wiki:Korea2017-05-CMS-EXO-16-052

Version 44 (modified by Sam, 7 years ago) ( diff )

--

The first MadAnalysis 5 workshop on LHC recasting @ Korea

Analysis 5: CMS-EXO16-052

Students Jeon Dajeong [mail:djeon@cern.ch] , Kim Seulgi [mail:pikachu0617@naver.com], Lee Sang Man [mail:daniel.s.lee4@gmail.com], Watson Ian James [mail:ian.james.watson@cern.ch]
Supervisor(s) Sam Bein, Jory Sonneveld [mail:jory.sonneveld@cern.ch, samuel.bein@gmail.com]

1. Analysis summary

2l+MET

2. Analysis global information

https://github.com/cms-sw/genproductions/tree/master/python/ThirteenTeV

https://gitlab.cern.ch/jsonneve/ma5recasts/blob/master/cms_sus_16_033/plots/plot_saf_file.py

Low-luminosity version of this analysis:

https://inspirehep.net/record/1618045/

it is already included in the PAD!

3. Selection description

3.1 Trigger selection

PS: This part cannot usually be described by the fast-simulation of detector.

3.2 Object definition

Electrons

  • pT > 20 GeV, |eta| < 2.4
  • Medium ID
  • PFiso04 < 0.077 (barrel electrons) or PFiso04 < 0.068) (endcap electrons)

Muons

  • pT > 20 GeV, |eta| < 2.4
  • Tight ID (Loose ID for vetoing 3 leptons events)
  • PFiso04 < 0.15

To get isolation:

double Riso = PHYSICS->Isol->eflow->relIsolation(el,event.rec(),0.4,0.,IsolationEFlow::ALL_COMPONENTS);

Jets

  • pT > 20 GeV, |eta| < 5 (PFloose ID)
  • Identified leptons in {delta}R < 0.4 disregarded
  • We categorize the selected events based on the number of selected jets with pT > 30 GeV for the corresponding requirements

b-jets

  • pT > 20 GeV, |eta| < 2.4
  • CSVv2>0.8484 (medium)

Taus

MET

  • Type-1 corrections are applied ?

3.3 Cut-flow chart

Event Selection:

  • Two same flavored lepton with opposite electric charge ( that pass lepton selection?)
  • el[0] pt > 25 and el[1] pt >20
  • mu[0] pt > 20 and mu[1] pt > 20
  • |dilep m - 91.1876( +- 0.0021)| < 15
  • dilep pt > 60
  • events having more than one jet with pT > 30 GeV - remove
  • b-tagged jet Veto : at least one bjet pt > 20, |eta| < 2.4, CSVv2 < 0.8484 - remove
  • if el[2 onwards] pt > 10 -remove
  • if mu[2 onwards] pt > 5 -remove
  • tau veto: loose id, tau pt > 18. (hadronically decaying tau leptons)

Optimization:

  • For each possible set of selections the full analysis, including the estimation of backgrounds from control samples in data and the systematic uncertainties, is repeated.

After optimization:

  • MET > 100
  • Delta phi ( dilep.pt , ptmiss) > 2.6 rad
  • |MET - dilep.pt| / dilep.pt < 0.4

If event has one jet:

  • Delta phi (jet.pt , ptmiss ) > 0.5
  • Delta R(ll) < 1.8

3.4 Definition of exotic observables

Like SUSY transverse observable. Notice that classical MT2 & MT2W are encoded within the MadAnalysis framework.

3. Detector simulation

Can we use the last release of the Delphes package with the CMS/ATLAS MA5tune-card? or do need to improve the simulation?

4. Validation material

4.1 Monte-Carlo samples

  1. LO/NLO? LO
  2. Model used:

1) dark matter production in a simplified model with a spin-1 mediator A

2) dark matter production in a simplified model with a spin-0 mediator phi

  1. MC generator program (name + version): MAD-GRAPH5_aMC@NLO 2.2.2
  2. Shower program (name + version): PYTHIA 8.205
  3. Tune of Pythia: CUETP8M1

+ The coupling g(kai) is set to one.

For 0.25 < g(q) < 1.0,

the width ranges 1-5% of the mediator mass for g(q) = 0.25,

30-50% of the mediator mass for g(q) = 1.00


Events for ADD extra-dimension scenario

LO, EFT implementaion in PYTHIA 8

M(D) = 1, 2 and 3 TeV, each with n = 2, 3, 4, 5, 6, 7


The events for the unparticle model

LO, PYTHIA 8, tune CUETP8M1

assuming cutoff scale ㅅ(U) =15TeV


ZH & gluon-gluon fusion = WZ & ZZ processes

4.1.1 Running Madgraph

import DMsimp_s_spin0
generate p p > xd xd~ z [QCD]
add process p p > xd xd~ z j [QCD]
output DM_spin0_noreallythistime
launch

Turn on madspin and the showering. Then fix:

madspin_card.dat -> set "z > l+ l-"

run_card.dat -> can set number of events, but also change shower to PYTHIA8, ickkw to 3 (FxFx merging)

shower_card.dat -> Qcut to 15 (50?) and njmax to 1 (since were generating 0 + 1 extra partons)

param_card.dat -> set the masses of Y0 (or Y1 for spin 1), the mediator, and Xd, the dark matter particle (the paper uses MY0=500, MXD=150GeV)

we will also have to set the pythia settings to match those of CMS (UE tune, etc.).

4.2 Reference plots / cut-flows

4. Validation results

5. Reinterpretation (optional)

6. Questions / issues

Note: See TracWiki for help on using the wiki.