But couldn't you mark the trees into subsets with some kind of variable.
Something like this
Tree 1 versus a set of three trees.
Putting in a variable to specify subsets is easy, right?
Tree 1 is marked as group 1
Trees 2 thru 4 are marked as group 2
Then you can have the option of comparing distances between tree 1 and
each of the others, or comparing distances between tree 1 and the
consensus of the three others, or both.
I just think it is always most user friendly to have only one file (and
also most programmer friendly.
Now I think you've gotten a nickel's worth
On 20 Apr 1995, Joe Felsenstein wrote:
> In article <Pine.OSF.3.91.950420081847.10168A-100000 at pioneer.nevada.edu>,
> Margaret Novack <novackm at NEVADA.EDU> wrote:
> >I think the "natural" method of input is one file with all the trees for
> >which the consensus is to be computed.
> >My .02
>> Well, the question is how to do it when we have one tree, and a bunch of
> others, and want to get distance measures between them. Or pairs of trees
> (produced, say, by two different methods) and want to measure distances
> between the members of each pair. You are right that one natural method
> is a single file of trees, but we want to do something in these other cases
> Joe Felsenstein joe at genetics.washington.edu (IP No. 220.127.116.11)
> Dept. of Genetics, Univ. of Washington, Box 357360, Seattle, WA 98195-7360