Hi Carl, Annemarie-
While it is possible in principle that Annemarie's results reflect true ML
estimate of lambda = 1, I think the practical reason this is occurring is that
the default bounds on lambda in fitContinuous are 1e-7 and 1. Because
optimization in fitContinuous uses a bounded BFGS
Hi Annemarie,
The only thing I would add to Carl's comment is that the theoretical
limit of lambda is not 1.0, but can be found (for an ultrametric tree)
by computing:
> C<-vcv.phylo(tree)
> maxLambda<-max(C)/max(C[upper.tri(C)])
You can then change the boundary condition for fitContinuous()
Hi Annemarie,
No problem, tried to give some answers below.
On Mon, May 23, 2011 at 8:05 AM, Annemarie Verkerk wrote:
> Dear Carl, Liam, and others,
>
> thanks for your explanation of what went wrong in the fitContinuous
> calculations. I set beta to a large number (100) in order to st
For some of your general questions, I'd suggest checking these two papers, the
appendix in the latter:
Garland, T., Jr., A. W. Dickerman, C. M. Janis, and J. A. Jones. 1993.
Phylogenetic analysis of covariance by computer simulation. Systematic Biology
42:265-292.
Lavin, S. R., W. H. Karasov,
Dear all,
I have one factor and several covariates and I would like to know
which of these explanatory variables are more likely to explain a
response variable in a comparative analysis. The factor is a dummy
variable (0-1).
At first, my strategy would be to use contrasts of all variables
Hi Emmanuel
I've done as you said but the positions of the branch support values in the
rooted tree remain a problem eg
library(ape)
## original tree
tree <- read.tree(text =
"(1:0.013072,2:0.005026,3:0.005003,(4:0.007930,5:0.002958)0.84:0.006529,6:0.013338,7:0.004297)1.00:0.055567,8:0.0158
Dear Carl, Liam, and others,
thanks for your explanation of what went wrong in the fitContinuous
calculations. I set beta to a large number (100) in order to
stop it from reading the maximum value. Then, I got exactly the same
results for lambda with the non-multiplied and the multipli