Dataframe group by only groups with 2 or more rows

Question

Is there a way to groupby only groups with 2 or more rows?

Or can I delete groups from a grouped dataframe that contains only 1 row?

Thank you very much for your help!

Answer 1

Yes there is a way. Here is an example below

df = pd.DataFrame(
    np.array([['A','A','B','C','C'],[1,2,1,1,2]]).T
    , columns= ['type','value']
    )
groups = df.groupby('type')
groups_without_single_row_df = [g for g in groups if len(g[1]) > 1]

groupby return a list of tuples. Here, 'type' (A, b or C) is the first element of the tuple and the subdataframe the second element.

You can check length of each subdataframe with len() as in [g for g in groups if len(g[1]) > 1] where we check the lengh of the second element of the tuple. If the the len() is greater than 1, it is include in the ouput list.

Hope it helps