402 merge from am3 params

[JhanSrbinovsky]

CABLE

Thank you for submitting a pull request to the CABLE Project.

Description

Merge modifications made during in AM3 development.

You can link issues by using a supported keyword in the pull request's description or in a commit message:

Fixes #402

Type of change

Merge modifications made during in AM3 development.

Checklist

• The new content is accessible and located in the appropriate section.
• I have checked that links are valid and point to the intended content.
• I have checked my code/text and corrected any misspellings

---

📚 Documentation preview 📚: https://cable--404.org.readthedocs.build/en/404/

[ccarouge]

Sorry for the delay. I forgot to final step of the review process.

[ccarouge]

The regression testing should happen in benchcab. The user guide is here: https://benchcab.readthedocs.io/en/latest/

I would expect the following config file to do what you need:

# Example configuration file for running the CABLE benchmarking.
# 
# Note, optional keys are available in this config file. See
# https://benchcab.readthedocs.io/en/stable/user_guide/config_options/
# for documentation on all the available keys.
# 
# This file is in YAML format. You can get information on the syntax here:
# https://yaml.org/spec/1.2.2/#chapter-2-language-overview
# Quick tips:
# You need a space after the ":"
#
# It uses the same syntax as Python for:
#     - lists (aka sequences)
#     - dictionaries (aka mappings with key/value pairs)
#
# Strings can be given with or without double or single quotes.

realisations:
  - repo:
      git:
        branch: main
  - repo:
     git:
       branch: 402-merge-from-am3-params
  
modules: [
  intel-compiler/2021.1.1,
  netcdf/4.7.4,
  openmpi/4.1.0
]

And you don't need to do the analysis on modelevaluation.org. Just check the log file to ensure the regression testing was all successful.

[JhanSrbinovsky]

@ccarouge, I can run benchcab and get results from dozens of runs but the documentation isnt clear WRT what the output is. So what is S0_RO, etc. I have two beaches A and B. How can I see that A is effectively identical to B?

[ccarouge]

@JhanSrbinovsky This part in the documentation should tell you where to look:

runs/fluxsite/analysis/bitwise-comparisons
directory that contains the standard output produced by the bitwise comparison command: benchcab fluxsite-bitwise-cmp. Standard output is only saved when the netcdf files being compared differ from each other

But I agree, we need to make it clearer where people can find the information about the results in the docs.

[JhanSrbinovsky]

Doh - so you're telling me that the only directory that sounded like what I thought might be relevant was empty by design

[ccarouge]

The PBS log file should also have human-friendly logs of success for the regression tests.

[JhanSrbinovsky]

I haven’t looked yet but there has to be an equivalent offline. They could even be hard-wired offline. They were previously hardwired here as well.

[JhanSrbinovsky]

DOH! - I've got myself in a muddle here. I've started to make changes pushing nCNP_pools index to coupled only, but I have done work in a new issue #412 (+ I have a local branch going) PLUS I have had to make the corresponding changes in AM3 to make sure it works there. Between juggling the kids and a new puppy I've got all these branches mixed up.

[JhanSrbinovsky]

BTW - renewed regression testing not only shows no unintended changes in output but absolutely zero changes in output