reading Newick trees is taking a long time

[Ben Morris]

Hi,

I'm trying to open some rather large trees using dendropy.Tree.get_tree_from_path (Dendropy version 3.12.0.) The trees are in Newick format and can be found here: 

http://www.evoio.org/wg/evoio/images/3/37/Smith_2011_angiosperms.txt
http://www.evoio.org/wg/evoio/images/f/f9/Greengenes2011.txt

Reading these trees with dendropy is taking unreasonably long - several hours, in fact. For comparison, ete2 and Bio.Phylo can read in the trees in less than a minute.

Any idea what might be causing these long load times?

[Jeet Sukumaran]

Yep, DendroPy is slow with big trees. Without significant (probably
breaking) changes in the core infrastructure, I do not see any speed-
up possible. Nonetheless, several hours seems unreasonably long with
the angiosperm dataset. Some recent benchmarks I've run clocked
parsing this dataset at ~23 minutes on the average. Still too long for
many applications, I know ...

On Dec 3, 6:17 pm, Ben Morris <b...@bendmorris.com> wrote:
> Hi,
>
> I'm trying to open some rather large trees using
> dendropy.Tree.get_tree_from_path (Dendropy version 3.12.0.) The trees are
> in Newick format and can be found here:
>

> http://www.evoio.org/wg/evoio/images/3/37/Smith_2011_angiosperms.txthttp://www.evoio.org/wg/evoio/images/f/f9/Greengenes2011.txt

[Ben Morris]

That's too bad. I do think DendroPy is very intuitive and will probably continue to use it for smaller trees. Thanks for your help!