space_eval: Evaluate Design Criterion on LHS

In GADGET: Gaussian Process Approximations for Designing Experiments

Description

This function simplifies the evaluation of design criteria on space-filling designs (matrices where each row is a possible design). The design criterion is parallelized as well using the parallel package. A progress bar gives a visual indicator when the design criterion is expensive.

Usage

space_eval(lhs, criterion, cluster = NULL)

Arguments

lhs	A matrix with n rows and d columns.
criterion	A function with vector input of length d.
cluster	A parallel cluster object.

Details

Only univariate design criteria can be optimized GADGET currently.

Value

A row vector of length n.

Examples

my_lhs   <- space_fill(c(0,0),c(10,10),20)
dc       <- function(x){sum(x^2)}
response <- space_eval(my_lhs,dc)
print(response) 

GADGET documentation built on Jan. 25, 2020, 1:06 a.m.