BUG: multiple grouping with a TimeGrouper requires sort

[jreback]

resampling has been fixed, so this is only with 2 or more groupers (#6516)

In [9]: 
df = DataFrame({
            'date' : pd.to_datetime([
                '20121002','20121007','20130130','20130202','20130305','20121002',
                '20121207','20130130','20130202','20130305','20130202','20130305']),
            'user_id' : [1,1,1,1,1,3,3,3,5,5,5,5],
            'whole_cost' : [1790,364,280,259,201,623,90,312,359,301,359,801],
            'cost1' : [12,15,10,24,39,1,0,90,45,34,1,12] }).set_index('date')

        expected = df.groupby('user_id')['whole_cost'].resample(
            'M', how='sum').dropna().reorder_levels(['date','user_id']).sortlevel().astype('int64')
        expected.name = 'whole_cost'

In [10]: expected
Out[10]: 
date        user_id
2012-10-31  1          2154
            3           623
2012-12-31  3            90
2013-01-31  1           280
            3           312
2013-02-28  1           259
            5           718
2013-03-31  1           201
            5          1102
Name: whole_cost, dtype: int64

These should be equivalent

In [11]: df.sort_index().groupby([pd.TimeGrouper(freq='M'), 'user_id'])['whole_cost'].sum()
Out[11]: 
date        user_id
2012-10-31  1          2154
            3           623
2012-12-31  3            90
2013-01-31  1           280
            3           312
2013-02-28  1           259
            5           718
2013-03-31  1           201
            5          1102
Name: whole_cost, dtype: int64

In [13]: df.groupby([pd.TimeGrouper(freq='M'), 'user_id'])['whole_cost'].sum()
ValueError: cannot reindex from a duplicate axis

[sinhrks]

Followings are results after #6908. Looks OK.

>>> df.sort_index().groupby([pd.TimeGrouper(freq='M'), 'user_id'])['whole_cost'].sum()
date        user_id
2012-10-31  1          2154
            3           623
2012-12-31  3            90
2013-01-31  1           280
            3           312
2013-02-28  1           259
            5           718
2013-03-31  1           201
            5          1102
Name: whole_cost, dtype: int64

>>> df.groupby([pd.TimeGrouper(freq='M'), 'user_id'])['whole_cost'].sum()
date        user_id
2012-10-31  1          2154
            3           623
2012-12-31  3            90
2013-01-31  1           280
            3           312
2013-02-28  1           259
            5           718
2013-03-31  1           201
            5          1102
Name: whole_cost, dtype: int64

[jreback]

gr8 I didn't looks thru you tests but pls add these on unless yours fully cover
and ref this issue