GlueX Offline Meeting, November 30, 2011
GlueX Offline Software Meeting
Wednesday, November 30, 2011
1:30 pm EDT
JLab: CEBAF Center F326
- Review of minutes from the last meeting: all
- Reconstruction sub-group reports
- Calorimeters: Mark for Matt
- Tracking: Simon
- PID: DC dE/dx Update Paul M.
- Too many photons: Beni
- Action Item Review
- Review of recent repository activity: all
- ESNet: 8542553
- Meeting URL: http://evo.caltech.edu/evoNext/koala.jnlp?meeting=ete8eavBvvanaaIiaiIt
- Phone Bridge ID: 11 6471
Talks can be deposited in the directory
/group/halld/www/halldweb/html/talks/2011-4Q on the JLab CUE. This directory is accessible from the web at https://halldweb.jlab.org/talks/2011-4Q/ . You have to be a member of the "halld" Unix group to put files in the directory.
- CMU: Will Levine, Paul Mattione, Curtis Meyer
- FSU: Nathan Sparks, Beni Zihlmann
- JLab: Mark Ito (chair), David Lawrence, Yi Qiang, Dmitry Romanov, Simon Taylor
- David told us that that decision on the segmentation of the BCAL read-out should be made soon. After that we should put the decided-upon segmentation into the reconstruction. Until now we have been running with "fine" segmentation. Smearing code will have to be updated.
- David also commented that the FCAL beam test gives us an opportunity to develop code to read in EVIO format into the JANA framework.
- Dmitry on the CCDB:
- He mentioned that Python 2.7 is now installed on the ifarm machines, a prerequisite for the CCDB shell. He will start work on installing a version that works there, for public consumption.
- He noted that the difficulties noted last time with table column names have been finessed.
Reconstruction sub-group reports
Matt Shepherd reported via email:
[With regard to] Beni's email regarding high increase in photons, I know Will had been checking in some changes to the BCAL code; however, I thought that all of his changes were in my reconstruction algorithm, which, as I understand it, is not the default. As far as I know, no one has made any changes to the FCAL reconstruction.
Will mentioned that he is putting together a report/summary of his changes to the BCAL code.
Simon reported an increase in the number of reconstructed π+'s, π-'s, and protons by 10-20% due to recent changes.
Paul had continued work on the particle ID using drift chamber dE/dx. See his wiki page for details. He has essentially calibrated the Monte Carlo dE/dx responses, measuring the mean and variance in dE/dx and uses those measurements in forming a χ2 statistic. He gets dramatically improved confidence level distributions as a result.
Too many photons
Last Monday Beni reported a change in the semi-weekly b1pi reconstruction job where the number of photons reconstructed increased by about 20%. The increase comes mostly in the forward region. The change occurred between 14th and 21st of November.
There was some speculation about what might have occurred, but no firm conclusions about the cause. David pointed out that changes in charged particle tracking can affect the photon yield but the increase in reconstruction efficiency that Simon reported earlier naively goes the other way; better tracking would identify more clusters as "charged" and thereby reduce the number of photons. More work is needed.
Mark mentioned that if we had suite of tests that looked at things at a lower level, these issues would be easier to understand. Relevant in this case would be, for example, the raw number of clusters, both "neutral" and "charged" or even the raw hit multiplicity and energy distributions. Developing a set of tests like this is a possible project for this working group.