Joint models in INLA using "current value" parameterisation

99 views
Skip to first unread message

Annabel Webb

unread,
Sep 20, 2023, 2:21:57 AM9/20/23
to R-inla discussion group
Hi all,

I am very new to INLA but interested to use it for joint modelling of longitudinal and survival outcomes. In particular I'm interesting in fitting joint models where the current value of the longitudinal outcome is included directly as a covariate in the survival model (as opposed to a shared random effect set-up). 

After following some examples found in this paper: https://doi.org/10.18637/jss.v100.i02 and elsewhere, I have tried to fit the same fairly simple joint model using both the JM package and via INLA (R code is attached), using the pbc.id and pbc2 datasets from the JM package. Looking at the output, it seems like some of the parameter estimates are quite different between my two fitted models, so I think I am still going wrong when trying to translate the model to specify it in INLA. 

Could anyone give me pointers on where I am going wrong here? 

On a related note - could anyone give some intuition as to why the appropriate value for the "w" variable used to copy the linear predictor from the longitudinal model to the survival model is -1?

Any help or comments would be appreciated!

Kind regards,
Annabel


JM_vs_INLA.R

Janet Van Niekerk

unread,
Sep 20, 2023, 2:37:48 AM9/20/23
to Annabel Webb, R-inla discussion group
Hi Annabel,

There is now a new package INLAjoint that is specifically for fitting joint models because as you can see in that paper, it is quite tedious.

--
You received this message because you are subscribed to the Google Groups "R-inla discussion group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to r-inla-discussion...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/r-inla-discussion-group/68bebdff-461f-4797-9243-d81852be63b5n%40googlegroups.com.
Reply all
Reply to author
Forward
0 new messages