ALCPG Sim/Reco Weekly Meeting: Weekly updates with an emphasis on vertexing and tracking.

Tuesday 3 Jul 2007, 13:30 → 15:15 US/Pacific

Time: 13:30 Pacific Time US (UT -8)

Connection Information:

Call: 510-665-5437 Meeting ID: 0873

Participation:

Please be prepared to participate via the ecs shared desktop.

Instructions can be found at:

http://confluence.slac.stanford.edu/display/ilc/Connecting+via+ESnet+Collaboration+Service

Mailing List:

lcd-dev

self-subscribe at:

http://www.slac.stanford.edu/cgi-bin/lwgate/LCD-DEV/subscribe.html

1 Java-Rave

Experiences with installing and using the JNI binding to connect the C++ RAVE vertexing toolkit with the org.lcsim framework.

Speaker: Mr Jan Strube (University of Oregon)

2 Effects of Tracking Limitations on Jet Mass Resolution

Speakers: Prof. Bruce Schumm (University of California at Santa Cruz), Chris Meyer

Slides meyer_0707.pdf meyer_0707.ppt

3 Characterization of AxialBarrelTracker Inefficiencies

Speakers: Prof. Bruce Schumm (University of California at Santa Cruz), Tyler Rice

Slides rice_0707.pdf rice_0707.ppt

4 Z segmentation studies

Speaker: Lori Stevens (SLAC)

Slides Z_Segmentation_Update.pdf Z_Segmentation_Update.ppt

5 Forward Tracking Studies

Status report and plans for studying the forward tracking region.

Speakers: Francisco Ruiz, Dr Hans Wenzel (FNAL)

Slides Francisco.pdf Francisco.ppt

6 Keeping Track of TrackerHits

Discussion on how best to do the bookkeeping for TrackerHits used by different track-finding algorithms.

Speaker: Dr Robert Kutschke (Fermilab)

Slides HitManagement.pdf HitManagement.ppt

Notes on Rob's talk:

• Page 4.
  • Tony suggested that we add another method: using the LCIO relation objects.
• Page 5.
  • Re option 1: Dima's code never writes out a container of unused hits. It uses method 4 to get its input and presumes that downstream code will do the same.
• Page 6
  • Rich wants to keep option 5 on the table. He sees each hit holding a list of tracks to which it belongs and each layer of pattern recognition adding to those lists as it consumes hits.
  • Rob replied that he does not like approach:
    • He prefered to keep the principle that objects in the event are immutable in order to facilitate the move to a proper data provenance system.
    • Another reason, not mentioned at the meeting, is that this is not full featured enough to implement things that we will want to do.
• Page 8.
  • Tony suggested that the last bullet ( the way to specify names of input and output containers ) should be made part of the framework.
• Future plans:
  • move the discussion to the ILC forum
  • Rob will continue to collect information and make a status report next week.

Summary of options

There are 6 options on the table, the 5 listed page 4 plus the LCIO relation objects.