Pandas_and_SQL_Notebook.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# SQL Take Home Assignment\n",
    "\n",
    "**Author: Joe Au-Yeung**\n",
    "\n",
    "\n",
    "## The database\n",
    "\n",
    "The SQLite3 database contains 3 tables.\n",
    "\n",
    "The first table `transactions`, contains details about each product that a customer has purchased. A transaction is labeled by a `trans_id` and can contain multiple products. Here are the columns in the `transactions` table:\n",
    "   * `trans_id`: the transaction id\n",
    "   * `trans_dt`: the date that the transaction took place (*this was not originally included in the provided README file*)\n",
    "   * `cust_id`: the customer id\n",
    "   * `prod_id`: the product id\n",
    "   * `item_qty`: the quantity of the product that is being purchased\n",
    "   * `item_price`: the per unit price of the product (NOTE: the total revenue\n",
    "     for a product is `item_qty * item_price`)\n",
    "\n",
    "The second table `products`, contains details about each product. The columns in the `products` table are:\n",
    "   * `prod_id`: the product id (same meaning as in `transactions`)\n",
    "   * `prod_name`: the product name\n",
    "   * `brand`: the brand of the product\n",
    "   * `category`: the category of the product\n",
    "   \n",
    "Finally the third table `segments`, contains the history of each customer and which market segment they belong to. Segmentation is calculated periodically for current customers and are appended to this table. The most recent segment is labeled in the `active-flag` column by a `Y`. The columns in this table are:\n",
    "   * `cust_id`: the customer id (same meaning as in `transactions`)\n",
    "   * `seg_name`: the segment of this customer\n",
    "   * `update_at`: the date when this segment was updated (*in the original text file this was labeled as `update_dt` we later discovered the column was actually called `update_at`*)\n",
    "   * `active_flag`: whether or not this segment is the active segment for this customer\n",
    "\n",
    "\n",
    "## Reading the database\n",
    "\n",
    "Before we can begin answering the questions let us import the `sqlite3` library and connecting to the database."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[(1, '2016-01-02 10:06:00', 9085146, 223029, 1, 42.99), (2, '2016-01-02 10:30:00', 1215814, 252270, 1, 103.95), (2, '2016-01-02 10:30:00', 1215814, 260383, 1, 74.99), (4, '2016-01-02 11:33:00', 18511160, 269119, 1, 51.99), (4, '2016-01-02 11:33:00', 18511160, 411162, 1, 59.99)]\n"
     ]
    }
   ],
   "source": [
    "import sqlite3\n",
    "\n",
    "conn = sqlite3.connect('sample.db')\n",
    "\n",
    "sample_query = 'SELECT * FROM transactions;'\n",
    "sample_pull = conn.execute(sample_query).fetchmany(5)\n",
    "print(sample_pull)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "When pulling our sample query from the `transactions` table we notice another column that was not in the provided README file. Let us pull the column headers of the `transactions` table."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[(0, 'trans_id', 'INTEGER', 0, None, 0), (1, 'trans_dt', 'TIMESTAMP', 0, None, 0), (2, 'cust_id', 'INTEGER', 0, None, 0), (3, 'prod_id', 'INTEGER', 0, None, 0), (4, 'item_qty', 'INTEGER', 0, None, 0), (5, 'item_price', 'REAL', 0, None, 0)]\n"
     ]
    }
   ],
   "source": [
    "print(conn.execute('PRAGMA TABLE_INFO(transactions);').fetchall())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "We see that this column is named `trans_dt`. Let us update our intro and begin answering the questions.\n",
    "\n",
    "## Question 1\n",
    "\n",
    "**Find the current active segment for each customer sorted by the segment\n",
    "   update date.  The output should contain three columns: `cust_id`,\n",
    "   `seg_name`, `updated_at`**\n",
    "   \n",
    "Let's assume that the `updated_at` column should be ordered in descending order. We will be using the `sqlite3` library."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The column 'update_dt' does not exist\n"
     ]
    }
   ],
   "source": [
    "#Pulling the data using the sqlite3 library\n",
    "\n",
    "query_1 = '''\n",
    "SELECT cust_id,\n",
    "       seg_name,\n",
    "       update_dt AS updated_at\n",
    "FROM segments\n",
    "WHERE active_flag = Y\n",
    "ORDER BY updated_at DESC;\n",
    "'''\n",
    "\n",
    "#We add this try and accept block in order for the notebook to continue running\n",
    "try:\n",
    "    question_1 = conn.execute(query_1).fetchall()\n",
    "except:\n",
    "    print(\"The column 'update_dt' does not exist\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "When we tried to run the query we got an error saying there was no such column `update_dt`. Let's query the column headers of the `segments` table to dive deeper."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[(0, 'cust_id', 'INTEGER', 0, None, 0), (1, 'seg_name', 'TEXT', 0, None, 0), (2, 'update_at', 'TIMESTAMP', 0, None, 0), (3, 'active_flag', 'TEXT', 0, None, 0)]\n"
     ]
    }
   ],
   "source": [
    "print(conn.execute('PRAGMA TABLE_INFO(segments);').fetchall())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "We see that the column column describing when the segments were updated is called `update_at` not `update_dt`. Let us update our query with this new information."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[(12064, 'INFREQUENT', '2016-06-01 00:00:00'), (106895, 'INFREQUENT', '2016-06-01 00:00:00'), (298867, 'ONE-OFFS', '2016-06-01 00:00:00')]\n"
     ]
    }
   ],
   "source": [
    "query_1 = '''\n",
    "SELECT cust_id,\n",
    "       seg_name,\n",
    "       update_at AS updated_at\n",
    "FROM segments\n",
    "WHERE active_flag = 'Y'\n",
    "ORDER BY updated_at DESC;\n",
    "'''\n",
    "\n",
    "question_1 = conn.execute(query_1).fetchall()\n",
    "print(question_1[:3])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "When working with data we I would be using `Pandas` to turn a pull into a dataframe to easily clean and analyze the data. Let's do that now."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "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>cust_id</th>\n",
       "      <th>seg_name</th>\n",
       "      <th>updated_at</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>12064</td>\n",
       "      <td>INFREQUENT</td>\n",
       "      <td>2016-06-01 00:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>106895</td>\n",
       "      <td>INFREQUENT</td>\n",
       "      <td>2016-06-01 00:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>298867</td>\n",
       "      <td>ONE-OFFS</td>\n",
       "      <td>2016-06-01 00:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>449146</td>\n",
       "      <td>INFREQUENT</td>\n",
       "      <td>2016-06-01 00:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>689576</td>\n",
       "      <td>ONE-OFFS</td>\n",
       "      <td>2016-06-01 00:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1317</th>\n",
       "      <td>11259901</td>\n",
       "      <td>VIP</td>\n",
       "      <td>2014-03-01 00:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1318</th>\n",
       "      <td>5764946</td>\n",
       "      <td>VIP</td>\n",
       "      <td>2014-01-01 00:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1319</th>\n",
       "      <td>8042053</td>\n",
       "      <td>VIP</td>\n",
       "      <td>2014-01-01 00:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1320</th>\n",
       "      <td>14075295</td>\n",
       "      <td>VIP</td>\n",
       "      <td>2014-01-01 00:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1321</th>\n",
       "      <td>15509765</td>\n",
       "      <td>VIP</td>\n",
       "      <td>2014-01-01 00:00:00</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1322 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       cust_id    seg_name           updated_at\n",
       "0        12064  INFREQUENT  2016-06-01 00:00:00\n",
       "1       106895  INFREQUENT  2016-06-01 00:00:00\n",
       "2       298867    ONE-OFFS  2016-06-01 00:00:00\n",
       "3       449146  INFREQUENT  2016-06-01 00:00:00\n",
       "4       689576    ONE-OFFS  2016-06-01 00:00:00\n",
       "...        ...         ...                  ...\n",
       "1317  11259901         VIP  2014-03-01 00:00:00\n",
       "1318   5764946         VIP  2014-01-01 00:00:00\n",
       "1319   8042053         VIP  2014-01-01 00:00:00\n",
       "1320  14075295         VIP  2014-01-01 00:00:00\n",
       "1321  15509765         VIP  2014-01-01 00:00:00\n",
       "\n",
       "[1322 rows x 3 columns]"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "\n",
    "question_1_df = pd.read_sql_query(query_1, conn)\n",
    "question_1_df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Question 2\n",
    "\n",
    "**For each product purchased between Jan 2016 and May 2016 (inclusive), find\n",
    "   the number of distinct transactions.  The output should contain `prod_id`,\n",
    "   `prod_name` and distinct transaction columns.**\n",
    "\n",
    "For this question we will use an inner join on the `transactions` table and the `products` table assuming that all product ids have an associated product name. We will also sort by the `count` column in descending order."
   ]
  },
  {
   "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>prod_id</th>\n",
       "      <th>prod_name</th>\n",
       "      <th>count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>226594814</td>\n",
       "      <td>Product 226594814</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>215734505</td>\n",
       "      <td>Product 215734505</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>138262084</td>\n",
       "      <td>Product 138262084</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>261231</td>\n",
       "      <td>Product 261231</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>220108</td>\n",
       "      <td>Product 220108</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1804</th>\n",
       "      <td>211200</td>\n",
       "      <td>Product 211200</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1805</th>\n",
       "      <td>209999</td>\n",
       "      <td>Product 209999</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1806</th>\n",
       "      <td>209732</td>\n",
       "      <td>Product 209732</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1807</th>\n",
       "      <td>207344</td>\n",
       "      <td>Product 207344</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1808</th>\n",
       "      <td>199922</td>\n",
       "      <td>Product 199922</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1809 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        prod_id          prod_name  count\n",
       "0     226594814  Product 226594814     10\n",
       "1     215734505  Product 215734505     10\n",
       "2     138262084  Product 138262084      9\n",
       "3        261231     Product 261231      9\n",
       "4        220108     Product 220108      9\n",
       "...         ...                ...    ...\n",
       "1804     211200     Product 211200      1\n",
       "1805     209999     Product 209999      1\n",
       "1806     209732     Product 209732      1\n",
       "1807     207344     Product 207344      1\n",
       "1808     199922     Product 199922      1\n",
       "\n",
       "[1809 rows x 3 columns]"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "query_2 = '''\n",
    "SELECT t.prod_id,\n",
    "       p.prod_name,\n",
    "       SUM(item_qty) AS count\n",
    "FROM transactions t\n",
    "INNER JOIN products p ON t.prod_id = p.prod_id\n",
    "WHERE trans_dt BETWEEN '2016-01-01' AND '2016-05-31'\n",
    "GROUP BY t.prod_id\n",
    "ORDER BY count DESC;\n",
    "'''\n",
    "\n",
    "question_2_df = pd.read_sql_query(query_2, conn)\n",
    "question_2_df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Question 3\n",
    "\n",
    "**Find the most recent segment of each customer as of 2016-03-01.\n",
    "   *Hint*: You cannot simply use `active_flag` since that is as of the current\n",
    "   date *not* 2016-03-01.  The output should contain the `cust_id`, `seg_name`\n",
    "   and `update_at`  columns and should have at most one row per customer.**\n",
    "   \n",
    "Once again we will read our query into a dataframe. Since the segment updates just get appended to the table we can can find the max value of the `update_at` column for each `cust_id` as long as we filter the data for dates before March 2016."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>cust_id</th>\n",
       "      <th>seg_name</th>\n",
       "      <th>update_at</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>4402</td>\n",
       "      <td>ONE-OFFS</td>\n",
       "      <td>2016-02-01 00:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>11248</td>\n",
       "      <td>LOYAL</td>\n",
       "      <td>2016-02-01 00:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>12064</td>\n",
       "      <td>INFREQUENT</td>\n",
       "      <td>2016-02-01 00:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>15088</td>\n",
       "      <td>ONE-OFFS</td>\n",
       "      <td>2016-02-01 00:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>66706</td>\n",
       "      <td>ONE-OFFS</td>\n",
       "      <td>2016-02-01 00:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>995</th>\n",
       "      <td>21233469</td>\n",
       "      <td>NEW</td>\n",
       "      <td>2016-02-01 00:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>996</th>\n",
       "      <td>21233549</td>\n",
       "      <td>NEW</td>\n",
       "      <td>2016-02-01 00:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>997</th>\n",
       "      <td>21233596</td>\n",
       "      <td>NEW</td>\n",
       "      <td>2016-02-01 00:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>998</th>\n",
       "      <td>21233911</td>\n",
       "      <td>NEW</td>\n",
       "      <td>2016-02-01 00:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>999</th>\n",
       "      <td>21233988</td>\n",
       "      <td>NEW</td>\n",
       "      <td>2016-02-01 00:00:00</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1000 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      cust_id    seg_name            update_at\n",
       "0        4402    ONE-OFFS  2016-02-01 00:00:00\n",
       "1       11248       LOYAL  2016-02-01 00:00:00\n",
       "2       12064  INFREQUENT  2016-02-01 00:00:00\n",
       "3       15088    ONE-OFFS  2016-02-01 00:00:00\n",
       "4       66706    ONE-OFFS  2016-02-01 00:00:00\n",
       "..        ...         ...                  ...\n",
       "995  21233469         NEW  2016-02-01 00:00:00\n",
       "996  21233549         NEW  2016-02-01 00:00:00\n",
       "997  21233596         NEW  2016-02-01 00:00:00\n",
       "998  21233911         NEW  2016-02-01 00:00:00\n",
       "999  21233988         NEW  2016-02-01 00:00:00\n",
       "\n",
       "[1000 rows x 3 columns]"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "query_3 = '''\n",
    "SELECT cust_id,\n",
    "       seg_name,\n",
    "       MAX(update_at) AS update_at\n",
    "FROM segments\n",
    "WHERE update_at < '2016-03-01'\n",
    "GROUP BY cust_id;\n",
    "'''\n",
    "\n",
    "question_3_df = pd.read_sql_query(query_3, conn)\n",
    "question_3_df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Question 4\n",
    "\n",
    "**Find the most popular category (by revenue) for each active segment.\n",
    "   *Hint*: The current (most up to date) active segment is specified by `active_flag = 'Y'` column in the segments table.**\n",
    "\n",
    "For this question let us create two new tables using the `WITH` clause. The first table will be taking the `transactions` table and calculating the revenue by multiplying the `item_qty` and the `item_price` columns. The second table we will be creating is using the `segments` table and filtering for the active segment for each customer using the `active_flag` column.\n",
    "\n",
    "Since we will be using multiple `WITH` clauses within our query we should test these out before creating our bigger query.\n",
    "\n",
    "We will also order the table by the `revenue` column in descending order."
   ]
  },
  {
   "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>cust_id</th>\n",
       "      <th>prod_id</th>\n",
       "      <th>revenue</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>4402</td>\n",
       "      <td>285204</td>\n",
       "      <td>24.99</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>11248</td>\n",
       "      <td>495237</td>\n",
       "      <td>24.99</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>11248</td>\n",
       "      <td>113495842</td>\n",
       "      <td>34.99</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>12064</td>\n",
       "      <td>238891</td>\n",
       "      <td>20.99</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>12064</td>\n",
       "      <td>192007896</td>\n",
       "      <td>218.97</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1858</th>\n",
       "      <td>21233469</td>\n",
       "      <td>166838318</td>\n",
       "      <td>72.99</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1859</th>\n",
       "      <td>21233549</td>\n",
       "      <td>268780</td>\n",
       "      <td>71.99</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1860</th>\n",
       "      <td>21233596</td>\n",
       "      <td>449428</td>\n",
       "      <td>88.99</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1861</th>\n",
       "      <td>21233911</td>\n",
       "      <td>257304</td>\n",
       "      <td>13.99</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1862</th>\n",
       "      <td>21233988</td>\n",
       "      <td>258205</td>\n",
       "      <td>137.98</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1863 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       cust_id    prod_id  revenue\n",
       "0         4402     285204    24.99\n",
       "1        11248     495237    24.99\n",
       "2        11248  113495842    34.99\n",
       "3        12064     238891    20.99\n",
       "4        12064  192007896   218.97\n",
       "...        ...        ...      ...\n",
       "1858  21233469  166838318    72.99\n",
       "1859  21233549     268780    71.99\n",
       "1860  21233596     449428    88.99\n",
       "1861  21233911     257304    13.99\n",
       "1862  21233988     258205   137.98\n",
       "\n",
       "[1863 rows x 3 columns]"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Our first WITH clause will be finding the revenue from the transaction table\n",
    "\n",
    "query_sales_revenue = '''\n",
    "SELECT cust_id,\n",
    "       prod_id,\n",
    "       SUM(item_qty * item_price) AS revenue\n",
    "FROM transactions\n",
    "GROUP BY prod_id\n",
    "ORDER BY cust_id; \n",
    "'''\n",
    "#We added the ORDER BY clause to ensure that if a customer bought mulitple items that they would show up in their own row\n",
    "#When we create our bigger query we can omit the ORDER BY clause\n",
    "\n",
    "\n",
    "revenue_df = pd.read_sql_query(query_sales_revenue, conn)\n",
    "revenue_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "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>cust_id</th>\n",
       "      <th>seg_name</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>4402</td>\n",
       "      <td>ONE-OFFS</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>11248</td>\n",
       "      <td>LOYAL</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>12064</td>\n",
       "      <td>INFREQUENT</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>12064</td>\n",
       "      <td>INFREQUENT</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>15088</td>\n",
       "      <td>ONE-OFFS</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1317</th>\n",
       "      <td>21233469</td>\n",
       "      <td>NEW</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1318</th>\n",
       "      <td>21233549</td>\n",
       "      <td>NEW</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1319</th>\n",
       "      <td>21233596</td>\n",
       "      <td>NEW</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1320</th>\n",
       "      <td>21233911</td>\n",
       "      <td>NEW</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1321</th>\n",
       "      <td>21233988</td>\n",
       "      <td>NEW</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1322 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       cust_id    seg_name\n",
       "0         4402    ONE-OFFS\n",
       "1        11248       LOYAL\n",
       "2        12064  INFREQUENT\n",
       "3        12064  INFREQUENT\n",
       "4        15088    ONE-OFFS\n",
       "...        ...         ...\n",
       "1317  21233469         NEW\n",
       "1318  21233549         NEW\n",
       "1319  21233596         NEW\n",
       "1320  21233911         NEW\n",
       "1321  21233988         NEW\n",
       "\n",
       "[1322 rows x 2 columns]"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Our second WITH clause will filter for the active segments\n",
    "\n",
    "query_active_segments = '''\n",
    "SELECT cust_id,\n",
    "        seg_name\n",
    "FROM segments\n",
    "WHERE active_flag = 'Y';\n",
    "'''\n",
    "\n",
    "active_segments_df = pd.read_sql_query(query_active_segments, conn)\n",
    "active_segments_df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "Now that we tested our two queries that we will be using in the `WITH` clause we can now create the correct query for question 4."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>seg_name</th>\n",
       "      <th>category</th>\n",
       "      <th>revenue</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>VIP</td>\n",
       "      <td>Men</td>\n",
       "      <td>43822.71</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>INFREQUENT</td>\n",
       "      <td>Make up</td>\n",
       "      <td>42013.49</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>LOYAL</td>\n",
       "      <td>Men</td>\n",
       "      <td>30921.36</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>ONE-OFFS</td>\n",
       "      <td>Women</td>\n",
       "      <td>23540.77</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>NEW</td>\n",
       "      <td>Make up</td>\n",
       "      <td>6088.92</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     seg_name category   revenue\n",
       "0         VIP      Men  43822.71\n",
       "1  INFREQUENT  Make up  42013.49\n",
       "2       LOYAL      Men  30921.36\n",
       "3    ONE-OFFS    Women  23540.77\n",
       "4         NEW  Make up   6088.92"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "query_4 = '''\n",
    "WITH revenue_table AS (\n",
    "SELECT cust_id,\n",
    "       prod_id,\n",
    "       SUM(item_qty * item_price) AS revenue\n",
    "FROM transactions\n",
    "GROUP BY prod_id\n",
    "),\n",
    "\n",
    "active_segments AS (\n",
    "SELECT cust_id,\n",
    "        seg_name\n",
    "FROM segments\n",
    "WHERE active_flag = 'Y'\n",
    ")\n",
    "\n",
    "SELECT  s.seg_name,\n",
    "        p.category,\n",
    "        SUM(rt.revenue) as revenue\n",
    "FROM revenue_table rt\n",
    "INNER JOIN active_segments s ON rt.cust_id = s.cust_id\n",
    "INNER JOIN products p ON rt.prod_id = p.prod_id\n",
    "GROUP BY s.seg_name\n",
    "ORDER BY revenue DESC;\n",
    "'''\n",
    "\n",
    "question_4_df = pd.read_sql_query(query_4, conn)\n",
    "question_4_df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Question 5\n",
    "\n",
    "**Use the current sample database to find insights. Please document your steps, include any code/documents you used (Excel, SQL, Python, etc), and have the final results in a Powerpoint format (limit 3 slides)\n",
    "   *Hint*: This is open ended, but you could look at sales trend, category insights, customers insights**\n",
    "   \n",
    "### Insight 1\n",
    "   \n",
    "The first thing I want to explore is how much each segment is spending per transaction. Although we should take into account the frequency of purchasing as well. With this in mind, we should find the average spend per transaction and multiply it by the number of transactions for each customer segment.\n",
    "\n",
    "We are labeling the customer's segment on the `active_flag`. We are going to take into account all transactions of a customer's lifetime since we are assuming all customers will start in the 'new' segment and gradually be updated into the other segments."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "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>cust_id</th>\n",
       "      <th>seg_name</th>\n",
       "      <th>trans_id</th>\n",
       "      <th>trans_dt</th>\n",
       "      <th>trans_total</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>9085146</td>\n",
       "      <td>ONE-OFFS</td>\n",
       "      <td>1</td>\n",
       "      <td>2016-01-02 10:06:00</td>\n",
       "      <td>42.99</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1215814</td>\n",
       "      <td>INFREQUENT</td>\n",
       "      <td>2</td>\n",
       "      <td>2016-01-02 10:30:00</td>\n",
       "      <td>178.94</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>18511160</td>\n",
       "      <td>ONE-OFFS</td>\n",
       "      <td>4</td>\n",
       "      <td>2016-01-02 11:33:00</td>\n",
       "      <td>501.91</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>15251041</td>\n",
       "      <td>INFREQUENT</td>\n",
       "      <td>6</td>\n",
       "      <td>2016-01-02 11:35:00</td>\n",
       "      <td>61.99</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>14769966</td>\n",
       "      <td>ONE-OFFS</td>\n",
       "      <td>7</td>\n",
       "      <td>2016-01-02 11:36:00</td>\n",
       "      <td>96.97</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1217</th>\n",
       "      <td>3737672</td>\n",
       "      <td>LOYAL</td>\n",
       "      <td>2649</td>\n",
       "      <td>2016-06-16 16:52:00</td>\n",
       "      <td>315.96</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1218</th>\n",
       "      <td>7049122</td>\n",
       "      <td>VIP</td>\n",
       "      <td>2654</td>\n",
       "      <td>2016-06-17 19:36:00</td>\n",
       "      <td>101.97</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1219</th>\n",
       "      <td>3649704</td>\n",
       "      <td>VIP</td>\n",
       "      <td>2662</td>\n",
       "      <td>2016-06-18 10:00:00</td>\n",
       "      <td>464.91</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1220</th>\n",
       "      <td>4095901</td>\n",
       "      <td>INFREQUENT</td>\n",
       "      <td>2665</td>\n",
       "      <td>2016-06-18 17:51:00</td>\n",
       "      <td>89.99</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1221</th>\n",
       "      <td>7989374</td>\n",
       "      <td>ONE-OFFS</td>\n",
       "      <td>2666</td>\n",
       "      <td>2016-06-18 12:34:00</td>\n",
       "      <td>46.99</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1222 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       cust_id    seg_name  trans_id             trans_dt  trans_total\n",
       "0      9085146    ONE-OFFS         1  2016-01-02 10:06:00        42.99\n",
       "1      1215814  INFREQUENT         2  2016-01-02 10:30:00       178.94\n",
       "2     18511160    ONE-OFFS         4  2016-01-02 11:33:00       501.91\n",
       "3     15251041  INFREQUENT         6  2016-01-02 11:35:00        61.99\n",
       "4     14769966    ONE-OFFS         7  2016-01-02 11:36:00        96.97\n",
       "...        ...         ...       ...                  ...          ...\n",
       "1217   3737672       LOYAL      2649  2016-06-16 16:52:00       315.96\n",
       "1218   7049122         VIP      2654  2016-06-17 19:36:00       101.97\n",
       "1219   3649704         VIP      2662  2016-06-18 10:00:00       464.91\n",
       "1220   4095901  INFREQUENT      2665  2016-06-18 17:51:00        89.99\n",
       "1221   7989374    ONE-OFFS      2666  2016-06-18 12:34:00        46.99\n",
       "\n",
       "[1222 rows x 5 columns]"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Since we are reusing a part of the queries from question 4 we do not need to test them again\n",
    "#The difference is we are adding trans_id and trans_dt to the revenue_table and main query\n",
    "#We are also finding the sum of each transaction\n",
    "\n",
    "\n",
    "query_customer_spending = '''\n",
    "WITH revenue_table AS (\n",
    "SELECT cust_id,\n",
    "       trans_id,\n",
    "       trans_dt,\n",
    "       prod_id,\n",
    "       SUM(item_qty * item_price) AS revenue\n",
    "FROM transactions\n",
    "GROUP BY prod_id\n",
    "),\n",
    "\n",
    "active_segments AS (\n",
    "SELECT cust_id,\n",
    "        seg_name\n",
    "FROM segments\n",
    "WHERE active_flag = 'Y'\n",
    ")\n",
    "\n",
    "SELECT r.cust_id,\n",
    "       s.seg_name,\n",
    "       r.trans_id,\n",
    "       r.trans_dt,\n",
    "       SUM(r.revenue) AS trans_total\n",
    "FROM revenue_table r\n",
    "INNER JOIN active_segments s ON r.cust_id = s.cust_id\n",
    "GROUP BY r.trans_id\n",
    "'''\n",
    "\n",
    "revenue_segments_df = pd.read_sql_query(query_customer_spending, conn)\n",
    "revenue_segments_df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "First let's define each segment by the number of transactions that have occurred over a year. To do this let us find the date range that we were given for transactions."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2016-01-02 00:00:00\n",
      "2016-06-18 17:51:00\n"
     ]
    }
   ],
   "source": [
    "print(revenue_segments_df['trans_dt'].min())\n",
    "print(revenue_segments_df['trans_dt'].max())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We see that the date range we were given for transactions is between Jan 2, 2016 to Jun 18, 2016.  Next we can find the average number of transactions for each segment."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'ONE-OFFS': 1.0364372469635628, 'INFREQUENT': 1.2072072072072073, 'LOYAL': 1.6785714285714286, 'NEW': 1.0769230769230769, 'VIP': 3.7}\n"
     ]
    }
   ],
   "source": [
    "def find_average_trans_per_segment(segment):\n",
    "    segment = revenue_segments_df[revenue_segments_df['seg_name'] == segment]\n",
    "    avg_trans = len(segment['trans_id']) / len(segment['cust_id'].unique())\n",
    "    return avg_trans\n",
    "\n",
    "segments_avg_trans = {s:find_average_trans_per_segment(s) for s in revenue_segments_df['seg_name'].unique()}\n",
    "\n",
    "print(segments_avg_trans)\n",
    "    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "Since we were given such a short time frame we can not see any differences in the 'one-offs', 'infrequent', and 'new' segments.\n",
    "\n",
    "Let us find the average transaction amount per segment."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'ONE-OFFS': 96.91372469635628, 'INFREQUENT': 123.63375375375374, 'LOYAL': 240.01564285714286, 'NEW': 102.66430769230769, 'VIP': 585.7735714285714}\n"
     ]
    }
   ],
   "source": [
    "def find_average_trans_amount_per_segment(segment):\n",
    "    segment = revenue_segments_df[revenue_segments_df['seg_name'] == segment]\n",
    "    avg_trans_amount = segment['trans_total'].sum() / len(segment['cust_id'].unique())\n",
    "    return avg_trans_amount\n",
    "\n",
    "segments_avg_trans_amount = {s:find_average_trans_amount_per_segment(s) for s in revenue_segments_df['seg_name'].unique()}\n",
    "\n",
    "print(segments_avg_trans_amount)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now we should find how much revenue each segment produced in the 6 month time frame."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "seg_name\n",
      "INFREQUENT    41170.04\n",
      "LOYAL         33602.19\n",
      "NEW            6673.18\n",
      "ONE-OFFS      23937.69\n",
      "VIP           41004.15\n",
      "Name: trans_total, dtype: float64\n"
     ]
    }
   ],
   "source": [
    "print(revenue_segments_df.groupby('seg_name')['trans_total'].sum())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We notice that the amount of spending from 'infrequent' and 'VIP' customers are similar but 'VIP' customers have more frequent and greater transactions than 'infrequent' customers. Let's create a scatter plot to determine any outliers."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "%matplotlib inline\n",
    "\n",
    "\n",
    "fig = plt.figure(figsize=(10,10))\n",
    "sns.scatterplot(x=revenue_segments_df['seg_name'], y=revenue_segments_df['trans_total'])\n",
    "plt.title('Transaction Totals Per Customer Segment')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We see some outliers for the 'one-offs', 'infrequent', 'loyal', and 'new' new segments. Let us filter out the upper outliers and recalculate our average number of transactions and average transaction total for each segment. We will assume anything in the 99th percentile is an outlier."
   ]
  },
  {
   "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>cust_id</th>\n",
       "      <th>seg_name</th>\n",
       "      <th>trans_id</th>\n",
       "      <th>trans_dt</th>\n",
       "      <th>trans_total</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>9085146</td>\n",
       "      <td>ONE-OFFS</td>\n",
       "      <td>1</td>\n",
       "      <td>2016-01-02 10:06:00</td>\n",
       "      <td>42.99</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>18511160</td>\n",
       "      <td>ONE-OFFS</td>\n",
       "      <td>4</td>\n",
       "      <td>2016-01-02 11:33:00</td>\n",
       "      <td>501.91</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>14769966</td>\n",
       "      <td>ONE-OFFS</td>\n",
       "      <td>7</td>\n",
       "      <td>2016-01-02 11:36:00</td>\n",
       "      <td>96.97</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>3769351</td>\n",
       "      <td>ONE-OFFS</td>\n",
       "      <td>10</td>\n",
       "      <td>2016-01-02 12:16:00</td>\n",
       "      <td>32.99</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>15088</td>\n",
       "      <td>ONE-OFFS</td>\n",
       "      <td>13</td>\n",
       "      <td>2016-01-02 12:25:00</td>\n",
       "      <td>288.94</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1210</th>\n",
       "      <td>14311317</td>\n",
       "      <td>VIP</td>\n",
       "      <td>2633</td>\n",
       "      <td>2016-06-11 10:51:00</td>\n",
       "      <td>163.96</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1213</th>\n",
       "      <td>8805421</td>\n",
       "      <td>VIP</td>\n",
       "      <td>2642</td>\n",
       "      <td>2016-06-13 00:00:00</td>\n",
       "      <td>231.96</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1215</th>\n",
       "      <td>9350306</td>\n",
       "      <td>VIP</td>\n",
       "      <td>2645</td>\n",
       "      <td>2016-06-15 12:44:00</td>\n",
       "      <td>60.99</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1218</th>\n",
       "      <td>7049122</td>\n",
       "      <td>VIP</td>\n",
       "      <td>2654</td>\n",
       "      <td>2016-06-17 19:36:00</td>\n",
       "      <td>101.97</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1219</th>\n",
       "      <td>3649704</td>\n",
       "      <td>VIP</td>\n",
       "      <td>2662</td>\n",
       "      <td>2016-06-18 10:00:00</td>\n",
       "      <td>464.91</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1207 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       cust_id  seg_name  trans_id             trans_dt  trans_total\n",
       "0      9085146  ONE-OFFS         1  2016-01-02 10:06:00        42.99\n",
       "2     18511160  ONE-OFFS         4  2016-01-02 11:33:00       501.91\n",
       "4     14769966  ONE-OFFS         7  2016-01-02 11:36:00        96.97\n",
       "6      3769351  ONE-OFFS        10  2016-01-02 12:16:00        32.99\n",
       "7        15088  ONE-OFFS        13  2016-01-02 12:25:00       288.94\n",
       "...        ...       ...       ...                  ...          ...\n",
       "1210  14311317       VIP      2633  2016-06-11 10:51:00       163.96\n",
       "1213   8805421       VIP      2642  2016-06-13 00:00:00       231.96\n",
       "1215   9350306       VIP      2645  2016-06-15 12:44:00        60.99\n",
       "1218   7049122       VIP      2654  2016-06-17 19:36:00       101.97\n",
       "1219   3649704       VIP      2662  2016-06-18 10:00:00       464.91\n",
       "\n",
       "[1207 rows x 5 columns]"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "segment_dfs = []\n",
    "\n",
    "for segment in revenue_segments_df['seg_name'].unique():\n",
    "    segment_df = revenue_segments_df[revenue_segments_df['seg_name'] == segment]\n",
    "    q_99 = segment_df['trans_total'].quantile(.99)\n",
    "    segment_df = segment_df[segment_df['trans_total'] < q_99]\n",
    "    segment_dfs.append(segment_df)\n",
    "    \n",
    "revenue_segments_cleaned_df = pd.concat(segment_dfs)\n",
    "revenue_segments_cleaned_df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now that we cleaned our dataset let us recalculate our summary statistics. \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Average number of transactions\n",
      "{'ONE-OFFS': 1.0368852459016393, 'INFREQUENT': 1.2066869300911853, 'LOYAL': 1.669064748201439, 'NEW': 1.078125, 'VIP': 3.657142857142857}\n",
      "\n",
      "\n",
      "Average amount per transaction\n",
      "{'ONE-OFFS': 89.72299180327869, 'INFREQUENT': 115.27711246200607, 'LOYAL': 225.51633093525183, 'NEW': 96.86343749999999, 'VIP': 553.6418571428571}\n",
      "\n",
      "\n",
      "Amount of revenue for each segment\n",
      "seg_name\n",
      "INFREQUENT    37926.17\n",
      "LOYAL         31346.77\n",
      "NEW            6199.26\n",
      "ONE-OFFS      21892.41\n",
      "VIP           38754.93\n",
      "Name: trans_total, dtype: float64\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#We need to readjust our functions we created to use our cleaned dataframe. \n",
    "\n",
    "def find_average_trans_per_segment(segment):\n",
    "    segment = revenue_segments_cleaned_df[revenue_segments_cleaned_df['seg_name'] == segment]\n",
    "    avg_trans = len(segment['trans_id']) / len(segment['cust_id'].unique())\n",
    "    return avg_trans\n",
    "\n",
    "\n",
    "def find_average_trans_amount_per_segment(segment):\n",
    "    segment = revenue_segments_cleaned_df[revenue_segments_cleaned_df['seg_name'] == segment]\n",
    "    avg_trans_amount = segment['trans_total'].sum() / len(segment['cust_id'].unique())\n",
    "    return avg_trans_amount\n",
    "\n",
    "\n",
    "print('Average number of transactions')\n",
    "segments_avg_trans = {s:find_average_trans_per_segment(s) for s in revenue_segments_df['seg_name'].unique()}\n",
    "print(segments_avg_trans)\n",
    "\n",
    "print('\\n')\n",
    "print('Average amount per transaction')\n",
    "segments_avg_trans_amount = {s:find_average_trans_amount_per_segment(s) for s in revenue_segments_df['seg_name'].unique()}\n",
    "print(segments_avg_trans_amount)\n",
    "\n",
    "print('\\n')\n",
    "print('Amount of revenue for each segment')\n",
    "print(revenue_segments_cleaned_df.groupby('seg_name')['trans_total'].sum())\n",
    "\n",
    "fig = plt.figure(figsize=(10,10))\n",
    "sns.scatterplot(x=revenue_segments_cleaned_df['seg_name'], y=revenue_segments_cleaned_df['trans_total'])\n",
    "plt.title('Transaction Totals Per Customer Segment')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Even though we removed the outliers we still see that 'infrequent' customers generated more revenue than 'loyal' customers. Perhaps we should find the number of transactions in each segment."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "            trans_id\n",
      "seg_name            \n",
      "INFREQUENT       397\n",
      "LOYAL            232\n",
      "NEW               69\n",
      "ONE-OFFS         253\n",
      "VIP              256\n"
     ]
    }
   ],
   "source": [
    "print(revenue_segments_cleaned_df[['trans_id', 'seg_name']].groupby('seg_name').count())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We now see that even though 'infrequent' customers spend less and visit the store less often than 'loyal' andd 'VIP' customers there was a higher number of transactions in this segment. With this in mind we can create our first slide.\n",
    "\n",
    "The first insight will be that 'infrequent' segment generated more revenue than the 'loyal' segment. I will recommend we focus marketing efforts on the 'infrequent' segment in order to grow this segment or hopefully it will lead to customers turning into 'loyal' or 'VIP customers.\n",
    "\n",
    "### Insight 2\n",
    "\n",
    "For the next insight I want to dive deeper into the products. First let us create a scatter plot with the price of the product and the amount of revenue generated. First let us pull from the database into a dataframe."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "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>cust_id</th>\n",
       "      <th>seg_name</th>\n",
       "      <th>trans_id</th>\n",
       "      <th>prod_id</th>\n",
       "      <th>category</th>\n",
       "      <th>brand</th>\n",
       "      <th>prod_name</th>\n",
       "      <th>item_qty</th>\n",
       "      <th>item_price</th>\n",
       "      <th>revenue</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>7820001</td>\n",
       "      <td>LOYAL</td>\n",
       "      <td>2009</td>\n",
       "      <td>199922</td>\n",
       "      <td>Men</td>\n",
       "      <td>N</td>\n",
       "      <td>Product 199922</td>\n",
       "      <td>1</td>\n",
       "      <td>25.99</td>\n",
       "      <td>25.99</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>13868709</td>\n",
       "      <td>INFREQUENT</td>\n",
       "      <td>1583</td>\n",
       "      <td>207344</td>\n",
       "      <td>Make up</td>\n",
       "      <td>G</td>\n",
       "      <td>Product 207344</td>\n",
       "      <td>2</td>\n",
       "      <td>15.99</td>\n",
       "      <td>31.98</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2345280</td>\n",
       "      <td>ONE-OFFS</td>\n",
       "      <td>2114</td>\n",
       "      <td>209732</td>\n",
       "      <td>Women</td>\n",
       "      <td>L</td>\n",
       "      <td>Product 209732</td>\n",
       "      <td>2</td>\n",
       "      <td>80.99</td>\n",
       "      <td>161.98</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3648035</td>\n",
       "      <td>ONE-OFFS</td>\n",
       "      <td>114</td>\n",
       "      <td>209999</td>\n",
       "      <td>Men</td>\n",
       "      <td>J</td>\n",
       "      <td>Product 209999</td>\n",
       "      <td>1</td>\n",
       "      <td>26.99</td>\n",
       "      <td>26.99</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2603787</td>\n",
       "      <td>INFREQUENT</td>\n",
       "      <td>2599</td>\n",
       "      <td>211115</td>\n",
       "      <td>Women</td>\n",
       "      <td>E</td>\n",
       "      <td>Product 211115</td>\n",
       "      <td>2</td>\n",
       "      <td>101.95</td>\n",
       "      <td>203.90</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1858</th>\n",
       "      <td>6533400</td>\n",
       "      <td>VIP</td>\n",
       "      <td>2452</td>\n",
       "      <td>398829613</td>\n",
       "      <td>Sun</td>\n",
       "      <td>J</td>\n",
       "      <td>Product 398829613</td>\n",
       "      <td>5</td>\n",
       "      <td>14.99</td>\n",
       "      <td>74.95</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1859</th>\n",
       "      <td>15098614</td>\n",
       "      <td>LOYAL</td>\n",
       "      <td>2586</td>\n",
       "      <td>403716082</td>\n",
       "      <td>Make up</td>\n",
       "      <td>L</td>\n",
       "      <td>Product 403716082</td>\n",
       "      <td>3</td>\n",
       "      <td>35.99</td>\n",
       "      <td>107.97</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1860</th>\n",
       "      <td>12423997</td>\n",
       "      <td>VIP</td>\n",
       "      <td>2581</td>\n",
       "      <td>407168795</td>\n",
       "      <td>Make up</td>\n",
       "      <td>M</td>\n",
       "      <td>Product 407168795</td>\n",
       "      <td>1</td>\n",
       "      <td>24.99</td>\n",
       "      <td>24.99</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1861</th>\n",
       "      <td>12423997</td>\n",
       "      <td>VIP</td>\n",
       "      <td>2581</td>\n",
       "      <td>407168814</td>\n",
       "      <td>Make up</td>\n",
       "      <td>M</td>\n",
       "      <td>Product 407168814</td>\n",
       "      <td>1</td>\n",
       "      <td>21.99</td>\n",
       "      <td>21.99</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1862</th>\n",
       "      <td>12788599</td>\n",
       "      <td>LOYAL</td>\n",
       "      <td>2592</td>\n",
       "      <td>407820712</td>\n",
       "      <td>Women</td>\n",
       "      <td>A</td>\n",
       "      <td>Product 407820712</td>\n",
       "      <td>1</td>\n",
       "      <td>1.99</td>\n",
       "      <td>1.99</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1863 rows × 10 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       cust_id    seg_name  trans_id    prod_id category brand  \\\n",
       "0      7820001       LOYAL      2009     199922      Men     N   \n",
       "1     13868709  INFREQUENT      1583     207344  Make up     G   \n",
       "2      2345280    ONE-OFFS      2114     209732    Women     L   \n",
       "3      3648035    ONE-OFFS       114     209999      Men     J   \n",
       "4      2603787  INFREQUENT      2599     211115    Women     E   \n",
       "...        ...         ...       ...        ...      ...   ...   \n",
       "1858   6533400         VIP      2452  398829613      Sun     J   \n",
       "1859  15098614       LOYAL      2586  403716082  Make up     L   \n",
       "1860  12423997         VIP      2581  407168795  Make up     M   \n",
       "1861  12423997         VIP      2581  407168814  Make up     M   \n",
       "1862  12788599       LOYAL      2592  407820712    Women     A   \n",
       "\n",
       "              prod_name  item_qty  item_price  revenue  \n",
       "0        Product 199922         1       25.99    25.99  \n",
       "1        Product 207344         2       15.99    31.98  \n",
       "2        Product 209732         2       80.99   161.98  \n",
       "3        Product 209999         1       26.99    26.99  \n",
       "4        Product 211115         2      101.95   203.90  \n",
       "...                 ...       ...         ...      ...  \n",
       "1858  Product 398829613         5       14.99    74.95  \n",
       "1859  Product 403716082         3       35.99   107.97  \n",
       "1860  Product 407168795         1       24.99    24.99  \n",
       "1861  Product 407168814         1       21.99    21.99  \n",
       "1862  Product 407820712         1        1.99     1.99  \n",
       "\n",
       "[1863 rows x 10 columns]"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "query_product_revenue = '''\n",
    "WITH active_segments AS (\n",
    "SELECT cust_id,\n",
    "        seg_name\n",
    "FROM segments\n",
    "WHERE active_flag = 'Y'\n",
    ")\n",
    "\n",
    "SELECT t.cust_id,\n",
    "       s.seg_name,\n",
    "       t.trans_id,\n",
    "       t.prod_id,\n",
    "       p.category,\n",
    "       p.brand,\n",
    "       p.prod_name,\n",
    "       SUM(t.item_qty) AS item_qty,\n",
    "       t.item_price,\n",
    "       SUM(t.item_qty * t.item_price) AS revenue\n",
    "FROM transactions t\n",
    "INNER JOIN active_segments s ON t.cust_id = s.cust_id\n",
    "INNER JOIN products p ON t.prod_id = p.prod_id\n",
    "GROUP BY t.prod_id\n",
    "'''\n",
    "\n",
    "\n",
    "product_revenue_df = pd.read_sql_query(query_product_revenue, conn)\n",
    "product_revenue_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 936x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(13,8))\n",
    "\n",
    "sns.scatterplot(data=product_revenue_df, x='item_price', y='revenue', hue='category')\n",
    "plt.title('Price vs Revenue', fontsize=20)\n",
    "plt.xlabel('Item Price ($)', fontsize=15)\n",
    "plt.ylabel('Revenue ($)', fontsize=15)\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "From the graph above we can make two observations:\n",
    "* There are a few make up products which have a lower price but high revenue\n",
    "* For for the Women category there seems to be for distinct groups\n",
    "\n",
    "For the purpose of this assessment let us explore the make up products with the higher revenue. If we plot just the make up products and add a trend line I expect these products to be outliers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "make_up_df = product_revenue_df[product_revenue_df['category'] == 'Make up']\n",
    "\n",
    "sns.lmplot(x='item_price', y='revenue', data=make_up_df)\n",
    "plt.title('Price vs Revenue', fontsize=20)\n",
    "plt.xlabel('Item Price ($)', fontsize=15)\n",
    "plt.ylabel('Revenue ($)', fontsize=15)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As expected there are three make up products that are far from the trendline. Let us explore these three products further."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "prod_id\n",
       "248225       602.83\n",
       "261231       449.82\n",
       "248743       399.84\n",
       "287826       351.89\n",
       "30317167     335.94\n",
       "              ...  \n",
       "222086         2.99\n",
       "114150033      2.50\n",
       "128663333      2.49\n",
       "166838310      2.45\n",
       "166838313      1.40\n",
       "Name: revenue, Length: 692, dtype: float64"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "top_make_up = make_up_df.groupby('prod_id')['revenue'].sum()\n",
    "top_make_up.sort_values(ascending=False, inplace=True)\n",
    "top_make_up"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can now take the top 3 'prod_id' and filter our makeup dataframe for these products."
   ]
  },
  {
   "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>cust_id</th>\n",
       "      <th>seg_name</th>\n",
       "      <th>trans_id</th>\n",
       "      <th>prod_id</th>\n",
       "      <th>category</th>\n",
       "      <th>brand</th>\n",
       "      <th>prod_name</th>\n",
       "      <th>item_qty</th>\n",
       "      <th>item_price</th>\n",
       "      <th>revenue</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>426</th>\n",
       "      <td>1264766</td>\n",
       "      <td>VIP</td>\n",
       "      <td>2133</td>\n",
       "      <td>248225</td>\n",
       "      <td>Make up</td>\n",
       "      <td>C</td>\n",
       "      <td>Product 248225</td>\n",
       "      <td>17</td>\n",
       "      <td>35.99</td>\n",
       "      <td>602.83</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>434</th>\n",
       "      <td>1802400</td>\n",
       "      <td>ONE-OFFS</td>\n",
       "      <td>286</td>\n",
       "      <td>248743</td>\n",
       "      <td>Make up</td>\n",
       "      <td>C</td>\n",
       "      <td>Product 248743</td>\n",
       "      <td>16</td>\n",
       "      <td>24.99</td>\n",
       "      <td>399.84</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>631</th>\n",
       "      <td>3568127</td>\n",
       "      <td>INFREQUENT</td>\n",
       "      <td>1819</td>\n",
       "      <td>261231</td>\n",
       "      <td>Make up</td>\n",
       "      <td>C</td>\n",
       "      <td>Product 261231</td>\n",
       "      <td>18</td>\n",
       "      <td>24.99</td>\n",
       "      <td>449.82</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     cust_id    seg_name  trans_id  prod_id category brand       prod_name  \\\n",
       "426  1264766         VIP      2133   248225  Make up     C  Product 248225   \n",
       "434  1802400    ONE-OFFS       286   248743  Make up     C  Product 248743   \n",
       "631  3568127  INFREQUENT      1819   261231  Make up     C  Product 261231   \n",
       "\n",
       "     item_qty  item_price  revenue  \n",
       "426        17       35.99   602.83  \n",
       "434        16       24.99   399.84  \n",
       "631        18       24.99   449.82  "
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "top_make_up = top_make_up.reset_index()\n",
    "top_make_up = top_make_up['prod_id'].head(3)\n",
    "top_make_up = top_make_up.tolist()\n",
    "\n",
    "make_up_df = make_up_df[make_up_df.prod_id.isin(top_make_up)]\n",
    "make_up_df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We see that these products that these products were outliers because each product was only purchased once in a large quantity.  This means we can not assume these products are popular or else we would be seeing multiple transactions that contain these products.\n",
    "\n",
    "### Repivot Insight 2\n",
    "\n",
    "This means we should repivot out thinking. Perhaps we should see which brands that customers seem to purchase the most. This way we can recommend to try and focus relationship efforts with that brand. We can even see the breakdown of which segments bought each brand.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>seg_name</th>\n",
       "      <th>brand</th>\n",
       "      <th>item_qty</th>\n",
       "      <th>revenue</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>INFREQUENT</td>\n",
       "      <td>A</td>\n",
       "      <td>145</td>\n",
       "      <td>2278.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>INFREQUENT</td>\n",
       "      <td>B</td>\n",
       "      <td>80</td>\n",
       "      <td>2712.83</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>INFREQUENT</td>\n",
       "      <td>C</td>\n",
       "      <td>267</td>\n",
       "      <td>10727.77</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>INFREQUENT</td>\n",
       "      <td>D</td>\n",
       "      <td>85</td>\n",
       "      <td>4548.73</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>INFREQUENT</td>\n",
       "      <td>E</td>\n",
       "      <td>47</td>\n",
       "      <td>1585.54</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>98</th>\n",
       "      <td>VIP</td>\n",
       "      <td>T</td>\n",
       "      <td>10</td>\n",
       "      <td>439.96</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99</th>\n",
       "      <td>VIP</td>\n",
       "      <td>U</td>\n",
       "      <td>33</td>\n",
       "      <td>1019.67</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>100</th>\n",
       "      <td>VIP</td>\n",
       "      <td>V</td>\n",
       "      <td>1</td>\n",
       "      <td>71.99</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>101</th>\n",
       "      <td>VIP</td>\n",
       "      <td>Y</td>\n",
       "      <td>33</td>\n",
       "      <td>1167.17</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>102</th>\n",
       "      <td>VIP</td>\n",
       "      <td>Z</td>\n",
       "      <td>3</td>\n",
       "      <td>96.70</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>103 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       seg_name brand  item_qty   revenue\n",
       "0    INFREQUENT     A       145   2278.00\n",
       "1    INFREQUENT     B        80   2712.83\n",
       "2    INFREQUENT     C       267  10727.77\n",
       "3    INFREQUENT     D        85   4548.73\n",
       "4    INFREQUENT     E        47   1585.54\n",
       "..          ...   ...       ...       ...\n",
       "98          VIP     T        10    439.96\n",
       "99          VIP     U        33   1019.67\n",
       "100         VIP     V         1     71.99\n",
       "101         VIP     Y        33   1167.17\n",
       "102         VIP     Z         3     96.70\n",
       "\n",
       "[103 rows x 4 columns]"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "query_product_revenue = '''\n",
    "WITH active_segments AS (\n",
    "SELECT cust_id,\n",
    "        seg_name\n",
    "FROM segments\n",
    "WHERE active_flag = 'Y'\n",
    ")\n",
    "\n",
    "SELECT s.seg_name,\n",
    "       p.brand,\n",
    "       SUM(t.item_qty) AS item_qty,\n",
    "       SUM(t.item_qty * t.item_price) AS revenue\n",
    "FROM transactions t\n",
    "INNER JOIN active_segments s ON t.cust_id = s.cust_id\n",
    "INNER JOIN products p ON t.prod_id = p.prod_id\n",
    "GROUP BY s.seg_name, p.brand\n",
    "'''\n",
    "\n",
    "\n",
    "product_df = pd.read_sql_query(query_product_revenue, conn)\n",
    "product_df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let us now plot the number of items bought for each brand."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(16,10))\n",
    "sns.barplot(x='brand', y='item_qty', hue='seg_name', data=product_df)\n",
    "plt.title('Items Purchased Form Each Brand', fontsize=20)\n",
    "plt.xlabel('Brand', fontsize=15)\n",
    "plt.ylabel('Items Purchased)', fontsize=15)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To build off of our recommendation from the first slide, let us see which brands 'infrequent' customers purchase from."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "brand\n",
      "C    267\n",
      "A    145\n",
      "M    142\n",
      "L     94\n",
      "D     85\n",
      "Name: item_qty, dtype: int64\n",
      "\n",
      "\n",
      "Total products purchased in the infrequent segment\n",
      "1171\n",
      "\n",
      "\n",
      "Total number of brands\n",
      "22\n"
     ]
    }
   ],
   "source": [
    "print(product_df[product_df['seg_name'] == 'INFREQUENT']\n",
    "                .groupby('brand')\n",
    "                ['item_qty'].sum()\n",
    "                .sort_values(ascending=False)\n",
    "                .head(5))\n",
    "print('\\n')\n",
    "print('Total products purchased in the infrequent segment')\n",
    "print(product_df[product_df['seg_name'] == 'INFREQUENT']['item_qty'].sum())\n",
    "print('\\n')\n",
    "print('Total number of brands')\n",
    "print(len(product_df['brand'].unique()))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Before we can create our slide let us figure out the market share in percentage for the top 5 brands."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "      seg_name brand  item_qty   revenue  market_share\n",
      "2   INFREQUENT     C       267  10727.77      0.228010\n",
      "0   INFREQUENT     A       145   2278.00      0.123826\n",
      "12  INFREQUENT     M       142   1933.60      0.121264\n",
      "11  INFREQUENT     L        94   3393.17      0.080273\n",
      "3   INFREQUENT     D        85   4548.73      0.072588\n"
     ]
    }
   ],
   "source": [
    "def find_market_share(segment):\n",
    "    segment_df = product_df[product_df['seg_name'] == segment].copy()\n",
    "    total_items_purchased = segment_df['item_qty'].sum()\n",
    "    segment_df['market_share'] = segment_df['item_qty'] / total_items_purchased\n",
    "    segment_df.sort_values('market_share', ascending=False, inplace=True)\n",
    "    return segment_df.head(5)\n",
    "\n",
    "\n",
    "print(find_market_share('INFREQUENT'))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We see that the top brands for infrequent customers are C, A, and M. We can suggest that we build a relationship with brand C. From the graph we also see that brand C holds the highest market share for the 'loyal' and 'VIP' customers. We should figure out the market share that brand C holds for these two segments to support our recommendation."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "   seg_name brand  item_qty  revenue  market_share\n",
      "24    LOYAL     C       189  7613.14      0.216247\n",
      "38    LOYAL     S        97  3735.37      0.110984\n",
      "23    LOYAL     B        75  2494.08      0.085812\n",
      "34    LOYAL     M        75  1191.19      0.085812\n",
      "22    LOYAL     A        74  1228.33      0.084668\n",
      "\n",
      "\n",
      "   seg_name brand  item_qty   revenue  market_share\n",
      "84      VIP     C       279  11486.31      0.217290\n",
      "83      VIP     B       160   5824.13      0.124611\n",
      "97      VIP     S       127   5587.24      0.098910\n",
      "86      VIP     E       105   4970.45      0.081776\n",
      "82      VIP     A       104   1048.13      0.080997\n"
     ]
    }
   ],
   "source": [
    "print(find_market_share('LOYAL'))\n",
    "print('\\n')\n",
    "print(find_market_share('VIP'))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Since brand C has such a big share of the market we should find out what category of items they are selling the most."
   ]
  },
  {
   "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>seg_name</th>\n",
       "      <th>brand</th>\n",
       "      <th>category</th>\n",
       "      <th>item_qty</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>VIP</td>\n",
       "      <td>C</td>\n",
       "      <td>Men</td>\n",
       "      <td>279</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>INFREQUENT</td>\n",
       "      <td>C</td>\n",
       "      <td>Women</td>\n",
       "      <td>267</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>LOYAL</td>\n",
       "      <td>C</td>\n",
       "      <td>Make up</td>\n",
       "      <td>189</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>ONE-OFFS</td>\n",
       "      <td>C</td>\n",
       "      <td>Make up</td>\n",
       "      <td>108</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>NEW</td>\n",
       "      <td>C</td>\n",
       "      <td>Men</td>\n",
       "      <td>41</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     seg_name brand category  item_qty\n",
       "0         VIP     C      Men       279\n",
       "1  INFREQUENT     C    Women       267\n",
       "2       LOYAL     C  Make up       189\n",
       "3    ONE-OFFS     C  Make up       108\n",
       "4         NEW     C      Men        41"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "query_brand_c = '''\n",
    "WITH active_segments AS (\n",
    "SELECT cust_id,\n",
    "        seg_name\n",
    "FROM segments\n",
    "WHERE active_flag = 'Y'\n",
    ")\n",
    "\n",
    "SELECT s.seg_name,\n",
    "       p.brand,\n",
    "       p.category,\n",
    "       SUM(t.item_qty) AS item_qty\n",
    "FROM transactions t\n",
    "INNER JOIN active_segments s ON t.cust_id = s.cust_id\n",
    "INNER JOIN products p ON t.prod_id = p.prod_id\n",
    "WHERE p.brand = 'C'\n",
    "GROUP BY s.seg_name, p.brand\n",
    "ORDER BY item_qty DESC\n",
    "'''\n",
    "\n",
    "brand_c_df = pd.read_sql_query(query_brand_c, conn)\n",
    "brand_c_df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We see that the category that brand C sells the most is 'Men' and 'Women'. With this information we can now create our slide."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Insight 3\n",
    "\n",
    "For our third insight we are going to see how long it takes a customer to go from 'infrequent' to either 'loyal' or 'VIP'.\n",
    "\n",
    "For our query we are going to assume that customers who are 'infrequent' can't revert back to 'new' or 'one-off'. We also will not include any customers who's active status is 'infrequent'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['INFREQUENT' 'LOYAL' 'VIP']\n",
      "   cust_id    seg_name            update_at\n",
      "0    11248  INFREQUENT  2016-01-01 00:00:00\n",
      "1    11248       LOYAL  2016-02-01 00:00:00\n",
      "2    12064       LOYAL  2014-01-01 00:00:00\n",
      "3    12064  INFREQUENT  2014-02-01 00:00:00\n",
      "4    12064  INFREQUENT  2014-03-01 00:00:00\n",
      "5    12064  INFREQUENT  2014-07-01 00:00:00\n",
      "6    12064  INFREQUENT  2014-09-01 00:00:00\n",
      "7    12064  INFREQUENT  2014-12-01 00:00:00\n",
      "8    12064  INFREQUENT  2015-06-01 00:00:00\n",
      "9    12064  INFREQUENT  2015-12-01 00:00:00\n"
     ]
    }
   ],
   "source": [
    "query_customers = '''\n",
    "SELECT cust_id,\n",
    "       seg_name,\n",
    "       update_at\n",
    "FROM segments\n",
    "WHERE seg_name = 'LOYAL' \n",
    "OR seg_name = 'VIP'\n",
    "OR (seg_name = 'INFREQUENT' AND active_flag != 'Y')\n",
    "ORDER BY cust_id, update_at;\n",
    "'''\n",
    "\n",
    "customer_df = pd.read_sql_query(query_customers, conn)\n",
    "print(customer_df['seg_name'].unique())\n",
    "print(customer_df.head(10))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For each customer ID we are going to find the minimum date for each segment. We are going to assume that at this date either they started off in that segment or that was the date that the segment was changed."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>cust_id</th>\n",
       "      <th>seg_name</th>\n",
       "      <th>update_at</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>11248</td>\n",
       "      <td>INFREQUENT</td>\n",
       "      <td>2016-01-01 00:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>11248</td>\n",
       "      <td>LOYAL</td>\n",
       "      <td>2016-02-01 00:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>12064</td>\n",
       "      <td>INFREQUENT</td>\n",
       "      <td>2014-02-01 00:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>12064</td>\n",
       "      <td>LOYAL</td>\n",
       "      <td>2014-01-01 00:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>66706</td>\n",
       "      <td>INFREQUENT</td>\n",
       "      <td>2014-01-01 00:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1112</th>\n",
       "      <td>20044027</td>\n",
       "      <td>INFREQUENT</td>\n",
       "      <td>2016-01-01 00:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1113</th>\n",
       "      <td>20132046</td>\n",
       "      <td>VIP</td>\n",
       "      <td>2016-02-01 00:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1114</th>\n",
       "      <td>20259860</td>\n",
       "      <td>LOYAL</td>\n",
       "      <td>2016-02-01 00:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1115</th>\n",
       "      <td>20330348</td>\n",
       "      <td>LOYAL</td>\n",
       "      <td>2016-03-01 00:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1116</th>\n",
       "      <td>20356677</td>\n",
       "      <td>VIP</td>\n",
       "      <td>2016-03-01 00:00:00</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1117 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       cust_id    seg_name            update_at\n",
       "0        11248  INFREQUENT  2016-01-01 00:00:00\n",
       "1        11248       LOYAL  2016-02-01 00:00:00\n",
       "2        12064  INFREQUENT  2014-02-01 00:00:00\n",
       "3        12064       LOYAL  2014-01-01 00:00:00\n",
       "4        66706  INFREQUENT  2014-01-01 00:00:00\n",
       "...        ...         ...                  ...\n",
       "1112  20044027  INFREQUENT  2016-01-01 00:00:00\n",
       "1113  20132046         VIP  2016-02-01 00:00:00\n",
       "1114  20259860       LOYAL  2016-02-01 00:00:00\n",
       "1115  20330348       LOYAL  2016-03-01 00:00:00\n",
       "1116  20356677         VIP  2016-03-01 00:00:00\n",
       "\n",
       "[1117 rows x 3 columns]"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "customer_pt = pd.pivot_table(customer_df, values=['update_at'], index=['cust_id', 'seg_name'], aggfunc=min)\n",
    "customer_pt = customer_pt.reset_index()\n",
    "customer_pt"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We see in this pivot table that for each customer the `seg_name` column starts with the 'infrequent' value first. In order to find how long a customer went from 'infrequent' to either 'loyal' or 'VIP' we can subtract the date associated with those two segments from the date associated with 'infrequent'\n",
    "\n",
    "If we get a positive number that means a customer moved up from 'infrequent' since the date that they became 'loyal' or 'VIP' should be greater than the date they were labeled as 'infrequent'. This means if we get a negative number then that means the customer actually moved down to 'infrequent'.\n",
    "\n",
    "First we should check if the `update_at` column is labeled as datetime."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "cust_id       int64\n",
      "seg_name     object\n",
      "update_at    object\n",
      "dtype: object\n"
     ]
    }
   ],
   "source": [
    "print(customer_pt.dtypes)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "cust_id               int64\n",
      "seg_name             object\n",
      "update_at    datetime64[ns]\n",
      "dtype: object\n"
     ]
    }
   ],
   "source": [
    "#Since the update_at column is not a datetime object we need to change it as such\n",
    "\n",
    "customer_pt['update_at'] = pd.to_datetime(customer_pt['update_at'])\n",
    "print(customer_pt.dtypes)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Example of a customer moving up from infrequent\n",
      "   cust_id    seg_name  update_at\n",
      "0    11248  INFREQUENT 2016-01-01\n",
      "1    11248       LOYAL 2016-02-01\n",
      "Notice that the time difference is positive\n",
      "31 days 00:00:00\n",
      "\n",
      "\n",
      "Example of a customer moving down into infrequent\n",
      "   cust_id    seg_name  update_at\n",
      "2    12064  INFREQUENT 2014-02-01\n",
      "3    12064       LOYAL 2014-01-01\n",
      "Notice that the time difference is negative\n",
      "-31 days +00:00:00\n"
     ]
    }
   ],
   "source": [
    "#Now that the update_at column is a datetime object we can continue\n",
    "#Below is an example of what the difference in times will look like\n",
    "\n",
    "import datetime as dt\n",
    "\n",
    "print('Example of a customer moving up from infrequent')\n",
    "cust = customer_pt[customer_pt['cust_id'] == 11248]\n",
    "print(cust)\n",
    "date = cust.iloc[1,2] - cust.iloc[0,2]\n",
    "print('Notice that the time difference is positive')\n",
    "print(date)\n",
    "\n",
    "print('\\n')\n",
    "print('Example of a customer moving down into infrequent')\n",
    "cust = customer_pt[customer_pt['cust_id'] == 12064]\n",
    "print(cust)\n",
    "date = cust.iloc[1,2] - cust.iloc[0,2]\n",
    "print('Notice that the time difference is negative')\n",
    "print(date)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[Timedelta('31 days 00:00:00'), Timedelta('-31 days +00:00:00'), Timedelta('334 days 00:00:00'), Timedelta('365 days 00:00:00'), Timedelta('700 days 00:00:00'), Timedelta('150 days 00:00:00'), Timedelta('-153 days +00:00:00'), Timedelta('-304 days +00:00:00'), Timedelta('365 days 00:00:00'), Timedelta('151 days 00:00:00'), Timedelta('424 days 00:00:00'), Timedelta('-59 days +00:00:00'), Timedelta('396 days 00:00:00'), Timedelta('61 days 00:00:00'), Timedelta('-90 days +00:00:00'), Timedelta('273 days 00:00:00'), Timedelta('577 days 00:00:00'), Timedelta('-31 days +00:00:00'), Timedelta('671 days 00:00:00'), Timedelta('-90 days +00:00:00'), Timedelta('-151 days +00:00:00'), Timedelta('638 days 00:00:00'), Timedelta('-546 days +00:00:00'), Timedelta('59 days 00:00:00'), Timedelta('424 days 00:00:00'), Timedelta('90 days 00:00:00'), Timedelta('-120 days +00:00:00'), Timedelta('-90 days +00:00:00'), Timedelta('212 days 00:00:00'), Timedelta('365 days 00:00:00'), Timedelta('-212 days +00:00:00'), Timedelta('424 days 00:00:00'), Timedelta('-273 days +00:00:00'), Timedelta('243 days 00:00:00'), Timedelta('-730 days +00:00:00'), Timedelta('-214 days +00:00:00'), Timedelta('-151 days +00:00:00'), Timedelta('-212 days +00:00:00'), Timedelta('212 days 00:00:00'), Timedelta('546 days 00:00:00'), Timedelta('424 days 00:00:00'), Timedelta('-151 days +00:00:00'), Timedelta('215 days 00:00:00'), Timedelta('365 days 00:00:00'), Timedelta('-120 days +00:00:00'), Timedelta('-181 days +00:00:00'), Timedelta('-92 days +00:00:00'), Timedelta('761 days 00:00:00'), Timedelta('89 days 00:00:00'), Timedelta('212 days 00:00:00'), Timedelta('-273 days +00:00:00'), Timedelta('-31 days +00:00:00'), Timedelta('-243 days +00:00:00'), Timedelta('549 days 00:00:00'), Timedelta('-59 days +00:00:00'), Timedelta('-90 days +00:00:00'), Timedelta('31 days 00:00:00'), Timedelta('212 days 00:00:00'), Timedelta('699 days 00:00:00'), Timedelta('-212 days +00:00:00'), Timedelta('-61 days +00:00:00'), Timedelta('730 days 00:00:00'), Timedelta('-151 days +00:00:00'), Timedelta('181 days 00:00:00'), Timedelta('151 days 00:00:00'), Timedelta('-31 days +00:00:00'), Timedelta('702 days 00:00:00'), Timedelta('485 days 00:00:00'), Timedelta('-181 days +00:00:00'), Timedelta('-59 days +00:00:00'), Timedelta('-30 days +00:00:00'), Timedelta('-90 days +00:00:00'), Timedelta('-61 days +00:00:00'), Timedelta('92 days 00:00:00'), Timedelta('365 days 00:00:00'), Timedelta('184 days 00:00:00'), Timedelta('-181 days +00:00:00'), Timedelta('427 days 00:00:00'), Timedelta('-59 days +00:00:00'), Timedelta('-396 days +00:00:00'), Timedelta('-181 days +00:00:00'), Timedelta('31 days 00:00:00'), Timedelta('120 days 00:00:00'), Timedelta('365 days 00:00:00'), Timedelta('181 days 00:00:00'), Timedelta('304 days 00:00:00'), Timedelta('-365 days +00:00:00'), Timedelta('396 days 00:00:00'), Timedelta('-31 days +00:00:00'), Timedelta('-334 days +00:00:00'), Timedelta('-212 days +00:00:00'), Timedelta('334 days 00:00:00'), Timedelta('455 days 00:00:00'), Timedelta('-120 days +00:00:00'), Timedelta('91 days 00:00:00'), Timedelta('-151 days +00:00:00'), Timedelta('424 days 00:00:00'), Timedelta('-334 days +00:00:00'), Timedelta('90 days 00:00:00'), Timedelta('-122 days +00:00:00'), Timedelta('153 days 00:00:00'), Timedelta('-151 days +00:00:00'), Timedelta('365 days 00:00:00'), Timedelta('-120 days +00:00:00'), Timedelta('181 days 00:00:00'), Timedelta('151 days 00:00:00'), Timedelta('120 days 00:00:00'), Timedelta('151 days 00:00:00'), Timedelta('243 days 00:00:00'), Timedelta('-31 days +00:00:00'), Timedelta('-396 days +00:00:00'), Timedelta('59 days 00:00:00'), Timedelta('396 days 00:00:00'), Timedelta('-122 days +00:00:00'), Timedelta('212 days 00:00:00'), Timedelta('-151 days +00:00:00'), Timedelta('761 days 00:00:00'), Timedelta('151 days 00:00:00'), Timedelta('761 days 00:00:00'), Timedelta('29 days 00:00:00'), Timedelta('-212 days +00:00:00'), Timedelta('-31 days +00:00:00'), Timedelta('-59 days +00:00:00'), Timedelta('-335 days +00:00:00'), Timedelta('365 days 00:00:00'), Timedelta('31 days 00:00:00'), Timedelta('-851 days +00:00:00'), Timedelta('334 days 00:00:00'), Timedelta('151 days 00:00:00'), Timedelta('396 days 00:00:00'), Timedelta('-181 days +00:00:00'), Timedelta('151 days 00:00:00'), Timedelta('485 days 00:00:00'), Timedelta('-546 days +00:00:00'), Timedelta('396 days 00:00:00'), Timedelta('-608 days +00:00:00'), Timedelta('120 days 00:00:00'), Timedelta('424 days 00:00:00'), Timedelta('-31 days +00:00:00'), Timedelta('59 days 00:00:00'), Timedelta('273 days 00:00:00'), Timedelta('184 days 00:00:00'), Timedelta('120 days 00:00:00'), Timedelta('-243 days +00:00:00'), Timedelta('31 days 00:00:00'), Timedelta('-31 days +00:00:00'), Timedelta('-31 days +00:00:00'), Timedelta('396 days 00:00:00'), Timedelta('485 days 00:00:00'), Timedelta('-365 days +00:00:00'), Timedelta('-455 days +00:00:00'), Timedelta('92 days 00:00:00'), Timedelta('62 days 00:00:00'), Timedelta('122 days 00:00:00'), Timedelta('183 days 00:00:00'), Timedelta('90 days 00:00:00'), Timedelta('334 days 00:00:00'), Timedelta('761 days 00:00:00'), Timedelta('365 days 00:00:00'), Timedelta('275 days 00:00:00'), Timedelta('-243 days +00:00:00'), Timedelta('-669 days +00:00:00'), Timedelta('-546 days +00:00:00'), Timedelta('212 days 00:00:00'), Timedelta('181 days 00:00:00'), Timedelta('730 days 00:00:00'), Timedelta('-365 days +00:00:00'), Timedelta('-396 days +00:00:00'), Timedelta('-485 days +00:00:00'), Timedelta('59 days 00:00:00'), Timedelta('151 days 00:00:00'), Timedelta('-151 days +00:00:00'), Timedelta('92 days 00:00:00'), Timedelta('-120 days +00:00:00'), Timedelta('151 days 00:00:00'), Timedelta('243 days 00:00:00'), Timedelta('730 days 00:00:00'), Timedelta('485 days 00:00:00'), Timedelta('-821 days +00:00:00'), Timedelta('365 days 00:00:00'), Timedelta('669 days 00:00:00'), Timedelta('59 days 00:00:00'), Timedelta('-577 days +00:00:00'), Timedelta('579 days 00:00:00'), Timedelta('-31 days +00:00:00'), Timedelta('424 days 00:00:00'), Timedelta('120 days 00:00:00'), Timedelta('273 days 00:00:00'), Timedelta('151 days 00:00:00'), Timedelta('31 days 00:00:00'), Timedelta('730 days 00:00:00'), Timedelta('151 days 00:00:00'), Timedelta('-273 days +00:00:00'), Timedelta('-640 days +00:00:00'), Timedelta('607 days 00:00:00'), Timedelta('30 days 00:00:00'), Timedelta('-334 days +00:00:00'), Timedelta('214 days 00:00:00'), Timedelta('-485 days +00:00:00'), Timedelta('214 days 00:00:00'), Timedelta('61 days 00:00:00'), Timedelta('61 days 00:00:00'), Timedelta('334 days 00:00:00'), Timedelta('702 days 00:00:00'), Timedelta('212 days 00:00:00'), Timedelta('151 days 00:00:00'), Timedelta('-334 days +00:00:00'), Timedelta('-851 days +00:00:00'), Timedelta('-243 days +00:00:00'), Timedelta('455 days 00:00:00'), Timedelta('-151 days +00:00:00'), Timedelta('-151 days +00:00:00'), Timedelta('275 days 00:00:00'), Timedelta('59 days 00:00:00'), Timedelta('59 days 00:00:00'), Timedelta('-150 days +00:00:00'), Timedelta('-546 days +00:00:00'), Timedelta('61 days 00:00:00'), Timedelta('-699 days +00:00:00'), Timedelta('-151 days +00:00:00'), Timedelta('-181 days +00:00:00'), Timedelta('761 days 00:00:00'), Timedelta('-31 days +00:00:00'), Timedelta('546 days 00:00:00'), Timedelta('516 days 00:00:00'), Timedelta('90 days 00:00:00'), Timedelta('-638 days +00:00:00'), Timedelta('-334 days +00:00:00'), Timedelta('-59 days +00:00:00'), Timedelta('-546 days +00:00:00'), Timedelta('485 days 00:00:00'), Timedelta('-31 days +00:00:00'), Timedelta('306 days 00:00:00'), Timedelta('-59 days +00:00:00'), Timedelta('455 days 00:00:00'), Timedelta('305 days 00:00:00'), Timedelta('120 days 00:00:00'), Timedelta('-31 days +00:00:00'), Timedelta('122 days 00:00:00'), Timedelta('151 days 00:00:00'), Timedelta('245 days 00:00:00'), Timedelta('-243 days +00:00:00'), Timedelta('31 days 00:00:00'), Timedelta('-424 days +00:00:00'), Timedelta('365 days 00:00:00'), Timedelta('31 days 00:00:00'), Timedelta('487 days 00:00:00'), Timedelta('-59 days +00:00:00'), Timedelta('730 days 00:00:00'), Timedelta('181 days 00:00:00'), Timedelta('273 days 00:00:00'), Timedelta('-424 days +00:00:00'), Timedelta('485 days 00:00:00'), Timedelta('-273 days +00:00:00'), Timedelta('-485 days +00:00:00'), Timedelta('-669 days +00:00:00'), Timedelta('-424 days +00:00:00'), Timedelta('62 days 00:00:00'), Timedelta('-334 days +00:00:00'), Timedelta('396 days 00:00:00'), Timedelta('-365 days +00:00:00'), Timedelta('-304 days +00:00:00'), Timedelta('61 days 00:00:00'), Timedelta('59 days 00:00:00'), Timedelta('610 days 00:00:00'), Timedelta('120 days 00:00:00'), Timedelta('59 days 00:00:00'), Timedelta('-90 days +00:00:00'), Timedelta('516 days 00:00:00'), Timedelta('59 days 00:00:00'), Timedelta('-365 days +00:00:00'), Timedelta('-638 days +00:00:00'), Timedelta('31 days 00:00:00'), Timedelta('-396 days +00:00:00'), Timedelta('-365 days +00:00:00'), Timedelta('-31 days +00:00:00'), Timedelta('-62 days +00:00:00'), Timedelta('31 days 00:00:00'), Timedelta('761 days 00:00:00'), Timedelta('761 days 00:00:00'), Timedelta('123 days 00:00:00'), Timedelta('518 days 00:00:00'), Timedelta('31 days 00:00:00'), Timedelta('59 days 00:00:00'), Timedelta('-396 days +00:00:00'), Timedelta('31 days 00:00:00'), Timedelta('120 days 00:00:00'), Timedelta('-90 days +00:00:00'), Timedelta('31 days 00:00:00'), Timedelta('304 days 00:00:00'), Timedelta('61 days 00:00:00'), Timedelta('-120 days +00:00:00'), Timedelta('-577 days +00:00:00'), Timedelta('882 days 00:00:00'), Timedelta('-608 days +00:00:00'), Timedelta('-120 days +00:00:00'), Timedelta('304 days 00:00:00'), Timedelta('-62 days +00:00:00'), Timedelta('31 days 00:00:00'), Timedelta('276 days 00:00:00'), Timedelta('396 days 00:00:00'), Timedelta('90 days 00:00:00'), Timedelta('243 days 00:00:00'), Timedelta('730 days 00:00:00'), Timedelta('-90 days +00:00:00'), Timedelta('579 days 00:00:00'), Timedelta('761 days 00:00:00'), Timedelta('243 days 00:00:00'), Timedelta('243 days 00:00:00'), Timedelta('-90 days +00:00:00'), Timedelta('31 days 00:00:00'), Timedelta('59 days 00:00:00'), Timedelta('-153 days +00:00:00'), Timedelta('-59 days +00:00:00'), Timedelta('396 days 00:00:00'), Timedelta('-59 days +00:00:00'), Timedelta('31 days 00:00:00'), Timedelta('-334 days +00:00:00'), Timedelta('61 days 00:00:00'), Timedelta('821 days 00:00:00'), Timedelta('610 days 00:00:00'), Timedelta('59 days 00:00:00'), Timedelta('-30 days +00:00:00'), Timedelta('243 days 00:00:00'), Timedelta('365 days 00:00:00'), Timedelta('669 days 00:00:00'), Timedelta('-304 days +00:00:00'), Timedelta('181 days 00:00:00'), Timedelta('335 days 00:00:00'), Timedelta('365 days 00:00:00'), Timedelta('184 days 00:00:00'), Timedelta('-151 days +00:00:00'), Timedelta('-546 days +00:00:00'), Timedelta('-31 days +00:00:00'), Timedelta('485 days 00:00:00'), Timedelta('-120 days +00:00:00'), Timedelta('59 days 00:00:00'), Timedelta('-90 days +00:00:00'), Timedelta('153 days 00:00:00'), Timedelta('-59 days +00:00:00'), Timedelta('365 days 00:00:00'), Timedelta('-393 days +00:00:00'), Timedelta('120 days 00:00:00'), Timedelta('90 days 00:00:00'), Timedelta('59 days 00:00:00'), Timedelta('-365 days +00:00:00'), Timedelta('120 days 00:00:00'), Timedelta('31 days 00:00:00'), Timedelta('-334 days +00:00:00'), Timedelta('89 days 00:00:00'), Timedelta('123 days 00:00:00'), Timedelta('-214 days +00:00:00'), Timedelta('122 days 00:00:00'), Timedelta('62 days 00:00:00'), Timedelta('243 days 00:00:00'), Timedelta('91 days 00:00:00'), Timedelta('123 days 00:00:00'), Timedelta('61 days 00:00:00'), Timedelta('30 days 00:00:00'), Timedelta('92 days 00:00:00'), Timedelta('61 days 00:00:00')]\n"
     ]
    }
   ],
   "source": [
    "date_changes = []\n",
    "\n",
    "for cust in customer_pt['cust_id'].unique():\n",
    "    cust_df = customer_pt[customer_pt['cust_id'] == cust]\n",
    "    #This line checks to see if the customer has moved segments\n",
    "    if len(cust_df) > 1:\n",
    "        date_change = cust_df.iloc[1,2] - cust_df.iloc[0,2]\n",
    "        date_changes.append(date_change)\n",
    "        \n",
    "print(date_changes)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let us now examine how many customers have moved from or to the 'infrequent' segment."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The total number of customers that moved into a new segment is 361.\n",
      "True     218\n",
      "False    143\n",
      "dtype: int64\n"
     ]
    }
   ],
   "source": [
    "#Positive timedeltas indicate a move out of infrequent\n",
    "date_changes = pd.Series(date_changes)\n",
    "num_of_changes = len(date_changes)\n",
    "print('The total number of customers that moved into a new segment is {}.'.format(num_of_changes))\n",
    "cust_timedeltas = date_changes > pd.Timedelta(0)\n",
    "num_of_changes = cust_timedeltas.value_counts()\n",
    "print(num_of_changes)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We see there were a total of 218 customers that moved out of the 'infrequent' segment and 143 customers that moved into the 'infrequent' segment. Now we need to find the average time that it takes for a customer to move."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "280 days 10:01:06.055045\n",
      "-250 days +12:15:06.293706\n"
     ]
    }
   ],
   "source": [
    "moved_up = date_changes[cust_timedeltas]\n",
    "avg_days_moved_up = moved_up.mean()\n",
    "print(avg_days_moved_up)\n",
    "moved_down = date_changes[~cust_timedeltas]\n",
    "avg_days_moved_down = moved_down.mean()\n",
    "print(avg_days_moved_down)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We see that the average time it took for a customer to move up from 'infrequent' was 280 days or 9 months. The average time it took for a customer to move down into 'infrequent' was 250 days or 8 months. We need to figure out the date range for the `update_at` column and then we can create our slide."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2014-01-01 00:00:00\n",
      "2016-06-01 00:00:00\n"
     ]
    }
   ],
   "source": [
    "print(customer_df['update_at'].min())\n",
    "print(customer_df['update_at'].max())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Next Steps / Conclusion\n",
    "\n",
    "Next steps would be seeing if we can obtain more demographic data about our customers and more historic data about transactions so we could create a machine learning model that could forecast possible sales.\n",
    "\n",
    "Thank you for reading through this notebook. If you have any questions I would be happy to answer them at the next stage of the interview process. Thank you once again and I hope to hear from you soon."
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}