Quentin Soubeyran

For those interested, I had the same problem where `__attrs_post_init__` needed a context. So I wrote [some helpers and a `@post_init` decorator (source code)](https://github.com/QuentinSoubeyran/pheres/blob/b334d8498de9c5b7624295c30ef3a479e1cc4c63/src/pheres/utils.py#L184) that generates the new `__init__()` by...

Since there are use case for both the old and the new behaviours, I would argue for a flag in the optimizer's `__init__` that defaults to the old behaviour. This...

> Also, in #302 a user said "Usually evaluating the same point multiple times isn't something you want to do, but if you have a (very) noisy objective it makes...

Should scikit-optimize leverage this PR to move to the black code-style, like [scikit-learn did](https://github.com/scikit-learn/scikit-learn/pull/18948) ?

Do you happen to have a more complete stacktrace showing were the error was raised ? If you haven't already done so, checking the total memory used with `n_jobs=1` and...

@kernc with the release of 0.9.0, feel free to close this. I can also update it to serve as information for people coming on the repo. Though an official, pinned...

@RNarayan73 you should use the above `unofficial` branch instead of the old mock branch if you want to test un-merged features.

Try to push a commit again to trigger the CI, perhaps ?

Well, I'm at a loss here... Have you run the tests locally using pytest ? Maybe the CI is straigh crashing on this PR, hence no info ?

#1009 is indeed the way to go. I believe you are misunderstanding what Real number are. When you specify `Real(low=0.0010, high=0.0050)`, this means all number in the continuous interval `[0.001;...