{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd #Dataframe, Series\n",
    "import numpy as np #Paquetes de Scientific computing, Arrays\n",
    "from matplotlib import pyplot as plt #Graficos\n",
    "\n",
    "from sklearn.model_selection import train_test_split #Dividir Dataset en train y test\n",
    "from sklearn.preprocessing import LabelEncoder #Pasar datos categoricos a numericos\n",
    "from sklearn import preprocessing #Normalizacion de datos\n",
    "\n",
    "#Imports dibujo arbol de decision\n",
    "import graphviz\n",
    "import StringIO as io\n",
    "import pydotplus\n",
    "import imageio\n",
    "\n",
    "import time #Medir tiempo de entrenamiento\n",
    "\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Dataset de conexiones KDD'99\n",
    "Dataset: kddcup.data.corrected (Completo)\n",
    "- Importacion del Dataset\n",
    "- Ingenieria de atributos\n",
    "- Visualización de datos y gráfico de 3 de los principales atributos\n",
    "- Entrenamiento de un clasificador (Arbol de decision)\n",
    "- Predecir objetivo usando el clasificador entrenado\n",
    "- Comparacion de resultados con progresion lineal"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "IMPORTACION DEL DATASET Y DESCRIPCION"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "col_names = [\"duration\",\"protocol_type\",\"service\",\"flag\",\"src_bytes\",\n",
    "    \"dst_bytes\",\"land\",\"wrong_fragment\",\"urgent\",\"hot\",\"num_failed_logins\",\n",
    "    \"logged_in\",\"num_compromised\",\"root_shell\",\"su_attempted\",\"num_root\",\n",
    "    \"num_file_creations\",\"num_shells\",\"num_access_files\",\"num_outbound_cmds\",\n",
    "    \"is_host_login\",\"is_guest_login\",\"count\",\"srv_count\",\"serror_rate\",\n",
    "    \"srv_serror_rate\",\"rerror_rate\",\"srv_rerror_rate\",\"same_srv_rate\",\n",
    "    \"diff_srv_rate\",\"srv_diff_host_rate\",\"dst_host_count\",\"dst_host_srv_count\",\n",
    "    \"dst_host_same_srv_rate\",\"dst_host_diff_srv_rate\",\"dst_host_same_src_port_rate\",\n",
    "    \"dst_host_srv_diff_host_rate\",\"dst_host_serror_rate\",\"dst_host_srv_serror_rate\",\n",
    "    \"dst_host_rerror_rate\",\"dst_host_srv_rerror_rate\",\"label\"]\n",
    "\n",
    "data = pd.read_csv('../dataset/kddcup.data.corrected', header=None, names = col_names)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>duration</th>\n",
       "      <th>src_bytes</th>\n",
       "      <th>dst_bytes</th>\n",
       "      <th>land</th>\n",
       "      <th>wrong_fragment</th>\n",
       "      <th>urgent</th>\n",
       "      <th>hot</th>\n",
       "      <th>num_failed_logins</th>\n",
       "      <th>logged_in</th>\n",
       "      <th>num_compromised</th>\n",
       "      <th>...</th>\n",
       "      <th>dst_host_count</th>\n",
       "      <th>dst_host_srv_count</th>\n",
       "      <th>dst_host_same_srv_rate</th>\n",
       "      <th>dst_host_diff_srv_rate</th>\n",
       "      <th>dst_host_same_src_port_rate</th>\n",
       "      <th>dst_host_srv_diff_host_rate</th>\n",
       "      <th>dst_host_serror_rate</th>\n",
       "      <th>dst_host_srv_serror_rate</th>\n",
       "      <th>dst_host_rerror_rate</th>\n",
       "      <th>dst_host_srv_rerror_rate</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>4.898431e+06</td>\n",
       "      <td>4.898431e+06</td>\n",
       "      <td>4.898431e+06</td>\n",
       "      <td>4.898431e+06</td>\n",
       "      <td>4.898431e+06</td>\n",
       "      <td>4.898431e+06</td>\n",
       "      <td>4.898431e+06</td>\n",
       "      <td>4.898431e+06</td>\n",
       "      <td>4.898431e+06</td>\n",
       "      <td>4.898431e+06</td>\n",
       "      <td>...</td>\n",
       "      <td>4.898431e+06</td>\n",
       "      <td>4.898431e+06</td>\n",
       "      <td>4.898431e+06</td>\n",
       "      <td>4.898431e+06</td>\n",
       "      <td>4.898431e+06</td>\n",
       "      <td>4.898431e+06</td>\n",
       "      <td>4.898431e+06</td>\n",
       "      <td>4.898431e+06</td>\n",
       "      <td>4.898431e+06</td>\n",
       "      <td>4.898431e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>4.834243e+01</td>\n",
       "      <td>1.834621e+03</td>\n",
       "      <td>1.093623e+03</td>\n",
       "      <td>5.716116e-06</td>\n",
       "      <td>6.487792e-04</td>\n",
       "      <td>7.961733e-06</td>\n",
       "      <td>1.243766e-02</td>\n",
       "      <td>3.205108e-05</td>\n",
       "      <td>1.435290e-01</td>\n",
       "      <td>8.088304e-03</td>\n",
       "      <td>...</td>\n",
       "      <td>2.329811e+02</td>\n",
       "      <td>1.892142e+02</td>\n",
       "      <td>7.537132e-01</td>\n",
       "      <td>3.071111e-02</td>\n",
       "      <td>6.050520e-01</td>\n",
       "      <td>6.464107e-03</td>\n",
       "      <td>1.780911e-01</td>\n",
       "      <td>1.778859e-01</td>\n",
       "      <td>5.792780e-02</td>\n",
       "      <td>5.765941e-02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>7.233298e+02</td>\n",
       "      <td>9.414311e+05</td>\n",
       "      <td>6.450123e+05</td>\n",
       "      <td>2.390833e-03</td>\n",
       "      <td>4.285434e-02</td>\n",
       "      <td>7.215084e-03</td>\n",
       "      <td>4.689782e-01</td>\n",
       "      <td>7.299408e-03</td>\n",
       "      <td>3.506116e-01</td>\n",
       "      <td>3.856481e+00</td>\n",
       "      <td>...</td>\n",
       "      <td>6.402094e+01</td>\n",
       "      <td>1.059128e+02</td>\n",
       "      <td>4.111860e-01</td>\n",
       "      <td>1.085432e-01</td>\n",
       "      <td>4.809877e-01</td>\n",
       "      <td>4.125978e-02</td>\n",
       "      <td>3.818382e-01</td>\n",
       "      <td>3.821774e-01</td>\n",
       "      <td>2.309428e-01</td>\n",
       "      <td>2.309777e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>4.500000e+01</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>...</td>\n",
       "      <td>2.550000e+02</td>\n",
       "      <td>4.900000e+01</td>\n",
       "      <td>4.100000e-01</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>5.200000e+02</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>...</td>\n",
       "      <td>2.550000e+02</td>\n",
       "      <td>2.550000e+02</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>1.032000e+03</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>...</td>\n",
       "      <td>2.550000e+02</td>\n",
       "      <td>2.550000e+02</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>4.000000e-02</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>5.832900e+04</td>\n",
       "      <td>1.379964e+09</td>\n",
       "      <td>1.309937e+09</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>3.000000e+00</td>\n",
       "      <td>1.400000e+01</td>\n",
       "      <td>7.700000e+01</td>\n",
       "      <td>5.000000e+00</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>7.479000e+03</td>\n",
       "      <td>...</td>\n",
       "      <td>2.550000e+02</td>\n",
       "      <td>2.550000e+02</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>1.000000e+00</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>8 rows × 38 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "           duration     src_bytes     dst_bytes          land  wrong_fragment  \\\n",
       "count  4.898431e+06  4.898431e+06  4.898431e+06  4.898431e+06    4.898431e+06   \n",
       "mean   4.834243e+01  1.834621e+03  1.093623e+03  5.716116e-06    6.487792e-04   \n",
       "std    7.233298e+02  9.414311e+05  6.450123e+05  2.390833e-03    4.285434e-02   \n",
       "min    0.000000e+00  0.000000e+00  0.000000e+00  0.000000e+00    0.000000e+00   \n",
       "25%    0.000000e+00  4.500000e+01  0.000000e+00  0.000000e+00    0.000000e+00   \n",
       "50%    0.000000e+00  5.200000e+02  0.000000e+00  0.000000e+00    0.000000e+00   \n",
       "75%    0.000000e+00  1.032000e+03  0.000000e+00  0.000000e+00    0.000000e+00   \n",
       "max    5.832900e+04  1.379964e+09  1.309937e+09  1.000000e+00    3.000000e+00   \n",
       "\n",
       "             urgent           hot  num_failed_logins     logged_in  \\\n",
       "count  4.898431e+06  4.898431e+06       4.898431e+06  4.898431e+06   \n",
       "mean   7.961733e-06  1.243766e-02       3.205108e-05  1.435290e-01   \n",
       "std    7.215084e-03  4.689782e-01       7.299408e-03  3.506116e-01   \n",
       "min    0.000000e+00  0.000000e+00       0.000000e+00  0.000000e+00   \n",
       "25%    0.000000e+00  0.000000e+00       0.000000e+00  0.000000e+00   \n",
       "50%    0.000000e+00  0.000000e+00       0.000000e+00  0.000000e+00   \n",
       "75%    0.000000e+00  0.000000e+00       0.000000e+00  0.000000e+00   \n",
       "max    1.400000e+01  7.700000e+01       5.000000e+00  1.000000e+00   \n",
       "\n",
       "       num_compromised            ...             dst_host_count  \\\n",
       "count     4.898431e+06            ...               4.898431e+06   \n",
       "mean      8.088304e-03            ...               2.329811e+02   \n",
       "std       3.856481e+00            ...               6.402094e+01   \n",
       "min       0.000000e+00            ...               0.000000e+00   \n",
       "25%       0.000000e+00            ...               2.550000e+02   \n",
       "50%       0.000000e+00            ...               2.550000e+02   \n",
       "75%       0.000000e+00            ...               2.550000e+02   \n",
       "max       7.479000e+03            ...               2.550000e+02   \n",
       "\n",
       "       dst_host_srv_count  dst_host_same_srv_rate  dst_host_diff_srv_rate  \\\n",
       "count        4.898431e+06            4.898431e+06            4.898431e+06   \n",
       "mean         1.892142e+02            7.537132e-01            3.071111e-02   \n",
       "std          1.059128e+02            4.111860e-01            1.085432e-01   \n",
       "min          0.000000e+00            0.000000e+00            0.000000e+00   \n",
       "25%          4.900000e+01            4.100000e-01            0.000000e+00   \n",
       "50%          2.550000e+02            1.000000e+00            0.000000e+00   \n",
       "75%          2.550000e+02            1.000000e+00            4.000000e-02   \n",
       "max          2.550000e+02            1.000000e+00            1.000000e+00   \n",
       "\n",
       "       dst_host_same_src_port_rate  dst_host_srv_diff_host_rate  \\\n",
       "count                 4.898431e+06                 4.898431e+06   \n",
       "mean                  6.050520e-01                 6.464107e-03   \n",
       "std                   4.809877e-01                 4.125978e-02   \n",
       "min                   0.000000e+00                 0.000000e+00   \n",
       "25%                   0.000000e+00                 0.000000e+00   \n",
       "50%                   1.000000e+00                 0.000000e+00   \n",
       "75%                   1.000000e+00                 0.000000e+00   \n",
       "max                   1.000000e+00                 1.000000e+00   \n",
       "\n",
       "       dst_host_serror_rate  dst_host_srv_serror_rate  dst_host_rerror_rate  \\\n",
       "count          4.898431e+06              4.898431e+06          4.898431e+06   \n",
       "mean           1.780911e-01              1.778859e-01          5.792780e-02   \n",
       "std            3.818382e-01              3.821774e-01          2.309428e-01   \n",
       "min            0.000000e+00              0.000000e+00          0.000000e+00   \n",
       "25%            0.000000e+00              0.000000e+00          0.000000e+00   \n",
       "50%            0.000000e+00              0.000000e+00          0.000000e+00   \n",
       "75%            0.000000e+00              0.000000e+00          0.000000e+00   \n",
       "max            1.000000e+00              1.000000e+00          1.000000e+00   \n",
       "\n",
       "       dst_host_srv_rerror_rate  \n",
       "count              4.898431e+06  \n",
       "mean               5.765941e-02  \n",
       "std                2.309777e-01  \n",
       "min                0.000000e+00  \n",
       "25%                0.000000e+00  \n",
       "50%                0.000000e+00  \n",
       "75%                0.000000e+00  \n",
       "max                1.000000e+00  \n",
       "\n",
       "[8 rows x 38 columns]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>duration</th>\n",
       "      <th>protocol_type</th>\n",
       "      <th>service</th>\n",
       "      <th>flag</th>\n",
       "      <th>src_bytes</th>\n",
       "      <th>dst_bytes</th>\n",
       "      <th>land</th>\n",
       "      <th>wrong_fragment</th>\n",
       "      <th>urgent</th>\n",
       "      <th>hot</th>\n",
       "      <th>...</th>\n",
       "      <th>dst_host_srv_count</th>\n",
       "      <th>dst_host_same_srv_rate</th>\n",
       "      <th>dst_host_diff_srv_rate</th>\n",
       "      <th>dst_host_same_src_port_rate</th>\n",
       "      <th>dst_host_srv_diff_host_rate</th>\n",
       "      <th>dst_host_serror_rate</th>\n",
       "      <th>dst_host_srv_serror_rate</th>\n",
       "      <th>dst_host_rerror_rate</th>\n",
       "      <th>dst_host_srv_rerror_rate</th>\n",
       "      <th>label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>tcp</td>\n",
       "      <td>http</td>\n",
       "      <td>SF</td>\n",
       "      <td>215</td>\n",
       "      <td>45076</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>normal.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>tcp</td>\n",
       "      <td>http</td>\n",
       "      <td>SF</td>\n",
       "      <td>162</td>\n",
       "      <td>4528</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>normal.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>tcp</td>\n",
       "      <td>http</td>\n",
       "      <td>SF</td>\n",
       "      <td>236</td>\n",
       "      <td>1228</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>2</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.50</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>normal.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>tcp</td>\n",
       "      <td>http</td>\n",
       "      <td>SF</td>\n",
       "      <td>233</td>\n",
       "      <td>2032</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>3</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.33</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>normal.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>tcp</td>\n",
       "      <td>http</td>\n",
       "      <td>SF</td>\n",
       "      <td>239</td>\n",
       "      <td>486</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.25</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>normal.</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 42 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   duration protocol_type service flag  src_bytes  dst_bytes  land  \\\n",
       "0         0           tcp    http   SF        215      45076     0   \n",
       "1         0           tcp    http   SF        162       4528     0   \n",
       "2         0           tcp    http   SF        236       1228     0   \n",
       "3         0           tcp    http   SF        233       2032     0   \n",
       "4         0           tcp    http   SF        239        486     0   \n",
       "\n",
       "   wrong_fragment  urgent  hot   ...     dst_host_srv_count  \\\n",
       "0               0       0    0   ...                      0   \n",
       "1               0       0    0   ...                      1   \n",
       "2               0       0    0   ...                      2   \n",
       "3               0       0    0   ...                      3   \n",
       "4               0       0    0   ...                      4   \n",
       "\n",
       "   dst_host_same_srv_rate  dst_host_diff_srv_rate  \\\n",
       "0                     0.0                     0.0   \n",
       "1                     1.0                     0.0   \n",
       "2                     1.0                     0.0   \n",
       "3                     1.0                     0.0   \n",
       "4                     1.0                     0.0   \n",
       "\n",
       "   dst_host_same_src_port_rate  dst_host_srv_diff_host_rate  \\\n",
       "0                         0.00                          0.0   \n",
       "1                         1.00                          0.0   \n",
       "2                         0.50                          0.0   \n",
       "3                         0.33                          0.0   \n",
       "4                         0.25                          0.0   \n",
       "\n",
       "   dst_host_serror_rate  dst_host_srv_serror_rate  dst_host_rerror_rate  \\\n",
       "0                   0.0                       0.0                   0.0   \n",
       "1                   0.0                       0.0                   0.0   \n",
       "2                   0.0                       0.0                   0.0   \n",
       "3                   0.0                       0.0                   0.0   \n",
       "4                   0.0                       0.0                   0.0   \n",
       "\n",
       "   dst_host_srv_rerror_rate    label  \n",
       "0                       0.0  normal.  \n",
       "1                       0.0  normal.  \n",
       "2                       0.0  normal.  \n",
       "3                       0.0  normal.  \n",
       "4                       0.0  normal.  \n",
       "\n",
       "[5 rows x 42 columns]"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 4898431 entries, 0 to 4898430\n",
      "Data columns (total 42 columns):\n",
      "duration                       int64\n",
      "protocol_type                  object\n",
      "service                        object\n",
      "flag                           object\n",
      "src_bytes                      int64\n",
      "dst_bytes                      int64\n",
      "land                           int64\n",
      "wrong_fragment                 int64\n",
      "urgent                         int64\n",
      "hot                            int64\n",
      "num_failed_logins              int64\n",
      "logged_in                      int64\n",
      "num_compromised                int64\n",
      "root_shell                     int64\n",
      "su_attempted                   int64\n",
      "num_root                       int64\n",
      "num_file_creations             int64\n",
      "num_shells                     int64\n",
      "num_access_files               int64\n",
      "num_outbound_cmds              int64\n",
      "is_host_login                  int64\n",
      "is_guest_login                 int64\n",
      "count                          int64\n",
      "srv_count                      int64\n",
      "serror_rate                    float64\n",
      "srv_serror_rate                float64\n",
      "rerror_rate                    float64\n",
      "srv_rerror_rate                float64\n",
      "same_srv_rate                  float64\n",
      "diff_srv_rate                  float64\n",
      "srv_diff_host_rate             float64\n",
      "dst_host_count                 int64\n",
      "dst_host_srv_count             int64\n",
      "dst_host_same_srv_rate         float64\n",
      "dst_host_diff_srv_rate         float64\n",
      "dst_host_same_src_port_rate    float64\n",
      "dst_host_srv_diff_host_rate    float64\n",
      "dst_host_serror_rate           float64\n",
      "dst_host_srv_serror_rate       float64\n",
      "dst_host_rerror_rate           float64\n",
      "dst_host_srv_rerror_rate       float64\n",
      "label                          object\n",
      "dtypes: float64(15), int64(23), object(4)\n",
      "memory usage: 1.5+ GB\n"
     ]
    }
   ],
   "source": [
    "data.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Histogramas (Normal / Ataque)\n",
    "- Ejemplo Histogramas (La totalidad se muestra en el Anexo 1 para cada tipo de ataque)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python2.7/dist-packages/matplotlib/cbook/deprecation.py:106: MatplotlibDeprecationWarning: Adding an axes using the same arguments as a previous axes currently reuses the earlier instance.  In a future version, a new instance will always be created and returned.  Meanwhile, this warning can be suppressed, and the future behavior ensured, by passing a unique label to each axes instance.\n",
      "  warnings.warn(message, mplDeprecation, stacklevel=1)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fa6b9111750>"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABKAAAAyXCAYAAADlyUnOAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvNQv5yAAAIABJREFUeJzs3X+s5fVd5/HXG8Z23DZSLJtZFnCna/mHmog6i2zcbO7aTWfaf8CkVuqmHZWIiZBoNBura6Tblo3+Uds0aUkwTJi6pdDUukWlGdnak6aJpdBaobS63FQJgxQUKO3QUMV89o/7HXM6c+7cOZf7Zi4zj0dyMud+vp/vj0Py+efJ95xvjTECAAAAAF3OOtUXAAAAAMDpTYACAAAAoJUABQAAAEArAQoAAACAVgIUAAAAAK0EKAAAAABaCVAAAAAAtBKgAAAaVNVPV9W9VXWkqh6tqk9U1X9qPueoqld3ngMAYDMEKACALVZVv5LkvUn+V5JdSb4vyQeSXHEqrwsA4FQRoAAAtlBVnZPkHUmuHWN8bIzxzBjjn8YYfzTG+O9V9dKqem9V/d30em9VvXTa92eq6jPHHO9f7mqqqluq6v1V9SdV9c2quruqvn/a9ulpl7+c7rr6qRfwYwMAnJAABQCwtf5jkp1J/nCd7f8jyeVJLk3yg0kuS/KbSxz/qiT/M8m5SVaT3JAkY4z/PG3/wTHGy8cYty9/6QAAPQQoAICt9cok/zDGeG6d7f8tyTvGGI+PMf4+azHpLUsc/w/HGJ+bjv+hrIUsAIBtTYACANhaTyQ5r6p2rLP93yZ5aO7vh6axk/W1ufffSvLy5S4PAOCFJ0ABAGytP0/y7SRXrrP975L8u7m/v28aS5Jnkvyroxuq6t90XCAAwAttvf8zBwDAJowxnq6q30ry/qp6LsmfJvmnJP81yX9J8uEkv1lV9yQZSX4ryf+edv/LJK+pqkuT/FWSty95+seS/Pus/TYUAMC24Q4oAIAtNsZ4d5JfydqPi/99koeTXJfk/yR5V5J7k9yX5P4kX5jGMsb4f1l7gt7/TfJgks8ce+wNvD3Jwar6elW96Xl/EACALVJjjFN9DQAAAACcxtwBBQAAAEArAQoAAACAVgIUAAAAAK0EKAAAAABaCVAAAAAAtNpxqi/ghXLeeeeN3bt3n+rL2BLPPPNMXvayl53qy4BtxbqAxawNWMzagONZF7CYtXFin//85/9hjPGvN5p3xgSo3bt359577z3Vl7ElZrNZVlZWTvVlwLZiXcBi1gYsZm3A8awLWMzaOLGqeuhk5vkKHgAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWu041RcAAAAA8IK59aeWm79zb3LrjVt7DT99+9Ye70XAHVAAAAAAtBKgAAAAAGglQAEAAADQSoACAAAAoJUABQAAAEArAQoAAACAVgIUAAAAAK0EKAAAAABaCVAAAAAAtBKgAAAAAGglQAEAAADQSoACAAAAoJUABQAAAEArAQoAAACAVgIUAAAAAK0EKAAAAABaCVAAAAAAtBKgAAAAAGglQAEAAADQSoACAAAAoJUABQAAAEArAQoAAACAVgIUAAAAAK0EKAAAAABaCVAAAAAAtBKgAAAAAGglQAEAAADQSoACAAAAoJUABQAAAEArAQoAAACAVgIUAAAAAK0EKAAAAABaCVAAAAAAtBKgAAAAAGglQAEAAADQSoACAAAAoJUABQAAAEArAQoAAACAVhsGqKq6qKo+VVVfrqoHquqXpvG3V9UjVfXF6fWGuX1+vapWq+qvq2rv3Pi+aWy1qt42N/6qqrp7Gr+9ql4yjb90+nt12r57o3MAAAAAsL2czB1QzyX51THGJUkuT3JtVV0ybXvPGOPS6XVnkkzbrkrymiT7knygqs6uqrOTvD/J65NckuTNc8f5nelYr07yVJKrp/Grkzw1jb9nmrfuOTb9XwEAAACANhsGqDHGo2OML0zvv5nkK0kuOMEuVyS5bYzx7THG3yRZTXLZ9FodY3x1jPGPSW5LckVVVZIfT/LRaf+DSa6cO9bB6f1Hk7x2mr/eOQAAAADYZpb6DajpK3A/lOTuaei6qrqvqg5U1bnT2AVJHp7b7fA0tt74K5N8fYzx3DHj33GsafvT0/z1jgUAAADANrPjZCdW1cuT/EGSXx5jfKOqbkzyziRj+vfdSX6u5So3qaquSXJNkuzatSuz2ezUXtAWOXLkyGnzWWCrWBewmLUBi1kbcDzrgjPGzuV+RvrIWedktuQ+GzoD19pJBaiq+q6sxacPjTE+liRjjMfmtv9ekj+e/nwkyUVzu184jWWd8SeSvKKqdkx3Oc3PP3qsw1W1I8k50/wTneNfjDFuSnJTkuzZs2esrKyczMfd9mazWU6XzwJbxbqAxawNWMzagONZF5wxbr1xqemznXuz8uyhrb2Gldu39ngvAifzFLxKcnOSr4wxfndu/Py5aT+R5EvT+zuSXDU9we5VSS5O8rkk9yS5eHri3Uuy9iPid4wxRpJPJXnjtP/+JB+fO9b+6f0bk/zZNH+9cwAAAACwzZzMHVA/luQtSe6vqi9OY7+RtafYXZq1r+D9bZJfSJIxxgNV9ZEkX87aE/SuHWP8c5JU1XVJDiU5O8mBMcYD0/F+LcltVfWuJH+RteCV6d/fr6rVJE9mLVqd8BwAAAAAbC8bBqgxxmeS1IJNd55gnxuS3LBg/M5F+40xvpoFT7EbYzyb5CeXOQcAAAAA28tST8EDAAAAgGUJUAAAAAC0EqAAAAAAaCVAAQAAANBKgAIAAACglQAFAAAAQCsBCgAAAIBWAhQAAAAArQQoAAAAAFoJUAAAAAC0EqAAAAAAaCVAAQAAANBKgAIAAACglQAFAAAAQCsBCgAAAIBWAhQAAAAArQQoAAAAAFoJUAAAAAC0EqAAAAAAaCVAAQAAANBKgAIAAACglQAFAAAAQCsBCgAAAIBWAhQAAAAArQQoAAAAAFoJUAAAAAC0EqAAAAAAaCVAAQAAANBKgAIAAACglQAFAAAAQCsBCgAAAIBWAhQAAAAArQQoAAAAAFoJUAAAAAC0EqAAAAAAaCVAAQAAANBKgAIAAACglQAFAAAAQCsBCgAAAIBWAhQAAAAArQQoAAAAAFoJUAAAAAC0EqAAAAAAaCVAAQAAANBKgAIAAACglQAFAAAAQCsBCgAAAIBWAhQAAAAArQQoAAAAAFoJUAAAAAC0EqAAAAAAaCVAAQAAANBKgAIAAACglQAFAAAAQCsBCgAAAIBWAhQAAAAArQQoAAAAAFoJUAAAAAC0EqAAAAAAaCVAAQAAANBKgAIAAACglQAFAAAAQCsBCgAAAIBWAhQAAAAArQQoAAAAAFoJUAAAAAC0EqAAAAAAaCVAAQAAANBKgAIAAACglQAFAAAAQCsBCgAAAIBWAhQAAAAArQQoAAAAAFoJUAAAAAC0EqAAAAAAaCVAAQAAANBKgAIAAACglQAFAAAAQCsBCgAAAIBWAhQAAAAArQQoAAAAAFoJUAAAAAC0EqAAAAAAaCVAAQAAANBKgAIAAACglQAFAAAAQCsBCgAAAIBWAhQAAAAArQQoAAAAAFoJUAAAAAC0EqAAAAAAaCVAAQAAANBKgAIAAACglQAFAAAAQCsBCgAAAIBWAhQAAAAArQQoAAAAAFoJUAAAAAC0EqAAAAAAaCVAAQAAANBKgAIAAACglQAFAAAAQCsBCgAAAIBWAhQAAAAArQQoAAAAAFoJUAAAAAC0EqAAAAAAaCVAAQAAANBKgAIAAACglQAFAAAAQCsBCgAAAIBWAhQAAAAArQQoAAAAAFoJUAAAAAC0EqAAAAAAaCVAAQAAANBKgAIAAACglQAFAAAAQCsBCgAAAIBWAhQAAAAArQQoAAAAAFoJUAAAAAC0EqAAAAAAaCVAAQAAANBKgAIAAACglQAFAAAAQCsBCgAAAIBWAhQAAAAArQQoAAAAAFoJUAAAAAC0EqAAAAAAaCVAAQAAANBKgAIAAACglQAFAAAAQCsBCgAAAIBWAhQAAAAArQQoAAAAAFoJUAAAAAC0EqAAAAAAaCVAAQAAANBKgAIAAACglQAFAAAAQCsBCgAAAIBWAhQAAAAArQQoAAAAAFoJUAAAAAC0EqAAAAAAaCVAAQAAANBKgAIAAACglQAFAAAAQCsBCgAAAIBWAhQAAAAArQQoAAAAAFoJUAAAAAC0EqAAAAAAaCVAAQAAANBKgAIAAACglQAFAAAAQKsNA1RVXVRVn6qqL1fVA1X1S9P491bVXVX14PTvudN4VdX7qmq1qu6rqh+eO9b+af6DVbV/bvxHqur+aZ/3VVVt9hwAAAAAbC8ncwfUc0l+dYxxSZLLk1xbVZckeVuST44xLk7yyenvJHl9koun1zVJbkzWYlKS65P8aJLLklx/NChNc35+br990/hS5wAAAABg+9kwQI0xHh1jfGF6/80kX0lyQZIrkhycph1McuX0/ookHxxrPpvkFVV1fpK9Se4aYzw5xngqyV1J9k3bvmeM8dkxxkjywWOOtcw5AAAAANhmlvoNqKraneSHktydZNcY49Fp09eS7JreX5Dk4bndDk9jJxo/vGA8mzgHAAAAANvMjpOdWFUvT/IHSX55jPGN6WeakiRjjFFVo+H6ntc5quqarH1FL7t27cpsNuu4tBfckSNHTpvPAlvFuoDFrA1YzNqA41kXnDF27l1q+pGzzslsyX02dAautZMKUFX1XVmLTx8aY3xsGn6sqs4fYzw6ff3t8Wn8kSQXze1+4TT2SJKVY8Zn0/iFC+Zv5hzfYYxxU5KbkmTPnj1jZWXl2CkvSrPZLKfLZ4GtYl3AYtYGLGZtwPGsC84Yty73M9KznXuz8uyhrb2Gldu39ngvAifzFLxKcnOSr4wxfndu0x1Jjj7Jbn+Sj8+Nv3V6Ut3lSZ6evkZ3KMnrqurc6cfHX5fk0LTtG1V1+XSutx5zrGXOAQAAAMA2czJ3QP1Ykrckub+qvjiN/UaS307ykaq6OslDSd40bbszyRuSrCb5VpKfTZIxxpNV9c4k90zz3jHGeHJ6/4tJbkny3Uk+Mb2y7DkAAAAA2H42DFBjjM8kqXU2v3bB/JHk2nWOdSDJgQXj9yb5gQXjTyx7DgAAAAC2l6WeggcAAAAAyxKgAAAAAGglQAEAAADQSoACAAAAoJUABQAAAEArAQoAAACAVgIUAAAAAK0EKAAAAABaCVAAAAAAtBKgAAAAAGglQAEAAADQSoACAAAAoJUABQAAAEArAQoAAACAVgIUAAAAAK0EKAAAAABaCVAAAAAAtBKgAAAAAGglQAEAAADQSoACAAAAoJUABQAAAEArAQoAAACAVgIUAAAAAK0EKAAAAABaCVAAAAAAtBKgAAAAAGglQAEAAADQSoACAAAAoJUABQAAAEArAQoAAACAVgIUAAAAAK0EKAAAAABaCVAAAAAAtBKgAAAAAGglQAEAAADQSoACAAAAoJUABQAAAEArAQoAAACAVgIUAAAAAK0EKAAAAABaCVAAAAAAtBKgAAAAAGglQAEAAADQSoACAAAAoJUABQAAAEArAQoAAACAVgIUAAAAAK0EKAAAAABaCVAAAAAAtBKgAAAAAGglQAEAAADQSoACAAAAoJUABQAAAEArAQoAAACAVgIUAAAAAK0EKAAAAABaCVAAAAAAtBKgAAAAAGglQAEAAADQSoACAAAAoJUABQAAAEArAQoAAACAVgIUAAAAAK0EKAAAAABaCVAAAAAAtBKgAAAAAGglQAEAAADQSoACAAAAoJUABQAAAEArAQoAAACAVgIUAAAAAK0EKAAAAABaCVAAAAAAtBKgAAAAAGglQAEAAADQSoACAAAAoJUABQAAAEArAQoAAACAVgIUAAAAAK0EKAAAAABaCVAAAAAAtBKgAAAAAGglQAEAAADQSoACAAAAoJUABQAAAEArAQoAAACAVgIUAAAAAK0EKAAAAABaCVAAAAAAtBKgAAAAAGglQAEAAADQSoACAAAAoJUABQAAAEArAQoAAACAVgIUAAAAAK0EKAAAAABaCVAAAAAAtBKgAAAAAGglQAEAAADQSoACAAAAoJUABQAAAEArAQoAAACAVgIUAAAAAK0EKAAAAABaCVAAAAAAtBKgAAAAAGglQAEAAADQSoACAAAAoJUABQAAAEArAQoAAACAVgIUAAAAAK0EKAAAAABaCVAAAAAAtBKgAAAAAGglQAEAAADQSoACAAAAoJUABQAAAEArAQoAAACAVgIUAAAAAK0EKAAAAABaCVAAAAAAtBKgAAAAAGglQAEAAADQSoACAAAAoJUABQAAAEArAQoAAACAVgIUAAAAAK0EKAAAAABaCVAAAAAAtBKgAAAAAGglQAEAAADQSoACAAAAoJUABQAAAEArAQoAAACAVgIUAAAAAK0EKAAAAABaCVAAAAAAtBKgAAAAAGglQAEAAADQSoACAAAAoJUABQAAAEArAQoAAACAVgIUAAAAAK0EKAAAAABaCVAAAAAAtBKgAAAAAGglQAEAAADQSoACAAAAoJUABQAAAEArAQoAAACAVgIUAAAAAK0EKAAAAABaCVAAAAAAtBKgAAAAAGglQAEAAADQSoACAAAAoJUABQAAAEArAQoAAACAVhsGqKo6UFWPV9WX5sbeXlWPVNUXp9cb5rb9elWtVtVfV9XeufF909hqVb1tbvxVVXX3NH57Vb1kGn/p9PfqtH33RucAAAAAYPs5mTugbkmyb8H4e8YYl06vO5Okqi5JclWS10z7fKCqzq6qs5O8P8nrk1yS5M3T3CT5nelYr07yVJKrp/Grkzw1jb9nmrfuOZb72AAAAAC8UDYMUGOMTyd58iSPd0WS28YY3x5j/E2S1SSXTa/VMcZXxxj/mOS2JFdUVSX58SQfnfY/mOTKuWMdnN5/NMlrp/nrnQMAAACAbej5/AbUdVV13/QVvXOnsQuSPDw35/A0tt74K5N8fYzx3DHj33GsafvT0/z1jgUAAADANrRjk/vdmOSdScb077uT/NxWXdRWqaprklyTJLt27cpsNju1F7RFjhw5ctp8Ftgq1gUsZm3AYtYGHM+64Iyxc7mfkj5y1jmZLbnPhs7AtbapADXGeOzo+6r6vSR/PP35SJKL5qZeOI1lnfEnkryiqnZMdznNzz96rMNVtSPJOdP8E53j2Ou8KclNSbJnz56xsrKy1OfcrmazWU6XzwJbxbqAxawNWMzagONZF5wxbr1xqemznXuz8uyhrb2Gldu39ngvApv6Cl5VnT/3508kOfqEvDuSXDU9we5VSS5O8rkk9yS5eHri3Uuy9iPid4wxRpJPJXnjtP/+JB+fO9b+6f0bk/zZNH+9cwAAAACwDW14B1RVfTjJSpLzqupwkuuTrFTVpVn7Ct7fJvmFJBljPFBVH0ny5STPJbl2jPHP03GuS3IoydlJDowxHphO8WtJbquqdyX5iyQ3T+M3J/n9qlrN2o+gX7XROQAAAADYfjYMUGOMNy8YvnnB2NH5NyS5YcH4nUnuXDD+1Sx4it0Y49kkP7nMOQAAAADYfp7PU/AAAAAAYEMCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0GrDAFVVB6rq8ar60tzY91bVXVX14PTvudN4VdX7qmq1qu6rqh+e22f/NP/Bqto/N/4jVXX/tM/7qqo2ew4AAAAAtp+TuQPqliT7jhl7W5JPjjEuTvLJ6e8keX2Si6fXNUluTNZiUpLrk/xoksuSXH80KE1zfn5uv32bOQcAAAAA29OGAWqM8ekkTx4zfEWSg9P7g0munBv/4Fjz2SSvqKrzk+xNctcY48kxxlNJ7kqyb9r2PWOMz44xRpIPHnOsZc4BAAAAwDa02d+A2jXGeHR6/7Uku6b3FyR5eG7e4WnsROOHF4xv5hwAAAAAbEM7nu8BxhijqsZWXMxWn6Oqrsna1/Sya9euzGazrb60U+LIkSOnzWeBrWJdwGLWBixmbcDxrAvOGDv3LjX9yFnnZLbkPhs6A9faZgPUY1V1/hjj0enrb49P448kuWhu3oXT2CNJVo4Zn03jFy6Yv5lzHGeMcVOSm5Jkz549Y2VlZdG0F53ZbJbT5bPAVrEuYDFrAxazNuB41gVnjFuX+ynp2c69WXn20NZew8rtW3u8F4HNfgXvjiRHn2S3P8nH58bfOj2p7vIkT09fozuU5HVVde704+OvS3Jo2vaNqrp8evrdW4851jLnAAAAAGAb2vAOqKr6cNbuXjqvqg5n7Wl2v53kI1V1dZKHkrxpmn5nkjckWU3yrSQ/myRjjCer6p1J7pnmvWOMcfSHzX8xa0/a++4kn5heWfYcAAAAAGxPGwaoMcab19n02gVzR5Jr1znOgSQHFozfm+QHFow/sew5AAAAANh+NvsVPAAAAAA4KQIUAAAAAK0EKAAAAABaCVAAAAAAtBKgAAAAAGglQAEAAADQSoACAAAAoJUABQAAAEArAQoAAACAVgIUAAAAAK0EKAAAAABaCVAAAAAAtBKgAAAAAGglQAEAAADQSoACAAAAoJUABQAAAEArAQoAAACAVgIUAAAAAK0EKAAAAABaCVAAAAAAtBKgAAAAAGglQAEAAADQSoACAAAAoJUABQAAAEArAQoAAACAVgIUAAAAAK0EKAAAAABaCVAAAAAAtBKgAAAAAGglQAEAAADQSoACAAAAoJUABQAAAEArAQoAAACAVgIUAAAAAK0EKAAAAABaCVAAAAAAtBKgAAAAAGglQAEAAADQSoACAAAAoJUABQAAAEArAQoAAACAVjtO9QWwvIee+FauvuWeU30Zufln/sOpvgQAAADgRcAdUAAAAAC0EqAAAAAAaCVAAQAAANBKgAIAAACglQAFAAAAQCsBCgAAAIBWAhQAAAAArQQoAAAAAFoJUAAAAAC0EqAAAAAAaCVAAQAAANBKgAIAAACglQAFAAAAQCsBCgAAAIBWAhQAAAAArQQoAAAAAFoJUAAAAAC0EqAAAAAAaCVAAQAAANBKgAIAAACglQAFAAAAQCsBCgAAAIBWAhQAAAAArQQoAAAAAFoJUAAAAAC0EqAAAAAAaCVAAQAAANBKgAIAAACglQAFAAAAQCsBCgAAAIBWAhQAAAAArQQoAAAAAFoJUAAAAAC0EqAAAAAAaCVAAQAAANBKgAIAAACglQAFAAAAQCsBCgAAAIBWAhQAAAAArQQoAAAAAFoJUAAAAAC0EqAAAAAAaCVAAQAAANBKgAIAAACglQAFAAAAQCsBCgAAAIDTzLRuAAAgAElEQVRWAhQAAAAArQQoAAAAAFoJUAAAAPD/2bv3eMvrut7j74+AlwRBRScDFD2ShT0SC41OHpu0BMuiHscjZCkaRRf1WFleykeRSke7iGXGiZJEDYEsk4xSjjrHLmqAkYiX40SoTAjKHW+Bfs8f67d1s9kzs2fYn1l7Zj+fj8d+7LV+67d+67vXWl/Y89q/328BrQQoAAAAAFoJUAAAAAC0EqAAAAAAaCVAAQAAANBKgAIAAACglQAFAAAAQCsBCgAAAIBWAhQAAAAArQQoAAAAAFoJUAAAAAC0EqAAAAAAaCVAAQAAANBKgAIAAACglQAFAAAAQCsBCgAAAIBWAhQAAAAArQQoAAAAAFoJUAAAAAC0EqAAAAAAaCVAAQAAANBKgAIAAACglQAFAAAAQCsBCgAAAIBWAhQAAAAArQQoAAAAAFoJUAAAAAC0EqAAAAAAaCVAAQAAANBKgAIAAACglQAFAAAAQCsBCgAAAIBWAhQAAAAArQQoAAAAAFoJUAAAAAC0EqAAAAAAaCVAAQAAANBKgAIAAACglQAFAAAAQCsBCgAAAIBWAhQAAAAArQQoAAAAAFoJUAAAAAC0EqAAAAAAaCVAAQAAANBKgAIAAACglQAFAAAAQCsBCgAAAIBWAhQAAAAArQQoAAAAAFoJUAAAAAC0EqAAAAAAaCVAAQAAANBKgAIAAACglQAFAAAAQCsBCgAAAIBWAhQAAAAArQQoAAAAAFoJUAAAAAC0EqAAAAAAaCVAAQAAANBKgAIAAACglQAFAAAAQCsBCgAAAIBWAhQAAAAArQQoAAAAAFoJUAAAAAC0EqAAAAAAaHWnAlRVXVFVl1bVJVV10bTsPlV1QVV9fPp+72l5VdXvV9XmqvpgVX3bou2cMK3/8ao6YdHyb5+2v3m6b23rMQAAAABYe1ZjD6jvGWMcMcY4crr+wiTvHGMcluSd0/UkeWKSw6avk5KclsxiUpJfT/IdSR6d5NcXBaXTkvzUovsds53HAAAAAGCN6TgE79gkZ06Xz0zyw4uWv37MvC/JAVX1gCRHJ7lgjHHdGOP6JBckOWa67V5jjPeNMUaS1y/Z1nKPAQAAAMAac2cD1Ejyjqq6uKpOmpZtGGNcNV3+dJIN0+WDknxq0X2vnJZta/mVyyzf1mMAAAAAsMbsfSfv/5gxxpaqun+SC6rqo4tvHGOMqhp38jG2aVuPMUWxk5Jkw4YN2bRpU+dQdpl997otj93vM/Mexh7zfLJnuOWWW7wnYRnmBizP3IA7Mi9YN+5+9A6tfstd9s+mHbzPdq3DuXanAtQYY8v0/Zqqektm53C6uqoeMMa4ajqM7ppp9S1JDll094OnZVuSbFyyfNO0/OBl1s82HmPp+E5PcnqSHHnkkWPjxo3LrbbbOfMvzs97br7fvIeR1/73R817CPBVmzZtyp4yx2E1mRuwPHMD7si8YN0467QdWn3T3Y/Oxi++fXXHsPGc1d3ebmCnD8GrqntW1X4Ll5M8IcmHkpyXZOGT7E5I8tbp8nlJnj59Gt5RSW6cDqN7e5InVNW9p5OPPyHJ26fbbqqqo6ZPv3v6km0t9xgAAAAArDF3Zg+oDUneMmtD2TvJWWOMv6uqC5OcW1UnJvlEkqdM65+f5PuTbE7y+STPTJIxxnVV9dIkF07rvWSMcd10+eeSvC7JPZL87fSVJC/fymMAAAAAsMbsdIAaY1ye5BHLLL82yeOXWT6SPGsr2zojyRnLLL8oybes9DEAAAAAWHvu7KfgAQAAAMA2CVAAAAAAtBKgAAAAAGglQAEAAADQSoACAAAAoJUABQAAAEArAQoAAACAVgIUAAAAAK0EKAAAAABaCVAAAAAAtBKgAAAAAGglQAEAAADQSoACAAAAoJUABQAAAEArAQoAAACAVgIUAAAAAK0EKAAAAABaCVAAAAAAtBKgAAAAAGglQAEAAADQSoACAAAAoJUABQAAAEArAQoAAACAVgIUAAAAAK0EKAAAAABaCVAAAAAAtBKgAAAAAGglQAEAAADQSoACAAAAoJUABQAAAEArAQoAAACAVgIUAAAAAK0EKAAAAABaCVAAAAAAtBKgAAAAAGglQAEAAADQSoACAAAAoJUABQAAAEArAQoAAACAVgIUAAAAAK0EKAAAAABaCVAAAAAAtBKgAAAAAGglQAEAAADQSoACAAAAoJUABQAAAEArAQoAAACAVgIUAAAAAK0EKAAAAABaCVAAAAAAtBKgAAAAAGglQAEAAADQSoACAAAAoJUABQAAAEArAQoAAACAVgIUAAAAAK0EKAAAAABaCVAAAAAAtBKgAAAAAGglQAEAAADQSoACAAAAoJUABQAAAEArAQoAAACAVgIUAAAAAK0EKAAAAABaCVAAAAAAtBKgAAAAAGglQAEAAADQSoACAAAAoJUABQAAAEArAQoAAACAVgIUAAAAAK0EKAAAAABaCVAAAAAAtBKgAAAAAGglQAEAAADQSoACAAAAoJUABQAAAEArAQoAAACAVgIUAAAAAK0EKAAAAABaCVAAAAAAtBKgAAAAAGglQAEAAADQSoACAAAAoJUABQAAAEArAQoAAACAVgIUAAAAAK0EKAAAAABaCVAAAAAAtBKgAAAAAGglQAEAAADQSoACAAAAoJUABQAAAEArAQoAAACAVgIUAAAAAK0EKAAAAABaCVAAAAAAtBKgAAAAAGglQAEAAADQSoACAAAAoJUABQAAAEArAQoAAACAVgIUAAAAAK0EKAAAAABaCVAAAAAAtBKgAAAAAGglQAEAAADQSoACAAAAoJUABQAAAEArAQoAAACAVgIUAAAAAK32nvcAANqcddy8R7B2PPWceY8AAABYx+wBBQAAAEArAQoAAACAVgIUAAAAAK0EKAAAAABaCVAAAAAAtBKgAAAAAGglQAEAAADQSoACAAAAoJUABQAAAEArAQoAAACAVgIUAAAAAK0EKAAAAABa7T3vAQCwC5x13LxHMPPUc+Y9AgAAYA7sAQUAAABAK3tAAcB6Zw85AACaCVAAwNqwVkLYWiDGAQB7GAEKAGCtWSsxTggDAFaJc0ABAAAA0MoeUAAALM+eWADAKhGgAAAAoIOQD18lQLHTTnzdhfMeQpLktc941LyHAAAAAGyDc0ABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKDV3vMeAAAAALAOnHXcvEfAHNkDCgAAAIBW9oACAGBt6/iL+d2PTs46bcfu89RzVn8cALBOCFAAMC+7Yjf0nflHNgAArDKH4AEAAADQyh5Qu6H737olz7n69+Y9jLx6w8vmPQQAAABgNyBAAatvHp9u4TAjAACANUuAAgAAgD3ZPP5ADEs4BxQAAAAArQQoAAAAAFoJUAAAAAC0EqAAAAAAaOUk5AAAAOxZnHQb1hwBit3eia+7cN5DyGuf8ah5DwHYEX4pBQCAXWq3DlBVdUyS30uyV5I/GWO8fM5DAmBbhB8AAFiXdtsAVVV7JXlNku9LcmWSC6vqvDHGh+c7svXjOVe/eN5DSJK8esPL5j2EtcM/7gEAAFiDdtsAleTRSTaPMS5Pkqo6O8mxSQSodWYthLBLXjHvEcwcccgB8x4CAAAA3EGNMeY9hp1SVU9OcswY4yen609L8h1jjGcvWuekJCdNVx+W5GO7fKA9Dkzy2XkPAtYY8wKWZ27A8swNuCPzApZnbmzbg8YY99veSrvzHlDbNcY4Pcnp8x7Haquqi8YYR857HLCWmBewPHMDlmduwB2ZF7A8c2N13GXeA7gTtiQ5ZNH1g6dlAAAAAKwhu3OAujDJYVX14Kq6a5Ljk5w35zEBAAAAsMRuewjeGOO2qnp2krcn2SvJGWOMy+Y8rF1ljzusEFaBeQHLMzdgeeYG3JF5AcszN1bBbnsScgAAAAB2D7vzIXgAAAAA7AYEKAAAAABaCVC7kao6pqo+VlWbq+qF8x4P7EpVdUZVXVNVH1q07D5VdUFVfXz6fu9peVXV709z5YNV9W3zGzn0qapDqurdVfXhqrqsqp47LTc3WNeq6u5V9c9V9a/T3PiNafmDq+r90xw4Z/ogm1TV3abrm6fbD53n+KFTVe1VVf9SVW+brpsXrHtVdUVVXVpVl1TVRdMyv0+tMgFqN1FVeyV5TZInJjk8yY9W1eHzHRXsUq9LcsySZS9M8s4xxmFJ3jldT2bz5LDp66Qkp+2iMcKudluS540xDk9yVJJnTf9vMDdY776U5HFjjEckOSLJMVV1VJJXJDl1jPHQJNcnOXFa/8Qk10/LT53Wgz3Vc5N8ZNF18wJmvmeMccQY48jput+nVpkAtft4dJLNY4zLxxj/meTsJMfOeUywy4wx3pPkuiWLj01y5nT5zCQ/vGj568fM+5IcUFUP2DUjhV1njHHVGOMD0+WbM/sHxUExN1jnpvf4LdPVfaavkeRxSd48LV86NxbmzJuTPL6qahcNF3aZqjo4yQ8k+ZPpesW8gK3x+9QqE6B2Hwcl+dSi61dOy2A92zDGuGq6/OkkG6bL5gvrznRoxCOTvD/mBiwcZnRJkmuSXJDk35LcMMa4bVpl8fv/q3Njuv3GJPfdtSOGXeJVSZ6f5CvT9fvGvIBk9keKd1TVxVV10rTM71OrbO95DwBgNYwxRlWNeY8D5qGq9k3yF0l+foxx0+I/UJsbrFdjjC8nOaKqDkjyliTfNOchwVxV1ZOSXDPGuLiqNs57PLDGPGaMsaWq7p/kgqr66OIb/T61OuwBtfvYkuSQRdcPnpbBenb1wu6u0/drpuXmC+tGVe2TWXz6szHGX06LzQ2YjDFuSPLuJN+Z2WESC3+AXfz+/+rcmG7fP8m1u3io0O27kvxQVV2R2ek8Hpfk92JeQMYYW6bv12T2R4tHx+9Tq06A2n1cmOSw6VMq7prk+CTnzXlMMG/nJTlhunxCkrcuWv706RMqjkpy46LdZ2GPMZ2L47VJPjLGeOWim8wN1rWqut+051Oq6h5Jvi+zc6S9O8mTp9WWzo2FOfPkJO8aY/hLN3uUMcaLxhgHjzEOzezfEu8aY/xYzAvWuaq6Z1Xtt3A5yROSfCh+n1p15b8hu4+q+v7MjtveK8kZY4xT5jwk2GWq6k1JNiY5MMnVSX49yV8lOTfJA5N8IslTxhjXTf8o/4PMPjXv80meOca4aB7jhk5V9Zgkf5/k0nztfB6/ktl5oMwN1q2q+tbMThi7V2Z/cD13jPGSqnpIZnt+3CfJvyT58THGl6rq7knekNl51K5LcvwY4/L5jB76TYfg/dIY40nmBevdNAfeMl3dO8lZY4xTquq+8fvUqhKgAAAAAGjlEDwAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAYMWq6oFVdUtV7dW0/cuqamPDdjdV1U+u9nYBAFgZAQoAdnNVdUVVfaGqbq6qG6rqn6rqZ6rqTv9/ftr29y5cH2N8coyx7xjjy6uw7ddV1csWLxtjPHyMsenObntXqKqTq+qN8x4HK1dVh1bVqKq9V3m7o6oeuprbBIA9jQAFAHuGHxxj7JfkQUlenuQFSV473yGxHtXMmvsdc2ej02rHKgBYr9bcLwcAwM4bY9w4xjgvyXFJTqiqb0mSqrpbVf1OVX2yqq6uqv9dVfeYbjuwqt427T11XVX9fVXdparekOSBSf56Ouzu+Uv3IJkObXtpVf3jtAfWO6rqwIXxVNWfV9Wnq+rGqnpPVT18Wn5Skh9L8vxp2389Lf/qHlfTmF9VVf8xfb2qqu423baxqq6squdV1TVVdVVVPXM7T8+DtjHOH5oO/7th+pm+edFtL6iqLdP9PlZVj6+qY5L8SpLjpvH/63IPuNx9p+WPrqr3To93VVX9QVXdddH9RlX9XFV9fLrvS6vqv0x7t91UVecuWf9JVXXJoj3gvnUr46mqOnV6zm6qqksXvUf2r6rXV9VnquoTVfXihZC0dG+vrbwPTqmqf0zy+SQPqar7VNWfTq/d9VX1Vzs63kXPxf+sqsur6rNV9duLxnWXaZyfmH6m11fV/kvGeGJVfTLJu5K8Z9rsDdPr9p3LPN7JVfXmqnpjVd2U5Bnber2qamGb/zpt87gd/RkBYD0QoABgDzTG+OckVyb5b9Oilyf5xiRHJHlokoOS/Np02/Omde+XZENmYWWMMZ6W5JOZ7V217xjjt7bycE9N8swk909y1yS/tOi2v01y2HTbB5L82TS+06fLvzVt+weX2e6vJjlqGvMjkjw6yYsX3f71SfaffpYTk7ymqu69jadl2XFW1TcmeVOSn5+eg/Mzi253raqHJXl2kkdNe5gdneSKMcbfJfnNJOdM43/E0gfb2n2nm7+c5BeSHJjkO5M8PsnPLdnE0Um+fXoOnp/k9CQ/nuSQJN+S5Eenx3lkkjOS/HSS+yb5oyTnLcS6JZ6Q5LGZvRf2T/KUJNdOt716WvaQJN+d5OnT87VST0tyUpL9knwiyRuSfF2Sh2f2nJ+6E+Nd8CNJjkzybUmOTfIT0/JnTF/fM4173yR/sOS+353kmzN7Ph87LTtget3eu5XHOzbJm5MckNn7dKuv1xhjYZuPmLZ5zk7+jACwRxOgAGDP9R9J7lNVlVkY+IUxxnVjjJsziyfHT+vdmuQBSR40xrh1jPH3Y4yxA4/zp2OM/zfG+EKSczMLRkmSMcYZY4ybxxhfSnJykkcs7KGyAj+W5CVjjGvGGJ9J8huZRY4Ft0633zrGOD/JLUkethPjPC7J34wxLhhj3Jrkd5LcI8l/zSw83C3J4VW1zxjjijHGv61w/Fu97xjj4jHG+8YYt40xrsgsUHz3kvv/1hjjpjHGZUk+lOQdY4zLxxg3Zhb2Hjmtd1KSPxpjvH+M8eUxxplJvpRZuFrq1swC0TclqTHGR8YYV9XspPLHJ3nR9HpdkeR3c/vne3teN8a4bIxxW2ah5olJfmaMcf30Gv3fnRjvgldM791PJnlVpviW2XvkldPzckuSFyU5vm5/2NzJY4zPTa/7Sr13jPFXY4yvjDG+sMLXa7Gd+RkBYI8mQAHAnuugJNdltlfP1yW5eDoc6IYkfzctT5LfTrI5yTumw5xeuIOP8+lFlz+f2V4oqaq9qurlVfVv06FMV0zrHJiV+YbM9qRZ8Ilp2YJrp9hxh8fekXEufZwxxleSfCrJQWOMzZntGXVykmuq6uyqWjyGrdrWfavqG2t22OOnp+fmN3PH5+XqRZe/sMz1hfE/KMnzFl7b6fU9JLd/rhbG9K7M9hB6zTSm06vqXtNj75M7Pt8HreRnnXxq0eVDklw3xrh+mfVWPN6tbHvx+2C598jeme3Jt9x9V+p291nh67XYzvyMALBHE6AAYA9UVY/KLB78Q5LPZhYsHj7GOGD62n+MsW+STHu8PG+M8ZAkP5TkF2s6V1GSHdkTaqmnZnYo0/dmdmjXoQvDW+G2/yOzf8gveOC0bLXd7nGmPcYOSbIlScYYZ40xHjOtM5K8Ylp1u8/NNu57WpKPJjlsjHGvzA57rOW3sl2fSnLKotf2gDHG140x3rSVMf3+GOPbkxye2aF4v5zZe+TW3PH53jJd/lxmEXPB1y+36SVjuk9VHXBnxzs5ZMm4Ft4Hy71HbsvtY93YyuVtWbrejr5eO/MzAsAeTYACgD1IVd2rqp6U5OwkbxxjXDrt0fPHSU6tqvtP6x1UVUdPl59UVQ+dwsuNmR069pVpk1dndm6dnbFfZocdXZtZvPjNJbdvb9tvSvLiqrpfzU4Y/mtJ3riN9XfWuUl+oGYnF98ns3NifSnJP1XVw6rqcdO5e76YWchb/NwcWlv5xLft3He/JDcluaWqvinJz96J8f9xkp+pqu+omXtW1Q9U1X7LjOlR03r7ZBaVvpjkK2OML0/PwylVtV9VPSjJL+Zrz/clSR5bVQ+cDqF80bYGNMa4KrPDBP+wqu5dVftU1cK5klY83kV+edrOIUmem+ScafmbkvxCVT24qvbN187LddtWtvOZzF6DHX1Pb+/1Wvpe3pmfEQD2aAIUAOwZ/rqqbs5sz4tfTfLK3P4E0i/I7DC7902HEP2ffO18SYdN129J8t4kfzjGePd02//KLALdUFWLTy6+Eq/P7JCoLUk+nOR9S25/bWbnR7qhFn1C2iIvS3JRkg8muTSzk5i/bAfHsF1jjI9ldnLvV2e2J9APZnbi9f/M7BxOL5+Wfzqzk2kvxJc/n75fW1UfWGbT27rvL2W2h9jNmcWKc5a5/0rHf1GSn8rs0LrrM3udn7GV1e81Pd71mb0212Z2CGaSPCezKHV5ZnvOnZXZibQzxrhgGuMHk1yc5G0rGNrTMtur6qNJrsnscMQdHe+Ct06Pe0mSv8nsvZNpfG/I7NPt/j2zoPacrW1kjPH5JKck+cfpfbfSczJt7/U6OcmZ0zafspM/IwDs0WrHzjEKAAC7TlWNzA592zzvsQAAO88eUAAAAAC0EqAAAAAAaOUQPAAAAABa2QMKAAAAgFYCFAAAAACt9p73AHaVAw88cBx66KHzHsaq+NznPpd73vOe8x4GrHnmCqyMuQIrY67AypgrsH170jy5+OKLPzvGuN/21ls3AerQQw/NRRddNO9hrIpNmzZl48aN8x4GrHnmCqyMuQIrY67AypgrsH170jypqk+sZD2H4AEAAADQSoACAAAAoJUABQAAAEArAQoAAACAVgIUAAAAAK0EKAAAAABaCVAAAAAAtBKgAAAAAGglQAEAAADQSoACAAAAoJUABQAAAEArAQoAAACAVgIUAAAAAK0EKAAAAABaCVAAAAAAtBKgAAAAAGglQAEAAADQSoACAAAAoJUABQAAAEArAQoAAACAVgIUAAAAAK0EKAAAAABaCVAAAAAAtBKgAAAAAGglQAEAAADQSoACAAAAoJUABQAAAEArAQoAAACAVgIUAAAAAK0EKAAAAABaCVAAAAAAtBKgAAAAAGglQAEAAADQSoACAAAAoJUABQAAAEArAQoAAACAVgIUAAAAAK0EKAAAAABaCVAAAAAAtBKgAAAAAGglQAEAAADQSoACAAAAoNXe8x4AAAAAwC5z1nHzHkHyDT877xHscvaAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFG0Hh0AACAASURBVFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKCVAAUAAABAKwEKAAAAgFYCFAAAAACtBCgAAAAAWglQAAAAALQSoAAAAABoJUABAAAA0EqAAgAAAKDVdgNUVR1SVe+uqg9X1WVV9dxp+clVtaWqLpm+vn/RfV5UVZur6mNVdfSi5cdMyzZX1QsXLX9wVb1/Wn5OVd11Wn636frm6fZDt/cYAAAAAKwtK9kD6rYkzxtjHJ7kqCTPqqrDp9tOHWMcMX2dnyTTbccneXiSY5L8YVXtVVV7JXlNkicmOTzJjy7aziumbT00yfVJTpyWn5jk+mn5qdN6W32MnX4WAAAAAGiz3QA1xrhqjPGB6fLNST6S5KBt3OXYJGePMb40xvj3JJuTPHr62jzGuHyM8Z9Jzk5ybFVVksclefN0/zOT/PCibZ05XX5zksdP62/tMQAAAABYY3boHFDTIXCPTPL+adGzq+qDVXVGVd17WnZQkk8tutuV07KtLb9vkhvGGLctWX67bU233zitv7VtAQAAALDG7L3SFatq3yR/keTnxxg3VdVpSV6aZEzffzfJT7SMcidV1UlJTkqSDRs2ZNOmTfMd0Cq55ZZb9pifBTqZK7Ay5gqsjLkCK2OusObdff6nkV6P82RFAaqq9sksPv3ZGOMvk2SMcfWi2/84ydumq1uSHLLo7gdPy7KV5dcmOaCq9p72clq8/sK2rqyqvZPsP62/rcf4qjHG6UlOT5IjjzxybNy4cSU/7pq3adOm7Ck/C3QyV2BlzBVYGXMFVsZcYc0767R5jyCb7vOz626erORT8CrJa5N8ZIzxykXLH7BotR9J8qHp8nlJjp8+we7BSQ5L8s9JLkxy2PSJd3fN7CTi540xRpJ3J3nydP8Tkrx10bZOmC4/Ocm7pvW39hgAAAAArDEr2QPqu5I8LcmlVXXJtOxXMvsUuyMyOwTviiQ/nSRjjMuq6twkH87sE/SeNcb4cpJU1bOTvD3JXknOGGNcNm3vBUnOrqqXJfmXzIJXpu9vqKrNSa7LLFpt8zEAAAAAWFu2G6DGGP+QpJa56fxt3OeUJKcss/z85e43xrg8y3yK3Rjji0n+x448BgAAAABryw59Ch4AAAAA7CgBCgAAAIBWAhQAAAAArQQoAAAAAFoJUAAAAAC0EqAAAAAAaCVAAQAAANBKgAIAAACglQAFAAAAQCsBCgAAAIBWAhQAAAAArQQoAAAAAFoJUAAAAAC0EqAAAAAAaCVAAQAAANBKgAIAAACglQAFAAAAQCsBCgAAAIBWAhQAAAAArQQoAAAAAFoJUAAAAAC0EqAAAAAAaCVAAQAAANBKgAIAAACglQAFAAAAQCsBCgAAAIBWAhQAAAAArQQoAAAAAFoJUAAAAAC0EqAAAAAAaCVAAQAAANBKgAIAAACglQAFAAAAQCsBCgAAAIBWAhQAAAAArQQoAAAAAFoJUAAAAAC0EqAAAAAAaCVAAQAAANBKgAIAAACglQAFAAAAQCsBCgAAAIBWAhQAAAAArQQoAAAAAFoJUAAAAAC0EqAAAAAAaCVAAQAAANBKgALg/7N3v6Gan3dexz/fTey26HabWB1iEnYLBqEWrO1sG1BktJikfZIKtbTCJltjI/0jCj6w+iTSulAf6GJhDWRNSCJm21J3aVhSY4h7WHyQNVVL/61LhrolCWmDTf841HXtcvlgfpHT8SRzZiafnHsmrxcc5r6v+3f/rusMXIR55/7dPwAAgCoBCgAAAIAqAQoAAACAKgEKAAAAgCoBCgAAAIAqAQoAAACAKgEKAAAAgCoBCgAAAIAqAQoAAACAKgEKAAAAgCoBCgAAAIAqAQoAAACAKgEKAAAAgCoBCgAAAIAqAQoAAACAKgEKAAAAgCoBCgAAAIAqAQoAAACAKgEKAAAAgCoBCgAAAIAqAQoAAACAKgEKAAAAgCoBCgAAAIAqAQoAAACAKgEKAAAAgCoBCgAAAIAqAQoAAACAKgEKAAAAgCoBCgAAAIAqAQoAAACAKgEKAAAAgCoBCgAAAIAqAQoAAACAKgEKAAAAgCoBCgAAAIAqAQoAAACAKgEKAAAAgCoBCgAAAIAqAQoAAACAKgEKAAAAgCoBCgAAAIAqAQoAAACAKgEKAAAAgCoBCgAAAIAqAQoAAACAKgEKAAAAgCoBCgAAAIAqAQoAAACAKgEKAAAAgCoBCgAAAIAqAQoAAACAKgEKAAAAgCoBCgAAAIAqAQoAAACAKgEKAAAAgCoBCgAAAIAqAQoAAACAKgEKAAAAgCoBCgAAAIAqAQoAAACAKgEKAAAAgCoBCgAAAIAqAQoAAACAKgEKAAAAgCoBCgAAAIAqAQoAAACAKgEKAAAAgCoBCgAAAIAqAQoAAACAKgEKAAAAgCoBCgAAAIAqAQoAAACAKgEKAAAAgCoBCgAAAIAqAQoAAACAKgEKAAAAgCoBCgAAAIAqAQoAAACAKgEKAAAAgCoBCgAAAIAqAQoAAACAKgEKAAAAgCoBCgAAAIAqAQoAAACAKgEKAAAAgCoBCgAAAIAqAQoAAACAKgEKAAAAgCoBCgAAAIAqAQoAAACAKgEKAAAAgCoBCgAAAIAqAQoAAACAKgEKAAAAgCoBCgAAAIAqAQoAAACAKgEKAAAAgCoBCgAAAIAqAQoAAACAqrMGqJm5dmZ+c2a+PjNfm5m/u41fOTOPzMwT259XbOMzM5+amZMz8+WZecu+c926Hf/EzNy6b/ytM/OV7T2fmpk53zkAAAAA2C2H+QTUj5L8/bXWG5Ncn+QjM/PGJB9L8uha67okj27Pk+SdSa7bfm5PcmdyOiYluSPJ25O8Lckdzwel7ZgP7nvfTdv4Oc0BAAAAwO45a4Baaz2z1vov2+P/meR3klyd5OYk922H3Zfk3dvjm5Pcv057LMnrZuaqJDcmeWSt9dxa67tJHkly0/baa9daj621VpL7zzjXucwBAAAAwI45p++AmpmfTfLnk/x2kmNrrWe2l76V5Nj2+OokT+5721Pb2IuNP3XAeM5jDgAAAAB2zOWHPXBm/liSf5vk7621frB9TVOSZK21ZmYV1ndBc8zM7Tl9iV6OHTuWvb29xtJedqdOnbpkfhdoslfgcOwVOBx7BQ7HXmHnvfrGo17BK3KfHCpAzcwfyen49G/WWr+2DX97Zq5aaz2zXf727Db+dJJr9739mm3s6SQnzhjf28avOeD485njx6y17kpyV5IcP358nThx4sxDLkp7e3u5VH4XaLJX4HDsFTgcewUOx15h5z1w9F8jvXflh15x++Qwd8GbJHcn+Z211j/f99KDSZ6/k92tST6/b/yW7U511yf5/nYZ3cNJbpiZK7YvH78hycPbaz+Ymeu3uW4541znMgcAAAAAO+Ywn4D6C0l+PslXZuZL29g/SvLJJJ+dmduSfDPJe7fXHkryriQnk/wwyQeSZK313Mx8Isnj23EfX2s9tz3+cJJ7k7wmyRe2n5zrHAAAAADsnrMGqLXWf0wyL/DyOw44fiX5yAuc654k9xww/sUkbzpg/DvnOgcAAAAAu+Wc7oIHAAAAAOdKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACg6qwBambumZlnZ+ar+8b+8cw8PTNf2n7ete+1fzgzJ2fmd2fmxn3jN21jJ2fmY/vG3zAzv72Nf2ZmXrWN/+T2/OT2+s+ebQ4AAAAAds9hPgF1b5KbDhj/pbXWm7efh5JkZt6Y5H1J/uz2nn85M5fNzGVJfjnJO5O8Mcn7t2OT5J9u5/rTSb6b5LZt/LYk393Gf2k77gXnOLdfGwAAAICXy1kD1Frrt5I8d8jz3Zzk02ut/73W+u9JTiZ52/Zzcq31jbXWHyT5dJKbZ2aS/JUkn9vef1+Sd+87133b488lecd2/AvNAQAAAMAOupDvgProzHx5u0Tvim3s6iRP7jvmqW3shcb/eJLvrbV+dMb4j51re/372/EvdC4AAAAAdtDl5/m+O5N8Isna/vxnSf7mS7Wol8rM3J7k9iQ5duxY9vb2jnZBL5FTp05dMr8LNNkrcDj2ChyOvQKHY6+w81599F8l/UrcJ+cVoNZa337+8cz8SpLf2J4+neTafYdes43lBca/k+R1M3P59imn/cc/f66nZubyJD+9Hf9ic5y5zruS3JUkx48fXydOnDin33NX7e3t5VL5XaDJXoHDsVfgcOwVOBx7hZ33wJ1HvYLsXfmhV9w+Oa9L8Gbmqn1P/1qS5++Q92CS9213sHtDkuuS/Kckjye5brvj3aty+kvEH1xrrSS/meQ92/tvTfL5fee6dXv8niT/YTv+heYAAAAAYAed9RNQM/OrSU4kef3MPJXkjiQnZubNOX0J3u8l+dtJstb62sx8NsnXk/woyUfWWn+4neejSR5OclmSe9ZaX9um+AdJPj0z/yTJf01y9zZ+d5J/PTMnc/pL0N93tjkAAAAA2D1nDVBrrfcfMHz3AWPPH/+LSX7xgPGHkjx0wPg3csBd7NZav5/kr5/LHAAAAADsngu5Cx4AAAAAnJUABQAAAECVAAUAAABAlQAFAAAAQJUABQAAAECVAAUAAABAlQAFAAAAQJUABQAAAECVAAUAAABAlQAFAAAAQJUABQAAAECVAAUAAABAlQAFAAAAQJUABQAAAECVAAUAAABAlQAFAAAAQJUABQAAAECVAAUAAABAlQAFAAAAQJUABQAAAECVAAUAAABAlQAFAAAAQJUABQAAAECVAAUAAABAlQAFAAAAQJUABQAAAECVAAUAAABAlQAFAAAAQJUABQAAAECVAAUAAABAlQAFAAAAQJUABQAAAECVAAUAAABAlQAFAAAAQJUABQAAAECVAAUAAABAlQAFAAAAQJUABQAAAECVAAUAAABAlQAFAAAAQJUABQAAAECVAAUAAABAlQAFAAAAQJUABQAAAECVAAUAAABAlQAFAAAAQJUABQAAAECVAAUAAABAlQAFAAAAQJUABQAAAECVAAUAAABAlQAFAAAAQJUABQAAAECVAAUAAABAlQAFAAAAQJUABQAAAECVAAUAAABAlQAFAAAAQJUABQAAAECVAAUAAABAlQAFAAAAQJUABQAAAECVAAUAAABAlQAFAAAAQJUABQAAAECVAAUAAABAlQAFAAAAQJUABQAAAECVAAUAAABAlQAFAAAAQJUABQAAAECVAAUAAABAlQAFAAAAQJUABQAAAECVAAUAAABAlQAFAAAAQJUABQAAAECVAAUAAABAlQAFAAAAQJUABQAAAECVAAUAAABAlQAFAAAAQJUABQAAAECVAAUAAABAlQAFAAAAQJUABQAAAECVAAUAAABAlQAFAAAAQJUABQAAAECVAAUAAABAlQAFAAAAQJUABQAAAECVAAUAAABAlQAFAAAAQJUABQAAAECVAAUAAABAlQAFAAAAQJUABQAAAECVAAUAAABAlQAFAAAAQJUABQAAAECVAAUAAABAlQAFAAAAQJUABQAAAECVAAUAAABAlQAFAAAAQJUABQAAAECVAAUAAABAlQAFAAAAQJUABQAAAECVAAUAAABAlQAFAAAAQJUABQAAAECVAAUAAABAlQAFAAAAQJUABQAAAECVAAUAAABAlQAFAAAAQJUABQAAAECVAAUAAABAlQAFAAAAQJUABQAAAEDV5Ue9AM7Dc99IHrjzqFeR/I3PHPUKAAAAgIuAT0ABAAAAUCVAAQAAAFAlQAEAAABQJUABAAAAUCVAAQAAAFAlQAEAAABQJUABAAAAUCVAAQAAAFAlQAEAAABQJUABAAAAUCVAAQAAAFAlQAEAAABQJUABAAAAUCVAAQAAAFAlQAEAAABQJUABAAAAUCVAAQAAAFAlQAEAAABQJUABAAAAUCVAAQAAAFAlQAEAAABQJUABAAAAUCVAAQAAAFB11gA1M/fMzLMz89V9Y1fOzCMz88T25xXb+MzMp2bm5Mx8eWbesu89t27HPzEzt+4bf+vMfGV7z6dmZs53DgAAAAB2z2E+AXVvkpvOGPtYkkfXWtcleXR7niTvTHLd9nN7kjuT0zEpyR1J3p7kbUnueD4obcd8cN/7bjqfOQAAAADYTWcNUGut30ry3BnDNye5b3t8X5J37xu/f532WJLXzcxVSW5M8sha67m11neTPJLkpu211661HltrrST3n3Guc5kDAAAAgB10vt8BdWyt9cz2+FtJjm2Pr07y5L7jntrGXmz8qQPGz2cOAAAAAHbQ5Rd6grXWmpn1UizmpZ5jZm7P6cv0cuzYsezt7b3USzsSp37ip7P36huPehnJJfL3yaXr1KlTl8y+hyZ7BQ7HXoHDsVfYeTvw7+lX4j453wD17Zm5aq31zHb527Pb+NNJrt133DXb2NNJTpwxvreNX3PA8eczx/9nrXVXkruS5Pjx4+vEiRMHHXbR2fu1e3Li9x8+6mUkJz5z1CuAF7W3t5dLZd9Dk70Ch2OvwOHYK+y8B47+q6T3rvzQK26fnO8leA8mef5Odrcm+fy+8Vu2O9Vdn+T722V0Dye5YWau2L58/IYkD2+v/WBmrt/ufnfLGec6lzkAAAAA2EFn/QTUzPxqTn966fUz81RO383uk0k+OzO3Jflmkvduhz+U5F1JTib5YZIPJMla67mZ+USSx7fjPr7Wev6LzT+c03fae02SL2w/Odc5AAAAANhNZw1Qa633v8BL7zjg2JXkIy9wnnuS3HPA+BeTvOmA8e+c6xwAAAAA7J7zvQQPAAAAAA5FgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKBKgAIAAACgSoACAAAAoEqAAgAAAKDq8qNeAOfuf/3BH+ZLT37vqJeRNx/1AgAAAICLgk9AAQAAAFAlQAEAAABQJUABAAAAUCVAAQAAAFAlQAEAAABQJUABAAAAUCVAAQAAAFAlQAEAAABQJUABAAAAUCVAAQAAAFAlQAEAAABQJUABAAAAUCVAAQAAAFAlQAEAAABQJUABAAAAUCVAAQAAAFAlQAEAAABQJUABAAAAUCVAAQAAAFAlQAEAAABQJUABAAAAUCVAAQAAAFAlQAEAAABQJUABAAAAUCVAAQAAAFAlQAEAAABQJUABAAAAUCVAAQAAAFAlQAEAAABQJUABAAAAUCVAAQAAAFAlQAEAAABQJUABAAAAUCVAAQAAAFAlQAEAAABQJUABAAAAUCVAAQAAAFAlQAEAAABQJUABAAAAUCVAAQAAAFAlQAEAAABQJUABAAAAUHVBAWpmfm9mvjIzX5qZL25jV87MIzPzxPbnFdv4zMynZubkzHx5Zt6y7zy3bsc/MTO37ht/63b+k9t758XmAAAAAGD3vBSfgPrLa603r7WOb88/luTRtdZ1SR7dnifJO5Nct/3cnuTO5HRMSnJHkrcneVuSO/YFpTuTfHDf+246yxwAAAAA7JjGJXg3J7lve3xfknfvG79/nfZYktfNzFVJbkzyyFrrubXWd5M8kuSm7bXXrrUeW2utJPefca6D5gAAAABgx1xogFpJ/v3M/OeZuX0bO7bWemZ7/K0kx7bHVyd5ct97n9rGXmz8qQPGX2wOAAAAAHbM5Rf4/r+41np6Zv5kkkdm5r/tf3GttWZmXeAcL+rF5tii2O1JcuzYsezt7TWX8rL5Pz95RZ75mfcc9TLyvUvk75NL16lTpy6ZfQ9N9gocjr0Ch2OvsPNefeNRr+AVuU8uKECttZ7e/nx2Zn49p7/D6dszc9Va65ntMrpnt8OfTnLtvrdfs409neTEGeN72/g1BxyfF5njzPXdleSuJDl+/Pg6ceLEQYdddL7w6V/JVd/83FEvI29+38NHvQR4UXt7e7lU9j002StwOPYKHI69ws574M6jXkH2rvzQK26fnPcleDPzR2fmp55/nOSGJF9N8mCS5+9kd2uSz2+PH0xyy3Y3vOuTfH+7jO7hJDfMzBXbl4/fkOTh7bUfzMz1293vbjnjXAfNAQAAAMCOuZBPQB1L8uun21AuT/LAWuvfzczjST47M7cl+WaS927HP5TkXUlOJvlhkg8kyVrruZn5RJLHt+M+vtZ6bnv84ST3JnlNki9sP0nyyReYAwAAAIAdc94Baq31jSR/7oDx7yR5xwHjK8lHXuBc9yS554DxLyZ502HnAAAAAGD3XOhd8AAAAADgRQlQAAAAAFQJUAAAAABUCVAAAAAAVAlQAAAAAFQJUAAAAABUCVAAAAAAVAlQAAAAAFQJUAAAAABUCVAAAAAAVAlQAAAAAFQJUAAAAABUCVAAAAAAVAlQAAAAAFQJUAAAAABUCVAAAAAAVAlQAAAAAFQJUAAAAABUCVAAAAAAVAlQzeH4vgAAIABJREFUAAAAAFQJUAAAAABUCVAAAAAAVAlQAAAAAFQJUAAAAABUCVAAAAAAVAlQAAAAAFQJUAAAAABUCVAAAAAAVAlQAAAAAFQJUAAAAABUCVAAAAAAVAlQAAAAAFQJUAAAAABUCVAAAAAAVAlQAAAAAFQJUAAAAABUCVAAAAAAVAlQAAAAAFQJUAAAAABUCVAAAAAAVAlQAAAAAFQJUAAAAABUCVAAAAAAVAlQAAAAAFQJUAAAAABUCVAAAAAAVAlQAAAAAFQJUAAAAABUCVAAAAAAVAlQAAAAAFQJUAAAAABUCVAAAAAAVAlQAAAAAFQJUAAAAABUCVAAAAAAVAlQAAAAAFQJUAAAAABUCVAAAAAAVAlQAAAAAFQJUAAAAABUCVAAAAAAVAlQAAAAAFQJUAAAAABUCVAAAAAAVAlQAAAAAFQJUAAAAABUCVAAAAAAVAlQAAAAAFQJUAAAAABUCVAAAAAAVAlQAAAAAFQJUAAAAABUCVAAAAAAVAlQAAAAAFQJUAAAAABUCVAAAAAAVAlQAAAAAFQJUAAAAABUCVAAAAAAVAlQAAAAAFQJUAAAAABUCVAAAAAAVAlQAAAAAFQJUAAAAABUCVAAAAAAVAlQAAAAAFQJUAAAAABUCVAAAAAAVAlQAAAAAFQJUAAAAABUCVAAAAAAVAlQAAAAAFQJUAAAAABUCVAAAAAAVAlQAAAAAFQJUAAAAABUCVAAAAAAVAlQAAAAAFQJUAAAAABUCVAAAAAAVAlQAAAAAFQJUAAAAABUCVAAAAAAVAlQAAAAAFQJUAAAAABUCVAAAAAAVAlQAAAAAFQJUAAAAABUCVAAAAAAVAlQAAAAAFQJUAAAAABUCVAAAAAAVAlQAAAAAFQJUAAAAABUCVAAAAAAVAlQAAAAAFQJUAAAAABUCVAAAAAAVAlQAAAAAFQJUAAAAABUCVAAAAAAVAlQAAAAAFQJUAAAAABUCVAAAAAAVAlQAAAAAFQJUAAAAABUCVAAAAAAVAlQAAAAAFQJUAAAAABUCVAAAAAAVF1+1Avg4nXbvY8f9RKSJHf/ws8d9RIAAAC4SHzpye8d9RKSP3XUC3j5+QQUAAAAAFUCFAAAAABVAhQAAAAAVQIUAAAAAFUCFAAAAABVAhQAAAAAVQIUAAAAAFUCFAAAAABVlx/1AuBC3Xbv40e9hNz9Cz931EsAAACAneUTUAAAAABUCVAAAAAAVLkED14Cu3AZYOJSQAAAAHaTT0ABAAAAUCVAAQAAAFDlEjwAXjYuV4XDsVcAgEvNRR2gZuamJP8iyWVJ/tVa65NHvCQgu/MPp7/0Uz/cmbUcNf+IBAAAjtJFG6Bm5rIkv5zkryZ5KsnjM/PgWuvrR7syODpiCwAAALvoog1QSd6W5ORa6xtJMjOfTnJzEgEKAAAAdsyu/A/zv3PUC3iFupgD1NVJntz3/Kkkbz+itQDstF35j/2u8Pfx43blclWXigJcGr75nd347wqwW2atddRrOC8z854kN621/tb2/OeTvH2t9dF9x9ye5Pbt6Z9J8rsv+0I7Xp/kfxz1IuAiYK/A4dgrcDj2ChyOvQJndyntk59Za/2Jsx10MX8C6ukk1+57fs029v+ste5KctfLuaiXw8x8ca11/KjXAbvOXoHDsVfgcOwVOBx7Bc7ulbhPfuKoF3ABHk9y3cy8YWZeleR9SR484jUBAAAAcIaL9hNQa60fzcxHkzyc5LIk96y1vnbEywIAAADgDBdtgEqStdZDSR466nUcgUvuskIosVfgcOwV/i979x+s6V3Wd/xzkQV0WPkZm6YhJVjC1JA64CLgqFMiDCSMY7BFTKwYJRproWML04FQOzAirbQWWkdARpMSFVgo2JKxYSKDq9YfiQShmICRJVITxVHMDwgoNnD1j3PHOSy7OWf3yXXOc8jrNfNMnv3e9/Xc37OZhw3vfX6wPZ4rsD2eK7C1+9zzZM9+CDkAAAAAe8Ne/gwoAAAAAPYAAWqNVdW5VXVjVR2uqpce5fgDq+pty/Frq+qMnd8l7L5tPFdeVFUfrqoPVdV7q+pRu7FP2E1bPU82nfdPq6qr6j71rSxwt+08V6rqucufKzdU1Vt2eo+wDrbx319/v6oOVdUHlv8Ge9Zu7BN2W1VdXlV/XlXXH+N4VdVPLc+lD1XV1+/0HneKALWmquqkJK9Lcl6Ss5JcWFVnHXHaxUlu6+7HJHltklfv7C5h923zufKBJE/s7q9L8o4k/3Fndwm7a5vPk1TVVyX5kSTX7uwOYT1s57lSVWcmuTTJN3X345L8qx3fKOyybf658qNJ3t7dT8jGN5a/fmd3CWvjTUnOvYfj5yU5c7ldkuQNO7CnXSFAra8nJTnc3Td1998kOZjk/CPOOT/JFcv9dyR5WlXVDu4R1sGWz5XuPtTdn11+eU2SR+7wHmG3befPlCR5ZTb+MuOvd3JzsEa281z5wSSv6+7bkqS7/3yH9wjrYDvPlU7y4OX+Q5L86Q7uD9ZGd/9Gklvv4ZTzk/x8b7gmyUOr6tSd2d3OEqDW12lJbt7061uWtaOe0913JbkjySN2ZHewPrbzXNns4iTvHt0RrJ8tnyfLy71P7+7/tZMbgzWznT9THpvksVX1W1V1TVXd099qw5er7TxXXpHke6rqlmx8c/m/3JmtwZ5zvP9/Zs/at9sbANgpVfU9SZ6Y5B/v9l5gnVTV/ZK8Jsn37fJWYC/Yl423STw1G6+o/Y2q+kfdffuu7grWz4VJ3tTd/7mqvjHJL1TV2d39hd3eGLA7vAJqff1JktM3/fqRy9pRz6mqfdl4aetf7sjuYH1s57mSqnp6kn+b5Nu7+3M7tDdYF1s9T74qydlJfq2qPp7kKUmu9EHk3Adt58+UW5Jc2d3/r7v/KMkfZiNIwX3Jdp4rFyd5e5J09+8k+YokJ+/I7mBv2db/n/lyIECtr/clObOqHl1VD8jGB/ddecQ5Vya5aLn/nCS/2t29g3uEdbDlc6WqnpDkjdmITz6rg/uie3yedPcd3X1yd5/R3Wdk47PSvr27r9ud7cKu2c5/f/3PbLz6KVV1cjbeknfTTm4S1sB2nit/nORpSVJVX5uNAPUXO7pL2BuuTPK9y7fhPSXJHd39id3e1ARvwVtT3X1XVb0wydVJTkpyeXffUFU/luS67r4yyWXZeCnr4Wx8qNkFu7dj2B3bfK78pyT7k/z35XP6/7i7v33XNg07bJvPE7jP2+Zz5eokz6iqDyf5fJJ/091egc59yjafKy9O8rNV9a+z8YHk3+cvy7kvqqq3ZuMvLk5ePhPt5UnunyTd/TPZ+Iy0ZyU5nOSzSb5/d3Y6r/xvAAAAAACTvAUPAAAAgFECFAAAAACjBCgAAAAARglQAAAAAIwSoAAAAAAYJUABAAAAMEqAAgAAAGCUAAUAAADAKAEKAAAAgFECFAAAAACjBCgAAAAARglQAAAAAIwSoAAAAAAYJUABAAAAMEqAAgAAAGCUAAUAAADAKAEKAAAAgFECFAAAAACjBCgAAAAARglQAAAAAIwSoAAAAAAYJUABAAAAMEqAAgAAAGCUAAUAAADAKAEKAAAAgFECFAAAAACjBCgAAAAARglQAAAAAIwSoAAAAAAYJUABAAAAMEqAAgAAAGCUAAUAAADAKAEKAAAAgFECFAAAAACjBCgAAAAARglQAAAAAIwSoAAAAAAYJUABAAAAMEqAAgAAAGCUAAUAAADAKAEKAAAAgFECFAAAAACjBCgAAAAARglQAAAAAIwSoAAAAAAYJUABAAAAMEqAAgAAAGCUAAUAAADAKAEKAAAAgFECFAAAAACjBCgAAAAARglQAAAAAIwSoAAAAAAYJUABAAAAMEqAAgAAAGCUAAUAAADAKAEKAAAAgFECFAAAAACjBCgAAAAARglQAAAAAIwSoAAAAAAYJUABAAAAMEqAAgAAAGCUAAUAAADAKAEKAAAAgFECFAAAAACjBCgAAAAARglQAAAAAIwSoAAAAAAYJUABAAAAMEqAAgAAAGCUAAUAAADAKAEKAAAAgFECFAAAAACjBCgAAAAARglQAAAAAIwSoAAAAAAYJUABAAAAMEqAAgAAAGCUAAUAAADAKAEKAAAAgFECFAAAAACjBCgAAAAARglQAAAAAIwSoAAAAAAYJUABAAAAMEqAAgAAAGCUAAUAAADAKAEKAAAAgFECFAAAAACjBCgAAAAARglQAAAAAIwSoAAAAAAYJUABAAAAMEqAAgAAAGCUAAUAAADAKAEKAAAAgFECFAAAAACjBCgAAAAARglQAAAAAIwSoAAAAAAYJUABAAAAMEqAAgAAAGCUAAUAAADAKAEKAAAAgFECFAAAAACjBCgAAAAARglQAAAAAIwSoAAAAAAYJUABAAAAMEqAAgAAAGCUAAUAAADAKAEKAAAAgFECFAAAAACjBCgAAAAARglQAAAAAIwSoAAAAAAYJUABAAAAMEqAAgAAAGCUAAUAAADAKAEKAAAAgFECFAAAAACjBCgAAAAARglQAAAAAIwSoAAAAAAYJUABAAAAMEqAAgAAAGCUAAUAAADAKAEKAAAAgFECFAAAAACjBCgAAAAARglQAAAAAIwSoAAAAAAYJUABAAAAMEqAAgAAAGCUAAUAAADAKAEKAAAAgFECFAAAAACjBCgAAAAARglQAAAAAIwSoAAAAAAYJUABAAAAMEqAAgAAAGCUAAUAAADAKAEKAAAAgFECFAAAAACjBCgAAAAARglQAAAAAIwSoAAAAAAYJUABAAAAMEqAAgAAAGCUAAUAAADAKAEKAGCPqKobquqpu70PAIDjJUABAKyoqr65qn67qu6oqlur6req6hvu7et09+O6+9fu7ccFAJi2b7c3AACwl1XVg5P8cpIfTvL2JA9I8i1JPnecj7Ovu++693cIALD7vAIKAGA1j02S7n5rd3++u/+qu3+luz+UJFX1/Kr6SFXdVlVXV9Wj7h6sqq6qF1TVR5N8tKreUFU/ufnBq+pdVfWi5f7Hq+rpy/2TquplVfWxqvp0Vb2/qk5fjv3DqnrP8mqsG6vquTv0ewEAcFQCFADAav4wyeer6oqqOq+qHnb3gao6P8nLkvyTJF+d5H8neesR889O8uQkZy3Hvquqapl/WJJnJDl4lOu+KMmFSZ6V5MFJnp/ks1X1oCTvSfKWJH8nyQVJXl9VZ907Py4AwPEToAAAVtDdn0ryzUk6yc8m+YuqurKqTknyz5P8h+7+yPL2un+f5PGbXwW1HL+1u/8qG4Gqs/EWviR5TpLf6e4/PcqlfyDJj3b3jb3h/3T3Xyb5tiQf7+7/1t13dfcHkrwzyXfe+z89AMD2CFAAACtaAtP3dfcjk5yd5O8l+S9JHpXkv1bV7VV1e5Jbk1SS0zaN37zpcTobr3a6cFn67iRvPsZlT0/ysaOsPyrJk+++5nLdf5bk757wDwgAsCIBCgDgXtTdf5DkTdkIUTcn+aHufuim21d2929vHjniId6a5DnLq6SenI1XLx3NzUn+wTHWf/2Ia+7v7h9e5ecCAFiFAAUAsILlA79fXFWPXH59ejZewXRNkp9JcmlVPW459pCquse3wi1vmftkkp9LcnV3336MU38uySur6sza8HVV9YhsfCPfY6vqeVV1/+X2DVX1tffKDwwAcAIEKACA1Xw6G69UuraqPpON8HR9khd39/9I8uokB6vqU8v6edt4zLckefryz2N5TZK3J/mVJJ9KclmSr+zuT2fjg8svSPKnSf5s2cMDj/9HAwC4d9TGRw0AAAAAwAyvgAIAAABglAAFAAAAwCgBCgAAAIBRAhQAAAAAowQoAAAAAEbt2+0N7JSTTz65zzjjjN3exr3iM5/5TB70oAft6OxuXHOV2b2231Vm7Xd2dq/td5VZ+13fWfudnd1r+11l1n5nZ/fafleZtd/1nbXf2dm9tt9VZu13dnaVa66b97///Z/s7q/e8sTuvk/cDhw40F8uDh06tOOzu3HNVWb32n5XmbXf2dm9tt9VZu13fWftd3Z2r+13lVn7nZ3da/tdZdZ+13fWfmdn99p+V5m139nZVa65bpJc19voMt6CBwAAAMAoAQoAAACAUQIUAAAAAKMEKAAAAABGCVAAAAAAjBKgAAAAABglQAEAAAAwSoACAAAAYJQABQAAAMAoAQoAAACAUQIUAAAAAKMEKAAAAABGCVAAAAAAjBKgAAAAABglQAEAAAAwSoACAAAAYJQABQAAAMAoAQoAAACAUQIUAAAAAKMEKAAAAABGCVAAAAAAjBKgAAAAABglQAEAAAAwSoACAAAAYJQABQAAAMAoAQoAAACAUQIUAAAAAKMEKAAAAABGCVAAAAAAjBKgAAAAABglQAEAAAAwSoACAAAAYJQABQAAAMAoAQoAAACAUQIUAAAAAKMEKAAAAABGCVAAAAAAjBKgAAAAABglQAEAAAAwSoACAAAAYJQABQAAAMAoAQoAAACAUft2ewOwsrd815eufcUzk7e84cQe70Rmv/ttJ3YtAAAAuA/wCigAAAAARglQAAAAAIwSoAAAAAAYJUABAAAAMEqAAgAAAGCUAAUAAADAKAEKAAAAgFECFAAAAACjBCgAAAAARglQAAAAAIwSoAAAAAAYJUABAAAAMEqAAgAAAGCUAAUAAADAKAEKAAAAgFECFAAAAACjBCgAAAAARglQAAAAAIwSoAAAAAAYJUABAAAAMEqAAgAAAGCUAAUAAADAKAEKAAAAgFECFAAAAACjBCgAAAAARglQAAAAAIwSoAAAAAAYJUABAAAAMEqAAgAAAGCUAAUAAADAKAEKAAAAgFECFAAAAACjBCgAAAAARglQAAAAAIwSoAAAAAAYJUABAAAAMEqAAgAAAGCUAAUAAADAKAEKAAAAgFECFAAAAACjBCgAAAAARglQAAAAAIwSoAAAAAAYJUABAAAAMEqAAgAAAGCUAAUAAADAKAEKAAAAgFECFAAAAACjBCgAAAAARglQAAAAAIwSoAAAAAAYJUABAAAAMEqAAgAAAGCUAAUAAADAKAEKAAAAgFECFAAAAACjBCgAAAAARglQAAAAAIwSoAAAAAAYJUABAAAAMEqAAgAAAGCUAAUAAADAKAEKAAAAgFECFAAAAACjBCgAAAAARglQAAAAAIwSoAAAAAAYJUABAAAAMEqAAgAAAGCUAAUAAADAKAEKAAAAgFECFAAAAACjBCgAAAAARglQAAAAAIwSoAAAAAAYJUABAAAAMEqAAgAAAGCUAAUAAADAKAEKAAAAgFECFAAAAACjBCgAAAAARglQAAAAAIwSoAAAAAAYJUABAAAAMEqAAgAAAGCUAAUAAADAKAEKAAAAgFECFAAAAACjBCgAAAAARglQAAAAAIwSoAAAAAAYJUABAAAAMEqAAgAAAGCUAAUAAADAKAEKAAAAgFFbBqiqOr2qDlXVh6vqhqr6kWX9FVX1J1X1weX2rE0zl1bV4aq6saqeuWn93GXtcFW9dNP6o6vq2mX9bVX1gGX9gcuvDy/Hz9jqGgAAAACsl+28AuquJC/u7rOSPCXJC6rqrOXYa7v78cvtqiRZjl2Q5HFJzk3y+qo6qapOSvK6JOclOSvJhZse59XLYz0myW1JLl7WL05y27L+2uW8Y17jhH8XAAAAABizZYDq7k909+8t9z+d5CNJTruHkfOTHOzuz3X3HyU5nORJy+1wd9/U3X+T5GCS86uqknxrkncs81ckefamx7piuf+OJE9bzj/WNQAAAABYM8f1GVDLW+CekOTaZemFVfWhqrq8qh62rJ2W5OZNY7csa8daf0SS27v7riPWv+ixluN3LOcf67EAAAAAWDPV3ds7sWp/kl9P8qru/qWqOiXJJ5N0klcmObW7n19VP53kmu7+xWXusiTvXh7m3O7+gWX9eUmenOQVy/mPWdZPT/Lu7j67qq5fZm5Zjn3siJkvukZ33/0qqrv3fEmSS5LklFNOOXDw4MHj/f1ZS3feeWf279+/o7O7cc1tz95605fO3e8h2f+FO07smicy+/Cv2Zj172Ytr7nK7F7b7yqz9ru+s/Y7O7vX9rvKrP3Ozu61/a4ya7/rO2u/s7N7bb+rzNrv7Owq11w355xzzvu7+4lbntjdW96S3D/J1UledIzjZyS5frl/aZJLNx27Osk3LrerN61futwqGyFr37L+t+fdPbvc37ecV8e6xj39DAcOHOgvF4cOHdrx2d245rZn3/zcL7kdeudlR13fzu2EZnfqZ70X53Zr1n7Xd9Z+13fWfmdn99p+V5m139nZvbbfVWbtd31n7Xd2dq/td5VZ+52dXeWa6ybJdb2NtrSdb8GrJJcl+Uh3v2bT+qmbTvuOJNcv969McsHyDXaPTnJmkt9N8r4kZy7fePeAbHyI+JXLZg8lec4yf1GSd216rIuW+89J8qvL+ce6BgAAAABrZt82zvmmJM9L8vtV9cFl7WXZ+Ba7x2fjLXgfT/JDSdLdN1TV25N8OBvfoPeC7v58klTVC7PxaqWTklze3Tcsj/eSJAer6seTfCAbwSvLP3+hqg4nuTUb0eoerwEAAADAetkyQHX3b2bjbW9HuuoeZl6V5FVHWb/qaHPdfVOO8i123f3XSb7zeK4BAAAAwHo5rm/BAwAAAIDjJUABAAAAMEqAAgAAAGCUAAUAAADAKAEKAAAAgFECFAAAAACjBCgAAAAARglQAAAAAIwSoAAAAAAYJUABAAAAMEqAAgAAAGCUAAUAAADAKAEKAAAAgFECFAAAAACjBCgAAAAARglQAAAAAIwSoAAAAAAYJUABAAAAMEqAAgAAAGCUAAUAAADAKAEKAAAAgFECFAAAAACjBCgAAAAARglQAAAAAIwSoAAAAAAYJUABAAAAMEqAAgAAAGCUAAUAAADAKAEKAAAAgFECFAAAAACjBCgAAAAARglQAAAAAIwSoAAAAAAYJUABAAAAMEqAAgAAAGCUAAUAAADAKAEKAAAAgFECFAAAAACjBCgAAAAARglQAAAAAIwSoAAAAAAYJUABAAAAMEqAAgAAAGCUAAUAAADAKAEKAAAAgFECFAAAAACjBCgAAAAARglQAAAAAIwSoAAAAAAYJUABAAAAMEqAAgAAAGCUAAUAAADAKAEKAAAAgFECFAAAAACjBCgAAAAARglQAAAAAIwSoAAAAAAYJUABAAAAMEqAAgAAAGCUAAUAAADAKAEKAAAAgFECFAAAAACjBCgAAAAARglQAAAAAIwSoAAAAAAYJUABAAAAMEqAAgAAAGCUAAUAAADAKAEKAAAAgFECFAAAAACjBCgAAAAARglQAAAAAIwSoAAAAAAYJUABAAAAMEqAAgAAAGCUAAUAAADAKAEKAAAAgFECFAAAAACjBCgAAAAARglQAAAAAIwSoAAAAAAYJUABAAAAMEqAAgAAAGCUAAUAAADAKAEKAAAAgFECFAAAAACjBCgAAAAARglQAAAAAIwSoAAAAAAYJUABAAAAMEqAAgAAAGCUAAUAAADAKAEKAAAAgFECFAAAAACjBCgAAAAARglQAAAAAIwSoAAAAAAYJUABAAAAMEqAAgAAAGCUAAUAAADAKAEKAAAAgFECFAAAAACjBCgAAAAARglQAAAAAIwSoAAAAAAYJUABAAAAMEqAAgAAAGCUAAUAAADAKAEKAAAAgFECFAAAAACjBCgAAAAARglQAAAAAIwSoAAAAAAYJUABAAAAMEqAAgAAAGCUAAUAAADAKAEKAAAAgFECFAAAAACjBCgAAAAARglQAAAAAIwSoAAAAAAYJUABAAAAMEqAAgAAAGCUAAUAAADAKAEKAAAAgFECFAAAAACjBCgAAAAARglQAAAAAIwSoAAAAAAYJUABAAAAMEqAAgAAAGCUAAUAAADAKAEKAAAAgFECFAAAAACjBCgAAAAARglQAAAAAIwSoAAAAAAYJUABAAAAMEqAAgAAAGCUAAUAAADAKAEKAAAAgFECFAAAAACjBCgAAAAARglQAAAAAIwSoAAAAAAYJUABAAAAMEqAAgAAAGDUlgGqqk6vqkNV9eGquqGqfmRZf3hVvaeqPrr882HLelXVT1XV4ar6UFV9/abHumg5/6NVddGm9QNV9fvLzE9VVZ3oNQAAAABYL9t5BdRdSV7c3WcleUqSF1TVWUlemuS93X1mkvcuv06S85KcudwuSfKGZCMmJXl5kicneVKSl98dlJZzfnDT3LnL+nFdAwAAAID1s2WA6u5PdPfvLfc/neQjSU5Lcn6SK5bTrkjy7OX++Ul+vjdck+ShVXVqkmcmeU9339rdtyV5T5Jzl2MP7u5ruruT/PwRj3U81wAAAABgzRzXZ0BV1RlJnpDk2iSndPcnlkN/luSU5f5pSW7eNHbLsnZP67ccZT0ncA0AAAAA1kxtvOhoGydW7U/y60le1d2/VFW3d/dDNx2/rbsfVlW/nOQnuvs3l/X3JnlJkqcm+Yru/vFl/d8l+askv7ac//Rl/VuSvKS7v+14r9Hd1x2x50uy8Ra9nHLKKQcOHjx4vL8/a+nOO+/M/v37d3R2N6657dlbb/rSufs9JPu/cMeJXfNEZh/+NRuz/t2s5TVXmd1r+11l1n7Xd9Z+Z2f32n5XmbXf2dm9tt9VZu13fWftd3Z2r+13lVn7nZ1d5Zrr5pxzznl/dz9xyxO7e8tbkvsnuTrJizat3Zjk1OX+qUluXO6/McmFR56X5MIkb9y0/sZl7dQkf7Bp/W/PO95r3NPPcODAgf5ycejQoR2f3Y1rbnv2zc/9ktuhd1521PXt3E5odqd+1ntxbrdm7Xd9Z+13fWftd3Z2r+13lVn7nZ3da/tdZdZ+13fWfmdn99p+V5m139nZVa65bpJc19toS9v5FrxKclmSj3T3azYdujLJ3d9kd1GSd21a/97lm+qekuSO3ngb3dVJnlFVD1s+fPwZSa5ejn2jmEh/AAAgAElEQVSqqp6yXOt7j3is47kGAAAAAGtm3zbO+aYkz0vy+1X1wWXtZUl+Isnbq+riJP83yXOXY1cleVaSw0k+m+T7k6S7b62qVyZ533Lej3X3rcv9f5HkTUm+Msm7l1uO9xoAAAAArJ8tA1RvfM5SHePw045yfid5wTEe6/Iklx9l/bokZx9l/S+P9xoAAAAArJfj+hY8AAAAADheAhQAAAAAowQoAAAAAEYJUAAAAACMEqAAAAAAGCVAAQAAADBKgAIAAABglAAFAAAAwCgBCgAAAIBRAhQAAAAAowQoAAAAAEYJUAAAAACMEqAAAAAAGCVAAQAAADBKgAIAAABglAAFAAAAwCgBCgAAAIBRAhQAAAAAowQoAAAAAEYJUAAAAACMEqAAAAAAGCVAAQAAADBKgAIAAABglAAFAAAAwCgBCgAAAIBRAhQAAAAAowQoAAAAAEYJUAAAAACMEqAAAAAAGCVAAQAAADBKgAIAAABglAAFAAAAwCgBCgAAAIBRAhQAAAAAowQoAAAAAEYJUAAAAACMEqAAAAAAGCVAAQAAADBKgAIAAABglAAFAAAAwCgBCgAAAIBRAhQAAAAAowQoAAAAAEYJUAAAAACMEqAAAAAAGCVAAQAAADBKgAIAAABglAAFAAAAwCgBCgAAAIBRAhQAAAAAowQoAAAAAEYJUAAAAACMEqAAAAAAGCVAAQAAADBKgAIAAABglAAFAAAAwCgBCgAAAIBRAhQAAAAAowQoAAAAAEYJUAAAAACMEqAAAAAAGCVAAQAAADBKgAIAAABglAAFAAAAwCgBCgAAAIBRAhQAAAAAowQoAAAAAEYJUAAAAACMEqAAAAAAGCVAAQAAADBKgAIAAABglAAFAAAAwCgBCgAAAIBRAhQAAAAAowQoAAAAAEYJUAAAAACMEqAAAAAAGCVAAQAAADBKgAIAAABglAAFAAAAwCgBCgAAAIBRAhQAAAAAowQoAAAAAEYJUAAAAACMEqAAAAAAGCVAAQAAADBKgAIAAABglAAFAAAAwCgBCgAAAIBRAhQAAAAAowQoAAAAAEYJUAAAAACMEqAAAAAAGCVAAQAAADBKgAIAAABglAAFAAAAwCgBCgAAAIBRAhQAAAAAowQoAAAAAEYJUAAAAACMEqAAAAAAGCVAAQAAADBKgAIAAABglAAFAAAAwCgBCgAAAIBRAhQAAAAAowQoAAAAAEYJUAAAAACMEqAAAAAAGCVAAQAAADBKgAIAAABglAAFAAAAwCgBCgAAAIBRAhQAAAAAowQoAAAAAEYJUAAAAACMEqAAAAAAGCVAAQAAADBKgAIAAABglAAFAAAAwCgBCgAAAIBRAhQAAAAAowQoAAAAAEYJUAAAAACMEqAAAAAAGCVAAQAAADBKgAIAAABglAAFAAAAwCgBCgAAAIBRAhQAAAAAowQoAAAAAEYJUAAAAACMEqAAAAAAGCVAAQAAADBKgAIAAABglAAFAAAAwCgBCgAAAIBRAhQAAAAAowQoAAAAAEYJUAAAAACMEqAAAAAAGCVAAQAAADBKgAIAAABglAAFAAAAwCgBCgAAAIBRAhQAAAAAo7YMUFV1eVX9eVVdv2ntFVX1J1X1weX2rE3HLq2qw1V1Y1U9c9P6ucva4ap66ab1R1fVtcv626rqAcv6A5dfH16On7HVNQAAAABYP9t5BdSbkpx7lPXXdvfjl9tVSVJVZyW5IMnjlpnXV9VJVXVSktclOS/JWUkuXM5Nklcvj/WYJLcluXhZvzjJbcv6a5fzjnmN4/uxAQAAANgpWwao7v6NJLdu8/HOT3Kwuz/X3X+U5HCSJy23w919U3f/TZKDSc6vqkryrUnescxfkeTZmx7riuX+O5I8bTn/WNcAAAAAYA2t8hlQL6yqDy1v0XvYsnZakps3nXPLsnas9Uckub277zpi/Yseazl+x3L+sR4LAAAAgDVU3b31SRufv/TL3X328utTknwySSd5ZZJTu/v5VfXTSa7p7l9czrssybuXhzm3u39gWX9ekicnecVy/mOW9dOTvLu7z14+c+rc7r5lOfaxI2a+6BrdfferqDbv+5IklyTJKaeccuDgwYPH9Zuzru68887s379/R2d345rbnr31pi+du99Dsv8Ld5zYNU9k9uFfszHr381aXnOV2b2231Vm7Xd9Z+13dnav7XeVWfudnd1r+11l1n7Xd9Z+Z2f32n5XmbXf2dlVrrluzjnnnPd39xO3PLG7t7wlOSPJ9VsdS3Jpkks3Hbs6yTcut6s3rV+63CobIWvfsv635909u9zft5xXx7rGVj/DgQMH+svFoUOHdnx2N6657dk3P/dLbofeedlR17dzO6HZnfpZ78W53Zq13/Wdtd/1nbXf2dm9tt9VZu13dnav7XeVWftd31n7nZ3da/tdZdZ+Z2dXuea6SXJdb6MtndBb8Krq1E2//I4kd39D3pVJLli+we7RSc5M8rtJ3pfkzOUb7x6QjQ8Rv3LZ6KEkz1nmL0ryrk2PddFy/zlJfnU5/1jXAAAAAGAN7dvqhKp6a5KnJjm5qm5J8vIkT62qx2fjLXgfT/JDSdLdN1TV25N8OMldSV7Q3Z9fHueF2Xi10klJLu/uG5ZLvCTJwar68SQfSHLZsn5Zkl+oqsPZ+BD0C7a6BgAAAADrZ8sA1d0XHmX5sqOs3X3+q5K86ijrVyW56ijrN+Uo32LX3X+d5DuP5xoAAAAArJ9VvgUPAAAAALYkQAEAAAAwSoACAAAAYJQABQAAAMAoAQoAAACAUQIUAAAAAKMEKAAAAABGCVAAAAAAjBKgAAAAABglQAEAAAAwSoACAAAAYJQABQAAAMAoAQoAAACAUQIUAAAAAKMEKAAAAABGCVAAAAAAjBKgAAAAABglQAEAAAAwSoACAAAAYJQABQAAAMAoAQoAAACAUQIUAAAAAKMEKAAAAABGCVAAAADw/9m78zBbisJu/N/iAiIiCCiIICiKKG6EHVEW2XHBLeIKKIoKGo1K1MQtalyymCgqxiACMYokalyiIi7EJfpG3E1Moq9ZhLgkosnPV/NLXu33j6rj9B1mOWfmFndGP5/nmefO6XOqu06f6uqqb/eZC3QlgAIAAACgKwEUAAAAAF0JoAAAAADoSgAFAAAAQFcCKAAAAAC6EkABAAAA0JUACgAAAICuBFAAAAAAdCWAAgAAAKArARQAAAAAXQmgAAAAAOhKAAUAAABAVwIoAAAAALoSQAEAAADQlQAKAAAAgK4EUAAAAAB0JYACAAAAoCsBFAAAAABdCaAAAAAA6EoABQAAAEBXAigAAAAAuhJAAQAAANCVAAoAAACArgRQAAAAAHQlgAIAAACgKwEUAAAAAF0JoAAAAADoSgAFAAAAQFcCKAAAAAC6EkABAAAA0JUACgAAAICuBFAAAAAAdCWAAgAAAKArARQAAAAAXQmgAAAAAOhKAAUAAABAVwIoAAAAALoSQAEAAADQlQAKAAAAgK4EUAAAAAB0JYACAAAAoCsBFAAAAABdCaAAAAAA6EoABQAAAEBXAigAAAAAuhJAAQAAANCVAAoAAACArgRQAAAAAHQlgAIAAACgKwEUAAAAAF0JoAAAAADoSgAFAAAAQFcCKAAAAAC6EkABAAAA0JUACgAAAICuBFAAAAAAdCWAAgAAAKArARQAAAAAXQmgAAAAAOhKAAUAAABAVwIoAAAAALoSQAEAAADQlQAKAAAAgK4EUAAAAAB0JYACAAAAoCsBFAAAAABdCaAAAAAA6EoABQAAAEBXAigAAAAAuhJAAQAAANCVAAoAAACArgRQAAAAAHQlgAIAAACgKwEUAAAAAF0JoAAAAADoSgAFAAAAQFcCKAAAAAC6EkABAAAA0JUACgAAAICuBFAAAAAAdCWAAgAAAKArARQAAAAAXQmgAAAAAOhKAAUAAABAVwIoAAAAALoSQAEAAADQlQAKAAAAgK4EUAAAAAB0JYACAAAAoCsBFAAAAABdCaAAAAAA6EoABQAAAEBXAigAAAAAuhJAAQAAANCVAAoAAACArgRQAAAAAHQlgAIAAACgKwEUAAAAAF0JoAAAAADoSgAFAAAAQFcCKAAAAAC6EkABAAAA0JUACgAAAICuBFAAAAAAdCWAAgAAAKArARQAAAAAXQmgAAAAAOhKAAUAAABAVwIoAAAAALoSQAEAAADQlQAKAAAAgK4EUAAAAAB0JYACAAAAoCsBFAAAAABdCaAAAAAA6EoABQAAAEBXAigAAAAAuhJAAQAAANDVlpu7AgDdvOW0ud+3OTF5ywWzr2Ol5cZlH/G2lZUHAAD4OeEOKAAAAAC6EkABAAAA0JUACgAAAICuBFAAAAAAdLVsAFVKuaiU8t1SyldGy3YqpVxZSvla+3fHtryUUl5dSvl6KeVLpZQDRmXOaK//WinljNHyA0spX25lXl1KKSvdBgAAAABrzzR3QF2c5KR5y56d5MPDMOyT5MPtcZKcnGSf9nN2kguSGiYleUGSQ5MckuQFk0Cpvebxo3InrWQbAAAAAKxNywZQwzB8LMl18xafmuSS9vslSR4wWn7pUH06yc1KKbslOTHJlcMwXDcMw/eTXJnkpPbc9sMwfHoYhiHJpfPWNcs2AAAAAFiDVvo3oHYdhuFb7fdvJ9m1/b57km+OXndNW7bU8msWWL6SbQAAAACwBpV649EyLyrlNkneOwzDXdrjHwzDcLPR898fhmHHUsp7k7x8GIZPtOUfTvKsJEcn2WYYhpe05c9L8uMkV7XXH9eW3yvJs4ZhuO+s2xiG4eoF6n126tf0suuuux542WWXzbRz1qof/vCH2W677W7Qsptjm1OXve4b1y+3xQ7Z7qf/sbJtrqTsTnvXsj6btbXNUdtYaZvYJG2ptY+Zyv68fzaboOx6q+9qyqpv37Lrrb6rKau+fcuut/qupqz6rt2y6tu37Hqr72rKqm/fsqvZ5lpzzDHHfHYYhoOWe92WK1z/d0opuw3D8K329bfvtuXXJrn16HV7tGXXpoZQ4+VXteV7LPD6lWzjeoZheEOSNyTJQQcdNBx99NELvWzdueqqq7LS97LSsptjm1OXfcv1/wzYVducmKP/64qVbXMlZY9+Wy3rs1lb2xy1jZW2iU3Sllr7mKnsz/tnswnKrrf6rqas+vYtu97qu5qy6tu37Hqr72rKqu/aLau+fcuut/qupqz69i27mm2uVyv9Ct67k0z+J7szkrxrtPz09j/VHZbkP9rX6K5IckIpZcf2x8dPSHJFe+4/SymHtf/97vR565plGwAAAACsQcveAVVKeWvq3Us3L6Vck/q/2b08yeWllLOS/HOSh7aXvy/JKUm+nuRHSR6TJMMwXFdKeXGSz7TXvWgYhskfNj8n9X/au3GS97efzLoNAAAAANamZQOoYRgevshTxy7w2iHJuYus56IkFy2w/Ookd1lg+fdm3QYAAAAAa89Kv4IHAAAAAFMRQAEAAADQlQAKAAAAgK4EUAAAAAB0JYACAAAAoCsBFAAAAABdCaAAAAAA6EoABQAAAEBXAigAAAAAuhJAAQAAANCVAAoAAACArgRQAAAAAHQlgAIAAACgKwEUAAAAAF0JoAAAAADoSgAFAAAAQFcCKAAAAAC6EkABAAAA0JUACgAAAICuBFAAAAAAdCWAAgAAAKArARQAAAAAXQmgAAAAAOhKAAUAAABAVwIoAAAAALoSQAEAAADQlQAKAAAAgK4EUAAAAAB0JYACAAAAoCsBFAAAAABdCaAAAAAA6EoABQAAAEBXAigAAAAAuhJAAQAAANCVAAoAAACArgRQAAAAAHQlgAIAAACgKwEUAAAAAF0JoAAAAADoSgAFAAAAQFcCKAAAAAC6EkABAAAA0JUACgAAAICuBFAAAAAAdCWAAgAAAKArARQAAAAAXQmgAAAAAOhKAAUAAABAVwIoAAAAALoSQAEAAADQlQAKAAAAgK4EUAAAAAB0JYACAAAAoCsBFAAAAABdCaAAAAAA6EoABQAAAEBXAigAAAAAuhJAAQAAANCVAAoAAACArgRQAAAAAHQlgAIAAACgKwEUAAAAAF0JoAAAAADoSgAFAAAAQFcCKAAAAAC6EkABAAAA0JUACgAAAICuBFAAAAAAdCWAAgAAAKArARQAAAAAXQmgAAAAAOhKAAUAAABAVwIoAAAAALoSQAEAAADQlQAKAAAAgK4EUAAAAAB0JYACAAAAoCsBFAAAAABdCaAAAAAA6EoABQAAAEBXAigAAAAAuhJAAQAAANCVAAoAAACArgRQAAAAAHQlgAIAAACgKwEUAAAAAF0JoAAAAADoSgAFAAAAQFcCKAAAAAC6EkABAAAA0JUACgAAAICuBFAAAAAAdCWAAgAAAKArARQAAAAAXQmgAAAAAOhKAAUAAABAVwIoAAAAALoSQAEAAADQlQAKAAAAgK4EUAAAAAB0JYACAAAAoCsBFAAAAABdCaAAAAAA6EoABQAAAEBXAigAAAAAuhJAAQAAANCVAAoAAACArgRQAAAAAHQlgAIAAACgKwEUAAAAAF0JoAAAAADoSgAFAAAAQFcCKAAAAAC62nJzVwDYfP75ez/KWRd/ZurXv/HMgzvWBgAAgJ9X7oACAAAAoCsBFAAAAABdCaAAAAAA6EoABQAAAEBXAigAAAAAuhJAAQAAANCVAAoAAACArgRQAAAAAHQlgAIAAACgKwEUAAAAAF0JoAAAAADoSgAFAAAAQFcCKAAAAAC6EkABAAAA0JUACgAAAICuBFAAAAAAdCWAAgAAAKArARQAAAAAXQmgAAAAAOhKAAUAAABAVwIoAAAAALoSQAEAAADQlQAKAAAAgK4EUAAAAAB0JYACAAAAoCsBFAAAAABdCaAAAAAA6EoABQAAAEBXAigAAAAAuhJAAQAAANCVAAoAAACArgRQAAAAAHS1qgCqlPJPpZQvl1K+UEq5ui3bqZRyZSnla+3fHdvyUkp5dSnl66WUL5VSDhit54z2+q+VUs4YLT+wrf/rrWxZahsAAAAArD2b4g6oY4Zh2H8YhoPa42cn+fAwDPsk+XB7nCQnJ9mn/Zyd5IKkhklJXpDk0CSHJHnBKFC6IMnjR+VOWmYbAAAAAKwxPb6Cd2qSS9rvlyR5wGj5pUP16SQ3K6XsluTEJFcOw3DdMAzfT3JlkpPac9sPw/DpYRiGJJfOW9dC2wAAAABgjVltADUk+WAp5bOllLPbsl2HYfhW+/3bSXZtv++e5Jujste0ZUstv2aB5UttAwAAAIA1ptSbi1ZYuJTdh2G4tpSyS+qdS09J8u5hGG42es33h2HYsZTy3iQvH4bhE235h5M8K8nRSbYZhuElbfnzkvw4yVXt9ce15fdK8qxhGO5bSvnBQttYoH5np37dL7vuuuuBl1122Yrf61rywx/+MNttt90NWnZzbHPqstd94/rlttgh2/30P1a2zZWU3WnvWnadfTbf+8F/5oc/2XLq1++187ar3u4Nuo9GbWOlbWKTtKXWPmYqu87akvr2Lau+fcuut/qupqz69i273uq7mrLqu3bLqm/fsuutvqspq759y65mm2vNMccc89nRn2Va1PQzzwUMw3Bt+/e7pZR3pv4Np++UUnYbhuFb7Wt0320vvzbJrUfF92jLrk0NocbLr2rL91jg9VliG/Pr94Ykb0iSgw46aDj66KMXetm6c9VVV2Wl72WlZTfHNqcu+5YLrl9umxNz9H9dsbJtrqTs0W+rZdfZZ3PJ29+Xj/1/t5j69W988MGr3u4Nuo9GbWOlbWKTtKXWPmYqu87akvr2Lau+fcuut/qupqz69i273uq7mrLqu3bLqm/fsuutvqspq759y65mm+vVir+CV0q5SSnlppPfk5yQ5CtJ3p1k8j/ZnZHkXe33dyc5vf1veIcl+Y/2NborkpxQStmx/fHxE5Jc0Z77z1LKYe1/vzt93roW2gYAAAAAa8xq7oDaNck7azaULZO8ZRiGD5RSPpPk8lLKWUn+OclD2+vfl+SUJF9P8qMkj0mSYRiuK6W8OMln2uteNAzDde33c5JcnOTGSd7ffpLk5YtsAwAAAIA1ZsUB1DAM30hy9wWWfy/JsQssH5Kcu8i6Lkpy0QLLr05yl2m3AQAAAMDas9r/BQ8AAAAAliSAAgAAAKArARQAAAAAXQmgAAAAAOhKAAUAAABAVwIoAAAAALoSQAEAAADQlQAKAAAAgK4EUAAAAAB0JYACAAAAoCsBFAAAAABdCaAAAAAA6EoABQAAAEBXAigAAAAAuhJAAQAAANCVAAoAAACArgRQAAAAAHQlgAIAAACgKwEUAAAAAF0JoAAAAADoSgAFAAAAQFcCKAAAAAC6EkABAAAA0JUACgAAAICuBFAAAAAAdCWAAgAAAKArARQAAAAAXQmgAAAAAOhKAAUAAABAVwIoAAAAALoSQAEAAADQlQAKAAAAgK4EUAAAAAB0JYACAAAAoCsBFAAAAABdCaAAAAAA6EoABQAAAEBXAigAAAAAuhJAAQAAANCVAAoAAACArgRQAAAAAHQlgAIAAACgKwEUAAAAAF0JoAAAAADoSgAFAAAAQFcCKAAAAAC6EkABAAAA0JUACgAAAICuBFAAAAAAdCWAAgAAAKArARQAAAAAXQmgAAAAAOhKAAUAAABAVwIoAAAAALoSQAEAAADQlQAKAAAAgK4EUAAAAAB0JYACAAAAoCsBFAAAAABdCaAAAAAA6EoABQAAAEBXAigAAAAAuhJAAQAAANCVAAoAAACArgRQAAAAAHQlgAIAAACgKwEUAAAAAF0JoAAAAADoSgAFAAAAQFcCKAAAAAC6EkABAAAA0JUACgAAAICuBFAAAAAAdCWAAgAAAKArARQAAAAAXQmgAAAAAOhKAAUAAABAVwIoAAAAALoSQAEAAADQlQAKAAAAgK4EUAAAAAB0JYACAAAAoCsBFAAAAABdCaAAAAAA6EoABQAAAEBXAigAAAAAuhJAAQAAANCVAAoAAACArgRQAAAAAHQlgAIAAACgKwEUAAAAAF0JoAAAAADoSgAFAAAAQFcCKAAAAAC6EkABAAAA0JUACgAAAICuBFAAAAAAdCWAAgAAAKArARQAAAAAXQmgAAAAAOhKAAUAAABAVwIoAAAAALoSQAEAAADQlQAKAAAAgK4EUAAAAAB0JYACAAAAoCsBFAAAAABdCaAAAAAA6EoABQAAAEBXAigAAAAAuhJAAQAAANCVAAoAAACArgRQAAAAAHQlgAIAAACgKwEUAAAAAF0JoAAAAADoSgAFAAAAQFcCKAAAAAC6EkABAAAA0JUACgAAAICuBFAAAAAAdLXl5q4AALM56+LPLPn8kTf90UaveeOZB/euEgAAwJLcAQUAAABAVwIoAAAAALoSQAEAAADQlQAKAAAAgK4EUAAAAAB0JYACAAAAoKstN3cFANayH//3T/KFb/5gpjL73/pmnWoDAACwPrkDCgAAAICuBFAAAAAAdCWAAgAAAKArARQAAAAAXfkj5GwWZ138maled+RNf5SzLv5M3njmwZ1rxFqxVNuYtIcxbQMAAGDtcwcUAAAAAF0JoAAAAADoSgAFAAAAQFcCKAAAAAC6EkABAAAA0JX/BQ/WqFn/p8DE/wgHAADA2uQOKAAAAAC6EkABAAAA0JUACgAAAICuBFAAAAAAdCWAAgAAAKArARQAAAAAXW25uSsA/Bx6y2mLP7fNiclbLlj06ad85weLPvetbR6Sp3znVfO2dbNZawcAAMANbF3fAVVKOamU8vellK+XUp69uesDAAAAwPWt2wCqlLIhyWuTnJxkvyQPL6Xst3lrBQAAAMB86/kreIck+fowDN9IklLKZUlOTfK3m7VW/ML4wjdHXxV7xYlJkh/v9ZB84RUvW/D1+996ma+Kzftq2lJfRRvb6GtpM34dbZctjrv+V9qWMl7/Ml+lg593Z138mZnLPPo2m74eAACwHqznAGr3JN8cPb4myaGbqS43rOu+sfKJ/xShwUbBSrNUsJIsE64ssM2Zw5UZg5Uf//dPFnwfU5Xda+Vll7LcOnttd3NauC1t2vc57bom2102COxhqb+JtZgljtXljp/r/a2saY+fVYSKP97iuCX7iInzd31JkuSNZx68ou1sMpugH522Hxv7wfanrWy7qwl8N0dZ9V27ZdW3b9n1Vt/VlFXftVtWffuWXW/1XU3Zn+f6PuJts6+fVSvDMGzuOqxIKeUhSU4ahuFx7fGjkxw6DMOTR685O8nZ7eG+Sf7+Bq9oHzdP8u83cNnNsc3VlF1v9V1NWfXtW3a91Xc1ZdV37ZZV375l11t9V1NWffuWXW/1XU1Z9V27ZdW3b9n1Vt/VlFXfvmVXs821Zq9hGG6x7KuGYViXP0kOT3LF6PFzkjxnc9frBnrvV9/QZTfHNn+R6vuL9F7Vd+2WVd+1W1Z91fcX9b2q79otq75rt6z6qu8v6nv9Rarvev1Zt3+EPMlnkuxTSrltKWXrJA9L8u7NXCcAAAAA5lm3fwNqGIb/W0p5cpIrkmxIctEwDH+zmasFAAAAwDzrNoBKkmEY3pfkfZu7HpvBGzZD2c2xzdWUXW/1XU1Z9e1bdr3VdzVl1XftllXfvmXXW31XU1Z9+5Zdb/VdTVn1Xbtl1bdv2fVW39WUVd++ZVezzXVp3f4RcgAAAADWh/X8N6AAAAAAWAcEUAAA60wpxRguSSllq/av/QHwc6yUUjZ3HZay1uu3VjhZr0OllCNKKU8opey4wvIOjimVUnba3HVYT0op26ygzIra47hcKWX7lazjhraSCVIp5c6llEf0qM8C2ypLPZ5xXRs21bp62VR1KqWs+O8pzt9P68WmOG6pVrJPSimnJnla+1+Ab1Cb8jNc7bpKKXdIckkpZZdhGH4qhPrFNm9ccKOVlt0cNtX5d5pym/u9rhWbYz/Y9ytXSrlLkpNLKTfejFhaHTkAACAASURBVHVYcLxXSjm8lLLXMAyDz3h5TtTrTCnllCTnJ9kqyc4zlj2hlHL8sIo//LWJB56bvP3NG3xsMfr9lqWUQ2Zc142TPKuUsudC659xXQeVUu62krK9Td7TJpgI7JPk0lLKibOUm7THUsph09ajlFJG5c5OctbkKvhK9Z64lFJ2SfKkFYQVhyS5TynltE1QhyX37WifbjV5vMLQ7OZJDm6/P7aUcqfV9DtTbvOWpZSDZ3j9uA1tsdK7KNp7vXDWyU4re4skL94EbXdDKWXb9vsum3rwM69fvUmyUVuZdVs3W+m2V2oTBo2HlVLuuCnW1dZXktn3ZSll9yS/l+RTSXbYVPVZZFsL1WmL9tzhrQ2vaL2llHsn2baUcv9S/1fjldTtf5J8O8lvl1J2XiqEWiuh6eYIyUopW5aOYfe8PmIypljV+5x1kjmvT39ykhdOOZ44OJk7DpfbRvt3k36G8+r+1FLKjWY9Z47P20ttZ/Lriiubn10Iv91q1rG5lVK2GO3zFfVjK9jm+HM+opSy3bTlpnzdhtHvy441Sylbl1Lu2X4/uZSy3zTbmdYmOn+P+61jkjw6yZGz9g/LbOOOpZRnTvG6HZKc2n4/sZRy0ujpByW5opSyZ+8QatL/rGA+sWYIoNaRUspBSX4/ybnDMLxmGIavz1D2Xkk+kOTPSiknrHD7447zppODfyUHWVvXT9vvx5RS9tkUJ4B20N+jlHLLyUC0HaBHJ/nPyeRpSv83yctaHc8brX+m91tK2T/J65L8aIHnJoOZnUsNKG5w7T2dkhoePbeUcv8VrupuSfZP8oBSygNmKVhKuVOSvyqlPGKafTw+gSc5OclFwzD8z4zb3LeU8pJSynNKDUh+OmP5yWd3QCnl7mX5gHGPJFck2bmUsmx43NZ73DAMb2rlji+lPHyFdbxFKWW7xQampZQ9Syl7tN9/JclrSimvTZIV3lWwIclTSinvSvKMJP85Y/lJvfafZl819007xssUd4WM2tCvJHl9avs/ZtZ2MAzDvyf5zSR7lFL2mqVs6v9Ee2GSW5ZSdpuxbJKfDUSOS3K/Usrzk5yXeoFilnVM2skdSikHl3lh2mhfPSnJ60opbyqlHFdKueksk6RSyhOTXFRKeVEp5Zxp6jXa9mNLKY+f4b3cuZRyaCnl5psi/Gx91O9llRO30fpK6+uOKaW8qpTysCQHTFn8B0nenOSxSd5fagC56nBhtO8OLKU8qJSy77zn79n6kZ+UemHmlSvdVvtMDkjy4SSvSPK3s9Z1qP4xySeS7JTk5WWJEKrt7xNKKS8opTygTHGH8+hzumcp5fGt/Kx31pxcSnlZKeV5pV4hv0Hv1Cql3DX1f1m6rJRy9KSv35TaPjqplPKc1FB9y/Y+Vxr63S/J+aWUqQPrUV9xWuoFkNcvd+yP+rTdp91GqRfZXtr6sd02xbE3qvuFSY7PjP9LeanBwaRvvutS2ymlHJfk9e0YmPm8U0o5PLUvnCo8mXHd4yBznzY+e2KpFzhnWc/keD12seN8NAd5euq+6xrmt21OPuenpI47lm3fZeOgbK8yuig+73VbJXlYKWX31gaeNkXb3CHJM0opVyR5epJvTf9urrf98Xhzx2S6UHeZdd0ryaNKKXctpdxvGIbzk3w8yWlJji6bIIRqffHpqRcxnjC/DvNfnjrW+2KSlyf56OR1wzCcl+TPk7yjdLwTqn22Hyml3GoYhv+7KfqfzUEAtb7slOSyYRg+VVrqOU3jbo3zDklOSvLI1BPPSe25qdpAO7lNJqjPSPL21Nvej571IJs3qTgryUVJ/iB1wnrgtOtZwgOSfLyUsms7wfw0yTuT/FuSV5WNE+tFDcPwP8Mw/CDJ7ZLcu5Tyq2351O+31NDwvCTvWCgwbOs6NcmHkrynlPIHpZRbTrPuaY068v3bAHGnsvHdYXdKDQk+neQ/kvxmKeXMFWzq60m+3NZzSinlqFItORkupWw1DMNXUwddrymlPHK5fVxqsLhP6gl8q9TAY2rtPf9xkv9KctMkLypzVxSmOiZaHe/f6vDAJL9blrj7axiGzyX51yS/luS5iw2KWh02JLlze92xwzBcmjrJOrbM8HW8Vsf7JXlH6pWZc0oNRMfb2jE1aH1gKeWRSR6V5E+S3KmU8qG2nqkmTKMT8XeSvD/JEUneOQzDtZOT5CzrSfLsJH+81L4avdcLk3w/dVJ87zJFCFVqoHH/1ABplyRnLldmkW3/Y+oA5v2LDQ4XKfetJN9Mck6SP1zJsd/6uB+n9jNnJLl8GIb/nqVPHrXlP03yxNS2slFfXEp5YKvnb6fecXJ8kgfPcA55VJKHp7b/g5Mse6V1dJ44L8njkvzVvHVe7z2OJohXJPnVJJ8opRxfVvc1ybsn+cvU8+9Xp2lby6xv1yQPafV8U+oFkgtSj/dFJ1qT9zAMw/9Jnbj8cuq55SctFFrVYHc0OX17aqB7eerdpZNJ6ulJ/qGUsv0wDP+SelHlR61uW8xwXpy87tLUifa3knxhlnHNqG08M8nTkvxd6jH8qlLKLcZ91ugceJfU9rtzkhOSPK8sc+Gr7ZOjW133Sp14P7XUr/5N817vnBqwfTfJjqkXWvZZTTizwDYOKqUcUha4y7v1nW9LclWSj6T272eWTXRH9mjfHpDk1amfwaNTP88VTcJKKcckeVGSP2ljsGnrUNqx+bDU/unatnzB8UEp5dgkT0jy4GEYrp2yboenXgj+UpIDU/uzI6YpO8W6fznJTYZhuG+Sn5ZSHlpKuf0U5fZLcnFq//ivSd5U5v05hNGxdViSF6cec2cneWSZFzQvs607pJ4H3j0Mwxc3VRueaO3lvqWUP01tr7dPsnfqvOVWU9bx8NTA9fDUOc9rR31YSr14cvf2+9lJHpLkKcMw/EepF9e7fq251DnIY5McPQzDNaVeLLnNQufS1j+d134/Jsm7U8/P55Z69/XPtIuw1yW5Osm7kvz5MAw/WaouwzD8W2q7OSTJF4dh+H7b1sz5wGgccVWSN5dS3jY+NlewrvskeWOSf0py7ySnlnqx/PVJPpd6/lt1CNXGUB9LnSc+tZTygvbU9b5R0/qj7ybZvdXrv1tdJ88/O3Xs8c6V9n9LaWOHC1PPYX9VStm9nfvXXwg1DIOfNfyTpIx+Py3JJ5NsN//5JHdNssMS69k+yc1H6/lGklNGz2+52PZTg8q3pjb6+6ROLA9KclZq6HBse+0WM763h6Z2JNum3j3zvNSB2gEr3Fdbjn5/aZLPJ7nlpG6pg8fzUgf6x86w3hslOTJ1cvbMhT6bJcreNTU4eHOSHRf43O7Q9uf+SXZNHSi+okM7ul9qOPSHbXv3afv9zkm+l3rynbz2XkmuTLLPCrbz+0kOTR2AXtbe+92XeP3prS43bo+PTA2FTlvqWBgtu3fbxqlJtpqyjtu39vzY9viA1BD00CR7t2UblllHaZ/XR1Ingk9qx+bNpyh7bOqVk5cm2WmJ1+2Q5BFJ3pfk+LbszHYcPmzK93pAko8muUuSU1KDlhck2WXe605McknqZPiM0fIPJ/nAlNsa91XbtOPtrkn+oh1327fnbjbFum46+v3SJH+WZOcp6/FbSV7T9vPWy7z2aUl2Sw0q3pc6Gd56/v6Zof0/NzWA3WvGcnunTrjemmS3FWx3cifVpan98q1nLH/b1Dtkd0wNU788v20meVaS35gcH60tvn1y7E6xjcen9nOPSfLBtP46C/QzSe6Ydh5IndC+r/1+i9QLDL+3ULtrj/dNDZcPb48fl9rnLdoPTVn/9yT536PHSx7ni6xjt9Rz0Wmp/eMHk3ym7ZN/Tv16/bMW+vzaZzPZJw9o+/Oc1H7kSWnn//n7Y8b63bEda0e2x6e0dnXfybpT7+b9uyQ3SZ3I3mxUftsptjE59920/btzkpeknvv2m+yn5cpPtpcart+qPb596sWsN6T2xdun9QFJjkrtr49qjw9N7Yd/P0sc722dlyd5QHt899Rxy5OneK/3SJ2MPG607LzUCeKy/eCU+/Go1NDhgiRfTfKMea/bJ8kVo8eHpfaRz09y29XUYbTOQ1LPHS9O8tn2Hq9NHR/efgXre3rqnf5JcqOl2vW89jA5BnZKPe9cPnpuw/wy7Th6Zft9q+WOndS+5c1Jnt0eb5N6Pr10NZ/h6PFp7Th4eeo57H+l9g+3WWIdW7b2fu5o2XuSPDl1fLffvLbwyVFbPiL1eH5mkjtNWefDUs9TfzFtmRnb875J/jo14Ht35uYbFyQ5ZIr1HJx6h8wR7fGe7TN6aeq5/YB2vNyxfeZPSg0sj0kdB3wytQ/eebn2sIrP+fBWp+ek9n1fTT13H7ZA2aNS++Dfbvv8dm1/vCvJr6TN60av3z3JV5J8Le3cmo3nRguNo/dt7/+jSV4wWn7zGd/nvq393iP1/PDe8TE447pukXph/m7t8XGpAdElmTsfPTl17H5iphyHLLCdLdq/B7T9uXuSLyb53batm2TjPubeqXf+/VLqnPVPUscoJ6b2e3dtr3tO6hx0z8X2+wrru3eSs9vvL01yTZLd2+Op5kBr5WezV8DPMh9QvRoy+X2X1DT4ZxOr0cHzgiSPnld29yXW+7DUEOqA1AnHcxfpmLZq/26dOsD4QJJfHz3/yCT/kOTkKd7LEUnOa7/fKHXg+M3R8wekDopek2T/GffTQal3S+w0WvaK1GR/l9GyPVsn84Yk955h/VunXqV+6/j9L/C6MqrPAamT8F1SA52nZuOJ9W1SO+u/nNQxdZLxpSRP2oRtaL/UdP+W7XP/dvssn5Y6gLoyyVdHr9+QOoGbOoBKnZhs1fbPnqkDn39NnYwvup9TT6oXpJ5cJhOS30y9a+3IRco8MfVK68vbPjwldWJx/ywTOIw+y7uNHn8odcDym6lXBRcMQDPq3FMnPpNJ/+NTQ7DJyf7YtAlkRqFsNj6JHZnkd1IH7Dsv9JpRXU/PxiHU6a3dPGSZ93mrtp+uHi07MHXSe+QC29o/dfL52owmJqmTpT+foS2cm3pSfnXqMXP7to/PSe1nPpJkmyXK3yE1jDlktOyy1DBs0RBq3r5+buok8d6Z6ysX6t9+J7X/euto2ZNSB6AzBwyt/PNbe5hpcpfaV7wkdYJzqxnK3Tn1Lr4tUycHr0vytPbcIZlikpA6YXtekl9P/btCkyB2vP/umzohuPuo3AfTBlxTbOP0JP+e5CPz9vVz5h1bN069K2Gn1P5wi9SJ2HuTvCV1YPjZJL8/+VwzN0ndJckfpQZoDx+t84Wpk7KpLpJk4378/mkTwNRB/ydGr1vwws1i62x12y81HHl0O7aua+9v97a/P5vkKfPrmjp5/J12LHw17UJUW8+rUvuh1YYaT2/rfu5o2eNSzx83Gi27MPXOp39r235H6jH/sixxbI/Kn9zKPCvJQW3Zy9vne07queMOi30u7ff7pQbrn03ymNE+PjPJ36Se49+QZNfR/vufJK8drePg1MDqtZkb65Qkt2i/3zr1vP3J1P5kco46InVysdQFhG1S7xr/WpK3tGWT8drFaRfHVvl53SN1nHN0e7xv23dnzfu83pXWJ7THh7b3fMr8/bqCOpTUu42+mnqMnTvaxj+nHnszXcxqZS6dt+ygLBGst3ZzSWofelTqBPby+esZH7epfedfp4WSbdkTkjxxoTKpF0rfnhqKjIOdv05y51Xsw0MyF549L3UsO+mDL0tyv8X2ffv3ZpP2lTqefHdbzx+nHrt3aZ/T7dvzV47a+2GtPT4nCwTIo23cPXWMsEtqX/X7qeeLmS9ULrEfDk09zz829e6WPVPvMHtf2++/McU6jkry/yd5cXu8obXPi9vj49ox88jUPuec1k6uSrs7MLXfWPTYnvX4GP1+69Rx2U6tnf5x6gXfW6Setx8yeu04NDoldaz4ucln1PbVO1NDj0l/NbnJYI/Ued0XMxfE7Z3RfLItOyu1/35ce/5uqXOm81IvDJ+f6frzLdr7+lTqmGDX0XOfTPL0Fe67V7XP5Y/avnpLan//h5kLoSaf33EzrnuX1JBxMve6UZI/bb8fk3p+++i8Muem9ueTUOm2rQ19PrWv+3rq2Py3Rsfy/86MFwQXqe+GzJ0/th8tn4RQkzpNPXbc3D+bvQJ+lvhwaqL6ntRJ+uRge2HqSfb4zJ10HpE64L7dqOz9kvx9kueMlm2RehKanFCOSf07R9dmgZNn6uB/kuYe3jqoN6R+x3WP0evOSvKF1En5ogOZ1NR498wl2tu3zuqy0WsOST35znQHQpI7tY7uvLQ7jVI74S+lDvJ/dnUrdaJ3buvQjp9hG1unhgvvyBKTy9SO+/Opg/mvpHbqd0sdCFzWOq3D27oemzqxenjmBsrnJjlnFe1m28xNyG7bPve7pg6a/yY1APh46p1GT2rPf7zVb6/UkOJfsswdA6P9uV3m7mC6X+qJ8u/bv6enhqa7LlS2/f4bqYPhk0bv/4VZYHLXnvtQ6sThk0l+py1/SNvnp0xT53nt5lGjx7+e5I8WKDe54+O41FDlXW2/XZga6N2xve7o1EHCODxZcNKbOvA4P/UEtuW8ffKA1iZ+qbW7R6Ze/ZrcbfjwzLtLIHMTgW3b+zordRB3VVrw2173yiRPHT0+O+2uu9TB2x+nXlm6zeg1e07Z9s5q7Wvf1L7rorb8bqkTvcuzTLjc2t+rUkP1A9uyLdt+vjRLXOXJxiHUc1IH1sfP27ePavtzv9SrW3+V5DWj+v9Nkn1Xevy19fxWar+z7OBtXrk9UwOWyzNdmPrk1EHfK9rPVqN29c7UW8Sv99ll7ti9SfvZkBpKfz41nNyi/fvN1InGIe11L0oNGe6fetfh59IGv4vU72Gpg7Dbp97988rUweSuqcfTF7PwuWdD6hXqC9rntHPqhYNJyHu/1AsVW6VOwk9KvXvgvNS2f37qAH/Sfo5OPW9OHSq293d1e78fydxk/Z1JvjDj5zrZ3yX1GLso9Q7H56b+fbS3tucPTf0a9CtGZcdt+g9Sb///tWwc2j0itZ89MysIE1In95M7Qp/c9u2DW333Tz2WT039iufkvTwvdfywf+rxflSmCF1Tz3tfadv8UOrk8iHtuSelfsVtwQn3aB1HJfl4+/2E1AtjD2yPH5ka1O3W2s3eSc5qz90x9Wu64zuZD8noeG/v57zU8/clqWOWR7R98rjRej6eRQLx1LHb77Xf904NZ349dUxyj9Tz64on7pmbiFzS3s9Jo+ful3oh6LmZm5DeP/XYe/DodY9p7XrZfmaJ9rxnatB2YqvHde3fA1LP0de0tvPsLHOHQuqY6ORW9iap/dpL23P3TJ3IHbLIfnhE+zxulXpBYRI+7JQaFl04KvPY1DDx3NTz+Vmp4emTUo+fqzN3J97kfd45dfK7Y+oFklentv8DU4PNv80MFxyy8fnootRzzsXZ+C6mbVLPIx/JAv1Wa4NPzMaT/Q3ZOEw7J/X42jdzfcytU/uKceh6jyR3WaK+J7X3+OzUiflt22fyitRzwvXC4hW0pa3a5/XF1En8JCTcKfWizj+kntNL6nhij3nrOT5zffSJqUHB5Li/Z/tcb5l6geNLrZ3eqT2/R+YC/VNSA8VVB8Tz6veM1hY/mTrmHV8sf0DqnW63H+2LU9tnfKfUPvnY1AuEL85cWHmP9vnepr3+A6nn1yenhlqPb/vhKanjwDuNtnlW6oWySX80uWh1x9bmPpUZQ9XU8eZftfczueP9KZliPjNqB3dpn2Vpn9fzWhs7M/Ui8SWtXZ8/+rwfnxnuHG/79NOpF/teldqPbZl6vD2svYeXpd6kMbnr+4jU8c4ksDog9TjYP/V89snUfuLg1PPP5Jz24qzyTtO2T16T2kf8Uq7/DYaXpYZfp6XOz1cdeN0QP5u9An4W+WDmOsGHZ+4K3aTTeVHqxPcLqZODr2R0FTr1BPMP7WB9U9rtwu25LUYH+rGpV6T3W6QOd009wVya5Ott2VapJ+vXjRt5RonsIuuaDBS2bHV7fXu8Q2qn+ubRa2+01LqWWPcdUwOt81K/FjW51fzgBcrcJvUOoLvMW35iRoOABcptnaWveu6S2tHfMnWg86n2+15tX38nNUj41mTbqVcfLmkdzKOS/GOSE1fRdk5Mnbg9qLWN3VM78ye2z/P5SX6SOqAYD7yvSv3jtr+b5B5TbuvU9n4+mXrSOT01yHpye363LHL7eDaeWD0rdVD0qdQrYD9rL/PKPD81YPmV1JP51pkL2+6TGb/6tECdTk+7JX+B5+6QOmi5Jm0gnDp5eXXqIOAJqQPJB6UOcF+5yHsdDz6PSguvRsuemHoCPDt1gndEalj7iLafj16kftumnjyvTh2s79GWPyj1itFFqcf832XuivmTUq8o7TNaz11Tj/nzZt2fre63buu9IrW/2CpzofBUx3bqSfblqSHU3VMH0G9Mcs+F9uMS7er5GR3jrY3+bdsfb0w9Ye+c2m+8OXXgtWB/uNA2s8QdNVnmqyeLlU3tm5Yd0KcO9j6R2tddlHo+uLTt793b84uup7WLK1ID9QNTB08fTR0IX556YeKfUs83325tZ7/Wzt+f2m8v9fXa09q6LkwNbR6YOhF4Ydvf75n32czft7dNDZFemXaXTFv+9NRJyuTiyHHtc/tmklPbstulXix5X+o58nNpXzuZsv3t1N7fdqP3cavR83+R9hW/KddX2s/hbZ9+r9Xtpq0tfqMt/7u23xf6KsYTU+9sOKe9tydk7tb77VMDo12nrdO8dR+T2uecMdrHf9nawadSz01fbMvelTYxafv2mix//h9PMp+fuUnG51Mnte/J3IB9wzLH931SQ7qHt8c7pl6A+GbqefRfMgoQ2nv7XJIz27I7pd4l9NxF1r9be9//nnYnTOp55vGp7f4jqefK+y5S/sjUycCRo2V3SJ30fjq1PS9Ydpp21P69+WjZa1LvaJlckPyV1PP4WaPX7JgauLxq9J4OSD32bzJjHSbn5vunhj5/mnoH31Gp7fva1PPkZe01B2d0x/Ei6zwstf1f1Pbx76T2yx9PDX6uTnKf0euPy9wV/w2pwcS9Uu8G/GDmgpUdUvvHyXFydur45KjW9p7RPu9T2mf+uiw8JvzD1HP7m1q97tra2pdSj+MTl2qzS7zvs1ODwhuntuvXJXl+e+4hqZPJSRCzYV7ZV7Z2dnrm3fWQuYvNh7f9t23quPLS9vze7T29KUtf0Cmp55KPpvapJ6T2UZNg89DU8c/MX7Oc155PTvKS9vsu7bN5xagOO6aG7uekhk9fSruzp73mlNSLnvceLTs2NRB9b0Z3kaWOJV6Z2pc9NXMXfm+UGsp+ZX4bWO1P6t3D72+/v77VafLeT211Gd+Vf6PUNv7F1H7ttm358akB/Qszd7zvkBqCXN3a5qWtTd5k1I7envYtlbY/t029sWHv1MD+yowugqb2dzsu854mrz0idby3f3v84NZefrut+xsZBeTLrPOE9jl+rr2P26WGaR9L/ZroT1LDlptk7o7HqdY92sbNW/t5ROp56OLM3UF2auq3L349c3cMHtWeOyi1/3xm+wy+lDp+Oji1Xx/3t7+W5HWbqO3cIvU4f2xqGHdhW//t573u7a3uD9qUbbfnz2avgJ95H0ht9LdOnXg+tS27Wepg+AGj122ZesI6PKNJYuqV3ru1g2KX1MHQ5dk4hJp8XeoxWeDrGdl4gvyc1D9wO75jYpvWOVyaeVchllnX5ES6Y+pE+vz2eIfUQcMb55eZYb+NQ6h3pJ5Yv5165Waxid624/KZu0tpqUnV/K9FjN9faR3ja1OvFH8ideBZWqf54tS/e/Xj1Mnu+GrEI1OvBv9hRieKVbSjK1PvcDopdXBzeGpY8p32Hr+VOuA7cl65i5N8cqH3t8A27p46aTmwtaU3pHbqd5yhnuOwYO/Ur6BsaPts/oCrtG18JcnbR8ufmOT0TXDsHZx64rveCW3Uvt6QOiCYTJa2bu33KakTlHu3Y+uA1DuJnrvIey0LLUsNH/6srfOxqZOdybY3pAZMS30N4V6pE6e/TLs6nzo5vX9rjx9MG6SlTvDfnBoqTK6WXZHaZ+zf2uKig5CF2kZqf/GdJO8ZLXtC6glz6q8rtXK/lBpAvb+11Y0G+Zn3daNsfCxuscCy3VP/xsPkbrWHpl4tfODoNTddpC7j2+Fvn42/frHQ1enx3xvZMKrzXlnmzod5bWKLeeuZTKx2SR2s7ZI6+Lsy9Ur8h1OvHi84qRzVY9vUIOGE1CuL17XPfJ/Uuyc+ltof7dfa5S+nDoYnVxy3yRJ/76fV63FpAVjq5PfC1ABwEsZsPb9e7fcTWzu+W+qx9ILUUPyo1OPt4tSB4+Qz3ip1EnxJaj86uUK5e+qE7uK0YGqGtnez1jZennrc3K4tPylL3PG1xPo2pE7Wvpga2Dw7tX/+eOo54vS2/GPtM5kfxt0/dcA7mXTfN7V/OSP1vHJBVnYny+TrlVuljh2ubOuc9GkXt33/gdFrfzf13D8JeX47i0xA2+d8xKjOp6Sej3ZMDfEmfdQkqLzeeGL+vmjLPpnRV4vbstuknpMmwfvP7m5t7+0vM3eX151TQ5q9s3BfcV5777+XURjY9s352fjrbJNjahKc/Vbmxm4bMjfu2TN1Yv3S0b5ZyVjn5NTz1KuT3Kstuyj1LvjfTfJ/0s6Pbft7ZS6AeWBq2Pfh1IsUD51hu+Ov9B2ReuFi99Rg5kepk/zJ30a5JvUrKcv+DZnUifab0i56pU4QJ3/DaEN7PP9ul8e0z3vylchHpR5bHxq95pkZ/T2s1DHzi1IDpzNS+7cNqWO/n339ct527t7eyxGpk9OXZeM7oc5PnRRO9fcJ5637l1InjJO7uG+aGpq8NnN3cG3UtuaVf3zqcXBh6hho/terjkm92PvG0bIvpP5h96RO7i/K9QO3rTN3N9D2qWPap7dt/K/M3YX6wKLovwAAIABJREFUwPbcdrO+93nbO761oXuMlu2c2u+eP1r2qNQw43PZOIzcIbUfPXrUNh/Z3t9RqcHMc9tz56bOh57fPrcPZ/RnEFLPPZvk76LNe4/3Th27PT/twlxbfqcs8jcnU89//5I6xp787b8t2+d6fupXr7Zs7feU1DD1pGz8FfrJHVXbzG/f7TO9OskHR8t+NbP1CfdN7Xeelxo6PTP1XHKf9jm9PqOLhsus686px/3kXPv61HPwu1PHqlel3pSxddr4IXWMPNOdaqnH7btGjz+fOg4+v+3HPdr6J+eFfTP3FfwXpAb+90ztn76cOl5/bOo5ePJNpdNTj61tsop53Gi/vHdeW3pF6hjiVqP39OXMXYDbJH9vqvfPZq+An3kfyNxgZfK/kk3utPjT1E72TamTuT2yQLCSesV43GlvmTqYvzzt63itsS57YLSD7IjUK+UXtoNqcrXgFq2DWPTgn9fZnZF62+nD2uPt2/v7g9HjJcOshdY7b/mkw3hY6l1a7550fvPLZG6wOBlU/7/2zjvcqupa+79FkaaAGHsBsXexISIiIEVAxYKKiIi994oaO4oFe0NFjRVjQI2JBbFi7NhbjDExUZOb5CYm+W5ucmPm98c7pmvuddbee+1zzuacA3M8z3rO2XuvMtcsY47xjraq9e287P1yrukBnJj5bS3SxK13ISHQC33bG5P7A1I6JiJGPR0YYud0Roj+7UjZK5tQvmAf7Y+sZ4/ZXAlDcu5GSs7hyD13U2Nql9i1C6mS8wcJf7MoZYyj7D29Z8J33nY51zdICJr5PUy0P5zUCtEXCSbeLXYKCm9otAu43WcwWl+75s01JLR6T6uNkPB+uH3elHQj8HNwCFIQFpIJgw37gBS09IplJ2QRvRUpfV5QOZky6yNooz93HWTZmxuMRV/kTXmF/XYUEsYuQYDVbLS5Hgu8YNeUDZmgdG3vbuPgk/7ehRTUdgiEeJ9GJitFa21VGgrJY5AwfTFBQvZMu8L8dWchheOrYNx6IhDqAVLPj7y5uBJyV++O1s2vbWxCwa199n9re+iZOtbmzXwEBDbwtAquXSac0zZXRto9TrNjacQ3ZpEKmreg9V0p/99gG//LM/zij0gBao8EqheQwOzn/Z5IGasWHnUSspL/HAvBtO+PRMaBsXn9bOcchXjIFcjqd4i15xwk4G1C6RraCwFm19hxJmkOmm6Ir92IeG1Zz7bMfPM5Ns6g1FtwR/tcOJk5Ah18ePIYpHD6UJh1kCfUa/b5QMqsE6T8TbX/vYww2sbnCQrmTCQ1SnVGQv2bpIBNR+uvN5AS0wkBMa+hvSs0gF2O9peKxgYbu2lozn9Cut+tgPj2IAToPkOO1wGl63kTgvx8SPl+Iu98e4/5yAg1KRi/+aQeQN1yrtscgZvd7f2/b/OuL+IDeyKvlVmIn4Zz0XscHGfzrUvw2wAUdtoXAfQnVOq3cv1gbXgIgVBnoz3Ge3ZcjfjLnchDYku0Nh5AvHJacL+tSIGEIsVUVrC+2NY+74rklOPQen3cnn038h7cEYWKnlXp/nbfY9G62Me3B4FSN+ScH/b3ynbd2tYvj6A9rQ+SAReSgqReFjsTreEwKftRyMuxAdiDPLNu8s9Ga/anSFbsbn18K1LcGwMAj0cGSc9jvCfU6UF7srKr/957Gg+z+Xgm4oErIWDjOSTT/AcDtOy6N1BlWsjs8UhfGIxk/smId66K9qv/It1Tt0K8o1D+vyr84R4kP3hDx7U2Hj0Q6L6enbsT4kPDcu4xw657EM35J4DTg3X/BQKTFyDwcAFaI6vbvLmCKl6cta7VzFgNsv56mFRGPAoB/l3zrg3Wxz4253yKij723brB/ddDANB7pLLkbohXhPmCBpDK0tsimcjrZPsgELeQARkB23OQfjEa7dczMWMjmsfzEK+qVgymG5rzvyHNddoZra1fIP13LpbmBYE/NRWrsr70wOonCAj8COlcAxCg+134Glrvx6N19CObW37vPQqBph+gdfFjtDd8iHTlz2hCTrictj8HHBp8HoLWpjfurEuaS7FRho2WOFq8AfEIBkMK3K+wEC+0MS9EDPVxFKt7EmkSxAaWB6Qg3Zf5rjMSOm83hvMeVSxTCGCaYZO8ozGRe9EG8X1kSSq04dpifQkpUH/FEp4jJetj4LKC98ndkDPnDETC1jgkFHwOjMhc75W8nkhg8iDfYcgqu3/AaPKumY82lHBzecbesR8SiG63MToKMXUfqvYFEt5WQ5aXaUi5+gnaGI+yPs/1xCjQNxtZG7ygcIe1d0skjD9rfe/Bj6ORovkmpYp8xZw/1taT7H6h8HYz+RXsrrD3PD/4LheEsnfwXkATEJN/BQn129q7vII2pZepIWTKvssDbruResZkhYDdbHwfIs0BsiWyxl2KvMkGB+fvgBT1cciV917ggvD5mfn0OBIifYjttWhd+Bh6733Sp8J77oY26B8hBaA7Ev4fQWDC3cibZR8kTC4kVTzHkirJw20u5uYuIrO5IVD6bVKhfDSa/3ehzfwp8nP8hDlxvECWNy4NgGOkyP0ACclHoPVyWIX+fQ4JQj3QGv+YFJhe1u5TCUgfZP14vI2lzys0j1KLewdKwaeXSQWEdWxs+tnny+xeoXAYtvlZ/5zgd29V/E64sXF+AoUTH2HvmmdN9f3dDwlfD2IhV8HYHwj8y+7VyfpqBvJq9PxkV8y6WqavBqJ1sqo9awEWWmG/H0yZJJkof8vPSMG0ddDa2gUpVacjITIJ+ulhBOgfgBThO0nDeX+LhOTeKFSsgecS2ue8V9dotD++goCIjdD6fgoBnR8TWN6LHNaWzYM2/xABED5pdj+0Jz5Ahf0U7b9PUBoyvQfij4VD1pHh415Shfc663MPoHeztiy0sR6OQIaLsaT+wb0uKjeW4Xq2cXwfraGlSPfWQxBvf5cqYQMIgH8O7WXf7Y9oDf4sc+4Au+9WSFZ5kNQTarjNyTxPq+Foj7wLgW59bI6dg/jb77FwfhTqFebeWRMBrusir5mHEQi2on1+g5RvrE2Qr7OGsRuAKej2eRm0f1+L5XZCPPgN0oIaVyH5bWWkyI2v9bnBOrkVyX2bIG/q2xEYPw7Jrh+hvcADYkOpEMKNDHPvBuvkG1Lvh91snLqTI/chHrU82uO+QPxmKyQjzEMK4SZof9kBM5ShfXsuVvWXNH/qepQWiBmE5MCDkYIZetxMQzK5N+gOpsbQV4J9FK3hv5Aq3uU8Q1fKXN+X1Fh0APA3BAh0QrrC58ire4z1USiDfEAZ0NrGdh4CA/a271ZBuslVaN2/TY1epdl3IuU5+yEZ9BkEEp1OGioY9kVPoH/w/puitbW0/T2B1CNwD8RvvOfPDoj37YH2uacolXlm0wjP1irveSLyZjsXyQZTkWw1AYE07xLIRkG/7IIAxdtJPYiPsDafhXjLukhmPJYUuJ2O9uu9EUDxHoGxCPGKz6wNPn/eBHvWMzYGFQHFoI1dSQHZzUm9IQ9EfOacoN2PUMBLDhnxL0D7zEQExnZC+9N/SEHk/Wz+leUtOfdeD62PK4LvhmKpYPz6QmCoB6BGWp93QPvI8/bO3ZHMvgDJB08g+Xo2kq8HUcEAWLC9fRC/904NeyMdKuRDpyI+16hiOa3haPEGxCMzIBL6PibNmXIM8HdSxpqXA2CYnTcQCUIvkSMYGnP6kirx+MH5WyMPiWuM4eyMNoiXKGBxRZvEWrZIPLjyMhLyvdvx0lRQrHPuOdIW+wTywwePCO7dwRbub0hdUkMF8Vkk2H6ANvNO1sarjJH4c70g3dMY0qDgeUOR4rOXMaUbjQF1sXsfSGr1TZAy5ON5V0ZK3hxKq180yvvJ5s7byEV8DqnA+zjydPod2ijOpDSJfC9Sq0luPoBg3m2LBEevAB+PXFLPt/f+nExuFCQYPIkE1RKlPby3/b8XEvCuRAr7HD/fSavGrY82gq4UqPqE1sUogkqNlAe/soDHYKSMfA9Zwj8GTrPf1rP5MihzzQRSb4Wl7fnPYJuyfe+9C5+2/vsIS0KK5u2jaJ4/gBSKst4BpCGtOyCB51Vk2euKBJRXMM8uG+vHrD+WR+vlZuQNdxiay5VCUEOvgZ1tPL0XobfAhkpqJS+qXdGGPzvbh9lxIgVA+iNL1TXB++xj7+09XwYgftgTCQnPAvODex5p/T007Mcq834fG595GIhk3z8JvJy5pgdSlD3ItT4CzN7EBGj7fm7wHu2Ca5/O6w97n9lIQd6HNGxgW/vuobBtwfzzfTgUCZ/9g364CilanW2uvOHHH/Gwo9FaPIgKIZSIt62LeM1Pg7ZtiICDq3KuyXqY9kQATVi1dBKmOFGadLu/tT9cU+Otjy9BXnnhem/A10gB+Jk2p55FCupJKCRsG5tfO9u9vfdHYeuiPWMVtP67IsX6OWQpnYKUqPn2uUGuwuA+3RHvuwQBxhOtjYUSWVPK4y5Da97Pg+mIR/RHoMUcxNtOxzy+bH6cYtdVzblBum62Rny/N1J0biLNZ9Lexrx3pX5FCsezQVu/RjzTW7MfJTCYIJ4Whh1Ntvc4wD43ML4hsOJKUsD4DCQX9LHPA6mSFxF5VLyNQKddEf/+KVKgfIL0msKQg3vvgGS3q5Ey5kPAu9p8vYnUYHEbWusTMve4qMjY5TzbA4ZjkPzzPFJkz0Cy1a+sj99C/OUiqiicaG3NJ8h1iYC+/0OA2r2UASVtPj0XzJupyLPVf+5J6T7VGe0zyyN+OAEBSPPRvrmx9eMLyIthfbQ/3IMUzBfQnnowaUjXiVjhilrWXg5v8OtkbxvXXB6A9oU/YPtF8P1JpIVPZtqcO9z6KFSu10J6xPcL8oiZCDg4hFR2Xg7xy4PJ6CM1zCX/vmMR/1oVyVaH2bN2RnzoNSQfl0tVEILFr1JagGkHBL6Mzjx7sM3bF4PvTkR8rmw4ea3vZv9vhHSd3RFwOwfJrAcgkOU6cjyN0Bp73a5/mKBAFJKNf2p9NwAZQE9H/PBkJDcejPbQO0mB4AQBkveQOjc8DDwXPLcPxXM+jUb7kN/jdyEN69wOySFhioKqnmVIhjmFNIXKN2g/Phqt2bPRmrzF5kbhHF12z+cQf3+GUhnsLlL5YhPEw3xi+i2tz8+iNK/cDkhXHEy6LyUojcE5NMITMtPedZDx7Bq0b51u73AyAqH8PtYf8cmKxR1a89HiDYhHzqCICX9GCkIdiyV5DM5JMufPQhvG07YQziZVOtuTVuqohnBPwUpI2uctbOFOJxVwygoX5GxIaOMaEizWgahM6pEF+8MrZ/2MkZyHNshppG6H3lKwD7IcLIsEme8hUChkiN7D4ABb4CGg1AUpNjOB/TLfP0/qvurzmMwgBSQ6IZDkKSoAdPYeH5Em6vYKfK05crqRCof9kKC0PFIW/2lzYU8kSM9HysN068OT7Ly3KZCnwZ4xFnninGHv6K2uRyHF70Esf0kwZlciRXvN4D5PklYICufxFLv/evY+JyCrsw8VWNP69zqquN+SbpbbIWZ+NRLSbso5xyvo3ZEg/V0SXOT2vS3aZF9Fit9ryCNglew72OcxSDhYM7jPg9Zn/l3OsTE6CwkoG5N62hxn12yBgKKy3mgIyNqN0jV7JVJ6+vj3yqwjDzQ9igSJoxCAcxBl3IZJwdM/kwIUxyCr0h5+HtvcuIvUmlsufGBD68/d7H1/mb0mGIdeSLDeHSk716FQVi8E90BAxa0o5GkNBIa8Yn06AoHgYT6QE9Gc7ZJtY867j0DrZhcbx6Mpnc/PkSrzyyAFZfvMteMQmHg8qYfPnpSGxXS3PskDnw6zfuqAwJDrMa+v4H07Zq7pjgQ2LygOQEqOT3LbBSkrtyDB9f6gbZ2Cc05AoEURQXJnJEBOCJ67CbISLp/X12i9e2v0TEpDG0+2cQ2Vte2QF9dPkLA+hFQ4nIR4q987K7qj29w60Z5xb/C9z3dSU2nnMs/ohBSCuTZ+x6LQmL8iK/hktDdXtJoihewIpITcT0FDUuYewxCI+g4CZz3gcj5SXL5A/DI0gL2NgLLOyBJ8NVUq3tq1u6K9x6/TLkhOuREpm78k31CW5RnrIwPRkQg8XxUpmHPQ/p49vz8CfMNwvR+ites9ELOenHOQwrh9MM9OJ0hontc+tCeFnilnWttWCn7vm/deNYzZ+kjR8PlFxtjcGW+fu5HjDUBpzqYtba4VTpxv13VAXof7IX55BNqTH0Ey5WZoL/na5sqKFFAO0Zr9FxlABK3Ff5B6JXakFHxYG+3jPn+o3yNORwmKw7DlAxHfuggB49tnntWTwNiH9pdXkZLq81GtZfd+DPH52fbOwxBPW6bSuFKaSiA332ow34ZRRga0eb4A7X9XB99PQzz9RPs8GslOW1ublw3ufykCsQ4vt+ZsTHuhtbo5kgtORjxsOZohrAjtye+Tyu47kcpoP0cyxi6ZdoUy2llI7va863QsDQOSQW4mXSvhWl3arrsCgXYHoDnd5ITjmecMQbLUxGAOz0Q8pmv4Pjlr7TzEa8ch+X2a9YdPpdAZ7afXkXqn9UZgXFjp2D8nQWtyLpIJtw7O+RHwSY3vOdzGbkDwXU8ERD+A9FYfKtggt15ef9nnrWwMr0N72+Von7oWyZQ+R+9m1FDhDe2Zb5B6HF9PAMxbX76E1vLP0Z61Pwrh3RB5eIZe7lPQ/tsdrZUX7R5jEW8oVC260jxCMrn3IOtjbTsN6QiTbF782Nq2V1Oe19JHizcgHmUGpiEIdZpN9oplvZHAc59N0GdIc0eNJKdMZQ4jWBPlAgnz1kxBm/50Crr72YI8ktRKOQxTLJAA9UOqJzAPXdw3QMKstyRuhVxbpyHF7D20AfW09/01EuwHIWXFV5zrgDa6wcgKF8b3hwlZD6PUPXZ1chKtoo3mFlL30A6kyUDLejIhcOGXxmAak4i0O7JijDemtard0+fwONbG8c+kOW/2R5v3pUgRuLUoA0OC3wsodPAQmw/PkubXONbeeS/SjWcNJLh+6c+z7weSCbu0OfY5Uu73QABPb7QhzSYFU9ZCAmVVd2kb+xmkwEZXG5vQFT0Me3qRNBxzPWRJ7mTH7aTKy23IytObVEAaipS4SdbG4xHAsyHaNJ+ye7azYzJprPgMu0cnJDzOxfJxlXkv/8z+aJ7/ECkXBwXn+KpVXWjoadIZCajeGjYB8YqqoTykFjdvJToeCbrDg3c4iUxIAqUC2mZ2TZiDaLzNEx+C0R4BFt5TcQLa+IfZ70cigcEDPT3RBr2xXfsICmvw1q2d0To9KXhmEe+5DVFYi3/OMASGHUkmjAatw81IBWt/7aDMtbOR+/RHlLrIjyRTZty+PwUpx2E+qd1tjOcg78ZKYW1rkXpM7IDy0/mQpC6IJ2yK1vTemes92FiJlx2A9qddSXM/3IfWvq+WFIZS7EQKWnrw8XbSsJgHEGB+PVJuQ+NBfySYeoH8QsQjBpOCUFXd3zPzsTtpPo6wTP2xiF+UrXpa9EDr4mablx2RUjAm6N/C+UeQktWYfDO9kcLl1+6FCBjz4OnS1q6s7HEc2lO2sPeo2h+I979OCsJvTOpxeAZSyBpUJcyMy2RrS3ubV7NIC3RciHiIBzkH2xgejACba9Ea2wPJCvNtjt2Ved6OyPtkDAI0z6DUA28qgeEvc+26aE88l1IQ6k60JpuUlzC4354IfLiatPLVKAQ87Fvl2gQpxO9SJXdb5ro1Ee/1hT8+xQrZIG+rX6A1vrb172tovfapct9ewTvsjHjggZlzTkRFeNbJfO/zjx6NZLpsvsbjSQ2RKyGg+lQEoP+F1HhxH2XCxxBv+oY0aXVHZHi4zOag78+PqG7MbW9zYyoCcmaVOa+k4ArlQahD0Bp8Acni3dB6HWC/j0C8fBri69cgGXEQaT6gw9CelFcQZRTyJLsLyYZdkdHiervmawLvkSbM511sXIYgmfFrtCcOtPZ9jryE8tJfvIjkgSxYfCoCi/uQGjPygI+VkdHlJwikblIOq5z7T0HGhDetv31ah47IA2kupQbarP7VCcnyL5HmvnoHyfJL25w6xH6/hHR/XQMZAy/N3M/3xdr2vidQWvH4bmpIuo5kqMPRWvCe5wcjcPIwqnuJ+vFaKvv+SPb9GeLrGyCjw1hkUD2fRlR4ReDQ4ODzwYiXhTx+aSS/e0+z3RAQ3dXG80tkeL7AxsLrkp2QPDwP6aZVc0wWbPOBiOd6+Xw16xOf06w9WtN+H28T+Z5y37WlGxCPCoOjDfrjYCJWqkgVbignAnfY/ysjkKQBk8ks/mOQgHw0EvL/EEz4fZHwUzTh+EFI4JmDvIb6GYO8CSm6HxLksih3P7QReoGipzGnBcE5/ZAS/ydSwdRvUk/YbwuwuFlSr6WdjLGNRMpPtpzrIZXeEVkTR6KNYhUk2B6AhLZ1ENjwCoHHRZn71ZTnKef68WgjDZXY04BT7P9pyKXdl7PtgJTCS+28HuF7VXnWqkih3h5ZxPuQKo8HISbuE+J+HymRD9q1ByOwYHckgN4E/DC497bIurgryrfwW1Jlck20+dxPCk5U9BQLxukCpGwfEPy2JVJk2lMaWjmP0rxe66HN1gu9dyCgZ7DNqbAyks/Ns7/d53JkNZ6KNrKXEDCXXW+T7ZzPSRXq9ggcuocKIJu19WpSj7xJaP1ODs6pliC4nY1NxbLDNPQWOBUJRD7Bq4/zz82PgxRm7/m1HlLcHkKAUp9grG8hBfmWQgLTKzbnvGv6zcEYHY6EVx8n79fe3VgBAiSseo+fkciTwHseVvPgWBfxqwXI8hcmmH8IARS5HlQVrvU5km6jYYhAg/LzSLDznjMrIV58I1pvGyOhb4PMfTqRJkTubn33Jqbwozn8Rxv7YxAPOR55CdxNWm1wIgKAynpIIoH2OWvH84i3dEXKzI8Rj8oCoKOR8nwCmsPr2Xc3kO45gxCPzoJ8I5ByepJ97oj4wyws1LkG/jkE8W3vSXI8CkkMqyIWtrYWeF4nxPu+y99IuifVXYhEgO6zBCG2NhcXoj0x61UdglCno321CEjdB/GxB2x8L7c5/DopiNGgKlPmHochGaJv8N1JNlcvsPfw3npDELhwFAJCTkH73MnIyr8AeXMMt/73XsP9kSwyFwEMl9pcPo2MsS7bTiRXnYOArqetf7zX0wRr5w7V+qpKP26K5SyxteC9VJex55+I9lTPP8tV+x1CClIUmmc2H/4cjP/biBesi4DM21FI12VoT33M+npkhXv6SqzvkMprw2xeHJY5NwST2iGF9D+knkmTrY/H5jznGGvPucBR9t1UxCM2QF5qlSrJ7o4Atgn2eTACYFdAxpv9KR76uhySTz8JvqtUSXlTMnIh6X43nTTM/G3rjzDB/ntIeb0H8eJeaA3chFVSszGYQ5Cv0K7fBM3/HdD6vRKt365Idt+fTPLvJszrITYWr6J9+n/RGvKV3srJaG8g2bIcWHwWBdcc2jdy00004b32tT7rhECDmSgSxQNJHcjRodCedprNy4TUc3mgjcflZAxTSN67E3lKef1wjXCMbOyfQkDVxgjQuZugEnCV9+lCml9qGwTm72Fzy4OgJyHZuGoIYzCPR1s/dQ6+8+DP1sioekpw3R4ElQoLjkW3zDPDKsY3IH7dnjIOFdZnnkdNtrl2KRm91ebR6jQ951MvxF862TjNQnqRB+pWRzpUA37Xlo8Wb0A8qgyQ0Ni3qFBRLOea3sA9NTzjKGMqq6HwPZ9D40uksH1BhSpWlG6gPZBi44Wxi5BgsjkSmnakIOJuDGJtLNYeCV5zsSpC9t0BWNJ1W7z3IOXFb1wrB+e2s/70eUJ+gsCBO5EyO9j6uoEiHdxvGLISzUJur72NaV6LFE4fQjaR6gBUY13yvaK6KRKY/0DqjjsGMfAzkJByNkLnvTDV3tpWsSoZpUDM8qQMfTKlCTyvRpbGxMZ+XWRh+JgAQEE5Vn6LwiAuJGWs7ZDAswoCuObaGOwdXLsaAtPuxKyQVdocPvd4GxMPXo1GQrBXALuiTXqwn7/2dyk092+wz6si0KCkUh7aGL5vbRyOhKRwznUN7umVzcPtPO81dxalFXvak6PoBe/X1/r9z6TV25ZHQuLdwMFF5hephafo2u5DquwchUAMb4U5EVkxu2auWcPm25loA/8SbbbLIOV3Bpqz26N11Q9t6t5TcQopoDoMCWZhKN3RWEJl+3wFCk/xc3QjtMa9S/NwCpTtRbzq+9a+x639oUfmMMrkyipw7WA0l48nI7jYu3sL20jrv+fROn8A8dP5wC3lxgvxuElofb4RrNVnST2PhqEQ0FeQkvQMUgYORYLOHUg5r5Tgf107rwNS+l9AHgbTES/eKef9PB89HymwtwRzcTCytl9G5YTcuyGetp997oD4Q9XwENI1NADta1ORN8PlyKPgaJtje4Xtba4DCd3XI6C+SbkianjXHqReJ9PQ/uw9k0bbvGhgEKIGA1hwzdIIqN8JKU8PozW3DFK0zq9yfTvEy54i9Q7zgNEG1vYwj1ln5Onhi0OsiEDEc4P10AMpaguDe25jY7Cpfd7f5sIlCBCZWmUOtke86Uybr/MRMHamPccrVI0GFtHe+wjm1YCMGNfacxYifnAl2k99UvZmm68IRP6ljdvTwP8ggK8f4i+/Q3v1zxH4fy6B4pi510aY5yIC6N4lBcRHIiPIqlQABRAo+QfSXGyT0F4+KjhnHOJDyyKe6fnLXgQJiAu8+y5IFn4E7Wuhka/avvpd3iKbf8ej/f68vPke/H+A9cvyZe63IZa2AslYX5IClBdj4L19vtL6IUwavTMCqTZB8tZ4BDJ0szF8Dq3fdmgPuwIZEXoVffcC/Rq+i5eLLkJhlz66IU9GewWBzDWDxYvisHGehsJKvdfTJqSpQnIBSwQy+GqvLyED24aID81CBraxwfnhfDkcAcF7E4Az1pZ1kd41yubfj5F39pqIJx8c6gHvAAAgAElEQVRNAQAOyTBfIH6zjo1FP1KD4jZIn2oQXVPmfqNtDnqP+fbIEPUaqc64DQInzyEFgboVub+dux4ybl5n/fhdOGKwzu6nVE491M4fiGSX/QicHeo8d9ZDfPQuBJy3Q/vIA8hI4o0AZ1NjEZTWfrR4A+JRYJAKVBDInN8TKQ9V3WUR2n4b2nCOQ/l57jPGNgwJ6mVD5TKL+GQkSHxCaU6ai5AA05icFT2REHy1fe5hC/N5cryYkAA6GitznGHYPY3xdEdu7S8hq9AwpIDdRpo367uNO7h+A2RN8uE4p6B8CD7Hgy/VPNr6odnKcOb0yyCkhG1rjPplJHx1QBaD20hjsUejTWJyjc8Ybu/nK3KsYn3+IbIyfkEKPnlQzFfmeBKBEn2C++2OvH18SFKYVPhwpEhcZe3/gDRu2yfsLhL6MQopL/cgJbcz8tj5HG0wD1EaZjPE2tkOKftzSEEdnxQ5tCp9jxRI2h6BW3ciBWQBaTL3MUhRbp9pXxckAIxEFo/DkeD+mR0V8xEgC+ZC0nwwH5ECFssjgLCwWzkFhTVk6ZqLvJIOCb77Knh+z8w1a1hb10VK098pTRrdC63HhUgpH2XzdyKa14NsvH5O6sW0M/IAmpr3HgiwnmT3nBj0y6+oknOOUl7mFcypSOB9Dgm75ULdarrWxv9mMko9slROt37+1L7rZGO9RnDt7ZTxRrF+fRMpUKE3zwSkBPqyy0eSVgUKy5Jvi4TnskIlEmS7IlByO6TsdCDNZzQtr3/QOutp82EW8mbazX5bCgEXl1ElxJaUpx1YC0+za7e2+eZB+06If16C1vWpVAHoy40/Ba3qiK/U9IzGHIgHzUegyj5I8bje+v77iH+V9RyjRgMYUmank/JQDzjuiAwiecadcO348+8lNRp40NuD3SMo9c45FfF2H463ElL4Vw/uO43SEFbvSXeyffbewRcg5X3bMu+3HSlfaW/XnI1yt5yFgICaq4Nl+sC/RzuknNyHhawj3vYJKRAxDO07N5MTJksK3nWlEfII2p/+ibxS9kOy1pdIJhuLlNGVEXj8c3LCDhGwNAurRGff7YV49PjwnXOuHWj39vLFFASE+VD5CZlxnkxppTNv6BpBmrulEEhn9/ku3w0FPBUz43g8kqE2R/v+QtKQ8AsI8oRa375MZU/kNazf/4hA8hcReLExkseyPPcB0jyVnWzueE/rrRAo2AsZPTaydz0muH4VJI/VVO6+XH8EfehltHuRwewdZMgoJ6Pta3O8UWBxPY7MOIfeNdcgsMxXUtvc3qlc4YOZpF567W2d3EDqQe9lq1CPCf8/xvowNHANRwZonxdsWQTePorWcW8qVC/NtHE1JJd+FHzXHvGmoQgILRTaa9ddjOS67yF95U3EN5azudEHGSu2Q9ELM2ocF5/r+BAkg19Dmm4g9Ph7l9RIMQnJHFchwO0RxNN+TJlomGacRyuhffFgtHfeSpqz9QgkB16B+NrvaKJXbWs7WrwB8ajDoGohn10Dk+mE0PFng+v/jATUQiWejWE8gDayPRH6Hm5mZ1MgnIFUkF+BtEx8D2PqPvHkfta+WZT3Yhqdc+9uaOO+2q7xwNFGdnTItGFTtEH0NIZ7JWmolReUT0YCmres9jTG1ayx5Tnvchhwa/B5fyQYek+ndpl32dUY8ypUEMCC83sgYGY7tAldgASGTghIOgMpgeFGfCtp/quzEUh4GhJyLrf7HIpyLITlxH2idL9Zn2Jz71MEULxElXxhwTj+Am1woxGwc7f9dhoScHy4lhdovSDWE1leRiB39cvsHU8hCOELntUfgaoDSBUev+Fvb20fXGHsfBLwy9BGcyZSpMqW57b+eZtAcbUxCYWzZnUrt3uOAJ63/1+wNeHnyVTr8wYCILLMPYIAoHOQknajfb9CsM7vRZb9pRHv6Ycsf79FwtLyNiYehNoFAaxlPSntnHeRkjQYKXFVwyYor2BegHhQWWG34LVhYuBct3VSsO54GgrwJ9scKMtfUM6nSaRVPlci5Qf7ofW0EvlVgU5Cgn0lT4RjbMxvRQrhQZjygwSlqyiTr4HUy+9QxLPGUppQuSMFq7pQytOq5iYM5uwx9t7nknqOrIpA86TSu1e5/84IfDnbxqABn7X7L23zocmVl6q0ZxOkpG5q/XyLjdfWCIyaRoGwGoqV0N4AAw1Jk8T7PXFt5AnogcZwzwj/3580X+HdwBPBb/vZPbxw7r1zeiCPgZloH+mJlJAFVAcxfZL00Dt4EpV5sC8w4UHcjkipegYZ8Mom3q1wz5XQnro82it/RZDIF+W0esbGqzNS3kMP4VuRAn8+pUpWmDvnSRqRo8T68lUEfHSz9/0I8ZDVg3PuJ79Sq/e0m4JktYmkgOK+aP2GynMIVHdEYOmNNqf8dVeh8LMGBk3KVzo7o8g8zrlfo/K/IcXxB/bcN22MV0WAyaPWf17ePMj6uOr4IO8N7+HsEyKPRJ7Yb9v8XR7JJ++Tn5ze9/EqaJ2dSprk+VMMELFzGgXsBM9YKvN9Vka7BfiJ/VZORmsUWFyvg1KedQICOe4gBTsvQny3d6U+JK3meGewltojIHIC0m9OIyfENtOGsALoFBv3n9pc8zpODyRPPEhBnc6u645ktPuRl9J3IXOIfw7LtqdcX9nnM6xd85AM9zCpQftsxGdmWv/lpo6p0Nb2SM7086SdPe+q4BzflzsheX87SpOMd0O6x4OID8+mkZVLC7a5N2ke4vYIjLsR6c9d0d52HOLxNVcwbe1HizcgHnUa2NorqjUqoz8SqDdFwMNt9l0XtCk+RFCZoYa27IbAi6cCZrIMEirvMmbYk4JeTJl7H2MMz7uADrZFny1f3ht5/0w2BjwKWeinITAlrCZxqr+ffa67NcYY6G0I1PFC7yNYJbwy11RN3G3njUIW1acRMJnYcy5CXjBhiJnfkE5BgniCrA9nImXsKWOeb5EKXDtmnjeVNMxqKQTI+Nww51EhXxilG/EWlCaVXw1t7l6guRgJWGtk2r6qzSufV2Z5ZEX6AXIN/zPaiMNnjUBVd3y790UgyY/QWigbq03DJOD729ytOG+QkncnAlq/jzbxHyCB42OqVOSpYW6tQBpCMhjFy0+25z5JKmh54aZX5vrQs+0rlOPB50E4wvp6DAKizkTC+TwUStIBCQELbKxGBO/+DKkFuUgS+lEIhHqDgt4mlFcwH0PrvqywW/TaamOEwIvdSXPLeY+I5ZB1tFq+rp8h4a07EkTPs+8HIC9Nf7+aqwIhoXOmtfFQe78ZSCGciUDw3L625//a5tH6SOkagayy/6FxniNF5kG4zr1Ssyfib0Nsvg1AxoVe1canzDN2ROtwM5sDV5MBsoJnfw94vKnrtEp7VkW84pngu+2Qt0FNldAK9OtaSBn0gO8GSBHy+3Bn0mS55RSVk2w+hIU/7kMK2iyktGyauWY04ntd0dq7FXl7vULgQVHlHWr2DkZA4zukIaA7IuNbozzaEBj4MKlHnvdyCj1vr0fy2brIe2g+WuOeJ4/G5K/MXPP5DQc3sm1rIJDiDCQP9UJ7wTuUJjPO875aBvG+8+zzQcjANIFUEQyTt4f7qzdQdEBA8TWknufjkIKY521Vt0pnNfRZP9K0EXMJEkMjXrM1pWDCtOzcLnPfpYNx3QbJGh6E2hatu58imeB1KssgI5BxYi+0H/vKu5sg48+xTXj/Srl+apLRgnvWDBYvgnE+FvGn76F97wlMNkb8/ykqpIyw83ZG3voHIaB+YwTw9rExuhHJSHkg1HdeqdYfo5DRyYcu3mxzwXuSdqdA4ZXsOAafH0Lg2GB7TkWDXtA2H2VwKQIOt0D65nNIH/sK7cWj0B71X2RywtXQ5pUoNW5vA/wop03tES993NqxSuY+ayIZrG5RLPac1ZDccQWSH3wl5+uRHuGTyHcM27+4HC3egHi0joMaMvrnLQK00b9IGgq3lG0a9xCUgi3QjiEIaf8esuj9gTR/Sw8kTDxIQS+mnPuviICUebboPyYTFoA24vEIQFnJmOf9CPTpggSc6WQEeRph/WzCeH3P+uFsa9dAZFlrlHJBaWLWn9l4voZAAR/qsDoSlrJg3QbIq+mqYC5NRMDSpUgg830TWmj9d+MQeBZWu3qeAl5Pdu5ABOJMQe7po4PfbiJwo0XK+JrB56wgdiyliS33oUxlDwSU/pzUsu8rsfkNvxrQ4JOAv0ex0tVLow37Zyg8YEPkTTWUJpZ/zTzHJ/O+FwnQvjz1s8E5JyEAtCOlioMPodsBbeLzkEX6puCcI5DH0wekSS57huNt62wgUmq9orcrAjD61PAuK1AQeA2uyVMwf0iQELlO1x6LPOJuRYreGCRA7kUa3lM2N1jweRxwlv2/FhKMr0eVoLKJzwtXBUKgxhdYNSe0zvdH/PFaa2fuPET7wWo2dxfaWF6FjAG9kNdgxcIUTZzTo5DwfDcCNpZDivDriI/OJacqWw33P8eeMQDxTW/R9mCt56890dqqKWF6jW3xXicHIsF+sp83aM87MG/eNPJZPiTYJ8g/w97vcwQYl9uH1yEFuVdBfKIT2uPHo32mA9o7dqTMmrc18iGp0rUxaX6SojJHIe/gnPn0X7auPgN2amI/no5kDO9BuTe2t5AWW3jKr1/E/2cjQNrPtacJKvUihfNFLOS9CW07HIWm32Vr9iXSZOCVvKnbIw/jOaTFBaagfWP/ctej/IIvIL55ebC+ZtlafZsKSX+pc6WznOdl+W8/pNg+SpB3Cu17Ya6eDnn3yblfgkDWF0hDWzci9TqdaM8baOf2ooIMggCAG0hTIQxFcvoxdv1mZIyEjeiTbK6fQWRkNFKeeBuWYsM+n0d+0aRGpZKo05h3tT70xQDmIHkpzO1ZqYhMKDPtaXPleSQHjyUFsobacxqAUKRgZA875xLr28ODe9+IwO3CnkQ5bQ1l9pvsfkXD7oYjeegMtB/8i9SD71MEuuwbvNvxSC7s00zjtBXwuv0/0NoRAnej0N4/kWaodtvINm5kbbudoBquzSe/ly9WwNN379jSDYhH6zmoMaO/LdrzkMDQHQn0TwcbW0dqSB5n13ghfhekrPhEqNfY/dpT0IupwjO62TXjME+m7AJHoNnXSGkbhIT5O0hBqBsREFW2PHkdx8kzz9UREHYHAggKl1kO7rU2aRjK6sjzZKp97oPCCS8g9dYpl3dmHHJZ9YBCO+QJdUzQ3nJVenoiUPBitGHtijbyqhs4sph8hAQoHyL5ChJAfY6mXFCOfEHsbpQbp2/e83Lu4SuMTW5E31dNAl7mOq/Ubm3roFmq02Se4ZN5H2Gfb7d5sLethbcoA5qhUMGvgd+QVhH8CPhBcM6KpEJWWW9N69/PkEL6HEHFynoeNEHBbMy1NueeRh5Kb5PmvNvHxuIlgpwhOdcPIAWK17Lx8R5XK2GeRhWuL1QVCCm+X5F6ebVHIOp5lBHgrG3nI0GrLxK2D0QeVP8hJzlvM49lX7Q/DEJWz9OxJLdoD3sJS85e8H5dSJNgb4jAlL3tvV4jDb3YHwvLtc/LIhChSYBAlbbleZ3ciICD4QQ5+JrhWT4k2Cfd7YG8oXsg48MTwEH2W6hw+cqmnUgTEf8QgYF3IOPST5BnRJG8Uzuj/BiNViCoEaS2a3wVyu2b2I+jkHL2GJJ1vJfHHmj//SuSfw5AnlHeg7E96T7Y38YiBPGnUCAPaHB+A+Aj+H8XZHS4E0sMXOE+25GGhLdHgMajpIUhDqXU0y1bffFdZNTqg2TAmcF9j6RABS87v9krnVXqs7BdyIjwcvD5WhvLakap0UgGOZvSYicJaSL9fW3ueS+rDgi0nY/JX+XaaffxMtLawfc7Io+qE8rNhxr6xOf62R4ZSk8H/mbty8poo+y3Qwveu2awuJnGuUH4G9oHtqQ03PNPCFArHBIezPvZNr/XQ7KAr3g4FPHw02ho0OiBQOYbkAy8BwLD9g3ufRUF07FUaF8IQvkw6AY50ZCssU3w+WpKDcAPofQCuyPQdLy9Q3fSKpk15wqu8A5rIuB6UyQTjUB89DR7XgckY96L9NeyFX/rNK++yx2JAH7P+7dDQFyLefgtkvdv6QbEo+0cxvS9S+Cxtokcj4Sjl5HlZX8EhhTyxAkWYM/gu05I0PHhN9ORktDbPlf1YmqGd22H3Jm/QJb9pRGgcivaNLuwCJLI5vWVb5/vK98n2XMK3nMgQSJxG8uPMGUXgVI+/0TFsE4kPL0dMNHcXB8V5tYxSEGbQ5kKY5lrtkGuwF7R7oti8h9EG/jtZLwaKCaIPWTzu3PBPtwNCUUrN6L/GxPu0x4JPq/SiLClgs/wybzfQha5XqRK9lVU9pQJQ+jGBt+9RVDBsoa2DLCxHFGPd63w3EYrmEWvDebduUhYPxkp352QcOQFzdz8JcH1NyAB/3yk8O2Hwl2qekw14t3GICXRg1DtCCx3OeevhnjnG6RhQ74a38HUUciyNbkJcKNva9Bf3tvkOLS/DCjSN4hXXYQUhlft/QYhRWASCjnZ3PrIV/HpgPhas3s+kVEUSL1OzrDv9kd87gHS3EJVFaQCzw1Dgs+2+feM8YsOCGTMFgvwFU2Psv6+EikHXW0cvMV3FAJwC7XT5mTdvMrqOD9XQJ4tW9nng21eTUA890rSNASJ9fE5KOWAB/B3RLx198y9C6VhCHhIXh6/rILZPu/74Pc1USjur0hzgC2FeOEvsDD3MnO3L5JFLsyc02QvrjqNXdj2O2yNPYX4zQ7Im/V9WyNPUd07fwMb36OR/PEWmSILCAh8G3lxfoT25J2RIj+eykaG7e3cKWhvPjbze9nKrrX0hX32uX5eRUaYexEgvRmpjPYcCl07s8Zn1QwWN+OY74fkVG/87os8aPohPeFGSj3s8zzQ2uf9jsCjV23cH0MeUb6Y0hAEpJxFaXjts2jv6WD9ORUBlQ8Dk2oZO5t/21Eaql4CvJWbu/Z7BwSOrk/qkXoxwR6A+PzTSGf7DwJdnrP58BAFKhTXOF49kDH0E7QvnWDPO8z691Jr0ygEEI9nEQKbmbbugBKv32LtXawq3uW+c0s3IB5t4zCmdB0SNrsgwah/8PtU0hxQR1NDSBDaVH9qjNzn5zkHCWJ7IqApG+5W1Yupmd67N9qsDzNGdSRy686t2tKMz/WbwnIEIWEFrqtZsUBW81+SelCcZWPhBcg1MAG5wL28NXp4re2w67tS0GsOWfS/JfXY6oisGldkN87Mdc0uiLGIhSKb/3V3z7X+fM/6ZDgCOKqOD2kI3TtYEnekaP4VeY3UCtRVFN7b6oE8cjqQCmIPBO96DArbyAvJ8Pwh9HhYH3n0vIs8Sj4iBaibtd9s/fwG2KuGazZDfP5F4OM69mlYROJmZO38Equqab9dQJCf0Ph6LXvW0ciSe0fw3cHI+v2EveNuwW8dqKN1lfJeJ8fZd1PQnr0PzeQVQsOQ4PWsHyfZ7ydgydaDMUkQMDUNeUbcaTxl4+C+JyDjQGOq5rYp/mD9MZfSapUXonxLX5Ia4dohpfQRG8t7sDwpSEbZJrhfLQnQy+bryTm3mgHKh2OuiACUD0mreI0gIzNmrj0SKYAHIuU0TEx+C00MCavzGJ5MmtfoauMB29qaH2L/e4U+WxU3TOr8JwJZBAEL80hDSv099kCVAL+xfrvf5kYD8Cm4v/cUvxcZdH+GvHSPaYb3z+b6OczefQsbz2+Rov8iAmx2tX56ldT7vNWvWwTkf4L41l9Jw+wvQt5jvyDHOw/tlZcgnaZ3dh4E/dcDeeoOQTxzKAJlvBf6TqRVDDsjAOVOUiPfCGQIWBEZemZTMC8oqXfdE0gWOYPyOQy7I4/LPHCtE9JZbkcGhvVRGhVvrBqIIgb2RUDb0Wi/6kaBIlWNGLPlEB8aiWStO2xunoKA/+tsPXSy/i0U/VOHdnrDmAcB+7WVddGk927pBsSjbRykguMVyJr9BHBa8PumWG6Qovezv72MkQ5Abspz0ca1BfIKeIVGhJY187tvbkzsaNsYCuUmaobn+hLYr6INu4F1INgUutGIKi+ZZ31km2BCmsujbNhPhXsNoBks7DW0+xekXleDkeVthZB5swgEscX1IE3mvZCC4Q/BtWNsfC5HylOjSzovbgdS3j9Fitl5SAHxwuaBxnPWyVwTzukxNn+nY9Um7fveJnC9jIVp1Kn9wymQ2ypzzQrIu+MVminPQ5nn7GL7yisopOVqBJhNRQrc29So1Gb6vq+N0W3ARcH36yABffXsNXV4R8/TqnmdnIzA+VOQIlTWW62R7QhDgt/DQk4R2LVJTnsPQvLDkyis63a01/dH8sDV1DlvT2s4SL2OT0b7rQcQ90SFLz4jLXLRjiDFALLmL2iOsaRhvp4Gezel+WYahOCRX6H1bluDFyIwLbeEOAIk3iU1Pp6PQkXHIQ/7t2rlM/Uet+D/oUjeOCf47mJ794GZc0PQYWnSXEFb2N95NCx3fzelyd6918tx9twi4d1ZT/G1kSx7LwqnPr8Z+iTM9XM3Ch/y+9HlqFL0lfZ5MPIM25E2omBbW+8g5a9DEQjlwZ+VCcALSoHFdxEwN8PWwXc56kh5wAb2jN+QFmzpjvjiQkorE7ZDfHIw0sU+tPsfg4xYW/rrC77bMsgw5MHi4Ugm2TNoZ7j+XycDJFNq8F0d5ay6FYE+/Wy8b7O/Y+y8kajS8d5F2pm3BrPzJ28+2ZpZE+luayId4AUERE22/ptW5/lTtb1l2t4i3liL6mjxBsSj9R+UCo4/tWOaMctD7LeJyP2+Z9FNBbkcjgVmBN9djCxx3iW9Z9iGFuyDLZBQ1OwofZm+9hUaNkcWjdnA9My5WVfcJlVsQILop6Qg1DlUcOkucL9FBULtgoSBh5Ar765lzqu7ILa4HqgyYKO8vBAgeRswMviuTQiedezPsJrckciqeofxmOtRSNiGmWuWJ7W6eqFuLRRO9rpdF1aV6k2Q3LU1HdQxN4vxywWkuYmORh5PF6HQlYup0agR8GVv5T8EKQGDkMV4qvHqy4Gui3geVfM6GWCfO2Ners3chsIhwUhOeMv66joENpyGkttehsKXFsm+0VoOpHxej7yaHkTeLdOtr+4oc81WyMOtSTkoaZivZxwKAfaeV6Hy2RPJeA3SK1C+QutdKExsaIU2HEHqxdw++O58FHZU10pUNfZXqDD2RkDSSTYWYfGTa4FzK9xnEyRHn4a8anrb9y8i5bi3rakvSEGPjjZHvNfQOFvvq1M5j2Kep/huiCeuSeNCzKvl+rkcyVrtUbjtORSQ0VrLgcAfn2ttDQTcv2zv4sPLhqEwslwABQEdTxHkt0JGgDdIow3aI/n1Hbv/Xcibann7fQ+bS7PJhKYG91zX+vcia8/sSvMhnMdofXdDuQsH23ddkEFsun323jk9UIjo9sF9OgT32gIZIdZCBpAT0fpdH4Fcq5Luyf6amo1YwbMr5UvbDnlZdUD74i8QIDYF6bE+gf9klFJixca0obnaG/zv+V/F6omLy9HiDYhH2zhoKDieawv3CxT3/D4FBIWA8QxAANYsVNkqdD2+EglBzVJWvhn7YJlF9Jw+tok8T1qOeFlkSTnSPocWifk0X2LZUShmepEnV29iu3dFFthT/TzLzh3qIIjFo/D4LJYhdI3si7xqcvshEOpi5CXUI3NNYjz4DhTqNQl52+yMXNrHIu/RmzAQCnkafUQze7y09sN45c9IK7J2RADoXIKQwVrnIrJQ/wp5Fr2EPDu2R8D2Qwg8HN1c71GgPYW9TqgzqEMmJLhc3xKEPiIl5QSUr+QsJFcs0iSwreVAecVGIUXpEvtuM+A++z/0MBiGZLFG5f/L2Rd9vp55CEi9Anmo9QjO6Wm/58oZ5FdoPRzzTK7Snp3teWH59D0w75LWeBg/8WOzmr37TIK8LXlrIFwbCKj7Fks2HZzzHCp+cwWZCrxY3q/gcyGjEPme4gtJiwDUErZZNNfPnQR5xSggo7WGw/jSCJQs/zKk33QGTrX/dyLNhTsoM2/92Ha2efFpdg1bv/g9enPkffgK2ks2QmHjryLD1G9svW+HdKXc1B9IhuiMwI11q7yfb+MwLEeqzd9bSHO/jrF2drbP3a1NIfi0kvXPctYnXyP542u7vjvyYLyPRqbkqPAOFfOl2fM/R0DaD5CX7Qk2T2cgoG4m2iPrnsu3WnvtnBDon1FurBeno8UbEI+2cZAvOM5GFpw+1CA4IpT8WlJXzFEoaWroZlqRibZQH9Rls0QWjInIw2GAbQwHobj+CaQJxo/O9FEPJPA1K3BizHvHlu7vRrR7BHLpLVvRimYUxOIRj8YeNKwm187W/CUEBRlyrjsZWZsPtPV/DamF/E4EQPjy9jsSlGVfkg7rp3NIvYFGIqPGXZjyUMO9EhufGaQhkr2Q8B167y7SMCGa4HXSguPiE+SGVdBeR8DHYilwE3gbZL/LnLcMQe4PlNbgA6ToeeV1MwTUjSt3n4JtycvX40ODVkOeOD4krhtBvp4q9w/DMd+nQIVWpKheaLxvLJKFXicTftwaxtD+vwJ5TXuFcQWkhB+Jcu9tV+a6LqTeiKNQmNJpyItph8zz7gReCj4XSihf5R0KeYoXvFfRXD+rZ/qgqozWGg5be68iAMgXbvBRATcgGblL5hq/tkaShhv2RUagqbautkPGp3WC53yGjO5TUYLuO5Gn4RTScOZOVPCsLcoHgjk7CuWbGmyf10Y63bvI8/BXBKG29h7bZu7VGeVtuwXpc0ODe3+O9Jhudt8mh1QH/VsoXxoyOH+A5KzB1of7I753LUoaXzjvY73ba397IAB6cEuvgUVxtHgD4tE2DsoLjtMpHmvsF+QxxvyON6beyZj2U1jS1CXlQK7Wv0dg3g+R9cArTYcgZeJ6Y5yfY2FM1m8HUEevnaKbWms6KODSSzMKYvGIR2MP8qvJlfWyNB75vB3zEfDwMtRCz54AABIgSURBVPJg2BoJzBsvira39gMJ+xeiMOaLUZjLNihUprGVno5HFl7vldoThS70aaF3bLTXSQuOS08bj4uRgjLW5nKLJH9dhO+9IwrVKuvVQxCWikAhr+D5ancDTWnxwFCj9mcq5+vZ134Lk6JvTsGE8DSyQivKoXMECku7v+jzFtHYZXM+XWLvOAmB0H+0dbc6Umpz87bYer0OeWJ+RBpKdzjylNnU39++Xwg83Mzv0iQvJBqR66fM/Gs1Ob3KjHM34623IsAizL92mR0NACGUVuQXBICi9c27yMtpRsgDEIh3Dg15+D557WrCu62NyRbIQ/glUsBoIAJ9d0Bg4kQMRCUIwQ3u1TG4Vy8b73dQzj8PQE8G5tj/NRl8ctrelHxp76F8evsG5x2MPKN61WkeNTW/2xITidHiDYhH2zjIFxyfporgGG5wlJb3HG/Xb2+fl0Ku2I3OOdTWDuubvZD1ehzwDxTjH4ZUTLR+uoW0pLfvzwYl1uNRuO/bhDt4PBbvg4LV5JCVfQGWFwpZ289FnqlvIOPA+JZ+n9Z0IM+KUSi0cROU5PpdCuR7CHjsFsDutv+tb3x4b5RzaR0EANY9f0SVttbsddLC7V0FKXfPIKNTqwEb6vS+/ZDXx5nIEn54zjleLlqW0qT2TyIFcaCt83GNeH6t+Xq+y/8U/q3heY2u0Gpy4FK1XLMIx/FG5DlxBgIMbkGeDDsgoHDp4Nx2mWt9CLrP6zSL0vxaR6PcT29iyrJ93+weGjTCC4km5vppS4ftrTPt/60RaPR9BJJuhbyYls9c48f3YuB4+7998P0aKOzK5x4qAXbqzcONf2xH6kl5EfLiux8Zvu+mQCJumwcjUbjgZFTdsAfyhLsZWMvOG4miWprDc69J+dLIN/TVLS1BU9vb0vN/UR4t3oB4tJ2DJgiOSBF42RjdvQj53w/F/g+xc9rcZtUMfdoL+B3wvwhsOgV5lfk+6YyEwtsRaNemcjO15oM24g4ej8X7oJjX3rLIgupz+nRE1tnHjSd7D4YljocW7OMh1n+FvZ+QN8KXKIfFL5ESMQ6FYMy3/WzPVvBujfI6aemDJlZubc0HqbK+BlJa97DP2yJvl8ODc8PcH0/bvPPXzzCZ6WXSlAV1z9cTjwb9eBLwaPB5leD/s5A3dW5YZfD/EGQ0ONT49pkEFZVNFvTebXUr0mD3L+yFRAvm+mmBcT4aebWGeZ3WQWGXc5CsvmbOdd3s73FIfu8S/DYAGSz6omiHE3KuD3l4XTzyETj4KxuvtW3MvKfTrsgzr6pRGzkgvIwiMvay77ogve5J41kLaKJcTTPmS6Ogoa+1tHdJOXxnRYpUmJIk6YYW2t8rnLM82uQeRsrTs8j18ffIorwLQskPRsxvvHPuz3VuequjJEnaIevZJkiRXIhccNsjj6htUfjdBGBj4HTn3N9aprWLHyVJMhz4zDn3y5ZuS6RIlShJkpOQe/cc59z7SZKMROD0qc65n7ds61o3JUmyMlKwf13lvMQ555IkWQZZT//qnFuQJMn5yOt3byTE9wWcc+4X/po6v0JFsj15Befc562hPZHA1qcvsvI/wGTn3JdJkvhQ0AudczfYuT1ROPiFzrkXg3sciELwxzvnHm9kOzohvnEZCpf5M7LGH+uceyBJkoHIs2cc8Ns4dxpSkiQXA792zs1MkqSTc+6fSZIsjQCl9RAQ/X9JkrRzzv3HrumKDKzXIg8MH3r3MTIg9Efz4C8IuNjJOffHRf1u1ShJks7IS+Yr4J8oLPCZJElGISDqEATuHwrMd86912KNbQIZD70aefV8i/STg1A+wfdQLp/fZWXFJEnWRMDLWATEnI/G3Odvux0Bzm8mSbI22jc+K/P8uvLwJEl2w6peOuf+Yt8NQjxmqnPuJxWu/a5NSZLcjgr3XAm85Zz7KkmSDojfrYRAtg8b+x5JknRBCdFftnnWD43JgSgP4wvBuXeiULaB9rmDc+7fOfesm6xfj/YuCRQBqEjNTkmSJMi7aSjKV/IWSp59pN+gkyS5ASVZvC9Jkr5LOgCQJElvZFWfhqxjuyCvsfuccw/ZOT2cc9+0XCsjRYrUUpQkyWooT8o2KP/eXsjC9nSLNmwxoyRJxiIl5K/Ay865k+37c1Bi2F2dc++3YBMjtXJKkmRLlBPlUeDfyOj2H+Ba59zXSZJsiypMPWcA0TzgHOfc85n7rIjSHCysVZnLgCGrI0NWXxRy1w15Vr2CjFynV1I+l3RKkmRvVMDkPOfcp/bdacib5Fzn3LdJkrR3zn2buW53FLL338AZzrl3kiTZH4XhLINCq78FnvRyXmuhJEk6ojn6tyRJfNGFrZGMOtc5968kSSYjr8s9kiTp4pz7R0u2uRbKW09JkpyFvKBeQeFSPjXIbs65v1a413kIwB2JgMXxyGOsJ3C5c25uawEakiQZjXKRbYkiL6YDTzvnflyOxwSGmfURmP4vNIdPAH7inLvH5siyyBvs3Sa2cWmUa2015L15mHPuxSRJDkdRInsC30Med2cmSbIQ+MI5N64pz11S2ttaKAJQkepGSZKcjOLEP0UlTM9yzt1hv10A/M05d3m02IqSJOmH3JhvcM5dnyTJUrbJt4qNK1KkSC1L5pkzAHlDvplVWCM1jZIk2QAJjA8jS+6WwNvOuZvt9/OAZ0KLZqRIUKKkJSiXz/8gsPK/zctgNAp3m+6c+8pfY98t55z7osj9C7SjA/CttWUL5E393ygE5WjkbX0ZCi/tjsLEPo5yWHky5fokFCb3IgIZ+qJk7d+GYF/OtcORd9s0k3c7IE/KTdG43OKc+6Y19b+1cRjw/5AMvwFSsGcA/4dAlc/My+9wYO+2KqMmSTIRhaT9FwpFXRP4vXPuT0mSDEYA4niXifgwz6d/OOd+Z5/PRIb34c6539nviXPul61pbAHMS+cO5L33L+fc/1ZrY5IkOyO+8SBKF7I14mljUGTLoSh1yFtNbFsH59y/DcSfi8KSD0b5q75NkuRoBO53Q/PwAbtujWo8tB7U1trbmigCUJHqQrYxTUXCzxeIue+JXHY/RjGyJzjnnmupNrZGMoHxIeQ99uvWtGlFihQp0uJISZK0R4DTa8BTzrkppnQORyWcP3HOXROc36oUikitg5Ik2R55tqyEcgNd45y7zn4bgpS125xzH9fp+SshoGQ6CgO5G4Gp41Co1IvIi68/cIdzbl492rE4UpIkKyBvmEHAN8ijqSTsrsK141Do00XOufuN3+wLLHTOfVTvtjeGkiQZhpJVr4TCvB+yUKPbkDfHB8gbd4Zzbk7LtbTxZODAJJSMuy/yBpzoFFp9JgIKJ2c9epIkWRf1w3wEIHoQ6k6U8HuMa+Vh8UmSjAH+XsSIlSh08B4EPPVHut32zrm/mDfnBsijZ34T2rOMs/Qixit3QGGfeyBvtLudc7+133uhnHZfJEnS0Tn3f4197pLS3tZIEYCK1OxkG/Uc5Ib4oTH5Fe3ntVFC11ecc4+1VBtbM4WMLVKkSJEi1YeyQFKSJBOQojjFOfd8kiQ9UI7CHVB+niXaYhmpIQWeT9uhfC8LUXGLQUjeucg5d62du6yrY67LZAnJ19OSFPKMvLC7CteNBi5EYZh31bONTaHM+9U118+ipiRNAeLX7M3ALOfca/b76cCGKPfTwcAC59yHmXusjNbR8nbuPOAu83qagEJdp7cVL9lqYXf2vy8a9XvkIbyfgXQjUSqVv1e7X5U2tKl8aW2tva2VIgAVqdkpSZJlUR6j05xzLySKJb8BbWSzgdu9q3pb2LQWNcV+iRQpUqT6UqCEDEUVqn4OPIbyeFyIQKgFiRJEd3HOfd2CzY3UiilRYvHpwJnOuVfMY2AkKn0+BMk859Tx+Yt1vp7FhZIk2RXll9sJJbSu6Dm1qCngiYsk109LkXkwfY6Agxecc5fZ95sCxznnDsmcXwI6AgcgAG5ZFKb9EioaNB6Y5Jz7YHGQ4y0EcQPkNHAzqqq5lnn+bYvCMg9xOUnVG/GsNpUvra21tzVSu5ZuQKTFj8zC9yCwY5IkG5u74Y9QPPkCz5TbOnOuF8V+iRQpUqT6kilaY1BJ5E+AyShP4QNIsH4oSZIdnHN/ieBTpCrUA3nJDbXPv0ZeUJ+hkJwn6/Vg80gZCmxmINNpqHLXGwj8Wt1O9WFCHSL41DLknHsUGOyc+6q1gU/wHU/cGXnRTUZl4z9EYZwjkiS5HIER3dsS+JQkyXZJkuxr/x8L/BSBs+8AxyVJcpCdugnQN0mSnkmSJP567+GYJMlE83j7AVrjf0Vhrf9ElSbPd8594K9ZRK/XrOTfO0mS/sCNwAgUhv4wyld3jPXhTcDVzQE+ATjn5gJnozDIEfb1AwgoBHkXnWKhoEnOLRYptbX2tkbq0NINiLTY0oOoYtOMJEnCik11yX0QKVKkSJEiFaVElcG2RCF2GyAQ4RoA59wskxmjkS5SVXLOzUuSZA/gyiRJPnfK8/MXVJr9SlfH0upOCXD/jarb+Xw93yRJcgzKU3NjkiRhvp42mSx6cSHn3B9aug3lyDz3zkU5w/qjCo7tnXOzkyT5NeKTe7omJppuAVoWuMQ8u1ZD3okjUBL+p4GLEhUBGgLs45z7S5l7XJgkybfOuQeSJPkhcB4qCvIwcEkY3lf/V6oPGdi2DXA+MME5926SJJOQZ89sYCPgfRThMq8539fuNwW4OEmS3xofnQ0kKF/aN76NzfG8plJba29roxiCF6lulMSKTZEiRYoUqZVQEGKyPQoLfwt5iHRCeS2+SJJkLFK6HgmvablWR2orlCTJLsC9wFPAf4B7zOulXs9bbPP1RFo0tChz/bQkJapGOAPlnz00SZJOqDDS6ghcmgl845z7U4V77IxCKKc75+5LkmRHZGg/37XSZPKNoSRJRiAvsdOdc1cmpZUbv3ZBQY46Pb9N5Evz1Nba21ooAlCRIkWKFClSpCWCLLTgYuAc5C7/GHCvc+4qA6buAA51sUJrpEaQ5fm5AM2py334RXMr60tKvp5I9adFmeunJSlJkt2AW1GepweSJGkHHIiKBVxWxvMpe49RKATvQVQV8XDn3NP1a3XLkPGxSyit3LgP8I4PM1wEz2+1+dKy1Nba2xooAlCRIkWKFClSpCWCzLr7OLLuXmG5QY4DvgbWRYmkY4XWSI0mm2OzkKJbtxL15pFxGVKGJ6LE46OBMciL5VBgSBsMmYpUZwoAzP5orn6CKnl1QUm2L0YheAcB53mP0LZOlvfvEmBaAEJ1czVUnk6SZGNUYOBD59yCOjW1xamlPXuSJFm+NYesZqmttbelKQJQkSJFihQpUqQlhswSfjnKl/NIokp3qwH/cM591tbCSyK1PrKQn8+cc7+s0/3XBu5BwFN/YCqwvXPuL+a1sgHwhXNufj2eH6ntk+X6uQDl8wlz/ayMwpLfBz5o7lw/LU0G3M4ETnSxQllFip49kepFEYCKFClSpEiRIi1RZJbwC4h5GyK1EVpS8vVEWjTU0rl+WpLqDRAvThQ9eyLVgyIAFSlSpEiRIkVa4sg8oS4BhiHrbhSIIrVqWlLy9URaNNTSuX4iRYq0ZFIEoCJFihQpUqRISyRF626k1k5Lar6eSIuGWjrXT6RIkZY8igBUpEiRIkWKFClSpEitlJbUfD2RFg3FXD+RIkValNSupRsQKVKkSJEiRYoUKVKkstQTgQPD7fP9KAzvb8B7zrmrnXPzACL4FKlWcs49Cgx2zn0VwadIkSLVmzq0dAMiRYoUKVKkSJEiRYqUT865p5Ik2QO4JEmSryxfz2z7+Z2WbFukxYNiKHKkSJEWFUUAKlKkSJEiRYoUKVKkVkzOuUeTJPk3cGGSJEtZvp77WrpdkSJFihQpUi0Uc0BFihQpUqRIkSJFitQGKObriRQpUqRIbZkiABUpUqRIkSJFihQpUhuhWL0xUqRIkSK1VYoAVKRIkSJFihQpUqRIkSJFihQpUqS6UqyCFylSpEiRIkWKFClSpEiRIkWKFKmuFAGoSJEiRYoUKVKkSJEiRYoUKVKkSHWlCEBFihQpUqRIkSJFihQpUqRIkSJFqitFACpSpEiRIkWKFClSpEiRIkWKFClSXSkCUJEiRYoUKVKkSJEiRYoUKVKkSJHqShGAihQpUqRIkSJFihQpUqRIkSJFilRXigBUpEiRIkWKFClSpEiRIkWKFClSpLrS/weWzDvXvUrKqwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fa6b9178d10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "pos_count = data[data['label'] == 'normal.']['count']\n",
    "neg_count = data[data['label'] != 'normal.']['count']\n",
    "\n",
    "pos_dst_host_same_src_port_rate = data[data['label'] == 'normal.']['dst_host_same_src_port_rate']\n",
    "neg_dst_host_same_src_port_rate = data[data['label'] != 'normal.']['dst_host_same_src_port_rate']\n",
    "\n",
    "pos_service = data[data['label'] == 'normal.']['service']\n",
    "neg_service = data[data['label'] != 'normal.']['service']\n",
    "\n",
    "fig = plt.figure(figsize=(20,200))\n",
    "\n",
    "#Count\n",
    "ax1 = fig.add_subplot(10,1,1)\n",
    "ax1.set_title(\"Count\")\n",
    "pos_count.hist(alpha = 0.7, bins = 30, label='positive')\n",
    "ax2 = fig.add_subplot(10,1,1)\n",
    "neg_count.hist(alpha = 0.7, bins = 30, label='negative')\n",
    "\n",
    "\n",
    "#Destination host same source port rate\n",
    "ax3 = fig.add_subplot(10,1,2)\n",
    "ax3.set_title(\"Destination host same source port rate\")\n",
    "pos_dst_host_same_src_port_rate.hist(alpha = 0.7, bins = 30, label='positive')\n",
    "ax4 = fig.add_subplot(10,1,2)\n",
    "neg_dst_host_same_src_port_rate.hist(alpha = 0.7, bins = 30, label='negative')\n",
    "\n",
    "\n",
    "#Service\n",
    "ax5 = fig.add_subplot(10,1,3)\n",
    "ax5.set_title(\"Service\")\n",
    "pos_service.hist(alpha = 0.7, bins = 30, label='positive')\n",
    "plt.setp(ax5.xaxis.get_majorticklabels(), rotation=45)\n",
    "ax6 = fig.add_subplot(10,1,3)\n",
    "plt.setp(ax6.xaxis.get_majorticklabels(), rotation=45)\n",
    "neg_service.hist(alpha = 0.7, bins = 30, label='negative')\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Tratamiento de datos antes de entrenar el modelo\n",
    "- Llamar attack a todo lo que no sea \"normal\" (SUPRIMIDO)\n",
    "- Transformar atributos categóricos en numéricos usando sklearn.preprocessing import LabelEncoder\n",
    "- Escalado de los datos entre 0 y 1 (SUPRIMIDO)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Llamar attack a los que no sean normales\n",
    "#data.loc[data['label']!='normal.','label'] = 'attack.'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "#TRANSFORMAR ATRIBUTOS categoricos EN NUMERO mediante from sklearn.preprocessing import LabelEncoder\n",
    "number = LabelEncoder()\n",
    "\n",
    "data_labels = data.label\n",
    "\n",
    "data['protocol_type'] = number.fit_transform(data['protocol_type'].astype('str'))\n",
    "data['service'] = number.fit_transform(data['service'].astype('str'))\n",
    "data['flag'] = number.fit_transform(data['flag'].astype('str'))\n",
    "data['label'] = number.fit_transform(data['label'].astype('str'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "#ESCALADO usando: from sklearn import preprocessing\n",
    "#minmax_scaler = preprocessing.MinMaxScaler(feature_range=(0,1))\n",
    "#data_minmax = minmax_scaler.fit_transform(data)\n",
    "#data_minmax = pd.DataFrame(data_minmax, columns=col_names)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "y = data.label\n",
    "#X = data_minmax.drop('label', axis=1)\n",
    "X = data.drop('label', axis=1)\n",
    "X = X.drop('is_host_login', axis=1)\n",
    "X = X.drop('num_outbound_cmds', axis=1)\n",
    "X = X.drop('urgent', axis=1)\n",
    "X = X.drop('su_attempted', axis=1)\n",
    "X = X.drop('num_shells', axis=1)\n",
    "X = X.drop('land', axis=1)\n",
    "X = X.drop('root_shell', axis=1)\n",
    "X = X.drop('num_failed_logins', axis=1)\n",
    "X = X.drop('num_file_creations', axis=1)\n",
    "X = X.drop('num_root', axis=1)\n",
    "#X = X.drop('is_guest_login', axis=1)\n",
    "#X = X.drop('num_access_files', axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.15)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training size: 4163666; Test size: 734765\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "(4898431, 31)"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print(\"Training size: {}; Test size: {}\".format(len(X_train),len(X_test)))\n",
    "X.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(4163666, 31)"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(734765, 31)"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Arbol de decision"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn import tree #Arboles de decision\n",
    "from sklearn.tree import DecisionTreeClassifier, export_graphviz\n",
    "\n",
    "#El split marca la complejidad del arbol, si ponemos 2 quedaría lo mas complejo posible (over trained)\n",
    "c = DecisionTreeClassifier(min_samples_split=10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "features =[\"duration\",\"protocol_type\",\"service\",\"flag\",\"src_bytes\",\n",
    "    \"dst_bytes\",\"wrong_fragment\",\"hot\",\n",
    "    \"logged_in\",\"num_compromised\",\n",
    "    \"num_access_files\",\n",
    "    \"is_guest_login\",\"count\",\"srv_count\",\"serror_rate\",\n",
    "    \"srv_serror_rate\",\"rerror_rate\",\"srv_rerror_rate\",\"same_srv_rate\",\n",
    "    \"diff_srv_rate\",\"srv_diff_host_rate\",\"dst_host_count\",\"dst_host_srv_count\",\n",
    "    \"dst_host_same_srv_rate\",\"dst_host_diff_srv_rate\",\"dst_host_same_src_port_rate\",\n",
    "    \"dst_host_srv_diff_host_rate\",\"dst_host_serror_rate\",\"dst_host_srv_serror_rate\",\n",
    "    \"dst_host_rerror_rate\",\"dst_host_srv_rerror_rate\"]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "23:31:05\n",
      "23:31:49\n"
     ]
    }
   ],
   "source": [
    "print time.strftime(\"%H:%M:%S\") \n",
    "dt = c.fit(X_train, y_train)\n",
    "print time.strftime(\"%H:%M:%S\") "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "def show_tree(tree, features, path):\n",
    "    f = io.StringIO()\n",
    "    export_graphviz(tree, out_file=f, feature_names=features)\n",
    "    pydotplus.graph_from_dot_data(f.getvalue()).write_png(path)\n",
    "    img = imageio.imread(path)\n",
    "    plt.rcParams[\"figure.figsize\"] = (20,20)\n",
    "    plt.imshow(img)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "show_tree(dt, features, 'dec_tree_01.png')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Predicciones con la parte de dataset de test\n",
    "y_pred = c.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import accuracy_score\n",
    "score = accuracy_score(y_test, y_pred) * 100\n",
    "print \"Accuracy using Decision Tree: \", score"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.14"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}