inlabru predictions with temporally varying spatial covariates

70 views
Skip to first unread message

Philip Mostert

unread,
Nov 19, 2021, 2:59:23 AM11/19/21
to R-inla discussion group
Dear INLA/inlabru team,

I'm working on a project where we are trying to model the distribution of species locations (coming from two datasets) across different time periods using a bunch of covariates. Some of the covariates considered are temporally varying (across two time periods: time 1 and time 2), and we are hoping to use this information to explain the range shifts of the studied species across different years.

To do this, we used inlabru, and constructed a point process SDM (as seen in the script attached). The model appears to run fine, and we obtain sensible looking results; the issue however comes when we try and use inlabru's predict function. When we predict our spatial model and our temporally varying covariate together, our prediction map becomes identical to our two temporally varying covariate at their respected time period (rather than the intensity of our species at a given location).

My question is whether we have set up our spatial model and predictions correctly in the script such that we should get the predicted intensity across our maps? If not, any advice on such a model would be greatly appreciated!

Kind regards
Philip 
Script_example.R
Reply all
Reply to author
Forward
0 new messages