Implement rapidity-based DeltaR calculations in TLorentzVector

[mattleblanc]

Hello,

At the moment, the ROOT::TLorentzVector class uses a pseudorapidity-based calculation for the commonly-used DeltaR distance between vectors. For massive objects, such as jets, the rapidity should often be used instead. I have implemented a small switch within the TLorentzVector::DeltaR function to allow users to make this calculation with the vector's rapidity, instead of pseudorapidity. This has a default value of kFALSE, which means that the current behaviour will be the default one.

I have cross-checked this new functionality with the fastjet::pseudojet::delta_R implementation and an internal ATLAS one, which all give the same results when performing the rapidity-based calculation.

Please let me know if I have violated any contribution guidelines, and I will update this PR accordingly!

🍻 Matt

[phsft-bot]

Can one of the admins verify this patch?

[mattleblanc]

Hi, I guess this should also be implemented in VectorUtil -- this is a work-in-progress until I push another commit. Sorry for not sending it all at once!

MLB

[lmoneta]

Hi Matt

Than you for your contribution!

If you have time, it would be nice to add also this also for the Genvector package for the ROOT::Math::LorentzVector classes. This can be added as a new function in the VectorUtil file, something like DeltaRapidityPhi.
I am not sure if it is better just having just the new function, DeltaRapidityPhi instead of DeltaR with a boolean parameter.

Cheers

Lorenzo

[mattleblanc]

I've added this functionality now also to ROOT::Math::GenVector, it gives the same results as the other implementation and fastjet::pseudojet. I'm not sure if this is exactly how you wanted it implemented, so please let me know if you have further comments -- I've tried to keep everything as consistent as possible between the two sets of changes.

I have un-marked the PR as a work-in-progress.

[mattleblanc]

Hi @lmoneta, this is just a ping in case you didn't see the other message -- there is obviously no rush, though.

[oshadura]

@phsft-bot build!

[phsft-bot]

Starting build on ROOT-debian10-i386/cxx14, ROOT-performance-centos8-multicore/default, ROOT-fedora30/cxx14, ROOT-fedora31/noimt, ROOT-ubuntu16/nortcxxmod, mac1014/python3, mac11.0/cxx17, windows10/cxx14
How to customize builds

[phsft-bot]

Build failed on ROOT-fedora31/noimt.
See console output.

[phsft-bot]

Build failed on ROOT-performance-centos8-multicore/default.
Running on null:/data/sftnight/workspace/root-pullrequests-build
See console output.

Errors:

• [2021-01-29T11:25:39.269Z] 968/2116 Test cling PR 144 #675: tutorial-math-mathcoreGenVector ...................................................................***Failed Error regular expression found in output. Regex=[: error:] 1.89 sec

Failing tests:

• projectroot.runtutorials.tutorial_math_mathcoreGenVector

[phsft-bot]

Build failed on mac11.0/cxx17.
Running on macphsft20.dyndns.cern.ch:/Users/sftnight/build/workspace/root-pullrequests-build
See console output.

Errors:

• [2021-01-29T11:34:33.103Z] 816/2176 Test cling PR 162 #702: tutorial-math-mathcoreGenVector ...................................................................***Failed Error regular expression found in output. Regex=[: error:] 0.69 sec

Failing tests:

• projectroot.bindings.pyroot.pythonizations.test.pyunittests_pyroot_pyz_rtensor
• projectroot.bindings.pyroot.pythonizations.test.pyunittests_pyroot_pyz_rdataframe_makenumpy
• projectroot.runtutorials.tutorial_math_mathcoreGenVector

[lmoneta]

In the case of VectorUtil we need two different functions:
DeltaR( v1,v2) for Eta
DeltaRapidityPhi (v1,v2) for Rapidity.

The reason is that we can call DeltaR also for 3D objects and if it is implemented in term of rapidity , it will not compile. You can see this from the failing error from the integration builds.

Can you please add this change ?

Thank you for this contribution !
Lorenzo

[mattleblanc]

Hi @lmoneta, that makes sense -- I have just pushed some more changes. Thanks for the review!

MLB

[lmoneta]

Thank you for the changes. They look good now.
It would be nice to add also a test for the new DeltaRapidityPhi function, for example adding few lines of code in
https://github.com/root-project/root/blob/master/tutorials/math/mathcoreGenVector.C#L424

If you don't have time doing this, I can do it later, after merging this PR.

[lmoneta]

@phsft-bot build!

[phsft-bot]

Starting build on ROOT-debian10-i386/cxx14, ROOT-performance-centos8-multicore/default, ROOT-fedora30/cxx14, ROOT-fedora31/noimt, ROOT-ubuntu16/nortcxxmod, mac1014/python3, mac11.0/cxx17, windows10/cxx14
How to customize builds

[phsft-bot]

Build failed on mac11.0/cxx17.
Running on macphsft20.dyndns.cern.ch:/Users/sftnight/build/workspace/root-pullrequests-build
See console output.

Failing tests:

• projectroot.bindings.pyroot.pythonizations.test.pyunittests_pyroot_pyz_rdataframe_makenumpy
• projectroot.bindings.pyroot.pythonizations.test.pyunittests_pyroot_pyz_rtensor

[mattleblanc]

Hi @lmoneta, I don't have time to implement this test at the moment, but I will add it to my list to revisit in the future, in case I get there before you do. Thanks once again for reviewing this PR!

[phsft-bot]

Build failed on windows10/cxx14.
Running on null:C:\build\workspace\root-pullrequests-build
See console output.

[lmoneta]

A test will then be added as a separate PR. Thank you @mattleblanc for this contribution!