Feature/script update#371
Conversation
|
@aerorahul This is the 1st version that I put together. All vital components are in place for a successful ATM-only experiment replicating the develop branch. I made it a draft because it wasn't tested yet. |
|
Please start following global-workflow commit message guidelines (see https://github.com/NOAA-EMC/global-workflow/wiki/Development-of-Global-Workflow#commit-message-standards) |
|
@WalterKolczynski-NOAA Appreciate bringing this up. I updated the description. |
|
@JianKuang-UMD this PR is already out of date with feature/coupled-crow after a PR from @WalterKolczynski-NOAA was merged yesterday. There is one (at most 2) more PRs from @WalterKolczynski-NOAA which has changes needed for p7 but actually will likely help by adding features in the forecast script that are in the develop branch but not in feature/coupled-crow yet. |
|
@JessicaMeixner-NOAA So should I update this PR now or after the other upcoming PR goes into coupled-crow? |
|
@WalterKolczynski-NOAA Should I update this PR now or after the other upcoming PR goes into coupled-crow? |
|
I'd wait until after coupled-crow settles. The next one has the bigger forecast script changes. |
|
@WalterKolczynski-NOAA |
|
This is still more work for the next prototype. |
|
How are they ready to run prototypes in 2 weeks if more work is required? |
|
We've said the whole time we're trying to have things ready by the end of this week, though now it may slip to Monday. |
|
I ran the develop branch on Orion and compared my output to Jian's run with the modular forecast script. Same ICs used for both runs. There is a reproducibility issue with this hash of ufs-weather-model but the forecast output up through f15 matched. See comparison below. |
aerorahul
left a comment
There was a problem hiding this comment.
The indents in the scripts are completely whacky.
Please use 2 space indents in the scripts. 4 spaces are acceptable too, but need to be consistent.
exglobal_forecast.sh is removed, why? And if it is removed, how are you running it.
Please don't add new exglobal scripts with _bk. What is the intention of having those?
Version control takes care of this stuff.
|
@KateFriedman-NOAA @JianKuang-UMD |
Co-authored-by: Rahul Mahajan <aerorahul@users.noreply.github.com>
|
For my tests: |
…ture/script_update
|
I updated the contents of feature/coupled-sprint into this PR, which will let us keep up with the develop branch. I created a tag named "coupled_sprint_start" pointing to the hash of coupled-crow when I initially started the merge. I will update this PR with coupled-crow branch later today. |
version of feature/coupled-crow, tag: coupled-sprint-0728 Updated template files in parm and ush to the newest version of feature/coupled-crow, tag: coupled-sprint-0728
…ture/script_update
|
@JianKuang-UMD |
|
@JianKuang-UMD |
|
Trailing spaces in the |
|
@aerorahul Unfortunately I wasn't able to see those trailing whitespaces. And for some reason I am unable to simply push a change for file permissions. |
WalterKolczynski-NOAA
left a comment
There was a problem hiding this comment.
A couple changes that were supposedly made haven't actually been made. Maybe forgot to push some commits? Also one syntax issue.
| 2) | ||
| cat >> input.nml << EOF | ||
| gwd_opt = 2 | ||
| do_ugwp = .true. |
There was a problem hiding this comment.
This still needs to be changed back to .false.
move OCNRES to config.fcst use DO_* switches to control parsing_namelists_FV3.sh eliminate duplicated variable ntasks_efcs
Improved readability. We can discuss ternary operators as a group, though I don't have a big issue with them since there are at least clearly assignments and they are designed to do that. But I also know I had trouble with them once upon a time, so I wouldn't say no to prohibiting them to keep code more readable. Bash doesn't have them anyway. |
…ayer in high latitudes (NOAA-EMC#371) from Huiya.
… suite definition files (NOAA-EMC#371) * address error entered in stochastic_physics_wrapper.F90 that broke the ca restart reproducibility * Remove scheme rayleigh_damp from all remaining suite definition files Co-authored-by: Lisa Bengtsson <Lisa.Bengtsson@noaa.gov>
* Remove obsolete OMPX switch from code Issue NOAA-EMC#371
Modularized forecast script reproducing the develop branch.
All vital components of the modularized forecast scripts are in place for a
successful ATM-only experiment replicating the develop branch. Please
comment on how could this PR be tested, as well as the test plan of the
"final" merge toward the develop.
Issue tracker: #365 and #366
Comment: I think maybe a specific hashtag should be used instead of the
vague "develop branch"
Jian Kuang's test contents for the modularized script:
ROTDIR: /work/noaa/marine/Jian.Kuang/comroot/modular-192
EXPDIR: /work/noaa/marine/Jian.Kuang/expdir/modular-192
RUNDIR: /work/noaa/stmp/jkuang/RUNDIRS/modular-192/2020090118/gfs/fcst.143393
07/31 test contents, this run is NOT completed successfully, due to the following error which also appear in feature/coupled-sprint branch as of that date:
FATAL from PE 0: mpp_domains_define.inc: At least one pe in pelist is not used by any tile in the mosaic
Repo: /work/noaa/marine/Jian.Kuang/coupled-sprint
EXPDIR: /work/noaa/marine/Jian.Kuang/expdir/modular_0731
ROTDIR: /work/noaa/marine/Jian.Kuang/comroot/modular_0731
RUNDIR: /work/noaa/stmp/jkuang/RUNDIRS/modular_0731/2020090118/gfs/fcst.355316