Design of Experiments, accelerate Latin Hypercube with Genetic Option

I would like to do a Latin Hypercube Sampling with genetic space filling and with 4 parameters over 10.000 samples. Unfortunately this takes very long (sampling was not finished after 14 h).

Are there any possibilities to accelerate this?

I tried to set the iterations from default = 100 to 4. This had no noticeable effect for the 10,000 samples.

The iterations are the number of outer loop iterations of the genetic optimization process. Changing this number will drastically effect the time and the number of mutations the algorithm tries. The inner loop iterations is also effected by the number of samples. The more samples, the more mutations are tried.

I would argue that with 10,000 samples and 4 parameter, using genetic optimization is not going to add much. If you had 100 samples and 4 parameters, then then using genetic optimization would help spread the samples out.

I would like to speed this algorithm up. Just need time to do it.

10,000 samples to fill a 4-D space sounds like overkill

Thank you very much for the answer.
Please let me know if the algorithm has been optimized.

Then I will test the “normal” LHS for the case with 10.000 samples and additionally compare the case 1000 samples with LHS “normal” as well as “genetic”.

Thanks a lot!