Wouter Kool

I just wanted to raise the discussion, I'm not saying that we should care that much. I think putting it into the diversity measure is a good idea, but as...

I am in favor of using something standard, and `numpy` would be a good option but I'm strongly in favor of avoiding global random state (this would become a mess...

Not completely sure if there are edge cases, but I think swapping of adjacent segments is equivalent to a relocate, so is this really needed? Maybe yes, since we use...

Some thoughts: * If penalties are float, but objectives are integral, we may get inconsistencies, i.e. if new_cost = cost + delta, it does not mean round(new_cost) = round(cost) +...

We can also investigate whether it is needed to update the diversity for all items already in the population when inserting a new individual. We could also choose to compute...