collinear_features.ipynb

{
 "cells": [
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In this project, we'll see how linear models work if there are collinear features present in the dataset. To check collinearity between the features, perturbation test will be done, weights and accuracy scores will be calculated and recorded, additional gaussian noise will be added to see these metrics and weights changes. SVM and Logistic Regression algorithms will be used."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# required imports\n",
    "\n",
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\")\n",
    "\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from sklearn.linear_model import SGDClassifier\n",
    "from sklearn.model_selection import GridSearchCV"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "          x         y         z       x*x       2*y  2*z+3*x*x         w  \\\n",
      "0 -0.581066  0.841837 -1.012978 -0.604025  0.841837  -0.665927 -0.536277   \n",
      "1 -0.894309 -0.207835 -1.012978 -0.883052 -0.207835  -0.917054 -0.522364   \n",
      "2 -1.207552  0.212034 -1.082312 -1.150918  0.212034  -1.166507  0.205738   \n",
      "3 -1.364174  0.002099 -0.943643 -1.280666  0.002099  -1.266540 -0.665720   \n",
      "4 -0.737687  1.051772 -1.012978 -0.744934  1.051772  -0.792746 -0.735054   \n",
      "\n",
      "   target  \n",
      "0       0  \n",
      "1       0  \n",
      "2       0  \n",
      "3       0  \n",
      "4       0  \n"
     ]
    }
   ],
   "source": [
    "# csv file\n",
    "data = pd.read_csv('E:/GITHUB REPOS/effect-of-collinear-features-on-linear-models/collinear_features_data.csv')\n",
    "\n",
    "print(data.head())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# splitting the data into X and y\n",
    "X = data.drop(['target'], axis=1)\n",
    "y = data['target'].values"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h1>Finding correlation between the features</h1>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                  x         y         z       x*x       2*y  2*z+3*x*x  \\\n",
      "x          1.000000 -0.205926  0.812458  0.997947 -0.205926   0.996252   \n",
      "y         -0.205926  1.000000 -0.602663 -0.209289  1.000000  -0.261123   \n",
      "z          0.812458 -0.602663  1.000000  0.807137 -0.602663   0.847163   \n",
      "x*x        0.997947 -0.209289  0.807137  1.000000 -0.209289   0.997457   \n",
      "2*y       -0.205926  1.000000 -0.602663 -0.209289  1.000000  -0.261123   \n",
      "2*z+3*x*x  0.996252 -0.261123  0.847163  0.997457 -0.261123   1.000000   \n",
      "w          0.583277 -0.401790  0.674486  0.583803 -0.401790   0.606860   \n",
      "target     0.728290 -0.690684  0.969990  0.719570 -0.690684   0.764729   \n",
      "\n",
      "                  w    target  \n",
      "x          0.583277  0.728290  \n",
      "y         -0.401790 -0.690684  \n",
      "z          0.674486  0.969990  \n",
      "x*x        0.583803  0.719570  \n",
      "2*y       -0.401790 -0.690684  \n",
      "2*z+3*x*x  0.606860  0.764729  \n",
      "w          1.000000  0.641750  \n",
      "target     0.641750  1.000000  \n"
     ]
    }
   ],
   "source": [
    "# prints pairwise correlation betweeen all the columns/features of the dataset\n",
    "# Note - correlation of a feature with itself is unit(1) value. (e.g. corr of x with x is 1)\n",
    "print(data.corr())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plotting heatmap of the correlation matrix of all the features/columns\n",
    "# annot=True gives correlation value of each cell # vmin=-1 and vmax=1 is setting limits of the heatmap (-1 to 1) # cmap=\"YlGnBu\" using different colour for heatmap\n",
    "heatmap = sns.heatmap(data.corr(), annot=True, vmin=-1, vmax=1, cmap=\"YlGnBu\")\n",
    "plt.show()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h1>Logistic Regression</h1>"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h3>1. Finding the best LR model for the given data</h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-1 {color: black;background-color: white;}#sk-container-id-1 pre{padding: 0;}#sk-container-id-1 div.sk-toggleable {background-color: white;}#sk-container-id-1 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-1 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-1 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-1 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-1 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-1 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-1 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-1 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-1 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-1 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-1 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-1 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-1 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-1 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-1 div.sk-item {position: relative;z-index: 1;}#sk-container-id-1 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-1 div.sk-item::before, #sk-container-id-1 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-1 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-1 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-1 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-1 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-1 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-1 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-1 div.sk-label-container {text-align: center;}#sk-container-id-1 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-1 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-1\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>SGDClassifier(loss=&#x27;log&#x27;)</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-1\" type=\"checkbox\" checked><label for=\"sk-estimator-id-1\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">SGDClassifier</label><div class=\"sk-toggleable__content\"><pre>SGDClassifier(loss=&#x27;log&#x27;)</pre></div></div></div></div></div>"
      ],
      "text/plain": [
       "SGDClassifier(loss='log')"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Train Logistic regression on data(X,y)\n",
    "# calling the SGDClassifier() for logistic regression\n",
    "clf = SGDClassifier(loss='log')\n",
    "\n",
    "clf.fit(X,y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-2 {color: black;background-color: white;}#sk-container-id-2 pre{padding: 0;}#sk-container-id-2 div.sk-toggleable {background-color: white;}#sk-container-id-2 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-2 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-2 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-2 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-2 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-2 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-2 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-2 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-2 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-2 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-2 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-2 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-2 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-2 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-2 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-2 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-2 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-2 div.sk-item {position: relative;z-index: 1;}#sk-container-id-2 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-2 div.sk-item::before, #sk-container-id-2 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-2 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-2 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-2 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-2 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-2 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-2 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-2 div.sk-label-container {text-align: center;}#sk-container-id-2 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-2 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-2\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>GridSearchCV(estimator=SGDClassifier(loss=&#x27;log&#x27;),\n",
       "             param_grid={&#x27;alpha&#x27;: array([1.00000000e-04, 7.74263683e-04, 5.99484250e-03, 4.64158883e-02,\n",
       "       3.59381366e-01, 2.78255940e+00, 2.15443469e+01, 1.66810054e+02,\n",
       "       1.29154967e+03, 1.00000000e+04])})</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-2\" type=\"checkbox\" ><label for=\"sk-estimator-id-2\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">GridSearchCV</label><div class=\"sk-toggleable__content\"><pre>GridSearchCV(estimator=SGDClassifier(loss=&#x27;log&#x27;),\n",
       "             param_grid={&#x27;alpha&#x27;: array([1.00000000e-04, 7.74263683e-04, 5.99484250e-03, 4.64158883e-02,\n",
       "       3.59381366e-01, 2.78255940e+00, 2.15443469e+01, 1.66810054e+02,\n",
       "       1.29154967e+03, 1.00000000e+04])})</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-3\" type=\"checkbox\" ><label for=\"sk-estimator-id-3\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">estimator: SGDClassifier</label><div class=\"sk-toggleable__content\"><pre>SGDClassifier(loss=&#x27;log&#x27;)</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-4\" type=\"checkbox\" ><label for=\"sk-estimator-id-4\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">SGDClassifier</label><div class=\"sk-toggleable__content\"><pre>SGDClassifier(loss=&#x27;log&#x27;)</pre></div></div></div></div></div></div></div></div></div></div>"
      ],
      "text/plain": [
       "GridSearchCV(estimator=SGDClassifier(loss='log'),\n",
       "             param_grid={'alpha': array([1.00000000e-04, 7.74263683e-04, 5.99484250e-03, 4.64158883e-02,\n",
       "       3.59381366e-01, 2.78255940e+00, 2.15443469e+01, 1.66810054e+02,\n",
       "       1.29154967e+03, 1.00000000e+04])})"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Find the best hyper prameter alpha with hyper parameter tuning using k-fold cross validation\n",
    "# params here are alpha values that are evenly separated on log scale # reference - https://www.javatpoint.com/numpy-logspace\n",
    "params = {'alpha': np.logspace(-4,4,10)}\n",
    "\n",
    "# calling the GridSearchCV() for k fold cross validation on trained Logistic Regression classifier(clf)\n",
    "gridsearch = GridSearchCV(clf, params)\n",
    "\n",
    "# training the GridSearchCV() on (X,y) data\n",
    "gridsearch.fit(X,y) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'alpha': 0.0001}\n"
     ]
    }
   ],
   "source": [
    "# printing the best/optimal value of alpha \n",
    "print(gridsearch.best_params_) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Creating a new Logistic regression with the best alpha\n",
    "# calling the SGDClassifier() for logistic regression with the best alpha obtained\n",
    "best_model = SGDClassifier(loss='log', alpha= 0.0001)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h3>2. Getting the accuracy and weights with the original data</h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-3 {color: black;background-color: white;}#sk-container-id-3 pre{padding: 0;}#sk-container-id-3 div.sk-toggleable {background-color: white;}#sk-container-id-3 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-3 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-3 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-3 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-3 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-3 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-3 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-3 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-3 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-3 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-3 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-3 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-3 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-3 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-3 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-3 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-3 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-3 div.sk-item {position: relative;z-index: 1;}#sk-container-id-3 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-3 div.sk-item::before, #sk-container-id-3 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-3 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-3 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-3 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-3 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-3 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-3 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-3 div.sk-label-container {text-align: center;}#sk-container-id-3 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-3 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-3\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>SGDClassifier(loss=&#x27;log&#x27;)</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-5\" type=\"checkbox\" checked><label for=\"sk-estimator-id-5\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">SGDClassifier</label><div class=\"sk-toggleable__content\"><pre>SGDClassifier(loss=&#x27;log&#x27;)</pre></div></div></div></div></div>"
      ],
      "text/plain": [
       "SGDClassifier(loss='log')"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# training the 'best_model' with X, y\n",
    "best_model.fit(X,y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1.0\n"
     ]
    }
   ],
   "source": [
    "# Check the accuracy of the model 'best_model_accuracy'\n",
    "# SGDClassifier() provides us with score() to calculate the accuracy of the classifier\n",
    "best_model_accuracy = best_model.score(X,y)\n",
    "\n",
    "# printing the best LR model's accuracy\n",
    "print(best_model_accuracy) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[ 7.19541163 -8.35844337 16.61595972  6.1516865  -8.35844337  7.54265566\n",
      "   4.02827796]]\n"
     ]
    }
   ],
   "source": [
    "# Get the weights W using best_model.coef_\n",
    "# coef() gives us the array of weights (estimated coefficients)\n",
    "W = best_model.coef_\n",
    "\n",
    "# printing the weights of the best model\n",
    "print(W) "
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h3>3. Modifying the original data</h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Add a noise(order of 10^-2) to each element of X and get the new data set X' (X' = X + e)\n",
    "# generating noise(order of 10^-2) of gaussian distribution with mean = 10^-2 and variance = 0 # reference - https://www.researchgate.net/post/How-to-add-some-noise-data-to-my-classification-datasets\n",
    "noise1 = np.random.normal(10**(-2),0,X.shape)\n",
    "\n",
    "# adding the generated noise to the original data\n",
    "X_dash = X + noise1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-4 {color: black;background-color: white;}#sk-container-id-4 pre{padding: 0;}#sk-container-id-4 div.sk-toggleable {background-color: white;}#sk-container-id-4 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-4 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-4 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-4 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-4 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-4 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-4 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-4 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-4 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-4 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-4 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-4 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-4 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-4 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-4 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-4 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-4 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-4 div.sk-item {position: relative;z-index: 1;}#sk-container-id-4 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-4 div.sk-item::before, #sk-container-id-4 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-4 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-4 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-4 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-4 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-4 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-4 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-4 div.sk-label-container {text-align: center;}#sk-container-id-4 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-4 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-4\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>SGDClassifier(loss=&#x27;log&#x27;)</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-6\" type=\"checkbox\" checked><label for=\"sk-estimator-id-6\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">SGDClassifier</label><div class=\"sk-toggleable__content\"><pre>SGDClassifier(loss=&#x27;log&#x27;)</pre></div></div></div></div></div>"
      ],
      "text/plain": [
       "SGDClassifier(loss='log')"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Train the same 'best_model' with data (X', y)\n",
    "best_model.fit(X_dash,y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1.0\n"
     ]
    }
   ],
   "source": [
    "# Check the accuracy of the model 'best_model_accuracy_edited'\n",
    "best_model_accuracy_edited = best_model.score(X_dash,y)\n",
    "\n",
    "# printing the best LR model's accuracy for modified data\n",
    "print(best_model_accuracy_edited) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[  9.61606699 -10.41200524  13.49470823   9.39133964 -10.41200524\n",
      "   10.08262991   4.2013575 ]]\n"
     ]
    }
   ],
   "source": [
    "# Get the weights W' using best_model.coef_\n",
    "# coef() gives us the array of weights (estimated coefficients)\n",
    "W_dash = best_model.coef_\n",
    "\n",
    "# printing the weights of the best model for modified data\n",
    "print(W_dash) "
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h3>4. Checking deviations in metric and weights</h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.0\n"
     ]
    }
   ],
   "source": [
    "# finding the difference between 'best_model_accuracy_edited' and 'best_model_accuracy'\n",
    "difference = float(best_model_accuracy - best_model_accuracy_edited)\n",
    "\n",
    "print(difference)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[2.42065536 2.05356187 3.12125149 3.23965314 2.05356187 2.53997425\n",
      "  0.17307955]]\n"
     ]
    }
   ],
   "source": [
    "# finding the absolute change between each value of W and W' ==> |(W-W')|\n",
    "# abs() used to calculate the absolute values(positive sign numbers only)\n",
    "absolute_change = abs(W - W_dash)\n",
    "\n",
    "print(absolute_change)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[3.2396531439655787, 3.1212514860779983, 2.5399742485031833, 2.420655361324876]\n",
      "x\n",
      "2*z+3*x*x\n",
      "z\n",
      "x*x\n"
     ]
    }
   ],
   "source": [
    "# printing the top 4 features which have higher % change in weights compared to the other features\n",
    "# sorting the features list in descending order(higher to lower) and then accessing the first 4 features(top 4 features)\n",
    "top4_features = sorted(absolute_change[0], reverse=True)[:4] \n",
    "print(top4_features)\n",
    "\n",
    "# code to print the names of top 4 features which are having higher % change in weights\n",
    "# storing the feature names in a list\n",
    "columns = list(data.columns)\n",
    "\n",
    "# argsort() used to get indices that would sort the list # reference - https://numpy.org/doc/stable/reference/generated/numpy.argsort.html\n",
    "for index in np.argsort(absolute_change[0])[-4:]: \n",
    "  print(columns[index])"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Observation:\n",
    "\n",
    "\n",
    "*   For Logistic Regression, we observe that the feature weights changes, after the addition of very little noise to the original data. This indicates the presence of multlicollinearity.\n",
    "*   Even in the change in weights(coefficients), the change in accuracy is zero.\n",
    "\n",
    "*   We cannot use these weights(coefficients) for feature importance directly, as they are highly collinear with each other"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h1>Support Vector Machines</h1>"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h3>1. Finding the best SVM model for given data</h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-5 {color: black;background-color: white;}#sk-container-id-5 pre{padding: 0;}#sk-container-id-5 div.sk-toggleable {background-color: white;}#sk-container-id-5 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-5 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-5 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-5 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-5 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-5 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-5 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-5 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-5 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-5 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-5 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-5 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-5 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-5 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-5 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-5 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-5 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-5 div.sk-item {position: relative;z-index: 1;}#sk-container-id-5 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-5 div.sk-item::before, #sk-container-id-5 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-5 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-5 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-5 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-5 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-5 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-5 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-5 div.sk-label-container {text-align: center;}#sk-container-id-5 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-5 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-5\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>SGDClassifier()</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-7\" type=\"checkbox\" checked><label for=\"sk-estimator-id-7\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">SGDClassifier</label><div class=\"sk-toggleable__content\"><pre>SGDClassifier()</pre></div></div></div></div></div>"
      ],
      "text/plain": [
       "SGDClassifier()"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Training Linear SVM on data(X,y)\n",
    "# calling the SGDClassifier() for Linear SVM # passing loss='hinge' for SVM\n",
    "clf = SGDClassifier(loss='hinge')\n",
    "\n",
    "clf.fit(X,y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-6 {color: black;background-color: white;}#sk-container-id-6 pre{padding: 0;}#sk-container-id-6 div.sk-toggleable {background-color: white;}#sk-container-id-6 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-6 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-6 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-6 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-6 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-6 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-6 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-6 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-6 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-6 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-6 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-6 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-6 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-6 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-6 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-6 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-6 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-6 div.sk-item {position: relative;z-index: 1;}#sk-container-id-6 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-6 div.sk-item::before, #sk-container-id-6 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-6 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-6 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-6 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-6 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-6 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-6 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-6 div.sk-label-container {text-align: center;}#sk-container-id-6 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-6 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-6\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>GridSearchCV(estimator=SGDClassifier(),\n",
       "             param_grid={&#x27;alpha&#x27;: array([1.00000000e-04, 7.74263683e-04, 5.99484250e-03, 4.64158883e-02,\n",
       "       3.59381366e-01, 2.78255940e+00, 2.15443469e+01, 1.66810054e+02,\n",
       "       1.29154967e+03, 1.00000000e+04])})</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-8\" type=\"checkbox\" ><label for=\"sk-estimator-id-8\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">GridSearchCV</label><div class=\"sk-toggleable__content\"><pre>GridSearchCV(estimator=SGDClassifier(),\n",
       "             param_grid={&#x27;alpha&#x27;: array([1.00000000e-04, 7.74263683e-04, 5.99484250e-03, 4.64158883e-02,\n",
       "       3.59381366e-01, 2.78255940e+00, 2.15443469e+01, 1.66810054e+02,\n",
       "       1.29154967e+03, 1.00000000e+04])})</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-9\" type=\"checkbox\" ><label for=\"sk-estimator-id-9\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">estimator: SGDClassifier</label><div class=\"sk-toggleable__content\"><pre>SGDClassifier()</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-10\" type=\"checkbox\" ><label for=\"sk-estimator-id-10\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">SGDClassifier</label><div class=\"sk-toggleable__content\"><pre>SGDClassifier()</pre></div></div></div></div></div></div></div></div></div></div>"
      ],
      "text/plain": [
       "GridSearchCV(estimator=SGDClassifier(),\n",
       "             param_grid={'alpha': array([1.00000000e-04, 7.74263683e-04, 5.99484250e-03, 4.64158883e-02,\n",
       "       3.59381366e-01, 2.78255940e+00, 2.15443469e+01, 1.66810054e+02,\n",
       "       1.29154967e+03, 1.00000000e+04])})"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Finding the best hyper prameter alpha with hyper parameter tuning using k-fold cross validation \n",
    "params = {'alpha': np.logspace(-4,4,10)}\n",
    "\n",
    "# calling the GridSearchCV() for k fold cross validation on trained Linear SVM classifier(clf)\n",
    "gridsearch = GridSearchCV(clf, params) \n",
    "\n",
    "gridsearch.fit(X,y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'alpha': 0.0001}\n"
     ]
    }
   ],
   "source": [
    "# printing the best/optimal value of alpha \n",
    "print(gridsearch.best_params_) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Creating a new Linear SVM with the best alpha \n",
    "best_model = SGDClassifier(loss='hinge', alpha= 0.0001)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h3>2. Getting the accuracy and the weights with the original data</h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-7 {color: black;background-color: white;}#sk-container-id-7 pre{padding: 0;}#sk-container-id-7 div.sk-toggleable {background-color: white;}#sk-container-id-7 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-7 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-7 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-7 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-7 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-7 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-7 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-7 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-7 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-7 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-7 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-7 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-7 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-7 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-7 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-7 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-7 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-7 div.sk-item {position: relative;z-index: 1;}#sk-container-id-7 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-7 div.sk-item::before, #sk-container-id-7 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-7 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-7 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-7 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-7 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-7 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-7 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-7 div.sk-label-container {text-align: center;}#sk-container-id-7 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-7 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-7\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>SGDClassifier()</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-11\" type=\"checkbox\" checked><label for=\"sk-estimator-id-11\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">SGDClassifier</label><div class=\"sk-toggleable__content\"><pre>SGDClassifier()</pre></div></div></div></div></div>"
      ],
      "text/plain": [
       "SGDClassifier()"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# train the 'best_model' with X, y\n",
    "best_model.fit(X,y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1.0\n"
     ]
    }
   ],
   "source": [
    "# Checking the accuracy of the model 'best_model_accuracy'\n",
    "best_model_accuracy = best_model.score(X,y)\n",
    "\n",
    "print(best_model_accuracy)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[ 6.96482656 -8.16865559 23.50976661  5.73875247 -8.16865559  8.00331548\n",
      "   2.38890647]]\n"
     ]
    }
   ],
   "source": [
    "# Getting the weights W using best_model.coef_\n",
    "W = best_model.coef_\n",
    "\n",
    "print(W)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h3>3. Modifying the original data</h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Adding a noise(order of 10^-2) to each element of X and get the new data set X' (X' = X + e)\n",
    "noise2 = np.random.normal(10**(-2),0,X.shape)\n",
    "\n",
    "# adding the generated noise to the original data\n",
    "X_dash = X + noise2 "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-8 {color: black;background-color: white;}#sk-container-id-8 pre{padding: 0;}#sk-container-id-8 div.sk-toggleable {background-color: white;}#sk-container-id-8 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-8 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-8 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-8 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-8 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-8 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-8 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-8 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-8 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-8 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-8 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-8 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-8 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-8 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-8 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-8 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-8 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-8 div.sk-item {position: relative;z-index: 1;}#sk-container-id-8 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-8 div.sk-item::before, #sk-container-id-8 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-8 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-8 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-8 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-8 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-8 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-8 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-8 div.sk-label-container {text-align: center;}#sk-container-id-8 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-8 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-8\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>SGDClassifier()</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-12\" type=\"checkbox\" checked><label for=\"sk-estimator-id-12\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">SGDClassifier</label><div class=\"sk-toggleable__content\"><pre>SGDClassifier()</pre></div></div></div></div></div>"
      ],
      "text/plain": [
       "SGDClassifier()"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Training the same 'best_model' with data (X', y)\n",
    "best_model.fit(X_dash,y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1.0\n"
     ]
    }
   ],
   "source": [
    "# Checking the accuracy of the model 'best_model_accuracy_edited'\n",
    "best_model_accuracy_edited = best_model.score(X_dash,y)\n",
    "\n",
    "print(best_model_accuracy_edited)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[ 9.21845423 -9.88508293 13.46697975 10.18228486 -9.88508293 10.7900483\n",
      "   8.03473596]]\n"
     ]
    }
   ],
   "source": [
    "# Getting the weights W' using best_model.coef_\n",
    "W_dash = best_model.coef_\n",
    "\n",
    "print(W_dash)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h3>4. Checking deviations in metric and weights</h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.0\n"
     ]
    }
   ],
   "source": [
    "# finding the difference between 'best_model_accuracy_edited' and 'best_model_accuracy'\n",
    "difference = float(best_model_accuracy - best_model_accuracy_edited)\n",
    "\n",
    "print(difference)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[ 2.25362767  1.71642734 10.04278686  4.44353239  1.71642734  2.78673282\n",
      "   5.64582949]]\n"
     ]
    }
   ],
   "source": [
    "# finding the absolute change between each value of W and W' ==> |(W-W')|\n",
    "absolute_change = abs(W - W_dash) # abs() used to calculate the absolute values(positive sign numbers only)\n",
    "\n",
    "print(absolute_change)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[10.042786860915847, 5.6458294855132465, 4.44353238807123, 2.786732824386819]\n",
      "2*z+3*x*x\n",
      "x*x\n",
      "w\n",
      "z\n"
     ]
    }
   ],
   "source": [
    "# printing the top 4 features which have higher % change in weights compared to the other features\n",
    "# sorting the features list in descending order(higher to lower) and then accessing the first 4 features(top 4 features)\n",
    "top4_features = sorted(absolute_change[0], reverse=True)[:4] \n",
    "print(top4_features)\n",
    "\n",
    "# code to print the names of top 4 features which are having higher % change in weights\n",
    "# storing the feature names in a list\n",
    "columns = list(data.columns)\n",
    "\n",
    "# argsort() used to get indices that would sort the list # reference - https://numpy.org/doc/stable/reference/generated/numpy.argsort.html\n",
    "for index in np.argsort(absolute_change[0])[-4:]: \n",
    "  print(columns[index])"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Observation:\n",
    "\n",
    "\n",
    "*   For Linear SVM also, we observe that the feature weights changes, after the addition of very little noise to the original data. This indicates the presence of multlicollinearity.\n",
    "*   Even in the change in weights(coefficients), the change in accuracy is zero.\n",
    "\n",
    "*   We cannot use these weights(coefficients) for feature importance directly, as they are highly collinear with each other\n"
   ]
  }
 ],
 "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": "3.10.9"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}