TST: indexing/test_floats.py::TestFloatIndexers::test_scalar_non_numeric

[simonjayhawkins]

precursor to #25750

This test has some commented out code with a pytest.raises not used as a context manager.

The goal here is primarily to remove the commented out code.

some refactoring but not enough to allow full parametrization.

[codecov]

Codecov Report

Merging #25748 into master will decrease coverage by <.01%.
The diff coverage is n/a.

@@            Coverage Diff             @@
##           master   #25748      +/-   ##
==========================================
- Coverage   91.25%   91.25%   -0.01%     
==========================================
  Files         172      172              
  Lines       52977    52977              
==========================================
- Hits        48343    48342       -1     
- Misses       4634     4635       +1

Flag	Coverage Δ
#multiple	89.82% <ø> (ø)	⬆️
#single	41.74% <ø> (ø)	⬆️

Impacted Files	Coverage Δ
pandas/util/testing.py	88.98% <0%> (-0.1%)	⬇️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 65c0441...34e98e6. Read the comment docs.

[codecov]

Codecov Report

Merging #25748 into master will decrease coverage by <.01%.
The diff coverage is n/a.

@@            Coverage Diff             @@
##           master   #25748      +/-   ##
==========================================
- Coverage   91.25%   91.25%   -0.01%     
==========================================
  Files         172      172              
  Lines       52977    52977              
==========================================
- Hits        48343    48342       -1     
- Misses       4634     4635       +1

Flag	Coverage Δ
#multiple	89.82% <ø> (ø)	⬆️
#single	41.74% <ø> (ø)	⬆️

Impacted Files	Coverage Δ
pandas/util/testing.py	88.98% <0%> (-0.1%)	⬇️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 65c0441...34e98e6. Read the comment docs.

[simonjayhawkins]

i guess that maybe this should have been assert 3.0 not in i

[jreback]

what is i?

[jreback]

quick looks seems this is not much more clear than the original. maybe split get/set will be helpful

[jreback]

i think we already have a fixture for these

[jreback]

what is i?

[simonjayhawkins]

quick looks seems this is not much more clear than the original.

i agree. the main change is to include the cases that were skipped (not in the pytest.skip sense but in the pass or commented out sense.

maybe split get/set will be helpful

i agree that there is basically only two tests here: idxr(obj)[3.0] and idxr(obj)[3.0] = 0
however, if we are checking for consistency of error messages then having them in the same parametrised test may be an advantage. (would also eliminate need for duplicating parametrisation or creating fixtures)
would you be adverse to something like..

@pytest.mark.parametrize('pass', ['getting', 'setting'])
def ...
    if pass == 'getting':
        # getting
         ...
         idxr(obj)[3.0]
    else:
        # setting
        ...
        idxr(obj)[3.0] = 0

also note that it is relatively straightforward to remove iloc completely from this test (and most of the other tests in this module) as it is tested at the start of the module, which make sense for float indexing with iloc. I felt that was out of scope for this PR. A precursor PR to refactor iloc would certainly make the test look less cluttered.

[jreback]

also note that it is relatively straightforward to remove iloc completely from this test (and most of the other tests in this module) as it is tested at the start of the module, which make sense for float indexing with iloc. I felt that was out of scope for this PR. A precursor PR to refactor iloc would certainly make the test look less cluttered.

if you don't mind that would be great.

[simonjayhawkins]

superseded by #25866