update_gaussian_process: Gaussian Process Update

In wzhorton/hortonlib: Collection of Useful Functions

Description

Updates and returns the mean vector for a gaussian process given evaluation points, observed data, a mean function, and a covariance function.

Argument model:
y(time) ~ GP(mnfun(x),covfun(x1-x2))

Usage

update_gaussian_process(x, y, time, mnfun, covfun, random = FALSE)

Arguments

x, y	coordinate vectors for the observed data points.
time	locations to evaluate the curve at.
mnfun	mean function that takes a single argument. If using a constant like 0 use function(z){0}.
covfun	covariance function that takes a single argument that represents an absolute distance. Distances are internally calculated using fields::rdist.
random	logical; FALSE will return the updated mean function. TRUE will generate a random curve.

Value

a vector corresponding to the time variable that represents the updated mean vector.

wzhorton/hortonlib documentation built on May 9, 2019, 8:19 a.m.