Chained groupby-resample object raises error when used twice with different column selector

[jer0nim0]

Going from 0.18.0 to 0.18.1, a chained groupby - resample object can now only be used once. A second similar operation raises a KeyError when a column that is not used in the first operation is selected.

Code Sample, a copy-pastable example if possible

N = 5
df = pd.DataFrame({'a': np.ones(N), 'b': np.ones(N), 'c': np.ones(N)}, 
                  index=pd.date_range('2016-01-01', periods=N))
group = df.groupby('a').resample('1D')
# OK:
print group[['b']].count()
# Error:
print group[['c']].count()

Actual Output

                b
a                
1.0 2016-01-01  1
    2016-01-02  1
    2016-01-03  1
    2016-01-04  1
    2016-01-05  1

---------------------------------------------------------------------------
KeyError                                  Traceback (most recent call last)
<ipython-input-101-875a81207428> in <module>()
----> 1 group[['c']].count()

C:\Miniconda2\lib\site-packages\pandas\tseries\resample.pyc in __getitem__(self, key)
    174     def __getitem__(self, key):
    175         try:
--> 176             return super(Resampler, self).__getitem__(key)
    177         except (KeyError, com.AbstractMethodError):
    178 

C:\Miniconda2\lib\site-packages\pandas\core\base.pyc in __getitem__(self, key)
    331                 bad_keys = list(set(key).difference(self.obj.columns))
    332                 raise KeyError("Columns not found: %s"
--> 333                                % str(bad_keys)[1:-1])
    334             return self._gotitem(list(key), ndim=2)
    335 

KeyError: "Columns not found: 'c'"

Expected Output

                b
a                
1.0 2016-01-01  1
    2016-01-02  1
    2016-01-03  1
    2016-01-04  1
    2016-01-05  1

                c
a                
1.0 2016-01-01  1
    2016-01-02  1
    2016-01-03  1
    2016-01-04  1
    2016-01-05  1

output of pd.show_versions()

INSTALLED VERSIONS

commit: None
python: 2.7.12.final.0
python-bits: 64
OS: Windows
OS-release: 7
machine: AMD64
processor: Intel64 Family 6 Model 58 Stepping 9, GenuineIntel
byteorder: little
LC_ALL: None
LANG: None

pandas: 0.18.1
nose: None
pip: 8.1.2
setuptools: 23.0.0
Cython: None
numpy: 1.11.1
scipy: 0.17.1
statsmodels: 0.6.1
xarray: None
IPython: 4.1.2
sphinx: 1.3.5
patsy: 0.4.1
dateutil: 2.5.3
pytz: 2016.6.1
blosc: None
bottleneck: None
tables: None
numexpr: None
matplotlib: 1.5.1
openpyxl: None
xlrd: None
xlwt: None
xlsxwriter: None
lxml: None
bs4: None
html5lib: None
httplib2: None
apiclient: None
sqlalchemy: None
pymysql: None
psycopg2: None
jinja2: 2.8
boto: None
pandas_datareader: None

[TomAugspurger]

Thanks for the report. This has been fixed on master (#13175 I believe).

The release candidate for 0.19 should be out in a few weeks.