Loan Default Prediction EDA.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1. Introduction\n",
    "\n",
    "* We aim to predict the likelihood of whether a person will default based information such as loan amount, annual income, and FICO score\n",
    "* The dataset has 74 variables and 887,379 rows\n",
    "* The response variable is **loan_status** which has 10 unique categories in the raw dataset\n",
    "* We will perform some exploratory data analysis here"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import seaborn as sns\n",
    "from sklearn.model_selection import train_test_split\n",
    "import matplotlib\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "# plt.rcParams[\"figure.figsize\"] = fig_size"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2. EDA on Response Variable"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/Mengxin/anaconda2/lib/python2.7/site-packages/IPython/core/interactiveshell.py:2717: DtypeWarning: Columns (19,55) have mixed types. Specify dtype option on import or set low_memory=False.\n",
      "  interactivity=interactivity, compiler=compiler, result=result)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "(887379, 74)"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# first get the dimension of the dataset\n",
    "loan = pd.read_csv(\"loan.csv\")\n",
    "loan.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2.1 Count by loan status type (raw dataset)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "loan_status\n",
       "Charged Off                                             45248\n",
       "Current                                                601779\n",
       "Default                                                  1219\n",
       "Does not meet the credit policy. Status:Charged Off       761\n",
       "Does not meet the credit policy. Status:Fully Paid       1988\n",
       "Fully Paid                                             207723\n",
       "In Grace Period                                          6253\n",
       "Issued                                                   8460\n",
       "Late (16-30 days)                                        2357\n",
       "Late (31-120 days)                                      11591\n",
       "dtype: int64"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# loan_status is the response variable\n",
    "statusByType = loan.groupby('loan_status').size()\n",
    "statusByType"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x135728d50>"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA1MAAANICAYAAAAiqbXEAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nJEmSJA2DxZQkSZIkDYPFlCRJkiQNw/8DkHjSF0Fvk0EAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11799df50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig_size=[]\n",
    "fig_size.append(10)\n",
    "fig_size.append(6)\n",
    "plt.rcParams[\"figure.figsize\"] = fig_size\n",
    "statusByType.plot(kind='bar')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Summary: \n",
    "\n",
    "As the problem at hand is to predict whether a person will default, loans whose status is \"current\", \"issued\", or \"late\" are not closed yet. Therefore, we should drop observations assosiated with these categories."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2.2 Recreate response variable"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "loan_status\n",
       "Charged Off                                             45248\n",
       "Default                                                  1219\n",
       "Does not meet the credit policy. Status:Charged Off       761\n",
       "Does not meet the credit policy. Status:Fully Paid       1988\n",
       "Fully Paid                                             207723\n",
       "dtype: int64"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## First drop categories that don't indicate whether a loan is defaulted\n",
    "mask = loan['loan_status'].isin(['Default', 'Charged Off', 'Does not meet the credit policy. Status:Charged Off',\n",
    "                                'Does not meet the credit policy. Status:Fully Paid', 'Fully Paid'])\n",
    "loan_new = loan[mask]\n",
    "loan_new.groupby('loan_status').size()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/Mengxin/anaconda2/lib/python2.7/site-packages/ipykernel/__main__.py:4: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "status\n",
       "paid      209711\n",
       "unpaid     47228\n",
       "dtype: int64"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Next, regroup response variable\n",
    "loan_new['status'] = np.where(loan_new['loan_status'].\n",
    "                              isin(['Default', 'Charged Off', 'Does not meet the credit policy. Status:Charged Off']),\n",
    "                              'unpaid', 'paid')\n",
    "loan_new.groupby('status').size()       "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### NOTE: By doing this filtering, we artifiially inflate the proportion of loans that default or are charged off. That's because defaulted and charged off loans reach theie termination (end status) faster than full paid loans in most cases."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'statusByType_train' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-8-20c909e2df40>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m      3\u001b[0m \u001b[0mfig_size\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mappend\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;36m6\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      4\u001b[0m \u001b[0mplt\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrcParams\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m\"figure.figsize\"\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mfig_size\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 5\u001b[0;31m \u001b[0mstatusByType_train\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mplot\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mkind\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m'bar'\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mwidth\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m0.3\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m: name 'statusByType_train' is not defined"
     ]
    }
   ],
   "source": [
    "fig_size=[]\n",
    "fig_size.append(10)\n",
    "fig_size.append(6)\n",
    "plt.rcParams[\"figure.figsize\"] = fig_size\n",
    "statusByType_train.plot(kind='bar', width=0.3)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2.3 Column Removal\n",
    "\n",
    "* There are several columns that have the majority of values being NA's, therefore we drop the columsn with more than 100,000 NA's\n",
    "* First let's count the # of NA's in each column"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "member_id                      0\n",
       "loan_amnt                      0\n",
       "funded_amnt                    0\n",
       "funded_amnt_inv                0\n",
       "term                           0\n",
       "int_rate                       0\n",
       "installment                    0\n",
       "grade                          0\n",
       "sub_grade                      0\n",
       "emp_title                  14165\n",
       "emp_length                     0\n",
       "home_ownership                 0\n",
       "annual_inc                     4\n",
       "verification_status            0\n",
       "issue_d                        0\n",
       "loan_status                    0\n",
       "pymnt_plan                     0\n",
       "url                            0\n",
       "desc                      166070\n",
       "purpose                        0\n",
       "title                         14\n",
       "zip_code                       0\n",
       "addr_state                     0\n",
       "dti                            0\n",
       "delinq_2yrs                   29\n",
       "earliest_cr_line              29\n",
       "inq_last_6mths                29\n",
       "mths_since_last_delinq    142645\n",
       "mths_since_last_record    224265\n",
       "open_acc                      29\n",
       "pub_rec                       29\n",
       "revol_bal                      0\n",
       "revol_util                   240\n",
       "total_acc                     29\n",
       "dtype: int64"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "loan_new.isnull().sum()[1:35]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "initial_list_status                 0\n",
       "out_prncp                           0\n",
       "out_prncp_inv                       0\n",
       "total_pymnt                         0\n",
       "total_pymnt_inv                     0\n",
       "total_rec_prncp                     0\n",
       "total_rec_int                       0\n",
       "total_rec_late_fee                  0\n",
       "recoveries                          0\n",
       "collection_recovery_fee             0\n",
       "last_pymnt_d                      548\n",
       "last_pymnt_amnt                     0\n",
       "next_pymnt_d                   252971\n",
       "last_credit_pull_d                 23\n",
       "collections_12_mths_ex_med        145\n",
       "mths_since_last_major_derog    209063\n",
       "policy_code                         0\n",
       "application_type                    0\n",
       "annual_inc_joint               256938\n",
       "dti_joint                      256938\n",
       "verification_status_joint      256938\n",
       "acc_now_delinq                     29\n",
       "tot_coll_amt                    66475\n",
       "tot_cur_bal                     66475\n",
       "open_acc_6m                    256795\n",
       "open_il_6m                     256795\n",
       "open_il_12m                    256795\n",
       "open_il_24m                    256795\n",
       "mths_since_rcnt_il             256799\n",
       "total_bal_il                   256795\n",
       "il_util                        256813\n",
       "open_rv_12m                    256795\n",
       "open_rv_24m                    256795\n",
       "max_bal_bc                     256795\n",
       "all_util                       256795\n",
       "total_rev_hi_lim                66475\n",
       "inq_fi                         256795\n",
       "total_cu_tl                    256795\n",
       "inq_last_12m                   256795\n",
       "status                              0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "loan_new.isnull().sum()[35:]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* Next let's drop columns with more than 100,000 NA's"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def drop_sparse_columns(df):\n",
    "    temp = df\n",
    "    drop_col = []\n",
    "    column_names = temp.columns.values\n",
    "    for name in column_names:\n",
    "        try:\n",
    "            if temp[name].isnull().sum() > 100000: # drop columns with more than 80% NA's\n",
    "                drop_col.append(name)\n",
    "                temp.drop(labels=name, axis=1, inplace=True)\n",
    "        except ValueError:\n",
    "            pass\n",
    "    return temp, drop_col"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 174,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/anaconda/lib/python2.7/site-packages/ipykernel/__main__.py:9: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "(256939, 53)"
      ]
     },
     "execution_count": 174,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "loan_new, drop_name = drop_sparse_columns(loan_new)\n",
    "loan_new.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 175,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['desc',\n",
       " 'mths_since_last_delinq',\n",
       " 'mths_since_last_record',\n",
       " 'next_pymnt_d',\n",
       " 'mths_since_last_major_derog',\n",
       " 'annual_inc_joint',\n",
       " 'dti_joint',\n",
       " 'verification_status_joint',\n",
       " 'open_acc_6m',\n",
       " 'open_il_6m',\n",
       " 'open_il_12m',\n",
       " 'open_il_24m',\n",
       " 'mths_since_rcnt_il',\n",
       " 'total_bal_il',\n",
       " 'il_util',\n",
       " 'open_rv_12m',\n",
       " 'open_rv_24m',\n",
       " 'max_bal_bc',\n",
       " 'all_util',\n",
       " 'inq_fi',\n",
       " 'total_cu_tl',\n",
       " 'inq_last_12m']"
      ]
     },
     "execution_count": 175,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "drop_name"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 176,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/anaconda/lib/python2.7/site-packages/ipykernel/__main__.py:16: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy\n"
     ]
    }
   ],
   "source": [
    "# further drop other columns\n",
    "loan_new.drop(labels=['collection_recovery_fee',\n",
    "                    'last_pymnt_amnt',\n",
    "                    'last_pymnt_d',\n",
    "                    'out_prncp',\n",
    "                    'out_prncp_inv',\n",
    "                    'pymnt_plan',\n",
    "                    'recoveries',\n",
    "                    'term',\n",
    "                    'title',\n",
    "                    'total_pymnt',\n",
    "                    'total_pymnt_inv',\n",
    "                    'total_rec_int',\n",
    "                    'total_rec_late_fee',\n",
    "                    'total_rec_prncp',\n",
    "                    'url'], axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 177,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(256939, 38)"
      ]
     },
     "execution_count": 177,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "loan_new.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 181,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\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>int_rate</th>\n",
       "      <th>installment</th>\n",
       "      <th>grade</th>\n",
       "      <th>sub_grade</th>\n",
       "      <th>emp_title</th>\n",
       "      <th>emp_length</th>\n",
       "      <th>home_ownership</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>2.569390e+05</td>\n",
       "      <td>2.569390e+05</td>\n",
       "      <td>256939.000000</td>\n",
       "      <td>256939.000000</td>\n",
       "      <td>256939.000000</td>\n",
       "      <td>256939.000000</td>\n",
       "      <td>256939.000000</td>\n",
       "      <td>256939</td>\n",
       "      <td>256939</td>\n",
       "      <td>242774</td>\n",
       "      <td>256939</td>\n",
       "      <td>256939</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>unique</th>\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>7</td>\n",
       "      <td>35</td>\n",
       "      <td>134759</td>\n",
       "      <td>12</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>top</th>\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>B</td>\n",
       "      <td>B3</td>\n",
       "      <td>Manager</td>\n",
       "      <td>10+ years</td>\n",
       "      <td>MORTGAGE</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>freq</th>\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>76617</td>\n",
       "      <td>18179</td>\n",
       "      <td>1959</td>\n",
       "      <td>77726</td>\n",
       "      <td>126598</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>1.221284e+07</td>\n",
       "      <td>1.352047e+07</td>\n",
       "      <td>13522.115950</td>\n",
       "      <td>13479.995641</td>\n",
       "      <td>13360.648099</td>\n",
       "      <td>13.780014</td>\n",
       "      <td>416.923016</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>1.476914e+07</td>\n",
       "      <td>1.573908e+07</td>\n",
       "      <td>8128.811481</td>\n",
       "      <td>8106.456843</td>\n",
       "      <td>8141.904114</td>\n",
       "      <td>4.389704</td>\n",
       "      <td>244.878760</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>5.473400e+04</td>\n",
       "      <td>7.047300e+04</td>\n",
       "      <td>500.000000</td>\n",
       "      <td>500.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>5.320000</td>\n",
       "      <td>15.670000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>1.438591e+06</td>\n",
       "      <td>1.683110e+06</td>\n",
       "      <td>7200.000000</td>\n",
       "      <td>7200.000000</td>\n",
       "      <td>7100.000000</td>\n",
       "      <td>10.740000</td>\n",
       "      <td>238.130000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>6.245475e+06</td>\n",
       "      <td>7.451894e+06</td>\n",
       "      <td>12000.000000</td>\n",
       "      <td>12000.000000</td>\n",
       "      <td>11950.000000</td>\n",
       "      <td>13.550000</td>\n",
       "      <td>364.290000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>1.643198e+07</td>\n",
       "      <td>1.847432e+07</td>\n",
       "      <td>18200.000000</td>\n",
       "      <td>18000.000000</td>\n",
       "      <td>18000.000000</td>\n",
       "      <td>16.550000</td>\n",
       "      <td>546.430000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>6.860466e+07</td>\n",
       "      <td>7.350742e+07</td>\n",
       "      <td>35000.000000</td>\n",
       "      <td>35000.000000</td>\n",
       "      <td>35000.000000</td>\n",
       "      <td>28.990000</td>\n",
       "      <td>1424.570000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  id     member_id      loan_amnt    funded_amnt  \\\n",
       "count   2.569390e+05  2.569390e+05  256939.000000  256939.000000   \n",
       "unique           NaN           NaN            NaN            NaN   \n",
       "top              NaN           NaN            NaN            NaN   \n",
       "freq             NaN           NaN            NaN            NaN   \n",
       "mean    1.221284e+07  1.352047e+07   13522.115950   13479.995641   \n",
       "std     1.476914e+07  1.573908e+07    8128.811481    8106.456843   \n",
       "min     5.473400e+04  7.047300e+04     500.000000     500.000000   \n",
       "25%     1.438591e+06  1.683110e+06    7200.000000    7200.000000   \n",
       "50%     6.245475e+06  7.451894e+06   12000.000000   12000.000000   \n",
       "75%     1.643198e+07  1.847432e+07   18200.000000   18000.000000   \n",
       "max     6.860466e+07  7.350742e+07   35000.000000   35000.000000   \n",
       "\n",
       "        funded_amnt_inv       int_rate    installment   grade sub_grade  \\\n",
       "count     256939.000000  256939.000000  256939.000000  256939    256939   \n",
       "unique              NaN            NaN            NaN       7        35   \n",
       "top                 NaN            NaN            NaN       B        B3   \n",
       "freq                NaN            NaN            NaN   76617     18179   \n",
       "mean       13360.648099      13.780014     416.923016     NaN       NaN   \n",
       "std         8141.904114       4.389704     244.878760     NaN       NaN   \n",
       "min            0.000000       5.320000      15.670000     NaN       NaN   \n",
       "25%         7100.000000      10.740000     238.130000     NaN       NaN   \n",
       "50%        11950.000000      13.550000     364.290000     NaN       NaN   \n",
       "75%        18000.000000      16.550000     546.430000     NaN       NaN   \n",
       "max        35000.000000      28.990000    1424.570000     NaN       NaN   \n",
       "\n",
       "       emp_title emp_length home_ownership  \n",
       "count     242774     256939         256939  \n",
       "unique    134759         12              6  \n",
       "top      Manager  10+ years       MORTGAGE  \n",
       "freq        1959      77726         126598  \n",
       "mean         NaN        NaN            NaN  \n",
       "std          NaN        NaN            NaN  \n",
       "min          NaN        NaN            NaN  \n",
       "25%          NaN        NaN            NaN  \n",
       "50%          NaN        NaN            NaN  \n",
       "75%          NaN        NaN            NaN  \n",
       "max          NaN        NaN            NaN  "
      ]
     },
     "execution_count": 181,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "loan_new.iloc[:, 0:12].describe(include='all')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 182,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>annual_inc</th>\n",
       "      <th>verification_status</th>\n",
       "      <th>issue_d</th>\n",
       "      <th>loan_status</th>\n",
       "      <th>purpose</th>\n",
       "      <th>zip_code</th>\n",
       "      <th>addr_state</th>\n",
       "      <th>dti</th>\n",
       "      <th>delinq_2yrs</th>\n",
       "      <th>earliest_cr_line</th>\n",
       "      <th>inq_last_6mths</th>\n",
       "      <th>open_acc</th>\n",
       "      <th>pub_rec</th>\n",
       "      <th>revol_bal</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>2.569350e+05</td>\n",
       "      <td>256939</td>\n",
       "      <td>256939</td>\n",
       "      <td>256939</td>\n",
       "      <td>256939</td>\n",
       "      <td>256939</td>\n",
       "      <td>256939</td>\n",
       "      <td>256939.000000</td>\n",
       "      <td>256910.000000</td>\n",
       "      <td>256910</td>\n",
       "      <td>256910.000000</td>\n",
       "      <td>256910.000000</td>\n",
       "      <td>256910.000000</td>\n",
       "      <td>2.569390e+05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>unique</th>\n",
       "      <td>NaN</td>\n",
       "      <td>3</td>\n",
       "      <td>103</td>\n",
       "      <td>5</td>\n",
       "      <td>14</td>\n",
       "      <td>889</td>\n",
       "      <td>51</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>646</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>top</th>\n",
       "      <td>NaN</td>\n",
       "      <td>Verified</td>\n",
       "      <td>Oct-2014</td>\n",
       "      <td>Fully Paid</td>\n",
       "      <td>debt_consolidation</td>\n",
       "      <td>945xx</td>\n",
       "      <td>CA</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Oct-2000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>freq</th>\n",
       "      <td>NaN</td>\n",
       "      <td>94002</td>\n",
       "      <td>8808</td>\n",
       "      <td>207723</td>\n",
       "      <td>150253</td>\n",
       "      <td>3316</td>\n",
       "      <td>43645</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2143</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>7.249885e+04</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>16.534986</td>\n",
       "      <td>0.250411</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.887821</td>\n",
       "      <td>10.935016</td>\n",
       "      <td>0.143354</td>\n",
       "      <td>1.530120e+04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>5.890043e+04</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>7.793541</td>\n",
       "      <td>0.742431</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.158745</td>\n",
       "      <td>4.902947</td>\n",
       "      <td>0.436027</td>\n",
       "      <td>1.970880e+04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>1.896000e+03</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>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000e+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>4.500000e+04</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>10.740000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>5.833000e+03</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>6.200000e+04</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>16.200000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.091800e+04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>8.700000e+04</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>21.990000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>14.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.908350e+04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>8.706582e+06</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>57.140000</td>\n",
       "      <td>29.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>33.000000</td>\n",
       "      <td>76.000000</td>\n",
       "      <td>15.000000</td>\n",
       "      <td>1.746716e+06</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          annual_inc verification_status   issue_d loan_status  \\\n",
       "count   2.569350e+05              256939    256939      256939   \n",
       "unique           NaN                   3       103           5   \n",
       "top              NaN            Verified  Oct-2014  Fully Paid   \n",
       "freq             NaN               94002      8808      207723   \n",
       "mean    7.249885e+04                 NaN       NaN         NaN   \n",
       "std     5.890043e+04                 NaN       NaN         NaN   \n",
       "min     1.896000e+03                 NaN       NaN         NaN   \n",
       "25%     4.500000e+04                 NaN       NaN         NaN   \n",
       "50%     6.200000e+04                 NaN       NaN         NaN   \n",
       "75%     8.700000e+04                 NaN       NaN         NaN   \n",
       "max     8.706582e+06                 NaN       NaN         NaN   \n",
       "\n",
       "                   purpose zip_code addr_state            dti    delinq_2yrs  \\\n",
       "count               256939   256939     256939  256939.000000  256910.000000   \n",
       "unique                  14      889         51            NaN            NaN   \n",
       "top     debt_consolidation    945xx         CA            NaN            NaN   \n",
       "freq                150253     3316      43645            NaN            NaN   \n",
       "mean                   NaN      NaN        NaN      16.534986       0.250411   \n",
       "std                    NaN      NaN        NaN       7.793541       0.742431   \n",
       "min                    NaN      NaN        NaN       0.000000       0.000000   \n",
       "25%                    NaN      NaN        NaN      10.740000       0.000000   \n",
       "50%                    NaN      NaN        NaN      16.200000       0.000000   \n",
       "75%                    NaN      NaN        NaN      21.990000       0.000000   \n",
       "max                    NaN      NaN        NaN      57.140000      29.000000   \n",
       "\n",
       "       earliest_cr_line  inq_last_6mths       open_acc        pub_rec  \\\n",
       "count            256910   256910.000000  256910.000000  256910.000000   \n",
       "unique              646             NaN            NaN            NaN   \n",
       "top            Oct-2000             NaN            NaN            NaN   \n",
       "freq               2143             NaN            NaN            NaN   \n",
       "mean                NaN        0.887821      10.935016       0.143354   \n",
       "std                 NaN        1.158745       4.902947       0.436027   \n",
       "min                 NaN        0.000000       0.000000       0.000000   \n",
       "25%                 NaN        0.000000       7.000000       0.000000   \n",
       "50%                 NaN        1.000000      10.000000       0.000000   \n",
       "75%                 NaN        1.000000      14.000000       0.000000   \n",
       "max                 NaN       33.000000      76.000000      15.000000   \n",
       "\n",
       "           revol_bal  \n",
       "count   2.569390e+05  \n",
       "unique           NaN  \n",
       "top              NaN  \n",
       "freq             NaN  \n",
       "mean    1.530120e+04  \n",
       "std     1.970880e+04  \n",
       "min     0.000000e+00  \n",
       "25%     5.833000e+03  \n",
       "50%     1.091800e+04  \n",
       "75%     1.908350e+04  \n",
       "max     1.746716e+06  "
      ]
     },
     "execution_count": 182,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "loan_new.iloc[:, 12:26].describe(include='all')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 183,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>revol_util</th>\n",
       "      <th>total_acc</th>\n",
       "      <th>initial_list_status</th>\n",
       "      <th>last_credit_pull_d</th>\n",
       "      <th>collections_12_mths_ex_med</th>\n",
       "      <th>policy_code</th>\n",
       "      <th>application_type</th>\n",
       "      <th>acc_now_delinq</th>\n",
       "      <th>tot_coll_amt</th>\n",
       "      <th>tot_cur_bal</th>\n",
       "      <th>total_rev_hi_lim</th>\n",
       "      <th>status</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>256699.000000</td>\n",
       "      <td>256910.000000</td>\n",
       "      <td>256939</td>\n",
       "      <td>256916</td>\n",
       "      <td>256794.000000</td>\n",
       "      <td>256939.0</td>\n",
       "      <td>256939</td>\n",
       "      <td>256910.000000</td>\n",
       "      <td>1.904640e+05</td>\n",
       "      <td>1.904640e+05</td>\n",
       "      <td>1.904640e+05</td>\n",
       "      <td>256939</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>unique</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "      <td>103</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>top</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>f</td>\n",
       "      <td>Jan-2016</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>INDIVIDUAL</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>paid</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>freq</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>185496</td>\n",
       "      <td>103973</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>256938</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>209711</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>54.315684</td>\n",
       "      <td>25.011732</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.006702</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.003106</td>\n",
       "      <td>2.033825e+02</td>\n",
       "      <td>1.381605e+05</td>\n",
       "      <td>2.969485e+04</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>24.827559</td>\n",
       "      <td>11.779200</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.088812</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.060409</td>\n",
       "      <td>2.103550e+04</td>\n",
       "      <td>1.523284e+05</td>\n",
       "      <td>2.949980e+04</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>36.200000</td>\n",
       "      <td>16.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>2.835575e+04</td>\n",
       "      <td>1.330000e+04</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>55.800000</td>\n",
       "      <td>23.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>8.076050e+04</td>\n",
       "      <td>2.230000e+04</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>73.900000</td>\n",
       "      <td>32.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>2.079905e+05</td>\n",
       "      <td>3.680000e+04</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>892.300000</td>\n",
       "      <td>150.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>9.152545e+06</td>\n",
       "      <td>8.000078e+06</td>\n",
       "      <td>2.013133e+06</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           revol_util      total_acc initial_list_status last_credit_pull_d  \\\n",
       "count   256699.000000  256910.000000              256939             256916   \n",
       "unique            NaN            NaN                   2                103   \n",
       "top               NaN            NaN                   f           Jan-2016   \n",
       "freq              NaN            NaN              185496             103973   \n",
       "mean        54.315684      25.011732                 NaN                NaN   \n",
       "std         24.827559      11.779200                 NaN                NaN   \n",
       "min          0.000000       1.000000                 NaN                NaN   \n",
       "25%         36.200000      16.000000                 NaN                NaN   \n",
       "50%         55.800000      23.000000                 NaN                NaN   \n",
       "75%         73.900000      32.000000                 NaN                NaN   \n",
       "max        892.300000     150.000000                 NaN                NaN   \n",
       "\n",
       "        collections_12_mths_ex_med  policy_code application_type  \\\n",
       "count                256794.000000     256939.0           256939   \n",
       "unique                         NaN          NaN                2   \n",
       "top                            NaN          NaN       INDIVIDUAL   \n",
       "freq                           NaN          NaN           256938   \n",
       "mean                      0.006702          1.0              NaN   \n",
       "std                       0.088812          0.0              NaN   \n",
       "min                       0.000000          1.0              NaN   \n",
       "25%                       0.000000          1.0              NaN   \n",
       "50%                       0.000000          1.0              NaN   \n",
       "75%                       0.000000          1.0              NaN   \n",
       "max                       6.000000          1.0              NaN   \n",
       "\n",
       "        acc_now_delinq  tot_coll_amt   tot_cur_bal  total_rev_hi_lim  status  \n",
       "count    256910.000000  1.904640e+05  1.904640e+05      1.904640e+05  256939  \n",
       "unique             NaN           NaN           NaN               NaN       2  \n",
       "top                NaN           NaN           NaN               NaN    paid  \n",
       "freq               NaN           NaN           NaN               NaN  209711  \n",
       "mean          0.003106  2.033825e+02  1.381605e+05      2.969485e+04     NaN  \n",
       "std           0.060409  2.103550e+04  1.523284e+05      2.949980e+04     NaN  \n",
       "min           0.000000  0.000000e+00  0.000000e+00      0.000000e+00     NaN  \n",
       "25%           0.000000  0.000000e+00  2.835575e+04      1.330000e+04     NaN  \n",
       "50%           0.000000  0.000000e+00  8.076050e+04      2.230000e+04     NaN  \n",
       "75%           0.000000  0.000000e+00  2.079905e+05      3.680000e+04     NaN  \n",
       "max           5.000000  9.152545e+06  8.000078e+06      2.013133e+06     NaN  "
      ]
     },
     "execution_count": 183,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "loan_new.iloc[:, 26:].describe(include='all')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 184,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(256939, 38)"
      ]
     },
     "execution_count": 184,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "loan_new.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "So we have removed columns with more than 100,000 missing values. We have also removed columns for extraneous reasons. Now of the columns left some comments:\n",
    "\n",
    "- Employee title has 242,744 values of which 134,759 are Manager\n",
    "- revol_util has only 256,699 values\n",
    "- last_credit_pull_d has only 256,916 values\n",
    "- collections_12_mths_ex_med might only have values equal to 6, but we will check"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here we see the number of unique employee titles:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 185,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "134760"
      ]
     },
     "execution_count": 185,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(list(loan_new['emp_title'].unique()))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To make any use of this we would have to map these to fewer catgories. Maybe a count by category will help."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 186,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "emp_title\n",
       "       NSA Industries llc                   1\n",
       "   Bus Operator and Instructor              1\n",
       "   Long Ilsand College Hospital             1\n",
       "  BUS OPERATER                              1\n",
       "  Golf Sevice Manager/ Caddie Master        1\n",
       "  old palm inc                              1\n",
       " (RED),  a division of The ONE Campaign     1\n",
       " A Bell Pest Services                       1\n",
       " A Red Thread Adoption Services, Inc.       1\n",
       " ABC Supply Company                         1\n",
       " ADM                                        1\n",
       " ADVANTAGE SEPTIC SERVICE                   1\n",
       " AFJROTC Teacher                            1\n",
       " Airway Transportation System Specialist    1\n",
       " Andover Sub-Acute &Rehab Center            1\n",
       " Area Sales Manager                         1\n",
       " Behavior Specialist                        1\n",
       " Best Buy                                   1\n",
       " Bice president fixed income sales          1\n",
       " Box Cleaning                               1\n",
       " Brocade Communications                     1\n",
       " BusAttendantdept.oftransportant            1\n",
       " Business Analyst                           1\n",
       " CASHIER                                    1\n",
       " CORRUGATOR SUPERVISOR                      1\n",
       " CRNA                                       1\n",
       " Care Manager                               1\n",
       " Cash management Team leader                1\n",
       " CenturyLink                                1\n",
       " Certified Dialysis Tech                    1\n",
       "                                           ..\n",
       "yrcw                                        1\n",
       "ysk corporation                             1\n",
       "yum brand                                   1\n",
       "yuma county general services                1\n",
       "yuma county road and bridge                 1\n",
       "z and s food and liquor                     1\n",
       "z/s contracting                             1\n",
       "zachry                                      1\n",
       "zachry industral inc                        1\n",
       "zashko inc.                                 1\n",
       "zaxbys                                      1\n",
       "zeck ford                                   1\n",
       "zeeco                                       1\n",
       "zeffert & associates                        1\n",
       "zellena inc. (pushkin Academy)              1\n",
       "zeller & wieliczko                          1\n",
       "zeno office solutions                       1\n",
       "zevuloni & associates                       1\n",
       "zia concrete supply                         1\n",
       "zifko tire and battery inc                  1\n",
       "zilkha residence                            1\n",
       "zimmerman equipment                         1\n",
       "zion lutheran school                        1\n",
       "zms fashion                                 1\n",
       "zoll medical corp                           1\n",
       "zone manager                                3\n",
       "zookeeper                                   1\n",
       "zozaya officiating                          1\n",
       "zs backroom                                 1\n",
       "zueck transportation                        1\n",
       "dtype: int64"
      ]
     },
     "execution_count": 186,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "loan_new.groupby('emp_title').size()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Nah, this is a task too large for the project.\n",
    "\n",
    "Now let's take a look at 'collections_12_mths_ex_med'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 187,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "collections_12_mths_ex_med\n",
       "0.0    255201\n",
       "1.0      1484\n",
       "2.0        97\n",
       "3.0         8\n",
       "4.0         2\n",
       "5.0         1\n",
       "6.0         1\n",
       "dtype: int64"
      ]
     },
     "execution_count": 187,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "loan_new.groupby('collections_12_mths_ex_med').size()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 188,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "purpose\n",
       "car                     3720\n",
       "credit_card            50649\n",
       "debt_consolidation    150253\n",
       "educational              422\n",
       "home_improvement       15190\n",
       "house                   1703\n",
       "major_purchase          6402\n",
       "medical                 2927\n",
       "moving                  2085\n",
       "other                  14766\n",
       "renewable_energy         270\n",
       "small_business          4926\n",
       "vacation                1615\n",
       "wedding                 2011\n",
       "dtype: int64"
      ]
     },
     "execution_count": 188,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "loan_new.groupby('purpose').size()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Feature Engineering\n",
    "\n",
    "First, which features are left?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 189,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['id', 'member_id', 'loan_amnt', 'funded_amnt', 'funded_amnt_inv',\n",
       "       'int_rate', 'installment', 'grade', 'sub_grade', 'emp_title',\n",
       "       'emp_length', 'home_ownership', 'annual_inc', 'verification_status',\n",
       "       'issue_d', 'loan_status', 'purpose', 'zip_code', 'addr_state',\n",
       "       'dti', 'delinq_2yrs', 'earliest_cr_line', 'inq_last_6mths',\n",
       "       'open_acc', 'pub_rec', 'revol_bal', 'revol_util', 'total_acc',\n",
       "       'initial_list_status', 'last_credit_pull_d',\n",
       "       'collections_12_mths_ex_med', 'policy_code', 'application_type',\n",
       "       'acc_now_delinq', 'tot_coll_amt', 'tot_cur_bal', 'total_rev_hi_lim',\n",
       "       'status'], dtype=object)"
      ]
     },
     "execution_count": 189,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "loan_new.columns.values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 192,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/anaconda/lib/python2.7/site-packages/ipykernel/__main__.py:2: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy\n",
      "  from ipykernel import kernelapp as app\n"
     ]
    }
   ],
   "source": [
    "# Let's add loan to income ratio\n",
    "loan_new['loan_income_ratio'] = loan_new['loan_amnt'] / loan_new['annual_inc']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 193,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/anaconda/lib/python2.7/site-packages/ipykernel/__main__.py:2: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy\n",
      "  from ipykernel import kernelapp as app\n"
     ]
    }
   ],
   "source": [
    "# Now add installment to monthly payment ratio\n",
    "loan_new['loan_income_ratio_month'] = loan_new['installment'] / (loan_new['annual_inc'] / 12)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4 Create training set and test set\n",
    "\n",
    "### 4.1 Split Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 194,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "response = loan_new['status']\n",
    "data = loan_new.drop(['status','loan_status'], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 195,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(data, response, test_size=0.2, random_state=0)\n",
    "train = pd.concat([X_train, y_train], axis=1)\n",
    "test = pd.concat([X_test, y_test], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 196,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\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>int_rate</th>\n",
       "      <th>installment</th>\n",
       "      <th>annual_inc</th>\n",
       "      <th>dti</th>\n",
       "      <th>delinq_2yrs</th>\n",
       "      <th>...</th>\n",
       "      <th>revol_bal</th>\n",
       "      <th>revol_util</th>\n",
       "      <th>total_acc</th>\n",
       "      <th>policy_code</th>\n",
       "      <th>acc_now_delinq</th>\n",
       "      <th>tot_coll_amt</th>\n",
       "      <th>tot_cur_bal</th>\n",
       "      <th>total_rev_hi_lim</th>\n",
       "      <th>loan_income_ratio</th>\n",
       "      <th>loan_income_ratio_month</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>collections_12_mths_ex_med</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0.0</th>\n",
       "      <td>12149083</td>\n",
       "      <td>13452089</td>\n",
       "      <td>13527.566702</td>\n",
       "      <td>13485.327742</td>\n",
       "      <td>13367.779659</td>\n",
       "      <td>13.770927</td>\n",
       "      <td>417.052085</td>\n",
       "      <td>72481.415444</td>\n",
       "      <td>16.536236</td>\n",
       "      <td>0.247664</td>\n",
       "      <td>...</td>\n",
       "      <td>15322.742419</td>\n",
       "      <td>54.367644</td>\n",
       "      <td>24.997915</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.003002</td>\n",
       "      <td>194.643884</td>\n",
       "      <td>138160.306045</td>\n",
       "      <td>29726.817722</td>\n",
       "      <td>0.208772</td>\n",
       "      <td>0.078019</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1.0</th>\n",
       "      <td>23253755</td>\n",
       "      <td>25407085</td>\n",
       "      <td>13371.276954</td>\n",
       "      <td>13371.276954</td>\n",
       "      <td>13365.397574</td>\n",
       "      <td>15.520640</td>\n",
       "      <td>418.177116</td>\n",
       "      <td>75018.495249</td>\n",
       "      <td>16.961489</td>\n",
       "      <td>0.657682</td>\n",
       "      <td>...</td>\n",
       "      <td>11987.352426</td>\n",
       "      <td>47.407345</td>\n",
       "      <td>27.603100</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.017520</td>\n",
       "      <td>1218.547170</td>\n",
       "      <td>136823.461590</td>\n",
       "      <td>25330.550539</td>\n",
       "      <td>0.199165</td>\n",
       "      <td>0.075343</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2.0</th>\n",
       "      <td>27281592</td>\n",
       "      <td>29627403</td>\n",
       "      <td>13012.628866</td>\n",
       "      <td>13012.628866</td>\n",
       "      <td>13009.536082</td>\n",
       "      <td>15.137835</td>\n",
       "      <td>406.310103</td>\n",
       "      <td>79253.300206</td>\n",
       "      <td>17.748454</td>\n",
       "      <td>0.979381</td>\n",
       "      <td>...</td>\n",
       "      <td>11415.701031</td>\n",
       "      <td>37.386598</td>\n",
       "      <td>30.412371</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.041237</td>\n",
       "      <td>1529.567010</td>\n",
       "      <td>155542.659794</td>\n",
       "      <td>32595.783505</td>\n",
       "      <td>0.185528</td>\n",
       "      <td>0.069663</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3.0</th>\n",
       "      <td>26221576</td>\n",
       "      <td>28711251</td>\n",
       "      <td>8512.500000</td>\n",
       "      <td>8512.500000</td>\n",
       "      <td>8493.750000</td>\n",
       "      <td>12.592500</td>\n",
       "      <td>269.581250</td>\n",
       "      <td>52200.000000</td>\n",
       "      <td>14.640000</td>\n",
       "      <td>1.125000</td>\n",
       "      <td>...</td>\n",
       "      <td>10379.875000</td>\n",
       "      <td>42.562500</td>\n",
       "      <td>27.375000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1207.250000</td>\n",
       "      <td>166917.250000</td>\n",
       "      <td>22796.875000</td>\n",
       "      <td>0.164286</td>\n",
       "      <td>0.062963</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4.0</th>\n",
       "      <td>33441592</td>\n",
       "      <td>36089893</td>\n",
       "      <td>13275.000000</td>\n",
       "      <td>13275.000000</td>\n",
       "      <td>13275.000000</td>\n",
       "      <td>16.690000</td>\n",
       "      <td>397.185000</td>\n",
       "      <td>68794.500000</td>\n",
       "      <td>19.165000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>14280.000000</td>\n",
       "      <td>46.500000</td>\n",
       "      <td>28.500000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>408.500000</td>\n",
       "      <td>33306.500000</td>\n",
       "      <td>44400.000000</td>\n",
       "      <td>0.241091</td>\n",
       "      <td>0.082969</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5.0</th>\n",
       "      <td>16281089</td>\n",
       "      <td>18383643</td>\n",
       "      <td>20000.000000</td>\n",
       "      <td>20000.000000</td>\n",
       "      <td>20000.000000</td>\n",
       "      <td>10.150000</td>\n",
       "      <td>646.760000</td>\n",
       "      <td>205000.000000</td>\n",
       "      <td>13.260000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>181363.000000</td>\n",
       "      <td>39.300000</td>\n",
       "      <td>30.000000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>574.000000</td>\n",
       "      <td>541960.000000</td>\n",
       "      <td>208200.000000</td>\n",
       "      <td>0.097561</td>\n",
       "      <td>0.037859</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6.0</th>\n",
       "      <td>33551290</td>\n",
       "      <td>36194558</td>\n",
       "      <td>20000.000000</td>\n",
       "      <td>20000.000000</td>\n",
       "      <td>19975.000000</td>\n",
       "      <td>14.310000</td>\n",
       "      <td>686.570000</td>\n",
       "      <td>54000.000000</td>\n",
       "      <td>35.440000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>14151.000000</td>\n",
       "      <td>41.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>6726.000000</td>\n",
       "      <td>40609.000000</td>\n",
       "      <td>34500.000000</td>\n",
       "      <td>0.370370</td>\n",
       "      <td>0.152571</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>7 rows × 23 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                  id  member_id     loan_amnt   funded_amnt  \\\n",
       "collections_12_mths_ex_med                                                    \n",
       "0.0                         12149083   13452089  13527.566702  13485.327742   \n",
       "1.0                         23253755   25407085  13371.276954  13371.276954   \n",
       "2.0                         27281592   29627403  13012.628866  13012.628866   \n",
       "3.0                         26221576   28711251   8512.500000   8512.500000   \n",
       "4.0                         33441592   36089893  13275.000000  13275.000000   \n",
       "5.0                         16281089   18383643  20000.000000  20000.000000   \n",
       "6.0                         33551290   36194558  20000.000000  20000.000000   \n",
       "\n",
       "                            funded_amnt_inv   int_rate  installment  \\\n",
       "collections_12_mths_ex_med                                            \n",
       "0.0                            13367.779659  13.770927   417.052085   \n",
       "1.0                            13365.397574  15.520640   418.177116   \n",
       "2.0                            13009.536082  15.137835   406.310103   \n",
       "3.0                             8493.750000  12.592500   269.581250   \n",
       "4.0                            13275.000000  16.690000   397.185000   \n",
       "5.0                            20000.000000  10.150000   646.760000   \n",
       "6.0                            19975.000000  14.310000   686.570000   \n",
       "\n",
       "                               annual_inc        dti  delinq_2yrs  \\\n",
       "collections_12_mths_ex_med                                          \n",
       "0.0                          72481.415444  16.536236     0.247664   \n",
       "1.0                          75018.495249  16.961489     0.657682   \n",
       "2.0                          79253.300206  17.748454     0.979381   \n",
       "3.0                          52200.000000  14.640000     1.125000   \n",
       "4.0                          68794.500000  19.165000     2.000000   \n",
       "5.0                         205000.000000  13.260000     2.000000   \n",
       "6.0                          54000.000000  35.440000     0.000000   \n",
       "\n",
       "                                     ...                 revol_bal  \\\n",
       "collections_12_mths_ex_med           ...                             \n",
       "0.0                                  ...              15322.742419   \n",
       "1.0                                  ...              11987.352426   \n",
       "2.0                                  ...              11415.701031   \n",
       "3.0                                  ...              10379.875000   \n",
       "4.0                                  ...              14280.000000   \n",
       "5.0                                  ...             181363.000000   \n",
       "6.0                                  ...              14151.000000   \n",
       "\n",
       "                            revol_util  total_acc  policy_code  \\\n",
       "collections_12_mths_ex_med                                       \n",
       "0.0                          54.367644  24.997915          1.0   \n",
       "1.0                          47.407345  27.603100          1.0   \n",
       "2.0                          37.386598  30.412371          1.0   \n",
       "3.0                          42.562500  27.375000          1.0   \n",
       "4.0                          46.500000  28.500000          1.0   \n",
       "5.0                          39.300000  30.000000          1.0   \n",
       "6.0                          41.000000  18.000000          1.0   \n",
       "\n",
       "                            acc_now_delinq  tot_coll_amt    tot_cur_bal  \\\n",
       "collections_12_mths_ex_med                                                \n",
       "0.0                               0.003002    194.643884  138160.306045   \n",
       "1.0                               0.017520   1218.547170  136823.461590   \n",
       "2.0                               0.041237   1529.567010  155542.659794   \n",
       "3.0                               0.000000   1207.250000  166917.250000   \n",
       "4.0                               0.000000    408.500000   33306.500000   \n",
       "5.0                               0.000000    574.000000  541960.000000   \n",
       "6.0                               0.000000   6726.000000   40609.000000   \n",
       "\n",
       "                            total_rev_hi_lim  loan_income_ratio  \\\n",
       "collections_12_mths_ex_med                                        \n",
       "0.0                             29726.817722           0.208772   \n",
       "1.0                             25330.550539           0.199165   \n",
       "2.0                             32595.783505           0.185528   \n",
       "3.0                             22796.875000           0.164286   \n",
       "4.0                             44400.000000           0.241091   \n",
       "5.0                            208200.000000           0.097561   \n",
       "6.0                             34500.000000           0.370370   \n",
       "\n",
       "                            loan_income_ratio_month  \n",
       "collections_12_mths_ex_med                           \n",
       "0.0                                        0.078019  \n",
       "1.0                                        0.075343  \n",
       "2.0                                        0.069663  \n",
       "3.0                                        0.062963  \n",
       "4.0                                        0.082969  \n",
       "5.0                                        0.037859  \n",
       "6.0                                        0.152571  \n",
       "\n",
       "[7 rows x 23 columns]"
      ]
     },
     "execution_count": 196,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "loan_new.groupby('collections_12_mths_ex_med').mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 197,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "paid      167863\n",
       "unpaid     37688\n",
       "Name: status, dtype: int64"
      ]
     },
     "execution_count": 197,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# loan_status is the response variable\n",
    "statusByType_train = train['status'].value_counts()\n",
    "statusByType_train"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 4.2 Impute Missing Values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 4.3 Write to Files"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 216,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "train.to_csv('loan_train.csv')\n",
    "test.to_csv('loan_test.csv')"
   ]
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python [default]",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}