LArSoft Coordination Meeting

US/Central
FNAL

FNAL

Erica Snider (Fermilab), Saba Sehrish (Fermilab)
Description

To connect via Zoom:  Meeting ID 831-443-820

(See instructions for setting Zoom default to join a meeting with audio and video off: https://larsoft.org/zoom-info)

PC, Mac, Linux, iOS, Android:  https://fnal.zoom.us/j/831443820

Phone:

H.323:

162.255.37.11 (US West)
162.255.36.11 (US East)
213.19.144.110 (EMEA)
See https://fnal.zoom.us/ for more information
 

At Fermilab:  WH7X (Racetrack)

Minutes for LArSoft coordination meeting on July 2, 2019

Present | Lynn Garren, Saba Sehrish, Erica Snider, Hans Wenzel|

Remote | Katherine Lato, David Rivera, Brett Viren|

Release and Project status report [Erica Snider]

  • LArSoft workshop debrief:
    • LArSoft workshop held on June 24-25, presentations were great, all the material is available (https://indico.fnal.gov/event/20453/other-view?view=standard)
  • Releases
    • Last week 08_23_00 on June 24th; it has TrajCluster3D updates and removal of deprecated modules (see Issue #20618)
    • Released v08_24_00 on June 27 with Pandora updates
    • Released v08_05_00_09 with Patch for MicroBooNE MCC9 production
  • We propose to skip this weeks integration release, there are no requests for this week
  • Genie v3 reminder: test release available since June 17, need experiment sign-of in order to complete migration

Conclusion

  • no release this week.

Migrating to the refactored larg4 [David Rivera]

  • Update on the status of migrating protodune simulation to re-factored larg4
  • PDSP (Protodune single phase)
  • Simulation chain was explain, see slides
  • There are key differences between the legacy code and re-factored larg4; one uses nutools, and use ConfigurablePhysicsList. While the re-factored larg4 is built on artg4t4, and you have reference lists and extensions.
  • There is a difference between optical physics as well; legacy has optical simulation code taken from Geant4 and adapted. Only one scintillation material is allowed. Used tricks to fake multiple materials (.e.g. adjusting LAr density). Model is essentially 20 yrs old.
    • Refactored uses updated OpticalPhysics in G4. Properties are attached to the materials, so arbitrary number is possible
  • Reference to Hans slides with details on LArG4 re-factoring: (https://indico.fnal.gov/event/18681/session/6/contribution/61/material/slides/0.pdf)
  • Goal for protodune is the customization of the physics list tailored to the interest of the physics under investigation * Low Energy Physics: * Solar Neutrinos * Neutron capture * Shielding
  • Another feature of the refactored larg4 is to be able to visualize better and debug geometry errors
  • Hans provided example fo refactored 3x1x1 DP
    • LArSoft feature #22466. Started with that
    • Example wiki is also provided
  • ToDO List: ignored the optical part of the simulation at this point for simplicity
    • Declared LAr volumes as charge sensitive detectors
      • pointed to relevant gdml files
      • Noted that he did no need to change much of the gdml from the exmple. Only a few lines, which was shown in slides. Mainly the sensitive volume, and if needed the step size and E-field
    • Redefined the protoDUNE services in the same spirit as the example
    • Created corresponding G4—> reco fhicl files
    • Also modified protoDUNE event display fhicls
  • Showed G4 display of geometry
    • One new feature is that you can add display colors to the materials in the gdml for use in G4 display
    • Good to be able to use that display to debug the geom
  • Event display of 6 GeV muon
  • Migration example can be found in larg4 wiki: ProtoDUNE Example wiki
  • Currently maintained in a feature branch of dunetpc:
    • feature/drivera_refactored_larg4_PDSP
    • Based on LArSoft v08_22_00
  • To do
    • Ignored optical part at first. Just wanted to check that charge part was right. So need to define optical props of the relevant materials in the geom file
    • Consider ways to provide various physical properties as configuration params for G4 stage
      • e.g., the e-field is hard-coded in the geom file. Would like that to be settable from fhicl configuration or some other means
      • Q: Was there a difference of opinion w/in PDSP as to whether this was needed?
        • A: In general, PDSP wants to have more parameters available via fhicl. No differences of opinion on this
        • Probably the most important at this point: E field and Rayleigh scattering length (the value of which is still under debate)
    • Purge refactored services
    • Continue validation process
    • Compare resource usage between the new the legacy frameworks
  • Noted benefits of migrating
    • Process of adding or customizing physics lists in the legacy version is really painful and convoluted and prone to breaking changes. Showed some examples developed in the new framework
  • Impressions
    • Low barrier of entry concerning the geometry
      • Modifying the gdml file for LArTPC expts, as a minimum case, only requires specifying the E field and declaring the active TPC volume as a SensitiveDetector
      • Straight-forward to define a separate set of fhicl files that takes advantage of the refactored LArG4
        • g4 stage: only need to become familiar w Refactored Physics Constructor (artg4tk)
        • post-g4 stages: only need to override the SimChannelLabel to match the one for the electDrift Module (or drift module of choice)
      • At a glance, the physics make a lot more sense
        • Ref phys lists are widely used and are supported by the geant collab
        • Saw improvements in results of certain processes where there were things that didn’t make sense in the legacy sim
  • Support from LArSoft team appreciated, and hope that other experiments will follow to migrate to the re-factored larg4.
    • Have been in close contact w Hans and Paul regarding issues of enabling fhicl parameters
    • Noted that Alex Himmel wants DUNE to switch over entirely (maybe) after TDR process
    • Also heard that SBND wants to migrate

Conclusion

  • Progress so far looks good and promising for the re-factored larg4 migration
There are minutes attached to this event. Show them.