Choose timezone
Your profile timezone:
Common Wire Tracker Digitisation
→
Europe/Zurich
42/R-032 (CERN)
Common Wire Tracker Digitisation
Quick recap
The meeting focused on reviewing two pull requests related to digitalization work. Andreas discussed Stefano's PR on wire tracker info, highlighting the need to generalize the number of layers per superlayer and addressing breaking changes in the k4RecTracker. For Muhammad's large PR on DCHdigi_v03, Andreas suggested converting it to draft status until the wire tracking tool PR is merged, recommended using existing EDM4hep data types instead of creating new ones, and advised making gas parameters more flexible. The discussion also covered implementation details around cluster size calculation, debug histograms, and the need for better documentation in header files. Andreas mentioned that Muhammad would need to present the new digitizer algorithm to the Digi-Reco conveners in the future, and the conversation ended with a brief discussion about potential future tracking approaches.
Next steps
Andreas
- Check that Straw Tube Simulation is unchanged by Stefano's PR
- Provide scripts and workflow command to Stefano to test DCHdigi_v02 with new WireTracker_info
- Look up how to handle root files (upload to FCC website and download via CMake) and inform Muhammad
- Start email thread with Andrea De Vita and Juan Miguel regarding root file handling approach
- Continue reviewing Muhammad's PR and provide detailed comments on code structure and potential further function separation
- Discuss with Digi-Reco conveners regarding Muhammad's presentation scheduling
Muhammad
- Mark PR as draft as soon as possible (convert to draft in GitHub)
- Refactor cluster position calculation to use existing EDM4hep SIM tracker hit data types instead of creating new EDM4hep data types
- Make average number of electron-ion pairs per cluster a Gaudi property (double value) instead of relying only on tabulated values
- Update comments in code to clarify ionization energy parameters
- Implement gas mixture selection options (at minimum include argon in addition to helium) with Gaudi property to switch between gas mixtures
- Ensure debug histograms are filled only with values that are actually used in the algorithm
- Add more documentation in header file explaining what each code section does
- Prepare presentation on new digitizer for DG Reco meeting (to be scheduled in coming weeks, not immediate)
Stefano
- Change wire tracker info to support different number of layers per superlayer (make nlayers_per_superlayer a vector instead of double) for both straw tube tracker and drift chamber
- Change layer info to use N_cells instead of N_wires to accommodate both drift chamber and straw tube tracker
- Fix rotation issue in PR by creating sister PR in K4Geo to address breaking changes in K4FTracker
- Update drift chamber geometry changes as separate PR after wire tracker info PR is merged
Summary
Pull Request Review Discussion
Andreas discussed two pull requests (PRs) with the team. For Stefano's PR on the wire tracker info, Andreas identified several issues including the need to allow different numbers of layers per superlayer and potential breaking changes in k4RecTracker that would require separate fixes. For Muhammad's large PR, Andreas recommended marking it as draft until after Stefano's wire tracking tool PR is merged, as it would be easier to adapt the drift chamber digitizer version 3 afterward.
Cluster Position Calculation Development Discussion
Andreas and Muhammad discussed the development of a cluster position calculation as a separate Gaudi functional independent of DCHdigi_v03. Muhammad had already started working on this and presented his progress, including creating new EDM4heo data types for cluster distribution. Andreas advised against creating new EDM data types and suggested using existing EDM4hep data types instead, specifically recommending that the algorithm should work with SIM tracker hits as input and output. They agreed to refine the approach, with Andreas suggesting that cluster-related information could be handled in the digitizer rather than as separate data types.
Digitization Code Updates Discussion
Muhammad and Andreas discussed updates to a digitization code related to ionization energy and cluster calculations. DG suggested making the average number of electron-ion pairs per cluster a Gaudi property rather than relying solely on tabulated values, to allow flexibility for different gas mixtures. They also addressed inconsistencies in the code regarding the calculation of cluster numbers, with Muhammad agreeing to streamline the approach by merging the two methods into one.
Digitizer Code PR Updates Discussion
Andreas and Muhammad discussed updates to Muhammad's digitizer code PR. They agreed to use a zero-truncated Poissonian distribution to avoid zero clusters and to implement gas mixture options beyond helium, including argon. Andreas informed Muhammad that he would need to present the digitizer to the FCC Physics Software and Computing DIGI-RECO Group before the PR is merged. They also discussed adding more documentation to the header file and handling root files appropriately. Andreas mentioned he would review the PR further and provide more specific feedback on code organization and structure.
There are minutes attached to this event.
Show them.