Question with BAPC+INLA analysis

[Song Cheng]

bapc_result <- BAPC(lc_HNC, predict = list(npredict = 30, retro = FALSE),secondDiff = FALSE, stdweight = wstand, verbose = TRUE)

Error in inla.check.control(control.compute, data) :
 Name `return.marginals.predictor' in control-argument `control.compute', is void.

  Valid ones are:
    config
    cpo
    dic
    gdensity
    graph
    hyperpar
    mlik
    openmp.strategy
    po
    q
    return.marginals
    smtp
    waic

I would like to use BPAC model to predict the cancer burden but constantly received this error. I do not know how to fix it? What is the meaning of Name..... is void?

[Song Cheng]

I reaaaaaaaaaaaaaaally need someone to help me. Have been stuck for several days! 

[pav]

I can't really help on the substantive question, but in this context "void" means "invalid" (as you might guess from the list of "valid" arguments that the error message helpfully supplies.) So what it means is that somehow inla() has received a specification of its control.compute argument that includes "return.marginals.predictor", which it thinks is unsupported (the supported argument is "return.marginals"). Since the control.compute argument doesn't  actually appear in the example you posted, I'd guess that this is generated by the BAPC function which presumably calls inla(), and you should probably take it up with the author of that package. But before you do, it might be a good idea to check whether the BAPC example at https://rdrr.io/rforge/BAPC/man/BAPC.html gives the same error message, and if not, why your case and the one there are different.

[Finn Lindgren]

Hi,

the latest INLA version does have return.marginals.predictor in control.compute.
Check what INLA version you have, and if you have something older than
22.09.15, upgrade it and test the BAPC estimation again.
Finn

> --
> 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/2227fa0d-7e62-4706-949d-c97c96648d4an%40googlegroups.com.



--
Finn Lindgren
email: finn.l...@gmail.com

[Havard Rue]

This is an issue with the BAPC package, that they do not update it along the developments of INLA.

--

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/11d4c357-820a-438d-9f95-0d2f5ac4b308n%40googlegroups.com.

[Finn Lindgren]

In this particular case, I believe it's just due needing to install
the latest INLA version; searching online, I found the BAPC package
was updated not long ago to support updated INLA.
Song Cheng, check what BAPC and INLA version you have, and make sure
they are both updated.
If you still get the error after that, you'll need to contact the BAPC
developers to get it fixed.
Finn

> To view this discussion on the web, visit https://groups.google.com/d/msgid/r-inla-discussion-group/DU0PR09MB5545C7DB0B7F6D8D77402969F0559%40DU0PR09MB5545.eurprd09.prod.outlook.com.