model doesn't run with larger mesh

[Sylvan Benaksas]

Hi,

I am running a model that has  ~1460 mesh nodes. If I try to make the triangles smaller and create a mesh with anything over this number of nodes the model fails after several  minutes with the following error message, is this a memory issue or something else?

arse section=[18] name=[INLA.Parameters] type=[INLA] inla_parse_INLA... section[INLA.Parameters] lincomb.derived.correlation.matrix = [No] global_node.factor = 2.000 global_node.degree = 2147483647 reordering = -1 constr.marginal.diagonal = 1.49e-08 Contents of ai_param 00000000091fed40 Optimiser: DEFAULT METHOD Option for GSL-BFGS2: tol = 0.1 Option for GSL-BFGS2: step_size = 1 Option for GSL-BFGS2: epsx = 0.01 Option for GSL-BFGS2: epsf = 0.02 Option for GSL-BFGS2: epsg = 0.01 R

Error in inla.inlaprogram.has.crashed() : The inla-program exited with an error. Unless you interupted it yourself, please rerun with verbose=TRUE and check the output carefully. If this does not help, please contact the developers at <he...@r-inla.org>. Error in inla.core.safe(formula = formula, family = family, contrasts = contrasts, : *** Fail to get good enough initial values. Maybe it is due to something else.

Cheers,

Sylvan

[INLA help]

Probably something else.  To small triangles can cause numerical singularities.   

I can rerun here to check if u share 

—

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 on the web, visit https://groups.google.com/d/msgid/r-inla-discussion-group/38e7f392-e57b-489e-9694-964fcd1db7dfn%40googlegroups.com.

[Sylvan Benaksas]

Hi Haavard,

Following another group post https://groups.google.com/g/r-inla-discussion-group/c/M7PfSQPkxuw I added control.inla = list(reordering = "metis") and the reduced size model is now running, I will test to see if it finishes fitting and then I will test the full size model,

Thanks,

Sylvan