How front is estimated and how to avoid duplication and invalid points

[jdzhu19]

First of all, I would like to ask how the target function value corresponding to pop is calculated in _NSGA2.py, and whether pop is the target function value or the fitness? (It seems to look like it, but I don't see from the code how it was calculated), I am a bit confused about this part.

Also, I would like to ask what you would suggest I do if I want to deal with duplicate(what I mean is that I manually approximate continuous points into discrete points, how to avoid duplication of discrete points after the change) and invalid points. My current method is to add a loop to _bayes.py and try to collect a random point immediately when a duplicate or invalid point is observed, but this does not work well, so I would like to ask if there is a way to avoid invalid points or points that have been sampled before when sampling the parameter points? Is it possible to do it with the constraints? Or do you have any better suggestions?

Thank you very much for your time and I hope to hear from you soon!