Pfthb

I have a DataFrame in this way:

 shop_id item_price item_cnt_day day month year
 59 9.00 1.0 02 01 2013
 59 8.00 2.0 02 01 2013
 25 10.00 4.0 05 02 2013
 25 17.0 1.0 06 01 2013
 25 10.00 1.0 15 01 2013

And I try to get the result like following DataFrame:

 shop_id all_revenue month year
 59 25.00 01 2013
 25 27.00 01 2013

I mean I want to get each shop's revenue in January 2013.

BUT, I don't know how to code in Pandas. Any help would be appreciated.

eval + groupby + sum

You can assign a series via eval, then use groupby:

res = df.eval('revenue=item_price * item_cnt_day')
 .groupby(['shop_id', 'month', 'year'], as_index=False)['revenue'].sum()

You can, if you wish, query for January 2013 (before or after the above operations):

res = res.query('month == 1 & year == 2013')

print(res)

 shop_id month year revenue
0 25 1 2013 27.0
2 59 1 2013 25.0

I like filtering the dataframe first, to reduce number of unnecessary calculations:

df.query('month == 1 and year == 2013')
 .assign(all_revenue = df.item_price * df.item_cnt_day)
 .groupby(['shop_id','month','year'], as_index=False)['all_revenue'].sum()

Output:

 shop_id month year all_revenue
0 25 1 2013 27.0
1 59 1 2013 25.0

Note: Because your column names are "friendly", no spaces nor special characters, you can use query method. If that doesn't work for your column naming then you need to use boolean indexing.

df[(df['month'] == 1) & (df['year'] == 2013)]
 .assign(all_revenue = df.item_price * df.item_cnt_day)
 .groupby(['shop_id','month','year'], as_index=False)['all_revenue'].sum()