Wjets bkg. Estimation in WW analysis Jun Gao , Yanwen Liu, Emmanuel - - PowerPoint PPT Presentation

▶
wjets bkg estimation in ww analysis
SMART_READER_LITE
LIVE PREVIEW

Wjets bkg. Estimation in WW analysis Jun Gao , Yanwen Liu, Emmanuel - - PowerPoint PPT Presentation

Feb 18, 2023 243 likes •485 views

Wjets bkg. Estimation in WW analysis Jun Gao , Yanwen Liu, Emmanuel Monnier U niversity of S cience and T echnology of C hina & C entre de P hysique des P articules de M arseille 6 March 2014 1 Introduction W+jets background for WW

slide-1
SLIDE 1

Wjets bkg. Estimation in WW analysis

1

Jun Gao , Yanwen Liu, Emmanuel Monnier University of Science and Technology of China &Centre de Physique des Particules de Marseille

6 March 2014

slide-2
SLIDE 2

Introduction

6 March 2014 2

  • W+jets background for WW analysis
  • Using fake factor method:

– Control region : tight lepton + loose lepton(non-tight) – W+jet contribution := Control region * “fake factor”.

  • Fake factor definition : f= Numerator /Denominator

numerator : tight lepton Denominator :loose lepton estimated using a jet-enriched sample.

slide-3
SLIDE 3

Main features/concerns

6 March 2014 3

1) The change of triggers in WW analysis ee/mm : Now using di-lepton trigger em: Now using the OR of single lepton trigger and em trigger Trigger may already have requirement on the quality of leptons. Electron-fake is the main concern. 2) Sample dependence (dijet-Wjet) Also take into account the possible bias introduced by event selection specifically used in nominal analysis.

slide-4
SLIDE 4

1) The choise of the trigger in fake factor study is related to the triggers used in nominal analysis 2) Triggers used in the Fake factor study

  • g20_etcut and g24_etcut

1)prescale 6.2e-6 and 1.0e-4 2) mainly for em(fake e+real muon) channel , will see later. 3) loose electron : fail medium

  • E22vh_loose1

1)Prescale 3.0e-3 2) for ee channel , di-electron trigger 3) loose electron : pass loose fail medium

  • Mu15
  • 1) Prescale 3.0e-3

4

Trigger

slide-5
SLIDE 5

Lepton Definition : Electron

6 March 2014 5

Electron Numerator Denominator Pt, eta, Nhits 20 GeV, |eta|<2.47 , exclude [1.37,1.52],Nhits>=4 Author , object quality author = 1 or 3 , (el_OQ & 1446) == 0 <0.4 mm <3 <3 Etcone30/Et <0.16 <0.28 <0.3 Ptcone30/Et <0.12 (20<pt<25) <0.16 (pt>25) <0.1 <0.16 IsEM Very tight (likelihoodID) (See the text below) Fail Numerator selection For e22vh_loose1 . Pass Veryloose_eID or loosePP fail Medium_eID For g20/24_etcut, fail Medium_eID

slide-6
SLIDE 6

Lepton Definition : Muon

6 March 2014 6

Muon Numerator = tight lepton Denominator Pt, eta, ID track combined muon 20 GeV, |eta|<2.4 Author , object quality author = 6 <1 mm <3 Cut removed Etcone30/Et <0.20 && <0.014*pt-0.15 <0.18 if 20<pt<25 <0.30 if pt>25 <0.25 if 20<pt<25 < 0.3 if pt>25 Ptcone30/Et <0.15 && <0.01pt-0.105 <0.12 removed Fail Numerator selection

slide-7
SLIDE 7

Jet-enriched fake lepton selection

6 March 2014 7

  • Lepton-jet overlap removal (compared with Wjets CR in the next slide)

1) tight electron/jet 2) jet/tight muon 3) loose muon/jet

  • Suppress W&Z contribution

Veto events if any lepton if mT > 30GeV

Veto events if in the Z-mass window |M(ll)-Mz|<13 GeV Veto events if MET_RefFinal>30 The remaining W&Z contribution is subtracted by MC estimation

  • f= Numerator /Denominator

Numverator = N(tight, data) – N(tight, MC ) Denominator = N(loose,data) – N(loose , MC)

slide-8
SLIDE 8
  • The OR stategy not exactly the same as jet-enriched CR.
  • The event selection requires jet-veto . The requirement of ‘0 jet’ also affected by

OR

6 March 2014 8

slide-9
SLIDE 9

6 March 2014 9

Pt distribution ( g20(24)_etcut )

  • Loose definition: !Medium
  • Tight electron : the contamination (mainly from Zee, Wenu)

will be subtracted

  • More reliability on MC subtraction in High-pT bins
slide-10
SLIDE 10

6 March 2014 10

Pt distribution ( e22vh_loose1 )

  • Loose definition: pass loose fail medium
  • Tight: Zee, Wenu.
  • Dut to trigger, Pt arount 22 GeV shows fluctuation.
slide-11
SLIDE 11
  • Central value of fake factor and the main

systematics

6 March 2014 11

slide-12
SLIDE 12

6 March 2014 12

Central value (low/high eta bins) Top:g20(24)etcut mid: e22vh_loose1 bottom :mu15

slide-13
SLIDE 13

Left: g20_etcut Middle:g24_etcut Right: e22vh_loose1

  • g20_etcut , <mu> is mainly within [10,25]

6 March 2014 13

<mu> distribution

derived from ATLAS Lumi calculator.

slide-14
SLIDE 14

6 March 2014 14

Left: g20(24)_etcut Right: e22vh_loose1 Pile Up

Check fake factor with averageIntPerXing < 20 or >20

slide-15
SLIDE 15

6 March 2014 15

1)Fake factor is estimated in jet-enriched CR:

i)Veto events if any lepton if mT > 30GeV

ii)Veto events if in the Z-mass window |M(ll)-Mz|<13 GeV iii)veto events if MET_RefFinal>30

Sample dependence

2) Compare dijetMC and Wjets MC When studying Wjets MC Not doing i) ii) iii) listed above. Use the truth information instead. 3) in order to truly estimate the difference between jet-enriched CR and Wjet CR. We inverse the cut iii) . For Wjets MC Wjets MC MET dependence

slide-16
SLIDE 16

6 March 2014 16

Sample dependence(left : photon trigger Right:single electron trigger)

Wjets MC: truth info + MET>30 + Wjet CR OR + N(goodjet)=0 Wjets MC: truth info + MET>30

For Wjets MC 1) Change the OR strategy in Fake factor CR to be the same as that in Wjets CR 2) Require jet-veto In this way two CRs becomes more similar. Wjets fake factor looks stable between two plots. For comparison

slide-17
SLIDE 17
  • Wjets CR
  • Show Em channel as an example
  • Em trigger together with single lepton trigger

is used

6 March 2014 17

slide-18
SLIDE 18

which trigger is fired? me ( muon is fake) em (electron is fake ) Case 1 em trigger only none 1.5% Case 2 electron trigger only 40% 3.8% Case 3 muon trigger only <1% 54% Case 4 em + electron only 60% 5.3% Case 5 em + muon only none 10.7% Case 6 electron + muon only none 3.1% Case 7 em+electron+muon (all triggers passed) none 21%

6 March 2014 18

Derived from W+jet MC. In most cases ( Red cells in the table. ~100% for me, ~90% for em) we rely on the tight lepton to fire trigger. Choose the fake factor wisely. E.g. In em channel , if the tight muon not firing the trigger while the em trigger is fired, we will use fake factor from e22vh_medium1

For OF channel (either electron or muon is fake)

Want to see which trigger is fired.

slide-19
SLIDE 19

6 March 2014 19

W+jet control region em(electron is fake) channel

tight pt tight eta loose eta loose pt

slide-20
SLIDE 20

6 March 2014 20

W+jet control region me(muon is fake) channel

tight pt tight eta loose eta loose pt

slide-21
SLIDE 21

6 March 2014 21

results

Summary ee mm em (fake e+m & fake m +e) Data-driven result 42.1 +/- 11.8 +/- 24.1 13.8 +/- 5.9 +/- 6.9 147.8 +/-5.6 +/- 78.2

In W+jets CR, the total background is the sum of estimation from data and subtraction of non- W+jets contributions predicted from MC Now comparing results with other mothod

slide-22
SLIDE 22
  • Back-up

6 March 2014 22

slide-23
SLIDE 23
  • Wz inclusive
  • mc12_8TeV.147800.PowhegPythia8_AU2CT10_Wplusenu.merge.NTUP_SMWZ.e1169_s1469_s1470_r3542_r3549_p1328/
  • mc12_8TeV.147801.PowhegPythia8_AU2CT10_Wplusmunu.merge.NTUP_SMWZ.e1169_s1469_s1470_r3542_r3549_p1328/
  • mc12_8TeV.147802.PowhegPythia8_AU2CT10_Wplustaunu.merge.NTUP_SMWZ.e1169_s1469_s1470_r3542_r3549_p1328/
  • mc12_8TeV.147803.PowhegPythia8_AU2CT10_Wminenu.merge.NTUP_SMWZ.e1169_s1469_s1470_r3542_r3549_p1328/
  • mc12_8TeV.147804.PowhegPythia8_AU2CT10_Wminmunu.merge.NTUP_SMWZ.e1169_s1469_s1470_r3542_r3549_p1328/
  • mc12_8TeV.147805.PowhegPythia8_AU2CT10_Wmintaunu.merge.NTUP_SMWZ.e1169_s1469_s1470_r3542_r3549_p1328/
  • mc12_8TeV.147806.PowhegPythia8_AU2CT10_Zee.merge.NTUP_SMWZ.e1169_s1469_s1470_r3542_r3549_p1328/
  • mc12_8TeV.147807.PowhegPythia8_AU2CT10_Zmumu.merge.NTUP_SMWZ.e1169_s1469_s1470_r3752_r3549_p1328/
  • mc12_8TeV.147808.PowhegPythia8_AU2CT10_Ztautau.merge.NTUP_SMWZ.e1169_s1469_s1470_r3542_r3549_p1328/

DY Low Mass samples included

Di-jet MC

  • MUON
  • mc12_8TeV.147910.Pythia8_AU2CT10_jetjet_JZ0W.merge.NTUP_SMWZ.e1126_s1469_s1470_r3542_r3549_p1328/
  • mc12_8TeV.147911.Pythia8_AU2CT10_jetjet_JZ1W.merge.NTUP_SMWZ.e1126_s1469_s1470_r3542_r3549_p1328/
  • mc12_8TeV.147912.Pythia8_AU2CT10_jetjet_JZ2W.merge.NTUP_SMWZ.e1126_s1469_s1470_r3542_r3549_p1328/
  • mc12_8TeV.147913.Pythia8_AU2CT10_jetjet_JZ3W.merge.NTUP_SMWZ.e1126_s1469_s1470_r3542_r3549_p1328/
  • mc12_8TeV.147914.Pythia8_AU2CT10_jetjet_JZ4W.merge.NTUP_SMWZ.e1126_s1469_s1470_r3542_r3549_p1328/
  • mc12_8TeV.147915.Pythia8_AU2CT10_jetjet_JZ5W.merge.NTUP_SMWZ.e1126_s1469_s1470_r3542_r3549_p1328/
  • mc12_8TeV.147916.Pythia8_AU2CT10_jetjet_JZ6W.merge.NTUP_SMWZ.e1126_s1469_s1470_r3542_r3549_p1328/
  • mc12_8TeV.147917.Pythia8_AU2CT10_jetjet_JZ7W.merge.NTUP_SMWZ.e1126_s1469_s1470_r3542_r3549_p1328/
  • mc12_8TeV.129390.Pythia8_AU2CT10_jetjet_JZ0W_mufilter.merge.NTUP_SMWZ.e1318_s1499_s1504_r3730_r3549_p1328/
  • mc12_8TeV.129391.Pythia8_AU2CT10_jetjet_JZ1W_mufilter.merge.NTUP_SMWZ.e1318_s1499_s1504_r3730_r3549_p1328/
  • mc12_8TeV.129392.Pythia8_AU2CT10_jetjet_JZ2W_mufilter.merge.NTUP_SMWZ.e1318_s1499_s1504_r3730_r3549_p1328/
  • mc12_8TeV.129393.Pythia8_AU2CT10_jetjet_JZ3W_mufilter.merge.NTUP_SMWZ.e1318_s1499_s1504_r3730_r3549_p1328/
  • ELECTRON
  • mc12_8TeV.129160.Pythia8_AU2CTEQ6L1_perf_JF17.merge.NTUP_SMWZ.e1130_s1468_s1470_r3542_r3549_p1328/

6 March 2014 23

MC used