcbFT/yL15ap5U4LMbrkvnMvs779u/v7AjRFnSjKpIT1ONw18iSyCd2tQG2djpf0UgE3fd5VQ6PIs5u5uCwSkRRoNXZsB0c1kTM73SkHdYftyI5HWmH+4YTO+bg4ODg4ODg4OCwRpxi2sHhDnwwlePvz04zm68CYJoWs/nq/8/eewdHkp9pek9mlvcFVBW8R3vvx3uaGQ7JIZdD7/b2dlcMrS5OoQidQlLoQiFF6EJxp5BOp9PFabW63ePekksO/XA5wxnODDm2e9p3ox0a3lWhvM1Kqz8SqG5Mo7uBbrSdfCImGhigKrOAQub3+37v977UNf0On5mNzeopVFUKVfWOHLuqaLx6eo53h9PcSxkHNneeuqaTLMoret8Uaiq/OTXLh2PZ23BmNja2zMPmHmciUyVbVdjSHsIprf3asKpo/HYoiWlCrqrwjQM9vDqU5MxskeaAi28e6LF10zb3DGPpMn/5hzEA/vzx/tueOPbBaJbTM0UAEiEPg4nAbT2+zb1DVdE4M1uiM+olFnDzg4OTZCsKm9tDfGpL6zUf+/aFNOeTJaBEZ9RHa9iW5dncWuxi2uaeJV2u85OjU1ahW1F4cuO1jUVN00Q3zCVBK9fDKYkE3A5KskbTggXecKrEhWSJTMWFoht4RFunZ3Nv8MFolvNJq5g9NJa97cW0zykxMl/G7ZQIeezbj83V+YeTc0xkq7gcIt/Y3022ogCQWtgh1BeSPJfzKo/6LJcll0PE7156fR5LVxiaLbK5LURvzH8rX4LNxwj7amZzz3L5bp9hmsiqzi+OzVBVNJ7b3kYieKkbUVN0fnhogkJN47ltrSsKvpnO1zg2kWdfbxMRn5POqG/hWFYk8+LF3MbmXiHkcTCdryEIAiHv7V8EVlWdeNCNQxJJFmVeP5tC1Q2e395Ok9/2a7e5xGJQmWmaeFwST2yI88vjM6i6wb9/6yJn5or0Ngf4+oFu4sGlQ60PDjTT1eQj5HUS/Ih96T+cmkNWdUbTFf7zJwdv2+uxub+xi2mbe5Z40M3ndrSTrShs74xwcb7MdL4GwOmZIokNl4rpVEkmt6ATvZAqr6iY/u3pOXJVleFUme89MdDogPTF/OiGScjrvCXSEhubW0WqVCficyEA80Xlth8/6nMR9DiRRIFsVWGuYHUZz8wWeXgwdtvPx+bu5dltbQzNFOmMevE4JQYSAYIeJ/mqwnsjGVySiGaYTGQrVxTTgiBcddcl6nMyW9CJ2mFbNmuIXUzb3NP0xwP0x62PO6Jegh4HsqozGF+qxWyPeOmN+chVVLZ3hlf03FG/i1xVJeR14LhsK/HZra1M52vEg247DtnmnmJnV5S3zqcRgO13ILRlZ1eEeNCN1ykhCnBuroSqm/TH7e12m6UE3A729zVd+tzloKvJR03R6Gv2k6sqdER8K2qMXM4XdncwV5BtHbXNmmIX0zZ3HaZpUqxpBDyOVRWrIY+TP3mkD8O8UkfnlES+sKtzVefxmW1tzORl4kH3kiFDhyTS02zf/G3uPXZ0Rfjnn92MeI3O3a2mI+JtfPynj/ZjmthDvDbXRRQFvrSnE003KNc1CjWVqN+16hRat0Oyr982a84dL6YFQfABPwL8QAH4smma9Tt7Vja3k0JN5fin4UyYAAAgAElEQVRknvaIl8FEgNfOpDg1XaAj4uXFvZ0IwspvtIIgIK3RfdkhiXQ3X1lwVOoa55IluqK+K7YXbWzuZiayVf7tG8MIgsB/8fQgHZE7U1AvIggCq/jztrnLyVYUxjIV1i1IMtbqOY9O5lA1g22dEToiXnwuBz87Ok2uqvLQQDMH+pvX5Fg2NjfK3SD4/DTwgWmaTwAHFz63+Rjx+pkkh8dzvHxilpKsMp6pANYAoHYXDvm9fHKWt87N86PDk6i6cadPx8Zmxfzh/DxjmSqj6QrvDGfu9OnY3EcYhsmPPpzkrXPz/PzYzJo8p2laz/nX74zxw0OT/OzoNKZpUpLVxgzMeLa6JseysbkZ7oZi+iKw2N6LAPYV/mPGYgSsQxJwiCKPrY/TEvLw6LrYXTngJ6sa45kKpZqKnTthcy+xuT2EyyHgdohsag3d6dOxucdQNMNqclyliaAvXBCNNbowmqb1nJIoUFN1RAGmcjWKssbuniitYQ8P2l1pm7uAOy7zAC4ABwRBOA2kgH92+RcFQfgz4M8Auru7b//Z2dxyntnUQm+zn5aQG69LYn1LkPWrHCq5nQzNlDg5VVhxGpeNzd2Cz+WgLxZAALyuu2+hanNrqWs6s3lr+G6xibEafnx4imRRprvJxx/tWTqDIooCX9rdyUi6wsbWtbl+Lz5nuiQzX1JIlxV+eGgSSRT4xOYWHl8fX5Pj2NjcLHfD1fQ7wCumaW4BXga+efkXTdP896Zp7jVNc288bv/h3I+4HCKb20M0B9ZWf2yaJqnSlbHfdU0nVZJX9PjXhpL8zXtjjKUrjf9/dq5IUVaZzNWoKNqanrONza2kqmjIqo6s6lQV/foPuAXkKgqVuv13cyf4+dEZfnp0mh8dnlr1Y03TJFO2xpnS5eXHmkbTFc7NlZjK1W7qPC9nMldlPFsjXa6TrdS5kCxxcDTDkYkc86U6J6byyOqdeS/b2CxyN3SmBSC78HEaWJlvmc3HBtM0qSg6fpe0qmHEN86lOD5ZIOpz8s0HenBIIopm8P33JyjWVHb3RK/Z2chWFE5OFwA4OJptpGW5JJG6puN1SrjuQhnK/UDvf/PyTT1+7F98Zo3O5P5iKlvl9+fnEQR4YkOc7Z231x5vaKbIK6fncDlEvra/2w5quc3ka5a3eKGqYJrmqoe7P7W1lTOzRbZ1XHmb1g2Tdy9aKs13L6bZusz3XA/DMHl1aI50WeHpTQnawl5yFZXeZj/TuRpuh8h7IxlEBA6OZinWVFTdZGS+wgu7OlZ8nLqmY5rcUHfexmY57oZi+j8BPxQE4VuACnzlDp+PzV3GL47PMDJvbR0+u61txY+bXQiEyFVVZM0gIIlUFY1izRpcmStcu3sS8jqJBd2kS3UGEpeslMJeJy0hL16XhD1/aHMv8f5olspCR/rgaI5nt7Xf1uPPFa2/OUUzyJTrdjF9m3l2axunpgtsaA2uqpBe5FoSPEkU6Iv5GU1X6I8Flv2e6zFTqHFmtgTAobEcn9vh5cGBZgzTJLTZyVvnUgTdTmqq1cxYnE9fzSB4qiTzow+nME2TL+zuXGLVaGNzo9zxYto0zTzwqTt9Hja3jg/HsuSrKg8ONON3r+4tZ5omY2lrWns0U7nOdy/l8fVxPhjJ0hvzEVg4bsTn4uHBGJPZKg8OWIMrmm5wZCKP2yGyvTPcuMk4JZFv7O+mrhl4XZc6GF/Y3clrQ0k2tAQJee/4n5CNzYrZ0xPht0NzAOzqvv2bgFvbw7w/kiHsddEXs71+bzddTb5b6i/++Z3t1FQdn+vGrouxgJuQ10lJVulfeH/43Q4+uaUVsLIETKwGyZMb4mxsDTGTr7H1GkFc+arCB6NZ2sNetnWGmc7VUDSr+J7IVO1i2mZNsCsBm1vKZLbKHy6kATCBT2xuWdXjBUHgkXUxhmYK7LhOYptumEvCWjqjPjr3XHnj2N/XxP6+JoZminwwmkEARhcKdp9LWpKoJYrCkkIa4LltbTy+Po7XuTrZiY3NnaYt7F3YohdovwMe04fHc0xma8xJdUYzFdYl7t5B448LH71u3gyCINxwIQ2W7OI7D/ag6mbjulvXdN48N4+umzy1KcHz29v57VCSM7MlWsNeHrpODP2b5+YZTVcYminSEfWysTXEaLqCZphs7bAdbWzWBruYtrml+N1WFLdmmIQ8N/Z229MTZU9P9KpfNwyTl45MMZ2v8ei6+DW/dxHdMPntUBLDNCnUVMJeK2DA5ViZBnq1HXYbm7uBsNfF1g5rURryrk2oxmqYztfIVhRrUKai3Pbj2yzl9TNJTkwV2NIeanR/7zQOScRxWf/i9EyRXxybJl1WmMpXeWJDovE15woSuhav7W6niMcp4nVJfHH36tJwbWyuh10R2NxSmvwuvvFAD2VZWzZNcC0oyVpjevzcXIl1LQGqdZ3WsOeqj5FEgVjQRapYZ1dXhI1tIdwOOybc5v5m0WdaEAQG4jema70ZHuhv5kKyhN/tYPAOHN9mKWfnSo1/75Zi+qPE/C4yZWvhlSkrrG8JIu0UME2TwRXsbDy+Pk5vzE+T34XXKTGTrxH2Ou2GiM2aYr+bbG45TX7XLR00CnkdbGoLMZmtsq7Fz9+8O0ZdNdjSEeKJDYmrTmy/uKeLbEUhHnSvapszVZQ5OpmnP+ZfIgmxsbnbUXWDmbyMKAj0NPlw3GY3GtMESRSRBAHbof3Os7cnyvGp/A05b9wKLiRLjKQr7OqKkAhZzZDuZj/feaiXD8ezPLbeknSsZiEoLgxGArx1fp4j4zl8LonvPNRru3nYrBl2MW2zamqKjlMSltyIK3WNqWyN1oinsa12uxAEgU9vtboqY+kKqm5yIVVmNFPhjXMpepr8PLY+fsUNw+UQr9m9vhqvDCVJl+qcnS3x500++4Jsc89wYqrAHy7MIwBBj+O6cwhrzYdjWS7OlxEFgdF0mdgaeMtrusFsQSYedNt/i6vkQH8zB1aRIPiHC/Ocmi6yqzvCA2ucPCirOr8+OYdhmqRKdb71QE/ja+eTJc7OllA0k729Tbgdy/+eZVVHEoVGcm5JVinJGibQHvaQLln+2FXF8lm33y82a4VdTN9GhlNlTk7n2dgaYlPbvTn4cHIqz2tnkoS8Lr5xoBuPUyJdrvOvXj3HyHyFze0h/ukz6297Qb1IT7OPfb1NzORrBD0OhmaKxAMejk6uXfcl5HGQLtXxuyUcazS4Y2NzOxhNl/nJQmDH5o7wbS+mWyMeYgEXTkkk4lmb3ap/ODXHcKpMxOfkOw/2Itp/k7cE0zQ5PJ7DNOHIRG7Ni2mHKOB3S5Rkbcl8jWGYnJwuoBkmI+kysmpcUUzrhslousLLJ2bxOEW+ur+bkqzyg4OTHJvMsy4R4BObW3hsfZz3RzK0Rzx3tS3jqekCF1IldnVFG/kGNnc3djG9RhSqKgGP45pygdfPJKkqOpPZGhtv0OfzTpEsylxMlfm7gxNkKgobW4Oky3U6oz7mS3WKNSvRLFdRKMnqssW0YZgcm8pjmrCrK3JLbnqL7h89zT6OTuSIBdwousHmNVy8PLetjYlslZaQ57Zvk9vY3AyHx3PIC7Zgx8bzvLBz5UEXa8H+nibeHU4TC7gZSKyNZnpxkLFY01ANA7dodxtvBYIgsKU9zOmZAlva114W4pCsIngxrnwRURT45OZW3jiX4kB/0xX3lneG0xwczTa8p6uKTrIoU5I1aqqOqhuU6xqZisJDQTef3dGOaZocncihGya7uqOrdjOZK8jMl+psbAs2uuDLUalrSKKwqg64qhu8diaJaUKuovKPHulb1bnZ3BnsYnoNeONsimOTeVrDHr6yt+uqRWIi5GYsXSURdN9ThXSlrvGjDyeZXojPdjtEAh4H7WHLn3MwEWBDa4DiqMoj62J0Rq0L4WS2imaYDb3a0GyRt87NA1YX4lZ2xS73U11t0tf1cEriHRnesrG5WRqeugK0R1YvcbpZfnp0mvPJMueTZR4aaGZf3813N5/Z3MKR8RyDicBVt/8vJ1mUCXocN2Xhdq8ynCrz+/MpHhqIsfEGGgyf2NzCM5sSt+z+FXA7CCxzbf3yvi5e3NvZOG6hqnJoLEtr2MOZ2aL1TaZJLOgi7HXR2+xHVnX6Yn78bpGAy8ngZYu3s3Ml3ly4FwmCsCIHqEWKssrffziJbpjMFGp86iqDmxfny/zq+CxOh8BX96087dMhCjQHrLCwROjmZVA2t4eP39XkFjC+ECYyV5CvCPi4nM9ubyddVmgO3L3bS8uhmya6AVG/CwPY3Bbi+e1tiKLA4fEcp6cL/P5CGkkQOJ8sA5Z2+adHpwHY1xtlLFOlWFOZL1kdg329K7943Sz30sLFxuZWsqUtzK5uaxG7qf32S81MLi1u12oA0SkJeJ0SbodIrqLw9nCaZr9rWf/h9y5meH8kg9cl8a0Hej52jg7/6tVzZCsKbw9n+H+/s3fV18aDo1lOTRfY1R1hV/ftu4bD0uv4m+dTjMxXODldYENrgF8en6U97OEzW9v5YDTD78/PM5GtUqipuB0OKorOf3xvjJ3dUT69tXWJBepqpXqGYWKY1rtX06/+Lp7MVjFMk7pqkizKKy6mBUHgK3svDcfb3Bt8vK4kt4hH1sU5OJplIO6/aiEN1jbWjQy83WlCHief29nObKHGlrYQuglRn5NCTeX35+dJl+sUF7yaz8wW+b/fvMhgwk+uqjAybw0Cdi10xMp1jbpm8H+8foEzcyW+/WBPo5u01h1kGxubpezuifD2cBBJEtnVdXuLIYAtbUF+fHiKoEe64cjpj/LKqTnSZYWh2SI9zT5G5isMAz0x/xXpdqmSDFhD1CVZu6KYni/J/OzoDIOJAE9uTHC/sVg33ojCzjRN3ruYwTBN3hvJ3LZiOlWSmcxW2dAaaiTZKprBaLpMW9iLS5JYlwggCAJ/+8E4F1Iloj4XqmESD7iZzFbxuSQuzleQRBGPU+JzO9r5/M52XjuT5I1zKaqK3kjEBUiX6/zs6DQOUeALuzuXSEsiPhef3dFOsiiz8xq7q7u6osyX6nhd0qp3Mm90ON7mzmEX02vAYCKwZAvpfkLTDaQFa6HeZh8/PDTJbEFmU1uQpza24HdLTGZ1NrYG6WryoWoGsqpTqmmkS3XSZQVF03GKAoOJAM1+Ny+fnEEUBN4bTrMuEWA0XcEpiWQrComgmz/a03lNHZqNjc2N8fNj0/z02DQCAgNxP1/Z131bj//+aJaIzypMjkzk1sTbOOBxkC4r+FwSLUEPI/MV3E5x2ZCoRxa61YmgZ9li5d++eZGhmSKvn03SH/ffd77zz29v55fHZ/jE5pZVNy4EQWAg4edCsnzV5MrxTIULyTJbO8JrUgxqusGPD09RVw2GU2W6m/ycnSsylq4Q8boo1zXGMhUOjmbxuiRqis5cUaYl5OF7jw9wZCKHSxIp1zV6mn00+Z0EFwryjqiXSl0HLAni3t4oDtHyYL+QLFOSrTmgkfnyFQuHgXjgugVy2Ofkxb1dN/0zsLk3sIvp+5CaomNirkoTeD5ZoiSrbO+MNArZC8kS/3BqjqDHwVf3deOQBOaKVmdncbo6HnDzbjVDra6xuyeK2yGRKdfZ2hFmLFMhU1EIe518ZV8Xe3qa+P4H4zQH3JRlFa9LYmS+QqGmcnauyEA8wGxBJlNW7FW5jc0t4PB4jlpdW/g4e8uL6alclV8cn8HnlHhxbxdPbEhwcqZAyONk7xpJvT6zrZ3xTIXWsIegx0lf3I/f7Wh0MS+nOeDm89cYuvQsbP9LgoDrPlzQv3YmSaai8PrZFC/s6lh1Qb2xNYSqGaxvubKQNAyTXx6fQdVNJrLVNRmcs2RB1seabvL+SAaAZLFOf9wqrOuqjs/loKvZy4mpAhGfE5dDoD3qpSCrjMxbMszntrZRVTW2LgxPnpgscHQih6IbPNjfzL/53TAdUS9f2t3JQMLPyek8onDJo9rG5lrYxfQ9jr6g31osgFNFmR8dnkI3TL6wq6MxhHctpnJVfnV8holsldZwkj99tJ/mgJvhVBndMJkv1RmaLbKnJ8pj6+NcSJYYTVsdiGxFoVrXEASBZEHmnz27qfG8fxrppy9mSV/29TYhCAJb28M8PBDDwOR7jw9yPlni4GiW7Z0RRAFaw567XiemGyazhRqxgO1ra3NvsaUjxCunkyDQKCpuJReSZeqqQV01mMxV2d0T5V9+aQdup7RssXsjuBzikvCkltCNL8T/4ql1/HZojv54gLaPSETuBkbTFY6M51jXEmB755USg0pdY2S+QneTj7BvGUelhcrUuAHBumma/ObULKpu+UD/+eMDS74uCOBzOSjULGerG6Uoq4ynq/TGfAQ9Tr60p5Mj4zl2dEU4OJplNF3h87va6W7yNRINS3UVUfDx+Lo4x6fyNPtdvHxilic3xKlrBomgm+l8leNTBc7Mlnh2ayu/OjFDXTMIeZzMFqxu9nSuRlnRSAQ9/NljA9c/2RskU67jcUofO83+/Yz9m8Qyds+UFbqbfPeUR2m+qvDDQ5MkizIP9jfz1KYWZgoyyoL11VSutqJiWhIFSrLGbMFKRnv3YpqWkAdBAI9TZDRd4c1zKXTDZH9fE7u7o7x0eIqJbJX1LQH6Y37eOp9irijz82PTfGZbGw5JpFBTOTqZR9UNNMPkgf5m9vRE2dwWwuUQkUSBQk1lV3eEfb1R/O474029Wl45Pce5uRIRn5NvP9i7alslG5s7xVS2etn1oXrLj7exLchwqozXJdHd5OP0TIEffziFzy3xJ4/0E/I4MEyWDITdKKemC+SqCnt7mq45u3It/G4HL+zqvOlzuVW8cTZFoaYymauysTV0xc/t58dmSBZlAm4H//jRvis6z998oIeDo1l2dEZuSOYR8bmYL9WXHaYTBIFPbrGcVR4auHGXlh9/OEWhptIccPHtB3t592Kanx6Z5pXTc/zz5zfxxIY4EZ91/Pawl3/+i1ME3Q4upsq8sLOd4VSJc8kSbqdEIuRh54JE4z+8MwpYbi5gLbqm8zXCPief3tLK8HyZ3mZ/QwZyPUzT5MRUAUU32NUVWbFN6ompPK+fSeFyiHzzQM+yi55FZFVntiDTHvGsyKnG5s7xsS+mZVXnbz+YoKbobO0I84nNLXf6lFbMVK5GsihzeqZItqLgdzvY3RNlPFNBN0y2dV7qPI3Ol/m/3hzG53LwF08Nkghe6t60hb18aW8nf//hJEG3kz+cTzOerdIV9fKpLW3IqnXzXZR4ANZAYl4mEXLz2pkkPpeDmbzM6ekiu7qidDf7mMxWOTdXolzXmMrVuJiq8NX9XY3O0WtDSf7qnVFiATeVus5ntret+LWnSjJvnp2nye/iqY2J27oISpetFK1CTUXVDSTb19bmHuH35+cxFj5+69w8/+1nbu3x2sJe/vSx/sbn71/MMDRbRBBgZ1eW4VQF3TB4YVdHw1LzRpjJ1/jtUBIAWTWueh3PlOu8PZwmHnTz0MCVbh/nkyVeO5OkNeTh8zs77rqFcmvYQ6GmEgu4cUpXnpuqW79dRTcwTatbfDl7eprY09N0w8d/cW8nyUJ9WRmeaZq8fGKWqqJTqmt840DPMs9wfRqvYWHRd2Q8T10zmMnX+H/eHsMpiTy2PsaeniZ+eXyGdy9mKNZU3E6RsXSFsM+JohmcmS0ylavSvrDDcLlRQNTv4tsP9fLZHe10RLyIosATHxk4/f35eU5OF9jRGeGRdVe+Vy6kyvzubKrx+b7elf1ck8V64/Vlq8o1i+kfH55ivlSnLezhq/tv73yDzer42BfTdc2gplhDCPmqcofP5ur84cI85+ZKHOhrbhTJg4kArSEPo+kKiZAHw7RiUte3BBmIB5Z0LX5+bIbJbA2AD0Ysf06vU2JrRxhNN2gNefivP7mRZFnmP7w9hmmaJIt1EiEX+91NzJfqtITcnJsrsaE1iCgIdDV5EQTLlqrJ7yJfU2nyOxvemG1hL/0xP5O5Ks1+F4ZpUpLVRjE9NFtcOI686unyQ6M5pvM1pvM1NiwMP94unt7UwuHxHP0xvy3zsLmn8F42R3En3rttES9RvwuXJFBXrWFlgPFMdUXF9Ey+xqun54j6XTy3ra0hb3M7RERBwDBNvNd4XW8Pp7mYKjMyX6E/FriiKDw5VaCuGoxnqmTKdRKrlIzIqs7ZuRLtYc+qH7sSnlgfJ+RxsLE1tGxn+fntbZyZLdEf9y/bYDAMk1xVIeJz3dBCwe2Q6G5e/vdkmpaNKlhSuBvlhV0dXEiWWd9q6bI/taWVdFkm7HGhaAYOUeBiqsKenibOJUuYi8cyTDwuCdO0gl5EQeAnR6bpjHppj/joavISD7Y1nDkCV9HVL3JsMk+hqvLzY9NsagvSHLgkP5wryPzq+AznkiXWJa4d3PJR9vc2UVU0Ql4nPde5bxVqKgD5hX9t7l4+NsX0VK5KpqywuT205I0f9jr5xOYWpnLVFa8swdKuDafKbO8M35RGbyXUNZ0Px3IAfDCaaRTTHqfEnz0+wBMbE5RkjQ0tAf7m/XHqqsG6lgDPb29vPMfGtiBHJnI4JIFMuc4PDk7gkEQeHWzm8ESesNfJY+vjPLetjcc3xBnLVqjIGj89Os1/+cwGZgo1fnNqDs0waB/zMl+uNwYLD/Q1UagpfH5XB3t7oo2LfG/Mz7ce7GU8U0EzDJoD7iUT0Ht7o5RljUTIzTOr3BHoavJyPlki4Hbcdt/ujoj3CsstG5t7gSfXxzkzZ3nBP70pftuPv6s7wtsX0kR9Th5bF1tIqDPZskLP62OTeXJVlVxVZTpXozfm5+hEjvdHsnREPWzvjDB4DZeFUk3j9TMpogEn332o94qvb24PMZOv0Rq+sbjpV07PMTJfwSkJ/Mkj/TcsN7kaL5+cZSpX48xsiT95pO+Kgrk54OaRdVefOfnVyVkupsp0Nfn40p61lbOIosAf7e7k4nyZTa3L/z5rik6qJNMR8V5VFtES8iy5p/Y0+9jQEuR8sky+phLyOnhqY4Lfn5+nt9mL3+3A4xTpjVnhYZ/f2cl7IxnePJdkOl/j1dNJvri7g//xl0MUaxqf29lGqqTgdUq8sKv9qoP6WztC/NXbY8QCLn5xfIY/fvjSQOXxqTyCINAS8rCzO8KOzpXPH4R9zmsOwV7OZ7a1cXauyOa26z//786m+MWxaTa1hfjuw70rkoVousHQbJGI13XVRZLNyrhvi2nDMPlgNItumKxvDfDS4WkM0yRdrvP0pqWF29aOMFs7rv9mPTNb5NBYlv5YgKMTOTTDZDpX5bsP3/zU8rVwOyQ0w+D0dHHZonP9wvBNpa6haiZ1TSdTXtpl72v2sb+viSc3xHn3YgbNMKlrGsenCmQrCjVFZzpnda4fHozxu7MphlNla+BlMke+onJwNEO+phLxWttofXE/F5JRZgs1xjM1JrMy6xPBJdtWqm5wdq6E2yny2PqlyVkPDcSW3WpdCds7I/TG/Lgdoq0ls7FZIUcm842PD40XbvvxXzk9x8X5MqIocD5VXtY67FrDWYOJABeSZYIeR6PgOjqRR1Z1JrM1nt/efk3J10i6jCRCTTGYK8pEP1Iwd0S9bOkI0RLyrFgDezmLHVnDvDTsdznHJ/O8PZymt9nPc9taV61bXuzky6qObpqIrO7xi9f4mXztlvj6f7QQvhzDMPnBoQnyVZX+uH/FBeVwqoxmQKaiWBINn4ufH5uxikCfs+EclS4r7HQ5GEwE2N4ZplhTwTQZmi3ywUsZzs6WaA64+fXJOdojHtIlhfaIhyc2XOknbhgmm9tCPLouRlXRr+g8D8QDnJsr0dPs58H+5luWj9Ab89O7AjeRuqbz0uEp0uU6hZrKc9vaGru1umHy2yHLi/3pTQnawpcaQe9czHBkPIcgwNcPdC+Rf16Lk1MF0uU6+/qa1myQ+F7nvv0pDM0WGzY6smZZxQFoN7H99OpQksNjWQQhye6FoYaAx8lwqkyqKLOzO7JklbtWFytra0tka0eIumZQlFUKVZXOqJfXzqQ4N1dkb28TD/Q3s68vyo8+nAIsf8z+eIDZfI3/5TdnqdZ1Tk4X+K8+sZ6T0wU8TpGoz+pidEZ99Cf8fDiWZVtnmGc2JRiZL6PpJienCsiqRleTj9JMEZdDZK4gU6pr/Pjw1DVt7BZ11nXVILdgk7dWhDz3xsCijc3dwlT+0tDhdK5y248vLVwPBSw972yhhqqZja7YkfEc/+ngBAG3g3/6zLrGoNki61uC9MX8SILQKJo3t4d4fyTDQDxwXemKUxLI11T8LnNZH+qfHZ3m7QtpAm4H//3zm1fdnf7kllZOThXoiHiXXQycmC6gaAbnkyUeWx8juMpr2LPb2jg5XWAgFrghL/4nN8Y5MVlgc/vyMpHrkSzKXEiW2dAaXLXrkm6aFGuWLWOusnJJ5Zb2EGOZCru7I0zmalQVnapiPY9pmmSqCtO5Gi6HwHsXM/THA7yws4NvPNDDobEsqfPzjGWqVFWdsGHy3PZWvv/uBKmSTLIo45JE9vY2LZFF/tU7o5ydLTGQ8PPIuhhBtwNFMxrfM5gI8L0nBpa8D28Vb19Ic3wqz86uCA9flup5YsrSku/qiuAQRXqafeRrCi3hpQuamby1kwHw4ViOz+64VEwbC/WQaYJhsCJSRZnXzljzCTVV57ltK591up+5b4vpy1dL7WEvfTE/6VKdHddILLoeDlFYsNJxsL+viXjQTcDj4AcHJzEMS2P8hd3Wavu9ixkOjmZZ3xLgU1taKSvaqoo/3TA5NV3A73YwEPfT1WQN9LWHPXx/QcqxuzvK8ckc55Jljk/l6W32E3A76F5YkaZKddrCXn50eIqZvAymtQ04ma3RFwtwZraIx+mgp9nHQCLAG2dTNPvd5KsqA4kAu3uifDiWYyxdweOU6Ih4GYgHkFWd1pCH8UwVj1MkEXTRHw+SCLqvGKY40NdETVrSCEsAACAASURBVNEJ+5yN87rXSRVljkzk6Y/7G7sCNjb3AhVZb3xcXgiluJ18flcHLodIyOuk2e/mh4cmMU345JYWtrSHeev8PJNZq+Afmi0uu3P10SLygf5mDvQ1rag4XN8SJFdV8bkcyxZB4xkrgrpc1yjJ6qqL6elcjdMzBYqy2pgpuZyt7SH+cCFNb8x/Qx29WMDNk8t0UlfKxtYQG68iwbgepmnykyPTyKrOuaQlM1kNTknk2W2tDKfKq7oPJ0Ie/vjhPrIVhb9+d4yZfBUEgSc2xJkryMzka4iigKqbqJrBcLLMsck8e3qiyO1hfnBwgnxFIRFw88hgM8lCnYjfiaIbzBZkfnN6jrpmNBIvUyWZ35+fp64Z1HUdv8tBpqLQEfXy5ct2Uhbfh9mKwmi6zEA8cMXib7VkKwovHZ7CxOSLuzuJBdyNXfCDoxke6G9GEgWGUyVeP2MNPxqGyYH+Zr73xADJokx3k3/JwiAWcONzSRRl9QrP7IcHYwQ8DiJe54qzHdxOCYcooBmmbe13GfftT6I35ufL+7rQ9Utdj9VGen6UL+xsZyxdoa7pvDOcpj8e4KmNCTDh1EyRkXSFziYv+3qbOD1TwDBNzs4VyVdV5ooyO7rCPLVxZdrgD0YzfDCSBeBLezr54q4OaqpORdF4/WyKal2jOeAi6nORKdeJBVz89MgUbREvbodIsabSHvaiaAaGabKvN8pYusrz29saBW+1rlmr1ppqRarWdQ70NbG9M4xTEpBVHZdkrbx3dkcaFxLTNBmZL3NsYVjnocH4VTXEEZ+LF3atbDvvXuGVoSTpUp1zcyW6m3z2EKLNPYNhXmo/6SttRa0hIY+zERQzNFNsBHIsFvabW4McHsviczvobVp+e7uqaDglsVHMqLpheeSHPNe9uT80GEPRrbCp1mXkCE9tTKBoOvHg1eUK1+LweI6SrDE0U+RAX9MVxdWu7iib2kK4HeItkwZci0OjWV4/m+SRwRiPrFudZl4QhIX7AriWcRJZCetbgjfcgGjyu+iMejk5naenyc9APIAkCLwzbOB2iDw82Ex7xIcoCEgivH4myc+OTmMaJp6FIfm6qjOcKhPxOhFFgZDHSkRcXFglizLzRZnOqJeDozkcosBIukzY67pqN/2lw1OU6xonpwo3LfkcTZcp1xeTFyvEAm62doZ57XSSsqLxN++N8bX93UsWlM6FwjnocS6706FoBrphIiDgcS5diLocIvt6m8hXFf7nXw0hazp/8eS6axbWYa+Trx/oJldV6bcDbRrct8U0sOZDYtMFmfaIl6GZAtmKiiRWUXWTZ7e1kizKjaCTfb1N7OyKcHAsS3/M39hiGUuvwtf1I2qUo5M5PhzLkSrJTOWqeJ0ib19I0xn1UFV0TkwXyJQV9vREOTtXYkdXhLcuzPOtB3r45OZWfnp0iq0dYS6kyjw0ELO2x8p1pvM1RuYrKJqO2ykxkrESCTXdRNFMupp8DMQC5CsKb52f59HBGCYwkAjSGwswk6/RfAODOvcyYa+TdKmO3y3Zsec29xRV5VIBXVH0a3zn2lGpa5yaLtAe8S5x3dnYGqQoW/aSi3HNQa+TrmYfHoeEw3FlwTY0U+TVoTkCbgdf29+N3+3g5ROzjMyXmSvK7O6O8vSmlqtKENrCXr52DYuxhwdjdEa9xPyuFS2SdcNK+4sFXAQ9Tja0BkgWrQG75QqbV0/P8fKJWda3BPneEwNrLhE4OpHj8HiOTW2hJZKARf7dWxcp1zVOTRd5aKAZUVzd9evFvV2MZyp3LBXwmU0tzBVkFN0gXa7zkyNTJIsyrSEPoiCyuyfCXKHOkfEcs/kamYrCXFGmrhkUZIXzqTIXU2U0w+SpjS08szHKYCLAlvYQk9kqLx2ZQtUNkvkayVKNoqyQCLvZ19vMtqvMVS1KSG9cQHqJwXiQk1MFTGBdwmr+PbkhQbGqMpKukK+qZCoKPc1+vrCrg/pV0igvJ1myXr8kCkxkqwwuEwX/zsU0J6etGYpfHJ/hzy6zs1yO5oB7ibuJzX1eTK+GmqLzi+PTyKrBZ7a3EVvmjbJouZQIedANA0UzEAVrtf2JLS1MZmvs77McQda3BhnPWMX2vt4oY5lq42vXIl2uYyyEo3hcVkpYS8jDjw9PMZwqk68qSIKAJIgomsHofIXZfA3NMDgva+SqKk5J4OycZY8ElqZwKtfE6Zkik9kqf/XOKDu7IrRHvRRllbDXQUURcEsiMb+bHxyc4FfHZ6ipOhtbQ4xkrBXyu8NpTs8UME3LvujYRJ7zyRJhr5PvPPTxCS95dmsrk9kqLSHPx+Y129wfeJ0Cat267fvW2Gniarx2JsnIfAVJFHhxTycfjGZxOUS2d4bZ2xNdMuhX143GEFRdvbJzPp6pYJpQkjXS5Tp+t4OirFKUNSayVRJBDx+OZXn2BnWch8ez/P685TbytQPd1x1u/t3ZFKemC3hdEt99qJc9PU1s74xcdZH965OzzBVlkiWZF/d2rrl93qGxLJW6zsHRbEMScDlRn4tyXSPsdS5bSNcUnamcZVO4nBNJ2OtcNnnxdhHyOvG7JbJZhROlPLKmo2gGcyVL7vF/vj5Mpa7hEAVcTontnWFMTGp1najPRXPAxcVUmZKscWgsS3PAxRd3d5CpKPz48CTDqQqqpnN0ukC1rlNXDdyixIH+poa++KN8cXcnI/OVRvF7M4R9zmW723v7mijIKh6HxFjacotZyWAiQLpUZyJbJR50s7Mruuz3DMYDuB0iJpZZgc3q+dgW06puMF+qEw+6cUoiI+mypSsGTi9s0ZVkrdHhWPxD2t0TpafJx8snZ1E0g1+fmuNbD/TwyGCct87PcyFZpiPi5cRkgdMzBWYLMk9vTPClPZ2NTseZ2SJvX0jT3ezj6Y0J8jUVhyjwhwvzHJ+0dNKf3dHWGHI0TZPWsMeawMZJLODm5FSeuWIdEWubR1NMRGFRmuGgLKtsarO0cYWaymyhhmGahLwORMEqtr++v5u5oszJqTxFWSMWcPHX747jkATKdY2OqI/WsIf1LUGOTeYJuB2NG9zIfJlsxTKfX+wufVzCS5ySSP9NSoZsbO4El/v/3owX8GpYlDOIgmVtN5qucGq6wNsX0ridIru7ozy1MYHf7eDB/mY03STqdy7rHb+pLcS7FzO0hNwNX+pPbWnlneE0o+kyJxYGta7G+bkSf/3eGD3NPv700f4rpBaLu4e5qkqhppIIXvuatphNIKs6dc3A45TQDRNJMJftOm9qC5GrKsSDnpvW1y7H+pYgRyfyDCYCyy70/4fPbubIRI4dVymIXzoy1bgvfvOBGwtduZUUayqFmuXR7HVJhDwuzBA0B914nRIXUmUU3cDnFPG5HJyZLRLyOHEIAl/Z20VV1XjjbApBsGqAxZ2F9y5mODqRpyRrdEQ9tATdqJpJ2OfgwEAT339/HNO0fLzXfUSmEgu4CbgdnJwu4MzC4bE8kijwme3ty+6Q1DUdAWFFqZ8X58vIqs6m1hDffrCXf/fmMOeTJY5O5EgE3QzPV3h6Y4ID/VdPnDw6mae7yYckCledAdjQGuJf/NF2KnWNwYQ1F3V5/WNzfe7LYrqu6bx1bh5REHhsfXzZN+1Lh6eYLcgNv83OqI+A24GiG3RGrSG/Yk1lZ3eE/X3NfDCS5vUz84S9TnqbfQiCpW9KFmWKssrovHWDAOumUaprnJ8rUayr/O0H4wynynzjgR7S5To/PjyF3yWRqdSZL1n/LUacTmSr7O6Okq1YJu3mgu56IltmZL5Ms9/FRKbCZLZGTdXxukRCbgduScQACjUF1TAYTATJV63nuJAska2oiIJAe9iHohvs7IoQ9buI+l1sagvx4ViWl45M0RvzM5OvMhAPsKs7yqe2thLyOHl4MIZmGPzi2AyKbrC5LcRAPGCFl8Tt8BIbm3sB5bKZQ2WZzu+t4JObWxiKemkPe8lVFc7NlahrBiVZ4ffn8xybyOOQBJ7d2sZousLpmQJep8T6liAhj5NzcyXeu2jNqMiqTpPfhaqbzBZqdEZ9tIQ87OmJWu5DhomiX/11LV6Lh1NlHuxvZttHisoD/U3Imk5b2EN8BdvYT21McGgsR2fUS9jr5OBolneG07SGPXx5b9cVBe0fP9zHY+vjtIU9axKh/lGe2JDgoYHYVZ875HUuawW3SGlBu166A8OpK2G+JJOt1HFIIl/f3823HuzhzXPzeBwiB0czuCQBv9uJ12mFt1xMVQh6HBiGyc+OTlPVNFwOCUkUaVtIk/zvfnqSkqwhCAKmaZIIuokH3GTKdR5bl2AwHlxYZJlM5qx7oyhajSu3Q6Sq6PzlH0ZJFmWGZouIAqxLBOmLF3k8eEmXXqlr/OjDSY5M5FjXEuSZjS1ohsF0XmZLe4j2iJe3zqco13We2BAnV1H4xbEZwFqsDcaDHBrPUZY1Ij4nhmFSlDUKVeueH/O7GUwEcEhiY6HgcUpsaAlycrrAhtZra9UXkyJlVec/vjdOua7xQH8zD95ENPxHmchUOTqZY31LsNHsu1+4L4vpE1MFTs8UAWgOuBp6vEVM02x4kzoWLnZhr5M/eaQPE2uitlhTOTVT4P2RDL86Pstkrkq+quJzSXxlXyf7e6OMp6tEfU7eOJtC1Y0FyYSToxM5XA6Jum6lK6bLSmPbL1WqAybHJ/O0hT2kinW6mnzkqyrrEgFU3WBbR5gdXZY+6/RMgX/zu2HLG1o3EARriKAoqxgGqJpOqaYhCgKCYCU6yprKwbFsw5S/p9nPoYXQlxd2dTRWm6Zpcmwyz0y+xrm5EpmyZWT/6S1tfPfh3iVdG5dDxIV4RaRpux1eYmNzz6Be1oyurYFkWtUNDo1lcTskdndHlh2q8zilxi5ba9hDe9jLkxsT/M27Y6iGSaokc3wyz7Nb25jMVjEXklznS3VCHifvj2TIVVUOj+foafJyaDRL0OPAdZmUoiXkIR60iqNraUh7Y37OzpXwuSTalrl2dUZ9q4rBbg64+fTW1sbnwykrEGeuIFOWtSvcjVwO8ZYXETdTpG/vDPPmufkVSRLvBL+/kKbJ70YQoC/mxyGJfG1/N8OpMn93cAJZM/C5JB4ejPHrk7OAtatcUzUupstkygqJoBuPS2JHZ4S3h9Oky3UErN9lvqrw6qkkgiAQDTipDs2xv7+JqM/JP5yyAnmSBZlcVWEqV2N/XzOCABdSJaZzNQTAJYmU6xpnZotcSJZ4blsb7REvx6fynJguMJquMJWvMZaukCzW2dwWYjxT4elNCY5PWg25oNuxJERF003yNYV1iQC5ikpPzMf5uRIlWWO+pPA//XKIqqLzxd2dfOehXr7//jhj6QrPbWvjmc0tPL4hvqL5Hk03+O3QHMcm84CJJAprWkz/9kySYk1lLF1l3ULhf79wXxbTLknk5HQBQYBPb73SPWNkQXOUr+nEgpe2PRa35eJBN9s6I42Lbr6qUlV0FkcMVN1kIB4k6HVwbDLPW+fnifpd9DX7+cq+Ll45NUeuqhILuIj7Xah6CUkUGMtUSJYsTXST38Xm9jCqbtDkc9EX89EUcPHNB7tpCXmZK8i4HBrDqQrzpTqiALJqYAKyoqNzyatVXBiBWGyCCKaJxykymqnwxMLr+fOFgYLLtx6HZou8eW6eqqJRljXWtwTZ0BLgmc2rDxOwsbH5+PHhWK7hOhTyOK7YAl+OsM/JDl+Ebz/cy2S+Zll79Vk37D09UdKlOhHfpajlwUSAg6NZOiJeZvIyFUVDMwxSpUtx3x6nxLcf7MEwueYcw3cf6uWBvmbiQfeyczGrRdMNRtMV4kE3EZ+L/X1NvH1hnq4mHyHv2t9ej0xYP+/1LYErwsfWgmOTeXwuiWOT+WUHGO80HREvZ+dKtH4kVGcmX2M8U8UwTEzTJF9V6Ip4EUWrU52vWsP2PqdI0GPprouyim/B3SLsc7GnJ8LxyQIzBZm6qiEvyDFOzhQ4O1diKldlKlflXLJIuqzgkkRUw+TRwRjtYQ+VusbW9jA9MR/dUR/vXLRyLk5NF3A7RCp1Hb9LoqLo+EyrU+2UBGRNJ+x10uR3L7ilGCRCbkqyhiRaLmTj2Qp/+YdRPC6Rx9bF+eSWVuqqTqZS55fHZhlJlzFMGMuUuZgq8dqZJIWqSqmusrsnumKLxxPTBYZTFZJFGUkUaKup5Bfi59eCWMBFsabS5Hfed/NGa/LXLgjCv77W103T/CdrcZyVIqs6A3E/AsKS6fVFgm4H8aCH5oCbDVfx3PzE5hZcDpETU3k03WSjHmAqWyMe9BDyOon6XWzrCFOuabx2JkmmrCAgWJZLYQ/nk2UeHozhcUps7ggT9DhIFusUZQ2/S8LncuB3STy1qZ0fHprkxFTesrEzDPrjQY5O5Lg4XwFMS2MXcFNXDeq6AaZVSIsCOEXLus4hinicArqugiDQEfHy6S2XOiaLRbRhmBimiUMSG50dn8vB4+vjdEZ9Kx5qsLGxsbl8SG21Uq8dnRH+txd3oOhGYwYhV1WYLytUVR1ZMwhIIg8PxtjRGcHvlvj/3hnF65RwO6UlnWmwtNnXc2wTBIFNK4wuXwmvnUlxZraI2ynyjx7uYzARYHANBtGuxpHxHLKqc2KqwCPrYqtOfz0xlefYZJ4t7SH29FzZfY4FXMzkZWKBu9Oh6dNbWznQ37wkcOeNsyleOjLFhtYgqWKdjW1BJrNV8rJKta4T9mnkqyrluk6Tz0XY6yRbVchMF2kJuvn6uhj9iSDbOkL81Ttj+JJl3JJAQdYYjAcYjAc4OJJF1U0UTadS11A0A0U0aAm4Wd8S5LUzKVpCbiRJ4BObW3FKAkOzRcp1jZ4mHz84NEld1elt9hP2utANA1EQ2N0TpbfZR0fUh2FanWBFt8wN3jw3T6Zc58h4nsqCPsuvO+iM+qzwM6+TRMhDk89NulInX1V5YWcHIa+lEReFhRphFTaGiz/XtrCHqN9FU8C1pgnDz29vZ65ovb/ut4bdWi2d/zPgFPD3wAysMuN0jemN+Yn6XQgI9CwzmZoIefj6ge5GmMi7F61413jQ3dAZZSsKu7oj7OyK8MNDE9RVga/u72awJdCYNt/TE+XsbJGo34nLIbGlPYRDErk4b3UqshWVf/J0b+O4Y+kypmlSqCqcT1VoCbnxuiSm81VG02WKska2UkcSBVwOCUU3yFcUVN1kXdxLS9jT2JJySCJ1zSDidSAJAm6nRE3V8XuswBqfy8HxqQItIU+jkC7JKj88NElN0fnsjnbWtQT57A5LJ7aSjpKNhazqjMxXaI/cmiEiG5t7hZ1dEQJuB26HuOzA4EdJl+u8ctqytvv0llaSJRlFM+lu8jWunYZpUpI1kkWZQDzA+xcz/Ob0HBvbgo1iIep3se46lmC3g3JdpSSr6IY1b3OrZ0e2tIf5YDTDukTwhoqcX52YYSpXYyRVYXd39IqC5gu7OkkW5Rvy2L4dCMLSIbp8VeHYZB6vU0JWdb68r5OjE3lMBBJBDwWHiqxqC51kq1D1uSWGZmUkSaA94iHodfH89nbOzhWJeF3s620iHnQR8jjZ0RlGFEWCbgndMNAME4coEPW5aIt4+NaDvfz9h5NousGp6TI11eDgSJrXz84T9Eh8YWcn7VEvNUXn5HQej8PBl/d10BryMpgIkKuqTGQrjKartEc8FKoqp6cLnJkpsK4lyPyCBauJSaGq0ZHwLpF/ALSEPfyvX9rR+LyqaDy3vY3xbJXPbW9fYtFYqKqcni3QF/MviRVfZDAR5Cv7HJiY1FWDZr97WVeXG0UShTW3LL5bWKtiug14EfgKoAE/BF4yTTO3Rs+/KlpCHv78sQFg6ZafbpjIqo7f7Wjohv/u4ARzBZl3LqRJFmWSpTqPr49RrutIgsCeniiVuiUurCgapZrK+xczvHo6idspsrMrzIP9MaJ+J1/c3UlV0djaHuLUTPEKX8qDozkEBMYyVQQBTs4UOTSatWLHETAMk3RZwcSkJeihougUZRXdgIvpGp/b2U6z340kwssnZpnO19B0k6Df2l4tyxqTmSpz5Tq16QIOSWB7R5j2qPXmnS1Y8anZisLRyRy9Mf+yXRTdMKmsMrHx48SvT84ynqnidUn840f67ivdl43NatANk6Ks4l6hTvf4ZJ5UsU6KOm+cT3FmxvLgl0SB/X1N7OgMc2KqQCzgaiSm/uToFOOZKhfnyzy7pZUv7em61iFuK26HxHy5TltIWPHP4GZ4cKCZB/pXlva4HOmysrCLaiGrOjP5Gu0RLx6nhGuFi6K7hcBl9/I/2t0BgsB8SUHRDeJBN1PZGlM5nSafg5pqEHBLXEiWUDQd0RAIuB3UVB3TtIJ8XA5reG8yV2M8neb4VAG/S+LMnKWFd0kirSE3j66P89j6BB6niEMSKNVVyyoXgX/9+jCjmQqabnJiqsj//tWdHOhrYipXRdNNXjo8zaPr4/z06DRHJ3KYwKPr4ii6gVMSmSvI+NwO0uU6L+7t4vhUngvJEhv7/n/23jM6svO88/zdWzkXUFVAIYfOORCdSDajKFKiskaysuPInrVnPbszu/buOfNhZnfCng3e3Zmz49HOeu21lSyPZGXREiWapESyE5vsnNBoZKAKldOtm/bDLRSBRuxuoAF0v79zeFiouuGtrhue+7zP8/8HKVY1/v3Pr/HhvS30dUdw2CTOj+QwTJM9bSFkWcLrtPPbj/egG+ase9P33hnhr08O0xJy0R3187tP9M5772oNezAMg5cuTuCyF3liS2zRe9zMZseHmRUJpk3TnAL+FPhTSZLagM8CFyRJ+iPTNP9yJfZxp9xej6PqBl8/MchUocrxLVH6uhtnLZdTVAZTJSRJ4rVrU+zvCKOZJh6njc5GLyVVZzxb4a2bKV69miCRq+CqSfE8vS1GWTX4818OWEoaMT+/c7wHl93KcE/X+50YSDGUKlGq6rSEXNhkGa9DZixdQTd0TEDVdWyS9cQtzxCDr2oG54YzfO5IF7+4nCBdUi0JJlliX3uY8byCZEJro4dUuUqhonFjssjNqWI9mO5s9DJWa4y5NTW/gYymG3zj5BCJvMLhnsYVr5u7NpHnZ5cmaQm5+fC+1g1ZN1VWrYcrVTPQTfPBbDwQCJbB6Vtpfnk9CVhlHku5zHZHfZwfsRQPqqpBQVGxSVJdw/+t/hRnh9K47Tae2d5EV8RH1O9iNFOhwevAtc5u2GVVpzdqfedK1VjRKfGFuJfp8b6uBnxOG70xP5Ik8Z23RxjPVtatFN5STDcgKpqO12knW1K5PJbjRiLPRK5CWdU42BliPKeQLlbxumzcmCxSrOq47Tb6uht4fHOMn16cYCRTZnPMyw/PjfOrG1PYJCsBdaw3QlnV8Lsc+N02HulqpDvi44fvjvGdt0fY1RLEZbdRUDRuJHJkylrddfD6ZJ5vnxnmo/taMUyT0WyZ/R1hzg1neOn8OLpp4nFad5AGrwOPy4bfbaei6sRDbg52NbCjNchf1+7JZwbTeBw2zg5m6Ix4iflduBy2ujX9tE27VHOBnGYyV+Frbw0ykauQKVfpjQXmHEeqblBRdQJuB986PcK3zwzXhA2kBS3scxWVr781WPPniNcNYXTD5JfXk1Q1g8e3RB+KQHtF4wBJkg5iBdLPAT8GTq/k9u+FbFllqmBpgvYni/Vg+kN7W7g6USAWcPKVV/sZTpX58L4WHDYbLrvMnrZQXQ3kf/u7K/zs4gRj2Qo2ydJ3liQYSZeRZYlMLcCN+FxkSirNQRu/uDzJYKrEpbEcfqe9rtXsssv43Q7+9NV+ZFnC5bDjMyBX1tAl0A0Nu03GNEEGKprB5fEC/8MPL+KQZQzT0mLd1RLiykSBkXQZJLNmb21HQmdHqx95xgnjdtjY3xEmX9EIe+bPOhdrXfQAA1PFFQ+m3xnOUlF1biaLTBWVesnMRuIDu1t4dzhDT9R3X26eAsF6ZeYN27aMIG9TzM/vPtnLX71xi7NDGS6P59nS5MfjtDZ0PVFA000KusZQqkxXxMeXjnVzZjBNd8THeLbCj86PEfY4+eQjbWt+/j21NcYb/VNWHat3/c/k2WWpFmhZv1W2rM76/0bEVsvGgtXcemxTpK5SpZsm2bJVbxx024kEXCQLVQJuOxG/i6e2NTE4VeLPXr9JRdMpKdZscLpYxW6T0E0YTJfIlVU6Gj18dH8bVd3g8nieGwkrW+1z2Yj6LcfM0ayCYVpNkD6nTMjjIF/W+D9evs5ETkHTTZIFhYtjOZRaRnd3q49PHGyjPezhzGCGx7dEKSk6H9jVwg/eHUVG4pGuMJM5hamiwlCqhNMuk66VgPpcNvyu93TFVd3gO2dGmMhVeP+uONviAaq6QaPPiaabbG7y8+m+9lmJrHJV56tv3SJf0Xh6exOVWsLIME10fWE9+olspSbOYGm0TwfTV8bznL5lFSZ4nTYeXYfNrCvNSjUg/gvgQ8Al4BvAf2ea5rKFKiVJ+hLw64AN+LxpmiMrMa6ZRHxO9nWEGMlUONrzntSL12mvi/z/i4/srmd7x7JlTt9Kc3k8z+5auUZLyI1dlmj0OYj4nLQ3egm67fQnigQ9DlrDbrxOB7vbQjQFXJy4meLnlydrNcl+drYFqOqWE1O2olItVEnmrYPRYZOxyTJOu4SimZgmGFjTPk67dQHUdANFNdBkExnQ7DI5RcMmWSUohmEykqnw1NYomgF72sMc6p4tC/gPHmmnP1lcMIMU8jg41N3IwFSRRzet/AmwsyXIaKZMPOSmcYPWGzf6nItqtQoEDwsHOhq4MJKlwedcdvOyyy6j6AYFRcMmS4S9TibzCpubrJroRF4h4nPS1/2enN4Ha46GPzo3hqIaTKgVxjKVOfvMlKq8fj1Jo8+5Ktev22kKuvno/rZV389KkSqp1v2npGKaJi/uaeHCaI4dLQ9Gz8xAssiPzo1jkyT8bjvxoCUGoJsmpmHyWDzIlx7t6ZJtEgAAIABJREFU5m/PjHBpLMdv//lJtjYHGM9VqGoGPpcdSQKbTcYmmbjsElP5KqmiSkHReO1qgu6oD1mCpoAl0dfodbKvPUS+ovHWzSmUqo4kSYS8luV20OtAliX6EwUKisrl8TxV1UDCUif51KF2fC47Fc3g4wfauJEosLMlyJWJPBdHc9xIFDjYGea/em4bz++Kc3OqwIn+NN88NcRUocpzO5t5YXec7TUd6alClZFMGYCLY5a+dFvYY2XwqzqPbonWHz6mSZeqdW3xW1NFPnGwDbss4XPZeHbHwve67qiPTU1+iorGgc73NNvDXgeyJNWSfhvzPn+nrFRm+p8D/cC+2n//ujaFIAGmaZp7F1qxVhbypGmaz97Njn9xZZLhdJkntkTpiix8MZckiae2NjGeqywqEzP9tPbypUkSeYXrkwWcNplbqRL72sNsjwe4lSqzpy1AuWpwfiSH12mnJ+Lno/tb2dsRZqjmqPSdt0dI5Cs47TKXxnKcGbSymZuafIxlSnzvnVGSeStbHnBbkj3lqoSEiddpNR3IkkRVM2hv8KDqlhJHrqyhGibFqs5QqsRjm6IkClUypSpOu4zHaWdHS5AXdrfMqXUKe50c7Fz84H58S5THt6zOjWhna5AdLXOnmAQCwcbjP73ez1/8agBZkgi4FzcEmUaSJNx2mYlchd6Ijx0tAQ7UbI57oj6+eKyLoNsx79TwrtYgt6ZKhL0OWsJzZ7V+eX2KaxNWxrAr4ntgm53ulvfvbObcSJYdLUEkSaKj0buhaqSXoqJZWdL2Ri/HYzGcNtnKGGcqHN8arWuI/+p6kpculKioOtmaA3HY6+DTfR3kKypv3JiiqlnyizeSBYqKCkicl7J4nHb2tIf4/JFG/tWPLnFlPE8sYJVbRH1ubpYLSJJEplTF57RhGAbP7WzmwmiWbFllKFXCIVvKX3/0gW00BTz85Ru3cNplPn+kk+NbLKMXb6rE6YEUOUWtlWiWef2aVVJ1fGuMly9PUFA0Tgyk+K3He+r31KjfSVfEy2imQlPAhaobvFxTnWkLe+YE0gDxoJs9bSESBYUjPRECbgefX0bZj8Mm85F9rZwfyfKjc2PsrKnEtIY9fPKRNmSkepnpg85KBdNzzeSXz/OATZKkl4GLwD8xTbNuJyBJ0peBLwN0ds42DEkXq5wdzADwZv/UgsG0aZr87NIEr11N4rDLtIY9fOlY16Ii5lG/k0ReIeCy89KFcSqazrdOWY6HLrvMSKbC9ngQJAh7HGxu9tMccvGdt4c5PZAm5HVQqFquSiGPnbNDWSRJIleu8oHdcYZTRVKFKtPCfdmKgWao6KaJARSqOq1BF3lFo6qbDKfL7GgJcqg7wttDaW4mChSqOi67jc1NfsI+yzymoFhGNAc6Gyz5nHWICKQFggeD/ppJiWGaXB3PLxlMa7rB2aEMPzo/jl2WGM8rvLC7pf75K1cn+c6ZEdwOG//0/VuJ36Y40BXx8Y+e2rTg9mMBF1cn8rgc8iz5NIFFqWrZRBeV9elweK9sbfLzqsvOZK7C/vYQXREfdpvE6YE08aDbavaXJPZ1hPnW6WFU3bDqrb1OYgEXe9tD/PL6FCGvk8vjObLlKsWqjmmA2ynhcdp4ZnuMZ3c088sbSQqKRqlq9VP1Rn3Eg25eOq8jyzKjmTJTRYX/+Pf9vD2YIVVUrRlk08Rht9He4KEl5GU0U2YoVaoFveMomsmxTRH+wys3GMtV6ufWf/z7G/ziSgJMeGF3ga6Ijzf7p2gNe3jpwjhfPNYNWHXknzjYzjdPDnLiZprRTKVexjOaLaPpxpwkmyxLvG/nXN1y07RmyeUleptev56kXNV57VqSAx0NnBxI8asbUzQFXfxaX8eqN+mfvpXmynieR7oalnR6XC1WqgHx1nzvS5JkAz4DzPt5jWbAaZrms5Ik/U/AR4Fvz9j2V4CvAPT19c0q3vG77UT9TpKFKt0RH9myitshz6mjOzWQ5s9ev0myUKUrYtmGK5oxJ5hOFhT8Ljtuh43ndsbZ1Roi6nfxP790mYtjOcpVnXxZBQmi/gDb4gGOb4nx2OYIkiTx/XdGeXcoy7XJAvs6wnxsfxsRn4OXLoxz4maaYlUj7LHzf758zer2nfFt7IBdsqRQZo4n4neiaDqaYTKZU3A5JByyRHPAjbeqYa/J/UhYcnou3WQ4bU1ZCVaHRF7hnaHMgmooAsHDwgf3xnnp4gROm8yLe1qWXP6tmyl+cXmCYkXD77ETD842Tjk3nK25xMKtqdKcYHopDvc00tnoxe+215uy1hMDySJv9k/RHfVxtHflnOWmeXc4w+lbabbHg/M6133/3VFG0mVuJArs75jfsXIjkyxWKSoaPpedU7cydEf95MoaboeNd4ez7G0PEwu4ONIT4Xef6OX1a0nyFQ3dMPE6bPz5rwY4NZBGkqCkaCi1+6jfJdPgd+N32nmjP0VFNTg/mqWqGQTdDqaKVfwuO7ppsrsthGaYhLwO3h3KYJgm7wxlOdoTpqSoqC47BqCaljPjpfEc50ez2GWJCyNZNNPkmycGMQDTBIcs47DLXBzNUanqlFWdN/un+MLRTiJ+J3KtpOV2pnufEgWFp7bG+MWVBI90NSwZ2Kq6QbpUxSHL/M3pYaq6VX7SGvagGyYTuQrRmvLJNN0RH5fGcnRFvFZJS7IIwGROoaBoC0rIWgY7KgG3/a4Dbk03eO1aAtOEV68mNnYwLUlSEPh9oA34HvBT4A+AfwacBb66yOpZ4O9rr38O9C13vw6bzOeOdFFWdW5MFviz12/ic9n4/JEufDMupMliBbtNxu+y0RJy8/T2pvqF9tJYjteuJhhKl3HYJFpCHr54rAu3w0ZHoxezpuixKebn4mgWWZYIehx84UgXj26Ocj1RYCxboTXsweu0EfQ42B4PcKy3kW1xyyb08ljeMmQxIep3M5mvkCtVZ30XDSipJg75vQi7akC2rBHzu5kqVTGBv3pjEJdDplS1jGmcdhsDyRIRv5P97WEqms6OliC7VtCYQDCbly6Mk8hbTSRffqL3oehUFgjm4/vvjuOt6dB+/91Rfu+pzYsuf2uqxLmRHF6XjQ/uifPB3a2zPj/c08hQqmxdc+/yQTUeWr9Nza9fT5LIK4xlK+xuC614wP9m/xRFxQq2Dvc0zlFLSuQUEnkFc+Gesg1NyOOgwesgXVLprukxdzZ6GUqVCHkcdVfKK+N5xrMKqWIVzTDRDYOhkRJFxcrae532uomKXZbY3hJiX4flijyerfDLapL+RIF0qUpJ0cgrKi67jYjPSTTg4up4DkO3hAY0w8TjsnGgo4GhjEJZ1fG7Hfgcdv7f1wcoKBpTBQVPzStC1Q3KVY3ORh+tYTdBt5O2sAeHXaKk6qi6ScTv4uSA1ay4uy3I5pgVQE7mKqRKVbY0BXh+V5yLYzl2t4U4P2IF/lcn8hzpaeTSWB7dMNnVGpyTdf7WqWEmcpX6eABuJAq0hj384N1R+ms+GjPVX57f1cxjmyP4aiUkR3oaef16ko4G76JeDL+4Msk7Q1magi4+e6hzyQz4fNhtVrXBSLq8piVLK3Um/yWQBt4Afgf4bwAn8FHTNM8use6vgH9Ye70fuHknO7bJllbkcNoquC8qOqlidVYwfbCzkeFUBUmCzx/prEvRgHXxOT2Y5vpkgajfhdth2YxOB0iSJLG1OUB/osjW5gBNATe6aXK4N8JbN1O82T+FicnTW5s43N2IXZbpinqJ+l387z+9yvnRLOPZCj6XnWaXnb3tIc6PwnC6jF3XkCVQdau4XDVMNANskiWHJ2M1SO5uD3J+JI+ESUnVKKsSUb+LHS1BNMMk4HIQcNv5/NEuVN2gJ+J74DIO64mA204ib1387BtQ2k8gWCkOtId548YUdlniYGfDkss3B130xnw4bJYt8u2B77FNUdobvHidNiIrYPe93mhv8JDIK0QDrroc4EqyucnPO0NZemO+eWVH4yEXE7lyrXnuwbt2uew2vnDUSrBNm5Uc7mlkWzyA12mrz0bHQ25cdhmvy47Paaei6UzmqxgmOB02moK1WKCsggmKpjOYKhL1uwh47ARddkbSZRw2mZDXTkU1CbhkDvU0cjNZpC3sZTxXoTfmQzPgw3tb+P75cW6lSkg1hZFsSaUl5MZpl5GQmCpaKiM22cbW5iAf3d/K87uaGUqX+f47ozhtMtvjASbyClpNk3okXeZARxinXSZbUvn6iUGqmkFfTyNPb2uqm7G9ciUBQLqocmksx0sXJgDQTbMuwABWlncyX6l/1hx0U60l6OC9bPdUoVoXawArTpppDtMb89ddTRdjOm6bzCkomnHXBjGfPNhOvqKuaWnrSgXTvaZp7gGQJOk/AUmg0zTN/FIrmqZ5VpKksiRJr9TW+5O7GcChngaKVY2Izzmn6aQ56GZrPMCZW2l+cmGcjx9or3+2Kebnzf4p2kJuvC4HfV2Nc+TantoWozXkwSab3ExatctOu1xvdrg2UWAoVUI3oCvi5eJYjuObIiQKVgagvcFDV8RHX1cjR3ob+YOvnSHotk7I41sjvHw5Ycn21bIFYa+dRzfF0A2rE3Z3WxCf087ViQJ+l510SWVfZ5h//qGd2GSJ8axC0GMXbnz3iQ/sbuHWVJF4yC0MWwQPNZ872kVfdwNep432xqXVPI70RqioBj6Xra7PPJPhdIkfnx8j4Hbw8QNtdzXrM5op43fb16Xp1FPbmtjXHq4FTSsfzD6zvZlHN0UXNJBxO2z0xgL4XPZ6/fCDht0mE7jtunx7kNUcdPPbx3v4rNLJYKrE69cSBF023A4bezvCbIr5UTSDkzdTvHotQbqoUtHKPLnVzz99bisnb6W5kSgiSRKHu63l4yEPTUE3Ya+Dr741SEUz2NkaoD3k4Y2BFIm85V4c9Njr/+5bmgPsbAnwF2/coqLp2GWZg51hHt0c5fiWGCcH0rxyZRK/y86+jjDvDmfxOmzYXXYMEzxOG821B1JF07kwmiNbVrHJs7Whn9vRzNnhDFubZ8vlJvMKl8ZybG0OWFl4m8yz25u5MpHnQGd4jurX+3Y0885wpr78vfL45ignbqbojfnvyWlxWhVoLVmpYLouUmmapi5J0s3lBNIz1vlnd7Iz3bAaCrMllae3NxELuGgKuPl038LOWNcmrOEMJEsoml6vq35ia4zdrSH6k9Y0Rus83d/fOjXMQLLIaLbCI10NPLPdOkiP9UZw1p4Oh9Ilpmo11wOVIi9fmkACOho9FKs6YY+D3W1BBpIFqNUJRQNOrk8WUXXTksCrHZshj5OPHWzl0d4YHqeNf/3DS4xkKnicNqI+LzudNp7b2czpWxn2dYTm2IsKVhenXRb264INz0qFUVvjyy8p87vsvLh34drqC6M5iopOUdEZTpfvuCfhzf4p3rgxhdMu11VB1hurLRW22ANId8THaKZCZ8TzQAbSd4LXaSWm3A4bv/V4L5LErBltgG3xAIOpEsPpJIZp4qzJOp7oT5GvqIznKrxyVePCWB6bJLGpyU93xIfHYQkDnBvO4rDZUFQdCYkGr4NtTQESRYWwz8nvHO+mJezlF1cSFCoqQY+DP3zfVrY0B7g+aTUdjmTKtDd4iYc8uOwy7w5nCHmcPNLVwMcPttWz7SGvg1jAhc9lq5deTdMZ8c6KE4w9Jsm8wqmBNCcGLBfmg50NPLezmT3tIfa0z3ZvnqY76lu2BOZyWG4GeyOwUsH0PkmScrXXEuCp/T0tjbeiBbzD6RIXR63dnRpI8YFlNL4c7Y1w4maKrc2BOQ2KjX4njf7GedczTZN8paYNqekoqs5kXkE3TX5+eZKKqnOkt5HSZSsrPpopU1A0wCRVUumOemkJeZjIVeifLKCaJopmWW86bTImlg2t0wYtYS/tDR7aG7xcGslzfjjPC7vjdEa8TOQrBNx2fv/pzRgm/M2pYQzTJFGozMq0L0SqaNVoz5QFtIxsFLoi808JCgSCB5eVOOPzFZWfXpjA45R53874ogpJ8zFVUNANk6aglV3bHg9wfbJA0G2/K1m76etcVTMoVLR1GUyvJdmKdZ/Klx9MNY87IVlQ6vfRvu6GuiTdTJoCblrDHva0hnA6ZN63oxmv047HaUPRrMb/QkWzgmzVoDviJVWscrArzKtXkmxp9uOwSXQ1+pAlmVjQyYH2BkazVtlpPOSlUtU50NnA9niAroiPoMfBd8+OMJYpoxsmLptMIq8wVVDY3hLk6kSegakiqq7z6UPvJRBddhsv7I5zbaLA4Z658YymG2g12+/t8SDjngonB9JcGc+Rr2ioNUOXO0kU6YbJu8MZXHYbO1uDVDWDK+N5moOu+jm9GO8OZ7g8ZmXBN3qCaqXUPO5rB1bEbz19lar6sgvOd7eF6uYrS2EYJhfHcjjtMlubA7SEPTQHXRQUDb/LziuXJ7kwZmlG7mkLs7ctxMcOtvGnr9xgIlfB77JbuqsuOw1eB1cmCgwki7x+Pcnh7ka2xf0omk5bg5eP7mvlK6/1oxsmv/loD41+J4ZhcqYm+TcwVeQDu+NEfE5awh5aQh6GUiX6kwUMw6R9GRqOg1Mlvv32MAAfP9BGV8RHRdX52luDVFSdna1Bnt8VX9a/jUAg2Lg4ZFBrQj/3MKta56cXJ/jWmWHskkTE7573Jj4T0zS5kSjic9nIlTT+1Y8voukmf/D0Zh7dbHkF/P7TizcxLsZjNaOWRp9z3lnGh52qZsmiVXXzgS3zWC6qbmDUOjEXU796bHMUmyyhGQZNARc2WeLXH+2mN+ajP1GoJdsKOG0GZ4eyHOlt5PpkASR4ZnsTrWEP/8tLVygoGkHVQcTvYjKv0BJyc2E0y8FOa7b7ykSegWSBf/z1M1RUg65GD+2NXgJuO9FajbvXYUPTTSbzVo3x4FRpVsb5+JbYnIeCK+N5BqYKXJsooBvw4t4WNjf5iYfcPLezmYtjWWyyxGCqdMeOxKdvpfnldUv72uWQuTKe58p4HodN4jcf65mT6Z+Jbpg1UzsrsSeC6TXA77Lz6492U9WMWUXvi3F9Ms+pgTRbmgM80rVwo8ypgRQ/eHeMoqLRGnZzpcmPohr0xvxWNtkh88vrU2RKVRw2Gc0w2BIPkC5VificTOQqOO02Pn+kk/FcBRlLEP38SA7dMLiRKPD+3ZZdeXPQhcMu88TWGDZZwmGXOb4lhqYb5BWNXFnjYGcDEb+rnn3PVVS+eXKIwakSAbe9fjFYjOnabbCexrsiPqo1fU3AarIQCAQPFysQRymajmGYaJJJVdOXXP70rTSvXUsiSVYdq1KL7C+M5lbEcjjkddTdEgVz+ci+Vq6M59nc5H+oA2mAlpCHF3bHSZeqizbPPrY5ylRB4epEgf98ZpgvPdpN0O3g/bUEVH+iwHfPjnJlPIfLbqOi6lwezxN0O7g6kWdTk59owEW+YiXjzo9myVc0zgym2dsepqjoPFfTeB6cKlJVDTTdIFW0AszuiI/moIv9HZaG8jsjGRIFheagi6mismiZZ7Kg8KNzY0wVlLrKx0CyWC+f2tUa5PiWGGM1neyhdIm/OZNiS5OfJ7bGqKg6pwbShL2OeZORMye0LYM5nZvJIhVVZzxXWdBpGaw655aQm9FM5YF48N2QwTRYT5Klqr7sYPrvrybJlVXGshX2tIVmaSTOxJIW0hhOl3DYZNLFtNUE2Brk2OYor1xJ0B72IEvwiYNtPL2tCafdRqBgY39nGFmCo5si/INH2ilWLQOVdKnK9cli7cD38VuPdXPiZgrThMe3RPC57KRLVY72RpjMVdBNkw/tbZ13fFXNwMSS6/O77IQ8S9ff7W4LMlVQaq+tEyLodvD+nXGG0yX6uhfPJgkEggeDtgY3A1NWt/5K1D6+b0ecTEnD67RxcJEkxTSlqhVwmyYc7AzTnyhS1Q0+sEfMjN0PYgEXscCDp5Jyt0yrVCyFLEvYZAlVN1FnZLHTxSo2SeLRTRGag27SJYXWsAePw85QqsTBrgaifhfP74ozli1ztCfC994ZpahouBxWDDIdi2xvCTCWLWO3yThsErtaQ7wzlLUsub1OdrYGyVdUPrKvlW3NARw2mV2ti8+2O2xyzd3RiddpqZTsn2H7LUkSn+7rIFHLlP9/b9wiV1Y5fSvN4Z5Gfnk9ybvDWcCq9b+99Gp/e5ibySIOm0xXoxdNN3j1arKWdc8tGkyDpcKRKas0ep1UVEuJLR5035VE3lqzIYPpbFnlr968RVUzeGJrbE6meShVYjKvsKs1WG/GaG/wcLGsEg+5cdgW/qG2NAeoalYmekuz1UAA0NbgJeh28EhXmGypyqYmPztbQuQrGl6nyXfOjHJhJIfTYSNb1shXtHqjic9l589+8xCDUyV6Yj78LjvP7njPbWhbPEC+ojGcKvHzK9a0x4t7W9g6z7RH1O/ixT2t7IgHaW/00Bb2cHYow+Ym/4KapS67rf4UPZOdrUF2Cj1qgeChoavRVw+mu5ahvrEUsYBrUUfC2znS24gsSXhdNvZ1NLCvY+kAXCBYa57aZnlTNAfddcnGbEnlq2/dQtVNjvZG+NyR2Q7N2bJa659iVvnTrx3uYCxTwWWXKFZ1y0mZ6ZrnllmOoL1RP5N5S5M8X1H5dz+/xq2pEk9ujfG5I0vbfYc8Dn7tUAfJQpWtzf551aemPTUAtjYHODmQwi7DLy5P1BsZZUmaVyHm0ni+Lm/3ytVJdreF2NESpKBoi5agqrrB5bE8jX4rQFd1g6++NUi6WGVTzMfHDy7dB7be2JDBdK6s1mucpnUPp8mWVb59ZsRqzstX6gfmke5G9rSGaAourq/51LYYkznLfrM74mNrUwDVMNgU83NtIs933h7hzK00O1uDnB1M09bgJR6y6qllWaJc1ZElCdttAXvI45i3Q/btwTTfPDnEyYEULpvM5iY/PTE/Vycso5cdLcE5zYHTQbBhmPzfr/VTqupcHM3NOZkFgoeV7j/+4T2tP/BvX1yhkawv3h3O1F+fGUzd9/277DYe33Lv5RwCwf2iVNW4Pllgd5vliDxNoWo17QFky7NN2GYm/B7bHJ0VTAfdDoLx5c2oz1ThSJeqXB7Lo2gGr19P8qm+jjkNv+liFY/TVk8ivnJlkp9fnqSgaBzubuQTB9vnnZU3TRNFM3h8S5TWsJt//t0LvH59iqO9ET64p4Wgxz7ru0/jsFuxyVC6RCKvcGW8wGcOdSBLEiHvwt/xtWsJ3hnKIksSXzrWhcMukylVOTeS4eJYjljAveGuExsymG5v8HC4p5F0qTqvZeo0Rq1O+Mp4nh+fH7McEw93zpEmKigaqUKV9gYPkzmFZME6Md7sn+L9u5rpDFgHc7JgTem46k5F1g5SRUuirzfmo9HrZEtzYNld5KlilUxZpVzVqUg61xIFOiNeroznuTZRIFtWeWyRWkKt9iU1Q9iHCwSCxZm+ZgFU1QfUBk8gWEF+8O4YI+kyboeNf3i8p57dbQt7OL4lSqo4Nw7JV95L+KWKypxt3g0tIQ8HOhu4NpFnX0d4jmHYyYEUr19LEnDb+cJRy8X53HCWyZxCtlxlLFthMl+hvWF2jbVhmHzr9BCjmQrHNkWI+JxoujX2oqKxLR6gqGjcmirS3uCdldzbHg9il2W+e3bEimFUHUUzlnQhnb4OmZjopkmDy87hnkauTRRoDbsZyZRW4p/svrIhg2lJkhYMMEMeBx/cE+fH58dRNavJbjRTxjSteuNEQZkVTFdUna++eYtSVWd3W4gnt8Zoa/BwdTxPplQlWVD42P42moNuMuUqDrvMZw+3s7s1TLpU5eJYjj1tYXa2Bmc5CS2XI70RcmUNVdMZzVQ42NmA12lHqZ2I08HyfMiyxCcPttOfKLB9mbVfAoHg4WVLk4+3hy3N/e0tD4a+q0Cwmui1e7Bhmtx+N16o36i9wcuxTRHSxSrHNi0vw5qrqLx5Y4qI37WgSMJ/+ewWJvMVIr65M+wjtXKLfEUjW7b0s/d3hsmULWfA9gYPzfPI1RWrGqMZq/Tr2mSBo0e7+OKxLm4mi3zqkQ5U3eDrJwbJVzS2xwPzSBGbVDWD0WyZF3bFlwykAZ7cGiPkcRD1O+sZ7+NbYlwdz3N1osCWO9SXXw9syGB6KdIlFd0w6U8WuTCa42BXA5lyFa/TTu9tTTdKrZHRWq+K0y7z6b4OTt9K8erVJKYJmbLKrVSJy2PWTWhzU4D2Ri/tjV72tN95AD0Tv8vOxw+28bEDrfz88iSDqRLHt1juh7mKyr4lth8PuZd18AoEAkE04MEmWdexWHDjd9ALBKvNB3e3cGEsS3fEd0c66kd7F541n4/XryW5Mm6dm21hz7z3dUsBY/7z9tFNEVTdoDnopinwXoA6n372TAJuB/s7wgxMFTlSK0eZKYBQqmr12u90aa7y17TxzZamwLKbmt0O25x/n0ypSq6iEQ+5uTZZ4GDXxhJGeCCD6eagC1mSkCRoCrgIeRwLGpuEPA7et6OZ4XSJQzPqmva0hcmVNSTJko+ZNomRJWlZJRypYpVMqUpP1Fd/gnyzf4ozg2l2t4Z4YuvsA1ySpFlNiQKBQLDSfGB3nDf7p5CAD+wSEnICwVKEvA4eXWZ2+V4I1yzPnXYZr+vOReCbgm4+1dfBZK7CYKpEV2T5DcZPb29a8DOv0877d8YZmCrSN0/GfF97mELFipV23jZDbpomL10Y52ayxOObows6K07vJ+hxkCurxBd4YFjPPJDBdFfEx2882o0kW8X+M51/5mM++0ynXZ51gO3rCBPxO3E7bPMW4s8kW1b56pu30AyTQ92N9UL6twczKKrBmcE0x7dEH3qdT4FAcH8xsZqaJCSMOZPWAoFgrTi2KUJ7g5egx37Xzp3j2QrfODmIaVoB8t2Uns7HYspft8dKMylWdS7VZvTPDqVnxVmGYVLRdLxOe307nz/SSa6i3rF5zHrggQymgXonaUHR+MaJQYpORWnIAAAgAElEQVSKzgu742yL373Lzu2F+wtRrur1Wud85b1pkV2tQc4MptnVGhKBtEAguO+YQKGisTJm4gKBYKWQJGlRA5blUFC0ukHbzNhjrfA5bfTGfAwkS+ycoYmtGyZ/fWqI8WyFI72N9cy/22FbMOm53nlgg2nTNBnNVsiWquQrVr3PzWSxHkwPpUr88NwYQbeDTxxsu6sfcHCqxKXxHDtbgkT8Ts6P5GgNu2lv8PLM9iaSBYUjM+qCntgaExlpgUCwZngcMhXVQJLAvYBxlUAguDMujeWoqDp728NzpGzvJ5tiPh7bHKVU1Ti0SmZs1yfzvHRhguagm4/tb51Xu3oaSZL46P62Odb1BUVjPGs1Pd5MFhcto6lqBpP5Ck0B94Jme9OcGkiRq6gc643icd7foPyBDaZfuZLg7FAGr8NGd9RLUdE5OMP558JojnJVp1zVGcmUl3TqmY8fnBtFUQ0GkkWag25uJovYZImnt8e4mSyyqzU4x0jlbgJp0zQ5P5Kjqhvs71jbk1VgTaW9PZimJ+arC+4LBBuBn5wfJ1FzQ/27C+M8J+qmBRsA0zR548YUparOY5vvf6C0GP2JAj85Pw5YQbXXaaevu2HWTPZwusRgqsTuttBdl3AsB0mSZmlarwbnRrJUNYOhVIlkobosAYTb456Qx8GBzjCDqRJHehZv1Pzbt0cYyZRpCbn5zOGFvTQGkkVeu5a09oe0aB34avDABtPTN4ySqvPMtuY5AuI7W4L0JwsE3Y45FplgdZZOFav0RHxzrC3TxSqj2TIuu8ylsRxOmw1fLWiWgJ9fmsQwrRNoyzwuhnfKtckCP7s0UfvL5JEN1uX6oPHTSxMk8wpXJwp0R3wbdlpK8PAxMFWq6+8PTG08LVfBw8mNRIG3blomQw67zJNbF1eouJ/ItUBR0w3ODmVoCXnIV1S+eKwbsOR3v3NmBM0wGU6X+XRfxxqO9t7Z1RpiJF2mKegm4ncuvcJtqLrBiZspBqdKbG7ys3kJGbxkTad7qlhddDmfy25ZvmsGUwWlns2+XzywwfRT22K81Z+ivcEzrxNPZ8TLf/HUZi6O5vj6iUE2N/nrEjJFReOrbw1S1Qz2dYR4Zvt7KhtVzeAbJ4eoqDpep42wx0nQY8fvsrOjJUhr2M3r15IMp8sr9kPOzETbZDE1u9aEPQ6SeQWfy3ZHUkkCwVoTnjFTFvY8sJd/wQNG0O1AliQM06RhEWe9taA76uPD+1ooVHTeHkyTKavEAu+JFEiS5QmBYc4xWtmIbG0OsPUekoQ/Pj/Od8+OoKg6iUIDO1uDhL1OTg6kODecZX9nmIOd76mGPL8rzoXR3BylkNuJBVx87kgnf3dhnKF0mW+eGOI3HusmsIozATN5YK+mTQE3H97XuuRyb92cIlNSOTWQpq+rEY/ThqIZdfei6XrraQzTRK25A7kdNtobPJSqOr0xH3trmtAfO9BGIq/MOqHuhU0xPx/e14qqG2y/hwZKwcrwgd1x62Ep6BIlN4INRUfUh3vQshRvbVi+dJZAsJY0Bd184WgnFc2YdyZ5rdncZN2Xt7cESJeqNM9IpLnsNj7d18FIpsy2FZip3ugUKhpBt4OxikbQ46jP6v/q+hRGrZxnZjC9KeZfdhlu1O8i6HEwkVPQTRNNv3+KRQ9sML1cNjf5OTWQpqPRi9thZRkbfU6e29nMeLYyS3sarAD6I/taGUyV2Nsewu2woajGrOy3wybTusIn/FJTIYL7h90mL1ucXiBYT/zek5sYTpeRJPjdJzat9XAEgmUTWUKSdj3gdtjmNVWJBVwrllzb6Dy3s5lYwEWj18Ge9nB9dndzk5+rE/l7jnWe3tZEyOOgOeie5Xa92jz0wfTxLTEOdTfissuziuR3t4XY3Ta/wHh31DcrmBI1swKBYCPQGfHx//zGISRYtAtfIBAIVoNYwMVzO+ca1L24t4Vn1aZ7jqd8LvuSro+rwUMfTINVk3xtskBTwEXYe/+eZASC9Ur3H/9wrYcgWCWmJak6Gu9N01YgEAiWYjJfIVfW2BTzLalmtpETkyKYBn56cYIr43ncDhu/+Vj3qv2gyYLCj8+P43PaeHFvCy77xj1wBALBxuPtwTT//ufXQYI/fHZLvc9DIBAIVppUsco3TgyhG+Ysc5bVIFdR+eG7Y8gSfGhva70W+34h5vmw1DsAFO0958LV4NxwlmRe4dZUiZvJ4qrtRyAQCObjwmiOgqJRqGhcGsut9XAEAsEDTFnV0WsxVVHRV3VfF0dzjGcrjGYqXB7Pr+q+5uOBDKZP3EzxvXdGSda0ppfifTua2dcR4oN7WuaYrKwkPVEfdlnC55q/SUEgEAhWk0d7IzhsEk6bvKpZIoFAIGgLe3hmexMHOsM8vnl1rzddES9Ou4zLIdO5jBK2K+N5vnt2ZMUSmw9cmUcir/DL65YLjm4YfPxA+5LrNPics7SkV4vuqI/ffXITNlkSkmoCgeC+kypVOVCTnUoWFFE3LRAIVpV9HfenlKwl5OHLT/Quq7naMExeujCObphM5Cp8eQWUjR64zLTfZcdbsxqN+e+f+81ycdplEUgLBII1Ieq3tNHtskx0A0iNCQQCwXJx2ORlqRTJslR3b1wpycIHLjPtcdr44rEusmWVeHB2MD2aKXNyIEVXxMf++/S0JBAIBOuFeMhNR4MXSYLm4PpLNggEgoeTG4kC50ey7GgJ3pPD4nL51CMdJAoKzSKYXhiv047XOfervXIlwUSuQn+iyOYm/6rWRwsEAsF64/xIlsFUCYALo9l6yYdAIBCsJT+9OEG5qjM4VWJLk39JGb17xWmXV9RN84Er81iM6XR+0OPAbX+ovrpAIBAQ9buQJJAlSZR5CASCdUOsdj2KBVyrHkivBg9VavbZ7U3sag3S6HMK9y+BQPDQ0R318aVj3UhwX612BQKBYDE+ur+VybyyYR/yH6qIUpYlWsOeZZuyZEpVNN2Y876mG2RK1ZUenkAgEKw6hmmgm3OvawKB4OFFN0wypSqmuXpeG4tht8m0hj0417BqIFdRUbS708N+qDLTd8IvrkxydjBDLODis4c76wocumHy9ZNDJPMKBzrDPLWtaY1HKhAIBMvj1ECKP/npVQD+2xe2sa9D1EwLBAL49plhhtNltsUDfHBPy1oP575zfiTLTy9O4HXa+PzRrjvuqVs3mWlJkv5rSZJeX+txTDNca9JJ5BUq6ntPKmVVJ5m3zGCG0uU1GZtAIBDcDRdGc2iGiWaYXBi9/y5hAoFg/WEYJiMZK54ZqsU+DxvT37tU1ZlapuHfTNZFZlqSJBewb63HMZPHt8R4s3+Knqhvlse732Xn2KYIN5NFjvZG1nCEAoFgtej+4x/e0/oD//bFFRrJyvLi3jjXJvLIssQLu+JrPRyBQLAOkGWJp7Y1cWksx4HOh1M2+FBPI3lFI+xx0NFw52ZW6yKYBn4H+AvgX671QKbpifroifrm/exob0QE0gKBYMMR9bv5Hz++Z62HIRAI1hn7O8IPtf9G1O/i030dd73+mpd5SJLkAJ40TfPnC3z+ZUmSTkmSdCqRSNzn0QkEAoFAIBAIBAuz5sE08EXgawt9aJrmV0zT7DNNsy8Wi93HYa1v3h5M8/13RpnMV9Z6KAKBYIOgaDo/uzjBy5cmqGpC0UMgWC0ujeX43jujD20N8sPGegimtwH/SJKknwC7JEn6x2s9oPVOtqTyypUE1ycLvHJFZOsFAsHyODec5dxIlneHs1wYza71cASCBxJVN/i7CxPcmCzws0sTaz0cwX1gzWumTdP8o+nXkiS9bprmv1vL8WwE3E4Zv8tOQdHqrkECwUzutYFO8GASqTkgAkR84tohEKwGdlki7HWQKlaJiHv0Q8GaB9MzMU3z8bUew0bAZbfx+aOdpEsqrSH3Wg9HIBDcxnpVA+mJ+vjC0S4kEDd5gWCVkCSJXzvUQbKg0BLyrPVwBPeBdRVMC5aP12nH6xQ/n0AguDM2ql2vQLCRcDtstN+FxJpgYyKiMYFAIFhnrNfMtkAgEAjmIoLph5BSVePVqwncDhvHt8TqVukCgeDBRtF0/v5KAlmSeGJrDKd9PfSgCwSrj26YvHotgaLqPLE1JmZ2BSuKOJoeQk4NpLk0ZlkJt4Q8bIsH1nhEAoHgfnBuOMuF0RwAEb+TA50NazwigeD+cG0yz9nBDGCVST6xVUjtClYOEUw/hEzXTNpkiQafY41HI1gIocghWGmEmofgYaXR68QmS+iGKfoGBCuOCKYfQna2BokGnLhsNkJeEUyvFiIYFqw3eqI+vni0C0mSaPQ513o4AsF9oyno5tePdaPoOk0BoYIlWFkk0zRXZ8OS1Ar8ANgJ+E3T1CRJ+hOgDzhjmuYf1pab895CRKNRs7u7e1XGKxDcKwMDA4jjU7AeEcemYL0ijk3Beub06dOmaZpLNpesZmY6BTwLfAdAkqSDgM80zeOSJP0HSZIOAfrt75mmeXKhDXZ3d3Pq1KlVHLJgKUYzZQaSRXa0BGkQma1Z9PX1ieNznWGaJu8MZzFMk/3tYeSHtNlWHJuC9crdHJuXxnLkyioHOhtEE61gVZEk6cxyllu1YNo0zQpQkaT6zesY8LPa658BRwFjnvcWDKYFa4tumHzn7RGqmsGNRIEvHute6yEJBItyYTTHLy5PAiBLEvs7wms8IoFAcC8MpUr85Pw4ABXN4EnRSChYB9zPR7owkKu9zgINC7w3C0mSvixJ0ilJkk4lEon7MlDB/EhYNqkADpvIBgjWPzOPU4ft4cxKCwQPEg6bXG+iFee0YL1wPxsQM0Cw9jpY+1uf571ZmKb5FeArAH19fatT4C1YFrIs8em+DgZTJTY3+dd6OALBkmyLB7DJYJiwtVlIQAoEG514yM0nDrSTV1R2xINLryAQ3AfuZ3rxDawaaoD3AW8u8J5gHdPgc7KvI4zPJYRgBBuDzU0BEUgLBA8QnREvu1pDD20PhGD9sWoRkSRJDuDHwD7gJeC/x6qhfg14xzTNE7Xl5rwn2FgMp0sYhnWBu53xbIVSVaM3tjEz2elilURBoTfqw14rGahqBjeTRVrCboJuIS240TBNk5vJIl6nnXjIksgyDJP+ZIGw17miGrSabtCfLBLzu2Y17CYLCplSld6of00CAmFXLtgoFBSNkXSZrogXt8NGRdW5NVWivcGDqhucvpVmRzxAa8Pc+49AcL9YzQZEFSvbPJO35lluUTk8wfqmP1Hgu2dHAfjAnjjbZ0y7jWXLfPPkEKYJT26LcXCDua2VqzpfOzFIVTPY2Rrk+V1xAH50boybySI+l43feqynHmQLNgZnBjO8ejWBJMFnDnUSD7l59VqCtwcz2GWJX3+se8Uekl6+PMnF0RxOu8xvPdaDx2kjW1L5+luDaIbJI10NwolNIFgA0zT55skhcmWVtrCHTx/q4DtvjzCerRBw2+lPFLmRKBDxu/iXH90lzFgEa4aIAgT3RKmq118XFW3OZ9My5rd/thGo6gaqbgBQqr43/kLtu1RUA32VdNoFq8f0sWia7/2u08exZphUVH3Bde92X6puUK0dS4qmoxnmrM8FAsFcTBPKtXN0+ro7fc7kKxpFRQWsc6pcXbnzViC4U0Thq+Ce2NkSpKBoGIbJvvbZsmO9UR9PbI1SVHQOdTeu0QjvnpDHwQu744xlKrOy6i/sjvPucIbuiA+X3baGIxTcDYd7rGPR57LRE/UB8MTWGB6njZjftaLuaM9ub+b0YIrWsIeQx8p2NwXdPLezmURB4fAGPC8EgvuFLEt8ZF8b1ybz7G4LAfDi3hYujubY2hwgW1Z5+fIk+9tDdDSKMg/B2iGCacE9IcsSR3sj834mSRKPdG3sYGF7PDirdAUg6nfxzPbmNRqR4F5xO2xzSiv8LjtPb2ta8X2FvI55j5XpwEAgECxOZ8Q7qx+nJeShJeQBoANxLgnWB6LMQ7Ag5apOtqzO+5lhmEwVFHRjeWUO2bK6JtNwqm6QKlZXdJumaX336RIQwfqjqi3/d59vWUXTSc+z/p0e9zPJltQVLSERCB4WdMPk+mSeym33kOlrcaqozCrFWw53co0QCJZCZKYF85IqVvn6iUFU3eCDe1rmSIv9+Pw4Vyfy9aaQxbg6kedH58Zw2GQ+e7iTxvtkQ64bJl8/MchUocr+zvCKZR5fvjTJuZEssYCLzx3uFPJM6wxVN/jaW7dIl9QlG/xmLtvX3cDxLTHKVZ2/evMWBUXj+JYofTNKMb73zig3k0W6Il4+cbB92WM6P5LlpxcncDtsfO5IZ73kQyAQLM2/+fElzg1n6Wj08m8+saduxvTShXF+dWOK0UyF/R0hPn2oY1llWlXN4Ktv3SJTUjnc08hjm6Or/RUEDzgiMy2Yl0ReoaoZmCaMZMpzPh/JlAAYzZYxlsjSjWTKmKZ1AUvklVUZ73yUVZ2pgpV5GEnP/Q53y/S/RyKv1JvKBOuHUlUnXbJmVOY7dmctq7y37HDtGMmW1Xqz0+3rT/89usR2b2d6vYqqM1W4f+eAQLDRMQyTgaR1vxnPlikp72Wnh9NlChWNfEVFUQ0mc8s7t4qKRmb6GrGC9wbBw4vITAvmZVPMx46WIGVVm1fS7qltTZwdzLC9JbBkZvZgZwOZUhWPw86mmG+1hjwHv8vOY5ujDCSLC9Z13w1PbI1x8maKTU0+3A7RgLjeCHkcHO2NMJQqcWzT4r97yDt32eagi4NdDUzmKhy77bh5elsT50ey7Gq7M+e1w92NFCoaIY+D7sj9OwcEgo2OLEt85lA7L12Y4HBPIyHve7M6T22L4bTLdEe9d2TO1OBzcqSnkeFMWWSlBSuCZG4gaa++vj7z1KlTaz0MgWBe+vr6EMenYD1y+7EpTFsE6wVx3RSsZyRJOm2aZt9Sy4kyD4FAIBAIBAKB4C4RZR4bEN0w+du3RxjJlHlme1NdGihZUPj2mWFkSeKTB9tn2RcLBBud65MFfnJ+jEafi08+0sZEVuH7744ScNv51CMdeJyi5EYgeNgpV3X+5vQQuYrGrtYg54azNAVdfOJge71xUSBYacSRtQHJllUGUyV0w+TCaLb+/o3JAkVFJ1/R6E8W13CEAsHKc3k8h6qbTOQqTOYUrkzkqWoGU4VqvSFWIBA83IxmyyQLVaqawatXE2iGyWimQlI0/gpWERFMb0DCHgebmvy4HDJ728MYhsmlsRw+l52Qx0GD18HmmH+th3lfSRYUzo9kUTSh47vR6U8UuDaRn/P+7tYQHqeNtgYPzUE3O1uD+Fw2moNu2hvWzv1sIlfh/EhW6I4LBKtEtqRybjhbtxJfjLawh3jIjddp49kdTbgdNjobvcT8LmtbZWtbhWVsSyBYLqLMYwNiWay21v9+48YUb/ZPIUnwmUOdxEMrZ4e8EaioOt88OURVM+hPFmf92wg2FtcnC3z/nVEAnttpzHI36476+L0nN9X/bgt7+PITm+Zs436Sr6j89ckhNMNkJFPm+V3xNR2PQPCgYZom3zo9RL6i0TTi4vNHuhZd3u2w8dnDnfW/H9s8W2f+W6esbcWGXXzh6OLbEgiWi8hMPwBMZ8RMk4cyO2aYZt2Rrqo9fN//QWLm76dsgN9SN0x0Uxx7AsFqMq3nf6/nmGmaK7YtgWAmIjN9n8lXVF66MIHDJvH8rjhOm8xPL02QKlbZ1uzn0ngeRTVIFhQkCZ7bGWd7PMBLF8ZRdZP372om6LZ0NofTJV69miTmd3Kkp4ErEwVeu5bkyW0x2sKeJcfy6tUEQ+kSYa+DdNFyi9vR8p5+biKv8PKlCcJeB8/tjGNbRE/69K00l8ZyHOxsYGdrkIqq18f8/K5mAu75Hd+uTxZ4s3+KTTH/kprAC+F12vnIvlaG02X2doSWXuEeuT6Z583+FJub/CuqX/2gMJmr8PLlSRp9Tp7b0cy1yQInB1JsiwcoKhqDqRKYVgbpfTub646Yb/ZPcX0yz6aYj6aAi5F0mQsjWUwg7HXw/K74orref/GrAU4PpnlxdwvP756dIba2XeBobyObm2Zr0WZLKj86P8a1yQKboj6e3x3H67Txf/3iBmPZMl861s2+jvC8+wx7nXxobysTuQr7F1jmVzeS3KxpnW96yMqvBIJ7ZapYxSZJjKRLJPM2/te/u0JfdwMj6QotITf9iQLvjGT56L5WntnexCtXEgynS5iA226jLezmP789Qjzo5p+8bysfP9DGjckim5v8/OT8OJlSlS3Nfi7//+zdd5Ak153Y+e/LzMqyXVVd7X33+BmMBWbgLSGSoAHJJbnkkutXFBm7dxsbcTppTcRKoZDuTnv3h3R32lBoT1pJa7TcJXdJ0ICgAQECIOzMYLyfnvbd1VXd5X1mvvujumvamzE904P3iUCgpyoz61X3y3y/fPne701k6G3w4zN1zo6lOdgVZm9HaF3tIMC1eI43r8bpifh5fLvKYf1BoYLpDXZ6NMXwdHWy1MWJDBG/ybmxdPW9kRSNAZN3rk3jSInb0HEbOroQ9MdytW1mk8y/0z9NNF0kmi7y0b0tMys6VXj76hSfe2DlpY4TuTLHBhNIKXnlwiT3tYd4/XJsXjB9bHCa8VSR8VSRXa1BehuXXmzCdiSvX44hJbx+Ocae9iCXopl5ZX50mcT4P78SZzpXJpYpcbArfMMZGXob/cuW71Z743KcRL5CLFPiQOeNl/le9d5AgolUkYlUkT1tQd64EiddqDA0lcc0NJL5MhPpIve1hzg+mOAf7WmhWLF56+oUAJYt2dka5K3+aYYTeSxb0tfo5+JEZtmgNlWo8NKZCRwp+fvjI/OC6bnHfuNyfFEwfWIkyfmxNBejGYplm8Y6N0GvwXsD0wB86/3RZT8XYFtzgG3NSwfJuZLFO/3V47x5Ja6CaUVZp2ODCTLF6k140XIIeV0MxHNsaQowEM9ydDCBS9f45rER9neGOTGcZCpXIpousactyI/OTZApWowmCpwYTvBgXwNtIS+DUznOj1fb3pPDSZqDHqKpIpYjcekar1+Os7cjNK8d3NkapG+VdubNq3Em0yUm0yX2d4VqnV/KvU0N89hgXfU+dE1gGhptYQ8NAZOA20AI2NsRRAhBR9hLg98k5HXR0+CjNeTBNDR0TdBZf73HuaehOukq7HPRVe8jPLMy1OzrKwl4DBoCJkKIWnDRHZl/keiK+BAC/G6dxjr3sseqlss389nVY7SFvHPKvHx5umfK2hry4DY2R3XsmbmYbqYyb6Sehmq9CbirdawnUv0b9zX6aZyp7811HoS4/vc3dY22mbH+PQ1+WoJuPC6d8MyE2tnzZTl+U6czUj03Fq6C5jbmH3uhrnovAY+Bx9Sp8xh0R3z0RKrnkyYE97Wvb7XDubyu6gRJgG618qGirFtPgw9DFzTVuWnwm4S9rlqnT3PQQ9fMeb+ztY6g16De5yJgGjTXudGE4GB3GAEEvca8QLgx4K61vfe1h2rHm73hnW1HuyP+WjvYtEI7WCvvTDvaWOfGb6r+yg8KtQLiHVAo2whB7ZF12XIo2w4Bt0G6WMGjaxQqNhIIelxomqBYsZGSRb2gmWIFr0vH0DUs26FQsZcdUrGQZTvkKzYB0yBTsgh6DISY/wgrW7IwdQ1zlaDRceSiYyxX5oXSxQp+01j18dndZKkyq5W8rptbb6SUpIsWdW4DR0ryFRu3oWHZEr/7emNjO5Jc2ar15JQsG8uWaELMO1+WU67YTGZLdIS9i+rxwmMvlC9b2E71s2bLlCtZZEoWrcGbm9C72mdvBLUConK3Wst1M1ey0IWgYNloQNDrIluy8Lp0HEfOO+9n20FzzjUmmi4QcBv43fPPwYVtr980ELCoLVtrOzhrM7ZpytLWugLiht42CSGeA/5g5p87gd8G/jvw/sxrn5VSTm9kme6EhcGlaVw/SWcbXHNB4LBcIDE3cDZ0jbp1JKU3dI3gzPYh79INfcC9tiqiaWLRMVYLfmZtxsdgm7HMG2luvRHiet3QELU6t7Bq6ZqY93utDnNa+2eaLn3ZpyALj72Qb4keJL/bmBfs36jVPltRlJXNnoeeOW1nre3TmXfez20HZ0/fluDSc4iWanthcXu41nZwljrfP3g29Bm1lPIlKeXTUsqngSHgJ8Dp2dc+CIH0XFdjWV46M8FYsnCni7KiiVSRl85McGVyce7flUgpebt/ip9eiFKsqPzPCkxlS/zw7ERtnsCtcmokyY/OTpDKV9a8z8WJDD88O8FkpnhLy6Ioyo0bnMrx0pmJ2tyi5aSLFX58LsrxocQGlezWKFZsfnohytv9U2ymkQHKyu7IgB4hxBYgKqXMCiF2CyFeB34O/KH8gNQuy3b4/qlxbEcynirwm4/13ekiLeuHZyeYzpW5FM3w20/717wk69VYrjbxy9A0ntzRtMoeyr3u5QuTjCYKnB9P0xXxrnlI0kqmsiVePj8JQKFi8+mDHavuU6zYtQmL8Wxp1dy1iqJsjO+dGqdsOQxO5fjaU8vnkf/55TgXJqodPO0h76ZZX+Hda9OcHK6uXNwYcC87eVnZXO7U7KnPAt+a+Xk78CRQDzy/cEMhxFeFEEeFEEdjsdgGFvH20jVBnad6L3O3PxKafeTldxvoYu1jwOo8BtrM9sFlhpEoHyyzdd3r0tc8/nA1njnHWms9MzSBb+aR8XJDnBRF2Xiz5/Bq5/Ls+y5dbKqMSrPl1sT1GEDZ/O7UX/J5qgE1s0M7hBDfBg4B35m7oZTyz4A/g+oExI0t5u0jhOCLR7qYSBXv6FLIa/GJ/W0MT+dpCXrQ1jGhoiXo4csPdVOs2HRF7u7vqGyMD+9pYUdLgKY6N27j1jSAfrfBrzzUw3S+XMscshpD1/jSQ91Mpot0q7qpKHeNX3ygk7FkgfZV1kp4dGsDHWEvIa9rU90QH+wK0+A38bjWlh1E2Rw2PJgWQrQCZSnllBDCDxSllDbwGHB6o+kmgEMAACAASURBVMtzJ/lMgy2bIO+sS9duuJzqYqHMpWvittT5kM9FyLe+BjXgNghsgvNPUT5IPC59TdcIIcSGrS1wq6nOpXvPnRjm8WnghZmftwPvzYyZ7gK+eQfKs6lci+d4fyix7LLh46kCxwanKZTXP+FvcKp6bLXMqnK3yZUsjg1WFym6EcWKzbHBBKNzJvvajuTUSHLdE2sVRbl75EoWb1yO8eNzE/OuD9O5MscGp0nmy3ewdMoHxYb3TEsp/9Ocn08A9290GTaryXSRF06MImV1xbendzbPe79Qtvn7YyNUbMnwdIHPHFp9ItaseLbEt96vHjuRL/OhXS23uviKcsN+MDO73zQ0vvJE37qHiPz4XJQrk1kMTfCbj/cRcBscHZjmzZkJsp+7X68tIKMoyubx4ulxfnwuSqZk8fCWCF97cisel87fHxshW7I4PZLiN+7iCf7KvUEt37aJSGA218lyg8evv7++4eVzc6g4qmNaucvcbJIfOef/s8eS896/Z6ZjKMoHimTO+T3nOjF7TqszW9kIairpJtIS9PCpg+0k8xX2d4YWve81dT47M3ljvUsgN9W5+fTBDqZzJfZ1hG9VkRXllvjYvjbOjVXT6d3IxMUP726hLeShNeippeM70hvBbWj4TGPJZcYVRbn7fXxfGy1BD/mSxcHucG2xsM/e30l/LMd2lXpO2QAqmN5ktq4yMaMj7KVjlVnQy+lr9NO3SSd0KPe2gNvgwb7IDe/vNXWO9M7fX9cEh7rrb7ZoiqLcQQG3wVNLrGHQGHDTGFAT4JWNoYZ5KIqiKIqiKMoNUsG0oiiKoiiKotwgFUwriqIoiqIoyg1SwbSiKIqiKIqi3CAVTCuKoiiKoijKDVLB9B3iOJJcybplxyuU7WVXRVyPbMnCcZbOzJlb4T1FWQspJdmSddN5oxfKlSok8+VFx71dn6coyp1XsmyKlepqv8WKTdlyav9fz76KcrNUarw7wHEkf3d0mPFUkcO99TyxfXFan/W4HM3w4ukJvKbGLz3YTXAmj+56/eRclNOjKbojPj73QOe89968Eueda9O0BD188UgXuiZuqszKB9OPz0U5O5amt9HHLxzqXH2HNTg9kuI/vXaVdKHCJ/e384UjXbX3fnh2gvPjGbY0+fn0wbWvCKooyt0tlinxd0eHcRzJ4d4I7w1MkytZaEJQ5zX44uEuGpZJjTd331+4v4POerX6qXJzVM/0HVCo2IynigAMxHM3fbyBqTyOlORKNpPp4k0cp1qWoek81oJe7v6ZckbTRXLlW9ejrnywXJupR4NT+Vv2lOPaVI5EvkzJcrgwkZ533GvxPAAD8bzqnVaUe8hoskDZcrAcyYnhBLYjiaaLpAoVShWn1sYuZWzOvkPT+Q0stXKvUj3Td4DfbXCkN8K1eJZHtjbc9PEOdYeJZ0sE3De3ktujWxs5NjjNztYghj7/PuuRrQ28eSVOT4P/hnu+FeWxbY28P5Rgd1sQ7RY93TjSW8+F8TSJXJnn9rbOO+7j2xo5MZxgT3sIIdTTFEW5V+xsqaM/lsVyJA/3RXjz6hStITe2U13IZdsKKx/uaKnj6sy+ezsWryasKOulguk75PHtjTy+vfGWHKsx4OZLD3bf9HH2tAfZs8wy5FubAquuvqgoq9nbEbrljVdbyMs/f27Xku/t6wyxr1M1lopyr/GaOp+9//pQse51dCQt3FdRbpYa5qEoiqIoiqIoN0j1TG8SP70Q5VI0y4N9Ee7vrq+9XqzYvHBilEzR4uP72mgPe2/L56fyFV44OYqU8OmD7YR95m35HEVZTSxT4rsnxzANjc8c6iDgvn4Zk1Lyw7NRBqZyPL6tsdYLPjiV40dno0T8Jp862I5LV/0IirLZZUsW335/lIrt8PyBdhqXmXAIS18bErky3zk5hibgUwc7CHnVEEblxqgWZRMoWTYnh1MUyjbHBxPz3htJ5BlLFskULc6OpW9bGS5PZpjKlpnOlbkUzd62z1GU1VyYSJMqVIhlSlyLzZ/AmylZnB9PUyjbvD+crL1+aiRFtmQxNJ1nYoWJSYqibB5XJ7PEMiWS+QoXJzIrbpsr29evDUPVdvRSNMN0rkw8W+bKpGrXlBungulNwG3obG8JYDkOQY9BoWxj2Q6Xoxn8pkG9z4VLF+xoCZDKV7gczSzKxnGz+hr9+Ewdr6nT13jjkxwVZS3yZYtL0QzFio3tSK5MZkjkygBsaw7gdmnUeQy6G3ykCtU6X7EdAqZBb6MPIWBP2/Xx/7ta6zA0QWOdm+bg8r1XiqJsHj0NPgJuA7dLW3VOj9/UifhNkvkybWEPVyaz9DT68Jo6frdOb8P19Hhly+FSNEO6WLndX0G5R6w4zEMI8WsrvS+l/ItbWxxlOZ/c3048U2I0WeRb748S9rm4OJHB49L5rcd7cWkaFcfhz98YoFix2dlax8f3td2yz28IuPnqk1sAVFYE5bb7xtERpnNlWoIemurcnBlNYRoav/lYL20hL7/91FYAKrbkz39+jULZZntLgE/ub+cXDnXiOHJeVo/tLXVsbQrcsgwiiqLceWGfyVee6ANWb5cS+QqpQgW3ofHDM1Haw2mO9Eb42hLt2ounx7kWz+F36/zWY32LslspykKrjZk+ssRrAnge6ADWFUwLIXqBd4DzQFlK+REhxD8DPg0MAr8hpVS3gkuQUpIrV1drypUsDL164pcsm4otcRuCSkVSsqrbZG/h6oqzVBCtbITZVQuhWtd9pg5AxXYoWw4+83pdtBybUsWpbTtrqaBZBdKKcu9Za7tUmHnKZTvUVj7Mlqwl95+9/hQrDraUanKZsqoV64iU8ndnfxbVGvfLwO8DbwP/2w1+5o+llL8yc8wm4Bkp5eNCiN8HPgN84waPe08TQvCpA+1cimbY3RbE69I5PpSgK+KrTcAKuA0+vq+N4en8vEmKirKZLKzrfrdBcHCatpB30cRXn2nw8X2tDE3nOaTqvKIoy+gIe3lmVzOJfBm/qZMr2zzYG1ly24/e18qpkSS9jX7chr7BJVU2o1VvuIQQBvAbwD+l2qv8eSnlxZv4zGeEEK8D/wBcAl6def0nwJdRwfSyuiI+uiLXx3U9u7tl0TY7WurY0VK3kcVSlFtuYV3/0K7FdX3W9pY6tqs6ryjKKg52hde0XVOde8n2VVGWs9qY6f8J+D3gZeA5KeXgTX7eOLADKAEvAEEgOvNeCljUtSSE+CrwVYDu7ptfmERRFEVRFEVRbpXVeqb/X2ASeBz47pyxRQKQUsr96/kwKWWJaiCNEOJ7QJrq2GuoBtbJJfb5M+DPAA4fPizX83mKoiiKoiiKcjutFkz/DvASsFQQ+8X1fpgQok5KOZsM8jGqwfqXgf8T+EdUx2IriqIoiqIoyqawWjD9p8DPgF+VUo7OfUMI8UtUg+D1eEII8a+p9k6/IaV8RwjxmhDiDWAI+PfrPJ5yGziO5PhQgoot6Y54uTSZZUdLHR0zqysWKzZHBxIEvQb7O1ceg/b+UIJ82eZwb31tIsfwdJ4rsSytQTcT6RLbmgLzxscqylpJWa2rJcvhSG+ktrJhKl/h2FCCRL5MV72Pwz31tWweJ4aTZIsWh3vr8bj0ecf69olRoukSv3i4kwa/yketKHfa4FSO/niO/R0hGhascFis2Lx4apxUscL25gCxbAmvS+dgVz2tIU9tu1MjSdKFxee8otwqqwXTp4C/Ad4WQvwvUsq5kwPXnWdKSvki8OKC1/4E+JP1Hku5fS5MZHj9chyAl86UCftMzo+n+e2ntiKE4K3+KU4MVUfk1PvMZQPh/liWVy/GAJASHt/eiO1IXjgxSsWWXIpm2NFSx7mx6rFV6jJlvS5PZnntUrWuakLw8JYGAH50boL3BqYZmqpm+fC6dPZ1hhicyvHKhUkAbCl5akdT7VjHB5N8/d1hoJpm759+ZOcGfxtFUeYqWw7fOTGG5UhGEwV+5eGeee+/fjnGN46NUKzYeFwamhDU+02Gpwv8k5n80cPTeV4+Xz3nK7bDM7uaN/x7KPe+1TKRSynl/wc8C/xzIcR/FULMRk5q/PI9ymtev3MPeKr3W16XXsvH6Z25s9eEwO1avgp5XDqzw+y9ZnU7MfM6UMsf7DY0VApr5UZ45/Qy+ebUW6+p49I1NE2giet1el6dXNBDFfDo6DM3dLPpJhVFuXN07Xobs/B8BfC5DHRNVNsiQ8fQBYYm8My5FrhngmyY37Ypyq20phZDSnlJCPEI8G+A91dbGVHZ3Poa/Xz+gU4qtkNryMPgVH5e7/NDfREa/CZ1HhfNdZ5lj9Me9vKFw13kyzZbm6pLkGua4AtHuhhNFGgNephIF+ms96oFYZQb0hXx8YuHOylZzrzlhD+yp5WtTQHKlkO9z6R7ZqnglqCHLx7pIleyFi0/vLM1yB99bBfj6SLP7FS9V4pyp+ma4ItHuhlLFuhr9C96/9FtDfhMnWzZorfBx1SugksX87Ztrque85lihW3NKy85rig3arVguhbhSCkt4A+EEC9RHfrRtOxeyqY3N3je3Rac954QYs15fdtnxlnPFfS4CLa5AKj3m4veV5T16KxfPMzINLRF9XZWW2hxnZy1tzPM3ltWMkVRblbI6yLkdS35nhCCQz3XM+p2Lb0GC60hz7wx1Ipyq60WTP+rhS9IKV8VQjwAfO32FElZTaFsczWWpSPsXXMwWrYcLkUzNAfdy/Ym58sW/bEcXfU+Qr6lL15QnaB4MZqhZNm4dI2+Rj9vXZnC59Z5eEuD6mVWbjvbkbxyYRKvqfPo1qXrXDRdJJ4t0eA3OTeeocFvsru1jsuxLBG/uWJQbTuSixMZwj7XkjeEiqJsDCkll6JZfKaO26UxmS6xo6WO8VSBkuWwvTmw6PwvWw4vnhqjYNl85mAHXrMa6iRyZUaTBbY2BdSQD+WWWm058W8v83oC+Le3pUTKqr53aoyRRAGvqfOVx/sw9NWGvsNPL0Q5P57BpQt+87E+/EuMCf3OiTHGU0X8bp2vPL5l2QmBb1+b4rVLMU4Op9jTHqwF6poQVGzJkzvUQwvl9vrW8VG+caw6WdCyJU/tnF/nUvkKf/veMMWKzUA8x1SuTMRvsrc9RNl20DXBrz3Ss2h58llvXIlzfDCBJgS/+kgPEfUERVHuiGODCV6/HKdiO1i2g9c0ODaYYDpXBuCZXc2LVjb8q7cH+K8/H0ACQ1MFfv9ju7Adyd8eHaZQtjlXn+YLh7vuwLdR7lWrR2HKXadsOwBYtoOzxmmgJWtmH0diLbPT7HErtlxxdmnZcpASHClxHEm+bAPVfxcr9toKpCg3IVe2aj/n5/w8y3IcHCmREoozdd92JPmKVft5tr4vpTyzjyMllRW2UxTl9irNORfLdrVlKpSvtzOlJdqcfNlGSomUsnatcKTEmjmXZ89vRblV1JT1Tejje9s4M5ait8GPaaztfujZ3S00+JO0hjzLjj/7xL42zo2n2dIUqGU1WMojWxswDY2DXWGCXhc7WwO8eGoCv1tXaYeUDfGFw50IAR5D58N7Wha93xBw8/F9bcQyJZ4/0Mb7Q0laQm4e6I5wfjxNQ2D54U4AT2xvxG/q1PtNWoJqrKWi3CkP9kXQNYHP1PG7DcaTRfZ3BumP5ylVbB6YM2Z61q890oNlS4qWzT9+vA8Al67x6YMdDEzluK89tNFfQ7nHCSk3T4a7w4cPy6NHj97pYijKkg4fPoyqn8rdaGHd7P2D79/U8Qb+7SdutkiKAqjrpnJ3E0Ick1IeXm07NcxjA6Tyldr4rqXkShaTmeIGlujWKVZsJlJFNtNNmbIx0sUK8Wxp1e1S+QqJFc6POyGZL5PM311lUhRlddO5MqlCZdXtciWLyfSta3crtsN4qoC91rGXyj1FDfO4zcZTBb5xdARHSj65v31RnstMscJfvj1IqeLw5I5GHuhZOrdPxXZI5is0+M0bWikwXaxeXIKe5bN0rFfZcvirtwfJFC22Nvn5yH2taqlWBYB4tsTX3x2iYkue29u6bJq60WSBbx4dQSL51IF2tjStPQ9ssWKTL9tE/Ca5kkXFdpadULhQrmRh2XLJrDVDU3m+9f4oAL9wqKOWo1pRlLtTvmxRqjjEsyW+f3ocTQi+eKRr2SFa2ZLFX741SLFi8/j2Ro70RkjlKxi6wO82cBzJVK5Mvc+1pgn+AN84OkI0XaSv0c9nDnXcyq+nbAIqmL7N4ply7U51MlNcFEwn8xVKlepkiGh66V48KSV/d3SYyXSJ3W11PLe3bV1lGJ6eHxzMzSFt2dULUMDjWveqb4WKTaZoMZrIc2YsxeBUnl9/tJfgMmOyZz8vW7LWHPQom9N0rkxlZrJQNF1cNpiOZ0o4cvb8KK05mM6VLP7q7UHyZZtD3WHOjqXJlSye29vK/s7wou3zZQspwe82iGVK/O17Q1iO5JP729jWPD9neixbolix0DRBLFusBdOOI0kXKwQ9rhu6oVUU5dZL5Sv89bvVDqmmOjdSgi0l8WyJlqCHsuVQqNh4XTqFik3I62I8VSBbqmBoGtF0kSuTGb53ahwpJZ+9v5OzY2kuTmRoDXn40oPdq5bBcWTtKdxmfcqs3BwVTN9mu9rqmEgXsWyHQ12LJ0p01ns50hthKlfikS0NSx6jYktimeqJOpZc/4k6mSnWAvpoulgLph1H8udvXOPVSzGa69z87rPbF60Kt5KQ18WTOxr51vFR/LbDG1eq6Yu+9tTWJVPv2Y7k6+8NE8uUONgdVqvM3cO2NgU42BUmX7Y53LvMSgpUFwSKpqv1c2F6q5Uk8uVaFpnL0SzRVJFL0QzRTJE/+thuGgLu2rYTqSLfPDaMI6s3k5miVQv0x5LFRcG0oQmGEwVcukZ35PpKat96f5Sh6TzbWwJ8cn/7msuqKMrtM5Ur1Tqkwl4XEb+JqWvsbKmjULb563cGa9eLoMfFlkY/1+K5ajvUFebRrY2cHUuRLVqcHU8Tz5YJeV249GqgbdnOqr3Tmib48J4WLkykl7yZV+59Kpi+zVy6tmS2gVlCCB7f3rjiMUxD4+mdzVyOZpacubya+9pDTKSqwfjejuuzmEuWw8BUHtuRJPMVRhOFdQXTAA/0ROiq9/EfXrmCrmkIIZjKlpcMposVu3ZTMDKdX/f3UDYPXRNryuxiGhofua913cfvCHs51B1mOlfmsa0N/I93hwn7XDQF3ETTpXnB9HiqMCd4LnB/Tz3DiSDFSrVXe6HJTKl2HqSLlZneLslIogDA8HRh3eVVFOX26G3ws78zRKZo8fj2xnlPPWPZAplidQjIeLJIsNXFufEUHpdBW8jLoe56In6TQ931nBxOEvGZRPwm3TMdTjtb69Y8zGN3W3DZJ3DKvU8F03eZ5e6CD3aF2dNWh65df89xqvmgS5aNz1z+T+lx6Xxi/+KhIV6z+vr3T43THfFxcInAYi2agx5+90PbeeXCJPV+k876pVeM87sNHt3aQH88x8PL9MIrH1zOzNOT5YZQOI7EdiSaJtA1wePbGtE1gRCC33q8jx+encBvGouGUu1uCzKcKOA4kp2tdbh0jY+uEMA/0FNPMl+mzuOit6HaMy2E4OmdTZwbT3NA9Twpyl1D0wTP7q52WC3MCd8a9LC3PUg8V2JPR5CK5XCwK8SpkTR+9/VrRcBt8E+e3MKPzkap2A4f3tNC3Qrzi6SUlC0H09Dmrb64ll5s5d6kgum7yPGhBK9ditEW8vD5B7rm5Xr+2/eG+N6pcVrqPPzus9vwuw2+cXSEH5+bQEp4dncz//OHtq/7Mx/b1shj21buGV+Lpjo3Xziy+opSD21p4CEVSCsLTGaKfPPYCALB5x/opKnOPe/9qWyJv3x7kBPDSXa11vHwlgZODqcI+1x88UgXZ0dTjCYK9DX6cenzg3GPS+dTB9r54dkJ/uvPB9jbEVrxaVHEb/KLS6yOdqArzIF1DEVRFGXjvHk1zjv90/Q2+vjMwQ6EEJwbT3NuPEMkYPK5fe04UvL1d4dIFSw+tq8V15zA123oPH9g9eFb2ZLFn75yhZPDSe7vrud3ntmK29D5+2MjjCYLPLmj6YaeICubm7qFuotcmsggZXUcZ3pOap+SZXNqJEXZcohmilyNZRmcypMtVoimixQrNseHEnew5IpycwbieUoVh2LFZnAqt+j9wek8k+kShXJ1qNA7/dM4UjKdKxPLlLgYzQBwLZ6rrZi20MWJzMz/07fviyiKckfMnt8D8XztGnA5msWRknimxFSuxGS6RCJfwZGSy9HsDX3OWLLA8HR1eOTgVI7xVJFMscJosjr869LMtUj5YFE903eRB3rq+dmlGIlcmb98a4DDfREe3dqI29B5emcz6UKFtpAHx5G8ciFKIl/hvvYQJcupPbZO5Sv8/fERbEfymUMdi3r4FOVutLO1jkvRDELAjtY63rwS5wdnxsmWLJ7c0cQzO5vZ2VpHvmLjcemkChUqtsPj2xtpC3k40hvh7f4ptjYHlk3P+FBfhNOjKTVBSFHuQQuvAdmSxbV4jjOjKZ7e2URznQdHSvoa/UznyhzoCnFmNMWrFyfpivh4fn/7skPMpJR8//Q4A/EcD21p4GBXmKODCQ5119NV78OlC/a0Bxmezqte6Q8oFUzfRba31LG1KcD//fJlbAmnRlI8urU6BOPDe1pqj6b//I1r6JpGU52bf/H8nnnBQ388W0tYf3kyo4JpZVMIeV38ysM9tX+fHEkxliqSLVoMxHPY2yW//mgvvw78lzeu1Z7cfPS+VgxdY29HaN7k2qWoIUaKcu9aeA2YfcK1tyNEV8SHrgl0xLwc0K9dqubC74/lSBcry6ZszZasWk/2hfE0X3liC195Yv42K83DUO59GzrMQwjxkBDiTSHE60KIfzfzWkoI8erMf8vn0PqA0DTB/s4Qxsz/l3Kgq/r+7ra6Rb1wW5oChH3VnNHbF6T8UpTN4kBniPaQl/awh77GwLxG7kDnbP0PqkWCFEVZUk+Dn4aAic/U2dW6dJaN/Z0hXLpgS5N/xQXNAm6D7S0BXLpgn3qypSxho3umB4EPSSmLQoi/FkLsA05LKZ/e4HLcNhXb4cXT46QLFT56XyvNy6zANDiV45ULk7SGvBzpreelsxN4jGp2jWd3t9RmJ88qWTYvnh4nV7LZ0uTHZWgcHUgQz5bQNQ3LkXz0vhaODSR4fyjJ1iY/5iqziq9MZnn9cozuiI8P7WqeNytZUW6WZTu8eGaCVL7Mh/e00hqqnguTmSLfPTnOlckMu9uCfPpAR20lwqMD0/zNu0OEfSa/88xWWoMeXj4/yV++NcATO5rY2hRgd1uQ/liOZL5cW0QlU6zw4ulxAD62t5V3riV49eIk9T4Xzx/oYGerurFUlHtZvmzxvVPj2I7k43vbCPlc/NojvQD8yQ8u8NrlGD0NPrY0+cmXbA521/Pxva2Lnmj99EKUt65OYeoaj21v5NGtjQgh1pVb/sJEmreuTrGlKcBTO5qA6lCRn5yfZDSR56mdzfQ1+lc5irKZbGgwLaWcmPNPC7CB3UKI14GfA38opdzUC9sPTuXpj1UfL50YTi6bQ/e9gQSJfIVEvkLJspmcWf3waizLfe2Le6QH4nkG4tXczGfHUjOTH/LkyzZly6aj3sebV6Z4q3+KaLpItmRxdvz6MJGlvHttmmS+QjKf4nBPZMmllRXlRg0nClydrD4aPTGc4LlQNT3jqeEUl6IZBuI5DE1wdizFozMZZV46O8FIosBoosD7Q0ke2dLA6dEUAO9dm2ZrU4ALE5naZJ/zY2ke2tLAhYlMbUGj9wYSvD+U4MpkljqPQUtwSgXTinKPuxTNMjqTC35u2zeRKvDKxUkyxQrRVJFUvoLlSOo8BqPJED0N14PaVL7CyeEUF8YzuAwNQ9d4sDey7nR37/RX29bjgwmO9NbjMw2mc2XOzLmWqWD63nJHsnkIIfYDjVLKc8B24EmgHnh+iW2/KoQ4KoQ4GovFNrik69ca8lDnMdBE9dHRcrbOvNcYMNnbXn1s7TN1OsJL52huC3vwu3V0TXB/dz1Br4uGgEljwKQt7EWfeey9rSmArgma68x5F4mlzObYbA15CHjU8Hnl1moJumvnQl/j9dzPfU1+6n0ufKZO2GfSM6dROdAZxm1oNATcbGnyzwTD1R7t2fraFfHidmmYhlar490RH6ZRfW1nax1tM+dhxG+ueyEiRVE2n656Lx6XjksX89q+xplriQDawl7aQh4aAiZBr7loTlHAY9Aa8lDvN4n4XPQ1+m8ob/Tstaqj3ovHqA5FC3pdtc/b2qyuSfcasdEdwTPjor8NfGFuT7UQ4mPAISnl/77cvocPH5ZHjx7dgFLeHNuRWI6D27g+njORK/PqpUlCXheP9DXw04sxypbNR+9rxec2KFsOsxOJX74wSbFi86FdzfMSx889brFio4nqYhKaELw/PM2f/vQqxYrNY1sb+MT+DvpmAvY3r8YZSxZ5bFsDbaH5wXqxYmPq2rxZzO9em2ZwqrqwyuzS45uBZTu1390zu5pXHAN3Oxw+fJjNUD9vlyuTGf77mwM0+N187amteE19yXMBqsOWkNX6axrVxur4UIKrk1nuaw/SHvbwxpUpBIIP7WxCaKI2Pvpv3h3inf4p2sIe9nWEeXZ3M25Dry3YYDuSn5yLUrZsntnVQr1//qSid/qnGJrO8+i2xmVvXu+UuXV44fl/MxbWzd4/+P5NHW/g337iZoukKMD6rpv/1w8v8O61aT7/QCdfPNK96H3LdnAkJPJlXr8cpy3koavex9v9cSJ+k6d3NmM5EpC4dH3eWg5QHab5k3NRhhN5XLrGfe1B3IbOxZnVh1e6MY9lSrx2KUZTnZsntjdSshzcCxZ1cRxJ2XZuaq5HoWzz8oUoAsGzu5vVvJHbTAhxTEp5eLXtNrQ7UghhAH8F/DMp5YQQwg8UpZQ28BhweiPLVPhPDAAAIABJREFUc7vomkDX5lfwd65N14ZpFMo2l2ZmBp8dT3OkN8LxoQTRdJEGv5vvnxrDsiUelz5vhvDc43pcOieHk/THsxzpjfDXbw1xYSJNpmgxlS0zni7xx5/cQ65k8U7/NABvXJaLFqNYeCKmixV+fiUOwOuX43z5ocUXrLvVlViWc2PVHMIhb4Knd66+nLVy6/z9sdGZep3lQHeYZ3Y2L3kuAEgJPz47wfmJ6rjpJ7Y38t0TY4ylClyZzPLhPS212fOtITcP9FTnJidzZV44MUoqX+HEcLI2dvrLD/XUFmA4M5rg8szwksuTWR7suz6vOZWv8ObVKQBevxTjlx68u+r35cnrdTjsS9bGWyrKB93wdJ7vnhzDdiT/7c0BPnd/57xe45Jl88qF6tPrVKHMWLLI8HSekyNJShWH0WSRR7c1LlotOF2s8LOLMQIeg0a/yYWJDK9djqELwch0HtOlYeo66UJlxWD6rZmb9KHpPNtbAoS9Jj84M4FL13h6ZxOumU4rzxLXw/U4M5aqXRvbwh7u71ap+O4GG/1s/xeBI8CfzNyt/SHwp0KIHNAP/MsNLs+G6Qh7OT+exuPS2dYc4Gosh5TV5U4nM0XemmngL06kGU8VSeQrvHBitJpLd1vjvGwG7w1MMzJd4PxEGq9LJ1u02NIc4LXLcRwpqdg2F6MZvv7eEF96sJug10W6UKF9Db1wXpdOvc/FdK7MVK7ES2cm8Jk6p0aSNNa5eXZXy4rp9t7pnyKeLfPYtgbCPpNUocyfvdbPQDzPx/a24Hbp5Es2UlS/+61cVrwp4MalC67FczjSYTJdZF9nmN1t82dynxhOMjydp7HOJJYpc7AzTHfD8j3wZ0ZTXIvnONIboTXk4f2hBC+cGOP+njCfOtCx7H4fJBOpIplShVzZojHgpmeJJxqxTIm3+6fwunROjSY5P5YhXawwnSvTGDCZzJRI5iuM6QXCc+rg49uv1xGf26C5zkOhbFOo2ETTJX50doKdLXU80FsNmluCHnRNICWEPC7+4fgI/bEcD2+N8GBvBF2Ds6NpYpkS9f4JntrRVLupjGWK/OfXrxH0uvjak1voj+f4+rtDbGsO8KUHu6nYktcuxZjMFPG4dPZ3htjWXMepkSRnx9I4jmRHax1HeiNMZoq80z9NZ72XQ2ts8Jrq3JiGxuBUjksTBvs6Qowk8gxO5XmwL1Ib8rLc73Y9n6Uom0HFdvjOiTHOjqVw6RqFcgWXpjGdLxP0uHjtUowLExkmkgViuRI9kWoWj8vRTHW4WKOfH52dwDQ02kOemblKJdJFC5cuMHUN26kOw3hsWyMlyyZbrFC2JOfG03z0vjbGkgVyJYv3BqY50hvh6MA0E+kij2xpoCFQbQ/bQx6uTmZxG4JjgwkGp3Ik89XFXKLpIr/8UDdCCDLFCm9cjhP0unh0a8O6J/63Bj1oQiAEy14PlI230RMQ/wb4mwUv37+RZbhT9nWG6JwZ0+U1ddrDXiQQ9LgolG38bp1cyWZfe5iBeJ5i2WZ4Os+xgerKhrMziSfTRd64HEdKyVS2RGe9j+agh94GPx/d20KxbDOSLCKA/liOVL7CrzzcTbZo1U76lbh0jS8/1MOJ4QQ/vxLnvYFpxlMFprJlIn4TQ9P4/AOdS+47lizUev0APrG/je+dHOcn56JkSxYTqQK724JM58r43QbtYS/dEd+agvy1aAi4eXZ3Cy+8P8qpkTQD8TzjqRLbmgO1XstUvsIrFyZxpOS7J6u9opPpIl95YsuSx8yXLX5yPoqU1R6MX36oh//yxjVimRIXJtI8tb2J0DK5ST9IXrk4SYPf5HBPhN96rJfuJcbrv3pxkpFEgdOjKZrqTOLZEi5dw+82aA15+djeVk6OJOms99JV76POYxBw65weSXOktxpQm4bG//HZvYwkClyNZfmLtwaxJbx+Jc79PfUIIWgPe/nNx3qRwImhJC+fnySeLZHMl+mJ+In43fjdBldjWYJegzqPUZus9LfvDXNiOAnAjpYAr1yIcTWW5exY9QlSumhxejTF0cFpOsJexpIFfv3RasaRixNpKrZkMlOip8HHqxdjjCaqPe1bGgNrmuDbGHDziX1tfP3dIbIlix+fm6hNrMyWLL60TE/67O92PZ+lKJvBmdEUPzgzTqZosbM5gGFo9ESq51dvg5/XL8c4N5YmXaygC4FA0Fnvo7Pei8vQODOWpGw5TKZLfOv9UVLFCtmiheVINAG9DdVx0Q/2RdjWHGBLo5+pTImr8RwtQU81Facuam1vwG3w+uXq01vLlrW81Yd7I2xpCnBuLMV7Awmmc2UG4lkqjuRaPMvwdIHuBh/v9E9zYWa1xo6wl951TkTsivj4zcd7EXDLhoEpN0/NOttAc8duzj0JvKbOrz7cS6ZYoTnowevWefXCJP3xHG6XRoP/ehDsdxt4XNUx088faGdXa5CSZfPy+ShuXae10ctDfQ30x3OEfSZ+t4Hb0HEH1v5oyTQ0ehv9vN0/jalLIj6TdKGC16XTEFgcODqOpFCxqfMYmIZG2XJq23XWe3G7qjcKDQE3XlPHZxn4TB3T0G75xMfWYHUypXfmpiXsc2HMGRfnMTUCboNsyaI5WP29Nq5wk2HqGnWeas/+7N+hNeQhlikR8rjwmuoUAoj4TSZSRTrrvbQtc3PUGHAzkihQ73PR6HfTucvHJ/e1Efab1PtNeiI+DvdGqPe7MHWNlqCH6Vx5UZ3zu13sbHWxo6WO8VSR0UT1Zu9SNFvL2jF7fs3mmRUC6rwugh4XLUEPYZ+LeLaMoVXPr0LZxqULOuurZdc1QVe9j/awh6uxLH53dTKjS9cQAnxmtY5F/G48Lp06j4HPNGrjIQNug8aAyWiigN+t4zHXPompOegm5DMpVmxagh4yRYtM0aLBv/xN2+zv1u/WcbvuyLxyRbktGvxufKZBpmjRUe8jEjApVhwaAyYRv1l7qiQldDR4+dSBdhoCbhL5Mm6XRnfEz7V4HtOwCflcOFLiOJKS5cxMSqwOlXj+QHvtuvEbj/Xx7ROjBNwuGuvc9FX8xDIlPC6dpjoTr6lTKNuLrk0Rv0lryIsQCRoDbvZ3hrgayxJwGwS91bYiMrOPSxeEvDcWDG/0fCBldRs+AfFmbJYJiLfCeKrA7J9mYc9ttmSRKlRqk6f+46vViYeWLfnKE32EfS7GUkVC3uriLTcqmS9Tshx8pk40XcRrGrSHPIsmVHzz2AijyQIP9NRzqDtMpmjNK/OlaIapbJHdbSFsRyKBsuVQ5zFuy511qlAhU6wgZfWx+cJx4fmyVRtaEMuUaQt5VpyxXSjbTOVKtIe8aJqgbDmcGkmyrXn+YiIf5AmIjiMZSxVo8FdvmJYipWQ0WSDsdZGcuTlZbluoTo6NZUor/n0qtsN/fr2fXMnG0AX/5Ikt+BfU+cl0kel8mfawl6DHVSurJgSGLpjKlvnh2QkCboMvPdjNQDxHwGPQ0+DHdiQnRxLVJ0B11Ueq8WwJy3GoWJKWoAfT0CiUbeLZEgio95kE3Ebt+0b85qJxmquZPcfbQx6KFWde/Vvpd7vcZ6kJiMrdai3XzVi6yES6WMuCMZ0r0xH2IoQglinxzWPDjCeLGIbg957dTshrMpYszDzdMnh/KDnT3hgULYdyxSGZL/GT85NUHMnn7u9clG86limha4KI30RKOa9NnXt+LjVMY+6+46kCPtOYFzhPpIp4Tf2Gg2ll49yVExCVtVuYdWOugNuYFyTXeQyKFZu2sKfW+30rshTMDRSXC3pLllPL+dsfy/LkjqZF2+5oqYOWjcvzG/K6VrxI+UyjFnB0RVY/BbymTqd5fQywaWgc7v3AL9Y5j6ZVH62uRIjr2wTWcBPlcemrZpNx6dX0eBcnMnhcOoa+uGFrDnrmLZ60sKzHBhJICZmiRSxT4r45jWo1FeX8v/VSTzK85uKyzv2+6zX3HF9Y/5ZyM5+lKHe7pqCHpjnn8NwbxqY6N32NAYqV6lOh2cxBczt07u9ZPI9gcErDaxp4qU5uXBhMz50bJISY16YubIMXlXfOvku15bMLWCn3DhVM3wM+d38no8k8HeGNb0y9ps6DfRGuxrI81HfrJhMqylp9ZE8LO1rqaA66F6XgW4v7e+qZzpcJe83aMA9FUTaPD+9pYVtzYMknkcvprPexo6WORL68ZLCtKOuhgul7gNfU2dZ851Z4e2xbI49tW36lRUW5nQxdqy2ScCNagh5++aGeW1giRVE2kq6JdV8DdE3wif1tt6lEygeNmqmiKIqiKIqiKDdI9UzfAqdGkrx1dYqtTQH+0Z4WoDpJ6bsnxzANjefua+XH56Lkyjaf3N9Wyw35/7x8meNDCT6yp3XR4ijv9E/x/nCSfR1BMsVqto6Ax+BTB9rRhODt/im2NQd4dnfLusr65pU4p0ZTHOoK89AtzPGsKBcnMrxwYpSJdJGHtzTwmYMdtdUNoTpJ7t98/xyvXoxxqDvMv/n0Xn7v6ye4Esvy+Qc6+e2ntyGl5IdnowxO5XhiexN72q/nCJ89z26k3iuKoiz00wtRLkezPLSlgVShwvnxNEd662uLRK3H65djnB1Lc393mFimxPdPj9Me8vJbj/etmP41mi7y/VPj+EydTx/sqE3KTuUrfOfkKBL49IGOJdNdposVXjgxhpSSTx1or81zuhrL8vL5KC1BD5/c375opce7xc8uxW7qd343UT3Tt8DxwQT5ss3p0RSFsg3A+fE0yXyFyXSJt65OMZ4qki5UODuWAqoZJX5+JV5dGvR8dNExjw4mKJRt3rw6xenRJEPTeQbiOY4NJjg283mnRlIUK/a6yvreQPW4RwcTN//FFWWO94eqCxWMJgpciWZrE1Nn5co2r12KU6zYvDeQ4OjANGfGqnX4xdMTAGRKFufH0+TLNseH5tfRm6n3iqIoc5Usm5PDKfJlm/cGpjk+0+YeHVh/2+g4kqMzbetbV6d4u3+aqWyZK7EsZ0ZTK+57bixNqlBhPFVkcDpXe/3yZIZ4tlxN+zmZWXLfy9Es8UyJqWyZixPXtzk5nCRXsumP5apZhu5CtiNv6nd+t1HB9C2wuy2IELClyY9nJsfr1qYAbpdGncfg/p56gl4XpqGxfWZss880arOH5y53PGvPzKp9BzrDbG0KEPGbNNW52dMWZE979fO2NgdwG+v7E8729O1ZsCqgotys3W1BGgNuwl4X7fUe2hbMWPebOvu7QoBgZ3Mdh3rq6Yr4EULwyMxTkoBp0B3xIQSLVq6ce56tt94riqLM5TaqqxELAfe1B9k1k6N+7tOwtdI0UbteHegKc19HEK9Lpy3oqWazWsG25gCmoRH0uuZl5Olt9OMzq+sl9C2zsEtvg+/6Nk3Xt9nVGkQTgtaQh8gK+envJF0TN/U7v9uoPNO3iOPIRTlgZ3+3s3kol9rGshyMZQKDuds7TvVYc/+9XM7ZGymrcvM+yHmmZy2sp0spl23MOfmlF/579jhLHUPV3Ruj8kwrd6s7fd1c2M7ezPVlpTZ7rfvNtTCGWMpy22yWa+XdXk6VZ3qDLVUZFlbupbZZLpBeuP3CfW+m8t3NFVfZ3NZStxYGzgv/vdJxVN29N6hgXrlbrNTObtSxltt2pSB6tW02y7Vys5RzNepZqaIoiqIoiqLcIBVMK4qiKIqiKMoNUsG0oiiKoiiKotwgFUwriqIoiqIoyg1SwbSiKIqiKIqi3CAVTCuKoiiKoijKDVLBtKIoiqIoiqLcoLsiz7QQ4t8Bh4HjUsrfu9PlURRFUZR71c3k+VY5vhVlsTseTAsh7gf8UsonhBD/UQhxREr53kr7pPJl/uV3zjGVLXFfRxC/abC9JcDVWI6A2+CnF6K8fXUKKeBwd4TOsIdvnxyjbEvchuBDu1owdcHJkRSGBkOJApYlMXSB29AIewzG0iUcCS5dIKXEcmA9a0VqgHNTv5mbI4AjvfX8+186hNvQ+MpfHOX8eBqXJmjwm0xmShi6xpZGPyXLYXAqB0KwvzOIhmBgKo8Q8FBfA//rR3fSHvYymszzj//bUSbSRZ7Z2cQXj3Tx3rUE7fVe+ieznB5L8ciWRn7r8V7chs5LZ8a5FM3yyNYGjvRWl0x/8dQ4/+PdQTrqffzxx3fxg7NRTo0kSRUqWI7kyR1NdNV7OTaYZEdLHc/tbeWVC5OcHk1xqDvME9ubALgay/K9U2MMxPP0Nvj45IF2tjYF5v0OTgwnef1SjJ5GPx/Z08w/HB8jkS/z8X1t85ZnnUgVeeHEKKah8fkHOqnzuAB4b2Cat65OsaMlwHN722rb3/cvXiJXttGA03/8NH7//KVe3702zdv9i/eb61o8x4unx6n3mXzugQ7cxvyFS7JFi3/xwhli2RL/+PE+nt7ZXHuvYjv8w/ER4tkyH9nTwvaWOoan83z31Bh1Hhef3NfG198b4u3+afZ3hvjqk1tq3wlgJJHnX3/3HMeGEkgpqdgOJUsS8ug8uq2J3gY/JcvmaixHW9jDQ30R/uy1fgan8jNlq1CZqdwuXdBS5yZftkjkrXWdI7eaYPE5KgBNVJcO9rg0HtoS4UosS/9kDgk01Zn0RPx01nuJpktcmMjQ1+jjX31qL9taAvzD8VFODidJ5sucG0/jdek8t7cVXdNq9XPWbH1pC3mIZ8v89HyUaKbEjpYAv//cLn52KYaU8Nn7O3inf5rvnBwj7HPxq4/0cF97CIBzYyn+wytXsB3Jg73VspYth6d2NvP8/rZlF2hwHMl3T41xcTxDybZpC3n5hUMdNATct+V3fafd7KIvN0sFlLeelJLvnBjje6fGaaozeWRrIwPxHOlihTcuxxhPFfEYGrmyTaHioAnwmDply6FScdA0sCTMLupsaOA489thfeZZvD3zopjZbva1tbTZN9K264C9zHsuUT2eveDitdT1bOH7AA0Bk3qfyVS2RKZYQQiBx6Xh0jV6GvxE00UmUkVsCaYuONxbT0fYR1+Tn3S+wk/OR4lnS+TLNroQeFyCbNnBdiRel86etgBtYR99jQEaAyYXJjKcHUvRGvTy0b3N/IefXiWaKhLw6OxtDwOQKVk4UtIUcGPogulchS8/1EXE7+bbJ0a5PJFhLFXE49J4sLeBJ3Y08t2To7xxZQq/afBHH9/NZw51APCzSzFODifZ3xma1w7O9aOzE7x7bZqTI0kEsLO1jv2dYX7hUAeGvvQAjJfPRzk7luZwTz2PbqvWte/PtMmfvb8Dj2vxYmIruePBNPAI8JOZn38CPAzUgmkhxFeBrwJ0d3cDcHo0xXiqQMV2ePfaNE9sb+K1y3GaAm4uJtKcHU1RcSRSwuXJNMOJHOWZmlqyJCeHk7hdGhXbYTRZwrIlEijbEkfaFMp2rWKXF9bwNbqTgTRUT8Lz42kuRjPoAq7Fsli2Q9mCsl3EsiVl22Fouvq7KVkOQsCF8Qw+0yBZKOPSNC5PZrg4kaE97OWtK1NE00Vs2+H9wSQ9ER8SwbGBaaayZUoVhwsTaaKpEi0hN+fHMwCcGknVgumfXY5RsSUD8Rzvj6QYSRQYSxaJpov43Qb9sRxjyQICwfnxNM/sbOLUSApHSk6NpGrB9PnxNIlchbFkgbDXxfnx9KJg+vRoCsuRXJ3McrWpelGZ3XduMH0pmiFftsmXbQan8uztCNXKbTuS8+MZntnVXAt4c+XqZdEBfnBuis8fmR9MnxpJLrnfXOfH05Qtp3ah62mYf4yzYylGkwUAXr0Ym3cRiWVKjCWr3+XceJrtLXWcH09TqjiUKiVOjSS5OJGhWLG5GsvO+06z3/dStPp+oWwjRLUhyZbg+FCCkNfFhYlMrR69eHqC6P/P3n1GyXWeB57/vzdUjl2dMxo5EiACCTCJpChSomTlOFbw2iN57Q+emR2f2Zld79kvu2OPd9dh5oxn5Z1jeVZhLUuyEiVRpCSKpEiCJEAQOXbOXTnXrRv2QzUK3ehuoNFo5Pd3Dg+7u27dfrtwq+5z3/u8z5OtUDIsDNOed2xXLYeJTBlFiFsaSMPiJx6H2kmqVLWwcTgyXLtwu/i2TuQNPJpGsWoxmSljmBZDySLvjKQJenQmM2XG0iXG0yVSBQPTo/OLU9M8ubmFUxNZntjUjGv2bHzxeHlzIImuCkZTRaq2w3CiyIunpygbtVfu1ESWd0ZSpIoGRcPkxHi2Hky/PZgikTcAeOnMNJqqUDFtLkznyVfMeRdFc+XKJv0zBWbyFRIFg4Bb58JM4YYF07c6mJVunRvVvbJctTkykiZbrmJYFi+fm6E97OWN/gTxvEHRsChUTCwHZjt1Uy2b4My+zy876ZqLnIQv38aB+sTAcq3k3L5UIA1QXeKD82qfpxcfT+QNqpZNoWLO/s0Opm3h1mrnvmJ1fjxzdjKHZYNHVzk9mSVRMMhXTCwbDBxmX1IAiobFZLZCtmwR9bk4PZmlZFjM5Cq4VIXvvD1GplilZFo4ZYejYxnawx6mcxVcmkLJsKiYFg1+Ny+fjbOuOcBkpsxgokDFtClUBP3xPFG/zjvDaaqmTdaq8vLZ6XowfWz2fHp0NLNoMG1aNifGswwni0xly2iKgj5ToMHvJlEwaAl5FjzHth2OjWVwHDg6luHAukZOXnZO7m30L3jeldwOwXQEuDD7dQbYOvdBx3G+CnwVYM+ePQ7AfV0ROiLe2ZnpMC5N4fGNzZybzrGtI8xUtsyr5xOAw6bWMN0NXv7x8CiG6eDVBbu7I+iawjsjadbEfAwli1RNB02rzUw3eHXGMmUsG1yawLZXMDMtLr3hbwUF2N4RYlNrELeqsK4lyImxDLqqXJqZVgS9sQAV02YwUQBga3sIRQgGEgIBbGoNsaktCMBD6xppD3uZyJTY01u7mntrMMW+NTEG4wWOjKbZ2h6iNezBpSlsaQ9xbirHzq5LgdwTG5v5enqIrqiX3d1RprMVUkWDoFfDtBzWNQfobvDx9mCSDS1B3LrKfV1hjo9l2NkVqe9na3uYgXiB7gYfDX5XPRiZ677OMK+ci9MT87G2yU9b2EOqWGVLW2jedptag5yZzOHWlXlvoJ1dYV6/kGB9S3BeQBx0q+QqFoqAT+ztXvB7d3ZFZmemg4sG0hdf56FEkahPpzW88M2+tT1Md4OPqVyZJzfP/wBpCrrpjHqZyVfqf/fmthD98QJBj8bO7ginJ3NkSlU2tAQXfCjU/k1DFAZT+HSVimlTMW1CHo19vQ3EAm729mqcnynQFvbwYF8D4+kShmnhd2vkSgazcSEuVdAS8lAyTJLF6i095peamVYFuHUVj0tlT28D52dynJ3M4wCNATdtEQ/dUS9Rr4tTk1nWxPzs7onQGvbQHvEQz/sIe7T6TM37trSgKIKNraF6IA2Xjpf9a2PE8xV6Yj6mshV6G/28b1MLvz4XxwG2tIfJVUzG0iUiXhfb51zo7FvTwJuDydrM9JoGzk/nqVo2G1qCBNxLf1wHPRrrmgNYtkPEpxPy6qxrDiy5vXR9bnU79LvxYsajK+zpjTKRKdMYdHFgbYyBeJGH1jXy8tkZTLt2/i5UbIqzkwA+l0bFrF3kq0otgL74GbTYzPTFt6s5Z2ZaVy5ddDvO1c/zmpjddpl/l6B2Pr7izLRYGOgvd2a6KVibmY7nK+RKVYQi8OgquirojfmZylYYT5ewZ2emN7WF6Ij66Ih6CXlUMsUqAshXTFRF4HMp5MoWpu3gc6m0hT20R2rn2Q0tQc5OZskbJo1BN+/f1spfv3ieimUTdKvs6IxgO+B1qdiOQ2PAja4qJAsG79nYRIPfzUC8QKFiMp4q4XWprG8OsKMzzEiyyK/PzhDwaLxn06Vz3s6uKEdGUuzojCz6OmiqwvaOMKWqRbZcBajHETG/a9HnKIrgvs4IJ8Yz3De737nn5LbIwnPy1QjHuYVnP0AI8YfAjOM43xZCfAzodBznrxfbds+ePc7bb799cwcoScu0Z88e5PEp3Y4uPzbvxmBMujNcfiEhPzel25kQ4pDjOHuuut1tEEzfD3zFcZyvCCH+M/A1x3HeXGzbxsZGp7e396aOT5KWa3BwEHl8SrcjeWxKtyt5bEq3s0OHDjmO41y18t0tT/NwHOewEKIshHgFeHepQBqgt7f3triCff1CgoF4gf1rY/Nyb6V722rOsNi2wy9OT5MsVHh8YzPNi+R9SdJyXTw2CxWTn5+cRCB4emsrXte1LbKRpNUmZ6al25kQ4vBytrvlwTTAnVQOL18xeaM/AcCr5+MymJZuiLF0ieNjGQDeHEzywR3tt3hE0t3gxHiWwXitKsvJiSy7e6K3eESSJEl3Ptm05Rp5dZWmYG2FfHeD7xaPRrpbxQKu+oIzeZxJq6U94kFTBLoqaF/BIhtJkiRpodtiZno1mZbN94+MM5kp8dSWVja2Bld1/6oi+MzeLgoVi7Bv8TJVknS9fC6NLx7opWxahJYohzbXSLLIj49OEPRofGJ35zXXyJTuDV5dRVUEihB4lqg0I0mSJF2buy6YThYMRpK125jHxzKrHkxDrRRL2Ccn9aUby6Up80qvXcmpiSzlqkW5ajGaKrKuefWPe+nOd346T2W2Llh/PM9uf8MtHpEkra5bXbZQujfddRFhg99Fd4MPl6bMa1QhSXezzW0hvC6V5pCbzqhMC5EWt74lSNCjEfbqC5ocSZIkSStz181Ma6rCx3d33uphSNJN1dXg4/cfW3urhyHd5hr8Ln7vkb5bPQxJkqS7yl03My1JkiRJkiRJN4sMpiVJkiRJkiRphWQwvQInx7P89NgE09nyrR6KJK3ITK7CT49N1GtZS/cGw7T51ZlpXjozTdWyb/VwJEmS7gp3Xc70jVY0ah3EHKe2Mn5Da5AtbSG6llEL2LRsjoyk8bpUtrbLxZHSjXNmMke6aLCzO4J7kRJovzo9zVi6xOnJHF0NXk5P5CgYJgfWNsqyenexQ0NJ/u7VAYSAgEtlz5rYrR6icMseAAAgAElEQVSSJEnSHe+eCabPT+eJ5yvs7IpcV7CgqwoBt0aubHJuOo9pOwzEC8ta/PXmYJKD/UkA/C6NXtk9UboBJjIlfnJsAoCiYfH4puYF20R8OmPpEn63yni6zGsXal09VUXhsQ1NN3yMx8cylKsWO7siaKq8QXazvDWQ5NREFoBDQykZTEuSJK2CeyKYjucr/PjoOI4DmVKVp7e2rnhfuqrw2X3dxPMVXjo7QzJvLKupBoCmXAoaVEWseAySdCWqIhACHAc0dfHj7L2bW9jYGiQWcFOsmChCYDsOIc+N/0gYiBd44eQUAFXLYf9aGdDdLC5NqX/26LJpiyRJ0qq4J4JpVQgEAgcHbRWCWL9bw+/W+PQeD2PpEh0R77Ket6cnit+t4nNpy0oLkaSVaA56+Pj9nWRKVTa3hRbdRlEEPbHanZGAW+NzD3RTrlo35bic+x5cKtiXboxHNjRxZiqHEIKH1jXe6uFI0m1HNn2RVuKeCKajfhef2NNJMm+wqW31OsN5dPWaGh8oipC50tJN0dXgo+satm8Kum/YWC7X1eDjI7s6KBkWm25Ah1JpaVvbw/z371mHoNbARZIkSbp+90QwDdAR8S57BlmSpBtrjVwvcMtskEG0JEnSqpIrfyRJkiRJkiRphe7amempbJk3+hN0Rr3s7mm41cORpNtW/0yeo6MZNrUF2dS6eI61dHcoVy1eOjODEPD4xmZcmpxPkSRJul53bTD967MzjKVK9M8UWNsUIOJz3eohSdJt6cVTUxQqFsPJIhuagyiy0sxd6/hYpl4aryno5v7u6C0ekSRJ0p3vrp2WuLigKujR8LpkCShJWkpz0ANAY8AtA+m7XGPAjRCgCEFT4OYtOpUkSbqb3bUz030xP2GPzpb20KId4FZD0TAZSZbobvDJgF26Y31wRxsz+Qox/9LB1UiyCNQqcVi2Q/9MnljATYNf3vG5k/Q2+tnWHkIRQpbnlCRJWiV3ZDDtOA6vno+TKlZ5bH0TYd/8pilnJnP1DnBhn14vX3dhJs9rFxL0xnw8sr7W5e38dJ43B5Ksaw6wb8215VZ/59AoibxBU9DNbz/Yswp/mSRdnW07vHBqini+wuMbm2m/xio16aLBCyen8Ls1ntrSgq4qtIWX3sf56Rw/erf2fvrgjjaGk0WOjmZIFgy2doR4eF0jnVEZmN0JvvnGIH/6szMA/C8f3MIn9lxLAUVJkiRpMTc1zUMIsU0I8ZoQ4hUhxN+Jmr+Y/f6vlrufkWSJtwdTXJjO83p/YsHj+YpZ/7pYsepfv34hQTxX4e3BFLlyFYBXz80wlS3zm/NxylVrwb6upDC778Kc3ydJN9pktszJ8SzT2QpvDSav+fmHh1OMpkqcmcwxEC9cdfvCnPdQvmJSMCxMy+bsVI7+mQK/PD19zWOQbo3njk9SrlqUqxbPzU44SJIkSdfnZs9Mn3Ec5wCAEOLvgH2A33GcR4QQfyOE2Os4zltX20nYp+PSFAzTpjlUuzU9GC/w/IlJGvwunt3eRqVqIYRgS3sIw7QxbZs1jX5mchVaQh58rtqf3h3zkSpmaA17cKnXdm3xofvaODOZW7TLXKZY5fkTk7g0hWe2teLRZRqItDoa/C7CXp1MqUpv7NrrNXdFfRwdzeDWVJova9Zi2w4FwyTouXS3Z1tHmKJh4eCwozPC2uYAfl3FcpwF+5jOlnnx1DQRn87TW1vrrasB3h5M8uZgko0tQZ7c3LKCv1y6Xs9sbeHQ7AXYs9tbr3t/6aLB8ycm8egqz2xrvWEpdZIkSbezmxpMO45TnfNtBXgv8OLs9y8CDwLzgmkhxJeBLwN0d3cDEPbqfPFAL0XDrC+eOj6eoWhYFI0SiYLBgdlWufmKyTcPDlE0LJ7e2so/f7QPr67WT/JPbGphd3cDAY+2YPGV4zgIsfSCrM6oj5FkiSMjafxujbD3UgBybCzDWLoE1FJJtnXcmM6Hx0YznJnKsas7ck3dGKU7l0dX+cL+Hiqmjd997W/h9S1Bfi/i5dhImlfPx3mwL0ZjwI1tO3z77REmMmV2dUd4z8ZmAFRFsH9trP58v0sDAeuaA+zojLBuznF3eDjFVLbMVLbMlrYQvXOas7wznKZStTk6muHRDU3o13jxKl0/r0sjNPs55VqFwPfoaIbxdBmofc7JDq+SJN2LbvrZTAjxW0KI40AztWA+O/tQBlhQp8lxnK86jrPHcZw9TU1N9Z8H3Fo9kAbY1BpCVwXNITfNITdj6RITmRLT2TKFioXjwFCiSMCtzZstg9pM99yfmZbNt98a4T/+8jwnxjNL/i0jySJv9Cc4M5nj9QvxeY91NXhRFYFbV2gLe5bYw/WxbIdfnJ5iJFnkV/JW+z1FU5UVBdIX/eZ8nL/8xTl+cmyCl8/OAFAxbSYytcBoKFFc8rkX61KPJEuMporzLkJ7Yn6EqL0/L29RvrWjtvBtc1vwugLpbLnK+ek8pmWveB/3qtfOxUkXq6SL1UVT5K5Vd4MPVRF4dPWKefeSJEl3s5u+ANFxnB8CPxRC/EfABC7mSISA9Er3u645wB8+vg4hBOemcvz4aC0f8EP3tbGxNUimVGV3z/JqqiaLRn1W+dREbsnZlpD3UrpJ7LIyUz0xP19+tA8huGG3PlVF0Bz0MJUt03qDAnbp7tQ/k0dXFRJ5o35HxetSeaCvgQszBfb3Lb0YNxZw14/71tD8AGpzW4iemA9dVRYEzAfWNrK/L3bFuz1XY5g23zo4TNGw2Nga5APb21a8r3tR2KejzL7+kcsWbq9Eb2Ptc04RQjaAkSTpnnVTg2khhNtxnMrst1nAAZ4Evk0t5eNr17n/2o7Ll7JJioa1rBNurlzF56rNWsf8bvqa/ExkytzXufRty7BX5/P7eyhUzEVnZS7PkzYtm+lchcbZYGQ1fHJPJ6mCQaOsGStdpmRYKMriF3P71sSomDbtYQ9PbGqu//zA2kYOrK2lSBUqJvmKSUto/oVag9/Flw70UjHtRUvjXVyPsJjrCaQBqpZNaXahcK5cvcrW0uUe29jM4aEUQhE8vLbp6k9YBrkeRJIu6f0fn7uu5w/+6bOrNBLpZrrZM9PPCCH+1ezX56jlQv+FEOIV4F3Hcd5cjV+yozNC0bBQhFhWDt9rF+Ic7E/SGHTz2b1daKrCh3d2LOt3hTw6Ic/yZnh++O44Q4kirWEPn93XvaznXI2uKjSH5Ky0NN9AvMCP3h1HUwWf2du9IOjd3RO94p2afMXk/319iHLV4sDaGA/0xeY97ndrXKEs9Q3jd2u8f1sbQ4kC9y/zTpN0iUsVVG0HYYNbziRLkiStipu9APEHwA8u+/Efrfbv0VWlXkd6OS7mh8ZzFQoVi7Bv8ZPMQLxAIl+hatmMp8vsW9NwTY0PZnKV+v+vtrhRkq7HSLKIZTtYtsNEpnTNzVVy5Wq9VORMvnKVra9dpljl9GSWNY3++sVg1bL55elpDNPmiU3NS+aEb2wNsrE1uOpjuhecmsiRK5uzX2fYc4219SVJkqSF7simLSuRLBicGM+wptG/oMHEg30xfnM+TneDb0EDmLnP/8GRMSpVm5FUkbVNAYpVi89f1qwlV65iWg7RRYKXp7a0cGwsw6bWkAykpRvqvs4Ik5kybl1hXfO1V3lpC3t5oK+BeN7godm0j6WUDIvDwymagm66G3xky9V5i4MX88Oj48RzFQ4Pp/nKo30oiuDMZI6T47X1yDG/q16RR1o9lm2TK1cRgI1zq4cjSZJ0V7hngunnjk0Qz1V4dyTN7z+2Fm3O4qg1jX7WNF65Xq8ABAJNFfUavO2XLfqbyVX4h7eGMW2HD2xvY0PL/NmzvqYAfbJ8nXQThH06n9p7fd3tDlwliL7o12enOTWRw7RtdEXBAR7oa7ji8y9eSs69pmwKutEUgeU4tMgFtTdEumhysQhKqihzziVJklbDHZ8098q5Gb715jAjyaVLeQG4Z4NnXVUWzApPZcscGUlfsQNi1O/i8Y1NFComa2I+PrKzfd7CLYB4vkLVcnAcmJwtMSZJdwPLdjg2mmE4UWQsXeJvX+7nW28OU65a9cW0puXUFwdOZa98/P/WznY2tgZRheDV87Wyki0hD196qJcvHeiVNdNvkLBPw3YcbMch4ru21B9JkiRpcXd0MJ0qGLw9mGIyU+a1y+o8X+6DO9rwu1Sqls1g4lIL5aJh8o9vj/Cr09O8cHLqivvIVWqd4QqGRbZsLgjK/S6VVNGgYlrs6o7Uf/7uSJrvHBplcBmtmyXpdvT6hQQvnprie++M8tr5OPmKSf9Mnr99pZ+SYfHezS38zkNr6Ih6mc6V6bnKWoKQRydfrlULOTSUIlkwAEjkDb53eIznjk5g2zINYbWdm85Tmm0nfmE6d6uHI0mSdFe4o4PpgEerL6xaaiGg49ROyMWqRcGwqFoOh4ZS9cdtp/Yf1GbfrmR9cwCvSyXo0RZt43xoOEXU58KtqZSM2gydYdr86sx0rbHKGdlYRbozmXYtN8BxoDfmx6Up5CsmZcPi7FQev1sl4tNJ5A2agx6OjWWvssdL79kGv4vA7GLDQ0MpMqUqZ6dyq7Lw8eL7X6qZnnPHYCKz+gtLJUmS7kV3dM60rip87oFuioY1r5U3QLpo8J1Do5i2w8d2ddDgd9ES8jCdK7OhJUiuXOXVc3FCXp0P72xnPF1mxxVqSgM0hzx85dG+JRcPdjf4GIwXCXn1esteXRU0BtzM5Co3rBOiJF2vyUyZtwaT9Mb8bF/kfXBgbSN+t0bIo7OxNcie3iinJ3M8f2ISj67SGHTjUhUaAy7ieWNZTYT2r42xpT2Ez6XWG7xsaAkykioSC7iJXmcawkyuwvcOj6IIwcd3d15zRZO70fu2tnJwIIUQ8P6tLbd6OJIkSXeFOzqYhlpAHfYunGAfTBTrJaDOz+Q5EGrks/u6qFoOLk3hhZNTnJ6s3ebsiHSwf21swT4Wc6UqHLt7GljfEsSrXwoOhBA8sKaBHx8dp2hYmJY9b/GjJN0Ofnl6mqlsmQszefqa/AvK0rk0hb29l8qoCSHY3BaiM+rFpSn1xjCf3ttNumTQtEgTobNTOV46M01n1MczW1tRFLHgInh7Z5hNbUE0RVx3xZv+mTzF2TtEA/GCDKaBfNmi0V97zdOzn4+SdDu53qYnknQr3LVRXV+Tn8aAi7BXZ+NsVQ0xp+VtLFA7serqwhM61G4PHxvNcHwsc023ikMefUEb5ZMTWUAwlCgynZO3VqXbz8VAM+DWFu3OWbVsDg0lOX9Znm3Qo8/rsOjSFJqDnkUD4XeGUxQqFmcmc6RLS1eSWGyR8EpsaAkS8enEAq4VlQe8G/ldCpYjsBAE3bJzoSRJ0mq442emlxLy6Hx+f2/9+8ubpNzfHaU97MXrUhcNpo+PZXnx1BQzuTJ7exv48M4OFGVlJ/iNrUEG40ViAZds+y3dlt63pYVtHSFifveCi0GA1y4kODy71uAz+zTawt76Y9PZMgPxAptaQ0vWaQfY1BpiIlOmLeRZ9D232qJ+F7/z0Job/nvuJF0xHw1+HSEEHZHlN5ySJEmSlnbXBtNzjSSL/PDdcXwulU/t6cLv1jBMm5BXw+da/CUQAhKFChdmCoCgp9HPrq4IiYJBxKsznatgmDa9V6lPDbUgYn1zEHWFwbgk3WiKIhY0M5r3+OyhKwQocy5KLdvhO4dHqVRtzk3n+e3LmhjNdV9XhAa/ix8dHedrrw3yyT2dhDwLg+qJTImJdJmtHaF5s95Qa3Pu0RSZKrVCw4kSiVwFBIykSrd6OJIkSXeFuy6YdhyHl87OMJkp8+iGJjoiXs5N5zBMG8O0GU2VaA15+NZbw1SqNs/uaGVdcxDHcRhOFon4XIQ8GpOZEpbl0Bry0BR0owjBz09McnIih64KDNNGCMFTW1rY1nHlhYvAkoF0PF/hhZNTBD0az2xtlUGCdFva3xcj5KktrFUVwZ98/ximDX/4nrWI2RYsV8vMMC2b1/sT5EomFc1mOFFc8N4ZS5X4d/90jErV4onNzXz50bX1x94cSPKb8zOEfToRrwvDtHl6a+ui3Ualxf3mwjRT+VoZwtcH4nxmX/ctHpEkSdKd75qDaSHE/+44zr+7EYNZDTP5CkeG0wAc7E/wsfs72dwWon+mgM+l0d3gYyRVrJeuG0oUWdcc5NdnZ3hnOE2uXKU55GYkWaI5VLvl/Z6NzYQ8Gn/9i3NYtk1r2ENTwIMqajNl1+PwUK1O9mQGNrUWZW6nNM/l6Um3iqYq3NdVq53+D28Oc3YqD8DPTkzyge2tfP/IOAGXRtWyF00TAXj+xBRnp3IMJYo8vK5x0a6jE5kSFdPCYWHjo1fPz/BmfxKAbR1h/G6NIyNpHr+seZK0tBNjl3Lej41kbuFIJEmS7h4rmZl+Brhtg+mwVyfs1cmUqvU6tm1hL7/3SF99mzWNftY1BygaJru6owBkZhdEXazwcXoyi64oPLg2xtb2EL86M03Eq/N6f4JCxSLk0WgJebl/9vkr1dvo59REDq9LIeDWGIwX6GrwyZSQe1zFtPjOoVGSeYP3b6/dPbkdTOfKNIfcuDQFx4Ft7WGGk0Vs26E/XuDsVI6t7YvfqUmXDPwuja3tIT61pwuv61IKh207PH9ikolMiQfXxCgaJv/sspQRr6bidal4NBWfS0URgu7YpdSUF05O8qszM9zXGebj93fKuzyLiPh0RtO1i5RGOaMvSZK0KlYSTKtCiCiwaLTnOE7y+oZ0dRXT4ruHxkgVDZ7d3jYvb9mtqXx+f8+C2tOZYnU2wPaiqwofuq993j4f29A0Lz+zM+qjPezFcWqtkScztUVWLlWQyFcYSpTwuXRy5Sqx61hUuKElSFfUh+XYfOONYYqGxZb2EE9vbV3xPqU733S2wnS2Vvnl1ETutgimR5JFvnt4FMeBP3pyPX2NfhoCbs5N5RACNEWgCPjx0XFGkiVMy+bBtbF6Sb2nNrdwaChFb6N/XiANMJkt1y9k+5r8fHhnx4Lf/8TmFhCCjqiX921uwYH6fpIFg58cm2Q8XWI6W2Z7R4Qt7SGgdqGcLhp0N/hui1n+W2luV8mqZd3CkUiSJN09VhJMbwIOsXgw7QB9i/x8VU2ky0zNdvI6OZFdsAjw8trTuXKVrx8cwjBt9q1p4KF1jQv2GfG5eGZbK09taSGRrzCWKvHGQJI1jX7+4a0R3h5KEfbpKAIMy6Yp6MLvVhfU410Jr0slV7YpVWsnt0xx6bJh0r2hNeyhq8FHIl9h+zJy8m+GTKmKM6dbaMPsReT6liBfCrpRFcHzJ6YYiOd5ezBJ2Ovi6FiGiFdnfUuQ5pCH929vW3TfDX4XUZ9OulRlbdPiqU4bW4NsbF38oiLo0WgLexhPl4j6XDTOlr4sVEy+cXCIStXm/p4oj21ous5X4c6mq5cuYnTtrlsyI0mSdEus5NP0pOM4u1Z9JNegLeKhNewhXayydXb26UqKhoVh1tohpxcJVAsVkxPjWdojHjqjPppDHppDHnb1RHEch1+ensIwbdyawv/1qZ1E/TpFwybs1fHoq1OrNejRee/mFkZTRfbMaY4h3Zt0VeETuztv9TDm2dwWIlU0qFo2O7sj8x6LzHYrjPldjCRrtaansxW6Yz7eHkrRGHBzdipHX1OApuDCOzkeXeXz+3upWvaK3lO6qvAvn9rAWKpI1O+u35UqVS0q1YvvfeOa93u32d0b5tRkFgHs6YlcdXtJkiTp6u7IqQm3pvLZa1iF3hLy8OiGJmZylQWdDlMFg//75QtUqjYNfhe/+8iaeeXyhBA8tqGZQ0NJdnVH6JmdBQ95WXXbOsLLqgwiSbeCqggeWX/lmd33bGxifUuAkEfnJ8cmmMiU2dAS4PtHxkgXqxwZSfPlR/sWTbdQFYGq1ALpWpOYFF5drS98vBpdVehtnD+r3Rhw856NTUxlKzzYJy9SD/Q18fOT0wgEe/uW1/VVkqSb53o7QA7+6bOrNBLpWqwkmP6rxX4ohPAAH3Ic5x+vb0g3xu6exRcK/vT4JAPxAtlSlZ6Yn2+/NcqDaxvY1HppxvvZHW08taUFVRGMpYo0Bt0L6t9atsObA0lsx2HfmoYlKxpcLluuUq5aNAc9S25zejLLUKLI7p6obPoi3bbyFZO3BpI0Btx0Rn18em8XVcvBpSkcH8sCDvmKSalq1S9Yc+UqJcOiOVQ7/h3H4aUzM3z30Ahl02ZDSxC/W7uuKjc7uyL3fK70RWPpItlSFQRMpctXf4IkSZJ0VdccTDuO87WLXwshVOB9wGeBp4FXgNsymF6KS1NY2xQgXTLIlaocH0tTNq15wfTF7f7yhbO8diFBxKfzpx/bXs8ZBTg1keWN/gQAHl1hd8+lWbAXTk5xbjrHA2ti84L6ZMHgmweHqFoOLSEPfrfKgbWN826DFyomPzs+iePUZtFlXVjpVjs9meX0RI4dnWH65uQ3f+fQCIPxImGvTnPITUvIg0urBbEf2dnBdw+PMJmp8PU3hvjC/l4qps3X36itZXhsYxP3d0c5P53nuaMT9McL2E4tl9q9SHvziwoVkyMjaZqDbroafHz/nTGy5Sof2N5Ge9jLD94dYzhR4pENjdddeedu8N3Do2TLtbUZ//TOKJ95YOkmO5IkSdLyrKh2lBDiUSHEfwEGgd+jFlCvcRznE6s4tlVn2w7j6RLl6qVV7B/c0cYHtrfx8V2dlKu1xi3vjqT5u98MMJIsznv+qcksiUKFE+MZ/v6NQSrmpf3MXYg49+uKaXF8LEOlanNkJD1vf7X8U4dCxeSN/jj9MwVePT8zbxtdVfDO5pCGbkILZkm6Ett2+H9eGeC7h0f52muD9Z8fH8twZDjNqYks2XIVz2V3bsI+nYjPhdelUqhY5CsmF6bzFGbrtF+sXBLwaIS8GiGPzrrmAJ/a01UvcbmYl87M8OZAkueOTXBqPMtEpkyhYnFiPEveMBmMF7Edh5Pj2dV/Me5AY3Nmo4cu+3yTJEmSVmYlTVtGgWHgb4A/dhwnJ4QYcBznqp/MQogHgL8ALOBtx3H+pRDij4EPA0PAlxzHWZVSFgPxAoqAntilSh+/OD3N8bEMIa/OF/b3oKsKHl1lW0eYkWSRTW1BEvkKVcsmXazy2oU4u6pR1jT60VWFj+7q4K9+cY6gR8e2a6kdAMOJIoeGUmxtD7G5LTTv5O/WVDa2Bjk3lWfbZYsl18T83N8TJZk3CKU0LJsF6R4uTeFzD3QznavQc4WgQpJullTBwDBtUoVLC/rKVYvuBh9Bj85Tm1sI+xZe+D2yvglVidMW9nJhOs9vzsdJ5A32rmlgf1+Mc1M5LMfh9x9bS7pYqxN/sfRdybB47ULtgrM55Obpra14dBXX7Ky1KgQdDV4aAy6yZZNNrUGCbo1NrUEGE0V2LjPv+m63uS3Eq+drd9C2LVEP/FrkKya/PD2NR1N4fFPzstPbJEmS7iYryZn+LvAR4NOAJYT4AbWSeMsxBDzhOE5ZCPENIcQjwOOO4zwshPg3s/u97jSRUxNZfnZ8EoAP3dfGuuYglu1wbjqHZdtkS1Uq5vxObV0NPj6xu5Nc2eStwSTJgsGZyRzj6TIbWoI8u6ONR9Y3cWw0y2i6SNTnwufSKBomL5yaJFuqzbA9Olt662B/gnje4KF1MT6wvQ1n28JOdooi6qW6ioZJrmzSElqYOx306AQ9clZaunkODaWYyJTY3xebV0ddUQQfuq+No6OZeYt5d3ZFMG0HTRHz0inKVYtkwaAp4MajK/X60T8+Oo4QgtawhwNrY0xmy/zk2AQAT21pWbAQ94fvjvGToxOMpEo0BlzkKyb/7IEe3rOxifaIh8aAm+agh8/v753XNXKpUnz3qqY5/5ZNgaXXaSzXO8MpLkzXumF2Rn312t6SJEn3kpXkTP+REOJfAI9Ty5X+cyAkhPgU8BPHcfJXeO7knG9NYAfw0uz3LwKfYxWC6dKcNI5y1SZdNPg/fn6GsWSJqN/F5/f3EFikPvTFWeyt7SEypSp//9oQtuPUb0XbTm0xVW/MDzh8/fUhhpKFWrAQdNMR8eJSFcbTJV67kKjv931bW7AdZ8Gixbl8Lm1eFZEbKVUw+PnJSTy6yjPbWq84Lunek8hXePlsLd2oatl8dFetRJ9tO/zi9DQV0+Z3H+6jI3qppI2mKjx4WXUI07L55sFh0kWDVLFKg99Vr/X80NpGHAdiARftES8nxi+1tjbthdfm5apNwKNhmBZuTaFi1N7juqos6Lh4+UXrSLLINw4OEfbqfHpvNw33cOe/0eSlj+eR1JIf1cvWGvLUG/Y0Bu/d11WSpHvbiqI3x3Ec4JfAL4UQOrUW458F/jOwsCPKZYQQO2a3S1NL+QDIAAtWCAkhvgx8GaC7e3mL7+7rjFA1bRRFsKUtxEtnpxlJFilXbTrd6pLtjuf8TiI+Fx/Y3srpiRwVy+KlM9N4dZVHNzRSNCwODiTon0mTL5vc1xXm4fWN3NcZRVEEQY+GS1MwTBtdFfztK/1YlsNH7++gM3rrUzXeGUkzPps72T9TYHObnE2SLvG5NLwulZJhEfNfmskcS5c4PlYLeo+MpOcF04sxLJtMqYppO4ymijT4XQzGCzy2oYmo3zWvC+mWthCW7WDZDjs6wli2g6pcCoo/sL2NngYfHl3h2FiWsXRp2X/P198Y4shIGlURbGwJ1jop3qMuLj4EyMzeTbse61uCfMHvQlMFIa8MpiVJujetJGf6a47jfOni97M5zj8CfiSEuGr1ZSFEA/CfgE8Bu4GLfYND1ILreRzH+SrwVYA9e/YsK51EVQQPzJkl64r66GsMEC9UeHZHWz3P8mrWtwQ5PZnj2GiGt4dSbGsP0dcU4J8/0sfJ8SxtYQ8pzWBja4hdXVHKpk3FtAh6ajnZubLJVLbMidnFT8OJ4i0Ppp8/MclvzsfJlqtsbAnRFr7+W73S3cXrUk6bGrMAACAASURBVPn8gz2kS1Xa5xwfDX4XAbdGvmLSvYz8fZ9L44lNzQwmCmxuC5IumeyfU+v5zGSOwUSBnZ0RmoJudnRGsG2H770zxmiqyCPrG+tVcZqCbh7b2MR/eP40hYrJi6em+fS+bjqj3qve0WkOetBUgaYo9Trx96pY0A3TBQCaQtdfZnM0VeQHR8ZxqQqf2ttVb5YjSZJ0L1nJzPSOpR5wHOeK00VCCA34OrWFi5NCiLeAPwD+A/Be4I0VjOeq1rcE+R+e3oimiGvurhbx6pydylOsmAynSqxrDqIogk/v62IsVaI35uOtwRR/8oPjjKZKrG8O8MUDvXREvHzjjWEOD6do9LvY0RVha3uYimnh1lRMq9aVTZtNCzk2lmFDS5A1yzzZn5rI8sq5GbobfDy9tXXZdXRPT+SI+ly1BjUPr1n2hYV0b/G7tXlVaS7+7IsHeusXjIuZe1wD3NcV4b6uWpD8xkCCkVSJrgYflu3w0+MTFComf//aIC5N4QPb22gJuvnx0XHawx5OTeTmlZgUQtAT83FyPIffrfJnPztNg8/FHz+zccHC3dcvJBhJFtm/NsZnH+hi75oobSHvqgSQdzL/nJQu/yp0b70wU8AwbQzTZiRZJHyNTacs28FxnPrxIkmSdCdaSTDtE0LsAhaN3hzHOXyF534S2Av82Wzw92+Bl4UQr1KrEPKXKxjPsiyWI70cD69v5NBwiulcmelshcaAi0rVolw2OTWR5e3BBP3xIkdHMpRNC5eqMJkp43WpvDmQpGCYOMBv7Wzn+++M8fLZOJtbg+i6guPAh3e286N3JyhXLc5O5viDx9fNu729lMPDKQoVi1MTOfb3NS5aPWExe9dEOT6WYWdXVAbS0jVzaQouTcFxHDKlKkGPXj9e4/kK33pzmAvTee7viRD06Hh1lSc2tdAfz3OwPwnATLaMTa1GdLpYJVOqEnBrHOxP1MtXjqVKfHrvwrSuP//Efbx6Ps7rFxKcnsyRLVV540KC39rZUd8mU6rWa76/ej7OZ/d1s6NTVvMAeGf00s2/Q0OZK2y5PFvaQgzM5PHoKn1N1zbrnyoY/MPbI1i2w0d2ddARuQFtZSVJkm6ClUSYHcD/yeLBtAM8sdQTHcf5FvCty378OvBnKxjHdTnYn+DYWIZd3ZF5s1+XE0LwlUfX8k9HRhnzlnhjIME/vj1KfzxPpWoS9bsJeXTKVZOq7dAUdLO9M4xHU1nbHOD0ZJZ1zQFcqsLPT05RMixeODXFE5uaMEyH//LSBaZzFZqDbtY2B1AEvDuSplS12N0TpVSt1anuivpoCXn40bvjpIoGnVEvM7kKnVEfQc/y/xkPrG3kwNqrprVL0hW9eGqaH707TtWy+fKjfezqjjKcLDKSLDKRKfPiyWlaQh7aI16mshWmcmVmchXaQh5+fXaGiUyZyUyJtc0BOiJeqrbDrq4oR0ZS2DZkylX+7jcD/PuPbcfr0rBth2y5Ssij8+TmFo6PZUgWDMDh6GiaLe0h1jUHAfC5VKI+nVSxKgO0y5QqlyqP5ivX3wGxKejmSw+tWdFzR1JFSrMLSQfjBflvJUnSHWslwfR5x3GWDJjvFG/011p/v9GfvGIwDbUc0l1dUaYyFSbSZfKVKrmyiXAcJjNlLMuhULVoDLhpDLjrOZz/9v2bGM+UiPndaKrCmpifNwcSbGgN0hbx8e5ImqJhsq7ZT1vEy8fv76Q/XuCXp6eBWvWQsVSJ0VSJQ0qK921pYXi20cJossTW9hB7extQljGTLUmrqX8mz1S2Foy9NZhkV3eUDS1BemN+4rkKCMiWq7Q7HkaSRfxujeagm4fWNXJsLMNoqki2ZGJN5Xlqcwv/8we34NFVHuhr4F//47tYts3J8Syvno/z1JZWvvnmEK+ei7OhJchH7+9EVRS2d4TIV0z8bp2hRJG2sJfvHBqlVLX4wPZWFMS8sn4SzF1zWDSW3u5mWNcc4PREDsOy2SIXQUuSdAe7ObXYbkMbWwOcmsixoSVYn03b1hGqB8KD8QIuTaF9drZkc1uImWyZwXieZMGgr8mPUbUom7WKBY5Tu219YjzDt98aJlmssq09xMPrm+q/szXsYd+aGBGfzu7uCOcnswzEy4S9Ln53Wxs+l4ZrTu6gW1PRVMFMrtbV7QPb22jwu5jOlhlOFclVTLIlk4/v7ry5L550z3tiUzMD8QKW7bCtPYwze2H5yT1d7OgM86sz08RzBooiqFQtUkWD929r482BJL2NPi7M5ClXbUzLxqvX7tr0xrxsaAnxh0+s489/dgZdUzg6ksbv0vinw2MIIciUUnxyTxcBj0bU52JTa4iwT2dXd5ShRHF2thoO9ieZyJRRBHxidxetcxZS1t6nWTqi3ntuNtSlwsWCHivMfFsgU6qiKWJBjv3V+Fwan9rbtTqDkCRJuoVW8nH6b+Z+M1sabxsw5jjO9KqM6iZ4ZlsbT2xqoWrZ/NdXB7Bsh4lMiQ/v7ODYaIYXT00hBHz8/k66GnycGM/w/SPjnJ/J0xSo3dqM+lz811cv8O5IhkLFxLJtgh6N//b6ELqq8PLZGfqaArSFPfUmGPmKSSJf4afHp3ju+CRFw0JTFDSlFkR3Nfj4+P2dlKoWG1oC9MZ8HOxP0hR08+Zgki8e6KVYMfna64NUqvY1n8AkaTWsbwnypx/fUS9h95vzcd4cSKIqgofXNTKeKnFiIsuhoST3dUXpa/Jzf3eE8zN5gm4dl6ogqOWKvT2UYiRVIlsyeWh9jM/u6+GLB3qYylSwHIeDAwksB+LZMg+ujdEUdON3qXhdGrGAu34x6dYUYgEXJcPC71JrpfaAiUxpXjD9/IlJhhJFNEXwe4/01bssXk26aBBwa3f0YjnzUmU8DPv693d+Os+Pj46jqwqf2tNFU1DeCZAk6d6zkkjsY0KIMcdxTgghwtRyni2gQQjxr2fzou8ILk3BtO168vfFihj52SYtjgNFwyKRr3ByPEvIq1EybAzL4bULCX77gR4+tacHwQhTuTIeXWF4tp61W6vtq2RY9McLvHIuzkS6TH88T1PQQ//JSRwcPLqCYVq8O5pmZ3eEkEenO3ap7FjIq7O2KUC+YhKdXWToc2t8bl83M7nKsqt/SNKNcHHxYaFiYjsOZydyzGTLuHSFqFenYjnkK1VifheqKtjZGeH8dI5CxcS0HSzHwefSiOcNQl4dw3SYypZ5cnMLPzgyhq4oTGTKuDWFnV0RdnSE8egqibyBqgimc5X6WPxujS/s76Vq2QwlChSrFrqqLKijrsy+z4Wo/bccvz47w+GhFI1BN5/b172sRcK3JQWwL315vcbTJRwHDNNmJleRwbQkSfeklQTTjziO8/uzX/8OcNZxnI8IIVqBn7JwgeFtzefS+MSeTsbTZbbOtsLd3RPFsmvdDiNejf/1hycYT5cIenQ2tQWxHYepTIm//sU5HtnQyGMbGjk0nCZTNCgZFo0xN1G/C1URvHYhzns3tyAE5A2T5qCH8UyZlqCLclXH51IJeXX+v7eG+cbBIZ7c1MyzO9rrM866qvC5B7qJ52uLDYuGiaYoRHwuIj7ZJEG68Szb4dRElqBHq3cJvdzD6xs5NZGlfybPYFywpzfKe7e24lYVHl7fRE/Mx7ffHuHEeJZjoylyFRO3JtjRGeH929vpinoZThbxuTTOTuVqtdBLJqmiQdSnE/ZqtEU8rGnyc3Yqh9eloimCRzY0LRjL998ZYzRVoiPi5cO7OxY8/vTWVk5OZOmIeJddKnNkdq1CPFehVLVWXB3oVvNpgqxRK9fvd11/OL2rO0KyYODRFda3BK57f5IkSXeilZwR5i5beYrZ9t+zdaNXZVA3k+M4eHWVnV2R+myTS1N4eH2t4sW5qRzTuQpFw6JUtXhPezN+l8pkpsypyRzuAYV8xaQ35idZqAACy3FoC3sQQpAsVGkKuvnsvm62toeZyZUZThYZThbRVZV1zX5e70+SLVWpWjaW7ZAtm3x0V0d98ZSmCry6ypnJLD8/OYXPpfK5BxZviS5Jq+3gQKJe1u4z+7poCy/MM/a5NHyuWgpEulhFVRQ+uKONfNni/FSWP/n+MfIVE69LZSpnoKuCRr+b/+mDW4j53TT4XRwAhhIFvnd4DNuByWwZlyqI5w22d4bpavDxwJoYf/PSBWzHIeqr3bW53MXZ6pl8ZcFjUFtQvLtnQbPVK3poXSNv9Cfojfnv6PddsXqp71V+FfI8gh6dxzY0oakC/Q5Of5EkSboeKzkrpIUQHwTGgIeA34V6Q5Y7bjXPL09Pc3Q0Q0vIw2f2di2ojLG2KcCTm5t59VycWMCFriq8d0sLvz47w4nZmbje2dm6h9Y2MZUr0xR0c6AvxomJLOubg2iqQkvIw1NbPExlyvztKxcIuFQcB0zLoVAxyZRrtXbPTNWaqvzw3XF+56E1mJbNNw8Oky5WMS0bTVUoVCyms2UCiwQSkrTaLPtSAGZaSzch3dsb5cJMHl0RfHRnOz84Mo5lOfzi9BTJgkHRsAh5NCzLxrQdbBy+/84YmqKwf22MB/tiNAc9hLw64NAUcNUW9wIhj05PzI+mCHRNMJ01WNO4eBfGp7e2cmI8U7/TtBrWNPrvipQqlwrmbEUPj3b9kx/npnI8d2wCVdQaWV3ePEeSJOlesJJg+ivAXwOtwL9wHGdy9udPAs+t1sBulou3b6eyZQzLxqOolKsWvzkfx6UpHFjbyJcfXct/99Aajoyk8bk0LBumMmW2tIXQNYXP7uvG79ZoD3sYTZVo8Ov43Tp9TQH643l+fWaGsFcjVzF55ewMR0czhL06e3sb6Ih4ODqWQVMUwj4NXRUMJQqUqhaf3tuF7UC6WKsNG/LqhLw6Ya++rHbOkrQaHuyL4dFVgh6NriWOO9t2UBWFZ7a1sqEliKYIHAcURdAa8pArm+iqjRC1wNjjUtjXGyNbMmnwuxhPl7BtmyMjKZoDOmPpAsdGs2xuDfHIhka2tIdpD3vIl03iuQqmZRP2Lp7mtK45wLpmeaG5mLBXo5irRdORJV6/azGZLdcmBRyHeM6QwbQkSfekaw6mHcc5CzyzyM+fF0JsXpVR3UQPr2/ircEka5sC9fzJw0Mpjo7WuoM1BtxsbguhqQp7ehsoGiZffbkfIQSlqsWTW1pY3xxACMFr5+McHEjic6l8YX8vX39jiB8fHcejq6iKoDnoZjpXwaurJIsGw8kCx8Yy2I7D+uYAz97XwqvnEpSqNj0NPvJlk+lcBY+uEHBrPL6pmc5oLZjJlask0yW6or76bHqqYKCpYslWzyt1cjxLPF9hd09UVg+5B+mqwt7eK9dif2ckxU+P1a6ruxt89DUFZmeqC/z7j23jP/3yPK8PpCgbJmGfTmfEy5NbWnj9QgLTdtjYGuR/+8kpXjkbJ1euYs6uWahaDp/f30NHxIttO3zn8CjHxzJ4dJXUdRRKdhyHw8MpKqbN3t6GeyZFoTLnzkJ5bmmPFbq/O0q6WMWtKWxYRs50xbR4ezCFz6Wyq/vaUm2kq5vKljk2mmF9S2DJ9Q2SJK2+1Y6M/hU3sCX4jXBxFsswbX58dJxy1a7P+ipCEPbOD0x1tRbYOk5toeLjG5vrj13M1RyIF/jBkTFePDXJQKKAUbVQVYFbVQl4NLa2h9nYFuJgfwJNrf2OJ7c08+KpGcZSJWzH4ZO7O9FVhRdOTgEQ8gqyJZNyoHYC/G+vDzGRLrG3t4EP7+q4YbdbZ3IVnj9RC5IKFZP3b29blf1Kd5fxdJljY7UL0MFEkdawh8PDaSzb4bljk7zWn6RctSgaFkGPTkfUw/cOj3FhJoeqKLwznGRgpkipaqKrClXLxq2r7O6JsL6l1tnw5ycneeVcnKJhMZ0rc24yR7porGgh7pmpHC+fjQOgCsEDfbHVezFuY5nSpQA6WTSvsOXy+N0aH7qvfdnbH+xPcmgoBUDE51qQOlOu1oLtkFeTLeBX4LmjE2RKVU5PZvmD96yTDb0k6SZZ7WD6jn3nnpvOcW4qD0Bz0M1n9nWhqwqNl3VQ09VaWsfF6hpzPby+kVTRYCxVqpXIM63arW4hEA6YtoPXpbKxNUjJsNjSFiRdMvHqCiGPxsH+JJmSQdCt8+tzM1RMm7JpMp4qc2Q4xfmpPFvaQ7xvaysnxjIkCgbJgsEz21uvers1nq/gOFxz6SqXpqApAtN28CyzHq905yhUTCYyZbobfLi0lc/OdkS99Mb8GJbNWKrIS6dNylUL07I5NJQkVahdaJp2rTPiOyMZbNthNFnCtB2C3lrFjqpl0RHx0BLyoimCZMHgJ0fH2dIe5tx0np4GL+emcgTdGudm8gzEC+zqvvZg2junisfFRk33grkp7+Yq1Jm+Vr7ZzxAh5v8bXPR6f4Ijw2kAoj7XkmlF0uICbo1MqYrXpS277KMkSddvtc8iS69Ous21hjy1utOWQ2fUu2jFgov8bm3RdIfGgJtnd7TVOyK+d2MLr+kJZnIVBOBxqWxqCVEyLDy6Sk9jlD09UX5+YoI//9kZ0sUKlg1l0+RnxyapmjamVWuK0R8vUDAsvC6VsFdnfXMQMZOjM+qjajnc3x0lkTcAZ8Ht1sF4ge8fGQPgwzs7rmkhVdir8+l9XaQKVZmHepexbYdvvTlMrmzS2+jjo7tW3klzR0eYcz15Xj47zY+OTrC5NUhHxEu8YDCZKWE7Ak0V9EQ8CBxifhfJgoFQIOZzEfW5EMLBE/aiqyoNfhepooGmKLx46v9n7z2D7LrPM8/fiTfHvrdzbuRMgAAJMZMiFSzJskxZcvZYtmyvZ8dbM1XeMK5xTdXuzOyud6bGtVUee9Y7nvGsozSygk1JVKTECBKByEB3o3Pf7r45nnz2w7l92UA3gAYBkAR5f1+IwEaf7r73nPf/vs/7PMu8MVPAdWGwI8SRkSTZqo4qiYy9zdfkUEeIZw/1Y9oOox+gRV6Bt27S78bR+NBQgrphkQiqVwXprLJaYIuCgE/5YEhv7iSfOtDLTL5ObzzAveiu1abNvcotF9OCIFTYuGgWuAfdPFbpCPv4wsMj2I57y7rg2XydN6YLjKZD7OuP8/nDg1Q0k8FkkF19MQQBhjtCmLbLQCLQSi+M+WUuL1X462NzZKteIS2LAqrkLUGeXShzaCiBYdnNh4+PAwPe6PPxHWkur1QQRbxIZlUiXzMoNUxOzBav0rjm6wZu8yeWrxm37ErQGfG3F4vuMLbjcjFTIRZU3rVIa9t1qRve2L/cuLWRf6bkhRStSixkSWRLZ5iZfI1i3aRh2rjAclmjYToEmxOZbV0Rnju9SK5W9iQfPomemJ+fPTKIbjn8+cvTFOoGj25PEZBlTswWWChq+BUJWRR5ZGuK/QMxCjWTPX0xorexH3Czrqfrurw4niNfN3h0a+p94evul2BV6RFU70yxla3qKKJILHjzn8WJ2SJvTBfJVnWOTRV4ZGuqJeMBeGAkSUdIJeJX2vect4Ffkdi25vvZpk2bd4a3s4D4vn2nbjbAYZVVfd+PLq/gVySmcjW2dUXojvlbXZendnZhOy4vXFqhbtj0JQL83JFBjs8U+Otjs4wvVakbFoblIIswnApi2C5BRcK0HU7NFpgreIExg8kQj25NU6wb/O3rc14n2oWLmQoj6RClhuf6MZuvX1VM7+mNUWwW1Hv7Yjf8mlzX5Tvnl5nO1Xhka5rt3e/bH/e7yksTWV6fKiAKAj//4OA6OdE7gSKJ/MS+HiaWq+wf2Lw+9cx8iefPLSGJAp9fo8/vivowLIcjo0me3tnFTL7OG9N5emN+ntzZyaHBJP/7Ny9QM2wMy8awHAzLwacYZKs63VE/XVEfpYbFn/xwElkS6I8H+afPbOfsQhmfLPKts0sYlsOTOzrpit7dYmuu0ODYlOevrYjC+2JfQJJEsD19hyTefuf31GyRP/7hBKos8rsf3U5v/MYHlGLNIFNucGGxgiQKGLZzVTG9VNZ5cTxLNKDwiX29tyU9arMe03bIlDQ6oz58clu216bNneKDIxa8C6zq++YKDTojPoY6giiSSKlhMrlSZTQVJhZUeHkyy9dPLdARVgmoIn3xIP/t+DxXsjVEERwHfIqEADRMB0EQ6AirnJ4v0zBsbMdFNx1eu5LjH/9lBUkQUGURRRIQRYGRdIh02Md9g3GWyhoPXrNMpcoiT+7o2tTXVNYszjQXyV6fzreL6buE0RSsOq6Lab8L4tUm/YkAVc2LAt8suZrnomE7LqW62SqmX5rI4VckHAd64wFsF0RBpDseYDAR4nymQsO0cVyHaEBBlURWKhrLZZ0/fmGSvniAuUKDmF8iVzPRLIdzCxWGUiH+hw9v40q2xslZT097IVPmjek8giDw9K6udfsLd4JYUMGniOimQ2f0/RGT3VgT1FLXb9/N46WJXOv1cGKmeNNiWhQFNMNBlgREoDvqw3Hc1qLcqbkihbpJoW4yk6+3pWV3mNV00M6oj59/YOjdvpw2bd43tIvp22BV37e1M8xTuzrZ1hVBEgW+cnyOQt3k+EyRz93fz396cYpLmQqiIDR9d+uEfTI+WUIUYGevj8WSRsOwcRzoT/jY0R3h0lKFmu4iCJAIKQgCrSXJgCLy6YN9/OzhQaIBBUEQeHyNs8jbJeyT6YsHmC822uPCu8hDW1IEVIlEUL2hPv9u873zy1zIVJBFgV95aHhTtoqHhxM0DJuQT7oqgbBYN3jtSp7OiA8BuLhYRpYE+mJBtnVHsDMVBppF78Nb0zw4kuRffv0s5YaJZtqML3uvbdd18UsimuX5UudrBvPFBgIuh4YSTOVqnJgucmahxPauCKmw764U01G/wi8dHaauW3Te5S74u8GdkNTu7YvyD6cX8MnipkJyVElkOBVisCPI4eEE5xYr/NEPJ/jMwT56YgG2dIa5mKkQ8sn0bKCpLtQMfnhphXjQS168VhecKWl898ISHSEfz+zqartZXEO+efDJVw1c123rqtu0uUO0i+nbYK2+L18z+H9/PMVAMkCmpDGVqzPYEeDychXNsFEkkZBPwi9LdIRUuqJ+9vXFODFbYDpXp6pZ1HQbcAn7JQ4OJXnuzBI+2WZLZ4iHtqQp1AzqeomKbhIJKDQMh7lCg++9PE0ipPLZ+/s3PbpzHJevnVpgrlDn8e2d7GnKPyRR4LP392PabnvEehfxKxIfGku925eB5XgJnJeWK/gViZ97YPCmOwNBVeaje7rX/bkiioiCQLaq89JEllcm8wjA3v4oW7si9MYCfOPUAjXdYqHYoDceYN9AnMtLVUzbJlc1cVwXRRLpifkQShqJkA/dtPjTH00S8Ss8tbMTMQ9nF0qUGgaXlyt84dGR1jWUNZM3Z0v0xv13ZLEw7JPv6fjwa/HJUG8lIN7+v+e4XqiPJAjUjJt3uo+MJPGrEmGfjGbaNJofM7lSoycWYCwd5rceH0NqNh6u5dUrOa5kawCMpsIMdlx9iDo2lWe5rLNc1tndG227gVzDR3Z3c3q+xM6eSLuQbtPmDvL+eUq8CwiC0NL7ff/iMpppc3mpiu26yJKXANefCHJgME6q0KAr6mcwGeThrWmqukWuubhTbJjeg0NwsR2XsmahGRaKJBD2yRSbT79k2Me//Zn9/KeXpqjpFr0xP5PZGpbjciVb4yvH5zk61rEps/5Sw2w9lM7Ml1rFNHgPyGNTeSzH5ehoR7uofh/z1M5OFooNTNuhqltM5Wrs7r2xrn6Vi5kK5xfL7O2PMZYO0xXzI4sCsaDKTL7OybkChuWwp8/rWGbKDaZzNVRZpFQ3iAQUIj6ZjpDKcCqEKonM5GssljRkSaIzGqAr6melYjCTb3BoKMnFTIXlso4oCgwkgxwaSrC9662O6HfOLTGdqyMKAl94ZGTDQth2XF6eyGG7Lh8a6/jABLYA6Gv2TJsrFrfFYDJIxK80D0A3n7DIksjBZlhLTbc4M1/GtB12rJGT3ejn0R0LcH7RO/jFQ+unKCOpEBMrVcI++V3ZQ3ivM5wKMXyLC+ht2rS5Oe1i+g6xuzfKSkWnPxFgOldjpaJj2g6KJPBPntqK63rdSNtx+YtXZzg9XyQWUChrFrGAQjKoMrFSxbQd/LLIV04uEA+quBjEAipXsjUyZY0TMwWqmoUqiwyngnTHAswV6kzldIKqxNdPLfBbj29Busl4MxZQGE2HmCs0riqkAc4vlnntird4FVAkjowkmS82yJQ0dvdGb3lRs817l6DqhW589aQ3qt9sTL3runz7bAbLccmUNcYeC/P49jS5qk7dsIkHFeqGjWbaXMxUsWyH//CDCQo1A81ySEd8/NmLVyjWvbTDVFjlwGCCY5MimbKOLAo8vbcX23XQTZvlqs5Awk/dMJnMegX/zp4Ij2/vvOq1vnrwkyUB6ZrO24mZAmfmSwRUidl8A4CQKnH/DdIda7rFhUyZ/kTwri88vhMIa7zx7kRjcjgV4ouPjiI29zhuhu24nJkvEfLJbOkM83MPDN7S5+uO+qgZFomgQmgDf/A9fTFG097BTP4AHZLatGnz7tIupu8Q+/rj7O2LIQgC3zqzSLlhEfLJZKs6Wzrf6rqcnC0wnasxV2gQUCR+/dFRvnPecyiIBmTmCw1qhs1Q0vOPth2XeEjhjakCiyWNmuG1loKqxJfemOPZ+weoahb5mkHdsOiK+PF88G78pBRFgZ880Lfh30X9nj7bdSEakKnqFl9+Yw7bcVksNfjEvs0nnrV579MbD/Cbj43e0thXEATSEU/r39kMAgqqMr94dBiA16fyhFQJ23YpN0wuZCqUdYvOiI9Cw2RbV4SKZhENKNRNG78iM5OrsVDSCKoSnVE/u/ui3D+U4HsXlnnz9Vn+8PI4jusy3BFid2+09bnW8syuboY7KnRF/QSuCRn68eVsS9ayKmWJBm6sEf+H04vMFRqossivPTJyzzsgyKKA1Uxuke/QxOlWDtevXsnx6qR3UH/2UP8tIEvzHwAAIABJREFUyzD+8rUZTs+VODNX4r7BOPsH1keSf5BCeNq0afPeoH3XuYOsFiMPjqaomzYhVWYk5ek2Sw2TmVwdy3YJqhI+WcC2HdIRlf/+ya3UDYvnzy0xsVLFcV2KdZNDQ3FOzpQwTIeBjoDn5yoJmLZDXbeYytZ54VKWMwtllsoaPllkSzrCpeUqO3tuvgx0PQY7gvzskUEsx6UvHqCivTUPdu7ZWJ42N+Lt6Cd/+lA/2apOOuzDdV3yNYOIX0GVRfYPxHlqZxfjy1XG0iE6wipHRzv4rmFju3Bmocw//4md4HqLi9mKwUS2SrFuEg+oDHUEuZAp81fHZmg05Sf5moEkCswW6usK5VVUWVw3aVllOBVifLnKfYMJDg0lsF33ph7f77eXe8QvodW8A3ki+C4cDFzPUvRmk7O15GsGb84VGeoIEQ94Xt+KLN5yHkCbNm3a3C3e0buRIAi9wDeAXUDYdV1LEIR/B9wPHHdd93feyeu5U7iuy0pFJx5UUWUvvODaNLkvvTFHuWGSCCrkagYnZ0tcWqqSq5tE/TLdMT+/+tAoL09k+crJeaqayZuzRT60JcXJmSKSCDu7o4gSnJ4rsVDSOJ8ps38whiwKdIRURFHAdt1bsjm7HmtH2hG/wk/d10emrLFnk3raNu9ParrFpaUKA8kgqbCvpZP9/sVlTs4U6Qir/NyRQbJVnUe3ptmSDvHyZJ5/++1LfOHhEfI1g1ebEqJCVee/vjrLxHKFsF9hW1cEURD40JYOjo528M//7jRzhQZ1wyLiUwABy3bAFag3dw5+PJ4lHfGtW+acytZQZPGqYvkT+3qo6hZhn7zpw8PH9/ZwbqHMQDJwz3elwbPeXKWqvfNHhVhAodgwCTaXEJfKGobl3LBD/a2zGTIljTfnSvzS0SEGOoKkQupVE782m2OprHFuscy2rsi7FhbV5u4y/D/9/W19/NS/+Yk7dCUfLN7po30eeAr4CoAgCAeBkOu6jwiC8EeCIBx2XffYO3xNt823zy1xbqFMR1jlFx4YWreF7rpuy1f4SrbOUknDdrw/OzVboKqZiKJETbdZKDV4aSKLabn4FJHTcyVyNQO/IuBXZH7x6BAnpj2vXdNyODtf5g8+u5/JbI2qbnFipsCX35ijXLc4uuVqv2nbcb1rsR1euJQloEo8siXVut4bWSUNJIPtzfg2/P2bi8wXG/gViV9/ZKSlS10oehrkXNVL4fzyG3OYtsvpuSKT2RqO6+KTRFwBqrrFYCLAH353nLMLZSQR8nWTZFBhqaxzbqFEqWFSbpg0DJuAIvHgaJJTcyUs28FxQZUlXpzIMblSY3KlxmgqTHfMj+u6nF0o8/y5JeAtKYFm2rxwaQVFFnlotANBFDa1eBj2yRwZub6m+l6jusZnuqjdWurlnWCpotHdPKifni9xvBkT/5Hd3ey6jrXeqgWp2uxGf3jn5jzz26zn66cWqGgW5xfL/NZjY21HjzZt7hDvaDHtuq4GaGvewEeB7zR//R3gQeCeK6YzJQ3wxpGG7eAXr+5gCYLAT93Xx/hyFUn0tJs100IzbCZWqlQNB1l0ObdY8pYLJQnL9pZswn4JF5WVikHDtHjuzCKC4PlMh30KfYkAr07msVwXy3K4sFDh4lKZly5n+Ref2s2Htngdu0LN4G9en8W0HXrjAc4vlpFEgURQYSAR5GunFqjqFp/Y17MpN5Dr8dJ4lmzN4JEtKRKh9fHLjuNyJVcjFfJtKn74vUixbnB6vsRwR+gDd8Cwmjofx3WvkkA8ujXNy5M5hjtCRPxK8yHtElAk/IqIZbtkKg0uL1U93bJmslzRvQMekFTF5r6AJ+kQEBjqCBHyyTw0luKZ3d1cXCzz569OE1QlBhIB1GaRpZk2f/CtCyxXdFRJpD8ZIKBIFOomr13JkQypvDlX4uxCGcN2ePHyasJezzr7vFLD5IVLK0QDCo9uTbWLjTvMtq4IP7i4Qtgn0zAsFosNXKBU9/yPLdvhzEKZqF9u/Ww+trebyZUavbHADQ9AU9kahbrBnr7YO+bQUjc8R5KemP+euBcEVImKZhFQpPZru02bO8i7LTqLAxPNX5eA3df+D4IgfBH4IsDg4K1tfr9TPLG9k2NTecY6w9ddxlkbMT6QDPLotjT/8utniQUUDMshFlTxKRI7u8OUzq8QUEQCikRH2OfpSIMKtuPy5lyJZMjHSCrMR/d0oZsuf3dyHkkQqOqe00GhblJsmPzJC5OtYtobl696ulZ5c66E7bjUDdsLkxE8OcfFTOVtF9MLxUZrhC8KbLio+INLy5yaLaHKIv/ooeF7clnoH05nWCprnJwp8sXHRt8X4//N8hN7ezi76B0k1hYs104uPnuon9lCg0/u7+GN6QKSKHBpqcKJmRIN0ybqEzEdz0KyLx5gJB3k4mKFimYyvlyl1LCQRYFtXRF+7ZEREiEfL1xaQZUE5goNvnx8jo/v7eUzB/v4vb87zZuzJXTLpiPkw3Qc4gGFiZUaDcPCRWBfvydPqusWkihgOy4TK7V1xfRrV/Kt8JjBZJCRto3YHeXyUpVEUOXNuSIzuRoT2RqdEV/rYPaDiyv83cl5fLLIP3tmG73xID5ZuukOyHJF4+9OzuO63oHoTgRYbYbnzy0xuVJDEgW+8PDIe17Hvbcvxg+qK+ztb8v12rS5k7zb7/wisHqXjDZ/fxWu6/4J8CcA999//3tyH2iwI7guPAC8LqwgrF/u6k8EeWUyT188wGTW4ehoB/m6Qa6ik6sYOK5LzbDRLJsXx3MokkBAkdjSGcbM1VEkL0L5Y3t7+b2vnOZKtoblOC1dqdwcYXes6QwPJAN0RX04LiSCCprpsFLVMS2HbFVHtxzuH0qy+zrLW5shGlDwKxKaabcipq+l3PBGy4bloJkOwfXN6/c8dcPTDXdF/YgfsO5OLKhsKmymM+pvpQaualu/cWqe/++VaUzbRTMh5pdJR/w8vasLQRA4NVtCt1wM20aSdLakvHTD/3ZigXTER023yNW8jrYxU2Rff5yHtnRQbpi4gO16nfN02E9XVGW5olM3HFzXZVtXhPiDCo4DL01kqeoW+zcoKFadSVRZJHGPTk5uxBpnPO7EEdB1Xc4tlvErV6dhXo90xAd4EjdFEqhqFlG/wky+zoeAc4slVio6AHP5xk3jycHzO//u+SWmszV64wHuwMrIplm9twvcGavBu81LEzlsx+Wl8RyHBhPt7nSbNneId7uYfhn4DeBvgA8Df/auXs0dZKHY4Csn5lElkZ+5f2CdpGEsHWJXb4xHt6XZ2x/j33/nMktlDdu2kUWRuqEjiQIiDqNpb4t9Z0+UVFglqMo0TJuLi2Vsx0uMEwSB/f0xXCDsl/ErEr/92BjgLZ186Y05AH76YD8Rv7eAJQpwaalKtmawuy/G/sH4VUspc4U63zyTIR5U+dT+3pv6yIZ9Mr90dIjaDeKXn9jeSciXpyfmJ7mBDOReQBQE4kEFnyw2v//v9hW993Fdl+9dWMa03ZZ1487eGH3xAHv7Y5QbJqmwj6pmIYoCO7ui/MKDQ5QaJvNFjXLD5NBwgpcmc5TqJq7rcjFT4V//w3kk0Ts4psIqW7oifO7wAJbt0hn10xMLtKYzqwe8zxy8ejl4uaLx928u4lckfvJALz1xPwFF2lS0+r1GUBaoWV61GfLdvhTi+EyBFy5lAfjMwb6bTrX29MXoivo5MpzkuxeWvYXtgMq2Lq8Qf2Ckg7lCg6AiMda5uQTLF8e9w9FsoYEkCnSE37n7yjO7ujiXCNAT898TU7a6bnF8psDOnmi7kG7T5g7yTrt5KMBzwH7gW8D/gqeh/hFwynXd197J67mbTKxUMSwHw3KYLdSJBd/qgs0V6vTEAvzW42OokogoCvzEvh6+eSaDaTlMZmsEFAlZFAn6JAaSQWq6zd++PuvZ7aVDqLLIUklDM72QlyOjSbqjAZ69v39d8tdsvt5agJwt1Dk8nOTje3sAeGDE4K+OeVrqWEDhSrbGcEcQQRA4PVeiollUNIv5YuO6I++qbpEpaQx1BAn55BuOOmNBhce2pTm7UGI6V7stffa7RXfMT6lhEgsoH6j0vM3iui5TuboXRtQ8MJ1bLLNY0lEkEdd1Ge0M0Z/wU9NtvnZyAVkSCDYLWFUW+eS+Xj6yp4dLSxUWSxlCPi9+/WKmwl++OkPDtFksNfApEp1RH9u7I6QjPgKKzFAyhCBAMqQy3BHaMJZ6Lefmy0xla/hkkYmVKgLeZGd79/uvmA75JGqWNx2K3AFJgmm5LJc1JEnwDkobMJOr41dFOiN+bMdlodjgR+NZIn6FkE/mI7u72d1cPnxgtIO+RICQKrd2LpYrGidmioykQmzrWu/gMZIKMZOvEQ8q9MYDzBUa7OuP3/bXthn8itRKdAS4kq3x4/Esg8kgj21LvyPXcCuIInRH/SiieMOF8zZt2twa7/QCoonXgV7Lq+/kNWyW8eUKL0/mGU2FeGjLzcfa17KrJ8rkihedPJp+q2A8v1jmm2cylBomu3ojPL2zmwuZCtmqwZZ0mMlslVhAIeKX0C2XbV1hEkGVH40v4NgOFdliOB0iGVQ5v1hhtDOI5TiEVJnRdIiK5sWEW7bL4eEEsiSyoyfKxaUKNc1i+zUPo0RI5QsPj7QWFC3H5chIkoe2pNjaFebyUoVYUKUndnWneXy5ysnZIls7Q7x2pUBVtxhNh64bBAOe7OX16QKvT+XRTBtRFPiFB4fuudjfj+zuZl9/jFTYd0t+uR8UXp7I8eqVPLIo8ItHh4gHVWRRZHtXhMViHVkSUWUJAZHpXIXZQh3bdhnqCDHYEcSvSOzpj2HZDhcyFQKqyJM7OgmqEotFDRcXWYRSw2J7NEBQlXh4S4rHt3UiSwKW4/JnL02xUGzQnwjws0cGid9AT3RitsiLEzk6Qir7B+JcWvI007Ik0BsL8Pp0nkRQva5/9b1EyCdB02c67L/9kYogNJdSBTZ8L5ycLfL9C8uIgsDnDg9wPlPiz1+eZqGgcXSsg4FkcN33tT9xtbTj+XNLLJd1LixWGEwG1+2lPLGjkwMDMX40niVT0inUDF6ZzPHg6NVuRncL23FbX/srkzmyFZ1sRedAf3zdRLJQMzg1V2QwGVyn138nKDa8Dr4si2+rkD63UObcYpn9/TG2bnCwadPmg8p7fy71LvHyRI5s1fBuigPxW14s6Qj7+OUPDa/784pmYTkOFzJldMsmVzXQTAdwSUV8KLJIoWbiIHJ0LEG5YbJS0fFJIjXbRRYFXIfmoqHGS+MN9vXHuJKtkasZ/PWxGVRZYmtnGEUSuH84iV8W0U2HmmHz3fNLDKdCjKbDxAIKL41nmVipMtYZbjk1VDSTimbyg4srCILAkzs61z3AvndhiZpuM52rtWKbKzex2jq7UObF8SyTK1WCqkTPJvWNL45nOTNf4sBAnAc28YB0HBfbde9a17jUMLm8XMWyXYbbC2rrKDdfB1ZzwTUehO3dEX7mcD9nF8uIApi2S0dIIepX0ExvMVYS4Xee3MoLl1f4m9fnkCU4PVcmqEq8diXPtq4Ii6UGsiTiuNAb9xMLyNQMm1cmcxybyhMPqnxyXy/zhQZn5ku8MpljOldna2eYaEDho3u6r5JvlBomEytVkkGFkE+6SgPvut5r7/R8CfA63b1rZFC6ZW+4fFpqmHzt1AK4Lp/a3/e2XGsuZMoslXUODSUI38GltrL+ljVeqWHf9r8nikLre+Ju8GYuN7zAJ8d1qeoW3zqT4dRsEdf1Ulw/uf/qJeVCzeB7F5aJBhSe3OFFxUf9CstlnaAqIeJppGMBpbXQfTFT4bUrObZ0RpBFkYuZCssVncFk8Kqf193gxEyBH15aoTcW4KcP9TPcESJT0khHfN7B5Rq+fS7DQtHzzP71R0avGz50t+iLBYj6ZHyK1MpHuLhUYWtnpPX9vB6u6/Kd80vYjku2qreL6TZt1tAupq/DcCpEtmrQHfO3fE7vBAcG4lQ1k1zNIBX2EQ+qnqWe5bCvP06mrLWCwLd1RcjVdFaqOsOpIJcyVRzgfKaM63rR3gKeu4EkCnRF/fhlERCQBIGP7OkGQLccSs2H2vPnlxjpCFGoz/HUzk5OzRYRBAHHrTKQDHB8usD+gThzhTq5qjeWH1+urrN96o4FmFiu0hsPcHg4weRKjQODNx6tBlQRx3XpiQXY0RPhvsFEcyHp+riuy+tTBRzX62rfrJiuGxZ/+dosNd3i43u770qww3NnFlku65yeK/EbG7h5LBQbnJwtsqUzvOFY+v3Ow1tTKJJA4pric1dvjKd2phlfrjGWDvHFR8d4ZSLL7/x1Gcdx2dMXbaUjTudqGJZDUJWRJYFYQKFQN0iFfAwmgyiSiCIJvHA5i+O62I6LKHhjd1kE03YQRYF4QCFb1XFcl1TYx9mF8jUdS5euiB/HcTg4lOQju7vpjQdwXJfBZJClsmd7KQoCvjU7A8+fW+LMfInt3ZGWZGqVy0sVss0luotLlVv2qc7XDJ47nQG8g+1Grjhvl9qaNNPqml+/XQ4OJlAkEZ8sbthpPTKSxHZdTwOdDlHVLBqGjSAIDHUE173/v3l2kb9/cxGfLNEX97OrN8ZH93Qzk6/TFfVzbLrAa1fyiILALzw4SEfYx4vjWVYqGisVncPN77Uqi4T96x9vs/k637+4TDrs4yO7u28qAboZx6byXFgss1Bs8PSuLrpjfiJ+mb54oOXBvpbVpoQiie/KVOvj+3o423zdCoLA15q+02cXyvxmc8fmegiCQHfUz3yxsW5S2eb9Qzv05e3RLqavwyNb09w3mCCoSLd9w12LKos8ubOLg0MJFksao+kQuuVQ0y16YgFKDRMJWK7q9CeDfGJ/L7bjYtoOf/3aLF89NU931M+ZhRKmTWvk7douhbrB9q4wuaqJYTtYTQ1jUJV4ZGuK+WID23GZK9TJ1wxOzRaRRIGKZjGQDFComaQjfk7NFphclvnO+WWCqsQTO662mXIcl9FUiMFkgN29nqfrRkVrsW5gOy7RgIJpO4ymwvhksbVsthnbMUEQ2NET4dxCmV2biEjPlLRWN2x8uXZXiulgs5vkk8VWV34t3z6boVA3ubxUZSQV+sDpqsM+maeuE6zxy0dHmMp5dmiG5fDQ1jT/7nMHmMvXeXpXF984tUim1KCqW4RUGcu26Y4GWCzrJEMqhbqJblm4Lmim4y3FuiBJApWGiWY5LXvHqF9GEMC23VbRPZYKwZpi+munFgmoInv64nzh4REEQcCvSDx3ZpEfXlzh0/f18RP7eogFFDrWyJEuLVUAz+rN3fOW9rSmW57PdfM18nas9RRJQGlqkO/kQR48nbRW9yYH0eDt3/4lUeDAwPUP0X5F4ok1NnXdMT+y1HQbCvtaQVKrhedUtsbZ+TI+RaRqeNepSGLLKaTRtPd0XBe9uQdS1kyOz3jSiYfGOhhLh4n45Q0XSN+YLpCrGuSqBgcG460Ez1VM23v9dEZ8112iXotmOtR0T7Im4Mk8KprFydkiBwcT66YSH9vTw8RKlZ6Y/6YL3XeDrogPLR1uBeesXoNvk9fymYN95GvGVe+FNm3atIvpG3Inx6vXEg+qLR2nT5aINm/8sYCCIosU6ybfv7BMOuLDL4v8/lfPcnahzJbOEAPJAJPZKrppI4sgiiK65SIJAssVHVEQGF+u8OXjc4ylQ3z11ALnF7xu9tauCF1RPzP5GpIo8ti2FN8+t8T5xQrp5hZ8XzzIqbkiyZCKr3ktqyyXNb7+5gLLZZ2AKtETC9AV9bNc0fje+WWSIZUP7+wiU9b429fn0C0b23GbaYtpTNslHlSZztWAzS3ofGR3Nx/e2bWpTs5A0xu41DBv+JC/HT6+t6cVIrFR9ynRLPqiAXnDYvuDjPea8fOXr81iWA6f2N/DQ2MpfuSs8IffG+dipsJYOkwyZCCLIpmyxnS+getCtmqwq9vHdL6G47j4FZFEUOVje3pIR1T+/XfHMW2bfM1kuVIgFpAREJjJ1xFESAV8zBbqV11PTbeQRG8JuKpbnFsoM5WrUaqbnF8sk68bfOHhEbqifk7OFjm74MmNHhzt4ORskT29b7kinJgp8IOLK8QCCr9ydAjf2wzGiPgVPnd4kFxNZ+sdPgw6a6QYjnOD//Eu4TgurivguFBtmPzHH01iOy6fvq+PvniA8eUapuNgGw5XVmocGfYOPrplo0oiR0aSzBRq9EQDramHbjr4ZG/qZdjuDaUdY+kwU7kaiaBKYgMd/fcuLHNuoYwsCvzKQ8M3dXTZ2x/DdlwifpmAT7qpzEOVxZt6Zt9N/rd/OM+5hTIjqRD/5qf38VP39TGdq7esXXNVnfHlKls6wxsWzLIkbuqQ0abNB412Mf0eZLWIl0TPX/pKtspsoU5VN5ktNHhkSwrNdBAE8Kky9w8lOHYlj2k7LBYbOC4EVBnbcpjJ11ku62TKGqbtkAh5Wm7T9h5A2arBSkVnJl9nJBXinz69jXTERzrio6KZ9MQCrU1723H50vE5LmYqlBsm+/rjGJZDuWHwu3/7JssVjcPDSXZ0RylrpueXrVtUdYv+RJCVqsYDI0kmsjWO3uJy0GZHoook8un7rr8E+XbQLZtS3SQd8SEIwk1DJD6+t4eFYoPOiP+OTjXeLyyWtJZOeiZfx3ZcXr3ijcsbzT8/PJzktak8QVUiVzNwXZeoqCBKAlG/TFW3sB1Ph5ur6ezoidIZ8ZGrGcwV6nTHAqiyxMRy1Su0TE/fu3rQCiiS5xqyv5cLi2V29ER57nSG+WID03aguVjXMG2yVZ3+eID/8MNxijWvyP5fP72XQ0OJq76uqVwN8DTTxYZF921Ypa2+B2H96+92qJtv6aRrxjsfJy6IIrGAjCSKzJc0KpqFi8tUtkZfPMDWrjAXMmVkSWS02dV/aSLLq5N5BpJBkkGFUt2iVK9wcChBV7Ow2+z9oT/hRxKgO+rbMGDLtL0Tht2UDt2Mx7el2dkdbVplShwd62Bvf4yAIt0VGcftSsguZSpUNYuJ5SqO4xDxK1ctgH7lxDwVzeL0fIlfe2T0Tl56mzbva9rF9HuQQ0MJBMHTbSdDKn45SjSgsFjSUCSBswslFElAFGTGUmF6YwF64gGmc3XMZrfJsh3OZSokgwo7e6JMrlTJVb1CI6hIhJoP6kRQpT8RIFPWqOkWlaZH9BM7OtfJOwAEBAYSAbSwj6NjHUzlalxZqZKt6pQaJtO5Oh1hlZ64n6Wyhm7amM1FtINDCToj/pbv772AZTv85aszFOome/tifHjXxvKFtSiSeE9a/r1TjKXD9MYDVDSDA/1xTMdpaUhX/3t2oUxPLEDDsAkoIpmyTsQn45MEemIBVioa80WNimZzYbHCrp4oAVXGKGt0BFXCPpltnWFSYZW6YZMIqvgViZcm8pxdKOO48JMH+tjVE2E2L3NlpYbtrvovy2zripCt6DQMm1TYR820cRwv8MS6jgXckZEO6oZNV8RPV/TOjMEt2+EvXp2hWDfZPxDjyR03f/3diKAq02ha44XeBV/kB0YTvDyRI6BI3D8U509fnMK0XT6629vv+GdPb2dXT5SuqJ/DI96B+3LTXWU2Xyfs8wpILwzL+zcPDSVQZZGemB+ffGPLt3/59XMcm8ojiyIDySC7eq92EnlieyeJoEpX1H9DB5hVBEFYt7h3s4nm7VjS3a6EbPW+PJIKIYrrPzZT0pjMVhlNvfNOI23eH3xQNdftYvo9yJeOz7Fc1lkq63z6vj6CPpn9/TFyVR2fLHJ2sYIqiwRVmaFkkBcncli2gyQKyCI4ruBZ4nVHOL9U5aN7uqloXtT42fkS//f3L/P5I4N0Rf1cyFQ4MBDjzIK3BHZuvsRCscFQMrQu1VESBZ491Mcrk3n298d44XKWFy6tsFLRUGWRwWSQX3hgsOV88vj2Tl64tIJr2Hxyfxdhn0xNtzgxU6Q75rsreuY7jW45FJoyl0xzGa3N7VHRTFYqGqbtMlfw7OtcF2IBlahfZjpfJx1RqWoWHWF/swNt4woCLlDVLBIhL82zolnMFOr855enAfj4nh5yNYMnd3TSnwiwvTuC48KrkznenCuhWTZm1WZipcbx6QLbuiIkwyqKJJIMKVQ0k2RQ5QeXlinUTfb0RT2rSp/MUzs6ObtQ5tlDG08++uIBfv6BoTv6vWqYdktmlSnpm/qYS0sVXNdzULkWeU239N3Q8suiyNO7utAtm+MzJUbTYVRJpN7UQouiwGhT87zKkZEkL03kGE2HeGRLiohfIR32tUJ4ntrZyeHhJGG/zMRKledOZ0iGVZ491L9uObimewcJu+kuci11w5tEyKLAlk2GxtwKb0wX+PHlLEMdQX7yQO8tF9W3KyGzbJeOsA/rOjZKPkWkI+TDp3yw9jzatLld2sX0ewzHcclVDYBWrO7ESpWyZuG4XjcmFlAYSQXpjPq5sFghVzUQBIgHFHRLwqcIdEb8KLLIlubiztGxDiZXvGADVZZaPtGTKzVKDYOKZuK68INLK/QngpycKfLrj462RqFLZY3XruQp1L3lnelcDVUSKTVMgqrMgYE4n9rfy/Y18oeJFe9zaKaNAHxkTzffvbDMxHIVQYBDQxrfPrtExC/zjx4aIeKX37Ut9+sR8sk8ui3NVLbWcmU4OVvkpYksY+kwH2l21NpsnkLdbAV8rFS1VpdxLB0i5JPpivmRRRGxKbUwLJuhjiAz+TrT2TqZik5/zE/YJ2M5LrmqRlWzGEuH2dsf476BOF89udD0is8RDyjs7ovy0wf7qRsW//q5C8wV6oRUmb5EAEUTEAWB2VyNmUKDgCrilyWGOoLcP5QkHlQpNUxWqjqJkEpZu31Luc0S8Ss8ui3FVLbOg2PXl0adXyw3rSy9lDtBELAch93XdF7LawrIQuP23Tw2Q1kzeXO2RG/cT388yJden2Oh1ODwcJKKZnF0rINtVlrFAAAgAElEQVT9zf2GVyZzvHYljyDA5w8P0h3zs7Mn2pJVnZkv8dqVPCGfRG8iQNjnpbnGggp1w+K1K3ksx2W5rLNc1te5EP3PH9vJH/1wgi2dYQ4Pr3dZ+cqJOX58OUvYJzOcCrVkJHeKVyZzXMiUyZQaPLWz85ZTNj801oHjuBwYjL8tCdlgMogoCqTDvg075Nu6PHvBtdkI13I9S8g2bT7ItIvp9xiiKPDM7i4uZiqtFC/bcUkEVfb1xyjWDFy8pcNndnZx7EoBUfQ6ThG/zIO9MaZydY6MJHhyR1frYTJfaLB67434Zfb0xfgvL09xeckLidmSDjNTqLNQbBALeCl0l5cqOK6X6jW5UiVT0jgxW6Q37mdHd5TPHxnAdl0uLVXI1wx+eHmFwY5Qy8kgGVKZydU4t1jiBxeXOTaVbz00JUHgxfFsyxVBkQRAIBlS+fyRgZverC3b4dximURQXffAvNMcGkpcpY99eSLLpUyFlbLOo1vT77hX7L3OaCrEwaEEVc3i8HASRfK8gQ3b4WN7uokHVWzHZSpX4/lzGW80XdJIhlSyVR1FFGhYDumAQl8iyCsTJrbrYtgOV7JVxlIh8jWD16fzZCs6flWmO+rz3DoA07JpGA66abBQbLC3P8au3ih/+sIkyxWNnpgf3XQwbZeQT+L8YpnvXVhmKlunLxGgfAuWck5Td3s72vlDQ0kODd3YXu/NuSKZUoOZfJ2hjiCxgLqh5jesSjTMppvHXVywXst3zi0xnas37QW9hdB8zaBYM9jZG7sq6GnV6951aclu1jKb9xZIa7pNrqq3JBVlzeS/vjLNUknDBfb1xzf0TR7tDPN/fnb/da91JlfnSrZG2CejW7d+aKrpns1cb9y/LnwGvM53uWEiNd0/bpU/e3GKMwtlXp7M8a9+au+GC9A3oi8e4PhMkX19sQ274p/c10tFs66aDKzlG6cWODVX5PBwkmc2aCRUdYuZXJ3RdGhDTXqbNu9X2sX0e5Ad3VF2dL/V4d3aGebDO7tYqeq8dsUbV4d8Mj+4lGVHTwTNtFtLLx1hlc6on8e3e9Z+q8zkveS5kVSYT9/Xhyx6G/URv8xIKkQ67GMyW0MQBFYqXgfuXz13gYZhM5IKesWO5aCZNrYDvfEAI6kwnzs8yO9/9QzjhSo+RaSim63i0rQd6qZNRbMxbIfjMwUe3pri6V1ddEZ8nJgpcma+RMSvYFguqiyQrxlkmjHpvXH/dTs3PxrPcnKmeJXf7M3QLRvdclrOKW+XswtlTswUSUdUwAHaD41bQRSFdVHLP32o/6rfS6LAWDpMdiDBSxM5GqZDIqgwnAoRDSgEVYmfOTRAzbCZLdRZKWtcXqpQN2ymsnUqmslyWUezHFwsCjWB719cpm541meaZSNLAn5FpKbbrJR1ig2LdNjHnt4YoWbHc6msc3ahjGE5SJLAzp7IphNRc1WdL70xh+26PHuwf50LQqFmkK3qjKRCVxVFpu1Q1+1bCnsZTYX5uxMLhH1eJPsTOzrZ07s+sTGovpWAGLwDCYibYfVgLEvevaXYMBEEz3nn2tCWo6MdBFXP3ahvA1eOwyNJKrpFPKAwsKZYLdZMdNMhHlTZ2RPho3t61n3sKisVrwjf6BBc0SxM26Fh2pjWrdudfPtchqlsHVkU+MIjIwSv0aWnwz5UWSTq94JTbpWpXB3NtJuLsi7X9hxm83VOz5fY1hXZUKbyzbMZTNvh+XNL/NLRoXW6aVEUrvu6c12Xr51aoNQwyZS1dcW067r8/lfPMleos7cvxu99Ytctf31t2tyrtIvpewBBENjb7z0Yk0GFUsMkHfZh2A49sQC9+wI8vCXVMuLfyBv76FgHP76cpSceIBX2PH5TYR+qJLK7L8YDI0meP79EqWFi2Q6W5eA4LookML5cJaTKiIJNUJWQBIG9fV6xP5Wr4bgus/k6uarBY9vTdG73ioaXJ3JUNYuwX8awHNIRX1PH6n3sR/Z088BoEt1yMCyH719cpjPi59UreeYLDSJ+mV99aGTDrp7dlAk4rrthB+taqrrFX7w6TU23+fDOTrZ3R1seq7bjcnK2gE+WrtpsX65ovDFdIB3xcf+azqAsCvTF/SiSiGnD3c1Y+2CTCHlLYFu7wuzvj3PfYIJEUEG3HEI+mYuZCvv7Y/z9Ke8gmK3quK7bkovopo3rguN68eM+WaI75sd2XfyKRH8ySFCV0JpSkomVKqbj6UrPLpSo6lbLVePwcPKGRVqpbnJ2scRwR4jeeICpXJ2KZtEwLC5kKlcV03XD4i9em8GwHHb3RluFSb6m8399+xKuC5/Y38OHxjZXuO/tj/HQlg5M22VLZ/i6tpAhvwJ48rGo7/YOlZvl6V1dDHUE6Yr6OTFdYCZfR5VFPrG/d11oiyqLG8ovVkmFffzM/QPr/nwgGeDgUIJCzeCBkevLYV6dzPHSRI6gKvGLR4fWFbt9iQADpSA+WSTwNhY0VxM0BcFb1l7/995ieVCR0Ez7lnXrn9zfw/cvrrC7N4p/A13zt85mqGgW48tVfvuJLRtK5uYLDQaSwQ0XEG9GLOgllsYC6187lu0yV6hhWC5XstUNP/5KtsbEcpV9A7GW5r1Nm/cD7WL6HuPAYAIHWC7rPDiapNywiDdjmQHGl6vMFercN3B1YEB/Isjnjwy2fq/KIj//wCBlzWzd1J492M+puSKpsBdr3hH20Z8MUKgbVDWbct3kzfkidcPihUsr9CWCWLaDT5bwKRLxoMLx6SJPbPcWjKayNXrjAQY6gnxqXy/xoNoqpFeRRIGpbI3BjiC/dHQYgP/80hTgBTQ4rou4wUPp4a0pIn7Z68SvuSlPrlRpmDY7u6PUDM/iaSARxHW90bDjuPzVsVlSYR+PbktxaCjJG9MFXhzPAuBX3gqg+capBb5/cQXTdviNx0Z5YrvnpNCXCHBsKs/Onuh1x6Ft7gzbuiJEjyhIotAqvOaLDS4teQ4eYZ9EVbcJ+hSSEc93ujceoGbYHI368fskXh7P4eK95n/7iTFOzRb5/sVlPryzk2d2dyOLIvPFBucXKwRVmbBPRmm6hoC3qPfzjwze1P3iG6c9//UTM0W++OgoharBc2cWPU9sWWI6V0MUBS/R0H3Lhm2tRd03z2Q4PV/CcVzG0sFNF9N+ReLzRwbJlDS2dl1/cc5dI/1w3iGjaVUWW4fUkXSIjrCPiF8mFVb5jy9MUDcsfvOxMXzK238vCcL6acdGLBQbLFc0gqpMuWGtK6Z/8/ExtpwNM5wKva0o8md2dXNu0ZN5bNT5Hk2HydUMuqL+dZ97M3x0Tw+PbevEr4gbyjQSQZWKZhELKBsW0tu7Iriuy0g6dMuuIoIg8MtHhzm3UG7J9daiyCIDyRDnFkqMpdcfiEzb4RunFrAcl4VSo3W/b9Pm/UC7ErgHObhGvrHWvqmmW/z9m4s4rstKReezG3Rw1uJXJPyK1IrEPTiU4OBQnL86NovZcNnaHeFT+3uxbIf5YoPlks6VvBfzXKib/M2xWSzH5UNbUiRDKuWG2Xr4q5JIfzKIIHgJadezw3vudIa5Qp3vnLf57x7fQiKk8rG93ZyeKzGWDl9XE+hXpKuixTXTbi5hLQMCpYbJG1MFypqJKAj8kye3sLs3ynyhgdbUQl5eqnJoKIksvfVAkdd0a2RJxLC8lL3l8ltOCq9dydMwbMaXKxRqJsnwzS203qssVzROzBQZSYXes9Hna7Wvruvy1ZPz6KbDxHKVXT1R+uIBUnu72dEVJRVROTZVwHWhpJk0LKf1c4w03WQWShojqTCiKOJvFnCKJBL1KwRUiYVigyd3dJIM+Rhf9jrfq3Ijy3Y4PV9qSk68olUzbcqaidJ87YiC15N8/vxSM0jEYTJXZaWmE/UrXFj0ipEDA3Es221FYIM3ZjcsB9NySARVrmRrvDyRYzh188I6FfaRuoncaa2Lg/22VLu3x3SuhiqLOC78Pz+6wpePzwGgWy6/+9Edd/3zu0C+amD6N06XjPoVPnON5OhWCKjSOv/xtTy8NcWBwfht+VDfaEfjUwd6WSg2rrs4OVtoePp10SvGbcclV/XSRTejvx5IBKg0zA0lOLbjMpAI0hcPbNhkkASBgCpR0ax3xZaxTZu7SfsV/T5CEgUUWUA33U0vxdmOy4vjWVwXXhzP8gsPDrHavDKamkG56ZvcGfHzzO5uprNVQj6Z85kyY6kwAUXiHz+5hRcuZ1vxwIIg8OzBfmqGtU73nClpTK5U2dkTRRA8qUi2avBXx2b41YdH6Iz4eWrn5keAZtOLdypXo9QwGe4I8cKlLJMrVcZXqnRFfLwwnuVje3pwXZfnzy0xV2gALn/8wwkODSX46J5uVFlkeE388+cODxBUJAzb4bFtb3lujy9XKdZNqrqFZVvAvVtMP39uieWyzoXFCoPJ4Ht+aUgQvCAj3XQIqBK7eqNcydWQRYGnd3dRN2wuZCreQqvg6XWf3JHmx+M5ZgsNvn1uCVkUKdR1/uZYgcmVGr98dBhB8N4/3VE/wx1B7htMNIudq3Whf/zCJC9cWiHql/kfP7aD/kSQ//LSFKfmSoRUiaNbUjy6NYUsiWzt8tL28jXPr3qprBNQJH7qYB9/+N3LZEoaH9qSumpkfv9QgrPzJQKqTG8iyI/Hs2QrOktljX398dtOZe2LB7m87C3x9cbu7uLuRhTqBi+NZ1Fliad2vPN+8w3D29/QLRvDtgFl3d+fnC3SFfUxmr47Xst3M1n3Zh73K1VvQTNf01sH0+lcnf5E4KbNF4B/8dWzZMoaQx1B/o9nr17klESBff0xXp7M8ci29T9bURT43OEBFkvex7dp836iXUy/j/ArEj97eJBMWdu0R6okCgwmg0zn6gx3hAiqMj95oJfZfIN9A1cvMAVUiV98cIhvn81wdqFMR0hlsCPIEzvS/PDiChPLVSaAoY5gU5MnrCukbcfly8fnMCyHiZUqnznYz1SuRmfEj2Y6aKZzS7ZLS2WNuUKdYt0gEVSJBTwrsTfnSjiOw2Kpwc6eKPmaZzcoCALP7O7Gdlz+8LuXATg+U+CLj46t+7d9ssSzGzxgDMtLn7Rdl6pm07l+z+ueIepXWC7rBFXpKg/i9zKfvX+A2aZrRVCVr/J2DvnkVnLb5aUKKxWdeFDmlck8Fc0CXH7yQC9/8K2L1AyLF8ez7OiO8Mzubu4bjPPcmQw13eb4TGFD7e5CsQF4GvyqblPRLPLNRUIz4NmzrS7D/uLRIfb0xvjO+SWWKzrgFRt1w+L4dAEXOHYlz7NrOqFbOiN8/sggmmkzlg5zeamKYTn0JQIbdlJvlaAi8f+3d+fRcZXnHce/P+3W4kWWLdtgLBuMF2xjQKEGnMQGB1wXDtA2AXJYQpvTQNKmKTTn9PS0/aM0af9gCU0LKU1Y0wQaEgeXBtIAMcQJBAw22IAhGANeZPAmyZZlrU//uHfEaCRZo2Xm3tE8n3N8PHNnNPPMnbs89533fd5EA2RZBFVo9jQe61k/p8+cRNW4ElrburhhRd/9LxNKioKuQBWlhT2VQ5I9s+0j3v7wMBJcf+7sIQ0CzQU1FWVB/faK4DtI1M7/MI0a+t3d3RwIj6P7w/KtvR833mhoZlxxIVt3NfUaZ5JQVVY85HKAzuUCT6bHmEkVJT2DttJ12dITaGnv7GkxmTW54ritG7MmV/BGQzMzq8tZvWga5SVFTB1fyra9hyktLjjuCUgEA/jaCVq8K0qLuO7cOl7ccZCZk8r7HdgykEMt7Tzy0k66uo3x44qprihh2ZzJQT/t6nK27GpiycyJtHV083tzeh/YCwvEvGlVvLX3cK/KKek466RJbN7VSHVFCdOH0a8yTlYvmsYHB49SO75syGW2olJZWnTc6dwT5tZWMbe2ipa2Ts6uq6axtYM1i2dQV1PB8rk1rHt1D6VFBcysDr7DWZMrmDbI9NRXL5vFus17qKspZ/EJEygsEMtPraHpWAeVpUW9tqXSokI+Mbua7fuPUFJUwLjaSpafUsPsmgrm1lax/0gb9XV9uwQkakPf9+sdNB7tCMpFfmLmqNRfP2/uFH7z7gFAfGpu9luG1yyezs6DweDiZSfXsGphduu0z62tYtehoPznpH5mOEys4wIJ5cbuMCRrFk9j654m5k6tRBKrFtSyZVcTp50w+P5UUFDAny6fzYbf7WfVwn5mx1XQva+9s3tYlUqcy2WyNCohxEV9fb1t3Lgx6jAcQTWC4sKCXqPRP2o+RnlpUU9SfrS9k0df3sXR9i4uOX1GTz+7gy3tfHDwKKdMrRzRT54fNh/jB7/9AKBXRYSh6Oq2IScpr+08xL0bdrBi3lQuO/PjVsX6+np8+4ynjq5uOrq6ewZ9dXZ18+Dz7/NR8zEuXDStZxzC9n1BS/D8sDJOfzq7umloOsaUqtJe3WKSt6XOsPzY4bZOVs6bQklhYa+LzIamVg62tDOvtmrAi5i712/nWEcXVWUft7YPV2LbfH77fr711O8oENx84Tzqj1M5I1O6uy2cqCeaX0Ja2jopLSrod723d3azbW8zU6pKewag5pLntx9gYzg4etXCvlPPmxnNrUEd6ZHUPh9I09EOdh46yuyaip6ZcAeTetwc6XTULreNdDrx0Z7OXNLLZlY/2N95y7Qblv5GoqfW0f0gLJcH8Nbe5p5kurqihOohtp73p3Z8GReeVsvBlvZ+f1JMx3Ba+x564QMamtv4yabdrFo4jUqv6BF7qRd+bZ3dNLV2UFpcyNbdTT3J9Mlp9JP93y0NvLuvheqKEq49Z1ZPUpi8Lb13oIVte4MJiV7b1cQFC3onNtMnjBs0Wbt06Qze/vBwWq3w6WpobKU5rPPc0NQ6aq87FJlI4obieEleSVFBz2RZuWjL7kY6u40tu5tYOX9qn+NbYrbITJlQXsyE8hzu9+Yil6sXU54FuIw5qbqcmsoSjrZ3DbkrRbpSp0vOhkR91+LCAnKkm7FLUV5SyLxpVby77wiLTxjaNpS4QGw82kFnWIs91ZTKMsqKC2nr7Op3Jrx0zJg4bljl2Y5nXElRTwWguA82dUO35MSJbHzvIPOnjR+VbkHOufR4Mu0yprykiGvGYC3RG1ecwoZ39jF/2njKszQlsxtdklizeOAJWI7nMwtr2byzkbm1lQNOujGhvJjrz6ujs9syWr1hqM5fMDWotlMAK+b17ffqctuyOZNZNmfgSWucc5kRn6O8czlixsRxfK7+pMGf6MakmdVBtZrBxLHlt7SokDVLZgz+ROecc2nLqQGIkvYB72fgpWuA/Rl43eGIUywQr3jiFAv0jedM4JVRfs184J8581K3zbivc49vZOIcXyaOm5kW5/XZH4939Mwys0GnV82pZDpTJG1MZ7RmNsQpFohXPHGKBTITT9w+Yzb4Z86/9x+MxzcycY4vzrENJNdi9nizbwxW0nTOOeeccy47PJl2zjnnnHNumDyZDtwTdQBJ4hQLxCueOMUCmYknbp8xG/wz59/7D8bjG5k4xxfn2AaSazF7vFnmfaadc84555wbJm+Zds4555xzbpg8mXbOOeecc26YPJl2zjnnnHNumHwGROdiRNJXzOzfo44jUyRNN7MGSQIuBRYAO4BHzawz2ugyQ1IxsBo4YGa/kXQ1MAH4LzNrzFIMZwHLgElAI/CCmW3Mxns759xYl5cDEONyYpFUCFyWGgvw06gSi7ismxjGMurflaRfAYkdUOH/pwFbzexTI4s4niQ9Y2bnS7oTaAWeAZYC9Wb2uWijywxJa4GXgInAWcDPCGb7+ryZXZSF978DKAWeApqA8cAqoMvMvprp9x+MpImJiwpJFwOLgO0EF1iRn6DieJxOFvf1B/E6lqcjl+KN+/bZn1xav+nKu2Q6TicWSQ8BrwFPp8Ryupldnc1YwnjitG5iE0sYz6h/V5JuApYA95vZ+nDZE2b2+6MSdAxJesrMViX+T1r+SzNbGWVsmZL82SRtNbNFqcsz/P7P9XdxNtDybEu6wPpngguOx4DzgBPN7Ppoo4vfcTpVDqy/WB3LB5OD8cZ6+0yVa+s3XfnYzeOsfk4gayU9F0EsdWZ2TcqyTWGLZRTitG7iFAtk4Lsys9sllQBflHQD8IMRRZgbHpD0XWCnpO8DzxJcUOR0q8QgWiT9HcEJpEHSzcBBoC1L779R0ncITl7NBCevC4BXsvT+6TrXzD4d3n5S0rORRvOxuB2nBxLX9Re3Y/lgci3eXNk+E3Jt/aYlH5PpOJ1Y1kl6HFifFMungXURxALxWjdxigUG/q7+ZyQvambtwF2S7gGuAV4dYZyxZmYPSXoauAioJTgGfdfMxvLn/ixBn+ntwDeB64Ay4IpsvLmZ3STpDOAc4FSCn1XvMbNN2Xj/NJwZnkgXJrosSCoAKqMOLJSRfX8UnRkmTgtiuv7idiwfTK7FG7c8YjC5tn7TknfdPACSTiwTCU4szwNFZvZSBLEsBxaHcTQR9K2cY2a/zXYsYTxnE2zYRUAnYGb2LxHFcgZBv6rE91RjZrdEEUsYTw1wNkG/13eAd6LYZpwbayQtIviZ983wfjmwxMxeiDayQJz3fUmXAr8ws6NJy8qBuXG5SE06507g4z6ycbmY6yNOOUI64pZHDCZOecZoybtkOrxi77MYeNLMPpPlWG4DpgJdwGTgT8xsX6IPXDZjCeP5XnizHZgC7CG4cpxqZn+W5VgSg/OUtHgh8HoU/TwlPWlmqyV9jaB/1+ME/RJ3m9nfZDse58aKuB0HU8V935e0B3gf+BBYC6wzs0PRRtVbOODsHIIBZ4eI8YCzOOUI6Yj7/pMqTnnGaMrHbh5HCEa6JhNBv81sq0/0cZO0BPiRpK9HEEfCKUnxbDGzPw5v/zKCWNYSr8F5JeH/lwMrzawb+I6kDRHF49xYEbfjYKq47/tvmdlKSbOBPyTof9oGPGZmd0UcW2LAWQnBALk3CX7Wv17StTEdcBanHCEdcd9/UsUpzxg1+ZhMvwlcbmZNyQsl/SKCWIoklZhZu5m9July4PsE5dGikLw9/G3SbaU+MdNiODhvoaQHgZMJBpK1hsvLogsp90nqArYQbHs7gGvCPp914f1/MrO/D59bAzQA/2Fmfx5NxC4D4nYcTJUT+76Z7QBuA26TVEtQxz0Ocm3AWZxyhHTEff9JFZs8YzTlYzeP6QSTJ7SnLC/Kdk3GsN/Qe2b2UdKyQuCzZvZwNmMJ3/s0YJuZdSUtKwFWm1lkgxkkFREMzpsX1c+qkmYl3d1jZh2SKoFPmtkTUcQ0Fkg6YmaV4e0HgLfN7BthMv000GxmZ4SP3wh8CdjgyfTYEbfjYKq47/uSLjKzn0cdx0Ak3Q6U03fAWZuZfS3K2PoTpxwhHXHff1LFNc8YqbxLpp1z8ZGSTN9AMOjsy2Ey/ThB/dTbzWyjpPXA/wEzPJl2Lnfk2oA+54YqH7t5OOdiJmxJuQD4XspDDwNXStpLMMBmDzAjy+E554YpHND3Kr3Lfgp4EojdgD7nhsOTaedclMZJ2gzUAS8Dqf0SnwRuIahU8Eh2Q3POjYJcG9Dn3JD1VwLG5RhJpmBK0cT9Ikn7wkLuSPpCeH9z0r+FkuoktUraJOlNSS9Kui78mzpJu1LLBIV/e3Z2P6Ebw1rNbCkwi2DE/1eSHwz7Lb4M3Az8OPvhOedGKDGg7/ykfyvJ8Uk6Mk3SkX6WzZO0PjwPvynpHkkXJZ3Xj0h6K7z9YNLf3Slpd+J8Lun6pL9pl7QlvJ3TtZ6j5C3TY0MLsEjSODNrJfjpbHfKcx5J7Wca9kvdnjTAaw7wE0kFZnafpJ3AJwmmfEbSfKDKzF7M6KdxecfMmiR9FXhM0t0pD98GPGtmB6ScHvDtXD66mI8roCSLqsxpLvtX4A4zewxA0mIz2wL8PLy/Hvjr5BreYQJ9ObAT+BSw3szuA+4LH3+PoOTj/ix+jjHHW6bHjieAPwhvXwX8cKgvYGbvAjcBidqfPwSuTHrKlcN5XefSEc6I9iq9tznM7HUzeyCaqJxzI2FmDamVMcLlsauMkQOmA7sSd8JEejArga3A3QS5gcsAT6bHjsRArTKCvmip04hekdLNY9wAr/MKMD+8/d/AZWFpOoArwvdxblQkKnkk3b/EzB4ys/fMbFE/z7/fK3k45/LUHcAzkp6Q9FeSJqbxN4nGtbXAxZKKMxphnvJkeowws9cIBnFdBfysn6c8YmZLk/7197MbJBVON7O9wOvABZKWAh1mtnWUQ3fOOefcIMLuGQuAHwErgBcklQ70/LB+8xrgp2bWTNDIdmEWQs07nkyPLeuAWxlZV4wzCAaMJCS6engXD+dc3pB0ecqveZsldUu6MRz0/RdJz/03SV+IMFyXJ8xsj5nda2aXAp1An1/wkqwGJgBbwr7Ry/GuHhnhyfTYci/wj2n2o+ojHJB4K/DtpMU/Jriy9S4ezrm8YWZrk3/NA+4CfkUw2Osj4C/Dlj/nskLS6kQ3DUnTgMn0LTaQ7Crgi2ZWZ2Z1wGzgQknlGQ82z3g1jzHEzHYBdw7w8BWSlifd/zLBBBgnS9oElAGHgW+HPyUlXrNR0gtArZntyFDozjkXW5JOBf4BOJegEWof8GvgOuA/IwzNjV3lknYl3b8dOBG4U9KxcNnXw+6YfYQJ80XAlxLLzKxF0gbgErxu/6jy6cSdc865AYQtgc8Dt5rZw0lT3V9CUEXpNIJGjI1mdn9EYTrnIuTdPJxzzrmB3QK8bma9urmFv9S9CHw+kqicc7Hh3Tycc865fkhaAfwRcOYAT/km8CjwXLZics7Fj7dMO+eccykkTSKYJe5aMzvc33PMbBvwBsEsf865POUt084551xfNwBTgbtTprFPLRH6DWBTtoJyzsWPD0B0zjnnnHNumLybh3POOeecc8PkybRzzjnnnHPD5Mm0c84555xzw+TJtHPOOeecc8PkybRzzjnnnHPD5Mm0c84555xzwwI3djcAAAAQSURBVOTJtHPOOeecc8P0/+vW/FSrkEyMAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 864x576 with 16 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# from pandas.tools.plotting import scatter_matrix # For older versions of Pandas\n",
    "from pandas.plotting import scatter_matrix\n",
    "\n",
    "attributes = [\"MEDV\", \"RM\", \"ZN\",\n",
    "              \"LSTAT\"]\n",
    "scatter_matrix(housing[attributes], figsize=(12, 8));"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Out of all the features, \"RM\", or the average number of rooms per dwelling, is the most positively correlated and the \"LSTAT\", or the percent lower status of the popualtion, is the most negatively correlated feature. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Because we are not missing any features, the data cleaning portion of this dataset is non-intensive. We will move on to scaling our data through a pipeline on our training set to be used in our model. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### We will use a standard scaler for Standarization, which will subtract the mean value and then divide it by the standard deviation so that the resulting distribution has unit variance. This will lower the effectiveness of outliers."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Prepare the Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.pipeline import Pipeline\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "\n",
    "full_pipeline = pipeline = Pipeline([\n",
    "    ('std_scaler', StandardScaler()) \n",
    "])\n",
    "\n",
    "housing_prepared = full_pipeline.fit_transform(housing)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### housing_prepared demonstrates the output of scaling the data. The real data training with occur below after train/ test split. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[-0.41978194,  0.28482986, -1.2879095 , ...,  0.44105193,\n",
       "        -1.0755623 ,  0.15968566],\n",
       "       [-0.41733926, -0.48772236, -0.59338101, ...,  0.44105193,\n",
       "        -0.49243937, -0.10152429],\n",
       "       [-0.41734159, -0.48772236, -0.59338101, ...,  0.39642699,\n",
       "        -1.2087274 ,  1.32424667],\n",
       "       ...,\n",
       "       [-0.41344658, -0.48772236,  0.11573841, ...,  0.44105193,\n",
       "        -0.98304761,  0.14880191],\n",
       "       [-0.40776407, -0.48772236,  0.11573841, ...,  0.4032249 ,\n",
       "        -0.86530163, -0.0579893 ],\n",
       "       [-0.41500016, -0.48772236,  0.11573841, ...,  0.44105193,\n",
       "        -0.66905833, -1.15724782]])"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "housing_prepared"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Train Test Split"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### To avoid creating a machine learning model that would just repeat the labels that it had just seen and fail to predict any new unseen data, we must divide and set aside a portion of our data as a test set (X_test, y_test). This will help avoid the situation of overfitting our model."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Since we are at first predicting the MEVD, we will create our test set by setting the MEDV equal to y and the rest of the features as X. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>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>0.00632</td>\n",
       "      <td>18.0</td>\n",
       "      <td>2.31</td>\n",
       "      <td>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</td>\n",
       "      <td>296.0</td>\n",
       "      <td>15.3</td>\n",
       "      <td>396.90</td>\n",
       "      <td>4.98</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</td>\n",
       "      <td>0.469</td>\n",
       "      <td>6.421</td>\n",
       "      <td>78.9</td>\n",
       "      <td>4.9671</td>\n",
       "      <td>2</td>\n",
       "      <td>242.0</td>\n",
       "      <td>17.8</td>\n",
       "      <td>396.90</td>\n",
       "      <td>9.14</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</td>\n",
       "      <td>0.469</td>\n",
       "      <td>7.185</td>\n",
       "      <td>61.1</td>\n",
       "      <td>4.9671</td>\n",
       "      <td>2</td>\n",
       "      <td>242.0</td>\n",
       "      <td>17.8</td>\n",
       "      <td>392.83</td>\n",
       "      <td>4.03</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</td>\n",
       "      <td>0.458</td>\n",
       "      <td>6.998</td>\n",
       "      <td>45.8</td>\n",
       "      <td>6.0622</td>\n",
       "      <td>3</td>\n",
       "      <td>222.0</td>\n",
       "      <td>18.7</td>\n",
       "      <td>394.63</td>\n",
       "      <td>2.94</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</td>\n",
       "      <td>0.458</td>\n",
       "      <td>7.147</td>\n",
       "      <td>54.2</td>\n",
       "      <td>6.0622</td>\n",
       "      <td>3</td>\n",
       "      <td>222.0</td>\n",
       "      <td>18.7</td>\n",
       "      <td>396.90</td>\n",
       "      <td>5.33</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  0.538  6.575  65.2  4.0900    1  296.0   \n",
       "1  0.02731   0.0   7.07     0  0.469  6.421  78.9  4.9671    2  242.0   \n",
       "2  0.02729   0.0   7.07     0  0.469  7.185  61.1  4.9671    2  242.0   \n",
       "3  0.03237   0.0   2.18     0  0.458  6.998  45.8  6.0622    3  222.0   \n",
       "4  0.06905   0.0   2.18     0  0.458  7.147  54.2  6.0622    3  222.0   \n",
       "\n",
       "   PTRATIO       B  LSTAT  \n",
       "0     15.3  396.90   4.98  \n",
       "1     17.8  396.90   9.14  \n",
       "2     17.8  392.83   4.03  \n",
       "3     18.7  394.63   2.94  \n",
       "4     18.7  396.90   5.33  "
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "y = housing[\"MEDV\"]\n",
    "X = housing.drop([\"MEDV\"], axis = 1)\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, random_state=0, test_size = 0.2)\n",
    "\n",
    "X.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Select and Train a Model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### We will train our model using linear regression with regularization. Because many of our features are strongely correlated to our prediction class (\"MEDV\"), we will use regularization that will reduce overfitting by constraining the degrees of freedom and by keeping the model weights as small as possible. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.linear_model import Ridge"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "ridge = Ridge(alpha=0.1, solver = \"cholesky\") #, l1_ratio=0.5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "lr = Pipeline(steps=[('preprocessor', full_pipeline),\n",
    "                      ('classifier', ridge)])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### We will then fit our model to our training data through the pipeline we created in the cell above. The training data first is scaled with the standard scaler to reduce the influence of outliers, then it is trained with ridge regression by computing the weighted sums of the biased term with alpha added to the cost function. If alpha is very large, then all weights end up closing in on zero and the result is a flat line going through the mean. \n",
    "\n",
    "#### When we do ridge regression, we're trying to minimize the following cost function:\n",
    "\n",
    "#### (y−Xβ)T(y−Xβ)+λβTβ"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Pipeline(memory=None,\n",
       "         steps=[('preprocessor',\n",
       "                 Pipeline(memory=None,\n",
       "                          steps=[('std_scaler',\n",
       "                                  StandardScaler(copy=True, with_mean=True,\n",
       "                                                 with_std=True))],\n",
       "                          verbose=False)),\n",
       "                ('classifier',\n",
       "                 Ridge(alpha=0.1, copy_X=True, fit_intercept=True,\n",
       "                       max_iter=None, normalize=False, random_state=None,\n",
       "                       solver='cholesky', tol=0.001))],\n",
       "         verbose=False)"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lr.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_pred = lr.predict(X_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Evaluating Our Trained Model With Root Mean Square Error(RMSE) and Mean Absolute Error (MAE):\n",
    "\n",
    "The mean squared error (MSE) measures the average of the squares of the errors, which is the average squared difference between the estimated values and the actual value. The RMSE and MAE measure the distance between two vectors: the vectors of the predictions and the vector of the target values. Usually, the RMSE score will be larger in datasets that have many outlier. This is clearly happening in our case because we have a higher norm index. \n",
    "\n",
    "**Root Mean Squared Error** (RMSE) is the square root of the mean of the squared errors:\n",
    "\n",
    "$$\\sqrt{\\frac 1n\\sum_{i=1}^n(y_i-\\hat{y}_i)^2}$$\n",
    "\n",
    "**Mean Absolute Error** (MAE) is the mean of the absolute value of the errors:\n",
    "\n",
    "$$\\frac 1n\\sum_{i=1}^n|y_i-\\hat{y}_i|$$"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "5.784276329477451"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.metrics import mean_squared_error\n",
    "\n",
    "y_pred = lr.predict(X_test)\n",
    "lin_mse = mean_squared_error(y_test, y_pred)\n",
    "lin_rmse = np.sqrt(lin_mse)\n",
    "lin_rmse"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3.8431047219898766"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.metrics import mean_absolute_error\n",
    "\n",
    "lin_mae = mean_absolute_error(y_test, y_pred)\n",
    "lin_mae"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Our MAE score tells us that our model can predict the price of the homes in boston worth under 50,000 dollars within 3,800 dollars of its actual value. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Gradient Boosted Regression Tree:\n",
    "\n",
    "\n",
    "#### We will select another model in an attempt to improve our MAE score and make a final selection. Our next model will be a Gradient Boosted Regression Tree (GBRT), which is a generalization of boosting to arbitrary differentiable loss functions. It works by sequentially adding predictors to an ensemble, each one correcting its predecessor. This method then tries to fit the new predictor to the residual errors made by the previous predictor. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.ensemble import GradientBoostingRegressor"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "GradientBoostingRegressor(alpha=0.9, criterion='friedman_mse', init=None,\n",
       "                          learning_rate=0.1, loss='ls', max_depth=2,\n",
       "                          max_features=None, max_leaf_nodes=None,\n",
       "                          min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "                          min_samples_leaf=1, min_samples_split=2,\n",
       "                          min_weight_fraction_leaf=0.0, n_estimators=60,\n",
       "                          n_iter_no_change=None, presort='auto',\n",
       "                          random_state=42, subsample=1.0, tol=0.0001,\n",
       "                          validation_fraction=0.1, verbose=0, warm_start=False)"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gbrt = GradientBoostingRegressor(max_depth=2, n_estimators=120, random_state=42)\n",
    "gbr = Pipeline(steps=[('preprocessor', full_pipeline),\n",
    "                      ('classifier', gbrt)])\n",
    "gbr.fit(X_train, y_train)\n",
    "\n",
    "errors = [mean_squared_error(y_test, y_pred)\n",
    "          for y_pred in gbrt.staged_predict(X_test)]\n",
    "bst_n_estimators = np.argmin(errors) + 1\n",
    "\n",
    "gbr_best = GradientBoostingRegressor(max_depth=2,n_estimators=bst_n_estimators, random_state=42)\n",
    "gbr_best.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "min_error = np.min(errors)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Regularization of Gradient Boosting Regressor with early stopping:\n",
    "\n",
    "#### Early stopping works by stopping the training process as soon as the validation error reaches a global minimum. As the algorithm runs through the epochs, the model learns, and the prediction error (RMSE) on the training set consequentially goes down. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "#gbrt = GradientBoostingRegressor(max_depth=2, warm_start=True, random_state=42)\n",
    "\n",
    "min_val_error = float(\"inf\")\n",
    "error_going_up = 0\n",
    "for n_estimators in range(1, 120):\n",
    "    gbr_best.n_estimators = n_estimators\n",
    "    gbr_best.fit(X_train, y_train)\n",
    "    y_pred = gbr_best.predict(X_test)\n",
    "    val_error = mean_squared_error(y_test, y_pred)\n",
    "    if val_error < min_val_error:\n",
    "        min_val_error = val_error\n",
    "        error_going_up = 0\n",
    "    else:\n",
    "        error_going_up += 1\n",
    "        if error_going_up == 5:\n",
    "            break  # early stopping"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "120\n"
     ]
    }
   ],
   "source": [
    "print(gbrt.n_estimators)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Minimum validation MSE: 23.01919387034353\n"
     ]
    }
   ],
   "source": [
    "print(\"Minimum validation MSE:\", min_val_error)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "4.803650837534946"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lin_mse = mean_squared_error(y_test, y_pred)\n",
    "lin_rmse = np.sqrt(lin_mse)\n",
    "lin_rmse"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### The Gradient Boosting Regressor with early stopping performed significantly better than the ridge model. We will keep this in mind when we make our final selection."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2.865109610833408"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lin_mae = mean_absolute_error(y_test, y_pred)\n",
    "lin_mae"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "feature_importance = gbrt.feature_importances_\n",
    "# make importances relative to max importance\n",
    "feature_importance = 100.0 * (feature_importance / feature_importance.max())\n",
    "sorted_idx = np.argsort(feature_importance)\n",
    "pos = np.arange(sorted_idx.shape[0]) + .5\n",
    "plt.subplot(1, 2, 2)\n",
    "plt.barh(pos, feature_importance[sorted_idx], align='center')\n",
    "plt.yticks(pos, X.columns)\n",
    "plt.xlabel('Relative Importance')\n",
    "plt.title('Variable Importance')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# XGBoost"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Finally, we will experiment with XGBoost. XGBoost is basically an optimized version of Gradient Boosted Trees because it considers the potential loss for all possible splits to create a new branch. XGBoost looks at the distribution of features across all data points in a leaf and uses them to reduce the search space of possible feature splits."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "try:\n",
    "    import xgboost\n",
    "except ImportError as ex:\n",
    "    print(\"Error: the xgboost library is not installed.\")\n",
    "    xgboost = None"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[02:50:30] WARNING: src/objective/regression_obj.cu:152: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "Validation MSE: 21.467616446970794\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/ryanmartin/anaconda3/lib/python3.7/site-packages/xgboost/core.py:587: FutureWarning: Series.base is deprecated and will be removed in a future version\n",
      "  if getattr(data, 'base', None) is not None and \\\n"
     ]
    }
   ],
   "source": [
    "if xgboost is not None:  # not shown in the book\n",
    "    xgb_reg = xgboost.XGBRegressor(random_state=42)\n",
    "    xgb_ = Pipeline(steps=[('preprocessor', full_pipeline),\n",
    "                      ('classifier', xgb_reg)])\n",
    "    xgb_.fit(X_train, y_train)\n",
    "    y_pred = xgb_.predict(X_test)\n",
    "    val_error = mean_squared_error(y_test, y_pred)\n",
    "    print(\"Validation MSE:\", val_error)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "4.63331592350131"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lin_mse = mean_squared_error(y_test, y_pred)\n",
    "lin_rmse = np.sqrt(lin_mse)\n",
    "lin_rmse"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Our MAE score improved slightly with XGBoost, so we will use this model as our final model."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2.7407673218671014"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lin_mae = mean_absolute_error(y_test, y_pred)\n",
    "lin_mae"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Evaluate our best model on the test set"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Below are the results of our final model using XGBoost. This model will basically predict the MEDV within 2.86 of actual MEDV without overfitting. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/ryanmartin/anaconda3/lib/python3.7/site-packages/xgboost/core.py:587: FutureWarning: Series.base is deprecated and will be removed in a future version\n",
      "  if getattr(data, 'base', None) is not None and \\\n",
      "/Users/ryanmartin/anaconda3/lib/python3.7/site-packages/xgboost/core.py:587: FutureWarning: Series.base is deprecated and will be removed in a future version\n",
      "  if getattr(data, 'base', None) is not None and \\\n",
      "/Users/ryanmartin/anaconda3/lib/python3.7/site-packages/xgboost/core.py:587: FutureWarning: Series.base is deprecated and will be removed in a future version\n",
      "  if getattr(data, 'base', None) is not None and \\\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[02:50:30] WARNING: src/objective/regression_obj.cu:152: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[02:50:30] WARNING: src/objective/regression_obj.cu:152: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[02:50:30] WARNING: src/objective/regression_obj.cu:152: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[02:50:30] WARNING: src/objective/regression_obj.cu:152: reg:linear is now deprecated in favor of reg:squarederror.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/ryanmartin/anaconda3/lib/python3.7/site-packages/xgboost/core.py:587: FutureWarning: Series.base is deprecated and will be removed in a future version\n",
      "  if getattr(data, 'base', None) is not None and \\\n",
      "/Users/ryanmartin/anaconda3/lib/python3.7/site-packages/xgboost/core.py:587: FutureWarning: Series.base is deprecated and will be removed in a future version\n",
      "  if getattr(data, 'base', None) is not None and \\\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[02:50:30] WARNING: src/objective/regression_obj.cu:152: reg:linear is now deprecated in favor of reg:squarederror.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/ryanmartin/anaconda3/lib/python3.7/site-packages/xgboost/core.py:587: FutureWarning: Series.base is deprecated and will be removed in a future version\n",
      "  if getattr(data, 'base', None) is not None and \\\n",
      "/Users/ryanmartin/anaconda3/lib/python3.7/site-packages/xgboost/core.py:587: FutureWarning: Series.base is deprecated and will be removed in a future version\n",
      "  if getattr(data, 'base', None) is not None and \\\n",
      "/Users/ryanmartin/anaconda3/lib/python3.7/site-packages/xgboost/core.py:587: FutureWarning: Series.base is deprecated and will be removed in a future version\n",
      "  if getattr(data, 'base', None) is not None and \\\n",
      "/Users/ryanmartin/anaconda3/lib/python3.7/site-packages/xgboost/core.py:587: FutureWarning: Series.base is deprecated and will be removed in a future version\n",
      "  if getattr(data, 'base', None) is not None and \\\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[02:50:30] WARNING: src/objective/regression_obj.cu:152: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[02:50:30] WARNING: src/objective/regression_obj.cu:152: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[02:50:30] WARNING: src/objective/regression_obj.cu:152: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[02:50:30] WARNING: src/objective/regression_obj.cu:152: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[02:50:30] WARNING: src/objective/regression_obj.cu:152: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "Scores: [2.59889548 2.40284154 2.17967535 4.42220455 2.42796543 2.92646844\n",
      " 3.42158029 3.04444426 2.80754835 2.41765229]\n",
      "Mean: 2.8649275997156125\n",
      "Standard deviation: 0.6260301927284697\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/ryanmartin/anaconda3/lib/python3.7/site-packages/xgboost/core.py:587: FutureWarning: Series.base is deprecated and will be removed in a future version\n",
      "  if getattr(data, 'base', None) is not None and \\\n"
     ]
    }
   ],
   "source": [
    "from sklearn.model_selection import cross_val_score\n",
    "\n",
    "def display_scores(scores):\n",
    "    print(\"Scores:\", scores)\n",
    "    print(\"Mean:\", scores.mean())\n",
    "    print(\"Standard deviation:\", scores.std())\n",
    "\n",
    "final_scores = cross_val_score(xgb_,  X_train, y_train,\n",
    "                                scoring=\"neg_mean_squared_error\", cv=10)\n",
    "final_rmse_scores = np.sqrt(-final_scores)\n",
    "display_scores(final_rmse_scores)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'Predicted Y')"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.scatter(y_test, y_pred)\n",
    "plt.xlabel('Y Test')\n",
    "plt.ylabel('Predicted Y')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "final_mse = mean_squared_error(y_test, y_pred)\n",
    "final_rmse = np.sqrt(final_mse)\n",
    "final_mae = mean_absolute_error(y_test, y_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "('MSE:', 21.467616446970794)"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\"MSE:\", final_mse"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "('RMSE:', 4.63331592350131)"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\"RMSE:\", final_rmse"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "('RMSE:', 2.7407673218671014)"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\"RMSE:\", final_mae"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Residual Plot:\n",
    "#### In regression analysis, the difference between the observed value of the dependent variable (y) and the predicted value (ŷ) is called the residual (e). Each data point has one residual.\n",
    "\n",
    "#### Residual = Observed value - Predicted value e = y - ŷ\n",
    "\n",
    "#### Both the sum and the mean of the residuals are equal to zero. That is, Σ e = 0 and e = 0."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.hist((y_test-y_pred),bins=50);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Predicting the NOX"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Because this model uses the same dataset as the housing price model, we will only have to change our X and y values and begin at selecting and training a model. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Select and Train a Model: \n",
    "\n",
    "#### We will train our model using linear regression with regularization. Because many of our features are strongely correlated to our prediction class (\"NOX\"), we will use regularization that will reduce overfitting by constraining the degrees of freedom and by keeping the model weights as small as possible. \n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "y = housing[\"NOX\"]\n",
    "X = housing.drop([\"NOX\"], axis = 1)\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, random_state=0, test_size = 0.2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.linear_model import Ridge"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [],
   "source": [
    "elastic_net = Ridge(alpha=0.1, solver = \"cholesky\") #, l1_ratio=0.5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [],
   "source": [
    "lr = Pipeline(steps=[('preprocessor', full_pipeline),\n",
    "                      ('classifier', elastic_net)])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### We will then fit our model to our training data through the pipeline we created in the cell above. The training data first is scaled with the standard scaler to reduce the influence of outliers, then it is trained with ridge regression by computing the weighted sums of the biased term with alpha added to the cost function. If alpha is very large, then all weights end up closing in on zero and the result is a flat line going through the mean. \n",
    "\n",
    "#### When we do ridge regression, we're trying to minimize the following cost function:\n",
    "\n",
    "#### (y−Xβ)T(y−Xβ)+λβTβ"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Pipeline(memory=None,\n",
       "         steps=[('preprocessor',\n",
       "                 Pipeline(memory=None,\n",
       "                          steps=[('std_scaler',\n",
       "                                  StandardScaler(copy=True, with_mean=True,\n",
       "                                                 with_std=True))],\n",
       "                          verbose=False)),\n",
       "                ('classifier',\n",
       "                 Ridge(alpha=0.1, copy_X=True, fit_intercept=True,\n",
       "                       max_iter=None, normalize=False, random_state=None,\n",
       "                       solver='cholesky', tol=0.001))],\n",
       "         verbose=False)"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lr.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_pred = lr.predict(X_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Evaluating Our Trained Model With Root Mean Square Error(RMSE) and Mean Absolute Error (MAE):\n",
    "\n",
    "The mean squared error (MSE) measures the average of the squares of the errors, which is the average squared difference between the estimated values and the actual value. The RMSE and MAE measure the distance between two vectors: the vectors of the predictions and the vector of the target values. Usually, the RMSE score will be larger in datasets that have many outliers. This is clearly happening in our case because we have a higher norm index. \n",
    "\n",
    "**Root Mean Squared Error** (RMSE) is the square root of the mean of the squared errors:\n",
    "\n",
    "$$\\sqrt{\\frac 1n\\sum_{i=1}^n(y_i-\\hat{y}_i)^2}$$\n",
    "\n",
    "**Mean Absolute Error** (MAE) is the mean of the absolute value of the errors:\n",
    "\n",
    "$$\\frac 1n\\sum_{i=1}^n|y_i-\\hat{y}_i|$$"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.06303098145118664"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.metrics import mean_squared_error\n",
    "\n",
    "y_pred = lr.predict(X_test)\n",
    "lin_mse = mean_squared_error(y_test, y_pred)\n",
    "lin_rmse = np.sqrt(lin_mse)\n",
    "lin_rmse"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### This MAE value seems really low, but it is because our mean NOX score in our dataset is about .55"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.0469936979347405"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.metrics import mean_absolute_error\n",
    "\n",
    "lin_mae = mean_absolute_error(y_test, y_pred)\n",
    "lin_mae"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### We will continue the same model selection process as the previous model because this dataset is the same. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Gradient Boosted Regression Tree:\n",
    "\n",
    "\n",
    "#### We will select another model in an attempt to improve our MAE score and make a final selection. Our next model will be a Gradient Boosted Regression Tree (GBRT), which is a generalization of boosting to arbitrary differentiable loss functions. It works by sequentially adding predictors to an ensemble, each one correcting its predecessor. This method then tries to fit the new predictor to the residual errors made by the previous predictor. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.ensemble import GradientBoostingRegressor"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "GradientBoostingRegressor(alpha=0.9, criterion='friedman_mse', init=None,\n",
       "                          learning_rate=0.1, loss='ls', max_depth=2,\n",
       "                          max_features=None, max_leaf_nodes=None,\n",
       "                          min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "                          min_samples_leaf=1, min_samples_split=2,\n",
       "                          min_weight_fraction_leaf=0.0, n_estimators=21,\n",
       "                          n_iter_no_change=None, presort='auto',\n",
       "                          random_state=42, subsample=1.0, tol=0.0001,\n",
       "                          validation_fraction=0.1, verbose=0, warm_start=False)"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gbrt = GradientBoostingRegressor(max_depth=2, n_estimators=120, random_state=42)\n",
    "gbr = Pipeline(steps=[('preprocessor', full_pipeline),\n",
    "                      ('classifier', gbrt)])\n",
    "gbr.fit(X_train, y_train)\n",
    "\n",
    "errors = [mean_squared_error(y_test, y_pred)\n",
    "          for y_pred in gbrt.staged_predict(X_test)]\n",
    "bst_n_estimators = np.argmin(errors) + 1\n",
    "\n",
    "gbr_best = GradientBoostingRegressor(max_depth=2,n_estimators=bst_n_estimators, random_state=42)\n",
    "gbr_best.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "min_error = np.min(errors)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Regularization of Gradient Boosting Regressor with early stopping:\n",
    "\n",
    "#### Early stopping works by stopping the training process as soon as the validation error reaches a global minimum. As the algorithm runs through the epochs, the model learns, and the prediction error (RMSE) on the training set consequentially goes down. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [],
   "source": [
    "#gbrt = GradientBoostingRegressor(max_depth=2, warm_start=True, random_state=42)\n",
    "\n",
    "min_val_error = float(\"inf\")\n",
    "error_going_up = 0\n",
    "for n_estimators in range(1, 120):\n",
    "    gbr.n_estimators = n_estimators\n",
    "    gbr.fit(X_train, y_train)\n",
    "    y_pred = gbrt.predict(X_test)\n",
    "    val_error = mean_squared_error(y_test, y_pred)\n",
    "    if val_error < min_val_error:\n",
    "        min_val_error = val_error\n",
    "        error_going_up = 0\n",
    "    else:\n",
    "        error_going_up += 1\n",
    "        if error_going_up == 5:\n",
    "            break  # early stopping"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "6\n"
     ]
    }
   ],
   "source": [
    "print(gbr.n_estimators)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Minimum validation MSE: 0.02640118175664736\n"
     ]
    }
   ],
   "source": [
    "print(\"Minimum validation MSE:\", min_val_error)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.16248440465671576"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lin_mse = mean_squared_error(y_test, y_pred)\n",
    "lin_rmse = np.sqrt(lin_mse)\n",
    "lin_rmse"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### The Gradient Boosting Regressor with early stopping performed significantly better than the ridge model. We will keep this in mind when we make our final selection."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.11479034467016486"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lin_mae = mean_absolute_error(y_test, y_pred)\n",
    "lin_mae"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# XGBoost"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Finally, we will experiment with XGBoost. XGBoost is basically an optimized version of Gradient Boosted Trees because it considers the potential loss for all possible splits to create a new branch. XGBoost looks at the distribution of features across all data points in a leaf and uses them to reduce the search space of possible feature splits."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [],
   "source": [
    "try:\n",
    "    import xgboost\n",
    "except ImportError as ex:\n",
    "    print(\"Error: the xgboost library is not installed.\")\n",
    "    xgboost = None"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[02:50:31] WARNING: src/objective/regression_obj.cu:152: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "Validation MSE: 0.002009773625211572\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/ryanmartin/anaconda3/lib/python3.7/site-packages/xgboost/core.py:587: FutureWarning: Series.base is deprecated and will be removed in a future version\n",
      "  if getattr(data, 'base', None) is not None and \\\n"
     ]
    }
   ],
   "source": [
    "if xgboost is not None:  # not shown in the book\n",
    "    xgb_reg = xgboost.XGBRegressor(random_state=42)\n",
    "    xgb_ = Pipeline(steps=[('preprocessor', full_pipeline),\n",
    "                      ('classifier', xgb_reg)])\n",
    "    xgb_.fit(X_train, y_train)\n",
    "    y_pred = xgb_.predict(X_test)\n",
    "    val_error = mean_squared_error(y_test, y_pred)\n",
    "    print(\"Validation MSE:\", val_error)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.044830498828493666"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lin_mse = mean_squared_error(y_test, y_pred)\n",
    "lin_rmse = np.sqrt(lin_mse)\n",
    "lin_rmse"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### As expected, XGBoost performed the best on the NOX class, just like it did with the MEDV class."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.027771743525710765"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lin_mae = mean_absolute_error(y_test, y_pred)\n",
    "lin_mae"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Evaluate our best model on the test set"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Below are the results of our final model using XGBoost. This model will basically predict the NOX within 0.036 of actual NOX without overfitting. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[02:50:31] WARNING: src/objective/regression_obj.cu:152: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[02:50:31] WARNING: src/objective/regression_obj.cu:152: reg:linear is now deprecated in favor of reg:squarederror.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/ryanmartin/anaconda3/lib/python3.7/site-packages/xgboost/core.py:587: FutureWarning: Series.base is deprecated and will be removed in a future version\n",
      "  if getattr(data, 'base', None) is not None and \\\n",
      "/Users/ryanmartin/anaconda3/lib/python3.7/site-packages/xgboost/core.py:587: FutureWarning: Series.base is deprecated and will be removed in a future version\n",
      "  if getattr(data, 'base', None) is not None and \\\n",
      "/Users/ryanmartin/anaconda3/lib/python3.7/site-packages/xgboost/core.py:587: FutureWarning: Series.base is deprecated and will be removed in a future version\n",
      "  if getattr(data, 'base', None) is not None and \\\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[02:50:31] WARNING: src/objective/regression_obj.cu:152: reg:linear is now deprecated in favor of reg:squarederror.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/ryanmartin/anaconda3/lib/python3.7/site-packages/xgboost/core.py:587: FutureWarning: Series.base is deprecated and will be removed in a future version\n",
      "  if getattr(data, 'base', None) is not None and \\\n",
      "/Users/ryanmartin/anaconda3/lib/python3.7/site-packages/xgboost/core.py:587: FutureWarning: Series.base is deprecated and will be removed in a future version\n",
      "  if getattr(data, 'base', None) is not None and \\\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[02:50:31] WARNING: src/objective/regression_obj.cu:152: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[02:50:31] WARNING: src/objective/regression_obj.cu:152: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[02:50:31] WARNING: src/objective/regression_obj.cu:152: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[02:50:31] WARNING: src/objective/regression_obj.cu:152: reg:linear is now deprecated in favor of reg:squarederror.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/ryanmartin/anaconda3/lib/python3.7/site-packages/xgboost/core.py:587: FutureWarning: Series.base is deprecated and will be removed in a future version\n",
      "  if getattr(data, 'base', None) is not None and \\\n",
      "/Users/ryanmartin/anaconda3/lib/python3.7/site-packages/xgboost/core.py:587: FutureWarning: Series.base is deprecated and will be removed in a future version\n",
      "  if getattr(data, 'base', None) is not None and \\\n",
      "/Users/ryanmartin/anaconda3/lib/python3.7/site-packages/xgboost/core.py:587: FutureWarning: Series.base is deprecated and will be removed in a future version\n",
      "  if getattr(data, 'base', None) is not None and \\\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[02:50:31] WARNING: src/objective/regression_obj.cu:152: reg:linear is now deprecated in favor of reg:squarederror.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/ryanmartin/anaconda3/lib/python3.7/site-packages/xgboost/core.py:587: FutureWarning: Series.base is deprecated and will be removed in a future version\n",
      "  if getattr(data, 'base', None) is not None and \\\n",
      "/Users/ryanmartin/anaconda3/lib/python3.7/site-packages/xgboost/core.py:587: FutureWarning: Series.base is deprecated and will be removed in a future version\n",
      "  if getattr(data, 'base', None) is not None and \\\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[02:50:31] WARNING: src/objective/regression_obj.cu:152: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "[02:50:32] WARNING: src/objective/regression_obj.cu:152: reg:linear is now deprecated in favor of reg:squarederror.\n",
      "Scores: [0.02901995 0.03758708 0.04023869 0.03255636 0.03842885 0.04105252\n",
      " 0.02625444 0.04190648 0.03890037 0.0344404 ]\n",
      "Mean: 0.03603851380121747\n",
      "Standard deviation: 0.005031741626112531\n"
     ]
    }
   ],
   "source": [
    "from sklearn.model_selection import cross_val_score\n",
    "\n",
    "def display_scores(scores):\n",
    "    print(\"Scores:\", scores)\n",
    "    print(\"Mean:\", scores.mean())\n",
    "    print(\"Standard deviation:\", scores.std())\n",
    "\n",
    "final_scores = cross_val_score(xgb_,  X_train, y_train,\n",
    "                                scoring=\"neg_mean_squared_error\", cv=10)\n",
    "final_rmse_scores = np.sqrt(-final_scores)\n",
    "display_scores(final_rmse_scores)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'Predicted Y')"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.scatter(y_test, y_pred)\n",
    "plt.xlabel('Y Test')\n",
    "plt.ylabel('Predicted Y')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [],
   "source": [
    "final_mse = mean_squared_error(y_test, y_pred)\n",
    "final_rmse = np.sqrt(final_mse)\n",
    "final_mae = mean_absolute_error(y_test, y_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "('MSE:', 0.002009773625211572)"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\"MSE:\", final_mse"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "('RMSE:', 0.044830498828493666)"
      ]
     },
     "execution_count": 70,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\"RMSE:\", final_rmse"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "('MAE:', 0.027771743525710765)"
      ]
     },
     "execution_count": 71,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\"MAE:\", final_mae"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Residual Plot:\n",
    "#### In regression analysis, the difference between the observed value of the dependent variable (y) and the predicted value (ŷ) is called the residual (e). Each data point has one residual.\n",
    "\n",
    "#### Residual = Observed value - Predicted value e = y - ŷ\n",
    "\n",
    "#### Both the sum and the mean of the residuals are equal to zero. That is, Σ e = 0 and e = 0."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.hist((y_test-y_pred),bins=50);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}