XGboost_binary.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "pd.options.mode.chained_assignment = None \n",
    "from sklearn.metrics import roc_auc_score\n",
    "from sklearn.metrics import confusion_matrix\n",
    "from sklearn.metrics import accuracy_score\n",
    "import xgboost as xgb\n",
    "from sklearn.metrics import mean_squared_error\n",
    "from sklearn.model_selection import train_test_split\n",
    "import pickle\n",
    "import sklearn.metrics as metrics\n",
    "import imblearn\n",
    "from imblearn.over_sampling import SMOTE\n",
    "from collections import Counter\n",
    "import matplotlib.pyplot as pyplot\n",
    "from numpy import where"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "#loading the filse\n",
    "flights = pd.read_csv('data/flights.csv')\n",
    "flights = flights[['fl_date','mkt_carrier','origin','dest'\n",
    "                  ,'taxi_out','taxi_in','cancelled','crs_elapsed_time','arr_delay']]\n",
    "\n",
    "#features and target\n",
    "X = flights.loc[:,flights.columns!='cancelled']\n",
    "y = flights[['cancelled']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "#lets create year, month and day\n",
    "X['fl_date'] = pd.to_datetime(X['fl_date'],format='%Y-%m-%d')\n",
    "X['year'] = pd.DatetimeIndex(X['fl_date']).year\n",
    "X['month'] = pd.DatetimeIndex(X['fl_date']).month\n",
    "X['day'] = pd.DatetimeIndex(X['fl_date']).day"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = X[['month','day','origin','dest','crs_elapsed_time']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "X= pd.get_dummies(X,columns=['origin','dest'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Ratio of Class 0 to Class 1: 60.72839506172839\n"
     ]
    }
   ],
   "source": [
    "#calculating the weights of class 0 to 1\n",
    "class_0 = y.loc[y['cancelled'] == 0].value_counts().values[0]\n",
    "class_1 = y.loc[y['cancelled'] == 1].value_counts().values[0]\n",
    "weight = class_0/class_1\n",
    "print('Ratio of Class 0 to Class 1:',weight)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Splitting into train test\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.30, random_state=123)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "#giving the ratio as a weight to XGBClassifier\n",
    "model = xgb.XGBClassifier(scale_pos_weight=60.72)\n",
    "model.fit(X_train.values,y_train.values.reshape(1,-1)[0])\n",
    "filename = 'XGBoost_model.sav'\n",
    "pickle.dump(model, open(filename, 'wb'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Roc_Auc Score: 0.6553020324612396\n",
      "accuracy score:  0.7326555555555555\n"
     ]
    }
   ],
   "source": [
    "y_pred = model.predict(X_test.values)\n",
    "print('Roc_Auc Score:',roc_auc_score(y_test,y_pred))\n",
    "print('accuracy score: ',accuracy_score(y_test,y_pred))\n",
    "arr = confusion_matrix(y_test,y_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "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": [
    "# calculate the fpr and tpr for all thresholds of the classification\n",
    "probs = model.predict_proba(X_test)\n",
    "preds = probs[:,1]\n",
    "fpr, tpr, threshold = metrics.roc_curve(y_test, y_pred)\n",
    "roc_auc = metrics.auc(fpr, tpr)\n",
    "\n",
    "plt.title('Receiver Operating Characteristic')\n",
    "plt.plot(fpr, tpr, 'b', label = 'AUC = %0.2f' % roc_auc)\n",
    "plt.legend(loc = 'lower right')\n",
    "plt.plot([0, 1], [0, 1],'r--')\n",
    "plt.xlim([0, 1])\n",
    "plt.ylim([0, 1])\n",
    "plt.ylabel('True Positive Rate')\n",
    "plt.xlabel('False Positive Rate')\n",
    "plt.show()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "bootcamp",
   "language": "python",
   "name": "bootcamp"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}