Problem_loan_prediction_code.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Problem loan prediction 2017 - Lending Club\n",
    "## Code\n",
    "### For more information about the results of this project, check the report.pdf file"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "from sklearn.preprocessing import LabelEncoder\n",
    "import seaborn as sns\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as pl\n",
    "%matplotlib inline\n",
    "l_encoder = LabelEncoder()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## STEP 1 Uploading the data\n",
    "\n",
    "The data is from the Lending Club. For 2017, currently (November 2017) there are two files, one for each of the first two quarters of the year.\n",
    "\n",
    "Upload each file separately and merge it into a single pandas data frame."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Uploading data into two dataframes\n",
    "\n",
    "data_1Q = pd.read_csv('data_files/LoanStats_2017Q1.csv', skiprows=1, low_memory=False)\n",
    "data_2Q = pd.read_csv('data_files/LoanStats3a.csv', skiprows=1, low_memory=False)\n",
    "\n",
    "# Merging the dataframes\n",
    "\n",
    "data_loans = pd.concat([data_1Q, data_2Q])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(139319, 145)\n",
      "HEAD\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>member_id</th>\n",
       "      <th>loan_amnt</th>\n",
       "      <th>funded_amnt</th>\n",
       "      <th>funded_amnt_inv</th>\n",
       "      <th>term</th>\n",
       "      <th>int_rate</th>\n",
       "      <th>installment</th>\n",
       "      <th>grade</th>\n",
       "      <th>sub_grade</th>\n",
       "      <th>...</th>\n",
       "      <th>hardship_payoff_balance_amount</th>\n",
       "      <th>hardship_last_payment_amount</th>\n",
       "      <th>disbursement_method</th>\n",
       "      <th>debt_settlement_flag</th>\n",
       "      <th>debt_settlement_flag_date</th>\n",
       "      <th>settlement_status</th>\n",
       "      <th>settlement_date</th>\n",
       "      <th>settlement_amount</th>\n",
       "      <th>settlement_percentage</th>\n",
       "      <th>settlement_term</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5000.0</td>\n",
       "      <td>5000.0</td>\n",
       "      <td>5000.0</td>\n",
       "      <td>36 months</td>\n",
       "      <td>8.24%</td>\n",
       "      <td>157.24</td>\n",
       "      <td>B</td>\n",
       "      <td>B1</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Cash</td>\n",
       "      <td>N</td>\n",
       "      <td>NaN</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>1</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>9000.0</td>\n",
       "      <td>9000.0</td>\n",
       "      <td>9000.0</td>\n",
       "      <td>36 months</td>\n",
       "      <td>10.49%</td>\n",
       "      <td>292.48</td>\n",
       "      <td>B</td>\n",
       "      <td>B2</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Cash</td>\n",
       "      <td>N</td>\n",
       "      <td>NaN</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>2</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>10150.0</td>\n",
       "      <td>10150.0</td>\n",
       "      <td>10150.0</td>\n",
       "      <td>36 months</td>\n",
       "      <td>7.24%</td>\n",
       "      <td>314.52</td>\n",
       "      <td>A</td>\n",
       "      <td>A3</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Cash</td>\n",
       "      <td>N</td>\n",
       "      <td>NaN</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>3</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4700.0</td>\n",
       "      <td>4700.0</td>\n",
       "      <td>4700.0</td>\n",
       "      <td>36 months</td>\n",
       "      <td>15.99%</td>\n",
       "      <td>165.22</td>\n",
       "      <td>C</td>\n",
       "      <td>C5</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Cash</td>\n",
       "      <td>N</td>\n",
       "      <td>NaN</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>4</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>20000.0</td>\n",
       "      <td>20000.0</td>\n",
       "      <td>20000.0</td>\n",
       "      <td>36 months</td>\n",
       "      <td>8.24%</td>\n",
       "      <td>628.95</td>\n",
       "      <td>B</td>\n",
       "      <td>B1</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Cash</td>\n",
       "      <td>N</td>\n",
       "      <td>NaN</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",
       "<p>5 rows × 145 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "    id  member_id  loan_amnt  funded_amnt  funded_amnt_inv        term  \\\n",
       "0  NaN        NaN     5000.0       5000.0           5000.0   36 months   \n",
       "1  NaN        NaN     9000.0       9000.0           9000.0   36 months   \n",
       "2  NaN        NaN    10150.0      10150.0          10150.0   36 months   \n",
       "3  NaN        NaN     4700.0       4700.0           4700.0   36 months   \n",
       "4  NaN        NaN    20000.0      20000.0          20000.0   36 months   \n",
       "\n",
       "  int_rate  installment grade sub_grade       ...        \\\n",
       "0    8.24%       157.24     B        B1       ...         \n",
       "1   10.49%       292.48     B        B2       ...         \n",
       "2    7.24%       314.52     A        A3       ...         \n",
       "3   15.99%       165.22     C        C5       ...         \n",
       "4    8.24%       628.95     B        B1       ...         \n",
       "\n",
       "  hardship_payoff_balance_amount hardship_last_payment_amount  \\\n",
       "0                            NaN                          NaN   \n",
       "1                            NaN                          NaN   \n",
       "2                            NaN                          NaN   \n",
       "3                            NaN                          NaN   \n",
       "4                            NaN                          NaN   \n",
       "\n",
       "  disbursement_method  debt_settlement_flag debt_settlement_flag_date  \\\n",
       "0                Cash                     N                       NaN   \n",
       "1                Cash                     N                       NaN   \n",
       "2                Cash                     N                       NaN   \n",
       "3                Cash                     N                       NaN   \n",
       "4                Cash                     N                       NaN   \n",
       "\n",
       "  settlement_status settlement_date settlement_amount  settlement_percentage  \\\n",
       "0               NaN             NaN               NaN                    NaN   \n",
       "1               NaN             NaN               NaN                    NaN   \n",
       "2               NaN             NaN               NaN                    NaN   \n",
       "3               NaN             NaN               NaN                    NaN   \n",
       "4               NaN             NaN               NaN                    NaN   \n",
       "\n",
       "  settlement_term  \n",
       "0             NaN  \n",
       "1             NaN  \n",
       "2             NaN  \n",
       "3             NaN  \n",
       "4             NaN  \n",
       "\n",
       "[5 rows x 145 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      " TAIL\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>member_id</th>\n",
       "      <th>loan_amnt</th>\n",
       "      <th>funded_amnt</th>\n",
       "      <th>funded_amnt_inv</th>\n",
       "      <th>term</th>\n",
       "      <th>int_rate</th>\n",
       "      <th>installment</th>\n",
       "      <th>grade</th>\n",
       "      <th>sub_grade</th>\n",
       "      <th>...</th>\n",
       "      <th>hardship_payoff_balance_amount</th>\n",
       "      <th>hardship_last_payment_amount</th>\n",
       "      <th>disbursement_method</th>\n",
       "      <th>debt_settlement_flag</th>\n",
       "      <th>debt_settlement_flag_date</th>\n",
       "      <th>settlement_status</th>\n",
       "      <th>settlement_date</th>\n",
       "      <th>settlement_amount</th>\n",
       "      <th>settlement_percentage</th>\n",
       "      <th>settlement_term</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>42533</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2525.0</td>\n",
       "      <td>2525.0</td>\n",
       "      <td>225.0</td>\n",
       "      <td>36 months</td>\n",
       "      <td>9.33%</td>\n",
       "      <td>80.69</td>\n",
       "      <td>B</td>\n",
       "      <td>B3</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Cash</td>\n",
       "      <td>N</td>\n",
       "      <td>NaN</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>42534</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6500.0</td>\n",
       "      <td>6500.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>36 months</td>\n",
       "      <td>8.38%</td>\n",
       "      <td>204.84</td>\n",
       "      <td>A</td>\n",
       "      <td>A5</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Cash</td>\n",
       "      <td>N</td>\n",
       "      <td>NaN</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>42535</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5000.0</td>\n",
       "      <td>5000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>36 months</td>\n",
       "      <td>7.75%</td>\n",
       "      <td>156.11</td>\n",
       "      <td>A</td>\n",
       "      <td>A3</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Cash</td>\n",
       "      <td>N</td>\n",
       "      <td>NaN</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>42536</th>\n",
       "      <td>Total amount funded in policy code 1: 460296150</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</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>42537</th>\n",
       "      <td>Total amount funded in policy code 2: 0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</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",
       "<p>5 rows × 145 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                    id  member_id  loan_amnt  \\\n",
       "42533                                              NaN        NaN     2525.0   \n",
       "42534                                              NaN        NaN     6500.0   \n",
       "42535                                              NaN        NaN     5000.0   \n",
       "42536  Total amount funded in policy code 1: 460296150        NaN        NaN   \n",
       "42537          Total amount funded in policy code 2: 0        NaN        NaN   \n",
       "\n",
       "       funded_amnt  funded_amnt_inv        term int_rate  installment grade  \\\n",
       "42533       2525.0            225.0   36 months    9.33%        80.69     B   \n",
       "42534       6500.0              0.0   36 months    8.38%       204.84     A   \n",
       "42535       5000.0              0.0   36 months    7.75%       156.11     A   \n",
       "42536          NaN              NaN         NaN      NaN          NaN   NaN   \n",
       "42537          NaN              NaN         NaN      NaN          NaN   NaN   \n",
       "\n",
       "      sub_grade       ...       hardship_payoff_balance_amount  \\\n",
       "42533        B3       ...                                  NaN   \n",
       "42534        A5       ...                                  NaN   \n",
       "42535        A3       ...                                  NaN   \n",
       "42536       NaN       ...                                  NaN   \n",
       "42537       NaN       ...                                  NaN   \n",
       "\n",
       "      hardship_last_payment_amount disbursement_method  debt_settlement_flag  \\\n",
       "42533                          NaN                Cash                     N   \n",
       "42534                          NaN                Cash                     N   \n",
       "42535                          NaN                Cash                     N   \n",
       "42536                          NaN                 NaN                   NaN   \n",
       "42537                          NaN                 NaN                   NaN   \n",
       "\n",
       "      debt_settlement_flag_date settlement_status settlement_date  \\\n",
       "42533                       NaN               NaN             NaN   \n",
       "42534                       NaN               NaN             NaN   \n",
       "42535                       NaN               NaN             NaN   \n",
       "42536                       NaN               NaN             NaN   \n",
       "42537                       NaN               NaN             NaN   \n",
       "\n",
       "      settlement_amount  settlement_percentage settlement_term  \n",
       "42533               NaN                    NaN             NaN  \n",
       "42534               NaN                    NaN             NaN  \n",
       "42535               NaN                    NaN             NaN  \n",
       "42536               NaN                    NaN             NaN  \n",
       "42537               NaN                    NaN             NaN  \n",
       "\n",
       "[5 rows x 145 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "print (data_loans.shape)\n",
    "print('HEAD')\n",
    "display(data_loans.head())\n",
    "print('\\n TAIL')\n",
    "display(data_loans.tail())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## STEP 2 Data Cleaning"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There are a few data descriptors at the end of the files than need to be remove. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>member_id</th>\n",
       "      <th>loan_amnt</th>\n",
       "      <th>funded_amnt</th>\n",
       "      <th>funded_amnt_inv</th>\n",
       "      <th>term</th>\n",
       "      <th>int_rate</th>\n",
       "      <th>installment</th>\n",
       "      <th>grade</th>\n",
       "      <th>sub_grade</th>\n",
       "      <th>...</th>\n",
       "      <th>hardship_payoff_balance_amount</th>\n",
       "      <th>hardship_last_payment_amount</th>\n",
       "      <th>disbursement_method</th>\n",
       "      <th>debt_settlement_flag</th>\n",
       "      <th>debt_settlement_flag_date</th>\n",
       "      <th>settlement_status</th>\n",
       "      <th>settlement_date</th>\n",
       "      <th>settlement_amount</th>\n",
       "      <th>settlement_percentage</th>\n",
       "      <th>settlement_term</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>42531</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3500.0</td>\n",
       "      <td>3500.0</td>\n",
       "      <td>225.0</td>\n",
       "      <td>36 months</td>\n",
       "      <td>10.28%</td>\n",
       "      <td>113.39</td>\n",
       "      <td>C</td>\n",
       "      <td>C1</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Cash</td>\n",
       "      <td>N</td>\n",
       "      <td>NaN</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>42532</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>36 months</td>\n",
       "      <td>9.64%</td>\n",
       "      <td>32.11</td>\n",
       "      <td>B</td>\n",
       "      <td>B4</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Cash</td>\n",
       "      <td>N</td>\n",
       "      <td>NaN</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>42533</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2525.0</td>\n",
       "      <td>2525.0</td>\n",
       "      <td>225.0</td>\n",
       "      <td>36 months</td>\n",
       "      <td>9.33%</td>\n",
       "      <td>80.69</td>\n",
       "      <td>B</td>\n",
       "      <td>B3</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Cash</td>\n",
       "      <td>N</td>\n",
       "      <td>NaN</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>42534</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6500.0</td>\n",
       "      <td>6500.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>36 months</td>\n",
       "      <td>8.38%</td>\n",
       "      <td>204.84</td>\n",
       "      <td>A</td>\n",
       "      <td>A5</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Cash</td>\n",
       "      <td>N</td>\n",
       "      <td>NaN</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>42535</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5000.0</td>\n",
       "      <td>5000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>36 months</td>\n",
       "      <td>7.75%</td>\n",
       "      <td>156.11</td>\n",
       "      <td>A</td>\n",
       "      <td>A3</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Cash</td>\n",
       "      <td>N</td>\n",
       "      <td>NaN</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>39786</th>\n",
       "      <td>Loans that do not meet the credit policy</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</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>96779</th>\n",
       "      <td>Total amount funded in policy code 1: 1437969475</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</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>42536</th>\n",
       "      <td>Total amount funded in policy code 1: 460296150</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</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>42537</th>\n",
       "      <td>Total amount funded in policy code 2: 0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</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>96780</th>\n",
       "      <td>Total amount funded in policy code 2: 520780182</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</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",
       "<p>10 rows × 145 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                     id  member_id  loan_amnt  \\\n",
       "42531                                               NaN        NaN     3500.0   \n",
       "42532                                               NaN        NaN     1000.0   \n",
       "42533                                               NaN        NaN     2525.0   \n",
       "42534                                               NaN        NaN     6500.0   \n",
       "42535                                               NaN        NaN     5000.0   \n",
       "39786          Loans that do not meet the credit policy        NaN        NaN   \n",
       "96779  Total amount funded in policy code 1: 1437969475        NaN        NaN   \n",
       "42536   Total amount funded in policy code 1: 460296150        NaN        NaN   \n",
       "42537           Total amount funded in policy code 2: 0        NaN        NaN   \n",
       "96780   Total amount funded in policy code 2: 520780182        NaN        NaN   \n",
       "\n",
       "       funded_amnt  funded_amnt_inv        term int_rate  installment grade  \\\n",
       "42531       3500.0            225.0   36 months   10.28%       113.39     C   \n",
       "42532       1000.0              0.0   36 months    9.64%        32.11     B   \n",
       "42533       2525.0            225.0   36 months    9.33%        80.69     B   \n",
       "42534       6500.0              0.0   36 months    8.38%       204.84     A   \n",
       "42535       5000.0              0.0   36 months    7.75%       156.11     A   \n",
       "39786          NaN              NaN         NaN      NaN          NaN   NaN   \n",
       "96779          NaN              NaN         NaN      NaN          NaN   NaN   \n",
       "42536          NaN              NaN         NaN      NaN          NaN   NaN   \n",
       "42537          NaN              NaN         NaN      NaN          NaN   NaN   \n",
       "96780          NaN              NaN         NaN      NaN          NaN   NaN   \n",
       "\n",
       "      sub_grade       ...       hardship_payoff_balance_amount  \\\n",
       "42531        C1       ...                                  NaN   \n",
       "42532        B4       ...                                  NaN   \n",
       "42533        B3       ...                                  NaN   \n",
       "42534        A5       ...                                  NaN   \n",
       "42535        A3       ...                                  NaN   \n",
       "39786       NaN       ...                                  NaN   \n",
       "96779       NaN       ...                                  NaN   \n",
       "42536       NaN       ...                                  NaN   \n",
       "42537       NaN       ...                                  NaN   \n",
       "96780       NaN       ...                                  NaN   \n",
       "\n",
       "      hardship_last_payment_amount disbursement_method  debt_settlement_flag  \\\n",
       "42531                          NaN                Cash                     N   \n",
       "42532                          NaN                Cash                     N   \n",
       "42533                          NaN                Cash                     N   \n",
       "42534                          NaN                Cash                     N   \n",
       "42535                          NaN                Cash                     N   \n",
       "39786                          NaN                 NaN                   NaN   \n",
       "96779                          NaN                 NaN                   NaN   \n",
       "42536                          NaN                 NaN                   NaN   \n",
       "42537                          NaN                 NaN                   NaN   \n",
       "96780                          NaN                 NaN                   NaN   \n",
       "\n",
       "      debt_settlement_flag_date settlement_status settlement_date  \\\n",
       "42531                       NaN               NaN             NaN   \n",
       "42532                       NaN               NaN             NaN   \n",
       "42533                       NaN               NaN             NaN   \n",
       "42534                       NaN               NaN             NaN   \n",
       "42535                       NaN               NaN             NaN   \n",
       "39786                       NaN               NaN             NaN   \n",
       "96779                       NaN               NaN             NaN   \n",
       "42536                       NaN               NaN             NaN   \n",
       "42537                       NaN               NaN             NaN   \n",
       "96780                       NaN               NaN             NaN   \n",
       "\n",
       "      settlement_amount  settlement_percentage settlement_term  \n",
       "42531               NaN                    NaN             NaN  \n",
       "42532               NaN                    NaN             NaN  \n",
       "42533               NaN                    NaN             NaN  \n",
       "42534               NaN                    NaN             NaN  \n",
       "42535               NaN                    NaN             NaN  \n",
       "39786               NaN                    NaN             NaN  \n",
       "96779               NaN                    NaN             NaN  \n",
       "42536               NaN                    NaN             NaN  \n",
       "42537               NaN                    NaN             NaN  \n",
       "96780               NaN                    NaN             NaN  \n",
       "\n",
       "[10 rows x 145 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "data_loans = data_loans.sort_values(by = 'id',na_position = 'first', ascending=True)\n",
    "display(data_loans.tail(10))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>member_id</th>\n",
       "      <th>loan_amnt</th>\n",
       "      <th>funded_amnt</th>\n",
       "      <th>funded_amnt_inv</th>\n",
       "      <th>term</th>\n",
       "      <th>int_rate</th>\n",
       "      <th>installment</th>\n",
       "      <th>grade</th>\n",
       "      <th>sub_grade</th>\n",
       "      <th>...</th>\n",
       "      <th>hardship_payoff_balance_amount</th>\n",
       "      <th>hardship_last_payment_amount</th>\n",
       "      <th>disbursement_method</th>\n",
       "      <th>debt_settlement_flag</th>\n",
       "      <th>debt_settlement_flag_date</th>\n",
       "      <th>settlement_status</th>\n",
       "      <th>settlement_date</th>\n",
       "      <th>settlement_amount</th>\n",
       "      <th>settlement_percentage</th>\n",
       "      <th>settlement_term</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>42531</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3500.0</td>\n",
       "      <td>3500.0</td>\n",
       "      <td>225.0</td>\n",
       "      <td>36 months</td>\n",
       "      <td>10.28%</td>\n",
       "      <td>113.39</td>\n",
       "      <td>C</td>\n",
       "      <td>C1</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Cash</td>\n",
       "      <td>N</td>\n",
       "      <td>NaN</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>42532</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>36 months</td>\n",
       "      <td>9.64%</td>\n",
       "      <td>32.11</td>\n",
       "      <td>B</td>\n",
       "      <td>B4</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Cash</td>\n",
       "      <td>N</td>\n",
       "      <td>NaN</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>42533</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2525.0</td>\n",
       "      <td>2525.0</td>\n",
       "      <td>225.0</td>\n",
       "      <td>36 months</td>\n",
       "      <td>9.33%</td>\n",
       "      <td>80.69</td>\n",
       "      <td>B</td>\n",
       "      <td>B3</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Cash</td>\n",
       "      <td>N</td>\n",
       "      <td>NaN</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>42534</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6500.0</td>\n",
       "      <td>6500.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>36 months</td>\n",
       "      <td>8.38%</td>\n",
       "      <td>204.84</td>\n",
       "      <td>A</td>\n",
       "      <td>A5</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Cash</td>\n",
       "      <td>N</td>\n",
       "      <td>NaN</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>42535</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5000.0</td>\n",
       "      <td>5000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>36 months</td>\n",
       "      <td>7.75%</td>\n",
       "      <td>156.11</td>\n",
       "      <td>A</td>\n",
       "      <td>A3</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Cash</td>\n",
       "      <td>N</td>\n",
       "      <td>NaN</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",
       "<p>5 rows × 145 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        id  member_id  loan_amnt  funded_amnt  funded_amnt_inv        term  \\\n",
       "42531  NaN        NaN     3500.0       3500.0            225.0   36 months   \n",
       "42532  NaN        NaN     1000.0       1000.0              0.0   36 months   \n",
       "42533  NaN        NaN     2525.0       2525.0            225.0   36 months   \n",
       "42534  NaN        NaN     6500.0       6500.0              0.0   36 months   \n",
       "42535  NaN        NaN     5000.0       5000.0              0.0   36 months   \n",
       "\n",
       "      int_rate  installment grade sub_grade       ...        \\\n",
       "42531   10.28%       113.39     C        C1       ...         \n",
       "42532    9.64%        32.11     B        B4       ...         \n",
       "42533    9.33%        80.69     B        B3       ...         \n",
       "42534    8.38%       204.84     A        A5       ...         \n",
       "42535    7.75%       156.11     A        A3       ...         \n",
       "\n",
       "      hardship_payoff_balance_amount hardship_last_payment_amount  \\\n",
       "42531                            NaN                          NaN   \n",
       "42532                            NaN                          NaN   \n",
       "42533                            NaN                          NaN   \n",
       "42534                            NaN                          NaN   \n",
       "42535                            NaN                          NaN   \n",
       "\n",
       "      disbursement_method  debt_settlement_flag debt_settlement_flag_date  \\\n",
       "42531                Cash                     N                       NaN   \n",
       "42532                Cash                     N                       NaN   \n",
       "42533                Cash                     N                       NaN   \n",
       "42534                Cash                     N                       NaN   \n",
       "42535                Cash                     N                       NaN   \n",
       "\n",
       "      settlement_status settlement_date settlement_amount  \\\n",
       "42531               NaN             NaN               NaN   \n",
       "42532               NaN             NaN               NaN   \n",
       "42533               NaN             NaN               NaN   \n",
       "42534               NaN             NaN               NaN   \n",
       "42535               NaN             NaN               NaN   \n",
       "\n",
       "       settlement_percentage settlement_term  \n",
       "42531                    NaN             NaN  \n",
       "42532                    NaN             NaN  \n",
       "42533                    NaN             NaN  \n",
       "42534                    NaN             NaN  \n",
       "42535                    NaN             NaN  \n",
       "\n",
       "[5 rows x 145 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The data frame has 139311 rows (loans) with 145 columns\n"
     ]
    }
   ],
   "source": [
    "data_loans.drop(data_loans.tail(5).index,inplace=True)\n",
    "display(data_loans.tail(5))\n",
    "print ('The data frame has %d rows (loans) with %d columns' % (data_loans.shape[0], data_loans.shape[1]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Remove the columns with only NaN and or with large number of NaN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>col_name</th>\n",
       "      <th>num_NaN</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>debt_settlement_flag</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>revol_bal</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>initial_list_status</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>out_prncp</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>total_pymnt</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>total_pymnt_inv</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>total_rec_prncp</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>addr_state</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>total_rec_int</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>recoveries</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>collection_recovery_fee</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>last_pymnt_amnt</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>policy_code</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>application_type</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>disbursement_method</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>total_rec_late_fee</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>purpose</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>out_prncp_inv</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>funded_amnt_inv</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>verification_status</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>emp_length</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>funded_amnt</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>loan_amnt</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>sub_grade</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>grade</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>home_ownership</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>installment</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>term</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>issue_d</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>pymnt_plan</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>115</th>\n",
       "      <td>sec_app_open_acc</td>\n",
       "      <td>138286</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>116</th>\n",
       "      <td>sec_app_collections_12_mths_ex_med</td>\n",
       "      <td>138286</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>117</th>\n",
       "      <td>sec_app_open_act_il</td>\n",
       "      <td>138286</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>118</th>\n",
       "      <td>sec_app_num_rev_accts</td>\n",
       "      <td>138286</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>119</th>\n",
       "      <td>sec_app_chargeoff_within_12_mths</td>\n",
       "      <td>138286</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>120</th>\n",
       "      <td>sec_app_revol_util</td>\n",
       "      <td>138292</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>121</th>\n",
       "      <td>sec_app_mths_since_last_major_derog</td>\n",
       "      <td>138955</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>122</th>\n",
       "      <td>hardship_last_payment_amount</td>\n",
       "      <td>139024</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>123</th>\n",
       "      <td>hardship_payoff_balance_amount</td>\n",
       "      <td>139024</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>124</th>\n",
       "      <td>hardship_loan_status</td>\n",
       "      <td>139024</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>125</th>\n",
       "      <td>hardship_dpd</td>\n",
       "      <td>139024</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>126</th>\n",
       "      <td>hardship_length</td>\n",
       "      <td>139024</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>127</th>\n",
       "      <td>payment_plan_start_date</td>\n",
       "      <td>139024</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>128</th>\n",
       "      <td>hardship_end_date</td>\n",
       "      <td>139024</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>129</th>\n",
       "      <td>hardship_amount</td>\n",
       "      <td>139024</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>130</th>\n",
       "      <td>deferral_term</td>\n",
       "      <td>139024</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>131</th>\n",
       "      <td>hardship_status</td>\n",
       "      <td>139024</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>132</th>\n",
       "      <td>hardship_reason</td>\n",
       "      <td>139024</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>133</th>\n",
       "      <td>hardship_start_date</td>\n",
       "      <td>139024</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>134</th>\n",
       "      <td>hardship_type</td>\n",
       "      <td>139024</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>135</th>\n",
       "      <td>orig_projected_additional_accrued_interest</td>\n",
       "      <td>139036</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>136</th>\n",
       "      <td>settlement_status</td>\n",
       "      <td>139101</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>137</th>\n",
       "      <td>settlement_date</td>\n",
       "      <td>139101</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>138</th>\n",
       "      <td>settlement_amount</td>\n",
       "      <td>139101</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>139</th>\n",
       "      <td>settlement_term</td>\n",
       "      <td>139101</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>140</th>\n",
       "      <td>settlement_percentage</td>\n",
       "      <td>139101</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>141</th>\n",
       "      <td>debt_settlement_flag_date</td>\n",
       "      <td>139159</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>142</th>\n",
       "      <td>url</td>\n",
       "      <td>139311</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>143</th>\n",
       "      <td>member_id</td>\n",
       "      <td>139311</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>144</th>\n",
       "      <td>id</td>\n",
       "      <td>139311</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>145 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                       col_name  num_NaN\n",
       "0                          debt_settlement_flag        0\n",
       "1                                     revol_bal        0\n",
       "2                           initial_list_status        0\n",
       "3                                     out_prncp        0\n",
       "4                                   total_pymnt        0\n",
       "5                               total_pymnt_inv        0\n",
       "6                               total_rec_prncp        0\n",
       "7                                    addr_state        0\n",
       "8                                 total_rec_int        0\n",
       "9                                    recoveries        0\n",
       "10                      collection_recovery_fee        0\n",
       "11                              last_pymnt_amnt        0\n",
       "12                                  policy_code        0\n",
       "13                             application_type        0\n",
       "14                          disbursement_method        0\n",
       "15                           total_rec_late_fee        0\n",
       "16                                      purpose        0\n",
       "17                                out_prncp_inv        0\n",
       "18                              funded_amnt_inv        0\n",
       "19                          verification_status        0\n",
       "20                                   emp_length        0\n",
       "21                                  funded_amnt        0\n",
       "22                                    loan_amnt        0\n",
       "23                                    sub_grade        0\n",
       "24                                        grade        0\n",
       "25                               home_ownership        0\n",
       "26                                  installment        0\n",
       "27                                         term        0\n",
       "28                                      issue_d        0\n",
       "29                                   pymnt_plan        0\n",
       "..                                          ...      ...\n",
       "115                            sec_app_open_acc   138286\n",
       "116          sec_app_collections_12_mths_ex_med   138286\n",
       "117                         sec_app_open_act_il   138286\n",
       "118                       sec_app_num_rev_accts   138286\n",
       "119            sec_app_chargeoff_within_12_mths   138286\n",
       "120                          sec_app_revol_util   138292\n",
       "121         sec_app_mths_since_last_major_derog   138955\n",
       "122                hardship_last_payment_amount   139024\n",
       "123              hardship_payoff_balance_amount   139024\n",
       "124                        hardship_loan_status   139024\n",
       "125                                hardship_dpd   139024\n",
       "126                             hardship_length   139024\n",
       "127                     payment_plan_start_date   139024\n",
       "128                           hardship_end_date   139024\n",
       "129                             hardship_amount   139024\n",
       "130                               deferral_term   139024\n",
       "131                             hardship_status   139024\n",
       "132                             hardship_reason   139024\n",
       "133                         hardship_start_date   139024\n",
       "134                               hardship_type   139024\n",
       "135  orig_projected_additional_accrued_interest   139036\n",
       "136                           settlement_status   139101\n",
       "137                             settlement_date   139101\n",
       "138                           settlement_amount   139101\n",
       "139                             settlement_term   139101\n",
       "140                       settlement_percentage   139101\n",
       "141                   debt_settlement_flag_date   139159\n",
       "142                                         url   139311\n",
       "143                                   member_id   139311\n",
       "144                                          id   139311\n",
       "\n",
       "[145 rows x 2 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Create df with number of NaN on each column\n",
    "use_col = pd.DataFrame(data_loans.isnull().sum().sort_values())\n",
    "use_col.reset_index(drop = False, inplace=True)\n",
    "use_col.columns = (['col_name','num_NaN'])\n",
    "display(use_col)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>debt_settlement_flag</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>revol_bal</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>initial_list_status</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>out_prncp</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>total_pymnt</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>total_pymnt_inv</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>total_rec_prncp</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>addr_state</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>total_rec_int</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>recoveries</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>collection_recovery_fee</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>last_pymnt_amnt</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>policy_code</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>application_type</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>disbursement_method</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>total_rec_late_fee</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>purpose</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>out_prncp_inv</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>funded_amnt_inv</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>verification_status</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>emp_length</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>funded_amnt</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>loan_amnt</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sub_grade</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>grade</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>home_ownership</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>installment</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>term</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>issue_d</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>pymnt_plan</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>loan_status</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>hardship_flag</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>int_rate</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>zip_code</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>annual_inc</th>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>last_credit_pull_d</th>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                         0\n",
       "debt_settlement_flag     0\n",
       "revol_bal                0\n",
       "initial_list_status      0\n",
       "out_prncp                0\n",
       "total_pymnt              0\n",
       "total_pymnt_inv          0\n",
       "total_rec_prncp          0\n",
       "addr_state               0\n",
       "total_rec_int            0\n",
       "recoveries               0\n",
       "collection_recovery_fee  0\n",
       "last_pymnt_amnt          0\n",
       "policy_code              0\n",
       "application_type         0\n",
       "disbursement_method      0\n",
       "total_rec_late_fee       0\n",
       "purpose                  0\n",
       "out_prncp_inv            0\n",
       "funded_amnt_inv          0\n",
       "verification_status      0\n",
       "emp_length               0\n",
       "funded_amnt              0\n",
       "loan_amnt                0\n",
       "sub_grade                0\n",
       "grade                    0\n",
       "home_ownership           0\n",
       "installment              0\n",
       "term                     0\n",
       "issue_d                  0\n",
       "pymnt_plan               0\n",
       "loan_status              0\n",
       "hardship_flag            0\n",
       "int_rate                 0\n",
       "zip_code                 1\n",
       "annual_inc               4\n",
       "last_credit_pull_d       6"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.DataFrame(data_loans.isnull().sum().sort_values()[data_loans.isnull().sum().sort_values()<10])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(139311, 142)"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_loans = data_loans.dropna(axis=1, how='all')\n",
    "data_loans.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def df_col_low_nan(df_data, max_nan): \n",
    "    # Returns a df with columns with less nan than min_nan\n",
    "    use_col = pd.DataFrame(df_data.isnull().sum().sort_values()[df_data.isnull().sum().sort_values()<max_nan])\n",
    "    use_col.reset_index(drop = False, inplace=True)\n",
    "    use_col.columns = (['col_name','num_NaN'])\n",
    "    new_data = pd.DataFrame()\n",
    "    for i in range(0,len(use_col.col_name)):\n",
    "        df = df_data[use_col.col_name[i]]\n",
    "        new_data = pd.concat([new_data, df], axis=1)\n",
    "    return new_data\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(139311, 36)"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "n_loan_df = df_col_low_nan(data_loans, 10)\n",
    "n_loan_df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#n_loan_df.to_csv('data_files/loans_2017.csv', index=False)\n",
    "n_loan_df = pd.read_csv('data_files/loans_2017.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    139311.000000\n",
       "mean      13707.767692\n",
       "std        9006.370993\n",
       "min         500.000000\n",
       "25%        6625.000000\n",
       "50%       12000.000000\n",
       "75%       19200.000000\n",
       "max       40000.000000\n",
       "Name: loan_amnt, dtype: float64"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "n_loan_df.loan_amnt.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>loan_amnt</th>\n",
       "      <th>revol_bal</th>\n",
       "      <th>out_prncp</th>\n",
       "      <th>out_prncp_inv</th>\n",
       "      <th>total_pymnt</th>\n",
       "      <th>total_pymnt_inv</th>\n",
       "      <th>total_rec_late_fee</th>\n",
       "      <th>recoveries</th>\n",
       "      <th>collection_recovery_fee</th>\n",
       "      <th>last_pymnt_amnt</th>\n",
       "      <th>policy_code</th>\n",
       "      <th>total_rec_int</th>\n",
       "      <th>total_rec_prncp</th>\n",
       "      <th>funded_amnt</th>\n",
       "      <th>installment</th>\n",
       "      <th>funded_amnt_inv</th>\n",
       "      <th>annual_inc</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>139311.000000</td>\n",
       "      <td>1.393110e+05</td>\n",
       "      <td>139311.000000</td>\n",
       "      <td>139311.000000</td>\n",
       "      <td>139311.000000</td>\n",
       "      <td>139311.000000</td>\n",
       "      <td>1.393110e+05</td>\n",
       "      <td>139311.000000</td>\n",
       "      <td>139311.000000</td>\n",
       "      <td>139311.000000</td>\n",
       "      <td>139311.0</td>\n",
       "      <td>139311.000000</td>\n",
       "      <td>139311.000000</td>\n",
       "      <td>139311.000000</td>\n",
       "      <td>139311.000000</td>\n",
       "      <td>139311.000000</td>\n",
       "      <td>1.393070e+05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>13707.767692</td>\n",
       "      <td>1.629335e+04</td>\n",
       "      <td>7707.743732</td>\n",
       "      <td>7705.556592</td>\n",
       "      <td>6960.615220</td>\n",
       "      <td>6743.681439</td>\n",
       "      <td>7.557048e-01</td>\n",
       "      <td>32.577107</td>\n",
       "      <td>4.595364</td>\n",
       "      <td>2014.531366</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1475.227889</td>\n",
       "      <td>5452.054549</td>\n",
       "      <td>13625.899068</td>\n",
       "      <td>411.721342</td>\n",
       "      <td>13414.719546</td>\n",
       "      <td>7.860105e+04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>9006.370993</td>\n",
       "      <td>2.296160e+04</td>\n",
       "      <td>8874.139235</td>\n",
       "      <td>8871.700828</td>\n",
       "      <td>7509.645493</td>\n",
       "      <td>7348.244653</td>\n",
       "      <td>5.913842e+00</td>\n",
       "      <td>412.842205</td>\n",
       "      <td>90.380138</td>\n",
       "      <td>4732.246409</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1722.640108</td>\n",
       "      <td>6447.174433</td>\n",
       "      <td>8965.774515</td>\n",
       "      <td>269.530222</td>\n",
       "      <td>9029.675138</td>\n",
       "      <td>1.827296e+05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>500.000000</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-4.000000e-10</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>500.000000</td>\n",
       "      <td>15.670000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000e+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>6625.000000</td>\n",
       "      <td>5.537000e+03</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>2298.170000</td>\n",
       "      <td>2234.330000</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>248.820000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>473.835000</td>\n",
       "      <td>1458.080000</td>\n",
       "      <td>6550.000000</td>\n",
       "      <td>211.120000</td>\n",
       "      <td>6225.000000</td>\n",
       "      <td>4.550000e+04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>12000.000000</td>\n",
       "      <td>1.088300e+04</td>\n",
       "      <td>4946.060000</td>\n",
       "      <td>4946.050000</td>\n",
       "      <td>4220.280000</td>\n",
       "      <td>4101.150000</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>437.630000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>966.230000</td>\n",
       "      <td>2800.000000</td>\n",
       "      <td>11975.000000</td>\n",
       "      <td>341.230000</td>\n",
       "      <td>11200.000000</td>\n",
       "      <td>6.500000e+04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>19200.000000</td>\n",
       "      <td>1.965600e+04</td>\n",
       "      <td>12638.900000</td>\n",
       "      <td>12629.485000</td>\n",
       "      <td>8413.577341</td>\n",
       "      <td>8095.720000</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>903.410000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1843.830000</td>\n",
       "      <td>6670.550000</td>\n",
       "      <td>19000.000000</td>\n",
       "      <td>544.170000</td>\n",
       "      <td>18250.000000</td>\n",
       "      <td>9.400000e+04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>40000.000000</td>\n",
       "      <td>1.698749e+06</td>\n",
       "      <td>38699.080000</td>\n",
       "      <td>38699.080000</td>\n",
       "      <td>58886.473429</td>\n",
       "      <td>58563.680000</td>\n",
       "      <td>3.119000e+02</td>\n",
       "      <td>29623.350000</td>\n",
       "      <td>7002.190000</td>\n",
       "      <td>41453.070000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>23886.470000</td>\n",
       "      <td>40000.000000</td>\n",
       "      <td>40000.000000</td>\n",
       "      <td>1715.420000</td>\n",
       "      <td>40000.000000</td>\n",
       "      <td>6.100000e+07</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           loan_amnt     revol_bal      out_prncp  out_prncp_inv  \\\n",
       "count  139311.000000  1.393110e+05  139311.000000  139311.000000   \n",
       "mean    13707.767692  1.629335e+04    7707.743732    7705.556592   \n",
       "std      9006.370993  2.296160e+04    8874.139235    8871.700828   \n",
       "min       500.000000  0.000000e+00       0.000000       0.000000   \n",
       "25%      6625.000000  5.537000e+03       0.000000       0.000000   \n",
       "50%     12000.000000  1.088300e+04    4946.060000    4946.050000   \n",
       "75%     19200.000000  1.965600e+04   12638.900000   12629.485000   \n",
       "max     40000.000000  1.698749e+06   38699.080000   38699.080000   \n",
       "\n",
       "         total_pymnt  total_pymnt_inv  total_rec_late_fee     recoveries  \\\n",
       "count  139311.000000    139311.000000        1.393110e+05  139311.000000   \n",
       "mean     6960.615220      6743.681439        7.557048e-01      32.577107   \n",
       "std      7509.645493      7348.244653        5.913842e+00     412.842205   \n",
       "min         0.000000         0.000000       -4.000000e-10       0.000000   \n",
       "25%      2298.170000      2234.330000        0.000000e+00       0.000000   \n",
       "50%      4220.280000      4101.150000        0.000000e+00       0.000000   \n",
       "75%      8413.577341      8095.720000        0.000000e+00       0.000000   \n",
       "max     58886.473429     58563.680000        3.119000e+02   29623.350000   \n",
       "\n",
       "       collection_recovery_fee  last_pymnt_amnt  policy_code  total_rec_int  \\\n",
       "count            139311.000000    139311.000000     139311.0  139311.000000   \n",
       "mean                  4.595364      2014.531366          1.0    1475.227889   \n",
       "std                  90.380138      4732.246409          0.0    1722.640108   \n",
       "min                   0.000000         0.000000          1.0       0.000000   \n",
       "25%                   0.000000       248.820000          1.0     473.835000   \n",
       "50%                   0.000000       437.630000          1.0     966.230000   \n",
       "75%                   0.000000       903.410000          1.0    1843.830000   \n",
       "max                7002.190000     41453.070000          1.0   23886.470000   \n",
       "\n",
       "       total_rec_prncp    funded_amnt    installment  funded_amnt_inv  \\\n",
       "count    139311.000000  139311.000000  139311.000000    139311.000000   \n",
       "mean       5452.054549   13625.899068     411.721342     13414.719546   \n",
       "std        6447.174433    8965.774515     269.530222      9029.675138   \n",
       "min           0.000000     500.000000      15.670000         0.000000   \n",
       "25%        1458.080000    6550.000000     211.120000      6225.000000   \n",
       "50%        2800.000000   11975.000000     341.230000     11200.000000   \n",
       "75%        6670.550000   19000.000000     544.170000     18250.000000   \n",
       "max       40000.000000   40000.000000    1715.420000     40000.000000   \n",
       "\n",
       "         annual_inc  \n",
       "count  1.393070e+05  \n",
       "mean   7.860105e+04  \n",
       "std    1.827296e+05  \n",
       "min    0.000000e+00  \n",
       "25%    4.550000e+04  \n",
       "50%    6.500000e+04  \n",
       "75%    9.400000e+04  \n",
       "max    6.100000e+07  "
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "n_loan_df.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Feature cleaning and selection\n",
    "\n",
    "### Target feature (loan_status)\n",
    "\n",
    "For the purpose of this project, problem loans are loans that have not been paid, and that have a been classified as:\n",
    "\n",
    "Late, Grace period, Default, or Charged Off"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['Current', 'Fully Paid', 'Late (31-120 days)', 'Charged Off',\n",
       "       'In Grace Period', 'Late (16-30 days)', 'Default',\n",
       "       'Does not meet the credit policy. Status:Fully Paid',\n",
       "       'Does not meet the credit policy. Status:Charged Off'], dtype=object)"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "n_loan_df.loan_status.unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['Fully Paid', 'Problem Loan'], dtype=object)"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Removing fully paid loans\n",
    "n_loan_df = n_loan_df[n_loan_df.loan_status.str.contains('Current')==False]\n",
    "\n",
    "# Collapsing all but Fully Paid into Problem Loan\n",
    "n_loan_df.loan_status = n_loan_df.loan_status.str.replace(r'(^.*Paid$)', 'Fully Paid')\n",
    "n_loan_df.loan_status = n_loan_df.loan_status.str.replace(r'(^.*Charged Off$)', 'Charged Off')\n",
    "n_loan_df.loan_status = n_loan_df.loan_status.str.replace(r'(^Late.*$)', 'Late')\n",
    "n_loan_df.loan_status = n_loan_df.loan_status.str.replace(r'(Late)|(Charged Off)|(In Grace Period)|(Default)', \n",
    "                                                          'Problem Loan')\n",
    "n_loan_df.loan_status.unique()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(58006, 36)"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "n_loan_df.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Other features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Custom functiones needed later\n",
    "def types_of_columns(df):\n",
    "    # creates two separate dataframes \n",
    "    # one with float column names and the other with object column names\n",
    "    non_num_cols = df.dtypes[df.dtypes == 'object']\n",
    "    float_cols   = df.dtypes[df.dtypes != 'object']\n",
    "    return non_num_cols, float_cols\n",
    "\n",
    "def column_checker (df1, df2, num):\n",
    "    # df1 dataframe whit the columns\n",
    "    # df2 dataframe with colum names\n",
    "    for col in range(0,len(df2)):\n",
    "        col = df2.index[col]\n",
    "        values = df1[col].unique()\n",
    "        if len(values) > num:\n",
    "           print('\\n', col, '\\n',len(values),'\\n',values)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Sorting the columns by type"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "col_types_original = types_of_columns(n_loan_df)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Non-Numeric Features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "hardship_flag           object\n",
       "initial_list_status     object\n",
       "disbursement_method     object\n",
       "application_type        object\n",
       "addr_state              object\n",
       "verification_status     object\n",
       "term                    object\n",
       "int_rate                object\n",
       "grade                   object\n",
       "sub_grade               object\n",
       "emp_length              object\n",
       "home_ownership          object\n",
       "purpose                 object\n",
       "issue_d                 object\n",
       "loan_status             object\n",
       "pymnt_plan              object\n",
       "debt_settlement_flag    object\n",
       "zip_code                object\n",
       "last_credit_pull_d      object\n",
       "dtype: object"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nn_col_list = col_types_original[0]\n",
    "(nn_col_list)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      " hardship_flag \n",
      " 2 \n",
      " ['N' 'Y']\n",
      "\n",
      " initial_list_status \n",
      " 2 \n",
      " ['w' 'f']\n",
      "\n",
      " disbursement_method \n",
      " 2 \n",
      " ['Cash' 'DirectPay']\n",
      "\n",
      " application_type \n",
      " 2 \n",
      " ['Individual' 'Joint App']\n",
      "\n",
      " addr_state \n",
      " 51 \n",
      " ['UT' 'VA' 'CA' 'NJ' 'PA' 'CT' 'FL' 'WA' 'AZ' 'MA' 'CO' 'MD' 'WI' 'OR' 'MO'\n",
      " 'NY' 'TX' 'IL' 'MN' 'VT' 'AL' 'OK' 'NE' 'MS' 'OH' 'IN' 'NH' 'SC' 'DC' 'GA'\n",
      " 'LA' 'TN' 'NC' 'MI' 'HI' 'AR' 'WY' 'NV' 'ID' 'DE' 'RI' 'MT' 'KY' 'AK' 'NM'\n",
      " 'SD' 'ND' 'KS' 'ME' 'WV' 'IA']\n",
      "\n",
      " verification_status \n",
      " 3 \n",
      " ['Not Verified' 'Source Verified' 'Verified']\n",
      "\n",
      " term \n",
      " 2 \n",
      " [' 36 months' ' 60 months']\n",
      "\n",
      " int_rate \n",
      " 420 \n",
      " ['  7.49%' ' 25.49%' ' 14.99%' ' 11.49%' ' 16.99%' ' 28.69%' ' 13.99%'\n",
      " ' 15.99%' ' 11.44%' ' 19.99%' ' 12.74%' ' 13.49%' '  5.32%' ' 18.99%'\n",
      " '  8.24%' ' 23.99%' '  7.99%' ' 22.74%' '  7.24%' ' 29.99%' ' 11.39%'\n",
      " ' 17.99%' ' 10.49%' ' 30.89%' '  6.99%' ' 29.49%' ' 30.79%' ' 24.74%'\n",
      " ' 21.49%' ' 26.24%' ' 30.74%' ' 30.49%' ' 30.84%' ' 30.94%' ' 30.99%'\n",
      " '  6.00%' ' 10.65%' ' 15.27%' ' 15.96%' ' 12.69%' '  7.90%' ' 18.64%'\n",
      " ' 21.28%' ' 14.65%' '  9.91%' ' 16.29%' '  6.03%' ' 11.71%' ' 12.42%'\n",
      " ' 14.27%' ' 16.77%' '  7.51%' '  8.90%' ' 18.25%' '  6.62%' ' 19.91%'\n",
      " ' 17.27%' ' 17.58%' ' 21.67%' ' 19.42%' ' 22.06%' ' 20.89%' ' 20.30%'\n",
      " ' 23.91%' ' 19.03%' ' 23.52%' ' 23.13%' ' 22.35%' ' 24.11%' ' 22.11%'\n",
      " ' 11.99%' '  5.99%' ' 10.99%' '  9.99%' ' 18.79%' '  8.49%' ' 16.49%'\n",
      " ' 12.99%' ' 15.23%' ' 14.79%' '  5.42%' ' 10.59%' ' 17.49%' ' 15.62%'\n",
      " ' 21.36%' ' 19.29%' ' 18.39%' ' 16.89%' ' 20.62%' ' 20.99%' ' 22.85%'\n",
      " ' 19.69%' ' 20.25%' ' 23.22%' ' 21.74%' ' 22.48%' ' 23.59%' ' 12.62%'\n",
      " ' 18.07%' ' 11.63%' '  7.91%' '  7.42%' ' 11.14%' ' 20.20%' ' 12.12%'\n",
      " ' 19.39%' ' 16.11%' ' 17.54%' ' 22.64%' ' 13.84%' ' 16.59%' ' 17.19%'\n",
      " ' 12.87%' ' 20.69%' '  9.67%' ' 21.82%' ' 19.79%' ' 18.49%' ' 22.94%'\n",
      " ' 24.59%' ' 24.40%' ' 21.48%' ' 14.82%' ' 14.17%' '  7.29%' ' 17.88%'\n",
      " ' 20.11%' ' 16.02%' ' 17.51%' ' 13.43%' ' 14.91%' ' 13.06%' ' 15.28%'\n",
      " ' 15.65%' ' 17.14%' ' 11.11%' ' 10.37%' ' 16.40%' '  7.66%' ' 10.00%'\n",
      " ' 18.62%' ' 10.74%' '  5.79%' '  6.92%' '  9.63%' ' 14.54%' ' 12.68%'\n",
      " ' 19.36%' ' 13.80%' ' 21.59%' ' 20.85%' ' 21.22%' ' 19.74%' ' 20.48%'\n",
      " '  6.91%' ' 12.23%' ' 12.61%' ' 10.36%' '  6.17%' '  6.54%' '  9.25%'\n",
      " ' 16.69%' ' 15.95%' '  8.88%' ' 13.35%' '  9.62%' ' 16.32%' ' 12.98%'\n",
      " ' 14.83%' ' 13.72%' ' 14.09%' ' 14.46%' ' 20.03%' ' 17.80%' ' 15.20%'\n",
      " ' 15.57%' ' 18.54%' ' 19.66%' ' 17.06%' ' 18.17%' ' 17.43%' ' 20.40%'\n",
      " ' 20.77%' ' 18.91%' ' 21.14%' ' 17.44%' ' 13.23%' '  7.88%' ' 11.12%'\n",
      " ' 13.61%' ' 10.38%' ' 17.56%' ' 17.93%' ' 15.58%' ' 13.98%' ' 14.84%'\n",
      " ' 15.21%' '  6.76%' '  6.39%' ' 11.86%' '  7.14%' ' 14.35%' ' 16.82%'\n",
      " ' 10.75%' ' 14.72%' ' 16.45%' ' 18.67%' ' 20.53%' ' 19.41%' ' 20.16%'\n",
      " ' 21.27%' ' 18.30%' ' 19.04%' ' 20.90%' ' 21.64%' ' 12.73%' ' 10.25%'\n",
      " ' 13.11%' ' 10.62%' ' 13.48%' ' 14.59%' ' 16.07%' ' 15.70%' '  9.88%'\n",
      " ' 11.36%' ' 15.33%' ' 13.85%' ' 14.96%' ' 14.22%' '  7.74%' ' 13.22%'\n",
      " ' 13.57%' '  8.59%' ' 17.04%' ' 14.61%' '  8.94%' ' 12.18%' ' 11.83%'\n",
      " ' 11.48%' ' 16.35%' ' 13.92%' ' 15.31%' ' 14.26%' ' 19.13%' ' 12.53%'\n",
      " ' 16.70%' ' 16.00%' ' 17.39%' ' 18.09%' '  7.40%' ' 18.43%' ' 17.74%'\n",
      " '  7.05%' ' 20.52%' ' 20.86%' ' 19.47%' ' 18.78%' ' 21.21%' ' 19.82%'\n",
      " ' 20.17%' ' 13.16%' '  8.00%' ' 13.47%' ' 12.21%' ' 16.63%' '  9.32%'\n",
      " ' 12.84%' ' 11.26%' ' 15.68%' ' 15.37%' ' 10.95%' ' 11.89%' ' 14.11%'\n",
      " ' 13.79%' '  7.68%' ' 11.58%' '  7.37%' ' 16.95%' ' 15.05%' ' 18.53%'\n",
      " ' 14.74%' ' 14.42%' ' 18.21%' ' 17.26%' ' 18.84%' ' 17.90%' ' 19.16%'\n",
      " ' 13.67%' '  9.38%' ' 12.72%' ' 13.36%' ' 11.46%' ' 10.51%' '  9.07%'\n",
      " ' 13.04%' ' 11.78%' ' 12.41%' ' 10.83%' ' 12.09%' ' 17.46%' ' 14.30%'\n",
      " ' 17.15%' ' 15.25%' ' 10.20%' ' 15.88%' ' 14.93%' ' 16.20%' ' 18.72%'\n",
      " ' 14.62%' '  8.32%' ' 14.12%' ' 10.96%' ' 10.33%' ' 10.01%' ' 12.86%'\n",
      " ' 11.28%' ' 11.59%' '  8.63%' ' 12.54%' ' 12.22%' ' 11.91%' ' 15.38%'\n",
      " ' 16.96%' ' 13.17%' '  9.70%' ' 16.33%' ' 14.75%' ' 15.07%' ' 16.01%'\n",
      " ' 10.71%' ' 10.64%' '  9.76%' ' 11.34%' ' 10.39%' ' 13.87%' ' 11.03%'\n",
      " ' 11.66%' ' 13.24%' ' 10.08%' '  9.45%' ' 13.55%' ' 12.29%' ' 11.97%'\n",
      " ' 12.92%' ' 15.45%' ' 14.50%' ' 14.18%' ' 15.13%' ' 16.08%' ' 15.76%'\n",
      " ' 17.03%' ' 17.34%' ' 16.71%' '  9.83%' ' 13.62%' ' 10.46%' '  9.51%'\n",
      " '  9.20%' ' 13.30%' ' 10.78%' '  7.75%' '  8.38%' ' 12.36%' ' 12.67%'\n",
      " ' 11.72%' ' 13.93%' '  8.07%' '  7.43%' ' 12.04%' ' 14.25%' ' 14.88%'\n",
      " ' 11.41%' ' 11.09%' ' 10.14%' ' 16.15%' ' 15.83%' '  7.12%' ' 18.36%'\n",
      " '  9.64%' '  9.96%' ' 11.22%' '  9.01%' '  9.33%' ' 11.54%' ' 12.17%'\n",
      " ' 12.80%' ' 14.38%' ' 13.75%' ' 14.70%' ' 12.49%' ' 14.07%' ' 10.91%'\n",
      " ' 13.12%' ' 10.28%' '  8.70%' ' 14.67%' ' 15.01%' ' 17.78%' ' 16.83%'\n",
      " ' 17.59%' ' 14.43%' ' 16.65%' ' 17.91%' ' 17.28%' ' 18.86%' ' 18.61%'\n",
      " ' 17.66%' ' 18.29%' ' 17.97%' ' 18.04%' ' 14.57%' ' 17.72%' ' 17.09%'\n",
      " ' 15.51%' ' 16.46%' ' 17.41%' ' 17.22%' ' 16.91%' ' 16.28%' ' 17.86%']\n",
      "\n",
      " grade \n",
      " 7 \n",
      " ['A' 'E' 'C' 'B' 'D' 'F' 'G']\n",
      "\n",
      " sub_grade \n",
      " 35 \n",
      " ['A4' 'E4' 'C4' 'B5' 'D1' 'F1' 'C3' 'C5' 'B4' 'D4' 'C1' 'C2' 'A1' 'D3' 'B1'\n",
      " 'E2' 'A5' 'E1' 'A3' 'F3' 'B3' 'D2' 'B2' 'G3' 'A2' 'F2' 'G1' 'E3' 'D5' 'E5'\n",
      " 'F5' 'F4' 'G2' 'G4' 'G5']\n",
      "\n",
      " emp_length \n",
      " 12 \n",
      " ['10+ years' '< 1 year' '4 years' '9 years' '5 years' '3 years' '1 year'\n",
      " '2 years' 'n/a' '6 years' '7 years' '8 years']\n",
      "\n",
      " home_ownership \n",
      " 6 \n",
      " ['MORTGAGE' 'RENT' 'OWN' 'ANY' 'OTHER' 'NONE']\n",
      "\n",
      " purpose \n",
      " 14 \n",
      " ['debt_consolidation' 'home_improvement' 'other' 'major_purchase'\n",
      " 'credit_card' 'medical' 'moving' 'car' 'vacation' 'small_business'\n",
      " 'renewable_energy' 'house' 'wedding' 'educational']\n",
      "\n",
      " issue_d \n",
      " 58 \n",
      " ['Mar-2017' 'Feb-2017' 'Jan-2017' 'Dec-2011' 'Nov-2011' 'Oct-2011'\n",
      " 'Sep-2011' 'Aug-2011' 'Jul-2011' 'Jun-2011' 'May-2011' 'Apr-2011'\n",
      " 'Mar-2011' 'Feb-2011' 'Jan-2011' 'Dec-2010' 'Nov-2010' 'Oct-2010'\n",
      " 'Sep-2010' 'Aug-2010' 'Jul-2010' 'Jun-2010' 'May-2010' 'Apr-2010'\n",
      " 'Mar-2010' 'Feb-2010' 'Jan-2010' 'Dec-2009' 'Nov-2009' 'Oct-2009'\n",
      " 'Sep-2009' 'Aug-2009' 'Jul-2009' 'Jun-2009' 'May-2009' 'Apr-2009'\n",
      " 'Mar-2009' 'Feb-2009' 'Jan-2009' 'Dec-2008' 'Nov-2008' 'Oct-2008'\n",
      " 'Sep-2008' 'Aug-2008' 'Jul-2008' 'Jun-2008' 'May-2008' 'Apr-2008'\n",
      " 'Mar-2008' 'Feb-2008' 'Jan-2008' 'Dec-2007' 'Nov-2007' 'Oct-2007'\n",
      " 'Sep-2007' 'Aug-2007' 'Jul-2007' 'Jun-2007']\n",
      "\n",
      " loan_status \n",
      " 2 \n",
      " ['Fully Paid' 'Problem Loan']\n",
      "\n",
      " pymnt_plan \n",
      " 2 \n",
      " ['n' 'y']\n",
      "\n",
      " debt_settlement_flag \n",
      " 2 \n",
      " ['N' 'Y']\n",
      "\n",
      " zip_code \n",
      " 877 \n",
      " ['840xx' '224xx' '945xx' '087xx' '194xx' '067xx' '320xx' '908xx' '982xx'\n",
      " '851xx' '027xx' '070xx' '802xx' '210xx' '537xx' '926xx' '900xx' '160xx'\n",
      " '971xx' '185xx' '993xx' '928xx' '841xx' '155xx' '630xx' '103xx' '785xx'\n",
      " '237xx' '927xx' '606xx' '930xx' '902xx' '026xx' '206xx' '207xx' '331xx'\n",
      " '551xx' '347xx' '906xx' '560xx' '220xx' '925xx' '330xx' '240xx' '950xx'\n",
      " '054xx' '338xx' '850xx' '339xx' '132xx' '805xx' '064xx' '361xx' '760xx'\n",
      " '744xx' '934xx' '687xx' '142xx' '359xx' '770xx' '390xx' '933xx' '973xx'\n",
      " '358xx' '082xx' '057xx' '088xx' '431xx' '941xx' '604xx' '023xx' '395xx'\n",
      " '787xx' '108xx' '655xx' '774xx' '161xx' '441xx' '460xx' '229xx' '346xx'\n",
      " '037xx' '917xx' '731xx' '019xx' '334xx' '182xx' '920xx' '295xx' '191xx'\n",
      " '342xx' '296xx' '158xx' '532xx' '119xx' '180xx' '473xx' '076xx' '021xx'\n",
      " '200xx' '775xx' '117xx' '060xx' '946xx' '146xx' '922xx' '750xx' '984xx'\n",
      " '128xx' '300xx' '708xx' '761xx' '380xx' '372xx' '112xx' '153xx' '780xx'\n",
      " '622xx' '433xx' '293xx' '228xx' '915xx' '952xx' '280xx' '484xx' '144xx'\n",
      " '531xx' '292xx' '765xx' '554xx' '499xx' '242xx' '078xx' '856xx' '104xx'\n",
      " '152xx' '809xx' '921xx' '272xx' '307xx' '968xx' '233xx' '722xx' '172xx'\n",
      " '462xx' '171xx' '740xx' '451xx' '786xx' '956xx' '949xx' '480xx' '379xx'\n",
      " '030xx' '782xx' '328xx' '706xx' '846xx' '944xx' '542xx' '221xx' '066xx'\n",
      " '827xx' '680xx' '275xx' '890xx' '294xx' '302xx' '270xx' '371xx' '466xx'\n",
      " '327xx' '118xx' '313xx' '333xx' '212xx' '799xx' '100xx' '907xx' '853xx'\n",
      " '344xx' '111xx' '120xx' '801xx' '478xx' '444xx' '238xx' '967xx' '704xx'\n",
      " '605xx' '832xx' '301xx' '793xx' '834xx' '486xx' '284xx' '107xx' '534xx'\n",
      " '197xx' '913xx' '741xx' '545xx' '985xx' '857xx' '773xx' '535xx' '641xx'\n",
      " '972xx' '940xx' '029xx' '349xx' '701xx' '705xx' '970xx' '378xx' '071xx'\n",
      " '613xx' '384xx' '454xx' '550xx' '743xx' '010xx' '150xx' '217xx' '559xx'\n",
      " '277xx' '911xx' '598xx' '410xx' '724xx' '958xx' '170xx' '986xx' '891xx'\n",
      " '988xx' '336xx' '557xx' '274xx' '190xx' '703xx' '218xx' '553xx' '208xx'\n",
      " '282xx' '756xx' '980xx' '236xx' '481xx' '472xx' '978xx' '737xx' '085xx'\n",
      " '080xx' '329xx' '020xx' '225xx' '914xx' '563xx' '341xx' '981xx' '751xx'\n",
      " '959xx' '400xx' '377xx' '736xx' '995xx' '456xx' '601xx' '335xx' '209xx'\n",
      " '960xx' '864xx' '110xx' '297xx' '403xx' '442xx' '562xx' '322xx' '350xx'\n",
      " '360xx' '955xx' '852xx' '847xx' '610xx' '996xx' '764xx' '368xx' '766xx'\n",
      " '325xx' '837xx' '115xx' '308xx' '752xx' '387xx' '337xx' '435xx' '288xx'\n",
      " '870xx' '074xx' '109xx' '720xx' '953xx' '201xx' '570xx' '402xx' '707xx'\n",
      " '303xx' '483xx' '136xx' '405xx' '728xx' '582xx' '935xx' '426xx' '018xx'\n",
      " '778xx' '989xx' '895xx' '600xx' '633xx' '495xx' '597xx' '123xx' '620xx'\n",
      " '489xx' '156xx' '450xx' '114xx' '961xx' '354xx' '396xx' '543xx' '496xx'\n",
      " '186xx' '546xx' '374xx' '077xx' '475xx' '245xx' '468xx' '321xx' '608xx'\n",
      " '810xx' '800xx' '135xx' '951xx' '028xx' '806xx' '148xx' '141xx' '386xx'\n",
      " '816xx' '271xx' '488xx' '650xx' '919xx' '366xx' '494xx' '640xx' '660xx'\n",
      " '617xx' '234xx' '239xx' '781xx' '461xx' '024xx' '214xx' '474xx' '860xx'\n",
      " '211xx' '721xx' '910xx' '658xx' '145xx' '626xx' '376xx' '223xx' '465xx'\n",
      " '446xx' '174xx' '983xx' '627xx' '711xx' '644xx' '616xx' '530xx' '547xx'\n",
      " '038xx' '990xx' '382xx' '283xx' '316xx' '184xx' '081xx' '544xx' '140xx'\n",
      " '373xx' '712xx' '113xx' '175xx' '232xx' '762xx' '305xx' '440xx' '625xx'\n",
      " '932xx' '281xx' '791xx' '634xx' '490xx' '923xx' '381xx' '216xx' '807xx'\n",
      " '804xx' '432xx' '453xx' '672xx' '193xx' '065xx' '585xx' '306xx' '662xx'\n",
      " '273xx' '784xx' '467xx' '904xx' '898xx' '783xx' '844xx' '727xx' '903xx'\n",
      " '829xx' '125xx' '469xx' '936xx' '581xx' '954xx' '173xx' '413xx' '714xx'\n",
      " '710xx' '034xx' '631xx' '730xx' '352xx' '974xx' '471xx' '957xx' '231xx'\n",
      " '470xx' '948xx' '685xx' '754xx' '069xx' '285xx' '241xx' '012xx' '063xx'\n",
      " '017xx' '089xx' '151xx' '227xx' '482xx' '276xx' '226xx' '607xx' '924xx'\n",
      " '808xx' '549xx' '768xx' '828xx' '015xx' '357xx' '286xx' '916xx' '310xx'\n",
      " '244xx' '370xx' '299xx' '363xx' '291xx' '874xx' '619xx' '195xx' '287xx'\n",
      " '939xx' '912xx' '121xx' '668xx' '599xx' '845xx' '235xx' '681xx' '700xx'\n",
      " '492xx' '670xx' '314xx' '882xx' '779xx' '351xx' '587xx' '062xx' '164xx'\n",
      " '797xx' '189xx' '574xx' '105xx' '324xx' '139xx' '937xx' '931xx' '871xx'\n",
      " '458xx' '776xx' '652xx' '116xx' '657xx' '393xx' '362xx' '143xx' '014xx'\n",
      " '443xx' '383xx' '463xx' '609xx' '075xx' '863xx' '583xx' '843xx' '279xx'\n",
      " '588xx' '693xx' '196xx' '430xx' '147xx' '875xx' '678xx' '072xx' '083xx'\n",
      " '448xx' '315xx' '615xx' '392xx' '977xx' '177xx' '457xx' '394xx' '124xx'\n",
      " '423xx' '084xx' '636xx' '427xx' '183xx' '391xx' '129xx' '667xx' '476xx'\n",
      " '061xx' '745xx' '159xx' '757xx' '584xx' '835xx' '397xx' '438xx' '309xx'\n",
      " '566xx' '168xx' '040xx' '831xx' '068xx' '045xx' '769xx' '187xx' '219xx'\n",
      " '388xx' '975xx' '016xx' '222xx' '319xx' '452xx' '290xx' '166xx' '612xx'\n",
      " '747xx' '992xx' '894xx' '137xx' '041xx' '590xx' '767xx' '716xx' '836xx'\n",
      " '815xx' '127xx' '881xx' '723xx' '594xx' '198xx' '487xx' '573xx' '812xx'\n",
      " '877xx' '539xx' '199xx' '013xx' '437xx' '122xx' '149xx' '897xx' '905xx'\n",
      " '880xx' '820xx' '079xx' '317xx' '571xx' '794xx' '278xx' '011xx' '684xx'\n",
      " '479xx' '047xx' '565xx' '725xx' '365xx' '618xx' '298xx' '638xx' '032xx'\n",
      " '726xx' '243xx' '746xx' '580xx' '790xx' '039xx' '436xx' '126xx' '053xx'\n",
      " '645xx' '629xx' '548xx' '656xx' '718xx' '614xx' '811xx' '073xx' '795xx'\n",
      " '401xx' '447xx' '558xx' '130xx' '162xx' '364xx' '230xx' '654xx' '666xx'\n",
      " '102xx' '326xx' '050xx' '493xx' '883xx' '407xx' '624xx' '997xx' '404xx'\n",
      " '755xx' '134xx' '653xx' '792xx' '591xx' '753xx' '311xx' '813xx' '133xx'\n",
      " '101xx' '411xx' '398xx' '689xx' '674xx' '713xx' '323xx' '188xx' '838xx'\n",
      " '025xx' '671xx' '178xx' '586xx' '647xx' '154xx' '176xx' '540xx' '086xx'\n",
      " '603xx' '181xx' '056xx' '878xx' '356xx' '421xx' '477xx' '719xx' '031xx'\n",
      " '788xx' '304xx' '048xx' '415xx' '823xx' '623xx' '688xx' '464xx' '686xx'\n",
      " '639xx' '637xx' '593xx' '385xx' '318xx' '855xx' '833xx' '179xx' '729xx'\n",
      " '138xx' '131xx' '661xx' '947xx' '564xx' '561xx' '044xx' '049xx' '169xx'\n",
      " '611xx' '859xx' '772xx' '497xx' '648xx' '165xx' '572xx' '763xx' '157xx'\n",
      " '664xx' '439xx' '777xx' '592xx' '577xx' '748xx' '422xx' '918xx' '058xx'\n",
      " '455xx' '826xx' '491xx' '758xx' '798xx' '420xx' '673xx' '814xx' '635xx'\n",
      " '106xx' '215xx' '051xx' '628xx' '043xx' '825xx' '665xx' '355xx' '434xx'\n",
      " '976xx' '789xx' '035xx' '796xx' '879xx' '498xx' '669xx' '679xx' '541xx'\n",
      " '646xx' '994xx' '735xx' '683xx' '651xx' '167xx' '424xx' '602xx' '246xx'\n",
      " '445xx' '991xx' '289xx' '803xx' '759xx' '538xx' '258xx' '042xx' '417xx'\n",
      " '033xx' '575xx' '717xx' '734xx' '893xx' '163xx' '022xx' '596xx' '263xx'\n",
      " '312xx' '265xx' '260xx' '830xx' '969xx' '739xx' '261xx' '262xx' '367xx'\n",
      " '052xx' '416xx' '998xx' '254xx' '576xx' '963xx' '749xx' '255xx' '663xx'\n",
      " '943xx' '094xx' '253xx' '528xx' '250xx' '884xx' '675xx' '256xx' '822xx'\n",
      " '449xx' '527xx' '257xx' '247xx' '203xx' '567xx' '999xx' '268xx' '865xx'\n",
      " '406xx' '676xx' '267xx' '595xx' '259xx' '264xx' '409xx' '375xx' '412xx'\n",
      " '425xx' '369xx' '266xx' '096xx' '251xx' '408xx' '824xx' '249xx' '485xx'\n",
      " '252xx' '556xx' '353xx' '677xx' '090xx' '007xx' '332xx' '093xx' '248xx'\n",
      " '414xx' '965xx' '522xx' '523xx' '738xx' '418xx' '204xx' '059xx' '514xx'\n",
      " '500xx' '503xx' '691xx' '036xx']\n",
      "\n",
      " last_credit_pull_d \n",
      " 125 \n",
      " ['Oct-2017' 'Jun-2017' 'Sep-2017' 'May-2017' 'Jul-2017' 'Aug-2017'\n",
      " 'Mar-2017' 'Apr-2017' 'Feb-2017' 'Jan-2017' 'Dec-2016' 'Oct-2016'\n",
      " 'Apr-2016' 'Sep-2016' 'Dec-2014' 'Aug-2013' 'May-2014' 'Jul-2015'\n",
      " 'Feb-2016' 'Jun-2016' 'Mar-2014' 'Sep-2012' 'Jan-2015' 'Jan-2016'\n",
      " 'Dec-2015' 'Jun-2012' 'Mar-2016' 'Mar-2015' 'Sep-2014' 'Apr-2014'\n",
      " 'Oct-2014' 'Feb-2014' 'Feb-2013' 'Nov-2015' 'Jul-2014' 'Dec-2012'\n",
      " 'Nov-2016' 'Jul-2016' 'Jun-2015' 'Mar-2013' 'Feb-2012' 'Nov-2013'\n",
      " 'Oct-2015' 'May-2016' 'Apr-2012' 'Sep-2015' 'May-2013' 'Aug-2015'\n",
      " 'Jun-2014' 'Aug-2012' 'Sep-2013' 'Jan-2012' 'Aug-2014' 'Oct-2013'\n",
      " 'Aug-2016' 'May-2012' 'Jul-2012' 'Jan-2014' 'Oct-2012' 'Jun-2013'\n",
      " 'Dec-2013' 'Nov-2014' 'Apr-2015' 'Apr-2013' 'Nov-2012' 'May-2015'\n",
      " 'Jul-2013' 'Feb-2015' 'Mar-2012' 'Nov-2011' 'Dec-2011' 'Jan-2013'\n",
      " 'Oct-2011' 'Sep-2011' 'Aug-2011' 'Jul-2011' 'Jun-2011' 'May-2011'\n",
      " 'Apr-2011' 'Mar-2011' 'Feb-2011' 'Jan-2011' 'Dec-2010' 'Nov-2010'\n",
      " 'Oct-2010' nan 'Sep-2010' 'Aug-2010' 'Jul-2010' 'Jun-2010' 'May-2010'\n",
      " 'Apr-2010' 'Mar-2010' 'Feb-2010' 'Aug-2007' 'Jan-2010' 'Dec-2009'\n",
      " 'Nov-2009' 'Oct-2009' 'Sep-2009' 'Jul-2009' 'Aug-2009' 'May-2009'\n",
      " 'Jun-2009' 'Apr-2009' 'Mar-2009' 'Feb-2009' 'Jan-2009' 'Dec-2008'\n",
      " 'Jun-2008' 'Sep-2008' 'May-2008' 'Aug-2008' 'Mar-2008' 'Oct-2008'\n",
      " 'Feb-2008' 'Jan-2008' 'Dec-2007' 'Jul-2008' 'Oct-2007' 'Sep-2007'\n",
      " 'Jun-2007' 'May-2007' 'Jul-2007' 'Nov-2007']\n"
     ]
    }
   ],
   "source": [
    "column_checker(n_loan_df, types_of_columns(n_loan_df)[0], 0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Dropping columns with repetitive or non-relevant information\n",
    "\n",
    "zip_code and addr_state have similar information - keeping addr_state\n",
    "\n",
    "grade and sub_grade also have similar information - keeping grade"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Dropping zip_code, sub_grade, and initial_list_status \n",
    "n_loan_df = n_loan_df.drop([\"zip_code\"], axis=1)\n",
    "n_loan_df = n_loan_df.drop([\"sub_grade\"], axis=1)\n",
    "n_loan_df = n_loan_df.drop([\"initial_list_status\"], axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Converting dates and other numbers into actual numeric variables"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# int_rate\n",
    "n_loan_df.int_rate = n_loan_df.int_rate.str.replace('%', \" \").astype(float)\n",
    "\n",
    "# emp_length\n",
    "# Removing 2110 rows with n/a in emp_length\n",
    "n_loan_df = n_loan_df[n_loan_df.emp_length.str.contains('n/a')==False]\n",
    "# less than one year is now 0\n",
    "n_loan_df.emp_length = n_loan_df.emp_length.str.replace(r'( years)|( year)|(\\+ years)', \"\").str.replace('< 1','0').astype(float)\n",
    "\n",
    "# Dates\n",
    "# issue_d is now delta time since issued\n",
    "n_loan_df.issue_d = ((max(pd.to_datetime(n_loan_df['issue_d'])) - pd.to_datetime(n_loan_df['issue_d'])).dt.days/365).astype(int)\n",
    "\n",
    "# last_credit_pull_d is now delta time since last credit check\n",
    "n_loan_df.last_credit_pull_d = (max(pd.to_datetime(n_loan_df['last_credit_pull_d'])) - pd.to_datetime(n_loan_df['last_credit_pull_d'])).dt.days/365\n",
    "n_loan_df.last_credit_pull_d = n_loan_df.last_credit_pull_d[n_loan_df.last_credit_pull_d.isnull()==False]\n",
    "n_loan_df = n_loan_df[n_loan_df.last_credit_pull_d.isnull() == False]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      " hardship_flag \n",
      " 2 \n",
      " ['N' 'Y']\n",
      "\n",
      " disbursement_method \n",
      " 2 \n",
      " ['Cash' 'DirectPay']\n",
      "\n",
      " application_type \n",
      " 2 \n",
      " ['Individual' 'Joint App']\n",
      "\n",
      " addr_state \n",
      " 51 \n",
      " ['UT' 'VA' 'CA' 'NJ' 'PA' 'CT' 'FL' 'WA' 'AZ' 'MA' 'CO' 'MD' 'WI' 'OR' 'MO'\n",
      " 'NY' 'TX' 'IL' 'MN' 'VT' 'AL' 'OK' 'NE' 'MS' 'OH' 'NH' 'SC' 'IN' 'DC' 'TN'\n",
      " 'NC' 'GA' 'HI' 'AR' 'MI' 'LA' 'WY' 'NV' 'ID' 'DE' 'RI' 'MT' 'KY' 'AK' 'NM'\n",
      " 'SD' 'ND' 'KS' 'ME' 'WV' 'IA']\n",
      "\n",
      " verification_status \n",
      " 3 \n",
      " ['Not Verified' 'Source Verified' 'Verified']\n",
      "\n",
      " term \n",
      " 2 \n",
      " [' 36 months' ' 60 months']\n",
      "\n",
      " grade \n",
      " 7 \n",
      " ['A' 'E' 'C' 'B' 'D' 'F' 'G']\n",
      "\n",
      " home_ownership \n",
      " 6 \n",
      " ['MORTGAGE' 'RENT' 'OWN' 'ANY' 'OTHER' 'NONE']\n",
      "\n",
      " purpose \n",
      " 14 \n",
      " ['debt_consolidation' 'home_improvement' 'other' 'major_purchase'\n",
      " 'credit_card' 'medical' 'moving' 'car' 'vacation' 'small_business'\n",
      " 'renewable_energy' 'house' 'wedding' 'educational']\n",
      "\n",
      " loan_status \n",
      " 2 \n",
      " ['Fully Paid' 'Problem Loan']\n",
      "\n",
      " pymnt_plan \n",
      " 2 \n",
      " ['n' 'y']\n",
      "\n",
      " debt_settlement_flag \n",
      " 2 \n",
      " ['N' 'Y']\n"
     ]
    }
   ],
   "source": [
    "column_checker(n_loan_df, types_of_columns(n_loan_df)[0], 0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(55892, 33)"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "n_loan_df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABI0AAAaYCAYAAAAKTCw1AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nf36IAXHRGU9lAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x118927c88>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "nn_col_list = types_of_columns(n_loan_df)[0].index.drop(['loan_status'])\n",
    "\n",
    "row = 4\n",
    "col = 3\n",
    "\n",
    "fig, axs = pl.subplots(row, col, figsize=(20,30))\n",
    "fig.subplots_adjust(wspace=0.5, hspace=0.25)\n",
    "axs[3, 2].axis('off')\n",
    "\n",
    "for num, name in enumerate(nn_col_list):\n",
    "    i = num%row\n",
    "    j = num%col\n",
    "    graph = sns.countplot(y = n_loan_df[name], hue = n_loan_df['loan_status'], ax = axs[i,j], linewidth = 1.5)\n",
    "    graph.set_xlabel(name, fontsize=15)\n",
    "    graph.set_ylabel('', fontsize=1)\n",
    "    graph.tick_params(labelsize=12)\n",
    "    \n",
    "pl.savefig('data_files/Categorical_variables.png', dpi=300)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      " hardship_flag\n",
      "N    55654\n",
      "Y      238\n",
      "Name: loan_status, dtype: int64\n",
      "\n",
      " disbursement_method\n",
      "Cash         55889\n",
      "DirectPay        3\n",
      "Name: loan_status, dtype: int64\n",
      "\n",
      " application_type\n",
      "Individual    55387\n",
      "Joint App       505\n",
      "Name: loan_status, dtype: int64\n",
      "\n",
      " addr_state\n",
      "AK     109\n",
      "AL     630\n",
      "AR     352\n",
      "AZ    1234\n",
      "CA    9356\n",
      "CO    1181\n",
      "CT    1000\n",
      "DC     250\n",
      "DE     170\n",
      "FL    4067\n",
      "GA    1897\n",
      "HI     266\n",
      "IA      12\n",
      "ID      40\n",
      "IL    2106\n",
      "IN     240\n",
      "KS     363\n",
      "KY     472\n",
      "LA     619\n",
      "MA    1723\n",
      "MD    1496\n",
      "ME      40\n",
      "MI    1091\n",
      "MN     905\n",
      "MO     975\n",
      "MS      90\n",
      "MT     123\n",
      "NC    1228\n",
      "ND      34\n",
      "NE      83\n",
      "NH     250\n",
      "NJ    2468\n",
      "NM     276\n",
      "NV     758\n",
      "NY    5096\n",
      "OH    1722\n",
      "OK     428\n",
      "OR     625\n",
      "PA    2078\n",
      "RI     271\n",
      "SC     653\n",
      "SD      91\n",
      "TN     261\n",
      "TX    4235\n",
      "UT     403\n",
      "VA    1885\n",
      "VT      86\n",
      "WA    1183\n",
      "WI     681\n",
      "WV     177\n",
      "WY     113\n",
      "Name: loan_status, dtype: int64\n",
      "\n",
      " verification_status\n",
      "Not Verified       22376\n",
      "Source Verified    16132\n",
      "Verified           17384\n",
      "Name: loan_status, dtype: int64\n",
      "\n",
      " term\n",
      " 36 months    41496\n",
      " 60 months    14396\n",
      "Name: loan_status, dtype: int64\n",
      "\n",
      " grade\n",
      "A    11384\n",
      "B    15496\n",
      "C    13475\n",
      "D     8321\n",
      "E     4653\n",
      "F     1844\n",
      "G      719\n",
      "Name: loan_status, dtype: int64\n",
      "\n",
      " home_ownership\n",
      "ANY            75\n",
      "MORTGAGE    25953\n",
      "NONE            8\n",
      "OTHER         136\n",
      "OWN          4752\n",
      "RENT        24968\n",
      "Name: loan_status, dtype: int64\n",
      "\n",
      " purpose\n",
      "car                    1742\n",
      "credit_card            7963\n",
      "debt_consolidation    27577\n",
      "educational             413\n",
      "home_improvement       4252\n",
      "house                   526\n",
      "major_purchase         2636\n",
      "medical                 964\n",
      "moving                  731\n",
      "other                  5345\n",
      "renewable_energy        111\n",
      "small_business         2128\n",
      "vacation                513\n",
      "wedding                 991\n",
      "Name: loan_status, dtype: int64\n",
      "\n",
      " pymnt_plan\n",
      "n    55668\n",
      "y      224\n",
      "Name: loan_status, dtype: int64\n",
      "\n",
      " debt_settlement_flag\n",
      "N    55787\n",
      "Y      105\n",
      "Name: loan_status, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "for col in nn_col_list:\n",
    "    series = n_loan_df.groupby(col)['loan_status'].count()\n",
    "    print ('\\n',series)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Numeric Features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "21"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(types_of_columns(n_loan_df)[1].index)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "loan_amnt 1208\n",
      "revol_bal 26976\n",
      "out_prncp 3447\n",
      "out_prncp_inv 3460\n",
      "total_pymnt 55342\n",
      "total_pymnt_inv 52965\n",
      "total_rec_late_fee 2627\n",
      "recoveries 4976\n",
      "collection_recovery_fee 2889\n",
      "last_pymnt_amnt 47818\n",
      "policy_code 1\n",
      "total_rec_int 48546\n",
      "total_rec_prncp 11752\n",
      "funded_amnt 1266\n",
      "int_rate 420\n",
      "installment 19462\n",
      "emp_length 11\n",
      "issue_d 6\n",
      "funded_amnt_inv 9307\n",
      "annual_inc 6415\n",
      "last_credit_pull_d 124\n"
     ]
    }
   ],
   "source": [
    "num_col_list = types_of_columns(n_loan_df)[1].index\n",
    "\n",
    "for col in num_col_list:\n",
    "    length = len(n_loan_df[col].unique())\n",
    "    print (col, length)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Column with a single value\n",
    "n_loan_df = n_loan_df.drop([\"policy_code\"], axis =1)\n",
    "\n",
    "# Columns with similar data\n",
    "n_loan_df = n_loan_df.drop([\"installment\"], axis =1)\n",
    "n_loan_df = n_loan_df.drop([\"funded_amnt\"], axis =1)\n",
    "n_loan_df = n_loan_df.drop([\"funded_amnt_inv\"], axis =1)\n",
    "n_loan_df = n_loan_df.drop([\"out_prncp_inv\"], axis =1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "num_col_list = types_of_columns(n_loan_df)[1].index\n",
    "\n",
    "cor_df = pd.DataFrame()\n",
    "for i in range(0,len(num_col_list)):\n",
    "    cor_df = pd.concat([cor_df, n_loan_df[num_col_list[i]]],axis=1)    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "loan_amnt                  float64\n",
       "hardship_flag               object\n",
       "revol_bal                  float64\n",
       "out_prncp                  float64\n",
       "total_pymnt                float64\n",
       "disbursement_method         object\n",
       "total_pymnt_inv            float64\n",
       "total_rec_late_fee         float64\n",
       "recoveries                 float64\n",
       "collection_recovery_fee    float64\n",
       "last_pymnt_amnt            float64\n",
       "application_type            object\n",
       "total_rec_int              float64\n",
       "addr_state                  object\n",
       "total_rec_prncp            float64\n",
       "verification_status         object\n",
       "term                        object\n",
       "int_rate                   float64\n",
       "grade                       object\n",
       "emp_length                 float64\n",
       "home_ownership              object\n",
       "purpose                     object\n",
       "issue_d                      int64\n",
       "loan_status                 object\n",
       "pymnt_plan                  object\n",
       "debt_settlement_flag        object\n",
       "annual_inc                 float64\n",
       "last_credit_pull_d         float64\n",
       "dtype: object"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "n_loan_df.dtypes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAoYAAAJBCAYAAADbZ3R1AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11fac6438>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "f, ax = pl.subplots(figsize=(10, 8))\n",
    "corr = cor_df.corr()\n",
    "sns.heatmap(corr, mask=np.zeros_like(corr, dtype=np.bool), cmap=sns.diverging_palette(250, 15, as_cmap=True),\n",
    "            square=True, ax=ax, annot=True, vmin=0, vmax=1, linewidths=2, xticklabels=1, yticklabels=True)\n",
    "pl.savefig('data_files/Correlations.png', dpi=300)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABJsAAANcCAYAAAD4va6fAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nskmSJEmSJEkd8/8B5Pec/PswVn8AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11ab2a860>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "row1 = 4\n",
    "col1 = 4\n",
    "\n",
    "fig, axs = pl.subplots(row1,col1, figsize=(20,15))\n",
    "fig.subplots_adjust(wspace=0.5, hspace=0.25)\n",
    "\n",
    "for num, name in enumerate(num_col_list):\n",
    "    i = num//row1\n",
    "    j = num%col1\n",
    "    graph = sns.boxplot(y = n_loan_df[name], ax = axs[i,j], linewidth = 1.5)\n",
    "    graph.set_xlabel(name, fontsize=15)\n",
    "    graph.set_ylabel('', fontsize=1)\n",
    "    graph.tick_params(labelsize=12)\n",
    "    \n",
    "pl.savefig('data_files/Outliers.png', dpi=300)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "n_loan_df = n_loan_df[n_loan_df.annual_inc.isnull()==False]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Removing outgroups"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "8383 outliers found in more than one feature will be remove\n",
      "(55888, 28) (47505, 28)\n"
     ]
    }
   ],
   "source": [
    "n_loans = n_loan_df.reset_index(drop=True)\n",
    "# ADAPTE FROM CUSTOMER SEGMENTATION PROJECT\n",
    "outliers_all = np.array([], dtype='int64')\n",
    "for column in types_of_columns(n_loans)[1].index:\n",
    "    Q1 = np.percentile(n_loans[column], 25)\n",
    "    Q3 = np.percentile(n_loans[column], 75)\n",
    "    step = (Q3 - Q1)*2\n",
    "    out_list_feat=list(n_loans[~((n_loans[column] >= Q1 - step) & (n_loans[column] <= Q3 + step))].index)\n",
    "    outliers_all = np.append(outliers_all, out_list_feat)\n",
    "\n",
    "from collections import Counter\n",
    "out_count = Counter(outliers_all)\n",
    "out_customer =[]\n",
    "for customer, count in out_count.items():\n",
    "    if count > 1:\n",
    "        out_customer.append(customer.astype(int))\n",
    "outliers  = list(sorted(out_customer))\n",
    "\n",
    "print(\"%d outliers found in more than one feature will be remove\" % len(outliers))\n",
    "\n",
    "n_loan_no_out = n_loans.drop(n_loans.index[outliers])\n",
    "    \n",
    "print(n_loan_df.shape, n_loan_no_out.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABJsAAANcCAYAAAD4va6fAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAIABJREFUeJzs3X+8XVV95//XG6KCBCgJgdpYSCEiDiJWI23HUoOiRS0V\nwbYUVKhfBXVEqnaYToEmI1JFwapUi3zFAioWFaTiD1pQ0Fod26AgZsykQYMNgoRAhRBAgc/8sffB\nk5Obm3uTc3POuXk9H4/zyN37s9c6a99HsrL3Z6+9VqoKSZIkSZIkqR+2G3QDJEmSJEmSNH2YbJIk\nSZIkSVLfmGySJEmSJElS35hskiRJkiRJUt+YbJIkSZIkSVLfmGySJEmSJElS35hskiRJkiRJUt+Y\nbJIkSZIkSVLfmGySJEmSJElS38wYdAP6bffdd6958+YNuhnS0Lvhhhvuqqo5g26H1mcfJk2Mfdjw\nsf+SJsb+a/jYf0kTM5n+a9olm+bNm8eSJUsG3Qxp6CW5ddBt6Jck1wO/CTzc7rqtqp7axl4AfBDY\nC/gWcEJV3drGArwLeG1b7iPAn1dVtfF5wN8BvwH8CHhTVV3b9b3HAu8EdgeuAV5TVXe3sScAfwu8\nAlgHvLuq3rupc7EPkyZmOvVh04X9lzQx9l/Dx/5LmpjJ9F++RidpunhTVc1sP51E0+7AFcAZwCxg\nCXBZV5kTgSOBg4BnAEcAJ3XFPwl8B5gNnAZ8Jsmctu4DgA8DrwL2pEkofair7GLgKcDewKHAqUkO\n7+P5SpIkDb0kT0nyYJKPd+17QZJlSdYluS7J3l2xJDk7yZr2c3b7gLATn9eWWdfWcVjP9x2b5NYk\n9ye5MsmsrXOmkrqZbJI0nR0FLK2qT1fVgzQJoIOS7N/GjwfOrapVVXUbcA5wAkCS/YBnAYuq6oGq\nuhz4LnB0W/Y44Kqq+lpVraVJaB2VZOeuus+sqnuq6vvABZ26JUmStiEfBP6tszHgh4GSthKTTZpy\nX/nKV1i4cCHXXXfdoJui6e2dSe5K8i9JFrb7DgBu6hxQVfcDK9r9G8Tbn7tjP6iq+8aJd9d9C/AQ\nsF+S3YAnjVP3epKcmGRJkiWrV6+e4Olqa1izZg1vfvObWbNmzaCbIkmTZh+mQUtyDPCfwJe7dg/y\nYaBGxIoVK3jpS1/KihUrBt0UbaZJJZscAqnN8Vd/9VcAnHXWWQNuiaax/wHsA8ylGUF0VZJ9gZnA\nT3uOvRfoXHD0xu8FZrb91GTLdsdnttu9dY95oVNVF1TVgqpaMGeO84UOk4svvpibb76ZSy65ZNBN\nkaRJsw/TICXZBXg78Nae0EAeBm7JuWjrO+2007j//vs57bTTBt0UbabJjmxyCKQm5Stf+QoPP9zM\n2fzwww87uklToqq+VVX3VdVDVXUx8C/AS4C1wC49h+8KdC5QeuO7AmvbCcInW7Y7vrbd7q37PjQy\n1qxZw9VXX01VcfXVVzsyQNJIsQ/TEDgTuLCqVvXsH9TDwPU4snx4rVixgp/85CcA/OQnP3F004ia\ncLLJIZDaHJ1RTR2ObtJWUkCApTTJbgCS7ATs2+6nN97+3B3bp6fP6Y13170v8HhgeVXdA9w+Tt0a\nARdffDGPPvooAI888ogjAySNFPswDVKSZwKHAX89RnhQDwPX48jy4dU7msnRTaNpQsmmYR8CaVZ6\neHVGNW1sW9pSSX4pye8m2SHJjCTHAb8DXA18Fnh6kqOT7AAsAm6qqmVt8UuAtyaZm2Qu8DbgIoCq\nWg7cCCxq6z4KOBC4vC37CeCIJIe0SawzgSu6+rRLgNOT7JbkacDrOnVrNFx77bXrjcy85pprBtwi\nSZo4+zAN2EJgHvCjJHcAfwYcneTbDOhhYD9OSltHZ1TTxrY1GiY6smmoh0CalR5eM2bMGHdb6oPH\nAe8AVgN3AScDR1bV8qpaTTNa8izgHuBg4Jiush8GrgJubj+fb/d1HAMsaMu+E3hFWydVtRR4PU3S\n6U5gJ+CNXWUXAbcAtwLXA++uqqv7ddKaeocddthjfdaMGTN44QtfOOAWSdLEHXLIIeNuS1PsApoE\n0jPbz/nAF4DfZbAPAyVtJZu88+8aAvnrY4Q3ewhkkr4NgdTw+ou/+Ave/va3P7btEEj1W5v8ec44\n8WuB/TcSK+DU9jNWfCXNk7mN1X0pcOlGYg8Br2k/GkHHH388V1/d5Ae33357Xv3qVw+4RZI0cQ8+\n+OB62w899NCAWqJtUVWto5lzF4D23u/BzkO7JEcDfwN8HPgWGz4M3IfmQSDAR9jwYeBFNA8Df0TP\nw8AknYeBs4FrgT/p8+lJmoCJjGxaiEMgtZme//znrzcy4NBDDx1wiyRpYmbPnv1Yn7Vw4UJmz549\n4BZJ0sR97WtfW2/7q1/96oBaIkFVLa6qV3ZtX1tV+1fVjlW1sH3A14lVVZ1aVbPaz6ntA8JOfGVb\nZseqemr7YLH7uy6tqr2qaqeqellV3b1VTlLSeiaSbHIIpLbIn/xJ8zDhta997YBbIkmT03VtK0kj\npbf/sj+TJG1Nm0w2VdW6qrqj86F5ve3Bqlo94PlQNCI6E1J2XkeRpFGwZs0arr/+egCuv/56lw2X\nJEmSJmjSszVX1eKe7YHMh6LRsGLFClauXAnAypUrWbFiBfPnzx9soyRpAsZaNvwtb3nLgFslSZIk\nDb+JrkYnbZZ3vOMd425L0rBy2XBJkiRp85hs0pTqjGra2LYkDavDDjtsvQUOXvjCFw64RZIkSdJo\nMNmkKTVv3rxxtyVpWB1//PFst13z3+T222/Pq1/96gG3SJIkSRoNJps0pU4//fRxtyVpWM2ePZvD\nDz+cJBx++OHMnj170E2SJEmSRoLJJk2p+fPnPzaaad68eU4OLmmkHH/88Rx44IGOapIkSZImwWST\nptxzn/tcAH7nd35nwC2RpMmZPXs2H/jABxzVJEmSJE2CySZNuUsvvRSAj33sYwNuiSRJkiRJmmom\nmzSl/uEf/oGqAqCquOqqqwbcIkmSJEmSNJVMNmlKve9971tv+73vfe+AWiJJkiRJkrYGk02aUp1R\nTRvbliRJkiRJ04vJJkmSJEmSJPWNySZJkqQhleTjSe5Icm+S5Ule2xV7QZJlSdYluS7J3l2xJDk7\nyZr2c3aSdMXntWXWtXUc1vO9xya5Ncn9Sa5MMmvrnLEkSZoOTDZJkiQNr3cB+1TVLsDvA+9I8uwk\nuwNXAGcAs4AlwGVd5U4EjgQOAp4BHAGc1BX/JPAdYDZwGvCZJHMAkhwAfBh4FbAnsA740FSdoCRJ\nmn5MNkmSJA2pqvpeVa3rbLaffYGjgKVV9emqehBYDByUZP/22OOBc6tqVVXdBpwDnACQZD/gWcCi\nqnqgqi4Hvgsc3ZY9Driqqr5WVWtpElpHJdl5ik9XkiRNEyabJEmShliSDyVZBywDbge+CBwA3NQ5\npqruB1a0++mNtz93x35QVfeNE++u+xbgIWC/Mdp2YpIlSZasXr16s89RkiRNLyabJEmShlhVvRHY\nGTiE5tW5h4CZwE97Dr23PY4x4vcCM9t5myZbtjfe3bYLqmpBVS2YM2fOZE5LkiRNYyabJEmShlxV\nPVJVXweeDLwBWAvs0nPYrkBntFJvfFdgbVXVZpTtjUuSJI3LZJMkSdLomEEzZ9NSmsm/AUiyU9d+\neuPtz92xfXrmYOqNd9e9L/B4YHnfzkKSJE1rJpskTRtJnpLkwSQf79o3kKXBkzwhyUfb5crvSPLW\nqT5/SdNLkj2SHJNkZpLtk/wu8MfAl4HPAk9PcnSSHYBFwE1Vtawtfgnw1iRzk8wF3gZcBFBVy4Eb\ngUVJdkhyFHAgcHlb9hPAEUkOaZNYZwJX9MzxJEmStFEmmyRNJx8E/q2zMeClwRcDTwH2Bg4FTk1y\neH9OU9I2omhemVsF3EOzotyfVtXnqmo1zepxZ7Wxg4Fjusp+GLgKuLn9fL7d13EMsKAt+07gFW2d\nVNVS4PU0Sac7gZ2AN07NKUqSpOloxqAbIEn9kOQY4D+BbwDz292PLQ3eHrMYuCvJ/u3T/8eWBm/j\n59AkoM7vWhr8RVX1AHB5klNobu7Op2tp8LbsGcD3k+zcPv0/Hjihqu4B7klyAc2y41dP8a9C0jTR\nJn+eN078WmD/jcQKOLX9jBVfCSwcp+5LgUsn3lpJkqRfcGSTpJGXZBfg7UDvq2oDWRo8yW7Ak8ap\nu7f9Lh0uSZIkadow2SRpOjgTuLAzQqnLoJYGn9lu99a9wbLh4NLhkiRJkqYXX6OTNNKSPBM4DPj1\nMcKbvTR4ki1ZGnxtu70L8OAYZSVJkqSBOO+881ixYsWgmzEpp5xyyqCbsIH58+dz8sknD7oZQ8uR\nTZJG3UJgHvCjJHcAfwYcneTbDGhp8HaeptvHqVuSJEmSpi1HNkkadRcAf9+1/Wc0yac3tNvvSXI0\n8AU2vjT4F9vttwEfgGZp8CSdpcFPB17ChkuDfzPJIcC32XBp8EuA05MsAX4ZeB3NBOGSJEnSwAz7\naJyFCxdusO/973//1m+ItogjmySNtKpaV1V3dD40r7A9WFWrB7w0+CLgFuBW4Hrg3VXlSnSSJEnS\nOObPn7/e9v77j7nwqoacySZJ00pVLa6qV3ZtX1tV+1fVjlW1sF3uuxOrqjq1qma1n1Pb5cI78ZVt\nmR2r6qntMuPd33VpVe1VVTtV1cuq6u6u2ENV9Zqq2qWq9qyq907xqUuSJA2NJB9PckeSe5MsT/La\nrtgLkixLsi7JdUn27oolydlJ1rSfs9vFWzrxeW2ZdW0dh/V877FJbk1yf5Irk8zaOmesfvnIRz6y\n3vb5558/oJZoS5hskiRJkiT127uAfapqF+D3gXckeXaS3YErgDOAWcAS4LKucicCR9LMd/kM4Ajg\npK74J4HvALOB04DPJJkDkOQAmlHqrwL2BNYBH5qqE9TUecITngA4qmmUTSjZZFZakiRJkjRRVfW9\nqlrX2Ww/+wJHAUur6tNV9SCwGDgoSSercDxwblWtqqrbgHNo571Msh/wLGBRVT1QVZcD36WZNgHg\nOOCqqvpaVa2lSWgd1bPgi0bA/vvvz0EHHeSophE20ZFNZqUlSZIkSROW5ENJ1gHLaFbq/SJwAHBT\n55iquh9Y0e6nN97+3B37QdeCLGPFu+u+BXgI2G+Mtp2YZEmSJatXr97sc5Q0tgklm8xKS5IkSZIm\no6reCOwMHEIzSOEhYCbw055D722PY4z4vcDM9g2ZyZbtjXe37YKqWlBVC+bMmTOZ05I0AROes2mY\ns9KSJEmSpOFTVY9U1deBJwNvoFk5eJeew3YFOveFvfFdgbXtIi6TLdsbl7SVTDjZNMxZaYdASpIk\nSdJQm0HzdsxSmmlWAEiyU9d+euPtz92xfXredumNd9e9L/B4YHnfzkLShExqNbphzUo7BFKSJEmS\nhkOSPZIck2Rmku2T/C7wx8CXgc8CT09ydJIdgEXATVW1rC1+CfDWJHOTzAXeBlwEUFXLgRuBRUl2\nSHIUcCBweVv2E8ARSQ5pk1hnAlf0vE0jaSuYsQXlOtnn4zs7x8lK/2u7PWZWuusf/0E0HUR32U7d\nZqXHcN5557FixYpBN2NSTjnllEE3YQPz58/n5JNPHnQzJEmSpOmgaAYnnE8zwOFW4E+r6nMASY4G\n/gb4OPAt4Jiush8G9gFubrc/0u7rOIYm+XQP8CPgFVW1GqCqliZ5Pc095WzgWuBP+n96kjZlk8mm\nJHsAzwc+DzwAHEaTlf5j4JvAe9rO4gtsPCv9xXb7bcAHoMlKJ+lkpU8HXsKGWelvJjkE+DZmpSVJ\nkiRp6LXJn+eNE78W2H8jsQJObT9jxVcCC8ep+1Lg0om3VtJUmMjIJrPSQ2zYR+MsXLhwg33vf//7\nt35DJEmSJEnSVrHJZJNZaW2Jv/zLv+Ttb3/7Y9uLFi0aYGskSZIkSdJUm9QE4dJkPf/5z19v+9BD\nDx1QSyRJkiRJ0tZgsklT7ld/9VcBRzVJkiRJkrQt2NzV6KQJmzVrFrNmzXJUkyRJkiRJ2wBHNkmS\nJEmSJKlvTDZJkiRJkiSpb0w2SZIkSZIkqW9MNkmSJEmSJKlvTDZJkiRJkiSpb0w2SZIkSZIkqW9M\nNkmSJEmSJKlvTDZJkiRJkiSpb0w2SRp5ST6e5I4k9yZZnuS1XbEXJFmWZF2S65Ls3RVLkrOTrGk/\nZydJV3xeW2ZdW8dhPd97bJJbk9yf5Moks7piT0jy0bZNdyR561T/HiRJkiRpGJhskjQdvAvYp6p2\nAX4feEeSZyfZHbgCOAOYBSwBLusqdyJwJHAQ8AzgCOCkrvgnge8As4HTgM8kmQOQ5ADgw8CrgD2B\ndcCHusouBp4C7A0cCpya5PD+nbIkSZIkDSeTTZJGXlV9r6rWdTbbz77AUcDSqvp0VT1IkwA6KMn+\n7bHHA+dW1aqqug04BzgBIMl+wLOARVX1QFVdDnwXOLotexxwVVV9rarW0iS0jkqyc1fdZ1bVPVX1\nfeCCTt2SJEmSNJ2ZbJI0LST5UJJ1wDLgduCLwAHATZ1jqup+YEW7n954+3N37AdVdd848e66bwEe\nAvZLshvwpHHq7m37iUmWJFmyevXqCZ+zJEmSJA0jk02SpoWqeiOwM3AIzatzDwEzgZ/2HHpvexxj\nxO8FZrbzNk22bHd8ZrvdW/fOjKGqLqiqBVW1YM6cORs7RUmSJEkaCSabJE0bVfVIVX0deDLwBmAt\nsEvPYbsCndFKvfFdgbVVVZtRtju+tt3urfs+NFLWrFnDm9/8ZtasWTPopkiSJEkjw2STpOloBs2c\nTUtpJv8GIMlOXfvpjbc/d8f26ZqDaax4d937Ao8HllfVPTSv8m2sbo2Iiy++mJtvvplLLrlk0E3R\nNqhd1fLCdtXL+5LcmOTFXfGBrLYpSZK0KSabJI20JHskOSbJzCTbJ/ld4I+BLwOfBZ6e5OgkOwCL\ngJuqallb/BLgrUnmJpkLvA24CKCqlgM3AouS7JDkKOBA4PK27CeAI5Ic0iaxzgSu6Jrj6RLg9CS7\nJXka8LpO3RoNa9as4eqrr6aquPrqqx3dpEGYAfwH8Dya0ZGnA59qE0WDXG1TkiRpXCabJI26onll\nbhVwD82Kcn9aVZ+rqtU0q8ed1cYOBo7pKvth4Crg5vbz+XZfxzHAgrbsO4FXtHVSVUuB19Mkne4E\ndgLe2FV2EXALcCtwPfDuqrq6XyetqXfxxRfz8MMPA/Dzn//c0U3a6qrq/qpaXFUrq+rRqvo88EPg\n2Qx2tU1JkqRxzRh0AyRpS7TJn+eNE78W2H8jsQJObT9jxVcCC8ep+1Lg0o3EHgJe0340gq699loe\nffRRAB599FGuueYa3vKWtwy4VdqWJdkT2I/mldw30LPaZpLOapvL2PLVNr/RVfctSR5qv/uGnjad\nSDOKir322msLz1CSJE0XjmySJGkMz3nOc9bbPvjggwfUEgmSPI5mJOXF7avAg1ptcz2upilJksZi\nskmSpDH8+7//+3rby5cvH1BLtK1Lsh3wMeBnwJva3YNabVOSJGmTTDZJkjSGH//4x+NuS1tDOxLp\nQpqJuo+uqp+3oYGsttmHU5IkSdsAk02SJI2ha5X4MbelreRvgacBR1TVA137B7napiRJ0rhMNkmS\nNIbnPe95425LUy3J3sBJwDOBO5KsbT/HDXi1TUmSpHG5Gp0kSWN45StfyfXXX7/etrQ1VdWtwEaH\n1A1qtU1JkqRNcWSTJElj+NznPrfe9lVXXTWglkiSNFqSPCHJhUluTXJfkhuTvLgr/oIky5KsS3Jd\nO5KzE0uSs5OsaT9np+td9iTz2jLr2joO6/nuY9vvvT/JlUlmbZ2zltTNZJMkSWO49tpr19u+5ppr\nBtQSSZJGzgzgP4Dn0axmeTrwqTZRtDtwBXAGMAtYAlzWVfZE4EiahQqeARxB80pxxyeB7wCzgdOA\nzySZA5DkAJpXhl9Fs7DCOuBDU3OKksZjskmSpDEccsgh425LkqSxVdX9VbW4qlZW1aNV9Xngh8Cz\ngaOApVX16ap6EFgMHJSk81rw8cC5VbWqqm4DzgFOAEiyH/AsYFFVPVBVlwPfpZnDDuA44Kqq+lpV\nraVJaB3Vs/qmpK1gk8kmh0BKkrZFzZQ3kiRpSyXZE9gPWAocANzUiVXV/cCKdj+98fbn7tgPelbH\n7I13130L8FD73ZK2oomMbHIIpCRpm/P1r399ve1//ud/HlBLJEkaXUkeR7O65cVVtQyYCfy057B7\ngc7oo974vcDMdtDCZMv2xrvbdWKSJUmWrF69enInJWmTNplscgikJGlb9JznPGe97YMPPnhALZEk\naTQl2Q74GPAz4E3t7rXALj2H7grct5H4rsDadpXNyZbtjT+mqi6oqgVVtWDOnDkTPidJEzPpOZuG\ncQikWWlJUr8tX7583G1JkrRx7UikC2neUjm6qn7ehpbSvPnSOW4nYN92/wbx9ufu2D49AxB64911\n7ws8HvA/cWkrm1SyaViHQJqVliT12+23377e9o9//OMBtUSSpJH0t8DTgCOq6oGu/Z8Fnp7k6CQ7\nAIuAm9r7S4BLgLcmmZtkLvA24CKAqloO3AgsSrJDkqOAA4HL27KfAI5IckibxDoTuKJngIOkrWDC\nyaZhHgIpSZIkSRoO7aJRJwHPBO5Isrb9HFdVq2mmTjkLuAc4GDimq/iHgauAm9vP59t9HccAC9qy\n7wRe0dZJVS0FXk+TdLoT2Al441Sdp6SNmzGRg3qGQL6kZwjk8V3HbWwI5L+222MOgezKNB9E0zF0\nl+3U7RBISZIkSRpyVXUrkHHi1wL7byRWwKntZ6z4SmDhOHVfClw68dZKmgoTHdnkEEhJkiRJkiRt\n0iaTTQ6BlCRtixYuXDjutiRJkqSxbfI1OodASpK2Rb/3e7/H9ddf/9j2EUccMbjGSJIkSSNkUqvR\nSZK0rXjf+9437rYkSZKksZlskiRpDKtWrVpv+z/+4z8G1BJJkiRptJhskiRJkiRJUt+YbJI00pI8\nIcmFSW5Ncl+SG5O8uCv+giTLkqxLcl276EEnliRnJ1nTfs5Okq74vLbMuraOw3q++9j2e+9PcmWS\nWT3t+miSe5PckeStU/27kCRJkqRhYLJJ0qibAfwH8DxgV+B04FNtomh34ArgDGAWsAS4rKvsicCR\nwEHAM4AjaFbf7Pgk8B1gNnAa8JkkcwCSHECzuuargD2BdcCHusouBp4C7A0cCpya5PB+nbQkSZIk\nDSuTTZJGWlXdX1WLq2plVT1aVZ8Hfgg8GzgKWFpVn66qB2kSQAcl6aygeTxwblWtqqrbgHOAEwCS\n7Ac8C1hUVQ9U1eXAd4Gj27LHAVdV1deqai1NQuuoJDt31X1mVd1TVd8HLujULUmSJEnTmckmSdNK\nkj2B/YClwAHATZ1YVd0PrGj30xtvf+6O/aCq7hsn3l33LcBDwH5JdgOeNE7dvW0+McmSJEtWr149\n8ZOVJEmSpCE0Y9ANkKR+SfI44BPAxVW1LMlMoDd7cy/QGX00E/hpT2xmO29Tb6wTn7uRst11z2y3\ne+vemTFU1QU0I59YsGBBbez8JEmSNPzOO+88VqxYMehmjLTO7++UU04ZcEtG2/z58zn55JMH8t0m\nmyRNC0m2Az4G/Ax4U7t7LbBLz6G7AvdtJL4rsLaqKslky3bH17bbuwAPjlFWkiRJ09SKFSu48Xvf\n55Enztr0wRrTdj9rnr/e8IOfDLglo2v7dXcP9PtNNkkaee1IpAtpJup+SVX9vA0tpZk7qXPcTsC+\n7f5O/CDgX9vtg3pi+yTZuetVuoNoRk51l+3UvS/weGB5Vd2X5PY2fs0YdUuSJGkae+SJs3hg/5cM\nuhnahu247IsD/X7nbJI0Hfwt8DTgiKp6oGv/Z4GnJzk6yQ7AIuCmqlrWxi8B3ppkbpK5wNuAiwCq\najlwI7AoyQ5JjgIOBC5vy34COCLJIW0S60zgiq7E1CXA6Ul2S/I04HWduiVJkiRpOjPZJGmkJdkb\nOAl4JnBHkrXt57iqWk2zetxZwD3AwcAxXcU/DFwF3Nx+Pt/u6zgGWNCWfSfwirZOqmop8HqapNOd\nwE7AG7vKLgJuAW4FrgfeXVVX9+/MJUmSJGk4+RrdJji525Zzcrf+GOTkbsOsqm4FMk78WmD/jcQK\nOLX9jBVfCSwcp+5LgUs3EnsIeE370QhKQvNX5BfbkiRJkjbNZNMmOLnblnNyty036MndpG3Rbrvt\nxt13/+Lf3qxZ/j8gSZIkTYTJpglwcjcN2qAnd5O2Rd2JJoA1a9YMqCWSJEnSaHHOJkmSJEmSJPWN\nySZJkiRJkiT1jckmSZIkSZIk9Y3JJkmSxrDddtuNuy1JkiRpbF45S5I0hkcffXTcbUmSJEljM9kk\nSZIkSZKkvjHZJEmSJEmSpL4x2SRJkjSEkrwpyZIkDyW5qCf2giTLkqxLcl2SvbtiSXJ2kjXt5+wk\n6YrPa8usa+s4rKfuY5PcmuT+JFcmmTXlJytJkqYVk02SJEnD6cfAO4CPdu9MsjtwBXAGMAtYAlzW\ndciJwJHAQcAzgCOAk7rinwS+A8wGTgM+k2ROW/cBwIeBVwF7AuuAD/X5vCRJ0jRnskmSJGkIVdUV\nVXUlsKYndBSwtKo+XVUPAouBg5Ls38aPB86tqlVVdRtwDnACQJL9gGcBi6rqgaq6HPgucHRb9jjg\nqqr6WlWtpUloHZVk5yk7UUmSNO2YbJIkSRotBwA3dTaq6n5gRbt/g3j7c3fsB1V13zjx7rpvAR4C\n9hurIUlObF/1W7J69erNPiFJkjS9mGySJEkaLTOBn/bsuxfYeSPxe4GZ7bxNky3bG19PVV1QVQuq\nasGcOXMmdRKSJGn6MtkkSZI0WtYCu/Ts2xW4byPxXYG1VVWbUbY3LkkT4iIH0rbNZJMkSdJoWUoz\n+TcASXYC9m33bxBvf+6O7dMzB1NvvLvufYHHA8v72H5J2wYXOZC2YRNKNpmVliRJ2rqSzEiyA7A9\nsH2SHZLMAD4LPD3J0W18EXBTVS1ri14CvDXJ3CRzgbcBFwFU1XLgRmBRW99RwIHA5W3ZTwBHJDmk\nTWKdCVxLCVqJAAAgAElEQVTRM8eTJG2SixxI27aJjmwyKy1JkrR1nQ48APw58Mr259OrajXNjdVZ\nwD3AwcAxXeU+DFwF3Nx+Pt/u6zgGWNCWfSfwirZOqmop8HqapNOdwE7AG6fm9CRto4ZikQMXOJCm\n1oyJHFRVVwAkWQA8uSv0WFa6jS8G7kqyf/t07bGsdBs/hyYBdX5XVvpFVfUAcHmSU2guns6nKyvd\nlj0D+H6SnX26JkmSpruqWkzzxH+s2LXA/huJFXBq+xkrvhJYOM73XgpcOpm2StIkzAR6sztbusjB\n3I2U7a37MVV1AXABwIIFC2pypzC+2267je3X/ZQdl32xn9VKk7L9ujXcdtvDA/v+CSWbxrFBVjpJ\nJyu9rDfO5LPS3+iq+5Yknaz0DVvY7gmzo9AwGHRHIUmSJPXJZi9ykMRFDqQRsaXJpqHISic5kWbE\nFHvttdfkzkCSJEmStLUspXkDBhh3kYN/bbfHXOSga9DCQTSv/naX7dQ9kEUO5s6dyx0PzeCB/V+y\nNb9WWs+Oy77I3Ll7Duz7tzTZNBRZ6akcAmlHoWEw6I5CkiRJmox2QYMZdC1yADxMs8jBe5IcDXyB\njS9y0Hm15G3AB6BZ5CBJZ5GD04GXsOEiB99McgjwbVzkQBqYiU4QvjEuvStpoIZxtcwkT0jy0ST3\nJrkjyVun8FcgSZI0jFzkQNqGTSjZ5NK7kobYMK6WuRh4CrA3cChwapLDt/hMJUmSRkRVLa6q9HwW\nt7Frq2r/qtqxqha2Cxd0ylVVnVpVs9rPqe3CB534yrbMjlX11HbBhO7vvbSq9qqqnarqZVV199Y6\nZ0m/MNGRTWalJQ2lqrqiqq4E1vSEHlsts6oepEkAHZSks3rTY6tlVtVtwDnACQBdq2UuqqoHqupy\n4Ls0/R10rZZZVWtpElpHdY3UPB44s6ruqarv07zme0K/z12SJEmShtGE5mxy6V1JI2ggq2Um+QHw\npDHqfvnGGuoiB5IkSZKmky2ds0mShtWmVrTcnNUyN1a2Oz6z3e6te4OVNDuq6oKqWlBVC+bMmbPR\nE5IkSZKkUWCySdJ0tdmrZW5G2e742na7t27nm5MkSZK0TTDZJGm6GshqmVV1D3D7OHVLkiRJ0rRm\nsknSSBvS1TIvAU5PsluSpwGv69QtSZIkSdOdySZJo24YV8tcBNwC3ApcD7y7qq7u1wlLkiRJ0jCb\n0Gp0kjSshnG1zKp6CHhN+5EkSZKkbYojmyRJkiRJktQ3JpskSZIkSZLUNyabJEmSJEmS1DcmmyRJ\nkiRJktQ3ThA+Aduvu5sdl31x0M0YWds9eC8Aj+6wy4BbMrq2X3c3sOegmyFJkiRpAryH3DLeQ265\nQd9DmmzahPnz5w+6CSNvxYr7AJi/j8mSzbenfxclSZKkEeB1+5bzHrIfBnsPabJpE04++eRBN2Hk\nnXLKKQC8//3vH3BLJEmSJGlqeQ+55byHHH3O2SRJkiRJkqS+MdkkSZIkSZKkvjHZJEmSJEmSpL4x\n2SRJkiRJkqS+MdkkSZIkSZKkvjHZJEmSJEmSpL6ZMegGSJK2Xeeddx4rVqwYdDMmrLMM77CZP3++\nyyxLkiRpaDiySZIkSZIkSX3jyCZJ0sAM82ichQsXbrDv/e9//9ZviCRJkjRiHNkkSdIYZs+evd72\nHnvsMaCWSJIkSaPFZJMkSWO4/PLL19v+1Kc+NaCWSJIkSaPFZJMkSRux3XbNf5OOapIkSZImzjmb\nJGmaGrWV3obRjjvuCMCTnvSkoV2JbhS4Wp4kSdK2xWSTJE1TK1as4MbvfZ9Hnjhr0E0ZWdv9rAC4\n4Qc/GXBLRtf26+4edBMkSZK0lZlskqRp6rbbbgNq0M0YaY/usMugmzANVPt3UZIkSdsK52ySpCmS\nZFaSzya5P8mtSY4ddJskaSLsvySNMvswafAc2SRJU+eDwM+APYFnAl9IclNVLd0aXz537lxW33Pv\n1viqaWu7B5vfnyOctkSYO3fuoBuhyRto/yVJW8g+TBqwoU42JZkFXAi8CLgL+J9VdelgWzVcRmEC\n4E77hnlyXSevVb8l2Qk4Gnh6Va0Fvp7kH4BXAX++Ndowf/78rfE1W+S2227jgQceGHQzNuqBRx4E\nYMdHM+CWjG/HHXcc4oTOniPxd1G/MAz913nnncfVV1+9Nb5qs61bt46q0XpVeeHChYNuwgaS8MQn\nPnHQzRjX4Ycf7nXiCBmGPmzYeQ/ZH95Djm+ok02YkZ4WOqs5SduY/YCHq2p5176bgIW9ByY5ETgR\nYK+99upbA0bhP79hv9jpzDU0vImchhc76rOB91+StAUm1IfZfw037yFHX4b1iUybkb6HJiO9vN13\nCfDjqtpoRnrBggW1ZMmSrdRKaXQluaGqFgy6HdNVkkOAT1fVL3ftex1wXFUt3Fg5+zBpYuzDpo79\n1/SwZs0aXvGKV1BVJOEzn/kMs2fPHnSzhP3XVNucPsz+S5qYyfRfwzxB+MYy0gf0HpjkxCRLkixZ\nvXr1VmugJI1jLdA70c+uwH0DaIskTYb91zQwe/ZsXvjCFwLwohe9yESTtiX2YdIQGOZk00ygd2bb\ne4Gdew+sqguqakFVLZgzZ85WaZwkbcJyYEaSp3TtOwjwNWBJw87+a5o46aSTeMYznsGJJ5446KZI\nW5N9mDQEhjnZZEZa0siqqvuBK4C3J9kpyW8Dvw98bLAtk6Tx2X9NH7Nnz+YDH/iAo5q0TbEPk4bD\nMCebzEhLGnVvBHYE7gQuBd7gAgeSRoT9l6RRZh8mDdjQrkZXVfcn6WSkXwv8Ok1G+r8OtmWSNDFV\ndTdw5KDbIUmTZf8laZTZh0mDN8wjm8CMtCRJkiRJ0khJVQ26DX2VZDVw66DboQ3sDtw16EZoPXtX\nlTPqDxn7sKFk/zWc7MOGjP3X0LIPGz72X0PG/mto2X8Nnwn3X9Mu2aThlGRJVS0YdDskabLsvySN\nMvswSaPK/mu0DftrdJIkSZIkSRohJpskSZIkSZLUNyabtLVcMOgGSNJmsv+SNMrswySNKvuvEeac\nTZIkSZIkSeobRzZJkiRJkiSpb0w2SZIkSZIkqW9MNkmSJEmSJKlvTDZJkiRJkiSpb0w2SZIkSZIk\nqW9MNkmSJElDJsmpSRYOuh2SJG0Ok03TTJKLkiwZdDuGVZL9kixO8kuDbouk0ZCkkrxpgscubI9/\neh++94S2rplbWpekkXQqsHDQjZA0fST5wyQnbEa5FyX50838zml7PeNDgfGZbNK2Zj9gEWCySZIk\nCUiy46DbIGmr+EPghM0o9yJgs5JN05wPBcZhskmSNDKSbJ/k8YNuhySNpx09cHOSh5L8R5Kzksxo\nY4uT3DVGmcdGUSZZCcwGFrX7ayJPzzt1J3lukm8neTDJjUl+u+e4lUnOTXJGklXAve3+i5IsSfLC\nJN9Ncn+Sryc5oKf89kn+Z5Ll7TmuSnJRV/z6JJ9JcmL7XQ8k+UKSuZP9XUqSRpPJpmkuyTOTfDnJ\nuiT3JPlEkj17jnlXe0G0tr1Y+ESSX+45ZmWSc5K8pT3mniR/P5nX0ZK8NMk1Se5Mcm+S/53kRT3H\ndC6SfqO92Hmgvcj5tSR7JLmybef3kzx/Mm1sL9Kuag//YXvhtnLiv01JW1vXjc+RSZYCDwK/kWSv\n9t/33W3/9o9JntpV7odJ3jNGfZ9O8vWu7V9r+5V7k9yX5Kok8/vQ9F9J8vn2Ru1HSV7f047fSvK5\nJLe3x9yY5Lg+fK+kAWuvbS4Dvg28DDgP+DPgbyZRzcuBnwIXAr/Vfr49wbJPBD4OnA/8AfCfwJd6\nr+2AY4HnAW8E/qhr/17Ae4CzgD8G9gAuS5KuYz4M/C/gU8DvAW9rv7fbbwEnA28F/j/gGcCVEzwH\nSX3WJoSPBp7XlcRe3MbelOTf2+TxiiRv6Sq3mObf+N5d5S5qY32/num59luWJmn+9ST/peuYTyW5\nfoyyi5P8JMnjksxr23pMkr9rr/VWJXlle+ypSX6cZHWSs5Ns11PPXUl+vb1nXZfkO0kO6TpmJZvx\nUGBbMmPQDdDUSTIHuB74Ps0FxUzgXcA1SRZU1c/aQ38ZOBtYBexO05l8JcnTq+rRrir/EPgucCLw\nZOC9wF/RXKRMxK8BXwTOBR4BXkxz8fM7VfUvXcc9EbgAeDdwP/AB4GPAQ8CXgA/RDFn8dJJfrap1\nE2zjt2ku9s4BjgJub+uUNNzm0fQHbwfuAG4Fvg6sAV4PrAP+HLg2yX5V9QDNDdAfAf+9U0mauQJe\nStN/kOQJwJeBnwOvAx6muXn6apIDq+ruLWjzhTT91nk0N41/m2RVVX2+65z+N01ftw54LvB3SR6t\nqk9uwfdKGry3A9dX1fHt9tVtnuadSd4xkQqq6jtJHgZWVdX/nuT37wicVlWXAiS5DvgRzSswf95z\n7O9V1YM9+2YBz62qf2/Lbwd8FngqsCzJ/jTJo1Oq6gNd5S7rqWcP4Leq6kdtPbcCX09yeFVdPclz\nkrTlzqRJJv8Sv7h/W5XkdTTXK+8F/hE4FDg3yROq6l3AR4CnAM+nuaYBWN3+OY+puZ7Zu23PGcAD\nNNdn/5jkKW2fdSHNfeSvVdUPAdqE+PHAx6vq51358bOBT9Ak2l4DXJzk19vveA3wbOAdwHeAv+9q\nwxOBi4G/prn+XARckWTv9v7z5cB1wGfa3xHA/9mCc552TDZNb29r//zdquoMj/53mg7haOCTAFV1\nQqdAku2Bb9Iknn4b+FpXfT8Hjqyqh9tj/wtwDBNMNlXVY0/02guX64ADaC5YupNNOwJvrqqvtsf+\nCvBBYFFVndPuWwUspXki96WJtLGq7k3yf9vjvlNVKyfSbkkDNxs4rKpuBEhyJrAT8MxOQijJvwAr\naS4aPkhzsXBqkt/sulE7Ang88Ol2+09oLrr2q6oftPV8C/gBcBLwzi1o85eq6i/an/8xyb7A6cDn\nAbovwNqLo6/RJMhfR9s3Sxo97XXUs9hwbpPLaG54fmsrNeWznR+qam2Sa4CDe4758hiJJoCVnURT\nq3Pz9GRgGc2NKMBFm2jDtzuJprYd/5LkzrYdJpukrayqbklyN7Bd59qovSdbDFxUVZ17x39Ksivw\nP5O8r6pWJbkdeKg3+T2F1zO7Ay+rqm+0dd8A3EIz39T5wDU096sn0CSBoOmb5gF/11PXVzrXZO11\n3iuA3wf2r6pHaB4IvIwmedSdbNoR+NOq+kpb9naahNTvAFdv4UOBbYKv0U1vBwP/1Ek0AVTVt2hu\nyB57dz/Ji5N8I8lPaZ7sr2pD+/XUd10nidP6P8AeSR43kcYkeXKSi5Pc1n7Pz2kmm+v9np8B/9y1\nvaL98ytj7Ot993+L2ihpKN3WSTS1DqO5yLg3yYw086DcB9wALIBmVACwnPVfDfkj4KtV9ZN2+2Ca\nm6EfdA6oqlU0ye/15jfZDJ/t2b4CeHZ7I0qS3ZJ8oH3S//P2cyIb9oeSRsvuwOOAn/Ts72zP2gpt\nWNuO8Ox2J/Cknn29bez4z57tzkj4Hdo/ZwP3d19fbsSdG9nX2w5Jg/Nk4Ff4xYO4jsuAXYADxys8\nhdczd3YSTQBVdSvNdd7B7fajNEmlV+cXQ5hOAJZU1fd66vpyVz330ozK+mqbaOpYwYb3lT+jeUuo\nozvxrgkw2TS9PYmxLyR+Qnuxk+Q5wOdoEkyvonni9pvtcTv0lBvr4iPAEzbVkDZr/jngvwJ/SZN5\nfg7NqKTe77mv5/W9zkXOY9/f9Qpg39ooaWj19mO70ySOft7zORT41a7jLgP+II1dgMNZ/4nVJvvI\nLdB7k3UnzWji3dvti2jO4T00SffnAB9lwz5N0mi5i6Y/2qNnf2e+zLtp5p5bb6GDJLv1sQ0zs+Hq\ncnvQTB/QrTaz/jXATm2/Op7e38HG2iFpcDrJ381NkF/E1FzPTCRZ/Xc0r8IdmmRnmjd3PjpGubHu\nD8faN+496Tj3n9oIX6Ob3m5n7P/o96TJDEMzXHA18EdVVQBJ9p6CtswHfh14cfd7+mNcDElSr94b\nortpktdnjnHsfV0/X0bzrv9v08wZtx3NCKOO22le5e21Z/sdW6K3792DZkTnXUl2oJlQ979V1fmd\nA7onppQ0mqrqkfZ1jz8A/rYr9IfAozRTFTwB2DnJ3Kq6rY2/iA2NdfMzUS8HOnM2zQReSDOnSj90\nRpq/mvEnPX9Wkr265mx6Lk1f+K99aoekLddJ/o6XIB/TFF/PbCxZvbSzUVUrk1xLM6Kpc53nVARD\nxGTT9PYt4A1Jdq6q++CxkUzzaCbXheZd1J93Ek2tqVgRqZNUemxC7jap9VyaCb23FjPS0uj7Ms2N\n29IxXhV5TFUtTfI9miduvwZcW1Vrug75Fs3w6+7JJefSjMBcvIVtfDnrzyf3cuCG9kZ0Js0FUXd/\nuDPN/AGbO9JA0vBYRDNX29/RjKY8kCY5/v+3c59cTTPh7UeTnEvTP71+jHqWAS9tj18L/N/O9dwm\nPACc1fY1P6ZZHOXxwPu38LwAqKr/m+QCmgmE96CZo+WXgFdU1TFdh64GvpBkEc1119k0ry47X5M0\nOL1J7FU0/cQfsP51yx8C9wI3b6QcNInzqbqe2SPJf+2as2kvmvnweudjupBmNNMBwJVV1Ttiaapt\nyUOBac+nqNPbe9s//zHJy9IsQ3kFTadxeRu7BtgryfuSvCDJGTSz+PfbMprO7NwkL01yDPBPwG3j\nF+u7zgThJyX5jSTjvocsaSi9l+bG6StJjk3yvCR/mOSDSf6459jLaJJNL2TDlZIuolmh6Utt+aNp\nLrTuolnWe0u8OMlZSV6U5Pz2+/8KoKp+Cvwb8JdJjk7ycuBammXOJY24qvonmsVJFgBX0UwWfi7w\npjZ+F83rHk8GrgReSbNqcK//TrMq7xdo+oxnT7AJ62hGHb2R5npvN+AlVdXP19feSLM61CtpVhp+\nX/u93b5Bs2DD+2huCL8HHNnHNkiavGXAgUmOTLKAZlXyxcAJSd7TXrecBbwBeGfXIgLLgD2TnJBk\nQZJ5U3w9cxfw8fY67+U0C6zcyYYLE1xJ82ryWImoraHzUGBh+3vZeQBtGFqObJrGqmp1kkNpLnA+\nSZN5/SLwls47p1X1xST/AziZZtWAb9IMh1ze57Y8lOQomouOz9Akns4CFgJP7+d3baIdtyb5M+DN\nNOe8imakl6QRUVV3JflNmj7kr2meqN9OM2Kzd6Tk39OMKHiI5oKku56HkhxGk7y6kGZ+t+uBozur\n3G2B19LcYL6FZgj6f6uqz3XFj6VJaF1CM//J39AssfumLfxeSUOgqi5jwwR3d/xLrD+KAJo+qPuY\nG/jFPJqT/f5/Bp45TnzeRvafMMa+lWO07RGaBPpfbaId59OsHCVpOHyIZmqTj9Ikov9XVS1uX4k7\npf2sAt5WVX/dVe5TNHNjvhuYA1xM8/raVF3P3ErTv7yLZl6mJcCxvStottdyX6JZIe7aLfzOzfHf\nae5vv0Bz3oey/qTi27Ss//aUJEmSpFGUZDHwpqrafVPHTnE7rgfuqqpXDLIdkkZPkouAp1fVggkc\nO4MmMfXRqjpjqtumyXFkkyRJkjTk2kl3x5sC45FxYpI0bSR5PHAQzciq2Wz59AeaAiab1BdJtqdn\niHW3qnp4KzZHkvouSYDtxznk0e4lciWpz/6SZvLxjTm0qhaz5QscbLGqWjjoNkgaPptKmk/invFX\naFa2vBM4qapW9aF56jNfo1NftMOln7exeFVtNBElSaMgyULgunEO+V/tjZ4k9V2SX6G5wdqYia5W\nJ0kD0b4iN95iVL/WzhOnacBkk/oiyVOBjc6+X1VLtmJzJKnv2hVGnjrOIT+uqh9vrfZIkiSNkiTz\ngPHmlPtuZyErjT6TTZIkSZIkSeqb8SYZlCRJkiRJkibFZJMkSZIkSZL6xmSTJEmSJEmS+sZkkyRJ\nkiRJkvrGZJMkSZIkSZL6xmSTJEmSJEmS+sZkkyRJkiRJkvrGZJMkSZIkSZL6xmSTJEmSJEmS+sZk\nkyRJkiRJkvrGZJMkSZIkSZL6xmSTJEmSJEmS+mZSyaYkT0nyYJKPd+17QZJlSdYluS7J3l2xJDk7\nyZr2c3aSdMXntWXWtXUc1vN9xya5Ncn9Sa5MMmtLTlaSJEmSJElTa8Ykj/8g8G+djSS7A1cArwWu\nAs4ELgN+sz3kROBI4CCggGuAHwLnt/FPAt8EXtJ+PpPkKVW1OskBwIeBlwLfBi4APgQcM14Dd999\n95o3b94kT0va9txwww13VdWcQbdD67MPkybGPmz42H9JE2P/NXzsv6SJmUz/NeFkU5JjgP8EvgHM\nb3cfBSytqk+3xywG7kqyf1UtA44Hzq2qVW38HJoE1PlJ9gOeBbyoqh4ALk9yCnA0TTLqOOCqqvpa\nW/YM4PtJdq6q+zbWznnz5rFkyZKJnpa0zUpy66DboA3Zh0kTYx/2/9i7/zirqvve/683M4kgKAkj\noRaj1B9o4g9MQpLbe2NDFCJJaupDcm+JRsZ+04vR64+q/dp8v5FC1NSYavqI3vyQWw1Dqn6NiklR\n79yA1eba5qahKhISQomCRUXH0QgDgpnh8/1jr4ObwzkzZ4aZ2ecw7+fjsR/MXmuvfT574CzO+ey1\n16o/7r/MauP+q/64/zKrTX/6r5oeo5N0KHAtcGVZ1YnA6tJORGwHNqTyferTz/m6Z8oSR+X1+XP/\nGtgFTK0lZjMzM7NGJ+nvJG2RtFXSekl/mqvzVAZmZmZWl2qds+k64PbSCKWcccDrZWVbgUOq1G8F\nxqUPO/1tW16/h6T5klZJWtXR0VHD5ZiZmZk1hK8CR0fEocCngeslfSA3lcECYAKwimwqg5L8VAan\nAGcBF+bq7waeBFqAL5FNZTARIDeVwfnAJGAH2VQGZmZmZjXpM9kk6VRgJvA3Faq7gEPLysYD26rU\njwe6IiIG0La8fo+IWBwR0yNi+sSJfvy53lx22WXMmDGDK68sHxhnZlbfOjs7ueyyy+js7Cw6FBuh\nIuLnEbGjtJu2Y8hNZRARO4FFwDRJJ6Rj90xlEBHPAzcBFwDkpjJYGBFvRMT9wNNkUxlAbiqDiOgi\nS2idI2mfG35W39yHmZlZUWoZ2TQDmAI8J2kL8OfAHElPAGvJ7pgBIGks2Qegtalor/r0c77u6LIP\nLuX1+XMfA7wdWF9DzFZHnn76aQCeeOKJgiMxM+uftrY21qxZw9KlS4sOxUYwSd+StANYB7wIPIyn\nMrAauA8zM7Oi1JJsWkyWQDo1bd8BHgLOBB4ATpI0R9JoYCGwOk0ODrAUuFLSZEmTgauAJQARsR54\nClgoabSkc4CTgftT2zuBsySdlpJY1wHLepsc3OrPZZddtte+RzeZWaPo7Oykvb2diKC9vd0jA6ww\nEXEx2TQCp5E9OrcLT2VgfXAfZmZmReoz2RQROyJiS2kje7xtZ0R0REQH2ZDrrwCvAR8C5uaa3wYs\nB9ak7cFUVjIXmJ7a3gB8Jp2TiFgLfIEs6fQyMBa4eD+u1QpQGtVU4tFNZtYo2tra2L17NwA9PT0e\nGWCFioieiHgcOAK4CE9lYH1wH2ZmZkWqdYLwPSJiUUR8Lre/MiJOiIgxETEjIjbm6iIiro6ICWm7\nOn3IKdVvTG3GRMTxEbGy7LXuiogjI2JsRPxRRLw6wOs0MzPrl5UrV9Ld3Q1Ad3c3K1asKDgiMwCa\neWvKAk9lYFW5DzMzsyL1O9lkZmY2EsycOZPm5mYAmpubmTVrVsER2Ugj6V2S5koaJ6lJ0pnAZ4FH\n8FQG1gf3YWZmViQnm8zMzCpobW1l1Kjsv8mmpibmzZtXcEQ2AgXZI3ObyaYcuAn4s4j4e09lYH1x\nH2ZmZkVqLjoAMzOzetTS0sLs2bNZvnw5s2fPpqWlpeiQbIRJyZ+P9lK/EjihSl0AV6etUv1GshWH\nq537LuCu2qO1euM+zMzMiuRkk5mZWRWtra1s3LjRIwLMrCG5DzMzs6I42WRmZlZFS0sLt9xyS9Fh\nmJkNiPswMzMriudsMjMzMzMzMzOzQeNkk5mZmZmZmZmZDRonm8zMzMzMzMzMbNA42WRmZmZmZmZm\nZoPGySYzMzMzMzMzMxs0TjaZmVUh6RJJqyTtkrQkVz5FUkjqym0LcvWSdKOkzrTdKEmFXISZmZmZ\nmdkwc7LJzKy6F4DrgTuq1L8jIsal7bpc+XzgbGAacApwFnDhkEZqZmZm1sAkTZD0gKTtkjZJOreX\nY6+QtEXSVkl3SDqowjHHSdop6e+GNnIzq8TJJjOzKiJiWUT8AOjsZ9NW4OaI2BwRzwM3ARcMdnxm\nZmZmB5BvAm8Ck4DzgG9LOrH8IElnAl8EzgCOAo4GvlzlfD8bsmjNrFdONpmZDdwmSZslfVfSYbny\nE4HVuf3VqawiSfPT43qrOjo6hipWG4DOzk4uu+wyOjv7m280MzOzWkkaC8wBFkREV0Q8DvwQOL/C\n4a3A7RGxNiJeA66l7KaepLnAb4BHhjRwM6vKySYzs/57Bfgg2d20DwCHAHfm6scBr+f2twLjqs3b\nFBGLI2J6REyfOHHiEIVsA9HW1saaNWtYunRp0aGYmZkdyKYC3RGxPldW7WZdpZt6kyS1AEg6lCwB\ndWVvL+ibfWZDy8kmM7N+SnfcVkVEd0S8BFwCfFzSIemQLuDQXJPxQFdExHDHagPX2dlJe3s7EUF7\ne7tHN5mZmQ2dcWQ35/K2kt3Qq3Rs+U09csdeRzbyaXNvL+ibfWZDy8kmM7P9V0oilfrUtWSTg5dM\nS2XWQNra2ti9ezcAPT09Ht1kZmY2dMpv1EF2s25bDceOT39uk3QqMBP4m0GP0Mz6xckmM7MqJDVL\nGg00AU2SRqeyD0s6XtKoNGT7FuCxiCjdZVsKXClpsqTJwFXAkkIuwgZs5cqVdHd3A9Dd3c2KFSsK\njsjMzOyAtR5olnRcrqzazbpKN/VeiohOYAYwBXhO0hbgz4E5kp4YiqDNrDonm8zMqrsGeINsxZPP\npVqtlAYAACAASURBVJ+vIVv1pJ3sbtvPgV3AZ3PtbgOWA2vS9mAqswYyc+ZMmpubAWhubmbWrFkF\nR2RmZnZgiojtwDLgWkljJX0E+DTwvQqHLwU+L+m9kt4JLOCtm3qLgWOAU9P2HeAh4MyhvQIzK+dk\nk5lZFRGxKCJUti2KiLsj4vciYmxEHB4R8yJiS65dRMTVETEhbVd7vqbG09rayqhR2X+TTU1NzJs3\nr+CIbKSRdJCk2yVtkrRN0lOSPpHqpkgKSV25bUGurSTdKKkzbTfmFylI7R+VtEPSOkkzy1773PS6\n2yX9QNKE4btyMxuhLgbGAC8DdwEXRcRaSUemPu5IgIhoB74GPApsAp4FFqa6HRGxpbSRPXK3MyI8\nA7jZMGsuOgAzM7N61NLSwuzZs1m+fDmzZ8+mpaWl6JBs5GkG/h34KPAc8Eng+5JOzh3zjojortB2\nPnA22eMlAawg+0L2nVR/N/CTdM5PAvdJOi4iOiSdSDYa81PAE2QjBb4FzB3cyzMze0tEvErWb5WX\nP0c2KXi+7OvA12s456LBis/M+scjm8zMzKpobW3l5JNP9qgmK0REbE+jKTdGxO6IeJAsYfSBGpq3\nAjdHxOaIeB64CbgAQNJU4P3Awoh4IyLuB54G5qS25wHLI+LHEdFF9ojKObkVN83MzMx65WSTmZlZ\nFS0tLdxyyy0e1WR1QdIkYCp7T5i7SdJmSd+VdFiu/ERgdW5/dSor1T0TEdt6qd/TNiJ+TTY33dRB\nuRAzMzM74DnZZGZmZlbnJL0NuBNoi4h1wCvAB4GjyEY6HZLqS8YBr+f2twLj0rxN5XWl+kOqtC2v\nz8c1X9IqSas6OjwlipmZmWWcbDIzMzOrY5JGka3I9CZwCUBEdEXEqojojoiXUvnHc4+6dQGH5k4z\nHuhKixWU15Xqt1VpW16/R0QsjojpETF94sSJA75GMzMzO7A42WRmZmZWp9JIpNuBScCciPhtlUNL\nK16WPtutJZscvGQabz1+txY4umwOpvL6PW0lHQO8HVg/wMswMzOzEcbJJjMzM7P69W3gPcBZEfFG\nqVDShyUdL2mUpBbgFuCxiCg9/rYUuFLSZEmTgauAJQARsR54ClgoabSkc4CTgftT2zuBsySdJmks\ncB2wrGyOJzMzM7OqmosOwMzMzMz2Jeko4EKyybm3ZIOcIJXtBv4KeBfZfEorgM/mmt8GHA2sSft/\nm8pK5pIln14DngM+ExEdABGxVtIXyJJOLcBK4E8G9+rMzMzsQOZkk5mZWRWdnZ18+ctfZuHChV6R\nzoZdRGwC1Mshd/fSNoCr01apfiMwo5f2dwF31RKnmZmZWTknmxrcrbfeyoYNG4oOo18uv/zyokPY\nx7HHHsull15adBhmVmfa2tpYs2YNS5cu5Yorrig6HDMzMzOzhlDTnE2S/k7SFklbJa2X9Ke5ujMk\nrZO0Q9Kjach3qU6SbpTUmbYblRsDLmlKarMjnWNm2eueK2mTpO2SfiBpwmBctJmZWV86Oztpb28n\nImhvb6ezs7PokMzMzMzMGkKtI5u+CsyPiB2STgAek/QksAlYBvwpsJxsAsl7gP+Q2s0HziZb0STI\n5hN4FvhOqr8b+AnwybTdJ+m4iOiQdCLZ3AKfAp4AFgPfIptjwJJ6H40zY8aMfcq+8Y1vDH8gZmb9\n1NbWxu7duwHo6enx6CYzMzMzsxrVNLIpIn4eETtKu2k7BjgHWBsR90bETmARMC0lpABagZsjYnNE\nPA/cBFwAIGkq8H5gYUS8ERH3A08Dc1Lb84DlEfHjiOgCFgDnlC3Ta3Vu7Nixe+0fcoj/+sysMaxc\nuZLu7m4Auru7WbFiRcERmZmZmZk1hpqSTQCSviVpB7AOeBF4GDgRWF06JiK2AxtSOeX16ed83TNl\ny+iW1+fP/Wuy1Vim1hqzFe+hhx7aa3/58uUFRWJm1j8zZ86kqakJgKamJmbNmlVwRGZmZmZmjaHm\nZFNEXAwcApxG9ujcLmAc8HrZoVvTcVSo3wqMS/M29bdtef0ekuZLWiVpVUdHR62XZMPMo5rMrJG0\ntraSLegFEcG8efMKjsjMzMzMrDHUnGwCiIieiHgcOAK4COgCDi07bDxQGq1UXj8e6ErL8fa3bXl9\nPq7FETE9IqZPnDixP5dkw2DatGlMmzbNo5rMrOHkk01mZmZmZlabfiWbcprJ5mxaSzb5NwCSxubK\nKa9PP+frji6bg6m8Pn/uY4C3A+sHGLOZmVnN2tra9ko2LV26tOCIzMzMzMwaQ5/JJknvkjRX0jhJ\nTZLOBD4LPAI8AJwkaY6k0cBCYHVErEvNlwJXSposaTJwFbAEICLWA08BCyWNlnQOcDJwf2p7J3CW\npNNSEus6YFnZHE9mZmZDonxC8B/96EcFRWJmZmZm1lhqGdkUZI/MbQZeI1tR7s8i4u8jooNs9biv\npLoPAXNzbW8DlgNr0vZgKiuZC0xPbW8APpPOSUSsBb5AlnR6GRgLXDygqzQzM+unww47rNd9MzMz\nMzOrrLmvA1Ly56O91K8ETqhSF8DVaatUvxGY0cu57wLu6itGMzOzwfbCCy/0um9mZmZmZpUNdM4m\nMzOzA9ru3bt73TczMzMzs8qcbDIzM6ugfAU6r0hnZmZmZlYbJ5vMzMzMzMysUJImSHpA0nZJmySd\n28uxV0jaImmrpDskHZTKD5J0e2q/TdJTkj4xfFdhZiVONpmZmVXQ1NTU676ZmZkNqm8CbwKTgPOA\nb0s6sfygtDr6F4EzgKOAo4Evp+pm4N/J5hweD1wDfF/SlCGO3czKONlkZmZWwamnnrrX/vve976C\nIjEzMzuwSRpLtsr5gojoiojHgR8C51c4vBW4PSLWRsRrwLXABQARsT0iFkXExojYHREPAs8CHxiW\nCzGzPZxsMjMzq+CXv/zlXvu/+MUvCorEzMzsgDcV6I6I9bmy1cA+I5tS2eqy4yZJaik/UNKkdO61\nFermS1olaVVHR8d+BW9m+3KyyczMrIKZM2cyalT23+SoUaOYNWtWwRHZSNPX3COSzpC0TtIOSY9K\nOipXJ0k3SupM242SlKufktrsSOeYWfba56bX3S7pB5ImDM9Vm9kINQ7YWla2FTikyrGvlx1H+bGS\n3gbcCbRFxLryk0TE4oiYHhHTJ06cOODAzawyJ5vMzMwqaG1t3Wt/3rx5BUViI1jVuUckHQYsAxYA\nE4BVwD25tvOBs4FpwCnAWcCFufq7gSeBFuBLwH2SJgKkOVJuI3t8ZRKwA/jW0FyimRkAXcChZWXj\ngW01HDs+/bnnWEmjgO+RzQF1yeCFaWa1crLJzMzMrA71MffIOcDaiLg3InYCi4Bpkk5IzVuBmyNi\nc0Q8D9xEmtNE0lTg/cDCiHgjIu4HniabLwWyiXmXR8SPI6KLLKF1jqRKIwzMzAbDeqBZ0nG5smlU\nePwtlU0rO+6liOiEbGQncDtZsnxORPx2aEI2s9442WRmZlZBW1sbpaeOJLF06dKCI7KRrmzukb3m\nLImI7cAG3prfpNKcJvm6ZyJiWy/1+XP/GtiVXrs8Js95Ymb7LfVhy4BrJY2V9BHg02Sjk8otBT4v\n6b2S3kmWEF+Sq/828B7grIh4Y2gjN7NqnGwyMzOrYOXKlfT09ADQ09PDihUrCo7IRrIKc4+Uz1kC\ne89vUmlOk3Hpjn9/25bX7+E5T8xsEF0MjAFeBu4CLoqItZKOlNQl6UiAiGgHvgY8CmwiG/G5ECDN\nXXchcCqwJbXrknTe8F+O2cjWXHQAZmZm9WjmzJk8/PDDdHd309zc7AnCrTBV5h7pa36TSnOadEVE\nSOpv2/J6M7NBFxGvks01V17+HFkSPF/2deDrFY7dBKi83MyGn0c2mZmZVZCfIFySJwi3QvQy98he\nc5ZIGgscw1vzm1Sa0yRfd3TZHEzl9flzHwO8nWxOFTMzM7M+OdlkZmZWQUtLC6NHjwbgoIMOoqWl\npeCIbISqNvfIA8BJkuZIGk32CMnq3PLeS4ErJU2WNBm4ijSnSUSsB54CFkoaLekc4GTg/tT2TuAs\nSaelJNZ1wLKyOZ7MzMzMqnKyyczMrIINGzbQ1dUFQFdXFxs2bCg4Ihtpept7JCI6yFaP+wrwGvAh\nYG6u+W3AcmBN2h5MZSVzgemp7Q3AZ9I5iYi1wBfIkk4vA2PJ5lIxMzMzq4nnbDIzM6vg+uuv32d/\nyZIlxQRjI1Jfc49ExErghCp1AVydtkr1G4EZvZz7LrIJes3MzMz6zSObzMzMKti4cWOv+2ZmZmZm\nVpmTTWZmVUi6RNIqSbskLSmrO0PSOkk7JD2aHncp1UnSjZI603ZjmuTXGsiUKVN63TczMzMzs8qc\nbDIzq+4F4HrgjnyhpMOAZcACYAKwCrgnd8h8sqV7pwGnAGeRzbtiDeSaa67pdd/MzMzMzCpzssnM\nrIqIWBYRPwA6y6rOAdZGxL0RsRNYBEyTVJo7pRW4OSI2R8TzwE3ABcMUtg2S3/zmN3vtv/766wVF\nYmZmZmbWWJxsMjPrvxOB1aWdiNgObEjl+9Snn0+kCknz0+N6qzo6OoYgXBuIRYsW7bW/cOHCYgIx\nMzMzM2swTjaZmfXfOKB8mMtW4JAq9VuBcdXmbYqIxRExPSKmT5w4cdCDtYHp6urqdd/MzMzMzCpz\nssnMrP+6gEPLysYD26rUjwe60lLk1iAOPvjgXvfNzMzMzKwyJ5vMzPpvLdnk3wBIGgsck8r3qU8/\nr8Uaynve85699t/73vcWFImZmZmZWWNxssnMrApJzZJGA01Ak6TRkpqBB4CTJM1J9QuB1RGxLjVd\nClwpabKkycBVwJICLsH2w1NPPbXX/pNPPllQJGZmZmZmjcXJJjOz6q4B3gC+CHwu/XxNRHQAc4Cv\nAK8BHwLm5trdBiwH1qTtwVRmDWT37t297puZmZmZWWXNRQdgZlavImIRsKhK3UrghCp1AVydNmtQ\nkshPs1VlfnczMzMzMyvjkU1mZmYV/O7v/m6v+2Zm9a6zs5PLLruMzs7OokMxM7MRxskmMzOzCjo6\nOnrdNzOrd21tbaxZs4alS5cWHYqZmY0wTjaZmZlV0Nzc3Ou+mVk96+zspL29nYigvb3do5vMzGxY\nOdlkZmZWwfbt23vdNzOrZ21tbXsWNujp6fHoJjMzG1Z9JpskHSTpdkmbJG2T9JSkT+Tqz5C0TtIO\nSY9KOipXJ0k3SupM243KzbAqaUpqsyOdY2bZa5+bXne7pB9ImjBYF25mZmZmdqBauXIl3d3dAHR3\nd7NixYqCIzIzs5GklpFNzcC/Ax8FxpMtBf79lCg6DFgGLAAmAKuAe3Jt5wNnA9OAU4CzgAtz9XcD\nTwItwJeA+yRNBJB0ItlS4ecDk4AdwLcGdJVmZmZmDUbSJZJWSdolaUmufIqkkNSV2xbk6n2zz5g5\nc+aex3+bm5uZNWtWwRGZmdlI0meyKSK2R8SiiNgYEbsj4kHgWeADwDnA2oi4NyJ2ki0RPk1SaTnw\nVuDmiNgcEc8DNwEXAEiaCrwfWBgRb0TE/cDTwJzU9jxgeUT8OCK6yBJa50g6ZHAu3czMzKyuvQBc\nD9xRpf4dETEubdflyn2zz2htbWXUqOyjflNTE/PmzSs4IrPeSZog6YGU6N4k6dxejr1C0hZJWyXd\nIemggZzHzIZOv+dskjQJmAqsBU4EVpfqImI7sCGVU16ffs7XPRMR23qpz5/718Cu9NrlMc1Pd/5W\nebUgMzMbDAcffHCv+2ZDLSKWRcQPgP7O7OybfUZLSwuzZ89GErNnz6alpaXokMz68k3gTbJE93nA\nt1MCfC+SzgS+CJwBHAUcDXy5v+cxs6HVr2STpLcBdwJtEbEOGAe8XnbYVqD0gaS8fiswLg3l7m/b\n8vo9ImJxREyPiOkTJ07szyWZmZlVtGvXrl73zerAJkmbJX03TW1QMiw3+6z+tba2cvLJJ3tUk9U9\nSWPJkt4LIqIrIh4Hfkg2yrJcK3B7RKyNiNeAa3krod6f85jZEKp5HWdJo4DvkWWJL0nFXcChZYeO\nB7ZVqR8PdEVESOpv2/J6MzOzIVNaxanavlmBXgE+CDxF9ijcN8luBp6Z6gdys29ylbal+oojmyTN\nJ3tsjyOPPHIAl2JDqaWlhVtuuaXoMMxqMRXojoj1ubLVwIwKx55IlkDKHzdJUgtwZD/OM2RuvfVW\n2tvbh/Ml+2XHjh1ERNFhNDxJdT/yffbs2Vx66aWFvHZNI5vSh5PbyYYizomI36aqtWTzAZSOGwsc\nk8r3qU8/5+uOLhuWXV6fP/cxwNuBfMdhZmZmNqKku/WrIqI7Il4iuwn48dxnqqo3+yrUleoHdLPP\no8vNbJCMI0ts51VLdFdKqJOOrfk8norFbGjVOrLp28B7gJkR8Uau/AHgryXNAR4CFgKr0yN2AEuB\nKyU9nPavAm4BiIj1kp4CFkq6BvgkcDJwfzr2TuAnkk4DngCuA5aVDfs2MzMbEr/zO7/Diy++uGf/\n8MMPLzAas16Vbk+XbiKWbtj9S9qveLMv95lqGtnnrnxbwDf7zGzY9CfRXSmhTjq25vNExGJgMcD0\n6dMHdZjPpZdeWthoErN60efIJklHka1gciqwJbfE7nkR0UH2TOxXgNeADwFzc81vA5YDa9L2YCor\nmQtMT21vAD6TzklErAW+QPbh52VgLHDxwC/VzMysdi+99NJe+1u2bCkoEhupJDVLGg00AU2SRqey\nD0s6XtKo9NjILcBjEVG601+62TdZ0mSym31LILvZR/b43cJ0vnPY92bfWZJOSyPWfbPPzIbDeqBZ\n0nG5snyiPK/S0zMvRURnP89jZkOoz5FNEbEJUC/1K4ETqtQFcHXaKtVvpJfnZyPiLuCuvmI0MzMb\nbJ6zyerANWSjxks+R7bi0q+AvwLeRfZ4yArgs7njbiNbnWlN2v9b9r3Zt4TsZt9zlN3sk1S62dcC\nrAT+ZDAvysysXERsl7QMuFbSnwLvAz4N/McKhy8Flki6E3iRbNXMJQM4j5kNoZonCDczMxtJmpqa\n6Onp2WvfbDhFxCJgUZXqu3tp55t9ZtaILgbuIHuqpRO4KCXAjwR+Abw3Ip6LiHZJXwMeBcaQjcxc\n2Nd5hvE6zAwnm8zMzCqaOnUqv/zlL/fsH3/88QVGY2ZmdmCLiFeBsyuUP0c28Xe+7OvA1/tzHjMb\nXjWtRmdmZjbS5BNNAL/4xS8KisTMzMzMrLE42WRmZmZmZmZmZoPGySYzMzMzMzMzMxs0TjaZmZmZ\nmR2AOjs7ueyyy+js7Cw6FDMzG2GcbDIzMzMzOwC1tbWxZs0ali5dWnQoZmY2wjjZZGZmVkFTU1Ov\n+2Zm9ayzs5P29nYigvb2do9uMjOzYeVkk5mZWQU9PT297puZ1bO2tjZ2794NZP2XRzeZmdlwcrLJ\nzMzMzOwAs3LlSrq7uwHo7u5mxYoVBUdkZmYjiZNNZmZmFYwePXqv/TFjxhQUiZlZ/82cOZPm5mYA\nmpubmTVrVsERmZnZSNJcdAD17tZbb2XDhg1Fh9HQSr+/yy+/vOBIGtuxxx7LpZdeWnQYZiPGrl27\n9trfuXNnQZGYmfVfa2srDz74IAC7d+9m3rx5BUdkZmYjiZNNfdiwYQNP/fyX9Bw8oehQGtaoNwOA\nf33mpYIjaVxNO14tOgSzESciet03MzMzM7PKnGyqQc/BE3jjhE8WHYaNYGPWPVx0CGZmZtZA2tra\nkASAJJYuXcoVV1xRcFRmZjZSeM4mMzMzM7MDzMqVK/esotnT0+MJws3MbFg52WRmZmZmdoDxBOFm\nZlYkJ5vMzMzMzA4wra2tjBqVfdRvamryBOFmZjasnGwyMzMzq0OSLpG0StIuSUvK6s6QtE7SDkmP\nSjoqVydJN0rqTNuNKk3ek9VPSW12pHPMLDv3uZI2Sdou6QeSvEpKA2ppaWH27NlIYvbs2bS0tBQd\nkpmZjSBONpmZmZnVpxeA64E78oWSDgOWAQuACcAq4J7cIfOBs4FpwCnAWcCFufq7gSeBFuBLwH2S\nJqZznwjcBpwPTAJ2AN8a5OuyYdLa2srJJ5/sUU1mZjbsnGwyMzMzq0MRsSwifgB0llWdA6yNiHsj\nYiewCJgm6YRU3wrcHBGbI+J54CbgAgBJU4H3Awsj4o2IuB94GpiT2p4HLI+IH0dEF1lC6xxJhwzZ\nhdqQaWlp4ZZbbvGoJjMzG3ZONpmZmZk1lhOB1aWdiNgObEjl+9Snn/N1z0TEtl7q8+f+NbALmDqI\n8ZuZmdkBzskmMzMzs8YyDni9rGwrcEiV+q3AuDRvU3/bltfvRdL8NK/Uqo6Ojn5dhJmZmR24nGwy\nMzMzayxdwKFlZeOBbVXqxwNdEREDaFtev5eIWBwR0yNi+sSJE/t1EWZmZnbgcrLJzGyAJD0maaek\nrrT9KldXdaUoM7P9tJZs8m8AJI0Fjknl+9Snn/N1R5fNwVRenz/3McDbgfWDGL+Z2R6SJkh6IK2A\nuUnSuX0cf4WkLZK2SrpD0kGp/CBJt6dzbJP0lKRPDM9VmFk5J5vMzPbPJRExLm3HQ00rRZmZ9UlS\ns6TRQBPQJGm0pGbgAeAkSXNS/UJgdUSsS02XAldKmixpMnAVsAQgItYDTwEL0/nOAU4G7k9t7wTO\nknRaSmJdBywrm+PJzGwwfRN4k2wFzPOAb6eVMfch6Uzgi8AZwFHA0cCXU3Uz8O/AR8lGZF4DfF/S\nlCGM3cyqcLLJzGzw9bVSlJlZLa4B3iD7YvW59PM1EdFBtnrcV4DXgA8Bc3PtbgOWA2vS9mAqK5kL\nTE9tbwA+k85JRKwFvkCWdHoZGAtcPDSXZ2YjXUpqzwEWRERXRDwO/BA4v0qTVuD2iFgbEa8B15JW\n24yI7RGxKCI2RsTuiHgQeBb4wJBfiJnto7noAMzMGtwNkr4K/Ar4UkQ8RoWVoiSVVopaV/EsVnea\nmpro6enZa99sOEXEIrJkdaW6lUDFBHaam+nqtFWq3wjM6OV17wLu6k+sZmYDNBXoTqMuS1ZTvY86\nkSwZlT92kqSWiOjMHyhpUjr/Wsxs2Hlkk5nZwP0F2fDtycBiYHma38SrOZmZWeE6Ozu57LLL6Ozs\n7Ptgs2KMI/uMlFf1MxOVV9uk/HhJbyMbodmWe8SYsmP8+ctsCDnZZGY2QBHx04jYFhG7IqIN+Cfg\nk3g1pwPC7t27e903M6t3bW1trFmzhqVLlxYdio1QaTGVqLI9Tj8/M1U4fnz6c8/xkkYB3yObB+qS\narH585fZ0HKyycxs8AQg+l4pyhpA9iRS9X0zs3rW2dlJe3s7EUF7e7tHN1khImJGRKjK9hGylS6b\nJR2Xa5ZfIbNcpdU2Xyo9QidJwO1kk43PiYjfDvpFmVlNako2SbokDTHcJWlJWV3V5b2VuVFSZ9pu\nTB1AqX5KarMjnWNm2bnPTUtXbpf0A0kT9vN6zcwGhaR3SDqztDqUpPOAPwDa6XulKDMzsyHV1ta2\nZ0RmT0+PRzdZXYqI7WQr+F4raaykjwCfJhuZVMlS4POS3ivpnWQr/y7J1X8beA9wVkS8MXSRm1lf\nah3Z9AJwPXBHvrCG5b3nA2eTZZxPAc4CLszV3w08CbQAXwLukzQxnftEspVTzifLTO8AvlX7pZmZ\nDam3kfWLHcArwKXA2RGxvoaVoszMzIbUypUr6e7uBqC7u5sVK1YUHJFZVRcDY8hWwLwLuCitjImk\nIyV1SToSICLaga8BjwKbyFabW5iOPYrsu+apwJbUrivdEDSzYVbTanQRsQxA0nTgiFzVnuW9U/0i\n4BVJJ6Q7+K3AzRGxOdXfRJaA+o6kqcD7gY+nrPP9ki4n+4L2HeA8YHlE/Di1XQD8UtIhEVHtGV4z\ns2GREkof7KW+6kpRZmZmQ23mzJk8/PDDdHd309zczKxZs4oOyayiiHiVbIBCpbrnyCYFz5d9Hfh6\nhWM3kU1nYGZ1YH/nbNpneW+gtLz3PvXp53zdM2WJo/L6/Ll/DewiW75yL15JwMzMzMzsLa2trYwa\nlX3Ub2pqYt68eQVHZGZmI8n+Jpv6Wt670tKU49K8Tf1tW16/h1cSMDMzMzN7S0tLC7Nnz0YSs2fP\npqWlpeiQzMxsBKnpMbpe9LVUZaWlKbsiIiT1t215vZmZmZmZVdHa2srGjRs9qsnMzIbd/o5s6mt5\n70pLU+brjpZ0SC/1+XMfA7ydbHlMMzMzMzPrRUtLC7fccotHNZmZ2bCraWSTpOZ0bBPQlJby7iZb\n3vuvJc0BHmLf5b2XAldKejjtXwXcAhAR6yU9BSyUdA3wSeBk4P507J3ATySdBjwBXAcsG+7JwZ9/\n/nmadrzOmHUP932w2RBp2tHJ8893Fx2GmZmZmZmZWZ9qHdl0DfAG8EXgc+nna2pY3vs2YDmwJm0P\nprKSucD01PYG4DPpnKTlLr9AlnR6GRhLtiymmZmZmZmZmZnVqZpGNkXEImBRlbqqy3tHRABXp61S\n/UZgRi+vexdwVy0xDpXJkyezZVczb5zwySLDsBFuzLqHmTx5UtFhmJmZmZmZmfVpf+dsMjMzMzMz\nMzMz28PJJjMzMzMzMzMzGzRONpmZmZmZmZmZ2aBxssnMzMzMzMzMzAaNk01mZmZmDUjSY5J2SupK\n269ydWdIWidph6RHJR2Vq5OkGyV1pu1GScrVT0ltdqRzzBzuazMzM7PG5mSTmZmZWeO6JCLGpe14\nAEmHAcuABcAEYBVwT67NfOBsYBpwCnAWcGGu/m7gSaAF+BJwn6SJQ30hNvg2bNjApz71KTZs2FB0\nKGZmNsI42WRmZmZ2YDkHWBsR90bETmARME3SCam+Fbg5IjZHxPPATcAFAJKmAu8HFkbEGxFxP/A0\nMGeYr8EGwfXXX8/27du5/vrriw7FzMxGGCebzMzMzBrXDZJekfRPkmakshOB1aUDImI7sCGV71Of\nfs7XPRMR26rU70XSfEmrJK3q6OjY74uxwbNhwwY2btwIwMaNGz26yczMhpWTTWZmZmaN6S+ACncB\ngAAAIABJREFUo4HJwGJguaRjgHHA62XHbgUOST+X128FxqV5m/pqu5eIWBwR0yNi+sSJftKunpSP\nZvLoJjMzG05ONpmZmZk1oIj4aURsi4hdEdEG/BPwSaALOLTs8PFAabRSef14oCsiooa21iBKo5qq\n7ZuZmQ0lJ5vMzMzMDgwBCFhLNvk3AJLGAsekcsrr08/5uqMlHVKl3hrEEUcc0eu+mZnZUHKyyczM\nzKzBSHqHpDMljZbULOk84A+AduAB4CRJcySNBhYCqyNiXWq+FLhS0mRJk4GrgCUAEbEeeApYmM59\nDnAycP+wXqDtt3e/+9177R955JEFRWJmZiNRc9EBmJmZmVm/vQ24HjgB6AHWAWenZBGS5gD/Hfg7\n4KfA3Fzb28jmelqT9v82lZXMJUs+vQY8B3wmIjz7d4P5l3/5l732f/rTnxYUiZmZjURONpmZmZk1\nmJT8+WAv9SvJElGV6gK4Om2V6jcCM/Y7SCvU7t27e903MzMbSk421aBpx6uMWfdw0WE0rFE7twKw\ne3T5fKNWq6YdrwKTig7DzMzMGoQksrziW/tmZmbDxcmmPhx77LFFh9DwNmzIFrA59mgnSwZukv8t\nmpmZWc1mzpzJj370oz37s2bNKjAas+okTQBuBz4OvAL8PxFxVy/HXwH8BXAwcB9wUUTsKjvmOLJH\nhe+LiM8NVexmVp2TTX249NJLiw6h4V1++eUAfOMb3yg4EjMzM7OR4cILL2TFihVEBJKYP39+0SGZ\nVfNN4E2yYfynAg9JWh0R+6yCKelM4IvA6cALZAsifDmVlZ/zZ0MZtJn1zqvRmZmZmZkdYFpaWjj8\n8MMBOPzww2lpaSk4IrN9SRoLzAEWRERXRDwO/BA4v0qTVuD2iFgbEa8B1wIXlJ1zLvAb4JEhC9zM\n+uRkk5mZmZnZAaazs5MXXngBgBdeeIHOzs6CIzKraCrQXVpJM1kNnFjl+BNTff7YSZJaACQdSpaA\nurKvF5Y0X9IqSas6Orzgptlgc7LJzMzMzOwAs3jx4l73zerEOGBrWdlW4JBejn+97Fhyx19HNvJp\nc18vHBGLI2J6REyfOHFiP0I2s1o42WRmZmZmdoBZuXJlr/tmw0HSY5KiyvY40AWUL1k9HthW5ZTl\nx49Pf26TdCowE/ibwbwGMxsYTxBuZmZmZnaA6enp6XXfbDhExIze6tOcTc2SjouIf0vF04B9JgdP\n1qb67+eOfSkiOiWdD0wBnpME2SioJknvjYj379eFmFm/eWSTmZmZmZmZDbuI2A4sA66VNFbSR4BP\nA9+r0mQp8HlJ75X0TmABsCTVLQaOIVvR7lTgO8BDwJlDdwVmVo2TTWZmZhVMmTKl130zMzMbFBcD\nY4CXgbuAiyJiLYCkIyV1SToSICLaga8BjwKbgGeBhaluR0RsKW1kj9ztjAjP/m1WAD9GZ2ZmVsHG\njRt73TczM7P9FxGvAmdXqXuO7HG4fNnXga/XcN5FgxGfmQ2MRzaZmZlV4JFNZmZmZmYD42STmZlZ\nBddcc02v+2Zm9eyUU07pdd/MzGwoOdlkZmZWwbHHHsvYsWMBGDt2LMcee2zBEZmZ1a6jY+9pal55\n5ZWCIjEzs5HIySYzM7MKOjs72b59OwDbt2+ns7Oz4IjMzGr34osv7rX/wgsvFBSJmZmNRE42mZmZ\nVXDbbbfttb948eKCIjEzMzMzayx1nWySNEHSA5K2S9ok6dyiYzIzq5X7sMb2yCOP7LW/cuXKgiIx\nG37uv8zMzGx/NBcdQB++CbwJTAJOBR6StDoi1hYblplZTdyHNbCenp5e980OcO6/zMzMbMDqdmST\npLHAHGBBRHRFxOPAD4Hzi43MzKxv7sPMrFG5/zIzM7P9Vc8jm6YC3RGxPle2GphRTDj16dZbb2XD\nhg1Fh9GrUnyXX355wZFUd+yxx3LppZcWHYYdWArvw2699Vba29uH6+UGZMeOHURE0WHUbMaMGUWH\nUJEkDj744KLDqGr27NnuYxuL+68aNFr/BfXZh9V7/wXuw8zMBqJuRzYB44CtZWVbgUPKD5Q0X9Iq\nSavKl3m14o0ZM4YxY8YUHYbZcHMfZmaNyv2XmZmZ7RfV6x0ZSe8D/ikiDs6V/Tnw0Yg4q1q76dOn\nx6pVq4YjRLOGJulfI2J60XEcqNyHNb6Pfexje41akMSjjz5aYESW5z5s6Lj/OjBUGsX02GOPDXsc\nti/3X/XH/ZdZbfrTf9XzyKb1QLOk43Jl0wBPTGlmjcB9WINbsGDBXvt/+Zd/WVAkZsPO/ZeZmZnt\nl7pNNkXEdmAZcK2ksZI+Anwa+F6xkZmZ9c19WOM7/fTTkQRko5o+9rGPFRyR2fBw/3VgKB/F5FFN\nZmY2nOo22ZRcDIwBXgbuAi7ykrtm1kDchzW40ugmj2qyEcj9l5mZmQ1YPa9GR0S8CpxddBxmZgPh\nPqzxnX766Zx++ulFh2E27Nx/HRg8msnMzIpS7yObzMzMzMzMzMysgdTtanQDJakD2FR0HLaPw4BX\nig7C9nJUREwsOgjbm/uwuuT+qz65D6sz7r/qlvuw+uP+q864/6pb7r/qT8391wGXbLL6JGmVl3g1\ns0bk/svMGpn7MDNrVO6/GpsfozMzMzMzMzMzs0HjZJOZmZmZmZmZmQ0aJ5tsuCwuOgAzswFy/2Vm\njcx9mJk1KvdfDcxzNpmZmZmZmZmZ2aDxyCYzMzMzMzMzMxs0TjaZmZmZmZmZmdmgcbLJzMzMzMzM\nzMwGjZNNZmZmZmZmZmY2aJxsMjMzMzMzMzOzQeNkk5mZmZmZmZmZDRonmwok6b9IumAA7T4u6c8G\n+JoXSApJ4wbSvp5JulrSjH62WSTplSEKyazhuZ/q3UBjHUh/1Y9z/ydJT0jaKSmG4jXMbOSRNCP1\ndycVHYvZgUTSlPTe+sNc2UZJNw3y67wrffeZUlbu93adkzRW0v8nqTP9XV1QdEy1cLKpWP8FuGAA\n7T4ODOhL3AHuamBGP9v8LXDm4IdidsBwPzU0BtJf1eo24DdkfdvvD9FrmNnI8wRZn/LrogMxswF5\nF7AQmFJW7vd2/bsIOAuYT/Z39VCx4dSmuegAzIoUEZuBzUXHYWaDR1IT0BQRbxYdS0FOABZHxD8W\nHYjZSHag9EWSBBwUEVuB/1N0PGY2uOrpvX2g9JuwV9+5cxBOdwLwq4i4fxDONWw8sqkgkpYAc4CP\npqFwIWlRqrtE0r9J2iVpg6Qrcu0WAVcBR+XaLUl1vy/p7yW9KGm7pKcknbe/cUpaJelsSevSYxmP\nS3pv7pjvS3qsQttFkl6S9Lbc8NC5kr4raaukzZI+l469WtILkjok3ShpVNl5XpH0Pkn/R9IOSU9K\nOi13zEagBViY+73MqOH69nqMLjeMdIakeyV1SXpG0sW5Yy6Q9Kakd5Sd68TUdmaNv16zutag/dRa\nYCfw4VR3ZBp2/GrqO/6XpOPL2o+R9DVJm9L1PCvphv2I56uS1qT+Y7OkOyX9Tq5+I1X6K0mjJH0x\n/U53SVovqbXG152h7LG5JuAb+d97qv9TSWvTeTdJurrCOU6T9I/pd9Up6X9IOmSgvwuzkaRaXzQY\n/ZCkpvSZ5blUv1bSubn6mj+bSPqjFOdOSVvS674tV1/63PURST9L1/GfVeFRm1r6rHSe/63ss9/W\n1O//5/3/jZvVD0l/IOnR9H//65Iek/S+VHeqpEfS+/+19Llg0gBeo8//oyUdJenu9B7eIelpSecq\ne3RuTTrs0dLnj9Sm0nv7YEm3pD5ip6SfSfp42Ws9Jum+dP4N6f39PyUd0Y9rGtLPcHXQd/6xsu+4\niypc+2OSHqjhd7QR+DzwvvzfW6o7SdJDkral7V7lPnOmYyZIWqzse/lOSf8s6cN9ve6giAhvBWzA\nMcA/kA1b/A9pOwL4r0AAN5M9hnIDsBv4Ymp3BHAn8GKu3TGp7rPA/wv8IXA6sAB4E/hs7nUvSOcf\nV2OcS4AO4BngPOAcso7q34HR6ZgzU4y/l2sn4Fng5rQ/Jb3uJuCvgFnA3UBPutb7gNnAl9Jxc3Pn\nWgTsAJ4G/gT4BFn2vQM4OB3zPrLHRv4293s5tIbrWwS8ktufkV7/34BrUpx3pLIPpWPeAewC/qTs\nXNcCW8iy8YX/G/PmbX+3BuunXgHWA58DZqYYJgDPAU+SPQ74h8Djqf8ak9oK+BGwjezRtjOAecD/\nqPG194k1xfO51J98BvgJ8AtgVKqv2l8B3wS6UiwzgRvJ+sk/rCGWQ9O5Arip7Pf+fwO/Bb6S+rUv\npn7sklz7/5TK7gE+CZwPPA/cV/S/RW/eGmGr0hcdORj9UHrv/pbss8mZwOL0Xv9sqq/ps0mKoQf4\nFln/fVHqj27KtVlE9rnr18CFwMeA43nrM9JJuWN77bNSv/QboC31PR8nuxnxX4v++/LmbbC29N74\nbXofzyH7TnNder9PTO+BnwBnp75hM9n3mren9lPSe+sPc+fcWPa+7PP/aLLH5F4ANpB9PjkDuBz4\nC+Ag4Nz0OheTPn/k4i9/b9+Z+qRLyb57LUvX+JHcMY+lvuyfgT8C5gIvAQ/343e3hCH8DEd99J03\nkH2XVu7Yo8k+O59Vw+/ofWSPzf2y7O/tWOB14JH0+59D9nnzZ6XXSn/vT6TXn0f2b/OH6Xf2O0P+\n3ij6zTmSN7IEy2O5/VFkncZ3y477VvqHVEru3ARs7OPcIntM8jbgH3LlF9D/L3EB/Mdc2VFAN/CF\nXNzPAV/OHXM6uU6LtzrR7+aOOTS9+f+NXIIG+Bfgntz+otT29FzZqalsdq7sFWBRP/8OFlE52XRt\nruxtZImtr+bKfgi0l53rV8B/L/rflTdvg7k1WD91aln5dUAnMCFX9s4U539L+2emtp8e4O+n11jJ\nRhlNTsf8Qa58n/6K7EPDbqC1rHwp8LN+xBTsnUQ6lOzL4MKy48o/SP1v4NGyY/bqy71581Z9q9QX\nDUY/RPala3uF9/DDZI9VlPZ7/WyS+txNFfrv/wt4A2hJ+4tSPH9UdtwM9v5s12efBUxPbQ4p+u/H\nm7eh2sgSSavIJRNydV8lS0ocmiv7MHsnPKbQd7Kpz/+jyZIa24HDq8R5Ujp+Rll5+Xv7PeXvbbLP\nfz8H/leu7LHUl70zV/Zn6Vxjavzd7dNvpvIDqe88LpV/LFdW+gzW3I/f06qysu+lON9e9lo9wKfS\n/ufJbuoelzummSwh9tdD/d7wY3T15Qjgd4F7y8rvIfuycHJvjSW9Mw133ESWxPkt2SRiU/czrpcj\n4p9LOxGxCfhX4ENpfzfwXWCeJKXDLiB7Q/y87FyP5M6zlSyJ848R0ZM7ZgPZl7O8N8k6tJJfpD9r\nHqbZTz8q/RARpYRY/rXuAc6Q1ALZ8Fiy3/M9QxSPWb2o137q+Yh4qqxsJrAC2CqpWVIz2Z2cfyX7\nAgTZB7VXI+Lv9/P195D0iTRE+XWyxHxpXri+rvEMsg93D5TiTTE/ApyqbB6Dgfh9YCxwb9l5/wGY\nBBwh6eB03PfLjnmc7O/oAwN8bbORprwvGox+6CTgYCr3u1MlTczt9/bZZCrZSKvy9/k/AKPT65QE\n8D/7uNZa+qxfkyW770qPoLyjl/OZNRxJY8mSR22RvsmX+RDwo/S9B4CI+ClZMukjNb5Grf9Hn06W\nNHlxoNeTfJAswbKnz0nf9+6tEPPPIuK13H7pO1r5d7neDNVnuLroOyPi34AfkxbcSd+X5wHfi4ju\nKrHXYibwALA7F9OzZP+2pueO+Vfg2dwxAP+YO2bIONlUXw5Pf75UVl7an9BH+yXAHwN/TTa874Nk\nj4CN3s+4Xq5Sdnhu/7tkI54+puzZ4Tnptcv9pmz/zSpl5TFvS50cAPHWpHH7e23V9BXT35N17nPS\n/h+TfaF8fIjiMasX9dpPlccDcFh6rd+WbR8D3p2OaSF73G9QSPogWf+wmWyI+++TDXmGvq/xMLKR\nUK+XxbuE7C7U4VVb9n1egLVl5300lb+b7G5hE9kItfwxu8hGd74bM6tFeV80GP1Qrf1uX59NSn3B\nw2WxPJvK8+/z16LvCXr77LPSl9BZZP3I94GONL/I0X2c26xRvJMsMVPtPXw4lT+jvETfn5nyr1HL\n/9GD9ZnmcKArInaUlb8EHCzpoFxZpe9M0L/PdUP1Ga6e+s7bgTmSxpElyY6i8nfl/jiM7BHJ8t/R\n0bmYDiP7HFp+zJ8wDJ/tvBpdfSm9Wd5VVl6aQO7Vag0ljSZ7lvW/RcR3cuWDkVAsj6dUtra0ExEb\nJa0ky9j+Hlki8+5BeO26FBFdkh4i64wWkz3He2+VOxpmB5J67acqvfdeJfsAcV2Fuv+fvbsPt6us\n7/z//iRRiYTHEBkbByJEpEXE0aht/alpZXSKQ0vBWnwqTC8L1kJxdGrtCBILVlHUQayF/BQBBbQK\n2qJiC2hGbdU2KIipkQYBmygSghJCwvN3/lhrw87OOcnJOfucfR7er+taV85a93r4rn2y77PWd93r\nvu9p/93A6JM4Q/ldmhabv9+pD5LsP8Jt76JpCfUCmtYCvYZK/I90v9B89kNd0P2wPV7RNAH/0hDr\n/GSUx5Zmmt66qB/1UHe9u6Fr+Vb17giuTTp1wQk0/aD0uqXr55Fcz4yozqqqbwH/LclcmqfsHwAu\n5bFEvDSV/Zzm//9w3+GfMvS91L40LU5G4heM7G90v65pfgrMS/LEnoTTvsDmqrq/D8foNl7XcJOp\n7vwM8KF2378BfLuqfrCd2EfiLpqWTR8douzOrnVW0vQx1avfv8dtmGwarN7WMmtpKovfY+vmd68E\nNvLYCAJDtfx5Ak2C59H/NG0Lo99mZBcM2/OkJL/eeZUuyX7As2laM3X7GE2G9hDg81XVm+keb0N9\nLuPpU8CnkxxJk0H+1AQeW5ooU6WeGsq1bVyrqmrLdtZ5a5L/XlVf6MMx5wIP9iSehxptb6jP5ys0\nTy73qKqr+xBLxzdp+hT4par64nArJfkW8PSq+ss+Hlua6fpRD32fptPZ36Pp56PjlcBNVbW+a9n2\nrk1+SNPn3qKq+v9HdTZb26k6qz3/K9OMePUXfTi+NHBVdW+Sb9N0J/LhIR48fxv44yS7VdU98Ggr\n6EWM8I2I9hgj+Rt9LfCnSfatqqEeLo201dG/0lyXvYKmD7bOq1+vGGnMfTCt6s6q2pLkMuBPgIOB\nN49mPz2upbnvvm47DR6upXmT4MdVNdqHlqNmsmmwVgO/k+QoHruBWwacn2QDzXuqL6bJRP7vqrqv\na7t9kxxP8yW6s21Z9K/AO5JspB0ZiqZp8+5jjPNO4JNJTqW5YXknzdOqC3vW+zxN885nM5iLiNXA\ny5N8maZ/gB92KvVx8iWaCux84Jaq+pdxPJY0KFOlnhrKB2hGNvlKknNpLhT2beP9RlVd1sb/DzT9\nifwlzYgdT6bpzPvEURzzauBNSf4PcCXw620MvYaqr36Y5DzgU0neS/MkaheaC4mDqur1o4iHqvpF\nmiF3z2lbWX2NJul3EE1nlb/brvpW4Nokj9B0DH8PTR8FLwfeXlU3jeb40gw35nqoqu5q65RTkzxE\nUzccTTMi1at6jjfstUlVPZLkLcAnkuxO88DgAZobq6OAVwzx2sywRlJnJXk5TSe6n6cZTGYhzShN\nXxnpcaQp4G3ANcBVSZbTdEr9azTfiQ/QXCP9Q5KzgHk0nYbfCFy+E8cYyd/oD9L0BfT1JO+iGbnt\nl4Fdq+q9NN/BLcBxbb+SD1bVyt4DVdUP2sTIh9uHgjfTjER8MEO3kBkP07Hu/BjwBprfQT8aKSyj\nGVjri0kuoLlnX0jz6vKFVbWCJln4BmBFkrNpRqWbT9OX2O1V9cE+xDG8sfYw7jT6ieYdys/RNG8r\n2pGJaIaYXEPzn/hHwP/s2W4XmlZFd7TbXdguX0yTvbyXpjJ5K9uOtnY8Oz/KU+eLeRNNi4R/YpiR\niYBPtsee1bN8ET2jLLTLb6VrpIXuY3bNb3UOXct7R1x6DvCt9vy3GWlhmHh7P5+lDDHyEk3n5NsM\n/92ebwHvHvT/Jyen8ZimUj01TNkvtXH8rK2/bm2/t4d0rTOXZvS8te06twDvGuGxt4m1Paf/aM/x\nGh4bhWSH9RVNvw9vonlN+X7aQRSAP9iJ39lWx+pa/lqaJvtbaJr9fxt4c886zwe+TNNK7V6ajj4/\nQNNyYeD/H52cJvM0XF3Uj3qIpgXRO9u65YH2u/maYeLY7rUJzTDmX2+/4xuB64EzaUdF6q2Tu7Zb\nyrbDo2+3zqIZ9vuzbdz3t+d3Hl0jTDk5TYeJJgnyNZqExS9o+kV8Vlv2X2gSrJ2yS4F9u7ZdxA5G\no2uX7fBvNE1fQJ9u/85vBm4Aju0qfw3NPd0DQLXLhvpuPxE4t6veWgm8rCeeFfTcHw21rx18bkPW\nm23ZtKg7e/axFvjkKP5/Dfk50SQAP0tznb6F5tr8fOApXevsAZzT9RmsBa4AXjDe34u0AUhDSnIh\nTWWxw97q297tbwMuqKrTxjs2SZIkSZImuyS/QpOYP7yqrt3R+tOBr9FpzJI8HjgMeDVNs7zzBxuR\nJEmSJEmDlWQ+TSvPM2i6lpgxrxGbbJrB2hGghh0FqqoeGuGufonmfdE7gBOram0fwhuzHZ0f8HDZ\ntE+a1PpYT43m2KFpfj2cCa1DRhDPI1U11IhQkiRJE2qQ13CTzJE0g2itBl7Xe+2YZDbNK8lDqap6\neJzjGzf9GG5aU9cFwIPDTUkWVdXxO3qFrqpurapU1b5VddG4Rz1y72A750fzbrWkyW2H9dQ4Hvu4\n7R27LZ9IL95BPO+Y4HgkSZKGM8hruEmjqi6sqllV9StVdd0Qq9zM8J/TzRMXaf/ZZ9MM1n7B99nO\nKt+rqge2Uz6pJfklmlZXwxnv0eokjdEg66m22fNTt7PKLVW1YTyOPUw8u9E0wx7OT6rqJxMVjyRJ\n0nCm+71mvyQ5FHjCMMX3V9WNExlPP5lskiRJkiRJUt/4Gp0kSZIkSZL6xmSTJEmSJEmS+sZkkyRJ\nkiRJkvrGZJMkSZIkSZL6xmSTJEmSJEmS+sZkkyRJkiRJkvrGZJMkSZIkSZL6xmSTJEmSJEmS+sZk\nkyRJkiRJkvrGZJMkSZIkSZL6xmSTJEmSJEmS+sZkkyRJkiRJkvrGZJMkSZIkSZL6Zs6gA+i3ffbZ\npxYtWjToMKRJ77rrrruzqhYMOg5tzTpMGhnrsMnH+ksaGeuvycf6SxqZnam/pl2yadGiRaxcuXLQ\nYUiTXpLbBh2DtmUdJo2MddjkY/0ljYz11+Rj/SWNzM7UX75GJ0mSJEmSpL4x2SRJkiRJkqS+Mdmk\ncbdhwwb+9E//lA0bNgw6FEnaKdZfkqayNWvW8PKXv5w1a9YMOhRJ2inHHHMMS5cu5ZWvfOWgQ9Eo\nmWzSuLvooou48cYbufjiiwcdiiTtlOXLl/O9732P5cuXDzoUSdppZ555Jvfeey9nnnnmoEORpJ3S\nedB3xx13DDgSjZbJJo2rDRs2cNVVV1FVXHXVVbYO0IRLclKSlUnuT3Jh1/JFSSrJpq7ptK7yJDkr\nyYZ2OitJerb/apLNSVYnObznuK9OcluSe5N8PsneE3LC6psNGzZw9dVXA3D11Vdbf0maUtasWcOt\nt94KwK233mrrJklTxjHHHLPVvK2bpiaTTRpXF110EQ8++CAADzzwgK2bNAg/Ac4ELhimfM+qmtdO\nZ3QtPwE4CjgMeCZwJHBiV/llwHeB+cDbgc8mWQCQ5BDgfOB1wL7AZuAjfTsjTYjly5fzyCOPAPDI\nI4/YuknSlNLbmsnWTZKmit4HfLZumppMNmlcdVoFdPzjP/7jgCLRTFVVV1TV54GdbZZyHPD+qlpb\nVeuAs4HjAZIcBDwbOL2qtlTV5cD3gM5jmNcAV1bV16pqE3AacHSS3cZ+Rpoo11577XbnJWky67Rq\nGm5ekqTxZLJJ42rPPffcan6vvfYaUCTSsG5LsjbJx5Ps07X8EOCGrvkb2mWdsh9V1T3bKX9026q6\nGbgfOKjfwWv8VNV25yVpMlu0aNF25yVJGk8mmzSubr/99q3mf/rTnw4oEmkbdwLPBfYHngPsBlzS\nVT4PuLtrfiMwr+23qbesU77bMNv2lm8lyQltv1Ir169fP4pT0Xh4yUtestX84YcfPsyakjT5rF27\ndqv5devWDSgSSdJMZLJJ48qWAZqsqmpTVa2sqoeq6mfAScBLu1512wTs3rXJHsCmav4T95Z1yu8Z\nZtve8t5YllfVkqpasmDBgtGflPrqxBNPZNas5s/krFmzOOGEEwYckSSN3EMPPbTVfKcPTWk6SPLJ\nJLcn2ZjkpiSv7yp7STt4y+Z2MJf9BxmrNFOZbNK4mjt37nbnpUmkkwnt1IuraDoH7zisXdYpO6Cn\nD6be8ke3TXIg8Hjgpj7HrHE0f/58XvSiFwHwohe9iPnz5w84IkmS1HoPcEBV7Q78NnBmkue0XSJc\nQdNf5t7ASuDTgwtTmrlMNmlc3Xfffdudl8ZbkjlJdgFmA7OT7NIue36SpyeZlWQ+8CFgRVV1Xn+7\nGHhzkoVJFgJvAS4EqKqbgOuB09v9HQ0cClzebnsJcGSSFybZFTgDuKKnjydNARs3bgTgnnv81UmS\nNFlU1feranNntp0OBI4GVlXVZ6rqPmAZcFiSgwcTqTRzmWzSuPI1Ok0CpwJbgLcBr21/PhU4APgy\nzatt36fpwPtVXdudD1wJ3NhOX2iXdRwLLAF+DrwbeEVVrQeoqlXAG2iSTncAuwJvHJez07jZsGED\n3/nOdwC47rrrthmGV5IkDU6SjyTZDKwGfgp8iW0HabkXWMNjg7h0b2+fmdI4MtkkaVqrqmVVlZ5p\nWVVdVlVPrapdq+rJVfUHVXV713ZVVW+tqr3b6a3VlS2tqluramlVza2qp1fVNT3HvbSq9mv3/ztV\ndddEnrfG7n3ve99W82efffaAIpGknbfbbluPSbH77r1dCUpTW1W9kWbwlRfSvDp3PzsA3L4AAAAg\nAElEQVQxSIt9Zkrjy2STJElD+Na3vrXV/De/+c0BRSJJO6/39d/Oa8HSdFJVD1fVN4CnAH/MTg7S\nImn8mGySJEmSJE1lc2j6bOodpGXXruWSJpDJJkmShjB79uztzkuSpImX5ElJjk0yL8nsJC+j6Xfz\nWuBzwDOSHNMOEHM6cENVrR5kzNJMZLJJkqQhPPLII9udlyRJA1E0r8ytpRmo5WzgTVX19+1gLccA\n72rLnkczqIukCTZn0AFIkjQZzZo1i4cffnireUmSNFhtQunF2ym/Bjh44iKSNBSvnCVJGsKee+65\n1fxee+01oEgkSZKkqcVkkyRJQ9iwYcNW83feeeeAIpEkSZKmFpNNkiRJkiRJ6huTTZIkSZIkSeob\nk02SJEmSJGlSWLFixXbnNTWMKNmU5JNJbk+yMclNSV7fVfaSJKuTbE7y1ST7d5UlyVlJNrTTWUnS\nVb6o3WZzu4/De4776iS3Jbk3yeeT7N2Pk5YkSZIkSdL4mDPC9d4DnFBVm5McDKxI8l3gNuAK4PXA\nlcAZwKeBX223OwE4CjgMKOBq4BbgvLb8MuCbwBHt9NkkT6uq9UkOAc4HXg58B1gOfAQ4dvSnK0mS\nJEmSJjNbM019I2rZVFXfr6rNndl2OhA4GlhVVZ+pqvuAZcBhbUIK4Djg/VW1tqrWAWcDxwMkOQh4\nNnB6VW2pqsuB7wHHtNu+Briyqr5WVZuA04Cjk+w2pjOWJEmSJEnSuBlxn01JPpJkM7Aa+CnwJeAQ\n4IbOOlV1L7CmXU5veftzd9mPquqe7ZR37/tm4H7goCFiOyHJyiQr169fP9JTkiRJkiRJUp+NONlU\nVW8EdgNeSPPq3P3APODunlU3tusxRPlGYF7bb9PObttb3h3b8qpaUlVLFixYMNJTkiRJkiRJUp/t\n1Gh0VfVwVX0DeArwx8AmYPee1fYAOq2Vesv3ADZVVY1i295ySZKkac1BWiRJ0lS0U8mmLnNo+mxa\nRdP5NwBJdu1aTm95+3N32QE9fTD1lnfv+0Dg8cBNo4xZ0gyU5KT2Ndv7k1zYtfxXk1yd5K4k65N8\nJsmTu8qXJXkwyaau6YCucm/UJE2E9wAHVNXuwG8DZyZ5TpJ9aFqanwbsDaykGaSlo3uQlmcCRwIn\ndpVfBnwXmA+8nWaQlgUAXYO0vA7YF9hMM0iLJEnSiOww2ZTkSUmOTTIvyewkLwNeBVwLfA54RpJj\nkuwCnA7cUFWr280vBt6cZGGShcBbgAsBquom4Hrg9CS7JDkaOBS4vN32EuDIJC9sk1hnAFf09PEk\nSTvyE+BM4IKe5XvRjHK5CNifptXkx3vW+XRVzeuaftRV5o2apHHnIC2SpJlo6dKlj06amkbSsqlo\nXplbC/yc5mLlTVX191W1nubC5F1t2fOAY7u2PR+4Erixnb7QLus4FljSbvtu4BXtPqmqVcAbaJJO\ndwC7Am8c1VlKmrGq6oqq+jywoWf5Ve1N2sb2Ru7DwAtGsk9v1CRNJAdpkSRJU82cHa3QJn9evJ3y\na4CDhykr4K3tNFT5rcDS7ez7UuDSHcUoSX3wIh57jbfjyCR30dzcfbiq/qZdPpIbtX/uFFTVzUk6\nN2rXjUfwkqavqnpjkpOBX6O5buoM0tKb3RnrIC0Lh9m2d9/dsS2naSXKkiVLasQnJUnSMHpbMy1d\nupQVK1YMJBaN3mj7bJKkaSPJM4F3AH/WtfhvgV8GFgB/BLwjyavasr6Nptke35YBkrbLQVokSdJU\nYrJJ0oyWZDFwFXBKVX29s7yq/q2qftLe4P0zcA7wira4rzdqVbW8qpZU1ZIFCxaM7YQkTXcO0iJJ\nkiY9k02SZqx2mPBrgDOq6hM7WL2AzrDh3qhJGncO0iJJkqYqk02SprUkc9obsdnA7PbGak578/UV\nmr6Yzhtiu99JslcazwNOAf4OvFGTNGEcpEWSJE1JO+wgXJKmuFNpnvh3vBZ4J81N3AHAsiTLOoVV\nNa/98VjgAuAJNDd676mqi7r2cyxNK4GfAz+m50YtSedGbT5N66n/0e8TkzS9OUiLJEmaqkw2SZrW\nqmoZsGyY4nduZ7tXDVfWlt+KN2qSNGOde+65rFmzZtBh7JRTTjll0CFsY/HixZx88smDDkNTSJIn\nAB8BDgf2Bm4G/qKqrkqyCLgFuLdrk7Oq6oyJjlOa6Uw2SZIkSZKmijnAf9C0/PwxcATwt0kO7Vpn\nz6p6aBDBSWqYbJIkSZJ20mRvjXPUUUfxi1/84tH5vfbai3POOWeAEUn9UVX3snWr9S8kuQV4DnDd\nQIKStA07CJckSZKmmbPPPnur+fe9730DikQaX0n2BQ7isVGBAW5LsjbJx5PsM6DQpBnNZJMkSZI0\nzSxevJhZs5pL/b322ovFixcPOCKp/5I8jmZAlouqajVwJ/BcYH+alk67teVDbXtCkpVJVq5fv36i\nQpZmDJNNkiRJ0jR04IEHMmvWLFs1aVpKMgv4BPAAcBJAVW2qqpVV9VBV/axd/tIku/VuX1XLq2pJ\nVS1ZsGDBhMYuzQT22SRJkiRNQ0984hM59NBDbdWkaSdJgI8B+wJHVNWDw6xa7b82spAmmMkmSZIk\nSdJU8jfALwOHV9WWzsIkzwd+Afw7sBfwIWBFVd09kCilGcwMryRJkiRpSkiyP3Ai8Czg9iSb2uk1\nwAHAl4F7gO8D9wOvGliw0gxmyyZJkiRJ0pRQVbcB2c4ql01ULJKGZ8smSZIkSZIk9Y3JJkmSJEmS\nJPWNySZJkiRJkiT1jckmSZIkSZIk9Y3JJkmSJEmSJPWNo9FJkiRJkjRDnHvuuaxZs2bQYeyUU045\nZdAhbGPx4sWcfPLJgw5j0rJlkyRJkiRJmhTmzp273XlNDbZskjStJTkJOB44FLisqo7vKnsJ8NfA\nfsC3geOr6ra2LMB7gNe3q38UeFtVVVu+CPg48Hzgx8BJVXVN175fDbwb2Ae4GvjDqrprnE5TkiRJ\nGpHJ3hpnw4YNHHPMMY/Of/KTn2T+/PkDjEijYcsmSdPdT4AzgQu6FybZB7gCOA3YG1gJfLprlROA\no4DDgGcCRwIndpVfBnwXmA+8HfhskgXtvg8BzgdeB+wLbAY+0ufzkiRJkqad+fPnP9qa6ZnPfKaJ\npinKZJOkaa2qrqiqzwMbeoqOBlZV1Weq6j5gGXBYkoPb8uOA91fV2qpaB5xN00KKJAcBzwZOr6ot\nVXU58D2g8wjmNcCVVfW1qtpEk9A6Oslu43aikiRJ0jTx1Kc+lV133ZXTTz990KFolEw2SZqpDgFu\n6MxU1b3Amnb5NuXtz91lP6qqe7ZT3r3vm4H7gYOGCiTJCUlWJlm5fv36UZ+QJEmSNB087nGPY/Hi\nxbZqmsJMNkmaqeYBd/cs2wjsNkz5RmBe25fTzm7bW76VqlpeVUuqasmCBQt26iQkSZIkabIx2SRp\nptoE7N6zbA/gnmHK9wA2tR2E7+y2veWSJEmSNG2ZbJI0U62i6fwbgCS7Age2y7cpb3/uLjugpw+m\n3vLufR8IPB64qY/xS5IkSdKkZLJJ0rSWZE6SXYDZwOwkuySZA3wOeEaSY9ry04Ebqmp1u+nFwJuT\nLEyyEHgLcCFAVd0EXA+c3u7vaOBQ4PJ220uAI5O8sE1inQFc0dPHkyRJkiRNSyabJE13pwJbgLcB\nr21/PrWq1tOMHvcu4OfA84Bju7Y7H7gSuLGdvtAu6zgWWNJu+27gFe0+qapVwBtokk53ALsCbxyf\n05MkSZKkyWXOoAOQpPFUVcuAZcOUXQMcPExZAW9tp6HKbwWWbue4lwKX7kyskiRJkjQd2LJJkiRJ\nkiRJfWOySZIkSZIkSX1jskmSJEmSJEl9s8NkU5InJPlYktuS3JPk+iS/1VX+kiSrk2xO8tUk+3eV\nJclZSTa001lJ0lW+qN1mc7uPw3uO/er2uPcm+XySvft14pIkSZKkqWUs96eSJs5IWjbNAf4DeDGw\nB83ITn/bJor2Aa4ATgP2BlYCn+7a9gTgKOAw4JnAkcCJXeWXAd8F5gNvBz6bZAFAkkNoRn56HbAv\nsBn4yKjOUpIkSZI0HYzl/lTSBNnhaHRVdS9bj+T0hSS3AM+hSRKtqqrPACRZBtyZ5OCqWg0cB7y/\nqta25WfTJKDOS3IQ8GzgpVW1Bbg8ySk0Q5GfB7wGuLKqvtZuexrwgyS7VdU9Yz91SZIkSdJUMsb7\nU0kTZKf7bEqyL3AQsAo4BLihU9Z+8de0y+ktb3/uLvtRT+Kot7x73zcD97fH7o3phCQrk6xcv379\nzp6SJEnSpGNXBpK0Yzt5fyppguxUsinJ44BLgIvazPA84O6e1TYCu7U/95ZvBOa1Fzs7u21v+aOq\nanlVLamqJQsWLNiZU5IkSZqs7MpAkrZjFPen3dvaYEEaRyNONiWZBXwCeAA4qV28Cdi9Z9U9gHuG\nKd8D2FRVNYpte8slSZKmraq6t6qWVdWtVfVIVX0B6LwqcjTtqyJVdR/NKyWHJTm43fzRrgyqah1w\nNnA8QFdXBqdX1Zaquhz4Hk1XBtDVlUFVbaJJaB2dZJubNUkalFHenz7KBgvS+BpRsqltifQxmqdb\nx1TVg23RKponZp31dgUObJdvU97+3F12QM+FS295974PBB4P3DSSmCVJkqaTydiVgSQNwhjuTyVN\nkJG2bPob4JeBI9vOvDs+BzwjyTFJdgFOB27o6nztYuDNSRYmWQi8BbgQoKpuAq4HTk+yS5KjgUOB\ny9ttLwGOTPLCtpI4A7jCzsElSdJMM1m7MvA1FEkDMtr7U0kTZIfJprazyROBZwG3J9nUTq+pqvU0\nTa7fBfwceB5wbNfm5wNXAje20xfaZR3HAkvabd8NvKLdJ1W1CngDzYXVHcCuwBtHf6qSJElTz2Tu\nysDXUCRNtDHen0qaIHN2tEJV3QZkO+XXAAcPU1bAW9tpqPJbgaXb2felwKU7ilGSJGk66nlV5Iie\nV0WO61pvuK4M/qWdH7Irg64W44fRPODr3razb7sykDRpjOX+VNLE2anR6CRJkjSh7MpAkiRNOSab\nJEmSJiG7MpAkSVPVDl+jkyRJ0sSzKwNJkjRV2bJJ0ozV1UqgMz2c5Ny2bFGS6ik/rWvbJDkryYZ2\nOqvtW4Wu7b+aZHOS1UkOH8Q5SpIkSdJEs2WTpBmrquZ1fk4yD7gd+EzPantW1UNDbH4CcBRNJ7oF\nXA3cApzXll8GfBM4op0+m+RpnddUJEmSJGm6smWTJDWOoemb5OsjXP844P1Vtbaq1gFnA8cDJDkI\neDZwelVtqarLge+1x5AkSZKkac1kkyQ1jgMubvs56XZbkrVJPp5kn67lhwA3dM3f0C7rlP2oZ+Sm\n7vKtJDkhycokK9evt+GTJEmSpKnNZJOkGa8d8enFwEVdi+8EngvsDzwH2I1mZKaOecDdXfMbgXlt\nv029ZZ3y3YY6flUtr6olVbVkwYIFYzkVSZIkSRo4+2ySJHgd8I2quqWzoKo2ASvb2Z8lOQn4aZLd\n2hZLm4Ddu/axB7CpqipJb1mn/B4kSZIkaZqzZZMkwR+wdaumoXRer+vUm6toOgfvOKxd1ik7IMlu\nw5RLkiRJ0rRlsknSjJbk14GF9IxCl+T5SZ6eZFaS+cCHgBVV1Xk97mLgzUkWJlkIvAW4EKCqbgKu\nB05PskuSo4FDgcsn5KQkSZIkaYB8jU7STHcccEVPZ94ABwB/BTyJpr+lq4FXdZWf365zYzv/0XZZ\nx7E0yaefAz8GXlFV9v4tSZIkadoz2SRpRquqE4dZfhlw2Xa2K+Ct7TRU+a3A0rFHKEmSJElTi6/R\nSZIkSZIkqW9MNkmSJEmSJKlvTDZJkiRJkiSpb0w2SZIkSZIkqW9MNkmSJEmSJKlvTDZJkiRJkiSp\nb0w2SZIkSZIkqW9MNkmSJEmSJKlvTDZJkiRJkqaEJCclWZnk/iQXdi1flKSSbOqaThtgqNKMNmfQ\nAWhszj33XNasWTPoMHbKKaecMugQtrF48WJOPvnkQYchSZKYmtc3k1HnM5yM115TideJk85PgDOB\nlwFzhyjfs6oemtiQJPUy2SRJkqRJZc2aNVz//R/w8BP3HnQoU9qsBwqA6370swFHMnXN3nzXoENQ\nj6q6AiDJEuApAw5H0jBMNk1xk/0py9KlS7dZds4550x8IJIkaUp5+Il7s+XgIwYdhma4uau/NOgQ\ntPNuS1LA1cCfVdWdQ62U5ATgBID99tuvrwHYOnPsbJnZH4NsmWmySZIkSZI01d0JPBe4HpgP/DVw\nCc3rdtuoquXAcoAlS5ZUPwOxdebY2TJz7AbdMtNkk8bVihUrtmrdtGLFioHFIkmSJGl6qqpNwMp2\n9mdJTgJ+mmS3qrpnouOxdaYGbdAtMx2NTpIkSZI03XRaK3nPKw2AXzyNu8MOO4zDDjvMVk2alJKs\nSHJf1xC5P+wqe0mS1Uk2J/lqkv27ypLkrCQb2umsJOkqX9Rus7ndx+ETfW6SJEnTTZI5SXYBZgOz\nk+zSLnt+kqcnmZVkPvAhYEVV3T3YiKWZyWSTJMFJVTWvnZ4OkGQf4ArgNGBvmmbZn+7a5gTgKOAw\n4JnAkcCJXeWXAd+l6TPg7cBnkywY7xORJEma5k4FtgBvA17b/nwqcADwZeAe4PvA/cCrBhSjNOPZ\nZ5MkDe1oYFVVfQYgyTLgziQHV9Vq4Djg/VW1ti0/myYBdV6Sg4BnAy+tqi3A5UlOAY4Bzpv4U5Ek\nSZoeqmoZsGyY4ssmLhJJ22PLJkmCdye5M8k/JVnaLjsEuKGzQlXdC6xpl29T3v7cXfajns4ou8sl\nSZIkadoy2SRppvtzmmbXC2mGv70yyYHAPKD3Hf+NwG7tz73lG4F5bb9NO9p2K0lOSLIyycr169eP\n5VwkSZIkaeBMNkma0arq21V1T1XdX1UXAf8EHAFsAnbvWX0Pmn4AGKJ8D2BTVdUItu2NYXlVLamq\nJQsW2K2TJEmSpKnNZJMkba2AAKtoOv8GIMmuwIHtcnrL25+7yw5Istsw5ZIkSZI0bY2og/AkJwHH\nA4cCl1XV8V1lLwH+GtgP+DZwfFXd1pYFeA/w+nb1jwJva5/8k2QR8HHg+cCPaUaEuqZr368G3g3s\nA1wN/GFV3TWqM5WkHkn2pKl//i/wEPD7wIuAU4CfA+9LcgzwReB04Ia2c3CAi4E3J/lSO/8WmiF2\nqaqbklwPnJ7kVJqWUocCl0/IiU0h5557LmvWrBl0GCN2yimnDDqEIS1evJiTTz550GFIkiRJwMhb\nNv0EOBO4oHvheA4NnuQQ4HzgdcC+wGbgIyM/NUnaocfR1G3rgTuBk4GjquqmqlpPM3rcu2gST88D\nju3a9nzgSuDGdvpCu6zjWGBJu+27gVe0+5SkEUlyUtuf2/1JLuwpe0mS1Uk2J/lqkv27ypLkrCQb\n2ums9gFgp3xRu83mdh+H9+z71UluS3Jvks8n2XvcT1aSJE0rI2rZVFVXACRZAjylq2g8hwZ/DXBl\nVX2t3fY04AdJdusZ4UmSRqVN/jx3O+XXAAcPU1bAW9tpqPJbgaVjDnKam8ytcZYuXbrNsnPOOWfi\nA9FM1nnY9zJgbmdh18O+19Mkvc+gedj3q+0q3Q/7iqZ1+C0011fQPOz7Jk2ryyNoHvY9rarWdz3s\neznwHZqBEz7C1sl2SZKk7Rprn03jOTR4775vBu4HDuoNwpGcJEnSdFNVV1TV54ENPUWPPuyrqvuA\nZcBhSTrJ8Ucf9lXVOuBsmu4Q6HrYd3pVbamqy4Hv0Tzsg66HfVW1iab1+tE9fdBJkiRt11iTTeM5\nNPiIhw53JCdJUr+tWLFiu/PSAE2Kh32SJEnDGWuyaTyHBt+pocMlSZJmiEnxsA9sXS5JkoY2oj6b\ntmMVTVNtYLtDg/9LOz/k0OBdT9cOAy7p2baz7wOBxwM3jTFmSZJG5LDDmj9D9tWkSWbUD/uS9PVh\nX1Utp+nXiSVLltROnIMkTVvr1q1j9ua7mbv6SzteWRonszdvYN26hwZ2/BG1bEoyJ8kuwGxgdpJd\nkswBPgc8I8kxbflwQ4MvTLKQZmjwC6EZGhzoDA2+S5Kj2Xpo8EuAI5O8sE1inQFcYefgkiRphut9\nIDfcw76OIR/2bafch32SJGlMRtqy6VSaRFLHa4F3VtWyJMcAHwY+CXybbYcGP4BmWHCAj7Lt0OAX\n0gwN/mO6hgavqlVJ3kCTdJoPXAP8jxGfmSRJ0hTWPtibQ9fDPuAhmod972uvwb7I8A/7Oo/U3wJ8\nCJqHfUk6D/tOpRmNrvdh3zeTvJBmNDof9knSTlq4cCG33z+HLQcfMehQNIPNXf0lFi7cd2DHH1Gy\nqaqW0Yx0MlTZuA0NXlWXApeOJEZJkqRpxod9kiRpShprn02SJEkaBz7skyRJU9VYR6OTJEmSJEmS\nHmWySZIkSZIkSX3ja3SSJEmaVBw2XJPFoIcOl6SpypZNkiRJkiRJ6htbNkmSJGlScdhwTRaDHjpc\nkqYqWzZJkiRJkqaEJCclWZnk/iQX9pS9JMnqJJuTfDXJ/gMKU5rxTDZJkiRJkqaKnwBnAhd0L0yy\nD3AFcBqwN7AS+PSERycJ8DU6SZIkSdIUUVVXACRZAjylq+hoYFVVfaYtXwbcmeTgqlo94YFKM5wt\nmyTNWEmekORjSW5Lck+S65P8Vlu2KEkl2dQ1nda1bZKclWRDO52VJF3li9rm25vb5tyHD+IcJUmS\nZohDgBs6M1V1L7CmXS5pgtmySdJMNgf4D+DFwI+BI4C/TXJo1zp7VtVQYx6fABwFHAYUcDVwC3Be\nW34Z8M12n0cAn03ytKpaPx4nIkmSNMPNA3qvszYCuw21cpITaK7n2G+//cY3MmkGMtkkacZqn3gt\n61r0hSS3AM8BrtvB5scB76+qtQBJzqa5YDkvyUHAs4GXVtUW4PIkpwDH8FgySpIkSf2zCdi9Z9ke\nwD1DrVxVy4HlAEuWLKl+BzN7813MXf2lfu92xph130YAHtml91eqkZq9+S5gcKNpmmzagXPPPZc1\na9YMOowprfP5nXLKKQOOZGpbvHgxJ5988qDDmNaS7AscBKzqWnxbkk7LpT+rqjvb5Vs11W5/PqSr\n7EdVdc8w5ZIkSeqvVTQPAwFIsitwIFtf102IxYsXT/Qhp501a5rL6MUHDC5ZMvXtO9D/iyabdmDN\nmjVc//0f8PAT9x50KFPWrAeaBwXX/ehnA45k6mqy0hpPSR4HXAJcVFWrk8wDngtcD8wH/rotf1m7\nyTzg7q5dbATmtf029ZZ1yhcOc2ybcUuSJI1Akjk097GzgdlJdgEeAj4HvC/JMcAXgdOBGwbRObgP\niMeu01DhnHPOGXAkGi2TTSPw8BP3ZsvBRww6DM1gNsEdX0lmAZ8AHgBOAqiqTTRD5gL8LMlJwE+T\n7Na2WOptqr0HsKmqKsmkasYtSZI0jZxKk0jqeC3wzqpa1iaaPgx8Evg2cOwA4pOEySZJM1zbEulj\nNC80H1FVDw6zaicJ1BnFcxVN5+D/0s4fxmPNtFcBB3Qlpjrll/Qz9h3xNeCx8zXg/vA1YElSv1TV\nMrbuc7O77Brg4ImMR9LQTDZJmun+Bvhl4PC2M28Akjwf+AXw78BewIeAFVXVeT3uYuDNSTrNzt7S\nrkNV3ZTkeuD0JKfSjEZ3KHD5BJzPo3wNeOx8DXjsfA1YkiRp5jHZJGnGSrI/cCJwP3B708gJ2mWP\nAH8FPImmv6WrgVd1bX4+cABwYzv/0XZZx7HAhcDPgR8Dr6iq3uF4x52vAWvQfA1YkiRp5jHZJGnG\nqqrbgGxnlcu2s20Bb22nocpvBZaOITxJmtEcNnzsHDp87AY9dLgkTVUmmyRJkjSpOGx4fzh0eD8M\nduhwSZqqTDZJkiRpUrFD+f5w6HBJ0qDM2vEqkiRJkiRJ0siYbJIkSZIkSVLfmGySJEmSJElS35hs\nkiRJkiRJUt+YbJIkSZIkSVLfmGySJEmSJElS38wZdACSpPGxbt06Zm++m7mrvzToUDSDzd68gXXr\nHhp0GJIkSZpAtmySJEmSJElS39iySZKmqYULF3L7/XPYcvARgw5FM9jc1V9i4cJ9Bx2GJEmSJpAt\nmyRJkiRJktQ3JpskSZIkSZLUN75GJ0nT2OzNd9lB+BjMum8jAI/ssvuAI5m6Zm++C/A1OkmSpJnE\nZNMOOJqTJgNHc9JoLF68eNAhTHlr1twDwOIDTJaM3r7+X5QkSZphTDZJ0jR18sknDzqEKe+UU04B\n4JxzzhlwJJIkSdLUMamTTUn2Bj4GvBS4E/iLqrp0ImNwNCdNBo7mNDVNhjpMkkbD+kuSJI3FpE42\nAX8NPEDT2cOzgC8muaGqVk1kEPZ5Mjb2eTJ29nkyZU2KOkySRsH6S5IkjdqkTTYl2RU4BnhGVW0C\nvpHk74DXAW+bqDgmez8T69atY8uWLYMOY7u2PHwfAHMfyYAjGd7cuXNZuHDhoMPYDvs8mWomSx02\n2Z177rmsWbNm0GEMqxNb53W6yWrx4sW+Nqm+sf4amclef8HUqMOsvzQekqwAfhXodHq6rqqePriI\nJhfrr/6w/tq+SZtsAg4CHqqqm7qW3QAs7V0xyQnACQD77bdfX4OY7P95pkJFsW7dOoBJncyxotA4\nmBR1mMZm7ty5gw5BGgTrr2nCOkwz3ElV9dFBB6HRsf6a+iZzsmkesLFn2UZgt94Vq2o5sBxgyZIl\nNf6hTR4mSKRJyzpsBKzDpEnJ+msErL8kTVXWX5oIswYdwHZsAno7+dkDuGcAsUjSzrIOkzRVWX9J\nmg7eneTOJP+UZGlvYZITkqxMsnL9+vUDCE+a3iZzsukmYE6Sp3UtOwywY0pJU4F1mKSpyvpL0lT3\n58ABwEKa1pdXJjmwe4WqWl5VS6pqyYIFCwYRozStTdpkU1XdC1wB/GWSXZP8f8BvA58YbGSStGPW\nYZKmKusvSVNdVX27qu6pqvur6iLgn4AjBh2XNJNM2mRT643AXOAO4FLgjx1yVwv8G9EAACAASURB\nVNIUYh0maaqy/pI0nRQweYfGlqahydxBOFV1F3DUoOOQpNGwDpM0VVl/SZqqkuwJPB/4v8BDwO8D\nLwJOGWRc0kwzqZNNkiRJkiTthMcBZwIHAw8Dq4GjquqmgUYlzTCpml6j1CZZD9w26Di0jX2AOwcd\nhLayf1XZG+IkYx02KVl/TU7WYZOM9dekZR02+Vh/TTLWX5OW9dfkM+L6a9olmzQ5JVlZVUsGHYck\n7SzrL0lTmXWYpKnK+mtqm+wdhEuSJEmSJGkKMdkkSZIkSZKkvjHZpImyfNABSNIoWX9JmsqswyRN\nVdZfU5h9NkmSJEmSJKlvbNkkSZIkSZKkvjHZJEmSJEmSpL4x2SRJkiRJkqS+MdkkSZIkSZKkvjHZ\nJEmSJEmSpL4x2SRJkiRJkqS+Mdk0QZJcmGRlH/d3UJJlSfbs1z4nkyRPas9v0aBj2VnT/XcjjUWS\nVyY5fhTbvTTJm0Z5zOOTVJJ5o9l+Io021rF8PtJMZp0kSSM3mnvaJCckOWq8YhrmmG9NsnQij6lt\nmWyaug4CTgema0LjSTTnt2jAcYzGdP/dSGPxSuD4UWz3UmAmJFO+CPwasHknt5spn4/Ub9ZJkjRy\nZ7DzdeYJwIQmm4C3Aksn+JjqMWfQAUiSNJklmQ3MrqoHxvtYVbUeWD/ex5E0dU1knTRaSR4HPFJV\nDw86Fkn9U1U3D+rYSXapqvsGdXztPFs2DUCSJye5IMmPkmxJclOSM5M8vme9v0iyJsl9SX6W5MtJ\n/lPbJPDKdrVb2qbYt47w2LcmOTvJaUluT7IpySVJ9mjLZyf5SZJlQ2y7Isnn2p87TcCf3S7fnOT6\ndn7XJB9Pcnd7jq8aYj+fTfLq9vw2JrkqyVPa8kXAje3qX22PUyM8v5cnuTrJHe1+v5XkpT3rLEty\nZ5LnJ1nZ/g6+keSp7et7n28/lx8k+c1hPr//mWRtkp8n+VTnlbmx/G6k6S7JhcAxwIs73+tOXZPk\npCT/nuT+tl74n13bLQPeAuzftd2FbdmvJfn7JD9Ncm9bD71mrHG2dcNRSVYB9wHPb8v2a7/zd7X1\n3j8keXrP9nOTvDfJbe353JLk3SM89lav1yRZ1M6/Msn5bb26Nsk7k8za0ecjaXjWSSM6dqdOem6S\nr+ex69bf7Vmvc213QpKb2xh/qeua67+012Sbk3w3yQuHONYfJbkxj133fjaPXZ92fwar23W+keRX\nRvepShqNdL1G11U/HJrm/uve9vt5dNf6K4DnAMd11ZfHj+A4nX0/r61ftgB/1pa9p60rNrXXRJck\n+U9d294KzAdO7zrm0rZsVpK3tfX6/W19dlzfPiBtxZZNg7EP8AuaL8ydNK9dLQMWACcCJPkD4H8D\nfw6sovnC/CawK/Ad4H8BZwNHAz8F7t+J478KWAP8EfBk4L3AR4Hfq6qHk1wE/EGSd1ZVtfEcALwI\n+J2efV0EfBg4C3gP8FngX4CbgVcAfwhcnOTrVbW2a7vnA79Ec7E2FzgHWA4c0Z7Pa4BLgD9pz3ek\nngp8CXg/8DDwW8BVSV5UVf/Utd4T2+O9F7gX+BDwCZrP8SrgIzTNLz+T5D9XVfcrLa8EvkfTJPQp\nwAeAvwLeyNh/N9J0dgawH80rpm9sl61N8kfAuTTfpX8AfgN4f5InVNV7aOqnp9HUgZ0bnE7rn0XA\nt2i+z5uBFwAfT/JIVV02hlgX0dQPfwncTpM83hv4BrABeEN7vLcB1yQ5qKq2JAnwdzSvwp0BXAcs\nBLa5sdpJ7wUup6lXXwK8g+Zvw9+y/c9H0vCsk0bu0zTXRn8FvJ7m+ug5VXVD1zovAA6kuXbdDNzd\nLn8izfXiB9vYTweuSLJ/5/oqyantuX2E5vr4icDLgXld+9mf5ndyGrAFeCfwD0meZmsHaaAupanz\n3gecDHwqyQHtvd8baa5ffkRTB0FznzhSl9HUC++kuX8G+E80955rae6r3wJ8JckzquoRmnr5qzT3\npR9tt/m39t9zgeNo6pvvAP8VuCDJhqr6ws6ctEagqpwmYAIuBFYOUzYHeDXNU6DHt8s+DFy+nf39\nd6CARTsZx63AXcC8rmWvAR4Bfrmdf1q779/oWqdzcTOnnT++Xee4rnWOaJdd0LVsD+BB4I+7lq2g\nuXDYq2vZm9pt57bzz2jnl47hM5/Vfrb/0BPTsnbfL+5a9sZ22Tu6lv1Ku+y3ej6/mzufQ7vs/wC3\nj/V34+Q0EyaaP/wruuZnAeuAj/es95G2ntilnT8buHUH+077nT8f+ErX8k59NW+EMV7Yrv+snuVn\n0NzU7d21bK82zj9p51/Wbvvbo/x8toqV5gazgIt71rse+FTX/A4/Hycnp20n66QdHrsT6//u+YxW\n99RBK2gSQPv2bL+s3f43u5Y9q13239r5PWmSUx8YwWfw613L9gceAt4w6P9HTk4zZaLrnrarfvjD\nrvL5vd9LYCVw4U4ep7PvU3aw3myaBHoBL+pafiewrGfdxTT3vMf1LL8Y+NdBf7bTcfI1ugFI401J\n/q1tEvggTSueJ9A8YYPmRuKINK9KPC/N+/n9cnVVbeqa/xzNBdFzAarq34Gv0Xb+1j4V+wPgE1X1\nUM++ru36eU3771c6C6rqbpqnfQt7tvvXqvp513wn29y73k5J8pQkFyVZR1PRPUjTiedBPas+AHx9\ne7F3LeuN6as9n8O/AU9K0z+BpJ3zFJpWjp/pWf5pYHfg0O1tnGSvJB9KchvN9/1BmlaHvd/5nbWu\nqq7vWXY4cDWwMcmcJHOAe2haCixp1/lN4K6q+vsxHr/XP/bM/xvNZyepv6yThva5zg/VtBz4O+B5\nPetcV1U/G2LbB2iSUR2da75OHfZrNK3cP76DGO6oqn/uiuM2mnPtjUPSxHr0GqWqNgB30L9rlC/2\nLkjyW0n+OcndNPd7nbdndlTPvoQm2fS5Tp3Z1pvXAs/q8/22sM+mQXkTzROxz9G8lvY8mtfFAHZp\n/72A5jW6VwLfBn6Wpl+nfnwJ7uieqaYJ8yaaV+o6PgYck6bfkN+keXp0wRD7+kXXzw8MsayzfJee\nZUOtwxDrjVia/kv+Hvh1mldMfoMmgXbVEPu9p71Y6j3+o3HVYx1vjiT20CQLJe2cTr3Te4PSmd97\nB9tfCPw+TdPtl9J85y9gDHXJMPFA01T793nsBrIz/Qbwn9t15tO8PttvI6lXJY2dddLQ7hhi/sk9\ny4aKEXquuYa4vprf/rujOHtjGC4OSRNrPK9RtqpXkjyX5n5vLfA6mmT1r7bFOzrmPjQtoe5m6zrz\nQppWqNYlfWafTYPxe8Bnq+rtnQW9HRy2f5Q/CHwwyX+medXtXTRfrPPGePwndc8keSLNO/Hdf+Q/\nQ9OP0StpLlq+XVU/GONxx9ti4L/QvPb25c7CJHMHF5KkHejUO0/qWb5v++9dw22YZBea11b/pKrO\n61rejwcpQw1KcBfNBc4ZQ5Td0/67AS9WpKnMOmloT2r31T3fmxwa0WAuQ+js98k0r75sL4ahlq0a\n5XElTX699crv0rw18/vVvgOXZP8R7usumpZQL6Bp4dRrqIS2xsCWTYMxl207jR52pJKq+o9qOqRc\nQ9OPEIytJdB/bVssdfwuzRd5Zdcxt9B0yPYnNB1d76hpc7+N5vw6SaVHP9u28nlBv4IaoTG30pKm\nsd6nXWuBn9Ak4bu9EtjIYyNTDvWU7Ak0f8e6v/O7Ab/dx3i7XQscAqyqqpU90w+71tk7yX8fpxiG\nY0snaXSsk0bm0dHn2uTZ79AMCNMP36Tp72lHI0I9Kcmvd8WxH/DsPsYhaXz08xplLvBgJ9HUGuo+\neqhjfoWmZdMeQ9SZK7taXapPbNk0GFcDf5rk2zSdTb+GplXOo5KcT5N9/RZNU7/foOm4+8/bVToX\nEScm+RSwuapuZGS2AF9M8j6ap0jvAz5XVf/Ws97HaEY32QJ8auSn1xc/bo97XPs+7oNVtXIH26ym\nuUh8f5LTgN1oRi5YN66RbmssvxtpulsN/E6So3jspm4ZcH6SDTT144uBP6bpkPa+ru32bYfL/T5w\nZ1XdmuRfgXck2UjzlOptNHXm7uMQ+weA19KMeHIuTd2ybxvvN6oZaepqmkEJLk3SGenkyTSdVp44\nDjF1DPn5jOPxpOnCOmlkXp/kAZpzfT3Ndeur+nESVfWLJGcA70ryeJpRhZ9AMxrdO6uqcx13J/DJ\nduS6zmh0d9C8AiNp8loNvCzJy2haMt7S9u00GlcDb0ryf4ArabpPee0wx3x5ki/TdBfzw6r6YZLz\naEbLey9NQ4tdaJL2B1XV60cZk4Zhsmkw/hJYAJzZzl8B/CnNF6bjm8AfASfSfAnWAH9UVZ+HplPE\nJP+r3e5kmgukRSM8/qdomld/jOb1ub+nuYjaSlWtbDvaXtF29D1hquq+dujh04H/CzyOpl+k7W1z\nf5Kjgb+mGV1mLc2rh0tpRrebEGP83UjT3UdoXne9gGbUpHdW1bL29ZNT2mkt8Jaq+mDXdn9Lk3R/\nL039eRHNIAavphnp6WKaC5gP0wyZfVK/A6+qO5P8Kk298kGaEZR+SjP0+PfadSrJ79K81vKmNtaf\n0AwLPJ6G+3wkbZ910sgc2x7jTOA/aF5h+W4/zqON891J7qL5vE8Efk4zWM09/6+9e4+Tq67vP/76\nZJdKuMMS0UYjtkFRQNCuWFtR1NBGWq2IFxQlWGuqaIhYL7WFykVttSpCvJRUgcQbUgVRiWkTFJX+\nvBBaLlIRIwRkEUiWS8iFyyaf3x/nLE42s8nuZmbOzM7r+XjsIztnvnPmPSH75ZzPfi81zW4DPgL8\nC8VaoiuAN9QUACW1pw9RbIJ1MUXh/c1MsEicmUsi4v0U91hvpbhn/kvg5hFN30txT3g5RR/8YoqN\nCt5Rtn0rxT35WopNC74wkTzatthyBJomu4hYRbFe1HvG0PaZFPPgZ2XmFdtrL0mSpMmjHLl1AbD7\niJ2MW53jQuDgzOzfXltJUntwZJO2EhF9wNMpfgv2c4r5rZIkSZIkSdtlsWkSiYgeRp9qlpm5aYyn\nejnFcPKbgDdlmwx/287nIzOHWhhHUgcqF7YddXOMZvYjEREUC1OOZlO79LeSWqPd+6Rmvbek7uX1\nUPdwN7rJ5dfAo6N8/RogM/ff3hS6zLwwM6dk5jMz85omZx6PKxj98z1aYS5JneN8ttGPRMT+TXzv\nOdt6b7a/E5Okyaet+6TymjCqnEIHkJknOoVOmjS8HuoSrtk0iUTEIRS7d9TzcKfviBYRT6fYYa6u\nMexWJ6nLlTdu+26jyfXN2vq2nKL81G002ZHdWSR1IPskSd3Gvqd7WGySJEmSJElSwziNTpIkSZIk\nSQ1jsUmSJEmSJEkNY7FJkiRJDRER50fEPRHx8zG0PTsiri2/bo6I+1uRUZIkNZ9rNkmSJKkhIuKF\nwDpgcWYePI7XzQOenZl/3bRwkiSpZRzZJEmSpIbIzB8C99Yei4g/jIilEXFNRPwoIg6s89LXA19t\nSUhJktR0vVUHkCRJ0qS2EHhbZv4qIp4HfBZ4yfCTEfEUim2wv1dRPkmS1GAWmyRJktQUEbEb8CfA\nf0TE8OHHjWh2HPD1zNzUymySJKl5LDZJkiSpWaYA92fmYdtocxzwjhblkSRJLeCaTZIkSWqKzFwL\n3BoRrwGIwqHDz5frN+0N/LiiiJIkqQksNkmSJKkhIuKrFIWjp0fEHRHxFuB44C0RcR1wI/BXNS85\nDrgo3R5ZkqRJJfx/uyRJkiRJkhrFkU2SJEmSJElqGItNkiRJkiRJaphJtxvdvvvum/vvv3/VMaS2\nd80116zJzGlV59CW7MOksbEPaz/2X9LY2H+1H/svaWzG039NumLT/vvvz4oVK6qOIbW9iLit6gza\nmn2YNDb2Ye3H/ksaG/uv9mP/JY3NePovp9FJkiRJkiSpYSw2SZIkSZIkqWEqLTZFxDsjYkVEPBwR\nF4547qURcVNEbIiI70fEUyqKKUmSJHWcwcFBTj75ZAYHB6uOIknjctlll3HkkUfy7W9/u+oomqCq\nRzbdCXwIOL/2YETsC1wCnAbsA6wAvtbydJIkSVKHWrRoETfccAOLFy+uOookjcvZZ58NwCc+8YmK\nk2iiKi02ZeYlmflNYOSvW14F3JiZ/5GZDwGnA4dGxIGtzihJkiR1msHBQZYuXUpmsnTpUkc3SeoY\nl1122RaPHd3Umdp1N7qDgOuGH2Tm+ohYWR6/qZVBFixYwMqVK1v5luMyMDDAxo0bq47R8aZOncr0\n6dOrjrFNM2fOZN68eVXHkBqqE/pYwP5BUsdZtGgRmzdvBmDTpk0sXryYU045peJUkrR9w6Oahn3i\nE5/g5S9/eUVpNFHtWmzaDVg94thaYPd6jSNiLjAXYMaMGQ0NsnLlSq79+S/YtMs+DT1vo0x5aAOx\n+dGqY3S8Bx9J7nr47qpjjKpnw71VR5C6ksV8SZ1q+fLlDA0NATA0NMSyZcssNkmSWqZdi03rgD1G\nHNsTeLBe48xcCCwE6O/vz0aH2bTLPmw88OhGn1Yas6k3Lak6gtQU7T4aZ/78+QCcc845FSeRpPGZ\nNWsWS5YsYWhoiN7eXo466qiqI0mSukjVC4SP5kbg0OEHEbEr8IflcUmSJI1TROwTEZdGxPqIuC0i\n3jBKu4iID0XEQEQ8EBFXRsRBrc6rHTNnzhymTCku9Xt6ejjhhBMqTiRJ6iaVFpsiojcidgZ6gJ6I\n2DkieoFLgYMj4tjy+Q8C12VmS9drkiRJmkQ+AzwC7AccD3xulCLSa4C/Bo6g2BX4x8AXWxVSjdHX\n18fs2bOJCGbPnk1fX1/VkaSGiIh3RsSKiHg4Ii4c8dxLI+KmiNgQEd+PiKdUFFPqelWPbDoV2Aj8\nPfDG8vtTM3M1cCzwYeA+4HDguKpCSpIkdbJylPixwGmZuS4zrwIuA95Up/lTgasy85bM3AR8CXhm\n69KqUebMmcMhhxziqCZNNncCHwLOrz0YEfsClwCnURTKVwBfa3k6SUDFazZl5unA6aM8txw4sJV5\nJEmSJqmnAUOZeXPNseuAI+u0vQh4bUQ8DbgVmAMsrXfSZm7Soh3X19fHueeeW3UMqaEy8xKAiOgH\nnlTz1KuAGzPzP8rnTwfWRMSBzpCRWq/qkU2SJElqvt0odvatNdpOv78FrgJ+STHq/DVA3W3MMnNh\nZvZnZv+0adMaGFeSxu0giiI6AJm5HlhZHt9KRMwtp+OtWL165EboknaUxSZJkqTJbzw7/f4TxRIG\nTwZ2Bs4AvhcRuzQ1oSTtmN2AB0YcG62obrFcajKLTZIkSZPfzUBvRBxQc+xQ6u/0exhwUWbekZlD\nmXkhsDeu2ySpvY2nqC6pySw2SZIkTXLldJJLgDMjYteIeAHwCurvMnc18JqI2C8ipkTEm4CdKKaj\nSFK7upGiiA48tjHCH1K/qC6pySw2SZIkdYeTgKnAPcBXgLdn5o0RMSMi1kXE8ArfH6VY9+Ra4H6K\n9ZqOzcz7qwgtSbUiojcidgZ6gJ6I2DkieoFLgYMj4tjy+Q8C17k4uFSNSnejkyRJUmtk5r3AK+sc\nv51irZPhxw8B7yi/JKndnEpRSBr2RuCMzDw9Io4FPg18CfgpcFwF+SRhsUmSJEmS1CEy83Tg9FGe\nWw4c2Mo8kupzGp0kSZIkSZIaxmKTJDVJROwfEUsi4r6IuCsiPl2uKSBJkiSpjpkzZ27x+IADDhil\npdqZxSZJap7PAquBJ1JsJf4iigV6JUmSJNVx6623bvH4lltuqSiJdoTFJklqnqcCX8vMhzLzLmAp\ncFDFmSRJkqS2tWnTpm0+Vmew2CRJzfMp4HURsUtETAdeRlFw2kJEzI2IFRGxYvXq1S0PKUmSJEmN\nZLFJkprnh8DBwFrgDmAF8M2RjTJzYWb2Z2b/tGnTWhxRkiRJkhrLYpMkNUFETKEYxXQJsCuwL7A3\n8NEqc0mSJElSs1lskqTm2AeYAXw6Mx/OzEHgAuDoamNJkiRJUnNZbJKkJsjMNcCtwNsiojci9gLm\nANdXm0ySJEmSmstikyQ1z6soFgVfDawEHgVOqTSRJEmSJDVZb9UBJGmyysxrgSOrziFJkiRJreTI\nJkmSJEmSJDWMxSZJkiRJkiQ1jMUmSZIkSZIkNUzbF5siYv+IWBIR90XEXRHx6YhwrSlJkiRJkqQ2\n1PbFJuCzFDs5PRE4DHgRcFKliSRJkiRJklRXJxSbngp8LTMfysy7gKXAQRVnkiRJkiRJUh2dUGz6\nFPC6iNglIqYDL6MoOEmSJEmSJKnNdEKx6YfAwcBa4A5gBfDN2gYRMTciVkTEitWrV1cQUZIkSZIk\nSdDmxaaImEIxiukSYFdgX2Bv4KO17TJzYWb2Z2b/tGnTWh9UkiRJkiRJQJsXm4B9gBnApzPz4cwc\nBC4Ajq42liRJkiSpHbmjuVS9ti42ZeYa4FbgbRHRGxF7AXOA66tNJkmSJElqU+5oLlWsrYtNpVdR\nLAq+GlgJPAqcUmkiSZIkSVK7ckdzqWJtP5QwM68Fjqw6hyRJkiSpIwzvaH4lxZq/LwNOq20QEXOB\nuQAzZsxodb5KLViwgJUrV1YdY1zmz59fdYStzJw5k3nz5lUdo211wsgmSZIkSZLGars7mrvJlNRc\nbT+ySZIkSZKksajZ0Xwh8CfAbsD5FDuav6/CaG2j3UfjHHPMMdx3332PPe7r6+Occ86pMJEmwpFN\nkiRJXSAi9omISyNifUTcFhFv2EbbP4iI70TEgxGxJiI+1sqsaozBwUFOPvlkBgcHq44itZI7mne4\nSy+9dIvH3/jGNypKoh1hsUmSJKk7fAZ4BNgPOB74XERstWBuRPwesAz4HvAE4EnAl1qYUw2yaNEi\nbrjhBhYvXlx1FKll3NF8cogIoBjVpM5ksUmSJGmSi4hdgWOB0zJzXWZeBVwGvKlO8xOBOzPzk5m5\nvtzNyZu0DjM4OMjSpUvJTJYuXeroJnUbdzTvcM961rM49NBDHdXUwSw2SZIkTX5PA4Yy8+aaY9dR\nfyvwPwZWRcR3yyl0V0bEIS1JqYZZtGgRmzdvBmDTpk2OblJXycxrM/PIzNw7M/fNzNdm5t1V55K6\nicUmSZKkyW83il2Zaq0Fdq/T9knAccC5wO8DlwOXldPrthARcyNiRUSsWL16dYMja0csX76coaEh\nAIaGhli2bFnFiSRJ3cRikyRJ0uS3DthjxLE9gQfrtN0IXJWZ383MR4CPA33AM0Y2dOvw9jVr1ix6\ne4uNp3t7eznqqKMqTiRJ6iYWmyRJkia/m4HeiDig5tihwI112l4PZEtSqWnmzJnDlCnFpX5PTw8n\nnHBCxYkkSd3EYpMkSdIkl5nrgUuAMyNi14h4AfAK4It1mn8J+OOImBURPcC7gDXAL1oWWDusr6+P\n2bNnExHMnj3bHZ0kSS1lsUmSJKk7nARMBe4BvgK8PTNvjIgZEbEuImYAZOYvgTcC/wbcB/wV8Ipy\nSp06yJw5czjkkEMc1SRJarneqgNI0mQWEccBHwRmAHcBJ2bmj6pNJakbZea9wCvrHL+dYgHx2mOX\nUIyEUgfr6+vj3HPPrTqGJKkLWWySpCaJiKOAjwKvA34GPLHaRJIkSZLUfBabJKl5zgDOzMyflI8H\nqgwjSZIkSa1gsUmSmqBcVLcf+FZErAR2Br4JvDczN7Yiw4IFC1i5cmUr3mrSGv77mz9/fsVJOtvM\nmTOZN29e1TEkSZLUIhabJKk59gN2Al4NHAE8ClwGnAr8Y23DiJgLzAWYMWNGwwKsXLmSa3/+Czbt\nsk/DztltpjxS7P5+zS13V5ykc/VsuLfqCJIkSWoxi02S1BzDo5cWZOZvASLik9QpNmXmQmAhQH9/\nfzYyxKZd9mHjgUc38pTSuEy9aUnVESRJktRiU6oOIEmTUWbeB9wB1BaPGlpIkiRJkqR2ZLFJkprn\nAmBeRDw+IvYGTgG+U3EmSZIkSWoqp9FJUvOcBewL3Aw8BFwMfLjSRJIkSZLUZBabJKlJMvNR4KTy\nS5IkSZK6gtPoJEmSJEmS1DAdUWyKiOMi4hcRsT4ifh0RR1SdSZIkSZIkSVtr+2l0EXEU8FHgdcDP\ngCdWm0iSJEmSJEmjaftiE3AGcGZm/qR8PFBlGEmSJEmSJI2urafRRUQP0A9Mi4iVEXFHRHw6IqaO\naDc3IlZExIrVq1dXE1aSJEmSJEntXWwC9gN2Al4NHAEcBjwbOLW2UWYuzMz+zOyfNm1a61NKkiRJ\nktqG6/5K1Wr3YtPG8s8FmfnbzFwDfBI4usJMkiRJkqQ2VbPu75uB3YEXArdUGkrqMm29ZlNm3hcR\ndwBZe7iqPJIkSZKktue6v1LF2n1kE8AFwLyIeHxE7A2cAnyn4kySJEmSpDbjur9Se+iEYtNZwNXA\nzcAvgP8FPlxpIkmSJElSO3LdX6kNtPU0OoDMfBQ4qfySJI3RwMAAPRseYOpNS6qOoi7Ws2GQgYGh\nqmNIkrrHFuv+AkTEJymKTf9YWSqpy3TCyCZJkiRJkrYrM+8DXPdXqljbj2ySJE3M9OnTuevhXjYe\n6Aaeqs7Um5Ywffp+VceQJHWX4XV/lwKP4rq/UstZbJIkSZIkTSZnAftSrPv7EHAxrvsrtZTFJkmS\nJEnSpOG6v1L1XLNJkiRJkiRJDWOxSZIkSZIkSQ1jsUmSJEmSJEkNY7FJkiRJkiRJDeMC4dsxMDBA\nz4YHmHrTkqqjqIv1bBhkYGCo6hiSpA4WEfsAXwD+DFgDfCAzv7Kd11wBvATYKTP9H5EkSRoTRzZJ\nkiR1h88AjwD7AccDn4uIg0ZrHBHHAzu1KJuaYHBwkJNPPpnBwcGqo0iSuowjm7Zj+vTp3PVwLxsP\nPLrqKOpiU29awvTp+1UdQ5LUoSJiV+BY4ODMXAdcFRGXAW8C/r5O+z2BDwInAD9uZVY1zqJFi7jh\nhhtYvHgxp5xyStVxJEldxJFNkiRJk9/TgKHMvLnm2HXAaCObPgJ8Drirtoyw0gAAG75JREFU2cHU\nHIODgyxdupTMZOnSpY5ukiS1lMUmSZKkyW83YO2IY2uB3Uc2jIh+4E+BBds7aUTMjYgVEbFi9erV\nDQmqxli0aBGbN28GYNOmTSxevLjiRJKkbmKxSZIkafJbB+wx4tiewIO1ByJiCvBZYP5YFgTPzIWZ\n2Z+Z/dOmTWtYWO245cuXMzRU/CccGhpi2bJlFSeSJHUTi02S1GQRcUBEPBQRX6o6i6SudTPQGxEH\n1Bw7FLhxRLs9gH7gaxFxF3B1efyOiDii+THVKLNmzaK3t1ietbe3l6OOOqriRJKkbmKxSZKa7zP8\n7oZNklouM9cDlwBnRsSuEfEC4BXAF0c0fQD4feCw8mt4h5Q/An7aorhqgDlz5jBlSnGp39PTwwkn\nnFBxIklSN7HYJElNFBHHAfcDV1SdRVLXOwmYCtwDfAV4e2beGBEzImJdRMzIwl3DX8DwQkx3Z+Yj\nVQXX+PX19TF79mwigtmzZ9PX11d1JElSF+mtOoAkTVYRsQdwJvAS4G+20W4uMBdgxowZrQknqetk\n5r3AK+scv51iAfF6r1kFRHOTqVnmzJnDqlWrHNUkSWo5RzZJUvOcBXwhM+/YViMX2JUkNUNfXx/n\nnnuuo5okSS3nyCZJaoKIOAyYBTy76iySJEmS1EoWmySpOY4E9gdujwgopqj0RMQzM/M5FeaSJEmS\npKbqiGl0bhsuqQMtBP6Q3+3o9G/A5cCfVxlKkiRJkpqtU0Y2uW24pI6SmRuADcOPI2Id8FBmrh79\nVZIkSZLU+dp+ZJPbhkuaDDLz9Mx8Y9U5JEmSuoGzY6RqtXWxqWbb8Hdvp93ciFgREStWr3bQgCRJ\nkiR1OWfHSBVq62ITbhsuSZIkSRoHZ8dI1WvbNZvcNlySJEmSNB41s2NeAvxNxXGkrtW2xSbcNlyS\nJEmSND6PzY4p7yPrioi5wFyAGTNmtCia1D3audi0ELio5vF7KIpPb68kjSRJkiSpbY1ndkxmLqS4\n56S/vz+bHE3qOm1bbHLbcEmSJEnSOByJs2OkttC2xaaRMvP0qjNIkiRJktqWs2OkNtExxSZJkiRJ\nkkbj7BipfVhskiRJkiRNOs6OkaozpeoAkiRJkiRJmjwsNkmSJEmSJKlhLDZJkiRJkiSpYSw2SZIk\nSZIkqWEsNkmSJEmSJKlh3I1OkiRJkqQGWbBgAStXrqw6Rkcb/vubP39+xUk628yZM5k3b14l722x\nSZIkSZKkBlm5ciXX/vwXbNpln6qjdKwpjyQA19xyd8VJOlfPhnsrfX+LTZIkSZIkNdCmXfZh44FH\nVx1DXWzqTUsqfX/XbJIkSZIkSVLDWGySJEmSJElSw1hskiRJkiRJUsNYbJIkSZImocHBQU4++WQG\nBwerjiJJ6jIWmyRJkqRJaNGiRdxwww0sXry46iiSpC5jsUmSJKkLRMQ+EXFpRKyPiNsi4g2jtJsT\nEddExNqIuCMiPhYR7mDcYQYHB1m6dCmZydKlSx3dJElqKYtNkiRJ3eEzwCPAfsDxwOci4qA67XYB\n3gXsCzwPeCnwnlaFVGMsWrSIzZs3A7Bp0yZHN0mSWspikyRJ0iQXEbsCxwKnZea6zLwKuAx408i2\nmfm5zPxRZj6SmQPAl4E/bW1i7ajly5czNDQEwNDQEMuWLas4kSSpm1hskqQmiIjHRcQXyqkqD0bE\ntRHxsqpzSepaTwOGMvPmmmPXAfVGNo30QuDGek9ExNyIWBERK1avXt2AmGqUI444YpuPJUlqJotN\nktQcvcBvgBcBewKnAhdHxP4VZpLUvXYD1o44thbYfVsvioi/BvqBj9d7PjMXZmZ/ZvZPmzatIUHV\nGJlZdQRJUhez2CRJTZCZ6zPz9MxclZmbM/M7wK3AH1WdTVJXWgfsMeLYnsCDo70gIl4J/DPwssxc\n08RsaoKrrrpqi8c/+tGPKkoiSepGbV1schqKpMkiIvajmMay1VQUp6FIaoGbgd6IOKDm2KGMPj1u\nNvDvwMsz84YW5FODzZo1i56eHgB6eno46qijKk4kSeombV1swmkokiaBiNiJYoHdRZl508jnnYYi\nqdkycz1wCXBmROwaES8AXgF8cWTbiHgJRZ91bGb+rLVJ1Shz5sx5rNjU29vLCSecUHEiqTUcsCC1\nh7YuNjkNRVKni4gpFDdzjwDvrDiOpO52EjAVuAf4CvD2zLwxImZExLqImFG2O43il3xLyuPrIuK7\nFWXWBPX19TF79mwigtmzZ9PX11d1JKlVHLAgtYHeqgOMx7amoTRTz4Z7mXrTkla+5aQy5aFiPdLN\nO49cKkJj1bPhXmC/qmNonCIigC9Q/Mc7OjMfrTiSpC6WmfcCr6xz/HaKBcSHH7+4lbnUPHPmzGHV\nqlWOalJXKUdynl5z6DsRMTxgYVUVmaRu1DHFpm1NQ4mIucBcgBkzZtR59cTNnDmzoefrRitXFmuP\nzvwDiyUTt5//FjvT54BnALMyc2PVYSRJkrpNVQMWpG7XEcWm7U1DycyFwEKA/v7+hu7zOm/evEae\nrivNnz8fgHPOOafiJFLrRMRTgL8FHgbuKgY5AfC3mfnlyoJJkrrGeeedx/XXX8/ChQv5wAc+UHUc\nqeWqGrAgqc3XbIKtpqEc6zQUSZ0gM2/LzMjMnTNzt5ovC02SpKYbHBxk+fLlACxbtozBwcGKE0mt\nNZYBC27QIjVP2xeb+N00lJc7DUWSJEnavvPOO4/NmzcDsHnzZhYuXFhxIql1HLAgVa+ti00101AO\no5iGMrwjyvEVR5MkSZLa1hVXXLHF4+FRTlKXcMCCVLG2XrMpM28DYrsNJUmSJD2mZq3Auo+lycp1\nM6X20NYjmyRJkiSN33Of+9wtHh9++OEVJZFay3UzpfZgsUmSJEmaZFatWrXNx5IkNZPFJkmSJGmS\n+e1vf7vF4zvvvLOiJJKkbmSxSZIkSZpkXLNJklQli02SJEnSJPOc5zxnm48lSWomi02SJEnSJDNy\n2tzIaXWSJDWTxSZJkiRpknHNJklSlSw2SZIkSZIkqWEsNkmSJEmSJKlhLDZJkiRJkiSpYSw2SZIk\nSZIkqWEsNkmSJEmSJKlhLDZJkiRJkiSpYSw2SZIkSZIkqWEsNkmSJEmSJKlheqsOIEmSJEnSZDEw\nMEDPhgeYetOSqqOoi/VsGGRgYKiy93dkkyRJkiRJkhrGkU2SJEmSJDXI9OnTuevhXjYeeHTVUdTF\npt60hOnT96vs/R3ZJEmSJEmSpIax2CRJkiRJkqSGsdgkSZIkSZKkhmnrNZsiYh/gC8CfAWuAD2Tm\nV6pNJUljYx8mqZ2Mp0+KiFOA9wO7AF8H3p6ZD7cqaydYsGABK1eurDrGuMyfP7/qCFuZOXMm8+bN\nqzqGJhmvwaTqtXWxCfgM8AiwH3AYcHlEXJeZN1YbS5LGxD5MUjsZU58UEX8O/D3wEuBO4FLgjPJY\nS3RCIWdgYICNGzdWHWNc2vHvdGBgoC1z1bIg1pG8BpMq1rbFpojYFTgWODgz1wFXRcRlwJto4cWO\nJE2EfZikdjLOPmkO8IXhm7KIOBP4Sp12TfODH/yANWvWtOrtusb69eurjrCV9evXt/1/64GBAYtN\nHcRrMKk9tG2xCXgaMJSZN9ccuw44spo47akTfvM3nK8dh24P8zdWaoK26MN6NtzL1JuWtPItx2XK\nQ2uJzY9WHaPj5ZSd2LzzHlXHqKtnw70Uv1hWxcbTJx0EXDai3X4R0ZeZg82L+Dt77bVX248aevjh\nh9m8eXPVMUZVL9uUKe23XOuUKVN43OMeV3WMbdprr72qjqDx8RpsO7z+aox2vv6C6q/B2rnYtBuw\ndsSxtcDuIxtGxFxgLsCMGTOan0zjMnXq1KojSFWovA+bOXNmw87VLAMDQ21/Q9kJpk6dyvTp7VrQ\n2a8j/i12gTH3SWXbB0a0o2y7RbGpWf3X5z//+Yadq1k64Rd+11133WPfH3rooRUmGZ2/8FMTjKm/\na+Y9ZLv/f8/rr8Zo7+svqPoarJ2LTeuAkWXCPYEHRzbMzIXAQoD+/v5sfrT24f+cpbZVeR9m/yCp\nxpj7pDpt9yz/9BqsRif0sUceeeRj359zzjnVBZFaa0z9XTP7r07oH6Rma+di081Ab0QckJm/Ko8d\nCriom6ROYB8mqZ2Mp0+6sXzu4pp2d7dqCp0a58orr6w6glQFr8GkNtB+E7dLmbkeuAQ4MyJ2jYgX\nAK8AvlhtMknaPvswSe1knH3SYuAtEfHMiNgbOA24sGVhJWkHeA0mtYe2LTaVTgKmAvdQ7ILydrer\nlNRB7MMktZO6fVJEzIiIdRExAyAzlwIfA74P3AbcCnywosySNBFeg0kVa+dpdGTmvcArq84hSRNh\nHyapnYzWJ2Xm7RQL6tYe+yTwyRZFk6SG8hpMql67j2ySJEmSJElSB7HYJEmSJEmSpIaJzMm1S21E\nrKZYX0DtZV9gTdUhtIWnZOa0qkNoS/Zhbcn+qz3Zh7UZ+6+2ZR/Wfuy/2oz9V9uy/2o/Y+6/Jl2x\nSe0pIlZkZn/VOSRpvOy/JHUy+zBJncr+q7M5jU6SJEmSJEkNY7FJkiRJkiRJDWOxSa2ysOoAkjRB\n9l+SOpl9mKROZf/VwVyzSZIkSZIkSQ3jyCZJkiRJkiQ1jMUmSZIkSZIkNYzFJkmSJEmSJDWMxSZJ\nkiRJkiQ1jMUmSZIkSZIkNYzFJjVNRFwYESsqzvC+iDiyzvGMiHdWEElSg7RDH9MoEbEmIk6vOoek\n7hIRJ5bXRLuNsf2k6XelTtDon7mIeFpEnB4RezXqnM0SEV+PiCtrHp8eEWtqHlf+WSLiyoj4es3j\nLTJO8Jx/WfbL++9ovqpZbNJk9z7gyKpDSGqKs4ATqw4hSV3EflfqbE8DPgi0fbGpjs8Df17zuJM/\nS1forTqAJEkTkZm/rjqDJHUT+11JYxERPUBPZj7SqHNm5h3AHY06n5rPkU0dLiKOiIgfRMSGiBiM\niH+PiN3L54aHRj+nHOK3ISKuLR/vGhEXRMQDEXFLRLx+xHmvLIcuzo2IVRGxMSIuj4jpO5h3RkRc\nFBH3lnn+MyKeXvP8/mXm10bEeWW+OyLijIiYMuJcr4mIX5XZvh8Rzy5fe2L5/CqgD/hgeTxHTKnr\niYiPRMTqiLgnIj4TEY/bkc8nqXVqh5ZHxF4R8fmIuDMiHoqI2yPi32vaPikiLi5/1jdGxK8j4qya\n57cYBl0eO7LsNw6uObZzRHwsIn4TEQ9HxHURcfQ4c7+wfN1DEXFNRPzJxP8WJLWDiHh+RHwrIn4b\nEevL663ja54fviY7JCKWlW1uiohXjTjP8PXXGyJiZUSsjYjvRsSTatps1TfVvnasmSb4ObeY0jPW\nz1W2PSYiflb2wYMRsSQinrIjeaRuEhFPjIjzy3u3jRFxc0R8KCJ+b0S7D5T9x0MRcXdELI2IJ5T3\nQd8um91a/uyuGsf7vzUibqg579cjYs/yuQsjYkVEvDIibgQeAp5XPrfN+7+yzZPLPmFjFPeef1Pn\n/R+borYjn6U8/8cj4rSIuCsi1kXEl4c/S9mm7hTj4deO9e9sDFmi/Fz3RMSDEbEY2KNR56+axaYO\nFhF/CiwH7gJeDbwLOBq4YETTRcBXgWOBAL4OfAG4s3zdT4HFtRcypecD84B3A28BngV8cwfy7gNc\nBTwdeBvwWmBXYHlETB3R/GPAujLfl4B/Kr8fPlc/cBHwP8AxwLeAr404xzHAA+VnfX759T81z/8d\n8PvAG4F/Bf4WmD/RzyepUp8EXgCcQjHE+h+ArHl+MfBkYC7wMuDDwESKy1+nmELyEeDlwNXAtyLi\nsLG8OCJ+H/gucC9Fn3Ye8GVglwlkkdQ+9gd+AryVom/4BnBBjPhlHvAVimuWY4BfARfVuf56HvBO\niuuUucBzgIVNzNQI2/xcEfEm4BLg1xTXf28GbgamNSGLNFntC9wPvBeYTXH/8mZgwXCDiDiB4hro\nkxTXQ28HVlLcc/0P8J6y6aso7o2OGcsbR8SpFNcsPwBeWZ73AaC2GLM/xT3cP1Nca906lvu/iAjg\nMuBginvOd1Pckz1/G5Em/FlKrwdmUfSP7wb+gmKaXqudTHGfu5DiunAjxd/hpOA0us72L8D/y8zX\nDR+IiAHgihG/7fp4Zi4qnw/gcuDKzPzH8tjPKP5xvxz4XM3rHg88PzNvL9vdBlwVEbMzc+kE8p5C\n0bkclpn3luf8b2AV8NfAZ2ra/jAz/678fllEzKboSC4uj70f+AVwXGYmsDQidgI+OnyCzPzfiBgC\n7sjMn9TJsyozTyy//8+yePcqJtEPuNRFDgc+k5m1RecvjXj+9Zk5/FuwK8f7BhHxUoqLkSMz8wfl\n4f+KiKcB/wi8ZgyneRfFb/v+IjM3lOddPyKrpA6TmV8d/r681voh8CSKG5mv1jQ9OzPPL9tdA9wN\n/CXwbzVt9qDoI+4r2z0BODsipmbmxiZkaoRRP1cUI9P/Bbg0M2sLXd9qcAZpUsvMGygKI8Bj91Hr\ngfMjYl45Ze1w4L8y87M1L72k5jW/LL/938xcNZb3jWIB7n8APpWZ76556pIRTfuAWZl5bc1rz2L7\n938vA54N/HFm/rRscw1FcfpX9TJl5tqJfJYaUyn62XXl+60HvhgRz8jMX4zzXBMSxVTD9wPnZeap\n5eH/jIhlwA7NJmoXjmzqUBGxC0UF9+KI6B3+oqgcPwr8UU3zK2q+X1n++b3hA5n5ALCarf9R/89w\noals99/APRSd2ETMApYBa2vyPghcA/SPaPtfIx7/H8UF0rDnAt8uC03DxnvRsr33kNQ5rgXeGxEn\nlcWfes//czksesYE32MWxUjS/x7R717B1n3YaA4Hlg0XmkqXTjCPpDYREXtHxLnlL+YeLb/mUixg\nW+uxa4/MHKS4rhp57XH1cKGp9H/ln+O6+RhHpkbY1ud6OsVI8pEj7yWNQznl6l0R8X8RsZHiZ/rL\nFCO1h69trgWOjmIJksPLgsaOej5FcWZ7P8MDtYWm0lju/w4H7h4uNAFk5m1lm2ZZNlxoKl1KMQPo\nuU18z5GeDDyRYlRXrZFFvI5lsalz7Q30AJ/ldxcQjwIPAztR/OMddn/N94/UOTZ8fOcRx+6p8773\nUPxQTMS+wOtG5H0UePGIvGPJ9wSKAlmtkY+3Zyx/B5I6wzsppvn+E/DLKNZzO67m+dcBK4Czgdui\nWLvkpeN8j30p+p6RfdjpbN2HjeYJjOhby8LTuvrNJXWICyn6mX8F/ozihuV8tr6uGMu1R7021GnX\nqEyNsK3P1Vf++dsmvK/UTd4FfJyiMPJXFEWad5TPDf+8nU8xCum1FEul3B3Fuk47UnQa68/w3XWO\njeX+b6tro1K9Y40y2rXYRO9zJ+IJ9bLUedyxnEbXue6nWI/kdGBJnefvpLiw2BGPH+XYRC8W7qUY\nfXRWneceHOe57mLref7O+5e6VGbeTzHv/eSIeBbwPuDLEXF9Zv5fZg4AJ5bTOQ6n6Du/FREzyt/C\nPwT83ojT7j3i8b3AAMVaBRN1FyP61nKk6m71m0tqdxGxM8WUsXdk5r/VHG/WL3UfKv+s12cNL57b\n6kzbMlj+2cqbOGkyeg3w9eGlUAAi4pm1DTJzM8Uv1s6OiCcDx1OsU3kHW07XHY/an+E122iXdY6N\n5f5vq2uj0uMp1jBqhtGuxYbvc7fVzzbKXfWy1HncsRzZ1KEycz3Foo9Pz8wVdb7ubMDbPKd2ukm5\nptHjgZ9N8HxXAAcBN9bJ+8vtvXiEq4GXl2sQDHtFnXaOVpK6TGZeT7F45hTgwBHPbS7XcDuDYlHu\n4Z2Q7hjZlq0L9ldQ/BZqXb1+d4zxrgaOKi9qho1nQUtJ7edxFP3Nw8MHotgZuN51SSMMb/39jJr3\nezJb9mGtzrQtv6Qo1M+p4L2lyWQqNT/TpVF3mMzM32Tmv1AsozJclJrISMkfUxR9JvIzPJb7v6uB\n/SLiecMvKu9Bn7Odc0901CcU12K1v+g7hqJYNnw9V6+ffR6N3SnuNxQFp78acXyr3Tw7lSObOtv7\nKBYD30yxQ9KDFPN1/4JisdodtRq4PCI+SPFD/FGKdZwmsjg4FLsivBH4XkQsoLjw2A94EXBV7UKW\nY/BRiqGhF0XEBRQdwVvL5zbXtLsJ+IuIWEoxNPKXmTneUVSS2lxEXEUxrPznFBcLb6VYNPNnUWxl\n+58UO9LdTHET9ncU/4MfXgTyUuAtEXE2xSYKL6bY6aXWsvI8yyLio8CNFBcdhwE7Z+YHxhD1UxRD\n3r8TEZ+kWMfkAzTvN3eSmiwzH4iIq4F/ioi1FNchf0+xU1PDt7DOzDsiYgVwVkRsoCgq/QPFCIJK\nMm0n7+aIGB5t+mWKxckTeAnw1XEU66Vut4xiBPdPKRbPPh6YWdsgIs6j6At+QvHz/mLgAIqFqKEo\n/gL8bURcBGwoFx4fVWbeXy70/eGI+D2KWTWPo7jnPKMcPT6asdz/LQGuA/4jIt5PUVA7g+1PJxv3\nZ6mxkeI+918pRmz9K8UmBsNr5P2szHpuRJwG7ENx7712jOffrszcFBEfAz4eEWuAH1HsHv+Mbb+y\ncziyqYNl5lXACymmj30R+DbFD8FvqD9ndrz+H8UOAZ8CvkBxEzfh6SOZuQb4Y4oC0NkUi0l+DNgT\nuH6c51pBsWXlH1Gs03IsxRacsGUn8F6KG87LKarmtQunS5o8fgycSFF4v5hijYCXZeYdFEOhb6DY\nRvdbwCJgA/Bnwzs7ZeblFDdrr6YoPD2lbP+YckOCV1Gsh/AuisLTeRQLZ141lpDlBdnRZb5vACdR\nXIRt2NbrJLW9NwC3UBS1z6H4+V7cxPd7PXA7xU6WHwHO5Hc3XlVlGlVmfoXiWu1Ain56cfn9eNfb\nlLrZmRTF2g+Vfz5CsYRArR9T3B9eQFHEOQZ4a2Z+Ex5bePs9FNcz/01x/7hdmfnPFPdasygWtD4P\n2IvtLIUylvu/8vrqFRSbIZxftvt0+Vm2de4JfZbSRcD3Ke5xPwV8F3hLzbkfofi7Gx7U8XcUn/++\nrc60Yz5F0Ye/jaKP3o3ifn5SiC0385IKEXElsCYzX111lrGKiDdSFN3+IDNvrTqPJEmSJKl9RMQq\nirWv3lN1lsnOaXTqWBHxOYrhpPdRzOk9FbjcQpMkSZIkSdWx2KQJKXczGXUaZmYOtSBGH/DZ8s9B\n4GtMomGHkjpLuWHBtrYW3pQOJ5bUptrk2k5Si0VEDxCjPJ2ZuamVeXZEO32W7WTpij7VaXSakIi4\nkG3vSPDUzFzVmjSSVL2IOJFijYTRvDkzL2xNGkkaH6/tpO5UTit7yihP35aZ+7cuzY5pp89SLkvz\notGez8xRC1GThcUmTUhE7E+xwO1ori8XVpOkrhARfcBTt9Hk1swcbFUeSRoPr+2k7hQRh1DsLlfP\nw+PY4a1y7fRZIuLpwO6jPd8NO3FabJIkSZIkSVLDjDovW5IkSZIkSRovi02SJEmSJElqGItNkiRJ\nkiRJahiLTZIkSZIkSWoYi02SJEmSJElqmP8PYa5hfIWF4VcAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11a0d72b0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "row1 = 4\n",
    "col1 = 4\n",
    "\n",
    "fig, axs = pl.subplots(row1,col1, figsize=(20,15))\n",
    "fig.subplots_adjust(wspace=0.5, hspace=0.25)\n",
    "\n",
    "for num, name in enumerate(num_col_list):\n",
    "    i = num//row1\n",
    "    j = num%col1\n",
    "    graph = sns.boxplot(y = n_loan_no_out[name], ax = axs[i,j], linewidth = 1.5)\n",
    "    graph.set_xlabel(name, fontsize=15)\n",
    "    graph.set_ylabel('', fontsize=1)\n",
    "    graph.tick_params(labelsize=12)\n",
    "    \n",
    "pl.savefig('data_files/Outliers_removed.png', dpi=300)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Encoding categorical columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>loan_amnt</th>\n",
       "      <th>revol_bal</th>\n",
       "      <th>out_prncp</th>\n",
       "      <th>total_pymnt</th>\n",
       "      <th>total_pymnt_inv</th>\n",
       "      <th>total_rec_late_fee</th>\n",
       "      <th>recoveries</th>\n",
       "      <th>collection_recovery_fee</th>\n",
       "      <th>last_pymnt_amnt</th>\n",
       "      <th>total_rec_int</th>\n",
       "      <th>total_rec_prncp</th>\n",
       "      <th>int_rate</th>\n",
       "      <th>emp_length</th>\n",
       "      <th>issue_d</th>\n",
       "      <th>annual_inc</th>\n",
       "      <th>last_credit_pull_d</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>47505.000000</td>\n",
       "      <td>47505.000000</td>\n",
       "      <td>47505.000000</td>\n",
       "      <td>47505.000000</td>\n",
       "      <td>47505.000000</td>\n",
       "      <td>47505.000000</td>\n",
       "      <td>47505.000000</td>\n",
       "      <td>47505.0</td>\n",
       "      <td>47505.000000</td>\n",
       "      <td>47505.000000</td>\n",
       "      <td>47505.000000</td>\n",
       "      <td>47505.000000</td>\n",
       "      <td>47505.000000</td>\n",
       "      <td>47505.000000</td>\n",
       "      <td>4.750500e+04</td>\n",
       "      <td>47505.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>11087.298179</td>\n",
       "      <td>12971.519819</td>\n",
       "      <td>752.672663</td>\n",
       "      <td>11270.860094</td>\n",
       "      <td>10757.236806</td>\n",
       "      <td>0.655155</td>\n",
       "      <td>2.770159</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3930.298062</td>\n",
       "      <td>1651.463334</td>\n",
       "      <td>9615.971530</td>\n",
       "      <td>12.443560</td>\n",
       "      <td>5.158068</td>\n",
       "      <td>4.448016</td>\n",
       "      <td>6.892530e+04</td>\n",
       "      <td>1.712473</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>7072.186212</td>\n",
       "      <td>15444.077279</td>\n",
       "      <td>3589.589835</td>\n",
       "      <td>7962.167911</td>\n",
       "      <td>7933.454146</td>\n",
       "      <td>4.941502</td>\n",
       "      <td>184.057959</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5742.542030</td>\n",
       "      <td>1849.042323</td>\n",
       "      <td>6782.609291</td>\n",
       "      <td>4.178995</td>\n",
       "      <td>3.620926</td>\n",
       "      <td>2.770891</td>\n",
       "      <td>6.656398e+04</td>\n",
       "      <td>2.137256</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>500.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>5.320000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.896000e+03</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>5600.000000</td>\n",
       "      <td>4103.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>5185.336427</td>\n",
       "      <td>4690.440000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>273.500000</td>\n",
       "      <td>470.770000</td>\n",
       "      <td>4500.000000</td>\n",
       "      <td>9.630000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>4.276500e+04</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>10000.000000</td>\n",
       "      <td>9103.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>9517.260678</td>\n",
       "      <td>8914.600000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>960.310000</td>\n",
       "      <td>1033.660000</td>\n",
       "      <td>8000.000000</td>\n",
       "      <td>12.180000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>6.000000e+04</td>\n",
       "      <td>0.586301</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>15000.000000</td>\n",
       "      <td>16854.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>15785.064998</td>\n",
       "      <td>15216.050000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5520.860000</td>\n",
       "      <td>2127.120000</td>\n",
       "      <td>14000.000000</td>\n",
       "      <td>14.960000</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>8.500000e+04</td>\n",
       "      <td>3.169863</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>40000.000000</td>\n",
       "      <td>487589.000000</td>\n",
       "      <td>34765.050000</td>\n",
       "      <td>39449.503184</td>\n",
       "      <td>38767.460000</td>\n",
       "      <td>165.690000</td>\n",
       "      <td>29623.350000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>34677.040000</td>\n",
       "      <td>15175.460000</td>\n",
       "      <td>36000.000000</td>\n",
       "      <td>30.990000</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>8.300000e+06</td>\n",
       "      <td>10.427397</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          loan_amnt      revol_bal     out_prncp   total_pymnt  \\\n",
       "count  47505.000000   47505.000000  47505.000000  47505.000000   \n",
       "mean   11087.298179   12971.519819    752.672663  11270.860094   \n",
       "std     7072.186212   15444.077279   3589.589835   7962.167911   \n",
       "min      500.000000       0.000000      0.000000      0.000000   \n",
       "25%     5600.000000    4103.000000      0.000000   5185.336427   \n",
       "50%    10000.000000    9103.000000      0.000000   9517.260678   \n",
       "75%    15000.000000   16854.000000      0.000000  15785.064998   \n",
       "max    40000.000000  487589.000000  34765.050000  39449.503184   \n",
       "\n",
       "       total_pymnt_inv  total_rec_late_fee    recoveries  \\\n",
       "count     47505.000000        47505.000000  47505.000000   \n",
       "mean      10757.236806            0.655155      2.770159   \n",
       "std        7933.454146            4.941502    184.057959   \n",
       "min           0.000000            0.000000      0.000000   \n",
       "25%        4690.440000            0.000000      0.000000   \n",
       "50%        8914.600000            0.000000      0.000000   \n",
       "75%       15216.050000            0.000000      0.000000   \n",
       "max       38767.460000          165.690000  29623.350000   \n",
       "\n",
       "       collection_recovery_fee  last_pymnt_amnt  total_rec_int  \\\n",
       "count                  47505.0     47505.000000   47505.000000   \n",
       "mean                       0.0      3930.298062    1651.463334   \n",
       "std                        0.0      5742.542030    1849.042323   \n",
       "min                        0.0         0.000000       0.000000   \n",
       "25%                        0.0       273.500000     470.770000   \n",
       "50%                        0.0       960.310000    1033.660000   \n",
       "75%                        0.0      5520.860000    2127.120000   \n",
       "max                        0.0     34677.040000   15175.460000   \n",
       "\n",
       "       total_rec_prncp      int_rate    emp_length       issue_d  \\\n",
       "count     47505.000000  47505.000000  47505.000000  47505.000000   \n",
       "mean       9615.971530     12.443560      5.158068      4.448016   \n",
       "std        6782.609291      4.178995      3.620926      2.770891   \n",
       "min           0.000000      5.320000      0.000000      0.000000   \n",
       "25%        4500.000000      9.630000      2.000000      0.000000   \n",
       "50%        8000.000000     12.180000      5.000000      5.000000   \n",
       "75%       14000.000000     14.960000     10.000000      6.000000   \n",
       "max       36000.000000     30.990000     10.000000      9.000000   \n",
       "\n",
       "         annual_inc  last_credit_pull_d  \n",
       "count  4.750500e+04        47505.000000  \n",
       "mean   6.892530e+04            1.712473  \n",
       "std    6.656398e+04            2.137256  \n",
       "min    1.896000e+03            0.000000  \n",
       "25%    4.276500e+04            0.000000  \n",
       "50%    6.000000e+04            0.586301  \n",
       "75%    8.500000e+04            3.169863  \n",
       "max    8.300000e+06           10.427397  "
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "n_loan_no_out.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>loan_amnt</th>\n",
       "      <th>hardship_flag</th>\n",
       "      <th>revol_bal</th>\n",
       "      <th>out_prncp</th>\n",
       "      <th>total_pymnt</th>\n",
       "      <th>disbursement_method</th>\n",
       "      <th>total_pymnt_inv</th>\n",
       "      <th>total_rec_late_fee</th>\n",
       "      <th>recoveries</th>\n",
       "      <th>collection_recovery_fee</th>\n",
       "      <th>...</th>\n",
       "      <th>grade</th>\n",
       "      <th>emp_length</th>\n",
       "      <th>home_ownership</th>\n",
       "      <th>purpose</th>\n",
       "      <th>issue_d</th>\n",
       "      <th>loan_status</th>\n",
       "      <th>pymnt_plan</th>\n",
       "      <th>debt_settlement_flag</th>\n",
       "      <th>annual_inc</th>\n",
       "      <th>last_credit_pull_d</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>17000.0</td>\n",
       "      <td>0</td>\n",
       "      <td>15406.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>17282.802792</td>\n",
       "      <td>0</td>\n",
       "      <td>17282.80</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>Fully Paid</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>120000.0</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>5000.0</td>\n",
       "      <td>0</td>\n",
       "      <td>10276.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>5168.906377</td>\n",
       "      <td>0</td>\n",
       "      <td>5168.91</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>2</td>\n",
       "      <td>10.0</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>Fully Paid</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>68000.0</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>10000.0</td>\n",
       "      <td>0</td>\n",
       "      <td>7936.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>10365.368358</td>\n",
       "      <td>0</td>\n",
       "      <td>10365.37</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>10.0</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>Fully Paid</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>145000.0</td>\n",
       "      <td>0.082192</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>16000.0</td>\n",
       "      <td>0</td>\n",
       "      <td>31664.0</td>\n",
       "      <td>15301.23</td>\n",
       "      <td>1560.040000</td>\n",
       "      <td>0</td>\n",
       "      <td>1560.04</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>3</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>Problem Loan</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>100000.0</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>12000.0</td>\n",
       "      <td>0</td>\n",
       "      <td>7348.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>12324.571709</td>\n",
       "      <td>0</td>\n",
       "      <td>12324.57</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>5</td>\n",
       "      <td>4.0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>Fully Paid</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>44460.0</td>\n",
       "      <td>0.419178</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 28 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   loan_amnt  hardship_flag  revol_bal  out_prncp   total_pymnt  \\\n",
       "0    17000.0              0    15406.0       0.00  17282.802792   \n",
       "2     5000.0              0    10276.0       0.00   5168.906377   \n",
       "3    10000.0              0     7936.0       0.00  10365.368358   \n",
       "4    16000.0              0    31664.0   15301.23   1560.040000   \n",
       "5    12000.0              0     7348.0       0.00  12324.571709   \n",
       "\n",
       "   disbursement_method  total_pymnt_inv  total_rec_late_fee  recoveries  \\\n",
       "0                    0         17282.80                 0.0         0.0   \n",
       "2                    0          5168.91                 0.0         0.0   \n",
       "3                    0         10365.37                 0.0         0.0   \n",
       "4                    0          1560.04                 0.0         0.0   \n",
       "5                    0         12324.57                 0.0         0.0   \n",
       "\n",
       "   collection_recovery_fee         ...          grade  emp_length  \\\n",
       "0                      0.0         ...              0        10.0   \n",
       "2                      0.0         ...              2        10.0   \n",
       "3                      0.0         ...              1        10.0   \n",
       "4                      0.0         ...              3         0.0   \n",
       "5                      0.0         ...              5         4.0   \n",
       "\n",
       "   home_ownership  purpose  issue_d   loan_status  pymnt_plan  \\\n",
       "0               1        2        0    Fully Paid           0   \n",
       "2               5        2        0    Fully Paid           0   \n",
       "3               4        2        0    Fully Paid           0   \n",
       "4               5        2        0  Problem Loan           0   \n",
       "5               1        2        0    Fully Paid           0   \n",
       "\n",
       "   debt_settlement_flag  annual_inc  last_credit_pull_d  \n",
       "0                     0    120000.0            0.000000  \n",
       "2                     0     68000.0            0.000000  \n",
       "3                     0    145000.0            0.082192  \n",
       "4                     0    100000.0            0.000000  \n",
       "5                     0     44460.0            0.419178  \n",
       "\n",
       "[5 rows x 28 columns]"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "for col in (nn_col_list):\n",
    "    n_loan_no_out[col]=l_encoder.fit_transform(n_loan_no_out[col])\n",
    "n_loan_no_out.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0, 1])"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "n_loan_no_out.loan_status = l_encoder.fit_transform(n_loan_no_out.loan_status)\n",
    "n_loan_no_out.loan_status.unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "n_loan_no_out.to_csv('data_files/loans_2017_clean.csv', index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "n_loan_no_out = pd.read_csv('data_files/loans_2017_clean.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(47505, 28)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>loan_amnt</th>\n",
       "      <th>hardship_flag</th>\n",
       "      <th>revol_bal</th>\n",
       "      <th>out_prncp</th>\n",
       "      <th>total_pymnt</th>\n",
       "      <th>disbursement_method</th>\n",
       "      <th>total_pymnt_inv</th>\n",
       "      <th>total_rec_late_fee</th>\n",
       "      <th>recoveries</th>\n",
       "      <th>collection_recovery_fee</th>\n",
       "      <th>...</th>\n",
       "      <th>grade</th>\n",
       "      <th>emp_length</th>\n",
       "      <th>home_ownership</th>\n",
       "      <th>purpose</th>\n",
       "      <th>issue_d</th>\n",
       "      <th>loan_status</th>\n",
       "      <th>pymnt_plan</th>\n",
       "      <th>debt_settlement_flag</th>\n",
       "      <th>annual_inc</th>\n",
       "      <th>last_credit_pull_d</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>17000.0</td>\n",
       "      <td>0</td>\n",
       "      <td>15406.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>17282.802792</td>\n",
       "      <td>0</td>\n",
       "      <td>17282.80</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>120000.0</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>5000.0</td>\n",
       "      <td>0</td>\n",
       "      <td>10276.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>5168.906377</td>\n",
       "      <td>0</td>\n",
       "      <td>5168.91</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>2</td>\n",
       "      <td>10.0</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>68000.0</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>10000.0</td>\n",
       "      <td>0</td>\n",
       "      <td>7936.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>10365.368358</td>\n",
       "      <td>0</td>\n",
       "      <td>10365.37</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>10.0</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>145000.0</td>\n",
       "      <td>0.082192</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>16000.0</td>\n",
       "      <td>0</td>\n",
       "      <td>31664.0</td>\n",
       "      <td>15301.23</td>\n",
       "      <td>1560.040000</td>\n",
       "      <td>0</td>\n",
       "      <td>1560.04</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>3</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>100000.0</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>12000.0</td>\n",
       "      <td>0</td>\n",
       "      <td>7348.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>12324.571709</td>\n",
       "      <td>0</td>\n",
       "      <td>12324.57</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>5</td>\n",
       "      <td>4.0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>44460.0</td>\n",
       "      <td>0.419178</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>8000.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1362.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>8269.725017</td>\n",
       "      <td>0</td>\n",
       "      <td>8269.73</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>2</td>\n",
       "      <td>9.0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>36000.0</td>\n",
       "      <td>0.082192</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>12000.0</td>\n",
       "      <td>0</td>\n",
       "      <td>3560.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>12219.091570</td>\n",
       "      <td>0</td>\n",
       "      <td>12219.09</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>10.0</td>\n",
       "      <td>1</td>\n",
       "      <td>9</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>100000.0</td>\n",
       "      <td>0.334247</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>12000.0</td>\n",
       "      <td>0</td>\n",
       "      <td>24815.0</td>\n",
       "      <td>11435.79</td>\n",
       "      <td>783.660000</td>\n",
       "      <td>0</td>\n",
       "      <td>783.66</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>10.0</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>58356.0</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>33425.0</td>\n",
       "      <td>0</td>\n",
       "      <td>23601.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>35921.187179</td>\n",
       "      <td>0</td>\n",
       "      <td>35921.19</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>2</td>\n",
       "      <td>5.0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>31872.0</td>\n",
       "      <td>0.082192</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>5000.0</td>\n",
       "      <td>0</td>\n",
       "      <td>92.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>5002.704306</td>\n",
       "      <td>0</td>\n",
       "      <td>5002.70</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>2</td>\n",
       "      <td>3.0</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>60000.0</td>\n",
       "      <td>0.082192</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>10 rows × 28 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   loan_amnt  hardship_flag  revol_bal  out_prncp   total_pymnt  \\\n",
       "0    17000.0              0    15406.0       0.00  17282.802792   \n",
       "1     5000.0              0    10276.0       0.00   5168.906377   \n",
       "2    10000.0              0     7936.0       0.00  10365.368358   \n",
       "3    16000.0              0    31664.0   15301.23   1560.040000   \n",
       "4    12000.0              0     7348.0       0.00  12324.571709   \n",
       "5     8000.0              0     1362.0       0.00   8269.725017   \n",
       "6    12000.0              0     3560.0       0.00  12219.091570   \n",
       "7    12000.0              0    24815.0   11435.79    783.660000   \n",
       "8    33425.0              0    23601.0       0.00  35921.187179   \n",
       "9     5000.0              0       92.0       0.00   5002.704306   \n",
       "\n",
       "   disbursement_method  total_pymnt_inv  total_rec_late_fee  recoveries  \\\n",
       "0                    0         17282.80                 0.0         0.0   \n",
       "1                    0          5168.91                 0.0         0.0   \n",
       "2                    0         10365.37                 0.0         0.0   \n",
       "3                    0          1560.04                 0.0         0.0   \n",
       "4                    0         12324.57                 0.0         0.0   \n",
       "5                    0          8269.73                 0.0         0.0   \n",
       "6                    0         12219.09                 0.0         0.0   \n",
       "7                    0           783.66                 0.0         0.0   \n",
       "8                    0         35921.19                 0.0         0.0   \n",
       "9                    0          5002.70                 0.0         0.0   \n",
       "\n",
       "   collection_recovery_fee         ...          grade  emp_length  \\\n",
       "0                      0.0         ...              0        10.0   \n",
       "1                      0.0         ...              2        10.0   \n",
       "2                      0.0         ...              1        10.0   \n",
       "3                      0.0         ...              3         0.0   \n",
       "4                      0.0         ...              5         4.0   \n",
       "5                      0.0         ...              2         9.0   \n",
       "6                      0.0         ...              1        10.0   \n",
       "7                      0.0         ...              1        10.0   \n",
       "8                      0.0         ...              2         5.0   \n",
       "9                      0.0         ...              2         3.0   \n",
       "\n",
       "   home_ownership  purpose  issue_d  loan_status  pymnt_plan  \\\n",
       "0               1        2        0            0           0   \n",
       "1               5        2        0            0           0   \n",
       "2               4        2        0            0           0   \n",
       "3               5        2        0            1           0   \n",
       "4               1        2        0            0           0   \n",
       "5               1        2        0            0           0   \n",
       "6               1        9        0            0           0   \n",
       "7               5        2        0            1           0   \n",
       "8               1        2        0            0           0   \n",
       "9               4        4        0            0           0   \n",
       "\n",
       "   debt_settlement_flag  annual_inc  last_credit_pull_d  \n",
       "0                     0    120000.0            0.000000  \n",
       "1                     0     68000.0            0.000000  \n",
       "2                     0    145000.0            0.082192  \n",
       "3                     0    100000.0            0.000000  \n",
       "4                     0     44460.0            0.419178  \n",
       "5                     0     36000.0            0.082192  \n",
       "6                     0    100000.0            0.334247  \n",
       "7                     0     58356.0            0.000000  \n",
       "8                     0     31872.0            0.082192  \n",
       "9                     0     60000.0            0.082192  \n",
       "\n",
       "[10 rows x 28 columns]"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print(n_loan_no_out.shape)\n",
    "n_loan_no_out.head(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "### Building the model\n",
    "#### Spliting the data for the model\n",
    "\n",
    "Spliting into features and target"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Loans original data 47505, target: 47505, features: 47505\n"
     ]
    }
   ],
   "source": [
    "target = n_loan_no_out.loan_status.values\n",
    "features = n_loan_no_out.drop(['loan_status'], axis=1).values\n",
    "print ('Loans original data %d, target: %d, features: %d' % (n_loan_no_out.shape[0], target.shape[0], features.shape[0]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Building an raw model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# FUNCTION FROM SKELEARN DOCUMENTATION WILL BE USED FOR THE CONFUSION MATRIX PLOT\n",
    "import itertools\n",
    "def plot_confusion_matrix(cm, classes,\n",
    "                          normalize=False,\n",
    "                          title='Confusion matrix',\n",
    "                          cmap=pl.cm.Blues):\n",
    "    \"\"\" FROM SCIKIT LEARN DOCUMENTATION\n",
    "    This function prints and plots the confusion matrix.\n",
    "    Normalization can be applied by setting `normalize=True`.\n",
    "    \"\"\"\n",
    "    if normalize:\n",
    "        cm = cm.astype('float') / cm.sum(axis=1)[:, np.newaxis]\n",
    "        print(\"Normalized confusion matrix\")\n",
    "    else:\n",
    "        print('Confusion matrix, without normalization')\n",
    "    pl.imshow(cm, interpolation='nearest', cmap=cmap)\n",
    "    pl.title(title)\n",
    "    pl.colorbar()\n",
    "    tick_marks = np.arange(len(classes))\n",
    "    pl.xticks(tick_marks, classes, rotation=45)\n",
    "    pl.yticks(tick_marks, classes)\n",
    "\n",
    "    fmt = '.2f' if normalize else 'd'\n",
    "    thresh = cm.max() / 2.\n",
    "    for i, j in itertools.product(range(cm.shape[0]), range(cm.shape[1])):\n",
    "        pl.text(j, i, format(cm[i, j], fmt),\n",
    "                 horizontalalignment=\"center\",\n",
    "                 color=\"white\" if cm[i, j] > thresh else \"black\")\n",
    "\n",
    "    pl.tight_layout()\n",
    "    pl.ylabel('True label')\n",
    "    pl.xlabel('Predicted label')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Split data into training, valdation, and testing subsets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "X_train, X_test, y_train, y_test = train_test_split(features, target, test_size=0.25, random_state=1)\n",
    "X_train, X_val, y_train, y_val   = train_test_split(X_train, y_train, test_size=0.2, random_state=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import StandardScaler\n",
    "sc = StandardScaler()\n",
    "X_train = sc.fit_transform(X_train)\n",
    "X_val = sc.transform(X_val)\n",
    "X_test = sc.transform(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "from keras.models import Sequential\n",
    "from keras.layers import Dense, Activation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "classifier = Sequential()\n",
    "classifier.add(Dense(units = 14, kernel_initializer = 'uniform', activation = 'relu', input_dim = 27))\n",
    "classifier.add(Dense(units = 14, kernel_initializer = 'uniform', activation = 'relu'))\n",
    "classifier.add(Dense(units = 1, kernel_initializer = 'uniform', activation = 'sigmoid'))\n",
    "classifier.compile(optimizer = 'adam', loss = 'binary_crossentropy', metrics = ['accuracy'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Fitting the model with the training set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/10\n",
      "28502/28502 [==============================] - 11s - loss: 0.0727 - acc: 0.9799    \n",
      "Epoch 2/10\n",
      "28502/28502 [==============================] - 10s - loss: 0.0217 - acc: 0.9941    \n",
      "Epoch 3/10\n",
      "28502/28502 [==============================] - 10s - loss: 0.0183 - acc: 0.9948    \n",
      "Epoch 4/10\n",
      "28502/28502 [==============================] - 10s - loss: 0.0156 - acc: 0.9958    \n",
      "Epoch 5/10\n",
      "28502/28502 [==============================] - 10s - loss: 0.0141 - acc: 0.9964    \n",
      "Epoch 6/10\n",
      "28502/28502 [==============================] - 10s - loss: 0.0122 - acc: 0.9967    \n",
      "Epoch 7/10\n",
      "28502/28502 [==============================] - 10s - loss: 0.0118 - acc: 0.9968    \n",
      "Epoch 8/10\n",
      "28502/28502 [==============================] - 11s - loss: 0.0109 - acc: 0.9969    \n",
      "Epoch 9/10\n",
      "28502/28502 [==============================] - 11s - loss: 0.0106 - acc: 0.9969    \n",
      "Epoch 10/10\n",
      "28502/28502 [==============================] - 12s - loss: 0.0093 - acc: 0.9974    \n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x130a80ef0>"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Training with training set\n",
    "classifier.fit(X_train, y_train, batch_size = 10, epochs = 10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "### Crossvalidation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/10\n",
      "25651/25651 [==============================] - 14s - loss: 0.0962 - acc: 0.9683    \n",
      "Epoch 2/10\n",
      "25651/25651 [==============================] - 11s - loss: 0.0253 - acc: 0.9936    \n",
      "Epoch 3/10\n",
      "25651/25651 [==============================] - 11s - loss: 0.0200 - acc: 0.9946    \n",
      "Epoch 4/10\n",
      "25651/25651 [==============================] - 11s - loss: 0.0176 - acc: 0.9955    \n",
      "Epoch 5/10\n",
      "25651/25651 [==============================] - 11s - loss: 0.0159 - acc: 0.9959    \n",
      "Epoch 6/10\n",
      "25651/25651 [==============================] - 11s - loss: 0.0144 - acc: 0.9966    \n",
      "Epoch 7/10\n",
      "25651/25651 [==============================] - 11s - loss: 0.0141 - acc: 0.9965    \n",
      "Epoch 8/10\n",
      "25651/25651 [==============================] - 11s - loss: 0.0127 - acc: 0.9970    \n",
      "Epoch 9/10\n",
      "25651/25651 [==============================] - 11s - loss: 0.0125 - acc: 0.9972    \n",
      "Epoch 10/10\n",
      "25651/25651 [==============================] - 11s - loss: 0.0119 - acc: 0.9973    \n",
      "2560/2851 [=========================>....] - ETA: 0sEpoch 1/10\n",
      "25651/25651 [==============================] - 14s - loss: 0.0971 - acc: 0.9676    \n",
      "Epoch 2/10\n",
      "25651/25651 [==============================] - 11s - loss: 0.0234 - acc: 0.9936    \n",
      "Epoch 3/10\n",
      "25651/25651 [==============================] - 11s - loss: 0.0186 - acc: 0.9949    \n",
      "Epoch 4/10\n",
      "25651/25651 [==============================] - 11s - loss: 0.0162 - acc: 0.9955    \n",
      "Epoch 5/10\n",
      "25651/25651 [==============================] - 11s - loss: 0.0137 - acc: 0.9962    \n",
      "Epoch 6/10\n",
      "25651/25651 [==============================] - 11s - loss: 0.0130 - acc: 0.9966    \n",
      "Epoch 7/10\n",
      "25651/25651 [==============================] - 11s - loss: 0.0112 - acc: 0.9971    \n",
      "Epoch 8/10\n",
      "25651/25651 [==============================] - 11s - loss: 0.0113 - acc: 0.9970    \n",
      "Epoch 9/10\n",
      "25651/25651 [==============================] - 11s - loss: 0.0107 - acc: 0.9970    \n",
      "Epoch 10/10\n",
      "25651/25651 [==============================] - 11s - loss: 0.0097 - acc: 0.9973    \n",
      "2850/2851 [============================>.] - ETA: 0sEpoch 1/10\n",
      "25652/25652 [==============================] - 14s - loss: 0.0956 - acc: 0.9695    \n",
      "Epoch 2/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0235 - acc: 0.9942    \n",
      "Epoch 3/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0191 - acc: 0.9950    \n",
      "Epoch 4/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0167 - acc: 0.9958    \n",
      "Epoch 5/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0148 - acc: 0.9963    \n",
      "Epoch 6/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0141 - acc: 0.9965    \n",
      "Epoch 7/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0126 - acc: 0.9968    \n",
      "Epoch 8/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0126 - acc: 0.9969    \n",
      "Epoch 9/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0120 - acc: 0.9968    \n",
      "Epoch 10/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0110 - acc: 0.9972    \n",
      "2510/2850 [=========================>....] - ETA: 0sEpoch 1/10\n",
      "25652/25652 [==============================] - 14s - loss: 0.0773 - acc: 0.9797    \n",
      "Epoch 2/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0233 - acc: 0.9931    \n",
      "Epoch 3/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0193 - acc: 0.9947    \n",
      "Epoch 4/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0170 - acc: 0.9953    \n",
      "Epoch 5/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0155 - acc: 0.9959    \n",
      "Epoch 6/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0144 - acc: 0.9961    \n",
      "Epoch 7/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0134 - acc: 0.9967    \n",
      "Epoch 8/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0133 - acc: 0.9964    \n",
      "Epoch 9/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0117 - acc: 0.9966    \n",
      "Epoch 10/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0111 - acc: 0.9972    \n",
      "2820/2850 [============================>.] - ETA: 0sEpoch 1/10\n",
      "25652/25652 [==============================] - 14s - loss: 0.0793 - acc: 0.9793    \n",
      "Epoch 2/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0233 - acc: 0.9938    \n",
      "Epoch 3/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0192 - acc: 0.9947    \n",
      "Epoch 4/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0173 - acc: 0.9947    \n",
      "Epoch 5/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0153 - acc: 0.9958    \n",
      "Epoch 6/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0140 - acc: 0.9963    \n",
      "Epoch 7/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0128 - acc: 0.9965    \n",
      "Epoch 8/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0118 - acc: 0.9968    \n",
      "Epoch 9/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0107 - acc: 0.9970    \n",
      "Epoch 10/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0099 - acc: 0.9967    \n",
      "2770/2850 [============================>.] - ETA: 0sEpoch 1/10\n",
      "25652/25652 [==============================] - 14s - loss: 0.0742 - acc: 0.9815    \n",
      "Epoch 2/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0230 - acc: 0.9935    \n",
      "Epoch 3/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0191 - acc: 0.9950    \n",
      "Epoch 4/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0160 - acc: 0.9956    \n",
      "Epoch 5/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0144 - acc: 0.9961    \n",
      "Epoch 6/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0126 - acc: 0.9964    \n",
      "Epoch 7/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0121 - acc: 0.9967    \n",
      "Epoch 8/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0110 - acc: 0.9970    \n",
      "Epoch 9/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0102 - acc: 0.9970    \n",
      "Epoch 10/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0102 - acc: 0.9968    \n",
      "2730/2850 [===========================>..] - ETA: 0sEpoch 1/10\n",
      "25652/25652 [==============================] - 14s - loss: 0.0761 - acc: 0.9788    \n",
      "Epoch 2/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0221 - acc: 0.9936    \n",
      "Epoch 3/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0182 - acc: 0.9948    \n",
      "Epoch 4/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0160 - acc: 0.9956    \n",
      "Epoch 5/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0144 - acc: 0.9959    \n",
      "Epoch 6/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0133 - acc: 0.9965    \n",
      "Epoch 7/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0115 - acc: 0.9965    \n",
      "Epoch 8/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0117 - acc: 0.9965    \n",
      "Epoch 9/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0103 - acc: 0.9972    \n",
      "Epoch 10/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0098 - acc: 0.9972    \n",
      "2780/2850 [============================>.] - ETA: 0sEpoch 1/10\n",
      "25652/25652 [==============================] - 14s - loss: 0.0756 - acc: 0.9802    \n",
      "Epoch 2/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0223 - acc: 0.9937    \n",
      "Epoch 3/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0190 - acc: 0.9953    \n",
      "Epoch 4/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0170 - acc: 0.9953    \n",
      "Epoch 5/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0153 - acc: 0.9956    \n",
      "Epoch 6/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0139 - acc: 0.9961    \n",
      "Epoch 7/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0123 - acc: 0.9966    \n",
      "Epoch 8/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0127 - acc: 0.9965    \n",
      "Epoch 9/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0117 - acc: 0.9966    \n",
      "Epoch 10/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0108 - acc: 0.9970    \n",
      "2740/2850 [===========================>..] - ETA: 0sEpoch 1/10\n",
      "25652/25652 [==============================] - 14s - loss: 0.1041 - acc: 0.9632    \n",
      "Epoch 2/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0249 - acc: 0.9938    \n",
      "Epoch 3/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0194 - acc: 0.9950    \n",
      "Epoch 4/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0165 - acc: 0.9963    \n",
      "Epoch 5/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0149 - acc: 0.9963    \n",
      "Epoch 6/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0138 - acc: 0.9964    \n",
      "Epoch 7/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0130 - acc: 0.9966    \n",
      "Epoch 8/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0118 - acc: 0.9972    \n",
      "Epoch 9/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0112 - acc: 0.9970    \n",
      "Epoch 10/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0109 - acc: 0.9972    \n",
      "2680/2850 [===========================>..] - ETA: 0sEpoch 1/10\n",
      "25652/25652 [==============================] - 14s - loss: 0.0818 - acc: 0.9799    \n",
      "Epoch 2/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0226 - acc: 0.9941    \n",
      "Epoch 3/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0179 - acc: 0.9953    \n",
      "Epoch 4/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0156 - acc: 0.9959    \n",
      "Epoch 5/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0143 - acc: 0.9961    \n",
      "Epoch 6/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0128 - acc: 0.9967    \n",
      "Epoch 7/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0118 - acc: 0.9968    \n",
      "Epoch 8/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0112 - acc: 0.9971    \n",
      "Epoch 9/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0100 - acc: 0.9973    \n",
      "Epoch 10/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0100 - acc: 0.9972    \n",
      "2710/2850 [===========================>..] - ETA: 0s"
     ]
    }
   ],
   "source": [
    "# Crossvalidation on validation set\n",
    "from keras.wrappers.scikit_learn import KerasClassifier # keras wrapper for sklearn\n",
    "from sklearn.model_selection import cross_val_score\n",
    "\n",
    "def build_classifier():\n",
    "    classifier_cv = Sequential()\n",
    "    classifier_cv.add(Dense(units = 14, kernel_initializer = 'uniform', activation = 'relu', input_dim = 27))\n",
    "    classifier_cv.add(Dense(units = 14, kernel_initializer = 'uniform', activation = 'relu'))\n",
    "    classifier_cv.add(Dense(units = 1, kernel_initializer = 'uniform', activation = 'sigmoid'))\n",
    "    classifier_cv.compile(optimizer = 'adam', loss = 'binary_crossentropy', metrics = ['accuracy'])\n",
    "    return classifier_cv\n",
    "\n",
    "classifier_cv = KerasClassifier(build_fn = build_classifier, batch_size = 10, epochs = 10)\n",
    "accuracies = cross_val_score(estimator = classifier_cv, X = X_train, y = y_train, cv = 10)\n",
    "\n",
    "mean = accuracies.mean()\n",
    "variance = accuracies.std()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean cv accuracy = 99.6562% +/- 0.1512\n"
     ]
    }
   ],
   "source": [
    "print('Mean cv accuracy = {:.4f}% +/- {:.4f}'.format(mean *100, variance *100) )"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Validation and optimization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy = 99.6913%, Loss = 0.0118\n"
     ]
    }
   ],
   "source": [
    "# Accuracy on the validation set\n",
    "loss, accuracy = classifier.evaluate(X_val, y_val,batch_size=128, verbose=0)\n",
    "print(\"Accuracy = {:.4f}%, Loss = {:.4f}\".format(accuracy* 100, loss))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/10\n",
      "25651/25651 [==============================] - 11s - loss: 0.1078 - acc: 0.9612    \n",
      "Epoch 2/10\n",
      "25651/25651 [==============================] - 11s - loss: 0.0258 - acc: 0.9937    \n",
      "Epoch 3/10\n",
      "25651/25651 [==============================] - 10s - loss: 0.0194 - acc: 0.9950    \n",
      "Epoch 4/10\n",
      "25651/25651 [==============================] - 10s - loss: 0.0170 - acc: 0.9956    \n",
      "Epoch 5/10\n",
      "25651/25651 [==============================] - 11s - loss: 0.0145 - acc: 0.9964    \n",
      "Epoch 6/10\n",
      "25651/25651 [==============================] - 11s - loss: 0.0134 - acc: 0.9963    \n",
      "Epoch 7/10\n",
      "25651/25651 [==============================] - 10s - loss: 0.0122 - acc: 0.9966    \n",
      "Epoch 8/10\n",
      "25651/25651 [==============================] - 10s - loss: 0.0109 - acc: 0.9975    \n",
      "Epoch 9/10\n",
      "25651/25651 [==============================] - 11s - loss: 0.0104 - acc: 0.9970    \n",
      "Epoch 10/10\n",
      "25651/25651 [==============================] - 11s - loss: 0.0099 - acc: 0.9972    \n",
      "25080/25651 [============================>.] - ETA: 0sEpoch 1/10\n",
      "25651/25651 [==============================] - 11s - loss: 0.0814 - acc: 0.9789    \n",
      "Epoch 2/10\n",
      "25651/25651 [==============================] - 10s - loss: 0.0235 - acc: 0.9933    \n",
      "Epoch 3/10\n",
      "25651/25651 [==============================] - 10s - loss: 0.0194 - acc: 0.9942    \n",
      "Epoch 4/10\n",
      "25651/25651 [==============================] - 10s - loss: 0.0167 - acc: 0.9954    \n",
      "Epoch 5/10\n",
      "25651/25651 [==============================] - 10s - loss: 0.0151 - acc: 0.9959    \n",
      "Epoch 6/10\n",
      "25651/25651 [==============================] - 10s - loss: 0.0136 - acc: 0.9961    \n",
      "Epoch 7/10\n",
      "25651/25651 [==============================] - 10s - loss: 0.0126 - acc: 0.9961    \n",
      "Epoch 8/10\n",
      "25651/25651 [==============================] - 10s - loss: 0.0113 - acc: 0.9966    \n",
      "Epoch 9/10\n",
      "25651/25651 [==============================] - 10s - loss: 0.0111 - acc: 0.9969    \n",
      "Epoch 10/10\n",
      "25651/25651 [==============================] - 10s - loss: 0.0101 - acc: 0.9974    \n",
      "25590/25651 [============================>.] - ETA: 0sEpoch 1/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0771 - acc: 0.9792    \n",
      "Epoch 2/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0226 - acc: 0.9937    \n",
      "Epoch 3/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0184 - acc: 0.9952    \n",
      "Epoch 4/10\n",
      "25652/25652 [==============================] - 14s - loss: 0.0171 - acc: 0.9955    \n",
      "Epoch 5/10\n",
      "25652/25652 [==============================] - 10s - loss: 0.0152 - acc: 0.9962    \n",
      "Epoch 6/10\n",
      "25652/25652 [==============================] - 10s - loss: 0.0141 - acc: 0.9962    \n",
      "Epoch 7/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0133 - acc: 0.9964    \n",
      "Epoch 8/10\n",
      "25652/25652 [==============================] - 10s - loss: 0.0120 - acc: 0.9970    \n",
      "Epoch 9/10\n",
      "25652/25652 [==============================] - 10s - loss: 0.0109 - acc: 0.9970    \n",
      "Epoch 10/10\n",
      "25652/25652 [==============================] - 10s - loss: 0.0108 - acc: 0.9970    \n",
      "25520/25652 [============================>.] - ETA: 0sEpoch 1/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.1249 - acc: 0.9595    \n",
      "Epoch 2/10\n",
      "25652/25652 [==============================] - 9s - loss: 0.0456 - acc: 0.9931     \n",
      "Epoch 3/10\n",
      "25652/25652 [==============================] - 12s - loss: 0.0293 - acc: 0.9945    \n",
      "Epoch 4/10\n",
      "25652/25652 [==============================] - 13s - loss: 0.0227 - acc: 0.9949    \n",
      "Epoch 5/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0189 - acc: 0.9958    \n",
      "Epoch 6/10\n",
      "25652/25652 [==============================] - 10s - loss: 0.0170 - acc: 0.9962    \n",
      "Epoch 7/10\n",
      "25652/25652 [==============================] - 14s - loss: 0.0146 - acc: 0.9965    \n",
      "Epoch 8/10\n",
      "25652/25652 [==============================] - 11s - loss: 0.0136 - acc: 0.9969    \n",
      "Epoch 9/10\n",
      "25652/25652 [==============================] - 9s - loss: 0.0126 - acc: 0.9972     \n",
      "Epoch 10/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0124 - acc: 0.9968     \n",
      "25520/25652 [============================>.] - ETA: 0sEpoch 1/10\n",
      "25652/25652 [==============================] - 9s - loss: 0.1044 - acc: 0.9661     \n",
      "Epoch 2/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0267 - acc: 0.9935     \n",
      "Epoch 3/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0207 - acc: 0.9947     \n",
      "Epoch 4/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0178 - acc: 0.9951     \n",
      "Epoch 5/10\n",
      "25652/25652 [==============================] - 9s - loss: 0.0157 - acc: 0.9961     \n",
      "Epoch 6/10\n",
      "25652/25652 [==============================] - 9s - loss: 0.0147 - acc: 0.9961     \n",
      "Epoch 7/10\n",
      "25652/25652 [==============================] - 7s - loss: 0.0130 - acc: 0.9968     \n",
      "Epoch 8/10\n",
      "25652/25652 [==============================] - 7s - loss: 0.0125 - acc: 0.9970     \n",
      "Epoch 9/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0121 - acc: 0.9970     \n",
      "Epoch 10/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0106 - acc: 0.9973     \n",
      "25490/25652 [============================>.] - ETA: 0sEpoch 1/10\n",
      "25652/25652 [==============================] - 9s - loss: 0.0869 - acc: 0.9756     \n",
      "Epoch 2/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0226 - acc: 0.9933     \n",
      "Epoch 3/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0186 - acc: 0.9945     \n",
      "Epoch 4/10\n",
      "25652/25652 [==============================] - 9s - loss: 0.0168 - acc: 0.9956     \n",
      "Epoch 5/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0152 - acc: 0.9955     \n",
      "Epoch 6/10\n",
      "25652/25652 [==============================] - 9s - loss: 0.0134 - acc: 0.9964     \n",
      "Epoch 7/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0124 - acc: 0.9965     \n",
      "Epoch 8/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0119 - acc: 0.9968     \n",
      "Epoch 9/10\n",
      "25652/25652 [==============================] - 7s - loss: 0.0108 - acc: 0.9970     \n",
      "Epoch 10/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0104 - acc: 0.9971     \n",
      "25060/25652 [============================>.] - ETA: 0sEpoch 1/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0746 - acc: 0.9804     \n",
      "Epoch 2/10\n",
      "25652/25652 [==============================] - 7s - loss: 0.0222 - acc: 0.9942     \n",
      "Epoch 3/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0185 - acc: 0.9950     \n",
      "Epoch 4/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0161 - acc: 0.9956     \n",
      "Epoch 5/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0144 - acc: 0.9961     \n",
      "Epoch 6/10\n",
      "25652/25652 [==============================] - 9s - loss: 0.0131 - acc: 0.9967     \n",
      "Epoch 7/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0117 - acc: 0.9971     \n",
      "Epoch 8/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0120 - acc: 0.9969     \n",
      "Epoch 9/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0110 - acc: 0.9971     \n",
      "Epoch 10/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0105 - acc: 0.9972     \n",
      "25180/25652 [============================>.] - ETA: 0sEpoch 1/10\n",
      "25652/25652 [==============================] - 10s - loss: 0.0822 - acc: 0.9787    \n",
      "Epoch 2/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0230 - acc: 0.9937     \n",
      "Epoch 3/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0186 - acc: 0.9949     \n",
      "Epoch 4/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0163 - acc: 0.9956     \n",
      "Epoch 5/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0150 - acc: 0.9959     \n",
      "Epoch 6/10\n",
      "25652/25652 [==============================] - 9s - loss: 0.0137 - acc: 0.9965     \n",
      "Epoch 7/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0128 - acc: 0.9966     \n",
      "Epoch 8/10\n",
      "25652/25652 [==============================] - 9s - loss: 0.0125 - acc: 0.9968     \n",
      "Epoch 9/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0116 - acc: 0.9969     \n",
      "Epoch 10/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0109 - acc: 0.9973     \n",
      "25020/25652 [============================>.] - ETA: 0sEpoch 1/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0833 - acc: 0.9806     \n",
      "Epoch 2/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0233 - acc: 0.9938     \n",
      "Epoch 3/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0193 - acc: 0.9948     \n",
      "Epoch 4/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0166 - acc: 0.9956     \n",
      "Epoch 5/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0144 - acc: 0.9964     \n",
      "Epoch 6/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0132 - acc: 0.9966     \n",
      "Epoch 7/10\n",
      "25652/25652 [==============================] - 9s - loss: 0.0122 - acc: 0.9965     \n",
      "Epoch 8/10\n",
      "25652/25652 [==============================] - 9s - loss: 0.0116 - acc: 0.9969     \n",
      "Epoch 9/10\n",
      "25652/25652 [==============================] - 9s - loss: 0.0111 - acc: 0.9971     \n",
      "Epoch 10/10\n",
      "25652/25652 [==============================] - 9s - loss: 0.0101 - acc: 0.9973     \n",
      "24890/25652 [============================>.] - ETA: 0sEpoch 1/10\n",
      "25652/25652 [==============================] - 9s - loss: 0.0751 - acc: 0.9809     \n",
      "Epoch 2/10\n",
      "25652/25652 [==============================] - 9s - loss: 0.0208 - acc: 0.9939     \n",
      "Epoch 3/10\n",
      "25652/25652 [==============================] - 9s - loss: 0.0173 - acc: 0.9954     \n",
      "Epoch 4/10\n",
      "25652/25652 [==============================] - 9s - loss: 0.0154 - acc: 0.9958     \n",
      "Epoch 5/10\n",
      "25652/25652 [==============================] - 9s - loss: 0.0138 - acc: 0.9963     \n",
      "Epoch 6/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0126 - acc: 0.9965     \n",
      "Epoch 7/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0119 - acc: 0.9967     \n",
      "Epoch 8/10\n",
      "25652/25652 [==============================] - 9s - loss: 0.0116 - acc: 0.9968     \n",
      "Epoch 9/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0107 - acc: 0.9971     \n",
      "Epoch 10/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0102 - acc: 0.9971     \n",
      "25230/25652 [============================>.] - ETA: 0s ETAEpoch 1/10\n",
      "25651/25651 [==============================] - 8s - loss: 0.0908 - acc: 0.9744     \n",
      "Epoch 2/10\n",
      "25651/25651 [==============================] - 8s - loss: 0.0278 - acc: 0.9934     \n",
      "Epoch 3/10\n",
      "25651/25651 [==============================] - 8s - loss: 0.0242 - acc: 0.9945     \n",
      "Epoch 4/10\n",
      "25651/25651 [==============================] - 7s - loss: 0.0223 - acc: 0.9951     \n",
      "Epoch 5/10\n",
      "25651/25651 [==============================] - 8s - loss: 0.0213 - acc: 0.9955     \n",
      "Epoch 6/10\n",
      "25651/25651 [==============================] - 8s - loss: 0.0199 - acc: 0.9957     \n",
      "Epoch 7/10\n",
      "25651/25651 [==============================] - 7s - loss: 0.0177 - acc: 0.9957     \n",
      "Epoch 8/10\n",
      "25651/25651 [==============================] - 9s - loss: 0.0176 - acc: 0.9959     \n",
      "Epoch 9/10\n",
      "25651/25651 [==============================] - 9s - loss: 0.0155 - acc: 0.9966     \n",
      "Epoch 10/10\n",
      "25651/25651 [==============================] - 9s - loss: 0.0156 - acc: 0.9969     \n",
      "25450/25651 [============================>.] - ETA: 0sEpoch 1/10\n",
      "25651/25651 [==============================] - 10s - loss: 0.0906 - acc: 0.9736    \n",
      "Epoch 2/10\n",
      "25651/25651 [==============================] - 9s - loss: 0.0273 - acc: 0.9935     \n",
      "Epoch 3/10\n",
      "25651/25651 [==============================] - 9s - loss: 0.0230 - acc: 0.9947     \n",
      "Epoch 4/10\n",
      "25651/25651 [==============================] - 9s - loss: 0.0222 - acc: 0.9951     \n",
      "Epoch 5/10\n",
      "25651/25651 [==============================] - 9s - loss: 0.0214 - acc: 0.9952     \n",
      "Epoch 6/10\n",
      "25651/25651 [==============================] - 9s - loss: 0.0202 - acc: 0.9958     \n",
      "Epoch 7/10\n",
      "25651/25651 [==============================] - 9s - loss: 0.0200 - acc: 0.9959     \n",
      "Epoch 8/10\n",
      "25651/25651 [==============================] - 11s - loss: 0.0192 - acc: 0.9959    \n",
      "Epoch 9/10\n",
      "25651/25651 [==============================] - 8s - loss: 0.0183 - acc: 0.9960     \n",
      "Epoch 10/10\n",
      "25651/25651 [==============================] - 8s - loss: 0.0179 - acc: 0.9965     \n",
      "25020/25651 [============================>.] - ETA: 0sEpoch 1/10\n",
      "25652/25652 [==============================] - 9s - loss: 0.1100 - acc: 0.9628     \n",
      "Epoch 2/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0280 - acc: 0.9938     \n",
      "Epoch 3/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0244 - acc: 0.9947     \n",
      "Epoch 4/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0231 - acc: 0.9952     \n",
      "Epoch 5/10\n",
      "25652/25652 [==============================] - 10s - loss: 0.0220 - acc: 0.9954    \n",
      "Epoch 6/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0196 - acc: 0.9956     \n",
      "Epoch 7/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0182 - acc: 0.9961     \n",
      "Epoch 8/10\n",
      "25652/25652 [==============================] - 10s - loss: 0.0163 - acc: 0.9965    \n",
      "Epoch 9/10\n",
      "25652/25652 [==============================] - 9s - loss: 0.0158 - acc: 0.9968     \n",
      "Epoch 10/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0159 - acc: 0.9967     \n",
      "25580/25652 [============================>.] - ETA: 0sEpoch 1/10\n",
      "25652/25652 [==============================] - 10s - loss: 0.0856 - acc: 0.9765    \n",
      "Epoch 2/10\n",
      "25652/25652 [==============================] - 9s - loss: 0.0276 - acc: 0.9934     \n",
      "Epoch 3/10\n",
      "25652/25652 [==============================] - 9s - loss: 0.0239 - acc: 0.9945     \n",
      "Epoch 4/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0219 - acc: 0.9949     \n",
      "Epoch 5/10\n",
      "25652/25652 [==============================] - 9s - loss: 0.0196 - acc: 0.9956     \n",
      "Epoch 6/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0187 - acc: 0.9961     \n",
      "Epoch 7/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0178 - acc: 0.9967     \n",
      "Epoch 8/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0169 - acc: 0.9966     \n",
      "Epoch 9/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0168 - acc: 0.9965     \n",
      "Epoch 10/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0163 - acc: 0.9970     \n",
      "25310/25652 [============================>.] - ETA: 0sEpoch 1/10\n",
      "25652/25652 [==============================] - 10s - loss: 0.1116 - acc: 0.9636    \n",
      "Epoch 2/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0299 - acc: 0.9931     \n",
      "Epoch 3/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0257 - acc: 0.9942     \n",
      "Epoch 4/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0239 - acc: 0.9951     \n",
      "Epoch 5/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0214 - acc: 0.9956     \n",
      "Epoch 6/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0206 - acc: 0.9957     \n",
      "Epoch 7/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0189 - acc: 0.9961     \n",
      "Epoch 8/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0187 - acc: 0.9965     \n",
      "Epoch 9/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0168 - acc: 0.9968     \n",
      "Epoch 10/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0171 - acc: 0.9965     \n",
      "24910/25652 [============================>.] - ETA: 0sEpoch 1/10\n",
      "25652/25652 [==============================] - 9s - loss: 0.1070 - acc: 0.9644     \n",
      "Epoch 2/10\n",
      "25652/25652 [==============================] - 9s - loss: 0.0278 - acc: 0.9935     \n",
      "Epoch 3/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0232 - acc: 0.9947     \n",
      "Epoch 4/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0211 - acc: 0.9953     \n",
      "Epoch 5/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0193 - acc: 0.9959     \n",
      "Epoch 6/10\n",
      "25652/25652 [==============================] - 9s - loss: 0.0167 - acc: 0.9961     \n",
      "Epoch 7/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0162 - acc: 0.9963     \n",
      "Epoch 8/10\n",
      "25652/25652 [==============================] - 9s - loss: 0.0155 - acc: 0.9965     \n",
      "Epoch 9/10\n",
      "25652/25652 [==============================] - 9s - loss: 0.0155 - acc: 0.9968     \n",
      "Epoch 10/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0133 - acc: 0.9968     \n",
      "25290/25652 [============================>.] - ETA: 0sEpoch 1/10\n",
      "25652/25652 [==============================] - 9s - loss: 0.1377 - acc: 0.9528     \n",
      "Epoch 2/10\n",
      "25652/25652 [==============================] - 7s - loss: 0.0336 - acc: 0.9928     \n",
      "Epoch 3/10\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "25652/25652 [==============================] - 8s - loss: 0.0260 - acc: 0.9942     \n",
      "Epoch 4/10\n",
      "25652/25652 [==============================] - 7s - loss: 0.0238 - acc: 0.9947     \n",
      "Epoch 5/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0217 - acc: 0.9955     \n",
      "Epoch 6/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0197 - acc: 0.9958     \n",
      "Epoch 7/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0191 - acc: 0.9959     \n",
      "Epoch 8/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0177 - acc: 0.9963     \n",
      "Epoch 9/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0176 - acc: 0.9964     \n",
      "Epoch 10/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0173 - acc: 0.9966     \n",
      "25410/25652 [============================>.] - ETA: 0sEpoch 1/10\n",
      "25652/25652 [==============================] - 9s - loss: 0.0956 - acc: 0.9727     \n",
      "Epoch 2/10\n",
      "25652/25652 [==============================] - 7s - loss: 0.0259 - acc: 0.9937     \n",
      "Epoch 3/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0229 - acc: 0.9944     \n",
      "Epoch 4/10\n",
      "25652/25652 [==============================] - 7s - loss: 0.0221 - acc: 0.9951     \n",
      "Epoch 5/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0203 - acc: 0.9958     \n",
      "Epoch 6/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0193 - acc: 0.9959     \n",
      "Epoch 7/10\n",
      "25652/25652 [==============================] - 7s - loss: 0.0180 - acc: 0.9961     \n",
      "Epoch 8/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0174 - acc: 0.9963     \n",
      "Epoch 9/10\n",
      "25652/25652 [==============================] - 7s - loss: 0.0170 - acc: 0.9965     \n",
      "Epoch 10/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0164 - acc: 0.9966     \n",
      "25550/25652 [============================>.] - ETA: 0sEpoch 1/10\n",
      "25652/25652 [==============================] - 9s - loss: 0.0823 - acc: 0.9787     \n",
      "Epoch 2/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0255 - acc: 0.9935     \n",
      "Epoch 3/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0221 - acc: 0.9950     \n",
      "Epoch 4/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0205 - acc: 0.9956     \n",
      "Epoch 5/10\n",
      "25652/25652 [==============================] - 7s - loss: 0.0187 - acc: 0.9959     \n",
      "Epoch 6/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0181 - acc: 0.9963     \n",
      "Epoch 7/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0167 - acc: 0.9963     \n",
      "Epoch 8/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0166 - acc: 0.9965     \n",
      "Epoch 9/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0157 - acc: 0.9968     \n",
      "Epoch 10/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0155 - acc: 0.9969     \n",
      "25190/25652 [============================>.] - ETA: 0sEpoch 1/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.1260 - acc: 0.9567     \n",
      "Epoch 2/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0290 - acc: 0.9935     \n",
      "Epoch 3/10\n",
      "25652/25652 [==============================] - 7s - loss: 0.0236 - acc: 0.9948     \n",
      "Epoch 4/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0218 - acc: 0.9954     \n",
      "Epoch 5/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0201 - acc: 0.9955     \n",
      "Epoch 6/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0193 - acc: 0.9956     \n",
      "Epoch 7/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0180 - acc: 0.9963     \n",
      "Epoch 8/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0166 - acc: 0.9964     \n",
      "Epoch 9/10\n",
      "25652/25652 [==============================] - ETA: 0s - loss: 0.0169 - acc: 0.996 - 8s - loss: 0.0168 - acc: 0.9965     \n",
      "Epoch 10/10\n",
      "25652/25652 [==============================] - 8s - loss: 0.0155 - acc: 0.9967     \n",
      "25580/25652 [============================>.] - ETA: 0s EEpoch 1/20\n",
      "25651/25651 [==============================] - 9s - loss: 0.0983 - acc: 0.9698     \n",
      "Epoch 2/20\n",
      "25651/25651 [==============================] - 9s - loss: 0.0249 - acc: 0.9937     \n",
      "Epoch 3/20\n",
      "25651/25651 [==============================] - 9s - loss: 0.0203 - acc: 0.9947     \n",
      "Epoch 4/20\n",
      "25651/25651 [==============================] - 8s - loss: 0.0171 - acc: 0.9956     \n",
      "Epoch 5/20\n",
      "25651/25651 [==============================] - 8s - loss: 0.0150 - acc: 0.9962     \n",
      "Epoch 6/20\n",
      "25651/25651 [==============================] - 9s - loss: 0.0131 - acc: 0.9965     \n",
      "Epoch 7/20\n",
      "25651/25651 [==============================] - 9s - loss: 0.0122 - acc: 0.9969     \n",
      "Epoch 8/20\n",
      "25651/25651 [==============================] - 9s - loss: 0.0113 - acc: 0.9967     \n",
      "Epoch 9/20\n",
      "25651/25651 [==============================] - 9s - loss: 0.0102 - acc: 0.9972     \n",
      "Epoch 10/20\n",
      "25651/25651 [==============================] - 9s - loss: 0.0099 - acc: 0.9973     \n",
      "Epoch 11/20\n",
      "25651/25651 [==============================] - 9s - loss: 0.0094 - acc: 0.9976     \n",
      "Epoch 12/20\n",
      "25651/25651 [==============================] - 9s - loss: 0.0086 - acc: 0.9977     \n",
      "Epoch 13/20\n",
      "25651/25651 [==============================] - 10s - loss: 0.0084 - acc: 0.9975    \n",
      "Epoch 14/20\n",
      "25651/25651 [==============================] - 9s - loss: 0.0082 - acc: 0.9978     \n",
      "Epoch 15/20\n",
      "25651/25651 [==============================] - 9s - loss: 0.0076 - acc: 0.9976     \n",
      "Epoch 16/20\n",
      "25651/25651 [==============================] - 9s - loss: 0.0073 - acc: 0.9982     \n",
      "Epoch 17/20\n",
      "25651/25651 [==============================] - 9s - loss: 0.0070 - acc: 0.9981     \n",
      "Epoch 18/20\n",
      "25651/25651 [==============================] - 9s - loss: 0.0073 - acc: 0.9979     \n",
      "Epoch 19/20\n",
      "25651/25651 [==============================] - 8s - loss: 0.0062 - acc: 0.9982     \n",
      "Epoch 20/20\n",
      "25651/25651 [==============================] - 8s - loss: 0.0063 - acc: 0.9984     \n",
      "25290/25651 [============================>.] - ETA: 0sEpoch 1/20\n",
      "25651/25651 [==============================] - 10s - loss: 0.0818 - acc: 0.9781    \n",
      "Epoch 2/20\n",
      "25651/25651 [==============================] - 9s - loss: 0.0229 - acc: 0.9935     \n",
      "Epoch 3/20\n",
      "25651/25651 [==============================] - 10s - loss: 0.0191 - acc: 0.9945    \n",
      "Epoch 4/20\n",
      "25651/25651 [==============================] - 9s - loss: 0.0163 - acc: 0.9954     \n",
      "Epoch 5/20\n",
      "25651/25651 [==============================] - 9s - loss: 0.0147 - acc: 0.9960     \n",
      "Epoch 6/20\n",
      "25651/25651 [==============================] - 9s - loss: 0.0128 - acc: 0.9963     \n",
      "Epoch 7/20\n",
      "25651/25651 [==============================] - 8s - loss: 0.0123 - acc: 0.9968     \n",
      "Epoch 8/20\n",
      "25651/25651 [==============================] - 9s - loss: 0.0114 - acc: 0.9966     \n",
      "Epoch 9/20\n",
      "25651/25651 [==============================] - 10s - loss: 0.0104 - acc: 0.9971    \n",
      "Epoch 10/20\n",
      "25651/25651 [==============================] - 11s - loss: 0.0102 - acc: 0.9970    \n",
      "Epoch 11/20\n",
      "25651/25651 [==============================] - 11s - loss: 0.0097 - acc: 0.9973    \n",
      "Epoch 12/20\n",
      "25651/25651 [==============================] - 11s - loss: 0.0087 - acc: 0.9975    \n",
      "Epoch 13/20\n",
      "25651/25651 [==============================] - 11s - loss: 0.0092 - acc: 0.9975    \n",
      "Epoch 14/20\n",
      "25651/25651 [==============================] - 11s - loss: 0.0090 - acc: 0.9975    \n",
      "Epoch 15/20\n",
      "25651/25651 [==============================] - 11s - loss: 0.0078 - acc: 0.9978    \n",
      "Epoch 16/20\n",
      "25651/25651 [==============================] - 11s - loss: 0.0078 - acc: 0.9977    \n",
      "Epoch 17/20\n",
      "25651/25651 [==============================] - 11s - loss: 0.0080 - acc: 0.9974    \n",
      "Epoch 18/20\n",
      "25651/25651 [==============================] - 11s - loss: 0.0076 - acc: 0.9976    \n",
      "Epoch 19/20\n",
      "25651/25651 [==============================] - 11s - loss: 0.0076 - acc: 0.9979    \n",
      "Epoch 20/20\n",
      "25651/25651 [==============================] - 11s - loss: 0.0080 - acc: 0.9977    \n",
      "25330/25651 [============================>.] - ETA: 0sEpoch 1/20\n",
      "25652/25652 [==============================] - 12s - loss: 0.0734 - acc: 0.9815    \n",
      "Epoch 2/20\n",
      "25652/25652 [==============================] - 11s - loss: 0.0227 - acc: 0.9938    \n",
      "Epoch 3/20\n",
      "25652/25652 [==============================] - 11s - loss: 0.0192 - acc: 0.9950    \n",
      "Epoch 4/20\n",
      "25652/25652 [==============================] - 11s - loss: 0.0165 - acc: 0.9954    \n",
      "Epoch 5/20\n",
      "25652/25652 [==============================] - 11s - loss: 0.0152 - acc: 0.9956    \n",
      "Epoch 6/20\n",
      "25652/25652 [==============================] - 11s - loss: 0.0138 - acc: 0.9965    \n",
      "Epoch 7/20\n",
      "25652/25652 [==============================] - 12s - loss: 0.0129 - acc: 0.9971    \n",
      "Epoch 8/20\n",
      "25652/25652 [==============================] - 11s - loss: 0.0121 - acc: 0.9969    \n",
      "Epoch 9/20\n",
      "25652/25652 [==============================] - 11s - loss: 0.0114 - acc: 0.9969    \n",
      "Epoch 10/20\n",
      "25652/25652 [==============================] - 12s - loss: 0.0114 - acc: 0.9971    \n",
      "Epoch 11/20\n",
      "25652/25652 [==============================] - 11s - loss: 0.0105 - acc: 0.9973    \n",
      "Epoch 12/20\n",
      "25652/25652 [==============================] - 11s - loss: 0.0094 - acc: 0.9974    \n",
      "Epoch 13/20\n",
      "25652/25652 [==============================] - 11s - loss: 0.0091 - acc: 0.9976    \n",
      "Epoch 14/20\n",
      "25652/25652 [==============================] - 11s - loss: 0.0085 - acc: 0.9973    \n",
      "Epoch 15/20\n",
      "25652/25652 [==============================] - 11s - loss: 0.0082 - acc: 0.9978    \n",
      "Epoch 16/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0079 - acc: 0.9975     \n",
      "Epoch 17/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0075 - acc: 0.9979     \n",
      "Epoch 18/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0074 - acc: 0.9977     \n",
      "Epoch 19/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0076 - acc: 0.9978     \n",
      "Epoch 20/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0062 - acc: 0.9981     \n",
      "25560/25652 [============================>.] - ETA: 0sEpoch 1/20\n",
      "25652/25652 [==============================] - 10s - loss: 0.0887 - acc: 0.9722    \n",
      "Epoch 2/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0237 - acc: 0.9937     \n",
      "Epoch 3/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0191 - acc: 0.9947     \n",
      "Epoch 4/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0169 - acc: 0.9956     \n",
      "Epoch 5/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0154 - acc: 0.9961     \n",
      "Epoch 6/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0141 - acc: 0.9965     \n",
      "Epoch 7/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0133 - acc: 0.9965     \n",
      "Epoch 8/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0122 - acc: 0.9966     \n",
      "Epoch 9/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0117 - acc: 0.9970     \n",
      "Epoch 10/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0115 - acc: 0.9970     \n",
      "Epoch 11/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0106 - acc: 0.9972     \n",
      "Epoch 12/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0100 - acc: 0.9975     \n",
      "Epoch 13/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0087 - acc: 0.9975     \n",
      "Epoch 14/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0094 - acc: 0.9976     \n",
      "Epoch 15/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0080 - acc: 0.9979     \n",
      "Epoch 16/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0086 - acc: 0.9975     \n",
      "Epoch 17/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0080 - acc: 0.9979     \n",
      "Epoch 18/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0077 - acc: 0.9978     \n",
      "Epoch 19/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0078 - acc: 0.9978     \n",
      "Epoch 20/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0069 - acc: 0.9979     \n",
      "25290/25652 [============================>.] - ETA: 0sEpoch 1/20\n",
      "25652/25652 [==============================] - 10s - loss: 0.0915 - acc: 0.9720    \n",
      "Epoch 2/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0240 - acc: 0.9940     \n",
      "Epoch 3/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0194 - acc: 0.9946     \n",
      "Epoch 4/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0168 - acc: 0.9959     \n",
      "Epoch 5/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0150 - acc: 0.9961     \n",
      "Epoch 6/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0137 - acc: 0.9965     \n",
      "Epoch 7/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0126 - acc: 0.9967     \n",
      "Epoch 8/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0119 - acc: 0.9967     \n",
      "Epoch 9/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0110 - acc: 0.9973     \n",
      "Epoch 10/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0111 - acc: 0.9972     \n",
      "Epoch 11/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0103 - acc: 0.9975     \n",
      "Epoch 12/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0098 - acc: 0.9975     \n",
      "Epoch 13/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0094 - acc: 0.9978     \n",
      "Epoch 14/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0094 - acc: 0.9974     \n",
      "Epoch 15/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0092 - acc: 0.9974     \n",
      "Epoch 16/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0084 - acc: 0.9979     \n",
      "Epoch 17/20\n",
      "25652/25652 [==============================] - 10s - loss: 0.0082 - acc: 0.9977    \n",
      "Epoch 18/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0070 - acc: 0.9980     \n",
      "Epoch 19/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0067 - acc: 0.9982     \n",
      "Epoch 20/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0072 - acc: 0.9979     \n",
      "25330/25652 [============================>.] - ETA: 0sEpoch 1/20\n",
      "25652/25652 [==============================] - 11s - loss: 0.0954 - acc: 0.9683    \n",
      "Epoch 2/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0220 - acc: 0.9945     \n",
      "Epoch 3/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0175 - acc: 0.9951     \n",
      "Epoch 4/20\n",
      "25652/25652 [==============================] - 10s - loss: 0.0149 - acc: 0.9956    \n",
      "Epoch 5/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0125 - acc: 0.9966     \n",
      "Epoch 6/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0116 - acc: 0.9968     \n",
      "Epoch 7/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0117 - acc: 0.9968     \n",
      "Epoch 8/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0101 - acc: 0.9975     \n",
      "Epoch 9/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0096 - acc: 0.9975     \n",
      "Epoch 10/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0094 - acc: 0.9975     \n",
      "Epoch 11/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0089 - acc: 0.9975     \n",
      "Epoch 12/20\n",
      "25652/25652 [==============================] - 10s - loss: 0.0081 - acc: 0.9977    \n",
      "Epoch 13/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0083 - acc: 0.9976     \n",
      "Epoch 14/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0081 - acc: 0.9979     \n",
      "Epoch 15/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0079 - acc: 0.9979     \n",
      "Epoch 16/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0074 - acc: 0.9980     \n",
      "Epoch 17/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0075 - acc: 0.9980     \n",
      "Epoch 18/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0073 - acc: 0.9980     \n",
      "Epoch 19/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0069 - acc: 0.9979     \n",
      "Epoch 20/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0069 - acc: 0.9980     \n",
      "25170/25652 [============================>.] - ETA: 0sEpoch 1/20\n",
      "25652/25652 [==============================] - 11s - loss: 0.0776 - acc: 0.9811    \n",
      "Epoch 2/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0234 - acc: 0.9935     \n",
      "Epoch 3/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0189 - acc: 0.9946     \n",
      "Epoch 4/20\n",
      "25652/25652 [==============================] - 10s - loss: 0.0159 - acc: 0.9960    \n",
      "Epoch 5/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0144 - acc: 0.9963     \n",
      "Epoch 6/20\n",
      "25652/25652 [==============================] - 10s - loss: 0.0131 - acc: 0.9966    \n",
      "Epoch 7/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0117 - acc: 0.9970     \n",
      "Epoch 8/20\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "25652/25652 [==============================] - 10s - loss: 0.0113 - acc: 0.9969    \n",
      "Epoch 9/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0108 - acc: 0.9970     \n",
      "Epoch 10/20\n",
      "25652/25652 [==============================] - 10s - loss: 0.0094 - acc: 0.9973    \n",
      "Epoch 11/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0098 - acc: 0.9973     \n",
      "Epoch 12/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0087 - acc: 0.9975     \n",
      "Epoch 13/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0084 - acc: 0.9978     \n",
      "Epoch 14/20\n",
      "25652/25652 [==============================] - 10s - loss: 0.0080 - acc: 0.9979    \n",
      "Epoch 15/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0081 - acc: 0.9975     \n",
      "Epoch 16/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0076 - acc: 0.9979     \n",
      "Epoch 17/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0079 - acc: 0.9977     \n",
      "Epoch 18/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0070 - acc: 0.9979     \n",
      "Epoch 19/20\n",
      "25652/25652 [==============================] - 10s - loss: 0.0072 - acc: 0.9981    \n",
      "Epoch 20/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0065 - acc: 0.9982     \n",
      "25630/25652 [============================>.] - ETA: 0sEpoch 1/20\n",
      "25652/25652 [==============================] - 11s - loss: 0.0761 - acc: 0.9801    \n",
      "Epoch 2/20\n",
      "25652/25652 [==============================] - 10s - loss: 0.0222 - acc: 0.9936    \n",
      "Epoch 3/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0182 - acc: 0.9946     \n",
      "Epoch 4/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0163 - acc: 0.9956     \n",
      "Epoch 5/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0143 - acc: 0.9960     \n",
      "Epoch 6/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0127 - acc: 0.9961     \n",
      "Epoch 7/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0124 - acc: 0.9963     \n",
      "Epoch 8/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0114 - acc: 0.9970     \n",
      "Epoch 9/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0103 - acc: 0.9971     \n",
      "Epoch 10/20\n",
      "25652/25652 [==============================] - 10s - loss: 0.0105 - acc: 0.9968    \n",
      "Epoch 11/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0099 - acc: 0.9970     \n",
      "Epoch 12/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0092 - acc: 0.9973     \n",
      "Epoch 13/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0088 - acc: 0.9972     \n",
      "Epoch 14/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0084 - acc: 0.9973     \n",
      "Epoch 15/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0079 - acc: 0.9976     \n",
      "Epoch 16/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0074 - acc: 0.9977     \n",
      "Epoch 17/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0085 - acc: 0.9975     \n",
      "Epoch 18/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0071 - acc: 0.9979     \n",
      "Epoch 19/20\n",
      "25652/25652 [==============================] - 10s - loss: 0.0071 - acc: 0.9978    \n",
      "Epoch 20/20\n",
      "25652/25652 [==============================] - 10s - loss: 0.0067 - acc: 0.9980    \n",
      "25240/25652 [============================>.] - ETA: 0sEpoch 1/20\n",
      "25652/25652 [==============================] - 11s - loss: 0.1020 - acc: 0.9662    \n",
      "Epoch 2/20\n",
      "25652/25652 [==============================] - 10s - loss: 0.0223 - acc: 0.9945    \n",
      "Epoch 3/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0179 - acc: 0.9954     \n",
      "Epoch 4/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0152 - acc: 0.9960     \n",
      "Epoch 5/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0138 - acc: 0.9964     \n",
      "Epoch 6/20\n",
      "25652/25652 [==============================] - 10s - loss: 0.0129 - acc: 0.9965    \n",
      "Epoch 7/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0122 - acc: 0.9970     \n",
      "Epoch 8/20\n",
      "25652/25652 [==============================] - 10s - loss: 0.0111 - acc: 0.9972    \n",
      "Epoch 9/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0110 - acc: 0.9971     \n",
      "Epoch 10/20\n",
      "25652/25652 [==============================] - 10s - loss: 0.0104 - acc: 0.9974    \n",
      "Epoch 11/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0100 - acc: 0.9974     \n",
      "Epoch 12/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0096 - acc: 0.9979     \n",
      "Epoch 13/20\n",
      "25652/25652 [==============================] - 10s - loss: 0.0094 - acc: 0.9976    \n",
      "Epoch 14/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0080 - acc: 0.9979     \n",
      "Epoch 15/20\n",
      "25652/25652 [==============================] - 10s - loss: 0.0082 - acc: 0.9977    \n",
      "Epoch 16/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0080 - acc: 0.9979     \n",
      "Epoch 17/20\n",
      "25652/25652 [==============================] - 10s - loss: 0.0076 - acc: 0.9980    \n",
      "Epoch 18/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0074 - acc: 0.9981     \n",
      "Epoch 19/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0069 - acc: 0.9978     \n",
      "Epoch 20/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0071 - acc: 0.9982     \n",
      "25110/25652 [============================>.] - ETA: 0sEpoch 1/20\n",
      "25652/25652 [==============================] - 11s - loss: 0.0742 - acc: 0.9800    \n",
      "Epoch 2/20\n",
      "25652/25652 [==============================] - 10s - loss: 0.0218 - acc: 0.9941    \n",
      "Epoch 3/20\n",
      "25652/25652 [==============================] - 10s - loss: 0.0179 - acc: 0.9952    \n",
      "Epoch 4/20\n",
      "25652/25652 [==============================] - 10s - loss: 0.0154 - acc: 0.9957    \n",
      "Epoch 5/20\n",
      "25652/25652 [==============================] - 10s - loss: 0.0141 - acc: 0.9965    \n",
      "Epoch 6/20\n",
      "25652/25652 [==============================] - 10s - loss: 0.0126 - acc: 0.9966    \n",
      "Epoch 7/20\n",
      "25652/25652 [==============================] - 10s - loss: 0.0120 - acc: 0.9967    \n",
      "Epoch 8/20\n",
      "25652/25652 [==============================] - 10s - loss: 0.0110 - acc: 0.9973    \n",
      "Epoch 9/20\n",
      "25652/25652 [==============================] - 10s - loss: 0.0105 - acc: 0.9969    \n",
      "Epoch 10/20\n",
      "25652/25652 [==============================] - 10s - loss: 0.0098 - acc: 0.9973    \n",
      "Epoch 11/20\n",
      "25652/25652 [==============================] - 10s - loss: 0.0098 - acc: 0.9972    \n",
      "Epoch 12/20\n",
      "25652/25652 [==============================] - 10s - loss: 0.0087 - acc: 0.9975    \n",
      "Epoch 13/20\n",
      "25652/25652 [==============================] - 10s - loss: 0.0082 - acc: 0.9979    \n",
      "Epoch 14/20\n",
      "25652/25652 [==============================] - 10s - loss: 0.0084 - acc: 0.9973    \n",
      "Epoch 15/20\n",
      "25652/25652 [==============================] - 10s - loss: 0.0075 - acc: 0.9977    \n",
      "Epoch 16/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0076 - acc: 0.9978     \n",
      "Epoch 17/20\n",
      "25652/25652 [==============================] - 10s - loss: 0.0065 - acc: 0.9981    \n",
      "Epoch 18/20\n",
      "25652/25652 [==============================] - 10s - loss: 0.0068 - acc: 0.9980    \n",
      "Epoch 19/20\n",
      "25652/25652 [==============================] - 10s - loss: 0.0063 - acc: 0.9981    \n",
      "Epoch 20/20\n",
      "25652/25652 [==============================] - 10s - loss: 0.0063 - acc: 0.9977    \n",
      "25600/25652 [============================>.] - ETA: 0sEpoch 1/20\n",
      "25651/25651 [==============================] - 9s - loss: 0.1076 - acc: 0.9646     \n",
      "Epoch 2/20\n",
      "25651/25651 [==============================] - 8s - loss: 0.0276 - acc: 0.9934     \n",
      "Epoch 3/20\n",
      "25651/25651 [==============================] - 8s - loss: 0.0247 - acc: 0.9943     \n",
      "Epoch 4/20\n",
      "25651/25651 [==============================] - 8s - loss: 0.0235 - acc: 0.9952     \n",
      "Epoch 5/20\n",
      "25651/25651 [==============================] - 8s - loss: 0.0214 - acc: 0.9953     \n",
      "Epoch 6/20\n",
      "25651/25651 [==============================] - 8s - loss: 0.0197 - acc: 0.9956     \n",
      "Epoch 7/20\n",
      "25651/25651 [==============================] - 8s - loss: 0.0186 - acc: 0.9963     \n",
      "Epoch 8/20\n",
      "25651/25651 [==============================] - 8s - loss: 0.0177 - acc: 0.9965     \n",
      "Epoch 9/20\n",
      "25651/25651 [==============================] - 8s - loss: 0.0174 - acc: 0.9968     \n",
      "Epoch 10/20\n",
      "25651/25651 [==============================] - 8s - loss: 0.0167 - acc: 0.9966     \n",
      "Epoch 11/20\n",
      "25651/25651 [==============================] - 8s - loss: 0.0137 - acc: 0.9967     \n",
      "Epoch 12/20\n",
      "25651/25651 [==============================] - 8s - loss: 0.0134 - acc: 0.9973     \n",
      "Epoch 13/20\n",
      "25651/25651 [==============================] - 8s - loss: 0.0141 - acc: 0.9970     \n",
      "Epoch 14/20\n",
      "25651/25651 [==============================] - 8s - loss: 0.0141 - acc: 0.9971     \n",
      "Epoch 15/20\n",
      "25651/25651 [==============================] - 8s - loss: 0.0131 - acc: 0.9973     \n",
      "Epoch 16/20\n",
      "25651/25651 [==============================] - 8s - loss: 0.0131 - acc: 0.9972     \n",
      "Epoch 17/20\n",
      "25651/25651 [==============================] - 8s - loss: 0.0130 - acc: 0.9972     \n",
      "Epoch 18/20\n",
      "25651/25651 [==============================] - 8s - loss: 0.0128 - acc: 0.9974     \n",
      "Epoch 19/20\n",
      "25651/25651 [==============================] - 8s - loss: 0.0126 - acc: 0.9975     \n",
      "Epoch 20/20\n",
      "25651/25651 [==============================] - 8s - loss: 0.0120 - acc: 0.9975     \n",
      "25380/25651 [============================>.] - ETA: 0sEpoch 1/20\n",
      "25651/25651 [==============================] - 9s - loss: 0.0848 - acc: 0.9777     \n",
      "Epoch 2/20\n",
      "25651/25651 [==============================] - 8s - loss: 0.0264 - acc: 0.9934     \n",
      "Epoch 3/20\n",
      "25651/25651 [==============================] - 8s - loss: 0.0239 - acc: 0.9947     \n",
      "Epoch 4/20\n",
      "25651/25651 [==============================] - 8s - loss: 0.0219 - acc: 0.9950     \n",
      "Epoch 5/20\n",
      "25651/25651 [==============================] - 8s - loss: 0.0206 - acc: 0.9951     \n",
      "Epoch 6/20\n",
      "25651/25651 [==============================] - 8s - loss: 0.0185 - acc: 0.9958     \n",
      "Epoch 7/20\n",
      "25651/25651 [==============================] - 8s - loss: 0.0173 - acc: 0.9959     \n",
      "Epoch 8/20\n",
      "25651/25651 [==============================] - 8s - loss: 0.0166 - acc: 0.9961     \n",
      "Epoch 9/20\n",
      "25651/25651 [==============================] - 8s - loss: 0.0155 - acc: 0.9964     \n",
      "Epoch 10/20\n",
      "25651/25651 [==============================] - 8s - loss: 0.0152 - acc: 0.9967     \n",
      "Epoch 11/20\n",
      "25651/25651 [==============================] - 8s - loss: 0.0147 - acc: 0.9970     \n",
      "Epoch 12/20\n",
      "25651/25651 [==============================] - 8s - loss: 0.0143 - acc: 0.9973     \n",
      "Epoch 13/20\n",
      "25651/25651 [==============================] - 8s - loss: 0.0144 - acc: 0.9973     \n",
      "Epoch 14/20\n",
      "25651/25651 [==============================] - 8s - loss: 0.0141 - acc: 0.9973     \n",
      "Epoch 15/20\n",
      "25651/25651 [==============================] - 8s - loss: 0.0127 - acc: 0.9973     \n",
      "Epoch 16/20\n",
      "25651/25651 [==============================] - 8s - loss: 0.0123 - acc: 0.9981     \n",
      "Epoch 17/20\n",
      "25651/25651 [==============================] - 8s - loss: 0.0125 - acc: 0.9976     \n",
      "Epoch 18/20\n",
      "25651/25651 [==============================] - 8s - loss: 0.0125 - acc: 0.9975     \n",
      "Epoch 19/20\n",
      "25651/25651 [==============================] - 8s - loss: 0.0121 - acc: 0.9974     \n",
      "Epoch 20/20\n",
      "25651/25651 [==============================] - 8s - loss: 0.0121 - acc: 0.9979     \n",
      "25410/25651 [============================>.] - ETA: 0sEpoch 1/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.1124 - acc: 0.9628     \n",
      "Epoch 2/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0282 - acc: 0.9939     \n",
      "Epoch 3/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0253 - acc: 0.9946     \n",
      "Epoch 4/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0235 - acc: 0.9952     \n",
      "Epoch 5/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0232 - acc: 0.9956     \n",
      "Epoch 6/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0212 - acc: 0.9963     \n",
      "Epoch 7/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0187 - acc: 0.9963     \n",
      "Epoch 8/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0169 - acc: 0.9966     \n",
      "Epoch 9/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0159 - acc: 0.9968     \n",
      "Epoch 10/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0161 - acc: 0.9968     \n",
      "Epoch 11/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0156 - acc: 0.9970     \n",
      "Epoch 12/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0152 - acc: 0.9970     \n",
      "Epoch 13/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0139 - acc: 0.9973     \n",
      "Epoch 14/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0135 - acc: 0.9973     \n",
      "Epoch 15/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0130 - acc: 0.9972     \n",
      "Epoch 16/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0113 - acc: 0.9976     \n",
      "Epoch 17/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0115 - acc: 0.9977     \n",
      "Epoch 18/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0127 - acc: 0.9975     \n",
      "Epoch 19/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0116 - acc: 0.9977     \n",
      "Epoch 20/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0124 - acc: 0.9978     \n",
      "25610/25652 [============================>.] - ETA: 0sEpoch 1/20\n",
      "25652/25652 [==============================] - 10s - loss: 0.1101 - acc: 0.9642    \n",
      "Epoch 2/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0297 - acc: 0.9931     \n",
      "Epoch 3/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0254 - acc: 0.9945     \n",
      "Epoch 4/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0236 - acc: 0.9951     \n",
      "Epoch 5/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0213 - acc: 0.9952     \n",
      "Epoch 6/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0200 - acc: 0.9959     \n",
      "Epoch 7/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0187 - acc: 0.9959     \n",
      "Epoch 8/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0189 - acc: 0.9962     \n",
      "Epoch 9/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0176 - acc: 0.9965     \n",
      "Epoch 10/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0173 - acc: 0.9965     \n",
      "Epoch 11/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0172 - acc: 0.9969     \n",
      "Epoch 12/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0175 - acc: 0.9968     \n",
      "Epoch 13/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0161 - acc: 0.9970     \n",
      "Epoch 14/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0168 - acc: 0.9969     \n",
      "Epoch 15/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0153 - acc: 0.9969     \n",
      "Epoch 16/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0153 - acc: 0.9968     \n",
      "Epoch 17/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0146 - acc: 0.9970     \n",
      "Epoch 18/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0141 - acc: 0.9972     \n",
      "Epoch 19/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0137 - acc: 0.9973     \n",
      "Epoch 20/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0136 - acc: 0.9972     \n",
      "25490/25652 [============================>.] - ETA: 0sEpoch 1/20\n",
      "25652/25652 [==============================] - 10s - loss: 0.1068 - acc: 0.9655    \n",
      "Epoch 2/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0295 - acc: 0.9933     \n",
      "Epoch 3/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0253 - acc: 0.9944     \n",
      "Epoch 4/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0232 - acc: 0.9953     \n",
      "Epoch 5/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0207 - acc: 0.9956     \n",
      "Epoch 6/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0184 - acc: 0.9960     \n",
      "Epoch 7/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0174 - acc: 0.9963     \n",
      "Epoch 8/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0172 - acc: 0.9964     \n",
      "Epoch 9/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0170 - acc: 0.9968     \n",
      "Epoch 10/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0159 - acc: 0.9970     \n",
      "Epoch 11/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0150 - acc: 0.9972     \n",
      "Epoch 12/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0154 - acc: 0.9970     \n",
      "Epoch 13/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0156 - acc: 0.9970     \n",
      "Epoch 14/20\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "25652/25652 [==============================] - 8s - loss: 0.0156 - acc: 0.9970     \n",
      "Epoch 15/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0155 - acc: 0.9973     \n",
      "Epoch 16/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0151 - acc: 0.9974     \n",
      "Epoch 17/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0146 - acc: 0.9974     \n",
      "Epoch 18/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0144 - acc: 0.9975     \n",
      "Epoch 19/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0129 - acc: 0.9976     \n",
      "Epoch 20/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0134 - acc: 0.9975     \n",
      "25470/25652 [============================>.] - ETA: 0sEpoch 1/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0849 - acc: 0.9759     \n",
      "Epoch 2/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0264 - acc: 0.9935     \n",
      "Epoch 3/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0233 - acc: 0.9949     \n",
      "Epoch 4/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0210 - acc: 0.9950     \n",
      "Epoch 5/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0186 - acc: 0.9956     \n",
      "Epoch 6/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0177 - acc: 0.9959     \n",
      "Epoch 7/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0166 - acc: 0.9960     \n",
      "Epoch 8/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0161 - acc: 0.9964     \n",
      "Epoch 9/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0161 - acc: 0.9965     \n",
      "Epoch 10/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0151 - acc: 0.9970     \n",
      "Epoch 11/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0150 - acc: 0.9968     \n",
      "Epoch 12/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0139 - acc: 0.9970     \n",
      "Epoch 13/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0135 - acc: 0.9973     \n",
      "Epoch 14/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0132 - acc: 0.9974     \n",
      "Epoch 15/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0145 - acc: 0.9973     \n",
      "Epoch 16/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0129 - acc: 0.9976     \n",
      "Epoch 17/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0136 - acc: 0.9975     \n",
      "Epoch 18/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0129 - acc: 0.9974     \n",
      "Epoch 19/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0112 - acc: 0.9977     \n",
      "Epoch 20/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0130 - acc: 0.9975     \n",
      "25390/25652 [============================>.] - ETA: 0sEpoch 1/20\n",
      "25652/25652 [==============================] - 9s - loss: 0.0826 - acc: 0.9786     \n",
      "Epoch 2/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0264 - acc: 0.9937     \n",
      "Epoch 3/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0243 - acc: 0.9948     \n",
      "Epoch 4/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0221 - acc: 0.9951     \n",
      "Epoch 5/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0213 - acc: 0.9953     \n",
      "Epoch 6/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0194 - acc: 0.9961     \n",
      "Epoch 7/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0185 - acc: 0.9959     \n",
      "Epoch 8/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0178 - acc: 0.9960     \n",
      "Epoch 9/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0171 - acc: 0.9964     \n",
      "Epoch 10/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0170 - acc: 0.9965     \n",
      "Epoch 11/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0161 - acc: 0.9968     \n",
      "Epoch 12/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0147 - acc: 0.9967     \n",
      "Epoch 13/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0142 - acc: 0.9971     \n",
      "Epoch 14/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0140 - acc: 0.9974     \n",
      "Epoch 15/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0134 - acc: 0.9973     \n",
      "Epoch 16/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0148 - acc: 0.9973     \n",
      "Epoch 17/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0136 - acc: 0.9973     \n",
      "Epoch 18/20\n",
      "25652/25652 [==============================] - ETA: 0s - loss: 0.0132 - acc: 0.997 - 8s - loss: 0.0134 - acc: 0.9973     \n",
      "Epoch 19/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0128 - acc: 0.9975     \n",
      "Epoch 20/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0122 - acc: 0.9976     \n",
      "25320/25652 [============================>.] - ETA: 0sEpoch 1/20\n",
      "25652/25652 [==============================] - 10s - loss: 0.0924 - acc: 0.9729    \n",
      "Epoch 2/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0265 - acc: 0.9936     \n",
      "Epoch 3/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0233 - acc: 0.9949     \n",
      "Epoch 4/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0218 - acc: 0.9952     \n",
      "Epoch 5/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0207 - acc: 0.9955     \n",
      "Epoch 6/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0187 - acc: 0.9959     \n",
      "Epoch 7/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0184 - acc: 0.9959     \n",
      "Epoch 8/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0180 - acc: 0.9959     \n",
      "Epoch 9/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0174 - acc: 0.9964     \n",
      "Epoch 10/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0169 - acc: 0.9963     \n",
      "Epoch 11/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0163 - acc: 0.9969     \n",
      "Epoch 12/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0164 - acc: 0.9971     \n",
      "Epoch 13/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0164 - acc: 0.9968     \n",
      "Epoch 14/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0159 - acc: 0.9969     \n",
      "Epoch 15/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0159 - acc: 0.9970     \n",
      "Epoch 16/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0151 - acc: 0.9970     \n",
      "Epoch 17/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0157 - acc: 0.9968     \n",
      "Epoch 18/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0143 - acc: 0.9974     \n",
      "Epoch 19/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0146 - acc: 0.9973     \n",
      "Epoch 20/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0146 - acc: 0.9970     \n",
      "25210/25652 [============================>.] - ETA: 0sEpoch 1/20\n",
      "25652/25652 [==============================] - 10s - loss: 0.1248 - acc: 0.9597    \n",
      "Epoch 2/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0328 - acc: 0.9933     \n",
      "Epoch 3/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0259 - acc: 0.9945     \n",
      "Epoch 4/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0232 - acc: 0.9950     \n",
      "Epoch 5/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0216 - acc: 0.9956     \n",
      "Epoch 6/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0200 - acc: 0.9960     \n",
      "Epoch 7/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0182 - acc: 0.9966     \n",
      "Epoch 8/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0178 - acc: 0.9968     \n",
      "Epoch 9/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0168 - acc: 0.9968     \n",
      "Epoch 10/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0174 - acc: 0.9968     \n",
      "Epoch 11/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0162 - acc: 0.9971     \n",
      "Epoch 12/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0148 - acc: 0.9973     \n",
      "Epoch 13/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0158 - acc: 0.9971     \n",
      "Epoch 14/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0151 - acc: 0.9972     \n",
      "Epoch 15/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0145 - acc: 0.9974     \n",
      "Epoch 16/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0153 - acc: 0.9971     \n",
      "Epoch 17/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0147 - acc: 0.9975     \n",
      "Epoch 18/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0144 - acc: 0.9975     \n",
      "Epoch 19/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0146 - acc: 0.9973     \n",
      "Epoch 20/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0142 - acc: 0.9977     \n",
      "25160/25652 [============================>.] - ETA: 0sEpoch 1/20\n",
      "25652/25652 [==============================] - 10s - loss: 0.0918 - acc: 0.9736    \n",
      "Epoch 2/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0260 - acc: 0.9937     \n",
      "Epoch 3/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0225 - acc: 0.9947     \n",
      "Epoch 4/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0209 - acc: 0.9952     \n",
      "Epoch 5/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0181 - acc: 0.9958     \n",
      "Epoch 6/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0167 - acc: 0.9964     \n",
      "Epoch 7/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0169 - acc: 0.9964     \n",
      "Epoch 8/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0161 - acc: 0.9969     \n",
      "Epoch 9/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0152 - acc: 0.9967     \n",
      "Epoch 10/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0153 - acc: 0.9968     \n",
      "Epoch 11/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0143 - acc: 0.9970     \n",
      "Epoch 12/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0138 - acc: 0.9971     \n",
      "Epoch 13/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0137 - acc: 0.9972     \n",
      "Epoch 14/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0135 - acc: 0.9973     \n",
      "Epoch 15/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0137 - acc: 0.9973     \n",
      "Epoch 16/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0133 - acc: 0.9977     \n",
      "Epoch 17/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0129 - acc: 0.9977     \n",
      "Epoch 18/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0125 - acc: 0.9978     \n",
      "Epoch 19/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0126 - acc: 0.9979     \n",
      "Epoch 20/20\n",
      "25652/25652 [==============================] - 8s - loss: 0.0123 - acc: 0.9979     \n",
      "25240/25652 [============================>.] - ETA: 0sEpoch 1/10\n",
      "25651/25651 [==============================] - 4s - loss: 0.1921 - acc: 0.9321     \n",
      "Epoch 2/10\n",
      "25651/25651 [==============================] - 3s - loss: 0.0335 - acc: 0.9916     \n",
      "Epoch 3/10\n",
      "25651/25651 [==============================] - 3s - loss: 0.0243 - acc: 0.9937     \n",
      "Epoch 4/10\n",
      "25651/25651 [==============================] - 3s - loss: 0.0211 - acc: 0.9942     \n",
      "Epoch 5/10\n",
      "25651/25651 [==============================] - 3s - loss: 0.0187 - acc: 0.9951     \n",
      "Epoch 6/10\n",
      "25651/25651 [==============================] - 3s - loss: 0.0172 - acc: 0.9955     \n",
      "Epoch 7/10\n",
      "25651/25651 [==============================] - 3s - loss: 0.0154 - acc: 0.9963     \n",
      "Epoch 8/10\n",
      "25651/25651 [==============================] - 3s - loss: 0.0145 - acc: 0.9964     \n",
      "Epoch 9/10\n",
      "25651/25651 [==============================] - 3s - loss: 0.0137 - acc: 0.9969     \n",
      "Epoch 10/10\n",
      "25651/25651 [==============================] - 3s - loss: 0.0128 - acc: 0.9970     \n",
      "25530/25651 [============================>.] - ETA: 0sEpoch 1/10\n",
      "25651/25651 [==============================] - 4s - loss: 0.1851 - acc: 0.9404     \n",
      "Epoch 2/10\n",
      "25651/25651 [==============================] - 3s - loss: 0.0325 - acc: 0.9916     \n",
      "Epoch 3/10\n",
      "25651/25651 [==============================] - 3s - loss: 0.0233 - acc: 0.9940     \n",
      "Epoch 4/10\n",
      "25651/25651 [==============================] - 3s - loss: 0.0201 - acc: 0.9947     \n",
      "Epoch 5/10\n",
      "25651/25651 [==============================] - 3s - loss: 0.0177 - acc: 0.9952     \n",
      "Epoch 6/10\n",
      "25651/25651 [==============================] - 3s - loss: 0.0162 - acc: 0.9958     \n",
      "Epoch 7/10\n",
      "25651/25651 [==============================] - 3s - loss: 0.0151 - acc: 0.9963     \n",
      "Epoch 8/10\n",
      "25651/25651 [==============================] - 3s - loss: 0.0142 - acc: 0.9964     \n",
      "Epoch 9/10\n",
      "25651/25651 [==============================] - 3s - loss: 0.0137 - acc: 0.9963     \n",
      "Epoch 10/10\n",
      "25651/25651 [==============================] - 3s - loss: 0.0129 - acc: 0.9970     \n",
      "25080/25651 [============================>.] - ETA: 0sEpoch 1/10\n",
      "25652/25652 [==============================] - 4s - loss: 0.1786 - acc: 0.9432     \n",
      "Epoch 2/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0297 - acc: 0.9927     \n",
      "Epoch 3/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0218 - acc: 0.9944     \n",
      "Epoch 4/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0187 - acc: 0.9951     \n",
      "Epoch 5/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0168 - acc: 0.9956     \n",
      "Epoch 6/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0155 - acc: 0.9964     \n",
      "Epoch 7/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0139 - acc: 0.9966     \n",
      "Epoch 8/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0135 - acc: 0.9967     \n",
      "Epoch 9/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0126 - acc: 0.9970     \n",
      "Epoch 10/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0125 - acc: 0.9967     \n",
      "25410/25652 [============================>.] - ETA: 0sEpoch 1/10\n",
      "25652/25652 [==============================] - 5s - loss: 0.1530 - acc: 0.9596     \n",
      "Epoch 2/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0299 - acc: 0.9919     \n",
      "Epoch 3/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0225 - acc: 0.9935     \n",
      "Epoch 4/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0195 - acc: 0.9949     \n",
      "Epoch 5/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0174 - acc: 0.9951     \n",
      "Epoch 6/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0165 - acc: 0.9954     \n",
      "Epoch 7/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0155 - acc: 0.9958     \n",
      "Epoch 8/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0142 - acc: 0.9961     \n",
      "Epoch 9/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0130 - acc: 0.9966     \n",
      "Epoch 10/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0125 - acc: 0.9966     \n",
      "24540/25652 [===========================>..] - ETA: 0sEpoch 1/10\n",
      "25652/25652 [==============================] - 5s - loss: 0.1510 - acc: 0.9645     \n",
      "Epoch 2/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0311 - acc: 0.9915     \n",
      "Epoch 3/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0253 - acc: 0.9929     \n",
      "Epoch 4/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0217 - acc: 0.9943     \n",
      "Epoch 5/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0190 - acc: 0.9945     \n",
      "Epoch 6/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0171 - acc: 0.9957     \n",
      "Epoch 7/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0155 - acc: 0.9957     \n",
      "Epoch 8/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0143 - acc: 0.9961     \n",
      "Epoch 9/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0132 - acc: 0.9964     \n",
      "Epoch 10/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0124 - acc: 0.9969     \n",
      "23670/25652 [==========================>...] - ETA: 0sEpoch 1/10\n",
      "25652/25652 [==============================] - 5s - loss: 0.1985 - acc: 0.9151     \n",
      "Epoch 2/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0572 - acc: 0.9890     \n",
      "Epoch 3/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0274 - acc: 0.9935     \n",
      "Epoch 4/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0213 - acc: 0.9950     \n",
      "Epoch 5/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0185 - acc: 0.9955     \n",
      "Epoch 6/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0164 - acc: 0.9959     \n",
      "Epoch 7/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0149 - acc: 0.9964     \n",
      "Epoch 8/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0138 - acc: 0.9966     \n",
      "Epoch 9/10\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "25652/25652 [==============================] - 3s - loss: 0.0128 - acc: 0.9969     \n",
      "Epoch 10/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0119 - acc: 0.9968     \n",
      "24630/25652 [===========================>..] - ETA: 0sEpoch 1/10\n",
      "25652/25652 [==============================] - 5s - loss: 0.1495 - acc: 0.9668     \n",
      "Epoch 2/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0288 - acc: 0.9922     \n",
      "Epoch 3/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0226 - acc: 0.9941     \n",
      "Epoch 4/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0202 - acc: 0.9945     \n",
      "Epoch 5/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0177 - acc: 0.9951     \n",
      "Epoch 6/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0166 - acc: 0.9958     \n",
      "Epoch 7/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0152 - acc: 0.9960     \n",
      "Epoch 8/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0140 - acc: 0.9961     \n",
      "Epoch 9/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0131 - acc: 0.9967     \n",
      "Epoch 10/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0126 - acc: 0.9965     \n",
      "25530/25652 [============================>.] - ETA: 0sEpoch 1/10\n",
      "25652/25652 [==============================] - 5s - loss: 0.1566 - acc: 0.9586     \n",
      "Epoch 2/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0301 - acc: 0.9920     \n",
      "Epoch 3/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0236 - acc: 0.9936     \n",
      "Epoch 4/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0206 - acc: 0.9942     \n",
      "Epoch 5/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0184 - acc: 0.9953     \n",
      "Epoch 6/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0168 - acc: 0.9958     \n",
      "Epoch 7/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0161 - acc: 0.9958     \n",
      "Epoch 8/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0148 - acc: 0.9963     \n",
      "Epoch 9/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0143 - acc: 0.9963     \n",
      "Epoch 10/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0136 - acc: 0.9966     \n",
      "24630/25652 [===========================>..] - ETA: 0sEpoch 1/10\n",
      "25652/25652 [==============================] - 5s - loss: 0.1782 - acc: 0.9458     \n",
      "Epoch 2/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0313 - acc: 0.9921     \n",
      "Epoch 3/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0228 - acc: 0.9940     \n",
      "Epoch 4/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0191 - acc: 0.9950     \n",
      "Epoch 5/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0168 - acc: 0.9956     \n",
      "Epoch 6/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0150 - acc: 0.9961     \n",
      "Epoch 7/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0133 - acc: 0.9966     \n",
      "Epoch 8/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0125 - acc: 0.9966     \n",
      "Epoch 9/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0114 - acc: 0.9970     \n",
      "Epoch 10/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0113 - acc: 0.9972     \n",
      "24810/25652 [============================>.] - ETA: 0sEpoch 1/10\n",
      "25652/25652 [==============================] - 5s - loss: 0.1449 - acc: 0.9636     \n",
      "Epoch 2/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0266 - acc: 0.9923     \n",
      "Epoch 3/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0210 - acc: 0.9942     \n",
      "Epoch 4/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0182 - acc: 0.9951     \n",
      "Epoch 5/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0161 - acc: 0.9958     \n",
      "Epoch 6/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0147 - acc: 0.9962     \n",
      "Epoch 7/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0133 - acc: 0.9966     \n",
      "Epoch 8/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0122 - acc: 0.9968     \n",
      "Epoch 9/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0120 - acc: 0.9969     \n",
      "Epoch 10/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0109 - acc: 0.9973     \n",
      "25260/25652 [============================>.] - ETA: 0sEpoch 1/10\n",
      "25651/25651 [==============================] - 4s - loss: 0.2023 - acc: 0.9300     \n",
      "Epoch 2/10\n",
      "25651/25651 [==============================] - 2s - loss: 0.0424 - acc: 0.9914     \n",
      "Epoch 3/10\n",
      "25651/25651 [==============================] - 3s - loss: 0.0262 - acc: 0.9935     \n",
      "Epoch 4/10\n",
      "25651/25651 [==============================] - 3s - loss: 0.0223 - acc: 0.9944     \n",
      "Epoch 5/10\n",
      "25651/25651 [==============================] - 3s - loss: 0.0198 - acc: 0.9950     \n",
      "Epoch 6/10\n",
      "25651/25651 [==============================] - 3s - loss: 0.0186 - acc: 0.9954     \n",
      "Epoch 7/10\n",
      "25651/25651 [==============================] - 3s - loss: 0.0170 - acc: 0.9959     \n",
      "Epoch 8/10\n",
      "25651/25651 [==============================] - 3s - loss: 0.0164 - acc: 0.9959     \n",
      "Epoch 9/10\n",
      "25651/25651 [==============================] - 3s - loss: 0.0145 - acc: 0.9962     \n",
      "Epoch 10/10\n",
      "25651/25651 [==============================] - 3s - loss: 0.0132 - acc: 0.9966     \n",
      "24900/25651 [============================>.] - ETA: 0sEpoch 1/10\n",
      "25651/25651 [==============================] - 4s - loss: 0.1657 - acc: 0.9584     \n",
      "Epoch 2/10\n",
      "25651/25651 [==============================] - 3s - loss: 0.0313 - acc: 0.9915     \n",
      "Epoch 3/10\n",
      "25651/25651 [==============================] - 3s - loss: 0.0249 - acc: 0.9933     \n",
      "Epoch 4/10\n",
      "25651/25651 [==============================] - 3s - loss: 0.0217 - acc: 0.9945     \n",
      "Epoch 5/10\n",
      "25651/25651 [==============================] - 3s - loss: 0.0198 - acc: 0.9952     \n",
      "Epoch 6/10\n",
      "25651/25651 [==============================] - 3s - loss: 0.0188 - acc: 0.9955     \n",
      "Epoch 7/10\n",
      "25651/25651 [==============================] - 3s - loss: 0.0173 - acc: 0.9958     \n",
      "Epoch 8/10\n",
      "25651/25651 [==============================] - 3s - loss: 0.0165 - acc: 0.9960     \n",
      "Epoch 9/10\n",
      "25651/25651 [==============================] - 3s - loss: 0.0156 - acc: 0.9962     \n",
      "Epoch 10/10\n",
      "25651/25651 [==============================] - 3s - loss: 0.0149 - acc: 0.9961     \n",
      "24570/25651 [===========================>..] - ETA: 0sEpoch 1/10\n",
      "25652/25652 [==============================] - 4s - loss: 0.1978 - acc: 0.9383     \n",
      "Epoch 2/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0319 - acc: 0.9922     \n",
      "Epoch 3/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0238 - acc: 0.9942     \n",
      "Epoch 4/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0209 - acc: 0.9950     \n",
      "Epoch 5/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0185 - acc: 0.9954     \n",
      "Epoch 6/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0169 - acc: 0.9958     \n",
      "Epoch 7/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0159 - acc: 0.9961     \n",
      "Epoch 8/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0146 - acc: 0.9962     \n",
      "Epoch 9/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0141 - acc: 0.9967     \n",
      "Epoch 10/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0138 - acc: 0.9971     \n",
      "24300/25652 [===========================>..] - ETA: 0sEpoch 1/10\n",
      "25652/25652 [==============================] - 4s - loss: 0.1923 - acc: 0.9356     \n",
      "Epoch 2/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0425 - acc: 0.9903     \n",
      "Epoch 3/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0267 - acc: 0.9936     \n",
      "Epoch 4/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0230 - acc: 0.9947     \n",
      "Epoch 5/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0204 - acc: 0.9953     \n",
      "Epoch 6/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0182 - acc: 0.9956     \n",
      "Epoch 7/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0169 - acc: 0.9959     \n",
      "Epoch 8/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0157 - acc: 0.9963     \n",
      "Epoch 9/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0147 - acc: 0.9965     \n",
      "Epoch 10/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0141 - acc: 0.9969     \n",
      "24060/25652 [===========================>..] - ETA: 0sEpoch 1/10\n",
      "25652/25652 [==============================] - 4s - loss: 0.1607 - acc: 0.9603     \n",
      "Epoch 2/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0310 - acc: 0.9920     \n",
      "Epoch 3/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0248 - acc: 0.9936     \n",
      "Epoch 4/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0224 - acc: 0.9942     \n",
      "Epoch 5/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0202 - acc: 0.9949     \n",
      "Epoch 6/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0189 - acc: 0.9951     \n",
      "Epoch 7/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0174 - acc: 0.9959     \n",
      "Epoch 8/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0164 - acc: 0.9959     \n",
      "Epoch 9/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0154 - acc: 0.9961     \n",
      "Epoch 10/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0140 - acc: 0.9966     \n",
      "24060/25652 [===========================>..] - ETA: 0sEpoch 1/10\n",
      "25652/25652 [==============================] - 4s - loss: 0.1762 - acc: 0.9591     \n",
      "Epoch 2/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0325 - acc: 0.9913     \n",
      "Epoch 3/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0248 - acc: 0.9935     \n",
      "Epoch 4/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0217 - acc: 0.9946     \n",
      "Epoch 5/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0197 - acc: 0.9952     \n",
      "Epoch 6/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0179 - acc: 0.9954     \n",
      "Epoch 7/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0167 - acc: 0.9957     \n",
      "Epoch 8/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0153 - acc: 0.9963     \n",
      "Epoch 9/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0144 - acc: 0.9964     \n",
      "Epoch 10/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0135 - acc: 0.9968     \n",
      "23910/25652 [==========================>...] - ETA: 0sEpoch 1/10\n",
      "25652/25652 [==============================] - 4s - loss: 0.1720 - acc: 0.9564     \n",
      "Epoch 2/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0318 - acc: 0.9913     \n",
      "Epoch 3/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0248 - acc: 0.9931     \n",
      "Epoch 4/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0218 - acc: 0.9942     \n",
      "Epoch 5/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0201 - acc: 0.9950     \n",
      "Epoch 6/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0191 - acc: 0.9951     \n",
      "Epoch 7/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0176 - acc: 0.9953     \n",
      "Epoch 8/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0166 - acc: 0.9959     \n",
      "Epoch 9/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0160 - acc: 0.9956     \n",
      "Epoch 10/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0156 - acc: 0.9960     \n",
      "25500/25652 [============================>.] - ETA: 0sEpoch 1/10\n",
      "25652/25652 [==============================] - 5s - loss: 0.1868 - acc: 0.9363     \n",
      "Epoch 2/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0687 - acc: 0.9886     \n",
      "Epoch 3/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0457 - acc: 0.9932     \n",
      "Epoch 4/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0338 - acc: 0.9940     \n",
      "Epoch 5/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0279 - acc: 0.9948     \n",
      "Epoch 6/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0247 - acc: 0.9953     \n",
      "Epoch 7/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0222 - acc: 0.9954     \n",
      "Epoch 8/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0204 - acc: 0.9956     \n",
      "Epoch 9/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0196 - acc: 0.9960     \n",
      "Epoch 10/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0184 - acc: 0.9962     \n",
      "25260/25652 [============================>.] - ETA: 0sEpoch 1/10\n",
      "25652/25652 [==============================] - 5s - loss: 0.1745 - acc: 0.9511     \n",
      "Epoch 2/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0316 - acc: 0.9918     \n",
      "Epoch 3/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0237 - acc: 0.9937     \n",
      "Epoch 4/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0202 - acc: 0.9949     \n",
      "Epoch 5/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0186 - acc: 0.9952     \n",
      "Epoch 6/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0178 - acc: 0.9955     \n",
      "Epoch 7/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0170 - acc: 0.9955     \n",
      "Epoch 8/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0158 - acc: 0.9960     \n",
      "Epoch 9/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0151 - acc: 0.9961     \n",
      "Epoch 10/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0147 - acc: 0.9962     \n",
      "25020/25652 [============================>.] - ETA: 0sEpoch 1/10\n",
      "25652/25652 [==============================] - 5s - loss: 0.1597 - acc: 0.9556     \n",
      "Epoch 2/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0288 - acc: 0.9923     \n",
      "Epoch 3/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0228 - acc: 0.9940     \n",
      "Epoch 4/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0200 - acc: 0.9950     \n",
      "Epoch 5/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0184 - acc: 0.9956     \n",
      "Epoch 6/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0169 - acc: 0.9960     \n",
      "Epoch 7/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0158 - acc: 0.9961     \n",
      "Epoch 8/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0156 - acc: 0.9962     \n",
      "Epoch 9/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0142 - acc: 0.9965     \n",
      "Epoch 10/10\n",
      "25652/25652 [==============================] - 3s - loss: 0.0140 - acc: 0.9966     \n",
      "24750/25652 [===========================>..] - ETA: 0sEpoch 1/20\n",
      "25651/25651 [==============================] - 5s - loss: 0.1387 - acc: 0.9688     \n",
      "Epoch 2/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0280 - acc: 0.9923     \n",
      "Epoch 3/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0232 - acc: 0.9936     \n",
      "Epoch 4/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0209 - acc: 0.9944     \n",
      "Epoch 5/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0188 - acc: 0.9950     \n",
      "Epoch 6/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0173 - acc: 0.9954     \n",
      "Epoch 7/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0159 - acc: 0.9957     \n",
      "Epoch 8/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0147 - acc: 0.9964     \n",
      "Epoch 9/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0136 - acc: 0.9967     \n",
      "Epoch 10/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0132 - acc: 0.9966     \n",
      "Epoch 11/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0121 - acc: 0.9972     \n",
      "Epoch 12/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0112 - acc: 0.9973     \n",
      "Epoch 13/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0105 - acc: 0.9972     \n",
      "Epoch 14/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0106 - acc: 0.9970     \n",
      "Epoch 15/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0095 - acc: 0.9975     \n",
      "Epoch 16/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0097 - acc: 0.9972     \n",
      "Epoch 17/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0088 - acc: 0.9977     \n",
      "Epoch 18/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0081 - acc: 0.9979     \n",
      "Epoch 19/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0079 - acc: 0.9980     \n",
      "Epoch 20/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0081 - acc: 0.9976     \n",
      "24870/25651 [============================>.] - ETA: 0sEpoch 1/20\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "25651/25651 [==============================] - 5s - loss: 0.1890 - acc: 0.9304     \n",
      "Epoch 2/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0447 - acc: 0.9912     \n",
      "Epoch 3/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0270 - acc: 0.9938     \n",
      "Epoch 4/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0222 - acc: 0.9945     \n",
      "Epoch 5/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0197 - acc: 0.9950     \n",
      "Epoch 6/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0175 - acc: 0.9956     \n",
      "Epoch 7/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0154 - acc: 0.9959     \n",
      "Epoch 8/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0140 - acc: 0.9964     \n",
      "Epoch 9/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0126 - acc: 0.9968     \n",
      "Epoch 10/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0117 - acc: 0.9972     \n",
      "Epoch 11/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0109 - acc: 0.9971     \n",
      "Epoch 12/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0106 - acc: 0.9974     \n",
      "Epoch 13/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0096 - acc: 0.9977     \n",
      "Epoch 14/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0094 - acc: 0.9975     \n",
      "Epoch 15/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0090 - acc: 0.9975     \n",
      "Epoch 16/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0084 - acc: 0.9978     \n",
      "Epoch 17/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0079 - acc: 0.9978     \n",
      "Epoch 18/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0071 - acc: 0.9982     \n",
      "Epoch 19/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0072 - acc: 0.9980     \n",
      "Epoch 20/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0074 - acc: 0.9981     \n",
      "24660/25651 [===========================>..] - ETA: 0sEpoch 1/20\n",
      "25652/25652 [==============================] - 5s - loss: 0.1490 - acc: 0.9644     \n",
      "Epoch 2/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0293 - acc: 0.9921     \n",
      "Epoch 3/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0234 - acc: 0.9939     \n",
      "Epoch 4/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0204 - acc: 0.9944     \n",
      "Epoch 5/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0183 - acc: 0.9952     \n",
      "Epoch 6/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0166 - acc: 0.9955     \n",
      "Epoch 7/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0151 - acc: 0.9963     \n",
      "Epoch 8/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0139 - acc: 0.9962     \n",
      "Epoch 9/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0129 - acc: 0.9966     \n",
      "Epoch 10/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0121 - acc: 0.9968     \n",
      "Epoch 11/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0109 - acc: 0.9970     \n",
      "Epoch 12/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0103 - acc: 0.9972     \n",
      "Epoch 13/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0095 - acc: 0.9972     \n",
      "Epoch 14/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0090 - acc: 0.9976     \n",
      "Epoch 15/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0082 - acc: 0.9976     \n",
      "Epoch 16/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0081 - acc: 0.9978     \n",
      "Epoch 17/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0079 - acc: 0.9977     \n",
      "Epoch 18/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0072 - acc: 0.9982     \n",
      "Epoch 19/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0070 - acc: 0.9980     \n",
      "Epoch 20/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0067 - acc: 0.9983     \n",
      "24630/25652 [===========================>..] - ETA: 0sEpoch 1/20\n",
      "25652/25652 [==============================] - 5s - loss: 0.1951 - acc: 0.9284     \n",
      "Epoch 2/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0346 - acc: 0.9922     \n",
      "Epoch 3/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0241 - acc: 0.9936     \n",
      "Epoch 4/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0199 - acc: 0.9950     \n",
      "Epoch 5/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0180 - acc: 0.9952     \n",
      "Epoch 6/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0153 - acc: 0.9960     \n",
      "Epoch 7/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0139 - acc: 0.9961     \n",
      "Epoch 8/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0128 - acc: 0.9967     \n",
      "Epoch 9/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0122 - acc: 0.9965     \n",
      "Epoch 10/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0115 - acc: 0.9967     \n",
      "Epoch 11/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0112 - acc: 0.9971     \n",
      "Epoch 12/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0104 - acc: 0.9973     \n",
      "Epoch 13/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0101 - acc: 0.9973     \n",
      "Epoch 14/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0096 - acc: 0.9974     \n",
      "Epoch 15/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0091 - acc: 0.9975     \n",
      "Epoch 16/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0089 - acc: 0.9977     \n",
      "Epoch 17/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0081 - acc: 0.9979     \n",
      "Epoch 18/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0081 - acc: 0.9979     \n",
      "Epoch 19/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0081 - acc: 0.9977     \n",
      "Epoch 20/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0076 - acc: 0.9980     \n",
      "25530/25652 [============================>.] - ETA: 0sEpoch 1/20\n",
      "25652/25652 [==============================] - 5s - loss: 0.1915 - acc: 0.9333     \n",
      "Epoch 2/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0452 - acc: 0.9904     \n",
      "Epoch 3/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0276 - acc: 0.9936     \n",
      "Epoch 4/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0220 - acc: 0.9944     \n",
      "Epoch 5/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0186 - acc: 0.9952     \n",
      "Epoch 6/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0164 - acc: 0.9959     \n",
      "Epoch 7/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0149 - acc: 0.9962     \n",
      "Epoch 8/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0135 - acc: 0.9966     \n",
      "Epoch 9/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0125 - acc: 0.9970     \n",
      "Epoch 10/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0114 - acc: 0.9968     \n",
      "Epoch 11/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0111 - acc: 0.9972     \n",
      "Epoch 12/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0101 - acc: 0.9973     \n",
      "Epoch 13/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0100 - acc: 0.9973     \n",
      "Epoch 14/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0093 - acc: 0.9977     \n",
      "Epoch 15/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0090 - acc: 0.9978     \n",
      "Epoch 16/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0086 - acc: 0.9979     \n",
      "Epoch 17/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0080 - acc: 0.9981     \n",
      "Epoch 18/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0079 - acc: 0.9979     \n",
      "Epoch 19/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0078 - acc: 0.9978     \n",
      "Epoch 20/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0075 - acc: 0.9984     \n",
      "25170/25652 [============================>.] - ETA: 0sEpoch 1/20\n",
      "25652/25652 [==============================] - 5s - loss: 0.1460 - acc: 0.9653     \n",
      "Epoch 2/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0313 - acc: 0.9913     \n",
      "Epoch 3/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0257 - acc: 0.9926     \n",
      "Epoch 4/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0222 - acc: 0.9940     \n",
      "Epoch 5/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0200 - acc: 0.9947     \n",
      "Epoch 6/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0180 - acc: 0.9952     \n",
      "Epoch 7/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0164 - acc: 0.9955     \n",
      "Epoch 8/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0146 - acc: 0.9961     \n",
      "Epoch 9/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0139 - acc: 0.9965     \n",
      "Epoch 10/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0123 - acc: 0.9964     \n",
      "Epoch 11/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0117 - acc: 0.9970     \n",
      "Epoch 12/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0108 - acc: 0.9970     \n",
      "Epoch 13/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0103 - acc: 0.9972     \n",
      "Epoch 14/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0095 - acc: 0.9974     \n",
      "Epoch 15/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0092 - acc: 0.9973     \n",
      "Epoch 16/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0085 - acc: 0.9974     \n",
      "Epoch 17/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0078 - acc: 0.9977     \n",
      "Epoch 18/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0076 - acc: 0.9975     \n",
      "Epoch 19/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0073 - acc: 0.9980     \n",
      "Epoch 20/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0073 - acc: 0.9975     \n",
      "25020/25652 [============================>.] - ETA: 0sEpoch 1/20\n",
      "25652/25652 [==============================] - 6s - loss: 0.1487 - acc: 0.9639     \n",
      "Epoch 2/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0279 - acc: 0.9922     \n",
      "Epoch 3/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0226 - acc: 0.9935     \n",
      "Epoch 4/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0195 - acc: 0.9947     \n",
      "Epoch 5/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0177 - acc: 0.9953     \n",
      "Epoch 6/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0157 - acc: 0.9955     \n",
      "Epoch 7/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0145 - acc: 0.9960     \n",
      "Epoch 8/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0136 - acc: 0.9965     \n",
      "Epoch 9/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0125 - acc: 0.9967     \n",
      "Epoch 10/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0122 - acc: 0.9968     \n",
      "Epoch 11/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0116 - acc: 0.9969     \n",
      "Epoch 12/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0109 - acc: 0.9973     \n",
      "Epoch 13/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0105 - acc: 0.9973     \n",
      "Epoch 14/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0099 - acc: 0.9973     \n",
      "Epoch 15/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0097 - acc: 0.9975     \n",
      "Epoch 16/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0090 - acc: 0.9976     \n",
      "Epoch 17/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0088 - acc: 0.9974     \n",
      "Epoch 18/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0083 - acc: 0.9979     \n",
      "Epoch 19/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0077 - acc: 0.9978     \n",
      "Epoch 20/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0076 - acc: 0.9979     \n",
      "24900/25652 [============================>.] - ETA: 0sEpoch 1/20\n",
      "25652/25652 [==============================] - 5s - loss: 0.1878 - acc: 0.9369     \n",
      "Epoch 2/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0334 - acc: 0.9917     \n",
      "Epoch 3/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0251 - acc: 0.9940     \n",
      "Epoch 4/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0220 - acc: 0.9944     \n",
      "Epoch 5/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0198 - acc: 0.9951     \n",
      "Epoch 6/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0180 - acc: 0.9956     \n",
      "Epoch 7/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0159 - acc: 0.9958     \n",
      "Epoch 8/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0146 - acc: 0.9965     \n",
      "Epoch 9/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0136 - acc: 0.9965     \n",
      "Epoch 10/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0126 - acc: 0.9966     \n",
      "Epoch 11/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0120 - acc: 0.9972     \n",
      "Epoch 12/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0112 - acc: 0.9971     \n",
      "Epoch 13/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0106 - acc: 0.9972     \n",
      "Epoch 14/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0101 - acc: 0.9975     \n",
      "Epoch 15/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0096 - acc: 0.9978     \n",
      "Epoch 16/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0096 - acc: 0.9975     \n",
      "Epoch 17/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0083 - acc: 0.9977     \n",
      "Epoch 18/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0076 - acc: 0.9979     \n",
      "Epoch 19/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0077 - acc: 0.9979     \n",
      "Epoch 20/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0071 - acc: 0.9981     \n",
      "24240/25652 [===========================>..] - ETA: 0sEpoch 1/20\n",
      "25652/25652 [==============================] - 5s - loss: 0.1504 - acc: 0.9639     \n",
      "Epoch 2/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0283 - acc: 0.9924     \n",
      "Epoch 3/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0226 - acc: 0.9936     \n",
      "Epoch 4/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0197 - acc: 0.9948     \n",
      "Epoch 5/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0174 - acc: 0.9952     \n",
      "Epoch 6/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0158 - acc: 0.9957     \n",
      "Epoch 7/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0145 - acc: 0.9961     \n",
      "Epoch 8/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0139 - acc: 0.9966     \n",
      "Epoch 9/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0128 - acc: 0.9968     \n",
      "Epoch 10/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0120 - acc: 0.9966     \n",
      "Epoch 11/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0114 - acc: 0.9970     \n",
      "Epoch 12/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0104 - acc: 0.9972     \n",
      "Epoch 13/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0099 - acc: 0.9975     \n",
      "Epoch 14/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0096 - acc: 0.9974     \n",
      "Epoch 15/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0095 - acc: 0.9975     \n",
      "Epoch 16/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0085 - acc: 0.9979     \n",
      "Epoch 17/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0081 - acc: 0.9979     \n",
      "Epoch 18/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0083 - acc: 0.9976     \n",
      "Epoch 19/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0077 - acc: 0.9979     \n",
      "Epoch 20/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0072 - acc: 0.9981     \n",
      "25650/25652 [============================>.] - ETA: 0s ETA: Epoch 1/20\n",
      "25652/25652 [==============================] - 6s - loss: 0.1423 - acc: 0.9655     \n",
      "Epoch 2/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0268 - acc: 0.9927     \n",
      "Epoch 3/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0210 - acc: 0.9938     \n",
      "Epoch 4/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0181 - acc: 0.9953     \n",
      "Epoch 5/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0160 - acc: 0.9958     \n",
      "Epoch 6/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0147 - acc: 0.9960     \n",
      "Epoch 7/20\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "25652/25652 [==============================] - 3s - loss: 0.0135 - acc: 0.9966     \n",
      "Epoch 8/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0130 - acc: 0.9965     \n",
      "Epoch 9/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0119 - acc: 0.9971     \n",
      "Epoch 10/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0116 - acc: 0.9969     \n",
      "Epoch 11/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0109 - acc: 0.9968     \n",
      "Epoch 12/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0103 - acc: 0.9972     \n",
      "Epoch 13/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0103 - acc: 0.9974     \n",
      "Epoch 14/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0093 - acc: 0.9976     \n",
      "Epoch 15/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0094 - acc: 0.9973     \n",
      "Epoch 16/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0085 - acc: 0.9976     \n",
      "Epoch 17/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0082 - acc: 0.9977     \n",
      "Epoch 18/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0079 - acc: 0.9975     \n",
      "Epoch 19/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0074 - acc: 0.9979     \n",
      "Epoch 20/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0073 - acc: 0.9979     \n",
      "25500/25652 [============================>.] - ETA: 0sEpoch 1/20\n",
      "25651/25651 [==============================] - 5s - loss: 0.1860 - acc: 0.9370     \n",
      "Epoch 2/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0686 - acc: 0.9905     \n",
      "Epoch 3/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0463 - acc: 0.9932     \n",
      "Epoch 4/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0346 - acc: 0.9939     \n",
      "Epoch 5/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0282 - acc: 0.9947     \n",
      "Epoch 6/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0247 - acc: 0.9949     \n",
      "Epoch 7/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0221 - acc: 0.9954     \n",
      "Epoch 8/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0205 - acc: 0.9953     \n",
      "Epoch 9/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0189 - acc: 0.9958     \n",
      "Epoch 10/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0181 - acc: 0.9959     \n",
      "Epoch 11/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0174 - acc: 0.9960     \n",
      "Epoch 12/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0165 - acc: 0.9961     \n",
      "Epoch 13/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0152 - acc: 0.9965     \n",
      "Epoch 14/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0148 - acc: 0.9968     \n",
      "Epoch 15/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0146 - acc: 0.9970     \n",
      "Epoch 16/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0145 - acc: 0.9970     \n",
      "Epoch 17/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0143 - acc: 0.9971     \n",
      "Epoch 18/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0135 - acc: 0.9971     \n",
      "Epoch 19/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0134 - acc: 0.9972     \n",
      "Epoch 20/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0132 - acc: 0.9973     \n",
      "25350/25651 [============================>.] - ETA: 0sEpoch 1/20\n",
      "25651/25651 [==============================] - 5s - loss: 0.1630 - acc: 0.9599     \n",
      "Epoch 2/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0320 - acc: 0.9913     \n",
      "Epoch 3/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0243 - acc: 0.9940     \n",
      "Epoch 4/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0215 - acc: 0.9946     \n",
      "Epoch 5/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0203 - acc: 0.9950     \n",
      "Epoch 6/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0191 - acc: 0.9953     \n",
      "Epoch 7/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0179 - acc: 0.9956     \n",
      "Epoch 8/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0168 - acc: 0.9959     \n",
      "Epoch 9/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0157 - acc: 0.9961     \n",
      "Epoch 10/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0152 - acc: 0.9963     \n",
      "Epoch 11/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0147 - acc: 0.9962     \n",
      "Epoch 12/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0137 - acc: 0.9966     \n",
      "Epoch 13/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0136 - acc: 0.9968     \n",
      "Epoch 14/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0127 - acc: 0.9970     \n",
      "Epoch 15/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0125 - acc: 0.9971     \n",
      "Epoch 16/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0125 - acc: 0.9970     \n",
      "Epoch 17/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0118 - acc: 0.9972     \n",
      "Epoch 18/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0113 - acc: 0.9975     \n",
      "Epoch 19/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0109 - acc: 0.9974     \n",
      "Epoch 20/20\n",
      "25651/25651 [==============================] - 3s - loss: 0.0097 - acc: 0.9975     \n",
      "25200/25651 [============================>.] - ETA: 0sEpoch 1/20\n",
      "25652/25652 [==============================] - 5s - loss: 0.1653 - acc: 0.9556     \n",
      "Epoch 2/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0303 - acc: 0.9919     \n",
      "Epoch 3/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0238 - acc: 0.9938     \n",
      "Epoch 4/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0212 - acc: 0.9946     \n",
      "Epoch 5/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0192 - acc: 0.9949     \n",
      "Epoch 6/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0179 - acc: 0.9954     \n",
      "Epoch 7/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0167 - acc: 0.9958     \n",
      "Epoch 8/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0159 - acc: 0.9959     \n",
      "Epoch 9/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0149 - acc: 0.9959     \n",
      "Epoch 10/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0138 - acc: 0.9963     \n",
      "Epoch 11/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0133 - acc: 0.9965     \n",
      "Epoch 12/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0134 - acc: 0.9967     \n",
      "Epoch 13/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0129 - acc: 0.9966     \n",
      "Epoch 14/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0116 - acc: 0.9970     \n",
      "Epoch 15/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0114 - acc: 0.9971     \n",
      "Epoch 16/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0106 - acc: 0.9973     \n",
      "Epoch 17/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0112 - acc: 0.9972     \n",
      "Epoch 18/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0105 - acc: 0.9972     \n",
      "Epoch 19/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0099 - acc: 0.9977     \n",
      "Epoch 20/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0100 - acc: 0.9977     \n",
      "25260/25652 [============================>.] - ETA: 0sEpoch 1/20\n",
      "25652/25652 [==============================] - 5s - loss: 0.1627 - acc: 0.9572     \n",
      "Epoch 2/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0308 - acc: 0.9915     \n",
      "Epoch 3/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0250 - acc: 0.9933     \n",
      "Epoch 4/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0222 - acc: 0.9944     \n",
      "Epoch 5/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0206 - acc: 0.9947     \n",
      "Epoch 6/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0191 - acc: 0.9952     \n",
      "Epoch 7/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0185 - acc: 0.9952     \n",
      "Epoch 8/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0166 - acc: 0.9960     \n",
      "Epoch 9/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0160 - acc: 0.9961     \n",
      "Epoch 10/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0155 - acc: 0.9959     \n",
      "Epoch 11/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0144 - acc: 0.9964     \n",
      "Epoch 12/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0138 - acc: 0.9964     \n",
      "Epoch 13/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0133 - acc: 0.9965     \n",
      "Epoch 14/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0132 - acc: 0.9964     \n",
      "Epoch 15/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0124 - acc: 0.9967     \n",
      "Epoch 16/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0120 - acc: 0.9966     \n",
      "Epoch 17/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0119 - acc: 0.9970     \n",
      "Epoch 18/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0111 - acc: 0.9967     \n",
      "Epoch 19/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0107 - acc: 0.9973     \n",
      "Epoch 20/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0107 - acc: 0.9973     \n",
      "24840/25652 [============================>.] - ETA: 0sEpoch 1/20\n",
      "25652/25652 [==============================] - 5s - loss: 0.1946 - acc: 0.9345     \n",
      "Epoch 2/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0425 - acc: 0.9904     \n",
      "Epoch 3/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0270 - acc: 0.9935     \n",
      "Epoch 4/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0232 - acc: 0.9943     \n",
      "Epoch 5/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0207 - acc: 0.9951     \n",
      "Epoch 6/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0192 - acc: 0.9952     \n",
      "Epoch 7/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0176 - acc: 0.9958     \n",
      "Epoch 8/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0161 - acc: 0.9962     \n",
      "Epoch 9/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0150 - acc: 0.9964     \n",
      "Epoch 10/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0141 - acc: 0.9966     \n",
      "Epoch 11/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0129 - acc: 0.9968     \n",
      "Epoch 12/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0122 - acc: 0.9970     \n",
      "Epoch 13/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0114 - acc: 0.9970     \n",
      "Epoch 14/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0109 - acc: 0.9973     \n",
      "Epoch 15/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0107 - acc: 0.9975     \n",
      "Epoch 16/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0102 - acc: 0.9975     \n",
      "Epoch 17/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0103 - acc: 0.9974     \n",
      "Epoch 18/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0099 - acc: 0.9977     \n",
      "Epoch 19/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0102 - acc: 0.9974     \n",
      "Epoch 20/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0093 - acc: 0.9976     \n",
      "24780/25652 [===========================>..] - ETA: 0sEpoch 1/20\n",
      "25652/25652 [==============================] - 5s - loss: 0.1667 - acc: 0.9612     \n",
      "Epoch 2/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0308 - acc: 0.9916     \n",
      "Epoch 3/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0241 - acc: 0.9942     \n",
      "Epoch 4/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0212 - acc: 0.9947     \n",
      "Epoch 5/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0193 - acc: 0.9953     \n",
      "Epoch 6/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0174 - acc: 0.9955     \n",
      "Epoch 7/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0162 - acc: 0.9961     \n",
      "Epoch 8/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0160 - acc: 0.9960     \n",
      "Epoch 9/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0150 - acc: 0.9962     \n",
      "Epoch 10/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0143 - acc: 0.9965     \n",
      "Epoch 11/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0137 - acc: 0.9966     \n",
      "Epoch 12/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0136 - acc: 0.9968     \n",
      "Epoch 13/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0130 - acc: 0.9969     \n",
      "Epoch 14/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0122 - acc: 0.9968     \n",
      "Epoch 15/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0120 - acc: 0.9972     \n",
      "Epoch 16/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0113 - acc: 0.9973     \n",
      "Epoch 17/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0107 - acc: 0.9972     \n",
      "Epoch 18/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0102 - acc: 0.9973     \n",
      "Epoch 19/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0097 - acc: 0.9974     \n",
      "Epoch 20/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0092 - acc: 0.9977     \n",
      "24570/25652 [===========================>..] - ETA: 0sEpoch 1/20\n",
      "25652/25652 [==============================] - 5s - loss: 0.1981 - acc: 0.9317     \n",
      "Epoch 2/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0717 - acc: 0.9888     \n",
      "Epoch 3/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0472 - acc: 0.9927     \n",
      "Epoch 4/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0344 - acc: 0.9942     \n",
      "Epoch 5/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0287 - acc: 0.9949     \n",
      "Epoch 6/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0250 - acc: 0.9950     \n",
      "Epoch 7/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0220 - acc: 0.9955     \n",
      "Epoch 8/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0209 - acc: 0.9955     \n",
      "Epoch 9/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0195 - acc: 0.9961     \n",
      "Epoch 10/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0185 - acc: 0.9961     \n",
      "Epoch 11/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0174 - acc: 0.9962     \n",
      "Epoch 12/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0163 - acc: 0.9965     \n",
      "Epoch 13/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0164 - acc: 0.9964     \n",
      "Epoch 14/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0160 - acc: 0.9965     \n",
      "Epoch 15/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0148 - acc: 0.9966     \n",
      "Epoch 16/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0145 - acc: 0.9969     \n",
      "Epoch 17/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0140 - acc: 0.9970     \n",
      "Epoch 18/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0136 - acc: 0.9971     \n",
      "Epoch 19/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0131 - acc: 0.9973     \n",
      "Epoch 20/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0123 - acc: 0.9970     \n",
      "24330/25652 [===========================>..] - ETA: 0sEpoch 1/20\n",
      "25652/25652 [==============================] - 5s - loss: 0.1918 - acc: 0.9362     \n",
      "Epoch 2/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0377 - acc: 0.9913     \n",
      "Epoch 3/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0257 - acc: 0.9936     \n",
      "Epoch 4/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0228 - acc: 0.9944     \n",
      "Epoch 5/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0206 - acc: 0.9952     \n",
      "Epoch 6/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0196 - acc: 0.9954     \n",
      "Epoch 7/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0183 - acc: 0.9957     \n",
      "Epoch 8/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0179 - acc: 0.9958     \n",
      "Epoch 9/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0169 - acc: 0.9960     \n",
      "Epoch 10/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0166 - acc: 0.9961     \n",
      "Epoch 11/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0159 - acc: 0.9965     \n",
      "Epoch 12/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0154 - acc: 0.9966     \n",
      "Epoch 13/20\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "25652/25652 [==============================] - 3s - loss: 0.0152 - acc: 0.9966     \n",
      "Epoch 14/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0148 - acc: 0.9968     \n",
      "Epoch 15/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0139 - acc: 0.9970     \n",
      "Epoch 16/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0140 - acc: 0.9970     \n",
      "Epoch 17/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0132 - acc: 0.9971     \n",
      "Epoch 18/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0130 - acc: 0.9973     \n",
      "Epoch 19/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0130 - acc: 0.9971     \n",
      "Epoch 20/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0122 - acc: 0.9974     \n",
      "24360/25652 [===========================>..] - ETA: 0sEpoch 1/20\n",
      "25652/25652 [==============================] - 5s - loss: 0.1999 - acc: 0.9318     \n",
      "Epoch 2/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0410 - acc: 0.9903     \n",
      "Epoch 3/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0256 - acc: 0.9936     \n",
      "Epoch 4/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0223 - acc: 0.9949     \n",
      "Epoch 5/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0199 - acc: 0.9953     \n",
      "Epoch 6/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0185 - acc: 0.9957     \n",
      "Epoch 7/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0167 - acc: 0.9961     \n",
      "Epoch 8/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0157 - acc: 0.9962     \n",
      "Epoch 9/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0149 - acc: 0.9965     \n",
      "Epoch 10/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0141 - acc: 0.9969     \n",
      "Epoch 11/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0131 - acc: 0.9970     \n",
      "Epoch 12/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0127 - acc: 0.9972     \n",
      "Epoch 13/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0127 - acc: 0.9972     \n",
      "Epoch 14/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0121 - acc: 0.9976     \n",
      "Epoch 15/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0113 - acc: 0.9977     \n",
      "Epoch 16/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0106 - acc: 0.9975     \n",
      "Epoch 17/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0101 - acc: 0.9976     \n",
      "Epoch 18/20\n",
      "25652/25652 [==============================] - ETA: 0s - loss: 0.0098 - acc: 0.997 - 3s - loss: 0.0100 - acc: 0.9979     \n",
      "Epoch 19/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0098 - acc: 0.9977     \n",
      "Epoch 20/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0095 - acc: 0.9977     \n",
      "25440/25652 [============================>.] - ETA: 0sEpoch 1/20\n",
      "25652/25652 [==============================] - 5s - loss: 0.1546 - acc: 0.9616     \n",
      "Epoch 2/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0293 - acc: 0.9920     \n",
      "Epoch 3/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0234 - acc: 0.9935     \n",
      "Epoch 4/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0211 - acc: 0.9949     \n",
      "Epoch 5/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0191 - acc: 0.9951     \n",
      "Epoch 6/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0178 - acc: 0.9959     \n",
      "Epoch 7/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0166 - acc: 0.9957     \n",
      "Epoch 8/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0156 - acc: 0.9959     \n",
      "Epoch 9/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0150 - acc: 0.9965     \n",
      "Epoch 10/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0141 - acc: 0.9966     \n",
      "Epoch 11/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0137 - acc: 0.9971     \n",
      "Epoch 12/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0137 - acc: 0.9970     \n",
      "Epoch 13/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0130 - acc: 0.9970     \n",
      "Epoch 14/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0122 - acc: 0.9971     \n",
      "Epoch 15/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0124 - acc: 0.9972     \n",
      "Epoch 16/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0116 - acc: 0.9976     \n",
      "Epoch 17/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0111 - acc: 0.9974     \n",
      "Epoch 18/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0116 - acc: 0.9976     \n",
      "Epoch 19/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0108 - acc: 0.9974     \n",
      "Epoch 20/20\n",
      "25652/25652 [==============================] - 3s - loss: 0.0102 - acc: 0.9977     \n",
      "25410/25652 [============================>.] - ETA: 0sEpoch 1/20\n",
      "28502/28502 [==============================] - 15s - loss: 0.0704 - acc: 0.9826    \n",
      "Epoch 2/20\n",
      "28502/28502 [==============================] - 12s - loss: 0.0223 - acc: 0.9940    \n",
      "Epoch 3/20\n",
      "28502/28502 [==============================] - 12s - loss: 0.0175 - acc: 0.9955    \n",
      "Epoch 4/20\n",
      "28502/28502 [==============================] - 12s - loss: 0.0157 - acc: 0.9959    \n",
      "Epoch 5/20\n",
      "28502/28502 [==============================] - 12s - loss: 0.0132 - acc: 0.9966    \n",
      "Epoch 6/20\n",
      "28502/28502 [==============================] - 12s - loss: 0.0127 - acc: 0.9965    \n",
      "Epoch 7/20\n",
      "28502/28502 [==============================] - 12s - loss: 0.0115 - acc: 0.9971    \n",
      "Epoch 8/20\n",
      "28502/28502 [==============================] - 12s - loss: 0.0107 - acc: 0.9971    \n",
      "Epoch 9/20\n",
      "28502/28502 [==============================] - 12s - loss: 0.0104 - acc: 0.9971    \n",
      "Epoch 10/20\n",
      "28502/28502 [==============================] - 12s - loss: 0.0098 - acc: 0.9975    \n",
      "Epoch 11/20\n",
      "28502/28502 [==============================] - 12s - loss: 0.0093 - acc: 0.9973    \n",
      "Epoch 12/20\n",
      "28502/28502 [==============================] - 12s - loss: 0.0090 - acc: 0.9974    \n",
      "Epoch 13/20\n",
      "28502/28502 [==============================] - 12s - loss: 0.0084 - acc: 0.9975    \n",
      "Epoch 14/20\n",
      "28502/28502 [==============================] - 12s - loss: 0.0087 - acc: 0.9973    \n",
      "Epoch 15/20\n",
      "28502/28502 [==============================] - 12s - loss: 0.0079 - acc: 0.9980    \n",
      "Epoch 16/20\n",
      "28502/28502 [==============================] - 12s - loss: 0.0076 - acc: 0.9979    \n",
      "Epoch 17/20\n",
      "28502/28502 [==============================] - 12s - loss: 0.0067 - acc: 0.9980    \n",
      "Epoch 18/20\n",
      "28502/28502 [==============================] - 12s - loss: 0.0069 - acc: 0.9981    \n",
      "Epoch 19/20\n",
      "28502/28502 [==============================] - 12s - loss: 0.0063 - acc: 0.9980    \n",
      "Epoch 20/20\n",
      "28502/28502 [==============================] - 12s - loss: 0.0068 - acc: 0.9979    \n"
     ]
    }
   ],
   "source": [
    "# Optimization\n",
    "from keras.wrappers.scikit_learn import KerasClassifier\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "def build_classifier_imp(optimizer):\n",
    "    classifier_op = Sequential()\n",
    "    classifier_op.add(Dense(units = 14, kernel_initializer = 'uniform', activation = 'relu', input_dim = 27))\n",
    "    classifier_op.add(Dense(units = 14, kernel_initializer = 'uniform', activation = 'relu'))\n",
    "    classifier_op.add(Dense(units = 1, kernel_initializer = 'uniform', activation = 'sigmoid'))\n",
    "    classifier_op.compile(optimizer = optimizer, loss = 'binary_crossentropy', metrics = ['accuracy'])\n",
    "    return classifier_op\n",
    "\n",
    "classifier_op = KerasClassifier(build_fn = build_classifier_imp)\n",
    "\n",
    "parameters = {'batch_size': [10, 30],\n",
    "              'epochs': [10, 20],\n",
    "              'optimizer': ['adam', 'rmsprop']}\n",
    "\n",
    "grid_search = GridSearchCV(estimator = classifier_op,\n",
    "                           param_grid = parameters,\n",
    "                           scoring = 'accuracy',\n",
    "                           cv = 10)\n",
    "\n",
    "grid_search = grid_search.fit(X_train, y_train)\n",
    "\n",
    "best_parameters = grid_search.best_params_\n",
    "best_accuracy = grid_search.best_score_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'batch_size': 10, 'epochs': 20, 'optimizer': 'adam'} 0.997438776226\n"
     ]
    }
   ],
   "source": [
    "print (best_parameters, best_accuracy )"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Fitting and prediction with optimized model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/20\n",
      "28502/28502 [==============================] - 12s - loss: 0.0092 - acc: 0.9973    \n",
      "Epoch 2/20\n",
      "28502/28502 [==============================] - 12s - loss: 0.0086 - acc: 0.9974    \n",
      "Epoch 3/20\n",
      "28502/28502 [==============================] - 12s - loss: 0.0083 - acc: 0.9975    \n",
      "Epoch 4/20\n",
      "28502/28502 [==============================] - 12s - loss: 0.0076 - acc: 0.9980    \n",
      "Epoch 5/20\n",
      "28502/28502 [==============================] - 12s - loss: 0.0075 - acc: 0.9978    \n",
      "Epoch 6/20\n",
      "28502/28502 [==============================] - 12s - loss: 0.0077 - acc: 0.9975    \n",
      "Epoch 7/20\n",
      "28502/28502 [==============================] - 12s - loss: 0.0073 - acc: 0.9978    \n",
      "Epoch 8/20\n",
      "28502/28502 [==============================] - 12s - loss: 0.0067 - acc: 0.9979    \n",
      "Epoch 9/20\n",
      "28502/28502 [==============================] - 12s - loss: 0.0069 - acc: 0.9979    \n",
      "Epoch 10/20\n",
      "28502/28502 [==============================] - 12s - loss: 0.0064 - acc: 0.9981    \n",
      "Epoch 11/20\n",
      "28502/28502 [==============================] - 12s - loss: 0.0074 - acc: 0.9979    \n",
      "Epoch 12/20\n",
      "28502/28502 [==============================] - 12s - loss: 0.0062 - acc: 0.9981    \n",
      "Epoch 13/20\n",
      "28502/28502 [==============================] - 12s - loss: 0.0061 - acc: 0.9981    \n",
      "Epoch 14/20\n",
      "28502/28502 [==============================] - 12s - loss: 0.0061 - acc: 0.9982    \n",
      "Epoch 15/20\n",
      "28502/28502 [==============================] - 12s - loss: 0.0056 - acc: 0.9983    \n",
      "Epoch 16/20\n",
      "28502/28502 [==============================] - 12s - loss: 0.0060 - acc: 0.9982    \n",
      "Epoch 17/20\n",
      "28502/28502 [==============================] - 12s - loss: 0.0057 - acc: 0.9982    \n",
      "Epoch 18/20\n",
      "28502/28502 [==============================] - 12s - loss: 0.0051 - acc: 0.9985    \n",
      "Epoch 19/20\n",
      "28502/28502 [==============================] - 12s - loss: 0.0057 - acc: 0.9980    \n",
      "Epoch 20/20\n",
      "28502/28502 [==============================] - 12s - loss: 0.0049 - acc: 0.9985    \n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x137891048>"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Training with optimize settings\n",
    "classifier.fit(X_train, y_train, batch_size = 10, epochs = 20)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from keras.models import load_model\n",
    "classifier.save('data_files/classifier_opt.hdf5') "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy = 99.7474%, Loss = 0.0123\n"
     ]
    }
   ],
   "source": [
    "# Accuracy on the validation set\n",
    "loss, accuracy = classifier.evaluate(X_val, y_val,batch_size=128, verbose=0)\n",
    "print(\"Accuracy = {:.4f}%, Loss = {:.4f}\".format(accuracy* 100, loss))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Implementation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Predicting on the test set \n",
    "y_pred = classifier.predict(X_test)\n",
    "y_pred = (y_pred > 0.5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy = 99.7137%, Loss = 0.0122\n"
     ]
    }
   ],
   "source": [
    "# Accuracy on the test set\n",
    "loss, accuracy = classifier.evaluate(X_test, y_test,batch_size=128, verbose=0)\n",
    "print(\"Accuracy = {:.4f}%, Loss = {:.4f}\".format(accuracy* 100, loss))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[10533,    12],\n",
       "       [   22,  1310]])"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.metrics import confusion_matrix\n",
    "cnf_matrix = confusion_matrix(y_test, y_pred, labels=[0, 1])\n",
    "cnf_matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion matrix, without normalization\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjcAAAI4CAYAAACWfsh7AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nSUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1378bd5f8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "pl.figure(figsize=(8,8))\n",
    "plot_confusion_matrix(cnf_matrix, classes=n_loan_no_out.loan_status.unique(),\n",
    "                      cmap=pl.cm.Reds, normalize=False)\n",
    "\n",
    "\n",
    "pl.savefig('data_files/Conf_matrix_2017.png', dpi = 300)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.98718914845516192"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.metrics import f1_score\n",
    "f1_score(y_test, y_pred.round(), average='binary')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['Current', 'Fully Paid', 'Late (31-120 days)', 'Charged Off',\n",
       "       'In Grace Period', 'Late (16-30 days)', 'Default',\n",
       "       'Does not meet the credit policy. Status:Fully Paid',\n",
       "       'Does not meet the credit policy. Status:Charged Off'], dtype=object)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "n_loan_df.loan_status.unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "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": "3.6.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}