熊猫如何在分组的同时在另一列上进行分组

Question

I have the following df ,

id    amount
1     20
2     8
1     3
1     2
2     7

I want to groupby the df by id , and sorting the groups by their sizes,

 df.groupby('id').size().sort_values(ascending=False)

but also aggregate on amount of each group to create a separate column total at the same time,

id    amount    total    size
1     20         25       3
1     3          25       3
1     2          25       3
2     8          15       2
2     7          15       2

Answer 1

You can use GroupBy + agg with a list, followed by pd.merge :

g = df.groupby('id')['amount'].agg(['size', 'sum'])

res = pd.merge(df, g, left_on='id', right_index=True)\
        .sort_values('size', ascending=False)

print(res)

   id  amount  size  sum
0   1      20     3   25
2   1       3     3   25
3   1       2     3   25
1   2       8     2   15
4   2       7     2   15