27_univariate_feature_selection.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd \n",
    "import numpy as np \n",
    "\n",
    "from sklearn.datasets import load_iris \n",
    "from sklearn.model_selection import train_test_split "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "X, y = load_iris(return_X_y=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "#generating noise, and adding it to X \n",
    "noise = np.random.RandomState(42).uniform(0, 0.1, size=(X.shape[0], 20))\n",
    "X = np.hstack((X, noise))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "#splitting dataset \n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, stratify=y, random_state=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Univariate feature Selection \n",
    "#with F-test for feature scoring \n",
    "\n",
    "from sklearn.feature_selection import SelectKBest, f_classif \n",
    "\n",
    "selector = SelectKBest(f_classif, k=4)\n",
    "selector.fit(X_train, y_train)\n",
    "scores = -np.log10(selector.pvalues_)\n",
    "scores /= scores.max()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([3.72189526e-01, 1.92112431e-01, 9.98924450e-01, 1.00000000e+00,\n",
       "       1.61525527e-02, 1.78421693e-03, 3.36072503e-02, 9.34126229e-04,\n",
       "       8.38039241e-03, 4.46122522e-03, 2.47289009e-03, 3.04121968e-03,\n",
       "       4.16671859e-03, 6.37306186e-03, 2.79418271e-03, 5.46727836e-03,\n",
       "       9.36025883e-04, 8.67509750e-03, 2.39127975e-03, 6.33147019e-04,\n",
       "       1.10538184e-03, 7.61511411e-03, 1.15995689e-03, 2.40400211e-02])"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "scores"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "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 matplotlib.pyplot as plt \n",
    "%matplotlib inline \n",
    "\n",
    "X_indices = np.arange(X.shape[-1])\n",
    "plt.figure(1)\n",
    "plt.clf()\n",
    "plt.bar(X_indices - 0.05, scores, width = 0.2)\n",
    "plt.title(\"Feature univariate score\")\n",
    "plt.xlabel(\"Feature number\")\n",
    "plt.ylabel(r\"Univariate score ($-Log(p_{value})$)\")\n",
    "plt.show()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Compare it with SVMs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Classification accuracy without selecting features: 0.789\n"
     ]
    }
   ],
   "source": [
    "from sklearn.pipeline import make_pipeline \n",
    "from sklearn.preprocessing import MinMaxScaler \n",
    "from sklearn.svm import LinearSVC \n",
    "\n",
    "clf = make_pipeline(MinMaxScaler(), LinearSVC())\n",
    "clf.fit(X_train, y_train)\n",
    "print(\"Classification accuracy without selecting features: {:.3f}\".format(clf.score(X_test, y_test)))\n",
    "svm_weights = np.abs(clf[-1].coef_).sum(axis=0)\n",
    "svm_weights /= svm_weights.sum()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "After \"univariate feature selection\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "classification accuracy after univariate feature selection: 0.868\n"
     ]
    }
   ],
   "source": [
    "clf_selected = make_pipeline(SelectKBest(f_classif, k=4), MinMaxScaler(), LinearSVC())\n",
    "clf_selected.fit(X_train, y_train)\n",
    "\n",
    "print(\"classification accuracy after univariate feature selection: {:.3f}\".format(clf_selected.score(X_test, y_test)))\n",
    "\n",
    "svm_weights_selected = np.abs(clf_selected[-1].coef_).sum(axis=0)\n",
    "svm_weights_selected /= svm_weights_selected.sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "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": [
    "plt.bar(X_indices - 0.45, scores, width=0.2, label=r\"Univariate score ($-Log(p_{value})$)\")\n",
    "\n",
    "plt.bar(X_indices - 0.25, svm_weights, width=0.2, label=\"SVM weight\")\n",
    "\n",
    "plt.bar(\n",
    "    X_indices[selector.get_support()] - 0.05,\n",
    "    svm_weights_selected,\n",
    "    width=0.2,\n",
    "    label=\"SVM weights after selection\",\n",
    ")\n",
    "\n",
    "plt.title(\"Comparing feature selection\")\n",
    "plt.xlabel(\"Feature number\")\n",
    "plt.yticks(())\n",
    "plt.axis(\"tight\")\n",
    "plt.legend(loc=\"upper right\")\n",
    "plt.show()"
   ]
  }
 ],
 "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": "3.9.9"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}