Various pandas deprecation warnings from test suite

[kandersolar]

We've started getting several types of deprecation warnings from the tests:

pvlib/tests/test_modelchain.py::test_aoi_model_interp
  /home/runner/work/pvlib-python/pvlib-python/pvlib/tests/test_modelchain.py:1476: FutureWarning: Series.__getitem__ treating keys as positions is deprecated. In a future version, integer keys will always be treated as labels (consistent with DataFrame behavior). To access a value by position, use `ser.iloc[pos]`
    assert mc.results.ac[1] < 1

pvlib/tests/test_clearsky.py::test_detect_clearsky_window
  /home/runner/work/pvlib-python/pvlib-python/pvlib/tests/test_clearsky.py:634: FutureWarning: Setting an item of incompatible dtype is deprecated and will raise in a future error of pandas. Value 'True' has dtype incompatible with int64, please explicitly cast to a compatible dtype first.
    expected.iloc[-3:] = True

pvlib/tests/test_singlediode.py: 14 warnings
  /home/runner/work/pvlib-python/pvlib-python/pvlib/tests/test_singlediode.py:118: FutureWarning: DataFrame.applymap has been deprecated. Use DataFrame.map instead.
    joined[is_number] = joined[is_number].applymap(np.float64)

  /home/runner/work/pvlib-python/pvlib-python/pvlib/scaling.py:279: FutureWarning: DataFrame.fillna with 'method' is deprecated and will raise in a future version. Use obj.ffill() or obj.bfill() instead.
    df = df.fillna(method='bfill').fillna(method='ffill')

Here's an example run: https://github.com/pvlib/pvlib-python/actions/runs/6550206298/job/17788805508?pr=1827#step:9:65

Addressing these would be good first PRs for anyone wanting to get started with contributing to pvlib's codebase! If you plan to submit a PR for one of these warnings, please comment here so that someone else does not duplicate the work.

[matsuobasho]

Is the intention to resolve these warnings by updating the deprecated methods with current ones, or rather to just hide the deprecation warnings?

[cwhanse]

@matsuobasho preferably update the methods to resolve the warnings. If updating breaks other code, perhaps hide them.

[matsuobasho]

Ok, I'm tackling this issue. Since is this my first issue, would like to ask about a nuance that I've encountered.
In the test_bird function of test_clearsky.py, we get a warning on lines 857-859:

irrads3 = clearsky.bird(
        zenith[12], airmass[12], aod_380nm, aod_500nm, h2o_cm, dni_extra=etr[12]
    )

Here's the warning:

FutureWarning: Series.__getitem__ treating keys as positions is deprecated. In a future version, integer keys will always be treated as labels (consistent with DataFrame behavior). To access a value by position, use `ser.iloc[pos]`

The interesting part is that the warning only pertains to etr[12]. The reason why is that out of the 3 objects that are being referenced by an index, only etr is a Pandas series. zenith is a Pandas index type, while airmass is a numpy array. Neither of those have an iloc attribute.

So changing to etr.iloc[12] gets rid of the warning. However, there is now an inconsistency in how we would be referencing these objects. I think it's fine but looking for confirmation especially as I'll potentially run into the same scenario in other warnings that are part of this issue. Thanks.

[kandersolar]

Wow, three different vector types going into a single function :P Don't worry about the inconsistency with not using iloc for all inputs in that case; changing to just etr.iloc[12] is fine. Maybe that test could be cleaned up a bit more generally, but that would be a separate PR. Thanks @matsuobasho!

[matsuobasho]

I made the updates to get rid of the errors but I get a bunch of failures when I run pytest and I can't be sure that the warnings in question don't also appear in the functions that fail. For example:

Please recommend a course of action, thanks.

[kandersolar]

Ah yeah, the SRML website has been down for a while and causing the associated tests to fail. If you check the recent CI test results for main you'll see the same failures there.

I suggest ignoring those failing tests for now. If any deprecation warnings come out of those tests whenever they eventually start passing again, we can just address them at that time.

[matsuobasho]

1. To double-check, are the warnings below in-scope for this issue as well?
   

2. The pull request submission form has the following:

• Maintainer: Appropriate GitHub Labels (including remote-data) and Milestone are assigned to the Pull Request and linked Issue.
  Do I need to add any additional labels to this PR / issue?

[cwhanse]

@matsuobasho we add the labels to the PR.