Difference between revisions of "February 1, 2012 Tracking CDC/FDC"

Revision as of 14:14, 1 February 2012

Meeting Time and Place

Wednesday February 1, 2012 at 11:00am At Jefferson Lab, the meeting will be held in F326

Connections

To connect from the outside, please use ESNET

1. ) ESNET: 8542553
   
2. )To connect by telephone, dial:

• You can look up toll-free number at http://www.readytalk.com/intl
• US and Canada: (866)740-1260 (toll free)
  
• International: (303)248-0285 (toll call)
• enter access code followed by the # sign: 3421244#
  

Hardware Agenda

1. Announcements
2. FDC Update logbook
3. CDC Update
   1. February 2012 GlueX-doc 1898

Tracking Software Agenda

1. Announcements.
2. Status of performance metrics.
3. Summary of current work on tracking.

Hardware Minutes

Participants: Curtis, Naomi, Dave, Mark, Eugene, Tim, Simon, Beni, Lubomir

1. FDC Update
2. CDC Update
   1. Stringing - this should be finished next week, there were 220 wires left to do this morning.
   2. HV - 500V is being applied to (one end of) each wire and the (very small) current measured.
   3. Resistance - now is a good opportunity to measure the resistance of each wire.
   4. Survey - wires need to settle for a week before finishing tension measurements. Tim will schedule the survey team between Feb 20 and March 5.
   5. Remaining hardware - the cheeseplate is on its way to CMU. Slava is working on the other hardware.
   6. Soldering - Kait and Amy are now expert at this.
3. Collaboration meeting
   1. One talk each will be scheduled for the CDC and FDC.

Software Minutes

The meeting started about five minutes early due to an early completion of the hardware meeting. Present at the meeting were: (CMU) N. Jarvis, P. Mattione and C.Meyer, (JLab) E. Chudakov, M. Ito, D. Lawrence, L. Pentchev, S. Taylor, B. Zihlmann (IU) R. Mitchell, K. Moriya, M. Shepherd (FSU) N. Sparks.

1. Performance Metrics:
   1. Simon reported on setting up an automated single track program similar to the the automated b1-pi program that now runs. This would run twice per week and allow us to track these more basic parameters. Mark is looking at this to make sure that he can get it to run before letting this go.
   2. Paul has some specific metrics that we can add to the single track code once it is running. Paul agreed to send this information to Kei who will implement it.
   3. Simon reported that currently about 1/2 of the reconstruction time is spent in the vertex section of the code---DVerteX Factory. This is even true for the single-track events. This factory caries out a Hough transform on the event to try and locate verticies.
2. Snapshot of Current Activity.
   1. Paul reported that he is looking at the code that reconstructs low-momentum pions in the drift chambers.
   2. Kei has agreed to work on adding additional metrics to Simon's automated scripts for monitoring software performance.
   3. Simon reported that he continues to work on all aspects of tracking. He has spent some time looking at the slow protons that stop in the CDC. Beni noted that with regard to this, recent updates have improved things by about 50%. Simon is also trying to look at regions where the tracking chi-square is poor.
   4. Beni is working on the slow protons which stop in the CDC. He noted that for these, the correct charge is found in 98% of the cases. Beni is also working on an improved display that will allow one to color code the tracks and hits in an event to be able to easily see how they were assigned. He noted that at the moment, the code appears to assign every hit to a track. There are no left-over hits in the chambers.
3. Other Important Activities
   1. Curtis brought up the question of other activities that should be addressed in future meetings. This opened up a brain-storming session on topics. We identified the following two areas that would be useful to address.
   2. The error matrices associated with tracks used to be wrong in many unclear ways. We have not made a push recently to see where we stand with these, and if it is now possible to correct them using standard scaling techniques. Will had tried to look at these for the case of b1-pi about 2 months ago, but at that time found that the photon errors building the pi0s was giving a lot of issues.
   3. How fast does our code run, and where is it spending its time? We should try to set up a method to track this between now and the software review in June. David and Simon have tools that do this, and it was suggested that it be added to the standard "weekly jobs", which would then let us track things.
4. Next Meetings
   1. The next scheduled meeting is the Wednesday before the collaboration meeting. We will not hold a meeting then.
   2. Mark agreed to add a talk on tracking to the offline session in the Collaboration meeting.
   3. The next meeting will be held on Wednesday February 29.