Skip to content
Prev 7030 / 29559 Next

Assessing residual spatial autocorrelation in a Poisson or Negative Binomial model

On Thu, 26 Nov 2009, Luis Iv?n Ortiz Valencia wrote:

            
Rather:

http://cran.r-project.org/view=Spatial

please, the sourceforge site is more for development, and is linked from 
the task view on your nearest CRAN mirror.

While lm.morantest() can be used on glm output objects, no work has been 
done to establish whether this is a sensible idea. It remains problematic 
to simulate spatially dependent discrete variables. However, it is 
possible that if you ignore the "test" of doubtful substance, you could 
track how Moran's I moves when adding variables in an exploratory way. Try 
with a smaller dataset first.

Hope this helps,

Roger