Virgil Chan

Thank you for the feedback! I just realized the term "logistic vector," and I think I can fix that. Regarding the optimal temperature being $\geq 1$, is there a reference...

I've made to my code in my fork of scikit-learn. Specifically, the "logistic vector" and "optimal temperature $\geq$ 1" parts are fixed. These changes can be found in the [`sklearn/calibration_temperature.py`](https://github.com/virchan/scikit-learn/blob/main/sklearn/calibration_temperature.py)...

Hello all, I have opened a PR #29517 implementing temperature scaling for multi-class classification within the `CalibratedClassifierCV` class. You can find the details and code in the PR. Comments and...

LGTM ping @adrinjalali

There's also an ongoing discussion in #31350 that touches on similar points.

I'll convert this back to draft and work on it.

Thanks @lucyleeow!

I renamed `_assemble_r2_explained_variance` to `_assemble_fraction_of_explained_deviance`. Doctest still failed when `average=True` is passed into `_weighted_percentile`: ```bash 1759 Examples 1760 -------- 1761 >>> from sklearn.metrics import d2_pinball_score 1762 >>> y_true = [1,...

I ran some benchmarks using `numpy.percentile` when `sample_weight` is `None` in this PR: ```bash # main branch Testing NumPy namespace... 100%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 200/200 [00:03

> Just reading the thread again, what happened to the failing test: > > ``` > FAILED sklearn/metrics/tests/test_common.py::test_multioutput_sample_weight_invariance[d2_pinball_score_01] - AssertionError: > Not equal to tolerance rtol=1e-07, atol=0 > Weighting d2_pinball_score_01...