Datatree design discussions - weekly meeting

[TomNicholas]

What is your issue?

In the bi-weekly dev meeting today we agreed that deliberate higher-level discussions of datatree's design would be useful. (i.e. we're not worried about our ability to write high-quality code, so let's focus review time more explicitly on the high-level design questions.)

This could take the form of me just talking through what I did in a certain part of the code and why, or a targeted discussion on specific design questions that I was never quite sure about. Some examples of the latter, as food for thought:

• Inheritance of dimension coordinates from parent nodes? Resolve dimension coordinates from parents xarray-contrib/datatree#297
• Symbolic links? Allow symbolic links between nodes xarray-contrib/datatree#5 (we decided this was overkill)
• Is dt.ds ugly? See also the difference between dt.ds and dt.to_dataset() Tree linkage + dataset in one object nonideal for UI tree view xarray-contrib/datatree#303 (comment)
• Which methods should map over the subtree and which shouldn't? (can't find the issue for this one)
• Ignore missing dims when mapping over subtree? Ignore missing dims when mapping over tree xarray-contrib/datatree#67
• API for sub-tree selection Tree-aware dataset handling/selection xarray-contrib/datatree#254
• API for merging leaves API for collapsing subtrees xarray-contrib/datatree#192
• Dict-like interface ambiguities Consistency between __contains__ and __getitem__ xarray-contrib/datatree#240
• The tree broadcasting rabbit hole Tree "broadcasting" xarray-contrib/datatree#199
• Relationship between datatree and catalogs Intake, catalogs, and datatree xarray-contrib/datatree#134
• Should xr.concat/xr.merge accept DataTree objects? (and map over them by default?) Support DataTree in Xarray's top level functions #9106

There was also this design doc I wrote at one point

@flamingbear are you free at 11:30am EST on Tuesday each week? @shoyer, @keewis and I are all free then. Others also welcome (e.g. @owenlittlejohns , @eni-awowale, @etienneschalk), but not required :)

[flamingbear]

@TomNicholas Yes, I'm good on 11:30am EST Every Tuesday.

[eni-awowale]

Thanks for putting this together @TomNicholas! 11:30 EST on Tuesdays should generally work for me.

[TomNicholas]

Great. 11:30am-12:00am EST every week it is then! (We can overrun if necessary, but prioritize agenda items so as to respect people's time who can't stay for longer.)

Google calendar event link (There is an xarray calendar but I don't think I have access to it)

We'll use the same zoom link as for the bi-weekly dev call.

Meeting notes

[flamingbear]

Google calendar event link (There is an xarray calendar but I don't think I have access to it)

FYI I can't open this calendar link.

[TomNicholas]

FYI I can't open this calendar link.

Ugh - I tried changing the account which owns the event, but if that still doesn't work then just forget about it. I'll post here if we need to cancel or change the time or anything.

[TomNicholas]

Following up from our discussion today (@shoyer @owenlittlejohns @keewis @flamingbear) the original issue about storing xr.Variable objects rather than xr.Dataset objects is here: xarray-contrib/datatree#2 (comment). My actual reasoning is recorded in this comment xarray-contrib/datatree#2 (comment), but this is not set in stone!

[TomNicholas]

I can't make the normal time of 11:30am EST tomorrow (it clashes with a job talk I need to attend).

The only other time tomorrow I could do is 1 hour earlier (10:30am EST). I'll be on zoom then anyway in case anyone shows up!

@shoyer @flamingbear @eni-awowale @owenlittlejohns

[schlich]

I'd like to pop in and be a fly on the wall of these meetings if they're still happening... lmk

[TomNicholas]

@schlich yes they are still happening! You missed one yesterday in fact. You're welcome to pop in.

[schlich]

Wonderful, it's in my calendar

[TomNicholas]

Now that we've released DataTree in xarray v2024.10.0, I'm going to cease these meetings, in favour of the normal bi-weekly xarray meetings (#4001). Anyone is still welcome to those, and we can discuss DataTree issues there!