variograms do not satisfy a legal model

Edzer,

Could you tell me how to override this check? I have been reading the
helpfile but I can't find it.

Thanks,

Thierry

------------------------------------------------------------------------
----
ir. Thierry Onkelinx
Instituut voor natuur- en bosonderzoek / Reseach Institute for Nature
and Forest
Cel biometrie, methodologie en kwaliteitszorg / Section biometrics,
methodology and quality assurance
Gaverstraat 4
9500 Geraardsbergen
Belgium
tel. + 32 54/436 185
Thierry.Onkelinx at inbo.be
www.inbo.be 

Do not put your faith in what statistics say until you have carefully
considered what they do not say.  ~William W. Watt
A statistical analysis, properly conducted, is a delicate dissection of
uncertainties, a surgery of suppositions. ~M.J.Moroney

 

  

-----Oorspronkelijk bericht-----
Van: Edzer J. Pebesma [mailto:e.pebesma at geo.uu.nl] 
Verzonden: maandag 7 mei 2007 17:00
Aan: ONKELINX, Thierry
CC: r-sig-geo at stat.math.ethz.ch
Onderwerp: Re: [R-sig-Geo] variograms do not satisfy a legal model

Thierry, you'll have to find out what the linear model of 
regionalization or intrinsic correlation mean; they are legal 
models, meaning that they lead to predictions with guaranteed 
non-negative prediction variances. There's a way to override 
the check (meaning it will not stop on error) but not to get 
that guarantee.
--
Edzer

ONKELINX, Thierry wrote:
    

Dear useRs,

I'm trying to do some cokriging. I try fit the variograms 
      

with fit.lmc 
    

(gstat package).

#Problem 1: how to get the model in an elegant way into the gstat 
object? This is how I do it now.
library("gstat")
data(meuse)
g <- gstat(id = "ln.zinc", formula = log(zinc)~1, locations = ~x+y, 
data = meuse) g <- gstat(g, id = "ln.lead", formula = log(lead)~1, 
locations = ~x+y, data = meuse) #fit the variogram models using 
fit.lmc model <- fit.lmc(variogram(g), g, model = vgm(.55, 
      

"Sph", 900, 
    

.05), fit.ranges = FALSE) g <- gstat(g, id = "ln.zinc", model = 
model[["model"]][["ln.zinc"]]) g <- gstat(g, id = 
      

"ln.lead", model = 
    

model[["model"]][["ln.lead"]]) g <- gstat(g, id = c("ln.zinc", 
"ln.lead"), model =
model[["model"]][["ln.zinc.ln.lead"]])
predict(g, newdata = meuse)

#Problem 2. the gstat object doesn't accept different 
      

ranges. Or am I 
    

doing something wrong?
library("gstat")
data(meuse)
g <- gstat(id = "ln.zinc", formula = log(zinc)~1, locations = ~x+y, 
data = meuse) g <- gstat(g, id = "ln.lead", formula = log(lead)~1, 
locations = ~x+y, data = meuse) # examine variograms and cross 
variogram:
plot(variogram(g))
# enter direct variograms:
g <- gstat(g, id = "ln.zinc", model = vgm(.55, "Sph", 800, 
      

.05)) g <- 
    

gstat(g, id = "ln.lead", model = vgm(.55, "Sph", 900, .05)) # enter 
cross variogram:
g <- gstat(g, id = c("ln.zinc", "ln.lead"), model = vgm(.47, "Sph", 
900,
.03))
predict(g, newdata = meuse)
# Error in predict.gstat(g, newdata = meuse) : gstat: value not 
allowed
for: variograms do not satisfy a legal model

Thanks,

Thierry


      

----------------------------------------------------------------------
    

--
----
ir. Thierry Onkelinx
Instituut voor natuur- en bosonderzoek / Reseach Institute 
      

for Nature 
    

and Forest Cel biometrie, methodologie en kwaliteitszorg / Section 
biometrics, methodology and quality assurance Gaverstraat 4 9500 
Geraardsbergen Belgium tel. + 32 54/436 185 
      

Thierry.Onkelinx at inbo.be 
    

www.inbo.be

Do not put your faith in what statistics say until you have 
      

carefully 
    

considered what they do not say.  ~William W. Watt A statistical 
analysis, properly conducted, is a delicate dissection of 
uncertainties, a surgery of suppositions. ~M.J.Moroney