Customer Lifetime Value.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "# Calculating Customer Lifetime Value  \n",
    "### Zach Brown"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "%matplotlib inline\n",
    "sns.set_style('darkgrid')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The purpose of this exercise is to use a list of historical customer transactions to estimate future lifetime value for customers.  The input data consists of 3 fields: a hashed ID for the customer that made the transaction, a timestamp in UTC of when the transaction occurred, and a monetary purchase value of the transaction.  The dates for the transactions range from November 30, 2016 to December 6, 2017.\n",
    "\n",
    "Before we can do anything, the data needs to be imported."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "df = pd.read_csv('data.csv',na_values=['None'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Id</th>\n",
       "      <th>Timestamp</th>\n",
       "      <th>PurchaseValue</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>4ec6fab0e2fca65b2950fbc9fffa45f8</td>\n",
       "      <td>2017-12-06T17:53:59Z</td>\n",
       "      <td>261.54</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>8170354d9b20f5edc7747886e7fbda90</td>\n",
       "      <td>2017-12-06T17:48:19Z</td>\n",
       "      <td>206.07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>5780f7258e491e462611d65f1dbd1956</td>\n",
       "      <td>2017-12-06T17:46:49Z</td>\n",
       "      <td>78.84</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>e5966041d824f1bdc3ac190d49a8bbec</td>\n",
       "      <td>2017-12-06T17:46:41Z</td>\n",
       "      <td>50.42</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>52aa2478574fe7a4a78162074eb6c987</td>\n",
       "      <td>2017-12-06T17:34:56Z</td>\n",
       "      <td>274.62</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                 Id             Timestamp  PurchaseValue\n",
       "0  4ec6fab0e2fca65b2950fbc9fffa45f8  2017-12-06T17:53:59Z         261.54\n",
       "1  8170354d9b20f5edc7747886e7fbda90  2017-12-06T17:48:19Z         206.07\n",
       "2  5780f7258e491e462611d65f1dbd1956  2017-12-06T17:46:49Z          78.84\n",
       "3  e5966041d824f1bdc3ac190d49a8bbec  2017-12-06T17:46:41Z          50.42\n",
       "4  52aa2478574fe7a4a78162074eb6c987  2017-12-06T17:34:56Z         274.62"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 50001 entries, 0 to 50000\n",
      "Data columns (total 3 columns):\n",
      "Id               50001 non-null object\n",
      "Timestamp        50001 non-null object\n",
      "PurchaseValue    49945 non-null float64\n",
      "dtypes: float64(1), object(2)\n",
      "memory usage: 1.1+ MB\n"
     ]
    }
   ],
   "source": [
    "df.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Two things jump out after importing the data and taking a quick look.  We will need to convert the Timestamp column to a datetime format.  Also, there are 56 null PurchaseValue entries that will need to be dealt with."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "df['Timestamp'] = pd.to_datetime(df['Timestamp'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Id</th>\n",
       "      <th>Timestamp</th>\n",
       "      <th>PurchaseValue</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>4ec6fab0e2fca65b2950fbc9fffa45f8</td>\n",
       "      <td>2017-12-06 17:53:59</td>\n",
       "      <td>261.54</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>8170354d9b20f5edc7747886e7fbda90</td>\n",
       "      <td>2017-12-06 17:48:19</td>\n",
       "      <td>206.07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>5780f7258e491e462611d65f1dbd1956</td>\n",
       "      <td>2017-12-06 17:46:49</td>\n",
       "      <td>78.84</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>e5966041d824f1bdc3ac190d49a8bbec</td>\n",
       "      <td>2017-12-06 17:46:41</td>\n",
       "      <td>50.42</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>52aa2478574fe7a4a78162074eb6c987</td>\n",
       "      <td>2017-12-06 17:34:56</td>\n",
       "      <td>274.62</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                 Id           Timestamp  PurchaseValue\n",
       "0  4ec6fab0e2fca65b2950fbc9fffa45f8 2017-12-06 17:53:59         261.54\n",
       "1  8170354d9b20f5edc7747886e7fbda90 2017-12-06 17:48:19         206.07\n",
       "2  5780f7258e491e462611d65f1dbd1956 2017-12-06 17:46:49          78.84\n",
       "3  e5966041d824f1bdc3ac190d49a8bbec 2017-12-06 17:46:41          50.42\n",
       "4  52aa2478574fe7a4a78162074eb6c987 2017-12-06 17:34:56         274.62"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 50001 entries, 0 to 50000\n",
      "Data columns (total 3 columns):\n",
      "Id               50001 non-null object\n",
      "Timestamp        50001 non-null datetime64[ns]\n",
      "PurchaseValue    49945 non-null float64\n",
      "dtypes: datetime64[ns](1), float64(1), object(1)\n",
      "memory usage: 1.1+ MB\n"
     ]
    }
   ],
   "source": [
    "df.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Ok, we've changed the data type of the TimeStamp column.  \n",
    "\n",
    "Let's take a look at the PurchaseValue column now."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x2a03401a6d8>"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAmEAAAFuCAYAAADJS88DAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x2a03401a0b8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(10,6))\n",
    "sns.distplot(df['PurchaseValue'].dropna())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "It looks like there are a few extremely large values.  Let's see if we can find out any more information about these outliers."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Id</th>\n",
       "      <th>Timestamp</th>\n",
       "      <th>PurchaseValue</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>46769</th>\n",
       "      <td>dca76db00cc59dfbcdcc97c8bbc7f9f1</td>\n",
       "      <td>2016-12-21 19:27:53</td>\n",
       "      <td>11599.68</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36562</th>\n",
       "      <td>5ac5ed64cd99ed2a8403b7a927e644ef</td>\n",
       "      <td>2017-03-31 20:57:25</td>\n",
       "      <td>5626.45</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6601</th>\n",
       "      <td>98f8e41f45721cbe49a3147f6cf62432</td>\n",
       "      <td>2017-10-31 17:12:30</td>\n",
       "      <td>4680.56</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14944</th>\n",
       "      <td>32df931a012f04da70265457d43e7da2</td>\n",
       "      <td>2017-08-26 17:49:58</td>\n",
       "      <td>4508.87</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8336</th>\n",
       "      <td>795a8a5e5beff24e1557a6cb49a82692</td>\n",
       "      <td>2017-10-16 22:15:53</td>\n",
       "      <td>4455.39</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5762</th>\n",
       "      <td>15f1f773f94ab0b1c63067f2b3d9f7be</td>\n",
       "      <td>2017-11-08 04:47:09</td>\n",
       "      <td>4181.89</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23413</th>\n",
       "      <td>de9957594c13e5c1846f68b01e832c07</td>\n",
       "      <td>2017-06-28 17:01:27</td>\n",
       "      <td>3545.43</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19155</th>\n",
       "      <td>c8fcdff9e08efe0b48f1e1f278c07e73</td>\n",
       "      <td>2017-07-26 03:03:46</td>\n",
       "      <td>3420.70</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17468</th>\n",
       "      <td>fe403ffcf47b4efdf39874d181ae6da4</td>\n",
       "      <td>2017-08-08 02:12:15</td>\n",
       "      <td>3344.83</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17307</th>\n",
       "      <td>d3a9042cdfe63cf068e9e6d92d42ba5c</td>\n",
       "      <td>2017-08-09 19:10:29</td>\n",
       "      <td>3307.93</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3368</th>\n",
       "      <td>0b937997e6f55b43f95a14c1c6fe94e2</td>\n",
       "      <td>2017-11-25 19:19:32</td>\n",
       "      <td>3208.34</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33011</th>\n",
       "      <td>a89a80ff2ae120c3b20d7a9154a318aa</td>\n",
       "      <td>2017-04-26 20:16:45</td>\n",
       "      <td>3125.16</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>840</th>\n",
       "      <td>7cfba6ffdb84c9fad97c0ff1263ea2af</td>\n",
       "      <td>2017-11-29 06:12:40</td>\n",
       "      <td>3084.98</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33113</th>\n",
       "      <td>5d9d6531ffb7756952118b67d21c2e18</td>\n",
       "      <td>2017-04-26 01:45:36</td>\n",
       "      <td>3034.97</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23339</th>\n",
       "      <td>4ccfbbe336ffe1bb928a31ea38fc13a1</td>\n",
       "      <td>2017-06-29 01:49:08</td>\n",
       "      <td>3000.63</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13782</th>\n",
       "      <td>127abd0244d04bed9ae082bd5dc60cb1</td>\n",
       "      <td>2017-09-07 21:46:50</td>\n",
       "      <td>2988.10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41524</th>\n",
       "      <td>58305961906f012793e4ab5320435868</td>\n",
       "      <td>2017-02-06 17:16:25</td>\n",
       "      <td>2967.37</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>47549</th>\n",
       "      <td>b538b73f8d43b82e55a31d99987a91fc</td>\n",
       "      <td>2016-12-16 23:51:25</td>\n",
       "      <td>2963.70</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4350</th>\n",
       "      <td>24f05bfab01fef56ec049a828ebe20ab</td>\n",
       "      <td>2017-11-19 20:24:56</td>\n",
       "      <td>2941.99</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7498</th>\n",
       "      <td>a8c6dd0d4f1b33878dfc2f3559e88f87</td>\n",
       "      <td>2017-10-23 10:52:52</td>\n",
       "      <td>2941.72</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25276</th>\n",
       "      <td>60b884e115b971ee52f550a8b64750f4</td>\n",
       "      <td>2017-06-17 23:02:46</td>\n",
       "      <td>2907.80</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37747</th>\n",
       "      <td>e49e3b7b0c37ea3ccf107deeff2ead1d</td>\n",
       "      <td>2017-03-18 17:51:05</td>\n",
       "      <td>2810.67</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20740</th>\n",
       "      <td>ba5cf27a059fb5e6357dc3e7f70dbab2</td>\n",
       "      <td>2017-07-15 22:02:55</td>\n",
       "      <td>2809.15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8147</th>\n",
       "      <td>68f9b81d631eb5ae00be3fcbf770bc51</td>\n",
       "      <td>2017-10-17 17:33:20</td>\n",
       "      <td>2799.18</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17332</th>\n",
       "      <td>cd4cb9ec252a085ed4d2d3af7c18280a</td>\n",
       "      <td>2017-08-09 17:26:27</td>\n",
       "      <td>2796.94</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                     Id           Timestamp  PurchaseValue\n",
       "46769  dca76db00cc59dfbcdcc97c8bbc7f9f1 2016-12-21 19:27:53       11599.68\n",
       "36562  5ac5ed64cd99ed2a8403b7a927e644ef 2017-03-31 20:57:25        5626.45\n",
       "6601   98f8e41f45721cbe49a3147f6cf62432 2017-10-31 17:12:30        4680.56\n",
       "14944  32df931a012f04da70265457d43e7da2 2017-08-26 17:49:58        4508.87\n",
       "8336   795a8a5e5beff24e1557a6cb49a82692 2017-10-16 22:15:53        4455.39\n",
       "5762   15f1f773f94ab0b1c63067f2b3d9f7be 2017-11-08 04:47:09        4181.89\n",
       "23413  de9957594c13e5c1846f68b01e832c07 2017-06-28 17:01:27        3545.43\n",
       "19155  c8fcdff9e08efe0b48f1e1f278c07e73 2017-07-26 03:03:46        3420.70\n",
       "17468  fe403ffcf47b4efdf39874d181ae6da4 2017-08-08 02:12:15        3344.83\n",
       "17307  d3a9042cdfe63cf068e9e6d92d42ba5c 2017-08-09 19:10:29        3307.93\n",
       "3368   0b937997e6f55b43f95a14c1c6fe94e2 2017-11-25 19:19:32        3208.34\n",
       "33011  a89a80ff2ae120c3b20d7a9154a318aa 2017-04-26 20:16:45        3125.16\n",
       "840    7cfba6ffdb84c9fad97c0ff1263ea2af 2017-11-29 06:12:40        3084.98\n",
       "33113  5d9d6531ffb7756952118b67d21c2e18 2017-04-26 01:45:36        3034.97\n",
       "23339  4ccfbbe336ffe1bb928a31ea38fc13a1 2017-06-29 01:49:08        3000.63\n",
       "13782  127abd0244d04bed9ae082bd5dc60cb1 2017-09-07 21:46:50        2988.10\n",
       "41524  58305961906f012793e4ab5320435868 2017-02-06 17:16:25        2967.37\n",
       "47549  b538b73f8d43b82e55a31d99987a91fc 2016-12-16 23:51:25        2963.70\n",
       "4350   24f05bfab01fef56ec049a828ebe20ab 2017-11-19 20:24:56        2941.99\n",
       "7498   a8c6dd0d4f1b33878dfc2f3559e88f87 2017-10-23 10:52:52        2941.72\n",
       "25276  60b884e115b971ee52f550a8b64750f4 2017-06-17 23:02:46        2907.80\n",
       "37747  e49e3b7b0c37ea3ccf107deeff2ead1d 2017-03-18 17:51:05        2810.67\n",
       "20740  ba5cf27a059fb5e6357dc3e7f70dbab2 2017-07-15 22:02:55        2809.15\n",
       "8147   68f9b81d631eb5ae00be3fcbf770bc51 2017-10-17 17:33:20        2799.18\n",
       "17332  cd4cb9ec252a085ed4d2d3af7c18280a 2017-08-09 17:26:27        2796.94"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.sort_values(by='PurchaseValue',ascending=False).head(25)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The highest purchase value is more than double that of the next highest purchase value.  Unfortunately, without any further information about these transactions there is no way to determine whether this is a legitimate purchase or not.  If the data set included more details, such as a description of the item being purchased, it might be possible to determine if any of these large purchase values are erroneous.  Since we do not have any further details though, we will need to leave these outliers in the data set."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now, let's deal with the null purchase values.  There are multiple ways to impute these missing values.  For example, maybe different sized purchases are typically made on different days of the week or times of the day.  We could take that into account when filling the null values.  For now though, let's just fill the null values with the mean PurchaseValue for that given customer."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "df['PurchaseValue'] = df.groupby('Id').transform(lambda x: x.fillna(x.mean()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 50001 entries, 0 to 50000\n",
      "Data columns (total 3 columns):\n",
      "Id               50001 non-null object\n",
      "Timestamp        50001 non-null datetime64[ns]\n",
      "PurchaseValue    49966 non-null float64\n",
      "dtypes: datetime64[ns](1), float64(1), object(1)\n",
      "memory usage: 1.1+ MB\n"
     ]
    }
   ],
   "source": [
    "df.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There are still a few null purchase values.  This means that the values for all purchases made by those customers are null.  For these customers, we will simply use the mean purchase value for all customers."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "df['PurchaseValue'].fillna(df['PurchaseValue'].mean(),inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 50001 entries, 0 to 50000\n",
      "Data columns (total 3 columns):\n",
      "Id               50001 non-null object\n",
      "Timestamp        50001 non-null datetime64[ns]\n",
      "PurchaseValue    50001 non-null float64\n",
      "dtypes: datetime64[ns](1), float64(1), object(1)\n",
      "memory usage: 1.1+ MB\n"
     ]
    }
   ],
   "source": [
    "df.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Earliest Timestamp: 2016-11-30 04:47:00\n",
      "Latest Timestamp: 2017-12-06 17:53:59\n"
     ]
    }
   ],
   "source": [
    "print('Earliest Timestamp:',df['Timestamp'].min())\n",
    "print('Latest Timestamp:',df['Timestamp'].max())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0,0.5,'Count')"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAtIAAAGGCAYAAABWnWbWAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x2a039c4de10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df_time = df.set_index('Timestamp').resample('1D').count()\n",
    "df_time.drop(columns='PurchaseValue',axis=1,inplace=True)\n",
    "fig = df_time.plot(figsize=(12,6),title='Purchase Count by Day')\n",
    "fig.set_ylabel('Count')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "When looking at the distribution of purchases by day, there is a large spike at the end of 2017.  Let's take a look at exactly which days account for the large spike in sales."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Timestamp\n",
       "2017-11-29    1092\n",
       "2017-11-28     887\n",
       "2017-11-25     448\n",
       "2017-09-14     379\n",
       "2017-07-25     338\n",
       "2017-09-28     336\n",
       "2017-11-26     318\n",
       "2017-07-26     274\n",
       "2017-08-10     272\n",
       "2017-08-15     264\n",
       "Name: Timestamp, dtype: int64"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['Timestamp'].groupby(df['Timestamp'].map(lambda x: x.date())).count().sort_values(ascending=False).head(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can see that the 3 highest sales days are at the end of November 2017.  Initially this seemed like there may have been an issue with the data.  However, this is right around the time of Black Friday and Cyber Monday where many retailers see their highest sales of the year.  The earliest data that we have is from right after this time period in 2016, so unfortunately we cannot compare with the previous year to see if there was a similar spike."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The raw data is in a transactional form, meaning that there is a separate row for every row and customers can be associated with multiple rows.  In order to use [Peter Fader and Bruce Hardie's BG/NBD Model](http://mktg.uni-svishtov.bg/ivm/resources/Counting_Your_Customers.pdf), this data will need to be reshaped into a Frequency/Recency Matrix.  This means that there will be one row per customer, with columns for frequency (the number of repeat purchases or total purchases minus 1), the age of the customer (time periods since the customer's first purchase), and recency (the age of the customer when they made their most recent purchase).  This is a very common issue, so luckily the [lifetimes library](https://github.com/CamDavidsonPilon/lifetimes) has a utility function that we can use to reshape our data into this form.  \n",
    "\n",
    "In order to validate our model, we will split the transactions into a train and test set.  The lifetimes library also includes functionality to do this."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from lifetimes.utils import summary_data_from_transaction_data, calibration_and_holdout_data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First, we will create the overall frequency/recency matrix.  This will be used later on to determine the 100 customers with the highest projected purchases and highest projected amount spent."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Index: 23743 entries, 0001117ff1305c1fe840697166e61564 to fffc0c75aa4e82c9856616b4c84e3f13\n",
      "Data columns (total 3 columns):\n",
      "frequency    23743 non-null float64\n",
      "recency      23743 non-null float64\n",
      "T            23743 non-null float64\n",
      "dtypes: float64(3)\n",
      "memory usage: 742.0+ KB\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>frequency</th>\n",
       "      <th>recency</th>\n",
       "      <th>T</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0001117ff1305c1fe840697166e61564</th>\n",
       "      <td>1.0</td>\n",
       "      <td>11.0</td>\n",
       "      <td>344.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>00028502859fd7e111d88c20456b59d5</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>342.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>000306d200fd6e9b03de48d15348f5c2</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>33.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0003f3458a6e7b495a975c2d9ddda559</th>\n",
       "      <td>2.0</td>\n",
       "      <td>285.0</td>\n",
       "      <td>343.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>000784b838b807ad589d4bc69c0c562f</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>210.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                  frequency  recency      T\n",
       "Id                                                         \n",
       "0001117ff1305c1fe840697166e61564        1.0     11.0  344.0\n",
       "00028502859fd7e111d88c20456b59d5        0.0      0.0  342.0\n",
       "000306d200fd6e9b03de48d15348f5c2        0.0      0.0   33.0\n",
       "0003f3458a6e7b495a975c2d9ddda559        2.0    285.0  343.0\n",
       "000784b838b807ad589d4bc69c0c562f        0.0      0.0  210.0"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "summary = summary_data_from_transaction_data(transactions=df, customer_id_col='Id', datetime_col='Timestamp')\n",
    "summary.info()\n",
    "summary.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now, we need to create another frequency/recency matrix using a calibration and holdout set.  The calibration set will be used to train the model and the holdout model will be used to validate the model."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Index: 18750 entries, 0001117ff1305c1fe840697166e61564 to fffc0c75aa4e82c9856616b4c84e3f13\n",
      "Data columns (total 5 columns):\n",
      "frequency_cal        18750 non-null float64\n",
      "recency_cal          18750 non-null float64\n",
      "T_cal                18750 non-null float64\n",
      "frequency_holdout    18750 non-null float64\n",
      "duration_holdout     18750 non-null int64\n",
      "dtypes: float64(4), int64(1)\n",
      "memory usage: 1.5+ MB\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>frequency_cal</th>\n",
       "      <th>recency_cal</th>\n",
       "      <th>T_cal</th>\n",
       "      <th>frequency_holdout</th>\n",
       "      <th>duration_holdout</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Id</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>0001117ff1305c1fe840697166e61564</th>\n",
       "      <td>1.0</td>\n",
       "      <td>11.0</td>\n",
       "      <td>248.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>97</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>00028502859fd7e111d88c20456b59d5</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>246.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>97</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0003f3458a6e7b495a975c2d9ddda559</th>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>247.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>97</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>000784b838b807ad589d4bc69c0c562f</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>114.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>97</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>000ad0f90e9fcb6ff5a0bc480cccbdb3</th>\n",
       "      <td>5.0</td>\n",
       "      <td>68.0</td>\n",
       "      <td>104.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>97</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                  frequency_cal  recency_cal  T_cal  \\\n",
       "Id                                                                    \n",
       "0001117ff1305c1fe840697166e61564            1.0         11.0  248.0   \n",
       "00028502859fd7e111d88c20456b59d5            0.0          0.0  246.0   \n",
       "0003f3458a6e7b495a975c2d9ddda559            1.0          2.0  247.0   \n",
       "000784b838b807ad589d4bc69c0c562f            0.0          0.0  114.0   \n",
       "000ad0f90e9fcb6ff5a0bc480cccbdb3            5.0         68.0  104.0   \n",
       "\n",
       "                                  frequency_holdout  duration_holdout  \n",
       "Id                                                                     \n",
       "0001117ff1305c1fe840697166e61564                0.0                97  \n",
       "00028502859fd7e111d88c20456b59d5                0.0                97  \n",
       "0003f3458a6e7b495a975c2d9ddda559                1.0                97  \n",
       "000784b838b807ad589d4bc69c0c562f                0.0                97  \n",
       "000ad0f90e9fcb6ff5a0bc480cccbdb3                0.0                97  "
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "summary_cal_holdout = calibration_and_holdout_data(df, 'Id', 'Timestamp',\n",
    "                                                  calibration_period_end='2017-09-01',\n",
    "                                                  observation_period_end='2017-12-07')\n",
    "summary_cal_holdout.info()\n",
    "summary_cal_holdout.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We now have a frequency/recency matrix that is split into a calibration set and a holdout set on September 1, 2017.  We will train the model on all transactions before that date and test it on the transactions after that date.  One point of note is that this matrix contains fewer customers than the overall matrix.  This is because any customers who had not purchased anything before September 1st were not included."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from lifetimes import BetaGeoFitter"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<lifetimes.BetaGeoFitter: fitted with 18750 subjects, a: 0.94, alpha: 48.72, b: 3.50, r: 0.37>"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bgf = BetaGeoFitter()\n",
    "bgf.fit(summary_cal_holdout['frequency_cal'],\n",
    "        summary_cal_holdout['recency_cal'],\n",
    "        summary_cal_holdout['T_cal'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Not every customer will have the same purchasing pattern.  For example, let's say that we have two customers that we'll call Customer A and Customer B.  customer A used to make purchases every day, but hasn't bought anything in several months.  Customer B doesn't make purchases as frequently, but generally buys something every month and their last purchase was last month.  In this example, Customer B is more likely to be \"alive\" (and therefore more valuable) than Customer A, even though Customer A used to make more purchases than Customer B.  This is an illustration of how frequency and recency are both important in determining customer lifetime value.  Let's visualize the relationship between frequency and recency."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from lifetimes.plotting import plot_frequency_recency_matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAaAAAAGoCAYAAAAAZTE0AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x2a034563748>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(6,6))\n",
    "plot_frequency_recency_matrix(bgf,cmap='viridis')\n",
    "plt.tight_layout()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Looking at this plot, we can see the best customers in the lower righthand corner.  This is where frequency and recency are both the highest.  \n",
    "\n",
    "Somebody like Customer A with a high frequency and low recency would be in the top right corner, which has a very low expected future purchase value.  Somebody like Customer B with a high recency and middling frequency would be at the bottom of the plot somewhere in the middle of the X-axis where the expected future purchases is non-zero.  \n",
    "\n",
    "Another interesting thing to notice is that the future value of customers starts to fan out a bit on the Y-axis (Recency) as the X-axis (Frequency) decreases.  This illustrates the uncertainty that comes with new customers.  \n",
    "\n",
    "Let's also take a look at how frequency and recency impact the probability that a customer is still active."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from lifetimes.plotting import plot_probability_alive_matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZUAAAGoCAYAAACOv3hIAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nA7lwrxRoEJTrAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x2a034184128>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(6,6))\n",
    "plot_probability_alive_matrix(bgf,cmap='viridis')\n",
    "plt.tight_layout()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As one might expect, a similar pattern is seen here.  As recency and frequency increase, so does the likelihood that the customer is still active.  And again, there is more uncertainty around customers with lower frequencies.  \n",
    "\n",
    "Next, let's compare how well our model predicts purchases for the holdout set compared to the actual values."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from lifetimes.plotting import plot_calibration_purchases_vs_holdout_purchases"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x2a03445f5c0>"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAesAAAF8CAYAAADvkJNaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x2a034aef7b8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_calibration_purchases_vs_holdout_purchases(bgf,summary_cal_holdout,figsize=(8,6))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The model appears to generalize fairly well when run against the holdout set.  \n",
    "\n",
    "Now, we can use our model to predict the number of purchases made by each customer in the next 12 months."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>frequency</th>\n",
       "      <th>recency</th>\n",
       "      <th>T</th>\n",
       "      <th>predicted_purchases</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>a5fadc51b1ae844ad9a70ad3cfa46a64</th>\n",
       "      <td>95.0</td>\n",
       "      <td>371.0</td>\n",
       "      <td>371.0</td>\n",
       "      <td>60.718377</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>48a503edbaded96a3be27deee11967a1</th>\n",
       "      <td>46.0</td>\n",
       "      <td>363.0</td>\n",
       "      <td>371.0</td>\n",
       "      <td>28.675147</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9f447f9415a380ac2eeee7df49c6ee7e</th>\n",
       "      <td>23.0</td>\n",
       "      <td>163.0</td>\n",
       "      <td>166.0</td>\n",
       "      <td>23.715042</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5f01420f0edda6555df5ce1cc62b986c</th>\n",
       "      <td>39.0</td>\n",
       "      <td>350.0</td>\n",
       "      <td>370.0</td>\n",
       "      <td>22.163748</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8d2ce54737dd404d20cadf1405d46dc8</th>\n",
       "      <td>33.0</td>\n",
       "      <td>360.0</td>\n",
       "      <td>362.0</td>\n",
       "      <td>21.424866</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>a62a17bb46864da2c6da691d838971b3</th>\n",
       "      <td>30.0</td>\n",
       "      <td>366.0</td>\n",
       "      <td>366.0</td>\n",
       "      <td>19.431837</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3b11478939967e896ae2619615650f97</th>\n",
       "      <td>20.0</td>\n",
       "      <td>190.0</td>\n",
       "      <td>194.0</td>\n",
       "      <td>18.971988</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2ad9a83ee23110d8c2f4c01600b94f20</th>\n",
       "      <td>29.0</td>\n",
       "      <td>357.0</td>\n",
       "      <td>364.0</td>\n",
       "      <td>18.508699</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75fda9ea22086bf3814ff8c3f53de8ca</th>\n",
       "      <td>28.0</td>\n",
       "      <td>348.0</td>\n",
       "      <td>355.0</td>\n",
       "      <td>18.174781</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30aa99d3357244cf38ca04eadef1473a</th>\n",
       "      <td>18.0</td>\n",
       "      <td>183.0</td>\n",
       "      <td>189.0</td>\n",
       "      <td>17.190421</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>a719d6643a7832535de9aded2f467825</th>\n",
       "      <td>26.0</td>\n",
       "      <td>360.0</td>\n",
       "      <td>368.0</td>\n",
       "      <td>16.444873</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>a1c8d419a97af1f7152e21c0dddfcbce</th>\n",
       "      <td>24.0</td>\n",
       "      <td>369.0</td>\n",
       "      <td>370.0</td>\n",
       "      <td>15.423088</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>cdf642859ae8d9a1e489c3a655cba827</th>\n",
       "      <td>23.0</td>\n",
       "      <td>343.0</td>\n",
       "      <td>350.0</td>\n",
       "      <td>15.110830</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ed2b4332b3ca253cfbb0ffb54d3f5ae0</th>\n",
       "      <td>23.0</td>\n",
       "      <td>337.0</td>\n",
       "      <td>348.0</td>\n",
       "      <td>14.952128</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>880474a82e33389ae7f377b2ea35bbdd</th>\n",
       "      <td>23.0</td>\n",
       "      <td>364.0</td>\n",
       "      <td>371.0</td>\n",
       "      <td>14.539901</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5ac5ed64cd99ed2a8403b7a927e644ef</th>\n",
       "      <td>23.0</td>\n",
       "      <td>222.0</td>\n",
       "      <td>250.0</td>\n",
       "      <td>14.475066</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16134915d822fe17588ae585935e1e81</th>\n",
       "      <td>22.0</td>\n",
       "      <td>356.0</td>\n",
       "      <td>362.0</td>\n",
       "      <td>14.186154</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>f6a4f156c817ecb376d678f2aafdc570</th>\n",
       "      <td>20.0</td>\n",
       "      <td>309.0</td>\n",
       "      <td>316.0</td>\n",
       "      <td>14.068079</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ca2202a96c2de6ca6b8a37a4a73fa730</th>\n",
       "      <td>7.0</td>\n",
       "      <td>12.0</td>\n",
       "      <td>20.0</td>\n",
       "      <td>13.592224</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>742d5a52d4df7cb14246d7f390de5d8a</th>\n",
       "      <td>21.0</td>\n",
       "      <td>340.0</td>\n",
       "      <td>356.0</td>\n",
       "      <td>13.205288</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>663b3df249f614305792ec7bf1dead30</th>\n",
       "      <td>20.0</td>\n",
       "      <td>347.0</td>\n",
       "      <td>352.0</td>\n",
       "      <td>13.191422</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>bb5d927e9f1aefa4db1881579e055a3e</th>\n",
       "      <td>20.0</td>\n",
       "      <td>348.0</td>\n",
       "      <td>355.0</td>\n",
       "      <td>13.045209</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>e675875db3648ab4dbbd52768296425c</th>\n",
       "      <td>12.0</td>\n",
       "      <td>145.0</td>\n",
       "      <td>152.0</td>\n",
       "      <td>12.929123</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>61d2370d7ccfe63adf6ab388d7415dc2</th>\n",
       "      <td>19.0</td>\n",
       "      <td>340.0</td>\n",
       "      <td>342.0</td>\n",
       "      <td>12.877917</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2b75f007b50d6b21e1501f47dc8d632b</th>\n",
       "      <td>19.0</td>\n",
       "      <td>355.0</td>\n",
       "      <td>355.0</td>\n",
       "      <td>12.612052</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>cd1751286a561d51655ee47246b3019d</th>\n",
       "      <td>19.0</td>\n",
       "      <td>351.0</td>\n",
       "      <td>355.0</td>\n",
       "      <td>12.500814</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9cb19c3fc7311aae01cf16571b528001</th>\n",
       "      <td>17.0</td>\n",
       "      <td>263.0</td>\n",
       "      <td>279.0</td>\n",
       "      <td>12.447227</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4f672b7c5d7a4214b08ec6906163c980</th>\n",
       "      <td>14.0</td>\n",
       "      <td>208.0</td>\n",
       "      <td>216.0</td>\n",
       "      <td>12.423322</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7108b2bb5964a79ee1e6c8a9acde8107</th>\n",
       "      <td>16.0</td>\n",
       "      <td>272.0</td>\n",
       "      <td>279.0</td>\n",
       "      <td>12.222910</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ca5cc4f526fcff10e85ce2685999e2c1</th>\n",
       "      <td>9.0</td>\n",
       "      <td>84.0</td>\n",
       "      <td>92.0</td>\n",
       "      <td>12.145760</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14f57af72e3d2a16b7b4c088cbe3196f</th>\n",
       "      <td>14.0</td>\n",
       "      <td>334.0</td>\n",
       "      <td>341.0</td>\n",
       "      <td>9.436900</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13cb5c06f8d3c1c18e14539c16088e21</th>\n",
       "      <td>15.0</td>\n",
       "      <td>356.0</td>\n",
       "      <td>370.0</td>\n",
       "      <td>9.377351</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>922bdaa119d44065b17b0447255646ae</th>\n",
       "      <td>8.0</td>\n",
       "      <td>144.0</td>\n",
       "      <td>144.0</td>\n",
       "      <td>9.324966</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1200a4df001756f0fc97e0142c069442</th>\n",
       "      <td>11.0</td>\n",
       "      <td>231.0</td>\n",
       "      <td>239.0</td>\n",
       "      <td>9.294833</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>47810b7c7fc24d3cf8287d03de62448b</th>\n",
       "      <td>14.0</td>\n",
       "      <td>342.0</td>\n",
       "      <td>349.0</td>\n",
       "      <td>9.292691</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>d0ad6b624dfc37784a755144be8c76a6</th>\n",
       "      <td>14.0</td>\n",
       "      <td>333.0</td>\n",
       "      <td>344.0</td>\n",
       "      <td>9.279398</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>b32f83615c36a983c147d57d68f20804</th>\n",
       "      <td>13.0</td>\n",
       "      <td>301.0</td>\n",
       "      <td>312.0</td>\n",
       "      <td>9.191790</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>80ed49605a244a6b8df7a0185125075d</th>\n",
       "      <td>12.0</td>\n",
       "      <td>265.0</td>\n",
       "      <td>276.0</td>\n",
       "      <td>9.177153</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5c7eec1661af8e5b46f1e227fae6e1ee</th>\n",
       "      <td>17.0</td>\n",
       "      <td>315.0</td>\n",
       "      <td>355.0</td>\n",
       "      <td>9.145926</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2da10b67a4882e14d257c728eda763c3</th>\n",
       "      <td>14.0</td>\n",
       "      <td>351.0</td>\n",
       "      <td>358.0</td>\n",
       "      <td>9.135718</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25e5fe3494dcb7d0de25fcd6f6b499d9</th>\n",
       "      <td>15.0</td>\n",
       "      <td>344.0</td>\n",
       "      <td>367.0</td>\n",
       "      <td>9.111223</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0b937997e6f55b43f95a14c1c6fe94e2</th>\n",
       "      <td>14.0</td>\n",
       "      <td>345.0</td>\n",
       "      <td>356.0</td>\n",
       "      <td>9.073500</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>bcf0ea738c6bb798b7d05f2603f8fccf</th>\n",
       "      <td>15.0</td>\n",
       "      <td>332.0</td>\n",
       "      <td>359.0</td>\n",
       "      <td>9.049101</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>52558a8c536649d21e274da275ce1c8e</th>\n",
       "      <td>13.0</td>\n",
       "      <td>297.0</td>\n",
       "      <td>313.0</td>\n",
       "      <td>9.013280</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>d88c16f53bf1f538e5dad005530611a9</th>\n",
       "      <td>10.0</td>\n",
       "      <td>200.0</td>\n",
       "      <td>211.0</td>\n",
       "      <td>9.000838</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>052a3c248db5edab01d32871d5acdbfa</th>\n",
       "      <td>14.0</td>\n",
       "      <td>359.0</td>\n",
       "      <td>366.0</td>\n",
       "      <td>9.000642</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>a92534133444b5028d12a129b1b128f7</th>\n",
       "      <td>6.0</td>\n",
       "      <td>66.0</td>\n",
       "      <td>75.0</td>\n",
       "      <td>8.987992</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>bcf1f968bf4bf2b9241f77366a4aee86</th>\n",
       "      <td>14.0</td>\n",
       "      <td>360.0</td>\n",
       "      <td>367.0</td>\n",
       "      <td>8.984043</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4278f21f39173ebb616faa118c22384a</th>\n",
       "      <td>11.0</td>\n",
       "      <td>258.0</td>\n",
       "      <td>261.0</td>\n",
       "      <td>8.948836</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>feeb6791f303fd34059c06d368f218e3</th>\n",
       "      <td>14.0</td>\n",
       "      <td>336.0</td>\n",
       "      <td>354.0</td>\n",
       "      <td>8.901579</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>51b3b64735d01118ff09427f8082277a</th>\n",
       "      <td>11.0</td>\n",
       "      <td>234.0</td>\n",
       "      <td>248.0</td>\n",
       "      <td>8.885969</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9459310958c7741a71359ef3a151b9ed</th>\n",
       "      <td>9.0</td>\n",
       "      <td>180.0</td>\n",
       "      <td>187.0</td>\n",
       "      <td>8.846049</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6116c18166249fed2e39ee8f3a0733cf</th>\n",
       "      <td>11.0</td>\n",
       "      <td>239.0</td>\n",
       "      <td>252.0</td>\n",
       "      <td>8.841880</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>51bf88a0aaf6da3d94ab246c9c857017</th>\n",
       "      <td>12.0</td>\n",
       "      <td>292.0</td>\n",
       "      <td>300.0</td>\n",
       "      <td>8.798157</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ea623f715c09492b8a565e2b8dfff305</th>\n",
       "      <td>14.0</td>\n",
       "      <td>333.0</td>\n",
       "      <td>355.0</td>\n",
       "      <td>8.746280</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>51ff794d843ce28fb7ca8885094979a7</th>\n",
       "      <td>6.0</td>\n",
       "      <td>86.0</td>\n",
       "      <td>90.0</td>\n",
       "      <td>8.693569</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8770527c316c8732cec34377f12e9299</th>\n",
       "      <td>9.0</td>\n",
       "      <td>188.0</td>\n",
       "      <td>195.0</td>\n",
       "      <td>8.645529</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>950e9dd87391b872ae7b1afe1c778bfd</th>\n",
       "      <td>13.0</td>\n",
       "      <td>346.0</td>\n",
       "      <td>353.0</td>\n",
       "      <td>8.576180</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7a04d32e6941f996df21c52d1ccbc7e8</th>\n",
       "      <td>13.0</td>\n",
       "      <td>347.0</td>\n",
       "      <td>355.0</td>\n",
       "      <td>8.522919</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7e70507512f04b9a1fc0594534443c8e</th>\n",
       "      <td>14.0</td>\n",
       "      <td>341.0</td>\n",
       "      <td>367.0</td>\n",
       "      <td>8.424063</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>100 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                  frequency  recency      T  \\\n",
       "Id                                                            \n",
       "a5fadc51b1ae844ad9a70ad3cfa46a64       95.0    371.0  371.0   \n",
       "48a503edbaded96a3be27deee11967a1       46.0    363.0  371.0   \n",
       "9f447f9415a380ac2eeee7df49c6ee7e       23.0    163.0  166.0   \n",
       "5f01420f0edda6555df5ce1cc62b986c       39.0    350.0  370.0   \n",
       "8d2ce54737dd404d20cadf1405d46dc8       33.0    360.0  362.0   \n",
       "a62a17bb46864da2c6da691d838971b3       30.0    366.0  366.0   \n",
       "3b11478939967e896ae2619615650f97       20.0    190.0  194.0   \n",
       "2ad9a83ee23110d8c2f4c01600b94f20       29.0    357.0  364.0   \n",
       "75fda9ea22086bf3814ff8c3f53de8ca       28.0    348.0  355.0   \n",
       "30aa99d3357244cf38ca04eadef1473a       18.0    183.0  189.0   \n",
       "a719d6643a7832535de9aded2f467825       26.0    360.0  368.0   \n",
       "a1c8d419a97af1f7152e21c0dddfcbce       24.0    369.0  370.0   \n",
       "cdf642859ae8d9a1e489c3a655cba827       23.0    343.0  350.0   \n",
       "ed2b4332b3ca253cfbb0ffb54d3f5ae0       23.0    337.0  348.0   \n",
       "880474a82e33389ae7f377b2ea35bbdd       23.0    364.0  371.0   \n",
       "5ac5ed64cd99ed2a8403b7a927e644ef       23.0    222.0  250.0   \n",
       "16134915d822fe17588ae585935e1e81       22.0    356.0  362.0   \n",
       "f6a4f156c817ecb376d678f2aafdc570       20.0    309.0  316.0   \n",
       "ca2202a96c2de6ca6b8a37a4a73fa730        7.0     12.0   20.0   \n",
       "742d5a52d4df7cb14246d7f390de5d8a       21.0    340.0  356.0   \n",
       "663b3df249f614305792ec7bf1dead30       20.0    347.0  352.0   \n",
       "bb5d927e9f1aefa4db1881579e055a3e       20.0    348.0  355.0   \n",
       "e675875db3648ab4dbbd52768296425c       12.0    145.0  152.0   \n",
       "61d2370d7ccfe63adf6ab388d7415dc2       19.0    340.0  342.0   \n",
       "2b75f007b50d6b21e1501f47dc8d632b       19.0    355.0  355.0   \n",
       "cd1751286a561d51655ee47246b3019d       19.0    351.0  355.0   \n",
       "9cb19c3fc7311aae01cf16571b528001       17.0    263.0  279.0   \n",
       "4f672b7c5d7a4214b08ec6906163c980       14.0    208.0  216.0   \n",
       "7108b2bb5964a79ee1e6c8a9acde8107       16.0    272.0  279.0   \n",
       "ca5cc4f526fcff10e85ce2685999e2c1        9.0     84.0   92.0   \n",
       "...                                     ...      ...    ...   \n",
       "14f57af72e3d2a16b7b4c088cbe3196f       14.0    334.0  341.0   \n",
       "13cb5c06f8d3c1c18e14539c16088e21       15.0    356.0  370.0   \n",
       "922bdaa119d44065b17b0447255646ae        8.0    144.0  144.0   \n",
       "1200a4df001756f0fc97e0142c069442       11.0    231.0  239.0   \n",
       "47810b7c7fc24d3cf8287d03de62448b       14.0    342.0  349.0   \n",
       "d0ad6b624dfc37784a755144be8c76a6       14.0    333.0  344.0   \n",
       "b32f83615c36a983c147d57d68f20804       13.0    301.0  312.0   \n",
       "80ed49605a244a6b8df7a0185125075d       12.0    265.0  276.0   \n",
       "5c7eec1661af8e5b46f1e227fae6e1ee       17.0    315.0  355.0   \n",
       "2da10b67a4882e14d257c728eda763c3       14.0    351.0  358.0   \n",
       "25e5fe3494dcb7d0de25fcd6f6b499d9       15.0    344.0  367.0   \n",
       "0b937997e6f55b43f95a14c1c6fe94e2       14.0    345.0  356.0   \n",
       "bcf0ea738c6bb798b7d05f2603f8fccf       15.0    332.0  359.0   \n",
       "52558a8c536649d21e274da275ce1c8e       13.0    297.0  313.0   \n",
       "d88c16f53bf1f538e5dad005530611a9       10.0    200.0  211.0   \n",
       "052a3c248db5edab01d32871d5acdbfa       14.0    359.0  366.0   \n",
       "a92534133444b5028d12a129b1b128f7        6.0     66.0   75.0   \n",
       "bcf1f968bf4bf2b9241f77366a4aee86       14.0    360.0  367.0   \n",
       "4278f21f39173ebb616faa118c22384a       11.0    258.0  261.0   \n",
       "feeb6791f303fd34059c06d368f218e3       14.0    336.0  354.0   \n",
       "51b3b64735d01118ff09427f8082277a       11.0    234.0  248.0   \n",
       "9459310958c7741a71359ef3a151b9ed        9.0    180.0  187.0   \n",
       "6116c18166249fed2e39ee8f3a0733cf       11.0    239.0  252.0   \n",
       "51bf88a0aaf6da3d94ab246c9c857017       12.0    292.0  300.0   \n",
       "ea623f715c09492b8a565e2b8dfff305       14.0    333.0  355.0   \n",
       "51ff794d843ce28fb7ca8885094979a7        6.0     86.0   90.0   \n",
       "8770527c316c8732cec34377f12e9299        9.0    188.0  195.0   \n",
       "950e9dd87391b872ae7b1afe1c778bfd       13.0    346.0  353.0   \n",
       "7a04d32e6941f996df21c52d1ccbc7e8       13.0    347.0  355.0   \n",
       "7e70507512f04b9a1fc0594534443c8e       14.0    341.0  367.0   \n",
       "\n",
       "                                  predicted_purchases  \n",
       "Id                                                     \n",
       "a5fadc51b1ae844ad9a70ad3cfa46a64            60.718377  \n",
       "48a503edbaded96a3be27deee11967a1            28.675147  \n",
       "9f447f9415a380ac2eeee7df49c6ee7e            23.715042  \n",
       "5f01420f0edda6555df5ce1cc62b986c            22.163748  \n",
       "8d2ce54737dd404d20cadf1405d46dc8            21.424866  \n",
       "a62a17bb46864da2c6da691d838971b3            19.431837  \n",
       "3b11478939967e896ae2619615650f97            18.971988  \n",
       "2ad9a83ee23110d8c2f4c01600b94f20            18.508699  \n",
       "75fda9ea22086bf3814ff8c3f53de8ca            18.174781  \n",
       "30aa99d3357244cf38ca04eadef1473a            17.190421  \n",
       "a719d6643a7832535de9aded2f467825            16.444873  \n",
       "a1c8d419a97af1f7152e21c0dddfcbce            15.423088  \n",
       "cdf642859ae8d9a1e489c3a655cba827            15.110830  \n",
       "ed2b4332b3ca253cfbb0ffb54d3f5ae0            14.952128  \n",
       "880474a82e33389ae7f377b2ea35bbdd            14.539901  \n",
       "5ac5ed64cd99ed2a8403b7a927e644ef            14.475066  \n",
       "16134915d822fe17588ae585935e1e81            14.186154  \n",
       "f6a4f156c817ecb376d678f2aafdc570            14.068079  \n",
       "ca2202a96c2de6ca6b8a37a4a73fa730            13.592224  \n",
       "742d5a52d4df7cb14246d7f390de5d8a            13.205288  \n",
       "663b3df249f614305792ec7bf1dead30            13.191422  \n",
       "bb5d927e9f1aefa4db1881579e055a3e            13.045209  \n",
       "e675875db3648ab4dbbd52768296425c            12.929123  \n",
       "61d2370d7ccfe63adf6ab388d7415dc2            12.877917  \n",
       "2b75f007b50d6b21e1501f47dc8d632b            12.612052  \n",
       "cd1751286a561d51655ee47246b3019d            12.500814  \n",
       "9cb19c3fc7311aae01cf16571b528001            12.447227  \n",
       "4f672b7c5d7a4214b08ec6906163c980            12.423322  \n",
       "7108b2bb5964a79ee1e6c8a9acde8107            12.222910  \n",
       "ca5cc4f526fcff10e85ce2685999e2c1            12.145760  \n",
       "...                                               ...  \n",
       "14f57af72e3d2a16b7b4c088cbe3196f             9.436900  \n",
       "13cb5c06f8d3c1c18e14539c16088e21             9.377351  \n",
       "922bdaa119d44065b17b0447255646ae             9.324966  \n",
       "1200a4df001756f0fc97e0142c069442             9.294833  \n",
       "47810b7c7fc24d3cf8287d03de62448b             9.292691  \n",
       "d0ad6b624dfc37784a755144be8c76a6             9.279398  \n",
       "b32f83615c36a983c147d57d68f20804             9.191790  \n",
       "80ed49605a244a6b8df7a0185125075d             9.177153  \n",
       "5c7eec1661af8e5b46f1e227fae6e1ee             9.145926  \n",
       "2da10b67a4882e14d257c728eda763c3             9.135718  \n",
       "25e5fe3494dcb7d0de25fcd6f6b499d9             9.111223  \n",
       "0b937997e6f55b43f95a14c1c6fe94e2             9.073500  \n",
       "bcf0ea738c6bb798b7d05f2603f8fccf             9.049101  \n",
       "52558a8c536649d21e274da275ce1c8e             9.013280  \n",
       "d88c16f53bf1f538e5dad005530611a9             9.000838  \n",
       "052a3c248db5edab01d32871d5acdbfa             9.000642  \n",
       "a92534133444b5028d12a129b1b128f7             8.987992  \n",
       "bcf1f968bf4bf2b9241f77366a4aee86             8.984043  \n",
       "4278f21f39173ebb616faa118c22384a             8.948836  \n",
       "feeb6791f303fd34059c06d368f218e3             8.901579  \n",
       "51b3b64735d01118ff09427f8082277a             8.885969  \n",
       "9459310958c7741a71359ef3a151b9ed             8.846049  \n",
       "6116c18166249fed2e39ee8f3a0733cf             8.841880  \n",
       "51bf88a0aaf6da3d94ab246c9c857017             8.798157  \n",
       "ea623f715c09492b8a565e2b8dfff305             8.746280  \n",
       "51ff794d843ce28fb7ca8885094979a7             8.693569  \n",
       "8770527c316c8732cec34377f12e9299             8.645529  \n",
       "950e9dd87391b872ae7b1afe1c778bfd             8.576180  \n",
       "7a04d32e6941f996df21c52d1ccbc7e8             8.522919  \n",
       "7e70507512f04b9a1fc0594534443c8e             8.424063  \n",
       "\n",
       "[100 rows x 4 columns]"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "t = 365\n",
    "summary['predicted_purchases'] = bgf.predict(t,summary['frequency'],summary['recency'],summary['T'])\n",
    "summary.sort_values(by='predicted_purchases',ascending=False).head(100)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "These are the customers with the top 100 projected numbers of purchases over the next 12 months.  \n",
    "\n",
    "Now that we have projections for the number of purchases that the customers will make in the next year, we can use [Fader and Hardie's Gamma-Gamma Model of Monetary Value](http://www.brucehardie.com/notes/025/gamma_gamma.pdf) to calculate Customer Lifetime Value.  First, we need to append the average transaction amount for each customer to the summary dataframe.  We will also create a subset of this data that only includes returning customers (customers with more than one transaction)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "df_purchase_avg = pd.DataFrame(df.groupby('Id')['PurchaseValue'].mean())\n",
    "summary = summary.merge(df_purchase_avg,left_index=True,right_index=True)\n",
    "summary.rename(columns={'PurchaseValue':'monetary_value'},inplace=True)\n",
    "returning_customers_summary = summary[summary['frequency'] > 0]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The Gamma-Gamma Model assumes that a customer's frequency and monetary value are independent of each other.  Let's check the correlation between these two fields to see if this holds true with our data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>monetary_value</th>\n",
       "      <th>frequency</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>monetary_value</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.011513</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>frequency</th>\n",
       "      <td>0.011513</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                monetary_value  frequency\n",
       "monetary_value        1.000000   0.011513\n",
       "frequency             0.011513   1.000000"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "returning_customers_summary[['monetary_value','frequency']].corr()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Pearson's Correlation Coefficient is nearly zero between frequency and monetary value.  Therefore, the assumption that these two variables are independent of each other is most likely true and we can proceed with the Gamma-Gamma Model."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "from lifetimes import GammaGammaFitter"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<lifetimes.GammaGammaFitter: fitted with 9652 subjects, p: 7.93, q: 3.87, v: 87.09>"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ggf = GammaGammaFitter()\n",
    "ggf.fit(returning_customers_summary['frequency'],returning_customers_summary['monetary_value'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now that the model has been fit, we can predict the top 100 customers in terms of projected total purchase value over the next 12 months."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "returning_customers_summary['predicted_avg_purchase'] = ggf.conditional_expected_average_profit(summary['frequency'],\n",
    "                                                                                                summary['monetary_value'])\n",
    "returning_customers_summary['predicted_total_spent'] = \\\n",
    "            returning_customers_summary['predicted_purchases'] * returning_customers_summary['predicted_avg_purchase']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>frequency</th>\n",
       "      <th>recency</th>\n",
       "      <th>T</th>\n",
       "      <th>predicted_purchases</th>\n",
       "      <th>monetary_value</th>\n",
       "      <th>predicted_avg_purchase</th>\n",
       "      <th>predicted_total_spent</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Id</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>a5fadc51b1ae844ad9a70ad3cfa46a64</th>\n",
       "      <td>95.0</td>\n",
       "      <td>371.0</td>\n",
       "      <td>371.0</td>\n",
       "      <td>60.718377</td>\n",
       "      <td>306.023481</td>\n",
       "      <td>305.775264</td>\n",
       "      <td>18566.177851</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5ac5ed64cd99ed2a8403b7a927e644ef</th>\n",
       "      <td>23.0</td>\n",
       "      <td>222.0</td>\n",
       "      <td>250.0</td>\n",
       "      <td>14.475066</td>\n",
       "      <td>771.444615</td>\n",
       "      <td>763.221185</td>\n",
       "      <td>11047.676642</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>60c19a709e3ced2d16d7100eb1069df5</th>\n",
       "      <td>10.0</td>\n",
       "      <td>292.0</td>\n",
       "      <td>300.0</td>\n",
       "      <td>7.371620</td>\n",
       "      <td>1285.566364</td>\n",
       "      <td>1249.068736</td>\n",
       "      <td>9207.659881</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>dca76db00cc59dfbcdcc97c8bbc7f9f1</th>\n",
       "      <td>19.0</td>\n",
       "      <td>339.0</td>\n",
       "      <td>364.0</td>\n",
       "      <td>11.286749</td>\n",
       "      <td>797.443333</td>\n",
       "      <td>787.035039</td>\n",
       "      <td>8883.067235</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ca2202a96c2de6ca6b8a37a4a73fa730</th>\n",
       "      <td>7.0</td>\n",
       "      <td>12.0</td>\n",
       "      <td>20.0</td>\n",
       "      <td>13.592224</td>\n",
       "      <td>655.932500</td>\n",
       "      <td>635.515389</td>\n",
       "      <td>8638.067559</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>a719d6643a7832535de9aded2f467825</th>\n",
       "      <td>26.0</td>\n",
       "      <td>360.0</td>\n",
       "      <td>368.0</td>\n",
       "      <td>16.444873</td>\n",
       "      <td>391.146429</td>\n",
       "      <td>389.079885</td>\n",
       "      <td>6398.369439</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9cb19c3fc7311aae01cf16571b528001</th>\n",
       "      <td>17.0</td>\n",
       "      <td>263.0</td>\n",
       "      <td>279.0</td>\n",
       "      <td>12.447227</td>\n",
       "      <td>516.350000</td>\n",
       "      <td>510.602272</td>\n",
       "      <td>6355.582328</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>742d5a52d4df7cb14246d7f390de5d8a</th>\n",
       "      <td>21.0</td>\n",
       "      <td>340.0</td>\n",
       "      <td>356.0</td>\n",
       "      <td>13.205288</td>\n",
       "      <td>482.105652</td>\n",
       "      <td>478.014357</td>\n",
       "      <td>6312.317481</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32df931a012f04da70265457d43e7da2</th>\n",
       "      <td>4.0</td>\n",
       "      <td>89.0</td>\n",
       "      <td>102.0</td>\n",
       "      <td>5.478943</td>\n",
       "      <td>1219.225000</td>\n",
       "      <td>1138.027875</td>\n",
       "      <td>6235.189612</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5f01420f0edda6555df5ce1cc62b986c</th>\n",
       "      <td>39.0</td>\n",
       "      <td>350.0</td>\n",
       "      <td>370.0</td>\n",
       "      <td>22.163748</td>\n",
       "      <td>271.127955</td>\n",
       "      <td>270.847452</td>\n",
       "      <td>6002.994703</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30aa99d3357244cf38ca04eadef1473a</th>\n",
       "      <td>18.0</td>\n",
       "      <td>183.0</td>\n",
       "      <td>189.0</td>\n",
       "      <td>17.190421</td>\n",
       "      <td>343.237368</td>\n",
       "      <td>341.214766</td>\n",
       "      <td>5865.625567</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>a62a17bb46864da2c6da691d838971b3</th>\n",
       "      <td>30.0</td>\n",
       "      <td>366.0</td>\n",
       "      <td>366.0</td>\n",
       "      <td>19.431837</td>\n",
       "      <td>300.672813</td>\n",
       "      <td>299.956983</td>\n",
       "      <td>5828.715235</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>48a503edbaded96a3be27deee11967a1</th>\n",
       "      <td>46.0</td>\n",
       "      <td>363.0</td>\n",
       "      <td>371.0</td>\n",
       "      <td>28.675147</td>\n",
       "      <td>202.915000</td>\n",
       "      <td>203.209346</td>\n",
       "      <td>5827.057904</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>eba458987dc67827871c1d4d92e646e1</th>\n",
       "      <td>7.0</td>\n",
       "      <td>266.0</td>\n",
       "      <td>266.0</td>\n",
       "      <td>5.753705</td>\n",
       "      <td>1043.908889</td>\n",
       "      <td>1004.418496</td>\n",
       "      <td>5779.128040</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ed2b4332b3ca253cfbb0ffb54d3f5ae0</th>\n",
       "      <td>23.0</td>\n",
       "      <td>337.0</td>\n",
       "      <td>348.0</td>\n",
       "      <td>14.952128</td>\n",
       "      <td>386.107778</td>\n",
       "      <td>383.853919</td>\n",
       "      <td>5739.432987</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>a92534133444b5028d12a129b1b128f7</th>\n",
       "      <td>6.0</td>\n",
       "      <td>66.0</td>\n",
       "      <td>75.0</td>\n",
       "      <td>8.987992</td>\n",
       "      <td>646.002000</td>\n",
       "      <td>622.940549</td>\n",
       "      <td>5598.984723</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>66162981fc95e268e45bbfc738059687</th>\n",
       "      <td>9.0</td>\n",
       "      <td>253.0</td>\n",
       "      <td>264.0</td>\n",
       "      <td>7.151782</td>\n",
       "      <td>784.979091</td>\n",
       "      <td>763.934954</td>\n",
       "      <td>5463.496360</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>f11a653807d6bde53e4757f53e098c96</th>\n",
       "      <td>16.0</td>\n",
       "      <td>316.0</td>\n",
       "      <td>329.0</td>\n",
       "      <td>10.799136</td>\n",
       "      <td>509.909444</td>\n",
       "      <td>503.952892</td>\n",
       "      <td>5442.255827</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2b75f007b50d6b21e1501f47dc8d632b</th>\n",
       "      <td>19.0</td>\n",
       "      <td>355.0</td>\n",
       "      <td>355.0</td>\n",
       "      <td>12.612052</td>\n",
       "      <td>423.311905</td>\n",
       "      <td>419.896988</td>\n",
       "      <td>5295.762728</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>98f8e41f45721cbe49a3147f6cf62432</th>\n",
       "      <td>3.0</td>\n",
       "      <td>39.0</td>\n",
       "      <td>71.0</td>\n",
       "      <td>4.147393</td>\n",
       "      <td>1401.225000</td>\n",
       "      <td>1276.283110</td>\n",
       "      <td>5293.247450</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2f486887c2edb2571d32c8cd15301711</th>\n",
       "      <td>9.0</td>\n",
       "      <td>336.0</td>\n",
       "      <td>343.0</td>\n",
       "      <td>6.108056</td>\n",
       "      <td>875.478000</td>\n",
       "      <td>850.935300</td>\n",
       "      <td>5197.560536</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>f09ff1c6c4ac8ea95d8621a94bb325fd</th>\n",
       "      <td>7.0</td>\n",
       "      <td>208.0</td>\n",
       "      <td>215.0</td>\n",
       "      <td>6.443771</td>\n",
       "      <td>818.107500</td>\n",
       "      <td>789.717714</td>\n",
       "      <td>5088.759759</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4f672b7c5d7a4214b08ec6906163c980</th>\n",
       "      <td>14.0</td>\n",
       "      <td>208.0</td>\n",
       "      <td>216.0</td>\n",
       "      <td>12.423322</td>\n",
       "      <td>411.254667</td>\n",
       "      <td>406.954712</td>\n",
       "      <td>5055.729472</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ede2a476c3894cf65d1619987d148422</th>\n",
       "      <td>9.0</td>\n",
       "      <td>240.0</td>\n",
       "      <td>242.0</td>\n",
       "      <td>7.741751</td>\n",
       "      <td>667.693636</td>\n",
       "      <td>651.183593</td>\n",
       "      <td>5041.301217</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7863835a6120bfa52ccd665bfe03e9c8</th>\n",
       "      <td>16.0</td>\n",
       "      <td>361.0</td>\n",
       "      <td>361.0</td>\n",
       "      <td>10.520608</td>\n",
       "      <td>443.895882</td>\n",
       "      <td>439.399523</td>\n",
       "      <td>4622.750048</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>922bdaa119d44065b17b0447255646ae</th>\n",
       "      <td>8.0</td>\n",
       "      <td>144.0</td>\n",
       "      <td>144.0</td>\n",
       "      <td>9.324966</td>\n",
       "      <td>497.575000</td>\n",
       "      <td>486.453563</td>\n",
       "      <td>4536.163053</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>cd4cb9ec252a085ed4d2d3af7c18280a</th>\n",
       "      <td>7.0</td>\n",
       "      <td>295.0</td>\n",
       "      <td>308.0</td>\n",
       "      <td>5.078936</td>\n",
       "      <td>916.495000</td>\n",
       "      <td>883.268393</td>\n",
       "      <td>4486.063482</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ca84dc4660f651a11a82041300367fae</th>\n",
       "      <td>4.0</td>\n",
       "      <td>73.0</td>\n",
       "      <td>78.0</td>\n",
       "      <td>6.368684</td>\n",
       "      <td>732.200000</td>\n",
       "      <td>691.412502</td>\n",
       "      <td>4403.387613</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>580a57e214a9d4df715accd2fe7065ba</th>\n",
       "      <td>6.0</td>\n",
       "      <td>46.0</td>\n",
       "      <td>51.0</td>\n",
       "      <td>10.478246</td>\n",
       "      <td>414.444286</td>\n",
       "      <td>404.555762</td>\n",
       "      <td>4239.034879</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>a63d0d4fe5bc662678bbbe6fbde1d900</th>\n",
       "      <td>7.0</td>\n",
       "      <td>285.0</td>\n",
       "      <td>314.0</td>\n",
       "      <td>4.789566</td>\n",
       "      <td>904.985000</td>\n",
       "      <td>872.324235</td>\n",
       "      <td>4178.054323</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1091a17e1def85799828dadf9c28f3f2</th>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>29.0</td>\n",
       "      <td>2.545383</td>\n",
       "      <td>1550.180000</td>\n",
       "      <td>1202.175329</td>\n",
       "      <td>3059.996252</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ba5cf27a059fb5e6357dc3e7f70dbab2</th>\n",
       "      <td>3.0</td>\n",
       "      <td>133.0</td>\n",
       "      <td>190.0</td>\n",
       "      <td>2.644494</td>\n",
       "      <td>1253.165000</td>\n",
       "      <td>1144.162132</td>\n",
       "      <td>3025.730066</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5911b41fe8c5cd07a67571724f073e20</th>\n",
       "      <td>10.0</td>\n",
       "      <td>227.0</td>\n",
       "      <td>261.0</td>\n",
       "      <td>7.064169</td>\n",
       "      <td>434.455833</td>\n",
       "      <td>427.685606</td>\n",
       "      <td>3021.243322</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>e4d423cc47a17c9ca7cea0c12e8e22be</th>\n",
       "      <td>4.0</td>\n",
       "      <td>357.0</td>\n",
       "      <td>364.0</td>\n",
       "      <td>2.697776</td>\n",
       "      <td>1189.218000</td>\n",
       "      <td>1110.510627</td>\n",
       "      <td>2995.909406</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>52cc702cd8c70995cf6ba1c762e341d4</th>\n",
       "      <td>9.0</td>\n",
       "      <td>187.0</td>\n",
       "      <td>214.0</td>\n",
       "      <td>7.398727</td>\n",
       "      <td>410.696667</td>\n",
       "      <td>404.121770</td>\n",
       "      <td>2989.986689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>bfd2558c6ca88693b5721952422b5b40</th>\n",
       "      <td>8.0</td>\n",
       "      <td>318.0</td>\n",
       "      <td>326.0</td>\n",
       "      <td>5.626780</td>\n",
       "      <td>542.847778</td>\n",
       "      <td>529.766854</td>\n",
       "      <td>2980.881555</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>e0d591d2bbc656a278b5675d9bcae6c3</th>\n",
       "      <td>11.0</td>\n",
       "      <td>325.0</td>\n",
       "      <td>337.0</td>\n",
       "      <td>7.418159</td>\n",
       "      <td>406.927857</td>\n",
       "      <td>401.630367</td>\n",
       "      <td>2979.357842</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>79d35a58b719d48ca15d7a573f8a2379</th>\n",
       "      <td>11.0</td>\n",
       "      <td>325.0</td>\n",
       "      <td>332.0</td>\n",
       "      <td>7.588608</td>\n",
       "      <td>394.165333</td>\n",
       "      <td>389.274377</td>\n",
       "      <td>2954.050742</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>970c6dfdd3bcc297b26100a96a9a06fe</th>\n",
       "      <td>12.0</td>\n",
       "      <td>368.0</td>\n",
       "      <td>368.0</td>\n",
       "      <td>7.818822</td>\n",
       "      <td>380.715385</td>\n",
       "      <td>376.613848</td>\n",
       "      <td>2944.676469</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2143bad4fa8805114a00a73853ac6ace</th>\n",
       "      <td>11.0</td>\n",
       "      <td>210.0</td>\n",
       "      <td>246.0</td>\n",
       "      <td>7.662536</td>\n",
       "      <td>388.269167</td>\n",
       "      <td>383.566025</td>\n",
       "      <td>2939.088593</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6f9ebbe87978c734ba71c0032e1f3e45</th>\n",
       "      <td>4.0</td>\n",
       "      <td>140.0</td>\n",
       "      <td>142.0</td>\n",
       "      <td>4.905500</td>\n",
       "      <td>629.202000</td>\n",
       "      <td>596.960491</td>\n",
       "      <td>2928.389497</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ac3537cd5ecca62eb010d5fb6f9c57d2</th>\n",
       "      <td>4.0</td>\n",
       "      <td>149.0</td>\n",
       "      <td>171.0</td>\n",
       "      <td>4.146901</td>\n",
       "      <td>745.435000</td>\n",
       "      <td>703.549362</td>\n",
       "      <td>2917.549303</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16d851d4eb417ae73e48392063df8fb6</th>\n",
       "      <td>3.0</td>\n",
       "      <td>131.0</td>\n",
       "      <td>160.0</td>\n",
       "      <td>3.257519</td>\n",
       "      <td>968.885000</td>\n",
       "      <td>890.485572</td>\n",
       "      <td>2900.774054</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>68f0db4c3018522a7c24607851afd762</th>\n",
       "      <td>14.0</td>\n",
       "      <td>305.0</td>\n",
       "      <td>315.0</td>\n",
       "      <td>9.849825</td>\n",
       "      <td>294.268000</td>\n",
       "      <td>292.916097</td>\n",
       "      <td>2885.172402</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4e7ef9579e693b03f57e5086ae7ebd6e</th>\n",
       "      <td>6.0</td>\n",
       "      <td>287.0</td>\n",
       "      <td>287.0</td>\n",
       "      <td>4.722597</td>\n",
       "      <td>632.825714</td>\n",
       "      <td>610.513840</td>\n",
       "      <td>2883.210769</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ab30a46cc9f7d41f0ee382ff9dd48ad0</th>\n",
       "      <td>5.0</td>\n",
       "      <td>148.0</td>\n",
       "      <td>159.0</td>\n",
       "      <td>5.488910</td>\n",
       "      <td>542.698333</td>\n",
       "      <td>522.308704</td>\n",
       "      <td>2866.905512</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>547bfa0c00ff82eb6f62d7a2b2cf4be7</th>\n",
       "      <td>2.0</td>\n",
       "      <td>28.0</td>\n",
       "      <td>38.0</td>\n",
       "      <td>4.384328</td>\n",
       "      <td>727.115000</td>\n",
       "      <td>652.569085</td>\n",
       "      <td>2861.076962</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>a8c6dd0d4f1b33878dfc2f3559e88f87</th>\n",
       "      <td>4.0</td>\n",
       "      <td>363.0</td>\n",
       "      <td>371.0</td>\n",
       "      <td>2.657352</td>\n",
       "      <td>1150.276667</td>\n",
       "      <td>1074.800349</td>\n",
       "      <td>2856.123041</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6d165923bdf6c3055311418ef905c58a</th>\n",
       "      <td>4.0</td>\n",
       "      <td>215.0</td>\n",
       "      <td>268.0</td>\n",
       "      <td>2.919019</td>\n",
       "      <td>1043.784000</td>\n",
       "      <td>977.143633</td>\n",
       "      <td>2852.300370</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42059a7ede026d409ff0f255635d7a08</th>\n",
       "      <td>20.0</td>\n",
       "      <td>268.0</td>\n",
       "      <td>326.0</td>\n",
       "      <td>6.412358</td>\n",
       "      <td>447.623000</td>\n",
       "      <td>443.943681</td>\n",
       "      <td>2846.725929</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0cb0b9356e0a85830bef4f4fda74f92b</th>\n",
       "      <td>11.0</td>\n",
       "      <td>295.0</td>\n",
       "      <td>296.0</td>\n",
       "      <td>8.298700</td>\n",
       "      <td>346.283846</td>\n",
       "      <td>342.918091</td>\n",
       "      <td>2845.774252</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34bff14a781e6eabd03f59ac8f610ed7</th>\n",
       "      <td>3.0</td>\n",
       "      <td>46.0</td>\n",
       "      <td>94.0</td>\n",
       "      <td>3.262338</td>\n",
       "      <td>945.807500</td>\n",
       "      <td>869.892422</td>\n",
       "      <td>2837.883245</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16134915d822fe17588ae585935e1e81</th>\n",
       "      <td>22.0</td>\n",
       "      <td>356.0</td>\n",
       "      <td>362.0</td>\n",
       "      <td>14.186154</td>\n",
       "      <td>199.011250</td>\n",
       "      <td>199.684684</td>\n",
       "      <td>2832.757666</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0425ae0d99a8e31c2758c25cca4ac2f7</th>\n",
       "      <td>8.0</td>\n",
       "      <td>93.0</td>\n",
       "      <td>94.0</td>\n",
       "      <td>11.316598</td>\n",
       "      <td>249.621818</td>\n",
       "      <td>249.232239</td>\n",
       "      <td>2820.461107</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9b306f73da56027310069b44074ef765</th>\n",
       "      <td>4.0</td>\n",
       "      <td>261.0</td>\n",
       "      <td>276.0</td>\n",
       "      <td>3.212192</td>\n",
       "      <td>932.238000</td>\n",
       "      <td>874.852871</td>\n",
       "      <td>2810.195774</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>de70325370400abbce654bb35511812d</th>\n",
       "      <td>13.0</td>\n",
       "      <td>331.0</td>\n",
       "      <td>371.0</td>\n",
       "      <td>7.246892</td>\n",
       "      <td>386.129412</td>\n",
       "      <td>382.188189</td>\n",
       "      <td>2769.676519</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6dce8b4af8ca65025bc7f84457ae13a1</th>\n",
       "      <td>5.0</td>\n",
       "      <td>225.0</td>\n",
       "      <td>287.0</td>\n",
       "      <td>3.249562</td>\n",
       "      <td>891.911250</td>\n",
       "      <td>847.950446</td>\n",
       "      <td>2755.467654</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>dc5cd67f8a03b032237af26b0c92c187</th>\n",
       "      <td>9.0</td>\n",
       "      <td>268.0</td>\n",
       "      <td>278.0</td>\n",
       "      <td>6.949175</td>\n",
       "      <td>401.204000</td>\n",
       "      <td>394.996076</td>\n",
       "      <td>2744.896692</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0994482188804e99710b7ee6d6a9e289</th>\n",
       "      <td>3.0</td>\n",
       "      <td>12.0</td>\n",
       "      <td>25.0</td>\n",
       "      <td>6.275799</td>\n",
       "      <td>458.485000</td>\n",
       "      <td>435.031385</td>\n",
       "      <td>2730.169671</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2e80c8b5d073f8310276f3ddfa2a1130</th>\n",
       "      <td>5.0</td>\n",
       "      <td>255.0</td>\n",
       "      <td>268.0</td>\n",
       "      <td>4.042078</td>\n",
       "      <td>702.280000</td>\n",
       "      <td>671.118927</td>\n",
       "      <td>2712.714794</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>100 rows × 7 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                  frequency  recency      T  \\\n",
       "Id                                                            \n",
       "a5fadc51b1ae844ad9a70ad3cfa46a64       95.0    371.0  371.0   \n",
       "5ac5ed64cd99ed2a8403b7a927e644ef       23.0    222.0  250.0   \n",
       "60c19a709e3ced2d16d7100eb1069df5       10.0    292.0  300.0   \n",
       "dca76db00cc59dfbcdcc97c8bbc7f9f1       19.0    339.0  364.0   \n",
       "ca2202a96c2de6ca6b8a37a4a73fa730        7.0     12.0   20.0   \n",
       "a719d6643a7832535de9aded2f467825       26.0    360.0  368.0   \n",
       "9cb19c3fc7311aae01cf16571b528001       17.0    263.0  279.0   \n",
       "742d5a52d4df7cb14246d7f390de5d8a       21.0    340.0  356.0   \n",
       "32df931a012f04da70265457d43e7da2        4.0     89.0  102.0   \n",
       "5f01420f0edda6555df5ce1cc62b986c       39.0    350.0  370.0   \n",
       "30aa99d3357244cf38ca04eadef1473a       18.0    183.0  189.0   \n",
       "a62a17bb46864da2c6da691d838971b3       30.0    366.0  366.0   \n",
       "48a503edbaded96a3be27deee11967a1       46.0    363.0  371.0   \n",
       "eba458987dc67827871c1d4d92e646e1        7.0    266.0  266.0   \n",
       "ed2b4332b3ca253cfbb0ffb54d3f5ae0       23.0    337.0  348.0   \n",
       "a92534133444b5028d12a129b1b128f7        6.0     66.0   75.0   \n",
       "66162981fc95e268e45bbfc738059687        9.0    253.0  264.0   \n",
       "f11a653807d6bde53e4757f53e098c96       16.0    316.0  329.0   \n",
       "2b75f007b50d6b21e1501f47dc8d632b       19.0    355.0  355.0   \n",
       "98f8e41f45721cbe49a3147f6cf62432        3.0     39.0   71.0   \n",
       "2f486887c2edb2571d32c8cd15301711        9.0    336.0  343.0   \n",
       "f09ff1c6c4ac8ea95d8621a94bb325fd        7.0    208.0  215.0   \n",
       "4f672b7c5d7a4214b08ec6906163c980       14.0    208.0  216.0   \n",
       "ede2a476c3894cf65d1619987d148422        9.0    240.0  242.0   \n",
       "7863835a6120bfa52ccd665bfe03e9c8       16.0    361.0  361.0   \n",
       "922bdaa119d44065b17b0447255646ae        8.0    144.0  144.0   \n",
       "cd4cb9ec252a085ed4d2d3af7c18280a        7.0    295.0  308.0   \n",
       "ca84dc4660f651a11a82041300367fae        4.0     73.0   78.0   \n",
       "580a57e214a9d4df715accd2fe7065ba        6.0     46.0   51.0   \n",
       "a63d0d4fe5bc662678bbbe6fbde1d900        7.0    285.0  314.0   \n",
       "...                                     ...      ...    ...   \n",
       "1091a17e1def85799828dadf9c28f3f2        1.0      2.0   29.0   \n",
       "ba5cf27a059fb5e6357dc3e7f70dbab2        3.0    133.0  190.0   \n",
       "5911b41fe8c5cd07a67571724f073e20       10.0    227.0  261.0   \n",
       "e4d423cc47a17c9ca7cea0c12e8e22be        4.0    357.0  364.0   \n",
       "52cc702cd8c70995cf6ba1c762e341d4        9.0    187.0  214.0   \n",
       "bfd2558c6ca88693b5721952422b5b40        8.0    318.0  326.0   \n",
       "e0d591d2bbc656a278b5675d9bcae6c3       11.0    325.0  337.0   \n",
       "79d35a58b719d48ca15d7a573f8a2379       11.0    325.0  332.0   \n",
       "970c6dfdd3bcc297b26100a96a9a06fe       12.0    368.0  368.0   \n",
       "2143bad4fa8805114a00a73853ac6ace       11.0    210.0  246.0   \n",
       "6f9ebbe87978c734ba71c0032e1f3e45        4.0    140.0  142.0   \n",
       "ac3537cd5ecca62eb010d5fb6f9c57d2        4.0    149.0  171.0   \n",
       "16d851d4eb417ae73e48392063df8fb6        3.0    131.0  160.0   \n",
       "68f0db4c3018522a7c24607851afd762       14.0    305.0  315.0   \n",
       "4e7ef9579e693b03f57e5086ae7ebd6e        6.0    287.0  287.0   \n",
       "ab30a46cc9f7d41f0ee382ff9dd48ad0        5.0    148.0  159.0   \n",
       "547bfa0c00ff82eb6f62d7a2b2cf4be7        2.0     28.0   38.0   \n",
       "a8c6dd0d4f1b33878dfc2f3559e88f87        4.0    363.0  371.0   \n",
       "6d165923bdf6c3055311418ef905c58a        4.0    215.0  268.0   \n",
       "42059a7ede026d409ff0f255635d7a08       20.0    268.0  326.0   \n",
       "0cb0b9356e0a85830bef4f4fda74f92b       11.0    295.0  296.0   \n",
       "34bff14a781e6eabd03f59ac8f610ed7        3.0     46.0   94.0   \n",
       "16134915d822fe17588ae585935e1e81       22.0    356.0  362.0   \n",
       "0425ae0d99a8e31c2758c25cca4ac2f7        8.0     93.0   94.0   \n",
       "9b306f73da56027310069b44074ef765        4.0    261.0  276.0   \n",
       "de70325370400abbce654bb35511812d       13.0    331.0  371.0   \n",
       "6dce8b4af8ca65025bc7f84457ae13a1        5.0    225.0  287.0   \n",
       "dc5cd67f8a03b032237af26b0c92c187        9.0    268.0  278.0   \n",
       "0994482188804e99710b7ee6d6a9e289        3.0     12.0   25.0   \n",
       "2e80c8b5d073f8310276f3ddfa2a1130        5.0    255.0  268.0   \n",
       "\n",
       "                                  predicted_purchases  monetary_value  \\\n",
       "Id                                                                      \n",
       "a5fadc51b1ae844ad9a70ad3cfa46a64            60.718377      306.023481   \n",
       "5ac5ed64cd99ed2a8403b7a927e644ef            14.475066      771.444615   \n",
       "60c19a709e3ced2d16d7100eb1069df5             7.371620     1285.566364   \n",
       "dca76db00cc59dfbcdcc97c8bbc7f9f1            11.286749      797.443333   \n",
       "ca2202a96c2de6ca6b8a37a4a73fa730            13.592224      655.932500   \n",
       "a719d6643a7832535de9aded2f467825            16.444873      391.146429   \n",
       "9cb19c3fc7311aae01cf16571b528001            12.447227      516.350000   \n",
       "742d5a52d4df7cb14246d7f390de5d8a            13.205288      482.105652   \n",
       "32df931a012f04da70265457d43e7da2             5.478943     1219.225000   \n",
       "5f01420f0edda6555df5ce1cc62b986c            22.163748      271.127955   \n",
       "30aa99d3357244cf38ca04eadef1473a            17.190421      343.237368   \n",
       "a62a17bb46864da2c6da691d838971b3            19.431837      300.672813   \n",
       "48a503edbaded96a3be27deee11967a1            28.675147      202.915000   \n",
       "eba458987dc67827871c1d4d92e646e1             5.753705     1043.908889   \n",
       "ed2b4332b3ca253cfbb0ffb54d3f5ae0            14.952128      386.107778   \n",
       "a92534133444b5028d12a129b1b128f7             8.987992      646.002000   \n",
       "66162981fc95e268e45bbfc738059687             7.151782      784.979091   \n",
       "f11a653807d6bde53e4757f53e098c96            10.799136      509.909444   \n",
       "2b75f007b50d6b21e1501f47dc8d632b            12.612052      423.311905   \n",
       "98f8e41f45721cbe49a3147f6cf62432             4.147393     1401.225000   \n",
       "2f486887c2edb2571d32c8cd15301711             6.108056      875.478000   \n",
       "f09ff1c6c4ac8ea95d8621a94bb325fd             6.443771      818.107500   \n",
       "4f672b7c5d7a4214b08ec6906163c980            12.423322      411.254667   \n",
       "ede2a476c3894cf65d1619987d148422             7.741751      667.693636   \n",
       "7863835a6120bfa52ccd665bfe03e9c8            10.520608      443.895882   \n",
       "922bdaa119d44065b17b0447255646ae             9.324966      497.575000   \n",
       "cd4cb9ec252a085ed4d2d3af7c18280a             5.078936      916.495000   \n",
       "ca84dc4660f651a11a82041300367fae             6.368684      732.200000   \n",
       "580a57e214a9d4df715accd2fe7065ba            10.478246      414.444286   \n",
       "a63d0d4fe5bc662678bbbe6fbde1d900             4.789566      904.985000   \n",
       "...                                               ...             ...   \n",
       "1091a17e1def85799828dadf9c28f3f2             2.545383     1550.180000   \n",
       "ba5cf27a059fb5e6357dc3e7f70dbab2             2.644494     1253.165000   \n",
       "5911b41fe8c5cd07a67571724f073e20             7.064169      434.455833   \n",
       "e4d423cc47a17c9ca7cea0c12e8e22be             2.697776     1189.218000   \n",
       "52cc702cd8c70995cf6ba1c762e341d4             7.398727      410.696667   \n",
       "bfd2558c6ca88693b5721952422b5b40             5.626780      542.847778   \n",
       "e0d591d2bbc656a278b5675d9bcae6c3             7.418159      406.927857   \n",
       "79d35a58b719d48ca15d7a573f8a2379             7.588608      394.165333   \n",
       "970c6dfdd3bcc297b26100a96a9a06fe             7.818822      380.715385   \n",
       "2143bad4fa8805114a00a73853ac6ace             7.662536      388.269167   \n",
       "6f9ebbe87978c734ba71c0032e1f3e45             4.905500      629.202000   \n",
       "ac3537cd5ecca62eb010d5fb6f9c57d2             4.146901      745.435000   \n",
       "16d851d4eb417ae73e48392063df8fb6             3.257519      968.885000   \n",
       "68f0db4c3018522a7c24607851afd762             9.849825      294.268000   \n",
       "4e7ef9579e693b03f57e5086ae7ebd6e             4.722597      632.825714   \n",
       "ab30a46cc9f7d41f0ee382ff9dd48ad0             5.488910      542.698333   \n",
       "547bfa0c00ff82eb6f62d7a2b2cf4be7             4.384328      727.115000   \n",
       "a8c6dd0d4f1b33878dfc2f3559e88f87             2.657352     1150.276667   \n",
       "6d165923bdf6c3055311418ef905c58a             2.919019     1043.784000   \n",
       "42059a7ede026d409ff0f255635d7a08             6.412358      447.623000   \n",
       "0cb0b9356e0a85830bef4f4fda74f92b             8.298700      346.283846   \n",
       "34bff14a781e6eabd03f59ac8f610ed7             3.262338      945.807500   \n",
       "16134915d822fe17588ae585935e1e81            14.186154      199.011250   \n",
       "0425ae0d99a8e31c2758c25cca4ac2f7            11.316598      249.621818   \n",
       "9b306f73da56027310069b44074ef765             3.212192      932.238000   \n",
       "de70325370400abbce654bb35511812d             7.246892      386.129412   \n",
       "6dce8b4af8ca65025bc7f84457ae13a1             3.249562      891.911250   \n",
       "dc5cd67f8a03b032237af26b0c92c187             6.949175      401.204000   \n",
       "0994482188804e99710b7ee6d6a9e289             6.275799      458.485000   \n",
       "2e80c8b5d073f8310276f3ddfa2a1130             4.042078      702.280000   \n",
       "\n",
       "                                  predicted_avg_purchase  \\\n",
       "Id                                                         \n",
       "a5fadc51b1ae844ad9a70ad3cfa46a64              305.775264   \n",
       "5ac5ed64cd99ed2a8403b7a927e644ef              763.221185   \n",
       "60c19a709e3ced2d16d7100eb1069df5             1249.068736   \n",
       "dca76db00cc59dfbcdcc97c8bbc7f9f1              787.035039   \n",
       "ca2202a96c2de6ca6b8a37a4a73fa730              635.515389   \n",
       "a719d6643a7832535de9aded2f467825              389.079885   \n",
       "9cb19c3fc7311aae01cf16571b528001              510.602272   \n",
       "742d5a52d4df7cb14246d7f390de5d8a              478.014357   \n",
       "32df931a012f04da70265457d43e7da2             1138.027875   \n",
       "5f01420f0edda6555df5ce1cc62b986c              270.847452   \n",
       "30aa99d3357244cf38ca04eadef1473a              341.214766   \n",
       "a62a17bb46864da2c6da691d838971b3              299.956983   \n",
       "48a503edbaded96a3be27deee11967a1              203.209346   \n",
       "eba458987dc67827871c1d4d92e646e1             1004.418496   \n",
       "ed2b4332b3ca253cfbb0ffb54d3f5ae0              383.853919   \n",
       "a92534133444b5028d12a129b1b128f7              622.940549   \n",
       "66162981fc95e268e45bbfc738059687              763.934954   \n",
       "f11a653807d6bde53e4757f53e098c96              503.952892   \n",
       "2b75f007b50d6b21e1501f47dc8d632b              419.896988   \n",
       "98f8e41f45721cbe49a3147f6cf62432             1276.283110   \n",
       "2f486887c2edb2571d32c8cd15301711              850.935300   \n",
       "f09ff1c6c4ac8ea95d8621a94bb325fd              789.717714   \n",
       "4f672b7c5d7a4214b08ec6906163c980              406.954712   \n",
       "ede2a476c3894cf65d1619987d148422              651.183593   \n",
       "7863835a6120bfa52ccd665bfe03e9c8              439.399523   \n",
       "922bdaa119d44065b17b0447255646ae              486.453563   \n",
       "cd4cb9ec252a085ed4d2d3af7c18280a              883.268393   \n",
       "ca84dc4660f651a11a82041300367fae              691.412502   \n",
       "580a57e214a9d4df715accd2fe7065ba              404.555762   \n",
       "a63d0d4fe5bc662678bbbe6fbde1d900              872.324235   \n",
       "...                                                  ...   \n",
       "1091a17e1def85799828dadf9c28f3f2             1202.175329   \n",
       "ba5cf27a059fb5e6357dc3e7f70dbab2             1144.162132   \n",
       "5911b41fe8c5cd07a67571724f073e20              427.685606   \n",
       "e4d423cc47a17c9ca7cea0c12e8e22be             1110.510627   \n",
       "52cc702cd8c70995cf6ba1c762e341d4              404.121770   \n",
       "bfd2558c6ca88693b5721952422b5b40              529.766854   \n",
       "e0d591d2bbc656a278b5675d9bcae6c3              401.630367   \n",
       "79d35a58b719d48ca15d7a573f8a2379              389.274377   \n",
       "970c6dfdd3bcc297b26100a96a9a06fe              376.613848   \n",
       "2143bad4fa8805114a00a73853ac6ace              383.566025   \n",
       "6f9ebbe87978c734ba71c0032e1f3e45              596.960491   \n",
       "ac3537cd5ecca62eb010d5fb6f9c57d2              703.549362   \n",
       "16d851d4eb417ae73e48392063df8fb6              890.485572   \n",
       "68f0db4c3018522a7c24607851afd762              292.916097   \n",
       "4e7ef9579e693b03f57e5086ae7ebd6e              610.513840   \n",
       "ab30a46cc9f7d41f0ee382ff9dd48ad0              522.308704   \n",
       "547bfa0c00ff82eb6f62d7a2b2cf4be7              652.569085   \n",
       "a8c6dd0d4f1b33878dfc2f3559e88f87             1074.800349   \n",
       "6d165923bdf6c3055311418ef905c58a              977.143633   \n",
       "42059a7ede026d409ff0f255635d7a08              443.943681   \n",
       "0cb0b9356e0a85830bef4f4fda74f92b              342.918091   \n",
       "34bff14a781e6eabd03f59ac8f610ed7              869.892422   \n",
       "16134915d822fe17588ae585935e1e81              199.684684   \n",
       "0425ae0d99a8e31c2758c25cca4ac2f7              249.232239   \n",
       "9b306f73da56027310069b44074ef765              874.852871   \n",
       "de70325370400abbce654bb35511812d              382.188189   \n",
       "6dce8b4af8ca65025bc7f84457ae13a1              847.950446   \n",
       "dc5cd67f8a03b032237af26b0c92c187              394.996076   \n",
       "0994482188804e99710b7ee6d6a9e289              435.031385   \n",
       "2e80c8b5d073f8310276f3ddfa2a1130              671.118927   \n",
       "\n",
       "                                  predicted_total_spent  \n",
       "Id                                                       \n",
       "a5fadc51b1ae844ad9a70ad3cfa46a64           18566.177851  \n",
       "5ac5ed64cd99ed2a8403b7a927e644ef           11047.676642  \n",
       "60c19a709e3ced2d16d7100eb1069df5            9207.659881  \n",
       "dca76db00cc59dfbcdcc97c8bbc7f9f1            8883.067235  \n",
       "ca2202a96c2de6ca6b8a37a4a73fa730            8638.067559  \n",
       "a719d6643a7832535de9aded2f467825            6398.369439  \n",
       "9cb19c3fc7311aae01cf16571b528001            6355.582328  \n",
       "742d5a52d4df7cb14246d7f390de5d8a            6312.317481  \n",
       "32df931a012f04da70265457d43e7da2            6235.189612  \n",
       "5f01420f0edda6555df5ce1cc62b986c            6002.994703  \n",
       "30aa99d3357244cf38ca04eadef1473a            5865.625567  \n",
       "a62a17bb46864da2c6da691d838971b3            5828.715235  \n",
       "48a503edbaded96a3be27deee11967a1            5827.057904  \n",
       "eba458987dc67827871c1d4d92e646e1            5779.128040  \n",
       "ed2b4332b3ca253cfbb0ffb54d3f5ae0            5739.432987  \n",
       "a92534133444b5028d12a129b1b128f7            5598.984723  \n",
       "66162981fc95e268e45bbfc738059687            5463.496360  \n",
       "f11a653807d6bde53e4757f53e098c96            5442.255827  \n",
       "2b75f007b50d6b21e1501f47dc8d632b            5295.762728  \n",
       "98f8e41f45721cbe49a3147f6cf62432            5293.247450  \n",
       "2f486887c2edb2571d32c8cd15301711            5197.560536  \n",
       "f09ff1c6c4ac8ea95d8621a94bb325fd            5088.759759  \n",
       "4f672b7c5d7a4214b08ec6906163c980            5055.729472  \n",
       "ede2a476c3894cf65d1619987d148422            5041.301217  \n",
       "7863835a6120bfa52ccd665bfe03e9c8            4622.750048  \n",
       "922bdaa119d44065b17b0447255646ae            4536.163053  \n",
       "cd4cb9ec252a085ed4d2d3af7c18280a            4486.063482  \n",
       "ca84dc4660f651a11a82041300367fae            4403.387613  \n",
       "580a57e214a9d4df715accd2fe7065ba            4239.034879  \n",
       "a63d0d4fe5bc662678bbbe6fbde1d900            4178.054323  \n",
       "...                                                 ...  \n",
       "1091a17e1def85799828dadf9c28f3f2            3059.996252  \n",
       "ba5cf27a059fb5e6357dc3e7f70dbab2            3025.730066  \n",
       "5911b41fe8c5cd07a67571724f073e20            3021.243322  \n",
       "e4d423cc47a17c9ca7cea0c12e8e22be            2995.909406  \n",
       "52cc702cd8c70995cf6ba1c762e341d4            2989.986689  \n",
       "bfd2558c6ca88693b5721952422b5b40            2980.881555  \n",
       "e0d591d2bbc656a278b5675d9bcae6c3            2979.357842  \n",
       "79d35a58b719d48ca15d7a573f8a2379            2954.050742  \n",
       "970c6dfdd3bcc297b26100a96a9a06fe            2944.676469  \n",
       "2143bad4fa8805114a00a73853ac6ace            2939.088593  \n",
       "6f9ebbe87978c734ba71c0032e1f3e45            2928.389497  \n",
       "ac3537cd5ecca62eb010d5fb6f9c57d2            2917.549303  \n",
       "16d851d4eb417ae73e48392063df8fb6            2900.774054  \n",
       "68f0db4c3018522a7c24607851afd762            2885.172402  \n",
       "4e7ef9579e693b03f57e5086ae7ebd6e            2883.210769  \n",
       "ab30a46cc9f7d41f0ee382ff9dd48ad0            2866.905512  \n",
       "547bfa0c00ff82eb6f62d7a2b2cf4be7            2861.076962  \n",
       "a8c6dd0d4f1b33878dfc2f3559e88f87            2856.123041  \n",
       "6d165923bdf6c3055311418ef905c58a            2852.300370  \n",
       "42059a7ede026d409ff0f255635d7a08            2846.725929  \n",
       "0cb0b9356e0a85830bef4f4fda74f92b            2845.774252  \n",
       "34bff14a781e6eabd03f59ac8f610ed7            2837.883245  \n",
       "16134915d822fe17588ae585935e1e81            2832.757666  \n",
       "0425ae0d99a8e31c2758c25cca4ac2f7            2820.461107  \n",
       "9b306f73da56027310069b44074ef765            2810.195774  \n",
       "de70325370400abbce654bb35511812d            2769.676519  \n",
       "6dce8b4af8ca65025bc7f84457ae13a1            2755.467654  \n",
       "dc5cd67f8a03b032237af26b0c92c187            2744.896692  \n",
       "0994482188804e99710b7ee6d6a9e289            2730.169671  \n",
       "2e80c8b5d073f8310276f3ddfa2a1130            2712.714794  \n",
       "\n",
       "[100 rows x 7 columns]"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "returning_customers_summary.sort_values(by='predicted_total_spent',ascending=False).head(100)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "These are the top 100 customers in terms of projected total amount spent over the next year.  This list looks different than the list of top 100 customers by projected number of purchases.  That is because customers different customers have different average purchase values.  For example, a customer who makes 100 purchases at \\$1 each will spend less than a customer who only makes ten \\$20 purchases.  \n",
    "\n",
    "In order to make it easy for these projections to be referenced later, we can export them to csv files."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "summary.sort_values(by='predicted_purchases',ascending=False).head(100).to_csv('top_100_predicted_purchases.csv')\n",
    "returning_customers_summary.sort_values(by='predicted_total_spent',ascending=False).head(100).to_csv('top_100_predicted_total_spent.csv')"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python [conda root]",
   "language": "python",
   "name": "conda-root-py"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.5.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}