Lab 2. Mikhail Brel.

Lab2.ipynb.json

{
 "metadata": {
  "name": "Lab2"
 },
 "nbformat": 3,
 "nbformat_minor": 0,
 "worksheets": [
  {
   "cells": [
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": "\u041b\u0430\u0431\u043e\u0440\u0430\u0442\u043e\u0440\u043d\u0430\u044f \u0440\u0430\u0431\u043e\u0442\u0430 2. \u0410\u0432\u0442\u043e\u0440 - \u0411\u0440\u0435\u043b\u044c \u041c\u0438\u0445\u0430\u0438\u043b <[email protected]> "
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": "import numpy as np\nimport pandas as pd",
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 291
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": "df = pd.read_csv('train.csv')",
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 292
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": "def convert_class(cls):\n    if cls == '50000+':\n        return 1\n    else:\n        return 0",
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 293
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": "df.CLASS = df.CLASS.apply(convert_class)",
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 294
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": "import pylab as pl",
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 295
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": "import pylab as py",
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 296
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": "\u0421\u0440\u0430\u0432\u043d\u0438\u043c \u0434\u043e\u043b\u0438 \u043c\u0443\u0436\u0447\u0438\u043d \u0438 \u0436\u0435\u043d\u0449\u0438\u043d, \u043f\u043e\u043b\u0443\u0447\u0430\u044e\u0449\u0438\u0445 \u0431\u043e\u043b\u044c\u0448\u0435 50\u041a \u0432 \u0433\u043e\u0434:"
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": "sums = df[['ASEX', 'CLASS']].groupby('ASEX').sum()\ntotals = df[['ASEX']].groupby('ASEX').count().values\ndistr = sums.divide(totals)\ndistr.plot(kind='bar', color=['pink', 'blue']);",
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAW8AAAEoCAYAAACXYXDAAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRGRADG8iIgP6P9sOVB2mlnmiAAAAAElFTkSuQmCC\n",
       "text": "<matplotlib.figure.Figure at 0x1092f5910>"
      }
     ],
     "prompt_number": 297
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": "    \u041c\u0443\u0436\u0447\u0438\u043d\u044b \u0447\u0430\u0449\u0435 \u043e\u043a\u0430\u0437\u044b\u0432\u0430\u044e\u0442\u0441\u044f \u0441\u0440\u0435\u0434\u0438 \u0442\u0435\u0445, \u0447\u0435\u0439 \u0434\u043e\u0445\u043e\u0434 \u0431\u043e\u043b\u044c\u0448\u0435 50\u041a"
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": "\u041f\u043e\u0441\u043c\u043e\u0442\u0440\u0438\u043c, \u043a\u0430\u043a \u0440\u0430\u0441\u043f\u0440\u0435\u0434\u0435\u043b\u044f\u044e\u0442\u0441\u044f \u043a\u043b\u0430\u0441\u0441\u044b \u0432 \u0437\u0430\u0432\u0438\u0441\u0438\u043c\u043e\u0441\u0442\u0438 \u043e\u0442 \u0433\u0440\u0430\u0436\u0434\u0430\u043d\u0441\u0442\u0432\u0430"
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": "sums = df[['PRCITSHP', 'CLASS']].groupby('PRCITSHP').sum()\ntotals = df[['PRCITSHP']].groupby('PRCITSHP').count().values\ndistr = sums.divide(totals)\ndistr.sort(columns=['CLASS'], ascending=True).plot(kind='barh');",
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAjkAAAD7CAYAAACIVIX7AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
       "text": "<matplotlib.figure.Figure at 0x10932cdd0>"
      }
     ],
     "prompt_number": 405
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": "\u0417\u0430\u0431\u0430\u0432\u043d\u043e, \u0447\u0442\u043e \u043d\u0430\u0442\u0443\u0440\u0430\u043b\u0438\u0437\u043e\u0432\u0430\u0432\u0448\u0438\u0435\u0441\u044f \u043f\u0435\u0440\u0435\u0441\u0435\u043b\u0435\u043d\u0446\u044b \u0447\u0430\u0449\u0435 \u043f\u0440\u0438\u0445\u043e\u0434\u044f\u0442 \u043a \u0443\u0441\u043f\u0435\u0445\u0443. \u0418 \u0447\u0442\u043e \u043d\u0435\u0433\u0440\u0430\u0436\u0434\u0430\u043d\u0435 \u0447\u0430\u0449\u0435 \u043f\u0440\u0438\u0445\u043e\u0434\u044f\u0442 \u043a \u0443\u0441\u043f\u0435\u0445\u0443, \u0447\u0435\u043c \u043f\u0443\u044d\u0440\u0442\u043e\u0440\u0438\u043a\u0430\u043d\u0446\u044b."
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": "\u041a\u0430\u0436\u0435\u0442\u0441\u044f, \u0447\u0442\u043e \u0441\u0440\u0435\u0434\u0438 self-emloyed \u0434\u043e\u043b\u0436\u043d\u043e \u0431\u044b\u0442\u044c \u0431\u043e\u043b\u044c\u0448\u0435 \u043f\u0440\u0435\u0434\u0441\u0442\u0430\u0432\u0438\u0442\u0435\u043b\u0435\u0439 50\u041a+:"
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": "seotr = df[['SEOTR', 'CLASS']]\nseotr.SEOTR = seotr.SEOTR.apply(lambda s: s > 0)\nsums = seotr.groupby('SEOTR').sum()\ncounts = seotr[['SEOTR']].groupby('SEOTR').count().values\ndistr = sums.divide(counts)\ndistr.plot(kind='bar', color=['green', 'red']);",
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAW8AAAEdCAYAAADD1ubeAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\neQAAAABJRU5ErkJggg==\n",
       "text": "<matplotlib.figure.Figure at 0x108d6fc50>"
      }
     ],
     "prompt_number": 404
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": "\u0422\u0430\u043a \u043e\u043d\u043e \u0438 \u0435\u0441\u0442\u044c."
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": "Baseline\n========"
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": "\u041d\u0430 \u0432\u0441\u044f\u043a\u0438\u0439 \u043f\u043e\u0436\u0430\u0440\u043d\u044b\u0439 \u043f\u0435\u0440\u0435\u0437\u0430\u0433\u0440\u0443\u0437\u0438\u043c \u0434\u0430\u0442\u0430\u0444\u0440\u0435\u0439\u043c"
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": "df = pd.read_csv('train.csv')\ndef convert_class(cls):\n    if cls == '50000+':\n        return 1\n    else:\n        return 0",
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 414
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": "df.CLASS = df.CLASS.apply(convert_class)",
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 415
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": "\u0417\u0430\u0434\u0443\u043c\u043c\u0438\u043a\u043e\u0434\u0438\u0440\u0443\u0435\u043c \u0441\u0442\u0440\u043e\u043a\u043e\u0432\u044b\u0435 \u0441\u0442\u043e\u043b\u0431\u0446\u044b..."
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": "def dummify(df, columns):\n    for column in columns:\n        df = pd.concat([df, pd.get_dummies(df[column], prefix=column)], axis=1)\n        df = df.drop(column, axis=1)\n    return df",
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 416
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": "columns = []\ncnt = 0\nfor t in df.dtypes:\n    if (t == 'object'):\n       columns.append(df.columns[cnt])\n    cnt += 1",
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 417
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": "df2 = dummify(df, columns)",
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 418
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": "df.columns\nprint(\"Size of incoming dataframe: {0}\".format(df.shape))\nprint(\"Size of dummified dataframe: {0}\".format(df2.shape))",
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": "Size of incoming dataframe: (10000, 42)\nSize of dummified dataframe: (10000, 373)\n"
      }
     ],
     "prompt_number": 419
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": "\u0426\u0435\u043d\u0442\u0440\u0438\u0440\u0443\u0435\u043c \u0438 \u043d\u043e\u0440\u043c\u0438\u0440\u0443\u0435\u043c..."
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": "from sklearn.preprocessing import scale\nM = scale(df2, axis=1)",
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 420
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": "\u041e\u0446\u0435\u043d\u0438\u043c \u043e\u0448\u0438\u0431\u043a\u0443 \u043e\u0431\u0443\u0447\u0435\u043d\u0438\u044f 1NN \u0441 \u043f\u043e\u043c\u043e\u0449\u044c\u044e \u043a\u0440\u043e\u0441\u0441-\u0432\u0430\u043b\u0438\u0434\u0430\u0446\u0438\u0438"
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": "from sklearn import cross_validation\nfrom sklearn.neighbors import KNeighborsClassifier as knn\nfrom sklearn.metrics import mean_absolute_error as MAE\n\nerrors = []\n#\u0411\u0435\u0440\u0435\u043c \u043d\u0435\u0441\u043a\u043e\u043b\u044c\u043a\u043e \u0440\u0430\u0437\u0431\u0438\u0435\u043d\u0438\u0439 \u0432\u044b\u0431\u043e\u0440\u043a\u0438 \u043d\u0430 \u043e\u0431\u0443\u0447\u0430\u044e\u0449\u0443\u044e \u0438 \u043a\u043e\u043d\u0442\u0440\u043e\u043b\u044c\u043d\u0443\u044e\nfor (learn, sup) in iter(cross_validation.StratifiedShuffleSplit(df2.CLASS, test_size=1/3, n_iter=5)):\n    X_in = M[learn, :]\n    Y_in = df2.CLASS.values[learn].astype(np.float32)\n    X_sup = M[sup, :]\n    Y_sup = df2.CLASS.values[sup].astype(np.float32)\n    #\u043e\u0431\u0443\u0447\u0430\u0435\u043c\u0441\u044f\n    A = knn(n_neighbors=1).fit(X_in, Y_in)\n    #\u043f\u0440\u0435\u0434\u0441\u043a\u0430\u0437\u044b\u0432\u0430\u0435\u043c \u0437\u043d\u0430\u0447\u0435\u043d\u0438\u044f \u043d\u0430 \u043a\u043e\u043d\u0442\u0440\u043e\u043b\u044c\u043d\u043e\u0439 \u0432\u044b\u0431\u043e\u0440\u043a\u0435\n    Y_pred = A.predict_proba(X_sup)\n    #\u0441\u0447\u0438\u0442\u0430\u0435\u043c \u043e\u0448\u0438\u0431\u043a\u0443\n    e = MAE(Y_pred[:, 1], Y_sup)\n    errors.append(e)",
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 422
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": "print(\"\u041e\u0448\u0438\u0431\u043a\u0430 \u043e\u0431\u0443\u0447\u0435\u043d\u0438\u044f: {0}\".format(pd.Series(errors).mean()))",
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": "\u041e\u0448\u0438\u0431\u043a\u0430 \u043e\u0431\u0443\u0447\u0435\u043d\u0438\u044f: 0.21589682063587282\n"
      }
     ],
     "prompt_number": 423
    }
   ],
   "metadata": {}
  }
 ]
}