https://github.com/cran/RandomFields
Tip revision: 6eca414de4c835af2032db4cae6c05e9cc684529 authored by Martin Schlather on 23 April 2016, 15:04:07 UTC
version 3.1.11
version 3.1.11
Tip revision: 6eca414
RFfitOptimiser.Rd
\name{RFfitoptimiser}
\alias{RFfitoptimiser}
\alias{RFfitOptimiser}
\title{Optimisers for
fitting model parameters to spatial data
}
\description{
See \command{\link{RFfit}} for a detailed description of
the fitting procedure.
}
\details{
Two parameters, see also \command{\link{RFoptions}}
can be passed to \command{\link{RFfit}} that allow
for choosing an optimiser different from \command{\link[stats]{optim}}:
\code{optimiser}takes one of the values
\code{"optim"}, \code{"optimx"}, \code{"soma"}, \code{"nloptr"},
\code{"GenSA"}, \code{"minqa"}, \code{"pso"} or \code{"DEoptim"},
see the corresponding packages for a description
If \code{optimiser="nloptr"}, then the additional parameter
\code{algorithm} must be given which takes the values
\code{"NLOPT_GN_DIRECT"},
\code{"NLOPT_GN_DIRECT_L"},
\code{"NLOPT_GN_DIRECT_L_RAND"},
\code{"NLOPT_GN_DIRECT_NOSCAL"},
\code{"NLOPT_GN_DIRECT_L_NOSCAL"},
\code{"NLOPT_GN_DIRECT_L_RAND_NOSCAL"},
\code{"NLOPT_GN_ORIG_DIRECT"},
\code{"NLOPT_GN_ORIG_DIRECT_L"},
\code{"NLOPT_LN_PRAXIS"},
\code{"NLOPT_GN_CRS2_LM"},
\code{"NLOPT_LN_COBYLA"},
\code{"NLOPT_LN_NELDERMEAD"},
\code{"NLOPT_LN_SBPLX"},
\code{"NLOPT_LN_BOBYQA"},
\code{"NLOPT_GN_ISRES"},
see \pkg{nloptr} for a description.
}
\author{Martin Schlather, \email{schlather@math.uni-mannheim.de}
\url{http://ms.math.uni-mannheim.de/de/publications/software}}
\seealso{
\command{\link{RFfit}}, \command{\link{RFoptions}}
}
\examples{
RFoptions(seed=0) ## *ANY* simulation will have the random seed 0; set
## RFoptions(seed=NA) to make them all random again
\dontshow{StartExample()}
## Here some alternative optimisers to 'optim' are considered.
## All but the \pkg{nloptr} algorithms are largely slower than 'optim'.
## Only a few of them return results as good as 'optim'.
data(soil)
str(soil)
soil <- RFspatialPointsDataFrame(
coords = soil[ , c("x.coord", "y.coord")],
data = soil[ , c("moisture", "NO3.N", "Total.N", "NH4.N", "DOC", "N20N")],
RFparams=list(vdim=6, n=1)
)
data <- soil["moisture"]
\dontshow{if (RFoptions()$internal$examples_red) {
warning("data have been reduced !")
All <- 1:7
data(soil)
soil <- RFspatialPointsDataFrame(
coords = soil[All, c("x.coord", "y.coord")],
data = soil[All, c("moisture", "NO3.N", "Total.N",
"NH4.N", "DOC", "N20N")],
RFparams=list(vdim=6, n=1)
)
data <- soil["moisture"]
}}
model <- ~1 + RMwhittle(scale=NA, var=NA, nu=NA) + RMnugget(var=NA)
\dontshow{if (RFoptions()$internal$examples_red){model<-~1+RMwhittle(scale=NA,var=NA,nu=1/2)}}
## standard optimiser 'optim'
print(unix.time(fit <- RFfit(model, data=data)))
print(fit)
opt <- "minqa" # 330 sec %ok
print(unix.time(fit2 <- try(RFfit(model, data=data, optimiser=opt))))
print(fit2)
opt <- "soma" # 450 sec % sehr schlecht
print(unix.time(fit2 <- try(RFfit(model, data=data, optimiser=opt))))
print(fit2)
opt <- "optimx" # 30 sec; better result
print(unix.time(fit2 <- try(RFfit(model, data=data, optimiser=opt))))
print(fit2)
opt <- "nloptr"
algorithm <- RC_NLOPTR_NAMES
\dontshow{if(!interactive()) algorithm <- RC_NLOPTR_NAMES[1]}
for (i in 1:length(algorithm)) { % all algorithms are fast
print(algorithm[i])
print(unix.time(fit2 <- try(RFfit(model, data=data, optimiser=opt,
algorithm=algorithm[i]))))
print(fit2)
}
if (interactive()) {
## the following two optimisers are too slow to be run on CRAN.
opt <- "pso" # 600 sec
print(unix.time(fit2 <- try(RFfit(model, data=data, optimiser=opt))))
print(fit2)
opt <- "GenSA" # 10^4 sec
print(unix.time(fit2 <- try(RFfit(model, data=data, optimiser=opt))))
print(fit2)
}
\dontshow{RFoptions(modus_operandi="normal")}
\dontshow{FinalizeExample()}
}
\keyword{spatial}
\keyword{optimize}