[Halld-offline] Minutes, GlueX Offline Software Meeting, May 18, 2011

[Mark M. Ito]

Find the minutes at 
http://www.jlab.org/Hall-D/software/wiki/index.php/GlueX_Offline_Meeting%2C_May_18%2C_2011#Minutes 
and below.

   ==

GlueX Offline Meeting, May 18, 2011
Minutes

    Present: Mark Ito (chair), David Lawrence, Paul Mattione, Kei Moriya,
    Nathan Sparks, Simon Taylor, Beni Zihlmann

Track Reconstruction Problems in Cascade Events

      * David: some tracks exhibit trajectories that curl tighter as they
        travel downstream.
      * We still need work on track candidate finding.
      * David has worked on a new hit selector that may help.
      * Mark speculated that curling tracks may not be one of the things we
        can do well initially.


Documentation on Getting Started

      * Beni did a major rework of the Offline Software wiki page. Many
        favorable comments were offered.
      * Paul has been working from the wiki documentation and finds it
        "pretty straightforward", to the surprise of many.


"b1pi" Problem

      * David pointed Paul to the howto on generating and reconstructing
        b1pi events. They discovered in the process that the reconstruction
        efficiency is not what it used to be: down from about 50% to about
        5%.
      * This needs to get attention.


Bad Hits and Kalman filter

    Simon got some questions on the sensitivity of the Kalman filter to bad
    hits on the track.
      * Bad hits can throw the Kalman filter off. The ALT1 is more robust
        against inclusion of anomalous hits.
      * Hits at the outer parts of the detector cause a particular problem
        since information about the track is refined from the outside in. A
        bad hit on the outside has more influence on the trajectory at the
        time it is added and can cause the fit to fail.
      * There are various options to reduce this sensitivity:
           + "annealing" algorithm
           + rejection of poor hits as they are considered for addition to
             the fit


Definition of Primary Particles

      * We will get a talk from Beni on this subject. It became an issue
        when he introduced changes to do studies correlating hits with
        generated particles.
      * Some changes have been made recently. They should help Matt
        Shepherd resolve his problem he was having with identifying primary
        photons.


IT Review

    Mark reminded us that there will be a review of "Information Technology
    at JLab in the 12 GeV Era" tomorrow. There is a [37]webpage for the
    review which contains a link to the [38]charge to the committee. Mark
    will present the view from Hall D.

    David asked how any comments/findings/recommendations of the review
    will impact our plans. Mark said that that remains to be seen, but we
    should probably take any feedback from the committee as an opportunity
    to look hard at what we are doing.

Figure Repository

    Kei wondered where we are with that. Zisis Papandreou announced a plan
    at the Collaboration meeting. We should track progress.

New Action Items

     1. Investigate problem reconstructing "b1pi" events.
     2. Give talk on the definition of primary and secondary particles in
        HDGeant. -> Beni

    Retrieved from
    
"http://www.jlab.org/Hall-D/software/wiki/index.php/GlueX_Offline_Meeting%2C_May_18%2C_2011"

References

   37. http://conferences.jlab.org/IT12GeV/
   38. http://conferences.jlab.org/IT12GeV/12GeVEraITReviewFinal.pdf