Hi Emmanuel,
thanks for information.
I really think it might be possible to obtain very similar results with both
packages (ape - phyreg)
However, I`m not sure that the math is the same.
"It uses branch lengths to account for recog- nised phylogeny (which makes the
errors of more closely rela
Hi Gustavo,
Grafen's method is partially implemented in ape. The function corGrafen
defines a correlation structure according to Grafen's method (see
?corClasses for all corStruct defined in ape). When used with nlme::gls
this makes possible to estimate the parameter of the branch length
trans
Hello Xavier,
Thanks for your answer.
I also might be wrong, but as far as I know, Ape has the function -
compute.brlen which can calculate branch lengths throght Grafen method but can
not do phylogenetic regression sensu (Grafen, 1989).
Grafen method uses a different approach to control phyloge
Hi Gustavo,
Thanks for that notification,
I may be wrong, but was'nt that method already implemented in the CAPER
package by David Orme (function pgls, brunch, crunch)?
If yes, then which new functionalities does phyreg bring?
Regards,
Xavier
2014-03-05 13:46 GMT+01:00 Gustavo Paterno :
> Hello
Hello all,
I just got the confirmation that the Grafen method for phylogenetic regression
(Grafen, 1989) was implemented in R !
The package “phyreg” has lots of details in help.
Best wishes for all.
Gustavo Paterno
___
R-sig-phylo mailing list - R-sig-