4.Feature_Selection/Week_9.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "86845900",
   "metadata": {},
   "source": [
    "# Week9"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from sklearn.preprocessing import StandardScaler"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "719c0519",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>sku</th>\n",
       "      <th>cost</th>\n",
       "      <th>retail</th>\n",
       "      <th>stype</th>\n",
       "      <th>quantity</th>\n",
       "      <th>orgprice</th>\n",
       "      <th>amt</th>\n",
       "      <th>style</th>\n",
       "      <th>color</th>\n",
       "      <th>size</th>\n",
       "      <th>packsize</th>\n",
       "      <th>vendor</th>\n",
       "      <th>brand</th>\n",
       "      <th>standardized_color</th>\n",
       "      <th>standardized_size</th>\n",
       "      <th>standardized_style</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>8936935</td>\n",
       "      <td>9.58</td>\n",
       "      <td>6.00</td>\n",
       "      <td>P</td>\n",
       "      <td>1</td>\n",
       "      <td>24.0</td>\n",
       "      <td>24.00</td>\n",
       "      <td>3592413641</td>\n",
       "      <td>WHEATNUBU</td>\n",
       "      <td>L</td>\n",
       "      <td>1</td>\n",
       "      <td>914938</td>\n",
       "      <td>KIDS HEA</td>\n",
       "      <td>Other</td>\n",
       "      <td>Standard Sizes</td>\n",
       "      <td>Other</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>8940963</td>\n",
       "      <td>4.31</td>\n",
       "      <td>2.75</td>\n",
       "      <td>P</td>\n",
       "      <td>1</td>\n",
       "      <td>11.0</td>\n",
       "      <td>8.25</td>\n",
       "      <td>005U61516</td>\n",
       "      <td>GREYHE</td>\n",
       "      <td>L</td>\n",
       "      <td>3</td>\n",
       "      <td>4012768</td>\n",
       "      <td>TOMMY HI</td>\n",
       "      <td>Grey</td>\n",
       "      <td>Standard Sizes</td>\n",
       "      <td>Other</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>8944045</td>\n",
       "      <td>28.05</td>\n",
       "      <td>17.00</td>\n",
       "      <td>P</td>\n",
       "      <td>1</td>\n",
       "      <td>68.0</td>\n",
       "      <td>29.99</td>\n",
       "      <td>CC412</td>\n",
       "      <td>MULTI</td>\n",
       "      <td>M</td>\n",
       "      <td>1</td>\n",
       "      <td>8043525</td>\n",
       "      <td>KORET OF</td>\n",
       "      <td>Multi</td>\n",
       "      <td>Standard Sizes</td>\n",
       "      <td>Other</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>8944045</td>\n",
       "      <td>25.00</td>\n",
       "      <td>14.50</td>\n",
       "      <td>P</td>\n",
       "      <td>1</td>\n",
       "      <td>58.0</td>\n",
       "      <td>29.00</td>\n",
       "      <td>CC412</td>\n",
       "      <td>MULTI</td>\n",
       "      <td>M</td>\n",
       "      <td>1</td>\n",
       "      <td>8043525</td>\n",
       "      <td>KORET OF</td>\n",
       "      <td>Multi</td>\n",
       "      <td>Standard Sizes</td>\n",
       "      <td>Other</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>8972921</td>\n",
       "      <td>14.72</td>\n",
       "      <td>17.50</td>\n",
       "      <td>P</td>\n",
       "      <td>1</td>\n",
       "      <td>35.0</td>\n",
       "      <td>35.00</td>\n",
       "      <td>TPILTRAQL</td>\n",
       "      <td>TERRA</td>\n",
       "      <td>BOU PIL</td>\n",
       "      <td>2</td>\n",
       "      <td>619402</td>\n",
       "      <td>STATUS Q</td>\n",
       "      <td>Other</td>\n",
       "      <td>Standard Sizes</td>\n",
       "      <td>Other</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       sku   cost  retail stype  quantity  orgprice    amt       style  \\\n",
       "0  8936935   9.58    6.00     P         1      24.0  24.00  3592413641   \n",
       "1  8940963   4.31    2.75     P         1      11.0   8.25   005U61516   \n",
       "2  8944045  28.05   17.00     P         1      68.0  29.99       CC412   \n",
       "3  8944045  25.00   14.50     P         1      58.0  29.00       CC412   \n",
       "4  8972921  14.72   17.50     P         1      35.0  35.00   TPILTRAQL   \n",
       "\n",
       "       color     size  packsize   vendor     brand standardized_color  \\\n",
       "0  WHEATNUBU        L         1   914938  KIDS HEA              Other   \n",
       "1     GREYHE        L         3  4012768  TOMMY HI               Grey   \n",
       "2      MULTI        M         1  8043525  KORET OF              Multi   \n",
       "3      MULTI        M         1  8043525  KORET OF              Multi   \n",
       "4      TERRA  BOU PIL         2   619402  STATUS Q              Other   \n",
       "\n",
       "  standardized_size standardized_style  \n",
       "0    Standard Sizes              Other  \n",
       "1    Standard Sizes              Other  \n",
       "2    Standard Sizes              Other  \n",
       "3    Standard Sizes              Other  \n",
       "4    Standard Sizes              Other  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "joined_drop = pd.read_pickle('../data/week8.pkl')\n",
    "joined_drop.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "35571046",
   "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>sku</th>\n",
       "      <th>retail</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>45338</th>\n",
       "      <td>994478</td>\n",
       "      <td>47677.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23027</th>\n",
       "      <td>5528349</td>\n",
       "      <td>35090.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32307</th>\n",
       "      <td>7388142</td>\n",
       "      <td>26491.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12551</th>\n",
       "      <td>3524026</td>\n",
       "      <td>23023.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22855</th>\n",
       "      <td>5494617</td>\n",
       "      <td>19221.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6024</th>\n",
       "      <td>224210</td>\n",
       "      <td>0.02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3963</th>\n",
       "      <td>1831106</td>\n",
       "      <td>0.01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34764</th>\n",
       "      <td>7858235</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7289</th>\n",
       "      <td>249386</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12325</th>\n",
       "      <td>3474289</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>45637 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "           sku    retail\n",
       "45338   994478  47677.00\n",
       "23027  5528349  35090.00\n",
       "32307  7388142  26491.00\n",
       "12551  3524026  23023.00\n",
       "22855  5494617  19221.00\n",
       "...        ...       ...\n",
       "6024    224210      0.02\n",
       "3963   1831106      0.01\n",
       "34764  7858235      0.00\n",
       "7289    249386      0.00\n",
       "12325  3474289      0.00\n",
       "\n",
       "[45637 rows x 2 columns]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "grouped_data = joined_drop.groupby('sku')['retail'].sum().reset_index()\n",
    "sorted_data = grouped_data.sort_values(by='retail', ascending=False)\n",
    "sorted_data"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9b3589e2",
   "metadata": {},
   "source": [
    "## Feature Engineering"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9f840957",
   "metadata": {},
   "source": [
    "we decide to select the following features:<br>\n",
    "and they could be divided into 2 categories\n",
    "\n",
    "categorical:\n",
    "- standardized_style\n",
    "- standardized_color\n",
    "- standardized_size\n",
    "- vendor\n",
    "- brand\n",
    "\n",
    "numerical:\n",
    "- retail price\n",
    "- packsize"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "31780a81",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "vendor\n",
       "5511283    39210\n",
       "113645     16114\n",
       "13031       6370\n",
       "3626213     5331\n",
       "13396       4196\n",
       "           ...  \n",
       "21131          1\n",
       "11536          1\n",
       "8016222        1\n",
       "2913323        1\n",
       "1713319        1\n",
       "Name: count, Length: 932, dtype: int64"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "joined_drop['vendor'].value_counts() "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "042b9411",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "932"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(joined_drop['vendor'].value_counts())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "35ea9e0c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "brand\n",
       "CLINIQUE    39210\n",
       "LANCOME     16942\n",
       "CABERNET    14782\n",
       "LIZ CLAI     8889\n",
       "LEVI STR     5638\n",
       "            ...  \n",
       "ADG/DESI        1\n",
       "SUMMER S        1\n",
       "WINSOME         1\n",
       "BROOKINS        1\n",
       "CONIFER         1\n",
       "Name: count, Length: 678, dtype: int64"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "joined_drop['brand'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "23aeced8",
   "metadata": {},
   "source": [
    "standardize retail price and packsize"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "2b467c9f",
   "metadata": {},
   "outputs": [],
   "source": [
    "# StandardScaler will normalize the features (each column of X, INDIVIDUALLY !!!) so that each column/feature/variable will have mean = 0 and standard deviation = 1.\n",
    "scaler = StandardScaler()\n",
    "\n",
    "# Scale the 'retail' and 'packsize' columns\n",
    "joined_drop[['retail', 'packsize']] = scaler.fit_transform(joined_drop[['retail', 'packsize']])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "547140e7",
   "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>retail</th>\n",
       "      <th>packsize</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-0.603488</td>\n",
       "      <td>-0.225073</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-0.730408</td>\n",
       "      <td>0.081489</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-0.173912</td>\n",
       "      <td>-0.225073</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.271543</td>\n",
       "      <td>-0.225073</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-0.154386</td>\n",
       "      <td>-0.071792</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>220727</th>\n",
       "      <td>3.047903</td>\n",
       "      <td>-0.225073</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>220728</th>\n",
       "      <td>3.047903</td>\n",
       "      <td>-0.225073</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>220729</th>\n",
       "      <td>3.047903</td>\n",
       "      <td>-0.225073</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>220730</th>\n",
       "      <td>3.047903</td>\n",
       "      <td>-0.225073</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>220731</th>\n",
       "      <td>3.047903</td>\n",
       "      <td>-0.225073</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>220276 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          retail  packsize\n",
       "0      -0.603488 -0.225073\n",
       "1      -0.730408  0.081489\n",
       "2      -0.173912 -0.225073\n",
       "3      -0.271543 -0.225073\n",
       "4      -0.154386 -0.071792\n",
       "...          ...       ...\n",
       "220727  3.047903 -0.225073\n",
       "220728  3.047903 -0.225073\n",
       "220729  3.047903 -0.225073\n",
       "220730  3.047903 -0.225073\n",
       "220731  3.047903 -0.225073\n",
       "\n",
       "[220276 rows x 2 columns]"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "joined_drop[['retail','packsize']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "f48dc236",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Create a histogram for the 'retail' column\n",
    "plt.hist(joined_drop['retail'], bins='auto')  # 'auto' lets matplotlib decide the number of bins\n",
    "plt.title('Histogram of Retail Prices')\n",
    "plt.xlabel('Retail Price')\n",
    "plt.ylabel('Frequency')\n",
    "\n",
    "# Show the plot\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6dd9bb3f",
   "metadata": {},
   "source": [
    "I don't think we need to drop ourliers here because it is normal that Dillard sells a wide range of packsize and retail prices."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1048a4e1",
   "metadata": {},
   "source": [
    "Now 'retail' and 'packsize' have been standardized."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "311f8c66",
   "metadata": {},
   "source": [
    "proceed with only selected features for x matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "33c4e1d3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "stype\n",
       "P    211919\n",
       "R      8357\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "joined_drop.stype.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "e51c2ca8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3.9434878420528596"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "return_percentage=(8357/211919)*100\n",
    "return_percentage"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "d41dba02",
   "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>sku</th>\n",
       "      <th>cost</th>\n",
       "      <th>retail</th>\n",
       "      <th>stype</th>\n",
       "      <th>quantity</th>\n",
       "      <th>orgprice</th>\n",
       "      <th>amt</th>\n",
       "      <th>style</th>\n",
       "      <th>color</th>\n",
       "      <th>size</th>\n",
       "      <th>packsize</th>\n",
       "      <th>vendor</th>\n",
       "      <th>brand</th>\n",
       "      <th>standardized_color</th>\n",
       "      <th>standardized_size</th>\n",
       "      <th>standardized_style</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>8936935</td>\n",
       "      <td>9.58</td>\n",
       "      <td>-0.603488</td>\n",
       "      <td>P</td>\n",
       "      <td>1</td>\n",
       "      <td>24.0</td>\n",
       "      <td>24.00</td>\n",
       "      <td>3592413641</td>\n",
       "      <td>WHEATNUBU</td>\n",
       "      <td>L</td>\n",
       "      <td>-0.225073</td>\n",
       "      <td>914938</td>\n",
       "      <td>KIDS HEA</td>\n",
       "      <td>Other</td>\n",
       "      <td>Standard Sizes</td>\n",
       "      <td>Other</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>8940963</td>\n",
       "      <td>4.31</td>\n",
       "      <td>-0.730408</td>\n",
       "      <td>P</td>\n",
       "      <td>1</td>\n",
       "      <td>11.0</td>\n",
       "      <td>8.25</td>\n",
       "      <td>005U61516</td>\n",
       "      <td>GREYHE</td>\n",
       "      <td>L</td>\n",
       "      <td>0.081489</td>\n",
       "      <td>4012768</td>\n",
       "      <td>TOMMY HI</td>\n",
       "      <td>Grey</td>\n",
       "      <td>Standard Sizes</td>\n",
       "      <td>Other</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>8944045</td>\n",
       "      <td>28.05</td>\n",
       "      <td>-0.173912</td>\n",
       "      <td>P</td>\n",
       "      <td>1</td>\n",
       "      <td>68.0</td>\n",
       "      <td>29.99</td>\n",
       "      <td>CC412</td>\n",
       "      <td>MULTI</td>\n",
       "      <td>M</td>\n",
       "      <td>-0.225073</td>\n",
       "      <td>8043525</td>\n",
       "      <td>KORET OF</td>\n",
       "      <td>Multi</td>\n",
       "      <td>Standard Sizes</td>\n",
       "      <td>Other</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>8944045</td>\n",
       "      <td>25.00</td>\n",
       "      <td>-0.271543</td>\n",
       "      <td>P</td>\n",
       "      <td>1</td>\n",
       "      <td>58.0</td>\n",
       "      <td>29.00</td>\n",
       "      <td>CC412</td>\n",
       "      <td>MULTI</td>\n",
       "      <td>M</td>\n",
       "      <td>-0.225073</td>\n",
       "      <td>8043525</td>\n",
       "      <td>KORET OF</td>\n",
       "      <td>Multi</td>\n",
       "      <td>Standard Sizes</td>\n",
       "      <td>Other</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>8972921</td>\n",
       "      <td>14.72</td>\n",
       "      <td>-0.154386</td>\n",
       "      <td>P</td>\n",
       "      <td>1</td>\n",
       "      <td>35.0</td>\n",
       "      <td>35.00</td>\n",
       "      <td>TPILTRAQL</td>\n",
       "      <td>TERRA</td>\n",
       "      <td>BOU PIL</td>\n",
       "      <td>-0.071792</td>\n",
       "      <td>619402</td>\n",
       "      <td>STATUS Q</td>\n",
       "      <td>Other</td>\n",
       "      <td>Standard Sizes</td>\n",
       "      <td>Other</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       sku   cost    retail stype  quantity  orgprice    amt       style  \\\n",
       "0  8936935   9.58 -0.603488     P         1      24.0  24.00  3592413641   \n",
       "1  8940963   4.31 -0.730408     P         1      11.0   8.25   005U61516   \n",
       "2  8944045  28.05 -0.173912     P         1      68.0  29.99       CC412   \n",
       "3  8944045  25.00 -0.271543     P         1      58.0  29.00       CC412   \n",
       "4  8972921  14.72 -0.154386     P         1      35.0  35.00   TPILTRAQL   \n",
       "\n",
       "       color     size  packsize   vendor     brand standardized_color  \\\n",
       "0  WHEATNUBU        L -0.225073   914938  KIDS HEA              Other   \n",
       "1     GREYHE        L  0.081489  4012768  TOMMY HI               Grey   \n",
       "2      MULTI        M -0.225073  8043525  KORET OF              Multi   \n",
       "3      MULTI        M -0.225073  8043525  KORET OF              Multi   \n",
       "4      TERRA  BOU PIL -0.071792   619402  STATUS Q              Other   \n",
       "\n",
       "  standardized_size standardized_style  \n",
       "0    Standard Sizes              Other  \n",
       "1    Standard Sizes              Other  \n",
       "2    Standard Sizes              Other  \n",
       "3    Standard Sizes              Other  \n",
       "4    Standard Sizes              Other  "
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "joined_drop.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c2c69f80",
   "metadata": {},
   "source": [
    "Since we want to investigate whether customer would return or purchase a product based on its features, we must make this sample into a balanced dataset first, we decide to use SMOTE - Synthetic Minority Over-sampling Technique to make it into a balanced dataset."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d1acac93",
   "metadata": {},
   "source": [
    "The regression would focus on whether the product would be returned or not and it could further predict some unknow data. \n",
    "\n",
    "Then, we move to K-means to decide What products features would be returned most if they are returned. If they are sold, what product features would be sold best.\n",
    "\n",
    "\n",
    "regression outline:<br>\n",
    "y:stype (P or R)<br>\n",
    "x matrix: retail price, style, standardized_color, standardized_size,packsize, vendor, brand"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "b043ef06",
   "metadata": {},
   "outputs": [],
   "source": [
    "# save data to pickle for later processings\n",
    "joined_drop.to_pickle('../data/week9.pkl')"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.11.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}