Johannes P. Dürholt
Johannes P. Dürholt
Hi, you are almost correct. In the first ask only the inputs are validated, in the second one in the `PredictiveStrategy` also the outputs. But we have to tidy this...
Details will follow soon ;)
We should orient on this one https://github.com/facebook/Ax/blob/5c0cbb698d303210ce1522c2cfc90f13f8859ca0/ax/plot/diagnostic.py#L520
I would also advocate for a method in `/plot` which takes `CvResults` as input.
Great! Regarding of the constrained objectives: Simple case: You have an optimization Problem in which you want to maximize one output but an output constrained has to be fulfilled like...
> Are there practical applications where 2 is not sufficient? Yes. We often run opts with the first use case.
Just as info, we will also bring back the random strategy and the sampling part from our side. We will orient very much on how you did it in `opti`.
> Just as info, we will also bring back the random strategy and the sampling part from our side. We will orient very much on how you did it in...
> Okay. Why do we need a separate base class and a flag? Isn't a class `ConstrainedObjective` for use case 1 sufficient? In use case 2 I do not see...
I really like the idea with the weights!