Photon purity measurement on JF17 Di jet sample using Direct photon working Group ntuple Z.Liang (Academia Sinica,TaiWan) 6/24/20161.

Slides:



Advertisements
Similar presentations
Recent Results on Radiative Kaon decays from NA48 and NA48/2. Silvia Goy López (for the NA48 and NA48/2 collaborations) Universitá degli Studi di Torino.
Advertisements

1 Search for the Flavor-Changing Neutral-Current Decay,   → p     HyangKyu Park University of Michigan, Ann Arbor for the HyperCP collaboration.
INTRODUCTION TO e/ ɣ IN ATLAS In order to acquire the full physics potential of the LHC, the ATLAS electromagnetic calorimeter must be able to identify.
Implementation of e-ID based on BDT in Athena EgammaRec Hai-Jun Yang University of Michigan, Ann Arbor (with T. Dai, X. Li, A. Wilson, B. Zhou) US-ATLAS.
1 N. Davidson E/p single hadron energy scale check with minimum bias events Jet Note 8 Meeting 15 th May 2007.
Top Turns Ten March 2 nd, Measurement of the Top Quark Mass The Low Bias Template Method using Lepton + jets events Kevin Black, Meenakshi Narain.
Kevin Black Meenakshi Narain Boston University
Status of  b Scan Jianchun Wang Syracuse University Representing L b scanners CLEO Meeting 05/11/02.
Update from the Photons + MET Group Bruce Schumm UC Santa Cruz / SCIPP 26 August 2010 SUSY/MET Meeting.
1 N. Davidson, E. Barberio E/p single hadron energy scale check with minimum bias event Hadronic Calibration Workshop 26 th -27 th April 2007.
Analysis Meeting – April 17 '07 Status and plan update for single hadron scale check with minimum bias events N. Davidson.
In order to acquire the full physics potential of the LHC, the ATLAS electromagnetic calorimeter must be able to efficiently identify photons and electrons.
1 N. Davidson Calibration with low energy single pions Tau Working Group Meeting 23 rd July 2007.
Sourav Tarafdar Banaras Hindu University For the PHENIX Collaboration Hard Probes 2012 Measurement of electrons from Heavy Quarks at PHENIX.
Using Track based missing Et tools to reject fake MET background Zhijun Liang,Song-Ming Wang,Dong liu, Rachid Mazini Academia Sinica 8/28/20151 TWiki page.
1 The Study of D and B Meson Semi- leptonic Decay Contributions to the Non-photonic Electrons Xiaoyan Lin CCNU, China/UCLA for the STAR Collaboration 22.
1 Shower maximum detector (SMD) is a wire proportional counter – strip readout detector using gas amplification. SMD is used to provide a spatial resolution.
JSPS Research Fellow / University of Tsukuba T. Horaguchi Oct for HAWAII /10/15HAWAII
Outline: (1) The data sample (2) Some news on the analysis method (3) Efficiency revised (4) Background revised (5) Data: spectrum + “phi-curve”
Preliminary comparison of ATLAS Combined test-beam data with G4: pions in calorimetric system Andrea Dotti, Per Johansson Physics Validation of LHC Simulation.
C.ClémentTile commissioning meeting – From susy group talk of last Wednesday  Simulation and digitization is done in version (8) 
Photon-jet reconstruction with the EEMC – Deuxième Partie Pibero Djawotho Indiana University Cyclotron Facility June 18, 2008 STAR.
Lepton efficiency & fake rate Yousuke Kataoka University of Tokyo Content definitions of leptons p2 efficiency and fake rate for SU3 ( ) p3, p4.
19/11/2010Inclusive Photon studies at ATLAS1 Inclusive photon studies at ATLAS L. Carminati (Universita’ e INFN Milano) On behalf of the ATLAS Collaboration.
W + /W - and l + /l - A Means to investigate PDFs T. Schörner-Sadenius, G. Steinbrück Hamburg University HERA-LHC Workshop, CERN, October 2004.
Study of Standard Model Backgrounds for SUSY search with ATLAS detector Takayuki Sasaki, University of Tokyo.
1 Direct Photon Studies in the ATLAS Detector Ivan Hollins 11/04/06 The University of Birmingham.
Higher harmonics flow measurement of charged hadrons and electrons in wide kinematic range with PHENIX VTX tracker Maki KUROSAWA for PHENIX collaboration.
1 Single top in e+jets channel Outline : - Data and MC samples - Overview of the analysis - Loose and topological cuts - MC efficiencies and expected number.
CALOR April Algorithms for the DØ Calorimeter Sophie Trincaz-Duvoid LPNHE – PARIS VI for the DØ collaboration  Calorimeter short description.
Software offline tutorial, CERN, Dec 7 th Electrons and photons in ATHENA Frédéric DERUE – LPNHE Paris ATLAS offline software tutorial Detectors.
Γ +Jet Analysis for the CMS Pooja Gupta, Brajesh Choudhary, Sudeep Chatterji, Satyaki Bhattacharya & R.K. Shivpuri University of Delhi, India.
Update on WH to 3 lepton Analysis And Electron Trigger Efficiencies with Tag And Probe Nishu 1, Suman B. Beri 1, Guillelmo Gomez Ceballos 2 1 Panjab University,
Fast Simulation and the Higgs: Parameterisations of photon reconstruction efficiency in H  events Fast Simulation and the Higgs: Parameterisations of.
Search for High-Mass Resonances in e + e - Jia Liu Madelyne Greene, Lana Muniz, Jane Nachtman Goal for the summer Searching for new particle Z’ --- a massive.
Issues with cluster calibration + selection cuts for TrigEgamma note Hardeep Bansil University of Birmingham Birmingham ATLAS Weekly Meeting 12/08/2010.
Calo Calibration Meeting 29/04/2009 Plamen Hopchev, LAPP Calibration from π 0 with a converted photon.
28/4/2006Chris Collins-Tooth tth, (h → bb) with EventViews Chris Collins-Tooth, Christian Shaw 03-May-2006.
Jet + Isolated Photon Triple Differential Cross Section Nikolay Skachkov: “Photon2007”, Paris, 9-13 July 2007 DO Measurement of Triple Differential Photon.
Tracking Variable Study Follow up Ryan Kelley Boris Mangano Vivek Sharma.
Update on Diffractive Dijets Hardeep Bansil University of Birmingham 12/07/2013.
L1Calo EM Efficiencies Hardeep Bansil University of Birmingham L1Calo Joint Meeting, Stockholm 29/06/2011.
Trigger study on photon slice Yuan Li Feb 27 th, 2009 LPNHE ATLAS group meeting.
Elliptic flow of D mesons Francesco Prino for the D2H physics analysis group PWG3, April 12 th 2010.
10 January 2008Neil Collins - University of Birmingham 1 Tau Trigger Performance Neil Collins ATLAS UK Physics Meeting Thursday 10 th January 2008.
Jet + Isolated Photon Triple Differential Cross Section Nikolay Skachkov: “Photon2007”, Paris, 9-13 July 2007 DO Measurement of Triple Differential Photon.
Current Status of e-ID based on BDT Algorithm Hai-Jun Yang University of Michigan (with X. Li, T. Dai, A. Wilson, B. Zhou) BNL Analysis Jamboree March.
Photon Reconstruction Efficiencies in Higgs → γγ Events Neil Cooper-Smith RHUL ATLAS UK Higgs Meeting - Durham 11/01/07.
LCFI physics studies meeting, 7 th December 04Sonja Hillertp. 1 Charge efficiency and leakage rates  purity vs efficiency plots give only part of the.
E. Soldatov Tight photon efficiency study using FSR photons from Z  ll  decays E.Yu.Soldatov* *National Research Nuclear University “MEPhI”
Search for Standard Model Higgs in ZH  l + l  bb channel at DØ Shaohua Fu Fermilab For the DØ Collaboration DPF 2006, Oct. 29 – Nov. 3 Honolulu, Hawaii.
Direct Photon v 2 Study in 200 GeV AuAu Collisions at RHIC Guoji Lin (Yale) For STAR Collaboration RHIC & AGS Users’ Meeting, BNL, June 5-9.
Zvi Citron Correlations Between Neutral Bosons and Jets in Pb+Pb Collisions at 2.76 TeV with the ATLAS Detector Zvi Citron for the ATLAS Collaboration.
Search for a new light boson in  decays J.Stepaniak, M.Berłowski, NCBJ Warsaw For WASA-at-COSY Collaboration Meson2014,Krakow
Methodology and examples to determine fake rate separate signal from background Using fit on sideband. Using independent control sample.
Some introduction Cosmics events can produce energetic jets and missing energy. They need to be discriminated from collision events with true MET and jets.
Searches for double partons
Converted photons efficiency
Converted photons efficiency
Measurement of the γ,W,Z with ATLAS for the ATLAS Collaboration
Validation of valid3 samples Zee(106050) and Jpsiee(105751)
NIKHEF / Universiteit van Amsterdam
Update of Electron Identification Performance Based on BDTs
Plans for checking hadronic energy
 discrimination with converted photons
VBF Higgs at LHC 方亚泉 高能物理研究所 April 2nd, 2018.
Performance of BDTs for Electron Identification
Samples and MC Selection
 discrimination with converted photons
Ivan Hollins 08/05/06 The University of Birmingham
Presentation transcript:

Photon purity measurement on JF17 Di jet sample using Direct photon working Group ntuple Z.Liang (Academia Sinica,TaiWan) 6/24/20161

Introduction to Direct photon ntuple Made by HiggsAnalysisUtils package, size 2.5k/evt Ntuple maker can run in DPD and AOD format Official tools are embedded into ntuple dumper,conversion recovery tool Truth matching information are done for photon and electron and jets objects. Wiki page : duction1 duction1 Central production of these ntuple for JF17 Dijet sample and photon jet sample are available on grid, more detailed are in the wiki page Trigger information are also included. 6/24/20162

Outline Photon purity measurement in JF17 dijet sample (Direct photon group ntuple) Two method to estimate photon purity after IsEM cut ph_isEM%65536==0) and track isolation cut (ph_PtIsolationCone==0)  Method1 : two dimensional side band  Method2 : photon conversion Comparison between method 1 and 2 and truth information 6/24/20163

Pt distribution of photon candidate with truth information Photon Pt[GeV] 6/24/20164 Photon candidate PT after IsEM cut ( ph_isEM%65536==0 ) Photon candidate Pt after IsEM cut ( ph_isEM%65536==0 ) and track isolation (ph_PtIsolationCone==0) Purity[Pt]= (Area under red curve )/ (Area under blue curve+ Area under red curve )[Pt] The question is : how we estimate the purity from data without any truth information ? Truth photon : photon from hard process and quark/gluon bremstrahlung

Method 1: 2 dimensional sideband method Select 2 different photon ID variable(ph_PtIsolation and ph_shwr_fracm), Try something like 2 dimensional sideband,3 control region, 1 signal region. Try to make use of information of 3 control region to get number of events in signal region. 6/24/20165

Photon ID variable 6/24/20166 Ph_shwr_fracm: fraction of Energy in LAr strip layer outside core, in formular [E(+-7)-E(+-3)]/E(+-7) Ph_PtIsolationCone: scalar sum of track pt (DR<0.3) around photon cluster Ph_shwr_fracm Ph_PtIsolationCone[GeV] Plots of Barrel region,Pt of photon >15GeV

Correlation between these two variables 6/24/ Ph_shwr_fracm Plots of Barrel region,Pt of photon >15GeV Ph_shwr_fracm is not sensitive to isolation cut for fake photon object Meaning that these two variables are not correlated to each other for fake photon object.

Control region and signal region 6/24/20168 C AD B Control region A : ph_shwr_fracm> 1.2* isemcut value && track isolation <1GeV Control region B : ph_shwr_fracm 2GeV Control region D : ph_shwr_fracm> 1.2* isemcut value && track isolation >2GeV Signal region C : ph_shwr_fracm< isemcut value && track isolation <1GeV

Purity measurement Assume number of background in region all region is NA’,NB’,NC’,ND’ Since two ID variables are not correlated for fake photon object. So NA’/NC’=ND’/NB’ Our signal (real photon ) only in region C,assume NC’’ is number of signal events in region C So NC=NC’+NC’’ Purity =NC’’/NC=1-NC’/NC=1-NA’*NB’/ND’/NC If A,D,B dominated by background (NB’=NB,ND’=ND,NA’=NA), purity=1-NA*NB/ND/NC 6/24/ C AD B

Validation of this method using truth information Pt~[5GeV,10GeV] 6/24/ Total eventsReal Photon (N’’)Truth fake photon (N’) Control region A (NA’’)4762 (NA’) Control region B (NB’’)55930 (NB’) Signal region C (NC’’)17491 (NC’) Control region D (ND’’)16146 (ND’) The truth information shows: Background events in region A B D is much at least 20 time larger than signal events NA’>>NA’’ and NB’>>NB’’ This formular for background events NA’/NC’=ND’/NB’ can be validated Table for photon candidate Pt~[5GeV,10GeV] C AD B

Validation of this method using truth information Pt~[30GeV,35GeV] 6/24/ Total eventsReal Photon (N’’)Truth fake photon (N’) Control region A1847 (NA’’)177 (NA’) Control region B (NB’’)1135 (NB’) Signal region C (NC’’)669 (NC’) Control region D3380 (ND’’)338 (ND’) The truth information shows: Background events in region A B D is much at least 20 time larger than signal events NA’>>NA’’ and NB’>>NB’’ This formular for background events NA’/NC’=ND’/NB’ can be validated Table for photon candidate Pt~[30GeV,35GeV] C AD B

Purity of method 1 for photon in barrel region(|eta|<1.37) 6/24/ purity Red: truth information Blue :method 1 measurement Photon Pt[MeV] Purity measured by method 1 increase with Pt, consistent with truth information

Method 2: conversion method After all IsEm cut and track isolation cut main background of photon is pi0 or other neutral meson They tend to decay to two photon or more. And the key point of method 2 is that two photon have more possibility to converse into at least one electron. Define converted fraction ( called it F ) is the probability of at least one electron was created by conversion of neutral em object before LAr presample layer F(data)=(1-purity)*F(fake photon) +purity*F(real photon ) 6/24/201613

Converted fraction Measurement using LAr presample layer 6/24/ We can measure conversion fraction using Pre-sample layer of Lar EM Calo. If neutral EM object converted into at least one electron before LAr Calo, it deposit its energy in presample layer like a landau distribution If not, neutral EM object can not loss energy by ionization,so energy loss is close to zero in pre-sample layer Ph_ES0/ph_E Peak Created by Converted neutral EM object Peak Created by Non-Converted neutral EM object

Template of Converted component In order to separate this two Component, we need template for One of the component to make a fit We can use inner detector vertexing Information to get this template The best choice of template for Converted component is photon candidate Matched with single track or double track conversion vertex( ph_Conv%10==1) 6/24/ Ph_ES0/ph_E Converted component Non-Converted component Ph_ES0/ph_E Ph_isEM%65536==0 && Track isolation ==0 && Ph_Conv%10>=1 Ph_isEM%65536==0 && Track isolation ==0 Template distribution for converted Component of ph_ES0/ph_E

Calculate converted fraction 6/24/ Converted neutral EM object Non-Converted neutral EM object Ph_ES0/ph_E 1.Plot ph_ES0/ph_E for Photon candidated That pass all the cuts 2. Get template distributon Of converted component Plot ph_ES0/ph_E with the cut ph_Conv%10==1 3. Fit data point of step 1 using Template from step 2 with the Help of TFractionFitter in Root 4. Conversion fraction F= red /(red+blue)

6/24/ With Truth info Real photon With Truth info Real Fake photon Data info(without truth ) photon candidate pass Isem and isolation cut Ph_ES0/ph_E

Control sample In real data,we can not trust distribution of Real Fake photon from MC simulation since jet simulation is not so easy. possible choice is to use a control sample which is dominated by Fake photon object Define a control sample with photon candidates which pass IsEM (itune1) cut but fail track isolation cut, This control sample is dominated by neutral meson background (has been validated in slide 9,10) Replace the third column of previous slide with control sample distribution 6/24/201618

6/24/ With Truth info Real photon Data info(without truth ) photon candidates pass Isem but fail isolation cut Real Fake photon Data info(without truth ) photon candidate pass Isem and isolation cut Ph_ES0/ph_E

Ph_ES0/ph_E VS Eta 6/24/ Eta Ph_ES0/ph_E In very central region |eta|<0.8 (As indicated by red line) Ph_ES0/ph_E do not have strong eta dependence in that region,relatively easy to handle. So we try method 2 in region |eta|<0.8 at very beginning. All the plots for method 2 in previous page is made in this very central region

Conversion fraction and purity in barrel region (|eta|<0.8) Pt of photon F(data)F(control sample)F(real phton) Truth info from MC Purity 5-10GeV GeV GeV GeV GeV GeV GeV /24/201621

(1-purity) of method 1,2 for photon in barrel region (|eta|<0.8) 6/24/ Red: truth information Blue :method 1 measurement Black :method 2 measurement Photon Pt [MeV]

Conversion fraction and purity in barrel region (|eta|<1.37) Pt of photon F(data)F(control sample)F(real phton) Truth info from MC Purity 5-10GeV GeV GeV GeV GeV GeV GeV /24/201623

Purity of method 1,2 for photon in barrel region (|eta|<1.37) 6/24/ Red: truth information Blue :method 1 measurement Black :method 2 measurement Photon Pt [MeV] try method 2 in full barrel, seems still reasonable

Summary Purity measurement Method 1 and 2 is more or less consistent with truth purity Purity increase with photon Pt increase. Next step : Running out of events in high Pt region,may try other high Pt dijet sample, or change binning of photon Pt to test methods in high Pt region Try Calo base isolation instead of track isolation 6/24/201625

Backup F(real photon)=1-Exp[-X/X0] F(fake photon from neutral meson)=1-Exp[-m*X/X0], where m is effective background neutral meson decay multiplicity. Since most of background is pi0, m is close to 2,but not exactly, m is lower than 2 in low pt region since two photon from pi0 is not collinear in low pt region m is larger than 2 in high pt region since back scattering will increase effective multiplicity Try to test method in endcap region 6/24/201626