Muon Trigger Validation issues Highlights of latest Validation

Muon Trigger Validation issues • Highlights of latest Validation Reports • Study on muons without magnetic field A. Ventura for the Muon Slice Validation Team 6 June 2008

Week 12 -18/05/2008 (M. B. ) • devval: only rel_0 working – rel_1 devval : strange crash in Atlas. Core (lib. SGTools) – rel_2 and rel_3 devval : No In. Det. Trig. Prep. Raw. Data. Format – rel_4 devval : conf problem in In. Det. Rec. Example – rel_5 devval : problem in Region. Selector – rel_6 devval : problems in Calo packages • dev: only rel_2, rel_5, rel_6 working – rel_1 dev : Trig. Muon. EF not building – rel_3, rel_4 dev, rel_0 : conf problems in ID • 14. 1. 0. Y, Atlas. Production rel_3 *not* working – rel_3 : problem in i. Pat fitter • Results were consistent across the working nightlies. • From rel 5, dev and rel 0, devval on, mu 20 i counts increased of ~20%, due to an update of Trigmu. Iso [Trigmu. Iso-00 -00 -20, including ID+calo isolation and endcap isolation].

Week 19 -25/05/2008 (A. S. ) • working devval: rel_2, rel_3, rel_5, rel_6, rel_0 NOT working devval: rel_1, rel 4 • working dev: rel_1, rel_3, rel_4, rel_5, rel_6, rel_0 NOT working dev: rel_2 • RTT Tests OK except: 0) no compiling (cf above) 1) devval rel_2 (crash) and rel_3 (crash), dev rel_4 (crash because of missing Input File), devval rel_0 (crash in EF) 2) Missing logs in devval rel_6 • 14. X. 0. MIG 2 saga: None compiled. Starting with Mu. Iso and Trig. Muon. Event inconsistencies (solved at the end of the week) Ending with Trigmu. Fast compiling error in mu. Fast: : hlt. Execute(const HLT: : Trigger. Element*, HLT: : Trigger. Element*)'

Week 26/05 -01/06/2008 (A. V. ) • Reference RTT file RTT_Muon. Slice_top. root updated to include the very relevant changes occurred so far in muon slice since the beginning of May (especially for LVL 2 isolation at high pt). • Memory leaks on muon slice mostly involved Trig. Muon. EFSegment. Finder and Trigmu. Fast • rel_1 dev: not working • rel_2 devval: not available • rel_4 devval: ATN not usable for tests involving EFID due to build failure in Atlas. Reco • rel_5 dev and devval: not available • rel_6 devval: ATN OK but crash in RTT: • rel_0 devval: RTT crash in Trig. Muon. EFCombiner. TMEFMatch. Maker , originating in Muid. Match. Maker matching. Procedure. • Many 14. 1. 0. Y builds failed during the week. bug #37467 Alg. Error. Auditor ERROR Illegal Return Code: Algorithm Trig. Ambiguity. Solver_Muon_EFID reported an ERROR, but returned a Status. Code "SUCCESS" Trig. Steer_EF WARNING Chain EF_mu 6 l aborting with error code ABORT_CHAIN GAUDI_EXCEPTION at step 0. . . FATAL 2008 -May-26 06: 33: 10 [static void ers: : Error. Handler: : Signal. Handler: : action(. . . ) at ers/src/Error. Handler. cxx: 88] Got signal 11 Segmentation fault (invalid memory reference) Import. Error: No module named Trig. Decision. Maker. Conf Py: Athena INFO leaving with code 3: "an error occurred when importing a file" RTT determined exit status to be: 3 bug #37448

This week 02 -08/06/2008 (A. V. ) • Many failures in RTT jobs during the week mostly due to resource problems • rel_3 devval initialization problems • Best releases on ATN so far looked rel_3, dev but… • Segmentation fault when running Trig. Muon. EFCombiner on higher statistics (RTT Trigger. Test), affecting also MET and Bphysics: there is a savannah bug about this (#37448) – Sergio prepared new tag: Trig. Muon. EF-00 -00 -23

Short term plans on Muon EF validation • New HLT trigger names were added or updated in ATN/RTT jobs (e. g. EF_mu 6_ef) • We are now going to validate the content as the algorithm develops in parallel to the usual Trig. Moore EF output, to be used as reference (e. g. pt/phi/eta spectra, efficiencies wrt LVL 2) • The goal is to implement automatic checks on Trig. Muon. EF output to be run daily both for ATN and for RTT

A glimpse at muons with B=0 • RDOs produced by O. Kortner under: /castor/cern. ch/user/k/kortner/straight_tracks/RDO Muons with p. T = 6 Ge. V • AODs and AANTs produced under /castor/cern. ch/user/a/anventur using 14. 1. 0 • Here plots obtained starting from AANTs (for LVL 1) and AODs (for HLT)

Generated eta, phi and LVL 1 resolution Generated eta (flat) eta resolution Generated phi (flat) phi resolution LVL 1 acceptance @ 6 Ge. V: 0. 829 +- 0. 024

LVL 1 pt thresholds barrel 4 6 10 11 20 40 endcap 4 6 10 15 20 • Behavior very similar to known LVL 1 performance on muons with standard magnetic field (…? ) • As known, better capability of threshold recognition in endcap than in barrel 40

mu. Fast and mu. Comb tracks Usual LVL 2 performance as known for 6 Ge. V muons in the standard ATLAS magnetic field: • is it really B=0? mu. Fast <pt> = 5. 95 Ge. V mu. Comb <pt> = 6. 82 Ge. V

Muid. SA reconstructed tracks Usual performance as known for 6 Ge. V muons in the standard ATLAS magnetic field: • is it really B=0? cot(theta) <pt> = 6. 02 Ge. V phi

Muid. CB reconstructed tracks Muid. CB efficiency at 6 Ge. V is 70% of Muid. SA, as known for standard magnetic field. <pt> = 5. 65 Ge. V The sample should be better understood or reproduced. cot(theta) phi