DataCleaning.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Data Cleaning"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Populating the interactive namespace from numpy and matplotlib\n"
     ]
    }
   ],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "from pandas import DataFrame, merge\n",
    "\n",
    "%pylab inline\n",
    "np.random.seed(127)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**You can go straight [here](#Impute-Missing-Value) for imputation**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 11456 entries, 0 to 11455\n",
      "Data columns (total 8 columns):\n",
      "ID                int64\n",
      "Name              object\n",
      "DateTime          object\n",
      "AnimalType        object\n",
      "SexuponOutcome    object\n",
      "AgeuponOutcome    object\n",
      "Breed             object\n",
      "Color             object\n",
      "dtypes: int64(1), object(7)\n",
      "memory usage: 716.1+ KB\n"
     ]
    }
   ],
   "source": [
    "animals = pd.read_csv(\"test.csv\", compression='gzip')\n",
    "animals.info(verbose=True, null_counts=False)"
   ]
  },
  {
   "cell_type": "raw",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "# categorical feature \"type\"\n",
    "typedict = {\"Dog\": 0, \"Cat\":1}\n",
    "animals[\"type\"] = animals[\"AnimalType\"].map(typedict)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Clean-up NULL values"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First, we want to know the columns that has `np.nan` values and the number of `np.nan`s"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Feature \"Name\" has 3225 missing values\n",
      "Feature \"AgeuponOutcome\" has 6 missing values\n"
     ]
    }
   ],
   "source": [
    "for att in animals.columns:\n",
    "    nums = animals[att].isnull().value_counts().as_matrix()\n",
    "    if (len(nums) != 1):\n",
    "        print \"Feature \\\"{}\\\" has {} missing values\".format(att, nums[-1])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Name NULL values\n",
    "\n",
    "Name might not seem like the first feature that matters when it comes to the outcome of an animal. However, having a name or not might affect the outcome of an animal, let's try it out."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true,
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "# we map nan and 'X' as '0' meaning no name, everything else has name\n",
    "animals['HasName'] = animals.Name.map(lambda x : 0 if x == 'X' else 1, na_action='ignore')\n",
    "animals['HasName'].fillna(value=0, inplace=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### DateTime Features\n",
    "We would like to add additional features indicating timeofday, weekofyear, etc for each DateTime value"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Our records are between the range 2013-10-01 10:44:00 to 2016-02-21 18:37:00\n"
     ]
    }
   ],
   "source": [
    "# we already know that \"DateTime\" column does not have null values\n",
    "animals['Timestamps'] = pd.to_datetime(animals['DateTime'])\n",
    "\n",
    "print \"Our records are between the range {} to {}\".format(animals.Timestamps.min(), animals.Timestamps.max())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# we convert timestamps to several potential groups\n",
    "animals['Year'] = animals.Timestamps.dt.year\n",
    "animals['Month'] = animals.Timestamps.dt.month\n",
    "animals['Weekday'] = animals.Timestamps.dt.weekday\n",
    "animals['Hour'] = animals.Timestamps.dt.hour\n",
    "animals['WeekofYear'] = animals.Timestamps.dt.weekofyear\n",
    "animals['DayofMonth'] = animals.Timestamps.dt.day"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "%load_ext autoreload\n",
    "%autoreload 2"
   ]
  },
  {
   "cell_type": "raw",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "import holidays\n",
    "\n",
    "from dateutil.relativedelta import relativedelta\n",
    "from datetime import date\n",
    "class MyHolidays(holidays.UnitedStates):\n",
    "    def _populate(self, year):\n",
    "        # Populate the holiday list with the default US holidays\n",
    "        holidays.UnitedStates._populate(self, year)\n",
    "        # Remove Columbus Day\n",
    "        # self.pop(date(year, 10, 1) + relativedelta(weekday=MO(+2)), None)\n",
    "        # Add Valentines Day\n",
    "        self[date(year, 2, 14)] = \"Valentines Day\"\n",
    "        self[date(year, 1, 1) + relativedelta(weekday=MO(+3))] = \"Martin Luther King Day\"\n",
    "        self[date(year, 2, 1) + relativedelta(weekday=MO(+3))] = \"President's Day\"\n",
    "        self[date(year, 5, 1) + relativedelta(weekday=SU(+2))] = \"Mother's Day\"\n",
    "        self[date(year, 6, 1) + relativedelta(weekday=SU(+3))] = \"Father's Day\"\n",
    "        self[date(year, 10, 31)] = \"Halloween\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Aside from all above features, we also want to see the influnce of public holidays and holiday seasons.\n",
    "# so we customized our own class \"MyHolidays\" based on Texas public holidays\n",
    "import MyHolidays\n",
    "tx_holidays = MyHolidays.MyHolidays(state=\"TX\", years =[2015])\n",
    "def is_holiday(dt):\n",
    "    if (dt.month in [1,12,7,8]) or (dt.date() in tx_holidays) or dt.weekday() in [5,6]:\n",
    "        return True\n",
    "    return False\n",
    "\n",
    "animals[\"isHoliday\"] = animals.Timestamps.apply(is_holiday)\n",
    "\n",
    "animals[\"isHoliday\"] = animals.Timestamps.apply(is_holiday)\n",
    "# plot_feature(\"isHoliday\", data=animals)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### AgeuponOutcome \n",
    "First we need to use the same unit on AgeuponOutcome feature. We decide to convert all of them into days. Note that some animals has an age of \"0 years\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The interesting finding is that all animals are missing age info happened in a very small time window."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "units = {\"year\" : 365, \"month\":31, \"day\":1, \"week\":7}\n",
    "\n",
    "def get_age_in_days(agestring):\n",
    "    tokens = agestring.split(\" \")\n",
    "    unit = tokens[1].replace(\"s\",\"\")\n",
    "    days = int(tokens[0]) * units[unit]\n",
    "    # return days\n",
    "    return days if days != 0 else -1\n",
    "\n",
    "animals[\"AgeinDays\"] = animals.AgeuponOutcome.map(get_age_in_days, na_action=\"ignore\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2016-02-10 18:49:00 to 2016-02-19 18:31:00\n"
     ]
    }
   ],
   "source": [
    "print \"{} to {}\".format(animals[animals.AgeinDays.isnull()].Timestamps.min(), \n",
    "                        animals[animals.AgeinDays.isnull()].Timestamps.max())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>ID</th>\n",
       "      <th>Name</th>\n",
       "      <th>DateTime</th>\n",
       "      <th>AnimalType</th>\n",
       "      <th>SexuponOutcome</th>\n",
       "      <th>AgeuponOutcome</th>\n",
       "      <th>Breed</th>\n",
       "      <th>Color</th>\n",
       "      <th>type</th>\n",
       "      <th>HasName</th>\n",
       "      <th>Timestamps</th>\n",
       "      <th>Year</th>\n",
       "      <th>Month</th>\n",
       "      <th>Weekday</th>\n",
       "      <th>Hour</th>\n",
       "      <th>WeekofYear</th>\n",
       "      <th>DayofMonth</th>\n",
       "      <th>isHoliday</th>\n",
       "      <th>AgeinDays</th>\n",
       "      <th>AgeInDaysCategorical</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>763</th>\n",
       "      <td>764</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2016-02-18 19:07:00</td>\n",
       "      <td>Cat</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Domestic Shorthair Mix</td>\n",
       "      <td>Black</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2016-02-18 19:07:00</td>\n",
       "      <td>2016</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>19</td>\n",
       "      <td>7</td>\n",
       "      <td>18</td>\n",
       "      <td>False</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1422</th>\n",
       "      <td>1423</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2014-04-25 15:17:00</td>\n",
       "      <td>Cat</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>0 years</td>\n",
       "      <td>Domestic Shorthair Mix</td>\n",
       "      <td>Blue</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2014-04-25 15:17:00</td>\n",
       "      <td>2014</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>15</td>\n",
       "      <td>17</td>\n",
       "      <td>25</td>\n",
       "      <td>False</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2133</th>\n",
       "      <td>2134</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2014-05-13 13:19:00</td>\n",
       "      <td>Cat</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>0 years</td>\n",
       "      <td>Domestic Shorthair Mix</td>\n",
       "      <td>Brown Tabby</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2014-05-13 13:19:00</td>\n",
       "      <td>2014</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>13</td>\n",
       "      <td>20</td>\n",
       "      <td>13</td>\n",
       "      <td>False</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2498</th>\n",
       "      <td>2499</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2016-02-16 18:54:00</td>\n",
       "      <td>Cat</td>\n",
       "      <td>Intact Male</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Domestic Longhair Mix</td>\n",
       "      <td>Blue/White</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2016-02-16 18:54:00</td>\n",
       "      <td>2016</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>18</td>\n",
       "      <td>7</td>\n",
       "      <td>16</td>\n",
       "      <td>False</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2890</th>\n",
       "      <td>2891</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2013-10-16 11:52:00</td>\n",
       "      <td>Cat</td>\n",
       "      <td>Intact Male</td>\n",
       "      <td>0 years</td>\n",
       "      <td>Domestic Shorthair Mix</td>\n",
       "      <td>Cream</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2013-10-16 11:52:00</td>\n",
       "      <td>2013</td>\n",
       "      <td>10</td>\n",
       "      <td>2</td>\n",
       "      <td>11</td>\n",
       "      <td>42</td>\n",
       "      <td>16</td>\n",
       "      <td>False</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4807</th>\n",
       "      <td>4808</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2014-04-25 15:18:00</td>\n",
       "      <td>Cat</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>0 years</td>\n",
       "      <td>Domestic Shorthair Mix</td>\n",
       "      <td>Blue</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2014-04-25 15:18:00</td>\n",
       "      <td>2014</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>15</td>\n",
       "      <td>17</td>\n",
       "      <td>25</td>\n",
       "      <td>False</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5059</th>\n",
       "      <td>5060</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2015-07-04 09:57:00</td>\n",
       "      <td>Dog</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>0 years</td>\n",
       "      <td>Chihuahua Shorthair Mix</td>\n",
       "      <td>Black</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2015-07-04 09:57:00</td>\n",
       "      <td>2015</td>\n",
       "      <td>7</td>\n",
       "      <td>5</td>\n",
       "      <td>9</td>\n",
       "      <td>27</td>\n",
       "      <td>4</td>\n",
       "      <td>False</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6092</th>\n",
       "      <td>6093</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2015-07-05 14:46:00</td>\n",
       "      <td>Cat</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>0 years</td>\n",
       "      <td>Domestic Shorthair Mix</td>\n",
       "      <td>Tortie</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2015-07-05 14:46:00</td>\n",
       "      <td>2015</td>\n",
       "      <td>7</td>\n",
       "      <td>6</td>\n",
       "      <td>14</td>\n",
       "      <td>27</td>\n",
       "      <td>5</td>\n",
       "      <td>False</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6323</th>\n",
       "      <td>6324</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2016-02-10 18:49:00</td>\n",
       "      <td>Cat</td>\n",
       "      <td>Intact Male</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Domestic Shorthair Mix</td>\n",
       "      <td>Black</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2016-02-10 18:49:00</td>\n",
       "      <td>2016</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>18</td>\n",
       "      <td>6</td>\n",
       "      <td>10</td>\n",
       "      <td>False</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6347</th>\n",
       "      <td>6348</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2016-02-12 18:27:00</td>\n",
       "      <td>Cat</td>\n",
       "      <td>Intact Male</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Domestic Shorthair Mix</td>\n",
       "      <td>Brown Tabby</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2016-02-12 18:27:00</td>\n",
       "      <td>2016</td>\n",
       "      <td>2</td>\n",
       "      <td>4</td>\n",
       "      <td>18</td>\n",
       "      <td>6</td>\n",
       "      <td>12</td>\n",
       "      <td>False</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6412</th>\n",
       "      <td>6413</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2013-11-04 15:00:00</td>\n",
       "      <td>Dog</td>\n",
       "      <td>Intact Female</td>\n",
       "      <td>0 years</td>\n",
       "      <td>Beagle/Chihuahua Shorthair</td>\n",
       "      <td>White</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2013-11-04 15:00:00</td>\n",
       "      <td>2013</td>\n",
       "      <td>11</td>\n",
       "      <td>0</td>\n",
       "      <td>15</td>\n",
       "      <td>45</td>\n",
       "      <td>4</td>\n",
       "      <td>False</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6671</th>\n",
       "      <td>6672</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2013-11-04 14:58:00</td>\n",
       "      <td>Dog</td>\n",
       "      <td>Intact Female</td>\n",
       "      <td>0 years</td>\n",
       "      <td>Beagle/Chihuahua Shorthair</td>\n",
       "      <td>White</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2013-11-04 14:58:00</td>\n",
       "      <td>2013</td>\n",
       "      <td>11</td>\n",
       "      <td>0</td>\n",
       "      <td>14</td>\n",
       "      <td>45</td>\n",
       "      <td>4</td>\n",
       "      <td>False</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7626</th>\n",
       "      <td>7627</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2016-02-19 18:31:00</td>\n",
       "      <td>Cat</td>\n",
       "      <td>Intact Female</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Domestic Shorthair Mix</td>\n",
       "      <td>Tortie</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2016-02-19 18:31:00</td>\n",
       "      <td>2016</td>\n",
       "      <td>2</td>\n",
       "      <td>4</td>\n",
       "      <td>18</td>\n",
       "      <td>7</td>\n",
       "      <td>19</td>\n",
       "      <td>False</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7745</th>\n",
       "      <td>7746</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2014-08-28 18:20:00</td>\n",
       "      <td>Cat</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>0 years</td>\n",
       "      <td>Domestic Shorthair Mix</td>\n",
       "      <td>White/Brown Tabby</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2014-08-28 18:20:00</td>\n",
       "      <td>2014</td>\n",
       "      <td>8</td>\n",
       "      <td>3</td>\n",
       "      <td>18</td>\n",
       "      <td>35</td>\n",
       "      <td>28</td>\n",
       "      <td>False</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9516</th>\n",
       "      <td>9517</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2014-08-28 18:20:00</td>\n",
       "      <td>Cat</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>0 years</td>\n",
       "      <td>Domestic Shorthair Mix</td>\n",
       "      <td>Blue Tabby/White</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2014-08-28 18:20:00</td>\n",
       "      <td>2014</td>\n",
       "      <td>8</td>\n",
       "      <td>3</td>\n",
       "      <td>18</td>\n",
       "      <td>35</td>\n",
       "      <td>28</td>\n",
       "      <td>False</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10702</th>\n",
       "      <td>10703</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2015-08-16 14:15:00</td>\n",
       "      <td>Cat</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>0 years</td>\n",
       "      <td>Domestic Shorthair Mix</td>\n",
       "      <td>Pink</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2015-08-16 14:15:00</td>\n",
       "      <td>2015</td>\n",
       "      <td>8</td>\n",
       "      <td>6</td>\n",
       "      <td>14</td>\n",
       "      <td>33</td>\n",
       "      <td>16</td>\n",
       "      <td>False</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10754</th>\n",
       "      <td>10755</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2016-02-12 18:33:00</td>\n",
       "      <td>Cat</td>\n",
       "      <td>Neutered Male</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Domestic Shorthair Mix</td>\n",
       "      <td>White/Black</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2016-02-12 18:33:00</td>\n",
       "      <td>2016</td>\n",
       "      <td>2</td>\n",
       "      <td>4</td>\n",
       "      <td>18</td>\n",
       "      <td>6</td>\n",
       "      <td>12</td>\n",
       "      <td>False</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10878</th>\n",
       "      <td>10879</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2013-10-16 11:51:00</td>\n",
       "      <td>Cat</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>0 years</td>\n",
       "      <td>Domestic Shorthair Mix</td>\n",
       "      <td>Gray</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2013-10-16 11:51:00</td>\n",
       "      <td>2013</td>\n",
       "      <td>10</td>\n",
       "      <td>2</td>\n",
       "      <td>11</td>\n",
       "      <td>42</td>\n",
       "      <td>16</td>\n",
       "      <td>False</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11096</th>\n",
       "      <td>11097</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2014-04-29 12:35:00</td>\n",
       "      <td>Cat</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>0 years</td>\n",
       "      <td>Domestic Shorthair Mix</td>\n",
       "      <td>Black</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2014-04-29 12:35:00</td>\n",
       "      <td>2014</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>12</td>\n",
       "      <td>18</td>\n",
       "      <td>29</td>\n",
       "      <td>False</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          ID Name             DateTime AnimalType SexuponOutcome  \\\n",
       "763      764  NaN  2016-02-18 19:07:00        Cat        Unknown   \n",
       "1422    1423  NaN  2014-04-25 15:17:00        Cat        Unknown   \n",
       "2133    2134  NaN  2014-05-13 13:19:00        Cat        Unknown   \n",
       "2498    2499  NaN  2016-02-16 18:54:00        Cat    Intact Male   \n",
       "2890    2891  NaN  2013-10-16 11:52:00        Cat    Intact Male   \n",
       "4807    4808  NaN  2014-04-25 15:18:00        Cat        Unknown   \n",
       "5059    5060  NaN  2015-07-04 09:57:00        Dog        Unknown   \n",
       "6092    6093  NaN  2015-07-05 14:46:00        Cat        Unknown   \n",
       "6323    6324  NaN  2016-02-10 18:49:00        Cat    Intact Male   \n",
       "6347    6348  NaN  2016-02-12 18:27:00        Cat    Intact Male   \n",
       "6412    6413  NaN  2013-11-04 15:00:00        Dog  Intact Female   \n",
       "6671    6672  NaN  2013-11-04 14:58:00        Dog  Intact Female   \n",
       "7626    7627  NaN  2016-02-19 18:31:00        Cat  Intact Female   \n",
       "7745    7746  NaN  2014-08-28 18:20:00        Cat        Unknown   \n",
       "9516    9517  NaN  2014-08-28 18:20:00        Cat        Unknown   \n",
       "10702  10703  NaN  2015-08-16 14:15:00        Cat        Unknown   \n",
       "10754  10755  NaN  2016-02-12 18:33:00        Cat  Neutered Male   \n",
       "10878  10879  NaN  2013-10-16 11:51:00        Cat        Unknown   \n",
       "11096  11097  NaN  2014-04-29 12:35:00        Cat        Unknown   \n",
       "\n",
       "      AgeuponOutcome                       Breed              Color  type  \\\n",
       "763              NaN      Domestic Shorthair Mix              Black     1   \n",
       "1422         0 years      Domestic Shorthair Mix               Blue     1   \n",
       "2133         0 years      Domestic Shorthair Mix        Brown Tabby     1   \n",
       "2498             NaN       Domestic Longhair Mix         Blue/White     1   \n",
       "2890         0 years      Domestic Shorthair Mix              Cream     1   \n",
       "4807         0 years      Domestic Shorthair Mix               Blue     1   \n",
       "5059         0 years     Chihuahua Shorthair Mix              Black     0   \n",
       "6092         0 years      Domestic Shorthair Mix             Tortie     1   \n",
       "6323             NaN      Domestic Shorthair Mix              Black     1   \n",
       "6347             NaN      Domestic Shorthair Mix        Brown Tabby     1   \n",
       "6412         0 years  Beagle/Chihuahua Shorthair              White     0   \n",
       "6671         0 years  Beagle/Chihuahua Shorthair              White     0   \n",
       "7626             NaN      Domestic Shorthair Mix             Tortie     1   \n",
       "7745         0 years      Domestic Shorthair Mix  White/Brown Tabby     1   \n",
       "9516         0 years      Domestic Shorthair Mix   Blue Tabby/White     1   \n",
       "10702        0 years      Domestic Shorthair Mix               Pink     1   \n",
       "10754            NaN      Domestic Shorthair Mix        White/Black     1   \n",
       "10878        0 years      Domestic Shorthair Mix               Gray     1   \n",
       "11096        0 years      Domestic Shorthair Mix              Black     1   \n",
       "\n",
       "       HasName          Timestamps  Year  Month  Weekday  Hour  WeekofYear  \\\n",
       "763        0.0 2016-02-18 19:07:00  2016      2        3    19           7   \n",
       "1422       0.0 2014-04-25 15:17:00  2014      4        4    15          17   \n",
       "2133       0.0 2014-05-13 13:19:00  2014      5        1    13          20   \n",
       "2498       0.0 2016-02-16 18:54:00  2016      2        1    18           7   \n",
       "2890       0.0 2013-10-16 11:52:00  2013     10        2    11          42   \n",
       "4807       0.0 2014-04-25 15:18:00  2014      4        4    15          17   \n",
       "5059       0.0 2015-07-04 09:57:00  2015      7        5     9          27   \n",
       "6092       0.0 2015-07-05 14:46:00  2015      7        6    14          27   \n",
       "6323       0.0 2016-02-10 18:49:00  2016      2        2    18           6   \n",
       "6347       0.0 2016-02-12 18:27:00  2016      2        4    18           6   \n",
       "6412       0.0 2013-11-04 15:00:00  2013     11        0    15          45   \n",
       "6671       0.0 2013-11-04 14:58:00  2013     11        0    14          45   \n",
       "7626       0.0 2016-02-19 18:31:00  2016      2        4    18           7   \n",
       "7745       0.0 2014-08-28 18:20:00  2014      8        3    18          35   \n",
       "9516       0.0 2014-08-28 18:20:00  2014      8        3    18          35   \n",
       "10702      0.0 2015-08-16 14:15:00  2015      8        6    14          33   \n",
       "10754      0.0 2016-02-12 18:33:00  2016      2        4    18           6   \n",
       "10878      0.0 2013-10-16 11:51:00  2013     10        2    11          42   \n",
       "11096      0.0 2014-04-29 12:35:00  2014      4        1    12          18   \n",
       "\n",
       "       DayofMonth  isHoliday  AgeinDays AgeInDaysCategorical  \n",
       "763            18      False        NaN                  NaN  \n",
       "1422           25      False       -1.0                 -1.0  \n",
       "2133           13      False       -1.0                 -1.0  \n",
       "2498           16      False        NaN                  NaN  \n",
       "2890           16      False       -1.0                 -1.0  \n",
       "4807           25      False       -1.0                 -1.0  \n",
       "5059            4      False       -1.0                 -1.0  \n",
       "6092            5      False       -1.0                 -1.0  \n",
       "6323           10      False        NaN                  NaN  \n",
       "6347           12      False        NaN                  NaN  \n",
       "6412            4      False       -1.0                 -1.0  \n",
       "6671            4      False       -1.0                 -1.0  \n",
       "7626           19      False        NaN                  NaN  \n",
       "7745           28      False       -1.0                 -1.0  \n",
       "9516           28      False       -1.0                 -1.0  \n",
       "10702          16      False       -1.0                 -1.0  \n",
       "10754          12      False        NaN                  NaN  \n",
       "10878          16      False       -1.0                 -1.0  \n",
       "11096          29      False       -1.0                 -1.0  "
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "animals[(animals.AgeinDays.isnull()) | (animals.AgeinDays==-1)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "animals.AgeinDays.fillna(value=-1, inplace=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We have more young animals than old. The probability distribution is greatly affected by the records (very discrete). We decide to also map ages to infant, child, young adult, adult and old groups"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "bins = [-5, 0, 5 * units[\"month\"], 12 * units[\"month\"], 3*units[\"year\"],\n",
    "        6 * units[\"year\"], 10*units[\"year\"], 20 * units[\"year\"]]\n",
    "labels = [\"Unknown\", \"Infant\", \"Puppy/Kitten\", \"Young Adult\", \"Adult\", \"Senior\", \"Geriatric\"]\n",
    "\n",
    "animals[\"LifeStage\"] = pd.Categorical(pd.cut(animals.AgeinDays, bins=bins, labels=labels), categories=labels)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### SexuponOutcome NULL and Unknown values"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As was shown before, \"Unknown\" sex has a disticntive distribution of outcome compared with other sex, we decide to\n",
    "leave it as is. Or, we could try [different encoding schemes](http://contrib.scikit-learn.org/categorical-encoding/) and see which one works the best on our project.\n",
    "\n",
    "So I'm planning to impute all these randomly with the top two sex within age range (0 days - 2 month) Or I should scale it idk."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "AnimalType Mode:  Cat\n",
      "AgeinDays Median:  21.0\n",
      "HasName Mode 0.0\n"
     ]
    }
   ],
   "source": [
    "unknowns = animals[animals.SexuponOutcome == \"Unknown\"]\n",
    "\n",
    "# so we try to see what their most shared features here\n",
    "print \"AnimalType Mode: \", unknowns.AnimalType.agg('mode')[0]\n",
    "print \"AgeinDays Median: \", unknowns.AgeinDays.median()\n",
    "print \"HasName Mode\", unknowns.HasName.mode()[0]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "These animals are mostly cats, 21 days old, no name and ended up being transfered. We pull out all records that fits this criteria:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index([u'ID', u'Name', u'DateTime', u'AnimalType', u'SexuponOutcome',\n",
       "       u'AgeuponOutcome', u'Breed', u'Color', u'HasName', u'Timestamps',\n",
       "       u'Year', u'Month', u'Weekday', u'Hour', u'WeekofYear', u'DayofMonth',\n",
       "       u'isHoliday', u'AgeinDays', u'LifeStage', u'BernoulliSex',\n",
       "       u'AvgWeight(pounds)', u'SizeScore(1to5)', u'Intelligent',\n",
       "       u'Friendliness', u'MixType', u'Hair', u'ReducedColor', u'BlendedColor',\n",
       "       u'ReducedPattern'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "animals.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total number of:  455\n",
      "Intact Male      0.506536\n",
      "Intact Female    0.472222\n",
      "Neutered Male    0.011438\n",
      "Spayed Female    0.009804\n",
      "Name: SexuponOutcome, dtype: float64\n"
     ]
    }
   ],
   "source": [
    "print \"Total number of: \", len(unknowns)\n",
    "p = animals[(animals.AnimalType==\"Cat\") & \n",
    "        (animals.AgeinDays<= 31) & \n",
    "        (animals.HasName==0) & \n",
    "        (animals.SexuponOutcome!=\"Unknown\" )].SexuponOutcome.value_counts(normalize=True)\n",
    "print p"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "So they are 53 percent intact male and 46 percent intact female. We either can use bernoulli/multinomial (if we really care) when assigning sex on these animals or leave as is."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from scipy.stats import bernoulli\n",
    "im_p = p[\"Intact Male\"] "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def unknown_sex(sexstring):\n",
    "    if sexstring == \"Unknown\":\n",
    "        sex = bernoulli.rvs(im_p)\n",
    "        if sex == 1:\n",
    "            return \"Intact Male\"\n",
    "        else:\n",
    "            return \"Intact Female\"\n",
    "    else:\n",
    "        return sexstring"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "After impute distribution: \n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "Intact Male      0.518681\n",
       "Intact Female    0.481319\n",
       "Name: BernoulliSex, dtype: float64"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "animals[\"BernoulliSex\"] = animals[\"SexuponOutcome\"].map(unknown_sex, na_action=\"ignore\")\n",
    "print \"After impute distribution: \"\n",
    "animals.loc[animals.SexuponOutcome==\"Unknown\", \"BernoulliSex\"].value_counts(normalize=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "While we are imputing ourselves and scikit-learn has the most basic imputer, we can also try [Orange](https://docs.orange.biolab.si/2/reference/rst/Orange.feature.imputation.html) as well. We tried but it didn't work."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Getting Breed information\n",
    "\n",
    "We need more information about \"Breed\". So we got some information from online sources [here](http://www.sorteaze.com/dog-breed-personality--social-traits.html) and [here](http://www.sorteaze.com/dog-breeds.html). Our script `load_dog_data()` is capable of scraping info from these sites from scratch. Unfortunately, this wevsite is now giving 404 error. So we are loading our backup copy.\n",
    "\n",
    "We think dog size and their characteristics are key features when people look for dogs. However, considering lots of shelter animals were abused, their characteristic may change. Not sure if dog personalities would be important in this case, but we are keeping it for now."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>BreedName</th>\n",
       "      <th>AvgWeight(pounds)</th>\n",
       "      <th>SizeScore(1to5)</th>\n",
       "      <th>Intelligent</th>\n",
       "      <th>Friendliness</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Affenpinscher</td>\n",
       "      <td>10</td>\n",
       "      <td>1</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Afghan Hound</td>\n",
       "      <td>55</td>\n",
       "      <td>4</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Airedale Terrier</td>\n",
       "      <td>53</td>\n",
       "      <td>3</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Akita</td>\n",
       "      <td>105</td>\n",
       "      <td>5</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Alaskan Klee Kai</td>\n",
       "      <td>15</td>\n",
       "      <td>2</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          BreedName  AvgWeight(pounds)  SizeScore(1to5)  Intelligent  \\\n",
       "0     Affenpinscher                 10                1          3.0   \n",
       "1      Afghan Hound                 55                4          2.0   \n",
       "2  Airedale Terrier                 53                3          4.0   \n",
       "3             Akita                105                5          2.0   \n",
       "4  Alaskan Klee Kai                 15                2          3.0   \n",
       "\n",
       "   Friendliness  \n",
       "0           3.6  \n",
       "1           2.8  \n",
       "2           3.0  \n",
       "3           2.0  \n",
       "4           3.2  "
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from customize_functions import load_dog_data, load_clean_dog_data\n",
    "\n",
    "# dog_data = load_dog_data()\n",
    "dog_data_clean = load_clean_dog_data()\n",
    "dog_data_clean.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/femianjc/anaconda2/envs/femiani/lib/python2.7/site-packages/numpy/core/_methods.py:59: RuntimeWarning: Mean of empty slice.\n",
      "  warnings.warn(\"Mean of empty slice.\", RuntimeWarning)\n"
     ]
    }
   ],
   "source": [
    "# since our record often give different names to dogs than what we downloaded online, our algorithm of matching \n",
    "# breed name search for tokens that narrows down relatable breeds instead of matching the entire breed string.\n",
    "import re\n",
    "import itertools\n",
    "from functools import wraps\n",
    "import sys\n",
    "\n",
    "def memoize(func):\n",
    "    cache = {}\n",
    "    @wraps(func)\n",
    "    def wrap(*args):\n",
    "        if args not in cache:\n",
    "            cache[args] = func(*args)\n",
    "        return cache[args]\n",
    "    return wrap\n",
    "\n",
    "@memoize\n",
    "def try_permutations(tokens):    \n",
    "    min_count = sys.maxint\n",
    "    best_string = None\n",
    "    for i in range(len(tokens)):\n",
    "        candidate = itertools.permutations(tokens, i+1)\n",
    "        for c in candidate:\n",
    "            string = ' '.join(x for x in c)            \n",
    "            count = len(dog_data_clean[dog_data_clean.BreedName.str.contains(\n",
    "                string, case=False)])\n",
    "            if count == 1:\n",
    "                return string\n",
    "            if count < min_count and count != 0:\n",
    "                min_count, best_string = count, string\n",
    "    return best_string\n",
    "            \n",
    "\n",
    "def add_dog_data(x):\n",
    "    name = x.Breed.replace(\" Mix\", \"\")\n",
    "    breeds= name.split(\"/\")\n",
    "    \n",
    "    matches = pd.DataFrame()\n",
    "    for br in breeds:\n",
    "        best_string = try_permutations(frozenset(br.split(\" \")))\n",
    "        if best_string:\n",
    "            matches = pd.concat(\n",
    "                [matches, \n",
    "                 dog_data_clean[dog_data_clean.BreedName.str.contains(\n",
    "                     best_string, case=False)]\n",
    "                ])\n",
    "    \n",
    "    return pd.Series(matches.iloc[:,1:]\n",
    "                     .as_matrix()\n",
    "                     .astype(int)\n",
    "                     .mean(axis=0)\n",
    "                     .round(0))                \n",
    "\n",
    "animals[dog_data_clean.columns[1:]] = (animals[animals.AnimalType=='Dog']\n",
    "                                 .apply(add_dog_data, axis=1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def mix_type(x):\n",
    "    if 'mix' in x.lower():\n",
    "        return 'Mix'\n",
    "    elif '/' in x:\n",
    "        return 'Cross'\n",
    "    else:\n",
    "        return 'Pure'\n",
    "\n",
    "animals['MixType'] = animals[\"Breed\"].map(mix_type)\n",
    "# plot_feature('MixType', data=animals)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "animals.loc[animals.AnimalType==\"Cat\", 'AvgWeight(pounds)'] = 9\n",
    "animals.loc[animals.AnimalType==\"Cat\", 'SizeScore(1to5)'] = 1\n",
    "animals.loc[animals.AnimalType==\"Cat\", \"Intelligent\"] = 3\n",
    "animals.loc[animals.AnimalType==\"Cat\", \"Friendliness\"] = 3"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Hair length and Color"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def hair(x):\n",
    "    if 'Short' in str(x):\n",
    "        return 'Short'\n",
    "    elif 'Medium' in str(x):\n",
    "        return 'Medium'\n",
    "    elif 'Long' in str(x):\n",
    "        return 'Long'\n",
    "    else:\n",
    "        return np.nan\n",
    "\n",
    "animals['Hair'] = animals[\"Breed\"].map(hair)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "So far these two are not very helpful. We need to dive deeper into these features.\n",
    "\n",
    "A lot of the color information for cats happens to be not just color but also the pattern (this also applies to dogs in some cases and others [like Merle]). A good resource for learning about cat colors and patterns is [here]( http://fixnation.org/2010/07/what-color-is-that-cat/). Using that resource let's map the colors and patterns to a smaller, more managable set."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "color_pattern = dict(\n",
    "    Tabby='Tabby',\n",
    "    Agouti='Tabby',\n",
    "    Tiger='Tabby',\n",
    "    Point='Point',\n",
    "    Tortie='TriColor',\n",
    "    Calico='TriColor',\n",
    "    Torbie='TriColor',\n",
    "    Brindle='TriColor',\n",
    "    Smoke='Smoke',\n",
    "    Merle='Merle')\n",
    "\n",
    "color_synonym =dict(\n",
    "    Flame='Orange',\n",
    "    Apricot='Orange',\n",
    "    Red='Red',\n",
    "    Ruddy='Red',\n",
    "    Gold='Yellow',\n",
    "    Cream='Tan',\n",
    "    Buff='Tan',\n",
    "    Fawn='Tan',\n",
    "    Liver='Brown',\n",
    "    Seal='Brown',\n",
    "    Sable='Brown',\n",
    "    Chocolate='Brown',\n",
    "    Blue='Grey',\n",
    "    Lilac='Grey',\n",
    "    goldenrod=\"yellow\",\n",
    "    darkorange=\"orange\",\n",
    "    pink=\"salmon\", \n",
    "    peru=\"brown\", \n",
    "    saddlebrown=\"brown\",\n",
    "    sandybrown=\"brown\",\n",
    "    yellowgreen = \"yellow\",\n",
    "    olive=\"darkolivegreen\",\n",
    "    chocolate=\"brown\",\n",
    "    darkgray=\"grey\",\n",
    "    gainsboro=\"grey\",\n",
    "    mistyrose=\"salmon\"\n",
    ")\n",
    "color = {\"goldenrod\": \"yellow\",\n",
    "         \"darkorange\": \"orange\",\n",
    "         \"pink\" : \"salmon\", \n",
    "         \"peru\" : \"brown\", \n",
    "         \"saddlebrown\" : \"brown\",\n",
    "         \"sandybrown\" : \"brown\",\n",
    "         \"yellowgreen\"\n",
    "         \"olive\" : \"darkolivegreen\",\n",
    "         \"chocolate\" : \"brown\",\n",
    "         \"darkgray\" : \"grey\",\n",
    "         \"gainsboro\" : \"grey\"}\n",
    "\n",
    "main_color = [\n",
    "    'Black', \n",
    "    'Brown', \n",
    "    'Yellow', \n",
    "    'Gray', \n",
    "    'Orange', \n",
    "    'Pink', \n",
    "    'Red', \n",
    "    'Silver', \n",
    "    'Tan',\n",
    "    'White']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def reduce_colors(color):\n",
    "    result = [] \n",
    "    for c in [c.strip() for c in re.split('/| ', color)]:\n",
    "        if c in main_color:\n",
    "            result.append(c)\n",
    "        elif c in color_synonym.keys():\n",
    "            result.append(color_synonym[c])\n",
    "    return 'Unknown' if not result else '/'.join(sort(list(set(result))))\n",
    "    \n",
    "animals['ReducedColor'] = animals.Color.map(reduce_colors)\n",
    "# plot_feature('ReducedColor', data=animals, tick_font_size='9', legend_offset=-0.4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Now we have reduced our color to 48 classes.\n"
     ]
    }
   ],
   "source": [
    "print \"Now we have reduced our color to {} classes.\".format(len(animals.ReducedColor.unique()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# to further reduce the number of colors, we blended our colors then grouped them\n",
    "import webcolors\n",
    "\n",
    "def closest_colour(requested_colour):\n",
    "    min_colours = {}\n",
    "    for key, name in webcolors.css3_hex_to_names.items():\n",
    "        r_c, g_c, b_c = webcolors.hex_to_rgb(key)\n",
    "        rd = (r_c - requested_colour[0]) ** 2\n",
    "        gd = (g_c - requested_colour[1]) ** 2\n",
    "        bd = (b_c - requested_colour[2]) ** 2\n",
    "        min_colours[(rd + gd + bd)] = name\n",
    "    return min_colours[min(min_colours.keys())]\n",
    "\n",
    "def get_colour_name(requested_colour):\n",
    "    try:\n",
    "        closest_name = webcolors.rgb_to_name(requested_colour)\n",
    "    except ValueError:\n",
    "        closest_name = closest_colour(requested_colour)\n",
    "    return closest_name\n",
    "\n",
    "def memoize(func):\n",
    "    cache = {}\n",
    "    @wraps(func)\n",
    "    def wrap(*args):\n",
    "        if args not in cache:\n",
    "            cache[args] = func(*args)\n",
    "        return cache[args]\n",
    "    return wrap\n",
    "@memoize\n",
    "def get_blend_color(colorstring):    \n",
    "    colors = colorstring.split(\"/\")\n",
    "    pallette = array([hstack(webcolors.name_to_rgb(c)) for c in colors])\n",
    "    blend = np.mean(pallette, axis=0)\n",
    "    return get_colour_name(blend)\n",
    "\n",
    "def color_name(colorstring):\n",
    "    if (colorstring == \"Unknown\"):\n",
    "        return np.nan\n",
    "    cs = get_blend_color(colorstring)\n",
    "    if cs in color_synonym.keys():\n",
    "        return color_synonym[cs]\n",
    "    return cs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Now we have reduced our color to 22 classes.\n"
     ]
    }
   ],
   "source": [
    "animals[\"BlendedColor\"] = animals.ReducedColor.map(color_name, na_action=\"ignore\")\n",
    "print \"Now we have reduced our color to {} classes.\".format(len(animals.BlendedColor.unique()))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There are 22 blended colors above and the mix of dogs and cats might affect the proportions of outcomes so we are just showing the 10 most popular for cats and then dogs."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def reduce_pattern(color):\n",
    "    result = None\n",
    "    for c in [c.strip() for c in re.split('/| ', color)]:\n",
    "        if c in color_pattern.keys():\n",
    "            return color_pattern[c]\n",
    "    return 'BiColor' if '/' in color else 'Solid'\n",
    "\n",
    "animals['ReducedPattern'] = animals.Color.map(reduce_pattern)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Impute-Missing-Value\n",
    "Unfortunately, scikit-learn only has the most basic imputer that supports \"mean\", \"median\" and \"mostfrequent\". We found two packages that do advanced imputation algorithms. We cannot get fancyimpute up and running on the machine we use. Orange does not quite give us what we want. Although Orange3 might work, it requires python3."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The autoreload extension is already loaded. To reload it, use:\n",
      "  %reload_ext autoreload\n"
     ]
    }
   ],
   "source": [
    "%load_ext autoreload\n",
    "%autoreload 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "from clean import make_features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "test_df = make_features(status=\"test\")\n",
    "train_df = make_features(status=\"train\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "from clean import impute_features\n",
    "test_df_imputed = impute_features(test_df)\n",
    "train_df_imputed = impute_features(train_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>ID</th>\n",
       "      <th>Name</th>\n",
       "      <th>DateTime</th>\n",
       "      <th>AnimalType</th>\n",
       "      <th>SexuponOutcome</th>\n",
       "      <th>AgeuponOutcome</th>\n",
       "      <th>Breed</th>\n",
       "      <th>Color</th>\n",
       "      <th>HasName</th>\n",
       "      <th>Year</th>\n",
       "      <th>...</th>\n",
       "      <th>BernoulliSex</th>\n",
       "      <th>AvgWeight(pounds)</th>\n",
       "      <th>SizeScore(1to5)</th>\n",
       "      <th>Intelligent</th>\n",
       "      <th>Friendliness</th>\n",
       "      <th>AvgAge</th>\n",
       "      <th>MixType</th>\n",
       "      <th>Hair</th>\n",
       "      <th>BlendedColor</th>\n",
       "      <th>ReducedPattern</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>Summer</td>\n",
       "      <td>2015-10-12 12:15:00</td>\n",
       "      <td>Dog</td>\n",
       "      <td>Intact Female</td>\n",
       "      <td>10 months</td>\n",
       "      <td>Labrador Retriever Mix</td>\n",
       "      <td>Red/White</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2015</td>\n",
       "      <td>...</td>\n",
       "      <td>Intact Female</td>\n",
       "      <td>68.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>13.0</td>\n",
       "      <td>Mix</td>\n",
       "      <td>Short</td>\n",
       "      <td>salmon</td>\n",
       "      <td>BiColor</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>Cheyenne</td>\n",
       "      <td>2014-07-26 17:59:00</td>\n",
       "      <td>Dog</td>\n",
       "      <td>Spayed Female</td>\n",
       "      <td>2 years</td>\n",
       "      <td>German Shepherd/Siberian Husky</td>\n",
       "      <td>Black/Tan</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2014</td>\n",
       "      <td>...</td>\n",
       "      <td>Spayed Female</td>\n",
       "      <td>59.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>12.0</td>\n",
       "      <td>Cross</td>\n",
       "      <td>Medium</td>\n",
       "      <td>darkolivegreen</td>\n",
       "      <td>BiColor</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>Gus</td>\n",
       "      <td>2016-01-13 12:20:00</td>\n",
       "      <td>Cat</td>\n",
       "      <td>Neutered Male</td>\n",
       "      <td>1 year</td>\n",
       "      <td>Domestic Shorthair Mix</td>\n",
       "      <td>Brown Tabby</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2016</td>\n",
       "      <td>...</td>\n",
       "      <td>Neutered Male</td>\n",
       "      <td>9.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>12.0</td>\n",
       "      <td>Mix</td>\n",
       "      <td>Short</td>\n",
       "      <td>brown</td>\n",
       "      <td>Tabby</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>Pongo</td>\n",
       "      <td>2013-12-28 18:12:00</td>\n",
       "      <td>Dog</td>\n",
       "      <td>Intact Male</td>\n",
       "      <td>4 months</td>\n",
       "      <td>Collie Smooth Mix</td>\n",
       "      <td>Tricolor</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2013</td>\n",
       "      <td>...</td>\n",
       "      <td>Intact Male</td>\n",
       "      <td>16.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>13.0</td>\n",
       "      <td>Mix</td>\n",
       "      <td>Short</td>\n",
       "      <td>Not_clear</td>\n",
       "      <td>Solid</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>Skooter</td>\n",
       "      <td>2015-09-24 17:59:00</td>\n",
       "      <td>Dog</td>\n",
       "      <td>Neutered Male</td>\n",
       "      <td>2 years</td>\n",
       "      <td>Miniature Poodle Mix</td>\n",
       "      <td>White</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2015</td>\n",
       "      <td>...</td>\n",
       "      <td>Neutered Male</td>\n",
       "      <td>14.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>Mix</td>\n",
       "      <td>Medium</td>\n",
       "      <td>white</td>\n",
       "      <td>Solid</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>6</td>\n",
       "      <td>Beau</td>\n",
       "      <td>2015-06-23 11:17:00</td>\n",
       "      <td>Dog</td>\n",
       "      <td>Neutered Male</td>\n",
       "      <td>3 years</td>\n",
       "      <td>Beagle Mix</td>\n",
       "      <td>Brown/White</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2015</td>\n",
       "      <td>...</td>\n",
       "      <td>Neutered Male</td>\n",
       "      <td>28.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>Mix</td>\n",
       "      <td>Short</td>\n",
       "      <td>rosybrown</td>\n",
       "      <td>BiColor</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>7</td>\n",
       "      <td>Bobo</td>\n",
       "      <td>2014-03-12 09:45:00</td>\n",
       "      <td>Cat</td>\n",
       "      <td>Neutered Male</td>\n",
       "      <td>13 years</td>\n",
       "      <td>Domestic Medium Hair Mix</td>\n",
       "      <td>Brown Tabby/White</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2014</td>\n",
       "      <td>...</td>\n",
       "      <td>Neutered Male</td>\n",
       "      <td>9.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>12.0</td>\n",
       "      <td>Mix</td>\n",
       "      <td>Medium</td>\n",
       "      <td>rosybrown</td>\n",
       "      <td>Tabby</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>8</td>\n",
       "      <td>Abby</td>\n",
       "      <td>2014-06-25 08:27:00</td>\n",
       "      <td>Cat</td>\n",
       "      <td>Spayed Female</td>\n",
       "      <td>6 months</td>\n",
       "      <td>Domestic Shorthair Mix</td>\n",
       "      <td>Brown Tabby</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2014</td>\n",
       "      <td>...</td>\n",
       "      <td>Spayed Female</td>\n",
       "      <td>9.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>12.0</td>\n",
       "      <td>Mix</td>\n",
       "      <td>Short</td>\n",
       "      <td>brown</td>\n",
       "      <td>Tabby</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>9</td>\n",
       "      <td>Ruby Grace</td>\n",
       "      <td>2014-11-12 18:05:00</td>\n",
       "      <td>Dog</td>\n",
       "      <td>Spayed Female</td>\n",
       "      <td>3 months</td>\n",
       "      <td>Cairn Terrier</td>\n",
       "      <td>Black/Cream</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2014</td>\n",
       "      <td>...</td>\n",
       "      <td>Spayed Female</td>\n",
       "      <td>14.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>Pure</td>\n",
       "      <td>Short</td>\n",
       "      <td>darkolivegreen</td>\n",
       "      <td>BiColor</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>10</td>\n",
       "      <td>Ruby</td>\n",
       "      <td>2014-04-07 17:41:00</td>\n",
       "      <td>Dog</td>\n",
       "      <td>Spayed Female</td>\n",
       "      <td>1 year</td>\n",
       "      <td>Pit Bull Mix</td>\n",
       "      <td>Brown/White</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2014</td>\n",
       "      <td>...</td>\n",
       "      <td>Spayed Female</td>\n",
       "      <td>60.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>12.0</td>\n",
       "      <td>Mix</td>\n",
       "      <td>Short</td>\n",
       "      <td>rosybrown</td>\n",
       "      <td>BiColor</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>11</td>\n",
       "      <td>Marlee</td>\n",
       "      <td>2014-01-01 17:03:00</td>\n",
       "      <td>Dog</td>\n",
       "      <td>Spayed Female</td>\n",
       "      <td>5 months</td>\n",
       "      <td>Labrador Retriever Mix</td>\n",
       "      <td>White/Black</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2014</td>\n",
       "      <td>...</td>\n",
       "      <td>Spayed Female</td>\n",
       "      <td>68.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>13.0</td>\n",
       "      <td>Mix</td>\n",
       "      <td>Short</td>\n",
       "      <td>grey</td>\n",
       "      <td>BiColor</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>12</td>\n",
       "      <td>Lala</td>\n",
       "      <td>2014-09-19 15:40:00</td>\n",
       "      <td>Dog</td>\n",
       "      <td>Intact Female</td>\n",
       "      <td>2 years</td>\n",
       "      <td>Pit Bull Mix</td>\n",
       "      <td>Blue/White</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2014</td>\n",
       "      <td>...</td>\n",
       "      <td>Intact Female</td>\n",
       "      <td>60.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>12.0</td>\n",
       "      <td>Mix</td>\n",
       "      <td>Long</td>\n",
       "      <td>silver</td>\n",
       "      <td>BiColor</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>13</td>\n",
       "      <td>Chloe</td>\n",
       "      <td>2015-01-25 15:13:00</td>\n",
       "      <td>Dog</td>\n",
       "      <td>Spayed Female</td>\n",
       "      <td>6 months</td>\n",
       "      <td>Pit Bull Mix</td>\n",
       "      <td>Brown Brindle</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2015</td>\n",
       "      <td>...</td>\n",
       "      <td>Spayed Female</td>\n",
       "      <td>60.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>12.0</td>\n",
       "      <td>Mix</td>\n",
       "      <td>Short</td>\n",
       "      <td>brown</td>\n",
       "      <td>TriColor</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>14</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2015-09-22 18:17:00</td>\n",
       "      <td>Dog</td>\n",
       "      <td>Neutered Male</td>\n",
       "      <td>2 years</td>\n",
       "      <td>Miniature Poodle Mix</td>\n",
       "      <td>White</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2015</td>\n",
       "      <td>...</td>\n",
       "      <td>Neutered Male</td>\n",
       "      <td>14.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>Mix</td>\n",
       "      <td>Short</td>\n",
       "      <td>white</td>\n",
       "      <td>Solid</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>15</td>\n",
       "      <td>Duffy</td>\n",
       "      <td>2014-07-20 19:24:00</td>\n",
       "      <td>Dog</td>\n",
       "      <td>Neutered Male</td>\n",
       "      <td>1 year</td>\n",
       "      <td>Flat Coat Retriever/German Shepherd</td>\n",
       "      <td>Black/Brown</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2014</td>\n",
       "      <td>...</td>\n",
       "      <td>Neutered Male</td>\n",
       "      <td>68.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>12.0</td>\n",
       "      <td>Cross</td>\n",
       "      <td>Short</td>\n",
       "      <td>maroon</td>\n",
       "      <td>BiColor</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>16</td>\n",
       "      <td>Jiggy</td>\n",
       "      <td>2014-03-26 18:33:00</td>\n",
       "      <td>Dog</td>\n",
       "      <td>Spayed Female</td>\n",
       "      <td>4 months</td>\n",
       "      <td>Pointer/Labrador Retriever</td>\n",
       "      <td>White</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2014</td>\n",
       "      <td>...</td>\n",
       "      <td>Spayed Female</td>\n",
       "      <td>62.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>Cross</td>\n",
       "      <td>Short</td>\n",
       "      <td>white</td>\n",
       "      <td>Solid</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>17</td>\n",
       "      <td>Ghengis</td>\n",
       "      <td>2015-05-29 17:04:00</td>\n",
       "      <td>Dog</td>\n",
       "      <td>Neutered Male</td>\n",
       "      <td>5 years</td>\n",
       "      <td>Doberman Pinsch Mix</td>\n",
       "      <td>Black/Tan</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2015</td>\n",
       "      <td>...</td>\n",
       "      <td>Neutered Male</td>\n",
       "      <td>77.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>Mix</td>\n",
       "      <td>Short</td>\n",
       "      <td>darkolivegreen</td>\n",
       "      <td>BiColor</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>18</td>\n",
       "      <td>Decker</td>\n",
       "      <td>2016-01-19 19:31:00</td>\n",
       "      <td>Dog</td>\n",
       "      <td>Neutered Male</td>\n",
       "      <td>3 years</td>\n",
       "      <td>English Bulldog</td>\n",
       "      <td>White/Tricolor</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2016</td>\n",
       "      <td>...</td>\n",
       "      <td>Neutered Male</td>\n",
       "      <td>53.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>Pure</td>\n",
       "      <td>Short</td>\n",
       "      <td>white</td>\n",
       "      <td>BiColor</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>19</td>\n",
       "      <td>Freckles</td>\n",
       "      <td>2014-06-16 12:49:00</td>\n",
       "      <td>Cat</td>\n",
       "      <td>Spayed Female</td>\n",
       "      <td>3 years</td>\n",
       "      <td>Domestic Shorthair Mix</td>\n",
       "      <td>Torbie</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2014</td>\n",
       "      <td>...</td>\n",
       "      <td>Spayed Female</td>\n",
       "      <td>9.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>13.0</td>\n",
       "      <td>Mix</td>\n",
       "      <td>Short</td>\n",
       "      <td>Not_clear</td>\n",
       "      <td>TriColor</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>20</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2014-02-11 19:27:00</td>\n",
       "      <td>Cat</td>\n",
       "      <td>Intact Female</td>\n",
       "      <td>4 years</td>\n",
       "      <td>Domestic Shorthair Mix</td>\n",
       "      <td>Torbie</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2014</td>\n",
       "      <td>...</td>\n",
       "      <td>Intact Female</td>\n",
       "      <td>9.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>Mix</td>\n",
       "      <td>Short</td>\n",
       "      <td>Not_clear</td>\n",
       "      <td>TriColor</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>21</td>\n",
       "      <td>Pumpkin</td>\n",
       "      <td>2014-11-14 19:32:00</td>\n",
       "      <td>Cat</td>\n",
       "      <td>Spayed Female</td>\n",
       "      <td>2 years</td>\n",
       "      <td>Domestic Shorthair Mix</td>\n",
       "      <td>Black</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2014</td>\n",
       "      <td>...</td>\n",
       "      <td>Spayed Female</td>\n",
       "      <td>9.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>13.0</td>\n",
       "      <td>Mix</td>\n",
       "      <td>Short</td>\n",
       "      <td>black</td>\n",
       "      <td>Solid</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>22</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2015-05-04 13:46:00</td>\n",
       "      <td>Cat</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>3 weeks</td>\n",
       "      <td>Domestic Shorthair Mix</td>\n",
       "      <td>Black</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2015</td>\n",
       "      <td>...</td>\n",
       "      <td>Intact Male</td>\n",
       "      <td>9.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>12.0</td>\n",
       "      <td>Mix</td>\n",
       "      <td>Short</td>\n",
       "      <td>black</td>\n",
       "      <td>Solid</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>23</td>\n",
       "      <td>Pluto</td>\n",
       "      <td>2014-06-07 14:14:00</td>\n",
       "      <td>Dog</td>\n",
       "      <td>Neutered Male</td>\n",
       "      <td>1 year</td>\n",
       "      <td>Norwich Terrier Mix</td>\n",
       "      <td>White/Tan</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2014</td>\n",
       "      <td>...</td>\n",
       "      <td>Neutered Male</td>\n",
       "      <td>12.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>Mix</td>\n",
       "      <td>Medium</td>\n",
       "      <td>wheat</td>\n",
       "      <td>BiColor</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>24</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2014-07-20 15:37:00</td>\n",
       "      <td>Cat</td>\n",
       "      <td>Intact Female</td>\n",
       "      <td>1 week</td>\n",
       "      <td>Domestic Shorthair Mix</td>\n",
       "      <td>Tortie</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2014</td>\n",
       "      <td>...</td>\n",
       "      <td>Intact Female</td>\n",
       "      <td>9.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>Mix</td>\n",
       "      <td>Short</td>\n",
       "      <td>Not_clear</td>\n",
       "      <td>TriColor</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>25</td>\n",
       "      <td>Portia</td>\n",
       "      <td>2013-10-26 16:23:00</td>\n",
       "      <td>Dog</td>\n",
       "      <td>Spayed Female</td>\n",
       "      <td>1 year</td>\n",
       "      <td>Black Mouth Cur/Boxer</td>\n",
       "      <td>Tan</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2013</td>\n",
       "      <td>...</td>\n",
       "      <td>Spayed Female</td>\n",
       "      <td>79.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>Cross</td>\n",
       "      <td>Long</td>\n",
       "      <td>tan</td>\n",
       "      <td>Solid</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>26</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2014-12-29 13:10:00</td>\n",
       "      <td>Cat</td>\n",
       "      <td>Neutered Male</td>\n",
       "      <td>1 year</td>\n",
       "      <td>Domestic Shorthair Mix</td>\n",
       "      <td>Brown Tabby</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2014</td>\n",
       "      <td>...</td>\n",
       "      <td>Neutered Male</td>\n",
       "      <td>9.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>13.0</td>\n",
       "      <td>Mix</td>\n",
       "      <td>Short</td>\n",
       "      <td>brown</td>\n",
       "      <td>Tabby</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>27</td>\n",
       "      <td>Lucy</td>\n",
       "      <td>2015-07-23 16:18:00</td>\n",
       "      <td>Cat</td>\n",
       "      <td>Spayed Female</td>\n",
       "      <td>2 months</td>\n",
       "      <td>Domestic Medium Hair Mix</td>\n",
       "      <td>Black Smoke</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2015</td>\n",
       "      <td>...</td>\n",
       "      <td>Spayed Female</td>\n",
       "      <td>9.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>Mix</td>\n",
       "      <td>Medium</td>\n",
       "      <td>black</td>\n",
       "      <td>Smoke</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>28</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2015-03-19 08:33:00</td>\n",
       "      <td>Dog</td>\n",
       "      <td>Intact Male</td>\n",
       "      <td>2 months</td>\n",
       "      <td>Norwich Terrier Mix</td>\n",
       "      <td>Brown</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2015</td>\n",
       "      <td>...</td>\n",
       "      <td>Intact Male</td>\n",
       "      <td>12.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>Mix</td>\n",
       "      <td>Long</td>\n",
       "      <td>brown</td>\n",
       "      <td>Solid</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>29</td>\n",
       "      <td>Jack</td>\n",
       "      <td>2014-10-09 17:22:00</td>\n",
       "      <td>Dog</td>\n",
       "      <td>Neutered Male</td>\n",
       "      <td>3 years</td>\n",
       "      <td>Labrador Retriever Mix</td>\n",
       "      <td>Black/White</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2014</td>\n",
       "      <td>...</td>\n",
       "      <td>Neutered Male</td>\n",
       "      <td>68.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>13.0</td>\n",
       "      <td>Mix</td>\n",
       "      <td>Short</td>\n",
       "      <td>grey</td>\n",
       "      <td>BiColor</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>30</td>\n",
       "      <td>Knight</td>\n",
       "      <td>2014-05-13 17:03:00</td>\n",
       "      <td>Dog</td>\n",
       "      <td>Neutered Male</td>\n",
       "      <td>2 years</td>\n",
       "      <td>Labrador Retriever Mix</td>\n",
       "      <td>Black</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2014</td>\n",
       "      <td>...</td>\n",
       "      <td>Neutered Male</td>\n",
       "      <td>68.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>13.0</td>\n",
       "      <td>Mix</td>\n",
       "      <td>Short</td>\n",
       "      <td>black</td>\n",
       "      <td>Solid</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11426</th>\n",
       "      <td>11427</td>\n",
       "      <td>Fatboy</td>\n",
       "      <td>2013-12-28 14:22:00</td>\n",
       "      <td>Dog</td>\n",
       "      <td>Neutered Male</td>\n",
       "      <td>5 years</td>\n",
       "      <td>Boxer Mix</td>\n",
       "      <td>Brown/White</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2013</td>\n",
       "      <td>...</td>\n",
       "      <td>Neutered Male</td>\n",
       "      <td>60.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>Mix</td>\n",
       "      <td>Short</td>\n",
       "      <td>rosybrown</td>\n",
       "      <td>BiColor</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11427</th>\n",
       "      <td>11428</td>\n",
       "      <td>Calypso</td>\n",
       "      <td>2013-12-21 15:55:00</td>\n",
       "      <td>Dog</td>\n",
       "      <td>Intact Female</td>\n",
       "      <td>2 months</td>\n",
       "      <td>American Pit Bull Terrier Mix</td>\n",
       "      <td>Brown Brindle</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2013</td>\n",
       "      <td>...</td>\n",
       "      <td>Intact Female</td>\n",
       "      <td>60.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>12.0</td>\n",
       "      <td>Mix</td>\n",
       "      <td>Medium</td>\n",
       "      <td>brown</td>\n",
       "      <td>TriColor</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11428</th>\n",
       "      <td>11429</td>\n",
       "      <td>Bolt</td>\n",
       "      <td>2013-11-15 12:32:00</td>\n",
       "      <td>Dog</td>\n",
       "      <td>Neutered Male</td>\n",
       "      <td>3 years</td>\n",
       "      <td>Siberian Husky</td>\n",
       "      <td>Red/White</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2013</td>\n",
       "      <td>...</td>\n",
       "      <td>Neutered Male</td>\n",
       "      <td>48.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>12.0</td>\n",
       "      <td>Pure</td>\n",
       "      <td>Short</td>\n",
       "      <td>salmon</td>\n",
       "      <td>BiColor</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11429</th>\n",
       "      <td>11430</td>\n",
       "      <td>Mary</td>\n",
       "      <td>2014-02-17 18:04:00</td>\n",
       "      <td>Dog</td>\n",
       "      <td>Spayed Female</td>\n",
       "      <td>2 years</td>\n",
       "      <td>German Shepherd Mix</td>\n",
       "      <td>Sable</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2014</td>\n",
       "      <td>...</td>\n",
       "      <td>Spayed Female</td>\n",
       "      <td>70.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>12.0</td>\n",
       "      <td>Mix</td>\n",
       "      <td>Short</td>\n",
       "      <td>brown</td>\n",
       "      <td>Solid</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11430</th>\n",
       "      <td>11431</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2014-12-08 12:40:00</td>\n",
       "      <td>Cat</td>\n",
       "      <td>Spayed Female</td>\n",
       "      <td>2 months</td>\n",
       "      <td>Domestic Shorthair Mix</td>\n",
       "      <td>Blue</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2014</td>\n",
       "      <td>...</td>\n",
       "      <td>Spayed Female</td>\n",
       "      <td>9.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>13.0</td>\n",
       "      <td>Mix</td>\n",
       "      <td>Short</td>\n",
       "      <td>grey</td>\n",
       "      <td>Solid</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11431</th>\n",
       "      <td>11432</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2015-06-30 09:00:00</td>\n",
       "      <td>Cat</td>\n",
       "      <td>Intact Male</td>\n",
       "      <td>2 years</td>\n",
       "      <td>Domestic Shorthair Mix</td>\n",
       "      <td>Blue Tabby/White</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2015</td>\n",
       "      <td>...</td>\n",
       "      <td>Intact Male</td>\n",
       "      <td>9.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>Mix</td>\n",
       "      <td>Short</td>\n",
       "      <td>silver</td>\n",
       "      <td>Tabby</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11432</th>\n",
       "      <td>11433</td>\n",
       "      <td>Fiona</td>\n",
       "      <td>2014-03-21 17:33:00</td>\n",
       "      <td>Dog</td>\n",
       "      <td>Spayed Female</td>\n",
       "      <td>5 years</td>\n",
       "      <td>Maltese</td>\n",
       "      <td>White</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2014</td>\n",
       "      <td>...</td>\n",
       "      <td>Spayed Female</td>\n",
       "      <td>6.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>17.0</td>\n",
       "      <td>Pure</td>\n",
       "      <td>Short</td>\n",
       "      <td>white</td>\n",
       "      <td>Solid</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11433</th>\n",
       "      <td>11434</td>\n",
       "      <td>Evanna</td>\n",
       "      <td>2015-08-16 16:40:00</td>\n",
       "      <td>Cat</td>\n",
       "      <td>Intact Female</td>\n",
       "      <td>2 months</td>\n",
       "      <td>Domestic Shorthair Mix</td>\n",
       "      <td>Black/White</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2015</td>\n",
       "      <td>...</td>\n",
       "      <td>Intact Female</td>\n",
       "      <td>9.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>Mix</td>\n",
       "      <td>Short</td>\n",
       "      <td>grey</td>\n",
       "      <td>BiColor</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11434</th>\n",
       "      <td>11435</td>\n",
       "      <td>April</td>\n",
       "      <td>2015-09-06 12:41:00</td>\n",
       "      <td>Dog</td>\n",
       "      <td>Spayed Female</td>\n",
       "      <td>6 years</td>\n",
       "      <td>Cardigan Welsh Corgi/German Shepherd</td>\n",
       "      <td>Black/Tan</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2015</td>\n",
       "      <td>...</td>\n",
       "      <td>Spayed Female</td>\n",
       "      <td>51.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>13.0</td>\n",
       "      <td>Cross</td>\n",
       "      <td>Short</td>\n",
       "      <td>darkolivegreen</td>\n",
       "      <td>BiColor</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11435</th>\n",
       "      <td>11436</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2015-08-13 17:09:00</td>\n",
       "      <td>Cat</td>\n",
       "      <td>Intact Male</td>\n",
       "      <td>3 weeks</td>\n",
       "      <td>Domestic Shorthair Mix</td>\n",
       "      <td>Brown Tabby</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2015</td>\n",
       "      <td>...</td>\n",
       "      <td>Intact Male</td>\n",
       "      <td>9.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>Mix</td>\n",
       "      <td>Short</td>\n",
       "      <td>brown</td>\n",
       "      <td>Tabby</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11436</th>\n",
       "      <td>11437</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2014-03-21 17:42:00</td>\n",
       "      <td>Cat</td>\n",
       "      <td>Neutered Male</td>\n",
       "      <td>5 months</td>\n",
       "      <td>Domestic Shorthair Mix</td>\n",
       "      <td>Black/White</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2014</td>\n",
       "      <td>...</td>\n",
       "      <td>Neutered Male</td>\n",
       "      <td>9.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>Mix</td>\n",
       "      <td>Short</td>\n",
       "      <td>grey</td>\n",
       "      <td>BiColor</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11437</th>\n",
       "      <td>11438</td>\n",
       "      <td>Peludita</td>\n",
       "      <td>2014-01-10 13:46:00</td>\n",
       "      <td>Dog</td>\n",
       "      <td>Intact Female</td>\n",
       "      <td>6 months</td>\n",
       "      <td>Chihuahua Longhair/Dachshund</td>\n",
       "      <td>White/Tan</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2014</td>\n",
       "      <td>...</td>\n",
       "      <td>Intact Female</td>\n",
       "      <td>13.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>Cross</td>\n",
       "      <td>Long</td>\n",
       "      <td>wheat</td>\n",
       "      <td>BiColor</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11438</th>\n",
       "      <td>11439</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2015-05-02 20:40:00</td>\n",
       "      <td>Dog</td>\n",
       "      <td>Neutered Male</td>\n",
       "      <td>1 month</td>\n",
       "      <td>Labrador Retriever Mix</td>\n",
       "      <td>Sable</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2015</td>\n",
       "      <td>...</td>\n",
       "      <td>Neutered Male</td>\n",
       "      <td>68.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>13.0</td>\n",
       "      <td>Mix</td>\n",
       "      <td>Short</td>\n",
       "      <td>brown</td>\n",
       "      <td>Solid</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11439</th>\n",
       "      <td>11440</td>\n",
       "      <td>Socks</td>\n",
       "      <td>2014-05-18 16:47:00</td>\n",
       "      <td>Dog</td>\n",
       "      <td>Spayed Female</td>\n",
       "      <td>2 years</td>\n",
       "      <td>Pit Bull Mix</td>\n",
       "      <td>White/Brown</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2014</td>\n",
       "      <td>...</td>\n",
       "      <td>Spayed Female</td>\n",
       "      <td>60.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>12.0</td>\n",
       "      <td>Mix</td>\n",
       "      <td>Short</td>\n",
       "      <td>rosybrown</td>\n",
       "      <td>BiColor</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11440</th>\n",
       "      <td>11441</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2015-10-20 17:13:00</td>\n",
       "      <td>Cat</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>1 week</td>\n",
       "      <td>Domestic Shorthair Mix</td>\n",
       "      <td>Brown Tabby</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2015</td>\n",
       "      <td>...</td>\n",
       "      <td>Intact Male</td>\n",
       "      <td>9.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>Mix</td>\n",
       "      <td>Short</td>\n",
       "      <td>brown</td>\n",
       "      <td>Tabby</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11441</th>\n",
       "      <td>11442</td>\n",
       "      <td>Doogie</td>\n",
       "      <td>2015-01-07 17:41:00</td>\n",
       "      <td>Dog</td>\n",
       "      <td>Neutered Male</td>\n",
       "      <td>4 years</td>\n",
       "      <td>Pit Bull Mix</td>\n",
       "      <td>Blue/White</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2015</td>\n",
       "      <td>...</td>\n",
       "      <td>Neutered Male</td>\n",
       "      <td>60.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>12.0</td>\n",
       "      <td>Mix</td>\n",
       "      <td>Short</td>\n",
       "      <td>silver</td>\n",
       "      <td>BiColor</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11442</th>\n",
       "      <td>11443</td>\n",
       "      <td>Sterling</td>\n",
       "      <td>2015-11-28 16:23:00</td>\n",
       "      <td>Dog</td>\n",
       "      <td>Neutered Male</td>\n",
       "      <td>2 years</td>\n",
       "      <td>Chihuahua Shorthair Mix</td>\n",
       "      <td>Tan/White</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2015</td>\n",
       "      <td>...</td>\n",
       "      <td>Neutered Male</td>\n",
       "      <td>6.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>Mix</td>\n",
       "      <td>Short</td>\n",
       "      <td>wheat</td>\n",
       "      <td>BiColor</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11443</th>\n",
       "      <td>11444</td>\n",
       "      <td>Bella</td>\n",
       "      <td>2014-10-04 12:17:00</td>\n",
       "      <td>Cat</td>\n",
       "      <td>Spayed Female</td>\n",
       "      <td>6 months</td>\n",
       "      <td>Domestic Longhair Mix</td>\n",
       "      <td>Torbie</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2014</td>\n",
       "      <td>...</td>\n",
       "      <td>Spayed Female</td>\n",
       "      <td>9.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>12.0</td>\n",
       "      <td>Mix</td>\n",
       "      <td>Long</td>\n",
       "      <td>Not_clear</td>\n",
       "      <td>TriColor</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11444</th>\n",
       "      <td>11445</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2014-06-01 17:15:00</td>\n",
       "      <td>Cat</td>\n",
       "      <td>Intact Female</td>\n",
       "      <td>2 years</td>\n",
       "      <td>Domestic Shorthair Mix</td>\n",
       "      <td>Calico</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2014</td>\n",
       "      <td>...</td>\n",
       "      <td>Intact Female</td>\n",
       "      <td>9.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>12.0</td>\n",
       "      <td>Mix</td>\n",
       "      <td>Short</td>\n",
       "      <td>Not_clear</td>\n",
       "      <td>TriColor</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11445</th>\n",
       "      <td>11446</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2015-04-14 14:34:00</td>\n",
       "      <td>Cat</td>\n",
       "      <td>Intact Male</td>\n",
       "      <td>2 weeks</td>\n",
       "      <td>Domestic Shorthair Mix</td>\n",
       "      <td>White</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2015</td>\n",
       "      <td>...</td>\n",
       "      <td>Intact Male</td>\n",
       "      <td>9.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>Mix</td>\n",
       "      <td>Short</td>\n",
       "      <td>white</td>\n",
       "      <td>Solid</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11446</th>\n",
       "      <td>11447</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2015-07-19 20:03:00</td>\n",
       "      <td>Cat</td>\n",
       "      <td>Neutered Male</td>\n",
       "      <td>2 months</td>\n",
       "      <td>Domestic Shorthair Mix</td>\n",
       "      <td>Black</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2015</td>\n",
       "      <td>...</td>\n",
       "      <td>Neutered Male</td>\n",
       "      <td>9.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>Mix</td>\n",
       "      <td>Short</td>\n",
       "      <td>black</td>\n",
       "      <td>Solid</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11447</th>\n",
       "      <td>11448</td>\n",
       "      <td>Sadie</td>\n",
       "      <td>2014-10-19 12:21:00</td>\n",
       "      <td>Dog</td>\n",
       "      <td>Spayed Female</td>\n",
       "      <td>1 year</td>\n",
       "      <td>Chihuahua Longhair Mix</td>\n",
       "      <td>Orange/Tan</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2014</td>\n",
       "      <td>...</td>\n",
       "      <td>Spayed Female</td>\n",
       "      <td>6.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>Mix</td>\n",
       "      <td>Long</td>\n",
       "      <td>brown</td>\n",
       "      <td>BiColor</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11448</th>\n",
       "      <td>11449</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2014-08-17 17:48:00</td>\n",
       "      <td>Dog</td>\n",
       "      <td>Spayed Female</td>\n",
       "      <td>2 months</td>\n",
       "      <td>Labrador Retriever Mix</td>\n",
       "      <td>Brown</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2014</td>\n",
       "      <td>...</td>\n",
       "      <td>Spayed Female</td>\n",
       "      <td>68.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>13.0</td>\n",
       "      <td>Mix</td>\n",
       "      <td>Short</td>\n",
       "      <td>brown</td>\n",
       "      <td>Solid</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11449</th>\n",
       "      <td>11450</td>\n",
       "      <td>Caicos</td>\n",
       "      <td>2015-07-20 14:55:00</td>\n",
       "      <td>Dog</td>\n",
       "      <td>Spayed Female</td>\n",
       "      <td>6 months</td>\n",
       "      <td>Bearded Collie Mix</td>\n",
       "      <td>Blue Merle/White</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2015</td>\n",
       "      <td>...</td>\n",
       "      <td>Spayed Female</td>\n",
       "      <td>50.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>13.0</td>\n",
       "      <td>Mix</td>\n",
       "      <td>Medium</td>\n",
       "      <td>silver</td>\n",
       "      <td>Merle</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11450</th>\n",
       "      <td>11451</td>\n",
       "      <td>Cosmo</td>\n",
       "      <td>2015-11-05 17:19:00</td>\n",
       "      <td>Dog</td>\n",
       "      <td>Neutered Male</td>\n",
       "      <td>10 months</td>\n",
       "      <td>Staffordshire Mix</td>\n",
       "      <td>Tan/White</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2015</td>\n",
       "      <td>...</td>\n",
       "      <td>Neutered Male</td>\n",
       "      <td>31.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>13.0</td>\n",
       "      <td>Mix</td>\n",
       "      <td>Short</td>\n",
       "      <td>wheat</td>\n",
       "      <td>BiColor</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11451</th>\n",
       "      <td>11452</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2014-07-08 14:50:00</td>\n",
       "      <td>Cat</td>\n",
       "      <td>Neutered Male</td>\n",
       "      <td>2 months</td>\n",
       "      <td>Domestic Shorthair Mix</td>\n",
       "      <td>Black</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2014</td>\n",
       "      <td>...</td>\n",
       "      <td>Neutered Male</td>\n",
       "      <td>9.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>Mix</td>\n",
       "      <td>Short</td>\n",
       "      <td>black</td>\n",
       "      <td>Solid</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11452</th>\n",
       "      <td>11453</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2014-10-21 12:57:00</td>\n",
       "      <td>Cat</td>\n",
       "      <td>Intact Female</td>\n",
       "      <td>2 weeks</td>\n",
       "      <td>Domestic Shorthair Mix</td>\n",
       "      <td>Blue</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2014</td>\n",
       "      <td>...</td>\n",
       "      <td>Intact Female</td>\n",
       "      <td>9.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>Mix</td>\n",
       "      <td>Short</td>\n",
       "      <td>grey</td>\n",
       "      <td>Solid</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11453</th>\n",
       "      <td>11454</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2014-09-29 09:00:00</td>\n",
       "      <td>Cat</td>\n",
       "      <td>Intact Female</td>\n",
       "      <td>1 year</td>\n",
       "      <td>Domestic Shorthair Mix</td>\n",
       "      <td>Calico</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2014</td>\n",
       "      <td>...</td>\n",
       "      <td>Intact Female</td>\n",
       "      <td>9.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>Mix</td>\n",
       "      <td>Short</td>\n",
       "      <td>Not_clear</td>\n",
       "      <td>TriColor</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11454</th>\n",
       "      <td>11455</td>\n",
       "      <td>Rambo</td>\n",
       "      <td>2015-09-05 17:16:00</td>\n",
       "      <td>Dog</td>\n",
       "      <td>Neutered Male</td>\n",
       "      <td>6 years</td>\n",
       "      <td>German Shepherd Mix</td>\n",
       "      <td>Black/Tan</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2015</td>\n",
       "      <td>...</td>\n",
       "      <td>Neutered Male</td>\n",
       "      <td>70.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>12.0</td>\n",
       "      <td>Mix</td>\n",
       "      <td>Short</td>\n",
       "      <td>darkolivegreen</td>\n",
       "      <td>BiColor</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11455</th>\n",
       "      <td>11456</td>\n",
       "      <td>Gusto</td>\n",
       "      <td>2014-07-12 18:40:00</td>\n",
       "      <td>Dog</td>\n",
       "      <td>Intact Male</td>\n",
       "      <td>4 years</td>\n",
       "      <td>Labrador Retriever</td>\n",
       "      <td>Cream</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2014</td>\n",
       "      <td>...</td>\n",
       "      <td>Intact Male</td>\n",
       "      <td>68.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>13.0</td>\n",
       "      <td>Pure</td>\n",
       "      <td>Short</td>\n",
       "      <td>tan</td>\n",
       "      <td>Solid</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>11456 rows × 28 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          ID        Name             DateTime AnimalType SexuponOutcome  \\\n",
       "0          1      Summer  2015-10-12 12:15:00        Dog  Intact Female   \n",
       "1          2    Cheyenne  2014-07-26 17:59:00        Dog  Spayed Female   \n",
       "2          3         Gus  2016-01-13 12:20:00        Cat  Neutered Male   \n",
       "3          4       Pongo  2013-12-28 18:12:00        Dog    Intact Male   \n",
       "4          5     Skooter  2015-09-24 17:59:00        Dog  Neutered Male   \n",
       "5          6        Beau  2015-06-23 11:17:00        Dog  Neutered Male   \n",
       "6          7        Bobo  2014-03-12 09:45:00        Cat  Neutered Male   \n",
       "7          8        Abby  2014-06-25 08:27:00        Cat  Spayed Female   \n",
       "8          9  Ruby Grace  2014-11-12 18:05:00        Dog  Spayed Female   \n",
       "9         10        Ruby  2014-04-07 17:41:00        Dog  Spayed Female   \n",
       "10        11      Marlee  2014-01-01 17:03:00        Dog  Spayed Female   \n",
       "11        12        Lala  2014-09-19 15:40:00        Dog  Intact Female   \n",
       "12        13       Chloe  2015-01-25 15:13:00        Dog  Spayed Female   \n",
       "13        14         NaN  2015-09-22 18:17:00        Dog  Neutered Male   \n",
       "14        15       Duffy  2014-07-20 19:24:00        Dog  Neutered Male   \n",
       "15        16       Jiggy  2014-03-26 18:33:00        Dog  Spayed Female   \n",
       "16        17     Ghengis  2015-05-29 17:04:00        Dog  Neutered Male   \n",
       "17        18      Decker  2016-01-19 19:31:00        Dog  Neutered Male   \n",
       "18        19    Freckles  2014-06-16 12:49:00        Cat  Spayed Female   \n",
       "19        20         NaN  2014-02-11 19:27:00        Cat  Intact Female   \n",
       "20        21     Pumpkin  2014-11-14 19:32:00        Cat  Spayed Female   \n",
       "21        22         NaN  2015-05-04 13:46:00        Cat        Unknown   \n",
       "22        23       Pluto  2014-06-07 14:14:00        Dog  Neutered Male   \n",
       "23        24         NaN  2014-07-20 15:37:00        Cat  Intact Female   \n",
       "24        25      Portia  2013-10-26 16:23:00        Dog  Spayed Female   \n",
       "25        26         NaN  2014-12-29 13:10:00        Cat  Neutered Male   \n",
       "26        27        Lucy  2015-07-23 16:18:00        Cat  Spayed Female   \n",
       "27        28         NaN  2015-03-19 08:33:00        Dog    Intact Male   \n",
       "28        29        Jack  2014-10-09 17:22:00        Dog  Neutered Male   \n",
       "29        30      Knight  2014-05-13 17:03:00        Dog  Neutered Male   \n",
       "...      ...         ...                  ...        ...            ...   \n",
       "11426  11427      Fatboy  2013-12-28 14:22:00        Dog  Neutered Male   \n",
       "11427  11428     Calypso  2013-12-21 15:55:00        Dog  Intact Female   \n",
       "11428  11429        Bolt  2013-11-15 12:32:00        Dog  Neutered Male   \n",
       "11429  11430        Mary  2014-02-17 18:04:00        Dog  Spayed Female   \n",
       "11430  11431         NaN  2014-12-08 12:40:00        Cat  Spayed Female   \n",
       "11431  11432         NaN  2015-06-30 09:00:00        Cat    Intact Male   \n",
       "11432  11433       Fiona  2014-03-21 17:33:00        Dog  Spayed Female   \n",
       "11433  11434      Evanna  2015-08-16 16:40:00        Cat  Intact Female   \n",
       "11434  11435       April  2015-09-06 12:41:00        Dog  Spayed Female   \n",
       "11435  11436         NaN  2015-08-13 17:09:00        Cat    Intact Male   \n",
       "11436  11437         NaN  2014-03-21 17:42:00        Cat  Neutered Male   \n",
       "11437  11438    Peludita  2014-01-10 13:46:00        Dog  Intact Female   \n",
       "11438  11439         NaN  2015-05-02 20:40:00        Dog  Neutered Male   \n",
       "11439  11440       Socks  2014-05-18 16:47:00        Dog  Spayed Female   \n",
       "11440  11441         NaN  2015-10-20 17:13:00        Cat        Unknown   \n",
       "11441  11442      Doogie  2015-01-07 17:41:00        Dog  Neutered Male   \n",
       "11442  11443    Sterling  2015-11-28 16:23:00        Dog  Neutered Male   \n",
       "11443  11444       Bella  2014-10-04 12:17:00        Cat  Spayed Female   \n",
       "11444  11445         NaN  2014-06-01 17:15:00        Cat  Intact Female   \n",
       "11445  11446         NaN  2015-04-14 14:34:00        Cat    Intact Male   \n",
       "11446  11447         NaN  2015-07-19 20:03:00        Cat  Neutered Male   \n",
       "11447  11448       Sadie  2014-10-19 12:21:00        Dog  Spayed Female   \n",
       "11448  11449         NaN  2014-08-17 17:48:00        Dog  Spayed Female   \n",
       "11449  11450      Caicos  2015-07-20 14:55:00        Dog  Spayed Female   \n",
       "11450  11451       Cosmo  2015-11-05 17:19:00        Dog  Neutered Male   \n",
       "11451  11452         NaN  2014-07-08 14:50:00        Cat  Neutered Male   \n",
       "11452  11453         NaN  2014-10-21 12:57:00        Cat  Intact Female   \n",
       "11453  11454         NaN  2014-09-29 09:00:00        Cat  Intact Female   \n",
       "11454  11455       Rambo  2015-09-05 17:16:00        Dog  Neutered Male   \n",
       "11455  11456       Gusto  2014-07-12 18:40:00        Dog    Intact Male   \n",
       "\n",
       "      AgeuponOutcome                                 Breed              Color  \\\n",
       "0          10 months                Labrador Retriever Mix          Red/White   \n",
       "1            2 years        German Shepherd/Siberian Husky          Black/Tan   \n",
       "2             1 year                Domestic Shorthair Mix        Brown Tabby   \n",
       "3           4 months                     Collie Smooth Mix           Tricolor   \n",
       "4            2 years                  Miniature Poodle Mix              White   \n",
       "5            3 years                            Beagle Mix        Brown/White   \n",
       "6           13 years              Domestic Medium Hair Mix  Brown Tabby/White   \n",
       "7           6 months                Domestic Shorthair Mix        Brown Tabby   \n",
       "8           3 months                         Cairn Terrier        Black/Cream   \n",
       "9             1 year                          Pit Bull Mix        Brown/White   \n",
       "10          5 months                Labrador Retriever Mix        White/Black   \n",
       "11           2 years                          Pit Bull Mix         Blue/White   \n",
       "12          6 months                          Pit Bull Mix      Brown Brindle   \n",
       "13           2 years                  Miniature Poodle Mix              White   \n",
       "14            1 year   Flat Coat Retriever/German Shepherd        Black/Brown   \n",
       "15          4 months            Pointer/Labrador Retriever              White   \n",
       "16           5 years                   Doberman Pinsch Mix          Black/Tan   \n",
       "17           3 years                       English Bulldog     White/Tricolor   \n",
       "18           3 years                Domestic Shorthair Mix             Torbie   \n",
       "19           4 years                Domestic Shorthair Mix             Torbie   \n",
       "20           2 years                Domestic Shorthair Mix              Black   \n",
       "21           3 weeks                Domestic Shorthair Mix              Black   \n",
       "22            1 year                   Norwich Terrier Mix          White/Tan   \n",
       "23            1 week                Domestic Shorthair Mix             Tortie   \n",
       "24            1 year                 Black Mouth Cur/Boxer                Tan   \n",
       "25            1 year                Domestic Shorthair Mix        Brown Tabby   \n",
       "26          2 months              Domestic Medium Hair Mix        Black Smoke   \n",
       "27          2 months                   Norwich Terrier Mix              Brown   \n",
       "28           3 years                Labrador Retriever Mix        Black/White   \n",
       "29           2 years                Labrador Retriever Mix              Black   \n",
       "...              ...                                   ...                ...   \n",
       "11426        5 years                             Boxer Mix        Brown/White   \n",
       "11427       2 months         American Pit Bull Terrier Mix      Brown Brindle   \n",
       "11428        3 years                        Siberian Husky          Red/White   \n",
       "11429        2 years                   German Shepherd Mix              Sable   \n",
       "11430       2 months                Domestic Shorthair Mix               Blue   \n",
       "11431        2 years                Domestic Shorthair Mix   Blue Tabby/White   \n",
       "11432        5 years                               Maltese              White   \n",
       "11433       2 months                Domestic Shorthair Mix        Black/White   \n",
       "11434        6 years  Cardigan Welsh Corgi/German Shepherd          Black/Tan   \n",
       "11435        3 weeks                Domestic Shorthair Mix        Brown Tabby   \n",
       "11436       5 months                Domestic Shorthair Mix        Black/White   \n",
       "11437       6 months          Chihuahua Longhair/Dachshund          White/Tan   \n",
       "11438        1 month                Labrador Retriever Mix              Sable   \n",
       "11439        2 years                          Pit Bull Mix        White/Brown   \n",
       "11440         1 week                Domestic Shorthair Mix        Brown Tabby   \n",
       "11441        4 years                          Pit Bull Mix         Blue/White   \n",
       "11442        2 years               Chihuahua Shorthair Mix          Tan/White   \n",
       "11443       6 months                 Domestic Longhair Mix             Torbie   \n",
       "11444        2 years                Domestic Shorthair Mix             Calico   \n",
       "11445        2 weeks                Domestic Shorthair Mix              White   \n",
       "11446       2 months                Domestic Shorthair Mix              Black   \n",
       "11447         1 year                Chihuahua Longhair Mix         Orange/Tan   \n",
       "11448       2 months                Labrador Retriever Mix              Brown   \n",
       "11449       6 months                    Bearded Collie Mix   Blue Merle/White   \n",
       "11450      10 months                     Staffordshire Mix          Tan/White   \n",
       "11451       2 months                Domestic Shorthair Mix              Black   \n",
       "11452        2 weeks                Domestic Shorthair Mix               Blue   \n",
       "11453         1 year                Domestic Shorthair Mix             Calico   \n",
       "11454        6 years                   German Shepherd Mix          Black/Tan   \n",
       "11455        4 years                    Labrador Retriever              Cream   \n",
       "\n",
       "       HasName  Year       ...         BernoulliSex  AvgWeight(pounds)  \\\n",
       "0          1.0  2015       ...        Intact Female               68.0   \n",
       "1          1.0  2014       ...        Spayed Female               59.0   \n",
       "2          1.0  2016       ...        Neutered Male                9.0   \n",
       "3          1.0  2013       ...          Intact Male               16.0   \n",
       "4          1.0  2015       ...        Neutered Male               14.0   \n",
       "5          1.0  2015       ...        Neutered Male               28.0   \n",
       "6          1.0  2014       ...        Neutered Male                9.0   \n",
       "7          1.0  2014       ...        Spayed Female                9.0   \n",
       "8          1.0  2014       ...        Spayed Female               14.0   \n",
       "9          1.0  2014       ...        Spayed Female               60.0   \n",
       "10         1.0  2014       ...        Spayed Female               68.0   \n",
       "11         1.0  2014       ...        Intact Female               60.0   \n",
       "12         1.0  2015       ...        Spayed Female               60.0   \n",
       "13         0.0  2015       ...        Neutered Male               14.0   \n",
       "14         1.0  2014       ...        Neutered Male               68.0   \n",
       "15         1.0  2014       ...        Spayed Female               62.0   \n",
       "16         1.0  2015       ...        Neutered Male               77.0   \n",
       "17         1.0  2016       ...        Neutered Male               53.0   \n",
       "18         1.0  2014       ...        Spayed Female                9.0   \n",
       "19         0.0  2014       ...        Intact Female                9.0   \n",
       "20         1.0  2014       ...        Spayed Female                9.0   \n",
       "21         0.0  2015       ...          Intact Male                9.0   \n",
       "22         1.0  2014       ...        Neutered Male               12.0   \n",
       "23         0.0  2014       ...        Intact Female                9.0   \n",
       "24         1.0  2013       ...        Spayed Female               79.0   \n",
       "25         0.0  2014       ...        Neutered Male                9.0   \n",
       "26         1.0  2015       ...        Spayed Female                9.0   \n",
       "27         0.0  2015       ...          Intact Male               12.0   \n",
       "28         1.0  2014       ...        Neutered Male               68.0   \n",
       "29         1.0  2014       ...        Neutered Male               68.0   \n",
       "...        ...   ...       ...                  ...                ...   \n",
       "11426      1.0  2013       ...        Neutered Male               60.0   \n",
       "11427      1.0  2013       ...        Intact Female               60.0   \n",
       "11428      1.0  2013       ...        Neutered Male               48.0   \n",
       "11429      1.0  2014       ...        Spayed Female               70.0   \n",
       "11430      0.0  2014       ...        Spayed Female                9.0   \n",
       "11431      0.0  2015       ...          Intact Male                9.0   \n",
       "11432      1.0  2014       ...        Spayed Female                6.0   \n",
       "11433      1.0  2015       ...        Intact Female                9.0   \n",
       "11434      1.0  2015       ...        Spayed Female               51.0   \n",
       "11435      0.0  2015       ...          Intact Male                9.0   \n",
       "11436      0.0  2014       ...        Neutered Male                9.0   \n",
       "11437      1.0  2014       ...        Intact Female               13.0   \n",
       "11438      0.0  2015       ...        Neutered Male               68.0   \n",
       "11439      1.0  2014       ...        Spayed Female               60.0   \n",
       "11440      0.0  2015       ...          Intact Male                9.0   \n",
       "11441      1.0  2015       ...        Neutered Male               60.0   \n",
       "11442      1.0  2015       ...        Neutered Male                6.0   \n",
       "11443      1.0  2014       ...        Spayed Female                9.0   \n",
       "11444      0.0  2014       ...        Intact Female                9.0   \n",
       "11445      0.0  2015       ...          Intact Male                9.0   \n",
       "11446      0.0  2015       ...        Neutered Male                9.0   \n",
       "11447      1.0  2014       ...        Spayed Female                6.0   \n",
       "11448      0.0  2014       ...        Spayed Female               68.0   \n",
       "11449      1.0  2015       ...        Spayed Female               50.0   \n",
       "11450      1.0  2015       ...        Neutered Male               31.0   \n",
       "11451      0.0  2014       ...        Neutered Male                9.0   \n",
       "11452      0.0  2014       ...        Intact Female                9.0   \n",
       "11453      0.0  2014       ...        Intact Female                9.0   \n",
       "11454      1.0  2015       ...        Neutered Male               70.0   \n",
       "11455      1.0  2014       ...          Intact Male               68.0   \n",
       "\n",
       "       SizeScore(1to5)  Intelligent  Friendliness  AvgAge  MixType    Hair  \\\n",
       "0                  4.0          5.0           4.0    13.0      Mix   Short   \n",
       "1                  4.0          4.0           3.0    12.0    Cross  Medium   \n",
       "2                  1.0          3.0           3.0    12.0      Mix   Short   \n",
       "3                  2.0          4.0           3.0    13.0      Mix   Short   \n",
       "4                  2.0          5.0           4.0    16.0      Mix  Medium   \n",
       "5                  2.0          2.0           4.0    15.0      Mix   Short   \n",
       "6                  1.0          3.0           3.0    12.0      Mix  Medium   \n",
       "7                  1.0          3.0           3.0    12.0      Mix   Short   \n",
       "8                  1.0          4.0           3.0    15.0     Pure   Short   \n",
       "9                  4.0          3.0           2.0    12.0      Mix   Short   \n",
       "10                 4.0          5.0           4.0    13.0      Mix   Short   \n",
       "11                 4.0          3.0           2.0    12.0      Mix    Long   \n",
       "12                 4.0          3.0           2.0    12.0      Mix   Short   \n",
       "13                 2.0          5.0           4.0    16.0      Mix   Short   \n",
       "14                 4.0          5.0           4.0    12.0    Cross   Short   \n",
       "15                 4.0          4.0           3.0    14.0    Cross   Short   \n",
       "16                 4.0          5.0           2.0    10.0      Mix   Short   \n",
       "17                 3.0          2.0           3.0     9.0     Pure   Short   \n",
       "18                 1.0          3.0           3.0    13.0      Mix   Short   \n",
       "19                 1.0          3.0           3.0    14.0      Mix   Short   \n",
       "20                 1.0          3.0           3.0    13.0      Mix   Short   \n",
       "21                 1.0          3.0           3.0    12.0      Mix   Short   \n",
       "22                 1.0          3.0           3.0    14.0      Mix  Medium   \n",
       "23                 1.0          3.0           3.0     8.0      Mix   Short   \n",
       "24                 4.0          4.0           3.0    10.0    Cross    Long   \n",
       "25                 1.0          3.0           3.0    13.0      Mix   Short   \n",
       "26                 1.0          3.0           3.0    16.0      Mix  Medium   \n",
       "27                 1.0          3.0           3.0    14.0      Mix    Long   \n",
       "28                 4.0          5.0           4.0    13.0      Mix   Short   \n",
       "29                 4.0          5.0           4.0    13.0      Mix   Short   \n",
       "...                ...          ...           ...     ...      ...     ...   \n",
       "11426              4.0          4.0           3.0     9.0      Mix   Short   \n",
       "11427              4.0          3.0           2.0    12.0      Mix  Medium   \n",
       "11428              3.0          2.0           3.0    12.0     Pure   Short   \n",
       "11429              4.0          5.0           3.0    12.0      Mix   Short   \n",
       "11430              1.0          3.0           3.0    13.0      Mix   Short   \n",
       "11431              1.0          3.0           3.0    14.0      Mix   Short   \n",
       "11432              1.0          2.0           4.0    17.0     Pure   Short   \n",
       "11433              1.0          3.0           3.0    10.0      Mix   Short   \n",
       "11434              3.0          4.0           3.0    13.0    Cross   Short   \n",
       "11435              1.0          3.0           3.0    14.0      Mix   Short   \n",
       "11436              1.0          3.0           3.0    15.0      Mix   Short   \n",
       "11437              1.0          3.0           3.0    15.0    Cross    Long   \n",
       "11438              4.0          5.0           4.0    13.0      Mix   Short   \n",
       "11439              4.0          3.0           2.0    12.0      Mix   Short   \n",
       "11440              1.0          3.0           3.0    14.0      Mix   Short   \n",
       "11441              4.0          3.0           2.0    12.0      Mix   Short   \n",
       "11442              1.0          3.0           3.0    16.0      Mix   Short   \n",
       "11443              1.0          3.0           3.0    12.0      Mix    Long   \n",
       "11444              1.0          3.0           3.0    12.0      Mix   Short   \n",
       "11445              1.0          3.0           3.0    16.0      Mix   Short   \n",
       "11446              1.0          3.0           3.0    16.0      Mix   Short   \n",
       "11447              1.0          3.0           3.0    16.0      Mix    Long   \n",
       "11448              4.0          5.0           4.0    13.0      Mix   Short   \n",
       "11449              3.0          5.0           4.0    13.0      Mix  Medium   \n",
       "11450              2.0          2.0           4.0    13.0      Mix   Short   \n",
       "11451              1.0          3.0           3.0    16.0      Mix   Short   \n",
       "11452              1.0          3.0           3.0    10.0      Mix   Short   \n",
       "11453              1.0          3.0           3.0     9.0      Mix   Short   \n",
       "11454              4.0          5.0           3.0    12.0      Mix   Short   \n",
       "11455              4.0          5.0           4.0    13.0     Pure   Short   \n",
       "\n",
       "         BlendedColor  ReducedPattern  \n",
       "0              salmon         BiColor  \n",
       "1      darkolivegreen         BiColor  \n",
       "2               brown           Tabby  \n",
       "3           Not_clear           Solid  \n",
       "4               white           Solid  \n",
       "5           rosybrown         BiColor  \n",
       "6           rosybrown           Tabby  \n",
       "7               brown           Tabby  \n",
       "8      darkolivegreen         BiColor  \n",
       "9           rosybrown         BiColor  \n",
       "10               grey         BiColor  \n",
       "11             silver         BiColor  \n",
       "12              brown        TriColor  \n",
       "13              white           Solid  \n",
       "14             maroon         BiColor  \n",
       "15              white           Solid  \n",
       "16     darkolivegreen         BiColor  \n",
       "17              white         BiColor  \n",
       "18          Not_clear        TriColor  \n",
       "19          Not_clear        TriColor  \n",
       "20              black           Solid  \n",
       "21              black           Solid  \n",
       "22              wheat         BiColor  \n",
       "23          Not_clear        TriColor  \n",
       "24                tan           Solid  \n",
       "25              brown           Tabby  \n",
       "26              black           Smoke  \n",
       "27              brown           Solid  \n",
       "28               grey         BiColor  \n",
       "29              black           Solid  \n",
       "...               ...             ...  \n",
       "11426       rosybrown         BiColor  \n",
       "11427           brown        TriColor  \n",
       "11428          salmon         BiColor  \n",
       "11429           brown           Solid  \n",
       "11430            grey           Solid  \n",
       "11431          silver           Tabby  \n",
       "11432           white           Solid  \n",
       "11433            grey         BiColor  \n",
       "11434  darkolivegreen         BiColor  \n",
       "11435           brown           Tabby  \n",
       "11436            grey         BiColor  \n",
       "11437           wheat         BiColor  \n",
       "11438           brown           Solid  \n",
       "11439       rosybrown         BiColor  \n",
       "11440           brown           Tabby  \n",
       "11441          silver         BiColor  \n",
       "11442           wheat         BiColor  \n",
       "11443       Not_clear        TriColor  \n",
       "11444       Not_clear        TriColor  \n",
       "11445           white           Solid  \n",
       "11446           black           Solid  \n",
       "11447           brown         BiColor  \n",
       "11448           brown           Solid  \n",
       "11449          silver           Merle  \n",
       "11450           wheat         BiColor  \n",
       "11451           black           Solid  \n",
       "11452            grey           Solid  \n",
       "11453       Not_clear        TriColor  \n",
       "11454  darkolivegreen         BiColor  \n",
       "11455             tan           Solid  \n",
       "\n",
       "[11456 rows x 28 columns]"
      ]
     },
     "execution_count": 85,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_df_imputed"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Categorical Encoder\n",
    "\n",
    "After all the features being extracted, we are ready to convert our categorical data into numerical data. One of the best things about the random forest model is that it **does not** require one-hot encoding for categorical data. We use [this package](http://contrib.scikit-learn.org/categorical-encoding/index.html) for categorical encoding. As one-hot increases dimension dramatically, after reading [this artical](http://www.willmcginnis.com/2015/11/29/beyond-one-hot-an-exploration-of-categorical-variables/), we decide to try \"Binary\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import category_encoders as ce"
   ]
  },
  {
   "cell_type": "raw",
   "metadata": {
    "collapsed": false
   },
   "source": [
    "for i, c in enumerate(train_df_imputed.columns):\n",
    "    print \"{}: {}\".format(i, c)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 151,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "train_df_imputed_copy = train_df_imputed.iloc[:, columns_of_interest].copy()\n",
    "test_df_imputed_copy = test_df_imputed.iloc[:, columns_of_interest].copy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 154,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "total_df = pd.concat([train_df_imputed_copy, test_df_imputed_copy], axis=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 155,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "encoder = ce.BinaryEncoder(verbose=1, cols=train_df_imputed_copy.columns.tolist(), drop_invariant=False, return_df=True, impute_missing=True, handle_unknown=\"impute\")\n",
    "encoder.fit(total_df, verbose=1)\n",
    "train_df_binary = encoder.transform(train_df_imputed_copy)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 156,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "test_df_binary = encoder.transform(test_df_imputed_copy)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 157,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "train_df_binary[\"OutcomeType\"] = train_df.OutcomeType"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 160,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "train_df_binary.to_csv(\"binary_imputed_train.csv\", index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 161,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "test_df_binary.to_csv(\"binary_imputed_test.csv\", index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 162,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "train_df_imputed.to_csv(\"imputed_train.csv\", index=False)\n",
    "test_df_imputed.to_csv(\"impute_test.csv\", index=False)"
   ]
  },
  {
   "cell_type": "raw",
   "metadata": {},
   "source": [
    "for c in test_df_binary.columns:\n",
    "    print \"{} : {}\".format(c, test_df_binary[c].isnull().sum())"
   ]
  },
  {
   "cell_type": "raw",
   "metadata": {},
   "source": [
    "data.to_csv(\"binary_categorical.csv\", index=False)"
   ]
  },
  {
   "cell_type": "raw",
   "metadata": {},
   "source": [
    "data = pd.read_csv(\"binary_categorical.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "data[\"Ordinal_Outcome\"] = pd.Categorical(animals.OutcomeType).codes"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Dealing with unbalanced data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Start training"
   ]
  },
  {
   "cell_type": "raw",
   "metadata": {},
   "source": [
    "data=pd.read_csv(\"binary_categorical_multinomial_imputed.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>type_0</th>\n",
       "      <th>HasName_0</th>\n",
       "      <th>Year_0</th>\n",
       "      <th>Year_1</th>\n",
       "      <th>Month_0</th>\n",
       "      <th>Month_1</th>\n",
       "      <th>Month_2</th>\n",
       "      <th>Month_3</th>\n",
       "      <th>Weekday_0</th>\n",
       "      <th>Weekday_1</th>\n",
       "      <th>...</th>\n",
       "      <th>AgeinDays_3</th>\n",
       "      <th>AgeinDays_4</th>\n",
       "      <th>AgeinDays_5</th>\n",
       "      <th>ReducedColor_0</th>\n",
       "      <th>ReducedColor_1</th>\n",
       "      <th>ReducedColor_2</th>\n",
       "      <th>ReducedColor_3</th>\n",
       "      <th>ReducedColor_4</th>\n",
       "      <th>ReducedColor_5</th>\n",
       "      <th>Ordinal_Outcome</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 77 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   type_0  HasName_0  Year_0  Year_1  Month_0  Month_1  Month_2  Month_3  \\\n",
       "0       0          0       0       0        0        0        0        0   \n",
       "1       1          0       0       1        0        0        0        1   \n",
       "2       0          0       1       0        0        0        1        0   \n",
       "3       1          1       0       0        0        0        1        1   \n",
       "4       0          1       0       1        0        1        0        0   \n",
       "\n",
       "   Weekday_0  Weekday_1       ...         AgeinDays_3  AgeinDays_4  \\\n",
       "0          0          0       ...                   0            0   \n",
       "1          0          0       ...                   0            0   \n",
       "2          0          1       ...                   0            0   \n",
       "3          0          1       ...                   0            1   \n",
       "4          0          1       ...                   0            0   \n",
       "\n",
       "   AgeinDays_5  ReducedColor_0  ReducedColor_1  ReducedColor_2  \\\n",
       "0            0               0               0               0   \n",
       "1            0               0               0               0   \n",
       "2            1               0               0               0   \n",
       "3            0               0               0               0   \n",
       "4            1               0               0               0   \n",
       "\n",
       "   ReducedColor_3  ReducedColor_4  ReducedColor_5  Ordinal_Outcome  \n",
       "0               0               0               0                3  \n",
       "1               0               0               1                2  \n",
       "2               0               1               0                0  \n",
       "3               0               1               1                4  \n",
       "4               0               0               1                4  \n",
       "\n",
       "[5 rows x 77 columns]"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from IPython.core import display as ICD\n",
    "from sklearn.model_selection import train_test_split, StratifiedKFold\n",
    "SPLIT_NUM = 8"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def run_skf(data, n_splits):\n",
    "    skf = StratifiedKFold(n_splits, random_state=8, shuffle=True)\n",
    "    X = data.iloc[:,3:-1]\n",
    "    y = data.iloc[:,-1]\n",
    "    train, test = [], []\n",
    "    for train_index, test_index in skf.split(X, y):\n",
    "        train.append(train_index)\n",
    "        test.append(test_index)\n",
    "    return train, test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of samples before splitting: 26729\n",
      "Number of validation samples: 2673\n",
      "Number of training samples: 21047\n",
      "\n",
      "Training sample size per fold:\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>Total</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Fold 1</th>\n",
       "      <td>8479</td>\n",
       "      <td>150</td>\n",
       "      <td>1202</td>\n",
       "      <td>3787</td>\n",
       "      <td>7429</td>\n",
       "      <td>21047</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Fold 1</th>\n",
       "      <td>8461</td>\n",
       "      <td>158</td>\n",
       "      <td>1203</td>\n",
       "      <td>3795</td>\n",
       "      <td>7430</td>\n",
       "      <td>21047</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Fold 1</th>\n",
       "      <td>8494</td>\n",
       "      <td>161</td>\n",
       "      <td>1203</td>\n",
       "      <td>3734</td>\n",
       "      <td>7456</td>\n",
       "      <td>21048</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Fold 1</th>\n",
       "      <td>8503</td>\n",
       "      <td>162</td>\n",
       "      <td>1205</td>\n",
       "      <td>3736</td>\n",
       "      <td>7442</td>\n",
       "      <td>21048</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Fold 1</th>\n",
       "      <td>8522</td>\n",
       "      <td>153</td>\n",
       "      <td>1191</td>\n",
       "      <td>3772</td>\n",
       "      <td>7411</td>\n",
       "      <td>21049</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Fold 1</th>\n",
       "      <td>8520</td>\n",
       "      <td>150</td>\n",
       "      <td>1188</td>\n",
       "      <td>3772</td>\n",
       "      <td>7421</td>\n",
       "      <td>21051</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Fold 1</th>\n",
       "      <td>8505</td>\n",
       "      <td>156</td>\n",
       "      <td>1212</td>\n",
       "      <td>3773</td>\n",
       "      <td>7405</td>\n",
       "      <td>21051</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Fold 1</th>\n",
       "      <td>8500</td>\n",
       "      <td>156</td>\n",
       "      <td>1200</td>\n",
       "      <td>3773</td>\n",
       "      <td>7422</td>\n",
       "      <td>21051</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           0    1     2     3     4  Total\n",
       "Fold 1  8479  150  1202  3787  7429  21047\n",
       "Fold 1  8461  158  1203  3795  7430  21047\n",
       "Fold 1  8494  161  1203  3734  7456  21048\n",
       "Fold 1  8503  162  1205  3736  7442  21048\n",
       "Fold 1  8522  153  1191  3772  7411  21049\n",
       "Fold 1  8520  150  1188  3772  7421  21051\n",
       "Fold 1  8505  156  1212  3773  7405  21051\n",
       "Fold 1  8500  156  1200  3773  7422  21051"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Validation samples per fold:\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>2</th>\n",
       "      <th>4</th>\n",
       "      <th>1</th>\n",
       "      <th>3</th>\n",
       "      <th>Total</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Fold 1</th>\n",
       "      <td>1233</td>\n",
       "      <td>170</td>\n",
       "      <td>1059</td>\n",
       "      <td>28</td>\n",
       "      <td>519</td>\n",
       "      <td>3009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Fold 1</th>\n",
       "      <td>1251</td>\n",
       "      <td>169</td>\n",
       "      <td>1058</td>\n",
       "      <td>20</td>\n",
       "      <td>511</td>\n",
       "      <td>3009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Fold 1</th>\n",
       "      <td>1218</td>\n",
       "      <td>169</td>\n",
       "      <td>1032</td>\n",
       "      <td>17</td>\n",
       "      <td>572</td>\n",
       "      <td>3008</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Fold 1</th>\n",
       "      <td>1209</td>\n",
       "      <td>167</td>\n",
       "      <td>1046</td>\n",
       "      <td>16</td>\n",
       "      <td>570</td>\n",
       "      <td>3008</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Fold 1</th>\n",
       "      <td>1190</td>\n",
       "      <td>181</td>\n",
       "      <td>1077</td>\n",
       "      <td>25</td>\n",
       "      <td>534</td>\n",
       "      <td>3007</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Fold 1</th>\n",
       "      <td>1192</td>\n",
       "      <td>184</td>\n",
       "      <td>1067</td>\n",
       "      <td>28</td>\n",
       "      <td>534</td>\n",
       "      <td>3005</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Fold 1</th>\n",
       "      <td>1207</td>\n",
       "      <td>160</td>\n",
       "      <td>1083</td>\n",
       "      <td>22</td>\n",
       "      <td>533</td>\n",
       "      <td>3005</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Fold 1</th>\n",
       "      <td>1212</td>\n",
       "      <td>172</td>\n",
       "      <td>1066</td>\n",
       "      <td>22</td>\n",
       "      <td>533</td>\n",
       "      <td>3005</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           0    2     4   1    3  Total\n",
       "Fold 1  1233  170  1059  28  519   3009\n",
       "Fold 1  1251  169  1058  20  511   3009\n",
       "Fold 1  1218  169  1032  17  572   3008\n",
       "Fold 1  1209  167  1046  16  570   3008\n",
       "Fold 1  1190  181  1077  25  534   3007\n",
       "Fold 1  1192  184  1067  28  534   3005\n",
       "Fold 1  1207  160  1083  22  533   3005\n",
       "Fold 1  1212  172  1066  22  533   3005"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "train_total, test = train_test_split(data, test_size=.1, random_state=8)\n",
    "print 'Number of samples before splitting: {}'.format(len(animals))\n",
    "print 'Number of validation samples: {}'.format(len(test))\n",
    "train, validation = run_skf(train_total, n_splits=SPLIT_NUM)\n",
    "visualize_folds(data, train, validation)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.metrics import confusion_matrix, classification_report, log_loss"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import itertools\n",
    "def plot_confusion_matrix(cm, classes,\n",
    "                          normalize=False,\n",
    "                          title='Confusion matrix',\n",
    "                          cmap=plt.cm.Blues):\n",
    "    \"\"\"\n",
    "    This function prints and plots the confusion matrix.\n",
    "    Normalization can be applied by setting `normalize=True`.\n",
    "    \"\"\"\n",
    "    if normalize:\n",
    "        cm = cm.astype('float') / cm.sum(axis=1)[:, np.newaxis]\n",
    "\n",
    "    figure()\n",
    "    plt.imshow(cm, interpolation='nearest', cmap=cmap)\n",
    "    plt.title(title)\n",
    "    plt.colorbar()\n",
    "    tick_marks = np.arange(len(classes))\n",
    "    plt.xticks(tick_marks, classes, rotation=45)\n",
    "    plt.yticks(tick_marks, classes)\n",
    "\n",
    "    fmt = '.2f' if normalize else 'd'\n",
    "    thresh = cm.max() / 2.\n",
    "    for i, j in itertools.product(range(cm.shape[0]), range(cm.shape[1])):\n",
    "        plt.text(j, i, format(cm[i, j], fmt),\n",
    "                 horizontalalignment=\"center\",\n",
    "                 color=\"white\" if cm[i, j] > thresh else \"black\")\n",
    "\n",
    "    plt.tight_layout()\n",
    "    plt.ylabel('True label')\n",
    "    plt.xlabel('Predicted label')\n",
    "    plt.xticks(rotation=90)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "freq = data.iloc[train[0]].Ordinal_Outcome.value_counts(normalize=True, sort=False).as_matrix()\n",
    "\n",
    "# weights = log(1./freq)\n",
    "weights = 1./freq\n",
    "weights = dict(enumerate(weights))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "metadata": {
    "collapsed": false,
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train: 0.182389185975, Validation: 0.920987863861\n",
      "Train: 0.182917640263, Validation: 0.91428733643\n",
      "Train: 0.182103537692, Validation: 0.956093900592\n",
      "Train: 0.182399848647, Validation: 0.860003685781\n",
      "Train: 0.182874781124, Validation: 0.927532178393\n",
      "Train: 0.182289807701, Validation: 0.888269792245\n",
      "Train: 0.183079447271, Validation: 0.927373259066\n",
      "Train: 0.182975667534, Validation: 0.891860066914\n",
      "Avg Train: 0.182628739526, Avg Validation: 0.91080101041\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAV0AAAE3CAYAAAAJ55KLAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xcba8748>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAWMAAAE3CAYAAAB/3SAAAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xe915a20>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAWMAAAE3CAYAAAB/3SAAAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xcbfec88>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAV0AAAE3CAYAAAAJ55KLAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nfwHzFyEw5xNJCgAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xe222f28>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAV0AAAE3CAYAAAAJ55KLAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xd2d02e8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAV0AAAE3CAYAAAAJ55KLAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xee16390>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAV0AAAE3CAYAAAAJ55KLAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n5AAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xdad4518>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAV0AAAE3CAYAAAAJ55KLAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xe2a50b8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "hard_set = set()\n",
    "train_loss = 0\n",
    "validation_loss = 0\n",
    "for i in range(SPLIT_NUM):\n",
    "    clf = RandomForestClassifier(n_estimators=100,\n",
    "                                 max_depth=100,\n",
    "                                 random_state=8,\n",
    "                                 n_jobs = 20,\n",
    "                                 # class_weight=weights,\n",
    "                                 max_features=None)  \n",
    "    clf.fit(data.iloc[train[i], :-1], data.iloc[train[i],-1])\n",
    "\n",
    "    pred_train = clf.predict_proba(data.iloc[train[i], :-1])\n",
    "    pred_validation = clf.predict_proba(data.iloc[validation[i], :-1])\n",
    "    pred_train_labels = pred_train.argmax(axis=1)\n",
    "    pred_validation_labels = pred_validation.argmax(axis=1)\n",
    "    train_true = pd.get_dummies(data.iloc[train[i],-1]).as_matrix()\n",
    "    validation_true = pd.get_dummies(data.iloc[validation[i], -1]).as_matrix()\n",
    "    \n",
    "\n",
    "    hard_set.update(set(train[i][index] for index in where(pred_train_labels ^ data.iloc[train[i],-1])[0]))\n",
    "    hard_set.update(set(validation[i][index] for index in where(pred_validation_labels ^ data.iloc[validation[i],-1])[0]))\n",
    "    \n",
    "    log_loss_train = log_loss(train_true, pred_train)\n",
    "    log_loss_validation = log_loss(validation_true, pred_validation)\n",
    "    \n",
    "    train_loss += log_loss_train\n",
    "    validation_loss += log_loss_validation\n",
    "    \n",
    "    cm = confusion_matrix(data.iloc[validation[i], -1], pred_validation.argmax(axis=1))\n",
    "    plot_confusion_matrix(cm, sorted(animals.OutcomeType.unique()))\n",
    "    print \"Train: {}, Validation: {}\".format(log_loss_train, log_loss_validation)\n",
    "print \"Avg Train: {}, Avg Validation: {}\".format(train_loss/SPLIT_NUM, validation_loss/SPLIT_NUM)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To begin with, the classifier did poorly on Died and Euthanasia"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "hard_df = animals.iloc[list(hard_set)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from scipy import interpolate "
   ]
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python [default]",
   "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": "2.7.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}