Marius Lindauer

@mfeurer could you please remind of the reason why it is not implemented in SMAC?

@KOLANICH we will consider this request for our next SMAC release.

We implicitly replaced that by `n_steps_plateau_walk` after introducing plateau walks. I don't know whether we want to consider `max_steps` again or simply dropping it entirely. (I recently read a paper...

So, how to solve this problem? I don't know whether we already agreed on something. I would propose to add a check in `Intensifier` that if `intensifiy` was called 10...

Dear Vivek, Unfortunately, SMAC does not handle such search spaces well if the constraints are unknown. The underlying model will try to learn the valid/invalid parts of the search space;...

Unfortunately, I think the answer is no. But could you please give us an example how you would like pass this information?

So, you can enumerate all possible configurations? That's quite unusual in our experiments, but we have also thought about whether it would make sense to have such a feature in...

Our forbidden constraints allow to define simple constraints in the configuration space. However, we strongly recommend to avoid such forbidden constraints because SMAC does not implement a constraint solver and...

I'm still hesitant to agree to this proposal. There are use cases in which average is not the correct aggregation. Nevertheless, the code assumes "average" for aggregation right now and...

Hi, Thank you for reporting this issue. However, I cannot reproduce it on my end. Running one of the examples, e.g., the svm example (https://github.com/automl/SMAC3/blob/master/examples/SMAC4HPO_svm.py) works fine on my machine....