Talks

• Status of the road algorithm for electrons in L3: Florian Beaudette.
  The road method has been demonstrated offline to give significantly better efficiency/background rejection than the simple cone-based electron finder for electrons in jets. In order to ease implementation into L3 and future maintenance, the job of finding the list of cells that are consistent with a given extrapolated path has been factorised out into a new stand-alone package "calroad". The L3-specific parts of the implementation (for example the job of finding the Level-3 energies of cells in the list) have been performed by Volker Buescher. Most of the necessary code is implemented in p12. The technical correctness of the L3 implementation has been tested extensively in data using J/Psi -> electron candidates selected offline using RECO information. Tests of timing and memory consumption still need to be performed. The following code modifications (in p12?) still need to be done:
  • Emfraction is calculated differently for road method (uses floors 1,2,3) and cone method (uses floors 1,2,3,4).
  • E/p cut needed as new parameter in trigger list.
  • Dealing with -2.9 cm shift of calorimeter wrt. central tracker. (Needs code from Steve Khan to handle calorimeter alignment constants to go into release?)
• L3 analysis_example and chunk documentation: Erich Varnes.
  Erich presented his "user"-level analysis examples and documentation of physics objects for L3. He has:
  • Written documentation of the access methods for all the quantities in all of the L3 physics_results objects in the current global trigger list.
  • Made a simple stripped down user function that makes loops over each of the L3 physics_results objects and gives an idea of how to use the access methods for a few interesting variables. This will be incorporated shortly by Marco Verzocchi into the analysis_example package.
  Please take a look at Erich's documentation. Try to use it and send him feedback!
  If you have documentation for the contents/access methods of other parts of the L3Chunk then please send them to Erich!
• Generic trigger list for Monte Carlo production: Terry Wyatt.
  In the next month or two we hope to have trigsim running as part of the standard MC production on the offsite farms. The idea behind the generic trigger list is primarily to get the physics objects produced by each level of the trigger available in the Monte Carlo output file for each event. This work is very closely connected with Erich's efforts to make it easier for non-experts to analyse L3 physics_results and Peter Tamburello's efforts to get L3 physics objects into the Thumbnail.
  Terry and Dan Claes came up with a proposed list of loose L3 filters/tools to be run as part of the generic trigger list. This proposal was discussed and a number of useful modifications were suggested. (The talk linked above has been updated to include the suggested improvements.) The following action items were identified:
  • P

Discussion Items

• We ran out of time and didn't have the planned discussion of plans/status for p13. Please take a look at the relevant part of our: overview of current activities and plans page and send comments. Needs offline discussion before next week's meeting.

Consumer survey:
If you read this please send Terry an email twyatt@fnal.gov so that he has some idea of whether or not these minutes get read and are or are not found to be useful!

Scribe: Terry Wyatt.