sc268/classification.ipynb

## classification.ipynb
{
 "metadata": {
  "name": "classfication.ipynb"
 },
 "nbformat": 3,
 "nbformat_minor": 0,
 "worksheets": [
  {
   "cells": [
    {
     "cell_type": "heading",
     "level": 1,
     "metadata": {},
     "source": [
      "Providers specialities classification based on peer group comparision"
     ]
    },
    {
     "cell_type": "raw",
     "metadata": {},
     "source": [
      "# ------------------------------------------------------------------------\n",
      "# Author     : Sean Chang\n",
      "# Email      : sean.chang@duke.edu\n",
      "# Filename   : classfication.ipynb\n",
      "# Description: Reclassification based on peer group comparison \n",
      "# ------------------------------------------------------------------------"
     ]
    },
    {
     "cell_type": "raw",
     "metadata": {},
     "source": [
      "ideas: Neural network/ SVM / Bayesian classfication/ ensemble learning\n",
      "\n",
      "future work based on this project: \n",
      "\n",
      "logestic regression + speciality \n",
      "if speciality is different, compare billing amounts\n",
      "Naive Bayes + VAR\n",
      "meta algorithms"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "# set working directory, load packages\n",
      "#% cd \"S:\\IIA\\Projects\\HCI\\Fraud Conditions\\specialty_practice\\SeanChang\"\n",
      "% cd \"\\\\iiapcfilep1\\S_Share\\IIA\\Projects\\HCI\\Fraud Conditions\\specialty_practice\\SeanChang\"\n",
      "\n",
      "import pandas as pd\n",
      "import numpy as np\n",
      "import scipy as sp"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "\\\\iiapcfilep1\\S_Share\\IIA\\Projects\\HCI\\Fraud Conditions\\specialty_practice\\SeanChang\n"
       ]
      }
     ],
     "prompt_number": 1
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "def random_sample(input, prop):\n",
      "    import random\n",
      "    index = random.sample(input.index, int(prop* input.shape[0]))\n",
      "    input1, input2 = data_management(input.ix[index], input.drop(index))\n",
      "    return input1, input2\n",
      "    \n",
      "# data preprocessing\n",
      "def data_management_single(dat):\n",
      "    \"\"\" \n",
      "    input: a pandas data frame\n",
      "    replace nan by zero\n",
      "    add a description column\n",
      "    remove all zero columns (i.e. no information)\n",
      "    \"\"\"\n",
      "    # create description if it's not there\n",
      "    if 'descr' not in dat.columns:\n",
      "        dat = add_descr(dat)\n",
      "    \n",
      "    # remove PRVSEQ (physician's ID)\n",
      "    #if 'PRVSEQ' in input.columns:\n",
      "    #    del input['PRVSEQ']\n",
      "        \n",
      "    # replace nan by zero\n",
      "    dat = dat.fillna(0)\n",
      "    \n",
      "    # remove \n",
      "    newcols = dat.columns[dat.std() > 1e-4]\n",
      "    dat = dat[newcols]\n",
      "    \n",
      "    # remove all zero column\n",
      "    if any(dat.sum(0)!=0):\n",
      "        dat = dat.loc[dat.index, dat.sum(0)!=0 ] \n",
      "        \n",
      "    # add description \n",
      "    spec_table = pd.read_csv('spec_description_table.csv')\n",
      "    spec = spec_table.SPECIALTY\n",
      "    descr = spec_table.DESCRIPTION\n",
      "    \n",
      "    #    must transfer spec1 to string for comparision\n",
      "    dat['spec1'] = map(lambda cat: str(cat).rjust(2, '0'), dat.spec1)   # note input =1, output = string '01'\n",
      "    dat['descr'] = \"\"\n",
      "\n",
      "    for i in xrange(0, len(spec)):\n",
      "        dat.descr[ dat.spec1 == spec[i]] = descr[i]\n",
      "    return dat\n",
      " \n",
      "def data_management(input1, input2):\n",
      "    \"\"\" input1,input2: two dataframes \n",
      "    two dataframes with common column names (covariates)\n",
      "    \"\"\"\n",
      "    input1 = data_management_single(input1)\n",
      "    input2 = data_management_single(input2)\n",
      "        \n",
      "    # remove those covariates that do not appear in both datasets\n",
      "    #input1_covariates = input1.columns\n",
      "    #input2_covariates = input2.columns\n",
      "    #common_covariates = pd.Series((input1.columns).intersection((input2.columns)))\n",
      "    common_covariates = list(set(input1.columns) & set(input2.columns))\n",
      "    input1 = input1[common_covariates]\n",
      "    input2 = input2[common_covariates]\n",
      "        \n",
      "    # keep common specialities (we are not able to predict unknown categories)\n",
      "    common_specialities = list(set(input1.spec1.unique()) & set(input2.spec1.unique()))\n",
      "    input1 = input1[map(lambda x: x in common_specialities, input1.spec1)]\n",
      "    input2 = input2[map(lambda x: x in common_specialities, input2.spec1)]\n",
      "    \n",
      "    output1 = input1\n",
      "    output2 = input2 \n",
      "    return output1, output2\n",
      "\n",
      "def add_descr(dat):    \n",
      "    # create description \n",
      "    spec_table = pd.read_csv('spec_description_table.csv')\n",
      "    spec = spec_table.SPECIALTY         #string                     \n",
      "    description = spec_table.DESCRIPTION\n",
      "    dat['descr'] = \"\"\n",
      "    \n",
      "    #change spec1 to two digits string\n",
      "    dat.spec1 = map(lambda cat: str(cat).rjust(2,'0'), dat.spec1)\n",
      "    \n",
      "    for i in xrange(0, len(spec)):\n",
      "        dat.descr[ dat.spec1 == spec[i]] = description [i]\n",
      "        \n",
      "    # delete unnecessary columns\n",
      "        delcols = ['estab_ind,individual','num_days_in_services','num_patients','per_0_11','per_18_44','estab_ind','individual','num_days_in_services'\n",
      "    ,'num_patients','SPECIALTY','cnt_0_11','cnt_0_4','cnt_12_17','cnt_18_44','cnt_45_54','cnt_55_64','cnt_5_11','cnt_65plus','cnt_bothgender'\n",
      "    ,'cnt_female','cnt_unique_d2','cnt_unique_proc','range_in_service','spec2','spec3','spec4','tot_claims','tot_lines','tot_unique_lines']\n",
      "    \n",
      "    for col in list(set(delcols) & set(dat.columns)):\n",
      "        del dat[col]\n",
      "    \n",
      "    return dat\n",
      "\n",
      "# empirical cdf\n",
      "class ECDF:\n",
      "    def __init__(self, observations):\n",
      "        self.observations = observations\n",
      "\n",
      "    def __call__(self, x):\n",
      "        counter = 0.0\n",
      "        for obs in self.observations:\n",
      "            if obs <= x:\n",
      "                counter += 1\n",
      "        return counter / len(self.observations)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 2
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "def cross_validation(true_category, pred_category, z):\n",
      "    \"\"\" input: pred categories and true categories\n",
      "        output: (a) cross table\n",
      "                (b) misclassification rate\n",
      "                (c) average of misclassification rate \n",
      "                (d) heat map \"\"\"\n",
      "    # cross table \n",
      "    output1 = pd.crosstab(true_category, pred_category, rownames = ['actual'], colnames =['preds'])\n",
      "    #print output1\n",
      "    \n",
      "    # misclassification rates\n",
      "    output = np.round(output1.apply(lambda r: r/r.sum(),axis = 1),2)\n",
      "    \n",
      "    #print output\n",
      "    misrate =[]\n",
      "    #for i in xrange(0,len(z)):\n",
      "    #    misrate.append(1.0 - output[z[i]][z[i]])\n",
      "    #print 'misclasssfication rates are' + str(np.round(misrate,2))\n",
      "    #print mean(misrate)\n",
      "    heatmap(output1)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 3
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "# writing everything in OOP \n",
      "# Random Forest class\n",
      "\n",
      "class Classifier:\n",
      "\n",
      "    def __init__(self, dat_train, dat_test, num_features):\n",
      "        #self.train, self.test = data_management(dat_train, dat_test)\n",
      "        self.num_features = num_features\n",
      "        self.train = dat_train\n",
      "        self.test = dat_test\n",
      "        #self.clf = RandomForestClassifier(n_jobs = 5)\n",
      "        self.z, self.output1, self.output2 = 0, [], []\n",
      "        \n",
      "    def __str__(self):\n",
      "        print self.z\n",
      "        \n",
      "    def fit(self, fit_type):\n",
      "        if fit_type not in ['spec1', 'descr']:\n",
      "            raise ValueError(\"fit type must be either spec1 or descr\")\n",
      "    \n",
      "        features = [col for col in self.train.columns if col not in ['spec1', 'descr', 'PRVSEQ','tot_adj_paid']]\n",
      "        import random\n",
      "        features_rnd = random.sample(features, self.num_features)\n",
      "        #clf = 0\n",
      "        #clf = RandomForestClassifier(n_jobs = 5)\n",
      "        #y, self.z = pd.factorize(self.train['spec1'])\n",
      "        y, self.z = pd.factorize(self.train[fit_type])\n",
      "        self.clf.fit(self.train[features_rnd], y)\n",
      "        preds = self.clf.predict(self.test[features_rnd])\n",
      "        preds = np.array(map(lambda cat: self.z[cat], preds))\n",
      "        \n",
      "        self.output1 = pd.crosstab(self.test[fit_type], preds, rownames=['actual'], colnames=['preds'])\n",
      "        #print self.output1\n",
      "        #self.output1 = pd.crosstab(self.test['descr'], preds, rownames=['actual'], colnames=['preds'])\n",
      "        # normalize\n",
      "        self.output2 = np.round(self.output1.apply(lambda r: r/r.sum(),axis = 1),2)\n",
      "        \n",
      "        #cross_validation(self.test['spec1'], preds,self.z)\n",
      "        #print self.output2, self.z\n",
      "        #return self.output1, z\n",
      "        self.test['preds'] = \"\"\n",
      "        self.test.preds = preds\n",
      "        return preds\n",
      "    \n",
      "    def save_preds(self, filename):\n",
      "        ''' add prediction data to the original test dataset '''\n",
      "        self.test.to_csv(str(filename))\n",
      "        \n",
      "\n",
      "    def misrate(self):\n",
      "        rates = []\n",
      "        #for i in xrange(0, self.output1.shape[0]):\n",
      "        #    rates.append(np.round(1-self.output2[self.z[i]][self.z[i]],2))            \n",
      "        for names in self.z:\n",
      "            rates.append(1-self.output2[names][names])\n",
      "        rates = map(lambda num: np.round(num,2), rates)\n",
      "        print rates, mean(rates)\n",
      "    \n",
      "    def misrate_overall(self):\n",
      "        count_mis = 0\n",
      "        for i in xrange(0, self.test.shape[0]):\n",
      "            print self.test.preds[i], self.test.spec1[i]\n",
      "            #if self.test.preds[i] != self.test.spec1[i]:\n",
      "            #    count_mis += 1\n",
      "        print  count_mis, self.test.shape[0], count_mis / self.test.shape[0]\n",
      "        return count_mis / self.test.shape[0]\n",
      "\n",
      "    def threshold(self, alpha):\n",
      "        ''' input: data, alpha (in [0,1]), z = all specialties\n",
      "            output: 100* alpha percentile of billing amount in each speciality.'''   \n",
      "        if 'tot_adj_paid' not in self.train.columns:\n",
      "            raise error(\"input dataframe does not have attribute 'tot_adj_paid'\")\n",
      "            \n",
      "        if (alpha > 1 or alpha<0):\n",
      "            raise error(\"percentile must in [0,1]\")\n",
      "        \n",
      "        threshold = []\n",
      "        for idx in xrange(0, len(self.z)):\n",
      "            temp = double(self.train[self.train.descr == self.z[idx]].tot_adj_paid)\n",
      "            threshold.append(np.percentile(temp, 100*alpha))\n",
      "        threshold = pd.Series(thresholds, index = self.z)\n",
      "        print threshold\n",
      "        return threshold\n",
      "        \n",
      "\n",
      "    def threshold1(self, alpha):\n",
      "        '''create speciality- alpha% quantile series (= dict)'''\n",
      "        spec_threshold = {}\n",
      "        for spec in self.train.descr.unique():\n",
      "            temp = self.train[self.train.descr == spec].tot_adj_paid\n",
      "            spec_threshold[spec] = round(np.percentile(temp, alpha),2)    \n",
      "        return spec_threshold\n",
      "    \n",
      "    def suspicious(self, alpha):\n",
      "        import csv\n",
      "        ''' suspicous if the provider is on top alpha% of its new speciality'''\n",
      "        suspicious_list = []\n",
      "        spec_threshold = self.threshold1(alpha)\n",
      "        #print spec_threshold\n",
      "        \n",
      "        for descr in self.test.descr.unique():\n",
      "            idx = (self.test.tot_adj_paid[self.test.preds == descr] > spec_threshold[descr])\n",
      "            suspicious_list.extend(self.test.PRVSEQ[self.test.preds == descr][idx].values)\n",
      "                \n",
      "        #print suspicious_list\n",
      "        return suspicious_list\n",
      "  \n",
      "  #  def suspicious1(self, alpha): \n",
      "  #      ''' print out those physicians whose billing is among the top 100*(1-alpha)% in new (predicted) groups'''\n",
      "  #      if 'preds' not in self.test.columns:\n",
      "  #          raise error[\"Please run 'fit()' module first to get predictions\"]\n",
      "        \n",
      "  #      threshold = self.threshold(alpha)\n",
      "  #      suspicious_list = []\n",
      "  #      for idx in self.test.index:\n",
      "  #          #if(self.test.tot_adj_paid[idx] > threshold[self.test.descr[idx]]):\n",
      "  #          if(self.test.tot_adj_paid[idx] > threshold[self.test.preds[idx]]):\n",
      "  #              suspicious_list.append(self.test.PRVSEQ[idx])\n",
      "  #      print suspicious_list, len(suspicious_list)\n",
      "        \n",
      "  #  def suspicious2(self, alpha):\n",
      "  #      ''' print out those physicians whose percentile change dramatically in new group'''\n",
      "  #      if 'preds' not in self.test.columns:\n",
      "  #          raise error[\"Please run 'fit()' module first to get predictions\"]\n",
      "  #      \n",
      "  #      threshold = self.threshold(alpha)\n",
      "  #      suspicious_list = []\n",
      "  #      for idx in self.test.index:\n",
      "  #          #if(self.test.tot_adj_paid[idx] > threshold[self.test.descr[idx]]):\n",
      "  #          if(self.test.tot_adj_paid[idx] > threshold[self.test.preds[idx]]):\n",
      "  #              suspicious_list.append(self.test.PRVSEQ[idx])\n",
      "  #      print suspicious_list, len(suspicious_list)\n",
      "\n",
      "    def heatmap(self):\n",
      "        # make a heatmap \n",
      "        import matplotlib.pyplot as plt\n",
      "        \n",
      "        # Plot it out\n",
      "        fig, ax = plt.subplots()\n",
      "        heatmap = ax.pcolor(self.output2, cmap=plt.cm.Blues, alpha=0.8)\n",
      "        \n",
      "        # Format\n",
      "        fig = plt.gcf()\n",
      "        fig.set_size_inches(5, 5)\n",
      "        \n",
      "        # turn off the frame\n",
      "        ax.set_frame_on(False)\n",
      "        \n",
      "        # put the major ticks at the middle of each cell\n",
      "        ax.set_yticks(np.arange(self.output2.shape[0]) + 0.5, minor=False)\n",
      "        ax.set_xticks(np.arange(self.output2.shape[1]) + 0.5, minor=False)\n",
      "        \n",
      "        # want a more natural, table-like display\n",
      "        #ax.invert_yaxis()\n",
      "        #ax.xaxis.tick_top()\n",
      "        \n",
      "        # Set the labels\n",
      "        \n",
      "        # label source:https://en.wikipedia.org/wiki/Basketball_statistics\n",
      "        #labels =\n",
      "        \n",
      "        # note I could have used nba_sort.columns but made \"labels\" instead\n",
      "        ax.set_xticklabels(self.output2.index, minor=False)\n",
      "        ax.set_yticklabels(self.output2.index, minor=False)\n",
      "        \n",
      "        # rotate the\n",
      "        plt.xticks(rotation=90)\n",
      "        \n",
      "        ax.grid(False)\n",
      "        \n",
      "        # Turn off all the ticks\n",
      "        ax = plt.gca()\n",
      "        \n",
      "        # add labels\n",
      "        plt.xlabel('predict category')\n",
      "        plt.ylabel('true category')\n",
      "        plt.colorbar(heatmap)\n",
      "        plt.show()\n",
      "        \n",
      "        for t in ax.xaxis.get_major_ticks():\n",
      "            t.tick1On = False\n",
      "            t.tick2On = False\n",
      "        for t in ax.yaxis.get_major_ticks():\n",
      "            t.tick1On = False\n",
      "            t.tick2On = False\n",
      "            \n",
      "        \n",
      "    def PCA_module(self):\n",
      "        ''' input: dat_train, dat_test\n",
      "            output: a merged dataframe with a new design matrix w.r.t eigenvectors'''\n",
      "        \n",
      "        # merge dat_train & dat_test to do PCA\n",
      "        dat = (self.train).append(self.test)\n",
      "        \n",
      "        # create new design matrix\n",
      "        cols = [col for col in dat.columns if col not in 'descr' ]\n",
      "        from sklearn.decomposition import PCA\n",
      "        newX = PCA(n_components = self.num_features).fit_transform(dat[cols])\n",
      "        \n",
      "        # create dataframe\n",
      "        newColnames = []\n",
      "        for i in xrange(0, num_features):\n",
      "            newColnames.append('Component' + str(i+1))\n",
      "        pcadat = pd.DataFrame(newX, columns =newColnames)\n",
      "        pcadat['descr'] = dat['descr']\n",
      "        \n",
      "        # move spec1 to 1st column\n",
      "        cols = pcadat.columns.tolist()\n",
      "        cols = cols[-1:] + cols[:-1]\n",
      "        pcadat = pcadat[cols]\n",
      "        return pcadat\n",
      "          \n",
      "\n",
      "\n",
      "class RandomForest(Classifier):\n",
      "    def __init__(self, dat_train, dat_test, num_features):\n",
      "        Classifier.__init__(self, dat_train, dat_test, num_features)\n",
      "        from sklearn.ensemble import RandomForestClassifier\n",
      "        self.clf = RandomForestClassifier()\n",
      "        \n",
      "class SVM(Classifier):\n",
      "    def __init__(self, dat_train, dat_test, num_features):\n",
      "        Classifier.__init__(self, dat_train, dat_test, num_features)\n",
      "        from sklearn.svm import SVC\n",
      "        SVC(C=1.0, cache_size=200, class_weight=None, coef0=0.0, degree=3,gamma=0.0, kernel='rbf', max_iter=-1, probability=False,\n",
      "        random_state=None, shrinking=True, tol=0.001, verbose=False)\n",
      "        self.clf = SVC()\n",
      "        \n",
      "class NaiveBayes(Classifier):\n",
      "    def __init__(self, dat_train, dat_test, num_features):\n",
      "        Classifier.__init__(self, dat_train, dat_test, num_features)\n",
      "        from sklearn.naive_bayes import GaussianNB\n",
      "        self.clf = GaussianNB()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 68
    },
    {
     "cell_type": "heading",
     "level": 3,
     "metadata": {},
     "source": [
      "Load datasets "
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "% cd \"\\\\iiapcfilep1\\S_Share\\IIA\\Projects\\HCI\\Fraud Conditions\\specialty_practice\\SeanChang\"\n",
      "dat = pd.read_csv('ctnfl_final_with_descr.csv')\n",
      "dat_train, dat_test = random_sample(dat, 0.8)\n",
      "dat_train, dat_test = data_management(dat_train, dat_test)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "\\\\iiapcfilep1\\S_Share\\IIA\\Projects\\HCI\\Fraud Conditions\\specialty_practice\\SeanChang\n"
       ]
      }
     ],
     "prompt_number": 18
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "dat.shape"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 14,
       "text": [
        "(21771, 1284)"
       ]
      }
     ],
     "prompt_number": 14
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "# generate datasets with predctions (_preds)\n",
      "\n",
      "# dataset2\n",
      "dat_train = pd.read_csv('ctnfl_final.csv')\n",
      "dat_test = pd.read_csv('ctnoh_final.csv')\n",
      "dat_train, dat_test = data_management(dat_train, dat_test)\n",
      "\n",
      "test = RandomForest(dat_train, dat_test, 1000)\n",
      "test.fit('spec1')\n",
      "#test.misrate_overall()\n",
      "figure(1)\n",
      "test.heatmap()\n",
      "#test.save_preds('ctnoh_final_preds.csv')\n",
      "\n",
      "# dataset2\n",
      "dat_test = pd.read_csv('ctnfl_final.csv')\n",
      "dat_train = pd.read_csv('ctnoh_final.csv')\n",
      "dat_train, dat_test = data_management(dat_train, dat_test)\n",
      "\n",
      "test = RandomForest(dat_train, dat_test, 1000)\n",
      "test.fit('spec1')\n",
      "figure(2)\n",
      "test.heatmap()\n",
      "#test.save_preds('ctnfl_final_preds.csv')"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "display_data",
       "text": [
        "<matplotlib.figure.Figure at 0x15ece940>"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAT0AAAFJCAYAAADpKPpKAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
       "text": [
        "<matplotlib.figure.Figure at 0x15eba4e0>"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "text": [
        "<matplotlib.figure.Figure at 0x12782940>"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAT0AAAFJCAYAAADpKPpKAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
       "text": [
        "<matplotlib.figure.Figure at 0x15ebd390>"
       ]
      },
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 32,
       "text": [
        "<matplotlib.text.Text at 0x15ebc668>"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAXcAAAEKCAYAAADpfBXhAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nABhE3AHAIOIOAAYRdwAw6P8AceSMxKCEBqwAAAAASUVORK5CYII=\n",
       "text": [
        "<matplotlib.figure.Figure at 0x80ac1978>"
       ]
      }
     ],
     "prompt_number": 32
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "# compute misclassification rates\n",
      "dat1 = pd.read_csv('ctnfl_final_preds.csv')\n",
      "dat2 = pd.read_csv('ctnoh_final_preds.csv')count = 0.\n",
      "dat = dat2\n",
      "for i in xrange(0, dat.shape[0]):\n",
      "    if dat.spec1[i] != dat.preds[i]:\n",
      "        count += 1\n",
      "print count, dat.shape[0], count/dat.shape[0]"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "8086.0 19521 0.414220582962\n"
       ]
      }
     ],
     "prompt_number": 24
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "# dataset3\n",
      "dat_train = pd.read_csv('ctnoh_final_with_descr.csv')\n",
      "dat_test = pd.read_csv('hmnm_6states_final.csv')\n",
      "dat_train, dat_test = data_management(dat_train, dat_test)\n",
      "\n",
      "test = RandomForest(dat_train, dat_test, 565)\n",
      "test.fit('spec1')\n",
      "test.save_preds('hmnm_6states_final_preds.csv')\n"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stderr",
       "text": [
        "C:\\Users\\i55319\\AppData\\Local\\Enthought\\Canopy\\User\\lib\\site-packages\\pandas\\io\\parsers.py:1070: DtypeWarning: Columns (23,24) have mixed types. Specify dtype option on import or set low_memory=False.\n",
        "  data = self._reader.read(nrows)\n"
       ]
      }
     ],
     "prompt_number": 91
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "dat = pd.read_csv('ctnfl_final.csv')\n",
      "dat_test, dat_test = random_sample(dat, 0.8)\n",
      "dat_train, dat_test = data_management(dat_train, dat_test)\n",
      "\n",
      "test = RandomForest(dat_train, dat_test, 1000)\n",
      "test.fit('spec1')\n",
      "#test.misrate_overall()\n",
      "figure(1)\n",
      "test.heatmap()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "display_data",
       "text": [
        "<matplotlib.figure.Figure at 0x1279a9e8>"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAATwAAAFJCAYAAAAG6pF0AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAAAElFTkSuQmCC\n",
       "text": [
        "<matplotlib.figure.Figure at 0x1279ac18>"
       ]
      }
     ],
     "prompt_number": 34
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "test = RandomForest(dat_train, dat_test, 1000)\n",
      "#test = SVM(dat_train, dat_test, 1000)\n",
      "test.fit('spec1')\n",
      "test.save_preds('ctnfl_final_with_descr_and_preds.csv')\n",
      "test.heatmap()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": "*"
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "print dat.shape, sum(dat.spec1=='35')"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "(21771, 1284) 48\n"
       ]
      }
     ],
     "prompt_number": 142
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "% cd \"\\\\iiapcfilep1\\S_Share\\IIA\\Projects\\HCI\\Fraud Conditions\\specialty_practice\\SeanChang\"\n",
      "\n",
      "temp = dat.groupby('descr').size().order(ascending = False)\n",
      "temp.to_csv('ctnfl_final_spec_size.csv')"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "\\\\iiapcfilep1\\S_Share\\IIA\\Projects\\HCI\\Fraud Conditions\\specialty_practice\\SeanChang\n"
       ]
      }
     ],
     "prompt_number": 151
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "dat_train = pd.read_csv('ctnfl_final.csv')\n",
      "dat_test = pd.read_csv('ctnoh_final.csv')\n",
      "dat_train, dat_test = data_management(dat_train, dat_test)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 52
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "dat_test.descr"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 55,
       "text": [
        "0     Ambulance, Medical Transportation\n",
        "1                          Pediatrician\n",
        "2                              Hospital\n",
        "3                 Obstetrics/Gynecology\n",
        "4                          Pediatrician\n",
        "5                     Internal Medicine\n",
        "6                             Pathology\n",
        "7                              Podiatry\n",
        "8                    Physical Therapist\n",
        "9                          Opthalmology\n",
        "10                     Vascular Surgery\n",
        "11                Obstetrics/Gynecology\n",
        "12                     Vascular Surgery\n",
        "13                           Cardiology\n",
        "14                 Diagnostic Radiology\n",
        "...\n",
        "19663             Neurology\n",
        "19664       General Surgery\n",
        "19665          Chiropractic\n",
        "19666    Nurse Practitioner\n",
        "19667     Internal Medicine\n",
        "19668          Neurosurgery\n",
        "19669     Internal Medicine\n",
        "19670    Emergency Medicine\n",
        "19671              Hospital\n",
        "19672              Hospital\n",
        "19673      Gastroenterology\n",
        "19674              Hospital\n",
        "19675    Nurse Practitioner\n",
        "19676          Pediatrician\n",
        "19677              Hospital\n",
        "Name: descr, Length: 19521, dtype: object"
       ]
      }
     ],
     "prompt_number": 55
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "## print out suspicious PRVSEQS\n",
      "\n",
      "percentile = 95\n",
      "\n",
      "dat_train = pd.read_csv('ctnfl_final.csv')\n",
      "dat_test = pd.read_csv('ctnoh_final.csv')\n",
      "dat_train, dat_test = data_management(dat_train, dat_test)\n",
      "\n",
      "test = RandomForest(dat_train, dat_test, 1000)\n",
      "test.fit('descr')\n",
      "suspicious_ID = test.suspicious(percentile)\n",
      "\n",
      "# save \n",
      "temp = pd.Series(suspicious_ID)\n",
      "temp.to_csv('suspicious_ctnoh_'+str(percentile)+'.csv')\n",
      "\n",
      "### dataset2\n",
      "\n",
      "dat_train = pd.read_csv('ctnoh_final_with_descr.csv')\n",
      "dat_test = pd.read_csv('ctnfl_final_with_descr.csv')\n",
      "dat_train, dat_test = data_management(dat_train, dat_test)\n",
      "\n",
      "test = RandomForest(dat_train, dat_test, 1000)\n",
      "test.fit('descr')\n",
      "suspicious_ID = test.suspicious(percentile)\n",
      "print 'ctnfl'\n",
      "\n",
      "#save\n",
      "temp = pd.Series(suspicious_ID)\n",
      "temp.to_csv('suspicious_ctnfl_'+str(percentile)+ '.csv')"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "ctnfl\n"
       ]
      }
     ],
     "prompt_number": 70
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "dat_test.shape[0]*0.02"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 29,
       "text": [
        "432.46000000000004"
       ]
      }
     ],
     "prompt_number": 29
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "len(suspicious_ID)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 217,
       "text": [
        "96"
       ]
      }
     ],
     "prompt_number": 217
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "dat_test.shape"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 218,
       "text": [
        "(4355, 1246)"
       ]
      }
     ],
     "prompt_number": 218
    },
    {
     "cell_type": "heading",
     "level": 2,
     "metadata": {},
     "source": [
      "Clustering"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "dat_train.descr.unique()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 160,
       "text": [
        "array(['CRNA (Nurse Ansesthetist)', 'Optometrist', 'Emergency Medicine',\n",
        "       'Family Practice', 'DME- Prosthetist/Orthotist',\n",
        "       'Diagnostic Radiology', 'Internal Medicine', 'General Surgery',\n",
        "       'Home Health Agency', 'Audiologist', 'Nurse Practitioner',\n",
        "       'Cardiology', 'Pulmonology', 'Hematology/Oncology', 'Nephrology',\n",
        "       'Obstetrics/Gynecology', 'Ambulatory Surgical Center',\n",
        "       'Ambulance, Medical Transportation', 'Anesthesiology', 'Pathology',\n",
        "       'Dialysis', 'Proctology, Colon & Rectal Surgery',\n",
        "       'Physician Assistant', 'Otolaryngology', 'Pediatrician',\n",
        "       'Physician Specialty Unknown', 'Oncology- Medical', 'Hospital',\n",
        "       'Podiatry', 'Gastroenterology', 'Oncology- Radiation',\n",
        "       'General Practice', 'Orthopaedic Surgery', 'Urology',\n",
        "       'Endocrinology, Diabetes, Metabolism',\n",
        "       'Pediatric Neonatal-Perinatal Medicine', 'Allergy/Immunology',\n",
        "       'Clinical Laboratory', 'Vascular Surgery', 'Urgent Care Center',\n",
        "       'Neurosurgery', 'Infectious Disease', 'Opthalmology',\n",
        "       'Anesthesiology Assistant', 'Neurology', 'Thoracic Surgery',\n",
        "       'Interventional Radiology', 'Certified Nurse Midwife',\n",
        "       'MRI, Independent DX Testing Center', 'Physical Therapist',\n",
        "       'Dermatology', 'Multispecialty Clinic or Group',\n",
        "       'Geriatric Medicine', 'Hematology', 'Critical Care (Intensivist)',\n",
        "       'Plastic and Reconstructive Surgery', 'Chiropractic',\n",
        "       'Interventional Pain Management', 'Sleep Disorders, Sleep Lab',\n",
        "       'Medical Genetics', 'Rheumatology', 'Nuclear Medicine',\n",
        "       'Preventive Medicine', 'Oncology- Gynecological',\n",
        "       'Maxillofacial Surgery', 'Rehabilitation & Physical Medicine',\n",
        "       'Occupational Therapist', 'Psychiatry', 'Orthopedic- Hand Surgery',\n",
        "       'Gynecology', 'Sports Medicine', 'Adolescent Medicine',\n",
        "       'Clinical Psychologist', 'Portable X-ray Supplier',\n",
        "       'Public Health/Welfare Clinic'], dtype=object)"
       ]
      }
     ],
     "prompt_number": 160
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "# see which group contains most physicians\n",
      "temp = dat_train.groupby('descr')\n",
      "temp.size().order(ascending = False)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 107,
       "text": [
        "descr\n",
        "Emergency Medicine             1802\n",
        "Internal Medicine              1503\n",
        "Diagnostic Radiology           1258\n",
        "Pediatrician                   1000\n",
        "Cardiology                      938\n",
        "Anesthesiology                  868\n",
        "Physician Specialty Unknown     832\n",
        "Family Practice                 796\n",
        "Hospital                        663\n",
        "CRNA (Nurse Ansesthetist)       655\n",
        "Nurse Practitioner              549\n",
        "Obstetrics/Gynecology           540\n",
        "General Surgery                 429\n",
        "Physician Assistant             417\n",
        "Pathology                       341\n",
        "...\n",
        "Proctology, Colon & Rectal Surgery    25\n",
        "Anesthesiology Assistant              23\n",
        "Maxillofacial Surgery                 19\n",
        "MRI, Independent DX Testing Center    19\n",
        "Gynecology                            16\n",
        "Oncology- Gynecological               15\n",
        "Preventive Medicine                   14\n",
        "Nuclear Medicine                      13\n",
        "Medical Genetics                       8\n",
        "Psychiatry                             7\n",
        "Adolescent Medicine                    7\n",
        "Sleep Disorders, Sleep Lab             6\n",
        "Orthopedic- Hand Surgery               6\n",
        "Clinical Psychologist                  3\n",
        "Public Health/Welfare Clinic           1\n",
        "Length: 73, dtype: int64"
       ]
      }
     ],
     "prompt_number": 107
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "# idea: try PCA for dimensional reduction then do clustering\n",
      "def PCA_module(dat, n_components):\n",
      "   \n",
      "    # create new design matrix\n",
      "    cols = [col for col in dat.columns if col not in ['spec1','PRVSEQ', 'descr']]\n",
      "    \n",
      "    # normalize data before doing PCA\n",
      "    design_mat = dat[cols]\n",
      "    #design_mat = (design_mat - design_mat.mean()) / design_mat.std()\n",
      "    \n",
      "    # PCA\n",
      "    from sklearn.decomposition import PCA\n",
      "    newX = PCA(n_components = n_components).fit_transform(design_mat)\n",
      "    \n",
      "    # create dataframe\n",
      "    newColnames = []\n",
      "    for i in xrange(0, n_components):\n",
      "        newColnames.append('Component' + str(i+1))\n",
      "    pcadat = pd.DataFrame(newX, columns =newColnames)\n",
      "    #pcadat['spec1'] = dat['spec1']\n",
      "    print pcadat.shape\n",
      "    pcadat['PRVSEQ'] = list(dat['PRVSEQ'])\n",
      "    \n",
      "    # move spec1 to 1st column\n",
      "    cols = pcadat.columns.tolist()\n",
      "    cols = cols[-1:] + cols[:-1]\n",
      "    pcadat = pcadat[cols]\n",
      "    return pcadat"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 154
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "names = [\n",
      "'Emergency Medicine',            \n",
      "'Internal Medicine',         \n",
      "'Diagnostic Radiology',          \n",
      "'Pediatrician']\n",
      "name = names[2]\n",
      "\n",
      "# \n",
      "dat_temp= dat_train[dat_train.descr==str(name)]\n",
      "\n",
      "# PCA\n",
      "dat_temp = PCA_module(dat_temp, 3)\n",
      "\n",
      "# add physician ID back\n",
      "dat_temp['PRVSEQ'] = physicians\n",
      "\n",
      "# 3D scatter plot\n",
      "import matplotlib.pyplot as plt\n",
      "from mpl_toolkits.mplot3d import Axes3D\n",
      "fig = plt.figure()\n",
      "ax = fig.add_subplot(111, projection='3d')\n",
      "ax.scatter(dat_temp['Component1'], dat_temp['Component2'] , dat_temp['Component3'])\n",
      "                                                                       \n",
      "ax.set_xlabel('1st p.c.')\n",
      "ax.set_ylabel('2nd p.c.')\n",
      "ax.set_zlabel('3rd p.c.')\n",
      "ax.set_title('Scatter plot of ' +name+ '\\n' + 'w.r.t. top 3 principal componants')\n",
      "plt.show()\n"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "(1267, 3)\n"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAV0AAAD5CAYAAACEcub7AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nNGhC/H+mfk3Z5dSt6AAAAABJRU5ErkJggg==\n",
       "text": [
        "<matplotlib.figure.Figure at 0xb304940>"
       ]
      }
     ],
     "prompt_number": 155
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "dat_temp.mean()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 109,
       "text": [
        "PRVSEQ        5.612414e+02\n",
        "Component1   -3.937751e-12\n",
        "Component2   -5.153495e-12\n",
        "Component3   -4.644188e-15\n",
        "dtype: float64"
       ]
      }
     ],
     "prompt_number": 109
    },
    {
     "cell_type": "heading",
     "level": 2,
     "metadata": {},
     "source": [
      "Fraud detection system 1"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "# reload data\n",
      "dat = pd.read_csv('ctnfl_final_with_descr.csv')\n",
      "dat_train, dat_test = random_sample(dat, 0.8)\n",
      "dat_train, dat_test = data_management(dat_train, dat_test)\n"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "(17390, 1284) (4353, 1284)\n"
       ]
      }
     ],
     "prompt_number": 25
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "dat_train.PRVSEQ.head()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 37,
       "text": [
        "839        1221\n",
        "3930       6631\n",
        "15051    101705\n",
        "3114       5293\n",
        "7438      12934\n",
        "Name: PRVSEQ, dtype: int64"
       ]
      }
     ],
     "prompt_number": 37
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "test = RandomForest(dat_train, dat_test, 290)\n",
      "test.fit('descr')\n",
      "test.suspicious1(0.95)\n",
      "#threshold = test.threshold(0.95)\n",
      "#test.suspicious1(demo, 0.95)\n",
      "\n",
      "#test.misrate()\n",
      "#test.heatmap()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "Obstetrics/Gynecology             30843.2450\n",
        "Orthopaedic Surgery               30759.5300\n",
        "Internal Medicine                  6078.1960\n",
        "Physician Specialty Unknown       90284.4200\n",
        "CRNA (Nurse Ansesthetist)          2781.4900\n",
        "Critical Care (Intensivist)       14395.7150\n",
        "Interventional Pain Management    23183.0050\n",
        "Thoracic Surgery                   7166.5150\n",
        "Nurse Practitioner                 5032.5055\n",
        "Pediatrician                      50683.3820\n",
        "Nephrology                         5353.3100\n",
        "Cardiology                         8999.9110\n",
        "Diagnostic Radiology               7438.9140\n",
        "Infectious Disease                 7384.6610\n",
        "Urology                           16744.6530\n",
        "...\n",
        "DME- Prosthetist/Orthotist            27622.9760\n",
        "Rehabilitation & Physical Medicine     8669.2450\n",
        "Dialysis                              55504.2960\n",
        "Urgent Care Center                    46451.0590\n",
        "Maxillofacial Surgery                  5694.7030\n",
        "Plastic and Reconstructive Surgery    11224.3500\n",
        "Multispecialty Clinic or Group        49593.0000\n",
        "Optometrist                           34677.9560\n",
        "Psychiatry                                   NaN\n",
        "Anesthesiology Assistant               4157.6875\n",
        "Ambulatory Surgical Center            62910.5440\n",
        "Gynecology                             5682.0240\n",
        "Osteopathy                                   NaN\n",
        "Sports Medicine                        8156.4845\n",
        "Adolescent Medicine                     678.5020\n",
        "Length: 70, dtype: float64\n",
        "[33, 76, 433, 552, 675, 676, 677, 794, 1058, 1076, 1206, 1498, 1818, 2112, 2232, 2592, 2642, 2938, 3045, 3072, 3080, 3082, 3094, 3135, 3152, 3188, 3191, 3283, 3288, 3337, 3340, 3373, 3442, 3459, 3823, 3954, 4200, 4526, 4543, 4575, 4599, 4614, 5032, 5314, 5367, 5380, 5431, 5496, 5498, 5522, 5686, 5703, 5827, 6087, 6152, 6281, 6282, 6379, 6566, 6708, 6710, 6779, 6791, 6853, 7032, 7082, 7105, 7491, 7529, 7535, 7573, 7578, 7583, 7594, 7607, 7608, 7647, 7946, 8528, 8870, 8883, 9349, 9726, 9766, 9772, 9794, 9907, 9938, 9950, 10228, 10415, 10438, 10527, 10598, 10715, 10809, 11414, 11569, 11599, 11692, 11988, 12245, 12296, 12387, 12509, 12773, 12932, 12985, 13176, 13264, 13347, 13399, 13628, 13813, 14337, 14748, 14849, 14877, 14930, 15081, 15445, 15533, 15603, 15906, 16035, 16346, 16529, 16612, 16623, 16764, 16855, 16919, 17036, 17237, 17339, 17737, 17805, 22240, 22342, 25569, 25581, 25731, 25736, 25746, 25820, 25872, 25910, 26063, 26277, 26294, 26318, 26373, 26598, 26618, 26646, 26673, 26855, 27014, 29325, 29362, 29427, 29564, 29846, 30322, 30348, 31017, 31394, 32491, 38613, 39141, 40607, 44138, 45097, 45652, 57665, 59663, 60209, 64151, 72650, 72659, 72661, 86184, 86190, 87684, 94691, 107204, 107217, 113493, 114507, 122024, 124004, 125604, 131641, 137159, 138192, 149690, 150192, 164684, 166197, 174746, 181762, 185773, 187310, 189788, 205280, 211311, 232283, 238290, 245303, 253283]"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        " 210\n"
       ]
      }
     ],
     "prompt_number": 124
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "input = dat_test\n",
      "print len(dat_test)\n",
      "suspicious_list = []\n",
      "for idx in input.index:\n",
      "    if input.tot_adj_paid[idx] > threshold[input.descr[idx]]:\n",
      "        suspicious_list.append(input.PRVSEQ[idx])\n",
      "print suspicious_list, len(suspicious_list)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "4355\n",
        "[86, 258, 443, 559, 607, 657, 735, 755, 913, 945, 1000, 1562, 1819, 1974, 2187, 2592, 3040, 3072, 3081, 3107, 3128, 3139, 3152, 3188, 3191, 3288, 3360, 3375, 3420, 3442, 3459, 3568, 3969, 4041, 4078, 4478, 4508, 4533, 4587, 4782, 4791, 4956, 5314, 5370, 5380, 5408, 5445, 5560, 5748, 5827, 6152, 6173, 6421, 6544, 6566, 6698, 6710, 6745, 6853, 7274, 7438, 7594, 7603, 7607, 7647, 7749, 7755, 7841, 8379, 8433, 8458, 8587, 8660, 8844, 8877, 8884, 8979, 9223, 9726, 9728, 9737, 9793, 9833, 10062, 10111, 10438, 10472, 10730, 11079, 11305, 11551, 11988, 12243, 12296, 12364, 12387, 12567, 12737, 12789, 12932, 13140, 13228, 13338, 13365, 13409, 13459, 13667, 13801, 13839, 13845, 13849, 13900, 14026, 14093, 14376, 14713, 14767, 14826, 14993, 15001, 15055, 15474, 15733, 15755, 15824, 15845, 16136, 16299, 16529, 16612, 16655, 16729, 16997, 17329, 22304, 22342, 25586, 25717, 25736, 25739, 25872, 25911, 25987, 26110, 26164, 26454, 26530, 26763, 26777, 26798, 26843, 27206, 27724, 28819, 29564, 29632, 29947, 30260, 30286, 31558, 32069, 41101, 42131, 42270, 42287, 45846, 70021, 71691, 86190, 89187, 98684, 101700, 108692, 108762, 109566, 110209, 114001, 121507, 124111, 137159, 143179, 146204, 158675, 159691, 200795, 232283, 244294, 246312, 249276, 251304]"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        " 190\n"
       ]
      }
     ],
     "prompt_number": 62
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "#dat_all = pd.read_csv('hmnm_6states_final_with_description.csv')\n",
      "#dat_all = pd.read_csv('hmnm_ilky_all_basic.csv')\n",
      "#dat_all = pd.read_csv('ctnfl_final_with_descr.csv')\n",
      "\n",
      "dat = pd.read_csv('ctnfl_final_with_descr.csv')\n",
      "dat_train, dat_test = random_sample(dat, 0.8)\n",
      "\n",
      "#dat_train = pd.read_csv('hmnm_il_basic_subset.csv')\n",
      "#dat_test = pd.read_csv('hmnm_ky_basic_subset.csv')\n",
      "\n",
      "\n",
      "num_features = int(dat_test.shape[1]*0.8)\n",
      "\n",
      "test = RandomForest(dat_train, dat_test, num_features)\n",
      "test.fit('spec1')\n",
      "#test.misrate()\n",
      "figure(1)\n",
      "test.heatmap()\n",
      "\n",
      "figure(2)\n",
      "test = NaiveBayes(dat_train, dat_test, num_features)\n",
      "test.fit('spec1')\n",
      "test.heatmap()\n",
      "\n",
      "figure(3)\n",
      "test = SVM(dat_train, dat_test, num_features)\n",
      "test.fit('spec1')\n",
      "test.heatmap()\n"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "display_data",
       "text": [
        "<matplotlib.figure.Figure at 0x508e1a58>"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAATwAAAFJCAYAAAAG6pF0AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
       "text": [
        "<matplotlib.figure.Figure at 0x508e1be0>"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "text": [
        "<matplotlib.figure.Figure at 0x67224c18>"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAATwAAAFKCAYAAACAfuPaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n/H/8w1Am4yPT8gAAAABJRU5ErkJggg==\n",
       "text": [
        "<matplotlib.figure.Figure at 0x6cfbe710>"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "text": [
        "<matplotlib.figure.Figure at 0xfcf72e8>"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAATwAAAFICAYAAADNtkLRAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAABJRU5ErkJggg==\n",
       "text": [
        "<matplotlib.figure.Figure at 0x483c7668>"
       ]
      }
     ],
     "prompt_number": 230
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "import numpy as np\n",
      "import scipy as sp\n",
      "import pandas as pd\n",
      "import random\n",
      "\n",
      "#%cd \"C:\\Users\\i55319\"\n",
      "% cd \"S:\\IIA\\Projects\\HCI\\Fraud Conditions\\specialty_practice\\SeanChang\"\n",
      "num_features = 240\n",
      "\n",
      "# TRAIN: hmnm_il_basic_subset \n",
      "# TEST: hmnm_ky_basic_subset:\n",
      "\n",
      "dat_train = pd.read_csv('hmnm_il_basic_subset.csv')\n",
      "dat_test = pd.read_csv('hmnm_ky_basic_subset.csv')\n",
      "dat_train, dat_test = data_management(dat_train, dat_test)\n",
      "test = RandomForest(dat_train, dat_test, num_features)\n",
      "test.fit('spec1')\n",
      "test.misrate()\n",
      "test.heatmap()\n",
      "\n",
      "\n",
      "# TRAIN: hmnm_il_all_subset \n",
      "# TEST: hmnm_ky_all_subset:\n",
      "\n",
      "dat_train = pd.read_csv('hmnm_il_all_subset.csv')\n",
      "dat_test = pd.read_csv('hmnm_ky_all_subset.csv')\n",
      "dat_train, dat_test = data_management(dat_train, dat_test)\n",
      "test = RandomForest(dat_train, dat_test,num_features)\n",
      "test.fit('spec1')\n",
      "test.misrate()\n",
      "test.heatmap()\n",
      "\n",
      "# TRAIN: alpha% of hmnm_ilky_all_subset\n",
      "# TEST: (1-alpha)% of hmnm_ilky_all_subset\n",
      "\n",
      "dat =pd.read_csv('hmnm_ilky_all_subset.csv')\n",
      "dat_train, dat_test = random_sample(dat, 0.9)\n",
      "test = RandomForest(dat_train, dat_test, num_features)\n",
      "test.fit('spec1')\n",
      "test.misrate()\n",
      "test.heatmap()\n",
      "\n",
      "# top num_features components of the PCA data generating from hmnm_ilky_all_subset\n",
      "# TRAIN: alpha % of above\n",
      "# TEST: (1-alpha)% of above\n",
      "\n",
      "#input = pd.read_csv('hmnm_ilky_all_subset.csv')\n",
      "#dat = pd.read_csv('hmnm_ilky_all_subset.csv')\n",
      "#dat_train, dat_test = random_sample(dat, 0.8)\n",
      "#dat_all = PCA_module(input, num_features) \n",
      "#result, z_result = random_forest_classifier(dat_train, dat_test, num_features)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "S:\\IIA\\Projects\\HCI\\Fraud Conditions\\specialty_practice\\SeanChang\n",
        "[0.52000000000000002, 0.070000000000000007, 0.14000000000000001, 0.14999999999999999, 0.56000000000000005, 0.070000000000000007, 0.19, 0.64000000000000001, 0.22, 0.040000000000000001]"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        " 0.26\n"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAToAAAFDCAYAAACqb8JVAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n/wPIbqri45HOOQAAAABJRU5ErkJggg==\n",
       "text": [
        "<matplotlib.figure.Figure at 0x22b143c8>"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "[0.70999999999999996, 0.089999999999999997, 0.70999999999999996, 0.59999999999999998, 0.19, 0.35999999999999999, 0.33000000000000002, 0.12, 0.31, 0.25] 0.367\n"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAToAAAFFCAYAAAB8NiFIAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nENLs/X/l/yxHz2ygfgAAAABJRU5ErkJggg==\n",
       "text": [
        "<matplotlib.figure.Figure at 0x19c63da0>"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "[0.56999999999999995, 0.17000000000000001, 0.25, 0.35999999999999999, 0.23999999999999999, 0.16, 0.45000000000000001, 0.70999999999999996, 0.33000000000000002, 0.26000000000000001] 0.35\n"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAToAAAFDCAYAAACqb8JVAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ngg==\n",
       "text": [
        "<matplotlib.figure.Figure at 0xb13e9e8>"
       ]
      }
     ],
     "prompt_number": 65
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "# TRAIN: two states data\n",
      "# TEST: il state\n",
      "# merge internal medicine/ general practice and general practice to the same group\n",
      "dat = pd.read_csv('hmnm_ilky_all_subset.csv')\n",
      "dat.spec1 = dat.spec1.replace( to_replace = ['1','8','11'], value = 'IM/GP/FP')\n",
      "dat.spec1.unique()\n",
      "dat_train, dat_test = random_sample(dat, 0.9)\n",
      "test = RandomForest(dat_train, dat_test, num_features)\n",
      "test.fit('spec1')\n",
      "test.misrate()\n",
      "test.heatmap()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "[0.34999999999999998, 0.19, 0.13, 0.29999999999999999, 0.16, 0.27000000000000002, 0.12, 0.34999999999999998] 0.23375\n"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAVkAAAFiCAYAAABcT6u1AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n6P8DtHH1DQw2tkYAAAAASUVORK5CYII=\n",
       "text": [
        "<matplotlib.figure.Figure at 0x1879ce48>"
       ]
      }
     ],
     "prompt_number": 68
    },
    {
     "cell_type": "heading",
     "level": 3,
     "metadata": {},
     "source": [
      "Load datasets"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "# merge two datasets from two states and split it to \n",
      "#alpha %, (1-alpha)% randomly as train and test data\n",
      "% cd \"S:\\IIA\\Projects\\HCI\\Fraud Conditions\\specialty_practice\\SeanChang\"\n",
      "dat_il = pd.read_csv('hmnm_il_all_basic.csv')\n",
      "dat_il['state'] = np.repeat('il', len(dat_il))\n",
      "\n",
      "dat_ky = pd.read_csv('hmnm_ky_all_basic.csv')\n",
      "dat_ky['state'] = np.repeat('ky', len(dat_ky))\n",
      "\n",
      "dat_il, dat_ky = data_managment(dat_il, dat_ky)\n",
      "dat_ilky = dat_il.append(dat_ky)\n",
      "print dat_ilky.shape\n",
      "#dat_all = pd.concat([dat_il, dat_ky], keys= ['il', 'ky'])\n",
      "dat_ilky.to_csv('hmnm_ilky_all_basic.csv')"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "S:\\IIA\\Projects\\HCI\\Fraud Conditions\\specialty_practice\\SeanChang\n",
        "(45603, 386)"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        " (17754, 386)\n",
        "(63357, 386)"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "\n"
       ]
      }
     ],
     "prompt_number": 23
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "# randomly select\n",
      "import random\n",
      "# seperate whole dataset as testing / training sets \n",
      "# randomly sample from each group\n",
      "index2 = random.sample(dat_all.index, int(0.7*dat_all.shape[0]))\n",
      "dat_train = dat_all.ix[index2]\n",
      "dat_test = dat_all.drop(index2)\n",
      "\n",
      "# see each group contains how many elements \n",
      "#print dat_train.shape, dat_test.shape\n",
      "#test = dat_train.groupby('spec1')\n",
      "#test2 = dat_test.groupby('spec1')\n",
      "#print test.size(), test2.size()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "# random sample keep the same number of each gruop\n",
      "from random import choice\n",
      "grouped = dat_ilky.groupby('spec1')\n",
      "subsampled = dat_ilky.ix[(choice(x) for x in grouped.groups.itervalues())]"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "#Loading data: \n",
      "#training: hmnm_il_all_subset; \n",
      "#testing: hmnm_ky_all_subset\n",
      "\n",
      "import numpy as np\n",
      "import scipy as sp\n",
      "import pandas as pd\n",
      "\n",
      "#%cd \"C:\\Users\\i55319\"\n",
      "% cd \"S:\\IIA\\Projects\\HCI\\Fraud Conditions\\specialty_practice\\SeanChang\"\n",
      "\n",
      "dat_train = pd.read_csv('hmnm_il_all_subset.csv')\n",
      "dat_test = pd.read_csv('hmnm_ky_all_subset.csv')\n",
      "\n",
      "#dat_train = pd.read_csv('hmnm_il_basic_subset.csv')\n",
      "#dat_test = pd.read_csv('hmnm_ky_basic_subset.csv')\n",
      "#dat_train = pd.read_csv('hmnm_il_all_basic.csv')\n",
      "#dat_test = pd.read_csv('hmnm_ky_all_basic.csv')\n",
      "print dat_train.shape"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "S:\\IIA\\Projects\\HCI\\Fraud Conditions\\specialty_practice\\SeanChang\n",
        "(1000, 2101)"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "\n"
       ]
      }
     ],
     "prompt_number": 2
    },
    {
     "cell_type": "heading",
     "level": 3,
     "metadata": {},
     "source": [
      "data preprocessing"
     ]
    },
    {
     "cell_type": "heading",
     "level": 3,
     "metadata": {},
     "source": [
      "random forest "
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "features = [col for col in dat_all.columns if col not in ['spec1']]\n",
      "import random\n",
      "index =random.sample(features, 10)\n",
      "#features_rnd = features[index]\n",
      "print index"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "['Component95', 'Component43', 'Component61', 'Component77', 'Component83', 'Component37', 'Component4', 'Component100', 'Component2', 'Component56']\n"
       ]
      }
     ],
     "prompt_number": 193
    },
    {
     "cell_type": "heading",
     "level": 3,
     "metadata": {},
     "source": [
      "Main codes"
     ]
    },
    {
     "cell_type": "heading",
     "level": 3,
     "metadata": {},
     "source": [
      "main code ends"
     ]
    },
    {
     "cell_type": "heading",
     "level": 3,
     "metadata": {},
     "source": [
      "full dataset\n"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "import pandas as pd\n",
      "\n",
      "#dat_all = pd.read_csv('hmnm_il_all_basic.csv')\n",
      "dat_all = pd.read_csv('hmnm_ilky_all_basic.csv')\n",
      "del dat_all['state'] \n",
      "\n",
      "import random\n",
      "index = random.sample(dat_all.index, int(dat_all.shape[0]* 0.8))\n",
      "dat_train = dat_all.loc[index]\n",
      "dat_test = dat_all.drop(index)\n",
      "dat_train, dat_test = data_managment(dat_train, dat_test)\n",
      "\n",
      "# modeling fitting \n",
      "result, z_result = random_forest_classifier(dat_train, dat_test, 100)\n",
      "output1 = pd.crosstab(dat_test['spec1'], result, rownames = ['actual'], colnames =['preds'])\n",
      "\n",
      "# add zero columns since #{prediction categories} <= #{true categories}\n",
      "# thesea are the difference\n",
      "# print  set(output2.index) - set(output2.columns) , len(set(output2.index) - set(output2.columns))\n",
      "for newname in list(set(output2.index) - set(output2.columns)):\n",
      "    #output1[newname] = output1[z_result[1]]\n",
      "    output2[newname] = zeros(output2.shape[0])\n",
      "print output2.shape\n",
      "\n",
      "# misclassification rate\n",
      "output2 = np.round(output2.apply(lambda r: r/r.sum(),axis = 1),2)\n",
      "heatmap(output2)\n"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "(50646, 386) (12672, 386)\n"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAUEAAAFCCAYAAABvmm+fAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\noAgRIm5riEJQhAgRtzX+P8Ud2M1NfzzYAAAAAElFTkSuQmCC\n",
       "text": [
        "<matplotlib.figure.Figure at 0x35da96d8>"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "(94, 94)\n"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAT4AAAFHCAYAAAA4FSA5AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
       "text": [
        "<matplotlib.figure.Figure at 0x2aab1be0>"
       ]
      }
     ],
     "prompt_number": 62
    },
    {
     "cell_type": "heading",
     "level": 2,
     "metadata": {},
     "source": [
      "Support Vector Machine"
     ]
    },
    {
     "cell_type": "heading",
     "level": 2,
     "metadata": {},
     "source": [
      "Neural Network"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "from pybrain.structure import FeedForwardNetwork\n",
      "n = FeedForwardNetwork()\n",
      "from pybrain.structure import LinearLayer, SigmoidLayer\n",
      "\n",
      "inLayer = LinearLayer(2)\n",
      "hiddenLayer = SigmoidLayer(3)\n",
      "outLayer = LinearLayer(1)\n",
      "\n",
      "n.addInputModule(inLayer)\n",
      "n.addModule(hiddenLayer)\n",
      "n.addOutputModule(outLayer)\n"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "ename": "SyntaxError",
       "evalue": "invalid syntax (<ipython-input-1-82c5365f8912>, line 10)",
       "output_type": "pyerr",
       "traceback": [
        "\u001b[1;36m  File \u001b[1;32m\"<ipython-input-1-82c5365f8912>\"\u001b[1;36m, line \u001b[1;32m10\u001b[0m\n\u001b[1;33m    from numpy.random import mmeans = [(-1,0),(2,4),(3,1)]\u001b[0m\n\u001b[1;37m                                    ^\u001b[0m\n\u001b[1;31mSyntaxError\u001b[0m\u001b[1;31m:\u001b[0m invalid syntax\n"
       ]
      }
     ],
     "prompt_number": 1
    },
    {
     "cell_type": "heading",
     "level": 2,
     "metadata": {},
     "source": [
      "Appendix"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "# random forest main module        \n",
      "from sklearn.ensemble import RandomForestClassifier\n",
      "\n",
      "# --- random forest --- #\n",
      "def random_forest_classifier(dat_train, dat_test, num_features):\n",
      "    features = [col for col in dat_train.columns if col not in ['spec1']]\n",
      "    import random\n",
      "    features_rnd = random.sample(features, num_features)\n",
      "    clf = RandomForestClassifier(n_jobs = 5)\n",
      "    y, z = pd.factorize(dat_train['spec1'])\n",
      "    clf.fit(dat_train[features_rnd], y)\n",
      "    preds = clf.predict(dat_test[features_rnd])\n",
      "    preds = np.array(map(lambda cat: z[cat], preds))\n",
      "\n",
      "    output1 = pd.crosstab(dat_test['spec1'], preds, rownames=['actual'], colnames=['preds'])\n",
      "    # call for outputs\n",
      "    cross_validation(dat_test['spec1'], preds, z)\n",
      "    return preds, z"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 199
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "test = dat_test.groupby('spec1')\n",
      "test.size()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 5,
       "text": [
        "spec1\n",
        "1        10\n",
        "5        11\n",
        "8        11\n",
        "11       11\n",
        "20       11\n",
        "24       11\n",
        "33       11\n",
        "35       11\n",
        "37       11\n",
        "65       11\n",
        "dtype: int64"
       ]
      }
     ],
     "prompt_number": 5
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "#dat_test[range(int(0.8 * dat_train.shape[1]), dat_train.shape[1])]\n",
      "#print dat_train.shape, dat_test.shape\n",
      "#dat_train.reindex( np.random.permutation(dat_train.index))\n",
      "#dat_train = dat_train[perm,:]\n",
      "\n",
      "#columns = dat_train.columns\n",
      "#type(columns)\n",
      "#dat_train[:5]\n",
      "#dat_train[[2,3,4]]  # this is how we select columns by index\n",
      "#print dat_train[500:]\n",
      "#dat_train[2:3]\n",
      "#dat_train.columns[2:5]\n",
      "#dat_train.shape[1]\n",
      "#dat_train[range(0,dat_train.shape[1])]\n",
      "#dat_test[dat_test.columns[1:50]].describe()  # select i-j th columns\n",
      "#dat_test.loc[0:100, dat_test.sum(0)!=0]       # drop those all zero columns\n",
      "#dat_test.loc[2,'spec1']                      # select i-th row, j-th column\n",
      "#dat_test[0:2]\n",
      "#dat_test.sum(0)!=0\n",
      "#dat_train.sum(0)!=0\n",
      "#dat_test.loc[0:100, dat_test.sum(0)!=0]\n",
      "\n",
      "#bygroup_test = dat_test.groupby('spec1')      # check number of entries in each group\n",
      "#bygroup_test.size()\n",
      "#bygroup_test.describe()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "index = [1,5,8,11,20,24,33,35,37,65]\n",
      "output.columns = index\n",
      "output2 = np.round(output.apply(lambda r: r/r.sum(),axis = 1),2)\n",
      "misrate = []\n",
      "print output2\n",
      "for i in xrange(0,len(index)):\n",
      "    misrate.append(1 - output2[index[i]][index[i]])\n",
      "print 'misclasssfication rates are' + str(np.round(misrate,2))"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "          1     5     8     11    20    24    33    35    37    65\n",
        "actual                                                            \n",
        "1       0.43  0.00  0.17  0.31  0.02  0.00  0.01  0.00  0.05  0.01\n",
        "5       0.00  0.95  0.00  0.00  0.00  0.00  0.02  0.00  0.00  0.03\n",
        "8       0.27  0.01  0.27  0.13  0.03  0.06  0.10  0.00  0.09  0.04\n",
        "11      0.37  0.01  0.12  0.36  0.04  0.00  0.02  0.00  0.08  0.00\n",
        "20      0.00  0.02  0.03  0.01  0.88  0.06  0.00  0.00  0.00  0.00\n",
        "24      0.00  0.00  0.08  0.00  0.03  0.82  0.06  0.00  0.01  0.00\n",
        "33      0.00  0.02  0.04  0.04  0.04  0.15  0.71  0.00  0.00  0.00\n",
        "35      0.00  0.00  0.00  0.01  0.01  0.00  0.00  0.98  0.00  0.00\n",
        "37      0.21  0.01  0.19  0.01  0.01  0.00  0.04  0.00  0.53  0.00\n",
        "65      0.03  0.00  0.02  0.02  0.02  0.00  0.03  0.01  0.00  0.87\n",
        "\n",
        "[10 rows x 10 columns]\n",
        "misclasssfication rates are[ 0.57  0.05  0.73  0.64  0.12  0.18  0.29  0.02  0.47  0.13]\n"
       ]
      }
     ],
     "prompt_number": 61
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "from sklearn.ensemble import RandomForestClassifier\n",
      "def random_forest(dat_train, dat_test):\n",
      "        #features = dat_train.columns[2:100]\n",
      "        #features = [col for col in dat_train.columns if col not in ['spec1']]\n",
      "    clf = RandomForestClassifier(n_jobs = 5)\n",
      "    y, z = pd.factorize(dat_train['spec1'])\n",
      "    clf.fit(dat_train[range(1,100)], y)\n",
      "    preds = clf.predict(dat_test[range(1,100)])\n",
      "    output = pd.crosstab(dat_test['spec1'], preds, rownames=['actual'], colnames=['preds'])\n",
      "    index = [1,5,8,11,20,24,33,35,37,65]\n",
      "    output.columns = index\n",
      "        # counts -> rates\n",
      "    output2 = np.round(output.apply(lambda r: r/r.sum(),axis = 1),2)\n",
      "    misrate = []\n",
      "    print output2\n",
      "        # rates -> misclassfication rates\n",
      "    for i in xrange(0,len(index)):\n",
      "        misrate.append(1 - output2[index[i]][index[i]])\n",
      "    print 'misclasssfication rates are' + str(np.round(misrate,2))\n",
      "    \n",
      "    return output2, misrate\n",
      "\n",
      "\n",
      "output, vec = random_forest(dat_train, dat_test)\n",
      "print mean(vec)\n",
      "# plot \n"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "ename": "KeyError",
       "evalue": "u'no item named __dummy__'",
       "output_type": "pyerr",
       "traceback": [
        "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m\n\u001b[1;31mKeyError\u001b[0m                                  Traceback (most recent call last)",
        "\u001b[1;32m<ipython-input-131-104cc7b31e06>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m()\u001b[0m\n\u001b[0;32m     22\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     23\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 24\u001b[1;33m \u001b[0moutput\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mvec\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mrandom_forest\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mdat_train\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mdat_test\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     25\u001b[0m \u001b[1;32mprint\u001b[0m \u001b[0mmean\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mvec\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     26\u001b[0m \u001b[1;31m# plot\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
        "\u001b[1;32m<ipython-input-131-104cc7b31e06>\u001b[0m in \u001b[0;36mrandom_forest\u001b[1;34m(dat_train, dat_test)\u001b[0m\n\u001b[0;32m      7\u001b[0m     \u001b[0mclf\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mfit\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mdat_train\u001b[0m\u001b[1;33m[\u001b[0m\u001b[0mrange\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;36m1\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;36m10\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0my\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      8\u001b[0m     \u001b[0mpreds\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mclf\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mpredict\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mdat_test\u001b[0m\u001b[1;33m[\u001b[0m\u001b[0mrange\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;36m1\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;36m10\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m----> 9\u001b[1;33m     \u001b[0moutput\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mpd\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcrosstab\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mdat_test\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;34m'spec1'\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mpreds\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mrownames\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;34m'actual'\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mcolnames\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;34m'preds'\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     10\u001b[0m     \u001b[0mindex\u001b[0m \u001b[1;33m=\u001b[0m \u001b[1;33m[\u001b[0m\u001b[1;36m1\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;36m5\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;36m8\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;36m11\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;36m20\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;36m24\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;36m33\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;36m35\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;36m37\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;36m65\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     11\u001b[0m     \u001b[0moutput\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcolumns\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mindex\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
        "\u001b[1;32mC:\\Users\\i55319\\AppData\\Local\\Enthought\\Canopy\\User\\lib\\site-packages\\pandas\\tools\\pivot.pyc\u001b[0m in \u001b[0;36mcrosstab\u001b[1;34m(rows, cols, values, rownames, colnames, aggfunc, margins, dropna)\u001b[0m\n\u001b[0;32m    368\u001b[0m         \u001b[0mdf\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;34m'__dummy__'\u001b[0m\u001b[1;33m]\u001b[0m \u001b[1;33m=\u001b[0m \u001b[1;36m0\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    369\u001b[0m         table = df.pivot_table('__dummy__', rows=rownames, cols=colnames,\n\u001b[1;32m--> 370\u001b[1;33m                                aggfunc=len, margins=margins, dropna=dropna)\n\u001b[0m\u001b[0;32m    371\u001b[0m         \u001b[1;32mreturn\u001b[0m \u001b[0mtable\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mfillna\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;36m0\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mastype\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mnp\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mint64\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    372\u001b[0m     \u001b[1;32melse\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
        "\u001b[1;32mC:\\Users\\i55319\\AppData\\Local\\Enthought\\Canopy\\User\\lib\\site-packages\\pandas\\tools\\pivot.pyc\u001b[0m in \u001b[0;36mpivot_table\u001b[1;34m(data, values, rows, cols, aggfunc, fill_value, margins, dropna)\u001b[0m\n\u001b[0;32m    138\u001b[0m     \u001b[1;31m# discard the top level\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    139\u001b[0m     \u001b[1;32mif\u001b[0m \u001b[0mvalues_passed\u001b[0m \u001b[1;32mand\u001b[0m \u001b[1;32mnot\u001b[0m \u001b[0mvalues_multi\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 140\u001b[1;33m         \u001b[0mtable\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mtable\u001b[0m\u001b[1;33m[\u001b[0m\u001b[0mvalues\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;36m0\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    141\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    142\u001b[0m     \u001b[1;32mif\u001b[0m \u001b[0mlen\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mrows\u001b[0m\u001b[1;33m)\u001b[0m \u001b[1;33m==\u001b[0m \u001b[1;36m0\u001b[0m \u001b[1;32mand\u001b[0m \u001b[0mlen\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mcols\u001b[0m\u001b[1;33m)\u001b[0m \u001b[1;33m>\u001b[0m \u001b[1;36m0\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
        "\u001b[1;32mC:\\Users\\i55319\\AppData\\Local\\Enthought\\Canopy\\User\\lib\\site-packages\\pandas\\core\\frame.pyc\u001b[0m in \u001b[0;36m__getitem__\u001b[1;34m(self, key)\u001b[0m\n\u001b[0;32m   1656\u001b[0m             \u001b[1;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_getitem_multilevel\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mkey\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1657\u001b[0m         \u001b[1;32melse\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m-> 1658\u001b[1;33m             \u001b[1;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_getitem_column\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mkey\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m   1659\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1660\u001b[0m     \u001b[1;32mdef\u001b[0m \u001b[0m_getitem_column\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mkey\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
        "\u001b[1;32mC:\\Users\\i55319\\AppData\\Local\\Enthought\\Canopy\\User\\lib\\site-packages\\pandas\\core\\frame.pyc\u001b[0m in \u001b[0;36m_getitem_column\u001b[1;34m(self, key)\u001b[0m\n\u001b[0;32m   1663\u001b[0m         \u001b[1;31m# get column\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1664\u001b[0m         \u001b[1;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcolumns\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mis_unique\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m-> 1665\u001b[1;33m             \u001b[1;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_get_item_cache\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mkey\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m   1666\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1667\u001b[0m         \u001b[1;31m# duplicate columns & possible reduce dimensionaility\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
        "\u001b[1;32mC:\\Users\\i55319\\AppData\\Local\\Enthought\\Canopy\\User\\lib\\site-packages\\pandas\\core\\generic.pyc\u001b[0m in \u001b[0;36m_get_item_cache\u001b[1;34m(self, item)\u001b[0m\n\u001b[0;32m   1003\u001b[0m         \u001b[0mres\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mcache\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mget\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mitem\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1004\u001b[0m         \u001b[1;32mif\u001b[0m \u001b[0mres\u001b[0m \u001b[1;32mis\u001b[0m \u001b[0mNone\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m-> 1005\u001b[1;33m             \u001b[0mvalues\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_data\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mget\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mitem\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m   1006\u001b[0m             \u001b[0mres\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_box_item_values\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mitem\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mvalues\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1007\u001b[0m             \u001b[0mcache\u001b[0m\u001b[1;33m[\u001b[0m\u001b[0mitem\u001b[0m\u001b[1;33m]\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mres\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
        "\u001b[1;32mC:\\Users\\i55319\\AppData\\Local\\Enthought\\Canopy\\User\\lib\\site-packages\\pandas\\core\\internals.pyc\u001b[0m in \u001b[0;36mget\u001b[1;34m(self, item)\u001b[0m\n\u001b[0;32m   2872\u001b[0m                 \u001b[1;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mget_for_nan_indexer\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mindexer\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   2873\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m-> 2874\u001b[1;33m             \u001b[0m_\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mblock\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_find_block\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mitem\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m   2875\u001b[0m             \u001b[1;32mreturn\u001b[0m \u001b[0mblock\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mget\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mitem\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   2876\u001b[0m         \u001b[1;32melse\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
        "\u001b[1;32mC:\\Users\\i55319\\AppData\\Local\\Enthought\\Canopy\\User\\lib\\site-packages\\pandas\\core\\internals.pyc\u001b[0m in \u001b[0;36m_find_block\u001b[1;34m(self, item)\u001b[0m\n\u001b[0;32m   3184\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   3185\u001b[0m     \u001b[1;32mdef\u001b[0m \u001b[0m_find_block\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mitem\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m-> 3186\u001b[1;33m         \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_check_have\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mitem\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m   3187\u001b[0m         \u001b[1;32mfor\u001b[0m \u001b[0mi\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mblock\u001b[0m \u001b[1;32min\u001b[0m \u001b[0menumerate\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mblocks\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   3188\u001b[0m             \u001b[1;32mif\u001b[0m \u001b[0mitem\u001b[0m \u001b[1;32min\u001b[0m \u001b[0mblock\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
        "\u001b[1;32mC:\\Users\\i55319\\AppData\\Local\\Enthought\\Canopy\\User\\lib\\site-packages\\pandas\\core\\internals.pyc\u001b[0m in \u001b[0;36m_check_have\u001b[1;34m(self, item)\u001b[0m\n\u001b[0;32m   3191\u001b[0m     \u001b[1;32mdef\u001b[0m \u001b[0m_check_have\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mitem\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   3192\u001b[0m         \u001b[1;32mif\u001b[0m \u001b[0mitem\u001b[0m \u001b[1;32mnot\u001b[0m \u001b[1;32min\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mitems\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m-> 3193\u001b[1;33m             \u001b[1;32mraise\u001b[0m \u001b[0mKeyError\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;34m'no item named %s'\u001b[0m \u001b[1;33m%\u001b[0m \u001b[0mcom\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mpprint_thing\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mitem\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m   3194\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   3195\u001b[0m     def reindex_axis(self, new_axis, indexer=None, method=None, axis=0,\n",
        "\u001b[1;31mKeyError\u001b[0m: u'no item named __dummy__'"
       ]
      }
     ],
     "prompt_number": 131
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "# Let's try SQL\n",
      "# join il & ky datasets\n",
      "from sql import *\n",
      "from sql.aggregate import *\n",
      "from sql.conditionals import *\n",
      "\n",
      "dat_il = pd.read_csv('hmnm_il_all_subset.csv')\n",
      "dat_ky = pd.read_csv('hmnm_ky_all_subset.csv')\n",
      "\n"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "dat_all = dat_train\n",
      "import random\n",
      "# seperate whole dataset as testing / training sets \n",
      "# randomly sample from each group\n",
      "#dat_all = pd.concat([dat_il, dat_ky])\n",
      "index2 = random.sample(dat_all.index, int(0.8*dat_all.shape[0]))\n",
      "dat_train = dat_all.ix[index2]\n",
      "dat_test = dat_all.drop(index2)\n",
      "print dat_train.shape, dat_test.shape"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "(800, 297) (200, 297)\n"
       ]
      }
     ],
     "prompt_number": 17
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "# random sample from data\n",
      "\n",
      "dat_all = dat_ilky #pd.read_csv('hmnm_il_all_subset.csv')\n",
      "print dat_all.shape\n",
      "# Method 1 #\n",
      "# generate test & training samples from the same dataset\n",
      "# 80% for training, 10% for testing\n",
      "#dat_train = dat_all[1:int(0.9 * dat_all.shape[0])]\n",
      "#dat_test = dat_all[int(0.9 * dat_all.shape[0]): dat_all.shape[0]]\n",
      "### randomly select training/ testing samples performs really bad ###\n",
      "\n",
      "# Method 2 # \n",
      "# from each speciality, select first 90% as training\n",
      "index = []\n",
      "for i in xrange(0,10):\n",
      "    index += map(lambda x: x+100*i, range(1,90))\n",
      "dat_train = dat_all.iloc[index] \n",
      "dat_test = dat_all.iloc[[rows for rows in range(1,dat_all.shape[0]) if rows not in index]]\n",
      "print dat_train.shape, dat_test.shape\n"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "(1990, 1528)\n",
        "(890, 1528)"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        " (1099, 1528)\n"
       ]
      }
     ],
     "prompt_number": 70
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "def SVM_classifier(dat_train, dat_test, num_features):\n",
      "    features = [col for col in dat_train.columns if col not in ['spec1']]\n",
      "    import random\n",
      "    features_rnd = random.sample(features, num_features)\n",
      "    X = dat_train[features_rnd]\n",
      "    y, z = pd.factorize(dat_train['spec1'])\n",
      "    from sklearn.svm import SVC\n",
      "    clf = SVC()\n",
      "    clf.fit(X, y) \n",
      "    SVC(C=1.0, cache_size=200, class_weight=None, coef0=0.0, degree=3,\n",
      "        gamma=0.0, kernel='rbf', max_iter=-1, probability=False,\n",
      "        random_state=None, shrinking=True, tol=0.001, verbose=False)\n",
      "    preds = clf.predict(dat_test[features_rnd])\n",
      "    preds = np.array(map(lambda cat: z[cat], preds))\n",
      "    cross_validation(dat_test['spec1'], preds, z)\n",
      "    return preds, z"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 55
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "def PCA_module(dat, num_features):\n",
      "    # create new design matrix\n",
      "    cols = [col for col in dat.columns if col not in 'spec1' ]\n",
      "    from sklearn.decomposition import PCA\n",
      "    newX = PCA(n_components = num_features).fit_transform(dat[cols])\n",
      "    \n",
      "    # create dataframe\n",
      "    newColnames = []\n",
      "    for i in xrange(0, num_features):\n",
      "        newColnames.append('Component' + str(i+1))\n",
      "    pcadat = pd.DataFrame(newX, columns =newColnames)\n",
      "    pcadat['spec1'] = dat['spec1']\n",
      "    \n",
      "    # move spec1 to 1st column\n",
      "    cols = pcadat.columns.tolist()\n",
      "    cols = cols[-1:] + cols[:-1]\n",
      "    pcadat = pcadat[cols]\n",
      "    return pcadat\n",
      "\n"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 9
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "\n",
      "import numpy as np\n",
      "import scipy as sp\n",
      "import pandas as pd\n",
      "\n",
      "#%cd \"C:\\Users\\i55319\"\n",
      "% cd \"S:\\IIA\\Projects\\HCI\\Fraud Conditions\\specialty_practice\\SeanChang\"\n",
      "\n",
      "# TRAIN: hmnm_il_basic_subset \n",
      "# TEST: hmnm_ky_basic_subset:\n",
      "\n",
      "dat_train = pd.read_csv('hmnm_il_basic_subset.csv')\n",
      "dat_test = pd.read_csv('hmnm_ky_basic_subset.csv')\n",
      "dat_train, dat_test = data_managment(dat_train, dat_test)\n",
      "result, z_result = SVM_classifier(dat_train, dat_test, 297)\n",
      "cross_validation(dat_test['spec1'], result, z_result)\n",
      "\n",
      "# TRAIN: hmnm_il_all_subset \n",
      "# TEST: hmnm_ky_all_subset:\n",
      "\n",
      "dat_train = pd.read_csv('hmnm_il_all_subset.csv')\n",
      "dat_test = pd.read_csv('hmnm_ky_all_subset.csv')\n",
      "dat_train, dat_test = data_managment(dat_train, dat_test)\n",
      "result, z_result = SVM_classifier(dat_train, dat_test, 297)\n",
      "cross_validation(dat_test['spec1'], result, z_result)\n",
      "\n",
      "# TRAIN: alpha% of hmnm_ilky_all_subset\n",
      "# TEST: (1-alpha)% of hmnm_ilky_all_subset\n",
      "\n",
      "alpha = 0.5\n",
      "dat_all = pd.read_csv('hmnm_ilky_all_subset.csv')\n",
      "index = random.sample(dat_all.index, int(alpha*dat_all.shape[0]))\n",
      "dat_train = dat_all.ix[index]\n",
      "dat_test = dat_all.drop(index)\n",
      "result, z_result = SVM_classifier(dat_train, dat_test, 297)\n",
      "cross_validation(dat_test['spec1'], result, z_result)\n"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "S:\\IIA\\Projects\\HCI\\Fraud Conditions\\specialty_practice\\SeanChang\n",
        "(1000, 297)"
       ]
      },
      {
       "ename": "ValueError",
       "evalue": "sample larger than population",
       "output_type": "pyerr",
       "traceback": [
        "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m\n\u001b[1;31mValueError\u001b[0m                                Traceback (most recent call last)",
        "\u001b[1;32m<ipython-input-207-7b97281245b8>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m()\u001b[0m\n\u001b[0;32m     13\u001b[0m \u001b[0mdat_test\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mpd\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mread_csv\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;34m'hmnm_ky_basic_subset.csv'\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     14\u001b[0m \u001b[0mdat_train\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mdat_test\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mdata_managment\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mdat_train\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mdat_test\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 15\u001b[1;33m \u001b[0mresult\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mz_result\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mSVM_classifier\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mdat_train\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mdat_test\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;36m297\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     16\u001b[0m \u001b[0mcross_validation\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mdat_test\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;34m'spec1'\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mresult\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mz_result\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     17\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
        "\u001b[1;32m<ipython-input-55-90aa72be6ebd>\u001b[0m in \u001b[0;36mSVM_classifier\u001b[1;34m(dat_train, dat_test, num_features)\u001b[0m\n\u001b[0;32m      2\u001b[0m     \u001b[0mfeatures\u001b[0m \u001b[1;33m=\u001b[0m \u001b[1;33m[\u001b[0m\u001b[0mcol\u001b[0m \u001b[1;32mfor\u001b[0m \u001b[0mcol\u001b[0m \u001b[1;32min\u001b[0m \u001b[0mdat_train\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcolumns\u001b[0m \u001b[1;32mif\u001b[0m \u001b[0mcol\u001b[0m \u001b[1;32mnot\u001b[0m \u001b[1;32min\u001b[0m \u001b[1;33m[\u001b[0m\u001b[1;34m'spec1'\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      3\u001b[0m     \u001b[1;32mimport\u001b[0m \u001b[0mrandom\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m----> 4\u001b[1;33m     \u001b[0mfeatures_rnd\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mrandom\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0msample\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mfeatures\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mnum_features\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m      5\u001b[0m     \u001b[0mX\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mdat_train\u001b[0m\u001b[1;33m[\u001b[0m\u001b[0mfeatures_rnd\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      6\u001b[0m     \u001b[0my\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mz\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mpd\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mfactorize\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mdat_train\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;34m'spec1'\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
        "\u001b[1;32mC:\\Users\\i55319\\AppData\\Local\\Enthought\\Canopy\\App\\appdata\\canopy-1.4.0.1938.win-x86_64\\lib\\random.pyc\u001b[0m in \u001b[0;36msample\u001b[1;34m(self, population, k)\u001b[0m\n\u001b[0;32m    319\u001b[0m         \u001b[0mn\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mlen\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mpopulation\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    320\u001b[0m         \u001b[1;32mif\u001b[0m \u001b[1;32mnot\u001b[0m \u001b[1;36m0\u001b[0m \u001b[1;33m<=\u001b[0m \u001b[0mk\u001b[0m \u001b[1;33m<=\u001b[0m \u001b[0mn\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 321\u001b[1;33m             \u001b[1;32mraise\u001b[0m \u001b[0mValueError\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;34m\"sample larger than population\"\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    322\u001b[0m         \u001b[0mrandom\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mrandom\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    323\u001b[0m         \u001b[0m_int\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mint\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
        "\u001b[1;31mValueError\u001b[0m: sample larger than population"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        " (990, 297)\n"
       ]
      }
     ],
     "prompt_number": 207
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "from sklearn import metrics\n",
      "from sklearn.cluster import KMeans\n",
      "from sklearn.datasets import load_digits\n",
      "from sklearn.decomposition import PCA\n",
      "from sklearn.preprocessing import scale\n",
      "\n",
      "kmeans = KMeans(init='k-means++', n_clusters=2 , n_init=2)\n",
      "kmeans.fit(dat_temp3)\n",
      "\n",
      "# Step size of the mesh. Decrease to increase the quality of the VQ.\n",
      "h = .02     # point in the mesh [x_min, m_max]x[y_min, y_max].\n",
      "x_min, x_max = dat_temp3[[0]].min() + 1, dat_temp3[[0]].max() - 1\n",
      "y_min, y_max = dat_temp3[[1]].min() + 1, dat_temp3[[1]].max() - 1\n",
      "xx, yy = np.meshgrid(np.arange(x_min, x_max, h), np.arange(y_min, y_max, h))\n",
      "\n",
      "# plot\n",
      "pl.clf()\n",
      "pl.imshow(dat_temp3, interpolation='nearest',\n",
      "          extent=(xx.min(), xx.max(), yy.min(), yy.max()),\n",
      "          cmap=pl.cm.Paired,\n",
      "          aspect='auto', origin='lower')\n",
      "\n",
      "pl.plot(dat_temp3[[0]], dat_temp3[[1]], 'k.', markersize=2)\n",
      "import pylab as pl\n",
      "\n",
      "figure(2)\n",
      "centroids = kmeans.cluster_centers_\n",
      "pl.scatter(centroids[:, 0], centroids[:, 1],\n",
      "           marker='x', s=169, linewidths=3,\n",
      "           color='b', zorder=10)\n",
      "pl.plot(dat_temp3[[0]], dat_temp3[[1]], 'k.', markersize=2)\n"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 59,
       "text": [
        "[<matplotlib.lines.Line2D at 0x6039f588>]"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAXUAAAEACAYAAABMEua6AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n7LH+rHqy7gAAAABJRU5ErkJggg==\n",
       "text": [
        "<matplotlib.figure.Figure at 0x15435ac8>"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAXYAAAEACAYAAACnJV25AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\niYhchsFOROQyDHYiIpdhsBMRucz/B1XPe/kQ5HYYAAAAAElFTkSuQmCC\n",
       "text": [
        "<matplotlib.figure.Figure at 0x6039f710>"
       ]
      }
     ],
     "prompt_number": 59
    }
   ],
   "metadata": {}
  }
 ]
}