Open
Conversation
Collaborator
|
Thanks @travissluka ! It's still not clear to me what atlas grid type would handle MOM6 tripolar grid (which https://github.com/ecmwf/atlas-orca Does tripolar MOM6 grid differ from ORCA dramatically that we don't need such add-on? @danholdaway suggested using an atmospheric gaussian grid that VADER tests currently work for and just replace existing values with ocean variables. I can do that but i would like to have a permanent solution eventually. |
Contributor
Author
|
@Dooruk I agree with Dan that it's probably easier to just use an existing gaussian grid. I don't think we need to go through the trouble of making sure the same tri-polar grid works with the vader ctests |
Contributor
|
Tests on this PR are disabled until you merge or run |
Collaborator
|
The code here is now updated to use vader recipe that @Dooruk introduced in https://github.com/JCSDA-internal/vader/pull/208 in linear variable changes. There are still a few hacks here and there that I borrowed from @travissluka's original code, that we may be able to remove / simplify in the future. Another important point to note is that this approach (using I think we should test all our current applications with this branch properly, before we proceed with merging, but please review when you have a minute. I'll start testing with our applications once I have some feedback. |
Collaborator
|
@shlyaeva thank you so much for moving this forward this is a lot of work in multiple repos. I like the idea of separating TL/AD from the NL part. I follow the changes in I tested the SOCA + VADER |
Collaborator
|
thank you so much for testing @Dooruk! can you share the logs from the two runs? I didn't expect the runtime to be much different, curious to see where it's happening. I'll test in our setup too. |
Collaborator
Sure, if you want the full logs I can email them but here are the OOPS bits: with vader:Prior to vader: |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Description
changes:
test_soca_3dvar_godasnow usesVertInterpinstead ofInsituTemperatureand so should make use of VADER for the TL/AD@Dooruk you should be able to use this to test your vader branch. In order for it to compile, I added a
changeVarAD()empty stub, and commented out all the code inchangeVarTL()andchangeVarTraj()because they currently contain bugs.Dependencies
build-group=https://github.com/JCSDA-internal/vader/pull/208