CODEWITHSUNDEEP
pythonpandas
NBC
NBC

Reputation: 1698

Groupby Pandas dataframe and drop values conditionally based on rank

I have a dataframe like this:

Col1 | Col2 | Col3
 a   |   8  |  9
 a   |   3  |  7
 a   |   1  |  3
 a   |   0  |  8
 b   |   6  |  18
 b   |   2  |  6

I would like to drop everything but the 2nd and 3rd top value for Col2, by grouping Col1, assuming it's possible

Output desired:

Col1 | Col2 | Col3
 a   |   3  |  7
 a   |   1  |  3
 b   |   2  |  6

Upvotes: 0

Views: 97

Answers (1)

BENY
BENY

Reputation: 323306

It is possible by using cumcount

df[df.groupby('Col1').cumcount().isin([1,2])]
Out[423]: 
  Col1  Col2  Col3
1    a     2     7
2    a     1     3
5    b     2     6

More information : 

df.groupby('Col1').cumcount()
Out[435]: 
0    0
1    1
2    2
3    3
4    0
5    1
dtype: int64

Upvotes: 1

Related Questions