statistical_analysis.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b style=\"font-size:150%;\"> This script makes a statistical analysis of the hard drive disks data. <br>\n",
    "</b> "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import scipy.stats\n",
    "import matplotlib\n",
    "import matplotlib.pyplot as plt\n",
    "from mpl_toolkits.mplot3d import Axes3D\n",
    "from matplotlib import ticker\n",
    "import seaborn\n",
    "%matplotlib inline\n",
    "np.random.seed(123456)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "## matplotlib settings\n",
    "matplotlib.style.use('ggplot')\n",
    "#plt.rcParams['figure.figsize']=16,10\n",
    "plt.rcParams.update({'font.size': 12, \n",
    "                            'font.family': 'STIXGeneral', \n",
    "                            'mathtext.fontset': 'stix'})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# seaborn settings\n",
    "seaborn.set_palette('deep', desat=.6)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# reading data set\n",
    "data= pd.read_csv('hard_drive_data_2015.csv')\n",
    "data= data.drop('Unnamed: 0', 1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p style= \"font-size: 120%;\">\n",
    "Failed disks:\n",
    "</p>\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "fail= pd.DataFrame(data[data.failure==1]).reset_index()\n",
    "fail= fail.drop('index', 1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p style= \"font-size: 120%;\">\n",
    "Success disks:\n",
    "</p>\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "success= data.groupby(['serial_number']).filter(lambda x: x['failure'].sum()==0) \n",
    "succ= pd.DataFrame(success.groupby(['serial_number']).tail(n=1)).reset_index()\n",
    "succ= succ.drop('index', 1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p style= \"font-size: 120%;\">\n",
    " Plot of the distributions\n",
    "</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/data-science/anaconda3/lib/python3.5/site-packages/statsmodels/nonparametric/kdetools.py:20: VisibleDeprecationWarning: using a non-integer number instead of an integer will result in an error in the future\n",
      "  y = X[:m/2+1] + np.r_[0,X[m/2+1:],0]*1j\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAsQAAALHCAYAAACJwVQtAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nHNr/AzUP5JYxpX3kAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x160364518>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig= plt.figure(figsize= (10,10))\n",
    "fig.subplots_adjust(hspace=0.3)\n",
    "\n",
    "ax= plt.subplot2grid((4,2), (0,0))\n",
    "#ax= fig.add_subplot(321)\n",
    "cax= seaborn.kdeplot(succ.smart_194_raw.dropna() , legend=False,\n",
    "                     color= 'red')\n",
    "ax.plot(succ.smart_194_raw.mean()*np.ones(100), np.linspace(0, 0.09, 100)    , 'r--' )\n",
    "cax.set_xlabel('Temperature (C)', fontsize= 14)\n",
    "for tick in cax.xaxis.get_major_ticks():\n",
    "                tick.label.set_fontsize(14)        \n",
    "#ax= fig.add_subplot(321)\n",
    "cax= seaborn.kdeplot(fail.smart_194_raw.dropna(), legend=False, \n",
    "                     color= 'green')\n",
    "ax.plot(fail.smart_194_raw.mean()*np.ones(100), np.linspace(0, 0.09, 100), 'g--' )\n",
    "cax.set_xlabel('Temperature (C)', fontsize= 14)\n",
    "cax.text(35, 0.08, r'$\\Delta\\mu= 0.24$ C', fontsize= 12)\n",
    "cax.text(35, 0.072\n",
    "         , r'$d= 0.05$', fontsize=12)\n",
    "for tick in cax.xaxis.get_major_ticks():\n",
    "               tick.label.set_fontsize(14) \n",
    "\n",
    "        \n",
    "ax= plt.subplot2grid((4,2), (0,1))\n",
    "#ax= fig.add_subplot(322)\n",
    "cax= seaborn.kdeplot(succ.smart_9_raw/(24.*365), color= 'red', legend=False)\n",
    "ax.plot((succ.smart_9_raw/(24.*365)).mean()*np.ones(100), np.linspace(0, 0.35, 100) , 'r--' )\n",
    "cax.set_xlabel('Running time (Years)', fontsize= 14)\n",
    "cax.set_xlim(0, 20)\n",
    "for tick in cax.xaxis.get_major_ticks():\n",
    "                tick.label.set_fontsize(14)\n",
    "#ax= fig.add_subplot(322)\n",
    "cax= seaborn.kdeplot(fail.smart_9_raw/(24.*365), color= 'green', legend=False)\n",
    "ax.plot((fail.smart_9_raw/(24.*365)).mean()*np.ones(100), np.linspace(0, 0.35, 100), 'g--' )\n",
    "cax.text(10, 0.35, r'$\\Delta\\mu= 0.15$ yr', fontsize=12)\n",
    "cax.text(10, 0.32, r'$d= 0.09$', fontsize=12)\n",
    "cax.set_xlim(0, 20)\n",
    "for tick in cax.xaxis.get_major_ticks():\n",
    "                tick.label.set_fontsize(14) \n",
    "\n",
    "ax= plt.subplot2grid((4,2), (1,0))\n",
    "#ax= fig.add_subplot(323)\n",
    "cax= seaborn.kdeplot(succ.smart_3_raw/1000., color= 'red', legend=False)\n",
    "ax.plot((succ.smart_3_raw/1000.).mean()*np.ones(100), np.linspace(0, 3.0, 100), 'r--' )\n",
    "cax.set_xlabel('Spin-up time (s)', fontsize= 14)\n",
    "for tick in cax.xaxis.get_major_ticks():\n",
    "                tick.label.set_fontsize(14)\n",
    "#ax= fig.add_subplot(323)\n",
    "cax= seaborn.kdeplot(fail.smart_3_raw/1000., color= 'green', legend=False)\n",
    "ax.plot((fail.smart_3_raw/1000.).mean()*np.ones(100), np.linspace(0, 3.0, 100), 'g--' )\n",
    "cax.text(6, 2.5, r'$\\Delta\\mu= 0.27$ s', fontsize=12)\n",
    "cax.text(6, 2.2, r'$d= 0.22$', fontsize=12)\n",
    "for tick in cax.xaxis.get_major_ticks():\n",
    "                tick.label.set_fontsize(14) \n",
    "        \n",
    "ax= plt.subplot2grid((4,2), (1,1))        \n",
    "#ax= fig.add_subplot(324)\n",
    "cax= seaborn.kdeplot(succ.smart_5_normalized.dropna(), legend=False, \n",
    "                     color= 'red')\n",
    "ax.plot(succ.smart_5_normalized.mean()*np.ones(100), np.linspace(0, 0.7, 100) , 'r--' )\n",
    "cax.set_xlabel('Counts bad sectors (Normalized)', fontsize= 14)\n",
    "\n",
    "for tick in cax.xaxis.get_major_ticks():\n",
    "                tick.label.set_fontsize(14)\n",
    "#ax= fig.add_subplot(324)\n",
    "cax= seaborn.kdeplot(fail.smart_5_normalized.dropna(), legend=False,\n",
    "                     color= 'green') \n",
    "cax.text(150, 0.6, r'$\\Delta\\mu= 6.8$ counts', fontsize=12)\n",
    "cax.text(150, 0.52, r'$d= 0.26$', fontsize=12)\n",
    "ax.plot(fail.smart_5_normalized.mean()*np.ones(100), np.linspace(0, 0.7, 100) , 'g--' )\n",
    "for tick in cax.xaxis.get_major_ticks():\n",
    "               tick.label.set_fontsize(14) \n",
    "        \n",
    "ax= plt.subplot2grid((4,2), (2,0))\n",
    "#ax= fig.add_subplot(325)\n",
    "cax= seaborn.kdeplot(succ.smart_1_normalized, legend=False,\n",
    "                     color= 'red')\n",
    "ax.plot((succ.smart_1_normalized).mean()*np.ones(100), np.linspace(0,0.12,100), 'r--' )\n",
    "cax.set_xlabel('Read error rate (Normalized)', fontsize= 14)\n",
    "for tick in cax.xaxis.get_major_ticks():\n",
    "                tick.label.set_fontsize(14)\n",
    "\n",
    "#ax= fig.add_subplot(325)\n",
    "cax= seaborn.kdeplot(fail.smart_1_normalized, legend=False,\n",
    "                     color= 'green')\n",
    "ax.plot((fail.smart_1_normalized).mean()*np.ones(100), np.linspace(0,0.12,100), 'g--' )\n",
    "cax.text(150, 0.12, r'$\\Delta\\mu= 8.6$ counts', fontsize=12)\n",
    "cax.text(150, 0.105, r'$d= 0.41$', fontsize=12)\n",
    "for tick in cax.xaxis.get_major_ticks():\n",
    "               tick.label.set_fontsize(14)   \n",
    "\n",
    "ax= plt.subplot2grid((4,2), (3,0))\n",
    "#ax= fig.add_subplot(326)\n",
    "ax.plot((0, 0.2), (0.5, 0.5) , 'r-')\n",
    "ax.plot((0, 0.2), (0.7, 0.7), 'g-')\n",
    "ax.text(0.3, 0.5, 'Working disks', color= 'red',fontsize= 14)\n",
    "ax.text(0.3, 0.7, 'Failed disks', color= 'green',fontsize= 14)\n",
    "ax.set_xlim(0,1)\n",
    "ax.set_ylim(0,1)\n",
    "ax.patch.set_visible(False)\n",
    "ax.axes.get_xaxis().set_visible(False)\n",
    "ax.axes.get_yaxis().set_visible(False)   \n",
    "\n",
    "ax= plt.subplot2grid((4,2), (2,1))\n",
    "#ax= fig.add_subplot(326)\n",
    "cax= seaborn.boxplot(succ.smart_12_raw, \n",
    "                     color= 'red', saturation=0.4)\n",
    "cax.set_xlabel('Power cycle counts of disks that work', fontsize= 14)\n",
    "cax.set_xlim(0, 200)\n",
    "cax.text(100, -0.2, r'$\\mu= 15.7$ counts', fontsize=12)\n",
    "for tick in cax.xaxis.get_major_ticks():\n",
    "                tick.label.set_fontsize(14)\n",
    "        \n",
    "ax= plt.subplot2grid((4,2), (3,1))        \n",
    "#ax= fig.add_subplot(212)\n",
    "cax= seaborn.boxplot(fail.smart_12_raw, \n",
    "                     color= 'green', saturation=0.4 )\n",
    "cax.set_xlabel('Power cycle counts of disks that fail', fontsize= 14)\n",
    "cax.set_xlim(0, 200)\n",
    "cax.text(100, -0.2, r'$\\mu= 22.3$ counts', fontsize=12)\n",
    "for tick in cax.xaxis.get_major_ticks():\n",
    "               tick.label.set_fontsize(14)      \n",
    "        \n",
    "        \n",
    "plt.tight_layout()\n",
    "plt.savefig('distributions.pdf')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": false
   },
   "source": [
    "<b style=\"font-size:150%;\"> PART I: Quantifying difference between distributions  <br>\n",
    "of failed and success disks\n",
    "</b> \n",
    "\n",
    "<p style=\"font-size:120%;\"> The Cohen's distance is the best measurement to quantify differences between the distributions.\n",
    "</p> \n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "class Cohen_distance(object):\n",
    "    '''\n",
    "    Input: specific S.M.A.R.T attribute of failed and success disks\n",
    "    '''\n",
    "    #def __init__(self, dist1, dist2):\n",
    "    #    self.dist1= dist1\n",
    "    #    self.dist2= dist2\n",
    "\n",
    "    def distance(self, dist1, dist2):\n",
    "        diff= abs(dist1.mean() - dist2.mean() )\n",
    "        n1, n2= len(dist1), len(dist2)\n",
    "        var1, var2= dist1.var(), dist2.var()\n",
    "        denominator= ((n1-1)*var1 + (n2-1)*var2)/ (n1+n2)\n",
    "        d= diff/np.sqrt(denominator)\n",
    "        return d\n",
    "\n",
    "    def overlap_superiority(self, dist1, dist2):\n",
    "        control_sample= dist1\n",
    "        treatment_sample= dist2\n",
    "        thresh = (control_sample.mean() + treatment_sample.mean()) / 2\n",
    "        control_above= sum(control_sample>thresh)/len(control_sample)\n",
    "        treatment_below= sum(treatment_sample <thresh)/ len(treatment_sample)\n",
    "        overlap= control_above+ treatment_below\n",
    "        superiority= np.size([x for x,y in zip(treatment_sample, control_sample) if x>y ])/len(treatment_sample)\n",
    "        return overlap, superiority\n",
    "    \n",
    "    def get_effect_size(self, dist1, dist2):\n",
    "        distance= self.distance(dist1, dist2)\n",
    "        overl_prob, sup_prob= self.overlap_superiority(dist1, dist2)\n",
    "        return distance, overl_prob, sup_prob"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>SMART_attribute</th>\n",
       "      <th>Cohens_distance</th>\n",
       "      <th>overlap_prob</th>\n",
       "      <th>superiority_prob</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>error_rate</td>\n",
       "      <td>0.410918</td>\n",
       "      <td>1.076156</td>\n",
       "      <td>0.013714</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>bad_sectors_count</td>\n",
       "      <td>0.265981</td>\n",
       "      <td>1.105345</td>\n",
       "      <td>0.005580</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>spin-up_time</td>\n",
       "      <td>0.220474</td>\n",
       "      <td>1.038184</td>\n",
       "      <td>0.003270</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>running_time</td>\n",
       "      <td>0.095970</td>\n",
       "      <td>1.052582</td>\n",
       "      <td>0.016853</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>internal_temp</td>\n",
       "      <td>0.052604</td>\n",
       "      <td>0.968214</td>\n",
       "      <td>0.010395</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>power_cycle_count</td>\n",
       "      <td>0.033592</td>\n",
       "      <td>1.159660</td>\n",
       "      <td>0.018529</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     SMART_attribute  Cohens_distance  overlap_prob  superiority_prob\n",
       "0         error_rate         0.410918      1.076156          0.013714\n",
       "2  bad_sectors_count         0.265981      1.105345          0.005580\n",
       "1       spin-up_time         0.220474      1.038184          0.003270\n",
       "3       running_time         0.095970      1.052582          0.016853\n",
       "5      internal_temp         0.052604      0.968214          0.010395\n",
       "4  power_cycle_count         0.033592      1.159660          0.018529"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dist= []\n",
    "stats= Cohen_distance()\n",
    "\n",
    "dist.append( stats.get_effect_size(fail.smart_1_normalized.dropna(), \n",
    "                                        succ.smart_1_normalized.dropna()))\n",
    "dist.append( stats.get_effect_size(fail.smart_3_raw.dropna()/1000., \n",
    "                                        succ.smart_3_raw.dropna()/1000.)) # secs\n",
    "dist.append( stats.get_effect_size(fail.smart_5_normalized.dropna(), \n",
    "                                        succ.smart_5_normalized.dropna()))\n",
    "dist.append( stats.get_effect_size(fail.smart_9_raw.dropna()/(24*365.),\n",
    "                                        succ.smart_9_raw.dropna()/(24*365.))) # hours\n",
    "dist.append( stats.get_effect_size(fail.smart_12_raw.dropna(),\n",
    "                                        succ.smart_12_raw.dropna()))\n",
    "dist.append( stats.get_effect_size(fail.smart_194_raw.dropna(),\n",
    "                                        succ.smart_194_raw.dropna()))\n",
    "\n",
    "dist= pd.DataFrame(dist, columns=['Cohens_distance', 'overlap_prob', 'superiority_prob'])\n",
    "\n",
    "smart_attr= pd.Series(['error_rate', 'spin-up_time', 'bad_sectors_count', 'running_time', 'power_cycle_count', 'internal_temp'])\n",
    "distances= pd.concat([smart_attr, dist], axis=1)\n",
    "distances= distances.rename(columns={0: 'SMART_attribute'} )\n",
    "distances.sort_values(by='Cohens_distance', ascending= False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "<p style=\"font-size:120%;\"> According to these results, the SMART attributes where there is more difference between failed and success disks are: <br> \n",
    "-- the bad sectors count <br> \n",
    "-- error rate <br>\n",
    "-- the spin-up time. \n",
    "</p> \n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b style=\"font-size:150%;\"> PART II: Quantifying the precision of the measurement  <br>\n",
    "of the Cohen's distances.\n",
    "</b> \n",
    "\n",
    "<p style=\"font-size:120%;\"> Here I will compute the 95% confidence intervals for the estimate of the Cohen's distance.\n",
    "</p> "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "class CohenSampler(Cohen_distance):\n",
    "    '''\n",
    "    Makes a sampling distribution of sampling distances.\n",
    "    Uses Boostraap\n",
    "    '''\n",
    "    def resample(self, dist1, dist2):\n",
    "        group1= np.random.choice(dist1, len(dist1), replace=True)\n",
    "        group2= np.random.choice(dist2, len(dist2), replace=True)\n",
    "        return self.distance(group1, group2)\n",
    "         \n",
    "    \n",
    "    def compute_sample_stats(self, dist1, dist2, niters= 1000):\n",
    "        sampling_dist= [self.resample(dist1, dist2) for i in range(niters)]\n",
    "        return np.array(sampling_dist)\n",
    "    \n",
    "    \n",
    "    def get_error_and_CI(self, dist1, dist2):\n",
    "        sampling= self.compute_sample_stats(dist1, dist2)\n",
    "        std_error= sampling.std()\n",
    "        CI= np.percentile(sampling, [2.5, 97.5])\n",
    "        \n",
    "        return np.round(CI, 2), np.round(std_error,2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "stats=[]\n",
    "sampling= CohenSampler()\n",
    "\n",
    "stats.append( sampling.get_error_and_CI(fail.smart_1_normalized.dropna(), \n",
    "                                        succ.smart_1_normalized.dropna()))\n",
    "stats.append( sampling.get_error_and_CI(fail.smart_3_raw.dropna()/1000., \n",
    "                                        succ.smart_3_raw.dropna()/1000.))\n",
    "stats.append( sampling.get_error_and_CI(fail.smart_5_normalized.dropna(), \n",
    "                                        succ.smart_5_normalized.dropna()))\n",
    "stats.append( sampling.get_error_and_CI(fail.smart_9_raw.dropna()/(24*365.), \n",
    "                                        succ.smart_9_raw.dropna()/(24*365.)))\n",
    "stats.append( sampling.get_error_and_CI(fail.smart_12_raw.dropna(), \n",
    "                                        succ.smart_12_raw.dropna()))\n",
    "stats.append( sampling.get_error_and_CI(fail.smart_194_raw.dropna(), \n",
    "                                        succ.smart_194_raw.dropna()) )\n",
    "stats= pd.DataFrame(stats, columns=['95% CI', 'std_error'])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>SMART_attribute</th>\n",
       "      <th>Cohens_distance</th>\n",
       "      <th>95% CI</th>\n",
       "      <th>std_error</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>error_rate</td>\n",
       "      <td>0.410918</td>\n",
       "      <td>[0.33, 0.49]</td>\n",
       "      <td>0.04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>spin-up_time</td>\n",
       "      <td>0.220474</td>\n",
       "      <td>[0.15, 0.29]</td>\n",
       "      <td>0.03</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>bad_sectors_count</td>\n",
       "      <td>0.265981</td>\n",
       "      <td>[0.18, 0.35]</td>\n",
       "      <td>0.05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>running_time</td>\n",
       "      <td>0.095970</td>\n",
       "      <td>[0.04, 0.15]</td>\n",
       "      <td>0.03</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>power_cycle_count</td>\n",
       "      <td>0.033592</td>\n",
       "      <td>[0.02, 0.24]</td>\n",
       "      <td>0.06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>internal_temp</td>\n",
       "      <td>0.052604</td>\n",
       "      <td>[0.01, 0.11]</td>\n",
       "      <td>0.03</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     SMART_attribute  Cohens_distance        95% CI  std_error\n",
       "0         error_rate         0.410918  [0.33, 0.49]       0.04\n",
       "1       spin-up_time         0.220474  [0.15, 0.29]       0.03\n",
       "2  bad_sectors_count         0.265981  [0.18, 0.35]       0.05\n",
       "3       running_time         0.095970  [0.04, 0.15]       0.03\n",
       "4  power_cycle_count         0.033592  [0.02, 0.24]       0.06\n",
       "5      internal_temp         0.052604  [0.01, 0.11]       0.03"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "stats_complete= pd.concat([distances['SMART_attribute'], distances['Cohens_distance'], stats], axis= 1 )\n",
    "stats_complete"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "<p style=\"font-size:120%;\"> As an example, the following is a plot of the sampling distribution of the error rate. The vertical gray line shows the Cohen's distance. The yellow lines, the 95% confidence interval of this measurement. \n",
    "</p> "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def plot_sample_stats(dist1, dist2):\n",
    "    s= CohenSampler()\n",
    "    distr= s.compute_sample_stats(dist1, dist2)\n",
    "    cohens_d= s.distance(dist1, dist2)\n",
    "    ci= np.percentile(distr, [2.5, 97.5])\n",
    "    plt.hist(distr)\n",
    "    plt.plot(cohens_d*np.ones(300), np.arange(300), color='0.6')\n",
    "    plt.plot(ci[0]*np.ones(300), np.arange(300), color='yellow')\n",
    "    plt.plot(ci[1]*np.ones(300), np.arange(300), color='yellow')\n",
    "    plt.xlabel('sampling distribution')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXsAAAEVCAYAAAABwEUhAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nyN9VY9Tg+AAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11e9f0208>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_sample_stats(fail.smart_1_normalized.dropna(),\n",
    "                  succ.smart_1_normalized.dropna())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b style=\"font-size:150%;\"> PART III: testing if the differences observed are due to chance.  \n",
    "</b> \n",
    "\n",
    "<p style=\"font-size:120%;\"> Here I will make some hypothesis testing\n",
    "</p> "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p style=\"font-size:120%;\">\n",
    "My null hypothesis is: <br>\n",
    "$H_0$: The distributions of S.M.A.R.T attributes \n",
    "of failed and working disks are the same. <br>\n",
    "\n",
    "This means that: $\\mu_1 -\\mu_2=0$. <br>\n",
    "We therefore expect that the sampling distribution of mean differences will be centered at cero. \n",
    "</p> "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "class Hypothesis_testing(object):\n",
    "    \n",
    "    def distance(self, dist1, dist2):\n",
    "        diff= abs(dist1.mean() - dist2.mean() )\n",
    "        n1, n2= len(dist1), len(dist2)\n",
    "        var1, var2= dist1.var(), dist2.var()\n",
    "        denominator= ((n1-1)*var1 + (n2-1)*var2)/ (n1+n2)\n",
    "        d= diff/np.sqrt(denominator)\n",
    "        return d\n",
    "    \n",
    "    def resample_with_model(self, n, pool):\n",
    "        '''\n",
    "        uses bootstraap to create samples from Ho\n",
    "        '''\n",
    "        np.random.shuffle(pool)\n",
    "        return self.distance(pool[:n], pool[n:])\n",
    "    \n",
    "    def test_hypothesis(self, dist1, dist2, niters=1000):\n",
    "        n, m = len(dist1), len(dist2)\n",
    "        pool= np.hstack((dist1, dist2))\n",
    "        sampling_dist= np.array([self.resample_with_model(n, pool) for i in range(niters)] )\n",
    "        observed_d= self.distance(dist1, dist2)\n",
    "        p_value= sum(sampling_dist> observed_d)/len(sampling_dist)\n",
    "        #plt.hist(sampling_dist)\n",
    "        #plt.plot(observed_d*np.ones(300), np.arange(300), color='gray')\n",
    "        return p_value\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "p_values=[]\n",
    "test= Hypothesis_testing()\n",
    "\n",
    "p_values.append(test.test_hypothesis(fail.smart_1_normalized.dropna(), \n",
    "                     succ.smart_1_normalized.dropna()) )\n",
    "p_values.append(test.test_hypothesis(fail.smart_3_raw.dropna()/1000.,\n",
    "                                    succ.smart_3_raw.dropna()/1000.))\n",
    "p_values.append(test.test_hypothesis(fail.smart_5_normalized.dropna(), \n",
    "                                    succ.smart_5_normalized.dropna() ))\n",
    "p_values.append(test.test_hypothesis(fail.smart_9_raw.dropna()/(24*365.), \n",
    "                                    succ.smart_9_raw.dropna()/(24*365.)))\n",
    "p_values.append(test.test_hypothesis(fail.smart_12_raw.dropna(), \n",
    "                                    succ.smart_12_raw.dropna()))\n",
    "p_values.append(test.test_hypothesis(fail.smart_194_raw.dropna(), \n",
    "                                    succ.smart_194_raw.dropna()))\n",
    "p_values= pd.Series(p_values)\n",
    "p_values_complete= pd.concat([distances['SMART_attribute'], distances['Cohens_distance'], p_values], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>SMART_attribute</th>\n",
       "      <th>Cohens_distance</th>\n",
       "      <th>p_value</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>error_rate</td>\n",
       "      <td>0.410918</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>spin-up_time</td>\n",
       "      <td>0.220474</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>bad_sectors_count</td>\n",
       "      <td>0.265981</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>running_time</td>\n",
       "      <td>0.095970</td>\n",
       "      <td>0.001</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>power_cycle_count</td>\n",
       "      <td>0.033592</td>\n",
       "      <td>0.075</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>internal_temp</td>\n",
       "      <td>0.052604</td>\n",
       "      <td>0.038</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     SMART_attribute  Cohens_distance  p_value\n",
       "0         error_rate         0.410918    0.000\n",
       "1       spin-up_time         0.220474    0.000\n",
       "2  bad_sectors_count         0.265981    0.000\n",
       "3       running_time         0.095970    0.001\n",
       "4  power_cycle_count         0.033592    0.075\n",
       "5      internal_temp         0.052604    0.038"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "p_values_complete= p_values_complete.rename(columns={0: 'p_value'})\n",
    "p_values_complete"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p style=\"font-size:120%;\">\n",
    "With a significance of 1%, I obtain the following: <br>\n",
    "- I reject $H_0$ in the first three distributions in the table of above. This means that I'm confident that the following distributions ARE DIFFERENT: <br>\n",
    "  1. error rate <br>\n",
    "  2. spin up time <br>\n",
    "  3. bad sectors count <br>\n",
    "  \n",
    "- I can not reject $H_0$ in the last two S.M.A.R.T attributes in the table. This means that the following distributions MIGHT BE THE SAME: <br>\n",
    "  1. power cycle count <br>\n",
    "  2. internal temperature <br>\n",
    "  \n",
    "- The p-value for the running time is equals to the significance level, so I CAN NOT SAY WHETHER THE DISTRIBUTIONS ARE THE SAME OR DIFFERENT.\n",
    "    \n",
    "</p> "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python [Root]",
   "language": "python",
   "name": "Python [Root]"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}