Part 2(a). Logistic Regression (classifying activities)1.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Part 2a. Logistic Regression(Classifying Activities)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 1. Load libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import seaborn as sb\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "\n",
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2. Split train-test datasets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_train = pd.read_csv(\"source/train.csv\")\n",
    "df_test = pd.read_csv(\"source/test.csv\")\n",
    "\n",
    "y_train = df_train.Activity\n",
    "x_train = df_train.drop(['subject', 'Activity'], axis=1)\n",
    "\n",
    "y_test = df_test.Activity\n",
    "x_test = df_test.drop(['subject', 'Activity'], axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3. Build Logistic Regression before Hyperparameter Tuning"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.metrics import confusion_matrix\n",
    "\n",
    "logistic_model1 = LogisticRegression()\n",
    "logistic_model1.fit(x_train, y_train)\n",
    "y_pred = logistic_model1.predict(x_test)\n",
    "cm = confusion_matrix(y_test, y_pred)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "####  a) Accuracy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.9579233118425518\n"
     ]
    }
   ],
   "source": [
    "accuracy = logistic_model1.score(x_test, y_test)\n",
    "print(\"Accuracy:\",accuracy)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### b) Confusion Table"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 864x576 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(figsize=(12, 8))\n",
    "ax.imshow(cm, cmap=plt.cm.Blues)\n",
    "ax.figure.colorbar(ax.imshow(cm, cmap=plt.cm.Blues), ax=ax)\n",
    "ax.set(\n",
    "    xticks=np.arange(cm.shape[1]),\n",
    "    yticks=np.arange(cm.shape[0]),\n",
    "    xticklabels=np.unique(y_pred),\n",
    "    yticklabels=np.unique(y_pred),\n",
    "    ylabel=\"True label\",\n",
    "    xlabel=\"Predicted label\",\n",
    ")\n",
    "plt.xticks(rotation = 90)\n",
    "thresh = cm.max() / 2.\n",
    "for i in range(cm.shape[0]):\n",
    "        for j in range(cm.shape[1]):\n",
    "            ax.text(j, i, int(cm[i, j]),ha=\"center\", va=\"center\",color=\"white\" if cm[i, j] > thresh else \"black\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 4. Build Logistic Regression with Hyperparameter Tuning"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "from scipy.stats import loguniform\n",
    "from sklearn.model_selection import RandomizedSearchCV\n",
    "space = dict()\n",
    "space['solver'] = ['lbfgs']\n",
    "space['penalty'] = ['l1', 'l2']\n",
    "space['C'] = loguniform(1e-5, 10)\n",
    "search = RandomizedSearchCV(logistic_model1, space, n_iter=30, scoring='accuracy', n_jobs=-1, cv=5, random_state=1)\n",
    "hypermodel = search.fit(x_train,y_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### a) Accuracy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.9596199524940617\n"
     ]
    }
   ],
   "source": [
    "y_pred1 = hypermodel.predict(x_test)\n",
    "cm1 = confusion_matrix(y_test, y_pred1)\n",
    "accuracy = hypermodel.score(x_test, y_test)\n",
    "print(\"Accuracy:\",accuracy)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### b) Confusion Table"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 864x576 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(figsize=(12, 8))\n",
    "ax.imshow(cm1, cmap=plt.cm.Blues)\n",
    "ax.figure.colorbar(ax.imshow(cm1, cmap=plt.cm.Blues), ax=ax)\n",
    "ax.set(\n",
    "    xticks=np.arange(cm1.shape[1]),\n",
    "    yticks=np.arange(cm1.shape[0]),\n",
    "    xticklabels=np.unique(y_pred1),\n",
    "    yticklabels=np.unique(y_pred1),\n",
    "    ylabel=\"True label\",\n",
    "    xlabel=\"Predicted label\",\n",
    ")\n",
    "plt.xticks(rotation = 90)\n",
    "thresh = cm1.max() / 2.\n",
    "for i in range(cm1.shape[0]):\n",
    "        for j in range(cm1.shape[1]):\n",
    "            ax.text(j, i, int(cm1[i, j]),ha=\"center\", va=\"center\",color=\"white\" if cm1[i, j] > thresh else \"black\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 5. Finding out feature importance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import math\n",
    "w=logistic_model1.coef_[0]\n",
    "df2 = pd.DataFrame(df_train.columns, columns = [\"feature\"])\n",
    "logistic_feature_importance = df2[:-2]\n",
    "logistic_feature_importance[\"importance\"] = pow(math.e, w)\n",
    "logistic_feature_importance = logistic_feature_importance.sort_values(by = [\"importance\"], ascending=False)\n",
    "sb.barplot(x=\"importance\", y=\"feature\", data=logistic_feature_importance.sort_values(by=\"importance\", ascending=False).head(30))\n",
    "plt.title('LightGBM Features (avg over folds)')\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 6. Grouping similiar feature, dropping them one by one and testing accuracy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "drop fBodyAcc\n",
      "Accuracy after removing current main feature: 0.9613165931455717\n",
      "\n",
      "\n",
      "drop fBodyGyro\n",
      "Accuracy after removing current main feature: 0.9541907024092298\n",
      "\n",
      "\n",
      "drop fBodyAccJerk\n",
      "Accuracy after removing current main feature: 0.9575839837122497\n",
      "\n",
      "\n",
      "drop tGravityAcc\n",
      "Accuracy after removing current main feature: 0.9521547336274178\n",
      "\n",
      "\n",
      "drop tBodyAcc\n",
      "Accuracy after removing current main feature: 0.9524940617577197\n",
      "\n",
      "\n",
      "drop tBodyGyroJerk\n",
      "Accuracy after removing current main feature: 0.9535120461486257\n",
      "\n",
      "\n",
      "drop tBodyGyro\n",
      "Accuracy after removing current main feature: 0.9412962334577536\n",
      "\n",
      "\n",
      "drop tBodyAccJerk\n",
      "Accuracy after removing current main feature: 0.9558873430607397\n",
      "\n",
      "\n",
      "drop tBodyAccMag\n",
      "Accuracy after removing current main feature: 0.9582626399728538\n",
      "\n",
      "\n",
      "drop tGravityAccMag\n",
      "Accuracy after removing current main feature: 0.9582626399728538\n",
      "\n",
      "\n",
      "drop tBodyAccJerkMag\n",
      "Accuracy after removing current main feature: 0.9606379368849678\n",
      "\n",
      "\n",
      "drop tBodyGyroMag\n",
      "Accuracy after removing current main feature: 0.9599592806243638\n",
      "\n",
      "\n",
      "drop tBodyGyroJerkMag\n",
      "Accuracy after removing current main feature: 0.9575839837122497\n",
      "\n",
      "\n",
      "drop fBodyAccMag\n",
      "Accuracy after removing current main feature: 0.9582626399728538\n",
      "\n",
      "\n",
      "drop fBodyBodyAccJerkMag\n",
      "Accuracy after removing current main feature: 0.9572446555819477\n",
      "\n",
      "\n",
      "drop fBodyBodyGyroMag\n",
      "Accuracy after removing current main feature: 0.9586019681031558\n",
      "\n",
      "\n",
      "drop fBodyBodyGyroJerkMag\n",
      "Accuracy after removing current main feature: 0.9592806243637597\n",
      "\n",
      "\n",
      "drop angle\n",
      "Accuracy after removing current main feature: 0.9548693586698337\n",
      "\n",
      "\n"
     ]
    }
   ],
   "source": [
    "main_feature_list = ['fBodyAcc','fBodyGyro','fBodyAccJerk','tGravityAcc','tBodyAcc','tBodyGyroJerk','tBodyGyro','tBodyAccJerk',\n",
    "                     'tBodyAccMag','tGravityAccMag','tBodyAccJerkMag','tBodyGyroMag','tBodyGyroJerkMag','fBodyAccMag',\n",
    "                     'fBodyBodyAccJerkMag','fBodyBodyGyroMag','fBodyBodyGyroJerkMag','angle']\n",
    "for feature in main_feature_list:\n",
    "    print(\"drop \" + feature)\n",
    "    X_train_drop_main_feature = x_train.drop(list(x_train.filter(regex = feature)),axis=1)\n",
    "    X_test_drop_main_feature = x_test.drop(list(x_test.filter(regex = feature)),axis=1)\n",
    "\n",
    "    logistic_model2 = LogisticRegression()\n",
    "    logistic_model2.fit(X_train_drop_main_feature, y_train)\n",
    "    y_pred = logistic_model2.predict(X_test_drop_main_feature)\n",
    "    accuracy = logistic_model2.score(X_test_drop_main_feature, y_test)\n",
    "    print(\"Accuracy after removing current main feature:\",accuracy)\n",
    "    print(\"\\n\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 7. Accuracy after removing top 10% feature importance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy after removing top 10 percent feature importance: 0.9582626399728538\n"
     ]
    }
   ],
   "source": [
    "top10p_features = logistic_feature_importance.sort_values(by=\"importance\", ascending=False).head(56)\n",
    "x_train_afterDrop10p = x_train.drop(top10p_features[\"feature\"],axis=1)\n",
    "x_test_afterDrop10p = x_test.drop(top10p_features[\"feature\"],axis=1)\n",
    "\n",
    "hypermodel3 = search.fit(x_train_afterDrop10p,y_train)\n",
    "accuracy = hypermodel3.score(x_test_afterDrop10p, y_test)\n",
    "print(\"Accuracy after removing top 10 percent feature importance:\",accuracy)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 8. Accuracy after removing top 30% feature importance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy after removing top 30 percent feature importance: 0.9467254835425857\n"
     ]
    }
   ],
   "source": [
    "top30p_features = logistic_feature_importance.sort_values(by=\"importance\", ascending=False).head(170)\n",
    "x_train_afterDrop30p = x_train.drop(top30p_features[\"feature\"],axis=1)\n",
    "x_test_afterDrop30p = x_test.drop(top30p_features[\"feature\"],axis=1)\n",
    "\n",
    "hypermodel4 = search.fit(x_train_afterDrop30p,y_train)\n",
    "accuracy = hypermodel4.score(x_test_afterDrop30p, y_test)\n",
    "print(\"Accuracy after removing top 30 percent feature importance:\",accuracy)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 9. Accuracy after removing top 50% feature importance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy after removing top 50 percent feature importance: 0.9507974211062097\n"
     ]
    }
   ],
   "source": [
    "top50p_features = logistic_feature_importance.sort_values(by=\"importance\", ascending=False).head(280)\n",
    "x_train_afterDrop50p = x_train.drop(top50p_features[\"feature\"],axis=1)\n",
    "x_test_afterDrop50p = x_test.drop(top50p_features[\"feature\"],axis=1)\n",
    "\n",
    "hypermodel5 = search.fit(x_train_afterDrop50p,y_train)\n",
    "accuracy = hypermodel5.score(x_test_afterDrop50p, y_test)\n",
    "print(\"Accuracy after removing top 50 percent feature importance:\",accuracy)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 10. Retesting by completely removing accelerometer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Acccuracy after removing accelerometer data: 0.9114353579911775\n"
     ]
    }
   ],
   "source": [
    "\n",
    "x_train_DropAcc = df_train.drop(['subject', 'Activity'], axis=1)\n",
    "x_train_DropAcc = x_train_DropAcc[x_train_DropAcc.columns.drop(list(x_train_DropAcc.filter(regex='Acc')))]\n",
    "x_test_DropAcc = df_test.drop(['subject', 'Activity'], axis=1)\n",
    "x_test_DropAcc = x_test_DropAcc[x_test_DropAcc.columns.drop(list(x_test_DropAcc.filter(regex='Acc')))]\n",
    "\n",
    "hypermodel7 = search.fit(x_train_DropAcc,y_train)\n",
    "accuracy = hypermodel7.score(x_test_DropAcc, y_test)\n",
    "print(\"Acccuracy after removing accelerometer data:\",accuracy)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 11. Retesting by completely removing gyroscope"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Acccuracy after removing gyroscope data: 0.9053274516457415\n"
     ]
    }
   ],
   "source": [
    "x_train_DropGyro = df_train.drop(['subject', 'Activity'], axis=1)\n",
    "x_train_DropGyro = x_train_DropGyro[x_train_DropGyro.columns.drop(list(x_train_DropGyro.filter(regex='Gyro')))]\n",
    "x_test_DropGyro = df_test.drop(['subject', 'Activity'], axis=1)\n",
    "x_test_DropGyro = x_test_DropGyro[x_test_DropGyro.columns.drop(list(x_test_DropGyro.filter(regex='Gyro')))]\n",
    "\n",
    "hypermodel8 = search.fit(x_train_DropGyro,y_train)\n",
    "accuracy = hypermodel8.score(x_test_DropGyro, y_test)\n",
    "print(\"Acccuracy after removing gyroscope data:\",accuracy)"
   ]
  }
 ],
 "metadata": {
  "interpreter": {
   "hash": "b3ba2566441a7c06988d0923437866b63cedc61552a5af99d1f4fb67d367b25f"
  },
  "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": "3.9.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}