[adegenet-forum] global rtest/ choice of spca axes of significance.

Jombart, Thibaut t.jombart at imperial.ac.uk
Mon May 24 18:13:12 CEST 2010


Hello,


> The global.rtest gives a test of whether there is significant global ( 
> ie cross-landscape) structure between sampled locations, correct?

Yes.
>
> In the spca tutorial example a single pca lambda1  clealy stands out .
>
> How does one decide where more than one axis is involved, should it 
> stand out ( ie have substantially greater values) on both variance and 
> SA axes?  or simply be an outlier on the variance axis?

Only variance = not a spatial structure. So you need both variance and autocorrelation.
>
> Goes the global.rtest test just the first spca axis , or for any 
> significant global structure ( that may be across the first few axes)
It tests for any global structure.
>
> If it does the latter is what objective approach can be taken to decide 
> how many axes should be considered as contributing to the structure?

See if several eigenvalues stand out. There are parts dedicated to this topic in the paper describing sPCA.

Cheers

Thibaut

>
> cheers
>
> Nevil Amos
> School of Biological Sciences
> Monash University
>
>
> _______________________________________________
> adegenet-forum mailing list
> adegenet-forum at lists.r-forge.r-project.org
> https://lists.r-forge.r-project.org/cgi-bin/mailman/listinfo/adegenet-forum


-- 
######################################
Dr Thibaut JOMBART
MRC Centre for Outbreak Analysis and Modelling
Department of Infectious Disease Epidemiology
Imperial College - Faculty of Medicine
St Mary’s Campus
Norfolk Place
London W2 1PG
United Kingdom
Tel. : 0044 (0)20 7594 3658
t.jombart at imperial.ac.uk
http://sites.google.com/site/thibautjombart/
http://adegenet.r-forge.r-project.org/


More information about the adegenet-forum mailing list