random_forest-without-pca.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import scipy as sp\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import accuracy_score \n",
    "from sklearn.model_selection import cross_val_score , KFold\n",
    "\n",
    "#from sklearn import linear_model, cross_validation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0\n"
     ]
    }
   ],
   "source": [
    "df = pd.read_csv('/home/amal/Téléchargements/train1.csv')\n",
    "\n",
    "print(df.isnull().values.sum().sum())  # check if any value is Null in DataFrame\n",
    "features = list(df.columns[1:94])\n",
    "classe = df['target']\n",
    "attribus = df[features]\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "#X_train, X_test, y_train, y_test = train_test_split( attribus, classe, test_size=0.1, random_state=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "\n",
    "from sklearn.feature_selection import SelectFromModel\n",
    "rfe = RandomForestClassifier()\n",
    "rfe.fit(attribus, classe)\n",
    "rfe.feature_importances_  \n",
    "model = SelectFromModel(rfe, prefit=True)\n",
    "X_new = model.transform(attribus)\n",
    "X_new.shape               \n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.78840757422321672, 0.78847148647734699, 0.7886168850722669, 0.78884313230591929, 0.78848826355842283, 0.78861739177852486]\n",
      "[0.78863368791483102, 0.7872431253103831, 0.79003954625060746, 0.78818193061795738, 0.78861786221094354, 0.7899428406759309]\n",
      "[0.79008824928731325, 0.78934527190210213, 0.78858561461858079, 0.78835912154738741, 0.78970053757996062, 0.7883584480010567]\n"
     ]
    }
   ],
   "source": [
    "#https://www.analyticsvidhya.com/blog/2015/06/tuning-random-forest-model/\n",
    "    \n",
    "max_features = [15, 16  , 30, 40 , 50 , 60, 70 , 80 , 90 ]\n",
    "k_scores = []\n",
    "k_scores1 = []\n",
    "k_scores2 = []\n",
    "for feature in max_features:\n",
    "    rf = RandomForestClassifier(n_estimators=1000 ,max_features=feature)\n",
    "    scores = cross_val_score(rf, attribus,classe,  cv=10, scoring = 'accuracy')\n",
    "    scores1 = cross_val_score(rf, attribus,classe,  cv=10, scoring = 'precision_micro')\n",
    "    scores2 = cross_val_score(rf, attribus,classe,  cv=10, scoring = 'recall_micro')\n",
    "    k_scores.append(scores.mean())\n",
    "    k_scores1.append(scores1.mean())\n",
    "    k_scores2.append(scores2.mean())\n",
    "print k_scores\n",
    "print k_scores1\n",
    "print k_scores2\n",
    "\n",
    "\n",
    "                       \n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'max_features' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-44-5e91e3cde310>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m      1\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mmatplotlib\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mpyplot\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0mplt\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      2\u001b[0m \u001b[0mget_ipython\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmagic\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34mu'matplotlib inline'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 3\u001b[0;31m \u001b[0mplt\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mplot\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mmax_features\u001b[0m \u001b[0;34m,\u001b[0m\u001b[0mk_scores\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      4\u001b[0m \u001b[0mplt\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mxlabel\u001b[0m \u001b[0;34m(\u001b[0m\u001b[0;34m'number of dimensions'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      5\u001b[0m \u001b[0mplt\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mylabel\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m'score'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mNameError\u001b[0m: name 'max_features' is not defined"
     ]
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline \n",
    "plt.plot(max_features ,k_scores)\n",
    "plt.xlabel ('number of dimensions')\n",
    "plt.ylabel('score')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.text.Text at 0x7f2b237f98d0>"
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjgAAAF5CAYAAACSmVCiAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nzMzMKuf/AcHv6xvEJKJcAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f2b31164d10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline \n",
    "plt.plot(max_features ,k_scores1)\n",
    "plt.xlabel ('number of dimensions')\n",
    "plt.ylabel('score')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.text.Text at 0x7f2b2372cd10>"
      ]
     },
     "execution_count": 73,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjgAAAF5CAYAAACSmVCiAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f2b237074d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline \n",
    "plt.plot(max_features ,k_scores2)\n",
    "plt.xlabel ('number of dimensions')\n",
    "plt.ylabel('Score')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(\"score d'entrainement:\", 0.97459517114321725)\n"
     ]
    }
   ],
   "source": [
    "#rf.fit(X_train, y_train) # fit the data to the algorithm\n",
    "#1.b. n_estimators :\n",
    "#This is the number of trees you want to build before taking the maximum voting or averages of predictions. Higher number of trees give you better performance but makes your code slower. You should choose as high value\n",
    "#as your processor can handle because this makes your predictions stronger and more stable.\n",
    "\n",
    "# choose the precision i need to improove who are true positive\n",
    "\n",
    "rfe = RandomForestClassifier(n_estimators=10 ,max_features=25, min_samples_leaf =1, min_samples_split =5)\n",
    "rfe.fit(attribus, classe)\n",
    "print(\"score d'entrainement:\", rfe.score(attribus,classe))\n",
    "#scores = cross_val_score(rfe, attribus,classe,  cv=10, scoring = 'accuracy')\n",
    "#print scores.mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "#rf.fit(X_train,y_train,)\n",
    "#dec = rf.predict(X_test)\n",
    "#print (' accuracy is ' , accuracy_score(y_test,dec))\n",
    "\n",
    "#print(rf.score(X_test, y_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "predict result is: \n",
      "[4 6 6 ..., 3 2 2]\n",
      "(144368,)\n"
     ]
    }
   ],
   "source": [
    "df1 = pd.read_csv('/home/amal/Téléchargements/test.csv')\n",
    "test_data = df1.as_matrix()\n",
    "test_data = np.delete(test_data, 0 , axis=1)\n",
    "#print(test_data)\n",
    "#print (test_data.shape)\n",
    "print 'predict result is: '\n",
    "prediction= rfe.predict(test_data)\n",
    "#print(prediction)\n",
    "print (prediction)\n",
    "print(prediction.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[     1      0      0 ...,      0      0      0]\n",
      " [     2      0      0 ...,      0      0      0]\n",
      " [     3      0      0 ...,      0      0      0]\n",
      " ..., \n",
      " [144366      0      0 ...,      0      0      0]\n",
      " [144367      0      1 ...,      0      0      0]\n",
      " [144368      0      1 ...,      0      0      0]]\n"
     ]
    }
   ],
   "source": [
    "#submission file\n",
    "tmp_array = [0,0,0,0,0,0,0,0,0,0]\n",
    "output_matrix = []\n",
    "for i in range(0,prediction.shape[0]):\n",
    "    tmp_array[prediction[i]]=1\n",
    "    tmp_array[0]=i+1\n",
    "    output_matrix.append(tmp_array[0:10])\n",
    "    tmp_array[prediction[i]] = 0\n",
    "print(np.array(output_matrix))\n",
    "np.savetxt(\"/home/amal/random_search_l10_.csv\", np.array(output_matrix), fmt='%i', delimiter=\",\" , header='id,Class_1,Class_2,Class_3,Class_4,Class_5,Class_6,Class_7,Class_8,Class_9', comments='')"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}