Dear Colleagues,

We wish to alert you to a new article, introducing phylogenetically aligned 
component analysis, currently in early release form:

Collyer, M.L. and D.C Adams. 2020 (in press). Phylogenetically aligned 
component analysis.  Methods in Ecology and Evolution.

If you do not have access to MEE, we have a link to the accepted article and 
supporting information here 
<https://www.researchgate.net/publication/344900621_Phylogenetically_Aligned_Component_Analysis>.
  (Or paste 
https://www.researchgate.net/publication/344900621_Phylogenetically_Aligned_Component_Analysis
 
<https://www.researchgate.net/publication/344900621_Phylogenetically_Aligned_Component_Analysis>
 in your web browser.)

Phylogenetically aligned component analysis (PACA) is an ordination method 
similar to principal component analysis (PCA), but aligns data to maximum 
phylogenetic signal rather than maximum variance.  In conjunction with PCA (or 
phylogenetic PCA), PACA provides a tool for visualizing phylogenetic signal in 
multivariate data, even if other signals (e.g., allometry, ontogenetic, 
ecological) might obscure phylogenetic signal in PCA.

We make PACA available to GM users in the RRPP and geomorph R packages, with 
RRPP::ordinate and geomorph::gm.prcomp functions.  These functions allow users 
to align data to either principal or phylogenetically aligned vectors, 
optionally transform data by phylogenetic covariances, project ancestral states 
and phylogenetic tree edges into a plot, and evaluate the amount of covariance 
between data and phylogeny, by vector.  Additionally, the physignal function in 
geomorph provides $PACA output, along with the amount of cumulative 
phylogenetic signal, by vector, which can inform if phylogenetic signal is 
especially strong in certain data dimensions.

We recommend installing the latest versions of RRPP and geomorph via Github; 
i.e.,

devtools::install_github(“mlcollyer/RRPP”, build_vignettes = TRUE)
devtools::install_github(“geomorphR/geomorph”, ref = “Stable”, build_vignettes 
= TRUE)

Happy computing!

Mike Collyer and Dean Adams

-- 
You received this message because you are subscribed to the Google Groups 
"Morphmet" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To view this discussion on the web visit 
https://groups.google.com/d/msgid/morphmet2/AC2158F3-4CD2-4416-BEE0-ABECD864CE41%40gmail.com.

Reply via email to