pandas groupby with value counts

Asked Apr 03 '19 at 12:45

Active Apr 03 '19 at 20:23

Viewed 29 times

I have a pandas dataframe.

df = pd.DataFrame({'A':['ant','ant','cherry', 'dog', 'ant'], 'B':['animal','animal', 'fruit', 'animal', 'animal'], 'C':['small','small','small', 'big', 'small']})

     A        B       C
0   ant    animal   small
1   ant    animal   small
2   cherry  fruit   small
3   dog    animal   big
4   ant   animal    small

I need to do a groupby so that the result is in the following form.

     A        B       C    Count_A
0   ant    animal   small    3
1   cherry  fruit   small    1
2   dog    animal   big      1

I have tried the following:

s = df.groupby(['A','B','C'])['A'].count()

The result is what I needed. However, I am unable to do any slicing on it since, all columns 'A','B' and 'C' have become the index now. the method .reset_index() also does not work. I get a value error ValueError: cannot insert A, already exists.

Would anyone be able to help me regarding this?

edited Apr 03 '19 at 20:23

jkdev

11,360
15
54
77

asked Apr 03 '19 at 12:45

Alain Michael Janith Schroter

2

`df.groupby(['A','B','C']).size() ` and your code work for me – BENY Apr 03 '19 at 12:47
3

Use `df.groupby(['A','B','C']).size().reset_index(name='Count_A')` – jezrael Apr 03 '19 at 12:48
ok. thanks. That helped. – Alain Michael Janith Schroter Apr 03 '19 at 13:56

pandas groupby with value counts

0 Answers0