[R-sig-Geo] retrieving fitted values from geoR model
Ken Nussear
knussear at usgs.gov
Wed Nov 7 13:39:30 CET 2007
Just in case anyone needs this in the future
Looks like I am able to get them using this call
fitted.likGRF(m4r)
Ken
>
>
> Message: 2
> Date: Tue, 6 Nov 2007 16:52:05 -0800
> From: Ken Nussear <knussear at usgs.gov>
> Subject: [R-sig-Geo] retrieving fitted values from geoR model
> To: r-sig-geo at stat.math.ethz.ch
> Message-ID: <AAC0BCC3-D58E-4082-AFD1-3FBEC440ED8E at usgs.gov>
> Content-Type: text/plain; charset=US-ASCII; format=flowed; delsp=yes
>
> I'm trying to obtain the fitted values for a spatial model using
> likfit under geoR. When I use the statement
>
> fitted(m4r, spatial = TRUE), per the instructions under help(likfit)
>
> I get the following error
>
>> fitted(m4r, spatial = TRUE)
> be patient ... this function currently require calling likfit again
> Error in eval(object.call) : object "object.call" not found
>
> If I ask the object for the call I get
>
>> m4r$call
> likfit(geodata = spUta, trend = ~TransectLength + all.road +
> Urban, ini.cov.pars = c(0.82, 1949), nugget = 0.55, cov.model =
> "exponential",
> method.lik = "REML")
>
>
>
> Can anyone provide help?
>
> Thanks
>
> Ken
>
>
>
> The model summary is:
>
>> summary(m4r)
> Summary of the parameter estimation
> -----------------------------------
> Estimation method: restricted maximum likelihood
>
> Parameters of the mean component (trend):
> beta0 beta1 beta2 beta3
> 1.1742 0.0008 -0.3758 -0.0001
>
> Parameters of the spatial component:
> correlation function: exponential
> (estimated) variance parameter sigmasq (partial sill) = 1.167
> (estimated) cor. fct. parameter phi (range parameter) = 1949
> anisotropy parameters:
> (fixed) anisotropy angle = 0 ( 0 degrees )
> (fixed) anisotropy ratio = 1
>
> Parameter of the error component:
> (estimated) nugget = 0.2518
>
> Transformation parameter:
> (fixed) Box-Cox parameter = 1 (no transformation)
>
> Maximised Likelihood:
> log.L n.params AIC BIC
> "-75.92" "7" "165.8" "180.3"
>
> non spatial model:
> log.L n.params AIC BIC
> "-81.95" "5" "173.9" "184.2"
>
> Call:
> likfit(geodata = spUta, trend = ~TransectLength + all.road +
> Urban, ini.cov.pars = c(0.82, 1949), nugget = 0.55, cov.model =
> "exponential",
> method.lik = "REML")
More information about the R-sig-Geo
mailing list