feat: Add a tool for writing B-fields to disk in CSV format

[stephenswat]

This PR adds a new example tool, CsvBFieldWriter, that dumps a B-field to disk in CSV format. The tool supports both grid-based and non-grid-based vector fields, it supports Cartesian and symmetrical cylindrical coordinates, and it supports user-specified ranges and bin counts.

Comes with free Python bindings.

[codecov]

Codecov Report

Merging #1470 (3535486) into main (b032bd2) will increase coverage by 0.07%.
The diff coverage is n/a.

❗ Current head 3535486 differs from pull request most recent head c9d82cc. Consider uploading reports for the commit c9d82cc to get more accurate results

@@            Coverage Diff             @@
##             main    #1470      +/-   ##
==========================================
+ Coverage   48.46%   48.54%   +0.07%     
==========================================
  Files         384      381       -3     
  Lines       21071    20785     -286     
  Branches     9702     9536     -166     
==========================================
- Hits        10213    10090     -123     
+ Misses       4130     4112      -18     
+ Partials     6728     6583     -145     

Impacted Files	Coverage Δ
Core/src/Surfaces/RectangleBounds.cpp	33.33% <0.00%> (-53.34%)	⬇️
Core/include/Acts/Surfaces/Surface.hpp	33.33% <0.00%> (-33.34%)	⬇️
Core/src/Definitions/Common.cpp	0.00% <0.00%> (-32.00%)	⬇️
Core/src/EventData/VectorMultiTrajectory.cpp	66.85% <0.00%> (-10.35%)	⬇️
Core/src/Utilities/Logger.cpp	30.43% <0.00%> (-8.70%)	⬇️
Core/include/Acts/EventData/MeasurementHelpers.hpp	80.00% <0.00%> (-5.72%)	⬇️
...e/include/Acts/EventData/VectorMultiTrajectory.hpp	52.89% <0.00%> (-5.62%)	⬇️
Core/include/Acts/EventData/MultiTrajectory.ipp	64.36% <0.00%> (-2.74%)	⬇️
Core/include/Acts/Propagator/ConstrainedStep.hpp	70.21% <0.00%> (-2.13%)	⬇️
Core/include/Acts/Propagator/Navigator.hpp	54.76% <0.00%> (-0.89%)	⬇️
... and 35 more

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

[andiwand]

looks great 👍

[asalzburger]

Is this good to go? There are still conversations unresolved.

[stephenswat]

I think we're waiting for some comments from @timadye.

[stephenswat]

Okay, unless Tim has any comments on the outstanding issue I think this is good to go from my side.

[timadye]

Sorry, I was away. I'll take a look today.

[timadye]

Are you planning to include a CSV reading tool? Apart from providing another good format, that would be useful to test this by writing and reading back.
I see there's a Examples/Detectors/MagneticField/src/FieldMapTextIo.cpp, but that looks like it is space-separated values.

[timadye]

Does this have a different interface compared to Examples/Io/Root/include/ActsExamples/Io/Root/RootBFieldWriter.hpp? It looks like this has the CsvBFieldWriter::CoordinateType coded as a template parameter, rather than an enum Config::gridType. That could be an improvement, but maybe needs considering. If it is better, we might consider changing RootBFieldWriter at a later stage. If so, do we need a more general BFieldWriter::CoordinateType or whatever?

Does this difference follow through to the Python API? It would also be nice to have a Python example and/or test to be sure it works.

Sorry for all the comments.

[stephenswat]

Hi Tim, I do have some changes in the pipeline to the Python B-field writing examples, but I opted to keep them separate from this PR for the time being. I like the idea of having a CSV reader, and I can work on implementing one.

Regarding the coordinate type enum, I agree that it might be useful to merge them, but there is also something to be said for embedding the capabilities of each writer into the class itself.

[paulgessinger]

@timadye @stephenswat can you advise if we can merge this?

[timadye]

I still had one issue outstanding, here. I hope @stephenswat can check that.

The definitive check would be to read the ATLAS B-field, write with this PR, and check the files have the same binning. If you are not able to do that, I can give it a go.

[stephenswat]

I have incorporated Tim's patches, and marked him appropriately as a co-author of the commits.

[andiwand]

this conflicts now with #1510

happy to re-approve after the conflict is resolved

[timadye]

I have incorporated Tim's patches, and marked him appropriately as a co-author of the commits.

Thanks @stephenswat! This looks good to me.

I guess you will make the other changes in a separate PR, namely:

1. harmonize the runXyzGrid() methods with the RootBFieldWriter and others.
2. replace my ugly dynamic_cast with a static test, eg. if constexpr.

We discussed this in the Tuesday dev meeting and were happy to go ahead with this PR as-is. I could still test the output is correct on the ATLAS B-field. Should I do that now, or wait for you to resolve the conflict with #1510?

[asalzburger]

Hi @stephenswat - can you resolve the conflict here? Then we should be able to get this in.

[stephenswat]

Conflict resolved.

[timadye]

Let me test this now, and then I hope we can merge this sucker.

[github-actions]

📊 Physics performance monitoring for c9d82cc

🟥 ERROR The result has missing elements!
This is likely a physmon job failure

Full report
CKF: ❌ seeded, ❌ truth smeared, ❌ truth estimated
IVF: ❌ seeded, ❌ truth smeared, ❌ truth estimated
❌ Ambiguity resolution
❌ Truth tracking

Vertexing ❌

❌ Vertexing vs. mu

❌ IVF seeded

❌ IVF truth_smeared

❌ IVF truth_estimated

CKF ❌

❌ CKF seeded

❌ CKF truth_smeared

❌ CKF truth_estimated

Ambiguity resolution ❌

❌ seeded

Truth tracking ❌

❌ Truth tracking

[timadye]

I tested with the ATLAS B-field map, and it looks the same as the root input.