+
Skip to content

Conversation

yantosca
Copy link
Contributor

This PR pegs the xarray version at 2023.2.0 in the environment file docs/environment_files/environment.yml. This is because the most recent version (2023.8.0) seems to have some under-the-hood shenanigans with dask that breaks backwards compatibility with the benchmark plotting scripts.

This is a temporary measure intended to give us a little more time to figure out what code modifications are needed so that GCPy can work with the latest xarray.

docs/environment_files/environment.yml
- The latest xarray (2023.8.0) seems to have some under-the-hood
  shenanigans with dask that break the benchmark plotting capability.
  For the time being, peg xarray to 2023.2.0, which seems to avoid
  this behavior.

Signed-off-by: Bob Yantosca <yantosca@seas.harvard.edu>
@yantosca yantosca added topic: User Environment Relating to python and/or conda environment category: Bug Fix Fixes a bug that was previously reported labels Aug 29, 2023
@yantosca yantosca added this to the 1.4.0 milestone Aug 29, 2023
@yantosca yantosca requested a review from msulprizio August 29, 2023 19:05
@yantosca yantosca self-assigned this Aug 29, 2023
@yantosca
Copy link
Contributor Author

Fixes #254

@yantosca yantosca merged commit 43bb419 into dev Aug 29, 2023
@yantosca yantosca deleted the bugfix/peg-xarray-to-2023-02 branch August 29, 2023 19:26
yantosca added a commit that referenced this pull request Aug 29, 2023
docs/environment_files/environment.yml
- In the prior commit we have added updates to plot.py that render
  the pegging of xarray to version 2023.2.0 unnecessary.  Restore
  the original code in the GCPy environment.yml file.

Signed-off-by: Bob Yantosca <yantosca@seas.harvard.edu>
yantosca added a commit that referenced this pull request Aug 30, 2023
This merge brings PR #257 (Update plot.py for more recent xarray; also
allow dask arrays to be passed to single_panel, six_plot routines, by
@yantosca) into the GCPy 1.4.0 development stream.

This PR does the following:
1. Now use "ds_new[lon_var].compute()" and "ds_new[lat_var].compute()"
   to evaluate the lon_var and lat_var fields.  This behavior is now
   necessary in the latest xarray (2023.8.0).
2. Allow routines "six_plot" and "single_panel" in gcpy/plot.py
   to also accept dask arrays (dask.array.Array)
3. Updated documentation
4. Reverted the update to environment.yml from PR #255; now allow
   latest xarray in the Python environment.

Signed-off-by: Bob Yantosca <yantosca@seas.harvard.edu>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

category: Bug Fix Fixes a bug that was previously reported topic: User Environment Relating to python and/or conda environment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

[BUG/ISSUE] Index error when creating 1-year benchmark plots for GCClassic vs GCHP

2 participants

点击 这是indexloc提供的php浏览器服务,不要输入任何密码和下载