What is the optimal lattice-based sampling method, and how many samples are required to sample the entire parameter space?
The current grid layout being used is not optimal. The correct way to determine an optimal lattice is the Petersen–Middleton theorem, which should allow us to potentially reduce the number of samples required compared to the hypercube sampling.