-
Notifications
You must be signed in to change notification settings - Fork 40
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Numpy version update #1164
Numpy version update #1164
Conversation
Conflict with xcdat: https://github.com/xCDAT/xcdat/blob/96bc649ad59750ec37b959d9d4890891b54c62c7/conda-env/dev.yml#L14. show details
@tomvothecoder do we have any plan to update numpy version dependency of xcdat? |
Just replied here, yes I will remove the |
Good to know, thank you! |
xcdat v0.7.3 released: https://github.com/xCDAT/xcdat/releases/tag/v0.7.3 |
xcdat v0.7.3 is now available on |
…lict with updated numpy version
Now I can create conda env from dev.yml without any conflict show details
|
Error from Demo 1a notebook Click here for Error message
|
(edited) Check boxes for demo notebook moved to the top of this PR. |
Error from MoV
Related issue: ajdawson/eofs#151 |
Error from MJO Click here for Error message
|
Error from ENSO:
|
Error from Demo 7 precip variavbilitiy
Error message:
|
Error from Demo 8:
This does not generate the output directory |
Not surprisingly, bunch of errors raised from demo notebooks, likely resulted from the numpy version update. Demo notebooks 0 to 3 run okay, but others are failing. @acordonez I will work on notebooks 4 to 6, and can I get you help on notebooks 7 to 9? Just a reminder, to test these, it is needed to generate a new conda env using the dev.yml in this branch. |
Regarding the error with ENSO, a PR. made for the ENSO package: CLIVAR-PRP/ENSO_metrics#80 |
Confirmed now demo 6 ENSO notebook working without issues after updating enso_metrics version to 1.1.2. |
Confirmed MoV notebook also working after updating eofs to 2.0.0. |
Update build_workflow.yml
@lee1043 I didn't find any problems with notebooks 8, 9, or 9b. In the comment above I documented some of the experimentation I did with the precip variability metrics and have one suggestion for a workaround. |
@acordonez thanks for the update. Great to hear there were no issue with 8, 9, and 9b. What do you consider as the most efficient and quick solution for the issue with 7? By the way, I found that CI build test was failing with the following message. Have you seen something like this? I am yet to have good idea how to address it.
|
@lee1043 For the precip variability metrics, this is the solution I found that worked in all cases: I'm not sure about the conda build failure. It seems like it might be having difficulty solving the environment and either failing or timing out as a result? |
@acordonez thank you for the suggestion. I will make that change to this PR. Regarding the conda build, I think the above changes in the yml files helped passing the test. In summary, I fixed some versions of dependencies, and set conda channel to be flexible, and set to use conda classic solver. |
Confirmed demo 7 working with no noticeable issue and the results are consistent. |
Merging the PR after reviewed by @lee1043 and @acordonez and all CI tests passed. |
This is one of steps toward xCDAT migration from CDAT.
numpy
version has been fixed as1.23.5
because of conflict with CDAT. Patch of CDAT particular for this conflict was made, which will work withnumpy >= 2.0
that is considered for the next version of E3SM Unified environment.PMP's step toward entire migration:
numpy
dependency and make sure it won't break anything.This PR is for the above step 1.
Demo notebook run check: