Roman Lutz

I believe the reason for using literalinclude is the plots. If you run normal code blocks and generate plots they won't show. There may be other ways of doing that,...

Do you have the example code which produces this output? Looks like the last 4 rows are added in the second one. Reading the code, it not obvious to me...

> Please consider merging and releasing this fix. Thanks! @fairlearn/fairlearn-maintainers

You can access them via `predictors_`. Note that they may not satisfy the fairness constraints individually. Were you looking for guidance on which one to select?

Using `expgrad.predictors_[i]` for whichever index you like. It's just an array. I recommend examining their fairness-related stats using MetricFrame, and again: they may not perform as well as ExponentiatedGradient because...

@sanvi0sh do you know what to do for this one? I'm afraid it's not super well-defined and leaves a lot of room for interpretation. There are other issues that don't...

@sanvi0sh that is not at all an unreasonable assumption. Other issues with "help wanted" are usually more clearly scoped. Feel free to look at them if you don't want to...

Sure, but I still suggest reading through the notebook first and suggesting additions to the user guide before actually creating a PR. Otherwise you may end up doing work and...

We should get @MiroDudik's take on this since he's one of the authors of the paper.

Yes, you're missing the `random_state` parameter on `predict`:  https://fairlearn.org/v0.8/api_reference/fairlearn.reductions.html#fairlearn.reductions.ExponentiatedGradient.predict Please reopen if this doesn't fully address your question.