Reconstruction Group Meeting

Monday 20 Apr 2026, 11:00 → 12:00 US/Eastern

Zoom coordinates:

• URL: https://lbnl.zoom.us/j/94910574347?pwd=c01YQlg5bG04dFR6UWs3NjZkS3ZSdz09
• Meeting ID: 949 1057 4347
• Password: recon

Meeting Notes

• See the google doc

11:00 → 11:05 Introduction

Speakers: Chandradoy Chatterjee (INFN Trieste), Dr Derek Anderson (Thomas Jefferson National Accelerator Facility), Shujie Li (Lawrence Berkeley National Laboratory)

11:05 → 11:30 Secondary Vertexing: AMVF Next Steps

Speakers: Rongrong Ma (Brookhaven National Laboratory), Dr Shyam Kumar (University and INFN Bari), Shyam Kumar (Universita e INFN, Bari (IT)), Xin Dong (Lawrence Berkeley National Laboratory)

AMVF PR

Next (Final) Steps:

• Replace ActsExamples::Trajectories with underlying storage containers (c.f. EICrecon#2311)
  • Update: Dima took care of it!
• Add description to algorithm CTOR
• Add units where needed in config
• Further improvements/concerns:
  • Deduplicating code for primary & secondary vertexing mode
  • Kaon mass peak not visible in last iteration, but is visible in helix swimmer... Not a showstopper, but would need further investigation. Better to catch now (esp. since kaon mass peak should be benchmark)

 

Action Item:

• Derek sends email to Xin, Dima, Rongrong, and Shyam to discuss workforce.
  • Especially useful for somebody to look into (potentially) missing kaon mass peak!
  • Validating performance useful for HF studies

11:30 → 11:55 Secondary Vertexing: Secondary Vertex EDM Next Steps

Speaker: Xin Dong (Lawrence Berkeley National Laboratory)

EDM PR

Next Steps:

• Suggestions:
  • Add the covariance matrix for the daughter 3-momenta
    • Action item: Xin will confirm with Shyam what would be useful in terms of errors.
  • Utilize existing types for some info (e.g. edm4eic::Vertex for vertex-specific info, edm4eic::ReconstructedParticle for daughters w/ momenta evaluated wrt secondary vertex) and use new type for new info
    • Action item: Derek will sketch out some ideas in the PR comments
  • Prefer relations to storing specific indices
    • Indices indicate which pair of particles the DCA is being calculated b/n. E.g. in a 3-body decay P --> A+B+C you would have entries for (A, B), (B, C), and (A, C)
    • Could possibly replace indices with additional oneToMany relations + helper methods, but needs some thought...
  • Potentially omit relation to primary vertex, or rename field to "parentVertex" or something similar
    • Decision: we'll keep primaryVertex as primaryVertex
• After resolving discussion:
  • Present type in weekly S&C meeting

11:55 → 12:00 AOB

Speakers: Chandradoy Chatterjee (INFN Trieste), Dr Derek Anderson (Thomas Jefferson National Accelerator Facility), Shujie Li (Lawrence Berkeley National Laboratory)