only use languages with known coordinates when importing data into clics about clics2 HOT 7 CLOSED

    glottolog = Glottolog()
    languoids = {l.id: l for l in glottolog.languoids()}
    l2geo = {}
    for lid, l_ in languoids.items():
        if l_.level == Level.language or l_.level == Level.dialect:
            while l_.latitude is None and l_.lineage:
                l_ = languoids[l_.lineage.pop()[1]]
            if l_.latitude is not None:
                l2geo[lid] = (l_.latitude, l_.longitude)

from clics2.

I think this mainly concerns dialects, and for these, we should default to using the coordinates of the corresponding Glottolog language.

from clics2.

Ah, is there a way to check for the corresponding language of a glottolog-dialect?

from clics2.

Just go up the glottolog tree until you find a language? I think I have written that code before, but I cannot remember for what application and for with which API version of the glottolog data.

@xrotwang, is the reason you suggest that that most Glottolog dialects don't have coordinates?

Then it should be “walk up the glottolog tree until you find something with coordinates”, just in case future super-dialects contain coordinates?

from clics2.

Is there a distinction between "dialect" and "language"?

from clics2.

That may in fact increase the data, as we are loosing some 100 languages now...

from clics2.

Yes, in fact, we can save almost all languages in the sample. I only kicked out the Bookkeeping families.

from clics2.