Unstacking boolean DataFrame fills 'True' for missing values

[warrenwjackson]

Is this a bug? If this is intentional, it is pretty dangerous to fill missing values with True.

>>> df = pd.DataFrame([False, False], index=pd.MultiIndex.from_arrays([['a', 'b'], ['c', 'l']]), columns=['col'])
>>> df
       col
a c  False
b l  False
>>> df.unstack()
     col
       c      l
a  False   True
b   True  False

[changhiskhan]

Unstack now converts bool dtype to object dtype