I want to assess local most important predictors for a specific event within a region. The same locations (grid) were all sampled in two different time occasions. I am assuming predictors to be spatially varying, but temporally stationary. The problem is that distance matrix calculation fails to run due to duplicate locations. Is there a way to carry this "atemporal" GWLR, or a workaround, interpreting the two different regression analyses as a global model?
I am running GWmodel, in R.