BUG: groupby with datetime and cythonized functions regression from 0.12

[jreback]

In [30]: DataFrame(dict(A = Timestamp('20130101'), B = np.arange(5))).groupby('A')['A'].max()
Out[30]: 
A
2013-01-01    1356998400000000000
Name: A, dtype: int64

should match

In [31]: DataFrame(dict(A = Timestamp('20130101'), B = np.arange(5))).groupby('A')['A'].apply(lambda x: x.max())
Out[31]: 
A
2013-01-01   2013-01-01 00:00:00
dtype: datetime64[ns]