Group-wise error variance and autocorrelation
Hi Tim, The nlme package allows to model the variance structure of the within-group errors. It also allows to include spatial correlation structures, in principle it is possible to define a grouping factor too, although I've never used it. Santiago Beguer?a Portugu?s
Date: Wed, 18 Jul 2007 10:09:34 -0500 From: "Tim Keitt" <tkeitt at gmail.com> Subject: [R-sig-Geo] Group-wise error variance and autocorrelation To: "R geostat list" <r-sig-geo at stat.math.ethz.ch> Message-ID: <6262c54c0707180809v4b7a1318o2547c793f5e45e48 at mail.gmail.com> Content-Type: text/plain; charset=ISO-8859-1; format=flowed I'm trying to specify a model where the fixed effects are dep ~ group + group:indep and the error variance and spatial autocorrelation vary by group. I've tried lme but AFAICT it only supports heteroscedasticity and autocorrelation within groups. Any pointers? THK
--------------------------------- _PLEASE, TAKE NOTE OF MY NEW EMAIL_ _ADDRESS AND CONTACT DETAILS!_ Santiago Beguer?a Portugu?s (PhD) Researcher Estaci?n Experimental de Aula Dei - CSIC (Aula Dei Experimental Station, Spanish National Research Council) Campus de Aula Dei, Avda. Monta?ana, 1005 50059 - Zaragoza (Spain) P.O. Box 202, 50080 - Zaragoza (Spain) T (+34) 976 716 107 T (+34) 976 716 100 / ext. 2356 F (+34) 976 716 145 http://www.santiagobegueria.net -------------- next part -------------- A non-text attachment was scrubbed... Name: sbegueria.vcf Type: text/x-vcard Size: 435 bytes Desc: not available URL: <https://stat.ethz.ch/pipermail/r-sig-geo/attachments/20070719/7cc52ba9/attachment.vcf>