Group by multiple columns pandas

Asked Feb 05 '20 at 13:59

Active Feb 05 '20 at 13:59

Viewed 18 times

I have this dataframe:

a = ["A", "A", "B", "B", "A", "A"]
b = [12, 15, 14, 17, 19, 20]
c = ["Area1", "Area1", "Area2", "Area2", "Area3", "Area3"]

df = pd.DataFrame({'Item': a, 'Quanity': b, 'Region': c})

I want to get the sum for each Item and for each Region, the result should be like this:

Item  Region  total
A     Area1   27
B     Area2   31
A     Area3   39

I have tried this method:

df['total'] = df.groupby(['Item', 'Region'])['Quanity'].transform('sum')
df[["Item", "Region", "total"]].drop_duplicates(keep='first')

However this method causes me to create another column and another dataframe (maybe). So are there any way faster (or more efficient) to do this task?

asked Feb 05 '20 at 13:59

Long_NgV

1

df.groupby(['Item','Region']).agg(total=('Quanity','sum')) – sammywemmy Feb 05 '20 at 14:02
1

Use `df.groupby(['Item', 'Region'], sort=False)['Quanity'].sum().reset_index(name='total')` – jezrael Feb 05 '20 at 14:02

Group by multiple columns pandas

0 Answers0