A Study on Cluster Resolution.ppt

1、A Study on Cluster Resolution,Neighborhood hit density gradients are proposed as a means for- identifying cluster boundaries- implementing a cluster split/merge strategyRelies on the inspection of “calorimeter domains” collections ofconnected cells ganged as projective towers. Currently coded in“box

2、 form” ( n x m x l ) cells as segmented in (theta,phi,layer).A tool to be used as a sup-port to clustering algorithms.Equally applicable to both analogue and digital readouts.,A. Maciel (NIU), Simulation & E-flow Wshop , NIU, November 7-9, 2002,1,2,neighborhood hit density gradients,EM shower profil

3、es suggest pre-clustering box of 4x4 cells.Locate (4 x 4 x all) domain with highest energy (or #hits).Inspect neighborhood w/ grads; i.e. find (Rmin, Gmin) and (Rmax,Gmax).If Gmax k*Gmin, then a secondary clusteris declared found (currently using k = 2 ).Rmin & Rmax determine a search area for“next-

4、hottest” (4 x 4 x all) domain.Cross check that 2 cluster also “sees” 1Use both sets of R&G(min,max) to re-size clusters.,1,search area for 2,Cluster ID Calorimeter Domain Methods,Rmin+1,Rmin,Rmax,Rmax + size,Gmin,Gmax,A. Maciel (NIU), Simulation & E-flow Wshop , NIU, November 7-9, 2002, distance, ge

5、nerated (cm), distance, reconstructed (#of cells),5000 events superimposed,(#of cells),1= layers 0 - 9 2= layers 10-19,0 ( 10GeV ) 5K single-0 events,neighborhood 2 is tested if neighborhood 1 fails,1 cell = 0.5 cm,A. Maciel (NIU), Simulation & E-flow Wshop , NIU, November 7-9, 2002,inefficiency,GeV

6、,Generated (upper) .vs. “Reconstructed” (lower),A. Maciel (NIU), Simulation & E-flow Wshop , NIU, November 7-9, 2002,Domain (Projective Box) Energy Resolution,A. Maciel (NIU), Simulation & E-flow Wshop , NIU, November 7-9, 2002,Layer energy centroid position fluctuations across layers (distances in

7、cell units),1 pre-cluster,1 re-sized,2 pre-cluster,2 re-sized,A. Maciel (NIU), Simulation & E-flow Wshop , NIU, November 7-9, 2002,A series of event displays for 10GeV 0 with the associated neighborhood gradientsanalysis as an illustration of the method.,Event 81,Evt# 81 Gamma1: ThetaBin=420 PhiBin=

8、213 Reco Energy = 4.87 MC Energy = 5.28 Gamma2: ThetaBin=417 PhiBin=204 Reco Energy = 4.39 MC Energy = 4.72,Example of a symmetric decay,Neighborhood Grads,Event 81,1 “looks at” 2 layers used: 0-9,2 “looks at” 1 layers used: 0-9,distance, in number of cells,distance, in number of cells,Event 36,Exam

9、ple of an asymmetric decayEvt# 36 Rmin1 = 4 , Rmax1 = 9 Rmin2 = 36, Rmax2 = 9 , DeltaR = 10Gamma1: ThetaBin=420 PhiBin=288 Box-Reco Energy=8.09 ( MC = 8.88 )Gamma2: ThetaBin=410 PhiBin=289 Box-Reco Energy=1.33 ( MC = 1.12 ),Neighborhood Grads,Event 36,1 “looks at” 2 layers used: 0-9,2 “looks at” 1 l

10、ayers used: 0-9,distance, in number of cells,distance, in number of cells,Event 48,Evt# 48 (how it can get tricky.) Gamma1 Energy=8.10 MC = 9.50 Gamma2 Energy=0.31 MC = 0.50,1,2,“dirt”,Neighborhood Grads,Event 48,1 “looks at” 2 layers used: 0-9,2 “looks at” 1 layers used: 0-9,distance, in number of

11、cells,distance, in number of cells,“dirt”,“dirt”,1,2,2,1,Note: “Grads” search was cylindrical (as opposed to directional, in theta and phi),Event 184,Neighborhood Grads,Evt# 184 ( A case of extreme asymmetry ) Gamma1: ThetaBin=419 PhiBin=1603 Reco Energy=9.45 MC Energy = 9.98Gamma2: ThetaBin=414 Phi

12、Bin=1588 Reco Energy=0.050 MC Energy=0.021,Event 184,1,2,2,layers 09 cannot detect gamma_2,layers 1019,layers 2029,layers 09,Rmin1 = 10 , Rmax1 = 13 hsize1 = 13 , DeltaR = 15,(*) clearly, any level of noise here turns this into an inefficiency.,(*),Event 284,Evt# 284 (Gamma_2 is a late shower)Gamma1

13、: ThetaBin=419 PhiBin=1492 Reco Energy=7.91 MC Energy=8.04Gamma2: ThetaBin=417 PhiBin=1483 Reco Energy=2.15 MC Energy=1.96,Neighborhood Grads,Event 284,Rmin2 = 4 Rmax2 = 8 DeltaR = 9,Rmin1 = 4 Rmax1 = 7 hsize1 = 5,1,layers 09 cannot detect gamma_2,1,2,2,1,layers 2029,layers 1019,layers 09,1,2,EM Sho

14、wer Characteristics in the “SD” Detector Model,The next two slides display shower profiles for electrons and gammas respectivelyProfiles are averaged over samples of 5000 monochromatic single particles (10GeV)All plots are longitudinal profiles, where the x-axis labels the layer number- Energy depos

15、ition per layer - Layer hosting hottest cell- Number of hits per layer - Layer mean square radius (*),(*) definition;,A. Maciel (NIU), Simulation & E-flow Wshop , NIU, November 7-9, 2002,SD 10GeV single s in Em-Cal (5k evts.),GeV,n.of cells,energy-weighted transverse shower radius (in #of cells),shower-max,hit-max,A. Maciel (NIU), Simulation & E-flow Wshop , NIU, November 7-9, 2002,