GlueX Offline Meeting, August 24, 2011

From GlueXWiki
Jump to: navigation, search

GlueX Offline Software Meeting
Wednesday, August 24, 2011
1:30 pm EDT
JLab: CEBAF Center F326


See Curtis's email for some background.

  1. Announcements
  2. Review of minutes from the last meeting: all
  3. Proposal to shrink the DST Data Format size (Richard/Discussion).
  4. Tracking chi-squared/error update: Paul Mattione.
  5. Tracking dE/dx update.
  6. FCal reconstruction update.
  7. BCAL reconstruction 8/24/11
  8. FDC Changes in GSIM[1] and restoration of default behavior [2].
    • Default behavior (same as old): only the time at the doca is reported (no explicit cluster generation).
    • Drift times and diffusion computed based on results of Garfield simulations for nominal gas mixtures for FDC and CDC.
  9. Treatment of magnetic field maps in the CCDB: Dmitry
  10. Action Item Review: all
  11. Review of recent repository activity: all

Communication Information


To connect by telephone:

  1. dial:
  2. enter access code followed by the # sign: 3421244#

Video Conferencing


Talks can be deposited in the directory /group/halld/www/halldweb/html/talks/2011-3Q on the JLab CUE. This directory is accessible from the web at . You have to be a member of the "halld" Unix group to do this.



CMU: Will Levine, Paul Mattione IU: Kei Moriya, Matt Shepherd JLab: David Lawrence, Mark Ito (chair), Dmitry Romanov, Sascha Somov, Simon Taylor, Elliott Wolin, Beni Zihlmann UConn: Richard Jones


Elliott will update the "DANA-EVIO stuff" now that the reconstruction class re-arrangement has settled down.

Review of minutes from the last meeting

We reviewed the minutes from the August 10 meeting.

  • Elliott wondered if using an ORM was overkill for the online since the queries are so simple.
  • Paul asked for collective wisdom about deleting DPhoton from the trunk. He will delete it.

Proposal to shrink the DST data format size

Richard objected to the title of this agenda item. Nonetheless...

He walked us through the recently circulated spreadsheet detailing the contents of the proposed DST and the anticipated savings in space for each reconstructed C++ class. The idea is to save enough information to reconstitute the targeted DANA classes minimizing the information while keeping the classes useful for data analysis.

Some random notes:

  • The particle list cuts only di-quarks from those generated by PYTHIA
  • A compression scheme is possible and there were several proposals on how to do it.
  • The exact format of the serialized data is yet to be determined.

Tracking chi-squared/error update

Paul led a discussion of his wiki page. We discussed his proposal, to have a formal framework for calculating and storing hit-by-hit measurement errors and then propagate them forward to get sensible errors on particle kinematics. Several of us thought that concentrating on high-level kinematic quantities (e. g. energy of a cluster, angle of a track), post-reconstruction might be more fruitful.

Tracking dE/dx update

Simon led us through his wiki page. Executive summary:

  • units problem
  • fixed now

FCAL reconstruction update

Beni had previously reported a surplus of split clusters in the FCAL, but on closer examination they turned out to be from photon conversions upstream of the FCAL producing separate, legitimate clusters. The reconstruction is behaving as expected.

BCAL reconstruction update

Will led us through his wiki page describing recent work understanding photons detected in the BCAL. His summary is reproduced here:

  • BCAL reconstruction seemed to be an obstacle in b1pi events
  • Two reconstruction algorithms in tree: KLOE algorithm=default, "new" algorithm by matt s.
  • Found two issues which lead to poor energy resolution at low E (fixes not checked in yet)
  • "New" algorithm seems to be doing well for single-photon events
  • Further study of b1pi, etc.
  • Further fine-tuning needed, but maybe should wait until BCAL segmentation finalized?

Note that this is work in progress.

FDC Changes in GSIM and restoration of default behavior

The title of this section says it all. The links in the title reference recent emails describing Simon's recent activity in the named areas. His summary (quoted from the agenda):

  • Default behavior (same as old): only the time at the doca is reported (no explicit cluster generation).
  • Drift times and diffusion computed based on results of Garfield simulations for nominal gas mixtures for FDC and CDC.

CCDB-related questions

Dmitry posed some questions to the group on how to proceed with some problem corners with the CCDB. See his talk for details. There were three main areas covered. All have to do with the conversion from the current flat-file-based calibration tree to the CCDB.

  • "calibration constants" that are in fact a list of reconstruction parameters
    • We decided that these should be converted to JANA parameters.
  • non-uniform comments in the current files (used to identify column names)
    • We decided that Dmitry should feel free to impose a uniform convention
  • very large calibration sets, in particular magnetic field maps
    • This will require more discussion.

New Action Item

  • Decide how to handle large calibration sets.