{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Цель вычислительного эксперимента:\n",
    "Цель эксперимента заключается в том, чтобы решить задачу классификации. Задача решается на выборке \"Lung Cancer Data Set \" с использованием моделей: kNN, SVM, логистическая регрессия. Структурные параметры: число и состав признаков. Используемый критерий качества AUC, F1, число признаков.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Описание выборки\n",
    "Используются данные из Lung Cancer Data Set (https://archive.ics.uci.edu/ml/datasets/Lung+Cancer). В выборке представлены медицинские описания онкобольных с 3 видами патологического рака лёгких. Первый столбец это метка предсказываемого класса (3 вида рака: 1-3), остальные 56 столбцов это целочисленные атрибуты, принимающие значения 0-3. Точного описания значения столбцов нет, поэтому будем отталкиваться от предположения, что это в целом описание здоровья больного, его персональные данные итд. \n",
    "Число объектов в выборке: 32.\n",
    "Число признаков: 56 признаков + признак класса.\n",
    "(32 примера очень малый объем выборки, уччитывыавя в том числе количество признаком по каждому обьекту)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Блок загрузки и предобработки выборок\n",
    "* Импортируем библиотеки\n",
    "* Выгружаем данные по адресу \"*/lung-cancer.data.txt\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "      <th>...</th>\n",
       "      <th>47</th>\n",
       "      <th>48</th>\n",
       "      <th>49</th>\n",
       "      <th>50</th>\n",
       "      <th>51</th>\n",
       "      <th>52</th>\n",
       "      <th>53</th>\n",
       "      <th>54</th>\n",
       "      <th>55</th>\n",
       "      <th>56</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>?</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>...</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 57 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   0   1   2   3  4   5   6   7   8   9  ...  47  48  49  50  51  52  53  54  \\\n",
       "0   1   0   3   0  ?   0   2   2   2   1 ...   2   2   2   2   2   1   1   1   \n",
       "1   1   0   3   3  1   0   3   1   3   1 ...   2   2   2   2   2   2   2   1   \n",
       "2   1   0   3   3  2   0   3   3   3   1 ...   2   2   2   2   2   2   2   2   \n",
       "3   1   0   2   3  2   1   3   3   3   1 ...   2   2   2   2   2   2   2   2   \n",
       "4   1   0   3   2  1   1   3   3   3   2 ...   2   2   2   2   2   2   2   1   \n",
       "\n",
       "   55  56  \n",
       "0   2   2  \n",
       "1   2   2  \n",
       "2   1   2  \n",
       "3   2   2  \n",
       "4   2   2  \n",
       "\n",
       "[5 rows x 57 columns]"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%matplotlib inline \n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from scipy.stats import norm\n",
    "import seaborn as sns\n",
    "import warnings\n",
    "warnings.filterwarnings('ignore')\n",
    "\n",
    "# Read data\n",
    "data = pd.read_csv('lung-cancer.csv', header=None)\n",
    "classes = data[0]\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 32 entries, 0 to 31\n",
      "Data columns (total 57 columns):\n",
      "0     32 non-null int64\n",
      "1     32 non-null int64\n",
      "2     32 non-null int64\n",
      "3     32 non-null int64\n",
      "4     32 non-null object\n",
      "5     32 non-null int64\n",
      "6     32 non-null int64\n",
      "7     32 non-null int64\n",
      "8     32 non-null int64\n",
      "9     32 non-null int64\n",
      "10    32 non-null int64\n",
      "11    32 non-null int64\n",
      "12    32 non-null int64\n",
      "13    32 non-null int64\n",
      "14    32 non-null int64\n",
      "15    32 non-null int64\n",
      "16    32 non-null int64\n",
      "17    32 non-null int64\n",
      "18    32 non-null int64\n",
      "19    32 non-null int64\n",
      "20    32 non-null int64\n",
      "21    32 non-null int64\n",
      "22    32 non-null int64\n",
      "23    32 non-null int64\n",
      "24    32 non-null int64\n",
      "25    32 non-null int64\n",
      "26    32 non-null int64\n",
      "27    32 non-null int64\n",
      "28    32 non-null int64\n",
      "29    32 non-null int64\n",
      "30    32 non-null int64\n",
      "31    32 non-null int64\n",
      "32    32 non-null int64\n",
      "33    32 non-null int64\n",
      "34    32 non-null int64\n",
      "35    32 non-null int64\n",
      "36    32 non-null int64\n",
      "37    32 non-null int64\n",
      "38    32 non-null object\n",
      "39    32 non-null int64\n",
      "40    32 non-null int64\n",
      "41    32 non-null int64\n",
      "42    32 non-null int64\n",
      "43    32 non-null int64\n",
      "44    32 non-null int64\n",
      "45    32 non-null int64\n",
      "46    32 non-null int64\n",
      "47    32 non-null int64\n",
      "48    32 non-null int64\n",
      "49    32 non-null int64\n",
      "50    32 non-null int64\n",
      "51    32 non-null int64\n",
      "52    32 non-null int64\n",
      "53    32 non-null int64\n",
      "54    32 non-null int64\n",
      "55    32 non-null int64\n",
      "56    32 non-null int64\n",
      "dtypes: int64(55), object(2)\n",
      "memory usage: 14.3+ KB\n",
      "None (32, 57)\n"
     ]
    }
   ],
   "source": [
    "print(data.info(), data.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    1\n",
       "1    1\n",
       "2    1\n",
       "3    1\n",
       "4    1\n",
       "Name: 0, dtype: int64"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Y = data[0]\n",
    "Y.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "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>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "      <th>10</th>\n",
       "      <th>...</th>\n",
       "      <th>47</th>\n",
       "      <th>48</th>\n",
       "      <th>49</th>\n",
       "      <th>50</th>\n",
       "      <th>51</th>\n",
       "      <th>52</th>\n",
       "      <th>53</th>\n",
       "      <th>54</th>\n",
       "      <th>55</th>\n",
       "      <th>56</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>?</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>...</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>...</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 56 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   1   2   3  4   5   6   7   8   9   10 ...  47  48  49  50  51  52  53  54  \\\n",
       "0   0   3   0  ?   0   2   2   2   1   1 ...   2   2   2   2   2   1   1   1   \n",
       "1   0   3   3  1   0   3   1   3   1   1 ...   2   2   2   2   2   2   2   1   \n",
       "2   0   3   3  2   0   3   3   3   1   1 ...   2   2   2   2   2   2   2   2   \n",
       "3   0   2   3  2   1   3   3   3   1   2 ...   2   2   2   2   2   2   2   2   \n",
       "4   0   3   2  1   1   3   3   3   2   2 ...   2   2   2   2   2   2   2   1   \n",
       "\n",
       "   55  56  \n",
       "0   2   2  \n",
       "1   2   2  \n",
       "2   1   2  \n",
       "3   2   2  \n",
       "4   2   2  \n",
       "\n",
       "[5 rows x 56 columns]"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X = data.drop(0, axis=1)\n",
    "X.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Анализ состава выборки\n",
    "* Анализ состава выборки\n",
    "* Анализ пропусков, статистика"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "      <th>10</th>\n",
       "      <th>...</th>\n",
       "      <th>47</th>\n",
       "      <th>48</th>\n",
       "      <th>49</th>\n",
       "      <th>50</th>\n",
       "      <th>51</th>\n",
       "      <th>52</th>\n",
       "      <th>53</th>\n",
       "      <th>54</th>\n",
       "      <th>55</th>\n",
       "      <th>56</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>32.000000</td>\n",
       "      <td>32.000000</td>\n",
       "      <td>32.000000</td>\n",
       "      <td>32.000000</td>\n",
       "      <td>32.000000</td>\n",
       "      <td>32.000000</td>\n",
       "      <td>32.000000</td>\n",
       "      <td>32.000000</td>\n",
       "      <td>32.000000</td>\n",
       "      <td>32.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>32.000000</td>\n",
       "      <td>32.000000</td>\n",
       "      <td>32.000000</td>\n",
       "      <td>32.000000</td>\n",
       "      <td>32.000000</td>\n",
       "      <td>32.000000</td>\n",
       "      <td>32.000000</td>\n",
       "      <td>32.000000</td>\n",
       "      <td>32.000000</td>\n",
       "      <td>32.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>2.031250</td>\n",
       "      <td>0.031250</td>\n",
       "      <td>2.375000</td>\n",
       "      <td>2.031250</td>\n",
       "      <td>0.281250</td>\n",
       "      <td>2.187500</td>\n",
       "      <td>2.125000</td>\n",
       "      <td>2.406250</td>\n",
       "      <td>1.156250</td>\n",
       "      <td>1.406250</td>\n",
       "      <td>...</td>\n",
       "      <td>2.062500</td>\n",
       "      <td>2.062500</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>1.843750</td>\n",
       "      <td>1.843750</td>\n",
       "      <td>1.562500</td>\n",
       "      <td>1.812500</td>\n",
       "      <td>1.718750</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.782237</td>\n",
       "      <td>0.176777</td>\n",
       "      <td>0.553581</td>\n",
       "      <td>1.031265</td>\n",
       "      <td>0.456803</td>\n",
       "      <td>0.737804</td>\n",
       "      <td>0.751343</td>\n",
       "      <td>0.756024</td>\n",
       "      <td>0.514899</td>\n",
       "      <td>0.559918</td>\n",
       "      <td>...</td>\n",
       "      <td>0.245935</td>\n",
       "      <td>0.245935</td>\n",
       "      <td>0.359211</td>\n",
       "      <td>0.359211</td>\n",
       "      <td>0.508001</td>\n",
       "      <td>0.447889</td>\n",
       "      <td>0.447889</td>\n",
       "      <td>0.504016</td>\n",
       "      <td>0.396558</td>\n",
       "      <td>0.456803</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>1.750000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>3.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>3.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>8 rows × 55 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "              0          1          2          3          5          6   \\\n",
       "count  32.000000  32.000000  32.000000  32.000000  32.000000  32.000000   \n",
       "mean    2.031250   0.031250   2.375000   2.031250   0.281250   2.187500   \n",
       "std     0.782237   0.176777   0.553581   1.031265   0.456803   0.737804   \n",
       "min     1.000000   0.000000   1.000000   0.000000   0.000000   1.000000   \n",
       "25%     1.000000   0.000000   2.000000   1.750000   0.000000   2.000000   \n",
       "50%     2.000000   0.000000   2.000000   2.000000   0.000000   2.000000   \n",
       "75%     3.000000   0.000000   3.000000   3.000000   1.000000   3.000000   \n",
       "max     3.000000   1.000000   3.000000   3.000000   1.000000   3.000000   \n",
       "\n",
       "              7          8          9          10    ...             47  \\\n",
       "count  32.000000  32.000000  32.000000  32.000000    ...      32.000000   \n",
       "mean    2.125000   2.406250   1.156250   1.406250    ...       2.062500   \n",
       "std     0.751343   0.756024   0.514899   0.559918    ...       0.245935   \n",
       "min     1.000000   1.000000   1.000000   1.000000    ...       2.000000   \n",
       "25%     2.000000   2.000000   1.000000   1.000000    ...       2.000000   \n",
       "50%     2.000000   3.000000   1.000000   1.000000    ...       2.000000   \n",
       "75%     3.000000   3.000000   1.000000   2.000000    ...       2.000000   \n",
       "max     3.000000   3.000000   3.000000   3.000000    ...       3.000000   \n",
       "\n",
       "              48         49         50         51         52         53  \\\n",
       "count  32.000000  32.000000  32.000000  32.000000  32.000000  32.000000   \n",
       "mean    2.062500   2.000000   2.000000   2.000000   1.843750   1.843750   \n",
       "std     0.245935   0.359211   0.359211   0.508001   0.447889   0.447889   \n",
       "min     2.000000   1.000000   1.000000   1.000000   1.000000   1.000000   \n",
       "25%     2.000000   2.000000   2.000000   2.000000   2.000000   2.000000   \n",
       "50%     2.000000   2.000000   2.000000   2.000000   2.000000   2.000000   \n",
       "75%     2.000000   2.000000   2.000000   2.000000   2.000000   2.000000   \n",
       "max     3.000000   3.000000   3.000000   3.000000   3.000000   3.000000   \n",
       "\n",
       "              54         55         56  \n",
       "count  32.000000  32.000000  32.000000  \n",
       "mean    1.562500   1.812500   1.718750  \n",
       "std     0.504016   0.396558   0.456803  \n",
       "min     1.000000   1.000000   1.000000  \n",
       "25%     1.000000   2.000000   1.000000  \n",
       "50%     2.000000   2.000000   2.000000  \n",
       "75%     2.000000   2.000000   2.000000  \n",
       "max     2.000000   2.000000   2.000000  \n",
       "\n",
       "[8 rows x 55 columns]"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a1b27b438>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "data_mean=data.loc[:,1:56]\n",
    "#построим гистограмму\n",
    "hist_mean=data_mean.hist(bins=4, figsize=(15, 10),grid=False,)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Ни в одной из гистрограмм не заметно аномальных выбросов."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "4\n",
      "38\n"
     ]
    }
   ],
   "source": [
    "for nul_col in data:\n",
    "    if data[nul_col].dtype != np.int64:\n",
    "        if (data[nul_col] == '?').any() == True:\n",
    "            print(nul_col)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Удалим 4 и 38 из-за наличия пропусков.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [],
   "source": [
    "data.drop(4,axis=1, inplace=True)\n",
    "data.drop(38,axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "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>groups</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   groups\n",
       "0        \n",
       "1       9\n",
       "2      13\n",
       "3      10"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "kind_cancer = data.groupby(0, axis=0)\n",
    "pd.DataFrame(kind_cancer.size(), columns=['groups'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Анализ мультикорреляции признаков\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a1ac25e48>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "corr = data.corr()\n",
    "sns.heatmap(corr, \n",
    "            xticklabels=corr.columns.values,\n",
    "            yticklabels=corr.columns.values);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Выбросим полностью коррелируемые признаки 31-32 и 47-48, а также сильно коррелируемые 51-52, 52-53."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [],
   "source": [
    "data.drop(31,axis=1, inplace=True)\n",
    "data.drop(47,axis=1, inplace=True)\n",
    "data.drop(52,axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Список моделей:\n",
    "Для решения поставленной задачи используем следующие модели:\n",
    "* kNN \n",
    "* SVM\n",
    "* Логистическая регрессия"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Список функций ошибки, критериев качества:\n",
    "В качестве критериев качества используются:\n",
    "* ROC AUC (не определен для мультиклассовой классификации)\n",
    "* F1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Способ разбиения выборки на обучение-контроль:\n",
    "Будем использовать кросс-валидацию с разбиением на 3 фолда."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# KNN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "from sklearn.metrics import f1_score, make_scorer\n",
    "neighbor_clf = KNeighborsClassifier()\n",
    "max_n_neighbors = 20"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 10 candidates, totalling 50 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Done  50 out of  50 | elapsed:    0.2s finished\n"
     ]
    }
   ],
   "source": [
    "parameters = {'n_neighbors': np.arange(1, max_n_neighbors, 2)}\n",
    "clf = GridSearchCV(neighbor_clf, parameters, cv = 5, n_jobs=4, \n",
    "                   verbose=1, scoring=make_scorer(f1_score,average='micro'))\n",
    "clf.fit(data, classes)\n",
    "knn_score_no_scaling = clf.cv_results_['mean_test_score']\n",
    "knn_score_std_no_scaling = clf.cv_results_['std_test_score']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "построим график"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No handles with labels found to put in legend.\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAtkAAAH4CAYAAACfVfGQAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvNQv5yAAAIABJREFUeJzs3XecVNX9//HXBxZY6rJ0EAEL3RiVoljBXrDEbjTGxBKTmKrJ16+/FI1fU01MM8UUS4xiiQXsBl1FbIidKgIi0tmFbbP9/P44d2QYZndn2dm5U97Px2MeMHfuzHz2zr13PnPu55xjzjlERERERCR1OoUdgIiIiIhIrlGSLSIiIiKSYkqyRURERERSTEm2iIiIiEiKKckWEREREUkxJdkiIiIiIimmJFs+ZWZ3mNkbraxTYmbOzP43wWNbzOz6mPvXB+s+nWDdB82spJ3xXhK8fq/2vE4b33NU8J4zY5b1NLNZZrY1eOyS4G/f0sGxrA7eL/52UfB4VzP7lZnNM7OImbVpvE4zO8vMVphZ5xTEWmJmD7b3dSQ1zOwNM7sj5v6tZvaPEEPKKM0c5983s+kJ1nVmdlVaA0wjM7vczFaZWUNwHO+ybZp53lVtPeeELdm/LcHzxgTn/L4dFVt76PwbHiXZsru+Y2Y9klz3eDOb0qHRpM96YBrwUsyyrwKnAlcEjz0O/B04IQ3x3BO8Z+ztqeCxHsBlQDXwclte1Mw6ATcAv3LONaYgzq8Bu/wwk4zxK+BCM9s37EAyRKLj/PvA9FCiCYmZDQH+DDwKHIU/jhNtm1yxu3/bGODHQEYm2RKegrADkKz0CnAQPqn8bSvrlgJrgf8HnNHBcXU451wt8Grc4nHAMufcf+KWr23v+5lZd+dcpIVV1jvn4uMBwDm3zcz6OeeiLW1Ht+GtjwH2wSfx7eacW9ze10hiW2Ss4GpAZ+dcXdixJOKcW21mL+F/MF4ddjzx0r39mjnO89G+QGfgn865d2OW5+S2yZTPPZvPdbIztWRLs4Jyg4fMbE1cC9c64HbgGjPr1srLOOCnwGlm9pndiOFIM3vezCrNbHtw2evAFtb/uZm9F6y/1sz+HbTGxK5zmpktNLMqMyszs9fM7KiYxy81s0VBicUWM3vBzCYGj+10OdHMVgOXAgdGyzWC5buUi5hZPzP7q5ltNLMaM3vZzA6OW8eZ2XfN7Ldmthl4r63bLJbb/Sldvwg845yriIktWp5zUPA5VJvZ28H9nmZ2e/AZrTSzC2JfLNHlSjPb38zmmNm24PN63cyOCx6bHrzXCWY228wqgT8Gj/Uws9+b2YZgOy4ws+Nb+4PM7Opg3e3BZzAndr82sxuC1+wU97yZQSyx614W7CO1ZvaRmX0/7jl3mC/JOMPMFgE1wMFmNtTM/hlso4iZLTez/zOzrnHPH2FmTwbrrAq2/S4lVma2n5k9bmYVwe2BBPv7fmY2P9hWS8zstGY20X/wrdktfi+0dHwEj3c3s18G26U2iP9nMY93Do6PNcHji8zs88lsv5htM8vMSoN98GkzG9tCvIXB+3w+ZtnPgs/0tJhlfzCz+cH/Ex3n/YEf246yrOkxb9PZzH5qZpvNbJP58psWz41mNi3Yt9eZPxe9bWYXtvScmOe2eF40swPMbG6wfcrMnwcHxzwe/fvONX9O2m7+fHlD9PM3X/o3L3jKO7ajFC5RKU03M/uj+WO51MxuAbokiDvZc+C3WtueZjbSzO4N9sFqM3s37jMuDPbDj4PP/x0zO7mV7Zrob1ttZjeb2XeCbVQW7H99g8enA3OC1VcFz18d8/wW99eY97zQzO4ys23AHDO708xeTxDjVeaPvV7B/RbPaxIuJdmSkJkVAg8DnwWOcM6tiFvlF8Bg4EtJvNwDwHJ8a3ZbYpgOzAXq8UnfefiT/h4tPG0QPqk/Bfg2sDfwnAV1xWa2D/Ag8By+xONC4DGgX/D4kcBfgLuBk4Av40stipp5v88BTwBL2VGukehv6Qb8FzgO+B6+VX8z8F+LS4qCx4cCXwC+2cLfGry0FcTc2l0/HTia5ktM7gTuBc4CDL89/4H/8XU28Bpwl5kNbyHoccB8/N95JX47PgzsGbfqP4B3gNOC/wP8Db/f3RQ872PgcTM7vJW/aTg+UT8duBzfQjffzKKf7Sz8Pn1U3PPOBRZGjwEz+x7+EvojwMzg/zfarnW5o4BfAj8DTgZWAQPwV3e+C5yIL9P4EvCHmG1jwGxgPH7/+y5+P4hPRvbFb8NC/L5yCTAR/wVtwTrdgaeBXsDngf/DX30akWD7vBz8/c3+GG7t+Aje91F8i/itwd/94+DvjvoJ/lxwG/5znQ/82+J+mJFg+5lZP/yl/LH4/eZcoCf+OOqeKGbnXA2wADgiZvGR+MQ9ftk8EvscsB2/D0aP8zdjHr8aGAZchP9MvwJ8q5nXihqJ/9svw5+L/gPcnmA77KS186KZDQRK8OVinwe+gd+nn7W4H3P47VuJP27vBn4U/B98ydvXg/9fyI5SuER+HvwdNwbrjiTuikgbz4Etbk8zG4S/ojoFuAa//f7BzuePB/HHxE+DxxcAs83sgGb+hpaci7+6dwXwP/jj/qfBY28GMQCcid9OnwvibMv+ejNQAZwTvPYsYIqZ7Z0glsedc5XB/dbOaxIm55xuuuGcA7gDeAN/cv4vPnHcI26dEuDBmPVXAQXB/S3A9THrXg9sCf5/CdAIjAnuPwiUtBLPK0E81szjl+Bbyns183hn/BePA44Mlp0NbG3hPa/BJ1TNPT4qeL2Z8dstbr1P//bg/qVAHTA6ZlkB8CG+7jm6zAFvJfl5rQ7Wj72tbWbdqwgatpN43WHBa53SzPb+Ysyyk4Nl/4xZVoRPAL6aaL8J7t+LL6fp3kwM04PXvSVu+XigKS6GTsD7wNNt2Nc7A93xX2oXxyx/B/hLzP1u+OTqmuB+H3xS8uO41/sJsAFf0hDdJxxwQCtxFOAToRqga7DslOC5U2PW2yPYpiUxy/4FLIs+L1g2Gn+cnRLc/1rwvOEx6xwWvP4dCWJpAC5vx/FxQvDapzXzeD+gKsH2ewJfchV7TO2y/fBJ3FagX8yy4uAz+noLcf0MeD/4fyFQi09MXg2W9Y3bbqPY9Tjf6fwWs9wBL8YteyT62knujxZs/78Cz7WybmvnxZ8D24A+McumBnFeEPf33RX33LeBWQmOw/1ilu20bfAt/BHgf+KOyaXEnHNo2zmwxe0ZfJ5VwNBmtsExwescFbf8ReCBFrZtos99dRBjQcyy3wIbYu7PDJ43qq37a8x7PpzgeNwCXBuzbA/8+e/sZuJv7rxWQsz5V7f03dSSLfF64jvODcafoD5pYd2f4lvEkrnEeTewhiQ7v5lZT3zL3Z0uOEsk+byTgkuQ2/EJQ7Quekzw73tAUXAp7vjgfWK9jS/9uMX8Jdn4lp/ddSywEN8aV2Bm0f4QLwCT49ZtrrUokbvxrTnRW4uXQ5MUbVVqbnSUuTH/j17heC66wDm3Hd9C1dIVh6OB+1zrdYfx22IKPiF5IOb9moL7LbZkm9khZvasmW3F7xvV+BbeMTGr3QecFfP5nAT0Bu4P7k/DHyMPxF5BwP/9g/GtSlGfOOfejovBzOzbZrbYzCL4BPjf+GQ+2ro8Bf8F/uml4uA4XBj3Jx2Lb/1violjFT4piO5TU/FJ8dqY15oPbIrfPs65BnxyFt+qGKu14+NooNQ5N7uZ5++H/xH/QNzy+4AxQQtl1C7bD/83PwuUx/zNFfhtE38cxZoHTAhaFg/BJ2h/Bg4y34E7uu/Mb+E1WvJM3P3F7Lwv7MLMis2XPX2E3w/q8S2lY1p4TjLnxan4Uq/y6IJgX1rNrsdIm+NO4DP4Hy6PxrxfU+z9QFvOga3FdTTwlHNufTMxHYv/0Ts/7jidm+C9kvF8cHzExjMoie+HtuyvO53rgvd7CH+lIuoc/L776bpJntckJEqyJd4w4FDgIefcxpZWdM4tx7dI/6+1UscZnDB+CVxkZiOTiKMYn0w1dxLdhfkRTGbjE+sv4BOiQ4KHC4M4luEvq+2Nbz3bYmb3BJdYcc79F3/5/kj8r/8tZvanBMl4Ww0IYqmPu32JXUskWtzu8es6596Iub3b+lNaVRj8W9vM49ti/l+XYFl0eSHN609yn238thgKVDrnqhOs18OaqYM1sxH4L27DX3o+DJ/MboqLcxb+s4p2Ej0PeMU5tya4Hy17WMTOn+PzwfLYzzLR5/ht4Nf45Ph0fEIUvSQfjWMI/kdKvPhlA/CXruP3qb1j4hhCgoS6mWXgP/NmP7ckjo/WPtehwb/x2yZ6vzjBslgD8J9J/N88g12Po1jz8a2Fh+NLRF5yzi3CtygeEix73zkXvx8nq637P/jW+vPw5RDH4/fHf7byvGTOi0NJvO02EpTFxdiduONFf5TF71Px99tyDmwtrtb2swFBXPHvdX2C90pGongMaC3Jbsv+mugzmwUcYGbRhPk8YHa0caIN5zUJiUYXkXgfAL8D7jCzDc65P7ey/k341q2zW1kP/BfID/CJQWvK8JfFhra2YozP4ROR86KtPIkSeufc4/ga3iL8pfnf4mtizw8evxO4M0i8zwRuAcqBa9sQS7xS/CXeryZ4LD6ZTbrlvoOUBv925HBUW0nus43fFuuBXmbWIy7RHgxUOz86QCIn4ltQT3fOVQEErUo7JR3OuZXmx4o/z/xoG6cC18WsEt02M0n8pbishdjBt0Q94Jz7tH+CmU2IW2cDMDDBcwfiy0piY3kYXzsbL3oVYgN+9Jt4gxIsA/+ZlzbzGNDq8dHa5xpNjAYF60ZFO+XFvnei7VeK/yF9Y4LHKhIsi8a83czexSfTB+Dr1MHXyx5By/XYKRf0eTkFuMo595eY5a01fCVzXlxP4s93MLteDUmFDcG/g9j584uPoS3nwNa0tp+VAp8Q/ohWbdlfE+3vJfjte56Z3YW/ivGzmMeTOq9JeJRkyy6cc/8y33P5j2ZW4Zy7u4V13zWzOfhExFp53Vozuxl/kliI/0Xf3LpVZvYacLGZ/THJkpHuQH3cus2WsgRlDfeYH1lklw6LzrnNwF/N7EwgPhFqq7n41qo1zrnmWhEzxSp8S81e+JN8R5gLnGtm/8/5jmnJWoD/MjobuAs+7Wx3Ni2Pbdsdn5zEXvI9l8TnwFn4jnnPBc+LLW14BV9/Oiz4sdZW3dk1oYjfRxfgR7GYGi0ZMbM9gEnsXM4wF19+sbCF42MBfsSQ4dGSETM7jARJWJA098B3Um5VM8fHXOD7ZjbTOfdYgqe9j7+cfQ6+jj3qXGB58JotmRusuyiJUqN48/AtiOPY0Qn7xSCWSbQ+HOnutPI2pxu+fvbTfcHMeuM7gjZ7rkvyvPga8FUz6+2C0YGCq3yj6Jixrd/D//g7HV+HHf2xcHrceqk8B84Fvmlmg5u54joX33my0jm3tJ3vlYzoFb34/aM9+yvOuSbzozKdh9/G5eyYBwHadl6TEOiDkIScc38OEu3bzazSOfdIC6vfhD+xJ+Ov+IT8UHwtXkuuxXfAfNLMbsPXok3DdzJM9AX+LPBtM/stfkilQ/G90z9lZl9hx4Qt6/Adxc5hR8J2A74VoATfGnggvmd+e1qxCV7/SqAk+KGxEn/Jcyq+/vaWdr5+QmZ2Er6G+IDgfvSKwwLn3EeJnhP8GFqITzxu74i48BPdLABeNLNf41umDsR3Sv1nc09yzi0xs3vxPwD74GvCL8cnTolayKKewyc1t5uf2XAivhNfovKA+/GX8H+F74D16WVp58cevx74XXCV5EV82d0YYIZz7nOt/N3P4pOD1/CdqS7Ej0Uc6wl8B8z7zc+sGsGP0LER/4UadT3wOv6qzD/x++se+NEb7nDOleA/vx8E61yP/1K+kcT19pPxCV6zExclcXw8i28lvsfMfoIfeWEovuPxV5xzpcHx+QMza8C3bJ6J70vQ4qgagd/gj+nnzOwP+NbK6IgwLznn7m3huS/iR9qoZMfIIPOC14TWE9ClwClm9lTwGstczBCXbRG0rC8AfmRm5fjP9Vp8+UqfVp7e2nnxN/hj4Wkz+wW+Pvfn+GQ4fiz/dnPObQ3iuCH4TBfhj8n4mXhTeQ68BbgYmGdmN+FHGBoP9HTO/ZId++GzwTZYhN+uBwCFzrlUT4wVvYL1FTObhb+q9h7t21+j7sN3XP8OvnNk7FjxbTmvSRhS2YtSt+y+kXiUjJ/gf0EfF9wvIUEvZXxdmKOZ0UXi1r0uWLckiZiOwn85VuNPHM8TjDhAgtFF8LOyfYz/4vkvPol2+MuysGMYqnXB37UKPxxht+DxmfjWh83B48vwX2oWPD6K3RhdJFhWhC/F+Rjf8rEW37HlsJh1Po01iW2zGrg5iXVcgtslrTzvGmBF3LJE23uX7ZEotkT7DbA/PqGsCG6vAccEj00nblSDmOf1wJf3bMS3BL4BnJDE9roYn9hG8BNOHNzcNsQnXA74SjOvdRH+akwEfwn/NeC7Le0TwfJe+MS3NLj9nR0jE8SO4DAS/0OwBvgI3yHuGeCRuNcbh+8XURrEsgL/QzZ2NJH98YlzLX5/PiPYZnfEvdbv8B28WtqGLR4fwTrd8cORrQ3ecxVwU8zjnfE/sqLHwWLgwtbORTGPDQu2YfTzX43vADyxldgHB9v5mbhYyoGVceuOYtfjfFKw31QFj01v7pilmXNf3Dr74pOkKnyn8O8n87zguc2eF4PHDwxeO/r4PcDgJI7bnbY7SYwuEizrBvwJ/yOhDH98fpe4EY3YzXNgou2CP0buC96vGv/D9Py4mG7AHxN1+LKLp4gbNSmJz301cecIEp8Lr8Yfqw3A6mT31+Y+i5jnW7B/OBKc50jivIZGFwntFk0cREQ+ZX7iijXA4c65BWHHk++C/gMrgT86537cAa/fGZ8gXOtaKA8TEZHkKckWkYTM7FagyDl3UasrS0qZ2ZX4EoIP8B0ev4uvv57ominzaef7nY8vIxnvdh6qTEREdpNqskWkOTcCl5pZZ+dcY9jB5Jla/Cg8I/CXiV8Hju2IBDtgwKVKsEVEUkct2SIiIiIiKabJaEREREREUkxJtoiIiIhIiuVETfaAAQPcqFGjwg4jq1VVVdGzZ3tnDpdY2qappe2ZetqmqaXtmXrapqmnbdp+Cxcu3OKcSzQz705yIskeNWoUb7zxRthhZLWSkhKmT58edhg5Rds0tbQ9U0/bNLW0PVNP2zT1tE3bz8yS6oSuchERERERkRRTki0iIiIikmJKskVEREREUkxJtoiIiIhIiinJFhERERFJMSXZIiIiIiIppiRbRERERCTFcmKcbBERERGRqJqaGjZv3kxNTQ0NDQ0trltQUEBhYSEDBw6ksLAwZTEoyRYRERGRnLF9+3Y2btzIwIEDGTJkCAUFBZhZwnWdczQ0NFBZWcmaNWsYPHgwRUVFKYlDSbaIiIiI5IwtW7YwfPhwevTo0eq6ZkaXLl0oLi6mW7dubNiwIWVJtmqyRURERCRn1NXV0b179zY/r3v37tTW1qYsDiXZIiIiIpJTmisPSfVzWqIkW0REREQkxZRki4iIiIikmJJsEREREZEUU5ItIiIiIpJiSrJFRERERFJMSbaISBqVAn8DKsMOREREOpSSbBGRNKgDfgOMBL4JHA5sCzUiEZHc5ZxLy3NaoiRbRKQDOWA2sBfwI3wLdg2wBJgKbA4vNBGRnNS1a1cikUibnxeJROjWrVvK4lCSLSLSQd4GDgY+D6wDqmIeqwNWA5OBT9IemYhI7howYABr166ltLSU+vr6FluonXPU19dTWlrK2rVr6d+/f8riKEjZK4mICADrgauBR/Ct1s2d3uvxCfYk4GVg77REJyKS24qKiujWrRubN29m69atNDQ0tLh+QUEBhYWFjBgxgsLCwpTFoSRbRCRFqoFfAL8CGvGt1a1pxJeMTAHmARM6LDoRkfxRWFjInnvuGWoMKhcREWmnJuBuYARwMxAhuQQ79vllwKHAmymPTkREwqAkW0SkHV4C9gOuBLbiW7N3hwO2A0cB81MTmoiIhEhJtojIblgJnAKcgB8ppKrl1ZNWCRwPPJui1xMRkXAoyRYRaYPtwLeBicAz7H7LdUuqgdOBRzvgtUVEJD2UZIuIJKEB+COwJ/BX/KghLfdXb58IcAHwrw58DxER6TgaXUREpBVPsqPmOlVlIcmIAF8BKoCvpfF9RUSk/ZRki4g0433gq8BbpDe5jhUBvgeUA9eGFIOIiLSdykVEROJsAi7BT3v+MuEl2FHVwI34JLv5ectERCSTKMkWEQnUAD8F9gJm4VuRm0KNaIdq4A/A11GiLSKSDVQuIiJ5zwEPAN/AD6HXESOGpEI1cCe+RvsOoHOo0YiISEuUZItIXnsNuAL4kPDLQpJRDTyEr9F+EOgSbjgiItIMlYuISF5aA3wOmAG8S3Yk2FHV+MlqTsSXuIiISOZRki0ieaUCP1rHOOAxfN11NorgO2VOx5e4iIhIZlGSLSJ5oRH4GzACuBWfpHbkZDLpUAO8DRwKlIUci4iI7ExJtojkvLnAWOA7wDayt/U6kVpgGX64wU0hxyIiIjsoyRaRnLUUOBo4jezp2Lg76oCPgEnAxyHHIiIinpJsEck5W/HTkR8IvEDmDsmXSvXAemAy/geFiIiES0m2iOSMOuBmYCR+POkaMmcymXRoBLYAU4BFIcciIpLvlGSLSNZzwMPAKOB6fFlIbYjxhKkJX3d+KPBGyLGIiOQzJdkiktXexLfcfgFfLpGrdddt4fCT1UwH5oUbiohI3lKSLSJZ6RPgPOBwfKKt5HpXVfgJa54OOxARkTykJFtEskoV8ENgNL5EJIJvuZXEqvEzWz4UdiAiInlGSbaIZI078ZPJ/AafXNeHG07WiAAX4befiIikh5JsEcl4L+JHy/g6UEp+DMmXahHgq8Afwg5ERCRPKMkWkYy1AjgpuNWguuv2igDXAjeFHYiISB4oCDsAEZF42/B11//Aj33dGG44OaUa+CmwHTg55FhERHKZWrJFJGPUA7/H113/Hd/yqgQ79aqBW4E15NdkPSIi6aQkW0RC54DHgX2A64AKfHmIdJxq/PTzF6IfMiIiHUFJtoiE6j3gMPyY1x+juut0agJmA2fgy3JERCR1lGSLSCg2ABcDBwOvouQ6LNXAXOAEfHmOiIikhpJsEUmrCHAjvjTkPjSZTCaI4H/oHIkv1RERkfZLe5JtZiea2TIzW2Fm1yZ4fISZPW9mb5nZu2amDvAiOcAB9+I7Nf4c34KqEoXMUYMv3ZmGH4tcRETaJ61Jtpl1xndqPwmYAFxgZhPiVvsBcL9z7kDgfOBP6YxRRFLvZWB/4HJgC5pMJlPVAh8AU4CNIcciIpLt0t2SPRVY4Zxb6ZyrA2YBp8et44A+wf+LgHVpjE9EUmgRcCxwHPA+qrvOBnX4DqiT8EP8iYjI7jHn0lcNaWZnAyc65y4L7n8BONg5d1XMOkOBZ4BioCdwrHNuYYLXugK4AmDw4MGTZs2alYa/IHdVVlbSq1evsMPIKfm8TeuAtfgJT1I1DvPwykrW5un27CitbdMCYBzQLW0RZbd8PuY7irZp6mmbtt+MGTMWOucmt7Zeumd8tATL4rP8C4A7nHO/NrNpwL/MbD/n3E7f1c6524DbACZPnuymT5/eEfHmjZKSErQNUysft+lW4Hr8TI0N+MllUuXmkhKuybPt2dFa26YG9AVeAD6TppiyWT4e8x1N2zT1tE3TJ93lImuBPWPuD2fXcpBLgfsBnHOvAIXAgLREJyK7pQq4ARjJjpkaU5lgSzgcUIYfx/z1kGMREck26U6yFwCjzWwvM+uK79g4O26dNcAxAGY2Hp9kb05rlCKSlHp8T+bhwC/xybZmasw9FcDR+BZtERFJTlqTbOdcA3AV8DSwBD+KyCIz+4mZnRasdjVwuZm9gx/x6xKXzsJxEWlVE77X8gjgf4BtaMSQXFeFHxbqybADERHJEumuycY59wTwRNyyH8X8fzH+6qSIZBgHPIv/pbweqAw3HEmzCHAWcCdwTsixiIhkurQn2SKSnRYA30BD8eW7CPBF/A+sL4Uci4hIJlOSLSItWgZ8F3geX2+t2i2J4K9mlAPfCjkWEZFMlfZp1UUkO3wCXAwciO9EEUEJtuxQDVwH3Bh2ICIiGUpJtojspAzfcj0a37kxAjSGGpFkqmrg5/je6voBJiKyM5WLiAjgk+lbgJ/ik2oNxSfJqAb+gi8d+StquRERiVKSLZLnGoDbgWvxibWG4pO2qgbuwSfa/0ZfLCIioEYHkbzlgP8AewHfAUpRgi27rxp4DDgNqA05FhGRTKAkWyQPPQ/shx+KbS0akk9SoxooAY5HP9hERJRki+SRt4AjgFOBxSi5ltSLAK/j97PykGMREQmTkmyRPPAhcAZ+KtX5KLmWjlUDLAIOAbaGHIuISFiUZIvksA3AZcBngMfRWNeSPrX4H3dTgPUhxyIiEgYl2SI5qBw/WsjewL/wyXVDqBFJPqoDPgYmAx+FHIuISLopyRbJIbXAzcBw4Pf45Lou1Igk3zUAG/GJ9vKQYxERSScl2SI5oBG4A9gTuB6owCfYIpmgEV+bfTDwbsixiIiki5JskSzmgDnAvsA3gM2oU6NkJgdsAw4HXgs5FhGRdFCSLZKlXgIOBC4AVgOVoUYjkpwK4Bj8WO0iIrlMSbZIlnkfOBo4AXgHtVxL9qkCZuJHvBERyVVKskWyxEfAucBU4AU0o55kt2rgHOC+sAMREekgSrJFMtwW4GvAOOBhfIfGplAjEkmNCPAl4JfoioyI5B4l2SIZqhL4MTAS+Cd+Fj2NdS25JgL8BBiE/zH5QbjhiIikjJJskQxThx/jejh+zOtq/PjXIrmqCr+f/x3YHzgMmI0f+k9EJFspyRbJEE3APcAI4DpgO6q7lvxSj79i8zJwETAEuAk/NKWISLZRki0SMgc8BYwFrsDPjqf6VMl3Ffj+CDfhf3ieA7weakQiIm2jJFskRK/hRws5G1iBkmuReBF86/ZD+KErx+FnN9WMpiKS6ZRki4RgKXAiMAN4AyXXIq1pwh8ny/Czmw4Cvg2sCjMoEZEWKMmRBrpYAAAgAElEQVQWSaO1+FrTg4BnUWucyO6oDG5/BiYA0/ElVxraUkQyiZJskTQoxbe6jQHuR2Ndi6RCHb6U5AV8zfYe+BF5ysIMSkQkoCRbpANVs6Pj1l/xyXV9qBGJ5KZKYAN+bPlhwIXAW6FGJCL5riDsAERyURN+2LHh+DGuNRSfSHpEj7X7gEeAvYFr8Z2Lu4UVlIjkJbVki3SA6/D112UowRYJQyP+2HsfuBLfUfJ/gI/DDEpE8oqSbJEUexo/Y6NqrkUyQyVQDvwW3y/ieGAufox6EZGOoiRbJIXWAueiUUNEMlG0o+SzwBn4vhK/xyfgIiKppiRbJEXqgVNReYhINqjE/yi+Dj99+5fwpSUiIqmiJFskRa4BlgMNYQciIkmrwl95+hd+9tWDgAfQKEAi0n5KskVSYDbwN9SKLZKtGvHJ9lvApfiOkj8E1ocZlIhkNSXZIu20Gj8mr+qwRXJDBbANP7HN3sBM4EXUUVJE2kZJtkg71AGnoARbJBfVBLcngJPxCfdf8PXcIiKtUZIt0g7fAFbhLzWLSG5y+Nrt1fi+F4PxY28vCzEmEcl8SrJFdtMDwN2oFVskn1Th+178AzgQOAQ/s6Q6PItIPCXZIrthBX7IL3V0FMlPDfgf2K8BF+OHAVwPbAozKBHJKEqyRdoogq/PVAu2iIDvKLkVn2SPBM4EXkEdJUXynZJskTa6Ej+JhaZNF5FYDt9R8lHgOGAs8E/0g1wkXynJFmmDu4AH0ZemiDSvCV+7/QHwLfyY298EPgwzKBFJOyXZIklaDHwV1WGLSPIqg9tfgP2AI4HH0ZUwkXygJFskCVWoDltEdl89vpRkHnA+MAz4JVAaZlAi0qGUZIu0wgGXABtRRyYRab9K/PnkemAP4AJgYZgBiUiHUJIt0oq/42d8qwk7EBHJKRH8eeV+fBnJROBf+FZvEcl+SrJFWvAOvuOS6rBFpKM04c8x0X4fx+BL1EQkuynJFmlGOXAKqsMWkfSpAhYAhwPbQo5FRNpHSbZIAg64ENgSdiAikndqgCXAVHzttohkJyXZ7bAVnQBz1R+A54DasAMRkbxUC6wGJgFrwg1FRHaTkux2+CswAT/hgOSOBcC1qA5bRMJVD2zAJ9rLQo5FRNpOSXY7NAJlwCEo0c4VZcBMVIctIpmhEX/V9GDgrZBjEZG2UZLdTo4difbykGOR9nHAuaizkYhkFgdsxw/zNy/kWEQkeWlPss3sRDNbZmYrzOzaBI/fYmZvB7flZpbxOU800Z6GEu1s9kvgZaAu7EBERBKoBE4Engo7EBFJSlqTbDPrDNwKnIQvZ77AzCbEruOc+45z7gDn3AH4/mcPpTPG3RXboq3auewzH7gB1WGLSGarBs4E7gs7EBFpVbpbsqcCK5xzK51zdcAs4PQW1r8AuDctkaWAw5caTEOJdjbZjN8JVYctItkgAnwJ3/leRDKXOefS92ZmZwMnOucuC+5/ATjYOXdVgnVHAq8Cw51zjQkevwK4AmDw4MGTZs2a1aGxJ7IeWNfMY52BcUBh+sJpl8rKSnr16hV2GKFYhp8AItVHwvDKStbm6TbtCNqeqadtmlrp3p6dgKHAkLS9Y/rl83dTR9E2bb8ZM2YsdM5Nbm29gnQEE8MSLGsutzkfeDBRgg3gnLsNuA1g8uTJbvr06SkJsC1uBH7UzGMG9MXX+I5LW0S7r6SkhDC2YdiuB26mY6YwvrmkhGvycJt2FG3P1NM2Ta0wtmcP4Gv4PiWJvmCzXb5+N3UkbdP0SXe5yFpgz5j7w2m+Mfh8sqhUJF5s6ciSkGORxJ7HfzF1RIItIpIO1cCfgMuBppBjEZGdpTvJXgCMNrO9zKwrPpGeHb+SmY0FioFX0hxfSkWHXToUWBxyLLKzDfjOQ6rDFpFsV41vkToHP4GNiGSGtCbZzrkG4CrgaXwD7/3OuUVm9hMzOy1m1QuAWS6dBeMdJJpoH4YS7UzRCJyGHw5LRCQXVOOH9jsJNR6IZIp012TjnHsCeCJu2Y/i7l+fzpg6WmyL9nxgYrjh5L3rgEVAQ9iBiIikUDW+H9BRwFygd7jhiOQ9zfiYJg4ox7dovx9yLPnsKfzg6xoPW0RyUQR4Fz8N+5aQYxHJd0qy0yjaon04SrTD8DFwHrqUKiK5rRb4EJgMfBJyLCL5TEl2CKKJ9nthB5JH6oFT0UgiIpIf6vAJ9iR8wi0i6ackOyTbgSPwl/Wk410NfIDv9Cgikg8a8DPaTkGNOiJhUJIdIiXa6fEo8A9Uhy0i+acJKMP3B3o15FhE8o2S7JCV4xPtd8IOJEetAi5CCbaI5LcK4Fj8qCMikh5KsjNAOXAkSrRTrRY4BSXYIiLg+6ScCjwcdiAieUJJdoaIJtpvhx1IDrkKWI2mGhYRiYoAFwK3hx2ISB5Qkp1Boon2W2EHkgPuA+5Bw/WJiMSL4Bshbgk7EJEcpyQ7w1TgZ+tSor37lgOXojIREZHmVAM/AH6In8NBRFJPSXYGiibab4YdSBaKACejFmwRkdZUA78Bvo7K6kQ6gpLsDFUBTEeJdltdAaxDXxgiIsmoBu7C12k3hByLSK5Rkp3Boon2wpDjyBZ3AA+hVmwRkbaoAmYDp+FHZRKR1FCSneGiifYbIceR6RbhL3mqDltEpO2qgRLgGHzSLSLtpyQ7C1QCM4AFYQeSoSpRHbaISHtF8FdOD8XPEiki7aMkO0tUAkcDr4cdSIZxwBeBTaiHvIhIe9UAS4EpwIaQYxHJdkqys0gl/lKeEu0dbgOewn8xiIhI+9UBa4BJwEchxyKSzZRkZ5loov1a2IFkgLeB76A6bBGRVKsHNuIT7SUhxyKSrZRkZ6Foov1q2IGEaDtwCqrDFhHpKI1AKTANjXIlsjuUZGepKuBY8jPRdsDnga1hByIikuMcvlHjKODFkGMRyTYFYQcguy+aaD+Lb2nIF78DXkDjuWYTi9Rjje2bIihS5+hUmRufelP3LtBZbRySPaqAk4D78VcRRaR1SrKzXBVwHPAMftilXPc6cB0qE8kWncuqKX5iCT3fa/84BX8H9nzuv+0PKgM09O7GtpPGUXXAHtDJwg5HJCnVwDnAP4ALQo5FJBsoyc4BVcDx5H6iXQrMRAl2NrC6BvqUfEifF1eCwfaj9qGhb2G7XvPM5R/w0JjRKYowPNbk6PnWJwy4/x16v/oRpadOpG7PvmGHJZKUCHAZfhztr4Uci0imU5KdI6KJ9tPAYSHH0hGa8C0o28MORFrmHD3eXU/xE0so2F5D1WeHUXbSOBr7dm/3S+9fu5q7po1qf4wZoGLaKHq+uZbip5Yx9Nb5VE4aTtmJY2nq3b4fIiLpUA18D98v5ochxyKSyZRk55Aq4AT8uNGHhxxLqv0CP2xhXdiBSLO6fLKdfnMWUbi6jNphfdhywYHUjuoXdliZqZNRNXlPqvcbQtHzK+jz0ip6vL+B7UfvS/lhe0GB6rUls1UDP8e3aP8aUNGTyK6UZOeYKuBE4EngiJBjSZWXgBtRmUim6lRZS99nltFrwcc09ejK1jM/Q+XkPVVrnARX2IVtJ42ncsoIih9bTPGTS+m14GPKThlPZNwgMG1DyVzV+AnByvB9JjqHG45IxlGSnYNiE+0jQ46lvTYDp6MEOyM1NtH75dX0nfsBVtdIxWF7se2Y0bjuXcKOLOs0DOjJ5kumULhsE/0eW8ygO98gMmYgpTMn0DCoV9jhiTSrCngAn2g/AOjoF9lBSXaOqsYPt5TNiXYj8DmgIuxAZBeFyzfTb84iumyuCpLB8TQM6h12WFmvZuwg1u07wP94+e8HDPvti1QcOoptx47GFSp9kcxUhR9K9njgcaBHuOGIZAwl2Tksmmg/gZ9IINvcgJ86vT7sQORTBVuqKH58MT2WbKK+fw82fXGyyhpSrXMnKo7Ym6oD96Dv08voPX8VPd/6hG0njFUZjmSsavzkaEcAzwFF4YYjkhGUZOe4auBkfOvC9HBDaZO5wM2oTCRTWG0DRc99QJ+XVuE6d6LspHGUHzYKClSF2VGaenWj9Kz9qTxkJMWzF9H/offo9dpHlJ06UR1KJSPVAIuAg4F5wMBwwxEJnZLsPFCNn6ErWxLt9cDZKMHOCMGYzn2fWkpBRS2Vk4az7YSxNPbRUHPpUrdHERuvnEaPd9ZR/MRShvzlFT804snjaCxq/9CIIqlUC6wEJgPzgeHhhiMSKiXZeSKaaD8GzAg5lpY0AKfha/wkXF3XlNFvzmK6fbyN2j37svniyZo0JSxmVB+wB5EJg+lT8iFFL66k++KNlE/fh/Ij98Z10RUFyRz1wCfAJPzoUNk/hZTI7lGSnUeq8TMmzgGODjmW5vwvsBjVYYepc3kNfZ9aSq83P6Ghdze2nPNZqg7U9N+ZwHUtYPvxvja7+Ikl9H12Ob3e+JjSU8YTmThEtfGSMRrxo0NNBUqAz4YajUg4lGTnmWrgVGA2cEzIscR7EvgTPkYJQUMjfV5aTdFzH2CNju3T92H7jH1x3XSayDSN/Xqw5aJJVK7YQvGcxQy6+00i+/Sn7NQJ1A/pE3Z4IgA4YBu+M+RTwKHhhiOSdvr2zEPV+JKMTEq0PwbORwl2KJyj+5JNFD++mC5bq6keP5iyU8bTMKBn2JFJK2r2HcD6bx5Or9fX0PfZ5Qz93TwqDhnJ9uPG0NSja9jhiQB+GNbjgYeB40KORSSdlGTnqWii/ShwbMix1OHrxVWHnX4FmyroN2cx3T/YQt2gXmz88lRqxmhMgKzSuROV00ZRvf8wiv67nN6vfkTPd9ax7bgxVE4dAZ01RbuErwo/sdhd+I7tIvlASXYeq8af9B4h3Fm6vgt8iK/hk/SwSD19/7uc3q98hOvamdKZE6iYNlIJWRZr6tmVstP3o/LgEfSbvZj+jy6i92trKD11ArX7DAg7PBEiwMXAduDSkGMRSQcl2XmuGjgDuDek938YuB2ViaRNk6PXgo/p+8wyOlXXUTl1BNuOG0NTr25hRyYpUj+kDxsvP5juizbQ7/ElDPnba1TtN4Syk8fT2E9z8Um4IsA38dOwXxNyLCIdTUm2UI1vSX4aOCGN77sS36qhBDs9uq0qpXjOIrqtK6dmVD9KT5tA/TDNy5aTzIjsN5R1YwfR58WV9Cn5kB5LN7H9yL0pn74PrqtO/RKeauDHQClwE6AxcSRX6UwrADQBZwL/AU5Mw/vV4GeiVILd8Tpvi1D85FJ6vrOOhqJCNl9wINX7D9Vwb3nAdenM9mNGUzl5OMVPLqXvcyvotXAtZSeNo/qzw8IOT/JYNfB7YCvwZ0CFapKLlGTLp6qBs0hPov11YA0+uZeOYfWN9HlhJX1eWAEOth0zmvKj9sF11cQl+aaxqDtbzj+QikNGUjxnEQNnvU3NKx+xaZh6Qkh4qoC78aUj96CERHKP9mnZSTU7WrRP6qD3uBeYhaZN7zDO0eO9DRQ/sYSCbRGqPjPUT8FdrHrcfFc7qh8bvn44vd74mL5PL+O+j+ro1/Au204Yq7p8CUU18Dh+hKlHgcJwwxFJKSXZsosIvkX7QXxJRyotAy5HZSIdpcv6cvrNWUThylLqhvRmw+WHULtP/7DDkkzSyaicOoKq/Ydy2h3P8dbCtfR8bz3bjhlNxbRRUKAL95Je1cA8YAbwLNAr3HBEUkZJtiQUwY9l+gC+hSEVqlEddkfpVFVH32eX0eu1NTR178LWM/ajcsqeGpJPmuUKu3DEuG48fuZk+j22mH6PL6H362sonTmBmrGDwg5P8kwEeBuYBrwA9As3HJGUUJItzYoA5wD3AzNT8HqXAevxU+1KijQ20fvVjyj67wd0qm2gYtooth87WrP9SdIaBvVi05em0H3pJoofW8zg2xdQPW4QZTMnaNZPSasaYDkwGZgPDA03HJF2U5ItLYoA59L+RPuf+Ho71WGnTuGKLRTPWUTXjZVE9u1P2cyJ1A/pHXZYko3MiIwfTGT0APrMX03RcysYdssLlB+2F9uP3hdXGOZ0VZJP6oCPgUn4RFskmynJllZFE+37gFN34/nvA99AZSKpUrC1muInFtNj0Ubq+3Vn0xcmEZkwWEPySfsVdKb8qH2oPGgPip9aRtGLK+n51idsO2EsVQcNh07ax6TjNQAb8S3a/w45FpH2UJItSYkA5+FHBTmtDc+rxNdhqwW7/ay2gaKSFfSZtwrXySg7YSzlh+8FXTQkn6RWU+9Ctp7zWSoOGUm/OYsY8OC7n07RXjeiOOzwJA804Yf2W4If3u/LwJFoPG3JLkqyJWkR4Hz8EHynJ7G+w8/ouBnVYbeLc/R8ex19n1xCQXktlQfuwbYTx9FYpMGupGPV7dmXDVceSs+3P6Hvk0sZ+qeX/f530jga+2j/k47l8Mn2v4FHgC7ARcCXgM+imSIl8ynJljaJABeQXKL9F+AZfGcW2T1d126jePYiCtdso3Z4EVsuPIjakep3L2nUyag6aDjVE4dQ9Ly/ktJj0Qa2H72vv5JSoCsp0rGagIrg/38C/oEffeQyfNK9d0hxibRGSba0WTTRvgc4o5l13gSuRmUiu6tTRS3FTy+l58K1NPXsypaz9qdqkmpiJTyuWwHbThxH5ZQ9KX58CcVPLaPXgo8pO3m8+gRI2jQEtyrgp8DPgH2Ar+BLGjX4pGQSJdmyWyLA5/FT4p4Z99h2/EgkSrB3Q0MTfV5eRdHcFVhDI+VH7K3RHSSjNPTvyeaLJ1P4wWaK5yxm0L8WEhk9gLKZE6gfrNFtJH1qg38XAdcC38N3lvwKvgFIe6OETUm27LYI/lLdXfiJa8DX0J0PlIYVVBYrXLqJfo8tpsuWKqrHDvTjFA/U3GeSmWpGD2T9t47w47Q/u5yhv5tHxbSRbD9mDE099KNQ0is6etV84B38zMLHBf+eCGjmAAmDkmxplwi+cyP4RPsW/PS4tc0+Q+IVbK6k+LHF9Fi2mfoBPdl4yRRqxumip2SBzp2oOGwvqj47jL7PLqf3y6vp+fY6th0/hsopI1TeJKGoDP59DHgRaATOAi4FDkcjlEj6pD3JNrMTgd8BnYG/O+d+nmCdc4Hr8Q2j7zjnPp/WIKVNoon2u8DNqEwkWVZTT9HcD+jz8mpcQWdKTx5PxaGjoEBfAZJdmnp1o/Rzn6Hi4BH0m7OY/g+/T+9X/ZB/tXv3Dzs8yWPlwb93Aw/jRyi5GLgE2B+NUCIdK61Jtpl1Bm7FX8VZCywws9nOucUx64wG/hc4zDlXZmZq0ssCEeAX+Nm6pBVNjp4L11L89FI6VdVRNWk4ZSeMo6l3t7AjE2mX+mFFbLziEHq8t57ix5cw5LZXqdp/KGUnj6exb/eww5M8FjtCyR+B24D++HKSi4C9QopLclu6W7KnAiuccysBzGwWfiS4xTHrXA7c6pwrA3DObUpzjLKblGC3rttHpRTPXky3T7ZTM6IvZZdMoW5437DDEkkdM6r3H0Zk3GD6vPAhfV74kO5LNlJ+1D6UH7kPrquG/JNwRUcoqQZuwo9Ssi87RigZGF5okmPMufRNE2JmZwMnOucuC+5/ATjYOXdVzDqPAMuBw/AlJdc7555K8FpXAFcADB48eNKsWbPS8BfsbD2wLu3v2jGGV1aytpc62aVS7DatrGni5WV1LFvfSM9uxmFjuzJmaGdMw54lTfto6qVjm1ZEmnhpWR0rNjTSu9Dv+/sOyc19X/to6qVzm3bC16j2xCfafcnN+u3Kykp6aT9tlxkzZix0zk1ubb10t2QnOqvGZ/kFwGhgOjAcmGdm+znntu30JOduw1/xYfLkyW769OkpD7Y1NwI/Svu7doybS0q4JoRtmMtuLinhmsOOoM+8lRTN/xBzjvIZ+7Bm+r4s7qY+x22lfTT10rZNT4JuK7dSN2cxT71TTk15P0pPnUj9sD4d/95ppH009cLapr2Benxt6xXA8eTOCCUlJSWEkTPlo3R/068F9oy5P5xdG4PXAq865+qBVWa2DJ90L0hPiCIp4Bwfbmxg2C0v0KU0QvXEwZSdPIGG/j3CjkwkFLV792f9Nw6n1+tr6PvMMob+YR6VU0ew7fixNPXMlfRFckW0fnsO8AK+pvts4Mv4y+y52MItqZf0fmJm+5vZfWb2oZnVmtlBwfKbzOykJF9mATDazPYys674IZVnx63zCDAjeO0BwBhgZbJxioSuvpGBd77BE2/V4rp0ZuNlB7P5C5OVYIt0MioPGcm6a2ZQMW0UvRZ8zLBfPU/v+augsSns6EQSKscPC3gXcDJ+Vsmr8SNqibQkqSQ7SKIXAkPw+1nsTAO1wDeSeR3nXANwFfA0sAS43zm3yMx+YmanBas9DWw1s8XA88D3nHNbk3l9kdA1OQY88A49lm7i8LFdWf/NI6jZd0DYUYlklKYeXSg7bSLrv3UEdcP70m/OYob+bh6FH2wJOzSRZjXhk+2twB+AacAofOfJ1aFFJZks2ZbsnwF3OOeOwu9Psd4GDkj2DZ1zTzjnxjjn9nHO3RQs+5Fzbnbwf+ec+65zboJz7jPOufT3aBTZTX2fWUbPd9dTdtI4DtyrC3TWRUWR5tQP7s2mS6ey6QuTsIYmBv/jNQbe9QYFW6tbf7JIiOrxo5N8BPwfMB4/7vatwOYQ45LMkmwGMA64L/h/fEfFcqBfyiISyVK9FqyhqORDKqaOoPzIvcMORyQ7mBGZOIR13zmSshPGUrhiC8N+8wJ9n1qK1TaEHZ1Iq2qC23vA/+A7nh0J/Jsds09Kfko2yd4ENJc1TATWpCYckexU+MEW+j38PpHRAyg9fSLk4PBkIh2qS2fKZ+zLuqunU7X/UIpKPmTYr0vo+dZaSONQsyLtUYWvoZ0HXIkfCvB04HF867fkl2ST7FnAT8zs8JhlzszG4H+4/TvlkYlkiS4bKxh490LqB/Vi84UHqUREpB0aiwrZet4BrP/qoTT2KWTAfe8w+M8v0/Xjba0/WSSDVOJbuGcDF+Av+X8ZeAlf3y25L9ls4IfAG/iRbKKt1o8C7+M72P409aGJZL5OFTUMun0BTV07s+mSKbjCLq0/SURaVTeymA1fO4wtZ+9Pl9IIQ2+dT/8H3qFTRU3YoYm0WQU+6b4TOAkYDFyDLzGR3JXUONnOuVpgppkdAxwDDABKgbnOuWc7MD6RjGV1jQy68w06VdWx8SvTaOzbPeyQRHJLJ6Nq8p5U7zeEoudW0Gf+Knq8v4Htx+xL+aF7QYGuGkl2iY5QUgn8HvgzfkjAy4ELgZHhhSYdoNUk28y64X9wPeacmwvM7fCoRDJdk2PAfW/R9ZPtbL5oEnXDi8KOSCRnucIubDt5PJVT9qT48SUUP7GUXq9/TNnM8UTGDQ47PJHdUh/cVuNnkL4RPzHIV4Bz8a2Zkt1abQYIWrH/H9C348MRyQ59n1pKj0UbKTtlApGJQ8IORyQvNAzsxeZLprDxS1MAGHTHGwy6/XUKNmsMB8lu0RFK3gW+h58O+yjgXnxnSslOyV5rew2Y1JGBiGSLXq9+RNGLKymfNpKKw0aFHY5I3qkZO4h13z6S0lPG0211GcNueZHixxZjNRq/QbJfNX6EkhfxrdoDgM8BT6ARSrJNUjXZwPeBe8ysDv85byRuvGznnGYPkJxXuGwT/R59n+pxgyibOUFD9YmEpaATFUfsTdUBe9D36WX0nr+Knm9/wrYTxlI5aU/opGNTsl9F8O8j+FrdHsCf8Em39vDM15aW7H3wdfof4CegqYi7ieS0LuvKGfjvN6kf0octFxyoofpEMkBT726Unr0/G75+OPX9e9L/P+8x5Nb5dPuoNOzQRFKqAt/CeTEwFT/dtmS2ZFuyv8yuMz2K5I3O5TUMunMBTYVd/FB93ZI9dEQkHeqGF7Hxymn0eGcdxU8sZcifX6HqgGGUnTSOxiKN/CO5owpYCBwKnAXcjB8SUDJPskP43dHBcYhkLKttYOAdC+gUqWfDlYfSWFQYdkgikogZ1QfsQWT8YIpKPqTPvJV0X7SR7TP2ofyIvaFL57AjFEkJB0SA+4GH8KNTXA10CzMo2UWbrneb2TAzO8vMLg/+HdZRgYlkhCbHgHvfouv6cjZ//iDqh/UJOyIRaYXrVsC2E8ay7rtHERkzkOJnljPslhfo/v4GTdEuOaUO31HyJmAUvnZbe3jmSKol28w6A3/Aj5ce2xTQaGa3Ad9wzmmWUMk5xY8tpsfSTWw9fSI14waFHY6ItEFDvx5s+cIkKldsoXjOYgbdvZDIPv0pO3Ui9UN6hx2eSMpUB7eLgAnA34H9Q41IIPmW7BvwddnX4X8sdQ/+vS5Yfn3qQxMJV+/5q+jz8mrKD9+Lymmjwg5HRHZTzb4DWP/Nwyk9bSJd15Uz9PfzKH70fTpV14UdmkhKVQFvAIcAXwQ2hRtO3ks2yb4Y+IFz7lfOuTXOudrg318BPwQu6bAIRULQffFGih9bTPWEwZSdPD7scESkvTp3ouLQUay7ZjqVU0fQ+9WPGHZzCb1e/QgadSFWcke0XnsWsDfwC3xZiaRfskn2IPxERIm8Gzyed7au2x52CNIBuq7dzoB736JujyK2nH+AxtsVySFNPbtSesZ+rP/mEdQP6U3/R95n6B9eotuHW8MOTSSl6vAt2z/Blx48iuq10y3ZJHs5cH4zj50PLEtNONnj+WWbeOT3L1H07HJ1pMkhnbdFGHjnApp6dmXTFyfjumqoPpFcVD+0DxsvP4TNFx5Ep5oGhvztVQb8eyGdyzSvmuSWamA9cCEwDd/KLemRbAbxf8AsMxsBPIgfD30QcA4wg+YT8Jx1+L4DGDN5OMvnfkBBWTVbz9wfCjQ5STazmnoG3bGATnWNbPjqwTT11lB9IjnNjOrPDCUybhB9XlxJn5IVdF+yifIj96Z8+r64rhryT3JHFfA6sBQ/9N8vgYGhRpT7ksoKnXP3AycCPYHfAf/Bz/7YAzjROfdAh0WYobp07sThZ+3PtuPG0JkprioAACAASURBVOvNTxh0++tYpD7ssGR3NTYx8J636LKpks0XHaSRB0TyiOvSme3HjGbd1dOJTBxC3+dWMOzXJfR4Z52uVEpOcUATcA+wF/ArVK/dkZJuenXOPeOcm4YfWWQI0N05d6hz7tkOiy7DmRnbjxnNlvM+S+HqUob85WU6b9OFmKzjHP1mL6L78s2UnrEfNaP1214kHzX27c6WCw5kw5XTaOzZlYH3vsXgv75Kl0/U/0ZyS7Re+wZ8sj0H1Wt3hKSSbDPrbWZDAZxzTc65TdFxsc1sqJn16sggM13VgcPZ+OWpFGyvYcit8+mqE3JW6T1vFb1fW8P2o/ahcuqIsMMRkZDVjurHhqsOZ+uZn6HL5kqG/vEl+j30Hp0qa8MOTSSlqoB1wAXAYcD74Ybz/9m77+g4q2vv498zqjPqXZZ7N8amuRtwIfSeEAi9h9BSCL439RLS3pubEAgklJBAIAQCSQgECCU02YALtjHGuGKDbdwk2ZIsS1bXef84IxBGsiVrNM+U32etWfLMPDOzdTwabZ1nn31iTndnsh/ALVDtzK24vudxrXF4Pjuumw4JPop+v4DUNepOGQ0CK7aT88Jq6sb3o/qk0V6HIyKRwmeonTyIrXNmsWf6UNKXfEz/20rJePMjtfyTmFMHLAIm4zY/2eltODGju0n2DODfXdz3fPD+uNdclMGO66fTXJBG4cOLXf9ViVjJm6vIe+JdmgZms+u8w9WqT0Q+x/qTqDpjLNu+dSyNg3LIfW4V/e58g9R1FV6HJhJSbbjOI4/hWv7dhuq1e6u7SXYWrgtMZxqAnNCEE/1aM1Mpu2Ya9aMLyXv6fbJfWANtqnSKNAmVeyn88xJaM1Mov3QiNkldBESkay2FGZRfMYnySydiWtsoevBtCh5eTOLOOq9DEwmpRtzM9q24zWyeQ/XaB6u7SfYHwGld3HcqsCE04cQGm5JIxSUT2DN1EFlzN5D/xLvQ3Op1WBJk6l2rPlraKL98Mm3pKV6HJCLRwBjqxxax7aYZVJ0yhtQNuyi5Yx7ZL6zBNLZ4HZ1ISNUBW3E9mo8FVnobTlTqbp/s3wL3GWOagIdwfc37AZcBNwDX9Ul00SzBR+VZ42jJCZDzwhoSdtdTcclE2tKSvY4svrW0UfDoUpJ21VF25RRaCuN6za6IHIzEBGpmDqfuyP5kv7iWrLkbSHtnC6uHWHfmUqVnEkPqgAXAJOBC4BdAvqcRRY/u9sn+A/Aj4HrcNuoVwa83AD8M3i/7MoaamcOpuPBIUrbspvje+STu0m5inrGWvKdX4F+/i11fOozG4XleRyQiUaw1M5Vd5x3O9uun05rt55UVTRTfO5/kzVVehyYSUu312n/Btfy7HdDOIAfWkz7ZPwNKcGUjlwa/llhrf9FHscWMvYeVUHb1FHx7myi+5y2SP672OqS4lFm6gfQlW6g+bgR1EwZ4HY6IxIimQTnsuG46x49PJqG6nn73zCfvb8tJqGnwOjSRkGoEaoFbcPXaz3sbTsTr0T7g1trd1toXrbWPBr+qIXQ3NQ7JZcd102lLSaTo/gX4V+7wOqS4Eli+jZyX1lJ3RAm7TxjldTgiEmt8hkP6J7Ftzix2zxxO2vJtlNxWSmbpBmjRmhyJLXXAFuA8XHu5Vd6GE7G6uxnNOcaYqzpcH2qMmW+MqTbGPGmMye67EGNHS0G6a/FXnEnBX5aS8dZHXocUF1I2VpL/9+U0DMlh55cPA6N6SRHpGzYlkepTxrDtphk0DM8n58U1lNwxD/+qMm3RLjGnDngLmAB8DdjlbTgRp7sz2T8EMjtc/y2u7v0XwFHAz0McV8xqS0+h7KtTqR9bRO6zq8h5dpVa/PWhxJ11FPx5CS3ZfioumQiJatUnIn2vJT+NissmUnblZGyCj8I/L6HwT4tJLN/jdWgiIdWG6+X8MK6/9h2oXrtdd5PsYcAKAGNMFnAicFOwHvsHwBl9E15ssskJVFw0gZqjh5D51kcUPLoU06TTiaHm29vkWvUB5ZdPUmcXEQm7hlEFbP/msVSePpaUzVWU/OYNcp5dhalXGiKxpb1e+3+AEcCL3oYTEXpSk90+3ToTaAVeCV7fAhSEMqi44DNUnXEolaePxb+qjKI/LMRX2+h1VLGjpZWCR5aSWFVPxaUTaclP8zoiEYlXCT72HDOUrXNmUTtxIBnzP6L/baWkv71ZZzIl5tQBm4Ev4xLGNd6G46nuJtnLgYuMMWnA1cDr1tr2jHAQUN4XwcWDPccMpeKiCSRtr3Et/ipqvQ4p+llL3pMrSP2okp3nHkbjkFyvIxIRoS09hcovjWf7jcfQXJBG3j9XUPy7N0nZWOl1aCIhVwe8iaspvhaIx3d5d5Ps7wNfBGpwf5j8uMN9ZwOLQhxXXKkfV0zZNVPxNbRQfO989IHbO1mvfkD6sq1UnzCKvUf09zocEZHPaO6fRdnXplFxwZEk1DVRfN8C8v+6jITqeq9DEwmp9v7aDwGDgbuAeNobtbub0byJm7GeDAy21nZMqh/ELYyUXmgalMOO66fTFkim6I+LCLy3zeuQolLasi1kv/IBtRMGsPu4EV6HIyLSOWPYe3gJ226eSfUXRuJfuYOSX5eS9coHmGat0ZHY0l6v/X1cvfZL3oYTNj3ZjGaPtXaptbZ6n9uft9auC31o8aclL40d102ncUAWBY8tI3PuBrV86oGUD3eR94/3aBiWx64vjlerPhGJeDY5kd0njGLbt2dSP6aQ7FfWUfLruQRWbNfnv8ScOmAT8CVgNrDW23D6XI82o5G+15aWTNlVU6g7rB85L6wh91/vQ2ub12FFvMSKWgoeWUpLboCKiydAot7aIhI9WnMD7LxoAju+OoW21EQKHn2Hoj8sJGl7jdehiYTcXmAecCRwPVDlbTh9RplIJEpKYOf5R7J75jAyFm6m4JGlmMZ4qmLqGV9tI4V/Wgw+Q/nlk2kLJHkdkojIQWkcns/2rx/DrrPHkbRjD/3ueoPcp9/HV9fkdWgiIdVer/0nXL3274i9em0l2ZHKZ6g+5RB2nT0O/9pyiu5fgG9Pg9dRRZ5m16ovoaaB8ssm0pIX8DoiEZHeSfBRO3Uw2+bMYs/UwaS/vZmS20rJmL9RZzYl5jQAe4DvAiOBl70NJ6SUZEe42qmDqbhsIkkVdfS7ez5JZdot7BNtlvy/Lyd1UxW7vnIETYNyvI5IRCRk2gLJVJ01ju3fOJam/pnkPrOSfne9Ser6nV6HJhJydcBGXMu644BYWOynJDsK1I8pouyaadDa5lr8bdAHLED2y2tJe287VaeMYe/4fl6HIyLSJ5qLMyi/agrlF0/ANLdQ9MdFFDyyhMTKvV6HJhJy7fXahwM3ANX7Pzyi9TrJNsbMMMa8FopgpGtNA7LYcf10WjJTKXrwbdKWbfE6JE+lLf6YrNc3sGfyQGpmDPM6HBGRvmUM9eOK2XbTTKpOGk3qup2U3D6X7JfWYppirZJV4l0rrozkQVy99j1EZ712KGayC3Ab1Egfa80JuBZ/g3PJf2I5Wa9+EJctnlLX7yTvqRXUj8yn8qxxatUnIvEjKYGa2SPYNmcWdeOKyXp9PSW3zSVt2da4/H0gsa0BtwvifwOjgFe9DafHEru6wxhzaTefY1KIYpFusP4kyq6cTN6T75H98joSquqp/OI4SIiPyp+ksj0U/GUpzQXpVFx0VNx83yIiHbVmpbLr/COpnTaYnGdWkf/Eu6Qv3ETVGYfSNCDL6/BEQqoO+Ag4E5gG3ItbJBnpukyycbtgWqA704T68zmcEn3sOu9wWnL8ZL+2nsTd9VRcdBQ2NbZb1/n2uFZ9bUkJlF8xKea/XxGRA2kcnMuOG44mbekWcl5aQ/Hdb1I7YSDVJ42mLSPF6/BEQmovUIqr174KuB2I5Exgf9OAO4A/AhkHuHR3xltCyRh2nzianeccRuqGXRTft4CE3fVeR9VnTFMrhX9egq+uiYrLJtKa7fc6JBGRyOAz1E0ayNY5s6g5Zijp72yh/22lZMz7EFrU8k9iSyuuv/aDwNsex3Ig+0uyFwATrLV1+7vgvlfxSN2kgZRfPonEqnqK754fm7uDtVnynniX5C3V7Dz/CJoGZHsdkYhIxLGpSVSfNpZtN82gcUgOuc+vpuTOeaSuLfc6NJGQS8JtaBPJ9pdkPwF82I3nWAX8JDThyMFoGFXAjmunAVB83wJS11V4HFFoZb+4hrSVO6g6bSz1hxZ7HY6ISERrKUin/IrJlF0+CSwU/WkxBX96m8SKWq9DE4krXSbZ1tq/WWvPO9ATWGtXW2t/HNqwpKea+2Wy44bptOT4KXxoMWmLP/Y6pJBIX7iJrHkfUjNtMHuOHuJ1OCIiUaNhTCHbvjWDqlPHkLqxipLfzCP7+dWYhmavQxOJC2rNEENas/zsuHYaDcPzyH/yPbL+szaqWzqlri0n95mV7B1dQNXpY9WqL0olAgHADyR4HItI3En0UTNjOFvnzKTuyP5kvvEh/W8rdRMxbdH7+0EkGnSZZBtj/mOMGb3PbccZY9L6Piw5WDY1ifLLJ7Fn4kCyX1tP3t+WR+XCl6TtNRQ8tozmogx2XqhWfdEqDZgFrAFWACfgEm4RCa+2jFR2fflwdtxwNM25AfKffI/ie94ieVOV16GJxKz9ZS7HA5802zTGJAAvA6O7fEQ3GGNONsasNcasN8Z8t5P7LzfGVBhj3g1eru7N68WlBB+V54yn6sRRpC/bStGDb2Pqo+f0YEJNA4UPLaYtJZHyyydhU/bXaVIiUSqQCzyC+9AYCAwHXgheRuMScBEJr6YB2ZRdN52dXzmchJoG+t07n7wn3iWhpsHr0ERiTk+nB3t1vj6YqN8NnAKMBS4wxozt5NAnrLVHBC9/7M1rxi1jqDluJDu/cjgpmyopvnc+CVV7vY7qgExjCwUPLcZX30z55RNpzUr1OiTpIT9wCW7jgC92cv8M3Grpe3CJuJoxioSZMdQdOYBtN89i9+zhpK3YTsltpWS+vh6aW72OTiRmhPsc/GRgvbX2Q2ttE/A4cFaYY4grdUcOoOzKKSTUNFB8z3ySt+z2OqSutVnyH19G8vYaKi48iuYS7VoWTQLACNxGAfcDmfs51odrsL8ZuBmXaEfyhgIiscimJFJ90hi23TSThhH55Ly0lpI75uFfuSOq1/OIRIoDJdmd/ZT15ievP9Cx7cWW4G37OscY854x5h/GmIG9eD0BGofnUXbddEjwUfT7BfjXlHkdUqdynltFYHU5lWceSsOYQq/DkW5KwCXY/wOsxv0l3V1pwE+B9cCXcMm2lreKhFdLXoCKSydSdtUUbJKPwkeWUvjA2ySV7fE6NJGoZmwXf60aY9qAaqClw835ndwGgLX2gFmRMeZc4CRr7dXB65cAk621X+9wTB5Qa61tNMZcC5xnrT2uk+e6BrgGoKioaMLjjz9+oJcPue3AtrC/6sGra2zjuaWNVNS0MXNsMuMHfTp3OKC2li3p6Z7FtnxjM/PWNHHE4ESOPSQ2tgL2ekzDwYdLsIcCySF4vr3AJqCBz28yEA/jGW4a09CKhfFsa7Os+LiFRR800dQK4wcmMmVkMqlJ3vz5GwtjGmliZUwTcGdPvfhOZs+evdRaO/FAx+0vyf5RT16wO72yjTHTgFuttScFr38v+Nj/7eL4BKDSWrvfuoGJEyfaJUuW9CTckPgpcEvYX7V3TGML+X9dRmBNObtnDKP65DHgM9xWWsqcWbM8icm/qoyCR5ZQf0gRFRdPAF9szGV6OaZ9LQU36/x74FxCO/tsgX8B1wM1QF3w9lgeT69oTEMrlsbTV9dE9strSV+0mTZ/EtUnjqZ28qCwfz7H0phGilgZ0yzgWeBYD17bGNOtJLvLtg19tMHMYmCkMWYosBU4H7iw4wHGmH7W2u3Bq2fizkBLiNiURCoumUDusyvJmvchidX17Dz3cM/iSd66m/y/LqOpfxY7zz8iZhLsWOYHzgN+A/TFBvcGOBs4FfgtcCsQPb1xRGJDW1oylWePZ8/kweQ+u5K8p98nY+EmKs88lMZheV6HJxIVwtobzVrbYoy5EXgJN9P/oLV2pTHmJ8ASa+0zwDeMMWfiSlIqgcvDGWNcSPBRedY4WnIC5LywhoSaBuqHhX+RS0J1PQUPLaYtLZnyyyZik9WqL5IFgELgUWB6GF4vGbco8nLg+7jSFB+fLyMRkb7TXJJJ2TVTCazYQc7zqym+fyF14/tRdeoYWnPU9V5kf8Ke1Vhrnwee3+e2Wzr8+3vA98IdV9wxhpqZw2nJ8ZP/t+X8o9ySOKGOlrzwdC82Dc0UPrQYX1MrO66bQluGWvVFKh+uPOQ7wHcJTe11T+ThylJewm1ssxBXuy0iYWIMew/rR/0hhWTO/ZDMuevxry6jZuZwamYOxyZrL1eRzmgbvTi397ASyq6eQkOzdS3+Nodh96/WNgoeW0ZSeS0VFx1Fc3FG37+mHJQ0YArwHm79QbgT7I5SgFeBZ4BhaDMbkXCzSQnsPn4k226eRf3YIrJf/YCSX5cSWL5NLf9EOqEkW2gcksuXp/ppS0mk6A8L8b+/o+9ezFpyn1mJf10FlWePo2FUQd+9lhy0ZFyf6/uAt3AruCPFF4B1wB24mnBtZiMSXq3ZfnZeeBQ7vjaNtkAyBX9dRtHvF5K0NYL3YRDxgJJsASAnzceO66fT3C+TgkeXkvHmR33yOhlvfETGos3snjncrVSXiOMHzgE+BC4mMvtWJwBfxW1mcyMuZlX0i4RX49Bctn/9GHZ9cTxJFbX0+92b5P5zBb7aRq9DE4kISrLlE23pKZR9dSr1Y4vIfW4VOc+shLbQnQL0v7+dnBdWUze+mOqTRofseSU0AsAA4AXgMVwtdKTLAH4JrAHOQLPaImHnM9ROGcS2m2exZ/oQ0pd8TMltpW6iplXLlCW+KcmWz7BJCVRcNIGao4eSOX8jBX9Zimlq7fXzJn9cTf4T79I0MJtd56lVXyTx4ZLTm3A7L870NpyDMgj4J/A6cBiq1xYJt7ZAElVnHMr2bx5L08Bscp9bRb873yB1XYXXoYl4Rkm2fJ7PUHXGWCrPGIt/dRlFf1jYq9N/CZV7KXx4Ma0ZKZRfOhGbpJXokSINOAp4B/gZbnFhNJsCvAv8CdduUA3GRMKruSiD8isnU37pRExrG0UPvk3Bw0tI3FV34AeLxBgl2dKlPUcPpeLiCSTtqKH4nvkkVtT2+DlMvWvVR0sb5ZdPoi092tO42JCMK7W4E3gbGONtOCFlcLtQbgJ+gEu0veyKIhJ3jKF+bBHbbppB1cljSN2wk5Lb55H94hpMY4vX0YmEjZJs2a/6Q4sp++pUfI0tFN87n5SNld1/cGsbBY8uJWlnHRWXTKClUK36IkEAV7+8AbiKyFzYGAqpuE1sPsJtK+tHH3giYZWYQM2s4WybM4u6w/uRVbqBkttKSVu6JaTrfUQilX7nyAE1Dcphx/VH05aWTNEfF7meqAdiLblPvY9//S52nXMYjcPz+z5Q2a8A0A/4F/APIF6aJxbiykcW43aqVL22SHi1Zqay67wj2H79dFqz/eT/fTnF984n+eNqr0MT6VNKsqVbWvIC7LhuOo0Dsij46zIySzfsd/OBzLkbyFjyMdXHjaBuwoAwRir7MrhZ3Btws9fHexuOZw4F3sD9gTEYJdsi4dY0KIcd101n57mHk1BdT7+73yLv78tJqGnwOjSRPqEkW7qtLZBM2VVTqDusHzkvriH36fc7bdEUWL6NnBfXUndECbtPGOVBpNIuDRiPm8X9JWpxB3AyrovK/+E23En1NhyR+OIz1E0YwLY5s9g9cxhp726l5LZSMudugJbed7ISiSRKsqVnkhLYef6R7J45nIxFmyl4ZOlnFrKkbKok/+/LaRiSw85zDgMTqxW/kS0Jl2D/CliGm8WVTyXiZvY3A9eizWxEws2mJFJ9yiFsu2kmDcPzyHlhDSV3zMO/ukxbtEvMUJItPeczVJ8yhl1fHId/bTlF9y8goaaBxF11FPx5KS1ZqVRcMhHUqs8TAT6drb0O/ZDvTxZue/b3gZNQyz+RcGvJT6PiskmUXTkZfIbCh5dQ+KfFJJb3vJuVSKTR5I0ctNopg2nJ8lPw2DsU3zMfm+gDaym/YjJtaWqaFm5+XFu+h4BTvA0l6gwDngPexG3X/jGgrr4i4dMwqoBt35pBxoJNZL+yjpLfzGPP9CE0pmhWW6KXJrmkVxrGFFL2tWnQ2kZiVT0Vl0ykJV9LysKpfWHjV4EPUYLdG8cAK4H7gFw0sy0SVgk+9hwzlK1zZlE7cQAZb33EI2/sJf3tzWr5J1FJSbb0WlP/LLZ/81i2f/0YGofmeh1OXEkDDgEW4DaW0Z83vecDLsbNZv8X7g+YJE8jEokvbekpVH7pMHbceAzZaT7y/rmC4t+9SY/2aRCJAEqyJSTa0lNoLtZmM+GSiEuofw68BxzubTgxKQDciqtt/zIu2dYyXpHwaeqfxTmTU6m44EgS6poovm8B+X9dRkJ1vdehiXSLarJFokwarqzhj4A6kPe9EuAx4F3gGmAVqtcWCRdjDHsPL6H+kEIySzeQOe9D/KvKqJk1nJoZw7BaYC8RTDPZIlEiFcgDHgVeRAl2uB0BLMIl3CWoNEcknGxyIrtPHM22b8+kfnQB2S+vo+T2uQRWbFfLP4lYSrJFooAfuBz4CDjL21DimgHOxP0//ARIB1I8jUgkvrTmBth58QR2fHUKbSmJFDz6DkV/WETSjhqvQxP5HCXZIhEsDRgFzAPuxbXoE+8lA98GNgFX4s4y6KS1SPg0Ds9n+9ePYddZh5K0o4Z+d75B7tPv46tr8jo0kU8oyRaJQIl8uvBuJTDR02ikK7nAPcByYDZq+ScSVgk+aqcNYducWeyZOpj0tzdTclsp6Qs2Qmub19GJKMkWiTQB4FjcArs5aHVyNBgFvAw8C4xE9doi4dQWSKbqrHFs/8axNJVkkvevlfS7601S1+/0OjSJc0qyRSJEKpADPAy8Cgz2Nhw5CMcBq4G7cP+Xfm/DEYkrzcUZlF89hfKLj8LX1ELRHxeR/8hSEiv3eh2axCkl2SIRwA9ciFtQ92XUjzmaJeDqtDcD30Sb2YiElTHUj+vH1m/PpOrEUfjXVVBy+1yy/rMW09TidXQSZ5Rki3goAAzDzVw/AGR5G46EUDrwv8Ba4Ay0mY1IWCUlUHPcSLbNmcneccVkv7aektvmEnh3q1r+SdgoyRbxQAIu6fo+sAaY5m040ocGAk8Cc3E7c6peWyR8WrP87Dz/SHZcO43WjGQKHn+XovsWkLxlt9ehSRxQki0SZmnAVOB94AeolCBeTALewdXcD8H9v2ehha0i4dA4JJcdNxzDrnPGk7SrjuK73yT3H+/hq230OjSJYfp8FwmTFNzixnuAC1DpQDwywDnBSxXwNrAAeAW3bbvFneXY41WAIrHMZ6idNIi68f3IfuUDMuZvJG3FdqqPH8meaUMgUfOOElpKskXCwI9LrO7E9VYWyQFOCl5uxSXYG3Bbt8/DlZd8iHvvNAENnkQpEntsahJVp49lz5RB5D63itx/rybj7c1Unj6WhtGFXocnMURJtkgfCgD5wF9wva9FumKAEcHLRcHbGnAz3Itwi2MXAdW4MyK1gLbbEDl4LQXplF8xGf+aMnKeW03Rnxazd0whVaePpSVfqyek95Rki/QRH3Azru46xeNYJDql4ur3p+LaAQJU4JLt+bjEewXuvWZwibeI9Ez9mCLqRxSQOf8jsl5dT8kdc6k5eii7jxuBTdWqGTl4SrJF+kA6bhfA07wORGJOAXB68AJuNnsdsBBXYvIGsAl3FqUxeBGRA0j0UTNjOLVH9ifnxbVkzfuQ9He2UnXyaOqOGgA+raKRnlOSLRJiabgttlVDK+HgA8YEL5cHb9uL62SyADfbvRiow51RUZmJSNfaMlLZde7h7Jk6mNxnV5L/j/fIWLiJyjMPpWlQjtfhCWAamvF/sJPUHXvghFFeh7NfSrJFQigN1yliKlDqbSgSxwLAMcHLfwVv244rM3kTeA1YhUvQ03AJuIh8qmlgNjuunU7a8q1kP7+GfvfMp/bI/lSfMobWzFSvw4sv1pJUXot/TTn+teWkbKzCtFnaAknUHzsUIrikR0m2SIi0z2BP9ToQkU70A84OXgBagf8AvwFexyXf23HdTOqBZg9iFIkoPkPdkQPYO7aYrNfXk/nGRwRW7mD37BHUHDMUkhK8jjBmmaZWUj/c6RLrNRUkVtcD0NQvk5oZw6gfU0jqwGz8CZHddlFJtkgIpOESFu3cKNGifdfRq4MXcP25l+Dqu18BluLaBybiyky0GbXEI5uSSPXJY6idNJCcf68m56W1pC/+mKrTDqF+bBEY1WuHQuKuvfjXluNfU07qh7swLW20JSfQMCKf3ceNoH50Aa1Z/k+Oj4bzCUqyRXopDXgJmO51ICK9lAHMDl6+h0uqP+azZSbrcLXdLbgZb5F40ZKXRsWlE0n9oIKcZ1dR+MhS6kfmU3X6WJqLMrwOL/q0tJG6sfKTMpCkCle41pyfxp4pg6kfU0jD0BxIjN4zBkqyRXohDXgRONrrQET6gAEGBS/nBm9rxrUNXIRLuhcA5bg68Dpc8i0SyxpGFrD9m8eSsXATWS+vo9+db7Bn6mB2Hz+KtkDk1gdHgoSahk9nqz/Yia+pFZvoo2FYHnumDqZ+dGFM9ShXki1ykALAC7jFZSLxIgk4Kni5LnhbFa6DSXs3k2W4DibaIl5iVoKPPUcPpe7wErJfXkfGgo2kvbuV6hNHUzt5kFr+tWuzpHxchX9NBf415SRvrwGgJSuVuiP7Uz+6kIYRedjk2ExHY/O7Eulj7Qm2dnEUcVvEnxi8/AhXZvIhn90ifgPaIl5iT1t6CpVfHO+2aH92FXlPv0/Gos1UnjGWxmF5XofnCV9dE/51U8kFzQAAIABJREFUFfjXlpO6roKEvc1Yn6FxcA5VJ4+hfkwhzUXpcVHLriRbpIfSgOeBGV4HIhKhDDA8eLkweFsjsBy3qPK14NcqtEW8xIbmkizKrplKYMV2cv69muL7F1J3WD+qThlDa07A6/D6lrUkbashsKac1LXlpHxcjbHQmp7sdtMcU0D9yAKsP/5KaZRki/RAAPg3SrBFeioFmBy8fCN4207cbPcPcQsq93oTmkhoGMPew0qoH1NE5twNZM7dgH9VGTUzh1Mzczg2OXoX8O3LNDSTun5nMLGuIHGP21u2cUAWu78wkvrRhTT1z4r7shkl2SLd1J5gz/Q6EJEYkQ+chisz+R5wD+pYItHPJiew+4RRruXf86vJfvUD0pduoerUMewd3y86yySsJbGiFv+aCgJryknZWOk2hElNpH5UAdWjC6kfVUBbRorXkUYUJdki3RAAngNmeRyHSCxKAm4DTgC+grqUSGxozfaz88Kj2DN1F7nPrqLgsWU0DN1E5RljaS7J8jq8AzLNraRs2OW6gawtJ6kyuCFMUQY1xw6jfnQBjYNzIMI3hPGSkmyRAwgAz+J6B4tI3zkJt9372cBKVD4isaFxWB7bv34M6W9vJvs/a+n32zepnTyI6hNH05aW7HV4n5FQ2WFDmA278LW00ZaUQMOIPGpmDqd+dCGt2f4DP5EASrJF9isAPAMc53UgInGiBNcK8BbgDlQ+IjHCZ6idOpi9h5WQ9co6MhZuIrB8G7tPGMWeqYO9mw1uaSNlUyX+tcEWe+W1ADTnBaidPCi4IUyutpA/SEqyRbrgB/4FfMHrQETiTALwc9zP3pdx3UeaPY1IJDTaAklUnXkotVMGkfPsKnKfXUX6os1UnTGWhpEFYYmhrqGNtMUfuxnrD3bia2zBJhgahuZROXkQ9aMLaClID0sssU5Jtkgn2hPs470ORCSOHQesBr6Ea/9X5204IiHTXJRB+VWT8a8qI+ffqyl64G32ji2i6rRDaMkL8Y6HbZbkLdWfbF/+4NZ68nmPlsxU6g7vF9wQJh+bopQw1DSivXAC8Evc6cxWj2OR0AkAT+H+f0XEW0XAG8DPgF+g8hGJIcZQf2gx9aMKyHzzI7JeX0/J7fOoOXYou2eP6FXS69vbROo6VwLib98QxkDj4BymjUziyVOn0lycEZ2dTqKIkuxemAqsAE4HPkKLdGJBe4J9oteBiMgnfLga7dnAF3FbtTd5GpFICCUlUDN7BHVHDSD7xTVklW4gbekWqk8ZQ90R/bvXa9pakrbXfFJbnbK5ym0Ik5ZM/ehCV1s9Mp+2QDITS0t5vF9m339foiS7t4YA7wDfBv6EEu1oFgD+iRJskUh1LLAGOBdYjMpHJLa0ZqWy6ytHsGfqYHKfXUn+35aTsXATlWccStPA7M8dbxpbSP1g5yct9hJrghvC9M9i9+wR1I8ppGlAdtxvCOMlJdkhkAz8DpecXYxLtFU+El38wJO4FmIiErnygVeBXwE/RuUjEnuaBuew4/qjSXtnCzkvrqXf3W9RO2EAVSePxtfQ8kltdepHlZhWS1tKIvWj8t2GMKMLaMtI9fpbkCAl2SF0Jq585AxgA5rVjhbtCfbJXgciIt3iA76D2331TKAGaPQ0IpEQ8xnqJg5k77hisl5bT+ZbH5G2bCumzQLQVJhOzdFDqR9dSOMQbQgTqcKeZBtjTgbuxHVp+qO19hddHPdl4O/AJGvtkjCG2CuDgaXAzcADKNGOdH7gH8ApXgciIj02FVgLnA+8hcpHJPbY1CSqTz2E2kkDyVi4meb8gNsQJjfgdWjSDWH908cYkwDcjctpxgIXGGPGdnJcBvANYFE44wuVJOAu4G9ABjpdEKn8uP+jU70OREQOWg7wIq77iPahk1jVUpBO1RljqZ02RAl2FAn3+YXJwHpr7YfW2ibgceCsTo77Ka47XkM4gwu103BbAx+KW1QnkaM9wT7d60BEpNcM8C1cq79+QIq34YiIAOFPsvsDH3e4viV42yeMMUcCA621z4UzsL4yELcK/mso0Y4UfuAJlGCLxJoJuM1rTkCftyLiPWOtDd+LGXMucJK19urg9UuAydbarwev+4DXgMuttRuNMaXAnM5qso0x1wDXABQVFU14/PHHw/RdHLwa4EOgDQjfqHfPgNpatqTH/jaqPmAYkBWG16qtrSU9DsY0XDSeoRfLY1qBm8VpC+NrxsvnaDhpTEMvVsY0ARgBePGdzJ49e6m1duKBjgt3ufAW3ORuuwHAtg7XM4BxQKlxuxAVA88YY87cN9G21t4P3A8wceJEO2vWrD4MO3S24FbDryWyFkXeVlrKnCgZw4PlB/5K5/VJfaG0tJRoeV9GA41n6MX6mL6HK9vbSXhqD+PhczTcNKahFytjmgU8i+ufH6nCXS6yGBhpjBlqjEnGLQp/pv1Oa+1ua22+tXaItXYIsBD4XIIdzQYAbwPXoUU64eQHHiN8CbaIeO8wXPnI6ah8RETCL6xJtrW2BbgReAn32fc3a+1KY8xPjDFnhjMWLyUCt+G2785C3Uf6mh94FDjb60BEJOzScYuc78Il2tr7TkTCJez5nbX2eeD5fW67pYtjZ4UjJq+chOs+chZuq2D1eA09P/AI8EWvAxERzxjgKmAarnxkB1HeukpEooK2CPJYf1xNzA2ofCTU/MCfgXO8DkREIsJY3MTGOah8RET6npLsCJAI/B/wL1z5SJK34cQEP/Aw8GWvAxGRiBIA/gL8Pvhv/RIUkb6iz5cIcgKwCjgCzbL0hh94CDjX4zhEJHJdDCzDtfTUWUQR6QtKsiNMCbAA+Cb64D8YfuBPwHleByIiEW8UsAK4AE1siEjoKcmOQAnA/8P1f8xG5SPd5QceBL7idSAiEjVSgQdwf5ynoV+KIhI6+jyJYF/A9Tk8CvfhL13z435Rnu91ICISlc7DbV4zGp1FFJHQUJId4YqBt4Cb0Ad/V/y4rT8v8DoQEYlqw3B12pej8hER6T0l2VEgAfgp8G8gB5WPdOTHdQm42OtARCQmpAD34HaIzcB9/oqIHAwl2VFkNq58ZBKaZQGXYN8HXOJ1ICISc87CLYo8FH3eisjBUZIdZYqAecAc4rt8pD3BvtTrQEQkZg0GlgDXEN+ftyJycJRkR6EE4MfAC0AukOxtOGHnx53OVYItIn0tCbgDeBLIxG0eJiLSHUqyo9hMXPnIZOKn+4gfuBu3MElEJFxOwW0WdhgqHxGR7lGSHeUKgbnAfxP7pzP9wO+AK7wORETiUn9gEXAjsf95KyK9pyQ7BviAW4CXiN3yET/wW+BKrwMRkbiWCPwf8AxuszCVj4hIV5Rkx5BjgTXAVGKrfMQP3AVc5XUgIiJBx+PK9Sai8hER6ZyS7BhTALwOfI/YOJ3pB+4ErvY6EBGRfRQDb6JuTyLSOSXZMcgH/AB4GcgjestH/MBvgK96HYiISBc6dntKJHo/b0Uk9JRkx7CjceUjRxN95SN+4HZcf1oRkUg3E7dxzRSi7/NWRPqGkuwYlw+8AvyQ6Dmd6Qd+DVzrdSAiIj2QCJQC3yd6Pm9FpO8oyY4DPuC7wKu4pDuST2f6gV8B13kdiIjIQfDhkuxXcGtkIvnzVkT6lpLsODINWIvrQhKJpzP9wC+BG7wORESkl6bjyvVmEJmftyLS95Rkx5lc3ILIW4is05l+XO/ZG70OREQkRHKB/+AWRkbS561ILGgC+nkdxAEoyY5DBrdD5GtExulMP/AL4OsexyEiEmoGuBm3M28xkOJtOCIxIYDbAXqE14EcgJLsODYVVz4yE+/eCAHgf4FvePT6IiLhMAlXPvIFVD4i0hsB4JtExw7QSrLjXA5uO/YSwn86MwD8HPfDIiIS67KA53Bn7lQ+ItJzfuBUXO4QDZRkCwYowp3OLCI8pzMDwE+Ab4XhtUREIoXBrT2ZD/QHUr0NRyRqJAPjgEdxP0fRQEm2fKL9dOYs+vZ0ZgC3EOjmPnwNEZFIdgSwGjgF95koIl3z4dY0vIT368h6Qkm2fEY2bnvgn9E3pzMDwK3AnD54bhGRaJIBPAn8BvfZGC2zcyLhlonb6CnH4zh6Skm2fI7BlXHMI7Sr4QPAj4D/CtHziYhEOwN8FVgEDETlIyL7CuBmsId6HchBUJItXZqIO515HL0vHwngenP/d2+DEhGJQeOAVcCFuFm7AJDuaUQi3vPjarAnex3IQVKSLfuVDfwb+H8cfPlIAPgh8J1QBSUiEoPSgAeAauB94D7gCtwMXhIu+U70LDqR8ArgSlfP9jqQXtDPqxyQwfWxPgY4HagEGrv52ADwA+B7fROaiEjMMbjEeihwUfC2GuBt4A3cLpLv4n6BtwF7PYhRpC8FgEuAb3sdSC9pJlu67Shc95ET6F75SACXXH+/L4MSEYkDmcDxuM5MC4BaXBvA24Ev8Wn71Uz0i12iWyowHbjb60BCQD+L0iOZwDPA/7H/8pEA8F1cmYiIiIRWAjAe+BquQ8kOYBPwMO7M4zhcq7NMtJW7RI8k3Bmcp3Hv8WinchHpMQPcAByNKx/ZBTR0uD+Aq7/+n/CHJiISt4pw9avtNayNwDLgTVx3hreBFtzsWq0XAYrsh8G16HuNvt2rI5w0ky0H7QjcaviT+PQHIoBr0XeLV0GJiAjgZrCn4vYleJnPLqi8Ei2olMiSBryOax0cK/RzJb2SCTwF/B53ivJm3GYzIiISWbSgUiKVH/gXMNbrQEJMSbb0mgGuBS5GfV1FRKJJ+4LK9kWVrbgzlPNxs99vAVW4WfFaXPItEkoB4He4PTlijZJsCRkl2CIi0a19QWX7okqAclzSPRd4BViH6wDRSPfbuYp0JoDbYfoKrwPpI0qyRUREpEuFdL6g8i3gRbSgUg5OADgVt+FMrFKSLSIiIt3WvqByKm4djgU24ma7XwNKgY9xdbZ7cQm4SEcpuDaTj+JKTmOVkmwRERE5aFpQKT3hw3UQeRHXyz2WKckWERGRkOpsQeVqXImJFlTGt0zc2Y4cj+MIByXZIiIi0qcScOUB4/jsgsoFuAWVLwNrcSUmWlAZuwK4MxtDPI4jXJRki4iISNgVAmcFL/DZBZX/ARahHfNiiR94DJjkdSBhpPeviIiIeK59QeXNuG3gq4BhuPawsbw4Lh4EgJ/z6R9U8UJJtoiIiEQcA2QBS4ABuCRcok8AuBS4yetAPKAkW0RERCLWaGAFcDQuYZPokYr7f/ud14F4REm2iIiIRLQs3OLIG3C1vRL5knDlPk/hFr7GIyXZIiIiEvF8wC+Bh9CMdqQzuBZ9rwFpHsfiJSXZIiIiEjXOw+0uWYSbLZXIk4brhV3kcRxeU5ItIiIiUeVw4H3gKDSrHWn8wDPAIV4HEgGUZIuIiEjUyQfexHWuUKIdGQLAPcBsrwOJEEqyRUREJColAvfiuldoQaS3Arg2fZd7HEckCXuSbYw52Riz1hiz3hjz3U7uv9YYs8IY864x5k1jzNhwxygiIiLR4wpcDXAu2sraC37gNOCnXgcSYcKaZBtjEoC7gVOAscAFnSTRj1lrx1trj8AtJL49nDGKiIhI9JmMq9Mei2a1wykZOAz4C9qZc1/hnsmeDKy31n5orW0CHmefXTattTUdrqYBNozxiYiISJTqB7wNfAnVaYdDAm7MX8Ql2/JZ4U6y+wMfd7i+JXjbZxhjbjDGbMDNZH8jTLGJiIhIlEsBHgF+gWa0+1omrkwn2+M4IpWxNnwTxcaYc4GTrLVXB69fAky21n69i+MvDB5/WSf3XQNcA1BUVDTh8ccf77vA40BtbS3p6elehxFTNKahpfEMPY1paGk8Q6+3Y1oLrAdaQxZR9BtQW8uWELxPfbgt7+PxjMHs2bOXWmsnHui4cK8P2AIM7HB9ALBtP8c/jls4/DnW2vuB+wEmTpxoZ82aFaIQ41NpaSkaw9DSmIaWxjP0NKahpfEMvVCM6SbgRGAz0BCCmKLdbaWlzOnlmPpxCdqpoQgohoW7XGQxMNIYM9QYkwycj+tZ/gljzMgOV08DPghjfCIiIhJDBgPLgJOIz1nXUAsA/w840+tAokBYk2xrbQtwI/ASsBr4m7V2pTHmJ8aY9v+vG40xK40x7wLfBj5XKiIiIiLSXQHgKeAHqE67NwK4pOxbXgcSJcLeTtJa+zzw/D633dLh398Md0wiIiIS2wzwfdxW7OcCe4E2TyOKLqnAMcBvvQ4kimjHRxEREYkbJwPv4BaFpXgcS7RIAobhzgYkeBxLNFGSLSIiInFlJLACOBbVaR+Iwe2k+Roaq55Ski0iIiJxJxO3QOwbqE57f9KB14EirwOJQkqyRUREJC75gP/FbQkeQNuC78uPawF3iNeBRCkl2SIiIhLXvgQsxM3WantwJwDcB8zyOI5opiRbRERE4t54YCUwAdUepwE3A5d6HUiUU5ItIiIiglvgNw+4gvhNtAPAGcCPvQ4kBijJFhEREQlKBH4H3EP8LYhMAQ4DHkb16aGgJFtERERkH5fhZrXz8GDnPg8kAMXAC6guPVSUZIuIiIh0YiLwPjCO2J/VzgTmAtleBxJDlGSLiIiIdKEYWITbij1W67QDwH+AwV4HEmOUZIuIiIjsRzKuTvmXxN6Mth94HDdrL6GlJFtERESkG27A7RKZhathjnYB3GY8Z3gdSIxSki0iIiLSTccCK4ARQKrHsfRGALgc+KbHccQyJdkiIiIiPTAQWAacSnTWaftxfyzc5XUgMU5JtoiIiEgP+YF/ALcQXXXaBhgG/JPYKHmJZEqyRURERA6CAb4DPA1kEPlJlcH1/H6V6JyBjzaR/n4QERERiWgnAu/gykhSPI5lf9KBUUCR14HECSXZIiIiIr00ArcgciaROUscAJ4luhdrRhsl2SIiIiIhkIHblvwmIqtOOwDch/sDQMJHSbaIiIhIiPiAnwGPAWm4OmgvpQFzgEs8jiMeKckWERERCbGzcdux98PtGOmFAG6jmVs9ev14pyRbREREpA8cCrwPTCb8ddopwOHAn/F+Nj1eKckWERER6SM5QClwNeFLtBNwM+gvAElhek35PCXZIiIiIn0oAbgTt/gwHIl2JjAXyArDa0nXlGSLiIiIhMElwDwgn76bYQ4ALwOD+uj5pfuUZIuIiIiEyQRgJTCe0M9q+4HHg68h3lOSLSIiIhJGhcAC4CuELtEOAL/AdRORyKAkW0RERCTMkoEHgdvp/cY1AeAK4Bu9DUpCSkm2iIiIiEe+hquhzgYSD+LxqcAM4K5QBiUhoSRbRERExENHAyuAkfRsVjsJGAE8iRK6SKT/ExERERGPDQCWAqfhtkI/EAPkAa8S/o1upHuUZIuIiIhEAD/wN+BHHHhGOx14HbeIUiKTkmwRERGRCGGA/wKeATLoPFHzA88BY8IYl/SckmwRERGRCHM88C4wGLe4sV0A+ANusaNENiXZIiIiIhFoGPAeMBtXp52Gm+W+yMugpNsOpluMiIiIiIRBOq405GfADly9tkQHJdkiIiIiEcwH3OJ1ENJjKhcREREREQkxJdkiIiIiIiGmJFtEREREJMSUZIuIiIiIhJiSbBERERGREFOSLSIiIiISYkqyRURERERCTEm2iIiIiEiIKckWEREREQkxJdkiIiIiIiGmJFtEREREJMSUZIuIiIiIhJiSbBERERGREFOSLSIiIiISYkqyRURERERCTEm2iIiIiEiIKckWEREREQkxJdkiIiIiIiFmrLVex9BrxpgKYJPXcUS5fGCn10HEGI1paGk8Q09jGloaz9DTmIaexrT3BltrCw50UEwk2dJ7xpgl1tqJXscRSzSmoaXxDD2NaWhpPENPYxp6GtPwUbmIiIiIiEiIKckWEREREQkxJdnS7n6vA4hBGtPQ0niGnsY0tDSeoacxDT2NaZioJltEREREJMQ0ky0iIiIiEmJKsuOIMWagMeZ1Y8xqY8xKY8w3OzlmljFmtzHm3eDlFi9ijSbGmI3GmBXB8VrSyf3GGHOXMWa9MeY9Y8xRXsQZDYwxozu89941xtQYY761zzF6jx6AMeZBY0y5Meb9DrflGmNeNsZ8EPya08VjLwse84Ex5rLwRR25uhjPXxlj1gR/pp8yxmR38dj9fj7Eqy7G9FZjzNYOP9undvHYk40xa4Ofqd8NX9SRq4vxfKLDWG40xrzbxWP1Hu0jKheJI8aYfkA/a+07xpgMYClwtrV2VYdjZgFzrLWnexRm1DHGbAQmWms77Tsa/EXxdeBUYApwp7V2SvgijE7GmARgKzDFWrupw+2z0Ht0v4wxM4Ba4M/W2nHB234JVFprfxFMTHKstd/Z53G5wBJgImBxnxETrLVVYf0GIkwX43ki8Jq1tsUY838A+45n8LiN7OfzIV51Maa3ArXW2tv287gEYB1wArAFWAxc0PH3WDzqbDz3uf/XwG5r7U86uW8jeo/2Cc1kxxFr7XZr7TvBf+8BVgP9vY0qLpyF++Cz1tqFQHbwDx7Zvy8AGzom2NI91tp5QOU+N58FPBz898PA2Z089CTgZWttZTCxfhk4uc8CjRKdjae19j/W2pbg1YXAgLAHFsW6eI92x2RgvbX2Q2ttE/A47r0d1/Y3nsYYA5wH/DWsQYmS7HhljBkCHAks6uTuacaY5caYF4wxh4Y1sOhkgf8YY5YaY67p5P7+wMcdrm9Bf9x0x/l0/UtB79GeK7LWbgf3BzdQ2Mkxeq8enCuBF7q470CfD/JZNwZLcB7soqRJ79GeOxYos9Z+0MX9eo/2ESXZccgYkw48CXzLWluzz93v4LYLPRz4LfB0uOOLQkdba48CTgFuCJ6268h08hjVae2HMSYZOBP4eyd36z3ad/Re7SFjzA+AFuDRLg450OeDfOpeYDhwBLAd+HUnx+g92nMXsP9ZbL1H+4iS7DhjjEnCJdiPWmv/ue/91toaa21t8N/PA0nGmPwwhxlVrLXbgl/LgadwpzM72gIM7HB9ALAtPNFFrVOAd6y1ZfveoffoQStrL1MKfi3v5Bi9V3sguDD0dOAi28UCp258PkiQtbbMWttqrW0D/kDnY6X3aA8YYxKBLwFPdHWM3qN9R0l2HAnWZT0ArLbW3t7FMcXB4zDGTMa9R3aFL8roYoxJCy4ixRiTBpwIvL/PYc8Al7omI2YqbvHJ9jCHGm26nHnRe/SgPQO0dwu5DPhXJ8e8BJxojMkJnqo/MXib7MMYczLwHeBMa+3eLo7pzueDBO2zVuWLdD5Wi4GRxpihwTNe5+Pe29K544E11totnd2p92jfSvQ6AAmro4FLgBUdWvl8HxgEYK29D/gycJ0xpgWoB87vaoZGACgCngrmfInAY9baF40x18InY/o8rrPIemAvcIVHsUYFY0wA1zngax1u6zieeo8egDHmr8AsIN8YswX4EfAL4G/GmKuAzcC5wWMnAtdaa6+21lYaY36KS2QAfmKtPZjFaTGli/H8HpACvBz8+V9orb3WGFMC/NFaeypdfD548C1EnC7GdJYx5ghc+cdGgp8BHcc02M3lRtwffwnAg9balR58CxGls/G01j5AJ2tb9B4NH7XwExEREREJMZWLiIiIiIiEmJJsEREREZEQU5ItIiIiIhJiSrJFREREREJMSbaIiIiISIgpyRYRiTHGmIeMMUsO4nGlxph/HOCYIcYYa4w5/eAjFBGJfeqTLSISe34K+L0OQkQkninJFhGJMdbaDV7HcLCMMX5rbb3XcYiI9JbKRUREQqC9RMMYc4Ix5j1jTJ0x5k1jzKHdfPytxpidxpgjjTELjTF7jTHLjDHHdnLs1caYlcaYRmPMJmPMf3cWyz63zQrG1WCMWWyMmRx8vVs7ef4LjTHrjTE1xpgXjDEDOgk50xjziDFmjzGm3Bjzo06e5zhjzKLga5YZY+4xxqTvE5M1xpxkjHnGGFML/C5431XB77E+GOfc7o6liEgkUJItIhI6g4BfAT8HLgAKcVuZm24+PgA8DPweOAdoxG15HGg/wBjzX8C9wNPA6cF//zS41XSnjDH9geeBcty29L8HHqXzkpIpwI3AzcA1wFHA/Z0c9ytgb/D5/gD8yBhzQ4fXHAu8COwMfi8/Ai4EOqv5fgBYDpwJPGCMmQHcB/wFOAW4EpgPZHX1PYqIRBqVi4iIhE4ucLS19gMAY4wPeAoYDazpxuP9wLesta8FH78dWAbMAF40xmTiktWfWWt/HHzMy8Ek/IfGmHutta2dPO+3cAnxGe2lGMaYGuCJTo7NBE6z1lYFjysG7uikjGOltfZrwX+/ZIwpBL4fjKENuAXYBJzZHpMxphJ4whgzzVq7oMNz/d1a+z/tV4wxc4D3rLX/2+GYZ/YzbiIiEUcz2SIiobOxPcEOWhX82lm5RWeagdL9PH4akAb83RiT2H4BXgOK9vM6k4CX90mSu0paF7cn2PvE0H+f457a5/o/gZIOMUwGnton6X8SaAGO2eex/97n+rvAkcaYO4wxM4wxyV3EKiISsZRki4iETvU+15uCX1O7+fia4CwwANbafR+fH/y6EpeQt19eD94+sIvnLQYqOt5grW0Aajs5trvfQ3kX1/t1+Fq2z2u2ArtwM/4d7XvcK8AVuBn8UmBnsJ47rZN4RUQikspFRESiR2Xw6+nsk5gGre3icTuAgo43GGNSgfTOD++Wwi6ub+/w9TPHGGMSgDw+/T7a/f/27hg0iigIwPA/IFhZipDGzkKr9KJEsBBBSCp7SwvRQtKIiiLEJqaSlKmsBMFKLQwqiMQ2iNYqgo0khTY6FrOB45K702TvYPX/4IqF925nu+ExMy/7/zwzV4CViDgIzAGLwAYwv4eYJWliPMmWpO54DXwHpjLz7Q6/zQH71oDTEdHb6Hhuj7HM9j3PUYn1x+b5DTDbJNa9a/YBr/70JZn5NTOXgZfA0d2HK0mT5Um2JHVEZn5rRu4tRcRh4AV1WHIEmMnM/sR3yz3gIvAqldSZAAABPUlEQVQ4Ihap8pF5qhny14A9oxyLiGWqzvoEcAG41FPucptq2nwUEfepWu0F4Elf0+M2EXGTKilZpaaTTAMn8RRbUoeYZEtSh2Tm3Yj4DFymxuz9AD6w86SQrT2fIuIssEQ1KL6jxuI9o0owduMqVbbysInhFs2M6+ad6xFxBrjTvHMDeNDsG2WN+r7zwAFqSsmNJn5J6oTI3FYKJ0n6x0XEcaoE41RmPh+1XpL0d0yyJek/EBELVPnGF2pu9zVq0sd070QTSVI7LBeRpDFrLqUZ1mj+M8d/4rGfuqXxELAJPAWumGBL0nh4ki1JY9Y0K14fsmQmM1cnE40kaRJMsiVpzCJiiroNcZD3Q8bvSZI6yCRbkiRJapmX0UiSJEktM8mWJEmSWmaSLUmSJLXMJFuSJElqmUm2JEmS1LLfrz1uo43JY1EAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x10fa37dd8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(12,8))   \n",
    "x = clf.param_grid['n_neighbors' ]\n",
    "y = clf.cv_results_['mean_test_score']\n",
    "plt.plot(x, y)\n",
    "plt.fill_between(x, y-clf.cv_results_['std_test_score'],\n",
    "                 y+clf.cv_results_['std_test_score'], \n",
    "                 color = 'cyan')\n",
    "plt.title('kNN classifier' +' F1 (micro averaged) score with a confidence interval', fontsize=15)\n",
    "plt.xlabel('n_neighbors' , fontsize=15)\n",
    "plt.ylabel('F1 score', fontsize=15)\n",
    "plt.legend(loc='best', prop={'size': 18})\n",
    "plt.grid()    \n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Исследуем зависимость от количества признаков. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No handles with labels found to put in legend.\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a1b924d30>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "dummy_param = {'n_neighbors':[5]}\n",
    "score = []\n",
    "score_std = []\n",
    "for sub_features in range(10,50,5):\n",
    "    tmp_score = []\n",
    "    tmp_score_std = []\n",
    "    for i in range(20):\n",
    "        cols = np.random.choice(data.columns, sub_features, replace=False)\n",
    "        data_small = data[cols]\n",
    "        classes_small = classes\n",
    "        clf = GridSearchCV(neighbor_clf, dummy_param, cv = 3, n_jobs=1, \n",
    "                   verbose=0, scoring=make_scorer(f1_score,average='micro'))\n",
    "        clf.fit(data_small, classes_small)\n",
    "        tmp_score.append(clf.cv_results_['mean_test_score'].mean())\n",
    "        tmp_score_std.append(clf.cv_results_['std_test_score'].mean())\n",
    "    score.append(np.mean(tmp_score))\n",
    "    score_std.append(np.mean(tmp_score_std))\n",
    "#graph    \n",
    "plt.figure(figsize=(12,8))\n",
    "subfeatures = True\n",
    "x = range(10,50,5) if subfeatures else range(12,36,4)\n",
    "y = np.array(score)\n",
    "plt.plot(x, y)\n",
    "plt.fill_between(x, y-score_std,\n",
    "                 y+score_std,\n",
    "                 color = 'cyan')\n",
    "plt.title('{} F1 (micro averaged) score with a confidence interval'.format(\"kNN\"), fontsize=15)\n",
    "if subfeatures:\n",
    "    plt.xlabel('Количество признаков', fontsize=15)\n",
    "else:\n",
    "    plt.xlabel('Объем выборки', fontsize=15)\n",
    "plt.ylabel('F1 score', fontsize=15)\n",
    "plt.legend(loc='best', prop={'size': 18})\n",
    "plt.grid()    \n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Вывод: \n",
    "* KNN сходится за одну итерацию\n",
    "* Оптимальное число соседей 8-9, в остальных значениях качество падает\n",
    "* На графике зависимости качества от числа признаков видно, что качество немного растет при увелечении признаков до 40.\n",
    "* Зависимость от размера выборки нет смысла смотреть, посколько он и так слишком мал (32)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# SVM"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No handles with labels found to put in legend.\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAt8AAAH3CAYAAABjHTMCAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvNQv5yAAAIABJREFUeJzs3Xd8W+XZ//HP5Rk7085qRiGBQiiBEiCjlBVK2S3QxWjZows6npa2dFE6H34tLe0DdFD28xTCaNmUTdiQwQhkQRJWQhIn8UhsyVP374/7GBRFtmVH0tH4vl8vvWydpUtHR0fXuc89zDmHiIiIiIhkXknYAYiIiIiIFAsl3yIiIiIiWaLkW0REREQkS5R8i4iIiIhkiZJvEREREZEsUfItIiIiIpIlSr4lL5nZGWa20My2mFmDmb1kZn8M5k03M2dmn+9h3bFm1mlmPwiezw2W/0eSZT9sZrFg/uw+YnI9PA4I5o82s/8xs3lm1m5mb/XzPX/PzB7vzzq9bOstM7s0HduS7WdmG83s4rjn95nZz0IMKaeY2ezgu7RH8LzCzC42s2kJy00Klvt0OJFmnpldZGZrgvPS9Yn7ppf1Lu3vOSdsqb63JOvNjP8+5Rqdf0XJt+QdM/sRcDXwIPA54DTgLuBYAOfcAuAN4KQeNvFF/LF/S9y0ZuDzZlaesOxJQEs/wvsDsF/C45Vg3gTgRGAd8HI/tomZDQF+CFzSn/V68Vngf9K0LUm/S4DvmtmIsAPJES/iv0srg+cVwM+BaT2uUYDMbDrwC+AKYH/gV2y7bwrJQN/bTPzxIZKTysIOQGQAzgf+7pz7cdy0e8zsF3HP5wAXmNkQ51xzwvonAc86596Om/YEcCBwBHBvwrJ3A19KMba3nHPP9zBvkXNuLPhSKOALKW4T4GSgDXioH+v0yDn30vZuw8wGOeda0xFPtgUXWTHnXFfYsSTjnHvKzDYBpwKXhx1PomzvP+fcZqCn71Ux2S34e2WwT7oV5L7Jlc/dzKqcc9Gw45DCoZJvyUcj8KXHW3FbD9d6M1AFHBe/jJl9GPhEMD9eK770/KS4ZXcB9sEn8tvNORfbjtVPB/4d/x6D2+4bzWyWmS0ws6iZPW1mk81sjJndaWbNZrbUzD4Zv7Fktz3N7CAzezxYpymojrN3MO+M4PbvzGB6FPh+MG+Umd1gZpvMLBLMn97XGzKzS8zs1eD1VpvZP83sQ3HzbzCzeUnWOz94r0OC5yVmdqGZrTCzNjN73cxOT1hnrpndbmZfMbOV+M97vJntZmZzzOzdIPbFZvYdMytJWP9jZvasmbUGyxwd7PPrE5Y7wMyeCLa1ycz+YWZDk+znV4JtLTSzT/Swi/6Fv6vT1378UfDeW81svZk9kLAfR5rZ381sbbDMcjP7Ttz8avPVodYF8+eb2eGp7L9g3h7mq8lsCR63xb9+knh3Co6lT8RNuzmY9rG4afeY2T+D/xOrH2wJ/l5nH1TvmhT3MtXBe24Kjq1fJH6mSeI6xsweNrM6M9tsZs8n7ode1v2s+epk0eBzv9/Mdoyb/0kzeyHuM/pL9/Gb8P5mB/uv2cxWmdk34pa5Hvjf4GlT3PLbVM0wsxFmdpOZtQSf+096iHuH4PivD47ZB81sStz87mo8J/S1P4PvyD1m1hjEP8/MDoubXxtsY32wH541s1l97Ndk782Z2bfN7LdmtiH4vK40s8pg/hkEF6xxx8bcuPV7PV7jXvMIM7vbzJqBK8x/r29NEuOlZvaOmVnwvNfzmggo+Zb89CLwTTM73cxGJlvAObcUX90jserJiUAMuC3JajcDx5lZVfD8ZOAF4M1+xFZiZmVxj9J+rJuUmQ0GZgHPJpldDVwFXBbEuwP+B/pm4Gl8tZw1wG1mVt3La8wGHgU68In+icBT+Koy8W7G3xk4mg/uENyJv2NwQbBeCfC4mX2kj7c2BvgtcAzwHWAn4LG4fTYHmGFmOyWsdwJwX9wdjcuBnwb74RjgDuBa27be7/7A1/HVdz4DNAXvbznwjeA9/QN/W/+HcfumGl/FqQq/j3+N3987xG/czPbH78N1+Lsa3wm2eV3cMuOB/wD1wTJ/B/6J/xwTPQvsa2Y1SeZ1b+804MfAH/GfwdeBFcDgYH4VMBc4Hl9F4Wh81ajxcZv5B3Am8Bt8daR3gfssaKsQZ5v9F3zGzwCD8KX0ZwBT8XeiLFnMzrlV+GPywLjJB+IT+gODuC14vad6eOvdF5O/5oPqXWvj5v8OX5XsC8D/ARfR952mycA9wfv4PH7//yf4XHtkZqcC/8ZXjTgBvy9fB0YH83cHHgA2Btv9Of5O2u1JNvcP/Hnrs/jP7UozmxnM+1XwfsG///3w58JkrgOOwh+DXwEOJ+FcaGa1+HPEFOBrQeyDgUfizoHdet2fZrYb/jgYF2zrs/jv4YeD+ZXAI8Bh+Iv244ENwWsNJDH9Hv4YPgX4PfBV4NvBvPvwxzh8cGx8I4ijP8frNfjP4tjg/znAp4Pzcff7Nnw1xlvjCkb6Oq+JgHNODz3y6gF8DFgFOHwivRj4JTAsYbkf4qtq1MRNWwA8mLDcXPwPYRn+B/KLwfTF+JPnHsFrze4jLpfk8XQPy16Kr6KSyvv9RLCtqQnTLw6mHxw37RvBtIvipu0eTDsqbtpbwKVxz58L9o31EMMZwTa+nTD9yCQxDMb/sP69H59pKT4RdsBBwbTuz+PCuOUmBJ/5F4LnHwmen56wvRuB+QmfcRT4UC8xWPCaPwZWxU0/D2gHJsRNmxnEen3ctKeAxxO2+clguT2C578DNgHVcct8OVjm4oR1JwXTD+sl5iuAf/Uy/6vB/pnWw/yPJu4//MXTa8R9T3raf/gLveVARdy0XYAu4Jhe4roZuDf4f6dg+b8Ac9wH3/H3j3lgdsJ+HBI8P6OHfXZjwvSXu7ed4vFYEhwLDwLX9rHcGvxdqZ6WmYNvg1IaN+2EIM79Et7fL+OWKcd/jy5J8j0cEjctcd9MDZ6fGLfMEPwF31tx034VHIu1cdNq8Bel5/Vnfwaf52qgqod9cDb+O7RL3LQy/AXL73vZd1u9t2CaA55MWO5O4Pm45+cT3Azt7/Ea95qXJaw7GugEToqbtl+w7PQe4t/mvBZMf4u4868exfdQybfkHefcInzScCz+B9uAnwEL4m/l4n/0yvGlMJjZzsC+bFvlpHu7nfhb/SeZv/29G7DNbcY+/B6YEfc4u5/rJ9NdMrQxybx2ti4dXBH8fSzJtMRSbGCrkvUbnHMu2TJx7kt4PhPY4Jx7onuCc64FXyqeWHKa+LpHBbeem/A/aquDWbsG2+nElyieGLfaF/ENYLvjOBSfPN4Rf8cBXwI9LaG0aaFzbqvqSmY2KLiFvgJ/odaBLwGeHGwH/Oe40Dm3Ju49zgPWx22nGv9DfGtCHE8H29w3bn897JyLxIXx7x52Uffn3VvJ4MvA0cF7mJmkdO2TwEvOuZ4a+M7Af3/evxPkfPWo29j289tm/wGfwpdwxuLe85v45KK3qkdPAfsHVRcOAhbhS527S8MPwieLS3rZRm8S20YsASb2toKZTTRf1WkN/njswJcY79rLalPwJbDX9bLMTOAOt3X9+H8Fr5G4j9+P2znXgU/ae407iRnB37vjttUMPJyw3KeCaZvjPrstwEK2/ez62p+fBG5xPdeL/lSw3TfjXgt8W5s+q6gl0e/PNy6OVI/Xrc51zrkN+PNq/PnoRGCl8438gb7PayKgaieSp5xzbc65e5xz5zvndgfOwZdgnB23zNv4Et3u260n4ROsO3rZ9Bz8rfmvAE85597rZ2jvOOcWxD2W93P9ZAYFf9uSzNvitq5L3h78beye4JzrnjaI5GrwCdjaHubHW5/wfFySad3L1fa0ETObgU8OVuNv/+4HfDxJnHPwSXT3D9eJwN1xP/Kj8KVLTfhkqftxPb5kbVwvsQP8P3x1mavwn/sMPri13x3Hh/AlkInip9UEcfwlIY42/AXgh+O2VRe/keC9JDYKhg8+754+N4Br8SX1J+CrSK03s1/FJeEj6f1zHQc0J1wMgN9X1d31aOOmJRqFv8PUkfDYiQ/eczJP4ttu7IFPuJ/CVwf4UFDN6ED8XaO+LgZ70pjwvJ1e9mNwEXA3/i7TRcAh+GPhP72th9+/0Pc+3mrfBYn4Jrb9jvQr7h58CH9eSEyE6xKej8J/nxI/u0PY9rPrK66+jrNR+O934mudmeS1UjHQ/dSf4zXZ8T4HOMrMhgXHzBeJ6zWrH+c1KXLq7UQKgnPuGjP7HR/0BtDtZuBPZjYGn3zf75xr6mVTTwAN+Lqt52Uk2P6rD/6OYNsfnXRowJcej+trQfzt03hr8XUcE43lg7iT+Sw+eT2xO8GyuAZqcebi61CfaGY34kvo/ztufj2+dGl//HtIFJ9wJEvkvghc7pz7XfcEMzsmYZl1+BLORKPj/m8Mtn8xcH+SZbsv4taRsL+C+rVDtlnDf97Qy34MLrwuAy4z35j4y/iS+zXA3/AJXm9179cCQ8ysOiEBHwtEnHPxF3zJ9l89/mL26iTzkt2p6bY4WPdAfCn3j5xzm81sUTDtQHw99mz5CLA3vmrWA90Tk9R9TrQp+Nvbd2eb70hwcTSS3r8jA7UOGGrb9tCR+D2txyeKv0qyjS1JpvVmE73vg3p8tbavJ5mXrFAhU/pzvCY73u8A/opvyP82/q5HfJe1qZ7XpMgp+Za8Y2ZjnHN1CdNGA8PZtrTiVuBP+NKsPfB1w3vknIuZ2W/xtyeTNYgKQ3fp+WT87dG0cs61mNkLwGlmdkU/SxtfAH5hZgc5556E96tgdDd87EkV0JHwWl9OElvMzG7Hl9C1Apvxjde6PYYvcR7unEu8rZ6KKuJ+/IOkKLGR7nzgy2Y2obvqSdAIbmxcnC1m9jwwxTnX2zE2HzgrIdn9XA/LTgr+vp7KG3HOvQtcYmZn4uv5g69+80Uz+1hQXStZPA7feO5GeL8R2RfwVWb68ij+e7WwP8eNc86Z2TP4EvuP4EvCCf6ehU/kempsCR/c4UlXaWJ3kh1/LOyIv6hLtt+6Lcdf6JyOrzaTzAvAZ83sx3FVTz6H//1NZR/31/zg77EEiWFQHe8w/Pen26P4/b+4l+oiqXoUOMHMfuKSdz/6KL4KzzuJ5+4MaYek3aEO6Hjt5pxrMLOH8Oejt4GlCd+rlM5rIkq+JR+9amZ34ev91QE74qsORIAb4hd0ztWZ2WP4hojNbN2Hd1LOuSvwDdnSzsy6ewjYFX9bv/v5E0GdwmTxvGlma/H1hh/PRFzAhfjeCP5jZlfh61XvByxwzvW4z5xzDwZJ1C1mdiG+BOwC/I/Q73t5vYeB75jZn/BJyyfwPRckcwu+AdV/4evOdideOOeWm9nfgDnBnY8F+IRsKrCrc+6cPt73w8B5QZ3vevzdjsqEZa7D96Zyr/m+5KvwPaJsYOvS9h8Aj5pZDH/htgXfI8oxwE+cc6/jLwTPC7b1R3zJ2Y/wjRkTTcdXp1ncU/Bm9vcg7ueDZQ/BV7/q7q3lxuD1HjI/4t9y/EXcrs65C51zS83sZnxXasPw7QPOxd9BSlZKmehiYB6+d5Rr8aWHE/CJ3vXOubm9rPsk/hhZHpeQPQV8C/9d7qknD5xz7Wb2Jj7hew1/YdZbktyXZfiqAn8wP7LoUPxnvKa3lYKLwx8A/zTfLeLN+IuZTwI3B3WBfw28BNxpZn/F103+f/gGrc9tR8w9xbTYzO4G/hp8pmvxPYwkVi36I/4795iZXY5/r2OBg/FVfpK2jenBL/BJ/5Nm9gf8eWBvYJNz7lr8cfg1YK75Lk5X4Uv+ZwLrnHOXDezd9mhZ8Pfbwfl/c1AF8GIGfrx2uwVf3auJbX8n+nNek2KWiVaceuiRyQdBMoG/ld+KLw2+Cdith+XPxP8g/l8P8+cCt/fyev3p7eT8FJZJ9uhr21cAjyZMuxjYmDBtNgm9AySLjSSt7fE/uk/if6Qb8Yn+tGDeGST0shC33mj8j2sDPol8ApiRwuf4A3y3di34xH+XZPsQXx/9nWDeEUm2Y/heaRbjSy43BDGc1tdnjE827sCXCK7H90ZybuJ7BfbCdz3Xhk9gj8eXSP8pYXuz8CXzm4P3tQSf5AxP+IwWBdt6GV+6upFtezu5C7iuj314Br6udH3wuS0Czk5YZiS+C7s6/PdlGfCtuPnV+O4a1wcxLUjczz3tv2DebviLjfrg81+B70JxYh+xzwr281UJn4cDHuvruMaXpC4K3pPD3ymYFPz/6YT1r8dfSPYWzwx8YhbFN3Q8I5X1gnU/h29Q2IpPPO8Ddoybfyi+BLw1+Bz+Qi89lvS030mht5NgWg2+fnJL8LleRJIelvigsWj3Z/8WvivB7l5mUt6f+B5q7sdfdG4J3u+hcfOHA3/Gf+fb8Rc7/wb272W/Jntvyc4RFxN3LsSfE36H/42IAXNTPV57+izi1h+K/645/J2ufp/XUG8nRf+w4EAQkRxmfrCb+fgfiG0GGJLsMrPJ+OT7K8653nq6GOj2u6tQfco5l4mqCSIiEhIl3yJ5wszuw3cb99OwYyk2ZvYjfCna2/iqJD/Cl+Tt5rYe5jtdr3chcKRzbna6ty0iIuFSnW+R/PE9/CiGkn0OPzLhePzt+aeACzKReAea8HWfRUSkwKjkW0REREQkSzTIjoiIiIhIlij5FhERERHJkoKu8z1q1Cg3adKkUF67paWFwYMHh/LahU77NnO0bzNH+zZztG8zR/s2c7RvMyesfbtw4cKNzrnRfS1X0Mn3pEmTWLBgQSivPXfuXGbPnh3Kaxc67dvM0b7NHO3bzNG+zRzt28zRvs2csPatmb2dynKqdiIiIiIikiVKvkVEREREskTJt4iIiIhIlij5FhERERHJEiXfIiIiIiJZouRbRERERCRLlHyLiIiIiGRJQffzLSIiIiISr6mpiY0bN9Le3t7rcqWlpQwdOpTa2loqKyvT9vpKvkVERESkKLS2trJ+/XomTpxIVVUVZpZ0OeccHR0dbN68mXfeeYcddtghbQm4qp2IiIiISFHYsGEDo0ePprq6usfEG8DMqKioYNSoUdTU1FBfX5+2GJR8i4iIiEhRaG1tZciQIf1aZ9iwYWzZsiVtMSj5FhEREZGi0NnZSVlZ/2pdl5eX09XVlbYYlHyLiIiISNHorbpJOpbvi5JvEREREZEsUfItIiIiIpIlSr5FRERERLJEybeIiIiISJYo+Za80gW0hh2EiIiIyABphEvJK5cBBlQDB4Uci4iIiOQf51y/ejBxzqX19VXyLXnjLeDnwf/HA2vDC0VERETyUHl5OdFotF/rRKPRtA0tD0q+JU844HSgLXi+BTgG6AgtIhEREck3Y8aMYc2aNUQikV5LtJ1zdHR0UF9fz+rVqxk5cmTaYlC1E8kLtwIL8XW+ATqB5cB/AVeEFZSIiIjklWHDhgHw3nvv0dHRexFeWVkZgwYNYocddmDQoEFpiyHrybeZHQn8GSgFrnbOXZIw/zLgkOBpNTDGOTcimHc68NNg3q+dczdkJ2oJUwPwVaAlYXoEuA5f9/uEbAclIiIieWnYsGHvJ+FhyGrybWalwJXAYcBqYL6Z3e2cW9K9jHPuv+KW/yawd/B/Lb7K73R8LYSFwboNWXwLEoJv0nMPJxHgLGBP4KNZi0hERERkYLJd53smsMI5t8o51w7MAY7rZfmTgZuD/48AHnbO1QcJ98PAkRmNVkL3BHAHH9T1TiaCPxCasxKRiIiIyMBlO/meALwb93x1MG0bZrYjMBl4rL/rSmFoA07BJ9e9cUAd8OXgfxEREZFcle0638k6VewpXzoJuN05193GLqV1zewrwFcAxo4dy9y5cwcQ5vZrbm4O7bULxXvAd4FYwvSJzc1cmmTfluAbZo7NeGSFS8dt5mjfZo72beZo32aO9m3m5Pq+zXbyvRr4cNzzifgcK5mTgPMS1p2dsO7cxJWcc1cBVwFMnz7dzZ49O3GRrJg7dy5hvXYhWAocDSTrifPSuXO5oId9W4Wvj7R/xiIrbDpuM0f7NnO0bzNH+zZztG8zJ9f3bbarncwHdjGzyWZWgU+w705cyMymADXAc3GTHwQON7MaM6sBDg+mSYGJAV9iYMPIR4HPAOvTGpGIiIhIemQ1+XbOdQLn45PmpcCtzrnFZvZLMzs2btGTgTkurvdz51w98Ct8Aj8f+GUwTQrMVcAbDLz+djM+Ae9MW0QiIiIi6ZH1fr6dc/cD9ydMuyjh+cU9rHstcG3GgpPQrQW+z7Z9evdHB7A42M5l6QhKREREJE00vLzklHPpvVvBVEXwJeh3pGFbIiIiIumi5Ftyxj3A4/iS63SIAKcCr6dpeyIiIiLbS8m35IQtwJn03ad3f0XxA/BsTzUWERERkXRR8i054QdkJkGO4euRn4YG4BEREZHwKfmW0M0DbmBgXQumohV4ALg8Q9sXERERSZWSbwlVB35Y+GSD6aRTBLgQeD7DryMiIiLSGyXfEqpL8dVCsiEKfBrYkKXXExEREUmk5FtCswo/alI2G0NuBo4FurL4miIiIiLdlHxLKBy+EWR7ll+3A1iEr4IiIiIikm1KviUU/wReJpwS6AjwF+DuEF5bREREipuSb8m6TcB5hNv3dgTf0HNFiDGIiIhI8VHyLVl3HpnrVrA/IvgBeNI9sI+IiIhIT5R8S1Y9hh9GPtt1vZOJAe/hR9bUADwiIiKSDUq+JWuiwCnkVklzFLgP+FvYgYiIiEhRUPItWXMR0BR2EEm0ABcA88MORERERAqekm/JileBK8mtUu94EeAYYGPYgYiIiEhBU/ItGRfD9yySC40se9MEfBYNwCMiIiKZo+RbMu5K/GiWud6osR14Cfhp2IGIiIhIwVLyLRm1BvgR4fbp3R8twJ/xjTBFRERE0k3Jt2TUWUBb2EH0UxQ4GXgz7EBERESk4Cj5loy5A3gG6Aw7kAHoHoAnGnYgIiIiUlCUfEtGbAbOIX+qmyTqAt4Fzg07EBERESkoSr4lI75L7nYrmKoovvT+H2EHIiIiIgVDybek3XPATeR+14KpiADfBhaGHYiIiIgUBCXfklbt+D69C6mudBQ4GqgPOxARERHJe0q+Ja0uAdaHHUQGNAKfxw8YJCIiIjJQSr4lbd7AJ9/5Xtc7mXZgPvDzsAMRERGRvKbkW9LCAaeQf31690cL8AfgwbADERERkbyl5FvS4gZgMYVfLSMKnAC8HXYgIiIikpeUfMt22wB8i/zt07u/WoCjKIzeXERERCS7lHzLdvsahV3dJFEX8Bb+fYuIiIj0h5Jv2S4PAQ/gGyQWkyhwG3Bd2IGIiIhIXlHyLQMWAU6jMHs3SUUEOB94OexAREREJG8o+ZYB+wmwOewgQhbB1/9uDDsQERERyQtKvmVAXgb+TmGNZDlQDcAXKPyeXkRERGT7KfmWfusCvoR6++jWBjwP/DrsQERERCTnKfmWfvsz8A5+YB3xWvCjez4SdiAiIiKS05R8S7+8A/yM4unTuz+i+Oon74YdiIiIiOQsJd+SMgecQfF1K9gfLcDRFFe/5yIiIpI6Jd+Ssn8B84DOsAPJYZ3AKuC8sAMRERGRnKTkW1LSCJyLqpukIgLcDPxf2IGIiIhIzlHyLSn5DupWsD8iwFeBV8MORERERHKKkm/p09P4odRVj7l/ovgBeJrCDkRERERyhpJv6VUb8GWKdwj57eGAjcCJqFtGERER8ZR8S69+jU8gZWDa8HcO/jvsQERERCQnKPmWHi0D/oBKvbdXC/4iZm7IcYiIiEj4lHxLUjF8dRPV806PKPBZYE3YgYiIiEiolHxLUlcDy/FJuKRHM3AMGqRIRESkmCn5lm2sA76H+vROt07gDeDbYQciUkBiqEGziOQXJd+yja+g6iaZEgFuxA/CIyLbZy7wUeDl4CEikg+UfMtW7gceBTrCDqSARYBzgCVhByKSp5YBh+Krcb2OL/0+CJgXZlAiIilS8i3vawbOQL2bZEMUOBLYHHYgInmkDjgb2Adf6h1/rtoCfBLftaeISC5T8i3vuxCfgEvmOXwicTKqryrSlyi+u87JwP8Fz5M1Bm/BX9Q+kr3QRET6LevJt5kdaWbLzWyFmV3YwzInmNkSM1tsZjfFTe8ys5eDx93Zi7rwLQSuxf+oSXa0AU8Avw87EJEcFQP+CeyAH6gqQt+9BbUAxwH3ZTY0EZEBK8vmi5lZKXAlcBiwGphvZnc755bELbML8CNgf+dcg5mNidtE1Dk3LZsxF4NO4Eso8Q5DC3Ax8HF8nVUR8Z4Avor/oehvz0sR4Iv4xs1fSHNcIiLbK9sl3zOBFc65Vc65dmAOvpAi3rnAlc65BgDnXF2WYyw6f0CDv4QpChwPrA07EJEcsBz4FHB08P9AuzyNAqcB/5umuERE0sWcy16NUzP7AnCkc+6c4PmpwCzn3Plxy9yJb8C+P1AKXOyceyCY14nvUaoTuMQ5d2eS1/gKvrc8xo4du++cOXMy+6Z60NzczJAhQ0J57f5oBxaTX4PpTGxuZnUe7Nv+MKAK2C34Pyz5ctzmI+3b3nXiCwHq6f/5qLdzQgnwYWDUdkVXvHTcZo72beaEtW8POeSQhc656X0tl9VqJyTPKxKz/zJgF2A2MBF4ysz2cM41Ajs4594zs52Ax8zsVefcyq025txVwFUA06dPd7Nnz07zW0jN3LlzCeu1U+WAg4Fnga6QY+mPS+fO5YIc37cDUQ2cCVwRYgz5cNzmK+3b5FqBPwK/xSfgAxljoK9zQjXwG+A7A9h2sdNxmznat5mT6/s229VOVuMLIbpNBN5LssxdzrkO59yb+DuPuwA4594L/q7C9zS1d6YDLmQ3Ay+SX4l3IYsA1wG3hR2ISBbEgJvwjSl/g69ekqnBvSLAT/AJvohI2LKdfM8HdjGzyWZWAZwEJPZacidwCICZjQJ2BVaZWY2ZVcZN3x+NUzJg9cA30BDyuSaCL/1eFnYgIhn0JLAHvn7gBrIztkAEn+T/BHXvKSLhymry7ZzrBM4HHgSWArc65xab2S/N7NhgsQeBTWa2BHgc+L5zbhN+FOEFZvZKMP2S+F5SpH/Ox9/uldwTwfdVrD7XpdDAAK3sAAAgAElEQVS8DhwOHIX/Acj2xX8E+BPwXygBF5HwZLvON865+/GjmMdPuyjufwd8N3jEL/MssGc2Yix0c4G7yNwtXtk+DlgPnALcQbgNMEXSYSPwY3zPIx2EW9UtAlyNT/z/jkaaE5Hs03mnyLTikzoNIZ/bWvGj9P0p7EBEtkMrvp71JHyf263kRhuTFnx981PJjXhEpLgo+S4yFwMNYQchKWnB1099JuxARPophm/QvQM++c5kY8qBiuAbGH0BXxovIpItSr6LyGLgf1Cpdz6JAsfiq6GI5IOn8I0pz8U3pszlRt0R4CHgM/Q9bL1IOq0EmtCdl2Kl5LtIxIAvo0aW+WgLPjnoDDsQkV68ARyBbywcRmPKgYrge185An+xK5JpdwHTgFX4u0PXobsvxUbJd5H4G7ACtfDPRx34uxbfDzsQkSQ2AV8DPgY8Sn7eWYsCLwCfRL0MSeZ04c/jJ+OPsxh+oJNvAROAK1EBWbFQ8l0E3gN+QP6URMm2IvhhW+8IOxCRQCvw/4AdgevJncaUAxUFXgYOxFcHEEmnjfhj6y9se4elGV9F64fAOOD36Pe60Cn5LgLnoPqMhSCC753hjbADkaLmgDn4pPtX5GZjyoFqxVeZ2Q8/EJlIOszDD1SykN7vDLUAjfiOEcYBv0QXgoVKyXeBuwt4AtUnKxRRfN1UlYpIGJ7BD7ZwDlBHYR6HbfjGcDPx71FkoBy+yudsfMl3qoVgEXxbn0vw1VEuDNaXwqHku4BtBs4mP+tgSnIxYC1wOqq/L9mzAj8q5eH49geFmHTHawfeAWYAa0KORfJTFN/JwfcYeEPeKP679md8w8xv4s//kv+UfBew71P4P5LFqBV4ALgi7ECk4G0Cvo5vTPkwxXUh34FvLzMDeDvkWCS/vInvzeRO0vOdacUn4lcBO+HvPL2Thu1KeJR8F6gX8EM5q+V0YWrBN855PuxApCC1Ab/D1+u+Dv/Dn8+NKQeqE1/1ZDq+9F+kL/8B9sJXXUp315Xt+N/0G4EpwJdQG6B8peS7AHXgb3epz9rCFgU+jW8lL5IODrgVf4v7FxRWY8qB6sI3vpwJLAk5FsldMeBnwOfx9bUzebHagU/Cb8PflToeeC2Dryfpp+S7AP0O1QsrFluA4yjOUklJr2fxP+Rn4Ut7i6mKSV9i+F4oPoHvjlAkXgNwKHAZ2S306sQn4ffgLw4PBxZk8fVl4JR8F5iVwG/QD2exaAdewbeGFxmIlcDRwGH40jO1E0nO4bt9OwjfdZwI+Iux3YHnCO+7E8Mn/Y8ABwMHAE+HFIukRsl3AXH4fqDVp3dxieBHRrsn7EAkr9QD5+G7DnwIXbCnagt+JMynwg5EQnc9sD+wjtyonuXw3+Nn8F3STsc3lFbPWLlHyXcB+V9gEaqCUIyi+MY3K8MORHJeG34EvR2AayjexpTbowU4El/SKMWnDV896zxy96I1gh/U57PAHvjCGSXhuUPJd4HYiO8DVLeMi1cEnxDk6o+BhMvhG2jtiBpTpkME397i3rADkaxajS9RvoX8ONe24BsKfwn4CL5BtS62w6fku0B8A3UrWOxi+AFBzkQlHLK15/Ddn50FrEcX6ekSAU7AX9RI4XsMX4q8lPxIvOM1A6vwA+/tCNyARr4Ok5LvAvAIcB+q6y2+CsG9wN/DDkRywip8d5SfAl7F/wBLekXxI87eGHYgkjEO+C3+u9REfpccN+MLac7HD13/F3QHLAxKvvNcBN/IMt+uwiVzIsB3gflhByKhacD/uO4BPIjOD5kWBb4G/C3sQCTtNgNH4ZPvQho7oxk/RsQPgHHAH9AdsWxS8p3nfoa/EheJFwWOwQ8PLsWjHf8jGt+YsjPUiIpHFH/Re1nYgUjaLAamAnMp3MS0BX+xfhE+Cf81yimyQcl3HlsE/JXCuhqX9GnCt3TP51ukkhoH3I6vy/lzfKmW2oBkXxT4KfCrsAOR7TYHP3DNGoqjWkYE343mf+Oro/wIFd5kkpLvPNWFH0JeP7DSk3bgRfzdESlczwPTgDPw/Q0XagldvogAl+AHvlLD5/zTge/A4Gz8Z1lsn2EEfw75E/Bh4Nv484qkl5LvPHUF8CbFd2KQ/uk+id4fdiCSdm8Cx+KHtV6Eku5cEgEuxycuOkfnj7XALHxPIMXeTqIVfyfn78Bk4BzgnVAjKixKvvPQu8CP0Y+tpCYKnIRP1iT/NeD79J8K/AclCbkqgq93fy6+G1DJbU/jv1Ovoe9UvDZ8In4jMAV/x31FqBEVBiXfecbh++pVt4LSHxF8i31VU8pf7cAf8Y0pr0aNKfNBBLgZOAW1vchVDv+9Ohx/Yau+r5PrwP9+3ArsiW9PtDjUiPKbku88cwd+wAz96Ep/dOFvGZ4TdiDSbw74NzAJ3yOBGlPmlwhwF/B5lNjlmmZ8EnkR6rggVZ348889+JE+j8APYy/9o+Q7jzThkydVN5GBiOIv3v4RdiCSsheAvYHT8PVR9d3PTxHgYeAzFEfPGfngdXwJ7oPoezUQXfgk/GHgQOAg4JlQI8ovSr7zyH+hq3PZPhF8I7AXww5EevUmcBzwSeAVlBwUggjwJL56g87j4boD2Ad4G91F2l4Ofzw/hT+2ZwCPoobGfVHynSeewfc7qhOFbK8ocDRQH3Ygso1G4Dv4hl/3o4ZfhSaKH3n2EHyVB8muTvxASKfgL2iVIKZXBFiALzjYE7gX7eOeKPnOA+34FsYqLZF0acTXQVUvDLmhHT8y4g74rr3UmLJwRfF3Mw5AIwlm0wb8Pv87uqjNtBZ8Y8yTgF2A21CD40RKvvPAb/EnDpF0acOXwF0cchzFzuFvgU/GD4a0Bd3dKgatwDLg42gUwWyYB+yOr26nxDt7WoCV+B7aJuG7K1SjY0/Jd457HfgdOmFI+rUAl+IbHEn2zQP2BU4F3kP1uotNG7AKP4T5+pBjKVQOuBKYDWxEiV9YmoHVwHnAROBvqOGxku8c5vB104r9IJXMiQIn4BseSXa8BRyPTwheQkl3MWvHdwE6A1gTciyFpntwsR+iKpu5ohmoAy4AxuGr2hVrwaKS7xx2LbAE1cuVzGrBN8DURV5mdeF7LNoduA8lBOJ14ruRnI6/MJPttwrYC98XtS5uc08LfkCjn+GT8N8Am0ONKPvKwg5AkqvD/1DrxCGZ1oXv2u6rwBnhhpIyh4+7A5+8dP/tTDKtr7/9WbYteHQEf9uDR0fc38RH9/pn42+3qk63JOrEn/NnAM/iG6nJwNyPL/FuQQVXua47v/kN8N/At/Cl4rWhRZQ9Sr5z1FfRj7RkTxTfIv1I4D/0P3ntTkLb4v7GJ6SJiWn3Op1snQB3xT3v6uERCx6Gv3XX/bC4v/EP4v4mcnF/ux+xJH9jbH+XWTH0nZaexfCNL2fh+0yeGm44eSeGL0m9DN1Vyjfdn9dlwJ/w+c+FwNjQIso8Jd856AHgIdQ4RLIrgr/9/bXgeXcyGv9/YkLa/QhDd+m3urCSQuHw3YDuDzyOH91U+lYPfA7fx7QS7/zVXTjxV/xdwtOAnwIfDi2izFGd7xzTApxO8TZCkHDF8H0PN+Hr4G0JHs34YzOKP0F2l1rrtq5Iejn89+9g4IWQY8kHL+HbUTyHqmkWijb878x1wK74jidWhhpR+in5zjE/wic7IiJSvLYAh+KHpJfkrsXfJViPLxCQwtKBT8JvAfbADwy3JNSI0kfJdw55Ebga3TYTERFfknsU8HDYgeSYNnzj8G+i38ti0IlPwu/C9wp0FP6ORz5T8p0jOvFDyKtBloiIdIvg+4W/J+xAcsQ7wD7Arah6ZrHpwl9sPYS/43EwvnegfKTkO0f8CXiX7e9RQURECksEOBGfcBazR4A9geWoxLuYxfCf/1PAYfhRYh8jv/InJd854G3g56ixSJ+cY8T9S1m6Rv3AiEhxieKrWtwQchxhiAG/Ao7FNwRXD0cCPtmOAPPxx8bH8AOY5UMSruQ7ZA5/QlVjkb5Vv7qO4U+u4pFX2xn80uqwwxERyaoo8HV8V2zFogk//sD/Q6Xd0rMW4DX8HaJdyf3CTPXzHbLb8FdtnWEHkuOsvZOa+5bQPm4YO7U1425bRKyynOjuhdwNv4jI1qL4UQAjwPdCjiXTXsMn3hvxjSxF+tICrMD3gJPLVPIdogbgK+T+FVouGDZ3JWVNrdQfN5Vj9hlE+/hhjLrpRSpXbQo7NBGRrIoAF+GrYhSqm/Cjfb6HEm8pPEq+Q/Rt1LtJKso2tTD8iVU07z2Btkm1VJQZdWfOpKumijE3LKBiTVPYIYqIZFUEuAT4IflRxzVV7fhRds/Fv8dCem8i3ZR8h+RJ4HZ0RZ+KmnuX4EqNxqN2e39abHAF68+ZRayqnDHXzqOsrjnECEVEsi8CXAF8i8JIUt/Dl3b/L+pGUAqbku8QtOH79Fbjkb4NWlZH9dI6mg7dha5hg7aa1zW8ivVnzwSDsde8QGmj9qiIFJcIfqTHs/G9guSrJ4Gp+HreSryl0Cn5DsEvgfqwg8gHnV3U3rOYjtGD2bz/5OSLjB7C+jNnUtLaydhrXqCkWfcSRKS4RPBDcH+Z/OuGzwG/xzesbESdD0hxUPKdZUuBy9CVfSqGPf0m5Zsi1H9mKpT1fKh2TBhO3RkzKG2IMua6eVir+gEXkeISAe4GPgfkyxmwGd8/88XoTrAUl6wn32Z2pJktN7MVZnZhD8ucYGZLzGyxmd0UN/10M3sjeJyevajTI4YvmVDZbN9Km6IMf2wFkd3H0rrr6D6Xb5tcy8ZT9qFi7RbG3LAA68i38h8Rke0TwY8CeQy5/zuzDF/N5BFUGCXFJ6vJt5mVAlcCRwG7Ayeb2e4Jy+wC/AjY3zk3FfhOML0WPxDkLPxooj83s5oshr/drgJeJ7/r5WVLzf3LIOZo+PTufS8ciO42lo0n7EXlW/WMuulF6NKeFpHiEgGexg+7natJ7b+B6cC7qMcvKU7ZLvmeCaxwzq1yzrUDc4DjEpY5F7jSOdcA4JyrC6YfATzsnKsP5j2MryaWF9biB0ZQn959q1y1icGvvMfmg3ems7a6X+tGpk2g/tipVC+tY+TtiyBWCH0AiIikLoofvG02vmpHrujEl6adiv8t1NlZipU5l73D38y+ABzpnDsneH4qMMs5d37cMnfiC4j3B0qBi51zD5jZBcAg59yvg+V+BkSdc5cmvMZX8GPXMHbs2H3nzJmThXe2rebmZoYMGfL+8xXAZnSy6Uss5pjzbJT2TjjlwCrKSm2bZSY2N7M6bt8mM39lO8+/0cHHdijjoI9WYLbtdmRbqexbGRjt28zRvk3OgEHAFPyP6UAk/pYNVCf+dzCK7v5203GbOZOamxkZwr495JBDFjrnpve1XLaHl0+WASXmo2XALviL9onAU2a2R4rr4py7Cl/Dg+nTp7vZs2dvR7gDN3fuXLpf+178gDq5egswlwx95k1qm5dQd8o+XLjHuKTLXDp3Lhf09bke7Ki5bymLnn6TJ3edTNOndk1/sAUopX0rA6J9mznatz2rBCbjq6KMHMD68b9lA/U8vh76FvKnMWg26LjNnGvmzuXzObxvs13tZDXw4bjnE/H96icuc5dzrsM59yawHJ+Mp7JuzmkGzkSJdypKmtsY/vDrRHcZRXTqh7ZvY2Y0HP1RmvedyIhH3mDoM2+mJ0gRkTzSBqwCZgDrs/zaDrgc+CS+e10l3iJetpPv+cAuZjbZzCqAk/C9I8W7EzgEwMxGAbvizx0PAoebWU3Q0PLwYFpO+wG5Vecul414cDkl7V2+a8F0VBMpMTZ9bk8iu4+l9p4lDH5x9fZvU0Qkz7TjS6+mB3+zIQKcAFyIuhEUSZTV5Ns51wmcj0+alwK3OucWm9kvzezYYLEHgU1mtgR4HPi+c26Tc64e+BU+gZ8P/DKYlrPmA9ej1typqHi3kSEL3mXz/pPoHJPGelqlJWw4eW+iO49k5O2LqFqS7bIfEZHwdeAb/s8AMn0fcCXwMXyVS931FdlW1vv5ds7d75zb1Tm3s3PuN8G0i5xzdwf/O+fcd51zuzvn9nTOzYlb91rn3EeCx3XZjr0/HPAldMWfkpij9u7FdA2ppOnQXdK//fJSNpw2nfbxwxh904tUrtyU/tcQEclxXUAdPgF/PUOvcS8wDZ/gq+BJJDmNcJkh6/GlDNK3wS+upvLdRhqP2g03qDwjr+Eqy6g7cyYdtdWMuXEBFaubMvI6IiK5LIavf/1x4LU0brcLX8XkBHxVS/VoItIzJd8Z0IRPvNWnd98s2kHNA8to3bGGlr0nZPS1YoMrqDt7JrGqcsZcN4+yOtXGF5Hi44BGfH++L6Vhe5vw3ZNdju72iqRCyXcGbA47gDwy4pHXKWlpp/7YNDWy7EPX8CrWnzMLDMZe8wKljfqpEJHi4/C/VQfhuwIcqIX44arnofrdIqlS8i2hKV+3haHPvU3zzB3omDA8a6/bOWow68+aSUlrJ2OvfoGS5rasvbaISC5pBj4FPDGAda8GDsTXI29PZ1AiBU7Jt4TD+UaWsUFlNB4+Jesv3zF+OHVnzqC0KcqYa+dhreqBVkSKUwtwNPBQisu3AqfhB4/TvUOR/lPyLaGofnUtg1ZtovHwKcQGV4QSQ9ukWjZ8eV8q1m1hzA0LsI6uUOIQEQlbBDiebQfeSPQ2sDdwO6pmIjJQSr4l66y9k5r7ltI2fhjNM3cINZbW3caw8YS9qHyrnlH/fBG61EZfRIpTFD/y3S09zH8Y33/3G6jEW2R7KPmWrBv++ErKmlppOHYqlGS+kWVfItMmUH/cHlQvq2Pk7Ysg5sIOSUQkFFHgTPwAcd1iwC+A4/CNNHWPUGT7lIUdgBSXso0tDHtyFc17T6BtUm3Y4byv+eM7UhJpp+ah14lVldPwmd2z0vuKiEiuiQLfwFcrmQIcDjyHSrtF0kUl35JVNfcuwZUajUftFnYo29h8yEfYfMBkhj37FsMfeSPscEREQhMFLgBeBZ5G9btF0kkl35I1VcvWU72sjoajd6Nr2KCww9mWGQ3HfJSSaAcjHn2DWFU5Ww6YHHZUIiKhiOKrmKgzVpH0UvIt2dHZRc09S+gYPZjNn8jhhNaMTZ/bk5JoB7X3LiFWVU7LvhPDjkpEREQKhKqdSFYMe+pNyjdFqP/MVCjL8cOutIQNJ+9NdOeRjPzXIqqWrA87IhERESkQOZ4FSSEobYoy/LEVRKaOpXXX0WGHk5ryUjacNp328cMYfdOLVK7cFHZEIiIiUgCUfEvG1dy/DJyj/pjdww6lX1xlGXVnzqSjtpoxNy6gYnVj2CGJiIhInlPyLRlVuXITg195j82zd6artjrscPotNriCurNnEasqZ8x18ymraw47JBEREcljSr4lc7pi1N6zmM6aKjYfvHPY0QxY1/BBrD9nFhiMveYFShvV262IiIgMjJJvyZihz79Nxbot1B+zO668NOxwtkvnqMGsP2smJW2djL36BUqa1fmWiIiI9J+Sb8mIkuY2Rjz8OtFdRhGdOjbscNKiY/xw6s6YQWlTlDHXzsNaO8IOSURERPKMkm/JiBEPLMfau3zXggU0THvbpFo2nLIvFeu2MOb6BVhHV9ghiYiISB5R8i1pV/FuI0MWvsvmAybTOWZI2OGkXeuUMWw8cRqVb9cz6p8vQlcs7JBEREQkTyj5lvSKOWrveo2uIZU0ffIjYUeTMZG9xlN/3B5UL6tj5G2vQMyFHZKIiIjkAQ0vL2k1eOFqKlc3sfHEvXCDysMOJ6OaP74jJdEOah5cTqyqnIZjC6uKjYiIiKSfkm9JG4t2UPPAMlp3rKFl2oSww8mKzbN3piTSzvCn3iRWXUHTYbuGHZKIiIjkMCXfkjYjHn6dkkg79cfNLJ4SYDMaj/4opZEORjz6BrGqcrYcMDnsqERERCRHKfmWtChft4Whz79N86wd6Bg/POxwssuMTZ/bk5LWDmrvXUKsqpyWfSeGHZWIiIjkIDW4lO3nHLV3v0ZsUBmNh00JO5pwlJaw4aS9iX5kJCP/tYiqxevCjkhERERykJJv2W7Vi9YyaFU9jYdPITa4IuxwwlNeyoZTp9M+YTijb36JypUbw45IREREcoySb9ku1t5Jzf1LaRs/jOaZO4QdTuhcZRl1Z8ygo7aaMTcsoGJ1Y9ghiYiISA5R8i3bZfjjKyhraqXhuKlQUiSNLPsQG1xB3dmziA2uYMy18yir2xJ2SCIiIpIjlHzLgJVtbGHYk2/SvPcE2nasDTucnNI1fBDrz54FJSWMvWYepQ2RsEMSERGRHKDkWwas5t4luLISGo/aLexQclLnqMGsP2smJW2djL1mHiXNbWGHJFJ4OrvCjkBEpF9STr7N7GNmdouZrTSzNjPbJ5j+GzM7KnMhSi6qWrae6mV1NB66C13DBoUdTs7qGD+MujNmUNoUZew187DWjrBDEikMXTGGPrWKD//iYe57sRWL6rslIvkhpeQ7SK4XAh8CbgTixw1vA76Z/tAkZ3V0UXPPEjpGD2bLJyaFHU3Oa5tUy4ZT9qV8/RbGXL8Aa1dJncj2KH+viQ/95Vlq71tK+/hhvLWhi3FXPE35e01hhyYi0qdUS77/G7jeOXcw8JuEeS8D09IaleS0YU+/SfmmCPXHToUy1VxKReuUMWw8cRqVb9cz6qYXoSsWdkgiecfauxjxn6WMu+IZyppa2fClvVn/tf343MxBWEeMcX95liHz3wk7TBGRXqWaOe0G3BL87xLmbQbU2q5IlDZGGf7YClqmfojWXUaHHU5eiew1nvrj9qB6WR2jbnsFYolfJRHpyaAVGxn35ycZ/sQqmveZyHvfPZjIx8aDGeNqSln7rQNonVTLyH+9ysjbXsE6dIdJRHJTqsPL1wE79TBvKqCihiJRc/9ScI6GYz4adih5qfnjO1IS7aDmweV0VZXTcOxUMHXRKNKTkpZ2au5bypAXV9MxajDrzv04bTuP3Ga52JBK6s6ayfBHXmfEYyuoeG8zG768D52jBocQtYhIz1JNvucAvzSzJcBzwTRnZrsCPwSuyURwklsqV25k8KK1NH5qV7pqq8MOJ29tnr0zJdEOhj+5ilhVOU2HTwk7JJHc4xzVr7xH7T1LKIl20DR7Z5oO3QVXXtrzOiVG0+FTaNuxhlG3vMy4y59m4wl7EZ36oezFLSLSh1ST758BuwNPAOuCaXfhG2A+BPw2/aFJTumKUXv3Ejprqth8cE83QSQlZjQetRulkXZGPLaCWHUFWw6YHHZUIjmjtCHCyDtfo2r5BtomDmfTObPoGDcs5fVbp4xh7TcPYPQ/X2TM/y6k6aCdaDxiCpSqjYqIhC+l5Ns51wZ82swOBQ4FRgH1wKPOuYczGJ/kiKHPv03F+i3Unbpv7yVPkhozNn12TyzaSe29S4hVldOy78SwoxIJV8wx9Nm3GPHQcgDqP72771FpAKPndtVUs+5r+1F77xKGP7mKyncb2Xjy3uoaVURC12fybWaVwAXAvc65R4FHMx6V5JSS5jZGPPw60V1HE919bNjhFI7SEjaePI2S6+cz8l+LiA0q0+1xKVrl721m5L8XUbm6iciU0dQfvwddNdtZva2slPrj96Rtx1pq//0q4/7naTZ8aW/adtq2zriISLb0eQ8uKPX+CTAi8+FILqp5YBnW0UX9Z3ZX48B0Kytlw6nTaZ8wnNE3vUTlyo1hRySSVdbRxYgHljHuiqcpa4iy4aRpbDhjxvYn3nFa9p7AuvP3J1ZVxth/PM+wuSvBqbchEQlHqhXgXgD2zWQgkpsq3m1kyILVbN5/Mp2jh4QdTkFylWXUnTGDjpHVjLlhARWrG8MOSSQrBq3YyLg/PcnwuStp2XuC7z5w2oSMXOR3jB3K2vMPILLnOGoeWMboGxdqVEwRCUWqyfcPgK+b2flmtpOZDTaz6vhHJoOUkMQctXe9RufQSpoO3SXsaApabHAFdWfPIja4gjHXzqOsbkvYIYlkTEmknZG3v8LYq18AYP05s9j0xb2IDa7I6Ou6yjI2nrw39Z/ZnarldYy7/GnK12hUTBHJrv6UfO8M/A/wBn5gnS0JDykwQxa+S+XqJhqP/iiuMtWOcWSguoYPYv3Zs6CkhLFXz6O0IRJ2SCLpFXQfOP6PTzD4xTU0Hbwza79zEK0fGZW9GMzYsv9k1n11P6wrxri/alRMEcmuVDOqs9h2ZEspYBbtYMQDy2mdVEPLtPFhh1M0OkcNZv3ZMxn79+cYe8081n11P2JDK8MOS2S7lTZGqb3zNaqX1fnuA8+aScf44aHF075jDWu/eQCjbnmZkf96lcq3Gqg/bg9chXpzkswqX7+FYXNX8uiGNobF3qCztvr9R2xwhdpWFYFUuxq8PsNxSI4Z8fDrlETaqT92pk4EWdYxbhgbzpjBmGteYOy181j31Y/jBpWHHZbIwMQcQ597ixEPLgcH9cd81HcfmAN9bseGVFJ35kyGP/oGIx59g4o1TWw4ZV+NiikZYe1dDH/sDYY9uQpXXspbroua1a9vtUysonSrZNw/quioraarplpd/RaIftUlMLPxwH5ALb6f7+ecc+9lIjAJT/m6zQx9/m2aZ+0YaslUMWubVMuGU/ZlzI0LGHP9AurOmqkSOck75es2+1LldxuJ7jqaTcfvkXuj45YYTYftStsOIzQqpmRM1bL11N61mLKGKM37TKTh6N343YLn+P4nDqCsIUrZpghl9RHKGoK/m1oY9MYGSjpiW22nc1glnTWJybl/dA2tHFCf+JJ9KSXfZlYKXA6cC8RnAF1mdhXwTedcLOnKkl+co/auxcQGldF4+K5hR1PUWqeMYeMJ0xg15yVG/XMhG06bnhOlhSJ96uhixGMrGPbESmJV5Ww4aRqRvcbn9F00jYopmVDaGKX2nsVUL15P+5ghrPvKx7fqZ95VlNExdigdY4duu7JzlDS3U1Yfobw+SMqDx6BVmyh9ecMpgNgAACAASURBVA0WVyHYlZXQWVNFZ201Hd1Jec0Hpee6g5o7Ui35/gW+3vePgVuA9cBY4ETgl8Am4KJMBCjZVb1oLYPerGfTZ/cgVp3Zngekb5G9xlPf2sHIO15j1K2vsPHEaSrZkJxWuWoTI//9KuUbW2jeZwINx+ye8V5M0uWDUTGX+lEx32lk45c0KqYMQFeMoc+8xYhHXoeYo+GIKWw+cCco68fFnBmxoZW0D62kfceabed3dlHW2LpVUt79GPJWAyVtnVuHNLji/eR8m8fwQbrQzKJUk+/TgJ865y6Nm/YO8Hszc8C3UPKd96ytk5r7ltI2YRjNM3YIOxwJNM/akZJoBzUPLCdWVU79cVNzugRRilNJpIMR/1nK0Pnv0lFbxfqzZ9K6y+iww+q/slLqj9+D1kk1jPxXMCrmyXvTtrNGxZTUVLzdwMg7XqVi3RYiU0bTcNwedGaiulVZKZ2jBidvo+AcJdEOyuqjWyfmDREq1jRR/do6LPZBsbkrMTpHVNFZm5CcByXnsepy/e6kUarJ9xhgUQ/zFgXzJc8Nf3wFZZtb2fDlfVS6mmM2H7wzJZEOhj+5iq7qcpoOnxJ2SCKec1S/uo7auxdTEmmn6aCdaPrUrnnfRiEybQId44Yx+v8WMvbq52k8Yjc2H7STzo3So5JIOyMeWMbQee/SOXwQdafs49sOhJG0mhGrrqC9uoL2iUnabnXFKG1qpawhvkqLT9SrF6+ntKV9q8VjlWVbNf7cOkGvgrL8/r5nW6rJ9+vAScBDSeadBCxP9QXN7Ejgz/i641c75y5JmH8G8HtgTTDpCufc1cG8LuDVYPo7zrljU31d6V3ZxhaGPfUmzftMSH57S8JlRuNRu/nSxcdWEKsqZ8uBO4UdlRS50qag+8CldbRNGMamM2fQMaFwGml3j4o58l+LqHlgGZVvN/jBgKpVd1biOMfgF9dQc/9SSqIdNB042V+A5vL4GKUldNVW01VbTdvO2862ts6k1VnKNrQwaPkGSjo/aObnDLqGDdq6jvnIqvefdw2tVKl5glSPjF8Dc8xsB+B2fJ3vMcAXgUPwCXifgoabVwKHAauB+WZ2t3NuScKitzjnzk+yiahzblqKMUs/1NyzGFdWQsNRu4UdivTEjPr/396dh0lZXXkc/57e96Yb6AZkdUEBcaMluKNxwai4GydGYxKjZkzMmGUSZxLjJHFiMmaZrGbRaCZGNGoiKtEYtaNGRXAXEEEFQdm7WRp6rzt/3LelaKo3uqreWn6f5+lHq+rtqtOXt7vOe+vec86ZSk5LO9UPLSFSnM/2ujFhRyXZKOIom7+SqoeXQiRCw0cmse2o8Rm5ZrSrK2bruCqq5i1hxM+eYsNF0zLqIkP2XP66bVT/5XWK3mmgZewQGs6eSvvIirDDGjRXmEf7yIrYP0vEkdvUGjM5L1q+kbytLbsenp/TY4WWjupiXEEKX6QkSH/rfN9tZpvxGy//F8gH2oEXgFnOuUf7+XrTgeXOubcBzGwOcCbQPfmWJCpeso6SpRtoOG0SkXJtLEppOcbGCw8h57aFDL33VSLF+SqJJkmVv24b1fe+StG7m2nebxgNZ02lY2iKlQ+Mt6ArZuuYIQy/40VG/vIZGmZPoenwMZrRy1LRNbsjhXlsOmcqTXVjsmNZUo7RWVFEZ0URreOrd3+8vdOXTwzWmOdv2rVKS05b5y6Hd5YV7LLG/INlLUNL/GbnDBzTfl9uOOf+BvzNzHKAYcDGPSgvuBewKur2auBDMY4718yOxS93ucY51/U9RWa2EOgAbnTO/WWAry/dtXdS9eBi2mrKfOMLSX15uWy4eBq1v53P8D++xPpPHp7c9tySndo7qXxiOZX/eItIYR4bLziY7YfulVXJZ9vYqK6Y971G4Up1xcxGxUvWUT1315rdkTJ1Iv5Afi4dNWV01JTt/phz5OxojzlrXriykZJX1+y6ETS3ayPo7rPm7dUluOL0XAJmzvXdNd7MyoEy59yaGI+NBLY555r68TznA6c45y4Lbl8MTHfOfT7qmKFAk3Ou1cyuBC5wzp0QPDbKOfe+me0NPA582Dn3VrfXuBy4HKC2tnbanDlz+vz54q0daGxqYlVZjBMvxSx4q43nlrVzZl0RY4elxxvI6KYmVqfB2CZac5vjvueb2dbsOHt6EbWVg//309gmTjqP7XsNnTy+qJXN2x37j8rlmAMKKS5InaQ72WMbcY7nl7ez4K12hpbn8JFDChlSmnlLbiC9z9t429Yc4ck32nh7XSdVpcbxUwrZq3rP/+5qbHfXGXE0tTi27HBsbY6wdYdjS/Dfrc0RWtp3Pb4wDypKcqgsNipKcqgoNipLjMNyWxhXlfyxPf74419wztX1dVx/k++7gS3Ouc/EeOxXQKVzrs9132Z2BHC9c+6U4Pa1AM657/ZwfC7Q4JzbbXGdmd0GPOicu6en16urq3MLFy7sK6y4WwXcU1/PF2fOTPprD0Tu5mZG/eAfNO8/nI0fnxZ2OP12U309X07xsU2W3K0tjPjlM1hrB+uuOCJ2o4YB0NgmTjqOrTW3U/XXNyh//l06qorZdPZUWiamXvnAsMa2aOl6ht31Mtbp2Hj+wTQfmHlLwNLxvI276JrdzrHlw/ux9egB1uyOQWM7cNayc9Y8P6pCi1/i0ox1+gUZZ04t4H8vOin58Zn1K/nu77KTY4Ere3hsHvDLfj7PAmA/M5uAr2ZyIfCx6APMbGTUDPtsYElwfxWwI5gRHwYcBXy/n68rMVTNWwI4Gk+bFHYosoc6K4pYd9mHqL35WWpueZ61Vx6Reu27JS0Vv76G6vsXkdvUytajJ7D55IlZuTGqN7t0xfyDumJmosKVDVT/+XVfs/uAGhpnT0lMzW7pF1eUT/uoStpHVdLc/cGII3dbC3mbdrDPslfCCK/f+vuXtBLY0cNjLUC/atM55zrM7HPAI/hSg7c65xaZ2beAhc65ucDVZjYbv667Abg0+PZJwK/MLALk4Nd8a6PmHipavpHSV9ew+aSJdFbpD0k66xhayvpPTaf2V89Se8t81l55JJFyrT+UPZO7pYXq+1+nZPE62kZWsOETdbSNHhJ2WCmrqytm1UPqiplJcrYHNbsXdNXsnkbzlNqs2uOQdnKMzspiOiuLKXs3tS+A+5t8LwNOI3ad748Ab8W4Pybn3Dz8bHn0fddF/f+1wLUxvu8ZYGp/X0d60Rmh6oFFtFcXs+VY1YrOBO0jK9jwycOp+e18am99nrWXz0jbjSgSkoij7Pl3qfrrG9AZofHUA9h69ATN4vZHXi6NZx5I6zh1xUx76VizW9JOf8+mnwI3m1kbcBuwBhgJfAK4CvhsQqKThCh/diUF65pYf/E0yE+PTZbSt9Zx1Wy4uI6a2xdQc/sC1n/qQ6rCIP2Sv24b1fe9RtHKRpr3HUrD2VPpGBqjZbX0aveumPuz9dh9MrJUWibK1Jrdknr6W+f7N2ZWi5+R/mLUQy3A151zv0lEcBJ/OdtaGfLomzRPHE7z5Nqww5E4a5k4nI0fPYRhd77E8DteYP3FdYPeFCQZrKOTyifeorJ+uS8feN5BbJ82Wh+tD8KuXTGXBl0xD1FXzBSW1TW7JRQDqfP9HTP7KXAEMBTYBDzrnNuSqOAk/qoefgPr6KThjMl6g81QOw4aRUNzB0P//BrD/vQKGz96iN5EZDeFKxqovu81CtY3sf3gUTScMVm1iuMkVlfMjRdNo01dMVNO8ZJ1VN+/iLzNzTRNG03jqarZLYk3oEVMQaL9cIJikQQreLeRshdWs+W4vekYrtqimazpQ2PJaW6j6uGlRIrzaThzii62BPCluqr++gbl89+lY0gx6z55OC3714QdVubp1hVzhLpippTczc1Uz13kNxbXlLH28hm07q01+pIc/Uq+zexcYIhz7pbg9gTgDmAy8Bjwaefc5oRFKYMXcVTPXURHeSFbTtgv7GgkCbYetw85O9qpfPJtOkvy2XLy/mGHJCErXrSW6vtfJ3dbK1uPCsoHaiNZQu3WFXNFIw1nqStmaDojVPzzHSr/vgyco3HW/nGp2S0yEP39q/t14PdRt3+KbzF/I3AFcAN+46WkqLKFqyhcvYUNFx6iN9tsYcbmUw8gZ0c7Qx5fTqQ4n23HqLpNNsrd2kLV/YsoXbSWthHlbLi4jrYxKh+YLJGyQtZ/cjqVjy2j8vFlFLy/hQ0fn0bHMG1qTabuNbsbZk9RXwQJRX+zsL2B1wDMrBI4GTjbOfeQmb2LT8KVfKeonB3tDHlkKS3jq9lx8Kiww5FkMqPhnKnktLZT/dASIsX5bK8bE3ZUkiwRR9mCoHxgR8TP8h2zt8oHhiHH2HLSRFrHDmHYXS8z8qdPs/H8g2g+cGTYkWW83Wp2XzzNFxzQ8h8JyUCmQLv60B8HdAJ/D26vBlKv37B8oPLvb5Kzo42G2Vr3m5VyjI0fPYScloUMvfdVIkX5GdkGW3aVt76Jofe9RtGKBlr2Hsqmc6ZqpjUFtOxfw5qrjwm6Yr7IlmMmsHnWAbogSgTnKH1hNVV/fcPX7D52b7Z8eD99+iuh6+8Z+ApwkZk9B1wGPOGcaw0eGwusT0RwMnj5a7ZS/uwKts0YR/so1SvNWnm5bPj4NGpvmc/wO19i/ScPp2XfYWFHJYnQEaGy/i0qn1hOpCBX5QNTUOeQYtZeMcN3xXzqHQpXbWbjxw5TV8w4yl+3jeo/v+4vPsdV0XDWgarZLSmjv8n3fwAP4JvqNOGXnXQ5C5gf57gkHpzfZBkpzmfLSRPDjkZC5grzWH/p4dT+6jmG/34h6z4zQ+t+M0zBykaG3vcqBeua2H7QSBrOmEKkXGXTUlJ0V8z7XmPkT54KumLqongwdqnZXZTHpnOn0jRNNbsltfS3yc7TZjYWmAi81a2yya3A8kQEJ4NT8uoait5pYNPZU4mUFIQdjqSASEkB6z89ndqbn6Hmd8+z7oojaK8tDzssGSRr8fs6yp9bSWdFEesvraP5ADXRSgddXTGH3fEitb+dz+aT92frceqKuSdUs1vSxUCa7GwDXohx/7y4RiRxYa0dVD20mNa9Kn1dWZFAZ0UR6z/9IWpvfpaaW+az9sojteM/jRUvXkf1X14nd1sL244Yz+ZT9tea1jTTXlvO2quOYuh9r1H1SNAV8wJ1xewv1eyWdKMdHhmq8onl5G1t9c1VNIMi3XQMLWX9p6djbZ3U3jKfnG2tfX+TpJTcrS0Mu+MFan6/kEhJPms/eySNs6co8U5TrjCPjRceQsPsKRQv28CInz1FwXtqIN2rzggVT77FqB/+g6JlG2icdQBrrj5GibekPP2VzkB5G5qoeOptmg4bTdvYqrDDkRTVPqKCDZ88nJrfPk/trc+z9vIZYYck/eEcZQtWUTVvCdYRofGU/dl6rMoHZgQzth05ntbRlTu7Yp4xhabp6orZXeGKoGb3OtXslvSj5DvTOEf1g4txebk0nqqOhtK71nHVbLh4GjW3L6DmtgW07OP6/iYJTd6GoHzgOw20TKj25QOHl4UdlsRZ29gq1lx9DMPmvMTQP79G4Up1xeySs72NIX99g/KFqtkt6UvJd4YpXrKe4qUbaDhtEpFyla2SvrVMHM7Gjx7KsDtf5DcrYfRzf6OjusR/VZXs/P/qEjqGFGmGNQwd/uP1IY8vJ5KXw6ZzptJUpwoOmSxSWqCumNG6anbPW0JOS4dqdkta01mbSdo7qXpwMW01ZWw7cnzY0Uga2XHQSNZWHsl5jy/k4aoR5DU0U/D+VkoWrcU6d86Guxyjs7KI9upuSXl1CR1VxURKCzQDFWcF7zYy9N7XKFi3je1TR9Iwe7IurLNFV1fMcVUMm/NS1nbF3K1m99kH0j5CNbslfQ06+TazY4HrnXMnxCEeGYSKp94mv2EH6y77kGYnZcDaxlVx2IR8/jhz6s47I47crS3kNezY7atkyXpym3bdqBkpyN0tKW+PSs7J18fm/WWtHb584LMrfIWaS+r8x+uSdVomDt+lK+bWoyfQeGrmd8W0tg4qH1tOxVOq2S2ZJR4z38PxLeclRLmbm6l8Yjnbp45Q50KJnxyjc0gxnUOKY1YQsLYO8hqbydsUJOWNwX83bado2QZy2iO7HN9RUbj7kpah/r+dZYV6Uw0ULwnKB25tYduMcb58YJHKzmUz3xXzCKoeWkzF0+9QsDqzu2IWL15H9dyomt0fmeQ/WRPJAD0m32Z2ST+f4/A4xSKDUPXQEgAaT5scciSSTVxBHu215bEb9ThHTlMbeQ07yO82a1701iZyt76HRe3vdHk5dFQVx17SUl2SFWs7c7a1Uv3AIkpfXUNbbRkbPnYkbeNUsUgCeTm+K+b4aobe+2pGdsXcpWZ3bRlrrziC1gnVYYclEle9vZvdBjigP1NRKpEQoqLlGyl9bQ2bT5pI55DisMMR8cyIlBfSVl4YO4Hs6CRvc+wlLUUrGslp7djl8M7Sgl3Wl0cva+msTPONoM5RujDYTNbWyeaTJrLluH0gL41/JkmYHQePon1EeWZ1xeyMUPHPd6h8dBngaJx1AFuPnqDfAclIvSXfa4EHgC/28RxnA7fHLSIZmM4IVXMX0V5dwpZj9w47GpH+y8ulY1hp7OoNzpHT3E5eQ/OuiXnjDgpWb6bktTVYZNeNoB1DiqNmyot3mTWPFOen7EbQvI3bffnAtzfRMj4oH1ij8oHSu0zqirlLze5JNTScoZrdktl6S76fBaY557b39gRm1hzfkGQgyp9dScH6JtZfUqfNbJI5zIiUFNBWUkDb6MrdH++MkLulhbzG6CUtPlEvWbSW3O1tuxweKczbuflz6K4z5x1VxZAXwu9OZ4SKJ99myGPLcHk5bDp7Kk2HazOZ9F9XV8zWcVVUPbSYkT99ig0XTYv9O5OCYtbsnjIi7LBEEq635Psu4Lx+PMdi4FvxCUcGImdbK0MefZPmicNpnlQTdjgiyZObQ2d1CZ3VJbTus/vD1toRczlL/oYmipeuJ6dj50ZQZ9BZUdRtjXnxB8l6pKww7rPmBas2M/TeVylYu43tU0bQeOaUjN04JwkWqyvm7BTvihlxlL6omt2SvXo8051zdwN39/UEzrklwH/FMyjpn6qH38A6Omk4Y3Lq/pEVCYErzKN9ZAXtI2PUAo44cptaY681X7aRvK0tux6en7N7s6GoJN0V9D9hsNYOhvxtKeXPrKCzvFAzfRI3H3TFvOvloCtmAw1nTU25rpj5a7dR/ZfXKFrRqJrdkrV0mZmmCt5tpOyF1Ww5bh+1lxYZiByjs6KIzooiWsfHqKLQ3unLJwZrzPM3RSXnb28ip61zl8M7ywp22fwZXUKxs6Log2UkKzZ0MOpHT5K3uZltM8bSOOsAlQ+UuIqUFrD+0sOpfHwZlY8to+D9rWy46LCUeI/wNbuXUfHUO0SK8th47kFsnzZay6wkK/VWavBvwOedc0uj7jsBmN/XOnBJsIij+v5FdFQUsuWEfcOORiSz5OfSUVMWe9Ojc+TsaI85a164spGSV7ttBM31G0EjJQU8sKqVSE0Za688InbSLxIPOcaWEyfSOjboivmzf7LpvIPYMTW8rpi71OyuG03jqarZLdmtt5nvE4EPdm2YWS7wKL6u94sJjkt6UbZwFYXvbWHDhYdojZxIMpkRKS2grbSAtjFDdn+8M7Jr+cSupkObm/nQvvncfenR4WzulKwT3RVz+B3hdMXcrWa3LjxFgIEvO9HnQyHL2dHOkEeW0jKhmh0Hjwo7HBGJlpvjq6kM3b1M2vT6eu5W4i1J9EFXzHlLfFfMVUFXzMoEb+7tjFDx9DtU/n0ZAI2nBjW707kWv0gcado0zVQ+upScHW00nDFFmyxFRKR3eTk0zp5C67gq3xXzp0+x8cJDadk3MV0xd63ZXUvD7Ml0Vqlmt0i0vpLvWJ0r1c0yJPnvb6X8uZVsmzGO9lHaHS4iIv2z4+BRtI0sZ/gfXqTmlvh3xVTNbpH+6yv5fsTMOrrd91iM+3DOqdB0IjlH9dxFREoK2HLS/mFHIyIiaaajJlZXzIOJlAxi82P3mt3HBTW7B1CCUyTb9PbbodrdKaTklfcpWtHApnOmpmX7YBERCV9XV8yW8VVUP7iYkT99mg0XHUbb6BgbiPuwS83u8VU0nDWV9hHlCYhaJLP01mRHyXeKsNYOquYtoXV0JU11Y8IOR0RE0pkZTUeMp22vrq6Yz9IwezJN08f2ay+RanaLDI4+F0oDlY8vJ29rKxs+Pk1/3EREJC527Yr5OoUrGmk4u/eumKrZLTJ4Sr5TXN6GJiqefpumaaNpG1sVdjgiIpJBdnbFXE7lY2/6rpgf370rZm7jDqrnLqZkiWp2iwyWku9U5hzVDyzG5eXSOOuAsKMREZFMlGNsOXE/WscO2a0rZmfEUfGPt1SzWySOlHynsOIl6yl+cwMNp08mUl4YdjgiIpLBunfF3Hb4GO5a0kxV0xvsmFxLwxmq2S0SD0q+U1V7J1UPLqatpoxtR4wLOxoREckCu3TFfGYFbUXG+kvqaJ5cG3ZoIhlDnxulqMon3ya/YQeNs6fo4z0REUmeoCvm+1cfw0VHFyvxFokzZXUpKLdxBxX1y9k+dWTCWgCLiIj0pn1UBfl5qrAlEm9KvlNQ1bwlADSeNinkSFKT3gpEREQkXSn5TjFFyzdS+tpats7cl84hxWGHk3KKgQJAW35EREQkHSn5TiWdEarmLqK9uoQtx+4ddjQpqQCYDFyPT8RFRERE0omS7xRS/swKCtY30Xj6ZMjvucNYtioF/ht/0n4FuAPNgIuIiEh6UfKdInK2tTDk78to3n84zZNqwg4nJVUDl0fdPht4AhiCTmQRERFJD8pZUkTVX5diHREazpgCpi2F3ZUCv2D3wvTTgReB0fglKSIiIiKpTMl3CihY2UjZi6vZeswEOoaVhh1OyjH8Ou/Tenh8AvAycDBaBy4iIiKpTcl32CKO6rmL6KgoYsvx+4YdTUoqAm6m9xKDVcDT+ARd68BFREQkVSn5DlnZglUUvreFxtMm4Qq7L6qQPOAjwGH9OLYAuBu4Bs2Ai4iISGpS8h2inB1tDHnkDVomVLPjoJFhh5OS8oEfDuB4A76DXx+uBFxERERSjZLvEFU++iY5ze00zNYmy1iKgX8Fxu7B914KzAPKUUdMERERSR1JT77NbJaZLTWz5Wb2tRiPX2pmG8zs5eDrsqjHPmFmy4KvTyQ38vjKf38r5c+tZNuMcbSPrAg7nJSUD1w3iO+fCcwHaoLnEhEREQlbUhcZm1ku8HPgJGA1sMDM5jrnFnc79C7n3Oe6fW818E2gDnDAC8H3NiYh9PhyfpNlpKSALSftH3Y0KakUuAEY7GXJJOBV4MPAcqBlkM8nIiIiMhjJnvmeDix3zr3tnGsD5gBn9vN7TwEedc41BAn3o8CsBMWZUCWvvE/RigY2n7I/kRLNycZSBVwRp+eqwc+Az0SVUERERCRcyU6+9wJWRd1eHdzX3blm9qqZ3WNmYwb4vSnNWjuomreE1tGVNNWN6fsbslAp/uOReF6WlAAPAZ9CCbiIiIiEx5xzyXsxs/OBU5xzlwW3LwamO+c+H3XMUKDJOddqZlcCFzjnTjCzrwCFzrnvBMd9A9jhnPtBt9e4nKALeW1t7bQ5c+Yk5WeL1g40NjWxqqxst8f+ubSNF99p5/wZRYwYkpv02NJBCX65SE+ampooizG2/bUeeA+I7PEzZK7RTU2sHsTYSs80tomjsU0cjW3iaGwTZ3xTE0NDGNvjjz/+BedcXV/HJbuw9Gogerp3NPB+9AHOuU1RN38DfC/qe2d2+9767i/gnPs18GuAuro6N3PmzO6HJNwq4J76er7c7bXzNjQx6tEn2T5tNDeddXDS40oHxcCT+IX9Pamvr2ew/64PARcAOwb1LJnnphjnrcSHxjZxNLaJo7FNHI1t4txSX8+5KTy2yV52sgDYz8wmmFkBcCEwN/oAM4sueD0bWBL8/yPAyWZWZWZVwMnBfenBOaofWIzLy6Vx1gFhR5OS8vCL+Pu8ZIyD0/AdMasBff4gIiIiyZLUmW/nXIeZfQ6fNOcCtzrnFpnZt4CFzrm5wNVmNhvoABrwJZtxzjWY2bfxCTzAt5xzDcmMfzCKF6+j+M0NNJw+mUh5YdjhpKR84EdJfL1DgZfxH6e8B7Qm8bVFREQkOyW9n7lzbh6+/0n0fddF/f+1wLU9fO+twK0JDTABrL2TqocW01ZbxrYjxoUdTkoqAq4Ekj06Y4CXgNOBF9AyFBEREUksdbhMgoon3ya/odl3sszVkMeSjy/iHoYK4DHgXFQJRURERBJLmWCC5TbuoOKJ5Ww/aCSt+wwLO5yUVAp8B6gMMYZ84Hb8Ry7FIcYhIiIimU3Jd4JVPbQEzGj8SG/F87JbJfDZsIMADPg6fl2TEnARERFJBCXfCVS0bCOlr69ly/H70DlE6VwsiWioM1gX4tunVqBfEBEREYkv5RYJ0hlxVD2wiPahJWw9Zu+ww0lZE4Ezww4ihqPwGzBHAgUhxyIiIiKZQ8l3gryysoOC9U00nj4Z8lVJOpZi4Gb8co9UtC/wCjAFLUMRERGR+FDynQCbtrUwf3kbO/YfTvOk2rDDSUl5+C5J08MOpA9DgWfwsZaGHIuIiIikPyXfCXDzX5fSGYHGM6aEHUrKygN+HHYQ/VQE3IffFKpShCIiIjIYSr7jrK0jwtrGHRw6IZ+OYZorjaUIuAIYH3IcA5ED/A/wv2gJioiIiOw5Jd9xVpCXw08un8GMfVOpfkdqyQeuDzuIPXQZMBctQREREZE9o+Q7AcyM3JxU3UYYrlLgv4AhYQcyCCcCzwHD8MtnREREJBz5+NLA+cBY4GNAqu+2U+4gSVUBXBV2EHFwIPAqcDywEmgJNxwREZGsUAZE8JXSDgVOwpcHPhyfYwDUhxJZ/yn5lqQpBX5G1Y2NlQAAIABJREFU5tTNHgksBM4CngV2hBuOiIhIRinA7xNrBsYBxwEzgRnAPqRuqeK+KPmWpNkXODvsIOKsDHgE+FfgDygBFxER2ROGf0/twCfdh+NntY8AppFZ1caUfEtSFAO/In2vUnuTi//ZJgLfwF+hi4iISM+K8El2M7AffhnncfhZ7dFkZr7QRcm3JFwefpPih8IOJMG+hP8Y7CI0Ay4iItIlBz+r3Ypflz0DnxfMAA4GCsMLLRRKviXh8vD1sbPBWfiNHicDW/GbQkRERLJJCf5T4TZgMvBh4Gh8sp3qlUiSQcm3JFQRvjb2hLADSaLDgZfwm0LW4P/4iIiIZKJcfEGFFmA4cCQ+2Z4BTEGJZiwaE0moPOBbYQcRgvHAy8AsfElCrQMXib8c/AV+KX6mTcu9RBKvq8lcBL9k5ER8qb/pQHVYQaUZJd+SMKX4TpZVIccRliHAk8DFwIMoMRCJpxJgDHA//hOm+4DLgU3A9hDjEskk+fiCCc3AXsAxwAn4We2JqFPjnlLyLQlTDnw+7CBCVgDMAb4J/AAl4CLxUIxfzvY/+N+xNcApwHLg5/iqQ+34zV0i0n/lQCd+Kck0dpb6qwsek/hQ8i0JUQr8lMxpqDMYhl96sy9wJVqCIrKnCvB/W+7Gf9TdXT7wb/hPm74K/BG/DtUlK0CRNFIYfDXj92XNZGepvwlkdqm/sCn5loTYGzg37CBSzCX4Dl1nAE0oIRAZiFJ8UnAnflNXb4YCvwW+AHwGeB0tRZHsZviZ6zb8J0fT2TmrfWhwnySPkm+JuxIyt6HOYB0HPI9vJrAJ/9G4iPTM8InBjcBVDOzvylTgWWAucAWwDS39kuxQjE/wWoH98eu0j8H32xgdYlziKfmWuMrDJ5ZHhB1ICjsAeAVfimk5/mNxEdldMTASnzxP2cPnMOBMfOWhHwA34C96deErmaKrgU0LvsBB9wY2+eGFJj1Q8i1xlU0NdQajBj8Dfi7wDzQbJ9JdCfBx/N+Tojg8XyHwH8CngGvwVVK0/0LSUQk+4e4ADmRnA5sP0feSLEkNSr4lborwb2z7hB1ImijGlyC8Br8+VQm4iJ+lKwHuAE5LwPOPwK8bXwh8GngLrQeX1JWH/31oxneGPJqdpf4m46uSSPpR8i1xkwd8O+wg0kwOfmZvIvAVNBMn2a0Uv/nrT/gkOZHq8I2w7gI+h7/41e+fhK0MvxnfAYcAJ+M7Rh6O7x0hmUHJt8RFKb6Wtbpb7Zmr8BVizkezcJKdivFNub5I8hp3GHAhMBu/FvxH+LXgHUl6fcluBlTgL/rGAsfiy/3NAPZDRQsymZJviYsy1FBnsE4Fnsav39uCb3QgkumKgWH4TZWHhBRDCT75vgL/d+zvaBmYJE4ZfsKqazNxHTtbtkt2UGdQGbSuhjqFYQeSAQ7BV0IZj8ZTMl8JcAHwBuEl3tHG4jdiPoxfCqaESOLF8OfTVOB24D188n0cOs+ykZJvGbQJwHlhB5FBRgMv4psglIQci0gi5OEbfvwBuI3UO8+PARYDPwYqiU+1FclO+fjzZxb+E5VXgXPQRslsp+RbBkUNdRKjAngcvwY81RITkcEoAQ7DJ7dnhxxLb3KBy4B38ctRilHCJP1XHHx9ClgEzMOv5RYBJd8yCLn4j8yODDuQDJUH/A74Omr9K5mhGLgWeIb06bJXgZ8BfwW/GU5LBKQ3ZfhGN1/HLy25Gb+ZXiSaNlzKHisAfhJ2EBnO8MnKPsClqBSapKcifCWkv+BLpqWj/fDLBh4FPgNsRJWJZKcyfB3ubwIfxb8/ivREM9+yRwrxyeC+IceRLS7Av/FXol9aSS8l+Pbub5C+iXe0k4DlwH/jEy5tjM5eufhPc47GX1guAy5Gibf0Te/jskfUUCf5jgReAEahP+6S+nLxyelvgTn4DZaZIg+4GliBT7aK0ZtpNinEf5pzAbAAeApfIlZ7n6S/9PdCBqwUuA4YGnYgWWgffFe+A9E6cEldJfiSaq8B/xJyLIk0FPgNPgGbgdaDZ7pS/AXlvwHvAH8EpoQakaQrJd8yYKXAF8IOIosNxW9Ym4Xe7CX1FAPX4BPS8eGGkjRT8A2y/oiv3awKRZmlDP+J4/eAdcCNwIhQI5J0p+RbBqQUv8lS6xzDVQjci29Lrzd6SQWFQA1+Q+J3yL7d/IZvU/8OvtJFCVoels4M/294CL4e/Sr091biR8m3DMh4/Do3CZ/hZ2L+Fy1BkXCV4D+JWQocFXIsYSvEVyh6CzgX/7uptcDpo6spzmlAPfASfsOwkiWJJ51P0m8l+JqleiNJLZcBc/EfjerfRpIpB/9p2M+BPwNDwg0npYzAL0N5CjgILRFLdcX497jPAEuAB8iM6jySmpR8S7/k4lsuHx12IBLTicBzwDCy7+N+CUcJcAB+A/Cl6MKvJ9Pws6e34H8/tWwhtZTha9BfD7yPv5AcH2I8kh2UfEu/5AM/DTsI6dUU4FV8M5CikGORzFYMfBafeKvWf98M33hlJX4zagm6SA5bKf7cvRlYC/w7vo+CSDIo+ZY+FQKX4JM6SW0j8FUmjkYzbBJ/BfhZwnnATfiLcum/Evxm1Dfwa4r1O5pcXU1xjsMvK3kTuAidx5J8Sr6lT3nADWEHIf1WCjyMv2DSm7vESylwAj5hmRluKGlvDL4j4t+A/dF68EQrCr4+hm9UVg8cj5ZKSXiUfEuvSoH/xK9VlPSRC/wSP8umSigyGF0l136An/FWc634OQpYjC/fOgT9rsZbKb6z6pfwS35+D0wKNSIRT8m39KoEv0ZR0tM1+NbemgGXPVGCXxf7AnAFmilMhBzgU/jk8Ep8Ap4bakTprwzYC3/BuA4/CVETakQiu1LyLT0qxdeQ1ua99DYb+Ad+Zk1v6tJfJcAn8S3iDwg5lmxQAfwQP94noAvmgcrBj9k04E7gXfwFoz5NkFSk5Ft6NBa/Q1/SXx2+MsUY1HVPeleAv1C7D/gZ6mabbPvg14Lfjy95p/XgvSvATxDNBp4EFgKno+RGUpvOT4mpq6GOTpDMMQ5fb/gwNKsmsZXi1yEvBU4JOZZsdyKwDLgRv25ZF0G7Kgm+rsCfr3/Gz3qLpAPlVrKbXPwb8LFhByJxNwS/BGU2SsBlJ8N/PH8D8BhaH5sq8oDPASuAT+D/jbL9TbsMXwDgemANfrPq2DADEtkDSf89NrNZZrbUzJab2dd6Oe48M3NmVhfcHm9mzWb2cvB1c/Kizi4F+I+bJTMV4NtefwUl4OITunHAfOALaFNlKqoGfoVfUnEE2bcUxfA/8wHAb/BJ91fw6+RF0lFSm2yZWS6+e+tJwGpggZnNdc4t7nZcOXA1/v0g2lvOuUOSEmyWKsQ3HZgYdiCSUIafOdoXuBxoDjUaCUsJ8C/47rXamJb6JgNPAQ/hf2+3ADtCjSix8oKvI4FvAsegi0PJDMme+Z4OLHfOve2ca8NXQTszxnHfBr4PtCQzOPFLTv477CAkaT6Ob8hTjt7UskkeftZwDvBblHinE8NvKFwBXIefEc60TdRdTXE+DryCXwp1LPobJZkj2cn3XsCqqNurg/s+YGaHAmOccw/G+P4JZvaSmf3DzI5JYJxZqQTfUGd42IFIUh2Lb0k/Ar25ZYNS4HBgCXBGyLHInisAvgq8BZyPv4BK99/fMvxF4b/jE4XfoU9hJTOZcy55L2Z2PnCKc+6y4PbFwHTn3OeD2znA48ClzrkVZlYPfNk5t9DMCoEy59wmM5uG7847xTm3tdtrXI7/RI7a2tppc+bMSdaP94F2oLGpiVVlZUl/7cHIAw4i9f+ANzU1UZZmY5sOOoCtTU28V1ZGO/48iIQcUyYZ3dTE6pDP2xxgJP5CK5Pob4JfOrYC/3FxPH9vk3He5uDff0biO6im+ntQvOi8TZywxvb4449/wTlX19dxSV3zjZ/pHhN1ezTwftTtcuBAoN7MwL9HzDWz2c65hUArgHPuBTN7C39RvDD6BZxzvwZ+DVBXV+dmzpyZmJ+kF6uAe+rr+XIIr72nSvEbeo4PO5B+qK+vJ4x/12xQX1/POTNn8i7wCPAn/BrTAmA70BlmcGnuphD/JhThK0Tcjy81mWn0N8Fz+N/Zz+F/X+OxHjxR520O/rw8EL+eexbZV8lF523ipPrYJvtcXwDsZ2YTzKwAuBCY2/Wgc26Lc26Yc268c2488BwwO5j5Hh5s2MTM9gb2A95OcvwZazR+45UI+NJdn8E3+9gKPAh8Eb9BswD/8bCkhxLgPOANMjPxlp0MuADfqv7L+H/7/FAj2l1XU5yzgH/iqyp8hOxLvCW7JfV8d8514C/KH8EvObzbObfIzL5lZrP7+PZjgVfN7BXgHuBK51xDYiPODmqoI73Jx1cZ+D6+6ce7wC/wO6XL8B9XpdobvPiPNcuB3wP/R/aVp8tmxcB/4ZvPnE5qlBQtwZ+D/4r/O3IvoNJlkq2SvewE59w8YF63+67r4diZUf9/L/73VeIoB183dmbIcUj6qAUuDr46gRfxpc/uwb+pFgLbQotOwCc6k/Et4sf0caxkrtH4c+AZ4DL8hfP2JMdQjp/pvjaIoTzJry+SijTZmeXUUEcGIxdfOeN64HVgLXALfj3ZEPxMl9piJ1cxvgrGcyjxFu9I/O/nz/C/l4kuLdnVFGcy/u/B+8A1KPEW6aLkO4sVAB/Ddw0TiYcqfNmzO4EGfAL4baAOf75VkD2VDJKtCL9D/Ql8/efccMORFJMDXIqf/f4sPgGP9zmShz8PT8SvLX0d//cg6R+xi6Q4Jd9ZLA/4bthBSMYyfCWDr+B3Wm8C7gA+BdTgl0aouUt8lODX9i4FPhRyLJLayoEf4BPjDxOf9eDF+KT7E8Br+I3aR6ELbZGe6II0S5UAX8MnQSLJUIZPEE8Pbi/Hd9e8G3gevzylCdUWH4hcfNJzM74boEh/7Y2fnX4cX9loHQNfD16GPwe/CFyFr9EtIn3TzHeWKsaXohIJy7740kdPApvxu6mvwpc5LELVOfrStanyNZR4y547Af+JyffYuTmyL2XAeOAn+KT9OpR4iwyEku8sVAr8EH3kL6mja53oT/A1it8EfgycjD9PK9DHdNGKgc/jK81MCDkWSX95+Avflfh14cXsnhzkBvcfga9s9DbwSbShWmRPKPnOQqPQTJmktjH4smSPAFvwpQy/hO+sVUD2Vk0oxHeqfAS4EV2QSHxVAb8EXsBXSCnFr9suAs7Fb6B+BjgFrecWGQz97c4yaqgj6SYfODr4uhFYj9/QdS/wWHBMK9AWSnTJUwocj2+aUxVyLJLZJuGXg83Dd7h9Cz9pIyLxoRwsi+TgKyGcEHYgIoNQg//k5s/4teKPA/8JTCUzZ8Vz8BfN/wvMRYm3JIcBpwEjUeItEm+a+c4iBcDPww5CJI5y8DXE6/Cbvhrxs+F/wc/ateG7cLaEFeAgleA3oN4PTAw5FhERiQ/NfGeJAnzXwUlhByKSQFXAecAf8HXF5+Ob/Ewn/Zr8FAOXA6+gxFtEJJNo5jtL5OHXy4pkCwOmBF9fxtcwrsfPIj+A38gJ0BxGcL0owK/v/hO+CYqIiGQWJd9ZoAT4d6A27EBEQlSKX8N6WnB7Ob5qyN34GfJCfILeGUp0Xim+ysQf8VVNREQk8yj5zgJF+BbfIrLTvsHXVfhqKU/jZ8T/gm8cksvAO/7tKcMvM/k+8FnSZ2mMiIgMnNZ8Z7hS4Af42W8Ria0Qv8Tjx8AKYBm+usgsEt/kpxjf6nsB8K8o8RYRyXRKvjPcSODisIMQSTOjgU8Df8XXOZ6HXzc+EZ+ox6ucYQm+o+Dr+FbxIiKS+bTsJIN1NdTJDTsQkTSWBxwVfH0X3+TnUXyTn78DDl/ScCBNfvLxv593AqfGM1gREUl5mvnOUDn48mqqliASXzXARcB9+CY//wC+ARzEznKGvSkFZgBLUeItIpKNlHxnqELUUEck0XKAw4Cv4+txrwd+h+/AWY1PtIuiji8GvoUveajqQyIi2UnLTjJQAXABWkMqkmyVwDnBlwOW4NeN/wmfhD8LHBxadCIikgo0852B1FBHJHyGvwD+EvAcvtmPEm8REVHynWFK8DW9R4QdiIiIiIjsRsl3hinEd7MUERERkdSj5DuDqKGOiIiISGpT8p1BRgCXhB2EiIiIiPRIyXeGKEUNdURERERSnZLvDNBVa/jEsAMRERERkV4p+c4AhcAvwg5CRERERPqk5DvNFQDnAQeGHYiIiIiI9EnJd5rLA74XdhAiIiIi0i9KvtNYCfBFYGTYgYiIiIhIvyj5TmMFwNfCDkJERERE+k3Jd5oqBW4K/isiIiIi6UHJd5qqBS4NOwgRERERGRAl32moFPglaqgjIiIikm6UfKcZAw4FTg47EBEREREZMCXfaaYINdQRERERSVdKvtNIAXA2MDXsQERERERkjyj5TiO5wP+EHYSIiIiI7DEl32miGLgGGBV2ICIiIiKyx5R8p4lC4NqwgxARERGRQVHynQZKge8BZWEHIiIiIiKDouQ7DQwHPh12ECIiIiIyaEq+U5wa6oiIiIhkDiXfKcyAg4BTwg5EREREROJCyXcKK8LPelvYgYiIiIhIXCj5TlH5wJnAwWEHIiIiIiJxo+Q7ReUBN4UdhIiIiIjElZLvFFQMfAHYK+xARERERCSulHynoALgP8IOQkRERETiTsl3iikFbgTKww5EREREROJOyXeKGQZcFnYQIiIiIpIQSU++zWyWmS01s+Vm9rVejjvPzJyZ1UXdd23wfUvNLOPKX5cCv8BvthQRERGRzJPUPM/McoGfAycBq4EFZjbXObe423HlwNXA/Kj7JgMXAlOAUcDfzWyic64zWfEnkgFTgVPDDkREREREEibZM9/TgeXOubedc23AHHw56+6+DXwfaIm670xgjnOu1Tn3DrA8eL6MoIY6IiIiIpkv2cn3XsCqqNur6VZRz8wOBcY45x4c6Pemq3zgDOCQsAMRERERkYRK9vLiWBO77oMHzXKAHwGXDvR7o57jcuBygNraWurr6/ckzkFpB/ZqauKmfr52DnAg0L+jpampKZR/12ygsU0cjW3iaGwTR2ObOBrbxEn1sU128r0aGBN1ezTwftTtcoI81MwARgBzzWx2P74XAOfcr4FfA9TV1bmZM2fGMfz+WQXcU1/Pl/vx2sXA54FrEhxTJqmvryeMf9dsoLFNHI1t4mhsE0djmzga28RJ9bFN9rKTBcB+ZjbBzArwGyjndj3onNvinBvmnBvvnBsPPAfMds4tDI670MwKzWwCsB/wfJLjj7sC4OthByEiIiIiSZHUmW/nXIeZfQ54BMgFbnXOLTKzbwELnXNze/neRWZ2N7AY6ACuSvdKJ6XAf6OGOiIiIiLZIuklpZ1z84B53e67rodjZ3a7fQNwQ8KCS7JqgsXpIiIiIpIV1OEyJGqoIyIiIpJ9lHyHwIDJwGlhByIiIiIiSaXkOwRFwM2ooY6IiIhItlHynWT5wEeAw8IORERERESSTsl3kuUBPww7CBEREREJhZLvJCoGrgLGhh2IiIiIiIRCyXcS5QPfCDsIEREREQmNku8k6WqoUxF2ICIiIiISGiXfSaKGOiIiIiKi5DsJSoGf45ediIiIiEj2UvKdYAZMAk4POxARERERCZ2S7wRTQx0RERER6aLkO4HygFOBaWEHIiIiIiIpQcl3gjj8Gu8fhR2IiIiIiKQMJd8J4oDPooY6IiIiIrKTku8EKAEKgOvCDkREREREUoqS7wQYCkwFKsMORERERERSipJvEREREZEkUfItIiIiIpIkSr5FRERERJJEybeIiIiISJIo+RYRERERSRIl3yIiIiIiSaLkW0REREQkSZR8i4iIiIgkiZJvEREREZEkUfItIiIiIpIkSr5FRERERJJEybeIiIiISJIo+RYRERERSRIl3yIiIiIiSaLkW0REREQkSZR8i4iIiIgkiZJvEREREZEkUfItIiIiIpIk5pwLO4aEMbMNwMqQXn4YsDGk1850GtvE0dgmjsY2cTS2iaOxTRyNbeKENbbjnHPD+zooo5PvMJnZQudcXdhxZCKNbeJobBNHY5s4GtvE0dgmjsY2cVJ9bLXsREREREQkSZR8i4iIiIgkiZLvxPl12AFkMI1t4mhsE0djmzga28TR2CaOxjZxUnpsteZbRERERCRJNPMtIiIiIpIkSr7jwMxuNbP1ZvZ61H3VZvaomS0L/lsVZozpqoexvd7M3jOzl4Ovj4QZYzoyszFm9oSZLTGzRWb2heB+nbeD1MvY6rwdJDMrMrPnzeyVYGz/K7h/gpnND87bu8ysIOxY000vY3ubmb0Tdd4eEnas6crMcs3sJTN7MLit8zZOYoxtSp+3Sr7j4zZgVrf7vgY85pzbD3gsuC0Ddxu7jy3Aj5xzhwRf85IcUyboAL7knJsEzACuMrPJ6LyNh57GFnTeDlYrcIJz7mDgEGCWmc0Avocf2/2ARuDTIcaYrnoaW4CvRJ23L4cXYtr7ArAk6rbO2/jpPraQwuetku84cM49CTR0u/tM4Pbg/28HzkpqUBmih7GVQXLOrXHOvRj8/zb8H6290Hk7aL2MrQyS85qCm/nBlwNOAO4J7td5uwd6GVuJAzMbDZwG/Da4bei8jYvuY5sOlHwnTq1zbg34N2OgJuR4Ms3nzOzVYFmKlkYMgpmNBw4F5qPzNq66jS3ovB204OPll4H1wKPAW8Bm51xHcMhqdLGzR7qPrXOu67y9IThvf2RmhSGGmM5+DPw7EAluD0Xnbbx0H9suKXveKvmWdPRLYB/8R6NrgB+EG076MrMy4F7g35xzW8OOJ5PEGFudt3HgnOt0zh0CjAamA5NiHZbcqDJD97E1swOBa4EDgMOBauCrIYaYlszsdGC9c+6F6LtjHKrzdoB6GFtI8fNWyXfirDOzkQDBf9eHHE/GcM6tC94kIsBv8G/AMkBmlo9PDu9wzt0X3K3zNg5ija3O2/hyzm0G6vHr6oeYWV7w0Gjg/bDiygRRYzsrWEblnHOtwO/QebsnjgJmm9kKYA5+ucmP0XkbD7uNrZn9IdXPWyXfiTMX+ETw/58A7g8xlozSlRwGzgZe7+lYiS1Yb3gLsMQ598Ooh3TeDlJPY6vzdvDMbLiZDQn+vxg4Eb+m/gngvOAwnbd7oIexfSPqYtzwa5J13g6Qc+5a59xo59x44ELgcefcRei8HbQexvbjqX7e5vV9iPTFzO4EZgLDzGw18E3gRuBuM/s08C5wfngRpq8exnZmUDbIASuAK0ILMH0dBVwMvBas8QT4D3TexkNPY/svOm8HbSRwu5nl4ieP7nbOPWhmi4E5ZvYd4CX8xY8MTE9j+7iZDccvk3gZuDLMIDPMV9F5myh3pPJ5qw6XIiIiIiJJomUnIiIiIiJJouRbRERERCRJlHyLiIiIiCSJkm8RERERkSRR8i0iIiIikiRKvkUkrZjZ9Wa2Mcb9x5tZq5nNCWq7ioiIpBwl3yKS9sxsKvBn4J/AJU41VEVEJEUp+RaRtGZmY4G/4hvXnOWcaws3IhERkZ4p+RaRtGVmVfjEuwM41Tm3tdvjJWb2EzNba2YtZrbAzE6O8TyXmpmL8TU+eHxm9O3gvlOD++qj7qs3s3u6PXes+442s3+Y2Q4z22RmvzGz8m7HjDOzO81sY3Dcq2b2seCxWLF2fa3oFnPXV6OZ/dXM9uv2OieY2fxgfNaZ2S/MrKyPcb++l9e/Leq428xsoZmdZWZvBK/xtJlN7vZ8zswujbo93My2mZmLum+Gmb1gZluC8XjZzC6Ierzr5z2w23Pf1DUmwe2RZnarmb1tZs1m9qaZfcfMCqKOGR8818yo+64ws7bo86c/51fw7981Nu1m9paZfam38RWRzKb28iKSlsysCJgL7Acc5JxbE+Ow3wCz8e3dlwOfAR4ys+Odc0/HOP4EoBnfIv6mXl7bgO8CnXsQ91HAY8BfgPOAocCNQFVwGzOrAZ4FdgBfBlYBBwJjgqc5IuopfwGsAb4d3G7t9pIXAW8H3/s/wP8BM4LXmQw8DDwKnBsccyOwNzCrjx9lS4xjfhfjuHHAD4Fv4Mf2v4BHzGw/51xLD8/9DaC4231bgZ8DK4EIcAZwp5ktcc691kes0YYBDcAXgUZgInA9MBy4ItY3mNmZwM+ATzrn/hb1UH/PryeCYwqA84GbzGyRc+7hAcQtIhlCybeIpKMc4A5gOpAPHAy8EX2AmU0C/gWfMN0e3PcI8Co+uTsl6vD84L/POudazGxYH6//MXyi+gA+ae7SDJT28b03As845z4aFet7wGNmdqBz7nXgGqASmBZ1UfFY1/HOueeivncrsCH6vm5eDZ7zOTPbF/h61GPX4ZPZ2c65zuD5GoC7zOwI59yzvfwcHd1f08y2xzhuGHCmc+6Z4JgXgLeAS4Gbux9sZhPwSfDtwKeifubFwGIzywMKgWr8OO0H9Dv5DhL1L0e93j+B7cCtZvb57suWzOxI4E7gP5xzf4i6fyDnV0PXWJnZU/gkfSr+wkdEsoyWnYhIOqoCzsInP3cBPzKzim7HHA4Y8KeuO5xzkeD20d2O7Zpl7XO9eLA84dvAf+Nnf6O9DhxrZtPNLC9IFC3qe0vws9Z3dz0eHPM00A5MCw49AXi4h9n8gcoNXmc8cCbwVNRj04E/dyXegXvxy3i6j9GeWt+VeAM451YCLwSvHcu3gb91izNaO9AE3APMZ/cENrfb2O5S+ca8fzOzxWbWHDzfHfiEfmy355qEv8B60Tn3P90eG8j5ZUE8JcAn8Rd7/+zh5xORDKfkW0TS1dXOufvws5+l7Fx20WUk0OSc29Ht/nVAiZkVRt1XDTQGyVNfrsB/avjzGI/dBCzDJ4XtwdexUY9XAbn4pSLtUV+t+ISsa1nJUPxSknh4OXiNd4Aa4HNRj43Ej8cHgkR8E35M4mF9D/eN7H6nmR2h3EJYAAAD5UlEQVQMfBS4tpfnOxw4ET8b/QD+04ZoXT9v19cXuz3+b8AP8NVxzsRfBFwVPFbU7dgf4z9ROTJYLhRtIOfXOUEs24HfAt+NviARkeyiZSciko42Oed+DuCcW2Nm3wB+aGa3OedeCo5ZA5SZWUm3BKkW2OGci14bvTe+WkpfyvDLNr4aLE/Z5UHn3DozOwzYF+iaif9V1CGbAYdfYzwvxvO/3/XzESM53UMX4pd5lAOfxy9vmeKca8KPUU30wWaWi0/+G+L0+jU93Lcoxv3fBe5wzr1uZnWxnsw5tzCI86kgxiXAfVGHdP28Xb4AHBN1+3zgT865/+y6o/sG0ChPAh8BbgN+Y2aHRp03Azm/Hge+ir/wmgz82MzWO+d+0sPrikgG08y3iGSCnwOvADebWdfftQX4RPe8roOCjZLn4Zd5dN1XBMzEb3Dsy5eBDcDvezrAORdxzr3pnFsYJIrboh7bDjwH7N/1eLevruT7MeAUM6vtR0x9WRQ89xPAN/FLK7qWt8wHzg4S7i7n4CdmYm1I3RM1wbpp4IPSkIcBz3c77jj8cpvr+vm8+fhktvtSkUXRY0q3mX38EqNYm1JjucE5146fLa8B/jPqsX6dX4HGIJ75zrnfAX8Hzu71pxORjKWZbxFJe865TjP7LD6Bvhy42Tm3xMzuBH4WrAfvqkZxAPBZgGB29Qb8bOUv+/FSF+NrifdneUpP/h0/+xzBr1vehk8gTwP+0zn3JvAj4BLgKTO7AV/tZBJQ6pz7/gBf7yDzpQPL8UtO2vBLYwC+A7wE/MXMfgmMBr4HPNLHZsuB2Aj8X/DpRDPwLfyyk9u6HXcJ8GPn3LuxnsTMvoJPmhcFP8s1+Amkx2Id34tHgavNbD5+hvwi/CcVPXLObTCzLwK/NbO7nHOL+nN+Rak2sxlBvJOBk/AbSkUkCyn5FpGM4Jx73sx+DXzXzO5zzq3HJ0Pfw1efGIKvinF6VBm4S/AzqCcGFUH68qxz7oFBxvm0mR2LL7n3f/jZ25X4jYPrgmM2BGuMv49fd1yIT5i/uwcveUfw3yZgMf7i4f3gdRaZ2an4zaP34cv53Ym/QIiXlcHz34gvO7gQ+JcYZQabguN6sgn4En6JUDP+3/K0AZYZBJ/8D8dfeID/ua/Grx/vkXPu92Z2ET4BPyq4AOvr/OpyPP7CMAKsBf5I7+vaRSSDmbowi4hIIphvuHOgcy7m+m0RkWykNd8iIiIiIkmi5FtEREREJEm07EREREREJEk08y0iIiIikiRKvkVEREREkkTJt4iIiIhIkij5FhERERFJEiXfIiIiIiJJouRbRERERCRJ/h/3wH0tMSzVDgAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a20582518>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.svm import LinearSVC\n",
    "SVM_clf = LinearSVC()\n",
    "dummy_param = {'penalty':['l2']}\n",
    "score = []\n",
    "score_std = []\n",
    "for sub_features in range(10,50,5):\n",
    "    tmp_score = []\n",
    "    tmp_score_std = []\n",
    "    for i in range(20):\n",
    "        cols = np.random.choice(data.columns, sub_features, replace=False)\n",
    "        data_small = data[cols]\n",
    "        classes_small = classes\n",
    "        clf = GridSearchCV(SVM_clf, dummy_param, cv = 3, n_jobs=1, \n",
    "                   verbose=0, scoring=make_scorer(f1_score,average='micro'))\n",
    "        clf.fit(data_small, classes_small)\n",
    "        tmp_score.append(clf.cv_results_['mean_test_score'].mean())\n",
    "        tmp_score_std.append(clf.cv_results_['std_test_score'].mean())\n",
    "    score.append(np.mean(tmp_score))\n",
    "    score_std.append(np.mean(tmp_score_std))\n",
    "#graph \n",
    "plt.figure(figsize=(12,8))\n",
    "subfeatures = True\n",
    "x = range(10,50,5) if subfeatures else range(12,36,4)\n",
    "y = np.array(score)\n",
    "plt.plot(x, y)\n",
    "plt.fill_between(x, y-score_std,\n",
    "                 y+score_std,\n",
    "                 color = 'cyan')\n",
    "plt.title('{} F1 (micro averaged) score with a confidence interval'.format(\"SVM\"), fontsize=15)\n",
    "if subfeatures:\n",
    "    plt.xlabel('Количество признаков', fontsize=15)\n",
    "else:\n",
    "    plt.xlabel('Объем выборки', fontsize=15)\n",
    "plt.ylabel('F1 score', fontsize=15)\n",
    "plt.legend(loc='best', prop={'size': 18})\n",
    "plt.grid()    \n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No handles with labels found to put in legend.\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a200712b0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "score = []\n",
    "score_std = []\n",
    "for s in range(12,36,4):\n",
    "    # Для каждого размера фитим 20 раз\n",
    "    tmp_score = []\n",
    "    tmp_score_std = []\n",
    "    for i in range(20):\n",
    "        data_small = data.sample(s)\n",
    "        classes_small = classes[data_small.index]\n",
    "        clf = GridSearchCV(SVM_clf, dummy_param, cv = 3, n_jobs=1, \n",
    "                   verbose=0, scoring=make_scorer(f1_score,average='micro'))\n",
    "        clf.fit(data_small, classes_small)\n",
    "        tmp_score.append(clf.cv_results_['mean_test_score'].mean())\n",
    "        tmp_score_std.append(clf.cv_results_['std_test_score'].mean())\n",
    "    score.append(np.mean(tmp_score))\n",
    "    score_std.append(np.mean(tmp_score_std))\n",
    "#graph \n",
    "plt.figure(figsize=(12,8))\n",
    "subfeatures = True\n",
    "x = range(10,40,5) if subfeatures else range(12,36,4)\n",
    "y = np.array(score)\n",
    "plt.plot(x, y)\n",
    "plt.fill_between(x, y-score_std,\n",
    "                 y+score_std,\n",
    "                 color = 'cyan')\n",
    "plt.title('{} F1 (micro averaged) score with a confidence interval'.format(\"SVM\"), fontsize=15)\n",
    "plt.xlabel('Объем выборки', fontsize=15)\n",
    "plt.ylabel('F1 score', fontsize=15)\n",
    "plt.legend(loc='best', prop={'size': 18})\n",
    "plt.grid()    \n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Вывод\n",
    "* SVM сходится за одну итерацию\n",
    "* Из графика зависимости от обьема выборки качество незначительно растет при увеличении выборки\n",
    "* Из графика зависимости от количества признаков видим, что качество зависит синусоидально, но перепад не большой"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Logistic Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No handles with labels found to put in legend.\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a205adac8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.linear_model import RidgeClassifier, RidgeClassifierCV, LogisticRegression\n",
    "linear_clf = LogisticRegression()\n",
    "dummy_param = {'penalty':['l2']}\n",
    "score = []\n",
    "score_std = []\n",
    "for sub_features in range(10,50,5):\n",
    "    tmp_score = []\n",
    "    tmp_score_std = []\n",
    "    # Для каждого размера фитим 20 раз c 5ю соседями и усредняем\n",
    "    for i in range(20):\n",
    "        cols = np.random.choice(data.columns, sub_features, replace=False)\n",
    "        data_small = data[cols]\n",
    "        classes_small = classes\n",
    "        clf = GridSearchCV(linear_clf, dummy_param, cv = 3, n_jobs=1, \n",
    "                   verbose=0, scoring=make_scorer(f1_score,average='micro'))\n",
    "        clf.fit(data_small, classes_small)\n",
    "        tmp_score.append(clf.cv_results_['mean_test_score'].mean())\n",
    "        tmp_score_std.append(clf.cv_results_['std_test_score'].mean())\n",
    "    score.append(np.mean(tmp_score))\n",
    "    score_std.append(np.mean(tmp_score_std))\n",
    "#graph\n",
    "plt.figure(figsize=(12,8))\n",
    "x = range(10,50,5) if subfeatures else range(12,36,4)\n",
    "y = np.array(score)\n",
    "plt.plot(x, y)\n",
    "plt.fill_between(x, y-score_std, \\\n",
    "                 y+score_std, \\\n",
    "                 color = 'cyan')\n",
    "plt.title('{} F1 (micro averaged) score with a confidence interval'.format(\"LogReg\"), fontsize=15)\n",
    "if subfeatures:\n",
    "    plt.xlabel('Количество признаков', fontsize=15)\n",
    "else:\n",
    "    plt.xlabel('Объем выборки', fontsize=15)\n",
    "plt.ylabel('F1 score', fontsize=15)\n",
    "plt.legend(loc='best', prop={'size': 18})\n",
    "plt.grid()    \n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 123,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No handles with labels found to put in legend.\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a205fd668>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "linear_clf = LogisticRegression()\n",
    "dummy_param = {'penalty':['l2']}\n",
    " \n",
    "score = []\n",
    "score_std = []\n",
    "for s in range(12,36,4):\n",
    "    # Для каждого размера фитим 20 раз\n",
    "    tmp_score = []\n",
    "    tmp_score_std = []\n",
    "    for i in range(20):\n",
    "        data_small = data.sample(s)\n",
    "        classes_small = classes[data_small.index]\n",
    "        clf = GridSearchCV(linear_clf, dummy_param, cv = 3, n_jobs=1, \n",
    "                   verbose=0, scoring=make_scorer(f1_score,average='micro'))\n",
    "        clf.fit(data_small, classes_small)\n",
    "        tmp_score.append(clf.cv_results_['mean_test_score'].mean())\n",
    "        tmp_score_std.append(clf.cv_results_['std_test_score'].mean())\n",
    "    score.append(np.mean(tmp_score))\n",
    "    score_std.append(np.mean(tmp_score_std))\n",
    "\n",
    "plt.figure(figsize=(12,8))\n",
    "x = range(10,40,5) if subfeatures else range(12,36,4)\n",
    "y = np.array(score)\n",
    "plt.plot(x, y)\n",
    "plt.fill_between(x, y-score_std, \\\n",
    "                 y+score_std, \\\n",
    "                 color = 'cyan')\n",
    "plt.title('{} F1 (micro averaged) score with a confidence interval'.format(\"LogReg\"), fontsize=15)\n",
    "plt.xlabel('Объем выборки', fontsize=15)\n",
    "plt.ylabel('F1 score', fontsize=15)\n",
    "plt.legend(loc='best', prop={'size': 18})\n",
    "plt.grid()    \n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Вывод:\n",
    "* сходится за одну итерацию\n",
    "* качество незначительно растет с увеличением количества признаков\n",
    "* качество незначительно растет при увеличении размера выборки, как и от количества признаков"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# МАЛЫЙ ОБЪЕМ ВЫБОРКИ!!!!!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}