Pandas groupby: get size of a group knowing its id (from .grouper.group_info[0])

Question

In the following snippet data is a pandas.DataFrame and indices is a set of columns of the data. After grouping the data with groupby I am interested in the ids of the groups, but only those with a size greater than a threshold (say: 3).

group_ids=data.groupby(list(data.columns[list(indices)])).grouper.group_info[0]

Now, how can I find which group has a size greater than or equal 3 knowing the id of the group? I only want ids of groups with a certain size.

#TODO: filter out ids from group_ids which correspond to groups with sizes < 3 

Answer 1

One way is to use the size method of the groupby:

g = data.groupby(...)
size = g.size()
size[size > 3]

For example, here there is only one group of size > 1:

In [11]: df = pd.DataFrame([[1, 2], [3, 4], [1,6]], columns=['A', 'B'])

In [12]: df
Out[12]:
   A  B
0  1  2
1  3  4
2  1  6 

In [13]: g = df.groupby('A')

In [14]: size = g.size()

In [15]: size[size > 1]
Out[15]:
A
1    2
dtype: int64

If you were interested in just restricting the DataFrame to those in large groups you could use the filter method:

In [21]: g.filter(lambda x: len(x) > 1)
Out[21]:
   A  B
0  1  2
2  1  6