INLA Posterior Sample

[Abi Riley]

Hi!

I have developed a spatiotemporal SPDE model for air pollution in INLA. The different models are described here: arxiv.org/abs/2506.10688

I selected a final model, that ended up being:

f.iid <- logPM25 ~ -

1 + Intercept + background + PM25PCM +  

  f(Time, NDVI_New, model = "ar1", hyper = hyper.ar13) +

  f(spatial.field, model = spde, group = spatial.field.group, control.group = list(model="ar1", hyper = hyper.ar1.rho)) +
  f(spatial.varying.field, model = spde)

A couple of linear covariates, a RW1 time varying coeff, a SPDE spatially varying coeff, and the ST SPDE. 

The data is: 3500 points * 72 times, so not too large.

However, now I am interested in calculating an exposure for cohort participants with uncertainty. However, inla.posterior.sample is not working, it works for some of the simpler models but otherwise just runs and doesn't crash. And I've tried through a cluster. I've also tried just extracting n = 1 samples.

Do you have any advice here of what the issue is, if it's size/complexity or something else?

Any help would be amazing here

Thanks,

Abi Riley

a.ri...@imperial.ac.uk

[INLA help]

Any chance of getting a copy of data and code so I could rerun it here to figure out why?

—

Haavard Rue

Helpdesk

he...@r-inla.org

--
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, visit https://groups.google.com/d/msgid/r-inla-discussion-group/427562e1-434a-4a14-9fb6-29b43f8b87edn%40googlegroups.com.