Data formats in RecoBTag

Event content definition

CMSSW_3_6_x More  CMSSW_3_6_x Less  (basic structure valid from CMSSW_3_6_0 onward)

b-tag intermediate collections (in RECO and AOD)
btagSoftElectrons	reco::Electron (lxr cvs)	Electron candidates identified by the dedicated btagging SoftElectronProducer, starting from reco::Tracks
softElectronTagInfos	reco::SoftLeptonTagInfo (lxr cvs)	soft electron dedicated TagInfo, containing informations used to b-tag jets due to the presence of a soft electron in the jet
impactParameterTagInfos	reco::TrackIPTagInfo (lxr cvs)	contains information used for btagging about track properties such as impact parameters, decay len, probability to originate from th primary vertex. Uses ak5JetTracksAssociatorAtVertex collection as input.
secondaryVertexTagInfos	reco::SecondaryVertexTagInfo (lxr cvs)	contains the reconstructed displaced secondary vertices in a jet and associated information, uses impactParameterTagInfos as input
ghostTrackVertexTagInfos
b-tag algorithm result collections (in RECO and AOD)
softMuonTagInfos	reco::SoftLeptonTagInfo (lxr cvs)	soft muon dedicated TagInfo, containing informations used to b-tag jets due to the presence of a soft muon in the jet
softElectronBJetTags	reco::JetTag (lxr cvs)	results of b-tagging a jet using the SoftElectronTagInfo and the default soft electron tagger, which uses a neural network to combine most electron properties to improve rejection of non-b jets
softMuonBJetTags	reco::JetTag (lxr cvs)	results of b-tagging a jet using the SoftMuonTagInfo and the default soft muon tagger, which uses a neural network to combine most muon properties to improve rejection of non-b jets
jetProbabilityBJetTags	reco::JetTag (lxr cvs)	result of jetProbability algorithm (based on TrackIPTagInfo).
jetBProbabilityBJetTags	reco::JetTag (lxr cvs)	result of jetProbability algorithm in the "jetBProbability" variant.
trackCountingHighPurBJetTags	reco::JetTag (lxr cvs)	Result of track counting algorithm (requiring three tracks to have a significance above the discriminator). To be used for high purity selection (B eff < 50%, mistag rate < 1% )
trackCountingHighEffBJetTags	reco::JetTag (lxr cvs)	Result of track counting algorithm (requiring two tracks to have a significance above the discriminator). To be used for high efficiency selection (B eff > 50%, mistag rate > 1% )
simpleSecondaryVertexHighPurBJetTags	reco::JetTag (lxr cvs)	Uses the flight distance (i.e. distance between a reconstructed secondary vertex and the primary vertex in a jet) as b-tagging discriminator. Secondary vertex is reconstructed with three or more tracks.
simpleSecondaryVertexHighEffBJetTags	reco::JetTag (lxr cvs)	Uses the flight distance (i.e. distance between a reconstructed secondary vertex and the primary vertex in a jet) as b-tagging discriminator. Secondary vertex is reconstructed with two or more tracks. Can be configured to return the value or significance in 2d and 3d, optionally corrected for the boost at the SV - works up to a maximum secondary vertex finding efficiency of ~70% in b-jets
combinedSecondaryVertexBJetTags	reco::JetTag (lxr cvs)	Result of application of a likelihood estimator to the tagging variables for the three possible algorithm outcomes (tracks only, pseudo vertex from at least two tracks or successful secondary vertex fit), obtained from impactParameterTagInfos and secondaryVertexTagInfos
combinedSecondaryVertexMVABJetTags	reco::JetTag (lxr cvs)	uses the PhysicsTools/MVAComputer framework to compute a discriminator from the impactParameterTagInfos and secondaryVertexTagInfos with an uptodate calibration from the the CMS conditions database, using a neural network instead of a likelihood ratio in case an actual secondary vertex was reconstructed
ghostTrackBJetTags

CMSSW_3_0_x More  CMSSW_3_0_x Less  (basic structure valid from CMSSW_3_0_0 onward upto 3_5_x)

b-tag intermediate collections (in RECO and AOD)
btagSoftElectrons	reco::Electron (lxr cvs)	Electron candidates identified by the dedicated btagging SoftElectronProducer, starting from reco::Tracks
softElectronTagInfos	reco::SoftLeptonTagInfo (lxr cvs)	soft electron dedicated TagInfo, containing informations used to b-tag jets due to the presence of a soft electron in the jet
impactParameterTagInfos	reco::TrackIPTagInfo (lxr cvs)	contains information used for btagging about track properties such as impact parameters, decay len, probability to originate from th primary vertex. Uses ak5JetTracksAssociatorAtVertex collection as input.
secondaryVertexTagInfos	reco::SecondaryVertexTagInfo (lxr cvs)	contains the reconstructed displaced secondary vertices in a jet and associated information, uses impactParameterTagInfos as input
b-tag algorithm result collections (in RECO and AOD)
softMuonTagInfos	reco::SoftLeptonTagInfo (lxr cvs)	soft muon dedicated TagInfo, containing informations used to b-tag jets due to the presence of a soft muon in the jet
softElectronBJetTags	reco::JetTag (lxr cvs)	results of b-tagging a jet using the SoftElectronTagInfo and the default soft electron tagger, which uses a neural network to combine most electron properties to improve rejection of non-b jets
softMuonBJetTags	reco::JetTag (lxr cvs)	results of b-tagging a jet using the SoftMuonTagInfo and the default soft muon tagger, which uses a neural network to combine most muon properties to improve rejection of non-b jets
jetProbabilityBJetTags	reco::JetTag (lxr cvs)	result of jetProbability algorithm (based on TrackIPTagInfo).
jetBProbabilityBJetTags	reco::JetTag (lxr cvs)	result of jetProbability algorithm in the "jetBProbability" variant.
trackCountingHighPurBJetTags	reco::JetTag (lxr cvs)	Result of track counting algorithm (requiring three tracks to have a significance above the discriminator). To be used for high purity selection (B eff < 50%, mistag rate < 1% )
trackCountingHighEffBJetTags	reco::JetTag (lxr cvs)	Result of track counting algorithm (requiring two tracks to have a significance above the discriminator). To be used for high efficiency selection (B eff > 50%, mistag rate > 1% )
simpleSecondaryVertexBJetTags	reco::JetTag (lxr cvs)	Uses the flight distance (i.e. distance between a reconstructed secondary vertex and the primary vertex in a jet) as b-tagging discriminator. Can be configured to return the value or significance in 2d and 3d, optionally corrected for the boost at the SV - works up to a maximum secondary vertex finding efficiency of ~70% in b-jets
combinedSecondaryVertexBJetTags	reco::JetTag (lxr cvs)	Result of application of a likelihood estimator to the tagging variables for the three possible algorithm outcomes (tracks only, pseudo vertex from at least two tracks or successful secondary vertex fit), obtained from impactParameterTagInfos and secondaryVertexTagInfos
combinedSecondaryVertexMVABJetTags	reco::JetTag (lxr cvs)	uses the PhysicsTools/MVAComputer framework to compute a discriminator from the impactParameterTagInfos and secondaryVertexTagInfos with an uptodate calibration from the the CMS conditions database, using a neural network instead of a likelihood ratio in case an actual secondary vertex was reconstructed

CMSSW_2_1_x More  CMSSW_2_1_x Less  (basic structure valid from CMSSW_1_7_0 onward)

b-tag intermediate collections (in RECO and AOD)
btagSoftElectrons	reco::Electron (lxr cvs)	Electron candidates identified by the dedicated btagging SoftElectronProducer, starting from reco::Tracks
softElectronTagInfos	reco::SoftLeptonTagInfo (lxr cvs)	soft electron dedicated TagInfo, containing informations used to b-tag jets due to the presence of a soft electron in the jet
impactParameterTagInfos	reco::TrackIPTagInfo (lxr cvs)	contains information used for btagging about track properties such as impact parameters, decay len, probability to originate from th primary vertex. Uses ic5JetTracksAssociatorAtVertex collection as input.
secondaryVertexTagInfos	reco::SecondaryVertexTagInfo (lxr cvs)	contains the reconstructed displaced secondary vertices in a jet and associated information, uses impactParameterTagInfos as input
b-tag algorithm result collections (in RECO and AOD)
softMuonTagInfos	reco::SoftLeptonTagInfo (lxr cvs)	soft muon dedicated TagInfo, containing informations used to b-tag jets due to the presence of a soft muon in the jet
softElectronBJetTags	reco::JetTag (lxr cvs)	results of b-tagging a jet using the SoftElectronTagInfo and the default soft electron tagger, which uses a neural network to combine most electron properties to improve rejection of non-b jets
softMuonBJetTags	reco::JetTag (lxr cvs)	results of b-tagging a jet using the SoftMuonTagInfo and the default soft muon tagger, which uses a neural network to combine most muon properties to improve rejection of non-b jets
softMuonNoIPBJetTags	reco::JetTag (lxr cvs)	results of b-tagging a jet using the SoftMuonTagInfo and the default soft muon tagger, which uses a neural network to combine most muon properties except the impact parameter to improve rejection of non-b jets
jetProbabilityBJetTags	reco::JetTag (lxr cvs)	result of jetProbability algorithm (based on TrackIPTagInfo).
jetBProbabilityBJetTags	reco::JetTag (lxr cvs)	result of jetProbability algorithm in the "jetBProbability" variant.
trackCountingHighPurBJetTags	reco::JetTag (lxr cvs)	Result of track counting algorithm (requiring three tracks to have a significance above the discriminator). To be used for high purity selection (B eff < 50%, mistag rate < 1% )
trackCountingHighEffBJetTags	reco::JetTag (lxr cvs)	Result of track counting algorithm (requiring two tracks to have a significance above the discriminator). To be used for high efficiency selection (B eff > 50%, mistag rate > 1% )
impactParameterMVABJetTags	reco::JetTag (lxr cvs)	uses the PhysicsTools/MVAComputer framework to compute a discriminator from the impactParameterTagInfos with an up-to-date calibration from the the CMS conditions database, making use of more general multivariate analysis techniques, like neural networks. Has been discontinued, do not use
simpleSecondaryVertexBJetTags	reco::JetTag (lxr cvs)	Uses the flight distance (i.e. distance between a reconstructed secondary vertex and the primary vertex in a jet) as b-tagging discriminator. Can be configured to return the value or significance in 2d and 3d, optionally corrected for the boost at the SV - works up to a maximum secondary vertex finding efficiency of ~70% in b-jets
combinedSecondaryVertexBJetTags	reco::JetTag (lxr cvs)	Result of application of a likelihood estimator to the tagging variables for the three possible algorithm outcomes (tracks only, pseudo vertex from at least two tracks or successful secondary vertex fit), obtained from impactParameterTagInfos and secondaryVertexTagInfos
combinedSecondaryVertexMVABJetTags	reco::JetTag (lxr cvs)	uses the PhysicsTools/MVAComputer framework to compute a discriminator from the impactParameterTagInfos and secondaryVertexTagInfos with an uptodate calibration from the the CMS conditions database, using a neural network instead of a likelihood ratio in case an actual secondary vertex was reconstructed

CMSSW_1_6_x More  CMSSW_1_6_x Less 

b-tag collections (in RECO and AOD)
btagSoftElectrons	reco::Electron (lxr cvs)	Electron candidates identified by the dedicated btagging SoftElectronProducer, starting from reco::Tracks
softElectronTagInfos	reco::SoftLeptonTagInfo (lxr cvs)	soft electron dedicated TagInfo, containing informations used to b-tag jets due to the presence of a soft electron in the jet
softMuonTagInfos	reco::SoftLeptonTagInfo (lxr cvs)	soft muon dedicated TagInfo, containing informations used to b-tag jets due to the presence of a soft muon in the jet
softElectronJetTags	reco::JetTag (lxr cvs)	results of b-tagging a jet using the SoftElectronTagInfo and the default soft electron tagger, which uses a neural network to combine most electron properties to improve rejection of non-b jets
softMuonJetTags	reco::JetTag (lxr cvs)	results of b-tagging a jet using the SoftMuonTagInfo and the default soft muon tagger, which uses a neural network to combine most muon properties to improve rejection of non-b jets
softMuonNoIPJetTags	reco::JetTag (lxr cvs)	results of b-tagging a jet using the SoftMuonTagInfo and the default soft muon tagger, which uses a neural network to combine most muon properties except the impact parameter to improve rejection of non-b jets
impactParameterTagInfos	reco::TrackIPTagInfo	contains information used for btagging about track properties such as impact parameters, decay len, probability to originate from th primary vertex.
jetProbabilityJetTags	reco::JetTag (lxr cvs)	result of jetProbability algorithm(based on TrackIPTagInfo).
trackCountingHighPurJetTags	reco::JetTag (lxr cvs)	Result of track counting algorithm (requiring three tracks to have a significance above the discriminator). To be used for high purity selection (B eff < 50%, mistag rate < 1% )
trackCountingHighEffJetTags	reco::JetTag (lxr cvs)	Result of track counting algorithm (requiring two tracks to have a significance above the discriminator). To be used for high efficiency selection (B eff > 50%, mistag rate > 1% )
impactParameterMVAJetTags	reco::JetTag (lxr cvs)	uses the PhysicsTools/MVAComputer framework to compute a discriminator from the TrackIPTagInfos with an up-to-date calibration from the the CMS conditions database, making use of more general multivariate analysis techniques, like neural networks
combinedSVTagInfos	reco::CombinedSVTagInfo (lxr cvs)	contains a list of tagging variables containing information about track impact parameters and results from an attempted secondary vertex fit
combinedSVJetTags	reco::JetTag (lxr cvs)	Result of application of a likelihood estimator to the tagging variables for the three possible algorithm outcomes (tracks only, pseudo vertex from at least two tracks or successful secondary vertex fit)
combinedSVMVAJetTags	reco::JetTag (lxr cvs)	uses the PhysicsTools/MVAComputer framework to compute a discriminator from the CombinedSVTagInfos with an uptodate calibration from the the CMS conditions database, making use of more general multivariate analysis techniques, like neural networks

Review status

Responsible: Andrea Rizzi, Thomas Speer
Last reviewed by: JyothsnaK - 26-May-2010