covTS-class {kergp}

Class "covTS"

Description

S4 class representing a Tensor Sum (TS) covariance kernel.

Objects from the Class

Objects can be created by call of the form new("covTS", ...) or by using the covTS function.

Slots

d:

Object of class "integer", the spatial dimension or number of inputs of the covariance.

inputNames:

Object of class "character", vector of input names. Length d.

kernel:

Object of class "covMan" representing a 1d kernel.

kernParNames:

Object of class "character", name of the kernel (among the allowed ones).

kernParCodes:

Object of class "integer", an integer code stating the dependence of the parameter to the input.

par:

Object of class "numeric", numeric vector of parameter values.

parN:

Object of class "integer", total number of parameters.

parInput:

Object of class "integer", the number of the inputs for each parameter. Same length as par, values between 1 and d.

parLower:

,

parUpper:

Object of class "numeric" numeric, vector of (possibly infinite) lower/upper bounds on parameters.

parBlock:

Object of class "integer"

Methods

coef

signature(object = "covTS"): extracts the numeric values of the covariance parameters.

coef<-

signature(object = "covTS"): replaces the whole vector of coefficients, as required during ML estimation.

coefLower

signature(object = "covTS"): extracts the numeric values of the lower bounds.

coefLower<-

signature(object = "covTS"): replacement method for lower bounds on covTS coefficients.

coefUpper

signature(object = "covTS"): ...

coefUpper<-

signature(object = "covTS"): replacement method for upper bounds on covTS coefficients.

covMat

signature(object = "covTS"): builds the covariance matrix, or the cross covariance matrix between two sets of locations for a covTS object.

kernelName

signature(object = "covTS"): return the character value of the kernel name.

parMap

signature(object = "covTS"): an integer matrix used to map the covTS parameters on the inputs and kernel parameters during the computations.

scores

signature(object = "covTS"): computes the scores.

show

signature(object = "covTS"): prints in a custom format.

simulPar

signature(object = "covTS"): simulates random values for the covariance parameters.

Note

The names of the methods strive to respect a camelCase naming convention.

While the coef<- replacement method is typically intended for internal use during likelihood maximization, the coefLower<- and coefUpper<- replacement methods can be used when some rough information exists on the possible values of the parameters.

Author(s)

Y. Deville, O. Roustant, D. Ginsbourger.

See Also

The covTS function providing a creator.

Examples

showClass("covTS")