feat: particle killing 4 Geant4 and refactor user actions

[benjaminhuth]

Out Geant4 algorithm allows to pass std::vectors of user actions. However, multiple actions per type seem not to be supported by Geant4. This changes this to allow only one user action per type (run, event, particle, step).

I'm not sure if this was possible in former Geant4 versions, there is also a comment in the Geant4 source code that indicates it is possible to add multiple actions per type. However, this is not the case.

[codecov]

Codecov Report

Merging #2057 (7cda786) into main (2a90742) will not change coverage.
The diff coverage is n/a.

@@           Coverage Diff           @@
##             main    #2057   +/-   ##
=======================================
  Coverage   49.37%   49.37%           
=======================================
  Files         427      427           
  Lines       24779    24779           
  Branches    11430    11430           
=======================================
  Hits        12234    12234           
  Misses       4484     4484           
  Partials     8061     8061           

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

[asalzburger]

Makes totally sense, thanks, @benjaminhuth

[asalzburger]

Out Geant4 algorithm allows to pass std::vectors of user actions. However, multiple actions per type seem not to be supported by Geant4. This changes this to allow only one user action per type (run, event, particle, step).

I'm not sure if this was possible in former Geant4 versions, there is also a comment in the Geant4 source code that indicates it is possible to add multiple actions per type. However, this is not the case.

As of your question, I think this comes from the singleton usage/desing of Geant4 in the past. But as of the chances you did, they are all in line what Geant4 expects to do - and were just missing on our side.

[asalzburger]

Out Geant4 algorithm allows to pass std::vectors of user actions. However, multiple actions per type seem not to be supported by Geant4. This changes this to allow only one user action per type (run, event, particle, step).
I'm not sure if this was possible in former Geant4 versions, there is also a comment in the Geant4 source code that indicates it is possible to add multiple actions per type. However, this is not the case.

As of your question, I think this comes from the singleton usage/desing of Geant4 in the past. But as of the chances you did, they are all in line what Geant4 expects to do - and were just missing on our side.

Oh - I misread this - I will check, I was under the impression that you can attach different UserActions, just not the the same one twice, I will check that.

[github-actions]

📊 Physics performance monitoring for 7cda786

Full report
Seeding: seeded, truth estimated, orthogonal
CKF: seeded, truth smeared, truth estimated, orthogonal
IVF: seeded, truth smeared, truth estimated, orthogonal
Ambiguity resolution: seeded, orthogonal
Truth tracking
Truth tracking (GSF)

Vertexing

Vertexing vs. mu

IVF seeded

IVF truth_smeared

IVF truth_estimated

IVF orthogonal

Seeding

Seeding seeded

Seeding truth_estimated

Seeding orthogonal

CKF

CKF seeded

CKF truth_smeared

CKF truth_estimated

CKF orthogonal

Ambiguity resolution

seeded

Truth tracking (Kalman Filter)

Truth tracking

Truth tracking (GSF)

Truth tracking

[andiwand]

nice catch - well done! left a few questions but looks good in general to me

[benjaminhuth]

@asalzburger I implemented the particle killing now with Acts::Volume, it feels much better now actually. Regarding other actions, I would like to wait until we need a similar list implementation if its ok for you. Is it good to go now from your perspective? (will need to update hashes, so formal approval is not yet required)

I also directly included the particle killing into full_chain_odd.py, so I also tag @andiwand as well. Is this okay or should I remove it again? I would like to execute this at least once in the CI.

[andiwand]

I also directly included the particle killing into full_chain_odd.py, so I also tag @andiwand as well. Is this okay or should I remove it again? I would like to execute this at least once in the CI.

I wonder if we could have a more dedicated test for that. Or will the full chain actually break if you remove it?

I only worry that we might silently remove rouge particles which should not be there in the first place and we will catch the error at a different place later in time