[R-sig-Geo] Comparing Estimating Spatial Simultaneous Autoregressive Models Using "spautolm" and "lagsarlm" functions

[Roger Bivand]

On Wed, 4 Apr 2012, Roosbeh Nowrouzian wrote:

> Hi dear List members:
>
> I estimated a Spatial Simultaneous Auto-regressive Model using "spautolm" [
> family="SAR"] and "lagsarlm" [type=lag].The estimated coefficients are
> approximately equal but the Lambda estimated are somehow different.

These do not fit the same model. spautolm(..., family="SAR") and 
errorsarlm() do, as the help page and examples make clear.

Roger

>
> In SAR using "spautolm":
>
> Lambda: 0.023787 *LR test value: 2.8569* *p-value: 0.090982*
>
>
> In SAR using "lagsarlm":
>
> Rho: 0.019479, *LR test value: 1.964*, *p-value: 0.16108*
> *
> *
> Setting significant level at 90% there would be significant Lambda for SAR
> using "spautolm"while insignificant if using "lagsarlm".
> I am wondering why it happens and which one is more accurate to report.
>
> Thanks for your help
> Roosbeh Nowrouzian
> PhD Student
> Department of Civil Engineering
> University of Florida
>
> 	[[alternative HTML version deleted]]
>
> _______________________________________________
> R-sig-Geo mailing list
> R-sig-Geo at r-project.org
> https://stat.ethz.ch/mailman/listinfo/r-sig-geo
>

-- 
Roger Bivand
Department of Economics, NHH Norwegian School of Economics,
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