Algorithms

Each algorithm will retrieve a container from either TEvent or TStore and if an output container is specified, it is saved to the TStore object such that the following algorithms can access the container. TStore will manage the memory for the user.

Event Selection

(moved to BasicEventSelection.h)

Jet Related

HLTJetGetter

The HLTJetGetter Algorithm retrieves jets from the TrigDecisionTool. The output is a shallow copy of the trigger feature that is requested via the combination of the name of the jet collection (m_inContainerName) and the jet trigger that is selected m_triggerList. The name of the copy is chosen via the m_outContainerName. The input container name should be given without any HLT_xAOD__JetContainer prefix if the collection comes directly from the xAOD. The list of triggers must be given as a regular expression using an or (|) if more than a trigger is requested. To request all triggers, use *.

JetCalibrator

The JetCalibrator Algorithm handles the calibration, cleaning, and common uncertainties associated with Jets. It interfaces with many JetETMiss packages. The output is a shallow copy of the input jet container (i.e. a new xAOD::JetCollection). A separate shallow copy is made for each systematic variation requested, and the names of the containers are saved in a vector for downstream algorithms to use. The algorithm does not apply any selection to the jets.

The parameters to control the Algorithm are described in the header documentation: https://xaodanahelpers.readthedocs.io/en/latest/JetCalibrator.html

Calibration

Jets are calibrated using JetCalibTools. The option CalibSequence determines the calibration sequence to be applied. “Insitu” in automatically added to data samples if it is not provided, and MC jobs will fail if it is included. Config file options can be provided with CalibConfigData, CalibConfigAFII, and CalibConfigFullSim. The config variable to use is determined from metadata.

1. Sort - Sort the output jets by pT

2. SetAFII - Force AFII configurations (in case metadata isn’t working)

3. InputContainer - Jet Collection to use

4. OutputContainer - Output Jet Collection for calibrated jets

5. JetAlgorithm - Input Jet Algorithm

6. OutputAlgo - Output Jet Algorithm

JES/JER Uncertainties

Outputs new containers for each systematic variation.

For JES: Uses JetUncertaintiesTool. JES is different for AFII, so m_setAFII may be required if metadata isn’t working. 1. JESUncertConfig - Configuration file for JES 2. JESUncertMCType - Type of JES Uncertainty to use, MC20 or MC21

For JER: Uses JERSmearingTool 1. JERUncertConfig - Configuration file for JER 2. JERFullSys - Run full list of systematics for data and MC. Not currently recommended by JER Group (10/08/2015). 3. JERApplyNominal - Apply a nominal JER Smearing.

Cleaning

Cleaning decisions are saved as auxiliary data; No selections are made to the container. The decision is saved as an auxdecor with names like clean_passLooseBadUgly clean_pass Settings: 1. JetCleanCutLevel - Cut Level, may be LooseBad or TightBad 2. JetCleanUgly - Clean Ugly Jets (in gap region) 3. SaveAllCleanDecisions - Save all four decisions (LooseBad, LooseBadUgly, TightBad, TightBadUgly) 4. CleanParent - Apply cleaning decision to the jet’s parent

JetSelector

Applies cuts on the input jet collection(s). The jets can be decorated with the decision of the cuts. A new xAOD::JetCollection can be made from the ones passing the cuts.

The JVT selection is OFF by default, it can be turned ON with `m_doJVT`. The default algorithm is NNJvt. If another algorithm is needed, use corresponding index for the enum here: https://acode-browser1.usatlas.bnl.gov/lxr/source/athena/PhysicsAnalysis/Interfaces/JetAnalysisInterfaces/JetAnalysisInterfaces/IJetJvtEfficiency.h#0022 (note: this link points to the latest r22 version, i.e. master, if a release is used, please check the corresponding enum for the given release: https://gitlab.cern.ch/atlas/athena/-/tags?search=release%2F22.2&sort=updated_desc) The default JVT working point is `FixedEffPt`, it can be changed with `m_WorkingPointJVT`. JVT is no longer recalculated in JetCalibrator, it is now done in JetSelector. It can be disabled by setting `m_recalculateJvtScores` to `false`. Furthermore, the JVT algorithm is now NNJvt (default for R22).

JetHists and JetHistsAlgo

JetHists is a class of histograms for jet variables. In the constructor one gives a name which is used as the first half of the name of all the histograms in the class and the detail string which determines which plots are made. One can declare a JetHists object in the event selection algorithm and fill the hists with every jet in a collection or pick jets by hand to fill histograms.This is done in the example event selection described below. JetHistsAlgo is a algorithm that holds a JetHists object and fills it with every jet in the user specified container.

BJetEfficiencyCorrector

This Algo wraps two flavor tagging tools - one returns the decision on if the jet is tagged or not for the given operating point (OP) and the other returns the efficiency scale factors needed for MC only. Like other Algos, “All” gives all the systematic variations. Two decorations are added to the input jets: 1. The B-Tag decision. Decoration name (default) “BTag_:math:{OP}” 2. The B-Tag efficiency scale factor. If the jet fails the cut the inefficiency scale factor is saved. Decoration name (default) “BTAG_SF_{OP}”. This decoration is not added if the OP has not been calibrated.

The available calibrated operating points in the August 2015 CDI file for fixed b-jet efficiencies: * FixedCutBEff_60 * FixedCutBEff_70 * FixedCutBEff_77 * FixedCutBEff_85

and the fixed cut un-calibrated operating points are: * FixedCutBEff_30 * FixedCutBEff_50 * FixedCutBEff_80 * FixedCutBEff_90

Finally, the flat efficiency un-calibrated operating points are: * FlatBEff_30 * FlatBEff_40 * FlatBEff_50 * FlatBEff_60 * FlatBEff_70 * FlatBEff_77 * FlatBEff_85

These decorations can be added to a TTree via HelpTreeBase. The info switch looks for “sfFTagFix” and “sfFTagFlt” for the fixed and flat efficiency OPs respectively. “sfFTagFix607785” will try to add the decision and efficiency/inefficiency scale factors for the fixed 60, 70, and 85% b-jet efficiency OPs to the tree. The user must also create and run BJetEfficiencyCorrector for each OP.

Note: To those paying attention. Currently the two flavor tagging tools use different names for the same operating point. This is handled internally in xAH until the flavor tagging group harmonizes the two.

Muon Related

MuonCalibrator

MuonEfficiencyCorrector

Produces a container of muons decorated with efficiencies and scale factors. This container is not simply the one in input, but is a deepCopy of that. This allows flexibility in decorating muons of systematically altered containers. The output container has a configurable name and is only created for MC events. When working with systematic uncertainties, a list of systematics is passed to this algorithm m_inputAlgoSystNames. This algorithm supports comma separated lists as inputs, which will be considered as a unique list. The systematic names in this list will be looked for to retrieve the muon containers to decorate. Systematic variations on the decorations themselves might be unnecessary for all systematic muon containers and might only be considered for the nominal container (need of deepCopy). This is the default configuration. Otherwise the option m_decorateWithNomOnInputSyscan be set to false. The algorithm features the option m_sysNamesForParContwhich is a list of systematic names. For each of them, a copy of the nominal muon container is put in the store carrying the name of the systematic. These containers are only decorated with the nominal efficiencies and scale factors. The use case of this are MET systematics for which one does not want systematic variations on efficiencies, but still wants to retrieve a nominal muon container in the tree algo. Retrieving the nominal would indeed carry all the unnecessary uncertainties.

MuonSelector

MuonHists and MuonHistsAlgo

Electron Related

ElectronCalibrator

ElectronEfficiencyCorrector

See the MuonEfficiencyCorrector

ElectronSelector

ElectronHists and ElectronHistsAlgo

Tau Related

TauCalibrator

TauSelector

Similar to other selectors. Tau identification/selection algorithms working points are decorated onto the tau automatically. Tha flag m_decorateWithTracks controls the decoration of the tau track information. It can be dumped in the ntuples by specifying the trackparams flag in the tau details string of the TreeAlgo.

TauEfficiencyCorrector

The TauEfficientyCorrector provides one cumulative decoration with one SF corresponding to the combination of working points used for the tau selection/identification algorithms. Several initialisations of the algorithm are needed with different combinations in order to dump in the ntuples different combined working points. Tau trigger SFs are saved separately and wrt said cumulative working point which has to be specified in the initialisation of a new instance of the algorithm together with the trigger menu.

TauJetMatching

This algorithm is introduced to match an arbitrary and configurable collection of jets with the TauJet object. This is useful for cases where the tau seedJet (LC topo jet) is not available in xAOD but one would need to get a handle on some original jet info. The algorithm has a configurable DeltaR matching criterion and for now decorates taus with the matched jet width. It should be executed before OLR. It can be used before tau selection and after tau calibration.

HelperFunctions

HelperClasses

NTuple Creation

HelpTreeBase is a class (not an algorithm) that creates and fills a TTree. When adding an object type i.e. jet, muon, or event level quantities, it can be easily configured with “info switches” ( See HelperClasses ) that take a space separated list of categories.