Work progress J Tao CMSIHEPBeijing 2008 05 30
Work progress J. Tao CMS-IHEP/Beijing 2008. 05. 30 1/13
Shower Splitting Method Fitting method B on in Whole Barrel EM shower 1 Calculating with one EM Shower: χ2_1 With Ehit, we can get COG(η, φ) Δη, Δφ, A, E 0 4 parameters (B on/off) Fitting with 2 EM showers Shower 1: η 1, φ1, A 1, f E 0, Δη 1, Δφ1_1. , Δφ1_2 Shower 2: η 2, φ2, A 2, (1. 0 -f) E 0, Δη 2, Δφ2_1. , Δφ2_2 f: energy factor 0. 0~1. 0 5 parameters EM shower 2 Original COG New 2 COG From 10, 20, 30, 50, 80, 120, 150 Ge. V in whole barrel samples if(χ if( 2_2< χ2_1) 2 EM shower invariant mass Lontitudinal: Gamaa-function Lateral: 2/13
20 Ge. V pi 0/Gamma Ecal. Only sample with B on in whole Barrel Gamma Sample ++++++++++ analyzed 4848 events ISTAT=3 fit 2 okevt: 2924 Fit 2 Chi 2 is better: 2921 n. Event keeped : 4364 (90%) with mass cut: 0. 05831 If Chi 2 of fitting with 2 EM shower is better Mass=Invariant mass of 2 EM shower, else Mass=0. 0. Pi 0 Sample ++++++++ analyzed 4833 events fit 2 okevt: 3892 Fit 2 Chi 2 is better: 3891 n. Event cut : 2971 with eff. : 61. 4732% ISTAT: a status integer indicating how good is the covariance matrix: 0= not calculated at all 1= approximation only, not accurate 2= full matrix, but forced positive-definite 3= full accurate covariance matrix 3/13
20 Ge. V pi 0/Gamma Ecal. Only sample with B on in whole Barrel Gamma Sample ++++++++++ analyzed 4848 events ISTAT=1, 2, 3 fit 2 okevt: 4409 Fit 2 Chi 2 is better: 4404 n. Event keeped : 4364 (90%) with mass cut: 0. 06115 If Chi 2 of fitting with 2 EM shower is better Mass=Invariant mass of 2 EM shower, else Mass=0. 0. Pi 0 Sample ++++++++ analyzed 4833 events fit 2 okevt: 4680 Fit 2 Chi 2 is better: 4670 n. Event cut : 3098 with eff. : 64. 101% 4/13
30 Ge. V in Barrel ISTAT=3 ISTAT=1, 2, 3 Gamma Sample ISTAT=3 ++++++++++ analyzed 4851 events fit 2 okevt: 2913 Fit 2 Chi 2 is better: 2909 n. Event keeped : 4366 (90%) with mass cut: 0. 08065 Pi 0 Sample ++++++++ analyzed 4838 events fit 2 okevt: 3793 Fit 2 Chi 2 is better: 3793 n. Event cut : 2785 with eff. : 57. 5651% Gamma Sample ISTAT=1, 2, 3 ++++++++++ analyzed 4851 events fit 2 okevt: 4440 Fit 2 Chi 2 is better: 4428 n. Event keeped : 4367 (90%) with mass cut: 0. 08425 Pi 0 Sample ++++++++ analyzed 4838 events fit 2 okevt: 4646 Fit 2 Chi 2 is better: 4633 n. Event cut : 2956 with eff. : 61. 0996% 5/13
50 Ge. V in Barrel ISTAT=3 ISTAT=1, 2, 3 Gamma Sample ISTAT=3 ++++++++++ analyzed 4847 events fit 2 okevt: 2998 Fit 2 Chi 2 is better: 2995 n. Event keeped : 4363 (90%) with mass cut: 0. 1303 Pi 0 Sample ++++++++ analyzed 4826 events fit 2 okevt: 3608 Fit 2 Chi 2 is better: 3608 n. Event cut : 1839 with eff. : 38. 1061% Gamma Sample ISTAT=1, 2, 3 ++++++++++ analyzed 4847 events fit 2 okevt: 4379 Fit 2 Chi 2 is better: 4370 n. Event keeped : 4363 (90%) with mass cut: 0. 1358 Pi 0 Sample ++++++++ analyzed 4826 events fit 2 okevt: 4646 Fit 2 Chi 2 is better: 4633 n. Event cut : 1850 with eff. : 38. 334% 6/13
80 Ge. V in Barrel ISTAT=3 ISTAT=1, 2, 3 Gamma Sample ISTAT=3 ++++++++++ analyzed 4845 events fit 2 okevt: 2977 Fit 2 Chi 2 is better: 2977 n. Event keeped : 4361 (90%) with mass cut: 0. 21283 Pi 0 Sample ++++++++ analyzed 4832 events fit 2 okevt: 3425 Fit 2 Chi 2 is better: 3608 n. Event cut : 698 with eff. : 14. 4454% Gamma Sample ISTAT=1, 2, 3 ++++++++++ analyzed 4845 events fit 2 okevt: 4350 Fit 2 Chi 2 is better: 4346 n. Event keeped : 4361 (90%) with mass cut: 0. 21977 Pi 0 Sample ++++++++ analyzed 4832 events fit 2 okevt: 4469 Fit 2 Chi 2 is better: 4663 n. Event cut : 705 with eff. : 14. 5902% 7/13
120 Ge. V in Barrel ISTAT=3 ISTAT=1, 2, 3 Gamma Sample ISTAT=3 ++++++++++ analyzed 4841 events fit 2 okevt: 2868 Fit 2 Chi 2 is better: 2867 n. Event keeped : 4357 (90%) with mass cut: 0. 30551 Pi 0 Sample ++++++++ analyzed 4823 events fit 2 okevt: 3207 Fit 2 Chi 2 is better: 3207 n. Event cut : 569 with eff. : 11. 7976% Gamma Sample ISTAT=1, 2, 3 ++++++++++ analyzed 4841 events fit 2 okevt: 4303 Fit 2 Chi 2 is better: 4294 n. Event keeped : 4357 (90%) with mass cut: 0. 31435 Pi 0 Sample ++++++++ analyzed 4832 events fit 2 okevt: 4456 Fit 2 Chi 2 is better: 4347 n. Event cut : 572 with eff. : 11. 8598% 8/13
π0 rejection For keeping 90% photon efficiency E [Ge. V] Rejection [%] 20 64. 1 30 61. 1 50 38. 3 80 14. 6 120 11. 9 Parameters value? From 10, 20, 30, 50, 80, 120, 150 Ge. V in whole barrel samples 9/13
Discrimination of γ/π0 using Neutral Network For keeping 90% photon efficiency -- For unconverted photons in Barrel -- Training-test sample. CMSSW_1_3_1 Pt [Ge. V] Rejection [%] 20 69. 9 30 57. 6 40 44. 9 50 36. 0 60 30. 5 Energy distribution of γ from H→γγ with m. H=120 Ge. V (Genlevel) Peak ~ 60 Ge. V 10/13
Transverse Energy distribution of γ from H→γγ with m. H=120 Ge. V (Gen-level) 10000 events 11/13
Backup 12/13
Ecal. Only, B on, Gamma --10, 20, 30, 50, 80, 120, 150 Ge. V in whole Barrel 13/13
14/13
15/13
- Slides: 15