Pars is used to update the coefficients of the model components
of
object and the individual (restricted) log-likelihood
contributions of each component are added together. The type of
log-likelihood (restricted or not) is determined by the
settings
attribute of
object.
USAGE:
logLik(object, Pars, conLin)
REQUIRED ARGUMENTS:
object
an object inheriting from class
glsStruct,
representing a list of linear model components, such as
corStruct and
varFunc objects.
Pars
the parameter values at which the (restricted)
log-likelihood is to be evaluated.
OPTIONAL ARGUMENTS:
conLin
an optional condensed linear model object, consisting of
a list with components
"Xy", corresponding to a regression
matrix (
X) combined with a response vector (
y), and
"logLik", corresponding to the log-likelihood of the
underlying linear model. Defaults to
attr(object, "conLin").
VALUE:
the (restricted) log-likelihood for the linear model described by
object
, evaluated at
Pars.