BUG: Inconsistent behavior in Index.difference · Issue #35217 · pandas-dev/pandas (original) (raw)

• I have checked that this issue has not already been reported.
• I have confirmed this bug exists on the latest version of pandas.
• (optional) I have confirmed this bug exists on the master branch of pandas.

Index.difference seems to be treating float vs object / int vs object differently.

Code Sample, a copy-pastable example

import pandas as pd float_index = pd.Index([1.0, 2, 3]) string_index = pd.Index(['1','2','3']) integer_index = pd.Index([1, 2, 3]) print(float_index.difference(string_index)) print(string_index.difference(float_index))

print(integer_index.difference(string_index)) print(string_index.difference(integer_index))

Float64Index([], dtype='float64') Index([], dtype='object') Int64Index([1, 2, 3], dtype='int64') Index(['1', '2', '3'], dtype='object')

Problem description

Index.difference should probably not type-cast and compare the elements to honor consistent behavior with other types where they are not being type-casted.

Expected Output

Float64Index([1.0, 2.0, 3.0], dtype='float64') Index(['1', '2', '3'], dtype='object') Int64Index([1, 2, 3], dtype='int64') Index(['1', '2', '3'], dtype='object')

or

All empty Index objects like in case of int vs object.

Output of pd.show_versions()

INSTALLED VERSIONS

commit : None

pandas : 1.0.5
numpy : 1.19.0
pytz : 2020.1
dateutil : 2.8.1
pip : 20.1.1
setuptools : 49.1.0
Cython : None
pytest : None
hypothesis : None
sphinx : None
blosc : None
feather : None
xlsxwriter : None
lxml.etree : None
html5lib : None
pymysql : None
psycopg2 : None
jinja2 : None
IPython : None
pandas_datareader: None
bs4 : None
bottleneck : None
fastparquet : None
gcsfs : None
lxml.etree : None
matplotlib : None
numexpr : None
odfpy : None
openpyxl : None
pandas_gbq : None
pyarrow : None
pytables : None
pytest : None
pyxlsb : None
s3fs : None
scipy : None
sqlalchemy : None
tables : None
tabulate : None
xarray : None
xlrd : None
xlwt : None
xlsxwriter : None
numba : None
None