Difference between revisions of "HDGeant4 Meeting, January 29, 2019"

From GlueXWiki
Jump to: navigation, search
(Agenda)
(add the minutes)
 
Line 13: Line 13:
 
# [https://github.com/JeffersonLab/HDGeant4/issues Issues on GitHub] (all)
 
# [https://github.com/JeffersonLab/HDGeant4/issues Issues on GitHub] (all)
 
# [https://github.com/JeffersonLab/HDGeant4/pulls Pull Requests on GitHub] (all)
 
# [https://github.com/JeffersonLab/HDGeant4/pulls Pull Requests on GitHub] (all)
 +
 +
== Minutes ==
 +
 +
Present:
 +
* ''' JLab: ''' Alex Austregesilo, Thomas Britton, Sean Dobbs, Mark Ito (chair), Simon Taylor, Beni Zihlmann
 +
* ''' UConn: ''' Richard Jones
 +
 +
There is a [https://bluejeans.com/s/LTqwb/ recording of this meeting] on the BlueJeans site. Use your JLab credentials.
 +
 +
=== Review of minutes from last time ===
 +
 +
We went over [[HDGeant4_Meeting,_January_15,_2019#Minutes|the minutes from January 15]] with no new issues raised.
 +
 +
=== How to run hdgeant without the DIRC ===
 +
 +
Richard led us through [https://mailman.jlab.org/pipermail/halld-offline/2019-January/003504.html Richard's email], sent earlier today, which addressed two points:
 +
# How to run pre-DIRC geometry, when the DIRC is present in the master branch of hdds
 +
#* Richard has embellished David Lawrence's mechanism whereby hdgeant can generate its HDDS FORTRAN code and corresponding libraries on the fly. The details are in [[HOWTO run hdgeant simulations using ccdb geometry|his new HOWTO]].
 +
#* We discussed the utility of having HDDS XML files in the source code repository at all. Richard stated that he finds the version stored there useful for development.
 +
#* We also discussed the need to built legacy (pre-DIRC) versions of HDDS on disk for general use. We thought that that might be more trouble than they are worth. [Added in press: second thoughts on this have been had by the secretary.]
 +
# How to run with the DIRC present, but not suffer the overhead of following CKOV photons
 +
#* CKOV turns of Cerenkov photons for all detectors, i. e., globally. Its historical default is CKOV=1 or on. We did not pay the computational price since no detector in the default geometry had an index of refraction defined, a necessary condition for optical photon tracking.
 +
#* Turning off optical photon tracking in the DIRC should speed things up by a factor of several.
 +
 +
=== Do-over on comparison studies ===
 +
 +
Thomas presented the first plots from his effort at repeated the studies we did when we first starting having HDGeant4 meetings. Find the plots [[Media:Pim_1_70.pdf|here]]. They are all for 1&nbsp;GeV/c &pi;<sup>&minus;</sup>'s at 70 degrees in &theta; and 0 degrees in &phi;, and they all show the results from HDG3 and HDG4 side-by-side (or top-and-bottom, as the case may be). The plots and accompanying discussion appear starting at 23:00 in the recording.
 +
 +
A more careful look needs to be taken. More particle settings are to come. [Added in press: some ROOT files have been posted by Thomas:
 +
 +
/w/halld-scifs17exp/halld2/home/tbritton/groupMC/G3G4Pure_v2/G3_pim_1_70.root
 +
/w/halld-scifs17exp/halld2/home/tbritton/groupMC/G3G4Pure_v2/G4_pim_1_70.root
 +
 +
File names have the form G[version number]_[particle]_[p GeV]_[theta deg].root]
 +
 +
=== Note on Upstream Bug Fixes ===
 +
 +
Richard reported that he has passed the information on bugs he has found in Geant4 up to the Geant4 team. He hopes to eventually be able to operate without the "G4fixes" part of the HDGeant4 distribution.
 +
 +
=== Issues and Pull Requests on GitHub ===
 +
 +
There weren't any to review.

Latest revision as of 14:30, 30 January 2019

HDGeant4 Meeting
Tuesday, January 29, 2019
3:00 pm
JLab: CEBAF Center, A110
BlueJeans: 968592007

Agenda

  1. Review of minutes from last time
  2. how to run hdgeant without the dirc (Richard)
  3. Do-over on comparison studies
    1. pim 1 GeV 70 degree
  4. Issues on GitHub (all)
  5. Pull Requests on GitHub (all)

Minutes

Present:

  • JLab: Alex Austregesilo, Thomas Britton, Sean Dobbs, Mark Ito (chair), Simon Taylor, Beni Zihlmann
  • UConn: Richard Jones

There is a recording of this meeting on the BlueJeans site. Use your JLab credentials.

Review of minutes from last time

We went over the minutes from January 15 with no new issues raised.

How to run hdgeant without the DIRC

Richard led us through Richard's email, sent earlier today, which addressed two points:

  1. How to run pre-DIRC geometry, when the DIRC is present in the master branch of hdds
    • Richard has embellished David Lawrence's mechanism whereby hdgeant can generate its HDDS FORTRAN code and corresponding libraries on the fly. The details are in his new HOWTO.
    • We discussed the utility of having HDDS XML files in the source code repository at all. Richard stated that he finds the version stored there useful for development.
    • We also discussed the need to built legacy (pre-DIRC) versions of HDDS on disk for general use. We thought that that might be more trouble than they are worth. [Added in press: second thoughts on this have been had by the secretary.]
  2. How to run with the DIRC present, but not suffer the overhead of following CKOV photons
    • CKOV turns of Cerenkov photons for all detectors, i. e., globally. Its historical default is CKOV=1 or on. We did not pay the computational price since no detector in the default geometry had an index of refraction defined, a necessary condition for optical photon tracking.
    • Turning off optical photon tracking in the DIRC should speed things up by a factor of several.

Do-over on comparison studies

Thomas presented the first plots from his effort at repeated the studies we did when we first starting having HDGeant4 meetings. Find the plots here. They are all for 1 GeV/c π's at 70 degrees in θ and 0 degrees in φ, and they all show the results from HDG3 and HDG4 side-by-side (or top-and-bottom, as the case may be). The plots and accompanying discussion appear starting at 23:00 in the recording.

A more careful look needs to be taken. More particle settings are to come. [Added in press: some ROOT files have been posted by Thomas:

/w/halld-scifs17exp/halld2/home/tbritton/groupMC/G3G4Pure_v2/G3_pim_1_70.root
/w/halld-scifs17exp/halld2/home/tbritton/groupMC/G3G4Pure_v2/G4_pim_1_70.root

File names have the form G[version number]_[particle]_[p GeV]_[theta deg].root]

Note on Upstream Bug Fixes

Richard reported that he has passed the information on bugs he has found in Geant4 up to the Geant4 team. He hopes to eventually be able to operate without the "G4fixes" part of the HDGeant4 distribution.

Issues and Pull Requests on GitHub

There weren't any to review.