Talks

• Review of p13 readiness for online running:
  • This has been progressing rather slower than hoped for due to a number of technical problems. Some "by hand" editing of the level3.sim file was needed to get a trigger list that is compatible with p13, but otherwise identical to global_CMT-9.31. Han Do has kinly agreed to take responsibility to run trigsim and produce l3fanalyze roottuples to allow comparison of online (p12.06) and offline (p13.05) results. Known features/problems:
    • Trigsim crashes after about 200 events. No traceback produced.
    • ET for calorimeter objects offline differ from online on an object-by-object basis at the few percent level. Possibly due to new calorimeter geometry (shift of 3 cm). To be checked. Statement from Marumi: calorimeter calibration/pedestal/thresholds should be identical in p12 and p13.
    • MEt objects appear to be filled with zeros.
  • Comparator: Han Do
    
• Comparison of missing HT and missing ET: Laurent Duflot
  The idea is to calculate MEt using jets rather than all calorimeter cells (L3FMHt). Hoped that this will be less sensitive to calorimeter noise and less luminosity dependent. Correlation of MHt and MEt is good for signal Monte Carlo with genuine jets + MEt. Could muon-tagged bbar events be used to compare MHt and MEt in real data with genuine MEt? For QCD (low ET) there is almost no correlation between MHt and MEt. MHt distribution is rather broader.
  Action item (done Thursday morning): Laurent and Terry entered the necessary filter and test trigger definitions into the trigger database. This will allow full testing of the new filter code and DB definitions with trigsim (as required by our new certification rules).
• Code for CPS L3 calibration: Yi Jiang
  Several stage process has been implemented:
  • gui to get calibration values to text files.
  • map to convert hardware to software strip id.
  • calibration data read data into unpacker.
  Currently determined by .rcp parameter whether to use calibration data from MC/flat file/DB. We need the default to be correctly set up for online use. Ideally, for offline we need correct behaviour to be automatic as set up by Volker for the calorimeter.
  Action item 1: Andre Turcot will look at the calorimeter implementation.
  Action item 2: Andre and Terry will look at what is implemented in the trigger DB for CPS and update if necessary. Some standalone tests are needed before CPS could be turned on in the default MC trigger list.

Discussion Item

• Chris Barnes has produced a new package l3ftrack_phys_results. This is the first step in the attempt to break the link between the L3 tool/filter code and the code for making and analysing TMBs.
  Action item: All physics tool responsible people - please take a look at this example. At next week's meeting we need to get an estimate of when you think you could implement such a change for your package. Since an interface change will be required for each package it would be very much simpler if we could make all the required changes at the same time! It is hoped to do this in time for p14, which would require the mods to go in mid-January 2003.

Scribe: Terry Wyatt.