feat: add support for RNTuples #1311
Conversation
|
Oh, cool! Thanks for starting this! |
|
Are we in a spot where we can start to wrap up this functionality @ariostas? |
|
I think so. I was told that there was some major-ish refactoring underway and that I should wait. If that's done, I'm happy to get back to looking into this. Although I'll probably wait a couple weeks to see what happens on the CMSSW side. |
|
Quick update on this. After cms-sw/cmssw#48071 we can now produce RNTuple NanoAODs with CMSSW. The format is not final, but we can probably start playing around with it. |
|
@ariostas do you have anything further you'd like to do with this PR, shall we mark it ready for review? It needs a rebase as well, there are some conflicts. |
|
I'll spend some time on this next week. It probably mostly works, but I might need to make some changes on the Uproot side. |
|
@ariostas OK - great, thanks! Please keep me informed! |
ariostas
force-pushed
the
ariostas/rntuple_support
branch
from
6ba606c to
ffcea34
Compare
|
Quick update on this. Now that coffea switched to virtual arrays by default, it seems like we don't need any changes on the Uproot side. I got it to read the form correctly, but it is failing to materialize the arrays. I'm guessing that it's because it's assuming that all fields are at the top level, but I'll have to look into it. |
|
Awesome - thanks for coming back to this. |
This is still a work in progress, but I'm opening this PR so I can ask questions about my changes, since I'm not familiar with how Coffea works.
This PR depends on scikit-hep/uproot5#1419