Talks

• Profiling and timing studies : Han Do
• Han pointed out the variety of syntax employed in error messages output to the log files.  This thwarted an initial plan to retrieve some timing info by parsing these files.  Additionally it appears that not all tools correctly implement calls to l3StatDevel.  Han suggests standardizing this output. He also showed some initial offline timing output from the l3Monitor package.  In the discussions that followed, it was pointed out that the timestamps in log files are neither reliable nor precise enough for our needs, and Han agreed to focus on the monitor statistics and perform profiler output as a consistency check.

• Tracking update : Daniel Whiteson
• Daniel updated last week's study of L3 track-matched electrons in a Z->ee data sample. These events are selected through calorimeter information (and a di-EM mass cut), unbiased by any track information. Matching was made to the nearest track in eta, phi.  Now correcting for track curvature (acheiving better propagation to the calorimeter) overall efficiciencies run 50-60% (70% in the later runs and 70-80% in the central eta region).  Event details of the inefficiencies have not yet been looked at.
• Daniel has pretty much concluded the disrepancy between reported and predicted rejections are due to the new addition of L2mu(5).  His original prediction of 56 was wrt to L1.  The online rejection of 15 is wrt L2.  Since the advertised L2mu(5) rejection rate is about 5, these values seem consistent.  

• L3MEt update: Lee Sawyer
• Lee reported on his first look at real data since, showing MEt distributions (x,y and totals) that compare before/after the zero suppression change was made offline.  He has not yet utilized L3NADA information, and still uses the null vertex.  Phi distributions show the vertexing offset. Since the current CFT bertexing at least provides z, he will check if it cleans up some of his Et distributions.
• Zero-usppression is available off-line, but should Marumi include the ped subtraction online?
  

Discussion Items

• Peter Tamburella asked about getting "charge" into the PhysResults base class to simplify its appearance in new track objects.  We discussed how to handle the tracking tool...elevate it to physics tool?   Write a physics tool wrapper that would control/create the PhysResults?  That opened a can of worms about streaming on track objects that went unresolved.
  

Scribe: Dan Claes.