SKL 1.5.0 changes

[Karla Saur (ksaur)]

Based on these 2 failures:

• Fixing the decomposition bug based on FIX PLSRegression.coef_ takes X and Y variances into account when scale=True scikit-learn/scikit-learn#28612
• TODO: Bagging issue, see [FEA] Add metadata routing to BaggingClassifier and BaggingRegressor scikit-learn/scikit-learn#28432

Closes #773

[Karla Saur (ksaur)]

Ok the decomposition bug is definitely fixed!

I am a bit confused about the bagging changes (scikit-learn/scikit-learn#28432) I don't see what would have broken our code. Can you please take a look Matteo Interlandi (@interesaaat); I think you know this op best.

=========================== short test summary info ============================
FAILED tests/test_sklearn_bagging.py::TestSklearnBagging::test_bagging_logistic_regression - AssertionError: 
Not equal to tolerance rtol=1e-06, atol=1e-06

Mismatched elements: 200 / 200 (100%)
Max absolute difference: 0.99875297
Max relative difference: 0.99875297
 x: array([[9.809476e-01, 1.905245e-02],
       [9.987530e-01, 1.247026e-03],
       [8.200397e-01, 1.799603e-01],...
 y: array([[0., 1.],
       [0., 1.],
       [0., 1.],...
====== 1 failed, 389 passed, 275 skipped, 59 warnings in 89.85s (0:01:29) ======

It seems that there is some type/rounding issue?

[Karla Saur (ksaur)]

Let's see where we are on this one, i think it's important to unpin this sooner than later. If necessary, we should mark this test as broken and add some warnings then move forward if possible