[phylocom-user] Different MNTD between phylocom "comtrait (-x 3)" vs. "picante 'ses.mntd"?
Bob Muscarella
bob.muscarella at gmail.com
Fri Dec 13 11:12:50 PST 2013
Hello -
I've been using the 'comtrait' function in Phylocom to take advantage
of it's speed compared to the 'ses.mpd' and 'ses.mntd' functions in
picante. The two calculations produce the same results for the MPD
metric but I am getting different results for the MNTD metric.
I am working with wood density values (continuous, from 0-1) and the
absolute value of trait differences in my sample are small (nearest
neighbors are generally ~.01). The 'ses.mntd' function gives
reasonable results ('mntd.obs' ~0.01). When I use the comtrait
function to calculate MNTD (with: -x 3), however, I get a larger value
('Metric' ~ 0.5). This doesn't seem reasonable unless I am missing
something.
I tried a small simulation by changing the trait values in the example
data file 'trait' that is distributed with Phylocom. I changed the
original (continuous) trait values such that they were all preceded by
"1." (e.g., trait value 2 became 1.2). The resulting output also
gives higher values than I expected for MNTD (i.e., the minimum
'Metric' is >1 where it seems like it should be 0.1).
Besides the observed MNTD values being different, the ranking of
communities and SES metrics are also different. Has anyone
encountered similar issues? Perhaps I'm missing something in how MNTD
is calculated in Phylocom?
Thanks for any input and please let me know if I can clarify my question.
Bob
More information about the phylocom-user
mailing list