https://github.com/cran/RandomFields
Tip revision: e5a7a2f272b7834f96c925ced7acfa0c6456a87f authored by Martin Schlather on 17 April 2017, 22:09:51 UTC
version 3.1.50
version 3.1.50
Tip revision: e5a7a2f
RMS.Rd
\name{RMS}
\alias{RMS}
\title{Scaling operator}
\description{
\command{\link{RMS}} is an operator that modifies the variance and the
coordinates or distances of a submodel \eqn{\phi} by
\deqn{C(h) = v * \phi(A*h/s).}
\bold{Most users will never call \command{\link{RMS}} directly, see the
details.}
}
\usage{
RMS(phi, var, scale, Aniso, proj, anisoT)
}
\arguments{
\item{phi}{submodel}
\item{var}{is the optional variance parameter \eqn{v},
It can be also an arbitrary non-negative function.
}
\item{scale}{scaling parameter \eqn{s} which is positive}
\item{Aniso}{matrix or \code{\link{RMmodel}}.
The optional anisotropy matrix \eqn{A}, multiplied from
the right by a distance vector \eqn{x}, i.e. \eqn{Ax}
}
\item{proj}{is the optional projection vector which defines a diagonal
matrix of zeros and ones and \code{proj} gives the
positions of the ones (integer values between 1 and teh dimension of
\eqn{x}).}
\item{anisoT}{the transpose of the anisotropy matrix \eqn{B},
multiplied from the left by a distance vector \eqn{x}, i.e. \eqn{x^\top
B}.
}
}
\value{
\command{\link{RMS}} returns an object of class \code{\link[=RMmodel-class]{RMmodel}}.
}
\details{
The call in the usage section is equivalent to
\code{phi(..., var, scale, anisoT, Aniso, proj)}, where \code{phi} has
to be replaced by a valid \command{\link{RMmodel}}
Most users will never call \command{\link{RMS}} directly.
}
\note{At most one of the arguments,
\code{Aniso}, \code{anisoT} and \code{proj} may be given at the same time.
}
\author{Martin Schlather, \email{schlather@math.uni-mannheim.de}
}
\seealso{
\command{\link{RMmodel}},
\command{\link{RMprod}} for an alternative way to define
an arbitrary, location dependent variance. There the standard
deviation is given so that \command{\link{RMprod}} might be used
even in the multivariate case.
}
\keyword{spatial}
\keyword{models}
\examples{
RFoptions(seed=0) ## *ANY* simulation will have the random seed 0; set
## RFoptions(seed=NA) to make them all random again
\dontshow{StartExample()}
model1 <- RMS(RMexp(), scale=2)
model2 <- RMexp(scale=2)
x <- seq(0, 10, 0.02)
print(all(RFcov(model1, x) == RFcov(model2, x))) # TRUE
\dontshow{FinalizeExample()}
}