loglik.linear.model: Profile log-likelihood or fixed parameters likelihood...

In barryrowlingson/PrevMap: Geostatistical Modelling of Spatially Referenced Prevalence Data

Description

Computes profile log-likelihood, or evaluatesx likelihood keeping the other paramaters fixed, for the scale parameter phi of the Matern function and the relative variance of the nugget effect nu2 in the linear Gaussian model.

Usage

loglik.linear.model(object, control.profile, plot.profile = TRUE,
  messages = TRUE)

Arguments

object	an object of class 'PrevMap', which is the fitted linear model obtained with the function linear.model.MLE.
control.profile	control parameters obtained with control.profile.
plot.profile	logical; if TRUE a plot of the computed profile likelihood is displayed.
messages	logical; if messages=TRUE then status messages are printed on the screen (or output device) while the function is running. Default is messages=TRUE.

Value

an object of class "profile.PrevMap" which is a list with the following values

eval.points.phi: vector of the values used for phi in the evaluation of the likelihood.

eval.points.rel.nugget: vector of the values used for nu2 in the evaluation of the likelihood.

profile.phi: vector of the values of the likelihood function evaluated at eval.points.phi.

profile.rel.nugget: vector of the values of the likelihood function evaluated at eval.points.rel.nugget.

profile.phi.rel.nugget: matrix of the values of the likelihood function evaluated at eval.points.phi and eval.points.rel.nugget.

fixed.par: logical value; TRUE is the evaluation if the likelihood is carried out by fixing the other parameters, and FALSE if the computation of the profile-likelihood was performed instead.

Author(s)

Emanuele Giorgi [e-mail]

Peter J. Diggle [e-mail]

barryrowlingson/PrevMap documentation built on May 11, 2019, 6:24 p.m.