Problems with the "bru_fill_missing" command #200
-
Good morning everyone, after the latest updates to INLA and INLABRU, the codes that I was previously using are no longer functioning. The issue specifically pertains to the calculation of covariates. Prior to including them in the model, I used the following code: Create mesh projection --------------------------------------------------xx = range(mesh_medit$loc[,1]) A = inla.spde.make.A(mesh = mesh_medit, loc = new_xy) ips <- inlabru::ipoints(domain = mesh_medit) ################COVARIATES############## load("data/BASEcovariates.RData") FIXED COVARIATESDepth -------------------------------------------------------------------depth_raster <- rasterFromXYZ(depthFinal) values <- eval_spatial(data = depth_SPDF_s, values2 <- inlabru:::bru_fill_missing(data = depth_SPDF_s, depth_SPDF_s = SpatialPixelsDataFrame(points = new_xy, Now I'm encountering the following error: I'm using version 2.8.0 of INLABRU; everything was working with version 2.7.0. |
Beta Was this translation helpful? Give feedback.
Replies: 1 comment 10 replies
-
Hi, I'm preparing to release version 2.9.0, so can you please check if the problem has been solved in the latest development version (2.8.0.9015)? Also, what INLA version did you use, and do you have The Can you show me the output from these commands in your current installation? |
Beta Was this translation helpful? Give feedback.
Thanks. The last two show that while the "user input" parts of the crs of the converted objects are the same, they used different methods to construct the WKT (which is the actual information used) that generated equivalent but text-wise different representations. That's annoying...
I think one can work around this by transforming one of the objects into the version of the crs of the other object.
Please try this:
Let me know if that works, and I can implement that workaround.