BUG/COMPAT: Series bool type not promoted to object when introducing NaNs

[domspad]

Code Sample, a copy-pastable example if possible

In [2]: import pandas as pd

In [3]: a = pd.Series([True,False,True])

In [4]: a
Out[4]: 
0     True
1    False
2     True
dtype: bool

In [6]: a[0] = pd.np.nan

In [7]: a
Out[7]: 
0    NaN
1    0.0
2    1.0
dtype: float64

Expected Output

based on type promotions mentioned here I would expect the series to get promoted to object to allow for NaNs.

When introducing NAs into an existing Series or DataFrame via reindex or some other means, boolean and integer types will be promoted to a different dtype in order to store the NAs. These are summarized by this table:

In [7]: a
Out[7]: 
0      NaN
1    False
2     True
dtype: object

output of pd.show_versions()

In [2]: pd.show_versions()

INSTALLED VERSIONS
------------------
commit: None
python: 2.7.11.final.0
python-bits: 64
OS: Darwin
OS-release: 15.4.0
machine: x86_64
processor: i386
byteorder: little
LC_ALL: en_US.UTF-8
LANG: en_US

pandas: 0.18.0
nose: 1.3.7
pip: 8.1.1
setuptools: 20.6.7
Cython: None
numpy: 1.10.4
scipy: 0.17.0
statsmodels: None
xarray: None
IPython: 4.1.2
sphinx: 1.3.5
patsy: None
dateutil: 2.5.2
pytz: 2016.3
blosc: None
bottleneck: None
tables: None
numexpr: None
matplotlib: 1.5.1
openpyxl: None
xlrd: None
xlwt: None
xlsxwriter: None
lxml: None
bs4: 4.4.1
html5lib: 0.999
httplib2: None
apiclient: None
sqlalchemy: None
pymysql: None
psycopg2: None
jinja2: 2.8
boto: 2.39.0```

[jreback]

yeah that should coerce to object. Right now we are promoting a boolean with nulls -> float64

this is somewhat deep in there. need to look inside core/internals/Block.setitem. where _maybe_promote is called. Bools have not gotten much love lately.

Though to be fair we do this similar to how we do ints with NA's, meaning we cast to floats.

So not sure how much this is relied upon in the real world.

[mroeschke]

This looks to work on master now. Could use a test

In [3]: In [2]: import pandas as pd
   ...:
   ...: In [3]: a = pd.Series([True,False,True])

In [4]: a[0] = np.nan

In [5]: a
Out[5]:
0      NaN
1    False
2     True
dtype: object