trigger production status (Fred Kral) July 1996
Overview
Trigger efficiency and background rate calculations use large BBSIM data sets.
Status
The simulation group, with the help of half a dozen institutions, has met almost all the pending requests, greatly helping the trigger design effort in addition to exercising the production system mechanics.
About a CPU year worth of BBSIM data were produced in a rather short time. There are tapes of most physics processes including B0 CP modes plus tags, B physics, Charm physics, continuum q q-bar, taus, mu pairs, wide-angle Bhabhas and two-photon final
states. In addition, there are large lost-particle beam backgrounds data sets with or without photoproduction cross-section enhancements (hadronics), beam-gas interactions, small-angle Bhabhas and cosmic rays. Data are available at SLAC.
Plans
We plan a new request for updated data samples that incorporate the following changes since the original production that used SRT release 0.6.4:
- Beampipe interaction region geometry change (for lost-particle runs)
- Drift chamber geometry change with new cell counts and endplates (for GHits)
- Calorimeter GHit crystal numbering starts at 0 instead of 1
- Silicon vertex tracker changes to GHits structure (for reco Digis)
- HEPEVT generated particle information written to dbio bbsim.xdr file, where this requires pointers to GEANT gEvent and gVertex structures (is this code already available?)
Starting late July, we'd like to have a complete re-run of the Spring-96 requests to date, with an additional wish to run more than one iteration of the default lost-particle runs, if feasible. Data sets need to be available no later than early
September for us to meet the next internal milestone of the late-October trigger workshop.
Manpower
Fred Kral, request preparation
Ed Frank, experienced production manager for Penn site
Needs from other parts of simulation
GHit and Digi structures: Coordination with each of the subsystems is crucial to make the bbsim output files be long-lived. We would like each subsystem to please indicate its plans for GHits in the short term. Note that both the structures
and the interpretation of the structures (i.e. the meaning of GHit words) have to be stable or backwards compatible. In addition, the SVT digi structures also need to be stable (it would be better to discontinue their use in BBSIM).
Production run: We need a lot of help from the simulation group and people at old and hopefully new production sites. Interest sites should contact Torre Wenaus and subscribe to the simuprod hypernews discussion group.
Post-production planning: We need to make plans for making the data accessible, useful and understandable. Some issues, that will hopefully be further discussed in
simuprod are:
- Clearly defining allowed tape formats and size limits before production and to prepare for any differences between tapes from different sites
- Checks on data quality
- Collating related runs into single files in a post-production stage
- Filtering data in post-production stage to make trigger summary tapes
- Communication of parameters that are not on tape: rate normalization factors for background data tapes
Help is needed to improve post-production mechanics.
Physics generators: Missing benchmark modes: we need switches in BEGET to select B+B- (in addition to B0B0~) and to generate B to tau decays.
|
