Starbucks_Capstone_notebook.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Starbucks Capstone Challenge\n",
    "\n",
    "### Introduction\n",
    "\n",
    "This data set contains simulated data that mimics customer behavior on the Starbucks rewards mobile app. Once every few days, Starbucks sends out an offer to users of the mobile app. An offer can be merely an advertisement for a drink or an actual offer such as a discount or BOGO (buy one get one free). Some users might not receive any offer during certain weeks. \n",
    "\n",
    "Not all users receive the same offer, and that is the challenge to solve with this data set.\n",
    "\n",
    "Your task is to combine transaction, demographic and offer data to determine which demographic groups respond best to which offer type. This data set is a simplified version of the real Starbucks app because the underlying simulator only has one product whereas Starbucks actually sells dozens of products.\n",
    "\n",
    "Every offer has a validity period before the offer expires. As an example, a BOGO offer might be valid for only 5 days. You'll see in the data set that informational offers have a validity period even though these ads are merely providing information about a product; for example, if an informational offer has 7 days of validity, you can assume the customer is feeling the influence of the offer for 7 days after receiving the advertisement.\n",
    "\n",
    "You'll be given transactional data showing user purchases made on the app including the timestamp of purchase and the amount of money spent on a purchase. This transactional data also has a record for each offer that a user receives as well as a record for when a user actually views the offer. There are also records for when a user completes an offer. \n",
    "\n",
    "Keep in mind as well that someone using the app might make a purchase through the app without having received an offer or seen an offer.\n",
    "\n",
    "### Example\n",
    "\n",
    "To give an example, a user could receive a discount offer buy 10 dollars get 2 off on Monday. The offer is valid for 10 days from receipt. If the customer accumulates at least 10 dollars in purchases during the validity period, the customer completes the offer.\n",
    "\n",
    "However, there are a few things to watch out for in this data set. Customers do not opt into the offers that they receive; in other words, a user can receive an offer, never actually view the offer, and still complete the offer. For example, a user might receive the \"buy 10 dollars get 2 dollars off offer\", but the user never opens the offer during the 10 day validity period. The customer spends 15 dollars during those ten days. There will be an offer completion record in the data set; however, the customer was not influenced by the offer because the customer never viewed the offer.\n",
    "\n",
    "### Cleaning\n",
    "\n",
    "This makes data cleaning especially important and tricky.\n",
    "\n",
    "You'll also want to take into account that some demographic groups will make purchases even if they don't receive an offer. From a business perspective, if a customer is going to make a 10 dollar purchase without an offer anyway, you wouldn't want to send a buy 10 dollars get 2 dollars off offer. You'll want to try to assess what a certain demographic group will buy when not receiving any offers.\n",
    "\n",
    "### Final Advice\n",
    "\n",
    "Because this is a capstone project, you are free to analyze the data any way you see fit. For example, you could build a machine learning model that predicts how much someone will spend based on demographics and offer type. Or you could build a model that predicts whether or not someone will respond to an offer. Or, you don't need to build a machine learning model at all. You could develop a set of heuristics that determine what offer you should send to each customer (i.e., 75 percent of women customers who were 35 years old responded to offer A vs 40 percent from the same demographic to offer B, so send offer A)."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Data Sets\n",
    "\n",
    "The data is contained in three files:\n",
    "\n",
    "* portfolio.json - containing offer ids and meta data about each offer (duration, type, etc.)\n",
    "* profile.json - demographic data for each customer\n",
    "* transcript.json - records for transactions, offers received, offers viewed, and offers completed\n",
    "\n",
    "Here is the schema and explanation of each variable in the files:\n",
    "\n",
    "**portfolio.json**\n",
    "* id (string) - offer id\n",
    "* offer_type (string) - type of offer ie BOGO, discount, informational\n",
    "* difficulty (int) - minimum required spend to complete an offer\n",
    "* reward (int) - reward given for completing an offer\n",
    "* duration (int) - time for offer to be open, in days\n",
    "* channels (list of strings)\n",
    "\n",
    "**profile.json**\n",
    "* age (int) - age of the customer \n",
    "* became_member_on (int) - date when customer created an app account\n",
    "* gender (str) - gender of the customer (note some entries contain 'O' for other rather than M or F)\n",
    "* id (str) - customer id\n",
    "* income (float) - customer's income\n",
    "\n",
    "**transcript.json**\n",
    "* event (str) - record description (ie transaction, offer received, offer viewed, etc.)\n",
    "* person (str) - customer id\n",
    "* time (int) - time in hours since start of test. The data begins at time t=0\n",
    "* value - (dict of strings) - either an offer id or transaction amount depending on the record\n",
    "\n",
    "**Note:** If you are using the workspace, you will need to go to the terminal and run the command `conda update pandas` before reading in the files. This is because the version of pandas in the workspace cannot read in the transcript.json file correctly, but the newest version of pandas can. You can access the termnal from the orange icon in the top left of this notebook.  \n",
    "\n",
    "You can see how to access the terminal and how the install works using the two images below.  First you need to access the terminal:\n",
    "\n",
    "<img src=\"pic1.png\"/>\n",
    "\n",
    "Then you will want to run the above command:\n",
    "\n",
    "<img src=\"pic2.png\"/>\n",
    "\n",
    "Finally, when you enter back into the notebook (use the jupyter icon again), you should be able to run the below cell without any errors."
   ]
  },
  {
   "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 datetime\n",
    "import pickle\n",
    "import matplotlib.pyplot as plt\n",
    "% matplotlib inline\n",
    "\n",
    "# read in the json files\n",
    "portfolio = pd.read_json('data/portfolio.json', orient='records', lines=True)\n",
    "profile = pd.read_json('data/profile.json', orient='records', lines=True)\n",
    "transcript = pd.read_json('data/transcript.json', orient='records', lines=True)"
   ]
  },
  {
   "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>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": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# explore the profile \n",
    "profile.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(17000, 5)"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "profile.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "age                    0\n",
       "became_member_on       0\n",
       "gender              2175\n",
       "id                     0\n",
       "income              2175\n",
       "dtype: int64"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#check for null values\n",
    "profile.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "age                 2175\n",
      "became_member_on    2175\n",
      "gender                 0\n",
      "id                  2175\n",
      "income                 0\n",
      "dtype: int64\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>gender</th>\n",
       "      <th>income</th>\n",
       "      <th>age</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>None</td>\n",
       "      <td>NaN</td>\n",
       "      <td>118</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>None</td>\n",
       "      <td>NaN</td>\n",
       "      <td>118</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>None</td>\n",
       "      <td>NaN</td>\n",
       "      <td>118</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>None</td>\n",
       "      <td>NaN</td>\n",
       "      <td>118</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>None</td>\n",
       "      <td>NaN</td>\n",
       "      <td>118</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  gender  income  age\n",
       "0   None     NaN  118\n",
       "2   None     NaN  118\n",
       "4   None     NaN  118\n",
       "6   None     NaN  118\n",
       "7   None     NaN  118"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# check count of age=118 value and corresponding gender and income columns\n",
    "print(profile[profile['age']==118].count())\n",
    "profile[['gender','income','age']][profile['age']==118].head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The gender and income have null values. And another thing to notice is whichever values are null in gender are also null in income."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As we can see above, the age = 118 column corresponds with the null gender and income columns. Thus, we can actually drop them during preprocessing if they do not take too large a proportion of our data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "def clean_profile(profile):\n",
    "    '''\n",
    "    data cleaning the profile dataframe\n",
    "    \n",
    "    INPUT:\n",
    "    profile - the profile dataframe to be cleaned\n",
    "    \n",
    "    OUTPUT:\n",
    "    profile - the cleaned profile dataframe\n",
    "    \n",
    "    '''\n",
    "    cleaned_profile = profile.copy()\n",
    "    # age: (numeric) missing value encoded as 118\n",
    "    cleaned_profile['age'] = cleaned_profile['age'].apply(lambda x: np.nan if x == 118 else x)\n",
    "    \n",
    "    # drop all missing values\n",
    "    profile.dropna(inplace=True)\n",
    "    \n",
    "    # the number of days since the user is a memeber of starbucks\n",
    "    cleaned_profile['memberdays'] = datetime.datetime.today().date() - pd.to_datetime(cleaned_profile['became_member_on'], format='%Y%m%d').dt.date\n",
    "    cleaned_profile['memberdays'] = cleaned_profile['memberdays'].dt.days\n",
    "    cleaned_profile.rename(columns={'id':'customer_id'}, inplace=True)\n",
    "    \n",
    "    return cleaned_profile"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "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>age</th>\n",
       "      <th>became_member_on</th>\n",
       "      <th>gender</th>\n",
       "      <th>customer_id</th>\n",
       "      <th>income</th>\n",
       "      <th>memberdays</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>NaN</td>\n",
       "      <td>20170212</td>\n",
       "      <td>None</td>\n",
       "      <td>68be06ca386d4c31939f3a4f0e3dd783</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1192</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>55.0</td>\n",
       "      <td>20170715</td>\n",
       "      <td>F</td>\n",
       "      <td>0610b486422d4921ae7d2bf64640c50b</td>\n",
       "      <td>112000.0</td>\n",
       "      <td>1039</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>NaN</td>\n",
       "      <td>20180712</td>\n",
       "      <td>None</td>\n",
       "      <td>38fe809add3b4fcf9315a9694bb96ff5</td>\n",
       "      <td>NaN</td>\n",
       "      <td>677</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>75.0</td>\n",
       "      <td>20170509</td>\n",
       "      <td>F</td>\n",
       "      <td>78afa995795e4d85b5d9ceeca43f5fef</td>\n",
       "      <td>100000.0</td>\n",
       "      <td>1106</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>NaN</td>\n",
       "      <td>20170804</td>\n",
       "      <td>None</td>\n",
       "      <td>a03223e636434f42ac4c3df47e8bac43</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1019</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    age  became_member_on gender                       customer_id    income  \\\n",
       "0   NaN          20170212   None  68be06ca386d4c31939f3a4f0e3dd783       NaN   \n",
       "1  55.0          20170715      F  0610b486422d4921ae7d2bf64640c50b  112000.0   \n",
       "2   NaN          20180712   None  38fe809add3b4fcf9315a9694bb96ff5       NaN   \n",
       "3  75.0          20170509      F  78afa995795e4d85b5d9ceeca43f5fef  100000.0   \n",
       "4   NaN          20170804   None  a03223e636434f42ac4c3df47e8bac43       NaN   \n",
       "\n",
       "   memberdays  \n",
       "0        1192  \n",
       "1        1039  \n",
       "2         677  \n",
       "3        1106  \n",
       "4        1019  "
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cleaned_profile = clean_profile(profile)\n",
    "cleaned_profile.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "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>age</th>\n",
       "      <th>became_member_on</th>\n",
       "      <th>income</th>\n",
       "      <th>memberdays</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>14825.000000</td>\n",
       "      <td>1.700000e+04</td>\n",
       "      <td>14825.000000</td>\n",
       "      <td>17000.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>54.393524</td>\n",
       "      <td>2.016703e+07</td>\n",
       "      <td>65404.991568</td>\n",
       "      <td>1180.449882</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>17.383705</td>\n",
       "      <td>1.167750e+04</td>\n",
       "      <td>21598.299410</td>\n",
       "      <td>411.223904</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>18.000000</td>\n",
       "      <td>2.013073e+07</td>\n",
       "      <td>30000.000000</td>\n",
       "      <td>663.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>42.000000</td>\n",
       "      <td>2.016053e+07</td>\n",
       "      <td>49000.000000</td>\n",
       "      <td>871.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>55.000000</td>\n",
       "      <td>2.017080e+07</td>\n",
       "      <td>64000.000000</td>\n",
       "      <td>1021.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>66.000000</td>\n",
       "      <td>2.017123e+07</td>\n",
       "      <td>80000.000000</td>\n",
       "      <td>1454.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>101.000000</td>\n",
       "      <td>2.018073e+07</td>\n",
       "      <td>120000.000000</td>\n",
       "      <td>2486.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                age  became_member_on         income    memberdays\n",
       "count  14825.000000      1.700000e+04   14825.000000  17000.000000\n",
       "mean      54.393524      2.016703e+07   65404.991568   1180.449882\n",
       "std       17.383705      1.167750e+04   21598.299410    411.223904\n",
       "min       18.000000      2.013073e+07   30000.000000    663.000000\n",
       "25%       42.000000      2.016053e+07   49000.000000    871.000000\n",
       "50%       55.000000      2.017080e+07   64000.000000   1021.000000\n",
       "75%       66.000000      2.017123e+07   80000.000000   1454.000000\n",
       "max      101.000000      2.018073e+07  120000.000000   2486.000000"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cleaned_profile.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "def display_customer_profile():\n",
    "    '''Display customer profile with histograms'''\n",
    "    \n",
    "    # display histogram of customer age\n",
    "    user_age = cleaned_profile['age'].plot(kind='hist', bins=20, title='Distribution of Customer Age')\n",
    "    user_age.set_xlabel(\"Customer Age\")\n",
    "       \n",
    "    # display histogram of the days being member\n",
    "    plt.figure()\n",
    "    memberdays = cleaned_profile['memberdays'].plot(kind='hist', bins=50, title='How long did the users become our members?')\n",
    "    memberdays.set_xlabel(\"Days\")\n",
    "    \n",
    "    # display histogram of user income\n",
    "    plt.figure()\n",
    "    user_income = cleaned_profile['income'].plot(kind='hist', bins=20, title='Distribution of Customer Income')\n",
    "    user_income.set_xlabel(\"Income\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAY4AAAEWCAYAAABxMXBSAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4wLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvpW3flQAAH0tJREFUeJzt3XuYHFW57/HvjwQINwkhATEJDGAUIkchhssWVDSA3INuQBAlIpjjEa94C8gWtsI2PF5Qt24wQiSAgoggEaMYgYgc5RIuAuEiMcRkSCADgSQQIATe/cdaLZ1Jz0zXTPd0z8zv8zz9dNWqVVXvVDr99qpatUoRgZmZWbU2aHQAZmbWtzhxmJlZIU4cZmZWiBOHmZkV4sRhZmaFOHGYmVkhThzWbZIulPQfNdrW9pKekzQoz8+RdEottp239ztJk2q1vQL7PUfSU5Ke6O19m9WLE4dVJGmhpBckrZL0rKS/SPqEpH99ZiLiExHxjSq3dUBndSJiUURsHhGv1CD2syVd3m77h0TEjJ5uu2Aco4EvAGMj4vUd1HmdpO9JWpQT5/w8P7yH++7ymDeD/APhGUkbNzoWq54Th3XmiIjYAtgBmAp8Bbi41juRNLjW22wSOwBPR8SySgslbQTcCLwFOBh4HfAO4Glgr94Ksp6UVPyekdQCvBMI4MheDMt6KiL88mu9F7AQOKBd2V7Aq8Buef4S4Jw8PRy4HngWWA78mfTD5LK8zgvAc8CXgRbSl8XJwCLglrKywXl7c4BvAncAK4DrgGF52f5Aa6V4SV/Aa4CX8/7+Vra9U/L0BsCZwD+BZcClwJZ5WSmOSTm2p4CvdnKctszrt+XtnZm3f0D+m1/NcVxSYd1TgCeBzTvZfgBvLJvv1jHP9Y8E5uX6c4Bd2x2/LwH3Ac+TfiBsC/wOWAX8EdiqrP4+wF/ytv4G7F+2bA5wLvD/cwxv7OBv+1qu813g+nbLtgZ+A6wE7gTOAW4tW74LMDv/3Y8Axzb6/8xAejU8AL+a80WFxJHLFwH/L0+Xf4l9E7gQ2DC/3gmo0rbKvpwvBTYDNqFy4ngc2C3X+RVweV62Px0kjjx9dqlu2fI5vJY4PgbMB3YCNgeuAS5rF9tPclxvA14q/5Jtt91LSUlti7zu34GTO4qz3bpXAjO6+HfoLHEUOeZvIiWEA3PdL+djsFFZ/dtIyWIkKaHeDewBbAzcBJyV644ktYoOJSWqA/P8iLJjvYjUkhoMbNjB3zYf+CTwdlKi37bdsbkS2BQYCywmJ478eVgMnJS3P46U4N/S6P83A+XlU1VW1BJgWIXyl4HtgB0i4uWI+HPk/+WdODsino+IFzpYfllEPBARzwP/ARxbunjeQycA342IBRHxHHA6cFy7U2b/GREvRMTfSL+o39Z+IzmWDwKnR8SqiFgIfAf4SJVxbA0s7cHfUeSYfxD4bUTMjoiXgW+TEuM7yur8d0Q8GRGPk1ovt0fEPRHxEnAtKYkAfBiYFRGzIuLViJgNzCUlkpJLImJeRKzN+1uHpP1Ip/Kuioi7gH8AH8rLBgH/TkpUqyPiQaD8+tThwMKI+Gne/t2kHxZHV3fYrKecOKyokaTTA+19i/QL8g+SFkiaUsW2FhdY/k/SL+UeXTTO3pC3V77twaRf2yXlvaBWk1om7Q0HNqqwrZFVxvE06Yu/u4oc83X+5oh4lXR8y2N9smz6hQrzpWOwA3BM7jTxrKRngf1Y92/p6t92EvCHiHgqz/88lwGMIP17lG+jfHoHYO92+z8BqNgBwWqvv16UtDqQtCfpi+bW9ssiYhWpB9EXJL0FuFnSnRFxI+l0SyVdtUhGl01vT/qF/RTplMumZXENIn3ZVLvdJaQvn/JtryV9UY7qYt1yT+WYdgAeLNvW41Wu/0fgHEmb5VZVJasp+1tJX46tUPiYLwH+T2lGkkjHt9pYyy0mtQY/3kmdDv8NJG0CHAsMKuumvDEwVNLbgAdI/x6jSKf+YN3PwmLgTxFxYDditxpwi8O6lLuMHk4653x5RNxfoc7hkt6Yv5BWAq/kF6Qv5J26sesPSxoraVPg68DVkbrr/h0YIukwSRuSLkiXd+d8EmjpqDcPcAXweUk7Stoc+C/gFxGxtkhwOZargHMlbSFpB+A04PLO1/yXy0hfgr+StIukDSRtLekMSaXTPvcCH5I0SNLBwLtLKxc85lcBh0makI/ZF0jXbv5S5G/OLgeOkPS+HNcQSftLqjbpHpXjHAvsnl+7kk6PnZiP6zXA2ZI2lbQLcGLZ+tcDb5L0EUkb5teeknbtxt9i3eDEYZ35jaRVpC+3r5J6v5zUQd0xpF/QzwF/Bf4nIubkZd8EzsynFb5YYP+XkS4GPwEMAT4DEBErSBdVLyL9Yn6e/Cs8+2V+f1rS3RW2Oz1v+xbgMeBF4NMF4ir36bz/BaSW2M/z9ruUrx0cADxM6iG0ktSLbDhwe672WeAIUu+lE4Bfl22i6mMeEY+Qrk38N6mldASpu/Waon9wRCwGJgJnkHqTLSb1yKr2+2QS8NNI9+48UXoBPwROyNeaPkXqsfYE6d/qClKiK7W0DgKOI7WkngDOY90fD1ZHpR4YZmZNS9J5wOsjotfv/rf1ucVhZk0nn7p7a76BcC/SPT/XNjouS3xx3Mya0Rak01NvIN1T8h3S/TLWBHyqyszMCvGpKjMzK6RfnqoaPnx4tLS0NDoMM7M+5a677noqIkZ0Va9fJo6Wlhbmzp3b6DDMzPoUSf/supZPVZmZWUFOHGZmVogTh5mZFeLEYWZmhThxmJlZIU4cZmZWiBOHmZkV4sRhZmaFOHGYmVkh/fLOcbN6a5ny226vu3DqYTWMxKz3ucVhZmaFOHGYmVkhThxmZlaIr3HYgNSTaxRmA13dWhySpktaJumBCsu+KCkkDc/zkvQDSfMl3SdpXFndSZIezS8/qN7MrMHqearqEuDg9oWSRgMHAovKig8BxuTXZOCCXHcYcBawN7AXcJakreoYs5mZdaFuiSMibgGWV1h0PvBloPxh5xOBSyO5DRgqaTvgfcDsiFgeEc8As6mQjMzMrPf06sVxSUcCj0fE39otGgksLptvzWUdlVfa9mRJcyXNbWtrq2HUZmZWrtcSh6RNga8CX6u0uEJZdFK+fmHEtIgYHxHjR4zo8pG5ZmbWTb3Z4tgZ2BH4m6SFwCjgbkmvJ7UkRpfVHQUs6aTczMwapNcSR0TcHxHbRERLRLSQksK4iHgCmAmcmHtX7QOsiIilwA3AQZK2yhfFD8plZmbWIPXsjnsF8FfgzZJaJZ3cSfVZwAJgPvAT4JMAEbEc+AZwZ359PZeZmVmD1O0GwIg4vovlLWXTAZzaQb3pwPSaBmdmZt3mIUfMzKwQJw4zMyvEicPMzApx4jAzs0KcOMzMrBAnDjMzK8SJw8zMCnHiMDOzQpw4zMysED861qyX9eSxtQunHlbDSMy6xy0OMzMrxInDzMwKceIwM7NCnDjMzKwQJw4zMyvEvaqsz+pJ7yQz6z63OMzMrBAnDjMzK8SJw8zMCnHiMDOzQuqWOCRNl7RM0gNlZd+S9LCk+yRdK2lo2bLTJc2X9Iik95WVH5zL5kuaUq94zcysOvVscVwCHNyubDawW0S8Ffg7cDqApLHAccBb8jr/I2mQpEHAj4BDgLHA8bmumZk1SN0SR0TcAixvV/aHiFibZ28DRuXpicCVEfFSRDwGzAf2yq/5EbEgItYAV+a6ZmbWII28xvEx4Hd5eiSwuGxZay7rqHw9kiZLmitpbltbWx3CNTMzaNANgJK+CqwFflYqqlAtqJzYotI2I2IaMA1g/PjxFeuY9XUekt2aQa8nDkmTgMOBCRFR+oJvBUaXVRsFLMnTHZWbmVkD9OqpKkkHA18BjoyI1WWLZgLHSdpY0o7AGOAO4E5gjKQdJW1EuoA+szdjNjOzddWtxSHpCmB/YLikVuAsUi+qjYHZkgBui4hPRMQ8SVcBD5JOYZ0aEa/k7XwKuAEYBEyPiHn1itnMzLpWt8QREcdXKL64k/rnAudWKJ8FzKphaGZm1gO+c9zMzApx4jAzs0KcOMzMrBAnDjMzK8SJw8zMCnHiMDOzQpw4zMysECcOMzMrxInDzMwKceIwM7NCnDjMzKyQhjyPw6ykJ8+XMLPGcIvDzMwKceIwM7NCnDjMzKwQJw4zMyvEicPMzApx4jAzs0KcOMzMrJC6JQ5J0yUtk/RAWdkwSbMlPZrft8rlkvQDSfMl3SdpXNk6k3L9RyVNqle8ZmZWnXq2OC4BDm5XNgW4MSLGADfmeYBDgDH5NRm4AFKiAc4C9gb2As4qJRszM2uMuiWOiLgFWN6ueCIwI0/PAI4qK780ktuAoZK2A94HzI6I5RHxDDCb9ZORmZn1ot6+xrFtRCwFyO/b5PKRwOKyeq25rKPy9UiaLGmupLltbW01D9zMzJJmuTiuCmXRSfn6hRHTImJ8RIwfMWJETYMzM7PX9HbieDKfgiK/L8vlrcDosnqjgCWdlJuZWYP09ui4M4FJwNT8fl1Z+ackXUm6EL4iIpZKugH4r7IL4gcBp/dyzGb9Qk9HIl449bAaRWJ9Xd0Sh6QrgP2B4ZJaSb2jpgJXSToZWAQck6vPAg4F5gOrgZMAImK5pG8Ad+Z6X4+I9hfczcysF9UtcUTE8R0smlChbgCndrCd6cD0GoZmZmY90CwXx83MrI9w4jAzs0KcOMzMrJCqEoek3eodiJmZ9Q3VtjgulHSHpE9KGlrXiMzMrKlVlTgiYj/gBNLNeHMl/VzSgXWNzMzMmlLV1zgi4lHgTOArwLuBH0h6WNIH6hWcmZk1n6ru45D0VtJNeYeRRqg9IiLulvQG4K/ANfUL0cyaQU/uPPdd5/1LtTcA/hD4CXBGRLxQKoyIJZLOrEtkZmbWlKpNHIcCL0TEKwCSNgCGRMTqiLisbtGZmVnTqfYaxx+BTcrmN81lZmY2wFSbOIZExHOlmTy9aX1CMjOzZlZt4nhe0rjSjKS3Ay90Ut/MzPqpaq9xfA74paTSQ5S2Az5Yn5DMzKyZVZU4IuJOSbsAbyY9zvXhiHi5rpGZmVlTKvI8jj2BlrzOHpKIiEvrEpWZmTWtam8AvAzYGbgXeCUXB+DEYWY2wFTb4hgPjM1P6jMzswGs2l5VDwCvr2cgZmbWN1Tb4hgOPCjpDuClUmFEHNmdnUr6PHAK6XTX/aRxsLYDrgSGAXcDH4mINZI2Jp0SezvwNPDBiFjYnf2amVnPVZs4zq7VDiWNBD5DOvX1gqSrgONIw5qcHxFXSroQOBm4IL8/ExFvlHQccB7uCmxm1jDVPo/jT8BCYMM8fSepVdBdg4FNJA0m3YG+FHgvcHVePgM4Kk9PzPPk5RMkqQf7NjOzHqi2V9XHgcmk00g7AyOBC4EJRXcYEY9L+jawiHT3+R+Au4BnI2Jtrtaa90F+X5zXXStpBbA18FS7GCfnGNl+++2LhmU90JPhts2s76n24vipwL7ASvjXQ5226c4OJW1FakXsCLwB2Aw4pELVUg+uSq2L9Xp3RcS0iBgfEeNHjBjRndDMzKwK1SaOlyJiTWkmn2LqbtfcA4DHIqIt331+DfAOYGjeLsAooDS8SSvpkbWl/W4JLO/mvs3MrIeqTRx/knQG6brEgcAvgd90c5+LgH0kbZqvVUwAHgRuBo7OdSYB1+XpmXmevPwm309iZtY41SaOKUAbqevs/wVmkZ4/XlhE3E66yH133t4GwDTSs8xPkzSfdA3j4rzKxcDWufy0HIuZmTVItYMcvkp6dOxParHTiDgLOKtd8QJgrwp1XwSOqcV+zcys56rtVfUYlS9I71TziMzMrKkVGauqZAipBTCs9uGYmVmzq/YGwKfLXo9HxPdIN+yZmdkAU+2pqnFlsxuQWiBb1CUiMzNratWeqvpO2fRa0vAjx9Y8GjMza3rV9qp6T70DMTOzvqHaU1WndbY8Ir5bm3DMzKzZFelVtSfpLm6AI4BbyIMPmpnZwFHkQU7jImIVgKSzgV9GxCn1CszMzJpTtUOObA+sKZtfA7TUPBozM2t61bY4LgPukHQt6Q7y95Me52pmZgNMtb2qzpX0O+CdueikiLinfmGZmVmzqvZUFaRHvK6MiO8DrZJ2rFNMZmbWxKpKHJLOIg17fnou2hC4vF5BmZlZ86r2Gsf7gT1Iz9AgIpZI8pAj/YifG25m1ar2VNWa/NS9AJC0Wf1CMjOzZlZt4rhK0o9JzwX/OPBHavRQJzMz61uq7VX17fys8ZXAm4GvRcTsukZmZmZNqcvEIWkQcENEHAA4WZiZDXBdnqqKiFeA1ZK27IV4zMysyVXbq+pF4H5Js4HnS4UR8Znu7FTSUOAiYDfSBfePAY8AvyANZbIQODYinpEk4PvAocBq4KMRcXd39mtmZj1XbeL4bX7VyveB30fE0ZI2It1ceAZwY0RMlTQFmEK6d+QQYEx+7Q1ckN/rpiddUxdOPayGkZiZNZ9OE4ek7SNiUUTMqNUOJb0OeBfwUYCIWAOskTQR2D9XmwHMISWOicCluTvwbZKGStouIpbWKiYzM6teV9c4fl2akPSrGu1zJ6AN+KmkeyRdlO8L2baUDPL7Nrn+SNZ97kdrLluHpMmS5kqa29bWVqNQzcysva4Sh8qmd6rRPgcD44ALImIP0jWTKVXGUBLrFURMi4jxETF+xIgRtYnUzMzW01XiiA6me6IVaI2I2/P81aRE8qSk7QDy+7Ky+qPL1h8FLKlRLGZmVlBXieNtklZKWgW8NU+vlLRK0sru7DAingAWS3pzLpoAPEh6LO2kXDYJuC5PzwROVLIPsMLXN8zMGqfTi+MRMahO+/008LPco2oBcBIpiV0l6WRgEXBMrjuL1BV3Pqk77kl1isnMzKpQbXfcmoqIe4HxFRZNqFA3gFPrHpSZmVWlIYmjP/M9IGbW3zlx9BN+noaZ9ZYij441MzNzi6OZuNVgZn2BWxxmZlaIE4eZmRXixGFmZoX4GoeZ1Z27qfcvbnGYmVkhThxmZlaIE4eZmRXixGFmZoU4cZiZWSFOHGZmVogTh5mZFeLEYWZmhThxmJlZIU4cZmZWiBOHmZkV0rDEIWmQpHskXZ/nd5R0u6RHJf1C0ka5fOM8Pz8vb2lUzGZm1tgWx2eBh8rmzwPOj4gxwDPAybn8ZOCZiHgjcH6uZ2ZmDdKQ0XEljQIOA84FTpMk4L3Ah3KVGcDZwAXAxDwNcDXwQ0mKiOjNmM2sMTyybvNpVIvje8CXgVfz/NbAsxGxNs+3AiPz9EhgMUBeviLXX4ekyZLmSprb1tZWz9jNzAa0Xk8ckg4HlkXEXeXFFapGFcteK4iYFhHjI2L8iBEjahCpmZlV0ohTVfsCR0o6FBgCvI7UAhkqaXBuVYwCluT6rcBooFXSYGBLYHnvh21mZtCAFkdEnB4RoyKiBTgOuCkiTgBuBo7O1SYB1+XpmXmevPwmX98wM2ucZrqP4yukC+XzSdcwLs7lFwNb5/LTgCkNis/MzGjwM8cjYg4wJ08vAPaqUOdF4JheDczMzDrUTC0OMzPrA5w4zMysECcOMzMrxInDzMwKceIwM7NCnDjMzKwQJw4zMyvEicPMzApx4jAzs0KcOMzMrBAnDjMzK8SJw8zMCnHiMDOzQpw4zMysECcOMzMrpKHP4zAzq6eWKb/t9roLpx5Ww0j6F7c4zMysECcOMzMrxInDzMwK6fXEIWm0pJslPSRpnqTP5vJhkmZLejS/b5XLJekHkuZLuk/SuN6O2czMXtOIFsda4AsRsSuwD3CqpLHAFODGiBgD3JjnAQ4BxuTXZOCC3g/ZzMxKej1xRMTSiLg7T68CHgJGAhOBGbnaDOCoPD0RuDSS24Chkrbr5bDNzCxr6DUOSS3AHsDtwLYRsRRScgG2ydVGAovLVmvNZe23NVnSXElz29ra6hm2mdmA1rDEIWlz4FfA5yJiZWdVK5TFegUR0yJifESMHzFiRK3CNDOzdhpyA6CkDUlJ42cRcU0uflLSdhGxNJ+KWpbLW4HRZauPApb0XrRmNhD15OZB6N83EDaiV5WAi4GHIuK7ZYtmApPy9CTgurLyE3Pvqn2AFaVTWmZm1vsa0eLYF/gIcL+ke3PZGcBU4CpJJwOLgGPyslnAocB8YDVwUu+Ga2Zm5Xo9cUTErVS+bgEwoUL9AE6ta1BmZlY13zluZmaFOHGYmVkhThxmZlaIE4eZmRXixGFmZoU4cZiZWSFOHGZmVogTh5mZFeLEYWZmhThxmJlZIQ0ZHdfMrL/ryei6zT6yrlscZmZWiBOHmZkV4sRhZmaFOHGYmVkhThxmZlaIE4eZmRXi7rhmZk2m2bvyusVhZmaFOHGYmVkhfSZxSDpY0iOS5kua0uh4zMwGqj6ROCQNAn4EHAKMBY6XNLaxUZmZDUx9InEAewHzI2JBRKwBrgQmNjgmM7MBqa/0qhoJLC6bbwX2Lq8gaTIwOc8+J+mROsUyHHiqTtvuL3yMquPj1DUfo66tc4x0Xo+2tUM1lfpK4lCFslhnJmIaMK3ugUhzI2J8vffTl/kYVcfHqWs+Rl1rxDHqK6eqWoHRZfOjgCUNisXMbEDrK4njTmCMpB0lbQQcB8xscExmZgNSnzhVFRFrJX0KuAEYBEyPiHkNCqfup8P6AR+j6vg4dc3HqGu9fowUEV3XMjMzy/rKqSozM2sSThxmZlaIE0cHJI2WdLOkhyTNk/TZXD5M0mxJj+b3rRodazOQNEjSPZKuz/M7Sro9H6df5E4NA5akoZKulvRw/kz9mz9L65L0+fx/7QFJV0ga4s8RSJouaZmkB8rKKn52lPwgD810n6Rx9YjJiaNja4EvRMSuwD7AqXmYkynAjRExBrgxzxt8FniobP484Px8nJ4BTm5IVM3j+8DvI2IX4G2kY+XPUiZpJPAZYHxE7EbqBHMc/hwBXAIc3K6so8/OIcCY/JoMXFCPgJw4OhARSyPi7jy9ivQffSRpqJMZudoM4KjGRNg8JI0CDgMuyvMC3gtcnasM6OMk6XXAu4CLASJiTUQ8iz9L7Q0GNpE0GNgUWIo/R0TELcDydsUdfXYmApdGchswVNJ2tY7JiaMKklqAPYDbgW0jYimk5AJs07jImsb3gC8Dr+b5rYFnI2Jtnm8lJd2BaiegDfhpPp13kaTN8GfpXyLiceDbwCJSwlgB3IU/Rx3p6LNTaXimmh8zJ44uSNoc+BXwuYhY2eh4mo2kw4FlEXFXeXGFqgO53/dgYBxwQUTsATzPAD4tVUk+Rz8R2BF4A7AZ6bRLewP5c1SNXvm/58TRCUkbkpLGzyLimlz8ZKnpl9+XNSq+JrEvcKSkhaRRi99LaoEMzaccwEPEtAKtEXF7nr+alEj8WXrNAcBjEdEWES8D1wDvwJ+jjnT02emV4ZmcODqQz9NfDDwUEd8tWzQTmJSnJwHX9XZszSQiTo+IURHRQrqYeVNEnADcDBydqw3o4xQRTwCLJb05F00AHsSfpXKLgH0kbZr/75WOkT9HlXX02ZkJnJh7V+0DrCid0qol3zneAUn7AX8G7ue1c/dnkK5zXAVsT/qwHxMR7S9cDUiS9ge+GBGHS9qJ1AIZBtwDfDgiXmpkfI0kaXdS54GNgAXASaQfbv4sZZL+E/ggqUfjPcAppPPzA/pzJOkKYH/S8OlPAmcBv6bCZycn3R+SemGtBk6KiLk1j8mJw8zMivCpKjMzK8SJw8zMCnHiMDOzQpw4zMysECcOMzMrxInDBgRJr5d0paR/SHpQ0ixJb+rGdo7Kg102lKTBkp6S9M1Gx2IDjxOH9Xu5b/u1wJyI2DkixpLuydm2G5s7CujVxCFpUIXig4BHgGPz32fWa5w4bCB4D/ByRFxYKoiIeyPiz5L2Lz1DBEDSDyV9NE9Pza2T+yR9W9I7gCOBb0m6V9LOknaXdFuuc23ZcxHmSDpf0i35+Rt7SromPz/hnLL9fVjSHXl7Py4lCUnPSfq6pNuBf6vwNx1PGqp9EWnY/9L2Ds3P/Lg1P5eh9HyUzfJzHe7MAy1OrNnRtQHHicMGgt1II61WTdIw4P3AWyLircA5EfEX0pAOX4qI3SPiH8ClwFdynftJd/WWrImIdwEXkoaEODXH8lFJW0valXSn9L4RsTvwCnBCXncz4IGI2Dsibm0X2yakITmuB64gJREkDQF+DBwSEfsBI8pW+yppOJg9SYn0W3mEXrPCnDjMKlsJvAhcJOkDpOEb1iFpS2BoRPwpF80gPXejZGZ+vx+Yl5/x8hJpyJHRpC//twN3Sro3z++U13mFNMBmJYcDN0fE6lzn/bmlsguwICIey/WuKFvnIGBK3s8cYAhpuAqzwgZ3XcWsz5vHawPltbeWdX9ADQGIiLWS9iJ9mR8HfIo08m8RpTGVXi2bLs0PJg2BPSMiTq+w7osR8UoH2z0e2DePSAzp+SfvAZ7uJBYB/x4Rj1QZu1mH3OKwgeAmYGNJHy8V5GsO7wb+CYyVtHFuQUzIyzcHtoyIWcDngN3zqquALQAiYgXwjKR35mUfAUqtj2rcCBwtaZu8z2GSduhshfw0wf2A7SOiJY9KfCopmTwM7JQfPAbpNFjJDcCnSxfSJe1RIE6zdThxWL8XaSTP9wMH5u6484CzgSURsZg0yuh9wM9II7BCSg7XS7qPlAw+n8uvBL6ULzDvTBrS+lu53u7A1wvE9SBwJvCHvP5soKvHfH6AdK2ivAVzHemi/avAJ4HfS7qVNJLqilznG8CGwH2SHsjzZt3i0XHN+hFJm0fEc7ll8SPg0Yg4v9FxWf/iFodZ//LxfAF8HrAlqZeVWU25xWFmZoW4xWFmZoU4cZiZWSFOHGZmVogTh5mZFeLEYWZmhfwvPakQWertkSkAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f1646969550>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f1646969668>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f16467c5860>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# visulize customer profile data\n",
    "display_customer_profile()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "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": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# explore the portfolio\n",
    "portfolio"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "channels      0\n",
       "difficulty    0\n",
       "duration      0\n",
       "id            0\n",
       "offer_type    0\n",
       "reward        0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#check for null values\n",
    "portfolio.isnull().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There are no missing values in this dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "10"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#check no. of unique offers\n",
    "portfolio.id.nunique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "offer_type\n",
       "bogo             4\n",
       "discount         4\n",
       "informational    2\n",
       "Name: id, dtype: int64"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "portfolio.groupby('offer_type')['id'].count()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There are 10 unique offer ids we will have to take note of, with 4 each of bogo and discount types, while 2 informational types."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "def clean_portfolio(portfolio):\n",
    "    '''\n",
    "    data cleaning the portfolio dataframe\n",
    "    \n",
    "    INPUT:\n",
    "    portfolio - the portfolio dataframe to be cleaned\n",
    "    \n",
    "    OUTPUT:\n",
    "    portfolio - the cleaned portfolio dataframe\n",
    "    \n",
    "    '''\n",
    "    # change the duration from day to hour\n",
    "    cleaned_portfolio = portfolio.copy()\n",
    "    cleaned_portfolio['duration'] = cleaned_portfolio['duration'] * 24\n",
    "    \n",
    "    # apply one hot encoding to channels column\n",
    "    cleaned_portfolio['web'] = cleaned_portfolio['channels'].apply(lambda x: 1 if 'web' in x else 0)\n",
    "    cleaned_portfolio['email'] = cleaned_portfolio['channels'].apply(lambda x: 1 if 'email' in x else 0)\n",
    "    cleaned_portfolio['mobile'] = cleaned_portfolio['channels'].apply(lambda x: 1 if 'mobile' in x else 0)\n",
    "    cleaned_portfolio['social'] = cleaned_portfolio['channels'].apply(lambda x: 1 if 'social' in x else 0)\n",
    "    \n",
    "    # apply one hot encoding to offer_type column\n",
    "    offer_type = pd.get_dummies(cleaned_portfolio['offer_type'])\n",
    "    \n",
    "    # drop the channels and offer_type column\n",
    "    cleaned_portfolio.drop(['channels', 'offer_type'], axis=1, inplace=True)\n",
    "    \n",
    "    # combine the portfolio and offer_type dataframe to form a cleaned dataframe\n",
    "    cleaned_portfolio = pd.concat([cleaned_portfolio, offer_type], axis=1, sort=False)\n",
    "    cleaned_portfolio.rename(columns={'id':'offer_id'}, inplace=True)\n",
    "    \n",
    "    return cleaned_portfolio"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "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>difficulty</th>\n",
       "      <th>duration</th>\n",
       "      <th>offer_id</th>\n",
       "      <th>reward</th>\n",
       "      <th>web</th>\n",
       "      <th>email</th>\n",
       "      <th>mobile</th>\n",
       "      <th>social</th>\n",
       "      <th>bogo</th>\n",
       "      <th>discount</th>\n",
       "      <th>informational</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>10</td>\n",
       "      <td>168</td>\n",
       "      <td>ae264e3637204a6fb9bb56bc8210ddfd</td>\n",
       "      <td>10</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>10</td>\n",
       "      <td>120</td>\n",
       "      <td>4d5c57ea9a6940dd891ad53e9dbe8da0</td>\n",
       "      <td>10</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>96</td>\n",
       "      <td>3f207df678b143eea3cee63160fa8bed</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>5</td>\n",
       "      <td>168</td>\n",
       "      <td>9b98b8c7a33c4b65b9aebfe6a799e6d9</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>20</td>\n",
       "      <td>240</td>\n",
       "      <td>0b1e1539f2cc45b7b9fa7c272da2e1d7</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>7</td>\n",
       "      <td>168</td>\n",
       "      <td>2298d6c36e964ae4a3e7e9706d1fb8c2</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>10</td>\n",
       "      <td>240</td>\n",
       "      <td>fafdcd668e3743c1bb461111dcafc2a4</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0</td>\n",
       "      <td>72</td>\n",
       "      <td>5a8bc65990b245e5a138643cd4eb9837</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>5</td>\n",
       "      <td>120</td>\n",
       "      <td>f19421c1d4aa40978ebb69ca19b0e20d</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>10</td>\n",
       "      <td>168</td>\n",
       "      <td>2906b810c7d4411798c6938adc9daaa5</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   difficulty  duration                          offer_id  reward  web  email  \\\n",
       "0          10       168  ae264e3637204a6fb9bb56bc8210ddfd      10    0      1   \n",
       "1          10       120  4d5c57ea9a6940dd891ad53e9dbe8da0      10    1      1   \n",
       "2           0        96  3f207df678b143eea3cee63160fa8bed       0    1      1   \n",
       "3           5       168  9b98b8c7a33c4b65b9aebfe6a799e6d9       5    1      1   \n",
       "4          20       240  0b1e1539f2cc45b7b9fa7c272da2e1d7       5    1      1   \n",
       "5           7       168  2298d6c36e964ae4a3e7e9706d1fb8c2       3    1      1   \n",
       "6          10       240  fafdcd668e3743c1bb461111dcafc2a4       2    1      1   \n",
       "7           0        72  5a8bc65990b245e5a138643cd4eb9837       0    0      1   \n",
       "8           5       120  f19421c1d4aa40978ebb69ca19b0e20d       5    1      1   \n",
       "9          10       168  2906b810c7d4411798c6938adc9daaa5       2    1      1   \n",
       "\n",
       "   mobile  social  bogo  discount  informational  \n",
       "0       1       1     1         0              0  \n",
       "1       1       1     1         0              0  \n",
       "2       1       0     0         0              1  \n",
       "3       1       0     1         0              0  \n",
       "4       0       0     0         1              0  \n",
       "5       1       1     0         1              0  \n",
       "6       1       1     0         1              0  \n",
       "7       1       1     0         0              1  \n",
       "8       1       1     1         0              0  \n",
       "9       1       0     0         1              0  "
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cleaned_portfolio = clean_portfolio(portfolio)\n",
    "cleaned_portfolio"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "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>event</th>\n",
       "      <th>person</th>\n",
       "      <th>time</th>\n",
       "      <th>value</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>offer received</td>\n",
       "      <td>78afa995795e4d85b5d9ceeca43f5fef</td>\n",
       "      <td>0</td>\n",
       "      <td>{'offer id': '9b98b8c7a33c4b65b9aebfe6a799e6d9'}</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>offer received</td>\n",
       "      <td>a03223e636434f42ac4c3df47e8bac43</td>\n",
       "      <td>0</td>\n",
       "      <td>{'offer id': '0b1e1539f2cc45b7b9fa7c272da2e1d7'}</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>offer received</td>\n",
       "      <td>e2127556f4f64592b11af22de27a7932</td>\n",
       "      <td>0</td>\n",
       "      <td>{'offer id': '2906b810c7d4411798c6938adc9daaa5'}</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>offer received</td>\n",
       "      <td>8ec6ce2a7e7949b1bf142def7d0e0586</td>\n",
       "      <td>0</td>\n",
       "      <td>{'offer id': 'fafdcd668e3743c1bb461111dcafc2a4'}</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>offer received</td>\n",
       "      <td>68617ca6246f4fbc85e91a2a49552598</td>\n",
       "      <td>0</td>\n",
       "      <td>{'offer id': '4d5c57ea9a6940dd891ad53e9dbe8da0'}</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            event                            person  time  \\\n",
       "0  offer received  78afa995795e4d85b5d9ceeca43f5fef     0   \n",
       "1  offer received  a03223e636434f42ac4c3df47e8bac43     0   \n",
       "2  offer received  e2127556f4f64592b11af22de27a7932     0   \n",
       "3  offer received  8ec6ce2a7e7949b1bf142def7d0e0586     0   \n",
       "4  offer received  68617ca6246f4fbc85e91a2a49552598     0   \n",
       "\n",
       "                                              value  \n",
       "0  {'offer id': '9b98b8c7a33c4b65b9aebfe6a799e6d9'}  \n",
       "1  {'offer id': '0b1e1539f2cc45b7b9fa7c272da2e1d7'}  \n",
       "2  {'offer id': '2906b810c7d4411798c6938adc9daaa5'}  \n",
       "3  {'offer id': 'fafdcd668e3743c1bb461111dcafc2a4'}  \n",
       "4  {'offer id': '4d5c57ea9a6940dd891ad53e9dbe8da0'}  "
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# explore transcript\n",
    "transcript.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['offer received', 'offer viewed', 'transaction', 'offer completed'], dtype=object)"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "transcript['event'].unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "17000"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#check number of unique people represented\n",
    "transcript['person'].nunique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "event     0\n",
       "person    0\n",
       "time      0\n",
       "value     0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#check for null values\n",
    "transcript.isnull().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There are no null values in this data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "def clean_transcript(transcript):\n",
    "    '''\n",
    "    data cleaning the transcript dataframe, return usable dataframe need contain offer events\n",
    "    \n",
    "    INPUT:\n",
    "    transcript - the transcript dataframe to be cleaned\n",
    "    \n",
    "    OUTPUT:\n",
    "    offer - the cleaned offer dataframe\n",
    "    \n",
    "    '''\n",
    "    offer = transcript.copy()\n",
    "    # extract the rows that are related to offer action \n",
    "    offer = offer[offer['value'].apply(lambda x: True if ('offer id' in x) or ('offer_id' in x) else False)]\n",
    "    \n",
    "    # extract the offer id from value column\n",
    "    offer['offer_id'] = offer['value'].apply(lambda x: x['offer id'] if ('offer id' in x) else x['offer_id'])\n",
    "    \n",
    "    offer.rename(columns={'person':'customer_id'}, inplace=True)\n",
    "    \n",
    "    return offer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "offer = clean_transcript(transcript)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "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>event</th>\n",
       "      <th>customer_id</th>\n",
       "      <th>time</th>\n",
       "      <th>value</th>\n",
       "      <th>offer_id</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>offer received</td>\n",
       "      <td>78afa995795e4d85b5d9ceeca43f5fef</td>\n",
       "      <td>0</td>\n",
       "      <td>{'offer id': '9b98b8c7a33c4b65b9aebfe6a799e6d9'}</td>\n",
       "      <td>9b98b8c7a33c4b65b9aebfe6a799e6d9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>offer received</td>\n",
       "      <td>a03223e636434f42ac4c3df47e8bac43</td>\n",
       "      <td>0</td>\n",
       "      <td>{'offer id': '0b1e1539f2cc45b7b9fa7c272da2e1d7'}</td>\n",
       "      <td>0b1e1539f2cc45b7b9fa7c272da2e1d7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>offer received</td>\n",
       "      <td>e2127556f4f64592b11af22de27a7932</td>\n",
       "      <td>0</td>\n",
       "      <td>{'offer id': '2906b810c7d4411798c6938adc9daaa5'}</td>\n",
       "      <td>2906b810c7d4411798c6938adc9daaa5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>offer received</td>\n",
       "      <td>8ec6ce2a7e7949b1bf142def7d0e0586</td>\n",
       "      <td>0</td>\n",
       "      <td>{'offer id': 'fafdcd668e3743c1bb461111dcafc2a4'}</td>\n",
       "      <td>fafdcd668e3743c1bb461111dcafc2a4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>offer received</td>\n",
       "      <td>68617ca6246f4fbc85e91a2a49552598</td>\n",
       "      <td>0</td>\n",
       "      <td>{'offer id': '4d5c57ea9a6940dd891ad53e9dbe8da0'}</td>\n",
       "      <td>4d5c57ea9a6940dd891ad53e9dbe8da0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            event                       customer_id  time  \\\n",
       "0  offer received  78afa995795e4d85b5d9ceeca43f5fef     0   \n",
       "1  offer received  a03223e636434f42ac4c3df47e8bac43     0   \n",
       "2  offer received  e2127556f4f64592b11af22de27a7932     0   \n",
       "3  offer received  8ec6ce2a7e7949b1bf142def7d0e0586     0   \n",
       "4  offer received  68617ca6246f4fbc85e91a2a49552598     0   \n",
       "\n",
       "                                              value  \\\n",
       "0  {'offer id': '9b98b8c7a33c4b65b9aebfe6a799e6d9'}   \n",
       "1  {'offer id': '0b1e1539f2cc45b7b9fa7c272da2e1d7'}   \n",
       "2  {'offer id': '2906b810c7d4411798c6938adc9daaa5'}   \n",
       "3  {'offer id': 'fafdcd668e3743c1bb461111dcafc2a4'}   \n",
       "4  {'offer id': '4d5c57ea9a6940dd891ad53e9dbe8da0'}   \n",
       "\n",
       "                           offer_id  \n",
       "0  9b98b8c7a33c4b65b9aebfe6a799e6d9  \n",
       "1  0b1e1539f2cc45b7b9fa7c272da2e1d7  \n",
       "2  2906b810c7d4411798c6938adc9daaa5  \n",
       "3  fafdcd668e3743c1bb461111dcafc2a4  \n",
       "4  4d5c57ea9a6940dd891ad53e9dbe8da0  "
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "offer.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['offer received', 'offer viewed', 'offer completed'], dtype=object)"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "offer['event'].unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "def merge_datasets(cleaned_portfolio, cleaned_profile, offer):\n",
    "    \"\"\" Merge the three data sets into one\n",
    "    Input:\n",
    "    - portfolio_clean\n",
    "    - profile_clean\n",
    "    - transcript_clean\n",
    "    Output:\n",
    "    - df: merged dataframe\n",
    "    \"\"\"\n",
    "    trans_prof = pd.merge(offer, cleaned_profile, on='customer_id',\n",
    "                          how=\"left\")\n",
    "    df = pd.merge(trans_prof, cleaned_portfolio, on='offer_id', how='left')\n",
    "    # Change the offer ids to a simplied form\n",
    "    offer_id = {'ae264e3637204a6fb9bb56bc8210ddfd': 'O0',\n",
    "                '4d5c57ea9a6940dd891ad53e9dbe8da0': 'O1',\n",
    "                '9b98b8c7a33c4b65b9aebfe6a799e6d9': 'O2',\n",
    "                'f19421c1d4aa40978ebb69ca19b0e20d': 'O3',\n",
    "                '0b1e1539f2cc45b7b9fa7c272da2e1d7': 'O4',\n",
    "                '2298d6c36e964ae4a3e7e9706d1fb8c2': 'O5',\n",
    "                'fafdcd668e3743c1bb461111dcafc2a4': 'O6',\n",
    "                '2906b810c7d4411798c6938adc9daaa5': 'O7',\n",
    "                '3f207df678b143eea3cee63160fa8bed': 'O8',\n",
    "                '5a8bc65990b245e5a138643cd4eb9837': 'O9'}\n",
    "    df.offer_id = df.offer_id.apply(lambda x: offer_id[x] if x else None)\n",
    "\n",
    "    return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "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>event</th>\n",
       "      <th>customer_id</th>\n",
       "      <th>time</th>\n",
       "      <th>value</th>\n",
       "      <th>offer_id</th>\n",
       "      <th>age</th>\n",
       "      <th>became_member_on</th>\n",
       "      <th>gender</th>\n",
       "      <th>income</th>\n",
       "      <th>memberdays</th>\n",
       "      <th>difficulty</th>\n",
       "      <th>duration</th>\n",
       "      <th>reward</th>\n",
       "      <th>web</th>\n",
       "      <th>email</th>\n",
       "      <th>mobile</th>\n",
       "      <th>social</th>\n",
       "      <th>bogo</th>\n",
       "      <th>discount</th>\n",
       "      <th>informational</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>offer received</td>\n",
       "      <td>78afa995795e4d85b5d9ceeca43f5fef</td>\n",
       "      <td>0</td>\n",
       "      <td>{'offer id': '9b98b8c7a33c4b65b9aebfe6a799e6d9'}</td>\n",
       "      <td>O2</td>\n",
       "      <td>75.0</td>\n",
       "      <td>20170509</td>\n",
       "      <td>F</td>\n",
       "      <td>100000.0</td>\n",
       "      <td>1106</td>\n",
       "      <td>5</td>\n",
       "      <td>168</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>offer received</td>\n",
       "      <td>a03223e636434f42ac4c3df47e8bac43</td>\n",
       "      <td>0</td>\n",
       "      <td>{'offer id': '0b1e1539f2cc45b7b9fa7c272da2e1d7'}</td>\n",
       "      <td>O4</td>\n",
       "      <td>NaN</td>\n",
       "      <td>20170804</td>\n",
       "      <td>None</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1019</td>\n",
       "      <td>20</td>\n",
       "      <td>240</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>offer received</td>\n",
       "      <td>e2127556f4f64592b11af22de27a7932</td>\n",
       "      <td>0</td>\n",
       "      <td>{'offer id': '2906b810c7d4411798c6938adc9daaa5'}</td>\n",
       "      <td>O7</td>\n",
       "      <td>68.0</td>\n",
       "      <td>20180426</td>\n",
       "      <td>M</td>\n",
       "      <td>70000.0</td>\n",
       "      <td>754</td>\n",
       "      <td>10</td>\n",
       "      <td>168</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>offer received</td>\n",
       "      <td>8ec6ce2a7e7949b1bf142def7d0e0586</td>\n",
       "      <td>0</td>\n",
       "      <td>{'offer id': 'fafdcd668e3743c1bb461111dcafc2a4'}</td>\n",
       "      <td>O6</td>\n",
       "      <td>NaN</td>\n",
       "      <td>20170925</td>\n",
       "      <td>None</td>\n",
       "      <td>NaN</td>\n",
       "      <td>967</td>\n",
       "      <td>10</td>\n",
       "      <td>240</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>offer received</td>\n",
       "      <td>68617ca6246f4fbc85e91a2a49552598</td>\n",
       "      <td>0</td>\n",
       "      <td>{'offer id': '4d5c57ea9a6940dd891ad53e9dbe8da0'}</td>\n",
       "      <td>O1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>20171002</td>\n",
       "      <td>None</td>\n",
       "      <td>NaN</td>\n",
       "      <td>960</td>\n",
       "      <td>10</td>\n",
       "      <td>120</td>\n",
       "      <td>10</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            event                       customer_id  time  \\\n",
       "0  offer received  78afa995795e4d85b5d9ceeca43f5fef     0   \n",
       "1  offer received  a03223e636434f42ac4c3df47e8bac43     0   \n",
       "2  offer received  e2127556f4f64592b11af22de27a7932     0   \n",
       "3  offer received  8ec6ce2a7e7949b1bf142def7d0e0586     0   \n",
       "4  offer received  68617ca6246f4fbc85e91a2a49552598     0   \n",
       "\n",
       "                                              value offer_id   age  \\\n",
       "0  {'offer id': '9b98b8c7a33c4b65b9aebfe6a799e6d9'}       O2  75.0   \n",
       "1  {'offer id': '0b1e1539f2cc45b7b9fa7c272da2e1d7'}       O4   NaN   \n",
       "2  {'offer id': '2906b810c7d4411798c6938adc9daaa5'}       O7  68.0   \n",
       "3  {'offer id': 'fafdcd668e3743c1bb461111dcafc2a4'}       O6   NaN   \n",
       "4  {'offer id': '4d5c57ea9a6940dd891ad53e9dbe8da0'}       O1   NaN   \n",
       "\n",
       "   became_member_on gender    income  memberdays  difficulty  duration  \\\n",
       "0          20170509      F  100000.0        1106           5       168   \n",
       "1          20170804   None       NaN        1019          20       240   \n",
       "2          20180426      M   70000.0         754          10       168   \n",
       "3          20170925   None       NaN         967          10       240   \n",
       "4          20171002   None       NaN         960          10       120   \n",
       "\n",
       "   reward  web  email  mobile  social  bogo  discount  informational  \n",
       "0       5    1      1       1       0     1         0              0  \n",
       "1       5    1      1       0       0     0         1              0  \n",
       "2       2    1      1       1       0     0         1              0  \n",
       "3       2    1      1       1       1     0         1              0  \n",
       "4      10    1      1       1       1     1         0              0  "
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = merge_datasets(cleaned_portfolio, cleaned_profile, offer)\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f164928c828>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.subplot(131)\n",
    "df[df.bogo == 1].event.value_counts().plot(kind='bar', rot=45, figsize=(16,5))\n",
    "plt.title('Offer Type - Bogo');\n",
    "\n",
    "plt.subplot(132)\n",
    "df[df.discount == 1].event.value_counts().plot(kind='bar', rot=45, figsize=(16,5))\n",
    "plt.title('Offer Type - Discount');\n",
    "\n",
    "plt.subplot(133)\n",
    "df[df.informational == 1].event.value_counts().plot(kind='bar', rot=45, figsize=(16,5))\n",
    "plt.title('Offer Type - Informational');"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f1646856cc0>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.subplot(131)\n",
    "df[df.event == 'offer received'].offer_id.value_counts().plot(kind='bar', figsize=(16,5))\n",
    "plt.title('Most Received Offers');\n",
    "\n",
    "plt.subplot(132)\n",
    "df[df.event == 'offer viewed'].offer_id.value_counts().plot(kind='bar', figsize=(16,5))\n",
    "plt.title('Most Viewed Offers');\n",
    "\n",
    "plt.subplot(133)\n",
    "df[df.event == 'offer completed'].offer_id.value_counts().plot(kind='bar', figsize=(16,5))\n",
    "plt.title('Most Completed Orders');"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f16466e74a8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.subplot(131)\n",
    "df[df.bogo == 1].offer_id.value_counts().plot(kind='pie', autopct='%1.1f%%', shadow=True, figsize=(16,5))\n",
    "plt.title('Bogo Offer Distribution');\n",
    "\n",
    "plt.subplot(132)\n",
    "df[df.discount == 1].offer_id.value_counts().plot(kind='pie', autopct='%1.1f%%', shadow=True, figsize=(16,5))\n",
    "plt.title('Discount Offer Distribution');\n",
    "\n",
    "plt.subplot(133)\n",
    "df[df.informational == 1].offer_id.value_counts().plot(kind='pie', autopct='%1.1f%%', shadow=True, figsize=(16,5))\n",
    "plt.title('Informational Offer Distribution');"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "O2    7677\n",
       "O4    7668\n",
       "O0    7658\n",
       "O5    7646\n",
       "O7    7632\n",
       "O9    7618\n",
       "O8    7617\n",
       "O6    7597\n",
       "O1    7593\n",
       "O3    7571\n",
       "Name: offer_id, dtype: int64"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[df.event == 'offer received'].offer_id.value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Each offer was received almost the same number of times. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "def create_user_item_matrix(offer, filename):\n",
    "    '''\n",
    "    Return the user item matrix that indicate the number of offer complete of a particular user\n",
    "    \n",
    "    INPUT:\n",
    "    offer - a cleaned transcript dataframe\n",
    "    filename(string) - the file name that save the user item matrix\n",
    "    \n",
    "    OUTPUT:\n",
    "    user_item_matrix - the user item matrix which \n",
    "        - row is user \n",
    "        - column is offer\n",
    "        - value is the number of offer complete by the user (NaN means no offer given)\n",
    "    \n",
    "    '''\n",
    "    # create an empty user item matrix\n",
    "    user_item_matrix = offer.groupby(['customer_id', 'offer_id'])['event'].agg(lambda x: np.nan).unstack()\n",
    "    \n",
    "    # we just focus on bogo and discount first\n",
    "    user_item_matrix.drop(list(cleaned_portfolio[cleaned_portfolio['informational']==1]['offer_id']), axis=1, inplace=True)\n",
    "    \n",
    "    for offer_id in user_item_matrix.columns:\n",
    "        print(\"Now processing: \", offer_id)\n",
    "        for customer_id in user_item_matrix.index:\n",
    "            events = []\n",
    "            for event in offer[(offer['offer_id']==offer_id) & (offer['customer_id']==customer_id)]['event']:\n",
    "                events.append(event)\n",
    "            if len(events) >= 3:\n",
    "                user_item_matrix.loc[customer_id, offer_id] = 0\n",
    "                for i in range(len(events)-2):\n",
    "                    # check if the transaction sequence is offer received -> offer viewed -> offer completed\n",
    "                    # if yes, we assume the user reacted positively with the offer we provided.\n",
    "                    if (events[i] == 'offer received') & (events[i+1] == 'offer viewed') & (events[i+2] == 'offer completed'):\n",
    "                        user_item_matrix.loc[customer_id, offer_id] += 1\n",
    "            elif len(events) > 0:\n",
    "                user_item_matrix.loc[customer_id, offer_id] = 0\n",
    "    \n",
    "    # store the large martix into file\n",
    "    fh = open(filename, 'wb')\n",
    "    pickle.dump(user_item_matrix,fh)\n",
    "    fh.close()\n",
    "    \n",
    "    return user_item_matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "# !! long process !!\n",
    "# created user_item_matrix on local machine and then uploaded to workspace to read directly\n",
    "\n",
    "# full_df = create_user_item_matrix(offer, 'user_item_matrix.p')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "# run this cell if you want to omit the long process of matrix creation\n",
    "full_df = pd.read_pickle('user_item_matrix.p')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "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>offer_id</th>\n",
       "      <th>0b1e1539f2cc45b7b9fa7c272da2e1d7</th>\n",
       "      <th>2298d6c36e964ae4a3e7e9706d1fb8c2</th>\n",
       "      <th>2906b810c7d4411798c6938adc9daaa5</th>\n",
       "      <th>4d5c57ea9a6940dd891ad53e9dbe8da0</th>\n",
       "      <th>9b98b8c7a33c4b65b9aebfe6a799e6d9</th>\n",
       "      <th>ae264e3637204a6fb9bb56bc8210ddfd</th>\n",
       "      <th>f19421c1d4aa40978ebb69ca19b0e20d</th>\n",
       "      <th>fafdcd668e3743c1bb461111dcafc2a4</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>person</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0009655768c64bdeb2e877511632db8f</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>00116118485d4dfda04fdbaba9a87b5c</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0011e0d4e6b944f998e987f904e8c1e5</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0020c2b971eb4e9188eac86d93036a77</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0020ccbbb6d84e358d3414a3ff76cffd</th>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "offer_id                          0b1e1539f2cc45b7b9fa7c272da2e1d7  \\\n",
       "person                                                               \n",
       "0009655768c64bdeb2e877511632db8f                               NaN   \n",
       "00116118485d4dfda04fdbaba9a87b5c                               NaN   \n",
       "0011e0d4e6b944f998e987f904e8c1e5                               1.0   \n",
       "0020c2b971eb4e9188eac86d93036a77                               NaN   \n",
       "0020ccbbb6d84e358d3414a3ff76cffd                               NaN   \n",
       "\n",
       "offer_id                          2298d6c36e964ae4a3e7e9706d1fb8c2  \\\n",
       "person                                                               \n",
       "0009655768c64bdeb2e877511632db8f                               NaN   \n",
       "00116118485d4dfda04fdbaba9a87b5c                               NaN   \n",
       "0011e0d4e6b944f998e987f904e8c1e5                               1.0   \n",
       "0020c2b971eb4e9188eac86d93036a77                               NaN   \n",
       "0020ccbbb6d84e358d3414a3ff76cffd                               1.0   \n",
       "\n",
       "offer_id                          2906b810c7d4411798c6938adc9daaa5  \\\n",
       "person                                                               \n",
       "0009655768c64bdeb2e877511632db8f                               0.0   \n",
       "00116118485d4dfda04fdbaba9a87b5c                               NaN   \n",
       "0011e0d4e6b944f998e987f904e8c1e5                               NaN   \n",
       "0020c2b971eb4e9188eac86d93036a77                               NaN   \n",
       "0020ccbbb6d84e358d3414a3ff76cffd                               NaN   \n",
       "\n",
       "offer_id                          4d5c57ea9a6940dd891ad53e9dbe8da0  \\\n",
       "person                                                               \n",
       "0009655768c64bdeb2e877511632db8f                               NaN   \n",
       "00116118485d4dfda04fdbaba9a87b5c                               NaN   \n",
       "0011e0d4e6b944f998e987f904e8c1e5                               NaN   \n",
       "0020c2b971eb4e9188eac86d93036a77                               1.0   \n",
       "0020ccbbb6d84e358d3414a3ff76cffd                               NaN   \n",
       "\n",
       "offer_id                          9b98b8c7a33c4b65b9aebfe6a799e6d9  \\\n",
       "person                                                               \n",
       "0009655768c64bdeb2e877511632db8f                               NaN   \n",
       "00116118485d4dfda04fdbaba9a87b5c                               NaN   \n",
       "0011e0d4e6b944f998e987f904e8c1e5                               1.0   \n",
       "0020c2b971eb4e9188eac86d93036a77                               NaN   \n",
       "0020ccbbb6d84e358d3414a3ff76cffd                               1.0   \n",
       "\n",
       "offer_id                          ae264e3637204a6fb9bb56bc8210ddfd  \\\n",
       "person                                                               \n",
       "0009655768c64bdeb2e877511632db8f                               NaN   \n",
       "00116118485d4dfda04fdbaba9a87b5c                               NaN   \n",
       "0011e0d4e6b944f998e987f904e8c1e5                               NaN   \n",
       "0020c2b971eb4e9188eac86d93036a77                               0.0   \n",
       "0020ccbbb6d84e358d3414a3ff76cffd                               NaN   \n",
       "\n",
       "offer_id                          f19421c1d4aa40978ebb69ca19b0e20d  \\\n",
       "person                                                               \n",
       "0009655768c64bdeb2e877511632db8f                               0.0   \n",
       "00116118485d4dfda04fdbaba9a87b5c                               0.0   \n",
       "0011e0d4e6b944f998e987f904e8c1e5                               NaN   \n",
       "0020c2b971eb4e9188eac86d93036a77                               NaN   \n",
       "0020ccbbb6d84e358d3414a3ff76cffd                               1.0   \n",
       "\n",
       "offer_id                          fafdcd668e3743c1bb461111dcafc2a4  \n",
       "person                                                              \n",
       "0009655768c64bdeb2e877511632db8f                               0.0  \n",
       "00116118485d4dfda04fdbaba9a87b5c                               NaN  \n",
       "0011e0d4e6b944f998e987f904e8c1e5                               NaN  \n",
       "0020c2b971eb4e9188eac86d93036a77                               1.0  \n",
       "0020ccbbb6d84e358d3414a3ff76cffd                               NaN  "
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "full_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(117306, 50275)"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_size = int(offer.shape[0]*0.7)\n",
    "test_size = offer.shape[0] - train_size\n",
    "train_size, test_size"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_offer = offer[:train_size]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "test_offer = offer[train_size:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "# !! long process !!\n",
    "# created train_df on local machine and then uploaded to workspace to read directly\n",
    "\n",
    "# train_df = create_user_item_matrix(train_offer, 'train_df.p')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "# run this cell if you want to omit the long process of matrix creation\n",
    "train_df = pd.read_pickle('train_df.p')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "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>offer_id</th>\n",
       "      <th>0b1e1539f2cc45b7b9fa7c272da2e1d7</th>\n",
       "      <th>2298d6c36e964ae4a3e7e9706d1fb8c2</th>\n",
       "      <th>2906b810c7d4411798c6938adc9daaa5</th>\n",
       "      <th>4d5c57ea9a6940dd891ad53e9dbe8da0</th>\n",
       "      <th>9b98b8c7a33c4b65b9aebfe6a799e6d9</th>\n",
       "      <th>ae264e3637204a6fb9bb56bc8210ddfd</th>\n",
       "      <th>f19421c1d4aa40978ebb69ca19b0e20d</th>\n",
       "      <th>fafdcd668e3743c1bb461111dcafc2a4</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>person</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0009655768c64bdeb2e877511632db8f</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>00116118485d4dfda04fdbaba9a87b5c</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0011e0d4e6b944f998e987f904e8c1e5</th>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0020c2b971eb4e9188eac86d93036a77</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0020ccbbb6d84e358d3414a3ff76cffd</th>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "offer_id                          0b1e1539f2cc45b7b9fa7c272da2e1d7  \\\n",
       "person                                                               \n",
       "0009655768c64bdeb2e877511632db8f                               NaN   \n",
       "00116118485d4dfda04fdbaba9a87b5c                               NaN   \n",
       "0011e0d4e6b944f998e987f904e8c1e5                               0.0   \n",
       "0020c2b971eb4e9188eac86d93036a77                               NaN   \n",
       "0020ccbbb6d84e358d3414a3ff76cffd                               NaN   \n",
       "\n",
       "offer_id                          2298d6c36e964ae4a3e7e9706d1fb8c2  \\\n",
       "person                                                               \n",
       "0009655768c64bdeb2e877511632db8f                               NaN   \n",
       "00116118485d4dfda04fdbaba9a87b5c                               NaN   \n",
       "0011e0d4e6b944f998e987f904e8c1e5                               1.0   \n",
       "0020c2b971eb4e9188eac86d93036a77                               NaN   \n",
       "0020ccbbb6d84e358d3414a3ff76cffd                               1.0   \n",
       "\n",
       "offer_id                          2906b810c7d4411798c6938adc9daaa5  \\\n",
       "person                                                               \n",
       "0009655768c64bdeb2e877511632db8f                               NaN   \n",
       "00116118485d4dfda04fdbaba9a87b5c                               NaN   \n",
       "0011e0d4e6b944f998e987f904e8c1e5                               NaN   \n",
       "0020c2b971eb4e9188eac86d93036a77                               NaN   \n",
       "0020ccbbb6d84e358d3414a3ff76cffd                               NaN   \n",
       "\n",
       "offer_id                          4d5c57ea9a6940dd891ad53e9dbe8da0  \\\n",
       "person                                                               \n",
       "0009655768c64bdeb2e877511632db8f                               NaN   \n",
       "00116118485d4dfda04fdbaba9a87b5c                               NaN   \n",
       "0011e0d4e6b944f998e987f904e8c1e5                               NaN   \n",
       "0020c2b971eb4e9188eac86d93036a77                               0.0   \n",
       "0020ccbbb6d84e358d3414a3ff76cffd                               NaN   \n",
       "\n",
       "offer_id                          9b98b8c7a33c4b65b9aebfe6a799e6d9  \\\n",
       "person                                                               \n",
       "0009655768c64bdeb2e877511632db8f                               NaN   \n",
       "00116118485d4dfda04fdbaba9a87b5c                               NaN   \n",
       "0011e0d4e6b944f998e987f904e8c1e5                               0.0   \n",
       "0020c2b971eb4e9188eac86d93036a77                               NaN   \n",
       "0020ccbbb6d84e358d3414a3ff76cffd                               0.0   \n",
       "\n",
       "offer_id                          ae264e3637204a6fb9bb56bc8210ddfd  \\\n",
       "person                                                               \n",
       "0009655768c64bdeb2e877511632db8f                               NaN   \n",
       "00116118485d4dfda04fdbaba9a87b5c                               NaN   \n",
       "0011e0d4e6b944f998e987f904e8c1e5                               NaN   \n",
       "0020c2b971eb4e9188eac86d93036a77                               0.0   \n",
       "0020ccbbb6d84e358d3414a3ff76cffd                               NaN   \n",
       "\n",
       "offer_id                          f19421c1d4aa40978ebb69ca19b0e20d  \\\n",
       "person                                                               \n",
       "0009655768c64bdeb2e877511632db8f                               0.0   \n",
       "00116118485d4dfda04fdbaba9a87b5c                               0.0   \n",
       "0011e0d4e6b944f998e987f904e8c1e5                               NaN   \n",
       "0020c2b971eb4e9188eac86d93036a77                               NaN   \n",
       "0020ccbbb6d84e358d3414a3ff76cffd                               1.0   \n",
       "\n",
       "offer_id                          fafdcd668e3743c1bb461111dcafc2a4  \n",
       "person                                                              \n",
       "0009655768c64bdeb2e877511632db8f                               0.0  \n",
       "00116118485d4dfda04fdbaba9a87b5c                               NaN  \n",
       "0011e0d4e6b944f998e987f904e8c1e5                               NaN  \n",
       "0020c2b971eb4e9188eac86d93036a77                               1.0  \n",
       "0020ccbbb6d84e358d3414a3ff76cffd                               NaN  "
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "# !! long process !!\n",
    "# created test_df on local machine and then uploaded to workspace to read directly\n",
    "\n",
    "# test_df = create_user_item_matrix(test_offer, 'test_df.p')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "# run this cell if you want to omit the long process of matrix creation\n",
    "test_df = pd.read_pickle('train_df.p')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "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>offer_id</th>\n",
       "      <th>0b1e1539f2cc45b7b9fa7c272da2e1d7</th>\n",
       "      <th>2298d6c36e964ae4a3e7e9706d1fb8c2</th>\n",
       "      <th>2906b810c7d4411798c6938adc9daaa5</th>\n",
       "      <th>4d5c57ea9a6940dd891ad53e9dbe8da0</th>\n",
       "      <th>9b98b8c7a33c4b65b9aebfe6a799e6d9</th>\n",
       "      <th>ae264e3637204a6fb9bb56bc8210ddfd</th>\n",
       "      <th>f19421c1d4aa40978ebb69ca19b0e20d</th>\n",
       "      <th>fafdcd668e3743c1bb461111dcafc2a4</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>person</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0009655768c64bdeb2e877511632db8f</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>00116118485d4dfda04fdbaba9a87b5c</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0011e0d4e6b944f998e987f904e8c1e5</th>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0020c2b971eb4e9188eac86d93036a77</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0020ccbbb6d84e358d3414a3ff76cffd</th>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "offer_id                          0b1e1539f2cc45b7b9fa7c272da2e1d7  \\\n",
       "person                                                               \n",
       "0009655768c64bdeb2e877511632db8f                               NaN   \n",
       "00116118485d4dfda04fdbaba9a87b5c                               NaN   \n",
       "0011e0d4e6b944f998e987f904e8c1e5                               0.0   \n",
       "0020c2b971eb4e9188eac86d93036a77                               NaN   \n",
       "0020ccbbb6d84e358d3414a3ff76cffd                               NaN   \n",
       "\n",
       "offer_id                          2298d6c36e964ae4a3e7e9706d1fb8c2  \\\n",
       "person                                                               \n",
       "0009655768c64bdeb2e877511632db8f                               NaN   \n",
       "00116118485d4dfda04fdbaba9a87b5c                               NaN   \n",
       "0011e0d4e6b944f998e987f904e8c1e5                               1.0   \n",
       "0020c2b971eb4e9188eac86d93036a77                               NaN   \n",
       "0020ccbbb6d84e358d3414a3ff76cffd                               1.0   \n",
       "\n",
       "offer_id                          2906b810c7d4411798c6938adc9daaa5  \\\n",
       "person                                                               \n",
       "0009655768c64bdeb2e877511632db8f                               NaN   \n",
       "00116118485d4dfda04fdbaba9a87b5c                               NaN   \n",
       "0011e0d4e6b944f998e987f904e8c1e5                               NaN   \n",
       "0020c2b971eb4e9188eac86d93036a77                               NaN   \n",
       "0020ccbbb6d84e358d3414a3ff76cffd                               NaN   \n",
       "\n",
       "offer_id                          4d5c57ea9a6940dd891ad53e9dbe8da0  \\\n",
       "person                                                               \n",
       "0009655768c64bdeb2e877511632db8f                               NaN   \n",
       "00116118485d4dfda04fdbaba9a87b5c                               NaN   \n",
       "0011e0d4e6b944f998e987f904e8c1e5                               NaN   \n",
       "0020c2b971eb4e9188eac86d93036a77                               0.0   \n",
       "0020ccbbb6d84e358d3414a3ff76cffd                               NaN   \n",
       "\n",
       "offer_id                          9b98b8c7a33c4b65b9aebfe6a799e6d9  \\\n",
       "person                                                               \n",
       "0009655768c64bdeb2e877511632db8f                               NaN   \n",
       "00116118485d4dfda04fdbaba9a87b5c                               NaN   \n",
       "0011e0d4e6b944f998e987f904e8c1e5                               0.0   \n",
       "0020c2b971eb4e9188eac86d93036a77                               NaN   \n",
       "0020ccbbb6d84e358d3414a3ff76cffd                               0.0   \n",
       "\n",
       "offer_id                          ae264e3637204a6fb9bb56bc8210ddfd  \\\n",
       "person                                                               \n",
       "0009655768c64bdeb2e877511632db8f                               NaN   \n",
       "00116118485d4dfda04fdbaba9a87b5c                               NaN   \n",
       "0011e0d4e6b944f998e987f904e8c1e5                               NaN   \n",
       "0020c2b971eb4e9188eac86d93036a77                               0.0   \n",
       "0020ccbbb6d84e358d3414a3ff76cffd                               NaN   \n",
       "\n",
       "offer_id                          f19421c1d4aa40978ebb69ca19b0e20d  \\\n",
       "person                                                               \n",
       "0009655768c64bdeb2e877511632db8f                               0.0   \n",
       "00116118485d4dfda04fdbaba9a87b5c                               0.0   \n",
       "0011e0d4e6b944f998e987f904e8c1e5                               NaN   \n",
       "0020c2b971eb4e9188eac86d93036a77                               NaN   \n",
       "0020ccbbb6d84e358d3414a3ff76cffd                               1.0   \n",
       "\n",
       "offer_id                          fafdcd668e3743c1bb461111dcafc2a4  \n",
       "person                                                              \n",
       "0009655768c64bdeb2e877511632db8f                               0.0  \n",
       "00116118485d4dfda04fdbaba9a87b5c                               NaN  \n",
       "0011e0d4e6b944f998e987f904e8c1e5                               NaN  \n",
       "0020c2b971eb4e9188eac86d93036a77                               1.0  \n",
       "0020ccbbb6d84e358d3414a3ff76cffd                               NaN  "
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "def FunkSVD(complete_mat, latent_features=12, learning_rate=0.0001, iters=100):\n",
    "    '''\n",
    "    This function performs matrix factorization using a basic form of FunkSVD with no regularization\n",
    "    \n",
    "    INPUT:\n",
    "    complete_mat - (numpy array) a matrix with users as rows, offers as columns, and offer completed as values\n",
    "    latent_features - (int) the number of latent features used\n",
    "    learning_rate - (float) the learning rate \n",
    "    iters - (int) the number of iterations\n",
    "    \n",
    "    OUTPUT:\n",
    "    user_mat - (numpy array) a user by latent feature matrix\n",
    "    offer_mat - (numpy array) a latent feature by offer matrix\n",
    "    '''\n",
    "    \n",
    "    # Set up useful values to be used through the rest of the function\n",
    "    n_users = complete_mat.shape[0]\n",
    "    n_offers = complete_mat.shape[1]\n",
    "    num_complete = np.count_nonzero(~np.isnan(complete_mat))\n",
    "    \n",
    "    # initialize the user and movie matrices with random values\n",
    "    user_mat = np.random.rand(n_users, latent_features)\n",
    "    offer_mat = np.random.rand(latent_features, n_offers)\n",
    "    \n",
    "    # initialize sse at 0 for first iteration\n",
    "    sse_accum = 0\n",
    "    \n",
    "    # keep track of iteration and MSE\n",
    "    print(\"Optimizaiton Statistics\")\n",
    "    print(\"Iterations | Mean Squared Error \")\n",
    "    \n",
    "    # for each iteration\n",
    "    for iteration in range(iters):\n",
    "\n",
    "        # update our sse\n",
    "        old_sse = sse_accum\n",
    "        sse_accum = 0\n",
    "        \n",
    "        # For each user-offer pair\n",
    "        for i in range(n_users):\n",
    "            for j in range(n_offers):\n",
    "                \n",
    "                # if the rating exists\n",
    "                if complete_mat[i, j] >= 0:\n",
    "                    \n",
    "                    # compute the error as the actual minus the dot product of the user and offer latent features\n",
    "                    diff = complete_mat[i, j] - np.dot(user_mat[i, :], offer_mat[:, j])\n",
    "                    \n",
    "                    # Keep track of the sum of squared errors for the matrix\n",
    "                    sse_accum += diff**2\n",
    "                    \n",
    "                    # update the values in each matrix in the direction of the gradient\n",
    "                    for k in range(latent_features):\n",
    "                        user_mat[i, k] += learning_rate * (2*diff*offer_mat[k, j])\n",
    "                        offer_mat[k, j] += learning_rate * (2*diff*user_mat[i, k])\n",
    "\n",
    "        # print results\n",
    "        print(\"%d \\t\\t %f\" % (iteration+1, sse_accum / num_complete))\n",
    "        \n",
    "    return user_mat, offer_mat"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Optimizaiton Statistics\n",
      "Iterations | Mean Squared Error \n",
      "1 \t\t 0.310734\n",
      "2 \t\t 0.267348\n",
      "3 \t\t 0.267007\n",
      "4 \t\t 0.266649\n",
      "5 \t\t 0.266287\n",
      "6 \t\t 0.265920\n",
      "7 \t\t 0.265549\n",
      "8 \t\t 0.265173\n",
      "9 \t\t 0.264791\n",
      "10 \t\t 0.264403\n",
      "11 \t\t 0.264009\n",
      "12 \t\t 0.263609\n",
      "13 \t\t 0.263201\n",
      "14 \t\t 0.262785\n",
      "15 \t\t 0.262362\n",
      "16 \t\t 0.261929\n",
      "17 \t\t 0.261486\n",
      "18 \t\t 0.261034\n",
      "19 \t\t 0.260571\n",
      "20 \t\t 0.260095\n",
      "21 \t\t 0.259608\n",
      "22 \t\t 0.259106\n",
      "23 \t\t 0.258591\n",
      "24 \t\t 0.258059\n",
      "25 \t\t 0.257512\n",
      "26 \t\t 0.256947\n",
      "27 \t\t 0.256362\n",
      "28 \t\t 0.255758\n",
      "29 \t\t 0.255132\n",
      "30 \t\t 0.254483\n",
      "31 \t\t 0.253810\n",
      "32 \t\t 0.253110\n",
      "33 \t\t 0.252383\n",
      "34 \t\t 0.251626\n",
      "35 \t\t 0.250837\n",
      "36 \t\t 0.250015\n",
      "37 \t\t 0.249157\n",
      "38 \t\t 0.248261\n",
      "39 \t\t 0.247326\n",
      "40 \t\t 0.246348\n",
      "41 \t\t 0.245325\n",
      "42 \t\t 0.244256\n",
      "43 \t\t 0.243136\n",
      "44 \t\t 0.241964\n",
      "45 \t\t 0.240738\n",
      "46 \t\t 0.239454\n",
      "47 \t\t 0.238109\n",
      "48 \t\t 0.236702\n",
      "49 \t\t 0.235229\n",
      "50 \t\t 0.233688\n",
      "51 \t\t 0.232076\n",
      "52 \t\t 0.230392\n",
      "53 \t\t 0.228633\n",
      "54 \t\t 0.226796\n",
      "55 \t\t 0.224880\n",
      "56 \t\t 0.222883\n",
      "57 \t\t 0.220805\n",
      "58 \t\t 0.218643\n",
      "59 \t\t 0.216397\n",
      "60 \t\t 0.214068\n",
      "61 \t\t 0.211654\n",
      "62 \t\t 0.209157\n",
      "63 \t\t 0.206578\n",
      "64 \t\t 0.203918\n",
      "65 \t\t 0.201179\n",
      "66 \t\t 0.198364\n",
      "67 \t\t 0.195476\n",
      "68 \t\t 0.192518\n",
      "69 \t\t 0.189494\n",
      "70 \t\t 0.186409\n",
      "71 \t\t 0.183268\n",
      "72 \t\t 0.180075\n",
      "73 \t\t 0.176837\n",
      "74 \t\t 0.173560\n",
      "75 \t\t 0.170249\n",
      "76 \t\t 0.166912\n",
      "77 \t\t 0.163554\n",
      "78 \t\t 0.160181\n",
      "79 \t\t 0.156801\n",
      "80 \t\t 0.153420\n",
      "81 \t\t 0.150044\n",
      "82 \t\t 0.146679\n",
      "83 \t\t 0.143332\n",
      "84 \t\t 0.140006\n",
      "85 \t\t 0.136709\n",
      "86 \t\t 0.133444\n",
      "87 \t\t 0.130217\n",
      "88 \t\t 0.127031\n",
      "89 \t\t 0.123891\n",
      "90 \t\t 0.120799\n",
      "91 \t\t 0.117758\n",
      "92 \t\t 0.114772\n",
      "93 \t\t 0.111842\n",
      "94 \t\t 0.108970\n",
      "95 \t\t 0.106158\n",
      "96 \t\t 0.103407\n",
      "97 \t\t 0.100717\n",
      "98 \t\t 0.098090\n",
      "99 \t\t 0.095525\n",
      "100 \t\t 0.093022\n",
      "101 \t\t 0.090582\n",
      "102 \t\t 0.088204\n",
      "103 \t\t 0.085888\n",
      "104 \t\t 0.083633\n",
      "105 \t\t 0.081437\n",
      "106 \t\t 0.079301\n",
      "107 \t\t 0.077223\n",
      "108 \t\t 0.075202\n",
      "109 \t\t 0.073237\n",
      "110 \t\t 0.071327\n",
      "111 \t\t 0.069470\n",
      "112 \t\t 0.067665\n",
      "113 \t\t 0.065912\n",
      "114 \t\t 0.064208\n",
      "115 \t\t 0.062552\n",
      "116 \t\t 0.060944\n",
      "117 \t\t 0.059382\n",
      "118 \t\t 0.057864\n",
      "119 \t\t 0.056389\n",
      "120 \t\t 0.054957\n",
      "121 \t\t 0.053565\n",
      "122 \t\t 0.052214\n",
      "123 \t\t 0.050901\n",
      "124 \t\t 0.049625\n",
      "125 \t\t 0.048386\n",
      "126 \t\t 0.047182\n",
      "127 \t\t 0.046012\n",
      "128 \t\t 0.044876\n",
      "129 \t\t 0.043771\n",
      "130 \t\t 0.042698\n",
      "131 \t\t 0.041656\n",
      "132 \t\t 0.040642\n",
      "133 \t\t 0.039658\n",
      "134 \t\t 0.038700\n",
      "135 \t\t 0.037770\n",
      "136 \t\t 0.036866\n",
      "137 \t\t 0.035986\n",
      "138 \t\t 0.035131\n",
      "139 \t\t 0.034300\n",
      "140 \t\t 0.033492\n",
      "141 \t\t 0.032705\n",
      "142 \t\t 0.031940\n",
      "143 \t\t 0.031196\n",
      "144 \t\t 0.030472\n",
      "145 \t\t 0.029767\n",
      "146 \t\t 0.029081\n",
      "147 \t\t 0.028413\n",
      "148 \t\t 0.027763\n",
      "149 \t\t 0.027130\n",
      "150 \t\t 0.026513\n",
      "151 \t\t 0.025912\n",
      "152 \t\t 0.025326\n",
      "153 \t\t 0.024755\n",
      "154 \t\t 0.024199\n",
      "155 \t\t 0.023656\n",
      "156 \t\t 0.023127\n",
      "157 \t\t 0.022610\n",
      "158 \t\t 0.022107\n",
      "159 \t\t 0.021615\n",
      "160 \t\t 0.021135\n",
      "161 \t\t 0.020666\n",
      "162 \t\t 0.020208\n",
      "163 \t\t 0.019761\n",
      "164 \t\t 0.019323\n",
      "165 \t\t 0.018896\n",
      "166 \t\t 0.018479\n",
      "167 \t\t 0.018070\n",
      "168 \t\t 0.017671\n",
      "169 \t\t 0.017280\n",
      "170 \t\t 0.016898\n",
      "171 \t\t 0.016524\n",
      "172 \t\t 0.016158\n",
      "173 \t\t 0.015800\n",
      "174 \t\t 0.015449\n",
      "175 \t\t 0.015105\n",
      "176 \t\t 0.014769\n",
      "177 \t\t 0.014439\n",
      "178 \t\t 0.014117\n",
      "179 \t\t 0.013800\n",
      "180 \t\t 0.013491\n",
      "181 \t\t 0.013187\n",
      "182 \t\t 0.012890\n",
      "183 \t\t 0.012598\n",
      "184 \t\t 0.012313\n",
      "185 \t\t 0.012033\n",
      "186 \t\t 0.011758\n",
      "187 \t\t 0.011489\n",
      "188 \t\t 0.011226\n",
      "189 \t\t 0.010967\n",
      "190 \t\t 0.010714\n",
      "191 \t\t 0.010466\n",
      "192 \t\t 0.010223\n",
      "193 \t\t 0.009985\n",
      "194 \t\t 0.009751\n",
      "195 \t\t 0.009522\n",
      "196 \t\t 0.009298\n",
      "197 \t\t 0.009078\n",
      "198 \t\t 0.008863\n",
      "199 \t\t 0.008652\n",
      "200 \t\t 0.008446\n",
      "201 \t\t 0.008244\n",
      "202 \t\t 0.008046\n",
      "203 \t\t 0.007852\n",
      "204 \t\t 0.007662\n",
      "205 \t\t 0.007476\n",
      "206 \t\t 0.007294\n",
      "207 \t\t 0.007116\n",
      "208 \t\t 0.006942\n",
      "209 \t\t 0.006771\n",
      "210 \t\t 0.006604\n",
      "211 \t\t 0.006441\n",
      "212 \t\t 0.006281\n",
      "213 \t\t 0.006125\n",
      "214 \t\t 0.005973\n",
      "215 \t\t 0.005823\n",
      "216 \t\t 0.005677\n",
      "217 \t\t 0.005535\n",
      "218 \t\t 0.005395\n",
      "219 \t\t 0.005259\n",
      "220 \t\t 0.005126\n",
      "221 \t\t 0.004996\n",
      "222 \t\t 0.004869\n",
      "223 \t\t 0.004745\n",
      "224 \t\t 0.004624\n",
      "225 \t\t 0.004506\n",
      "226 \t\t 0.004391\n",
      "227 \t\t 0.004278\n",
      "228 \t\t 0.004168\n",
      "229 \t\t 0.004061\n",
      "230 \t\t 0.003956\n",
      "231 \t\t 0.003854\n",
      "232 \t\t 0.003755\n",
      "233 \t\t 0.003658\n",
      "234 \t\t 0.003563\n",
      "235 \t\t 0.003470\n",
      "236 \t\t 0.003380\n",
      "237 \t\t 0.003293\n",
      "238 \t\t 0.003207\n",
      "239 \t\t 0.003123\n",
      "240 \t\t 0.003042\n",
      "241 \t\t 0.002963\n",
      "242 \t\t 0.002886\n",
      "243 \t\t 0.002810\n",
      "244 \t\t 0.002737\n",
      "245 \t\t 0.002666\n",
      "246 \t\t 0.002596\n",
      "247 \t\t 0.002528\n",
      "248 \t\t 0.002462\n",
      "249 \t\t 0.002398\n",
      "250 \t\t 0.002336\n"
     ]
    }
   ],
   "source": [
    "# Create user-by-item matrix - nothing to do here\n",
    "train_data_np = np.array(train_df)\n",
    "\n",
    "# Fit FunkSVD with the specified hyper parameters to the training data\n",
    "user_mat, offer_mat = FunkSVD(train_data_np, latent_features=15, learning_rate=0.005, iters=250)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Optimizaiton Statistics\n",
      "Iterations | Mean Squared Error \n",
      "1 \t\t 0.298519\n",
      "2 \t\t 0.265832\n",
      "3 \t\t 0.265351\n",
      "4 \t\t 0.264870\n",
      "5 \t\t 0.264385\n",
      "6 \t\t 0.263898\n",
      "7 \t\t 0.263406\n",
      "8 \t\t 0.262911\n",
      "9 \t\t 0.262412\n",
      "10 \t\t 0.261908\n",
      "11 \t\t 0.261400\n",
      "12 \t\t 0.260886\n",
      "13 \t\t 0.260367\n",
      "14 \t\t 0.259842\n",
      "15 \t\t 0.259310\n",
      "16 \t\t 0.258772\n",
      "17 \t\t 0.258227\n",
      "18 \t\t 0.257674\n",
      "19 \t\t 0.257113\n",
      "20 \t\t 0.256543\n",
      "21 \t\t 0.255963\n",
      "22 \t\t 0.255374\n",
      "23 \t\t 0.254774\n",
      "24 \t\t 0.254162\n",
      "25 \t\t 0.253538\n",
      "26 \t\t 0.252901\n",
      "27 \t\t 0.252250\n",
      "28 \t\t 0.251584\n",
      "29 \t\t 0.250902\n",
      "30 \t\t 0.250204\n",
      "31 \t\t 0.249488\n",
      "32 \t\t 0.248753\n",
      "33 \t\t 0.247997\n",
      "34 \t\t 0.247221\n",
      "35 \t\t 0.246421\n",
      "36 \t\t 0.245598\n",
      "37 \t\t 0.244750\n",
      "38 \t\t 0.243874\n",
      "39 \t\t 0.242971\n",
      "40 \t\t 0.242038\n",
      "41 \t\t 0.241073\n",
      "42 \t\t 0.240076\n",
      "43 \t\t 0.239044\n",
      "44 \t\t 0.237976\n",
      "45 \t\t 0.236870\n",
      "46 \t\t 0.235724\n",
      "47 \t\t 0.234537\n",
      "48 \t\t 0.233308\n",
      "49 \t\t 0.232034\n",
      "50 \t\t 0.230714\n",
      "51 \t\t 0.229345\n",
      "52 \t\t 0.227928\n",
      "53 \t\t 0.226460\n",
      "54 \t\t 0.224940\n",
      "55 \t\t 0.223366\n",
      "56 \t\t 0.221738\n",
      "57 \t\t 0.220054\n",
      "58 \t\t 0.218314\n",
      "59 \t\t 0.216516\n",
      "60 \t\t 0.214661\n",
      "61 \t\t 0.212748\n",
      "62 \t\t 0.210776\n",
      "63 \t\t 0.208747\n",
      "64 \t\t 0.206659\n",
      "65 \t\t 0.204515\n",
      "66 \t\t 0.202313\n",
      "67 \t\t 0.200056\n",
      "68 \t\t 0.197745\n",
      "69 \t\t 0.195382\n",
      "70 \t\t 0.192967\n",
      "71 \t\t 0.190503\n",
      "72 \t\t 0.187993\n",
      "73 \t\t 0.185438\n",
      "74 \t\t 0.182842\n",
      "75 \t\t 0.180207\n",
      "76 \t\t 0.177536\n",
      "77 \t\t 0.174834\n",
      "78 \t\t 0.172102\n",
      "79 \t\t 0.169345\n",
      "80 \t\t 0.166567\n",
      "81 \t\t 0.163770\n",
      "82 \t\t 0.160960\n",
      "83 \t\t 0.158140\n",
      "84 \t\t 0.155313\n",
      "85 \t\t 0.152484\n",
      "86 \t\t 0.149657\n",
      "87 \t\t 0.146836\n",
      "88 \t\t 0.144023\n",
      "89 \t\t 0.141223\n",
      "90 \t\t 0.138440\n",
      "91 \t\t 0.135676\n",
      "92 \t\t 0.132936\n",
      "93 \t\t 0.130221\n",
      "94 \t\t 0.127536\n",
      "95 \t\t 0.124882\n",
      "96 \t\t 0.122263\n",
      "97 \t\t 0.119679\n",
      "98 \t\t 0.117135\n",
      "99 \t\t 0.114630\n",
      "100 \t\t 0.112168\n",
      "101 \t\t 0.109749\n",
      "102 \t\t 0.107374\n",
      "103 \t\t 0.105045\n",
      "104 \t\t 0.102762\n",
      "105 \t\t 0.100526\n",
      "106 \t\t 0.098337\n",
      "107 \t\t 0.096196\n",
      "108 \t\t 0.094102\n",
      "109 \t\t 0.092056\n",
      "110 \t\t 0.090057\n",
      "111 \t\t 0.088106\n",
      "112 \t\t 0.086201\n",
      "113 \t\t 0.084342\n",
      "114 \t\t 0.082528\n",
      "115 \t\t 0.080760\n",
      "116 \t\t 0.079035\n",
      "117 \t\t 0.077354\n",
      "118 \t\t 0.075715\n",
      "119 \t\t 0.074117\n",
      "120 \t\t 0.072561\n",
      "121 \t\t 0.071043\n",
      "122 \t\t 0.069564\n",
      "123 \t\t 0.068123\n",
      "124 \t\t 0.066719\n",
      "125 \t\t 0.065350\n",
      "126 \t\t 0.064015\n",
      "127 \t\t 0.062715\n",
      "128 \t\t 0.061447\n",
      "129 \t\t 0.060211\n",
      "130 \t\t 0.059005\n",
      "131 \t\t 0.057829\n",
      "132 \t\t 0.056683\n",
      "133 \t\t 0.055564\n",
      "134 \t\t 0.054473\n",
      "135 \t\t 0.053408\n",
      "136 \t\t 0.052368\n",
      "137 \t\t 0.051353\n",
      "138 \t\t 0.050362\n",
      "139 \t\t 0.049395\n",
      "140 \t\t 0.048450\n",
      "141 \t\t 0.047527\n",
      "142 \t\t 0.046625\n",
      "143 \t\t 0.045744\n",
      "144 \t\t 0.044882\n",
      "145 \t\t 0.044040\n",
      "146 \t\t 0.043217\n",
      "147 \t\t 0.042412\n",
      "148 \t\t 0.041624\n",
      "149 \t\t 0.040854\n",
      "150 \t\t 0.040101\n",
      "151 \t\t 0.039363\n",
      "152 \t\t 0.038641\n",
      "153 \t\t 0.037935\n",
      "154 \t\t 0.037243\n",
      "155 \t\t 0.036566\n",
      "156 \t\t 0.035903\n",
      "157 \t\t 0.035254\n",
      "158 \t\t 0.034618\n",
      "159 \t\t 0.033994\n",
      "160 \t\t 0.033384\n",
      "161 \t\t 0.032785\n",
      "162 \t\t 0.032199\n",
      "163 \t\t 0.031624\n",
      "164 \t\t 0.031060\n",
      "165 \t\t 0.030507\n",
      "166 \t\t 0.029966\n",
      "167 \t\t 0.029434\n",
      "168 \t\t 0.028913\n",
      "169 \t\t 0.028402\n",
      "170 \t\t 0.027900\n",
      "171 \t\t 0.027408\n",
      "172 \t\t 0.026925\n",
      "173 \t\t 0.026451\n",
      "174 \t\t 0.025986\n",
      "175 \t\t 0.025530\n",
      "176 \t\t 0.025082\n",
      "177 \t\t 0.024642\n",
      "178 \t\t 0.024210\n",
      "179 \t\t 0.023786\n",
      "180 \t\t 0.023369\n",
      "181 \t\t 0.022960\n",
      "182 \t\t 0.022558\n",
      "183 \t\t 0.022164\n",
      "184 \t\t 0.021776\n",
      "185 \t\t 0.021395\n",
      "186 \t\t 0.021021\n",
      "187 \t\t 0.020653\n",
      "188 \t\t 0.020292\n",
      "189 \t\t 0.019937\n",
      "190 \t\t 0.019589\n",
      "191 \t\t 0.019246\n",
      "192 \t\t 0.018909\n",
      "193 \t\t 0.018578\n",
      "194 \t\t 0.018252\n",
      "195 \t\t 0.017932\n",
      "196 \t\t 0.017618\n",
      "197 \t\t 0.017308\n",
      "198 \t\t 0.017004\n",
      "199 \t\t 0.016706\n",
      "200 \t\t 0.016412\n",
      "201 \t\t 0.016123\n",
      "202 \t\t 0.015839\n",
      "203 \t\t 0.015560\n",
      "204 \t\t 0.015285\n",
      "205 \t\t 0.015015\n",
      "206 \t\t 0.014750\n",
      "207 \t\t 0.014489\n",
      "208 \t\t 0.014232\n",
      "209 \t\t 0.013980\n",
      "210 \t\t 0.013732\n",
      "211 \t\t 0.013488\n",
      "212 \t\t 0.013248\n",
      "213 \t\t 0.013012\n",
      "214 \t\t 0.012780\n",
      "215 \t\t 0.012552\n",
      "216 \t\t 0.012328\n",
      "217 \t\t 0.012107\n",
      "218 \t\t 0.011891\n",
      "219 \t\t 0.011677\n",
      "220 \t\t 0.011468\n",
      "221 \t\t 0.011262\n",
      "222 \t\t 0.011059\n",
      "223 \t\t 0.010860\n",
      "224 \t\t 0.010664\n",
      "225 \t\t 0.010472\n",
      "226 \t\t 0.010283\n",
      "227 \t\t 0.010097\n",
      "228 \t\t 0.009914\n",
      "229 \t\t 0.009734\n",
      "230 \t\t 0.009558\n",
      "231 \t\t 0.009384\n",
      "232 \t\t 0.009213\n",
      "233 \t\t 0.009046\n",
      "234 \t\t 0.008881\n",
      "235 \t\t 0.008719\n",
      "236 \t\t 0.008560\n",
      "237 \t\t 0.008403\n",
      "238 \t\t 0.008249\n",
      "239 \t\t 0.008098\n",
      "240 \t\t 0.007950\n",
      "241 \t\t 0.007804\n",
      "242 \t\t 0.007661\n",
      "243 \t\t 0.007520\n",
      "244 \t\t 0.007381\n",
      "245 \t\t 0.007246\n",
      "246 \t\t 0.007112\n",
      "247 \t\t 0.006981\n",
      "248 \t\t 0.006852\n",
      "249 \t\t 0.006725\n",
      "250 \t\t 0.006601\n"
     ]
    }
   ],
   "source": [
    "# Test for the best number of latent feature. (with latent features 10)\n",
    "user_mat_10, offer_mat_10 = FunkSVD(train_data_np, latent_features=10, learning_rate=0.005, iters=250)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Optimizaiton Statistics\n",
      "Iterations | Mean Squared Error \n",
      "1 \t\t 0.278910\n",
      "2 \t\t 0.268717\n",
      "3 \t\t 0.267871\n",
      "4 \t\t 0.267028\n",
      "5 \t\t 0.266186\n",
      "6 \t\t 0.265342\n",
      "7 \t\t 0.264497\n",
      "8 \t\t 0.263652\n",
      "9 \t\t 0.262805\n",
      "10 \t\t 0.261958\n",
      "11 \t\t 0.261110\n",
      "12 \t\t 0.260261\n",
      "13 \t\t 0.259410\n",
      "14 \t\t 0.258559\n",
      "15 \t\t 0.257706\n",
      "16 \t\t 0.256852\n",
      "17 \t\t 0.255997\n",
      "18 \t\t 0.255140\n",
      "19 \t\t 0.254282\n",
      "20 \t\t 0.253422\n",
      "21 \t\t 0.252561\n",
      "22 \t\t 0.251697\n",
      "23 \t\t 0.250832\n",
      "24 \t\t 0.249964\n",
      "25 \t\t 0.249094\n",
      "26 \t\t 0.248221\n",
      "27 \t\t 0.247346\n",
      "28 \t\t 0.246468\n",
      "29 \t\t 0.245587\n",
      "30 \t\t 0.244702\n",
      "31 \t\t 0.243814\n",
      "32 \t\t 0.242922\n",
      "33 \t\t 0.242026\n",
      "34 \t\t 0.241126\n",
      "35 \t\t 0.240221\n",
      "36 \t\t 0.239311\n",
      "37 \t\t 0.238395\n",
      "38 \t\t 0.237474\n",
      "39 \t\t 0.236547\n",
      "40 \t\t 0.235613\n",
      "41 \t\t 0.234672\n",
      "42 \t\t 0.233724\n",
      "43 \t\t 0.232768\n",
      "44 \t\t 0.231804\n",
      "45 \t\t 0.230831\n",
      "46 \t\t 0.229849\n",
      "47 \t\t 0.228856\n",
      "48 \t\t 0.227854\n",
      "49 \t\t 0.226840\n",
      "50 \t\t 0.225815\n",
      "51 \t\t 0.224778\n",
      "52 \t\t 0.223728\n",
      "53 \t\t 0.222664\n",
      "54 \t\t 0.221587\n",
      "55 \t\t 0.220494\n",
      "56 \t\t 0.219386\n",
      "57 \t\t 0.218261\n",
      "58 \t\t 0.217120\n",
      "59 \t\t 0.215960\n",
      "60 \t\t 0.214783\n",
      "61 \t\t 0.213586\n",
      "62 \t\t 0.212368\n",
      "63 \t\t 0.211130\n",
      "64 \t\t 0.209871\n",
      "65 \t\t 0.208589\n",
      "66 \t\t 0.207284\n",
      "67 \t\t 0.205956\n",
      "68 \t\t 0.204603\n",
      "69 \t\t 0.203224\n",
      "70 \t\t 0.201820\n",
      "71 \t\t 0.200390\n",
      "72 \t\t 0.198932\n",
      "73 \t\t 0.197447\n",
      "74 \t\t 0.195935\n",
      "75 \t\t 0.194393\n",
      "76 \t\t 0.192823\n",
      "77 \t\t 0.191224\n",
      "78 \t\t 0.189596\n",
      "79 \t\t 0.187938\n",
      "80 \t\t 0.186251\n",
      "81 \t\t 0.184534\n",
      "82 \t\t 0.182788\n",
      "83 \t\t 0.181014\n",
      "84 \t\t 0.179210\n",
      "85 \t\t 0.177379\n",
      "86 \t\t 0.175520\n",
      "87 \t\t 0.173634\n",
      "88 \t\t 0.171723\n",
      "89 \t\t 0.169786\n",
      "90 \t\t 0.167825\n",
      "91 \t\t 0.165842\n",
      "92 \t\t 0.163837\n",
      "93 \t\t 0.161811\n",
      "94 \t\t 0.159767\n",
      "95 \t\t 0.157706\n",
      "96 \t\t 0.155630\n",
      "97 \t\t 0.153539\n",
      "98 \t\t 0.151437\n",
      "99 \t\t 0.149325\n",
      "100 \t\t 0.147204\n",
      "101 \t\t 0.145078\n",
      "102 \t\t 0.142947\n",
      "103 \t\t 0.140813\n",
      "104 \t\t 0.138680\n",
      "105 \t\t 0.136548\n",
      "106 \t\t 0.134420\n",
      "107 \t\t 0.132298\n",
      "108 \t\t 0.130183\n",
      "109 \t\t 0.128078\n",
      "110 \t\t 0.125984\n",
      "111 \t\t 0.123903\n",
      "112 \t\t 0.121836\n",
      "113 \t\t 0.119786\n",
      "114 \t\t 0.117753\n",
      "115 \t\t 0.115739\n",
      "116 \t\t 0.113746\n",
      "117 \t\t 0.111774\n",
      "118 \t\t 0.109826\n",
      "119 \t\t 0.107901\n",
      "120 \t\t 0.106000\n",
      "121 \t\t 0.104126\n",
      "122 \t\t 0.102277\n",
      "123 \t\t 0.100456\n",
      "124 \t\t 0.098663\n",
      "125 \t\t 0.096897\n",
      "126 \t\t 0.095160\n",
      "127 \t\t 0.093452\n",
      "128 \t\t 0.091773\n",
      "129 \t\t 0.090123\n",
      "130 \t\t 0.088503\n",
      "131 \t\t 0.086912\n",
      "132 \t\t 0.085351\n",
      "133 \t\t 0.083820\n",
      "134 \t\t 0.082318\n",
      "135 \t\t 0.080845\n",
      "136 \t\t 0.079401\n",
      "137 \t\t 0.077986\n",
      "138 \t\t 0.076600\n",
      "139 \t\t 0.075243\n",
      "140 \t\t 0.073913\n",
      "141 \t\t 0.072612\n",
      "142 \t\t 0.071338\n",
      "143 \t\t 0.070091\n",
      "144 \t\t 0.068871\n",
      "145 \t\t 0.067677\n",
      "146 \t\t 0.066510\n",
      "147 \t\t 0.065368\n",
      "148 \t\t 0.064251\n",
      "149 \t\t 0.063159\n",
      "150 \t\t 0.062091\n",
      "151 \t\t 0.061048\n",
      "152 \t\t 0.060027\n",
      "153 \t\t 0.059030\n",
      "154 \t\t 0.058056\n",
      "155 \t\t 0.057104\n",
      "156 \t\t 0.056174\n",
      "157 \t\t 0.055265\n",
      "158 \t\t 0.054377\n",
      "159 \t\t 0.053510\n",
      "160 \t\t 0.052662\n",
      "161 \t\t 0.051835\n",
      "162 \t\t 0.051027\n",
      "163 \t\t 0.050237\n",
      "164 \t\t 0.049466\n",
      "165 \t\t 0.048713\n",
      "166 \t\t 0.047978\n",
      "167 \t\t 0.047260\n",
      "168 \t\t 0.046559\n",
      "169 \t\t 0.045875\n",
      "170 \t\t 0.045206\n",
      "171 \t\t 0.044553\n",
      "172 \t\t 0.043916\n",
      "173 \t\t 0.043294\n",
      "174 \t\t 0.042686\n",
      "175 \t\t 0.042092\n",
      "176 \t\t 0.041513\n",
      "177 \t\t 0.040947\n",
      "178 \t\t 0.040394\n",
      "179 \t\t 0.039854\n",
      "180 \t\t 0.039327\n",
      "181 \t\t 0.038812\n",
      "182 \t\t 0.038309\n",
      "183 \t\t 0.037817\n",
      "184 \t\t 0.037337\n",
      "185 \t\t 0.036868\n",
      "186 \t\t 0.036410\n",
      "187 \t\t 0.035962\n",
      "188 \t\t 0.035524\n",
      "189 \t\t 0.035096\n",
      "190 \t\t 0.034678\n",
      "191 \t\t 0.034269\n",
      "192 \t\t 0.033870\n",
      "193 \t\t 0.033479\n",
      "194 \t\t 0.033096\n",
      "195 \t\t 0.032722\n",
      "196 \t\t 0.032356\n",
      "197 \t\t 0.031998\n",
      "198 \t\t 0.031648\n",
      "199 \t\t 0.031305\n",
      "200 \t\t 0.030969\n",
      "201 \t\t 0.030641\n",
      "202 \t\t 0.030319\n",
      "203 \t\t 0.030004\n",
      "204 \t\t 0.029695\n",
      "205 \t\t 0.029392\n",
      "206 \t\t 0.029096\n",
      "207 \t\t 0.028805\n",
      "208 \t\t 0.028520\n",
      "209 \t\t 0.028241\n",
      "210 \t\t 0.027967\n",
      "211 \t\t 0.027698\n",
      "212 \t\t 0.027434\n",
      "213 \t\t 0.027175\n",
      "214 \t\t 0.026921\n",
      "215 \t\t 0.026672\n",
      "216 \t\t 0.026427\n",
      "217 \t\t 0.026186\n",
      "218 \t\t 0.025950\n",
      "219 \t\t 0.025718\n",
      "220 \t\t 0.025490\n",
      "221 \t\t 0.025265\n",
      "222 \t\t 0.025045\n",
      "223 \t\t 0.024828\n",
      "224 \t\t 0.024614\n",
      "225 \t\t 0.024404\n",
      "226 \t\t 0.024197\n",
      "227 \t\t 0.023994\n",
      "228 \t\t 0.023794\n",
      "229 \t\t 0.023596\n",
      "230 \t\t 0.023402\n",
      "231 \t\t 0.023211\n",
      "232 \t\t 0.023022\n",
      "233 \t\t 0.022837\n",
      "234 \t\t 0.022653\n",
      "235 \t\t 0.022473\n",
      "236 \t\t 0.022295\n",
      "237 \t\t 0.022119\n",
      "238 \t\t 0.021946\n",
      "239 \t\t 0.021775\n",
      "240 \t\t 0.021606\n",
      "241 \t\t 0.021440\n",
      "242 \t\t 0.021275\n",
      "243 \t\t 0.021113\n",
      "244 \t\t 0.020953\n",
      "245 \t\t 0.020795\n",
      "246 \t\t 0.020638\n",
      "247 \t\t 0.020484\n",
      "248 \t\t 0.020331\n",
      "249 \t\t 0.020180\n",
      "250 \t\t 0.020031\n"
     ]
    }
   ],
   "source": [
    "# Test for the best number of latent feature. (with latent features 5)\n",
    "user_mat_5, offer_mat_5 = FunkSVD(train_data_np, latent_features=5, learning_rate=0.005, iters=250)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [],
   "source": [
    "def predict_reaction(user_matrix, offer_matrix, user_id, offer_id):\n",
    "    '''\n",
    "    We can use this to make predictions for how users would react with the offers\n",
    "    \n",
    "    INPUT:\n",
    "    user_matrix - user by latent factor matrix\n",
    "    offer_matrix - latent factor by offer matrix\n",
    "    user_id - the user_id from the reviews df\n",
    "    offer_id - the offer_id according the offers df\n",
    "    \n",
    "    OUTPUT:\n",
    "    pred - the predicted reaction for user_id-offer_id according to FunkSVD\n",
    "    '''\n",
    "    try:\n",
    "        # Use the training data to create a series of users and offers that matches the ordering in training data\n",
    "        user_ids_series = np.array(train_df.index)\n",
    "        offer_ids_series = np.array(train_df.columns)\n",
    "\n",
    "        # User row and Offer Column\n",
    "        user_row = np.where(user_ids_series == user_id)[0][0]\n",
    "        offer_col = np.where(offer_ids_series == offer_id)[0][0]\n",
    "\n",
    "        # Take dot product of that row and column in U and V to make prediction\n",
    "        pred = np.dot(user_matrix[user_row, :], offer_matrix[:, offer_col])\n",
    "        \n",
    "        return pred\n",
    "    \n",
    "    except:\n",
    "        # print(\"Sorry, but a prediction cannot be made for this user-offer pair.  It looks like one of these items does not exist in our current database.\")\n",
    "    \n",
    "        return None"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [],
   "source": [
    "def validation(test_df, user_mat, offer_mat):\n",
    "    '''Measure the squared errors for the prediction'''\n",
    "    num_complete = np.count_nonzero(~np.isnan(test_df))\n",
    "    \n",
    "    sse_accum = 0\n",
    "    \n",
    "    for user_id in test_df.index:\n",
    "        for offer_id in test_df.columns:\n",
    "            if ~np.isnan(test_df.loc[user_id, offer_id]):\n",
    "                predict_value = predict_reaction(user_mat, offer_mat, user_id, offer_id)\n",
    "                if predict_value != None:\n",
    "                    # compute the error as the actual minus the dot product of the user and offer latent features\n",
    "                    diff = test_df.loc[user_id, offer_id] - predict_reaction(user_mat, offer_mat, user_id, offer_id)\n",
    "\n",
    "                    # keep track of the sum of squared errors for the matrix\n",
    "                    sse_accum += diff**2\n",
    "    \n",
    "    print(sse_accum / num_complete)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.00727034246899\n"
     ]
    }
   ],
   "source": [
    "# Evaluation for latent features of 15\n",
    "validation(test_df, user_mat, offer_mat)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.00905600476228\n"
     ]
    }
   ],
   "source": [
    "# Evaluation for latent features of 10\n",
    "validation(test_df, user_mat_10, offer_mat_10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.0208896664067\n"
     ]
    }
   ],
   "source": [
    "# Evaluation for latent features of 5\n",
    "validation(test_df, user_mat_5, offer_mat_5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [],
   "source": [
    "def best_sale_offer(user_item_matrix):\n",
    "    # Find out which offer sells the best\n",
    "\n",
    "    offer_gain = []\n",
    "    for offer_id in user_item_matrix.columns:\n",
    "        offer_gain.append([offer_id, transcript[(transcript['person'].isin(list(user_item_matrix[user_item_matrix[offer_id]>=1].index)))&(transcript['event']=='transaction')]['value'].apply(lambda x: x['amount']).sum()])\n",
    "\n",
    "    offer_value = pd.DataFrame(offer_gain, columns=['offer_id', 'gain'])\n",
    "    offer_value['gain'] = pd.to_numeric(offer_value['gain'])\n",
    "    offer_value.sort_values(by='gain', ascending=False, inplace=True)\n",
    "    print(offer_value)\n",
    "    \n",
    "    best_sale = offer_value.plot(kind='bar', title='Sale Amount of Different Offers')\n",
    "    best_sale.set_xlabel('Offer number')\n",
    "    best_sale.set_ylabel('Sale Amount')\n",
    "    \n",
    "    return offer_value"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "def recommend_offer(user_id, user_mat, offer_mat):\n",
    "    recommend = {}\n",
    "    for offer_id in train_df.columns:\n",
    "        pred_val = predict_reaction(user_mat, offer_mat, user_id, offer_id)\n",
    "        if pred_val != None:\n",
    "            recommend[offer_id] = pred_val\n",
    "        else:\n",
    "            break\n",
    "    if pred_val == None:\n",
    "        print(\"Because this user wasn't in our database, we are giving back the top best sale offer recommendations for all users.\")\n",
    "        top_offer = best_sale_offer(full_df)\n",
    "        for offer_id in top_offer['offer_id']:\n",
    "            print(\"offer id: \", offer_id)\n",
    "    else:\n",
    "        print(\"recommend offer for user \", user_id)\n",
    "        for offer_id, pred_val in sorted(recommend.items(), key=lambda kv:(kv[1], kv[0]), reverse=True):\n",
    "            print(\"offer id: \", offer_id, \" predicted value: \", round(pred_val,2))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "recommend offer for user  0610b486422d4921ae7d2bf64640c50b\n",
      "offer id:  2906b810c7d4411798c6938adc9daaa5  predicted value:  1.24\n",
      "offer id:  fafdcd668e3743c1bb461111dcafc2a4  predicted value:  1.15\n",
      "offer id:  0b1e1539f2cc45b7b9fa7c272da2e1d7  predicted value:  0.76\n",
      "offer id:  f19421c1d4aa40978ebb69ca19b0e20d  predicted value:  0.69\n",
      "offer id:  ae264e3637204a6fb9bb56bc8210ddfd  predicted value:  0.56\n",
      "offer id:  4d5c57ea9a6940dd891ad53e9dbe8da0  predicted value:  0.37\n",
      "offer id:  9b98b8c7a33c4b65b9aebfe6a799e6d9  predicted value:  0.02\n",
      "offer id:  2298d6c36e964ae4a3e7e9706d1fb8c2  predicted value:  -0.25\n"
     ]
    }
   ],
   "source": [
    "# recommend offer for a particular user\n",
    "recommend_offer('0610b486422d4921ae7d2bf64640c50b', user_mat, offer_mat)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Because this user wasn't in our database, we are giving back the top best sale offer recommendations for all users.\n",
      "                           offer_id       gain\n",
      "7  fafdcd668e3743c1bb461111dcafc2a4  596220.84\n",
      "1  2298d6c36e964ae4a3e7e9706d1fb8c2  538027.98\n",
      "6  f19421c1d4aa40978ebb69ca19b0e20d  480555.03\n",
      "3  4d5c57ea9a6940dd891ad53e9dbe8da0  441918.93\n",
      "5  ae264e3637204a6fb9bb56bc8210ddfd  423683.27\n",
      "2  2906b810c7d4411798c6938adc9daaa5  305568.42\n",
      "4  9b98b8c7a33c4b65b9aebfe6a799e6d9  304361.77\n",
      "0  0b1e1539f2cc45b7b9fa7c272da2e1d7  218905.94\n",
      "offer id:  fafdcd668e3743c1bb461111dcafc2a4\n",
      "offer id:  2298d6c36e964ae4a3e7e9706d1fb8c2\n",
      "offer id:  f19421c1d4aa40978ebb69ca19b0e20d\n",
      "offer id:  4d5c57ea9a6940dd891ad53e9dbe8da0\n",
      "offer id:  ae264e3637204a6fb9bb56bc8210ddfd\n",
      "offer id:  2906b810c7d4411798c6938adc9daaa5\n",
      "offer id:  9b98b8c7a33c4b65b9aebfe6a799e6d9\n",
      "offer id:  0b1e1539f2cc45b7b9fa7c272da2e1d7\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZsAAAETCAYAAADge6tNAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4wLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvpW3flQAAIABJREFUeJzt3X+8VVWd//HXW/wB/sCfaMglMMVSK1GvYOmUZSJqipmWNjOgUzE1WtY0M1LTd/yBlj7Gxh81aZikNP7ISUtMTRmNSlPhIv5CUlApbvgDARFEFPDz/WOtm4fruefue7mbA8f38/E4j7v32mvv9dnnwvnctfc6aysiMDMzK9Mm9Q7AzMwan5ONmZmVzsnGzMxK52RjZmalc7IxM7PSOdmYmVnpnGxsvZA0T9In6h3HhkzSeyXNlLRM0le7eYxvSfpxxfqnJM2XtFzSfj3RxoZA0pclvZDPa0dJB0uak9ePq3d89nZONlaYpEMk/UHSUkmLJd0n6cD11PZukt6U9MP10V53SJoq6QvrcIh/A6ZGxDYRcVkHx1+ZE8UrkmZIGidpi7Y6EfGdiKiM4SLg9IjYOiJmdtZGmSQdKqm1QL0PS7onn+dSSbdK2rti+2bAfwEj8nktAs4FfpDXf1neWVh3OdlYIZL6Ar8Cvg/sAAwAzgFeX08hjAaWACdVfrg2mEHArE7qnB4R2wD9gW8AJwG3S1LBYxZpoypJm3Znvy628SHgLuAWYFdgN+AR4D5J78nVdgF6sxGdlwER4Zdfnb6AZuDlGtt3B+4BFgEvAdcC21Vsnwd8Ii9vAowDns71bwR26KT9p4EvAy8AJ7TbFsA/AXOAZcD4HM/9wCv5+JtX1P8iMBdYDEwGds3lg/OxNq2oOxX4Ql4+BbiX1FtYAjwLHJm3nQ+sAVYCy0l/ZVc7j2NJH4ov52Pvlcvvabf/nlX2/WssFWXvBlYAn8zrZwP/A2yRjxPAq/n9e1sbud5FwJ/ze3sF0Ccf61CgFTgTeB74aS7/JPBwPoc/AB9s93v+F+BRYCnwM1Ji2Ap4DXgzt7287X1vdz6/B35YpfwOYFKO+dV8XsvzOT2dj/taLtsC2Ba4CngO+AtwHtCr4vd4H3Bx/jdwHrAH8Nsc80vAz+r9f67RXnUPwK+N4wX0JSWGa4Ajge3bbd8DODz/R+8H/A64pGL7PN5KNl8DHgCacv0fAdfXaPtvSD2o7Uk9q8nttgcpafQF9sl17wbekz90ngDG5Lofzx8m++e2vw/8Lm8bTOfJZhUpWfUiJb8FgNrX7eA82j4oDwc2I13SmktOhAX2r7o9v9cX5uWzgf9p997s0dExgEvye7cDsA1wK/DdvO1QYDVwYX6v+uT37UVgeH4PxuTf7RYVv+dppF7JDsBs4EsVx2utcX5bkpLhx6psOxV4rsbvaR7531de/yXp39VWwM45pn+s+D2uBr4CbJrP63rg30l/CPUGDqn3/7lGe/kymhUSEa8Ah5D+k18JLJQ0WdIuefvciJgSEa9HxELSNfWPdnC4fwT+PSJaI+J10gfkCTUuZ4wB7oiIJcB1wJGSdm5X58KIeCUiZgGPA3dFxDMRsZT0V/F+ud7fAhMj4qHc9jeBD0kaXPCt+FNEXBkRa0iJtz/psk4RnwVuy+/TKlKPog/w4YL7d2QB6YO9S/Klty8CX4+IxRGxDPgO6dJcmzeBs/Lv9bVc/0cR8WBErImIa0jJ/aCKfS6LiAURsZiUvIYWDGkH0of9c1W2PQfsVPC8diH9QfS1iHg1Il4k9WIqz2tBRHw/Ilbn81pFuhS3a0SsjIh7C8ZsBTnZWGERMTsiTomIJuD9pL9eLwGQtLOkGyT9RdIrpEs5HX04DAJ+IellSS+T/vpdQ5UPbUl9gBNJl+WIiPtJl3w+167qCxXLr1VZ3zov7wr8qeKclpN6bAM6Of02z1fsuyIvbt1B3fbat/0mML8LbXdkAOlyUFf1I/UmZlT8Ln6dy9ssjIiVFeuDgG+01c/7DCSdW5vnK5ZXUPz9WUJKbv2rbOtP6pEWMYjUc3yuIsYfkXo4bea32+ffAAHTJM2S9A8F27KCnGysWyLij8DVpKQD8F1Sr+eDEdEX+DvSf95q5pPudWxX8eodEX+pUvdTpMtjP5T0vKTnSR+uo7sZ+gLShxEAkrYCdiRd1381F29ZUf9dXTh2Z1Oot29bpA/qauddiKSBwAGkex1d9RIpEe9T8XvYNiIqk0P7c5oPnN/ud7dlRFxfoL2a709EvEq6z3Zilc2fIV0aLWI+qbe1U0WMfSNin45iiYjnI+KLEbErqef9Q0l7FGzPCnCysUIkvU/SNyQ15fWBwMmkey+QrvcvB16WNAD41xqHuwI4X9KgfKx+kkZ1UHcMMBH4AOlyzFDgYGCopA9041SuA06VNDSPavsO8GBEzMuX//4C/J2kXvmv2927cOwXSPeJOnIjcLSkw/Lw3W+QPhT/0NWTkLSlpI+SRm1NA27v6jFyz+pK4OK2y5KSBkg6osZuVwJfkjRcyVaSjpa0TYEmXwB2lLRtjTrjgDGSvippG0nbSzoP+BBp9GOR83qONKLte5L6StpE0u75/apK0olt/7ZJPawg9bathzjZWFHLSDeFH5T0KinJPE76wIT0QbA/aTTPbcDNNY51Kemm9F2SluVjDW9fKSetw0gDDZ6veM0gXe4Z09WTiIi7gf8H3ES6D7A7a1/L/yIpUS4iDTboSiK4lHTvaYmkt32HJSKeJPX4vk/qVRwDHBMRb3ShjR/k9+wF0iXMm4CROXF0x5mkQQoP5Muf/we8t6PKEdFCeo9+QPpQnku64d6p3Bu+HngmX97atUqde4EjgONJv58/ke63HRIRc4qfFqOBzUmDQ5YAP6f65bk2B5L+bS8n/ds8IyKe7UJ71om2UTRmZmalcc/GzMxK52RjZmalc7IxM7PSOdmYmVnpnGzMzKx0nu0022mnnWLw4MH1DsPMbKMyY8aMlyKiX2f1nGyywYMH09LSUu8wzMw2KpL+1HktX0YzM7P1wMnGzMxK52RjZmalK/WejaTtgB+TZgYO4B+AJ0lP7xtMeuDRZyJiSZ4B91LgKNK05KdExEP5OGOAb+fDnpefoYGkA0gzD/chTUR4RkSEpB2qtVHmuZrZO9eqVatobW1l5cqVnVfeSPXu3ZumpiY222yzbu1f9gCBS4FfR8QJkjYnTd3+LeDuiLhA0jjSLK9nkh52NCS/hgOXA8Nz4jiL9FjiID17Y3JOHpcDY0kTOd4OjCQ9KGtcB22YmfW41tZWttlmGwYPHkz6u7mxRASLFi2itbWV3XbbrVvHKO0ymqS+wEdIzwEnIt6IiJeBUaQnHJJ/HpeXRwGTInkA2E5Sf9IMsFPykwSXAFOAkXlb34i4P9JsopPaHataG2ZmPW7lypXsuOOODZloACSx4447rlPPrcx7Nu8BFgI/kTRT0o/zg6p2yc+baHvuRNvT8waw9tPzWnNZrfLWKuXUaMPMrBSNmmjarOv5lZlsNiU93+TyiNiP9BTEcTXqVzuT6EZ5YZLGSmqR1LJw4cKu7GpmttG64oormDRp0npts8x7Nq1Aa0Q8mNd/Tko2L0jqHxHP5UthL1bUH1ixfxPpMbqtwKHtyqfm8qYq9anRxloiYgIwAaC5ublQoho87rYi1bpk3gVH9/gxzax+evpzoqc/I770pS/16PGKKK1nExHPA/MltT317zDSU/Mm89YTFseQHmtLLh+dHzV7ELA0XwK7ExiRHw+7PTACuDNvWybpoDySbXS7Y1Vrw8ysIY0fP573ve99HH744Zx88slcdNFFXHnllRx44IHsu+++fPrTn2bFihUAnH322Vx00UUAHHrooZx55pkMGzaMPffck9///velxFf292y+Alwr6VHSs+O/A1wAHC5pDnB4Xoc0muwZ0mNmrwT+CSAiFgPjgen5dW4uA/gyaWj1XOBp0kg0arRhZtZwWlpauOmmm5g5cyY333zzX6feOv7445k+fTqPPPIIe+21F1dddVXV/VevXs20adO45JJLOOecc0qJsdShzxHxMGnIcnuHVakbwGkdHGciMLFKeQvpOzztyxdVa8PMrBHde++9jBo1ij59+gBwzDHHAPD444/z7W9/m5dffpnly5dzxBFHVN3/+OOPB+CAAw5g3rx5pcToiTjNzDZy6W/1tzvllFP45S9/yb777svVV1/N1KlTq9bbYostAOjVqxerV68uJUZPV2NmtpE75JBDuPXWW1m5ciXLly/nttvSAIVly5bRv39/Vq1axbXXXlvXGN2zMTPbyB144IEce+yx7LvvvgwaNIjm5ma23XZbxo8fz/Dhwxk0aBAf+MAHWLZsWd1iVEfdr3ea5ubmKPI8Gw99NrP2Zs+ezV577VXXGJYvX87WW2/NihUr+MhHPsKECRPYf//9e7SNaucpaUZEVLs3vxb3bMzMGsDYsWN54oknWLlyJWPGjOnxRLOunGzMzBrAddddV+8QavIAATMzK52TjZlZD2j0+9/ren5ONmZm66h3794sWrSoYRNO2/Nsevfu3e1j+J6Nmdk6ampqorW1lUaePb7tSZ3d5WRjZraONttss24/wfKdwpfRzMysdE42ZmZWOicbMzMrnZONmZmVzsnGzMxK52RjZmal89DnBuXZqc1sQ+KejZmZlc7JxszMSudkY2ZmpXOyMTOz0jnZmJlZ6ZxszMysdE42ZmZWOicbMzMrXanJRtI8SY9JelhSSy7bQdIUSXPyz+1zuSRdJmmupEcl7V9xnDG5/hxJYyrKD8jHn5v3Va02zMysPtZHz+ZjETE0Iprz+jjg7ogYAtyd1wGOBIbk11jgckiJAzgLGA4MA86qSB6X57pt+43spA0zM6uDelxGGwVck5evAY6rKJ8UyQPAdpL6A0cAUyJicUQsAaYAI/O2vhFxf6QHf09qd6xqbZiZWR2UnWwCuEvSDEljc9kuEfEcQP65cy4fAMyv2Lc1l9Uqb61SXqsNMzOrg7In4jw4IhZI2hmYIumPNeqqSll0o7ywnADHArz73e/uyq5mZtYFpfZsImJB/vki8AvSPZcX8iUw8s8Xc/VWYGDF7k3Agk7Km6qUU6ON9vFNiIjmiGju169fd0/TzMw6UVqykbSVpG3aloERwOPAZKBtRNkY4Ja8PBkYnUelHQQszZfA7gRGSNo+DwwYAdyZty2TdFAehTa63bGqtWFmZnVQ5mW0XYBf5NHImwLXRcSvJU0HbpT0eeDPwIm5/u3AUcBcYAVwKkBELJY0Hpie650bEYvz8peBq4E+wB35BXBBB22YmVkdlJZsIuIZYN8q5YuAw6qUB3BaB8eaCEysUt4CvL9oG2ZmVh+eQcDMzErnZGNmZqUre+izWU2Dx93W48ecd8HRPX5MM1s37tmYmVnpnGzMzKx0TjZmZlY6JxszMyudk42ZmZXOycbMzErnZGNmZqVzsjEzs9I52ZiZWemcbMzMrHRONmZmVjonGzMzK50n4jQrwBOGmq0b92zMzKx0TjZmZlY6JxszMyudk42ZmZXOycbMzErn0WhmDcSj5mxD5Z6NmZmVzsnGzMxK52RjZmalc7IxM7PSlZ5sJPWSNFPSr/L6bpIelDRH0s8kbZ7Lt8jrc/P2wRXH+GYuf1LSERXlI3PZXEnjKsqrtmFmZvWxPno2ZwCzK9YvBC6OiCHAEuDzufzzwJKI2AO4ONdD0t7AScA+wEjghzmB9QL+GzgS2Bs4Odet1YaZmdVBqclGUhNwNPDjvC7g48DPc5VrgOPy8qi8Tt5+WK4/CrghIl6PiGeBucCw/JobEc9ExBvADcCoTtowM7M66DTZSDqxSFkHLgH+DXgzr+8IvBwRq/N6KzAgLw8A5gPk7Utz/b+Wt9uno/JabbQ/j7GSWiS1LFy4sOApmZlZVxXp2XyzYNlaJH0SeDEiZlQWV6kanWzrqfK3F0ZMiIjmiGju169ftSpmZtYDOpxBQNKRwFHAAEmXVWzqC6yuvtdaDgaOlXQU0DvvdwmwnaRNc8+jCViQ67cCA4FWSZsC2wKLK8rbVO5TrfylGm2YmVkd1OrZLABagJXAjIrXZOCIGvsBEBHfjIimiBhMusF/T0T8LfAb4IRcbQxwS16enNfJ2++JiMjlJ+XRarsBQ4BpwHRgSB55tnluY3Lep6M2zMysDjrs2UTEI8Ajkq6LiFU92OaZwA2SzgNmAlfl8quAn0qaS+rRnJTjmCXpRuAJUo/qtIhYAyDpdOBOoBcwMSJmddKGmZnVQZGJOIdJOhsYlOsLiIh4T9FGImIqMDUvP0MaSda+zkqg6sCDiDgfOL9K+e3A7VXKq7ZhZmb1USTZXAV8nXQJbU254ZiZWSMqkmyWRsQdpUdiZmYNq0iy+Y2k/wRuBl5vK4yIh0qLyszMGkqRZDM8/2yuKAvSt/TNzMw61WmyiYiPrY9AzMyscXWabCT9R7XyiDi358MxM7NGVOQy2qsVy72BT7L2LM5mZmY1FbmM9r3KdUkXkb7Vb2ZmVkh3HjGwJVD4C51mZmZF7tk8xluzJvcC+gG+X2NmZoUVuWfzyYrl1cALFc+KMTMz61Snl9Ei4k/AdsAxwKdIj2A2MzMrrMiTOs8ArgV2zq9rJX2l7MDMzKxxFLmM9nlgeES8CiDpQuB+4PtlBmZmZo2jyGg0sfZsz2uo/uhlMzOzqor0bH4CPCjpF3n9OPwwMjMz64IiX+r8L0lTgUNIPZpTI2Jm2YGZmVnjKNKzAXiWNOx5U0CS9vcjBszMrKgiX+ocD5wCPM1bX+70IwbMzKywIj2bzwC7R8QbZQdjZmaNqchotMdJX+o0MzPrliI9m+8CMyU9ztqPhT62tKjMzKyhFEk21wAXAo8Bb5YbjpmZNaIiyealiLis9EjMzKxhFUk2MyR9l/TAtMrLaB76bGZmhRQZILAfcBDwHeB7+XVRZztJ6i1pmqRHJM2SdE4u303Sg5LmSPqZpM1z+RZ5fW7ePrjiWN/M5U9KOqKifGQumytpXEV51TbMzKw+ijxi4GPtX8DJBY79OvDxiNgXGAqMlHQQ6f7PxRExBFhCmuiT/HNJROwBXJzrIWlv4CRgH2Ak8ENJvST1Av4bOJL02IOTc11qtGFmZnVQdAYBJG0LfBr4HLAXMKBW/YgIYHle3Sy/2r4M+rlcfg1wNnA5MCovA/wc+IEk5fIbIuJ14FlJc4Fhud7ciHgmx3cDMErS7BptmNkGYPC423r0ePMuOLpHj2c9r2aykdQHOJb0wb0/sA1pIs7fFTl47n3MAPYg9UKeBl6ueNJnK28lrQHAfICIWC1pKbBjLn+g4rCV+8xvVz4879NRG2ZmhfR0QoR3dlLs8DKapGuBp4ARwA+AwaTLXFMjotAQ6IhYExFDgSZSb2SvatXamuxgW0+Vv42ksZJaJLUsXLiwWhUzM+sBte7ZvJ90v2M28MeIWEMHH9qdiYiXgamkgQbbSWrrUTUBC/JyKzAQIG/fFlhcWd5un47KX6rRRvu4JkREc0Q09+vXrzunZmZmBXSYbPKN/c8AfYH/k/R7YBtJ7ypyYEn9JG2Xl/sAnyAlrt8AJ+RqY4Bb8vLkvE7efk++7zMZOCmPVtsNGAJMA6YDQ/LIs81Jgwgm5306asPMzOqg5j2biPgj8B/Af0hqJo1CmyapNSI+3Mmx+wPX5Ps2mwA3RsSvJD0B3CDpPGAmbz2I7Srgp3kAwGJS8iAiZkm6EXiC9JiD03IvC0mnA3cCvYCJETErH+vMDtowM7M6KDwaLSJagBZJ/wJ8pED9R0nf0Wlf/gxvjSarLF8JnNjBsc4Hzq9Sfjtwe9E2zMysPgonmzb5MtVvS4jFzMwaVJEZBMzMzNaJk42ZmZWu02QjaRdJV0m6I6/vLcnTv5iZWWFFejZXk0Z87ZrXnwK+VlZAZmbWeIokm50i4kbyg9PyNDBrSo3KzMwaSpFk86qkHcmzB+SZm5eWGpWZmTWUIkOf/5n0Lf7dJd0H9OOtb+ebmZl1qtNkExEPSfoo8F7SJJdPRsSq0iMzM7OG0WGykXR8B5v2lERE3FxSTGZm1mBq9WyOqbEtACcbMzMrpMNkExGnrs9AzMyscRWaG03S0cA+QO+2sog4t6ygzMyssRSZQeAK4LPAV0gDBE4EBpUcl5mZNZAi37P5cESMJj0S+hzgQ6z9hEwzM7OaiiSb1/LPFZJ2BVYBu5UXkpmZNZoi92x+lR/v/J/AQ6SRaFeWGpWZmTWUIl/qHJ8Xb5L0K6B3RHi6GjMzK6zDy2iSDpT0ror10cCNwHhJO6yP4MzMrDHUumfzI+ANAEkfAS4AJpEm4ZxQfmhmZtYoal1G6xURi/PyZ4EJEXET6XLaw+WHZmZmjaJWz6aXpLZkdBhwT8W2Ql8GNTMzg9pJ43rgt5JeIg1//j2ApD3w82zMzKwLas2Ndr6ku4H+wF0REXnTJqTZBMzMrM4Gj7utx48574Kje/yYNS+HRcQDVcqe6vEozMysoRWZQcDMzGydlJZsJA2U9BtJsyXNknRGLt9B0hRJc/LP7XO5JF0maa6kRyXtX3GsMbn+HEljKsoPkPRY3ucySarVhpmZ1UeZPZvVwDciYi/gIOA0SXsD44C7I2IIcHdeBzgSGJJfY4HLISUO4CxgODAMOKsieVye67btNzKXd9SGmZnVQWnJJiKei4iH8vIyYDYwABgFXJOrXQMcl5dHAZMieQDYTlJ/4AhgSkQsjoglwBRgZN7WNyLuz4MXJrU7VrU2zMysDtbLPRtJg4H9gAeBXSLiOUgJCdg5VxsAzK/YrTWX1SpvrVJOjTbMzKwOSk82krYGbgK+FhGv1KpapSy6Ud6V2MZKapHUsnDhwq7samZmXVBqspG0GSnRXBsRN+fiF/IlMPLPF3N5K2s/lK0JWNBJeVOV8lptrCUiJkREc0Q09+vXr3snaWZmnSpzNJqAq4DZEfFfFZsmA20jysYAt1SUj86j0g4CluZLYHcCIyRtnwcGjADuzNuWSTootzW63bGqtWFmZnVQ5hxnBwN/DzxWMXHnt0izR98o6fPAn4ET87bbgaOAucAK4FSAiFgsaTwwPdc7t2KC0C8DVwN9gDvyixptmJlZHZSWbCLiXqrfV4E0sWf7+gGc1sGxJgITq5S3AO+vUr6oWhtmZlYfnkHAzMxK52RjZmalc7IxM7PSOdmYmVnpnGzMzKx0TjZmZlY6JxszMyudk42ZmZXOycbMzErnZGNmZqVzsjEzs9I52ZiZWemcbMzMrHRONmZmVjonGzMzK52TjZmZlc7JxszMSudkY2ZmpXOyMTOz0jnZmJlZ6ZxszMysdE42ZmZWOicbMzMrnZONmZmVzsnGzMxK52RjZmalKy3ZSJoo6UVJj1eU7SBpiqQ5+ef2uVySLpM0V9Kjkvav2GdMrj9H0piK8gMkPZb3uUySarVhZmb1U2bP5mpgZLuyccDdETEEuDuvAxwJDMmvscDlkBIHcBYwHBgGnFWRPC7Pddv2G9lJG2ZmVielJZuI+B2wuF3xKOCavHwNcFxF+aRIHgC2k9QfOAKYEhGLI2IJMAUYmbf1jYj7IyKASe2OVa0NMzOrk/V9z2aXiHgOIP/cOZcPAOZX1GvNZbXKW6uU12rDzMzqZEMZIKAqZdGN8q41Ko2V1CKpZeHChV3d3czMClrfyeaFfAmM/PPFXN4KDKyo1wQs6KS8qUp5rTbeJiImRERzRDT369ev2ydlZma1re9kMxloG1E2Brilonx0HpV2ELA0XwK7Exghafs8MGAEcGfetkzSQXkU2uh2x6rWhpmZ1cmmZR1Y0vXAocBOklpJo8ouAG6U9Hngz8CJufrtwFHAXGAFcCpARCyWNB6YnuudGxFtgw6+TBrx1ge4I7+o0YaZmdVJackmIk7uYNNhVeoGcFoHx5kITKxS3gK8v0r5omptmJlZ/WwoAwTMzKyBOdmYmVnpnGzMzKx0TjZmZlY6JxszMyudk42ZmZXOycbMzErnZGNmZqVzsjEzs9I52ZiZWemcbMzMrHRONmZmVjonGzMzK52TjZmZlc7JxszMSudkY2ZmpXOyMTOz0jnZmJlZ6ZxszMysdE42ZmZWOicbMzMrnZONmZmVzsnGzMxK52RjZmalc7IxM7PSOdmYmVnpGjbZSBop6UlJcyWNq3c8ZmbvZA2ZbCT1Av4bOBLYGzhZ0t71jcrM7J2rIZMNMAyYGxHPRMQbwA3AqDrHZGb2jqWIqHcMPU7SCcDIiPhCXv97YHhEnN6u3lhgbF59L/BkD4eyE/BSDx+zDBtDnBtDjOA4e5rj7FllxDkoIvp1VmnTHm50Q6EqZW/LqhExAZhQWhBSS0Q0l3X8nrIxxLkxxAiOs6c5zp5Vzzgb9TJaKzCwYr0JWFCnWMzM3vEaNdlMB4ZI2k3S5sBJwOQ6x2Rm9o7VkJfRImK1pNOBO4FewMSImFWHUEq7RNfDNoY4N4YYwXH2NMfZs+oWZ0MOEDAzsw1Lo15GMzOzDYiTjZmZlc7JxszMSudk00MkfVXSwM5rWmckDZfUNy/3kXSOpFslXShp23rHV0nSMEkH5uW9Jf2zpKPqHVctkg7JcY6odyztSXqfpMMkbd2ufGS9YuqMpEn1jqGa/F6eKekySZfm5b3qFo8HCPQMSUuBV4GngeuB/42IhfWNqmsknRoRP9kA4pgF7JtHFU4AVgA/Bw7L5cfXNcBM0lmk+fc2BaYAw4GpwCeAOyPi/PpF9xZJ0yJiWF7+InAa8AtgBHBrRFxQz/jaSPoqKbbZwFDgjIi4JW97KCL2r2d8OY72X6EQ8DHgHoCIOHa9B1WFpDOBk0lTdbXm4ibS10BuqMfv3Mmmh0iaCRxA+qD5LHAsMIOUeG6OiGV1DK8QSX+OiHdvAHHMjoi98vJaHzKSHo6IofWL7i2SHiN9KG4BPA80RcQrkvoAD0bEB+saYCZpZkTsl5enA0dFxEJJWwEPRMQH6hthkt/PD0XEckmDSX9g/DQiLq08h3qS9BDwBPBj0qwkIv0fPwkgIn5bv+jeIukpYJ+IWNWufHNgVkQMWd8xNeT3bOokIuJN4C7gLkmbkf7qPRm4COh07qD1QdKjHW0CdllV4vawAAAEBElEQVSfsdTweEUv6xFJzRHRImlPYFVnO69HqyNiDbBC0tMR8QpARLwm6c06x1ZpE0nbky6bq63HHRGvSlpd39DW0isilgNExDxJhwI/lzSI6lNQ1UMzcAbw78C/RsTDkl7bUJJMhTeBXYE/tSvvn7etd042PWet/wz5L4rJwOT8l+6GYhfgCGBJu3IBf1j/4VT1BeBSSd8mTRp4v6T5wPy8bUPxhqQtI2IFqVcLQL6vtCElm21JvWwBIeldEfF8vi+yoXyIAzwvaWhEPAyQezifBCYCG0TvK/9BebGk/80/X2DD/Bz9GnC3pDmk/zcA7wb2AE7vcK8S+TJaD5G0Z0Q8Ve84OiPpKuAnEXFvlW3XRcTn6hBWVZK2Ad5D+s/cGhEv1DmktUjaIiJer1K+E9A/Ih6rQ1iFSdoS2CUinq13LACSmki9xeerbDs4Iu6rQ1g1SToaODgivlXvWNqTtAnpcSsDSH9UtALTc298/cfjZGNmZmXz0GczMyudk42ZmZXOycasE5KaJN0iaY6kp/MX5Dav2H69pEclfT1/ke5hSTMl7V7PuNuTdIqkH9Q7DntncrIxq0GSgJuBX+bvJuwJbA2cn7e/C/hwRHwwIi4GjgNuiYj9IuLpgm30Kif6nrWxxGkbJicbs9o+Dqxsm1khj+T5OvAPeTTXXcDOuTdzFmnI6Rck/QZA0t9Jmpa3/6jtA1vScknnSnoQ+FBlg5Km5ql5pkl6StLf5PK1eiaSfpW/i9J2vAslzZD0f3kanamSnpFU+a32gZJ+LenJHG/bsbocp1lXONmY1bYP6Tsqf5W/vPln0ncWjgWejoihEXEOcAVwcUR8LM9D9VnS0NihwBrgb/NhtgIej4jh1YahA5vmKWa+BpxVZXt7WwFTI+IAYBlwHnA48Cng3Ip6w3IMQ4ETJTWvY5xmhWyIX0Yy25CINC1J0fJKh5G+7Dk9XY2jD/Bi3rYGuKnGvjfnnzOAwQXifAP4dV5+DHg9IlblKWAq958SEYsAJN0MHAKsXoc4zQpxsjGrbRbw6coCpRmpB5ImXd25xr4CromIb1bZtrKTL9e1fVl0DW/9P13N2lcjelcsr4q3vjT3Ztv+EfGmpMr/5+0TZNv8Xt2N06wQX0Yzq+1uYEtJo+GvN8m/B1ydp6npbN8TJO2c990hz/PVXfOAoZI2UXqcxbBuHOPwHEcf0mCG+0qI0+xtnGzMasi9hU+R7m/MAZ4CVgKdTk8SEU8A3yZNzPoo6TEE/dchnPuAZ0mXyS4CHurGMe4Ffgo8DNwUES0lxGn2Np6uxszMSueejZmZlc7JxszMSudkY2ZmpXOyMTOz0jnZmJlZ6ZxszMysdE42ZmZWOicbMzMr3f8Hc8T1+rSTKXYAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f16466957f0>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# we recommend top best sale offer to a new user\n",
    "recommend_offer('new_user', user_mat, offer_mat)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [],
   "source": [
    "def gender_offer(full_df):\n",
    "    '''\n",
    "    Find out which type of offer (bogo/discount) attracts most to different gender\n",
    "    Return with a plotted bar chart\n",
    "    '''\n",
    "    \n",
    "    # initialize the gender response and the gender sent dataframe\n",
    "    gender_response = pd.DataFrame(data=[[0,0],[0,0],[0,0]], index=['M', 'F', 'O'], columns=['bogo', 'discount'])\n",
    "    gender_sent = pd.DataFrame(data=[[0,0],[0,0],[0,0]], index=['M', 'F', 'O'], columns=['bogo', 'discount'])\n",
    "\n",
    "    # We count the number of offer completed base on the customer gender\n",
    "    for offer_type in ['bogo', 'discount']:\n",
    "        for offer_id in portfolio[portfolio['offer_type']==offer_type]['id']:\n",
    "            # count the total number of customers responsed our offers.\n",
    "            gender_response[offer_type] = gender_response[offer_type].add(profile[profile['id'].isin(full_df[full_df[offer_id]>0].index)]['gender'].value_counts(), fill_value=0)\n",
    "            # count the total number of offer sent\n",
    "            gender_sent[offer_type] = gender_sent[offer_type].add(profile[profile['id'].isin(full_df[full_df[offer_id]>=0].index)]['gender'].value_counts(), fill_value=0)\n",
    "        \n",
    "    gender_offer = (gender_response / gender_sent * 100).plot(kind='bar', title='Percentage of customer response')\n",
    "    gender_offer.set_xlabel(\"Gender\")\n",
    "    gender_offer.set_ylabel(\"Percentage\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f16465febe0>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "gender_offer(full_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "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>difficulty</th>\n",
       "      <th>duration</th>\n",
       "      <th>offer_id</th>\n",
       "      <th>reward</th>\n",
       "      <th>web</th>\n",
       "      <th>email</th>\n",
       "      <th>mobile</th>\n",
       "      <th>social</th>\n",
       "      <th>bogo</th>\n",
       "      <th>discount</th>\n",
       "      <th>informational</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>10</td>\n",
       "      <td>168</td>\n",
       "      <td>ae264e3637204a6fb9bb56bc8210ddfd</td>\n",
       "      <td>10</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>10</td>\n",
       "      <td>120</td>\n",
       "      <td>4d5c57ea9a6940dd891ad53e9dbe8da0</td>\n",
       "      <td>10</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>96</td>\n",
       "      <td>3f207df678b143eea3cee63160fa8bed</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>5</td>\n",
       "      <td>168</td>\n",
       "      <td>9b98b8c7a33c4b65b9aebfe6a799e6d9</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>20</td>\n",
       "      <td>240</td>\n",
       "      <td>0b1e1539f2cc45b7b9fa7c272da2e1d7</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>7</td>\n",
       "      <td>168</td>\n",
       "      <td>2298d6c36e964ae4a3e7e9706d1fb8c2</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>10</td>\n",
       "      <td>240</td>\n",
       "      <td>fafdcd668e3743c1bb461111dcafc2a4</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0</td>\n",
       "      <td>72</td>\n",
       "      <td>5a8bc65990b245e5a138643cd4eb9837</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>5</td>\n",
       "      <td>120</td>\n",
       "      <td>f19421c1d4aa40978ebb69ca19b0e20d</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>10</td>\n",
       "      <td>168</td>\n",
       "      <td>2906b810c7d4411798c6938adc9daaa5</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   difficulty  duration                          offer_id  reward  web  email  \\\n",
       "0          10       168  ae264e3637204a6fb9bb56bc8210ddfd      10    0      1   \n",
       "1          10       120  4d5c57ea9a6940dd891ad53e9dbe8da0      10    1      1   \n",
       "2           0        96  3f207df678b143eea3cee63160fa8bed       0    1      1   \n",
       "3           5       168  9b98b8c7a33c4b65b9aebfe6a799e6d9       5    1      1   \n",
       "4          20       240  0b1e1539f2cc45b7b9fa7c272da2e1d7       5    1      1   \n",
       "5           7       168  2298d6c36e964ae4a3e7e9706d1fb8c2       3    1      1   \n",
       "6          10       240  fafdcd668e3743c1bb461111dcafc2a4       2    1      1   \n",
       "7           0        72  5a8bc65990b245e5a138643cd4eb9837       0    0      1   \n",
       "8           5       120  f19421c1d4aa40978ebb69ca19b0e20d       5    1      1   \n",
       "9          10       168  2906b810c7d4411798c6938adc9daaa5       2    1      1   \n",
       "\n",
       "   mobile  social  bogo  discount  informational  \n",
       "0       1       1     1         0              0  \n",
       "1       1       1     1         0              0  \n",
       "2       1       0     0         0              1  \n",
       "3       1       0     1         0              0  \n",
       "4       0       0     0         1              0  \n",
       "5       1       1     0         1              0  \n",
       "6       1       1     0         1              0  \n",
       "7       1       1     0         0              1  \n",
       "8       1       1     1         0              0  \n",
       "9       1       0     0         1              0  "
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cleaned_portfolio"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [],
   "source": [
    "def gender_offer_channel(full_df):\n",
    "    '''\n",
    "    Find out which type of offer (bogo/discount) attracts most to different gender\n",
    "    Return with a plotted bar chart\n",
    "    '''\n",
    "    \n",
    "    # initialize the gender response and the gender sent dataframe\n",
    "    gender_response = pd.DataFrame(data=[[0,0,0,0],[0,0,0,0],[0,0,0,0]], index=['M', 'F', 'O'], columns=['web', 'email', 'mobile', 'social'])\n",
    "    gender_sent = pd.DataFrame(data=[[0,0,0,0],[0,0,0,0],[0,0,0,0]], index=['M', 'F', 'O'], columns=['web', 'email', 'mobile', 'social'])\n",
    "\n",
    "    # We count the number of offer completed base on the customer gender\n",
    "    for channel_type in ['web', 'email', 'mobile', 'social']:\n",
    "        for offer_id in cleaned_portfolio[(cleaned_portfolio[channel_type]==1)&(cleaned_portfolio['informational']!=1)]['offer_id']:\n",
    "            # count the total number of customers responsed our offers.\n",
    "            gender_response[channel_type] = gender_response[channel_type].add(profile[profile['id'].isin(full_df[full_df[offer_id]>0].index)]['gender'].value_counts(), fill_value=0)\n",
    "            # count the total number of offer sent\n",
    "            gender_sent[channel_type] = gender_sent[channel_type].add(profile[profile['id'].isin(full_df[full_df[offer_id]>=0].index)]['gender'].value_counts(), fill_value=0)\n",
    "        \n",
    "    gender_offer = (gender_response / gender_sent * 100).plot(kind='bar', title='Percentage of customer response')\n",
    "    gender_offer.set_xlabel(\"Gender\")\n",
    "    gender_offer.set_ylabel(\"Percentage\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f16465df198>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "gender_offer_channel(full_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [],
   "source": [
    "def best_sale_offer(user_item_matrix):\n",
    "    # Find out which offer sells the best\n",
    "\n",
    "    offer_gain = []\n",
    "    for offer_id in user_item_matrix.columns:\n",
    "        offer_gain.append([offer_id, transcript[(transcript['person'].isin(list(user_item_matrix[user_item_matrix[offer_id]>=1].index)))&(transcript['event']=='transaction')]['value'].apply(lambda x: x['amount']).sum()])\n",
    "    offer_value = pd.DataFrame(offer_gain, columns=['offer_id', 'gain'])\n",
    "    offer_value['gain'] = pd.to_numeric(offer_value['gain'])\n",
    "    offer_value.sort_values(by='gain', ascending=False, inplace=True)\n",
    "    print(offer_value)\n",
    "    \n",
    "    best_sale = offer_value.plot(kind='bar', title='Sale Amount of Different Offers')\n",
    "    best_sale.set_xlabel('Offer number')\n",
    "    best_sale.set_ylabel('Sale Amount')\n",
    "    \n",
    "    return offer_value"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                           offer_id       gain\n",
      "7  fafdcd668e3743c1bb461111dcafc2a4  596220.84\n",
      "1  2298d6c36e964ae4a3e7e9706d1fb8c2  538027.98\n",
      "6  f19421c1d4aa40978ebb69ca19b0e20d  480555.03\n",
      "3  4d5c57ea9a6940dd891ad53e9dbe8da0  441918.93\n",
      "5  ae264e3637204a6fb9bb56bc8210ddfd  423683.27\n",
      "2  2906b810c7d4411798c6938adc9daaa5  305568.42\n",
      "4  9b98b8c7a33c4b65b9aebfe6a799e6d9  304361.77\n",
      "0  0b1e1539f2cc45b7b9fa7c272da2e1d7  218905.94\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>offer_id</th>\n",
       "      <th>gain</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>fafdcd668e3743c1bb461111dcafc2a4</td>\n",
       "      <td>596220.84</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2298d6c36e964ae4a3e7e9706d1fb8c2</td>\n",
       "      <td>538027.98</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>f19421c1d4aa40978ebb69ca19b0e20d</td>\n",
       "      <td>480555.03</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4d5c57ea9a6940dd891ad53e9dbe8da0</td>\n",
       "      <td>441918.93</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>ae264e3637204a6fb9bb56bc8210ddfd</td>\n",
       "      <td>423683.27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2906b810c7d4411798c6938adc9daaa5</td>\n",
       "      <td>305568.42</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>9b98b8c7a33c4b65b9aebfe6a799e6d9</td>\n",
       "      <td>304361.77</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0b1e1539f2cc45b7b9fa7c272da2e1d7</td>\n",
       "      <td>218905.94</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                           offer_id       gain\n",
       "7  fafdcd668e3743c1bb461111dcafc2a4  596220.84\n",
       "1  2298d6c36e964ae4a3e7e9706d1fb8c2  538027.98\n",
       "6  f19421c1d4aa40978ebb69ca19b0e20d  480555.03\n",
       "3  4d5c57ea9a6940dd891ad53e9dbe8da0  441918.93\n",
       "5  ae264e3637204a6fb9bb56bc8210ddfd  423683.27\n",
       "2  2906b810c7d4411798c6938adc9daaa5  305568.42\n",
       "4  9b98b8c7a33c4b65b9aebfe6a799e6d9  304361.77\n",
       "0  0b1e1539f2cc45b7b9fa7c272da2e1d7  218905.94"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZsAAAETCAYAAADge6tNAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4wLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvpW3flQAAIABJREFUeJzt3X+8VVWd//HXW/wB/sCfaMglMMVSK1GvYOmUZSJqipmWNjOgUzE1WtY0M1LTd/yBlj7Gxh81aZikNP7ISUtMTRmNSlPhIv5CUlApbvgDARFEFPDz/WOtm4fruefue7mbA8f38/E4j7v32mvv9dnnwvnctfc6aysiMDMzK9Mm9Q7AzMwan5ONmZmVzsnGzMxK52RjZmalc7IxM7PSOdmYmVnpnGxsvZA0T9In6h3HhkzSeyXNlLRM0le7eYxvSfpxxfqnJM2XtFzSfj3RxoZA0pclvZDPa0dJB0uak9ePq3d89nZONlaYpEMk/UHSUkmLJd0n6cD11PZukt6U9MP10V53SJoq6QvrcIh/A6ZGxDYRcVkHx1+ZE8UrkmZIGidpi7Y6EfGdiKiM4SLg9IjYOiJmdtZGmSQdKqm1QL0PS7onn+dSSbdK2rti+2bAfwEj8nktAs4FfpDXf1neWVh3OdlYIZL6Ar8Cvg/sAAwAzgFeX08hjAaWACdVfrg2mEHArE7qnB4R2wD9gW8AJwG3S1LBYxZpoypJm3Znvy628SHgLuAWYFdgN+AR4D5J78nVdgF6sxGdlwER4Zdfnb6AZuDlGtt3B+4BFgEvAdcC21Vsnwd8Ii9vAowDns71bwR26KT9p4EvAy8AJ7TbFsA/AXOAZcD4HM/9wCv5+JtX1P8iMBdYDEwGds3lg/OxNq2oOxX4Ql4+BbiX1FtYAjwLHJm3nQ+sAVYCy0l/ZVc7j2NJH4ov52Pvlcvvabf/nlX2/WssFWXvBlYAn8zrZwP/A2yRjxPAq/n9e1sbud5FwJ/ze3sF0Ccf61CgFTgTeB74aS7/JPBwPoc/AB9s93v+F+BRYCnwM1Ji2Ap4DXgzt7287X1vdz6/B35YpfwOYFKO+dV8XsvzOT2dj/taLtsC2Ba4CngO+AtwHtCr4vd4H3Bx/jdwHrAH8Nsc80vAz+r9f67RXnUPwK+N4wX0JSWGa4Ajge3bbd8DODz/R+8H/A64pGL7PN5KNl8DHgCacv0fAdfXaPtvSD2o7Uk9q8nttgcpafQF9sl17wbekz90ngDG5Lofzx8m++e2vw/8Lm8bTOfJZhUpWfUiJb8FgNrX7eA82j4oDwc2I13SmktOhAX2r7o9v9cX5uWzgf9p997s0dExgEvye7cDsA1wK/DdvO1QYDVwYX6v+uT37UVgeH4PxuTf7RYVv+dppF7JDsBs4EsVx2utcX5bkpLhx6psOxV4rsbvaR7531de/yXp39VWwM45pn+s+D2uBr4CbJrP63rg30l/CPUGDqn3/7lGe/kymhUSEa8Ah5D+k18JLJQ0WdIuefvciJgSEa9HxELSNfWPdnC4fwT+PSJaI+J10gfkCTUuZ4wB7oiIJcB1wJGSdm5X58KIeCUiZgGPA3dFxDMRsZT0V/F+ud7fAhMj4qHc9jeBD0kaXPCt+FNEXBkRa0iJtz/psk4RnwVuy+/TKlKPog/w4YL7d2QB6YO9S/Klty8CX4+IxRGxDPgO6dJcmzeBs/Lv9bVc/0cR8WBErImIa0jJ/aCKfS6LiAURsZiUvIYWDGkH0of9c1W2PQfsVPC8diH9QfS1iHg1Il4k9WIqz2tBRHw/Ilbn81pFuhS3a0SsjIh7C8ZsBTnZWGERMTsiTomIJuD9pL9eLwGQtLOkGyT9RdIrpEs5HX04DAJ+IellSS+T/vpdQ5UPbUl9gBNJl+WIiPtJl3w+167qCxXLr1VZ3zov7wr8qeKclpN6bAM6Of02z1fsuyIvbt1B3fbat/0mML8LbXdkAOlyUFf1I/UmZlT8Ln6dy9ssjIiVFeuDgG+01c/7DCSdW5vnK5ZXUPz9WUJKbv2rbOtP6pEWMYjUc3yuIsYfkXo4bea32+ffAAHTJM2S9A8F27KCnGysWyLij8DVpKQD8F1Sr+eDEdEX+DvSf95q5pPudWxX8eodEX+pUvdTpMtjP5T0vKTnSR+uo7sZ+gLShxEAkrYCdiRd1381F29ZUf9dXTh2Z1Oot29bpA/qauddiKSBwAGkex1d9RIpEe9T8XvYNiIqk0P7c5oPnN/ud7dlRFxfoL2a709EvEq6z3Zilc2fIV0aLWI+qbe1U0WMfSNin45iiYjnI+KLEbErqef9Q0l7FGzPCnCysUIkvU/SNyQ15fWBwMmkey+QrvcvB16WNAD41xqHuwI4X9KgfKx+kkZ1UHcMMBH4AOlyzFDgYGCopA9041SuA06VNDSPavsO8GBEzMuX//4C/J2kXvmv2927cOwXSPeJOnIjcLSkw/Lw3W+QPhT/0NWTkLSlpI+SRm1NA27v6jFyz+pK4OK2y5KSBkg6osZuVwJfkjRcyVaSjpa0TYEmXwB2lLRtjTrjgDGSvippG0nbSzoP+BBp9GOR83qONKLte5L6StpE0u75/apK0olt/7ZJPawg9bathzjZWFHLSDeFH5T0KinJPE76wIT0QbA/aTTPbcDNNY51Kemm9F2SluVjDW9fKSetw0gDDZ6veM0gXe4Z09WTiIi7gf8H3ES6D7A7a1/L/yIpUS4iDTboSiK4lHTvaYmkt32HJSKeJPX4vk/qVRwDHBMRb3ShjR/k9+wF0iXMm4CROXF0x5mkQQoP5Muf/we8t6PKEdFCeo9+QPpQnku64d6p3Bu+HngmX97atUqde4EjgONJv58/ke63HRIRc4qfFqOBzUmDQ5YAP6f65bk2B5L+bS8n/ds8IyKe7UJ71om2UTRmZmalcc/GzMxK52RjZmalc7IxM7PSOdmYmVnpnGzMzKx0nu0022mnnWLw4MH1DsPMbKMyY8aMlyKiX2f1nGyywYMH09LSUu8wzMw2KpL+1HktX0YzM7P1wMnGzMxK52RjZmalK/WejaTtgB+TZgYO4B+AJ0lP7xtMeuDRZyJiSZ4B91LgKNK05KdExEP5OGOAb+fDnpefoYGkA0gzD/chTUR4RkSEpB2qtVHmuZrZO9eqVatobW1l5cqVnVfeSPXu3ZumpiY222yzbu1f9gCBS4FfR8QJkjYnTd3+LeDuiLhA0jjSLK9nkh52NCS/hgOXA8Nz4jiL9FjiID17Y3JOHpcDY0kTOd4OjCQ9KGtcB22YmfW41tZWttlmGwYPHkz6u7mxRASLFi2itbWV3XbbrVvHKO0ymqS+wEdIzwEnIt6IiJeBUaQnHJJ/HpeXRwGTInkA2E5Sf9IMsFPykwSXAFOAkXlb34i4P9JsopPaHataG2ZmPW7lypXsuOOODZloACSx4447rlPPrcx7Nu8BFgI/kTRT0o/zg6p2yc+baHvuRNvT8waw9tPzWnNZrfLWKuXUaMPMrBSNmmjarOv5lZlsNiU93+TyiNiP9BTEcTXqVzuT6EZ5YZLGSmqR1LJw4cKu7GpmttG64oormDRp0npts8x7Nq1Aa0Q8mNd/Tko2L0jqHxHP5UthL1bUH1ixfxPpMbqtwKHtyqfm8qYq9anRxloiYgIwAaC5ublQoho87rYi1bpk3gVH9/gxzax+evpzoqc/I770pS/16PGKKK1nExHPA/MltT317zDSU/Mm89YTFseQHmtLLh+dHzV7ELA0XwK7ExiRHw+7PTACuDNvWybpoDySbXS7Y1Vrw8ysIY0fP573ve99HH744Zx88slcdNFFXHnllRx44IHsu+++fPrTn2bFihUAnH322Vx00UUAHHrooZx55pkMGzaMPffck9///velxFf292y+Alwr6VHSs+O/A1wAHC5pDnB4Xoc0muwZ0mNmrwT+CSAiFgPjgen5dW4uA/gyaWj1XOBp0kg0arRhZtZwWlpauOmmm5g5cyY333zzX6feOv7445k+fTqPPPIIe+21F1dddVXV/VevXs20adO45JJLOOecc0qJsdShzxHxMGnIcnuHVakbwGkdHGciMLFKeQvpOzztyxdVa8PMrBHde++9jBo1ij59+gBwzDHHAPD444/z7W9/m5dffpnly5dzxBFHVN3/+OOPB+CAAw5g3rx5pcToiTjNzDZy6W/1tzvllFP45S9/yb777svVV1/N1KlTq9bbYostAOjVqxerV68uJUZPV2NmtpE75JBDuPXWW1m5ciXLly/nttvSAIVly5bRv39/Vq1axbXXXlvXGN2zMTPbyB144IEce+yx7LvvvgwaNIjm5ma23XZbxo8fz/Dhwxk0aBAf+MAHWLZsWd1iVEfdr3ea5ubmKPI8Gw99NrP2Zs+ezV577VXXGJYvX87WW2/NihUr+MhHPsKECRPYf//9e7SNaucpaUZEVLs3vxb3bMzMGsDYsWN54oknWLlyJWPGjOnxRLOunGzMzBrAddddV+8QavIAATMzK52TjZlZD2j0+9/ren5ONmZm66h3794sWrSoYRNO2/Nsevfu3e1j+J6Nmdk6ampqorW1lUaePb7tSZ3d5WRjZraONttss24/wfKdwpfRzMysdE42ZmZWOicbMzMrnZONmZmVzsnGzMxK52RjZmal89DnBuXZqc1sQ+KejZmZlc7JxszMSudkY2ZmpXOyMTOz0jnZmJlZ6ZxszMysdE42ZmZWOicbMzMrXanJRtI8SY9JelhSSy7bQdIUSXPyz+1zuSRdJmmupEcl7V9xnDG5/hxJYyrKD8jHn5v3Va02zMysPtZHz+ZjETE0Iprz+jjg7ogYAtyd1wGOBIbk11jgckiJAzgLGA4MA86qSB6X57pt+43spA0zM6uDelxGGwVck5evAY6rKJ8UyQPAdpL6A0cAUyJicUQsAaYAI/O2vhFxf6QHf09qd6xqbZiZWR2UnWwCuEvSDEljc9kuEfEcQP65cy4fAMyv2Lc1l9Uqb61SXqsNMzOrg7In4jw4IhZI2hmYIumPNeqqSll0o7ywnADHArz73e/uyq5mZtYFpfZsImJB/vki8AvSPZcX8iUw8s8Xc/VWYGDF7k3Agk7Km6qUU6ON9vFNiIjmiGju169fd0/TzMw6UVqykbSVpG3aloERwOPAZKBtRNkY4Ja8PBkYnUelHQQszZfA7gRGSNo+DwwYAdyZty2TdFAehTa63bGqtWFmZnVQ5mW0XYBf5NHImwLXRcSvJU0HbpT0eeDPwIm5/u3AUcBcYAVwKkBELJY0Hpie650bEYvz8peBq4E+wB35BXBBB22YmVkdlJZsIuIZYN8q5YuAw6qUB3BaB8eaCEysUt4CvL9oG2ZmVh+eQcDMzErnZGNmZqUre+izWU2Dx93W48ecd8HRPX5MM1s37tmYmVnpnGzMzKx0TjZmZlY6JxszMyudk42ZmZXOycbMzErnZGNmZqVzsjEzs9I52ZiZWemcbMzMrHRONmZmVjonGzMzK50n4jQrwBOGmq0b92zMzKx0TjZmZlY6JxszMyudk42ZmZXOycbMzErn0WhmDcSj5mxD5Z6NmZmVzsnGzMxK52RjZmalc7IxM7PSlZ5sJPWSNFPSr/L6bpIelDRH0s8kbZ7Lt8jrc/P2wRXH+GYuf1LSERXlI3PZXEnjKsqrtmFmZvWxPno2ZwCzK9YvBC6OiCHAEuDzufzzwJKI2AO4ONdD0t7AScA+wEjghzmB9QL+GzgS2Bs4Odet1YaZmdVBqclGUhNwNPDjvC7g48DPc5VrgOPy8qi8Tt5+WK4/CrghIl6PiGeBucCw/JobEc9ExBvADcCoTtowM7M66DTZSDqxSFkHLgH+DXgzr+8IvBwRq/N6KzAgLw8A5gPk7Utz/b+Wt9uno/JabbQ/j7GSWiS1LFy4sOApmZlZVxXp2XyzYNlaJH0SeDEiZlQWV6kanWzrqfK3F0ZMiIjmiGju169ftSpmZtYDOpxBQNKRwFHAAEmXVWzqC6yuvtdaDgaOlXQU0DvvdwmwnaRNc8+jCViQ67cCA4FWSZsC2wKLK8rbVO5TrfylGm2YmVkd1OrZLABagJXAjIrXZOCIGvsBEBHfjIimiBhMusF/T0T8LfAb4IRcbQxwS16enNfJ2++JiMjlJ+XRarsBQ4BpwHRgSB55tnluY3Lep6M2zMysDjrs2UTEI8Ajkq6LiFU92OaZwA2SzgNmAlfl8quAn0qaS+rRnJTjmCXpRuAJUo/qtIhYAyDpdOBOoBcwMSJmddKGmZnVQZGJOIdJOhsYlOsLiIh4T9FGImIqMDUvP0MaSda+zkqg6sCDiDgfOL9K+e3A7VXKq7ZhZmb1USTZXAV8nXQJbU254ZiZWSMqkmyWRsQdpUdiZmYNq0iy+Y2k/wRuBl5vK4yIh0qLyszMGkqRZDM8/2yuKAvSt/TNzMw61WmyiYiPrY9AzMyscXWabCT9R7XyiDi358MxM7NGVOQy2qsVy72BT7L2LM5mZmY1FbmM9r3KdUkXkb7Vb2ZmVkh3HjGwJVD4C51mZmZF7tk8xluzJvcC+gG+X2NmZoUVuWfzyYrl1cALFc+KMTMz61Snl9Ei4k/AdsAxwKdIj2A2MzMrrMiTOs8ArgV2zq9rJX2l7MDMzKxxFLmM9nlgeES8CiDpQuB+4PtlBmZmZo2jyGg0sfZsz2uo/uhlMzOzqor0bH4CPCjpF3n9OPwwMjMz64IiX+r8L0lTgUNIPZpTI2Jm2YGZmVnjKNKzAXiWNOx5U0CS9vcjBszMrKgiX+ocD5wCPM1bX+70IwbMzKywIj2bzwC7R8QbZQdjZmaNqchotMdJX+o0MzPrliI9m+8CMyU9ztqPhT62tKjMzKyhFEk21wAXAo8Bb5YbjpmZNaIiyealiLis9EjMzKxhFUk2MyR9l/TAtMrLaB76bGZmhRQZILAfcBDwHeB7+XVRZztJ6i1pmqRHJM2SdE4u303Sg5LmSPqZpM1z+RZ5fW7ePrjiWN/M5U9KOqKifGQumytpXEV51TbMzKw+ijxi4GPtX8DJBY79OvDxiNgXGAqMlHQQ6f7PxRExBFhCmuiT/HNJROwBXJzrIWlv4CRgH2Ak8ENJvST1Av4bOJL02IOTc11qtGFmZnVQdAYBJG0LfBr4HLAXMKBW/YgIYHle3Sy/2r4M+rlcfg1wNnA5MCovA/wc+IEk5fIbIuJ14FlJc4Fhud7ciHgmx3cDMErS7BptmNkGYPC423r0ePMuOLpHj2c9r2aykdQHOJb0wb0/sA1pIs7fFTl47n3MAPYg9UKeBl6ueNJnK28lrQHAfICIWC1pKbBjLn+g4rCV+8xvVz4879NRG2ZmhfR0QoR3dlLs8DKapGuBp4ARwA+AwaTLXFMjotAQ6IhYExFDgSZSb2SvatXamuxgW0+Vv42ksZJaJLUsXLiwWhUzM+sBte7ZvJ90v2M28MeIWEMHH9qdiYiXgamkgQbbSWrrUTUBC/JyKzAQIG/fFlhcWd5un47KX6rRRvu4JkREc0Q09+vXrzunZmZmBXSYbPKN/c8AfYH/k/R7YBtJ7ypyYEn9JG2Xl/sAnyAlrt8AJ+RqY4Bb8vLkvE7efk++7zMZOCmPVtsNGAJMA6YDQ/LIs81Jgwgm5306asPMzOqg5j2biPgj8B/Af0hqJo1CmyapNSI+3Mmx+wPX5Ps2mwA3RsSvJD0B3CDpPGAmbz2I7Srgp3kAwGJS8iAiZkm6EXiC9JiD03IvC0mnA3cCvYCJETErH+vMDtowM7M6KDwaLSJagBZJ/wJ8pED9R0nf0Wlf/gxvjSarLF8JnNjBsc4Hzq9Sfjtwe9E2zMysPgonmzb5MtVvS4jFzMwaVJEZBMzMzNaJk42ZmZWu02QjaRdJV0m6I6/vLcnTv5iZWWFFejZXk0Z87ZrXnwK+VlZAZmbWeIokm50i4kbyg9PyNDBrSo3KzMwaSpFk86qkHcmzB+SZm5eWGpWZmTWUIkOf/5n0Lf7dJd0H9OOtb+ebmZl1qtNkExEPSfoo8F7SJJdPRsSq0iMzM7OG0WGykXR8B5v2lERE3FxSTGZm1mBq9WyOqbEtACcbMzMrpMNkExGnrs9AzMyscRWaG03S0cA+QO+2sog4t6ygzMyssRSZQeAK4LPAV0gDBE4EBpUcl5mZNZAi37P5cESMJj0S+hzgQ6z9hEwzM7OaiiSb1/LPFZJ2BVYBu5UXkpmZNZoi92x+lR/v/J/AQ6SRaFeWGpWZmTWUIl/qHJ8Xb5L0K6B3RHi6GjMzK6zDy2iSDpT0ror10cCNwHhJO6yP4MzMrDHUumfzI+ANAEkfAS4AJpEm4ZxQfmhmZtYoal1G6xURi/PyZ4EJEXET6XLaw+WHZmZmjaJWz6aXpLZkdBhwT8W2Ql8GNTMzg9pJ43rgt5JeIg1//j2ApD3w82zMzKwLas2Ndr6ku4H+wF0REXnTJqTZBMzMrM4Gj7utx48574Kje/yYNS+HRcQDVcqe6vEozMysoRWZQcDMzGydlJZsJA2U9BtJsyXNknRGLt9B0hRJc/LP7XO5JF0maa6kRyXtX3GsMbn+HEljKsoPkPRY3ucySarVhpmZ1UeZPZvVwDciYi/gIOA0SXsD44C7I2IIcHdeBzgSGJJfY4HLISUO4CxgODAMOKsieVye67btNzKXd9SGmZnVQWnJJiKei4iH8vIyYDYwABgFXJOrXQMcl5dHAZMieQDYTlJ/4AhgSkQsjoglwBRgZN7WNyLuz4MXJrU7VrU2zMysDtbLPRtJg4H9gAeBXSLiOUgJCdg5VxsAzK/YrTWX1SpvrVJOjTbMzKwOSk82krYGbgK+FhGv1KpapSy6Ud6V2MZKapHUsnDhwq7samZmXVBqspG0GSnRXBsRN+fiF/IlMPLPF3N5K2s/lK0JWNBJeVOV8lptrCUiJkREc0Q09+vXr3snaWZmnSpzNJqAq4DZEfFfFZsmA20jysYAt1SUj86j0g4CluZLYHcCIyRtnwcGjADuzNuWSTootzW63bGqtWFmZnVQ5hxnBwN/DzxWMXHnt0izR98o6fPAn4ET87bbgaOAucAK4FSAiFgsaTwwPdc7t2KC0C8DVwN9gDvyixptmJlZHZSWbCLiXqrfV4E0sWf7+gGc1sGxJgITq5S3AO+vUr6oWhtmZlYfnkHAzMxK52RjZmalc7IxM7PSOdmYmVnpnGzMzKx0TjZmZlY6JxszMyudk42ZmZXOycbMzErnZGNmZqVzsjEzs9I52ZiZWemcbMzMrHRONmZmVjonGzMzK52TjZmZlc7JxszMSudkY2ZmpXOyMTOz0jnZmJlZ6ZxszMysdE42ZmZWOicbMzMrnZONmZmVzsnGzMxK52RjZmalKy3ZSJoo6UVJj1eU7SBpiqQ5+ef2uVySLpM0V9Kjkvav2GdMrj9H0piK8gMkPZb3uUySarVhZmb1U2bP5mpgZLuyccDdETEEuDuvAxwJDMmvscDlkBIHcBYwHBgGnFWRPC7Pddv2G9lJG2ZmVielJZuI+B2wuF3xKOCavHwNcFxF+aRIHgC2k9QfOAKYEhGLI2IJMAUYmbf1jYj7IyKASe2OVa0NMzOrk/V9z2aXiHgOIP/cOZcPAOZX1GvNZbXKW6uU12rDzMzqZEMZIKAqZdGN8q41Ko2V1CKpZeHChV3d3czMClrfyeaFfAmM/PPFXN4KDKyo1wQs6KS8qUp5rTbeJiImRERzRDT369ev2ydlZma1re9kMxloG1E2Brilonx0HpV2ELA0XwK7Exghafs8MGAEcGfetkzSQXkU2uh2x6rWhpmZ1cmmZR1Y0vXAocBOklpJo8ouAG6U9Hngz8CJufrtwFHAXGAFcCpARCyWNB6YnuudGxFtgw6+TBrx1ge4I7+o0YaZmdVJackmIk7uYNNhVeoGcFoHx5kITKxS3gK8v0r5omptmJlZ/WwoAwTMzKyBOdmYmVnpnGzMzKx0TjZmZlY6JxszMyudk42ZmZXOycbMzErnZGNmZqVzsjEzs9I52ZiZWemcbMzMrHRONmZmVjonGzMzK52TjZmZlc7JxszMSudkY2ZmpXOyMTOz0jnZmJlZ6ZxszMysdE42ZmZWOicbMzMrnZONmZmVzsnGzMxK52RjZmalc7IxM7PSOdmYmVnpGjbZSBop6UlJcyWNq3c8ZmbvZA2ZbCT1Av4bOBLYGzhZ0t71jcrM7J2rIZMNMAyYGxHPRMQbwA3AqDrHZGb2jqWIqHcMPU7SCcDIiPhCXv97YHhEnN6u3lhgbF59L/BkD4eyE/BSDx+zDBtDnBtDjOA4e5rj7FllxDkoIvp1VmnTHm50Q6EqZW/LqhExAZhQWhBSS0Q0l3X8nrIxxLkxxAiOs6c5zp5Vzzgb9TJaKzCwYr0JWFCnWMzM3vEaNdlMB4ZI2k3S5sBJwOQ6x2Rm9o7VkJfRImK1pNOBO4FewMSImFWHUEq7RNfDNoY4N4YYwXH2NMfZs+oWZ0MOEDAzsw1Lo15GMzOzDYiTjZmZlc7JxszMSudk00MkfVXSwM5rWmckDZfUNy/3kXSOpFslXShp23rHV0nSMEkH5uW9Jf2zpKPqHVctkg7JcY6odyztSXqfpMMkbd2ufGS9YuqMpEn1jqGa/F6eKekySZfm5b3qFo8HCPQMSUuBV4GngeuB/42IhfWNqmsknRoRP9kA4pgF7JtHFU4AVgA/Bw7L5cfXNcBM0lmk+fc2BaYAw4GpwCeAOyPi/PpF9xZJ0yJiWF7+InAa8AtgBHBrRFxQz/jaSPoqKbbZwFDgjIi4JW97KCL2r2d8OY72X6EQ8DHgHoCIOHa9B1WFpDOBk0lTdbXm4ibS10BuqMfv3Mmmh0iaCRxA+qD5LHAsMIOUeG6OiGV1DK8QSX+OiHdvAHHMjoi98vJaHzKSHo6IofWL7i2SHiN9KG4BPA80RcQrkvoAD0bEB+saYCZpZkTsl5enA0dFxEJJWwEPRMQH6hthkt/PD0XEckmDSX9g/DQiLq08h3qS9BDwBPBj0qwkIv0fPwkgIn5bv+jeIukpYJ+IWNWufHNgVkQMWd8xNeT3bOokIuJN4C7gLkmbkf7qPRm4COh07qD1QdKjHW0CdllV4vawAAAEBElEQVSfsdTweEUv6xFJzRHRImlPYFVnO69HqyNiDbBC0tMR8QpARLwm6c06x1ZpE0nbky6bq63HHRGvSlpd39DW0isilgNExDxJhwI/lzSI6lNQ1UMzcAbw78C/RsTDkl7bUJJMhTeBXYE/tSvvn7etd042PWet/wz5L4rJwOT8l+6GYhfgCGBJu3IBf1j/4VT1BeBSSd8mTRp4v6T5wPy8bUPxhqQtI2IFqVcLQL6vtCElm21JvWwBIeldEfF8vi+yoXyIAzwvaWhEPAyQezifBCYCG0TvK/9BebGk/80/X2DD/Bz9GnC3pDmk/zcA7wb2AE7vcK8S+TJaD5G0Z0Q8Ve84OiPpKuAnEXFvlW3XRcTn6hBWVZK2Ad5D+s/cGhEv1DmktUjaIiJer1K+E9A/Ih6rQ1iFSdoS2CUinq13LACSmki9xeerbDs4Iu6rQ1g1SToaODgivlXvWNqTtAnpcSsDSH9UtALTc298/cfjZGNmZmXz0GczMyudk42ZmZXOycasE5KaJN0iaY6kp/MX5Dav2H69pEclfT1/ke5hSTMl7V7PuNuTdIqkH9Q7DntncrIxq0GSgJuBX+bvJuwJbA2cn7e/C/hwRHwwIi4GjgNuiYj9IuLpgm30Kif6nrWxxGkbJicbs9o+Dqxsm1khj+T5OvAPeTTXXcDOuTdzFmnI6Rck/QZA0t9Jmpa3/6jtA1vScknnSnoQ+FBlg5Km5ql5pkl6StLf5PK1eiaSfpW/i9J2vAslzZD0f3kanamSnpFU+a32gZJ+LenJHG/bsbocp1lXONmY1bYP6Tsqf5W/vPln0ncWjgWejoihEXEOcAVwcUR8LM9D9VnS0NihwBrgb/NhtgIej4jh1YahA5vmKWa+BpxVZXt7WwFTI+IAYBlwHnA48Cng3Ip6w3IMQ4ETJTWvY5xmhWyIX0Yy25CINC1J0fJKh5G+7Dk9XY2jD/Bi3rYGuKnGvjfnnzOAwQXifAP4dV5+DHg9IlblKWAq958SEYsAJN0MHAKsXoc4zQpxsjGrbRbw6coCpRmpB5ImXd25xr4CromIb1bZtrKTL9e1fVl0DW/9P13N2lcjelcsr4q3vjT3Ztv+EfGmpMr/5+0TZNv8Xt2N06wQX0Yzq+1uYEtJo+GvN8m/B1ydp6npbN8TJO2c990hz/PVXfOAoZI2UXqcxbBuHOPwHEcf0mCG+0qI0+xtnGzMasi9hU+R7m/MAZ4CVgKdTk8SEU8A3yZNzPoo6TEE/dchnPuAZ0mXyS4CHurGMe4Ffgo8DNwUES0lxGn2Np6uxszMSueejZmZlc7JxszMSudkY2ZmpXOyMTOz0jnZmJlZ6ZxszMysdE42ZmZWOicbMzMr3f8Hc8T1+rSTKXYAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f164656a4e0>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "best_sale_offer(full_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "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": "3.6.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}