[R-sig-Geo] help for fit.vaiogram.gls

[Saman Monfared]

Hi.
Itry to estmate prameters of vaigrm by  fit.vaiogram.gls:

rm(list=ls())
library(gstat)
memory.limit(4000)
data(meuse)
coordinates(meuse)<-~x+y
v.fit4<-fit.variogram.gls(log(zinc)~1, meuse[1:40,], vgm(1, "Sph", 900,1))

the results are:

[1] -1.939367e-17  6.324797e-01  4.123851e+02
[1] 4.846680e-05 7.312376e-01 3.661078e+02
[1] 4.846680e-05 2.421149e-01 5.450078e+02
[1] 4.846680e-05 2.968638e-01 7.078753e+02
[1] 4.846680e-05 3.525650e-01 8.629175e+02
[1] 4.846680e-05 4.078170e-01 1.013148e+03
[1] 4.846680e-05 4.624658e-01 1.160053e+03
[1] 4.846680e-05 5.165518e-01 1.304490e+03
[1] 4.846680e-05 5.701438e-01 1.361356e+03
[1] 4.846680e-05 5.912904e-01 1.361355e+03
[1] 4.846680e-05 5.912685e-01 1.361355e+03(**)
Warning message:
In fit.variogram.gls(log(zinc) ~ 1, meuse[1:40, ], vgm(0.04, "Mat",  :
  range parameter at search space boundary

Does the las model (**) is  my answer??
I have changed the start values several times but The warning remain.

Thanks.


-- 

Saman Monfared
Msc Student,
Department of Statistics, Shiraz University,
Shiraz 71454, Iran
Email: Samanmonfared1 at gmail.com

Tel: +98 917 5305167