Suavesito

No, it does not work. But... if you change your algorithm to this t works ```julia k(x, y) = (((x+y) - 2)^6 + 0.2) * log(1 + (x+y)^2) global_min, minimisers...

Ok, now I see what you're trying to do. I don't know is this kind of algorithms are really suited for finding `minmax` of functions. Maybe this is a theoretical...

Still to add tests and discuss the proper algorithm to unify boxes.

Don't worry, happy to help. :D

Sure, I'll add test. I just want to know what do you think about the "problem" of using union-find that I described in the comment in the original issue.

Hi! I decided to take a chance into this and found there is something weird with the implementation described above for 2 dimensions and above. The weird case is easily...

Is because the `B₁∩B₃` and `B₂∩B₃` is empty, even if `(B₁∪B₂)∩B₃` is not, and the algorithm is just testing pairwise the intersections.

Maybe we can add three options for unify, `:hull`, `:connected` and `:none`?

I can add the algorithm for the hull too. I'll add it to the PR and make the changes on the documentation and add tests.

I can check the differences between my inplementation and theirs, also see performan differences and reply back later.