├── orders data analysis.py
└── sql_code.sql


/orders data analysis.py:
--------------------------------------------------------------------------------
 1 | #!/usr/bin/env python
 2 | # coding: utf-8
 3 | 
 4 | # In[76]:
 5 | 
 6 | 
 7 | #import libraries
 8 | #!pip install kaggle
 9 | import kaggle
10 | 
11 | !kaggle datasets download ankitbansal06/retail-orders -f orders.csv
12 | 
13 | 
14 | # In[77]:
15 | 
16 | 
17 | #extract file from zip file
18 | import zipfile
19 | zip_ref = zipfile.ZipFile('orders.csv.zip') 
20 | zip_ref.extractall() # extract file to dir
21 | zip_ref.close() # close file
22 | 
23 | 
24 | # In[145]:
25 | 
26 | 
27 | #read data from the file and handle null values
28 | import pandas as pd
29 | df = pd.read_csv('orders.csv',na_values=['Not Available','unknown'])
30 | df['Ship Mode'].unique()
31 | 
32 | 
33 | # In[154]:
34 | 
35 | 
36 | #rename columns names ..make them lower case and replace space with underscore
37 | #df.rename(columns={'Order Id':'order_id', 'City':'city'})
38 | #df.columns=df.columns.str.lower()
39 | #df.columns=df.columns.str.replace(' ','_')
40 | df.head(5)
41 | 
42 | 
43 | # In[159]:
44 | 
45 | 
46 | #derive new columns discount , sale price and profit
47 | #df['discount']=df['list_price']*df['discount_percent']*.01
48 | #df['sale_price']= df['list_price']-df['discount']
49 | df['profit']=df['sale_price']-df['cost_price']
50 | df
51 | 
52 | 
53 | # In[162]:
54 | 
55 | 
56 | #convert order date from object data type to datetime
57 | df['order_date']=pd.to_datetime(df['order_date'],format="%Y-%m-%d")
58 | 
59 | 
60 | # In[167]:
61 | 
62 | 
63 | #drop cost price list price and discount percent columns
64 | df.drop(columns=['list_price','cost_price','discount_percent'],inplace=True)
65 | 
66 | 
67 | # In[169]:
68 | 
69 | 
70 | #load the data into sql server using replace option
71 | import sqlalchemy as sal
72 | engine = sal.create_engine('mssql://ANKIT\SQLEXPRESS/master?driver=ODBC+DRIVER+17+FOR+SQL+SERVER')
73 | conn=engine.connect()
74 | 
75 | 
76 | # In[172]:
77 | 
78 | 
79 | #load the data into sql server using append option
80 | df.to_sql('df_orders', con=conn , index=False, if_exists = 'append')
81 | 
82 | 
83 | 


--------------------------------------------------------------------------------
/sql_code.sql:
--------------------------------------------------------------------------------
 1 | --find top 10 highest reveue generating products 
 2 | select top 10 product_id,sum(sale_price) as sales
 3 | from df_orders
 4 | group by product_id
 5 | order by sales desc
 6 | 
 7 | 
 8 | 
 9 | 
10 | --find top 5 highest selling products in each region
11 | with cte as (
12 | select region,product_id,sum(sale_price) as sales
13 | from df_orders
14 | group by region,product_id)
15 | select * from (
16 | select *
17 | , row_number() over(partition by region order by sales desc) as rn
18 | from cte) A
19 | where rn<=5
20 | 
21 | 
22 | 
23 | --find month over month growth comparison for 2022 and 2023 sales eg : jan 2022 vs jan 2023
24 | with cte as (
25 | select year(order_date) as order_year,month(order_date) as order_month,
26 | sum(sale_price) as sales
27 | from df_orders
28 | group by year(order_date),month(order_date)
29 | --order by year(order_date),month(order_date)
30 | 	)
31 | select order_month
32 | , sum(case when order_year=2022 then sales else 0 end) as sales_2022
33 | , sum(case when order_year=2023 then sales else 0 end) as sales_2023
34 | from cte 
35 | group by order_month
36 | order by order_month
37 | 
38 | 
39 | 
40 | 
41 | 
42 | --for each category which month had highest sales 
43 | with cte as (
44 | select category,format(order_date,'yyyyMM') as order_year_month
45 | , sum(sale_price) as sales 
46 | from df_orders
47 | group by category,format(order_date,'yyyyMM')
48 | --order by category,format(order_date,'yyyyMM')
49 | )
50 | select * from (
51 | select *,
52 | row_number() over(partition by category order by sales desc) as rn
53 | from cte
54 | ) a
55 | where rn=1
56 | 
57 | 
58 | 
59 | 
60 | 
61 | 
62 | --which sub category had highest growth by profit in 2023 compare to 2022
63 | with cte as (
64 | select sub_category,year(order_date) as order_year,
65 | sum(sale_price) as sales
66 | from df_orders
67 | group by sub_category,year(order_date)
68 | --order by year(order_date),month(order_date)
69 | 	)
70 | , cte2 as (
71 | select sub_category
72 | , sum(case when order_year=2022 then sales else 0 end) as sales_2022
73 | , sum(case when order_year=2023 then sales else 0 end) as sales_2023
74 | from cte 
75 | group by sub_category
76 | )
77 | select top 1 *
78 | ,(sales_2023-sales_2022)
79 | from  cte2
80 | order by (sales_2023-sales_2022) desc
81 | 


--------------------------------------------------------------------------------