Muon Trigger report A Ventura INFN Lecce and

Muon Trigger report A. Ventura INFN Lecce and University of Salento on behalf of the Muon Trigger Signature Group Muon Combined Performance Meeting CERN, January 27, 2010 A. Ventura Muon Trigger report

Outline ¢ Trigger reprocessing status ¢ Performance on reprocessed data (run 142191) ¢ Muon offline-EF comparison ¢ Menus at restart ¢ Conclusions 2 A. Ventura Muon Trigger report

Trigger reprocessing ¢ ¢ At the end of 2009, a reprocessed pilot run has been done including the patch fixing Mu. CTPI problem All jobs have been launched by January 21 after another pilot run was produced. The schedule is constrained by validation of release 15. 5. 5 for the forthcoming restart of data taking The reprocessing is still going on and more new results will be available asap More details available on https: //twiki. cern. ch/twiki/bin/view/Atlas/Trigger. Repr ocessing 2009 3 A. Ventura Muon Trigger report

Trigger performance on reprocessed run 142191 ¢ Trigger rate (Stream: Muonsw. Beam) l l Black: L 1 MU 6 Blue: L 2 mu 6_Msonly Green: EF mu 6_Msonly Red: L 1_MBTS_1_1_COL 4 A. Ventura Muon Trigger report

Trigger performance on reprocessed run 142191 ¢ MU 4 Trigger chain flow (Stream: Muonsw. Beam) mu mu 4_M mu 4_p So 4_L nly 2 M So nly _pa ass ss. L L 2 2 5 A. Ventura Muon Trigger report

Trigger performance on reprocessed run 142191 ¢ Mu. Fast w. r. t. EF SA using L 2 MSonly_pass. L 2 Black: mu. Fast reconstruct OK Blue: L 2_mu 6 passed (additional cut by p. T value) EF SA pointing cut |d 0| < 500 mm 6 A. Ventura Muon Trigger report

Trigger performance on reprocessed run 142191 ¢ Offline muons w/ MBTS_1_1_Col required Muid. SA Muid. CB 7 A. Ventura Muon Trigger report

Trigger performance on reprocessed run 142191 ¢ EF muons wrt Muid. SA and wrt Muid. CB (w/ MBTS) 8 A. Ventura Muon Trigger report

Trigger performance on reprocessed run 142191 ¢ ¢ Muonsw. Beam stream EF_mu 4_MSonly 16 events pass COLLCAND selection, out of which 6 pass L 2 and 4 pass EF Correlation EFMS-Moore is perfect for 3 of the 4 events ¢ Distance EFMS-Muid. MS 9 A. Ventura Muon Trigger report

Trigger performance on reprocessed run 142191 ¢ A. Ventura EFMS-Muid. MS correlation: eta and phi Muon Trigger report

Trigger performance on reprocessed run 142191 ¢ Correlation EFMS-Moore: number of hits per track 11 A. Ventura Muon Trigger report

Muon offline and muon EF comparison (1) ¢ ¢ ¢ A fruitful discussion was held last week involving Domizia within the Muon Trigger Signature group. Definition of quality cuts for muon offline tracks to be used for comparison with muon HLT objects requires feedback from MCP group and cooperation with offline experts Studies shown so far were performed by means of (re)processed trigger ntuples: very useful tool, found in agreement with what observed in ESDs l Still some work is needed for including offline info and solve few issues left in the ntuple writing procedure 12 A. Ventura Muon Trigger report

Muon offline and muon EF comparison (2) ¢ ¢ Also the use of minimum bias stream (instead of Muonsw. Beam) can reduces the large contamination from cosmics. As of now, comparison has been performed with Moore/Muid, next steps will involve also Muonboy tracks. It was agreed to prepare a list of golden trigger events, based on the requirements imposed at offline level (at least for the limited statistics presently available). First step is to define and check L 1 efficiencies, then to validate muon HLT The use of the MCP and Trigger/TDAQ sharepoints is going to be fundamental for these purposes 13 A. Ventura Muon Trigger report

Trigger menus at restart (very preliminary) ¢ ¢ ¢ In 2009 menu, special commissioning chains were: l _pass. L 2 MS (Mu. Fast in PT, Mu. Comb, Combination with ID at EF) -> Mu. Comb commissioning l _L 2 MSonly_pass. L 2 (Mu. Fast in PT, Combination with ID at EF) -> EF Commissioning We would now configure a new chain (_pass. HLT) replacing the five chains L 2 MSonly_pass. L 2, _pass. L 2 MS, i_loose_MSonly_Si. Track More details in this afternoon discussion (see Margherita Primavera’s talk) http: //indicobeta. cern. ch/conference. Display. py? conf. Id=74170 14 A. Ventura Muon Trigger report

Conclusions ¢ Trigger reprocessing ongoing, updates on muon L 1 and HLT performance expected soon l ¢ Central Trigger Validation moving towards new tools to validate/compare data to Monte Carlo (input to come from all trigger slices). l ¢ Studies are based on both ESDs and Trigger Ntuple: overall agreement, with minor issues to be addressed Strategies for offline comparison are under study Discussion ongoing on startup menus (see this afternoon’s session) 15 A. Ventura Muon Trigger report

Backup slides A. Ventura Muon Trigger report

Trigger performance on reprocessed run 142191 ¢ mu. Comb Muons wrt Muid. CB (w/ MBTS) 17 A. Ventura Muon Trigger report

Update on the new RPC cabling ¢ ¢ On Jan 12 a new tag of Muon. Cnv. Example (02 -03 -19) was introduced into 16. 0. 0 nightlies which sets the new RPC cabling (Muon. RPC_cabling) to be used by default with real data even if the trigger algorithms are active. The algorithmic behavior of the new package does not differ w. r. t. the old one and the only difference lies in the access to COOL for the cabling maps. The new cabling has been used at Tier 0 for months and last year it was also proved to work correctly in 15. 5. X. Y, Atlas. P 1 HLT caches in athena. MT with stop-start transitions. Now the idea is to test the new cabling as soon as data taking with collisions will restart; this is going to be done over the express stream offline comparing performance of the HLT with old and with new code. 18 A. Ventura Muon Trigger report