Swapi proton moments

[hafarooki]

Change Summary

New response function for SWAPI + new proton/alpha fitting algorithm to use the new response function. See new Markdown document file included in the PR for extensive details about the changes.

[jtniehof]

Can you rebase or merge main into the PR and resolve conflicts?

[hafarooki]

Should be all good now

[pleasant-menlo]

We manage the Python dependencies in the Docker containers using uv. Since this PR adds a dependency on numba, can you install uv and run uv add numba, then commit the updated pyproject.toml and uv.lock files?

After making this change locally, we were able to run the integration test in Docker. It returned a failure:

AssertionError:
Not equal to tolerance rtol=0.01, atol=0
proton_sw_speed regression
Mismatched elements: 1 / 3 (33.3%)
Max absolute difference among violations: 30.87983496
Max relative difference among violations: 0.06362372
 ACTUAL: array([474.574554, 475.23233 , 516.230835])
 DESIRED: array([474.576, 475.232, 485.351])

Is this expected and the test just needs to be updated, or does it indicate an error in the code?

[hafarooki]

will do!

That is expected, I made some algorithm changes over the past few days and need to update the assertions

[hafarooki]

Just to double check, I plotted the values around that time and compared with another spacecraft and it does look like the correct speed should be 516, not 485. So I'll go ahead and update those tests

[hafarooki]

Fixed the integration test, please try again when you get a chance @pleasant-menlo

[pleasant-menlo]

We have created two pull requests to configure the SDS triggering system.

• Update SWAPI l3a dependencies for new algorithm sds-data-manager#1319
  • Adds the new ancillary, MAG, and SPICE requirements
• SWAPI remove unused ancillary dependencies sds-data-manager#1320
  • Removes the no-longer-used lookup tables; to be merged only after the new processing code is deployed.

Before getting the first one merged, we would like to upload the new ancillary files to production so that the current processor continues to be triggered. These are:

• azimuthal-transmission
• central-effective-area
• passband-fit-coefficients

Are the versions in this PR suitable, or would you like us to wait to upload until you produce finalized versions?

[hafarooki]

The current versions are thoroughly validated and not likely to change much if at all for the upcoming data release. I think you can go ahead and upload this version of them for now. Will it be possible to upload a new version with minor changes later on?

[pleasant-menlo]

We renamed the files to 20250924, following the conventions, and uploaded them to production.

If you have an update in the future that should apply over the whole mission, you can upload v002 with the same 20250924 start date (mission launch date). If an update should only be applied for data after a specific time, you can upload v002 with a later start date and it will only affect data from the specified time onward.

We ran the integration test in Docker with your latest changes, and it passed.