Sorry, I didn't see that this had already been addressed.
--
Liam J. Revell
University of Massachusetts Boston
web: http://faculty.umb.edu/liam.revell/
email: liam.rev...@umb.edu
blog: http://phytools.blogspot.com
On 5/17/2011 12:39 PM, David Bapst wrote:
Alanna-
It's because when multi2di() resolves the polytomies, it puts in zero-length
edges. This causes the phylogenetic variance-covariance matrix calculated
from your tree to be singular.
One solution would be to add a small arbitrary constant to your zero-length
edges, although I would caution you to try several different values and see
how your results differ.
-Dave Bapst, UChicago
On Tue, May 17, 2011 at 10:23 AM, Alanna Maltbyalanna.mal...@ioz.ac.ukwrote:
Dear all
I've been trying to use Maximum Likelihood to estimate ancestral
characters using ace in APE, but I get an error:
?Error in solve.default(out$hessian) :
Lapack routine dgesv: system is exactly singular?
My code looks like this:
tree-read.nexus(tree.nex)
data-read.csv(data.csv)
BW-data.frame(data[,c(2,3)])
rownames(BW)-data[,2]
na.omit(BW)-BW
name.check(tree,BW)-overlap
drop.tip(tree,overlap$Tree.not.data)-tree2
tree3-multi2di(tree2)
BW2-BW[tree3$tip.label,]
BW3-data.frame(BW2[,2])
as.matrix(BW3)-BWM
as.vector(BWM)-BWV
names(BWV)-BW2[,1]
asrBW-ace(BWV, tree3, type=continuous)
asrBW$ace
which is possibly not super-elegant, but works when I use pic instead of
ML.
I notice that someone brought up the same problem last year but there
were no solutions. Any ideas?
Many thanks,
Alanna
-
Alanna Maltby
PhD Student - Evolution of Echolocation in Bats
University College London and Institute of Zoology
Tel: +44 (0) 20 7449 6322
Website: www.zsl.org/alannamaltbyhttp://www.zsl.org/alannamaltby
The Zoological Society of London is incorporated by Royal Charter
Principal Office England. Company Number RC000749
Registered address:
Regent's Park, London, England NW1 4RY
Registered Charity in England and Wales no. 208728
_
This e-mail has been sent in confidence to the named a...{{dropped:21}}
___
R-sig-phylo mailing list
R-sig-phylo@r-project.org
https://stat.ethz.ch/mailman/listinfo/r-sig-phylo
___
R-sig-phylo mailing list
R-sig-phylo@r-project.org
https://stat.ethz.ch/mailman/listinfo/r-sig-phylo
