varVec-methods: Covariance Matrix for a Covariance Kernel Object

In kergp: Gaussian Process Laboratory

Description

Covariance matrix for a covariance kernel object.

Usage

## S4 method for signature 'covMan'
varVec(object, X, compGrad = FALSE, 
       checkNames = NULL, index = -1L, ...)

## S4 method for signature 'covTS'
varVec(object, X, compGrad = FALSE, 
       checkNames = TRUE, index = -1L, ...)

Arguments

object	An object with S4 class corresponding to a covariance kernel.
X	The usual matrix of spatial design points, with n rows and d cols where n is the number of spatial points and d is the 'spatial' dimension.
compGrad	Logical. If TRUE a derivative with respect to the vector of parameters will be computed and returned as an attribute of the result. For the covMan class, this is possible only when the gradient of the kernel is computed and returned as a "gradient" attribute of the result.
checkNames	Logical. If TRUE (default), check the compatibility of X with object, see checkX.
index	Integer giving the index of the derivation parameter in the official order.
...	Not used yet.

Details

The variance vector is computed in a C program using the .Call interface. The R kernel function is evaluated within the C code using eval.

Value

A vector of length nrow(X) with general element V[i] = K(X[i, ], X[i, ], θ) where K(x1, x2, θ) is the covariance kernel function.

Note

The value of the parameter θ can be extracted from the object with the coef method.

See Also

coef method

Examples

myCov <- covTS(inputs = c("Temp", "Humid", "Press"),
               kernel = "k1PowExp",
               dep = c(range = "cst", shape = "cst"),
               value = c(shape = 1.8, range = 1.1))
n <- 100; X <- matrix(runif(n*3), nrow = n, ncol = 3)
try(V1 <- varVec(myCov, X = X)) ## bad colnames
colnames(X) <- inputNames(myCov)
V2 <- varVec(myCov, X = X)

Xnew <- matrix(runif(n * 3), nrow = n, ncol = 3)
colnames(Xnew) <- inputNames(myCov)
V2 <- varVec(myCov, X = X)

kergp documentation built on March 18, 2021, 5:06 p.m.