[R-sig-Geo] spatial probit

Eelke Folmer E.O.Folmer at rug.nl
Wed May 28 13:53:58 CEST 2008

Roger Bivand mentions LeSage's matlab toolbox
(http://www.spatial-econometrics.com/), including very clear manuals with
theory. I am unable to judge the theoretical validity of the routines but I
experienced that I was able to estimate parameters of spatial probit
regression models (with MCMC) that reflected the parameters used in the
data-generating process.
Eelke Folmer

-----Original Message-----
From: Bjarke Christensen [mailto:Bjarke.Christensen at sydbank.dk] 
Sent: 27 May 2008 12:41
To: r-sig-geo at stat.math.ethz.ch
Subject: Re: [R-sig-Geo] spatial probit


I recently asked a very similar question, and got the following thoughtful
reply from Roger Bivand:

For my purpose, which is to estimate probabilities for use in an inverse
probability weighted regression, I've decided to explore alternative
routes, but if you find a way to estimate a probit model with a spatially
lagged limited dependent variable as a predictor, I'd be interested in
hearing about it as well.

Bjarke Christensen

More information about the R-sig-Geo mailing list