[R-sig-Geo] SDM and SDEM models with single aliased feature and lag

Roger Bivand Roger@B|v@nd @end|ng |rom nhh@no
Mon May 2 20:15:27 CEST 2022


On Mon, 2 May 2022, Ari via R-sig-Geo wrote:

> Hello everyone,

> I am working with a SDM and SDEM models from the spatialreg package in a 
> geographical pattern containing some islands. While simulating the 
> regressions I encountered aliased variables, but to my surprise it only 
> reported a single feature (and its lag). In previous cases when I 
> encountered aliased variables it always was between a pair of features, 
> as it usually meant that they both had perfect collinearity. I know how 
> to adapt the data to deal with perfect colinearity, but I am not sure on 
> how to proceed if only one feature (and its lag) are reported as 
> aliased? Could it be caused by the "isolated nature" of the dataset 
> (despite working with the zero.policy=TRUE argument)?

Could you please provide reproducible examples of these cases, for example 
using the Boston data set ?spData::boston ? Please also report whether the 
variables are categorical (in R factors), or not. Without a minimal worked 
example, it isn't obvious what the problem might be. By feature, do you 
mean a single geometry/observation, or several observations on one 
subgraph?

Roger

>
> Thanks the help.
> Best regards,
> Ari
> _______________________________________________
> R-sig-Geo mailing list
> R-sig-Geo using r-project.org
> https://eur02.safelinks.protection.outlook.com/?url=https%3A%2F%2Fstat.ethz.ch%2Fmailman%2Flistinfo%2Fr-sig-geo&data=05%7C01%7CRoger.Bivand%40nhh.no%7C1ac78a5aa03a4072452808da2c5af02c%7C33a15b2f849941998d56f20b5aa91af2%7C0%7C0%7C637871066516191175%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=rfyCTdzM517VkdaG39ViNhdbgpSWRqsZD4MBmerZ6TU%3D&reserved=0
>

-- 
Roger Bivand
Emeritus Professor
Department of Economics, Norwegian School of Economics,
Postboks 3490 Ytre Sandviken, 5045 Bergen, Norway.
e-mail: Roger.Bivand using nhh.no
https://orcid.org/0000-0003-2392-6140
https://scholar.google.no/citations?user=AWeghB0AAAAJ&hl=en



More information about the R-sig-Geo mailing list