SALES DATA

ANALYSIS
(Python + SQL)
DATA CLEANING (PYTHON)
SALES DATA ANALYSIS USING (PYTHON AND POSTGRES SQL)
Connect to Kaggle API and download file
import kaggle

!kaggle datasets download ankitbansal06/retail-orders -f orders.csv

Dataset URL: https://www.kaggle.com/datasets/ankitbansal06/retail-orders License(s): CC0-1.0

Downloading orders.csv.zip to C:\Users\kumar\LEARNING\Python + Postgres

0%| | 0.00/200k [00:00<?, ?B/s]

100%|##########| 200k/200k [00:00<00:00, 232kB/s]
100%|##########| 200k/200k [00:00<00:00, 232kB/s]

Extract csv file from zipped download file

import zipfile
zip_ref = zipfile.ZipFile('orders.csv.zip') zip_ref.extractall() # extract file to dir
zip_ref.close() # close file

Import Libraries
import pandas as pd
READ Dataset
data=pd.read_csv('orders.csv',na_values=['Not Available','unknown'])
data.head(10)

Order Id Order Date Ship Mode Segment Country \

United States
0 1 2023-03-01 Second Class Consumer United States
1 2 2023-08-15 Second Class Consumer United States
2 3 2023-01-10 Second Class Corporate United States
3 4 2022-06-18 Standard Class Consumer United States
4 5 2022-07-13 Standard Class Consumer United States
5 6 2022-03-13 NaN Consumer United States
6 7 2022-12-28 Standard Class Consumer United States
7 8 2022-01-25 Standard Class Consumer United States
8 9 2023-03-23 NaN Consumer United States
City State Postal Code Region Category \
0 Henderson Kentucky 42420 South Furniture Furniture
Henderson Los Kentucky South Office Supplies
1 42420 West
Angeles California
South Furniture
2 Fort Lauderdale Fort Florida 90036 Office Supplies
South
3 Lauderdale Los Florida 33311 West Furniture
Angeles California West
4 33311 Office Supplies
California West
Los Angeles Los
California West Technology
5 Angeles Los 90032
California West Office Supplies Office
Angeles Los
Sub Category Product Id cost price List Price Quantity \
0 Bookcases FUR-BO-10001798 240 260 2
1 Chairs FUR-CH-10000454 600 730 3
2 Labels OFF-LA-10000240 10 10 2
3 Tables FUR-TA-10000577 780 960 5
4 Storage OFF-ST-10000760 20 20 2
5 Furnishings FUR-FU-10001487 50 50 7
6 Art OFF-AR-10002833 10 10 4
7 Phones TEC-PH-10002275 860 910 6
8 Binders OFF-BI-10003910 20 20 3
9 Appliances OFF-AP-10002892 90 110 5

Discount Percent
0 2
1 3
2 5
3 2
4 5
5 3
6 3
7 5
8 2
9 3

data.dtypes

Order Id int64
Order Date object
Ship Mode object
Segment object
Country object
City object
State object
Postal Code int64
Region object
Category object
Sub Category object
Product Id object
cost price int64
Formatting columns
data.columns=data.columns.str.lower().str.replace(' ','_')

Date Time format change

data['order_date']=pd.to_datetime(data['order_date'],format='%Y-%m-%d')

Column addition
data['profit']=data['list_price']-data['cost_price']
data['discount']=data['list_price']*data['discount_percent']*.01
data['sale_price']=data['list_price']-data['discount']

drop useless columns

data.drop(['list_price','cost_price','discount_percent'],axis='columns',inplace=True) data.head()

order_id order_date ship_mode segment country \

0 United States
1 2023-03-01 Second Class Consumer United States
1 2 2023-08-15 Second Class Consumer United States
3 2023-01-10 Second Class Corporate United States
2
4 2022-06-18 Standard Class Consumer United States
city
5 2022-07-13 state
Standard Class postal_code region
Consumer category \
0 Henderson Kentucky 42420 South Furniture
1 Henderson Kentucky 42420 South Furniture
2 Los Angeles California 90036 West Office Supplies
3 Fort Lauderdale Florida 33311 South Furniture
4 Fort Lauderdale Florida 33311 South Office Supplies

sub_category product_id quantity profit discount sale_price

0 Bookcases FUR-BO-10001798 2 20 5.2 254.8
1 Chairs FUR-CH-10000454 3 130 21.9 708.1
2 Labels OFF-LA-10000240 2 0 0.5 9.5
3 Tables FUR-TA-10000577 5 180 19.2 940.8
4 Storage OFF-ST-10000760 2 0 1.0 19.0

CREATE TABLE IN POSTGRES (using psycopg2 libarary)

import psycopg2

# connect and create a new database

try:
conn=psycopg2.connect(host='localhost',password='harsh',user='postgres')

# auto commit
conn.autocommit=True

# cursor
cur=conn.cursor()
cur.execute('''CREATE DATABASE Orders;''')

conn.close()
except psycopg2.errors as e:
print(e)

# connect to newly connected database

try:
 conn=psycopg2.connect(database='orders',host='localhost',password='harsh',user='postgres')
except psycopg2.errors as e:
print(e)

# create cursor conn.autocommit=True

cur=conn.cursor()

CREATE TABLE Orders_table

query = """
CREATE TABLE Orders_table (
order_id INT,
order_date DATE,
ship_mode VARCHAR,
segment VARCHAR,
country VARCHAR,
city VARCHAR,
state VARCHAR,
postal_code INT,
region VARCHAR,
category VARCHAR,
sub_category VARCHAR,
product_id VARCHAR,
quantity INT,
profit INT,
discount FLOAT,
sale_price FLOAT
);
"""

try:
cur.execute(query)
print("Table Created")
except psycopg2.errors as e:
print(e)

INSERT DATA In New TAble

insert_query = """
INSERT INTO Orders_table
(order_id, order_date, ship_mode, segment, country, city, state, postal_code,
region, category, sub_category, product_id, quantity, profit, discount, sale_price)
VALUES (%s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s)
"""
try:
for index, row in data.iterrows():
# Execute the insert query for each row
cur.execute(insert_query, row)
print("Successfull insesrtion of data")
except psycopg2.errors as e:
print(e)
cur.execute('''SELECT * from Orders_table''')

pd.DataFrame(cur.fetchall())

0 1 2 3 4 \
0 1 2023-03-01 Second Class Consumer United States
1 2 2023-08-15 Second Class Consumer United States
2 3 2023-01-10 Second Class Corporate United States
3 4 2022-06-18 Standard Class Consumer United States
4 5 2022-07-13 Standard Class Consumer United States
... ... ... ... ... ...
9989 9990 2023-02-18 Second Class Consumer United States
9990 9991 2023-03-17 Standard Class Consumer United States
9991 9992 2022-08-07 Standard Class Consumer United States
9992 9993 2022-11-19 Standard Class Consumer United States
9993 9994 2022-07-17 Second Class Consumer United States
 5 6 7 8 9 10 \
0 Henderson Kentucky 42420 South Furniture Bookcases
1 Henderson Kentucky 42420 South Furniture Chairs
2 Los Angeles California 90036 West Office Supplies Labels
3 Fort Lauderdale Florida 33311 South Furniture Tables
4 Fort Lauderdale Florida 33311 South Office Supplies Storage
... ... ... ... ... ... ...
9989 Miami Florida 33180 South Furniture Furnishings
9990 Costa Mesa California 92627 West Furniture Furnishings
9991 Costa Mesa California 92627 West Technology Phones
9992 Costa Mesa California 92627 West Office Supplies Paper
9993 Westminster California 92683 West Office Supplies Appliances

[9994 rows x 16 columns]

CLOSE CONNECTION
conn.close()

CONCLUSION
1. Import dataset from Kaggle API

i. Unzip file and load dataset

ii. Data Cleaning
iii. Featuree Creation

2. Connection to PostgreSQL

i. Creating database and table

ii. Insertion of data into PostgreSQL
 SQL QUERIES
SALES DATA ANALYSIS USING (PYTHON AND POSTGRES SQL)
 FIND TOP 10 HIGHEST REVENUE GENERATING PRODUCTS
product_id total_revenue
TEC-CO-10004722 59514
SELECT OFF-BI-10003527 26525.3
PRODUCT_ID, ROUND(SUM(SALE_PRICE)::NUMERIC,1) AS TOTAL_REVENUE TEC-MA-10002412 21734.4
FROM ORDERS_TABLE FUR-CH-10002024 21096.2
GROUP BY 1 OFF-BI-10001359 19090.2
ORDER BY SUM(SALE_PRICE) DESC OFF-BI-10000545 18249
LIMIT 10; TEC-CO-10001449 18151.2
TEC-MA-10001127 17906.4
OFF-BI-10004995 17354.8
OFF-SU-10000151 16325.8
FIND TOP 5 HIGHEST SELLING PRODUCTS IN EACH REGION
region product_id total_sales rank
WITH CTE AS ( Central TEC-CO-10004722 16975 1
SELECT Central TEC-MA-10000822 13770 2
REGION, PRODUCT_ID, ROUND(SUM(SALE_PRICE)::NUMERIC,1) AS TOTAL_SALES, Central OFF-BI-10001120 11056.5 3
ROW_NUMBER () OVER (PARTITION BY REGION ORDER BY SUM(SALE_PRICE) Central OFF-BI-10000545 10132.7 4
DESC) AS RN Central OFF-BI-10004995 8416.1 5
FROM ORDERS_TABLE East TEC-CO-10004722 29099 1
GROUP BY 1,2 East TEC-MA-10001047 13767 2
) East FUR-BO-10004834 11274.1 3
SELECT
East OFF-BI-10001359 8463.6 4
REGION, PRODUCT_ID, TOTAL_SALES, RN AS RANK
East TEC-CO-10001449 8316 5
FROM CTE
WHERE RN<=5 South TEC-MA-10002412 21734.4 1
ORDER BY REGION, RN ASC; South TEC-MA-10001127 11116.4 2
South OFF-BI-10001359 8053.2 3
South TEC-MA-10004125 7840 4
South OFF-BI-10003527 7391.4 5
West TEC-CO-10004722 13440 1
West OFF-SU-10000151 12592.3 2
West FUR-CH-10001215 9604 3
West OFF-BI-10003527 7804.8 4
West TEC-AC-10003832 7722.7 5
 FIND TOP 5 HIGHEST SELLING PRODUCTS IN EACH REGION

WITH CTE AS ( month sales_2022 sales_2023

SELECT 1 94712.5 88632.6
REGION, PRODUCT_ID, ROUND(SUM(SALE_PRICE)::NUMERIC,1) AS TOTAL_SALES, 2 90091 128124.2
ROW_NUMBER () OVER (PARTITION BY REGION ORDER BY SUM(SALE_PRICE) DESC) AS RN 3 80106 82512.3
FROM ORDERS_TABLE 4 95451.6 111568.6
GROUP BY 1,2 5 79448.3 86447.9
) 6 94170.5 68976.5
SELECT 7 78652.2 90563.8
8 104808 87733.6
REGION, PRODUCT_ID, TOTAL_SALES, RN AS RANK
9 79142.2 76658.6
FROM CTE
10 118912.7 121061.5
WHERE RN<=5
11 84225.3 75432.8
ORDER BY REGION, RN ASC;
12 95869.9 102556.1

FOR EACH CATEGORY WHICH MONTH HAD HIGHEST SALES

WITH CTE AS (
SELECT CATEGORY,
EXTRACT (YEAR FROM ORDER_DATE) AS YEAR, category year month total_sales
EXTRACT (MONTH FROM ORDER_DATE) AS MONTH, Furniture 2022 10 42888.9
ROUND(SUM(SALE_PRICE)::NUMERIC,1) AS TOTAL_SALES
Office Supplies 2023 2 44118.5
FROM ORDERS_TABLE
GROUP BY 1,2,3 Technology 2023 10 53000.1
ORDER BY 1,2,3),
CTE2 AS (
SELECT
CATEGORY, YEAR, MONTH, TOTAL_SALES,
MAX(TOTAL_SALES) OVER (PARTITION BY CATEGORY) AS MAX_SALES
FROM CTE
)
SELECT
CATEGORY, YEAR, MONTH, TOTAL_SALES
FROM CTE2
WHERE TOTAL_SALES=MAX_SALES;
 WHICH SUB CATEGORY HAD HIGHEST GROWTH BY PROFIT IN 2023 COMPARE TO 2022

WITH CTE AS(

SELECT SUB_CATEGORY, EXTRACT (YEAR FROM ORDER_DATE) AS YEAR, SUM(PROFIT) AS TOTAL_PROFIT
FROM ORDERS_TABLE
GROUP BY 1,2),
CTE2 AS(
SELECT
SUB_CATEGORY,
SUM (CASE WHEN YEAR=2022 THEN TOTAL_PROFIT ELSE 0 END) AS TOTAL_PROFIT_2022,
SUM (CASE WHEN YEAR=2023 THEN TOTAL_PROFIT ELSE 0 END) AS TOTAL_PROFIT_2023,
SUM (CASE WHEN YEAR=2023 THEN TOTAL_PROFIT ELSE 0 END)-SUM (CASE WHEN YEAR=2022 THEN TOTAL_PROFIT ELSE 0 END) AS PROFIT_INCREASE
FROM CTE
GROUP BY SUB_CATEGORY)
SELECT *,
CONCAT (ROUND (((PROFIT_INCREASE *100)/TOTAL_PROFIT_2022)::NUMERIC,2),' %') AS GROWTH_PCT
FROM CTE2
ORDER BY PROFIT_INCREASE DESC
LIMIT 1;

sub_category total_profit_2022 total_profit_2023 profit_increase growth_pct

Machines 9980 14500 4520 45.29%
THANK YOU

Link to Github