Find identical levels for variables to fix GFS-FV3 update #423
Conversation
|
Before saving and committing, make sure you "Clear All Output" under the edit mode. We do that to avoid confusing the learners when we teach. |
|
Makes perfect sense - hopefully got it cleaned up. |
| "_, _, common_ind = np.intersect1d(lev_uwnd, lev_hght, return_indices=True)\n", | ||
| "temperature = temperature[common_ind, :, :]\n", | ||
| "# Reassign units to variables\n", | ||
| "temperature.attrs['units'] = 'kelvin'\n", |
There was a problem hiding this comment.
Is xarray dropping units here or are the original units bad?
There was a problem hiding this comment.
It's dropping units there.
There was a problem hiding this comment.
😞 That one doesn't seem sensible.
|
This has added a bunch of pycharm-specific json to the notebook. I'm unsure how to feel about that... |
|
I can create a new branch and edit the notebook just through Jupyter to keep that out if that would be better. |
|
It looks like this is primarily failing on the build against MetPy master because it was using |
different number of vertical levels between temperature and other variables, caused by the FV3 update to the GFS
|
Updated as requested |
|
Let me know if the errors with Travis are something I need to fix or if those are known issues. They don't look related to the notebooks already updated within this PR. |
|
Looks like current errors will be fixed by #425. |
Update to the Isentropic Analysis and NCSS notebooks to find identical vertical levels between temperature and other variables due to the differing number of vertical levels caused by the FV3 update to the GFS.
Fixes #422.