[R] Help on phylogenetic ANOVA functions

Guido Buezas buezas.guido at gmail.com
Tue May 12 21:15:11 CEST 2015

Hi everyone:

I'm trying to do a phylogenetic ANOVA analysis, but i'm stuck with both
possibilities: phylANOVA from phytools package, and aov.phylo from geiger

Here is what i have

Input data:
CSV file with columns:
Spp (categorical, for taxonomic species,, 29 species)
Fam (categorical, for taxonomic families)
IF (continuous, length ratio data ranging from 1 to 3 approx, 91 cases)

Phylogenetic tree. (Newick format, with distances, 29 species). I'm sure
that all species are both in phylo tree and in data.

My problems are:

phylANOVA didn't calculate post-hoc tests. All i got was an F and p, but
all results from post-hocs are NA.

aov.phylo won't stop saying:

'formula' must be of the form 'dat~group', where 'group' is a named factor
> vector and 'dat' is a data matrix or named vector"

When were made this way:

dat <- as.vector(data$IF)
group <- as.vector(data$Spp)

I've googled and tried several ways to name the damned vectors, but it
keeps saying that. Like: names(OBJECT)<-c("name1", "name2",....)

Eventually, i could run it, i don't know how (and that bothers me). BUT,
something bad happened:

The following tips were not found in 'phy' and were dropped from 'data':

And viceversa for 'data' and 'phy'. So, analysis is restricted to F and p
values. When i try to make the juicy post-hoc tests (even when i don't know
if it will work), it fails saying that Tukey HSD is not appliable to lm

Hope you understood me and my english, and can help me!

Thank you very much in advance

Guido, from Mar del Plata
Lic. Guido Buezas
Laboratorio de Morfología Funcional y Comportamiento

	[[alternative HTML version deleted]]

More information about the R-help mailing list