kernCompute: Kernel Matrix for '"lineqGP"' Models.

In lineqGPR: Gaussian Process Regression Models with Linear Inequality Constraints

Description

Compute the kernel matrix for "lineqGP" models. attr: "gradient".

Usage

kernCompute(x1, x2 = NULL, type, par, d = 1L)

Arguments

x1	a vector with the first input locations.
x2	a vector with the second input locations.
type	a character string corresponding to the type of the kernel. Options: "gaussian", "matern32", "matern52", "exponential".
par	the values of the kernel parameters (variance, lengthscale).
d	a number corresponding to the dimension of the input space.

Value

Kernel matrix K(x1,x2) (or K(x1,x1) if x2 is not defined).

Author(s)

A. F. Lopez-Lopera.

Examples

x <- seq(0, 1, 0.01)
K <- kernCompute(x, type = "gaussian", par =  c(1, 0.1))
image(K, main = "covariance matrix")

lineqGPR documentation built on Jan. 11, 2020, 9:23 a.m.