Customer Segments.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 1. Business Problem\n",
    "## 1.1 Problem Context\n",
    "Our client is an online retailer based in the UK. They sell all-occasion gifts, and many of their customers are wholesalers.\n",
    "* Most of their customers are from the UK, but they have a small percent of customers from other countries.\n",
    "* They want to create groups of these international customers based on their previous purchase patterns.\n",
    "* Their goal is to provide more tailored services and improve the way they market to these international customers.\n",
    "\n",
    "## 1.2 Problems with current approach\n",
    "Currently, the retailer simply groups their international customers by country. As you'll see in the project, this is quite inefficient because:\n",
    "* There's a large number of countries (which kind of defeats the purpose of creating groups).\n",
    "* Some countries have very few customers.\n",
    "* This approach treats large and small customers the same, regardless of their purchase patterns.\n",
    "\n",
    "## 1.3 Problem Statement\n",
    "The retailer has hired us to help them create customer clusters, a.k.a **\"customer segments\"** through a data-driven approach.\n",
    "* They've provided us a dataset of past purchase data at the transaction level.\n",
    "* Our task is to build a clustering model using that dataset.\n",
    "* Our clustering model should factor in both aggregate sales patterns and specific items purchased.\n",
    "\n",
    "## 1.4 Business Objectives and constraints\n",
    "\n",
    "# 2. Machine Learning Problem\n",
    "## 2.1 Data Overview\n",
    "For this project:\n",
    "* The dataset has 35116 observations for previous international transactions.\n",
    "* The observations span 37 different countries.\n",
    "* **Note:** There is no target variable.\n",
    "\n",
    "We have the following features:\n",
    "\n",
    "Invoice information\n",
    "* 'InvoiceNo' – Unique ID for invoice\n",
    "* 'InvoiceDate' – Invoice date\n",
    "\n",
    "Item information\n",
    "* 'StockCode' – Unique ID for item\n",
    "* 'Description' – Text description for item\n",
    "* 'Quantity' – Units per pack for item\n",
    "* 'UnitPrice' – Price per unit in GBP\n",
    "\n",
    "Customer information\n",
    "* 'CustomerID' – Unique ID for customer\n",
    "* 'Country' – Country of customer\n",
    "\n",
    "## 2.2 Mapping Buisness problem to ML problem\n",
    "### 2.2.1 Type of ML Problem\n",
    "It is an unsupervised learning task, where given the features about each transaction, we need to segment the customers based on their buying patterns.\n",
    "* It is importnat to note that the given data is transaction-level while the clusters (or segmenst) we need to create are customer-level."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 3. Exploratory Data Analysis\n",
    "Import the libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# NumPy for numerical computing\n",
    "import numpy as np\n",
    "\n",
    "# Pandas for DataFrames\n",
    "import pandas as pd\n",
    "pd.set_option('display.max_columns', 100)\n",
    "\n",
    "# Matplotlib for visualization\n",
    "from matplotlib import pyplot as plt\n",
    "# display plots in the notebook\n",
    "%matplotlib inline \n",
    "\n",
    "# Seaborn for easier visualization\n",
    "import seaborn as sns\n",
    "\n",
    "# Ignore Warnings\n",
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\")\n",
    "\n",
    "# StandardScaler from Scikit-Learn\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "\n",
    "# PCA from Scikit-Learn\n",
    "from sklearn.decomposition import PCA\n",
    "\n",
    "# Scikit-Learn's KMeans algorithm\n",
    "from sklearn.cluster import KMeans\n",
    "\n",
    "# Adjusted Rand index\n",
    "from sklearn.metrics import adjusted_rand_score"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3.1 Load the dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Load international online transactions data from CSV\n",
    "df = pd.read_csv('Files/int_online_tx.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Even though we eventually want customer-level data, it's still helpful to do some basic exploratory analysis at the transaction level."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(35116, 8)"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Dataframe dimensions\n",
    "df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>InvoiceNo</th>\n",
       "      <th>StockCode</th>\n",
       "      <th>Description</th>\n",
       "      <th>Quantity</th>\n",
       "      <th>InvoiceDate</th>\n",
       "      <th>UnitPrice</th>\n",
       "      <th>CustomerID</th>\n",
       "      <th>Country</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>536370</td>\n",
       "      <td>22728</td>\n",
       "      <td>ALARM CLOCK BAKELIKE PINK</td>\n",
       "      <td>24</td>\n",
       "      <td>12/1/10 8:45</td>\n",
       "      <td>3.75</td>\n",
       "      <td>12583.0</td>\n",
       "      <td>France</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>536370</td>\n",
       "      <td>22727</td>\n",
       "      <td>ALARM CLOCK BAKELIKE RED</td>\n",
       "      <td>24</td>\n",
       "      <td>12/1/10 8:45</td>\n",
       "      <td>3.75</td>\n",
       "      <td>12583.0</td>\n",
       "      <td>France</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>536370</td>\n",
       "      <td>22726</td>\n",
       "      <td>ALARM CLOCK BAKELIKE GREEN</td>\n",
       "      <td>12</td>\n",
       "      <td>12/1/10 8:45</td>\n",
       "      <td>3.75</td>\n",
       "      <td>12583.0</td>\n",
       "      <td>France</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>536370</td>\n",
       "      <td>21724</td>\n",
       "      <td>PANDA AND BUNNIES STICKER SHEET</td>\n",
       "      <td>12</td>\n",
       "      <td>12/1/10 8:45</td>\n",
       "      <td>0.85</td>\n",
       "      <td>12583.0</td>\n",
       "      <td>France</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>536370</td>\n",
       "      <td>21883</td>\n",
       "      <td>STARS GIFT TAPE</td>\n",
       "      <td>24</td>\n",
       "      <td>12/1/10 8:45</td>\n",
       "      <td>0.65</td>\n",
       "      <td>12583.0</td>\n",
       "      <td>France</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>536370</td>\n",
       "      <td>10002</td>\n",
       "      <td>INFLATABLE POLITICAL GLOBE</td>\n",
       "      <td>48</td>\n",
       "      <td>12/1/10 8:45</td>\n",
       "      <td>0.85</td>\n",
       "      <td>12583.0</td>\n",
       "      <td>France</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>536370</td>\n",
       "      <td>21791</td>\n",
       "      <td>VINTAGE HEADS AND TAILS CARD GAME</td>\n",
       "      <td>24</td>\n",
       "      <td>12/1/10 8:45</td>\n",
       "      <td>1.25</td>\n",
       "      <td>12583.0</td>\n",
       "      <td>France</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>536370</td>\n",
       "      <td>21035</td>\n",
       "      <td>SET/2 RED RETROSPOT TEA TOWELS</td>\n",
       "      <td>18</td>\n",
       "      <td>12/1/10 8:45</td>\n",
       "      <td>2.95</td>\n",
       "      <td>12583.0</td>\n",
       "      <td>France</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>536370</td>\n",
       "      <td>22326</td>\n",
       "      <td>ROUND SNACK BOXES SET OF4 WOODLAND</td>\n",
       "      <td>24</td>\n",
       "      <td>12/1/10 8:45</td>\n",
       "      <td>2.95</td>\n",
       "      <td>12583.0</td>\n",
       "      <td>France</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>536370</td>\n",
       "      <td>22629</td>\n",
       "      <td>SPACEBOY LUNCH BOX</td>\n",
       "      <td>24</td>\n",
       "      <td>12/1/10 8:45</td>\n",
       "      <td>1.95</td>\n",
       "      <td>12583.0</td>\n",
       "      <td>France</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   InvoiceNo StockCode                          Description  Quantity  \\\n",
       "0     536370     22728            ALARM CLOCK BAKELIKE PINK        24   \n",
       "1     536370     22727            ALARM CLOCK BAKELIKE RED         24   \n",
       "2     536370     22726           ALARM CLOCK BAKELIKE GREEN        12   \n",
       "3     536370     21724      PANDA AND BUNNIES STICKER SHEET        12   \n",
       "4     536370     21883                     STARS GIFT TAPE         24   \n",
       "5     536370     10002          INFLATABLE POLITICAL GLOBE         48   \n",
       "6     536370     21791   VINTAGE HEADS AND TAILS CARD GAME         24   \n",
       "7     536370     21035      SET/2 RED RETROSPOT TEA TOWELS         18   \n",
       "8     536370     22326  ROUND SNACK BOXES SET OF4 WOODLAND         24   \n",
       "9     536370     22629                  SPACEBOY LUNCH BOX         24   \n",
       "\n",
       "    InvoiceDate  UnitPrice  CustomerID Country  \n",
       "0  12/1/10 8:45       3.75     12583.0  France  \n",
       "1  12/1/10 8:45       3.75     12583.0  France  \n",
       "2  12/1/10 8:45       3.75     12583.0  France  \n",
       "3  12/1/10 8:45       0.85     12583.0  France  \n",
       "4  12/1/10 8:45       0.65     12583.0  France  \n",
       "5  12/1/10 8:45       0.85     12583.0  France  \n",
       "6  12/1/10 8:45       1.25     12583.0  France  \n",
       "7  12/1/10 8:45       2.95     12583.0  France  \n",
       "8  12/1/10 8:45       2.95     12583.0  France  \n",
       "9  12/1/10 8:45       1.95     12583.0  France  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# First 10 rows of data\n",
    "df.head(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here are some questions to consider:\n",
    "* In the first 10 observations, how many different customers are there?\n",
    "* How many different invoices are there?\n",
    "* Given answers to the first two questions, how many unique purchases are shown?\n",
    "* Technically, aren't these observations actually line-items within each transaction? \n",
    "* Do you expect the customer-level dataset to be much smaller?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Display the distribution of transactions by country."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x15db2f812e8>"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAokAAAJQCAYAAADi5v7GAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x15db4309e80>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Make figsize\n",
    "plt.figure(figsize=(9,10))\n",
    "\n",
    "# Bar plot by country\n",
    "sns.countplot(y='Country', data=df)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There are many **sparse classes**. Countries like Lithuania, Brazil, and even the USA have a tiny number of transactions.\n",
    "\n",
    "**Note:** This is at the transaction/line-item level. The number of customers for each country is even smaller because each customer has multiple transactions! Therefore, it's plain to see that clustering by country is not very efficient."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 4. Transaction-level Cleaning\n",
    "Before we aggregate to the customer level, we need to tidy up a few things at the transaction level.\n",
    "* Technically, this is the **\"line-item\" level** because one invoice (a.k.a. transaction) spans multiple rows. However, we'll just refer to it as the \"transaction level\" for simplicity.\n",
    "* Also the terms **\"aggegrating up\"** and **\"rolling up\"** are used interchangeability.\n",
    "\n",
    "First, we need to check for any missing data before rolling up. If we don't, we might get confusing results later because rolling up to a higher level can sometimes hide the fact that data was missing at the lower level.\n",
    "\n",
    "Display the number of missing observations for each feature."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "InvoiceNo         0\n",
       "StockCode         0\n",
       "Description       0\n",
       "Quantity          0\n",
       "InvoiceDate       0\n",
       "UnitPrice         0\n",
       "CustomerID     1418\n",
       "Country           0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Missing data by feature\n",
    "df.isnull().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "'CustomerID' has missing observations\n",
    "* Should we **label them as missing** (as for categorical features) or should we **flag and fill** them (as for numeric features)?\n",
    "* We should do neither. Instead, we simply need to drop transactions with missing CustomerID.\n",
    "* Think back to the project scope: We are trying to cluster customers in order to provide more tailored service!\n",
    "* That means transactions with missing 'CustomerID' are actually pointless to keep.\n",
    "* In other words, they should be considered \"unwanted observations\" instead of \"missing data!\".\n",
    "\n",
    "Drop observations with missing customer ID's."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Only keep transactions with CustomerID's\n",
    "df = df[df.CustomerID.notnull()]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Next, just for clarity, convert the CustomerID's from floats into integers. (This is technically not required, but it's good practice to save ID's as either strings or integers so they don't get mixed up with other numeric features.)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    12583\n",
       "1    12583\n",
       "2    12583\n",
       "3    12583\n",
       "4    12583\n",
       "Name: CustomerID, dtype: int32"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Convert customer ID's into integers\n",
    "df['CustomerID'] = df.CustomerID.astype(int)\n",
    "\n",
    "# Display first 5 CustomerID's in the transaction dataset\n",
    "df.CustomerID.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There's one feature we need to create at the transaction level.\n",
    "\n",
    "Look back at the first 10 observations you displayed earlier.\n",
    "* Are there any features that tell you how much money the customer spent on each transaction?\n",
    "* Well, we have 'Quantity' and 'UnitPrice', but those are for individual units, not for the transaction.\n",
    "* We need to multiply them together to get the Sales amount for that transaction."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    90.0\n",
       "1    90.0\n",
       "2    45.0\n",
       "3    10.2\n",
       "4    15.6\n",
       "Name: Sales, dtype: float64"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Create 'Sales' interaction feature\n",
    "df['Sales'] = df.Quantity * df.UnitPrice\n",
    "\n",
    "# Display first 5 Sales values in the transaction dataset\n",
    "df.Sales.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Before moving on, save your cleaned transaction-level data as **cleaned_transactions.csv.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Save cleaned transaction-level data\n",
    "df.to_csv('Files/cleaned_transactions.csv', index=None)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 5. Customer-level feature engineering\n",
    "Now that we have a cleaned transaction-level dataset, it's time to **roll it up** (aggregate up) to the customer level, which we'll feed into our machine learning algorithms later.\n",
    "\n",
    "We want 1 customer per row, and we want the features to represent information such as:\n",
    "* Number of unique purchases by the customer\n",
    "* Average cart value for the customer\n",
    "* Total sales for the customer\n",
    "* Etc.\n",
    "\n",
    "To do so, we'll use two tools seen already:\n",
    "* groupby() to roll up by customer.\n",
    "* agg() to engineer aggregated features.\n",
    "\n",
    "Aggegrate invoice data by customer. We'll engineer 1 feature:\n",
    "* 'total_transactions' - the total number of unique transactions for each customer."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>total_transactions</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CustomerID</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>12347</th>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12348</th>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12349</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12350</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12352</th>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            total_transactions\n",
       "CustomerID                    \n",
       "12347                        7\n",
       "12348                        4\n",
       "12349                        1\n",
       "12350                        1\n",
       "12352                        8"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Aggegrate invoice data\n",
    "invoice_data = df.groupby('CustomerID').InvoiceNo.agg({ 'total_transactions' : 'nunique' })\n",
    "\n",
    "# Display invoice data for first 5 customers\n",
    "invoice_data.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Explanation:\n",
    "* df.groupby('CustomerID') - group data by CustomerID.\n",
    "* df.groupby('CustomerID').InvoiceNo - to get number of unique transactions, we used the Invoice ID column.\n",
    "* df.groupby('CustomerID').InvoiceNo.agg({ 'total_transactions' : 'nunique' }) - finally, we passed in a dictionary of aggregations to the .agg() function.\n",
    "* The keys for the dictionary should be the names of the features to engineer and the values can be any aggregation function.\n",
    "* Many common functions can be referred to by a simple string, e.g.  'nunique', 'count', 'sum', 'mean', 'min', 'max'\n",
    "* 'nunique' counts the number of unique values within each group.\n",
    "\n",
    "Next, using the same technique, aggregate product data by customer. Engineer 2 features:\n",
    "* 'total_products' - the total number of products purchased by each customer (not the unit Quantity within each product!).\n",
    "* 'total_unique_products' - the number of unique products purchased by each customer."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>total_products</th>\n",
       "      <th>total_unique_products</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CustomerID</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>12347</th>\n",
       "      <td>182</td>\n",
       "      <td>103</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12348</th>\n",
       "      <td>31</td>\n",
       "      <td>22</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12349</th>\n",
       "      <td>73</td>\n",
       "      <td>73</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12350</th>\n",
       "      <td>17</td>\n",
       "      <td>17</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12352</th>\n",
       "      <td>85</td>\n",
       "      <td>59</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            total_products  total_unique_products\n",
       "CustomerID                                       \n",
       "12347                  182                    103\n",
       "12348                   31                     22\n",
       "12349                   73                     73\n",
       "12350                   17                     17\n",
       "12352                   85                     59"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Aggregate product data\n",
    "product_data = df.groupby('CustomerID').StockCode.agg( { 'total_products' : 'count', \n",
    "                                                     'total_unique_products' : 'nunique' } )\n",
    "\n",
    "# Display product data for first 5 customers\n",
    "product_data.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "By definition, 'total_unique_products' should always be less than or equal to 'total_products'.\n",
    "\n",
    "Finally, aggregate sales data by customer. Engineer 2 features:\n",
    "* 'total_sales' - the total sales for each customer.\n",
    "* 'avg_product_value' - the average value of the products purchased by the customer (not the UnitPrice!)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>total_sales</th>\n",
       "      <th>avg_product_value</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CustomerID</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>12347</th>\n",
       "      <td>4310.00</td>\n",
       "      <td>23.681319</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12348</th>\n",
       "      <td>1797.24</td>\n",
       "      <td>57.975484</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12349</th>\n",
       "      <td>1757.55</td>\n",
       "      <td>24.076027</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12350</th>\n",
       "      <td>334.40</td>\n",
       "      <td>19.670588</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12352</th>\n",
       "      <td>2506.04</td>\n",
       "      <td>29.482824</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            total_sales  avg_product_value\n",
       "CustomerID                                \n",
       "12347           4310.00          23.681319\n",
       "12348           1797.24          57.975484\n",
       "12349           1757.55          24.076027\n",
       "12350            334.40          19.670588\n",
       "12352           2506.04          29.482824"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Roll up sales data\n",
    "sales_data = df.groupby('CustomerID').Sales.agg( { 'total_sales' : 'sum', \n",
    "                                                  'avg_product_value' : 'mean' } )\n",
    "\n",
    "# Display sales data for first 5 customers\n",
    "sales_data.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 6. Intermediary levels\n",
    "You won't always be able to easily roll up to customer-level directly. Sometimes, it will be easier to create intermediary levels first.\n",
    "\n",
    "For example, let's say we wanted to calculate the average cart value for each customer.\n",
    "* 'avg_product_value' isn't the same thing because it doesn't first group products that were purchased within the same \"cart\" (i.e. invoice).\n",
    "\n",
    "Therefore, let's first aggregate cart data at the \"cart-level.\"\n",
    "* We'll group by 'CustomerID' AND by 'InvoiceID'. Remember, we're treating each invoice as a \"cart.\"\n",
    "* Then, we'll calculate 'cart_value' by taking the sum of the Sales column. This is the total sales by invoice (i.e. cart).\n",
    "* Finally, we'll call .reset_index() to turn CustomerID and InvoiceID back into regular columns so we can perform another aggregation."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>cart_value</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CustomerID</th>\n",
       "      <th>InvoiceNo</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"7\" valign=\"top\">12347</th>\n",
       "      <th>537626</th>\n",
       "      <td>711.79</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>542237</th>\n",
       "      <td>475.39</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>549222</th>\n",
       "      <td>636.25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>556201</th>\n",
       "      <td>382.52</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>562032</th>\n",
       "      <td>584.91</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>573511</th>\n",
       "      <td>1294.32</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>581180</th>\n",
       "      <td>224.82</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"4\" valign=\"top\">12348</th>\n",
       "      <th>539318</th>\n",
       "      <td>892.80</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>541998</th>\n",
       "      <td>227.44</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>548955</th>\n",
       "      <td>367.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>568172</th>\n",
       "      <td>310.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12349</th>\n",
       "      <th>577609</th>\n",
       "      <td>1757.55</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12350</th>\n",
       "      <th>543037</th>\n",
       "      <td>334.40</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"7\" valign=\"top\">12352</th>\n",
       "      <th>544156</th>\n",
       "      <td>296.50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>545323</th>\n",
       "      <td>144.35</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>545332</th>\n",
       "      <td>840.30</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>546869</th>\n",
       "      <td>120.33</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>547390</th>\n",
       "      <td>160.33</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>567505</th>\n",
       "      <td>366.25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>568699</th>\n",
       "      <td>266.25</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                      cart_value\n",
       "CustomerID InvoiceNo            \n",
       "12347      537626         711.79\n",
       "           542237         475.39\n",
       "           549222         636.25\n",
       "           556201         382.52\n",
       "           562032         584.91\n",
       "           573511        1294.32\n",
       "           581180         224.82\n",
       "12348      539318         892.80\n",
       "           541998         227.44\n",
       "           548955         367.00\n",
       "           568172         310.00\n",
       "12349      577609        1757.55\n",
       "12350      543037         334.40\n",
       "12352      544156         296.50\n",
       "           545323         144.35\n",
       "           545332         840.30\n",
       "           546869         120.33\n",
       "           547390         160.33\n",
       "           567505         366.25\n",
       "           568699         266.25"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Aggregate cart-level data (i.e. invoice-level)\n",
    "cart_data = df.groupby(['CustomerID' , 'InvoiceNo']).Sales.agg( { 'cart_value' : 'sum' })\n",
    "\n",
    "# Display cart data for first 20 CARTS\n",
    "cart_data.head(20)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>CustomerID</th>\n",
       "      <th>InvoiceNo</th>\n",
       "      <th>cart_value</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>12347</td>\n",
       "      <td>537626</td>\n",
       "      <td>711.79</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>12347</td>\n",
       "      <td>542237</td>\n",
       "      <td>475.39</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>12347</td>\n",
       "      <td>549222</td>\n",
       "      <td>636.25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>12347</td>\n",
       "      <td>556201</td>\n",
       "      <td>382.52</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>12347</td>\n",
       "      <td>562032</td>\n",
       "      <td>584.91</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>12347</td>\n",
       "      <td>573511</td>\n",
       "      <td>1294.32</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>12347</td>\n",
       "      <td>581180</td>\n",
       "      <td>224.82</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>12348</td>\n",
       "      <td>539318</td>\n",
       "      <td>892.80</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>12348</td>\n",
       "      <td>541998</td>\n",
       "      <td>227.44</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>12348</td>\n",
       "      <td>548955</td>\n",
       "      <td>367.00</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   CustomerID  InvoiceNo  cart_value\n",
       "0       12347     537626      711.79\n",
       "1       12347     542237      475.39\n",
       "2       12347     549222      636.25\n",
       "3       12347     556201      382.52\n",
       "4       12347     562032      584.91\n",
       "5       12347     573511     1294.32\n",
       "6       12347     581180      224.82\n",
       "7       12348     539318      892.80\n",
       "8       12348     541998      227.44\n",
       "9       12348     548955      367.00"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Reset index\n",
    "cart_data.reset_index(inplace=True)\n",
    "\n",
    "# Display cart data for first 10 CARTS\n",
    "cart_data.head(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now that we have cart-level cart data, all we need to do is roll up by CustomerID again to get customer-level cart data.\n",
    "\n",
    "Aggregate cart data by customer. Engineer 3 features:\n",
    "* 'avg_cart_value' - average cart value by customer.\n",
    "* 'min_cart_value' - minimum cart value by customer.\n",
    "* 'max_cart_value' - maximum cart value by customer."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>min_cart_value</th>\n",
       "      <th>max_cart_value</th>\n",
       "      <th>avg_cart_value</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CustomerID</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>12347</th>\n",
       "      <td>224.82</td>\n",
       "      <td>1294.32</td>\n",
       "      <td>615.714286</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12348</th>\n",
       "      <td>227.44</td>\n",
       "      <td>892.80</td>\n",
       "      <td>449.310000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12349</th>\n",
       "      <td>1757.55</td>\n",
       "      <td>1757.55</td>\n",
       "      <td>1757.550000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12350</th>\n",
       "      <td>334.40</td>\n",
       "      <td>334.40</td>\n",
       "      <td>334.400000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12352</th>\n",
       "      <td>120.33</td>\n",
       "      <td>840.30</td>\n",
       "      <td>313.255000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            min_cart_value  max_cart_value  avg_cart_value\n",
       "CustomerID                                                \n",
       "12347               224.82         1294.32      615.714286\n",
       "12348               227.44          892.80      449.310000\n",
       "12349              1757.55         1757.55     1757.550000\n",
       "12350               334.40          334.40      334.400000\n",
       "12352               120.33          840.30      313.255000"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Aggregate cart data (at customer-level)\n",
    "agg_cart_data = cart_data.groupby('CustomerID').cart_value.agg( { 'avg_cart_value' : 'mean', \n",
    "                                                                 'min_cart_value' : 'min',\n",
    "                                                                 'max_cart_value' : 'max'})\n",
    "\n",
    "# Display cart data for first 5 CUSTOMERS\n",
    "agg_cart_data.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 7. Joining various customer level dataframes\n",
    "We have multiple dataframes that each contain customer-level features:\n",
    "* invoice_data\n",
    "* product_data\n",
    "* sales_data\n",
    "* agg_cart_data\n",
    "\n",
    "Let's join the various customer-level datasets together with the .join() function.\n",
    "* Just pick one of the customer-level dataframes and join it to a list of the others.\n",
    "* By default, it will join the dataframes on their index... In this case, it will join by CustomerID, which is exactly what we want.\n",
    "* You can read more about the .join() function in the official documentation."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>total_transactions</th>\n",
       "      <th>total_products</th>\n",
       "      <th>total_unique_products</th>\n",
       "      <th>total_sales</th>\n",
       "      <th>avg_product_value</th>\n",
       "      <th>min_cart_value</th>\n",
       "      <th>max_cart_value</th>\n",
       "      <th>avg_cart_value</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CustomerID</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>12347</th>\n",
       "      <td>7</td>\n",
       "      <td>182</td>\n",
       "      <td>103</td>\n",
       "      <td>4310.00</td>\n",
       "      <td>23.681319</td>\n",
       "      <td>224.82</td>\n",
       "      <td>1294.32</td>\n",
       "      <td>615.714286</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12348</th>\n",
       "      <td>4</td>\n",
       "      <td>31</td>\n",
       "      <td>22</td>\n",
       "      <td>1797.24</td>\n",
       "      <td>57.975484</td>\n",
       "      <td>227.44</td>\n",
       "      <td>892.80</td>\n",
       "      <td>449.310000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12349</th>\n",
       "      <td>1</td>\n",
       "      <td>73</td>\n",
       "      <td>73</td>\n",
       "      <td>1757.55</td>\n",
       "      <td>24.076027</td>\n",
       "      <td>1757.55</td>\n",
       "      <td>1757.55</td>\n",
       "      <td>1757.550000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12350</th>\n",
       "      <td>1</td>\n",
       "      <td>17</td>\n",
       "      <td>17</td>\n",
       "      <td>334.40</td>\n",
       "      <td>19.670588</td>\n",
       "      <td>334.40</td>\n",
       "      <td>334.40</td>\n",
       "      <td>334.400000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12352</th>\n",
       "      <td>8</td>\n",
       "      <td>85</td>\n",
       "      <td>59</td>\n",
       "      <td>2506.04</td>\n",
       "      <td>29.482824</td>\n",
       "      <td>120.33</td>\n",
       "      <td>840.30</td>\n",
       "      <td>313.255000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            total_transactions  total_products  total_unique_products  \\\n",
       "CustomerID                                                              \n",
       "12347                        7             182                    103   \n",
       "12348                        4              31                     22   \n",
       "12349                        1              73                     73   \n",
       "12350                        1              17                     17   \n",
       "12352                        8              85                     59   \n",
       "\n",
       "            total_sales  avg_product_value  min_cart_value  max_cart_value  \\\n",
       "CustomerID                                                                   \n",
       "12347           4310.00          23.681319          224.82         1294.32   \n",
       "12348           1797.24          57.975484          227.44          892.80   \n",
       "12349           1757.55          24.076027         1757.55         1757.55   \n",
       "12350            334.40          19.670588          334.40          334.40   \n",
       "12352           2506.04          29.482824          120.33          840.30   \n",
       "\n",
       "            avg_cart_value  \n",
       "CustomerID                  \n",
       "12347           615.714286  \n",
       "12348           449.310000  \n",
       "12349          1757.550000  \n",
       "12350           334.400000  \n",
       "12352           313.255000  "
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Join together customer-level data\n",
    "customer_df = invoice_data.join([product_data, sales_data, agg_cart_data])\n",
    "\n",
    "# Display customer-level data for first 5 customers\n",
    "customer_df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Finally, let's save customer_df as our **analytical base table** to use later.\n",
    "\n",
    "**Very Important:** We will not set index=None because we want to keep the CustomerID's as the index (this will be important and we'll see later)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Save analytical base table\n",
    "customer_df.to_csv('Files/analytical_base_table.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 8. Curse of Dimensionality.\n",
    "Let's import the cleaned dataset (not the analytical base table) that we saved in previously."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Read cleaned_transactions.csv\n",
    "df = pd.read_csv('Files/cleaned_transactions.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 8.1 So what is \"The Curse of Dimensionality?\"\n",
    "\n",
    "\"dimensionality\" refers to the number of features in your dataset. The basic idea is that as the number of features increases, you'll need more and more observations to build any sort of meaningful model, especially for clustering.\n",
    "\n",
    "Because cluster models are based on the \"distance\" between two observations, and distance is calculated by taking the differences between feature values, every observation will seem \"far away\" from each other if the number of features increases.\n",
    "\n",
    "The reference link provides an excellent analogy : https://www.quora.com/What-is-the-curse-of-dimensionality\n",
    "\n",
    "    Let's say you have a straight line 100 yards long and you dropped a penny somewhere on it. It wouldn't be too hard to find. You walk along the line and it takes two minutes.\n",
    "\n",
    "    Now let's say you have a square 100 yards on each side and you dropped a penny somewhere on it. It would be pretty hard, like searching across two football fields stuck together. It could take days.\n",
    "\n",
    "    Now a cube 100 yards across. That's like searching a 30-story building the size of a football stadium. Ugh.\n",
    "\n",
    "    The difficulty of searching through the space gets a lot harder as you have more dimensions.\n",
    "\n",
    "For our practical purposes, it's enough to remember that when you have many features (high dimensionality), it makes clustering especially hard because every observation is \"far away\" from each other.\n",
    "\n",
    "The amount of \"space\" that a data point could potentially exist in becomes larger and larger, and clusters become very hard to form.\n",
    "\n",
    "## 8.2 Item Data\n",
    "So how does The Curse of Dimensionality arise in this problem?\n",
    "\n",
    "Well, in the previous module, we created a customer-level analytical base table with important features such as total sales by customer and average cart value by customer.\n",
    "\n",
    "However, remember, the client would also like to to include information about individual items that were purchased.\n",
    "* For example, if two customers purchased similar items, our model should be more likely to group them into the same cluster.\n",
    "* In other words, we care not just about how much a customer purchases, but also what they purchase.\n",
    "* In every observation, along with the data that we hav built, we need to include the information of what are the products that are purchased by each customer, i.e. we need to find some way to represent each unique item.\n",
    "\n",
    "One way is to create a vector of unique values of the StockCode column and if the customer has purchased a particular product, fill it by 1, else fill it by 0.\n",
    "\n",
    "This is like the binary CountVectorizer technique."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>10002</th>\n",
       "      <th>10120</th>\n",
       "      <th>10125</th>\n",
       "      <th>10133</th>\n",
       "      <th>10135</th>\n",
       "      <th>11001</th>\n",
       "      <th>15034</th>\n",
       "      <th>15036</th>\n",
       "      <th>15039</th>\n",
       "      <th>15044A</th>\n",
       "      <th>15044B</th>\n",
       "      <th>15044C</th>\n",
       "      <th>15044D</th>\n",
       "      <th>15056BL</th>\n",
       "      <th>15056N</th>\n",
       "      <th>15056P</th>\n",
       "      <th>15058A</th>\n",
       "      <th>15058B</th>\n",
       "      <th>15058C</th>\n",
       "      <th>15060B</th>\n",
       "      <th>16008</th>\n",
       "      <th>16011</th>\n",
       "      <th>16012</th>\n",
       "      <th>16014</th>\n",
       "      <th>16016</th>\n",
       "      <th>16045</th>\n",
       "      <th>16048</th>\n",
       "      <th>16054</th>\n",
       "      <th>16156L</th>\n",
       "      <th>16156S</th>\n",
       "      <th>16161G</th>\n",
       "      <th>16161P</th>\n",
       "      <th>16161U</th>\n",
       "      <th>16168M</th>\n",
       "      <th>16169E</th>\n",
       "      <th>16169K</th>\n",
       "      <th>16169M</th>\n",
       "      <th>16218</th>\n",
       "      <th>16219</th>\n",
       "      <th>16225</th>\n",
       "      <th>16235</th>\n",
       "      <th>16236</th>\n",
       "      <th>16237</th>\n",
       "      <th>16238</th>\n",
       "      <th>16258A</th>\n",
       "      <th>16259</th>\n",
       "      <th>17003</th>\n",
       "      <th>17011F</th>\n",
       "      <th>17012A</th>\n",
       "      <th>17012B</th>\n",
       "      <th>...</th>\n",
       "      <th>85231B</th>\n",
       "      <th>85232A</th>\n",
       "      <th>85232B</th>\n",
       "      <th>85232D</th>\n",
       "      <th>90001B</th>\n",
       "      <th>90001D</th>\n",
       "      <th>90013A</th>\n",
       "      <th>90013C</th>\n",
       "      <th>90018C</th>\n",
       "      <th>90019A</th>\n",
       "      <th>90024B</th>\n",
       "      <th>90030A</th>\n",
       "      <th>90030B</th>\n",
       "      <th>90030C</th>\n",
       "      <th>90031</th>\n",
       "      <th>90036E</th>\n",
       "      <th>90037B</th>\n",
       "      <th>90057</th>\n",
       "      <th>90070</th>\n",
       "      <th>90087</th>\n",
       "      <th>90098</th>\n",
       "      <th>90099</th>\n",
       "      <th>90108</th>\n",
       "      <th>90114</th>\n",
       "      <th>90120B</th>\n",
       "      <th>90145</th>\n",
       "      <th>90160A</th>\n",
       "      <th>90160B</th>\n",
       "      <th>90160C</th>\n",
       "      <th>90160D</th>\n",
       "      <th>90161B</th>\n",
       "      <th>90161C</th>\n",
       "      <th>90161D</th>\n",
       "      <th>90162A</th>\n",
       "      <th>90162B</th>\n",
       "      <th>90164A</th>\n",
       "      <th>90170</th>\n",
       "      <th>90173</th>\n",
       "      <th>90184B</th>\n",
       "      <th>90184C</th>\n",
       "      <th>90192</th>\n",
       "      <th>90201A</th>\n",
       "      <th>90201B</th>\n",
       "      <th>90201C</th>\n",
       "      <th>90201D</th>\n",
       "      <th>90202D</th>\n",
       "      <th>90204</th>\n",
       "      <th>C2</th>\n",
       "      <th>M</th>\n",
       "      <th>POST</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 2574 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   10002  10120  10125  10133  10135  11001  15034  15036  15039  15044A  \\\n",
       "0      0      0      0      0      0      0      0      0      0       0   \n",
       "1      0      0      0      0      0      0      0      0      0       0   \n",
       "2      0      0      0      0      0      0      0      0      0       0   \n",
       "3      0      0      0      0      0      0      0      0      0       0   \n",
       "4      0      0      0      0      0      0      0      0      0       0   \n",
       "\n",
       "   15044B  15044C  15044D  15056BL  15056N  15056P  15058A  15058B  15058C  \\\n",
       "0       0       0       0        0       0       0       0       0       0   \n",
       "1       0       0       0        0       0       0       0       0       0   \n",
       "2       0       0       0        0       0       0       0       0       0   \n",
       "3       0       0       0        0       0       0       0       0       0   \n",
       "4       0       0       0        0       0       0       0       0       0   \n",
       "\n",
       "   15060B  16008  16011  16012  16014  16016  16045  16048  16054  16156L  \\\n",
       "0       0      0      0      0      0      0      0      0      0       0   \n",
       "1       0      0      0      0      0      0      0      0      0       0   \n",
       "2       0      0      0      0      0      0      0      0      0       0   \n",
       "3       0      0      0      0      0      0      0      0      0       0   \n",
       "4       0      0      0      0      0      0      0      0      0       0   \n",
       "\n",
       "   16156S  16161G  16161P  16161U  16168M  16169E  16169K  16169M  16218  \\\n",
       "0       0       0       0       0       0       0       0       0      0   \n",
       "1       0       0       0       0       0       0       0       0      0   \n",
       "2       0       0       0       0       0       0       0       0      0   \n",
       "3       0       0       0       0       0       0       0       0      0   \n",
       "4       0       0       0       0       0       0       0       0      0   \n",
       "\n",
       "   16219  16225  16235  16236  16237  16238  16258A  16259  17003  17011F  \\\n",
       "0      0      0      0      0      0      0       0      0      0       0   \n",
       "1      0      0      0      0      0      0       0      0      0       0   \n",
       "2      0      0      0      0      0      0       0      0      0       0   \n",
       "3      0      0      0      0      0      0       0      0      0       0   \n",
       "4      0      0      0      0      0      0       0      0      0       0   \n",
       "\n",
       "   17012A  17012B  ...   85231B  85232A  85232B  85232D  90001B  90001D  \\\n",
       "0       0       0  ...        0       0       0       0       0       0   \n",
       "1       0       0  ...        0       0       0       0       0       0   \n",
       "2       0       0  ...        0       0       0       0       0       0   \n",
       "3       0       0  ...        0       0       0       0       0       0   \n",
       "4       0       0  ...        0       0       0       0       0       0   \n",
       "\n",
       "   90013A  90013C  90018C  90019A  90024B  90030A  90030B  90030C  90031  \\\n",
       "0       0       0       0       0       0       0       0       0      0   \n",
       "1       0       0       0       0       0       0       0       0      0   \n",
       "2       0       0       0       0       0       0       0       0      0   \n",
       "3       0       0       0       0       0       0       0       0      0   \n",
       "4       0       0       0       0       0       0       0       0      0   \n",
       "\n",
       "   90036E  90037B  90057  90070  90087  90098  90099  90108  90114  90120B  \\\n",
       "0       0       0      0      0      0      0      0      0      0       0   \n",
       "1       0       0      0      0      0      0      0      0      0       0   \n",
       "2       0       0      0      0      0      0      0      0      0       0   \n",
       "3       0       0      0      0      0      0      0      0      0       0   \n",
       "4       0       0      0      0      0      0      0      0      0       0   \n",
       "\n",
       "   90145  90160A  90160B  90160C  90160D  90161B  90161C  90161D  90162A  \\\n",
       "0      0       0       0       0       0       0       0       0       0   \n",
       "1      0       0       0       0       0       0       0       0       0   \n",
       "2      0       0       0       0       0       0       0       0       0   \n",
       "3      0       0       0       0       0       0       0       0       0   \n",
       "4      0       0       0       0       0       0       0       0       0   \n",
       "\n",
       "   90162B  90164A  90170  90173  90184B  90184C  90192  90201A  90201B  \\\n",
       "0       0       0      0      0       0       0      0       0       0   \n",
       "1       0       0      0      0       0       0      0       0       0   \n",
       "2       0       0      0      0       0       0      0       0       0   \n",
       "3       0       0      0      0       0       0      0       0       0   \n",
       "4       0       0      0      0       0       0      0       0       0   \n",
       "\n",
       "   90201C  90201D  90202D  90204  C2  M  POST  \n",
       "0       0       0       0      0   0  0     0  \n",
       "1       0       0       0      0   0  0     0  \n",
       "2       0       0       0      0   0  0     0  \n",
       "3       0       0       0      0   0  0     0  \n",
       "4       0       0       0      0   0  0     0  \n",
       "\n",
       "[5 rows x 2574 columns]"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Get item_dummies - creates the vector of StockCode\n",
    "item_dummies = pd.get_dummies(df.StockCode)\n",
    "\n",
    "item_dummies.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now, add 'CustomerID' to this new dataframe so that we can roll up (aggregate) by customer later."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>10002</th>\n",
       "      <th>10120</th>\n",
       "      <th>10125</th>\n",
       "      <th>10133</th>\n",
       "      <th>10135</th>\n",
       "      <th>11001</th>\n",
       "      <th>15034</th>\n",
       "      <th>15036</th>\n",
       "      <th>15039</th>\n",
       "      <th>15044A</th>\n",
       "      <th>15044B</th>\n",
       "      <th>15044C</th>\n",
       "      <th>15044D</th>\n",
       "      <th>15056BL</th>\n",
       "      <th>15056N</th>\n",
       "      <th>15056P</th>\n",
       "      <th>15058A</th>\n",
       "      <th>15058B</th>\n",
       "      <th>15058C</th>\n",
       "      <th>15060B</th>\n",
       "      <th>16008</th>\n",
       "      <th>16011</th>\n",
       "      <th>16012</th>\n",
       "      <th>16014</th>\n",
       "      <th>16016</th>\n",
       "      <th>16045</th>\n",
       "      <th>16048</th>\n",
       "      <th>16054</th>\n",
       "      <th>16156L</th>\n",
       "      <th>16156S</th>\n",
       "      <th>16161G</th>\n",
       "      <th>16161P</th>\n",
       "      <th>16161U</th>\n",
       "      <th>16168M</th>\n",
       "      <th>16169E</th>\n",
       "      <th>16169K</th>\n",
       "      <th>16169M</th>\n",
       "      <th>16218</th>\n",
       "      <th>16219</th>\n",
       "      <th>16225</th>\n",
       "      <th>16235</th>\n",
       "      <th>16236</th>\n",
       "      <th>16237</th>\n",
       "      <th>16238</th>\n",
       "      <th>16258A</th>\n",
       "      <th>16259</th>\n",
       "      <th>17003</th>\n",
       "      <th>17011F</th>\n",
       "      <th>17012A</th>\n",
       "      <th>17012B</th>\n",
       "      <th>...</th>\n",
       "      <th>85232A</th>\n",
       "      <th>85232B</th>\n",
       "      <th>85232D</th>\n",
       "      <th>90001B</th>\n",
       "      <th>90001D</th>\n",
       "      <th>90013A</th>\n",
       "      <th>90013C</th>\n",
       "      <th>90018C</th>\n",
       "      <th>90019A</th>\n",
       "      <th>90024B</th>\n",
       "      <th>90030A</th>\n",
       "      <th>90030B</th>\n",
       "      <th>90030C</th>\n",
       "      <th>90031</th>\n",
       "      <th>90036E</th>\n",
       "      <th>90037B</th>\n",
       "      <th>90057</th>\n",
       "      <th>90070</th>\n",
       "      <th>90087</th>\n",
       "      <th>90098</th>\n",
       "      <th>90099</th>\n",
       "      <th>90108</th>\n",
       "      <th>90114</th>\n",
       "      <th>90120B</th>\n",
       "      <th>90145</th>\n",
       "      <th>90160A</th>\n",
       "      <th>90160B</th>\n",
       "      <th>90160C</th>\n",
       "      <th>90160D</th>\n",
       "      <th>90161B</th>\n",
       "      <th>90161C</th>\n",
       "      <th>90161D</th>\n",
       "      <th>90162A</th>\n",
       "      <th>90162B</th>\n",
       "      <th>90164A</th>\n",
       "      <th>90170</th>\n",
       "      <th>90173</th>\n",
       "      <th>90184B</th>\n",
       "      <th>90184C</th>\n",
       "      <th>90192</th>\n",
       "      <th>90201A</th>\n",
       "      <th>90201B</th>\n",
       "      <th>90201C</th>\n",
       "      <th>90201D</th>\n",
       "      <th>90202D</th>\n",
       "      <th>90204</th>\n",
       "      <th>C2</th>\n",
       "      <th>M</th>\n",
       "      <th>POST</th>\n",
       "      <th>CustomerID</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>12583</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>12583</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>12583</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>12583</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>12583</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 2575 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   10002  10120  10125  10133  10135  11001  15034  15036  15039  15044A  \\\n",
       "0      0      0      0      0      0      0      0      0      0       0   \n",
       "1      0      0      0      0      0      0      0      0      0       0   \n",
       "2      0      0      0      0      0      0      0      0      0       0   \n",
       "3      0      0      0      0      0      0      0      0      0       0   \n",
       "4      0      0      0      0      0      0      0      0      0       0   \n",
       "\n",
       "   15044B  15044C  15044D  15056BL  15056N  15056P  15058A  15058B  15058C  \\\n",
       "0       0       0       0        0       0       0       0       0       0   \n",
       "1       0       0       0        0       0       0       0       0       0   \n",
       "2       0       0       0        0       0       0       0       0       0   \n",
       "3       0       0       0        0       0       0       0       0       0   \n",
       "4       0       0       0        0       0       0       0       0       0   \n",
       "\n",
       "   15060B  16008  16011  16012  16014  16016  16045  16048  16054  16156L  \\\n",
       "0       0      0      0      0      0      0      0      0      0       0   \n",
       "1       0      0      0      0      0      0      0      0      0       0   \n",
       "2       0      0      0      0      0      0      0      0      0       0   \n",
       "3       0      0      0      0      0      0      0      0      0       0   \n",
       "4       0      0      0      0      0      0      0      0      0       0   \n",
       "\n",
       "   16156S  16161G  16161P  16161U  16168M  16169E  16169K  16169M  16218  \\\n",
       "0       0       0       0       0       0       0       0       0      0   \n",
       "1       0       0       0       0       0       0       0       0      0   \n",
       "2       0       0       0       0       0       0       0       0      0   \n",
       "3       0       0       0       0       0       0       0       0      0   \n",
       "4       0       0       0       0       0       0       0       0      0   \n",
       "\n",
       "   16219  16225  16235  16236  16237  16238  16258A  16259  17003  17011F  \\\n",
       "0      0      0      0      0      0      0       0      0      0       0   \n",
       "1      0      0      0      0      0      0       0      0      0       0   \n",
       "2      0      0      0      0      0      0       0      0      0       0   \n",
       "3      0      0      0      0      0      0       0      0      0       0   \n",
       "4      0      0      0      0      0      0       0      0      0       0   \n",
       "\n",
       "   17012A  17012B     ...      85232A  85232B  85232D  90001B  90001D  90013A  \\\n",
       "0       0       0     ...           0       0       0       0       0       0   \n",
       "1       0       0     ...           0       0       0       0       0       0   \n",
       "2       0       0     ...           0       0       0       0       0       0   \n",
       "3       0       0     ...           0       0       0       0       0       0   \n",
       "4       0       0     ...           0       0       0       0       0       0   \n",
       "\n",
       "   90013C  90018C  90019A  90024B  90030A  90030B  90030C  90031  90036E  \\\n",
       "0       0       0       0       0       0       0       0      0       0   \n",
       "1       0       0       0       0       0       0       0      0       0   \n",
       "2       0       0       0       0       0       0       0      0       0   \n",
       "3       0       0       0       0       0       0       0      0       0   \n",
       "4       0       0       0       0       0       0       0      0       0   \n",
       "\n",
       "   90037B  90057  90070  90087  90098  90099  90108  90114  90120B  90145  \\\n",
       "0       0      0      0      0      0      0      0      0       0      0   \n",
       "1       0      0      0      0      0      0      0      0       0      0   \n",
       "2       0      0      0      0      0      0      0      0       0      0   \n",
       "3       0      0      0      0      0      0      0      0       0      0   \n",
       "4       0      0      0      0      0      0      0      0       0      0   \n",
       "\n",
       "   90160A  90160B  90160C  90160D  90161B  90161C  90161D  90162A  90162B  \\\n",
       "0       0       0       0       0       0       0       0       0       0   \n",
       "1       0       0       0       0       0       0       0       0       0   \n",
       "2       0       0       0       0       0       0       0       0       0   \n",
       "3       0       0       0       0       0       0       0       0       0   \n",
       "4       0       0       0       0       0       0       0       0       0   \n",
       "\n",
       "   90164A  90170  90173  90184B  90184C  90192  90201A  90201B  90201C  \\\n",
       "0       0      0      0       0       0      0       0       0       0   \n",
       "1       0      0      0       0       0      0       0       0       0   \n",
       "2       0      0      0       0       0      0       0       0       0   \n",
       "3       0      0      0       0       0      0       0       0       0   \n",
       "4       0      0      0       0       0      0       0       0       0   \n",
       "\n",
       "   90201D  90202D  90204  C2  M  POST  CustomerID  \n",
       "0       0       0      0   0  0     0       12583  \n",
       "1       0       0      0   0  0     0       12583  \n",
       "2       0       0      0   0  0     0       12583  \n",
       "3       0       0      0   0  0     0       12583  \n",
       "4       0       0      0   0  0     0       12583  \n",
       "\n",
       "[5 rows x 2575 columns]"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Add CustomerID to item_dummies\n",
    "item_dummies['CustomerID'] = df.CustomerID\n",
    "\n",
    "# Display first 5 rows of item_dummies\n",
    "item_dummies.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Next, roll up the item dummies data into customer-level item data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>10002</th>\n",
       "      <th>10120</th>\n",
       "      <th>10125</th>\n",
       "      <th>10133</th>\n",
       "      <th>10135</th>\n",
       "      <th>11001</th>\n",
       "      <th>15034</th>\n",
       "      <th>15036</th>\n",
       "      <th>15039</th>\n",
       "      <th>15044A</th>\n",
       "      <th>15044B</th>\n",
       "      <th>15044C</th>\n",
       "      <th>15044D</th>\n",
       "      <th>15056BL</th>\n",
       "      <th>15056N</th>\n",
       "      <th>15056P</th>\n",
       "      <th>15058A</th>\n",
       "      <th>15058B</th>\n",
       "      <th>15058C</th>\n",
       "      <th>15060B</th>\n",
       "      <th>16008</th>\n",
       "      <th>16011</th>\n",
       "      <th>16012</th>\n",
       "      <th>16014</th>\n",
       "      <th>16016</th>\n",
       "      <th>16045</th>\n",
       "      <th>16048</th>\n",
       "      <th>16054</th>\n",
       "      <th>16156L</th>\n",
       "      <th>16156S</th>\n",
       "      <th>16161G</th>\n",
       "      <th>16161P</th>\n",
       "      <th>16161U</th>\n",
       "      <th>16168M</th>\n",
       "      <th>16169E</th>\n",
       "      <th>16169K</th>\n",
       "      <th>16169M</th>\n",
       "      <th>16218</th>\n",
       "      <th>16219</th>\n",
       "      <th>16225</th>\n",
       "      <th>16235</th>\n",
       "      <th>16236</th>\n",
       "      <th>16237</th>\n",
       "      <th>16238</th>\n",
       "      <th>16258A</th>\n",
       "      <th>16259</th>\n",
       "      <th>17003</th>\n",
       "      <th>17011F</th>\n",
       "      <th>17012A</th>\n",
       "      <th>17012B</th>\n",
       "      <th>...</th>\n",
       "      <th>85231B</th>\n",
       "      <th>85232A</th>\n",
       "      <th>85232B</th>\n",
       "      <th>85232D</th>\n",
       "      <th>90001B</th>\n",
       "      <th>90001D</th>\n",
       "      <th>90013A</th>\n",
       "      <th>90013C</th>\n",
       "      <th>90018C</th>\n",
       "      <th>90019A</th>\n",
       "      <th>90024B</th>\n",
       "      <th>90030A</th>\n",
       "      <th>90030B</th>\n",
       "      <th>90030C</th>\n",
       "      <th>90031</th>\n",
       "      <th>90036E</th>\n",
       "      <th>90037B</th>\n",
       "      <th>90057</th>\n",
       "      <th>90070</th>\n",
       "      <th>90087</th>\n",
       "      <th>90098</th>\n",
       "      <th>90099</th>\n",
       "      <th>90108</th>\n",
       "      <th>90114</th>\n",
       "      <th>90120B</th>\n",
       "      <th>90145</th>\n",
       "      <th>90160A</th>\n",
       "      <th>90160B</th>\n",
       "      <th>90160C</th>\n",
       "      <th>90160D</th>\n",
       "      <th>90161B</th>\n",
       "      <th>90161C</th>\n",
       "      <th>90161D</th>\n",
       "      <th>90162A</th>\n",
       "      <th>90162B</th>\n",
       "      <th>90164A</th>\n",
       "      <th>90170</th>\n",
       "      <th>90173</th>\n",
       "      <th>90184B</th>\n",
       "      <th>90184C</th>\n",
       "      <th>90192</th>\n",
       "      <th>90201A</th>\n",
       "      <th>90201B</th>\n",
       "      <th>90201C</th>\n",
       "      <th>90201D</th>\n",
       "      <th>90202D</th>\n",
       "      <th>90204</th>\n",
       "      <th>C2</th>\n",
       "      <th>M</th>\n",
       "      <th>POST</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CustomerID</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>12347</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12348</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12349</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12350</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12352</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 2574 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            10002  10120  10125  10133  10135  11001  15034  15036  15039  \\\n",
       "CustomerID                                                                  \n",
       "12347           0      0      0      0      0      0      0      0      0   \n",
       "12348           0      0      0      0      0      0      0      0      0   \n",
       "12349           0      0      0      0      0      0      0      0      0   \n",
       "12350           0      0      0      0      0      0      0      0      0   \n",
       "12352           0      0      0      0      0      0      0      0      0   \n",
       "\n",
       "            15044A  15044B  15044C  15044D  15056BL  15056N  15056P  15058A  \\\n",
       "CustomerID                                                                    \n",
       "12347            0       0       0       0        0       0       0       0   \n",
       "12348            0       0       0       0        0       0       0       0   \n",
       "12349            0       0       0       0        0       0       0       0   \n",
       "12350            0       0       0       0        0       0       0       0   \n",
       "12352            0       0       0       0        0       0       0       0   \n",
       "\n",
       "            15058B  15058C  15060B  16008  16011  16012  16014  16016  16045  \\\n",
       "CustomerID                                                                     \n",
       "12347            0       0       0      1      0      0      0      0      0   \n",
       "12348            0       0       0      0      0      0      0      0      0   \n",
       "12349            0       0       0      0      0      0      0      0      0   \n",
       "12350            0       0       0      0      0      0      0      0      0   \n",
       "12352            0       0       0      0      0      0      0      0      0   \n",
       "\n",
       "            16048  16054  16156L  16156S  16161G  16161P  16161U  16168M  \\\n",
       "CustomerID                                                                 \n",
       "12347           0      0       0       0       0       0       0       0   \n",
       "12348           0      0       0       0       0       0       0       0   \n",
       "12349           0      0       0       0       0       0       0       0   \n",
       "12350           0      0       0       0       0       0       0       0   \n",
       "12352           0      0       0       0       0       0       0       0   \n",
       "\n",
       "            16169E  16169K  16169M  16218  16219  16225  16235  16236  16237  \\\n",
       "CustomerID                                                                     \n",
       "12347            0       0       0      0      0      0      0      0      0   \n",
       "12348            0       0       0      0      0      0      0      0      0   \n",
       "12349            0       0       0      0      0      0      0      0      0   \n",
       "12350            0       0       0      0      0      0      0      0      0   \n",
       "12352            0       0       0      0      0      0      0      0      0   \n",
       "\n",
       "            16238  16258A  16259  17003  17011F  17012A  17012B  ...   85231B  \\\n",
       "CustomerID                                                       ...            \n",
       "12347           0       0      0      0       0       0       0  ...        0   \n",
       "12348           0       0      0      0       0       0       0  ...        0   \n",
       "12349           0       0      0      0       0       0       0  ...        0   \n",
       "12350           0       0      0      0       0       0       0  ...        0   \n",
       "12352           0       0      0      0       0       0       0  ...        0   \n",
       "\n",
       "            85232A  85232B  85232D  90001B  90001D  90013A  90013C  90018C  \\\n",
       "CustomerID                                                                   \n",
       "12347            0       0       1       0       0       0       0       0   \n",
       "12348            0       0       0       0       0       0       0       0   \n",
       "12349            0       0       0       0       0       0       0       0   \n",
       "12350            0       0       0       0       0       0       0       0   \n",
       "12352            0       0       0       0       0       0       0       0   \n",
       "\n",
       "            90019A  90024B  90030A  90030B  90030C  90031  90036E  90037B  \\\n",
       "CustomerID                                                                  \n",
       "12347            0       0       0       0       0      0       0       0   \n",
       "12348            0       0       0       0       0      0       0       0   \n",
       "12349            0       0       0       0       0      0       0       0   \n",
       "12350            0       0       0       0       0      0       0       0   \n",
       "12352            0       0       0       0       0      0       0       0   \n",
       "\n",
       "            90057  90070  90087  90098  90099  90108  90114  90120B  90145  \\\n",
       "CustomerID                                                                   \n",
       "12347           0      0      0      0      0      0      0       0      0   \n",
       "12348           0      0      0      0      0      0      0       0      0   \n",
       "12349           0      0      0      0      0      0      0       0      0   \n",
       "12350           0      0      0      0      0      0      0       0      0   \n",
       "12352           0      0      0      0      0      0      0       0      0   \n",
       "\n",
       "            90160A  90160B  90160C  90160D  90161B  90161C  90161D  90162A  \\\n",
       "CustomerID                                                                   \n",
       "12347            0       0       0       0       0       0       0       0   \n",
       "12348            0       0       0       0       0       0       0       0   \n",
       "12349            0       0       0       0       0       0       0       0   \n",
       "12350            0       0       0       0       0       0       0       0   \n",
       "12352            0       0       0       0       0       0       0       0   \n",
       "\n",
       "            90162B  90164A  90170  90173  90184B  90184C  90192  90201A  \\\n",
       "CustomerID                                                                \n",
       "12347            0       0      0      0       0       0      0       0   \n",
       "12348            0       0      0      0       0       0      0       0   \n",
       "12349            0       0      0      0       0       0      0       0   \n",
       "12350            0       0      0      0       0       0      0       0   \n",
       "12352            0       0      0      0       0       0      0       0   \n",
       "\n",
       "            90201B  90201C  90201D  90202D  90204  C2  M  POST  \n",
       "CustomerID                                                      \n",
       "12347            0       0       0       0      0   0  0     0  \n",
       "12348            0       0       0       0      0   0  0     4  \n",
       "12349            0       0       0       0      0   0  0     1  \n",
       "12350            0       0       0       0      0   0  0     1  \n",
       "12352            0       0       0       0      0   0  3     5  \n",
       "\n",
       "[5 rows x 2574 columns]"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Create item_data by aggregating at customer level\n",
    "item_data = item_dummies.groupby('CustomerID').sum()\n",
    "\n",
    "# Display first 5 rows of item_data\n",
    "item_data.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As you can see, even after rolling up to the customer level, most of the values are still 0. That means that most customers are not buying a huge array of different items, which is to be expected."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Finally, let's display the total number times each item was purchased."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "10002        12\n",
       "10120         1\n",
       "10125        13\n",
       "10133         5\n",
       "10135         4\n",
       "11001         8\n",
       "15034         5\n",
       "15036        19\n",
       "15039         3\n",
       "15044A        6\n",
       "15044B        3\n",
       "15044C        2\n",
       "15044D        4\n",
       "15056BL      50\n",
       "15056N       35\n",
       "15056P       24\n",
       "15058A        9\n",
       "15058B        8\n",
       "15058C        4\n",
       "15060B       12\n",
       "16008        11\n",
       "16011         3\n",
       "16012         4\n",
       "16014        10\n",
       "16016        16\n",
       "16045         8\n",
       "16048         8\n",
       "16054         2\n",
       "16156L        6\n",
       "16156S       12\n",
       "           ... \n",
       "90098         1\n",
       "90099         2\n",
       "90108         1\n",
       "90114         1\n",
       "90120B        1\n",
       "90145         2\n",
       "90160A        1\n",
       "90160B        1\n",
       "90160C        1\n",
       "90160D        1\n",
       "90161B        1\n",
       "90161C        1\n",
       "90161D        1\n",
       "90162A        1\n",
       "90162B        1\n",
       "90164A        1\n",
       "90170         1\n",
       "90173         1\n",
       "90184B        1\n",
       "90184C        1\n",
       "90192         1\n",
       "90201A        1\n",
       "90201B        3\n",
       "90201C        2\n",
       "90201D        1\n",
       "90202D        1\n",
       "90204         1\n",
       "C2            6\n",
       "M            34\n",
       "POST       1055\n",
       "Length: 2574, dtype: int64"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Total times each item was purchased\n",
    "item_data.sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As you can see, most items were purchased less than a handful of times!\n",
    "\n",
    "First of all, we've just created 2574 customer-level item features, which leads to The Curse of Dimensionality.\n",
    "To make matters even worse, most of the values for many of those features are 0!\n",
    "\n",
    "So, we'll introduce a couple of strategies for reducing the number of item features that we actually keep."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Before moving on, let's save this customer-level item dataframe as 'item_data.csv'. We'll use it again in the next module."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Save item_data.csv\n",
    "item_data.to_csv('Files/item_data.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 8.3 Method 1 - Thresholding\n",
    "One very simple and straightforward way to reduce the dimensionality of this item data is to set a threshold for keeping features.\n",
    "* The rationale is that you might only want to keep **popular items.**\n",
    "* For example, let's say item A was only purchased by 2 customers. Well, the feature for item A will be 0 for almost all observations, which isn't very helpful.\n",
    "* On the other hand, let's say item B was purchased by 100 customers. The feature for item B will allow more meaningful comparisons.\n",
    "\n",
    "To make this concrete, assume we only wish to keep item features for the 20 most popular items.\n",
    "\n",
    "First, we can see which items those are and the number of times they were purchased.\n",
    "* Take the sum by column.\n",
    "* Sort the values.\n",
    "* Look at the last 20 (since they are sorted in ascending order by default)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 185,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "22748        50\n",
       "22847        50\n",
       "15056BL      50\n",
       "23159        50\n",
       "85014B       50\n",
       "23243        51\n",
       "23231        51\n",
       "22898        51\n",
       "23388        51\n",
       "23255        52\n",
       "22966        52\n",
       "22558        53\n",
       "21558        53\n",
       "22356        53\n",
       "84997C       53\n",
       "23292        53\n",
       "22699        54\n",
       "22331        54\n",
       "22385        54\n",
       "23480        55\n",
       "84375        55\n",
       "21156        55\n",
       "21242        55\n",
       "22417        56\n",
       "21122        56\n",
       "84997B       57\n",
       "22895        58\n",
       "22962        58\n",
       "22557        58\n",
       "22907        59\n",
       "           ... \n",
       "23206       103\n",
       "22352       104\n",
       "22727       104\n",
       "22138       104\n",
       "22720       104\n",
       "21559       106\n",
       "22728       108\n",
       "22555       109\n",
       "20724       110\n",
       "20712       110\n",
       "22961       114\n",
       "22630       115\n",
       "22139       117\n",
       "21080       122\n",
       "85099B      123\n",
       "20726       123\n",
       "20719       128\n",
       "20750       132\n",
       "23084       140\n",
       "20725       141\n",
       "21212       143\n",
       "22551       158\n",
       "22629       160\n",
       "22328       166\n",
       "21731       169\n",
       "22556       179\n",
       "22554       197\n",
       "22423       222\n",
       "22326       271\n",
       "POST       1055\n",
       "Length: 120, dtype: int64"
      ]
     },
     "execution_count": 185,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Display most popular 20 items\n",
    "item_data.sum().sort_values().tail(120)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Next, if we take the .index of the above series, we can get just a list of the StockCodes for those 20 items."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 186,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['22748', '22847', '15056BL', '23159', '85014B', '23243', '23231',\n",
       "       '22898', '23388', '23255',\n",
       "       ...\n",
       "       '21212', '22551', '22629', '22328', '21731', '22556', '22554', '22423',\n",
       "       '22326', 'POST'],\n",
       "      dtype='object', length=120)"
      ]
     },
     "execution_count": 186,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Get list of StockCodes for the 20 most popular items\n",
    "top_20_items = item_data.sum().sort_values().tail(120).index\n",
    "\n",
    "top_20_items"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Finally, we can keep only the features for those 20 items."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 187,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(414, 120)"
      ]
     },
     "execution_count": 187,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Keep only features for top 20 items\n",
    "top_20_item_data = item_data[top_20_items]\n",
    "\n",
    "# Shape of remaining dataframe\n",
    "top_20_item_data.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 188,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>22748</th>\n",
       "      <th>22847</th>\n",
       "      <th>15056BL</th>\n",
       "      <th>23159</th>\n",
       "      <th>85014B</th>\n",
       "      <th>23243</th>\n",
       "      <th>23231</th>\n",
       "      <th>22898</th>\n",
       "      <th>23388</th>\n",
       "      <th>23255</th>\n",
       "      <th>22966</th>\n",
       "      <th>22558</th>\n",
       "      <th>21558</th>\n",
       "      <th>22356</th>\n",
       "      <th>84997C</th>\n",
       "      <th>23292</th>\n",
       "      <th>22699</th>\n",
       "      <th>22331</th>\n",
       "      <th>22385</th>\n",
       "      <th>23480</th>\n",
       "      <th>84375</th>\n",
       "      <th>21156</th>\n",
       "      <th>21242</th>\n",
       "      <th>22417</th>\n",
       "      <th>21122</th>\n",
       "      <th>84997B</th>\n",
       "      <th>22895</th>\n",
       "      <th>22962</th>\n",
       "      <th>22557</th>\n",
       "      <th>22907</th>\n",
       "      <th>23307</th>\n",
       "      <th>20676</th>\n",
       "      <th>21977</th>\n",
       "      <th>21976</th>\n",
       "      <th>23293</th>\n",
       "      <th>22661</th>\n",
       "      <th>23290</th>\n",
       "      <th>22627</th>\n",
       "      <th>84879</th>\n",
       "      <th>20682</th>\n",
       "      <th>22367</th>\n",
       "      <th>23236</th>\n",
       "      <th>22355</th>\n",
       "      <th>22197</th>\n",
       "      <th>22553</th>\n",
       "      <th>20685</th>\n",
       "      <th>23108</th>\n",
       "      <th>22384</th>\n",
       "      <th>21936</th>\n",
       "      <th>20728</th>\n",
       "      <th>...</th>\n",
       "      <th>22631</th>\n",
       "      <th>22090</th>\n",
       "      <th>23199</th>\n",
       "      <th>21915</th>\n",
       "      <th>23203</th>\n",
       "      <th>84997D</th>\n",
       "      <th>22029</th>\n",
       "      <th>23209</th>\n",
       "      <th>23254</th>\n",
       "      <th>22077</th>\n",
       "      <th>22960</th>\n",
       "      <th>23256</th>\n",
       "      <th>22492</th>\n",
       "      <th>22467</th>\n",
       "      <th>22726</th>\n",
       "      <th>22382</th>\n",
       "      <th>23240</th>\n",
       "      <th>23245</th>\n",
       "      <th>21094</th>\n",
       "      <th>21086</th>\n",
       "      <th>23206</th>\n",
       "      <th>22352</th>\n",
       "      <th>22727</th>\n",
       "      <th>22138</th>\n",
       "      <th>22720</th>\n",
       "      <th>21559</th>\n",
       "      <th>22728</th>\n",
       "      <th>22555</th>\n",
       "      <th>20724</th>\n",
       "      <th>20712</th>\n",
       "      <th>22961</th>\n",
       "      <th>22630</th>\n",
       "      <th>22139</th>\n",
       "      <th>21080</th>\n",
       "      <th>85099B</th>\n",
       "      <th>20726</th>\n",
       "      <th>20719</th>\n",
       "      <th>20750</th>\n",
       "      <th>23084</th>\n",
       "      <th>20725</th>\n",
       "      <th>21212</th>\n",
       "      <th>22551</th>\n",
       "      <th>22629</th>\n",
       "      <th>22328</th>\n",
       "      <th>21731</th>\n",
       "      <th>22556</th>\n",
       "      <th>22554</th>\n",
       "      <th>22423</th>\n",
       "      <th>22326</th>\n",
       "      <th>POST</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CustomerID</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>12347</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12348</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12349</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</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>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12350</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12352</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 120 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            22748  22847  15056BL  23159  85014B  23243  23231  22898  23388  \\\n",
       "CustomerID                                                                     \n",
       "12347           0      0        0      0       0      0      0      0      0   \n",
       "12348           0      0        0      0       0      0      0      0      0   \n",
       "12349           0      0        0      0       1      0      0      0      0   \n",
       "12350           0      0        0      0       0      0      0      0      0   \n",
       "12352           0      0        0      0       0      1      0      0      0   \n",
       "\n",
       "            23255  22966  22558  21558  22356  84997C  23292  22699  22331  \\\n",
       "CustomerID                                                                   \n",
       "12347           0      0      0      0      0       1      0      2      0   \n",
       "12348           0      0      0      0      0       0      0      0      0   \n",
       "12349           0      0      0      0      0       0      0      0      0   \n",
       "12350           0      0      0      0      0       0      0      0      0   \n",
       "12352           0      0      0      0      0       0      0      0      0   \n",
       "\n",
       "            22385  23480  84375  21156  21242  22417  21122  84997B  22895  \\\n",
       "CustomerID                                                                   \n",
       "12347           0      1      0      0      0      2      0       1      0   \n",
       "12348           0      0      0      0      0      0      0       0      0   \n",
       "12349           0      0      0      0      0      0      0       0      0   \n",
       "12350           0      0      0      0      0      0      0       0      0   \n",
       "12352           0      0      0      0      0      0      0       0      0   \n",
       "\n",
       "            22962  22557  22907  23307  20676  21977  21976  23293  22661  \\\n",
       "CustomerID                                                                  \n",
       "12347           0      0      0      0      0      0      2      0      0   \n",
       "12348           0      0      0      0      0      2      0      0      0   \n",
       "12349           0      1      0      0      0      0      0      1      0   \n",
       "12350           0      1      0      0      0      0      0      0      0   \n",
       "12352           0      0      0      0      0      0      0      0      0   \n",
       "\n",
       "            23290  22627  84879  20682  22367  23236  22355  22197  22553  \\\n",
       "CustomerID                                                                  \n",
       "12347           0      0      0      0      0      0      0      0      0   \n",
       "12348           0      0      0      0      0      0      0      0      0   \n",
       "12349           0      0      0      0      0      1      0      0      1   \n",
       "12350           0      0      0      0      0      0      0      0      0   \n",
       "12352           0      1      0      0      0      0      0      0      0   \n",
       "\n",
       "            20685  23108  22384  21936  20728  ...   22631  22090  23199  \\\n",
       "CustomerID                                     ...                         \n",
       "12347           0      0      0      0      0  ...       0      0      0   \n",
       "12348           0      0      0      0      0  ...       0      0      0   \n",
       "12349           1      1      0      0      0  ...       0      0      0   \n",
       "12350           0      0      0      0      0  ...       0      0      0   \n",
       "12352           0      0      0      0      0  ...       0      0      0   \n",
       "\n",
       "            21915  23203  84997D  22029  23209  23254  22077  22960  23256  \\\n",
       "CustomerID                                                                   \n",
       "12347           0      0       1      0      0      0      0      0      0   \n",
       "12348           0      0       0      0      0      0      0      0      0   \n",
       "12349           0      0       0      0      0      0      0      1      0   \n",
       "12350           1      0       0      0      0      0      0      0      0   \n",
       "12352           0      0       0      0      0      0      0      0      0   \n",
       "\n",
       "            22492  22467  22726  22382  23240  23245  21094  21086  23206  \\\n",
       "CustomerID                                                                  \n",
       "12347           3      0      3      0      0      0      0      0      0   \n",
       "12348           0      0      0      0      0      0      0      0      0   \n",
       "12349           0      0      0      0      1      0      0      1      0   \n",
       "12350           0      0      0      0      0      0      0      0      0   \n",
       "12352           0      0      0      0      0      2      0      0      0   \n",
       "\n",
       "            22352  22727  22138  22720  21559  22728  22555  20724  20712  \\\n",
       "CustomerID                                                                  \n",
       "12347           0      5      0      0      0      3      0      0      0   \n",
       "12348           0      0      0      0      0      0      0      0      0   \n",
       "12349           0      0      0      1      0      0      1      0      0   \n",
       "12350           0      0      0      0      0      0      0      0      0   \n",
       "12352           0      0      1      2      0      0      0      0      0   \n",
       "\n",
       "            22961  22630  22139  21080  85099B  20726  20719  20750  23084  \\\n",
       "CustomerID                                                                   \n",
       "12347           0      0      0      0       0      0      4      0      3   \n",
       "12348           0      0      0      0       0      0      0      0      0   \n",
       "12349           0      0      0      0       0      0      0      0      0   \n",
       "12350           0      0      0      0       0      0      0      0      0   \n",
       "12352           0      1      0      0       0      0      0      0      0   \n",
       "\n",
       "            20725  21212  22551  22629  22328  21731  22556  22554  22423  \\\n",
       "CustomerID                                                                  \n",
       "12347           0      0      0      0      0      5      0      0      4   \n",
       "12348           0      0      0      0      0      0      0      0      0   \n",
       "12349           0      0      0      0      0      0      1      1      1   \n",
       "12350           0      0      1      0      0      0      0      0      0   \n",
       "12352           0      0      0      0      0      1      0      0      2   \n",
       "\n",
       "            22326  POST  \n",
       "CustomerID               \n",
       "12347           0     0  \n",
       "12348           0     4  \n",
       "12349           1     1  \n",
       "12350           0     1  \n",
       "12352           0     5  \n",
       "\n",
       "[5 rows x 120 columns]"
      ]
     },
     "execution_count": 188,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "top_20_item_data.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "These 20 features are much more manageable than the 2574 from earlier, and they are arguably the most important features because they are the most popular items.\n",
    "\n",
    "Finally, save this top 20 items dataframe as 'threshold_item_data.csv'."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 189,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Save threshold_item_data.csv\n",
    "top_20_item_data.to_csv('Files/threshold_item_data.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 8.4 PCA\n",
    "Let's import the full item data that we saved in the previous module (before applying thresholds)\n",
    "\n",
    "This time, we'll also pass in the argument index_col=0 to tell Pandas to treat the first column (CustomerID) as the index."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 135,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Read item_data.csv\n",
    "item_data = pd.read_csv('Files/item_data.csv', index_col=0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Just to confirm, this dataset should have 2574 features:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(414, 2574)"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Display item_data's shape\n",
    "item_data.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Principal Component Analysis (PCA)** is an Unsupervised Learning task that creates a sequence of new, uncorrelated features that each try to maximize its \"explained variance\" of the original dataset.\n",
    "* It does so by generating linear combinations from your original features.\n",
    "* These new features are meant to replace the original ones.\n",
    "\n",
    "Here's where dimensionality reduction comes into play, and it's brilliantly simple:\n",
    "* You don't need to keep all of the principal components!\n",
    "* You can just keep some number of the \"best\" components, a.k.a. the ones that explain the most variance.\n",
    "* Remember, PCA creates a sequence of principal components and each one tries to capture the most variance after accounting for the ones before it.\n",
    "\n",
    "First, **scale item_data**, which you imported."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[-0.13303802, -0.04920678, -0.1119152 , ..., -0.10482848,\n",
       "        -0.15568183, -0.68068072],\n",
       "       [-0.13303802, -0.04920678, -0.1119152 , ..., -0.10482848,\n",
       "        -0.15568183,  0.38776219],\n",
       "       [-0.13303802, -0.04920678, -0.1119152 , ..., -0.10482848,\n",
       "        -0.15568183, -0.41356999],\n",
       "       [-0.13303802, -0.04920678, -0.1119152 , ..., -0.10482848,\n",
       "        -0.15568183, -0.41356999],\n",
       "       [-0.13303802, -0.04920678, -0.1119152 , ..., -0.10482848,\n",
       "         5.53128397,  0.65487292]])"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Initialize instance of StandardScaler\n",
    "scaler = StandardScaler()\n",
    "\n",
    "# Fit and transform item_data\n",
    "item_data_scaled = scaler.fit_transform(item_data)\n",
    "\n",
    "# Display first 5 rows of item_data_scaled\n",
    "item_data_scaled[:5]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Next, initialize and fit an instance of the PCA transformation.\n",
    "\n",
    "Keep all of the components for now (just don't pass in any argument)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "PCA(copy=True, iterated_power='auto', n_components=None, random_state=None,\n",
       "  svd_solver='auto', tol=0.0, whiten=False)"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Initialize and fit a PCA transformation\n",
    "pca = PCA()\n",
    "pca.fit(item_data_scaled)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Finally, generate new \"principal component features\" from item_data_scaled."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[  5.87614052e+00,   4.82891689e+00,  -4.95545069e-01, ...,\n",
       "         -6.62646278e-04,  -4.43629806e-05,   5.39585737e-15],\n",
       "       [ -4.21983747e+00,   1.22621722e+00,  -1.88191211e+00, ...,\n",
       "         -2.11471782e-04,  -2.69154340e-04,   2.28810026e-15],\n",
       "       [ -1.26084676e+00,   2.38766371e+00,  -7.31483440e-02, ...,\n",
       "          2.57750179e-04,  -2.05682324e-04,   1.98105421e-15],\n",
       "       [ -4.92931983e+00,  -7.14858724e-01,  -1.58093026e+00, ...,\n",
       "         -1.98603886e-03,   9.92316312e-05,   4.06272238e-15],\n",
       "       [ -2.05311348e+00,   2.11095022e+00,   1.84807021e-01, ...,\n",
       "         -3.43415269e-04,  -1.32981672e-04,   1.27120536e-14]])"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Generate new features\n",
    "PC_items = pca.transform(item_data_scaled)\n",
    "\n",
    "# Display first 5 rows\n",
    "PC_items[:5]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Explained Variance\n",
    "It's very helpful to calculate and plot the cumulative explained variance.\n",
    "* This will tell us the total amount of variance we'd capture if we kept up to the n-th component.\n",
    "* First, we'll use np.cumsum() to calculate the cumulative explained variance.\n",
    "* Then, we'll plot it so we can see how many PC features we'd need to keep in order to capture most of the original variance."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x15db65e20f0>]"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXcAAAD8CAYAAACMwORRAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x15db65e29b0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Cumulative explained variance\n",
    "cumulative_explained_variance = np.cumsum(pca.explained_variance_ratio_)\n",
    "\n",
    "# Plot cumulative explained variance\n",
    "plt.grid()\n",
    "plt.plot(range(len(cumulative_explained_variance)), cumulative_explained_variance)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This chart is saying: To capture about 98% of the variance, we'd need to keep around 300 components.\n",
    "\n",
    "We can confirm:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.98242418319166058"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# How much variance we'd capture with the first 125 components\n",
    "cumulative_explained_variance[300]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Reducing 2574 features down to 300 (about 88% fewer features) while capturing almost 80% of the original variance is certainly not bad!"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Initialize and fit another PCA transformation.\n",
    "* This time, only keep 300 components.\n",
    "* Generate the principal component features from the fitted instance and name the new matrix PC_items.\n",
    "* Then, display the shape of PC_items to confirm it only has 300 features."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 162,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(414, 300)"
      ]
     },
     "execution_count": 162,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pca = PCA(n_components=300)\n",
    "\n",
    "# Fit and transform item_data_scaled\n",
    "PC_items = pca.fit_transform(item_data_scaled)\n",
    "\n",
    "# Display shape of PC_items\n",
    "PC_items.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Next, for convenience, let's put PC_items into a new dataframe.\n",
    "\n",
    "We'll also name the columns and update its index to be the same as the orginal item_data's index."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 163,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>PC1</th>\n",
       "      <th>PC2</th>\n",
       "      <th>PC3</th>\n",
       "      <th>PC4</th>\n",
       "      <th>PC5</th>\n",
       "      <th>PC6</th>\n",
       "      <th>PC7</th>\n",
       "      <th>PC8</th>\n",
       "      <th>PC9</th>\n",
       "      <th>PC10</th>\n",
       "      <th>PC11</th>\n",
       "      <th>PC12</th>\n",
       "      <th>PC13</th>\n",
       "      <th>PC14</th>\n",
       "      <th>PC15</th>\n",
       "      <th>PC16</th>\n",
       "      <th>PC17</th>\n",
       "      <th>PC18</th>\n",
       "      <th>PC19</th>\n",
       "      <th>PC20</th>\n",
       "      <th>PC21</th>\n",
       "      <th>PC22</th>\n",
       "      <th>PC23</th>\n",
       "      <th>PC24</th>\n",
       "      <th>PC25</th>\n",
       "      <th>PC26</th>\n",
       "      <th>PC27</th>\n",
       "      <th>PC28</th>\n",
       "      <th>PC29</th>\n",
       "      <th>PC30</th>\n",
       "      <th>PC31</th>\n",
       "      <th>PC32</th>\n",
       "      <th>PC33</th>\n",
       "      <th>PC34</th>\n",
       "      <th>PC35</th>\n",
       "      <th>PC36</th>\n",
       "      <th>PC37</th>\n",
       "      <th>PC38</th>\n",
       "      <th>PC39</th>\n",
       "      <th>PC40</th>\n",
       "      <th>PC41</th>\n",
       "      <th>PC42</th>\n",
       "      <th>PC43</th>\n",
       "      <th>PC44</th>\n",
       "      <th>PC45</th>\n",
       "      <th>PC46</th>\n",
       "      <th>PC47</th>\n",
       "      <th>PC48</th>\n",
       "      <th>PC49</th>\n",
       "      <th>PC50</th>\n",
       "      <th>...</th>\n",
       "      <th>PC251</th>\n",
       "      <th>PC252</th>\n",
       "      <th>PC253</th>\n",
       "      <th>PC254</th>\n",
       "      <th>PC255</th>\n",
       "      <th>PC256</th>\n",
       "      <th>PC257</th>\n",
       "      <th>PC258</th>\n",
       "      <th>PC259</th>\n",
       "      <th>PC260</th>\n",
       "      <th>PC261</th>\n",
       "      <th>PC262</th>\n",
       "      <th>PC263</th>\n",
       "      <th>PC264</th>\n",
       "      <th>PC265</th>\n",
       "      <th>PC266</th>\n",
       "      <th>PC267</th>\n",
       "      <th>PC268</th>\n",
       "      <th>PC269</th>\n",
       "      <th>PC270</th>\n",
       "      <th>PC271</th>\n",
       "      <th>PC272</th>\n",
       "      <th>PC273</th>\n",
       "      <th>PC274</th>\n",
       "      <th>PC275</th>\n",
       "      <th>PC276</th>\n",
       "      <th>PC277</th>\n",
       "      <th>PC278</th>\n",
       "      <th>PC279</th>\n",
       "      <th>PC280</th>\n",
       "      <th>PC281</th>\n",
       "      <th>PC282</th>\n",
       "      <th>PC283</th>\n",
       "      <th>PC284</th>\n",
       "      <th>PC285</th>\n",
       "      <th>PC286</th>\n",
       "      <th>PC287</th>\n",
       "      <th>PC288</th>\n",
       "      <th>PC289</th>\n",
       "      <th>PC290</th>\n",
       "      <th>PC291</th>\n",
       "      <th>PC292</th>\n",
       "      <th>PC293</th>\n",
       "      <th>PC294</th>\n",
       "      <th>PC295</th>\n",
       "      <th>PC296</th>\n",
       "      <th>PC297</th>\n",
       "      <th>PC298</th>\n",
       "      <th>PC299</th>\n",
       "      <th>PC300</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CustomerID</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>12347</th>\n",
       "      <td>5.876141</td>\n",
       "      <td>4.828917</td>\n",
       "      <td>-0.495545</td>\n",
       "      <td>3.093849</td>\n",
       "      <td>-0.633706</td>\n",
       "      <td>-11.534092</td>\n",
       "      <td>2.161996</td>\n",
       "      <td>7.116957</td>\n",
       "      <td>5.336009</td>\n",
       "      <td>4.424414</td>\n",
       "      <td>4.637002</td>\n",
       "      <td>2.316503</td>\n",
       "      <td>12.310783</td>\n",
       "      <td>1.986696</td>\n",
       "      <td>-11.431269</td>\n",
       "      <td>5.617791</td>\n",
       "      <td>-2.177268</td>\n",
       "      <td>7.665832</td>\n",
       "      <td>-12.135817</td>\n",
       "      <td>-16.049000</td>\n",
       "      <td>-5.949238</td>\n",
       "      <td>7.814288</td>\n",
       "      <td>40.518194</td>\n",
       "      <td>-35.575039</td>\n",
       "      <td>-17.557418</td>\n",
       "      <td>6.869259</td>\n",
       "      <td>20.886868</td>\n",
       "      <td>2.296657</td>\n",
       "      <td>-26.160394</td>\n",
       "      <td>-32.580135</td>\n",
       "      <td>7.880285</td>\n",
       "      <td>12.358371</td>\n",
       "      <td>6.826718</td>\n",
       "      <td>-32.957865</td>\n",
       "      <td>-4.405057</td>\n",
       "      <td>-17.238628</td>\n",
       "      <td>-11.272274</td>\n",
       "      <td>-14.487767</td>\n",
       "      <td>-2.766330</td>\n",
       "      <td>4.736717</td>\n",
       "      <td>-6.985884</td>\n",
       "      <td>-1.464059</td>\n",
       "      <td>10.603071</td>\n",
       "      <td>1.239909</td>\n",
       "      <td>3.086324</td>\n",
       "      <td>5.336237</td>\n",
       "      <td>-0.410776</td>\n",
       "      <td>1.616712</td>\n",
       "      <td>-6.383660</td>\n",
       "      <td>3.065176</td>\n",
       "      <td>...</td>\n",
       "      <td>0.064395</td>\n",
       "      <td>-0.415426</td>\n",
       "      <td>-0.206658</td>\n",
       "      <td>-0.005371</td>\n",
       "      <td>0.188138</td>\n",
       "      <td>0.242410</td>\n",
       "      <td>-0.195482</td>\n",
       "      <td>-0.162702</td>\n",
       "      <td>-0.200203</td>\n",
       "      <td>-0.170627</td>\n",
       "      <td>0.118951</td>\n",
       "      <td>-0.124982</td>\n",
       "      <td>-0.092571</td>\n",
       "      <td>-0.209632</td>\n",
       "      <td>-0.122009</td>\n",
       "      <td>0.112249</td>\n",
       "      <td>0.097985</td>\n",
       "      <td>0.047909</td>\n",
       "      <td>0.191655</td>\n",
       "      <td>-0.245912</td>\n",
       "      <td>-0.183734</td>\n",
       "      <td>-0.034925</td>\n",
       "      <td>0.108376</td>\n",
       "      <td>-0.060922</td>\n",
       "      <td>-0.060365</td>\n",
       "      <td>-0.033723</td>\n",
       "      <td>0.029060</td>\n",
       "      <td>-0.006391</td>\n",
       "      <td>-0.001222</td>\n",
       "      <td>0.040427</td>\n",
       "      <td>0.089004</td>\n",
       "      <td>-0.218621</td>\n",
       "      <td>0.158177</td>\n",
       "      <td>0.049166</td>\n",
       "      <td>-0.037660</td>\n",
       "      <td>-0.024126</td>\n",
       "      <td>0.094181</td>\n",
       "      <td>-0.109886</td>\n",
       "      <td>-0.035857</td>\n",
       "      <td>0.015897</td>\n",
       "      <td>-0.071372</td>\n",
       "      <td>-0.046723</td>\n",
       "      <td>0.109823</td>\n",
       "      <td>0.021534</td>\n",
       "      <td>-0.073246</td>\n",
       "      <td>-0.005525</td>\n",
       "      <td>-0.046094</td>\n",
       "      <td>0.134262</td>\n",
       "      <td>-0.046511</td>\n",
       "      <td>0.068503</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12348</th>\n",
       "      <td>-4.219837</td>\n",
       "      <td>1.226217</td>\n",
       "      <td>-1.881912</td>\n",
       "      <td>-0.174340</td>\n",
       "      <td>0.817996</td>\n",
       "      <td>1.160834</td>\n",
       "      <td>-0.871683</td>\n",
       "      <td>-0.846634</td>\n",
       "      <td>0.460240</td>\n",
       "      <td>0.854274</td>\n",
       "      <td>-0.401793</td>\n",
       "      <td>0.981803</td>\n",
       "      <td>-1.104237</td>\n",
       "      <td>0.770190</td>\n",
       "      <td>-0.092975</td>\n",
       "      <td>-0.313393</td>\n",
       "      <td>-0.448633</td>\n",
       "      <td>1.207302</td>\n",
       "      <td>0.381678</td>\n",
       "      <td>-0.469152</td>\n",
       "      <td>0.185789</td>\n",
       "      <td>0.054585</td>\n",
       "      <td>0.499987</td>\n",
       "      <td>0.373514</td>\n",
       "      <td>1.226615</td>\n",
       "      <td>-0.773072</td>\n",
       "      <td>-0.270125</td>\n",
       "      <td>0.850016</td>\n",
       "      <td>-0.370481</td>\n",
       "      <td>-0.573939</td>\n",
       "      <td>1.012802</td>\n",
       "      <td>-0.580828</td>\n",
       "      <td>0.949802</td>\n",
       "      <td>-0.733970</td>\n",
       "      <td>0.353890</td>\n",
       "      <td>-0.387479</td>\n",
       "      <td>-0.298541</td>\n",
       "      <td>0.396418</td>\n",
       "      <td>0.154924</td>\n",
       "      <td>0.333479</td>\n",
       "      <td>-0.466261</td>\n",
       "      <td>0.703454</td>\n",
       "      <td>1.200580</td>\n",
       "      <td>-0.202163</td>\n",
       "      <td>-0.871794</td>\n",
       "      <td>-0.055069</td>\n",
       "      <td>-0.421733</td>\n",
       "      <td>-1.623011</td>\n",
       "      <td>-0.397691</td>\n",
       "      <td>0.452590</td>\n",
       "      <td>...</td>\n",
       "      <td>0.129957</td>\n",
       "      <td>1.022734</td>\n",
       "      <td>-0.456650</td>\n",
       "      <td>-1.224744</td>\n",
       "      <td>0.530329</td>\n",
       "      <td>-0.994510</td>\n",
       "      <td>-0.371455</td>\n",
       "      <td>-0.468078</td>\n",
       "      <td>0.489332</td>\n",
       "      <td>-0.620042</td>\n",
       "      <td>0.995225</td>\n",
       "      <td>-1.389717</td>\n",
       "      <td>1.171395</td>\n",
       "      <td>-0.617378</td>\n",
       "      <td>0.078084</td>\n",
       "      <td>1.336674</td>\n",
       "      <td>-0.120785</td>\n",
       "      <td>0.330687</td>\n",
       "      <td>-0.235496</td>\n",
       "      <td>0.112896</td>\n",
       "      <td>0.654065</td>\n",
       "      <td>0.388106</td>\n",
       "      <td>-0.654803</td>\n",
       "      <td>0.410856</td>\n",
       "      <td>-0.836039</td>\n",
       "      <td>1.419597</td>\n",
       "      <td>0.055308</td>\n",
       "      <td>0.657981</td>\n",
       "      <td>0.559558</td>\n",
       "      <td>0.225912</td>\n",
       "      <td>-0.892273</td>\n",
       "      <td>-0.111400</td>\n",
       "      <td>-0.459792</td>\n",
       "      <td>0.133336</td>\n",
       "      <td>-0.166611</td>\n",
       "      <td>0.015979</td>\n",
       "      <td>-1.039514</td>\n",
       "      <td>-0.153822</td>\n",
       "      <td>-0.368017</td>\n",
       "      <td>-0.717055</td>\n",
       "      <td>-0.187247</td>\n",
       "      <td>-0.006443</td>\n",
       "      <td>0.247051</td>\n",
       "      <td>0.526344</td>\n",
       "      <td>0.048497</td>\n",
       "      <td>0.309830</td>\n",
       "      <td>-0.974179</td>\n",
       "      <td>-0.156291</td>\n",
       "      <td>0.451286</td>\n",
       "      <td>1.135782</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12349</th>\n",
       "      <td>-1.260847</td>\n",
       "      <td>2.387664</td>\n",
       "      <td>-0.073148</td>\n",
       "      <td>-1.119161</td>\n",
       "      <td>-0.034339</td>\n",
       "      <td>-3.715390</td>\n",
       "      <td>-0.343602</td>\n",
       "      <td>-0.149045</td>\n",
       "      <td>-0.790001</td>\n",
       "      <td>-0.379234</td>\n",
       "      <td>1.570090</td>\n",
       "      <td>0.183310</td>\n",
       "      <td>0.223639</td>\n",
       "      <td>-0.051066</td>\n",
       "      <td>3.179090</td>\n",
       "      <td>-0.512679</td>\n",
       "      <td>-1.839577</td>\n",
       "      <td>0.383216</td>\n",
       "      <td>0.317492</td>\n",
       "      <td>-0.003006</td>\n",
       "      <td>-0.420161</td>\n",
       "      <td>0.456437</td>\n",
       "      <td>-0.789675</td>\n",
       "      <td>1.751367</td>\n",
       "      <td>-1.481236</td>\n",
       "      <td>0.757191</td>\n",
       "      <td>-0.912292</td>\n",
       "      <td>0.771948</td>\n",
       "      <td>0.503370</td>\n",
       "      <td>-1.345637</td>\n",
       "      <td>0.286381</td>\n",
       "      <td>0.668565</td>\n",
       "      <td>0.085125</td>\n",
       "      <td>-0.216750</td>\n",
       "      <td>-0.667408</td>\n",
       "      <td>-1.535679</td>\n",
       "      <td>0.783801</td>\n",
       "      <td>-1.418823</td>\n",
       "      <td>-0.098817</td>\n",
       "      <td>-0.929870</td>\n",
       "      <td>0.090828</td>\n",
       "      <td>-0.095242</td>\n",
       "      <td>-0.163216</td>\n",
       "      <td>0.700639</td>\n",
       "      <td>1.381549</td>\n",
       "      <td>-0.243736</td>\n",
       "      <td>-0.105134</td>\n",
       "      <td>-0.176474</td>\n",
       "      <td>0.862919</td>\n",
       "      <td>1.685211</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.603672</td>\n",
       "      <td>0.174326</td>\n",
       "      <td>-0.233102</td>\n",
       "      <td>-0.452149</td>\n",
       "      <td>0.927693</td>\n",
       "      <td>0.320566</td>\n",
       "      <td>0.737518</td>\n",
       "      <td>0.138547</td>\n",
       "      <td>1.013321</td>\n",
       "      <td>-0.503070</td>\n",
       "      <td>0.170520</td>\n",
       "      <td>-0.236994</td>\n",
       "      <td>0.403923</td>\n",
       "      <td>0.352233</td>\n",
       "      <td>-0.010432</td>\n",
       "      <td>-0.446412</td>\n",
       "      <td>-0.494003</td>\n",
       "      <td>0.712903</td>\n",
       "      <td>0.281747</td>\n",
       "      <td>-0.043063</td>\n",
       "      <td>0.183140</td>\n",
       "      <td>0.226364</td>\n",
       "      <td>0.131696</td>\n",
       "      <td>0.180204</td>\n",
       "      <td>0.144474</td>\n",
       "      <td>0.145077</td>\n",
       "      <td>0.045493</td>\n",
       "      <td>-0.119915</td>\n",
       "      <td>0.117218</td>\n",
       "      <td>0.620796</td>\n",
       "      <td>-0.186910</td>\n",
       "      <td>-0.194374</td>\n",
       "      <td>0.630890</td>\n",
       "      <td>0.425284</td>\n",
       "      <td>0.027759</td>\n",
       "      <td>-0.493276</td>\n",
       "      <td>0.415432</td>\n",
       "      <td>-0.243451</td>\n",
       "      <td>0.811091</td>\n",
       "      <td>0.186713</td>\n",
       "      <td>0.097599</td>\n",
       "      <td>0.228464</td>\n",
       "      <td>-0.240718</td>\n",
       "      <td>0.009613</td>\n",
       "      <td>-0.257763</td>\n",
       "      <td>0.246641</td>\n",
       "      <td>0.406052</td>\n",
       "      <td>-0.316100</td>\n",
       "      <td>-0.559927</td>\n",
       "      <td>-0.311160</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12350</th>\n",
       "      <td>-4.929320</td>\n",
       "      <td>-0.714859</td>\n",
       "      <td>-1.580930</td>\n",
       "      <td>-0.469767</td>\n",
       "      <td>-0.116290</td>\n",
       "      <td>0.647835</td>\n",
       "      <td>0.112904</td>\n",
       "      <td>-1.336313</td>\n",
       "      <td>1.547069</td>\n",
       "      <td>-1.190409</td>\n",
       "      <td>-0.213170</td>\n",
       "      <td>1.057289</td>\n",
       "      <td>-0.475002</td>\n",
       "      <td>0.856574</td>\n",
       "      <td>-1.305371</td>\n",
       "      <td>0.242700</td>\n",
       "      <td>0.268312</td>\n",
       "      <td>1.952558</td>\n",
       "      <td>-0.227722</td>\n",
       "      <td>0.569825</td>\n",
       "      <td>-0.186512</td>\n",
       "      <td>0.903530</td>\n",
       "      <td>1.364748</td>\n",
       "      <td>0.178125</td>\n",
       "      <td>-0.260398</td>\n",
       "      <td>-0.168825</td>\n",
       "      <td>-0.681952</td>\n",
       "      <td>1.081305</td>\n",
       "      <td>-0.155239</td>\n",
       "      <td>-0.397164</td>\n",
       "      <td>0.240051</td>\n",
       "      <td>0.634068</td>\n",
       "      <td>-1.447081</td>\n",
       "      <td>-0.186097</td>\n",
       "      <td>-0.315379</td>\n",
       "      <td>-0.306429</td>\n",
       "      <td>-1.359020</td>\n",
       "      <td>-0.001473</td>\n",
       "      <td>0.579056</td>\n",
       "      <td>-0.248313</td>\n",
       "      <td>0.588129</td>\n",
       "      <td>-0.217986</td>\n",
       "      <td>0.284601</td>\n",
       "      <td>0.525266</td>\n",
       "      <td>-0.531672</td>\n",
       "      <td>-0.101395</td>\n",
       "      <td>-1.472187</td>\n",
       "      <td>0.130116</td>\n",
       "      <td>0.140861</td>\n",
       "      <td>-0.668311</td>\n",
       "      <td>...</td>\n",
       "      <td>1.556940</td>\n",
       "      <td>2.420670</td>\n",
       "      <td>0.865762</td>\n",
       "      <td>0.453298</td>\n",
       "      <td>-0.993826</td>\n",
       "      <td>-1.640322</td>\n",
       "      <td>1.547916</td>\n",
       "      <td>1.346862</td>\n",
       "      <td>-0.364600</td>\n",
       "      <td>-1.070965</td>\n",
       "      <td>-0.477111</td>\n",
       "      <td>0.513733</td>\n",
       "      <td>1.054107</td>\n",
       "      <td>-0.190973</td>\n",
       "      <td>-0.051314</td>\n",
       "      <td>-0.076611</td>\n",
       "      <td>-0.034641</td>\n",
       "      <td>0.851705</td>\n",
       "      <td>-0.620095</td>\n",
       "      <td>0.348728</td>\n",
       "      <td>0.521268</td>\n",
       "      <td>1.447778</td>\n",
       "      <td>0.489761</td>\n",
       "      <td>1.487142</td>\n",
       "      <td>-0.401093</td>\n",
       "      <td>-0.761589</td>\n",
       "      <td>-0.079544</td>\n",
       "      <td>0.569086</td>\n",
       "      <td>-0.275222</td>\n",
       "      <td>-0.320405</td>\n",
       "      <td>0.523359</td>\n",
       "      <td>0.833541</td>\n",
       "      <td>0.124539</td>\n",
       "      <td>-0.004853</td>\n",
       "      <td>0.350597</td>\n",
       "      <td>-0.079295</td>\n",
       "      <td>1.062622</td>\n",
       "      <td>0.150736</td>\n",
       "      <td>-0.752001</td>\n",
       "      <td>0.890077</td>\n",
       "      <td>0.892657</td>\n",
       "      <td>1.049206</td>\n",
       "      <td>0.155618</td>\n",
       "      <td>0.172248</td>\n",
       "      <td>-0.171932</td>\n",
       "      <td>0.179725</td>\n",
       "      <td>-0.358419</td>\n",
       "      <td>-2.150534</td>\n",
       "      <td>-1.023392</td>\n",
       "      <td>-1.995684</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12352</th>\n",
       "      <td>-2.053113</td>\n",
       "      <td>2.110950</td>\n",
       "      <td>0.184807</td>\n",
       "      <td>0.128240</td>\n",
       "      <td>0.087787</td>\n",
       "      <td>-1.943579</td>\n",
       "      <td>2.383307</td>\n",
       "      <td>0.899610</td>\n",
       "      <td>-0.671980</td>\n",
       "      <td>0.081452</td>\n",
       "      <td>0.876840</td>\n",
       "      <td>1.051124</td>\n",
       "      <td>-0.741349</td>\n",
       "      <td>-2.071853</td>\n",
       "      <td>-0.637244</td>\n",
       "      <td>0.019206</td>\n",
       "      <td>0.114969</td>\n",
       "      <td>0.573977</td>\n",
       "      <td>-0.030594</td>\n",
       "      <td>0.105880</td>\n",
       "      <td>-0.693458</td>\n",
       "      <td>-0.911360</td>\n",
       "      <td>-2.330069</td>\n",
       "      <td>0.517448</td>\n",
       "      <td>-1.757740</td>\n",
       "      <td>0.224012</td>\n",
       "      <td>-1.062572</td>\n",
       "      <td>-1.806766</td>\n",
       "      <td>-0.929000</td>\n",
       "      <td>2.098583</td>\n",
       "      <td>-0.155669</td>\n",
       "      <td>2.220819</td>\n",
       "      <td>2.238957</td>\n",
       "      <td>-0.573106</td>\n",
       "      <td>2.444062</td>\n",
       "      <td>1.377984</td>\n",
       "      <td>0.187290</td>\n",
       "      <td>-0.400572</td>\n",
       "      <td>1.804934</td>\n",
       "      <td>1.208779</td>\n",
       "      <td>-0.960939</td>\n",
       "      <td>-0.863664</td>\n",
       "      <td>-1.128681</td>\n",
       "      <td>-0.058438</td>\n",
       "      <td>-0.884060</td>\n",
       "      <td>-2.207028</td>\n",
       "      <td>-1.590184</td>\n",
       "      <td>-1.454667</td>\n",
       "      <td>0.929104</td>\n",
       "      <td>0.282129</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.028970</td>\n",
       "      <td>0.244557</td>\n",
       "      <td>0.306695</td>\n",
       "      <td>-0.543993</td>\n",
       "      <td>-0.426855</td>\n",
       "      <td>0.544659</td>\n",
       "      <td>-0.509539</td>\n",
       "      <td>-0.745640</td>\n",
       "      <td>-0.149256</td>\n",
       "      <td>-0.031816</td>\n",
       "      <td>0.357092</td>\n",
       "      <td>-0.003698</td>\n",
       "      <td>0.502374</td>\n",
       "      <td>-0.074565</td>\n",
       "      <td>0.383567</td>\n",
       "      <td>-0.014870</td>\n",
       "      <td>-0.105848</td>\n",
       "      <td>0.316509</td>\n",
       "      <td>-0.137248</td>\n",
       "      <td>-0.025613</td>\n",
       "      <td>0.196455</td>\n",
       "      <td>-0.259308</td>\n",
       "      <td>-0.078183</td>\n",
       "      <td>0.202068</td>\n",
       "      <td>-0.164799</td>\n",
       "      <td>-0.010466</td>\n",
       "      <td>0.258529</td>\n",
       "      <td>0.329032</td>\n",
       "      <td>0.275869</td>\n",
       "      <td>0.388376</td>\n",
       "      <td>0.294109</td>\n",
       "      <td>0.129528</td>\n",
       "      <td>0.022517</td>\n",
       "      <td>-0.442529</td>\n",
       "      <td>-0.298672</td>\n",
       "      <td>-0.136678</td>\n",
       "      <td>0.186785</td>\n",
       "      <td>-0.233479</td>\n",
       "      <td>0.094408</td>\n",
       "      <td>-0.005697</td>\n",
       "      <td>-0.224264</td>\n",
       "      <td>0.002607</td>\n",
       "      <td>-0.179418</td>\n",
       "      <td>-0.396663</td>\n",
       "      <td>0.141766</td>\n",
       "      <td>-0.075031</td>\n",
       "      <td>0.038654</td>\n",
       "      <td>-0.078615</td>\n",
       "      <td>0.288926</td>\n",
       "      <td>-0.113879</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 300 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                 PC1       PC2       PC3       PC4       PC5        PC6  \\\n",
       "CustomerID                                                                \n",
       "12347       5.876141  4.828917 -0.495545  3.093849 -0.633706 -11.534092   \n",
       "12348      -4.219837  1.226217 -1.881912 -0.174340  0.817996   1.160834   \n",
       "12349      -1.260847  2.387664 -0.073148 -1.119161 -0.034339  -3.715390   \n",
       "12350      -4.929320 -0.714859 -1.580930 -0.469767 -0.116290   0.647835   \n",
       "12352      -2.053113  2.110950  0.184807  0.128240  0.087787  -1.943579   \n",
       "\n",
       "                 PC7       PC8       PC9      PC10      PC11      PC12  \\\n",
       "CustomerID                                                               \n",
       "12347       2.161996  7.116957  5.336009  4.424414  4.637002  2.316503   \n",
       "12348      -0.871683 -0.846634  0.460240  0.854274 -0.401793  0.981803   \n",
       "12349      -0.343602 -0.149045 -0.790001 -0.379234  1.570090  0.183310   \n",
       "12350       0.112904 -1.336313  1.547069 -1.190409 -0.213170  1.057289   \n",
       "12352       2.383307  0.899610 -0.671980  0.081452  0.876840  1.051124   \n",
       "\n",
       "                 PC13      PC14       PC15      PC16      PC17      PC18  \\\n",
       "CustomerID                                                                 \n",
       "12347       12.310783  1.986696 -11.431269  5.617791 -2.177268  7.665832   \n",
       "12348       -1.104237  0.770190  -0.092975 -0.313393 -0.448633  1.207302   \n",
       "12349        0.223639 -0.051066   3.179090 -0.512679 -1.839577  0.383216   \n",
       "12350       -0.475002  0.856574  -1.305371  0.242700  0.268312  1.952558   \n",
       "12352       -0.741349 -2.071853  -0.637244  0.019206  0.114969  0.573977   \n",
       "\n",
       "                 PC19       PC20      PC21      PC22       PC23       PC24  \\\n",
       "CustomerID                                                                   \n",
       "12347      -12.135817 -16.049000 -5.949238  7.814288  40.518194 -35.575039   \n",
       "12348        0.381678  -0.469152  0.185789  0.054585   0.499987   0.373514   \n",
       "12349        0.317492  -0.003006 -0.420161  0.456437  -0.789675   1.751367   \n",
       "12350       -0.227722   0.569825 -0.186512  0.903530   1.364748   0.178125   \n",
       "12352       -0.030594   0.105880 -0.693458 -0.911360  -2.330069   0.517448   \n",
       "\n",
       "                 PC25      PC26       PC27      PC28       PC29       PC30  \\\n",
       "CustomerID                                                                   \n",
       "12347      -17.557418  6.869259  20.886868  2.296657 -26.160394 -32.580135   \n",
       "12348        1.226615 -0.773072  -0.270125  0.850016  -0.370481  -0.573939   \n",
       "12349       -1.481236  0.757191  -0.912292  0.771948   0.503370  -1.345637   \n",
       "12350       -0.260398 -0.168825  -0.681952  1.081305  -0.155239  -0.397164   \n",
       "12352       -1.757740  0.224012  -1.062572 -1.806766  -0.929000   2.098583   \n",
       "\n",
       "                PC31       PC32      PC33       PC34      PC35       PC36  \\\n",
       "CustomerID                                                                  \n",
       "12347       7.880285  12.358371  6.826718 -32.957865 -4.405057 -17.238628   \n",
       "12348       1.012802  -0.580828  0.949802  -0.733970  0.353890  -0.387479   \n",
       "12349       0.286381   0.668565  0.085125  -0.216750 -0.667408  -1.535679   \n",
       "12350       0.240051   0.634068 -1.447081  -0.186097 -0.315379  -0.306429   \n",
       "12352      -0.155669   2.220819  2.238957  -0.573106  2.444062   1.377984   \n",
       "\n",
       "                 PC37       PC38      PC39      PC40      PC41      PC42  \\\n",
       "CustomerID                                                                 \n",
       "12347      -11.272274 -14.487767 -2.766330  4.736717 -6.985884 -1.464059   \n",
       "12348       -0.298541   0.396418  0.154924  0.333479 -0.466261  0.703454   \n",
       "12349        0.783801  -1.418823 -0.098817 -0.929870  0.090828 -0.095242   \n",
       "12350       -1.359020  -0.001473  0.579056 -0.248313  0.588129 -0.217986   \n",
       "12352        0.187290  -0.400572  1.804934  1.208779 -0.960939 -0.863664   \n",
       "\n",
       "                 PC43      PC44      PC45      PC46      PC47      PC48  \\\n",
       "CustomerID                                                                \n",
       "12347       10.603071  1.239909  3.086324  5.336237 -0.410776  1.616712   \n",
       "12348        1.200580 -0.202163 -0.871794 -0.055069 -0.421733 -1.623011   \n",
       "12349       -0.163216  0.700639  1.381549 -0.243736 -0.105134 -0.176474   \n",
       "12350        0.284601  0.525266 -0.531672 -0.101395 -1.472187  0.130116   \n",
       "12352       -1.128681 -0.058438 -0.884060 -2.207028 -1.590184 -1.454667   \n",
       "\n",
       "                PC49      PC50    ...        PC251     PC252     PC253  \\\n",
       "CustomerID                        ...                                    \n",
       "12347      -6.383660  3.065176    ...     0.064395 -0.415426 -0.206658   \n",
       "12348      -0.397691  0.452590    ...     0.129957  1.022734 -0.456650   \n",
       "12349       0.862919  1.685211    ...    -0.603672  0.174326 -0.233102   \n",
       "12350       0.140861 -0.668311    ...     1.556940  2.420670  0.865762   \n",
       "12352       0.929104  0.282129    ...    -0.028970  0.244557  0.306695   \n",
       "\n",
       "               PC254     PC255     PC256     PC257     PC258     PC259  \\\n",
       "CustomerID                                                               \n",
       "12347      -0.005371  0.188138  0.242410 -0.195482 -0.162702 -0.200203   \n",
       "12348      -1.224744  0.530329 -0.994510 -0.371455 -0.468078  0.489332   \n",
       "12349      -0.452149  0.927693  0.320566  0.737518  0.138547  1.013321   \n",
       "12350       0.453298 -0.993826 -1.640322  1.547916  1.346862 -0.364600   \n",
       "12352      -0.543993 -0.426855  0.544659 -0.509539 -0.745640 -0.149256   \n",
       "\n",
       "               PC260     PC261     PC262     PC263     PC264     PC265  \\\n",
       "CustomerID                                                               \n",
       "12347      -0.170627  0.118951 -0.124982 -0.092571 -0.209632 -0.122009   \n",
       "12348      -0.620042  0.995225 -1.389717  1.171395 -0.617378  0.078084   \n",
       "12349      -0.503070  0.170520 -0.236994  0.403923  0.352233 -0.010432   \n",
       "12350      -1.070965 -0.477111  0.513733  1.054107 -0.190973 -0.051314   \n",
       "12352      -0.031816  0.357092 -0.003698  0.502374 -0.074565  0.383567   \n",
       "\n",
       "               PC266     PC267     PC268     PC269     PC270     PC271  \\\n",
       "CustomerID                                                               \n",
       "12347       0.112249  0.097985  0.047909  0.191655 -0.245912 -0.183734   \n",
       "12348       1.336674 -0.120785  0.330687 -0.235496  0.112896  0.654065   \n",
       "12349      -0.446412 -0.494003  0.712903  0.281747 -0.043063  0.183140   \n",
       "12350      -0.076611 -0.034641  0.851705 -0.620095  0.348728  0.521268   \n",
       "12352      -0.014870 -0.105848  0.316509 -0.137248 -0.025613  0.196455   \n",
       "\n",
       "               PC272     PC273     PC274     PC275     PC276     PC277  \\\n",
       "CustomerID                                                               \n",
       "12347      -0.034925  0.108376 -0.060922 -0.060365 -0.033723  0.029060   \n",
       "12348       0.388106 -0.654803  0.410856 -0.836039  1.419597  0.055308   \n",
       "12349       0.226364  0.131696  0.180204  0.144474  0.145077  0.045493   \n",
       "12350       1.447778  0.489761  1.487142 -0.401093 -0.761589 -0.079544   \n",
       "12352      -0.259308 -0.078183  0.202068 -0.164799 -0.010466  0.258529   \n",
       "\n",
       "               PC278     PC279     PC280     PC281     PC282     PC283  \\\n",
       "CustomerID                                                               \n",
       "12347      -0.006391 -0.001222  0.040427  0.089004 -0.218621  0.158177   \n",
       "12348       0.657981  0.559558  0.225912 -0.892273 -0.111400 -0.459792   \n",
       "12349      -0.119915  0.117218  0.620796 -0.186910 -0.194374  0.630890   \n",
       "12350       0.569086 -0.275222 -0.320405  0.523359  0.833541  0.124539   \n",
       "12352       0.329032  0.275869  0.388376  0.294109  0.129528  0.022517   \n",
       "\n",
       "               PC284     PC285     PC286     PC287     PC288     PC289  \\\n",
       "CustomerID                                                               \n",
       "12347       0.049166 -0.037660 -0.024126  0.094181 -0.109886 -0.035857   \n",
       "12348       0.133336 -0.166611  0.015979 -1.039514 -0.153822 -0.368017   \n",
       "12349       0.425284  0.027759 -0.493276  0.415432 -0.243451  0.811091   \n",
       "12350      -0.004853  0.350597 -0.079295  1.062622  0.150736 -0.752001   \n",
       "12352      -0.442529 -0.298672 -0.136678  0.186785 -0.233479  0.094408   \n",
       "\n",
       "               PC290     PC291     PC292     PC293     PC294     PC295  \\\n",
       "CustomerID                                                               \n",
       "12347       0.015897 -0.071372 -0.046723  0.109823  0.021534 -0.073246   \n",
       "12348      -0.717055 -0.187247 -0.006443  0.247051  0.526344  0.048497   \n",
       "12349       0.186713  0.097599  0.228464 -0.240718  0.009613 -0.257763   \n",
       "12350       0.890077  0.892657  1.049206  0.155618  0.172248 -0.171932   \n",
       "12352      -0.005697 -0.224264  0.002607 -0.179418 -0.396663  0.141766   \n",
       "\n",
       "               PC296     PC297     PC298     PC299     PC300  \n",
       "CustomerID                                                    \n",
       "12347      -0.005525 -0.046094  0.134262 -0.046511  0.068503  \n",
       "12348       0.309830 -0.974179 -0.156291  0.451286  1.135782  \n",
       "12349       0.246641  0.406052 -0.316100 -0.559927 -0.311160  \n",
       "12350       0.179725 -0.358419 -2.150534 -1.023392 -1.995684  \n",
       "12352      -0.075031  0.038654 -0.078615  0.288926 -0.113879  \n",
       "\n",
       "[5 rows x 300 columns]"
      ]
     },
     "execution_count": 163,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Put PC_items into a dataframe\n",
    "items_pca = pd.DataFrame(PC_items)\n",
    "\n",
    "# Name the columns\n",
    "items_pca.columns = ['PC{}'.format(i + 1) for i in range(PC_items.shape[1])]\n",
    "\n",
    "# Update its index\n",
    "items_pca.index = item_data.index\n",
    "\n",
    "# Display first 5 rows\n",
    "items_pca.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* Now we have a dataframe of 300 customer-level principal component features.\n",
    "* These were generated from the 300 principal components that explained the most variance for the original features.\n",
    "* The index of this PCA item dataframe contains CustomerID's, which will make it possible for us to join this to our analytical base table.\n",
    "\n",
    "Finally, save this item dataframe with PCA features as 'pca_item_data.csv'.\n",
    "* Next, we'll compare the clusters made from using these features against those in 'threshold_item_data.csv'.\n",
    "* Do not set index=None because we want to keep the CustomerID's as the index."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 164,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Save pca_item_data.csv\n",
    "items_pca.to_csv('Files/pca_item_data.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 9. KMeans Clustering\n",
    "Let's import 3 CSV files we've saved throughout this project.\n",
    "* Let's import 'analytical_base_table.csv' as base_df.\n",
    "* Let's import 'threshold_item_data.csv' as threshold_item_data.\n",
    "* Let's import 'pca_item_data.csv' as pca_item_data.\n",
    "* Set index_col=0 for each one to use CustomerID as the index."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 190,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Import analytical base table\n",
    "base_df = pd.read_csv('Files/analytical_base_table.csv', index_col=0)\n",
    "\n",
    "# Import thresholded item features\n",
    "threshold_item_data = pd.read_csv('Files/threshold_item_data.csv', index_col=0)\n",
    "\n",
    "# Import PCA item features\n",
    "pca_item_data = pd.read_csv('Files/pca_item_data.csv', index_col=0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Print the shape of each one to make sure we're on the same page."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 191,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(414, 8)\n",
      "(414, 120)\n",
      "(414, 300)\n"
     ]
    }
   ],
   "source": [
    "# Print shape of each dataframe\n",
    "print( base_df.shape )\n",
    "print( threshold_item_data.shape )\n",
    "print( pca_item_data.shape )"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Because K-Means creates clusters based on distances, and because distances are calculated by between observations defined by their feature values, **the features you choose to input into the algorithm heavily influence the clusters that are created.**\n",
    "\n",
    "For this project, we will look at 3 possible feature sets and compare the clusters created from them. We'll try:\n",
    "1. Only purchase pattern features (\"Base DF\")\n",
    "2. Purchase pattern features + item features chosen by thresholding (\"Threshold DF\")\n",
    "3. Purchase pattern features + principal component features from items (\"PCA DF\")\n",
    "\n",
    "Create a threshold_df by joining base_df with threshold_item_data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 192,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>total_transactions</th>\n",
       "      <th>total_products</th>\n",
       "      <th>total_unique_products</th>\n",
       "      <th>total_sales</th>\n",
       "      <th>avg_product_value</th>\n",
       "      <th>min_cart_value</th>\n",
       "      <th>max_cart_value</th>\n",
       "      <th>avg_cart_value</th>\n",
       "      <th>22748</th>\n",
       "      <th>22847</th>\n",
       "      <th>15056BL</th>\n",
       "      <th>23159</th>\n",
       "      <th>85014B</th>\n",
       "      <th>23243</th>\n",
       "      <th>23231</th>\n",
       "      <th>22898</th>\n",
       "      <th>23388</th>\n",
       "      <th>23255</th>\n",
       "      <th>22966</th>\n",
       "      <th>22558</th>\n",
       "      <th>21558</th>\n",
       "      <th>22356</th>\n",
       "      <th>84997C</th>\n",
       "      <th>23292</th>\n",
       "      <th>22699</th>\n",
       "      <th>22331</th>\n",
       "      <th>22385</th>\n",
       "      <th>23480</th>\n",
       "      <th>84375</th>\n",
       "      <th>21156</th>\n",
       "      <th>21242</th>\n",
       "      <th>22417</th>\n",
       "      <th>21122</th>\n",
       "      <th>84997B</th>\n",
       "      <th>22895</th>\n",
       "      <th>22962</th>\n",
       "      <th>22557</th>\n",
       "      <th>22907</th>\n",
       "      <th>23307</th>\n",
       "      <th>20676</th>\n",
       "      <th>21977</th>\n",
       "      <th>21976</th>\n",
       "      <th>23293</th>\n",
       "      <th>22661</th>\n",
       "      <th>23290</th>\n",
       "      <th>22627</th>\n",
       "      <th>84879</th>\n",
       "      <th>20682</th>\n",
       "      <th>22367</th>\n",
       "      <th>23236</th>\n",
       "      <th>...</th>\n",
       "      <th>22631</th>\n",
       "      <th>22090</th>\n",
       "      <th>23199</th>\n",
       "      <th>21915</th>\n",
       "      <th>23203</th>\n",
       "      <th>84997D</th>\n",
       "      <th>22029</th>\n",
       "      <th>23209</th>\n",
       "      <th>23254</th>\n",
       "      <th>22077</th>\n",
       "      <th>22960</th>\n",
       "      <th>23256</th>\n",
       "      <th>22492</th>\n",
       "      <th>22467</th>\n",
       "      <th>22726</th>\n",
       "      <th>22382</th>\n",
       "      <th>23240</th>\n",
       "      <th>23245</th>\n",
       "      <th>21094</th>\n",
       "      <th>21086</th>\n",
       "      <th>23206</th>\n",
       "      <th>22352</th>\n",
       "      <th>22727</th>\n",
       "      <th>22138</th>\n",
       "      <th>22720</th>\n",
       "      <th>21559</th>\n",
       "      <th>22728</th>\n",
       "      <th>22555</th>\n",
       "      <th>20724</th>\n",
       "      <th>20712</th>\n",
       "      <th>22961</th>\n",
       "      <th>22630</th>\n",
       "      <th>22139</th>\n",
       "      <th>21080</th>\n",
       "      <th>85099B</th>\n",
       "      <th>20726</th>\n",
       "      <th>20719</th>\n",
       "      <th>20750</th>\n",
       "      <th>23084</th>\n",
       "      <th>20725</th>\n",
       "      <th>21212</th>\n",
       "      <th>22551</th>\n",
       "      <th>22629</th>\n",
       "      <th>22328</th>\n",
       "      <th>21731</th>\n",
       "      <th>22556</th>\n",
       "      <th>22554</th>\n",
       "      <th>22423</th>\n",
       "      <th>22326</th>\n",
       "      <th>POST</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CustomerID</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>12347</th>\n",
       "      <td>7</td>\n",
       "      <td>182</td>\n",
       "      <td>103</td>\n",
       "      <td>4310.00</td>\n",
       "      <td>23.681319</td>\n",
       "      <td>224.82</td>\n",
       "      <td>1294.32</td>\n",
       "      <td>615.714286</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12348</th>\n",
       "      <td>4</td>\n",
       "      <td>31</td>\n",
       "      <td>22</td>\n",
       "      <td>1797.24</td>\n",
       "      <td>57.975484</td>\n",
       "      <td>227.44</td>\n",
       "      <td>892.80</td>\n",
       "      <td>449.310000</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12349</th>\n",
       "      <td>1</td>\n",
       "      <td>73</td>\n",
       "      <td>73</td>\n",
       "      <td>1757.55</td>\n",
       "      <td>24.076027</td>\n",
       "      <td>1757.55</td>\n",
       "      <td>1757.55</td>\n",
       "      <td>1757.550000</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12350</th>\n",
       "      <td>1</td>\n",
       "      <td>17</td>\n",
       "      <td>17</td>\n",
       "      <td>334.40</td>\n",
       "      <td>19.670588</td>\n",
       "      <td>334.40</td>\n",
       "      <td>334.40</td>\n",
       "      <td>334.400000</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12352</th>\n",
       "      <td>8</td>\n",
       "      <td>85</td>\n",
       "      <td>59</td>\n",
       "      <td>2506.04</td>\n",
       "      <td>29.482824</td>\n",
       "      <td>120.33</td>\n",
       "      <td>840.30</td>\n",
       "      <td>313.255000</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 128 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            total_transactions  total_products  total_unique_products  \\\n",
       "CustomerID                                                              \n",
       "12347                        7             182                    103   \n",
       "12348                        4              31                     22   \n",
       "12349                        1              73                     73   \n",
       "12350                        1              17                     17   \n",
       "12352                        8              85                     59   \n",
       "\n",
       "            total_sales  avg_product_value  min_cart_value  max_cart_value  \\\n",
       "CustomerID                                                                   \n",
       "12347           4310.00          23.681319          224.82         1294.32   \n",
       "12348           1797.24          57.975484          227.44          892.80   \n",
       "12349           1757.55          24.076027         1757.55         1757.55   \n",
       "12350            334.40          19.670588          334.40          334.40   \n",
       "12352           2506.04          29.482824          120.33          840.30   \n",
       "\n",
       "            avg_cart_value  22748  22847  15056BL  23159  85014B  23243  \\\n",
       "CustomerID                                                                \n",
       "12347           615.714286      0      0        0      0       0      0   \n",
       "12348           449.310000      0      0        0      0       0      0   \n",
       "12349          1757.550000      0      0        0      0       1      0   \n",
       "12350           334.400000      0      0        0      0       0      0   \n",
       "12352           313.255000      0      0        0      0       0      1   \n",
       "\n",
       "            23231  22898  23388  23255  22966  22558  21558  22356  84997C  \\\n",
       "CustomerID                                                                   \n",
       "12347           0      0      0      0      0      0      0      0       1   \n",
       "12348           0      0      0      0      0      0      0      0       0   \n",
       "12349           0      0      0      0      0      0      0      0       0   \n",
       "12350           0      0      0      0      0      0      0      0       0   \n",
       "12352           0      0      0      0      0      0      0      0       0   \n",
       "\n",
       "            23292  22699  22331  22385  23480  84375  21156  21242  22417  \\\n",
       "CustomerID                                                                  \n",
       "12347           0      2      0      0      1      0      0      0      2   \n",
       "12348           0      0      0      0      0      0      0      0      0   \n",
       "12349           0      0      0      0      0      0      0      0      0   \n",
       "12350           0      0      0      0      0      0      0      0      0   \n",
       "12352           0      0      0      0      0      0      0      0      0   \n",
       "\n",
       "            21122  84997B  22895  22962  22557  22907  23307  20676  21977  \\\n",
       "CustomerID                                                                   \n",
       "12347           0       1      0      0      0      0      0      0      0   \n",
       "12348           0       0      0      0      0      0      0      0      2   \n",
       "12349           0       0      0      0      1      0      0      0      0   \n",
       "12350           0       0      0      0      1      0      0      0      0   \n",
       "12352           0       0      0      0      0      0      0      0      0   \n",
       "\n",
       "            21976  23293  22661  23290  22627  84879  20682  22367  23236  \\\n",
       "CustomerID                                                                  \n",
       "12347           2      0      0      0      0      0      0      0      0   \n",
       "12348           0      0      0      0      0      0      0      0      0   \n",
       "12349           0      1      0      0      0      0      0      0      1   \n",
       "12350           0      0      0      0      0      0      0      0      0   \n",
       "12352           0      0      0      0      1      0      0      0      0   \n",
       "\n",
       "            ...   22631  22090  23199  21915  23203  84997D  22029  23209  \\\n",
       "CustomerID  ...                                                             \n",
       "12347       ...       0      0      0      0      0       1      0      0   \n",
       "12348       ...       0      0      0      0      0       0      0      0   \n",
       "12349       ...       0      0      0      0      0       0      0      0   \n",
       "12350       ...       0      0      0      1      0       0      0      0   \n",
       "12352       ...       0      0      0      0      0       0      0      0   \n",
       "\n",
       "            23254  22077  22960  23256  22492  22467  22726  22382  23240  \\\n",
       "CustomerID                                                                  \n",
       "12347           0      0      0      0      3      0      3      0      0   \n",
       "12348           0      0      0      0      0      0      0      0      0   \n",
       "12349           0      0      1      0      0      0      0      0      1   \n",
       "12350           0      0      0      0      0      0      0      0      0   \n",
       "12352           0      0      0      0      0      0      0      0      0   \n",
       "\n",
       "            23245  21094  21086  23206  22352  22727  22138  22720  21559  \\\n",
       "CustomerID                                                                  \n",
       "12347           0      0      0      0      0      5      0      0      0   \n",
       "12348           0      0      0      0      0      0      0      0      0   \n",
       "12349           0      0      1      0      0      0      0      1      0   \n",
       "12350           0      0      0      0      0      0      0      0      0   \n",
       "12352           2      0      0      0      0      0      1      2      0   \n",
       "\n",
       "            22728  22555  20724  20712  22961  22630  22139  21080  85099B  \\\n",
       "CustomerID                                                                   \n",
       "12347           3      0      0      0      0      0      0      0       0   \n",
       "12348           0      0      0      0      0      0      0      0       0   \n",
       "12349           0      1      0      0      0      0      0      0       0   \n",
       "12350           0      0      0      0      0      0      0      0       0   \n",
       "12352           0      0      0      0      0      1      0      0       0   \n",
       "\n",
       "            20726  20719  20750  23084  20725  21212  22551  22629  22328  \\\n",
       "CustomerID                                                                  \n",
       "12347           0      4      0      3      0      0      0      0      0   \n",
       "12348           0      0      0      0      0      0      0      0      0   \n",
       "12349           0      0      0      0      0      0      0      0      0   \n",
       "12350           0      0      0      0      0      0      1      0      0   \n",
       "12352           0      0      0      0      0      0      0      0      0   \n",
       "\n",
       "            21731  22556  22554  22423  22326  POST  \n",
       "CustomerID                                           \n",
       "12347           5      0      0      4      0     0  \n",
       "12348           0      0      0      0      0     4  \n",
       "12349           0      1      1      1      1     1  \n",
       "12350           0      0      0      0      0     1  \n",
       "12352           1      0      0      2      0     5  \n",
       "\n",
       "[5 rows x 128 columns]"
      ]
     },
     "execution_count": 192,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Join base_df with threshold_item_data\n",
    "threshold_df = base_df.join(threshold_item_data)\n",
    "\n",
    "# Display first 5 rows of threshold_df\n",
    "threshold_df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Create a pca_df by joining base_df with pca_item_data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 193,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>total_transactions</th>\n",
       "      <th>total_products</th>\n",
       "      <th>total_unique_products</th>\n",
       "      <th>total_sales</th>\n",
       "      <th>avg_product_value</th>\n",
       "      <th>min_cart_value</th>\n",
       "      <th>max_cart_value</th>\n",
       "      <th>avg_cart_value</th>\n",
       "      <th>PC1</th>\n",
       "      <th>PC2</th>\n",
       "      <th>PC3</th>\n",
       "      <th>PC4</th>\n",
       "      <th>PC5</th>\n",
       "      <th>PC6</th>\n",
       "      <th>PC7</th>\n",
       "      <th>PC8</th>\n",
       "      <th>PC9</th>\n",
       "      <th>PC10</th>\n",
       "      <th>PC11</th>\n",
       "      <th>PC12</th>\n",
       "      <th>PC13</th>\n",
       "      <th>PC14</th>\n",
       "      <th>PC15</th>\n",
       "      <th>PC16</th>\n",
       "      <th>PC17</th>\n",
       "      <th>PC18</th>\n",
       "      <th>PC19</th>\n",
       "      <th>PC20</th>\n",
       "      <th>PC21</th>\n",
       "      <th>PC22</th>\n",
       "      <th>PC23</th>\n",
       "      <th>PC24</th>\n",
       "      <th>PC25</th>\n",
       "      <th>PC26</th>\n",
       "      <th>PC27</th>\n",
       "      <th>PC28</th>\n",
       "      <th>PC29</th>\n",
       "      <th>PC30</th>\n",
       "      <th>PC31</th>\n",
       "      <th>PC32</th>\n",
       "      <th>PC33</th>\n",
       "      <th>PC34</th>\n",
       "      <th>PC35</th>\n",
       "      <th>PC36</th>\n",
       "      <th>PC37</th>\n",
       "      <th>PC38</th>\n",
       "      <th>PC39</th>\n",
       "      <th>PC40</th>\n",
       "      <th>PC41</th>\n",
       "      <th>PC42</th>\n",
       "      <th>...</th>\n",
       "      <th>PC251</th>\n",
       "      <th>PC252</th>\n",
       "      <th>PC253</th>\n",
       "      <th>PC254</th>\n",
       "      <th>PC255</th>\n",
       "      <th>PC256</th>\n",
       "      <th>PC257</th>\n",
       "      <th>PC258</th>\n",
       "      <th>PC259</th>\n",
       "      <th>PC260</th>\n",
       "      <th>PC261</th>\n",
       "      <th>PC262</th>\n",
       "      <th>PC263</th>\n",
       "      <th>PC264</th>\n",
       "      <th>PC265</th>\n",
       "      <th>PC266</th>\n",
       "      <th>PC267</th>\n",
       "      <th>PC268</th>\n",
       "      <th>PC269</th>\n",
       "      <th>PC270</th>\n",
       "      <th>PC271</th>\n",
       "      <th>PC272</th>\n",
       "      <th>PC273</th>\n",
       "      <th>PC274</th>\n",
       "      <th>PC275</th>\n",
       "      <th>PC276</th>\n",
       "      <th>PC277</th>\n",
       "      <th>PC278</th>\n",
       "      <th>PC279</th>\n",
       "      <th>PC280</th>\n",
       "      <th>PC281</th>\n",
       "      <th>PC282</th>\n",
       "      <th>PC283</th>\n",
       "      <th>PC284</th>\n",
       "      <th>PC285</th>\n",
       "      <th>PC286</th>\n",
       "      <th>PC287</th>\n",
       "      <th>PC288</th>\n",
       "      <th>PC289</th>\n",
       "      <th>PC290</th>\n",
       "      <th>PC291</th>\n",
       "      <th>PC292</th>\n",
       "      <th>PC293</th>\n",
       "      <th>PC294</th>\n",
       "      <th>PC295</th>\n",
       "      <th>PC296</th>\n",
       "      <th>PC297</th>\n",
       "      <th>PC298</th>\n",
       "      <th>PC299</th>\n",
       "      <th>PC300</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CustomerID</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>12347</th>\n",
       "      <td>7</td>\n",
       "      <td>182</td>\n",
       "      <td>103</td>\n",
       "      <td>4310.00</td>\n",
       "      <td>23.681319</td>\n",
       "      <td>224.82</td>\n",
       "      <td>1294.32</td>\n",
       "      <td>615.714286</td>\n",
       "      <td>5.876141</td>\n",
       "      <td>4.828917</td>\n",
       "      <td>-0.495545</td>\n",
       "      <td>3.093849</td>\n",
       "      <td>-0.633706</td>\n",
       "      <td>-11.534092</td>\n",
       "      <td>2.161996</td>\n",
       "      <td>7.116957</td>\n",
       "      <td>5.336009</td>\n",
       "      <td>4.424414</td>\n",
       "      <td>4.637002</td>\n",
       "      <td>2.316503</td>\n",
       "      <td>12.310783</td>\n",
       "      <td>1.986696</td>\n",
       "      <td>-11.431269</td>\n",
       "      <td>5.617791</td>\n",
       "      <td>-2.177268</td>\n",
       "      <td>7.665832</td>\n",
       "      <td>-12.135817</td>\n",
       "      <td>-16.049000</td>\n",
       "      <td>-5.949238</td>\n",
       "      <td>7.814288</td>\n",
       "      <td>40.518194</td>\n",
       "      <td>-35.575039</td>\n",
       "      <td>-17.557418</td>\n",
       "      <td>6.869259</td>\n",
       "      <td>20.886868</td>\n",
       "      <td>2.296657</td>\n",
       "      <td>-26.160394</td>\n",
       "      <td>-32.580135</td>\n",
       "      <td>7.880285</td>\n",
       "      <td>12.358371</td>\n",
       "      <td>6.826718</td>\n",
       "      <td>-32.957865</td>\n",
       "      <td>-4.405057</td>\n",
       "      <td>-17.238628</td>\n",
       "      <td>-11.272274</td>\n",
       "      <td>-14.487767</td>\n",
       "      <td>-2.766330</td>\n",
       "      <td>4.736717</td>\n",
       "      <td>-6.985884</td>\n",
       "      <td>-1.464059</td>\n",
       "      <td>...</td>\n",
       "      <td>0.064395</td>\n",
       "      <td>-0.415426</td>\n",
       "      <td>-0.206658</td>\n",
       "      <td>-0.005371</td>\n",
       "      <td>0.188138</td>\n",
       "      <td>0.242410</td>\n",
       "      <td>-0.195482</td>\n",
       "      <td>-0.162702</td>\n",
       "      <td>-0.200203</td>\n",
       "      <td>-0.170627</td>\n",
       "      <td>0.118951</td>\n",
       "      <td>-0.124982</td>\n",
       "      <td>-0.092571</td>\n",
       "      <td>-0.209632</td>\n",
       "      <td>-0.122009</td>\n",
       "      <td>0.112249</td>\n",
       "      <td>0.097985</td>\n",
       "      <td>0.047909</td>\n",
       "      <td>0.191655</td>\n",
       "      <td>-0.245912</td>\n",
       "      <td>-0.183734</td>\n",
       "      <td>-0.034925</td>\n",
       "      <td>0.108376</td>\n",
       "      <td>-0.060922</td>\n",
       "      <td>-0.060365</td>\n",
       "      <td>-0.033723</td>\n",
       "      <td>0.029060</td>\n",
       "      <td>-0.006391</td>\n",
       "      <td>-0.001222</td>\n",
       "      <td>0.040427</td>\n",
       "      <td>0.089004</td>\n",
       "      <td>-0.218621</td>\n",
       "      <td>0.158177</td>\n",
       "      <td>0.049166</td>\n",
       "      <td>-0.037660</td>\n",
       "      <td>-0.024126</td>\n",
       "      <td>0.094181</td>\n",
       "      <td>-0.109886</td>\n",
       "      <td>-0.035857</td>\n",
       "      <td>0.015897</td>\n",
       "      <td>-0.071372</td>\n",
       "      <td>-0.046723</td>\n",
       "      <td>0.109823</td>\n",
       "      <td>0.021534</td>\n",
       "      <td>-0.073246</td>\n",
       "      <td>-0.005525</td>\n",
       "      <td>-0.046094</td>\n",
       "      <td>0.134262</td>\n",
       "      <td>-0.046511</td>\n",
       "      <td>0.068503</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12348</th>\n",
       "      <td>4</td>\n",
       "      <td>31</td>\n",
       "      <td>22</td>\n",
       "      <td>1797.24</td>\n",
       "      <td>57.975484</td>\n",
       "      <td>227.44</td>\n",
       "      <td>892.80</td>\n",
       "      <td>449.310000</td>\n",
       "      <td>-4.219837</td>\n",
       "      <td>1.226217</td>\n",
       "      <td>-1.881912</td>\n",
       "      <td>-0.174340</td>\n",
       "      <td>0.817996</td>\n",
       "      <td>1.160834</td>\n",
       "      <td>-0.871683</td>\n",
       "      <td>-0.846634</td>\n",
       "      <td>0.460240</td>\n",
       "      <td>0.854274</td>\n",
       "      <td>-0.401793</td>\n",
       "      <td>0.981803</td>\n",
       "      <td>-1.104237</td>\n",
       "      <td>0.770190</td>\n",
       "      <td>-0.092975</td>\n",
       "      <td>-0.313393</td>\n",
       "      <td>-0.448633</td>\n",
       "      <td>1.207302</td>\n",
       "      <td>0.381678</td>\n",
       "      <td>-0.469152</td>\n",
       "      <td>0.185789</td>\n",
       "      <td>0.054585</td>\n",
       "      <td>0.499987</td>\n",
       "      <td>0.373514</td>\n",
       "      <td>1.226615</td>\n",
       "      <td>-0.773072</td>\n",
       "      <td>-0.270125</td>\n",
       "      <td>0.850016</td>\n",
       "      <td>-0.370481</td>\n",
       "      <td>-0.573939</td>\n",
       "      <td>1.012802</td>\n",
       "      <td>-0.580828</td>\n",
       "      <td>0.949802</td>\n",
       "      <td>-0.733970</td>\n",
       "      <td>0.353890</td>\n",
       "      <td>-0.387479</td>\n",
       "      <td>-0.298541</td>\n",
       "      <td>0.396418</td>\n",
       "      <td>0.154924</td>\n",
       "      <td>0.333479</td>\n",
       "      <td>-0.466261</td>\n",
       "      <td>0.703454</td>\n",
       "      <td>...</td>\n",
       "      <td>0.129957</td>\n",
       "      <td>1.022734</td>\n",
       "      <td>-0.456650</td>\n",
       "      <td>-1.224744</td>\n",
       "      <td>0.530329</td>\n",
       "      <td>-0.994510</td>\n",
       "      <td>-0.371455</td>\n",
       "      <td>-0.468078</td>\n",
       "      <td>0.489332</td>\n",
       "      <td>-0.620042</td>\n",
       "      <td>0.995225</td>\n",
       "      <td>-1.389717</td>\n",
       "      <td>1.171395</td>\n",
       "      <td>-0.617378</td>\n",
       "      <td>0.078084</td>\n",
       "      <td>1.336674</td>\n",
       "      <td>-0.120785</td>\n",
       "      <td>0.330687</td>\n",
       "      <td>-0.235496</td>\n",
       "      <td>0.112896</td>\n",
       "      <td>0.654065</td>\n",
       "      <td>0.388106</td>\n",
       "      <td>-0.654803</td>\n",
       "      <td>0.410856</td>\n",
       "      <td>-0.836039</td>\n",
       "      <td>1.419597</td>\n",
       "      <td>0.055308</td>\n",
       "      <td>0.657981</td>\n",
       "      <td>0.559558</td>\n",
       "      <td>0.225912</td>\n",
       "      <td>-0.892273</td>\n",
       "      <td>-0.111400</td>\n",
       "      <td>-0.459792</td>\n",
       "      <td>0.133336</td>\n",
       "      <td>-0.166611</td>\n",
       "      <td>0.015979</td>\n",
       "      <td>-1.039514</td>\n",
       "      <td>-0.153822</td>\n",
       "      <td>-0.368017</td>\n",
       "      <td>-0.717055</td>\n",
       "      <td>-0.187247</td>\n",
       "      <td>-0.006443</td>\n",
       "      <td>0.247051</td>\n",
       "      <td>0.526344</td>\n",
       "      <td>0.048497</td>\n",
       "      <td>0.309830</td>\n",
       "      <td>-0.974179</td>\n",
       "      <td>-0.156291</td>\n",
       "      <td>0.451286</td>\n",
       "      <td>1.135782</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12349</th>\n",
       "      <td>1</td>\n",
       "      <td>73</td>\n",
       "      <td>73</td>\n",
       "      <td>1757.55</td>\n",
       "      <td>24.076027</td>\n",
       "      <td>1757.55</td>\n",
       "      <td>1757.55</td>\n",
       "      <td>1757.550000</td>\n",
       "      <td>-1.260847</td>\n",
       "      <td>2.387664</td>\n",
       "      <td>-0.073148</td>\n",
       "      <td>-1.119161</td>\n",
       "      <td>-0.034339</td>\n",
       "      <td>-3.715390</td>\n",
       "      <td>-0.343602</td>\n",
       "      <td>-0.149045</td>\n",
       "      <td>-0.790001</td>\n",
       "      <td>-0.379234</td>\n",
       "      <td>1.570090</td>\n",
       "      <td>0.183310</td>\n",
       "      <td>0.223639</td>\n",
       "      <td>-0.051066</td>\n",
       "      <td>3.179090</td>\n",
       "      <td>-0.512679</td>\n",
       "      <td>-1.839577</td>\n",
       "      <td>0.383216</td>\n",
       "      <td>0.317492</td>\n",
       "      <td>-0.003006</td>\n",
       "      <td>-0.420161</td>\n",
       "      <td>0.456437</td>\n",
       "      <td>-0.789675</td>\n",
       "      <td>1.751367</td>\n",
       "      <td>-1.481236</td>\n",
       "      <td>0.757191</td>\n",
       "      <td>-0.912292</td>\n",
       "      <td>0.771948</td>\n",
       "      <td>0.503370</td>\n",
       "      <td>-1.345637</td>\n",
       "      <td>0.286381</td>\n",
       "      <td>0.668565</td>\n",
       "      <td>0.085125</td>\n",
       "      <td>-0.216750</td>\n",
       "      <td>-0.667408</td>\n",
       "      <td>-1.535679</td>\n",
       "      <td>0.783801</td>\n",
       "      <td>-1.418823</td>\n",
       "      <td>-0.098817</td>\n",
       "      <td>-0.929870</td>\n",
       "      <td>0.090828</td>\n",
       "      <td>-0.095242</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.603672</td>\n",
       "      <td>0.174326</td>\n",
       "      <td>-0.233102</td>\n",
       "      <td>-0.452149</td>\n",
       "      <td>0.927693</td>\n",
       "      <td>0.320566</td>\n",
       "      <td>0.737518</td>\n",
       "      <td>0.138547</td>\n",
       "      <td>1.013321</td>\n",
       "      <td>-0.503070</td>\n",
       "      <td>0.170520</td>\n",
       "      <td>-0.236994</td>\n",
       "      <td>0.403923</td>\n",
       "      <td>0.352233</td>\n",
       "      <td>-0.010432</td>\n",
       "      <td>-0.446412</td>\n",
       "      <td>-0.494003</td>\n",
       "      <td>0.712903</td>\n",
       "      <td>0.281747</td>\n",
       "      <td>-0.043063</td>\n",
       "      <td>0.183140</td>\n",
       "      <td>0.226364</td>\n",
       "      <td>0.131696</td>\n",
       "      <td>0.180204</td>\n",
       "      <td>0.144474</td>\n",
       "      <td>0.145077</td>\n",
       "      <td>0.045493</td>\n",
       "      <td>-0.119915</td>\n",
       "      <td>0.117218</td>\n",
       "      <td>0.620796</td>\n",
       "      <td>-0.186910</td>\n",
       "      <td>-0.194374</td>\n",
       "      <td>0.630890</td>\n",
       "      <td>0.425284</td>\n",
       "      <td>0.027759</td>\n",
       "      <td>-0.493276</td>\n",
       "      <td>0.415432</td>\n",
       "      <td>-0.243451</td>\n",
       "      <td>0.811091</td>\n",
       "      <td>0.186713</td>\n",
       "      <td>0.097599</td>\n",
       "      <td>0.228464</td>\n",
       "      <td>-0.240718</td>\n",
       "      <td>0.009613</td>\n",
       "      <td>-0.257763</td>\n",
       "      <td>0.246641</td>\n",
       "      <td>0.406052</td>\n",
       "      <td>-0.316100</td>\n",
       "      <td>-0.559927</td>\n",
       "      <td>-0.311160</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12350</th>\n",
       "      <td>1</td>\n",
       "      <td>17</td>\n",
       "      <td>17</td>\n",
       "      <td>334.40</td>\n",
       "      <td>19.670588</td>\n",
       "      <td>334.40</td>\n",
       "      <td>334.40</td>\n",
       "      <td>334.400000</td>\n",
       "      <td>-4.929320</td>\n",
       "      <td>-0.714859</td>\n",
       "      <td>-1.580930</td>\n",
       "      <td>-0.469767</td>\n",
       "      <td>-0.116290</td>\n",
       "      <td>0.647835</td>\n",
       "      <td>0.112904</td>\n",
       "      <td>-1.336313</td>\n",
       "      <td>1.547069</td>\n",
       "      <td>-1.190409</td>\n",
       "      <td>-0.213170</td>\n",
       "      <td>1.057289</td>\n",
       "      <td>-0.475002</td>\n",
       "      <td>0.856574</td>\n",
       "      <td>-1.305371</td>\n",
       "      <td>0.242700</td>\n",
       "      <td>0.268312</td>\n",
       "      <td>1.952558</td>\n",
       "      <td>-0.227722</td>\n",
       "      <td>0.569825</td>\n",
       "      <td>-0.186512</td>\n",
       "      <td>0.903530</td>\n",
       "      <td>1.364748</td>\n",
       "      <td>0.178125</td>\n",
       "      <td>-0.260398</td>\n",
       "      <td>-0.168825</td>\n",
       "      <td>-0.681952</td>\n",
       "      <td>1.081305</td>\n",
       "      <td>-0.155239</td>\n",
       "      <td>-0.397164</td>\n",
       "      <td>0.240051</td>\n",
       "      <td>0.634068</td>\n",
       "      <td>-1.447081</td>\n",
       "      <td>-0.186097</td>\n",
       "      <td>-0.315379</td>\n",
       "      <td>-0.306429</td>\n",
       "      <td>-1.359020</td>\n",
       "      <td>-0.001473</td>\n",
       "      <td>0.579056</td>\n",
       "      <td>-0.248313</td>\n",
       "      <td>0.588129</td>\n",
       "      <td>-0.217986</td>\n",
       "      <td>...</td>\n",
       "      <td>1.556940</td>\n",
       "      <td>2.420670</td>\n",
       "      <td>0.865762</td>\n",
       "      <td>0.453298</td>\n",
       "      <td>-0.993826</td>\n",
       "      <td>-1.640322</td>\n",
       "      <td>1.547916</td>\n",
       "      <td>1.346862</td>\n",
       "      <td>-0.364600</td>\n",
       "      <td>-1.070965</td>\n",
       "      <td>-0.477111</td>\n",
       "      <td>0.513733</td>\n",
       "      <td>1.054107</td>\n",
       "      <td>-0.190973</td>\n",
       "      <td>-0.051314</td>\n",
       "      <td>-0.076611</td>\n",
       "      <td>-0.034641</td>\n",
       "      <td>0.851705</td>\n",
       "      <td>-0.620095</td>\n",
       "      <td>0.348728</td>\n",
       "      <td>0.521268</td>\n",
       "      <td>1.447778</td>\n",
       "      <td>0.489761</td>\n",
       "      <td>1.487142</td>\n",
       "      <td>-0.401093</td>\n",
       "      <td>-0.761589</td>\n",
       "      <td>-0.079544</td>\n",
       "      <td>0.569086</td>\n",
       "      <td>-0.275222</td>\n",
       "      <td>-0.320405</td>\n",
       "      <td>0.523359</td>\n",
       "      <td>0.833541</td>\n",
       "      <td>0.124539</td>\n",
       "      <td>-0.004853</td>\n",
       "      <td>0.350597</td>\n",
       "      <td>-0.079295</td>\n",
       "      <td>1.062622</td>\n",
       "      <td>0.150736</td>\n",
       "      <td>-0.752001</td>\n",
       "      <td>0.890077</td>\n",
       "      <td>0.892657</td>\n",
       "      <td>1.049206</td>\n",
       "      <td>0.155618</td>\n",
       "      <td>0.172248</td>\n",
       "      <td>-0.171932</td>\n",
       "      <td>0.179725</td>\n",
       "      <td>-0.358419</td>\n",
       "      <td>-2.150534</td>\n",
       "      <td>-1.023392</td>\n",
       "      <td>-1.995684</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12352</th>\n",
       "      <td>8</td>\n",
       "      <td>85</td>\n",
       "      <td>59</td>\n",
       "      <td>2506.04</td>\n",
       "      <td>29.482824</td>\n",
       "      <td>120.33</td>\n",
       "      <td>840.30</td>\n",
       "      <td>313.255000</td>\n",
       "      <td>-2.053113</td>\n",
       "      <td>2.110950</td>\n",
       "      <td>0.184807</td>\n",
       "      <td>0.128240</td>\n",
       "      <td>0.087787</td>\n",
       "      <td>-1.943579</td>\n",
       "      <td>2.383307</td>\n",
       "      <td>0.899610</td>\n",
       "      <td>-0.671980</td>\n",
       "      <td>0.081452</td>\n",
       "      <td>0.876840</td>\n",
       "      <td>1.051124</td>\n",
       "      <td>-0.741349</td>\n",
       "      <td>-2.071853</td>\n",
       "      <td>-0.637244</td>\n",
       "      <td>0.019206</td>\n",
       "      <td>0.114969</td>\n",
       "      <td>0.573977</td>\n",
       "      <td>-0.030594</td>\n",
       "      <td>0.105880</td>\n",
       "      <td>-0.693458</td>\n",
       "      <td>-0.911360</td>\n",
       "      <td>-2.330069</td>\n",
       "      <td>0.517448</td>\n",
       "      <td>-1.757740</td>\n",
       "      <td>0.224012</td>\n",
       "      <td>-1.062572</td>\n",
       "      <td>-1.806766</td>\n",
       "      <td>-0.929000</td>\n",
       "      <td>2.098583</td>\n",
       "      <td>-0.155669</td>\n",
       "      <td>2.220819</td>\n",
       "      <td>2.238957</td>\n",
       "      <td>-0.573106</td>\n",
       "      <td>2.444062</td>\n",
       "      <td>1.377984</td>\n",
       "      <td>0.187290</td>\n",
       "      <td>-0.400572</td>\n",
       "      <td>1.804934</td>\n",
       "      <td>1.208779</td>\n",
       "      <td>-0.960939</td>\n",
       "      <td>-0.863664</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.028970</td>\n",
       "      <td>0.244557</td>\n",
       "      <td>0.306695</td>\n",
       "      <td>-0.543993</td>\n",
       "      <td>-0.426855</td>\n",
       "      <td>0.544659</td>\n",
       "      <td>-0.509539</td>\n",
       "      <td>-0.745640</td>\n",
       "      <td>-0.149256</td>\n",
       "      <td>-0.031816</td>\n",
       "      <td>0.357092</td>\n",
       "      <td>-0.003698</td>\n",
       "      <td>0.502374</td>\n",
       "      <td>-0.074565</td>\n",
       "      <td>0.383567</td>\n",
       "      <td>-0.014870</td>\n",
       "      <td>-0.105848</td>\n",
       "      <td>0.316509</td>\n",
       "      <td>-0.137248</td>\n",
       "      <td>-0.025613</td>\n",
       "      <td>0.196455</td>\n",
       "      <td>-0.259308</td>\n",
       "      <td>-0.078183</td>\n",
       "      <td>0.202068</td>\n",
       "      <td>-0.164799</td>\n",
       "      <td>-0.010466</td>\n",
       "      <td>0.258529</td>\n",
       "      <td>0.329032</td>\n",
       "      <td>0.275869</td>\n",
       "      <td>0.388376</td>\n",
       "      <td>0.294109</td>\n",
       "      <td>0.129528</td>\n",
       "      <td>0.022517</td>\n",
       "      <td>-0.442529</td>\n",
       "      <td>-0.298672</td>\n",
       "      <td>-0.136678</td>\n",
       "      <td>0.186785</td>\n",
       "      <td>-0.233479</td>\n",
       "      <td>0.094408</td>\n",
       "      <td>-0.005697</td>\n",
       "      <td>-0.224264</td>\n",
       "      <td>0.002607</td>\n",
       "      <td>-0.179418</td>\n",
       "      <td>-0.396663</td>\n",
       "      <td>0.141766</td>\n",
       "      <td>-0.075031</td>\n",
       "      <td>0.038654</td>\n",
       "      <td>-0.078615</td>\n",
       "      <td>0.288926</td>\n",
       "      <td>-0.113879</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 308 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            total_transactions  total_products  total_unique_products  \\\n",
       "CustomerID                                                              \n",
       "12347                        7             182                    103   \n",
       "12348                        4              31                     22   \n",
       "12349                        1              73                     73   \n",
       "12350                        1              17                     17   \n",
       "12352                        8              85                     59   \n",
       "\n",
       "            total_sales  avg_product_value  min_cart_value  max_cart_value  \\\n",
       "CustomerID                                                                   \n",
       "12347           4310.00          23.681319          224.82         1294.32   \n",
       "12348           1797.24          57.975484          227.44          892.80   \n",
       "12349           1757.55          24.076027         1757.55         1757.55   \n",
       "12350            334.40          19.670588          334.40          334.40   \n",
       "12352           2506.04          29.482824          120.33          840.30   \n",
       "\n",
       "            avg_cart_value       PC1       PC2       PC3       PC4       PC5  \\\n",
       "CustomerID                                                                     \n",
       "12347           615.714286  5.876141  4.828917 -0.495545  3.093849 -0.633706   \n",
       "12348           449.310000 -4.219837  1.226217 -1.881912 -0.174340  0.817996   \n",
       "12349          1757.550000 -1.260847  2.387664 -0.073148 -1.119161 -0.034339   \n",
       "12350           334.400000 -4.929320 -0.714859 -1.580930 -0.469767 -0.116290   \n",
       "12352           313.255000 -2.053113  2.110950  0.184807  0.128240  0.087787   \n",
       "\n",
       "                  PC6       PC7       PC8       PC9      PC10      PC11  \\\n",
       "CustomerID                                                                \n",
       "12347      -11.534092  2.161996  7.116957  5.336009  4.424414  4.637002   \n",
       "12348        1.160834 -0.871683 -0.846634  0.460240  0.854274 -0.401793   \n",
       "12349       -3.715390 -0.343602 -0.149045 -0.790001 -0.379234  1.570090   \n",
       "12350        0.647835  0.112904 -1.336313  1.547069 -1.190409 -0.213170   \n",
       "12352       -1.943579  2.383307  0.899610 -0.671980  0.081452  0.876840   \n",
       "\n",
       "                PC12       PC13      PC14       PC15      PC16      PC17  \\\n",
       "CustomerID                                                                 \n",
       "12347       2.316503  12.310783  1.986696 -11.431269  5.617791 -2.177268   \n",
       "12348       0.981803  -1.104237  0.770190  -0.092975 -0.313393 -0.448633   \n",
       "12349       0.183310   0.223639 -0.051066   3.179090 -0.512679 -1.839577   \n",
       "12350       1.057289  -0.475002  0.856574  -1.305371  0.242700  0.268312   \n",
       "12352       1.051124  -0.741349 -2.071853  -0.637244  0.019206  0.114969   \n",
       "\n",
       "                PC18       PC19       PC20      PC21      PC22       PC23  \\\n",
       "CustomerID                                                                  \n",
       "12347       7.665832 -12.135817 -16.049000 -5.949238  7.814288  40.518194   \n",
       "12348       1.207302   0.381678  -0.469152  0.185789  0.054585   0.499987   \n",
       "12349       0.383216   0.317492  -0.003006 -0.420161  0.456437  -0.789675   \n",
       "12350       1.952558  -0.227722   0.569825 -0.186512  0.903530   1.364748   \n",
       "12352       0.573977  -0.030594   0.105880 -0.693458 -0.911360  -2.330069   \n",
       "\n",
       "                 PC24       PC25      PC26       PC27      PC28       PC29  \\\n",
       "CustomerID                                                                   \n",
       "12347      -35.575039 -17.557418  6.869259  20.886868  2.296657 -26.160394   \n",
       "12348        0.373514   1.226615 -0.773072  -0.270125  0.850016  -0.370481   \n",
       "12349        1.751367  -1.481236  0.757191  -0.912292  0.771948   0.503370   \n",
       "12350        0.178125  -0.260398 -0.168825  -0.681952  1.081305  -0.155239   \n",
       "12352        0.517448  -1.757740  0.224012  -1.062572 -1.806766  -0.929000   \n",
       "\n",
       "                 PC30      PC31       PC32      PC33       PC34      PC35  \\\n",
       "CustomerID                                                                  \n",
       "12347      -32.580135  7.880285  12.358371  6.826718 -32.957865 -4.405057   \n",
       "12348       -0.573939  1.012802  -0.580828  0.949802  -0.733970  0.353890   \n",
       "12349       -1.345637  0.286381   0.668565  0.085125  -0.216750 -0.667408   \n",
       "12350       -0.397164  0.240051   0.634068 -1.447081  -0.186097 -0.315379   \n",
       "12352        2.098583 -0.155669   2.220819  2.238957  -0.573106  2.444062   \n",
       "\n",
       "                 PC36       PC37       PC38      PC39      PC40      PC41  \\\n",
       "CustomerID                                                                  \n",
       "12347      -17.238628 -11.272274 -14.487767 -2.766330  4.736717 -6.985884   \n",
       "12348       -0.387479  -0.298541   0.396418  0.154924  0.333479 -0.466261   \n",
       "12349       -1.535679   0.783801  -1.418823 -0.098817 -0.929870  0.090828   \n",
       "12350       -0.306429  -1.359020  -0.001473  0.579056 -0.248313  0.588129   \n",
       "12352        1.377984   0.187290  -0.400572  1.804934  1.208779 -0.960939   \n",
       "\n",
       "                PC42    ...        PC251     PC252     PC253     PC254  \\\n",
       "CustomerID              ...                                              \n",
       "12347      -1.464059    ...     0.064395 -0.415426 -0.206658 -0.005371   \n",
       "12348       0.703454    ...     0.129957  1.022734 -0.456650 -1.224744   \n",
       "12349      -0.095242    ...    -0.603672  0.174326 -0.233102 -0.452149   \n",
       "12350      -0.217986    ...     1.556940  2.420670  0.865762  0.453298   \n",
       "12352      -0.863664    ...    -0.028970  0.244557  0.306695 -0.543993   \n",
       "\n",
       "               PC255     PC256     PC257     PC258     PC259     PC260  \\\n",
       "CustomerID                                                               \n",
       "12347       0.188138  0.242410 -0.195482 -0.162702 -0.200203 -0.170627   \n",
       "12348       0.530329 -0.994510 -0.371455 -0.468078  0.489332 -0.620042   \n",
       "12349       0.927693  0.320566  0.737518  0.138547  1.013321 -0.503070   \n",
       "12350      -0.993826 -1.640322  1.547916  1.346862 -0.364600 -1.070965   \n",
       "12352      -0.426855  0.544659 -0.509539 -0.745640 -0.149256 -0.031816   \n",
       "\n",
       "               PC261     PC262     PC263     PC264     PC265     PC266  \\\n",
       "CustomerID                                                               \n",
       "12347       0.118951 -0.124982 -0.092571 -0.209632 -0.122009  0.112249   \n",
       "12348       0.995225 -1.389717  1.171395 -0.617378  0.078084  1.336674   \n",
       "12349       0.170520 -0.236994  0.403923  0.352233 -0.010432 -0.446412   \n",
       "12350      -0.477111  0.513733  1.054107 -0.190973 -0.051314 -0.076611   \n",
       "12352       0.357092 -0.003698  0.502374 -0.074565  0.383567 -0.014870   \n",
       "\n",
       "               PC267     PC268     PC269     PC270     PC271     PC272  \\\n",
       "CustomerID                                                               \n",
       "12347       0.097985  0.047909  0.191655 -0.245912 -0.183734 -0.034925   \n",
       "12348      -0.120785  0.330687 -0.235496  0.112896  0.654065  0.388106   \n",
       "12349      -0.494003  0.712903  0.281747 -0.043063  0.183140  0.226364   \n",
       "12350      -0.034641  0.851705 -0.620095  0.348728  0.521268  1.447778   \n",
       "12352      -0.105848  0.316509 -0.137248 -0.025613  0.196455 -0.259308   \n",
       "\n",
       "               PC273     PC274     PC275     PC276     PC277     PC278  \\\n",
       "CustomerID                                                               \n",
       "12347       0.108376 -0.060922 -0.060365 -0.033723  0.029060 -0.006391   \n",
       "12348      -0.654803  0.410856 -0.836039  1.419597  0.055308  0.657981   \n",
       "12349       0.131696  0.180204  0.144474  0.145077  0.045493 -0.119915   \n",
       "12350       0.489761  1.487142 -0.401093 -0.761589 -0.079544  0.569086   \n",
       "12352      -0.078183  0.202068 -0.164799 -0.010466  0.258529  0.329032   \n",
       "\n",
       "               PC279     PC280     PC281     PC282     PC283     PC284  \\\n",
       "CustomerID                                                               \n",
       "12347      -0.001222  0.040427  0.089004 -0.218621  0.158177  0.049166   \n",
       "12348       0.559558  0.225912 -0.892273 -0.111400 -0.459792  0.133336   \n",
       "12349       0.117218  0.620796 -0.186910 -0.194374  0.630890  0.425284   \n",
       "12350      -0.275222 -0.320405  0.523359  0.833541  0.124539 -0.004853   \n",
       "12352       0.275869  0.388376  0.294109  0.129528  0.022517 -0.442529   \n",
       "\n",
       "               PC285     PC286     PC287     PC288     PC289     PC290  \\\n",
       "CustomerID                                                               \n",
       "12347      -0.037660 -0.024126  0.094181 -0.109886 -0.035857  0.015897   \n",
       "12348      -0.166611  0.015979 -1.039514 -0.153822 -0.368017 -0.717055   \n",
       "12349       0.027759 -0.493276  0.415432 -0.243451  0.811091  0.186713   \n",
       "12350       0.350597 -0.079295  1.062622  0.150736 -0.752001  0.890077   \n",
       "12352      -0.298672 -0.136678  0.186785 -0.233479  0.094408 -0.005697   \n",
       "\n",
       "               PC291     PC292     PC293     PC294     PC295     PC296  \\\n",
       "CustomerID                                                               \n",
       "12347      -0.071372 -0.046723  0.109823  0.021534 -0.073246 -0.005525   \n",
       "12348      -0.187247 -0.006443  0.247051  0.526344  0.048497  0.309830   \n",
       "12349       0.097599  0.228464 -0.240718  0.009613 -0.257763  0.246641   \n",
       "12350       0.892657  1.049206  0.155618  0.172248 -0.171932  0.179725   \n",
       "12352      -0.224264  0.002607 -0.179418 -0.396663  0.141766 -0.075031   \n",
       "\n",
       "               PC297     PC298     PC299     PC300  \n",
       "CustomerID                                          \n",
       "12347      -0.046094  0.134262 -0.046511  0.068503  \n",
       "12348      -0.974179 -0.156291  0.451286  1.135782  \n",
       "12349       0.406052 -0.316100 -0.559927 -0.311160  \n",
       "12350      -0.358419 -2.150534 -1.023392 -1.995684  \n",
       "12352       0.038654 -0.078615  0.288926 -0.113879  \n",
       "\n",
       "[5 rows x 308 columns]"
      ]
     },
     "execution_count": 193,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Join base_df with pca_item_data\n",
    "pca_df = base_df.join(pca_item_data)\n",
    "\n",
    "# Display first 5 rows of pca_df\n",
    "pca_df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Number of clusters\n",
    "So, how many clusters should you set?\n",
    "* As with much of Unsupervised Learning, there's no right or wrong answer.\n",
    "* Typically, you should consider how your client/key stakeholder will use the clusters.\n",
    "* For example, let's say our client, the online gift retailer, employs 3 customer service reps, and they want to assign one cluster to each rep.\n",
    "* In that case, the obvious answer is 3.\n",
    "* For this project, we'll set the number of clusters to 3. However, you should always feel free to adapt this number depending on what you need.\n",
    "\n",
    "First scale both the dataframes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 194,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Initialize instance of StandardScaler\n",
    "t_scaler = StandardScaler()\n",
    "p_scaler = StandardScaler()\n",
    "\n",
    "# Fit and transform\n",
    "threshold_df_scaled = t_scaler.fit_transform(threshold_df)\n",
    "pca_df_scaled = p_scaler.fit_transform(pca_df)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "K-Means with threshold_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 195,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "t_kmeans = KMeans(n_clusters = 3, init = 'k-means++', random_state = 123)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 196,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "t_kmeans.fit(threshold_df_scaled)\n",
    "threshold_df['cluster'] = t_kmeans.fit_predict(threshold_df_scaled)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 197,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<seaborn.axisgrid.FacetGrid at 0x15dbb97ffd0>"
      ]
     },
     "execution_count": 197,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYoAAAFgCAYAAACloT70AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nSUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x15dbb97f978>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Scatterplot, colored by cluster\n",
    "sns.lmplot(x='total_sales', y='avg_cart_value', hue='cluster', data=threshold_df, fit_reg=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "K-Means with pca_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 198,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "p_kmeans = KMeans(n_clusters = 3, init = 'k-means++', random_state = 123)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 199,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "p_kmeans.fit(pca_df_scaled)\n",
    "pca_df['cluster'] = p_kmeans.fit_predict(pca_df_scaled)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 200,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<seaborn.axisgrid.FacetGrid at 0x15dbb97ccf8>"
      ]
     },
     "execution_count": 200,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYoAAAFgCAYAAACloT70AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nLZiI7GwAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x15dbb95ec18>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Scatterplot, colored by cluster\n",
    "sns.lmplot(x='total_sales', y='avg_cart_value', hue='cluster', data=pca_df, fit_reg=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 201,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.59626023895962021"
      ]
     },
     "execution_count": 201,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Similary between base_df.cluster and threshold_df.cluster\n",
    "adjusted_rand_score(pca_df.cluster, threshold_df.cluster)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "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": "3.6.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}