[adegenet-forum] DAPC a-score

Jombart, Thibaut t.jombart at imperial.ac.uk
Fri Jan 24 17:17:57 CET 2014


Hello, 

saving the object should not (read: very probably does not) change anything, but results differing from one run to another would just reflect the small number of repetitions for any given solution (argument nsim).

In any case, the cross-validation procedure is a better option. See ?xvalDapc for more information.

All the best
Thibaut

________________________________________
From: adegenet-forum-bounces at lists.r-forge.r-project.org [adegenet-forum-bounces at lists.r-forge.r-project.org] on behalf of Margarita Lopez Uribe [mayalopez at gmail.com]
Sent: 24 January 2014 16:07
To: adegenet-forum at lists.r-forge.r-project.org
Subject: [adegenet-forum] DAPC a-score

Dear adegenet users,

I am trying to optimize the number of PCA to retain for a DAPC analysis.

I am using the optim.a.score() command but I am getting consistently different results if I directly call optim.a.score() or if I call it by saving it in a tmp (e.g. tmp<-optim.a.score(x)).

With the first syntax I get around 35 optimal PCA and with the second syntax I get around 20. Is this a negligible difference? Any idea why I am getting this results consistently different?

Thanks in advance!

Margarita


More information about the adegenet-forum mailing list