Customer Segmentation.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "import plotly.express as px\n",
    "\n",
    "import datetime"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>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",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>536365</td>\n",
       "      <td>85123A</td>\n",
       "      <td>WHITE HANGING HEART T-LIGHT HOLDER</td>\n",
       "      <td>6</td>\n",
       "      <td>2010-12-01 08:26:00</td>\n",
       "      <td>2.55</td>\n",
       "      <td>17850.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>536365</td>\n",
       "      <td>71053</td>\n",
       "      <td>WHITE METAL LANTERN</td>\n",
       "      <td>6</td>\n",
       "      <td>2010-12-01 08:26:00</td>\n",
       "      <td>3.39</td>\n",
       "      <td>17850.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>536365</td>\n",
       "      <td>84406B</td>\n",
       "      <td>CREAM CUPID HEARTS COAT HANGER</td>\n",
       "      <td>8</td>\n",
       "      <td>2010-12-01 08:26:00</td>\n",
       "      <td>2.75</td>\n",
       "      <td>17850.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>536365</td>\n",
       "      <td>84029G</td>\n",
       "      <td>KNITTED UNION FLAG HOT WATER BOTTLE</td>\n",
       "      <td>6</td>\n",
       "      <td>2010-12-01 08:26:00</td>\n",
       "      <td>3.39</td>\n",
       "      <td>17850.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>536365</td>\n",
       "      <td>84029E</td>\n",
       "      <td>RED WOOLLY HOTTIE WHITE HEART.</td>\n",
       "      <td>6</td>\n",
       "      <td>2010-12-01 08:26:00</td>\n",
       "      <td>3.39</td>\n",
       "      <td>17850.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   InvoiceNo StockCode                          Description  Quantity  \\\n",
       "0     536365    85123A   WHITE HANGING HEART T-LIGHT HOLDER         6   \n",
       "1     536365     71053                  WHITE METAL LANTERN         6   \n",
       "2     536365    84406B       CREAM CUPID HEARTS COAT HANGER         8   \n",
       "3     536365    84029G  KNITTED UNION FLAG HOT WATER BOTTLE         6   \n",
       "4     536365    84029E       RED WOOLLY HOTTIE WHITE HEART.         6   \n",
       "\n",
       "           InvoiceDate  UnitPrice  CustomerID  \n",
       "0  2010-12-01 08:26:00       2.55     17850.0  \n",
       "1  2010-12-01 08:26:00       3.39     17850.0  \n",
       "2  2010-12-01 08:26:00       2.75     17850.0  \n",
       "3  2010-12-01 08:26:00       3.39     17850.0  \n",
       "4  2010-12-01 08:26:00       3.39     17850.0  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_csv('online retail v2.csv', index_col=0)\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.InvoiceDate = pd.to_datetime(df.InvoiceDate)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 348867 entries, 0 to 541893\n",
      "Data columns (total 7 columns):\n",
      " #   Column       Non-Null Count   Dtype         \n",
      "---  ------       --------------   -----         \n",
      " 0   InvoiceNo    348867 non-null  int64         \n",
      " 1   StockCode    348867 non-null  object        \n",
      " 2   Description  348867 non-null  object        \n",
      " 3   Quantity     348867 non-null  int64         \n",
      " 4   InvoiceDate  348867 non-null  datetime64[ns]\n",
      " 5   UnitPrice    348867 non-null  float64       \n",
      " 6   CustomerID   348867 non-null  float64       \n",
      "dtypes: datetime64[ns](1), float64(2), int64(2), object(2)\n",
      "memory usage: 21.3+ MB\n"
     ]
    }
   ],
   "source": [
    "df.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['12-2010', '01-2011', '02-2011', '03-2011', '04-2011', '05-2011',\n",
       "       '06-2011', '07-2011', '08-2011', '09-2011', '10-2011', '11-2011',\n",
       "       '12-2011'], dtype=object)"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.InvoiceDate.dt.strftime('%m-%Y').unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>InvoiceDate</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CustomerID</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>12747.0</th>\n",
       "      <td>2011-12-07 14:34:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12748.0</th>\n",
       "      <td>2011-12-09 12:20:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12749.0</th>\n",
       "      <td>2011-12-06 09:56:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12820.0</th>\n",
       "      <td>2011-12-06 15:12:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12821.0</th>\n",
       "      <td>2011-05-09 15:51:00</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   InvoiceDate\n",
       "CustomerID                    \n",
       "12747.0    2011-12-07 14:34:00\n",
       "12748.0    2011-12-09 12:20:00\n",
       "12749.0    2011-12-06 09:56:00\n",
       "12820.0    2011-12-06 15:12:00\n",
       "12821.0    2011-05-09 15:51:00"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_customer = df.groupby('CustomerID')['InvoiceDate'].max().to_frame()\n",
    "df_customer.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "def recency_score(invdate):\n",
    "    if invdate >= datetime.datetime(2011, 12, 1):\n",
    "        return 10;\n",
    "\n",
    "    elif invdate >= datetime.datetime(2011, 11, 1):\n",
    "        return 9;\n",
    "\n",
    "    elif invdate >= datetime.datetime(2011, 10, 1):\n",
    "        return 8;\n",
    "\n",
    "    elif invdate >= datetime.datetime(2011, 9, 1):\n",
    "        return 7;\n",
    "\n",
    "    elif invdate >= datetime.datetime(2011, 8, 1):\n",
    "        return 6;\n",
    "\n",
    "    elif invdate >= datetime.datetime(2011, 7, 1):\n",
    "        return 5;\n",
    "\n",
    "    elif invdate >= datetime.datetime(2011, 6, 1):\n",
    "        return 4;\n",
    "\n",
    "    elif invdate >= datetime.datetime(2011, 4, 1):\n",
    "        return 3;\n",
    "\n",
    "    elif invdate >= datetime.datetime(2011, 2, 1):\n",
    "        return 2;\n",
    "\n",
    "    else:\n",
    "        return 1;"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>recency</th>\n",
       "      <th>DaysElapsed</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>12747.0</th>\n",
       "      <td>10</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12748.0</th>\n",
       "      <td>10</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12749.0</th>\n",
       "      <td>10</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12820.0</th>\n",
       "      <td>10</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12821.0</th>\n",
       "      <td>3</td>\n",
       "      <td>213</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18280.0</th>\n",
       "      <td>2</td>\n",
       "      <td>277</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18281.0</th>\n",
       "      <td>4</td>\n",
       "      <td>180</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18282.0</th>\n",
       "      <td>10</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18283.0</th>\n",
       "      <td>10</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18287.0</th>\n",
       "      <td>8</td>\n",
       "      <td>42</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3909 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            recency  DaysElapsed\n",
       "CustomerID                      \n",
       "12747.0          10            1\n",
       "12748.0          10            0\n",
       "12749.0          10            3\n",
       "12820.0          10            2\n",
       "12821.0           3          213\n",
       "...             ...          ...\n",
       "18280.0           2          277\n",
       "18281.0           4          180\n",
       "18282.0          10            7\n",
       "18283.0          10            3\n",
       "18287.0           8           42\n",
       "\n",
       "[3909 rows x 2 columns]"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_customer['recency'] = df_customer.InvoiceDate.map(recency_score)\n",
    "df_customer['DaysElapsed'] = (df_customer.InvoiceDate.max() - df_customer.InvoiceDate).map(lambda x : x.days)\n",
    "df_customer.drop('InvoiceDate', axis=1, inplace=True)\n",
    "df_customer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "CustomerID\n",
       "12747.0      33\n",
       "12748.0    2090\n",
       "12749.0      71\n",
       "12820.0      34\n",
       "12821.0       0\n",
       "Name: PurchaseFrequency, dtype: int64"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "purchase_frequency = df.groupby('CustomerID').apply(lambda x : x[x['InvoiceDate'] >= datetime.datetime(2011, 10, 1)].count())['InvoiceDate']\n",
    "purchase_frequency.name = 'PurchaseFrequency'\n",
    "purchase_frequency.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>recency</th>\n",
       "      <th>DaysElapsed</th>\n",
       "      <th>PurchaseFrequency</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>12747.0</th>\n",
       "      <td>10</td>\n",
       "      <td>1</td>\n",
       "      <td>33</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12748.0</th>\n",
       "      <td>10</td>\n",
       "      <td>0</td>\n",
       "      <td>2090</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12749.0</th>\n",
       "      <td>10</td>\n",
       "      <td>3</td>\n",
       "      <td>71</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12820.0</th>\n",
       "      <td>10</td>\n",
       "      <td>2</td>\n",
       "      <td>34</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12821.0</th>\n",
       "      <td>3</td>\n",
       "      <td>213</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            recency  DaysElapsed  PurchaseFrequency\n",
       "CustomerID                                         \n",
       "12747.0          10            1                 33\n",
       "12748.0          10            0               2090\n",
       "12749.0          10            3                 71\n",
       "12820.0          10            2                 34\n",
       "12821.0           3          213                  0"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_customer = pd.concat([df_customer, purchase_frequency], axis=1)\n",
    "df_customer.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    3909.000000\n",
       "mean       29.646457\n",
       "std       100.336952\n",
       "min         0.000000\n",
       "25%         0.000000\n",
       "50%         9.000000\n",
       "75%        34.000000\n",
       "max      4226.000000\n",
       "Name: PurchaseFrequency, dtype: float64"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_customer.PurchaseFrequency.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Axes: xlabel='PurchaseFrequency', ylabel='Density'>"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.kdeplot(df_customer, x='PurchaseFrequency')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Frequency scores:\n",
    "5 : > 90 percentile\\\n",
    "4 : > 75 percentile\\\n",
    "3 : > 50 percentile\\\n",
    "2 : > 25 percentile\\\n",
    "1 : > 0 percentile"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "def frequency_score(val, series):\n",
    "    if val >= series.quantile(0.9):\n",
    "        return 5\n",
    "    \n",
    "    elif val >= series.quantile(0.75):\n",
    "        return 4\n",
    "    \n",
    "    elif val >= series.quantile(0.5):\n",
    "        return 3\n",
    "    \n",
    "    elif val >= series.quantile(0.408):\n",
    "        return 2\n",
    "    \n",
    "    else:\n",
    "        return 1\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1    1595\n",
       "3     977\n",
       "4     601\n",
       "5     392\n",
       "2     344\n",
       "Name: frequency, dtype: int64"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_customer['frequency'] = df_customer.PurchaseFrequency.map(lambda x: frequency_score(x, df_customer.PurchaseFrequency))\n",
    "df_customer['frequency'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>recency</th>\n",
       "      <th>DaysElapsed</th>\n",
       "      <th>PurchaseFrequency</th>\n",
       "      <th>frequency</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CustomerID</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>12747.0</th>\n",
       "      <td>10</td>\n",
       "      <td>1</td>\n",
       "      <td>33</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12748.0</th>\n",
       "      <td>10</td>\n",
       "      <td>0</td>\n",
       "      <td>2090</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12749.0</th>\n",
       "      <td>10</td>\n",
       "      <td>3</td>\n",
       "      <td>71</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12820.0</th>\n",
       "      <td>10</td>\n",
       "      <td>2</td>\n",
       "      <td>34</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12821.0</th>\n",
       "      <td>3</td>\n",
       "      <td>213</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            recency  DaysElapsed  PurchaseFrequency  frequency\n",
       "CustomerID                                                    \n",
       "12747.0          10            1                 33          3\n",
       "12748.0          10            0               2090          5\n",
       "12749.0          10            3                 71          4\n",
       "12820.0          10            2                 34          4\n",
       "12821.0           3          213                  0          1"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_customer.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "df['TotalPrice'] = df.UnitPrice * df.Quantity"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>recency</th>\n",
       "      <th>DaysElapsed</th>\n",
       "      <th>PurchaseFrequency</th>\n",
       "      <th>frequency</th>\n",
       "      <th>AmountSpent</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",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>12747.0</th>\n",
       "      <td>10</td>\n",
       "      <td>1</td>\n",
       "      <td>33</td>\n",
       "      <td>3</td>\n",
       "      <td>4196.01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12748.0</th>\n",
       "      <td>10</td>\n",
       "      <td>0</td>\n",
       "      <td>2090</td>\n",
       "      <td>5</td>\n",
       "      <td>32937.99</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12749.0</th>\n",
       "      <td>10</td>\n",
       "      <td>3</td>\n",
       "      <td>71</td>\n",
       "      <td>4</td>\n",
       "      <td>4090.88</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12820.0</th>\n",
       "      <td>10</td>\n",
       "      <td>2</td>\n",
       "      <td>34</td>\n",
       "      <td>4</td>\n",
       "      <td>942.34</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12821.0</th>\n",
       "      <td>3</td>\n",
       "      <td>213</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>92.72</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            recency  DaysElapsed  PurchaseFrequency  frequency  AmountSpent\n",
       "CustomerID                                                                 \n",
       "12747.0          10            1                 33          3      4196.01\n",
       "12748.0          10            0               2090          5     32937.99\n",
       "12749.0          10            3                 71          4      4090.88\n",
       "12820.0          10            2                 34          4       942.34\n",
       "12821.0           3          213                  0          1        92.72"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_customer['AmountSpent'] = df.groupby('CustomerID')['TotalPrice'].sum()\n",
    "df_customer.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Axes: xlabel='AmountSpent', ylabel='Density'>"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.kdeplot(df_customer, x = 'AmountSpent')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2.9\n",
      "294.25\n",
      "640.76\n",
      "1557.7600000000002\n"
     ]
    }
   ],
   "source": [
    "print(df_customer.AmountSpent.quantile(0))\n",
    "print(df_customer.AmountSpent.quantile(0.25))\n",
    "print(df_customer.AmountSpent.quantile(0.5))\n",
    "print(df_customer.AmountSpent.quantile(0.75))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count      3909.000000\n",
       "mean       1716.297543\n",
       "std        5993.615163\n",
       "min           2.900000\n",
       "25%         294.250000\n",
       "50%         640.760000\n",
       "75%        1557.760000\n",
       "max      238169.000000\n",
       "Name: AmountSpent, dtype: float64"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_customer.AmountSpent.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "def monetary_score(val, series):\n",
    "    if val >= series.quantile(0.9):\n",
    "        return 5\n",
    "    \n",
    "    elif val >= series.quantile(0.75):\n",
    "        return 4\n",
    "    \n",
    "    elif val >= series.quantile(0.5):\n",
    "        return 3\n",
    "    \n",
    "    elif val >= series.quantile(0.25):\n",
    "        return 2\n",
    "    \n",
    "    else:\n",
    "        return 1\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>recency</th>\n",
       "      <th>DaysElapsed</th>\n",
       "      <th>PurchaseFrequency</th>\n",
       "      <th>frequency</th>\n",
       "      <th>AmountSpent</th>\n",
       "      <th>monetary</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",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>12747.0</th>\n",
       "      <td>10</td>\n",
       "      <td>1</td>\n",
       "      <td>33</td>\n",
       "      <td>3</td>\n",
       "      <td>4196.01</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12748.0</th>\n",
       "      <td>10</td>\n",
       "      <td>0</td>\n",
       "      <td>2090</td>\n",
       "      <td>5</td>\n",
       "      <td>32937.99</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12749.0</th>\n",
       "      <td>10</td>\n",
       "      <td>3</td>\n",
       "      <td>71</td>\n",
       "      <td>4</td>\n",
       "      <td>4090.88</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12820.0</th>\n",
       "      <td>10</td>\n",
       "      <td>2</td>\n",
       "      <td>34</td>\n",
       "      <td>4</td>\n",
       "      <td>942.34</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12821.0</th>\n",
       "      <td>3</td>\n",
       "      <td>213</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>92.72</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            recency  DaysElapsed  PurchaseFrequency  frequency  AmountSpent  \\\n",
       "CustomerID                                                                    \n",
       "12747.0          10            1                 33          3      4196.01   \n",
       "12748.0          10            0               2090          5     32937.99   \n",
       "12749.0          10            3                 71          4      4090.88   \n",
       "12820.0          10            2                 34          4       942.34   \n",
       "12821.0           3          213                  0          1        92.72   \n",
       "\n",
       "            monetary  \n",
       "CustomerID            \n",
       "12747.0            5  \n",
       "12748.0            5  \n",
       "12749.0            5  \n",
       "12820.0            3  \n",
       "12821.0            1  "
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_customer['monetary'] = df_customer.AmountSpent.map(lambda x: monetary_score(x, df_customer.AmountSpent))\n",
    "df_customer.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_rfm = df_customer[['recency', 'frequency', 'monetary']]\n",
    "df_kmeans = df_customer[['DaysElapsed', 'PurchaseFrequency', 'AmountSpent']]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 1) Customer Segmentation Using RFM Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>recency</th>\n",
       "      <th>frequency</th>\n",
       "      <th>monetary</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>12747.0</th>\n",
       "      <td>10</td>\n",
       "      <td>3</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12748.0</th>\n",
       "      <td>10</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12749.0</th>\n",
       "      <td>10</td>\n",
       "      <td>4</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12820.0</th>\n",
       "      <td>10</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12821.0</th>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            recency  frequency  monetary\n",
       "CustomerID                              \n",
       "12747.0          10          3         5\n",
       "12748.0          10          5         5\n",
       "12749.0          10          4         5\n",
       "12820.0          10          4         3\n",
       "12821.0           3          1         1"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_rfm.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "def segmenter(row):\n",
    "    if row['recency'] in [9, 10] and row['frequency'] in [5] and row['monetary'] in [5]:\n",
    "        return 'Soulmates'\n",
    "    \n",
    "    elif row['recency'] in [7, 8, 9, 10] and row['frequency'] in [3, 4, 5] and row['monetary'] in [3, 4, 5]:\n",
    "        return 'Lovers'\n",
    "    \n",
    "    elif row['recency'] in [9, 10] and row['frequency'] in [1] and row['monetary'] in [4, 5]:\n",
    "        return 'New passions'\n",
    "    \n",
    "    elif row['recency'] in [7, 8] and row['frequency'] in [5] and row['monetary'] in [5]:\n",
    "        return 'Flirting'\n",
    "    \n",
    "    elif row['recency'] in [7, 8] and row['frequency'] in [1] and row['monetary'] in [1]:\n",
    "        return 'Apprentice'\n",
    "    \n",
    "    elif row['recency'] in [9, 10] and row['frequency'] in [1, 2] and row['monetary'] in [9, 10]:\n",
    "        return 'Potential Lovers'\n",
    "    \n",
    "    elif row['recency'] in [9, 10] and row['frequency'] in [5, 6, 7, 8] and row['monetary'] in [5, 6, 7, 8]:\n",
    "        return 'Platonic Friends'\n",
    "    \n",
    "    elif row['recency'] in [3, 4, 5, 6] and row['frequency'] in [1,2, 3, 4, 5, 6, 7, 8, 9, 10] and row['monetary'] in [1,2, 3, 4, 5, 6, 7, 8, 9, 10]:\n",
    "        return 'ATDU'\n",
    "    \n",
    "    elif row['recency'] in [1, 2] and row['frequency'] in [9, 10] and row['monetary'] in [9, 10]:\n",
    "        return 'Ex-Lovers'\n",
    "    \n",
    "    elif row['recency'] in [1, 2] and row['frequency'] in [1,2] and row['monetary'] in [9, 10]:\n",
    "        return 'One Night Stand'\n",
    "    \n",
    "    elif row['recency'] in [1, 2] and row['frequency'] in [3, 4] and row['monetary'] in [1,2]:\n",
    "        return 'Break Up'\n",
    "    \n",
    "    else:\n",
    "        return 'Unclassified'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/tmp/ipykernel_2129/842370909.py:1: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  df_rfm['segment'] = df_rfm.apply(segmenter, axis=1)\n"
     ]
    }
   ],
   "source": [
    "df_rfm['segment'] = df_rfm.apply(segmenter, axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Unclassified    1671\n",
       "Lovers          1212\n",
       "ATDU             799\n",
       "Soulmates        153\n",
       "Apprentice        74\n",
       "Name: segment, dtype: int64"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_rfm['segment'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>recency</th>\n",
       "      <th>frequency</th>\n",
       "      <th>monetary</th>\n",
       "      <th>segment</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CustomerID</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>12822.0</th>\n",
       "      <td>7</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>Unclassified</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12823.0</th>\n",
       "      <td>7</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>Unclassified</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12824.0</th>\n",
       "      <td>8</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>Unclassified</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12827.0</th>\n",
       "      <td>10</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>Unclassified</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12829.0</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>Unclassified</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18274.0</th>\n",
       "      <td>9</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>Unclassified</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18276.0</th>\n",
       "      <td>8</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>Unclassified</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18277.0</th>\n",
       "      <td>8</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>Unclassified</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18280.0</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>Unclassified</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18282.0</th>\n",
       "      <td>10</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>Unclassified</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1671 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            recency  frequency  monetary       segment\n",
       "CustomerID                                            \n",
       "12822.0           7          1         3  Unclassified\n",
       "12823.0           7          1         4  Unclassified\n",
       "12824.0           8          3         2  Unclassified\n",
       "12827.0          10          3         2  Unclassified\n",
       "12829.0           1          1         1  Unclassified\n",
       "...             ...        ...       ...           ...\n",
       "18274.0           9          3         1  Unclassified\n",
       "18276.0           8          3         2  Unclassified\n",
       "18277.0           8          2         1  Unclassified\n",
       "18280.0           2          1         1  Unclassified\n",
       "18282.0          10          2         1  Unclassified\n",
       "\n",
       "[1671 rows x 4 columns]"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_rfm[df_rfm.segment == 'Unclassified']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 1) Customer Segmentation Using K-Means"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>DaysElapsed</th>\n",
       "      <th>PurchaseFrequency</th>\n",
       "      <th>AmountSpent</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>12747.0</th>\n",
       "      <td>1</td>\n",
       "      <td>33</td>\n",
       "      <td>4196.01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12748.0</th>\n",
       "      <td>0</td>\n",
       "      <td>2090</td>\n",
       "      <td>32937.99</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12749.0</th>\n",
       "      <td>3</td>\n",
       "      <td>71</td>\n",
       "      <td>4090.88</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12820.0</th>\n",
       "      <td>2</td>\n",
       "      <td>34</td>\n",
       "      <td>942.34</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12821.0</th>\n",
       "      <td>213</td>\n",
       "      <td>0</td>\n",
       "      <td>92.72</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            DaysElapsed  PurchaseFrequency  AmountSpent\n",
       "CustomerID                                             \n",
       "12747.0               1                 33      4196.01\n",
       "12748.0               0               2090     32937.99\n",
       "12749.0               3                 71      4090.88\n",
       "12820.0               2                 34       942.34\n",
       "12821.0             213                  0        92.72"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_kmeans.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Ipy",
   "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": "3.10.9"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}