# LHD v0.1.0

0

0th

Percentile

## Latin Hypercube Designs (LHDs) Algorithms

Contains functions for finding space-filling Latin Hypercube Designs (LHDs), e.g. maximin distance LHDs. Unlike other packages, our package is particularly useful in the area of Design and Analysis of Experiments (DAE). More specifically, it is very useful in design of computer experiments. One advantage of our package is its comprehensiveness. It contains a variety of heuristic algorithms (and their modifications) for searching maximin distance LHDs. In addition to that, it also contains other useful tools for developing and constructing maximin distance LHDs. In the future, algebraic construction methods will be added. Please refer to the function documentations for the detailed references of each function. Among all the references we used, one reference should be highlighted here, which is Ruichen Jin, Wei Chen, Agus Sudjianto (2005) <doi:10.1016/j.jspi.2004.02.014>. They provided a new form of phi_p criterion, which does not lose the space-filling property and simultaneously reduces the computational complexity when evaluating (or re-evaluating) an LHD. Their new phi_p criterion is a fundamental component of our many functions. Besides, the computation nature of the new phi_p criterion enables our functions to have less CPU time.

## Functions in LHD

 Name Description dij Calculate the Inter-site Distance SA2008 Simulated Annealing for LHD with Multi-objective Optimization Approach OA2LHD Transfer an Orthogonal Array (OA) into a LHD phi_p Calculate the phi_p Criterion SLHD Sliced Latin Hypercube Design (SLHD) exchange Exchange two random elements OASA Orthogonal-Array-Based Simulated Annealing GA Genetic Algorithm for LHD SA Simulated Annealing for LHD rLHD Generate a random Latin Hypercube Design (LHD) LaPSO Particle Swarm Optimization for LHD No Results!