BUG: TypeError when trying to set values in empty DataFrame using boolean key

[rekcahpassyla]

Found this bug through #10188 - it caused build failure

import numpy as np
import pandas as pd

pd.show_versions()

df = pd.DataFrame([[2, 2]], index=pd.DatetimeIndex(['2011-01-01']))
df2 = pd.DataFrame([[1, 1]], index=pd.DatetimeIndex(['2011-01-01']))

df[df > df2] = 47

assert np.all(df.values==47)

df = pd.DataFrame()
df2 = pd.DataFrame()

df[df > df2] = 47

results in this error (in both 0.15.2 and 0.16.1)

INSTALLED VERSIONS
------------------
commit: None
python: 2.7.9.final.0
python-bits: 64
OS: Windows
OS-release: 7
machine: AMD64
processor: Intel64 Family 6 Model 26 Stepping 5, GenuineIntel
byteorder: little
LC_ALL: None
LANG: None

pandas: 0.16.1-46-g0aceb38
nose: 1.3.6
Cython: 0.22
numpy: 1.9.2
scipy: 0.14.0
statsmodels: 0.6.1
IPython: 3.1.0
sphinx: 1.3.1
patsy: 0.3.0
dateutil: 2.4.2
pytz: 2015.4
bottleneck: 0.8.0
tables: 3.1.1
numexpr: 2.3.1
matplotlib: 1.4.3
openpyxl: None
xlrd: 0.9.3
xlwt: None
xlsxwriter: 0.7.2
lxml: 3.4.4
bs4: 4.3.2
html5lib: 0.999
httplib2: None
apiclient: None
sqlalchemy: 1.0.4
pymysql: None
psycopg2: None
---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
c:\test_set_empty_frame_with_boolean.py in <module>()
     14 df2 = pd.DataFrame()
     15
---> 16 df[df > df2] = 47
     17

c:\python\envs\pandas-0.16.1\lib\site-packages\pandas\core\frame.pyc in __setitem__(self, key, value)
   2119             self._setitem_array(key, value)
   2120         elif isinstance(key, DataFrame):
-> 2121             self._setitem_frame(key, value)
   2122         else:
   2123             # set column

c:\python\envs\pandas-0.16.1\lib\site-packages\pandas\core\frame.pyc in _setitem_frame(self, key, value)
   2153         # df[df > df2] = 0
   2154         if key.values.dtype != np.bool_:
-> 2155             raise TypeError('Must pass DataFrame with boolean values only')
   2156
   2157         self._check_inplace_setting(value)

TypeError: Must pass DataFrame with boolean values only

Proposed #10196

[jreback]

this is a dupe of #10126