Error when key-only Grouper is passed to groupby in a list

[jmmease]

Overview

A Grouper object configured with only a key specification may be passed to groupby to group a DataFrame by a particular column. For example:

In [19]: import pandas as pd

In [20]: df = pd.DataFrame({'A': [0, 0, 0, 1, 1, 1], 
    ...:                    'B': [1, 1, 2, 2, 3, 3],
    ...:                    'C': [1, 2, 3, 4, 5, 6]})

In [26]: df.groupby(pd.Grouper(key='A')).count()
Out[26]: 
   B  C
A      
0  3  3
1  3  3

However, when this key-only Grouper is passed to groupby inside a list, a TypeError is thrown. The following example is for a scalar list containing the same Grouper object as in the previous example.

In [27]: df.groupby([pd.Grouper(key='A')]).count()
---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
<ipython-input-27-f4f86763ebfc> in <module>()
----> 1 df.groupby([pd.Grouper(key='A')]).count()

/Users/measejm1/anaconda/lib/python3.5/site-packages/pandas/core/generic.py in groupby(self, by, axis, level, as_index, sort, group_keys, squeeze, **kwargs)
   3776         return groupby(self, by=by, axis=axis, level=level, as_index=as_index,
   3777                        sort=sort, group_keys=group_keys, squeeze=squeeze,
-> 3778                        **kwargs)
   3779 
   3780     def asfreq(self, freq, method=None, how=None, normalize=False):

/Users/measejm1/anaconda/lib/python3.5/site-packages/pandas/core/groupby.py in groupby(obj, by, **kwds)
   1425         raise TypeError('invalid type: %s' % type(obj))
   1426 
-> 1427     return klass(obj, by, **kwds)
   1428 
   1429 

/Users/measejm1/anaconda/lib/python3.5/site-packages/pandas/core/groupby.py in __init__(self, obj, keys, axis, level, grouper, exclusions, selection, as_index, sort, group_keys, squeeze, **kwargs)
    352                                                     level=level,
    353                                                     sort=sort,
--> 354                                                     mutated=self.mutated)
    355 
    356         self.obj = obj

/Users/measejm1/anaconda/lib/python3.5/site-packages/pandas/core/groupby.py in _get_grouper(obj, key, axis, level, sort, mutated)
   2400                         sort=sort,
   2401                         in_axis=in_axis) \
-> 2402             if not isinstance(gpr, Grouping) else gpr
   2403 
   2404         groupings.append(ping)

/Users/measejm1/anaconda/lib/python3.5/site-packages/pandas/core/groupby.py in __init__(self, index, grouper, obj, name, level, sort, in_axis)
   2197 
   2198                 # get the new grouper
-> 2199                 grouper = self.grouper._get_binner_for_grouping(self.obj)
   2200                 self.obj = self.grouper.obj
   2201                 self.grouper = grouper

/Users/measejm1/anaconda/lib/python3.5/site-packages/pandas/core/groupby.py in _get_binner_for_grouping(self, obj)
    290         group_axis = obj._get_axis(self.axis)
    291         return Grouping(group_axis, None, obj=obj, name=self.key,
--> 292                         level=self.level, sort=self.sort, in_axis=False)
    293 
    294     @property

/Users/measejm1/anaconda/lib/python3.5/site-packages/pandas/core/groupby.py in __init__(self, index, grouper, obj, name, level, sort, in_axis)
   2213                     t = self.name or str(type(self.grouper))
   2214                     raise ValueError("Grouper for '%s' not 1-dimensional" % t)
-> 2215                 self.grouper = self.index.map(self.grouper)
   2216                 if not (hasattr(self.grouper, "__len__") and
   2217                         len(self.grouper) == len(self.index)):

/Users/measejm1/anaconda/lib/python3.5/site-packages/pandas/indexes/base.py in map(self, mapper)
   2238         applied : array
   2239         """
-> 2240         return self._arrmap(self.values, mapper)
   2241 
   2242     def isin(self, values, level=None):

pandas/src/generated.pyx in pandas.algos.arrmap_int64 (pandas/algos.c:94003)()

TypeError: 'NoneType' object is not callable

The same error occurs for lists containing multiple key-only Groupers. e.g. [pd.Grouper(key='A'), pd.Grouper(key='B')]

Expected Output

In [27]: df.groupby([pd.Grouper(key='A')]).count()
Out[27]: 
   B  C
A      
0  3  3
1  3  3

Output of pd.show_versions()

## INSTALLED VERSIONS

commit: None
python: 3.5.2.final.0
python-bits: 64
OS: Darwin
OS-release: 15.6.0
machine: x86_64
processor: i386
byteorder: little
LC_ALL: None
LANG: en_US.UTF-8

pandas: 0.18.1
nose: 1.3.7
pip: 8.1.2
setuptools: 23.0.0
Cython: 0.24
numpy: 1.11.1
scipy: 0.17.1
statsmodels: 0.6.1
xarray: None
IPython: 4.2.0
sphinx: 1.4.1
patsy: 0.4.1
dateutil: 2.5.3
pytz: 2016.4
blosc: 1.4.1
bottleneck: 1.1.0
tables: 3.2.2
numexpr: 2.6.0
matplotlib: 1.5.1
openpyxl: 2.3.2
xlrd: 1.0.0
xlwt: 1.1.2
xlsxwriter: 0.9.2
lxml: 3.6.0
bs4: 4.4.1
html5lib: 0.999
httplib2: None
apiclient: None
sqlalchemy: 1.0.13
pymysql: None
psycopg2: None
jinja2: 2.8
boto: 2.40.0
pandas_datareader: None

[chris-b1]

This probably should work, thanks for the report! But I'm curious, why are using a Grouper in this case instead of df.groupby(['A']), df.groupby(['A','B']), etc?

[jmmease]

Yeah, good question.

I'm in the process of working on implementing #5677 which will allow the A in df.groupby(['A']) to refer to an index level in df named A if df has no column named A. Per the conversation in #5677, If a frame has both a column named A and an index level named A then df.groupby(['A']) will group by the column (this maintains backwards compatibility). I came across this bug in the process of writing tests to verify that this ambiguity is resolved properly.