[R-sig-eco] Problem with Capscale and Phylobetadiversity distance matrix

[Alexandre F. Souza]

Hello everyone,

I have a problem trying to use vegan's capscale() on a phylobetadiversity
distance matrix, reproduced below using the example:

library(psych)  # For the PCA test involved

# Opens a full matrix of 245 x 245 phylobetadiversity distances
bnti.ses = read.csv(file.choose(), header = T, row.names=1);
dim(bnti.ses); # this gives the dimensions
bnti.ses[1:5,1:5]; # this gives a look at the first 5 rows and columns

# Test PCA scores
test = data.frame(replicate(10,sample(0:100,245,rep=TRUE)))
test
pca = principal(teste, nfactors = 3, covar = FALSE, scores = TRUE)
meta.pca.scores = pca$scores
rownames(meta.pca.scores)=rownames(bnti.ses)


## dbRDA

# betaNTI standard effect size


bnti.ses = bnti.ses + abs(min(bnti.ses));

bnti.ses = bnti.ses/max(bnti.ses);

range(bnti.ses);


identical(rownames(bnti.ses),colnames(bnti.ses)); # must be TRUE to proceed

identical(rownames(bnti.ses),rownames(meta.pca.scores)); # must be TRUE to
proceed


# Here is where the error occurs
bnti.ses.dbrda.int = capscale(as.dist(bnti.ses) ~ 1,data=meta.pca.scores);

bnti.ses.dbrda.full = capscale(as.dist(bnti.ses) ~ .,data=meta.pca.scores);

bnti.ses.select.rsq = ordiR2step(bnti.ses.dbrda.int, scope =
formula(bnti.ses.dbrda.full),direction = "both",pstep=10000, perm.max =
100000); bnti.ses.select.rsq;

I get the message:

Error in formula.default(object, env = baseenv()) : invalid formula


I tried the same code using the varespec and varechem example files present
in capscale's help and it worked. I checked the dimensions of the
phylobetadiversity matrix and they are ok (245 x 245). I tried to omit the
row names but the problem persists.

For me the very last test that should have worked was

bnti.ses = data.frame(replicate(245,sample(0:100,245,rep=TRUE)))

rownames(bnti.ses)=rownames(meta.pca.scores)colnames(bnti.ses)=rownames(bnti.ses)


and then working this artificial bnti.ses through the subsequent
standardization and capscale steps. However, I got the same error mesage
with this example.

Am I missing something, or is there a workaround?

Thanks,

Alexandre
-- 
Dr. Alexandre F. Souza
Professor Ajunto II
Universidade Federal do Rio Grande do Norte
CB, Departamento de Ecologia
Campus Universitário - Lagoa Nova
59072-970 - Natal, RN - Brasil
lattes: lattes.cnpq.br/7844758818522706
http://www.docente.ufrn.br/alexsouza



-- 
Dr. Alexandre F. Souza
Professor Ajunto II
Universidade Federal do Rio Grande do Norte
CB, Departamento de Ecologia
Campus Universitário - Lagoa Nova
59072-970 - Natal, RN - Brasil
lattes: lattes.cnpq.br/7844758818522706
http://www.docente.ufrn.br/alexsouza

	[[alternative HTML version deleted]]