import pandas as pd
import numpy as np
rand = np.random.RandomState(1)
df = pd.DataFrame({'A': ['foo', 'bar', 'baz'] * 10,
                   'B': [rand.choice(['cat', 'dog', 'fish', 'pig', 'cow']) for i in range(30)],
                   'C': 1})

>> df.head(5)
     A    B  C
0  foo  pig  1
1  bar  cow  1
2  baz  cat  1
3  foo  dog  1
4  bar  pig  1

>> d = df.groupby(['A','B']).sum();
>> d = d.groupby(level=0, group_keys=False).apply(lambda x: x.sort_values('C', ascending=False)); d

          C
A   B
bar dog   4
    cow   2
    fish  2
    cat   1
    pig   1
baz cow   4
    cat   3
    fish  2
    dog   1
foo dog   4
    cow   3
    pig   2
    cat   1

def summarise(l, n=10, name='Other'):
    h = l.head(n)

    idx = l.index[0]
    if isinstance(idx, (list, tuple)):
        prefix = list(idx[:-1])
    else:
        prefix = []
    return h.append(pd.DataFrame([l.tail(-n).sum()], columns=l.columns, index=[tuple(prefix+[name])]))

>> summarise(d, n=2)
            C
A   B
bar dog     4
    cow     2
    Other  24

A     B    C
bar   dog    4
bar   cow    2
bar   Other  4
baz   cow    4
baz   cat    3
baz   Other  3
foo   dog    4
foo   cow    3
foo   Other  3