gistfile1.txt

{
 "metadata": {
  "name": "",
  "signature": "sha256:fca7e57a6ba82559b16f1971ff93611fb126632a4782cf40ad0c33ed5eb46262"
 },
 "nbformat": 3,
 "nbformat_minor": 0,
 "worksheets": [
  {
   "cells": [
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "### EyeQuant Hiring Task\n",
      "\n",
      "CSV-File, 5 Columns, 960 Rows\n",
      "\n",
      "+ How are the different columns distributed?\n",
      "+ Are the columns related? Correlated? How much?\n",
      "+ Is there a anything special about one (or several) columns as compared to the others? If so, why?\n",
      "+ Is there a column that can be predicted specially good from the others?\n",
      "+ If you were told to predict a particular column using information from the others"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "import pandas as pd\n",
      "data = pd.read_csv('eq_data.csv')"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 18
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "data.columns"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 19,
       "text": [
        "Index([u'sc0', u'sc1', u'sc2', u'sc3', u'sc4', u'sc5'], dtype=object)"
       ]
      }
     ],
     "prompt_number": 19
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "### Correlations"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "from scipy.stats import linregress"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 21
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "for i, col1 in enumerate(data.columns[:-1]):\n",
      "        for j, col2 in enumerate(data.columns[i+1:]):\n",
      "            slope, intercept, r_value, p_value, std_err = linregress(data[col1], data[col2])\n",
      "            print '{} vs {} r = {}, p = {}'.format(col1, col2, r_value, p_value)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "sc0 vs sc1 r = 0.278302762359, p = 1.55244553324e-18\n",
        "sc0 vs sc2 r = 0.638549596231, p = 4.33284394596e-111\n",
        "sc0 vs sc3 r = 0.785580884627, p = 6.21274063713e-202\n",
        "sc0 vs sc4 r = 0.810795227895, p = 4.697653532e-225\n",
        "sc0 vs sc5 r = 0.859687106329, p = 9.95606876821e-282\n",
        "sc1 vs sc2 r = 0.425000586559, p = 2.18181008594e-43\n",
        "sc1 vs sc3 r = 0.187118970676, p = 5.16435701926e-09\n",
        "sc1 vs sc4 r = 0.3186917072, p = 4.20831833518e-24\n",
        "sc1 vs sc5 r = 0.254797880313, p = 1.08338746549e-15\n",
        "sc2 vs sc3 r = 0.58144635172, p = 6.48086955899e-88\n",
        "sc2 vs sc4 r = 0.666065567099, p = 4.08590370726e-124\n",
        "sc2 vs sc5 r = 0.662588253529, p = 2.15741594221e-122\n",
        "sc3 vs sc4 r = 0.7728576345, p = 2.01525955777e-191\n",
        "sc3 vs sc5 r = 0.925258582497, p = 0.0\n",
        "sc4 vs sc5 r = 0.922128225264, p = 0.0\n"
       ]
      }
     ],
     "prompt_number": 27
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "### Histograms "
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "%matplotlib inline\n",
      "import pylab as pl\n",
      "for i, col in enumerate(data.columns):\n",
      "    pl.figure()\n",
      "    pl.hist(data[col], 50, color='k', alpha=0.8)\n",
      "    pl.grid('on', 'both')"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAXQAAAEACAYAAACj0I2EAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
       "text": [
        "<matplotlib.figure.Figure at 0x7f2487644290>"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAXQAAAEACAYAAACj0I2EAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAElFTkSuQmCC\n",
       "text": [
        "<matplotlib.figure.Figure at 0x7f2487a0a5d0>"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAXQAAAEACAYAAACj0I2EAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nIiIiIiIiIiIiIiIiIiIiIiIiIjJs/w8fb744QjggLgAAAABJRU5ErkJggg==\n",
       "text": [
        "<matplotlib.figure.Figure at 0x7f2487a1c590>"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAXQAAAEACAYAAACj0I2EAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nERERERERERERERERERERKdr/A4l8GsOKEDuVAAAAAElFTkSuQmCC\n",
       "text": [
        "<matplotlib.figure.Figure at 0x7f24873dac50>"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAXQAAAEACAYAAACj0I2EAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\niIiIiIiIiIiM2v8DGyyMl6YXtbsAAAAASUVORK5CYII=\n",
       "text": [
        "<matplotlib.figure.Figure at 0x7f2487250150>"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAXQAAAEACAYAAACj0I2EAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nIiIiIiIiItJv/w+nfeCDPRE5XgAAAABJRU5ErkJggg==\n",
       "text": [
        "<matplotlib.figure.Figure at 0x7f24870cde10>"
       ]
      }
     ],
     "prompt_number": 32
    }
   ],
   "metadata": {}
  }
 ]
}