v1.0-DataExploration.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# CapitalOne Data Challenge - Data Exploration"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Firstly, I loaded required packages, then imported the dataset by using pandas package. Before importing, I changed Null values to NaN in '.txt' file since Python does not convert Null values as NaN.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "#load package\n",
    "import pandas as pd\n",
    "import seaborn as sns\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "file_path= './Desktop/DS-master/cardtransactions.txt' # path of the file"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "#create import function\n",
    "def import_data(path):\n",
    "    dfList = []\n",
    "    df=pd.DataFrame()\n",
    "    for iot_chunk in pd.read_json(file_path, lines=True,chunksize=100000) :\n",
    "        dfList.append(iot_chunk)\n",
    "    df = pd.concat(dfList,sort=False)\n",
    "    return df"
   ]
  },
  {
   "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>accountNumber</th>\n",
       "      <th>customerId</th>\n",
       "      <th>creditLimit</th>\n",
       "      <th>availableMoney</th>\n",
       "      <th>transactionDateTime</th>\n",
       "      <th>transactionAmount</th>\n",
       "      <th>merchantName</th>\n",
       "      <th>acqCountry</th>\n",
       "      <th>merchantCountryCode</th>\n",
       "      <th>posEntryMode</th>\n",
       "      <th>posConditionCode</th>\n",
       "      <th>merchantCategoryCode</th>\n",
       "      <th>currentExpDate</th>\n",
       "      <th>accountOpenDate</th>\n",
       "      <th>dateOfLastAddressChange</th>\n",
       "      <th>cardCVV</th>\n",
       "      <th>enteredCVV</th>\n",
       "      <th>cardLast4Digits</th>\n",
       "      <th>transactionType</th>\n",
       "      <th>echoBuffer</th>\n",
       "      <th>currentBalance</th>\n",
       "      <th>merchantCity</th>\n",
       "      <th>merchantState</th>\n",
       "      <th>merchantZip</th>\n",
       "      <th>cardPresent</th>\n",
       "      <th>posOnPremises</th>\n",
       "      <th>recurringAuthInd</th>\n",
       "      <th>expirationDateKeyInMatch</th>\n",
       "      <th>isFraud</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>737265056</td>\n",
       "      <td>737265056</td>\n",
       "      <td>5000</td>\n",
       "      <td>5000.0</td>\n",
       "      <td>2016-08-13T14:27:32</td>\n",
       "      <td>98.55</td>\n",
       "      <td>Uber</td>\n",
       "      <td>US</td>\n",
       "      <td>US</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>rideshare</td>\n",
       "      <td>06/2023</td>\n",
       "      <td>2015-03-14</td>\n",
       "      <td>2015-03-14</td>\n",
       "      <td>414</td>\n",
       "      <td>414</td>\n",
       "      <td>1803</td>\n",
       "      <td>PURCHASE</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>False</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>737265056</td>\n",
       "      <td>737265056</td>\n",
       "      <td>5000</td>\n",
       "      <td>5000.0</td>\n",
       "      <td>2016-10-11T05:05:54</td>\n",
       "      <td>74.51</td>\n",
       "      <td>AMC #191138</td>\n",
       "      <td>US</td>\n",
       "      <td>US</td>\n",
       "      <td>9.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>entertainment</td>\n",
       "      <td>02/2024</td>\n",
       "      <td>2015-03-14</td>\n",
       "      <td>2015-03-14</td>\n",
       "      <td>486</td>\n",
       "      <td>486</td>\n",
       "      <td>767</td>\n",
       "      <td>PURCHASE</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>True</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>737265056</td>\n",
       "      <td>737265056</td>\n",
       "      <td>5000</td>\n",
       "      <td>5000.0</td>\n",
       "      <td>2016-11-08T09:18:39</td>\n",
       "      <td>7.47</td>\n",
       "      <td>Play Store</td>\n",
       "      <td>US</td>\n",
       "      <td>US</td>\n",
       "      <td>9.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>mobileapps</td>\n",
       "      <td>08/2025</td>\n",
       "      <td>2015-03-14</td>\n",
       "      <td>2015-03-14</td>\n",
       "      <td>486</td>\n",
       "      <td>486</td>\n",
       "      <td>767</td>\n",
       "      <td>PURCHASE</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>False</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   accountNumber  customerId  creditLimit  availableMoney  \\\n",
       "0      737265056   737265056         5000          5000.0   \n",
       "1      737265056   737265056         5000          5000.0   \n",
       "2      737265056   737265056         5000          5000.0   \n",
       "\n",
       "   transactionDateTime  transactionAmount merchantName acqCountry  \\\n",
       "0  2016-08-13T14:27:32              98.55         Uber         US   \n",
       "1  2016-10-11T05:05:54              74.51  AMC #191138         US   \n",
       "2  2016-11-08T09:18:39               7.47   Play Store         US   \n",
       "\n",
       "  merchantCountryCode  posEntryMode  posConditionCode merchantCategoryCode  \\\n",
       "0                  US           2.0               1.0            rideshare   \n",
       "1                  US           9.0               1.0        entertainment   \n",
       "2                  US           9.0               1.0           mobileapps   \n",
       "\n",
       "  currentExpDate accountOpenDate dateOfLastAddressChange  cardCVV  enteredCVV  \\\n",
       "0        06/2023      2015-03-14              2015-03-14      414         414   \n",
       "1        02/2024      2015-03-14              2015-03-14      486         486   \n",
       "2        08/2025      2015-03-14              2015-03-14      486         486   \n",
       "\n",
       "   cardLast4Digits transactionType  echoBuffer  currentBalance  merchantCity  \\\n",
       "0             1803        PURCHASE         NaN             0.0           NaN   \n",
       "1              767        PURCHASE         NaN             0.0           NaN   \n",
       "2              767        PURCHASE         NaN             0.0           NaN   \n",
       "\n",
       "   merchantState  merchantZip  cardPresent  posOnPremises  recurringAuthInd  \\\n",
       "0            NaN          NaN        False            NaN               NaN   \n",
       "1            NaN          NaN         True            NaN               NaN   \n",
       "2            NaN          NaN        False            NaN               NaN   \n",
       "\n",
       "   expirationDateKeyInMatch  isFraud  \n",
       "0                     False    False  \n",
       "1                     False    False  \n",
       "2                     False    False  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.set_option('display.max_columns', None) # pandas setting\n",
    "transactions=import_data(file_path)\n",
    "transactions.head(3)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Dimension of the dataset is 786363 x 29.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Dataset Dimensions: 786363 x 29\n"
     ]
    }
   ],
   "source": [
    "print(\"Dataset Dimensions: {} x {}\".format(transactions.shape[0], transactions.shape[1])) # size of the dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>accountNumber</th>\n",
       "      <th>customerId</th>\n",
       "      <th>creditLimit</th>\n",
       "      <th>availableMoney</th>\n",
       "      <th>transactionAmount</th>\n",
       "      <th>posEntryMode</th>\n",
       "      <th>posConditionCode</th>\n",
       "      <th>cardCVV</th>\n",
       "      <th>enteredCVV</th>\n",
       "      <th>cardLast4Digits</th>\n",
       "      <th>echoBuffer</th>\n",
       "      <th>currentBalance</th>\n",
       "      <th>merchantCity</th>\n",
       "      <th>merchantState</th>\n",
       "      <th>merchantZip</th>\n",
       "      <th>posOnPremises</th>\n",
       "      <th>recurringAuthInd</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>7.863630e+05</td>\n",
       "      <td>7.863630e+05</td>\n",
       "      <td>786363.000000</td>\n",
       "      <td>786363.000000</td>\n",
       "      <td>786363.000000</td>\n",
       "      <td>782309.000000</td>\n",
       "      <td>785954.000000</td>\n",
       "      <td>786363.000000</td>\n",
       "      <td>786363.000000</td>\n",
       "      <td>786363.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>786363.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>5.372326e+08</td>\n",
       "      <td>5.372326e+08</td>\n",
       "      <td>10759.464459</td>\n",
       "      <td>6250.725369</td>\n",
       "      <td>136.985791</td>\n",
       "      <td>9.049943</td>\n",
       "      <td>3.271980</td>\n",
       "      <td>544.467338</td>\n",
       "      <td>544.183857</td>\n",
       "      <td>4757.417799</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4508.739089</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>2.554211e+08</td>\n",
       "      <td>2.554211e+08</td>\n",
       "      <td>11636.174890</td>\n",
       "      <td>8880.783989</td>\n",
       "      <td>147.725569</td>\n",
       "      <td>16.774130</td>\n",
       "      <td>9.809023</td>\n",
       "      <td>261.524220</td>\n",
       "      <td>261.551254</td>\n",
       "      <td>2996.583810</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6457.442068</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>1.000881e+08</td>\n",
       "      <td>1.000881e+08</td>\n",
       "      <td>250.000000</td>\n",
       "      <td>-1005.630000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>100.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>3.301333e+08</td>\n",
       "      <td>3.301333e+08</td>\n",
       "      <td>5000.000000</td>\n",
       "      <td>1077.420000</td>\n",
       "      <td>33.650000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>310.000000</td>\n",
       "      <td>310.000000</td>\n",
       "      <td>2178.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>689.910000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>5.074561e+08</td>\n",
       "      <td>5.074561e+08</td>\n",
       "      <td>7500.000000</td>\n",
       "      <td>3184.860000</td>\n",
       "      <td>87.900000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>535.000000</td>\n",
       "      <td>535.000000</td>\n",
       "      <td>4733.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2451.760000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>7.676200e+08</td>\n",
       "      <td>7.676200e+08</td>\n",
       "      <td>15000.000000</td>\n",
       "      <td>7500.000000</td>\n",
       "      <td>191.480000</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>785.000000</td>\n",
       "      <td>785.000000</td>\n",
       "      <td>7338.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5291.095000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>9.993896e+08</td>\n",
       "      <td>9.993896e+08</td>\n",
       "      <td>50000.000000</td>\n",
       "      <td>50000.000000</td>\n",
       "      <td>2011.540000</td>\n",
       "      <td>90.000000</td>\n",
       "      <td>99.000000</td>\n",
       "      <td>998.000000</td>\n",
       "      <td>998.000000</td>\n",
       "      <td>9998.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>47498.810000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       accountNumber    customerId    creditLimit  availableMoney  \\\n",
       "count   7.863630e+05  7.863630e+05  786363.000000   786363.000000   \n",
       "mean    5.372326e+08  5.372326e+08   10759.464459     6250.725369   \n",
       "std     2.554211e+08  2.554211e+08   11636.174890     8880.783989   \n",
       "min     1.000881e+08  1.000881e+08     250.000000    -1005.630000   \n",
       "25%     3.301333e+08  3.301333e+08    5000.000000     1077.420000   \n",
       "50%     5.074561e+08  5.074561e+08    7500.000000     3184.860000   \n",
       "75%     7.676200e+08  7.676200e+08   15000.000000     7500.000000   \n",
       "max     9.993896e+08  9.993896e+08   50000.000000    50000.000000   \n",
       "\n",
       "       transactionAmount   posEntryMode  posConditionCode        cardCVV  \\\n",
       "count      786363.000000  782309.000000     785954.000000  786363.000000   \n",
       "mean          136.985791       9.049943          3.271980     544.467338   \n",
       "std           147.725569      16.774130          9.809023     261.524220   \n",
       "min             0.000000       2.000000          1.000000     100.000000   \n",
       "25%            33.650000       2.000000          1.000000     310.000000   \n",
       "50%            87.900000       5.000000          1.000000     535.000000   \n",
       "75%           191.480000       9.000000          1.000000     785.000000   \n",
       "max          2011.540000      90.000000         99.000000     998.000000   \n",
       "\n",
       "          enteredCVV  cardLast4Digits  echoBuffer  currentBalance  \\\n",
       "count  786363.000000    786363.000000         0.0   786363.000000   \n",
       "mean      544.183857      4757.417799         NaN     4508.739089   \n",
       "std       261.551254      2996.583810         NaN     6457.442068   \n",
       "min         0.000000         0.000000         NaN        0.000000   \n",
       "25%       310.000000      2178.000000         NaN      689.910000   \n",
       "50%       535.000000      4733.000000         NaN     2451.760000   \n",
       "75%       785.000000      7338.000000         NaN     5291.095000   \n",
       "max       998.000000      9998.000000         NaN    47498.810000   \n",
       "\n",
       "       merchantCity  merchantState  merchantZip  posOnPremises  \\\n",
       "count           0.0            0.0          0.0            0.0   \n",
       "mean            NaN            NaN          NaN            NaN   \n",
       "std             NaN            NaN          NaN            NaN   \n",
       "min             NaN            NaN          NaN            NaN   \n",
       "25%             NaN            NaN          NaN            NaN   \n",
       "50%             NaN            NaN          NaN            NaN   \n",
       "75%             NaN            NaN          NaN            NaN   \n",
       "max             NaN            NaN          NaN            NaN   \n",
       "\n",
       "       recurringAuthInd  \n",
       "count               0.0  \n",
       "mean                NaN  \n",
       "std                 NaN  \n",
       "min                 NaN  \n",
       "25%                 NaN  \n",
       "50%                 NaN  \n",
       "75%                 NaN  \n",
       "max                 NaN  "
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "transactions.describe() # Descriptive statistics of the features in the dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['accountNumber', 'customerId', 'creditLimit', 'availableMoney',\n",
       "       'transactionDateTime', 'transactionAmount', 'merchantName',\n",
       "       'acqCountry', 'merchantCountryCode', 'posEntryMode',\n",
       "       'posConditionCode', 'merchantCategoryCode', 'currentExpDate',\n",
       "       'accountOpenDate', 'dateOfLastAddressChange', 'cardCVV',\n",
       "       'enteredCVV', 'cardLast4Digits', 'transactionType', 'echoBuffer',\n",
       "       'currentBalance', 'merchantCity', 'merchantState', 'merchantZip',\n",
       "       'cardPresent', 'posOnPremises', 'recurringAuthInd',\n",
       "       'expirationDateKeyInMatch', 'isFraud'], dtype=object)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.array(transactions.columns) # Columns names as list"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**There are 29 features in the dataset (6 of them are Null) and their data types are as follows:**\n",
    "\n",
    "   **Timestamp: 4 columns,**\n",
    "   **Boolean : 3 columns (False vs True),**\n",
    "   **Categorical: 7 columns,**\n",
    "   **Numerical: 9 columns**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 786363 entries, 0 to 786362\n",
      "Data columns (total 29 columns):\n",
      " #   Column                    Non-Null Count   Dtype  \n",
      "---  ------                    --------------   -----  \n",
      " 0   accountNumber             786363 non-null  int64  \n",
      " 1   customerId                786363 non-null  int64  \n",
      " 2   creditLimit               786363 non-null  int64  \n",
      " 3   availableMoney            786363 non-null  float64\n",
      " 4   transactionDateTime       786363 non-null  object \n",
      " 5   transactionAmount         786363 non-null  float64\n",
      " 6   merchantName              786363 non-null  object \n",
      " 7   acqCountry                781801 non-null  object \n",
      " 8   merchantCountryCode       785639 non-null  object \n",
      " 9   posEntryMode              782309 non-null  float64\n",
      " 10  posConditionCode          785954 non-null  float64\n",
      " 11  merchantCategoryCode      786363 non-null  object \n",
      " 12  currentExpDate            786363 non-null  object \n",
      " 13  accountOpenDate           786363 non-null  object \n",
      " 14  dateOfLastAddressChange   786363 non-null  object \n",
      " 15  cardCVV                   786363 non-null  int64  \n",
      " 16  enteredCVV                786363 non-null  int64  \n",
      " 17  cardLast4Digits           786363 non-null  int64  \n",
      " 18  transactionType           785665 non-null  object \n",
      " 19  echoBuffer                0 non-null       float64\n",
      " 20  currentBalance            786363 non-null  float64\n",
      " 21  merchantCity              0 non-null       float64\n",
      " 22  merchantState             0 non-null       float64\n",
      " 23  merchantZip               0 non-null       float64\n",
      " 24  cardPresent               786363 non-null  bool   \n",
      " 25  posOnPremises             0 non-null       float64\n",
      " 26  recurringAuthInd          0 non-null       float64\n",
      " 27  expirationDateKeyInMatch  786363 non-null  bool   \n",
      " 28  isFraud                   786363 non-null  bool   \n",
      "dtypes: bool(3), float64(11), int64(6), object(9)\n",
      "memory usage: 158.2+ MB\n"
     ]
    }
   ],
   "source": [
    "transactions.info() # Non-Null count and object type"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**There are 11 features including at least 1 Null value. In addition, 5 out of these 11 features does not include any value in the dataset.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "accountNumber                    0\n",
       "customerId                       0\n",
       "creditLimit                      0\n",
       "availableMoney                   0\n",
       "transactionDateTime              0\n",
       "transactionAmount                0\n",
       "merchantName                     0\n",
       "acqCountry                    4562\n",
       "merchantCountryCode            724\n",
       "posEntryMode                  4054\n",
       "posConditionCode               409\n",
       "merchantCategoryCode             0\n",
       "currentExpDate                   0\n",
       "accountOpenDate                  0\n",
       "dateOfLastAddressChange          0\n",
       "cardCVV                          0\n",
       "enteredCVV                       0\n",
       "cardLast4Digits                  0\n",
       "transactionType                698\n",
       "echoBuffer                  786363\n",
       "currentBalance                   0\n",
       "merchantCity                786363\n",
       "merchantState               786363\n",
       "merchantZip                 786363\n",
       "cardPresent                      0\n",
       "posOnPremises               786363\n",
       "recurringAuthInd            786363\n",
       "expirationDateKeyInMatch         0\n",
       "isFraud                          0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "transactions.isnull().sum() # Check each column for Null values"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**I removed those columns including all Null values in the dataset to reduce dimensionality.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "transactions.drop(['echoBuffer','merchantCity','merchantState','merchantZip','posOnPremises','recurringAuthInd'],\n",
    "                  axis=1,inplace=True) # drop several columns"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**In addition to that, I checked the relationship between merchantCategoryCode and cardPresent. Online transactions accounted for 55% of all transactions.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "False    433495\n",
       "True     352868\n",
       "Name: cardPresent, dtype: int64"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "transactions.cardPresent.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.5513"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "round(transactions.cardPresent.value_counts()[0]/len(transactions),4)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**There are 5000 unique customers in the dataset.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "5000"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "transactions.accountNumber.nunique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "5000"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# There are 5000 unique customers in the dataset\n",
    "transactions['customerId'].nunique()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**I checked whether account number and customerID have one to one relation or not and observed that each customer has 1 unique account number in the database.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "account=transactions.accountNumber.unique()\n",
    "customer=transactions['customerId'].unique()\n",
    "comparison = account== customer\n",
    "equal_arrays = comparison.all()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "True\n"
     ]
    }
   ],
   "source": [
    "print(equal_arrays)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Capital One provides 10 different Credit Limits to its customers starting from 250 to $20000.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 5000,  2500, 50000, 15000, 10000,   250,   500,  1000,  7500,\n",
       "       20000])"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "transactions.creditLimit.unique() # Credit Limit types"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Capital One offers 5000, 15000 or 7500 as credit limit to its customers mostly.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "z=pd.DataFrame(transactions.groupby('creditLimit').count()['isFraud']).reset_index(drop=False)\n",
    "z.columns=['creditLimit','Freq'] # rename columns\n",
    "plt.figure(figsize=(14,6)) # adjust figure size\n",
    "sns.set_theme() # set theme\n",
    "ax = sns.barplot(x=\"creditLimit\", y=\"Freq\", data=z) # draw barplot"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**There are 12417 fraudulent cases among all population.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "False    773946\n",
       "True      12417\n",
       "Name: isFraud, dtype: int64"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "transactions['isFraud'].value_counts() # distribution of fraud vs non-fraud cases"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Ratio of fraudulent cases to all population is 1.58 % so the dataset is highly imbalanced.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.0158"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Dataset is highly imbalanced\n",
    "round(transactions['isFraud'].value_counts() [1] / len(transactions),4) # ratio: 1.579 %"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**There are 19 merchant categories such as rideshare, auto,hotels and fastfood.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['rideshare', 'entertainment', 'mobileapps', 'fastfood',\n",
       "       'food_delivery', 'auto', 'online_retail', 'gym', 'health',\n",
       "       'personal care', 'food', 'fuel', 'online_subscriptions',\n",
       "       'online_gifts', 'hotels', 'airline', 'furniture', 'subscriptions',\n",
       "       'cable/phone'], dtype=object)"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "transactions['merchantCategoryCode'].unique() # list of merchant categories"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "19"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "transactions['merchantCategoryCode'].nunique() # size: 19 categories"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Top 3 categories are online, fastfood and entertainment in terms of the number of transactions.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1728x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "t=pd.DataFrame(transactions['merchantCategoryCode'].value_counts()).reset_index(drop=False)\n",
    "t.columns=['MerchantCategory','Freq'] # rename columns\n",
    "plt.figure(figsize=(24,6)) # adjust figure size\n",
    "sns.set_theme() # set theme\n",
    "ax = sns.barplot(x=\"MerchantCategory\", y=\"Freq\", data=t) # draw barplot"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**The dataset includes transactions in 4 countries such as US, Canada, Mexico and Puerto Rico.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['US', None, 'CAN', 'MEX', 'PR'], dtype=object)"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "transactions.acqCountry.unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['US', 'CAN', None, 'PR', 'MEX'], dtype=object)"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "transactions.merchantCountryCode.unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "US     774709\n",
       "MEX      3130\n",
       "CAN      2424\n",
       "PR       1538\n",
       "Name: acqCountry, dtype: int64"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "transactions.acqCountry.value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**There are 3 different transaction types such as Purchase, Address Verification and Reversal. Also, it includes some Null values in that feature.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['PURCHASE', 'ADDRESS_VERIFICATION', 'REVERSAL', None], dtype=object)"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "transactions['transactionType'].unique() # list of transaction type categories"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "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>index</th>\n",
       "      <th>transactionType</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>PURCHASE</td>\n",
       "      <td>745193</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>REVERSAL</td>\n",
       "      <td>20303</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>ADDRESS_VERIFICATION</td>\n",
       "      <td>20169</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  index  transactionType\n",
       "0              PURCHASE           745193\n",
       "1              REVERSAL            20303\n",
       "2  ADDRESS_VERIFICATION            20169"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "e=pd.DataFrame(transactions.transactionType.value_counts()).reset_index(drop=False)\n",
    "e"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "e=pd.DataFrame(transactions.transactionType.value_counts()).reset_index(drop=False)\n",
    "e.columns=['Category','Freq'] # rename columns\n",
    "plt.figure(figsize=(14,6)) # adjust figure size\n",
    "sns.set_theme() # set theme\n",
    "ax = sns.barplot(x=\"Freq\", y=\"Category\", data=e) # draw barplot"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**I created 'Date' column by splitting original 'TransactionDate' column.The dataset obtains transactions in year 2016.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "transactions['Date']=transactions['transactionDateTime'].apply(lambda x: x.split('T')[0]) # create 'Date' column"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'2016-01-01T00:01:02'"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Dataset includes all transactions during year 2016\n",
    "transactions.transactionDateTime.min() # Start point: 2016-01-01 T 00:01:02"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'2016-12-30T23:59:45'"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "transactions.transactionDateTime.max() # End point:   2016-12-30 T 23:59:45"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**I checked the key statistics of  transaction amounts. Maximum transaction amount was $2011. Also, average of transaction amount is 136.98 standard deviation of transaction amount was 147.72.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    786363.000000\n",
       "mean        136.985791\n",
       "std         147.725569\n",
       "min           0.000000\n",
       "25%          33.650000\n",
       "50%          87.900000\n",
       "75%         191.480000\n",
       "max        2011.540000\n",
       "Name: transactionAmount, dtype: float64"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "transactions.transactionAmount.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**I plotted the transactions amounts by applying histogram. This shows that there is right skewed (positive skewness) distribution. Avg. of transaction amounts (136.98) is much higher than median of transaction amounts (87.90)**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/anaconda3/lib/python3.7/site-packages/seaborn/distributions.py:2557: FutureWarning: `distplot` is a deprecated function and will be removed in a future version. Please adapt your code to use either `displot` (a figure-level function with similar flexibility) or `histplot` (an axes-level function for histograms).\n",
      "  warnings.warn(msg, FutureWarning)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1a241ce090>"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAt8AAAGsCAYAAAAFVMs2AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAgAElEQVR4nOzdfXhUd53//9eZm0wymQm5YSaBQCm2ttQipW3aItqoq0sQwbJsvXaFvei1+1v8bfemC6usfkuXFldFXZTaden3Z73ZX7UoqP2G5ScGat21tlALaAtt6Q0tUiAw5I6SZJLJzDnn90fI0JSbBpgz58zk+bgurjLnnJl5T+Z4+cqH9+fzMWzbtgUAAADAcT63CwAAAABGC8I3AAAAkCeEbwAAACBPCN8AAABAnhC+AQAAgDwhfAMAAAB5QvgGAAAA8iTgdgH51NXVK8vy5rLmNTURdXT0uF0GigD3EnKFewm5wr2EXCmEe8nnM1RVVX7O86MqfFuW7dnwLcnTtaGwcC8hV7iXkCvcS8iVQr+XaDsBAAAA8oTwDQAAAOQJ4RsAAADIE8I3AAAAkCeEbwAAACBPCN8AAABAnhC+AQAAgDwhfAMAAAB5QvgGAAAA8oTwDQAAAOQJ4RsAAADIE8I3AAAAkCeEbwAAACBPCN8AAABAnhC+AQAAgDwJuF0ABnUnB9Sbyozo2lAwoAC/NgEAABQcwrdH9PVntHNfYkTX3nRNrQIhvjoAAIBCw/gpAAAAkCeODp9u3rxZDz74oDKZjO644w4tWrRo2Pl9+/ZpxYoV6u3tVUNDg1atWqVAIKDW1lYtX75cHR0dmjx5stasWaPy8nItWLBApmlKkvr7+3Xo0CE98cQTGjt2rJMfAwAAAMgJx0a+E4mE1q5dq/Xr16u5uVkbNmzQ/v37h12zfPlyrVy5Ulu3bpVt29q4caMkadWqVVq4cKFaWlo0depUrVu3TpL06KOPatOmTdq0aZOuu+463XXXXQRvAAAAFAzHwvf27ds1Y8YMVVZWKhwOq6mpSS0tLdnzR44cUX9/v6ZPny5JWrBggVpaWpROp7Vz5041NTUNO/5WO3bs0EsvvaQlS5Y4VT4AAACQc46F7+PHjysWi2Ufx+NxJRKJc56PxWJKJBLq6upSJBJRIBAYdvytHnjgAS1btkx+v9+p8gEAAICcc6zn27IsGYaRfWzb9rDH5zr/9uskDXv86quvqqurSx/+8IcvuKaamsgFPydfjncmFY2UjujacDikWHXY4YpQyGKxqNsloEhwLyFXuJeQK4V+LzkWvuvq6rRr167s47a2NsXj8WHn29raso/b29sVj8dVXV2t7u5umaYpv99/xvN++ctfas6cORdVU0dHjyzLvqjnOs7vV3dP/4guTSZTajs18RR4u1gsqra2brfLQBHgXkKucC8hVwrhXvL5jPMO+DrWdjJz5kzt2LFDnZ2d6uvr07Zt29TY2Jg9X19fr1AopN27d0uSNm3apMbGRgWDQTU0NGjLli2SpObm5mHPe/bZZ9XQ0OBU2QAAAIBjHAvftbW1WrZsmRYvXqz58+dr7ty5mjZtmpYsWaK9e/dKktasWaPVq1dr9uzZSiaTWrx4sSTp3nvv1caNGzVnzhzt2rVLS5cuzb7uoUOHVFtb61TZAAAAgGMM27Y92oeRe15uO7H9fv169xsjuvama2pVzg6XOIdC+Cc5FAbuJeQK9xJypRDuJdfaTgAAAAAMR/gGAAAA8oTwDQAAAOQJ4RsAAADIE8I3AAAAkCeEbwAAACBPCN8AAABAnhC+AQAAgDwhfAMAAAB5QvgGAAAA8oTwDQAAAOQJ4RsAAADIE8K3xxzvSurFP3S6XQYAAAAcQPj2mP1HTup3L7fJtm23SwEAAECOEb49JpOxZNlSXyrjdikAAADIMcK3x2RMS5LU05d2uRIAAADkGuHbY9KEbwAAgKJF+PaYTGaw17unj7YTAACAYkP49hjaTgAAAIoX4dtjaDsBAAAoXoRvj8lkToXvJOEbAACg2BC+PcS27ezId29/WhZrfQMAABQVwreHWLYt25bKSwOybamvn0mXAAAAxYTw7SHpUyudVEZDkuj7BgAAKDaEbw8ZWumkMkL4BgAAKEaEbw85Hb5LJBG+AQAAig3h20OGVjoJlfhVFgoQvgEAAIoM4dtDhlY6Cfp9ipQRvgEAAIoN4dtDMubghMuA36dIWVC9bDEPAABQVAjfHpI+1XYSDJwK3/1pWRZrfQMAABQLwreHDE24HBr5tm0pmWL0GwAAoFgQvj1kaMJlwG+ovCwoiW3mAQAAignh20PSbxv5llhuEAAAoJgQvj0kY1ry+wz5fG8Z+SZ8AwAAFA3Ct4ekM7YC/sGvxO8zFC4NqJfwDQAAUDQI3x6SMS0FA6e/kkhZkJFvAACAIkL49pCMaSngN7KPCd8AAADFhfDtIemMlW07kQbDd7I/w1rfAAAARYLw7SFvbzspLwvKltTbz+g3AABAMSB8e8iZI98BSax4AgAAUCwI3x6SMe0zJlxKUk8fu1wCAAAUA8K3h7x9wmV5aVCGxHKDAAAARcLR8L1582bNmTNHs2bN0iOPPHLG+X379mnBggVqamrSihUrlMkMjvC2trZq0aJFmj17tu6880719vZKknp6evSZz3xG8+fP1/z58/XCCy84WX7eDYbv01+J79Ra37SdAAAAFAfHwncikdDatWu1fv16NTc3a8OGDdq/f/+wa5YvX66VK1dq69atsm1bGzdulCStWrVKCxcuVEtLi6ZOnap169ZJklavXq1x48apublZ//RP/6T77rvPqfLzzrLsM9pOJJYbBAAAKCaOhe/t27drxowZqqysVDgcVlNTk1paWrLnjxw5ov7+fk2fPl2StGDBArW0tCidTmvnzp1qamoadty2bW3btk2f/vSnJUmNjY368pe/7FT5eZdKm5I0bORbInwDAAAUk4BTL3z8+HHFYrHs43g8rj179pzzfCwWUyKRUFdXlyKRiAKBwLDjHR0dKikp0fr16/Xf//3fCoVCuvvuuy+oppqayCV+Kue8crBLkhQpDykaKc0erx5TptdaTyocLpHfNxjMw+GQYtVhV+pEYYjFom6XgCLBvYRc4V5CrhT6veRY+LYsS4ZxevKgbdvDHp/r/NuvkyTDMGSaptrb2xWNRrVhwwY99dRT+ru/+zs9/vjjI66po6PHsxvWpNKD/e5mxlR3T3/2eElg8GdxrK1HFeUlkqRkMqU208x/kSgIsVhUbW3dbpeBIsC9hFzhXkKuFMK95PMZ5x3wdaztpK6uTm1tbdnHbW1tisfj5zzf3t6ueDyu6upqdXd3yzwVLoeeV1VVpUAgoLlz50qS3v/+9yuZTKqjo8Opj5BX/QNDbSfDf/Eozy43SOsJAABAoXMsfM+cOVM7duxQZ2en+vr6tG3bNjU2NmbP19fXKxQKaffu3ZKkTZs2qbGxUcFgUA0NDdqyZYskqbm5WY2NjSopKdHMmTP185//XJL07LPPqqysTFVVVU59hLxKDZy751tiuUEAAIBi4Fj4rq2t1bJly7R48WLNnz9fc+fO1bRp07RkyRLt3btXkrRmzRqtXr1as2fPVjKZ1OLFiyVJ9957rzZu3Kg5c+Zo165dWrp0qSTpS1/6kp544gnNnTtX9913n9auXSufrziWKh8K329f7SQcCsgwGPkGAAAoBoZt295sgnaAl3u+n3mlXf/70T2a9/7LVRUNDTv36K9fV6yyVLdeN16SdNM1tSoPOdaujwJXCP1wKAzcS8gV7iXkSiHcS671fOPCpAYGJ1wG/Wd+JSw3CAAAUBwI3x6R7fkOGGecKy8LqKcvk++SAAAAkGOEb48YWu3kbCPf0bKg+lIZmZaV77IAAACQQ4Rvj0gNmDKMwT6htyvPrnjC6DcAAEAhI3x7RH86o4Dfd8YGQ9Lp5Qbp+wYAAChshG+PSA2YZ205kQjfAAAAxYLw7RGpAVOBwNm/jrLSgHyG1JMkfAMAABQywrdHpNKmgv4zW04kyWcYKme5QQAAgIJH+PaI1IB5xtbyb0X4BgAAKHyEb4/oH8ics+1EGuz77u0nfAMAABQywrdHDLadnPvrGFzr21Q6w1rfAAAAhYrw7RHnm3ApSWMiJZKkEz2pfJUEAACAHCN8e0T/eZYalKTqaKkkqesk4RsAAKBQEb49wLbtUxMuz77aiSSVlwUUDPjU2U34BgAAKFSEbw/ImJYs2z7vaieGYag6GlJXd38eKwMAAEAuEb49oH/AlKTz9nxLUlU0pK7ulCzbzkdZAAAAyDHCtwekToXv8/V8S1JVRakypq32E4x+AwAAFCLCtwf0p0c28l0dDUmSjrT1OF4TAAAAco/w7QGnR77PPeFSkiojJTIM6Uh7bz7KAgAAQI4Rvj0gO/L9Dm0nfr9PY8pLdOQ4I98AAACFiPDtAakRTriUBiddMvINAABQmAjfHjDSCZfS6RVPevrSTpcFAACAHCN8e8BI204kqerUTpeHaD0BAAAoOIRvD8iOfI+g7aS6YnDFE8I3AABA4SF8e0Dq1Mi3/x1WO5GkslBA0XBQhxLdTpcFAACAHCN8e0BqwFRJ0Cef8c7hW5LqYxFGvgEAAAoQ4dsD+tOmSksCI75+QqxcrR29ypiWg1UBAAAg1wjfHpAayKgk6B/x9ePHlitj2kp0Jh2sCgAAALlG+PaAjGkrMIJ+7yF11WFJ0tEOwjcAAEAhIXx7gGnZ8vtG/lXET4Xv1g422wEAACgkhG8PME1Lft/IR75DQb9qKkrVyk6XAAAABYXw7QGmZY9omcG3Gj+2nLYTAACAAkP49oALbTuRpHE1YR3rTMqybIeqAgAAQK4Rvj3ANK2LGvlOZyy1v9nnUFUAAADINcK3BwyOfF9g+K4plyS10noCAABQMAjfHpC5iPA9buzQcoNMugQAACgUhG8PMM0L7/kuLw2qoryEFU8AAAAKCOHbAyz7wlc7kaTxNWFWPAEAACgghG8PuNB1voeMG1uuox29sm1WPAEAACgEhG8PGFzn+8K/ivE15epLmTrRM+BAVQAAAMg1wrcHXMxqJ9Jg24nENvMAAACFgvDtAZfSdiJJR5l0CQAAUBAcDd+bN2/WnDlzNGvWLD3yyCNnnN+3b58WLFigpqYmrVixQplMRpLU2tqqRYsWafbs2brzzjvV2zsYLp955hndcsstuu2223Tbbbfpf/2v/+Vk+XlzsW0nY8pLVBYKsNY3AABAgXAsfCcSCa1du1br169Xc3OzNmzYoP379w+7Zvny5Vq5cqW2bt0q27a1ceNGSdKqVau0cOFCtbS0aOrUqVq3bp0k6fnnn9df/dVfadOmTdq0aZNWr17tVPl5dTHrfEuSYRgaPzbMyDcAAECBcCx8b9++XTNmzFBlZaXC4bCamprU0tKSPX/kyBH19/dr+vTpkqQFCxaopaVF6XRaO3fuVFNT07DjkrR37149+eSTmjdvnv7mb/5GR48edar8vDLNixv5lqRxNeVstAMAAFAgAk698PHjxxWLxbKP4/G49uzZc87zsVhMiURCXV1dikQiCgQCw45LUjQa1cc+9jHNmjVLP/rRj7Rs2TL9+Mc/HnFNNTWRS/1YjrAsS37DUDRSOqLrw+GQYtWDky3ffVm1ntxzVKFwSBXlJU6WiQISi0XdLgFFgnsJucK9hFwp9HvJsfBtWZYM43QrhW3bwx6f6/zbr5OUffyFL3whe+xTn/qUvv71r6u7u1vR6Mi+hI6OHlmWt9bEtmxbli35/Ya6e/pH9JxkMqU205QkRUKDI+b79rfpXeMrHKsThSMWi6qtrdvtMlAEuJeQK9xLyJVCuJd8PuO8A76OtZ3U1dWpra0t+7itrU3xePyc59vb2xWPx1VdXa3u7m6Zp8Ll0PMsy9KDDz6YPT7E7/c79RHyYuiXgYvp+ZakmorB0fLOkyML7gAAAHCPY+F75syZ2rFjhzo7O9XX16dt27apsbExe76+vl6hUEi7d++WJG3atEmNjY0KBoNqaGjQli1bJEnNzc1qbGyUz+fTY489pq1bt2aPX3fddQqHw059hLwwzVPh+yJ7vmvGDIbvDsI3AACA5zkWvmtra7Vs2TItXrxY8+fP19y5czVt2jQtWbJEe/fulSStWbNGq1ev1uzZs5VMJrV48WJJ0r333quNGzdqzpw52rVrl5YuXSpJ+upXv6qHH35YH//4x/Wzn/1MX/ziF50qP29My5J08SPf4VBAoRI/4RsAAKAAGLZte6sJ2kFe7Pk+mRzQ0gee1MJZVyswwl+FbrqmVuWh0+3693znt6qtKtM//Ok0h6pEISmEfjgUBu4l5Ar3EnKlEO4l13q+MTKX2nYiDfZ9d55M5aokAAAAOITw7bJLbTuRpJqKEG0nAAAABYDw7TJzaLUT/8WH7+qKUvX0pZVKm+98MQAAAFxD+HZZtu3Ed2ltJxLLDQIAAHgd4dtl5iWu8y2x3CAAAEChIHy7LBc939UVIUli0iUAAIDHEb5dloue78pISIYhdbzJyDcAAICXEb5dloue74Dfp8pIiJ5vAAAAjyN8uywXI9/SYN83Pd8AAADeRvh22eme70v7KmoqCN8AAABeR/h22em2k0sb+a6uCKnzZEqWbeeiLAAAADiA8O2ynLWdVJTKtGyd7B3IRVkAAABwQMDtAka70+t8j/z3IMNnqDeVGXasvCwoSTrc3qtg0D/sXCgYUIBfswAAAFxH+HaZaQ72fAcuYOQ7lTb13Cttw451dQ+u8b1rX0LtJ/qGnbvpmloFQnzVAAAAbmM81GVDI9++S+z5Li8dDNe9/Zl3uBIAAABuIXy7LBfby0tSSdCvYMCn3r50LsoCAACAAwjfLruYnu9zKS8NMPINAADgYYRvlw31fF/qaifS4KTLHka+AQAAPIvw7bJctZ1IUnlpUL39hG8AAACvIny77PQ63zloOykLaCBtKZ2xLvm1AAAAkHuEb5dl205yMPIdKR1c65vRbwAAAG8ifLssl20n4bJTyw32MekSAADAiwjfLjMtW36fIcPITc+3JCUZ+QYAAPAkwrfLTNPOyai3JIVDbLQDAADgZYRvl2Us65J3txzi8xkqCwXo+QYAAPAowrfLhtpOcoWNdgAAALyL8O0yy7JzsszgkPKyoJJstAMAAOBJhG+X5bLnWzo98m3bds5eEwAAALlB+HaZaVk5Dd/h0oBMy1YqzUY7AAAAXkP4dpmZ67YTlhsEAADwLMK3y0zTViDHbScSyw0CAAB4EeHbZble7STMFvMAAACeRfh2Wcay5PfnLnyXhfzyGVKSLeYBAAA8h/DtssHVTnL3NRiGoXBpkJFvAAAADyJ8u8y07JztcDkkXBpQkp5vAAAAzyF8u8zKcc+3xC6XAAAAXkX4dpmZ455vaXDSZbI/zUY7AAAAHkP4dtngUoO5/RrKSwOybKkvZeb0dQEAAHBpCN8uy/VSg5JUXsZGOwAAAF5E+HZZxrJz3nbCRjsAAADeNKLw/Q//8A/avn2707WMSqZp5Xzkm412AAAAvGlE4fuP//iPtW7dOjU1Nem73/2uTpw4MaIX37x5s+bMmaNZs2bpkUceOeP8vn37tGDBAjU1NWnFihXKZAZHaltbW7Vo0SLNnj1bd955p3p7e4c979ixY7r55pt1+PDhEdXhZYNtJ7n9B4hQ0Ce/z2C5QQAAAI8ZUer7xCc+oR/+8Idat26dOjo6dPvtt2v58uXas2fPOZ+TSCS0du1arV+/Xs3NzdqwYYP2798/7Jrly5dr5cqV2rp1q2zb1saNGyVJq1at0sKFC9XS0qKpU6dq3bp12edYlqUVK1YonS6OUV3TgbYTwzBYbhAAAMCDRjzkalmWDh48qD/84Q8yTVM1NTW677779MADD5z1+u3bt2vGjBmqrKxUOBxWU1OTWlpasuePHDmi/v5+TZ8+XZK0YMECtbS0KJ1Oa+fOnWpqahp2fMh3vvMdzZw5U1VVVRf1gb3GibYTSQqXBdXbVxy/oAAAABSLEYXvtWvX6oMf/KC+853vaM6cOdq2bZs+//nP64c//OFZ20kk6fjx44rFYtnH8XhciUTinOdjsZgSiYS6uroUiUQUCASGHZek559/Xk8//bT+8i//8sI/qUc5scOlNDjpkrYTAAAAbwmM5KLOzk499NBDmjJlyrDj4XBYX//618/6HMuyZBinQ6Vt28Men+v826+TBtso+vr6tGrVKn3zm9+U7yJ7pGtqIhf1PCdZthSNlEo6/d93EgwG3vHaqooyvX7kpMLhkMLhkGLV4UuuFYUjFou6XQKKBPcScoV7CblS6PfSiMK3aZpnBO+77rpLDzzwgD7wgQ+c9Tl1dXXatWtX9nFbW5vi8fiw821tbdnH7e3tisfjqq6uVnd3t0zTlN/vzz5v165d6ujo0J133ilpcOT805/+tL71rW/pXe9614g+bEdHjyzLW7s+mqalgdRge0h3T/+InpNOZ97x2oBPsiUd7+hRMlmpNpMNd0aLWCyqtrZut8tAEeBeQq5wLyFXCuFe8vmM8w74njd833vvvUokEtq9e7c6OzuzxzOZjA4dOnTeN545c6b+/d//XZ2dnSorK9O2bdv0r//6r9nz9fX1CoVC2r17t2688UZt2rRJjY2NCgaDamho0JYtWzRv3jw1NzersbFRt956q371q19ln/9Hf/RH+va3v60JEya84w/Bq2zbdmSTHUkqZ7lBAAAAzzlv+L799tv16quv6uWXX85OgJQkv9+fnSh5LrW1tVq2bJkWL16sdDqt22+/XdOmTdOSJUt011136b3vfa/WrFmje+65Rz09Pbr22mu1ePFiSYOh//Of/7wefPBBjRs3Tt/4xjdy8FG9xzw1Cu9E+I6c2uWyJ0n4BgAA8ArDtu137MNIJBKqra3NRz2O8lrbSSpt6s6v/1q3f+gKfez979Kvd78xouddd1VMz73Sdt5rTMvW+m2v6L1X1GjJJ65VeWhEHUYoAoXwT3IoDNxLyBXuJeRKIdxLl9R28o//+I/65je/qb/+678+6/nNmzdfWnWjnGk6N/Lt9xkqLwvqZO9Azl8bAAAAF+e84XvJkiWSpH/5l3/JSzGjjWlZkpwJ35IUDQfVTdsJAACAZ5x3zb6pU6dKkm6++WaNGzdON998s5LJpHbu3KlrrrkmLwUWs2zPtz+328sPqSgv0cnkgEbQWQQAAIA8GFHqW7lypR566CG99tpruueee3T48GHdfffdTtdW9JxsO5EGR77TGUs97HQJAADgCSMK388//7zuu+8+PfbYY/qTP/kTrV69WkeOHHG6tqLndNtJRbhEktR+YmTrhwMAAMBZIwrftm3L5/Ppqaee0owZMyRJ/f0Eukvl5FKDkhQ9Fb6Pn+hz5PUBAABwYUYUvi+77DItWbJEhw8f1s0336zPfOYzuvrqq52ureg53fMdCQdlSGonfAMAAHjCiBZ/Xr16tR577DHdeOON2R0o58+f73RtRc/pnu+h5QYZ+QYAAPCGEQ25hsNhNTQ06OTJk3rhhRc0bdo0vf76607XVvScbjuRBiddMvINAADgDSMa+f7mN7+p733ve6qpqckeMwxDjz/+uGOFjQbZCZd+58J3RXmJ3kh0y7ZtGYZz7wMAAIB3NqLwvWnTJm3btq0otpj3ktNtJ870fEuDI999KVM9fensBEwAAAC4Y0Spb9y4cQRvB+Sj7WRoucFEF60nAAAAbhvRyPf73vc+fe1rX9NHPvIRlZaWZo9fe+21jhU2GuSj7SS73GBXUlfWj3HsfQAAAPDORhS+H330UUlSS0tL9hg935duqO0k4GDbSSQclGFIiU5GvgEAANw2ovD9q1/9yuk6RqWhthOfg20nfp+h6mgpyw0CAAB4wIiGXHt7e/WFL3xBd9xxh06cOKGVK1eqt7fX6dqKXj56viUpVlmqRGfS0fcAAADAOxtR+P7iF7+oaDSqjo4OhUIh9fT0aOXKlU7XVvTy0fMtSbGqMiW6+mTbtqPvAwAAgPMbUfjet2+fli1bpkAgoLKyMq1Zs0b79u1zurai5/QOl0PGVpapL5VRT1/a0fcBAADA+Y0ofPveNiHQNM0zjuHCnW47cfZnGa8sk8RygwAAAG4bUeq76aab9G//9m/q7+/Xb37zG/393/+9brnlFqdrK3rZ8O1w20m8ajB8H22nTx8AAMBNIwrfn/3sZxUOhxWNRnX//fdrypQp+ud//menayt6pjnY8x3IQ9tJqMSvNxI9jr4PAAAAzu8dlxp87LHH9N3vflcvv/yySktLdfXVV+uGG25QKBTKR31FLV9tJz7D0GXxiA4muh19HwAAAJzfecP3L37xC61du1Z33XWXpkyZIsMwtHfvXn3pS19SKpXSrFmz8lVnUcrkqe1EkibVRvXEnlZZlu3ouuIAAAA4t/OG74cfflj/+Z//qfHjx2ePXXHFFbruuut09913E74v0VDbST7C8KS6qAZ2WzrWmdT4seWOvx8AAADOdN5+h97e3mHBe8jkyZOVSqUcK2q0MC1bhjHYFuK0SXVRSaL1BAAAwEXnDd9+v/+c59iw5dJZlu14v/eQcTVhBQM+HTxG+AYAAHALi3W7yLTsvPR7S4OTOifGI3qDkW8AAADXnLfn++WXX9YNN9xwxnHbtjUwMOBYUaOFadqOLzP4VpNqo3r6xWOybDsvrS4AAAAY7rzh+7HHHstXHaOSaVmOby3/VpPqovrv3x9R+4k+xavCeXtfAAAADDpv+K6vr89XHaNSxrLl9+ev82dS7dCkyx7CNwAAgAvo+XaRadp5HfkeP7Zcfp9B3zcAAIBLCN8uynfbSTDgU32snBVPAAAAXEL4dpGZ57YTabD15GCim6UiAQAAXED4dlG+206kwUmX3cm0urrZJAkAACDfCN8uMi07L1vLv1V20iWtJwAAAHlH+HaRZVl5XedbkibGIwr4fXrl8Im8vi8AAAAI364yrfy3nZQE/bqyvkL7Dnbl9X0BAABA+HZVvtf5HnLNpCodSvSopy+d9/cGAAAYzQjfLnJjwqUkXTOpWraklxj9BgAAyCvCt4vyvc73kMvHRRUq8dN6AgAAkGeEb1KxbaoAACAASURBVBe5sc63JAX8Pl09sZLwDQAAkGcBtwsYzfLVdmL4DPWmMsOOXVE/Rnte69CRjl5VRkLDzoWCAQX4tQwAACDnHA3fmzdv1oMPPqhMJqM77rhDixYtGnZ+3759WrFihXp7e9XQ0KBVq1YpEAiotbVVy5cvV0dHhyZPnqw1a9aovLxc+/fv1z333KNkMqkxY8boK1/5iurr6538CI4yLUt+v/PhO5U29dwrbcOOpTOmJKnl6YO6on7MsHM3XVOrQIjfywAAAHLNsfHNRCKhtWvXav369WpubtaGDRu0f//+YdcsX75cK1eu1NatW2XbtjZu3ChJWrVqlRYuXKiWlhZNnTpV69atyx7/27/9W/3Xf/2X5syZo2984xtOlZ8Xbiw1OKQqGlIo6NexzqQr7w8AADAaORa+t2/frhkzZqiyslLhcFhNTU1qaWnJnj9y5Ij6+/s1ffp0SdKCBQvU0tKidDqtnTt3qqmpadhxSfr+97+vxsZGWZal1tZWVVRUOFV+Xri12okkGYahuuoyHe1IyrZtV2oAAAAYbRzrLTh+/LhisVj2cTwe1549e855PhaLKZFIqKurS5FIRIFAYNhxSQoEAjp58qTmzJmj/v5+/eAHP3Cq/LwYHPl2r7m6riasg4kedSfTqigvca0OAACA0cKx8G1Zlgzj9KiubdvDHp/r/NuvkzTscUVFhZ588kk98cQTuvPOO/X444/L7/ePqKaamsjFfhxH2LatSCSkWCyq451JRSOlI3peMBgY8bXnu/7KidX67YvH1dUzoPra0/+KEA6HFKsOj/j14T2xWNTtElAkuJeQK9xLyJVCv5ccC991dXXatWtX9nFbW5vi8fiw821tpycBtre3Kx6Pq7q6Wt3d3TJNU36/f9jztmzZoo997GMyDEONjY3q7+/Xm2++qerq6hHV1NHRI8vyTotF2rQ0kEqrra1b8vvV3dM/suelMyO+9nzX+2QrUhbUa4dPaFLt6V9MksmU2kxzxK8Pb4nFooP3FHCJuJeQK9xLyJVCuJd8PuO8A76O9TzMnDlTO3bsUGdnp/r6+rRt2zY1NjZmz9fX1ysUCmn37t2SpE2bNqmxsVHBYFANDQ3asmWLJKm5uTn7vO9973t67LHHJElPP/20qqqqRhy8vWiw59u9thPDMDQhVq6jHUllTMu1OgAAAEYLx5JfbW2tli1bpsWLF2v+/PmaO3eupk2bpiVLlmjv3r2SpDVr1mj16tWaPXu2ksmkFi9eLEm69957tXHjRs2ZM0e7du3S0qVLJUlf+cpX9P3vf1+33XabvvWtb+mBBx5wqnzH2bbt6monQybEIzItW0c7WPUEAADAaY4u5jxv3jzNmzdv2LGHHnoo+/cpU6bopz/96RnPq6+vP+tkyiuvvFI/+tGPcl+oC6xTK4zkY53v86mtDivo9+nw8R5NjHurJx4AAKDYsI+hS0zzVPh2eeTb7zM0fmxYh9t6WHIQAADAYYRvl5jWUPh2/yuYEI+oL2Wq42TK7VIAAACKmvvJb5TKhm+X204kqT5WLkPS4eM9bpcCAABQ1AjfLjFPrS7idtuJJJWWBDS2skyH2wjfAAAATiJ8u+R024n74VuSJsbL1Xkypd7+tNulAAAAFC3Ct0u81PMtDfZ9S7SeAAAAOMkbyW8U8lLPtySNKS9RNBzUGwnCNwAAgFMI3y7xUs+3NLjb5eV1UR3rTKqnj9YTAAAAJxC+XeK1thNJuqwuKtuW9uxvd7sUAACAouSd5DfKeK3tRJKqoyFFw0H9/lXCNwAAgBMI3y4Z2uEy4JG2E2mw9eSy2qheeaOL1hMAAAAHEL5dYlre6vkecnldVJYt/f6VNrdLAQAAKDqEb5dksm0n3voKqitCqq4IadfLhG8AAIBc81byG0WG2k58Hhv5NgxD1787phf/0MmGOwAAADlG+HaJ5bEdLt9q+lVjZVq2nmXiJQAAQE4Rvl3i1Z5vSZpUG1VNRUg7XzrudikAAABFhfDtEtOjPd/SYOvJjVfH9cKBTiX7M26XAwAAUDS8l/xGCS8uNfhWN02JD7ae7GfiJQAAQK4Qvl2S8XDbiSRNHl+hqmhIu14ifAMAAOQK4dslXm47kSSfYajh6rieP9ChvhStJwAAALngzeQ3Cgy1nXh15FsabD3JmLae3c+qJwAAALlA+HbJ6ZFv74bvd9UPtZ6w6gkAAEAuEL5d4uWlBof4DEM3XhXT3tc7aT0BAADIAcK3Swqh7USSGqbElTEt7Xmtw+1SAAAACh7h2yWnd7j09ldw5YQxGhMp0TP7Em6XAgAAUPC8nfyKWMayZUjyeXzk22cYumlKXHtfZ8MdAACAS0X4dolpWZ6ebPlWt1xTq4xp6XevsOY3AADApSB8u8Q0bc+3nAx51/gKjR1Tqt/SegIAAHBJAm4XMFqZlu3ZyZaGz1Dv21Y3uf6qmB7fdUjHupKKhkuGnQsFAwoUxu8RAAAAriJ8u8S0bM+2naTSpp57W4tJaYlfli01P/G6pkyqGnbupmtqFQhxKwEAALwTxitdYpqWZ0e+z6YqGlJlpEQHjna7XQoAAEDBIny7ZLDtpLB+/JePq1DbiT719KXdLgUAAKAgFVb6KyJebjs5l8njopKkPxxj9BsAAOBiEL5dUmhtJ5IUDZdo7JhSvXbkTdm27XY5AAAABYfw7RIvr3ZyPu+eOEZv9gyo7USf26UAAAAUHMK3Swqx51uSLq+rUNDv0yuH3nS7FAAAgIJTeOmvSKQzloLBwvvxBwM+TR5foYPHupVKm26XAwAAUFAKL/0ViXTGUtBfmD/+qyaOkWnZer31pNulAAAAFJTCTH9FYCBjqqRAt4WsrihVzZhSvXroBBMvAQAALkBhpr8ikM5YChZo+JakqyaM0YmeAbWf6He7FAAAgIJRuOmvwA2Gb7/bZVy0y8cNTbw84XYpAAAABYPw7ZJCH/kemnh54Fi3etnxEgAAYEQcTX+bN2/WnDlzNGvWLD3yyCNnnN+3b58WLFigpqYmrVixQplMRpLU2tqqRYsWafbs2brzzjvV29srSXrttde0aNEi3XbbbfqzP/sz7du3z8nyHVXo4VuSplxWKcuyteP5Y26XAgAAUBAcS3+JREJr167V+vXr1dzcrA0bNmj//v3Drlm+fLlWrlyprVu3yrZtbdy4UZK0atUqLVy4UC0tLZo6darWrVsnSbrnnnu0ZMkSbdq0SUuXLtXnPvc5p8p33EDGKtgJl0MqoyHVVpfpN3taZVlMvAQAAHgnjqW/7du3a8aMGaqsrFQ4HFZTU5NaWlqy548cOaL+/n5Nnz5dkrRgwQK1tLQonU5r586dampqGnZckj75yU/q1ltvlSRdffXVOnr0qFPlO8q2bWXMwh/5lqQpl1Wp82RKe17rcLsUAAAAz3Ms/R0/flyxWCz7OB6PK5FInPN8LBZTIpFQV1eXIpGIAoHAsOPSYBD3+wcnKT7wwAP66Ec/6lT5jsqYliQVRfieGI+oMlKiX/3usNulAAAAeF7AqRe2LEuGYWQf27Y97PG5zr/9OklnXPe1r31Nzz33nB5++OELqqmmJnKhH8MRPckBSVJVZVixWFSSdLwzqWikdETPDwYDI742H9d/6IaJan7iNQ3IUH3MGz/j0W7ovgIuFfcScoV7CblS6PeSY+G7rq5Ou3btyj5ua2tTPB4fdr6trS37uL29XfF4XNXV1eru7pZpmvL7/cOel8lk9LnPfU6JREIPP/ywotEL++F3dPR4ojf5RE9KkpTqT6utrXvwoN+v7p6RrZmdTmdGfG0+rr9pSkybn3xdP/3ly1r40atG/Dw4IxaLnr6vgEvAvYRc4V5CrhTCveTzGecd8HWs72HmzJnasWOHOjs71dfXp23btqmxsTF7vr6+XqFQSLt375Ykbdq0SY2NjQoGg2poaNCWLVskSc3NzdnnffWrX1VPT4++973vXXDw9pJ0ZrDtpNAnXA6pKC/RTVPienLPUSX7M26XAwAA4FmOpb/a2lotW7ZMixcv1vz58zV37lxNmzZNS5Ys0d69eyVJa9as0erVqzV79mwlk0ktXrxYknTvvfdq48aNmjNnjnbt2qWlS5eqs7NTjzzyiA4cOKBPfvKTuu2223Tbbbc5Vb6jBjLF0/M9pOnmy9Q/YOrXzx5xuxQAAADPcqztRJLmzZunefPmDTv20EMPZf8+ZcoU/fSnPz3jefX19frBD35wxvEXX3wx90W6IFOE4XtSXVTvubxKj+06pI82TCyqzwYAAJArJCQXDGRMScUVviVp9i2X6UTPgJ5+kU13AAAAzqa40l+BGOr5DvqL68d/7eXVmhiPaOszh2TZ7k9sBQAA8JriSn8FIjvhMuh3uZLcMgxDs2+5TK3tvdrLpjsAAABnIHy7oFhHviXppilxVVeE9IunD8pm9BsAAGCY4kt/BSAbvoPF9+MP+H362C2T9MrhN7XvYJfb5QAAAHhK8aW/ApCdcFkkI9+Gz1BvKpP9c+OUuKqiIf3s16+rpz897FxvKqNTv3sAAACMOo4uNYizSxfZUoOptKnnXmkbduzqyyr19AsJbfrN62dsOX/TNbUKhLj1AADA6FMc6a/ApM2hHS6La8LlW11ZP0aRsqCefbWd3m8AAIBTCN8uSKeLa+T7bHw+Q9OuqFHHyZQOHe9xuxwAAABPKN7052Fp05LfZ8jnM9wuxVHvGl+hijCj3wAAAEMI3y4YSFtFPeo9xOczdN2VY3WiZ0B/ONbtdjkAAACuK/4E6EFpc3SEb0m6fFxUlZES7dnfwa6XAABg1BsdCdBj0hlTJaMkfBvG4Oj3m70DOtB60u1yAAAAXDU6EqDHpDOWAkW80snbXVYbUVU0pD2vdciyGP0GAACjF+HbBemMNWpGvqXB0e/p7x6r7mRarx150+1yAAAAXDN6EqCHDGRGT8/3kAmxco0dU6o9r3VkNxkCAAAYbUZXAvSIdMYqmq3lR2po9Lu3P6PfPNfqdjkAAACuGF0J0CPSGUvB4Oj70Y8fW65xNWFt/e0b6u1Pu10OAABA3o2+BOgB6Yw56ka+h9x4dUx9qYy27DjodikAAAB5NzoToMvSGUslwdGz2slbVVeUquGauB7bdVgdb/a7XQ4AAEBeEb5dMDAKe77fau7MyyXZav7N626XAgAAkFejNwG6KD0KVzt5q+qKUn3kxgna/vwxHTjKxjsAAGD0GL0J0EWjaXv5c5k3c7Ki5SX64baX2XgHAACMGqM7AboknSZ8h0sD+rMPX6kDR7v1xB6WHgQAAKPD6E6ALjAtS5Ztj6odLs9lxrW1umpipX72P6+pOzngdjkAAACOIwHm2dDujsHA6Fzt5K0Mw9BfzLpKfSlTP/v1a26XAwAA4DjCd54NZMM3P3pJmhCL6KMNE/TEc0f1yqETbpcDAADgKBJgnmUI3zJ8hnpTmeyfWTdfpuqKkL63ZZ9O9A4MO9ebyujUjwwAAKDgBdwuYLRh5FtKpU0990rbsGM3XBXTL3cd1vd//qKuvyo27NxN19QqEOJWBQAAhW/0JkCXDPV8M+FyuPFjy3XF+Ao9f6BTnSfZ+RIAABQnEmCepRn5PqeGKXGFgn5tf/4Ya38DAICiRALMs3TGlKRRvb38uYRK/LrlPbXqPJnS8wc63S4HAAAg50iAeZYd+Q6y1ODZTKqL6vK6qJ7b366ON2k/AQAAxYXwnWfZCZeMfJ/TLe+pVWmJX0/tPSrTZKkTAABQPEiAeZadcBnkR38uoRK/Zk6t04meAf3+1Xa3ywEAAMgZEmCepRn5HpH6WERXTRyjF//QpZcOdrldDgAAQE6QAPMsO+GS1U7e0Y1Xx1UZKdH3f75Px7uSbpcDAABwyUiAeXZ6qUEmXL6TYMCnD99QL0n690f3qn8g43JFAAAAl4bwnWfscHlhouES/eXHr1Fre6+++/N9sm3W/wYAAIWLBJhn6YwlQ1LAb7hdSsGYMqlKn/zQldr9cpv+z28OuF0OAADARQu4XcBokzYtBQM+GQbh+0I03TxRxzp79f9t/4NiY0p163Xj3S4JAADgghG+8yydtmg5uQiGYegvZl2tjpMpPbz1ZVVXlOraydVulwUAAHBBHE2Bmzdv1pw5czRr1iw98sgjZ5zft2+fFixYoKamJq1YsUKZzOCEutbWVi1atEizZ8/WnXfeqd7e3mHP+8lPfqLPf/7zTpbumLRpEr4vUsDv09/On6pxNWGta96rw8d73C4JAADggjiWAhOJhNauXav169erublZGzZs0P79+4dds3z5cq1cuVJbt26VbdvauHGjJGnVqlVauHChWlpaNHXqVK1bt06SlEqltGbNGn35y192qmzHDWQY+b5Qhs9Qbyqj3lRGlqRP3zZVJQG/vrHxWbV29GbP9aYyyrAhJgAA8DDHUuD27ds1Y8YMVVZWKhwOq6mpSS0tLdnzR44cUX9/v6ZPny5JWrBggVpaWpROp7Vz5041NTUNOy5JO3fulGVZWr58uVNlOy6dsVTCMoMXJJU2tXNfIvtn/+ETuvW6cerpS+vrP35W2/cezZ5LpVmOEAAAeJdjPd/Hjx9XLBbLPo7H49qzZ885z8diMSUSCXV1dSkSiSgQCAw7Lkkf+MAH9IEPfECPPvroRdVUUxO5qOflkuHzqaw0oFgsOuz48c6kopHSEb1GMBgY8bXFen00UqrZM3z6+fYDeur5Y/r4zMny+QyFwyHFqsMjfu1i9fb7C7hY3EvIFe4l5Eqh30uOhW/Lsoat6GHb9rDH5zr/9usk5WxlkI6OHlmWu+tE9yYHJEltbd3DT/j96u7pH9FrpNOZEV9bzNdXR0t0y3tq9fQLCT2+86BueU+tksmU2kxzxK9djGKx6Jn3F3ARuJeQK9xLyJVCuJd8PuO8A76OtZ3U1dWpra0t+7itrU3xePyc59vb2xWPx1VdXa3u7m6ZpwLU259X6AbbTuj5zpWrJlZq6uRqvXLoTb1woNPtcgAAAM7LsRQ4c+ZM7dixQ52dnerr69O2bdvU2NiYPV9fX69QKKTdu3dLkjZt2qTGxkYFg0E1NDRoy5YtkqTm5uZhzyt0AxlTQT/hO5euv2qsLq+L6nevtOt3L7e98xMAAABc4lgKrK2t1bJly7R48WLNnz9fc+fO1bRp07RkyRLt3btXkrRmzRqtXr1as2fPVjKZ1OLFiyVJ9957rzZu3Kg5c+Zo165dWrp0qVNl5l06YykYZMJlLhmGofe/t07xqjL9YOtLevmNLrdLAgAAOCtHN9mZN2+e5s2bN+zYQw89lP37lClT9NOf/vSM59XX1+sHP/jBOV93wYIFWrBgQe4KzaN0xmLk2wF+v08fur5eTzzbqvt/skdLPzlNV19W5XZZAAAAw5AC8yzNOt+OKS3x6x9un6bqipDu/8keRsABAIDnkALzjAmXzqooL9E/f+p6VVeEtPYnzxHAAQCAp5AC84wdLp03JhLSP3/qetVUlBLAAQCAp5AC88i2bWVMwnc+EMABAIAXkQLzKGNakkT4zhMCOAAA8BpSYB4NZIbCN0sNOsXwGepNZbJ/AkG//v5Pp6k6Wqq1G5/Ts/vbh50/9ZUAAADkhaNLDWK49Kmkx4RL56TSpp575cyNdm69bpy2PXNI6/7PXv3RDRNUVxOWJN10Ta0CIf5nAAAA8oMUmEfpDG0nbikLBTTr5okqLw3qV787rKMdvW6XBAAARiFSYB4NEL5dNRTAI2VBPb7rsPYfftPtkgAAwChDCsyjDOHbdWWhgGbfcpni1WFtf/6YNj91QJZtu10WAAAYJUiBeTSQMSURvt1WEvTrozdO0JUTxmjbM4f0wE/3qPNkv9tlAQCAUYAUmEenJ1yy2onbfD5D77u2Vrd/6Aq9dLBL//Ld3+p/nj3CKDgAAHAU4TuPmHDpLYZh6IPX1+sL/9fNmlQb1cMtL+srj/xOB46edLs0AABQpEiBeUT49qZ4VVjLP3W9/vJjU3S8M6l//X936aHNL9CKAgAAco4FjvOI8O09Q5vySNINU+K6ZnK1frnrkH61+7B2v9KmeTMnq3H6ePl8hiQpFAyIrw8AAFwswnceZSdc+klvXnG2TXnqqsP6xPsn67cvJvSzX7+m/3n2iGZOrVVVtJRNeQAAwCUhBeZRdsJlkAmXXhcJB/VHN9br1mnj1NuX1s+3H9TzBzqZkAkAAC4JQ3h5lDZPtZ0w8l0QDMPQ5PEVGje2XE+/cEy/e7lN3ckB/d/zrlV1Ranb5QEAgAJECsyjdJqe70JUWuLXB6eP1/um1ungsW79y3ef0f/8nmUJAQDAhSMF5lHatOT3GdnJeygchmHo3RPG6PN/caMm1Ub08NaX9dVHfqfW9l63SwMAAAWE8J1HA2mLUe8CF6ssG1yWcM4Utbb36t7vPaNHtr2ik8kBt0sDAAAFgJ7vPEqblkoI3wXPMAzdOm28rrtirJp/87r++/dHtP2Fo/rYLZP0kRsnqIzVUAAAwDmQEvIonTEZ+S5wb10X3B/w6U8/fKXeP228Nj35uh594nX94rdvqPG6cfrg9fWKhktYFxwAAAxD+M6jdMZSMMAyg4XsbOuCS9INV8V0WW1Ez7/eqa3PHNIvdx3WuyeM0af++CpNjEVcqBQAAHgR4TuPBsM3w6DFauyYMn3o+nq92ZPS8wc69fKhE1r1/Z2a8Z5a/XHDRE2qi7pdIgAAcBnhO48GCN+jwphISO9/7zhNv3KsOk6mtP35o9r+/DFdWT9GH7lxgm68OqYAa70DADAqEb7zKJU2mXA5ipSXBfWhGyboTz/4Lj2156h+9bsj+n/+6wWNiZTow9Pr9cHp4zUmEnK7TAAAkEeE7zyxbVvHOpK64aqxbpeCPDJ8hmxLev914/W+aeP04oFOPfFcq5qfPKDN2/+g668aq8bp9bq8LirDMJigCQBAkSN858nJZFo9fWnVj2Xy3WhytgmaN02J6+qJlXrpjS4992qHdr3UppqKUk2ZVKk//dCVqiwvcalaAADgNMJ3nhxu65EkTYiVu1wJvKCivEQ3X1Or698d02utb+rlgyf01N5jem5/hz44fbw+MG2caqvCbpcJAAByjPCdJ0eOD4bv+jgj3zgtGPBpymVVunpipY51JnWss09bnj6on+84qMnjorrlPXW67ooaxavKZBiG2+UCAIBLRPjOk8NtvaooL1FFmJYCnMkwDI2rKdcnPvAupVIZPbPvuH77YkI/fvxV/fjxV1UZKRkM6ZdVasplVYRxAAAKFOE7T46096h+LC0nOD/DZygUCujW6eN16/TxajvRp1feOKFXD5/Qi3/o0tMvJiRJY8pL9O6JY3TNpGq9ZxJhHACAQkH4zgPLtnWkvVcfvK7e7VLgcWeboFkS9OnaydV6z+VV6k6mT7WnJPXCgU7temnw2spIiaZMqtKUy6r0/usnyGfbhHEAADyI8J0HbSf6NJC2mGyJS2IYxmDrUnmJrppYKdu2dfm4Ch082q2X3ujSiwc69fQLCf3nL15STUVIUyZV6aoJlZo8rkLjx5bL5yOMAwDgNsJ3Hhw+3itJqo8x2RK5YxiGamvKFa8K66b31A6uJd+Z1MFEr154rV2/f7VdT+09JkkqCfg0qS6qyeMqTv2JKlZJqwoAAPlG+M6DI+09MiR6vpFzZ2tTqYyWatqVNXrvFdU62ZtWx8k+tb/Zr9SApV/97ogy5iFJUnlpQJPHVWhibUS1VWHVVpUpVlmmymhIPkI5AACOIHznweG2XsUqyxQq8btdCkYRwzA0JlKiMZESvWv8GN18bZ0yaVOtHUm9caxbBxPdOnisWy8e7JJl2dnnBf0+1YwpVW11WONrwqqrDmtcTbnqasKKlAVd/EQAABQ+wnceHGnrUT393nDZW0fJS4I+vXvCGL17whhZlq3e/rS6k0N/BtSdTCvRmdQLBzqUMU8H80hZUOPeEshrq8tUU1GqymhI0bIgbSwAALwDwrfD0hlTic4+3Xh13O1SgLPy+QxFwyWKvm0N+puvrVMmY6nzZL8SnUkluvp0vDOpRFdSv3+1Xb/Zc3TY9QG/ocpISFXRwT+VkZCqoyFVnnpcFRn8e8Dvy+fHAwDAUwjfDjvakZRl26x0goLz9n7yinBQFeExunLCmMHzA6ZOJgeU7M8o2Z/RmGhIJ7pTOtGT0oGjJ3WiZ0DpjHXG60bDQY05FcyrK0pVFSk5HdCjpaqKhFQW8jOKDgAoSoRvhx1uG9xWfgIrnaDIhEr8ipWUZR9fd1VsWFi3bVsDGSsbzpOp9Fv+nlHHyX693npSPX3pM17b7zMULg0oHAooXBpQWej038OhoMpKA4qUBhQNl2SXX6wIB1UWChDaAQCe5mj43rx5sx588EFlMhndcccdWrRo0bDz+/bt04oVK9Tb26uGhgatWrVKgUBAra2tWr58uTo6OjR58mStWbNG5eXlOnnypD772c/q0KFDqq6u1v33369YLObkR7hkh9t6FfAbileVvfPFQBExDEOhoF+hoF9V0dAZ52++tk62ZSudsfRmb0onugcG/9szoN6+tPpSmbf8MdXZnVKyf/Dx2UbUpcHQXl4aVKhk8H1LS/wKBX0qCfqzx0oCfoVKfIP/DfpVEvRl6ywp8Q+G/FBAZafCP20yAIBccix8JxIJrV27Vo8++qhKSkr053/+57rlllt05ZVXZq9Zvny5vvjFL2r69Om6++67tXHjRi1cuFCrVq3SwoUL9fGPf1z/8R//oXXr1mn58uW6//771dDQoG9/+9tqbm7Wl770Jd1///1OfYScONLWq3E15fwfOPA2Z1smURpqbzlzVZW3jqyblq3UgKn+gYz6h/6bMtU3YCqVNjUmEtLAqb/3pjLq7E5pIG1pIGMqnbGUSpuy7TPe4qxKgj6VhQIK+n3y+30K+P7/9u49KKry/wP4+yy7qARpzEAaYhdTU2qkJAe1UJkRFC/OZgAADxhJREFUVFhFxsoJUIcix9ExbxCUjooXaHMUEWdyusxkMWOaikGKmoqNSt4SMfKCo6KiIhdDLgt7+/z+kO/mAmv4K3ZJ3q8ZZvd5znOe8znsZ858dvc5oMDFRYGLSgW1iwK1iwouqvuPGrXKtphvelS7KFCp7v+4KA88VylQ2Wm3eK5SoFJgf9uDcykKFAXWR34bQETUcbRb8X306FEEBgaiR48eAIDQ0FDk5uZi9uzZAIDS0lI0NDTA398fABAZGYn09HS89dZbOHHiBDZs2GDtj46ORnx8PPLy8pCZmQkACA8PR3JyMoxGIzSatv35M2f8h7+Xnn0KT3RT/+2xRaXArWvbzkPtomrz2M42viPF4qzx3bqoYTa1nOO/EPujjPdwsz924POeOHelyu52EYGIwGQWmM0WvNjnKRgMJhhMlqai3gy9wWR93mAwwWQWWCwCs0UgAphMZphFYG7qN1kEdY1G/FknMBrNMJgsMJlb/4TeGf5XjKOpGL//9H5Br+B+G0pTu2m7i0oBBE37/K/fdgygQIW/2irVA3M1HVdp2q5q6leavzmAgqap7Af/YNP+pgf67e/zYONhV2bF3rgWO7U+y8PjVFrd1nKY0vq2h+yjtOUElebN1uN52G72jtPaqXbrqoG+wdjm49iLu8W5tuH1f9jLZS9L2nqch+Wm7XH4Bvjf8sQTlairM7RprEatwmv9vdDVwX/q+e9qvnYrvu/cuWOzJMTb2xuFhYV2t3t5eaGsrAx3796Fu7s71Gq1TX/zfdRqNdzd3VFVVYWnn366TTE99ZTjb3qcGu7X5rFhb/Zt89gXej/1SHF0pvEdKZaONr4jxdLe4x91biIiIkdot7UQFovF5p2niNi07W1vPg6w/5WpiECl4nIOIiIiIvpvaLfKtWfPnigv/2s9Z3l5Oby9ve1ur6iogLe3Nzw9PVFTUwOz2dxiP29vb1RUVAAATCYT6urqrMtaiIiIiIg6unYrvocPH478/HxUVVVBr9dj7969CAoKsm738fFBly5dcOrUKQDAzp07ERQUBI1Gg4CAAOzatQsAkJWVZd1v5MiRyMrKAgDs2rULAQEBbV7vTURERETkbIpIW+/5f3TZ2dnYuHEjjEYjJk+ejLi4OMTFxWHOnDl45ZVXcP78eSxatAi1tbXw8/NDSkoKXF1dUVpaisTERFRWVqJXr15Ys2YNunfvjj///BOJiYm4fv06PDw8sHr1avTu3bu9wiciIiIi+le1a/FNRERERER/4d2KREREREQOwuKbiIiIiMhBWHwTERERETkIi28iIiIiIgdh8e1k2dnZGD9+PEJCQpCZmenscOg/ICYmBmFhYZg4cSImTpyIM2fO2M2jo0ePQqvVIiQkBGvXrnVi1NSR1NbWIjw8HDdu3ABgP0/OnTuHyMhIhIaG4pNPPoHJZAIA3Lx5E1FRURg7dixmzpyJuro6p5wHOV/zXEpKSkJISIj1+rRv3z4Aj55j1LlkZGQgLCwMYWFh0Ol0AB7z65KQ09y+fVtGjx4td+/elbq6OtFqtVJcXOzssKgDs1gs8sYbb4jRaLT22csjvV4vI0eOlGvXronRaJTY2FjJy8tzYvTUERQUFEh4eLj4+fnJ9evXH5onYWFhcvr0aRERSUpKkszMTBER+eCDDyQnJ0dERDIyMkSn0znnZMipmueSiEh4eLiUlZXZjPv/5Bh1HkeOHJF33nlHGhsbxWAwyNSpUyU7O/uxvi7xk28nOnr0KAIDA9GjRw+4ubkhNDQUubm5zg6LOrDLly8DAGJjYzFhwgR89913dvOosLAQzz77LHx9faFWq6HVaplfhC1btmDJkiXW/xxsL09KS0vR0NAAf39/AEBkZCRyc3NhNBpx4sQJhIaG2vRT59M8l/R6PW7evImPP/4YWq0W6enpsFgsj5xj1Ll4eXkhMTERrq6u0Gg06Nu3L65evfpYX5fUzg6gM7tz5w68vLysbW9vbxQWFjoxIuro7t27h2HDhmHx4sUwGo2YOnUqxo0b12oetZZfZWVlzgibOpCVK1fatO3lSfN+Ly8vlJWV4e7du3B3d4darbbpp86neS5VVFQgMDAQS5YsgYeHB2bMmIEffvgBbm5uj5Rj1Ln069fP+vzq1avYvXs3oqOjH+vrEj/5diKLxQJFUaxtEbFpEzX36quvQqfTwcPDA56enpg8eTLS09NbzSPmF7WFvTyx199aHjGvCAB8fX2xYcMGeHt7o1u3boiJicGhQ4ceOceocyouLkZsbCwSEhLg6+v7WF+XWHw7Uc+ePVFeXm5tl5eXW7++I2rNyZMnkZ+fb22LCHx8fFrNI+YXtYW9PGneX1FRAW9vb3h6eqKmpgZms9lmPNGFCxewZ88ea1tEoFarHznHqPM5deoUpk+fjgULFmDSpEmP/XWJxbcTDR8+HPn5+aiqqoJer8fevXsRFBTk7LCoA6upqYFOp0NjYyNqa2uxY8cOfPbZZ63m0eDBg3HlyhWUlJTAbDYjJyeH+UUt2MsTHx8fdOnSBadOnQIA7Ny5E0FBQdBoNAgICMCuXbsAAFlZWcwrAnC/2F61ahWqq6thNBrx/fffY8yYMY+cY9S53Lp1C7NmzcLq1asRFhYG4PG/LikiIs4OojPLzs7Gxo0bYTQaMXnyZMTFxTk7JOrg0tLSsGfPHlgsFrz77ruYNm2a3TzKz89HSkoKGhsbMXLkSCQlJXXor+LIcYKDg7Fp0yb07t3bbp6cP38eixYtQm1tLfz8/JCSkgJXV1eUlpYiMTERlZWV6NWrF9asWYPu3bs7+5TISR7MpczMTGRmZsJkMiEkJAQLFy4EYP9aZC/HqPNYsWIFtm3bhj59+lj7pkyZgueee+6xvS6x+CYiIiIichAuOyEiIiIichAW30REREREDsLim4iIiIjIQVh8ExERERE5CItvIiIiIiIHYfFNRNSOYmNjUVVV5ewwWsjIyMDPP/8MAFi3bh2ysrL+8ZzffvstBgwYgIKCgn88178hLy8P69atc3YYREQ2WHwTEbWjI0eOODuEVh07dgwmkwkA8OGHHyIiIuIfz7l582ZotVp88803/3iuf8PZs2dRXV3t7DCIiGyonR0AEdHjKikpCQAwbdo0XLp0CaGhobhw4QLmz58PtVqNjRs3wmAwoKqqChEREZg7dy6OHTuGtWvXwtfXF8XFxTCZTFi2bBmGDBmCkydPIjU1FRaLBQAwY8YMhIaG4sqVK0hOTkZdXR3Ky8vx0ksvIS0tDV26dMGZM2ewYsUK6PV6aDQaJCQk4PLly/j999+h0+ng4uKC/fv3o1+/fnjvvfdw8uRJ6HQ66/i5c+ciKCgI27dvx759+6BSqVBSUoKuXbvi008/Rd++fQHcL+arq6sRHx+PMWPG4NatW+jVqxcAICYmBn5+figoKEBVVRXefvttVFRU4Pjx49Dr9UhLS8OAAQNw+/ZtLF26FKWlpRARRERE4P3338eNGzeg1Wpx+vRpALBp24urtrYWmzdvhtlshoeHB+bNm+eEDCAiaoUQEVG76d+/v1RWVsro0aMlIyNDREQsFotER0fLlStXRETk9u3bMnDgQKmsrJRff/1VBg4cKH/88YeIiHz11VcSFRUlIiJTp06VnJwcERE5d+6cLF26VEREUlNTJSsrS0REDAaDhIeHS25urhgMBhkxYoQcPHhQRETOnj0r4eHhYjabJTo6Wnbv3i0iIh999JF8+eWXUlVVJcOGDZOCggIREbl48aIMHTpUrl27Jtu2bZMhQ4bIrVu3REQkOTlZEhISrOc5Z84cSU1NFRGRuLg40el01m3R0dEye/ZsEREpKCiQ/v37y/79+0VEZOXKlbJo0SIREYmKipKvv/5aRETu3bsnWq1WcnJy5Pr16+Lv72+d78H2w+JKT0+XZcuWPfqLRkTUjrjshIjIQQICAgAAiqLg888/R1FRETIyMpCamgoRgV6vBwA888wzGDhwIABg0KBB1qUT48aNQ3JyMhYsWICioiLMnz8fABAfHw9PT0988cUXWLp0Ke7cuYP6+npcvHgRKpUKo0aNAgC8/PLLyM7OhkrV+qW/sLAQffr0weDBgwEA/fr1w2uvvYbjx48DAPz8/NCzZ88WcZWXl2P//v3WpSsRERHYunUr6uvrrXOPGTMGAODr6wsAePPNNwEAffr0QXV1Nerr6/Hbb78hKioKAODh4YHIyEj88ssvf/t7tRcXEVFHxOKbiMhB3NzcAAD19fWYNGkSioqKMGjQICQkJECtVkNEAABdu3a17qMoirV/ypQp+PHHHzFixAgcPnwYEyZMQGNjI+bPn48tW7bAx8cH06dPh5+fH0QELi4uUBTFJoaLFy9a13o3ZzabW4wXEet4e3Ft2bIFADBz5kwEBwdDp9OhtrYWO3bssI53dXW1mVej0di0LRaLdb4H+0wmk82xAMBoNNqMsxcXEVFHxOKbiKgdubi4tCh2S0pKUFtbi7lz5yI4OBjHjh2DwWCwruW2Z8qUKTh37hwiIyOxfPly3Lt3D+Xl5Th8+DBmzZqF8ePHAwDOnDkDs9mMF154AYqiWG/6LCoqwrRp02CxWFqNy9/fH5cvX0ZhYSEAoLi4GCdOnMDQoUPtxmQ2m7F161YsW7YMBw4cwIEDB5CXl4cZM2Zg06ZNbS6E3d3dMXjwYGRmZgIAampqkJWVheHDh+PJJ5+E0WjEpUuXAAA//fRTm+Zs7RyJiJyNN1wSEbWjsWPHIiYmBnV1dda+AQMGYNSoURg3bhxcXV3Rv39/vPjiiygpKWnxCfGDFi5ciFWrViEtLQ2KomD27Nno3bs35s2bh1mzZsHNzQ3u7u54/fXXce3aNbi6umL9+vVYtWoVdDodNBoN1q9fD1dXVwQHB2PNmjU2nyJ7enpi3bp1WL58ORoaGqAoClJSUvD8889bb3Zs7uDBg7BYLNBqtTb906dPx6ZNm3Do0KE2/65Wr16N5ORkbN++HQaDAVqtFpGRkVAUBfHx8YiLi4OnpyfGjh3bpvkCAwOxcOFCLF++HIsXL25zHERE7UkRfj9HREREROQQXHZCREREROQgLL6JiIiIiByExTcRERERkYOw+CYiIiIichAW30REREREDsLim4iIiIjIQVh8ExERERE5CItvIiIiIiIH+T+P9EpCt205NwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 864x504 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(12, 7))\n",
    "sns.distplot(transactions.transactionAmount)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[Text(0, 0.5, 'Freq')]"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x504 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# visualize transaction amounts by using the histogram\n",
    "sns.set()\n",
    "plt.figure(figsize=(12, 7))\n",
    "plt.title('Distribution of Transaction Amounts', weight='bold')\n",
    "ax=sns.histplot(data=transactions, x=\"transactionAmount\", color='blue',bins=15)\n",
    "ax.set(xlabel='Amount')\n",
    "ax.set(ylabel='Freq')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**61% of all transactions are less than specific transaction amount ($125) in the dataset.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "z=transactions[(transactions.transactionAmount < 125)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.61"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "round(len(z)/len(transactions),2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**This table indicates how many transactions each account had for each merchant in year 2016.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "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>customerId</th>\n",
       "      <th>merchantName</th>\n",
       "      <th>Count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>100088067</td>\n",
       "      <td>1st BBQ</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>100088067</td>\n",
       "      <td>1st Pub</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>100088067</td>\n",
       "      <td>1st Sandwitch Bar #758805</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>100088067</td>\n",
       "      <td>34th BBQ #412895</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>100088067</td>\n",
       "      <td>AMC #191138</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   customerId               merchantName  Count\n",
       "0   100088067                    1st BBQ      1\n",
       "1   100088067                    1st Pub      1\n",
       "2   100088067  1st Sandwitch Bar #758805      1\n",
       "3   100088067           34th BBQ #412895      1\n",
       "4   100088067                AMC #191138      1"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cust_length=transactions.groupby(['customerId','merchantName']).count()['Date']\n",
    "cust_length=pd.DataFrame(cust_length).reset_index()\n",
    "cust_length.columns=['customerId','merchantName','Count']\n",
    "#cust_length=cust_length.sort_values(by='Count',ascending=False)\n",
    "cust_length.head()\n",
    "#cust_length.to_csv('zazaza.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "-----"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}