I have a question regarding the gamlss package. I am attempting to fit a mixed effects model using the Befa Inflated distribution as follows
gam <- gamlss(y~time+re(random=list(ID=pdDiag(~1+time)),method="REML"), sigma.formula=~1, nu.formula=~1, tau.formula=~1, family= BEINF, data=dat)
I wanted to
1) derive estimates for the variance components of the model...i.e. for the random intercept and random coefficient associated with time
2) get predicted values based on new subjects that takes into consideration the random effects so I can do a visual predictive check
What functions calls in gamlss are there to achieve this?
In regard to 1, I am aware of the getSmo function which produces the output below. But from this output it is not entirely clear what are the variance estimates corresponding to intercept and time.
> getSmo(gam2,what="mu") Linear mixed-effects model fit by maximum likelihood Data: Data Log-likelihood: -19035.68 Fixed: fix.formula (Intercept) 0.3913904 Random effects: Formula: ~1 + time | ID Structure: Diagonal (Intercept) time Residual StdDev: 1.331172 0.3315697 0.678072 Variance function: Structure: fixed weights Formula: ~W.var Number of Observations: 2750 Number of Groups: 250