Bitwise operations have inconsistent behavior, different from numpy

[cyrusmaher]

Code Sample

# succeeds
pd.Series([False]) & pd.Series([6.])

# Example: order matters
# fails: ufunc 'bitwise_and' not supported for the input types
pd.Series([6.]) & pd.Series([False])

# Example: behavior is different from numpy
# fails: ufunc 'bitwise_and' not supported for the input types
np.array([False]) & np.array([6.])

Problem description

Bitwise operations between floats and bools error out in numpy. They error out in pandas too if the first argument is a float, but not if the first argument is a bool.

Expected Output

ufunc 'bitwise_and' not supported for the input types

Output of pd.show_versions()

INSTALLED VERSIONS

commit: None
python: 3.6.3.final.0
python-bits: 64
OS: Darwin
OS-release: 17.7.0
machine: x86_64
processor: i386
byteorder: little
LC_ALL: None
LANG: en_US.UTF-8
LOCALE: en_US.UTF-8

pandas: 0.23.4
pytest: None
pip: 18.0
setuptools: 36.5.0.post20170921
Cython: 0.28.2
numpy: 1.15.2
scipy: 1.1.0
pyarrow: None
xarray: None
IPython: 6.1.0
sphinx: None
patsy: 0.5.0
dateutil: 2.6.1
pytz: 2017.2
blosc: None
bottleneck: None
tables: None
numexpr: None
feather: None
matplotlib: 2.2.2
openpyxl: None
xlrd: 1.1.0
xlwt: None
xlsxwriter: None
lxml: None
bs4: None
html5lib: 0.999999999
sqlalchemy: None
pymysql: None
psycopg2: None
jinja2: 2.9.6
s3fs: 0.1.5
fastparquet: None
pandas_gbq: None
pandas_datareader: None

[mfenner1]

In a similar vein:

>>> df = pd.DataFrame({'likes_hockey' : [True, True, False],
...                    'likes_soccer' : [True, False, False]})


>>> df.likes_hockey & df.likes_soccer # works for series
0     True
1    False
2    False
dtype: bool


>>> # df.bitwise_and(df.likes_soccer, axis='rows') # desirable, but not implemented

>>> df.mul(df.likes_soccer, axis='rows') # implemented, but complains and recommends '&'
/Users/mfenner/anaconda3/lib/python3.7/site-packages/pandas/core/computation/expressions.py:178: 
UserWarning: evaluating in Python space because the '*' operator is not supported by numexpr for the 
bool dtype, use '&' instead
f"evaluating in Python space because the {repr(op_str)} "

   likes_hockey  likes_soccer
0          True          True
1         False         False
2         False         False

>>> # but neither of these broadcast/align appropriately (which is expected):
>>> # df & df.likes_soccer  # fails
>>> # df.likes_soccer & df  # fails

Much of this might apply to np.logical_and as well (and the other bitwise_ and logical_ operators/functions).