dprice_binned_classifier_comparison.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import extract\n",
    "import visualize\n",
    "import seaborn as sns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn import preprocessing\n",
    "from sklearn.model_selection import train_test_split, GridSearchCV\n",
    "from sklearn.metrics import confusion_matrix\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.ensemble import RandomForestClassifier, AdaBoostClassifier\n",
    "from sklearn.naive_bayes import GaussianNB"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "d = extract.get_data()\n",
    "d = extract.clean_data(d)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "s0 = pd.read_csv(\"price_change_data.csv\", usecols = ['d_index', 'tradetype',\n",
    "\t\t'state', 'coupon', 'maturity', 'issuetype', 'issuesource', 'rtg',\n",
    "\t\t'tradedate', 'dprice', 'holdtime'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "d_index = s0.pop('d_index')\n",
    "s_state_features = extract.build_state_features(s0, num_states = 20)\n",
    "s_other_features = extract.build_other_text_features(s0)\n",
    "s = pd.concat([s0.drop(['state', 'issuetype', 'issuesource', 'tradetype'], axis = 1), s_state_features, s_other_features], axis = 1)\n",
    "s = s.dropna()\n",
    "s = s[s.holdtime < 7] #pick out only data with holdtime < 7 days"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "dprice = s.pop('dprice')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Bin First, Then Classify"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Binarize the dprice variable: 0 means loss, 1 means a gain:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "dprice_binary = dprice.apply(lambda x: x > 0).astype(np.int)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "s_train, s_test, dp_train, dp_test = train_test_split(s, dprice_binary, test_size=0.20, random_state=203)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Set the various classifiers to try:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "names = [\"Decision Tree\", \"Random Forest\", \"Naive Bayes\"]\n",
    "classifiers = [DecisionTreeClassifier(max_depth=25),\n",
    "\t\tRandomForestClassifier(max_depth=25, n_estimators=100),\n",
    "\t\tGaussianNB()]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Decision Tree: number correct predictions/total samples = 0.902369668246\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAcMAAAGOCAYAAADrQGhsAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nIm5PYSgiIm7v/wEuy5ZDAQnvzQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fc237832050>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Random Forest: number correct predictions/total samples = 0.912606635071\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAcMAAAGOCAYAAADrQGhsAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nCEMAgNv7/wCkVEOETKrmogAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fc237433410>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Naive Bayes: number correct predictions/total samples = 0.746161137441\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAcMAAAGOCAYAAADrQGhsAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fc23712b610>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "for name, clf in zip(names, classifiers):\n",
    "\t\tclf = clf.fit(s_train, dp_train)\n",
    "\t\tdp_pred = clf.predict(s_test)\n",
    "        \tscore = clf.score(s_test, dp_test)\n",
    "\t\tprint(name + \": number correct predictions/total samples = {}\".format(score))\n",
    "\t\t# Compute confusion matrix\n",
    "\t\tcnf_matrix = confusion_matrix(dp_test, dp_pred)\n",
    "\t\tplt.figure()\n",
    "\t\tvisualize.plot_confusion_matrix(cnf_matrix, classes=[0, 1], title='Confusion matrix, without normalization')\n",
    "\t\tplt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Now we can try to predict more precisely how much of a gain or loss one can potentially get, using more bins:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "bins = [-3, -1, 0, 1, 3]\n",
    "dprice_binned = np.digitize(dprice.values, bins)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([2, 3, 3, 2, 3])"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dprice_binned[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([-0.762,  0.376,  0.473, -0.665,  0.473])"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dprice[:5].values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "s_train, s_test, dp_train, dp_test = train_test_split(s, dprice_binned, test_size=0.20, random_state=53)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Decision Tree: number correct predictions/total samples = 0.637725118483\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAcQAAAGOCAYAAAAJnHMVAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nKogiIiKACqKIiAiggigiIgLA/wE7nuAs4U0+EQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fc2377a2950>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Random Forest: number correct predictions/total samples = 0.652511848341\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAcQAAAGOCAYAAAAJnHMVAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n9eJG9HehqQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fc23753ba50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Naive Bayes: number correct predictions/total samples = 0.349004739336\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAcQAAAGOCAYAAAAJnHMVAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nQRQREQHgfwE7/y7/LhPumAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fc237226e50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "for name, clf in zip(names, classifiers):\n",
    "\t\tclf = clf.fit(s_train, dp_train)\n",
    "\t\tdp_pred = clf.predict(s_test)\n",
    "\t\tscore = clf.score(s_test, dp_test)\n",
    "\t\tprint(name + \": number correct predictions/total samples = {}\".format(score))\n",
    "\t\t# Compute confusion matrix\n",
    "\t\tcnf_matrix = confusion_matrix(dp_test, dp_pred)\n",
    "\t\tplt.figure()\n",
    "\t\tvisualize.plot_confusion_matrix(cnf_matrix, classes=bins, title='Confusion matrix, without normalization')\n",
    "\t\tplt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Let us try to see more closely which bonds are yielding positive returns (bin index 3-5 represent positive gains):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "clf = RandomForestClassifier(max_depth=25, n_estimators=100)\n",
    "clf = clf.fit(s_train, dp_train)\n",
    "dp_pred = clf.predict(s)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[2 2 3 2 2 3 2 2 3 1 1 1 0 0 0 1 1 1 1 2]\n"
     ]
    }
   ],
   "source": [
    "print dp_pred[:20]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[2 3 3 2 3 3 2 2 4 1 1 1 0 0 0 1 1 1 1 1]\n"
     ]
    }
   ],
   "source": [
    "print dprice_binned[:20]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[ -7.62000000e-01   3.76000000e-01   4.73000000e-01  -6.65000000e-01\n",
      "   4.73000000e-01   5.70000000e-01  -3.00000000e-03  -1.50000000e-02\n",
      "   1.12600000e+00  -1.96700000e+00  -1.81700000e+00  -2.20400000e+00\n",
      "  -3.22800000e+00  -3.22800000e+00  -3.30000000e+00  -2.75400000e+00\n",
      "  -2.07000000e+00  -1.93100000e+00  -2.22400000e+00  -1.52600000e+00]\n"
     ]
    }
   ],
   "source": [
    "print dprice[:20].values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "I = s[dprice_binned >= 3].index #respresents index set of rows from s that yields a positive return"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1      0.376\n",
       "2      0.473\n",
       "10     0.473\n",
       "11     0.570\n",
       "26     1.126\n",
       "129    0.452\n",
       "130    0.332\n",
       "222    0.582\n",
       "279    3.817\n",
       "280    2.879\n",
       "Name: dprice, dtype: float64"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dprice[I][:10]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "good_d_rows = list(set(d_index[I].values)) #get unique row indices from d where an initial sale lead to positive gains"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>tradeid</th>\n",
       "      <th>cusip</th>\n",
       "      <th>amount</th>\n",
       "      <th>price</th>\n",
       "      <th>yield</th>\n",
       "      <th>tradedate</th>\n",
       "      <th>tradetype</th>\n",
       "      <th>name</th>\n",
       "      <th>state</th>\n",
       "      <th>coupon</th>\n",
       "      <th>maturity</th>\n",
       "      <th>issuesize</th>\n",
       "      <th>issuetype</th>\n",
       "      <th>issuesource</th>\n",
       "      <th>bidcount</th>\n",
       "      <th>rtg</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>317.0</td>\n",
       "      <td>45471AKG6</td>\n",
       "      <td>5000.0</td>\n",
       "      <td>99.507</td>\n",
       "      <td>4.037</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Sale_to_Customer</td>\n",
       "      <td>indiana st fin auth-a</td>\n",
       "      <td>Indiana</td>\n",
       "      <td>4.00</td>\n",
       "      <td>6635.0</td>\n",
       "      <td>88930000.0</td>\n",
       "      <td>REVENUE BONDS</td>\n",
       "      <td>HLTH, HOSP, NURSHOME REV.</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>318.0</td>\n",
       "      <td>45471AKG6</td>\n",
       "      <td>30000.0</td>\n",
       "      <td>99.410</td>\n",
       "      <td>4.045</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Sale_to_Customer</td>\n",
       "      <td>indiana st fin auth-a</td>\n",
       "      <td>Indiana</td>\n",
       "      <td>4.00</td>\n",
       "      <td>6635.0</td>\n",
       "      <td>88930000.0</td>\n",
       "      <td>REVENUE BONDS</td>\n",
       "      <td>HLTH, HOSP, NURSHOME REV.</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24578</th>\n",
       "      <td>63754.0</td>\n",
       "      <td>914460LE8</td>\n",
       "      <td>40000.0</td>\n",
       "      <td>95.230</td>\n",
       "      <td>4.090</td>\n",
       "      <td>21.0</td>\n",
       "      <td>Sale_to_Customer</td>\n",
       "      <td>univ of mn-b-txbl</td>\n",
       "      <td>Minnesota</td>\n",
       "      <td>3.75</td>\n",
       "      <td>7642.0</td>\n",
       "      <td>2325000.0</td>\n",
       "      <td>REVENUE BONDS</td>\n",
       "      <td>COLLEGE &amp; UNIV. REV.</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16388</th>\n",
       "      <td>42492.0</td>\n",
       "      <td>64972GFA7</td>\n",
       "      <td>45000.0</td>\n",
       "      <td>113.550</td>\n",
       "      <td>2.946</td>\n",
       "      <td>16.0</td>\n",
       "      <td>Sale_to_Customer</td>\n",
       "      <td>nyc muni wtr-dd-ref</td>\n",
       "      <td>New York</td>\n",
       "      <td>5.00</td>\n",
       "      <td>7046.0</td>\n",
       "      <td>174030000.0</td>\n",
       "      <td>REVENUE BONDS</td>\n",
       "      <td>WATER REVENUE</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>823.0</td>\n",
       "      <td>575579HB6</td>\n",
       "      <td>10000.0</td>\n",
       "      <td>118.093</td>\n",
       "      <td>2.350</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Sale_to_Customer</td>\n",
       "      <td>ma bay trn</td>\n",
       "      <td>Massachusetts</td>\n",
       "      <td>5.00</td>\n",
       "      <td>2679.0</td>\n",
       "      <td>93065000.0</td>\n",
       "      <td>REVENUE BONDS</td>\n",
       "      <td>SALES TAX REVENUE</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>1583.0</td>\n",
       "      <td>57583R5N1</td>\n",
       "      <td>250000.0</td>\n",
       "      <td>112.250</td>\n",
       "      <td>1.639</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Sale_to_Customer</td>\n",
       "      <td>ma dev-b1-harvard unv</td>\n",
       "      <td>Massachusetts</td>\n",
       "      <td>5.00</td>\n",
       "      <td>8629.0</td>\n",
       "      <td>260675000.0</td>\n",
       "      <td>REVENUE BONDS</td>\n",
       "      <td>COLLEGE &amp; UNIV. REV.</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>1584.0</td>\n",
       "      <td>57583R5N1</td>\n",
       "      <td>250000.0</td>\n",
       "      <td>112.250</td>\n",
       "      <td>1.639</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Sale_to_Customer</td>\n",
       "      <td>ma dev-b1-harvard unv</td>\n",
       "      <td>Massachusetts</td>\n",
       "      <td>5.00</td>\n",
       "      <td>8629.0</td>\n",
       "      <td>260675000.0</td>\n",
       "      <td>REVENUE BONDS</td>\n",
       "      <td>COLLEGE &amp; UNIV. REV.</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>1585.0</td>\n",
       "      <td>57583R5N1</td>\n",
       "      <td>200000.0</td>\n",
       "      <td>112.209</td>\n",
       "      <td>1.650</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Sale_to_Customer</td>\n",
       "      <td>ma dev-b1-harvard unv</td>\n",
       "      <td>Massachusetts</td>\n",
       "      <td>5.00</td>\n",
       "      <td>8629.0</td>\n",
       "      <td>260675000.0</td>\n",
       "      <td>REVENUE BONDS</td>\n",
       "      <td>COLLEGE &amp; UNIV. REV.</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>1586.0</td>\n",
       "      <td>57583R5N1</td>\n",
       "      <td>400000.0</td>\n",
       "      <td>112.000</td>\n",
       "      <td>1.703</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Sale_to_Customer</td>\n",
       "      <td>ma dev-b1-harvard unv</td>\n",
       "      <td>Massachusetts</td>\n",
       "      <td>5.00</td>\n",
       "      <td>8629.0</td>\n",
       "      <td>260675000.0</td>\n",
       "      <td>REVENUE BONDS</td>\n",
       "      <td>COLLEGE &amp; UNIV. REV.</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24613</th>\n",
       "      <td>64639.0</td>\n",
       "      <td>927676MX1</td>\n",
       "      <td>15000.0</td>\n",
       "      <td>100.000</td>\n",
       "      <td>4.897</td>\n",
       "      <td>21.0</td>\n",
       "      <td>Sale_to_Customer</td>\n",
       "      <td>vi fin gross rcpts ln</td>\n",
       "      <td>Virgin Islands</td>\n",
       "      <td>5.00</td>\n",
       "      <td>3501.0</td>\n",
       "      <td>19530000.0</td>\n",
       "      <td>REVENUE BONDS</td>\n",
       "      <td>SALES TAX REVENUE</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       tradeid      cusip    amount    price  yield  tradedate  \\\n",
       "0        317.0  45471AKG6    5000.0   99.507  4.037        0.0   \n",
       "1        318.0  45471AKG6   30000.0   99.410  4.045        0.0   \n",
       "24578  63754.0  914460LE8   40000.0   95.230  4.090       21.0   \n",
       "16388  42492.0  64972GFA7   45000.0  113.550  2.946       16.0   \n",
       "14       823.0  575579HB6   10000.0  118.093  2.350        0.0   \n",
       "27      1583.0  57583R5N1  250000.0  112.250  1.639        0.0   \n",
       "28      1584.0  57583R5N1  250000.0  112.250  1.639        0.0   \n",
       "29      1585.0  57583R5N1  200000.0  112.209  1.650        0.0   \n",
       "30      1586.0  57583R5N1  400000.0  112.000  1.703        0.0   \n",
       "24613  64639.0  927676MX1   15000.0  100.000  4.897       21.0   \n",
       "\n",
       "              tradetype                   name           state  coupon  \\\n",
       "0      Sale_to_Customer  indiana st fin auth-a         Indiana    4.00   \n",
       "1      Sale_to_Customer  indiana st fin auth-a         Indiana    4.00   \n",
       "24578  Sale_to_Customer      univ of mn-b-txbl       Minnesota    3.75   \n",
       "16388  Sale_to_Customer    nyc muni wtr-dd-ref        New York    5.00   \n",
       "14     Sale_to_Customer             ma bay trn   Massachusetts    5.00   \n",
       "27     Sale_to_Customer  ma dev-b1-harvard unv   Massachusetts    5.00   \n",
       "28     Sale_to_Customer  ma dev-b1-harvard unv   Massachusetts    5.00   \n",
       "29     Sale_to_Customer  ma dev-b1-harvard unv   Massachusetts    5.00   \n",
       "30     Sale_to_Customer  ma dev-b1-harvard unv   Massachusetts    5.00   \n",
       "24613  Sale_to_Customer  vi fin gross rcpts ln  Virgin Islands    5.00   \n",
       "\n",
       "       maturity    issuesize      issuetype                issuesource  \\\n",
       "0        6635.0   88930000.0  REVENUE BONDS  HLTH, HOSP, NURSHOME REV.   \n",
       "1        6635.0   88930000.0  REVENUE BONDS  HLTH, HOSP, NURSHOME REV.   \n",
       "24578    7642.0    2325000.0  REVENUE BONDS       COLLEGE & UNIV. REV.   \n",
       "16388    7046.0  174030000.0  REVENUE BONDS              WATER REVENUE   \n",
       "14       2679.0   93065000.0  REVENUE BONDS          SALES TAX REVENUE   \n",
       "27       8629.0  260675000.0  REVENUE BONDS       COLLEGE & UNIV. REV.   \n",
       "28       8629.0  260675000.0  REVENUE BONDS       COLLEGE & UNIV. REV.   \n",
       "29       8629.0  260675000.0  REVENUE BONDS       COLLEGE & UNIV. REV.   \n",
       "30       8629.0  260675000.0  REVENUE BONDS       COLLEGE & UNIV. REV.   \n",
       "24613    3501.0   19530000.0  REVENUE BONDS          SALES TAX REVENUE   \n",
       "\n",
       "       bidcount  rtg  \n",
       "0           NaN    6  \n",
       "1           NaN    6  \n",
       "24578       3.0    2  \n",
       "16388       NaN    2  \n",
       "14          NaN    2  \n",
       "27          NaN    1  \n",
       "28          NaN    1  \n",
       "29          NaN    1  \n",
       "30          NaN    1  \n",
       "24613       NaN    4  "
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "good_d = d.loc[good_d_rows, :]\n",
    "good_d[:10]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "buckeye tob asset-a-2    52\n",
       "il st txb-pension        38\n",
       "california st            32\n",
       "ca stwd cmntys-a-ref     29\n",
       "fl hurricane-ser a       24\n",
       "ca txb-var purp          21\n",
       "ca dept wtr resources    20\n",
       "pennsylvania-1st ser     18\n",
       "ca st                    13\n",
       "ca st univ-a             13\n",
       "nj transprtn trust-aa    12\n",
       "liberty dev goldman      12\n",
       "ysleta isd-bldg          12\n",
       "salt verde fnl corp      11\n",
       "ca econ-a-preref-ref     10\n",
       "california st-ref        10\n",
       "tob settlement ser 1a    10\n",
       "ny dorm auth-a            9\n",
       "univ of ct-a              9\n",
       "miami-dade co wtr-ref     9\n",
       "Name: name, dtype: int64"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "good_d.name.value_counts()[:20]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.0    620\n",
       "6.0    568\n",
       "3.0    458\n",
       "5.0    446\n",
       "4.0    404\n",
       "2.0    393\n",
       "Name: holdtime, dtype: int64"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "s.loc[I,:].holdtime.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "I_pred = s[dp_pred >= 3].index #respresents predicted index set of rows from s1 that yields a positive return"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1932    0.568\n",
       "1939    0.375\n",
       "1945    0.913\n",
       "1972    0.684\n",
       "3930   -0.925\n",
       "3931    1.498\n",
       "4044    1.430\n",
       "4045    1.347\n",
       "4055    1.290\n",
       "4056    1.207\n",
       "Name: dprice, dtype: float64"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dprice[I_pred][50:60]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "good_d_rows_pred = list(set(d_index[I_pred].values)) #get unique row indices from d where an initial sale is predicted to lead to positive gains"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>tradeid</th>\n",
       "      <th>cusip</th>\n",
       "      <th>amount</th>\n",
       "      <th>price</th>\n",
       "      <th>yield</th>\n",
       "      <th>tradedate</th>\n",
       "      <th>tradetype</th>\n",
       "      <th>name</th>\n",
       "      <th>state</th>\n",
       "      <th>coupon</th>\n",
       "      <th>maturity</th>\n",
       "      <th>issuesize</th>\n",
       "      <th>issuetype</th>\n",
       "      <th>issuesource</th>\n",
       "      <th>bidcount</th>\n",
       "      <th>rtg</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>317.0</td>\n",
       "      <td>45471AKG6</td>\n",
       "      <td>5000.0</td>\n",
       "      <td>99.507</td>\n",
       "      <td>4.037</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Sale_to_Customer</td>\n",
       "      <td>indiana st fin auth-a</td>\n",
       "      <td>Indiana</td>\n",
       "      <td>4.00</td>\n",
       "      <td>6635.0</td>\n",
       "      <td>88930000.0</td>\n",
       "      <td>REVENUE BONDS</td>\n",
       "      <td>HLTH, HOSP, NURSHOME REV.</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>318.0</td>\n",
       "      <td>45471AKG6</td>\n",
       "      <td>30000.0</td>\n",
       "      <td>99.410</td>\n",
       "      <td>4.045</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Sale_to_Customer</td>\n",
       "      <td>indiana st fin auth-a</td>\n",
       "      <td>Indiana</td>\n",
       "      <td>4.00</td>\n",
       "      <td>6635.0</td>\n",
       "      <td>88930000.0</td>\n",
       "      <td>REVENUE BONDS</td>\n",
       "      <td>HLTH, HOSP, NURSHOME REV.</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6146</th>\n",
       "      <td>57118.0</td>\n",
       "      <td>794665FP9</td>\n",
       "      <td>15000.0</td>\n",
       "      <td>118.924</td>\n",
       "      <td>2.320</td>\n",
       "      <td>6.0</td>\n",
       "      <td>Sale_to_Customer</td>\n",
       "      <td>sales tax asset-a</td>\n",
       "      <td>New York</td>\n",
       "      <td>5.00</td>\n",
       "      <td>3880.0</td>\n",
       "      <td>129315000.0</td>\n",
       "      <td>REVENUE BONDS</td>\n",
       "      <td>SALES TAX REVENUE</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8047</th>\n",
       "      <td>20030.0</td>\n",
       "      <td>011415PT7</td>\n",
       "      <td>15000.0</td>\n",
       "      <td>103.654</td>\n",
       "      <td>1.600</td>\n",
       "      <td>8.0</td>\n",
       "      <td>Sale_to_Customer</td>\n",
       "      <td>alamon clg-ref</td>\n",
       "      <td>Texas</td>\n",
       "      <td>5.00</td>\n",
       "      <td>351.0</td>\n",
       "      <td>4415000.0</td>\n",
       "      <td>GENERAL OBLIGATION LTD</td>\n",
       "      <td>AD VALOREM PROPERTY TAX</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28677</th>\n",
       "      <td>53852.0</td>\n",
       "      <td>74529JAC9</td>\n",
       "      <td>420000.0</td>\n",
       "      <td>67.625</td>\n",
       "      <td>9.771</td>\n",
       "      <td>23.0</td>\n",
       "      <td>Sale_to_Customer</td>\n",
       "      <td>pr s/tax-ser b</td>\n",
       "      <td>Puerto Rico</td>\n",
       "      <td>6.05</td>\n",
       "      <td>7093.0</td>\n",
       "      <td>575000000.0</td>\n",
       "      <td>REVENUE BONDS</td>\n",
       "      <td>SALES TAX REVENUE</td>\n",
       "      <td>NaN</td>\n",
       "      <td>19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7169</th>\n",
       "      <td>27286.0</td>\n",
       "      <td>13066YQC6</td>\n",
       "      <td>50000.0</td>\n",
       "      <td>104.906</td>\n",
       "      <td>1.200</td>\n",
       "      <td>7.0</td>\n",
       "      <td>Sale_to_Customer</td>\n",
       "      <td>ca wtr</td>\n",
       "      <td>California</td>\n",
       "      <td>5.00</td>\n",
       "      <td>426.0</td>\n",
       "      <td>181785000.0</td>\n",
       "      <td>REVENUE BONDS</td>\n",
       "      <td>ELEC. PWR. &amp; LT. REVS.</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16708</th>\n",
       "      <td>33668.0</td>\n",
       "      <td>59334DBV5</td>\n",
       "      <td>25000.0</td>\n",
       "      <td>102.587</td>\n",
       "      <td>1.200</td>\n",
       "      <td>16.0</td>\n",
       "      <td>Sale_to_Customer</td>\n",
       "      <td>miami etc w/s-ref</td>\n",
       "      <td>Florida</td>\n",
       "      <td>5.00</td>\n",
       "      <td>214.0</td>\n",
       "      <td>3215000.0</td>\n",
       "      <td>REVENUE BONDS</td>\n",
       "      <td>WATER REVENUE</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25260</th>\n",
       "      <td>53926.0</td>\n",
       "      <td>74529JAG0</td>\n",
       "      <td>75000.0</td>\n",
       "      <td>23.155</td>\n",
       "      <td>6.060</td>\n",
       "      <td>22.0</td>\n",
       "      <td>Sale_to_Customer</td>\n",
       "      <td>pr s/tax-cabs-a</td>\n",
       "      <td>Puerto Rico</td>\n",
       "      <td>0.00</td>\n",
       "      <td>8919.0</td>\n",
       "      <td>114697901.1</td>\n",
       "      <td>REVENUE BONDS</td>\n",
       "      <td>SALES TAX REVENUE</td>\n",
       "      <td>6.0</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2061</th>\n",
       "      <td>29727.0</td>\n",
       "      <td>429335GZ9</td>\n",
       "      <td>135000.0</td>\n",
       "      <td>116.117</td>\n",
       "      <td>2.880</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Sale_to_Customer</td>\n",
       "      <td>hidalgo drain dt #1</td>\n",
       "      <td>Texas</td>\n",
       "      <td>5.00</td>\n",
       "      <td>3836.0</td>\n",
       "      <td>4725000.0</td>\n",
       "      <td>GENERAL OBLIGATION UNLTD</td>\n",
       "      <td>AD VALOREM PROPERTY TAX</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10161</th>\n",
       "      <td>4398.0</td>\n",
       "      <td>13937KGV2</td>\n",
       "      <td>100000.0</td>\n",
       "      <td>98.248</td>\n",
       "      <td>3.630</td>\n",
       "      <td>10.0</td>\n",
       "      <td>Sale_to_Customer</td>\n",
       "      <td>cape coral-ref</td>\n",
       "      <td>Florida</td>\n",
       "      <td>3.50</td>\n",
       "      <td>6788.0</td>\n",
       "      <td>2720000.0</td>\n",
       "      <td>REVENUE BONDS</td>\n",
       "      <td>APPROPRIATIONS</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       tradeid      cusip    amount    price  yield  tradedate  \\\n",
       "0        317.0  45471AKG6    5000.0   99.507  4.037        0.0   \n",
       "1        318.0  45471AKG6   30000.0   99.410  4.045        0.0   \n",
       "6146   57118.0  794665FP9   15000.0  118.924  2.320        6.0   \n",
       "8047   20030.0  011415PT7   15000.0  103.654  1.600        8.0   \n",
       "28677  53852.0  74529JAC9  420000.0   67.625  9.771       23.0   \n",
       "7169   27286.0  13066YQC6   50000.0  104.906  1.200        7.0   \n",
       "16708  33668.0  59334DBV5   25000.0  102.587  1.200       16.0   \n",
       "25260  53926.0  74529JAG0   75000.0   23.155  6.060       22.0   \n",
       "2061   29727.0  429335GZ9  135000.0  116.117  2.880        1.0   \n",
       "10161   4398.0  13937KGV2  100000.0   98.248  3.630       10.0   \n",
       "\n",
       "              tradetype                   name        state  coupon  maturity  \\\n",
       "0      Sale_to_Customer  indiana st fin auth-a      Indiana    4.00    6635.0   \n",
       "1      Sale_to_Customer  indiana st fin auth-a      Indiana    4.00    6635.0   \n",
       "6146   Sale_to_Customer      sales tax asset-a     New York    5.00    3880.0   \n",
       "8047   Sale_to_Customer         alamon clg-ref        Texas    5.00     351.0   \n",
       "28677  Sale_to_Customer         pr s/tax-ser b  Puerto Rico    6.05    7093.0   \n",
       "7169   Sale_to_Customer                 ca wtr   California    5.00     426.0   \n",
       "16708  Sale_to_Customer      miami etc w/s-ref      Florida    5.00     214.0   \n",
       "25260  Sale_to_Customer        pr s/tax-cabs-a  Puerto Rico    0.00    8919.0   \n",
       "2061   Sale_to_Customer    hidalgo drain dt #1        Texas    5.00    3836.0   \n",
       "10161  Sale_to_Customer         cape coral-ref      Florida    3.50    6788.0   \n",
       "\n",
       "         issuesize                 issuetype                issuesource  \\\n",
       "0       88930000.0             REVENUE BONDS  HLTH, HOSP, NURSHOME REV.   \n",
       "1       88930000.0             REVENUE BONDS  HLTH, HOSP, NURSHOME REV.   \n",
       "6146   129315000.0             REVENUE BONDS          SALES TAX REVENUE   \n",
       "8047     4415000.0    GENERAL OBLIGATION LTD    AD VALOREM PROPERTY TAX   \n",
       "28677  575000000.0             REVENUE BONDS          SALES TAX REVENUE   \n",
       "7169   181785000.0             REVENUE BONDS     ELEC. PWR. & LT. REVS.   \n",
       "16708    3215000.0             REVENUE BONDS              WATER REVENUE   \n",
       "25260  114697901.1             REVENUE BONDS          SALES TAX REVENUE   \n",
       "2061     4725000.0  GENERAL OBLIGATION UNLTD    AD VALOREM PROPERTY TAX   \n",
       "10161    2720000.0             REVENUE BONDS             APPROPRIATIONS   \n",
       "\n",
       "       bidcount  rtg  \n",
       "0           NaN    6  \n",
       "1           NaN    6  \n",
       "6146        NaN    1  \n",
       "8047        NaN    1  \n",
       "28677       NaN   19  \n",
       "7169        NaN    3  \n",
       "16708       NaN    4  \n",
       "25260       6.0    4  \n",
       "2061        NaN    3  \n",
       "10161       NaN    4  "
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "good_d_pred = d.loc[good_d_rows_pred, :]\n",
    "good_d_pred[:10]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "REVENUE BONDS                23546\n",
       "GENERAL OBLIGATION UNLTD      9066\n",
       "GENERAL OBLIGATION LTD        1604\n",
       "CERTIFICATE PARTICIPATION      798\n",
       "TAX ALLOCATION                 306\n",
       "SPECIAL ASSESSMENT              85\n",
       "SPECIAL TAX                     51\n",
       "Name: issuetype, dtype: int64"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "d.issuetype.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "REVENUE BONDS                763\n",
       "GENERAL OBLIGATION UNLTD     288\n",
       "GENERAL OBLIGATION LTD        51\n",
       "CERTIFICATE PARTICIPATION     25\n",
       "TAX ALLOCATION                 5\n",
       "SPECIAL TAX                    2\n",
       "SPECIAL ASSESSMENT             1\n",
       "Name: issuetype, dtype: int64"
      ]
     },
     "execution_count": 70,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "good_d_pred.issuetype.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "REVENUE BONDS                1120\n",
       "GENERAL OBLIGATION UNLTD      376\n",
       "GENERAL OBLIGATION LTD         56\n",
       "CERTIFICATE PARTICIPATION      32\n",
       "TAX ALLOCATION                  5\n",
       "SPECIAL TAX                     2\n",
       "SPECIAL ASSESSMENT              1\n",
       "Name: issuetype, dtype: int64"
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "good_d.issuetype.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "California           5973\n",
       "New York             4172\n",
       "Texas                2445\n",
       "Illinois             2300\n",
       "New Jersey           2186\n",
       "Florida              2129\n",
       "Pennsylvania         1722\n",
       "Puerto Rico          1698\n",
       "Massachusetts        1377\n",
       "Ohio                 1076\n",
       "Washington            824\n",
       "Arizona               571\n",
       "South Carolina        569\n",
       "Connecticut           561\n",
       "Michigan              535\n",
       "Maryland              483\n",
       "Louisiana             435\n",
       "Georgia               406\n",
       "Virginia              398\n",
       "Dist. of Columbia     361\n",
       "Name: state, dtype: int64"
      ]
     },
     "execution_count": 73,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "d.state.value_counts()[:20]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "California        340\n",
       "New York          206\n",
       "Florida           108\n",
       "Texas              99\n",
       "Pennsylvania       85\n",
       "Illinois           83\n",
       "Ohio               78\n",
       "Massachusetts      76\n",
       "New Jersey         72\n",
       "Puerto Rico        49\n",
       "Arizona            39\n",
       "Washington         36\n",
       "Connecticut        28\n",
       "Maryland           26\n",
       "South Carolina     22\n",
       "Michigan           19\n",
       "Georgia            17\n",
       "Wisconsin          14\n",
       "Virgin Islands     14\n",
       "Kentucky           14\n",
       "Name: state, dtype: int64"
      ]
     },
     "execution_count": 74,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "good_d.state.value_counts()[:20]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}