Hello everyone,
I would like to fit a glm regression model on spatial lattice data.
Kaluzny et al, in s+spatialstats users manual refer that rather than fitting
mju (the mean) with slm, one could model the mean using other functions in
s-plus, and then model the remaining correlated errors using slm. He then
goes on to an example of fitting a spatially lagged model.
Since I am not so familiar with the method. Could anyone elaborate on this
(perhaps using a logistic model as an example)?
Many thanks
Miguel
____________________________________________
Dr Miguel Repas
Centro de Ecologia e Biologia Vegetal
Faculdade de Ciências de Lisboa
http://cebv.fc.ul.pt
____________________________________________
|