Gradient Boosting.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "68e780cb",
   "metadata": {},
   "source": [
    "# 5. (a) Preprocessing the Data"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "83cbe3e4",
   "metadata": {},
   "source": [
    "### Importing the libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "ff8b3a40",
   "metadata": {},
   "outputs": [],
   "source": [
    "#importing the required libraries\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "import time\n",
    "from datetime import datetime\n",
    "from sklearn.preprocessing import LabelEncoder\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.ensemble import RandomForestClassifier,GradientBoostingClassifier\n",
    "from sklearn.metrics import plot_confusion_matrix,classification_report,accuracy_score"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ac2c9189",
   "metadata": {},
   "source": [
    "### Loading the train and test data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "04068d53",
   "metadata": {},
   "outputs": [],
   "source": [
    "#loading the training and testing data\n",
    "train = pd.read_csv('loan_train.csv')\n",
    "test = pd.read_csv('loan_test.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "700cd872",
   "metadata": {},
   "source": [
    "### Checking for null values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "55b4d18b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "51.26763629103722"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#checking the total % of null values in the training data\n",
    "(train.isnull().sum().sum())/(train.shape[0]*train.shape[1])*100"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "09af7457",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "51.46263262855191"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#checking the total % of null values in the testing data\n",
    "(test.isnull().sum().sum())/(test.shape[0]*test.shape[1])*100"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "47abcacb",
   "metadata": {},
   "outputs": [],
   "source": [
    "#51.26% of the training data and 51.46% of the test data are null values, so have to effectively process the null values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "d5fae25a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "id                                0\n",
       "member_id                         0\n",
       "loan_amnt                         0\n",
       "funded_amnt                       0\n",
       "funded_amnt_inv                   0\n",
       "                              ...  \n",
       "tax_liens                        17\n",
       "tot_hi_cred_lim               24999\n",
       "total_bal_ex_mort             24999\n",
       "total_bc_limit                24999\n",
       "total_il_high_credit_limit    24999\n",
       "Length: 111, dtype: int64"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#count of null values in each column\n",
    "train.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "19ea65cb",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "67"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#total number of columns containing some null values\n",
    "train.isnull().any().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "1b913c1b",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Only considering those columns which are having less than 20% of its data as null and dropping the rest of the columns\n",
    "train = train[train.columns[(train.isnull().sum())/len(train) < 0.2]]\n",
    "test = test[test.columns[(test.isnull().sum())/len(test) < 0.2]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "469601d0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(24999, 53)"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "48b92ac8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(14718, 53)"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "4aa9661a",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Now we are effectively left with 53 columns in both train and test data"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9f4516cf",
   "metadata": {},
   "source": [
    "### Dropping un-necessary features from the train and test dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "61173587",
   "metadata": {},
   "outputs": [],
   "source": [
    "#dropping some unnecessary columns which doesn't really add any value to our observation\n",
    "train = train.drop(['emp_title','pymnt_plan','title', 'zip_code', 'addr_state', 'initial_list_status',\n",
    "                    'application_type'],axis=1)\n",
    "test = test.drop(['emp_title','pymnt_plan','title', 'zip_code', 'addr_state', 'initial_list_status',\n",
    "                  'application_type'],axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "de063d56",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(24999, 46)"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "2dff7f1c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(14718, 46)"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "3dc97b61",
   "metadata": {},
   "outputs": [],
   "source": [
    "#dropping some unnecessary numerical columns\n",
    "train = train.drop(['acc_now_delinq','chargeoff_within_12_mths','collection_recovery_fee',\n",
    "                    'collections_12_mths_ex_med','delinq_amnt','policy_code','tax_liens'],axis=1)\n",
    "test = test.drop(['acc_now_delinq','chargeoff_within_12_mths','collection_recovery_fee',\n",
    "                  'collections_12_mths_ex_med','delinq_amnt','policy_code','tax_liens'],axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "d17fa48f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(24999, 39)"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "9f42aacf",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(14718, 39)"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d238fea4",
   "metadata": {},
   "source": [
    "### Categorical features into binary features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "1550d0d7",
   "metadata": {},
   "outputs": [],
   "source": [
    "#now we have to convert some of the categorical features into binary features\n",
    "#we can achieve this either by one hot encoding or label encoding depending upon the type of data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "e0833ffe",
   "metadata": {},
   "outputs": [],
   "source": [
    "# One hot encoding on categorical columns\n",
    "train = pd.get_dummies(train,columns = ['home_ownership','verification_status','purpose'],drop_first = True)\n",
    "test = pd.get_dummies(test,columns = ['home_ownership','verification_status','purpose'],drop_first = True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "006fd1db",
   "metadata": {},
   "outputs": [],
   "source": [
    "Lab_enc = LabelEncoder()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "31b8afeb",
   "metadata": {},
   "outputs": [],
   "source": [
    "#converting the ordinal columns into label encoded columns\n",
    "#encoding for the train set\n",
    "Lab_enc.fit(train['grade'])\n",
    "train['grade'] = Lab_enc.transform(train['grade'])\n",
    "Lab_enc.fit(train['sub_grade'])\n",
    "train['sub_grade'] = Lab_enc.transform(train['sub_grade'])\n",
    "\n",
    "#encoding for the test set\n",
    "Lab_enc.fit(test['grade'])\n",
    "test['grade'] = Lab_enc.transform(test['grade'])\n",
    "Lab_enc.fit(test['sub_grade'])\n",
    "test['sub_grade'] = Lab_enc.transform(test['sub_grade'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "5282efec",
   "metadata": {},
   "outputs": [],
   "source": [
    "#now we have to convert the date time columns into numerical columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "d0cf52cc",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0        11-Mar\n",
       "1        10-Mar\n",
       "2        11-Dec\n",
       "3        16-May\n",
       "4        15-Sep\n",
       "          ...  \n",
       "24994    12-May\n",
       "24995    14-Jul\n",
       "24996    13-Jan\n",
       "24997    11-Oct\n",
       "24998    14-Feb\n",
       "Name: last_pymnt_d, Length: 24999, dtype: object"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train['last_pymnt_d']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "39ede36a",
   "metadata": {},
   "outputs": [],
   "source": [
    "#these columns were having nan values beforehand, so have to drop those rows first\n",
    "train.dropna(subset=['last_pymnt_d'], inplace=True)\n",
    "test.dropna(subset=['last_pymnt_d'], inplace=True)\n",
    "test.dropna(subset=['last_credit_pull_d'], inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "29c7b08a",
   "metadata": {},
   "outputs": [],
   "source": [
    "# first we have to Convert objects to datetime columns\n",
    "\n",
    "def get_month(data):\n",
    "    y, m = data.split('-')\n",
    "    return datetime.strptime(str(2000 + int(y))+'-'+m+'-1', '%Y-%b-%d').month\n",
    "\n",
    "def get_year(data):\n",
    "    y, m = data.split('-')\n",
    "    return datetime.strptime(str(2000 + int(y))+'-'+m+'-1', '%Y-%b-%d').year"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "fcdc1c4f",
   "metadata": {},
   "outputs": [],
   "source": [
    "#further dividing the issue_d , last_pymt_d and last_credit_pull_d into months and year for the train set\n",
    "\n",
    "\n",
    "train['issue_d_year'] = train['issue_d'].apply(get_year)  \n",
    "train['issue_d_month'] = train['issue_d'].apply(get_month)  \n",
    "train['last_pymnt_d_year'] = train['last_pymnt_d'].apply(get_year)  \n",
    "train['last_pymnt_d_month'] = train['last_pymnt_d'].apply(get_month) \n",
    "train['last_credit_pull_d_year'] = train['last_credit_pull_d'].apply(get_year)  \n",
    "train['last_credit_pull_d_month'] = train['last_credit_pull_d'].apply(get_month)\n",
    "\n",
    "#further dividing the issue_d , last_pymt_d and last_credit_pull_d into months and year for the test set\n",
    "\n",
    "\n",
    "test['issue_d_year'] = test['issue_d'].apply(get_year)  \n",
    "test['issue_d_month'] = test['issue_d'].apply(get_month)  \n",
    "test['last_pymnt_d_year'] = test['last_pymnt_d'].apply(get_year)  \n",
    "test['last_pymnt_d_month'] = test['last_pymnt_d'].apply(get_month) \n",
    "test['last_credit_pull_d_year'] = test['last_credit_pull_d'].apply(get_year)  \n",
    "test['last_credit_pull_d_month'] = test['last_credit_pull_d'].apply(get_month)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "b8f53842",
   "metadata": {},
   "outputs": [],
   "source": [
    "#func to strip term and int_term columns\n",
    "def strip_col(data):\n",
    "    return data[:3]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "72ec5434",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Converting term and int_rate to float types\n",
    "train['term'] = train['term'].apply(strip_col)\n",
    "train['term'] = train['term'].astype('float')\n",
    "train['int_rate'] = train['int_rate'].apply(strip_col)\n",
    "train['int_rate'] = train['int_rate'].astype('float')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "2e5cbc60",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Converting term and int_rate to numerical columns\n",
    "test['term'] = test['term'].apply(strip_col)\n",
    "test['term'] = test['term'].astype('float')\n",
    "test['int_rate'] = test['int_rate'].apply(strip_col)\n",
    "test['int_rate'] = test['int_rate'].astype('float')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "c96b4045",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 800x600 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Checking correlation using heatmap\n",
    "plt.figure(figsize=(8,6),dpi=100)\n",
    "sns.heatmap(train.corr())\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "70a4d483",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(24954, 60)"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "ff4a5417",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(14690, 61)"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "ad416954",
   "metadata": {},
   "outputs": [],
   "source": [
    "#dropping some more unnecessary features\n",
    "train.drop(['issue_d','last_pymnt_d','last_credit_pull_d','earliest_cr_line','url'],axis=1,inplace=True)\n",
    "test.drop(['home_ownership_NONE','issue_d','last_pymnt_d','last_credit_pull_d',\n",
    "           'earliest_cr_line','url'],axis=1,inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "932c71d7",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Remove % symbol from revol_util column\n",
    "train['revol_util'] = train['revol_util'].str.rstrip('%')\n",
    "train['revol_util'] = train['revol_util'].astype(float)\n",
    "\n",
    "test['revol_util'] = test['revol_util'].str.rstrip('%')\n",
    "test['revol_util'] = test['revol_util'].astype(float)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "bd82d0c9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "id                                       0\n",
       "member_id                                0\n",
       "loan_amnt                                0\n",
       "funded_amnt                              0\n",
       "funded_amnt_inv                          0\n",
       "term                                     0\n",
       "int_rate                                 0\n",
       "installment                              0\n",
       "grade                                    0\n",
       "sub_grade                                0\n",
       "emp_length                              41\n",
       "annual_inc                               0\n",
       "loan_status                              0\n",
       "dti                                      0\n",
       "delinq_2yrs                              0\n",
       "inq_last_6mths                           0\n",
       "open_acc                                 0\n",
       "pub_rec                                  0\n",
       "revol_bal                                0\n",
       "revol_util                              26\n",
       "total_acc                                0\n",
       "out_prncp                                0\n",
       "out_prncp_inv                            0\n",
       "total_pymnt                              0\n",
       "total_pymnt_inv                          0\n",
       "total_rec_prncp                          0\n",
       "total_rec_int                            0\n",
       "total_rec_late_fee                       0\n",
       "recoveries                               0\n",
       "last_pymnt_amnt                          0\n",
       "pub_rec_bankruptcies                   417\n",
       "home_ownership_OTHER                     0\n",
       "home_ownership_OWN                       0\n",
       "home_ownership_RENT                      0\n",
       "verification_status_Source Verified      0\n",
       "verification_status_Verified             0\n",
       "purpose_credit_card                      0\n",
       "purpose_debt_consolidation               0\n",
       "purpose_educational                      0\n",
       "purpose_home_improvement                 0\n",
       "purpose_house                            0\n",
       "purpose_major_purchase                   0\n",
       "purpose_medical                          0\n",
       "purpose_moving                           0\n",
       "purpose_other                            0\n",
       "purpose_renewable_energy                 0\n",
       "purpose_small_business                   0\n",
       "purpose_vacation                         0\n",
       "purpose_wedding                          0\n",
       "issue_d_year                             0\n",
       "issue_d_month                            0\n",
       "last_pymnt_d_year                        0\n",
       "last_pymnt_d_month                       0\n",
       "last_credit_pull_d_year                  0\n",
       "last_credit_pull_d_month                 0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#checking for remaining null values\n",
    "train.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "1100984e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0        10+ years\n",
       "1           1 year\n",
       "2          4 years\n",
       "3        10+ years\n",
       "4          5 years\n",
       "           ...    \n",
       "24994      7 years\n",
       "24995      7 years\n",
       "24996    10+ years\n",
       "24997    10+ years\n",
       "24998     < 1 year\n",
       "Name: emp_length, Length: 24954, dtype: object"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train['emp_length']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "ace5cafd",
   "metadata": {},
   "outputs": [],
   "source": [
    "#converting the emp_length column to have only integers\n",
    "train['emp_length'].fillna('0',inplace=True)\n",
    "train['emp_length']=train['emp_length'].str.extract('(\\d+)')\n",
    "train['emp_length'] = train['emp_length'].astype(int)\n",
    "\n",
    "test['emp_length'].fillna('0',inplace=True)\n",
    "test['emp_length']=test['emp_length'].str.extract('(\\d+)')\n",
    "test['emp_length'] = test['emp_length'].astype(int)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "77add171",
   "metadata": {},
   "outputs": [],
   "source": [
    "#replacing the nan values with the mean of the particular column\n",
    "train['revol_util'] = train['revol_util'].fillna(train['revol_util'].mean())\n",
    "test['revol_util'] = test['revol_util'].fillna(test['revol_util'].mean())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "94204df7",
   "metadata": {},
   "outputs": [],
   "source": [
    "#dropping some more unnecessary features on the way\n",
    "train.drop(['total_pymnt','total_pymnt_inv','total_rec_prncp','total_rec_int',\n",
    "        'total_rec_late_fee','recoveries','pub_rec_bankruptcies'],inplace=True,axis=1)\n",
    "\n",
    "test.drop(['total_pymnt','total_pymnt_inv','total_rec_prncp','total_rec_int',\n",
    "        'total_rec_late_fee','recoveries','pub_rec_bankruptcies'],inplace=True,axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "a1e206e2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Fully Paid     83.461569\n",
       "Charged Off    13.741284\n",
       "Current         2.797147\n",
       "Name: loan_status, dtype: float64"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#loan_status in comparison\n",
    "(train['loan_status'].value_counts()*100)/len(train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "dd3f6c0a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "698"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# number of 'current' loan_status which are not relevent to us for this problem\n",
    "(train['loan_status'] == 'Current').sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "0df4a625",
   "metadata": {},
   "outputs": [],
   "source": [
    "#dropping the 'current' loan_status\n",
    "train = train.drop(train[train['loan_status'] == 'Current'].index)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "987cc0ff",
   "metadata": {},
   "outputs": [],
   "source": [
    "train['loan_status'] = train['loan_status'].map({'Fully Paid':'+1','Charged Off':'-1'})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "dd4fde02",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "+1    85.863292\n",
       "-1    14.136708\n",
       "Name: loan_status, dtype: float64"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#loan_status in comparison\n",
    "(train['loan_status'].value_counts()*100)/len(train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "0ce26c14",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "+1    85.078608\n",
       "-1    14.921392\n",
       "Name: loan_status, dtype: float64"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#dropping the 'current' loan_status\n",
    "test = test.drop(test[test['loan_status'] == 'Current'].index)\n",
    "test['loan_status'] = test['loan_status'].map({'Fully Paid':'+1','Charged Off':'-1'})\n",
    "#loan_status in comparison\n",
    "(test['loan_status'].value_counts()*100)/len(test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "cb77023c",
   "metadata": {},
   "outputs": [],
   "source": [
    "#features with only 0's \n",
    "train.drop(['out_prncp','out_prncp_inv'],axis=1,inplace=True)\n",
    "test.drop(['out_prncp','out_prncp_inv'],axis=1,inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "1bb9b4e1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(24256, 46)"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "aa5abd99",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(14248, 46)"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "7e6b6fb5",
   "metadata": {},
   "outputs": [],
   "source": [
    "# At the end we successfully shrinked our data down to 46 features for both train and test dataset"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b4b87450",
   "metadata": {},
   "source": [
    "# 5. (b) Training the model and making predictions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "711ae9ba",
   "metadata": {},
   "outputs": [],
   "source": [
    "#making the train and test splits\n",
    "X_train = train.drop('loan_status', axis=1)\n",
    "X_test = test.drop('loan_status', axis=1)\n",
    "y_train = train['loan_status']\n",
    "y_test = test['loan_status']"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1e909beb",
   "metadata": {},
   "source": [
    "### Reporting best accuracy, optimal hyperparameters, precision and recall of the models"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "c25c5e16",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Applied a grid search crossvalidation to find the optimal parameters to get the best acccuracy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "id": "c7b7aee7",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Please don't try to run this cell as it takes hours to complete\n",
    "\n",
    "\n",
    "# from sklearn.model_selection import GridSearchCV\n",
    "# gradient_model = GradientBoostingClassifier()\n",
    "\n",
    "# n_estimators = [10,50,100,200,300]\n",
    "# max_features = [5,6,7,8,9,10]\n",
    "# max_depth = [None,5,10,15,20]\n",
    "\n",
    "# param_grid = {'n_estimators': n_estimators,'max_features': max_features,'max_depth': max_depth}\n",
    "# grid_model = GridSearchCV(estimator = gradient_model, param_grid = param_grid, verbose = 2)\n",
    "# grid_model.fit(X_train,y_train)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "29a15ab9",
   "metadata": {},
   "source": [
    "### Model 1 (n_estimators=50,max_features=5,max_depth=5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "11614c21",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy Score:  92.96041549691185\n",
      "Time Taken: 1.566 secs\n",
      "\n",
      "\n",
      "Classification Report\n",
      "\n",
      "\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "          +1       0.93      0.99      0.96     12122\n",
      "          -1       0.93      0.57      0.71      2126\n",
      "\n",
      "    accuracy                           0.93     14248\n",
      "   macro avg       0.93      0.78      0.83     14248\n",
      "weighted avg       0.93      0.93      0.92     14248\n",
      "\n"
     ]
    }
   ],
   "source": [
    "start_time = time.time()\n",
    "\n",
    "model1 = GradientBoostingClassifier(n_estimators=50,max_features=5,max_depth=5)\n",
    "model1.fit(X_train, y_train)\n",
    "pred = model1.predict(X_test)\n",
    "print('Accuracy Score: ',accuracy_score(y_test, pred)*100)\n",
    "\n",
    "end_time=time.time()\n",
    "print(\"Time Taken: {} secs\".format(round(end_time-start_time,3)))\n",
    "print('\\n')\n",
    "# Checking classification_report\n",
    "print('Classification Report')\n",
    "print('\\n')\n",
    "print(classification_report(y_test, pred))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9c44affb",
   "metadata": {},
   "source": [
    "### Model 2 (n_estimators=100,max_features=5,max_depth=5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "0a20dc3a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy Score:  95.12914093206064\n",
      "Time Taken: 2.937 secs\n",
      "\n",
      "\n",
      "Classification Report\n",
      "\n",
      "\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "          +1       0.96      0.99      0.97     12122\n",
      "          -1       0.92      0.74      0.82      2126\n",
      "\n",
      "    accuracy                           0.95     14248\n",
      "   macro avg       0.94      0.86      0.90     14248\n",
      "weighted avg       0.95      0.95      0.95     14248\n",
      "\n"
     ]
    }
   ],
   "source": [
    "start_time = time.time()\n",
    "\n",
    "model2 = GradientBoostingClassifier(n_estimators=100,max_features=5,max_depth=5)\n",
    "model2.fit(X_train, y_train)\n",
    "pred = model2.predict(X_test)\n",
    "print('Accuracy Score: ',accuracy_score(y_test, pred)*100)\n",
    "\n",
    "end_time=time.time()\n",
    "print(\"Time Taken: {} secs\".format(round(end_time-start_time,3)))\n",
    "print('\\n')\n",
    "# Checking classification_report\n",
    "print('Classification Report')\n",
    "print('\\n')\n",
    "print(classification_report(y_test, pred))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f574f621",
   "metadata": {},
   "source": [
    "### Model 3 (n_estimators=150,max_features=5,max_depth=5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "0ca34fed",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy Score:  95.66956765861875\n",
      "Time Taken: 4.492 secs\n",
      "\n",
      "\n",
      "Classification Report\n",
      "\n",
      "\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "          +1       0.96      0.99      0.97     12122\n",
      "          -1       0.91      0.79      0.85      2126\n",
      "\n",
      "    accuracy                           0.96     14248\n",
      "   macro avg       0.93      0.89      0.91     14248\n",
      "weighted avg       0.96      0.96      0.96     14248\n",
      "\n"
     ]
    }
   ],
   "source": [
    "start_time = time.time()\n",
    "\n",
    "model3 = GradientBoostingClassifier(n_estimators=150,max_features=5,max_depth=5)\n",
    "model3.fit(X_train, y_train)\n",
    "pred = model3.predict(X_test)\n",
    "print('Accuracy Score: ',accuracy_score(y_test, pred)*100)\n",
    "\n",
    "end_time=time.time()\n",
    "print(\"Time Taken: {} secs\".format(round(end_time-start_time,3)))\n",
    "print('\\n')\n",
    "# Checking classification_report\n",
    "print('Classification Report')\n",
    "print('\\n')\n",
    "print(classification_report(y_test, pred))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "23de8ef0",
   "metadata": {},
   "source": [
    "### Final Model (n_estimators=500,max_features=8,max_depth=10)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bd714403",
   "metadata": {},
   "source": [
    "#### n_estimators = 500, max_features = 8, max_depth = 10 leads to the best accuracy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "id": "f01d9373",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy Score:  96.80656934306569\n",
      "Time Taken: 54.273 secs\n"
     ]
    }
   ],
   "source": [
    "start_time = time.time()\n",
    "\n",
    "final_model = GradientBoostingClassifier(n_estimators=500,max_features=8,max_depth=10)\n",
    "final_model.fit(X_train, y_train)\n",
    "pred = final_model.predict(X_test)\n",
    "print('Accuracy Score: ',accuracy_score(y_test, pred)*100)\n",
    "\n",
    "end_time=time.time()\n",
    "print(\"Time Taken: {} secs\".format(round(end_time-start_time,3)))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c46eb199",
   "metadata": {},
   "source": [
    "#### Best test accuracy achieved is ≈ 97%"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "id": "7b2f5030",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Classification Report\n",
      "\n",
      "\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "          +1       0.97      0.99      0.98     12122\n",
      "          -1       0.93      0.85      0.89      2126\n",
      "\n",
      "    accuracy                           0.97     14248\n",
      "   macro avg       0.95      0.92      0.93     14248\n",
      "weighted avg       0.97      0.97      0.97     14248\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# Checking classification_report\n",
    "print('Classification Report')\n",
    "print('\\n')\n",
    "print(classification_report(y_test, pred))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "id": "ec358cee",
   "metadata": {},
   "outputs": [],
   "source": [
    "#The precision and recall for the model is listed above in the classification report"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "id": "e35c918c",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Checking for the Important features having atleast 1% corr with our class label as per our model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "id": "885b5361",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1000x400 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "feat_imp = pd.DataFrame(index = X_train.columns, data = final_model.feature_importances_, columns = ['Importance'])\n",
    "feat_imp = feat_imp[feat_imp['Importance'] > 0.01]\n",
    "\n",
    "plt.figure(figsize=(10,4),dpi=100)\n",
    "sns.barplot(data = feat_imp.sort_values('Importance'), x = feat_imp.sort_values('Importance').index, y = 'Importance')\n",
    "plt.xticks(rotation=90)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a68e4db8",
   "metadata": {},
   "source": [
    "### Effect of increasing the number of trees on the test error"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "id": "bc82c7fd",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Time Taken: 222.68\n"
     ]
    }
   ],
   "source": [
    "start_time = time.time()\n",
    "test_error=[]\n",
    "n_estimators = [50,100,150,200,250,300,400,500]\n",
    "\n",
    "for n in n_estimators:\n",
    "    gradient_model = GradientBoostingClassifier(n_estimators=n)\n",
    "    gradient_model.fit(X_train,y_train)\n",
    "    pred = gradient_model.predict(X_test)\n",
    "    test_error.append(1-accuracy_score(y_test,pred))\n",
    "\n",
    "end_time=time.time()\n",
    "print(\"Time Taken: {}\".format(round(end_time-start_time,3)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "id": "3d864342",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(n_estimators,test_error)\n",
    "plt.xlabel('Number of Trees')\n",
    "plt.ylabel('Error')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "65a9ee62",
   "metadata": {},
   "source": [
    "### Comparing GradientBoosting Performance against a simple decision tree"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "id": "e2caf590",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9390791690061763"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#simple decision tree built using information gain \n",
    "dt_model = DecisionTreeClassifier(criterion = 'entropy')\n",
    "dt_model.fit(X_train,y_train)\n",
    "pred = dt_model.predict(X_test)\n",
    "accuracy_score(pred,y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "20784785",
   "metadata": {},
   "source": [
    "#### Test accuracy achieved using decision tree is ≈ 94%"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "id": "8f8c106a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Classification Report\n",
      "\n",
      "\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "          +1       0.96      0.96      0.96     12122\n",
      "          -1       0.80      0.79      0.79      2126\n",
      "\n",
      "    accuracy                           0.94     14248\n",
      "   macro avg       0.88      0.88      0.88     14248\n",
      "weighted avg       0.94      0.94      0.94     14248\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# Checking classification_report\n",
    "print('Classification Report')\n",
    "print('\\n')\n",
    "print(classification_report(y_test, pred))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "id": "c13eadfc",
   "metadata": {},
   "outputs": [],
   "source": [
    "#The precision and recall of the decision tree model is listed above in the classification report"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}