make_cov_mat_ardC: Make a covariance matrix with covariance function

In nategarton13/sparseRGPs: Fit Sparse Gaussian Processes

Description

Make a covariance matrix with covariance function

Usage

make_cov_mat_ardC(x, x_pred, cov_par, cov_fun, delta, lnames)

Arguments

x	Matrix of observed data values
x_pred	Potentially an empty matrix. If not empty, this returns the covariance between x and x_pred.
cov_par	Named list of covariance parameters. Need "sigma", "tau", and "l1", ..., "ld" where d is the input dimension
cov_fun	A character string specifying the covariance function. Currently only works with "ard".
delta	Scalar value that functions as a fixed nugget used to stabilize matrix inverses and decompositions.
lnames	Vector of the names of the length scale parameters, i.e. "l1", ..., "ld"

Value

Covariance matrix at the observed data locations if x_pred = matrix(), or the covariances between x and x_pred if x_pred is not an empty matrix.

nategarton13/sparseRGPs documentation built on May 27, 2020, 9:46 a.m.