How to get a specific row from a groupby object?

If I have a dataframe groupby object thats been grouped by id –

sl.no   ID      hour
50834   1       6
50833   1       6
50832   1       6
19188   2       9
19187   2       9
19186   2       9
13380   3       294
13379   3       294
13378   3       294

which can be created by:

data = {
    'sl.no': [50834, 50833, 50832, 19188, 19187, 19186, 13380, 13379, 13378],
    'ID': [1, 1, 1, 2, 2, 2, 3, 3, 3],
    'hour': [6, 6, 6, 9, 9, 9, 294, 294, 294]
     }
df = pd.DataFrame(data)

dfnew = df.sort_values(['sl.no'],ascending=False).groupby('ID').head(3)

I want to extract the nth row of each group, so for example if I wanted the 2nd row the output would be-

sl.no   ID      hour
50833   1       6
19187   2       9
13379   3       294

I tried using iloc(2) which I found out wasn’t possible. I found a convoluted way to do this by getting head(2) and finding the difference with head(1) but is there an easier / direct way to do this?

>Solution :

Code

boolean indexing by groupby + cumcount

out = df[df.groupby('ID').cumcount().eq(1)]

out:

   sl.no  ID  hour
1  50833   1     6
4  19187   2     9
7  13379   3   294

or you can use groupby + nth

out = df.groupby('ID').nth(1)

same result

both codes will not extract anything for groups where the 2nd row does not exist.