@massich thanks so much for resurrecting this PR-- I'm still using the feature in multiple projects and so would love to see it incorporated. I just pushed some flake8 fixes to my original PR #5181.

What's your plan for tests? I have a set of real-world benchmarks in my brf-example repo. Let me know how I can help.

What's your plan for tests?

@potash We want to extend the benchmark on several imbalanced dataset.
In this regard, we will use the datasets which we recently added in the imbalanced-learning as well as the additional metrics available.

I'm still using the feature in multiple projects and so would love to see it incorporated. I just pushed some flake8 fixes to my original PR #5181.

Could you reset the history to avoid merging master into your branch. Rebasing should be the way.
In fact, an easy thing could be to checkout the current branch and cherry-pick your PEP8 commits and make a PR on @massich branch.

Let me know how I can help.

I am sure that we will need support to make sense of the benchmarking results.

@glemaitre I cherry-picked the flake8 fixes and put them in a PR to @masich.

This work is followed here #8732 'cos I messed up somehow and I don't find howto retake the PR