Add config file for diagnostic scripts and use it to generalize sst_eval.py #74
Conversation
|
Hi @gabyneg and @theresa-morrison, @uwagura is currently working on generalizing our diagnostic scripts to support multiple regional domains. Could you please share the post-processing (PP) paths on archive for the Hawaii and Arctic domains? They don't need to be the best results; we just want to test the script and ensure it functions correctly across different domains. Thanks! |
|
Hi @yichengt900 and @uwagura, my PP step has not been functioning properly, so I have been using my output from my history files e.g. : "/archive/g1n/fre/cefi/MHI/CEFI_MHI_COBALT_V2/gfdl.ncrc5-intel22-prod/history" |
|
I can definitely look into the pp issue @gabyneg . And thanks for sharing the path to the history files! |
|
@uwagura Thank you! Here are my xml files (physics only) "/gpfs/f5/cefi/proj-shared/MHI/xmls/MHI/MHI_physics.xml", & (physics + bgc): "/gpfs/f5/cefi/proj-shared/MHI/xmls/MHI/CEFI_MHI_cobalt.xml". |
|
@yichengt900 Here is a path with pp for the full simulation period: Note that Kate has adjusted a lot of parameters since I ran this experiment, so it is not the up-to-date state of the Arctic configuration! |
|
@theresa-morrison, thanks for sharing the path. No worries, the main goal here is to make sure the script can work with all the different domains we have, especially the Arctic one considering its unique grid projection. @uwagura, please give it a try for these two domains too, so we can ensure it works for all our existing regional domains except the Great Lakes. |
…e_map keys instead of separate list
…option to select data projection
…rom process_glorys
…rom sst_eval import list
…line from annotate function, addressed label_mode error
There was a problem hiding this comment.
@uwagura , thanks for your great work! The new plots look good. Approve.
This pull request adds a config.yaml file to the physics diagnostics directory to make it easier to run the diagnostics scripts across domains by removing hardcoded variables. As a first test for the new config file, this PR also modifies the sst_eval.py script to use variables supplied by the config file. The script produces the same output as before for the NWA domain, but it can now be used across other domains as well by modifying the variables within the config file. Below is are the outputs from the script in the NWA and NEP domains using the new config file