Difference between revisions of "August 2, 2017, Production & Analysis Working Group"

From GlueXWiki
Jump to: navigation, search
(This Week's Topics)
(Upcoming Study Schedule)
 
(13 intermediate revisions by 5 users not shown)
Line 34: Line 34:
 
# Processing Update --- Alex Austregesilo
 
# Processing Update --- Alex Austregesilo
 
#* Analysis Launch --- Paul Mattione
 
#* Analysis Launch --- Paul Mattione
 +
#** [[File:analysis_library_comparison.png|150px|thumb|left|Memory Profile]]<br style="clear:both;"/>
 
#** [https://halldweb.jlab.org/wiki-private/index.php/Spring_2017_Analysis_Launch Proposal for ReactionFilter channels]
 
#** [https://halldweb.jlab.org/wiki-private/index.php/Spring_2017_Analysis_Launch Proposal for ReactionFilter channels]
 
#** [https://halldweb.jlab.org/wiki/index.php/Spring_2017_Analysis_Launch_Cuts Analysis Launch Cuts]
 
#** [https://halldweb.jlab.org/wiki/index.php/Spring_2017_Analysis_Launch_Cuts Analysis Launch Cuts]
 +
#** Launch started 08/01/2017, 12 threads per job, 9gb ram, up to 500 jobs in parallel, already at 20%
 
# Calibration Update --- Sean Dobbs
 
# Calibration Update --- Sean Dobbs
 
# Simulations Update --- Sean Dobbs
 
# Simulations Update --- Sean Dobbs
Line 43: Line 45:
  
 
=== Analysis Focus: Total Cross Sections, Acceptance Corrections ===
 
=== Analysis Focus: Total Cross Sections, Acceptance Corrections ===
# J/&Psi;, &phi; --- Lubomir
+
# J/&Psi;, &phi;[http://www.jlab.org/Hall-D/detector/fdc/reports/jp/Jpsi_anal_aug2_2017.pdf Update] --- Lubomir
 
# &eta;, &rho;, &omega;, &phi; --- Simon
 
# &eta;, &rho;, &omega;, &phi; --- Simon
 
# &phi; --- Thomas
 
# &phi; --- Thomas
Line 53: Line 55:
 
# Flux --- Justin
 
# Flux --- Justin
 
# Beam energy --- Beamline Group
 
# Beam energy --- Beamline Group
# Polarization (TPOL & lineshape) --- Beamline Group
+
# [https://userweb.jlab.org/~dugger/triPol/eBeamPolStudyv2.pdf Dependence of polarization on energy cut] --Mike D
 
# Beam Asymmetries --- Alex Austregesilo
 
# Beam Asymmetries --- Alex Austregesilo
 +
#* [[File:PSigma_vs_run.png|150px|thumb|left|P Sigma]] [[File:psi0_vs_run.png|150px|thumb|left|Phase]]<br style="clear:both;"/>
  
 
= Upcoming Study Schedule =
 
= Upcoming Study Schedule =
Line 87: Line 90:
 
# Channel/Analysis Studies: Branching ratios, cross sections, SDMEs, beam asymmetries
 
# Channel/Analysis Studies: Branching ratios, cross sections, SDMEs, beam asymmetries
 
# Other reconstruction/analysis issues
 
# Other reconstruction/analysis issues
 +
 +
== Simulation Reports ==
 +
 +
15 Minute summaries of progress and future plans on the [[2017 mcsmear Workplan]].
 +
 +
# August 9th
 +
## BCAL - Mark Dalton
 +
## CDC - Naomi Jarvis
 +
## SC - Mahmoud Kamel
 +
# August 17th
 +
## FCAL - Matt Shepherd
 +
## FDC - Lubomir Pentchev / Alex Austregesilo
 +
## TOF - Beni Zhilmann
 +
 +
= Minutes =
 +
* Analysis launch started Tuesday, analyzing 60 channels, and should finish in a few days (~500 jobs running at once):
 +
** It's using the new library, which has not been merged into the master because it's not 100% functional yet (e.g. no detached vertices yet).  It has been tagged though: [https://github.com/JeffersonLab/sim-recon/releases/tag/analysis-2017_01-ver01-batch01 analysis-2017_01-ver01-batch01]
 +
** It's using a set of built-in cuts that are applied to ALL channels: [https://halldweb.jlab.org/wiki/index.php/Spring_2017_Analysis_Launch_Cuts Analysis Launch Cuts]
 +
** Running with 12 threads, it's 4x faster than the old library (30 min vs 2hr) and uses 1/2 the RAM (9 vs 18 GB)
 +
** With tighter cuts than before, the trees are 27% of the size of the REST files (vs. 62% with old cuts), and is expected to take 32 TB of space.
 +
 +
* Lubomir is trying to understand why the background is different for his e+/- analysis between the 2016 and 2017 data.  The response from the kinematic fit looks like it may be different for the background, although it's uncertain why this might be.  He's still looking into it.
 +
 +
* Mike Dugger showed that the extracted values of the polarization depend heavily on the width of the &Delta;E matching cut between the tagger and the PS, at least for narrow-width cuts. The same narrow cut yields a different polarization for the different PARA/PERP/45/135/AMO settings. However, everything looks consistent with wide cuts, or by not cutting at all, so that's what he's doing for now.
 +
 +
* Alex Austregesilo showed the extracted P*Sigma from the Spring 2017 data as a function of run # for the different polarization orientations, showing that they are relatively stable and in agreement with each other.

Latest revision as of 14:38, 3 August 2017

Meeting Time and Place

The meeting will be on Wednesday August 2, 2017 at 2:00 pm EDT. For those people at Jefferson Lab, the meeting will be in room F326.

Meeting Connections

  1. To join via a Web Browser, go to the page [1] https://bluejeans.com/115815824.
  2. To join via Polycom room system go to the IP Address: 199.48.152.152 (bjn.vc) and enter the meeting ID: 115815824.
  3. To join via phone, use one of the following numbers and the Conference ID: 115815824.
    • US or Canada: +1 408 740 7256 or
    • US or Canada: +1 888 240 2560
  4. More information on connecting to bluejeans is available.

Reconstruction & Analysis Studies Status

Agenda

  1. Announcements
  2. Data Production
  3. Analysis Focus
  4. This Week's Studies
  5. Any other studies
  6. Upcoming Study Schedule

This Week's Topics

  • Talks: Summary/overviews only. Detailed discussions should be held in the appropriate working groups.
  • Talks should be limited to 10 + 5 minutes.

Data Production

  1. Monitoring Update --- Thomas Britton
  2. Processing Update --- Alex Austregesilo
  3. Calibration Update --- Sean Dobbs
  4. Simulations Update --- Sean Dobbs

Analysis Focus: Total Cross Sections, Acceptance Corrections

  1. J/Ψ, φUpdate --- Lubomir
  2. η, ρ, ω, φ --- Simon
  3. φ --- Thomas
  4. ω --- Cristiano
  5. ω --- Mike
  6. Any others?

Studies: Beamline & Triggering

  1. Flux --- Justin
  2. Beam energy --- Beamline Group
  3. Dependence of polarization on energy cut --Mike D
  4. Beam Asymmetries --- Alex Austregesilo
    • P Sigma
      Phase

Upcoming Study Schedule

  • Updates on experiment, simulation, and the comparison between the two.
  • Summary/overviews only. Detailed discussions should be held in the appropriate working groups.
  • All talks should be limited to 10 + 5 minutes.

Next Week: Hit Efficiencies (i.e. Is the detector working?) Updates on experiment, simulation, and the comparison between the two.

  1. CDC Hit Efficiencies --- Naomi Jarvis
  2. FDC Hit Efficiencies --- Alex Austregesilo
  3. BCAL Hit Efficiencies --- Elton Smith
  4. FCAL Hit Efficiencies --- Jon Zarling
  5. Trigger emulation --- Alex Somov
  6. Triggering efficiency --- Alex Somov

The week after next: Alignment & Track/Shower Efficiencies (i.e. Is the reconstruction working?) Updates on experiment, simulation, and the comparison between the two.

  1. Drift chamber alignment --- Mike Staib, Alex Barnes
  2. Tracking Efficiencies --- Paul Mattione, Cristiano Fanelli
  3. BCAL Shower Efficiencies --- Elton Smith
  4. FCAL Shower Efficiencies --- Jon Zarling

The following week: Other Updates on experiment, simulation, and the comparison between the two.

  1. Track / Hit Matching: BCAL, FCAL, TOF, SC --- Paul Mattione, Simon
  2. Efficiencies: TOF, SC --- Beni, Mahmoud
  3. Means & Resolutions (time, energy, dE/dx): Tracking, BCAL, FCAL, SC, TOF
  4. Uncertainties: PID (BCAL, FCAL, TOF, dE/dx), Kinfit (BCAL, FCAL, tracking)
  5. Channel/Analysis Studies: Branching ratios, cross sections, SDMEs, beam asymmetries
  6. Other reconstruction/analysis issues

Simulation Reports

15 Minute summaries of progress and future plans on the 2017 mcsmear Workplan.

  1. August 9th
    1. BCAL - Mark Dalton
    2. CDC - Naomi Jarvis
    3. SC - Mahmoud Kamel
  2. August 17th
    1. FCAL - Matt Shepherd
    2. FDC - Lubomir Pentchev / Alex Austregesilo
    3. TOF - Beni Zhilmann

Minutes

  • Analysis launch started Tuesday, analyzing 60 channels, and should finish in a few days (~500 jobs running at once):
    • It's using the new library, which has not been merged into the master because it's not 100% functional yet (e.g. no detached vertices yet). It has been tagged though: analysis-2017_01-ver01-batch01
    • It's using a set of built-in cuts that are applied to ALL channels: Analysis Launch Cuts
    • Running with 12 threads, it's 4x faster than the old library (30 min vs 2hr) and uses 1/2 the RAM (9 vs 18 GB)
    • With tighter cuts than before, the trees are 27% of the size of the REST files (vs. 62% with old cuts), and is expected to take 32 TB of space.
  • Lubomir is trying to understand why the background is different for his e+/- analysis between the 2016 and 2017 data. The response from the kinematic fit looks like it may be different for the background, although it's uncertain why this might be. He's still looking into it.
  • Mike Dugger showed that the extracted values of the polarization depend heavily on the width of the ΔE matching cut between the tagger and the PS, at least for narrow-width cuts. The same narrow cut yields a different polarization for the different PARA/PERP/45/135/AMO settings. However, everything looks consistent with wide cuts, or by not cutting at all, so that's what he's doing for now.
  • Alex Austregesilo showed the extracted P*Sigma from the Spring 2017 data as a function of run # for the different polarization orientations, showing that they are relatively stable and in agreement with each other.