scaleDesign: Scale a Design

In DiceDesign: Designs of Computer Experiments

Scale a Design

Description

This function scales the values of the design points to values comprised in [0,1]. The scaling can be made by the Rosenblatt transformation (uniformization by applying the empirical cumulative distribution function) or by translating the design from maximum and minimum values (given for each variable).

Usage

scaleDesign(design, min=NULL, max=NULL, uniformize=FALSE)

Arguments

design	a matrix (or a data.frame) corresponding to the design of experiments to scale
min	the vector of minimal bounds of each design variable. If not given, the minimal value of each variable is taken
max	the vector of maximal bounds of each design variable. If not given, the maximal value of each variable is taken
uniformize	boolean: TRUE to use the Rosenblatt transformation (the min and max vectors are useless in this case). If FALSE (default value), the translation from max and min values is applied

Value

A list containing:

design	the scaled design
min	the vector of minimal bounds that has been used
max	the vector of maximal bounds that has been used
uniformize	the value of this boolean argument
InitialDesign	the starting design

Author(s)

B. Iooss

Examples

  d <- 2
  n <- 100
  x <- matrix(rnorm(d*n), ncol=d)
  xscale1 <- scaleDesign(x, uniformize=FALSE)
  xscale2 <- scaleDesign(x, uniformize=TRUE)
  par(mfrow=c(1,2))
  plot(xscale1$design) ; plot(xscale2$design)

DiceDesign documentation built on May 29, 2024, 10:35 a.m.