Hi Grace,
Sorry for not responding earlier. The prior for the fixed effects looks
OK - it can be useful to use such priors in the case of complete separation.
Without seeing the data it is hard to diagnose why the mixing is poor.
My guess is that a phylogenetic heritability (signal) close to
So, just to recap, no .xyz email address is subscribed, and they have been
summarily banned from doing for 4 months. For close to a month, the archives
have been private. The list archives remain publicly indexed at
mail-archive.com, but they hide email addresses. It’s possible that we have
Grace,
Maybe this easiest thing to do is just try phylogenetic logistic regression,
which is appropriate for binary dependent variables regardless of the form of
the independent variables. You can use phyloglm in the phylolm package, or
binaryPGLMM in the ape package.
Cheers, Tony
On
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 Department of
Grace,
The package phytools includes a function fitPagel if that is the method
that you want to use. He has a post on his blog that discusses his
implementation so it won't be a black box for you:
http://blog.phytools.org/2014/12/r-function-for-pagels-1994-correlation.html
cheers
Heath
On
There's also corDISC in corHMM (disclosure: I'm a coauthor on the package).
It does Pagel 1994 with two or three binary characters, allows for user set
root states (I believe the canonical Pagel (1994) assumes equal freq at the
root, I don't know if phytools uses this same assumption [a lot of
Dear Grace,
the problem may derive from your specification of the priors. Usually you don’t
specify the prior for B in MCMCglmm. The problem may also be related to the
size of your dataset. Estimation of effects can be difficult with binary data,
when the dataset is small. Below is a small
I have received phishing spam through the R-sig-phylo mailing list
(pretending I had expressed interest in renting something from them) from
anya_phill...@casetotours.xyz
disguised as a reply to my comment. So that address should immediately be
removed from the list.
Joe
-
Hello,
I have characterized a few hundred bacteria from two environments and want to
know if the bacteria from one environment is more likely to show a trait than
the bacteria isolated from the other environment. So my data is binary in both
the independent and in the dependent variable:
