search

airr-community / airr-standards

AIRR Community Data Standards
https://docs.airr-community.org
Creative Commons Attribution 4.0 International
35 stars 23 forks source link

DataFile and RepertoireGroup modifications #629

Closed javh closed 2 years ago

javh commented 2 years ago

Per #581, DataFile changes:

Per the August 8, 2022 call:

schristley commented 2 years ago

hopefully you can merge this and #581 without too much hassle

javh commented 2 years ago

Yes, I hope so. Once this is in, then I can finish the R package and #581 should be good to go. Everything else is done with it. (As near as I can remember anyway - I'm sure I forgot something and we'll need to patch release.)

scharch commented 2 years ago

I don't think I understand why Clone is a top-level object but Tree isn't. Neither is an object which is "within" another object (at least if I am reading the schema correctly).

javh commented 2 years ago

Good point. It doesn't look like it's setup that way in the schema. Should it be?

Though, my motivation for cutting Tree is more so I wouldn't have to figure out how to parse Tree.properties.nodes.additionalProperties.

javh commented 2 years ago

Also, @schristley and I were on the fence about cutting Clone, so I went with "no".

scharch commented 2 years ago

Should it be?

I.... need to think more. Let's leave it for now and revisit for 2.0?

javh commented 2 years ago

I.... need to think more. Let's leave it for now and revisit for 2.0?

Seems wise.

schristley commented 2 years ago

Should it be?

I.... need to think more. Let's leave it for now and revisit for 2.0?

It's that way now because there was the use case that multiple trees could be defined for a clone, so instead of embedding a single tree into the clone object, Tree allows multiple to be defined without introducing lots of redundancy.