DESY Hbb Analysis Framework
Analysis/MssmHbb

Extension of the Analysis Framework for the MSSM Hbb all-hadronic and semileptonic analyses in CMS

See also the code documentation Doxygen page

This package inherits from the Analysis/Tools package

Installation

The codes as well as the ntuples are independent of CMSSW. However, in order to compile it uses scram. So the latest version in any architecture should be fine.

1 cmsrel CMSSW_X_Y_Z
2 cd CMSSW_X_Y_Z/src
3 cmsenv
4 
5 git clone https://github.com/desy-cms/analysis-tools.git Analysis/Tools
6 git clone https://github.com/desy-cms/analysis-calibrations.git Analysis/Tools/data/calibrations
7 git clone https://github.com/desy-cms/analysis-ntuples.git Analysis/Tools/data/ntuples
8 git clone https://github.com/desy-cms/analysis-mssmhbb.git Analysis/MssmHbb
9 
10 scram b -j4 USER_CXXFLAGS="-Wno-misleading-indentation"
11 hash -r

:zzz: The USER_CXXFLAGS="-Wno-misleading-indentation" prevents a large number of warnings from misleading identation in modules of the boost library.

:zzz: More information about the additional packages, calibration, ntuples etc can be found in the Analysis/Tools https://github.com/desy-cms/analysis-tools/blob/develop/README.md "README.md".

Running the analysis

A macro with the full analysis workflow for both semileptonic and all-hadronic channels can be found in Analysis/MssmHbb/bin/MssmHbbAnalysis.cc

Semileptonic channel

The configuration file for the semileptonic analysis with 2017 data is available in the file Analysis/MssmHbb/test/mssmhbb_semilep_2017.cfg.

The semileptonic standard selection

:warning: Unfortunately Github markdown does not render Latex equations :rage:

All-hadronic channel

The configuration file for the semileptonic analysis is available in the file Analysis/MssmHbb/test/mssmhbb_semilep_2017.cfg.

The all-hadronic standard selection

TO DO

:warning: Unfortunately Github markdown does not render Latex equations :rage:

Main macro

The structure of the macro is as follows:

#include "Analysis/MssmHbb/interface/MssmHbbAnalyser.h"
using namespace analysis;
using namespace analysis::tools;
using namespace analysis::mssmhbb;
int main(int argc, char ** argv)
{
TH1::SetDefaultSumw2(); // proper treatment of errors when scaling histograms
MssmHbbAnalyser mssmhbb(argc,argv);
for ( int i = 0 ; i < mssmhbb.nEvents() ; ++i )
{
if ( ! mssmhbb.event(i) ) continue; // read event, run selection/json
if ( ! mssmhbb.triggerSelection () ) continue; // trigger
if ( ! mssmhbb.preselection() ) continue; // preselection
if ( ! mssmhbb.jetCorrections() ) continue; // jet corrections
if ( ! mssmhbb.muonSelection() ) continue; // muon
if ( ! mssmhbb.jetSelection() ) continue; // jets
if ( ! mssmhbb.muonJet() ) continue; // muon-jet association
if ( ! mssmhbb.btagSelection() ) continue; // btagging
mssmhbb.endSelection();
}
} // end main

Compared to the Analysis/Tools example this is much simplified, for the individual selections were grouped and included in methods of the MssmHbbAnalyser class. However some flexibility to re-order the workflow is still allowed. Of course, the user can also replace any of the if statements above and enable more detailed steps of the selections, or add other steps to the workflow.

Event selection detailed description

  1. MSSMHBBAnalyser::event(i)

When calling the method event(i), the entry i is read from the TTree of the ntuple. If running on data, the event is checked whether it is valid from a list of certified data given by the provided JSON file.

When running on Monte Carlo, generated event weights are applied. Pileup re-weighing is also applied if re-weigh file is provided in the configuration file in the [Corrections] section, e.g.:

1 [Corrections]
2 Pileup.reweight = calibrations/2017/pileup/PileupWeight_Run2017_Mix_2017.root
  1. MSSMHBBAnalyser::triggerSelection()

It returns true if the event passes both the HLT and L1 triggers, which are given in the configuration file section and whose results are in the TriggerResults collection, thus

1 [Trigger]
2 results = TriggerResults
3 hltPath = HLT_Mu12_DoublePFJets40MaxDeta1p6_DoubleCaloBTagCSV_p33_v
4 l1Seed = L1_Mu12er2p3_Jet40er2p3_dR_Max0p4_DoubleJet40er2p3_dEta_Max1p6
  1. MSSMHBBAnalyser::preselection()

This method will prepare a vector of muons and a vector of jets that pass the identification criteria, namely muon id for muons, and jet id and jet pileup id for jets. It will return false if any of those vectors has zero entries.

The identification criteria is defined in the configuration file as, for example

1 [Jets]
2 id = tight
3 puId = loose
1 [Muons]
2 id = tight
  1. MSSMHBBAnalyser::jetCorrections()

This method will apply b-jet energy regression corrections to all the jets in the event, if requested in the configuration file:

1 [Corrections]
2 Jets.bRegression = true

It will also smear the jet $p_{t}$ according to the jet energy resolution scale factors, if the relevant calibration files are provided in the configuration, e.g.:

1 [Corrections]
2 Jets.bRegression = true
3 Jets.jerPtRes = calibrations/2017/jer/Fall17_V3_MC_PtResolution_AK4PFchs.txt
4 Jets.jerSF = calibrations/2017/jer/Fall17_V3_MC_SF_AK4PFchs.txt

In the jetCorrections method the returned value is always true.

  1. MSSMHBBAnalyser::muonSelection()

The muonSelection method selects events according to the minimum number of muons required. It also skims the vector of muons to contain only muons with the required kinematics and matching to muon trigger objects.

Example of the requirements give in the configuration:

1 [Muons]
2 ptMin = 13.
3 etaMax = 2.2
1 [Trigger.Objects]
2 Muons.L1 = hltL1Mu12er2p3Jet40er2p3dRMax0p4DoubleJet40er2p3dEtaMax1p6_l1muon
3 Muons.L3 = hltL3fL1sMu12Dijet40L1f0L2f8L3Filtered12
  1. MSSMHBBAnalyser::jetSelection()

The jetSelection method selects events according to the minimum number of jets required. It also checks whether the event fulfils the required kinematic selection and matching of the two leading jets to jet trigger objects. It returns false if any of the requirements is not fulfilled.

For a selection, in which:

the configuration reads

1 [Jets]
2 ptMin = 60.
3 ptMin = 50.
4 ptMin = 30.
5 etaMax = 2.2
6 etaMax = 2.2
7 etaMax = 2.2
8 ...

:warning: Notice that the order matters! First ptMin entry is for the leading jet, second ptMin is for the second leading jet, ...

1 [Trigger.Objects]
2 Jets.L1 = hltL1Mu12er2p3Jet40er2p3dRMax0p4DoubleJet40er2p3dEtaMax1p6_l1jet
3 Jets.Calo = hltDoubleCaloBJets30eta2p3
4 Jets.PF = hltDoublePFBJets40Eta2p3
  1. MSSMHBBAnalyser::muonJet()

The muonJet method returns true if either of the two leading jets contains at least one of the selected muons.

  1. MSSMHBBAnalyser::btagSelection()

If the leading jets fulfil the required b-tag working point selection and matching of the two leading jets to the corresponding trigger objects, the btagSelection method returns true. The requirements are controlled in the configuration file. For example, select events with three leading jets passing the medium working point of the deepflavour algorithm.

1 [BTag]
2 nMin = 3
3 wp = medium
4 wp = medium
5 wp = medium
6 algorithm = deepflavour

where wp if the working point of each required b-tagged jet.

:warning: Notice that the order matters! First wp entry is for the leading jet, second wp is for the second leading jet, ...

The values of the WP's must be given in the configuration file:

1 [BTag]
2 loose = 0.0521
3 medium = 0.3033
4 tight = 0.7489

Because in the MSSM Hbb analysis, the background modelling is performed in a control region, in which on the b-tagged jets has its b-tagging requirement reversed, one can apply this possibility in the configuration files by setting the flag

1 [Info]
2 signalRegion = false

and defininig reverse b tag WP and the rank of the jet to have the reverse b-tag applied, e.g. reverse b tag of the 3rd leading jet with loose WP:

1 [BTag]
2 revWP = loose
3 revBJet = 3

The matching to trigger object is done given the appropriate trigger object collection, e.g.

1 [Trigger.Objects]
2 BTag.Calo = hltBTagCalo30x8CSVp0p92DoubleWithMatching

Configuration file

The structure of the configuration file is defined by the Analysis/Tools package. More details can be found in the Analysis/Tools README.md.

Workflow

The workflow of the main macro for DATA is

1 +----------------------------------------------------------------------------------------------+
2 | Workflow DATA |
3 +----------------------------------------------------------------------------------------------+
4 | 0 - Total events read |
5 | 1 - Certified data: Cert_294927-306462_13TeV_EOY2017ReReco_Collisions17_JSON.txt |
6 | 2 - Using Jet collection: MssmHbb/Events/updatedPatJets |
7 | 3 - Using Muon collection: MssmHbb/Events/slimmedMuons |
8 | 4 - HLT_Mu12_DoublePFJets40MaxDeta1p6_DoubleCaloBTagCSV_p33_v |
9 | 5 - L1_Mu12er2p3_Jet40er2p3_dR_Max0p4_DoubleJet40er2p3_dEta_Max1p6 |
10 | 6 - MuonId: tight |
11 | 7 - JetId: tight |
12 | 8 - JetPileupId: loose |
13 | 9 - b jet energy regression |
14 | 10 - NMuons >= 1 |
15 | 11 - Muons selected: pt > 13.0 GeV and |eta| < 2.2 |
16 | 12 - Muons selected: Online muon matching (L1 and L3) |
17 | 13 - NJets >= 3 |
18 | 14 - Jet 1: pt > 60.0 GeV and |eta| < 2.2 |
19 | 15 - Jet 1: online jet match |
20 | 16 - Jet 2: pt > 50.0 GeV and |eta| < 2.2 |
21 | 17 - Jet 2: online jet match |
22 | 18 - Jet 3: pt > 30.0 GeV and |eta| < 2.2 |
23 | 19 - MSSMHbb Semileptonic: Jet-muon association |
24 | 20 - Jet 1: deepflavour btag > 0.3033 (medium) |
25 | 21 - Jet 1: online b jet match |
26 | 22 - Jet 2: deepflavour btag > 0.3033 (medium) |
27 | 23 - Jet 2: online b jet match |
28 | 24 - Jet 3: deepflavour btag < 0.0521 (loose) [reverse btag] |
29 | 25 - *** Filling jets histograms - final_selection |
30 | 26 - Fill MssmHbb Histograms |
31 | 27 - Fill MssmHbb tree |
32 +----------------------------------------------------------------------------------------------+

where as for the Monte Carlo it is

1 +----------------------------------------------------------------------------------------------+
2 | Workflow Monte Carlo |
3 +----------------------------------------------------------------------------------------------+
4 | 0 - Total events read |
5 | 1 - Generated weighted events (sign of weights) |
6 | 2 - Using GenParticles collection: MssmHbb/Events/prunedGenParticles |
7 | 3 - Using GenJets collection: MssmHbb/Events/slimmedGenJets |
8 | 4 - Using Jet collection: MssmHbb/Events/updatedPatJets |
9 | 5 - Using Muon collection: MssmHbb/Events/slimmedMuons |
10 | 6 - Pileup weight (PileupWeight_Run2017_Mix_2017.root) |
11 | 7 - HLT_Mu12_DoublePFJets40MaxDeta1p6_DoubleCaloBTagCSV_p33_v |
12 | 8 - L1_Mu12er2p3_Jet40er2p3_dR_Max0p4_DoubleJet40er2p3_dEta_Max1p6 |
13 | 9 - MuonId: tight |
14 | 10 - JetId: tight |
15 | 11 - JetPileupId: loose |
16 | 12 - b jet energy regression |
17 | 13 - JER smearing (Fall17_V3_MC_PtResolution_AK4PFchs.txt,Fall17_V3_MC_SF_AK4PFchs.txt) |
18 | 14 - NMuons >= 1 |
19 | 15 - Muons selected: pt > 13.0 GeV and |eta| < 2.2 |
20 | 16 - Muons selected: Online muon matching (L1 and L3) |
21 | 17 - NJets >= 3 |
22 | 18 - Jet 1: pt > 60.0 GeV and |eta| < 2.2 |
23 | 19 - Jet 1: online jet match |
24 | 20 - Jet 2: pt > 50.0 GeV and |eta| < 2.2 |
25 | 21 - Jet 2: online jet match |
26 | 22 - Jet 3: pt > 30.0 GeV and |eta| < 2.2 |
27 | 23 - MSSMHbb Semileptonic: Jet-muon association |
28 | 24 - Jet 1: deepflavour btag > 0.3033 (medium) |
29 | 25 - Jet 1: btag SF applied (deepflavour medium WP) |
30 | 26 - Jet 1: online b jet match |
31 | 27 - Jet 2: deepflavour btag > 0.3033 (medium) |
32 | 28 - Jet 2: btag SF applied (deepflavour medium WP) |
33 | 29 - Jet 2: online b jet match |
34 +----------------------------------------------------------------------------------------------+

Output

The macro produces a root file containing a directory with jet-related histograms and a TTree containing the dijet mass $m_{bb}$ of the two leading b-tagged jets and the event weights after the full selection.

MssmHbbAnalyser class

:warning: To do

Generated by   doxygen 1.8.11