110 likes | 179 Views
Addressing bug in assigning primary vertex to di-hadron candidate, leading to inefficient Pointing Variable calculation as #PV increases. Detailed analysis and proposed upgrades for improved efficiency in trigger selection.
E N D
Update on Bsff Efficiency S. Blusk Nov 18, 2009
BUG ! Found a bug in the way I was assigning a primary vertex to a di-hadron candidate. Led to improper value for the Pointing Variable Loss of efficiency as #PV increases.
Pointing Variable for DC06 Offline Selected Bsff K K f1 K f2 Up to 4 combinations/event K Proposed cut in LHCb-INT-2009-008 Up to 2 combinations/event High Efficiency when kaonsfrom different f, but ~35%efficient when from same f.
Trigger Efficiency for DC06 Offline Selected Events Bug fixed …with Bug Small (if any) loss in trigger efficiencyIF ALL THE DAUGHTERS ARE RECONSTRUCTED
Bsff Trigger Efficiency with Minimal Upgrade Layout P(4 long tracks)24%, 21%, 21%, 15%
Bsff Trigger Efficiency with Minimal Upgrade Layout P(trigger)33%, 31%, 20%, 25% P(4 long tracks)24%, 21%, 21%, 15%
Bsff Trigger Efficiency with Minimal Upgrade Layout P(trigger)33%, 31%, 20%, 25% P(4 long tracks)24%, 21%, 21%, 15% P(trigger)10.3%, 8.4%, 8.8%, 6.1%
Bsff Trigger Efficiency with Minimal Upgrade Layout 2e32 20e32: • ~38% loss in 4-track efficiency • ~25% loss in trigger efficiency(less sensitive, since only need 2 tracks) • ~40% loss in combined (trigger+4-track recon. Eff) • BOTTOM LINE: • Efficiency dominated by 4-track recon. efficiency P(trigger)33%, 31%, 20%, 25% P(4 long tracks)24%, 21%, 21%, 15% P(trigger)10.3%, 8.4%, 8.8%, 6.1%
Quick check on MinBias after Bug Fix • On 2e32: No difference (expected, since mainly affects events with many PVs) • Running on higher lumi’s now.
Summary • Bug fix trigger efficiency robust with increased lumi IF THE TRACKS ARE FOUND. • Need to run on full stats Min Bias • Bring rate at 2e33 down to XXX kHz forhigher level trigger (exclusive & inclusive b trigger) • Need cuts that suppress heavy flavor, as these dominate the trigger rate.