search

spatialstatisticsupna / bigDM

R package for scalable Bayesian disease mapping models for high-dimensional data
13 stars 4 forks source link

Merging the results of K-order models with 5 covariates #3

Closed seyiaros closed 11 months ago

seyiaros commented 11 months ago

Hello

I am using the CAR_INLA function to fit a spatial model with a spatial domain of 6791 divided into 9 sub-regions. I am following your example of fitting spatial models but with a vector object of 5 covariates (X). The model runs fine from steps 1 - 2 but I get an error message "Error in checkforRemoteError(val): 10 nodes produced errors; first error: invalid 'row.names' length. My model strategy is simplified.laplace and K = 1. I am not sure of why I am getting this error message. Thank you

aritz-adin commented 11 months ago

Dear @seyiaros,

Could you please send me a reproducible example in order to check what is happening?

Best, Aritz

seyiaros commented 11 months ago

Dear Aritz

Thank you for getting back. I resolved the issue - All the grouping variables should be of a character format variable: I initially had a numeric format grouping variable.

Regards Seyi

aritz-adin commented 11 months ago

Nice!

I will try to modify the code to accept numeric variables as grouping variables as well.

Best, Aritz