Feature/mesoscale mpmd g2g #670
Conversation
parm/metplus_config/stats/mesoscale/precip/PcpCombine_obsCCPA.conf
Outdated
Show resolved
Hide resolved
|
Testing just jevs_mesoscale_nam_precip_stats.sh with vhr=01 to make sure everything looks good with the MPMD. 1. jevs_mesoscale_nam_precip_stats.sh (vhr=01)Log File: /lfs/h2/emc/vpppg/noscrub/mallory.row/verification/EVS_PRs/pr670/EVS/dev/drivers/scripts/stats/mesoscale/jevs_mesoscale_nam_precip_stats_00.o181623068 It looks like the code needs to be updated with the change of the CCPA file output path to
|
|
The ccpa observation is now linked to the job_num_work_dir and then copied to the COMOUT directory like it should be. You can now test again. |
|
I have manually submitted the NAM and RAP precip jobs for vhr=00-12, and a cronjob is set up in place to continue running the rest of the vhrs for today. I can't run for VDATE=PDYm3 because the MRMS data isn't around for that long. I ran into that problem yesterday. |
|
That sounds good, cron is the best way to get all the hours in there. Then I can look at the results tomorrow. |
|
@malloryprow Can you send me the path to where the .o files are located, so I can at least follow along there to check for errors? And the path to the working directories so I can check to see that the job subdirectories are working as intended? |
|
Yes the log files are in /lfs/h2/emc/vpppg/noscrub/mallory.row/verification/EVS_PRs/pr670/EVS/dev/drivers/scripts/stats/mesoscale. I ran the snowfall jobs too for vhr=00,06,12 but waiting until 18Z to do vhr=18. |
|
@malloryprow Everything looks good from what I've seen thus far. |
|
COMOUT is /lfs/h2/emc/vpppg/noscrub/mallory.row/verification/EVS_PRs/pr670/evs/v2.0/stats/mesoscale. 1. jevs_mesoscale_nam_snowfall_stats.shLog Files: /lfs/h2/emc/vpppg/noscrub/mallory.row/verification/EVS_PRs/pr670/EVS/dev/drivers/scripts/stats/mesoscale/jevs_mesoscale_nam_snowfall_stats_00.o* 2. jevs_mesoscale_rap_snowfall_stats.shLog Files: /lfs/h2/emc/vpppg/noscrub/mallory.row/verification/EVS_PRs/pr670/EVS/dev/drivers/scripts/stats/mesoscale/jevs_mesoscale_rap_snowfall_stats_00.o* |
|
Snowfall is good! Final stats files match what is in emc.vpppg. All the individual job directories are being used. I think we're good with the snowfall. |
|
Agreed! I just got done looking through everything myself. |
|
@malloryprow I noticed that the gather job failed, and it was due to a missing curly bracket in the StatAnalysis config file. I corrected that and made the correction. Please re-run the vhr=23 job for VDATE=20250224, precip jobs. |
|
Rerun! |
|
Ah, excellent! Precip is good. The precip stats files match what is in emc.vpppg. The job subdirectories look good too as well. |
|
Yup, looked good to me too! |
There was a problem hiding this comment.
Code changes are good and testing successful.
Thank you @PerryShafran-NOAA!
|
@AndrewBenjamin-NOAA Would you please review the code so we can get this PR merged? |
Note to developers: You must use this PR template!
Description of Changes
Relating to issue #557 :
This PR completes the fixing of the Bugzilla 1547, ensuring that scripts that use MPMD have child processes where the work is done in. This particular script adds the child processes for NAM/RAP mesoscale precip and snowfall stats. The rest of the mesoscale component, stats and plots, was completed earlier in PR #619 .
Developer Questions and Checklist
Yes, it is a bugzilla fix.
No
No
${USER}where necessary throughout the code.HOMEevsare removed from the code.dev/drivers/scriptsordev/modulefileshave been made in the correspondingecf/scriptsandecf/defs/evs-nco.def?Testing Instructions
https://github.com/PerryShafran-NOAA/EVS.gitfeature/mesoscale_mpmd_g2gdev/drivers/scripts/stats/mesoscale.jevs_mesoscale_nam_precip_stats.shandjevs_mesoscale_rap_precip_stats.shusingqsub -v vhr=${vhr},where ${vhr}=1 to 22. When all 22 scripts are completed, then run usingqsub -v vhr=23to do the gather step.jevs_mesoscale_nam_snowfall_stats.shandjevs_mesoscale_rap_snowfall_stats.shusingqsub -v vhr=${vhr}, where ${vhr}=00, 06, 12. When these three runs are completed, then run usingqsub -v vhr=18to do the gather step.