Setting values on Series using .loc with a TZ-aware DatetimeIndex fails

[coroa]

In [1]: import pandas as pd

In [2]: index = pd.date_range('2011-01-01', '2011-01-02', tz='utc')   # without tz='utc' it works

In [3]: index
Out[3]: DatetimeIndex(['2011-01-01', '2011-01-02'], dtype='datetime64[ns, UTC]', freq='D')

In [4]: s = pd.Series(range(2), index=index)

In [5]: s.loc[index] = 5
---------------------------------------------------------------------------
KeyError                                  Traceback (most recent call last)
<ipython-input-5-992315c98f0d> in <module>()
----> 1 s.loc[index] = 5

/usr/lib/python2.7/dist-packages/pandas/core/indexing.pyc in __setitem__(self, key, value)
    114 
    115     def __setitem__(self, key, value):
--> 116         indexer = self._get_setitem_indexer(key)
    117         self._setitem_with_indexer(indexer, value)
    118 

/usr/lib/python2.7/dist-packages/pandas/core/indexing.pyc in _get_setitem_indexer(self, key)
    109 
    110         try:
--> 111             return self._convert_to_indexer(key, is_setter=True)
    112         except TypeError:
    113             raise IndexingError(key)

/usr/lib/python2.7/dist-packages/pandas/core/indexing.pyc in _convert_to_indexer(self, obj, axis, is_setter)
   1148                 mask = check == -1
   1149                 if mask.any():
-> 1150                     raise KeyError('%s not in index' % objarr[mask])
   1151 
   1152                 return _values_from_object(indexer)

KeyError: "['2011-01-01T01:00:00.000000000+0100' '2011-01-02T01:00:00.000000000+0100'] not in index"

On python 2.7.11, pandas 0.17.1, linux.

[jreback]

so getitem and .loc getting working as well as setitem, but setting with .loc seems broken (only with list-likes).

In [9]: s[index]
Out[9]: 
2011-01-01 00:00:00+00:00    0
2011-01-02 00:00:00+00:00    1
Freq: D, dtype: int64

In [10]: s.loc[index]
Out[10]: 
2011-01-01 00:00:00+00:00    0
2011-01-02 00:00:00+00:00    1
Freq: D, dtype: int64

In [11]: s[index] = 5

In [12]: s
Out[12]: 
2011-01-01 00:00:00+00:00    5
2011-01-02 00:00:00+00:00    5
Freq: D, dtype: int64

# setting single items
In [14]: s.loc[index[0]] = 10

In [15]: s
Out[15]: 
2011-01-01 00:00:00+00:00    10
2011-01-02 00:00:00+00:00     5
Freq: D, dtype: int64

[jreback]

closed by #12054