part1.ipynb

{
 "cells": [
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h2>\n",
    "Team leader: Jiaqin Kang<br>\n",
    "Student ID: 33656491<br>\n",
    "Team member : Aromal Shiney<br>\n",
    "student id: 33659221<br>\n",
    "</h2>"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Part 1: A\n",
    "<p> The task given was to create a Simple Nearest Neighbour algorithm with Minkowski distance that would successfully classify on the Sonar data and would seperate Rocks from Metals.After classifying the records on the test data, record its precision, accuracy, recall and F1 measure with respect to class \"M\". The algorithm should be able to produce results for Manhatten Distance and Euclidian Distance based on Minkowski power q.</p>\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "\n",
    "df = pd.read_csv('sonar_train.csv')\n",
    "df2 = pd.read_csv('sonar_test.csv')"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p> The code above creates two dataframes for both training set and test set using the pandas library. These dataframes will be used for data cleaning,preprocessing and building predictive models.</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "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>A1</th>\n",
       "      <th>A2</th>\n",
       "      <th>A3</th>\n",
       "      <th>A4</th>\n",
       "      <th>A5</th>\n",
       "      <th>A6</th>\n",
       "      <th>A7</th>\n",
       "      <th>A8</th>\n",
       "      <th>A9</th>\n",
       "      <th>A10</th>\n",
       "      <th>...</th>\n",
       "      <th>A52</th>\n",
       "      <th>A53</th>\n",
       "      <th>A54</th>\n",
       "      <th>A55</th>\n",
       "      <th>A56</th>\n",
       "      <th>A57</th>\n",
       "      <th>A58</th>\n",
       "      <th>A59</th>\n",
       "      <th>A60</th>\n",
       "      <th>Class</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.0079</td>\n",
       "      <td>0.0086</td>\n",
       "      <td>0.0055</td>\n",
       "      <td>0.0250</td>\n",
       "      <td>0.0344</td>\n",
       "      <td>0.0546</td>\n",
       "      <td>0.0528</td>\n",
       "      <td>0.0958</td>\n",
       "      <td>0.1009</td>\n",
       "      <td>0.1240</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0176</td>\n",
       "      <td>0.0127</td>\n",
       "      <td>0.0088</td>\n",
       "      <td>0.0098</td>\n",
       "      <td>0.0019</td>\n",
       "      <td>0.0059</td>\n",
       "      <td>0.0058</td>\n",
       "      <td>0.0059</td>\n",
       "      <td>0.0032</td>\n",
       "      <td>R</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.0599</td>\n",
       "      <td>0.0474</td>\n",
       "      <td>0.0498</td>\n",
       "      <td>0.0387</td>\n",
       "      <td>0.1026</td>\n",
       "      <td>0.0773</td>\n",
       "      <td>0.0853</td>\n",
       "      <td>0.0447</td>\n",
       "      <td>0.1094</td>\n",
       "      <td>0.0351</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0013</td>\n",
       "      <td>0.0005</td>\n",
       "      <td>0.0227</td>\n",
       "      <td>0.0209</td>\n",
       "      <td>0.0081</td>\n",
       "      <td>0.0117</td>\n",
       "      <td>0.0114</td>\n",
       "      <td>0.0112</td>\n",
       "      <td>0.0100</td>\n",
       "      <td>M</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.0093</td>\n",
       "      <td>0.0269</td>\n",
       "      <td>0.0217</td>\n",
       "      <td>0.0339</td>\n",
       "      <td>0.0305</td>\n",
       "      <td>0.1172</td>\n",
       "      <td>0.1450</td>\n",
       "      <td>0.0638</td>\n",
       "      <td>0.0740</td>\n",
       "      <td>0.1360</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0212</td>\n",
       "      <td>0.0091</td>\n",
       "      <td>0.0056</td>\n",
       "      <td>0.0086</td>\n",
       "      <td>0.0092</td>\n",
       "      <td>0.0070</td>\n",
       "      <td>0.0116</td>\n",
       "      <td>0.0060</td>\n",
       "      <td>0.0110</td>\n",
       "      <td>R</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.0151</td>\n",
       "      <td>0.0320</td>\n",
       "      <td>0.0599</td>\n",
       "      <td>0.1050</td>\n",
       "      <td>0.1163</td>\n",
       "      <td>0.1734</td>\n",
       "      <td>0.1679</td>\n",
       "      <td>0.1119</td>\n",
       "      <td>0.0889</td>\n",
       "      <td>0.1205</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0061</td>\n",
       "      <td>0.0015</td>\n",
       "      <td>0.0084</td>\n",
       "      <td>0.0128</td>\n",
       "      <td>0.0054</td>\n",
       "      <td>0.0011</td>\n",
       "      <td>0.0019</td>\n",
       "      <td>0.0023</td>\n",
       "      <td>0.0062</td>\n",
       "      <td>R</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.0317</td>\n",
       "      <td>0.0956</td>\n",
       "      <td>0.1321</td>\n",
       "      <td>0.1408</td>\n",
       "      <td>0.1674</td>\n",
       "      <td>0.1710</td>\n",
       "      <td>0.0731</td>\n",
       "      <td>0.1401</td>\n",
       "      <td>0.2083</td>\n",
       "      <td>0.3513</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0201</td>\n",
       "      <td>0.0248</td>\n",
       "      <td>0.0131</td>\n",
       "      <td>0.0070</td>\n",
       "      <td>0.0138</td>\n",
       "      <td>0.0092</td>\n",
       "      <td>0.0143</td>\n",
       "      <td>0.0036</td>\n",
       "      <td>0.0103</td>\n",
       "      <td>R</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 61 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       A1      A2      A3      A4      A5      A6      A7      A8      A9  \\\n",
       "0  0.0079  0.0086  0.0055  0.0250  0.0344  0.0546  0.0528  0.0958  0.1009   \n",
       "1  0.0599  0.0474  0.0498  0.0387  0.1026  0.0773  0.0853  0.0447  0.1094   \n",
       "2  0.0093  0.0269  0.0217  0.0339  0.0305  0.1172  0.1450  0.0638  0.0740   \n",
       "3  0.0151  0.0320  0.0599  0.1050  0.1163  0.1734  0.1679  0.1119  0.0889   \n",
       "4  0.0317  0.0956  0.1321  0.1408  0.1674  0.1710  0.0731  0.1401  0.2083   \n",
       "\n",
       "      A10  ...     A52     A53     A54     A55     A56     A57     A58  \\\n",
       "0  0.1240  ...  0.0176  0.0127  0.0088  0.0098  0.0019  0.0059  0.0058   \n",
       "1  0.0351  ...  0.0013  0.0005  0.0227  0.0209  0.0081  0.0117  0.0114   \n",
       "2  0.1360  ...  0.0212  0.0091  0.0056  0.0086  0.0092  0.0070  0.0116   \n",
       "3  0.1205  ...  0.0061  0.0015  0.0084  0.0128  0.0054  0.0011  0.0019   \n",
       "4  0.3513  ...  0.0201  0.0248  0.0131  0.0070  0.0138  0.0092  0.0143   \n",
       "\n",
       "      A59     A60  Class  \n",
       "0  0.0059  0.0032      R  \n",
       "1  0.0112  0.0100      M  \n",
       "2  0.0060  0.0110      R  \n",
       "3  0.0023  0.0062      R  \n",
       "4  0.0036  0.0103      R  \n",
       "\n",
       "[5 rows x 61 columns]"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "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>A1</th>\n",
       "      <th>A2</th>\n",
       "      <th>A3</th>\n",
       "      <th>A4</th>\n",
       "      <th>A5</th>\n",
       "      <th>A6</th>\n",
       "      <th>A7</th>\n",
       "      <th>A8</th>\n",
       "      <th>A9</th>\n",
       "      <th>A10</th>\n",
       "      <th>...</th>\n",
       "      <th>A52</th>\n",
       "      <th>A53</th>\n",
       "      <th>A54</th>\n",
       "      <th>A55</th>\n",
       "      <th>A56</th>\n",
       "      <th>A57</th>\n",
       "      <th>A58</th>\n",
       "      <th>A59</th>\n",
       "      <th>A60</th>\n",
       "      <th>Class</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.0125</td>\n",
       "      <td>0.0152</td>\n",
       "      <td>0.0218</td>\n",
       "      <td>0.0175</td>\n",
       "      <td>0.0362</td>\n",
       "      <td>0.0696</td>\n",
       "      <td>0.0873</td>\n",
       "      <td>0.0616</td>\n",
       "      <td>0.1252</td>\n",
       "      <td>0.1302</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0041</td>\n",
       "      <td>0.0074</td>\n",
       "      <td>0.0030</td>\n",
       "      <td>0.0050</td>\n",
       "      <td>0.0048</td>\n",
       "      <td>0.0017</td>\n",
       "      <td>0.0041</td>\n",
       "      <td>0.0086</td>\n",
       "      <td>0.0058</td>\n",
       "      <td>R</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.0530</td>\n",
       "      <td>0.0885</td>\n",
       "      <td>0.1997</td>\n",
       "      <td>0.2604</td>\n",
       "      <td>0.3225</td>\n",
       "      <td>0.2247</td>\n",
       "      <td>0.0617</td>\n",
       "      <td>0.2287</td>\n",
       "      <td>0.0950</td>\n",
       "      <td>0.0740</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0244</td>\n",
       "      <td>0.0199</td>\n",
       "      <td>0.0257</td>\n",
       "      <td>0.0082</td>\n",
       "      <td>0.0151</td>\n",
       "      <td>0.0171</td>\n",
       "      <td>0.0146</td>\n",
       "      <td>0.0134</td>\n",
       "      <td>0.0056</td>\n",
       "      <td>M</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.0368</td>\n",
       "      <td>0.0279</td>\n",
       "      <td>0.0103</td>\n",
       "      <td>0.0566</td>\n",
       "      <td>0.0759</td>\n",
       "      <td>0.0679</td>\n",
       "      <td>0.0970</td>\n",
       "      <td>0.1473</td>\n",
       "      <td>0.2164</td>\n",
       "      <td>0.2544</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0105</td>\n",
       "      <td>0.0024</td>\n",
       "      <td>0.0018</td>\n",
       "      <td>0.0057</td>\n",
       "      <td>0.0092</td>\n",
       "      <td>0.0009</td>\n",
       "      <td>0.0086</td>\n",
       "      <td>0.0110</td>\n",
       "      <td>0.0052</td>\n",
       "      <td>M</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.0164</td>\n",
       "      <td>0.0173</td>\n",
       "      <td>0.0347</td>\n",
       "      <td>0.0070</td>\n",
       "      <td>0.0187</td>\n",
       "      <td>0.0671</td>\n",
       "      <td>0.1056</td>\n",
       "      <td>0.0697</td>\n",
       "      <td>0.0962</td>\n",
       "      <td>0.0251</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0090</td>\n",
       "      <td>0.0223</td>\n",
       "      <td>0.0179</td>\n",
       "      <td>0.0084</td>\n",
       "      <td>0.0068</td>\n",
       "      <td>0.0032</td>\n",
       "      <td>0.0035</td>\n",
       "      <td>0.0056</td>\n",
       "      <td>0.0040</td>\n",
       "      <td>R</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.0216</td>\n",
       "      <td>0.0124</td>\n",
       "      <td>0.0174</td>\n",
       "      <td>0.0152</td>\n",
       "      <td>0.0608</td>\n",
       "      <td>0.1026</td>\n",
       "      <td>0.1139</td>\n",
       "      <td>0.0877</td>\n",
       "      <td>0.1160</td>\n",
       "      <td>0.0866</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0052</td>\n",
       "      <td>0.0049</td>\n",
       "      <td>0.0096</td>\n",
       "      <td>0.0134</td>\n",
       "      <td>0.0122</td>\n",
       "      <td>0.0047</td>\n",
       "      <td>0.0018</td>\n",
       "      <td>0.0006</td>\n",
       "      <td>0.0023</td>\n",
       "      <td>R</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 61 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       A1      A2      A3      A4      A5      A6      A7      A8      A9  \\\n",
       "0  0.0125  0.0152  0.0218  0.0175  0.0362  0.0696  0.0873  0.0616  0.1252   \n",
       "1  0.0530  0.0885  0.1997  0.2604  0.3225  0.2247  0.0617  0.2287  0.0950   \n",
       "2  0.0368  0.0279  0.0103  0.0566  0.0759  0.0679  0.0970  0.1473  0.2164   \n",
       "3  0.0164  0.0173  0.0347  0.0070  0.0187  0.0671  0.1056  0.0697  0.0962   \n",
       "4  0.0216  0.0124  0.0174  0.0152  0.0608  0.1026  0.1139  0.0877  0.1160   \n",
       "\n",
       "      A10  ...     A52     A53     A54     A55     A56     A57     A58  \\\n",
       "0  0.1302  ...  0.0041  0.0074  0.0030  0.0050  0.0048  0.0017  0.0041   \n",
       "1  0.0740  ...  0.0244  0.0199  0.0257  0.0082  0.0151  0.0171  0.0146   \n",
       "2  0.2544  ...  0.0105  0.0024  0.0018  0.0057  0.0092  0.0009  0.0086   \n",
       "3  0.0251  ...  0.0090  0.0223  0.0179  0.0084  0.0068  0.0032  0.0035   \n",
       "4  0.0866  ...  0.0052  0.0049  0.0096  0.0134  0.0122  0.0047  0.0018   \n",
       "\n",
       "      A59     A60  Class  \n",
       "0  0.0086  0.0058      R  \n",
       "1  0.0134  0.0056      M  \n",
       "2  0.0110  0.0052      M  \n",
       "3  0.0056  0.0040      R  \n",
       "4  0.0006  0.0023      R  \n",
       "\n",
       "[5 rows x 61 columns]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df2.head()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p> The code blocks above displays the first 5 rows of the training dataset and test data. By displaying the datasets we are able to see the structure of the dataset and the class labels. This code block is particularly useful to ensure that the dataset is loaded correctly.</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(139, 61)"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(69, 61)"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df2.shape"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p> The code blocks above checks the shape of both training and test dataset. The code returns the number of rows and columns of each set.</p>"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data Cleansing and Preprocessing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "A1       0\n",
       "A2       0\n",
       "A3       0\n",
       "A4       0\n",
       "A5       0\n",
       "        ..\n",
       "A57      0\n",
       "A58      0\n",
       "A59      0\n",
       "A60      0\n",
       "Class    0\n",
       "Length: 61, dtype: int64"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#check for null values\n",
    "df.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "A1       0\n",
       "A2       0\n",
       "A3       0\n",
       "A4       0\n",
       "A5       0\n",
       "        ..\n",
       "A57      0\n",
       "A58      0\n",
       "A59      0\n",
       "A60      0\n",
       "Class    0\n",
       "Length: 61, dtype: int64"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df2.isnull().sum()\n"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p> The next step will be to check if the datasets need to be cleansed from null values,duplicated values, handling outliers, transforming categorical data. The code blocks above checks for null values in the dataset.</p>"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Seperation of Predictors and Target Variable"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "x_train = df.drop('Class', axis=1)\n",
    "y_train = df['Class']\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "x_test = df2.drop('Class', axis=1)\n",
    "y_test = df2['Class']"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p> The code blocks above is splitting the x_train(the predictors) from the y_train(the target variable) on both datasets. After dropping the target variable called \"class\" from x_train, a new object variable is created to fill the labels(\"class\") called y_train.</p>"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Feature Scaling\n",
    "<p>Feature Scaling is a Machine Learning technique that is used to standardise the range of independent variables in the dataset. This is an important prepreocessing task because some of the algorithms require scaled input features to give acccurate results. We have chossen to use the default StandardScaler from the sklearn's library to scale the x_train and x_test values. Using the StandardScaler() function we can transform any mean and variance units into a scale. The code below demonstrates the steps.</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "#scale the x values\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "scaler = StandardScaler()\n",
    "x_train = scaler.fit_transform(x_train)\n",
    "x_test = scaler.transform(x_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[-0.93165451, -0.8731433 , -0.94288121, ..., -0.33592537,\n",
       "        -0.32740932, -0.73095381],\n",
       "       [ 1.37735619,  0.30647505,  0.21154975, ...,  0.49024938,\n",
       "         0.55903847,  0.75325143],\n",
       "       [-0.86948884, -0.31677691, -0.5207191 , ...,  0.51975562,\n",
       "        -0.31068389,  0.97151691],\n",
       "       ...,\n",
       "       [-1.21584044, -0.56911795, -0.33309149, ..., -1.04407515,\n",
       "        -0.77899592, -0.40355559],\n",
       "       [ 0.54256001, -0.29245489,  0.48777928, ..., -0.45395033,\n",
       "         0.10745186, -0.46903524],\n",
       "       [-0.08353712, -0.63904378, -0.19758267, ...,  0.19518697,\n",
       "         0.44196046,  0.6004656 ]])"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[-0.72739587, -0.67248657, -0.51811316, ..., -0.58672841,\n",
       "         0.12417729, -0.16346357],\n",
       "       [ 1.07096823,  1.55601922,  4.11785224, ...,  0.96234923,\n",
       "         0.92699793, -0.20711666],\n",
       "       [ 0.35162259, -0.28637438, -0.81779614, ...,  0.077162  ,\n",
       "         0.52558761, -0.29442285],\n",
       "       ...,\n",
       "       [ 0.08519828, -0.10395917, -0.55980818, ...,  0.60827434,\n",
       "        -0.37758561, -0.99287238],\n",
       "       [ 1.05320661,  0.57705761,  2.09043173, ...,  0.32796505,\n",
       "         4.77384682,  3.11051858],\n",
       "       [-0.47873318, -0.69072809, -1.01845344, ..., -0.48345657,\n",
       "         0.17435358,  0.42585322]])"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(139, 60)\n",
      "(139,)\n",
      "(69, 60)\n",
      "(69,)\n"
     ]
    }
   ],
   "source": [
    "print(x_train.shape)\n",
    "print(y_train.shape)\n",
    "print(x_test.shape)\n",
    "print(y_test.shape)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p> The code above prints the shape of the newly transformed x_train and x_test to ensure that they are having the same shapes so that they can be fed to the Machine Learning Algorithms.</p>"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Handling Categorical Data\n",
    "<p> Handling Categorical Data is an important part of data prepreocesing. We have categorical data in our dataset. The last column \"class\" in our dataset denotes a categorical data. This column must be converted into numerical values so they are correcly interpreted by the algorithms. There are several ways to encode the categorical data into numerical values. We are using scikit-learn's LabelEncoder function. With the help of this function we are able to convert the \"class\" category into numerical values. Each of the Rock instances will now have \"0\" and Metals will have \"1\" assigned to them. </p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "# label encoding y\n",
    "from sklearn.preprocessing import LabelEncoder\n",
    "le = LabelEncoder()\n",
    "y_train = le.fit_transform(y_train)\n",
    "y_test = le.transform(y_test)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "###  create a simple Nearest Neighbour algorithm with Minkowski Distance with power of Q "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "# cereating the minikowski distance function\n",
    "#d(x,y) = (∑|xi-yi|^p)^(1/p)\n",
    "def minkowski_distance(x, y, p):\n",
    "    #\"zip(x,y)\" = pairs up the elements of \"x\" and \"y\" for each corresponding index creating a new iteratiable tuple\n",
    "    #The \"for\" loop iterates the two tuples produced by the \"zip\" assigning \"i\" th element of \"x\" and \"xi\" and the \"i\" th element of \"y\" to \"yi\"\n",
    "    #The expression of \"abs(xi-yi)**p\" computes the absolute distance between the \"i\" th element of \"x\" and \"y\" raised by the power of \"p\"\n",
    "    #The \"sum\" function adds up all the values to produce the final value of the distance\n",
    "    distance =  (sum(abs(xi-yi)**p for xi, yi in zip(x,y))) # calculating the distance between x and y\n",
    "    distance = distance ** (1/p) # raise the distance to the power of 1/p\n",
    "    return distance # return the distance"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p> The code block above computes the minkowski distance between two datapoints(x and y) using p which is the order of the distance metric. Minkowski Distance is a distance metric that generalises both Euclidian and Manahatten distance. Minkowski distance calculates the distance between two data points in n dimensional space. As the value of \"p\" changes the distance measure is also changed. when p = 1, we are measuring in Manhatten Distance whereas when p = 2, it is set to Euclidian Distance.</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "# creating the Nearest Neighbour Algorithm\n",
    "def nearest_neighbour(x_train, x_test, y_train,  q):\n",
    "    y_pred = [] # creating an empty list to store the predicted values\n",
    "    # for each row in the test set\n",
    "    for i in range(len(x_test)):\n",
    "        distances = []\n",
    "        # for each row in the training set\n",
    "        for j in range(len(x_train)):\n",
    "            distance = minkowski_distance(x_test[i], x_train[j],p = q ) # calculate the distance between the test and training set\n",
    "            distances.append(distance) # append the distance to the list\n",
    "         #get the indices for the knn\n",
    "        indices = np.argsort(distances)[:q]\n",
    "        #get the corresponding labels for the knn\n",
    "        labels = y_train[indices]\n",
    "        #predict the lables on the test set\n",
    "        predict_label =  np.argmax(np.bincount(labels))\n",
    "        y_pred.append(predict_label)\n",
    "    return y_pred"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p> The code block above implements the Simplest Nearest Neighbour algorithm. This algorithm takes four paraemeters, these paraemeters are used to make perdictions on the test set.</p>"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Manhattan Distance\n",
    "<p> Manhattan Distance is a metric used to measure the distance between two data points with respective of the sum of absolute differences of the Cartesian coordinates.</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "q= 1 # manhatten distance\n",
    "y_pred = nearest_neighbour(x_train, x_test, y_train, q) "
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p> The code block above executes the Simplest Nearest Neighbour algorithm using the Manhatten Distance</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1, 0, 0, 1, 1, 0, 0, 0, 0, 0, 1, 1, 1, 1, 0, 0, 0, 1, 0, 0, 0, 1, 1, 1, 1, 0, 1, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 1, 1, 1, 0, 1, 1, 0, 1, 1, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0, 1]\n"
     ]
    }
   ],
   "source": [
    "print(y_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "x1 = pd.DataFrame({'True': y_test, 'Predicted': y_pred})\n",
    "x2 = pd.crosstab(x1['True'], x1['Predicted'])"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p> Using the pandas dataframe we are creating two arrays \"y_test\" and \"y_pred\". The \"y_test\" array contains the true class labels for the test set. whereas the \"y_pred\" array contains the predicted class labels. The pandas crosstab is used to create contingency table for the true and predicted labels.</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "    True  Predicted\n",
      "0      1          1\n",
      "1      0          0\n",
      "2      0          0\n",
      "3      1          1\n",
      "4      1          1\n",
      "..   ...        ...\n",
      "64     0          0\n",
      "65     0          0\n",
      "66     1          1\n",
      "67     0          0\n",
      "68     1          1\n",
      "\n",
      "[69 rows x 2 columns]\n"
     ]
    }
   ],
   "source": [
    "print(x1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predicted   0   1\n",
      "True             \n",
      "0          36   1\n",
      "1           4  28\n"
     ]
    }
   ],
   "source": [
    "print(x2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# create a heatmap for the confusion matrix\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "sns.heatmap(x2, annot=True, cmap='Blues')\n",
    "plt.show()\n"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p> Using the Seaborn library to visualise the confusion matrix of true and predicted variables.</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "TP:  36\n",
      "TN:  28\n",
      "FP:  4\n",
      "FN:  1\n"
     ]
    }
   ],
   "source": [
    "# get the TP, TN, FP, FN\n",
    "TP = x2.iloc[0,0]\n",
    "TN = x2.iloc[1,1]\n",
    "FP = x2.iloc[1,0]\n",
    "FN = x2.iloc[0,1]\n",
    "print(\"TP: \", TP)\n",
    "print(\"TN: \", TN)\n",
    "print(\"FP: \", FP)\n",
    "print(\"FN: \", FN)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Precision:  0.9\n",
      "Recall:  0.972972972972973\n",
      "F1-Score:  0.935064935064935\n",
      "Accuracy:  0.927536231884058\n"
     ]
    }
   ],
   "source": [
    "# calcuate the precision, recall, f1-score and accuracy\n",
    "precision = TP/(TP+FP)\n",
    "recall = TP/(TP+FN)\n",
    "f1_score = 2*(precision*recall)/(precision+recall)\n",
    "accuracy = (TP+TN)/(TP+TN+FP+FN)\n",
    "print(\"Precision: \", precision)\n",
    "print(\"Recall: \", recall)\n",
    "print(\"F1-Score: \", f1_score)\n",
    "print(\"Accuracy: \", accuracy)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p> The code block above calculates the precision, recall, F1-score and accuracy of the Simplest Neighbour Algorithm using confusion matrix. </p>"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Euclidean Distance\n",
    "<p> Euclidean Distance refers to the shortest points between two data points. It is the default distance measure used in K-Nearest Neighbour algorithm."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "q =2 # euclidean distance\n",
    "y_pred = nearest_neighbour(x_train, x_test, y_train, q)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p> Simplest Neighbour Algorithm is taking the Euclidian Distance to measure the distance between the predictors and the target variables.</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1, 0, 0, 1, 1, 0, 0, 0, 0, 0, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 1]\n"
     ]
    }
   ],
   "source": [
    "print(y_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "x1 = pd.DataFrame({'True': y_test, 'Predicted': y_pred})\n",
    "x2 = pd.crosstab(x1['True'], x1['Predicted'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "    True  Predicted\n",
      "0      1          1\n",
      "1      0          0\n",
      "2      0          0\n",
      "3      1          1\n",
      "4      1          1\n",
      "..   ...        ...\n",
      "64     0          0\n",
      "65     0          0\n",
      "66     1          0\n",
      "67     0          0\n",
      "68     1          1\n",
      "\n",
      "[69 rows x 2 columns]\n"
     ]
    }
   ],
   "source": [
    "print(x1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predicted   0   1\n",
      "True             \n",
      "0          37   0\n",
      "1          12  20\n"
     ]
    }
   ],
   "source": [
    "print(x2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# create a heatmap for the confusion matrix\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "sns.heatmap(x2, annot=True, cmap='Blues')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "TP:  37\n",
      "TN:  20\n",
      "FP:  12\n",
      "FN:  0\n"
     ]
    }
   ],
   "source": [
    "# get the TP, TN, FP, FN\n",
    "TP = x2.iloc[0,0]\n",
    "TN = x2.iloc[1,1]\n",
    "FP = x2.iloc[1,0]\n",
    "FN = x2.iloc[0,1]\n",
    "print(\"TP: \", TP)\n",
    "print(\"TN: \", TN)\n",
    "print(\"FP: \", FP)\n",
    "print(\"FN: \", FN)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Precision:  0.7551020408163265\n",
      "Recall:  1.0\n",
      "F1-Score:  0.8604651162790697\n",
      "Accuracy:  0.8260869565217391\n"
     ]
    }
   ],
   "source": [
    "# calcuate the precision, recall, f1-score and accuracy\n",
    "precision = TP/(TP+FP)\n",
    "recall = TP/(TP+FN)\n",
    "f1_score = 2*(precision*recall)/(precision+recall)\n",
    "accuracy = (TP+TN)/(TP+TN+FP+FN)\n",
    "print(\"Precision: \", precision)\n",
    "print(\"Recall: \", recall)\n",
    "print(\"F1-Score: \", f1_score)\n",
    "print(\"Accuracy: \", accuracy)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p> The code block above calculates the precison, recall, F1-score and accuracy using the confusion matrix. The results are displayed above. </p>"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Part 1: B\n",
    "***"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Calculate q from 1 - 20\n",
    "<p> The next task was to measure the accuracy, precision, recall and f1-score based on q values from 1-20. Report the results in a chart and evaluate which of the q values lead to the best accuracy on the test set. </p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "# create a list to store values\n",
    "accuracy_list = []\n",
    "precision_list = []\n",
    "recall_list = []\n",
    "f1_score_list = []"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p> Creating a list for measuring the accuracy, precision, recall and f1-score metrics. </p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "for i in range(1,21):\n",
    "    y_pred = nearest_neighbour(x_train, x_test, y_train, i)\n",
    "    x1 = pd.DataFrame({'True': y_test, 'Predicted': y_pred})\n",
    "    x2 = pd.crosstab(x1['True'], x1['Predicted'])\n",
    "    # get the TP, TN, FP, FN\n",
    "    TP = x2.iloc[0,0]\n",
    "    TN = x2.iloc[1,1]\n",
    "    FP = x2.iloc[1,0]\n",
    "    FN = x2.iloc[0,1]\n",
    "    # calcuate the precision, recall, f1-score and accuracy\n",
    "    precision = TP/(TP+FP)\n",
    "    recall = TP/(TP+FN)\n",
    "    f1_score = 2*(precision*recall)/(precision+recall)\n",
    "    accuracy = (TP+TN)/(TP+TN+FP+FN)\n",
    "    accuracy_list.append(accuracy)\n",
    "    precision_list.append(precision)\n",
    "    recall_list.append(recall)\n",
    "    f1_score_list.append(f1_score)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p> The code block above executes a for loop that goes through all the values of q from 1-20. On each iterations it calls over the nearest neighbour algorithm to get the predictions. It then creates a dataframe for storing the true and predicted labels. It also creates a confusion matrix using the crosstab function. After extracting the information from the confusion matrix it then calculates all the evaluation metrics. </p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "# crate a result table to store the values\n",
    "result = pd.DataFrame({'q': range(1,21),'Accuracy': accuracy_list, 'Precision': precision_list, 'Recall': recall_list, 'F1-Score': f1_score_list})"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p> The result of the evaluation metrics is stored on a pandas dataframe with the corresponding variable names</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "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>q</th>\n",
       "      <th>Accuracy</th>\n",
       "      <th>Precision</th>\n",
       "      <th>Recall</th>\n",
       "      <th>F1-Score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>0.927536</td>\n",
       "      <td>0.900000</td>\n",
       "      <td>0.972973</td>\n",
       "      <td>0.935065</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>0.826087</td>\n",
       "      <td>0.755102</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.860465</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>0.869565</td>\n",
       "      <td>0.818182</td>\n",
       "      <td>0.972973</td>\n",
       "      <td>0.888889</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>0.753623</td>\n",
       "      <td>0.692308</td>\n",
       "      <td>0.972973</td>\n",
       "      <td>0.808989</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>0.782609</td>\n",
       "      <td>0.750000</td>\n",
       "      <td>0.891892</td>\n",
       "      <td>0.814815</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>6</td>\n",
       "      <td>0.739130</td>\n",
       "      <td>0.686275</td>\n",
       "      <td>0.945946</td>\n",
       "      <td>0.795455</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>7</td>\n",
       "      <td>0.724638</td>\n",
       "      <td>0.695652</td>\n",
       "      <td>0.864865</td>\n",
       "      <td>0.771084</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>8</td>\n",
       "      <td>0.695652</td>\n",
       "      <td>0.660000</td>\n",
       "      <td>0.891892</td>\n",
       "      <td>0.758621</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>9</td>\n",
       "      <td>0.637681</td>\n",
       "      <td>0.630435</td>\n",
       "      <td>0.783784</td>\n",
       "      <td>0.698795</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>10</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.640000</td>\n",
       "      <td>0.864865</td>\n",
       "      <td>0.735632</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>11</td>\n",
       "      <td>0.681159</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.810811</td>\n",
       "      <td>0.731707</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>12</td>\n",
       "      <td>0.652174</td>\n",
       "      <td>0.638298</td>\n",
       "      <td>0.810811</td>\n",
       "      <td>0.714286</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>13</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.652174</td>\n",
       "      <td>0.810811</td>\n",
       "      <td>0.722892</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>14</td>\n",
       "      <td>0.681159</td>\n",
       "      <td>0.653061</td>\n",
       "      <td>0.864865</td>\n",
       "      <td>0.744186</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>15</td>\n",
       "      <td>0.695652</td>\n",
       "      <td>0.681818</td>\n",
       "      <td>0.810811</td>\n",
       "      <td>0.740741</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>16</td>\n",
       "      <td>0.681159</td>\n",
       "      <td>0.653061</td>\n",
       "      <td>0.864865</td>\n",
       "      <td>0.744186</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>17</td>\n",
       "      <td>0.681159</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.810811</td>\n",
       "      <td>0.731707</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>18</td>\n",
       "      <td>0.681159</td>\n",
       "      <td>0.653061</td>\n",
       "      <td>0.864865</td>\n",
       "      <td>0.744186</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>19</td>\n",
       "      <td>0.652174</td>\n",
       "      <td>0.651163</td>\n",
       "      <td>0.756757</td>\n",
       "      <td>0.700000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>20</td>\n",
       "      <td>0.637681</td>\n",
       "      <td>0.625000</td>\n",
       "      <td>0.810811</td>\n",
       "      <td>0.705882</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     q  Accuracy  Precision    Recall  F1-Score\n",
       "0    1  0.927536   0.900000  0.972973  0.935065\n",
       "1    2  0.826087   0.755102  1.000000  0.860465\n",
       "2    3  0.869565   0.818182  0.972973  0.888889\n",
       "3    4  0.753623   0.692308  0.972973  0.808989\n",
       "4    5  0.782609   0.750000  0.891892  0.814815\n",
       "5    6  0.739130   0.686275  0.945946  0.795455\n",
       "6    7  0.724638   0.695652  0.864865  0.771084\n",
       "7    8  0.695652   0.660000  0.891892  0.758621\n",
       "8    9  0.637681   0.630435  0.783784  0.698795\n",
       "9   10  0.666667   0.640000  0.864865  0.735632\n",
       "10  11  0.681159   0.666667  0.810811  0.731707\n",
       "11  12  0.652174   0.638298  0.810811  0.714286\n",
       "12  13  0.666667   0.652174  0.810811  0.722892\n",
       "13  14  0.681159   0.653061  0.864865  0.744186\n",
       "14  15  0.695652   0.681818  0.810811  0.740741\n",
       "15  16  0.681159   0.653061  0.864865  0.744186\n",
       "16  17  0.681159   0.666667  0.810811  0.731707\n",
       "17  18  0.681159   0.653061  0.864865  0.744186\n",
       "18  19  0.652174   0.651163  0.756757  0.700000\n",
       "19  20  0.637681   0.625000  0.810811  0.705882"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "result"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAkAAAAGwCAYAAABB4NqyAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjYuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8o6BhiAAAACXBIWXMAAA9hAAAPYQGoP6dpAADhzklEQVR4nOydd3hUZfbHP3cmmUx6JZU0OiH0DiLSQUGxggUEwborll13ZV1pFuS3ilgWlK5YQEVYRFoo0qUEQu8tkEJI72Vm7u+PYYaETJJJMi3wfp7nPknufe/7npkkM98557znSLIsywgEAoFAIBDcRSjsbYBAIBAIBAKBrRECSCAQCAQCwV2HEEACgUAgEAjuOoQAEggEAoFAcNchBJBAIBAIBIK7DiGABAKBQCAQ3HUIASQQCAQCgeCuw8neBjgiOp2O5ORkPD09kSTJ3uYIBAKBQCAwA1mWycvLIzQ0FIWieh+PEEAmSE5OJjw83N5mCAQCgUAgqANXr16lcePG1Y4RAsgEnp6egP4J9PLysrM1AoFAIBAIzCE3N5fw8HDj+3h1CAFkAkPYy8vLSwgggUAgEAgaGOakr4gkaIFAIBAIBHcdQgAJBAKBQCC46xACSCAQCAQCwV2HyAESCAQCwV2JVqulrKzM3mYIaolKpapxi7s5CAEkEAgEgrsKWZZJTU0lOzvb3qYI6oBCoSA6OhqVSlWveYQAEggEAsFdhUH8BAYG4ubmJgreNiAMhYpTUlKIiIio1+9OCCCBQCAQ3DVotVqj+PH397e3OYI60KhRI5KTk9FoNDg7O9d5HpEELRAIBIK7BkPOj5ubm50tEdQVQ+hLq9XWax4hgAQCgUBw1yHCXg0XS/3uRAjsDiAlP4Wskqwqr/u6+BLiEWJDiwQCgUAgcGzs6gHasWMHI0aMIDQ0FEmSWL16dY33bN++nc6dO6NWq2nSpAlfffVVpTErV64kJiYGFxcXYmJiWLVqlRWsdwxS8lMYvno4o9aOqvIYvno4Kfkp9jZVIBAIBAKHwa4CqKCggPbt2/Pll1+aNf7SpUvcf//99OnTh8OHD/Ovf/2LSZMmsXLlSuOYvXv3MmrUKMaMGcORI0cYM2YMTzzxBPv27bPWw7ArWSVZlGpLqx1Tqi2t1kMkEAgEgtqh1cnsvZDB/xKS2HshA61OtrdJglpi1xDYsGHDGDZsmNnjv/rqKyIiIpgzZw4ArVu35uDBg3z88cc8+uijAMyZM4dBgwYxefJkACZPnsz27duZM2cOP/74o8Ufg0AgEAjuLjYcT2H6bydJySk2ngvxVjN1RAxDY62bbrBnzx769OnDoEGD2LBhg1XXutNpUEnQe/fuZfDgwRXODRkyhIMHDxoz+6sas2fPnirnLSkpITc3t8IhEAgEAsHtbDiewsvfHaogfgBSc4p5+btDbDhu3XSDxYsX8+qrr7Jr1y4SExOtulZ13AkVtBuUAEpNTSUoKKjCuaCgIDQaDenp6dWOSU1NrXLemTNn4u3tbTzCw8Mtb7xAIBAIHA5Zliks1Zh15BWXMXXNCUwFuwznpq05SV5xmVnzyXLtwmYFBQX89NNPvPzyywwfPpylS5dWuL5mzRq6dOmCWq0mICCARx55xHitpKSEf/zjH4SHh+Pi4kLz5s1ZtGgRAEuXLsXHx6fCXKtXr66w22ratGl06NCBxYsX06RJE1xcXJBlmQ0bNnDPPffg4+ODv78/w4cP58KFCxXmunbtGqNHj8bPzw93d3e6dOnCvn37uHz5MgqFgoMHD1YY/8UXXxAZGVnr56e2NLhdYLdvfzM8QeXPmxpT3ba5yZMn8+abbxp/zs3NFSJIIBAI7gKKyrTETNlokblkIDW3mLbTNpk1/uSMIbipzH8bXrFiBS1btqRly5Y888wzvPrqq7z77rtIksTvv//OI488wjvvvMOyZcsoLS3l999/N947duxY9u7dy+eff0779u25dOmS0XFgLufPn+enn35i5cqVKJVKQC/K3nzzTdq2bUtBQQFTpkzh4YcfJiEhAYVCQX5+Pn379iUsLIw1a9YQHBzMoUOH0Ol0REVFMXDgQJYsWUKXLl2M6yxZsoRx48ZZvVRBgxJAwcHBlTw5aWlpODk5GSt6VjXmdq9QeVxcXHBxcbG8wTbg94u/1zxIIBAIBA2eRYsW8cwzzwAwdOhQ8vPz2bJlCwMHDuSDDz5g9OjRTJ8+3Ti+ffv2AJw9e5affvqJuLg4Bg4cCECTJk1qvX5paSnLli2jUaNGxnOG/NvyNgYGBnLy5EliY2P54YcfuHHjBgcOHMDPzw+AZs2aGcdPnDiRl156idmzZ+Pi4sKRI0dISEjg119/rbV9taVBCaCePXvy22+/VTi3adMmunTpYiyH3bNnT+Li4njjjTcqjOnVq5dNbbUFP57+kW9PfmvWWGu7EgUCgaAh4uqs5OSMIWaN3X8pk3FLDtQ4bun4rnSL9jNrbXM5c+YM+/fvNwoDJycnRo0axeLFixk4cCAJCQk8//zzJu9NSEhAqVTSt29fs9czRWRkZAXxA3DhwgXeffdd/vzzT9LT09HpdAAkJiYSGxtLQkICHTt2NIqf2xk5ciR//etfWbVqFaNHj2bx4sX069ePqKioetlqDnYVQPn5+Zw/f97486VLl0hISMDPz4+IiAgmT55MUlIS336rf5N/6aWX+PLLL3nzzTd5/vnn2bt3L4sWLaqwu+u1117j3nvvZdasWTz00EP873//Y/Pmzezatcvmj8+a/Hz2Zz7c9yEASkmJVq6+JPhvF3+jTUAbW5gmEAgEDQZJkswOQ/Vp3ogQbzWpOcUm84AkINhbTZ/mjVAqLBu+WbRoERqNhrCwMOM5WZZxdnYmKysLV1fXKu+t7hrou6vf/iHZVJKzu7t7pXMjRowgPDycBQsWEBoaik6nIzY2ltLSUrPWVqlUjBkzhiVLlvDII4/www8/GHd6Wxu7CqCDBw/Sr18/48+GPJxnn32WpUuXkpKSUiHLPTo6mnXr1vHGG2/w3//+l9DQUD7//PMKLrhevXqxfPly/v3vf/Puu+/StGlTVqxYQffu3W33wKzMqnOrmLF3BgDPxjzLU62eIrs02+TY3y/8zrenvuX7U9/j6+LLi+1ftKGlAoFAcOegVEhMHRHDy98dQoIKIsggd6aOiLG4+NFoNHz77bd88sknlXY5P/roo3z//fe0a9eOLVu2MH78+Er3t23bFp1Ox/bt240hsPI0atSIvLw8CgoKjCInISGhRrsyMjI4deoUX3/9NX369AGo5Gxo164dCxcuJDMzs0ov0MSJE4mNjWXu3LmUlZVVSN62KrKgEjk5OTIg5+Tk2NuUSqw5v0Zuu7StHLs0Vv5o30eyTqer8Z5FxxbJsUtj5dilsfKiY4tsYKVAIBA4JkVFRfLJkyfloqKiOs+x/liy3OPDzXLkP9cajx4fbpbXH0u2oKW3WLVqlaxSqeTs7OxK1/71r3/JHTp0kLdt2yYrFAp5ypQp8smTJ+WjR4/Ks2bNMo4bN26cHB4eLq9atUq+ePGivG3bNnnFihWyLMtyRkaG7O7uLk+aNEk+d+6c/P3338uhoaFyeYkwdepUuX379hXW1mq1sr+/v/zMM8/I586dk7ds2SJ37dpVBuRVq1bJsizLJSUlcosWLeQ+ffrIu3btki9cuCD/8ssv8p49eyrM1atXL1mlUskvvfRSjc9Hdb/D2rx/CwFkAkcVQL9f+F1u9007OXZprPze3vfMEj8Gvj7ytVEEfXP8GytaKRAIBI6LJQSQLMuyRquT95xPl1cfvibvOZ8ua7Tmvx7XluHDh8v333+/yWvx8fEyIMfHx8srV66UO3ToIKtUKjkgIEB+5JFHjOOKiorkN954Qw4JCZFVKpXcrFkzefHixcbrq1atkps1ayar1Wp5+PDh8vz582sUQLIsy3FxcXLr1q1lFxcXuV27dvIff/xRQQDJsixfvnxZfvTRR2UvLy/Zzc1N7tKli7xv374K8yxatEgG5P3799f4fFhKAEmyLLJjbyc3Nxdvb29ycnLw8vKytzkAbLy8kX/u+CdaWcujzR9lSs8pKKTalXGamzCXeUfmAfB2t7d5uvXT1jBVIBAIHJbi4mIuXbpEdHQ0arXa3uYIbvLBBx+wfPlyjh07VuPY6n6HtXn/blCFEO9WtiRu4e0db6OVtTzU9KE6iR+Al9u/zPNt9bsEPtr/ET+d+cnSpgoEAoFAYDb5+fkcOHCAL774gkmTJtl0bSGAHJztV7fz9+1/RyNrGN5kONN7Ta+T+AH9bodXO77K+Db6JLn3/nyPlWdX1nCXQCAQCATW4a9//Sv33HMPffv25bnnnrPp2kIAOTC7knbxxh9voNFpGBY1jPd6v4dSYX7dCFNIksQbnd/gmdb6YlrT907nf+f/ZwlzBQKBQCCoFUuXLqWkpIQVK1YYq0vbCiGAHJQ9yXt4betrlOnKGBQ5iA/7fIiTwjJVCyRJ4h9d/8HolqORkXl397usvbjWInMLBAKBQNAQEALIAdmfsp9JWydRqiulX3g/Zt07y2Lix4AkSUzuPpnHWjyGjMw7u95hw+UNFl1DIBAIBAJHRQggByP+ejx/3fpXSrQl9Anrw8d9P8ZZ4WyVtRSSgnd7vMvDzR5GJ+t4e8fbbL6y2SprCQQCgUDgSAgB5EAkpCXwyuZXKNIU0Su0F5/2+xSVUmXVNRWSgqk9pzKiyQi0spa3tr/FtsRtVl1TIBAIBAJ7IwSQg3DsxjFe2vwShZpCuod057N+n+GitE2HeqVCyXu932NY9DA0soY3t7/Jjms7bLK2QCAQCAT2QAggB+BExglejHuRgrICugR14Yv+X6B2sm2BLqVCyYf3fMjgyMFodBre2PYGe5L22NQGgUAgEDgWUVFRZjcnrc1YR8CuzVAFcDrzNC9seoG8sjw6BXbivwP+i6tT9d1zrYWTwomP7v0I7XYtWxK3MGnbJL4c8CU9QnrYxR6BQCBwOHKuQUF61dfdG4F3WNXX68G4ceP45ptvAHByciI8PJxHHnmE6dOnm+zUbgkOHDhg9ty1GesICAFkA1LyU8gqyap0PjE3kel7p5Nflk+7Ru2YO3Aubs5udrDwFs4KZ/5z739444832H5tO3/d/Ffe7v42Mf4xJsf7uvgS4hFS5/Wqem4sNb9AIBBYDE0JzO8HBWlVj/EIhNePg5N1UhiGDh3KkiVLKCsrY+fOnUycOJGCggLmzZtXYVxZWRnOzvXfQNOoUSOrjHUEhACyMin5KQxfPZxSbWmVYyQkpvWchruzYyhnZ6Uzs++bzYtxL3Lw+kGm751e5ViVUsXakWvrJFLMeW7qM79AIBBYFKVK790pSAd0JgYowCtMP85KuLi4EBwcDMBTTz3Ftm3bWL16NUFBQaxevZpJkybx/vvvc/nyZbRaLbm5ubz11lusXr2a4uJiunTpwqeffkr79u2Nc65Zs4YZM2Zw/PhxPDw8uPfee/n1118BfVjr9ddf5/XXXwdg2rRpLF68mOvXr+Pv789jjz3G559/bnJsYmIir776Klu2bEGhUDB06FC++OILgoKCjHOtXr2av/3tb7z77rtkZWUxbNgwFixYgKenp9WeQwMiB8jKZJVkVfsGDyAjU6Yrs5FF5qFSqnit02s1jivVllbrwakOc56b+swvEAgENSLLUFpg3lFWCPe+hWnxg/78vW/px5kznwV6kbu6ulJWpn//OH/+PD/99BMrV64kISEBgAceeIDU1FTWrVtHfHw8nTp1YsCAAWRmZgLw+++/88gjj/DAAw9w+PBhtmzZQpcuXUyu9csvv/Dpp5/y9ddfc+7cOVavXk3btm1NjpVlmZEjR5KZmcn27duJi4vjwoULjBo1qsK4CxcusHr1atauXcvatWvZvn07H330Ub2fF3MQHiBBlVh7C75AIBDYnbJC+DDUcvMtf8r8sf9KBlXdPf/79+/nhx9+YMCAAQCUlpaybNkyYyhq69atHDt2jLS0NFxc9CG5jz/+mNWrV/PLL7/wwgsv8MEHHzB69GimT7/l6S/vHSpPYmIiwcHBDBw4EGdnZyIiIujWrZvJsZs3b+bo0aNcunSJ8PBwAJYtW0abNm04cOAAXbt2BUCn07F06VKjx2fMmDFs2bKFDz74oM7Pi7kIASSoNy/GvVinStUancYK1ggEAsGdy9q1a/Hw8ECj0VBWVsZDDz3EF198wdy5c4mMjKyQhxMfH09+fj7+/v4V5igqKuLChQsAJCQk8Pzzz5u19uOPP86cOXNo0qQJQ4cO5f7772fEiBE4OVV+/T916hTh4eFG8QMQExODj48Pp06dMgqgqKioCuGukJAQ0tKqybGyIEIACepNdkm2vU0QCASCuuHspvfE1AZZhqX3Q+pxkLUgKSE4FsatA0mq3dq1pF+/fsybNw9nZ2dCQ0MrJDrfvgNLp9MREhLCH3/8UWkeHx8fQB9CM5fw8HDOnDlDXFwcmzdv5pVXXuE///kP27dvr5RwLcsykonn4vbzt98nSRI6XVUhRssiBJCg3vxfn/+jiU+TWt93Mfsi/9j5DytYJBAIBGYiSXULQw2YAt89qv9e1up/dvGwrG0mcHd3p1mzZmaN7dSpE6mpqTg5OREVFWVyTLt27diyZQvjx483a05XV1cefPBBHnzwQf7yl7/QqlUrjh07RqdOnSqMi4mJITExkatXrxq9QCdPniQnJ4fWrVubtZa1EQJIUG8ivSNp6dey1vdpZa0VrBEIBAIb0HQAhHaE5MP6r00H2NuiSgwcOJCePXsycuRIZs2aRcuWLUlOTmbdunWMHDmSLl26MHXqVAYMGEDTpk0ZPXo0Go2G9evX849/VP5wunTpUrRaLd27d8fNzY1ly5bh6upKZGSkybXbtWvH008/zZw5c9BoNLzyyiv07du3yiRrWyN2gQkEAoFAUFskCQZMhYCW+q+1CX3ZCEmSWLduHffeey/PPfccLVq0YPTo0Vy+fNm4Ff2+++7j559/Zs2aNXTo0IH+/fuzb98+k/P5+PiwYMECevfubfQc/fbbb5VyjAxrr169Gl9fX+69914GDhxIkyZNWLFihVUfc22QZNkC+/DuMHJzc/H29iYnJwcvL696zdWQa91Y2/aG/NwIBIKGSXFxMZcuXSI6Ohq12rYthwSWobrfYW3ev0UIzMqEeISwduTaBlnt2Nq2m5p/X8o+ZsfPJtA1kM/7f46f2s8hnxuBQCAQNGyEALIBIR4hDfZN3Nq23z5/lFcUcxPmklaUhkJSNNjnTSAQCASOjcgBEjgUbs5u9GncB4C4K3F2tkYgEAgEdypCAAkcjoERAwG9ABIpagKBQCCwBkIACRyOexvfi0qh4nLuZc5nn7e3OQKBQCC4AxECSOBweKg86BXWC4DNVzbb2RqBQCAQ3IkIASRwSAZFDgJg05VNdrZEIBAIBHciQgAJHJK+jfvipHDifPZ5LuVcsrc5AoFAILjDEAJI4JB4u3jTI6QHIMJgAoFAILA8QgAJHBZDGExshxcIBALHICoqijlz5hh/NrS8aIgIASRwWPqF90MpKTmVeYqreVftbY5AIBCQkp/CyYyTVR4p+SlWW3vcuHFIkoQkSTg5OREREcHLL79MVlbV1foFVSMqQQscFl+1L12Cu7AvZR+br2xmfOx4e5skEAjuYhyhf+HQoUNZsmQJGo2GkydP8txzz5Gdnc2PP/5olfXuZOzuAZo7d66xoVnnzp3ZuXNnteP/+9//0rp1a1xdXWnZsiXffvtthetLly41KuTyR3FxsTUfhsBKDI4cDIgwmEAgsD9ZJVnVih+AUm1ptf0T64uLiwvBwcE0btyYwYMHM2rUKDZturVbdsmSJbRu3Rq1Wk2rVq2YO3duhfuvXbvG6NGj8fPzw93dnS5duhi7v1+4cIGHHnqIoKAgPDw86Nq1K5s337k5mHb1AK1YsYLXX3+duXPn0rt3b77++muGDRvGyZMniYiIqDR+3rx5TJ48mQULFtC1a1f279/P888/j6+vLyNGjDCO8/Ly4syZMxXuFV1/Gyb9I/rz/p/vcyz9GCn5KaI3mEAgsCiyLFOkKTJrbLHGvA/SxZpiCssKaxzn6uSKJElmzWmKixcvsmHDBpydnQFYsGABU6dO5csvv6Rjx44cPnyY559/Hnd3d5599lny8/Pp27cvYWFhrFmzhuDgYA4dOoROpwMgPz+f+++/n/fffx+1Ws0333zDiBEjOHPmjMn35IaOXQXQ7NmzmTBhAhMnTgRgzpw5bNy4kXnz5jFz5sxK45ctW8aLL77IqFGjAGjSpAl//vkns2bNqiCAJEkiODjYNg9CYFUCXAPoFNSJ+OvxbE7czJiYMfY2yUhKfkq1n/R8XXyFYBMIHJwiTRHdf+hu0Tmf3fCsWeP2PbUPN2e3Ws29du1aPDw80Gq1xsjG7NmzAXjvvff45JNPeOSRRwCIjo7m5MmTfP311zz77LP88MMP3LhxgwMHDuDn5wdAs2bNjHO3b9+e9u3bG39+//33WbVqFWvWrOGvf/1rrexsCNhNAJWWlhIfH8/bb79d4fzgwYPZs2ePyXtKSkoqeXJcXV3Zv38/ZWVlRhWcn59PZGQkWq2WDh068N5779GxY8cqbSkpKaGkpMT4c25ubl0flsAKDIocRPz1eOKuxDmMAHKEXACBQHD30a9fP+bNm0dhYSELFy7k7NmzvPrqq9y4cYOrV68yYcIEnn/+eeN4jUaDt7c3AAkJCXTs2NEofm6noKCA6dOns3btWpKTk9FoNBQVFZGYmGiTx2Zr7CaA0tPT0Wq1BAUFVTgfFBREamqqyXuGDBnCwoULGTlyJJ06dSI+Pp7FixdTVlZGeno6ISEhtGrViqVLl9K2bVtyc3P57LPP6N27N0eOHKF58+Ym5505cybTp0+3+GMUWIaBEQP5aP9HJKQlkFaYRqBboL1NqlUugBBAAoHj4urkyr6n9pk19nTmabO8O98M/YZWfq3MWru2uLu7G702n3/+Of369WP69OlGD82CBQvo3r2iR0upVOrXc61+vbfeeouNGzfy8ccf06xZM1xdXXnssccoLa3+ta6hYvddYLfHP2VZrjIm+u6775KamkqPHj2QZZmgoCDGjRvH//3f/xl/wT169KBHjx7Ge3r37k2nTp344osv+Pzzz03OO3nyZN58803jz7m5uYSHh9f3oQksRJB7EO0btefIjSNsSdzCk62etLdJAoHgDkGSJLPDUGon83JJ1U7qWoe26srUqVMZNmwYL7/8MmFhYVy8eJGnn37a5Nh27dqxcOFCMjMzTXqBdu7cybhx43j44YcBfTTl8uXL1jTfrthtF1hAQABKpbKStyctLa2SV8iAq6srixcvprCwkMuXL5OYmEhUVBSenp4EBASYvEehUNC1a1fOnTtXpS0uLi54eXlVOASOhSiKKBAIBJW57777aNOmDR9++CHTpk1j5syZfPbZZ5w9e5Zjx46xZMkSY47Qk08+SXBwMCNHjmT37t1cvHiRlStXsnfvXkCfD/Trr7+SkJDAkSNHeOqpp4wJ0ncidhNAKpWKzp07ExdX8Q0tLi6OXr16VXuvs7MzjRs3RqlUsnz5coYPH45CYfqhyLJMQkICISEiDNGQGRg5EID46/FkFGXY2RqBQHA34uvii0qpqnaMSqnC18XXRhbpefPNN1mwYIExTcSQBtK3b1+WLl1KdHS03jaVik2bNhEYGMj9999P27Zt+eijj4wRlE8//RRfX1969erFiBEjGDJkCJ06dbLpY7ElkizLsr0WX7FiBWPGjOGrr76iZ8+ezJ8/nwULFnDixAkiIyOZPHkySUlJxlo/Z8+eZf/+/XTv3p2srCxmz55NXFwc8fHxREVFATB9+nR69OhB8+bNyc3N5fPPP2fZsmXs3r2bbt26mWVXbm4u3t7e5OTkCG+QAzF67WhOZJxgSs8pPN7icbvacjLjJKPWjqpx3IrhK4jxj7GBRQKBwByKi4u5dOmSsf5cbRG7P+1Pdb/D2rx/2zUHaNSoUWRkZDBjxgxSUlKIjY1l3bp1REZGApCSklIh+1yr1fLJJ59w5swZnJ2d6devH3v27DGKH4Ds7GxeeOEFUlNT8fb2pmPHjuzYscNs8SNwXAZFDuJExgk2X9lsdwEkEAjuTkI8QoTAuUOwqwfIUREeIMckMTeRB1Y9gFJS8scTf+Cj9rGbLcIDJBA0TOrrARLYH0t5gOzeCuNuI3f9es7e04fcDRvsbUqDI8Irgpa+LdHKWrZd3WZXWxw1F0AgEAgE5iEEkA3RZGSQMmUq2vR0UqZMRZMhknlri2E32OZE+/anCfEI4ZO+nwCgQMHn/T7nh/t/wMPZA4B3e7wriiAKBAKBAyMEkI2QZZnUadPQFer7w+gKCkidJoov1haDANqTvIe80jy72rLy3EoAhjcdTr+IfrRt1JbBUfrmrWezzgrxIxAIBA6MEEA2Im/9evLiNoNWqz+h1ZIXF0fu+vX2NayB0cSnCU29m6LRafjj6h92s+N05mn+uPoHEhIT2040njcItC2JW9DqtHayTiAQCAQ1IQSQDdBkZJAydRrcXuFakkiZOk2EwmqJoSbQ5iv2C4PNPzofgKFRQ4n2jjae7x7cHU+VJ+lF6STcSLCTdQKBQCCoCSGArEyF0NftG+5kWYTC6oDBy7IraRcFZQU2X/9c1jljReoX2r1Q4Zqz0pl+4f0AUbVaIBAIHBkhgKxMyblzFUNft3MzFFZSTasOQUVa+LYg0iuSUl0pO6/ttPn6C44uAPRCrJlvs0rXjYnaVzajk+/cMvICgUDQkBECyMq4NG+O56CBcLPUeCWUSjwHDcKlik71gspIksTACH0YzNZelos5F9lwWV/C4Hbvj4GeoT1xc3LjeuF1jqUfs6V5AoHAhoiyJg0bIYCsjCRJBE+bhsLNzWQOkMLdneBpU+1jXANmUJTey7IzaSdFmiKbrbvw6EJkZO4Lv49Wfq1MjnFRutA3vC9g3zwlgUBgPexR1mTcuHFIklTpOH/+PDt27GDEiBGEhoYiSRKrV682a87Dhw8zfPhwAgMDUavVREVFMWrUKNLT0637YBwAIYBsgJO/P9cnvm4yB+j6xNdx8ve3i10NmRi/GMI8wijSFLEnaY9N1kzMTWTdpXUAvNTupWrHDo7Ub4ePuxKHKLYuENxZ2LOsydChQ0lJSalwREdHU1BQQPv27fnyyy/NnistLY2BAwcSEBDAxo0bOXXqFIsXLyYkJITCm4/NGpSVlVlt7togBJAN2HA8hWcverE7JBatdOspv+IRyLMXvdhwPMWO1jVMyofBNl3ZZJM1Fx5biFbWck/YPbQJaFPt2N5hvXF1ciUpP4mTmSdtYp9AILAN9ixr4uLiQnBwcIVDqVQybNgw3n//fR555BGz59qzZw+5ubksXLiQjh07Eh0dTf/+/ZkzZw4RERHGcSdOnOCBBx7Ay8sLT09P+vTpw4ULFwDQ6XTMmDGDxo0b4+LiQocOHdhQLiR4+fJlJEnip59+4r777kOtVvPdd98BsGTJElq3bo1araZVq1bMnTvXQs+SeQgBZGW0Opnpv51EliS+6PAoRUoVBn+Ak04Lsv66Vie8BLXFEAbbfm07pdpSq66VlJ/Ebxd+A+DFdi/WON7VyZV7wu4BRBhMIHBkZFlGV1ho9lF67RopU6aaLmsyZSql166ZPZe9vcPBwcFoNBpWrVpVpS1JSUnce++9qNVqtm7dSnx8PM899xwajQaAzz77jE8++YSPP/6Yo0ePMmTIEB588EHO3bax55///CeTJk3i1KlTDBkyhAULFvDOO+/wwQcfcOrUKT788EPeffddvvnmG6s/bgN27QZ/N7D/UiYpOcUA5Lh48nnHx3j5yGo8ygoJK8ygaU4S56XG7L+USc+mIhRWG9oGtCXQLZC0wjT2Ju815t1Yg0XHFqGRNfQI6UGHwA5m3TM4cjBxV+LYdHkTkzpOQrr9BVMgENgduaiIM506W2AiGV1eHhcGDjL7lpaH4pHc3Gq1zNq1a/Hw8DD+PGzYMH7++edazWGgR48e/Otf/+Kpp57ipZdeolu3bvTv35+xY8cSFBQEwH//+1+8vb1Zvnw5zs7OALRo0cI4x8cff8w///lPRo8eDcCsWbPYtm0bc+bM4b///a9x3Ouvv17BO/Xee+/xySefGM9FR0dz8uRJvv76a5599tk6PZ7aIjxAViYtr7jCzzvDOvDU/dPYE9oWgH5XD5kcJ6gZhaQwbjm3ZhgstSCV1edXA+Z5fwz0adwHlUJFYl4iZ7POWsk6gUBwN9GvXz8SEhKMx+eff27WfR9++CEeHh7GIzExEYAPPviA1NRUvvrqK2JiYvjqq69o1aoVx47pd7AmJCTQp08fo/gpT25uLsnJyfTu3bvC+d69e3Pq1KkK57p06WL8/saNG1y9epUJEyZUsOn99983htZsgfAAWZlAT7XJ89sad6Jv0hH6JiWwKHZ4leME1TMochDfn/qebVe3UaYtw1lZ+Z+0viw5voQyXRldgrrQJbhLzTfcxN3Znd5hvdl2dRubEzfT0q+lxW0TCAT1Q3J1peWheLPGyrJM8t/fIn/HDtO13ZRKPPr2Jezj/5i9dm1xd3enWbPK9cdq4qWXXuKJJ54w/hwaGmr83t/fn8cff5zHH3+cmTNn0rFjRz7++GO++eYbXM2w8XbvtizLlc65u7sbv9fp9PXRFixYQPfu3SuMU1ZVMsYKCA+QlekW7UeIt5rbgx/xQS3Jc3bFvziXfoWJdIv2s4t9DZ0OjTrgr/YnrzSP/an7LT7/jcIb/HL2FwBebG++98eAwUMVd1lUhRYIHBFJklC4uZl1KN3dCXn/vWrLmoS8N8Ps+WwZFvfz86NZs2bGw8nJtP9DpVLRtGlTCgr0VfbbtWvHzp07Te7c8vLyIjQ0lF27dlU4v2fPHlq3bl2lLUFBQYSFhXHx4sUKNjVr1ozo6Ogq77M0QgBZGaVCYuqIGIAKIkijcGJnWHsAXtKcQ6kQ+SF1QalQGnuDWaMo4tITSynVldKhUQe6B3ev+Ybb6BveFyeFExdyLnAx+6LF7RMIBLbFyd+fkOnTTJY1CZk+zW5lTfLz841hMYBLly6RkJBgDHWZYu3atTzzzDOsXbuWs2fPcubMGT7++GPWrVvHQw89BMBf//pXcnNzGT16NAcPHuTcuXMsW7aMM2fOAPDWW28xa9YsVqxYwZkzZ3j77bdJSEjgtddeq9beadOmMXPmTD777DPOnj3LsWPHWLJkCbNnz7bME2IGQgDZgKGxIcx7phPB3hXDXAkt9G+oXvt3oSsWOUB1pXwHdo1OY7F5M4oy+OnMT4De+1OXT2teKi96hvQERG8wgeBOwXPYsIoV/m9W9PcaNsxuNh08eJCOHTvSsWNHAN588006duzIlClTqrwnJiYGNzc3/va3v9GhQwd69OjBTz/9xMKFCxkzZgygD49t3bqV/Px8+vbtS+fOnVmwYIExJ2jSpEn87W9/429/+xtt27Zlw4YNrFmzhuY1dDeYOHEiCxcuZOnSpbRt25a+ffuydOlSm3qAJNne+/AckNzcXLy9vcnJycHLy8ti82p1MvsvZfLPlUdJzCxk1sNt6PT2c2hSUgibMwevoUMsttbdhEanod9P/cguyWbh4IV0D6m9p8YUn8Z/yuLji2nj34YfH/ixzu7qVedWMWXPFFr6tuSXB3+xiG0CgaBuFBcXc+nSJaKjo1Gr6557qcnI4MLQYejy8lB4edF0/TpR1NZGVPc7rM37t/AA2RClQqJnU3+GtwsBYN/lbLyHPwBAztrf7Glag8ZJ4cSAiAGA5bws2cXZ/Hj6RwBeav9SvWL1/cL7oZSUnMk6Q2Ju1e5ogUDQcHDy9ydkxnSUAQGEzJguxE8DRAggO9C7WQAAuy+k4zl8OAAF23egzcmxp1kNGkMe0JbELWh1JnZn1JJlp5ZRpCmilV8r+jauX30hH7UP3YK7ASIMJhDcSXgNG0aLXTvxGjrU3qYI6oAQQHagc6QvKicF13NLSPINw6VFC+SyMnI3brS3aQ2W7sHd8VR5kl6UTsKNhHrNlVuayw+nfgD0dX8ssVPDmonaAoFAIKg9QgDZAbWzki6RvgDsuZCO1wi9Fyj3t7X2NKtB46x0pl94P6D+IuP7U9+TX5ZPM59m9I/obwnz6B/RH4Wk4ETGCZLzky0yp0AgEAjqjhBAdsIYBjufjvcD+jygwgMHKEsRjVHrimE32OYrm9HJujrNkV+az3cn9Y36Xmz3IgrJMv8iAa4BdA7Sl9sXXiCBwP6I/T8NF0v97oQAshO9bvb92nshA0VwCG43y4Tn/v67Pc1q0PQM7Ym7szvXC69zLP1YneZYfmY5uaW5RHlFGQWVpTB0rxcCSCCwH4bt24WFhXa2RFBXSkv1za/rWzVatMKwE23DvPF0cSK3WMPJ5FwajxhB4cGD5Py2Fv+JE+1tXoPERelC38Z9WXdpHXGX42jfqH2t7i8sK+TbE98C8EK7F1AqLFuSfWDkQGbun8mRG0e4XnCdIPcgi84vEAhqRqlU4uPjQ1paGgBuNq7ILKgfOp2OGzdu4ObmVmU1a3MRAshOOCkVdG/ix+ZTaey+kM7zQwaT+v77lJw5Q/HZs6jLddsVmM+gyEGsu7SOzYmb+VuXv9Xqhe2nMz+RVZJFuGc4w6ItX9As0C2QDo06kHAjgc2Jm3m69dMWX0MgENRMcHAwgFEECRoWCoWCiIiIegtXIYDsSK+mAXoBdD6dl/o2xaNPH/K3biV37e+o3xQCqC70DuuNq5MrSflJnMw8SRv/NmbdV6QpYumJpQA83/Z5nBTW+dcYFDlIL4CuCAEkENgLSZIICQkhMDDQZI8rgWOjUqlQKOqfwSMEkB0xJEIfuJxJiUaL94jhNwXQWhq9/hqSBX7BdxuuTq7cE3YPcVfiiLscZ7YAWnl2JRnFGYS6hzK86XCr2TcochD/Ofgf4q/Hk16UToBrgNXWEggE1aNUKm3afVzgWIh3WDvSIsiDAA8Xist0HE7MxqNfPxTu7pQlJ1N0+LC9zWuwDI4cDOiTjc3ZLVCiLWHJ8SUATGg7AWeFs9VsC/EIIdY/FhmZrYlbrbaOQCAQCKpHCCA7IkmScTfYnvPpKNRqPAfpdx7l/CZaY9SVPo374KJ0ITEvkbNZZ2scv/rcatKK0ghyC2Jks5FWt29QlP53LHaDCQQCgf0QAsjOGAXQhQwAY1HEvPUbkG9u9RPUDndnd3qF9gJqFhll2jIWHl8IwHOxz6FSqqxu36AIvQA6kHqArOIsq68nEAgEgsoIAWRnDHlACVezKSjR4N6jB8pGAWhzcsjftdvO1jVcyhdFrI41F9aQWpBKgGsAjzR/xBamEe4VTiu/VmhlLduubrPJmgKBQCCoiBBAtiDnGiQnmDzCi8/SyacAjU5m/6VMJKUS7/vvByBXdIivM33D++KkcOJCzgUuZl80OaZMV8aCYwsAGN9mPGontc3sMwg0EQYTCAQC+2B3ATR37lyio6NRq9V07tyZnTt3Vjv+v//9L61bt8bV1ZWWLVvy7bffVhqzcuVKYmJicHFxISYmhlWrVlnL/JrRlMD8fjC/b5XHN5p/oqKM3efTAfAaPgKAvK3b0OYX2M/2BoyXyoueIT2BqkXGuovrSMpPwk/tx+MtH7eleUYB9GfKn+SW5tp0bYFAIBDYWQCtWLGC119/nXfeeYfDhw/Tp08fhg0bRmJiosnx8+bNY/LkyUybNo0TJ04wffp0/vKXv/BbuYThvXv3MmrUKMaMGcORI0cYM2YMTzzxBPv27bPVw6qIUgXeYVT9VCvQeIRSihO7b+YBqWPboIqKQi4uJn9L9SEcQdVU52XR6rRG78+zbZ7F1cnVprZFe0fTzKcZGp2G7Ve323RtgUAgENhZAM2ePZsJEyYwceJEWrduzZw5cwgPD2fevHkmxy9btowXX3yRUaNG0aRJE0aPHs2ECROYNWuWccycOXMYNGgQkydPplWrVkyePJkBAwYwZ84cGz2q25Ak6P9voKrmnDqkAf8GJE6l5JKRX4IkSXgN1ydD54gO8XWmX3g/lJKSM1lnSMytKKo3XN7AldwreLt4M6rlKLvYZxBom65sssv6AoFAcDdjNwFUWlpKfHw8gwcPrnB+8ODB7Nmzx+Q9JSUlqNUV8zRcXV3Zv3+/sZrn3r17K805ZMiQKuc0zJubm1vhsChNB0BoR7i9s7ikhNCO+MQOpWWQJwB/XswEwPvmbrCCPXvQpKdb1p67BB+1D92CuwEVvUA6Wcf8o/MBGBszFndnd7vYNzBS3xx1T9IeCspEqFMgEAhsid0EUHp6OlqtlqCgig0hg4KCSE1NNXnPkCFDWLhwIfHx8ciyzMGDB1m8eDFlZWWk3xQJqamptZoTYObMmXh7exuP8PDwej662zB4geTbvECyVn9ekujVTL8dfvcF/eNQRUaibtcOdDpy1623rD13EaZq7sRdieNizkU8VZ482epJe5lGc5/mRHlFUaorZce1HXazQyAQCO5G7J4EfXszM1mWq2xw9u677zJs2DB69OiBs7MzDz30EOPGjQOoUM68NnMCTJ48mZycHONx9erVOj6aajB4gYxGKvQ/Nx0AQO+m+u3we87f8vZ4G8Jga0UYrK70D++PQlJwIuMESflJFbw/T7d+Gk+Vp91skyRJ7AYTCAQCO2E3ARQQEIBSqazkmUlLS6vkwTHg6urK4sWLKSws5PLlyyQmJhIVFYWnpycBAXoBERwcXKs5AVxcXPDy8qpwWBxjLtBNZJ3R+wPQvYkfSoXE5YxCkrKLAPC6fxgolRQfPUrp5cuWt+kuoFRbSiu/VgB8f/J7vj35LWezzqJ2UtM1qCsp+Sl2tc8QBtt5bSeFZYU2XTslP4WTGSerPOz93AgEAoE1sVszVJVKRefOnYmLi+Phhx82no+Li+Ohhx6q9l5nZ2caN24MwPLlyxk+fLixM2zPnj2Ji4vjjTfeMI7ftGkTvXr1ssKjqCVNB4BHMOSngqu/0fsD4Kl2pl1jbw4nZrP7fDpPdAnHKSAA9549Kdi1i5y1v9Por3+xo/GQu349qR98SPC/38Fr6FC72mIOKfkpDF89nFKtvqL2slPLjNeKNcVM2DQBlVLF2pFrCfEIsYuNrf1aE+YRRlJ+EruTdxs9Qtbm9ufGFPZ+bgQCgcCa2DUE9uabb7Jw4UIWL17MqVOneOONN0hMTOSll14C9KGpsWPHGsefPXuW7777jnPnzrF//35Gjx7N8ePH+fDDD41jXnvtNTZt2sSsWbM4ffo0s2bNYvPmzbz++uu2fniVkSToNEb/vZuP0ftjwGQY7GYydO5vv5nV2NNaaDIySJkyFW16OilTpqLJyLCbLeaSVZJV7Rs86D1EWSX2a0chSdKt5q2XbRcGawjPjUAgEFgTuwqgUaNGMWfOHGbMmEGHDh3YsWMH69atIzIyEoCUlJQKNYG0Wi2ffPIJ7du3Z9CgQRQXF7Nnzx6ioqKMY3r16sXy5ctZsmQJ7dq1Y+nSpaxYsYLu3bvb+uGZpt1o/decJNBpK1wq3xfMIHY8BgxEUqspvXKF4uMnbGqqAVmWSZ02DV2hPkSjKyggddp0u9hyJ2IIg22/tp0SbYmdrREIBIK7A7uFwAy88sorvPLKKyavLV26tMLPrVu35vDhwzXO+dhjj/HYY49ZwjzL4xcNzm5QVggZF6BRC+OlTpG+uDgpSMsr4cKNfJoFeqL0cMezf39y160jd+1vuLaNtbnJeevXkxdXriCjVkteXBy569fjNWyYze2502gb0JZg92BSC1LZk7SHfhH97G2SQCAQ3PHYfRfYXYdCCYEx+u+vH6twSe2spEuULwC7z98KMRk6xOesW4esreg1sjaajAxSpk6rFK5DkkiZOq1BhMIcHUmSGBih9wKJ3WACgUBgG4QAsgfBN704qccrXep1Mw9od7k8II977kHp44P2RjoFf/5pExPhttDX7flHsixCYRbEkPz8x9U/KNOW2dcYgUAguAsQAsgeBN0UQNcrC6DezfQC6M+LGWh1etEhOTvjOXQIALk2bI1Rcu6cPvRVldfpZiis5Nw5m9l0p9IhsAONXBuRV5bHnynWF7lnM89afQ2BQCBwZIQAsgdGAVQ5qbltmDeeaidyizUcT8oxnvcecbNDfFwcuuJim5jp0rw5noMGgqKKPxOlEs9Bg3Bp3twm9tzJKCQF/SP6A9YPgx27cYwP9n9g1TUEAoHA0RECyB4EtdF/zU2CwswKl5QKiR5NKrbFAHDt2BHn0FB0BQXkb9tmEzMlScLvhRcqh7/0F1G4uxM8bapNbKkLvi6+qJSqaseolCp8XXxtZFH1GLbDb726lTKddcJgJzJO8GLcixRripGoujo6ONZzIxAIBJbG7rvA7krUXuATCdlX9GGw6HsrXO7V1J+4k9fZeyGDV+5rBoCkUOA1fDgZ8+eT89tam+y+KktLI/nNv5kWQLJMyPRpOPn7W92OuhLiEcLakWurrWXj6+LrMIX+OgV1wk/tR2ZxJgdTD9IztKdF5z+deZoXNr1AXlkenQI7MbXnVIq1t7yJ7//5PsfSjzG+zXiGRg91qOdGIBAILI0QQPYiuK1eAKVWFkCGPKADlzMp0WhxcdL3OfMeoRdA+Tt3os3ORunjYzXztNnZXJ0wkbKrV3Fq3BiXJtEU7N5jzAdS+vs3iC3wIR4hDeZN3EnhRL/wfqw8t5K4K3EWFUBns87y/KbnyS3NpV2jdswdOBd3Z/cKY+4Lv49j6cdIyk8ixj/GYmsLBAKBIyJCYPbCEAYzkQjdPNCDRp4uFJfpOHQl23jepXlzXFq1grIycjduspppuoICrr74EiXnzuHUqBGRSxYTOnMmCjc34xhtdjba7OyqJxHUCUMYbEviFrQ6y5Q8uJB9gec3PU92STax/rF8NfCrSuIHoHNQZwAOXj9o16rjAoFAYAuEALIX1ewEkySpXFXo9ArXyrfGsAa60lKuvTqJoiNHUHh7E75oIarwcJz8/QmZMR1lQABOoaH6HWBbtljFhruZriFd8VJ5kVmcyaG0Q/We71LOJSZsnEBmcSat/Vrz1aCv8FR5mhzbNqAtLkoXMoszuZR7qd5rCwQCgSMjBJC9MNQCSjsNWk2ly71N1AMC8HrgAZAkCg8epCw52aImyRoNyX9/i4I9e5Dc3IiY/zXqFrcqVXsNG0aLXTvxfVxfZTt3w0aLri8AZ4Uz/cL1laA3X9lcw+jqScxNZOLGiWQUZ9DStyXzB83H28W7yvEqpYp2jdoBcDD1YL3WFggEAkdHCCB74RMFKg/QlkBG5To6vZrpPUBHruWQV3xrR5BzcDBuXbsCkPP77xYzR5ZlUqZOJW/TJiRnZ8L/+yWu7dubHOs5RN8JvmDvXhEGswKDo/RhsM1XNqOTdXWa42reVZ7b+BxpRWk082nG/MHz8VH71Hhfl6AuAMRfj6/TugKBQNBQEALIXigUt/KATFSEbuzrRoSfG1qdzIHLFbfKexnDYJYpiijLMmmz/o+clb+CQkHo7E9w71l1Aq5Lk2h97R+NhrytttmSfzfRI6QHHs4epBWlcfTG0Vrfn5yfzISNE7heeJ1o72gWDF6An9rPrHtFHpBAILhbEALInhgToY+ZvNz7pheofF8wAK/Bg5GcnSk5e5biM/Wv6Jvx9ddk3mw8G/L++3gNGlTjPcbK1Bs31Ht9QUVUShV9w/sCtS+KmFqQynMbnyOlIIVIr0gWDV5EgGuA2fe3a9QOJ4UTaYVpXMu7Vqu1BQKBoCEhBJA9qaYiNJjuCwag9PbGva9+63zu2volQ2f+8AM35nymN2fy2/g88rBZ93kNvRkG27MXbW5uvWwQVMbQGyzuSpzZnpi0wjQmbJxAUn4SjT0as3DwQhq5NarVuq5OrrQNaAvovUACgUBwpyIEkD0J1r/RmAqBAcadYKdT80jPL6lwzXu4vjVGztrfkXV1yxPJ+W0t1997H4CAV17G79lnzb7XpWlTXJo3g7Iy8rZsrdP6gqrpHdobVydXUgpSOJFhWiCXJ70onQkbJ5CYl0iYRxiLhywm2D24TmuXD4MJBALBnYoQQPYkMAaQID8VCtIrXfb3cKFVsH7L8t4LFcNgHv3uQ+HhgSYlhaJDtd8unbdtG8lvvw2yjO/TTxPw6qvVjtfqZPZeyOB/CUnsvaBv1GpIhs7bIMJglkbtpObexnovX01hsIyiDCZunMjl3MsEuwezaMiiehV/FInQAoHgbkAIIHvi4gF+0frvU6vKA9KHwW6vB6RwccFzsH63UE4tk6EL9u8n6fU3QKvF68ERBL3zLySp6r5QG46ncM+srTy54E9eW57Akwv+5J5ZWznatBMA+Xv2iDCYFTAnDJZVnMXzcc9zIecCgW6BLB68mDCPsHqt2yGwA0pJSVJ+Ein5KfWaSyAQCBwVIYDsjTERuqo8IENBxIxK1wxFEfM2bEAuLTVruaLjJ7j28ivIJSV49OtH6AcfIFXV7R29+Hn5u0Ok5FTsQJ+aU8zE7ZmUNo7Uh8G2ijCYpekT1gcXpQtX865yNqtysntOSQ4vxL3AuaxzNHJtxKLBiwj3Cq/3uu7O7sZWGCIMJhAI7lSEALI3QTfzgExUhAboFu2HUiFxJaOQa1mFFa65deuGU6NGaHNyyN+1q8alSi5e5Orzz6MrKMCta1fCPp2N5Oxc5XitTmb6bycx5XswnFt/840yz4qtOe5W3JzduCfsHgA2Xan4/OaW5vJC3AuczjyNn9qPhUMWEuUdZbG1DXlAIgwmEAjuVEQzVHtjqAhdRSK0p9qZ9o29OZSYzZ7zGTzR9VY/LkmpxOuBB8hcupSc337Ds3//KpcpS0oi8bkJaLOyUMfG0njeXBRqdbWm7b+UWcnzUx4ZvQB6iPUU7NqFNi8PpafpNguC2pOSn0KMfwxbErew9uJaBkQMAKCorIj3973P+ezz+Lr4smjwIpp4N7Ho2l2CurD0xFKH9QCl5KeQVZJV5XXRyd56NPTnviHb35Btd0SEALI3hq3wN06DphScVJWG9G4WwKHEbHZfSOeJrhVDHF4jhpO5dCn5W7ehzc9H6eFR6X5NRgaJz01Ak5qKqkkTwhfMNznudtLyqhY/Bq54BlMSGoFLciL527bh/eCDNd4jqJmU/BSGrx5OqVYf2kzOT2bU2lGVxn14z4c0821m8fU7BnVEQuJK7hVuFN6o9XZ6a3L7c2MKlVLF2pFrxZuBhWnoz31Dtr8h2+6oiBCYvfGJABcv0JVBuumihoZ6QHsuZFRKhlXHxKBq0gS5pIS8uMq9o7S5uSROfJ7SK1dwDg0lYvEinHx9zTIt0LN6DxEAkoR8r97zJHqDWY6skqxqX+gM+LmaV+G5tnipvGjp1xJwvDCYOc9Nqba02k/KgrrR0J/7hmx/Q7bdURECyN5IUo2J0J0ifXBxUnAjr4Rzafm33S7hNfwBQN8hPnf9es7e04fcDRvQFRVx9eVXKDl1CqW/PxGLF+EcbH5tmG7RfoR4Vy2CJCDEW03L0SMB9GGw/PwqxwsaFobt8I4aBhMIBIL6IASQI2CsCG16K7yLk5KuUfpP+nvOV64X5D1cvxusYO9eUt6dgjY9nZR3p3D15Zcpio9H4elJxKKFqKKiamWWUiHxbK/q75k6IgbXli1QRUcjl5aSv030BrtTMAog0RleIBDcgQgB5AjUkAgNt7rD7zaxHV4VEYG6XTuQZXQFBQDo8vMp/HMfklpN+NdfoW7VqtZmaXUy647p68ConSv+qXipnZj3TCeGxoYgSdKt3mAiDHbH0ClIX+fpQs4FMoszaxgtEAgEDQshgByBGrbCA/S+mQf058UMNNrKrS9cmjbVf2PIEbr51W/sWNw6daqTWT/sT+TotRw81U5s+/t9/Ph8Dx5oqw+h3dM8gKGxtxLtjL3Bdu5Em19Qp/UEjoWv2pdmPvoE60PXa19tXCAQCBwZIYAcgcBWgAQFNyDvuskhsWHeeKmdyCvWcDy5YtVlTUYGeXGm2yVkLV+OJqOy16gmbuSV8H8bTgPw1pCWhHi70rOpP8/0iALg0JXsCgnZLi1aoIqM1IfB/vij1usJHBPRF0wgENypCAHkCKjcwf+mB6cKL5BSIdGjyc0wWLk8IFmWSZ02DV1Rkcn7dAUFpE6bXmuTZq4/RV6xhtgwL57uHmk83z7cG6VCIjW3mORyNYL0YbCbvcE2it5gdwpdgkUekEAguDMRAshRMCZCVxMGM9EXrOTcOf32d63W9E1aLXlxcZScO2e2KX9ezODXQ0lIErw/si1Kxa0+YW4qJ9qEegFw8HLFvBCvm3lA+Tt2GnORBHXD18UXlbJyTajyqJQqfF3MK2lQVwyJ0GezzpJTkmPVtczFrOdGYf3n5m7E18UXlcL+f5d1xdfFF4VU/dueo9rv6+KLs6Lqyv3guLY7KqIQoqMQHAsnV1ebCN37ZiL0wctZFJdpUTsrcWneHM9BA8nbus20CFIq8ezfH5fmzc0yo0yr493Vehue6hZBh3CfSmM6Rfhy9FoOh65k8VCHW403XVq1wjkygrIrieT98QfeDzxg1pqCyoR4hLB25Fq7V30NcA0gyiuKy7mXOZx2mPvC77PqeuZgeG4WHV/EijMr6NCoA5O7Tya7JJu/bvkrZboypveaLorBWYEQjxAebf4oP575kSivKD7q8xGSJJFSkMIb295ARubT+z512OfeWemMk+REqVzKO93foV2jdgB8evBT/kz9k54hPR32byfEI4TBkYP5/dLvxAbE8m6Pd/nfhf/xw6kf6BLUhbe6viUqQdcS4QFyFMxIhG7ayINATxdKNDoOJerfGCVJInjaNBRubvqaQuWRJBTu7gRPm2q2GYt3XeJcWj7+7ir+McT0zrEuUfpPGAevVHxzliQJryE3w2BiN1i9CfEIIcY/psrDVi90xjwgBwqDhXiEcOTGEQAeaf4IMf4x9ArtxRMtnwDgl3O/2NO8O5ackhx+u/gbAK93ep02AW2I8Y9hQMQA7m9yPwC/nvvVniZWyzcnvqFUV0q7gHaMajnK+L/0965/B+DPlD8p0ppOJ7A31wuuG3sCvtn5TWL8Y3ggWv8h81z2OVr5tRLip5YIAeQoGIohpp8FTYnJIZIk3eoOf/5WYrOTvz8h06fd2gFmQJYJmT4NJ39/s0xIzi5izmZ9qOztYa3wdjPtbu0cqRdAp1JyKSjRVLh2Kwy2Q4TB7hAMeUCOVBH6au5VTmeeRikp6Rfez3h+fJvxOCucib8ez4HUA3a08M7kh1M/kF+WT3Pf5vSL6Ffh2gttX0BCYkviFs5mma5qb08yizNZcWYFAC+2fxGp3AfGln4t6R/eHxmZhUcX2svEallyYgllujI6B3Wma3BXAFr7t8bVyZWckhzOZZmf5iDQIwSQo+DdGNTeoNPAjTNVDut1Mw9o94WKBRE9hw3Dc9BAUCr1J5RKPAcNwmvYMLNNmPHbSYrKtHSN8uXRTo2rHBfi7UqYjys6GRKuZle45tK6Nc4REcglJeTv2GH22gLHxZAHdCrzFAVljiFq4xL1ux67BnfFR+1jPB/kHsTDzR4G4OujX9vDtDuW/NJ8lp1aBsAL7V6olEvTxKcJg6MGAzD/6Hyb21cTy04uo0hTRIx/DH3C+lS6/kL7FwD4/dLvJOYm2tq8akkvSueXs3qv5ovtXjSed1Y40zGwI+BYH1AaCkIAOQqSZF49oJsC6Oi1HPKKy8rdXi4UBrUOfW07ncaGE6koFRLvjYxFoZCqHd/pphco3mQYTBRFvJMIdg8mzCMMrazlcNphe5sDQNxlvQAaFDmo0rUJbSfgJDmxL2UfCWkJNrbszmX5meXkleYR7R3NoIjKzzvohRHApsubuJh90ZbmVUtOSQ4/nv4R0Nso3Z4uALTxb0OfsD7oZB0LjzmWF2jp8aWUaEto36g9PUJ6VLgmSlXUHbsLoLlz5xIdHY1araZz587s3Lmz2vHff/897du3x83NjZCQEMaPH09GuTo3S5cuRZKkSkdxcc2dze2OGRWhw3xcifJ3Q6uT2Xex4i4sJ39/QmZMRxkQQMiM6WaHvorLtExdo+9D9lzvKFoFe9V4T5dI03lAAJ43BVD+9u3oCgvNskHg2DhSW4zk/GSOZxxHQqJ/RP9K10M9Qnmw2YMAfHX0K1ubd0dSWFbINye+AfQCQqlQmhzXwrcFAyIGICMz/5jjeIG+O/UdBWUFtPBtUSFkejsvttd7V3678BtJ+Um2Mq9aMosz+ensT4De+3O7eDP8b8Zfj6/ULFtQPXYVQCtWrOD111/nnXfe4fDhw/Tp04dhw4aRmGja/bhr1y7Gjh3LhAkTOHHiBD///DMHDhxg4sSJFcZ5eXmRkpJS4VCrzehsbm9q6AlmoFezW93hb8dr2DBa7NpprMxsDnP/uEBiZiHBXmpeG9jCrHsMeUCHr2Sh093Wob5NDM6NGyMXF4sw2B2C4VOmI7jZN1/ZDOhtCnANMDlmYuxElJKS3Um7OZ5e9QcKgXn8dOYnskuyifCMYGhU9a8tBi/Q+kvruZJ7xRbmVUteaR7fn/weMB26K0/7Ru3pGdITjaxh0bFFtjKxWr498S1FmiLa+LfhnrB7Kl2PDYjFRelCZnEml3Iv2cHChotdBdDs2bOZMGECEydOpHXr1syZM4fw8HDmzZtncvyff/5JVFQUkyZNIjo6mnvuuYcXX3yRgwcrfiqVJIng4OAKR3WUlJSQm5tb4bALhkTo1OOVE5rLYUyEvlC5MWptuZRewFd/XABgyogYPFzMq4zQKtgTN5WSvBINZ9PyKlyTJMmYDC3CYHcGhkTo4xnHKdLYd5dM3BV9+Gtg5MAqx4R7hfNAE/0Oma+PiFyg+lCkKWLJiSUAPN/ueZwU1b9GxPjH0LdxX3SyjgVHF9jCxGr54dQP5JXl0dS7qcmQ6e281P4lAFadX0VqQaq1zauW7OJsY+jOlPcH9LV/DNv5HcFD25CwmwAqLS0lPj6ewYMHVzg/ePBg9uzZY/KeXr16ce3aNdatW4csy1y/fp1ffvmFB26rN5Ofn09kZCSNGzdm+PDhHD5cfd7CzJkz8fb2Nh7h4eH1e3B1JbA1SAooyoS8qv/xet6sCH06NY/0fNM7xsxBlmWm/O84pVod97ZoxLDY6oVieZyUCmONoNvzgAA8b26Hz9++vcoq1YKGQ2OPxgS5BaHRaYzbz+3B9YLrJNxIAGBgRNUCCGBi24koJAV/XPuD05mnbWDdncnKsyvJLM4kzCPMKCprwpCou/biWq7lXbOmedVSUFZgTNx+vt3zNRZBBH0T4K7BXdHoNCw+vtjaJlbLd6e+o1BTSEvfltXW4DKGqEUeUK2wmwBKT09Hq9USFBRU4XxQUBCpqabf/Hv16sX333/PqFGjUKlUBAcH4+PjwxdffGEc06pVK5YuXcqaNWv48ccfUavV9O7dm3PVVEKePHkyOTk5xuPq1auWeZC1xdkV/G8WLKwmEdrfw4XWIfo8HVNhMHNZdyyVnefSUTkpmPFgG5OfLqqjSxWJ0ADq2DY4h4UhFxWRv6P6vC6B4yNJkkNsh9+SuAWADo06EOQeVO3YaO9ohkTpPZGOuCupIVCiLTGKgIltJ9ZYidhA20Zt6R3aG62stWtC8fLTy8kpySHKK6rG0F15Xmqn9wKtPLuSG4U3rGVeteSW5vL9KX3o7vZt+7djzANKFXlAtcHuSdC3/1JlWa7yF33y5EkmTZrElClTiI+PZ8OGDVy6dImXXnrJOKZHjx4888wztG/fnj59+vDTTz/RokWLCiLpdlxcXPDy8qpw2A1jInT1eUC9jfWA6hYGyy/RMGOtPvH55b5NiQpwr/UcVe0EA0NvMP2bj+gNdmfgCAURzQl/lcdQmybuSpyok1IHVp1bxY2iGwS7B/NQ04dqda8hofh/F/5HSn6KNcyrlsKyQr49+S2gF29VJW6bomtwVzoGdqRUV2oM/9kaQ82lZj7NGBAxoNqx7Rq1w0nhRFpRml09bg0NuwmggIAAlEplJW9PWlpaJa+QgZkzZ9K7d2/eeust2rVrx5AhQ5g7dy6LFy8mJcX0P5hCoaBr167VeoAcCjN6gsGt7fC31wMylzlxZ7meW0Kkvxsv39e0TnN0jPBFkuBKRiE38iqH4gyJ2Hl/bEfXEHbhCarF8Cnz6I2jlGjrHnqtK+lF6RxKOwSY3v5uima+zYxiyRHyURoSZdoyFh3XJwJPiJ2As9I874+BjoEd6R7c3W6hpF/O/kJmcSaNPRobq1SbiyRJRi/Qz2d+JqOo7p72ulBQVsCyk1XXXLodtZOatgH6MioiDGY+dhNAKpWKzp07ExcXV+F8XFwcvXr1MnlPYWEhCkVFk5U3C/9V5faTZZmEhARCQhpIifCgmrfCA3SN9sNJIXE1s4irmbXban4qJZcley4DMO3BNqidzf9kVB5vV2daBHoCVYXBYnEODUUuLBS7we4Aoryi8Ff7U6or5diN6j2U1mBr4lZ0so5Y/1hCPULNvs+Qj7Lh8gYu5jhObRpH538X/kdqQSqNXBvxcPOH6zSHwQv067lfSStMs6R51VKsKTZ6bmoTuitPz9CetA1oS7G2mG9OfmNpE6vlx9M/kluaS5RXFIMjB9d8AyIPqC7YNQT25ptvsnDhQhYvXsypU6d44403SExMNIa0Jk+ezNixY43jR4wYwa+//sq8efO4ePEiu3fvZtKkSXTr1o3QUP0L4vTp09m4cSMXL14kISGBCRMmkJCQUCFM5tAYQmAZ56Csaq+Jh4sT7W8mIddmN5hOJ/Pv1cfR6mSGxQbTr2Vgfayl882+YIbeZOWRJMlYE0j0Bmv4SJJk1+3whu3v5oa/DLT0a0m/8H4O3ebA0SjTlRlzd8bHjsdF6VKneboEdaFTYCd9KOm47UJJv577lfSidELcQ3iw6YN1mkOSJKN4Xn56OVnFVTcmtiSFZYV8e0Ifuquu5tLtlK8HJDAPuwqgUaNGMWfOHGbMmEGHDh3YsWMH69atIzIyEoCUlJQKNYHGjRvH7Nmz+fLLL4mNjeXxxx+nZcuW/PrrreZ72dnZvPDCC7Ru3ZrBgweTlJTEjh076Natm80fX53wDAFXP5B1cONUtUMNeUC7z5vvnv3l0DXir2ThplLy7vCYepkK0DniZkHEy5kmrxu2w+f98YcIg90BGBKhbf0pM7s4m/2p+wHzw1/lMXgiHLHNgSPy+8XfScpPwk/tx2MtHqvzPJIkGZ/7n8/+THpR/Ut31ESptrReobvy3Nv4Xlr7taZIU2QMSVmbn8/+TFZJFuGe4QyLNr+VUfvA9iglJUn5SXbJuWqI2D0J+pVXXuHy5cuUlJQQHx/Pvffea7y2dOlS/vjjjwrjX331VU6cOEFhYSHJycl89913hIWFGa9/+umnXLlyhZKSEtLS0ti4cSM9e/a01cOpP5JkVkVoqFgQ0ZzM/6yCUmau04uq1wc2J9THtX62cqsg4vGkXIrLtJWuq9u1wyk0RB8Gq6HKt8DxMXzKPHLjCGW6shpGW45tV7ehlbW08mtFhFdEre935DYHjoZGpzHmS41rMw5Xp/q9TvQM6Um7gHaUaEuMng1rsvr8atIK0wh0DWRk85H1mqu8F+jH0z+SU5JjAQurplhTbPSUTWw7scaaS+Vxd3Ynxl//oVaEwczD7gJIYAIzeoIBdIzwQe2sID2/hLPX82uc9v82niGrsIwWQR6M7x1tCUuJ9HcjwENFqVbH8aTKLw6SJOE12LAbbJNF1hTYj6Y+TfFx8aFIU8SJ9BM2W9e4+6uG2j/V4YhtDhyRDZc3kJiXiI+LD6Najqr3fOW9QMvPWDeUVKYrM1Zwfq7tc3UO3ZWnX0Q/mvs2J78snx9O/VDv+apj5bmVZBRnEOoeyogmI2p9v8gDqh1CADki5StCV4OLk5KuUX4A7K5hO/yhxCyWH9C7/t8f2RZnpWV+9ZIk0Smi6u3wcCsMlr9tG7oS2+8eElgOhaSgU2AnwHa5BrmluexN2QvAoKjah78MOGKbA0dDq9MavT9jY8bi5uxmkXn7hPUhxj/G6qGktRfWklyQjL/an0ebP2qRORWSwtjeY9mpZeSX1vxhsy6UaEtYfEy/W25C27qF7hyhVldDQgggRyS43Fb4GkJbvavpC2ZAo9Xx7urjyDI82qkx3aL9LGYqQJeoqhujAqjbt8cpJARdQQEFu3ZZdG2B7bF1HtD2q9vR6DQ09W5KE+8m9ZrL4IlwhDYHjkhcYhwXcy7iqfLkyVZPWmze8qGkH07/YJVQkkanMRa8HB87HrWT5fo/DooYRBPvJuSV5hlbU1ia1edWk1aURqBbICObjazTHB0COyAhcSX3it0KODYkhAByRBq1AoUTFGdDbvWuekNfsH0XM9BodSbHfPfnFU4k5+KldmLy/a0sba0xD+jQlSyTuUj6MJj+k7voDdbwMbjZD6cdRqPTWH09Q/irPt4fA52DOtMlqItDtDlwNHSyziggxrQeg4fKw6Lz9wvvRwvfFhSUFRgrHFuS9ZfWcy3/Gr4uvjze4nGLzq1UKHm+3fMAfHvyWwrLald6pCbK11x6LvY5VEpVnebxUnnRyk//Gi/CYDUjBJAj4uQCATe7stcQBmsT6o2X2om8Eg3HTOTgpOUW88mmswD8Y2grAjzqHxO/ndgwb1RKBRkFpVzOMP3CYOwNtnWrCIM1cFr4tsDT2ZOCsgLOZJ2x6loFZQXsTtoN1G33lykMzS7t2ebAEdmWuI1zWefwcPbgqdZPWXz+8l6g7059R15pXg13mI9WpzWKt7FtLBe6K8/QqKFEekWSXZLNT2d+sujcv138jZSCFAJcA+odurNnqYqGhhBAjoqxInT1BeeUComexu7wlcNgH6w7RV6JhvaNvXmyW+13z5iDi5OSto29garzgFw7tMcpKEgfBtu92yp2CGyDUqGkY1BHwPptMXZe20mprpRIr0ia+zS3yJzdgrvRoVEHu7Y5cDRkWebro18D8GSrJ/F28bbKOgMjB9LUu6nFQ0mbrmzicu5lvFReFg3dlcdJ4cTEthMBWHJiCUUayzR5vn3XXX1Dd8ZEaNEZvkaEAHJUDInQ12veaWNsi3FbIvSe8+n8LyEZSdInPisVtWt2WhtuNUY1XQ9IUijwHKKvaJq7QfQGa+jYarfJpiv6nYODIgfVullvVUiSZPQC2aPNgSOy49oOTmWewtXJlTExY6y2jkJSVAglFZQV1HvO8qG7Z2Kewd259n0NzeWBJg8Q5hFGZnEmK8+utMic6y6t41r+NfzUfhYJ3XUK0m9SuJBzgcxi06/HAj1CADkqZtYCAujVVC+ADl7JMtbiKdXo+Pf/9PeO6RFp9NBYi+oaoxow9AbL37oNXWmpVe0RWBeDADp0/RA62XTuWX0p0hSxK0mfNG+p8JeBXqG9iPWPtUubA0ejvPdndKvR+Kp9rbre0KihRHlFkVOSw4ozK+o935bELZzPPo+HswdPt37aAhZWjbPC2egFWnx8cb174llj152v2pdmPs0AEQarCSGAHBVDLaDMC1BafcJd00buBHm5UKrRGQXIgp0XuXijgAAPFX8b3NLa1hoToc9ezyen0HSBPNcOHfRhsPx8EQZr4LTyb4Wrkyu5pblW67K+O2k3RZoiwjzCaO3X2qJzl/cC2bLNgSOyJ3kPx9KPoVaqeTbmWauvp1QojSLimxPf1CuUJMsyXx/Ri7enWj+Fl8rLIjZWx0NNHyLYPZgbRTdYdW5VvebaeHkjl3Mv4+3izehWoy1koWiLYS5CADkqnkHg3kjfEiOt+pYYkiTR+6YX6OeDV1my+xJzNusTn995oDXernUvBW8uAR4uRPnrP70cumr6zURSKPAcrA+Did5gDRtnhTMdA2/mAVkpDGaN8Fd57NHmwNGQZZmvjnwFwOMtH8ff1d8m697f5H4aezQmsziTn8/8XOd5/rj6B2eyzuDm5MaY1tYL3ZXHWenMhNgJACw6vogybd0qot++686SobvOwfpEaJEHVD11EkAajYbNmzfz9ddfk5enz+RPTk4mP986BaLuWsxMhAbwUOtLpq9OSGb6bycp08qolBJqp7p1eq8LnSP19YXiL1cXBrtZFXrrVhEGa+BY81NmibaE7Ve3A7VvfmoukiQZC9wZum/fbexP3U/CjQRUChXj24y32brlQ0lLTyytUyjp9tCdj9rHkiZWy8PNH6aRayNSC1L534X/1WmOzVc2cyHnAp7OnhbfdWf43zybddbq7TsaMrUWQFeuXKFt27Y89NBD/OUvf+HGDf020v/7v//j73//u8UNvKsxMxF6w/EUvt17pdL5Uq3MK98fYsNx2zTG62xGHpBrx444NWqELi+Pgj17bGKXwDqU325rTi+62rA3eS+FmkKC3IJoG9DWonOXp39Ef5r5NCO/LN8qtWkcHYOAeLTFozRya2TTtR9s+iAh7iHcKLrBr+d+rfmG29idvJsTGSdwdXJlbMxYK1hYNS5KF56LfQ6AhccW1rovnk7WGZ/7p2OexlPlaVH7AlwDiPKKQkbmcNphi859J1FrAfTaa6/RpUsXsrKycHW91STv4YcfZsuWLRY17q4n+OYLfzWJ0FqdzPTfTlY7zfTfTqLVWfYNyhSGitAJV7Mpq6IoowiD3TnEBsTionQhsziTSzmXLDq3sfhh5CAUkvUi9QpJYaxNs+yk9docOCLx1+M5kHoAJ4WT8c3cllQIJR1bRKnWfI9whdBdC9uF7srzaItH8VP7kZSfxO8Xf6/VvX9c/YOzWWdxc3LjmdbPWMU+wwcUEQarmlq/suzatYt///vfqFQVK1VGRkaSlCQaDFoUYwjsRJUtMfZfyiQlp7jKKWQgJaeY/Zesvx2yWSMPvNROFJVpOZ1SdZGz8mEwWYTBGiwqpYr2jdoDls0DKtOWse3qNsB64a/yDIocRLR3NHmleSw/s9zq6zkKhuThh5s9TLB7sF1seLj5wwS6BXK98HqtQkn7Uvdx5MYRXJQujGszznoGVoOrk6sxbLjw2EKzq6KXD9091fopq9VcsnXLmoZIrQWQTqdDq9VWOn/t2jU8PS3rxrvrCWgBCmcoyYHsRJND0vKqFj91GVcfFArJuB3+YBX1gABcO3VC2SgAXW4uBX/+aXW7BNbDGkXX9qXuI680jwDXADo06mCxeatCqVDyfFt9bZpvTnxj8TYHjsiRG0fYm7IXJ8mJCW0n2M0OlVJl9D4tOrbI7FCSwfvzaHPbh+7K80TLJ/Bx8eFK7hU2XDavvtnOpJ2czDhp9ZpLhv/NU5mn7irPZm2otQAaNGgQc+bMMf4sSRL5+flMnTqV+++/35K2CZxU+r5goG+MaoJAT/Oqhpo7rr50rqEzPICkVOI1yFAUUYTBGjLWyAMyhL8GRAxAqbBNEv+w6GFEeEZYpc2BI2Lw/gxvOpwwjzC72vJo80fxV/uTlJ/E2gtraxx/IPUA8dfjcVY4Mz7WdonbpnBzdjPmHy04ugCtrrJzoDzlt+0/0eIJ/NSWbUxdnmD3YBp7NEYn60i4kWC1dRoytRZAn376Kdu3bycmJobi4mKeeuopoqKiSEpKYtasWdaw8e6mhkTobtF+hHirqWqTsASEeKvr3gE+5xokJ1R95FQMe3aOqlkAAXgawmBbtiCX1W0bqcD+tGvUDmeFM2lFaVzNu1rv+TQ6DVsTtwKWL35YHdZqc+CInMg4wc6knfqqzDc9X/ZE7aQ2CpkFxxbUGEoyhI/sGborz5OtnsRT5cnFnIvEJcZVO3Zvyl6Oph/Vh+5ix1ndNpEHVD21FkChoaEkJCTw97//nRdffJGOHTvy0UcfcfjwYQIDA61h492NsSK06a3wSoXE1BExAJVEkOHnqSNi6tYGQ1MC8/vB/L5VHwvu04+7SYdwH5QKiZScYpKzq34TcevcGWVAALqcHBEGa8CondTGXVqWyDU4eP0g2SXZ+Lr4Gl+8bYXBG2LJNgeOiMEDcX/0/UR4Wac/YG15vMXj+Lr4cjXvKusvra9yXEJaAvtS9tk9dFceD5WHsQbR/KPzq6yMXt7781iLxwhwDbC6bSIPqHrqtL3C1dWV5557ji+//JK5c+cyceLECjvCBBbEmAhd9U6wobEhzHumE8HeFcNcwd5q5j3TiaGxIXVbW6kC7zCq/jNRgFeYftxN3FROxIToq7EerCEM5jlIn+Bqq95guevXc/aePqIXmYWxZPfpuMv6T9D9I/rjpHCq93y1wVnhbHxTXXJ8Sb3bHDgiZzLPsO3qNiQkY08uR8DN2Y2xbfShpPlH51cZSvrqqD7358FmDxLqEWoz+2riqdZP4eHswbmsc2xL3GZyzMHrBzmUdkgfurNRzSVDHtCJ9BN3tFezrtT6Febbb7+t9vrYsbatx3DHY9gKn3kJSvLBxcPksKGxIQyKCWb/pUzS8ooJ9NSHverVAFWSoP+/4btHqxig01+/rUpv50hfjiXlcOhKFg+2r/pFymvIULJ/XE7+5i3I06YhOVuvYrUmI4OUKVPR5eWRMmUqbl274uRv+62zdyJdgrqw4NiCervZtTotWxL1pTRsGf4qz0NNH+LrI19zvfA6q86tsmh7AkfAED4aEjWEJt5N7GxNRZ5s9SRLji/hcu5l4q7EMTR6aIXrx9OPsztpN0pJycTYiXay0jTeLt482epJFhxbwNdHv6Z/RP9K1csN3p9Hmj9CkHuQTewK8wgjyC2I64XXOXLjCD1Cethk3YZCneoAlT9eeeUVxo0bxwsvvMDrr79uBRPvctwDwCMYkCGt+no/SoVEz6b+PNQhjJ5N/S3T/b3pAAjtCNJtyaiSUn++6YBKt3Q2YycYgFvXLij9/dHm5FCwb3/9ba0CWZZJnTYNXaF+d4+uoIDUadOttt7dRofADiglJckFySTnJ9d5nsNph8kozsBL5UW3kG4WtNB8VEqV0QtUnzYHjsj5rPNsvrIZwFgB25Fwd3Y37or6+ujXlUJJBvH2QJMHCPcKt7l9NTE2ZiyuTq6cyjzFzqSdFa4dTjvMvtR9OCmcjLWPbIEkSbfCYCIPqBK1FkBZWVkVjvz8fM6cOcM999zDjz/+aA0bBcZE6Jo7w1scgxdIvs0lLWtNen/glgA6lZJHQUnVCY3lw2B5G60Xlspbv568uM1gKN+g1ZIXF0fu+qpzDQTm4+bsRht//d9ofXINNifq35zvC78PZ4X1+9dVxSPNH6l3mwNHZP6x+cjIDIwYSHPf5vY2xySGUNL57PPGZHiA05mn+ePqH0hIxmR1R8NH7WP0GH515KsKuyIN3p+Hmj5EiEcdUxLqiLFUhcgDqoRFguzNmzfno48+4plnnuH06dOWmFJQnuBYuLCl2orQVqXpAHAPhIK0W+eq8P4AhPq4EuqtJjmnmCPXsunVtOpkP6+hQ8levoK8uM0ET5li8TCYJiODlKnT9EKt/DZtSSJl6jTcunUToTAL0DmoM0fTjxJ/PZ4Hmz5Y6/t1ss64/X1w5GBLm1crXJQuPNr8Ub46+hVzE+bSwrdFpXwkXxffer2RpeSnkFVSdY6cpedPzk9mwyX9h4xBkYNIyU+x+RuxOXipvHiw6YP8cPoHPjv0GSHuIUiSxOyDswHoGdoTtdI2JT3qwrMxz/LDyR84ln6MFadX0C6wHeezzrM7eTcKFPRt3Nfmz71BAB27cYwSbQkuShebrX071v67ry0WyzJUKpUkJ9fd/S2ohqCbeUD28AABlBVBaUHFc/1Me38MdIr0JfloCvGXs6oVQG5duqD080ObmcmN//6X7F9WEvzvd/AaOrTKe8xBm51N4ZEjpH00C12eiarUsmwMhTX+4vN6rSXQ7zZZcmJJnd3sR28cJa0wDXdnd3qG9rSwdbUjJT+FxccXA3Cj6AZPr3u60hiVUsXakWvr9GKdkp/C8NXDq239YM35/7nzn/Wa35qk5Kfwy9lfALice5nRv1fMwdqTvIfhq4c7pO0ApdpSymR92PSD/R9UuKZDx6Rtk2z+3Ed6ReKv9iejOINjN44ZQ2K2xtp/93Wh1gJozZo1FX6WZZmUlBS+/PJLevfubTHDBOUILtcSQ6cDhfV6I5nk+EooKwClMxhyIvybVXtLl0hf1h5NIT6x+npAkpMTnoMGkb1iBRmLFkNZWa2TlGWNhpJz5yg6coSihCMUJSRQevlyzTfeDIWVnDuHS3PHDAk0FDoGdkQhKUjMSyStMI1At9qVxDDkpvRt3BeVUlXDaOuSVZJFqa76Fi2l2lKySrLq9EKdVZJVY98rR57fmlj7ubc2WSVZVW6DN2Br+w15QBsvb+Tg9YN2E0CO+HdZawE0cuTICj9LkkSjRo3o378/n3zyiaXsEpTHv7l+q3lpPmRfBj8b7t6QZdg/X/99+6fgyHLQlkDyIfCLqvK2zpH6wouHrmSh08koqknI9hwymOwVK+BmQcSaPDOa9PRbYufIEYqOH0curNy+wDkyErm0FE1qapW91Nzvu0+IHwvgqfKkpW9LTmWeIv56PMOih5l9ryzLDhP+EgjuRLoE6QWQJUpV3EnUWgDpdNWrW4EVUDrpW2KkHtV7gWwpgK4d0K/rpIaB00DhBAcX6QVQ7CNV3tYqxBNXZyW5xRrO38inRVDVfeK0GbftFiuXpOw5YADFZ85QdDhBL3aOHKHs2rVKcyg8PHBt1w7XDu1x7dABddu2OPn6osnI4MLQYejy802KoOLjxyn4cx/uPbqb/ZQITNM5qDOnMk9xMPVgrQTQyYyTJBck4+rkSu+whuNF/vrI1/iqfWt9X1Zx9V5RW80vuHsw1Oo6cuMIZboyu24ycCRsW2lMUHeC2+qFSOpxaD3CdusavD+xj4GbH4R11gugpEPV3uasVNAh3Ie9FzM4eDmrSgGkycggdcYMk9eS/v6WPtx3e6sMScKlWVNcO3TAtX17XNu3R9W0KZKJ0KCTvz8h06eR9ObfKl1TBgaiTUsjcfx4/F96kUZ/+QuSk/iXqCtdgrvw3anvar3bxOD96RPWB7WT4ya43s7Wq1trHuTA8wvuHpr6NMXHxYfskmxOpJ+gQ2AHe5vkEJj1av/mm2+aPeHs2bPrbIygGsyoCG1x8tPgxGr9991ubj0Nu9meIPkwaDV671QVdI70Ze/FDOKvZPFU98ol92+vz1MJrRa0WpTe3qg76IWO203vjtKzao/S7XgOG4bn+vXkbd2mn1OpxLN/f0JnfUTqhx+S88tKMuZ9ReG+/YR9/B+cQx2nwmxDonOg/m/jYs5FMooy8HetOYerfPhrUJR9ih/WldEtR9c61wkgrTCN5WeW231+wd2DQlLQOagzWxK3cPD6QSGAbmKWADp8+LBZk91e+VJgQWroCWYV4r8BXRk07qrf9g4Q0BxUnlCaB+lnbtUoMsGtxqimCyKWnDunr89TAxHLvkXdokXt7b+JJEkET5tGwZ/70OXloXB3J3jaVBRuboS+/z4evXqRMmUqRYcOcXHkw4S8/x5eg0UuSm3xUfvQzKcZ57PPcyjtkFnVnM9mnSUxLxEXpQv3ht1rAystx8PNHybGP6bW953MOGmWQLH2/IK7iy5BXdiSuIX46/EOW0vJ1pglgLZtM93bRGBDDB6g7CtQnAtqL+uup9XAQf1WYLqW6xmkUEJoB7i8E5LiqxVAncL1AuhyRiHp+SUEeFSsP+HSvDmegwbe8szczk1PTX3EjwEnf39CZkwn9YMPCf73OxV2mHndfz/qtm1J+tvfKT56lKRJr1Hw5GiC/vlPFOqGE5JxBLoEdeF89nkOph40SwAZvD+9Q3vj5uxmbfMEgrsWQx7Q4bTDaHQam/fac0RsvJ9aUGfc/MDzZmimhpYYFuHM75CXDG4B0GZkxWuGMFhS9TsKvN2caRGk710Wb6IxqsEzo3Bzq1xTSJKMnhpL4TVsGC127TRZY0gVHk7U99/hP1Ffpj77x+VcfmIUJefPW2z9u4Hadp92xPCXr4tvjVvxVUoVvi61T1C+E+a3Jg3ZdnBs+1v4tsDT2ZOCsgLOZJ6x+fq+Lr4ob2+pdBu2fm7qJAEPHDjAzz//TGJiIqWlFff1//rrrxYxTGCC4Fi9KEk9BhFWbmq3f4H+a+dnwem2yqFhnfRfaxBAoM8DOns9n0NXshjSJrjS9SqTlGWZkOnTbFqlWXJ2JvDvf8etR0+S336bkrNnufTY4wT9azI+jz8uQrxmYPiUeS7rHDklOXi7eFc59kL2BS7mXMRJ4UTfxn1tZWKNhHiEsHbkWqtVrG3o81uThmw7OLb9SoWSTkGd2H5tOwevH6RNQNXee2vg7+qPj4sPGcUZTIidwOCoymkGDl8Jevny5YwdO5bBgwcTFxfH4MGDOXfuHKmpqTz88MPWsFFgICgWzm2yfiJ02il9iEtSQJfnKl83eICun4TSQlBVHbroHOnHj/uvctCEB8hAVUnKXsPM30ptSTzu6U2T1atI/ufbFOzeTeqUqRTs2UvIjOkovawcemzgBLgGEOUVxeXcyxy6foh+Ef2qHGvw/vQK7YWnyvykdlsQ4hFi1Rfihj6/NWnItoNj2985qLNRAD3b5lmbrr36/GoyijMIdAvklQ6v2L3gKdQhBPbhhx/y6aefsnbtWlQqFZ999hmnTp3iiSeeICKi8k6fmpg7dy7R0dGo1Wo6d+7Mzp07qx3//fff0759e9zc3AgJCWH8+PFkZGRUGLNy5UpiYmJwcXEhJiaGVatW1douh8SYCG1lAXRgof5ry/vBu3Hl615h4BGkb4haQ1K2oTHqsWs5lGhM5PlwWygMLB76qgtOAQGEL5hP4Ft/Bycn8jZs4NLDj1CUkGBXuxoC5obBjOEvM3KFBAJB/TH0BTt0/VCNFastSZmujEXHFgHwXOxzDiF+oA4C6MKFCzzwwAMAuLi4UFBQgCRJvPHGG8yfP79Wc61YsYLXX3+dd955h8OHD9OnTx+GDRtGYmKiyfG7du1i7NixTJgwgRMnTvDzzz9z4MABJk68ldG+d+9eRo0axZgxYzhy5AhjxozhiSeeYN++fbV9qI6HoSdY2kl9SwxrUJyrr/YM0O0F02Mkyew8oCh/N/zdVZRqdRxPyqlynCFJWRkQQMiM6Q7RoFRSKPCfMIGoH77HOTycsqQkLj/9DOnzFyDf9vznrl/P2Xv6kLvBel3tGwrmdJ++knuFs1lncZKc6BdetZdIIBBYjtb+rXFzciO3NJdzWedstu7aC2tJLkjGX+3Po80ftdm6NVFrAeTn50fezeaSYWFhHD+u90ZkZ2dTWFU9lyqYPXs2EyZMYOLEibRu3Zo5c+YQHh7OvHnzTI7/888/iYqKYtKkSURHR3PPPffw4osvcvDgrRfaOXPmMGjQICZPnkyrVq2YPHkyAwYMYM6cOVXaUVJSQm5uboXDIfFroq/IXFYIWZess8aR5fqWGwEtIbqabclm5gFJkkSnSMN2+Oor1FaXpGxPXNu1I/rXlXjdfz9otdyYPZurEyeiuXEDuNlxfspUtOnppEyZiuY2j+TdhiEP6HTmafJL802OMXh/uoV0qzZPSCAQWA4nhRMdA/UlTWpbsLSuaHQaFhzT55SOjx3vUMVOzRZACTdd/3369CEuTv/i9cQTT/Daa6/x/PPP8+STTzJgwACzFy4tLSU+Pp7Bt9VbGTx4MHv27DF5T69evbh27Rrr1q1DlmWuX7/OL7/8YvRIgd4DdPucQ4YMqXJOgJkzZ+Lt7W08wsPDzX4cNkXpBIGt9d9box6QLMOBm8nP3Z6vttu7uR4g0DdGBTh4ueGW6Fd6ehL6yceEfPA+kqsrBXv2cvGhkeTt2FGhmKOhj9ndTLB7MI09GqOTdRxOM11DzND8VIS/BALbYviAYqu+YOsvredq3lV8XXx5vMXjNlnTXMwWQJ06daJz5860bt2aJ598EoDJkyfz97//nevXr/PII4+waNEisxdOT09Hq9USFBRU4XxQUBCpqakm7+nVqxfff/89o0aNQqVSERwcjI+PD1988YVxTGpqaq3mNDyOnJwc43H16lWzH4fNsWZF6EvbIf0sqDyg3ajqxxoKI2ZdgkLThQ4NGPKADiVmIVfRlLQhIEkSPo8+SvTKX3Bp2RJtZibXXnhRX8zRUMeoXB+zu5nq8oCS8pM4kXEChaSgf0R/W5smENzVGP4346/HW/31WKvTMv+oPjVmbJuxDlfry2wBtHv3bjp16sTHH39M06ZNeeaZZ9i+fTv/+Mc/WLNmDbNnz8bXt/b792/fWizLcpXbjU+ePMmkSZOYMmUK8fHxbNiwgUuXLvHSSy/VeU7Q5zJ5eXlVOByW4Jt5QNZIhDZsfW//ZM2FFl19wa+p/vsa+oLFhnmjUipIzy/lSkbtwqSOiEuTJkT9tALvR6uIZUsSKVOn3dWhsOrygAzeny5BXfBT+9nULoHgbifWPxYXpQuZxZlcyrFSKsVNNl3ZxOXcy3ipvHiy1ZNWXasumC2AevbsyYIFC0hNTWXevHlcu3aNgQMH0rRpUz744AOumejQXR0BAQEolcpKnpm0tLRKHhwDM2fOpHfv3rz11lu0a9eOIUOGMHfuXBYvXkxKSgoAwcHBtZqzwWH0AJ2w7LzZV+HMOv33Xc0sk25mGEztrCQ2TC+oasoDaihIKhW63Bx9s9bbkeW7PhRmcLOfTD9JYVlF0WvI/xkYOdDmdgkEdzvOSmfaN2oPWDcPSCfrjN6fMTFjcHd2t9padaXWSdCurq48++yz/PHHH5w9e5Ynn3ySr7/+mujoaO6//36z51GpVHTu3NmYT2QgLi6OXr16mbynsLAQxW1vOEqlvrKkwZXXs2fPSnNu2rSpyjkbHEE3ewPlJEJRtuXmPbgYZJ0+8TmwlXn3GBujVu8BAugSpf+kX109oIaEsY9ZVbvxbobCSs7ZbqeFIxHmEUawezAaWcORG0eM51MLUjly4wgSEgMizM8ZFAgElsPooU21ngDakriF89nn8XD24KnWT1ltnfpQr1YYTZs25e233+add97By8uLjRs31ur+N998k4ULF7J48WJOnTrFG2+8QWJiojGkNXnyZMaOHWscP2LECH799VfmzZvHxYsX2b17N5MmTaJbt26E3uzg/dprr7Fp0yZmzZrF6dOnmTVrFps3b+b111+vz0N1HFx9wftmkralvEBlxXDoG/335ft+1UR5D1ANseROETfzgO4QAWToY4ayitLukoTnoEG4NG9uW8McBEmSTIbBtiRuAaBjYMc6dToXCAT1p3yOnjXygGRZ5usjXwPwdOun8VI5ZlpJnQXQ9u3befbZZwkODuYf//gHjzzyCLt3767VHKNGjWLOnDnMmDGDDh06sGPHDtatW0dkZCQAKSkpFWoCjRs3jtmzZ/Pll18SGxvL448/TsuWLSu03+jVqxfLly9nyZIltGvXjqVLl7JixQq6d+9e14fqeFg6EfrkaijM0Bc4bGm+F4/gtqBwgoIbkFN94rghEfpsWh45RWX1MNYxqLaPGYAs4xQc3KCTvuuLqd0mIvwlENiftgFtcVY4c6PoBlfzLL/p54+rf3Am6wxuTm6MiRlj8fktRa1aYVy9epWlS5eydOlSLl26RK9evfjiiy944okncHevW3zvlVde4ZVXXjF5benSpZXOvfrqq7z66qvVzvnYY4/x2GOP1cmeBkFwLJxdb7mt8Ibk5y7j9VvtzcVZrRdjKQl6L5BP1ZXAG3m6EOnvxpWMQg4nZnFfy4b/6b/KPmY3yVq2DLm4mOBpU5Gq8hTdwRg8QMduHKNEW0JeaR6HruvDpQMjhAASCOyF2klN24C2HEo7xMHrB4nwqn0Xh6qQZZmvj+q9P6NbjXboOl9me4AGDRpEdHQ0c+fO5bHHHuPUqVPs2rWL8ePH11n8COqIJROhk+Ih6SAoVdBpXO3vr0U9oM5mFkRsSHgOG1YxFKZU4jloEMHvzQCFguyffybpjTfR3dY0+G4g0iuSANcASnWlHL1xlK2JW5GRaRfQzmF7JQkEdwsGD62l84B2J+/mRMYJXJ1cGRsztuYb7IjZAsjV1ZWVK1dy7do1Zs2aRcuWLa1pl6A6DAIo7RToTPfXMpv9N/t+xYwEj0a1v98ogGpOhL4TBVBVfcx8H3+csDmfIjk7k7dpE1dfeBFtfoGdrbUtt+cBbbqyCRDhL4HAEShfD8hSyLLMV0e+AuDxFo/j72r/lkbVYbYAWrNmDQ899JBx15XAjvhFg7MbaIog40Ld5ynIgOMr9d9X1ferJgwtMZIPg1ZT7dAukfqdYAlXs9FobdeIz9pU1cfMa/BgwhfMR+HmRuGff5I4bhyazOqLRt5JpOSnEOqh35yw7uI6DqQcACDKK4qTGSdJyU+xp3kCwV1Nh0YdUEpKkguSSc5Ptsic+1L3ceTGEVyULoxrM84ic1qTeu0CE9gJhRICb26Hv16PPKDD34K2BELaQ+MudZsjoIW+cnRZIaSfqXZo80APPNVOFJZqOZ2aV7f16ohWJ7P3Qgb/S0hi74UMtDrLJidX1cfMvUcPIr75BqWvL8XHj3Pl6WcoS7bMi40jk5KfwvDVw1l8fDEAl3Mvo0Mveidtm8SotaMYvnq4EEECgZ1wc3ajjX8bwHL1gAzen0ebP0ojtzpEFGyMEEANleCbYbC6VoTWaeGA/s2Jbi9U3/erOhTKW20xagiDKRSScTu8LcNgG46ncM+srTy54E9eW57Akwv+5J5ZW9lw3DZvvq5tY4n8/nucQkMovXSJy089TcmFenjuGgBZJVmUaqvPeyrVlpJVcueEQwWChkbnYMvlAR1IPUD89XicFc6Mjx1f7/lsgRBADZX6JkKf3agvpujqC7FVtHQwFzM7w8OtPCBbFUTccDyFl787REpOcYXzqTnFvPzdIZuJIJcm0UT98AOqpk3RpKZy5amnKTpypOYbBQKBwEoYcvQskQdk2Pn1cLOHCXYPrvd8tkAIoIZKfWsBGbq+dxwDzq71s6UOneFtURBRq5OZ/ttJTAW7DOem/3bS4uGwqnAODibyu2Wo27dDm5PDlfHPkV/L2lkCgUBgKToGdkQhKUjMSyStMK3O8ySkJbAvZR9OkhMT2k6woIXWRQighkqQPnZLblKN3dgrkX4eLmwFJOhqgT9WgwC6fgLKiqod2j7cB4UESdlFpORUP7a+7L+UWcnzUx4ZSMkpZv8l2yUmO/n6Erl4Me69eiEXFnL1pZfv+s7xAoHAPniqPGnpq9/RXZ8wmMH782CzB40bHxoCQgA1VNRe4KOvmF1rL9CBm1vfWwwB36j62+IVBh5BIGsh5Wi1Q91dnGgdYpvGqGl5VYufuoyzFAp3dxp/NQ/PYUOhrIykN/9G1o8/2tQGgUAggPpvhz+efpxdSbtQSkomxprZSNtBEAKoIRPcVv+1NonQJfmQ8L3++9r0/aoOSYJQ8/OADGGwg5etK4ACPdUWHWdJFCoVYR9/jM+To0GWSZ0+g/R58+7q1hkCgcD2GAsi1nEnmMH780CTBwj3CreYXbZACKCGTF0SoY/9BCW54NcEmva3nC21yAPqZMgDSrSuAKpJTEhAiLeabtF+VrWjyvWVSoKnTCHgZiuYG599zvWZM5Gr6jAvEAgEFqZzoP61+2LORTKKMmp17+nM0/xx9Q8kJCa2bVjeHxACqGFjyAMytxaQLN/q+9V1Iigs+Os3FkSsuSJ0lyi94DiRnEthafXFE+vKieQcXlx2S4yZ2uQvA1NHxKBU1LEEgAWQJIlGk14l6F//AiDr22Ukv/02clnDbhjr6+KLSqmqdoxKqcLXxddGFgkEAlP4qH1o7tscqH0YbP7R+QAMjRpKtHe0xW2zNrVqhipwMAy1gNJO66sw19TI9MoeSDupryLd4SnL2mKoBZR5UZ+U7Va1VyXUW02wl5rU3GKOXM2hZ1PLlku/klHAs4sPkFeioXu0H093j2Dm+tOVEqIDPFQO05TVb+wYlL4+JE/+F7lrfkObk0PjOXNQuOp36OWuX0/qBx8S/O93KhVbdERCPEJYO3JttXV+fF18RU8wgcAB6BLUhXNZ54i/Hs/gqMFm3XMu6xxxV+IAeKFdHTsJ2BkhgBoyPlH6Ksyl+ZBxDgJbVz9+v16t0/Zxff0fS+LmB35NIfOC3gvUrOp+T5Ik0TnKl9+PphB/JdOiAigtr5gxi/aTnl9C6xAvFjzbBS+1Mw+0C2X/pUzS8orxVDszeeVRrueVMO+PC7wxqIXF1q8P3iNGoPTy4tprr1OwfQeJEyYSPm8uskZDypSp6PLySJkyFbeuXY3tNhyZEI8QIXAEggZA56DO/Hj6x1rlAS04qo8mDIocRDPfZtYyzaqIEFhDRqG4FQarKRE6NxlOr9V/381Cyc+3U5vGqFaoCJ1bXMaziw+QmFlIhJ8b3zzXFS+1MwBKhUTPpv481CGM/q0Cmfqg/nmbt/0Cl9Mdp0mpR9++RCxehMLLi6JDh7j8zBiSJ09GV1gIgK6ggNRp0+1spUAguJMwJEKfyzpHTklOjeMv5lxkw+UNQMP1/oAQQA0fcwsixi8FnQYiet7aPWZpalMQMeqWANJZoBBhcZmWid8c5FRKLgEeLiyb0K3a3V3DYoO5t0UjSjU6pqw54VC7r9w6dSJy2TKcGjWi9Nw5CnbsBK1Wf1GrJS8uTtQOEggEFiPANYBo72hkZA5dr/kD7MKjC5GRuS/8Plr5tbKBhdZBCKCGjjERuhoBpCnVCyCwnvcHKgqgGgRF6xAvXJ2V5BZruHAjv17LarQ6Jv14mP2XMvF0ceKb57oS6e9e7T2SJDH9wTaolAp2nL3BhuOp9bLB0qhbtiBs7n9N92iTJFKmTkOTUbsdGwKBQFAV5m6HT8xNZN2ldQC81O4lq9tlTYQAauiYUwvo1BrIv64vVthqhHVtUThBwQ3IuVrtUGelgvbh3kD9+oLJssw7q46z6eR1VE4KFjzbhTah3mbdGx3gzkv3NQX0LTHyS6yzI60uyLJM5vz5pgWQLItQmEAgsCiGvmA1CaCFxxailbXcE3YPbQLa2MI0qyEEUEMnMAaQID8VCtJNjzFUfu48Hpyq35pcL5zVtzxStWiMWp88oP9sPMOKg1dRSPDFkx3p0aR2ycGv3NeUCD83UnOL+XzLuTrbYWlKzp0jL24zVFUT6GYorOSc49gsEAgaLgYBdDrzNHmleSbHJOUn8duF3wB4sd2LNrPNWggB1NBx8QC/m/UXUk3UA0o9Bol79Z6ZzuOsb08tEqG7ROq3ytdVAC3ceZG5f1wAYOYjbRnSpvYdiNXOSqY/pBdti3Zd4kyq6X98W+PSvDmegwaCUlntuPQFCyg+e9ZGVgkEgjuVIPcgwj3D0ck6EtISTI5ZdGwRGllDj5AedAjsYFP7rIEQQHcC1VWENhQ+bD0CvGywJbkWAqhjhA8Al9ILyMgvqdUyqw5f4/3fTwHw1pCWjOoaUav7y9OvZSBD2wSj1cn8e/Uxh0iIliSJ4GnTULi5mQ6D3RRGuWt+49KDD3H15VcoPHTYxlYKBII7ierygFILUll1fhVwZ3h/QAigO4OqdoIVZcHRn/TfW6rvV00YBFDyYdBpqx3q46aieaAHUDsv0LbTabz1s77p6nO9o3nlZh5PfZgyIgY3lZIDl7NYeSip3vNZAid/f0KmTzOZUB728X+I+vlnPIcMAUkif9s2rjz1FFeeGUP+jh0OIeIEAkHDoro8oMXHF6PRaegS1MXYQLWhIwTQnYChIvTtidAJP4CmCALbQGQv29gS0EJfnLGsAG6cqXG4MQ/IzL5g8VeyePn7eDQ6mZEdQvn3A62RTHlIakmojyuvDdCXg5+57hTZhaX1ntMSeA4bVjEUplTiOWgQXsOG4do2lsafzaHJ77/j/dij4OxM4cGDXH3hRS49/Ag5v/+OrHGcxG6BQODYGITNyfSTFJYVGs/fKLzByrMrAXix/Z3h/QEhgO4MDB6gG6f1W95BnzxrCH91m2g6jGINFMpbbTFqkwhtRmf4s9fzeG7pAYrLdNzXshH/ebw9Cgv28XrunmiaB3qQUVDKfzbWLN5sQYVQGKBwdyd42tQKY1yaRBP6/vs02xyH37hxSG5ulJw+TfLf/s6F+x8ga/kKdCXVhxhz16/n7D19yN2wwWqPRSAQODah7qEEuwejkTUcuXHEeH7piaWU6kpp36g93YO729FCyyIE0J2ATwS4eIOuTN8SA+DCVsi6pD/f9gnb2lMHAXQ0KYcSTdUhs2tZhYxdtJ+cojI6Rvgw9+lOOCst++frrFTw/ki9mPxhfyIJV7MtOn9dcfL3J2TGdJQBAYTMmF5lGwznoCCC3v4nzbduIeDVv6L08aEsMZHUadM4P3AgGQsXos2vXHNJk5FBypSpaNPTSZkyVdQXEgjuUiRJqhQGyyjK4Kcz+lSKl9q/ZBGPu6MgBNCdgCRVbolh6PvV4Sn9TjFbUouK0NEB7vi5qyjV6DielGtyTEZ+CWMX7Sc1t5jmgR4sGdcVN5V12th1b+LPI53CkGX49+pjaC1QpdoSeA0bRotdO81qhKr08aHRX/5Cs61bCPrXZJyCg9HeSCft4084368/aZ/OMYocWZZJnTZNtNoQCARAuTygVL0A+vbktxRri2nj34beob3taZrFEQLoTsFYEfoYZF6Cc5v0P3edaHtbDAIo7SSUFVU7VJIkOt3sC3bIRCJ0fomG8UsPcDG9gFBvNd9O6IaPmxVrGQGTh7XGS+3E8aRcvt93xaprWROFmxt+Y8fSbNNGQj78EFWTJujy8sj4+mvO9x9A6oz3yPrue329IdFqQyC460nJT8FH7QPA0RtH2XVtF9+f+h6AB5o8QGqBY1XMry9CAN0J5FwD9c3qx1f2wh8zARkad9V3is+x8a4m78bgHqjvPWaqNtFtGPqCHbySWeF8iUbLS8viOXotB183Z76d0J0Qb1ermFyeRp4uvDVU39/mPxvPkJZXbPU1rYmkUuHzyMM0WfsbYV98jrptW+SSErJ++IHrH3xg4gbRakMguNtIyU9h+OrhvL7tdQA0soaXt7xMiVafP/h/B/6P4auHk5KfYkcrLYsQQA0dTQnM7wc7P9b/nHQQjq7Qf3/tAMzvCwvu04+zFZJUqzDYrYrQ2cbt21qdzN9+OsKu8+m4qZQsGd+NZoG2C+U91S2Cdo29ySvWMHPdaZuta00khQKvQYOI+mkF4UsWo/TzNT1QtNoQCO46skqyKNVWv/u1VFtKVkndK/c7GkIANXSUKvAOA6pKTFOAV5h+nC2phQBqG+aNkwLS80tYvPsSey+kM3XNcdYeTcFZKfH1mM50CPexrr23oVRIvD8yFkmCVYeT2HvhzvGGSJKEk78/2sxqXshEqw2BQHCHIwRQQ0eSoP+/gaqSdXX667bO3A/rpP9qhgD640yacWfBe2tP8eSCfXz3ZyIAs5/oQJ/mjaxmZnW0a+zDM90jAXj3f8cp1VTRl6sBUmOrjZv1hlyaN7etYQKBQGAjhAC6E2g64NbW8/JIN2vyNB1ge5sM9mRehMLMKodtOJ7Cy98dokxrWsA5K+275fLvg1sS4KHifFo+i3ZdsqstlqTaVhuSZLLekEAgENxJCAF0J2D0At2GrLWP9wfAzQ/8mui/TzbdF0yrk5n+28kqfVcSMP23k/Xbip5zDZITqj5qSBD3dnPmX/e3BuDzLedIyq5+V1tDospWG7JMyPRpVdYbEggEgjsBuwuguXPnEh0djVqtpnPnzuzcubPKsePGjUOSpEpHmzZtjGOWLl1qckxxccPeyVMjTQdASEeQbv5K7en9MVBDY9T9lzJJyan69yIDKTnF7L9UtQepWgwJ4vP7Vn2YkSD+cMcwukX7UVSmZcZvJhrONmAqtdoAUCpx73Ov/YwSCAQCG2BXAbRixQpef/113nnnHQ4fPkyfPn0YNmwYiYmJJsd/9tlnpKSkGI+rV6/i5+fH448/XmGcl5dXhXEpKSmo1WpbPCT7IUkw4N8g38xTsaf3x0ANAsjc7eV13oZuTBCv6s/cvARxSdInRDspJDaeuM7W09frZo8DcnurDRQK0GrJ+fVX+xomEAgEVsauAmj27NlMmDCBiRMn0rp1a+bMmUN4eDjz5s0zOd7b25vg4GDjcfDgQbKyshg/fnyFcZIkVRgXHBxsi4djf8rnAtnb+wMVd4KZ6E4e6GmeKDV3XCWMocGqkpfNTxBvEeTJhHuiAZi65gTFZdV3um9IlG+14fPoIwBkfvstsvbOeYwCgaB6fF18UdXwYVClVOHrUkX5jAaI3QRQaWkp8fHxDB48uML5wYMHs2fPHrPmWLRoEQMHDiQyMrLC+fz8fCIjI2ncuDHDhw/n8OHD1c5TUlJCbm5uhaNBIkkwYCoEtNR/tXfPluC2oHCCgjR9Ls5tdIv2I8RbXeUGfgkI8VbTLdqv7jYYRKF02596HUKEkwY0J8RbzdXMIuZuO193mxwQQ6uNoH/9C6W3N2XXrpG3dau9zRI4OA29iW5Dtt/Stod4hLB25FpWDF9R5bF25FpCPEIssp4jYDcBlJ6ejlarJSgoqML5oKAgUlNrLredkpLC+vXrmTixYquHVq1asXTpUtasWcOPP/6IWq2md+/enKumnsnMmTPx9vY2HuHh4XV7UI5A037w1/36r/bG2fVWiw4T2+GVCompI2KAylWMDD9PHRGDsj4d3w1eIPk2L1AdQoTuLk5Ge7/afpGLNyo3Fm3oKFxd8Rk9GoDMb76xszUCR6ahN9FtyPZby/YQjxBi/GOqPO4k8QMOkAR9e2dZWZbN6ja7dOlSfHx8GDlyZIXzPXr04JlnnqF9+/b06dOHn376iRYtWvDFF19UOdfkyZPJyckxHlevXq3TYxGYoIaCiENjQ5j3TCeCvSuGuYK91cx7phNDYy3wD+cbfdsJqc4hwiFtgrmvZSNKtTqm/O+EsXL1nYTvU0+BszNFB+MpOnbc3uYIHJCG3kS3IdvfkG13NOwmgAICAlAqlZW8PWlpaZW8QrcjyzKLFy9mzJgxqFTVxywVCgVdu3at1gPk4uKCl5dXhUNgIUINBRFNJ0KDXgTt+md/fny+B5+N7sCPz/dg1z/7W0b8APw597YTMtz7zzqFCCVJYvqDbVA5Kdh1Pp3fj905fXEMOAcF4n3/MEB4gQSmyVu/3iZNdK0VorKV/dagIdvuaNhNAKlUKjp37kxcXFyF83FxcfTq1avae7dv38758+eZMGFCjevIskxCQgIhIXeW667BYPAAJR8GXdVJtUqFRM+m/jzUIYyeTf3rF/YqT0E6HNZ3M8a/2a3zRXXvZxPp785f7tPPNeO3k+QVl9XHQofE79lnAcjdsIEyM0LSgrsHTUYGKVOnmSygackmutYK89jKfmvQkG13ROwaAnvzzTdZuHAhixcv5tSpU7zxxhskJiby0ksvAfrQ1NixYyvdt2jRIrp3705sbGyla9OnT2fjxo1cvHiRhIQEJkyYQEJCgnFOgY1p1BKc3aGsAG6csf36+xeApkgf8rr/Y3AL0J8/uKhe077YtwlR/m6k5ZUwZ/Od1y9LHRODW9euoNGQ9f339jZH4CBUCL+YKKCpy8sj8YUXKNi3n7KkpDrvJLRkmEfW6Si7nkZhfDzZq1dzZcxYdPn5pu134HBSjc+9A9vuqDjZc/FRo0aRkZHBjBkzSElJITY2lnXr1hl3daWkpFSqCZSTk8PKlSv57LPPTM6ZnZ3NCy+8QGpqKt7e3nTs2JEdO3bQrVs3qz8egQkUN3dbXdmlrwgdFGO7tUsLYf98/fe9JukTw1/ZC7NjIOmg3itlqoWIGaidlcx4KJaxi/ezZPclmgd64KpSEuip37VmMQ+WHfEbP47CAwfI+ulnAl5++VatIMFdS8m5c/rwS1XIMiUnTpJ404OIkxPOoaGoGofhHNYY5/Bw/feN9d8rfXxM5nwawzwGyoV5vIYNqzRem5tL2bVrlF69Rtm1a5Qllf8+Cbm0+i7nt69Tcu6cw/XBKzlbw3PvwLY7KpJ8J2Zx1pPc3Fy8vb3JyckR+UCWYNO7sOdz6PIcDP/UduvuXwDr/g6+UfDXeFDe1Pu/TIDjv0DHZ+Ch/9ZriUfm7uZQYnaFcyHeaqaOiLFcDpOdkHU6LgwbRtmVRIKmvIvfU0/Z2ySBnZFlmcTxz1H455+mB0gSSl9flJ6elCYnQ1n14WGFm5teDDVujCq8Mc5hjVF4e3N9xozKng5JQnJ1JeCVl9FmZunFzbVrlF67hq6m0iVKJc7BwTiFhVF27RqalBSTtckAJE9PQqdPw3PIEKSqmgXbEFmnI3/bNtK/nk/x0aNVD1Qq8ezfn8ZffG474xyQ2rx/CwFkAiGALMyJ1fDzsxDSHl7cYZs1tRr4sjNkXdaHvro9f+ta4p+weAg4qeFvp8G1boW9NhxP4aXvKid3Gz7PWmwXmx3J/O57rr//PqrISJqsX4eksPvG0Urkrl9P6gcfEvzvd/AaOtTe5tzR5G7aRNI/34YiEz3xJAmFpydN16/Dyd8fWatFk5ZW2TNzLYmyq1fRpKVZ1Dalvz/OjcNQ3fQ0OTcOQ3XT0+QcFITk7Azo82guDB1mOgxWDueICPwnTMD74ZEoathsYw3ksjJyfv+djIULKT1/4aZR+seARlPJdoWnJ003rL/re/gJAVRPhACyMNlXYU6sviji5Gv6+kDW5viv8Mt4cPWDN06Aqlz4Rpbhq3vg+nEY/AH0+uv/t3ff4U2VbQCHf0n3LoVOyihl7z1lb1BBtjIFAbcIKjg+ARcgDkRlyhRcCCjIRih7L9mz7A5aoHsm5/vjtIW2SZo0s+S9r6sXIT055z3NOE/e8TwG716llnhqxg6ttcwUyEv5907sUKKHw9QpKVxu3wF1YiKhc+bg1cEG8ks9Ju9ilpSE0ts77+IrmJakVnPv+++JnzsPAOfKlcm8UjgZaNlvv9E4RKWJOiODrDt3ybp9i8zbt8m6fYf08+dJPXCgyMd6P/sMbrXr5PQelcW5bFmUHh56n0/ixo3cGT+h0P1Bn35KdkwMD37+GVVCAgCO/v74jRiO78BBOHjqf4ziUqel8XD1Gu4vXkzW3bsAKD09KfX88/gNG0rqkSMa2+7VpQuhszVPDbEnhly/be/rnPDk8QkFD39QZ0P0afMfT5JgX84HQdMx+YMfkFdQNMlJoHl0Eai1lcrQzuyFXG2E0sODUgPkWnv3ly61bmMKEPlQLEOVlMTtV1/LC378RowgbO2a/EV0HRzw6txZ7+AHQOnigkulMDzbtMHvhRcIfO9dyi9eVLg47+NyjlP2yy/xGzYUrw7tca1a1aDgBzQUAc7Zb6n+/fB//TUq79xB4PuTcAwKIvvePWJnfsWVDh2I/e47su+b5z2tSkggbt48rnTsRMxnn5F19y4OpUvj//bbVN65g4AJ43H09y/c9pxe2eSICDJFDjuDiABIMD+FosiEiCZ1fQ9EnQRHt/xDX4+r0x9cvOH+NbhmeMkHvQu5JhazkKsNKTV4MDg4kHr4MOnnz1u7OXlKei6akiDj2jWu9x9AckQEChcXQr6cQeCkiSidnPIV0VV6eBA0ZbLRx8tXnFfDUm+zHIfC7Ve6u+M3fDiVt24h+PPPcQ4LQ52YSPzceVzp0JHozz7P653RRZ/XTlZsLDEzZ3KlQ0fuzfoO1f37OJUtS+DH/6Pyv9spM3YMDl5e2tvu6Yl748ZImZnEfP5Fcf8kdkkEQIJlWDIA2pczCbDBYPAoo3kbF0+onzOp94jhS+L1LdD6/c7LHLlesnuBnIKD8+bW3F9qG4kRS3oumpIgaccOrvcfQOb16zgGB1Nh5Up8nn027/ePF9EN/mSqyYYeHUuXJnjqFI1LvYOnTjHtcYpov8LZGd++faj0z3rKzv4O19q1kdLTebBiBVe6dOXuxElkaBgKhKJfO5k3bxL18WSuduzE/UWLUaek4FKlCiEzvyR8y2b8XngBpavmz5mCbQ/6ZCo4OZEcEUHSjp3G/3HshJgDpIGYA2QGV7bDir7gFw5vas8KbbSYszC3pVz89I1j4FdJ+7b3LsGPTeRt3zoFvuX1PkzuHKDohHS0vYEUkPe7zjUDmditOpUDPPU+hi1JO32a6/0HgJMTlf/djlNAgNXaIkkSd958U/6g15JnxjEwEN++fXEOq4hzxYo4V6iAg4Hv5ULHsZNVNpJaTdzcucR9/wMA7o0bU/a7WRadW2Wrf3tJkkg9eJC4BQtIPfBoJZxnx46UGTMat3r18rbT1v708+eJX/iT3CuUM/zu1qABpceMxrNt22IvNIj9+mviF/6EU2golf5ZrzV4etKJSdBGEgGQGaTehy9zanK9FwnuRlR412Xty3DqV6jZGwbo0Vux7BmI3A1PjYdOhnWtbz4TxSs5q8AefxPl9klM71OHU3cS+P3ILVRqCQelgkFNyvFWpyp69yDZkusvDCbt+HFKvzyWgHHjrNaO9EuXiHy2l8GPc/Dzk4OhvJ8K8r/ly2u8WGibKGvIRN+SRpWcwt1JE0ne/i8gD38GTpqYt4LKkmx9gnva6dPEL1hI0vbteb1V7s2aUXr0aFQJCdydUPi141KjOhnnL+T936NNa8qMGYN748ZGt0edksLVnk+THR1Nmddfx//114zeZ0kkAiAjiQDITGY3kOfcDFkDlQ0vRFqkhNvwXT15svXoHY+G3XQ5tw7+GCpniB5/DhxdDDrk5jNRTF1/Lt+E6IJ5gK7EJjFj80W2nYsBwN3ZgTFtKjG6dSU8XKyai9QgiVu3cufNt3Dw9aXyzh0o3Sywmk+DvG/X/+7QPIFdqcQ5PBz3+vXJvH6dzOvXyb53T/sOFQqcgoPzBUcOpf2I/niyxlw0jy/1fpJkXr/OrddfJ/PKVRQ5c3x8+/axaptKQoqDjKtXiV+0mIR16+Tl6SBPTNa2uEKhwLt7d0qPGY1r9eombUvipk3ceXs8ChcXKv2zHudy5Uy6/5JABEBGEgGQmax+CU6vgvYfQdt3Tb//LR/CgR+gYmsY8Y9+j1Flw6w6kHQX+iyEugMMPqxKLXE48j6xSek6M0EfjrzPFxvPc/LWQwDKeLrwducqDGxcDkcH25+OJ6lUXO3ajazbtwmaMoVSgwZarS1ZUVFc6dip8EVGS4CiSk4h6+YNMq9fJyMnKMq8foPMyEjUSUmGHdxGhmNMKXnXLu688y7qpCQcAwII/X523nCOoJ+sqCjiFi/m4cpfdAY/Hq1aUv6nn8zSBkmSuDlyJKkHDuLZvj3l5hYsBP3kEwGQkUQAZCYH58LmSVC1O7zwm2n3nfYQvq0Fmckw+E+o0ln/x+76EnZ+DuWawaitpm1XAZIkselMNF9uvsD1eHn5dri/BxO7VadzzUCNZQFsyf1ly4iZNh3nSpWo9M96qyVGjP3qK+J/0jx53ZAhKkmSUD14IAdEkXJglHbmjF65aCqtX1fiSw5IkkT8goXcmzULJAm3Bg0Inf0djv7+1m5aiaTv8Kw5XzsZV69yrVdvyM4mdO4cvNrbVu4ucxN5gATbFNJQ/vfOMZ0ZWIvl6GI5+AmoCZU7GfbYhsPkJI23DkGUjlTzJqBQKOhRJ5itb7dl6rO18PNw5uq9FMb8fIwB8w9w/Gbxq9Rbgk/fvig9Pcm8do2UvXut0oaU/fvzgh/XevWMykWjUChw9PPDvWFDfPv2IWDC+KJz0QAerZ8q8cGPOiWFO+Pe5t6334Ik4TtwIBWWLRXBjxFcqlTRK4+ROV87LuHhlB4h12KL+WIa6owMsx2rpBMBkGA5wXVB4QApsfJ8HVPJzoBDcpI2Wr5ZeGl0UbyCoMYz8u0j5umaLsjZUcnwlhXZ9W47Xm9fGVcnJUeuP6DPnP28uvIYkXEpeduq1BIHrsbz98k7HLgaj0ptvU5bB09PfPv1A6yTGDH7wQPuTpwEgO+ggZSb86Nlc9HkSD16jPu//IJUjCSalqYpF03mrVtcf/4FkrZsAScngqZOJXjqFBRWKPnwJLFUHqOilHnlFRwDA8m6dYt4Mw23PQnEEJgGYgjMjOa1huj/oP8yqNXbNPs8vhzWvQHeZeHNk+BYjA/x6/tgaQ9wcofx58HN1zRt01N0QjrfbLvIn8duo5bAUalgcLPy1C7rwzfbLumcZG1pmbfvcLVLF1CrCfv7b1yrVbXIcSVJ4varr5G8cyfO4eGE/bkKpZub2SbKalsF5hQWRlZkJADuTZoQ/NmnOFeoYLLjmpKmlVTpFy5wZ/wE1AkJOPiXIfS72bg3bGDtpj5RbGEFYb4J0Rv+wTk01CLHtTYxBCbYrtyVWXdNlAtIrYb938u3m79SvOAHoEJL8K8BWanyMnoLC/Jx5ct+9dj4VmvaV/MnWy2x7MAN3v3zv0IlN6IT0nllxXE2n4myeDsBnEPL4tWlCwD3l1suMeKDX38leedOFE5OlP36q7xVaN7du1N17x6TrxLSVi4hfMM/BH74IQo3N1KPHOFar97EL12KpCUnkbVoKhVyY8SL3Bo9BnVCAq516xL2558i+DEDba8dS6ZP8OrWDffmzZEyMoj5YprFjluSiABIsKy8jNAmCoAubYa4S+DiAw2HF38/CgU0zakPduQn089R0lP1IG+WvNiUFSOb4qSliGpuy6auP2e14TC/4cMASFy3nuy4OLMfL/3SJWJnfAlAwLvvmHz5sCbayiUolEr8hg6h0rq/5QtMejqx02dwY/AQMq5dM3u79KWpVEjm5cugVuPTtw8Vfl6OU2CgdRv5hCqq1IbF2vC/j8DRkeQdO0iKiLDo8UsCEQAJlpXXA3QC1Cb4xrw/Zyly4xfB1cjhyroDwdkL4q9A5C7j22YEBwclWTqCG2sXW3Vv0ADXenWRsrJ48KuJV/QVoE5P5+6Ed5AyMvBo05pSQ4ea9XiP01UuwblcOcovWUzQ1KkoPTxIO3mSyN7PEbdwIVJuPhgr0VoqBFC4uuL/9tsoXQzLeSUYxlylQgzhEh6e92Ul5vMvxIToAkQAJFiWfzVw8pBXbMVdMm5ftw7DzQPg4AzNXja+bS5eUG+QfPvwQuP3ZwS9i63quZ05lB4xApCHpsz5wRr71ddkXL6MQ+nShHzxhcVTBegaYlMoFJQaOIBK/6zHo3VrpMxM7n39DdcHDiL9opGv72KSJImoyVNQp6Ro7MmUsrKImfqJFVpmf8w1PGuIMq+8imNAgJgQrYEIgATLUjpASH35trGFUfd9J/9bdwB4m2hCcJNR8r8XN0LCHdPssxj0LZVhzZIaXp074xgSjOr+fRLXrzfLMZIiIniwYgUAIdOn4VhGS3FbK3MKDqbcgvkET5uG0tub9LNniezXj3s//oiUmWmRNkhqNanHT3B30vskb9+uPRmfSkXStm1kXL5skXYJ1uXg6UHgpIkAxC9YSOZtE67ALeFEACRYXtnH8gEVV9wVuLBBvt3yTePblCugBlR4CiQ1HFtiuv0aqGmYH8E+rujq6/DzcKZpmJlqqulB4eiI3+AhgJwg0dQLSrNiY4l6/wNAnnPk2bq1SfdvagqFAt/nelNp/Xo8O3SArCzivv+ByP4DSDt71izHlLKzSTl4iOhPPuVKu/bceOEFEv/+W/eDLJCLRrAtXt27iwnRGogASLC8vInQRgRAB74HJDmrtH81kzQrT+5k6GPLINsy394LclAqmPxMTQCtQVBKRjbn7iZarlEa+Pbvh9LdnYzLV0jZv99k+5XUaqLe/wDVgwe4VK+Ov4bCkrbKKTCA0B9/IOTrr3Dw9SXj4kWuDxhI7LezUGvoDdKUp0cXKSuL5D17ifrfx1xu05abI0bw4JdfyI6NRenpifczzxD0xecovTytmotGsB0KhYKgjz4UE6ILEAGQYHm5AVDMWcgqxhyW5Fg4mbNUvdVbpmtXrupPg2eQnLDxgnmGdvTRrXYwc4c0JMgn/zBXkI8r1QI9ychWM2LJYa7dS7ZSC8HB2xufvn0BuL/UdEvi7y9bTsq+fShcXSn71UyUJSxBn0KhwKdnTypt+Aev7t1ApSJ+/nwin+tD2qlTedtlx8cT9fFkVHFxRH08mez4eI37U2dkkLRjB3cnTuJSq6e4NXo0D1etQnX/Pg4+Pvj06UPovLlU2b+PsjO/pFSfPgRPnVp4DpAkETx1yhNXyFUomkvlyvgNExOiHycSIWogEiGamSTBV1Ug5R6M2g7lmhj2+H8/hT1fQWgTGLXN8MzP+tg5DXZNh/ItYeQm0+/fAJqKraZlqXh+wUFO30mgrK8ba15tSaC3deYDZd68ydWu3UCSqLThH1zCw43aX/q5c0QOHARZWVYvumoqiVu3Ev3Jp6ji4kCpxG/4cMq88TpREyeStGOnvFS9QJFVdWoqybv3kLR1K8kREXn5fAAcypTBq1NHvLt0wb1JExROToWOKUkSd958U+v+BfujSk7hWo8eZMfGUubNN/B/9VVrN8nkRDFUI4kAyAJ+GSjn8Ok2A5obsIIrI1kuepr+EAaueFTCwtQS78K3tUFSwSv7IbCWeY5jhLjkDPrPO0BkXArVAr34Y2wLfNwLXwgt4fYbb5C0bTu+AwYQ/MnUYu9HnZpKZN9+ZEZG4tmpI6Hff2/zBWL1lf3gAbHTp5Pw9zoAHMqURhVXuMen1JDBZMfEkLxnL1L6ox5Sx6AgOZlel864NWyIQketsrxjasgELXp/7FvChg3cnfDOE5shWmSCFmxfcecBnfhZDn78wqFaD5M3K493CFTvKd+2UH0wQ5XxdGH5yKYEeLlwMSaJUcuOkJZpnWzEfsPlJJQJf/9N9oPiF3SNmTadzMhIHAMCCP700ycm+AFwLFWKkBkzCJ03Fwd/f43BD8CDFStJ2rYdKT0dp3Ll8Bs1koq//0blHf8S9OEHco+PHsEP2EYuGsG2ePfogXuzZvKE6GnTrd0cqxIBkGAdIcVYCabKggM/yrdbviEvqTenpqPlf0/9DunWnWysTTk/d5aPaoq3qyNHbzzgtV+Ok6WyfIFOt0aNcK1VCykjg4e/FS8xYuKWrTxctQoUCkK+nIFjqVImbqVt8GzbFrfatXQO3TqFhRG2ZjXhW7cQ+O67uNWrh0JZvI9rW8hFI9iOfBmi//2X5F3WTfpqTSIAEqwjdyn8/auQqmc247N/QcIt8PCHes+brWl5KraGMtUgKwX++938xyum6kHeLBrRBBdHJTsuxDJx9X+oLVwiQ6FQ4JeTGPH+L79oXO2kS1ZUFFEffwxA6ZdewqN5c1M30WZkXL5M8s4IneVWsiIjUTg5PVE9YILteHxCdLQdT4gWAZBgHe5+UCpMvn33RNHbS9KjxIfNxoKTBSb8KhTQJGdJ/OGFVqsPpo8mFf348YWGOCgVrDl+h+mbL1i8Dd7duuIYGIjqXhyJGzbq/ThJpeLuexPlAp116uD/5htmbKX1uVSpkr9QZkEiT49gAWVezckQffMm9xcvtvjxDU3/YA4iABKsx5DCqNd2QsxpuYxG41Hmbdfj6g2Ujxl3Ea7vtdxxi6FTzUBm9K0LwILd15i/66pFj69wcqLUkMGAYYkR4xcuJPXIEZTu7pT9aqbGFU1PknyFMkWeHsFKHDw9CHjvPQDi5s0n87blMt/rm/7B3EQAJFhPXmFUPQKg3N6fhsPk3iNLcfWRS20AHLFufTB99GsUygc95Erp0zZdYNXRWxY9fqkBA1C4uZFx4QKphw4XuX3ayZPc+/4HAAI//h/OFSqYu4k2wbF0aYKnThF5egSr8u7ZA/emTeUJ0dMtkyFakiSip0zJS+ugTkkhekrxV44aQwRAgvXkBkC3j+oeXoo6BdciQOEALayQtyJ3GOz8P5AYZfnjG2hMm3DGtqkEwKQ1p9l+LsZix3bw8cH3ud4A3F+6VOe2quRk7rzzLqhUePfsiU+vXuZvoA3x6t49/1BYztCXd/fu1m2YYDfyTYje/i/Ju3eb/ZhJmzaRtG27nJsK8mrTJW6yfL41EQAJ1hNcVw5qUmIhUUf3676cxG21+4Bvecu07XFBtaF8Czkn0HHTZTs2p0ndq9OvUSgqtcRrvxznyHU9J5qbQKmhQ0GhIDkigozISK3bRX/yCVm3b+NUtixBUybb3YTffENhIIa+BKtwqVIFv6FDAYj+7HOzTojOjo8navIUjUO/UZOnWHwoTARAgvU4uT1KMKhtOfyDG3B2rXzblEVPDZXbC3R0ibwc38YpFAqm96lDx+oBZGSrGbn0COejLLOU3yUsDM927QB48PPPGrdJWLeOxHXrQakkZOZMHLy8LNI2WyPy9Ai2oMxrr+Ho72/WCdGSJBE1eQrqlBSNQ7/WGAoTAZBgXUVVhj84R+55qdRe7jGylhrPgkcAJEc/qkJv4xwdlPzwQkOaVCxFUno2wxcf5tb91KIfaAK5iREfrv0L1cOH+X6XeesW0VM/AaDMa6/i3rCBRdpkq0SeHsHaHDw9CJg4EYC4+QvIvH3HJKu01OnppB47RvyiRdwcOZLk7dtBrSVPWc5QWMbly8U+nqFEACRYl66VYKn34fhy+bY5ip4awtEZGskXdVvNDK2Jm7MDPw1rQvUgL2KTMhi66BBxyebP+eHerCkuNWogpaXx4I9VeR+mCf9s4M4776BOScGtUSPKjB1r9rYIglA07549cG/SBCk9nahPphq8SkuSJDJv3yHhnw1Ef/Y5kf36c7FxE24MHkLszK9IPXBQ9w6skP7B6gHQnDlzCAsLw9XVlUaNGrFnzx6t244YMQKFQlHop1at/HWaVq9eTc2aNXFxcaFmzZqsXbvW3KchFFfeSrAToC5QxuHIIshKhaA6UKmdxZtWSKMRoFDC9T0Qa/k8O8Xl4+7EspFNCS3lxvX4VEYsOUxSunmH8RQKBX7D5URr95cvf/Rh+sEHpJ/6D6WXF2W/nIHC0dGs7RAEQT8KhYKgj/8HSiWpu/fIQ1VoX6WlTk8n9ehR4n/6idtvvMHlNm242qkTd995hwcrVpB+5gxkZ+PgXwavzp0IePcdQuf8iNLLy2bSP1i1GOrvv//O0KFDmTNnDq1atWL+/Pn89NNPnDt3jvLlC092TUhIIC0tLe//2dnZ1KtXjzfeeIMpU6YAcODAAVq3bs2nn37Kc889x9q1a/n444/Zu3cvzZo106tdohiqBalVMK2cnG351UMQIC/hJitNLkaaGgd9F0GdftZtZ67fBsOFf6DpGOgx09qtMci1e8n0n3eA+JRMWlQqzZIXm+DqZL5yIlJmJpc6dEAdFw9KZb6u77LffiNWOwmCDbo59mVSNJTHCPzoIxx8fUk7eZK0kydJv3ABsrPzb+ToiGuNGrjVr49bvXq41a+PU9mQfAscEjdu5M74CYX2b6rPhBJTDb5Zs2Y0bNiQuXPn5t1Xo0YNevfuzbRpReck+Ouvv+jTpw+RkZFUyMkfMnDgQBITE9n02JK6bt26UapUKX799Ve92iUCIAtb0gNu7INec6CBnEiPo4vhn7fBpzy8eQIcbKSn4OpO+Lk3OHvBhPPgUrIm756+ncCgBQdIyVTRvXYQP+RkjzaX22+PJ0nD8lYRAAmC7cmOj+dq126ok5P12t7Bvwzu9evLAU/9+rjWqoXSVXeWfkmSuPPmmyTt2CkvhXdwwKtDB0K/n22KUzDo+m21q0pmZibHjh1j0qRJ+e7v0qUL+/fv12sfixYtolOnTnnBD8g9QG+//Xa+7bp27cqsWbO07icjI4OMx5b+JSbaZuHLJ1bZhnIAdOeYHACpVbD/e/l3LV6zneAHIKwtlK4M8Vfgvz+giRmzUifchpQ47b/38Aefsgbtsk6oDwuHNWbEkiNsOhPN//4+wyfP1uLI9QfEJqUT4OVK0zA/kwRF2fHxpGga0s5Z8uretKlY9SRopVJLHI68b/LXpaBZXoLCx0ZZClJ6e+PTqxdu9evhXr8+jiEhBqevyE3/kHLwEOqkJKumf7DalSUuLg6VSkVgYGC++wMDA4mOji7y8VFRUWzatIlffvkl3/3R0dEG73PatGlMnWqdTJQCj02EzlkJdmED3L8Grr7QYIjVmqWRUikvid88SZ4M3XikzqrexZadAQvayzmStPEMgHFnwNHFoF23rFyGWYPq89ovx/nl0E3WnbxDcsaj+VfBPq5MfqYm3WoHF7f1uj9MH1vyaqpvfcKTZfOZKKauP0dUQnrefaZ4XQraZVy+LCco1EGdmEipAf2Nnqicm/4h+vMvCProQ6t9EbL6JOiC0aMkSXpFlEuXLsXX15fevXsbvc/333+fhISEvJ9btyxbPsDuheQshY85A1npj8peNB0NLp7Wa5c29Z4HRzeIPQc3D5jnGA7OOb072t6iSvAuK29XDD3qBDOoSTmAfMEPQHRCOq+sOM7mM8XPep33YapSad7ACktehZJh85koXllxPF/wA6Z5XQraWbpIry2kf7BaAFSmTBkcHBwK9czExsYW6sEpSJIkFi9ezNChQ3F2zn8BCAoKMnifLi4ueHt75/sRLCThNqTdl3t71NmwfTLcOQpKJ6jQChIsV6BPb26+ULe/fNtcS+IVCujwEaAlZwZq+ffF7H1SqSUiLt7T+LvcSYFT159DpS7eFEFR8VwoDpVaYur6c2h61ZnidSloZ49Feq0WADk7O9OoUSO2bduW7/5t27bRsmVLnY/dtWsXV65cYdSowvMvWrRoUWifW7duLXKfghXkDvMsaAfpD+X7Ds2T/1VnyZONF7aTt7M1TUbL/55bB0lmqrXlGQSeGgJ3hQOENIDwjsXe9eHI+4W+YT9OAqIS0jkcWbwSGvb4YSoYz9yvS0E3eyvSa9UhsPHjx/PTTz+xePFizp8/z9tvv83Nmzd5+eWXAXloatiwYYUet2jRIpo1a0bt2rUL/e6tt95i69atzJgxgwsXLjBjxgy2b9/OuHHjzH06gqHMPMxjVsF1IbSpHKjlJms0hYc3Yc83MKclzGsFyRqCK0llVO8PQGyS9otMcbbTxN4+TAXjXYrRbwGKMa9LQTd7KtJr1QBo4MCBzJo1i08++YT69euze/duNm7cmLeqKyoqips3b+Z7TEJCAqtXr9bY+wPQsmVLfvvtN5YsWULdunVZunQpv//+u945gAQLMvMwj9k1zekFOrYEVNm6t9Ul9b6c9HFxN5hVB/6dCrFn5cCvWk8oFSb3+uRycIayTYxqeoCX7qWqhm6njT19mArFl5iexYzNF/jsn/N6bf/3yTvEJoogyBzsqUivVfMA2SqRB8iCJAkWtoeoUyA9FggpHOReltE7bTcAys6Ab2rKyRoHroAaz+j/2MxUuLgRTq+CK9vl+U8AKKDiU1CnP9R8FtxKyb9f0Tf/42v1gX6LjZoD9NSMHUQnpGucb6EAgnxc2Tuxg9FLj7Pj47narbu85NXbm/BNG0XvjwBAZraalYduMPvfyzxIlbOTOzsoyFQVfVlyc3JgdOswxrQNx9PFhlJlPCESN23KW6VVkurUlZhEiLZKBEAWpukCDzBkNVTuZPn26CvhNuz8Ak6ulJfy9/wm/+8L5ulRZcO1CDj9B5z/R85+nSuoDtQZALX7Fs7tkxsk3j0BZarJOYgkFTzznVyeo5hyV9sAhYIgBTB3SEOTLTkuqR+mgnlIksSG01F8ufkiN3MK9Ib7ezCpew2yVWpeXVn4dZkbhk/oUpUdF2I5fvMhAGU8nXmrYxUGNS2Pk4PVFzYLViYCICOJAMjC8nqB/pMv7CWl9+fb2kXn6XnrNESfloOeM2vk3qJcvhXknp46/R+VANHm6k7YNBG6z5D3t+1/4OgKo3dAYC3dj9VBU74VgDGtw/igZ81i71cQtDl0LZ4vNl3g1K2HAPh7ufB2p6oMaByKY04AU1QeIEmS2HI2mhmbLxIZJ3+RqFTGg/e6VadrrUCDk/MJTw4RABlJBEBWULAXyNZ7f/J6ZU6heQ6TQl7B5egCD288utu9tDx8Vac/lGtavABPrYZfB8LlrVCmqhwoGpEv6fGMuzsvxPLXybvUL+fL2ldbiguJYDJXYpOYvukC28/LXxrcnR0Y2yacl1qH4aFhCEufTNBZKjW/Hb7JrO2XiU/JBKBRhVJ80KM6jSr4mf+kBJsjAiAjiQDICh4f5glpYNu9P7m0Dd0V5OQB1XtC3QFyVXsHJ+OPnRIvrxJLioJ6L8Bzc4t+jB7uJWXQasYOMrPV/DG2BU3DxEVEME5sYjrfbr/M70duopbAQang+ableKtjVfy9DMtirk1SehYLd19j4Z5I0rLk5JvdagXxXrdqVPK3wWSqgtmIAMhIIgCykseHecLbW7s1RSs4dFdQ5S5QbyBU6w7OHqY//vV9sOxpefJ477lQ/wWT7Pb9Nf/x6+FbdKoRwE/DjVttJtiv5IxsFuy+xsLd1/KCkq61AnmvW3XCdQUlRtTAi0lM59ttl/jj6K28YOuFpuV5s2MVkwVbgm0TAZCRRAAk6E1bL1C/pVD7OfMff9dM2PkZOLnDmF3gX9XoXV69l0ynb3YhSbB9fBsqB5SsivemJApyaqftb5OlUvPbkVt8t/0SccnysFTD8r580KMGjSsW0aOo79y6ImrgXYpJYsamC/x7Qd6Ph7MDY9vKw23uzo46228q5ty/eF1qJwIgI4kASNCbtSdwq1Xw83MQuQsCasHof8HJzejdjl5+lG3nYhjYuBwz+tU1QUNLHlGQUztNf5sgH1d61Qth27kYruVMTA4r48HEbtXoWitIv/lkRc6tU0JIPb3fXweuxjNt03n+u50AQICXC293roq3qyOfbThvtufWnK8d8brUTQRARhIBkGAQa0/gToqBeU/J35obvQjPzDJ6l0ev36ffvAM4OyjZO7E9Ad7GJUQsaXJTBGhKDwCmTRFQ0mj72zyutIcz4zoVc2l6UXPrDHx/qdU5S+63XODW/TSt25nquTXna0e8LosmAiAjiQBIMIgtTOC+ulPuCUKSEyTW1mNydhH6zNnH8ZsPebVdOO91K2KZ/hMkN0mktppUpkwSWdIU9bcB8HRxZO/E9vi6F7OEjba5dUb2rmZkq/j5wA0+33i+UHWWx3m7OvJGxyooi3EMtSQx+9/LJKVrzwxf3P0XtW97fl0+zpDrt0ifKQjGUiig42R5AnfHydZZvRbeHtq8A7tnwrq3ILg+lA43apdj2oTz8opjrDh4g1fbV7abbLuGFORsEW5fGa2L+tuAPPn5fFRS8f82CgW0eRd+KzCp38gaeC6ODtQK8dEZ/AAkpmfz+Qb9SnIUh7n2b8+vy+Kyj080QTC38Pbw+mHrtqHtJHll2M398OeLMGqbzomiRelcM5CwMh5ExqXwx5FbjHwqzISNtV1X7iXptZ09FuS8/SBVr+2M+tuosuDYssL3O7oZXQNP33Y1quBLaCl3g/d/+0Eqx248NMv+9d23Pb4ui0sEQILwpHBwhL4/yfOBok7Bto/llALF3Z1SwUutw/hw7RkW7Y1kWIsKeZl6n0QpGdks3HONuRFX9dre2EKxJUm2Ss2fx24zfdMFvbYv9t9GrYa/X4PLW0DpBOqsxxqRJvcKDVkNTsXbv77teqdL9WL1ohy4Gs/zCw+aZf/67tueXpfGenI/zQTBHvmUhefmy7cPzZNrjhmhb8NQSns4c+dhGhtOR5mggbYnWyUX5Gw7M4JZ2y+Tka3GyUH3MEtpD2e7SBIpSRI7LsTQY/YeJq05zcO0LHT9aRTIK5KK9beRJNjyAfz3uzzfZ+AKeU4dgH81cPKEG3vhz5FyXb1iaBrmR7CPK9pOwaj2m3n/Re0bwNfdyS5el6YiAiBBeNJU7QIt35Rv//0qPLihe3sdXJ0cGN6yIgALdl/jSVozIUkSW89G02XWbj5ce4a45AwqlHZnzuCGzB7UAAVovdikZGZzKUa/obKS6tSthzy/8CAjlx7lUkwyvu5OfPx0TWYNqq/xb5P7/8nP1CzeJNw9X8GhnIzmvedCtW7ynLoy1aDbDHjhN3BwgYsbYP1bFDmZRwMHpYLJz9TM116Ttd/M+9e171xJ6VkcuBpv8L7tlVgFpoFYBSaUeKosWNIdbh+Bso1h5OZil+B4kJJJy+k7SMtSsfKlZrSqXMbEjbW84zcfMG3jeY5cfwCAn4dcUfz5puVxdtRekDPI2xUPFweu3kvB38uFNa+0pJyf4XNFbNnN+FRmbr3I+lN3AXB2VDKyVRivtAvHx01+DZk8F83RxfDP2/LtbjOg+cuat7uwAX4fImc/b/kmdPnU8GNh/lw61sgDFOzjyvGbD3F3duDX0c2pV87XqOOUVGIZvJFEACQ8ER7cgPmtIT3BqIsFwOS/z7DswA3aVPVn+cimJmykZUXGpTBzywU2no4GwNVJyUtPVWJs20p4uRYOEDVl3E1Oz2bgggNciE6iYml3Vr3c8okos/AgJZPvd1zh54PXyVJJKBTQp0Eo47tUpaxv4eSaJstGfHYtrHoRkOTVXx0+0r39iRXyPCGATlPhqXGGH5MnLxN0tlrNyKVH2HclHj8PZ1a93EJ3yZEnlAiAjCQCIOGJcX69/I0Z4IVV8vBYMdy6n0rbmTtRS7DprdbUCC5Z74v45Axm/3uZlYdukq2WUCqgX6NQxneuRpCP4ZNGYxLT6Tt3P7cfpFErxJvfxjTXGECVBOlZKpbsu86ciCt5OWbaVPVnUrfq1Awx8/N8dQesHCBPdm48Enp+o98y932zYdv/5NvPzIZGw83bzhIiOSOb5xcc5PSdBMr6urH6lZbFen2XZIZcv8UcIEF4ktV4BpqOlW+vHQsJd4q1m3J+7nSvI3fdL9x9zVStM7u0TBU/7rxC25kRLDtwg2y1RPtq/mx8qzVf9qtX7ItDoLcrP49qRmkPZ87eTWTM8mOkZ2koiGvDVGqJP4/dpv1XEczYfIGk9GxqBnvz86imLB/Z1PzBz+2j8NsQOfip9Rz0+Er/HD+t3oRW4+Tb/4yDc+vM1coSxdPFkaUvNqFSGQ/uPExj2OJDPEzNtHazbJboAdJA9AAJT5TsDFjUWV4aX74lDF8vL5k30KlbD+n14z4clQp2v9eeEA3DIpambZhBpZZYfew2X2+7SExiBgB1yvrwfvfqtNRnDpOeFcnP3Elg0IKDJGdk061WED8ObmhTWXg1/X2UCth9OY5pG89zIVqeyB3i48o7XavRu35ZlJZo/72LsLgrpD2A8A7w/O/gaGDmaEmCdW/AiZ/BwRkG/wmV2pqnvYYwopq9qdy6n0q/efuJScygYXlfVr7UHDdnB7Me01aIITAjiQBIeOLEX4X5bSEzSb95FloMnH+AQ5H3Gd06jA971jRxIw2jrSBn34Zl2XYuhksxyQCElnLj3a7VeKZuiH4XdwMrku+/EseIJUfIVKl5vml5vniutn6FP81M09+njKczZTxd8gIfL1dHXm9fmeEtK+LqZKEL5MObsKgrJN2VJ+gP+xtcijlXRZUNf46Qh3qdPeXgvmxDkzbXICaqZm8KF6OT6D9vP4np2bSv5s+CYY0Nr8tWAokhMEEQ8isd/qhI6u6v5NphxTC2bSUAfj18i8T0rCK2Np/copAFyzJEJ6Tz486rXIpJxsfNiY961uDfCW3pZUjPhoNzzjd0bR+PSvAuK28HtKxchu8G1UehgF8P3+SbbZeKf2Imou3vE5ecyYXoJByVCl56Kozd77ZnbNtwywU/KXFyzbqku/Ly9sGrih/8gNyT2ecnCGsDmcmwsh/EXTZdew1uj2GvHXOqFuTF4hFNcHVSsvPiPd778z/UatHf8TgRAAmCvajTDxqNACRYM0auIm+gdlUDqBLgSXJGNr8cumnyJupDpZaYuv6czmrkHi4O7JzQjpdaV8LF0cCLu0KR00Om1rKBulBNqu51gvmsd20Avt9xhaX7Ig07pgnp8/fx83Dm/R41KOVh/gtxnowkucp7/BXwKQdD14K7CZL2ObnCoF/kpImp8bC8tzwMZQ3FeO2YU+OKfszJGZZde+JOTiFYEQTlEgGQINiTbtMhoJbcRb9mNKgNm7irVCoY3UbuBVqyL5LMbG0f9OajT0HOlAwVF41JVBjeUb6gKgp8RCoc5PvDOxZ6yOBmFZjQuSoAU9af4++TxZtwbix9/j6xSRkcjrxvoRYBWelyGYuok+BeBob+Zdp5MC5e8hyg0lUg8bbcy5RipYSAQfXBI6Dw/Qql1teOOXWoHsiXfesCsGhvJPN2lZxFDOYmAiBBsCdObtB/KTi5Q+Qu2PuNwbvoVT+EAC8XYhIzrHKR17fYo1FFIRUKeQWdVCDAK6Ii+esdKjMiJ3P2hD9OsevSveK3oRjUaolNZ/QrWWKxopmqbFg9CiJ3g7MXDPkTylQ2/XE8ysi9St5lIe4S/NIfMpJNfxxt0h7Cjs9gdj3Nc4AkNfiEyr1UFta3USgf9awBwIzNF/jjyC2Lt8EWiQBIEOyNf1U53wrAjs/lytt3Txb+0bJk3sXRgRdbyZXhF+6xfHkMfYs9GlUU8loERGgpJJusfYWPQqHg46dr8ky9ELLVEq+sOMaJmw+K3w4D7L8Sx7M/7mX5Af1Kn1ikaKYkycvUL/wjz3t5/pdH9b3MwTdnaM3ND+4ck3NgZWeY73ggD+3tmgmz6sLumfJcpMA64Bcu9xg+7vx6ebt/P4FUC/bAAS+1rsTLbcMBmLTmP7adM3wI/EkjVoFpIFaBCU+87AyYXh6ydfQC6FitkpCWRctp/5KSqWLJiCa0r66hy99MsrLV1Jm6hfQszcNvCuTVYHsndijekvRru+CXgXL18bKN4c7RAgfIKdRZvYfWXWRmqxm17Ah7LsdRyt2JVS+3oHKAl+Ft0cOF6ESmb7pAxEW5t8nD2QGFQkFKRrbGeUBG/30MsW0y7JslD/8MWC73qlnC7WOw7BnISoGavaHfYlCaeKJ3ZiocWQh7Z0FaTjDjXwM6fAjVn4ar/8pznnK1/0gOBKNOyv938YYWr0PzV8DVMtcZSZJ478//WHXsNi6OSpaPbEqzSoZXvbdlYhWYIAi6OTjLFba10r1axcfNieeblgdg/u6rZmigZpIk8dmGczqDHzCioGXk7kfBT5WuMGLDox6LkAZQ93l5GGzVCHlbLZwdlcwb0oh65Xx5kJrFsEWHufswzfD26BCVkMa7q07R/bs9RFy8h6NSwYiWFdn9Xnu+6i/P+TBHwU+97ZstBz8gZ2u2VPADENoIBq0EpROc+ws2TChW8VSNstLh4Dz4rh5s+1gOfkpXhr6L4JV98nkqFI/mkYH8b5t3YEwEDFwpz8PLSISIL+C7urD3W8hMMU37dFAoFEzrU4dONQLJyFbz0vKjnLubaPbj2ioRAAmCPVIooOPHOjYoerXKyKfCcFQqOHjtPv/dfmjyJmry/Y4rLDtwA4UCRj0VRnCBTM5BPq7MHdKweAUnI/fIZRmy06BKFxj4s7zCKLciecfJ0OsH+du9KgN+fV4eZtHCw8WRJSOaUMnfg7sJ6QxbfJgHKcZn5U1Mz2Lmlgu0/yqCVcduI0nQs04w28e3ZcqztSjt6UK32sHMHdKwUKZro/4+hjix4lGpis6fQMOh5j2eJuHtoe9PgAKOLZHn5xgjO1Mu2vp9Q9g8UZ7n41sBes2BVw/Jqywf72VSKPK/dhSKnLllT8PLe+VeqdJV5GSQ26fIAdWBOXKAZUaODkp+eKEBTSqWIik9m+FLDnMzPtWsx7RVYghMAzEEJtgFSYKF7eX5Po8PligcILgujN5Z5HLdt38/ydoTd+hZN5gfXzBvAroVB2/w0V9nAJj6bC2Gt6xouoKT1/fJOWSyUqFyZ3mIy0nLHJmsdHmCbeRucCsFL26CgBpad33nYRp95+wnOjGd+uV8+WV0M9ydDc/EnZmt5pdDN5i94wr3cwKpJhVL8X6PGjQsX0rjY8xd8FOj8//AH0PlSb+t3pIDIGt6vNJ812nQ4lXDHq/Khv9+h10z4GHO/CrvsnKPTv0hhmewLrjv06tg13R4cF2+zytY3neDYcbtuwgJaVkMnC8X9a1Q2p0/n5CiviITtJFEACTYjSvb889TyDVkNVTuVOTDz91NpMfsPSgVEPFOe8qXdjdDI2Hj6She++U4kgRvdqzC+Jzl5iZxYz+s6CfPFwnvKOeU0Rb85MpIguW95B4gr2AYuRlKVdS6+eWYJPrNO0BCWhZtqvrz07DGODvq1wEvSRIbT0fz5ZYL3Mj5ph7u78Gk7jXoVCPAJrJO54ncI7+eVBnQYAg8+4PFct7otPsr2PGpfPu5+VBvUNGPUavh7BqImCbnLgJ5eXvrCXI+raJeI4ZQZcHJlfJk6sScHEY+5aHte1Dv+WKVrtFHbGI6fXKK+tYM9ua3sc3xLqFFfXOJAMhIIgAS7EZuL1DUf/LcFpB7gF45AAG65gg9MnTRIfZcjmN4iwpM7VXb5E3cdyWOF3NKTQxuVp7Pepuw1MSNA/IFOytFrkk16Ff9L2yp92FJD7h3Xg5+Rm4BryCtmx+78YAhPx0iLUtFr/ohfDugfpHZqQ9di+eLTRc4deshAGU8XRjfuSoDGofiaK2yBtpqXcVdgnVvQXaqPEzYf5nZLtwGkyTY8iEc/BFQQtfPoUJLzdt6lIE7x+XAJ/acfJ+bHzw1DpqMBmfzBPmAvDjh2DLY8xUk56zS8qsEbSfJQ2xJUSavMxYZl0L/efuJS86keSU/lr7Y1HKZwc1ABEBGEgGQYFc09QL5lINR28C76Lkiey/HMWTRIdycHNg/qYNJswufvp3AoAUHSMlU0aNOEN8/b8JiozcPyuedmQyV2sPzv8p5kgyRGAVLusnDFwE15UnTOrIb77wYy+hlR8lWS4xoWZGPetbgyPUHhYaorsQmMX3TRbafly+C7s4OjG0Tzkutw/Bw0SOoMFdBTn1qXSmd4N2r4OZj+P7NSa2GtWPh9B+6t1M6gjpbvu3iAy3fgOYvy8kWLSUzFY4ukidH5+YNKl1FDooydExaLmadsdyivl4ZMfSs7MjErjU4H5XI/dRM/NydqVXWGweFwiKFXI0lAiAjiQBIsCt5c4FOQFBdeTXK/asQWBte3Aiuui9kkiTRc/ZezkUlMr5zVd7sWMUkzYqMS6Hf3P3Ep2TSMrw0S15sYnhZC21uHoIVfXKCn3bw/G+GBz+57kfC4m6QHK1Xcc+/Ttxh3O8nAfB0cSQ5IzvvdwFeLlQN9GL/1TjUEjgoFTzftBxvdqyif94ecxbkzHutnEJruYegejB2l20MfRWUnQkzK0NGgu7tnDzkuUItXpPneVlLRjIcmgf7Z0N6EW1GCSH19Jq7p8mBS3eosrIFZRQ6jmOhQq7GEMvgBUHQ3+OrVTp/AkPXgGcgxJyB3wYXmUhOoVDkFUldtv866VmGldfQJCYxnaGLDhGfkkmdsj4sGNbYdMHPrcOPen7C2uQMexUz+AHwC5OT77n6yjmDfntB59+sd4Oy9G8UCpAv+AG5RMXeK3Lw06VmIFvGteGz3nUMS1po6oKcajVkpclDfknR0OQltNe6Ajp9bJvBD8iTip+bp3ubGr1g3Gl5FaQ1gx+QA+k278Bb/0HbieCo63VqXJ2xFlVCyPQMQS1pfrxaUvDQKcAihVwtxeo9QHPmzGHmzJlERUVRq1YtZs2aRevWrbVun5GRwSeffMKKFSuIjo4mNDSUDz/8kJEjRwKwdOlSXnzxxUKPS0tLw9VVvw8R0QMk2L2oU7CkJ2QmQa3noO9iUGr/vpSlUtNuZgR3Hqbx+XO1GdysQrEPnZCaxYD5B7gYk0RYGQ9WvdyCMp4m+sZ564hcJyozCSq2hhf+MN2cjttHYdmz8nwiHXNgVGqJp2bs0Fmvq7SHM4c/7FT84T5tk9tzhbUFN185UMtKk//NTpNXuGU/9pOVLk9o1odCCcHF74GwGEmC+W0g+r8Cv1DIvZ4v77Hd9qfEw9wWj+YHFeTgIg/BuvnJwZubb87/S+X8PHY7734/cHJFpZaYMO1rZmV9qvXw45z+x9fvTzD/SkIjGHL9tuoMtd9//51x48YxZ84cWrVqxfz58+nevTvnzp2jfPnyGh8zYMAAYmJiWLRoEZUrVyY2Npbs7Pzfory9vbl48WK++/QNfgRBQL6QDfwZVvaHs2vBMwi6TdN6YXByUDLyqTA+/eccP+2JZFCT8sX6kEzLVDFq2REuxiQR4OXC8pFNTRf83D6aM+yVG/z8btoJraGN5VIPK/vLGX/XvymvgioQOOpTrDQ+JZPDkfdpEV7MLL1lm8gTspOiNf8+clfx9qtwkHvLFAp5JdzjJMtWOi82hQI6TdYQIErQeYptt9+jNPSeoz24VWXIE6WT9KsHl8fRjWxnH8ZmOJGscMWddB5/+2ZLSs5KFfkrqToDjXld2hirBkDffPMNo0aN4qWXXgJg1qxZbNmyhblz5zJt2rRC22/evJldu3Zx7do1/PzkiYYVK1YstJ1CoSAoSPtqDEEQ9BDeXh4uWD0KDs2VJ0S3ekvr5oOalOO77ZeIjEth27kYutU27D2YpVLz+i/HOXrjAd6ujiwf1ZRyfqbqnTkm9/xkJEKFVjnBj4dp9v24Su2g3xL4Y5i8rNnVB7p+ke+iatZirllpcHihPHk2tzxDQXUHQWAtOZBxdJGHVRxdcv7vKv84uT52+7Htcnu0Cq4ezM0dZeFK58WWm6W5JLZfW9uH/g3pD+TEimkP5CHLtAdykda0+xruz/mRVJCdhkt2GjW0dPI6KtR8ndUfUFiuiK4FWC0AyszM5NixY0yaNCnf/V26dGH//v0aH7Nu3ToaN27Ml19+yc8//4yHhwfPPvssn376KW5uj8ZGk5OTqVChAiqVivr16/Ppp5/SoIH2AnwZGRlkZDzq5k1MtN/U4IKQT+7S260fyWn/PYOg3kCNm3q4ODKkeQXmRFxl4Z5rBgVAkiQxafVp/r0Qi4ujkkUjmlA9yETDz3eOPwp+yrfMGfYyQ/CTq8bT0OtH+OtlODhHnhvUbmLer81SzFWVBceXy8U4c7/9+1WWL24Pb+a/UD43z/heDoVC7u3J7YmQVCWj9ydXSW6/tra7+cg/OvJRFSJJ8vsi7QH/XY5k5l8H8SWJiY6/EaKIz+sFipF82a2uA8iT+JtXKk2gd8kfVbHaJOi4uDhUKhWBgYH57g8MDCQ6WnO37bVr19i7dy9nzpxh7dq1zJo1iz///JPXXnstb5vq1auzdOlS1q1bx6+//oqrqyutWrXi8uXLWtsybdo0fHx88n7KlStnmpMUhCdByzfkoo0Af78KV/7VuumIlhVxdlBy7MYDjt3Qv9r19E0XWH38Ng5KBT++0JAmFbUvJTfI3RPwc2951U/5FjB4lc4VWiZT/3nollNNPuILuXZUjqZhfgT7uBaq05VLAQT7yEvii6RWwanf4YfGsGG8HPz4lJPLM7x2CHp+9Si/k6kv8gVrXZWE3pPHleT2m6rtCoXcS1mqIrUat+OKV1P+Ubfig+yX8g2BBSoe8p7jH4DEzov3aDtzJ19vvUhSepbRp2JNVl8FVjChmSRJWpOcqdVqFAoFK1eupGnTpvTo0YNvvvmGpUuXkpYmFxps3rw5Q4YMoV69erRu3Zo//viDqlWr8v3332ttw/vvv09CQkLez61bt0x3goLwJOj8KdTuJ+dH+WNYTvmMwgK8XXmugZwnZP6ua3rtev6uq8zfLW87o29dOtUMLOIRerp7Us7WnJ4A5ZpbLvjJ1fxlaPe+fHvzRDj5KyAvbZ/8TE3AiGKlkgQXNsC8p2DtGDkPkUcAdJ8JbxyDBoPl4SpzXuQ11boqSUpy+83Q9sdfl3vUdTmllld23lHL831edVzHmsqbaVTel/QsNd/vuEK7mREsP3CdLJWOVYE2zGoBUJkyZXBwcCjU2xMbG1uoVyhXcHAwZcuWxcfnUV6SGjVqIEkSt2/f1vgYpVJJkyZNdPYAubi44O3tne9HEITHKJXy5MuwtvLy8ZX94L7mAGd0mzAAtp2P4dq9ZJ27XXX0FtM2XQDggx7V6ZezPNxoUaceC36awZA/LZvILlfbidDsFfn236/JQQsYV6z0WgT81FFebh97Tv4G3/FjeOskNBuTP0eLuS/y4e3h9cPyvyVRSW6/Gdr+6HXpxpfZA7msLst72WP4ynE0AA1v/8yflTYwb3BDwsp4EJ+Sycd/n6XLt7vZdDqKkpZW0KrL4Js1a0ajRo2YM2dO3n01a9akV69eGidBL1iwgHHjxhEbG4unp/xN7u+//6ZPnz4kJyfnmweUS5IkmjZtSp06dVi8eLFe7RLL4AVBi/REWNoDok/LKfpHbgVP/0KbvbTsCNvPx/J80/JM61NH4662n4th7IpjqNQSY9tU4v0e2guKFqIr03HcZdg4QQ5+QpvKdc1crfg+Vqth3evypGgHZxj8J1RqCxhYrPTWEdjxiVyEFcDJHZq/Ig9RWjtfjfBE0fi6PLZYHmYFaPYKWZ0/57cjt5i1/TLxOcV5G5b35YMeNWhsqiHsYigxmaB///13hg4dyrx582jRogULFixg4cKFnD17lgoVKvD+++9z584dli9fDsiTm2vUqEHz5s2ZOnUqcXFxvPTSS7Rt25aFCxcCMHXqVJo3b06VKlVITExk9uzZ/Pzzz+zbt4+mTZvq1S4RAAmCDknRsKizPLk2pCEMX19oaOlw5H0GzD+As6OSfRM7FKoyfeT6fYb8dIiMbDX9GoUys19d/et76ZPpGKBsIxj6l3WDn1yqbFg1XF4e7+Qh/81CG+n32JizsOMzuLhR/r+DMzQeKRfl9AwwX5sFoaCjS+CfcfLtpmOg+5ckZ6pYsOsqC/dEkpaTBLVrrUDe61adcH8LDjnnKDF5gAYOHEh8fDyffPIJUVFR1K5dm40bN1KhgpxELSoqips3b+Zt7+npybZt23jjjTdo3LgxpUuXZsCAAXz22Wd52zx8+JAxY8YQHR2Nj48PDRo0YPfu3XoHP4IgFMErCIasgUVd4O5xWDVCrqPl8KiKdJOKpahfzpeTtx6y/MB1JnR5VFj1fFQiI5ceISNbTcfqAUzvU8ew4qa5mY5T4tCakdjJXW6jLQQ/IM/H6bsIfhkg5+D5+Tl45js5i7QmHv5yIsKIaXD6T0CSEw3We0FeUearOU+aIJhV4xdB6QDr3oTDC0CtwrPHV4zvUo3BzSswa/slfj9yiy1nY3J6gMvxVseqhb4A2QqrZ4K2RaIHSBD0cPsoLH1aziBc7wV5jtBjgcym01G8svI4Pm6OfDeoAQlpWSiAT/85x73kTJpULMXykc1wcy5GiYuiMh33Xw61ehm+X3PLSIZlz8iBoy6OrnLdqtwAr2ZvaP8h+Fc1dwsFoWgnVspz2pCg0Qjo+W1ews/LMUnM2HyB7eflHlpNhXwNGvo1UIkZArNVIgASBD1d3CxPxpVU8pBMx4/zfqVSSzT/Yjv3kjMLPaysrysb32yDj7tTod/ppWAivscF1YWxu213VU9KPHxbU+7hKUrlzvLS9ZD6Zm+WIBjk1G/w1ytyBvCGw+Dp7/JlPT94LZ5pG89z6rZcXNXfy4W3O1XFx82Rzzacz5cNPdjHlcnP1NQ9+V9PohiqIAiWUa0bPDNLvr3nazkLcY5t56I1Bj8Adx6mc+CalknMRZEkuHNMLthaMPgBucyBrQY/IJcz6PWD7m38a8CLm+TVayL4EWxRvUHw3AJ5aPb4clj3hpyXKkfzSqX567VW/PBCA8r7uXMvKYMP1p7mtV9OFCoFE52QzisrjrP5jIElPIwkAiBBEIzTcJg8PAOw8V049zcqtcTU9ee0PkQBTF1/DpXagA7o+9cgYgZ830heBn5pc4GdKktOQrva/eRyFJr4hcMr+6FCS8u2SRAMVbc/9FkoZxk/uUIeFnssCFIoFDxdN4Tt49vyv6draP1ekvspYPBngpFEACQIgvHavAuNXgQkWD2a8wc36yz4KQFRCekcjiwiW3TqfTjyE/zUGWY3kLMq378qT3KuM0AeHsrbaQkpxglyGzt/ovl3Pb4sVEBVEGxWnX7Qb5EcBJ36Fda+LK96fIyzo5KawT7omnCj92eCCVl1FZggCE8IhQJ6fg0p9+DCP1TdOYaqig+5JOkuK6OxsGJWuty7898fcHkrqHPS7SuUcrHRugOh+tPy0vvcjMh3T5Sc3p9cJbkgpyA8rtZz8vvzz5Fw+g/59fzcgkfFczFzEeBiEgGQIAimoXSAvj/B8t443zrIz85f8F7mWOLw0bh5vOT9qOCnWg03D8B/v8HZv+XaXbmC6spBT51+8hL8x+VmOt40sWSWMyipBTkFoaCavaD/UjktxpnVco9sn4V56THMUgTYSGIVmAZiFZggGCH1PtKiLijitZefAYjHF98x63E4/xf8twoSHuX8wjtUnl9QdyAEGJAhuqTJXc2W24M1eqcIgISS7cJGuV6gOgtqPAv9FoODEyq1xFMzdhCdkI6moEOBXApm78QORi2JF8vgjSQCIEEw0oObqGc3QClla/y1WgK1oyuOqse6u1285W+RdQdChVb2Mw/m6k65B6v7jJJZk0oQCrq0BX4fAqpMebi63xJwdGbzmSheWSHnwHo88MgNd4qsg6cHEQAZSQRAgmACRxY9qh2kjdJRznVTbyBU7QZOhev5CYJQAl3eBr8NBlUGVOshD485urD5TBRT15+ziTxAIgDSQARAgmACkgQ/NEGKv0yhDm0nd+g0FWr3lfPiCILw5LmyXQ6CstOhSlcY+DM4uohM0LZMBECCYCLaSlYMWQ2VO1m+PYIgWNbVnfDrIDkIqtBazhbv6Kx5Ww9/uc6fEUQAZCQRAAmCiRQsWZG71FtM9hUE+3FtF6wcAKoilrh7BsC4M+BY/OKpohSGIAi2IXepd27JCrHUWxDsT6W2clmXwoPhj1GCd1lw0NI7ZAYiABIEwbxyE/5ByUtWKAiCaYS1hi6f6tjA8pncRQAkCIJ55SYrLFOt5CUrFATBdFq8DmWqFr5f4WCVL0ciABIEwfzC28Prh0WeG0GwZwoFdJtW+H4rDY2LAEgQBEEQBMvIHRJX5IQfVur9AREACYIgCIJgKXkLI9Ty/624MEIEQIIgCIIgWI6NLIwQAZAgCIIgCJZjIwsjHK1yVEEQBEEQ7FfuwggrEj1AgiAIgiDYHREACYIgCIJgd0QAJAiCIAiC3REBkCAIgiAIdkcEQIIgCIIg2B0RAAmCIAiCYHdEACQIgiAIgt0RAZAgCIIgCHZHBECCIAiCINgdEQAJgiAIgmB3RCkMDSRJAiAxMdHKLREEQRAEQV+51+3c67guIgDSICkpCYBy5cpZuSWCIAiCIBgqKSkJHx8fndsoJH3CJDujVqu5e/cuXl5eKKxUpdYSEhMTKVeuHLdu3cLb29vazTE7ezpfca5PLns6X3GuTy5zna8kSSQlJRESEoJSqXuWj+gB0kCpVBIaGmrtZliMt7e3XbzhctnT+YpzfXLZ0/mKc31ymeN8i+r5ySUmQQuCIAiCYHdEACQIgiAIgt0RAZAdc3FxYfLkybi4uFi7KRZhT+crzvXJZU/nK871yWUL5ysmQQuCIAiCYHdED5AgCIIgCHZHBECCIAiCINgdEQAJgiAIgmB3RAAkCIIgCILdEQHQE2ratGk0adIELy8vAgIC6N27NxcvXtT5mIiICBQKRaGfCxcuWKjVxTdlypRC7Q4KCtL5mF27dtGoUSNcXV2pVKkS8+bNs1BrjVOxYkWNz9Nrr72mcfuS9Lzu3r2bZ555hpCQEBQKBX/99Ve+30uSxJQpUwgJCcHNzY127dpx9uzZIve7evVqatasiYuLCzVr1mTt2rVmOgPD6DrfrKwsJk6cSJ06dfDw8CAkJIRhw4Zx9+5dnftcunSpxuc7PT3dzGejW1HP7YgRIwq1uXnz5kXu1xaf26LOVdPzo1AomDlzptZ92urzqs+1xlbftyIAekLt2rWL1157jYMHD7Jt2zays7Pp0qULKSkpRT724sWLREVF5f1UqVLFAi02Xq1atfK1+/Tp01q3jYyMpEePHrRu3ZoTJ07wwQcf8Oabb7J69WoLtrh4jhw5ku88t23bBkD//v11Pq4kPK8pKSnUq1ePH374QePvv/zyS7755ht++OEHjhw5QlBQEJ07d86r36fJgQMHGDhwIEOHDuXUqVMMHTqUAQMGcOjQIXOdht50nW9qairHjx/nf//7H8ePH2fNmjVcunSJZ599tsj9ent753uuo6KicHV1Nccp6K2o5xagW7du+dq8ceNGnfu01ee2qHMt+NwsXrwYhUJB3759de7XFp9Xfa41Nvu+lQS7EBsbKwHSrl27tG6zc+dOCZAePHhguYaZyOTJk6V69erpvf17770nVa9ePd99Y8eOlZo3b27ilpnfW2+9JYWHh0tqtVrj70vq8wpIa9euzfu/Wq2WgoKCpOnTp+fdl56eLvn4+Ejz5s3Tup8BAwZI3bp1y3df165dpUGDBpm8zcYoeL6aHD58WAKkGzduaN1myZIlko+Pj2kbZ2KaznX48OFSr169DNpPSXhu9Xlee/XqJXXo0EHnNiXheZWkwtcaW37fih4gO5GQkACAn59fkds2aNCA4OBgOnbsyM6dO83dNJO5fPkyISEhhIWFMWjQIK5du6Z12wMHDtClS5d893Xt2pWjR4+SlZVl7qaaTGZmJitWrGDkyJFFFu4tqc9rrsjISKKjo/M9by4uLrRt25b9+/drfZy251rXY2xVQkICCoUCX19fndslJydToUIFQkNDefrppzlx4oRlGmikiIgIAgICqFq1KqNHjyY2Nlbn9k/CcxsTE8OGDRsYNWpUkduWhOe14LXGlt+3IgCyA5IkMX78eJ566ilq166tdbvg4GAWLFjA6tWrWbNmDdWqVaNjx47s3r3bgq0tnmbNmrF8+XK2bNnCwoULiY6OpmXLlsTHx2vcPjo6msDAwHz3BQYGkp2dTVxcnCWabBJ//fUXDx8+ZMSIEVq3KcnP6+Oio6MBND5vub/T9jhDH2OL0tPTmTRpEi+88ILO4pHVq1dn6dKlrFu3jl9//RVXV1datWrF5cuXLdhaw3Xv3p2VK1eyY8cOvv76a44cOUKHDh3IyMjQ+pgn4bldtmwZXl5e9OnTR+d2JeF51XStseX3ragGbwdef/11/vvvP/bu3atzu2rVqlGtWrW8/7do0YJbt27x1Vdf0aZNG3M30yjdu3fPu12nTh1atGhBeHg4y5YtY/z48RofU7DHRMpJil5UT4otWbRoEd27dyckJETrNiX5edVE0/NW1HNWnMfYkqysLAYNGoRarWbOnDk6t23evHm+ycOtWrWiYcOGfP/998yePdvcTS22gQMH5t2uXbs2jRs3pkKFCmzYsEFncFDSn9vFixczePDgIufylITnVde1xhbft6IH6An3xhtvsG7dOnbu3EloaKjBj2/evLlNfcPQl4eHB3Xq1NHa9qCgoELfJGJjY3F0dKR06dKWaKLRbty4wfbt23nppZcMfmxJfF5zV/Vpet4KflMs+DhDH2NLsrKyGDBgAJGRkWzbtk1n748mSqWSJk2alLjnOzg4mAoVKuhsd0l/bvfs2cPFixeL9R62tedV27XGlt+3IgB6QkmSxOuvv86aNWvYsWMHYWFhxdrPiRMnCA4ONnHrzC8jI4Pz589rbXuLFi3yVk/l2rp1K40bN8bJyckSTTTakiVLCAgIoGfPngY/tiQ+r2FhYQQFBeV73jIzM9m1axctW7bU+jhtz7Wux9iK3ODn8uXLbN++vVjBuSRJnDx5ssQ93/Hx8dy6dUtnu0vycwtyD26jRo2oV6+ewY+1lee1qGuNTb9vTTadWrApr7zyiuTj4yNFRERIUVFReT+pqal520yaNEkaOnRo3v+//fZbae3atdKlS5ekM2fOSJMmTZIAafXq1dY4BYNMmDBBioiIkK5duyYdPHhQevrppyUvLy/p+vXrkiQVPtdr165J7u7u0ttvvy2dO3dOWrRokeTk5CT9+eef1joFg6hUKql8+fLSxIkTC/2uJD+vSUlJ0okTJ6QTJ05IgPTNN99IJ06cyFv1NH36dMnHx0das2aNdPr0aen555+XgoODpcTExLx9DB06VJo0aVLe//ft2yc5ODhI06dPl86fPy9Nnz5dcnR0lA4ePGjx8ytI1/lmZWVJzz77rBQaGiqdPHky3/s4IyMjbx8Fz3fKlCnS5s2bpatXr0onTpyQXnzxRcnR0VE6dOiQNU4xj65zTUpKkiZMmCDt379fioyMlHbu3Cm1aNFCKlu2bIl8bot6HUuSJCUkJEju7u7S3LlzNe6jpDyv+lxrbPV9KwKgJxSg8WfJkiV52wwfPlxq27Zt3v9nzJghhYeHS66urlKpUqWkp556StqwYYPlG18MAwcOlIKDgyUnJycpJCRE6tOnj3T27Nm83xc8V0mSpIiICKlBgwaSs7OzVLFiRa0fRLZoy5YtEiBdvHix0O9K8vOau2S/4M/w4cMlSZKX1E6ePFkKCgqSXFxcpDZt2kinT5/Ot4+2bdvmbZ9r1apVUrVq1SQnJyepevXqNhP86TrfyMhIre/jnTt35u2j4PmOGzdOKl++vOTs7Cz5+/tLXbp0kfbv32/5kytA17mmpqZKXbp0kfz9/SUnJyepfPny0vDhw6WbN2/m20dJeW6Leh1LkiTNnz9fcnNzkx4+fKhxHyXledXnWmOr71tFzgkIgiAIgiDYDTEHSBAEQRAEuyMCIEEQBEEQ7I4IgARBEARBsDsiABIEQRAEwe6IAEgQBEEQBLsjAiBBEARBEOyOCIAEQRAEQbA7IgASBEEQBMHuiABIEARBEAS7IwIgQRAEQRDsjgiABEEQBEGwOyIAEgTBbqSkpDBs2DA8PT0JDg7m66+/pl27dowbN87aTRMEwcJEACQIgt1499132blzJ2vXrmXr1q1ERERw7NgxazdLEAQrcLR2AwRBECwhOTmZRYsWsXz5cjp37gzAsmXLCA0NtXLLBEGwBtEDJAiCXbh69SqZmZm0aNEi7z4/Pz+qVatmxVYJgmAtIgASBMEuSJJk7SYIgmBDRAAkCIJdqFy5Mk5OThw8eDDvvgcPHnDp0iUrtkoQBGsRc4AEQbALnp6ejBo1infffZfSpUsTGBjIhx9+iFIpvgcKgj0SAZAgCHZj5syZJCcn8+yzz+Ll5cWECRNISEiwdrMEQbAChSQGxgVBsGPt2rWjfv36zJo1y9pNEQTBgkTfryAIgiAIdkcEQIIgCIIg2B0xBCYIgiAIgt0RPUCCIAiCINgdEQAJgiAIgmB3RAAkCIIgCILdEQGQIAiCIAh2RwRAgiAIgiDYHREACYIgCIJgd0QAJAiCIAiC3REBkCAIgiAIduf/vhdMInzWPIYAAAAASUVORK5CYII=",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#plot the accuracy, precision, recall and f1-score against the value of p\n",
    "import matplotlib.pyplot as plt\n",
    "plt.plot(result['q'], result['Accuracy'], label='Accuracy', marker = 'o')\n",
    "plt.plot(result['q'], result['Precision'], label='Precision',marker = 'v')\n",
    "plt.plot(result['q'], result['Recall'], label='Recall', marker = 's')\n",
    "plt.plot(result['q'], result['F1-Score'], label='F1-Score',marker = 'd')\n",
    "plt.xlabel('q')\n",
    "plt.ylabel('Value')\n",
    "plt.legend()\n",
    "plt.show()\n"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Summary\n",
    "<p> The graph above displays the score on evaluation metrics based on the power of q from 1-20. We can clearly see that from the graph that as the amount of q increases the metrics rate decreases. This shows that the optimum power of q for the given task should be 1 which means the Minkowski Distance should be measured in Manhattan. The best accuracy throughout was 0.927536 which was achieved when power of q was 1. This pattern can also be observed with other metrics. </p>"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "DM",
   "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": "3.9.16"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}