https://github.com/cran/spatstat
Tip revision: 1b6f7ce253524f978d8945645a567ffe709bf90a authored by Adrian Baddeley on 13 October 2016, 01:00:59 UTC
version 1.47-0
version 1.47-0
Tip revision: 1b6f7ce
logLik.slrm.Rd
\name{logLik.slrm}
\Rdversion{1.1}
\alias{logLik.slrm}
\title{
Loglikelihood of Spatial Logistic Regression
}
\description{
Computes the (maximised) loglikelihood of a fitted
Spatial Logistic Regression model.
}
\usage{
\method{logLik}{slrm}(object, ..., adjust = TRUE)
}
\arguments{
\item{object}{
a fitted spatial logistic regression model.
An object of class \code{"slrm"}.
}
\item{\dots}{
Ignored.
}
\item{adjust}{
Logical value indicating whether to adjust the loglikelihood
of the model to make it comparable with a point process
likelihood. See Details.
}
}
\details{
This is a method for \code{\link[stats]{logLik}} for fitted spatial logistic
regression models (objects of class \code{"slrm"}, usually obtained
from the function \code{\link{slrm}}). It computes the log-likelihood
of a fitted spatial logistic regression model.
If \code{adjust=FALSE}, the loglikelihood is computed
using the standard formula for the loglikelihood of a
logistic regression model for a finite set of (pixel) observations.
If \code{adjust=TRUE} then the loglikelihood is adjusted so that it
is approximately comparable with the likelihood of a point process
in continuous space, by subtracting the value
\eqn{n \log(a)}{n * log(a)}
where \eqn{n} is the number of points in the original point pattern
dataset, and \eqn{a} is the area of one pixel.
}
\value{
A numerical value.
}
\seealso{
\code{\link{slrm}}
}
\examples{
X <- rpoispp(42)
fit <- slrm(X ~ x+y)
logLik(fit)
logLik(fit, adjust=FALSE)
}
\author{\adrian
\email{adrian@maths.uwa.edu.au}
and \rolf
}
\keyword{spatial}
\keyword{models}
\keyword{methods}