[R-sig-Geo] help with likfit.glsm in package geoRglm

[Ken Nussear]

I'm trying to fit parameters in a spatial model using likfit.glsm and  
I'm getting an error. I wonder if anyone out there can help me figure  
out how to fix it.

the object WMgeo is a geospatial dataset and I have run successful  
models using likfit in the geoR package. I just cant get this one to go.

Please help!

 > WMmcmccontrl4 <- mcmc.control(S.scale = .00015, thin=1, phi.scale  
= 0.01)
 > WMliktrend4 <- trend.spatial(~TRAN_LNTH + GRSPSTBG + STAll, WMgeo)
 > WMmcmcmodel4 <- list(cov.pars = c(1,1), beta = c(1,1,1,1),  
family="binomial", trend= WMliktrend4)
 > WMmcmctest4 <- glsm.mcmc(WMgeo, model = WMmcmcmodel4,  
mcmc.input=WMmcmccontrl4)
iter. numb. 1000  : Acc.-rate =  0.66
MCMC performed: n.iter. =  1000 ; thinning =  1 ; burn.in =  0
 > WMmcmcprep4 <- prepare.likfit.glsm(WMmcmctest4, use.intensity=FALSE)
 > WMmcmclikfit4 <- likfit.glsm(WMmcmcprep4, cov.model="spherical",  
ini.phi=.1)
--------------------------------------------------------------------
likfit.glsm: likelihood maximisation using the function optim.
phi =  0.1 tausq.rel =  0
    Min. 1st Qu.  Median    Mean 3rd Qu.    Max.
       0       0       0     Inf     Inf     Inf
Error in .func.val(SivS, SivD, DivD, beta.hat[ll, ], sigmasq.hat[ll],  :
	Some function values are not finite




Ken

--
Kenneth E. Nussear
Research Herpetologist, USGS
(702) 564-4515  Fax  (702) 564-4600
knussear at usgs.gov