Transport improvements for 2024-W31

[khaeru]

• Investigate scale-1 values/pass-through of values for c=methanol. Next PR.

Housekeeping/code improvement:

• Adjust the "transport" CI workflow to dispatch a run in message_data.
  • For security reasons, GitHub discourages allowing workflows in public repositories to run on private, self-hosted runners. This is because a malicious actor could open a PR that runs the workflow with modifications, accessing private data via the self-hosted runner.
  • This can be enabled in settings, but instead this PR takes a different approach.
  • The transport.yaml workflow is reduced to merely dispatch a workflow file that remains in message_data. That file in turn clones iiasa/message-ix-models (the specific branch/ref that is sending the dispatch) and runs the usual workflow.
• Clean up .model.transport after migration #210
• Adjust packaging so that .model.transport is included.
• Migrate reserve_margin() #211 Next PR.

How to review

• Read the diff.
• Note that the CI checks all pass.
• See that the transport.yaml workflow here dispatches a run of the corresponding transport.yaml workflow in message_data.

PR checklist

• Continuous integration checks all ✅
• Add or expand tests; coverage checks both ✅
• Add, expand, or update documentation. N/A, mostly CI changes.
• Update doc/whatsnew. N/A

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 58.8%. Comparing base (2fcae67) to head (de2446f).
Report is 311 commits behind head on main.

Additional details and impacted files

@@          Coverage Diff          @@
##            main    #212   +/-   ##
=====================================
  Coverage   58.8%   58.8%           
=====================================
  Files        194     194           
  Lines      15159   15159           
=====================================
  Hits        8922    8922           
  Misses      6237    6237           

Files with missing lines	Coverage Δ
message_ix_models/model/transport/CHN_IND.py	30.3% <100.0%> (ø)
message_ix_models/model/transport/roadmap.py	56.0% <100.0%> (ø)
...e_ix_models/tests/model/transport/test_emission.py	100.0% <100.0%> (ø)

[glatterf42]

Thanks for addressing these issues so quickly! The one failing test is likely just flaky, so looks good to me :)

[glatterf42]

Great, I especially like the dispatch summary linking directly to the message_data run :)