Skip to content

Update Muon MVA ID model and working points in MiniAOD and NanoAOD #41903

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
cmsbuild merged 5 commits into from
Jul 11, 2023
Merged

Update Muon MVA ID model and working points in MiniAOD and NanoAOD #41903

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
709d8b4
updating Muon MVA ID working points for the new training
Jun 8, 2023
01ab854
Updating working points of the Muon MVA ID for the new model in NanoAOD
Jun 8, 2023
773a5fa
defining wps in config file
Jul 5, 2023
5180509
fixing format
Jul 6, 2023
4443ff6
code formats
Jul 6, 2023
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion PhysicsTools/NanoAOD/python/muons_cff.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -40,7 +40,7 @@
inputTensorName= cms.string("float_input"),
outputNames = cms.vstring(["probGOOD", "wpMedium", "wpTight"]),
batch_eval =cms.bool(True),
outputFormulas = cms.vstring(["at(1)", "? at(1) > 0.14 ? 1 : 0", "? at(1) > 0.35 ? 1 : 0"]),
outputFormulas = cms.vstring(["at(1)", "? at(1) > 0.08 ? 1 : 0", "? at(1) > 0.20 ? 1 : 0"]),
Copy link
Contributor

@simonepigazzini simonepigazzini Jun 23, 2023

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Should we preserve the old formula for Run2 reprocessing by adding it back through the run2_nanoAOD_ANY modifier?

Copy link
Contributor Author

@andrea21z andrea21z Jun 23, 2023

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

No, the new formula should be used also for Run 2 reprocessing

Copy link
Contributor

@simonepigazzini simonepigazzini Jun 23, 2023

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

ok cool

variablesOrder = cms.vstring(["LepGood_global_muon","LepGood_validFraction","Muon_norm_chi2_extended","LepGood_local_chi2","LepGood_kink","LepGood_segmentComp","Muon_n_Valid_hits_extended","LepGood_n_MatchedStations","LepGood_Valid_pixel","LepGood_tracker_layers","LepGood_pt","LepGood_eta"]),
variables = cms.PSet(
LepGood_global_muon = cms.string("isGlobalMuon"),
Expand Down
16 changes: 9 additions & 7 deletions PhysicsTools/PatAlgos/plugins/PATMuonProducer.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -156,7 +156,9 @@ namespace pat {
std::vector<double> effectiveAreaVec_;
std::vector<double> miniIsoParams_;
double relMiniIsoPUCorrected_;

/// working points of the muon MVA ID
double mvaIDtightCut_;
double mvaIDmediumCut_;
/// embed the track from best muon measurement (global pflow)
bool embedBestTrack_;
/// embed the track from best muon measurement (muon only)
Expand Down Expand Up @@ -347,6 +349,10 @@ PATMuonProducer::PATMuonProducer(const edm::ParameterSet& iConfig, PATMuonHeavyO
geometryToken_{esConsumes()},
transientTrackBuilderToken_{esConsumes(edm::ESInputTag("", "TransientTrackBuilder"))},
patMuonPutToken_{produces<std::vector<Muon>>()} {
// Muon MVA ID wps
mvaIDmediumCut_ = iConfig.getParameter<double>("mvaIDwpMedium");
mvaIDtightCut_ = iConfig.getParameter<double>("mvaIDwpTight");

// input source
muonToken_ = consumes<edm::View<reco::Muon>>(iConfig.getParameter<edm::InputTag>("muonSource"));
// embedding of tracks
Expand Down Expand Up @@ -990,19 +996,15 @@ void PATMuonProducer::produce(edm::Event& iEvent, const edm::EventSetup& iSetup)
// MVA ID
float mvaID = 0.0;
constexpr int MVAsentinelValue = -99;
constexpr float mvaIDmediumCut = 0.08;
constexpr float mvaIDtightCut = 0.12;
if (computeMuonIDMVA_) {
const double dz = std::abs(muon.muonBestTrack()->dz(primaryVertex.position()));
const double dxy = std::abs(muon.muonBestTrack()->dxy(primaryVertex.position()));
if (muon.isLooseMuon()) {
mvaID = globalCache()->muonMvaIDEstimator().computeMVAID(muon)[1];
} else {
mvaID = MVAsentinelValue;
}
muon.setMvaIDValue(mvaID);
muon.setSelector(reco::Muon::MvaIDwpMedium, muon.mvaIDValue() > mvaIDmediumCut);
muon.setSelector(reco::Muon::MvaIDwpTight, muon.mvaIDValue() > mvaIDtightCut and dz < 0.5 and dxy < 0.2);
muon.setSelector(reco::Muon::MvaIDwpMedium, muon.mvaIDValue() > mvaIDmediumCut_);
muon.setSelector(reco::Muon::MvaIDwpTight, muon.mvaIDValue() > mvaIDtightCut_);
}

//SOFT MVA
Expand Down
2 changes: 2 additions & 0 deletions PhysicsTools/PatAlgos/python/producersLayer1/muonProducer_cfi.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -111,6 +111,8 @@
# Don't forget to set flags properly in miniAOD_tools.py
computeMuonIDMVA = cms.bool(False),
mvaIDTrainingFile = cms.FileInPath("RecoMuon/MuonIdentification/data/mvaID.onnx"),
mvaIDwpMedium = cms.double(0.08),
mvaIDwpTight = cms.double(0.20),
recomputeBasicSelectors = cms.bool(True),
useJec = cms.bool(True),
mvaDrMax = cms.double(0.4),
Expand Down