ENH: updating subset types in drop_duplicates/duplicated #986
Conversation
|
@Dr-Irv This is the follow-up to pandas-dev/pandas#59392. |
|
I have added a test for the supported arguments. |
tests/test_frame.py
Outdated
| @pytest.mark.parametrize( | ||
| "drop_arg", | ||
| [ | ||
| {"AAA"}, # set | ||
| ["AAA"], # list | ||
| ("AAA",), # tuple | ||
| {"AAA": None}, # dict | ||
| "AAA", # str | ||
| ] | ||
| ) | ||
| def test_types_drop_duplicates(drop_arg) -> None: |
There was a problem hiding this comment.
For the purposes of testing type checks, you have to explicitly put in each of these arguments and can't parameterize. So you'd do:
check(assert_type(df.drop_duplicates({"AAA"}), pd.DataFrame), pd.DataFrame)
check(assert_type(df.drop_duplicates(["AAA"]), pd.DataFrame), pd.DataFrame)
# and similar for the other 3Then you don't need to do assert_frame_equal()
|
Updated based on your feedback. Thanks! |
tests/test_frame.py
Outdated
| check(assert_type(df.drop_duplicates({"AAA"}), pd.DataFrame), pd.DataFrame) | ||
| check(assert_type(df.drop_duplicates(["AAA"]), pd.DataFrame), pd.DataFrame) | ||
| check(assert_type(df.drop_duplicates(("AAA",)), pd.DataFrame), pd.DataFrame) | ||
| check(assert_type(df.drop_duplicates({"AAA": None}), pd.DataFrame), pd.DataFrame) | ||
| check(assert_type(df.drop_duplicates("AAA"), pd.DataFrame), pd.DataFrame) |
There was a problem hiding this comment.
With pandas 2.2, the set and dict arguments don't work. But they will work in a future release of pandas.
So you need to surround the set and dict tests with
if PD_LTE_22:
# Put the 2 tests hereThere was a problem hiding this comment.
Thanks for the tip!
Co-authored-by: Irv Lustig <[email protected]>
There was a problem hiding this comment.
You need to run black on your changed files, or follow the instructions at https://github.com/pandas-dev/pandas-stubs/blob/main/docs/setup.md to set up your environment and install the pre-commit checks.
assert_type()to assert the type of any return value