[R-sig-ME] compare two GAMM4 models using AICs

dani orchidn at live.com
Tue Oct 24 21:49:50 CEST 2017


Hello everyone,


I am fitting two gamm4 models because I would like to see whether there is justification for including a spline term for x1. Can this be done by comparing the AICs for the underlying mixed models (i.e., the "mer" part) of the two models?


b1 <- gamm4(y~x1+offset(e),data=dat,random=~(1|fac))

b2 <- gamm4(y~x1+s(x1)+offset(e),data=dat,random=~(1|fac))


summary(b1$gam)

summary(b1$mer)


summary(b2$gam)

summary(b2$mer)


AIC(b1$mer)

AIC(b2$mer)


Thank you very much!

Best,

DaniNM

<http://aka.ms/weboutlook>

	[[alternative HTML version deleted]]



More information about the R-sig-mixed-models mailing list