etwork for Experimental Research on Evolution
http://nere.bio.uci.edu/
(A University of California Multicampus Research Project)
Original message
Date: Fri, 13 Mar 2009 09:36:38 +
From: "Tom Oliver"
Subject: Re: [R-sig-phylo] compar.gee- Generalized
esti
Dear Emmanuel & Helplist,
Thanks for your response. So would a valid approach be to test for phylogenetic
autocorrelation beforehand (e.g. Using Gittleman and Kot's method implemented
by the gearymoran function in R package ade4) and if there is significant
autocorrelation, then to use the GEE
Actually, I have already observed the same thing with simulated data.
When you shuffle your data (independently for both vectors), you
cancel the (potential) relationship that exists between the variables,
but you also remove the covariance among species. So using GEEs is no
more valid beca
Hi Emmanuel,
Here is the data summarised if this is what you mean? If the independent
resampling did not completely cancel the actual relationship between the
variables wouldn't that make the non-phylogenetic ANOVAs significant too?
If it helps to isolate what is going on, I seem to be getting
Can you look at the summary of the distribution of these variables? It
could be that the independent resampling does not completely cancel
the actual relationship they seem to have.
EP
Tom Oliver a écrit :
Hi Emmanuel,
Sorry the script should be:
for (i in (1:30)){
resp<-data[,5]
expla
Hi Emmanuel,
Sorry the script should be:
for (i in (1:30)){
resp<-data[,5]
explanatory<-data[,36]
resp<-sample(resp,replace=T)
explanatory<-sample(explanatory,replace=T)
names(resp)<-names(explanatory)<-data[,2]
data2<-data.frame(resp,explanatory)
data2<-na.omit(data2)
plot(resp~explanatory)
pri
Tom Oliver a écrit :
Hi Emmanuel,
Here is the script I used for randomizing independently on both
vectors. Similar results occur for sampling without replacement.
for (i in (1:30)){
resp<-data[,5]
explanatory<-data[,36]
resp<-sample(slope,replace=T)
Is that normal that you sampled 'slope'
Hi Emmanuel,
Here is the script I used for randomizing independently on both
vectors. Similar results occur for sampling without replacement.
for (i in (1:30)){
resp<-data[,5]
explanatory<-data[,36]
resp<-sample(slope,replace=T)
explanatory<-sample(explanatory,replace=T)
names(resp)<-names(exp
Tom Oliver a écrit :
Hi Emmanuel,
Thanks for your post. I can accept that sometimes normal OLS
regressions will give non-significant correlations yet after
accounting for covariance among species using phylogenetic methods
then the traits may show significant correlations. However, I
Hi Emmanuel,
Thanks for your post. I can accept that sometimes normal OLS regressions will
give non-significant correlations yet after accounting for covariance among
species using phylogenetic methods then the traits may show significant
correlations. However, I am confused how I can randomly
Hi Tom,
We had a discussion related to this topic last year. Here's my main comment:
https://stat.ethz.ch/pipermail/r-sig-phylo/2008-April/70.html
You may have a look also at other messages in the same thread, of course.
HTH
EP
Tom Oliver a écrit :
Hello Helplist,
I was wondering i
Hello Helplist,
I was wondering if anyone could shed some light on this problem..
I have been using the compar.gee package (Package ape version 2.2) for a
comparative analaysis using a binary categorical variable.
My phylogenetic tree has 42 tips with branch lengths set to 1 and the response
v
12 matches
Mail list logo
