SPOT (version 2.11.10)

buildPCA: buildPCA

Description

buildPCA builds principal components of given dataset. It is used inside plotPCA function to build necessary object to perform principal components analysis.

Usage

buildPCA(x, control = list())

Arguments

x

dataset of parameters to be transformed

control

control list

Value

returns a list with the following elements: sdev the standard deviations of the principal components (i.e., the square roots of the eigenvalues of the covariance/correlation matrix, though the calculation is actually done with the singular values of the data matrix). rotation the matrix of variable loadings (i.e., a matrix whose columns contain the eigenvectors). x transformed matrix. center,scale the centering and scaling used, or FALSE.

Examples

# NOT RUN {
#define objective function

objFun <- function(x) 2*(x[1] - 1)^2 + 5*(x[2] - 3)^2 + (10*x[3] - x[4]/3)   

spotConfig <- 
list(types = c('numeric', 'numeric', 'numeric', 'numeric'), 
funEvals = 15, #budget
noise = TRUE,
seedFun = 1,
replicated = 2,
seedSPOT = 1,
design = designLHD,
model = buildRandomForest, #surrogate model
optimizer = optimLHD, #LHD to optimize model
optimizerControl = list(funEvals=100)) #100 model evals in each step

lower <- c(-20, -20, -20, -20) 
upper <- c(20, 20, 20, 20)

res <- spot(x=NULL, 
   fun=objFun, 
   lower=lower, 
   upper=upper, 
   control=spotConfig) 

resPCA <- buildPCA(res$x) 

# }