01 Data Exploration.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This is notebook is for exploring and understanding the datasets and preparing data for later machine learning models.\n",
    "\n",
    "> ### Key parts in this notebook\n",
    "\n",
    "### Analyze `Portfolio` data\n",
    "\n",
    "In this part we will take a look at the promotion offers. There are 10 different promotions that belong to 3 general types. There are 4 features that describe the promotions: channels, difficulty, duration and reward. We found out that discount promotions usually have long durations and high difficulties while buy-one-get-one promotions have shorter durations and lower difficulties. Discount promotions will have higher reward. \n",
    "\n",
    "### Analyze `Profile` data\n",
    "\n",
    "In this part we will take a look at customers' demographic data.\n",
    "\n",
    "`Age`: 12.8% of the data are missing. About half the customers are between 50 and 80 years old.\n",
    "<br>`Gender`: 1.4% of the data are missing. 57.2% of the customers are males.\n",
    "<br>`Customer age (how long the customer has been with us)`: 56% of the customers joined in the past 3 years.\n",
    "<br>`Income`: 12.8% of the data are missing. Income seems to be corrrelated with age. In this simulated dataset, different age buckets have different limits of maximum incomes. For people under 40 years old, their income won't exceed \n",
    "80k and for people under 50 years old, their income won't exceed 100k. For older people, the maximum income is about 120k. About 40% of the income numbers are between 50k and 80k.\n",
    "\n",
    "### Analyze `Transcript` data\n",
    "\n",
    "There are 4 types of transcripts: `normal transaction`, `offer received`, `offer viewed` and `offer complete`. \n",
    "\n",
    "In total we offered 76,277 promotions and there are 63,288 unique person + offer_type combinations. Only 6 people didn't revceive any offer.\n",
    "\n",
    "**Each person can receive the same type of offer for multiple times** and the offer id only tells what type of offer that is ---- it can't be used to recognize a unique offer. Our challenge here is to match the various actions to tell if a certain offer was viewed or completed. Also, customers might not receive all types of offers.\n",
    "\n",
    "**We solve this problem by giving an order number of the transactions for each customer. The offer id plus the rank can be seen as a \"unique ID\" for that offer.** We discovered that until the current offer expires, the app won't send another offer under the same type to that customer (we will prove this later). We will attribute the earlist following actions (view the offer or complete the offer) to the earliest offer (under the same offer id) with an extra condition that the following actions need to happen before that offer expires.\n",
    "\n",
    "*After matching offer_received with offer_viewed and offer_complete, we build out the funnel for each promotion sent:*\n",
    "\n",
    "offer_received \n",
    "<br>|__ offer viewed \n",
    "<br>&nbsp; &nbsp; &nbsp; |__ offer completed\n",
    "<br>&nbsp; &nbsp; &nbsp; |__ offer not completed\n",
    "<br>|__ offer not viewed\n",
    "<br>&nbsp; &nbsp; &nbsp; |__ offer completed\n",
    "<br>&nbsp; &nbsp; &nbsp; |__ offer not completed\n",
    "\n",
    "74.16% of the offers were viewed (while a small portion of them were viewed after they were completed); 12.86% of the offers were viewed without being completed. Among the viewed offers, 66.96% of the offers were viewed within 1 day after the offer were received. 13.54% of the offer were completed within one day after being viewed.\n",
    "\n",
    "In general, offers that are viewed are more likely to be completed, especially for discount promotions. Female customers, older (> 35) customers, wealthier customers and customers who joined earlier are more valuable."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 1. Load Data and Packages"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import math\n",
    "import json\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "from datetime import datetime\n",
    "\n",
    "# read in the json files\n",
    "portfolio = pd.read_json('portfolio.json', orient='records', lines=True)\n",
    "profile = pd.read_json('profile.json', orient='records', lines=True)\n",
    "transcript = pd.read_json('transcript.json', orient='records', lines=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 2. Understand the data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 2.1 Analyze `portfolio` data\n",
    "\n",
    "Portfolio dataset shows the information about the promos.\n",
    "\n",
    "Explanation of the data:\n",
    "\n",
    "**reward: (numeric)** money awarded for the amount spent\n",
    "<br>**channels: (list)** web, email, mobile, social\n",
    "<br>**difficulty: (numeric)** money required to be spent to receive reward\n",
    "<br>**duration: (numeric)** time for offer to be open, in days\n",
    "<br>**offer_type: (string)** bogo, discount, informational"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "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>channels</th>\n",
       "      <th>difficulty</th>\n",
       "      <th>duration</th>\n",
       "      <th>id</th>\n",
       "      <th>offer_type</th>\n",
       "      <th>reward</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>[email, mobile, social]</td>\n",
       "      <td>10</td>\n",
       "      <td>7</td>\n",
       "      <td>ae264e3637204a6fb9bb56bc8210ddfd</td>\n",
       "      <td>bogo</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>[web, email, mobile, social]</td>\n",
       "      <td>10</td>\n",
       "      <td>5</td>\n",
       "      <td>4d5c57ea9a6940dd891ad53e9dbe8da0</td>\n",
       "      <td>bogo</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>[web, email, mobile]</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>3f207df678b143eea3cee63160fa8bed</td>\n",
       "      <td>informational</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>[web, email, mobile]</td>\n",
       "      <td>5</td>\n",
       "      <td>7</td>\n",
       "      <td>9b98b8c7a33c4b65b9aebfe6a799e6d9</td>\n",
       "      <td>bogo</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>[web, email]</td>\n",
       "      <td>20</td>\n",
       "      <td>10</td>\n",
       "      <td>0b1e1539f2cc45b7b9fa7c272da2e1d7</td>\n",
       "      <td>discount</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>[web, email, mobile, social]</td>\n",
       "      <td>7</td>\n",
       "      <td>7</td>\n",
       "      <td>2298d6c36e964ae4a3e7e9706d1fb8c2</td>\n",
       "      <td>discount</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>[web, email, mobile, social]</td>\n",
       "      <td>10</td>\n",
       "      <td>10</td>\n",
       "      <td>fafdcd668e3743c1bb461111dcafc2a4</td>\n",
       "      <td>discount</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>[email, mobile, social]</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>5a8bc65990b245e5a138643cd4eb9837</td>\n",
       "      <td>informational</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>[web, email, mobile, social]</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>f19421c1d4aa40978ebb69ca19b0e20d</td>\n",
       "      <td>bogo</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>[web, email, mobile]</td>\n",
       "      <td>10</td>\n",
       "      <td>7</td>\n",
       "      <td>2906b810c7d4411798c6938adc9daaa5</td>\n",
       "      <td>discount</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                       channels  difficulty  duration  \\\n",
       "0       [email, mobile, social]          10         7   \n",
       "1  [web, email, mobile, social]          10         5   \n",
       "2          [web, email, mobile]           0         4   \n",
       "3          [web, email, mobile]           5         7   \n",
       "4                  [web, email]          20        10   \n",
       "5  [web, email, mobile, social]           7         7   \n",
       "6  [web, email, mobile, social]          10        10   \n",
       "7       [email, mobile, social]           0         3   \n",
       "8  [web, email, mobile, social]           5         5   \n",
       "9          [web, email, mobile]          10         7   \n",
       "\n",
       "                                 id     offer_type  reward  \n",
       "0  ae264e3637204a6fb9bb56bc8210ddfd           bogo      10  \n",
       "1  4d5c57ea9a6940dd891ad53e9dbe8da0           bogo      10  \n",
       "2  3f207df678b143eea3cee63160fa8bed  informational       0  \n",
       "3  9b98b8c7a33c4b65b9aebfe6a799e6d9           bogo       5  \n",
       "4  0b1e1539f2cc45b7b9fa7c272da2e1d7       discount       5  \n",
       "5  2298d6c36e964ae4a3e7e9706d1fb8c2       discount       3  \n",
       "6  fafdcd668e3743c1bb461111dcafc2a4       discount       2  \n",
       "7  5a8bc65990b245e5a138643cd4eb9837  informational       0  \n",
       "8  f19421c1d4aa40978ebb69ca19b0e20d           bogo       5  \n",
       "9  2906b810c7d4411798c6938adc9daaa5       discount       2  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "portfolio"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(figsize=(10,5))\n",
    "ax.scatter(portfolio['difficulty'], portfolio['duration'])\n",
    "plt.xlabel('difficulty')\n",
    "plt.ylabel('duration')\n",
    "plt.title(\"Promos\")\n",
    "\n",
    "for i, txt in enumerate(portfolio['offer_type']):\n",
    "    ax.annotate(txt, (portfolio['difficulty'][i], portfolio['duration'][i]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Discount promotions usually have long duration and high difficulties while buy-one-get-one promotions have shorter duration and lower difficulties. Discount promotions will have higher reward. In general, longer promotions will have higher difficulties."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "portfolio['web'] = portfolio['channels'].apply(lambda x: 1 if 'web' in x else 0)\n",
    "portfolio['mobile'] = portfolio['channels'].apply(lambda x: 1 if 'mobile' in x else 0)\n",
    "portfolio['email'] = portfolio['channels'].apply(lambda x: 1 if 'email' in x else 0)\n",
    "portfolio['social'] = portfolio['channels'].apply(lambda x: 1 if 'social' in x else 0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "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>channels</th>\n",
       "      <th>difficulty</th>\n",
       "      <th>duration</th>\n",
       "      <th>id</th>\n",
       "      <th>offer_type</th>\n",
       "      <th>reward</th>\n",
       "      <th>web</th>\n",
       "      <th>mobile</th>\n",
       "      <th>email</th>\n",
       "      <th>social</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>[email, mobile, social]</td>\n",
       "      <td>10</td>\n",
       "      <td>7</td>\n",
       "      <td>ae264e3637204a6fb9bb56bc8210ddfd</td>\n",
       "      <td>bogo</td>\n",
       "      <td>10</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>[web, email, mobile, social]</td>\n",
       "      <td>10</td>\n",
       "      <td>5</td>\n",
       "      <td>4d5c57ea9a6940dd891ad53e9dbe8da0</td>\n",
       "      <td>bogo</td>\n",
       "      <td>10</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>[web, email, mobile]</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>3f207df678b143eea3cee63160fa8bed</td>\n",
       "      <td>informational</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>[web, email, mobile]</td>\n",
       "      <td>5</td>\n",
       "      <td>7</td>\n",
       "      <td>9b98b8c7a33c4b65b9aebfe6a799e6d9</td>\n",
       "      <td>bogo</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>[web, email]</td>\n",
       "      <td>20</td>\n",
       "      <td>10</td>\n",
       "      <td>0b1e1539f2cc45b7b9fa7c272da2e1d7</td>\n",
       "      <td>discount</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>[web, email, mobile, social]</td>\n",
       "      <td>7</td>\n",
       "      <td>7</td>\n",
       "      <td>2298d6c36e964ae4a3e7e9706d1fb8c2</td>\n",
       "      <td>discount</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>[web, email, mobile, social]</td>\n",
       "      <td>10</td>\n",
       "      <td>10</td>\n",
       "      <td>fafdcd668e3743c1bb461111dcafc2a4</td>\n",
       "      <td>discount</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>[email, mobile, social]</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>5a8bc65990b245e5a138643cd4eb9837</td>\n",
       "      <td>informational</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>[web, email, mobile, social]</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>f19421c1d4aa40978ebb69ca19b0e20d</td>\n",
       "      <td>bogo</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>[web, email, mobile]</td>\n",
       "      <td>10</td>\n",
       "      <td>7</td>\n",
       "      <td>2906b810c7d4411798c6938adc9daaa5</td>\n",
       "      <td>discount</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                       channels  difficulty  duration  \\\n",
       "0       [email, mobile, social]          10         7   \n",
       "1  [web, email, mobile, social]          10         5   \n",
       "2          [web, email, mobile]           0         4   \n",
       "3          [web, email, mobile]           5         7   \n",
       "4                  [web, email]          20        10   \n",
       "5  [web, email, mobile, social]           7         7   \n",
       "6  [web, email, mobile, social]          10        10   \n",
       "7       [email, mobile, social]           0         3   \n",
       "8  [web, email, mobile, social]           5         5   \n",
       "9          [web, email, mobile]          10         7   \n",
       "\n",
       "                                 id     offer_type  reward  web  mobile  \\\n",
       "0  ae264e3637204a6fb9bb56bc8210ddfd           bogo      10    0       1   \n",
       "1  4d5c57ea9a6940dd891ad53e9dbe8da0           bogo      10    1       1   \n",
       "2  3f207df678b143eea3cee63160fa8bed  informational       0    1       1   \n",
       "3  9b98b8c7a33c4b65b9aebfe6a799e6d9           bogo       5    1       1   \n",
       "4  0b1e1539f2cc45b7b9fa7c272da2e1d7       discount       5    1       0   \n",
       "5  2298d6c36e964ae4a3e7e9706d1fb8c2       discount       3    1       1   \n",
       "6  fafdcd668e3743c1bb461111dcafc2a4       discount       2    1       1   \n",
       "7  5a8bc65990b245e5a138643cd4eb9837  informational       0    0       1   \n",
       "8  f19421c1d4aa40978ebb69ca19b0e20d           bogo       5    1       1   \n",
       "9  2906b810c7d4411798c6938adc9daaa5       discount       2    1       1   \n",
       "\n",
       "   email  social  \n",
       "0      1       1  \n",
       "1      1       1  \n",
       "2      1       0  \n",
       "3      1       0  \n",
       "4      1       0  \n",
       "5      1       1  \n",
       "6      1       1  \n",
       "7      1       1  \n",
       "8      1       1  \n",
       "9      1       0  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "portfolio"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 2.2 Analyze `profile` data\n",
    "\n",
    "Profile data has customers' demographic information.\n",
    "\n",
    "**gender: (categorical)** M, F, O, or null\n",
    "<br>**age: (numeric)** missing value encoded as 118\n",
    "<br>**id**: (string/hash)\n",
    "<br>**became_member_on**: (date) format YYYYMMDD\n",
    "<br>**income**: (numeric)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(17000, 5)"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "profile.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "scrolled": true
   },
   "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>age</th>\n",
       "      <th>became_member_on</th>\n",
       "      <th>gender</th>\n",
       "      <th>id</th>\n",
       "      <th>income</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>118</td>\n",
       "      <td>20170212</td>\n",
       "      <td>None</td>\n",
       "      <td>68be06ca386d4c31939f3a4f0e3dd783</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>55</td>\n",
       "      <td>20170715</td>\n",
       "      <td>F</td>\n",
       "      <td>0610b486422d4921ae7d2bf64640c50b</td>\n",
       "      <td>112000.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>118</td>\n",
       "      <td>20180712</td>\n",
       "      <td>None</td>\n",
       "      <td>38fe809add3b4fcf9315a9694bb96ff5</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>75</td>\n",
       "      <td>20170509</td>\n",
       "      <td>F</td>\n",
       "      <td>78afa995795e4d85b5d9ceeca43f5fef</td>\n",
       "      <td>100000.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>118</td>\n",
       "      <td>20170804</td>\n",
       "      <td>None</td>\n",
       "      <td>a03223e636434f42ac4c3df47e8bac43</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   age  became_member_on gender                                id    income\n",
       "0  118          20170212   None  68be06ca386d4c31939f3a4f0e3dd783       NaN\n",
       "1   55          20170715      F  0610b486422d4921ae7d2bf64640c50b  112000.0\n",
       "2  118          20180712   None  38fe809add3b4fcf9315a9694bb96ff5       NaN\n",
       "3   75          20170509      F  78afa995795e4d85b5d9ceeca43f5fef  100000.0\n",
       "4  118          20170804   None  a03223e636434f42ac4c3df47e8bac43       NaN"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "profile.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2.2.1 Age"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Age distribution')"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.hist(profile['age'])\n",
    "plt.title(\"Age distribution\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Some people have wierd age numbers. We need to consider those values as missing values."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.12823529411764706"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sum(profile['age'] > 100) / len(profile)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We will now create age buckets for future visulizations and analysis."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [],
   "source": [
    "def bucket_age(x):\n",
    "    x = float(x)\n",
    "    if x < 25 :\n",
    "        return \"0 ~ 24\"\n",
    "    elif x >= 25 and x <= 34:\n",
    "        return \"25 ~ 34\"\n",
    "    elif x >= 35 and x <= 44:\n",
    "        return \"35 ~ 44\"\n",
    "    elif x >= 45 and x <= 54:\n",
    "        return \"45 ~ 54\"\n",
    "    elif x >=55 and  x <= 64:\n",
    "        return \"55 ~ 64\"\n",
    "    elif x is None:\n",
    "        return float('nan')\n",
    "    else:\n",
    "        return \"65 ~\"\n",
    "\n",
    "profile.loc[profile['age'] > 100, 'age'] = float('nan')\n",
    "profile['age_bucket'] = profile['age'].apply(bucket_age)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "65 ~       0.378882\n",
       "55 ~ 64    0.201235\n",
       "45 ~ 54    0.177235\n",
       "35 ~ 44    0.109941\n",
       "25 ~ 34    0.081176\n",
       "0 ~ 24     0.051529\n",
       "Name: age_bucket, dtype: float64"
      ]
     },
     "execution_count": 78,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "profile['age_bucket'].value_counts(normalize=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.4958823529411765"
      ]
     },
     "execution_count": 79,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sum((profile['age']>=50) & (profile['age']<=80)) / len(profile)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2.2.2 Gender"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "M    0.572277\n",
       "F    0.413423\n",
       "O    0.014300\n",
       "Name: gender, dtype: float64"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "profile['gender'].value_counts(normalize=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "1.4% of the values are missing."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2.2.3 Customer age (how many months since the customer first joined)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "now = pd.to_datetime(datetime.strftime(datetime.now(), '%Y-%m-%d'), format=\"%Y/%m/%d\")\n",
    "\n",
    "def format_date(x):\n",
    "    x = str(x)\n",
    "    that_day = pd.to_datetime(x, format=\"%Y%m%d\")\n",
    "    return (now - that_day).days/30"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "profile['customer_age'] = profile['became_member_on'].apply(format_date)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 0, 'months')"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYAAAAEWCAYAAABv+EDhAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvOIA7rQAAGXZJREFUeJzt3X2wZVV95vHvI43iC9i8NIjdrY3aUcFSJB0kwYoGDC+iQiwZURN7LEzPVLAiiZkEranBNwxmHHXMi4YBIhoV0aigGLELZJKYEW0EUUCKFondNqEbGvAtoMBv/tjryqG9t++93Zf74vp+qm6dvddeZ++1zjn3PGevvc8+qSokSf152Fw3QJI0NwwASeqUASBJnTIAJKlTBoAkdcoAkKROGQDSHElyeZLXtulXJfniDK772iTPb9NvTvL3M7juNyU5e6bWp7ljAIgkr0yyLsmPktyS5B+TPHcn1zmjbzq/7KrqI1V11GT1knwwydunsL6DqurynW1Xkucn2bjNut9RVa/d2XVr7hkAnUvyx8B7gXcA+wFPAP4GOH4u2zWTkiya6zbMlp76qhlQVf51+gc8FvgRcOJ26nwQePvI/POBjSPzfwZ8H/ghcANwJHAM8FPgZ23932h1Hw9cBGwF1gO/P7KeNwOfAP6+reubwK8AbwQ2AxuAo7Zp+znALW37bwd2acv+M/Bl4D1tW2/fTv+eDFwG3A7cBnwEWDyy/BDgqtamTwAf3+bxeBFwNXAn8K/AM7ezrd8Gvg3cBfwV8H+B1460+V/adFrbN7e61wDPANa0x/Sn7XH9bKt/c3sergHuARa1sheMPLafbG3/IfB14Fkj7SrgKds+58Cjgf8A7m/b+1F7Dt8M/P1I/ZcA17bH4HLg6SPLbgb+pLXtrtaG3eb6te/f8OceQN9+HdgN+PSO3DnJU4HXAb9WVbsDRwM3V9UXGPYoPl5Vj6mqZ7W7fAzYyPAm8jLgHUmOHFnli4EPA3syvOlewrCXuhR4K/C3I3XPA+4FngI8GzgKGB2WeA5wE7AvcMb2ugH8eWvT04HlDG9wJHk4w2PzQWCv1v7fGen/IcC5wH8B9m7tuyjJI8Z5rPYB/gH478A+wHeAwydo01HAbzIE4GLg5cDtVXUWQ0D9RXtcXzxyn1cAxzGE173jrPN4hgDbC/go8Jkku078sEBV/Rg4FtjUtveYqtq0Tb9+heFxORVYAnwe+Gx77Mb8J4YPBQcAz2QIO80DBkDf9gZum+ANYyruAx4BHJhk16q6uaq+M17FJMuB5wJ/VlV3V9XVwNnA741U++equqS15xMMbyhnVtXPgPOBFUkWJ9mP4Y3p1Kr6cVVtZvjEfNLIujZV1V9W1b1V9R8TdaCq1lfV2qq6p6q2AO8GntcWH8bwafp9VfWzqvoU8NWRu/8+8LdVdUVV3VdV5zF8Aj9snE29ELiuqj7Z+vNe4N8naNbPgN2BpwGpquur6paJ+tC8r6o2bKevV45s+90MwT9eO6fr5cDF7TH8GfAu4JHAb2zTtk1VtRX4LHDwDGxXM8AA6NvtwD47Om5cVesZPvm9Gdic5Pwkj5+g+uOBrVX1w5Gyf2P4dD/m1pHp/2AIp/tG5gEeAzwR2BW4JcmdSe5k+PS978j9N0ylD0n2be3+fpIfMAxB7TPS5u9X1egVE0fX+0TgDWNtaO1Y3u63rceP3retc9w2VtVlDENEfw3cmuSsJHtM0pXJ+ju67ft5YE9sZz2e4XkcXfcGHvy8jgbdTxieQ80DBkDf/h9wN3DCdur8GHjUyPzjRhdW1Uer6rkMb4YFvHNs0Tbr2QTslWT3kbInMIzfT9cGhk/a+1TV4va3R1UdNNq0Ka7rz1vdZ1bVHsDvMgwLwXB8YWmSjNRfvk07zhhpw+KqelRVfWyc7dwyet+2zuXj1BsaX/W+qvpV4CCGoaD/Nkm/Juvv6LYfBixjeE5geFOe6DmebL2bGJ77sXWP9WtHnlfNMgOgY1V1F/A/gL9OckKSRyXZNcmxSf6iVbsaeGGSvZI8juETPzAcA0hyRBvzvpvhU/rYJ/ZbGYZsHta2tYHhIOmfJ9ktyTOBkxnGtKfb7luALwL/K8keSR6W5MlJnjfZfcexO8PBzTuTLOWBN1oYAvI+4HVJFiU5Hjh0ZPn/Af5rkudk8Ogkx20TcmMuBg5K8tK2x/WHbBOmY5L8WlvnrgwBfDcPflyftAP9/NWRbZ/KEKBfacuuBl6ZZJckx/DAENjY9vZO8tgJ1nsBcFySI1t739DW/a870EbNMgOgc1X1buCPGQ5ObmH4VPs64DOtyoeBbzCczfFFhrM4xjwCOJPh7Jl/ZxiCeVNb9ol2e3uSr7fpVwArGD41fho4varW7mDTXw08HLgOuIPhLJf9d2A9b2E40+cuhjfpT40tqKqfAi9lCKo7GfYOPsfwBkdVrWM4DvBXrQ3rmeAAZ1XdBpzI8HjdDqxkOFNpPHswhMsdDMMrtzOMrcNw5tOBbcjpMxPcfzwXMozX38Fw3OWlbcwe4PUMB+DvBF7FA889VfVthoO8N7VtPmjYqKpuYHhc/pLhdfBi4MXtsdM8lwcPb0raniRXAB+oqr+b67ZIO8s9AGk7kjwvyePaENBqhtMYvzDX7ZJmggGgX3pJPtAuc7Ht3wemcPenMgyB3cUwvv2yKZySKS0IDgFJUqfcA5CkTs3rC0fts88+tWLFirluhiQtKFdeeeVtVbVksnrzOgBWrFjBunXr5roZkrSgJPm3yWs5BCRJ3TIAJKlTBoAkdcoAkKROGQCS1CkDQJI6ZQBIUqcMAEnqlAEgSZ2a198EXqhWnHbxnGz35jOPm5PtSlqY3AOQpE5NKQCS3Jzkm0muTrKule2VZG2SG9vtnq08Sd6XZH2Sa5IcMrKe1a3+je3HNSRJc2Q6ewC/VVUHV9WqNn8acGlVrQQubfMAxzL83ulKYA3wfhgCAzgdeA7DD2ufPhYakqTZtzNDQMcD57Xp84ATRso/VIOvAIuT7A8cDaytqq1VdQewFjhmJ7YvSdoJUw2AAr6Y5Moka1rZfmM/jddu923lS4ENI/fd2MomKn+QJGuSrEuybsuWLVPviSRpWqZ6FtDhVbUpyb7A2iTf3k7djFNW2yl/cEHVWcBZAKtWrfL3KiXpITKlPYCq2tRuNwOfZhjDv7UN7dBuN7fqG4HlI3dfBmzaTrkkaQ5MGgBJHp1k97Fp4CjgW8BFwNiZPKuBC9v0RcCr29lAhwF3tSGiS4CjkuzZDv4e1cokSXNgKkNA+wGfTjJW/6NV9YUkXwMuSHIy8D3gxFb/88ALgfXAT4DXAFTV1iRvA77W6r21qrbOWE8kSdMyaQBU1U3As8Ypvx04cpzyAk6ZYF3nAudOv5mSpJnmN4ElqVMGgCR1ygCQpE4ZAJLUKQNAkjplAEhSpwwASeqUASBJnTIAJKlTBoAkdcoAkKROGQCS1CkDQJI6ZQBIUqcMAEnqlAEgSZ0yACSpUwaAJHXKAJCkThkAktQpA0CSOmUASFKnDABJ6pQBIEmdMgAkqVMGgCR1ygCQpE4ZAJLUKQNAkjplAEhSpwwASerUlAMgyS5JrkryuTZ/QJIrktyY5ONJHt7KH9Hm17flK0bW8cZWfkOSo2e6M5KkqZvOHsDrgetH5t8JvKeqVgJ3ACe38pOBO6rqKcB7Wj2SHAicBBwEHAP8TZJddq75kqQdNaUASLIMOA44u80HOAL4ZKtyHnBCmz6+zdOWH9nqHw+cX1X3VNV3gfXAoTPRCUnS9E11D+C9wJ8C97f5vYE7q+reNr8RWNqmlwIbANryu1r9n5ePc5+fS7Imybok67Zs2TKNrkiSpmPSAEjyImBzVV05WjxO1Zpk2fbu80BB1VlVtaqqVi1ZsmSy5kmSdtCiKdQ5HHhJkhcCuwF7MOwRLE6yqH3KXwZsavU3AsuBjUkWAY8Fto6Ujxm9jyRplk26B1BVb6yqZVW1guEg7mVV9SrgS8DLWrXVwIVt+qI2T1t+WVVVKz+pnSV0ALAS+OqM9USSNC1T2QOYyJ8B5yd5O3AVcE4rPwf4cJL1DJ/8TwKoqmuTXABcB9wLnFJV9+3E9iVJO2FaAVBVlwOXt+mbGOcsnqq6GzhxgvufAZwx3UZKkmae3wSWpE4ZAJLUKQNAkjplAEhSpwwASeqUASBJnTIAJKlTBoAkdcoAkKROGQCS1CkDQJI6ZQBIUqd25mqg896K0y6e6yZI0rzlHoAkdcoAkKROGQCS1CkDQJI6ZQBIUqcMAEnqlAEgSZ0yACSpUwaAJHXKAJCkThkAktQpA0CSOmUASFKnDABJ6pQBIEmdMgAkqVMGgCR1ygCQpE4ZAJLUqUkDIMluSb6a5BtJrk3yllZ+QJIrktyY5ONJHt7KH9Hm17flK0bW9cZWfkOSox+qTkmSJjeVPYB7gCOq6lnAwcAxSQ4D3gm8p6pWAncAJ7f6JwN3VNVTgPe0eiQ5EDgJOAg4BvibJLvMZGckSVM3aQDU4Edtdtf2V8ARwCdb+XnACW36+DZPW35kkrTy86vqnqr6LrAeOHRGeiFJmrYpHQNIskuSq4HNwFrgO8CdVXVvq7IRWNqmlwIbANryu4C9R8vHuc/ottYkWZdk3ZYtW6bfI0nSlEwpAKrqvqo6GFjG8Kn96eNVa7eZYNlE5dtu66yqWlVVq5YsWTKV5kmSdsC0zgKqqjuBy4HDgMVJFrVFy4BNbXojsBygLX8ssHW0fJz7SJJm2VTOAlqSZHGbfiTwAuB64EvAy1q11cCFbfqiNk9bfllVVSs/qZ0ldACwEvjqTHVEkjQ9iyavwv7Aee2MnYcBF1TV55JcB5yf5O3AVcA5rf45wIeTrGf45H8SQFVdm+QC4DrgXuCUqrpvZrsjSZqqSQOgqq4Bnj1O+U2McxZPVd0NnDjBus4Azph+MyVJM81vAktSpwwASeqUASBJnTIAJKlTBoAkdcoAkKROGQCS1CkDQJI6ZQBIUqcMAEnqlAEgSZ0yACSpUwaAJHXKAJCkThkAktQpA0CSOmUASFKnDABJ6pQBIEmdMgAkqVMGgCR1ygCQpE4ZAJLUKQNAkjplAEhSpwwASeqUASBJnTIAJKlTBoAkdcoAkKROGQCS1KlJAyDJ8iRfSnJ9kmuTvL6V75VkbZIb2+2erTxJ3pdkfZJrkhwysq7Vrf6NSVY/dN2SJE1mKnsA9wJvqKqnA4cBpyQ5EDgNuLSqVgKXtnmAY4GV7W8N8H4YAgM4HXgOcChw+lhoSJJm36QBUFW3VNXX2/QPgeuBpcDxwHmt2nnACW36eOBDNfgKsDjJ/sDRwNqq2lpVdwBrgWNmtDeSpCmb1jGAJCuAZwNXAPtV1S0whASwb6u2FNgwcreNrWyi8m23sSbJuiTrtmzZMp3mSZKmYcoBkOQxwD8Ap1bVD7ZXdZyy2k75gwuqzqqqVVW1asmSJVNtniRpmqYUAEl2ZXjz/0hVfaoV39qGdmi3m1v5RmD5yN2XAZu2Uy5JmgNTOQsowDnA9VX17pFFFwFjZ/KsBi4cKX91OxvoMOCuNkR0CXBUkj3bwd+jWpkkaQ4smkKdw4HfA76Z5OpW9ibgTOCCJCcD3wNObMs+D7wQWA/8BHgNQFVtTfI24Gut3lurauuM9EKSNG2TBkBV/Qvjj98DHDlO/QJOmWBd5wLnTqeBkqSHht8ElqROGQCS1KmpHAOQ5q0Vp108Z9u++czj5mzb0kxwD0CSOuUewC8RPw1Lmg73ACSpU+4BaEbM5d6HpB3jHoAkdco9AGkHzdVej8dbNFPcA5CkThkAktQpA0CSOmUASFKnDABJ6pQBIEmdMgAkqVMGgCR1ygCQpE4ZAJLUKQNAkjplAEhSpwwASeqUASBJnTIAJKlTBoAkdcoAkKROGQCS1CkDQJI6ZQBIUqcMAEnqlAEgSZ2aNACSnJtkc5JvjZTtlWRtkhvb7Z6tPEnel2R9kmuSHDJyn9Wt/o1JVj803ZEkTdVU9gA+CByzTdlpwKVVtRK4tM0DHAusbH9rgPfDEBjA6cBzgEOB08dCQ5I0NyYNgKr6J2DrNsXHA+e16fOAE0bKP1SDrwCLk+wPHA2sraqtVXUHsJZfDBVJ0iza0WMA+1XVLQDtdt9WvhTYMFJvYyubqPwXJFmTZF2SdVu2bNnB5kmSJjPTB4EzTlltp/wXC6vOqqpVVbVqyZIlM9o4SdIDdjQAbm1DO7Tbza18I7B8pN4yYNN2yiVJc2RHA+AiYOxMntXAhSPlr25nAx0G3NWGiC4BjkqyZzv4e1QrkyTNkUWTVUjyMeD5wD5JNjKczXMmcEGSk4HvASe26p8HXgisB34CvAagqrYmeRvwtVbvrVW17YFlSdIsmjQAquoVEyw6cpy6BZwywXrOBc6dVuskSQ8ZvwksSZ0yACSpUwaAJHXKAJCkThkAktQpA0CSOmUASFKnDABJ6pQBIEmdMgAkqVMGgCR1ygCQpE4ZAJLUKQNAkjplAEhSpwwASeqUASBJnTIAJKlTBoAkdcoAkKROTfqj8JLmlxWnXTxn2775zOPmZLtz1ee56u9sMQAkaQK/7GFrAEiasrl8Q9TM8xiAJHXKAJCkThkAktQpA0CSOmUASFKnDABJ6pQBIEmdMgAkqVMGgCR1atYDIMkxSW5Isj7JabO9fUnSYFYDIMkuwF8DxwIHAq9IcuBstkGSNJjtPYBDgfVVdVNV/RQ4Hzh+ltsgSWL2Lwa3FNgwMr8ReM5ohSRrgDVt9kdJbphgXfsAt814C2ef/Zhf7Mf80m0/8s6d2t4Tp1JptgMg45TVg2aqzgLOmnRFybqqWjVTDZsr9mN+sR/zi/14aM32ENBGYPnI/DJg0yy3QZLE7AfA14CVSQ5I8nDgJOCiWW6DJIlZHgKqqnuTvA64BNgFOLeqrt3B1U06TLRA2I/5xX7ML/bjIZSqmryWJOmXjt8ElqROGQCS1Kl5HwBJlif5UpLrk1yb5PWtfK8ka5Pc2G73nOu2bk+S3ZJ8Nck3Wj/e0soPSHJF68fH28HxeS/JLkmuSvK5Nr9Q+3Fzkm8muTrJula2oF5bAEkWJ/lkkm+3/5VfX2j9SPLU9jyM/f0gyakLrR8ASf6o/Z9/K8nH2v//vPsfmfcBANwLvKGqng4cBpzSLh9xGnBpVa0ELm3z89k9wBFV9SzgYOCYJIcB7wTe0/pxB3DyHLZxOl4PXD8yv1D7AfBbVXXwyHnaC+21BfC/gS9U1dOAZzE8NwuqH1V1Q3seDgZ+FfgJ8GkWWD+SLAX+EFhVVc9gOOHlJObj/0hVLag/4ELgt4EbgP1b2f7ADXPdtmn04VHA1xm+BX0bsKiV/zpwyVy3bwrtX8bwj3gE8DmGL/gtuH60tt4M7LNN2YJ6bQF7AN+lndSxUPuxTduPAr68EPvBA1c82IvhTMvPAUfPx/+RhbAH8HNJVgDPBq4A9quqWwDa7b5z17KpacMmVwObgbXAd4A7q+reVmUjw4tnvnsv8KfA/W1+bxZmP2D4JvoXk1zZLkMCC++19SRgC/B3bVju7CSPZuH1Y9RJwMfa9ILqR1V9H3gX8D3gFuAu4Erm4f/IggmAJI8B/gE4tap+MNft2RFVdV8Nu7fLGC6M9/Txqs1uq6YnyYuAzVV15WjxOFXndT9GHF5VhzBcofaUJL851w3aAYuAQ4D3V9WzgR8zz4dJtqeNjb8E+MRct2VHtGMUxwMHAI8HHs3w+trWnP+PLIgASLIrw5v/R6rqU6341iT7t+X7M3yqXhCq6k7gcoZjGouTjH0hbyFcGuNw4CVJbma4musRDHsEC60fAFTVpna7mWG8+VAW3mtrI7Cxqq5o859kCISF1o8xxwJfr6pb2/xC68cLgO9W1Zaq+hnwKeA3mIf/I/M+AJIEOAe4vqrePbLoImB1m17NcGxg3kqyJMniNv1IhhfJ9cCXgJe1avO+H1X1xqpaVlUrGHbTL6uqV7HA+gGQ5NFJdh+bZhh3/hYL7LVVVf8ObEjy1FZ0JHAdC6wfI17BA8M/sPD68T3gsCSPau9fY8/HvPsfmfffBE7yXOCfgW/ywJjzmxiOA1wAPIHhAT+xqrbOSSOnIMkzgfMYzgh4GHBBVb01yZMYPknvBVwF/G5V3TN3LZ26JM8H/qSqXrQQ+9Ha/Ok2uwj4aFWdkWRvFtBrCyDJwcDZwMOBm4DX0F5nLKx+PIrhAOqTququVrYQn4+3AC9nOIvxKuC1DGP+8+p/ZN4HgCTpoTHvh4AkSQ8NA0CSOmUASFKnDABJ6pQBIEmdMgCkGdCuxvkHI/PPH7tSqjRfGQDSzFgM/MGktaR5xABQd5KsaNfNP7tdr/0jSV6Q5MvtWu2HtmvQfybJNUm+0r7IR5I3Jzk3yeVJbkryh221ZwJPbtex/5+t7DEj1+j/SPtWKEnOTHJdW/e75uAhkIBZ/lF4aR55CnAisAb4GvBK4LkMFyF7E8O3Ua+qqhOSHAF8iOF3HACeBvwWsDtwQ5L3M1x87RntYn9j35J+NnAQwzVfvgwcnuQ64HeAp1VVjV0eRJoL7gGoV9+tqm9W1f3AtQw/OFIMlxxZwRAGHwaoqsuAvZM8tt334qq6p6puY7gw2X4TbOOrVbWxbePqtt4fAHcDZyd5KcOPnkhzwgBQr0avwXL/yPz9DHvG27vE9eh972PiPelfqNeuB38ow9VtTwC+ML1mSzPHAJDG90/Aq+Dnwzm3TfI7FD9kGBLarva7Fo+tqs8Dp/LAsJI06zwGII3vzQy/sHUNwzDN6u1Vrqrb20HkbwH/CFw8QdXdgQuT7Mawl/FHM9dkaXq8GqgkdcohIEnqlAEgSZ0yACSpUwaAJHXKAJCkThkAktQpA0CSOvX/AanW4+emzz+LAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.hist(profile['customer_age'])\n",
    "plt.title(\"Customer_age distribution\")\n",
    "plt.xlabel('months')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.5477647058823529"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sum(profile['customer_age'] <= 36) / len(profile)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The \"oldest\" customer joined about 6 years ago! "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2.2.4 Income"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/tianyiwang/anaconda3/lib/python3.7/site-packages/numpy/lib/histograms.py:754: RuntimeWarning: invalid value encountered in greater_equal\n",
      "  keep = (tmp_a >= first_edge)\n",
      "/Users/tianyiwang/anaconda3/lib/python3.7/site-packages/numpy/lib/histograms.py:755: RuntimeWarning: invalid value encountered in less_equal\n",
      "  keep &= (tmp_a <= last_edge)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Income')"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.hist(profile['income'])\n",
    "plt.title('Income')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.12794117647058822"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "profile['income'].isna().sum() / len(profile)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.41611764705882354"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sum((profile['income']>50000) & (profile['income']<80000))/len(profile)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Relationship between age and income')"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.scatter(profile['age'], profile['income'], c=profile['customer_age'])\n",
    "plt.xlabel('age')\n",
    "plt.ylabel('income')\n",
    "plt.title(\"Relationship between age and income\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 2.3 Analyze `Transcript` data\n",
    "\n",
    "Transcript table records customers' transactions.\n",
    "\n",
    "Explanation on the data:\n",
    "\n",
    "**person**: (string/hash)\n",
    "<br>**event**: (string) offer received, offer viewed, transaction, offer completed\n",
    "<br>**value**: (dictionary) different values depending on event type\n",
    "<br>**offer id**: (string/hash) not associated with any \"transaction\"\n",
    "<br>**amount**: (numeric) money spent in \"transaction\"\n",
    "<br>**reward**: (numeric) money gained from \"offer completed\"\n",
    "<br>**time**: (numeric) hours after start of test\n",
    "\n",
    "There are 4 types of transcript: normal transaction, offer received, offer viewed and offer complete. \n",
    "\n",
    "**Each person can receive the same type of offer for multiple times** and the offer id only tells what type of offer that is ---- it can't be used to recognize a unique offer. Our challenge here is to match the various actions to tell if a certain offer was viewed or completed. Also, customers might not received all types of offers.\n",
    "\n",
    "**We solve this problem by giving an order number of the transactions for each customer. The offer id plus the rank can be seen as a \"unique ID\" for that offer.** We discovered that until the current offer expires, the app won't send another offer under the same type to that customer (we will prove this later). We will attribute the earlist following actions (view the offer or complete the offer) to the earliest offer (under the same offer id) with an extra condition that the following actions need to happen before that offer expires."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "transaction        0.453304\n",
       "offer received     0.248837\n",
       "offer viewed       0.188315\n",
       "offer completed    0.109544\n",
       "Name: event, dtype: float64"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Distribution of transaction types\n",
    "\n",
    "transcript['event'].value_counts(normalize=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "transcript = pd.read_json('transcript.json', orient='records', lines=True).reset_index()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "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>index</th>\n",
       "      <th>event</th>\n",
       "      <th>person</th>\n",
       "      <th>time</th>\n",
       "      <th>value</th>\n",
       "      <th>offer_id</th>\n",
       "      <th>amount</th>\n",
       "      <th>reward</th>\n",
       "      <th>rank_in_user</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>112281</th>\n",
       "      <td>112281</td>\n",
       "      <td>offer received</td>\n",
       "      <td>6006641a7b15444fae1ae0a7e27a88dd</td>\n",
       "      <td>336</td>\n",
       "      <td>{'offer id': '0b1e1539f2cc45b7b9fa7c272da2e1d7'}</td>\n",
       "      <td>0b1e1539f2cc45b7b9fa7c272da2e1d7</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>8.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>167074</th>\n",
       "      <td>167074</td>\n",
       "      <td>offer viewed</td>\n",
       "      <td>b68bc936296144569115d7cbe8077f51</td>\n",
       "      <td>408</td>\n",
       "      <td>{'offer id': 'fafdcd668e3743c1bb461111dcafc2a4'}</td>\n",
       "      <td>fafdcd668e3743c1bb461111dcafc2a4</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>16.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>267219</th>\n",
       "      <td>267219</td>\n",
       "      <td>offer completed</td>\n",
       "      <td>d8ec9896d2a34515af66ccf7be5fbe66</td>\n",
       "      <td>588</td>\n",
       "      <td>{'offer_id': 'ae264e3637204a6fb9bb56bc8210ddfd...</td>\n",
       "      <td>ae264e3637204a6fb9bb56bc8210ddfd</td>\n",
       "      <td>NaN</td>\n",
       "      <td>10.0</td>\n",
       "      <td>16.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>264754</th>\n",
       "      <td>264754</td>\n",
       "      <td>transaction</td>\n",
       "      <td>b1d541fa105e4b6f8f6c6a14831e938f</td>\n",
       "      <td>582</td>\n",
       "      <td>{'amount': 75.65}</td>\n",
       "      <td>NaN</td>\n",
       "      <td>75.65</td>\n",
       "      <td>NaN</td>\n",
       "      <td>17.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41932</th>\n",
       "      <td>41932</td>\n",
       "      <td>transaction</td>\n",
       "      <td>ec77e19f18ae42e5ac50e450f0e3f3e3</td>\n",
       "      <td>96</td>\n",
       "      <td>{'amount': 19.52}</td>\n",
       "      <td>NaN</td>\n",
       "      <td>19.52</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         index            event                            person  time  \\\n",
       "112281  112281   offer received  6006641a7b15444fae1ae0a7e27a88dd   336   \n",
       "167074  167074     offer viewed  b68bc936296144569115d7cbe8077f51   408   \n",
       "267219  267219  offer completed  d8ec9896d2a34515af66ccf7be5fbe66   588   \n",
       "264754  264754      transaction  b1d541fa105e4b6f8f6c6a14831e938f   582   \n",
       "41932    41932      transaction  ec77e19f18ae42e5ac50e450f0e3f3e3    96   \n",
       "\n",
       "                                                    value  \\\n",
       "112281   {'offer id': '0b1e1539f2cc45b7b9fa7c272da2e1d7'}   \n",
       "167074   {'offer id': 'fafdcd668e3743c1bb461111dcafc2a4'}   \n",
       "267219  {'offer_id': 'ae264e3637204a6fb9bb56bc8210ddfd...   \n",
       "264754                                  {'amount': 75.65}   \n",
       "41932                                   {'amount': 19.52}   \n",
       "\n",
       "                                offer_id  amount  reward  rank_in_user  \n",
       "112281  0b1e1539f2cc45b7b9fa7c272da2e1d7     NaN     NaN           8.0  \n",
       "167074  fafdcd668e3743c1bb461111dcafc2a4     NaN     NaN          16.0  \n",
       "267219  ae264e3637204a6fb9bb56bc8210ddfd     NaN    10.0          16.0  \n",
       "264754                               NaN   75.65     NaN          17.0  \n",
       "41932                                NaN   19.52     NaN           4.0  "
      ]
     },
     "execution_count": 85,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "transcript.sample(frac=1).head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "#We will use some custom functions to extract offer_id, reward or amount in the 'value' column\n",
    "\n",
    "def get_offer_name1(x):\n",
    "    try:\n",
    "        return x['offer id']\n",
    "    except:\n",
    "        return float('nan')\n",
    "\n",
    "def get_offer_name2(x):\n",
    "    try:\n",
    "        return x['offer_id']\n",
    "    except:\n",
    "        return float('nan') \n",
    "    \n",
    "def get_amount(x):\n",
    "    try:\n",
    "        return x['amount']\n",
    "    except:\n",
    "        return float('nan')\n",
    "    \n",
    "def get_reward(x):\n",
    "    try:\n",
    "        return x['reward']\n",
    "    except:\n",
    "        return float('nan')\n",
    "\n",
    "    \n",
    "transcript['offer_id'] = transcript['value'].apply(get_offer_name1)\n",
    "transcript.loc[(transcript['offer_id'].isna()), \"offer_id\"] = transcript['value'].apply(get_offer_name2)\n",
    "transcript['amount'] = transcript['value'].apply(get_amount)\n",
    "transcript['reward'] = transcript['value'].apply(get_reward)\n",
    "transcript['rank_in_user'] = transcript.groupby('person')['index'].rank() #add a rank to help us attribute offer related activities"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "In total we offered 76,277 promotions.\n"
     ]
    }
   ],
   "source": [
    "print(\"In total we offered {:,} promotions.\"\\\n",
    "      .format(len(transcript[transcript['event'] == \"offer received\"])))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "In total there are 63,288 unique person + offer_type combinations.\n"
     ]
    }
   ],
   "source": [
    "print(\"In total there are {:,} unique person + offer_type combinations.\"\\\n",
    "      .format(len(transcript[transcript['event'] == \"offer received\"][['person','offer_id']].drop_duplicates())))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "> Customers can receive the same type of offer for multiple times."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "scrolled": 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>person</th>\n",
       "      <th>index</th>\n",
       "      <th>event</th>\n",
       "      <th>time</th>\n",
       "      <th>value</th>\n",
       "      <th>offer_id</th>\n",
       "      <th>amount</th>\n",
       "      <th>reward</th>\n",
       "      <th>rank_in_user</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0009655768c64bdeb2e877511632db8f</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>00116118485d4dfda04fdbaba9a87b5c</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0011e0d4e6b944f998e987f904e8c1e5</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0020c2b971eb4e9188eac86d93036a77</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0020ccbbb6d84e358d3414a3ff76cffd</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                             person  index  event  time  value  offer_id  \\\n",
       "0  0009655768c64bdeb2e877511632db8f      5      5     5      5         5   \n",
       "1  00116118485d4dfda04fdbaba9a87b5c      2      2     2      2         2   \n",
       "2  0011e0d4e6b944f998e987f904e8c1e5      5      5     5      5         5   \n",
       "3  0020c2b971eb4e9188eac86d93036a77      5      5     5      5         5   \n",
       "4  0020ccbbb6d84e358d3414a3ff76cffd      4      4     4      4         4   \n",
       "\n",
       "   amount  reward  rank_in_user  \n",
       "0       0       0             5  \n",
       "1       0       0             2  \n",
       "2       0       0             5  \n",
       "3       0       0             5  \n",
       "4       0       0             4  "
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "transcript[transcript['event'] == \"offer received\"].groupby(['person']).count().reset_index().head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "transcript.to_csv(\"transcript_cleaned.csv\", index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "> Customers might not receive all the offers."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2.3.1 Match offer_received with offer_viewed"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "#from_received_to_viewed is table that matches the offer_received and offer_viewed activities\n",
    "\n",
    "from_received_to_viewed = transcript[transcript['event'] == \"offer received\"][['person', 'offer_id', 'event','time','rank_in_user']]\\\n",
    "                                        .merge(transcript[transcript['event'] == \"offer viewed\"][['person', 'offer_id', 'event','time','rank_in_user']], \n",
    "                                                  on=['person', 'offer_id'], how='left')\\\n",
    "                                        .merge(portfolio[['id','duration']].rename({'id':'offer_id'}, axis=1))\n",
    "\n",
    "from_received_to_viewed['time_received_to_viewed'] = from_received_to_viewed['time_y'] - from_received_to_viewed['time_x']\n",
    "\n",
    "#the offer_viewed action must happen after the offer_received action\n",
    "#the offer_viewed action must be within the active duration of that offer\n",
    "from_received_to_viewed = from_received_to_viewed[(from_received_to_viewed['rank_in_user_x'] < from_received_to_viewed['rank_in_user_y']) & \\\n",
    "                                                                           (from_received_to_viewed['time_received_to_viewed'] <= from_received_to_viewed['duration']*24)]\n",
    "\n",
    "#After the filters above, match the first offer_view with the first offer_received, the 2nd offer view with the 2nd offer_completed, ..., etc\n",
    "from_received_to_viewed = from_received_to_viewed\\\n",
    "                                        .sort_values('rank_in_user_y').groupby(['person','offer_id','rank_in_user_x']).first().reset_index()\\\n",
    "                                        .sort_values('rank_in_user_x').groupby(['person','offer_id','rank_in_user_y']).first().reset_index()\n",
    "\n",
    "from_received_to_viewed['time_received_to_viewed'] = (from_received_to_viewed['time_y'] - from_received_to_viewed['time_x']) / 24"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "56567\n",
      "56567\n",
      "56567\n"
     ]
    }
   ],
   "source": [
    "#check whether the matches are unique; the below 3 numbers should be the same\n",
    "print(len(from_received_to_viewed))\n",
    "print(len(from_received_to_viewed[['person','offer_id','rank_in_user_x']].drop_duplicates()))\n",
    "print(len(from_received_to_viewed[['person','offer_id','rank_in_user_y']].drop_duplicates()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Days taken to view the offer')"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.hist(from_received_to_viewed['time_received_to_viewed'])\n",
    "plt.title(\"Days taken to view the offer\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "66.96% of the offer can be viewed within 1 day\n"
     ]
    }
   ],
   "source": [
    "print(\"{:.2%} of the offer can be viewed within 1 day\"\\\n",
    "      .format(sum(from_received_to_viewed['time_received_to_viewed'] <= 1)/len(from_received_to_viewed)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2.3.2 Match offer_received with offer_completed"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "#use the same logic to match offer_received and offer_completed\n",
    "\n",
    "from_received_to_completed = transcript[transcript['event'] == \"offer received\"][['person', 'offer_id', 'event','time','rank_in_user']]\\\n",
    "                                        .merge(transcript[transcript['event'] == \"offer completed\"][['person', 'offer_id', 'event','time','rank_in_user']], \n",
    "                                                  on=['person', 'offer_id'], how='left')\\\n",
    "                                        .merge(portfolio[['id','duration']].rename({'id':'offer_id'}, axis=1))\n",
    "\n",
    "from_received_to_completed['time_received_to_completed'] = from_received_to_completed['time_y'] - from_received_to_completed['time_x']\n",
    "\n",
    "from_received_to_completed = from_received_to_completed[(from_received_to_completed['rank_in_user_x'] < from_received_to_completed['rank_in_user_y']) & \\\n",
    "                                                                           (from_received_to_completed['time_received_to_completed'] <= from_received_to_completed['duration']*24)]\n",
    "\n",
    "\n",
    "from_received_to_completed = from_received_to_completed\\\n",
    "                                        .sort_values('rank_in_user_y').groupby(['person','offer_id','rank_in_user_x']).first().reset_index()\\\n",
    "                                        .sort_values('rank_in_user_x').groupby(['person','offer_id','rank_in_user_y']).first().reset_index()\n",
    "\n",
    "from_received_to_completed['time_received_to_completed'] = (from_received_to_completed['time_y'] - from_received_to_completed['time_x']) / 24"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "33101\n",
      "33101\n",
      "33101\n"
     ]
    }
   ],
   "source": [
    "#check whether the matches are unique; the below 3 numbers should be the same\n",
    "print(len(from_received_to_completed))\n",
    "print(len(from_received_to_completed[['person','offer_id','rank_in_user_x']].drop_duplicates()))\n",
    "print(len(from_received_to_completed[['person','offer_id','rank_in_user_y']].drop_duplicates()))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2.3.3 Create a table to reflect the funnel\n",
    "\n",
    "offer_received \n",
    "<br>|__ offer viewed \n",
    "<br>&nbsp; &nbsp; &nbsp; |__ offer completed\n",
    "<br>&nbsp; &nbsp; &nbsp; |__ offer not completed\n",
    "<br>|__ offer not viewed\n",
    "<br>&nbsp; &nbsp; &nbsp; |__ offer completed\n",
    "<br>&nbsp; &nbsp; &nbsp; |__ offer not completed"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "offer_pipe = transcript[transcript['event'] == \"offer received\"][['person', 'offer_id','time','rank_in_user','event']]\\\n",
    "                    .merge(from_received_to_viewed[['person','offer_id','rank_in_user_x','event_y','time_received_to_viewed']]\\\n",
    "                              .rename({'rank_in_user_x':'rank_in_user'}, axis=1), \n",
    "                              on=['person','offer_id','rank_in_user'], how=\"left\")\\\n",
    "                    .merge(from_received_to_completed[['person','offer_id','rank_in_user_x','event_y','time_received_to_completed']]\\\n",
    "                              .rename({'rank_in_user_x':'rank_in_user'}, axis=1), \n",
    "                              on=['person','offer_id','rank_in_user'], how=\"left\")\n",
    "offer_pipe['viewed'] = offer_pipe['event_y_x'].notna()\n",
    "offer_pipe['time_viewed_to_completed'] = offer_pipe['time_received_to_completed'] - offer_pipe['time_received_to_viewed']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "76277\n",
      "76277\n",
      "76277\n"
     ]
    }
   ],
   "source": [
    "#check whether the matches are unique; the below 3 numbers should be the same\n",
    "#in offer pipe, each row is a unique offer\n",
    "print(len(offer_pipe))\n",
    "print(len(offer_pipe[['person','offer_id','rank_in_user']].drop_duplicates()))\n",
    "print(len(transcript[transcript['event'] == \"offer received\"][['person','offer_id','rank_in_user']].drop_duplicates()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/tianyiwang/anaconda3/lib/python3.7/site-packages/numpy/lib/histograms.py:754: RuntimeWarning: invalid value encountered in greater_equal\n",
      "  keep = (tmp_a >= first_edge)\n",
      "/Users/tianyiwang/anaconda3/lib/python3.7/site-packages/numpy/lib/histograms.py:755: RuntimeWarning: invalid value encountered in less_equal\n",
      "  keep &= (tmp_a <= last_edge)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Days taken to complete the offer')"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.hist(offer_pipe['time_received_to_completed'])\n",
    "plt.title(\"Days taken to complete the offer\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Days taken from viewing the offer to completing the offer')"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.hist(offer_pipe['time_received_to_completed'] - offer_pipe['time_received_to_viewed'])\n",
    "plt.title(\"Days taken from viewing the offer to completing the offer\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Notice that the negative numbers mean that the offers were completed before viewed. This means that the customer made the purchase that already qualified for the promotion although they didn't know about the promotion."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "74.16% offers were viewed.\n"
     ]
    }
   ],
   "source": [
    "print(\"{:.2%} offers were viewed.\"\\\n",
    "      .format((sum(offer_pipe[offer_p]['viewed']))/len(offer_pipe)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "12.86% offers were completed without being viewed.\n"
     ]
    }
   ],
   "source": [
    "print(\"{:.2%} offers were completed without being viewed.\"\\\n",
    "       .format(sum((offer_pipe['event_y_x'].isna() & offer_pipe['event_y_y'].notna()) | \\\n",
    "       (offer_pipe['time_received_to_completed'] < offer_pipe['time_received_to_viewed']))/len(offer_pipe)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "13.54% of the offer will be completed within one day after being viewed\n"
     ]
    }
   ],
   "source": [
    "print(\"{:.2%} of the offer will be completed within one day after being viewed\"\\\n",
    "        .format(sum((offer_pipe['time_received_to_completed'] - offer_pipe['time_received_to_viewed'] <= 1) & \\\n",
    "                   (offer_pipe['time_received_to_completed'] - offer_pipe['time_received_to_viewed'] >= 0)) / len(offer_pipe)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "view_complete = offer_pipe['event_y_x'].notna() & offer_pipe['event_y_y'].notna()\n",
    "view_not_complete = offer_pipe['event_y_x'].notna() & offer_pipe['event_y_y'].isna()\n",
    "not_view_complete = offer_pipe['event_y_x'].isna() & offer_pipe['event_y_y'].notna()\n",
    "not_view_not_complete = offer_pipe['event_y_x'].isna() & offer_pipe['event_y_y'].isna()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "offer_pipe['viewed'] = offer_pipe['event_y_x'].notna()\n",
    "offer_pipe['completed'] = offer_pipe['event_y_y'].notna()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "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>viewed</th>\n",
       "      <th>completed</th>\n",
       "      <th>person</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>13995</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>5715</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>29181</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>27386</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   viewed  completed  person\n",
       "0   False      False   13995\n",
       "1   False       True    5715\n",
       "2    True      False   29181\n",
       "3    True       True   27386"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "offer_pipe.groupby(['viewed','completed']).count()['person'].reset_index()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.07492428910418605"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(5715)/(13995+5715+29181+27386)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "6 people didn't receive any offer\n"
     ]
    }
   ],
   "source": [
    "print(\"{} people didn't receive any offer\".format(len(transcript['person'].unique()) - len(offer_pipe['person'].unique())))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 3. Put things together -- describe our customers\n",
    "\n",
    "Right now we have customers' demographics, their transactions and promotion related data. We will put these things together to see whether there are any interesting insights."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "final_table = offer_pipe.drop(['event','event_y_x','event_y_y'], axis=1)\\\n",
    ".merge(portfolio[['id','offer_type','difficulty','duration','reward']].rename({'id':'offer_id'}, axis=1), on='offer_id')\\\n",
    ".merge(profile.rename({'id':'person'}, axis=1)[['person','age','gender','age_bucket','customer_age','income']], on='person')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "76277\n",
      "76277\n"
     ]
    }
   ],
   "source": [
    "print(len(offer_pipe))\n",
    "print(len(final_table))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Besides offers, we also want to create some features around the transactions."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "t1 = transcript[transcript['event'] == \"transaction\"]\n",
    "t2 = transcript[transcript['event'] == \"offer completed\"]\n",
    "\n",
    "per_person = profile.rename({'id':'person'}, axis=1)[['person','age','gender','age_bucket','customer_age']]\\\n",
    ".merge(t1.groupby('person').count()['index'].reset_index().rename({'index':'numbers_of_purchases'}, axis=1).fillna(0), on='person', how='left')\\\n",
    ".merge(t1.groupby('person')['amount'].sum().reset_index().rename({'amount':'total_purchase_amount'}, axis=1).fillna(0), on='person', how='left')\\\n",
    ".merge(t2.groupby('person')['reward'].sum().reset_index().rename({'reward':'total_reward_amount'}, axis=1).fillna(0), on='person', how='left')\n",
    "\n",
    "per_person['net_revenue'] = per_person['total_purchase_amount'] - per_person['total_reward_amount']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "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>person</th>\n",
       "      <th>age</th>\n",
       "      <th>gender</th>\n",
       "      <th>age_bucket</th>\n",
       "      <th>customer_age</th>\n",
       "      <th>numbers_of_purchases</th>\n",
       "      <th>total_purchase_amount</th>\n",
       "      <th>total_reward_amount</th>\n",
       "      <th>net_revenue</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>68be06ca386d4c31939f3a4f0e3dd783</td>\n",
       "      <td>NaN</td>\n",
       "      <td>None</td>\n",
       "      <td>65 ~</td>\n",
       "      <td>39.366667</td>\n",
       "      <td>9.0</td>\n",
       "      <td>20.40</td>\n",
       "      <td>5.0</td>\n",
       "      <td>15.40</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0610b486422d4921ae7d2bf64640c50b</td>\n",
       "      <td>55.0</td>\n",
       "      <td>F</td>\n",
       "      <td>55 ~ 64</td>\n",
       "      <td>34.266667</td>\n",
       "      <td>3.0</td>\n",
       "      <td>77.01</td>\n",
       "      <td>5.0</td>\n",
       "      <td>72.01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>38fe809add3b4fcf9315a9694bb96ff5</td>\n",
       "      <td>NaN</td>\n",
       "      <td>None</td>\n",
       "      <td>65 ~</td>\n",
       "      <td>22.200000</td>\n",
       "      <td>6.0</td>\n",
       "      <td>14.30</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>78afa995795e4d85b5d9ceeca43f5fef</td>\n",
       "      <td>75.0</td>\n",
       "      <td>F</td>\n",
       "      <td>65 ~</td>\n",
       "      <td>36.500000</td>\n",
       "      <td>7.0</td>\n",
       "      <td>159.27</td>\n",
       "      <td>20.0</td>\n",
       "      <td>139.27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>a03223e636434f42ac4c3df47e8bac43</td>\n",
       "      <td>NaN</td>\n",
       "      <td>None</td>\n",
       "      <td>65 ~</td>\n",
       "      <td>33.600000</td>\n",
       "      <td>3.0</td>\n",
       "      <td>4.65</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                             person   age gender age_bucket  customer_age  \\\n",
       "0  68be06ca386d4c31939f3a4f0e3dd783   NaN   None       65 ~     39.366667   \n",
       "1  0610b486422d4921ae7d2bf64640c50b  55.0      F    55 ~ 64     34.266667   \n",
       "2  38fe809add3b4fcf9315a9694bb96ff5   NaN   None       65 ~     22.200000   \n",
       "3  78afa995795e4d85b5d9ceeca43f5fef  75.0      F       65 ~     36.500000   \n",
       "4  a03223e636434f42ac4c3df47e8bac43   NaN   None       65 ~     33.600000   \n",
       "\n",
       "   numbers_of_purchases  total_purchase_amount  total_reward_amount  \\\n",
       "0                   9.0                  20.40                  5.0   \n",
       "1                   3.0                  77.01                  5.0   \n",
       "2                   6.0                  14.30                  NaN   \n",
       "3                   7.0                 159.27                 20.0   \n",
       "4                   3.0                   4.65                  NaN   \n",
       "\n",
       "   net_revenue  \n",
       "0        15.40  \n",
       "1        72.01  \n",
       "2          NaN  \n",
       "3       139.27  \n",
       "4          NaN  "
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "per_person.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "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>person</th>\n",
       "      <th>offer_id</th>\n",
       "      <th>time</th>\n",
       "      <th>rank_in_user</th>\n",
       "      <th>time_received_to_viewed</th>\n",
       "      <th>time_received_to_completed</th>\n",
       "      <th>viewed</th>\n",
       "      <th>time_viewed_to_completed</th>\n",
       "      <th>completed</th>\n",
       "      <th>offer_type</th>\n",
       "      <th>difficulty</th>\n",
       "      <th>duration</th>\n",
       "      <th>reward</th>\n",
       "      <th>age</th>\n",
       "      <th>gender</th>\n",
       "      <th>age_bucket</th>\n",
       "      <th>customer_age</th>\n",
       "      <th>income</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>78afa995795e4d85b5d9ceeca43f5fef</td>\n",
       "      <td>9b98b8c7a33c4b65b9aebfe6a799e6d9</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.25</td>\n",
       "      <td>5.50</td>\n",
       "      <td>True</td>\n",
       "      <td>5.25</td>\n",
       "      <td>True</td>\n",
       "      <td>bogo</td>\n",
       "      <td>5</td>\n",
       "      <td>7</td>\n",
       "      <td>5</td>\n",
       "      <td>75.0</td>\n",
       "      <td>F</td>\n",
       "      <td>65 ~</td>\n",
       "      <td>36.500000</td>\n",
       "      <td>100000.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>78afa995795e4d85b5d9ceeca43f5fef</td>\n",
       "      <td>f19421c1d4aa40978ebb69ca19b0e20d</td>\n",
       "      <td>504</td>\n",
       "      <td>13.0</td>\n",
       "      <td>3.25</td>\n",
       "      <td>0.25</td>\n",
       "      <td>True</td>\n",
       "      <td>-3.00</td>\n",
       "      <td>True</td>\n",
       "      <td>bogo</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>75.0</td>\n",
       "      <td>F</td>\n",
       "      <td>65 ~</td>\n",
       "      <td>36.500000</td>\n",
       "      <td>100000.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>78afa995795e4d85b5d9ceeca43f5fef</td>\n",
       "      <td>ae264e3637204a6fb9bb56bc8210ddfd</td>\n",
       "      <td>408</td>\n",
       "      <td>11.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>4.25</td>\n",
       "      <td>True</td>\n",
       "      <td>4.25</td>\n",
       "      <td>True</td>\n",
       "      <td>bogo</td>\n",
       "      <td>10</td>\n",
       "      <td>7</td>\n",
       "      <td>10</td>\n",
       "      <td>75.0</td>\n",
       "      <td>F</td>\n",
       "      <td>65 ~</td>\n",
       "      <td>36.500000</td>\n",
       "      <td>100000.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>78afa995795e4d85b5d9ceeca43f5fef</td>\n",
       "      <td>5a8bc65990b245e5a138643cd4eb9837</td>\n",
       "      <td>168</td>\n",
       "      <td>6.0</td>\n",
       "      <td>2.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>True</td>\n",
       "      <td>NaN</td>\n",
       "      <td>False</td>\n",
       "      <td>informational</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>75.0</td>\n",
       "      <td>F</td>\n",
       "      <td>65 ~</td>\n",
       "      <td>36.500000</td>\n",
       "      <td>100000.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>ebe7ef46ea6f4963a7dd49f501b26779</td>\n",
       "      <td>9b98b8c7a33c4b65b9aebfe6a799e6d9</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>False</td>\n",
       "      <td>NaN</td>\n",
       "      <td>False</td>\n",
       "      <td>bogo</td>\n",
       "      <td>5</td>\n",
       "      <td>7</td>\n",
       "      <td>5</td>\n",
       "      <td>59.0</td>\n",
       "      <td>M</td>\n",
       "      <td>55 ~ 64</td>\n",
       "      <td>64.466667</td>\n",
       "      <td>41000.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                             person                          offer_id  time  \\\n",
       "0  78afa995795e4d85b5d9ceeca43f5fef  9b98b8c7a33c4b65b9aebfe6a799e6d9     0   \n",
       "1  78afa995795e4d85b5d9ceeca43f5fef  f19421c1d4aa40978ebb69ca19b0e20d   504   \n",
       "2  78afa995795e4d85b5d9ceeca43f5fef  ae264e3637204a6fb9bb56bc8210ddfd   408   \n",
       "3  78afa995795e4d85b5d9ceeca43f5fef  5a8bc65990b245e5a138643cd4eb9837   168   \n",
       "4  ebe7ef46ea6f4963a7dd49f501b26779  9b98b8c7a33c4b65b9aebfe6a799e6d9     0   \n",
       "\n",
       "   rank_in_user  time_received_to_viewed  time_received_to_completed  viewed  \\\n",
       "0           1.0                     0.25                        5.50    True   \n",
       "1          13.0                     3.25                        0.25    True   \n",
       "2          11.0                     0.00                        4.25    True   \n",
       "3           6.0                     2.00                         NaN    True   \n",
       "4           1.0                      NaN                         NaN   False   \n",
       "\n",
       "   time_viewed_to_completed  completed     offer_type  difficulty  duration  \\\n",
       "0                      5.25       True           bogo           5         7   \n",
       "1                     -3.00       True           bogo           5         5   \n",
       "2                      4.25       True           bogo          10         7   \n",
       "3                       NaN      False  informational           0         3   \n",
       "4                       NaN      False           bogo           5         7   \n",
       "\n",
       "   reward   age gender age_bucket  customer_age    income  \n",
       "0       5  75.0      F       65 ~     36.500000  100000.0  \n",
       "1       5  75.0      F       65 ~     36.500000  100000.0  \n",
       "2      10  75.0      F       65 ~     36.500000  100000.0  \n",
       "3       0  75.0      F       65 ~     36.500000  100000.0  \n",
       "4       5  59.0      M    55 ~ 64     64.466667   41000.0  "
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "final_table.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "> We now have 2 tables:\n",
    "<br>`final_table`: This table is on the offer level --- each row is a unique offer. We can use this table to see the statuses of the offers --- was the offer viewed? Was the offer completed?...\n",
    "<br>`per_person`: This table is on customer level --- each row is a unique customer. We can use this table to get an overview of each customer --- how many purchases they have made so far? What are their net revenues?...\n",
    "\n",
    "> In this data exploration section, we will answer these 2 questions: \n",
    "> #### (1) What types of customers are more reponsive to each type of campaigns? (2) What types of customers are more valuable?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [],
   "source": [
    "final_table_not_viewed = final_table[~final_table['viewed']]\n",
    "final_table_viewed = final_table[final_table['viewed']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [],
   "source": [
    "def complete_rate_compare(x):\n",
    "    final_table_not_viewed.groupby(x)['completed'].mean().reset_index()\\\n",
    "    .merge(final_table_viewed.groupby(x)['completed'].mean().reset_index(), on=x)\\\n",
    "    .rename({'completed_x':'not_viewed', 'completed_y':'viewed'}, axis=1)\\\n",
    "    .set_index(x).plot.bar()\n",
    "    plt.title(\"Offer completion rate\")\n",
    "    plt.xticks(rotation = 0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "complete_rate_compare('offer_type')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    ">In general, offers that are viewed are more likely to be completed, especially for discount promotions. This is also because discount offers usually require higher amount of purchases, while buy-one-get-one is easier to achieve."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "complete_rate_compare('gender')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "> Female customers are more likely to complete the offers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "complete_rate_compare('age_bucket')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "> Older customers are more likely to complete the offers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.DataFrame(np.random.rand(10, 5), columns=['A', 'B', 'C', 'D', 'E'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Income and whether the customer completed the offer')"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "final_table[['completed','income']].pivot(columns=\"completed\")['income'].plot.box()\n",
    "plt.xlabel('completed the offer')\n",
    "plt.ylabel('income')\n",
    "plt.title(\"Income and whether the customer completed the offer\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    ">Income level is a positive factor"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, \"'Customer age' and whether the customer completed the offer\")"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZUAAAEWCAYAAACufwpNAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvOIA7rQAAIABJREFUeJzt3XeYXWW5/vHvnQRIgFASBiRAElAERJpEilg4gFHxUH6KgiIGDoqFi6IgCRx/ggrHWMF2RBAlgPRDE48EiASUntBJVCSEUEKYAIFQpCTP+eN9h6xs9sysmaw1k5ncn+va1179ffZqz3pX24oIzMzMqjCgtwMwM7P+w0nFzMwq46RiZmaVcVIxM7PKOKmYmVllnFTMzKwyTirWlKRdJT2+jNMYLSkkDaoopoMl/bWKaXWhzKmSvtCTZVp1qliPu1BWl9b3qrePTsraRdJDkl6UtK+k9STdJGmhpB9XWVblSUXSbEmjC+07SPpfSQskPSvpDkmHLGMZPbaiLG/yTm7X3o6jbj25wRXKPEnSeT1V3rLqjSTbn9W9/PO+cY+6pt+J7wC/iIjVI+IK4DBgPrBGRBxTZUG11lQk7Qz8GbgReAcwHPgK8LE6y+1JPbnTs+p4uVXL83O5Nwp4sKF9RnTj6fdOl3VEVPoBZgOjc/NfgV92MOzBwF8bugXwjty8JzADWAg8ARwLrAa8AiwGXsyfEcAqwGnAk/lzGrBKns6uwOPAccDTwFxg3zz9fwDPAicUYhgATAAeBp4BLgaG5X6jc4yHAnOAm5r8rrWBq4FW4LncvGGh/8bATfl3XQ/8Ejiv0H8n4BZgAXAvsGuh39Rie0O5HwfuBl4AHgNOKvRri3tcjns+8J+F/kOAs3O8M4BvAI+3U863gZ/n5pWAl4AfFKbzrzwPOiuzo/k8J4/btox3bltfgB/lOB8BPlaY3prAWXn5PgGcDAwsrGs3A6fm5X1yw2/6KPAa8Hou797C/P5uHnchcC2wTpll1WS+bQRclteLZ0hHjgAnNSz/tvk2qBD7rFz+I8CBwBZ5Pi/K8S4ozINzchmPAt8EBjSZBwvyNN+Xuz9G2jbGFeJYJc/rOcA84HRgSMM2NR54Cji3nd/8RWBmjn0G8J7cfYs8bxeQdnZ7F8Y5G/hv4E/5t90MvI20TT8H/A3YrmGfc3ye/nPA74DBxTgLw44A/ifPn0eAIztZ/h2tUwPz/Jmf5+XhxeXWMB/OJe2zXsnTP45l2D46mNf/JK3fVwEjcveHG8q+IP/O13L7Hh2VRYl93lJxVJ1UCj9wVdIK/2/LkFTmAh8o7KjbVsilVpTc7TvAbcC6QAtpQ/9uYfg3gG+RdoJfzCvV+cBQYEvSBrpJHv7oPK0NSRvWr4ELGmbwOaQEN6TJ7xoOfDLPg6HAJcAVhf635pVxZeD9pCRwXu63QV6oe+YF/eHc3lJinu8KbJXH25q0I9i3Ie4zSTv+bYBXgS1y/4nAX4BhpJ3fA43zuFDObsD9ufl9eUW8vdDv3pJllpnPgwrlHkzaGL5I2qC/QjqAUO5/RZ7Gank9uAP4UmHcN4AjgEHtLLeTKOzcc7ep+fe9M/+GqcDEri6rHO+9pB36asBg4P3Nyi3+9jzsC8Bmud/6wJYdbD/nAFeS1rvRpIOmQxvmwSE5npNJO4lf5vk/lrTzXz0Pfxpp5zQsT+8PwPcatqnv53Gbzc9PkXbE7wVEOlsxirQN/hM4gbQN7JbLbfuNZ5N2sNvn+fRnUgL4fCHuGwrlzCatrxvlWG8mHzRQ2FfkZTSdtB9YGdiElAw+0sHy72id+jIpwbWVewPtJJVCnHs0Wc5d3j7a2SbnA+/Jw/6cws6/SdlnUziw6qgsSuzzloqlu0mjxA5ugxzI5h0MczAdJ5U5wJdI5/0ad56NSeVhYM9C+0eA2YXhX2HJEcbQXM6OheGns2QHPBPYvdBvfdLObFBhBm/ShXmxLfBcbh5J2hhXLfQ/jyVJZTwNR33AZApHkF0o9zTg1IYVo1hjugM4IDfPAj5a6HdY4zwu9GurjQwnHd2cQDpqXZ1Ui/lZyTLLzOfGpPLPQvuqeZi3AeuRNsghhf6fIe988rhzOplfJ9E8qXyz0P5V4JquLitSTauV5kexS5XLW5PKAtJBypCG8Q6msP2QdrivAu8qdPsSMLUw/EOFflvlctYrdHuGtL6KVAN9e8NveKSwTb1GrhG0Mz8nA0c16f4BUu1mQKHbBeSaNWmHd2ah3xHAzIa4FxTaZwNfLrTvCTxciLMtqezYuA6Qaji/a2c5dLZO/bmh3LF0L6l0eftoMu2zyGcLcvvqedjR7ZR9Nksnlcr2eXWeB32OVOVan5TNu+OTpOr7REn3ARMi4tZ2hh1Bqu63eTR3a/NMRCzKza/k73mF/q+QFgSko6nLJS0u9F9EWsnaPNZe0JJWJR2RfpRUwwIYKmlgjunZiHi5YVobFcr+lKS9Cv1XIh0FdUjSjqQax7tJR2KrkGpJRU8Vml9myW8e0fCbivNyKRHxiqRpwIeADwKnkHZEu+RuPy9ZZpn53OjNaUXEy5LI0xtGmk9zczdIR6bF39TuMutER/GXXVYbAY9GxBtdKTgiXpK0P+nU71mSbgaOiYhm29Q6pOXeuB1sUGhvXOeJiGbbQQspaU8vzE+REleb1oj4Vwfhb0Q62Gs0AngsIorLvbM429tW2zSuuyN4q1HACEkLCt0GkmrozbTVqtpbp0pvM53ozvbxRMM0RgB3tbVExIuSniHN09klYlimfV5RbRfq807zVlJiaM9LpBUXAElva5jGnRGxD6naeQXpPB+krNnoSdKMaTMyd+uOx0jn6tcqfAZHRHFBNouhzTHAZqSa0BqkHS+kjXIuMCwnnjYbFZofIx39FsteLSImloj7fNLpio0iYk3SOXB1PMqb5jbEMbKT4W8kVbm3A+7M7R8BdiBdLyqjo/nc0fxtb1qvkq53tE1rjYjYsjBMZ9PsTplll9VjwMh2LnIutR2Qal5LgoqYHBEfZskB2pntxDufdHTZuB007oDKmE/aeW9Z+G1rRkRxZ97Z/HoMeHuT7k8CG0kq7n+6G2ebxnW32bb/GKmmVVxeQyNiz9y/8fd0tk51dZvpzvrV2X6ozVL7P0mrkc4klJ2ny7rPe1Pdz6kcBxws6RuShgNI2kbShbn/vcCWkraVNJhU/SQPt7KkAyWtGRGvk84rt9U05gHDJa1ZKOsC4JuSWiStQzpv2t3bA08HTpE0KsfSImmfLow/lLRBLpA0DDixrUdEPApMA07Kv3FnoHikex6wl6SPSBooaXC+hXrDkuU+GxH/krQD8NkuxHwxcLyktXNZR3Qy/I2kc9wzIuI10mmiL5A22taSZXY0n1tJNd1NykwoIuaSLqL/WNIakgZIerukD5WMBdJ6NbphZ9eRriyrO0g7oYmSVsvD7pL73QN8UNLIvE4f3zZSfp5g77yTeJV0YbW4HWwoaeU8DxaRluMpkobm+fp1urEd5FrEmcCpktbNsWwg6SNdmMxvgGMlba/kHTmm20mJ9DhJK+Vb5PcCLuxgWp05XNKGeXs7AbioyTB3AC9IGi9pSF5m75b03tx/qeVfYp26GDgyl7s26VRwR+ZRcn3OurIfOh84JO9LVwH+i3Sdc3YNZXWo1qQSEbeQjmZ3A2ZJehY4A/jf3P8fpAvs1wMPke7sKToImC3pBdJFsc/l8f5GSiKzlJ5/GUG6eDcNuA+4n1QVPLmbof+UdMR/raSFpAtYO3Zh/NNI1x3m53Gvaeh/IOn89DM5xotIOwwi4jFgH9KG0Uo6gvgG5ZbVV4Hv5Ji/xZKaXRnfJlXfHyFtSOd2MvwtpN/YViuZQbrOUraWAh3M51zTPQW4OS/jnUpM7/Ok0z9tdwFdSjq6L6vtVOEzku7qcEi6tqzyDn8v0sXqOaRrUPvnfteR1oH7SNf2ri6MOoBU832SdFfPh0jLGdI5/QeBpyTNz92OIO2wZ5G2p/OB33b6y5sbT7qgflveBq8n1cBLiYhLSMvwfNKF+CtIdxS9BuxNerRgPulOr8+3c0qvrPNJ6+2s/HnLtl9YBtuS1vP5pMTXdnDabPl3tE6dSbpudC9pf3NZJzF+j3Tgu0DSsSV+U+n9UERMAf4/6c62uaQa4gElyuhyWZ1pu2vGepGki4C/RcSJnQ5sZkuRNBv4QkRc39uxmF/T0iskvTdXowdI+ijpaPeK3o7LzGxZ+SnY3vE2UlV5OOk0yFci4u7eDcnMbNn59JeZmVXGp7/MzKwyfeL01zrrrBOjR4/u7TDMzPqU6dOnz4+Ilp4ss08kldGjRzNt2rTeDsPMrE+R1N2n/LvNp7/MzKwyTipmZlYZJxUzM6uMk4qZmVXGScXMzCrTJ+7+smoMGDCA4sOukli8eHEHY5j1DK+b/YdrKiuIxo0W0r9+DhjgVcB6l9fN/sVLbQXR3ut4/Joe621eN/sXJxUzM6uMk4qZmVXGScXMzCrjpGJmZpVxUjEzs8rUmlQkfU3Sg5IekHSBpMGSNpZ0u6SHJF0kaeU6YzAzs55TW1KRtAFwJDAmIt4NDAQOAL4PnBoRmwLPAYfWFYOZmfWsuk9/DQKGSBoErArMBXYDLs39JwH71hyDmZn1kNqSSkQ8AfwImENKJs8D04EFEfFGHuxxYINm40s6TNI0SdNaW1vrCtPMzCpU5+mvtYF9gI2BEcBqwMeaDNr0sdmIOCMixkTEmJaWHv03TDMz66Y6Xyi5B/BIRLQCSLoMeB+wlqRBubayIfBkjTGY2XJIUreG86tbln91XlOZA+wkaVWlNWN3YAZwA7BfHmYccGWNMZjZcigi3vyUHc4JpW+o85rK7aQL8ncB9+eyzgDGA1+X9E9gOHBWXTGYmVnPqvX/VCLiRODEhs6zgB3qLNfM+o6IaHo6zDWTvslP1JtZr2s7vTVq/NU+1dXHOamYmVllnFTMzKwyTipmZlYZJxUzM6uMk4qZmVXGScXMzCrjpGJmZpVxUjEzs8o4qZiZWWWcVMzMrDJOKmZmVhknFTMzq4yTipmZVcZJxczMKuOkYmZmlXFSMTOzyjipmJlZZZxUzMysMk4qZmZWGScVMzOrjJOKmZlVZlBvB2D1kdTl4SKirnDMbAXgpNKPFRNERwnGicTMqlLb6S9Jm0m6p/B5QdLRkoZJuk7SQ/l77bpisCXGjh3bpe5mZt1RW1KJiL9HxLYRsS2wPfAycDkwAZgSEZsCU3K71Wzy5MmMHTv2zRqLJMaOHcvkyZN7OTIz60966kL97sDDEfEosA8wKXefBOzbQzGs8CZPnszixYsZNf5qFi9e7IRiZpXrqaRyAHBBbl4vIuYC5O91m40g6TBJ0yRNa21t7aEwzcxsWdSeVCStDOwNXNKV8SLijIgYExFjWlpa6gnOzMwq1RM1lY8Bd0XEvNw+T9L6APn76R6IwczMekBPJJXPsOTUF8BVwLjcPA64sgdiMDOzHlBrUpG0KvBh4LJC54nAhyU9lPtNrDMGMzPrObU+/BgRLwPDG7o9Q7obzMzM+hm/+8vMzCrjpGJmZpVxUjEzs8o4qZiZWWWcVMzMrDJOKmZmVhknFTMzq4yTipmZVcZJxczMKuOkYmZmlXFSMTOzyjipmJlZZZxUzMysMk4qZmZWGScVMzOrjJOKmZlVxknFzMwq46RiZmaVcVIxM7PKOKmYmVllOk0qkt4paYqkB3L71pK+WX9oZmbW15SpqZwJHA+8DhAR9wEH1BmUmZn1TWWSyqoRcUdDtzfqCMbMzPq2MkllvqS3AwEgaT9gbpmJS1pL0qWS/iZppqSdJQ2TdJ2kh/L32ssQv5mZLUfKJJXDgV8Dm0t6Ajga+ErJ6f8UuCYiNge2AWYCE4ApEbEpMCW3m5lZPzCoswEiYhawh6TVgAERsbDMhCWtAXwQODhP5zXgNUn7ALvmwSYBU4HxXQ3czMyWP50mFUlfb2gHeB6YHhH3dDDqJkAr8DtJ2wDTgaOA9SJiLkBEzJW0bjvlHgYcBjBy5MjOf4mZmfW6Mqe/xgBfBjbIn8NINY0zJR3XwXiDgPcAv4qI7YCX6MKprog4IyLGRMSYlpaWsqOZmVkvKpNUhgPviYhjIuIYUpJpoXBqqx2PA49HxO25/VJSkpknaX2A/P10N2M3M7PlTJmkMhJ4rdD+OjAqIl4BXm1vpIh4CnhM0ma50+7ADOAqYFzuNg64sqtBm5nZ8qnTayrA+cBtktp2/nsBF+QL9zM6GfcI4PeSVgZmAYeQEtnFkg4F5gCf6lbkZma23Clz99d3Jf0J2AUQ8OWImJZ7H9jJuPeQTpc12r2rgZqZ2fKvTE2FiJgmaQ4wGEDSyIiYU2tkZmbW55R5oeTekh4CHgFuzN9/qjswMzPre8pcqP8usBPwj4jYGNgDuLnWqMzMrE8qk1Rej4hngAGSBkTEDcC2NcdlZmZ9UJlrKgskrQ7cRLqT62n8lmIzM2uiTE1lH+Bl4GvANcDDpNuKzczMllLmluKXcuNi0gsglyLp1ojYuerAzMys76niP+oHVzANMzPrB6pIKlHBNMzMrB+oIqmYmZkB1SQVVTANMzPrBzpMKpIGSrq+k2kcVGE8ZmbWh3WYVCJiEfCypDU7GOaByqMyM7M+qczDj/8C7pd0HenfGwGIiCNri8rMzPqkMknlj/ljZmbWoTIPP06SNAQYGRF/74GYzMysjyrz6vu9gHtIr2hB0raSrqo7MDMz63vK3FJ8ErADsADe/DfHjWuMyczM+qgySeWNiHi+oZufojczs7coc6H+AUmfBQZK2hQ4Eril3rDMzKwvKlNTOQLYEngVuAB4ATi6zqDMzKxvKnP318vAf+aPmZlZuzpNKpLGACcAo4vDR8TW9YVlZmZ9UZlrKr8HvgHcT/qjLjMzs6bKJJXWiOjWcymSZgMLgUWku8jGSBoGXESq+cwGPh0Rz3Vn+mZmtnwpk1ROlPQbYArpYj0AEXFZyTL+LSLmF9onAFMiYqKkCbl9fNmAzcxs+VUmqRwCbA6sxJLTXwGUTSqN9gF2zc2TgKk4qZj1S9t8+1qef+X1Lo0zekLXXjW45pCVuPfEsV0ax+pTJqlsExFbdXP6AVwrKYBfR8QZwHoRMRcgIuZKWrfZiJIOAw4DGDlyZDeLN7Pe9PwrrzN74sdrLaOrScjqVSap3CbpXRExoxvT3yUinsyJ4zpJfys7Yk5AZwCMGTPGT/CbmfUBZZLK+4Fxkh4hXVMREGVuKY6IJ/P305IuJ71DbJ6k9XMtZX3g6e6Hb2Zmy5MySeWj3ZmwpNWAARGxMDePBb4DXAWMAybm7yu7M30zM1v+lHmi/lFJ2wAfyJ3+EhH3lpj2esDlktrKOT8irpF0J3CxpEOBOcCnuhe6mZktb8o8UX8U8EWW3O11nqQzIuLnHY0XEbOAbZp0fwbYvRuxmpnZcq7M6a9DgR0j4iUASd8HbgU6TCpmZrbiKfOWYpGeiG+zKHczMzNbSpmayu+A2/PdWwD7Ar+tLyQzM+urylyo/4mkqaRbiwUcEhF31x2YmZn1PWUu1J8bEQcBdzXpZmZm9qYy11S2LLZIGghsX084ZmbWl7WbVCQdL2khsLWkF/JnIekJeD+waGZmb9FuUomI70XEUOCHEbFG/gyNiOERcXwPxmhmZn1EmdNfV+fXrCDpc5J+ImlUzXGZmVkfVCap/Ap4Ob+q5TjgUeCcWqMyM7M+qcxzKm9EREjaB/hpRJwlaVzdgVk53fkTJOjaf1D4T5DMrKwySWWhpOOBzwEfzHd/rVRvWFaW/wTJzJYnZU5/7U/6H5VDI+IpYAPgh7VGZWZmfVKZJ+qfAn5SaJ+Dr6mYmVkTZZ6oX0j6r3mAlUmnvl6MiDXrDMzMzPqeMjWVocV2SfuS/hbYzMxsKWWuqSwlIq4AdqshFjMz6+PKnP76RKF1ADCGJafDzMzM3lTmluK9Cs1vALOBvWuJxszM+rQySWUAcFRELACQtDbwY+A/6gzMzMz6njLXVLZuSygAEfEcsF19IZmZWV9VJqkMyLUTACQNo1wNx8zMVjBlksOPgVskXUq6QP9p4JRaozIzsz6p05pKRJwDfBKYB7QCn4iIc8sWIGmgpLslXZ3bN5Z0u6SHJF0kaeXuBm9mZsuXUs+pRMSMiPhFRPw8ImZ0sYyjgJmF9u8Dp0bEpsBzwKFdnJ6ZmS2nuvzwY1dI2hD4OPCb3C7Sg5OX5kEmAfvWGYOZmfWcWpMKcBrpj70W5/bhwIKIeCO3P0566/FbSDpM0jRJ01pbW2sO08zMqlBbUpH078DTETG92LnJoE2fzo+IMyJiTESMaWlpqSVGMzOrVp23Bu8C7C1pT2AwsAap5rKWpEG5trIh8GSNMZiZWQ+qLalExPHA8QCSdgWOjYgDJV0C7AdcCIwDrqwrBjPrXUO3mMBWkybUXAakS7e2POiNhxjHAxdKOhm4GzirF2Iwsx6wcOZE/931CqZHkkpETAWm5uZZ+P9YzMz6pbrv/jIzsxWIk4qZmVXGScXMzCrjpGJmZpVxUjEzs8o4qZiZWWWcVMzMrDJOKmZmVhknFTMzq4yTipmZVcZJxczMKuOkYmZmlXFSMTOzyjipmJlZZZxUzMysMk4qZmZWGScVMzOrjJOKmZlVxknFzMwq46RiZmaVcVIxM7PKDOrtAGzZDN1iAltNmlBzGQAfr7UMM+sfnFT6uIUzJzJ7Yr07/NET/ljr9M2s//DpLzMzq0xtNRVJg4GbgFVyOZdGxImSNgYuBIYBdwEHRcRrdcVhZr2r7prumkNWqnX61jV1nv56FdgtIl6UtBLwV0l/Ar4OnBoRF0o6HTgU+FWNcZhZL+nqqdnRE/5Y++lcq1dtp78ieTG3rpQ/AewGXJq7TwL2rSsGMzPrWbVeU5E0UNI9wNPAdcDDwIKIeCMP8jiwQTvjHiZpmqRpra2tdYZpZmYVqTWpRMSiiNgW2BDYAdii2WDtjHtGRIyJiDEtLS11hmlmZhXpkbu/ImIBMBXYCVhLUtu1nA2BJ3siBjMzq19tSUVSi6S1cvMQYA9gJnADsF8ebBxwZV0xmJlZz6rz7q/1gUmSBpKS18URcbWkGcCFkk4G7gbOqjEGMzPrQbUllYi4D9iuSfdZpOsrZmbWz/iJejMzq4yTipmZVcZJxczMKuO3FPcDfreSmS0vnFT6uO68J8nvVzKzuvj0l5mZVcZJxczMKuOkYmZmlXFSMTOzyjipmJlZZZxUzMysMk4qZmZWGScVMzOrjJOKmZlVxknFzMwq46RiZmaVcVIxM7PKOKmYmVllnFTMzKwyTipmZlYZJxUzM6uM/6SrH5PUfr/vN+8eETVFY7ZEd9ZN8PrZFzip9GPeAG155XWz/6rt9JekjSTdIGmmpAclHZW7D5N0naSH8vfadcVgZmY9q85rKm8Ax0TEFsBOwOGS3gVMAKZExKbAlNxuZmb9QG1JJSLmRsRduXkhMBPYANgHmJQHmwTsW1cMZmbWs3rk7i9Jo4HtgNuB9SJiLqTEA6zbEzGYmVn9ak8qklYH/gc4OiJe6MJ4h0maJmlaa2trfQGamVllak0qklYiJZTfR8RlufM8Sevn/usDTzcbNyLOiIgxETGmpaWlzjDNzKwidd79JeAsYGZE/KTQ6ypgXG4eB1xZVwxmZtaz6nxOZRfgIOB+SffkbicAE4GLJR0KzAE+VWMMZmbWg9QXHkKS1Ao82ttx9CPrAPN7OwizJrxuVmtURPTo9YM+kVSsWpKmRcSY3o7DrJHXzb7PL5Q0M7PKOKmYmVllnFRWTGf0dgBm7fC62cf5moqZmVXGNRUzM6uMk4qZmVXGf9LVD0haBNxf6LRvRMxuZ9jRwNUR8e76IzMDScNJf3MB8DZgEdD2Qr8dIuK1XgnMauGk0j+8EhHb9nYQZs1ExDPAtgCSTgJejIgfFYfJr3VSRCzu+QitSj791U9JGi3pL5Luyp/3NRlmS0l3SLpH0n2SNs3dP1fo/mtJA3v+F1h/J+kdkh6QdDpwF7CRpAWF/gdI+k1uXk/SZfnN5XdI2qm34raOOan0D0NyArhH0uW529PAhyPiPcD+wM+ajPdl4Ke5ljMGeFzSFnn4XXL3RcCB9f8EW0G9CzgrIrYDnuhguJ8BP8hP238a+E1PBGdd59Nf/UOz018rAb+Q1JYY3tlkvFuB/5S0IXBZRDwkaXdge+DOdEaCIbTz9wRmFXg4Iu4sMdwewGZ5nQRYW9KQiHilvtCsO5xU+q+vAfOAbUg10n81DhAR50u6Hfg4MFnSFwABkyLi+J4M1lZYLxWaF5PWvzaDC83CF/X7BJ/+6r/WBObmC58HAW+5LiJpE2BWRPyM9D83W5Pu0tlP0rp5mGGSRvVc2Laiyuvqc5I2lTQA+H+F3tcDh7e15Bq4LYecVPqv/wbGSbqNdOrrpSbD7A88kP/vZnPgnIiYAXwTuFbSfcB1wPo9FLPZeOAa0sHN44XuhwO75BtKZgBf7I3grHN+TYuZmVXGNRUzM6uMk4qZmVXGScXMzCrjpGJmZpVxUjEzs8o4qdgKQ9LZkvbrZJiDJY3o4nRHS3qgne6fbZj2L7oy7ZLlb55f0XO3pLdLOlLSTEm/r7oss844qZgt7WCgS0mlA6OBz3Y2UAX2Ba6MiO0i4mHgq8CeEVHqnW2S/GYNq4yTivUqSZ/PD7TdK+nc3G2UpCm5+xRJI3P3syX9StINkmZJ+pCk3+aj8rML03xR0o/z25mnSGppUu72km6UNF3SZEnr51rMGOD3+ch/SLPhCuPfK+lWCk96N5gIfCBP62u52whJ10h6SNIPCvGMlXRrjvkSSas3iXlbSbfl+XK5pLUl7QkcDXwhz5fTgU2AqyR9TdJqeR7dmWsy++RpHZzL+QNwbdeWmlkHIsIff3rlA2wJ/B1YJ7cPy99/AMbl5v8ArsjNZwMXkt4DtQ/wArAV6eBoOrBtHi6AA3Pzt4BfFMbfj/SyzVuAltx9f+C3uXkqMCY3dzTcfcCHcvMPgQea/L5dSX+I1tZ+MDCL9AqdwcCjwEbAOsBNwGp5uPHAt5pMr1jmd4DTcvNJwLGF4WYX5ul/AZ/LzWsB/wBWy7E83jbP/fGnqo8SeN98AAAB+UlEQVSrvdabdgMujYj5ABHxbO6+M/CJ3Hwu8IPCOH+IiJB0PzAvIu4HkPQg6XTTPaQXE16Uhz8PuKyh3M2AdwPX5bfeDgTmNomv6XCS1gTWiogbCzF+rORvnhIRz+eYZwCjSDv7dwE353JWJr1B+k1NypwEXFKivLHA3pKOze2DgZG5+brCPDerhJOK9SaRahWdKQ7zav5eXGhua29vfW4sQ8CDEbFzifjeMpyktZpMs6xizItIMYu0g/9MN6fZEQGfjIi/L9VR2pHm74MzWya+pmK9aQrwaaX/MEfSsNz9FuCA3Hwg8NcuTncA6TQXpAvljeP/HWiRtHMudyVJW+Z+C4GhHQ0XEQuA5yW9vxBjM8VpdeQ20ssS35HLWVXSUv9/k2s3z0n6QO50EHAjnZsMHKFcBZK0XYlxzLrNNRXrNRHxoKRTgBslLQLuJp3rPxL4raRvAK3AIV2c9EvAlpKmA8+TroUUy30tX5T/WT6tNAg4DXiQdN3ldEmvkE7DtTfcITnGl0k77mbuA96QdG+e7nPtzIdWSQcDF0haJXf+Jun6R9G4HNuqpGszZebLd3PM9+XEMhv49xLjmXWL31Js/Y6kFyPiLXdPmVn9fPrLzMwq45qKmZlVxjUVMzOrjJOKmZlVxknFzMwq46RiZmaVcVIxM7PK/B+w/LORrc41hAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "final_table[['completed','customer_age']].pivot(columns=\"completed\")['customer_age'].plot.box()\n",
    "plt.xlabel('completed the offer')\n",
    "plt.ylabel('customer_age')\n",
    "plt.title(\"'Customer age' and whether the customer completed the offer\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "offer_type\n",
       "bogo             30499\n",
       "discount         30543\n",
       "informational    15235\n",
       "Name: person, dtype: int64"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "final_table.groupby('offer_type').count()['person']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "offer_id                          offer_type   \n",
       "0b1e1539f2cc45b7b9fa7c272da2e1d7  discount         7668\n",
       "2298d6c36e964ae4a3e7e9706d1fb8c2  discount         7646\n",
       "2906b810c7d4411798c6938adc9daaa5  discount         7632\n",
       "3f207df678b143eea3cee63160fa8bed  informational    7617\n",
       "4d5c57ea9a6940dd891ad53e9dbe8da0  bogo             7593\n",
       "5a8bc65990b245e5a138643cd4eb9837  informational    7618\n",
       "9b98b8c7a33c4b65b9aebfe6a799e6d9  bogo             7677\n",
       "ae264e3637204a6fb9bb56bc8210ddfd  bogo             7658\n",
       "f19421c1d4aa40978ebb69ca19b0e20d  bogo             7571\n",
       "fafdcd668e3743c1bb461111dcafc2a4  discount         7597\n",
       "Name: person, dtype: int64"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "final_table.groupby(['offer_id','offer_type']).count()['person']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "> The distribution among offers is quite balanced."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [],
   "source": [
    "def average_net_revenue_by_group(x):\n",
    "    per_person.groupby(x)['net_revenue'].mean().plot.bar()\n",
    "    plt.title(\"Average Net Revenue\")\n",
    "    plt.xticks(rotation = 0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "average_net_revenue_by_group('gender')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "average_net_revenue_by_group('age_bucket')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'net revenue')"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.scatter(per_person['age'], per_person['net_revenue'])\n",
    "plt.title('Age and net revenue')\n",
    "plt.xlabel('age')\n",
    "plt.ylabel('net revenue')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'net revenue')"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.scatter(per_person['customer_age'], per_person['net_revenue'])\n",
    "plt.title('Customer age and net revenue')\n",
    "plt.xlabel('customer_age')\n",
    "plt.ylabel('net revenue')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'net_revenue')"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.scatter(per_person['total_reward_amount'], per_person['net_revenue'])\n",
    "plt.title(\"Total reward amount and net revenue\")\n",
    "plt.xlabel('total_reward_amount')\n",
    "plt.ylabel('net_revenue')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 4. Prepare data for machine learning models"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [],
   "source": [
    "offer_pipe['time_viewed_to_completed'] = offer_pipe['time_received_to_completed'] - offer_pipe['time_received_to_viewed']\n",
    "offer_pipe['completed'] = offer_pipe['event_y_y'].notna()\n",
    "offer_pipe['effectively_viewed'] = offer_pipe['viewed'] & \\\n",
    "                                               ((offer_pipe['time_viewed_to_completed'] >= 0) | (~offer_pipe['completed']))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "68.79% of the offers were effectively viewed (not viewed after they had completed the offer)\n"
     ]
    }
   ],
   "source": [
    "print(\"{:.2%} of the offers were effectively viewed (not viewed after they had completed the offer)\"\\\n",
    "       .format(sum(offer_pipe['effectively_viewed']) / len(offer_pipe)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [],
   "source": [
    "def numeric_gender(x):\n",
    "    if x is None:\n",
    "        return 0.5\n",
    "    if x == \"F\":\n",
    "        return 0\n",
    "    else:\n",
    "        return 1\n",
    "\n",
    "profile['gender'] = profile['gender'].apply(numeric_gender)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [],
   "source": [
    "offer_pipe_for_modeling = offer_pipe.drop(['rank_in_user', \n",
    "                                           'event', \n",
    "                                           'event_y_x', \n",
    "                                           'time_received_to_viewed', \n",
    "                                           'time_received_to_completed',\n",
    "                                           'time_viewed_to_completed',\n",
    "                                           'viewed',\n",
    "                                           'event_y_y'], axis=1)\\\n",
    "                                     .merge(portfolio[portfolio['offer_type'] != 'informational']\\\n",
    "                                               .rename({'id':'offer_id'}, axis=1).drop('channels', axis=1), on=\"offer_id\")\\\n",
    "                                     .merge(profile[['age', 'gender', 'income', 'customer_age','id']].rename({'id':'person'}, axis=1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "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>person</th>\n",
       "      <th>offer_id</th>\n",
       "      <th>time</th>\n",
       "      <th>completed</th>\n",
       "      <th>effectively_viewed</th>\n",
       "      <th>difficulty</th>\n",
       "      <th>duration</th>\n",
       "      <th>offer_type</th>\n",
       "      <th>reward</th>\n",
       "      <th>web</th>\n",
       "      <th>mobile</th>\n",
       "      <th>email</th>\n",
       "      <th>social</th>\n",
       "      <th>age</th>\n",
       "      <th>gender</th>\n",
       "      <th>income</th>\n",
       "      <th>customer_age</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>78afa995795e4d85b5d9ceeca43f5fef</td>\n",
       "      <td>9b98b8c7a33c4b65b9aebfe6a799e6d9</td>\n",
       "      <td>0</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>5</td>\n",
       "      <td>7</td>\n",
       "      <td>bogo</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>75.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>100000.0</td>\n",
       "      <td>36.500000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>78afa995795e4d85b5d9ceeca43f5fef</td>\n",
       "      <td>f19421c1d4aa40978ebb69ca19b0e20d</td>\n",
       "      <td>504</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>bogo</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>75.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>100000.0</td>\n",
       "      <td>36.500000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>78afa995795e4d85b5d9ceeca43f5fef</td>\n",
       "      <td>ae264e3637204a6fb9bb56bc8210ddfd</td>\n",
       "      <td>408</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>10</td>\n",
       "      <td>7</td>\n",
       "      <td>bogo</td>\n",
       "      <td>10</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>75.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>100000.0</td>\n",
       "      <td>36.500000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>ebe7ef46ea6f4963a7dd49f501b26779</td>\n",
       "      <td>9b98b8c7a33c4b65b9aebfe6a799e6d9</td>\n",
       "      <td>0</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>5</td>\n",
       "      <td>7</td>\n",
       "      <td>bogo</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>59.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>41000.0</td>\n",
       "      <td>64.466667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>ebe7ef46ea6f4963a7dd49f501b26779</td>\n",
       "      <td>9b98b8c7a33c4b65b9aebfe6a799e6d9</td>\n",
       "      <td>576</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>5</td>\n",
       "      <td>7</td>\n",
       "      <td>bogo</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>59.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>41000.0</td>\n",
       "      <td>64.466667</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                             person                          offer_id  time  \\\n",
       "0  78afa995795e4d85b5d9ceeca43f5fef  9b98b8c7a33c4b65b9aebfe6a799e6d9     0   \n",
       "1  78afa995795e4d85b5d9ceeca43f5fef  f19421c1d4aa40978ebb69ca19b0e20d   504   \n",
       "2  78afa995795e4d85b5d9ceeca43f5fef  ae264e3637204a6fb9bb56bc8210ddfd   408   \n",
       "3  ebe7ef46ea6f4963a7dd49f501b26779  9b98b8c7a33c4b65b9aebfe6a799e6d9     0   \n",
       "4  ebe7ef46ea6f4963a7dd49f501b26779  9b98b8c7a33c4b65b9aebfe6a799e6d9   576   \n",
       "\n",
       "   completed  effectively_viewed  difficulty  duration offer_type  reward  \\\n",
       "0       True                True           5         7       bogo       5   \n",
       "1       True               False           5         5       bogo       5   \n",
       "2       True                True          10         7       bogo      10   \n",
       "3      False               False           5         7       bogo       5   \n",
       "4      False                True           5         7       bogo       5   \n",
       "\n",
       "   web  mobile  email  social   age  gender    income  customer_age  \n",
       "0    1       1      1       0  75.0     0.0  100000.0     36.500000  \n",
       "1    1       1      1       1  75.0     0.0  100000.0     36.500000  \n",
       "2    0       1      1       1  75.0     0.0  100000.0     36.500000  \n",
       "3    1       1      1       0  59.0     1.0   41000.0     64.466667  \n",
       "4    1       1      1       0  59.0     1.0   41000.0     64.466667  "
      ]
     },
     "execution_count": 73,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "offer_pipe_for_modeling.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [],
   "source": [
    "offer_pipe_for_modeling.to_csv('offer_pipe_for_modeling.csv', index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "> # Appendix\n",
    "Prove that until the current offer expires, the app won't send another offer under the same type to that customer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [],
   "source": [
    "t1 = transcript[transcript['event'] == \"offer received\"][['person','time','offer_id','rank_in_user']]\\\n",
    ".merge(portfolio[['id','duration']].rename({'id':'offer_id'}, axis=1))\\\n",
    ".merge(transcript[transcript['event'] == \"offer received\"][['person','time','offer_id','rank_in_user']], on=['person','offer_id'])\n",
    "\n",
    "t1['duration_calc'] = t1['time_y'] - t1['time_x'] #duration between 2 offers under the same type\n",
    "\n",
    "t1['expiration'] = t1['time_x' ] + t1['duration'] #expiration time of the first offer \n",
    "\n",
    "t1 = t1[(t1['duration_calc'] > 0) & (t1['expiration'] >= t1['time_y'])] #the offer on the right should be later than the offer on the left \n",
    "#and we add a filter to get the later offer that comes before the earlier offer expires"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "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>person</th>\n",
       "      <th>time_x</th>\n",
       "      <th>offer_id</th>\n",
       "      <th>rank_in_user_x</th>\n",
       "      <th>duration</th>\n",
       "      <th>time_y</th>\n",
       "      <th>rank_in_user_y</th>\n",
       "      <th>duration_calc</th>\n",
       "      <th>expiration</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "Empty DataFrame\n",
       "Columns: [person, time_x, offer_id, rank_in_user_x, duration, time_y, rank_in_user_y, duration_calc, expiration]\n",
       "Index: []"
      ]
     },
     "execution_count": 76,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "t1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}