[prev in list] [next in list] [prev in thread] [next in thread] 

List:       r-sig-geo
Subject:    [R-sig-Geo] Prediction using Autologistic Regression
From:       "=?iso-8859-1?Q?Jaime_R._Garc=EDa_M=E1rquez?=" <jaime.garcia () uni-bonn ! de>
Date:       2008-07-21 13:15:01
Message-ID: 000c01c8eb33$c81f8ef0$585eacd0$ () garcia () uni-bonn ! de
[Download RAW message or body]

Dear list,

 

 

 

I am working with presence-absence data and testing, among others, the
autologistic regression model.

 

 

 

The first thing I do is to calculate the autocovariate term using the
function autocov_dist() (package spdep) ONLY for the pixels defined by the
location of my presence-absence data.

 

 

 

Then I run a GLM, including the autocovariate term as an explanatory
variable.

 

 

 

Now I want to predict to the entire study area using the GLM object.  The
problem here is that when I specify as newdata the environmental variables
for the entire study area, the autocovariate term is not included and the
function predict.glm() does not work.

 

 

 

My interpretation of the given error is that the model (GLM object) includes
the autocovariate term (which appears highly significant) but this term is
not included in the set of environmental variables for the entire study
area.

 

 

 

What options do I have?  Should I create and add a new autocovariate term to
the environmental variables? Should I predict leaving out from the GLM
object the autocovariate term? How?

 

 

 

I appreciate any help you can give me

 

 

 

Best wishes,

 

 

 

Jaime

 


	[[alternative HTML version deleted]]

_______________________________________________
R-sig-Geo mailing list
R-sig-Geo@stat.math.ethz.ch
https://stat.ethz.ch/mailman/listinfo/r-sig-geo
[prev in list] [next in list] [prev in thread] [next in thread]