[R-sig-Geo] Moran I for generalized linear model

Roger Bivand Roger.Bivand at nhh.no
Thu Jun 17 16:48:56 CEST 2010

On Thu, 17 Jun 2010, elaine kuo wrote:

> This is Elaine.
>> I am try the R package spdep to calculate Moran I for each of the 7
>> generalized linear
>> models generated by 3 predictors respectively.

There is no accepted authority for any test for the residuals of a glm() 
fit for spatial autocorrelation (a paper in Geographical Analysis is not 
adequate authority - one in Biometrika or Statistics in Medicine might 
be). Because a glm object is also an lm object, you may - at your own risk 
- use lm.morantest() as if the fitted model was an lm object. Since it is 
very hard to simulate correlated multivariate counts, we have no idea 
whether this test behaves as it should.

Hope this helps,


>    The models were generated by bird richness and the habitual conditions,
> based on a grid system with the number corresponding to the variables of the
> models.
>> I read the pdf manual but still confounded which function should be used to
>> achieved the goal.
> Thanks
> 	[[alternative HTML version deleted]]
> _______________________________________________
> R-sig-Geo mailing list
> R-sig-Geo at stat.math.ethz.ch
> https://stat.ethz.ch/mailman/listinfo/r-sig-geo

Roger Bivand
Economic Geography Section, Department of Economics, Norwegian School of
Economics and Business Administration, Helleveien 30, N-5045 Bergen,
Norway. voice: +47 55 95 93 55; fax +47 55 95 95 43
e-mail: Roger.Bivand at nhh.no

More information about the R-sig-Geo mailing list