[Phylobase-devl] a graphics challenge ...
bcomeara at nescent.org
Sat Mar 21 03:13:23 CET 2009
C++ is a hurdle (R plays better with C), but phylobase already uses
RCpp to talk between R and C++. I wouldn't advise using TreeLib,
though, because it's not really distributed in any public way -- it's
in some of Rod Page's open source programs, and he gives copies and
example files upon request (at least he did to me), but there's no
public repository or even static website for it. It has accessors for
branch lengths and tree traversal, but that is already written
directly in R for phylobase.
Incorporating Bio++ is interesting -- that project has really
developed well in the past few years. I think it would be an extensive
project, and one that, once the investment was made to get the tree
part working, it would be logical to get the other parts (likelihood
calculations, bootstrapping, etc.) incorporated, as the hard parts
would be done. I personally think it would take more time than people
have been able to put into phylobase recently, and perhaps would be a
bit chewy for a summer student, though the potential benefits could be
large. For me, a bigger priority would be stabilizing the current
phylobase functions and getting it ready to go to CRAN, and I don't
think adding Bio++ would accelerate this. But that's just one vote.
On Mar 20, 2009, at 7:49 PM, Hilmar Lapp wrote:
> On Mar 18, 2009, at 5:19 PM, Brian O'Meara wrote:
>> For tree manipulation (traversal, etc.) in C++ I use Rod Page's
>> TreeLib, which
>> integrates well with NCL.
> Would this be a good candidate to use for optimizing the data
> representation? Or would C++ be a major hurdle for that?
> Another possibility that comes to mind (though it is also written in C
> ++) could be the PhylLib part of Bio++:
> : Hilmar Lapp -:- Durham, NC -:- hlapp at duke dot edu :
> Phylobase-devl mailing list
> Phylobase-devl at lists.r-forge.r-project.org
More information about the Phylobase-devl