I only just read some of the posts regarding tree-comparison issues.
Somehow much of the preceding thread did not appear in
my newreader spool (e.g., Joe's post). As I am in Canada (vicariously
reading from VA - no Usnet here) I am curious as to whether people
are specifying only "USA" in their distributions.
In response, an alternative way to examine whether trees in a file
fit some user-specified tree is to use BPA to code the user-tree
as a large multistate character (by way of additive binary coding)
e.g.,
((a b) (c d)) represented as
a 1000101
b 0100101
c 0010011
d 0001011
And then examing fits and stats of this data set to all targetted trees.
That is, this data set will fit its target tree with 6 steps and a CI
and RI of 1.00. Ostensibly, these 3 criteria are both necessary
and sufficient to identify trees matching ((a b)(c d)) in any arrangement
of the taxa [e.g. ((d c) (b a))].
The "xs" command in Hennig86 could easily be manipulated in batch mode
to examine for this on large numbers of trees.
With some considerable experience in writing software to make Hennig86
do things not considered by Farris, I'd be happy to give it a shot
if there's a demand for it and place it on the zoo.toronto.edu ftp site.
Your humble servant,
Mark
--
Mark E. Siddall "I don't mind a parasite...
mes at vims.edu I object to a cut-rate one"
Virginia Inst. Marine Sci. - Rick
Gloucester Point, VA, 23062
