BUG: Series.where doesn't work with empty lists

[jluttine]

• I have checked that this issue has not already been reported.

• I have confirmed this bug exists on the latest version of pandas. (Checked 1.0.3)

• (optional) I have confirmed this bug exists on the master branch of pandas.

---

Note: Please read this guide detailing how to provide the necessary information for us to reproduce your bug.

Code Sample, a copy-pastable example

>>> pd.Series([], dtype=float).where([])
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/nix/store/vqd2pmsja84h36wz6c878bgbpz46cy2h-python3.7-pandas-1.0.3/lib/python3.7/site-packages/pandas/core/generic.py", line 8919, in where
    cond, other, inplace, axis, level, errors=errors, try_cast=try_cast
  File "/nix/store/vqd2pmsja84h36wz6c878bgbpz46cy2h-python3.7-pandas-1.0.3/lib/python3.7/site-packages/pandas/core/generic.py", line 8673, in _where
    raise ValueError(msg.format(dtype=cond.dtype))
ValueError: Boolean array expected for the condition, not float64

For comparison, if the list isn't empty, it works:

>>> pd.Series([42], dtype=float).where([True])
0    42
dtype: int64

Problem description

[this should explain why the current behaviour is a problem and why the expected output is a better solution]

If where accepts lists as its mask argument, it should work for empty lists too. Either don't support lists at all or support empty lists too.

Expected Output

>>> pd.Series([], dtype=float).where([])
Series([], dtype: float64)

Output of pd.show_versions()

pd.show_versions()

INSTALLED VERSIONS

commit : None
python : 3.7.7.final.0
python-bits : 64
OS : Linux
OS-release : 5.4.41
machine : x86_64
processor :
byteorder : little
LC_ALL : None
LANG : en_US.UTF-8
LOCALE : en_US.UTF-8

pandas : 1.0.3
numpy : 1.18.3
pytz : 2019.3
dateutil : 2.8.1
pip : None
setuptools : None
Cython : None
pytest : None
hypothesis : None
sphinx : None
blosc : None
feather : None
xlsxwriter : None
lxml.etree : 4.5.0
html5lib : 1.0.1
pymysql : None
psycopg2 : None
jinja2 : None
IPython : None
pandas_datareader: None
bs4 : 4.8.2
bottleneck : 1.3.1
fastparquet : None
gcsfs : None
lxml.etree : 4.5.0
matplotlib : None
numexpr : 2.7.1
odfpy : None
openpyxl : 3.0.2
pandas_gbq : None
pyarrow : None
pytables : None
pytest : None
pyxlsb : None
s3fs : None
scipy : 1.4.1
sqlalchemy : 1.3.13
tables : None
tabulate : None
xarray : None
xlrd : 1.2.0
xlwt : 1.3.0
xlsxwriter : None
numba : None

[simonjayhawkins]

Thanks @jluttine for the report.

It appears that #21947 fixed this issue for the DataFrame case.

numpy returns the same array in this case.

>>> np.where([], np.array([], dtype=float), [])
array([], dtype=float64)
>>>

have opened a PR with similar behaviour for Series.

>>> pd.Series([], dtype=float).where([])
Series([], dtype: float64)