geostatsp (version 0.4.4)

loglik.GRF: Log-Likelihood for a Gaussian Random Field

Description

This function computes the value of the log-likelihood for a Gaussian random field.

Usage

loglik.GRF(geodata, ...)## S3 method for class 'SpatialPointsDataFrame':
loglik.GRF(geodata, formula, sigmasq, 
  phi, psiA = 1, psiR = 1, kappa = 1, lambda = 1, nugget = 0, ...)

Arguments

geodata

A SpatialPointsDataFrame containing the data to be fit, or an object suitable for passing to loglik.GRF in geoR.

formula

A formula object, specifies the response variable and covariates

sigmasq

Variance of the Gaussian random field

phi

Range of the GRF

psiA

Anisotropy rotation angle

psiR

Ratio of major to minor axis lengths for anisotropy, defaults to 1 for an isotropic process.

kappa

Roughness parameter for Matern correlation

lambda

Box-Cox parameter.

nugget

Observation-level variance

...

Additional arguments for loglik.GRF in geoR.

Value

• the value of the log likelihood.

Details

This function is essentially a wrapper for loglik.GRF in geoR.

Examples

data(swissRain)
swissRain$lograin = log(swissRain$rain)
swissRain$elevation = extract(swissAltitude, swissRain)
loglik.GRF(swissRain, lograin ~ elevation, 
	cov.model="matern", kappa = 2, 
	phi=10000,sigmasq=0.5^2)

loglik.GRF(swissRain, lograin ~ elevation, 
	cov.model="matern", kappa = 2, 
	phi=5000,sigmasq=1)