mixexp (version 1.2.7)

Xvert: This function creates an extreme vertices design in a constrained mixture space.

Description

This function calls the function crvtave to create an extreme vertices design in a constrained mixture space. If there are only three factors the function DesignPoints is called to plot the results.

Usage

Xvert(nfac=3,uc=c(0,0),lc=c(0,0),nlc=0,lb=c(0,0),ub=c(0,0),coef,ndm=0,plot=TRUE,
                 cornerlabs = c("x1","x2","x3"), axislabs = c("x1","x2","x3"),
                 pseudo=TRUE)

Arguments

nfac

an integer representing the number of mixture variables in the design. Maximum nfac=12

uc

a vector of length nfac containing upper constraints on each mixture component

lc

a vector of length nfac containing lower constraints on each mixture component

nlc

the number of linear constraints, the default is zero

lb

a vector of length nlc containing the lower bounds for the linear constraints

ub

a vector of length nlc containing the upper bounds for the linear constraints

coef

an nlc by nfac matrix containing the coefficients of the components of the linear constraints

ndm

an integer representing the highest order of centroids requested. An overall centroid is always included, 0 indicates no other centroids will be created, 1 indicates edge centroids are requested, etc.

plot

a logical variable indicating whether a plot of the design is desired when there are only 3 components. Default is TRUE

cornerlabs

This is a vector of text labels for the x1, x2 and x3 vertices. Use when there are only 3 components for plotting.

axislabs

This is a vector of text labels for the x1, x2 and x3 axis. Use when there are only 3 components for plotting.

pseudo

logical variable, when TRUE plot in pseudo component space when there are lower constraints.

References

1. Piepel, G. F. "Programs for Generating Extreme Vertices and Centroids of Linearly Consrtained Experimental Regions" Journal of Quality Technology, Vol 20, No. 2, pp. 125-139, 1988.

2. "John Lawson, Cameron Willden (2016).", "Mixture Experiments in R Using mixexp.", "Journal of Statistical Software, Code Snippets, 72(2), 1-20.", "doi:10.18637/jss.v072.c02"

Examples

# NOT RUN {
# Polvoron Example from Lawson
des<-Xvert(3,uc=c(.8,.95,.50),lc=c(0,.10,.05),ndm=1,plot=FALSE)


#Snee Marquardt(1976) example
Xvert(8,uc=c(.45,.50,.10,.4,.6,.2,.05,.05),lc=c(.1,.05,0,0,.1,.05,0,0),ndm=0)
    
# Example page 465
exvert<-Xvert(4,uc=c(.188,.128,.438,.438),lc=c(.124,.064,.374,.374),ndm=2)

# Example from Piepel 1988
coef<-matrix(c(.85,.9,1,.7,0,1),ncol=3,byrow=TRUE)
des<-Xvert(3,lc=c(.1,.1,0),uc=c(.5,.7,.7),nlc=2,lb=c(.9,.4),ub=c(.95,0),coef,ndm=1,plot=FALSE)

# }