t the simulation is doing Brownian Motion on some
other scale, such as a log scale? If one does BM on the logs and then
looks at the original phenotype scale, you *would* expect that as the
variance among species increases, so does the mean of the species
means.
Joe
Joe Felsenstein j...@g
t noted.
Someone please educate me.
Joe
Joe Felsenstein, j...@gs.washington.edu
Dept. of Genome Sciences, Univ. of Washington
Box 355065, Seattle, WA 98195-5065 USA
___
R-sig-phylo mailing list
R-sig-phylo@r-project.org
https://stat.ethz.ch/mailman/listinfo/r-sig-phylo
Not only can't I understand the underlying model here, I can't type
correctly either: Julien Claude, not Clause. If it's any excuse,
S is next to D on my keyboard.
Joe
Joe Felsenstein, j...@gs.washington.edu
Dept. of Genome Sciences, Univ. of Washington
Box 3550
effort of doing this
for all interior nodes will be no worse than about twice
that of a single pass through the tree.
However people may prefer to use PGLS, which if
properly done should give the proper estimates for
all nodes. There is some discussion of this in
Rohlf's 2001 paper in Evol
Anyone considering this issue should consider these two papers carefully. Of
course mixing within- and between-species analyses is more difficult yet. I
hope to release an R package later this year to do the one-character analysis
(it is not too hard to put one together yourself in the
d has these
large
changes (and that fits with known Mendelian and Darwinian mechanisms).
Just complaining that the Brownian stochastic process is no good is
insufficient.
If we want to add the fossils to the calculation, then they will of
course
pressure the Brownian Motion process to change mor
o do much more
complicated models that are impossibly hard (and they
are not aware how hard they are).
Just understand, when you raise legitimate concerns, that us
model-analyzers are also used to getting a lot of these unreasonable
demands too, and may be grumpy as a result.
Joe
Joe Felsenstein
phenotypes in the fossils
that seem incompatible with the Brownian Motion assumption?
Joe
Joe Felsenstein j...@gs.washington.edu
Dept of Genome Sciences and Dept of Biology, Univ. of Washington,
Box 5065, Seattle Wa 98195-5065
[[alternative HTML version deleted
it evolved at the
> same rate of its sister species?
Good way to approach it. If you can calculate the likelihood of
trees, one way would be to not bother fitting any ancestral values:
just try different lengths for the branch that connects the fossil to
the tree, and see which one m
s 999 simulated
cases.
Joe
Joe Felsenstein, j...@gs.washington.edu
Dept. of Genome Sciences, Univ. of Washington
Box 355065, Seattle, WA 98195-5065 USA
___
R-sig-phylo mailing list
R-sig-phylo@r-project.org
https://stat.ethz.ch/mailman/listinfo/r-sig-phylo
ry unpleasant surprises.
J.F.
Joe Felsenstein j...@gs.washington.edu
Department of Genome Sciences and Department of Biology,
University of Washington, Box 355065, Seattle, WA 98195-5065 USA
[[alternative HTML version deleted]]
___
R-sig-
ample.
Eliminating that character, not telling the statistical method you did
that, and then going ahead with the analysis is a Big Mistake.
J.F.
Joe Felsenstein j...@gs.washington.edu
Department of Genome Sciences and Department of Biology,
University of Was
nable statistical
analysis.
Joe
----
Joe Felsenstein, j...@gs.washington.edu
Dept. of Genome Sciences, Univ. of Washington
Box 355065, Seattle, WA 98195-5065 USA
___
R-sig-phylo mailing list
R-sig-phylo@r-project.org
https://stat.ethz.ch/mailman/listinfo/
t (as a fan of REML) that for most tests you can do ML
for two models and compare
the likelihoods, or you can do REML for two models and compare those
likelihoods. The results will usually not be very different.
Joe
Joe Felsenstein j...@gs.washington.edu
Dept of Genome Sciences
ould work as well as ML. If we
are trying to test some hypothesis about the value of the grand mean, of course
REML is inappropriate.
An example would be simple one-way ANOVA which is actually a REML analysis, and
it does test means of the groups.
Joe
Joe Felsenstein j...@
e is evolving along the phylogeny according to a
Browian motion process. This may well not be a reasonable assumption.
Joe
Joe Felsenstein j...@gs.washington.edu
Department of Genome Sciences and Department of Biology,
University of Washington, Box 355065, S
we should be OK to use PIC regression.
I worry. Why are we to assume that current phenotypes are distributed around
optima that are based on *current* environmental variables?
Joe
---
Joe Felsenstein, j...@gs.washington.edu
Department of Genome Sciences and Department of Biology
University of
there would be no problem.
But I do think it is a Big Mistake (and apparently a frequent one),
when people measure the residual of a character regressed on
environmental variables,
then casually assume that it is undergoing Brownian Motion, when the
environmental variables may have been differe
given the true phylogeny). In that
case each column of the data matrix is drawn i.i.d. from a distribution.
Of course, even correlations of rates of evolution among neighboring
sites violates this.
Whether and how number of species comes in is trickier.
Joe
Joe Felsenstein j...@
used in the
statistical field of graphical model, and phylogeny models are graphical
models..
Joe
Joe Felsenstein j...@gs.washington.edu
Department of Genome Sciences and Department of Biology,
University of Washington, Box 355065, Seattle, WA 98195-5065 USA
[[alternativ
) and only subtract D, and call the
result the FIC, But that works only for nested hypotheses, and the main
point of the AIC is to deal with non-nested hypotheses. To make matters
worse, in my field the AIC has the reputation of too easily favoring the
most complex hypothesis, so maybe we should b
wrong too? That is what
disturbs me and makes me feel there is something I don't
understand about the AIC argument.
Joe
Joe Felsenstein j...@gs.washington.edu
Department of Genome Sciences and Department of Biology,
University of Washington, Box 355065, Seattle, WA 98195-5065 U
ed complaints that one
ought not regress present-day y's on present-day x's. Lots of people are
doing it -- and they're all wrong.
*** whining off ***
Joe
Joe Felsenstein j...@gs.washington.edu
Dept of Genome Sciences and Dept of Biology, Univ.
*not* "fairly straightforward to prove". In which case it is not
necessary to bring speciation rates or priors into it.
A reversible two-state model should be able to have its parameters
estimated on a given tree, clocklike or not.
Joe
Joe Felsenstein, j...@gs.washington
formally correct,
but it is a case where one is not being very Bayesian. For my money,
you are a Bayesian not just if you use a prior, but if you are willing to
use a controversial prior. And in this case the prior is pretty
uncontroversial.
Joe
Joe Felsenstein j...@gs.washingto
characters using the threshold model. American Naturalist 179: 145-156.
Joe
Joe Felsenstein j...@gs.washington.edu
Dept of Genome Sciences and Dept of Biology, Univ. of Washington, Box 5065,
Seattle Wa 98195-5065
___
R-sig-phylo mailin
, and not necessarily very
quickly. So it seems that an ordinary comparative
methods analysis is justified.
It is very similar to having a character such as time to
escape a predator. The fact that it does not take millions
of years to escape one attack is really irrelevant.
Joe`
Joe Felsenst
Then run any distance
method.
J.F.
Joe Felsenstein j...@gs.washington.edu
Department of Genome Sciences and Department of Biology,
University of Washington, Box 355065, Seattle, WA 98195-5065 USA
(from 1 October 2012 to 10 December 2012 on sabbatical leave at)
Department of Statis
nch lengths on
that tree. Daniel has already expressed his legitimate concerns in such a case
as to whether it takes (for example) trifurcations as if they were real rather
than an expression of our uncertainty.
J.F.
Joe Felsenstein j...@gs.washington.edu
Department of Genome Sci
r issue with parsimony methods -- the set of most parsimonious
trees may have a consensus, which may well not be a most parsimonious tree.
People who see the consensus of most parsimonious trees may not realize that
the particular tree they are looking at is not most parsimonious.
J.F.
Joe
s biology
more interesting but makes molecular change less clocklike.
Joe
Joe Felsenstein j...@gs.washington.edu
Department of Genome Sciences and Department of Biology,
University of Washington, Box 355065, Seattle, WA 98195-5065 USA
[[alternative HTML
considering different
transforms of the data (there are parameterized families of them)
and use the likelihood to test values of the transform parameters (with
appropriate correction of the likelihood by having a Jacobian term).
Joe
Joe Felsenstein j...@gs.washington.edu
Department
nuous characters.
American Journal of Human Genetics 25: 471-492.
(which can be accessed at that journal free of charge)
** bragging mode off **
Joe
Joe Felsenstein, j...@gs.washington.edu
Dept. of Genome Sciences, Univ. of Washington
Box 355065, Seattle, WA 98195-5065 USA
_
ick up some data paperwork for him for a joint project the two
of you were working on.
(Okay, the last part is "Poisson d'Avrile")
Joe Felsenstein, j...@gs.washington.edu
Dept. of Genome Sciences, Univ. of Washington
Box 355065, Seattle, WA 98195-5065 USA
___
Zharkikh had a very good paper in 1994 covering many of the distances
that had been invented up to that date, and explaining relationships between
them:
Zharkikh, A. 1994. Estimation of evolutionary
distances between nucleotide sequences.
Journal of Molecular Evolution 39: 315.329
Joe
Joe F
iterion. In a sense what I am raising is the
likelihood and modeling equivalent problem.)
Joe
Joe Felsenstein j...@gs.washington.edu
Department of Genome Sciences and Department of Biology,
University of Washington, Box 355065, Seattle, WA 98195-5065 USA
[[al
g and adding to the
tree any phenotypes for the most recent common ancestor.
Thus if we find (chimp, bonobo) to be the "cherry" we
remove them, leaving a tree such as
(macacque, (gibbon, (orang, gorilla)));
so now (orang, gorilla) is a cherry.
Joe
Joe Felsenstein, j...@gs.washing
);
so now (gorilla, human) is a cherry.
Joe
Joe Felsenstein, j...@gs.washington.edu
Dept. of Genome Sciences, Univ. of Washington
Box 355065, Seattle, WA 98195-5065 USA
___
R-sig-phylo mailing list - R-sig-phylo@r-project.org
https://stat.ethz.
in, J. 2012. A comparative method for both discrete and continuous
characters using the threshold model. American Naturalist 179: 145-156
General packages such a MCMCGlimm can (probably) be made to do
the equivalent, with some torture.
Joe
Joe Felsenstein j...@gs.washington.e
hen one species has A and
the other has A/G that you just count a match of 1/2, as for a small distance
most of the likelihood is contributed by the A possibility. For longer
branches the contributions of the two terms are more nearly equal.
Joe
Joe Felsenstein j...@gs.washingto
27;s
phytools package can call our program.
Joe
Joe Felsenstein j...@gs.washington.edu
Department of Genome Sciences and Department of Biology,
University of Washington, Box 355065, Seattle, WA 98195-5065 USA
___
R-sig-phylo mailing list
of Y may actually be the result of
optimum selection which is affected by past values of X which
we do not observe directly.
I've complained about this here in the past, to no avail,
Thomas Hansen, in a recent paper, made the same point, with
evidence too.
/* crankiness off */
J.F.
Systematic Biology 61 (3): 413-425. ISSN
1063-5157.
J.F.
Joe Felsenstein j...@gs.washington.edu
Department of Genome Sciences and Department of Biology,
University of Washington, Box 355065, Seattle, WA 98195-5065 USA
___
R-sig-phylo mailing li
ime most traits are at something
> close to adaptive equilibrium is generic to evolutionary biology.
I guess I'm not an evolutionary biologist then. If an adaptive peak is moving
around, I am unwilling to assume that the population always succeeds in staying
on top of it.
Joe
Joe Fel
e run in Windows using the Command Prompt tool, which you will
find in the Accessories menu that is found in the menu
opened by the All Programs tab in the Start Menu.
Joe
----
Joe Felsenstein, j...@gs.washington.edu
Dept. of Genome Sciences, Univ. of Washington
Box 355065, Seattle, WA 98195-506
>> R-sig-phylo mailing list - R-sig-phylo@r-project.org
> >> https://stat.ethz.ch/mailman/listinfo/r-sig-phylo
> >> Searchable archive at http://www.mail-archive.com/r-
> >> sig-ph...@r-project.org/
> >>
> >>
>
&g
think, for labeled histories.
Joe
Joe Felsenstein j...@gs.washington.edu
Department of Genome Sciences and Department of Biology,
University of Washington, Box 355065, Seattle, WA 98195-5065 USA
___
R-sig-phylo mailing list - R-sig-phylo@r-
see of course also retree.html.
Joe
Joe Felsenstein j...@gs.washington.edu
Department of Genome Sciences and Department of Biology,
University of Washington, Box 355065, Seattle, WA 98195-5065 USA
___
R-sig-phylo mailing list - R-sig-ph
two levels of menus.
Joe
Joe Felsenstein j...@gs.washington.edu
Department of Genome Sciences and Department of Biology,
University of Washington, Box 355065, Seattle, WA 98195-5065 USA
___
R-sig-phylo mailing list - R-sig-phylo@r-pro
framework.
The issue of what entities should be sampled in cross-validation
depends on how, at what level, you expect the model to depart from
multivariate normality with Brownian Motion. Diego and Ted seem to
have some such expectation but I can't see what that alternative
m
the
model is wrong so some of the changes attributed to
between-species evolution are actually within-species
sampling variation (phenotypic variance).
Joe
Joe Felsenstein j...@gs.washington.edu
Department of Genome Sciences and Department of Biology,
University of Washington, Box 3
ne major phylogeny program has very recently added code to
model this. We have a version of Dnaml for PHYLIP that can do it, and hope
to release it soon.
As George Shireff saw, it is actually easy to do, and the computations are
not any slower.
J.F.
Joe Felsenstein j...@gs.washington
four observations given that the base really is A,
and second row is the probabilities given that it really is G, etc., then
the four quantities you want at one site at one tip of the tree are the
four numbers in a column. (Which column depends on what you observed).
The rows have to add to 1; the
ou don't
want to print the resulting tree on paper, it would then be possible to
view the tree in an application such as Adobe Acrobat Reader and zoom in on
it and see the tiny branches and their labels. Making multiple plots for
one tree would probably confuse the matter.
Or is there
d line. It is the selection made by typing the character
"#".
J.F.
Joe Felsenstein j...@gs.washington.edu
Department of Genome Sciences and Department of Biology,
University of Washington, Box 355065, Seattle, WA 98195-5065 USA
[[alternative HTML version deleted]]
__
is just briefly mentioned by them, I gather
because the authors forgot about it until the last minute when writing the
review. Nevertheless it is getting increasing use.
Joe
Joe Felsenstein j...@gs.washington.edu
Department of Genome Sciences and Department of Biology,
University of
model implementations do not allow three or
more states. But the possibilty should be mentioned here.
Joe Felsenstein j...@gs.washington.edu
Department of Genome Sciences and Department of Biology,
University of Washington, Box 355065, Seattle, WA 98195-5065 USA
___
hange in log(size) or in
sqrt(size) have constant variance? There are parameterized
transformation such as y = (x^p - x^{-p})/(2p) or elsey = ( x^p
- 1) / p + 1 for which you could estimate the parameter p by ML.
Joe
Joe Felsenstein j...@gs.washington.edu
Department of Ge
s close to the true P value, one needs 4 times as many
permutations. And this need for more and more samples continues
indefinitely. There is no sudden change as one reaches a threshold number
of permutations.
But that's what you really meant, right?
Joe
---
Joe Felsenstein j...@gs
alue when its assumptions are met.
J.F.
Joe Felsenstein j...@gs.washington.edu
Department of Genome Sciences and Department of Biology,
University of Washington, Box 355065, Seattle, WA 98195-5065 USA
___
R-sig-phylo mailing list - R-sig-
Warning: You'd have to ensure that the traits for which you are comparing
rates are evolving independently, so that they do not covary in their
evolutionary changes.
I assume Dean Adams's paper involves some way of coping with this. The
issue of log-transforms that Ted raised is very important, o
I do not know offhand whether there is an R implementation, but how about
Mark Pagel's 1994 method for testing whether two 0/1 characters changing
along a ohylogeny are changing independently?
J.F.
-
j...@gs.washington.edu
Joe Felsenstein, Department of Genome Sciences and Departme
...@gs.washington.edu
Joe Felsenstein, Department of Genome Sciences and Department of Biology
Box 355065, University of Washington, Seattle, WA 98195-5065
[[alternative HTML version deleted]]
___
R-sig-phylo mailing list - R-sig-phylo@r-project.org
https
osest way I figured out is the function dist.dna from the ape
> package. But I think it does not estimate distances between groups.
>
You want to use distances between groups? But you don't want to think
about coalescents?
J.F.
Joe Felsenstein j...@gs.washington.edu
ive discussions in the book by John Wakeley. Also in the
elementary population genetics text by Nielsen and Slatkin.
J.F.
----
Joe Felsenstein j...@gs.washington.edu
Department of Genome Sciences and Department of Biology,
University of Washington, Box 355065, Seattle, WA 98195-5065
some, but the type I
error rate stays the same.
If the 100 trees are something else, such as the personal opinions of 100
of your friends, then there is no statistical justification for this.
J.F.
Joe Felsenstein j...@gs.washington.edu
Department of Genome Sciences and Department of
d in a multivariate statistics book by Morrison.
I am not sure where Paul published this, but I think he did in an appendix to a
paper of his in some multiauthor volume.
J.F.
Joe Felsenstein j...@gs.washington.edu
Department of Genome Sciences and Department of Biology,
U
continuous and another discrete.
It is implemented in my program Threshml, which should be callable from
Liam Revell's "phytools" R package.
However it does not precisely answer the question you posed, but just asks
whether the two traits evolve in a correlated fashion.
J
model. American Naturalist 179:
145-156.
which is implemented in my program Threshml which can be called from Liam
Revell's Phytools R package. It also works for multiple threshold
characters and multiple continuous characters.
Joe
----
Joe Felsenstein j...@gs.washington.edu
Departme
Liam --
Thanks, I had a "senior moment" and have been corrected: Rphylip, not
Phytools.
Joe
Joe Felsenstein j...@gs.washington.edu
Department of Genome Sciences and Department of Biology,
University of Washington, Box 355065, Seattle, WA 98195-5065 USA
[[altern
ach.
>
> 1) Diversitree package BiSSE
> 2) Caper package using MacroCAIC
>
> Any suggestions would be greatly appreciated-
And what do these methods assume about how that "discrete binary predictor"
evolves along the tree?
Joe
-
j...@gs.washington.edu
Joe Felsenstein,
ow the environments changed along the tree. Not much
effect so far so you can expect more shots across the bow.
Joe
----
Joe Felsenstein j...@gs.washington.edu
Department of Genome Sciences and Department of Biology,
University of Washington, Box 355065, Seattle, WA 98195-5065 USA
[[alter
this with my own C programs, but it can be done in R too.
But what does it mean to be "scaled similarly" ?
Joe
Joe Felsenstein j...@gs.washington.edu
Department of Genome Sciences and Department of Biology,
University of Washington, Box 355065, Seattle, WA 98195-50
about any covariation between traits? Not knowing that is
a positive objective of your study?
The vectors of the individual standardized contrasts are independent
multivariate quantities. The contrasts for different traits covary, but
the multivariate vectors are i.i.d.
J.F.
Joe Felse
.
Application deadline March 1, 2017.
Application forms and details here:
http://tinyurl.com/EQG2017
Web page:
http://depts.washington.edu/fhl/studentSummer2017.html#SumB-genetics
Instructors:
Dr. Joe Felsenstein
Department of Genome Sciences
University of Washington, Seattle
j...@gs.washington.edu
Dr
problem that one either
estimates those separately for each pair of sequences, or jointly
estimates them from the whole dataset, without using a tree in the
process.
J.F.
Joe Felsenstein j...@gs.washington.edu
Department of Genome Sciences and Department of Biology,
University of Washi
distances after the
distance matrix is computed.
J.F.
Joe Felsenstein j...@gs.washington.edu
Department of Genome Sciences and Department of Biology,
University of Washington, Box 355065, Seattle, WA 98195-5065 USA
___
R-sig-phylo mailing li
pressibility should be much lower.
Joe
Joe Felsenstein j...@gs.washington.edu
Department of Genome Sciences and Department of Biology,
University of Washington, Box 355065, Seattle, WA 98195-5065 USA
___
R-sig-phylo mailing list - R-si
.
Is described as working from image files, and it is available for
various operating systems.
Does it solve the problem?
Joe
Joe Felsenstein [EMAIL PROTECTED]
Department of Genome Sciences and Department of Biology,
University of Washington, Box 3550
ell me please.
This is an illegitimate answer, but ... have APE call PHYLIP's
program Drawgram. It can do a Circular tree, and by playing
around with the options for use or nonuse of branch lengths,
and the placement of ancestral nodes, you can get the circular
tournament-format tree.
x27;s not the traditional analysis for allometry
but it ought to become the new tradition.
Joe
Joe Felsenstein [EMAIL PROTECTED]
Department of Genome Sciences and Department of Biology,
University of Washington, Box 355065, Seattle, WA 98195-5065 USA
__
covariances.
My 2008 paper in American Naturalist is another paper that is relevant.
Joe
Joe Felsenstein [EMAIL PROTECTED]
Department of Genome Sciences and Department of Biology,
University of Washington, Box 355065, Seattle, WA 98195-5065 USA
__
shares a common ancestor with "Athene_noctua"
> and parse out the names of the tips in that clade. Is the question
> clear? I hope so. Any advice would be much appreciated!
It's easy. Don't do anything! Everything in the tree already shares
a common ancestor with &
s the entire
matrix, that might explain why the bootstrap support is so much lower.
Those are two interesting bootstrap methods ... but I would say both
are completely wrong. Isn't it more correct to bootstrap the original
data from which the distances are derived, then compute distances
for
Emmanuel Paradis wrote --:
Joe Felsenstein a écrit :
Monica Poelchaum wrote --
The bootstrap values this function returns are MUCH lower than
what is
returned in another program, "Past". I'm wondering at what level the
bootstrap algorithm randomizes the samples, since thi
presume this alternative is a good approach)
Some people answered questioning this as a good resampling
procedure. Just to be the Voice Of Orthodoxy, let me add ...
Whatever its merits, it isn't the same as parametric
bootstrapping. That involves simulating data along a tree,
not shuffl
and what needs to be done when more elaborate
analyses than contrasts from species means are to be done.
I'll be giving a short (13-15 minute) talk on this at this
year's Evolution meetings.
Joe Felsenstein j...@gs.washington.edu
Department of Genome Sciences
ve natural limits at zero. Then the regression
might just be a simple linear one.
Joe
Joe Felsenstein j...@gs.washington.edu
Department of Genome Sciences and Department of Biology,
University of Washington, Box 355065, Seattle, WA 98195-5065 USA
__
covariance matrices. When you have both phylogenetic covariances
and within-species phenotypic covariances, these may have different
axes, of course.
J.F.
----
Joe Felsenstein j...@gs.washington.edu
Department of Genome Sciences and Department of Biology,
University of Washington, Box
es, but as posters here have
pointed out, the between-species ("phylogenetic") covariances can
be used to calculate the RMA slope. In the next major release we will
explicitly calculate RMA stuff too.
Joe Felsenstein joe (at) gs.washington.edu
Department of Genome Sciences an
washington.edu/papers/spectrum/
As for integrating this with between-species comparative methods, I am
pretty sure that can be done but it will be even messier.
For combinations of gene-flow-thinking and tree-thinking see Rasmus Nielsen
and Jody Hey's program IMa, which is still in its earl
f how to classify is less important anyway.
J.F.
Joe Felsenstein, j...@gs.washington.edu
Dept. of Genome Sciences, Univ. of Washington
Box 355065, Seattle, WA 98195-5065 USA
___
R-sig-phylo mailing list
R-sig-phylo@r-project.org
https://stat.ethz.ch/ma
group of interest.
If one simply measures the same character in different species,
is that not good enough? My curiousity is because I suspect that
if Stephane had coded his characters 0/1 before trying to analyze
them, they would then be accepted as "made of homology hypotheses".
many systematists have
been strong -- they are really outraged, and figure that this is just
some arbitrary opinion of mine, which they are (barely) willing to tolerate.
I suppose the matter will become one of open discussion some day.
Anyway, back to R.
J.F.
Joe Felsenstein
shing phylogenetic
comparative methods.
(It's taking a lot longer to wean physiologists of two-species
comparisons, but I have not had to do any of that advocacy -- others
have borne the brunt.)
J.F.
Joe Felsenstein j...@gs.washington.edu
Department of Genome Sciences and Depar
t it'll be straightforward to simulate (continous) traits under
the available models of evolution.
Keep in mind that one can't always assume that the variances (say
of Brownian motion) are equal in all characters, so the correlations
aren't quite enough.
Joe
Joe Felsenstein, j..
root of the covariances among characters, and multiply
the values at each node by that.)
This way you don't need a nodes x nodes covariance matrix.
Joe
Joe Felsenstein j...@gs.washington.edu
Department of Genome Sciences and Department of Biology,
University of Washington, B
provable that this will be identical (getting
the same P values and estimates) to a REML ("reduced" or "restricted" maximum
likelihood) on the full covarying Brownian motion model. If PGLS gets
identical results to that, then that proves the identity to contrasts analysis.
Joe
on2009 meeting (for 13 minutes,
anyway) and we hope to have more methods available soon.
Obviously I prefer statistically-based methods such as likelihood (or Bayesian)
ones instead of parsimony, but I won't bore people with that.
Joe
Joe Felsenstein j...@gs.washington.edu
Departmen
ares fit of the distances.
Joe
---
Joe Felsenstein, j...@gs.washington.edu
Department of Genome Sciences and Department of Biology
University of Washington
Box 355065
Seattle WA 98195-5065
___
R-sig-phylo mailing list
R-sig-phylo@r-project.org
https://stat.
1 - 100 of 137 matches
Mail list logo
