Feature Selection (Test) -malgenome.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "#import data \n",
    "data=pd.read_csv(\"C:/Users/Ayushi Chaudhuri/Desktop/Malware 2.0/Malware-drebin - Copy/data/malgenome215dataset1260malware2539benign.csv\",)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(3799, 216)"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "ro,col = data.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = data.iloc[:,range(0,col-1)] #Feature set\n",
    "y = data.iloc[:,col-1] #Class labels S, B --> Binary Classification \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>transact</th>\n",
       "      <th>bindService</th>\n",
       "      <th>onServiceConnected</th>\n",
       "      <th>ServiceConnection</th>\n",
       "      <th>android.os.Binder</th>\n",
       "      <th>READ_SMS</th>\n",
       "      <th>attachInterface</th>\n",
       "      <th>WRITE_SMS</th>\n",
       "      <th>TelephonyManager.getSubscriberId</th>\n",
       "      <th>Ljava.lang.Class.getCanonicalName</th>\n",
       "      <th>...</th>\n",
       "      <th>READ_EXTERNAL_STORAGE</th>\n",
       "      <th>Ljava.lang.Object.getClass</th>\n",
       "      <th>SET_ORIENTATION</th>\n",
       "      <th>DEVICE_POWER</th>\n",
       "      <th>EXPAND_STATUS_BAR</th>\n",
       "      <th>GET_TASKS</th>\n",
       "      <th>GLOBAL_SEARCH</th>\n",
       "      <th>GET_PACKAGE_SIZE</th>\n",
       "      <th>SET_PREFERRED_APPLICATIONS</th>\n",
       "      <th>android.intent.action.PACKAGE_CHANGED</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3794</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3795</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3796</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3797</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3798</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3799 rows × 215 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      transact  bindService  onServiceConnected  ServiceConnection  \\\n",
       "0            0            0                   0                  0   \n",
       "1            0            0                   0                  0   \n",
       "2            0            0                   0                  0   \n",
       "3            0            0                   0                  0   \n",
       "4            1            1                   1                  1   \n",
       "...        ...          ...                 ...                ...   \n",
       "3794         0            1                   1                  1   \n",
       "3795         1            1                   1                  1   \n",
       "3796         0            0                   0                  0   \n",
       "3797         1            1                   1                  1   \n",
       "3798         0            0                   0                  0   \n",
       "\n",
       "      android.os.Binder  READ_SMS  attachInterface  WRITE_SMS  \\\n",
       "0                     0         0                0          0   \n",
       "1                     0         1                0          0   \n",
       "2                     0         0                0          0   \n",
       "3                     0         1                0          1   \n",
       "4                     1         1                1          1   \n",
       "...                 ...       ...              ...        ...   \n",
       "3794                  0         0                0          0   \n",
       "3795                  1         0                1          0   \n",
       "3796                  0         0                0          0   \n",
       "3797                  1         0                1          0   \n",
       "3798                  0         0                0          0   \n",
       "\n",
       "      TelephonyManager.getSubscriberId  Ljava.lang.Class.getCanonicalName  \\\n",
       "0                                    0                                  0   \n",
       "1                                    1                                  0   \n",
       "2                                    0                                  0   \n",
       "3                                    1                                  0   \n",
       "4                                    1                                  1   \n",
       "...                                ...                                ...   \n",
       "3794                                 0                                  1   \n",
       "3795                                 0                                  1   \n",
       "3796                                 0                                  0   \n",
       "3797                                 0                                  1   \n",
       "3798                                 0                                  0   \n",
       "\n",
       "      ...  READ_EXTERNAL_STORAGE  Ljava.lang.Object.getClass  SET_ORIENTATION  \\\n",
       "0     ...                      0                           1                0   \n",
       "1     ...                      0                           1                0   \n",
       "2     ...                      0                           0                0   \n",
       "3     ...                      0                           1                0   \n",
       "4     ...                      0                           1                0   \n",
       "...   ...                    ...                         ...              ...   \n",
       "3794  ...                      0                           1                0   \n",
       "3795  ...                      0                           1                0   \n",
       "3796  ...                      0                           1                0   \n",
       "3797  ...                      0                           1                0   \n",
       "3798  ...                      0                           0                0   \n",
       "\n",
       "      DEVICE_POWER  EXPAND_STATUS_BAR  GET_TASKS  GLOBAL_SEARCH  \\\n",
       "0                0                  0          0              0   \n",
       "1                0                  0          0              0   \n",
       "2                0                  0          0              0   \n",
       "3                0                  0          0              0   \n",
       "4                1                  0          0              0   \n",
       "...            ...                ...        ...            ...   \n",
       "3794             0                  0          0              0   \n",
       "3795             0                  0          0              0   \n",
       "3796             0                  0          0              0   \n",
       "3797             0                  0          0              0   \n",
       "3798             0                  0          0              0   \n",
       "\n",
       "      GET_PACKAGE_SIZE  SET_PREFERRED_APPLICATIONS  \\\n",
       "0                    0                           0   \n",
       "1                    0                           0   \n",
       "2                    0                           0   \n",
       "3                    0                           0   \n",
       "4                    0                           0   \n",
       "...                ...                         ...   \n",
       "3794                 0                           0   \n",
       "3795                 0                           0   \n",
       "3796                 0                           0   \n",
       "3797                 0                           0   \n",
       "3798                 0                           0   \n",
       "\n",
       "      android.intent.action.PACKAGE_CHANGED  \n",
       "0                                         0  \n",
       "1                                         0  \n",
       "2                                         0  \n",
       "3                                         0  \n",
       "4                                         0  \n",
       "...                                     ...  \n",
       "3794                                      0  \n",
       "3795                                      0  \n",
       "3796                                      0  \n",
       "3797                                      0  \n",
       "3798                                      0  \n",
       "\n",
       "[3799 rows x 215 columns]"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Univariate Feature Selection for Logistic Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "--------------------------------\n",
      "Percentage: 0.4\n",
      "Selected feature:86\n",
      "Accuracy:0.9855263157894737\n",
      "F1-score:0.9780439121756487\n",
      "MCC-score:0.9672533665614411\n",
      "--------------------------------\n",
      "--------------------------------\n",
      "Percentage: 0.43\n",
      "Selected feature:92\n",
      "Accuracy:0.9855263157894737\n",
      "F1-score:0.978131212723658\n",
      "MCC-score:0.9673194209594288\n",
      "--------------------------------\n",
      "--------------------------------\n",
      "Percentage: 0.46\n",
      "Selected feature:98\n",
      "Accuracy:0.9842105263157894\n",
      "F1-score:0.9760956175298805\n",
      "MCC-score:0.9643077982764424\n",
      "--------------------------------\n",
      "--------------------------------\n",
      "Percentage: 0.48\n",
      "Selected feature:103\n",
      "Accuracy:0.9868421052631579\n",
      "F1-score:0.9799196787148595\n",
      "MCC-score:0.9702051076094816\n",
      "--------------------------------\n",
      "--------------------------------\n",
      "Percentage: 0.51\n",
      "Selected feature:109\n",
      "Accuracy:0.9868421052631579\n",
      "F1-score:0.98\n",
      "MCC-score:0.9702135483728812\n",
      "--------------------------------\n",
      "--------------------------------\n",
      "Percentage: 0.55\n",
      "Selected feature:118\n",
      "Accuracy:0.9881578947368421\n",
      "F1-score:0.9819639278557115\n",
      "MCC-score:0.9731884944120393\n",
      "--------------------------------\n"
     ]
    }
   ],
   "source": [
    "import math\n",
    "from sklearn.feature_selection import SelectKBest\n",
    "from sklearn.feature_selection import chi2\n",
    "\n",
    "from sklearn import preprocessing\n",
    "lb = preprocessing.LabelBinarizer()\n",
    "yb = lb.fit_transform(y).ravel()\n",
    "\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "\n",
    "from sklearn.metrics import confusion_matrix\n",
    "from sklearn.metrics import matthews_corrcoef\n",
    "from sklearn.metrics import f1_score, accuracy_score\n",
    "\n",
    "\n",
    "#selection_percentage = [.25,.40,.55,.70,.85,1]\n",
    "selection_percentage = [.40,.43,.46,.48,.51,.55]\n",
    "results_all =[]\n",
    "\n",
    "all_importance = []\n",
    "\n",
    "for i in selection_percentage:\n",
    "    sl = math.floor((col-1)*i)\n",
    "    X_new = SelectKBest(chi2, k=sl).fit_transform(X, y)\n",
    "    X_train,X_test, y_train, y_test =train_test_split(X_new,yb,test_size=0.20, random_state = 40)\n",
    "    clf = LogisticRegression(random_state=100,solver='saga',max_iter=300).fit(X_train, y_train)\n",
    "    \n",
    "    # get importance\n",
    "    importance = clf.coef_[0]\n",
    "    \n",
    "    all_importance.append(importance)\n",
    "    \n",
    "    y_pred_lr =clf.predict(X_test)\n",
    "    \n",
    "    mc_lr=matthews_corrcoef(y_test, y_pred_lr)\n",
    "    f1_lr=f1_score(y_test, y_pred_lr)\n",
    "\n",
    "    acc_lr= accuracy_score(y_test, y_pred_lr)\n",
    "    print('--------------------------------')\n",
    "    print('Percentage: {}'.format(i))\n",
    "    print('Selected feature:{}'.format(sl))\n",
    "    print('Accuracy:{}'.format(acc_lr))\n",
    "    print('F1-score:{}'.format(f1_lr))\n",
    "    print('MCC-score:{}'.format(mc_lr))\n",
    "    print('--------------------------------')\n",
    "    \n",
    "    results_all.append([i, sl, acc_lr,f1_lr,mc_lr])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.40</td>\n",
       "      <td>86</td>\n",
       "      <td>0.985526</td>\n",
       "      <td>0.978044</td>\n",
       "      <td>0.967253</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.43</td>\n",
       "      <td>92</td>\n",
       "      <td>0.985526</td>\n",
       "      <td>0.978131</td>\n",
       "      <td>0.967319</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.46</td>\n",
       "      <td>98</td>\n",
       "      <td>0.984211</td>\n",
       "      <td>0.976096</td>\n",
       "      <td>0.964308</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.48</td>\n",
       "      <td>103</td>\n",
       "      <td>0.986842</td>\n",
       "      <td>0.979920</td>\n",
       "      <td>0.970205</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.51</td>\n",
       "      <td>109</td>\n",
       "      <td>0.986842</td>\n",
       "      <td>0.980000</td>\n",
       "      <td>0.970214</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.55</td>\n",
       "      <td>118</td>\n",
       "      <td>0.988158</td>\n",
       "      <td>0.981964</td>\n",
       "      <td>0.973188</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      0    1         2         3         4\n",
       "0  0.40   86  0.985526  0.978044  0.967253\n",
       "1  0.43   92  0.985526  0.978131  0.967319\n",
       "2  0.46   98  0.984211  0.976096  0.964308\n",
       "3  0.48  103  0.986842  0.979920  0.970205\n",
       "4  0.51  109  0.986842  0.980000  0.970214\n",
       "5  0.55  118  0.988158  0.981964  0.973188"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results_all = pd.DataFrame(results_all)\n",
    "results_all"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt \n",
    "plt.plot(results_all.iloc[:,0],results_all.iloc[:,2], 'o--') #percentage selection vs accuracy\n",
    "plt.xlabel('% of feature selected')\n",
    "plt.ylabel('Accurcay')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Max accurcay is 98.82% for 55.00000000000001% of the whole features\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "acc= results_all.iloc[:,2] #aacuracy values\n",
    "max_indx = np.argmax(acc)\n",
    "importance  = all_importance[max_indx] #fetching max accurcay parameters of LR\n",
    "\n",
    "#summarize feature importance\n",
    "# for i,v in enumerate(importance):\n",
    "#     print('Feature: %0d, Score: %.5f' % (i,v))\n",
    "\n",
    "print('Max accurcay is {}% for {}% of the whole features'.format(round(acc[max_indx]*100,2),selection_percentage[max_indx]*100))\n",
    "#plot feature importance\n",
    "plt.bar([X for X in range(len(importance))], importance)\n",
    "plt.xlabel('Feature index')\n",
    "plt.ylabel('Importance based on LR weights')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Selected indices based on importance Features are:\n",
      "[  5   7   8  11  13  16  20  21  23  24  25  26  29  31  35  36  37  39\n",
      "  45  46  49  52  53  55  59  71  72  73  78  80  89  91  95  98  99 107\n",
      " 110 111 113 116]\n"
     ]
    }
   ],
   "source": [
    "indx = np.where(importance >0)\n",
    "np.size(indx)\n",
    "selection=[]\n",
    "selection.append(indx)\n",
    "selection = selection[0][0]\n",
    "print('Selected indices based on importance Features are:')\n",
    "print(selection)\n",
    "\n",
    "# X_selected = X.iloc[:,selection]\n",
    "# y_selected = yb.iloc[selection,0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.6.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}