search

lexibank / baf2

Bangime and Friends 2
Creative Commons Attribution 4.0 International
0 stars 0 forks source link

Please add MORPHEMES column to Edictor #8

Closed IndianaTones closed 2 years ago

IndianaTones commented 2 years ago

Could you please add a (permanent) MORPHEMES column to our Edictor Dataset so that I can begin adding partial cognates there for noun classes among the Dogon languages? This will be very interesting.

LinguList commented 2 years ago

Sorry, only saw this issue now.

LinguList commented 2 years ago

Adding columns is actually a no-brainer, but you need to adjust the URL you use to open edictor to SEE them.

LinguList commented 2 years ago

So if you use this URL (copy and paste it directly from here into the browser), you see all morphemes as a column:

https://digling.org/edictor/?remote_dbase=bangime&file=bangime&columns=DOCULECT|SUBGROUP|FAMILY|CONCEPT|FRENCH|VALUE|FORM|TOKENS|MORPHEMES|COGID|BORID|COGIDS|ALIGNMENT|COMMENT&basics=DOCULECT|CONCEPT|VALUE|FORM|TOKENS|MORPHEMES|COGID|BORID|COGIDS
IndianaTones commented 2 years ago

Got it - I wonder if I should re-run the code to see how it does with partial morphemes, or do this only on the new dataset?

LinguList commented 2 years ago

The code takes full cognates, no partial cognates. Have you drastically modified the data?

IndianaTones commented 2 years ago

I have modified the ID's yeah...

LinguList commented 2 years ago

This means we have to re-run the code and write a short update for ORE. This does not need to be long, just make sure numbers match, etc., and indicate that the underlying data is now in a new zenodo version, which we'd fix then.

IndianaTones commented 2 years ago

If you think it's worth it at this point? If not, can also be done on the new dataset (LanguageIslands).

LinguList commented 2 years ago

This depends on what you want to do. ORE offers to make a revision whenever you want. But if we consider this as closed, I'd say: no more modifications here.

IndianaTones commented 2 years ago

Yeah, I think we can close this one and focus on the new one. I had wondered about doing this some time ago; seems like it's best to work on the new dataset now though I think.

LinguList commented 2 years ago

That is perfectly fine, and more consistent. We can alter easily compared with this dataset.