Retail_17.8.2.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Business Objective"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- I am a store manager, what can I do to enhance my revenue across 100 stores?  \n",
    "\n",
    "- When I discount some items, will that drive up average sales per basket？\n",
    "\n",
    "- Some product may have **elasticity of demand**, other product may not (Once I step into the store, I am going to start browsing and I am gonna buying more stuff I probably won't need. So that's what they want, they want to change people's behavior. If you markdown the price of merchandise, it's going to drive up sales of products.) (Percent change in demand is more than percent change in price) \n",
    "\n",
    "- Cross-selling: what are the prodcuts that people tend to buy together. Association Rule.  \n",
    "- Upselling: Salesperson tries to persuade customers to buy a higher-end product, an upgrade or the newest version of an item.\n",
    "\n",
    "- Shall we markdown? How much shall we markdown? Does it make sense to markdown? Product Category?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<br>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Look at Data First"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "from datetime import datetime"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0:00:01.052109\n"
     ]
    }
   ],
   "source": [
    "start_time = datetime.now()\n",
    "data = pd.read_csv('RetailerPromotionStrategy_data.csv')\n",
    "print datetime.now()-start_time"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1048575, 7)"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "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>StoreID</th>\n",
       "      <th>TransactionID</th>\n",
       "      <th>Line</th>\n",
       "      <th>Sales</th>\n",
       "      <th>Count</th>\n",
       "      <th>Date</th>\n",
       "      <th>SKU</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00057</td>\n",
       "      <td>36</td>\n",
       "      <td>850.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A21773</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00057</td>\n",
       "      <td>39</td>\n",
       "      <td>1500.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A18009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00057</td>\n",
       "      <td>40</td>\n",
       "      <td>1500.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A25592</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00057</td>\n",
       "      <td>41</td>\n",
       "      <td>1600.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A1449</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00057</td>\n",
       "      <td>42</td>\n",
       "      <td>1600.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A1449</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00057</td>\n",
       "      <td>52</td>\n",
       "      <td>1700.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A12434</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00057</td>\n",
       "      <td>57</td>\n",
       "      <td>1695.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A13081</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00057</td>\n",
       "      <td>58</td>\n",
       "      <td>2095.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A5985</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00057</td>\n",
       "      <td>59</td>\n",
       "      <td>1650.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A21387</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00057</td>\n",
       "      <td>60</td>\n",
       "      <td>1795.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A5987</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  StoreID        TransactionID  Line   Sales  Count           Date     SKU\n",
       "0     A29  11-7270-00007-00057    36   850.0    1.0  3/11/12 00:00  A21773\n",
       "1     A29  11-7270-00007-00057    39  1500.0    1.0  3/11/12 00:00  A18009\n",
       "2     A29  11-7270-00007-00057    40  1500.0    1.0  3/11/12 00:00  A25592\n",
       "3     A29  11-7270-00007-00057    41  1600.0    1.0  3/11/12 00:00   A1449\n",
       "4     A29  11-7270-00007-00057    42  1600.0    1.0  3/11/12 00:00   A1449\n",
       "5     A29  11-7270-00007-00057    52  1700.0    1.0  3/11/12 00:00  A12434\n",
       "6     A29  11-7270-00007-00057    57  1695.0    1.0  3/11/12 00:00  A13081\n",
       "7     A29  11-7270-00007-00057    58  2095.0    1.0  3/11/12 00:00   A5985\n",
       "8     A29  11-7270-00007-00057    59  1650.0    1.0  3/11/12 00:00  A21387\n",
       "9     A29  11-7270-00007-00057    60  1795.0    1.0  3/11/12 00:00   A5987"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "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>Line</th>\n",
       "      <th>Sales</th>\n",
       "      <th>Count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>1048575.000</td>\n",
       "      <td>1048575.000</td>\n",
       "      <td>1048575.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>10.727</td>\n",
       "      <td>1720.513</td>\n",
       "      <td>0.896</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>15.550</td>\n",
       "      <td>3646.003</td>\n",
       "      <td>1.129</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>1.000</td>\n",
       "      <td>-99995.000</td>\n",
       "      <td>-120.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>3.000</td>\n",
       "      <td>395.000</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>7.000</td>\n",
       "      <td>1295.000</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>14.000</td>\n",
       "      <td>2195.000</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>823.000</td>\n",
       "      <td>99995.000</td>\n",
       "      <td>120.000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             Line       Sales       Count\n",
       "count 1048575.000 1048575.000 1048575.000\n",
       "mean       10.727    1720.513       0.896\n",
       "std        15.550    3646.003       1.129\n",
       "min         1.000  -99995.000    -120.000\n",
       "25%         3.000     395.000       1.000\n",
       "50%         7.000    1295.000       1.000\n",
       "75%        14.000    2195.000       1.000\n",
       "max       823.000   99995.000     120.000"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.set_option('display.float_format', lambda x: '%.3f' % x)\n",
    "data.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1      138198\n",
       "2      103886\n",
       "3       84615\n",
       "4       72704\n",
       "5       63255\n",
       "6       56578\n",
       "7       50227\n",
       "8       45197\n",
       "9       40427\n",
       "10      36305\n",
       "11      32797\n",
       "12      29470\n",
       "13      26529\n",
       "14      23595\n",
       "15      21240\n",
       "16      19211\n",
       "17      17401\n",
       "18      15649\n",
       "19      14274\n",
       "20      12758\n",
       "21      11479\n",
       "22      10460\n",
       "23       9544\n",
       "24       8744\n",
       "25       7832\n",
       "26       7065\n",
       "27       6486\n",
       "28       6029\n",
       "29       5405\n",
       "30       4995\n",
       "        ...  \n",
       "285         1\n",
       "796         1\n",
       "236         1\n",
       "748         1\n",
       "750         1\n",
       "239         1\n",
       "240         1\n",
       "754         1\n",
       "755         1\n",
       "756         1\n",
       "246         1\n",
       "758         1\n",
       "760         1\n",
       "763         1\n",
       "764         1\n",
       "765         1\n",
       "256         1\n",
       "265         1\n",
       "267         1\n",
       "268         1\n",
       "269         1\n",
       "270         1\n",
       "272         1\n",
       "277         1\n",
       "278         1\n",
       "279         1\n",
       "280         1\n",
       "281         1\n",
       "282         1\n",
       "182         1\n",
       "Name: Line, Length: 604, dtype: int64"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# How many observations in each Line\n",
    "data['Line'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.000       850239\n",
       "0.000       171448\n",
       "-1.000        8577\n",
       "2.000         5810\n",
       "10.000        2943\n",
       "4.000         2471\n",
       "3.000         1687\n",
       "6.000         1624\n",
       "5.000         1271\n",
       "8.000          356\n",
       "-2.000         345\n",
       "20.000         283\n",
       "9.000          222\n",
       "12.000         175\n",
       "-3.000         150\n",
       "30.000         117\n",
       "7.000          104\n",
       "15.000          78\n",
       "-4.000          76\n",
       "18.000          56\n",
       "-10.000         52\n",
       "-5.000          46\n",
       "-6.000          46\n",
       "16.000          44\n",
       "40.000          42\n",
       "24.000          35\n",
       "11.000          26\n",
       "14.000          23\n",
       "13.000          23\n",
       "50.000          22\n",
       "             ...  \n",
       "-16.000          2\n",
       "80.000           2\n",
       "29.000           2\n",
       "54.000           2\n",
       "-25.000          2\n",
       "-18.000          2\n",
       "72.000           2\n",
       "-54.000          2\n",
       "-48.000          1\n",
       "39.000           1\n",
       "91.000           1\n",
       "-9.000           1\n",
       "-120.000         1\n",
       "-40.000          1\n",
       "-91.000          1\n",
       "-36.000          1\n",
       "-11.000          1\n",
       "-14.000          1\n",
       "26.000           1\n",
       "65.000           1\n",
       "-33.000          1\n",
       "-7.000           1\n",
       "120.000          1\n",
       "33.000           1\n",
       "-31.000          1\n",
       "52.000           1\n",
       "-39.000          1\n",
       "44.000           1\n",
       "42.000           1\n",
       "0.010            1\n",
       "Name: Count, Length: 87, dtype: int64"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# How many oberservations in each Count category\n",
    "data['Count'].value_counts().sort_values(ascending=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "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>StoreID</th>\n",
       "      <th>TransactionID</th>\n",
       "      <th>Line</th>\n",
       "      <th>Sales</th>\n",
       "      <th>Count</th>\n",
       "      <th>Date</th>\n",
       "      <th>SKU</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>304917</th>\n",
       "      <td>A22</td>\n",
       "      <td>12-7234-00007-00457</td>\n",
       "      <td>1</td>\n",
       "      <td>99995.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3/12/12 00:00</td>\n",
       "      <td>A19551</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>304918</th>\n",
       "      <td>A22</td>\n",
       "      <td>12-7234-00007-00457</td>\n",
       "      <td>2</td>\n",
       "      <td>-99995.000</td>\n",
       "      <td>-1.000</td>\n",
       "      <td>3/12/12 00:00</td>\n",
       "      <td>A19551</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>378017</th>\n",
       "      <td>A27</td>\n",
       "      <td>12-7070-00007-00002</td>\n",
       "      <td>763</td>\n",
       "      <td>3195.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3/12/12 00:00</td>\n",
       "      <td>A3503</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>378029</th>\n",
       "      <td>A27</td>\n",
       "      <td>12-7070-00007-00002</td>\n",
       "      <td>823</td>\n",
       "      <td>-646.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>3/12/12 00:00</td>\n",
       "      <td>A25525</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>653883</th>\n",
       "      <td>A46</td>\n",
       "      <td>12-7027-00013-00092</td>\n",
       "      <td>3</td>\n",
       "      <td>60000.000</td>\n",
       "      <td>120.000</td>\n",
       "      <td>3/12/12 00:00</td>\n",
       "      <td>A10809</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>653885</th>\n",
       "      <td>A46</td>\n",
       "      <td>12-7027-00013-00092</td>\n",
       "      <td>6</td>\n",
       "      <td>-60000.000</td>\n",
       "      <td>-120.000</td>\n",
       "      <td>3/12/12 00:00</td>\n",
       "      <td>A10809</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       StoreID        TransactionID  Line      Sales    Count           Date  \\\n",
       "304917     A22  12-7234-00007-00457     1  99995.000    1.000  3/12/12 00:00   \n",
       "304918     A22  12-7234-00007-00457     2 -99995.000   -1.000  3/12/12 00:00   \n",
       "378017     A27  12-7070-00007-00002   763   3195.000    1.000  3/12/12 00:00   \n",
       "378029     A27  12-7070-00007-00002   823   -646.000    0.000  3/12/12 00:00   \n",
       "653883     A46  12-7027-00013-00092     3  60000.000  120.000  3/12/12 00:00   \n",
       "653885     A46  12-7027-00013-00092     6 -60000.000 -120.000  3/12/12 00:00   \n",
       "\n",
       "           SKU  \n",
       "304917  A19551  \n",
       "304918  A19551  \n",
       "378017   A3503  \n",
       "378029  A25525  \n",
       "653883  A10809  \n",
       "653885  A10809  "
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.loc[(data['Count']==-120)|(data['Count']==120)| (data['Sales']==-99995)|(data['Sales']==99995)|\n",
    "        (data['Line']==823)|(data['Line']==763)]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<br><br>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Exploratory Data Analysis"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Explore cases when Count=0"
   ]
  },
  {
   "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>StoreID</th>\n",
       "      <th>TransactionID</th>\n",
       "      <th>Line</th>\n",
       "      <th>Sales</th>\n",
       "      <th>Count</th>\n",
       "      <th>Date</th>\n",
       "      <th>SKU</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00057</td>\n",
       "      <td>63</td>\n",
       "      <td>-786.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A256</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00057</td>\n",
       "      <td>65</td>\n",
       "      <td>-770.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A8481</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00057</td>\n",
       "      <td>75</td>\n",
       "      <td>-500.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A2341</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00059</td>\n",
       "      <td>3</td>\n",
       "      <td>-800.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A24593</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00060</td>\n",
       "      <td>9</td>\n",
       "      <td>-475.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A15257</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   StoreID        TransactionID  Line    Sales  Count           Date     SKU\n",
       "12     A29  11-7270-00007-00057    63 -786.000  0.000  3/11/12 00:00    A256\n",
       "14     A29  11-7270-00007-00057    65 -770.000  0.000  3/11/12 00:00   A8481\n",
       "20     A29  11-7270-00007-00057    75 -500.000  0.000  3/11/12 00:00   A2341\n",
       "34     A29  11-7270-00007-00059     3 -800.000  0.000  3/11/12 00:00  A24593\n",
       "39     A29  11-7270-00007-00060     9 -475.000  0.000  3/11/12 00:00  A15257"
      ]
     },
     "execution_count": 85,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.loc[data['Count']==0].head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "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>StoreID</th>\n",
       "      <th>TransactionID</th>\n",
       "      <th>Line</th>\n",
       "      <th>Sales</th>\n",
       "      <th>Count</th>\n",
       "      <th>Date</th>\n",
       "      <th>SKU</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00057</td>\n",
       "      <td>62</td>\n",
       "      <td>2095.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A256</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00057</td>\n",
       "      <td>63</td>\n",
       "      <td>-786.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A256</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   StoreID        TransactionID  Line    Sales  Count           Date   SKU\n",
       "11     A29  11-7270-00007-00057    62 2095.000  1.000  3/11/12 00:00  A256\n",
       "12     A29  11-7270-00007-00057    63 -786.000  0.000  3/11/12 00:00  A256"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# When Count=0, it means it's a discount applied to the purchase\n",
    "data.loc[(data['SKU']=='A256')&(data['TransactionID']=='11-7270-00007-00057')]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "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>StoreID</th>\n",
       "      <th>TransactionID</th>\n",
       "      <th>Line</th>\n",
       "      <th>Sales</th>\n",
       "      <th>Count</th>\n",
       "      <th>Date</th>\n",
       "      <th>SKU</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1161</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00110</td>\n",
       "      <td>29</td>\n",
       "      <td>845.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A6007</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1234</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00116</td>\n",
       "      <td>20</td>\n",
       "      <td>234.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A2452</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4065</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00008-00358</td>\n",
       "      <td>6</td>\n",
       "      <td>670.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A13774</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5634</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00014-00241</td>\n",
       "      <td>10</td>\n",
       "      <td>333.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A21307</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9361</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00006-00209</td>\n",
       "      <td>33</td>\n",
       "      <td>475.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A15257</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     StoreID        TransactionID  Line   Sales  Count           Date     SKU\n",
       "1161     A29  11-7270-00007-00110    29 845.000  0.000  3/11/12 00:00   A6007\n",
       "1234     A29  11-7270-00007-00116    20 234.000  0.000  3/11/12 00:00   A2452\n",
       "4065     A29  11-7270-00008-00358     6 670.000  0.000  3/11/12 00:00  A13774\n",
       "5634     A29  11-7270-00014-00241    10 333.000  0.000  3/11/12 00:00  A21307\n",
       "9361     A29  11-7270-00006-00209    33 475.000  0.000  3/11/12 00:00  A15257"
      ]
     },
     "execution_count": 108,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Why there are rows where Count=0 but Sales>0?\n",
    "data.loc[(data['Count']==0)&(data['Sales']>0)].head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Select rows where transactionID and SKU match transactionID and SKU when Count=0 but Sales>0\n",
    "q = '''\n",
    "    select\n",
    "        d1.StoreID, d1.TransactionID, d1.Line, d1.Sales, d1.Count, d1.Date, d1.SKU\n",
    "    from \n",
    "        data d1\n",
    "    inner join\n",
    "        (select \n",
    "            TransactionID, SKU\n",
    "        from\n",
    "            data\n",
    "        where Sales>0 and Count=0) as d2\n",
    "    on \n",
    "        d1.TransactionID = d2.TransactionID and\n",
    "        d1.SKU = d2.SKU\n",
    "    '''\n",
    "data1 = pysqldf(q)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "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>StoreID</th>\n",
       "      <th>TransactionID</th>\n",
       "      <th>Line</th>\n",
       "      <th>Sales</th>\n",
       "      <th>Count</th>\n",
       "      <th>Date</th>\n",
       "      <th>SKU</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00110</td>\n",
       "      <td>22</td>\n",
       "      <td>2195.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A6007</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00110</td>\n",
       "      <td>23</td>\n",
       "      <td>-845.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A6007</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00110</td>\n",
       "      <td>24</td>\n",
       "      <td>2195.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A6007</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00110</td>\n",
       "      <td>25</td>\n",
       "      <td>-845.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A6007</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00110</td>\n",
       "      <td>26</td>\n",
       "      <td>2195.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A6007</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00110</td>\n",
       "      <td>27</td>\n",
       "      <td>-845.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A6007</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00110</td>\n",
       "      <td>28</td>\n",
       "      <td>-2195.000</td>\n",
       "      <td>-1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A6007</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00110</td>\n",
       "      <td>29</td>\n",
       "      <td>845.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A6007</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00116</td>\n",
       "      <td>15</td>\n",
       "      <td>1423.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A2452</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00116</td>\n",
       "      <td>16</td>\n",
       "      <td>-234.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A2452</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  StoreID        TransactionID  Line     Sales  Count           Date    SKU\n",
       "0     A29  11-7270-00007-00110    22  2195.000  1.000  3/11/12 00:00  A6007\n",
       "1     A29  11-7270-00007-00110    23  -845.000  0.000  3/11/12 00:00  A6007\n",
       "2     A29  11-7270-00007-00110    24  2195.000  1.000  3/11/12 00:00  A6007\n",
       "3     A29  11-7270-00007-00110    25  -845.000  0.000  3/11/12 00:00  A6007\n",
       "4     A29  11-7270-00007-00110    26  2195.000  1.000  3/11/12 00:00  A6007\n",
       "5     A29  11-7270-00007-00110    27  -845.000  0.000  3/11/12 00:00  A6007\n",
       "6     A29  11-7270-00007-00110    28 -2195.000 -1.000  3/11/12 00:00  A6007\n",
       "7     A29  11-7270-00007-00110    29   845.000  0.000  3/11/12 00:00  A6007\n",
       "8     A29  11-7270-00007-00116    15  1423.000  1.000  3/11/12 00:00  A2452\n",
       "9     A29  11-7270-00007-00116    16  -234.000  0.000  3/11/12 00:00  A2452"
      ]
     },
     "execution_count": 116,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data1.head(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### We know from the above table: some customers have bought the discounted items but he told the counter cashier that he didn't want this item anymore. In order to not include the discount from that item, we have to add that discount back. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<br><br>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Explore cases when Count=-1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "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>StoreID</th>\n",
       "      <th>TransactionID</th>\n",
       "      <th>Line</th>\n",
       "      <th>Sales</th>\n",
       "      <th>Count</th>\n",
       "      <th>Date</th>\n",
       "      <th>SKU</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>100</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00066</td>\n",
       "      <td>5</td>\n",
       "      <td>-2013.000</td>\n",
       "      <td>-1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A2729</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>354</th>\n",
       "      <td>A19</td>\n",
       "      <td>11-7287-00025-00075</td>\n",
       "      <td>4</td>\n",
       "      <td>-1721.000</td>\n",
       "      <td>-1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A16911</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>446</th>\n",
       "      <td>A27</td>\n",
       "      <td>11-7069-00020-00010</td>\n",
       "      <td>6</td>\n",
       "      <td>-4252.000</td>\n",
       "      <td>-1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A7293</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>458</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00072</td>\n",
       "      <td>37</td>\n",
       "      <td>-2395.000</td>\n",
       "      <td>-1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A19365</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>622</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00089</td>\n",
       "      <td>19</td>\n",
       "      <td>-3995.000</td>\n",
       "      <td>-1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A21166</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    StoreID        TransactionID  Line     Sales  Count           Date     SKU\n",
       "100     A29  11-7270-00007-00066     5 -2013.000 -1.000  3/11/12 00:00   A2729\n",
       "354     A19  11-7287-00025-00075     4 -1721.000 -1.000  3/11/12 00:00  A16911\n",
       "446     A27  11-7069-00020-00010     6 -4252.000 -1.000  3/11/12 00:00   A7293\n",
       "458     A29  11-7270-00007-00072    37 -2395.000 -1.000  3/11/12 00:00  A19365\n",
       "622     A29  11-7270-00007-00089    19 -3995.000 -1.000  3/11/12 00:00  A21166"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Count=-1\n",
    "data.loc[data['Count']==-1].head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from pandasql import sqldf\n",
    "pysqldf = lambda q: sqldf(q, globals())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "q = '''\n",
    "    select\n",
    "        d1.StoreID, d1.TransactionID, d1.Line, d1.Sales, d1.Count, d1.Date, d1.SKU\n",
    "    from \n",
    "        data d1\n",
    "    inner join\n",
    "        (select \n",
    "            TransactionID, SKU\n",
    "        from\n",
    "            data\n",
    "        where Count=-1) as d2\n",
    "    on \n",
    "        d1.TransactionID = d2.TransactionID and\n",
    "        d1.SKU = d2.SKU\n",
    "    '''\n",
    "data1 = pysqldf(q)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "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>StoreID</th>\n",
       "      <th>TransactionID</th>\n",
       "      <th>Line</th>\n",
       "      <th>Sales</th>\n",
       "      <th>Count</th>\n",
       "      <th>Date</th>\n",
       "      <th>SKU</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00066</td>\n",
       "      <td>3</td>\n",
       "      <td>2013.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A2729</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00066</td>\n",
       "      <td>4</td>\n",
       "      <td>2013.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A2729</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00066</td>\n",
       "      <td>5</td>\n",
       "      <td>-2013.000</td>\n",
       "      <td>-1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A2729</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>A19</td>\n",
       "      <td>11-7287-00025-00075</td>\n",
       "      <td>3</td>\n",
       "      <td>1721.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A16911</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>A19</td>\n",
       "      <td>11-7287-00025-00075</td>\n",
       "      <td>4</td>\n",
       "      <td>-1721.000</td>\n",
       "      <td>-1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A16911</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>A27</td>\n",
       "      <td>11-7069-00020-00010</td>\n",
       "      <td>1</td>\n",
       "      <td>4252.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A7293</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>A27</td>\n",
       "      <td>11-7069-00020-00010</td>\n",
       "      <td>6</td>\n",
       "      <td>-4252.000</td>\n",
       "      <td>-1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A7293</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00072</td>\n",
       "      <td>33</td>\n",
       "      <td>2395.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A19365</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00072</td>\n",
       "      <td>35</td>\n",
       "      <td>2395.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A19365</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00072</td>\n",
       "      <td>36</td>\n",
       "      <td>2395.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A19365</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00072</td>\n",
       "      <td>37</td>\n",
       "      <td>-2395.000</td>\n",
       "      <td>-1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A19365</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00089</td>\n",
       "      <td>4</td>\n",
       "      <td>3995.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A21166</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00089</td>\n",
       "      <td>19</td>\n",
       "      <td>-3995.000</td>\n",
       "      <td>-1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A21166</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00089</td>\n",
       "      <td>20</td>\n",
       "      <td>1995.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A21166</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00009-00038</td>\n",
       "      <td>4</td>\n",
       "      <td>2695.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A24548</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00009-00038</td>\n",
       "      <td>5</td>\n",
       "      <td>2695.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A24548</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00009-00038</td>\n",
       "      <td>6</td>\n",
       "      <td>-2695.000</td>\n",
       "      <td>-1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A24548</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00105</td>\n",
       "      <td>16</td>\n",
       "      <td>250.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A3664</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00105</td>\n",
       "      <td>17</td>\n",
       "      <td>250.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A3664</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00105</td>\n",
       "      <td>18</td>\n",
       "      <td>-250.000</td>\n",
       "      <td>-1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A3664</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   StoreID        TransactionID  Line     Sales  Count           Date     SKU\n",
       "0      A29  11-7270-00007-00066     3  2013.000  1.000  3/11/12 00:00   A2729\n",
       "1      A29  11-7270-00007-00066     4  2013.000  1.000  3/11/12 00:00   A2729\n",
       "2      A29  11-7270-00007-00066     5 -2013.000 -1.000  3/11/12 00:00   A2729\n",
       "3      A19  11-7287-00025-00075     3  1721.000  1.000  3/11/12 00:00  A16911\n",
       "4      A19  11-7287-00025-00075     4 -1721.000 -1.000  3/11/12 00:00  A16911\n",
       "5      A27  11-7069-00020-00010     1  4252.000  1.000  3/11/12 00:00   A7293\n",
       "6      A27  11-7069-00020-00010     6 -4252.000 -1.000  3/11/12 00:00   A7293\n",
       "7      A29  11-7270-00007-00072    33  2395.000  1.000  3/11/12 00:00  A19365\n",
       "8      A29  11-7270-00007-00072    35  2395.000  1.000  3/11/12 00:00  A19365\n",
       "9      A29  11-7270-00007-00072    36  2395.000  1.000  3/11/12 00:00  A19365\n",
       "10     A29  11-7270-00007-00072    37 -2395.000 -1.000  3/11/12 00:00  A19365\n",
       "11     A29  11-7270-00007-00089     4  3995.000  1.000  3/11/12 00:00  A21166\n",
       "12     A29  11-7270-00007-00089    19 -3995.000 -1.000  3/11/12 00:00  A21166\n",
       "13     A29  11-7270-00007-00089    20  1995.000  1.000  3/11/12 00:00  A21166\n",
       "14     A29  11-7270-00009-00038     4  2695.000  1.000  3/11/12 00:00  A24548\n",
       "15     A29  11-7270-00009-00038     5  2695.000  1.000  3/11/12 00:00  A24548\n",
       "16     A29  11-7270-00009-00038     6 -2695.000 -1.000  3/11/12 00:00  A24548\n",
       "17     A29  11-7270-00007-00105    16   250.000  1.000  3/11/12 00:00   A3664\n",
       "18     A29  11-7270-00007-00105    17   250.000  1.000  3/11/12 00:00   A3664\n",
       "19     A29  11-7270-00007-00105    18  -250.000 -1.000  3/11/12 00:00   A3664"
      ]
     },
     "execution_count": 120,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data1.head(20)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### We can see from the table, what count=-1 means that this customer didn't want this item at all. \n",
    "### There are several cases:\n",
    "###  1. The customer bought just 1 item and he didn't want this item during the  transaction; \n",
    "### 2. The customer bought a discouted item and he didn't want this item during the same transaction;\n",
    "### 3. The customer bought 2 or over 2 items of this product and he didn't want one of this item during the same transaction.\n",
    "### 4. The customer bought 1 item and he didn't want this item during the transaction, since he found the same item in a cheaper price.\n",
    "### 5. Cashier wrongly scanned one item twice."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<br><br>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Explore cases when count=-2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Count=-2\n",
    "q = '''\n",
    "    select\n",
    "        d1.StoreID, d1.TransactionID, d1.Line, d1.Sales, d1.Count, d1.Date, d1.SKU\n",
    "    from \n",
    "        data d1\n",
    "    inner join\n",
    "        (select \n",
    "            TransactionID, SKU\n",
    "        from\n",
    "            data\n",
    "        where Count=-2) as d2\n",
    "    on \n",
    "        d1.TransactionID = d2.TransactionID and\n",
    "        d1.SKU = d2.SKU\n",
    "    '''\n",
    "data2 = pysqldf(q)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "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>StoreID</th>\n",
       "      <th>TransactionID</th>\n",
       "      <th>Line</th>\n",
       "      <th>Sales</th>\n",
       "      <th>Count</th>\n",
       "      <th>Date</th>\n",
       "      <th>SKU</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>A17</td>\n",
       "      <td>11-7720-00017-00144</td>\n",
       "      <td>6</td>\n",
       "      <td>1600.000</td>\n",
       "      <td>2.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A5678</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>A17</td>\n",
       "      <td>11-7720-00017-00144</td>\n",
       "      <td>7</td>\n",
       "      <td>-1600.000</td>\n",
       "      <td>-2.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A5678</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>A19</td>\n",
       "      <td>11-7287-00024-00014</td>\n",
       "      <td>9</td>\n",
       "      <td>1400.000</td>\n",
       "      <td>2.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A6653</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>A19</td>\n",
       "      <td>11-7287-00024-00014</td>\n",
       "      <td>10</td>\n",
       "      <td>-1400.000</td>\n",
       "      <td>-2.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A6653</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00012-00113</td>\n",
       "      <td>2</td>\n",
       "      <td>-17990.000</td>\n",
       "      <td>-2.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A22728</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00012-00113</td>\n",
       "      <td>3</td>\n",
       "      <td>11495.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A22882</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00012-00113</td>\n",
       "      <td>4</td>\n",
       "      <td>11495.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A22882</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00012-00113</td>\n",
       "      <td>6</td>\n",
       "      <td>-22990.000</td>\n",
       "      <td>-2.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A22882</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00012-00113</td>\n",
       "      <td>7</td>\n",
       "      <td>17990.000</td>\n",
       "      <td>2.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A22882</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00018-00076</td>\n",
       "      <td>1</td>\n",
       "      <td>2800.000</td>\n",
       "      <td>2.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A1891</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00018-00076</td>\n",
       "      <td>3</td>\n",
       "      <td>-2800.000</td>\n",
       "      <td>-2.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A1891</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>A55</td>\n",
       "      <td>11-1105-00010-00112</td>\n",
       "      <td>1</td>\n",
       "      <td>8504.000</td>\n",
       "      <td>2.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A7293</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>A55</td>\n",
       "      <td>11-1105-00010-00112</td>\n",
       "      <td>3</td>\n",
       "      <td>-8504.000</td>\n",
       "      <td>-2.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A7293</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>A60</td>\n",
       "      <td>11-0160-00021-00086</td>\n",
       "      <td>5</td>\n",
       "      <td>2624.000</td>\n",
       "      <td>2.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A1092</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>A60</td>\n",
       "      <td>11-0160-00021-00086</td>\n",
       "      <td>6</td>\n",
       "      <td>-2624.000</td>\n",
       "      <td>-2.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A1092</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>A60</td>\n",
       "      <td>11-0160-00021-00182</td>\n",
       "      <td>1</td>\n",
       "      <td>3400.000</td>\n",
       "      <td>2.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A12968</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>A60</td>\n",
       "      <td>11-0160-00021-00182</td>\n",
       "      <td>3</td>\n",
       "      <td>-3400.000</td>\n",
       "      <td>-2.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A12968</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>A65</td>\n",
       "      <td>11-3571-00008-00008</td>\n",
       "      <td>2</td>\n",
       "      <td>-19990.000</td>\n",
       "      <td>-2.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A11920</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>A65</td>\n",
       "      <td>11-3571-00008-00048</td>\n",
       "      <td>1</td>\n",
       "      <td>-9000.000</td>\n",
       "      <td>-2.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A25153</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>A65</td>\n",
       "      <td>11-3571-00012-00215</td>\n",
       "      <td>1</td>\n",
       "      <td>2622.000</td>\n",
       "      <td>2.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A2401</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   StoreID        TransactionID  Line      Sales  Count           Date     SKU\n",
       "0      A17  11-7720-00017-00144     6   1600.000  2.000  3/11/12 00:00   A5678\n",
       "1      A17  11-7720-00017-00144     7  -1600.000 -2.000  3/11/12 00:00   A5678\n",
       "2      A19  11-7287-00024-00014     9   1400.000  2.000  3/11/12 00:00   A6653\n",
       "3      A19  11-7287-00024-00014    10  -1400.000 -2.000  3/11/12 00:00   A6653\n",
       "4      A29  11-7270-00012-00113     2 -17990.000 -2.000  3/11/12 00:00  A22728\n",
       "5      A29  11-7270-00012-00113     3  11495.000  1.000  3/11/12 00:00  A22882\n",
       "6      A29  11-7270-00012-00113     4  11495.000  1.000  3/11/12 00:00  A22882\n",
       "7      A29  11-7270-00012-00113     6 -22990.000 -2.000  3/11/12 00:00  A22882\n",
       "8      A29  11-7270-00012-00113     7  17990.000  2.000  3/11/12 00:00  A22882\n",
       "9      A29  11-7270-00018-00076     1   2800.000  2.000  3/11/12 00:00   A1891\n",
       "10     A29  11-7270-00018-00076     3  -2800.000 -2.000  3/11/12 00:00   A1891\n",
       "11     A55  11-1105-00010-00112     1   8504.000  2.000  3/11/12 00:00   A7293\n",
       "12     A55  11-1105-00010-00112     3  -8504.000 -2.000  3/11/12 00:00   A7293\n",
       "13     A60  11-0160-00021-00086     5   2624.000  2.000  3/11/12 00:00   A1092\n",
       "14     A60  11-0160-00021-00086     6  -2624.000 -2.000  3/11/12 00:00   A1092\n",
       "15     A60  11-0160-00021-00182     1   3400.000  2.000  3/11/12 00:00  A12968\n",
       "16     A60  11-0160-00021-00182     3  -3400.000 -2.000  3/11/12 00:00  A12968\n",
       "17     A65  11-3571-00008-00008     2 -19990.000 -2.000  3/11/12 00:00  A11920\n",
       "18     A65  11-3571-00008-00048     1  -9000.000 -2.000  3/11/12 00:00  A25153\n",
       "19     A65  11-3571-00012-00215     1   2622.000  2.000  3/11/12 00:00   A2401"
      ]
     },
     "execution_count": 87,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data2.head(20)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "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>StoreID</th>\n",
       "      <th>TransactionID</th>\n",
       "      <th>Line</th>\n",
       "      <th>Sales</th>\n",
       "      <th>Count</th>\n",
       "      <th>Date</th>\n",
       "      <th>SKU</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>5003</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00012-00113</td>\n",
       "      <td>2</td>\n",
       "      <td>-17990.000</td>\n",
       "      <td>-2.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A22728</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     StoreID        TransactionID  Line      Sales  Count           Date  \\\n",
       "5003     A29  11-7270-00012-00113     2 -17990.000 -2.000  3/11/12 00:00   \n",
       "\n",
       "         SKU  \n",
       "5003  A22728  "
      ]
     },
     "execution_count": 92,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.loc[(data['SKU']=='A22728')&(data['TransactionID']=='11-7270-00012-00113')]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "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>StoreID</th>\n",
       "      <th>TransactionID</th>\n",
       "      <th>Line</th>\n",
       "      <th>Sales</th>\n",
       "      <th>Count</th>\n",
       "      <th>Date</th>\n",
       "      <th>SKU</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>5003</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00012-00113</td>\n",
       "      <td>2</td>\n",
       "      <td>-17990.000</td>\n",
       "      <td>-2.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A22728</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>223347</th>\n",
       "      <td>A15</td>\n",
       "      <td>12-7054-00006-00199</td>\n",
       "      <td>7</td>\n",
       "      <td>11495.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3/12/12 00:00</td>\n",
       "      <td>A22728</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>604225</th>\n",
       "      <td>A43</td>\n",
       "      <td>12-7285-00006-00058</td>\n",
       "      <td>5</td>\n",
       "      <td>11495.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3/12/12 00:00</td>\n",
       "      <td>A22728</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       StoreID        TransactionID  Line      Sales  Count           Date  \\\n",
       "5003       A29  11-7270-00012-00113     2 -17990.000 -2.000  3/11/12 00:00   \n",
       "223347     A15  12-7054-00006-00199     7  11495.000  1.000  3/12/12 00:00   \n",
       "604225     A43  12-7285-00006-00058     5  11495.000  1.000  3/12/12 00:00   \n",
       "\n",
       "           SKU  \n",
       "5003    A22728  \n",
       "223347  A22728  \n",
       "604225  A22728  "
      ]
     },
     "execution_count": 101,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.loc[(data['SKU']=='A22728')]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "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>StoreID</th>\n",
       "      <th>TransactionID</th>\n",
       "      <th>Line</th>\n",
       "      <th>Sales</th>\n",
       "      <th>Count</th>\n",
       "      <th>Date</th>\n",
       "      <th>SKU</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>23312</th>\n",
       "      <td>A65</td>\n",
       "      <td>11-3571-00008-00008</td>\n",
       "      <td>2</td>\n",
       "      <td>-19990.000</td>\n",
       "      <td>-2.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A11920</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      StoreID        TransactionID  Line      Sales  Count           Date  \\\n",
       "23312     A65  11-3571-00008-00008     2 -19990.000 -2.000  3/11/12 00:00   \n",
       "\n",
       "          SKU  \n",
       "23312  A11920  "
      ]
     },
     "execution_count": 95,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.loc[(data['SKU']=='A11920')&(data['TransactionID']=='11-3571-00008-00008')]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "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>StoreID</th>\n",
       "      <th>TransactionID</th>\n",
       "      <th>Line</th>\n",
       "      <th>Sales</th>\n",
       "      <th>Count</th>\n",
       "      <th>Date</th>\n",
       "      <th>SKU</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>17859</th>\n",
       "      <td>A65</td>\n",
       "      <td>11-3571-00002-00340</td>\n",
       "      <td>1</td>\n",
       "      <td>9995.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A11920</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20424</th>\n",
       "      <td>A65</td>\n",
       "      <td>11-3571-00001-00209</td>\n",
       "      <td>1</td>\n",
       "      <td>9995.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A11920</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20425</th>\n",
       "      <td>A65</td>\n",
       "      <td>11-3571-00001-00209</td>\n",
       "      <td>2</td>\n",
       "      <td>9995.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A11920</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23312</th>\n",
       "      <td>A65</td>\n",
       "      <td>11-3571-00008-00008</td>\n",
       "      <td>2</td>\n",
       "      <td>-19990.000</td>\n",
       "      <td>-2.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A11920</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>814110</th>\n",
       "      <td>A65</td>\n",
       "      <td>12-3572-00001-00226</td>\n",
       "      <td>1</td>\n",
       "      <td>9995.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3/12/12 00:00</td>\n",
       "      <td>A11920</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       StoreID        TransactionID  Line      Sales  Count           Date  \\\n",
       "17859      A65  11-3571-00002-00340     1   9995.000  1.000  3/11/12 00:00   \n",
       "20424      A65  11-3571-00001-00209     1   9995.000  1.000  3/11/12 00:00   \n",
       "20425      A65  11-3571-00001-00209     2   9995.000  1.000  3/11/12 00:00   \n",
       "23312      A65  11-3571-00008-00008     2 -19990.000 -2.000  3/11/12 00:00   \n",
       "814110     A65  12-3572-00001-00226     1   9995.000  1.000  3/12/12 00:00   \n",
       "\n",
       "           SKU  \n",
       "17859   A11920  \n",
       "20424   A11920  \n",
       "20425   A11920  \n",
       "23312   A11920  \n",
       "814110  A11920  "
      ]
     },
     "execution_count": 99,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.loc[(data['SKU']=='A11920')&(data['StoreID']=='A65')]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Several Cases:\n",
    "### 1. Customers bought 2 or over 2 items of a product and didn't want 2 of them immediately.\n",
    "### 2. Customers bought 2 or over 2 items of a product and decided to return items after the transaction.\n",
    "### 3. Missing Records: There are some obervations only indicate they didn't want to buy them, but no observations indicate they already have bought them (when count is positive)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<br><br>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Check Missing Values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 123,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "StoreID          0\n",
       "TransactionID    0\n",
       "Line             0\n",
       "Sales            0\n",
       "Count            0\n",
       "Date             0\n",
       "SKU              0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 123,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.isnull().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<br><br>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Explore Outliers"
   ]
  },
  {
   "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>Line</th>\n",
       "      <th>Sales</th>\n",
       "      <th>Count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>1048575.000</td>\n",
       "      <td>1048575.000</td>\n",
       "      <td>1048575.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>10.727</td>\n",
       "      <td>1720.513</td>\n",
       "      <td>0.896</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>15.550</td>\n",
       "      <td>3646.003</td>\n",
       "      <td>1.129</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>1.000</td>\n",
       "      <td>-99995.000</td>\n",
       "      <td>-120.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>3.000</td>\n",
       "      <td>395.000</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>7.000</td>\n",
       "      <td>1295.000</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>14.000</td>\n",
       "      <td>2195.000</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>823.000</td>\n",
       "      <td>99995.000</td>\n",
       "      <td>120.000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             Line       Sales       Count\n",
       "count 1048575.000 1048575.000 1048575.000\n",
       "mean       10.727    1720.513       0.896\n",
       "std        15.550    3646.003       1.129\n",
       "min         1.000  -99995.000    -120.000\n",
       "25%         3.000     395.000       1.000\n",
       "50%         7.000    1295.000       1.000\n",
       "75%        14.000    2195.000       1.000\n",
       "max       823.000   99995.000     120.000"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "def find_outliers_tukey(x):\n",
    "    q1 = np.percentile(x,25)\n",
    "    q3 = np.percentile(x,75)\n",
    "    iqr = q3 - q1\n",
    "    floor = q1 - 1.5*iqr\n",
    "    ceiling = q3 + 1.5*iqr\n",
    "    outlier_indices = list(x.index[(x < floor) | (x > ceiling)])\n",
    "    outlier_values = list(x[outlier_indices])\n",
    "    \n",
    "    return outlier_indices, outlier_values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Feature 'Line' first and last 10 outliers：[31 32 33 34 35 36 37 38 39 40][796 797 805 806 808 809 811 812 822 823]\n",
      "\n",
      "Number of outliers: 66220\n",
      "\n",
      "Feature 'Sales' first and last 10 outliers：[-99995.00 -99900.00 -89900.00 -88800.00 -86400.00 -84995.00 -84900.00\n",
      " -80000.00 -79995.00 -79950.00][86900.00 88000.00 89900.00 90000.00 92120.00 94900.00 96000.00 99000.00\n",
      " 99900.00 99995.00]\n",
      "\n",
      "Number of outliers: 85599\n",
      "\n",
      "Feature 'Count' first and last 10 outliers：[-120.00 -91.00 -60.00 -54.00 -48.00 -40.00 -39.00 -36.00 -33.00 -32.00][58.00 60.00 65.00 70.00 72.00 80.00 90.00 91.00 99.00 120.00]\n",
      "\n",
      "Number of outliers: 198336\n",
      "\n"
     ]
    }
   ],
   "source": [
    "np.set_printoptions(formatter={'float_kind':lambda x: \"%.2f\" % x})\n",
    "outlier_var = ['Line','Sales','Count']\n",
    "for x in outlier_var:\n",
    "    tukey_indices, tukey_values = find_outliers_tukey(data[x])\n",
    "    a = np.sort(tukey_values)\n",
    "    print \"Feature '{0}' first and last 10 outliers：{1}{2}\\n\".format(x,np.unique(a)[:10],np.unique(a)[-10:])\n",
    "    print \"Number of outliers: {0}\\n\".format(len(tukey_values))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 151,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "5500.0"
      ]
     },
     "execution_count": 151,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.percentile(data['Sales'],95)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "-825.0"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.percentile(data['Sales'],5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 154,
   "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>StoreID</th>\n",
       "      <th>TransactionID</th>\n",
       "      <th>Line</th>\n",
       "      <th>Sales</th>\n",
       "      <th>Count</th>\n",
       "      <th>Date</th>\n",
       "      <th>SKU</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>653883</th>\n",
       "      <td>A46</td>\n",
       "      <td>12-7027-00013-00092</td>\n",
       "      <td>3</td>\n",
       "      <td>60000.000</td>\n",
       "      <td>120.000</td>\n",
       "      <td>3/12/12 00:00</td>\n",
       "      <td>A10809</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>653885</th>\n",
       "      <td>A46</td>\n",
       "      <td>12-7027-00013-00092</td>\n",
       "      <td>6</td>\n",
       "      <td>-60000.000</td>\n",
       "      <td>-120.000</td>\n",
       "      <td>3/12/12 00:00</td>\n",
       "      <td>A10809</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       StoreID        TransactionID  Line      Sales    Count           Date  \\\n",
       "653883     A46  12-7027-00013-00092     3  60000.000  120.000  3/12/12 00:00   \n",
       "653885     A46  12-7027-00013-00092     6 -60000.000 -120.000  3/12/12 00:00   \n",
       "\n",
       "           SKU  \n",
       "653883  A10809  \n",
       "653885  A10809  "
      ]
     },
     "execution_count": 154,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.loc[(data['Count']==120)|(data['Count']==-120)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "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>StoreID</th>\n",
       "      <th>TransactionID</th>\n",
       "      <th>Line</th>\n",
       "      <th>Sales</th>\n",
       "      <th>Count</th>\n",
       "      <th>Date</th>\n",
       "      <th>SKU</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>655356</th>\n",
       "      <td>A47</td>\n",
       "      <td>12-7304-00002-00016</td>\n",
       "      <td>10</td>\n",
       "      <td>89900.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3/12/12 00:00</td>\n",
       "      <td>A17041</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>655357</th>\n",
       "      <td>A47</td>\n",
       "      <td>12-7304-00002-00016</td>\n",
       "      <td>11</td>\n",
       "      <td>89900.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3/12/12 00:00</td>\n",
       "      <td>A17041</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>655358</th>\n",
       "      <td>A47</td>\n",
       "      <td>12-7304-00002-00016</td>\n",
       "      <td>12</td>\n",
       "      <td>-89900.000</td>\n",
       "      <td>-1.000</td>\n",
       "      <td>3/12/12 00:00</td>\n",
       "      <td>A17041</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       StoreID        TransactionID  Line      Sales  Count           Date  \\\n",
       "655356     A47  12-7304-00002-00016    10  89900.000  1.000  3/12/12 00:00   \n",
       "655357     A47  12-7304-00002-00016    11  89900.000  1.000  3/12/12 00:00   \n",
       "655358     A47  12-7304-00002-00016    12 -89900.000 -1.000  3/12/12 00:00   \n",
       "\n",
       "           SKU  \n",
       "655356  A17041  \n",
       "655357  A17041  \n",
       "655358  A17041  "
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.loc[(data['SKU']=='A17041')&(data['TransactionID']=='12-7304-00002-00016')]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "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>StoreID</th>\n",
       "      <th>TransactionID</th>\n",
       "      <th>Line</th>\n",
       "      <th>Sales</th>\n",
       "      <th>Count</th>\n",
       "      <th>Date</th>\n",
       "      <th>SKU</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>844638</th>\n",
       "      <td>A67</td>\n",
       "      <td>12-3097-00008-00087</td>\n",
       "      <td>2</td>\n",
       "      <td>-88800.000</td>\n",
       "      <td>-2.000</td>\n",
       "      <td>3/12/12 00:00</td>\n",
       "      <td>A24153</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       StoreID        TransactionID  Line      Sales  Count           Date  \\\n",
       "844638     A67  12-3097-00008-00087     2 -88800.000 -2.000  3/12/12 00:00   \n",
       "\n",
       "           SKU  \n",
       "844638  A24153  "
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# There are some cases where only have negative sales but no corresponding positive sales with the same transactionID and SKU\n",
    "data.loc[(data['SKU']=='A24153')&(data['TransactionID']=='12-3097-00008-00087')]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 185,
   "metadata": {
    "collapsed": false,
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1720.5125972246444"
      ]
     },
     "execution_count": 185,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Check how much the mean changes after we drop the outliers in 'Sales'\n",
    "data['Sales'].mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 191,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "tukey_indices, tukey_values = find_outliers_tukey(data['Sales'])\n",
    "data.drop(tukey_indices,inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 193,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1208.8544643218809"
      ]
     },
     "execution_count": 193,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# After Drop\n",
    "data['Sales'].mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### The mean of sales appears to change a lot if we drop those outliers, so we can't treat them as negligible.\n",
    "### We can't also replace those outliers with 99 percentile value of that feature. Since there are both positive and negative outliers and some absolute value are same, which means it doesn't infulence average revenue."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<br><br>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Average Revenue per Basket"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "data = pd.read_csv('RetailerPromotionStrategy_data.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 301,
   "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>StoreID</th>\n",
       "      <th>TransactionID</th>\n",
       "      <th>Line</th>\n",
       "      <th>Sales</th>\n",
       "      <th>Count</th>\n",
       "      <th>Date</th>\n",
       "      <th>SKU</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00057</td>\n",
       "      <td>36</td>\n",
       "      <td>850.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A21773</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00057</td>\n",
       "      <td>39</td>\n",
       "      <td>1500.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A18009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00057</td>\n",
       "      <td>40</td>\n",
       "      <td>1500.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A25592</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00057</td>\n",
       "      <td>41</td>\n",
       "      <td>1600.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A1449</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00057</td>\n",
       "      <td>42</td>\n",
       "      <td>1600.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A1449</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  StoreID        TransactionID  Line    Sales  Count           Date     SKU\n",
       "0     A29  11-7270-00007-00057    36  850.000  1.000  3/11/12 00:00  A21773\n",
       "1     A29  11-7270-00007-00057    39 1500.000  1.000  3/11/12 00:00  A18009\n",
       "2     A29  11-7270-00007-00057    40 1500.000  1.000  3/11/12 00:00  A25592\n",
       "3     A29  11-7270-00007-00057    41 1600.000  1.000  3/11/12 00:00   A1449\n",
       "4     A29  11-7270-00007-00057    42 1600.000  1.000  3/11/12 00:00   A1449"
      ]
     },
     "execution_count": 301,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 162,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "197714"
      ]
     },
     "execution_count": 162,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(data.TransactionID.unique())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 222,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# Select transactions who have discount items and total sales>0\n",
    "q='''\n",
    "  select\n",
    "      a.StoreID,\n",
    "      a.TransactionID,\n",
    "      sum(a.Count) as Counts,\n",
    "      sum(a.Sales) as Total\n",
    "  from\n",
    "      data a\n",
    "  inner join\n",
    "      (select StoreID, TransactionID\n",
    "       from data\n",
    "       where Count=0) b\n",
    "  on\n",
    "      a.StoreID=b.StoreID and a.TransactionID=b.TransactionID\n",
    "  group by\n",
    "      a.StoreID, \n",
    "      a.TransactionID\n",
    "  having\n",
    "      Total>0\n",
    "  '''\n",
    "data1= pysqldf(q)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 274,
   "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>StoreID</th>\n",
       "      <th>TransactionID</th>\n",
       "      <th>Counts</th>\n",
       "      <th>Total</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>A01</td>\n",
       "      <td>12-7149-00001-00004</td>\n",
       "      <td>12.000</td>\n",
       "      <td>6960.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>A01</td>\n",
       "      <td>12-7149-00001-00005</td>\n",
       "      <td>144.000</td>\n",
       "      <td>331200.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>A01</td>\n",
       "      <td>12-7149-00001-00006</td>\n",
       "      <td>24.000</td>\n",
       "      <td>33420.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>A01</td>\n",
       "      <td>12-7149-00001-00013</td>\n",
       "      <td>20.000</td>\n",
       "      <td>39960.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>A01</td>\n",
       "      <td>12-7149-00001-00016</td>\n",
       "      <td>4.000</td>\n",
       "      <td>6648.000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  StoreID        TransactionID  Counts      Total\n",
       "0     A01  12-7149-00001-00004  12.000   6960.000\n",
       "1     A01  12-7149-00001-00005 144.000 331200.000\n",
       "2     A01  12-7149-00001-00006  24.000  33420.000\n",
       "3     A01  12-7149-00001-00013  20.000  39960.000\n",
       "4     A01  12-7149-00001-00016   4.000   6648.000"
      ]
     },
     "execution_count": 274,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data1.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 327,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Average sales per basket for those transactions include discount items is 45388.6339343.\n"
     ]
    }
   ],
   "source": [
    "# Average sales per basket (there are discount items in transactions)\n",
    "avg_rev_discount = data1.Total.sum() / data1.TransactionID.size\n",
    "\n",
    "print 'Average sales per basket for those transactions include discount items is {0}.'.format(avg_rev_discount)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 352,
   "metadata": {
    "collapsed": false,
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "# Average sales per basket in different store\n",
    "avg_rev_discount_store = data1.groupby('StoreID').Total.sum() / data1.groupby('StoreID').TransactionID.size()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 319,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# Select transactions which don't have any discount items\n",
    "\n",
    "q = '''\n",
    "    select\n",
    "      t1.StoreID,\n",
    "      t1.TransactionID,\n",
    "      sum(t1.Count) as Counts,\n",
    "      sum(t1.Sales) as Total\n",
    "    from\n",
    "    (select\n",
    "        a.StoreID, a. TransactionID, a.Line, a.Sales, a.Count, a.Date, a.SKU\n",
    "    from \n",
    "        data a\n",
    "    left join\n",
    "        data1 b\n",
    "    on \n",
    "        a.StoreID=b.StoreID and a.TransactionID=b.TransactionID\n",
    "    where\n",
    "        b.StoreID is null) t1\n",
    "    group by\n",
    "      t1.StoreID, \n",
    "      t1.TransactionID\n",
    "    having\n",
    "      Total>0    \n",
    "    '''\n",
    "data2 = pysqldf(q)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 320,
   "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>StoreID</th>\n",
       "      <th>TransactionID</th>\n",
       "      <th>Counts</th>\n",
       "      <th>Total</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>A01</td>\n",
       "      <td>12-7149-00001-00002</td>\n",
       "      <td>4.000</td>\n",
       "      <td>6645.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>A01</td>\n",
       "      <td>12-7149-00001-00003</td>\n",
       "      <td>1.000</td>\n",
       "      <td>1000.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>A01</td>\n",
       "      <td>12-7149-00001-00007</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3300.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>A01</td>\n",
       "      <td>12-7149-00001-00008</td>\n",
       "      <td>4.000</td>\n",
       "      <td>3871.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>A01</td>\n",
       "      <td>12-7149-00001-00009</td>\n",
       "      <td>7.000</td>\n",
       "      <td>6864.000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  StoreID        TransactionID  Counts    Total\n",
       "0     A01  12-7149-00001-00002   4.000 6645.000\n",
       "1     A01  12-7149-00001-00003   1.000 1000.000\n",
       "2     A01  12-7149-00001-00007   1.000 3300.000\n",
       "3     A01  12-7149-00001-00008   4.000 3871.000\n",
       "4     A01  12-7149-00001-00009   7.000 6864.000"
      ]
     },
     "execution_count": 320,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data2.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 409,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "55631"
      ]
     },
     "execution_count": 409,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(data1.TransactionID.unique())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 321,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "136226"
      ]
     },
     "execution_count": 321,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(data2.TransactionID.unique())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 326,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Average sales per basket for those transactions include discount items is 7490.94903763.\n"
     ]
    }
   ],
   "source": [
    "# Average sales per basket (there aren't any discount items in transactions)\n",
    "avg_rev_discount1 = data2.Total.sum() / data2.TransactionID.size\n",
    "\n",
    "print 'Average sales per basket for those transactions include discount items is {0}.'.format(avg_rev_discount1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 353,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Average sales per basket without discount items in different store\n",
    "avg_rev_discount_store1 = data2.groupby('StoreID').Total.sum() / data2.groupby('StoreID').TransactionID.size()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Visualization\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 345,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZUAAAEICAYAAACXo2mmAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nyfOfrAZOpqEAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x15889a450>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots()\n",
    "x=np.arange(2)\n",
    "width=0.1\n",
    "rects1 = ax.bar(x,[avg_rev_discount1,avg_rev_discount])\n",
    "ax.set_ylabel('Average Revenue Per Basket')\n",
    "ax.set_ylim(0,50000)\n",
    "ax.set_title('Will Discount Items Drive Up Average Revenue?')\n",
    "ax.set_xticks(x+width/2 )\n",
    "ax.set_xticklabels(('without discount','with discount'))\n",
    "#annotate\n",
    "def autolabel(rects):\n",
    "    for rect in rects:\n",
    "        height=rect.get_height()\n",
    "        ax.text(rect.get_x()+rect.get_width()/2, 1.035*height, '{:.2f}$'.format(height), ha='center', va='bottom')\n",
    "autolabel(rects1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 410,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x17876cb10>"
      ]
     },
     "execution_count": 410,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABKEAAAJaCAYAAAD3QlTHAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nFAAAAABF9z97M3Y/rDi9lwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1764a6d90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualize average sales per basket in different store\n",
    "fig, ax = plt.subplots(figsize=(20,10))\n",
    "x = np.arange(len(avg_rev_discount_store))\n",
    "width=0.5\n",
    "rects1 = ax.bar(x,avg_rev_discount_store,width=(x[1]-x[0])*0.8)\n",
    "ax.set_ylabel('Average Revenue Per Basket')\n",
    "ax.set_ylim(0,175000)\n",
    "ax.set_xlabel('Store')\n",
    "ax.set_title('Will Discount Items Drive Up Average Revenue in Each Store?')\n",
    "\n",
    "rects2 = ax.bar(x,avg_rev_discount_store1)\n",
    "\n",
    "ax.legend((rects1[0],rects2[0]),('With Discount','Without Discount'))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### According to above plots, we can conclude that discount items will drive up average sales per basket."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<br><br>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Average Sales Per Item for Each Product"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "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>StoreID</th>\n",
       "      <th>TransactionID</th>\n",
       "      <th>Line</th>\n",
       "      <th>Sales</th>\n",
       "      <th>Count</th>\n",
       "      <th>Date</th>\n",
       "      <th>SKU</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00057</td>\n",
       "      <td>36</td>\n",
       "      <td>850.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A21773</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00057</td>\n",
       "      <td>39</td>\n",
       "      <td>1500.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A18009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00057</td>\n",
       "      <td>40</td>\n",
       "      <td>1500.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A25592</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00057</td>\n",
       "      <td>41</td>\n",
       "      <td>1600.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A1449</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00057</td>\n",
       "      <td>42</td>\n",
       "      <td>1600.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A1449</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  StoreID        TransactionID  Line    Sales  Count           Date     SKU\n",
       "0     A29  11-7270-00007-00057    36  850.000  1.000  3/11/12 00:00  A21773\n",
       "1     A29  11-7270-00007-00057    39 1500.000  1.000  3/11/12 00:00  A18009\n",
       "2     A29  11-7270-00007-00057    40 1500.000  1.000  3/11/12 00:00  A25592\n",
       "3     A29  11-7270-00007-00057    41 1600.000  1.000  3/11/12 00:00   A1449\n",
       "4     A29  11-7270-00007-00057    42 1600.000  1.000  3/11/12 00:00   A1449"
      ]
     },
     "execution_count": 92,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# Select discount items and group by SKU\n",
    "q='''\n",
    "  select\n",
    "      a.SKU,\n",
    "      sum(a.Count) as Counts,\n",
    "      sum(a.Sales) as Total\n",
    "  from\n",
    "      data a\n",
    "  inner join\n",
    "      (select SKU, TransactionID\n",
    "       from data\n",
    "       where Count=0) b\n",
    "  on\n",
    "      a.SKU=b.SKU and a.TransactionID=b.TransactionID\n",
    "  group by\n",
    "      a.SKU\n",
    "  having\n",
    "      Total>0\n",
    "  '''\n",
    "data1= pysqldf(q)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "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>SKU</th>\n",
       "      <th>Counts</th>\n",
       "      <th>Total</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>A1</td>\n",
       "      <td>26.000</td>\n",
       "      <td>30392.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>A10004</td>\n",
       "      <td>36.000</td>\n",
       "      <td>54334.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>A10022</td>\n",
       "      <td>3.000</td>\n",
       "      <td>5000.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>A10045</td>\n",
       "      <td>14.000</td>\n",
       "      <td>18708.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>A10048</td>\n",
       "      <td>90.000</td>\n",
       "      <td>112500.000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      SKU  Counts      Total\n",
       "0      A1  26.000  30392.000\n",
       "1  A10004  36.000  54334.000\n",
       "2  A10022   3.000   5000.000\n",
       "3  A10045  14.000  18708.000\n",
       "4  A10048  90.000 112500.000"
      ]
     },
     "execution_count": 95,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data1.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "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>SKU</th>\n",
       "      <th>Counts</th>\n",
       "      <th>Total</th>\n",
       "      <th>Average_sales</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>A1</td>\n",
       "      <td>26.000</td>\n",
       "      <td>30392.000</td>\n",
       "      <td>1168.923</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>A10004</td>\n",
       "      <td>36.000</td>\n",
       "      <td>54334.000</td>\n",
       "      <td>1509.278</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>A10022</td>\n",
       "      <td>3.000</td>\n",
       "      <td>5000.000</td>\n",
       "      <td>1666.667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>A10045</td>\n",
       "      <td>14.000</td>\n",
       "      <td>18708.000</td>\n",
       "      <td>1336.286</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>A10048</td>\n",
       "      <td>90.000</td>\n",
       "      <td>112500.000</td>\n",
       "      <td>1250.000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      SKU  Counts      Total  Average_sales\n",
       "0      A1  26.000  30392.000       1168.923\n",
       "1  A10004  36.000  54334.000       1509.278\n",
       "2  A10022   3.000   5000.000       1666.667\n",
       "3  A10045  14.000  18708.000       1336.286\n",
       "4  A10048  90.000 112500.000       1250.000"
      ]
     },
     "execution_count": 97,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# average sales per item\n",
    "data1['Average_sales'] = data1.Total / data1.Counts\n",
    "data1.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Select same items in data1 that are not discounted\n",
    "q='''\n",
    "  select\n",
    "      a.SKU,\n",
    "      sum(a.Count) as Counts,\n",
    "      sum(a.Sales) as Total\n",
    "  from\n",
    "      data a\n",
    "  inner join\n",
    "      (select SKU, TransactionID\n",
    "       from data\n",
    "       where Count=0) b\n",
    "  on\n",
    "      a.SKU = b.SKU and a.TransactionID != b.TransactionID\n",
    "  group by\n",
    "      a.SKU\n",
    "  having\n",
    "      Total>0\n",
    "  '''\n",
    "data2= pysqldf(q)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "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>SKU</th>\n",
       "      <th>Counts</th>\n",
       "      <th>Total</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>A1</td>\n",
       "      <td>1534.000</td>\n",
       "      <td>1863248.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>A10004</td>\n",
       "      <td>3744.000</td>\n",
       "      <td>8491286.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>A10022</td>\n",
       "      <td>30.000</td>\n",
       "      <td>42955.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>A10023</td>\n",
       "      <td>297.000</td>\n",
       "      <td>206415.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>A10025</td>\n",
       "      <td>3.000</td>\n",
       "      <td>75000.000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      SKU   Counts       Total\n",
       "0      A1 1534.000 1863248.000\n",
       "1  A10004 3744.000 8491286.000\n",
       "2  A10022   30.000   42955.000\n",
       "3  A10023  297.000  206415.000\n",
       "4  A10025    3.000   75000.000"
      ]
     },
     "execution_count": 99,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data2.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "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>StoreID</th>\n",
       "      <th>TransactionID</th>\n",
       "      <th>Line</th>\n",
       "      <th>Sales</th>\n",
       "      <th>Count</th>\n",
       "      <th>Date</th>\n",
       "      <th>SKU</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>611272</th>\n",
       "      <td>A43</td>\n",
       "      <td>12-7285-00011-00011</td>\n",
       "      <td>1</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>3/12/12 00:00</td>\n",
       "      <td>A10023</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       StoreID        TransactionID  Line  Sales  Count           Date     SKU\n",
       "611272     A43  12-7285-00011-00011     1  0.000  0.000  3/12/12 00:00  A10023"
      ]
     },
     "execution_count": 111,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data[(data['SKU']=='A10023')& (data['Count']==0)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "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>StoreID</th>\n",
       "      <th>TransactionID</th>\n",
       "      <th>Line</th>\n",
       "      <th>Sales</th>\n",
       "      <th>Count</th>\n",
       "      <th>Date</th>\n",
       "      <th>SKU</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>611272</th>\n",
       "      <td>A43</td>\n",
       "      <td>12-7285-00011-00011</td>\n",
       "      <td>1</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>3/12/12 00:00</td>\n",
       "      <td>A10023</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       StoreID        TransactionID  Line  Sales  Count           Date     SKU\n",
       "611272     A43  12-7285-00011-00011     1  0.000  0.000  3/12/12 00:00  A10023"
      ]
     },
     "execution_count": 113,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Some cases will have Sales=0 when Count=0, we shoulde exclude these SKUs\n",
    "data[(data['TransactionID']=='12-7285-00011-00011')& (data['SKU']=='A10023')]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(4141, 4)"
      ]
     },
     "execution_count": 107,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data2.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "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>SKU</th>\n",
       "      <th>Counts</th>\n",
       "      <th>Total</th>\n",
       "      <th>Average_sales</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>A1</td>\n",
       "      <td>1534.000</td>\n",
       "      <td>1863248.000</td>\n",
       "      <td>1214.634</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>A10004</td>\n",
       "      <td>3744.000</td>\n",
       "      <td>8491286.000</td>\n",
       "      <td>2267.972</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>A10022</td>\n",
       "      <td>30.000</td>\n",
       "      <td>42955.000</td>\n",
       "      <td>1431.833</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>A10023</td>\n",
       "      <td>297.000</td>\n",
       "      <td>206415.000</td>\n",
       "      <td>695.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>A10025</td>\n",
       "      <td>3.000</td>\n",
       "      <td>75000.000</td>\n",
       "      <td>25000.000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      SKU   Counts       Total  Average_sales\n",
       "0      A1 1534.000 1863248.000       1214.634\n",
       "1  A10004 3744.000 8491286.000       2267.972\n",
       "2  A10022   30.000   42955.000       1431.833\n",
       "3  A10023  297.000  206415.000        695.000\n",
       "4  A10025    3.000   75000.000      25000.000"
      ]
     },
     "execution_count": 104,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# average sales per item\n",
    "data2['Average_sales'] = data2.Total / data2.Counts\n",
    "data2.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 128,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# Inner Join data1 and data2 with same SKU\n",
    "q='''\n",
    "  select\n",
    "    a.SKU, a.Average_sales as Discount_Average_Sales , b.Average_sales as Non_Discount_Average_Sales\n",
    "  from\n",
    "      data1 a\n",
    "  inner join\n",
    "      data2 b\n",
    "  on\n",
    "      a.SKU = b.SKU\n",
    "  '''\n",
    "data3= pysqldf(q)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "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>SKU</th>\n",
       "      <th>Discount_Average_Sales</th>\n",
       "      <th>Non_Discount_Average_Sales</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>A1</td>\n",
       "      <td>1168.923</td>\n",
       "      <td>1214.634</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>A10004</td>\n",
       "      <td>1509.278</td>\n",
       "      <td>2267.972</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>A10022</td>\n",
       "      <td>1666.667</td>\n",
       "      <td>1431.833</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>A10045</td>\n",
       "      <td>1336.286</td>\n",
       "      <td>1602.636</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>A10048</td>\n",
       "      <td>1250.000</td>\n",
       "      <td>1376.146</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      SKU  Discount_Average_Sales  Non_Discount_Average_Sales\n",
       "0      A1                1168.923                    1214.634\n",
       "1  A10004                1509.278                    2267.972\n",
       "2  A10022                1666.667                    1431.833\n",
       "3  A10045                1336.286                    1602.636\n",
       "4  A10048                1250.000                    1376.146"
      ]
     },
     "execution_count": 129,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data3.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 143,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABKIAAAI/CAYAAABETHEtAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nrkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x135006450>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualize total sales for each product \n",
    "fig, ax = plt.subplots(figsize=(20,10))\n",
    "x = np.arange(20)\n",
    "rects1 = ax.bar(x,data3.Discount_Average_Sales[:20],width=(x[1]-x[0])*0.3)\n",
    "ax.set_ylabel('Average Sales Per Item')\n",
    "ax.set_xlabel('First 20 SKU')\n",
    "ax.set_xticks(x+0.3/2)\n",
    "rects2 = ax.bar(x+0.3,data3.Non_Discount_Average_Sales[:20],width=(x[1]-x[0])*0.3) #another bar\n",
    "ax.legend((rects1[0],rects2[0]),('With Discount','Without Discount')) #legend\n",
    "\n",
    "plt.tick_params(axis='x',bottom='off',labelbottom='off')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### It appears some SKU has higher average sales per item for product who are discounted than for the same product who are not discounted. So we may not to implement markdown stategies in those products."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<br><br>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Elasticity of Demand\n",
    "#### Percent change in demand / Percnet change in price"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "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>StoreID</th>\n",
       "      <th>TransactionID</th>\n",
       "      <th>Line</th>\n",
       "      <th>Sales</th>\n",
       "      <th>Count</th>\n",
       "      <th>Date</th>\n",
       "      <th>SKU</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00057</td>\n",
       "      <td>36</td>\n",
       "      <td>850.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A21773</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00057</td>\n",
       "      <td>39</td>\n",
       "      <td>1500.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A18009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00057</td>\n",
       "      <td>40</td>\n",
       "      <td>1500.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A25592</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00057</td>\n",
       "      <td>41</td>\n",
       "      <td>1600.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A1449</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00057</td>\n",
       "      <td>42</td>\n",
       "      <td>1600.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A1449</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  StoreID        TransactionID  Line    Sales  Count           Date     SKU\n",
       "0     A29  11-7270-00007-00057    36  850.000  1.000  3/11/12 00:00  A21773\n",
       "1     A29  11-7270-00007-00057    39 1500.000  1.000  3/11/12 00:00  A18009\n",
       "2     A29  11-7270-00007-00057    40 1500.000  1.000  3/11/12 00:00  A25592\n",
       "3     A29  11-7270-00007-00057    41 1600.000  1.000  3/11/12 00:00   A1449\n",
       "4     A29  11-7270-00007-00057    42 1600.000  1.000  3/11/12 00:00   A1449"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "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>StoreID</th>\n",
       "      <th>TransactionID</th>\n",
       "      <th>Line</th>\n",
       "      <th>Sales</th>\n",
       "      <th>Count</th>\n",
       "      <th>Date</th>\n",
       "      <th>SKU</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00057</td>\n",
       "      <td>64</td>\n",
       "      <td>2051.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A8481</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00057</td>\n",
       "      <td>65</td>\n",
       "      <td>-770.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A8481</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8227</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00006-00161</td>\n",
       "      <td>13</td>\n",
       "      <td>2051.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A8481</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8228</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00006-00161</td>\n",
       "      <td>14</td>\n",
       "      <td>-770.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A8481</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11696</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00006-00312</td>\n",
       "      <td>11</td>\n",
       "      <td>2051.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A8481</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11697</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00006-00312</td>\n",
       "      <td>12</td>\n",
       "      <td>-770.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A8481</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12444</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00008-00235</td>\n",
       "      <td>7</td>\n",
       "      <td>2051.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A8481</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12445</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00008-00235</td>\n",
       "      <td>8</td>\n",
       "      <td>-774.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A8481</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19516</th>\n",
       "      <td>A78</td>\n",
       "      <td>11-5634-00003-00090</td>\n",
       "      <td>10</td>\n",
       "      <td>2051.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A8481</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19517</th>\n",
       "      <td>A78</td>\n",
       "      <td>11-5634-00003-00090</td>\n",
       "      <td>11</td>\n",
       "      <td>-814.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A8481</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      StoreID        TransactionID  Line    Sales  Count           Date    SKU\n",
       "13        A29  11-7270-00007-00057    64 2051.000  1.000  3/11/12 00:00  A8481\n",
       "14        A29  11-7270-00007-00057    65 -770.000  0.000  3/11/12 00:00  A8481\n",
       "8227      A29  11-7270-00006-00161    13 2051.000  1.000  3/11/12 00:00  A8481\n",
       "8228      A29  11-7270-00006-00161    14 -770.000  0.000  3/11/12 00:00  A8481\n",
       "11696     A29  11-7270-00006-00312    11 2051.000  1.000  3/11/12 00:00  A8481\n",
       "11697     A29  11-7270-00006-00312    12 -770.000  0.000  3/11/12 00:00  A8481\n",
       "12444     A29  11-7270-00008-00235     7 2051.000  1.000  3/11/12 00:00  A8481\n",
       "12445     A29  11-7270-00008-00235     8 -774.000  0.000  3/11/12 00:00  A8481\n",
       "19516     A78  11-5634-00003-00090    10 2051.000  1.000  3/11/12 00:00  A8481\n",
       "19517     A78  11-5634-00003-00090    11 -814.000  0.000  3/11/12 00:00  A8481"
      ]
     },
     "execution_count": 84,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data[data['SKU']=='A8481'].head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Association Analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "from mlxtend.frequent_patterns import apriori\n",
    "from mlxtend.frequent_patterns import association_rules"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "data = pd.read_csv('RetailerPromotionStrategy_data.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 412,
   "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>InvoiceNo</th>\n",
       "      <th>StockCode</th>\n",
       "      <th>Description</th>\n",
       "      <th>Quantity</th>\n",
       "      <th>InvoiceDate</th>\n",
       "      <th>UnitPrice</th>\n",
       "      <th>CustomerID</th>\n",
       "      <th>Country</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>536365</td>\n",
       "      <td>85123A</td>\n",
       "      <td>WHITE HANGING HEART T-LIGHT HOLDER</td>\n",
       "      <td>6</td>\n",
       "      <td>2010-12-01 08:26:00</td>\n",
       "      <td>2.550</td>\n",
       "      <td>17850.000</td>\n",
       "      <td>United Kingdom</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>536365</td>\n",
       "      <td>71053</td>\n",
       "      <td>WHITE METAL LANTERN</td>\n",
       "      <td>6</td>\n",
       "      <td>2010-12-01 08:26:00</td>\n",
       "      <td>3.390</td>\n",
       "      <td>17850.000</td>\n",
       "      <td>United Kingdom</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>536365</td>\n",
       "      <td>84406B</td>\n",
       "      <td>CREAM CUPID HEARTS COAT HANGER</td>\n",
       "      <td>8</td>\n",
       "      <td>2010-12-01 08:26:00</td>\n",
       "      <td>2.750</td>\n",
       "      <td>17850.000</td>\n",
       "      <td>United Kingdom</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>536365</td>\n",
       "      <td>84029G</td>\n",
       "      <td>KNITTED UNION FLAG HOT WATER BOTTLE</td>\n",
       "      <td>6</td>\n",
       "      <td>2010-12-01 08:26:00</td>\n",
       "      <td>3.390</td>\n",
       "      <td>17850.000</td>\n",
       "      <td>United Kingdom</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>536365</td>\n",
       "      <td>84029E</td>\n",
       "      <td>RED WOOLLY HOTTIE WHITE HEART.</td>\n",
       "      <td>6</td>\n",
       "      <td>2010-12-01 08:26:00</td>\n",
       "      <td>3.390</td>\n",
       "      <td>17850.000</td>\n",
       "      <td>United Kingdom</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  InvoiceNo StockCode                          Description  Quantity  \\\n",
       "0    536365    85123A   WHITE HANGING HEART T-LIGHT HOLDER         6   \n",
       "1    536365     71053                  WHITE METAL LANTERN         6   \n",
       "2    536365    84406B       CREAM CUPID HEARTS COAT HANGER         8   \n",
       "3    536365    84029G  KNITTED UNION FLAG HOT WATER BOTTLE         6   \n",
       "4    536365    84029E       RED WOOLLY HOTTIE WHITE HEART.         6   \n",
       "\n",
       "          InvoiceDate  UnitPrice  CustomerID         Country  \n",
       "0 2010-12-01 08:26:00      2.550   17850.000  United Kingdom  \n",
       "1 2010-12-01 08:26:00      3.390   17850.000  United Kingdom  \n",
       "2 2010-12-01 08:26:00      2.750   17850.000  United Kingdom  \n",
       "3 2010-12-01 08:26:00      3.390   17850.000  United Kingdom  \n",
       "4 2010-12-01 08:26:00      3.390   17850.000  United Kingdom  "
      ]
     },
     "execution_count": 412,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_excel('http://archive.ics.uci.edu/ml/machine-learning-databases/00352/Online%20Retail.xlsx')\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "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>StoreID</th>\n",
       "      <th>TransactionID</th>\n",
       "      <th>Line</th>\n",
       "      <th>Sales</th>\n",
       "      <th>Count</th>\n",
       "      <th>Date</th>\n",
       "      <th>SKU</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00057</td>\n",
       "      <td>36</td>\n",
       "      <td>850.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A21773</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00057</td>\n",
       "      <td>39</td>\n",
       "      <td>1500.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A18009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00057</td>\n",
       "      <td>40</td>\n",
       "      <td>1500.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A25592</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00057</td>\n",
       "      <td>41</td>\n",
       "      <td>1600.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A1449</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>A29</td>\n",
       "      <td>11-7270-00007-00057</td>\n",
       "      <td>42</td>\n",
       "      <td>1600.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3/11/12 00:00</td>\n",
       "      <td>A1449</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  StoreID        TransactionID  Line   Sales  Count           Date     SKU\n",
       "0     A29  11-7270-00007-00057    36   850.0    1.0  3/11/12 00:00  A21773\n",
       "1     A29  11-7270-00007-00057    39  1500.0    1.0  3/11/12 00:00  A18009\n",
       "2     A29  11-7270-00007-00057    40  1500.0    1.0  3/11/12 00:00  A25592\n",
       "3     A29  11-7270-00007-00057    41  1600.0    1.0  3/11/12 00:00   A1449\n",
       "4     A29  11-7270-00007-00057    42  1600.0    1.0  3/11/12 00:00   A1449"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# Explore StoreID = A29\n",
    "basket = data[data['StoreID']=='A29'].groupby(['TransactionID','SKU'])['Count'].sum().unstack().reset_index().fillna(0).set_index('TransactionID')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "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>SKU</th>\n",
       "      <th>A1</th>\n",
       "      <th>A10004</th>\n",
       "      <th>A10006</th>\n",
       "      <th>A10007</th>\n",
       "      <th>A10008</th>\n",
       "      <th>A10016</th>\n",
       "      <th>A10023</th>\n",
       "      <th>A10024</th>\n",
       "      <th>A10025</th>\n",
       "      <th>A10045</th>\n",
       "      <th>...</th>\n",
       "      <th>A9933</th>\n",
       "      <th>A9938</th>\n",
       "      <th>A9941</th>\n",
       "      <th>A9942</th>\n",
       "      <th>A9946</th>\n",
       "      <th>A9947</th>\n",
       "      <th>A9963</th>\n",
       "      <th>A9964</th>\n",
       "      <th>A9981</th>\n",
       "      <th>A9982</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>TransactionID</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>11-7270-00005-00002</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11-7270-00005-00003</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11-7270-00005-00004</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11-7270-00005-00005</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11-7270-00005-00006</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 4325 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "SKU                   A1  A10004  A10006  A10007  A10008  A10016  A10023  \\\n",
       "TransactionID                                                              \n",
       "11-7270-00005-00002  0.0     0.0     0.0     0.0     0.0     0.0     0.0   \n",
       "11-7270-00005-00003  0.0     0.0     0.0     0.0     0.0     0.0     0.0   \n",
       "11-7270-00005-00004  0.0     0.0     0.0     0.0     0.0     0.0     0.0   \n",
       "11-7270-00005-00005  0.0     0.0     0.0     0.0     0.0     0.0     0.0   \n",
       "11-7270-00005-00006  0.0     0.0     0.0     0.0     0.0     0.0     0.0   \n",
       "\n",
       "SKU                  A10024  A10025  A10045  ...    A9933  A9938  A9941  \\\n",
       "TransactionID                                ...                          \n",
       "11-7270-00005-00002     0.0     0.0     0.0  ...      0.0    0.0    0.0   \n",
       "11-7270-00005-00003     0.0     0.0     0.0  ...      0.0    0.0    0.0   \n",
       "11-7270-00005-00004     0.0     0.0     0.0  ...      0.0    0.0    0.0   \n",
       "11-7270-00005-00005     0.0     0.0     0.0  ...      0.0    0.0    0.0   \n",
       "11-7270-00005-00006     0.0     0.0     0.0  ...      0.0    0.0    0.0   \n",
       "\n",
       "SKU                  A9942  A9946  A9947  A9963  A9964  A9981  A9982  \n",
       "TransactionID                                                         \n",
       "11-7270-00005-00002    0.0    0.0    0.0    0.0    0.0    0.0    0.0  \n",
       "11-7270-00005-00003    0.0    0.0    0.0    0.0    0.0    0.0    0.0  \n",
       "11-7270-00005-00004    0.0    0.0    0.0    0.0    0.0    0.0    0.0  \n",
       "11-7270-00005-00005    0.0    0.0    0.0    0.0    0.0    0.0    0.0  \n",
       "11-7270-00005-00006    0.0    0.0    0.0    0.0    0.0    0.0    0.0  \n",
       "\n",
       "[5 rows x 4325 columns]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "basket.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def encode_units(x):\n",
    "    if x <= 0:\n",
    "        return 0\n",
    "    if x >= 1:\n",
    "        return 1\n",
    "\n",
    "basket_sets = basket.applymap(encode_units)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "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>antecedants</th>\n",
       "      <th>consequents</th>\n",
       "      <th>support</th>\n",
       "      <th>confidence</th>\n",
       "      <th>lift</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>(A2854)</td>\n",
       "      <td>(A3664)</td>\n",
       "      <td>0.045802</td>\n",
       "      <td>0.254902</td>\n",
       "      <td>1.624225</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>(A3664)</td>\n",
       "      <td>(A2854)</td>\n",
       "      <td>0.156938</td>\n",
       "      <td>0.074392</td>\n",
       "      <td>1.624225</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>(A25312)</td>\n",
       "      <td>(A3664)</td>\n",
       "      <td>0.039291</td>\n",
       "      <td>0.325714</td>\n",
       "      <td>2.075438</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>(A3664)</td>\n",
       "      <td>(A25312)</td>\n",
       "      <td>0.156938</td>\n",
       "      <td>0.081545</td>\n",
       "      <td>2.075438</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  antecedants consequents   support  confidence      lift\n",
       "0     (A2854)     (A3664)  0.045802    0.254902  1.624225\n",
       "1     (A3664)     (A2854)  0.156938    0.074392  1.624225\n",
       "2    (A25312)     (A3664)  0.039291    0.325714  2.075438\n",
       "3     (A3664)    (A25312)  0.156938    0.081545  2.075438"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "frequent_itemsets = apriori(basket_sets, min_support=0.01, use_colnames=True)\n",
    "rules = association_rules(frequent_itemsets, metric=\"lift\", min_threshold=1)\n",
    "rules.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "280.0"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "basket['A25312'].sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "811.0"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "basket['A3664'].sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### In store A29, since the confidence and lift are not high in each rule, we conclude there is no strong relationship between two items."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<br><br>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "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>antecedants</th>\n",
       "      <th>consequents</th>\n",
       "      <th>support</th>\n",
       "      <th>confidence</th>\n",
       "      <th>lift</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>(A5483)</td>\n",
       "      <td>(A13675)</td>\n",
       "      <td>0.010922</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>51.500000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>(A13675)</td>\n",
       "      <td>(A5483)</td>\n",
       "      <td>0.019417</td>\n",
       "      <td>0.562500</td>\n",
       "      <td>51.500000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>(A25312)</td>\n",
       "      <td>(A8397)</td>\n",
       "      <td>0.064320</td>\n",
       "      <td>0.188679</td>\n",
       "      <td>6.477987</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>(A8397)</td>\n",
       "      <td>(A25312)</td>\n",
       "      <td>0.029126</td>\n",
       "      <td>0.416667</td>\n",
       "      <td>6.477987</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>(A2854)</td>\n",
       "      <td>(A3664)</td>\n",
       "      <td>0.058252</td>\n",
       "      <td>0.229167</td>\n",
       "      <td>1.072917</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  antecedants consequents   support  confidence       lift\n",
       "0     (A5483)    (A13675)  0.010922    1.000000  51.500000\n",
       "1    (A13675)     (A5483)  0.019417    0.562500  51.500000\n",
       "2    (A25312)     (A8397)  0.064320    0.188679   6.477987\n",
       "3     (A8397)    (A25312)  0.029126    0.416667   6.477987\n",
       "4     (A2854)     (A3664)  0.058252    0.229167   1.072917"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Explore StoreID = A79\n",
    "basket = data[data['StoreID']=='A79'].groupby(['TransactionID','SKU'])['Count'].sum().unstack().reset_index().fillna(0).set_index('TransactionID')\n",
    "def encode_units(x):\n",
    "    if x <= 0:\n",
    "        return 0\n",
    "    if x >= 1:\n",
    "        return 1\n",
    "\n",
    "basket_sets = basket.applymap(encode_units)\n",
    "\n",
    "frequent_itemsets = apriori(basket_sets, min_support=0.01, use_colnames=True)\n",
    "rules = association_rules(frequent_itemsets, metric=\"lift\", min_threshold=1)\n",
    "rules.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "14.0\n",
      "25.0\n"
     ]
    }
   ],
   "source": [
    "print basket['A5483'].sum()\n",
    "\n",
    "print basket['A13675'].sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### We can see in Store A79, there are a quite few rules with a high lift value which means that it occurs more frequently than would be expected given the number of transactions and product combinatons. \n",
    "### We can see several where the confidence is high as well.\n",
    "### In looking at above rules, it seems that product A5483 and A13675 are purchased together, product A13675 and A5483 are purchased together, product A8397 and A25312 are purchased together.\n",
    "### We can drive up sales through recommendation from their antecendants.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python [conda root]",
   "language": "python",
   "name": "conda-root-py"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}