Adding DataFrames with offsets yields naive index

[craustin]

Adding two DataFrames with a DateRange index (with BusinessMonthEnd offset) yields a DataFrame with a simple 'Index'.

[craustin]

from pandas import DateRange, DataFrame
from pandas.core.datetools import BMonthEnd
a = DataFrame({}, index=DateRange('2011-01-01', '2011-11-30', offset=BMonthEnd()))
b = DataFrame({}, index=DateRange('2011-01-01', '2011-11-29', offset=BMonthEnd()))
print a+b

Yields:
Empty DataFrame
Columns: Index([], dtype=object)
Index: Index([2011-01-31 00:00:00, 2011-02-28 00:00:00, 2011-03-31 00:00:00,
2011-04-29 00:00:00, 2011-05-31 00:00:00, 2011-06-30 00:00:00,
2011-07-29 00:00:00, 2011-08-31 00:00:00, 2011-09-30 00:00:00,
2011-10-31 00:00:00, 2011-11-30 00:00:00], dtype=object)

In Pandas 0.4.0, yields:
Empty DataFrame
class 'pandas.core.daterange.DateRange'
offset: 1 BusinessMonthEnd, tzinfo: None
[2011-01-31 00:00:00, ..., 2011-11-30 00:00:00]
length: 11

Note, this doesn't just happen with empty DataFrames - I just did that for a simple repro.

[wesm]

Fixed in the above commit