MAINT rename base_estimator in _BaseChain subclasses

[SuccessMoses]

Reference Issues/PRs

Fixes #29620 using the process followed in a similar pull request #23819

What does this implement/fix? Explain your changes.

This PR continues from the stalled PR #29682.

• Renames the base_estimator parameter to estimator in RegressorChain and ClassifierChain for consistency, see base_estimator vs estimator for meta-estimators #9104. The former parameter is deprecated and will be removed in 1.7
• A deprecation warning is added in the .fit method of the _BaseChain class so that for classes that inherit from _BaseChain, a warning message is displayed anytime .fit is called with base_estimator.
• Added test for deprecation warning in RegressorChain and ClassifierChain.
• Followed the process in a similar merged PR, MAINT rename base_estimator and base_estimator_ in ensemble classes #23819, to ensure no error.

[github-actions]

✔️ Linting Passed

All linting checks passed. Your pull request is in excellent shape! ☀️

_{Generated for commit: b382bdd. Link to the linter CI: here}

[SuccessMoses]

@adrinjalali , please help me fix the issue with changelog. This is my first time making this kind of pull request. I added my changes to doc/whats_new/v1.5.rst but still having issues

[adrinjalali]

We've changed our changelog system, here's the new info: https://github.com/scikit-learn/scikit-learn/blob/main/doc/whats_new/upcoming_changes/README.md

You also need to enable pre-commit hooks so that the linters run before you commit.

[SuccessMoses]

@glemaitre, I would appreciate it if you could review my pull request

[glemaitre]

Could you fix the CI.

[SuccessMoses]

CI is failing because of test/test_multioutput.py. I tried to run pytest locally before pushing, but pytest won't work.

How can i build scikit-learn?

.

[glemaitre]

You should follow the instructions in the contributing guideline:

https://scikit-learn.org/dev/developers/contributing.html#how-to-contribute

[SuccessMoses]

@glemaitre there was a bug in the new code. I fixed it. The code passes pytest locally but not in the Azure pipeline. I don't know how to fix this. I am using windows 10.

[adrinjalali]

You need to run the tests with SKLEARN_WARNINGS_AS_ERRORS=1 to get the same error as the CI.

You should read the log of the CI to see where the issue comes from. The stack trace of the CI shows you where the issue is.

Maybe @StefanieSenger could pair with you to help you move forward.

[StefanieSenger]

Hi @SuccessMoses,
thanks for working on this! :)

In order to make the CI tests pass, you need to fix the failing tests (use estimator instead of base_estimator). If you run pytest with the SKLEARN_WARNINGS_AS_ERRORS=1 flag, you should see which tests are failing locally.

Let me know if that doesn't work or you run into problems.

[adrinjalali]

Otherwise LGTM I think.

[thomasjpfan]

Minor nits to update the versions, otherwise LGTM

[SuccessMoses]

@thomasjpfan thanks for the review. I fixed the version number

[adrinjalali]

You've got merge conflicts here @SuccessMoses

[glemaitre]

Just pushing to fix the test. It was one remaining self.base_estimator due to the merge conflict. Let me review.

[glemaitre]

LGTM if the CI get through this time.