First INLA fit in katrina-slm.R fails

[rsbivand]

Any ideas?

> semm1<-inla(
+    update(f1, ~.+f(idx, model="slm", args.slm=args.slm, hyper=hyper.slm)),
+    data=Katrina, family="binomial",
+    control.family = list(link="probit", hyper=zero.variance),
+    control.compute=list(dic=TRUE, cpo=TRUE)
+ )
Error in inla.set.hyper(family[i.family], "likelihood", cont.family[[i.family]]$hyper,  : 
  Model binomial [ likelihood ], has none hyperparameters, but 'hyper' is  list(prec = list(initial = 25, fixed = TRUE))

[becarioprecario]

I thought I had removed that. The binomial likelihood does not have a ‘prec' hyper parameter so zero.variance is not required here and, in fact, it should not be used. Simply remove hyper=zero.variance El 3 feb 2021, a las 20:21, Roger Bivand <notifications@github.com<mailto:notifications@github.com>> escribió: Any ideas?

semm1<-inla(

+ update(f1, ~.+f(idx, model="slm", args.slm=args.slm, hyper=hyper.slm)), + data=Katrina, family="binomial", + control.family = list(link="probit", hyper=zero.variance), + control.compute=list(dic=TRUE, cpo=TRUE) + ) Error in inla.set.hyper(family[i.family], "likelihood", cont.family[[i.family]]$hyper, : Model binomial [ likelihood ], has none hyperparameters, but 'hyper' is list(prec = list(initial = 25, fixed = TRUE)) — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub<#2>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/ABYD6WSXHWA2SFCK63J2I6DS5GO4DANCNFSM4XBNMXHQ>.

[rsbivand]

Thanks, done.