10_DataEngineering.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<div style=float:right><img src=\"assets/img/appliedai-logo.png\" width=100 style=\"margin: 0px 20px\"></img></div>\n",
    "\n",
    "\n",
    "##### Jonathan Sedar Personal Project\n",
    "## PyMC3 vs PyStan Comparison\n",
    "_Spring 2016_\n",
    "\n",
    "This set of Notebooks and scripts comprise the **pymc3_vs_pystan** personal project by Jonathan Sedar of Applied AI Ltd, written primarily for presentation at the PyData London 2016 Conference.\n",
    "\n",
    "The project demonstrates hierarchical linear regression using two Bayesian inference frameworks: PyMC3 and PyStan. The project borrows heavily from code written for Applied AI Ltd and is supplied here for educational purposes only. No copyright or license is extended to users.\n",
    "\n",
    "    \n",
    "# 10_DataEngineering\n",
    "\n",
    "#### Acquisition, Basic Cleaning, Feature Creation\n",
    "    \n",
    "    \n",
    "    \n",
    "+ [Setup](#Setup)\n",
    "    + [Local Functions](#Local-Functions)\n",
    "    + [Load Raw Data](#Load-Raw-Data)\n",
    "\n",
    "\n",
    "+ [Prepare Data](#Prepare-Data)\n",
    "    + [Cleaning and Exclusions](#Cleaning-and-Exclusions)\n",
    "    + [Feature Creation](#Feature-Creation)\n",
    "    + [Check size of data of interest](#Check-size-of-data-of-interest)\n",
    "    + [Save to store](#Save-to-store)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Setup"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "## Interactive magics\n",
    "%matplotlib inline\n",
    "%qtconsole --colors=linux"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# general packages\n",
    "import string\n",
    "import regex as re\n",
    "import requests\n",
    "import sqlite3\n",
    "from zipfile import ZipFile\n",
    "from io import BytesIO\n",
    "from convenience_functions import *\n",
    "from ipywidgets import interactive, fixed\n",
    "\n",
    "# scientific packages\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "sns.set(style=\"darkgrid\", palette=\"muted\")\n",
    "# plt.style.use('ggplot')\n",
    "plt.rcParams['figure.figsize'] = 12, 6\n",
    "np.random.seed(0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Local Functions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "def get_cardata(sqlcnx, tblnm='raw'):\n",
    "    \"\"\" Extract car data from local sqlite db or go download it \n",
    "        Deals with zipfiles nicely\n",
    "    \"\"\"\n",
    "\n",
    "    zip_url = 'http://carfueldata.direct.gov.uk/additional/aug2015/'\n",
    "    zip_fn = 'download-data-for-Aug-2015-Euro-6.zip'\n",
    "    \n",
    "    csr = cnxsql.cursor()\n",
    "    tblexist = csr.execute((\"select * from sqlite_master \" +\n",
    "                           \"where type = 'table' and name = ?\"), (tblnm,)).fetchall()\n",
    "    \n",
    "    if len(tblexist) == 0:  # go get info from interwebs\n",
    "        try:         \n",
    "            r = requests.get(zip_url + zip_fn)\n",
    "            f = ZipFile(BytesIO(r.content), mode='r')\n",
    "            fn = f.namelist()[0]\n",
    "            dfraw = pd.read_csv(f.open(fn), encoding='cp1252', sep=',') # <- Excel encoding!   \n",
    "            dfraw.rename(columns=lambda x: snakey_lowercase(x), inplace=True)\n",
    "            dfraw.drop([c for c in dfraw.columns if c[:3] == 'unn'], inplace=True, axis=1)\n",
    "            dfraw.to_sql('raw', cnxsql, if_exists='replace', index=None)\n",
    "            print('Retrived dataset from web and stored in db')\n",
    "        \n",
    "        except Exception as e:\n",
    "            raise e     # TODO make error better\n",
    "    else:\n",
    "        dfraw = pd.read_sql('select * from raw', cnxsql, index_col=None)\n",
    "        print('Used dataset from existing db')\n",
    "       \n",
    "    return dfraw\n",
    "\n",
    "\n",
    "def clean_transmission(x):\n",
    "    \"\"\" extract info from transmission: manual / auto / semiauto \"\"\"\n",
    "    try:\n",
    "        s = re.sub('^[0-9]{1}','', x.lower())\n",
    "    except:\n",
    "        return ''\n",
    "       \n",
    "    if (s[0] == 'a') | (s[:2] == 'qa'):\n",
    "        return 'auto'\n",
    "    \n",
    "    elif s[:3] =='mul':\n",
    "        return 'auto'\n",
    "    \n",
    "    elif (s[0] == 'm') | (s[:2] == 'qm'):\n",
    "        if len(s) == 1:\n",
    "            print(x)\n",
    "        if s[1] != 'u':\n",
    "            return 'manual'\n",
    "        \n",
    "    elif (s[0] == 'd') | (s[:2] == 'qd') | \\\n",
    "        (re.search('ssg$',s) != None) | (re.search('cvt$',s) != None) | \\\n",
    "        (re.search('sat[0-9]',s) != None):\n",
    "        return 'semiauto'\n",
    "    \n",
    "    else:\n",
    "        return 'unknown'\n",
    "    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load Raw Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "cnxsql = sqlite3.connect('data/car_emissions.db')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Retrived dataset from web and stored in db\n",
      "(2666, 27)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>2607</th>\n",
       "      <th>1653</th>\n",
       "      <th>835</th>\n",
       "      <th>count</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>min</th>\n",
       "      <th>25%</th>\n",
       "      <th>50%</th>\n",
       "      <th>75%</th>\n",
       "      <th>max</th>\n",
       "      <th>dtype</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>manufacturer</th>\n",
       "      <td>VOLVO</td>\n",
       "      <td>MERCEDES-BENZ</td>\n",
       "      <td>BMW</td>\n",
       "      <td>2666</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>object</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>model</th>\n",
       "      <td>V40, MY16</td>\n",
       "      <td>C-Class Saloon, Model Year 2015</td>\n",
       "      <td>7 Series F01/F02</td>\n",
       "      <td>2666</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>object</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>description</th>\n",
       "      <td>T3</td>\n",
       "      <td>C 200 with 18\" rear wheels</td>\n",
       "      <td>730d Saloon</td>\n",
       "      <td>2666</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>object</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>transmission</th>\n",
       "      <td>6AT</td>\n",
       "      <td>M6</td>\n",
       "      <td>A8</td>\n",
       "      <td>2657</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>object</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>engine_capacity</th>\n",
       "      <td>1498</td>\n",
       "      <td>1991</td>\n",
       "      <td>2993</td>\n",
       "      <td>2659</td>\n",
       "      <td>2072.17</td>\n",
       "      <td>902.38</td>\n",
       "      <td>395.0</td>\n",
       "      <td>1560.00</td>\n",
       "      <td>1984.0</td>\n",
       "      <td>1999.00</td>\n",
       "      <td>6752.0</td>\n",
       "      <td>float64</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fuel_type</th>\n",
       "      <td>Petrol</td>\n",
       "      <td>Petrol</td>\n",
       "      <td>Diesel</td>\n",
       "      <td>2666</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>object</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>electric_energy_consumption_mileskwh</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>19</td>\n",
       "      <td>4.96</td>\n",
       "      <td>1.95</td>\n",
       "      <td>3.4</td>\n",
       "      <td>3.95</td>\n",
       "      <td>4.6</td>\n",
       "      <td>5.45</td>\n",
       "      <td>12.4</td>\n",
       "      <td>float64</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>whkm</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>19</td>\n",
       "      <td>135.21</td>\n",
       "      <td>31.38</td>\n",
       "      <td>50.0</td>\n",
       "      <td>114.00</td>\n",
       "      <td>135.0</td>\n",
       "      <td>156.50</td>\n",
       "      <td>182.0</td>\n",
       "      <td>float64</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>maximum_range_km</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>19</td>\n",
       "      <td>103.05</td>\n",
       "      <td>73.69</td>\n",
       "      <td>31.0</td>\n",
       "      <td>34.50</td>\n",
       "      <td>53.0</td>\n",
       "      <td>172.50</td>\n",
       "      <td>212.0</td>\n",
       "      <td>float64</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>maximum_range_miles</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>19</td>\n",
       "      <td>63.89</td>\n",
       "      <td>45.90</td>\n",
       "      <td>19.0</td>\n",
       "      <td>21.50</td>\n",
       "      <td>33.0</td>\n",
       "      <td>107.50</td>\n",
       "      <td>132.0</td>\n",
       "      <td>float64</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>metric_urban_cold</th>\n",
       "      <td>7.3</td>\n",
       "      <td>7.1</td>\n",
       "      <td>6.5</td>\n",
       "      <td>2647</td>\n",
       "      <td>7.10</td>\n",
       "      <td>3.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5.20</td>\n",
       "      <td>6.2</td>\n",
       "      <td>8.10</td>\n",
       "      <td>24.7</td>\n",
       "      <td>float64</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>metric_extra_urban</th>\n",
       "      <td>4.5</td>\n",
       "      <td>4.6</td>\n",
       "      <td>5.1</td>\n",
       "      <td>2647</td>\n",
       "      <td>4.85</td>\n",
       "      <td>2.35</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4.00</td>\n",
       "      <td>4.5</td>\n",
       "      <td>5.30</td>\n",
       "      <td>107.0</td>\n",
       "      <td>float64</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>metric_combined</th>\n",
       "      <td>5.5</td>\n",
       "      <td>5.5</td>\n",
       "      <td>5.6</td>\n",
       "      <td>2659</td>\n",
       "      <td>5.68</td>\n",
       "      <td>2.73</td>\n",
       "      <td>0.6</td>\n",
       "      <td>4.40</td>\n",
       "      <td>5.1</td>\n",
       "      <td>6.30</td>\n",
       "      <td>108.0</td>\n",
       "      <td>float64</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>imperial_urban_cold</th>\n",
       "      <td>38.7</td>\n",
       "      <td>39.8</td>\n",
       "      <td>43.5</td>\n",
       "      <td>2647</td>\n",
       "      <td>44.72</td>\n",
       "      <td>13.30</td>\n",
       "      <td>0.0</td>\n",
       "      <td>34.90</td>\n",
       "      <td>45.6</td>\n",
       "      <td>54.30</td>\n",
       "      <td>91.1</td>\n",
       "      <td>float64</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>imperial_extra_urban</th>\n",
       "      <td>62.8</td>\n",
       "      <td>61.4</td>\n",
       "      <td>55.4</td>\n",
       "      <td>2647</td>\n",
       "      <td>61.87</td>\n",
       "      <td>13.18</td>\n",
       "      <td>0.0</td>\n",
       "      <td>53.30</td>\n",
       "      <td>62.8</td>\n",
       "      <td>70.60</td>\n",
       "      <td>108.6</td>\n",
       "      <td>float64</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>imperial_combined</th>\n",
       "      <td>51.4</td>\n",
       "      <td>51.4</td>\n",
       "      <td>50.4</td>\n",
       "      <td>2659</td>\n",
       "      <td>54.52</td>\n",
       "      <td>16.43</td>\n",
       "      <td>17.6</td>\n",
       "      <td>44.80</td>\n",
       "      <td>55.4</td>\n",
       "      <td>64.20</td>\n",
       "      <td>470.8</td>\n",
       "      <td>float64</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>co2_gkm</th>\n",
       "      <td>129</td>\n",
       "      <td>128</td>\n",
       "      <td>148</td>\n",
       "      <td>2666</td>\n",
       "      <td>137.41</td>\n",
       "      <td>41.98</td>\n",
       "      <td>0.0</td>\n",
       "      <td>112.00</td>\n",
       "      <td>128.0</td>\n",
       "      <td>151.00</td>\n",
       "      <td>370.0</td>\n",
       "      <td>int64</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fuel_cost_12000_miles</th>\n",
       "      <td>£1,189</td>\n",
       "      <td>£1,189</td>\n",
       "      <td>£1,277</td>\n",
       "      <td>2630</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>object</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>electricity_cost</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>22</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>object</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>total_cost_12000_miles</th>\n",
       "      <td>£1,189</td>\n",
       "      <td>£1,189</td>\n",
       "      <td>£1,277</td>\n",
       "      <td>2633</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>object</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                           2607  \\\n",
       "manufacturer                              VOLVO   \n",
       "model                                 V40, MY16   \n",
       "description                                  T3   \n",
       "transmission                                6AT   \n",
       "engine_capacity                            1498   \n",
       "fuel_type                                Petrol   \n",
       "electric_energy_consumption_mileskwh        NaN   \n",
       "whkm                                        NaN   \n",
       "maximum_range_km                            NaN   \n",
       "maximum_range_miles                         NaN   \n",
       "metric_urban_cold                           7.3   \n",
       "metric_extra_urban                          4.5   \n",
       "metric_combined                             5.5   \n",
       "imperial_urban_cold                        38.7   \n",
       "imperial_extra_urban                       62.8   \n",
       "imperial_combined                          51.4   \n",
       "co2_gkm                                     129   \n",
       "fuel_cost_12000_miles                    £1,189   \n",
       "electricity_cost                            NaN   \n",
       "total_cost_12000_miles                   £1,189   \n",
       "\n",
       "                                                                 1653  \\\n",
       "manufacturer                                            MERCEDES-BENZ   \n",
       "model                                 C-Class Saloon, Model Year 2015   \n",
       "description                                C 200 with 18\" rear wheels   \n",
       "transmission                                                       M6   \n",
       "engine_capacity                                                  1991   \n",
       "fuel_type                                                      Petrol   \n",
       "electric_energy_consumption_mileskwh                              NaN   \n",
       "whkm                                                              NaN   \n",
       "maximum_range_km                                                  NaN   \n",
       "maximum_range_miles                                               NaN   \n",
       "metric_urban_cold                                                 7.1   \n",
       "metric_extra_urban                                                4.6   \n",
       "metric_combined                                                   5.5   \n",
       "imperial_urban_cold                                              39.8   \n",
       "imperial_extra_urban                                             61.4   \n",
       "imperial_combined                                                51.4   \n",
       "co2_gkm                                                           128   \n",
       "fuel_cost_12000_miles                                          £1,189   \n",
       "electricity_cost                                                  NaN   \n",
       "total_cost_12000_miles                                         £1,189   \n",
       "\n",
       "                                                   835  count     mean  \\\n",
       "manufacturer                                       BMW   2666      NaN   \n",
       "model                                 7 Series F01/F02   2666      NaN   \n",
       "description                                730d Saloon   2666      NaN   \n",
       "transmission                                        A8   2657      NaN   \n",
       "engine_capacity                                   2993   2659  2072.17   \n",
       "fuel_type                                       Diesel   2666      NaN   \n",
       "electric_energy_consumption_mileskwh               NaN     19     4.96   \n",
       "whkm                                               NaN     19   135.21   \n",
       "maximum_range_km                                   NaN     19   103.05   \n",
       "maximum_range_miles                                NaN     19    63.89   \n",
       "metric_urban_cold                                  6.5   2647     7.10   \n",
       "metric_extra_urban                                 5.1   2647     4.85   \n",
       "metric_combined                                    5.6   2659     5.68   \n",
       "imperial_urban_cold                               43.5   2647    44.72   \n",
       "imperial_extra_urban                              55.4   2647    61.87   \n",
       "imperial_combined                                 50.4   2659    54.52   \n",
       "co2_gkm                                            148   2666   137.41   \n",
       "fuel_cost_12000_miles                           £1,277   2630      NaN   \n",
       "electricity_cost                                   NaN     22      NaN   \n",
       "total_cost_12000_miles                          £1,277   2633      NaN   \n",
       "\n",
       "                                         std    min      25%     50%      75%  \\\n",
       "manufacturer                             NaN    NaN      NaN     NaN      NaN   \n",
       "model                                    NaN    NaN      NaN     NaN      NaN   \n",
       "description                              NaN    NaN      NaN     NaN      NaN   \n",
       "transmission                             NaN    NaN      NaN     NaN      NaN   \n",
       "engine_capacity                       902.38  395.0  1560.00  1984.0  1999.00   \n",
       "fuel_type                                NaN    NaN      NaN     NaN      NaN   \n",
       "electric_energy_consumption_mileskwh    1.95    3.4     3.95     4.6     5.45   \n",
       "whkm                                   31.38   50.0   114.00   135.0   156.50   \n",
       "maximum_range_km                       73.69   31.0    34.50    53.0   172.50   \n",
       "maximum_range_miles                    45.90   19.0    21.50    33.0   107.50   \n",
       "metric_urban_cold                       3.00    0.0     5.20     6.2     8.10   \n",
       "metric_extra_urban                      2.35    0.0     4.00     4.5     5.30   \n",
       "metric_combined                         2.73    0.6     4.40     5.1     6.30   \n",
       "imperial_urban_cold                    13.30    0.0    34.90    45.6    54.30   \n",
       "imperial_extra_urban                   13.18    0.0    53.30    62.8    70.60   \n",
       "imperial_combined                      16.43   17.6    44.80    55.4    64.20   \n",
       "co2_gkm                                41.98    0.0   112.00   128.0   151.00   \n",
       "fuel_cost_12000_miles                    NaN    NaN      NaN     NaN      NaN   \n",
       "electricity_cost                         NaN    NaN      NaN     NaN      NaN   \n",
       "total_cost_12000_miles                   NaN    NaN      NaN     NaN      NaN   \n",
       "\n",
       "                                         max    dtype  \n",
       "manufacturer                             NaN   object  \n",
       "model                                    NaN   object  \n",
       "description                              NaN   object  \n",
       "transmission                             NaN   object  \n",
       "engine_capacity                       6752.0  float64  \n",
       "fuel_type                                NaN   object  \n",
       "electric_energy_consumption_mileskwh    12.4  float64  \n",
       "whkm                                   182.0  float64  \n",
       "maximum_range_km                       212.0  float64  \n",
       "maximum_range_miles                    132.0  float64  \n",
       "metric_urban_cold                       24.7  float64  \n",
       "metric_extra_urban                     107.0  float64  \n",
       "metric_combined                        108.0  float64  \n",
       "imperial_urban_cold                     91.1  float64  \n",
       "imperial_extra_urban                   108.6  float64  \n",
       "imperial_combined                      470.8  float64  \n",
       "co2_gkm                                370.0    int64  \n",
       "fuel_cost_12000_miles                    NaN   object  \n",
       "electricity_cost                         NaN   object  \n",
       "total_cost_12000_miles                   NaN   object  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dfraw = get_cardata(cnxsql)\n",
    "custom_describe(dfraw)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Prepare Data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Cleaning and Exclusions"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Rename cols"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "dfraw.rename(columns={'manufacturer':'mfr', 'description':'desc'\n",
    "                     ,'transmission':'trans'}, inplace=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Select features of interest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "ft_cat = ['mfr','model','desc','trans','fuel_type']\n",
    "ft_num_shared = ['metric_combined', 'metric_extra_urban', 'metric_urban_cold'\n",
    "                 ,'engine_capacity', 'emissions_co_mgkm', 'emissions_nox_mgkm']\n",
    "ft_num_diesel = ['particulates_no_mgkm', 'thc_nox_emissions_mgkm']\n",
    "ft_num_petrol = ['thc_emissions_mgkm']     "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(2666, 14)"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = dfraw[ft_cat + ft_num_shared +  ft_num_diesel + ft_num_petrol].copy()\n",
    "df.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Drop rows containing nulls in shared features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "mfr (2666, 14)\n",
      "model (2666, 14)\n",
      "desc (2666, 14)\n",
      "trans (2657, 14)\n",
      "fuel_type (2657, 14)\n",
      "metric_combined (2657, 14)\n",
      "metric_extra_urban (2647, 14)\n",
      "metric_urban_cold (2647, 14)\n",
      "engine_capacity (2647, 14)\n",
      "emissions_co_mgkm (2647, 14)\n",
      "emissions_nox_mgkm (2647, 14)\n"
     ]
    }
   ],
   "source": [
    "for ft in ft_cat + ft_num_shared:\n",
    "    df = df.loc[~pd.isnull(df[ft])]\n",
    "    print('{} {}'.format(ft, df.shape))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Clean strings"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "for col in ['mfr','model','desc','fuel_type']:\n",
    "    df[col] = df[col].apply(lambda x: x.lower())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Change dtypes and remove junk"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(2646, 14)\n"
     ]
    }
   ],
   "source": [
    "# some cols contain numerics with comma decimals, European-style. Remove rows: can't trust.\n",
    "\n",
    "colswcommadecimal = ['emissions_nox_mgkm','thc_nox_emissions_mgkm','particulates_no_mgkm']\n",
    "\n",
    "for col in colswcommadecimal:\n",
    "    df = df.loc[df[col].apply(lambda x: re.search('[0-9]\\,[0-9]',str(x)) == None)]\n",
    "\n",
    "print(df.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# a couple of numeric cols still report as object type, coerce to numeric\n",
    "\n",
    "for col in ['emissions_nox_mgkm', 'particulates_no_mgkm','thc_nox_emissions_mgkm']:\n",
    "    df[col] = df[col].astype(np.float)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Remove duplicate rows"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "df = df.drop_duplicates()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": false,
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(2643, 14)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>763</th>\n",
       "      <th>1741</th>\n",
       "      <th>1036</th>\n",
       "      <th>count</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>min</th>\n",
       "      <th>25%</th>\n",
       "      <th>50%</th>\n",
       "      <th>75%</th>\n",
       "      <th>max</th>\n",
       "      <th>dtype</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>mfr</th>\n",
       "      <td>bmw</td>\n",
       "      <td>mercedes-benz</td>\n",
       "      <td>citroen</td>\n",
       "      <td>2643</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>abarth</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>volvo</td>\n",
       "      <td>object</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>model</th>\n",
       "      <td>5 series f10/f11, from march 2010</td>\n",
       "      <td>e-class coupé, model year 2015</td>\n",
       "      <td>ds4</td>\n",
       "      <td>2643</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1 seri</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>zafira</td>\n",
       "      <td>object</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>desc</th>\n",
       "      <td>520i touring - 245/40 r19 tyres front &amp; rear</td>\n",
       "      <td>e 220 bluetec with 17\" rear wheels</td>\n",
       "      <td>bluehdi 150 s&amp;s (18'' wheels)</td>\n",
       "      <td>2643</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>'1 air</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>zc 1.6</td>\n",
       "      <td>object</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>trans</th>\n",
       "      <td>A8</td>\n",
       "      <td>A7</td>\n",
       "      <td>M6</td>\n",
       "      <td>2643</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5MT</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>SAT6</td>\n",
       "      <td>object</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fuel_type</th>\n",
       "      <td>petrol</td>\n",
       "      <td>diesel</td>\n",
       "      <td>diesel</td>\n",
       "      <td>2643</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>diesel</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>petrol</td>\n",
       "      <td>object</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>metric_combined</th>\n",
       "      <td>6.7</td>\n",
       "      <td>4.7</td>\n",
       "      <td>3.9</td>\n",
       "      <td>2643</td>\n",
       "      <td>5.65</td>\n",
       "      <td>1.87</td>\n",
       "      <td>3.1</td>\n",
       "      <td>4.4</td>\n",
       "      <td>5.1</td>\n",
       "      <td>6.3</td>\n",
       "      <td>16</td>\n",
       "      <td>float64</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>metric_extra_urban</th>\n",
       "      <td>5.3</td>\n",
       "      <td>4</td>\n",
       "      <td>3.6</td>\n",
       "      <td>2643</td>\n",
       "      <td>4.85</td>\n",
       "      <td>2.35</td>\n",
       "      <td>0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.5</td>\n",
       "      <td>5.3</td>\n",
       "      <td>107</td>\n",
       "      <td>float64</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>metric_urban_cold</th>\n",
       "      <td>9.1</td>\n",
       "      <td>5.8</td>\n",
       "      <td>4.4</td>\n",
       "      <td>2643</td>\n",
       "      <td>7.10</td>\n",
       "      <td>3.00</td>\n",
       "      <td>0</td>\n",
       "      <td>5.2</td>\n",
       "      <td>6.2</td>\n",
       "      <td>8.1</td>\n",
       "      <td>24.7</td>\n",
       "      <td>float64</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>engine_capacity</th>\n",
       "      <td>1997</td>\n",
       "      <td>2143</td>\n",
       "      <td>1997</td>\n",
       "      <td>2643</td>\n",
       "      <td>2073.12</td>\n",
       "      <td>903.98</td>\n",
       "      <td>395</td>\n",
       "      <td>1560.0</td>\n",
       "      <td>1984.0</td>\n",
       "      <td>1999.0</td>\n",
       "      <td>6752</td>\n",
       "      <td>float64</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>emissions_co_mgkm</th>\n",
       "      <td>285</td>\n",
       "      <td>143</td>\n",
       "      <td>292</td>\n",
       "      <td>2643</td>\n",
       "      <td>268.21</td>\n",
       "      <td>140.57</td>\n",
       "      <td>30</td>\n",
       "      <td>162.0</td>\n",
       "      <td>243.0</td>\n",
       "      <td>341.0</td>\n",
       "      <td>913</td>\n",
       "      <td>float64</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>emissions_nox_mgkm</th>\n",
       "      <td>32</td>\n",
       "      <td>43</td>\n",
       "      <td>51</td>\n",
       "      <td>2643</td>\n",
       "      <td>37.40</td>\n",
       "      <td>18.63</td>\n",
       "      <td>1</td>\n",
       "      <td>23.0</td>\n",
       "      <td>35.0</td>\n",
       "      <td>52.0</td>\n",
       "      <td>158</td>\n",
       "      <td>float64</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>particulates_no_mgkm</th>\n",
       "      <td>NaN</td>\n",
       "      <td>1.1</td>\n",
       "      <td>0.2</td>\n",
       "      <td>1306</td>\n",
       "      <td>0.35</td>\n",
       "      <td>0.35</td>\n",
       "      <td>0</td>\n",
       "      <td>0.1</td>\n",
       "      <td>0.2</td>\n",
       "      <td>0.5</td>\n",
       "      <td>2</td>\n",
       "      <td>float64</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>thc_nox_emissions_mgkm</th>\n",
       "      <td>NaN</td>\n",
       "      <td>58</td>\n",
       "      <td>70</td>\n",
       "      <td>1316</td>\n",
       "      <td>78.20</td>\n",
       "      <td>23.33</td>\n",
       "      <td>11</td>\n",
       "      <td>64.0</td>\n",
       "      <td>78.0</td>\n",
       "      <td>90.0</td>\n",
       "      <td>184</td>\n",
       "      <td>float64</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>thc_emissions_mgkm</th>\n",
       "      <td>26</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1350</td>\n",
       "      <td>42.89</td>\n",
       "      <td>11.89</td>\n",
       "      <td>7</td>\n",
       "      <td>35.0</td>\n",
       "      <td>43.0</td>\n",
       "      <td>51.0</td>\n",
       "      <td>95</td>\n",
       "      <td>float64</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                                 763  \\\n",
       "mfr                                                              bmw   \n",
       "model                              5 series f10/f11, from march 2010   \n",
       "desc                    520i touring - 245/40 r19 tyres front & rear   \n",
       "trans                                                             A8   \n",
       "fuel_type                                                     petrol   \n",
       "metric_combined                                                  6.7   \n",
       "metric_extra_urban                                               5.3   \n",
       "metric_urban_cold                                                9.1   \n",
       "engine_capacity                                                 1997   \n",
       "emissions_co_mgkm                                                285   \n",
       "emissions_nox_mgkm                                                32   \n",
       "particulates_no_mgkm                                             NaN   \n",
       "thc_nox_emissions_mgkm                                           NaN   \n",
       "thc_emissions_mgkm                                                26   \n",
       "\n",
       "                                                      1741  \\\n",
       "mfr                                          mercedes-benz   \n",
       "model                       e-class coupé, model year 2015   \n",
       "desc                    e 220 bluetec with 17\" rear wheels   \n",
       "trans                                                   A7   \n",
       "fuel_type                                           diesel   \n",
       "metric_combined                                        4.7   \n",
       "metric_extra_urban                                       4   \n",
       "metric_urban_cold                                      5.8   \n",
       "engine_capacity                                       2143   \n",
       "emissions_co_mgkm                                      143   \n",
       "emissions_nox_mgkm                                      43   \n",
       "particulates_no_mgkm                                   1.1   \n",
       "thc_nox_emissions_mgkm                                  58   \n",
       "thc_emissions_mgkm                                     NaN   \n",
       "\n",
       "                                                 1036  count     mean     std  \\\n",
       "mfr                                           citroen   2643      NaN     NaN   \n",
       "model                                             ds4   2643      NaN     NaN   \n",
       "desc                    bluehdi 150 s&s (18'' wheels)   2643      NaN     NaN   \n",
       "trans                                              M6   2643      NaN     NaN   \n",
       "fuel_type                                      diesel   2643      NaN     NaN   \n",
       "metric_combined                                   3.9   2643     5.65    1.87   \n",
       "metric_extra_urban                                3.6   2643     4.85    2.35   \n",
       "metric_urban_cold                                 4.4   2643     7.10    3.00   \n",
       "engine_capacity                                  1997   2643  2073.12  903.98   \n",
       "emissions_co_mgkm                                 292   2643   268.21  140.57   \n",
       "emissions_nox_mgkm                                 51   2643    37.40   18.63   \n",
       "particulates_no_mgkm                              0.2   1306     0.35    0.35   \n",
       "thc_nox_emissions_mgkm                             70   1316    78.20   23.33   \n",
       "thc_emissions_mgkm                                NaN   1350    42.89   11.89   \n",
       "\n",
       "                           min     25%     50%     75%     max    dtype  \n",
       "mfr                     abarth     NaN     NaN     NaN   volvo   object  \n",
       "model                   1 seri     NaN     NaN     NaN  zafira   object  \n",
       "desc                    '1 air     NaN     NaN     NaN  zc 1.6   object  \n",
       "trans                      5MT     NaN     NaN     NaN    SAT6   object  \n",
       "fuel_type               diesel     NaN     NaN     NaN  petrol   object  \n",
       "metric_combined            3.1     4.4     5.1     6.3      16  float64  \n",
       "metric_extra_urban           0     4.0     4.5     5.3     107  float64  \n",
       "metric_urban_cold            0     5.2     6.2     8.1    24.7  float64  \n",
       "engine_capacity            395  1560.0  1984.0  1999.0    6752  float64  \n",
       "emissions_co_mgkm           30   162.0   243.0   341.0     913  float64  \n",
       "emissions_nox_mgkm           1    23.0    35.0    52.0     158  float64  \n",
       "particulates_no_mgkm         0     0.1     0.2     0.5       2  float64  \n",
       "thc_nox_emissions_mgkm      11    64.0    78.0    90.0     184  float64  \n",
       "thc_emissions_mgkm           7    35.0    43.0    51.0      95  float64  "
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "custom_describe(df)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Feature Creation"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Derive transmission"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "manual      1280\n",
      "auto        1033\n",
      "semiauto     330\n",
      "Name: trans_cln, dtype: int64\n",
      "0\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAssAAAGlCAYAAAAMHrtsAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nMBgMBoPBYLQCc5YZDAaDwWAwGIxW+D9C1MmkpVk3SgAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11647f438>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df['trans_cln'] = df['trans'].apply(lambda x: clean_transmission(x))\n",
    "print(df['trans_cln'].value_counts())\n",
    "print(df['trans'].value_counts().sum() - df['trans_cln'].value_counts().sum())\n",
    "_ = sns.heatmap(df.groupby(['trans_cln','trans']).size().unstack()\n",
    "                 ,cbar=False, annot=True, fmt='.0f',annot_kws={'rotation':'90'})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# seems good, overwrite transmission and remove transmission_cln\n",
    "df.drop('trans', axis=1, inplace=True)\n",
    "df.rename(columns={'trans_cln':'trans'}, inplace=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Derive TDi"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "collapsed": false,
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "298"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def get_tdi(r):\n",
    "    if r['fuel_type'] in set(['diesel', 'diesel electric']):\n",
    "        if re.search('tdi',r['desc']) != None:\n",
    "            return True\n",
    "        else:\n",
    "            return False\n",
    "    else:\n",
    "        return False\n",
    "\n",
    "df['is_tdi'] = df.apply(lambda r: get_tdi(r), axis=1)\n",
    "df['is_tdi'].sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th>is_tdi</th>\n",
       "      <th>False</th>\n",
       "      <th>True</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fuel_type</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>diesel</th>\n",
       "      <td>994.0</td>\n",
       "      <td>298.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>diesel electric</th>\n",
       "      <td>14.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>petrol</th>\n",
       "      <td>1302.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>petrol electric</th>\n",
       "      <td>10.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>petrol hybrid</th>\n",
       "      <td>25.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "is_tdi            False  True \n",
       "fuel_type                     \n",
       "diesel            994.0  298.0\n",
       "diesel electric    14.0    NaN\n",
       "petrol           1302.0    NaN\n",
       "petrol electric    10.0    NaN\n",
       "petrol hybrid      25.0    NaN"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# check against fuel type\n",
    "df.groupby(['fuel_type','is_tdi']).size().unstack()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Simplify fuel type"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "fuel_type\n",
       "diesel    1292\n",
       "hybrid      49\n",
       "petrol    1302\n",
       "dtype: int64"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def simplify_fuel_type(r):\n",
    "    if r['fuel_type'] in set(['petrol', 'diesel']):\n",
    "        return r['fuel_type']\n",
    "    else:\n",
    "        return 'hybrid'\n",
    "\n",
    "df['fuel_type'] = df.apply(lambda r: simplify_fuel_type(r), axis=1)\n",
    "df.groupby(['fuel_type']).size()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Add hierarchical parent company ownership structure"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "parent = {\n",
    "    'aston martin lagonda':['aston martin lagonda']\n",
    "    ,'bmw':['bmw','mini','rolls royce']\n",
    "    ,'daimler-ag':['mercedes-benz','smart']          \n",
    "    ,'fiat':['fiat','abarth','alfa romeo','chrysler jeep','ferrari','lancia','maserati','ram']\n",
    "    ,'ford':['ford']\n",
    "    ,'gm':['buick','cadillac','chevrolet','vauxhall']\n",
    "    ,'honda':['honda','acura']\n",
    "    ,'hyundai':['hyundai','kia']\n",
    "    ,'mazda':['mazda']\n",
    "    ,'mclaren':['mclaren']\n",
    "    ,'mitsubishi':['mitsubishi']\n",
    "    ,'peugeot-citroen':['peugeot','citroen','ds']\n",
    "    ,'renault-nissan':['renault','dacia','nissan']\n",
    "    ,'saab':['saab']\n",
    "    ,'shanghai-auto':['ssangyong']\n",
    "    ,'subaru':['subaru']\n",
    "    ,'suzuki':['suzuki']\n",
    "    ,'tata':['jaguar','land rover']\n",
    "    ,'toyota':['toyota','lexus']\n",
    "    ,'volkswagen':['volkswagen','audi','bentley motors','lamborghini','porsche','seat','skoda']\n",
    "    ,'geely':['volvo']\n",
    "}\n",
    "\n",
    "# and invert dictionary for reverse lookup\n",
    "parent_inv = {}\n",
    "for k, vs in parent.items():\n",
    "    for v in vs:\n",
    "        parent_inv[v] = k"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "df['parent'] = df['mfr'].apply(lambda x: parent_inv[x])\n",
    "ft_cat.append('parent')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# check that every car has a parent\n",
    "df.shape[0] == df.groupby(['mfr','parent']).size().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Plot ownership structure"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAyAAAAHRCAYAAACSId39AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x117516cf8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "f, ax1d = plt.subplots(1,1,figsize=(12,6))\n",
    "ax = sns.heatmap(df.groupby(['mfr','parent']).size().unstack(0)\n",
    "                 ,cmap=\"OrRd\", ax=ax1d, annot=True, fmt='.0f', cbar=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Add boolean for mfr == volkswagen, and parent == volkswagen"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "df['mfr_is_vw'] = df['mfr'] == 'volkswagen'\n",
    "df['parent_is_vw'] = df['parent'] == 'volkswagen'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th>parent_is_vw</th>\n",
       "      <th>False</th>\n",
       "      <th>True</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mfr_is_vw</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>False</th>\n",
       "      <td>2000.0</td>\n",
       "      <td>536.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>True</th>\n",
       "      <td>NaN</td>\n",
       "      <td>107.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "parent_is_vw   False  True \n",
       "mfr_is_vw                  \n",
       "False         2000.0  536.0\n",
       "True             NaN  107.0"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.groupby(['mfr_is_vw','parent_is_vw']).size().unstack()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Final check"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(2643, 18)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>277</th>\n",
       "      <th>1788</th>\n",
       "      <th>1839</th>\n",
       "      <th>count</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>min</th>\n",
       "      <th>25%</th>\n",
       "      <th>50%</th>\n",
       "      <th>75%</th>\n",
       "      <th>max</th>\n",
       "      <th>dtype</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>mfr</th>\n",
       "      <td>audi</td>\n",
       "      <td>mercedes-benz</td>\n",
       "      <td>mercedes-benz</td>\n",
       "      <td>2643</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>abarth</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>volvo</td>\n",
       "      <td>object</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>model</th>\n",
       "      <td>a7 sportback</td>\n",
       "      <td>gla class, model year 2015</td>\n",
       "      <td>s-class, model year 2015½</td>\n",
       "      <td>2643</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1 seri</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>zafira</td>\n",
       "      <td>object</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>desc</th>\n",
       "      <td>3.0 v6 tdi ultra 218ps 7speed s tronic with 19...</td>\n",
       "      <td>gla 250 4matic with 19\" rear wheels</td>\n",
       "      <td>s 350 bluetec with 19'' rear wheels</td>\n",
       "      <td>2643</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>'1 air</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>zc 1.6</td>\n",
       "      <td>object</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fuel_type</th>\n",
       "      <td>diesel</td>\n",
       "      <td>petrol</td>\n",
       "      <td>diesel</td>\n",
       "      <td>2643</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>diesel</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>petrol</td>\n",
       "      <td>object</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>metric_combined</th>\n",
       "      <td>4.8</td>\n",
       "      <td>6.6</td>\n",
       "      <td>5.9</td>\n",
       "      <td>2643</td>\n",
       "      <td>5.65</td>\n",
       "      <td>1.87</td>\n",
       "      <td>3.1</td>\n",
       "      <td>4.4</td>\n",
       "      <td>5.1</td>\n",
       "      <td>6.3</td>\n",
       "      <td>16</td>\n",
       "      <td>float64</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>metric_extra_urban</th>\n",
       "      <td>4.4</td>\n",
       "      <td>5.6</td>\n",
       "      <td>5.1</td>\n",
       "      <td>2643</td>\n",
       "      <td>4.85</td>\n",
       "      <td>2.35</td>\n",
       "      <td>0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.5</td>\n",
       "      <td>5.3</td>\n",
       "      <td>107</td>\n",
       "      <td>float64</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>metric_urban_cold</th>\n",
       "      <td>5.5</td>\n",
       "      <td>8.3</td>\n",
       "      <td>7.3</td>\n",
       "      <td>2643</td>\n",
       "      <td>7.10</td>\n",
       "      <td>3.00</td>\n",
       "      <td>0</td>\n",
       "      <td>5.2</td>\n",
       "      <td>6.2</td>\n",
       "      <td>8.1</td>\n",
       "      <td>24.7</td>\n",
       "      <td>float64</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>engine_capacity</th>\n",
       "      <td>2967</td>\n",
       "      <td>1991</td>\n",
       "      <td>2987</td>\n",
       "      <td>2643</td>\n",
       "      <td>2073.12</td>\n",
       "      <td>903.98</td>\n",
       "      <td>395</td>\n",
       "      <td>1560.0</td>\n",
       "      <td>1984.0</td>\n",
       "      <td>1999.0</td>\n",
       "      <td>6752</td>\n",
       "      <td>float64</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>emissions_co_mgkm</th>\n",
       "      <td>171</td>\n",
       "      <td>144</td>\n",
       "      <td>277</td>\n",
       "      <td>2643</td>\n",
       "      <td>268.21</td>\n",
       "      <td>140.57</td>\n",
       "      <td>30</td>\n",
       "      <td>162.0</td>\n",
       "      <td>243.0</td>\n",
       "      <td>341.0</td>\n",
       "      <td>913</td>\n",
       "      <td>float64</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>emissions_nox_mgkm</th>\n",
       "      <td>50</td>\n",
       "      <td>27</td>\n",
       "      <td>66</td>\n",
       "      <td>2643</td>\n",
       "      <td>37.40</td>\n",
       "      <td>18.63</td>\n",
       "      <td>1</td>\n",
       "      <td>23.0</td>\n",
       "      <td>35.0</td>\n",
       "      <td>52.0</td>\n",
       "      <td>158</td>\n",
       "      <td>float64</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>particulates_no_mgkm</th>\n",
       "      <td>0.4</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.8</td>\n",
       "      <td>1306</td>\n",
       "      <td>0.35</td>\n",
       "      <td>0.35</td>\n",
       "      <td>0</td>\n",
       "      <td>0.1</td>\n",
       "      <td>0.2</td>\n",
       "      <td>0.5</td>\n",
       "      <td>2</td>\n",
       "      <td>float64</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>thc_nox_emissions_mgkm</th>\n",
       "      <td>73</td>\n",
       "      <td>NaN</td>\n",
       "      <td>90</td>\n",
       "      <td>1316</td>\n",
       "      <td>78.20</td>\n",
       "      <td>23.33</td>\n",
       "      <td>11</td>\n",
       "      <td>64.0</td>\n",
       "      <td>78.0</td>\n",
       "      <td>90.0</td>\n",
       "      <td>184</td>\n",
       "      <td>float64</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>thc_emissions_mgkm</th>\n",
       "      <td>NaN</td>\n",
       "      <td>30</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1350</td>\n",
       "      <td>42.89</td>\n",
       "      <td>11.89</td>\n",
       "      <td>7</td>\n",
       "      <td>35.0</td>\n",
       "      <td>43.0</td>\n",
       "      <td>51.0</td>\n",
       "      <td>95</td>\n",
       "      <td>float64</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>trans</th>\n",
       "      <td>semiauto</td>\n",
       "      <td>auto</td>\n",
       "      <td>auto</td>\n",
       "      <td>2643</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>auto</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>semiau</td>\n",
       "      <td>object</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>is_tdi</th>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>2643</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>False</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>True</td>\n",
       "      <td>bool</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>parent</th>\n",
       "      <td>volkswagen</td>\n",
       "      <td>daimler-ag</td>\n",
       "      <td>daimler-ag</td>\n",
       "      <td>2643</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>aston</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>volksw</td>\n",
       "      <td>object</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mfr_is_vw</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>2643</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>False</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>True</td>\n",
       "      <td>bool</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>parent_is_vw</th>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>2643</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>False</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>True</td>\n",
       "      <td>bool</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                                      277  \\\n",
       "mfr                                                                  audi   \n",
       "model                                                        a7 sportback   \n",
       "desc                    3.0 v6 tdi ultra 218ps 7speed s tronic with 19...   \n",
       "fuel_type                                                          diesel   \n",
       "metric_combined                                                       4.8   \n",
       "metric_extra_urban                                                    4.4   \n",
       "metric_urban_cold                                                     5.5   \n",
       "engine_capacity                                                      2967   \n",
       "emissions_co_mgkm                                                     171   \n",
       "emissions_nox_mgkm                                                     50   \n",
       "particulates_no_mgkm                                                  0.4   \n",
       "thc_nox_emissions_mgkm                                                 73   \n",
       "thc_emissions_mgkm                                                    NaN   \n",
       "trans                                                            semiauto   \n",
       "is_tdi                                                               True   \n",
       "parent                                                         volkswagen   \n",
       "mfr_is_vw                                                           False   \n",
       "parent_is_vw                                                         True   \n",
       "\n",
       "                                                       1788  \\\n",
       "mfr                                           mercedes-benz   \n",
       "model                            gla class, model year 2015   \n",
       "desc                    gla 250 4matic with 19\" rear wheels   \n",
       "fuel_type                                            petrol   \n",
       "metric_combined                                         6.6   \n",
       "metric_extra_urban                                      5.6   \n",
       "metric_urban_cold                                       8.3   \n",
       "engine_capacity                                        1991   \n",
       "emissions_co_mgkm                                       144   \n",
       "emissions_nox_mgkm                                       27   \n",
       "particulates_no_mgkm                                    NaN   \n",
       "thc_nox_emissions_mgkm                                  NaN   \n",
       "thc_emissions_mgkm                                       30   \n",
       "trans                                                  auto   \n",
       "is_tdi                                                False   \n",
       "parent                                           daimler-ag   \n",
       "mfr_is_vw                                             False   \n",
       "parent_is_vw                                          False   \n",
       "\n",
       "                                                       1839  count     mean  \\\n",
       "mfr                                           mercedes-benz   2643      NaN   \n",
       "model                             s-class, model year 2015½   2643      NaN   \n",
       "desc                    s 350 bluetec with 19'' rear wheels   2643      NaN   \n",
       "fuel_type                                            diesel   2643      NaN   \n",
       "metric_combined                                         5.9   2643     5.65   \n",
       "metric_extra_urban                                      5.1   2643     4.85   \n",
       "metric_urban_cold                                       7.3   2643     7.10   \n",
       "engine_capacity                                        2987   2643  2073.12   \n",
       "emissions_co_mgkm                                       277   2643   268.21   \n",
       "emissions_nox_mgkm                                       66   2643    37.40   \n",
       "particulates_no_mgkm                                    0.8   1306     0.35   \n",
       "thc_nox_emissions_mgkm                                   90   1316    78.20   \n",
       "thc_emissions_mgkm                                      NaN   1350    42.89   \n",
       "trans                                                  auto   2643      NaN   \n",
       "is_tdi                                                False   2643      NaN   \n",
       "parent                                           daimler-ag   2643      NaN   \n",
       "mfr_is_vw                                             False   2643      NaN   \n",
       "parent_is_vw                                          False   2643      NaN   \n",
       "\n",
       "                           std     min     25%     50%     75%     max  \\\n",
       "mfr                        NaN  abarth     NaN     NaN     NaN   volvo   \n",
       "model                      NaN  1 seri     NaN     NaN     NaN  zafira   \n",
       "desc                       NaN  '1 air     NaN     NaN     NaN  zc 1.6   \n",
       "fuel_type                  NaN  diesel     NaN     NaN     NaN  petrol   \n",
       "metric_combined           1.87     3.1     4.4     5.1     6.3      16   \n",
       "metric_extra_urban        2.35       0     4.0     4.5     5.3     107   \n",
       "metric_urban_cold         3.00       0     5.2     6.2     8.1    24.7   \n",
       "engine_capacity         903.98     395  1560.0  1984.0  1999.0    6752   \n",
       "emissions_co_mgkm       140.57      30   162.0   243.0   341.0     913   \n",
       "emissions_nox_mgkm       18.63       1    23.0    35.0    52.0     158   \n",
       "particulates_no_mgkm      0.35       0     0.1     0.2     0.5       2   \n",
       "thc_nox_emissions_mgkm   23.33      11    64.0    78.0    90.0     184   \n",
       "thc_emissions_mgkm       11.89       7    35.0    43.0    51.0      95   \n",
       "trans                      NaN    auto     NaN     NaN     NaN  semiau   \n",
       "is_tdi                     NaN   False     NaN     NaN     NaN    True   \n",
       "parent                     NaN  aston      NaN     NaN     NaN  volksw   \n",
       "mfr_is_vw                  NaN   False     NaN     NaN     NaN    True   \n",
       "parent_is_vw               NaN   False     NaN     NaN     NaN    True   \n",
       "\n",
       "                          dtype  \n",
       "mfr                      object  \n",
       "model                    object  \n",
       "desc                     object  \n",
       "fuel_type                object  \n",
       "metric_combined         float64  \n",
       "metric_extra_urban      float64  \n",
       "metric_urban_cold       float64  \n",
       "engine_capacity         float64  \n",
       "emissions_co_mgkm       float64  \n",
       "emissions_nox_mgkm      float64  \n",
       "particulates_no_mgkm    float64  \n",
       "thc_nox_emissions_mgkm  float64  \n",
       "thc_emissions_mgkm      float64  \n",
       "trans                    object  \n",
       "is_tdi                     bool  \n",
       "parent                   object  \n",
       "mfr_is_vw                  bool  \n",
       "parent_is_vw               bool  "
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "custom_describe(df)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### And write to db"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "df.reset_index(inplace=True)\n",
    "del df['index']\n",
    "df.to_sql('cars', cnxsql, if_exists='replace', index=None)\n",
    "cnxsql.close()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Quickly check size of data of interest: Volkswagen"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Manufacturer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "VW Cars are 107 / 2643 = 4.0%\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAArkAAAD/CAYAAAD1wM7sAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nhKhzpMgVQgghhBB1jhS5QgghhBCizvl/HbpYQRSbQLMAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x119aac2b0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "grp = df.groupby('mfr').size().sort_values()#[::-1]\n",
    "ax = grp.plot(kind='barh', figsize=(10,4))\n",
    "print('VW Cars are {} / {} = {:.1%}'.format(grp['volkswagen'], grp.sum()\n",
    "                                       ,grp['volkswagen'] / grp.sum()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "VW Group are 643 / 2643 = 24.3%\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAArkAAAD9CAYAAAC4CG/nAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n9bwH+R8AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x119a9edd8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "grp = df.groupby('parent').size().sort_values()#[::-1]\n",
    "ax = grp.plot(kind='barh', figsize=(10,4))\n",
    "print('VW Group are {} / {} = {:.1%}'.format(grp['volkswagen'], grp.sum()\n",
    "                                       ,grp['volkswagen'] / grp.sum()))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Observe**: \n",
    "\n",
    "+ There's a long tail of underrepresented cars, we might need to account for this later\n",
    "+ Volkswagen is reasonably well represented in the full set with 4% of the total"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---\n",
    "**Applied AI Ltd &copy; 2016**  \n",
    "<a href='http://www.applied.ai'>applied.ai</a>"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.5.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}