Occupancy_Detection.ipynb

{
 "nbformat": 4,
 "nbformat_minor": 0,
 "metadata": {
  "colab": {
   "provenance": []
  },
  "kernelspec": {
   "name": "python3",
   "display_name": "Python 3"
  },
  "language_info": {
   "name": "python"
  }
 },
 "cells": [
  {
   "cell_type": "markdown",
   "source": [
    "#Importing Libraies"
   ],
   "metadata": {
    "id": "HbPJD0qi1O-N"
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.svm import LinearSVC\n",
    "from sklearn.naive_bayes import GaussianNB\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import confusion_matrix, accuracy_score, classification_report, roc_curve, roc_auc_score"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "import warnings\n",
    "warnings.filterwarnings('ignore')"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "#Data Reading "
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% md\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "data = pd.read_csv('datatraining.txt', error_bad_lines=False)\n",
    "data_test_1 = pd.read_csv('datatest1.txt', error_bad_lines=False)\n",
    "data_test_2 = pd.read_csv('datatest2.txt', error_bad_lines=False)"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "#Data Nature"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% md\n"
    }
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "**Number of (rows, columns)**"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% md\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "data.shape"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "data_test_1.shape"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "data_test_2.shape"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "**Snippets form data**"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% md\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "data.head()"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "**Data information**"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% md\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "data.info()"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "**Data statistics**"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% md\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "data.describe()"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "**Count valuew of Occupancy**"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% md\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "data['Occupancy'].value_counts()"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "data_test_1['Occupancy'].value_counts()"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "data_test_2['Occupancy'].value_counts()"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "**Drop column date**"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% md\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "data.drop('date', axis=1, inplace=True)\n",
    "data_test_1.drop('date', axis=1, inplace=True)\n",
    "data_test_2.drop('date', axis=1, inplace=True)"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "#Data Distribution"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% md\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "plt.figure(figsize=(20,7))\n",
    "sns.countplot(x=data['Temperature'])\n",
    "plt.title('Distribution Plot for Temperature')\n",
    "plt.xlabel('Temperature')\n",
    "plt.ylabel('Count')\n",
    "plt.savefig('Distribution Plot for Temperature.png')\n",
    "plt.show()"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "plt.figure(figsize=(20,7))\n",
    "sns.countplot(x=data['Humidity'])\n",
    "plt.title('Distribution Plot for Humidity')\n",
    "plt.xlabel('Humidity')\n",
    "plt.ylabel('Count')\n",
    "plt.savefig('Distribution Plot for Humidity.png')\n",
    "plt.show()"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "plt.figure(figsize=(20,7))\n",
    "sns.countplot(x=data['Light'])\n",
    "plt.ylim(0, 250)\n",
    "plt.title('Distribution Plot for Light')\n",
    "plt.xlabel('Light')\n",
    "plt.ylabel('Count')\n",
    "plt.savefig('Distribution Plot for Light.png')\n",
    "plt.show()"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "plt.figure(figsize=(20,7))\n",
    "sns.countplot(x=data['CO2'])\n",
    "plt.title('Distribution Plot for CO2')\n",
    "plt.xlabel('CO2')\n",
    "plt.ylabel('Count')\n",
    "plt.savefig('Distribution Plot for CO2.png')\n",
    "plt.show()"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "plt.figure(figsize=(20,7))\n",
    "sns.countplot(x=data['HumidityRatio'])\n",
    "plt.title('Distribution Plot for HumidityRatio')\n",
    "plt.xlabel('HumidityRatio')\n",
    "plt.ylabel('Count')\n",
    "plt.savefig('Distribution Plot for HumidityRatio.png')\n",
    "plt.show()"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "plt.figure(figsize=(20,7))\n",
    "sns.countplot(x=data['Occupancy'])\n",
    "plt.title('Distribution Plot for Occupancy')\n",
    "plt.xlabel('Occupancy')\n",
    "plt.ylabel('Count')\n",
    "plt.savefig('Distribution Plot for Occupancy.png')\n",
    "plt.show()"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "**Heatmap of correlation**"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% md\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "plt.figure(figsize=(10,7))\n",
    "plt.title('Correlation Table for Features', fontdict={'fontsize':18})\n",
    "ax = sns.heatmap(data.corr(), annot=True, linewidths=.2)\n",
    "plt.savefig('Correlation Table for Feature.png')"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "**Box Plot for Features**"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% md\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "plt.figure(figsize=(20,7))\n",
    "plt.title('Box Plot for Features', fontdict={'fontsize':18})\n",
    "ax = sns.boxplot(data=data.drop(['Occupancy'], axis=1), orient=\"h\", palette=\"Set2\")\n",
    "plt.savefig('Box Plot for Features.png')"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "#Data Scaling "
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% md\n"
    }
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "**Scale data using Standard Scaler**"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% md\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "scaler = StandardScaler()\n",
    "columns = ['Temperature', 'Humidity', 'Light', 'CO2', 'HumidityRatio']\n",
    "scaler.fit(np.array(data[columns]))\n",
    "\n",
    "data[columns] = scaler.transform(np.array(data[columns]))\n",
    "data_test_1[columns] = scaler.transform(np.array(data_test_1[columns]))\n",
    "data_test_2[columns] = scaler.transform(np.array(data_test_2[columns]))"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "**Box Plot after Scaling**"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% md\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "plt.figure(figsize=(20,7))\n",
    "plt.title('Box Plot for Features', fontdict={'fontsize':18})\n",
    "ax = sns.boxplot(data=data.drop(['Occupancy'], axis=1), orient=\"h\", palette=\"Set2\")\n",
    "plt.savefig('Box Plot for Features After scaling.png')"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "#Data Preprocessing"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% md\n"
    }
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "**Spliting data into X, Y**"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% md\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "y_values = data['Occupancy'].values\n",
    "x_values = data[['Temperature',\t'Humidity',\t'Light',\t'CO2',\t'HumidityRatio']].values"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "x_values.shape, y_values.shape"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "**Spliting test data 1 & 2 into X,Y**"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% md\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "y_test_1 = data_test_1['Occupancy'].values\n",
    "x_test_1 = data_test_1[['Temperature',\t'Humidity',\t'Light',\t'CO2',\t'HumidityRatio']].values\n",
    "\n",
    "y_test_2 = data_test_2['Occupancy'].values\n",
    "x_test_2 = data_test_2[['Temperature',\t'Humidity',\t'Light',\t'CO2',\t'HumidityRatio']].values"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "y_test_1.shape, x_test_1.shape"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "y_test_2.shape, x_test_2.shape"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "**Split data into train & test data**"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% md\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "x_train, x_test, y_train, y_test = train_test_split(x_values, y_values, test_size=0.3, random_state=40)"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "x_train.shape, y_train.shape"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "x_test.shape, y_test.shape"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "#Logistic Regression Algorithm"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% md\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "LR_Model = LogisticRegression(penalty='l2', dual=False, tol=1e-4, C=1, fit_intercept=True, intercept_scaling=1, class_weight=None, random_state=None, \n",
    "                           solver='lbfgs', max_iter=100, multi_class='auto', verbose=0, warm_start=False, n_jobs=None, l1_ratio=None)\n",
    "LR_Model = LR_Model.fit(x_train, y_train)"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "lr_y_pred   = LR_Model.predict(x_test)\n",
    "lr_y_pred_1 = LR_Model.predict(x_test_1)\n",
    "lr_y_pred_2 = LR_Model.predict(x_test_2)"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "print('---------------------------------------------------------------------------------------')\n",
    "print(\"{:20}\\t|\\t{:20}\\t|\\t{:20}\".format('Original', 'Test 1', 'Test 2'))\n",
    "print('---------------------------------------------------------------------------------------')\n",
    "print(\"{:>15}\\t|\\t{:>15}\\t|\\t{:>15}\".format(\n",
    "    accuracy_score(y_test, lr_y_pred)*100, \n",
    "    accuracy_score(y_test_1, lr_y_pred_1)*100, \n",
    "    accuracy_score(y_test_2, lr_y_pred_2)*100)\n",
    ")\n",
    "print('---------------------------------------------------------------------------------------')"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "print(classification_report(y_test, lr_y_pred))"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "print(classification_report(y_test_1, lr_y_pred_1))"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "print(classification_report(y_test_2, lr_y_pred_2))"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "plt.figure(figsize=(20,10))\n",
    "sns.heatmap(confusion_matrix(y_test, lr_y_pred), annot=True, cmap='Oranges', fmt=\"d\")\n",
    "plt.title('Confusion Matrix of Data')\n",
    "plt.savefig('Confusion Matrix of Data Logistic Regression.png')\n",
    "plt.show()"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "plt.figure(figsize=(20,10))\n",
    "sns.heatmap(confusion_matrix(y_test_1, lr_y_pred_1), annot=True, cmap='Reds', fmt=\"d\")\n",
    "plt.title('Confusion Matrix of Test 1')\n",
    "plt.savefig('Confusion Matrix of Test 1 Logistic Regression.png')\n",
    "plt.show()"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "plt.figure(figsize=(20,10))\n",
    "sns.heatmap(confusion_matrix(y_test_2, lr_y_pred_2), annot=True, cmap='Blues', fmt=\"d\")\n",
    "plt.title('Confusion Matrix of Test 2')\n",
    "plt.savefig('Confusion Matrix of Test 2 Logistic Regression.png')\n",
    "plt.show()"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "plt.figure(figsize=(20,10))\n",
    "\n",
    "y_pred_proba = LR_Model.predict(x_test)\n",
    "fpr, tpr, _ = roc_curve(y_test,  y_pred_proba)\n",
    "plt.plot(fpr, tpr, label=\"Data\")\n",
    "\n",
    "y_pred_proba_1 = LR_Model.predict(x_test_1)\n",
    "fpr, tpr, _ = roc_curve(y_test_1,  y_pred_proba_1)\n",
    "plt.plot(fpr, tpr, label=\"Test 1\")\n",
    "\n",
    "y_pred_proba_2 = LR_Model.predict(x_test_2)\n",
    "fpr, tpr, _ = roc_curve(y_test_2,  y_pred_proba_2)\n",
    "plt.plot(fpr, tpr, label=\"Test 2\")\n",
    "\n",
    "plt.title('ROC Logistic Regression')\n",
    "plt.legend(loc=4)\n",
    "plt.savefig('ROC Logistic Regression.png')\n",
    "plt.show()"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "#Support Vector Machine Algorithm"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% md\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "LSVC_Model = LinearSVC()\n",
    "LSVC_Model = LSVC_Model.fit(x_train, y_train)"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "svc_y_pred   = LSVC_Model.predict(x_test)\n",
    "svc_y_pred_1 = LSVC_Model.predict(x_test_1)\n",
    "svc_y_pred_2 = LSVC_Model.predict(x_test_2)"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "print('---------------------------------------------------------------------------------------')\n",
    "print(\"{:20}\\t|\\t{:20}\\t|\\t{:20}\".format('Original', 'Test 1', 'Test 2'))\n",
    "print('---------------------------------------------------------------------------------------')\n",
    "print(\"{:>15}\\t|\\t{:>15}\\t|\\t{:>15}\".format(\n",
    "    accuracy_score(y_test,   svc_y_pred)  *100, \n",
    "    accuracy_score(y_test_1, svc_y_pred_1)*100, \n",
    "    accuracy_score(y_test_2, svc_y_pred_2)*100)\n",
    ")\n",
    "print('---------------------------------------------------------------------------------------')"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "print(classification_report(y_test, svc_y_pred))"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "print(classification_report(y_test_1, svc_y_pred_1))"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "print(classification_report(y_test_2, svc_y_pred_2))"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "plt.figure(figsize=(20,10))\n",
    "sns.heatmap(confusion_matrix(y_test, svc_y_pred), annot=True, cmap='Oranges', fmt=\"d\")\n",
    "plt.title('Confusion Matrix of Data')\n",
    "plt.savefig('Confusion Matrix of Data Support Vector Machine.png')\n",
    "plt.show()"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "plt.figure(figsize=(20,10))\n",
    "sns.heatmap(confusion_matrix(y_test_1, svc_y_pred_1), annot=True, cmap='Reds', fmt=\"d\")\n",
    "plt.title('Confusion Matrix of Test 1')\n",
    "plt.savefig('Confusion Matrix of Test 1 Support Vector Machine.png')\n",
    "plt.show()"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "plt.figure(figsize=(20,10))\n",
    "sns.heatmap(confusion_matrix(y_test_2, svc_y_pred_2), annot=True, cmap='Blues', fmt=\"d\")\n",
    "plt.title('Confusion Matrix of Test 2')\n",
    "plt.savefig('Confusion Matrix of Test 2 Support Vector Machine.png')\n",
    "plt.show()"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "plt.figure(figsize=(20,10))\n",
    "\n",
    "y_pred_proba = LSVC_Model.predict(x_test)\n",
    "fpr, tpr, _ = roc_curve(y_test,  y_pred_proba)\n",
    "plt.plot(fpr, tpr, label=\"Data\")\n",
    "\n",
    "y_pred_proba_1 = LSVC_Model.predict(x_test_1)\n",
    "fpr, tpr, _ = roc_curve(y_test_1,  y_pred_proba_1)\n",
    "plt.plot(fpr, tpr, label=\"Test 1\")\n",
    "\n",
    "y_pred_proba_2 = LSVC_Model.predict(x_test_2)\n",
    "fpr, tpr, _ = roc_curve(y_test_2,  y_pred_proba_2)\n",
    "plt.plot(fpr, tpr, label=\"Test 2\")\n",
    "\n",
    "plt.title('ROC Support Vector Machine')\n",
    "plt.legend(loc=4)\n",
    "plt.savefig('ROC Support Vector Machine.png')\n",
    "plt.show()"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "#Naive Bayes Algorithm"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% md\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "GNB_Model = GaussianNB()\n",
    "GNB_Model = GNB_Model.fit(x_train, y_train)"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "gnb_y_pred   = GNB_Model.predict(x_test)\n",
    "gnb_y_pred_1 = GNB_Model.predict(x_test_1)\n",
    "gnb_y_pred_2 = GNB_Model.predict(x_test_2)"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "print('---------------------------------------------------------------------------------------')\n",
    "print(\"{:20}\\t|\\t{:20}\\t|\\t{:20}\".format('Original', 'Test 1', 'Test 2'))\n",
    "print('---------------------------------------------------------------------------------------')\n",
    "print(\"{:>15}\\t|\\t{:>15}\\t|\\t{:>15}\".format(\n",
    "    accuracy_score(y_test,   gnb_y_pred)  *100, \n",
    "    accuracy_score(y_test_1, gnb_y_pred_1)*100, \n",
    "    accuracy_score(y_test_2, gnb_y_pred_2)*100)\n",
    ")\n",
    "print('---------------------------------------------------------------------------------------')"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "print(classification_report(y_test, gnb_y_pred))"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "print(classification_report(y_test_1, gnb_y_pred_1))"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "print(classification_report(y_test_2, gnb_y_pred_2))"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "plt.figure(figsize=(20,10))\n",
    "sns.heatmap(confusion_matrix(y_test, gnb_y_pred), annot=True, cmap='Oranges', fmt=\"d\")\n",
    "plt.title('Confusion Matrix of Data')\n",
    "plt.savefig('Confusion Matrix of Data Naive Bayes.png')\n",
    "plt.show()"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "plt.figure(figsize=(20,10))\n",
    "sns.heatmap(confusion_matrix(y_test_1, gnb_y_pred_1), annot=True, cmap='Reds', fmt=\"d\")\n",
    "plt.title('Confusion Matrix of Test 1')\n",
    "plt.savefig('Confusion Matrix of Test 1 Naive Bayes.png')\n",
    "plt.show()"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "plt.figure(figsize=(20,10))\n",
    "sns.heatmap(confusion_matrix(y_test_2, gnb_y_pred_2), annot=True, cmap='Blues', fmt=\"d\")\n",
    "plt.title('Confusion Matrix of Test 2')\n",
    "plt.savefig('Confusion Matrix of Test 2 Naive Bayes.png')\n",
    "plt.show()"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "plt.figure(figsize=(20,10))\n",
    "\n",
    "y_pred_proba = GNB_Model.predict(x_test)\n",
    "fpr, tpr, _ = roc_curve(y_test,  y_pred_proba)\n",
    "plt.plot(fpr, tpr, label=\"Data\")\n",
    "\n",
    "y_pred_proba_1 = GNB_Model.predict(x_test_1)\n",
    "fpr, tpr, _ = roc_curve(y_test_1,  y_pred_proba_1)\n",
    "plt.plot(fpr, tpr, label=\"Test 1\")\n",
    "\n",
    "y_pred_proba_2 = GNB_Model.predict(x_test_2)\n",
    "fpr, tpr, _ = roc_curve(y_test_2,  y_pred_proba_2)\n",
    "plt.plot(fpr, tpr, label=\"Test 2\")\n",
    "\n",
    "plt.title('ROC Naive Bayes')\n",
    "plt.legend(loc=4)\n",
    "plt.savefig('ROC Naive Bayes.png')\n",
    "plt.show()"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "source": [
    "plt.figure(figsize=(20,10))\n",
    "sns.heatmap(confusion_matrix(y_test_1, gnb_y_pred_1), annot=True, cmap='Reds', fmt=\"d\")\n",
    "plt.title('Confusion Matrix of Test 1')\n",
    "plt.savefig('Confusion Matrix of Test 1 Naive Bayes.png')\n",
    "plt.show()"
   ],
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "owXhjHwtbbB9",
    "outputId": "ae949e34-2c73-41f4-f360-d59d8bb01daf"
   },
   "execution_count": null,
   "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"
     }
    }
   ]
  },
  {
   "cell_type": "code",
   "source": [
    "plt.figure(figsize=(20,10))\n",
    "sns.heatmap(confusion_matrix(y_test_2, gnb_y_pred_2), annot=True, cmap='Blues', fmt=\"d\")\n",
    "plt.title('Confusion Matrix of Test 2')\n",
    "plt.savefig('Confusion Matrix of Test 2 Naive Bayes.png')\n",
    "plt.show()"
   ],
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "fOa6CQKIbi3N",
    "outputId": "d3b180fa-1ab1-4275-947d-1fcbda9508b3"
   },
   "execution_count": null,
   "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"
     }
    }
   ]
  },
  {
   "cell_type": "code",
   "source": [
    "plt.figure(figsize=(20,10))\n",
    "\n",
    "y_pred_proba = GNB_Model.predict(x_test)\n",
    "fpr, tpr, _ = roc_curve(y_test,  y_pred_proba)\n",
    "plt.plot(fpr, tpr, label=\"Data\")\n",
    "\n",
    "y_pred_proba_1 = GNB_Model.predict(x_test_1)\n",
    "fpr, tpr, _ = roc_curve(y_test_1,  y_pred_proba_1)\n",
    "plt.plot(fpr, tpr, label=\"Test 1\")\n",
    "\n",
    "y_pred_proba_2 = GNB_Model.predict(x_test_2)\n",
    "fpr, tpr, _ = roc_curve(y_test_2,  y_pred_proba_2)\n",
    "plt.plot(fpr, tpr, label=\"Test 2\")\n",
    "\n",
    "plt.title('ROC Naive Bayes')\n",
    "plt.legend(loc=4)\n",
    "plt.savefig('ROC Naive Bayes.png')\n",
    "plt.show()"
   ],
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "lYEtTAP5PxgC",
    "outputId": "c5f4b40f-1452-4c0c-aa33-192d1d22a317"
   },
   "execution_count": null,
   "outputs": [
    {
     "output_type": "display_data",
     "data": {
      "text/plain": [
       "<Figure size 1440x720 with 1 Axes>"
      ],
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     }
    }
   ]
  }
 ]
}