HDGeant4 Meeting, February 12, 2019

From GlueXWiki
Revision as of 16:59, 12 February 2019 by Marki (Talk | contribs) (Minutes)

(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search

HDGeant4 Meeting
Tuesday, February 12, 2019
3:00 pm
JLab: CEBAF Center, A110
BlueJeans: 968592007


  1. Review of minutes from the last HDGeant4
  2. Review of minutes from the last Software Meeting
  3. Talks at Collaboration Meeting
  4. Do-over on comparison studies
  5. http://nuclear.gla.ac.uk/~pauli/pres.pdf
  6. Issues on GitHub (all)
  7. Pull Requests on GitHub (all)



  • CMU: Naomi Jarvis
  • JLab: Thomas Britton, Mark Ito (chair), Simon Taylor, Beni Zihlmann
  • UConn: Richard Jones

Today's meeting did not follow the published agenda.

Mark has tagged a version of HDDS that uses the pre-DIRC geometry found in the CCDB. Recall that this version has overlaps removed and includes the Lucite shield in front of the TOF. The new tag is 3.14. Mark will prepare a version set using it for general use.

Thomas has looked at the plots of the recent HDG3 vs 4 comparison simulation and does not see any concerning differences. These were done with most physics processes turned off (no multiple scattering, etc.). He will re-run with all physics processes turned on as a further check. Naomi reported that she had seen HDG4 executing many times more slowly that HDG3 several months ago. Thomas will look at the execution times when he does the running.

Peter Pauli has done a detailed high-level comparison of results from HDG3 and HDG4. He has been on owl shift and could not be at the meeting to present his results. He did send his slides to Mark. Find them here. We will hear from him at a future meeting.

We discussed the nature of the advice Thomas should give at his talk at the Collaboration Meeting vis-a-vis readiness of HDG4 for prime time. We agreed that we can recommend use of HDG4 going forward now; no known problems exist. In particular, tuning of the smearing should now be done against HDG4 so that that effort will transmit to future efforts. Richard made the point that we are now in a situation where, if HDG3 and HDG4 give different results, it may not be the case that HDG3 is "right" and HDG4 is "wrong." For example, there are processes in Geant4 that are better modeled than they were in GEANT.