Check all EDAnalyzers in the 2015MiniAOD branch

[caredg]

In the spirit of #56, check the EDAnalyzers that were initially implemented in the 2015MiniAOD branch to verify that:

• they are Aligning with the documentation found in the WorkBookMiniAOD2015; in partular, with the implementation suggested by these explicit examples
• they run on real collisions data and MC samples
• the output gives leafs that are filled properly and gives a reasonable output

The EDAnalyzers that need to be checked are:

• [ ] ElectronAnalyzer
• [ ] MuonAnalyzer
• [ ] GenParticleAnalyzer
• [ ] JetAnalyzer
• [ ] MetAnalyzer
• [ ] MuonAnalyzer.cc
• [ ] PhotonAnalyzer
• [ ] TauAnalyzer
• [ ] TriggObjectAnalyzer
• [ ] VertexAnalyzer

It is very important that, as one checks this information, the corresponding documentation is input in the CMS Open Data Guide, following the work already started, and almost finished, for Run 1 objects.

Also, these two analyzers should be created if possible:

• [ ] EventAnalyzer
• [ ] TriggerAnalyzer
• [ ] TrackAnalyzer

The master branch of the TriggerInfoTool could be used to create the TriggerAnalyzer.

NOTE: Please create a separate issue for each EDAnalyzer test and, if possible, assign it to yourself

[katilp]

Recording the very useful list from @jmhogan (in the WS mattermost) here:

• Jets in 2015 data:

  • Jet ID for 2015: https://twiki.cern.ch/twiki/bin/view/CMS/JetID13TeVRun2015
  • JEC page for GTs and tar files: https://twiki.cern.ch/twiki/bin/viewauth/CMS/JECDataMC#2015_Data (MC recs are in the table above, GT = 76X_mcRun2_asymptotic_RunIIFall15DR76_v1)
  • JER info: https://twiki.cern.ch/twiki/bin/view/CMS/JetResolution#2015_data

  Internal for our use: test code to make sure JEC/JER are done right https://twiki.cern.ch/twiki/bin/view/CMS/JERCReference#CMSSW_7_6_X_updated_09_05_2016

  • Btagging:

  • https://twiki.cern.ch/twiki/bin/viewauth/CMS/BtagRecommendation76X

  • Wtagging:

  • https://twiki.cern.ch/twiki/bin/view/CMS/JetWtagging#2016_scale_factors_and_correctio (go to the bottom of this section and "show" the 76X scale factors. They are documented in an AN)

  • Top tagging:

  • https://twiki.cern.ch/twiki/bin/view/CMS/JetTopTagging#Previous_working_points

• Electron ID in 2015 data:

  • MVA-based (I actually have code for implementing the MVA prediction outside of VID): https://twiki.cern.ch/twiki/bin/view/CMS/MultivariateElectronIdentificationRun2Archive#MVA_recipes_for_2015_data_50ns_2

  • Cut-based: https://twiki.cern.ch/twiki/bin/view/CMS/CutBasedElectronIdentificationRun2Archive#Working_points_for_2015_data_50n

  • Efficiencies/SFs: https://twiki.cern.ch/twiki/bin/view/CMS/EgammaIDRecipesRun2#Older_Efficiencies_and_Scale_Fac (scroll down to 76X series)

• Muon ID in 2015 data: (these are in general not really changes w.r.t. 2012 in terms of code)

  • Cut-based: https://twiki.cern.ch/twiki/bin/viewauth/CMS/SWGuideMuonIdRun2#Muon_Identification

  • Isolation: https://twiki.cern.ch/twiki/bin/viewauth/CMS/SWGuideMuonIdRun2#Muon_Isolation

  • Efficiencies/SFs: https://twiki.cern.ch/twiki/bin/viewauth/CMS/MuonReferenceEffsRun2#Results_for_2015_data

• Photon ID in 2015 data:

  • MVA-based: https://twiki.cern.ch/twiki/bin/view/CMS/MultivariatePhotonIdentificationRun2Archive#MVA_recipes_for_2015_data_50ns_2

  • Cut-based: https://twiki.cern.ch/twiki/bin/view/CMS/CutBasedPhotonIdentificationRun2Archive#Working_points_for_2015_data_50n

  • Efficiencies/SFs: https://twiki.cern.ch/twiki/bin/view/CMS/EgammaIDRecipesRun2#Older_Efficiencies_and_Scale_AN1 (scroll down to 76X series)

• MET filters in 2015 data:

  • https://twiki.cern.ch/twiki/bin/viewauth/CMS/MissingETOptionalFiltersRun2#76X_Recommendations_and_Recipes

• Taus

  • we will have to ask them.

[caredg]

From Julie: Some references for jet substructure variable storage in early MiniAOD: https://github.com/jmhogan/Ljmet-Com/blob/CMSSW_7_6_X/src/JetSubCalc.cc

thePrunedMass = corrunc(double)l2l3jet.userFloat("ak8PFJetsCHSPrunedMass"); theSoftDropMass = corrunc(double)l2l3jet.userFloat("ak8PFJetsCHSSoftDropMass"); theNjettinessTau1 = (double)corrak8.userFloat("NjettinessAK8:tau1"); theNjettinessTau2 = (double)corrak8.userFloat("NjettinessAK8:tau2");

"Early" MiniAOD certainly means 2015 and I think also 2016 before they switched the default AK8 jet collection to Puppi jets.

[jmhogan]

To be a little more general: variable = thisPatJetObject.userFloat("specificName"); Where the "specificName" strings are in the comment above.

JEC/JER don't get propagated to userFloat content, so the code link shows our method to calculate and apply the corrections onto those mass variables.

[nsmith-]

A possible direction we can go in is to port some of these corrections to correctionlib. An introductory tutorial is at https://cms-nanoaod.github.io/correctionlib/correctionlib_tutorial.html