I'm not very well-versed in the technique you used to optimize Latin Hypercubes, but I'd be interested in using something similar to optimize randomized QMC sequences, which are similar, but stratify as many marginals as possible (e.g. all 2-d projections when n > d^2) rather than just one-dimensional marginals. I suspect it should be possible, since randomized Faure sequences have similar properties to LHS.
Randomized QMC sequences can be found in QuasiMonteCarlo.jl, where I've gone through the work of adding them.
I'm not very well-versed in the technique you used to optimize Latin Hypercubes, but I'd be interested in using something similar to optimize randomized QMC sequences, which are similar, but stratify as many marginals as possible (e.g. all 2-d projections when
n > d^2) rather than just one-dimensional marginals. I suspect it should be possible, since randomized Faure sequences have similar properties to LHS.Randomized QMC sequences can be found in QuasiMonteCarlo.jl, where I've gone through the work of adding them.