possible pandas bug on agg via dict

gistfile1.txt

In [13]: %cpaste
Pasting code; enter '--' alone on the line to stop or use Ctrl-D.
:In [1]: df = pd.DataFrame({'A' : ['foo', 'bar', 'foo', 'bar',
:   ...:                           'foo', 'bar', 'foo', 'foo'],
:   ...:                    'B' : ['one', 'one', 'two', 'three',
:   ...:                           'two', 'two', 'one', 'three'],
:   ...:                    'C' : np.random.randn(8),
:   ...:                    'D' : np.random.randn(8)})
:--

In [14]: df
Out[14]:
     A      B         C         D
0  foo    one -0.508647 -1.037872
1  bar    one  0.232630 -1.498335
2  foo    two -0.294682  0.506412
3  bar  three -0.254762 -0.229931
4  foo    two  0.374847 -0.803666
5  bar    two  0.946507 -0.855568
6  foo    one -0.358013  0.433566
7  foo  three  0.555966  0.525245

In [15]: grouped = df.groupby(['A', 'B'])

In [19]: grouped['D'].agg({'r':np.sum, 'r2':np.mean})
Out[19]: 
            r        r2
A                      
bar -2.583835 -0.861278
foo -0.376315 -0.075263


In [20]: grouped[['D','C']].agg({'r':np.sum, 'r2':np.mean})
---------------------------------------------------------------------------
NotImplementedError                       Traceback (most recent call last)
<ipython-input-20-f26edb6d56b9> in <module>()
----> 1 grouped[['D','C']].agg({'r':np.sum, 'r2':np.mean})

/Users/aterrel/miniconda/envs/boldmetrics/lib/python2.7/site-packages/pandas/core/groupby.pyc in agg(self, func, *args, **kwargs)
    683     @Appender(_agg_doc)
    684     def agg(self, func, *args, **kwargs):
--> 685         return self.aggregate(func, *args, **kwargs)
    686 
    687     def _iterate_slices(self):

/Users/aterrel/miniconda/envs/boldmetrics/lib/python2.7/site-packages/pandas/core/groupby.pyc in aggregate(self, arg, *args, **kwargs)
   2657                 subset = obj
   2658                 if isinstance(subset, DataFrame):
-> 2659                     raise NotImplementedError("Aggregating on a DataFrame is "
   2660                                               "not supported")
   2661 

NotImplementedError: Aggregating on a DataFrame is not supported

In [21]: grouped[['D','C']].agg([np.sum, np.mean])
Out[21]: 
            D                   C          
          sum      mean       sum      mean
A                                          
bar -2.583835 -0.861278  0.924375  0.308125
foo -0.376315 -0.075263 -0.230529 -0.046106