Skip to contents

Generates candidates using Latin Hypercube Sampling, which provides good space-filling properties.

Usage

candidate_generator_lhs()

Value

A candidate generator function.

Details

LHS ensures that each variable's marginal distribution is uniformly covered, even with a relatively small number of samples. Good default choice for most problems.

Uses paradox::generate_design_lhs() internally.