DataCleaning_price.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Data Science Portfolio 1#\n",
    "Repository containing portfolio of data science projects completed by me for self learning. Presented in the form of jupyter Notebooks.\n",
    "\n",
    "## Electronic Product Price Elasticity Project -\n",
    "This portfolio main purpose is to analyze price data in the sense of analyzing the impact of different pricing data through time in various e-commerce platforms and measure the impression demand for brands that are direct competitors among different electronic products.\n",
    "\n",
    "i.e. Does impression demand of competitor X increases if competitor Y increase their prices in the same or similar electronic products within a timeframe? Does free shipping boost impressions in electronic products? if yes, which products tend to have better results?\n",
    "\n",
    "\n",
    "**Part 1 Data Cleaning and Preprocessing**\n",
    "\n",
    "<font color='red'>1.1 Data Cleaning </font>\n",
    "\n",
    "- Manage Null values\n",
    "- Normalize Data (Condition and Shipping type label cleaning)\n",
    "- Deleted non-used columns\n",
    "- Filtered USD price products    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Part 1  Data Cleaning and Preprocessing "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1.1 Data Cleaning "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import datetime\n",
    "import math\n",
    "import matplotlib.pyplot as plt\n",
    "import re\n",
    "import seaborn as sns\n",
    "from IPython.display import display, HTML, Image"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "pricing_data = 'DatafinitiElectronicsProductsPricingData.csv'\n",
    "df = pd.read_csv (pricing_data)\n",
    "pd.options.display.width = 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Dataframe Dimensions: (14592, 31)\n"
     ]
    }
   ],
   "source": [
    "print(\"Dataframe Dimensions:\",df.shape)\n",
    "pd.set_option('display.max_columns', 500)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "---------- Display information about column types and number of null values ----------\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>prices.amountMax</th>\n",
       "      <th>prices.amountMin</th>\n",
       "      <th>prices.availability</th>\n",
       "      <th>prices.condition</th>\n",
       "      <th>prices.currency</th>\n",
       "      <th>prices.dateSeen</th>\n",
       "      <th>prices.isSale</th>\n",
       "      <th>prices.merchant</th>\n",
       "      <th>prices.shipping</th>\n",
       "      <th>prices.sourceURLs</th>\n",
       "      <th>asins</th>\n",
       "      <th>brand</th>\n",
       "      <th>categories</th>\n",
       "      <th>dateAdded</th>\n",
       "      <th>dateUpdated</th>\n",
       "      <th>ean</th>\n",
       "      <th>imageURLs</th>\n",
       "      <th>keys</th>\n",
       "      <th>manufacturer</th>\n",
       "      <th>manufacturerNumber</th>\n",
       "      <th>name</th>\n",
       "      <th>primaryCategories</th>\n",
       "      <th>sourceURLs</th>\n",
       "      <th>upc</th>\n",
       "      <th>weight</th>\n",
       "      <th>Unnamed: 26</th>\n",
       "      <th>Unnamed: 27</th>\n",
       "      <th>Unnamed: 28</th>\n",
       "      <th>Unnamed: 29</th>\n",
       "      <th>Unnamed: 30</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>column type</th>\n",
       "      <td>object</td>\n",
       "      <td>float64</td>\n",
       "      <td>float64</td>\n",
       "      <td>object</td>\n",
       "      <td>object</td>\n",
       "      <td>object</td>\n",
       "      <td>object</td>\n",
       "      <td>bool</td>\n",
       "      <td>object</td>\n",
       "      <td>object</td>\n",
       "      <td>object</td>\n",
       "      <td>object</td>\n",
       "      <td>object</td>\n",
       "      <td>object</td>\n",
       "      <td>object</td>\n",
       "      <td>object</td>\n",
       "      <td>object</td>\n",
       "      <td>object</td>\n",
       "      <td>object</td>\n",
       "      <td>object</td>\n",
       "      <td>object</td>\n",
       "      <td>object</td>\n",
       "      <td>object</td>\n",
       "      <td>object</td>\n",
       "      <td>object</td>\n",
       "      <td>object</td>\n",
       "      <td>object</td>\n",
       "      <td>object</td>\n",
       "      <td>float64</td>\n",
       "      <td>object</td>\n",
       "      <td>object</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>null values (nb)</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1925</td>\n",
       "      <td>1403</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>772</td>\n",
       "      <td>5799</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>10384</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>6974</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>14553</td>\n",
       "      <td>14574</td>\n",
       "      <td>14586</td>\n",
       "      <td>14580</td>\n",
       "      <td>14586</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>null values (%)</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>13.1922</td>\n",
       "      <td>9.61486</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5.29057</td>\n",
       "      <td>39.741</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>71.1623</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>47.7933</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>99.7327</td>\n",
       "      <td>99.8766</td>\n",
       "      <td>99.9589</td>\n",
       "      <td>99.9178</td>\n",
       "      <td>99.9589</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                      id prices.amountMax prices.amountMin  \\\n",
       "column type       object          float64          float64   \n",
       "null values (nb)       0                0                0   \n",
       "null values (%)        0                0                0   \n",
       "\n",
       "                 prices.availability prices.condition prices.currency  \\\n",
       "column type                   object           object          object   \n",
       "null values (nb)                1925             1403               0   \n",
       "null values (%)              13.1922          9.61486               0   \n",
       "\n",
       "                 prices.dateSeen prices.isSale prices.merchant  \\\n",
       "column type               object          bool          object   \n",
       "null values (nb)               0             0             772   \n",
       "null values (%)                0             0         5.29057   \n",
       "\n",
       "                 prices.shipping prices.sourceURLs   asins   brand categories  \\\n",
       "column type               object            object  object  object     object   \n",
       "null values (nb)            5799                 0       0       0          0   \n",
       "null values (%)           39.741                 0       0       0          0   \n",
       "\n",
       "                 dateAdded dateUpdated      ean imageURLs    keys  \\\n",
       "column type         object      object   object    object  object   \n",
       "null values (nb)         0           0    10384         0       0   \n",
       "null values (%)          0           0  71.1623         0       0   \n",
       "\n",
       "                 manufacturer manufacturerNumber    name primaryCategories  \\\n",
       "column type            object             object  object            object   \n",
       "null values (nb)         6974                  0       0                 0   \n",
       "null values (%)       47.7933                  0       0                 0   \n",
       "\n",
       "                 sourceURLs     upc  weight Unnamed: 26 Unnamed: 27  \\\n",
       "column type          object  object  object      object      object   \n",
       "null values (nb)          0       0       0       14553       14574   \n",
       "null values (%)           0       0       0     99.7327     99.8766   \n",
       "\n",
       "                 Unnamed: 28 Unnamed: 29 Unnamed: 30  \n",
       "column type          float64      object      object  \n",
       "null values (nb)       14586       14580       14586  \n",
       "null values (%)      99.9589     99.9178     99.9589  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "# gives some infomation on columns types and number of null values\n",
    "tab_info = pd.DataFrame(df.dtypes).T.rename(index={0:'column type'})\n",
    "tab_info = tab_info.append(pd.DataFrame(df.isnull().sum()).T.rename(index={0:'null values (nb)'}))\n",
    "tab_info = tab_info.append(pd.DataFrame(df.isnull().sum()/df.shape[0]*100).T.\n",
    "                         rename(index={0:'null values (%)'}))\n",
    "print ('-' * 10 + \" Display information about column types and number of null values \" + '-' * 10 )\n",
    "print \n",
    "display(tab_info)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Drop columns that are not going to be used for the project\n",
    "df = df[df.columns.drop(['sourceURLs','prices.availability','asins','manufacturerNumber','upc', 'primaryCategories','keys','id','ean'])]\n",
    "#Drop columns that started with Unnamed\n",
    "df = df[df.columns.drop(list(df.filter(regex='Unnamed*')))]\n",
    "\n",
    "#Change name of columns\n",
    "columns_to_change = {}\n",
    "for column in df.columns:\n",
    "    if column.startswith(\"prices.\"):\n",
    "        columns_to_change[column] = column[7:]\n",
    "df.rename(columns=columns_to_change, inplace=True)\n",
    "df.rename(columns = {'amountMax':'price', 'amountMin':'disc_price','categories':'p_description'}, inplace = True) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['price', 'disc_price', 'condition', 'currency', 'dateSeen', 'isSale',\n",
       "       'merchant', 'shipping', 'sourceURLs', 'brand', 'p_description',\n",
       "       'dateAdded', 'dateUpdated', 'imageURLs', 'manufacturer', 'name',\n",
       "       'weight'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Doublecheck Sale Prices\n",
    "df['isSale'] = np.where(df['price']> df['disc_price'],\"Yes\", \"No\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Text Normalization (Product Condition, Merchant Website and Shipping method)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Product Condition Percentage\n",
      "--------------------------------------------------\n",
      "New                         85.649671\n",
      "None                         9.614857\n",
      "Used                         3.125000\n",
      "Refurbished                  1.576206\n",
      "For parts or not working     0.034265\n",
      "Name: condition, dtype: float64\n",
      "Product Condition values count\n",
      "--------------------------------------------------\n",
      "New                         12498\n",
      "None                         1403\n",
      "Used                          456\n",
      "Refurbished                   230\n",
      "For parts or not working        5\n",
      "Name: condition, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "#Filtered Product Condition by Used, New and Refurbished\n",
    "condition= df['condition'].fillna('None')\n",
    "df['condition'] = pd.np.where(condition.str.contains('new', flags=re.IGNORECASE), \"New\",\n",
    "                           pd.np.where(condition.str.contains('refurbished',flags=re.IGNORECASE), \"Refurbished\", \n",
    "                           pd.np.where(condition.str.contains('pre-owned|used',flags=re.IGNORECASE), \"Used\",condition)))\n",
    "                  \n",
    "c = df['condition'].value_counts()\n",
    "print(\"Product Condition Percentage\\n{}\\n{}\\nProduct Condition values count\\n{}\\n{}\".format(\n",
    "      50*\"-\", c / len(df.index) * 100, 50*\"-\", c))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# for the Merchants accuracy, we will use the source URL to include the correct merchant\n",
    "#df['sourceURLs'][(df['merchant']== 'Not Specified') | (df['merchant']== 'Other')]\n",
    "merchant_ab= df['sourceURLs']\n",
    "df['merchant'] = pd.np.where(merchant_ab.str.contains('bhphotovideo', flags=re.IGNORECASE), \"bhphotovideo.com\",\n",
    "                 pd.np.where(merchant_ab.str.contains('eBay|e bay', flags=re.IGNORECASE), \"ebay.com\",\n",
    "                 pd.np.where(merchant_ab.str.contains('Amazon',flags=re.IGNORECASE), \"Amazon.com\", \n",
    "                 pd.np.where(merchant_ab.str.contains('Bestbuy',flags=re.IGNORECASE), \"Bestbuy.com\",\n",
    "                 pd.np.where(merchant_ab.str.contains('Homedepot',flags=re.IGNORECASE), \"homedepot.com\",\n",
    "                 pd.np.where(merchant_ab.str.contains('newegg',flags=re.IGNORECASE), \"newegg.com\",\n",
    "                 pd.np.where(merchant_ab.str.contains('kmart',flags=re.IGNORECASE), \"kmart.com\",\n",
    "                 pd.np.where(merchant_ab.str.contains('frys',flags=re.IGNORECASE), \"frys.com\",\n",
    "                 pd.np.where(merchant_ab.str.contains('cdw',flags=re.IGNORECASE), \"cdw.com\",\n",
    "                 pd.np.where(merchant_ab.str.contains('target',flags=re.IGNORECASE), \"target.com\",\n",
    "                 pd.np.where(merchant_ab.str.contains('overstock',flags=re.IGNORECASE), \"overstock.com\",\n",
    "                 pd.np.where(merchant_ab.str.contains('barcodable',flags=re.IGNORECASE), \"barcodable.com\",\n",
    "                 pd.np.where(merchant_ab.str.contains('kohls',flags=re.IGNORECASE), \"kohls.com\",\n",
    "                 pd.np.where(merchant_ab.str.contains('sears',flags=re.IGNORECASE), \"sears.com\",\n",
    "                 pd.np.where(merchant_ab.str.contains('Wal-mart|Walmart',flags=re.IGNORECASE), \"Walmart.com\",\"Other\")))))))))))))))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Merchant Percentage\n",
      "--------------------------------------------------\n",
      "Bestbuy.com         30.838816\n",
      "Walmart.com         25.479715\n",
      "bhphotovideo.com    18.893914\n",
      "ebay.com            14.624452\n",
      "Amazon.com           7.147752\n",
      "target.com           0.842928\n",
      "frys.com             0.507127\n",
      "barcodable.com       0.322094\n",
      "newegg.com           0.308388\n",
      "cdw.com              0.294682\n",
      "Other                0.260417\n",
      "homedepot.com        0.157621\n",
      "kmart.com            0.123355\n",
      "kohls.com            0.075384\n",
      "sears.com            0.068531\n",
      "overstock.com        0.054825\n",
      "Name: merchant, dtype: float64\n",
      "Merchant values count\n",
      "--------------------------------------------------\n",
      "Bestbuy.com         4500\n",
      "Walmart.com         3718\n",
      "bhphotovideo.com    2757\n",
      "ebay.com            2134\n",
      "Amazon.com          1043\n",
      "target.com           123\n",
      "frys.com              74\n",
      "barcodable.com        47\n",
      "newegg.com            45\n",
      "cdw.com               43\n",
      "Other                 38\n",
      "homedepot.com         23\n",
      "kmart.com             18\n",
      "kohls.com             11\n",
      "sears.com             10\n",
      "overstock.com          8\n",
      "Name: merchant, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "m = df['merchant'].value_counts()\n",
    "print(\"Merchant Percentage\\n{}\\n{}\\nMerchant values count\\n{}\\n{}\".format(\n",
    "      50*\"-\", m / len(df.index) * 100, 50*\"-\", m))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Our data shows the date that an advertisment with specific price was seen, often named this action in tech companies and marketing as impressions. \n",
    "With the main purpose to do further analysis in price elasticities model by date, each row is extended by dateseen (impression_date)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Dataframe Dimensions without date impressions rows: (14592, 17)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>name</th>\n",
       "      <th>dateSeen</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Sanus VLF410B1 10-Inch Super Slim Full-Motion ...</td>\n",
       "      <td>2017-03-30T06:00:00Z,2017-03-10T22:00:00Z,2017...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Boytone - 2500W 2.1-Ch. Home Theater System - ...</td>\n",
       "      <td>2017-12-14T06:00:00Z</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Boytone - 2500W 2.1-Ch. Home Theater System - ...</td>\n",
       "      <td>2017-09-08T05:00:00Z</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Boytone - 2500W 2.1-Ch. Home Theater System - ...</td>\n",
       "      <td>2017-10-10T05:00:00Z</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Boytone - 2500W 2.1-Ch. Home Theater System - ...</td>\n",
       "      <td>2017-08-28T07:00:00Z</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                name  \\\n",
       "0  Sanus VLF410B1 10-Inch Super Slim Full-Motion ...   \n",
       "1  Boytone - 2500W 2.1-Ch. Home Theater System - ...   \n",
       "2  Boytone - 2500W 2.1-Ch. Home Theater System - ...   \n",
       "3  Boytone - 2500W 2.1-Ch. Home Theater System - ...   \n",
       "4  Boytone - 2500W 2.1-Ch. Home Theater System - ...   \n",
       "\n",
       "                                            dateSeen  \n",
       "0  2017-03-30T06:00:00Z,2017-03-10T22:00:00Z,2017...  \n",
       "1                               2017-12-14T06:00:00Z  \n",
       "2                               2017-09-08T05:00:00Z  \n",
       "3                               2017-10-10T05:00:00Z  \n",
       "4                               2017-08-28T07:00:00Z  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(\"Dataframe Dimensions without date impressions rows:\", df.shape)\n",
    "display(df[['name','dateSeen']][:5])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Split dates for further individual row processing as df['Date_imp] column, add count of impressions per product as df['Imp_count']\n",
    "time_scratch=[]\n",
    "date_count=[]\n",
    "\n",
    "for i in df['dateSeen']:\n",
    "    time_= i.split(\",\")\n",
    "    time_scratch.append(time_)\n",
    "    date_count.append(len(time_))\n",
    "                      \n",
    "df['Imp_count']= date_count                     \n",
    "df['Date_imp'] = time_scratch"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Impressions by date expansion\n",
    "def explode(df, lst_cols, fill_value='', preserve_index=False):\n",
    "    # make sure `lst_cols` is list-alike\n",
    "    if (lst_cols is not None\n",
    "        and len(lst_cols) > 0\n",
    "        and not isinstance(lst_cols, (list, tuple, np.ndarray, pd.Series))):\n",
    "        lst_cols = [lst_cols]\n",
    "    # all columns except `lst_cols`\n",
    "    idx_cols = df.columns.difference(lst_cols)\n",
    "    # calculate lengths of lists\n",
    "    lens = df[lst_cols[0]].str.len()\n",
    "    # preserve original index values    \n",
    "    idx = np.repeat(df.index.values, lens)\n",
    "    # create \"exploded\" DF\n",
    "    res = (pd.DataFrame({\n",
    "                col: np.repeat(df[col].values, lens)\n",
    "                for col in idx_cols},\n",
    "                index=idx)\n",
    "             .assign(**{col:np.concatenate(df.loc[lens>0, col].values)\n",
    "                            for col in lst_cols}))\n",
    "    # append those rows that have empty lists\n",
    "    if (lens == 0).any():\n",
    "        # at least one list in cells is empty\n",
    "        res = (res.append(df.loc[lens==0, idx_cols], sort=False)\n",
    "                  .fillna(fill_value))\n",
    "    # revert the original index order\n",
    "    res = res.sort_index()\n",
    "    # reset index if requested\n",
    "    if not preserve_index:        \n",
    "        res = res.reset_index(drop=True)\n",
    "    return res\n",
    "\n",
    "\n",
    "df = explode(df , ['Date_imp'], fill_value='')\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Add Month, day, weeknumber, day of the week for further analysis separately **"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Dataframe dimensions with impressions: (43238, 26)\n"
     ]
    }
   ],
   "source": [
    "#Convert df['Date_imp'] to Datetime\n",
    "df['Date_imp']= pd.to_datetime(df['Date_imp'], errors='coerce')\n",
    "df['Date_imp_d']= df['Date_imp'].dt.strftime('%Y/%m/%d')\n",
    "df['Day_n']= df['Date_imp'].dt.strftime('%A')\n",
    "df['year']= df['Date_imp'].dt.year\n",
    "df['month']= df['Date_imp'].dt.month\n",
    "df['month_n']= df['Date_imp'].dt.strftime('%B')\n",
    "df['day']= df['Date_imp'].dt.day\n",
    "df['Week_Number'] = df['Date_imp'].dt.week\n",
    "#( Check weekday) df['WeekDay'] = df['Date_imp'].dt.weekday\n",
    "\n",
    "\n",
    "print(\"Dataframe dimensions with impressions:\" ,df.shape)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "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>Date_imp</th>\n",
       "      <th>Date_imp_d</th>\n",
       "      <th>Day_n</th>\n",
       "      <th>month_n</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2017-03-30 06:00:00</td>\n",
       "      <td>2017/03/30</td>\n",
       "      <td>Thursday</td>\n",
       "      <td>March</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2017-03-10 22:00:00</td>\n",
       "      <td>2017/03/10</td>\n",
       "      <td>Friday</td>\n",
       "      <td>March</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2017-03-04 10:00:00</td>\n",
       "      <td>2017/03/04</td>\n",
       "      <td>Saturday</td>\n",
       "      <td>March</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2017-03-03 14:00:00</td>\n",
       "      <td>2017/03/03</td>\n",
       "      <td>Friday</td>\n",
       "      <td>March</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2017-12-14 06:00:00</td>\n",
       "      <td>2017/12/14</td>\n",
       "      <td>Thursday</td>\n",
       "      <td>December</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             Date_imp  Date_imp_d     Day_n   month_n\n",
       "0 2017-03-30 06:00:00  2017/03/30  Thursday     March\n",
       "1 2017-03-10 22:00:00  2017/03/10    Friday     March\n",
       "2 2017-03-04 10:00:00  2017/03/04  Saturday     March\n",
       "3 2017-03-03 14:00:00  2017/03/03    Friday     March\n",
       "4 2017-12-14 06:00:00  2017/12/14  Thursday  December"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[['Date_imp','Date_imp_d','Day_n','month_n']][:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "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>Imp_count</th>\n",
       "      <th>brand</th>\n",
       "      <th>condition</th>\n",
       "      <th>currency</th>\n",
       "      <th>dateAdded</th>\n",
       "      <th>dateSeen</th>\n",
       "      <th>dateUpdated</th>\n",
       "      <th>disc_price</th>\n",
       "      <th>imageURLs</th>\n",
       "      <th>isSale</th>\n",
       "      <th>manufacturer</th>\n",
       "      <th>merchant</th>\n",
       "      <th>name</th>\n",
       "      <th>p_description</th>\n",
       "      <th>price</th>\n",
       "      <th>shipping</th>\n",
       "      <th>sourceURLs</th>\n",
       "      <th>weight</th>\n",
       "      <th>Date_imp</th>\n",
       "      <th>Date_imp_d</th>\n",
       "      <th>Day_n</th>\n",
       "      <th>year</th>\n",
       "      <th>month</th>\n",
       "      <th>month_n</th>\n",
       "      <th>day</th>\n",
       "      <th>Week_Number</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>4</td>\n",
       "      <td>Sanus</td>\n",
       "      <td>New</td>\n",
       "      <td>USD</td>\n",
       "      <td>2015-04-13T12:00:51Z</td>\n",
       "      <td>2017-03-30T06:00:00Z,2017-03-10T22:00:00Z,2017...</td>\n",
       "      <td>2018-05-12T18:59:48Z</td>\n",
       "      <td>104.99</td>\n",
       "      <td>https://images-na.ssl-images-amazon.com/images...</td>\n",
       "      <td>No</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Bestbuy.com</td>\n",
       "      <td>Sanus VLF410B1 10-Inch Super Slim Full-Motion ...</td>\n",
       "      <td>Audio &amp; Video Accessories,TV Mounts,TV Accesso...</td>\n",
       "      <td>104.99</td>\n",
       "      <td>NaN</td>\n",
       "      <td>http://www.bestbuy.com/site/sanus-tv-wall-moun...</td>\n",
       "      <td>32.8 pounds</td>\n",
       "      <td>2017-03-30 06:00:00</td>\n",
       "      <td>2017/03/30</td>\n",
       "      <td>Thursday</td>\n",
       "      <td>2017</td>\n",
       "      <td>3</td>\n",
       "      <td>March</td>\n",
       "      <td>30</td>\n",
       "      <td>13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>4</td>\n",
       "      <td>Sanus</td>\n",
       "      <td>New</td>\n",
       "      <td>USD</td>\n",
       "      <td>2015-04-13T12:00:51Z</td>\n",
       "      <td>2017-03-30T06:00:00Z,2017-03-10T22:00:00Z,2017...</td>\n",
       "      <td>2018-05-12T18:59:48Z</td>\n",
       "      <td>104.99</td>\n",
       "      <td>https://images-na.ssl-images-amazon.com/images...</td>\n",
       "      <td>No</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Bestbuy.com</td>\n",
       "      <td>Sanus VLF410B1 10-Inch Super Slim Full-Motion ...</td>\n",
       "      <td>Audio &amp; Video Accessories,TV Mounts,TV Accesso...</td>\n",
       "      <td>104.99</td>\n",
       "      <td>NaN</td>\n",
       "      <td>http://www.bestbuy.com/site/sanus-tv-wall-moun...</td>\n",
       "      <td>32.8 pounds</td>\n",
       "      <td>2017-03-10 22:00:00</td>\n",
       "      <td>2017/03/10</td>\n",
       "      <td>Friday</td>\n",
       "      <td>2017</td>\n",
       "      <td>3</td>\n",
       "      <td>March</td>\n",
       "      <td>10</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Imp_count  brand condition currency             dateAdded  \\\n",
       "0          4  Sanus       New      USD  2015-04-13T12:00:51Z   \n",
       "1          4  Sanus       New      USD  2015-04-13T12:00:51Z   \n",
       "\n",
       "                                            dateSeen           dateUpdated  \\\n",
       "0  2017-03-30T06:00:00Z,2017-03-10T22:00:00Z,2017...  2018-05-12T18:59:48Z   \n",
       "1  2017-03-30T06:00:00Z,2017-03-10T22:00:00Z,2017...  2018-05-12T18:59:48Z   \n",
       "\n",
       "   disc_price                                          imageURLs isSale  \\\n",
       "0      104.99  https://images-na.ssl-images-amazon.com/images...     No   \n",
       "1      104.99  https://images-na.ssl-images-amazon.com/images...     No   \n",
       "\n",
       "  manufacturer     merchant  \\\n",
       "0          NaN  Bestbuy.com   \n",
       "1          NaN  Bestbuy.com   \n",
       "\n",
       "                                                name  \\\n",
       "0  Sanus VLF410B1 10-Inch Super Slim Full-Motion ...   \n",
       "1  Sanus VLF410B1 10-Inch Super Slim Full-Motion ...   \n",
       "\n",
       "                                       p_description   price shipping  \\\n",
       "0  Audio & Video Accessories,TV Mounts,TV Accesso...  104.99      NaN   \n",
       "1  Audio & Video Accessories,TV Mounts,TV Accesso...  104.99      NaN   \n",
       "\n",
       "                                          sourceURLs       weight  \\\n",
       "0  http://www.bestbuy.com/site/sanus-tv-wall-moun...  32.8 pounds   \n",
       "1  http://www.bestbuy.com/site/sanus-tv-wall-moun...  32.8 pounds   \n",
       "\n",
       "             Date_imp  Date_imp_d     Day_n  year  month month_n  day  \\\n",
       "0 2017-03-30 06:00:00  2017/03/30  Thursday  2017      3   March   30   \n",
       "1 2017-03-10 22:00:00  2017/03/10    Friday  2017      3   March   10   \n",
       "\n",
       "   Week_Number  \n",
       "0           13  \n",
       "1           10  "
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11479fc50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x115011630>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Setting background color for all graphs\n",
    "bg_color = (0.98,0.98,0.98)\n",
    "plt.rcParams['figure.facecolor'] = bg_color\n",
    "plt.rcParams['axes.facecolor'] = bg_color\n",
    "\n",
    "#Create impressions pie chart data frame\n",
    "pie_prod = df.groupby('year').agg({'name':'nunique','Date_imp':'count'}).reset_index()\n",
    "pie_prod = pie_prod.rename(columns = {'year':'Year','name':'Product', 'Date_imp':'Impressions'})\n",
    "x = pie_prod['Year']\n",
    "unique_prod = pie_prod['Product']\n",
    "Imp_count = pie_prod['Impressions']\n",
    "percent_unique = 100.*unique_prod/unique_prod.sum()\n",
    "percent_imp = 100.*Imp_count/Imp_count.sum()\n",
    "\n",
    "\n",
    "#Pie Chart subplots \n",
    "plt.rcParams['figure.figsize'] = [8, 5]\n",
    "plt.subplot(121)\n",
    "patches, texts = plt.pie(unique_prod  ,startangle=100, shadow = True, wedgeprops={\"edgecolor\":\"k\",'linewidth': 1, 'antialiased': True})\n",
    "labels = ['{0} - {1:1.2f} %'.format(i,j) for i,j in zip(x, percent_unique)]\n",
    "sort_legend = True\n",
    "if sort_legend:\n",
    "    patches, labels, dummy =  zip(*sorted(zip(patches, labels, unique_prod),\n",
    "                                          key=lambda x: x[2],\n",
    "                                          reverse=True))\n",
    "plt.legend(patches, labels, loc='upper center', title = 'Year / %', bbox_to_anchor = (.9, .85),\n",
    "           fontsize = 10)\n",
    "    \n",
    "plt.title('Unique Products per Year',fontsize = 15, y=.9)\n",
    "plt.axis('equal')\n",
    "\n",
    "plt.subplot(122)\n",
    "patches, texts = plt.pie(Imp_count , startangle=100, shadow = True, wedgeprops={\"edgecolor\":\"k\",'linewidth': 1, 'antialiased': True})\n",
    "labels = ['{0} - {1:1.2f} %'.format(i,j) for i,j in zip(x, percent_imp)]\n",
    "sort_legend = True\n",
    "if sort_legend:\n",
    "    patches, labels, dummy =  zip(*sorted(zip(patches, labels, Imp_count),\n",
    "                                          key=lambda x: x[2],\n",
    "                                          reverse=True))\n",
    "plt.legend(patches, labels, loc='upper center', title = 'Year / %', bbox_to_anchor = (.9, .85),\n",
    "           fontsize = 10)\n",
    "plt.title('Total Impressions per Year',fontsize = 15, y=.9)\n",
    "plt.axis('equal')\n",
    "plt.show()\n",
    "\n",
    "\n",
    "pie_prod.plot.bar(x ='Year', rot=0, grid = True, edgecolor='k')\n",
    "plt.title('Count of Unique Producst and Total Impressions per Year', fontsize=15)\n",
    "plt.tick_params(axis='x', labelsize=12)\n",
    "plt.tick_params(axis='y', labelsize=10)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In previous charts, we can vizualize that about 80% of the impressions in the dataset ocurred in 2017. For this main reason our analysis would cover only year 2017"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "New DataFrame Dimensions with only 2017 Impressions: 34389\n"
     ]
    }
   ],
   "source": [
    "#Select only 2017 Impressions for analysis\n",
    "df = df[df['year'] == 2017]\n",
    "print(\"New DataFrame Dimensions with only 2017 Impressions:\",len(df))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Add Discount percentage of products with impressions\n",
    "df['Disc_percentage']= round(1-(df['disc_price']/df['price']),2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "---------- Display information about column types and number of null values, unique values and total count----------\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Imp_count</th>\n",
       "      <th>brand</th>\n",
       "      <th>condition</th>\n",
       "      <th>currency</th>\n",
       "      <th>dateAdded</th>\n",
       "      <th>dateSeen</th>\n",
       "      <th>dateUpdated</th>\n",
       "      <th>disc_price</th>\n",
       "      <th>imageURLs</th>\n",
       "      <th>isSale</th>\n",
       "      <th>manufacturer</th>\n",
       "      <th>merchant</th>\n",
       "      <th>name</th>\n",
       "      <th>p_description</th>\n",
       "      <th>price</th>\n",
       "      <th>shipping</th>\n",
       "      <th>sourceURLs</th>\n",
       "      <th>weight</th>\n",
       "      <th>Date_imp</th>\n",
       "      <th>Date_imp_d</th>\n",
       "      <th>Day_n</th>\n",
       "      <th>year</th>\n",
       "      <th>month</th>\n",
       "      <th>month_n</th>\n",
       "      <th>day</th>\n",
       "      <th>Week_Number</th>\n",
       "      <th>Disc_percentage</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>column type</th>\n",
       "      <td>int64</td>\n",
       "      <td>object</td>\n",
       "      <td>object</td>\n",
       "      <td>object</td>\n",
       "      <td>object</td>\n",
       "      <td>object</td>\n",
       "      <td>object</td>\n",
       "      <td>float64</td>\n",
       "      <td>object</td>\n",
       "      <td>object</td>\n",
       "      <td>object</td>\n",
       "      <td>object</td>\n",
       "      <td>object</td>\n",
       "      <td>object</td>\n",
       "      <td>float64</td>\n",
       "      <td>object</td>\n",
       "      <td>object</td>\n",
       "      <td>object</td>\n",
       "      <td>datetime64[ns]</td>\n",
       "      <td>object</td>\n",
       "      <td>object</td>\n",
       "      <td>int64</td>\n",
       "      <td>int64</td>\n",
       "      <td>object</td>\n",
       "      <td>int64</td>\n",
       "      <td>int64</td>\n",
       "      <td>float64</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Unique Values</th>\n",
       "      <td>54</td>\n",
       "      <td>296</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>1077</td>\n",
       "      <td>5785</td>\n",
       "      <td>983</td>\n",
       "      <td>3697</td>\n",
       "      <td>1142</td>\n",
       "      <td>2</td>\n",
       "      <td>266</td>\n",
       "      <td>13</td>\n",
       "      <td>1129</td>\n",
       "      <td>1118</td>\n",
       "      <td>3546</td>\n",
       "      <td>131</td>\n",
       "      <td>6415</td>\n",
       "      <td>748</td>\n",
       "      <td>2217</td>\n",
       "      <td>270</td>\n",
       "      <td>7</td>\n",
       "      <td>1</td>\n",
       "      <td>12</td>\n",
       "      <td>12</td>\n",
       "      <td>31</td>\n",
       "      <td>51</td>\n",
       "      <td>74</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Total Count</th>\n",
       "      <td>34389</td>\n",
       "      <td>34389</td>\n",
       "      <td>34389</td>\n",
       "      <td>34389</td>\n",
       "      <td>34389</td>\n",
       "      <td>34389</td>\n",
       "      <td>34389</td>\n",
       "      <td>34389</td>\n",
       "      <td>34389</td>\n",
       "      <td>34389</td>\n",
       "      <td>18638</td>\n",
       "      <td>34389</td>\n",
       "      <td>34389</td>\n",
       "      <td>34389</td>\n",
       "      <td>34389</td>\n",
       "      <td>16111</td>\n",
       "      <td>34389</td>\n",
       "      <td>34389</td>\n",
       "      <td>34389</td>\n",
       "      <td>34389</td>\n",
       "      <td>34389</td>\n",
       "      <td>34389</td>\n",
       "      <td>34389</td>\n",
       "      <td>34389</td>\n",
       "      <td>34389</td>\n",
       "      <td>34389</td>\n",
       "      <td>34389</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>null values (nb)</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>15751</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>18278</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>null values (%)</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>45.8024</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>53.1507</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 Imp_count   brand condition currency dateAdded dateSeen  \\\n",
       "column type          int64  object    object   object    object   object   \n",
       "Unique Values           54     296         5        5      1077     5785   \n",
       "Total Count          34389   34389     34389    34389     34389    34389   \n",
       "null values (nb)         0       0         0        0         0        0   \n",
       "null values (%)          0       0         0        0         0        0   \n",
       "\n",
       "                 dateUpdated disc_price imageURLs  isSale manufacturer  \\\n",
       "column type           object    float64    object  object       object   \n",
       "Unique Values            983       3697      1142       2          266   \n",
       "Total Count            34389      34389     34389   34389        18638   \n",
       "null values (nb)           0          0         0       0        15751   \n",
       "null values (%)            0          0         0       0      45.8024   \n",
       "\n",
       "                 merchant    name p_description    price shipping sourceURLs  \\\n",
       "column type        object  object        object  float64   object     object   \n",
       "Unique Values          13    1129          1118     3546      131       6415   \n",
       "Total Count         34389   34389         34389    34389    16111      34389   \n",
       "null values (nb)        0       0             0        0    18278          0   \n",
       "null values (%)         0       0             0        0  53.1507          0   \n",
       "\n",
       "                  weight        Date_imp Date_imp_d   Day_n   year  month  \\\n",
       "column type       object  datetime64[ns]     object  object  int64  int64   \n",
       "Unique Values        748            2217        270       7      1     12   \n",
       "Total Count        34389           34389      34389   34389  34389  34389   \n",
       "null values (nb)       0               0          0       0      0      0   \n",
       "null values (%)        0               0          0       0      0      0   \n",
       "\n",
       "                 month_n    day Week_Number Disc_percentage  \n",
       "column type       object  int64       int64         float64  \n",
       "Unique Values         12     31          51              74  \n",
       "Total Count        34389  34389       34389           34389  \n",
       "null values (nb)       0      0           0               0  \n",
       "null values (%)        0      0           0               0  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "# gives some infomation on columns types and number of null values and unique values \n",
    "summary_tab = pd.DataFrame(df.dtypes).T.rename(index={0:'column type'})\n",
    "summary_tab= summary_tab.append(pd.DataFrame(df.nunique()).T.rename(index={0:'Unique Values'}))\n",
    "summary_tab= summary_tab.append(pd.DataFrame(df.count()).T.rename(index={0:'Total Count'}))\n",
    "summary_tab= summary_tab.append(pd.DataFrame(df.isnull().sum()).T.rename(index={0:'null values (nb)'}))\n",
    "summary_tab = summary_tab.append(pd.DataFrame(df.isnull().sum()/df.shape[0]*100).T.rename(index={0:'null values (%)'}))\n",
    "print ('-' * 10 + \" Display information about column types and number of null values, unique values and total count\" + '-' * 10 )\n",
    "print \n",
    "display(summary_tab)"
   ]
  }
 ],
 "metadata": {
  "hide_input": false,
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.4"
  },
  "widgets": {
   "application/vnd.jupyter.widget-state+json": {
    "state": {},
    "version_major": 2,
    "version_minor": 0
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}