reference trees about autocogphylo HOT 9 OPEN

I totally see your point. I just wanted to make sure, we are on the same page.

from autocogphylo.

I am working on it. I will try to get the reference trees asap.

from autocogphylo.

It seems that some of the iso-codes in the data set appear as inner nodes of our reference tress. What should we do here?

from autocogphylo.

remove the inner nodes if you have outer ones. But I'd need to see an example, as it should not happen too often, maybe only with dialects in a few cases.

BTW: glottolog api works for reference trees now, I checked, just make sure to have the most recent version, there's no bug in newick output.

from autocogphylo.

I do not fully get your suggestion here. We have cognate data for a language which appears as an inner node in the glottolog tree. But we have also cognate data for its daughters.

In the AA case "bahn1262" describes an internal node which has as its daughter "gola1254" and "kont1244" so we can not make it a terminal node easily.

from autocogphylo.

Ah, yes, sorry, you are right! Making it a terminal node seems to be best.

from autocogphylo.

So making it a sister to its daughters?

from autocogphylo.

Yes, Id' say so. You see: we only measure topology. If you have an internal node, say, Latin, with Italian and French, then the normal tree will make Latin the outgroup, but assign it zero branch lengths (ideally, Taraka has a paper on this). But since we only measure topology, branch lengths are not crucial for us, right?

from autocogphylo.

done with the last commit 20d6063

from autocogphylo.