[R-sig-eco] residuals in RDA, and test for spatial autocorrelation

2014-04-01 Thread Pinney, Tracy A
Hello List, I have two questions. 1.)How do I generate a matrix of residuals in RDA (I am using rda() in vegan)? 2.)How can I test for spatial dependence (spatial autocorrelation) in the response matrix of a multivariate analysis (RDA in my case)? I have read that you can look at

Re: [R-sig-eco] residuals in RDA, and test for spatial autocorrelation

2014-04-01 Thread amommendes
Hi Tracy, You can try this: require (vegan) data(mite) data(mite.env) data(mite.xy) mod- rda (mite~., data=mite.env) res-residuals(mod) require (spacemakeR) vario-variogmultiv(res, mite.xy) plot(vario$d, vario$var, ty='b',pch=20,xlab=Distance, ylab=C(distance)) But, I'm not sure. Maybe to

Re: [R-sig-eco] residuals in RDA, and test for spatial autocorrelation

2014-04-01 Thread Jari Oksanen
Dear Tracy, You can see if Helene Wagner's mso() in function in vegan satisfies your needs for analysing spatial dependence. Reference and further description in ?mso. Cheers, Jari Oksanen On 02/04/2014, at 00:12 AM, Pinney, Tracy A wrote: Hello List, I have two questions. 1.)How