mhp: Multivatriate hyperparameter (mhp) objects

In multivator: A Multivariate Emulator

Multivatriate hyperparameter (mhp) objects

Description

Create and manipulate multivatriate hyperparameter (mhp) objects

Usage

mhp(M, B, levels = NULL, names = NULL)
is.mhp(x)
M(x)
M(x) <- value
B(x)
B(x) <- value
levels(x)
summary(object,...)

Arguments

M	Variance matrix (must be positive definite)
B	Array of roughness parameters. Each slice (ie B[,,i]) must be positive-definite
levels	Character vector holding the levels. Default NULL means to use rownames(M) or dimnames(B[[3]])
names	Character vector holding the names of the dimensions. Default of NULL means to use dimnames(B[[1]])
x,object	Object of class mhp
value	Replacement object
...	Further arguments passed to the summary method

Details

An mhp object must have names and levels, so either provide them explicitly with the eponymous arguments, or give named arrays to M and B.

Value

Returns an object of class mhp

Author(s)

Robin K. S. Hankin

See Also

mdm

Examples

hp <- mhp(M=diag(2),B=array(c(diag(3),diag(3)),c(3,3,2)),
names=letters[1:3],levels=c("oak","ash"))
M(hp)
B(hp)[1,1,1] <- 30  # try a negative value and see what happens
names(hp)
names(hp) <-  c("Alice","Zachy","Annabel")
levels(hp) <- c("squid","snail")
summary(hp)

multivator documentation built on Aug. 22, 2023, 9:10 a.m.