[Phylobase-devl] edge data?

[Ben Bolker]

Brian O'Meara wrote:
> On Mar 24, 2008, at 3:50 PM, Peter Cowan wrote:
>> Hi all,
>>
>> I've been trying to write up an example function for the vignette.  I
>> thought that I would do a brownian trait simulator, because they are
>> so popular.  I was basing the S4 version on the one by David Ackerly
>> in the picante package [^1] called evolve.trait().  This nifty
>> function keeps track of the trait value for each unit of branch
>> length.  That is, if a branch has length 5, there will be five values
>> along the branch.  This data seems most logically stored as edge data.
>>
>> As of now the phylo4d class does not have an edge data slot.
>> Furthermore, the commented out version suggests a data frame, which
>> would not work with the variable length data described above, trees
>> with differing branch lengths.
>>
>> I forget why we decided against edge data, is this too esoteric of a
>> use case?  Any suggestions of a better way to handle the above case?
> 

    This is a bit of a leap, but we could add singleton nodes along the
branch, and there would then be a one-to-one correspondence between
nodes and the edges (this is similar to the way OUCH handles shifts
in ancestral environments (? can't think of the right word)).
One problem is that at least the plotting code in ape (which we're
still using) doesn't like such trees -- but I do think this is a useful
generalization for the future (and ape::plot.phylo doesn't crash,
it just stops with a warning).

   Ben

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 252 bytes
Desc: OpenPGP digital signature
Url : http://lists.r-forge.r-project.org/pipermail/phylobase-devl/attachments/20080324/ffed8c2d/attachment.pgp