I am checking to see if there have been any updates on setting anisotropy in INLA-SPDE. I inquired with your team a few years ago, and there were no such settings available at that time. There are some packages inspired by R-INLA, but based on the TMB approach, that include an anisotropy option. Typically, these packages add a rotation matrix, H, to the range parameter. In my opinion, it is similar to the Barrier model within the INLA framework (when I say "similar", I mean they both adjust the range parameter, but of course, differently).
I saw a post from Finn last year mentioning that your team is working on a better approach to address the anisotropy issue. I would like to get the most recent updates on this.
Thanks,
To view this discussion, visit https://groups.google.com/d/msgid/r-inla-discussion-group/c50d321f5cd53ef1f71175f6fbeb13aab3ed618d.camel%40r-inla.org.