10. Credit Card Transactions.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import warnings\n",
    "warnings.simplefilter('ignore')\n",
    "\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.decomposition import PCA\n",
    "from sklearn.cluster import DBSCAN\n",
    "from sklearn.ensemble import IsolationForest\n",
    "\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Load Dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>credit_card</th>\n",
       "      <th>city</th>\n",
       "      <th>state</th>\n",
       "      <th>zipcode</th>\n",
       "      <th>credit_card_limit</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1280981422329509</td>\n",
       "      <td>Dallas</td>\n",
       "      <td>PA</td>\n",
       "      <td>18612</td>\n",
       "      <td>6000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>9737219864179988</td>\n",
       "      <td>Houston</td>\n",
       "      <td>PA</td>\n",
       "      <td>15342</td>\n",
       "      <td>16000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>4749889059323202</td>\n",
       "      <td>Auburn</td>\n",
       "      <td>MA</td>\n",
       "      <td>1501</td>\n",
       "      <td>14000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>9591503562024072</td>\n",
       "      <td>Orlando</td>\n",
       "      <td>WV</td>\n",
       "      <td>26412</td>\n",
       "      <td>18000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2095640259001271</td>\n",
       "      <td>New York</td>\n",
       "      <td>NY</td>\n",
       "      <td>10001</td>\n",
       "      <td>20000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        credit_card      city state  zipcode  credit_card_limit\n",
       "0  1280981422329509    Dallas    PA    18612               6000\n",
       "1  9737219864179988   Houston    PA    15342              16000\n",
       "2  4749889059323202    Auburn    MA     1501              14000\n",
       "3  9591503562024072   Orlando    WV    26412              18000\n",
       "4  2095640259001271  New York    NY    10001              20000"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "info = pd.read_csv('./data/credit_card/cc_info.csv')\n",
    "info.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 984 entries, 0 to 983\n",
      "Data columns (total 5 columns):\n",
      "credit_card          984 non-null int64\n",
      "city                 984 non-null object\n",
      "state                984 non-null object\n",
      "zipcode              984 non-null int64\n",
      "credit_card_limit    984 non-null int64\n",
      "dtypes: int64(3), object(2)\n",
      "memory usage: 38.5+ KB\n"
     ]
    }
   ],
   "source": [
    "info.info()"
   ]
  },
  {
   "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>credit_card</th>\n",
       "      <th>zipcode</th>\n",
       "      <th>credit_card_limit</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>9.840000e+02</td>\n",
       "      <td>984.000000</td>\n",
       "      <td>984.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>5.410432e+15</td>\n",
       "      <td>17895.316057</td>\n",
       "      <td>12321.138211</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>2.545234e+15</td>\n",
       "      <td>23778.651105</td>\n",
       "      <td>7398.449174</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>1.003715e+15</td>\n",
       "      <td>690.000000</td>\n",
       "      <td>2000.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>3.316062e+15</td>\n",
       "      <td>3280.000000</td>\n",
       "      <td>7000.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>5.365218e+15</td>\n",
       "      <td>5820.000000</td>\n",
       "      <td>10000.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>7.562153e+15</td>\n",
       "      <td>18101.250000</td>\n",
       "      <td>16000.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>9.999757e+15</td>\n",
       "      <td>98401.000000</td>\n",
       "      <td>55000.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        credit_card       zipcode  credit_card_limit\n",
       "count  9.840000e+02    984.000000         984.000000\n",
       "mean   5.410432e+15  17895.316057       12321.138211\n",
       "std    2.545234e+15  23778.651105        7398.449174\n",
       "min    1.003715e+15    690.000000        2000.000000\n",
       "25%    3.316062e+15   3280.000000        7000.000000\n",
       "50%    5.365218e+15   5820.000000       10000.000000\n",
       "75%    7.562153e+15  18101.250000       16000.000000\n",
       "max    9.999757e+15  98401.000000       55000.000000"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "info.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "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>credit_card</th>\n",
       "      <th>date</th>\n",
       "      <th>transaction_dollar_amount</th>\n",
       "      <th>Long</th>\n",
       "      <th>Lat</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1003715054175576</td>\n",
       "      <td>2015-09-11 00:32:40</td>\n",
       "      <td>43.78</td>\n",
       "      <td>-80.174132</td>\n",
       "      <td>40.267370</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1003715054175576</td>\n",
       "      <td>2015-10-24 22:23:08</td>\n",
       "      <td>103.15</td>\n",
       "      <td>-80.194240</td>\n",
       "      <td>40.180114</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1003715054175576</td>\n",
       "      <td>2015-10-26 18:19:36</td>\n",
       "      <td>48.55</td>\n",
       "      <td>-80.211033</td>\n",
       "      <td>40.313004</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1003715054175576</td>\n",
       "      <td>2015-10-22 19:41:10</td>\n",
       "      <td>136.18</td>\n",
       "      <td>-80.174138</td>\n",
       "      <td>40.290895</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1003715054175576</td>\n",
       "      <td>2015-10-26 20:08:22</td>\n",
       "      <td>71.82</td>\n",
       "      <td>-80.238720</td>\n",
       "      <td>40.166719</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        credit_card                date  transaction_dollar_amount       Long  \\\n",
       "0  1003715054175576 2015-09-11 00:32:40                      43.78 -80.174132   \n",
       "1  1003715054175576 2015-10-24 22:23:08                     103.15 -80.194240   \n",
       "2  1003715054175576 2015-10-26 18:19:36                      48.55 -80.211033   \n",
       "3  1003715054175576 2015-10-22 19:41:10                     136.18 -80.174138   \n",
       "4  1003715054175576 2015-10-26 20:08:22                      71.82 -80.238720   \n",
       "\n",
       "         Lat  \n",
       "0  40.267370  \n",
       "1  40.180114  \n",
       "2  40.313004  \n",
       "3  40.290895  \n",
       "4  40.166719  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "transaction = pd.read_csv('./data/credit_card/transactions.csv', parse_dates=['date'])\n",
    "transaction.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 294588 entries, 0 to 294587\n",
      "Data columns (total 5 columns):\n",
      "credit_card                  294588 non-null int64\n",
      "date                         294588 non-null datetime64[ns]\n",
      "transaction_dollar_amount    294588 non-null float64\n",
      "Long                         294588 non-null float64\n",
      "Lat                          294588 non-null float64\n",
      "dtypes: datetime64[ns](1), float64(3), int64(1)\n",
      "memory usage: 11.2 MB\n"
     ]
    }
   ],
   "source": [
    "transaction.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>credit_card</th>\n",
       "      <th>transaction_dollar_amount</th>\n",
       "      <th>Long</th>\n",
       "      <th>Lat</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>2.945880e+05</td>\n",
       "      <td>294588.000000</td>\n",
       "      <td>294588.000000</td>\n",
       "      <td>294588.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>5.424562e+15</td>\n",
       "      <td>86.008036</td>\n",
       "      <td>-76.235238</td>\n",
       "      <td>40.937669</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>2.555803e+15</td>\n",
       "      <td>124.655954</td>\n",
       "      <td>20.135015</td>\n",
       "      <td>5.391695</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>1.003715e+15</td>\n",
       "      <td>0.010000</td>\n",
       "      <td>-179.392887</td>\n",
       "      <td>-68.046553</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>3.344214e+15</td>\n",
       "      <td>29.970000</td>\n",
       "      <td>-80.209708</td>\n",
       "      <td>40.487726</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>5.353426e+15</td>\n",
       "      <td>58.470000</td>\n",
       "      <td>-73.199737</td>\n",
       "      <td>42.403066</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>7.646245e+15</td>\n",
       "      <td>100.400000</td>\n",
       "      <td>-72.091933</td>\n",
       "      <td>43.180015</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>9.999757e+15</td>\n",
       "      <td>999.970000</td>\n",
       "      <td>179.917513</td>\n",
       "      <td>78.914330</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        credit_card  transaction_dollar_amount           Long            Lat\n",
       "count  2.945880e+05              294588.000000  294588.000000  294588.000000\n",
       "mean   5.424562e+15                  86.008036     -76.235238      40.937669\n",
       "std    2.555803e+15                 124.655954      20.135015       5.391695\n",
       "min    1.003715e+15                   0.010000    -179.392887     -68.046553\n",
       "25%    3.344214e+15                  29.970000     -80.209708      40.487726\n",
       "50%    5.353426e+15                  58.470000     -73.199737      42.403066\n",
       "75%    7.646245e+15                 100.400000     -72.091933      43.180015\n",
       "max    9.999757e+15                 999.970000     179.917513      78.914330"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "transaction.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Data Processing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>credit_card</th>\n",
       "      <th>date</th>\n",
       "      <th>transaction_dollar_amount</th>\n",
       "      <th>Long</th>\n",
       "      <th>Lat</th>\n",
       "      <th>city</th>\n",
       "      <th>state</th>\n",
       "      <th>zipcode</th>\n",
       "      <th>credit_card_limit</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1003715054175576</td>\n",
       "      <td>2015-09-11 00:32:40</td>\n",
       "      <td>43.78</td>\n",
       "      <td>-80.174132</td>\n",
       "      <td>40.267370</td>\n",
       "      <td>Houston</td>\n",
       "      <td>PA</td>\n",
       "      <td>15342</td>\n",
       "      <td>20000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1003715054175576</td>\n",
       "      <td>2015-10-24 22:23:08</td>\n",
       "      <td>103.15</td>\n",
       "      <td>-80.194240</td>\n",
       "      <td>40.180114</td>\n",
       "      <td>Houston</td>\n",
       "      <td>PA</td>\n",
       "      <td>15342</td>\n",
       "      <td>20000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1003715054175576</td>\n",
       "      <td>2015-10-26 18:19:36</td>\n",
       "      <td>48.55</td>\n",
       "      <td>-80.211033</td>\n",
       "      <td>40.313004</td>\n",
       "      <td>Houston</td>\n",
       "      <td>PA</td>\n",
       "      <td>15342</td>\n",
       "      <td>20000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1003715054175576</td>\n",
       "      <td>2015-10-22 19:41:10</td>\n",
       "      <td>136.18</td>\n",
       "      <td>-80.174138</td>\n",
       "      <td>40.290895</td>\n",
       "      <td>Houston</td>\n",
       "      <td>PA</td>\n",
       "      <td>15342</td>\n",
       "      <td>20000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1003715054175576</td>\n",
       "      <td>2015-10-26 20:08:22</td>\n",
       "      <td>71.82</td>\n",
       "      <td>-80.238720</td>\n",
       "      <td>40.166719</td>\n",
       "      <td>Houston</td>\n",
       "      <td>PA</td>\n",
       "      <td>15342</td>\n",
       "      <td>20000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        credit_card                date  transaction_dollar_amount       Long  \\\n",
       "0  1003715054175576 2015-09-11 00:32:40                      43.78 -80.174132   \n",
       "1  1003715054175576 2015-10-24 22:23:08                     103.15 -80.194240   \n",
       "2  1003715054175576 2015-10-26 18:19:36                      48.55 -80.211033   \n",
       "3  1003715054175576 2015-10-22 19:41:10                     136.18 -80.174138   \n",
       "4  1003715054175576 2015-10-26 20:08:22                      71.82 -80.238720   \n",
       "\n",
       "         Lat     city state  zipcode  credit_card_limit  \n",
       "0  40.267370  Houston    PA    15342              20000  \n",
       "1  40.180114  Houston    PA    15342              20000  \n",
       "2  40.313004  Houston    PA    15342              20000  \n",
       "3  40.290895  Houston    PA    15342              20000  \n",
       "4  40.166719  Houston    PA    15342              20000  "
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = pd.merge(left=transaction, right=info, on='credit_card', how='left')\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>credit_card</th>\n",
       "      <th>transaction_dollar_amount</th>\n",
       "      <th>Long</th>\n",
       "      <th>Lat</th>\n",
       "      <th>city</th>\n",
       "      <th>state</th>\n",
       "      <th>zipcode</th>\n",
       "      <th>credit_card_limit</th>\n",
       "      <th>month</th>\n",
       "      <th>day</th>\n",
       "      <th>weekday</th>\n",
       "      <th>hour</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1003715054175576</td>\n",
       "      <td>43.78</td>\n",
       "      <td>-80.174132</td>\n",
       "      <td>40.267370</td>\n",
       "      <td>Houston</td>\n",
       "      <td>PA</td>\n",
       "      <td>15342</td>\n",
       "      <td>20000</td>\n",
       "      <td>9</td>\n",
       "      <td>11</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1003715054175576</td>\n",
       "      <td>103.15</td>\n",
       "      <td>-80.194240</td>\n",
       "      <td>40.180114</td>\n",
       "      <td>Houston</td>\n",
       "      <td>PA</td>\n",
       "      <td>15342</td>\n",
       "      <td>20000</td>\n",
       "      <td>10</td>\n",
       "      <td>24</td>\n",
       "      <td>5</td>\n",
       "      <td>22</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1003715054175576</td>\n",
       "      <td>48.55</td>\n",
       "      <td>-80.211033</td>\n",
       "      <td>40.313004</td>\n",
       "      <td>Houston</td>\n",
       "      <td>PA</td>\n",
       "      <td>15342</td>\n",
       "      <td>20000</td>\n",
       "      <td>10</td>\n",
       "      <td>26</td>\n",
       "      <td>0</td>\n",
       "      <td>18</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1003715054175576</td>\n",
       "      <td>136.18</td>\n",
       "      <td>-80.174138</td>\n",
       "      <td>40.290895</td>\n",
       "      <td>Houston</td>\n",
       "      <td>PA</td>\n",
       "      <td>15342</td>\n",
       "      <td>20000</td>\n",
       "      <td>10</td>\n",
       "      <td>22</td>\n",
       "      <td>3</td>\n",
       "      <td>19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1003715054175576</td>\n",
       "      <td>71.82</td>\n",
       "      <td>-80.238720</td>\n",
       "      <td>40.166719</td>\n",
       "      <td>Houston</td>\n",
       "      <td>PA</td>\n",
       "      <td>15342</td>\n",
       "      <td>20000</td>\n",
       "      <td>10</td>\n",
       "      <td>26</td>\n",
       "      <td>0</td>\n",
       "      <td>20</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        credit_card  transaction_dollar_amount       Long        Lat     city  \\\n",
       "0  1003715054175576                      43.78 -80.174132  40.267370  Houston   \n",
       "1  1003715054175576                     103.15 -80.194240  40.180114  Houston   \n",
       "2  1003715054175576                      48.55 -80.211033  40.313004  Houston   \n",
       "3  1003715054175576                     136.18 -80.174138  40.290895  Houston   \n",
       "4  1003715054175576                      71.82 -80.238720  40.166719  Houston   \n",
       "\n",
       "  state  zipcode  credit_card_limit  month  day  weekday  hour  \n",
       "0    PA    15342              20000      9   11        4     0  \n",
       "1    PA    15342              20000     10   24        5    22  \n",
       "2    PA    15342              20000     10   26        0    18  \n",
       "3    PA    15342              20000     10   22        3    19  \n",
       "4    PA    15342              20000     10   26        0    20  "
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# extract month, weekday, and hour information\n",
    "data['month'] = data['date'].apply(lambda x: x.month)\n",
    "data['day'] = data['date'].apply(lambda x: x.day)\n",
    "data['weekday'] = data['date'].apply(lambda x: x.dayofweek)\n",
    "data['hour'] = data['date'].apply(lambda x: x.hour)\n",
    "\n",
    "# drop useless features\n",
    "data = data.drop(labels='date', axis=1)\n",
    "\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Question 1\n",
    "Your boss wants to identify those users that in your dataset never went above the monthly credit card limit (calendar month). The goal of this is to automatically increase their limit. Can you send him the list of Ids?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "def monthly_transaction(df):\n",
    "    \"\"\" function to calculate montly total transactions \"\"\"\n",
    "    total_trans = df['transaction_dollar_amount'].sum()\n",
    "    credit_limit = df['credit_card_limit'].values[0]\n",
    "    \n",
    "    return pd.Series([total_trans, credit_limit], index=['total_transaction', 'credit_limit'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Valid users:\t 862\n"
     ]
    },
    {
     "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>credit_card</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1003715054175576</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1013870087888817</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1023820165155391</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1073931538936472</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1077622576192810</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        credit_card\n",
       "0  1003715054175576\n",
       "1  1013870087888817\n",
       "2  1023820165155391\n",
       "3  1073931538936472\n",
       "4  1077622576192810"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# find users who never exceed credit limit\n",
    "grouped = data.groupby(['credit_card', 'month']).apply(monthly_transaction)\n",
    "grouped = grouped.reset_index()\n",
    "\n",
    "grouped['exceed'] = (grouped['total_transaction'] > grouped['credit_limit']).astype(int)\n",
    "grouped_sum = grouped.groupby('credit_card')['exceed'].sum().reset_index()\n",
    "\n",
    "lists = grouped_sum[grouped_sum['exceed'] == 0][['credit_card']]\n",
    "\n",
    "print('Valid users:\\t', len(lists))\n",
    "lists.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "So, there are 862 credit cards in total that never exceed their credit limit."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Question 2\n",
    "On the other hand, she wants you to implement an algorithm that as soon as a user goes above her monthly limit, it triggers an alert so that the user can be notified about that. We assume here that at the beginning of the new month, user total money spent gets reset to zero (i.e. she pays the card fully at the end of each month). Build a function that for each day, returns a list of users who went above their credit card monthly limit on that day."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "def alert(data, month, day):\n",
    "    \"\"\" function to return the list of credit cards that exceed the limit \"\"\"\n",
    "    data = data[(data['month'] == month) & (data['day'] <= day)]\n",
    "    limit = data.groupby('credit_card')['credit_card_limit'].min().reset_index()\n",
    "    transaction = data.groupby('credit_card')['transaction_dollar_amount'].sum().reset_index()\n",
    "    \n",
    "    merged = pd.merge(left=transaction, right=limit, on='credit_card', how='left')\n",
    "    result = merged[merged['transaction_dollar_amount'] > merged['credit_card_limit']]\n",
    "    \n",
    "    return result"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>credit_card</th>\n",
       "      <th>transaction_dollar_amount</th>\n",
       "      <th>credit_card_limit</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>194</th>\n",
       "      <td>2850146878241916</td>\n",
       "      <td>4012.81</td>\n",
       "      <td>4000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>319</th>\n",
       "      <td>3936887050542904</td>\n",
       "      <td>5352.07</td>\n",
       "      <td>4000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>341</th>\n",
       "      <td>4118286032166087</td>\n",
       "      <td>3416.52</td>\n",
       "      <td>3000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>383</th>\n",
       "      <td>4462290711836916</td>\n",
       "      <td>3119.53</td>\n",
       "      <td>2000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>525</th>\n",
       "      <td>5723635641134781</td>\n",
       "      <td>2161.30</td>\n",
       "      <td>2000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>698</th>\n",
       "      <td>7299183791723634</td>\n",
       "      <td>7074.49</td>\n",
       "      <td>7000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>772</th>\n",
       "      <td>7850942767136368</td>\n",
       "      <td>5704.00</td>\n",
       "      <td>4000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>834</th>\n",
       "      <td>8522875529951473</td>\n",
       "      <td>5860.89</td>\n",
       "      <td>5000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>874</th>\n",
       "      <td>8972201384562696</td>\n",
       "      <td>5266.08</td>\n",
       "      <td>5000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          credit_card  transaction_dollar_amount  credit_card_limit\n",
       "194  2850146878241916                    4012.81               4000\n",
       "319  3936887050542904                    5352.07               4000\n",
       "341  4118286032166087                    3416.52               3000\n",
       "383  4462290711836916                    3119.53               2000\n",
       "525  5723635641134781                    2161.30               2000\n",
       "698  7299183791723634                    7074.49               7000\n",
       "772  7850942767136368                    5704.00               4000\n",
       "834  8522875529951473                    5860.89               5000\n",
       "874  8972201384562696                    5266.08               5000"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# example\n",
    "alert(data, month=8, day=20)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Question 3\n",
    "Finally, your boss is very concerned about frauds cause they are a huge cost for credit card companies. She wants you to implement an unsupervised algorithm that returns all transactions that seem unusual and are worth being investigated further."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>credit_card</th>\n",
       "      <th>transaction_dollar_amount</th>\n",
       "      <th>Long</th>\n",
       "      <th>Lat</th>\n",
       "      <th>city</th>\n",
       "      <th>state</th>\n",
       "      <th>zipcode</th>\n",
       "      <th>credit_card_limit</th>\n",
       "      <th>month</th>\n",
       "      <th>day</th>\n",
       "      <th>weekday</th>\n",
       "      <th>hour</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1003715054175576</td>\n",
       "      <td>43.78</td>\n",
       "      <td>-80.174132</td>\n",
       "      <td>40.267370</td>\n",
       "      <td>Houston</td>\n",
       "      <td>PA</td>\n",
       "      <td>15342</td>\n",
       "      <td>20000</td>\n",
       "      <td>9</td>\n",
       "      <td>11</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1003715054175576</td>\n",
       "      <td>103.15</td>\n",
       "      <td>-80.194240</td>\n",
       "      <td>40.180114</td>\n",
       "      <td>Houston</td>\n",
       "      <td>PA</td>\n",
       "      <td>15342</td>\n",
       "      <td>20000</td>\n",
       "      <td>10</td>\n",
       "      <td>24</td>\n",
       "      <td>5</td>\n",
       "      <td>22</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1003715054175576</td>\n",
       "      <td>48.55</td>\n",
       "      <td>-80.211033</td>\n",
       "      <td>40.313004</td>\n",
       "      <td>Houston</td>\n",
       "      <td>PA</td>\n",
       "      <td>15342</td>\n",
       "      <td>20000</td>\n",
       "      <td>10</td>\n",
       "      <td>26</td>\n",
       "      <td>0</td>\n",
       "      <td>18</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1003715054175576</td>\n",
       "      <td>136.18</td>\n",
       "      <td>-80.174138</td>\n",
       "      <td>40.290895</td>\n",
       "      <td>Houston</td>\n",
       "      <td>PA</td>\n",
       "      <td>15342</td>\n",
       "      <td>20000</td>\n",
       "      <td>10</td>\n",
       "      <td>22</td>\n",
       "      <td>3</td>\n",
       "      <td>19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1003715054175576</td>\n",
       "      <td>71.82</td>\n",
       "      <td>-80.238720</td>\n",
       "      <td>40.166719</td>\n",
       "      <td>Houston</td>\n",
       "      <td>PA</td>\n",
       "      <td>15342</td>\n",
       "      <td>20000</td>\n",
       "      <td>10</td>\n",
       "      <td>26</td>\n",
       "      <td>0</td>\n",
       "      <td>20</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        credit_card  transaction_dollar_amount       Long        Lat     city  \\\n",
       "0  1003715054175576                      43.78 -80.174132  40.267370  Houston   \n",
       "1  1003715054175576                     103.15 -80.194240  40.180114  Houston   \n",
       "2  1003715054175576                      48.55 -80.211033  40.313004  Houston   \n",
       "3  1003715054175576                     136.18 -80.174138  40.290895  Houston   \n",
       "4  1003715054175576                      71.82 -80.238720  40.166719  Houston   \n",
       "\n",
       "  state  zipcode  credit_card_limit  month  day  weekday  hour  \n",
       "0    PA    15342              20000      9   11        4     0  \n",
       "1    PA    15342              20000     10   24        5    22  \n",
       "2    PA    15342              20000     10   26        0    18  \n",
       "3    PA    15342              20000     10   22        3    19  \n",
       "4    PA    15342              20000     10   26        0    20  "
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There are some features that cannot be directly used, such as `Long`, `Lat`, `city`, `state`, `zipcode`. However, these information is important towards anomaly detection. One idea is to replace `city`, `state`, `zipcode` with the user's home latitude and longitude. In this work, to simplify the work, I use the median of `Long` and `Lat` as the indicator of user's home address."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "# calculate the median for Longitude and Latitude\n",
    "address = data.groupby('credit_card')['Long', 'Lat'].mean().reset_index()\n",
    "address = address.rename(columns={'Long': 'Long_median', 'Lat': 'Lat_median'})\n",
    "data = pd.merge(left=data, right=address, on='credit_card', how='left')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "# calculate the distance\n",
    "data['distance'] = np.sqrt((data['Long'] - data['Long_median'])**2 + (data['Lat'] - data['Lat_median'])**2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For transaction and credit limit information, two more useful information are extracted:\n",
    "* ratio of transaction and transaction median\n",
    "* ratio of transaction and credit limit"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "# calculate user transaction median\n",
    "median = data.groupby('credit_card')['transaction_dollar_amount'].median().reset_index()\n",
    "median = median.rename(columns={'transaction_dollar_amount': 'transaction_median'})\n",
    "data = pd.merge(left=data, right=median, on='credit_card', how='left')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>credit_card</th>\n",
       "      <th>transaction_dollar_amount</th>\n",
       "      <th>Long</th>\n",
       "      <th>Lat</th>\n",
       "      <th>city</th>\n",
       "      <th>state</th>\n",
       "      <th>zipcode</th>\n",
       "      <th>credit_card_limit</th>\n",
       "      <th>month</th>\n",
       "      <th>day</th>\n",
       "      <th>weekday</th>\n",
       "      <th>hour</th>\n",
       "      <th>Long_median</th>\n",
       "      <th>Lat_median</th>\n",
       "      <th>distance</th>\n",
       "      <th>transaction_median</th>\n",
       "      <th>transaction_to_median</th>\n",
       "      <th>transaction_to_limit</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1003715054175576</td>\n",
       "      <td>43.78</td>\n",
       "      <td>-80.174132</td>\n",
       "      <td>40.267370</td>\n",
       "      <td>Houston</td>\n",
       "      <td>PA</td>\n",
       "      <td>15342</td>\n",
       "      <td>20000</td>\n",
       "      <td>9</td>\n",
       "      <td>11</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>-77.583829</td>\n",
       "      <td>39.610299</td>\n",
       "      <td>2.672342</td>\n",
       "      <td>93.73</td>\n",
       "      <td>0.467086</td>\n",
       "      <td>0.002189</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1003715054175576</td>\n",
       "      <td>103.15</td>\n",
       "      <td>-80.194240</td>\n",
       "      <td>40.180114</td>\n",
       "      <td>Houston</td>\n",
       "      <td>PA</td>\n",
       "      <td>15342</td>\n",
       "      <td>20000</td>\n",
       "      <td>10</td>\n",
       "      <td>24</td>\n",
       "      <td>5</td>\n",
       "      <td>22</td>\n",
       "      <td>-77.583829</td>\n",
       "      <td>39.610299</td>\n",
       "      <td>2.671878</td>\n",
       "      <td>93.73</td>\n",
       "      <td>1.100501</td>\n",
       "      <td>0.005158</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1003715054175576</td>\n",
       "      <td>48.55</td>\n",
       "      <td>-80.211033</td>\n",
       "      <td>40.313004</td>\n",
       "      <td>Houston</td>\n",
       "      <td>PA</td>\n",
       "      <td>15342</td>\n",
       "      <td>20000</td>\n",
       "      <td>10</td>\n",
       "      <td>26</td>\n",
       "      <td>0</td>\n",
       "      <td>18</td>\n",
       "      <td>-77.583829</td>\n",
       "      <td>39.610299</td>\n",
       "      <td>2.719557</td>\n",
       "      <td>93.73</td>\n",
       "      <td>0.517977</td>\n",
       "      <td>0.002427</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1003715054175576</td>\n",
       "      <td>136.18</td>\n",
       "      <td>-80.174138</td>\n",
       "      <td>40.290895</td>\n",
       "      <td>Houston</td>\n",
       "      <td>PA</td>\n",
       "      <td>15342</td>\n",
       "      <td>20000</td>\n",
       "      <td>10</td>\n",
       "      <td>22</td>\n",
       "      <td>3</td>\n",
       "      <td>19</td>\n",
       "      <td>-77.583829</td>\n",
       "      <td>39.610299</td>\n",
       "      <td>2.678229</td>\n",
       "      <td>93.73</td>\n",
       "      <td>1.452897</td>\n",
       "      <td>0.006809</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1003715054175576</td>\n",
       "      <td>71.82</td>\n",
       "      <td>-80.238720</td>\n",
       "      <td>40.166719</td>\n",
       "      <td>Houston</td>\n",
       "      <td>PA</td>\n",
       "      <td>15342</td>\n",
       "      <td>20000</td>\n",
       "      <td>10</td>\n",
       "      <td>26</td>\n",
       "      <td>0</td>\n",
       "      <td>20</td>\n",
       "      <td>-77.583829</td>\n",
       "      <td>39.610299</td>\n",
       "      <td>2.712572</td>\n",
       "      <td>93.73</td>\n",
       "      <td>0.766243</td>\n",
       "      <td>0.003591</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        credit_card  transaction_dollar_amount       Long        Lat     city  \\\n",
       "0  1003715054175576                      43.78 -80.174132  40.267370  Houston   \n",
       "1  1003715054175576                     103.15 -80.194240  40.180114  Houston   \n",
       "2  1003715054175576                      48.55 -80.211033  40.313004  Houston   \n",
       "3  1003715054175576                     136.18 -80.174138  40.290895  Houston   \n",
       "4  1003715054175576                      71.82 -80.238720  40.166719  Houston   \n",
       "\n",
       "  state  zipcode  credit_card_limit  month  day  weekday  hour  Long_median  \\\n",
       "0    PA    15342              20000      9   11        4     0   -77.583829   \n",
       "1    PA    15342              20000     10   24        5    22   -77.583829   \n",
       "2    PA    15342              20000     10   26        0    18   -77.583829   \n",
       "3    PA    15342              20000     10   22        3    19   -77.583829   \n",
       "4    PA    15342              20000     10   26        0    20   -77.583829   \n",
       "\n",
       "   Lat_median  distance  transaction_median  transaction_to_median  \\\n",
       "0   39.610299  2.672342               93.73               0.467086   \n",
       "1   39.610299  2.671878               93.73               1.100501   \n",
       "2   39.610299  2.719557               93.73               0.517977   \n",
       "3   39.610299  2.678229               93.73               1.452897   \n",
       "4   39.610299  2.712572               93.73               0.766243   \n",
       "\n",
       "   transaction_to_limit  \n",
       "0              0.002189  \n",
       "1              0.005158  \n",
       "2              0.002427  \n",
       "3              0.006809  \n",
       "4              0.003591  "
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# extract useful information\n",
    "data['transaction_to_median'] = data['transaction_dollar_amount'] / data['transaction_median']\n",
    "data['transaction_to_limit'] = data['transaction_dollar_amount'] / data['credit_card_limit']\n",
    "\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "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>credit_card</th>\n",
       "      <th>weekday</th>\n",
       "      <th>hour</th>\n",
       "      <th>distance</th>\n",
       "      <th>transaction_to_median</th>\n",
       "      <th>transaction_to_limit</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1003715054175576</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>2.672342</td>\n",
       "      <td>0.467086</td>\n",
       "      <td>0.002189</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1003715054175576</td>\n",
       "      <td>5</td>\n",
       "      <td>22</td>\n",
       "      <td>2.671878</td>\n",
       "      <td>1.100501</td>\n",
       "      <td>0.005158</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1003715054175576</td>\n",
       "      <td>0</td>\n",
       "      <td>18</td>\n",
       "      <td>2.719557</td>\n",
       "      <td>0.517977</td>\n",
       "      <td>0.002427</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1003715054175576</td>\n",
       "      <td>3</td>\n",
       "      <td>19</td>\n",
       "      <td>2.678229</td>\n",
       "      <td>1.452897</td>\n",
       "      <td>0.006809</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1003715054175576</td>\n",
       "      <td>0</td>\n",
       "      <td>20</td>\n",
       "      <td>2.712572</td>\n",
       "      <td>0.766243</td>\n",
       "      <td>0.003591</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        credit_card  weekday  hour  distance  transaction_to_median  \\\n",
       "0  1003715054175576        4     0  2.672342               0.467086   \n",
       "1  1003715054175576        5    22  2.671878               1.100501   \n",
       "2  1003715054175576        0    18  2.719557               0.517977   \n",
       "3  1003715054175576        3    19  2.678229               1.452897   \n",
       "4  1003715054175576        0    20  2.712572               0.766243   \n",
       "\n",
       "   transaction_to_limit  \n",
       "0              0.002189  \n",
       "1              0.005158  \n",
       "2              0.002427  \n",
       "3              0.006809  \n",
       "4              0.003591  "
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# drop useless features\n",
    "labels = ['transaction_dollar_amount', 'Long', 'Lat', 'city', 'state', 'zipcode', 'credit_card_limit', \n",
    "          'month', 'day', 'Long_median', 'Lat_median', 'transaction_median']\n",
    "data = data.drop(labels=labels, axis=1)\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now, let's first visualize the data in 2D throught PCA."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "# normalize the data\n",
    "x = data[['weekday', 'hour', 'distance', 'transaction_to_median', 'transaction_to_limit']]\n",
    "# x = data[['distance', 'transaction_to_median', 'transaction_to_limit']]\n",
    "scaler = StandardScaler()\n",
    "norm_x = scaler.fit_transform(x)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# apply PCA\n",
    "pca = PCA(n_components=2, random_state=42)\n",
    "pca_x = pca.fit_transform(norm_x)\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(10, 8))\n",
    "ax.plot(pca_x[:, 0], pca_x[:, 1], '.', markersize=1)\n",
    "ax.set_xlabel('PCA Component 1', fontsize=12)\n",
    "ax.set_ylabel('PCA Component 2', fontsize=12)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There are different methods for unsupervised anomaly detection, such as Isolation Forest, One-Class SVM, DBSCAN and so on.\n",
    "\n",
    "Here, for simplicity, only DBSCAN method with default parameters is implemented."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "# build isolation forest model\n",
    "forest = IsolationForest(n_estimators=100, max_samples='auto', contamination=0.1, \n",
    "                         max_features=1.0, bootstrap=False, n_jobs=-1, random_state=42)\n",
    "forest = forest.fit(norm_x)\n",
    "score = forest.decision_function(norm_x)\n",
    "label = forest.predict(norm_x)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABQgAAAHwCAYAAADq7oIJAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvhp/UCwAAIABJREFUeJzs3X10HNWZ5/Ff2ZZfJLABS4ZgMDaYxbxtcLBJwDiYROZFJkoyMEOUkBDBLOzG4uzEwcpmz3glL5mzE5nxLhN5kzADPjDJOM4OkHHAeIiSsWEgBkQgCQFr4mAMOA6oDYEgATZ27R8Pl6oudbeq37vV3885faq76tatW9XCXD269z6e7/sCAAAAAAAAUJvGlbsBAAAAAAAAAMqHACEAAAAAAABQwwgQAgAAAAAAADWMACEAAAAAAABQwwgQAgAAAAAAADWMACEAAAAAAABQwwgQAgAAAAAAADWMACEAAAAAAABQwwgQAgAAAAAAADVsQrkbkI/GxkZ/9uzZ5W4GAABARXviiScSvu83lbsdhUZfEAAAILO4/cCqDhDOnj1b/f395W4GAABARfM8b3e521AM9AUBAAAyi9sPZIoxAAAAAAAAUMMIEAIAAAAAAAA1jAAhAAAAAAAAUMMIEAIAAAAAAAA1jAAhAAAAAAAAUMOqOosxAACoTm+88YZeeeUVHThwoNxNqXp1dXWaMWOGpk6dWu6mAACAGkS/rnwK2Q8kQAgAAErqjTfe0Msvv6yZM2dqypQp8jyv3E2qWr7v66233tKePXskiSAhAAAoKfp15VPofiBTjAEAQEm98sormjlzpurr6+lE5snzPNXX12vmzJl65ZVXyt0cAABQY+jXlU+h+4EECAEAQEkdOHBAU6ZMKXczxpQpU6YwrQcAAJQc/bryK1Q/kAAhAAAoOf7CXFg8TwAAUC70Q8qrUM+fACEAAAAAAABQwwgQAgAAFMEXv/hFXXbZZWk/AwAAYGx6/vnn5Xme+vv7U36uRAQIAQAAYsg3wHfLLbfou9/9bgFbBAAAgFzt2bNH1113nY477jhNnDhRM2fO1H/6T/9JL730Ulb1LFmyRB0dHRnLHH/88dq7d6/OOuusfJpcVAQIAQAASmDatGk64ogj8qqDRCQAAAD527VrlxYsWKCnn35ad9xxh3bu3Knvfve7+vWvf62FCxfq+eefL+j1xo8fr2OOOUYTJkzIuY79+/cXsEUjESAEAADIkhtNeMstt2jmzJk68sgj1d7eruHh4VHPcXzfV09Pj0466SRNmTJFZ555ZtIIQzcVZcOGDfrYxz6mKVOm6Dvf+U5R7wsAAKAWLF++XOPGjVNfX58+/vGPa9asWbrwwgvV19encePGafny5ZJSjw4M9+m++MUvatu2bVq3bp08z5PneSmDi6mmGD/zzDNatmyZDj/8cM2YMUNtbW36/e9/P+I63/jGN3TcccfpuOOOK8KTCBAgBAAAyMFDDz2kp59+Wn19fdq4caPuuece3XLLLbHP/8u//EvddtttWrdunZ555hl97Wtf0/XXX6/77rsvqdzXvvY1felLX9IzzzyjT33qU4W+DQAAgJry6quvasuWLVq+fLnq6+uTjtXX1+tLX/qS7r//fr322muj1nXLLbfo3HPPVXt7u/bu3au9e/fq+OOPH/W8vXv36qMf/ajOOOMMPfbYY+rr69Obb76p1tZWHTp06P1y27Zt0y9/+Utt2bJFP/nJT7K/2SzkPrYRAACgjBIJaf16qb1damws/fWnTp2qb33rW5owYYJOPfVU/emf/ql+8pOf6Gtf+9qo5w4NDWnt2rV64IEHtHjxYknSnDlz9Nhjj2ndunVatmzZ+2VvuOEGXXHFFUW7DwAAgLIrYcfuN7/5jXzf16mnnpry+GmnnSbf9/Wb3/xm1LqmTZumiRMnqr6+Xsccc0zsNnzrW9/SBz/4QX3jG994f9+dd96po446Sv39/TrnnHMkSZMnT9btt9+uSZMmxa47VwQIAQBAVVq/XurstPcrV5b++qeddlrSOjLHHnusHn300VjnPvPMM3r77bd1ySWXyPO89/cfOHBAs2fPTiq7YMGCgrQXAACgYpWhYxfug4X5vp/xeCE88cQTevDBB3XYYYeNOPbb3/72/QDhGWecUZLgoESAEAAAVKn29uRtqdXV1SV99jwvaUpIJq7cj370I82aNStjvQ0NDXm0EgAAoAqUsGN38skny/M8/frXv065fMuzzz4rz/N00kknady4ce8HDJ1CJI07dOiQli1bpptvvnnEsaOPPvr996XsBxIgBAAAVamxsTwjBwvhtNNO06RJk7R792597GMfK3dzAAAAyquEHbujjjpKF198sf7v//2/+vKXv5y0DuHw8LDWrVunSy+9VEcddZSampq0d+/epPN/8YtfJM34mDhxog4ePJhVGz70oQ/pBz/4gU444YQRfxwuF5KUAAAAlNjhhx+uG2+8UTfeeKNuv/127dy5U0899ZS+/e1v69Zbby138wAAAMa03t5evfvuu2pubtZPf/pTvfjii9q6dauWLl0q3/fV29srSfrYxz6m+++/X5s2bdLAwIBWrFihF198Mamu2bNn67HHHtPzzz+vRCIRa0bJ8uXL9frrr+vKK6/Uo48+queee059fX267rrr9Mc//rEo9zwaAoQAAABlcNNNN6m7u1s333yzTj/9dC1dulR33XWX5syZU+6mAQAAjGknnXSS+vv7dfrpp+vzn/+8TjzxRH32s5/Vqaeeqscff/z9/tg111zz/mvRokU67LDD9OlPfzqprhtvvFETJ07UaaedpqamJr3wwgujXv/YY4/Vww8/rHHjxumSSy7R6aefruXLl2vSpEklW3MwyovOpa4mCxYs8Pv7+8vdDAAAkIVnn302bdY45C7Tc/U87wnf98dcthP6ggAAlBf9uspQiH4gIwgBAAAAAACAGkaAEAAAAAAAAKhhBAgBAAAAAACAGkaAEAAAAAAAAKhhBAgBAEDJVXOStErE8wQAAOVCP6S8CvX8CRACAICSqqur01tvvVXuZowpb731lurq6srdDAAAUGPo15VfofqBBAgBAEBJzZgxQ3v27NHw8DB/cc6T7/saHh7Wnj17NGPGjHI3BwAA1Bj6deVT6H7ghAK0CQAAILapU6dKkn73u9/pwIEDZW5N9aurq9PRRx/9/nMFAAAoFfp15VXIfiABQgAAUHJTp04loAUAADAG0K8bG5hiDAAAAAAAANQwAoQAAAAAAABADSNACAAAAAAAANQwAoQAAAAAAABADSNACAAAAAAAANQwAoQAAAAAAABADSNACAAAAAAAANQwAoQAAAAAAABADSNACAAAAAAAANQwAoQAAAAAAABADSNACAAAAAAAANSwigkQep432fO8xzzP+4Xneb/2PG91udsEAACA0qAvCAAAUD4Tyt2AkHckfcz3/Tc9z6uT9G+e593v+/72cjcMAAAARUdfEAAAoEwqJkDo+74v6c33Pta99/LL1yIAAACUCn1BAACA8qmYKcaS5HneeM/znpL0iqQf+77/aIoy13me1+95Xv/g4GDpGwkAAICioC8IAABQHhUVIPR9/6Dv+2dJOk7SOZ7nnZGizK2+7y/wfX9BU1NT6RsJAACAoqAvCAAAUB4VFSB0fN//g6Stki4pc1MAAABQYvQFAQAASqtiAoSe5zV5nnfEe++nSGqWtKO8rQIAAEAp0BcEAAAon4pJUiLpA5Lu8DxvvCxw+QPf9+8tc5sAAABQGvQFAQAAyqRiAoS+7/9S0vxytwMAAAClR18QAACgfCpmijEAAAAAAACA0iNACAAAAAAAANQwAoQAAAAAAABADSNACAAAAAAAANQwAoQAAAAAAABADSNACAAAAAAAANQwAoQAAAAAAABADSNACAAAAAAAANQwAoQAAAAAAABADSNACAAAABRTIiGtWSMNDNg2kUhfxh2Lfo5Tf5yy2bQ3n/oSCam7217Z1jPa9XNpXz7PN1f5XCPXc0txX5WknPdbLc+6WtoJoOxqLkDoefFe06bxbygAAMCYEbcTmO3rxBOl44+XFi6ULrtMqq+XLrrI9nd0WFCwrU3q7LTjnZ1Sb+/IYNXVV9uxq6+2c6Kfr71WOvJIu8727cnn9/YG9W7fbtc++WRry9q10gUXSPPn2/7t24Nrdnfbee7V3W3HFy+2z+vXjwwuDAxIy5bZ1hkYsGssWWLv3f2sXm2v9evjf0/hZxF+TuF2rF8ftC+u6Dlx6hgtqDha4CXdNeIEbOLeY7SuXJ5NNYveb6pnW6wAWbU8a9fO6L87ABAxodwNqFRvvGH/lq5cWe6WAAAAoGLt2mXbl14K9v34x7Zdt86O9/XZ5507pZYWe9/ZaduVK63TuXmzNG+ebaXUnyWpv19qb5d27AjOD2tvD9rkjh86lHz82WftmqtXj7yfjRut7nnzrKwLLri6VqwI2nLffbZdsUJ68MHg/ZIlVqa5WVq0yOqJyz2L6HMKv3f1ZVNv9Jw4dUTvfbTPo10zXb1x2hu3jbk8m2oWvd9UzzbO8y7EtSuVa9/QUHGeA4AxgwBhGlOnVv6/9QAAACizOXOkAwekY46Rjj5a+ulPpfPPD4KBN9xgo/6Gh21EX0eHndfQMDLA0Noqbdpk2yVLkj8fc4x0993S3LnSN78pPfRQcF5HR1DfJZdIn/2sNH68tGeP9PWvS//8z/bX79dfD0Y7tbdbwGB4OLiX+no7/6abbORhY+PINq5dm7x1799800ZUrl0rTZ8enNPYmN3zjF4v/JzCdWYb4IieE6eO0YKKowWI0l0jTmAp7j1G68rl2VSz6P2merbFCuRVy7N27UwkRv73BAAhnu/75W5DzhYsWOD39/eXuxkAAAAVzfO8J3zfX1DudhRaTn1BN0U1l+BVIesAAAAogbj9QEYQAgAAYOxzQb3BQVuHa3DQRsyFR/Y1Ngbl3Oi9VEHAahk5BAAAEBMBQgAAAIx9bh2y5mb7/E//lLxWn2RJOly5rVttLbytW6U77mCkIDDWMTLY8ByAmlVzAcKTT7YlYUZz+eXWbwQAAMAYsGqVbV3CkGhw0GXcdcLJQZqaUtfpeZLv23p/PT3Sc89JP/qRNHOm9D/+h/Q3f2PZgzs77RftgQHpuuukd9+1/Y2NySMXe3ttROPwsPT009I55wTnOuFybuTjzp1BQpG5c5N/uQ+XTyc8glKydnZ0JLc9fP1UIyzddaTkusLBBin9udkEJDKN8oxTV7RMnFGjmZ7/aG1O92yKqVaCPKnuM9d7L1Yyk3LhOQDIlu/7Vfs6++yz/WxZLy7eCwAAYCyQ1O9XQN+t0K+s+oLZdAJzeY0bl/y5qSl439NjbWhpGXmeO9bTk7ped9yJluvp8f158+z9vHnB8dHqzXSdcDvTXd+VSXWd8Dnh9ox2bvRa6aSrJ25d6Z5RqvrSnZvu+Yx2Ttx7zFe2z7RapbrPXO99cNDOGRwsbBvLhecA4D1x+4E1N4Jw7tz4IwgBAAAwRkyeLL39dmHrjDuCMJwB+M03k0cQhjOsuqzC4RGE0Yyj4XL19fZ58eLkEYSuXLR8Oq4eZ+1aaf/+5LaHry8FmZaj1wmXibbDSXVutEwm6doQt650101VX6pzo88/TntTPZtiKlbm3kpTyKzFY21tUZ4DgCyRxRgAAGCMI4txSCIhtbXZVOO5c6WLL5bWrZO6umwNQsmSmHR2WtCPX5QBAEAVI4sxAAAAENXYKJ1yigUId+6Ujj02CAImEiNH9QEAANQAAoQAAACoLU8/Hbx/8EHpxRctacnDD9u02uFhS9TR2pp/godaSRYBAACq2rhyN6DUrrjClosZ7bVlS7lbCgAAgIKZPj3o6G3blnzMZTTu67PpxevW2ft580Z2EsePt3XqTjxRuuoqaeFC6ZhjpMMPl447zjqRnZ3S0qUWZHQZQdevH9mmRMKuNzBg20Qi9bHubquju9v2R88Ln++ORd+7Oru7g3qcgQFp2TLbppLueKr2R6+ZqnyqY7mKtq0Y18hVnLZUUnsBADWt5kYQ3nVXvHJf+IL0yivFbQsAAABK5NVXC1PPoUNBgNEFFp0337RO5OCgfV6xwhJ+bN1qoxGjXPBw61Zp82bb56Y7pzomSQ0Ntg0fi5bp7Bz5fuVKq3P16qAed60VK4Lz77tvZDvTHU/XxvA1U91vqmO5iratGNfIVZy2VFJ7AQA1reYChJdfHi9IeOedxW8LAAAASqRQWYw9T5ozxzIWz59vIwpffNGy1E6bJv3930s//an05JMWHNy0yQJYS5aMDADFycbb2mqjFFNlzXXnZTo//D5dNt21a5O3UemOx2l/qvKFXNsx2rZKWj8yl4zKAACUSc1nMWZZGAAAMNaRxVjW6fvc56QHHohX/oMflC66KFiP8KabpIceCgI5cTuQdDYBAEAZxe0H1nyAEAAAYKwjQKggUNfaaqP62tuD6Z3z5tn7hx5KPk5ADwAAVLm4/cCam2IMAACAGuSCgUNDtv7evn3S7t3S3Ll2bO5cCxBOn85acAAAoOYQIAQAAMDY19pq04u3brUkI6tXB+vxrVghvfaatGOHfSZAWHuYCg4AqHHjyt2AUrvqKltbOtNrwgRp+/ZytxQAAAAFc9FFUl9fkIHYBQcl6Wc/C4KDnZ2pO4hHHSV94hPS4YdLhx0mHX+8dRgHBqSlS6WODqm72z6vWWMBJyeRSN6X6XP0WDbcuanakIt0bcl2fzHakG/ZKDfCdP367M8FAGAMqLkRhN/73uhlDh60Px4++2zx2wMAAIASeOGF/M5/7TXp3nuDz0ND1mE88UQLPPb12f7HH7esxVIwEtEFn9y+TJ+l5GPZcPVs3TqyDbmItjPX/fnIps58rk82YQBAjau5AOHnPjd6kHD8eP54CAAAMKacd570yCO5n3/kkdKiRRZ88337vH69bffvl045xaamtrVJS5YkB5qiwafRttH3cblzWltHtiEX6YJm2e4vRhvyLRvV2MjUcgBATSOLMQAAwBhHFuOQ8FpzktTba9uODtaeAwAAY07cfmDNrUEIAACAGuZGijU2WqBw9Wrbv359/msAFkK5rw8AAGpSzU0xBgAAQI1yowdbW6UNG6ThYQsWPvxwsIagVPh19LJRjHX8AAAARlFTU4ybmnL/Y+z3vy9deWVu5wIAAJQTU4wlnXpqkKm4GMaNkyZPli691DqcJ50UXG/RIunkk6Uvf1m67Tbp4x9Pntq8b5903XXSu+9K8+fb6MaODjve22v1Pf201blzp/ThDwdBRBfw3LRJOvNM6T//Z+ntt60t3/62ZVp212lstLp6e6WXXpJ++lOppUW64QY739XjAqiSral4223Sv/6r9Oqr0rp1dt6TT0o33SQ99FDydO1EwrIou2OuzsWLpVWr7P5StT26bW8P2uumhKf6HBUNArt7d9dzbY1bZ7g+1659+2y7Z49lsD7uuOQp6tH6BgakFSuktWul6dOra1p7dEp+pmefTV3h80f7TjE28D0DZRO7H+j7ftW+zj77bD8btqJ0bq/6+qwuBQAAUDEk9fsV0Hcr9CurvmA+HcFCvOrqgk5lV1ewv6fH95ubk8v29Fibe3rS19fTExx359fXJ5cZPz54v3Kl77e02DZalzu/pcW28+YFx+bOTS7b1DTyWFdX8j2F63B1NjQEx7q6gv1u69rgttFnEP3c1WXvBweTv2d33NUbfVbR96mukam+np7kuqPfWar6wvca/k5TXa8cBgdTP0vfT/3cUj37HTvs/nbsSF+/+xlpaUk+N9PzL+S9oLwK+T0DyErcfiBTjGP69KfL3QIAAABUjeOOs1F6kjRnjvShD0l33SV96lNBmeZmG00zOGhTnE84QfqzP0vOxjs0JD3wgPSznwV1XXFFcqZed/7wsHTEETYS8c03pYMHg+s8+aSV2b9f6uqytt19t/Taazaq76KLbJScJG3ebOdIwdTrmTOlSZOSRxCecoqNaAw791ypoSF5BOFzz9loyunTpWuuCa7R0mKj6pYsCe7BtWW0TM9DQ6mnYoczOS9cmLwv3ftM2Y9TZYZubbXnFh5BmKnetWuD7fTp1vZ01yuHTNPaUz2bVM9+xQr7TiXpvvtS19/VZd/55s22z51byOzXTNGvXMXIcg6goGpqirHn5XadCy6Q/umfGAkNAACqE1OMJY0fLx06VJgLT51qwbjJky0g9uijtu/mm6W//3sLcl17rU3NffJJm1I6fXpuU1ylYFqwlHpaavS4JPX0WLvcdOR9+4JprqecEpwXvW42WZ7jTj8NT7E95ZTRrxun0810xcIpxLOPfsfpykvF/d74uQCAEeL2A2sqQAgAAFCLCBBGuIDa8LBUX18da8EBAADkIG4/kCnGAAAAqA1udNHQkLR6dbC/p8dGPS1aZAFDFzSUghF0bW3JyTMAYKxhBCZQ08aVuwHlsGWLzQjxvOA1bpx0+eW5ZzkGAABABUskbO24zs7k4KAkvfWW9NRTtr7emjV2fN48adYse796tbRggZ3b1iZ1d+fWaUwk7Nw45ycS1pZU5TIdy7XOQsn3GqVoYy6ybVc57qNSn12tqsbvw63huH59uVuCsGr8WUJVqqkAoQsGXnqp9M47ycd839ZpbmpKDhx6njRxogUVAQAAUKWamoJEH3Hs22eBQ+fNN23b12cBw6Ym+wtzQ4N0/PHWaZwwQTrySEvmMX++rcsWDgr29gYBxw9/2NYpXLrUyknBL4Hbt9vahi4gGf2lsLc3+VgiYSMeZ8+WzjnHjrlzBgakZcuszquvtmM9PbbPta+z0xbd7uhIDl66tnd0pG7nwICdGz6Wb4Ahm/PD7Rjtl+d8grMDA9Lixdndl7uPq68uXVAx7rMj2FA4mZ5l+Puolmfe3m7/PpQzkUi1PKtSInCLUomT6rhSX2effXaWqZ1zfzU1ZXUpAACAiiGp36+AvluhX1n1BfPpCOb6amnx/Z6e4HNXl+83N6cu5/tB2Xnzko/39CTfS1dX8rHwNaLntLQk19nSErQh2r7oudFj0Xa6usPHBgft+OBg/O8mLJvzo+2IPqdUZUcrFy6b6hnGva/BwXjtynTdbMV9dvleB4FMzzL8ffDM4+NZjZTvv6uoeXH7gaxBGENdnXTnneVuBQAAACqK50lTpkhHHSW99JJlSj78cEt+ctppltF1+nRb81CykXgdHTZC57HHpJNOkl54wcpJwaidxYulVatsXcTGxpGjedz6iOFzdu+W7r1XmjFDWrIk2O/qXrVKeugh2x/OaDx9ujQ4aBmPzzwz+Xrt7db2RMJG0UXb2doqnX66ZWp2xxobpZUrc3+m2Zwfbkf4ntOVdd/DaKOjwvcvBfe2dm38ddkaG6U77kjO3jua6HWzFffZ5XsdBDI9y/D3wTOPj2c1Ur7/rgIxkcUYAABgjCOL8Xtc9mJn9WqbWrttm31uabGgjpsK3NVlwTgW7QcAAFWKLMYAAABAmAv8SdLMmTbyr7VV2rnTRpfdcENQNhwc7Oy0fYzgAAAAY1RNJSkBAABADXOL3k+aJO3ZI736qnTjjfb+D3+Qrr/epve6IOL69fa5pcUCia4OFtAHAABjTE0FCKPZieO+GhrIYgwAAFDVmpqkdevs/TvvBPvDy+288IK0Y4e9X73aRg6ee660ebM0b551DJuabH9Tk60/eNpplsX4rLNsHby/+ztb12/hQutALl0aZBV2GYOXLrWswi5jcbhMlMu+29kZvDJl4g2X7+5OzvAbzpIczmIcze6bKgga3RfNIBzeRjMbx5Uu+BqnPZnOz6cNmerMpmwlq9Z2A4XGfwuocUwxjmF4WPrCF6RXXil3SwAAAJCTYvzC9/bb0rPP2vuXXrLtQw9Jhw7Z6MQvfMESgPT1WUBRsl8+XfkdOyzwuGNHUCY6jXn9+mBEY1hDQ+opz9Hyjz9uAU6nszO4pmRBTVfe1ZlqWnV0n/u8davVH91KlgjlvvvSP79UbU81nTtOezKdn41095mqzmzKVrJqbTdQaPy3gBpHgDCG+nqyGAMAAFS1o46yoF2hjBtnIwoPHpSmTpXmzJHeeEPatctGFp5+unTTTdLf/I00f36QkXNw0LL+3nSTBRNdxuJwmTCXfXd4ONhXX58+w2e4fH291NY2MsPv4sV2/WiW5Wj20PA56Y65DMLhbTSzcVzpspfGaU+m8/NpQ6Y6sylbyaq13UCh8d8CahxZjAEAAMY4shjLRu51dlryka1bLXPxtGnSVVfZVNi+PhtZd+WVNqKup2f0ESSJRHKG4+hnAACAMiOLMQAAAOC4ESFDQxYcbGmxEW6bNlmQcOLEYERdQ0P6ESTRIGA4iBj9DAAAUCUIEAIAAGDsc8G77duljRttWu+GDTZacM4cCxjedptNy+3oSD8CkDWqAADAGFRTWYwlm0Fy5pmZsxZPmmRLyVxxRW4J2AAAAFCBNm60rMQ7dkjnny9997u2f9cuy3C8Zo0FDGfOlA47TDr1VOmEE6Rrr7Wg4ZIl0u7d0jXXSL290nnnBZmA42buLVZW3Hyzb1Zi9s6BgSDbcqFke5/Fei6V+LwBADWtYkYQep53vKQ7JR0j6ZCkW33fv6WQ1/jEJ6R77x293P799rrrLvt86aXSY4+xlAwAAECxFL0vuHGj9JnPBJ8PHpR++9vUZV1n0GX6vf324Ni2bTYdef9+6YUXpIsushGI27bZ8euvtzUOo9zU5MFBCwwNDUnd3cnHr746yAAczYrb3h5MbZZGrnWY78jGfM4v1tqLK1YEzyObbMiZZHufxRoxykhUAECFqZgAoaR3JX3F9/2fe553uKQnPM/7se/7zxTqAnGCg6ns2mX/D+f/3QAAAEVT3L7gtdcWpBpJFhx0XnjBXk66BIAuINTcnP745s021TlVVtxwQEkaGVzKN/tmPucXK9jlsiBnmw05k2zvs1hZTcmWCgCoMBWbxdjzvH+W1Ov7/o/Tlck2i3HcEYTO+PE2HXnpUuvzMIIQAABUo2rMYlzwvuDGjdJnPysdOmTrDB46JL39tnTUUTbt+N/+TTriCOnFFy3I5/rIdXX2/t137bjvW5Bv7147Pn++1Tc8bFNhe3vm/cDDAAAgAElEQVSlU04ZeX03yq611RKjREfbjTYKL3xcqqxsyWRvBgCgYsXtB1ZkgNDzvNmSHpR0hu/7b0SOXSfpOkmaNWvW2bt37y55+wAAAKpJtQUIS9IXzBTUih4jAAYAAKpU3H5gxSUp8TzvMEl3SfqLaIdQknzfv9X3/QW+7y9oamoqfQMBAABQNEXtC0YTQ7h1AKPJRVzGYxcMjH4GAAAYYyoqQOh5Xp2sQ/g93/fvLvb1SB4GAABQOYreF3Rr5fX2WkKQ1aste3Ffn2UpLkbWXACVjV8KAUBSBSUp8TzPk3SbpGd93y/gSsTha6TeH17vOZUJE6Qf/Ui65JLCtwkAAAAl6AuGO4KrV4883tcnzZtn713m3DgmTLCMyJ4nTZliHcbdu6U9e2xB609+UrrhBmnDBgtAPPaY9LvfSU1Ndm59vfSlL0lf/aqtiTh5svTtb0s//amti/juu1bu/PMt0cqGDXbdtjZ7PzxsdXR0SPv22XbWLEuc0tsrTZ8u9fRITz4ZrI+YSNj7REL6+c+tPlf/pk12/g03SH/1V9IPfmDrLLpj7e12neuvt/UYb701dZ11dcExyYKuHR1WV6bFvV09kpV35dz5p5xi+9raUq/lGK4n1bTwVPtzXf9xtPfp1pvMJJ/p7KWYCl9p1yhEe8goDQDG9/2KeEk6X5Iv6ZeSnnrv1ZLpnLPPPtvPRrDidPavpqasLgUAAFAxJPX7FdDfy/Qqel8wn45gvq+WlszHGxpGdjxHqydaZ0/PyH0tLbY//Nn3k/elqt+1p65u5LHodTLV6Y75fvI5PT3pv6dwPeFyqe4tU12unujxVPvTlU11PJv3o7Uxm3YX+9xqvUYh2jM4aOcPDuZeB4qL7wjIS9x+YMWMIPR9/98kpRnjV14TJkh33lnuVgAAAIxdldwXzCg8gnD8eOnAAWnmTMuSHB5BuHBhYUYQLlxo121rs/duBGF7u41Y278/GEG4dq2NIBwctBGEa98bmNnebusvphpBuGRJ+hGES5YE1xkaslBdujrr6oJjkr3fv9/qcqPsUnH1uPfR88MjCF170tUTrSPd/nRl45yT6X1ra+Y2xr1WKc6t1msUoj1ujVFULkZ5AiVRkVmM41qwYIHf399f7mYAAABUtGrLYhxXVn3B7m6bWjx3rnTxxbb24Jw50rnnStu3S2edJf3yl9I//INNEb72Wum226SPf7y00x0BAMn4txXIS9x+YMWMIAQAAACKpqNDevhhW2vw4oullhYblebWyHMSCemhh6Tnn7fP2fxSykikwiIoAEDi31agRAgQAgAAYOxrbJQWLbIA4Y9+ZFNwH3lE+tznpMWLpeXLbTrw8ccnJzHp7JS2bpXuuIMgVakxrRAAgJIZV+4GlNLAgDRjhi0RM9rrssvIdA8AADBmbN9uWXXHj7fgoCT94Q821fgzn7HMvLffLv2v/2XH1q2zANX48ZbV+PLLbZpyImGvNWtSdxYTCQtqLV1q1+zuts/u3FTl09WVSa7nZVtHujIDA9KyZbYNlxsYiNeuONdub7cMzKOtL5fvsyj2+dHj+Tz3ajNW7gMAakDNjCBMJKR58+KXv+8+WztakqZNk7ZskT7ykeK0DQAAAEW2aJElDhnN/v223b07ef+DD9orOrrwvPMsMcnvficdeaT00ksWeJSkf//3IBgpSY8/bsk/brjBpjl3dwej5FxiDkm65BJp1SpL6NHZaSMXBwak66+3xCC33ipt2mTHHnjA7q2jIxjhGJ6aK0m9vbbt6LDt+vWWQGPFCgt+PvCAXUuyhCfhusKj+Nrbg3o7Omw05v790o9/HJR74AHbPzRk95dOb689y8FBu6ZrX3iUZtxpheFn2NCQ/ZTkfEcqjna+O+5Gosa5XrFGT0Z/Noo9hbuUo0BrYUp6LdwjgLKpmQDh+vW5n/v66/Zv8LPPFq49AAAAKKE4wcFcPPJI8H7PnuRjXigp89y5Fox7/HELiu3cab/gu6Dd0FAQfNy4UdqxwwJtTU0WWFmxQtq2zY6vWGGBpq1brc6+PguMuQBMOCgjBfU2NNjWBas2b7a/oPf12csJ1xXOEhuud/58O8cFFl25wcHkukbz5JNB+fB1s+GuPTSUWzAq30y4cbIgu+cdDs6Nls05nzalE/3ZKHbwrhRZj51amJJeC/cIoGxqJkDY3p78/8JsTJuWX4ARAAAAZXb44dIf/5j++Lhx2QURXfnoCMI33pCam20tw/BIwGuvtVF/Z54pXXWVTWmWglFy4SmY4fNcYGXtWguA+b69b2y0IKEbHRgOwESDMm50YrhMa6u0ZIltN2yQhodtf319crnwKL5ovU1NwfvwfYT3p9PRYQFBd/1o+7IRvrYbQZjL+bka7Xz3XYVHfo12vWIlZUgVsCtm8K6UySVKGYwsl1q4RwBl4/m+X+425GzBggV+f39/uZsBAABQ0TzPe8L3/QXlbkehZdUXTCSkq6+2UVzNzTYt95JLpJtukk4/3dZJ6+mxsp2dUleXjfbbvNn2FzLIMTBgowBTZVEGAAAooLj9wJpKUgIAAIAa1dhoAbmWFht119YmffWrNrru7LOl446zYOEdd0gXXGAj6jZvtvLZjtYZLTHDpk1W96ZN+d8XUGokHgGAMalmphg7qdZsTiTsD7m9vdL06az7CgAAMCZt2GCBuTlzpO9/P5jm+4tfBIlFfv1r255xho0izMVo64QxTTAzEjEURrGeI+vgAcCYVDMBQrccipNqPcJwluPw8VNPle65hxkgAAAAVeuEE4KMwuvWJR97442R5cNlwpmLJVt/cNIkqa7Ogojf+pb0H/+j9NRT0qWXWpZiKX4AMJFIzjTc2DgyuDMwYMfCmY1T1ZMqIOT2h9f7a2uzEYzhbLbh49GMwqnqjhuASnV/mWQKQLm6hodtvcTwfWTbrkz3VgzZXKcQbSpWIK/UAW6m5ANASdRMgDCfJCPPPmv/T7rvvsK1BwAAACXkgoOpZJvh+NAh6a237LVypX3eudOOrVtnwchMAZlo4Gb9+uRMw25fuMyKFUG2YZfZeLR6o/tdJl0pWF/RiR6PZhROVXfcAFSq+8skUwAqXFf0PrJtV7jOUoyIy+Y6hWhTsQJ5pUw8ItnPvvuO+YUMAIqmZgKE+WQxPvVU+4MVAAAAqtSsWemDhJMnS2+/Hb+u0UYQjhaQiQZu2ttHZhqObteulfbvT85sPFq90f2trdLChfa+rc2yGEczG7vj6erIlC05nVT3l0mmAJSrKzyCMHof2QbGSjUiLpvrFKJNpQ7kFYv7RYxfyACgqMhiDAAAMMaRxfg927dbwKW7W/ra16Rdu2z/ypU2Ks9NsQ0vUM2URgAAUMXIYgwAAACEfeQj0kMPSX//90FwcO5c6U/+xEalbdggXXKJHbvpJlvbLpEgaysAABjzamaKcVQ+6ywDAACgyiQSUk+PZZ5z6wVOmmTvP/c56bnnbN9f/ZX07rvSv/97UE4iayuywy8WAIAqU1MjCOfPlzzPXk1N1s9rasq878QTbR9/MAYAAKhiF15oowB37pQmTrR977xj2+ees2ChZMFBKQgOdnYGwcG/+itp9mxbg3DOHFv7bvv2YHRhImHTl7u7bf+yZbZ1+wrRoXSjGQcG4o9qDI+ADLcxXR3R8qUePRmnjanaGqe+Ut2HSzKyfn31jEDNtZ3Vcn8AgIxqagThU09lf86uXfb/u3TJ4gAAAFAFnn46eL9/v/0l2K3FHTdJyeuv20uSnn/eXu3t0o4dQRmXYXfjRtv/3HPB8TgZfEeTKiNxNhlxw21MlQE4VflSj54MZypO18Zw2dHal20W5UIIJxkpVZbkfOXazmq5v3wxKhTAGFdTAcKzzso+SDhnjnTFFcVPagYAAIAiuvxy6a677P348dLBg8GxI4+UDh2SXn45cx2u3OuvS0ccIX3wg9Jf/7Wta+g6iy5b7yWX2DqGq1ZJW7bYvkJ0KMMZiaPZe0c7J9rGVBmAU5WPvi+2cNbjdG0Mlw1vR6uvVPcRziBcqizJ+cq1ndVyf/mqlUAogJpFFmMAAIAxjizG7xkYkD71KRvRd8QR0h/+EBxrbpYWLZKGh6X6egtMbdo0crQQo4iA2sR/+wCqFFmMAQAAgLBTTrHRfj09tjZgc7Ptnz7dRvt1d9sIoYYG27dypQUCwmusuZFhBAiA2sJ/+wDGOAKEAAAAqB2NjTYCaNMmqbfXgoT79kl3321BwN7eILmEE044gfyQ0AIAgIpUUwHCgQGbOTJ9epClOPyaPduWOBkYKHdLAQAAUFADA9LSpRboc0HATZuscyhJf/u3tu/mm6Wjj5b+9/+WzjnH9rW22qjDxYstM7HrLGYb7BrLwbFM9xY+RrAVAICKVDMBwrVrpXnzpEcekV59NXWZ3bulbdusnAsaTphgfUOChgAAAFXszDOlvj4LVLmMtp2dwft33rHt0JAlK9m71zLorlljncPOTunccy2j7rx50tSp0gc+YPubmqS6OumYY6Q//3Pp2mul44+3V0eHTWdetsyCjJ2dtg0HGiV7v2yZlV2zxj53d9trYCA5+OYCbm6/K9vZaddbssTasHSp1dfZae+jHdpEIriGa2O64GcikVxP9Hg48Bc+lkhIV18dHGtvt/sPJ7SIEziNPp9w2ej1Uj0/9z6RCOpK1cEvVxA31XXzCUCH34fvt1D3F7eedG3Kth4AQPH5vl+1r7PPPtuPa9w435dyf7W0xL4UAABARZHU71dA363Qr2z6gnl1BPN9zZtn2+Zm3+/psW20g9nSklzWfQ6/7+mxsj09yfvDZdNdO1WH1tUTvW74WPSarkz0+OCgvXdbdyzc1sHB1N9NtK5Uos8nXDbd9aLPz5UJP7dc2lIMqa6bbVtSPYfo/Rbq/uLWk65N2dYDAMhZ3H7ghHIHKEtlzRrpK1/J/rzx46UPfchGIAIAAKBKzZghvfJK4eo7/HDp4MFg5KHn2To2l11moagHHrD9n/ykdNVVlgRl7VpLlNLaKq1YkdzBdO9XrbJEKq2t0sKFtq+tzUYFulF3btvaavtd2eFhez39tHTSSdILL9h1775bevLJkR3a9nYbMSlJl1wStDF8jfB2cDCoZ/r05OMugUOqc937dMkdUpWPij6faN3R86PPz71vb7dj4TqzbUsxpHtm2bQlXR3h+41+b7mK27ZM95VNPQCAovMsmFidFixY4Pf395e7GQAAABXN87wnfN9fUO52FFrWfcFEwtYflGwqbjRg5dbICwezUu0DAACoEnH7gTUzghAAAAA1zAX6Ojosa/HVVwcj+jIFDsMj4wAAAMaomgkQJhL2h997741Xfvx4W8v6Qx+y2Rm9vdZ/BAAAQBXq7pbWrZOeekq6/37ptddsKu6VV9q02b4+K/fP/yw9/7wlGJk0yZKPnH++Jf3YtMmma27aFEyJzDS6kNGHlY/vqLR43gBQsWomi3FPT/zgoGRLyjz1lHT77dZfXLGieG0DAABAkX3/+7b9x3+04KBkfwVesyYIDkrWAfzDH6Rf/Urq75d+9rPkTMZuO2OGdNpp9v6cc2zkYTTjcG9vcM6WLUGm4Wi5VBmDw1l3nVQZeqPZX6MZjtNlPo5m/E11Tvi8aDszZadN14ZU7cw1e+1o58et331HbgRppnOj30upMwKXQ6HbFs52DQCoKDUzgvDJJ/M7nyQlAAAAVWzfvsLW5/uWtEOSdu2y0YmS9Pjj0ubNI6/9hS8E5VOV6+y07cqVFjxZvdo+NzQEU5xdcGXr1uRzGxqCEVmpyrg6o/uHhuw6DzxgQdLoOeFrRtuZ7n26a0WnaYfrzWUK92jn51N/qnMTCZuW7u6nocG2+dxDIdpabIVuG0lJAKBi1UyAsLfX/nibizPOYHoxAAAAQsaNk/7Df5AWLZJ+8xtbm6axcWTG4URC+pd/kb75TWn7dssyXF8/spyUHDxx2YVTHQ9n6JWSAzjRDMfpMh+3twej5ubPly66aOQ50TZF96V7n64NYfkGikY7P279HR1BgDXTuevXW3Cwudm+83TPIBeVHDQrdNtY0xMAKlZNZTFeu1b6ylfi13/iibb8zHe+Q4AQAABUL7IYS/rzP5duu0267DLpiSdsmnFzs/TyyzYV9o03pIkTLVj0xz/aCMGDB+3cSZOkO+6w9QorTT5rurEeXHw8KwBAlYrbD6ypACEAAEAtIkAYsXSpTamdPj2YejxrlgUAOzuTA0AEhgAAQBWL2w+smSQlUQMD0rJlto2Ks64yAAAAqpSbGrJvn3TkkfbeJSxx025dh7CnJzmJRaakDamOVXICimLJ557jPt9KfK6V2CYAAGKqmTUIN26UPvOZkfuja0iHueVcduywGSkAAACoUtu3S5/7nHT44ZalWJLmzLH1AF1WY0m69VZLOOKCPJ5n27/9W+nFFy2Q6BJ63HGHjSocGJBWrLD61q2z9QPb2mzf6adb0EjKvPZapY5UzKVdhU4QkuqYlP01iv2Ms73vSv3OUZ34eQKQp5oJEF57be7n3n03AUIAAICq1tqanEVYsuzDUXv3Jn92y/G89pp0++32fsoU+yvz2WdL558v/fCHlnzkxBPt+PCwdP310rZt0q9/LXV1pU7yEP6FPhxccglEhoeDsvX1llAj+ot/IhGMboweL0TAwLVraCg5W3Km+l2SlaEhK5PNtTMlxRgtYUrce5GKE8DLNqFHJWcvLgQCVqU11n+eABRdzQQIb7st9QjCOObOLWxbAAAAUGKFnPb51lu2feEF6R//0d43NEiXXmojCOvrg8Di7t12LFWAJBoUdNv166XVq0eWb2hIParOlY0eL0TAwLVraGhkXenqb2y0tnR2pm5zJpmy3EaPZXtPxQ7gZZuht5KzFxcCAavSGus/TwCKrmYChFdeKf3FX0i//312582eLX33u0VpEgAAAErl5pulr3zF3k+bJr3+ur2fMkX6+telb31LmjdPuu++ILiXiudJLS3Sr35lAcKzzpLefFP6h3+wvyqfcIL9gt7WZiP65s9P/wt7+Bf6cHDJjcCLjiBMN6puaCi5vlT158q1K5EIRhDGqb8SgxWVFsDLtj3VphJ/Bsaysf7zBKDoyGIMAAAwxpHFOGJgQLrsMmnnTmn5cptqvHatJS9JNy1yzRobDdXSYmsPSkyfBAAAFY8sxgAAAEAq06fbNBHJgnybN1tCEcmCfW6ab3hacnu7BQc3b7ZjbrQOwUEAADAG1EyAcONGmxGS66u+3hKdFHL5GgAAAJTYwIB07rmWibi+3qbxHnWUJSHp7LTOXne3vT/2WOmYY6Qzz5Q++EHpl7+UPvtZ6Z57pEWLbApxZ2fyq7s76DC6usL73P41a5LLhT9HpTo+2jlxy+RTPt96CnW9dHUVsv5CqLT2AAAQUjNrEOaTxViytahvv92WpmFpBwAAgCqzdq114iZNCpKMDA/bX4Lfflv62c/sddtt0quv2vEDB6SXX7aXs2FDsEbhI4+kvtbjj9v1VqywEYdScrKOaPIG93nrVpu+HB2VGE1msn69rTvokpOk65z29lqZBx6wdqca7Zguk3I+Hd649RQyiUWquoqZJCOXDL0k7QAAVLBRA4Se5x0haZGk1yT9zA8tWuh5XoOkr/i+/z+L18TCyCeLsWTrV7e1scYuAACoLWOlL6iVK6VDh4LgoOP7QTKQI48MgoPphNfvPuII6Q9/kE48UXruOds3d24QFNy8WWputtGGmZJ7tLdbcNBNX44Gj6IZjjs7pa4uqacnXue0ry91vVL6TMr5iFtPIZNYpKqrmEkycgn2kbQDAFDBMiYp8TzvdEl9kppk05F/Luly3/d3v3f8aEm/831/fAnaOkK2C1PfcIP9ITUOz5NOO82WmunsZHkZAABQvXJNUjKm+oJuBOHUqRbUk6SJEy3gd+CAdN55lnH4/vuDYJ9z+eXSn/6pJTS54AJp716prk76xjekhx6SWltthJ5kf1HetMn2bdoUf4RZ3BFp2YxcSySCzm9Hx+gjCOnwxsdzAwBUibj9wNEChJskvSvp85KmSrpF0nmSLvR9/zdV1SkEAACoUXkECMdeX/CCC6QHH5RmzpROPVW66SYL8rkpuy5LMUEfAAAwBsTtB442xfgjsg7gkKQhSX/med5aSVs9z7tQ0uv5NxUAAAAVauz1BT/8YQsQvvmmTb2VpB//2EaEPf548jRfRokBAIAaMVoW40mSkoYY+r6/QtIPJG2TdGqR2lVwV12VXxZjz5MmTLBsyAAAADVizPQF388ge+21ti7g6+/FNvfvl7Zvl66+Wlq1ytb227FDOuEEm3Lc2WkjDSdMsDUHp06VTj7Ztscea1N3BwaCbMXbt0vLltk+ybbLltn+uBlss8l2Gy6b7rxoNuVM5fLJPhxnf9xsw7k+g0pRzvZX4vMAAFS80UYQDkhaIOmZ8E7f97/sed44Sf9crIYV2ve+l38dBw9an/LKK/OvCwAAoAqMmb7g+0klbr9d+tCHbF9Dg40mvOIKac8ey/Y7e7a0c2fyuS7Q4oKKf/xjsF23zoKAbjTixo0WYJSk++4LMhk/95ztHxqy62Yalejamk1ZJ1XijPXrg4zHDQ3BdOqhIQsapqorl+zDcfa7NobvLd8MxJWYHbic7S9kfZlG0TLCFgDGlNEChPdIapN0Z/SA7/v/1fO8CZL+SzEaVmgf+ICtJ52v227Lvw4AAIAqMWb6gvr61227Y0cQwBsasu2ePbZ9992RwcE4/u3fbNvQYCMN58yx0Yjbt0tPPy0dfbQlRjn3XAuqrF5twcje3iChiUty0tFhn7dutezKq1fb+zvukPbts4Dj2rXS9OkWnGlttfsYGpIuucTWUGxtDdqWSEiDg5aEZf9+6Yc/HNl+F+gJ1zUwkNymcAAonI3XJUJxmaC7upKz9LrrNzdb/a+9Js2bFzyHaH0DA3aPq1ZZlubWVhsN54JQqYJS+WQHdtdbu1Y65ZTUZXIJhLW3B88ykch8XqGzG49WXzb3E81yHT4vGois5IBhJbet0vCsgNrl+37Vvs4++2w/rh07fN/S1GX3Ouww31++3Pe7unx/cDD25QAAACqGpH6/AvpuhX5l0xfMqSOYz6unx/fnzRu5v7k5eN/Skrx15/X02PuuruBYT09yeVcmXD5c1nHHoq+WlqBzm6muaH1R0fqjZcPHw/fQ3Gyfox3s8D1G25bqc76i18t0j9les9BtLZRs2jU4GHxP0fPCx7Ktt9QquW2VhmcFjDlx+4GjjSAcM045xf742dub3Xn9/en/mAgAAIAqccQRNoqvUM46S/rVryz0NXmyjaD7wAds9OCiRTb6ZvFi6fOflz74QennP7fRfTfcEIzMa2uTliyxUXILF9q+8Kgv996N5nEjA90Iwmj51larL1rH0JC1b3jY2vzhD9vIr0wj8NK1KSpcf339yLLuuHsfvodUney1a5O30bYVerRd9Hqp5HrNQre1ULJpV2NjME05el74WLb1llolt63S8KyAmuVZMLE6LViwwO/v7y93MwAAACqa53lP+L6/oNztKLSs+4LhqbQbNgRBs6efls45JzloFi4/2lQ7puQBAIAKFbcfOFoWYwAAAGBs6OmxIGBzs61/t2aNJRnZts3ef/KT0tKl0pYtlnnYlW9rS84AHM4IHLZvX+Gyx1ZCJto4bchUJpfsxMW472idpXi2xb5Gts99LMn35xIAkFLNBAgvvljyvMK8xo2TDjvM+o4AAACoEo8+atuXXrJtXZ1tp0617SOPWDbiZcss8/Cdd9pU3r4+CyjOmmUJR1avtteMGdKxx9q2s1O66irbhjMDS5YIY9myIPC4ZYsFIjs6UgcaEwnp8sutrp6e1Pfi6ty+PQiEZBOQc4FO1153fmentW1gIEhCsXixfQ7XNzBg295eK9PbO/I67lhbW7Df1ZmqfPj4+vWp73u0+0olWmfca+QqkZCuvtqucfXVxQlSZbqHYt9fLgoZsItzf5X4DACgwsVag9DzvFmSXvQj85E9z/MkHe/7/guFaIznebdLukzSK77vn1GIOp0HHihcXb5vS6l84QvSK68Url4AAIBKVIq+YDH7gZKk+fOlp55K3nfggG3feCN5/6FDtn355eT9b72V/Nn3pb17g88//7ltv/e95LUGH37YgoyPP24Zfbdts85kX58d37jRsgu7zMRDQ9KDD9qxu+6ytf06Ouyzyxh8zz2WcfnRR23kolvnb/Vqe9/RYXUNDlpgZnBQamqy6dWbNlkZl0VYsgzMkpWVpE99ytp0++2W9XnFCum++4LAy6232vVXrrTMxe4eXcbl8FTrvj4LdPb0BOuaDQ0lZ8B10q1/5rIlS8G9pTo/lWKvYxi1fr0FmOfNs62bfl7IaeiZ7iHVsfA0+HA27FIttu6+r1Q/H9mK8/3F/Y5zXR6AZQUAjEVxMplIOihpRor90yUdjFNHzOt8VNKHJD0dp3w2metOPrlwSek8z/cbGnz//vtjXx4AAKBslGcW41L0BbPtB/pZ9gWLkqk40yucBXjlSvt8//1BZuO5c31/+XLbRjMTd3XZOe5YNMOwezU1Be+7uuzl3ruyLmuy27p2dXUlHxsctFf4uj09vr9jh52zY4c9x8HB5HsLX8vdWzjLrauruTn5+4hmwB1NNBtytueXkmvbjh3pMwCXWvj6cTI3F1r456aSsuOOtQzVAJBC3H5g3CzGnqRU2UwOk/R2LoHJVHzff9DzvNmFqi/skUekiy6Snnwyu/M+9znpu98tRosAAACqRtH7gsXsB0qyrMPREYS5GjdOmj3bRmJddpllSP75z21EYl2ddP750rXXBlmAOzqCUUYLFiSPPBoYCEZzhTMTNzYGU3HdPik5Y3BbWzBK0Y0wbGgYmY1406bkrct07EbihUdBuXUXw/vvuy+or7HRRoCFR/NFrxXOcnvvvcH9hUUz4I4mmmPF7r0AACAASURBVA052/NLKdy2dBmASy18/XAm6VJxPzfu56pSjLUM1QCQh4xZjD3P+9v33i6XtF7ScOjweEnnSNrv+/6igjXIOob3+mmmlnied52k6yRp1qxZZ+/evbtQlwYAABiTcs1iXOq+4Gj9wPfK5N4XdGsLStLMmdKePdKJJ9qU4K9+1ab1nniiJS755jeDoB1TCUuPKZwAABRE3H7gaCMIz3T1STpV0v7Qsf2Sfi7p5pxamCPf92+VdKskLViwIH10EwAAAPkaW33BVaukZ5+1oN+BAxYgXLLE9rs1/557zhJ/hEfNVepItbEsmzUGMbYUIzhMwBkARpUxi7Hv+xf6vn+hpDskXeo+v/e62Pf9633f/01pmpqfadMKl8XY82z2BlmMAQDAWDaW+oJKJCwQuGuX1N8v/eIXtv///b8gWUi47NKlNk14/nzr+I0bZ9N6m5qCjuXJJ9v+GTNsu2iRjUC87DLpyCNtivGWLRaEvOACCzx2dydnLg5nd82UcdjtD2cv7u62Kb4XXGDbzk7bLl0aZEweGAgyFqfLVpwqI3L4ulu2SKeeasdTtWfpUqvPZTZOdw1Xzh1P9SzcvtbWIKlJqrJxvu9wtuVsz8s22+5o58U9nm178213qcXJtF2MDMS51hnnv8lqNlbuA0BhxFmosJQvSbNVhCQlxVh7uqkp9uUBAADKRnkmKSnVK5t+oJ9lX3BEgo90rxNO8P1zzy1sh9G9d0k8wskNwskO0iU+SJVgIpzAJNN1w8lPUiU7mTcvSFYSTTLiyrm65s1L355wApSeniBhitufLrGJOx5O5BHeF/3u4iaFcNd39xb3vGIlrYh7PNdEHtWSNCNVO6P7ipGAJtc64/w3mU/95VYtPzcA8hK3HxgrSYnneZMl/VdJH5c0Q5GRh77v/8dCBCs9z9sgaYmkRs/zXpLU5fv+bYWoe9Ik6Z13ClGTqa+X7ryzcPUBAABUqlL0BYvZD5QUrD04mt277RWH51nYylm4UNq50xKRPP64NHeudOON0n//79LRR9uahm4KSjTJQTjZQTTxgUvQMTgozZkjNTdLp5xi1zrvPGnCBOnMM62DOjxso9C+8hVbR3HVKrvmypV23NU9OCjdc4+0Y4f0iU9IEycGZYeGbESRK3vmmdKXvxyMvnLtGRqyc/bvt5GW114bJEBxSUzmzrUyd98tPf209MIL1saODjv/4Yftu3HTP6P7Vq5MTlDS2mojnuJOFZ0/3zIVxk0mUaykFeHjqaa7uuPhJDKppJsqWy1JMzL9vIeT2xR6WnmudaZ7rtH91Tolvlp+bgCURpwooqTbJb0mW++lW1JX+BWnjmK8svmr8eBgbn8Mnj69+v4QBAAAEKY8RxCOhb5gQaeRpBqtt2NHvNFRuYqOoos7YinT9VPVEbe9o5UL1+3KutF8XV2Z25Dp3uK2r5JHdOXzM8GIr8pUyT9vAGpe3H5gxizGjud5r0r6M9/3+0YtXEILFizw+/v7Y5dPJGzZmLiOPNL+ePmRj+TQOAAAgAqRaxbj0PnV3xe8+GLpgQfyu+C0aVJdnfTGG7bW4GGH2fqE3/mOjehLNbqrUMkREolgVF5HR/y6sr1+3PLZ1OvKtrZKmzbl9yzGQrKJfO5hLNw/AKCk4vYD4wYIX5L0cd/3BwrRuELJNkAIAABQiwoQIBwbfcE1a2wa4LRp0uuvJx+bNcum2e7cKbW02F+J582zKbiSJcyopqmDAFANCHoDRRe3H5gxi3FIj6QVnufFLV9xGhoKm8XY82wZl7/7uyBBHAAAwBhV9X1BSTaCbfr0IDg4ebJ0xhn2/p13LDj40Y9Kp59u6/zt2GHblSttzb5rr7WRg1dcYdNSxo+3DuGZZ1qW38mTpQ98wNbx6+iwsh0d1lHs7LS15To6LBuvy1a7fbtl93X7U2UTjWY67u5Ozljssg9HM+C67MRbtthahSedZGVdfdu3J2c6TpXBOJphuLNzZDujGYozcW0Kd54zZXIeLdtxquyy7tjAQObsx+nanS6za7Ttoz2z0cQtn0um2VJl3yULLvJVjKzVAHITZx6ypB9J+oOk3ZLul7Qp/IpTRzFe5c5iLPl+XZ3/fsIxAACASqT81yCs+r6gv2OH78+Zk9yRW7LE9ydMSN7nsgNPn+77y5fbeW7tvLiv+vrUdYZf0czB0Sy9g4O2Vp97uQ5nODtweNHscJ3NzVbOfY5mUnZ1uPPmzk1eHzCchdi1J3rd8PqC4efj2rtypb26upLXZXNZesOd53SZnAcHk7MfR7Mrp1qLL5oJOV25cFtc2WjW5Og50bZHy4UzEbs1KTOtSZfveo+5rNUYfb75rptXzDURWdevvErx/MP/zlXb98zPJ6pI3H5grCzGkhKS7ilUUHKsmDBBWrdO+uEPLSkdAADAGFX9fcGrrpJ27Uret3Vr8L6uTjpwwEYRStK+fdbRW7cuXv0TJkjvvmvTTIaHbd+UKdJbb1mdc+bY9WfNsky+Lrvu4sWW5XfWLOm3v7UMyt3ddv7q1bZdudKmO2/ebJmSu7rsGsPD0r/8S9Dm+fNtmvTmzVJfn503caJ0ww3S//yf0ssv2ygdl0l53z4bCblzZzC1WrLRPF1dNq3aZQ1293TBBUGWXTfyZ+VKe3a+b+XWrEl+Ng0NwfRs12kOd57TZbZdv97upblZWrQouZzLapwu47MktbXZ80pVzrVh/35rezRrcqpzom1PlYl669YgY7bbppuaHjeDbLpymTLnxsm+W4jMu8XMglutmYHHilI8//Xr7d+5np7qm17MzyfGoFhrEFaqbNadueIK6a678r/mkUdKn/60dPzx2a0PDQAAUC75rkFYqbJag3D6dOnVV4vXmPPOkx55xN67gFZbm7Rhg+1ra7MEHekSdbj1EZ2uruT6V6+2AN4ddySfF01eIo2ezCR8TriN7vzwemCuXV1dFuhLl4DF/bLs2u0CivX1uXeaS7U2WSGvU8iELHGvles1Kn3tt0pv31hXiudfzd9xNbcdNaegSUpClS6QdJKke33fH/I8r0HSO77vv5t7U3OXS5KSCy9M/mNxOlOm2B9UN24kizEAAKhuhQoQVnVfcMsWC7BF+74TJkjjxlnHb8+e9OfX1Umf+ISt47dvn/Sv/yp97GPSUUdZECwaaMslO21bm438a262ugqdCTkXxchqDAAASqbQWYyPlq0xs1CSL+lk3/ef8zzvO5Le9n3/v+bb4FyQxRgAAGB0BchiXP19wURCuvxy6cEHUx+fOdOm4+7aZQG6H//Yzrn66iCj8Q9/aCMRixUII8gGAAAKrNBZjP+3pN9Lmi5pOLT//0m6KPvmAQAAoIpUf19w/fogOHjYYZZtOGzPnmCNwuFhy1TrgoNNTZbReMWK0TNuRjMEuwzDcbIT9/YG6+fFkSmbbz5cu6JZkeOcky4Tcbr6MmXBzVTnaG3IVrmy8ca9fyCKnw8AhRYnk4mklyWd8d77P0o68b33cyQNxamjGK9KyGK8ZIkluGtutmRlAAAAlUb5ZzGu+r6gPzjo+0ceaR24SZNSd+wOPzx47zIPz5vn+6tX+35dne9fcIHvH320748fH5QbP973p02zOhsb09e5cKGdL/n+lCn2vq7OPh9xRPK1zz3Xji9fbtk9d+ywjMALF/r+ccfZ1mUJdud89rNWrqXF93/2Mzvvmmt8f+ZM3581y/f/5m8sm/H3v2+ZN3fsCLINX3ONlTnrLN//6Efts+T7kyfbtrnZ6nTZeV32Tpep113XlY1mUw5nYXbP1NUTzgwczga6Y0fwHcyda8/CZUl236fLIOoyobrrZpNR19Xj2uiyOI+WmTR6/VyzmWbKAlzoDMHRdubT7tHOLUaG11zrLFe22TjXzVRmtPPdz0fcn9lySHUP5fjZAWpc3H5g3E7hG5L+gz+yU3iOpH1x6ijGqxIChOFXS0vs5gAAAJRMAQKEVd8X9L///SAgl+0rHBAsx8sF0aIvFxCTfL++PjkAFy07bpxtGxoy1yn5/oQJI/e5OltagsBEeJ9kAchogFWy4N68eb5///2Z6wkHwqLtc/W5AGE4eObepwo0jiYaZHFBV3cdJxq0CJ/n2ppLIC+fAFG2ogHHfAKQo51b6OBmPnUWoy2Fum4+AeJocLvU9xdHqnsox88OUOPi9gMnxBxo+KCkL0r6727goed54yV9VdJPshiwOOYsWSKdfrrNmFi7ttytAQAAKIrq7wt+5jO5n3vw4OhlPM+SnRw8aNvx46UDB6Sjj5beftumGz//vGU69jxb8/D3v7dpy9//vvTuu7bv5JOlbdukN9+UXn9dOvdcm3Z8zTXSr34l7d1r06PPOceSo0yaJD3wgPR//o/00Y9K+/dLs2bZdWfOtGQqb70lXXedTWG+8ELpjTekOXOklSut7fv2WR2vvy798Y/WlmnTpMMPl445xs75kz+Rbrop6PDefrtNu25qkm64wfZt3hxkcB4etumPzc3WUd6xQ/rmN20dxxUrrJ7p0+08l/G3tdXOaW+340NDVk99vfTXfy099JAdk0Zu3dTsTAliUq3xGK6nsdGma6fippZL9tzC19282RLguH3prpVKY2PwPWRzLBfpnl243bnWle3xXITrzGa9znRtKfaan3GeQaYy0WPR9rqfj0QiyDBeaVLdXzl+dgDEEyeKKOk0SYOSfixpv6S7JQ3I1qI5KU4dxXhl9Vdj3/5wnM0fa2fPDqYNu5kLXV2MdgYAANVF+Y8grP6+YKlH/YVHlqUa6RbeHxY+p6UlGCUYnari6oqOXgtfI3zNVCMGU43queCCYO2cdCN53H43YtCN2ks3fdVNQR5tPZ5CjnCLU38q6UbtZbufUVDFVYjnW23fUbW1F0DFiNsPjDWC0Pf9ZzzPO1PSf5H0jqTJskWp1/m+v7dAscqiu/JKe3le5nKTJkkvvZT8h6RMf1AEAAAYy8ZEX/CYY2zEXiHV11vHcupUG2W3a5d9XrTIRrJJNmpPSh5xNjwcnB8dJbN2rY0CnD/fRqzt2xeMuAtz57W22pSW8KgbN5ouXHdrq7RwYfprR0fSRfeluvbixcGowuhot/DnxkbpvvtGPL4RCjnCLU79qaQbtZftfkZBFVchnm+1fUfV1l4AVcezYGJ1WrBggd/f31/uZgAAAFQ0z/Oe8H1/QbnbUWhZ9QW7u6XVq7O7QE+PBX/WrLFgXU+P/XLushu740AlCk9JlYo7nXYsKvYUZAAokbj9wLhrEMrzvHpJZ0maIWlc+Jjv+3dn3cISGzfO5lIUus5Dh6QjjpBOPdX+eOqWRuH/IQAAYCyp9r6gEonRy3iedRiPO0765CelwUEbCbhtmzRxogUKb7/dppocdpgFD/76r6VXX02up75eeucdacYM6ZVXpMmTrb6TT7Z1Af/iL6RLLpHOPNPq7++X2tqkKVOkSy+Vjj8+GIHY3S39y79IX/+69K1v2ejCE0+0TufRR9vowT/5E2nVKumUU+zaw8PS009LZ5xhndKODmnnTunzn7e1C3ftkurqpC99SfrLv5Q+/GHpuefsHr/zHVsbsKdHevRRa/Pu3TZScMsWa5Nb529gwNY2PHDAOsO/+pWNePxv/83Ov/hiW59ww4ZgLcG2tmC9wQ0brL62NnufSFidN91k9+fWJkzVud6yRfrCF6Q777RnOTBgIy1XrRrZIXfH3ChM9/6UU0b/mQkHiBKJYERourUO3Tmp2h5uR/jamQJRqdoQN2gVXjdRSl5DMa4491xsqe65FMG76LqT5fT/2Tv38KqqM/9/dy4QEq4mEeUWg6GEIkWEoIAQdIJiwNhipxWvE7HtM5P49CdOop2WSRg70zaMOFNjZ2zFjJdKbUedogJSrCgq8VLRgm2ieAHxgjkooInKbf3++LLc6+zsc87e55ZzDu/nefazz9l77bXetc4R37Pyvu9XNisFQUgGXvKQAVSBdWeOuhxHvPSRiCPVVIzdBNgEQRAEQRD6GsRegzDtfcEgxV8vx8yZyXEgW1qUGjbM/bpZT1CrD4dzQsON4dbGrU9TXTjUGLrmXqhxTbVoZ+1D/d687vxsnOrIejyz1p+uf1hczPdOBWc3ReTqavt1VVVkheBQNRF1X27P6jZ6TqYasmlHuHG82ODlB4e5ZuFqRIbDWdMy0jiJIBol3HiQ6Hn5QeoPCoIQA179QK8RhP8J4FEA/6SUei9Oe5MZg1sEoSAIgiAIQgaR/r7g8OH+2j/7bPRj6UhE53nAACoKWxYwcSKjBWtrgd/+llGEubm086KLbIdy5073CMJNm9jPRRcBl1/O6MAdO/jMtGmM5DvnHEYj1tayXuDChbw+ZgwjGUtLgaefZsTg0aO819XFMevqGBE4ejTXYuZM4MIL2X93N6PKOjrYx0knBUcQ1tUB//ZvtK2x0a59qCMI585lhN3Bg8DGjXY03ZgxwK5ddrRiTw/XSY+3fDnnfdddjBzUEYSAHR1oRhBq9D2zjuPEiZGjw9xqIXZ3A888wxTztrbgZwMB3m9qou0bNwb352aH2zi6Lx2JGMoWE7cIM2d9RPO11+i4UDUtTRIdaReNEm48iLeKdCxI/UFBEJKBl11EAN3oQ4W6UIefvxovWRL+D6tejuJipWbMUKquTtSMBUEQBEFIHxB7BGHa+4JJiQY0o97MKLiGBr6++mper6tTQdFAWuW3stJ+JhLOiKKuLjqo+nCLNjIjopzqxs5Iv0iqyE1NoaOrvEY7aTvc7HXaZ6pCxyOKKpboMC/KxbFGn/mNGPPbPp7RcakUaScIgiD0wqsf6DWC8BkA4wG8kYA9yqRwxx08AgGguDhy+9NPB/7wBynxIAiCIAiCgAzwBVFfb9dSi5bsbNYn3L2bkXPZ2Yy2GzWK0XAXXsiaezpCTtegA+iAmnXkSkrse+PHU+XXWSsvHM6IoqIi1isE2H9BQe9oIzMiyqmCXFPDqLrnn2edwkiqyOFqoXmNdtL2uNkbSqXZFN2IhViiw7woF8cafeY3Ysxv+3hGx6VSpJ0gCIIQNZ5UjC3LWgTgxwBWAtgG4JB5Xyn1UkKsi4CoGAuCIAiCIEQmVhXjjPAFAwGmtzrTPk20At1llwF79nDT76mnbLXiRAsFxNK/iBgIgiAIguBCvFWM//fY+Zcu9xSAbK+G9RWjRgHvvhu//gYPZu3B229nuRXxxQRBEARByGDS3hdESws3B7WysBtHj/L8wAPA55/zdVkZowTr64F77gEOHGC9NV1XcOBA4MgR4PBh1vJrawPa27lht3Ur25SWAlu2sI7fW2/x2tixVAm+5x7grLNsG1esAF5+mbUBgd7OZns76w3OmcPxAFu5eMUKRiLOnRusNKzVfbUK8fz5tuoxwDbXXw/cfLN9zamCDIRWszXr5WllYt0m1MalVsc1axOuWsU1a22NrDDsHNeM1oy0URpOaThemPMDOEcvKsChFHu9KAmHW2s/1/3O08taRjsv2fgWBEFIHl7ykAGUhDu89JGII5VUjGfOZFmUjg7PJgmCIAiCICQFxF6DMO19QTVtmnfHTivkmmq8Xg/L8te+sFCpLVuoejtiRO/7OTk8l5aylmFpqXs//fsHv58zh33OmRM8J0Cp4cP92V1ZyULc+n15OW1uaaHzqxV7x4yx2+gahbq+oLano4O16rRdZp/m+uv+TdVdXV+xo4NnPa6eW3W1PZ4ev6Ojd308Xa9P13ysq3NX+42ltp6bErRbfUDnGG61BPWc3PpwqyvpbOP3ejTzjFQfMpwSsd/nhNCY/51ITUhBEI7h1Q/0FEGolNoZz03JTMQUunv00b6zQxAEQRAEId5khC/4+uve23Z18XzoUPh2bigV+t7JJ1PFeN8++9revYyO6uhwf+bwYZ7feosRgqNG8f2kScDHH7MeIgB88YX9TH4+U6MBoLISKC9n/zrSMBDgvSFDqIB8++0cJycHGDGCSsJjx1Kd+NlngSeftPseMIB9aZs3bLDTtnft4nnYMKoNP/kkFX2rq6n6CzBSrF8/2z4AyMtjX3PmAK++yvU316ShgVFky5fz/Qsv2P0VF7N9bi6vVVQwErO7m5GemzbZbXWdvJoaXv/4Y75/7DFbARqwFXnN135r7NXW2mtTWRlcR9HEqQAcrpZgVVXv6+bzoZ4Nd727m0cgEF2EXqhakaHaOetMRlJIFvVef5j/nRQUZGZtSIkqFYSE4TXFGJZlfQ3APwL4KphK8hcA/66U2pYg29IC7ZPU1wNvvumtprQgCIIgCEK6kfa+4P79yRknJ4ebbcOH84fskSOsbXjGGcC997LNBRdww2/GDOC88+yU3zFjgHXrgPffZ3rvoEFsP2ECN7Asy96sGz6cqccANw2zs9m+uxv4xS+AP/7RTtfVqc+TJgHXXUcxkgce4OZgayuwcCFw5ZUUWbnzTm7o3XUX+9Z1G8eM4Qbg4sXABx/Q3s2b6QjrTbDTTrM32558kv3o9OSWFtozZQo3OufMYSr1gQN87lvf4g/+vXvtlOjNm93FUhYv5kYgwLWrqaEd5eXBqc0FBe4bV2vWcNOwoYGbmc6x9Hhur71SVMR060gbGW5iM85Nnfp6W8TF2Y8XYZRw1wsKuMEY7WaS2Xe4591sMIV1/DwnhMbLpmu649xUFwQhbngVKakB8CCAzQCOFUTB2ceORUqphxNmYRiiESlZvx74xjfssjJujBpFP8dL6RNBEARBEIRUJw4iJenvC44bFxwl5hfL4jk7m5tKXV1AXR2j8p5+mlFxt9/ONqYSsX5tOpbhImAi3dM126Kp2RdpjFB14sLVmHM+E6munNlebwY618cvpvqzn9qFEoEkayGkH/KdFQTfePUDvW4Q/hnAQ0qpJsf1fwFwkVJqctSWxoCoGAuCIAiCIEQmDhuE6e8Lrl/PTS4vacNagARgpNsDD8Rf3EEQBEEQBCEJePUDszz29xUA97hcvweAxNkJgiAIgiBkNunvC956q/eagiecYL9+6ilbXVjX7tNpj4naHAwEgsdLRwIBpo82N/eeR7znlwnrFY5Mn5/QG/nMhXRHvsNpidcNwg8BTHW5PhXAnviZkzgsK7HHwIHA/fezHMK8ecx0EARBEARByBDS3hfEe+95b7t3L+sGalasoJNXXBzaGczNBQYPZgqyvnbKKaxdU1rK+9deC5x4ou00zprFWnrTpwMzZwLf/CY3J885h/dPPZX36uuBJUsoHFJfz2jIceOAyy5jbcOBA4HRo3m9vp6vR48GLr6Y448ezbYVFRQeGTKE98aNo539+rEeYXY2awQOGMC6gI2NtLWoiPUJS0pYD/GEE3h9xQo6vfpHYGcnaxHOmsXaisuX82hpsX8strez9mJjI+0aNYp2NzcD11zD96ecwnbmJmN7O7BgAcfo7KTDXV/Pfi64gGddz06PZdrm5ceq2SYQYJ9z5/JsPudsZ/arn3P+IAhnk3nN2T9g11xra/P+HY40v2ifddoaTb9uaxarXX6e7ey0v0vRPJ+MjY94fObxWte+2OiRzaX0J17/bgnJxYvUMYBlAPYB+CGAcwDMBfAjfc1LH4k4pk6d6kPWOfFHfr79urras2mCIAiCIAgJBcCLKgafKxN8waQ4g16PgoLYni8u9nc9lmPAAPfrubk8V1Xx3NREB9itbVWVUi0tfF1e7s3u8nL7GfO56urQ45SV8bPWz+l2LS32tZaW0N8Rs405tvM5t3b6vvmc+YMgnE3mNWf/SinV1cXXXV3ev++R5qdx69vtWihbQ/Xrxw6/z0eaUyT0HKqro3s+Fnu9Eo/PPF7rmoz5hhtfSE/i9e+WEBe8+oFeVYx/DOBTANcDuOnYtfcANAH4eUw7lBnE178OjBzJWtGiZiwIgiAIQgaR/r5geTnQ0ZH8cbOzqWQMMILv7beZ7vyLXzB9uV8/4OBBoH9/4IsvqII8fz6wZw9Vgz//nNFOeXnAE09QGXjyZOD//T/g3HOBF15g20GDgLvvBh55BPj97zne0KHA9u2MGFywAHjtNeCdd4DPPgNmz2ZU3t69jG6sqQEefJAKxzt2MHqxuhr4wx+oNlxcTBs//pgqqYcO8f7EiVT3A+gAf/opVZxLSyneMmKEraQMcNxlyxht+fjjjH684w7asns3owlzchh1UlZmK7LOnw/cdJPtZB88SEGS/HzO4amngHuOZcFr9VY3BeOaGkYmRVIEBihE8/zzjOJ0UzUOda2rq/cPgkg2VVRwroGAN2XjaHCz200R1u1aOPvd+nVi1u10to/0fLian17GdqI/l5Ur7e+ln+ejGdMv8fjMY7Ez3Hc8GSRjjYXEIgrk6YmXXUTzADAIwCC/zyXi8PNX48sui/4Pp+aRlaXUokVKNTTwD6UdHTw3NcnmuCAIgiAIqQlijCA0j3T1BeMWXecWvTZmDM91dXbUREeH3basTH0ZZafR7erq7Ged0VlmpJOJeb2pqXffzjHcnFSvUVx6Hh0ddpuqKtv5TbcokVSNTOoru7xGEMZKsqMEBUEQhC/x6gd6UjHWWJZ1KoAJx97+RSn1Zlx3K30SrYrxNdcAq1a53xs3jpGAjY0iSicIgiAIQmYQq4qx0U/6+oL33w9ccom/AQoLgX37WBdv/35g0SI6iatXM6rppZcYAfjTnwKbN/eOcNKRTzU1wJo17hFQZnQUEBwp1dkJLF3KSKfxhhaMeb2w0D26KpLSsvO+F2XmTFBvTtU5pKpd8SKW+WX62giCICQYr36gpw1Cy7IKAawCUAPgqL4M4BEAVyul9sZga9REu0FoIv+/EQRBEAQh04l1gzAjfEEtNOKVkhLgN79x3/gDUt+J1PNtaYkuzSte80v1dRIEQRCEDMerH+hVxfgOAGUAZgPIO3bMAVAK4FfRGplMJkxwF5wrLo4sShfu6NePZVtGjgR+9StGINbXi+CSIAiCIAgZRdr7gqipoQqwFwYMAHbu5KZWq+EVsgAAIABJREFUYyNw1VW91TxNhcbOTtZlmzLFrtV34olUJl6/nvcqK23V1OZmPltfb6vkdnby/bhxtlN52WWsf3fqqazRp8eZNYvPmMq+TvXcSZMYXdjebl9zqv/W1/N9e7ttU3Mz2y1ezPetrbbN9fWc09y5fEbbqxWH3VRPW1t796PH0crH8+bx2vr1dNq1qrFu4ySSirDzvR7TXPP6enuu4cYKNa4XW7w875VQ44RTF47n+F6ej4fybCJUhUURVxAEwRte8pAB9ACY4XJ9BoBuL30k4kg1FWPLsl9LiQxBEARBEFIFxK5inPa+oGpo8O7UTZ7MOnvr1gWr5+p6fwUFSt18s1IjRiiVl6dUSUnovnJygl9Pnuzebs6c8DaVltq1DvUxbJj6sgZiZaV9vaqqd83FsjK7FqLzcCoVm3UWdb1B5zNmXyNG2H0sWWKvWUtLcI1EN2Vgc6z8fJ5N290Ud3Wf5mei2zlrMjrHDFdT0unAO2vxOfvu6oqslGz2EamWnlnv0UkoVdlw6sLh+ogG8/lwSsex/BCKh6qw07ZUqu/o5348x8pUEjnv43VNhYzEqx/oVcW4C0C3y/UeAH2SUpKKKMXzjBkiuCQIgiAIQkaR/r7gAw94b/vKKzxrdV4AWLuWB0C12euvt+/t3Bm6r8OHg1/rvp288w5w9dXAnXfa1xYtooru7t3ASScBW7bwel4eoxX1+127eABU/p0yhbbn5HDMrCwqEwNUFzb7AqhqPGwYFYrnzKEy8cSJVOLduJH1D8vKqDj88st85vzz+cwLLwAffAAcPZZ5vno10NPDSMqaGkYxFhTYznF3N+/n51PR+NFHgTFjaL9+7u67GZUIuCvuNjVRQXntWqr/trSwrwULOD+T2lp7TIDn7duB005jyvPixfZca2qCn9XjdXdzDroP8/7atbTFnJ9WIy4qClYEDqfMGgiwELpW2n700d7zcD7rVEJ2qgs7iVUZ1nw+nNJxLD+EounD+YzTtr5SxHVbIz/34zlWppLIeR+vayoc13itQbgEwGUArlBKvXvs2kgAdwH4jVLqjoRaGQI/dWcsKzE2FBfTpxo8GJg/Hxg9mhkLUmJFEARBEIRUIQ41CNPeF0yYMxiOvDzg88+BQYOATz6hDdr3HjCATmRJCa/t2sWNr+3b7c3Mqipu9uXncyPr1luBxx4D7rmHabjLlzNt+oIL2CY/n44owE2zX/6S4ioA7/X08Idufj7w+uvc3MrNBbq6gCVLuBHY3c1+W1rsjaANG7hRWFXF9GaA47S2sq2eT2EhNzMLC4G9ezmf5ubQ67NgAW2orKQdU6a4KwW6ib0AwbUN582jjXPmAAsX+hNtCVWvUT+j16Spyd7sdAq7aHucbcMJ1LjZUF4O/N//BYvSpGIdx1S0SZMqtvkVCkrkWJlKIud9vK6pkJF49gO9hBkC2AbgEwCHALx97Dh07NqfzcNLf/E6fKWVGCxfHjrLYPhwpbZsiapbQRAEQRCElASxpxinvy+YqBoz+hg8mOfcXJ51mm15uVJ1dXw9YwbPOjW4uprpa2bqqn5fVWX37ZYaGSn9TaecDhum1Nix9niRUnLd+g2V+qpt1XbrZ3U6t+4zFLpf3T5UCqiXFFHdR0OD/z7ikQqq+9ep1M51jkS4MfoqRVYQBEHICLz6gV4jCJu87UsCSqnlXtvGSjxUjAVBEARBEDKdOEQQpr8vePLJTIWNFssCJk8G/vxnRv0VFjLqbcIERv/ddBMVj2fP5utlyxjlBzD6b80aO5rMGVXmFqmiBT6A6NJTOjuBpUuBlStpqzPSLdzYsRJqPqHGiUekVTKjtbyML9FhgiAIQorg1Q/0tEGYqsgGoSAIgiAIQmRi3SBMVXz5gs3NdjqsX6qqWFvP3PgxU3HNtFQzXRVwT13tC/p6kylUGm8q4Hdt+nozMlb7Uqnfvl4rQRCEeJOC/6559QO9ipSYHecByDKvKaV6QjRPGZJRdqZ/f/6BVmoRCoIgCIKQqaSrL4hAwP8zw4cDe/awHlxLC/Dcc8ChQxT4mDYNGDUKuPlm4Gc/Y62/I0fsGoP/+q+2qEVLCzcoe3qAoUPZbvhw4MMP6Sy++y5wyinACScAZ5zBayNHAtdeS/GP889nDcNDh4Bnn2XkYlYW8MUXjIz86CPWrrv/fgqNXHop6x9+8AHwox8Bt93Gtu++C9x4IyMgDx3iOAMGAF/9KmsSnnkmhVGuvRZ47TW2y8qyBTfy8ykC0t5OQYzRo+n43ngj+9P1Ent6gJde4txPPhl44gmu18CBwL/8C9dx507aV1jIZ84+m2MvW8Z+lixxj7Y0P0+zLuHs2Xx2/Hj2B9i1G7Vwiq7R6Izwc4oRdHayna6JCAT/2HMKj7S2BvevazN2d9s1GKP5wRjqGR1dqsfUcwQ4fjhxhVCRjl5qJTr7jUfUZKoKQchmaDDpbHtfIOt1fJOq/655wUseMoASAL8HcADAEefhpY9EHH7qziS67IzbIWVCBEEQBEFIBRB7DcK09wWDavp5PQYN4nno0OQ7krqWoZ+jutqufaiPrCx/fRQX+2vvHC+Ww+xL11DUZ6djrevy6fuh7ND3nQ66WdfPWf/PfKalpXdbs+6ivme2d9Z1dI7nlVDPOMd02uunnqFzHcPZ5+w3VF9+5uilvmNf4GUufm1P51qS6Wx7sjC/D32xXqn639LxSAp+Fl79QK8RhPcCyANwLYA9AFRcdiczDGcEoS7zIgiCIAiCkOakvy94wQVUufXDJ5/wvG9f5LamQnF2NpWLe3qAgwcZ5XX0KBWNBw5ktN3IkXZknlsEYVYWlXABYNw4Xnv1VeDAAd7Lze0dQbhyZegIwsGDGeX49tsc+8ABRg/u2cNIyNGjgX79gCeftO/n5XFe2dnAiSdyHp99Brz8MlBRAVRXh48gPHQI+PRTjjl7Nus3fvwxlZfz8jj/cBGEc+cysm3u3GDHOhBgdF5TE6Pn5s4NH0FYUWFH2+l+9Lmmpnekz8qV/NymTOH9VauYZq7b6tTyoiI+193du3+tdqxxjusF8xkzIsk5pjnH7m4+EypqRT/b3c0+zXVwrrOToqLgfp1zMvtascJbZKGzz1TBy+flN0oomu9AqpDOticrms8ZWWyek0E6R61lGqn675oXvOwiAvgUwAQvbZN5+Pmr8ZYtShUURPfHzBEjlCotVWrdOgqjVVX1FnETBEEQBEFIVRB7BGHa+4JRO4JeD91/pEg3L5FaSgVHhWllXP18NFEJZtSZ7rupKdgetwgYp61+IiOcfYTqM9q5xCM6J1JfTpv7KjLEy5y9rkuio5viEVmYDqRglJDgQrK+f339fejr8YWUxqsf6DWC8BUAxQD+moA9yqRw1ln8AybAPyJcfDHw1FMsA3PZZSy1YpbuCPXHhfnzk2OvIAiCIAhCCpH2viBWrQIuuST2fgYNYhRd//6M3nv/fSAnB7j9dqoZOyNGnBFaXiK1AEaxdXdza2rxYkba6f6iiYLRkWOAHXXnpmqsox7M9qatfiIj3Ppw69Mv8YzOidSXcw59FRniZc5e1yXR0U2hIgvTMfosHOkcJXQ8kazvX19/H/p6fCEj8KRibFnWRAA/P3ZsB3DIvK+U2pUQ6yIQi4qx1A0VBEEQBOF4IVYV44zxBceOBd56y98gQ4YwfXPjRqbU3nVX9M6jXwc0lZV/BSHVkR98giAIAOKvYpwF4EQADyG45ox17H22bwuTTCgVY52mHw/y8/mH6Z/9TP4fJAiCIAhCRpH2viACAdbE88v+/axFBwBr1wLTp3Oj8OOPgXXrgMmTgaefZo3BU09l7bqdO1m3b8IEKgO/+CLw5puMBvz0U+DHP2ZNwMmTgWeeoVrwtGnAX//KNroe3yOPADNmsL/GRmDvXioCT5wIbN7MGoKzZgEvvEBV5OnT6ZDqdtXVwOWXs0bg3r2sdfhf/8Vxegzh6Z4eYNs2YNIkvu/sBL71LeCHPwRuvdVuv3cvN0qHDbPrBi5Zwg3M//s/YOpU1h78yU/4ftky2qnr+D33HMcoKmJazk032W10dE9rKz8rbU9+vq0ODNgbPgDH3bqV/ZjjbN3KfsaPD94k2rHDVi0+66ze349QisHhNpnWrweuvBK4++7IqUbhFIM7O4GlSxk5WlgYPE8/m1zOMbyoE8faV6I34qJVS/ZTky3Zm4myeRkdsm6CkFi85CEDeAnAHwEsADANwFTz8NJHIo5UVTFuaJASAIIgCIIgpA6IvQZh2vuCXyrLRuvc1dUl16EcNiw+/YwdG/w+J8fbc5bFc35++HZOlWDAVmAuKwvdJjub59JSnquqwitNO9WEzdqAbuOUl/dWE9VKx+Xlvb8foVSN9femocHdudeqz8XFkb+D4RSDzRqVbvMMVT8tlLKw35qXJh0dfK6hwb0vZx3MVKppaK5HNPUyk1UnsS/rMibrh2oixnFTCBcEISJe/UCvEYTlAE5XSr0Wr43JTGbrVhEREgRBEAQho0h/X/D116N/dsWK+NnhRkGBXedO8/HHvduNHMkoQDcGDbJVl812w4ezr48/pvrx4cOR7SksZLQgAHzjG0BZGbBpExWO8/KoxjxmDPDtbzOC8L33qGwMMKrxgguABx8EiotZ7HvxYkY9/u//2ineR47wXFTESMu1a/m+qoqRf2vXsm1lJSMju7vZD2BH1j36KG1yjvPQQ0BHR3AUXm0tlY51BCHQWxlYtzMdec3WrbYKtunc3323HUEYiXCKwStX2mez3qTzWSfa1k2bmP7upeZlpCispUu5/gcPMkrT7AvgvbY2ex1qaji+vu9GLJFffmoaOn+E+amXGarPRNCXdRmT9UNVfhALQvrhZRcRwCYA53tpm8wjVSII+/VTX/6Bddo0/tFNIggFQRAEQUgVEHsEYdr7gqqwMLEOoekUnn66HXVYUaHUmDFKLVrEc0WFUldfrdTMmUrNmMF2OlKrstJ+7zyamuhkNjSwXV0dnzef05F4up1+rSPCtmyxIwqHDrWfN6MEKyvZvqmJh3ZmtXPb0dHbyTWjM/UzbpFrOiLv6quVGjJEfRmZp6+7jeeMAjQJN44XRzxcv84oNLd5pwKh1iAckaLX9Pelo8N9POc6xFNhOVbkR1hk0jmCUD5fQYgKr36gV5GSbwNoBnAzgG3oXZj6pfhtWXonWpGS9nbgiitY9qW9nX9IzcoCRo3iH8f0HzqkvIEgCIIgCJlAHERK0t8XXL+ekW2xUl9PB3LHDuDccxkZpevqLV4cXb23aKOrzOcAb3241bqrqQFWr7bnF02tutbW4OcjzcnPnMO1jSUyLVPqmfmdR7zn7aW/TFlrQRCENMSrH+h1g/BomNtKKdUnhaljUTEWBEEQBEE4XojDBmH6+4KdncDVV1PQI5JYSU4ONzFGj2Z6a2cn00tFTTg85ubj+PF9bY0gCKlCPP9gIAiCb7z6gVke+ysNc4yN1shkcsMNVDKOxzFgAMu8DB8OVFQw4jAQ6OsZCoIgCIIgJIy09wWxdCnw7LPelIwPHwY++ICbidu2MTquoQHYsIHRgwsXUqX4wgt5PuUU1v0bNYo1+Zqb2W7BAm6adXYCM2cCJ5/MTcfLLgPmzeN1gI5kczOdyuZmvg8EWPuws5PX9HWNvq/HaW/n8/Pm8bV+1mxz//2s99febo/Z3Gy3czq0nZ32HJztGxtZH3DJEjsF57vfZX26hQvtvrSdgQCfmzePbd3GNMdwm6t+JtSzoe7FA3Mebu/jPZZeB/N7FKpNqO9Iou3si3ESQTrbng4EAsBVV/G/e13/04muVxjqviAIycFLHnKqHqlSg9A8cnIo3DZhglJz5tilO8KV8hAEQRAEQUgkiLEGYaoevmoQrlunVFZWdA5eU5OtVqvVdyMdWgV41CilBg92b3PCCTyfeGLw9cmT3VWTBw1SqqCAtuTl8ZqurWjWEezfP9hWfTbnb77WqsOVlVQEzspSasQIW/HYsoLbjxzpPh9dg1HXYZwzh30CVCseOtS+f9ppPOflKbVgAe/rawCVhrds4drrPk4/neeKCvVlPTtdk0yvlx6jrs52vp11y5w1Bpuagus8uqngmuqpHR22InJTU+/2W7b4c/ydNRjdFJrLyoJrNJptTPVmZ40/3a6qKvh55zqEu+Zmr3M9I9VBDKUu7FZ/MtJ40bYJRaLrI6Zz3bx42G4qa3v5jNN5vQQhRfHqB3pVMYZlWV8D8I8AvgpAAfgLgH9XSm2L12alZVnzAfwngGwAdyilfhqvvpOFFob76195LF1KgTUtBgbwvSAIgiAIQjqR9r7grbcCR8NlSodgwAAqIHd18b1W342EOlbGZ/fu0G0++ojnDz8E8vOBnh6+f+WV3qrGgK1SbN7TasM9PUyNPnwY+OKLYFuPHAFyc4OjJ/VaZGXxem4uFYE1770XPBc9HyC0kvLBg/ZrrWoMUGVYqxdrOjp4/vzz3s5xbi7vL1xozw+wo+h27GC6t6k4XFbGe/v28fzYY2wHBBcZb2jorVK8fLn9+oUXbKcd6K1mDNCx7+gAyst7t2ls5HU9P+fc3FIp29psGwoKeG/DBqa1n38+P9eODrYpKOAcamvt78HixUxrAnqr4tbWUmF47Vr2p5/X4zpVZr0ozzrbtLWx/+rqyGrL5jppzLm7jRmNTX5ItKJwOqv5xsN2c31DpQ8XFdn9r1iRvuslCGmOpw1Cy7JqADwIYDOAdccunw3gJcuyFimlHo7VEMuysgHcBmAegN0AXrAsa41S6i+x9p1McnKYhlxWRl9o5Uped54FQRAEQRDShYzwBa+8MnjjxyuffQbcd1/wNcvihllWVvCm46BB3Iw7eBDo358O4dSpwJ//DLzzDvDpp0BeHjf6Dh4Exo4F3nyTab8rVwI/+AE3ghYtAq69Fvj7vwe2b2ca85w5fH3gADcTP/+c/TQ20patW4FrrgH+4R+48XjaacDOnbRn6FBg/nxgzx6KtRw9yjlMnMhxfvhDbqD+6U/AQw/RppNO4qbj5MlMzQYoZjJ2LDftHn8cOPFECrRs3Ajs2gUMGcIU6gMHgNmzOeczz+R8briBm5elpcBzz9Hu5cu5JmefDfzlL8AZZ3Bz9Mc/Zrqs3mSrrOQm34QJtPfuuzkfwN58mD3bHmPWLI550022GIvZ1nnu7uZc8/O52TZ3bvBmkX6tN+/0pqXuW1/XzJ5tj+3EbcPF3OzTmyirV9sbic3NtgiMHqeoiNc15muToiLgrrt6P++2DqGuOQm1juE2gMKNZc7d67PRtAmFuTmVCBK9AZlI4mG73/VN5/UShDTHq0jJnwE8pJRqclz/FwAXKaUmx2yIZc0A0KyUOv/Y+x8AgFLqJ6GeiUakJBAAvv99lmBx+wNwYSH/ULd7N/2ZxkapkyoIgiAIQnoTB5GS9PcFJ0ywN5yiJS+PG3MAo6X0hmNZGaPVmppCb9ToqBgz8i2Wgvxmf+aPb7cINd1W26zP8RJdSYTAgJsyciYgYgyCIAhCkom3ivHnAE5TSu1wXB8HYJtSKi9qS+2+vglgvlLqmmPvrwBwplKqPtQzsaoYh/KrBEEQBEEQMok4bBCmvy+4fj1wwQXRGsfosltuAX77W2DKFIpzrF7N+4sXA2vW2BEvbhtA8d4Y8tOfbltTQzv1OZy9giAIgiBkBPFWMf4QwFSX61MB7PFjWBgsl2u9di8ty/quZVkvWpb1YpeuBRMltbX2H3EFQRAEQRCEkKS/L9je7t+iIUOY3rp2Lc+vvMJ02vx8YPx4RrYVFDAFpaGBG2yh1Dh1ml28NuH89Kfbjh8ffA5nryBEg6l8LQiCIKQVXkVKfgXgdsuyygA8CzprZ4OFqlfEyZbdAEYb70cBeM/ZSCn1SwC/BPhXY6+dW24u5zHcag/7xSxBU1DA7JPSUmDkSOCyy1jaxSyXIgiCIAiCkEakvS/4pQCIH/bvp3DHiy+yTpoWwvj1rxk1uHo16+h1dXHTMBBgncBLLwVuuw343e+A6dNZx/Dxx1mg+swzuTE3YQLwne9QUGThQuDUU207t27l6ylTeH7pJaBfP+DGG1mnsLOTdfrWrGH9wZ/9DLj3XuDhh4ETTqDIR04O6wU+9xzrC95xB/D73wPnnstahG+/zXqBeXmsRfhv/8YU7MbG4MjIVauAZ57h+1NOAbZsYR8nnGDX7LvpJtYuvPNObqpeeSUd39JS4HvfowjKrFmMutQRjKtXc72ef57n++4DzjqL47S3cw1PPpl9FhYy3TgQ4NxbW7nJqdHpyGYdwdWrg9+vWRNcG7Cw0I6q1POdPx9Ytozrrn8g6DRnM0pUb8qGiuJ0XndGcJqRm+b49fX29VARnZ2dFElZtgzYvDn4eaeNnZ3sc/x4vnfr38scItlk4kWZMV7RtIlO105U6nwqR+ymun2CICQWL1LH4F90rwMdt6PHjt0Avo9jacqxHuBm5ZsASgH0A/AKgInhnpk6daoPWee+O3JzeS4u9myuIAiCIAhC3ADwoorNT0t7X1A1NPh34GbMUGrMGL4fNkypRYuUGjCA70eOVGrgQL7W52gcRD9HYaH92rLs16Wl4Z/Lz/c+Rnm5/XrOnMjtq6vt1wUFto15ecH2AkqVlfV+xpxbV5dSHR3sR1+vqurdvro6+LNtaQm+X1XlbqOeW3W1/Yw5X/N1S0twv7rPlhaO2dVl96uvOe3R1/V7s73zmvO6s0+N21yc82xpoX3mfAClmppC2xZuDpFsMunooB0dHaHb+OkvHPHqRymul163RPSvaWqyP4tE4TYXryRizoIg9Dle/UBPEYTHOrwFwC2WZQ06du2T6LclXcc4bFlWPYDHAGQDuFMp9Wo8x0gW4SIIBUEQBEEQ0o2M8AU/+shf+0OHGC1XUMD3H38MPPigff/dd+3Xn34auT+tfGz2r1NQJk6kcu62bRQ7ee89KhcfOQKMGsV2I0cy6vDOO/m82ddJJ1F45N57GfXYvz8jEwcOpG09PWzzwQfBNhUUUKV4/37g1Vc5VkeHLbqiU3BKSoB9+9huwAAqJL/wAlBVxWi8YcMYQbhqFfCHP/D8+ec8hg0DBg+movKOHbRz5UqgogLYsIFrDFAZuK0N2LSJ0ZoDBjCSb8oUFg6vqmIkXGdnb3VgrQK8aRMjPqdMYcSiU5nYGUG4aROj3aqq2N6MINSRcxs2MK18yhTgvPOCI+q04IuzXlEold+amt4KyTU1XAuznfO1iZ67GUGonzcVmFtb+VmOHQsMH26vczgF4nBzCGeTyfjxoSMHw/UdDfFUuw2lLh2v/pOJ21y8kq5zFgQhLoQVKTnmAH4DwP8ppQ447g0BcBGAB5RS3Qm1MgR+ClPPmgU8+2x8x8/Kom82fjzwN39jR+5LNLYgCIIgCKlEtCIlmeQLYtw4blDFytixFDv57DMKn+zfD3zlK/zL8MaN3JzLyuL9A8eWrLIS+Pu/B+rqgPPP52afmfqqVY0bG4E5c4CnngKuvpobehMncoNMF87WqbQAz2bKrU4P1Bthy5bRRsBOF37+eaYz79rV+zmniImZErt3L9NHzfRctzREne5rpgPrFGEg2Fl2pgbX1wePY9rmR4zFa3qkl/Ze04hTjeZmpr83NXFdU9nWVCBZn2cyxkn176YgCEknLirGlmU1AJirlFoQ4v4jADYppf49aktjIFYVY7dyIPJvqCAIgiAImUYMG4SZ4wuuXw9cdBFw8KD3AbKyGMnW3Q0MHcq0kOZm75tPbpti4dq7Oabx/LEvGwfHD/JZC4IgCMeI1wbhiwCWKaXWhbg/H8C/KqXcVO0STqwbhIIgCIIgCMcDMWwQZo4vWF9P4ZBIOFOBAUYHjh4N/PKXweIYmmj/6pzsTZwVKxilqKMRZQNJEARBEDIer35gVoT74wCEq/3yFwCn+jEsnQkE6FsOH07BtylTgPvvZ7mWQYN4feZMCq+tWMH2giAIgiAIaUzm+IKPPeatnXNzcPBg4K23mParVWADAdvZCwSYvtvYCHz3uzxXVVHl95vfBCZNYvrsr37F6MPGRp47O4GrruL7q66iAzlvHt+3tzMtee5ctjPHc6LvdXYGn3Vb89na2uBUZW3L3Lkcr7MzdL96rk47AgH2M2+euxOs78+dy3NnJ8fUa+C0Xa9RYyNVj8eNY7967PZ2YMGCYFvd6OwMbmfOR48fbj2dc/Dr3Ht5xk+/bp9HLEQzp3g+76fPeI/l/O/XT9+JmHcyiWR/X80v3ddVSB7yXUks4RRMABwAMC3M/WkADnhRQ0nEkeoqxg0NFIDasiWymJcgCIIgCEKiQJQqxpnkC6p165QaPDg2527kSKVmzqS6sVa21aqkgFIVFd770mrIpmKv8x6g1JAhSo0YwddZWVQkvvpqzmfMGN4HlBo0iGetHDx5slKXXqpU//58X1ZGtWWtNKxVlJ1j1dWxf71Wo0bxPG2a3XdJCcdetCjY/uxsex1mzuS5pCR4bjNn2q+1+rNW2500yX2tyspsFWo9RkkJba2qorPd0BD8WttaXGw7407V36YmW/G1o4PnujrbkddoZVetAqzVlt0cfN2f/l7oNk1N9njOfvX3qKuLh9nWrT9TqdhUq/WqXmsqMOs56fnrszm2W/9e1W6jscnZp557VVWwfX4IZbtf1V5TfToaleBkEo0ys9t3PRm42RWLGrOQuYjSdlR49QMjqRhvB3AegFC5G/MR/q/KxzVbt7JWdXk5RcSAyKJegiAIgiAIKUTm+ILt7bZoiBfy81l/8PBhqtE98QSVi031Yq1u29TE9xs22Pe0QnEoPvuM524XfRd9D6AIyv79fH30KAU97rwTePhhoKvLbvfJMVHpvXt5fuUVHhpToMUc0zmWMw17926edSp3bi4ViQEKnZgcOcLzCy/0npPm/fft11r9+bXXeP7rX3keNgz42teoSAxQ2GXr1uAxdu60bd2923a2zdcvG7r8AAAgAElEQVS5uVyj2lpe0wrKBw/yswNscRitaFxWxut6PMBWdO3uttVhdXsg2MHX/TU1cTzdRp8LCoJVcnU/GzfaitnLl9ttAbu/lpZgJWSnWq1X9VpTgVn3r+0w56XvufXvVe02GptC9blxIwVstH1+FHoj2e5Vtdf8zNra/KsEJ5NolJndvuvJmKObXbGoMQuZiyhtJ5RIG4R3AvhPy7JeVUr93rxhWdbXAfwAwPcTZVy6kp8PXHIJ/z1bs8YWklu5sq8tEwRBEARB8MXx6wv29NhqwZs3Ax9/zNc5Odw01OTnM11VP7NlC2sW3ncfhVF27OBmXl4eUFLCWoZ//COVfffuBc4+m8rC+/YBAwey7zPP5CbbqFHcpDx0iBuT+/Zx0+uSS4C//Vvge9/jpuGcOcBf/gK8/TZQUQH86U/A/Pm06eGHeT7rLNqmFG35/HNuYp5xBvDcc8AppzAtesoUbhpu2MC28+fz+t69TLP+8Y+Bn/+cG3HTpgHPPMO1GTSICs4jRzKtuqODdufmcj02b+a9lSu5LoEA8PTT3GT80Y+Ygq3Vi9vauFFnirzs3Utl5127gJNPBqZP573OTjraDz7ITT3z9fXXA7feSjXnzZvteourV9v1FzV6483NcS8q4gaFTmnr7mafQG8H3/njVdemrKgIvq77vesue57m5oyzrVkrMtRmj9cfzs52BQX2/M0NyFDttS1eNm2isclZE1On9gNM53fa54VwtvvZfNKfmfP7k4q4rX2kz838rhcUJG+ObnbJRpDghtd/e4SoCCtSAgCWZd0L4FIAnQCO/TkOE8CaNPcrpS5NqIVh8CtS0tkJXH45/zj52We9/6ibkwMMGcJz//6sL3jWWXE2WhAEQRAEIclEK1Jy7NnM8AUDAWDq1N5Rb6HIzma0Wn4+8B//Afz2t9w8O/dc4NpruSk3enSwQnE6KceagiXyY8s7sm6CIAhCmuHVD4wUQQil1OWWZa0BHcOvALBA5/CflVK/jdnSJDJ+fPiMB0EQBEEQBCGYjPEF9+5laqlXjhxh9F5PDzeDduwAzjuPEXWvv+7+TFFRb3XgRG4adnbaUXdu6srhkOic6NDrVVPDzcJ02AwWBEEQBA9EUjEGACilfquU+rpSaqJS6qvHXqePQygIgiAIgiBETUb4gvX1wAcf+Htm6FCed+xgUemamsgKirpuVlub+/t4ouuxLV3q/1mdppUKm1vppEqp123NmsR9rqFIp3USBEEQ0o6IEYSZgmUlb6xBg1gWprCQWSk68ySdsk4EQRAEQRAyiilTbGEKr7z/PjBzJuvPPPUUsGoV8OqrtpjDXXcxMvGKK3hdC35YFjePGhtZgw9gFKIWz9iwgfX5LItpyuPGAY8/zlp+JSWsQ/iTnwB3302BlB/+0BZMOeEEpknrMkG5ucBjj9k1ci64AHjjDUY5KsVU6UmTgGefZc3Bb3wDeOghCoFUVlJ8pH9/9jluHCMR//hH9jNsGOsjvvsu+29pYf3BN96gwMigQVyfrVtZy/CVV1jz73e/4xjnnMM11LUIzz6bKdrf+Q7w0Uecf34+7+/bxzVduZLrvG4dN2arqlgv8frrgZtvBsaM4fijRnFO55zDtXnpJaBfP+CnP7XrDe7Ywc9mzhw65gDHGzkSuO46jvPtbzMSs74eOPFEzveee4LrDAUCnPvzzwOnnUZHfvFi3quttZ38mhrWN+zp4TiLFwe/N9PRdb/OHwf6mlkLUUeHmoIqd93V+xk9PtA79V3XOVy8mPN+7jnWumxs9PbDJNIPmWT90DHnumaN/LBKV5zfF/mhLAipgRep41Q9pk6d6kPWue8OrcAtityCIAiCIPQFAF5UKeC7xfvw4wuqdeuUys727sANHsxzWZlSDQ18XVLCc2kpz5WVShUWJsaBLCjg2bL61pF1syncUVwc3X3dd3V1dP3qo7xcfelwl5W5t8nNtcdUqveY5eXB352mpt59mA69dvKd/TjfO38EuP040Nf0PKqr7XtdXXafkcZ3u+/FplCYtnZ12edwc0kEzrnKD6v0xPl9kR/KgpBQvPqBx00EYTJxRhD6FRYTBEEQBEEQ4sx117GuoFfGjgUOHGAUWk8PU4x1BOCYMcCECYwkBBiB98UXfK3TVpQK7u/EE4GFC/n6f/+XfQNUyNu/nwrGBw8GRxA2N3PMnBw6mG4RhMOHMxrv6FFGCB46RJXfAwfYJhERhK+/TpuHDmVUoFsE4YMP2irMSoWPIMzPB/77v4Ft2xgZNnFifCIId+5kHxUVVL4FekcQAozSO3jQjiAMlTZcWgpUV9u1JjVmXcKKiuAIQvO980eA248D/TqUmrKbgq5zfP1a10isrbWVkRcv5vrqCEKvP0xMW3UkIxBaUTlRmHONRs1YSA2iVeAWBCGhRFQxTmX8KNddfjnw619HP9awYfQpS0vp2xQV9c4SEARBEARBSEViUTFOZXypGK9fDyxaZKcBe2HOHKYWV1UxPXnmTODDD5mCWlbGDbOtW7mRozelzHS5UGmQzpTPUKmSftPukpWm50XJNxXUfmNZD/NZIP3SHxO5/pIOKgiCkFZ49QNj2iC0LKsUwBKl1I+i7iQGfDmFgiAIgiAIxymJ2iBMK1+wsZGbJuGwLDva7dAhbiiedZYdzTVxIvvoy02vVMDLBlG6byKlwgZnLKT7+guCIAhxw6sf6EnF2NFxP8uyFluWtRHADgCXRWOgIAiCIAiCkH6krS/43HOR2+g/nB86xPO6ddxgWb+e6cQ9PUwxralJnJ3JJFpVXC8KyH5VklNNobe2lpuD6ZrymEoq1YIgCEJa4LkGoWVZXwOwBMDlAIYCuAPAPymlnk+QbXElkSrGlgVceinwH/8h/w8WBEEQBCEzSXdfEGVlTBf2w2efMQW4p4fvt21jH93dPHbupPqurm149CiQl8d6gp99xtp42dnceNTOaF4e8Jvf8PW3v83ad6NGsR5cUREjFq+9lunNo0ez1mFdHXD++aydt3cvNyz372f9wrlzgbfftpWCy8uBP/2Jts2aBbz8MscfMQL42tdYa/Caa1jTLxBg7cSuLkbLtbYC77xDldyzzgJ276ZdS5YwDVpHUi5bxpTq2bOBG27g2KWlwJYtvWsC3n47C3Ob6bqtrfaaap57jmv7059yzZub3dvq2n46LRsITueuqWFtwa1b+ez48ZxnczM3ec85x1Y01v3V13Ndly7l2t96K2v/NTTYG5duSsPmtc5OPm+qDofCTDHXNYvCRfwlQ/E1Hn3G2kcsCsXhFKHjlaLf11GZfsbva1uTyfE0V0FINOEUTAAMAvA9AC8A+ATA/wCYB+AQgK96UUFJ5JFqKsbZ2UqddppSdXUUu2tqUqqjo7fIlyAIgiAIQjJBlCrGmeQLqspK/87d0KHB7ysqQivj+jny8pQaODD0PfN9Vlb8ndb+/XvPK9z6DB/O86BBPI8dy3N+fuSxqqpshdKmJr73YqOpHu08Zsyw7db9aVVbU6W3uppOeKTPvrra7kerJetnw6kGm4q+c+bYdkdy/E1V4aam3n2Gaq/Hc7PJTVnYD6Y6sNc+nGNqteeGBl4P90MonBJyNArF4RSho1FqjuZ+tHj97PyM72wb6/cjGURrY7IVkNNhLQXBgVc/MFIE4W4A7QD+E8CDSqkeALASGY6Xxhw5AmzfzkPzwgu2wF06li8RBEEQBOG4JnN8wdNOA5580nv7ggJGBwKM+vv8czp2AKPl3norels+/9z7vaNHe7cZPBj45BNuMZnoGoqR6NePkYO6fSDgPp8BAxgJuWcP33/yCc/Dh7MPreoMUCHZzdYpU+xIv+5uir04+x82jFGNTzxBVWWA6sNbt9oCMZWVwPTpjPi7+2620Z9HdTUj9+bOtVWQt27ltbY2+3MfMIDRh2YE4datdNYbGjgnM4KwrY33qqvdVYNNRd/KStvutrbwjn9tLbBhQ/BahFNxdY7nZpObsrAfamsZObp2bWT7I425dSvnpvtzsymcEnI0CsXhFKGjUWqO5n60eP3s/IzvbBvr9yMZRGtjshWQ02EtBSFKwoqUWJbVBeANAPcAWK2U+ujY9UMAJiul/pIUK0PgpzB1ov3YYcOY/TBuHP2D/PzeGRAS8SwIgiAIQl8QrUhJJvmCaG4Gli+PfrCqKm52aQdv1SpugHz4ob35ePQoN9yOHmWK7tatQE4ON8YWLABOPpmpt5YFnHQSU4V7evh+/Hjgb/6GY23bxvTiZ58NnxL7ve8xRXnCBDt1+Ac/4ObmgQPAhRcyDVinKOflMY331ls5hpky/OCDwPPPM415yxZuQF1+Oe8NHszNu3PPZWqzmR5bUsJ05muuAX70I6ZGa3u1rWbKp04Zzs/nGl13HX9wn3WWfT8QYMpua2twerLup70duOIKOw3bHMNJIGCrTeuUY+d9r6m9ofrXabGrV9vrkqgU0FDP9UWKcKjUZzPdO14K3ZlMMtYiHdY7HWwE0sdOQTCIi4qxZVm5AL4O1pupBPAH0EH8NYDT08kpNP87Bmw/4ZprgH/6J/o9OTn0p26/PXLpEEEQBEEQhHQhhg3CjPEFEQhwQ+3BB4Ov5+QAhw/ztY6C698fKC4GzjuPG1TmRpdzk8vLZpAgCIIgCEIf4dUPDJtirJQ6BOB3AH5nWVYJgFoAK449t9yyrP8BsOFYu5RGC3lpWlrs19/+dvLtEQRBEARBSHUyyRdEUZGdvmqiNwcBO0X2iy8o0PHXvwJnnx3cvq0tOBKxoCAz0swkKkYQBEEQjmuyvDZUSu1USjUDKAVwAQAF4AEAXYkxTRAEQRAEQUgV0t4XDASYDuyHLVuoYLt8OXDVVUx77e7mhmBDA9DUlLy6V4CtqBsIxL8/XVerrS0+fSeLeK9JJiJrdPyhlbubm+VzF4RMI4H/pkcSKenFMQWUxwA8ZllWIYAr425VAkhmLe0BA4AhQ1i6JVJZFEEQBEEQhHQiXX1BtLaGFwdxY9AgOnVZWbbYwtq1TEVpaGAtvBkzWJtmyxYKKwQCdDyfegqYNg149VWmIxcWslD1iy+yzt/8+cCPfwxUVFDhrqCANf7efJPjnHEGBTw2bmT7lStZD3DjRv7of+ABiqVcfjmFMRYsAIYOZR3B99+nE/r97wPXXw98+inrG77zDmsVfvIJbTxyhGuyfDn7GjECuOUW4KGH+Ow//iPw0Udsm5cHHDoEfOUrrFt4882cd34+5/fSS7bdAO3Ys4cRmK+9xs3Zqio7ZVsX6p40icIg55/P8+rVXMNt24Azz7TFAHR9oJtuYpr4c8/x2e3bWf/xl78E7rnHruvolv4dqWae11p54eoAmvUZw/0ACJeq7tZ/ZyfbTJnCNfHz46K1lZ9xdze/O6HscVubaOoHuq2Ds95TLNGqXj63RJEKkbZebDAjnTMlylkQBJJIoZxwEscAigEsAzDY5d6QY/cKvcglJ+KYOnWqD1nnvjuSpbguCIIgCILgBoAXVRS+Vib5guqyy2Jz6KqqlKqr43nLFqWampQqLIy+P8vy1768PPh9cbFS1dXhn8nPj96+goLQ94qLY1tLwLbd7MttPi0tPEKtQ7g1cjrhuh99Xb/X47o57c5nQl0zr3v5AWC29dK/uTZ+f1w0NfG5pqbI9jjXxuuahJqbW3/m664u++wVL59boog091SxoauLn3dTk7+1jYVoPstUHidaUt0+If2J4jvm1Q+MFEH4fQAlSqkDLhuL+y3LGgfg/x1zDoVjOCMIk5l5IgiCIAiCEEcyxxdcv95bO8vi1gZgi5aUlgKzZtkROf362RGFubmMrAOAgQMZrTd2LPDWW4z8272b93Sk3dChVB12iyAcPhx4+WVgzBhg1CiqGOv2ZgRhfj5w992068MPe0cQdnQwSvDrX2d0XSDAuoqjRtFWHUHYvz9VkAsKqKrc1QW89x7Hb2mJLYLw6aepbnzuuYxa3LED2LePUZRz5zKCcO7c3hGEFRXBEYTake7qco8g1ON3dvKeGUHodML1e+e5poa2uDntzrahrun33d3u99z67e52t9Wt/5Ur+VlNmeL/x0V9PT/jcM+FWhuva+K871yHUOsYTSSMl88tUUSae6rYUFQUOlo0USQyqqkvxomWVLdPSH+cAhtxJJKK8SsAvq+U2hTi/lwAP1dKfS0h1kXAj3Ld/fcDl1wS/Vj5+czCKC1ltklODjMm/Eb4C4IgCIIgJJsYVIwzxhfE+vXAN79pb1y4oTf48vK4CXjgAFNiH3kEKCuz0yYXL2Yq7DPPcMNuzhy2v+kmYPNm+4e7mWYJRE4LjJSG6TW90U+6aLjnYmHFCjrKOh07FVIzhdRDvheZQ7I+y1T/zqS6fcJxiVc/MNIG4acAJiil3glxfwyAV5VSg6K2NAZ8OYWCIAiCIAjHKTFsEGaOL6hruFkW8PjjtmKxE8sCbr8dePddO2JQb3I5CfdD0LlBdrwhP5IFQRAEISXw6gdGSjE+BGA0AFenEMAoAId92iYIgiAIgiCkB5njC37ve0y3jYRSbDtxIlBSAixcyE2uzk5g6VKmeo4fz7bh0nxSIRWxL0lgCpQgCIIgCPEn0gbhSwC+AeDZEPcvBrA1rhYliGSpGOfmArfdxj869/TYJVFaW21fUhAEQRAEIU3IGF8QBw96b6sU6wICrIXX1gZs2MB04oMHgfPOc4+Ma29nTZv9+1lvr6iINQU//BCYPZs1B6+8ErjvPuDUU4Hp04E33uB56lSgro41/F56Cbj3Xta2qa9nTcCODqY879rFGofz5gFLljDVWTudug7g2Wfz3q23Ag8/zFqI//zPfH/ttcBPfgIcPgyccgptnjuXasbr1lF5ub2dNXXmz2dx7e3bgdNOY9/bt9P2XbuCHVwdoanfv/QSHfDSUtZSnDGD9Ri1MrGbuvDs2UzTNjdhNW4qvk41W32ePZv1Gt0Uf93Sr83Ucd3P6tW85kwPN1+H69cLpv3meLEo+/pJIw81Fz/tzPbhlIuda+13npkakZqp8xIEIT0Jp2ACYBH4V+HvA8g2rueABakPAbjYixpKIo5UVTHOzXUXahMEQRAEQegLEL2Kccb4gr5VjLXK8LRpPM+cqdSMGUqNGsX3M2Yo1dCg1NVXK1VaqtTChf6Vic0jL6/3+7Fjwz8zZ07oe26qx0BsysbOo6xMqXXrOJZeJy9HdbWtvtjVRWVo3Z/Tce7o4Htzrlq9VavzVlYGn3U/bkq5TgVY3Ye5RqZicHW13cZ8XV5ONetQ/XrBqcbrZm8kdFttl1f1ZFNN2Pw8/LZztjdtCKWMDIRXVI5lfulGPFWRw31v0lFZNx1tTnVkTY9bvPqBYSMIlVIPWpb1MwC3ALjJsqw3jt06FUABgBVKqQfitluZAYSKIFy5sq8tEwRBEARB8EdG+YLPPeevvVI86zSUZ58FysttVeItW3ho3nqrdx+mInIotFLysGHA++/bqryffw68+SbbTJtGFd+dO/k+O5vqeZYFVFUxsnHmTNq2axej9jo6GLX37rt2BOF11/E6QMXjffuAE05gRKKu5ThuHPDBB3z9ySeMXty1i+9LSmjDyJHsd8cORkR2dbEfgHYoZa/NiBFURh40iP0NG0YF6LY2piC3tdF+gErGX/lKsOO8dCnbV1Zyrm4qvnqN9fn88xkdqduaqqKhUr8LCzmP8nKOX1FBEZq1a/m6utp+XV7OdayttdczmpRyU423osK+5kcFVbdtamK9Sy/qyeZ506bgz8NvO7N9JOXi2lo7EtcrfueXbsSzFEG47006Kuumo82pjqypEIGwIiVfNrKs6QAuA1AGwALwGoD7lFLPJ9a88PgpTJ2oFGPLor/wP/8jUeGCIAiCIKQm0YqUGM+nvS+I9nZuxHR18f3Mmdz0C0VeHjfp5swBzjmH1+bPB268kWnGZ5zBzby9e4EnnmDNws2bgZwcpu9efDHwne/Y7c8+G1i0iCm0y5Yxhfixx5h2fPfdvLZ5s50eO348+8/Pt9NcW1qArVuB669nuvDKldzYcqaC6lRbtzRYZzqtm+KyM31Xp7+6peCedRY3Hm+5henYzv70Mzp9WM/Ta9qpW+1Hcz5uKcZ+0n/DpTeHSrPdu5c2OecSL/ykncaaoppshWu//UgKrne8fM/TaR3T0eZUR9b0uCUuKsapTqwqxs3NFKdrauJrQRAEQRCETCTWDcJUxbcvOG9e6OilrCxuxpWVsW7gzJl876WQtPzoEgRBEAQhRfHqB2ZF6CTfsqzbLMt617KsDy3Lus+yrIzxeurr+YdY/UdZQRAEQRAEwSajfMFAgEcojh4FPv2Uab5jxzK68NAhRsrp5wIBYMWK3v3otK22tsTZ74VQ9qXrOIIgCIIgJI2wG4QAlgP4OwCPAvgNgHkA/ivBNiWNoiKm3ssfegVBEARBEFzJHF+wrY2RgZHYs8eu/ffkk0w3aW3lhlhrq/tGYG1t/Ouj+dmE021D2RdvW6LZEI00H6/zDQSY+tPc7G+DMpGbmp2dwIIFPCfTjuNpo1bPtbPTfc6h1qKv1uh4+mwEQcgcwimYAHgDwCXG++mgWl22FwWURB+ppmKcl0dhu3XrRBxIEARBEITUAdGrGGeML6i6upQaM8abU5eTo9SCBVQnrquzFVSbmqhcXFVFBduGBqUqKpQ64QSlsrKC+xgwgIrBublUvB03TqnsbCrslpfzel6eUv36UQG5slKpSy+ljZMnK3XSSexn4EClTjtNqZEjqRR89dUcs7iYfUyaxPtahXf4cKUKCvh6zBilfvlLPltQwKO0lOMuWqTUxRcr1b+/UieeyH5nzmQ//fvTzgED2PbSS3m9sFCpefN4bcoUKgxrleEtW7hWJSVKnX46+2po4PXKSnvtteLzgAFUlt6yhes5Y4a9xlppU9/Tn0FHh614DNifRV0dX+tnzc9cO+X6M6yq6u2kO8e7+mqeOzp692O+7+jgWduk1ZfDtdeKxeb7jo7I6qLmfb/KvqHsWbeOn3NdXeg1cT4Trx84fvtzKj475xxKDTieKsF+MP/NiBa/a+SlfTJVbI8nxdxUmWuq2CGkHF79wEhO4UEAIx3XPgMw2kvniT5SbYNQH8XFffP/IUEQBEEQBDdi2CDMGF9QdXVx4yc315tDl5/Pc3m5vclVUaHUkCHqy82yZDqY0R6WFXsf+fm2g2seLS32hk15ufuzoa6Hul9VZW+uOO/psXQ7c7PQtEljbg7pPt2cdN3ObTxnP+Z7bU9Dg73R56V9dTW/j+b7SBtZ5n39uqnJ22ZAKHvMzzTSxlq8N9r89ufclI20oRnpeqLxs0EYyka/a+SlfTI3TPtqc7YvSJW5poodQsrh1Q/MiRBgmH3MMTQ5DER87rgkL4/pyr/6VbCImyAIgiAIQpqSOb5gWxtw552h7xcUAN3d9vsLLgBefRXo6KC67ZNPBrffuxcoLQXeegvIzgaOHAm+n50N9O/POoZDh1I92bKAU0+l0vFrr7Huoaa0FJgxA3j6aWDYMF7r6gIOHGBtxMGDga98Bfja1+hovv02sG8fUF4OjBsHPP44MHo0n9u5k47pJ59Qibl/f44JACeeyDqL1dW055FHgCFDgIULOddnn2X7gQM5x5wcrltJCXDFFbTz6aepyFxbS+VgwFZmfuQR2p+fD8yaxXY33sh1KC8H3ngDGDSI9i5axGLgWrW5s5MiMrNmMWXbVHQuKqIickUFx6uvp3319cGKz6YDrl87nfJQ7/V4Y8YAu3ZR0ditH32uqQHmzu0tTuO1ve5fq1G72RZuLl5FcULZM2kScO21wPnnh16TUOdY8dufrg0F2OdQ971cTzT19fw3xcv8dMo+EGyr3zXy0j7en2Os9mQKqTLXVLFDSFvCqhhblnUUwB8AfGFcvgDAkwB69AWlVE2iDAyHH+W6zk6gqgrYvdv/OHl59Ouqq+l31ddL3UJBEARBENKHaFWMM8kXRCAA3HADN7EOHgQGDAA++4z3Tj+dm1K33cYNscOHgaYmOn1tbdzYWbUKeOYZxludcYa9YbVmDe+vWWP/KHMqGrupHOtr5rNuDmYsCsmdncDSpdyAiqTEHM9xo0XUoIXjEfneC4KQYLz6gZE2CNu8DKaU6pMtal9OoYH8GywIgiAIwvFEDBuEmeMLBgKMEOvo6H0vNxcYPpwRd/v3M/Jn1SrgjjuAKVOAJUt6bwSKEykIgiAIQhrg1Q8Mmx7SV85eoumrSHNBEARBEIR0IqN8wbY2981BgOmvOs0kK4upxnV1TGHduBHYupXnTZuAtWt5vusuu99oNgzlL9ZCphDP77L8d3H8Ip+9IPQ56Vc/JkosK7H9Z2UBEyYADzzgP4NDEARBEARBSDC1tcA//zNr8oVD1wUsKmIk4a5drNtXWsqagDNmcJNwzBjW5evoYP0wy2L6scnw4dxknD+fbbduBXp6gI8/Zv2aV15hvcCiIv443r4dmD4dOPdc4LrrgFtuAf74R+CJJ4APPgBOOol1CJ95hjZNmgTs2MEzwNqE+vX27cBpp7Hv+fNZB7C7m7US338f6NcPuP9+1jFcvJg1Ew8epJ0A7SopAU44AfjBD4BbbwW+/nXghz/k61277JTq5mbgscd4/Y9/5DyvuYbXm5uBu+8OXaNwyRJg9WquS36+Xcuns7N3fUGzzo++b0Z4antaW4F33uGm7siR3HQYPz54AwJw34xwS8t2poPX1NBmwLYpEOC4ANd72bJg22bPZi3LZcuA9evZbvFiRqpu3cpno/0RYY7trIXk3HRJxCZMqBp6Xux22qL76u62a/iFS9X323+spOImVqJsSvZco/0eCYIQP7womaTqkYoqxjNmeDZJEARBEAQhKSBKFeNUP3ypGHd0KJWT48+xq6jgefBg+1r//vF1HocO7X0tO5tnraQc6xFKSXjYMG9jDBjAc1ZWsF0NDcFKwqays35Gq7M9kiUAACAASURBVEaHssFUJjZVX53Xncqc5n392lT4dY6hlH2vqoqq1OZ4WklWz0c/Yz5nqg47bTJVks25OlWe3e45xzPxosJrzjnZasRebXTDzRbdl17PUKrU0fYfK4lUiY3nOsaDZCvi9pXitJBayPcgIXj1A4+bCMJksWdPX1sgCIIgCIIg9GLpUoqP+CEQ4PnAAfvaF1+4t3VDRxVOmMCIwS1bgBEjWPPw0CHgvfcYwVdWxkhALZCiFZF7eoDKSkZTffAB+3r3Xd4bNQo47zzgoYfsiMR9+3rbUFbGyBwdQfjeezwAPgfY4wKMerQsjqfRYi5HjzKq66KLgPvus1OvNRddxGjHt95iNNy77zLqsrycNtx7L/DrX9POkhLgW99ihN3Bg8H9AIzgO3gwtEKxvq+j9LQ6MMB5mhGEphqxThN3oqOXGhoYXamf0c8BtgpxTY2tpuxUC62qsiMFTdtCRRBOnMh1NMdzswsIHVVVW2srcCdbjRiIvn6Tmy26r0CgtwpwIlR9/ZJIldhoI+gSZVOyFXGlDpgASCRpHxNWpCTV8VOYul8/+mHxICuLflNeHrNNBg8GTjkFeP554J57gLPOis84giAIgiAI8SBakZJUx5dISWcncMUVwAsvhG6Tl8cU5KFDga9+Fbj5Zm7mBALcxDl0iJtppaXAunXcYCssZNosACxaxJTiMWO4mXbbbUz7dUtpdUuRnTQJ+Id/4P3p03k2U0bdUkl1SqzeeOo5Ji7d08N7ra200UytbW3lRt5TT3FDq7U1OGVWt+n5Uqia6DRfPRedamumB+/da6fomuNGUnQOlSIbb/RYzpTmWNMpMyXNU+g75LMWBPnvIEHERcU41YlWxVgQBEEQBOF4QjYIj9HZCSxcyGg9gPX1PvqIr08+mbX5CguBSy5xr3tnkk4/YlasYERGS4sdkZFO9guCIAiCEDVxUTEWBEEQBEEQhIwgEAAuuIDprxq9OQgwjbZfP0bA3Xabff2FFxgNpyPsFi+2BTHSJf0pXCqnV2RDUUgk8v3KLOTzFIS0JKuvDRAEQRAEQRCEhNPWFrw56GTfPta0MykpYb26+npg+XIeX/86o/Ha2hJrbzzRm4Gx/FDXdaHa2vjjf8UKu0aj0Jto1yjSc4nqt68xv1/xJNXnnakk6vMUBCGhHDcRhDk5dr3nRDBsGGtFf/op8Itf2OVm5A8mgiAIgiAIKUBtLfDb3wJ+UpK18zh+PDBrFvDoo3x+4EDgv/8b+MEPgDPOoON36qnAmWcCzz4LvP4605M//ZQiIUpxc7K0lPUNDx0CPvyQha3PPx8YPdpOZW5vp61LlwI/+QnH2r8fuPBC4NprGb1YU8Pz7NmsPTh+PO3cto11DAHgpZcorvLOO8CgQcAdd/D+7NnADTewoPaNNwI//SmjJnftAsaO5Tw/+4z1CW+9lfb09PCoquLYN9wA3Hkn0NEB/Oxndj1CHVkJBNc8NOsVAsD8+bbdus4i4F4bEAiORjL7Dpf6bdqjawyadQ5Nu8x2gF3X0ayjaNrvpU6i3iDZtAm4667e66DnaUakAsBVV3FTuquLdS+nTGE/e/dy3J4efse6u4Hm5vA2mGui+wUSG/kaTeRYIMD5NDX5E8TwMpbzc0j2j7N4RdKlW0ResgVOBMEL6fbfUV/gReo4VY+pU6f6kHVO3lFcnFxFeEEQBEEQhHAAeFGlgO8W78OPL6iUUmrixOicu9NPV2rOHKWGDEmcA6kdx/Jyvs/N7d2mujr4rNv6cVDNZ/S1SM847Sws5OvCQr43bWppsa+Zr/V9N7vNds71UMq9v1COtps95vVQdpn9mevsZr8XJ7+ry31NzH5CrVt1tVJVVcHjOZ9ravL+vTf77ery/lw0RPp84vWM1+ecn0NXl31OBtHOLVH9CMLxzHH835FXP/C4iSDMzu6bCEJBEARBEAQhRXjzzeiee/ll+/WwYVQp7ulhfwMG0AHMz6e4SbQRhM7ouFARhHPnMgpt7tzkRhACnFNtLce48krg7ruBacdqnmubTAfYfF1TA1RU8LUzglC36+62I+vc+gnVt4m+7rSntpb9u9nltHvlSvtcWNjbfi9OflERI9bMyEezHzOC0G3ddMTglCl8X1PDFHhn1KUXzDVIdNRMNJFj0UabeXnO+TnoiEIgOTVE4xVJJxF5ghA78t9RRI47FePmZpaPufRS4OGHeW3ePPpPra38f64zA0GiTwVBEARBSGdExfgYv/oV8N3vems7fDhw0knAK69wEyw/nynEjY3BzqGZjqo36iIhaU6C0DfIf3uCIByHiIpxCOrrgYIC/j/h1792b1NU5L2khyAIgiAIgpAmvP66/Tory47yAxhRd+KJwJ49wIwZwJYtjJK77LLwmwnjx7M2IeB98yHZUUyCIBC/6t2CIAjHEcedinE8RNwEQRAEQRCENCMQAO6/335/9Ki9OQgwDXjPHqbZlpby2hNPUCyitTW8CqpWSm1t9abcWVsLtLRImpMgCETUlgVBSAGOmw1Cy0r8ceqpwCmnMEpR/m0XBEEQBEFIIdraWGcvEm++CTz0EF+/+CJ/tC9fbtef6ewEFizg2bkxeMstrPd3333AkiWsHzdoEOsITpoEzJxpRw4OHQqcfDIwYQKvtbez7s2SJcD06UBJCZ3K9evZ5v77Oe769UBlJevWtbcz7aW+ntcaG2lXczNfNzfTxvZ2YNw44OKL6bDW17O/wkLWw+vs5DF3Lvtpb2ebceP4Ws9b3+/stPu8/HLardejsdF+r9dK26nt0eum56bH0Oj7nZ08t7fb/TQ20oYlSzhOe7v/jZVAINget/tuffrdxAnX3rznnK95ze1Z8zsYLYnekJINL3/oqOJIf1zoC+L5vRcEIbXxomSSqkeqqhgfp8I4giAIgiCkKBAVY6W2bFGqoCB6566ujsqnY8fyfXa2UsOH8/XJJyuVleXPUXSqFJeVubfV7bTtWkE41DMlJcHv58xxb2eOX10drJKr5whQydhUgtXjmnY4FX91Gz2uqVpsqvXqOZWXB39WTsVf/bybarNWWo5GNTfUc06lS61829Tkb6xwipnhlJ4jqTWbKsvRkmhl4+NYLTQqkq2u7Ae/n6V89oKQcnj1A1OiBqFlWX8LoBnABADTlVL+lEdShLFjqZS8cKFkjAiCIAiCIHglKb7gsmW2im00dHYyukcrIR85wpRkAHj//eC2gwdTAQ8AsrPZVpOTQ0fxueeABx4AiouBv/s7qtru2MGIvvffB3bvZpThvn0soL1qFVWDTzoJuPNO9nX++Uxf2biR6soffwzs3GmPc/gw1Yjr6nhv717eGzPGjqYcOdJW7f3rX6m2PHw455mfzxTrtja26e7mczt2sH1hIRWJ9+yxFX+7uhiBqduUl/P59ev5XjvJmzYBa9dy/s6oKacS8ezZwE038TN88EGuXVkZ57ZjB1Bd7V81103R2Dm+qSzd2Ag0NflLDQ+nmOl2L5IStMZUWY6W2lr7M2hri39dPlEL9Ucq10b0+1nKZy8IaUtKqBhbljUBwFEAtwP4R69OoR/luilTgJdfjt5GJ8OGUQn52muBNWtECEsQBEEQhNQl1VWMk+ELorGRaW9eycoCRo0C9u9nKu2993IDrKUF2LCBG2ljxgADBrD9wYPchLvwQjqIq1fz+uLFfL1tGzfJ7rwT+Pa3ewuamO8BvtYbY6ZCciBgpzvX19tta2o4TiAAPPYYN87KyoBvfMNOa9bPaZt0H9qJ1TbU1NDB1WfT0XWO7+YAe20Tq5psshRpM1X5NlPnJQiCIATh1Q9MiQ1CjWVZm5Aop1AQBEEQBOE4JdU3CDUJ9QUDAeCqqxgxBTDCLjsb+OILbgYePWpH7DU12fXp0nEDJV3tFgRBEAQh7nj1A9NOpMSyrO9alvWiZVkvdnV19bU5giAIgiAIQhKJyResqGBKLcD02y++4OvcXJ737eN57VpuEJqqxKlQeN+rDTpdMZ6bg6kwf0EQBEEQEkbSNggty9poWdZ2l+Oi/9/evYfJVZX5Hv8tkpCkO3JJOlwEExKTIRFQEYJIgMSxA9hAc5FxjESZJuoZj83jCJPGy8R0yDhoR5p5DmHmCAMt+nCbC45RYpCMBBBoLhoYZEiPXCbhpnQFYqC5Rtb54+11atfuqu6q7rrv7+d59lNV+7LW2nsXYfVba6+3kHK891d574/23h89ffr0Mberr88SoR1xhDRpkv2A3NhoWYkPOMBex42zJ0z23NP6j9OnS4cfblO+nH66NHWqdO65uROhAQAAJF3F+4I9PZaN+Lnnhm4LgcLgwQdtXyk951wIFoZHZ6XM7LPRzMF9ffZ+8eLMbLshw+/NN9tjy8uXZx4TzdgbzZIcOphhLrzzzkuvi2fAjWbCjbczV3bckYJ/qZQ9lpzr/MNx0ey68Uy78czB8WNH+pytjJGMlEU4nu251CoVZB3NPS9UMTIrV4NaDITXYpsBVKd8MpmUa5G0WdLR+e5fjVmMV63Ku0kAAABloRrJYlzKvqC/6abRde4aGryfPDn9ecYM7xcs8H7RIu/PP9/WTZmSecz48UPLmTVr+O0HHphZ1kEHWV2S98cdZ3UddJD3e+9t65qbvf/Zz9JZfBsa7DW0ddYs71essOzNoe5Fiywbs+T9Xnt5f/jhVmbIxvyRj3i/davtM2eOHbt1a2b24KOPtmXGDO8/8AFb9+lPWx2hbfPmWftCltx4GfEsvdEsydk+e2/rTjwxs4yR5JNFONqJD0uuTLLRTLP9/Zn79/fb9V60yF6zlZGrPaHcrVutvBUr8m9HPp/j1zqakXnFivQ9ip9fPKtu/JyjdUXvdy75ZuqtZEbfUmbgLdV5jabN1Zw1GUDR5dsPrHhHMKMxBAgBAACKjgChTwfSirlMm5bffg0NFuQrdv3hnPbYI/c+0cCcZIG/4coMgblwbPg83LmG4KTkfWOjHxJ4CmXMmZMZUIsGrELd2T57nxnQmzMnv8DGcEGQEOwKwbgQMBsu0BINxETb09KSeXyuMnK1J3q+I5URb0c+n7Nd61Wrhgb2sp1ftA3Rc47XFb3fueQbyCplkG4kpQycleq8RtPmSl5jAGWXbz9wfHnHK2bnnDtL0hWSpku61Tn3sPf+5Ao3qyDTpkmf+lQ6mRwAAADyU5a+4A9+IH3842MrY/Vq6Re/kF57TWpokL76VenSS6Vdu6QXXpBeecXmp5k0SXr7bctC3Ncn/fCH0r77Wkfx0EPt2B07pM2bpWOPlZ591uaz6e1Nf54zR9q61cIxH/qQ9Prr0qZN0v77S8ccY/MLHnus9JWvSBdeaI9Cv//90pYt0nHHSc88I334w9LZZ0sXX2xzLi5caJ8vukjats3a9PLL9rprVzoD86xZlgm5p8e2SdLKlZaFOZWS7r9fevFF6fjj7fVzn5O+9jXpwAOlyy6T7r47M0FKd3f6NWRjlmyeRCmduTkcE/8c3vf32/mtW5ff/IphLsZc2zo705+jj2eG+uOi7ZKkgQHpnntszsoFC6yuBx6w+5OtjFztCfu2tlo54fuVbztGeo3WG7+2ra32/enutj9o4ucffz8wkLuuke5J/Jix7lcKw31nxqpU5zWaNlfyGgOoWlWVxbhQo8linEqlp04JfarLL5cefZREbwAAoD7VShbjQhXcF1y+XLr2WstgvHt3OnuxZIG9ww6zwF5fnwXJdu2SrrxS2r49s6NIlmAEfBcAAFUu335gVYwgLKf4j4WnnJL5CgAAgDoVRkjt3m2vITgo2Ui9bdvSn52zUX4hOBgNAoVkIVLpRhvVmqQGyko54gwAgDIqWxZjAAAAoGJSKXs8dTivvGKLJL3rXVJzsz0mfM45FhDs6rJtbW3p7MYwIWja01PpltQfstRm4noMj+sDYJQSEyB0rnTL+PHSzTdX+gwBAACQU0+PzeHnXH77P/KI7b9mjXTnnbZuyxZ7/PjjH7dHUg44YPhO4n772WPM++wjHXSQzS/X0WFldHTY59mzbe5BydYvWWJzFXZ22vrFi22/446zR6SPOcbmKzz8cDt240YLBvT2SosWSUceaeuvvlqaP9/W33yzzYs4ZYrNMzh7tgU9DzlE+uAHrexzz5Vmzky38eqrpcZGWx/au3ixtSG8Llli9Z96qnTEERZQ3bbN5iScOdPmRJw50/bt6LD1c+dam/r6rL3HHWfbUilbt3ixrb/6amnqVDvfUEd47e1NB0BSKbtWnZ3pgEgIkPT1Ze6XK2gS3Rbfb7ht+WwfTj77p1LS0qV2jcI8SaVSC4GlUl2PWjj3fNVTsL6e7gtQC/LJZFKtSzVlMW5szLspAAAAZaUayWJc6FJQFuOOjrF3+IbLFlzIEs9WO2+etTG+ftas/Dqh2bIVT5jg/3/G3/C+kCV6TLxd0SVkUo7XP9wSzY4crSNk1I3XH+qI17VqVWY58cy6YVs0K280M3IwXPbefDP75to+nHz2j2YOXrUqv3JHqxYy25bqeozm3EuZ8XgsqrVdo1EL30mgBuTbD0zcHISlMG6cdM01lW4FAAAAclq7duxlROcszMf06TbyZeJE6Y03bBThpz9tI+oOO0y64w6b5zCM9Onult56y7LY3nuvZQV++mk7bsYMGxH46KPS88/bqMTnn7csxC0tlmX44ostscof/mBZhbu7pY99zBKtjBtnowg/+lHpsccsM/JDD1k5DQ2WlGXDBmnnTunEE6Vly6S/+ivpzDOlb37T2vvAA9J73ys9+aS9bt9uGZGvuCKd5fi116SXXpJuv13ae29rS3Ozzf/40ks2GjNkR371VUsKM2GC1b1ihX323uq/+GLL5nzJJVbHBRek67r7bsuou2GDlb9w4dDMrK2tNiIxfN682fbv6cmcNzBbRtdcGYFH2i/+fjj5ZJKNZg5ub8+v3NGqhcy2pboeozn3ap2LtJ7mxayF7yRQRxKTxTg8fXDllfmVPWmS9UE6OpI1zzIAAKg/ZDGWBZlWrRpdRc5Z0GrcOPu8774WVPubv5H+9m8tcHjSSdJ73mOZ79asseDcoYfa/oUm8Aj7t7ZK69fnPi6fcgupu1KJRkZb72iva9ISqaA0+D4BqBH59gMTEyAcDv+2AwCAekaAcFBfnyUd6e+3zxMm2Ii1xkbp8sulSy+1UXvXXpsO7o1VuYJfAAAAWeTbD+QRY9XXKGwAAADk0NaWDg5KFhyU7JHFL3zB3j/9tPSRj0jz5tljuE1N9ijjaIN0o30MsVofXwQAAHUpMVmMAQAAkGB9fZaFOB8vvyzdd5/NTbN69diygba1SV1d9potu24+xxVipKyf+WYFLUf20LFkAK7GOkdz7Qtpw2jbO5bzTFIW2SSdK4DSqeF/SxIzgjDMi1wqs2ZJ55zDnIUAAABVafFiSxQyGh0d6dF8s2fb/INPPWXlTZggff7z0tVX2xyFU6ZI7363NHmyJRV54gnpwx+217PPll54wYKOAwPStm1W1pYt0ic/aYlFzjjD6rrxxnT90TkJb7zREoE8+aT0s59ZgpLZsy3ZyV13WYKRa6+1kZLTp2ce09AgPfOMbf/Rj2x+w6YmC1i2t9tj1Q0NtjzxhHT99dINN1ibonMrTpuWbs8VV0i33Sb98IeWUCT+WHQqJa1bZ6+PPmrXInSYwyjJgQHpwQctgYg0dMRk/PylzFGdubZLQ9sTHZkZ7u3AgD1mHvYb7ePdoezNm6Xrrht6bHxUaColnXde5nkPV/dwo0qjx0l2zcN1GMto1HodyZrtOtfrudaqXP8tMP0Cql0t/1uST6rjal2OOuqoAtI6l2chAzsAAKg2kh7yVdB3K/ZSSF+wbJ3BXMu8eUPXzZmTfj9hQvp9S0tm57Kra+j6XMu0afba3Jz9mLA92nHNVm5DQ/b2t7Rkb8+8een10Q5xWJetw9zfb+9XrUqX198/9N5lqy9bHbmuW3TfUGd//9D6w37ZjstHf3+6DdmOjdYdb3d8XT7HZ7tG0fMOn4c7Lp9zGu2x1Wyk7wYqL9d/C6P97xMolyr8tyTffmBiRhBOnVqeEYRkYAcAAKhC555rI+KGM368tHv38PuMdgTh2WdLX/yi9Pjj1nHcf3/p29+Wbrkl+wjCBQusvmjnsrXV1g83gvCKK2ykXsiAHD2moUE69ljpggukk09Ol93dLb31VuYIwvnzpS99yfabO3foCMLQnlmzbARhT4+NIIy3ua3NRuhFRxCG7WEi8FQqcwRfXNg/nEu2OvLZHq0ziNcf3b/Qjn1Tk40cjI7ki2+P1h2tJ5z3cHUPN3F6/LiBgcyyRzuKpV4na892nev1XGtVrv8WRvvfJ1AuNfxvCVmMAQAA6hxZjGVBoKOOkrZvT6/bay9p1y4Lcu21l/TIIxa4+/3vpaOPlk49dWwJSgAAACqMLMYAAABA0NOTGRyULDgoSXvvLT38sL1/800bkffQQ7ZElTNYyDxbAACgjMhiDAAAgPrX2pp+BDbukUfS73futMeHJam52V5Xrx57NuNChUnOy1knAAAoXA1nLo5KTIBwyhTJudIt++8v9fbWzfcCAACgvpx/vs0HmE2uKXc2bbLAYNDRkdkBbGiQPvEJ6UMfsjnsJk60+fre/W7pgAOkQw6RLrnE5iN897uts3jzzbbvySdLM2dKxx1n8/pNmWKvp55q+23bZiMZt261kYsLF0qLFlnG4d5eO/aQQ+y9lO6E9vZKS5ZYu6ZOlZYvt2M6O6397e1WTnu7bZs7V9q40bZ3dqb37ey0sk491dZJ9rp4sbW5o8POZb/97PhcUql0ealUup0bN9o8hxs32ue+PisztC26f7RtYd9oh7uvL33dwvpsnfKwLnp8KL+jI11nrvOIlpetXbnqHa3hzqFYf2xU4o+XWviDqRbaWAlcl9HhutW/evlRL59MJtW6VFsW41zJ2wAAACpJZDEuT2cw2+JcZmexsXHoPnvskfkazW4cX1paMjMiz5tn5xc6odmyJY+U/Tia2TiemTis835oOeFcpk0bObtuPMNuOHb69NxtjGfkDfvEMwWHz6G9uTIYx7Mdx8sfrhMfLy9bu3LVO1rDnUMxyh8p63IpVKLO0eCPuuyyZd7OpgqzuFZUyJS+alWlW4JSqfLvfL79wMTMQdjYmE7mVQr77Zc7eRsAAADq1OGHWybjvj7LgDxjhnU633lHmjTJRi5eeqm0777WWdy2zdYdf7yNDjzoIOmcc6RvflP66Eeln/7URheedZb0wANWvmSZjsePt1GGL78s/fmf2yjGMFohdD5POEFaudKSrtxxh5UTsiK/9potjz4qHXGE9Prrlvn4xBOla6+1R6q7u9OZgKOZiyV7HRiwjMfHH29JXy64wB7f7uiwfeKZG0MW42gbN2+WNmyQpk+XfvCDdNblww6T7r/f2tbUNDQj79KlNoKxtdVeo1mYJTvvu+/OncE4mu04evzAQDrLc65OfLYMx/F25ap3tLJlay1mBteeHrsPLS3l++OlEnWOBplys2trS//329OTO1NrGE0l1Ww2V6AgNZy5OCpRWYxnz5aefrqwOq66Svr85wtsGAAAQBUhi7EsiHTRRfZ+yhTp1Vct4LZ7t61zTho3zhKWvPOOzUU4caK9P+cc6TOfsWDYfvtJ995rwbSFC4ubuKRSiUnGWm+hx5OApTpU4j5w72tfPveQ+5yJ64EKy7cfmKgAIQAAQBIRIIzo7MycV3DKFBsBeNZZNoruxBNtVN2XvmTz/MX/oOMPPQAAUEPy7QcmJkkJAAAAEi6Vkn7848x1r75qj+zedJONDrznHlv/k5/YI3Lr1mXuHx597elhwnkAAFA3EjMH4ZQppZmDcPx4my7miSdsKo3OTn5MBgAAqErr1kkPP5x9W5ifL3jjDXt99lnLPhkdMbhunY1CvOoqCyJefLE0bZr0u99ZxmLJ5uk77TRpn31sfsJ166RDD01nv/3Zz2zeusmTbfuaNTZ/XmurtH69zSW4Zs3QefXWrUvPlxceb06lsq+Xho54TKWkri6b0zDapp4em/vvK1+RLr/cMgKH6xHKlLKPnoyXOW1aZntOOcXO48gjLXPy+vXZR2DGz2Pp0tz7xo+RbP8bb8xsd64ywjmH612OEaHVNPq0FG2JlikN/d5Vy7nnUgttBIASSkyAsFQJSnbvlm6/3d5feaU9iVIHc1MCAADUl9NPtwQg+XrxRXu95hp7DRPuRz3xhPSFL9j7l1+211dfTW+//vr0+/Z2C2SdeaYlJ5Gkp55Kb3/kEam/3x5z3rpVmjfPXn/zG2n7dktu0tSU+Xh0cM890qZN6c+plAXHtmyxoNzatba+rc0CZmHf88+Xliyxzuy996bnZFy2TNqxI7OOkPFv9Wpr5/Tp6UBKT0+6jgsvtMBntJ3r1ll5mzbZyMxw/vFOc09P5nHXX2/XeGDAgqpRITAYPfcHH7TkCVFh3cCAnUMIvB52mLU5JFyIticaPLzxRls3lrkmQ3nh+mU791LKFgwtRRKJUObmzZboJnquYVu4D6UKTI6lzEKvSSH1EnysX9xb1JHEBAhLJT6CkERXAAAAVaiQ4GApHHmkBc+2brVMx/39lkV4xgxpzz2tMxmCgi0tNuJuzRrpscfs+Ntuk849N11ec7O9hiBMc7P05JOWke+226y8oKsrHRSKBhJfeCEzILd7twVvzjwzHRgNZbe1pUfqbdmSLmfFCtvW32/ru7ttBOHPf277zJmTbsvs2enzy9ZpbmvLflw20WBiSBizdKkFpn7wA7sOc+ZYexYvtsBUR0f6Gr/1ll2XeEbjUHYIdIXgYWPj6ANpobxVq9L3opyynU8psvRGM9wuWJB5rtGM0aUKTI61zEKvSSH1ktW3fnFvUUcSEyA87jj7YXQsyGgMAABQo047LR0kHDdOmjQp/YjJhAn2ed99pWeekUZK4jdhgjR1qvT731v243Ds4sUWmHr74YwalwAAIABJREFUbdtn/nwLcn34w/YHZBiV191trxdemA6oZXvc9dZb7VHfENybMyfdhvDIb/Tzjh1W5sqV0i23ZD5GLFk5AwM24iU81rxxo33+9a8tUPm971l7Dj546CPL7e0WKIu2U7JtXV2Z7bnxxqGj8EZ6ZLipKftx8XONnkvYHsrr7LR6wrU99FD7oz2VyhxBGLZJQ/+oD+fV2mqBrui60YgGnioxwih6PiEY2tRU/GBGU5N03XXZR1OF+sJ9KHZgMvo6WoVek0LqLUVAFtWBe4s6QhZjAACAOkcW44i+PunYY6WdO6W99rLl2Wdt26pVFmBauzY9IqSri1Eho8WjdwAAVFy+/cDEjCAEAABAwqVSNpJw5077vGtXZmKS733PPi9fnh6dxqiQ0ePROwAAasYelW4AAAAAUBZdXUPntdu9O/3+d7+z0YOf+lRm4LCvTzr1VHuthFTK2pVKDb+u2rS1VWbOvZEU49pV+juRzUjnVQ3fmWpoAyqH+w9UtcQECBsabIqYYi7Tptm0MAAAAKhyxxyTzrQ7kocftn1Xr7ZsvfPmWeKFefNydwwbGqQpU+z9HntIl1xi89dNmWLzEe63nzR3rjR5svSJT9ijzGEuwBBo6u21fdrb7f3ixTaRdmurjcTr7LTPM2fato4O6eMft/0XLrRty5dbZuLeXtu+YIElB9m4Mf2HefgjvbfX6u7tHbqtry/92t5u7YqWEaRS1q7QvnBcKpWe023HDmvT5z4nLVpk5YXzl9LXILQ5nNtwQYRooCG0ob09fe5hW7ZAXhjZeN552c8lfh7ZXHihfSfa2zP3S6Ws7CVLrIzhAiLxbfkEHePnHT0+nFdPz9D9+/rsfOPbs5U93LlH64y3N58A5UhtqGf5BsdKEUSrlsBctu8ogOrhva/Z5aijjvL5stmmi7/Mm5d3EwAAACpC0kO+CvpuxV4K6QuWrDOYa3Fu5H26urxvabH3LS3WsYx2MuP7z5lTWCc1+nn69HSdXV2Z+4TX6LZou7KVEYT9wxL2j+4TLSN+/tHt8TZHy4gL9UbbHD/3+PUN+vuztzNaTrbtUVu32j4rVmTuFy8j2s7hziF6HaJtzee8w/H9/fa+v3/o/tGyo9uzlT3cuUfrjLd3uHONl5+rDfVspOtT6H6lqLvUsn1HAZRcvv3AxMxBOHmy9PrrxS1z6lR+/AAAAKgJCxZIDz6Y//7jx0t//KOFevI1YYJlMHbORqH95CfS449Lb75pWY733tsSorS0SEccYY/etrbasd3d0ssvS5/5jHTyydKyZdJXvyq99Zb0oQ/ZaLxTTrGRa889Jx19tI10XLzYOrpbtlhb58+Xtm+3TL233CLdcYeN4PuHf5AefTTzcd+Q0XflSunuuzO3hYy3ra3SrFnSbbdJV1wxtIyQTThkPF66NJ0pN+jutvOYOVP67W/t3Jua0vuErM4h+/IDD9iIz+EeTY5nDo1nZw7nE72+QTzbbvxcpOznEXXooZZlOpWyUabR9vT32/0IGaqj7RzuHEIbo20d6Zjo+2xZeHNlMB6u7Oi+w9Ufv7YjZXOtdDbnSss3220psuJWS6bdUmTPBlA0ZDEGAACoc2QxHtTZaY8Nz5wp7bOP9OKLFuB5+21pzz0tsPfmm7bviSdKd95pj9UuWyadeab07W9nD2xUS7beamkHAACoGvn2AxMzByEAAAASrr1dOv98Cww+8oj0wgsWHJRshNv++6f3dc4CbqecIl18sXTNNbkfHQmjYiodlGN+LwAAMEoECAEAAFD/QgKKnp70vDMTJqS3T5xoj+YedJA9DnznnbZvKmWPjDY32yOV1TLZfzbVmjUYAABUvcQECHt7LaFcMbIX7723dMghZDAGAACoGSedJF15ZeacgmH0oJR+tPi552wuwPHjpcsus/nl1q6VNm2yLMbTp9sovf32s32ckxobpcMPt7n19tnH3i9ebJ3FkOE3ZPkdLjNsyB67cWM6m3G2TL/xzLdLllj5O3bYuh070tl4e3stc/DChbZPX196W7aMxtHMu2HfbO2PZ9MNmXvjGZFz1RXaHK03m95ey8J8zDGZmY2zlR39HG9jrms+XKA3n6zChZaZre1jLa9Y7SpF3dWunjMGA0Ch8slkUq1LIZnr9tor/6Rv+S5z5uRdPQAAQMWILMbF7wjms0ybNnRdNINryOi5apVta2621333Te+fK9NvfP1ImYejHdho2fGMxtHMu6E92TILR7OihvbHy4muz1ZXfP9s4pmNm5vT1y1b2bnaGJdPVtd8sgoXWmZ0v5H2Ham8fDPCjiaDbbVkvS21Yp9nrgzZAFBB+fYDE5PFeNeu4pd58snFLxMAAAAl4FxhGYlHY//9bdTQH/9on3fssEeTjzzSPm/ZIm3YYI8ur1iRnjNw1Sp7NLi/30YqzpljGZc/8pHsmX6jGYdDhuAjj5SWL09nHl6wQLrnHitv331tVOTs2dITT9hxzc3ZMxpHM++G9px4ovThD1uW4ng21LY2ad06ez9njp3TmjV2ngsW2LlF9w91rVxpbT777HS92fT0SJ/+tI32fPZZa09IxBIyDkczGcfbFq87yCeraz5ZhQstM2yPt3U05YXvjzR8ZtjRZLCtlqy3pVbs8+zpse9+S0v9XzsAdScxWYw/9zmbW3osFiywvtUVV0iPPkqCOAAAUBvIYizp5pulpUuHBgknTZLeeMPev+tdNpfMSSdJkydLDzwgPf+8ZTjets0eH/Ze2rnTgmVnnGFzFkr2ub3dgoLt7dKhh1pHsb093WGMZxnO9bm1VVq/fuydzXh5ra3SjTfatmi7Rjp+pHaMdF7FkEqlA5H5tD0JyFpdfbgnAKpQvv3AxAQIAQAAkooA4aDOTmn1ahsRd/DB0o9+lE5YEjQ3WyAqGqDjj34AAFCj8u0HJuYRYwAAACTc0qX26O4FF0jLlqWDgw0N0muv2ftNm2yE2qZN9ohtQ0P6Ud3Nm6Xrrqu9ICEBTgAAMILEZDEGAABAgqVSFvjbsMHmvQsZf6V0cPDAA2104X772edf/tJGHIYMxmH+wFoT5qqrxbYXSzEyy2YrYyzlkkEXKB6++/WF+1kRiQkQLltmc1OPdWlqko47Ttq4UTr1VKmvr9JnBgAAgBGdcIIF+qShjxUHL7wg3XVXep6+++5Lb9u61V47OtIdwz33tNfJk6UpUyxJyXHHSb29ljBk771tjsL2dvsjJ/zB09trHclcHcroH0Z9fdKSJVZGR4c9Jh39g6mvz8q4+WZp/nwrO5TR2WlLa6slC9m2zcrq67P93vte6ZhjhpadStnnJUusjfH6Q/tztbuvL3sgrb1dmjvXjo+2L/4H4Gj+MMx1TFi/bl06SBrdt5C6sgVaCwm+xusqduA2lZLOO688weDh7l89qbYgRbW1p5rwQ0h94X5WRj6pjqt1OeqoowpI61zcZfp0e21pybsJAAAAFSHpIV8FfbdiL4X0BYveGRxumTdv6LquLlui23N1KMN+XV22LVtZQdje2JguO1pGvO5QX7Y2Zts3tDHb+eVqd2hTtJ3RMufNG9q+XOefr1zHhPWrVtn7/v7MfQupKxzb3z/8unzbWMix+Yhe/2KVOVJdhd6nWjOa72IpVVt7qkmx/3tCZXE/iyrffmBi5iA891zp+uvHXs60adKf/In0zW9aNuPu7rGXCQAAgBI77TTppz/Nvs25odmN8zFhgvT225YJedw4qbHRRuV1d0tXXy3967/ayMKzzrL5/4ITTpDWrLG5ELN1KMO+bW02+u+ttywrckODLdGywrGf/ayN5gqjLdrapIGBzPL6+6UtW+yYl1+2DvK0aTaKMF522Peii6TLLsus/5RTrP252t3aKi1enNnOtjYbwXjbbdbGOXOGti/b+ecr1zHR9WH+xWz75lNXU5O0YsXI6/JtYyHHFlp+qeeazPb9qkej+S6WUrW1p5oU+78nVBb3syLIYgwAAFDnyGIseyRv6VJ7zHjmTFu3bZsFyHbssOzFRx5p6xsa7HFYEnoAAIAal28/MDFzEAIAACDBenrScxB+8pO2zJ5twcHZsy04ePbZ0mOPWSBxpOAgc4EBAIA6kphHjAEAAJBgbW322Oz990sPPCDdeWd621NPWbDv5pul7dulnTulM8+0R2XXr8/+yGaYQF3iMSiglqVS9t9zOR7NBoAqlqgRhGPNYLzHHtL73299QX4sBgAAqCH7729BwLvusuDgpElD99m+3V7vvdc6fPPm2ev06UM7hn/3d7bvV7+anntw4kQLMMycacHGkF34ve9NZzeOjjoMGYj7+oZmDl60yObxC9uix0X3DZmEw7rFi+3x6JBtOGQg7uy0fTs6rOyQlThkJs6VeThkqo1vj2dk7u3NvW84z+g+I2UQzrZtuOzDo81MHK+zs3Notuh4edH7Vo+KmUW6XMePBdlSAcDkk8mkWpeCMtf54ianI3EUAACoFSKLce5O3YQJ3jtX3I5iNKtweI1m/w0dyZDtt6Uld+bg6LZwXDwjcXxdrmzK2TIix7flyjwc3x7PyBytK75v+BzdZ6QMwtm2DZd9eLSZibPVGe/wx8uL3rd6VMws0uU6fizIlgqgzuXbD+QR4wI4Jx1+uCVuI3EUAABADdljD+mdd9KfZ82SWlqkm26yeQjHjZP++Mf8y5s6VXrpJTtu4kTpgAOkZ5+V3vUuG1HY1SX94AeWXfjrX7cRjN3d0t13pzuSIQtwd7clS4lmDr70Uut8hm1SZgbTaEbi6LoHHrAOa1OTdVpXrrQMxE1NNrfiYYfZY9ZHHGHJWCR7Xbo0e+bhkKk2vj28hozMK1fayMds+4Y2RvcZKYPwcJmGh1tXaGbieJ0DA9Jrr2VmdI6/Ru9bPSpmFulyHT8WZEsFAElkMQYAAKh7ZDGOiM43Fh4tbGyUrrlGevxx297XJ61bZ4G1+DHMUQYAAGpIvv1ARhACAAAgOZqaLNDX1SX98pfSQQdJzz0n/epXti4EA6dNS7/v77f50QYGbG46AACAOkOAEAAAAMkQD/hFdXdboPC55yyJycCAjSwMST0ke/QUAACgDiUqi3HQ1ye95z25sxVPnCiNHy9dfXWlWwoAAICi6OuT5s+3gF88OCjZ/IM33GDBQUlavdr2lSzzsWTHzZ4tnX66NHmyzePX0WHz6s2dKy1bZsHEc86xOQqXLcueDTdbxuC44TIIj0W03Hzqb28fmi05ni04V/uyZRrOVfdIWWzzzTCcq5x89ssnk26+2XbLmZW3khmAUd34bgAoRD6ZTKp1KTSL8WWXjS4J3VVXFVQNAABAVRFZjEfXCcx3mTp1+O3NzelMqeFzdH022TIIr1qVXrZuzcy82t+f3tbfnz0za39/Zibjri4rp6XF+/vuS3/u6rJy4lmOo+1qaUnvE83qG60vbF+1aug5xbPVjpTFNp5hOF5mrozD8eNDG4fLkhw9j+h1jF6/kbLtjnQ+xcycW2gG4HzqLlVm3/j1rPXswdV+Dtm+0wASJ99+YFU8YuycWyvpdElvSXpSUpv3fmex6xltcqovfUn6/OeL2xYAAACYcvUFi27CBOntt21U4UsvSQceKL3wgmURfvJJ6fXXbb9Nm9JJTiR7xHnTpvT6bJ3UbBmEBwZsZKMkPfigtGGDvV+xwsoJ2xob7TWMgAzl9/TYMc3N0sKFVsd559m6p56Stm6VNm+2z6tW2RKStkSzJYd9FiyweRtbW23bhg25zyd+TvFstSNlsW1rk37+c7tmwx2Xq5xou6P3Ip71OLpPuK7hOoZzbGkZOdvuSOcTLXesGXQLzQCcT93FbF+ucqXS1FFOpbpOxZLtOw0AOVRFFmPn3EmSfuG93+2c+44kee8vHum4QjPXdXdLF11UePuuuooAIQAAqF3VnsW4LH3BxsbizSE4bpw9XvyP/yjdfbcFyNavT7+GQM26dVZnQ4M9qhsyIKdStk3KXD+S6HFLl6bramoaWqY0NPNytmzMfX3ShRdKK1dmnstwGZuzlZPvurEY7XUbrt0j7RP9LBXvfCqZGXs016EUdUu1nx28FjKc10IbAZRUvv3AqggQRjnnzpJ0jvf+3JH2LTRACAAAkETVHiCMKktfMB4UyxYMyvYZAACgxuTbD6zGJCXnS/pZro3OuS845x5yzj3U399fxmYBAACgDErbF0ylLDi4YYMF/jo6LKnIokWW/CI8AiulHx884YR0ko5qUM+JB+r53AAAqGJlm4PQObdJ0gFZNn3De//jwX2+IWm3pOtzleO9v0rSVZL9alyCpgIAAKDIqqYvGObhmzbN5tybNi2duThkK06lpJkz7XHba6+1/S68ULr11oKrK4lqn/dsLOr53AAAqGJlCxB675uH2+6cO0/SaZI+5qvtuWcAAACMSVX0BVMpSxAya5b09NO2bscOaeJE6c03pXe9S3rlFem226QnnrDJ/S+/XPrKV+xx5GpRaFKK0ajU49XlODcAADBEtWQxPkXSxZIWee+LNHs0AAAAakHZ+oI9Pfb4ajB+vLR7twUHJcs+3NgorVljS8gSvHWrzVV47LEla1pBmppKP7quUiP5ynFuAABgiKoIEEpaJ2mipNudc5LU673/y8o2CQAAAGVSnr5gW5vNMfjaa9I//ZP08svSpEnSG2/YqML77pO6uiwQeN11FiRrbZUWL7Zjk5S0hJF8AAAkStVlMS4EWYwBAABGVktZjAsxpizGJ50kbd8unX++NG+eBQLXrx8++Ld2rY2q6+pilBsAAKgJ+fYDq2UEIQAAAFAeF15owcF58yzgt369JSsZKejHqDoAAFCn9qh0AwAAAICySKWkzk57nHjRIun006Ubb7QgYU+PbV+71l6zCfPj1fvjxQAAIHEIEAIAACAZ1q2TVq+WrrxS+s1v0sHAlhZ7xDgk5jjvvHSQcKSgIQAAQB0gQAgAAIBkmTZN2rHD3vf1WbbiMP9gc7N9XrfOtoegYU9P5doLAABQYsxBCAAAgGRob5caG6UTTpBWrpSOPFJavtweMx4YsH0WLpQ2bUofw7yDAAAgAQgQAgAAIBnCHIKSdPvt6fWNjTZKsLExHUQMAcHoMQAAAHWKACEAAACSLTpKkIAgAABIIAKEAAAASDaCggAAIOFIUgIAAAAAAAAkGAFCAAAAAAAAIMEIEAIAAAAAAAAJRoAQAAAAAAAASDAChAAAAAAAAECCESAEAAAAAAAAEowAIQAAAAAAAJBgBAgBAAAAAACABCNACAAAAAAAACQYAUIAAAAAAAAgwQgQAgAAAAAAAAlGgBAAAAAAAABIMAKEAAAAAAAAQIIRIAQAAAAAAAASjAAhAAAAAAAAkGAECAEAAAAAAIAEI0AIAAAAAAAAJBgBQgAAAAAAACDBCBACAAAAAAAACUaAEAAAAAAAAEgwAoQAAAAAAABAghEgBAAAAAAAABKMACEAAAAAAACQYAQIAQAAAAAAgAQjQAgAAAAAAAAkGAFCAAAAAAAAIMEIEAIAAAAAAAAJRoAQAAAAAAAASDAChAAAAAAAAECCESAEAAAAAAAAEowAIQAAAAAAAJBgBAgBAAAAAACABCNACAAAAAAAACQYAUIAAAAAAAAgwQgQAgAAAAAAAAlGgBAAAAAAAABIMAKEAAAAAAAAQIIRIAQAAAAAAAASjAAhAAAAAAAAkGAECAEAAAAAAIAEI0AIAAAAAAAAJBgBQgAAAAAAACDBCBACAAAAAAAACUaAEAAAAAAAAEgwAoQAAAAAAABAghEgBAAAAAAAABKMACEAAAAAAACQYFURIHTOrXHO/adz7mHn3M+dc++udJsAAABQHvQFAQAAKqsqAoSS1nrv3++9/6Ckn0r6ZqUbBAAAgLKhLwgAAFBBVREg9N7vinxslOQr1RYAAACUF31BAACAyhpf6QYEzrlvSfqspD9I+ugw+31B0hckacaMGeVpHAAAAEqKviAAAEDlOO/L8wOtc26TpAOybPqG9/7Hkf2+JmmS937VSGUeffTR/qGHHipiKwEAAOqPc+5X3vujK9wG+oIAAABllm8/sGwjCL33zXnueoOkWyWN2CkEAABAbaAvCAAAUL2qYg5C59zcyMdWSVsr1RYAAACUF31BAACAyqqWOQi/7Zw7VNI7krZJ+ssKtwcAAADlQ18QAACggqoiQOi9/0Sl2wAAAIDKoC8IAABQWVXxiDEAAAAAAACAyiBACAAAAAAAACQYAUIAAAAAAAAgwQgQAgAAAAAAAAlGgBAAAAAAAABIMOe9r3QbRs051y9pWxGLbJKUKmJ5qE7c52TgPicD9zkZuM9jN9N7P73SjSg2+oIYJe5zMnCfk4H7nAzc57HJqx9Y0wHCYnPOPeS9P7rS7UBpcZ+TgfucDNznZOA+o1z4riUD9zkZuM/JwH1OBu5zefCIMQAAAAAAAJBgBAgBAAAAAACABCNAmOmqSjcAZcF9TgbuczJwn5OB+4xy4buWDNznZOA+JwP3ORm4z2XAHIQAAAAAAABAgjGCEAAAAAAAAEgwAoQAAAAAAABAghEgHOScO8U51+ece8I599VKtwfF4Zy71jn3onPuN5F1U51ztzvnfjv4um8l24ixc869xzl3h3PucefcY865Lw+u517XEefcJOfcA865Rwbv8+rB9bOcc/cP3uebnXN7VrqtGBvn3Djn3Bbn3E8HP3OPUVL0A+sXfcH6Rz8wGegHJgt9wcogQCj78km6UtLHJb1P0lLn3Psq2yoUyfclnRJb91VJ/+G9nyvpPwY/o7btlnSR936+pGMlfWnwv2HudX15U9Kfeu8/IOmDkk5xzh0r6TuSLh+8zy9LWl7BNqI4vizp8chn7jFKhn5g3fu+6AvWO/qByUA/MFnoC1YAAUJzjKQnvPdPee/fknSTpDMq3CYUgff+LkkvxVafIem6wffXSTqzrI1C0XnvX/De/3rw/Suy/5kcJO51XfHm1cGPEwYXL+lPJf3r4Hruc41zzh0s6VRJ/zT42Yl7jNKiH1jH6AvWP/qByUA/MDnoC1YOAUJzkKRnIp+fHVyH+rS/9/4FyToUkvarcHtQRM65QyQdKel+ca/rzuDjBg9LelHS7ZKelLTTe797cBf+/a59fy+pQ9I7g5+niXuM0qIfmDz0D+oU/cD6Rj8wMegLVggBQuOyrPNlbwWAMXHOTZH0b5L+ynu/q9LtQfF57//ovf+gpINlo37mZ9utvK1CsTjnTpP0ovf+V9HVWXblHqOY+I4BdYB+YP2jH1j/6AtW1vhKN6BKPCvpPZHPB0t6vkJtQen93jl3oPf+BefcgbJfoFDjnHMTZJ3C6733twyu5l7XKe/9TufcZtlcQ/s458YP/qrIv9+1baGkVudci6RJkvaS/YrMPUYp0Q9MHvoHdYZ+YLLQD6xr9AUriBGE5kFJcwcz4+wp6VOS1le4TSid9ZLOG3x/nqQfV7AtKILBeSmukfS49747sol7XUecc9Odc/sMvp8sqVk2z9Adks4Z3I37XMO891/z3h/svT9E9v/iX3jvzxX3GKVFPzB56B/UEfqByUA/MBnoC1aW856RmZI0GKH+e0njJF3rvf9WhZuEInDO3ShpsaQmSb+XtErSv0v6Z0kzJG2X9Gfe+/jk1aghzrnjJd0t6VGl56r4umz+Ge51nXDOvV82KfE42Q9c/+y9v8Q5N1uWVGCqpC2Slnnv36xcS1EMzrnFkv7ae38a9xilRj+wftEXrH/0A5OBfmDy0BcsPwKEAAAAAAAAQILxiDEAAAAAAACQYAQIAQAAAAAAgAQjQAgAAAAAAAAkGAFCAAAAAAAAIMEIEAIAAAAAAAAJRoAQAAAAAAAASDAChABKwjn3feecH1zeds495Zz7rnOuMbbf2c65XzjndjrnBpxzjzrnvuWc2y+2357OuX7n3CvOub3zbINzzn3OOXff4HG7nHO/ds51OOf2Kub51jrn3Gbn3Lo89jvRObfeOffc4L39izI0DwAA1BD6gbWFfiAAiQAhgNLaJOlASbMl/Y2k/y3pu2Gjc+5bkv5F0sOSTpP0PklflnSIpC/GyjpT0tOSeiV9Os/6fyjpCkkbJH1M0vslrZT0UUlnj+J8IE2R9BvZfXq9wm0BAADVi35g/aEfCNQz7z0LCwtL0RdJ35f009i6qyW9MPj+GEle0oU5jt8n9vk2SRdI+oykh/Ko/5OD5Z89XPmyH0pWSnpG0puSHpV0RmS/QwbL+ZSkO2WdoS2yTubhku6VNCDpl5JmRY7rlHWgPidp++Bx/y6pKbJPvnV/QtLtkl6T9F+SlsTO5X2SbpX0iqQXJd0o6YD4vZB15p6T9LKkHkkNke0+thySxzV+VdJfVPq7xsLCwsLCwlJdC/1A+oEsLCy1tzCCEEA5vS5pwuD7c2Udqiuy7ei93xneO+dmSlos6SZJt0ia55z74Ah1nSvpv733t4xQ/pclrZB0saQjJP1I0i1Zyl8t6TuSjpS0U9INg23/hqyTO0nS/4kdc4ikZZLOkNQsaa6kayPb8637W4Nlf0DSg5Jucs5NkSTn3IGS7pJ1Qo8ZrGeKpPXOuei/8SfIOrLNkv5c0lmD9Yd23CfrLB44uDwTv2YAAABjQD+QfiCAKkaAEEBZOOeOkT0S8h+Dq+ZKetJ7/3Yeh58v6Xbvfb/3fkDWOVw+wjFzJW3No+y/lvRd7/0N3vv/9t5/U9Ldg+ujur33G7z3WyVdJukwSVd47+/w3j8maZ3skZWoyZI+673f4r2/R9L/knS6c25ugXVf7r3/iff+t5K+LmmqpNB5/KKkR7z3F3vvH/fe/6ekz0paIOnoSBm7JH1xcJ+fyx7p+Zgkee//IOktSa957383uPwxj2sHAAAwIvqB9AMBVD8ChABK6RTn3KvOuTdkv0zeJXs8RJJcPgUM/vr5F7J5ZIIfSjrXOTdpuEPzKHsvSe+WdE9s0y9lj2tE/Wfk/e8HXx8L4nIYAAAC7UlEQVSNrWt0zjVE1j3nvd8e+Xy/pHckzR9D3c8PvobJu4+SdOLgdX7VOfeq0r/6vjdy3H9573fHysmYABwAAKCI6AfSDwRQQ8ZXugEA6tpdkr4g6W1Jz8d+Jf5vSSc45/b03r81TBknSZoh6Xrn3PWR9eNkc7Jcn/UoK39+nu30eax7O8u2bOsK/eGloLq99945F61nD9m8M/Ffm6V0BzajjEgd/EgEAABKhX5gkeumHwiglPhHAUApvea9f8J7vy3LIyQ3SGqU1J7tQOfcPoNvl8seJflgbLlawz9ecoOkuc65rFnqnHP7eO93yX5BPT62+XjZJNBjdZBz7j2Rz8fI/t19vIh1/1r2mMu2wWsdXV4poJy3ZJ1tAACAYqAfSD8QQA1hBCGAivDe3++c65K01jl3sKR/k/SspFmyDt8TzrkrJbVK+jPv/W+ixzvnrpF0n3Puvd77J7NU8c+SzpT94vx3kjbKMrvNlz3e8i+yrG1rJV3inPutpF/JJpM+QfbIxli9Luk659yFsnlo/q+kWwfnkFGR6r5S0ucl3eyc+46kfkmzZdn7Liqgc/g/ko5xzh0iy0r3kvf+nfhOg5Nizxn8uIekGYOTab8Ue4wGAAAgK/qBUpHqph8IoGgYQQigYrz3F0v6lKQPSdog+8V0naTtkv5BNsnym5Juy3Ls/bI5VrL+euy997LJsL8s6TRJd8jmirlU0p2yjqhkWeHWSuqSZYA7S9InvPcPF+EU/0eWce8nkn4h6SlJbZHtY67be/+8pIWyOW02SnpM1ll8c3DJ13dlvx7/l6xzOSPHfkdL2jK4TJZl9dsi6ZIC6gIAAAlHP5B+IIDq4uzfTgBAMTnnOiWd470/vNJtAQAAQPnQDwRQixhBCAAAAAAAACQYAUIAAAAAAAAgwXjEGAAAAAAAAEgwRhACAAAAAAAACUaAEAAAAAAAAEgwAoQAAAAAAABAghEgBAAAAAAAABKMACEAAAAAAACQYP8PpV2a8xApuGIAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 1296x504 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# visualization\n",
    "inlier = pca_x[label == 1]\n",
    "outlier = pca_x[label == -1]\n",
    "xmin = pca_x[:, 0].min() - 1\n",
    "xmax = pca_x[:, 0].max() + 1\n",
    "ymin = pca_x[:, 1].min() - 1\n",
    "ymax = pca_x[:, 1].max() + 1\n",
    "\n",
    "fig, ax = plt.subplots(nrows=1, ncols=2, figsize=(18, 7))\n",
    "ax[0].plot(inlier[:, 0], inlier[:, 1], 'b.', markersize=2, label='Inlier')\n",
    "ax[0].set_xlabel('PCA Component 1', fontsize=14)\n",
    "ax[0].set_ylabel('PCA Component 2', fontsize=14)\n",
    "ax[0].legend(fontsize=14)\n",
    "ax[0].set_xlim([xmin, xmax])\n",
    "ax[0].set_ylim([ymin, ymax])\n",
    "ax[1].plot(outlier[:, 0], outlier[:, 1], 'r.', markersize=2, label='Outlier')\n",
    "ax[1].set_xlabel('PCA Component 1', fontsize=14)\n",
    "ax[1].set_ylabel('PCA Component 2', fontsize=14)\n",
    "ax[1].legend(fontsize=14)\n",
    "ax[1].set_xlim([xmin, xmax])\n",
    "ax[1].set_ylim([ymin, ymax])\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For other methods, please refer to: [Link](https://github.com/stasi009/TakeHomeDataChallenges/blob/master/10.CreditCard/credit_card_fraud.ipynb)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}