modeling-expected-loss-time-to-default.ipynb

{
 "cells": [
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": [
    "#  Load Libraries"
   ]
  },
  {
   "metadata": {
    "pycharm": {
     "is_executing": false,
     "name": "#%%\n"
    },
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "# This Python 3 environment comes with many helpful analytics libraries installed\n",
    "# It is defined by the kaggle/python docker image: https://github.com/kaggle/docker-python\n",
    "# For example, here's several helpful packages to load in \n",
    "\n",
    "import numpy as np # linear algebra\n",
    "import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.dummy import DummyClassifier\n",
    "from sklearn.metrics import accuracy_score, f1_score, confusion_matrix, recall_score\n",
    "# Input data files are available in the \"../input/\" directory.\n",
    "# For example, running this (by clicking run or pressing Shift+Enter) will list all files under the input directory\n",
    "\n",
    "import os\n",
    "for dirname, _, filenames in os.walk('/kaggle/input'):\n",
    "    for filename in filenames:\n",
    "        print(os.path.join(dirname, filename))\n",
    "\n",
    "# Any results you write to the current directory are saved as output."
   ],
   "execution_count": 1,
   "outputs": [
    {
     "output_type": "stream",
     "text": "/kaggle/input/lending-club-loan-data/database.sqlite\n/kaggle/input/lending-club-loan-data/LCDataDictionary.xlsx\n/kaggle/input/lending-club-loan-data/loan.csv\n",
     "name": "stdout"
    }
   ]
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": [
    "# Notebook Goals\n",
    "The goal of this notebook is the use the Five C's of credit framework (Character, Capacity, Cash Flow, Collateral, and Condition) to select the proxy features from the Kaggle Lending Club Dataset to predict loan default probability.\n",
    "\n",
    "Using the predicted loan default probability, the notebook will the generate expected loss on loans within the portfolio.\n",
    "\n",
    "Expected Loss (EL) on a given is calculuated using the loan amount multiplied by a loan's probability of default (PD) mulitplied by the the loan's loss given default (LGD). Loss Given Default is the amount of the loan that is deemed unrecoverable and is represented as a number between 0 (no loss) and 1 (100% of the loan is a loss)\n",
    "\n",
    "The equation for EL is given as:\n",
    "\n",
    "** EL = PD x LGD x Loan Amount*\n",
    "\n",
    "After selected features using the 5 C's of Credit Framework and adjusting for the skewedness in the data (there is a greater proportion of loans in the portfolio that have NOT defaulted versus loans that HAVE defaulted) we will model proabilities of default using classification models. \n",
    "\n",
    "With proabilties of default we will then model Expected Loss using an assumed LGD of 1 (a LGD of 1 is a conservative assumption given the unsecured nature of the loans) "
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "raw_data = pd.read_csv(\"/kaggle/input/lending-club-loan-data/loan.csv\",low_memory=False)\n",
    "raw_data.head()"
   ],
   "execution_count": 2,
   "outputs": [
    {
     "output_type": "execute_result",
     "execution_count": 2,
     "data": {
      "text/plain": "   id  member_id  loan_amnt  funded_amnt  funded_amnt_inv        term  \\\n0 NaN        NaN       2500         2500           2500.0   36 months   \n1 NaN        NaN      30000        30000          30000.0   60 months   \n2 NaN        NaN       5000         5000           5000.0   36 months   \n3 NaN        NaN       4000         4000           4000.0   36 months   \n4 NaN        NaN      30000        30000          30000.0   60 months   \n\n   int_rate  installment grade sub_grade  ... hardship_payoff_balance_amount  \\\n0     13.56        84.92     C        C1  ...                            NaN   \n1     18.94       777.23     D        D2  ...                            NaN   \n2     17.97       180.69     D        D1  ...                            NaN   \n3     18.94       146.51     D        D2  ...                            NaN   \n4     16.14       731.78     C        C4  ...                            NaN   \n\n  hardship_last_payment_amount disbursement_method  debt_settlement_flag  \\\n0                          NaN                Cash                     N   \n1                          NaN                Cash                     N   \n2                          NaN                Cash                     N   \n3                          NaN                Cash                     N   \n4                          NaN                Cash                     N   \n\n  debt_settlement_flag_date settlement_status settlement_date  \\\n0                       NaN               NaN             NaN   \n1                       NaN               NaN             NaN   \n2                       NaN               NaN             NaN   \n3                       NaN               NaN             NaN   \n4                       NaN               NaN             NaN   \n\n  settlement_amount  settlement_percentage settlement_term  \n0               NaN                    NaN             NaN  \n1               NaN                    NaN             NaN  \n2               NaN                    NaN             NaN  \n3               NaN                    NaN             NaN  \n4               NaN                    NaN             NaN  \n\n[5 rows x 145 columns]",
      "text/html": "<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>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>2500</td>\n      <td>2500</td>\n      <td>2500.0</td>\n      <td>36 months</td>\n      <td>13.56</td>\n      <td>84.92</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>1</th>\n      <td>NaN</td>\n      <td>NaN</td>\n      <td>30000</td>\n      <td>30000</td>\n      <td>30000.0</td>\n      <td>60 months</td>\n      <td>18.94</td>\n      <td>777.23</td>\n      <td>D</td>\n      <td>D2</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>5000</td>\n      <td>5000</td>\n      <td>5000.0</td>\n      <td>36 months</td>\n      <td>17.97</td>\n      <td>180.69</td>\n      <td>D</td>\n      <td>D1</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>4000</td>\n      <td>4000</td>\n      <td>4000.0</td>\n      <td>36 months</td>\n      <td>18.94</td>\n      <td>146.51</td>\n      <td>D</td>\n      <td>D2</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>30000</td>\n      <td>30000</td>\n      <td>30000.0</td>\n      <td>60 months</td>\n      <td>16.14</td>\n      <td>731.78</td>\n      <td>C</td>\n      <td>C4</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>"
     },
     "metadata": {}
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "preprocess_df = raw_data[['emp_length', 'loan_status', 'home_ownership', 'issue_d',\n",
    "                          'earliest_cr_line', 'purpose', 'term', 'annual_inc', 'dti',\n",
    "                          'loan_amnt', 'int_rate', 'pub_rec_bankruptcies',]].copy()"
   ],
   "execution_count": 3,
   "outputs": []
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "preprocess_df = raw_data[['emp_length', 'loan_status', 'home_ownership', 'issue_d',\n",
    "                          'earliest_cr_line', 'purpose', 'term', 'annual_inc', 'dti',\n",
    "                          'loan_amnt', 'int_rate', 'pub_rec_bankruptcies',]].copy()\n",
    "# Create a list of columns that are NOT numeric values\n",
    "not_numeric_cols = ['emp_length', 'loan_status', 'home_ownership', 'issue_d',\n",
    "                    'earliest_cr_line', 'purpose', 'term']\n",
    "\n",
    "# Create list of columns that ARE numeric values and print\n",
    "numeric_cols = [col for col in preprocess_df.columns if col not in not_numeric_cols]\n",
    "print(numeric_cols)\n",
    "\n",
    "# Convert numeric cols into numeric data types\n",
    "preprocess_df[numeric_cols] = preprocess_df[numeric_cols].apply(pd.to_numeric)"
   ],
   "execution_count": 4,
   "outputs": [
    {
     "output_type": "stream",
     "text": "['annual_inc', 'dti', 'loan_amnt', 'int_rate', 'pub_rec_bankruptcies']\n",
     "name": "stdout"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "# Create list of datetime columns\n",
    "datetime_cols = ['earliest_cr_line', 'issue_d']\n",
    "\n",
    "# Convert to datetime\n",
    "preprocess_df[datetime_cols] = preprocess_df[datetime_cols].apply(pd.to_datetime)"
   ],
   "execution_count": 5,
   "outputs": []
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": [
    "# Missing Data / Missing Observations"
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "preprocess_df.isnull().sum()"
   ],
   "execution_count": 6,
   "outputs": [
    {
     "output_type": "execute_result",
     "execution_count": 6,
     "data": {
      "text/plain": "emp_length              146907\nloan_status                  0\nhome_ownership               0\nissue_d                      0\nearliest_cr_line            29\npurpose                      0\nterm                         0\nannual_inc                   4\ndti                       1711\nloan_amnt                    0\nint_rate                     0\npub_rec_bankruptcies      1365\ndtype: int64"
     },
     "metadata": {}
    }
   ]
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": [
    "# Character: pub_rec_bankruptcies, earliest_cr_line, issue_d\n",
    "What it is: A lender’s opinion of a borrower’s general trustworthiness, credibility and personality. Not only the willingness of the oblgior to repay an obligation, and live up to its terms, but also honesty relative to the risk of fraud. \n",
    "\n",
    "Why it matters: Banks want to lend to people who are responsible and keep commitments.\n",
    "\n",
    "How it’s assessed: From your work experience, credit history, credentials, references, reputation and interaction with lenders."
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "character_df = preprocess_df[['pub_rec_bankruptcies', 'earliest_cr_line', 'issue_d']].copy()"
   ],
   "execution_count": 7,
   "outputs": []
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "print(character_df.head())\n",
    "print(character_df.describe())\n",
    "print(character_df.dtypes)"
   ],
   "execution_count": 8,
   "outputs": [
    {
     "output_type": "stream",
     "text": "   pub_rec_bankruptcies earliest_cr_line    issue_d\n0                   1.0       2001-04-01 2018-12-01\n1                   1.0       1987-06-01 2018-12-01\n2                   0.0       2011-04-01 2018-12-01\n3                   0.0       2006-02-01 2018-12-01\n4                   0.0       2000-12-01 2018-12-01\n       pub_rec_bankruptcies\ncount          2.259303e+06\nmean           1.281935e-01\nstd            3.646130e-01\nmin            0.000000e+00\n25%            0.000000e+00\n50%            0.000000e+00\n75%            0.000000e+00\nmax            1.200000e+01\npub_rec_bankruptcies           float64\nearliest_cr_line        datetime64[ns]\nissue_d                 datetime64[ns]\ndtype: object\n",
     "name": "stdout"
    }
   ]
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": [
    "<code>'earliest_cr_line'</code> shows the first credit line on the customer credit report. \n",
    "\n",
    "Will create a new feature that shows the number of months between the first credit line in the customers history and the issue date of the loan."
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "# fill the missing value for earliest_cr_line with most frequently occuring\n",
    "character_df['earliest_cr_line'].fillna(character_df['earliest_cr_line'].value_counts().index[0], inplace=True)"
   ],
   "execution_count": 9,
   "outputs": []
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "# count months between now and 'earliest_cr_line'\n",
    "character_df['credit_hist_in_months'] = ((character_df['issue_d'] - character_df['earliest_cr_line'])/np.timedelta64(1, 'M'))\n",
    "character_df['credit_hist_in_months'] = character_df['credit_hist_in_months'].astype(int)\n",
    "\n",
    "character_df.head()"
   ],
   "execution_count": 10,
   "outputs": [
    {
     "output_type": "execute_result",
     "execution_count": 10,
     "data": {
      "text/plain": "   pub_rec_bankruptcies earliest_cr_line    issue_d  credit_hist_in_months\n0                   1.0       2001-04-01 2018-12-01                    212\n1                   1.0       1987-06-01 2018-12-01                    378\n2                   0.0       2011-04-01 2018-12-01                     92\n3                   0.0       2006-02-01 2018-12-01                    153\n4                   0.0       2000-12-01 2018-12-01                    215",
      "text/html": "<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>pub_rec_bankruptcies</th>\n      <th>earliest_cr_line</th>\n      <th>issue_d</th>\n      <th>credit_hist_in_months</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>1.0</td>\n      <td>2001-04-01</td>\n      <td>2018-12-01</td>\n      <td>212</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>1.0</td>\n      <td>1987-06-01</td>\n      <td>2018-12-01</td>\n      <td>378</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>0.0</td>\n      <td>2011-04-01</td>\n      <td>2018-12-01</td>\n      <td>92</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>0.0</td>\n      <td>2006-02-01</td>\n      <td>2018-12-01</td>\n      <td>153</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>0.0</td>\n      <td>2000-12-01</td>\n      <td>2018-12-01</td>\n      <td>215</td>\n    </tr>\n  </tbody>\n</table>\n</div>"
     },
     "metadata": {}
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "# Create a new binary feature of whether or not there is a bankruptcy on file in customers credit history\n",
    "character_df['cb_person_bk_on_file_Y'] = character_df['pub_rec_bankruptcies'].apply(lambda x: 1 if x >= 1 else 0)\n",
    "character_df.head()"
   ],
   "execution_count": 11,
   "outputs": [
    {
     "output_type": "execute_result",
     "execution_count": 11,
     "data": {
      "text/plain": "   pub_rec_bankruptcies earliest_cr_line    issue_d  credit_hist_in_months  \\\n0                   1.0       2001-04-01 2018-12-01                    212   \n1                   1.0       1987-06-01 2018-12-01                    378   \n2                   0.0       2011-04-01 2018-12-01                     92   \n3                   0.0       2006-02-01 2018-12-01                    153   \n4                   0.0       2000-12-01 2018-12-01                    215   \n\n   cb_person_bk_on_file_Y  \n0                       1  \n1                       1  \n2                       0  \n3                       0  \n4                       0  ",
      "text/html": "<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>pub_rec_bankruptcies</th>\n      <th>earliest_cr_line</th>\n      <th>issue_d</th>\n      <th>credit_hist_in_months</th>\n      <th>cb_person_bk_on_file_Y</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>1.0</td>\n      <td>2001-04-01</td>\n      <td>2018-12-01</td>\n      <td>212</td>\n      <td>1</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>1.0</td>\n      <td>1987-06-01</td>\n      <td>2018-12-01</td>\n      <td>378</td>\n      <td>1</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>0.0</td>\n      <td>2011-04-01</td>\n      <td>2018-12-01</td>\n      <td>92</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>0.0</td>\n      <td>2006-02-01</td>\n      <td>2018-12-01</td>\n      <td>153</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>0.0</td>\n      <td>2000-12-01</td>\n      <td>2018-12-01</td>\n      <td>215</td>\n      <td>0</td>\n    </tr>\n  </tbody>\n</table>\n</div>"
     },
     "metadata": {}
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "# drop the old features from the character_df\n",
    "character_df.drop(['pub_rec_bankruptcies', 'earliest_cr_line', 'issue_d'], axis=1, inplace=True)\n",
    "character_df.head()"
   ],
   "execution_count": 12,
   "outputs": [
    {
     "output_type": "execute_result",
     "execution_count": 12,
     "data": {
      "text/plain": "   credit_hist_in_months  cb_person_bk_on_file_Y\n0                    212                       1\n1                    378                       1\n2                     92                       0\n3                    153                       0\n4                    215                       0",
      "text/html": "<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>credit_hist_in_months</th>\n      <th>cb_person_bk_on_file_Y</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>212</td>\n      <td>1</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>378</td>\n      <td>1</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>92</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>153</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>215</td>\n      <td>0</td>\n    </tr>\n  </tbody>\n</table>\n</div>"
     },
     "metadata": {}
    }
   ]
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": [
    "# Capacity/Cash flow: annual_inc & dti\n",
    "What it is: Your ability to repay the loan. The ability to generate cash to repay all oblgiations, when due. Assessing this is the primary goal of credit analysis\n",
    "\n",
    "Why it matters: Lenders want to be assured that your business generates enough cash flow to repay the loan in full."
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "capacity_df = preprocess_df[['annual_inc', 'dti']].copy()\n",
    "capacity_df.head()"
   ],
   "execution_count": 13,
   "outputs": [
    {
     "output_type": "execute_result",
     "execution_count": 13,
     "data": {
      "text/plain": "   annual_inc    dti\n0     55000.0  18.24\n1     90000.0  26.52\n2     59280.0  10.51\n3     92000.0  16.74\n4     57250.0  26.35",
      "text/html": "<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>annual_inc</th>\n      <th>dti</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>55000.0</td>\n      <td>18.24</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>90000.0</td>\n      <td>26.52</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>59280.0</td>\n      <td>10.51</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>92000.0</td>\n      <td>16.74</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>57250.0</td>\n      <td>26.35</td>\n    </tr>\n  </tbody>\n</table>\n</div>"
     },
     "metadata": {}
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "# fill missing values for annual income with the mean\n",
    "capacity_df['annual_inc'] = capacity_df['annual_inc'].fillna(capacity_df['annual_inc'].mean())\n",
    "\n",
    "# fill missing values for dti with the mean\n",
    "capacity_df['dti'] = capacity_df['dti'].fillna(capacity_df['dti'].mean())"
   ],
   "execution_count": 14,
   "outputs": []
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "# describe the capacity/cash flow proxy features\n",
    "print(capacity_df[['annual_inc', 'dti']].describe())"
   ],
   "execution_count": 15,
   "outputs": [
    {
     "output_type": "stream",
     "text": "         annual_inc           dti\ncount  2.260668e+06  2.260668e+06\nmean   7.799243e+04  1.882420e+01\nstd    1.126961e+05  1.417796e+01\nmin    0.000000e+00 -1.000000e+00\n25%    4.600000e+04  1.190000e+01\n50%    6.500000e+04  1.785000e+01\n75%    9.300000e+04  2.448000e+01\nmax    1.100000e+08  9.990000e+02\n",
     "name": "stdout"
    }
   ]
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": [
    "# Capital\n",
    "What it is: The amount of money invested by the business owner or management team. It is required during periods of weak cash flow generation for an obligor to sustain itself and meeet obligations. If there isn't sufficient cash flow to meet obligations, then assets may need to be sold to meet obligations.\n",
    "\n",
    "Why it matters: Banks are more willing to lend to owners who have invested some of their own money into the venture. It shows you have some “skin in the game.”\n",
    "\n",
    "How it’s assessed: From the amount of money the borrower or management team has invested in the business."
   ]
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": [
    "# <i>NO PROXY FEATURES USED FOR CAPITAL<i>"
   ]
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": [
    "# Conditions : loan_amnt, int_rate, term\n",
    "What it is: The condition of a business/individual — whether it is growing or faltering — as well as what you’ll use the funds for. Refers to the macro-economic and competitive environment which will impact an obligor's future performance and its ability to generate cash. \n",
    "\n",
    "Why it matters: To ensure that loans are repaid, banks want to lend to businesses operating under favorable conditions. They aim to identify risks and protect themselves accordingly.\n",
    "\n",
    "How it’s assessed: From a review of the competitive landscape, supplier and customer relationships, and macroeconomic and industry-specific issues."
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "conditions_df = preprocess_df[['loan_amnt', 'int_rate', 'term']].copy()"
   ],
   "execution_count": 16,
   "outputs": []
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "conditions_df.term.value_counts()"
   ],
   "execution_count": 17,
   "outputs": [
    {
     "output_type": "execute_result",
     "execution_count": 17,
     "data": {
      "text/plain": " 36 months    1609754\n 60 months     650914\nName: term, dtype: int64"
     },
     "metadata": {}
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "# Convert values of term to 0, 1 where 0 = 36 months and 1 = 60 months\n",
    "conditions_df['term'] = conditions_df['term'].replace({' 36 months': '0',\n",
    "                                                       ' 60 months': '1'})\n",
    "\n",
    "# convert term into an integer data type \n",
    "conditions_df['term'] = conditions_df['term'].astype(int)\n",
    "\n",
    "# Rename term column\n",
    "conditions_df = conditions_df.rename(columns={'term': 'term_60'})\n",
    "\n",
    "print(conditions_df.term_60.value_counts())\n",
    "conditions_df.head()"
   ],
   "execution_count": 18,
   "outputs": [
    {
     "output_type": "stream",
     "text": "0    1609754\n1     650914\nName: term_60, dtype: int64\n",
     "name": "stdout"
    },
    {
     "output_type": "execute_result",
     "execution_count": 18,
     "data": {
      "text/plain": "   loan_amnt  int_rate  term_60\n0       2500     13.56        0\n1      30000     18.94        1\n2       5000     17.97        0\n3       4000     18.94        0\n4      30000     16.14        1",
      "text/html": "<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>loan_amnt</th>\n      <th>int_rate</th>\n      <th>term_60</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>2500</td>\n      <td>13.56</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>30000</td>\n      <td>18.94</td>\n      <td>1</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>5000</td>\n      <td>17.97</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>4000</td>\n      <td>18.94</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>30000</td>\n      <td>16.14</td>\n      <td>1</td>\n    </tr>\n  </tbody>\n</table>\n</div>"
     },
     "metadata": {}
    }
   ]
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": [
    "# Collateral: home_ownership\n",
    "What it is: Assets that are used to guarantee or secure a loan.\n",
    "\n",
    "Why it matters: Collateral is a backup source if the borrower cannot repay a loan."
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "collateral_df = preprocess_df[['home_ownership']].copy()\n",
    "collateral_df.head()"
   ],
   "execution_count": 19,
   "outputs": [
    {
     "output_type": "execute_result",
     "execution_count": 19,
     "data": {
      "text/plain": "  home_ownership\n0           RENT\n1       MORTGAGE\n2       MORTGAGE\n3       MORTGAGE\n4       MORTGAGE",
      "text/html": "<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>home_ownership</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>RENT</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>MORTGAGE</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>MORTGAGE</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>MORTGAGE</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>MORTGAGE</td>\n    </tr>\n  </tbody>\n</table>\n</div>"
     },
     "metadata": {}
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "# create emp_length dummy data frame\n",
    "home_ownership = pd.DataFrame(pd.get_dummies(collateral_df['home_ownership'], prefix='home_ownership'))\n",
    "\n",
    "# join the loan_amnt dummy dataframe to conditions_df\n",
    "collateral_df = pd.concat([collateral_df, home_ownership], axis=1, sort=False)\n",
    "\n",
    "# drop original emp_length feature\n",
    "collateral_df.drop(['home_ownership'], axis=1, inplace=True)\n",
    "\n",
    "collateral_df.head()"
   ],
   "execution_count": 20,
   "outputs": [
    {
     "output_type": "execute_result",
     "execution_count": 20,
     "data": {
      "text/plain": "   home_ownership_ANY  home_ownership_MORTGAGE  home_ownership_NONE  \\\n0                   0                        0                    0   \n1                   0                        1                    0   \n2                   0                        1                    0   \n3                   0                        1                    0   \n4                   0                        1                    0   \n\n   home_ownership_OTHER  home_ownership_OWN  home_ownership_RENT  \n0                     0                   0                    1  \n1                     0                   0                    0  \n2                     0                   0                    0  \n3                     0                   0                    0  \n4                     0                   0                    0  ",
      "text/html": "<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>home_ownership_ANY</th>\n      <th>home_ownership_MORTGAGE</th>\n      <th>home_ownership_NONE</th>\n      <th>home_ownership_OTHER</th>\n      <th>home_ownership_OWN</th>\n      <th>home_ownership_RENT</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>1</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n    </tr>\n  </tbody>\n</table>\n</div>"
     },
     "metadata": {}
    }
   ]
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": [
    "# Loan Status\n",
    "\n",
    "Current Status of the Loan. Target variable"
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "loan_status_df = preprocess_df[['loan_status']].copy()\n",
    "loan_status_df.loan_status.value_counts()"
   ],
   "execution_count": 21,
   "outputs": [
    {
     "output_type": "execute_result",
     "execution_count": 21,
     "data": {
      "text/plain": "Fully Paid                                             1041952\nCurrent                                                 919695\nCharged Off                                             261655\nLate (31-120 days)                                       21897\nIn Grace Period                                           8952\nLate (16-30 days)                                         3737\nDoes not meet the credit policy. Status:Fully Paid        1988\nDoes not meet the credit policy. Status:Charged Off        761\nDefault                                                     31\nName: loan_status, dtype: int64"
     },
     "metadata": {}
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "loan_status_df['loan_status'] = loan_status_df['loan_status'].replace({'Fully Paid': 0, 'Current': 0, 'Charged Off': 1,\n",
    "                                                                       'Late (31-120 days)': 1, 'In Grace Period': 1,\n",
    "                                                                       'Late (16-30 days)':1,\n",
    "                                                                       'Does not meet the credit policy. Status:Fully Paid': 0,\n",
    "                                                                       'Does not meet the credit policy. Status:Charged Off': 1,\n",
    "                                                                       'Default':1})\n",
    "loan_status_df.loan_status.value_counts()"
   ],
   "execution_count": 22,
   "outputs": [
    {
     "output_type": "execute_result",
     "execution_count": 22,
     "data": {
      "text/plain": "0    1963635\n1     297033\nName: loan_status, dtype: int64"
     },
     "metadata": {}
    }
   ]
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": [
    "# Combine all the processed dataframes together into a single one"
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "# Concatenate all the processed dataframes into a single one\n",
    "processed_df = pd.concat([character_df,\n",
    "                          capacity_df,\n",
    "                          conditions_df,\n",
    "                          collateral_df,\n",
    "                          loan_status_df,\n",
    "                          raw_data['grade']], axis=1, sort=False)\n",
    "processed_df.head()"
   ],
   "execution_count": 23,
   "outputs": [
    {
     "output_type": "execute_result",
     "execution_count": 23,
     "data": {
      "text/plain": "   credit_hist_in_months  cb_person_bk_on_file_Y  annual_inc    dti  \\\n0                    212                       1     55000.0  18.24   \n1                    378                       1     90000.0  26.52   \n2                     92                       0     59280.0  10.51   \n3                    153                       0     92000.0  16.74   \n4                    215                       0     57250.0  26.35   \n\n   loan_amnt  int_rate  term_60  home_ownership_ANY  home_ownership_MORTGAGE  \\\n0       2500     13.56        0                   0                        0   \n1      30000     18.94        1                   0                        1   \n2       5000     17.97        0                   0                        1   \n3       4000     18.94        0                   0                        1   \n4      30000     16.14        1                   0                        1   \n\n   home_ownership_NONE  home_ownership_OTHER  home_ownership_OWN  \\\n0                    0                     0                   0   \n1                    0                     0                   0   \n2                    0                     0                   0   \n3                    0                     0                   0   \n4                    0                     0                   0   \n\n   home_ownership_RENT  loan_status grade  \n0                    1            0     C  \n1                    0            0     D  \n2                    0            0     D  \n3                    0            0     D  \n4                    0            0     C  ",
      "text/html": "<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>credit_hist_in_months</th>\n      <th>cb_person_bk_on_file_Y</th>\n      <th>annual_inc</th>\n      <th>dti</th>\n      <th>loan_amnt</th>\n      <th>int_rate</th>\n      <th>term_60</th>\n      <th>home_ownership_ANY</th>\n      <th>home_ownership_MORTGAGE</th>\n      <th>home_ownership_NONE</th>\n      <th>home_ownership_OTHER</th>\n      <th>home_ownership_OWN</th>\n      <th>home_ownership_RENT</th>\n      <th>loan_status</th>\n      <th>grade</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>212</td>\n      <td>1</td>\n      <td>55000.0</td>\n      <td>18.24</td>\n      <td>2500</td>\n      <td>13.56</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>C</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>378</td>\n      <td>1</td>\n      <td>90000.0</td>\n      <td>26.52</td>\n      <td>30000</td>\n      <td>18.94</td>\n      <td>1</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>D</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>92</td>\n      <td>0</td>\n      <td>59280.0</td>\n      <td>10.51</td>\n      <td>5000</td>\n      <td>17.97</td>\n      <td>0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>D</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>153</td>\n      <td>0</td>\n      <td>92000.0</td>\n      <td>16.74</td>\n      <td>4000</td>\n      <td>18.94</td>\n      <td>0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>D</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>215</td>\n      <td>0</td>\n      <td>57250.0</td>\n      <td>26.35</td>\n      <td>30000</td>\n      <td>16.14</td>\n      <td>1</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>C</td>\n    </tr>\n  </tbody>\n</table>\n</div>"
     },
     "metadata": {}
    }
   ]
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": [
    "# Count total Non-Defualt and Defaulted Loans in the Data Set\n",
    "\n",
    "Where \n",
    "\n",
    "'0' is a loan that has NOT defaulted \n",
    "\n",
    "and \n",
    "\n",
    "'1' is a loan that HAS defaulted"
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "print(processed_df.loan_status.value_counts())"
   ],
   "execution_count": 24,
   "outputs": [
    {
     "output_type": "stream",
     "text": "0    1963635\n1     297033\nName: loan_status, dtype: int64\n",
     "name": "stdout"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "ax = processed_df.loan_status.value_counts().plot(kind='bar')\n",
    "labels = ['Non-Default', 'Default']\n",
    "ax.set_xticklabels(labels, rotation='horizontal')\n",
    "ax.set_ylabel('Number of Loans')"
   ],
   "execution_count": 25,
   "outputs": [
    {
     "output_type": "execute_result",
     "execution_count": 25,
     "data": {
      "text/plain": "Text(0, 0.5, 'Number of Loans')"
     },
     "metadata": {}
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "<Figure size 432x288 with 1 Axes>",
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAaEAAAD8CAYAAADezxtfAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvnQurowAAHVFJREFUeJzt3X+0XGV97/H3x0SUUiFBDjTNDxP1WAXUAKcQL1VRagigBDDUcNUcMe1BFlSs9tZgtUEiq6ALqFSFxpKSuCwJBYTcGhpjRNRVQQ4QAwFpDhHhmDQJJCaxIN6E7/1jP2N2DnNm5kwy5wHm81prr9nz3c+vyZrkm/3sZ/ZWRGBmZpbDy3IPwMzM2peTkJmZZeMkZGZm2TgJmZlZNk5CZmaWjZOQmZll4yRkZmbZOAmZmVk2TkJmZpbNyNwDeKE75JBDYuLEibmHYWb2onLvvfc+GREd9co5CdUxceJEent7cw/DzOxFRdIvGinn6TgzM8umZUlI0nhJd0h6WNIaSRem+MGSVkham15Hp7gkXS2pT9JqSUeX2upO5ddK6i7Fj5H0QKpztSQ124eZmQ2/Vp4J7QQ+FRFvAqYA50s6HJgDrIyITmBleg9wMtCZth7gGigSCjAXOA44FphbSSqpTE+p3rQUH1IfZmaWR8uSUERsiIj70v4O4GFgLDAdWJiKLQROT/vTgUVRuAsYJWkMcBKwIiK2RMRWYAUwLR07MCJ+HMXzKBYNaGsofZiZWQbDck1I0kTgKOBu4LCI2ABFogIOTcXGAk+UqvWnWK14f5U4TfRhZmYZtDwJSfp94GbgExGxvVbRKrFoIl5zOI3UkdQjqVdS7+bNm+s0aWZmzWppEpL0cooE9M2IuCWFN1amwNLrphTvB8aXqo8D1teJj6sSb6aPPUTE/Ijoioiujo66y9zNzKxJrVwdJ+A64OGIuLJ0aClQWeHWDdxWis9KK9imANvSVNpyYKqk0WlBwlRgeTq2Q9KU1NesAW0NpQ8zM8uglT9WPR74MPCApFUp9hngMuBGSbOBx4Gz0rFlwClAH/A0cA5ARGyRNA+4J5W7JCK2pP3zgOuB/YHb08ZQ+zAzszxULCyzwXR1dcWL4Y4JE+d8O/cQXlIeu+zU3EMwe1GTdG9EdNUr5zsmmJlZNk5CZmaWjZOQmZll4yRkZmbZOAmZmVk2TkJmZpaNk5CZmWXjJGRmZtk4CZmZWTZOQmZmlo2TkJmZZeMkZGZm2TgJmZlZNk5CZmaWjZOQmZll4yRkZmbZOAmZmVk2TkJmZpZNy5KQpAWSNkl6sBRbImlV2h6TtCrFJ0p6pnTs2lKdYyQ9IKlP0tWSlOIHS1ohaW16HZ3iSuX6JK2WdHSpre5Ufq2k7lZ9djMza0wrz4SuB6aVAxHxgYiYHBGTgZuBW0qHH60ci4iPleLXAD1AZ9oqbc4BVkZEJ7AyvQc4uVS2J9VH0sHAXOA44FhgbiVxmZlZHi1LQhHxA2BLtWPpbObPgBtqtSFpDHBgRPw4IgJYBJyeDk8HFqb9hQPii6JwFzAqtXMSsCIitkTEVmAFA5KkmZkNr1zXhN4ObIyItaXYJEn3S7pT0ttTbCzQXyrTn2IAh0XEBoD0emipzhNV6gwWfx5JPZJ6JfVu3rx56J/OzMwakisJnc2eZ0EbgAkRcRTwSeBfJR0IqErdqNP2YHUabisi5kdEV0R0dXR01OnOzMyaNexJSNJI4ExgSSUWEc9GxFNp/17gUeANFGcr40rVxwHr0/7GNM1WmbbblOL9wPgqdQaLm5lZJjnOhP4U+FlE/G6aTVKHpBFp/7UUiwrWpWm2HZKmpOtIs4DbUrWlQGWFW/eA+Ky0Sm4KsC21sxyYKml0WpAwNcXMzCyTka1qWNINwAnAIZL6gbkRcR0wk+cvSHgHcImkncAu4GMRUVnUcB7FSrv9gdvTBnAZcKOk2cDjwFkpvgw4BegDngbOAYiILZLmAfekcpeU+jAzswxaloQi4uxB4h+pEruZYsl2tfK9wJFV4k8BJ1aJB3D+IG0tABbUGreZmQ0f3zHBzMyycRIyM7NsnITMzCwbJyEzM8vGScjMzLJxEjIzs2ychMzMLBsnITMzy8ZJyMzMsnESMjOzbJyEzMwsGychMzPLxknIzMyycRIyM7NsnITMzCwbJyEzM8vGScjMzLJpWRKStEDSJkkPlmIXS/qlpFVpO6V07CJJfZIekXRSKT4txfokzSnFJ0m6W9JaSUsk7Zfir0jv+9LxifX6MDOzPFp5JnQ9MK1K/KqImJy2ZQCSDgdmAkekOl+TNELSCOCrwMnA4cDZqSzA5amtTmArMDvFZwNbI+L1wFWp3KB97OPPbGZmQ9CyJBQRPwC2NFh8OrA4Ip6NiJ8DfcCxaeuLiHUR8VtgMTBdkoB3Azel+guB00ttLUz7NwEnpvKD9WFmZpnkuCZ0gaTVabpudIqNBZ4olelPscHirwZ+FRE7B8T3aCsd35bKD9aWmZllMtxJ6BrgdcBkYANwRYqrStloIt5MW88jqUdSr6TezZs3VytiZmb7wLAmoYjYGBG7IuI54Ovsng7rB8aXio4D1teIPwmMkjRyQHyPttLxgyimBQdrq9o450dEV0R0dXR0NPNRzcysAcOahCSNKb09A6isnFsKzEwr2yYBncBPgHuAzrQSbj+KhQVLIyKAO4AZqX43cFupre60PwP4Xio/WB9mZpbJyPpFmiPpBuAE4BBJ/cBc4ARJkymmwR4DzgWIiDWSbgQeAnYC50fErtTOBcByYASwICLWpC4+DSyW9AXgfuC6FL8O+IakPoozoJn1+jAzszxUnCTYYLq6uqK3tzf3MOqaOOfbuYfwkvLYZafmHoLZi5qkeyOiq1453zHBzMyycRIyM7NsnITMzCybuklI0vGSDkj7H5J0paTXtH5oZmb2UtfImdA1wNOS3gr8DfALYFFLR2VmZm2hkSS0M/3OZjrw5Yj4MvCq1g7LzMzaQSO/E9oh6SLgQ8A70p2nX97aYZmZWTto5EzoA8CzwOyI+G+Km35+qaWjMjOztlD3TCglnitL7x/H14TMzGwfaGR13Jnp6aXbJG2XtEPS9uEYnJmZvbQ1ck3oi8D7IuLhVg/GzMzaSyPXhDY6AZmZWSs0cibUK2kJcCvFAgUAIuKWlo3KzMzaQiNJ6EDgaWBqKRaAk5CZme2VRlbHnTMcAzEzs/ZTNwlJeiUwGzgCeGUlHhEfbeG4zMysDTSyMOEbwB8AJwF3AuOAHa0clJmZtYdGktDrI+JzwP9ExELgVODN9SpJWiBpk6QHS7EvSfqZpNWSviVpVIpPlPSMpFVpu7ZU5xhJD0jqk3S1JKX4wZJWpN8wrZA0OsWVyvWlfo4utdWdyq+V1N3oH5KZmbVGI0no/6XXX0k6EjgImNhAveuBaQNiK4AjI+ItwH8BF5WOPRoRk9P2sVL8GqAH6Exbpc05wMqI6ARWpvcAJ5fK9qT6SDoYmAscBxwLzK0kLjMzy6ORJDQ//WP9OWAp8BBweb1KEfEDYMuA2HciYmd6exfF1N6gJI0BDoyIH6c7eS8CTk+HpwML0/7CAfFFUbgLGJXaOQlYERFbImIrRUIcmCTNzGwYNbI67p/T7p3Aa/dh3x8FlpTeT5J0P7Ad+GxE/JDiZqn9pTL9KQZwWERsSGPcIOnQFB8LPFGlzmBxMzPLpJHVcQcBFwNvT6HvA/MiYluznUr6W2An8M0U2gBMiIinJB0D3CrpCEBVqke95gep03BbknoopvKYMGFCne7MzKxZjUzHLaA4O/mztO0A/qXZDtOCgPcCH0xTbETEsxHxVNq/F3gUeAPF2Up5ym4csD7tb0zTbJVpu00p3g+Mr1JnsPjzRMT8iOiKiK6Ojo5mP6qZmdXRSBJ6XUTMjYh1afs8TU7LSZoGfBo4LSKeLsU70sPykPRaikUF69J02w5JU9KquFnAbanaUqCywq17QHxWWiU3BdiW2lkOTJU0Ol3jmppiZmaWSSO37XlG0p9ExI8AJB0PPFOvkqQbgBOAQyT1U6xMuwh4BbAirbS+K62EewdwiaSdwC7gYxFRWdRwHsVKu/2B29MGcBlwo6TZwOPAWSm+DDgF6KO43dA5ABGxRdI84J5U7pJSH2ZmlkEjSehjwKJ0bQhgK7vPQAYVEWdXCV83SNmbgZsHOdYLHFkl/hRwYpV4AOcP0tYCiulFMzN7AWhkddxPgbdKOjC93y7p/cDqVg/OzMxe2hq5JgQUySciKk9UvapF4zEzszbScBIaoNpyZzMzsyFpNgnV+62OmZlZXYNeE5L0ANWTjYDDWjYiMzNrG7UWJrx32EZhZmZtadAkFBG/GM6BmJlZ+2n2mpCZmdlecxIyM7NsBk1Cklam17rPDjIzM2tGrYUJYyS9EzhN0mIG/DYoIu5r6cjMzOwlr1YS+juKR2aPA64ccCyAd7dqUGZm1h5qrY67CbhJ0uciYt4wjsnMzNpEIzcwnSfpNIrHLQB8PyL+vbXDMjOzdlB3dZykvwcuBB5K24UpZmZmtlcaeZ7QqcDkiHgOQNJC4H6KB9SZmZk1rdHfCY0q7R80aCkzM7MhaORM6O+B+yXdQbFM+x34LMjMzPaBumdCEXEDMAW4JW1vi4jFjTQuaYGkTZIeLMUOlrRC0tr0OjrFJelqSX2SVks6ulSnO5VfK6m7FD9G0gOpztWS1GwfZmY2/BqajouIDRGxNCJui4j/HkL71wPTBsTmACsjohNYmd4DnAx0pq0HuAaKhALMBY4DjgXmVpJKKtNTqjetmT7MzCyPlt47LiJ+AGwZEJ4OLEz7C4HTS/FFUbgLGCVpDHASsCIitkTEVmAFMC0dOzAifhwRASwa0NZQ+jAzswxy3MD0sIjYAMUZFnBoio8FniiV60+xWvH+KvFm+tiDpB5JvZJ6N2/ePOQPaGZmjamZhCS9rHw9p8VUJRZNxJvpY89AxPyI6IqIro6OjjpNmplZs2omofTboJ9KmrAP+9xYmQJLr5tSvB8YXyo3DlhfJz6uSryZPszMLINGpuPGAGskrZS0tLLtRZ9LgcoKt27gtlJ8VlrBNgXYlqbSlgNTJY1OCxKmAsvTsR2SpqRVcbMGtDWUPszMLINGfif0+WYbl3QDcAJwiKR+ilVulwE3SpoNPA6clYovA04B+oCngXMAImKLpHnAPancJRFRWexwHsUKvP2B29PGUPswM7M8GrmB6Z2SXgN0RsR3Jf0eMKKRxiPi7EEOnVilbADnD9LOAmBBlXgvcGSV+FND7cPMzIZfIzcw/QvgJuCfUmgscGsrB2VmZu2hkWtC5wPHA9sBImItu5c8m5mZNa2RJPRsRPy28kbSSOovhTYzM6urkSR0p6TPAPtLeg/wb8D/be2wzMysHTSShOYAm4EHgHMpVph9tpWDMjOz9tDI6rjn0oPs7qaYhnskrTIzMzPbK3WTkKRTgWuBRyluezNJ0rkRcXvtmmZmZrU18mPVK4B3RUQfgKTXAd9m9w9DzczMmtLINaFNlQSUrGP3vdjMzMyaNuiZkKQz0+4aScuAGymuCZ3F7lvomJmZNa3WdNz7SvsbgXem/c3A6OcXNzMzG5pBk1BE+OaeZmbWUo2sjpsE/CUwsVw+Ik5r3bDMzKwdNLI67lbgOoq7JDzX2uGYmVk7aSQJ/SYirm75SMzMrO00koS+LGku8B3g2UowIu5r2ajMzKwtNJKE3gx8GHg3u6fjIr03MzNrWiM/Vj0DeG1EvDMi3pW2phOQpD+StKq0bZf0CUkXS/plKX5Kqc5FkvokPSLppFJ8Wor1SZpTik+SdLektZKWSNovxV+R3vel4xOb/RxmZrb3GklCPwVG7asOI+KRiJgcEZOBY4CngW+lw1dVjkXEMgBJhwMzgSOAacDXJI2QNAL4KnAycDhwdioLcHlqqxPYCsxO8dnA1oh4PXBVKmdmZpk0koQOA34mabmkpZVtH/V/IvBoRPyiRpnpwOKIeDYifg70AcemrS8i1qWH7i0GpksSxVThTan+QuD0UlsL0/5NwImpvJmZZdDINaG5Lex/JnBD6f0FkmYBvcCnImIrMBa4q1SmP8UAnhgQPw54NfCriNhZpfzYSp2I2ClpWyr/5D77RGZm1rBGnid0Zys6TtdpTgMuSqFrgHkUix7mUdy9+6MUj4943rCofhYXNcpT51h5bD1AD8CECRMG/QxmZrZ36k7HSdqRFg9sl/QbSbskbd8HfZ8M3BcRGwEiYmNE7IqI54CvU0y3QXEmM75Ubxywvkb8SWCUpJED4nu0lY4fBGwZOLCImB8RXRHR1dHRsdcf1MzMqqubhCLiVRFxYNpeCbwf+Mo+6PtsSlNxksaUjp0BPJj2lwIz08q2SUAn8BOKO3l3ppVw+1FM7S1NT329A5iR6ncDt5Xa6k77M4Dv+SmxZmb5NHJNaA8RcWt5OXQzJP0e8B7g3FL4i5ImU0yPPVY5FhFrJN0IPATsBM6PiF2pnQuA5cAIYEFErEltfRpYLOkLwP0Utx0ivX5DUh/FGdDMvfkcZma2dxq5gemZpbcvA7qoch1lKCLiaYoFAeXYh2uUvxS4tEp8GbCsSnwdu6fzyvHfUDwPyczMXgAaORMqP1doJ8VZyvSWjMbMzNpKI6vj/FwhMzNriVqP9/67GvUiIua1YDxmZtZGap0J/U+V2AEUt755NcVveczMzJpW6/HeV1T2Jb0KuBA4h+L2OFcMVs/MzKxRNa8JSToY+CTwQYp7rh2dbqVjZma212pdE/oScCYwH3hzRPx62EZlZmZtodYdEz4F/CHwWWB96dY9O/bRbXvMzKzN1bom1MhjHszMzJrmRGNmZtk4CZmZWTZOQmZmlo2TkJmZZeMkZGZm2TgJmZlZNk5CZmaWjZOQmZllky0JSXpM0gOSVknqTbGDJa2QtDa9jk5xSbpaUp+k1ZKOLrXTncqvldRdih+T2u9LdVWrDzMzG365z4TeFRGTI6IrvZ8DrIyITmBleg9wMtCZth7gGvjdDVbnAsdRPM57bimpXJPKVupNq9OHmZkNs9xJaKDpFHfrJr2eXoovisJdwChJY4CTgBURsSXd3XsFMC0dOzAifhwRASwa0Fa1PszMbJjlTEIBfEfSvZJ6UuywiNgAkF4PTfGxwBOluv0pViveXyVeqw8zMxtmNZ8n1GLHR8R6SYcCKyT9rEZZVYlFE/GGpKTYAzBhwoRGq5mZ2RBlOxOKiPXpdRPwLYprOhvTVBrpdVMq3g+ML1UfB6yvEx9XJU6NPspjmx8RXRHR1dHRsTcf08zMasiShCQdkB4ZjqQDgKnAg8BSoLLCrRu4Le0vBWalVXJTgG1pKm05MFXS6LQgYSqwPB3bIWlKWhU3a0Bb1fowM7Nhlms67jDgW2nV9EjgXyPiPyTdA9woaTbwOHBWKr8MOAXoA54GzgGIiC2S5gH3pHKXRMSWtH8ecD2wP3B72gAuG6QPMzMbZlmSUESsA95aJf4UcGKVeADnD9LWAmBBlXgvcGSjfZiZ2fB7oS3RNjOzNuIkZGZm2TgJmZlZNk5CZmaWjZOQmZll4yRkZmbZOAmZmVk2TkJmZpaNk5CZmWXjJGRmZtk4CZmZWTZOQmZmlo2TkJmZZeMkZGZm2TgJmZlZNk5CZmaWTa4nq5pZG5k459u5h/CS8dhlp+Yewj7lMyEzM8tm2JOQpPGS7pD0sKQ1ki5M8Ysl/VLSqrSdUqpzkaQ+SY9IOqkUn5ZifZLmlOKTJN0taa2kJZL2S/FXpPd96fjE4fvkZmY2UI4zoZ3ApyLiTcAU4HxJh6djV0XE5LQtA0jHZgJHANOAr0kaIWkE8FXgZOBw4OxSO5entjqBrcDsFJ8NbI2I1wNXpXJmZpbJsCehiNgQEfel/R3Aw8DYGlWmA4sj4tmI+DnQBxybtr6IWBcRvwUWA9MlCXg3cFOqvxA4vdTWwrR/E3BiKm9mZhlkvSaUpsOOAu5OoQskrZa0QNLoFBsLPFGq1p9ig8VfDfwqInYOiO/RVjq+LZUfOK4eSb2Sejdv3rxXn9HMzAaXLQlJ+n3gZuATEbEduAZ4HTAZ2ABcUSlapXo0Ea/V1p6BiPkR0RURXR0dHTU/h5mZNS9LEpL0cooE9M2IuAUgIjZGxK6IeA74OsV0GxRnMuNL1ccB62vEnwRGSRo5IL5HW+n4QcCWffvpzMysUTlWxwm4Dng4Iq4sxceUip0BPJj2lwIz08q2SUAn8BPgHqAzrYTbj2LxwtKICOAOYEaq3w3cVmqrO+3PAL6XypuZWQY5fqx6PPBh4AFJq1LsMxSr2yZTTI89BpwLEBFrJN0IPESxsu78iNgFIOkCYDkwAlgQEWtSe58GFkv6AnA/RdIjvX5DUh/FGdDMVn5QMzOrbdiTUET8iOrXZpbVqHMpcGmV+LJq9SJiHbun88rx3wBnDWW8ZmbWOr5jgpmZZeMkZGZm2TgJmZlZNk5CZmaWjZOQmZll4yRkZmbZOAmZmVk2TkJmZpaNk5CZmWXjJGRmZtk4CZmZWTZOQmZmlo2TkJmZZeMkZGZm2TgJmZlZNk5CZmaWjZOQmZll05ZJSNI0SY9I6pM0J/d4zMzaVdslIUkjgK8CJwOHA2dLOjzvqMzM2lPbJSHgWKAvItZFxG+BxcD0zGMyM2tL7ZiExgJPlN73p5iZmQ2zkbkHkIGqxGKPAlIP0JPe/lrSIy0fVfs4BHgy9yDq0eW5R2AZ+Lu5b72mkULtmIT6gfGl9+OA9eUCETEfmD+cg2oXknojoiv3OMwG8nczj3acjrsH6JQ0SdJ+wExgaeYxmZm1pbY7E4qInZIuAJYDI4AFEbEm87DMzNpS2yUhgIhYBizLPY425WlOe6HydzMDRUT9UmZmZi3QjteEzMzsBcJJqI1JCklXlN7/taSL91HbF0v6paRVktZKuqWRO1NIemOqc7+k1zXZ71+n/Y9I+sNmxm8vLpJ2pe/NGkk/lfRJSXX/fZP0pVTnS032++v0OlHS/26mjXbnJNTengXOlHRIi9q/KiImR0QnsAT4nqSOOnVOB26LiKMi4tG97P8jgJNQe3gmfdeOAN4DnALMbaDeucDREfF/9rL/iYCTUBOchNrbToqLsX818ICk10haKWl1ep2Q4tdLulrSf0paJ2lGIx1FxBLgO6S/qJKOkXSnpHslLZc0RtIpwCeAP5d0Ryp3ayqzJv2IuDK+X5f2Z0i6fsD4ZwBdwDfT/5D3H9KfjL1oRcQmih+bX6DCiHTGc0/6Pp8LIGkpcABwt6QPSHqfpLvTWfh3JR2Wyv3u7Dq9f1DSxAHdXga8PX3Xnvf3yQbnJGRfBT4o6aAB8a8AiyLiLcA3gatLx8YAfwK8l+IvX6PuA94o6eXAPwIzIuIYYAFwaVq1eC3FGdS7Up2PpjJdwMclvbqRjiLiJqAX+GD6H/IzQxinvchFxDqKf98OBWYD2yLij4E/Bv5C0qSIOI3dZ1BLgB8BUyLiKIp7Sv7NELqcA/wwtXXVPv0wL3FtuUTbdouI7ZIWAR8Hyv9Qvw04M+1/A/hi6ditEfEc8FDlf4sNqtwy6Y+AI4EVkqD4vdaGQep8XNIZaX880Ak8NYQ+rX1Vvm9TgbeUztoPovge/XxA+XHAEkljgP2qHLcWcBIygH+gOEv5lxplymv5ny3tC0DSpcCpABExeZA2jqI4OxGwJiLeVmtQkk4A/hR4W0Q8Len7wCurjOeVmJVIei2wC9hE8X37y4hYXqfaPwJXRsTS9N27OMV3sueskb9v+5Cn44yI2ALcSDFtUfGfFLc0AvggxVRFrTb+Nk1FVE1Akt5P8T/SG4BHgA5Jb0vHXi7piCrVDgK2pgT0RmBK6dhGSW9KK6DOqFIXYAfwqlrjtpeetPjlWuArUfwQcjlwXpoGRtIbJB1QpepBwC/Tfncp/hhwdKp7NDCpSl1/15rkJGQVV1DcRbji48A5klYDHwYubKLNv6os0QY+BLw7Ijan5zjNAC6X9FNgFfC/qtT/D2BkGsM84K7SsTnAvwPfY/CpvOuBa70woS3sX1miDXyXYhHM59OxfwYeAu6T9CDwT1SfBboY+DdJP2TPu2nfDBwsaRVwHvBfVequBnam5eFemDAEvmOCmZll4zMhMzPLxknIzMyycRIyM7NsnITMzCwbJyEzM8vGScjMzLJxEjIzs2ychMzMLJv/D5TH/NNnmgbFAAAAAElFTkSuQmCC\n"
     },
     "metadata": {
      "needs_background": "light"
     }
    }
   ]
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": [
    " # EDA"
   ]
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": [
    "### Defaults by Grade\n",
    "\n",
    "Loans rated as \"C\" show the highest instances of defaults"
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "grade_default = pd.crosstab(processed_df['grade'], processed_df['loan_status'])\n",
    "\n",
    "fig, ax = plt.subplots()\n",
    "\n",
    "grade_default.plot.bar(legend=True, alpha=0.7, ax=ax)\n",
    "plt.title(\"Non-Default vs Default by Grade\")\n",
    "ax.legend([\"Non-Default\", \"Default\"])"
   ],
   "execution_count": 26,
   "outputs": [
    {
     "output_type": "execute_result",
     "execution_count": 26,
     "data": {
      "text/plain": "<matplotlib.legend.Legend at 0x7f4f38dcd8d0>"
     },
     "metadata": {}
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "<Figure size 432x288 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     }
    }
   ]
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": [
    "# Correlation in Credit Features"
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "# Sample figsize in inches\n",
    "fig, ax = plt.subplots(figsize=(20,10))\n",
    "\n",
    "# Imbalanced DataFrame Correlation\n",
    "corr = processed_df.corr()\n",
    "sns.heatmap(corr, cmap='YlGnBu', annot_kws={'size':30}, ax=ax)\n",
    "ax.set_title(\"Imbalanced Correlation Matrix\", fontsize=14)\n",
    "plt.show()"
   ],
   "execution_count": 27,
   "outputs": [
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "<Figure size 1440x720 with 2 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     }
    }
   ]
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": [
    "# Undersampling\n",
    "\n",
    "The process where you randomly delete some of the observations from the majority class in order to match the numbers with the minority class"
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "# Shuffle the Dataset.\n",
    "random_data = processed_df.sample(frac=1,random_state=4)\n",
    "\n",
    "# Put all the fraud class in a separate dataset.\n",
    "default_df = random_data.loc[random_data['loan_status'] == 1]\n",
    "\n",
    "#Randomly select 297033 observations from the non-fraud (majority class)\n",
    "non_default_df = random_data.loc[random_data['loan_status'] == 0].sample(n=297033, random_state=42)"
   ],
   "execution_count": 28,
   "outputs": []
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "# Concatenate both dataframes again\n",
    "US_df = pd.concat([default_df, non_default_df])\n",
    "\n",
    "#plot the dataset after the undersampling\n",
    "plt.figure(figsize=(8, 8))\n",
    "sns.countplot('loan_status', data=US_df)\n",
    "plt.title('Undersampled Balanced Loan Status')\n",
    "plt.xticks(ticks=(0,1), labels=('Non-Default', 'Default'))\n",
    "plt.show()"
   ],
   "execution_count": 29,
   "outputs": [
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "<Figure size 576x576 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     }
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "# Create X and y using undersampled dataframe\n",
    "X = US_df.drop(['loan_status', 'grade'], axis=1)\n",
    "y = US_df['loan_status']\n",
    "\n",
    "X_train_US, X_test_US, y_train_US, y_test_US = train_test_split(X, y, test_size=.4, random_state=123)"
   ],
   "execution_count": 30,
   "outputs": []
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "# Create, train, and fit a logistic regression model\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "clf_logistic = LogisticRegression(solver='lbfgs').fit(X_train_US, np.ravel(y_train_US))\n",
    "\n",
    "# Create predictions of probability for loan status using test data\n",
    "# .predict_proba creates an array of probabilities of default: [[non-defualt, default]]\n",
    "lr_preds = clf_logistic.predict_proba(X_test_US)\n",
    "\n",
    "# # Create dataframes of predictions and true labels\n",
    "lr_preds_df = pd.DataFrame(lr_preds[:,1][0:], columns = ['lr_pred_PD'])\n",
    "true_df = y_test_US\n",
    "\n",
    "# Concatenate and print the two data frames for comparison\n",
    "print(pd.concat([true_df.reset_index(drop = True), lr_preds_df], axis = 1))"
   ],
   "execution_count": 31,
   "outputs": [
    {
     "output_type": "stream",
     "text": "        loan_status  lr_pred_PD\n0                 0    0.530770\n1                 0    0.510444\n2                 0    0.419373\n3                 0    0.631293\n4                 1    0.639652\n...             ...         ...\n237622            1    0.742460\n237623            1    0.571265\n237624            1    0.327558\n237625            0    0.528750\n237626            0    0.753022\n\n[237627 rows x 2 columns]\n",
     "name": "stdout"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "# Reassign loan status based on the threshold and print the predictions\n",
    "lr_preds_df['lr_pred_loan_status_60'] = lr_preds_df['lr_pred_PD'].apply(lambda x: 1 if x > 0.60 else 0)\n",
    "print(\"Non-Default / Default predictions at 60% Threshhold: \")\n",
    "print(lr_preds_df['lr_pred_loan_status_60'].value_counts())\n",
    "\n",
    "# Print the confusion matrix\n",
    "from sklearn.metrics import confusion_matrix\n",
    "print(\"Confusion Matrix at 60% Threshhold: \")\n",
    "print(confusion_matrix(y_test_US, lr_preds_df['lr_pred_loan_status_60']))\n",
    "\n",
    "# Print the classification report\n",
    "from sklearn.metrics import classification_report\n",
    "target_names = ['Non-Default', 'Default']\n",
    "print(classification_report(y_test_US, lr_preds_df['lr_pred_loan_status_60'], target_names=target_names))"
   ],
   "execution_count": 32,
   "outputs": [
    {
     "output_type": "stream",
     "text": "Non-Default / Default predictions at 60% Threshhold: \n0    176631\n1     60996\nName: lr_pred_loan_status_60, dtype: int64\nConfusion Matrix at 60% Threshhold: \n[[98568 20453]\n [78063 40543]]\n              precision    recall  f1-score   support\n\n Non-Default       0.56      0.83      0.67    119021\n     Default       0.66      0.34      0.45    118606\n\n    accuracy                           0.59    237627\n   macro avg       0.61      0.58      0.56    237627\nweighted avg       0.61      0.59      0.56    237627\n\n",
     "name": "stdout"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "# Reassign loan status based on the threshold and print the predictions\n",
    "lr_preds_df['lr_pred_loan_status_50'] = lr_preds_df['lr_pred_PD'].apply(lambda x: 1 if x > 0.50 else 0)\n",
    "print(\"Non-Default / Default t predictions at 50% Threshhold: \")\n",
    "print(lr_preds_df['lr_pred_loan_status_50'].value_counts())\n",
    "\n",
    "# Print the confusion matrix\n",
    "from sklearn.metrics import confusion_matrix\n",
    "print(\"Confusion Matrix at 50% Threshhold: \")\n",
    "print(confusion_matrix(y_test_US, lr_preds_df['lr_pred_loan_status_50']))\n",
    "\n",
    "# Print the classification report\n",
    "from sklearn.metrics import classification_report\n",
    "target_names = ['Non-Default', 'Default']\n",
    "print(classification_report(y_test_US, lr_preds_df['lr_pred_loan_status_50'], target_names=target_names))"
   ],
   "execution_count": 33,
   "outputs": [
    {
     "output_type": "stream",
     "text": "Non-Default / Default t predictions at 50% Threshhold: \n1    138916\n0     98711\nName: lr_pred_loan_status_50, dtype: int64\nConfusion Matrix at 50% Threshhold: \n[[62234 56787]\n [36477 82129]]\n              precision    recall  f1-score   support\n\n Non-Default       0.63      0.52      0.57    119021\n     Default       0.59      0.69      0.64    118606\n\n    accuracy                           0.61    237627\n   macro avg       0.61      0.61      0.60    237627\nweighted avg       0.61      0.61      0.60    237627\n\n",
     "name": "stdout"
    }
   ]
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": [
    "# Plotting AUC for Logistic Regression Classification with Undersampled Data"
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "# Print the accuracy score the model\n",
    "print(clf_logistic.score(X_test_US, y_test_US))\n",
    "\n",
    "# Plot the ROC curve of the probabilities of default\n",
    "from sklearn.metrics import roc_curve\n",
    "\n",
    "lr_prob_default = lr_preds[:, 1]\n",
    "fallout, sensitivity, thresholds = roc_curve(y_test_US, lr_prob_default)\n",
    "plt.plot(fallout, sensitivity, color = 'darkorange')\n",
    "plt.plot([0, 1], [0, 1], linestyle='--')\n",
    "plt.title(\"ROC Chart for LR on PD\")\n",
    "plt.xlabel(\"Fall-out\")\n",
    "plt.ylabel(\"Sensitivity\")\n",
    "plt.show()\n",
    "\n",
    "# Compute the AUC and store it in a variable\n",
    "from sklearn.metrics import roc_auc_score\n",
    "\n",
    "lr_auc = roc_auc_score(y_test_US, lr_prob_default)"
   ],
   "execution_count": 34,
   "outputs": [
    {
     "output_type": "stream",
     "text": "0.6075193475488897\n",
     "name": "stdout"
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "<Figure size 432x288 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     }
    }
   ]
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": [
    "# Oversampling\n",
    "\n",
    "The process of generating synthetic data that tries to randomly generate a sample of the attributes from observations in the minority class. There are a number of methods used to oversample a dataset for a typical classification problem. The most common technique is called SMOTE (Synthetic Minority Over-sampling Technique)"
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "# Create X and y using processed_df\n",
    "\n",
    "X = processed_df.drop(['loan_status', 'grade'], axis=1)\n",
    "y = processed_df['loan_status']\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=.4, random_state=123)"
   ],
   "execution_count": 35,
   "outputs": []
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "from imblearn.over_sampling import SMOTE\n",
    "\n",
    "# Resample the minority class. You can change the strategy to 'auto' if you are not sure.\n",
    "sm = SMOTE(sampling_strategy='minority', random_state=7)\n",
    "\n",
    "# Fit the model to generate the data.\n",
    "oversampled_trainX, oversampled_trainY = sm.fit_sample(X_train, y_train)\n",
    "oversampled_train = pd.concat([pd.DataFrame(oversampled_trainY), pd.DataFrame(oversampled_trainX)], axis=1)\n",
    "oversampled_train.columns = processed_df.drop(['grade'], axis=1).columns"
   ],
   "execution_count": 36,
   "outputs": [
    {
     "output_type": "stream",
     "text": "Using TensorFlow backend.\n",
     "name": "stderr"
    }
   ]
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": [
    "# Correlation in Credit Features AFTER oversampling"
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "# Sample figsize in inches\n",
    "fig, ax = plt.subplots(figsize=(20,10))\n",
    "\n",
    "# Imbalanced DataFrame Correlation\n",
    "corr = oversampled_train.corr()\n",
    "sns.heatmap(corr, cmap='YlGnBu', annot_kws={'size':30}, ax=ax)\n",
    "ax.set_title(\"Imbalanced Correlation Matrix after oversampling\", fontsize=14)\n",
    "plt.show()"
   ],
   "execution_count": 37,
   "outputs": [
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "<Figure size 1440x720 with 2 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     }
    }
   ]
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": [
    "# Creating, Training, and Fitting a Logistic Regression Classfier Model to Oversampled Data"
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "# Create, train, and fit a logistic regression model\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "clf_logistic = LogisticRegression(solver='lbfgs').fit(oversampled_trainX, np.ravel(oversampled_trainY))\n",
    "\n",
    "# Create predictions of probability for loan status using test data\n",
    "# .predict_proba creates an array of probabilities of default: [[non-defualt, default]]\n",
    "lr_preds = clf_logistic.predict_proba(X_test)\n",
    "\n",
    "# # Create dataframes of predictions and true labels\n",
    "lr_preds_df = pd.DataFrame(lr_preds[:,1][0:], columns = ['lr_pred_PD'])\n",
    "true_df = y_test\n",
    "\n",
    "# Concatenate and print the two data frames for comparison\n",
    "print(pd.concat([true_df.reset_index(drop = True), lr_preds_df], axis = 1))"
   ],
   "execution_count": 38,
   "outputs": [
    {
     "output_type": "stream",
     "text": "        loan_status  lr_pred_PD\n0                 0    0.494650\n1                 0    0.438002\n2                 0    0.516289\n3                 0    0.550872\n4                 0    0.438621\n...             ...         ...\n904263            0    0.478588\n904264            1    0.549889\n904265            0    0.408460\n904266            0    0.529360\n904267            0    0.656664\n\n[904268 rows x 2 columns]\n",
     "name": "stdout"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "# Reassign loan status based on the threshold and print the predictions\n",
    "lr_preds_df['lr_pred_loan_status_60'] = lr_preds_df['lr_pred_PD'].apply(lambda x: 1 if x > 0.60 else 0)\n",
    "print(\"Non-Default / Default predictions at 60% Threshhold: \")\n",
    "print(lr_preds_df['lr_pred_loan_status_60'].value_counts())\n",
    "\n",
    "# Print the confusion matrix\n",
    "from sklearn.metrics import confusion_matrix\n",
    "print(\"Confusion Matrix at 60% Threshhold: \")\n",
    "print(confusion_matrix(y_test, lr_preds_df['lr_pred_loan_status_60']))\n",
    "\n",
    "# Print the classification report\n",
    "from sklearn.metrics import classification_report\n",
    "target_names = ['Non-Default', 'Default']\n",
    "print(classification_report(y_test, lr_preds_df['lr_pred_loan_status_60'], target_names=target_names))"
   ],
   "execution_count": 39,
   "outputs": [
    {
     "output_type": "stream",
     "text": "Non-Default / Default predictions at 60% Threshhold: \n0    881276\n1     22992\nName: lr_pred_loan_status_60, dtype: int64\nConfusion Matrix at 60% Threshhold: \n[[765309  19895]\n [115967   3097]]\n              precision    recall  f1-score   support\n\n Non-Default       0.87      0.97      0.92    785204\n     Default       0.13      0.03      0.04    119064\n\n    accuracy                           0.85    904268\n   macro avg       0.50      0.50      0.48    904268\nweighted avg       0.77      0.85      0.80    904268\n\n",
     "name": "stdout"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "# Reassign loan status based on the threshold and print the predictions\n",
    "lr_preds_df['lr_pred_loan_status_50'] = lr_preds_df['lr_pred_PD'].apply(lambda x: 1 if x > 0.50 else 0)\n",
    "print(\"Non-Default / Default t predictions at 50% Threshhold: \")\n",
    "print(lr_preds_df['lr_pred_loan_status_50'].value_counts())\n",
    "\n",
    "# Print the confusion matrix\n",
    "from sklearn.metrics import confusion_matrix\n",
    "print(\"Confusion Matrix at 50% Threshhold: \")\n",
    "print(confusion_matrix(y_test, lr_preds_df['lr_pred_loan_status_50']))\n",
    "\n",
    "# Print the classification report\n",
    "from sklearn.metrics import classification_report\n",
    "target_names = ['Non-Default', 'Default']\n",
    "print(classification_report(y_test, lr_preds_df['lr_pred_loan_status_50'], target_names=target_names))"
   ],
   "execution_count": 40,
   "outputs": [
    {
     "output_type": "stream",
     "text": "Non-Default / Default t predictions at 50% Threshhold: \n0    545854\n1    358414\nName: lr_pred_loan_status_50, dtype: int64\nConfusion Matrix at 50% Threshhold: \n[[485141 300063]\n [ 60713  58351]]\n              precision    recall  f1-score   support\n\n Non-Default       0.89      0.62      0.73    785204\n     Default       0.16      0.49      0.24    119064\n\n    accuracy                           0.60    904268\n   macro avg       0.53      0.55      0.49    904268\nweighted avg       0.79      0.60      0.67    904268\n\n",
     "name": "stdout"
    }
   ]
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": [
    "# Plotting AUC for Logistic Regression Classification with Oversampled Data"
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "# Print the accuracy score the model\n",
    "print(clf_logistic.score(X_test, y_test))\n",
    "\n",
    "# Plot the ROC curve of the probabilities of default\n",
    "from sklearn.metrics import roc_curve\n",
    "\n",
    "lr_prob_default = lr_preds[:, 1]\n",
    "fallout, sensitivity, thresholds = roc_curve(y_test, lr_prob_default)\n",
    "plt.plot(fallout, sensitivity, color = 'darkorange')\n",
    "plt.plot([0, 1], [0, 1], linestyle='--')\n",
    "plt.title(\"ROC Chart for LR on PD\")\n",
    "plt.xlabel(\"Fall-out\")\n",
    "plt.ylabel(\"Sensitivity\")\n",
    "plt.show()\n",
    "\n",
    "# Compute the AUC and store it in a variable\n",
    "from sklearn.metrics import roc_auc_score\n",
    "\n",
    "lr_auc = roc_auc_score(y_test, lr_prob_default)"
   ],
   "execution_count": 41,
   "outputs": [
    {
     "output_type": "stream",
     "text": "0.6010297832058638\n",
     "name": "stdout"
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "<Figure size 432x288 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     }
    }
   ]
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": [
    "# Creating, Training, and Fitting a Balanced Bagging Classfier Model to Oversampled Data"
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "from imblearn.ensemble import BalancedBaggingClassifier\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "\n",
    "# Create an object of the classifier and fit oversampled training data to the object\n",
    "bbc = BalancedBaggingClassifier(base_estimator=DecisionTreeClassifier(),\n",
    "                                sampling_strategy='auto',\n",
    "                                replacement=False,\n",
    "                                random_state=0).fit(oversampled_trainX, np.ravel(oversampled_trainY))\n",
    "\n",
    "# Create predictions of probability for loan status using test data\n",
    "bbc_preds = bbc.predict_proba(X_test)"
   ],
   "execution_count": 42,
   "outputs": []
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "# Create dataframes of predictions and true labels\n",
    "bbc_preds_df = pd.DataFrame(bbc_preds[:,1][0:], columns = ['bbc_pred_PD'])\n",
    "true_df = y_test\n",
    "\n",
    "# Concatenate and print the two data frames for comparison\n",
    "print(pd.concat([true_df.reset_index(drop = True), bbc_preds_df], axis = 1))"
   ],
   "execution_count": 43,
   "outputs": [
    {
     "output_type": "stream",
     "text": "        loan_status  bbc_pred_PD\n0                 0          0.0\n1                 0          0.0\n2                 0          0.0\n3                 0          0.7\n4                 0          0.0\n...             ...          ...\n904263            0          0.0\n904264            1          0.4\n904265            0          0.0\n904266            0          0.0\n904267            0          0.4\n\n[904268 rows x 2 columns]\n",
     "name": "stdout"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "# Reassign loan status based on the threshold and print the predictions\n",
    "bbc_preds_df['bbc_pred_loan_status_60'] = bbc_preds_df['bbc_pred_PD'].apply(lambda x: 1 if x > 0.60 else 0)\n",
    "print(\"Non-Default / Default predictions at 60% Threshhold: \")\n",
    "print(bbc_preds_df['bbc_pred_loan_status_60'].value_counts())\n",
    "\n",
    "# Print the confusion matrix\n",
    "from sklearn.metrics import confusion_matrix\n",
    "print(\"Confusion Matrix at 60% Threshhold: \")\n",
    "print(confusion_matrix(y_test, bbc_preds_df['bbc_pred_loan_status_60']))\n",
    "\n",
    "# Print the classification report\n",
    "from sklearn.metrics import classification_report\n",
    "target_names = ['Non-Default', 'Default']\n",
    "print(classification_report(y_test, bbc_preds_df['bbc_pred_loan_status_60'], target_names=target_names))\n"
   ],
   "execution_count": 44,
   "outputs": [
    {
     "output_type": "stream",
     "text": "Non-Default / Default predictions at 60% Threshhold: \n0    882996\n1     21272\nName: bbc_pred_loan_status_60, dtype: int64\nConfusion Matrix at 60% Threshhold: \n[[769926  15278]\n [113070   5994]]\n              precision    recall  f1-score   support\n\n Non-Default       0.87      0.98      0.92    785204\n     Default       0.28      0.05      0.09    119064\n\n    accuracy                           0.86    904268\n   macro avg       0.58      0.52      0.50    904268\nweighted avg       0.79      0.86      0.81    904268\n\n",
     "name": "stdout"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "# Reassign loan status based on the threshold and print the predictions\n",
    "bbc_preds_df['bbc_pred_loan_status_50'] = bbc_preds_df['bbc_pred_PD'].apply(lambda x: 1 if x > 0.50 else 0)\n",
    "print(\"Non-Default / Default predictions at 50% Threshhold: \")\n",
    "print(bbc_preds_df['bbc_pred_loan_status_50'].value_counts())\n",
    "\n",
    "# Print the confusion matrix\n",
    "from sklearn.metrics import confusion_matrix\n",
    "print(\"Confusion Matrix at 50% Threshhold: \")\n",
    "print(confusion_matrix(y_test, bbc_preds_df['bbc_pred_loan_status_50']))\n",
    "\n",
    "# Print the classification report\n",
    "from sklearn.metrics import classification_report\n",
    "target_names = ['Non-Default', 'Default']\n",
    "print(classification_report(y_test, bbc_preds_df['bbc_pred_loan_status_50'], target_names=target_names))\n"
   ],
   "execution_count": 45,
   "outputs": [
    {
     "output_type": "stream",
     "text": "Non-Default / Default predictions at 50% Threshhold: \n0    861159\n1     43109\nName: bbc_pred_loan_status_50, dtype: int64\nConfusion Matrix at 50% Threshhold: \n[[753502  31702]\n [107657  11407]]\n              precision    recall  f1-score   support\n\n Non-Default       0.87      0.96      0.92    785204\n     Default       0.26      0.10      0.14    119064\n\n    accuracy                           0.85    904268\n   macro avg       0.57      0.53      0.53    904268\nweighted avg       0.79      0.85      0.81    904268\n\n",
     "name": "stdout"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "# Print the accuracy score the model\n",
    "print(bbc.score(X_test, y_test))\n",
    "\n",
    "# Plot the ROC curve of the probabilities of default\n",
    "from sklearn.metrics import roc_curve\n",
    "\n",
    "bbc_prob_default = bbc_preds[:, 1]\n",
    "fallout, sensitivity, thresholds = roc_curve(y_test, bbc_prob_default)\n",
    "plt.plot(fallout, sensitivity, color = 'darkorange')\n",
    "plt.plot([0, 1], [0, 1], linestyle='--')\n",
    "plt.title(\"ROC Chart for BBC on PD\")\n",
    "plt.xlabel(\"Fall-out\")\n",
    "plt.ylabel(\"Sensitivity\")\n",
    "plt.show()\n",
    "\n",
    "# Compute the AUC and store it in a variable\n",
    "from sklearn.metrics import roc_auc_score\n",
    "\n",
    "bbc_auc = roc_auc_score(y_test, bbc_prob_default)"
   ],
   "execution_count": 46,
   "outputs": [
    {
     "output_type": "stream",
     "text": "0.8458875023776137\n",
     "name": "stdout"
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "<Figure size 432x288 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     }
    }
   ]
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": [
    "# Creating, Training, and Fitting a GBT Model to Oversampled Data"
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "# Train a model\n",
    "import xgboost as xgb\n",
    "clf_gbt = xgb.XGBClassifier(tree_method='gpu_hist').fit(oversampled_trainX, np.ravel(oversampled_trainY))\n",
    "\n",
    "# Predict with a model\n",
    "# .predict_proba creates an array of probabilities of default: [[non-default, default]]\n",
    "gbt_preds = clf_gbt.predict_proba(X_test)\n",
    "\n",
    "# Create dataframes of first five predictions, and first five true labels\n",
    "gbt_preds_df = pd.DataFrame(gbt_preds[:,1][0:], columns = ['gbt_pred_PD'])\n",
    "true_df = y_test\n",
    "\n",
    "# Concatenate and print the two data frames for comparison\n",
    "print(pd.concat([true_df.reset_index(drop = True), gbt_preds_df], axis = 1))"
   ],
   "execution_count": 47,
   "outputs": [
    {
     "output_type": "stream",
     "text": "        loan_status  gbt_pred_PD\n0                 0     0.134561\n1                 0     0.013553\n2                 0     0.071101\n3                 0     0.112374\n4                 0     0.046831\n...             ...          ...\n904263            0     0.057313\n904264            1     0.484802\n904265            0     0.043729\n904266            0     0.095952\n904267            0     0.073000\n\n[904268 rows x 2 columns]\n",
     "name": "stdout"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "# Reassign loan status based on the threshold and print the predictions\n",
    "gbt_preds_df['gbt_pred_loan_status_60'] = gbt_preds_df['gbt_pred_PD'].apply(lambda x: 1 if x > 0.60 else 0)\n",
    "print(\"Non-Default / Default  predictions at 60% Threshhold: \")\n",
    "print(gbt_preds_df['gbt_pred_loan_status_60'].value_counts())\n",
    "\n",
    "# Print the confusion matrix\n",
    "print(\"Confusion Matrix at 60% Threshhold: \")\n",
    "print(confusion_matrix(y_test, gbt_preds_df['gbt_pred_loan_status_60']))\n",
    "\n",
    "# Print the classification report\n",
    "target_names = ['Non-Default', 'Default']\n",
    "print(classification_report(y_test, gbt_preds_df['gbt_pred_loan_status_60'], target_names=target_names))"
   ],
   "execution_count": 48,
   "outputs": [
    {
     "output_type": "stream",
     "text": "Non-Default / Default  predictions at 60% Threshhold: \n0    889204\n1     15064\nName: gbt_pred_loan_status_60, dtype: int64\nConfusion Matrix at 60% Threshhold: \n[[774802  10402]\n [114402   4662]]\n              precision    recall  f1-score   support\n\n Non-Default       0.87      0.99      0.93    785204\n     Default       0.31      0.04      0.07    119064\n\n    accuracy                           0.86    904268\n   macro avg       0.59      0.51      0.50    904268\nweighted avg       0.80      0.86      0.81    904268\n\n",
     "name": "stdout"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "# Reassign loan status based on the threshold and print the predictions\n",
    "gbt_preds_df['gbt_pred_loan_status_50'] = gbt_preds_df['gbt_pred_PD'].apply(lambda x: 1 if x > 0.50 else 0)\n",
    "print(\"Non-Default / Default predictions at 50% Threshhold: \")\n",
    "print(gbt_preds_df['gbt_pred_loan_status_50'].value_counts())\n",
    "\n",
    "# Print the confusion matrix\n",
    "print(\"Confusion Matrix at 50% Threshhold: \")\n",
    "print(confusion_matrix(y_test, gbt_preds_df['gbt_pred_loan_status_50']))\n",
    "\n",
    "# Print the classification report\n",
    "target_names = ['Non-Default', 'Default']\n",
    "print(classification_report(y_test, gbt_preds_df['gbt_pred_loan_status_50'], target_names=target_names))"
   ],
   "execution_count": 49,
   "outputs": [
    {
     "output_type": "stream",
     "text": "Non-Default / Default predictions at 50% Threshhold: \n0    863459\n1     40809\nName: gbt_pred_loan_status_50, dtype: int64\nConfusion Matrix at 50% Threshhold: \n[[757276  27928]\n [106183  12881]]\n              precision    recall  f1-score   support\n\n Non-Default       0.88      0.96      0.92    785204\n     Default       0.32      0.11      0.16    119064\n\n    accuracy                           0.85    904268\n   macro avg       0.60      0.54      0.54    904268\nweighted avg       0.80      0.85      0.82    904268\n\n",
     "name": "stdout"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "# Print the accuracy score the model\n",
    "print(clf_gbt.score(X_test, y_test))\n",
    "\n",
    "# Plot the ROC curve of the probabilities of default\n",
    "from sklearn.metrics import roc_curve\n",
    "\n",
    "xgb_prob_default = gbt_preds[:, 1]\n",
    "fallout, sensitivity, thresholds = roc_curve(y_test, xgb_prob_default)\n",
    "plt.plot(fallout, sensitivity, color = 'darkorange')\n",
    "plt.plot([0, 1], [0, 1], linestyle='--')\n",
    "plt.title(\"ROC Chart for XGB on PD\")\n",
    "plt.xlabel(\"Fall-out\")\n",
    "plt.ylabel(\"Sensitivity\")\n",
    "plt.show()\n",
    "\n",
    "# Compute the AUC and store it in a variable\n",
    "from sklearn.metrics import roc_auc_score\n",
    "\n",
    "xgb_auc = roc_auc_score(y_test, xgb_prob_default)\n"
   ],
   "execution_count": 50,
   "outputs": [
    {
     "output_type": "stream",
     "text": "0.8516910915790451\n",
     "name": "stdout"
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "<Figure size 432x288 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     }
    }
   ]
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": [
    "# Modeling Five C's Portfolio Expected Loss using Balanced Bagging Classifier and XGB Classfier predicted probabilities of default"
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "# Creating a portfolio datafram\n",
    "portfolio_5c = pd.DataFrame(gbt_preds[:,1][0:], columns = ['gbt_prob_default'])\n",
    "portfolio_5c.index = X_test.index\n",
    "portfolio_5c['bbc_prob_default'] = bbc_preds[:,1][0:]\n",
    "portfolio_5c['lr_prob_default'] = lr_preds[:,1][0:]\n",
    "portfolio_5c['lgd'] = 1 # Assumes that given a default, entire loan is a loss with no recoverable amounts\n",
    "portfolio_5c['loan_amnt'] = X_test.loan_amnt\n",
    "portfolio_5c.head()"
   ],
   "execution_count": 51,
   "outputs": [
    {
     "output_type": "execute_result",
     "execution_count": 51,
     "data": {
      "text/plain": "         gbt_prob_default  bbc_prob_default  lr_prob_default  lgd  loan_amnt\n739904           0.134561               0.0         0.494650    1       6000\n798388           0.013553               0.0         0.438002    1       7000\n96237            0.071101               0.0         0.516289    1      15000\n1749328          0.112374               0.7         0.550872    1      24000\n1849506          0.046831               0.0         0.438621    1       6500",
      "text/html": "<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>gbt_prob_default</th>\n      <th>bbc_prob_default</th>\n      <th>lr_prob_default</th>\n      <th>lgd</th>\n      <th>loan_amnt</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>739904</th>\n      <td>0.134561</td>\n      <td>0.0</td>\n      <td>0.494650</td>\n      <td>1</td>\n      <td>6000</td>\n    </tr>\n    <tr>\n      <th>798388</th>\n      <td>0.013553</td>\n      <td>0.0</td>\n      <td>0.438002</td>\n      <td>1</td>\n      <td>7000</td>\n    </tr>\n    <tr>\n      <th>96237</th>\n      <td>0.071101</td>\n      <td>0.0</td>\n      <td>0.516289</td>\n      <td>1</td>\n      <td>15000</td>\n    </tr>\n    <tr>\n      <th>1749328</th>\n      <td>0.112374</td>\n      <td>0.7</td>\n      <td>0.550872</td>\n      <td>1</td>\n      <td>24000</td>\n    </tr>\n    <tr>\n      <th>1849506</th>\n      <td>0.046831</td>\n      <td>0.0</td>\n      <td>0.438621</td>\n      <td>1</td>\n      <td>6500</td>\n    </tr>\n  </tbody>\n</table>\n</div>"
     },
     "metadata": {}
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "# Create expected loss columns for each model using the formula\n",
    "portfolio_5c['gbt_expected_loss'] = portfolio_5c['gbt_prob_default'] * portfolio_5c['lgd'] * portfolio_5c['loan_amnt']\n",
    "portfolio_5c['bbc_expected_loss'] = portfolio_5c['bbc_prob_default'] * portfolio_5c['lgd'] * portfolio_5c['loan_amnt']\n",
    "portfolio_5c['lr_expected_loss'] = portfolio_5c['lr_prob_default'] * portfolio_5c['lgd'] * portfolio_5c['loan_amnt']\n",
    "\n",
    "# Print the total portfolio size\n",
    "print('Portfolio size: $' + \"{:,.2f}\".format(np.sum(portfolio_5c['loan_amnt'])))\n",
    "\n",
    "# Print the sum of the expected loss for bbc\n",
    "print('BBC expected loss: $' + \"{:,.2f}\".format(np.sum(portfolio_5c['bbc_expected_loss'])))\n",
    "\n",
    "# Print the sum of the expected loss for gbt\n",
    "print('GBT expected loss: $' + \"{:,.2f}\".format(np.sum(portfolio_5c['gbt_expected_loss'])))\n",
    "\n",
    "# Print the sum of the expected loss for lr \n",
    "print('LR expected loss: $' + \"{:,.2f}\".format(np.sum(portfolio_5c['lr_expected_loss'])))\n",
    "\n",
    "# Print portfolio first five rows\n",
    "portfolio_5c.head()"
   ],
   "execution_count": 52,
   "outputs": [
    {
     "output_type": "stream",
     "text": "Portfolio size: $13,613,538,150.00\nBBC expected loss: $2,281,645,884.17\nGBT expected loss: $2,849,133,626.67\nLR expected loss: $6,914,929,459.48\n",
     "name": "stdout"
    },
    {
     "output_type": "execute_result",
     "execution_count": 52,
     "data": {
      "text/plain": "         gbt_prob_default  bbc_prob_default  lr_prob_default  lgd  loan_amnt  \\\n739904           0.134561               0.0         0.494650    1       6000   \n798388           0.013553               0.0         0.438002    1       7000   \n96237            0.071101               0.0         0.516289    1      15000   \n1749328          0.112374               0.7         0.550872    1      24000   \n1849506          0.046831               0.0         0.438621    1       6500   \n\n         gbt_expected_loss  bbc_expected_loss  lr_expected_loss  \n739904          807.364494                0.0       2967.901671  \n798388           94.872819                0.0       3066.016730  \n96237          1066.514030                0.0       7744.334095  \n1749328        2696.973860            16800.0      13220.921515  \n1849506         304.403611                0.0       2851.037149  ",
      "text/html": "<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>gbt_prob_default</th>\n      <th>bbc_prob_default</th>\n      <th>lr_prob_default</th>\n      <th>lgd</th>\n      <th>loan_amnt</th>\n      <th>gbt_expected_loss</th>\n      <th>bbc_expected_loss</th>\n      <th>lr_expected_loss</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>739904</th>\n      <td>0.134561</td>\n      <td>0.0</td>\n      <td>0.494650</td>\n      <td>1</td>\n      <td>6000</td>\n      <td>807.364494</td>\n      <td>0.0</td>\n      <td>2967.901671</td>\n    </tr>\n    <tr>\n      <th>798388</th>\n      <td>0.013553</td>\n      <td>0.0</td>\n      <td>0.438002</td>\n      <td>1</td>\n      <td>7000</td>\n      <td>94.872819</td>\n      <td>0.0</td>\n      <td>3066.016730</td>\n    </tr>\n    <tr>\n      <th>96237</th>\n      <td>0.071101</td>\n      <td>0.0</td>\n      <td>0.516289</td>\n      <td>1</td>\n      <td>15000</td>\n      <td>1066.514030</td>\n      <td>0.0</td>\n      <td>7744.334095</td>\n    </tr>\n    <tr>\n      <th>1749328</th>\n      <td>0.112374</td>\n      <td>0.7</td>\n      <td>0.550872</td>\n      <td>1</td>\n      <td>24000</td>\n      <td>2696.973860</td>\n      <td>16800.0</td>\n      <td>13220.921515</td>\n    </tr>\n    <tr>\n      <th>1849506</th>\n      <td>0.046831</td>\n      <td>0.0</td>\n      <td>0.438621</td>\n      <td>1</td>\n      <td>6500</td>\n      <td>304.403611</td>\n      <td>0.0</td>\n      <td>2851.037149</td>\n    </tr>\n  </tbody>\n</table>\n</div>"
     },
     "metadata": {}
    }
   ]
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": [
    "# TIME TO DEFAULT PREDICTION\n",
    "\n",
    "We can use the following features contained in the raw data to predict time to default:\n",
    "\n",
    "*'emp_length', 'chargeoff_within_12_mths', 'mths_since_last_delinq', 'mths_since_last_record', 'mths_since_last_major_derog', 'mths_since_recent_revol_delinq'  *\n",
    "\n",
    "These feaure provide an indication of time in months since a negative credit event (i.e. an event that has a negative impact on a customers credit history such as late payment, bankruptcy, loan default, civil lawsuit, etc).\n",
    "\n",
    "These features will be used with the processed_df features to model time to default using Kaplan Meier Estimates and Cox Proportional Hazard Models."
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "preprocess_TTD_df = raw_data[['emp_length', 'chargeoff_within_12_mths',\n",
    "                              'mths_since_last_delinq', 'mths_since_last_record',\n",
    "                              'mths_since_last_major_derog', 'mths_since_recent_revol_delinq']].copy()"
   ],
   "execution_count": 144,
   "outputs": []
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "preprocess_TTD_df.head()"
   ],
   "execution_count": 145,
   "outputs": [
    {
     "output_type": "execute_result",
     "execution_count": 145,
     "data": {
      "text/plain": "  emp_length  chargeoff_within_12_mths  mths_since_last_delinq  \\\n0  10+ years                       0.0                     NaN   \n1  10+ years                       0.0                    71.0   \n2    6 years                       0.0                     NaN   \n3  10+ years                       0.0                     NaN   \n4  10+ years                       0.0                     NaN   \n\n   mths_since_last_record  mths_since_last_major_derog  \\\n0                    45.0                          NaN   \n1                    75.0                          NaN   \n2                     NaN                          NaN   \n3                     NaN                          NaN   \n4                     NaN                          NaN   \n\n   mths_since_recent_revol_delinq  \n0                             NaN  \n1                             NaN  \n2                             NaN  \n3                             NaN  \n4                             NaN  ",
      "text/html": "<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>emp_length</th>\n      <th>chargeoff_within_12_mths</th>\n      <th>mths_since_last_delinq</th>\n      <th>mths_since_last_record</th>\n      <th>mths_since_last_major_derog</th>\n      <th>mths_since_recent_revol_delinq</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>10+ years</td>\n      <td>0.0</td>\n      <td>NaN</td>\n      <td>45.0</td>\n      <td>NaN</td>\n      <td>NaN</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>10+ years</td>\n      <td>0.0</td>\n      <td>71.0</td>\n      <td>75.0</td>\n      <td>NaN</td>\n      <td>NaN</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>6 years</td>\n      <td>0.0</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>10+ years</td>\n      <td>0.0</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>10+ years</td>\n      <td>0.0</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>"
     },
     "metadata": {}
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "preprocess_TTD_df.isnull().sum()"
   ],
   "execution_count": 146,
   "outputs": [
    {
     "output_type": "execute_result",
     "execution_count": 146,
     "data": {
      "text/plain": "emp_length                         146907\nchargeoff_within_12_mths              145\nmths_since_last_delinq            1158502\nmths_since_last_record            1901512\nmths_since_last_major_derog       1679893\nmths_since_recent_revol_delinq    1520309\ndtype: int64"
     },
     "metadata": {}
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "preprocess_TTD_df.emp_length.value_counts()"
   ],
   "execution_count": 147,
   "outputs": [
    {
     "output_type": "execute_result",
     "execution_count": 147,
     "data": {
      "text/plain": "10+ years    748005\n2 years      203677\n< 1 year     189988\n3 years      180753\n1 year       148403\n5 years      139698\n4 years      136605\n6 years      102628\n7 years       92695\n8 years       91914\n9 years       79395\nName: emp_length, dtype: int64"
     },
     "metadata": {}
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "# fill in missing values with the most frequently occuring employment length\n",
    "preprocess_TTD_df['emp_length'].fillna(preprocess_TTD_df['emp_length'].value_counts().index[0], inplace=True)\n",
    "\n",
    "# create a dataframe of employment length for customers in dataset\n",
    "emp_length_df = pd.DataFrame(pd.get_dummies(preprocess_TTD_df.emp_length, prefix='emp_length'))\n",
    "emp_length_df.head()\n",
    "\n",
    "# concat the emp_length dataframe with preprocess_TTD_df, dropping original columns\n",
    "preprocess_TTD_df = pd.concat([preprocess_TTD_df, emp_length_df], axis=1, sort=False)\n",
    "preprocess_TTD_df.drop(['emp_length'], axis=1, inplace=True)\n",
    "preprocess_TTD_df.head()"
   ],
   "execution_count": 149,
   "outputs": [
    {
     "output_type": "execute_result",
     "execution_count": 149,
     "data": {
      "text/plain": "   chargeoff_within_12_mths  mths_since_last_delinq  mths_since_last_record  \\\n0                       0.0                     NaN                    45.0   \n1                       0.0                    71.0                    75.0   \n2                       0.0                     NaN                     NaN   \n3                       0.0                     NaN                     NaN   \n4                       0.0                     NaN                     NaN   \n\n   mths_since_last_major_derog  mths_since_recent_revol_delinq  \\\n0                          NaN                             NaN   \n1                          NaN                             NaN   \n2                          NaN                             NaN   \n3                          NaN                             NaN   \n4                          NaN                             NaN   \n\n   emp_length_1 year  emp_length_10+ years  emp_length_2 years  \\\n0                  0                     1                   0   \n1                  0                     1                   0   \n2                  0                     0                   0   \n3                  0                     1                   0   \n4                  0                     1                   0   \n\n   emp_length_3 years  emp_length_4 years  emp_length_5 years  \\\n0                   0                   0                   0   \n1                   0                   0                   0   \n2                   0                   0                   0   \n3                   0                   0                   0   \n4                   0                   0                   0   \n\n   emp_length_6 years  emp_length_7 years  emp_length_8 years  \\\n0                   0                   0                   0   \n1                   0                   0                   0   \n2                   1                   0                   0   \n3                   0                   0                   0   \n4                   0                   0                   0   \n\n   emp_length_9 years  emp_length_< 1 year  \n0                   0                    0  \n1                   0                    0  \n2                   0                    0  \n3                   0                    0  \n4                   0                    0  ",
      "text/html": "<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>chargeoff_within_12_mths</th>\n      <th>mths_since_last_delinq</th>\n      <th>mths_since_last_record</th>\n      <th>mths_since_last_major_derog</th>\n      <th>mths_since_recent_revol_delinq</th>\n      <th>emp_length_1 year</th>\n      <th>emp_length_10+ years</th>\n      <th>emp_length_2 years</th>\n      <th>emp_length_3 years</th>\n      <th>emp_length_4 years</th>\n      <th>emp_length_5 years</th>\n      <th>emp_length_6 years</th>\n      <th>emp_length_7 years</th>\n      <th>emp_length_8 years</th>\n      <th>emp_length_9 years</th>\n      <th>emp_length_&lt; 1 year</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>0.0</td>\n      <td>NaN</td>\n      <td>45.0</td>\n      <td>NaN</td>\n      <td>NaN</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>0.0</td>\n      <td>71.0</td>\n      <td>75.0</td>\n      <td>NaN</td>\n      <td>NaN</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>0.0</td>\n      <td>NaN</td>\n      <td>NaN</td>\n      <td>NaN</td>\n      <td>NaN</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>0.0</td>\n      <td>NaN</td>\n      <td>NaN</td>\n      <td>NaN</td>\n      <td>NaN</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>0.0</td>\n      <td>NaN</td>\n      <td>NaN</td>\n      <td>NaN</td>\n      <td>NaN</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n    </tr>\n  </tbody>\n</table>\n</div>"
     },
     "metadata": {}
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "# fill in missing values with the most frequently occuring chargeoff\n",
    "preprocess_TTD_df['chargeoff_within_12_mths'].fillna(preprocess_TTD_df['chargeoff_within_12_mths'].value_counts().index[0], inplace=True)"
   ],
   "execution_count": 150,
   "outputs": []
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": [
    "# mths_since_last_delinq, mths_since_last_record, mths_since_last_major_derog, mths_since_recent_revol_delinq\n",
    "\n",
    "LC Data Dictionary defines \n",
    "\n",
    "\n",
    "* <b>mths_since_last_delinq:</b> The number of months since the borrower's last delinquency.\n",
    "* <b>mthsSinceLastRecord:</b> The number of months since the last public record.\n",
    "* <b>mths_since_last_major_derog:</b> Months since most recent 90-day or worse rating\n",
    "* <b>mths_since_recent_revol_delinq:</b> Months since most recent revolving delinquency\n",
    "\n",
    "\n",
    "Important to note that current consumer credit practices allows for delinquencies to stay on a consumer credit report for 7 years, so while information maybe missing, it does not mean that a delinquency has not occured.\n",
    "\n",
    "With no addtional data given on these features, we will assume that delinquencies, records, major derogatory remarks, and months since last revolving delinquencies have since fallen off the consumer credit reports and can no longer be factored into consideration. \n",
    "\n",
    "Will fill the  missing these features with 0 months"
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "preprocess_TTD_df.mths_since_last_delinq\n",
    "print(preprocess_TTD_df.mths_since_last_delinq.describe())"
   ],
   "execution_count": 151,
   "outputs": [
    {
     "output_type": "stream",
     "text": "count    1.102166e+06\nmean     3.454092e+01\nstd      2.190047e+01\nmin      0.000000e+00\n25%      1.600000e+01\n50%      3.100000e+01\n75%      5.000000e+01\nmax      2.260000e+02\nName: mths_since_last_delinq, dtype: float64\n",
     "name": "stdout"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "preprocess_TTD_df.mths_since_last_record \n",
    "print(preprocess_TTD_df.mths_since_last_record .describe())"
   ],
   "execution_count": 152,
   "outputs": [
    {
     "output_type": "stream",
     "text": "count    359156.000000\nmean         72.312842\nstd          26.464094\nmin           0.000000\n25%          55.000000\n50%          74.000000\n75%          92.000000\nmax         129.000000\nName: mths_since_last_record, dtype: float64\n",
     "name": "stdout"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "preprocess_TTD_df.mths_since_last_major_derog\n",
    "print(preprocess_TTD_df.mths_since_last_major_derog.describe())"
   ],
   "execution_count": 153,
   "outputs": [
    {
     "output_type": "stream",
     "text": "count    580775.000000\nmean         44.164220\nstd          21.533121\nmin           0.000000\n25%          27.000000\n50%          44.000000\n75%          62.000000\nmax         226.000000\nName: mths_since_last_major_derog, dtype: float64\n",
     "name": "stdout"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "preprocess_TTD_df.mths_since_recent_revol_delinq\n",
    "print(preprocess_TTD_df.mths_since_recent_revol_delinq.describe())"
   ],
   "execution_count": 154,
   "outputs": [
    {
     "output_type": "stream",
     "text": "count    740359.000000\nmean         35.782223\nstd          22.307239\nmin           0.000000\n25%          17.000000\n50%          33.000000\n75%          51.000000\nmax         202.000000\nName: mths_since_recent_revol_delinq, dtype: float64\n",
     "name": "stdout"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "preprocess_TTD_df.fillna(0, inplace=True)"
   ],
   "execution_count": 155,
   "outputs": []
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "preprocess_TTD_df"
   ],
   "execution_count": 156,
   "outputs": [
    {
     "output_type": "execute_result",
     "execution_count": 156,
     "data": {
      "text/plain": "         chargeoff_within_12_mths  mths_since_last_delinq  \\\n0                             0.0                     0.0   \n1                             0.0                    71.0   \n2                             0.0                     0.0   \n3                             0.0                     0.0   \n4                             0.0                     0.0   \n...                           ...                     ...   \n2260663                       0.0                     0.0   \n2260664                       0.0                    13.0   \n2260665                       0.0                     0.0   \n2260666                       0.0                     7.0   \n2260667                       0.0                     0.0   \n\n         mths_since_last_record  mths_since_last_major_derog  \\\n0                          45.0                          0.0   \n1                          75.0                          0.0   \n2                           0.0                          0.0   \n3                           0.0                          0.0   \n4                           0.0                          0.0   \n...                         ...                          ...   \n2260663                     0.0                          0.0   \n2260664                     0.0                         46.0   \n2260665                     0.0                          0.0   \n2260666                     0.0                          0.0   \n2260667                     0.0                          0.0   \n\n         mths_since_recent_revol_delinq  emp_length_1 year  \\\n0                                   0.0                  0   \n1                                   0.0                  0   \n2                                   0.0                  0   \n3                                   0.0                  0   \n4                                   0.0                  0   \n...                                 ...                ...   \n2260663                             0.0                  0   \n2260664                            17.0                  0   \n2260665                             0.0                  0   \n2260666                             7.0                  0   \n2260667                             0.0                  0   \n\n         emp_length_10+ years  emp_length_2 years  emp_length_3 years  \\\n0                           1                   0                   0   \n1                           1                   0                   0   \n2                           0                   0                   0   \n3                           1                   0                   0   \n4                           1                   0                   0   \n...                       ...                 ...                 ...   \n2260663                     1                   0                   0   \n2260664                     0                   0                   0   \n2260665                     1                   0                   0   \n2260666                     1                   0                   0   \n2260667                     0                   0                   1   \n\n         emp_length_4 years  emp_length_5 years  emp_length_6 years  \\\n0                         0                   0                   0   \n1                         0                   0                   0   \n2                         0                   0                   1   \n3                         0                   0                   0   \n4                         0                   0                   0   \n...                     ...                 ...                 ...   \n2260663                   0                   0                   0   \n2260664                   0                   0                   0   \n2260665                   0                   0                   0   \n2260666                   0                   0                   0   \n2260667                   0                   0                   0   \n\n         emp_length_7 years  emp_length_8 years  emp_length_9 years  \\\n0                         0                   0                   0   \n1                         0                   0                   0   \n2                         0                   0                   0   \n3                         0                   0                   0   \n4                         0                   0                   0   \n...                     ...                 ...                 ...   \n2260663                   0                   0                   0   \n2260664                   0                   0                   0   \n2260665                   0                   0                   0   \n2260666                   0                   0                   0   \n2260667                   0                   0                   0   \n\n         emp_length_< 1 year  \n0                          0  \n1                          0  \n2                          0  \n3                          0  \n4                          0  \n...                      ...  \n2260663                    0  \n2260664                    1  \n2260665                    0  \n2260666                    0  \n2260667                    0  \n\n[2260668 rows x 16 columns]",
      "text/html": "<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>chargeoff_within_12_mths</th>\n      <th>mths_since_last_delinq</th>\n      <th>mths_since_last_record</th>\n      <th>mths_since_last_major_derog</th>\n      <th>mths_since_recent_revol_delinq</th>\n      <th>emp_length_1 year</th>\n      <th>emp_length_10+ years</th>\n      <th>emp_length_2 years</th>\n      <th>emp_length_3 years</th>\n      <th>emp_length_4 years</th>\n      <th>emp_length_5 years</th>\n      <th>emp_length_6 years</th>\n      <th>emp_length_7 years</th>\n      <th>emp_length_8 years</th>\n      <th>emp_length_9 years</th>\n      <th>emp_length_&lt; 1 year</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>45.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>0.0</td>\n      <td>71.0</td>\n      <td>75.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>...</th>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n    </tr>\n    <tr>\n      <th>2260663</th>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>2260664</th>\n      <td>0.0</td>\n      <td>13.0</td>\n      <td>0.0</td>\n      <td>46.0</td>\n      <td>17.0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>1</td>\n    </tr>\n    <tr>\n      <th>2260665</th>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>2260666</th>\n      <td>0.0</td>\n      <td>7.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>7.0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>2260667</th>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n    </tr>\n  </tbody>\n</table>\n<p>2260668 rows × 16 columns</p>\n</div>"
     },
     "metadata": {}
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "preprocess_TTD_df.isnull().sum()"
   ],
   "execution_count": 157,
   "outputs": [
    {
     "output_type": "execute_result",
     "execution_count": 157,
     "data": {
      "text/plain": "chargeoff_within_12_mths          0\nmths_since_last_delinq            0\nmths_since_last_record            0\nmths_since_last_major_derog       0\nmths_since_recent_revol_delinq    0\nemp_length_1 year                 0\nemp_length_10+ years              0\nemp_length_2 years                0\nemp_length_3 years                0\nemp_length_4 years                0\nemp_length_5 years                0\nemp_length_6 years                0\nemp_length_7 years                0\nemp_length_8 years                0\nemp_length_9 years                0\nemp_length_< 1 year               0\ndtype: int64"
     },
     "metadata": {}
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "process_TTD_df = preprocess_TTD_df.copy()\n",
    "process_TTD_df"
   ],
   "execution_count": 158,
   "outputs": [
    {
     "output_type": "execute_result",
     "execution_count": 158,
     "data": {
      "text/plain": "         chargeoff_within_12_mths  mths_since_last_delinq  \\\n0                             0.0                     0.0   \n1                             0.0                    71.0   \n2                             0.0                     0.0   \n3                             0.0                     0.0   \n4                             0.0                     0.0   \n...                           ...                     ...   \n2260663                       0.0                     0.0   \n2260664                       0.0                    13.0   \n2260665                       0.0                     0.0   \n2260666                       0.0                     7.0   \n2260667                       0.0                     0.0   \n\n         mths_since_last_record  mths_since_last_major_derog  \\\n0                          45.0                          0.0   \n1                          75.0                          0.0   \n2                           0.0                          0.0   \n3                           0.0                          0.0   \n4                           0.0                          0.0   \n...                         ...                          ...   \n2260663                     0.0                          0.0   \n2260664                     0.0                         46.0   \n2260665                     0.0                          0.0   \n2260666                     0.0                          0.0   \n2260667                     0.0                          0.0   \n\n         mths_since_recent_revol_delinq  emp_length_1 year  \\\n0                                   0.0                  0   \n1                                   0.0                  0   \n2                                   0.0                  0   \n3                                   0.0                  0   \n4                                   0.0                  0   \n...                                 ...                ...   \n2260663                             0.0                  0   \n2260664                            17.0                  0   \n2260665                             0.0                  0   \n2260666                             7.0                  0   \n2260667                             0.0                  0   \n\n         emp_length_10+ years  emp_length_2 years  emp_length_3 years  \\\n0                           1                   0                   0   \n1                           1                   0                   0   \n2                           0                   0                   0   \n3                           1                   0                   0   \n4                           1                   0                   0   \n...                       ...                 ...                 ...   \n2260663                     1                   0                   0   \n2260664                     0                   0                   0   \n2260665                     1                   0                   0   \n2260666                     1                   0                   0   \n2260667                     0                   0                   1   \n\n         emp_length_4 years  emp_length_5 years  emp_length_6 years  \\\n0                         0                   0                   0   \n1                         0                   0                   0   \n2                         0                   0                   1   \n3                         0                   0                   0   \n4                         0                   0                   0   \n...                     ...                 ...                 ...   \n2260663                   0                   0                   0   \n2260664                   0                   0                   0   \n2260665                   0                   0                   0   \n2260666                   0                   0                   0   \n2260667                   0                   0                   0   \n\n         emp_length_7 years  emp_length_8 years  emp_length_9 years  \\\n0                         0                   0                   0   \n1                         0                   0                   0   \n2                         0                   0                   0   \n3                         0                   0                   0   \n4                         0                   0                   0   \n...                     ...                 ...                 ...   \n2260663                   0                   0                   0   \n2260664                   0                   0                   0   \n2260665                   0                   0                   0   \n2260666                   0                   0                   0   \n2260667                   0                   0                   0   \n\n         emp_length_< 1 year  \n0                          0  \n1                          0  \n2                          0  \n3                          0  \n4                          0  \n...                      ...  \n2260663                    0  \n2260664                    1  \n2260665                    0  \n2260666                    0  \n2260667                    0  \n\n[2260668 rows x 16 columns]",
      "text/html": "<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>chargeoff_within_12_mths</th>\n      <th>mths_since_last_delinq</th>\n      <th>mths_since_last_record</th>\n      <th>mths_since_last_major_derog</th>\n      <th>mths_since_recent_revol_delinq</th>\n      <th>emp_length_1 year</th>\n      <th>emp_length_10+ years</th>\n      <th>emp_length_2 years</th>\n      <th>emp_length_3 years</th>\n      <th>emp_length_4 years</th>\n      <th>emp_length_5 years</th>\n      <th>emp_length_6 years</th>\n      <th>emp_length_7 years</th>\n      <th>emp_length_8 years</th>\n      <th>emp_length_9 years</th>\n      <th>emp_length_&lt; 1 year</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>45.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>0.0</td>\n      <td>71.0</td>\n      <td>75.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>...</th>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n    </tr>\n    <tr>\n      <th>2260663</th>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>2260664</th>\n      <td>0.0</td>\n      <td>13.0</td>\n      <td>0.0</td>\n      <td>46.0</td>\n      <td>17.0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>1</td>\n    </tr>\n    <tr>\n      <th>2260665</th>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>2260666</th>\n      <td>0.0</td>\n      <td>7.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>7.0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>2260667</th>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n    </tr>\n  </tbody>\n</table>\n<p>2260668 rows × 16 columns</p>\n</div>"
     },
     "metadata": {}
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "# create a dataframe of employment length for customers in dataset\n",
    "grade_df = pd.DataFrame(pd.get_dummies(processed_df.grade, prefix='grade'))\n",
    "grade_df.head()\n",
    "\n",
    "# concat the emp_length dataframe with preprocess_TTD_df, dropping original columns\n",
    "processed_df = pd.concat([processed_df, grade_df], axis=1, sort=False)\n",
    "processed_df.drop(['grade'], axis=1, inplace=True)\n",
    "processed_df.head()"
   ],
   "execution_count": 170,
   "outputs": [
    {
     "output_type": "execute_result",
     "execution_count": 170,
     "data": {
      "text/plain": "   credit_hist_in_months  cb_person_bk_on_file_Y  annual_inc    dti  \\\n0                    212                       1     55000.0  18.24   \n1                    378                       1     90000.0  26.52   \n2                     92                       0     59280.0  10.51   \n3                    153                       0     92000.0  16.74   \n4                    215                       0     57250.0  26.35   \n\n   loan_amnt  int_rate  term_60  home_ownership_ANY  home_ownership_MORTGAGE  \\\n0       2500     13.56        0                   0                        0   \n1      30000     18.94        1                   0                        1   \n2       5000     17.97        0                   0                        1   \n3       4000     18.94        0                   0                        1   \n4      30000     16.14        1                   0                        1   \n\n   home_ownership_NONE  ...  home_ownership_OWN  home_ownership_RENT  \\\n0                    0  ...                   0                    1   \n1                    0  ...                   0                    0   \n2                    0  ...                   0                    0   \n3                    0  ...                   0                    0   \n4                    0  ...                   0                    0   \n\n   loan_status  grade_A  grade_B  grade_C  grade_D  grade_E  grade_F  grade_G  \n0            0        0        0        1        0        0        0        0  \n1            0        0        0        0        1        0        0        0  \n2            0        0        0        0        1        0        0        0  \n3            0        0        0        0        1        0        0        0  \n4            0        0        0        1        0        0        0        0  \n\n[5 rows x 21 columns]",
      "text/html": "<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>credit_hist_in_months</th>\n      <th>cb_person_bk_on_file_Y</th>\n      <th>annual_inc</th>\n      <th>dti</th>\n      <th>loan_amnt</th>\n      <th>int_rate</th>\n      <th>term_60</th>\n      <th>home_ownership_ANY</th>\n      <th>home_ownership_MORTGAGE</th>\n      <th>home_ownership_NONE</th>\n      <th>...</th>\n      <th>home_ownership_OWN</th>\n      <th>home_ownership_RENT</th>\n      <th>loan_status</th>\n      <th>grade_A</th>\n      <th>grade_B</th>\n      <th>grade_C</th>\n      <th>grade_D</th>\n      <th>grade_E</th>\n      <th>grade_F</th>\n      <th>grade_G</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>212</td>\n      <td>1</td>\n      <td>55000.0</td>\n      <td>18.24</td>\n      <td>2500</td>\n      <td>13.56</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>...</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>378</td>\n      <td>1</td>\n      <td>90000.0</td>\n      <td>26.52</td>\n      <td>30000</td>\n      <td>18.94</td>\n      <td>1</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>...</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>92</td>\n      <td>0</td>\n      <td>59280.0</td>\n      <td>10.51</td>\n      <td>5000</td>\n      <td>17.97</td>\n      <td>0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>...</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>153</td>\n      <td>0</td>\n      <td>92000.0</td>\n      <td>16.74</td>\n      <td>4000</td>\n      <td>18.94</td>\n      <td>0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>...</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>215</td>\n      <td>0</td>\n      <td>57250.0</td>\n      <td>26.35</td>\n      <td>30000</td>\n      <td>16.14</td>\n      <td>1</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>...</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n    </tr>\n  </tbody>\n</table>\n<p>5 rows × 21 columns</p>\n</div>"
     },
     "metadata": {}
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "final_processed = processed_df.join(process_TTD_df)"
   ],
   "execution_count": 171,
   "outputs": []
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "final_processed.columns"
   ],
   "execution_count": 172,
   "outputs": [
    {
     "output_type": "execute_result",
     "execution_count": 172,
     "data": {
      "text/plain": "Index(['credit_hist_in_months', 'cb_person_bk_on_file_Y', 'annual_inc', 'dti',\n       'loan_amnt', 'int_rate', 'term_60', 'home_ownership_ANY',\n       'home_ownership_MORTGAGE', 'home_ownership_NONE',\n       'home_ownership_OTHER', 'home_ownership_OWN', 'home_ownership_RENT',\n       'loan_status', 'grade_A', 'grade_B', 'grade_C', 'grade_D', 'grade_E',\n       'grade_F', 'grade_G', 'chargeoff_within_12_mths',\n       'mths_since_last_delinq', 'mths_since_last_record',\n       'mths_since_last_major_derog', 'mths_since_recent_revol_delinq',\n       'emp_length_1 year', 'emp_length_10+ years', 'emp_length_2 years',\n       'emp_length_3 years', 'emp_length_4 years', 'emp_length_5 years',\n       'emp_length_6 years', 'emp_length_7 years', 'emp_length_8 years',\n       'emp_length_9 years', 'emp_length_< 1 year'],\n      dtype='object')"
     },
     "metadata": {}
    }
   ]
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": [
    "# Survival Analysis with KMF"
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": [
    "# Go to settings tab and turn internet on to download lifelines and instantiate KMF\n",
    "\n",
    "!pip install lifelines\n",
    "from lifelines import KaplanMeierFitter"
   ],
   "execution_count": 169,
   "outputs": [
    {
     "output_type": "stream",
     "text": "Collecting lifelines\n  Downloading lifelines-0.24.6-py3-none-any.whl (326 kB)\n\u001B[K     |████████████████████████████████| 326 kB 7.9 MB/s eta 0:00:01\n\u001B[?25hRequirement already satisfied: scipy>=1.2.0 in /opt/conda/lib/python3.6/site-packages (from lifelines) (1.4.1)\nCollecting autograd>=1.3\n  Downloading autograd-1.3.tar.gz (38 kB)\nCollecting autograd-gamma>=0.3\n  Downloading autograd_gamma-0.4.2-py2.py3-none-any.whl (3.8 kB)\nRequirement already satisfied: pandas>=0.23.0 in /opt/conda/lib/python3.6/site-packages (from lifelines) (0.25.3)\nRequirement already satisfied: matplotlib>=3.0 in /opt/conda/lib/python3.6/site-packages (from lifelines) (3.0.3)\nRequirement already satisfied: numpy>=1.14.0 in /opt/conda/lib/python3.6/site-packages (from lifelines) (1.18.1)\nRequirement already satisfied: future>=0.15.2 in /opt/conda/lib/python3.6/site-packages (from autograd>=1.3->lifelines) (0.18.2)\nRequirement already satisfied: pytz>=2017.2 in /opt/conda/lib/python3.6/site-packages (from pandas>=0.23.0->lifelines) (2019.3)\nRequirement already satisfied: python-dateutil>=2.6.1 in /opt/conda/lib/python3.6/site-packages (from pandas>=0.23.0->lifelines) (2.8.0)\nRequirement already satisfied: cycler>=0.10 in /opt/conda/lib/python3.6/site-packages (from matplotlib>=3.0->lifelines) (0.10.0)\nRequirement already satisfied: kiwisolver>=1.0.1 in /opt/conda/lib/python3.6/site-packages (from matplotlib>=3.0->lifelines) (1.1.0)\nRequirement already satisfied: pyparsing!=2.0.4,!=2.1.2,!=2.1.6,>=2.0.1 in /opt/conda/lib/python3.6/site-packages (from matplotlib>=3.0->lifelines) (2.4.6)\nRequirement already satisfied: six>=1.5 in /opt/conda/lib/python3.6/site-packages (from python-dateutil>=2.6.1->pandas>=0.23.0->lifelines) (1.14.0)\nRequirement already satisfied: setuptools in /opt/conda/lib/python3.6/site-packages (from kiwisolver>=1.0.1->matplotlib>=3.0->lifelines) (45.2.0.post20200210)\nBuilding wheels for collected packages: autograd\n  Building wheel for autograd (setup.py) ... \u001B[?25ldone\n\u001B[?25h  Created wheel for autograd: filename=autograd-1.3-py3-none-any.whl size=47990 sha256=14a9d7d54cadd9b9259c3817b03955b1ee44464aee57674eea805a3b98dd04ad\n  Stored in directory: /root/.cache/pip/wheels/b2/a5/3c/929b91003a3b75a175ccba7cef35200a1890cdb46903f18072\nSuccessfully built autograd\nInstalling collected packages: autograd, autograd-gamma, lifelines\nSuccessfully installed autograd-1.3 autograd-gamma-0.4.2 lifelines-0.24.6\n",
     "name": "stdout"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "raw_data.grade.value_counts()",
   "execution_count": 190,
   "outputs": [
    {
     "output_type": "execute_result",
     "execution_count": 190,
     "data": {
      "text/plain": "B    663557\nC    650053\nA    433027\nD    324424\nE    135639\nF     41800\nG     12168\nName: grade, dtype: int64"
     },
     "metadata": {}
    }
   ]
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": "# Instantiate KMF object"
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "kmf1 = KaplanMeierFitter() ## instantiate the class to create an object\n\ntenure = final_processed['credit_hist_in_months']\nevent = final_processed['loan_status']",
   "execution_count": null,
   "outputs": []
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": "## Time to Default (or negative credit event) for Grade A rated customers"
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "## Two Cohorts are compared. Cohort 1: Not Grade A Customers, and Cohort  2: Grade A CUstomers\ngrade_A = final_processed['grade_A']    \nnot_grade_A = (grade_A == 0)      ## Cohort Not Grade A Customers, having the pandas series  for the 1st cohort\nyes_grade_A = (grade_A == 1)     ## Cohort Grade A Customers, having the pandas series  for the 2nd cohort\n\n\n## fit the model for 1st cohort\nkmf1.fit(tenure[not_grade_A], event[not_grade_A], label='Not Grade A')\na1 = kmf1.plot()\n\n## fit the model for 2nd cohort\nkmf1.fit(tenure[yes_grade_A], event[yes_grade_A], label='Grade A')\nkmf1.plot(ax=a1)\n# a1.set_xlim(0, 100)",
   "execution_count": 203,
   "outputs": [
    {
     "output_type": "execute_result",
     "execution_count": 203,
     "data": {
      "text/plain": "<matplotlib.axes._subplots.AxesSubplot at 0x7f4e8c5e6400>"
     },
     "metadata": {}
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "<Figure size 432x288 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     }
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "type(a1)",
   "execution_count": 197,
   "outputs": [
    {
     "output_type": "execute_result",
     "execution_count": 197,
     "data": {
      "text/plain": "matplotlib.axes._subplots.AxesSubplot"
     },
     "metadata": {}
    }
   ]
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": "## Time to Default (or negative credit event) for Grade B rated customers"
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "## Two Cohorts are compared. Cohort 1: Not Grade B Customers, and Cohort  2: Grade B CUstomers\ngrade_B = final_processed['grade_B']    \nnot_grade_B = (grade_B == 0)      ## Cohort Not Grade B Customers, having the pandas series  for the 1st cohort\nyes_grade_B = (grade_B == 1)     ## Cohort Grade B Customers, having the pandas series  for the 2nd cohort\n\n\n## fit the model for 1st cohort\nkmf1.fit(tenure[not_grade_B], event[not_grade_B], label='Not Grade B')\na1 = kmf1.plot()\n\n## fit the model for 2nd cohort\nkmf1.fit(tenure[yes_grade_B], event[yes_grade_B], label='Grade B')\nkmf1.plot(ax=a1)",
   "execution_count": 183,
   "outputs": [
    {
     "output_type": "execute_result",
     "execution_count": 183,
     "data": {
      "text/plain": "<matplotlib.axes._subplots.AxesSubplot at 0x7f4e8cb80358>"
     },
     "metadata": {}
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "<Figure size 432x288 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     }
    }
   ]
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": "## Time to Default (or negative credit event) for Grade C rated customers"
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "## Two Cohorts are compared. Cohort 1: Not Grade C Customers, and Cohort  2: Grade C CUstomers\ngrade_C = final_processed['grade_C']    \nnot_grade_C = (grade_C == 0)      ## Cohort Not Grade C Customers, having the pandas series  for the 1st cohort\nyes_grade_C = (grade_C == 1)     ## Cohort Grade C Customers, having the pandas series  for the 2nd cohort\n\n\n## fit the model for 1st cohort\nkmf1.fit(tenure[not_grade_C], event[not_grade_C], label='Not Grade C')\na1 = kmf1.plot()\n\n## fit the model for 2nd cohort\nkmf1.fit(tenure[yes_grade_C], event[yes_grade_C], label='Grade C')\nkmf1.plot(ax=a1)",
   "execution_count": 184,
   "outputs": [
    {
     "output_type": "execute_result",
     "execution_count": 184,
     "data": {
      "text/plain": "<matplotlib.axes._subplots.AxesSubplot at 0x7f4e8cb3b898>"
     },
     "metadata": {}
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "<Figure size 432x288 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     }
    }
   ]
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": "## Time to Default (or negative credit event) for Grade D rated customers"
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "## Two Cohorts are compared. Cohort 1: Not Grade D Customers, and Cohort  2: Grade D CUstomers\ngrade_D = final_processed['grade_D']    \nnot_grade_D = (grade_D == 0)      ## Cohort Not Grade D Customers, having the pandas series  for the 1st cohort\nyes_grade_D = (grade_D == 1)     ## Cohort Grade D Customers, having the pandas series  for the 2nd cohort\n\n\n## fit the model for 1st cohort\nkmf1.fit(tenure[not_grade_D], event[not_grade_D], label='Not Grade D')\na1 = kmf1.plot()\n\n## fit the model for 2nd cohort\nkmf1.fit(tenure[yes_grade_D], event[yes_grade_D], label='Grade D')\nkmf1.plot(ax=a1)",
   "execution_count": 185,
   "outputs": [
    {
     "output_type": "execute_result",
     "execution_count": 185,
     "data": {
      "text/plain": "<matplotlib.axes._subplots.AxesSubplot at 0x7f4e8cb53d30>"
     },
     "metadata": {}
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "<Figure size 432x288 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     }
    }
   ]
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": "## Time to Default (or negative credit event) for Grade E rated customers"
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "## Two Cohorts are compared. Cohort 1: Not Grade E Customers, and Cohort  2: Grade E CUstomers\ngrade_E = final_processed['grade_E']    \nnot_grade_E = (grade_E == 0)      ## Cohort Not Grade E Customers, having the pandas series  for the 1st cohort\nyes_grade_E = (grade_E == 1)     ## Cohort Grade E Customers, having the pandas series  for the 2nd cohort\n\n\n## fit the model for 1st cohort\nkmf1.fit(tenure[not_grade_E], event[not_grade_E], label='Not Grade D')\na1 = kmf1.plot()\n\n## fit the model for 2nd cohort\nkmf1.fit(tenure[yes_grade_E], event[yes_grade_E], label='Grade D')\nkmf1.plot(ax=a1)",
   "execution_count": 186,
   "outputs": [
    {
     "output_type": "execute_result",
     "execution_count": 186,
     "data": {
      "text/plain": "<matplotlib.axes._subplots.AxesSubplot at 0x7f4e8a9a31d0>"
     },
     "metadata": {}
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "<Figure size 432x288 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     }
    }
   ]
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": "## Time to Default (or negative credit event) for Grade F rated customers"
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "## Two Cohorts are compared. Cohort 1: Not Grade F Customers, and Cohort  2: Grade F CUstomers\ngrade_F = final_processed['grade_F']    \nnot_grade_F = (grade_F == 0)      ## Cohort Not Grade F Customers, having the pandas series  for the 1st cohort\nyes_grade_F = (grade_F == 1)     ## Cohort Grade F Customers, having the pandas series  for the 2nd cohort\n\n\n## fit the model for 1st cohort\nkmf1.fit(tenure[not_grade_F], event[not_grade_F], label='Not Grade F')\na1 = kmf1.plot()\n\n## fit the model for 2nd cohort\nkmf1.fit(tenure[yes_grade_F], event[yes_grade_F], label='Grade F')\nkmf1.plot(ax=a1)",
   "execution_count": 187,
   "outputs": [
    {
     "output_type": "execute_result",
     "execution_count": 187,
     "data": {
      "text/plain": "<matplotlib.axes._subplots.AxesSubplot at 0x7f4e8ca12588>"
     },
     "metadata": {}
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "<Figure size 432x288 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     }
    }
   ]
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": "## Time to Default (or negative credit event) for Grade G rated customers"
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "## Two Cohorts are compared. Cohort 1: Not Grade G Customers, and Cohort  2: Grade G CUstomers\ngrade_G = final_processed['grade_G']    \nnot_grade_G = (grade_G == 0)      ## Cohort Not Grade G Customers, having the pandas series  for the 1st cohort\nyes_grade_G = (grade_G == 1)     ## Cohort Grade G Customers, having the pandas series  for the 2nd cohort\n\n\n## fit the model for 1st cohort\nkmf1.fit(tenure[not_grade_G], event[not_grade_G], label='Not Grade G')\na1 = kmf1.plot()\n\n## fit the model for 2nd cohort\nkmf1.fit(tenure[yes_grade_G], event[yes_grade_G], label='Grade G')\nkmf1.plot(ax=a1)",
   "execution_count": 191,
   "outputs": [
    {
     "output_type": "execute_result",
     "execution_count": 191,
     "data": {
      "text/plain": "<matplotlib.axes._subplots.AxesSubplot at 0x7f4e8c97d908>"
     },
     "metadata": {}
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "<Figure size 432x288 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     }
    }
   ]
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": "# Intepreting the Plots\n\nThe y-axis represent the probablilty that a default/credit negative event will occur at a specific point on the timeline (the x-axis). \n\nAs we move along the timeline, we can see the survival of loans not rated G (blue line) has a better survival probability then loans G. This is consistent with the fact that G-rated loans are the weakest in credit rating. Compare the G-rated loan plot to the loans rated A, the reverse is true. Loans rate A (higher financial and credit strength) have improved survival probability versus all other loans."
  }
 ],
 "metadata": {
  "kernelspec": {
   "language": "python",
   "display_name": "Python 3",
   "name": "python3"
  },
  "language_info": {
   "pygments_lexer": "ipython3",
   "nbconvert_exporter": "python",
   "version": "3.6.4",
   "file_extension": ".py",
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "name": "python",
   "mimetype": "text/x-python"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}