1.5. Bagging & Randon Forest.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn.datasets import load_digits\n",
    "from sklearn import ensemble, cross_validation, learning_curve, metrics, tree\n",
    "import numpy as np\n",
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Populating the interactive namespace from numpy and matplotlib\n"
     ]
    }
   ],
   "source": [
    "%pylab inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "digits = load_digits()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "X = digits.data\n",
    "y = digits.target"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.828694317164\n"
     ]
    }
   ],
   "source": [
    "classifier = tree.DecisionTreeClassifier()\n",
    "scores = cross_validation.cross_val_score(classifier, X, y, cv=10)\n",
    "print(scores.mean())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def write_array_to_file(answers, filename):\n",
    "    with open(filename, \"w\") as fout:\n",
    "        fout.write(\" \".join([str(num) for num in answers]))\n",
    "        \n",
    "def write_one_to_file(answer, filename):\n",
    "    with open(filename, \"w\") as fout:\n",
    "        fout.write(answer)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "write_one_to_file(str(scores.mean()), \"4_1.txt\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.928751810932\n"
     ]
    }
   ],
   "source": [
    "begging = ensemble.BaggingClassifier(n_estimators=100)\n",
    "score = cross_validation.cross_val_score(begging, X, y, cv=10).mean()\n",
    "print(score)\n",
    "write_one_to_file(str(score), \"4_2.txt\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.935466581733\n"
     ]
    }
   ],
   "source": [
    "beg = ensemble.BaggingClassifier(n_estimators=100, max_features = 8)\n",
    "score = cross_validation.cross_val_score(beg, X, y, cv=10).mean()\n",
    "print(score)\n",
    "write_one_to_file(str(score), \"4_3.txt\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.954400868763\n"
     ]
    }
   ],
   "source": [
    "beg = ensemble.BaggingClassifier(tree.DecisionTreeClassifier(max_features = 8), n_estimators=100)\n",
    "score = cross_validation.cross_val_score(beg, X, y, cv=10).mean()\n",
    "print(score)\n",
    "write_one_to_file(str(score), \"4_4.txt\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "?ensemble.RandomForestClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "rf_classifier_low_depth = ensemble.RandomForestClassifier(n_estimators = 50, max_depth = 2, random_state = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "train_sizes, train_scores, test_scores = learning_curve.learning_curve(rf_classifier_low_depth, X, y, \n",
    "                                                                       train_sizes=np.arange(0.1,1., 0.2), \n",
    "                                                                       cv=3, scoring='accuracy')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[ 119  358  597  836 1075]\n",
      "[ 0.95238095  0.89571695  0.87493021  0.87320574  0.83813953]\n",
      "[ 0.6426317   0.74119521  0.7689446   0.79291802  0.79686082]\n"
     ]
    }
   ],
   "source": [
    "print train_sizes\n",
    "print train_scores.mean(axis = 1)\n",
    "print test_scores.mean(axis = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0xa3f4d30>"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX0AAAD7CAYAAACG50QgAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xa337b70>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "pylab.grid(True)\n",
    "pylab.plot(train_sizes, train_scores.mean(axis = 1), 'g-', marker='o', label='train')\n",
    "pylab.plot(train_sizes, test_scores.mean(axis = 1), 'r-', marker='o', label='test')\n",
    "pylab.ylim((0.0, 1.05))\n",
    "pylab.legend(loc='lower right')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "?ensemble.RandomForestClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 142,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "rf_classifier_big_depth = ensemble.RandomForestClassifier(n_estimators = 20, max_features = 10, max_depth = 100, random_state = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 143,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "train_sizes, train_scores, test_scores = learning_curve.learning_curve(rf_classifier_big_depth, X, y, \n",
    "                                                                       train_sizes=np.arange(0.1,1., 0.2), \n",
    "                                                                       cv=3, scoring='accuracy')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 144,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[ 119  358  597  836 1075]\n",
      "[ 1.          1.          0.99944165  1.          1.        ]\n",
      "[ 0.74016012  0.83756568  0.89707115  0.90763889  0.91876592]\n"
     ]
    }
   ],
   "source": [
    "print train_sizes\n",
    "print train_scores.mean(axis = 1)\n",
    "print test_scores.mean(axis = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 145,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0xd660588>"
      ]
     },
     "execution_count": 145,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX0AAAD7CAYAAACG50QgAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nSoiKvohICVHRFxEpISr6IiIl5P8D8DesGh8gLlAAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xd190cf8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "pylab.grid(True)\n",
    "pylab.plot(train_sizes, train_scores.mean(axis = 1), 'g-', marker='o', label='train')\n",
    "pylab.plot(train_sizes, test_scores.mean(axis = 1), 'r-', marker='o', label='test')\n",
    "pylab.ylim((0.0, 1.05))\n",
    "pylab.legend(loc='lower right')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "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": 0
}