Support for empty slice in multi indices

[SylvainCorlay]

It is currently possible to pass a list of labels for a given dataframe to DataFrame.loc in the case of a multi index, but the list cannot be empty unlike in the case of a non-hierarchical index.

[SylvainCorlay]

Ping @SSunkara .

[jreback]

hmm this should work
can you give a small example (copy/pastable) of what you are doing and print of.show_versions()

[SylvainCorlay]

Here it is:

import pandas as pd
import numpy as np
multi_index = pd.MultiIndex.from_product((['foo', 'bar', 'baz'], ['alpha', 'beta']))
df = pd.DataFrame(np.random.randn(5, 6), index=range(5), columns=multi_index)
df = df.sortlevel(0, axis=1)
df.loc[:, ([], slice(None))]

The version is 0.14.0.

Here is the error:

---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
<ipython-input-5-15460543620a> in <module>()
      5 df = pd.DataFrame(np.random.randn(5, 6), index=range(5), columns=multi_index)
      6 df = df.sortlevel(0, axis=1)
----> 7 df.loc[:, ([], slice(None))]

/opt/anaconda20/lib/python2.7/site-packages/pandas/core/indexing.pyc in __getitem__(self, key)
   1125     def __getitem__(self, key):
   1126         if type(key) is tuple:
-> 1127             return self._getitem_tuple(key)
   1128         else:
   1129             return self._getitem_axis(key, axis=0)

/opt/anaconda20/lib/python2.7/site-packages/pandas/core/indexing.pyc in _getitem_tuple(self, tup)
    643     def _getitem_tuple(self, tup):
    644         try:
--> 645             return self._getitem_lowerdim(tup)
    646         except IndexingError:
    647             pass

/opt/anaconda20/lib/python2.7/site-packages/pandas/core/indexing.pyc in _getitem_lowerdim(self, tup)
    751         # we may have a nested tuples indexer here
    752         if self._is_nested_tuple_indexer(tup):
--> 753             return self._getitem_nested_tuple(tup)
    754 
    755         # we maybe be using a tuple to represent multiple dimensions here

/opt/anaconda20/lib/python2.7/site-packages/pandas/core/indexing.pyc in _getitem_nested_tuple(self, tup)
    823 
    824             current_ndim = obj.ndim
--> 825             obj = getattr(obj, self.name)._getitem_axis(key, axis=axis, validate_iterable=True)
    826             axis += 1
    827 

/opt/anaconda20/lib/python2.7/site-packages/pandas/core/indexing.pyc in _getitem_axis(self, key, axis, validate_iterable)
   1254             return self._getitem_iterable(key, axis=axis)
   1255         elif _is_nested_tuple(key, labels):
-> 1256             locs = labels.get_locs(key)
   1257             indexer = [ slice(None) ] * self.ndim
   1258             indexer[axis] = locs

/opt/anaconda20/lib/python2.7/site-packages/pandas/core/index.pyc in get_locs(self, tup)
   3579                 ranges.append(reduce(
   3580                     np.logical_or,[ _convert_indexer(self._get_level_indexer(x, level=i)
-> 3581                                                      ) for x in k ]))
   3582             elif k == slice(None):
   3583                 # include all from this level

TypeError: reduce() of empty sequence with no initial value

Printout of show_versions:

INSTALLED VERSIONS
------------------
commit: None
python: 2.7.8.final.0
python-bits: 64
OS: Linux
OS-release: 3.13.0-24-generic
machine: x86_64
processor: x86_64
byteorder: little
LC_ALL: None
LANG: None

pandas: 0.14.0
nose: 1.3.3
Cython: 0.20.1
numpy: 1.8.1
scipy: 0.14.0
statsmodels: 0.5.0
IPython: 3.0.0-dev
sphinx: 1.2.2
patsy: 0.2.1
scikits.timeseries: None
dateutil: 1.5
pytz: 2014.3
bottleneck: None
tables: 3.1.1
numexpr: 2.3.1
matplotlib: 1.3.1
openpyxl: 1.8.5
xlrd: 0.9.3
xlwt: 0.7.5
xlsxwriter: 0.5.5
lxml: 3.3.5
bs4: 4.3.1
html5lib: None
bq: None
apiclient: None
rpy2: 2.4.3
sqlalchemy: 0.9.4
pymysql: None
psycopg2: None

[jreback]

I see it..ok will see if can get in for 0.15.1 (releasing this weekend). thanks for the report.

[jreback]

so this should return an empty frame (with the same index, and column index metadata preserved)

e.g.

DataFrame(index=df.index,columns=multi_index.reindex([])[0])

yes?

[SylvainCorlay]

This is what I was expecting from the documentation, but wasn't sure if I was doing something wrong. Hopefully next time, we will open a PR with a correction!

[jreback]

done (was pretty trivial), sadly don't always catch all of the cases. Note that you should for sure upgrade to 0.14.1 as lots of changes in mi slicing. (of course 0.15.1 even better) :)

[jreback]

closed by #8739