search

m2lines / convection-parameterization-in-CAM

Deployment of the convection parameterization scheme (previously from within SAM) deployed into the CAM model
MIT License
3 stars 2 forks source link

Understand the build process of the source repo (for SAM) #3

Closed dorchard closed 1 year ago

dorchard commented 1 year ago

This may not be strictly necessary, but would probably be useful from the point of view of understanding the dependencies.

dorchard commented 1 year ago

Should we close this now?

jatkinson1000 commented 1 year ago

I'd say probably yes, since Paul seemed to think we don't need to build here. It might be nice to be able to verify the removed module works however? Perhaps via a testbed fortran program and comparison to the original pytorch? This would be something for #9 (or possibly a new issue). @jdenholm and I could work together, him from the pytorch end, me from the Fortran?

jatkinson1000 commented 1 year ago

Closing this as discussed above.

jatkinson1000 commented 1 year ago

Reopening based on discussions associated with trying to run on CSD3 for purposes of testing refactored parameterisation code.

jatkinson1000 commented 1 year ago

It would be useful to have SAM running with the old parameterisation and then switch for our refactored version to ensure everything still works as expected before moving to CAM (#ToBeOpened).

I have managed to compile the SAM code provided here on CSD3 after much effort (build scripts and makefiles etc. were not provided). However, I’m unable to run as there are no testcases or run instructions provided. Based on the documentation for the latest version of SAM I have created a CaseName file and copied the prm file from sam_code_NN/sam_cases/run_files_x8_5_layers/ to a new case, but it suggests that I also need, at minimum, an snd file, and possibly others such as grd, lsf, sfc etc. It also appears that the case is run from a RESTART which has not been provided. I have requested these files and instructions from the authors.

mondus commented 1 year ago

Janni has provided a link to these files on cheyenne but these are in a home directory so may not be accessible to Jack (even if he has access to the cheyenne system).

mondus commented 1 year ago

@paogorman Can provide build scripts that accompany Yannis original code. @jatkinson1000 can diff this with the build scrips that he has extracted from SAM. We will then fork Yannis repo and add the necessary build scripts and binary files so that the paper can be reproduced.

jatkinson1000 commented 1 year ago

This is now ready for review and merge on the ICCS fork on Janni's code: PR #3 Create a version of SAM(SON) code that successfully compiles

jatkinson1000 commented 1 year ago

Reviewed by Dominic and merged. This can now be closed.