Tutorial AOD 100305

60 %
40 %
Information about Tutorial AOD 100305

Published on October 29, 2007

Author: Nikita

Source: authorstream.com

AOD Tutorial:  AOD Tutorial These slides can be found in ppt in http://hepunx.rl.ac.uk/atlasuk/simulation/level2/egamma/tutorial/Tutorial_AOD_100305.ppt In pdf http://hepunx.rl.ac.uk/atlasuk/simulation/level2/egamma/tutorial/Tutorial_AOD_100305.pdf AOD Tutorial:  AOD Tutorial Outline Transferring files via DonQuixote Exercise 1: Copy a file via DonQuixote Overview of ESD/AOD production Interactive AOD session Exercise 2: Basic set-up Exercise 3: Look at AOD interactively Exercise 4: Basic set-up retrieve electron container and plot some quantities Useful tools four-momentum, particle, navigable Exercise 5: plot invariant Z-mass Some more useful tools Exercise 6: Use some tools to plot Z-mass Transferring files via DonQuixote:  Transferring files via DonQuixote Most of ESD/AOD prod for Rome will be centrally produced via GRID To find out about the Status of Rome productions and available samples https://uimon.cern.ch/twiki/bin/view/Atlas/RomeStatusWiki http://atlfarm003.mi.infn.it/%7Enegri/rome_dataset.htm Files can be transferred via DonQuixote https://uimon.cern.ch/twiki/bin/view/Atlas/RomeGetFilesWiki What is it ATLAS Experiment Data Management System integrate all Grid data management services used by ATLAS provide production managers and physicists access to file-resident event data implementing data flow as defined by the ATLAS Computing Model In short: mean to access “official” ATLAS data produced on the GRID Documentation how to use it in https://uimon.cern.ch/twiki/bin/view/Atlas/DonQuijoteEndUserClient To transfer files you need a GRID certificate and be registered with ATLAS VO You can install DonQuixote in your lab on a machine with GRID UI Basic GRID set-up:  Basic GRID set-up At CERN lxplus can act as a GRID machine To be able to run there we have to export the keys and get them on lxplus This was the exercise to do prior to the tutorial. If you havn’t done so, follow the instructions and watch over someone else’s shoulder http://www.grid-support.ac.uk/ca/user-documentation/uk-esc-documentation.pdf http://msmizans.home.cern.ch/msmizans/production/dc2/grid0.html To enable your lxplus session to act as UI source /afs/cern.ch/project/gd/LCG-share/2.3.0/sl3/etc/ profile.d/grid_env.sh source /afs/cern.ch/project/gd/LCG-share/2.3.0/rh73/etc/ profile.d/grid_env.sh Obtain temporary grid proxy certificate grid-proxy-init -valid Hours:Minutes (if you don’t give the time, you’ll get 8h) At the end of your session grid-proxy-destroy Export your key to CERN:  Export your key to CERN Follow indications “how to export your key” in http://www.grid-support.ac.uk/ca/user-documentation/uk-esc-documentation.pdf Then let’s install it to CERN (instructions in web page out-dated, so let’s do it together) http://msmizans.home.cern.ch/msmizans/production/dc2/grid0.html Log on lxplus7 and in your home directory, do the following, assuming the key is called: MyKeys.pfx source /afs/cern.ch/project/gd/LCG-share/2.3.0/rh73/etc/ profile.d/grid_env.sh openssl pkcs12 -in MyKeys.pfx -clcerts -nokeys -out usercert.pem openssl pkcs12 -in MyKeys.pfx -nocerts -out userkey.pem chmod 400 userkey.pem chmod 444 usercert.pem mkdir .globus/ mv *.pem .globus/ How to use DonQuixote at CERN:  How to use DonQuixote at CERN Set up lxplus7 to act as GRID UI and identify yourself to the grid Note: DonQuixote version 2 only works under RH7, version 3 about to come source /afs/cern.ch/project/gd/LCG-share/2.3.0/rh73/etc/ profile.d/grid_env.sh grid-proxy-init The python script you need to use is installed at CERN /afs/cern.ch/atlas/offline/external/DQClient/dms2/dms2.py Basic commands you are likely to use are: Search for a dataset dms2.py search ‘rome.003025.digit.B4_jets17filter*’ dms2.py search ‘rome.003025.digit.B4_jets17filter*’ > dijet.out (more convenient) Copy data to castor (for LCG files) dms2.py get ‘rome.003025.digit.B4_jets17filter._02010.pool.root’ -–remote dms2.py get ‘rome.003025.digit.B4_jets17filter*’ --copy –-remote -–multiple For files from grid3 or NorduGrid (ng) dms2.py –g grid3 or dms2.py –g ng Exercise 1: copy a file via Grid:  Exercise 1: copy a file via Grid Install your GRID key at CERN Execute the grid-init Ok, let’s try to see which files for dataset “rome.003025.digit.B4_jets17filter*’” was produced Now copy one of the files to CERN Note: in the CERN setup the datafile will automatically end up in the castor grid directory /castor/cern.ch/grid/atlas/datafiles/<prod>/<type> <prod> = dc1, dc2, rome <type> = evgen, simul, digit, reco Note: at the end of the copy there will be some error messages, however, the copy probably worked Check you see it in the castor directory Don Quixote is a bit buggy, so sometimes you have to try again, some datasets such as digitcalib doesn’t seem to be accepted by the tool. Very likely we don’t have to worry for the large scale AOD Rome productions The files will be available at CERN from a castor scratch area Brief Overview of ESD/AOD production:  Brief Overview of ESD/AOD production ESD : Event Summary Data Detailed output of reconstruction Contains main reconstruction objects, such a EM clusters, track particles, muons, taus, jets… Gives as well list of cells belonging to a cluster, space-points belonging to a track…. Thus refinements of combined reco. e.g. particle id, track refitting, jet calibration possible Things such as re-doing pattern recognition, re-calibration of cells not possible AOD: Analysis Object Data Only necessary information to perform most analysis available data organised into analysis objects whose classes have physical meaning Electron, photon, jet, tau, muon, … Also TrackParticle essential aspect is truth association, which directly links the original Monte Carlo data to the reconstructed objects AOD is formed from the ESD through the use of AOD Builders Back-navigation to ESD information possible Slide12:  LVL1 calo variables, LVL2 calo + track variables How to produce an ESD:  How to produce an ESD Some pre-defined jobOptions available get_files optRecExtoESD.py OptRectoESD.py doWriteESD = True # ESD persistency flag DetDescrVersion = "Rome-Initial" # Detector description EvtMax = 5 #number of Event to process doCBNT = False # suppress ntuple doHist = False # suppress histos PoolRDOInput = ["rfio:/castor/cern.ch/xxxx/ MyAOD.pool.root" ] # the input raw data file PoolESDOutput = “ESD.pool.root” # output ESD name To run athena.py optRecExToESD.py ../share/RecExCommon_topOptions.py Similar you can run together with the Trigger_forRecExCommon_jobOptions.py file from the TriggerRelease area and produce an ESD with offline and trigger information How to produce an AOD:  How to produce an AOD Some pre-defined jobOptions available get_files optESDtoAOD.py athena optESDtoAOD.py ../share/RecExCommon_topOptions.py OptESDtoAOD.py AllAlgs = False # Do not re-run the reco. readESD = True # Read the ESD doWriteAOD = True # AOD persistency output DetDescrVersion = "Rome-Initial" # Detector description doHist = False # suppress histograms PoolESDInput = [ "ESD.pool.root" ] # ESD input file PoolAODOutput = "AOD.pool.root“ # AOD output file from AthenaCommon.DetFlags import DetFlags # Detector Flags DetFlags.detdescr.ID_setOn() DetFlags.detdescr.Calo_setOn() DetFlags.detdescr.Muon_setOn() # switch AOD streaming - we are not interested in that here from ParticleEventAthenaPool.AODFlags import AODFlags AODFlags.Streaming = False How to produce an AOD:  How to produce an AOD In addition something similar exists for ATLFAST I don’t think we should try to do this here, but have a look at http://www.usatlas.bnl.gov/PAT/tutorial904.html#Prod_ESD_AOD Ok, let’s try to look at an AOD and do basic set-up:  Ok, let’s try to look at an AOD and do basic set-up This tutorial is based on release 10 Though AOD’s are from 9.0.4 Not everything can be tested due to changes in persistified classes E.g. back navigation doesn’t work Let’s start with the basic set-up Ask for Release 10.0.0 in cmt requirements file source /afs/cern.ch/sw/contrib/CMT/v1r16p20040901/mgr/setup.sh cmt config source setup.sh –tag=opt create Tutorial area (where ever it’s convenient) cd to your tutorial area cmt co -r UserAnalysis-00-02-00 PhysicsAnalysis/AnalysisCommon/UserAnalysis cmt co Control/AthenaServices cd PhysicsAnalysis/AnalysisCommon/UserAnalysis/ UserAnalysis-00-02-00/src cp ~mwielers/maxidisk/tutorial_files/AnalysisSkeleton.cxx . cd ../UserAnalysis Exercise 2: Basic set-up for tutorial:  Exercise 2: Basic set-up for tutorial cp ~mwielers/maxidisk/tutorial_files/AnalysisSkeleton.h . cd ../cmt Add AthenaServices to your requirements Use AthenaServices AthenaServices-* Control cmt bro cmt config source setup.sh cmt bro gmake cd ../run cp ~mwielers/maxidisk/tutorial_files/PoolFileCatalog.xml . get_files PDGTABLE.MeV cp ~mwielers/maxidisk/tutorial_files/AnalysisSkeleton_jobOptions.py . cp ~mwielers/maxidisk/tutorial_files/zee.aod.py . Let’s look at an ESD and AOD by just opening the pool file and see what persistified object are in there Open ESD file with root ~mwielers/maxidisk/tutorial_files/ESD.pool.root Do you see some trigger stuff? Open AOD file with root ~mwielers/maxidisk/tutorial_files/AOD.pool.root Interactive AOD session:  Interactive AOD session Unfortunately it only works under RH73 and doesn’t run out of the box in 10 cmt co Control/AthenaServices (already done) Only works on AOD’s not directly on ESD’s (though possible via back-navigation) get_files -jo Interactive_topO.py pool_insertFileToCatalog /afs/cern.ch/user/m/mwielers/maxidisk/tutorial_files/AOD.pool.root edit the jobO (if needed) e.g., change input file EventSelector.InputCollections = [ "AOD.pool.root" ] To EventSelector.InputCollections = [ “/afs/cern.ch/user/m/mwielers/maxidisk/tutorial_files/AOD.pool.root" ] Or EventSelector.InputCollections = [ “rfio/castor/cern.ch/user/m/mwielers/aod/AOD.pool.root" ] EventSelector.InputCollections = ["AOD.pool_1.root“, “AOD.pool_2.root”] start an interactive session:  start an interactive session athena -i Interactive_topO.py -l FATAL Don’t forget -i option!! MCTruth::TruthStrategyManager: registered strategy DMuonCatchAll MCTruth::TruthStrategyManager: registered strategy IDETIonization MCTruth::TruthStrategyManager: registered strategy IDETDecay MCTruth::TruthStrategyManager: registered strategy IDETConversion MCTruth::TruthStrategyManager: registered strategy IDETBrems MCTruth::TruthStrategyManager: registered strategy CALOCatchAll ==> New TileCablingService created Loaded dictionary PyAnalysisCoreDict Loaded dictionary PyParticleToolsDict Loaded dictionary PyTriggerToolsDict Loaded dictionary PyKernelDict Initialize application manager athena> theApp.initialize() Run 1 event athena> theApp.nextEvent() Basic commands in interactive AOD session:  Basic commands in interactive AOD session Retrieve ElectronCollection athena> econ = PyParticleTools.getElectrons("ElectronCollection") where “ElectronCollection” is key name in the StoreGate. Other methods are defined in PyParticleTools e.g., for Muon, PyParticleTools.getMuons(”MuonCollection”) Number of electrons athena> len(econ) 5 Get the first electron athena> e = econ[0] See pT athena> e.pt() 33299.1563339692 Create and fill a ROOT histogram on the fly: plot("ElectronContainer#ElectronCollection","$x.pt()",nEvent=5) nEvent has to be less than the number of events in the file…  Only works once, then you’re at the end of the file and we don’t know how to re-start, unless you re-start your interactive session Get a list of methods:  Get a list of methods athena> dir(e) ['Electron', 'Navigable', 'P4EEtaPhiM', '_C_instance', '_C_metaclass', '__class__', '__del__', '__delattr__', '__dict__', '__doc__', '__getattribute__', '__hash__', '__init__', '__module__', '__new__', '__reduce__', '__reduce_ex__', '__repr__', '__setattr__', '__str__', '__weakref__', '_empty_instance', '_made_in_Python', 'author', 'begin', 'charge', 'contains', 'cosTh', 'cotTh', 'd0wrtPrimVtx', 'dataType', 'e', 'eg', 'end', 'et', 'eta', 'fillToken', 'getContainer', 'getEgammaWeight', 'getIndex', 'getParameter', 'hasCharge', 'hasPdgId', 'hasTrack', 'hlv', 'iPt', 'isEM', 'm', 'm_author', 'm_charge', 'm_constituents', 'm_dataType', 'm_e', 'm_eta', 'm_hasCharge', 'm_hasPdgId', 'm_hasTrack', 'm_isEM', 'm_m', 'm_origin', 'm_parameters', 'm_pdgId', 'm_phi', 'm_track', 'numberOfBLayerHits', 'numberOfPixelHits', 'numberOfSCTHits', 'numberOfTRTHighThresholdHits', 'numberOfTRTHits', 'origin', 'p', 'parameter', 'pdgId', 'phi', 'pt', 'putElement', 'px', 'py', 'pz', 'remove', 'removeAll', 'set4Mom', 'setE', 'setEta', 'setM', 'setPhi', 'set_all', 'set_charge', 'set_dataType', 'set_eg', 'set_isEM', 'set_origin', 'set_parameter', 'set_pdgId', 'set_track', 'sinTh', 'size', 'track', 'z0wrtPrimVtx', '~Electron', '~I4Momentum', '~INavigable', '~INavigable4Momentum', '~IParticle', '~Navigable', '~P4EEtaPhiM', '~P4EEtaPhiMBase', '~ParticleBase'] Unfortunately you don’t just see the members, but as well all method provided by the class See AOD contents using PyPoolBrowser:  See AOD contents using PyPoolBrowser athena> include ("PyAnalysisExamples/PyPoolBrowser.py")  double-click “/” click “+” of ElectronCollection → “+” on 0 → double click on “pt”, then you will see a histogram for pT distribution of electrons Doesn’t really work in 9.0.4 Be patient… very slow Unfortunately only works on AOD’s right now, not on ESD’s Use Ctrl-D to exit Exercise 2: Interactive AOD session:  Exercise 2: Interactive AOD session Do basic set-up Run interactive AOD session and try out some of the commands For more examples/possibilities, see http://tmaeno.home.cern.ch/tmaeno/Inter.htm Allows navigation via ElementLink going from electron  track  track quantities Allows navigation back to ESD e.g. electron  egammaObject in ESD How to access the Container:  How to access the Container Just as you do in the reconstruction Retrieve via StoreGate Example: std::string m_electronContainerName = "ElectronCollection"; ... const ElectronContainer* elecTES; sc=m_storeGate->retrieve( elecTES, m_electronContainerName ); if( sc.isFailure() || !elecTES ) { mLog << MSG::WARNING << "No AOD electron container" << endreq; return StatusCode::SUCCESS; } mLog << MSG::DEBUG << "ElectronContainer retrieved" << endreq; ... // iterators over the container ElectronContainer::const_iterator elecItr = elecTES->begin(); ElectronContainer::const_iterator elecItrE = elecTES->end(); for (; elecItr != elecItrE; ++elecItr) { double pT = (*elecItr)->pt(); ... StoreGate keys for AOD Container (1):  StoreGate keys for AOD Container (1) StoreGate keys for AOD Container (2):  StoreGate keys for AOD Container (2) StoreGate keys for Atlfast AOD Container :  StoreGate keys for Atlfast AOD Container Glimpse at one of the classes: Electron:  Glimpse at one of the classes: Electron Electron.h in PhysicsAnalysis/AnalysisCommon/ParticleEvent class Electron : public ParticleBase, public P4EEtaPhiM, public Navigable<egammaContainer,double> { int isEM() const { return m_isEM; } bool hasTrack() const { return m_hasTrack; } const Rec::TrackParticle* track() const { return ((m_hasTrack) ? (*m_track) : 0) ; } Via navigation to track (e.g. (*m_track)->trackSummary()->get( Trk::numberOfBLayerHits ) ) double z0wrtPrimVtx() const ; double d0wrtPrimVtx() const ; int numberOfBLayerHits() const int numberOfPixelHits() const; int numberOfSCTHits() const; int numberOfTRTHits() const; int numberOfTRTHighThresholdHits() const; Glimpse at one of the classes: Electron:  Glimpse at one of the classes: Electron ElectronParamDefs.h enum ParamDef { // common enums EoverP = 0, // Enum's for egamma etaBE2 = 1, et37 = 2, e237 = 3, e277 = 4, ethad1 = 5, weta1 = 6, weta2 = 7, f1 = 8, e2tsts1 = 9, emins1 = 10, wtots1 = 11, fracs1 = 12, epiNN = 13, // Enum's for softe etaCorrMag = 1, f1core = 2, f3core = 3, Pos7 = 4, Iso = 5, Widths2 = 6, emWeight = 7, pionWeight = 8 Access via (*elec)->parameter(ElectronParameters::EoverP) To find out which objects in each class:  To find out which objects in each class Hopefully soon via PAT web page https://uimon.cern.ch/twiki/bin/view/Atlas/PhysicsAnalysisTools At least that’s promised for in the next weeks Via interactive python session Unfortunately you get more than just the class members, also all methods But probably the easiest way right now Via doxigen Look directly at class in release area Some members in Kyle’s talk http://agenda.cern.ch/fullAgenda.php?ida=a045107#s2 Ask an “expert” More about 4-momentum:  More about 4-momentum Most particles such as electrons, photons, muon, tau, jets inherit from I4Momentum For example: for (; elecItr != elecItrE; ++elecItr) { if( (*elecItr)->pt()> m_etElecCut ) { double electronEta = (*elecItr)->eta(); ... Complete list double px(), py(), pz(), m() double p(), eta(), phi() double e(), et(), pt(), double iPt() double cosTh(), sinTh(), cotTh() HepLorentzVector hlv() HepLorentzVector:  HepLorentzVector xxx Exercise 4:  Exercise 4 Modify AnalysisSkeleton.cxx and plot some histograms of quantities in the ElectronContainer Retrieve ElectronContainer Plot size of it Plot pT, eta and EoverP of the electron candidates with pT>20GeV Run and check your histograms with root Watch out for the fix me in the code! More about Particle:  More about Particle Particle have a 4-momentum representation (see more on next slide) Particles have constituents (e.g. are navigable) Daughters from particle decay Reconstruction-level object methods of IParticle Slide35:  in Event/EventKernel Navigation:  Navigation Navigation to Constituents from AOD back to ESD: CompositeParticle* W = new CompositeParticle(); W->add(jetContainer, jet1, jet2); If (!W->contains(jet3)) { do something;} NavigationToken<CaloCell,double>* cellToken = new NavigationToken<CaloCell,double>(); jet->fillToken(*cellToken); NavigationToken<CaloCell, double>::const_iterator c = cellToken->begin(); NavigationToken<CaloCell,double>::const_iterator cend = cellToken->end(); for(; c != cend; ++c) { const CaloCell* thisCell = *c; double weight = jet->getParameter(thisCell); // double weight = (*c).second; double et = weight * thisCell->et(); …. Exercise 5: Plot inv. Z-mass :  Exercise 5: Plot inv. Z-mass Add to AnalysisSkeleton.cxx Use 4-vector method to combine 2 electrons with pT>20GeV and put it in histo Run and check histo Analysis Tools, Utilities:  Analysis Tools, Utilities The following tools are currently available – Look in CVS: offline/PhysicsAnalysis/AnalysisCommon/ AnalysisUtils AnalysisTools Main tools Combination Permutation DeltaR matching Selector Filter Special Utilities Navigation Association IAnalysisTools:  IAnalysisTools Delta phi between two particles deltaPhi (const INavigable4Momentum *p1, const INavigable4Momentum *p2) Delta R between 2 particles double deltaR (const INavigable4Momentum *p1, const INavigable4Momentum *p2) Invariant mass between 2 particles imass2 (const INavigable4Momentum *p1, const INavigable4Momentum *p2) Invariant mass between 4 particles imass4 (const INavigable4Momentum *p1, const INavigable4Momentum *p2, const INavigable4Momentum *p3, const INavigable4Momentum *p4) IAnalysisTools:  IAnalysisTools find the closest (in R) element in a collection to an INavigable4Momentum bool matchR (const INavigable4Momentum *t, COLL *coll, int &index, double &deltaR) bool matchR (const INavigable4Momentum *t, COLL *coll, ELEMENT *element, double &deltaR) bool matchR (const INavigable4Momentum *t, COLL *coll, int &index, double &deltaR, const int pdg) sort by pT, eta, phi sortPT (COLL *coll) sortEta (COLL *coll) sortPhi (COLL *coll) classify by charge (pos, neg) classifyCharge (const COLL *coll, std::vector<typename COLL::value_type> &pos, std::vector<typename COLL::value_type> &neg) DeltaR Matching:  DeltaR Matching Get a handle on the tools in the initialize() method: IAlgTool *tmp_analysisTools; sc = toolSvc->retrieveTool("AnalysisTools",tmp_analysisTools); if (StatusCode::SUCCESS != sc) { mLog << MSG::ERROR << "Can't get handle on ana tools" << endreq; return StatusCode::FAILURE; } Then in execute m_analysisTools=dynamic_cast<IAnalysisTools *>(tmp_analysisTools); int index = -1; double deltaRMatch; /// find a match to this electron in the MC truth container /// the index and deltaR are returned bool findAMatch = m_analysisTools->matchR((*elecItr), mcpartTES, index, deltaRMatch, (*elecItr)->pdgId()); /// the matched object and deltaR are returned bool findAMatch = m_analysisTools->matchR((*elecItr), mcpartTES, element, deltaRMatch, (*elecItr)->pdgId()); Combination Example: Wjj:  Combination Example: Wjj AnalysisUtils::Combination<const ParticleJetContainer> comb(jetTDS,2); get all the combinations of 2 jets std::vector<const IParticleContainer::base_value_type*> jj; while (comb.get(jj)) { double mjj = (jj[0]->hlv()+jj[1]->hlv()).m(); m_jj->fill(mjj,m_eventWeight); if ( fabs(mjj-mW) < m_deltaMjj ) { CompositeParticle * Wjj = new CompositeParticle(); Jet* newJet = new Jet( oldjet ) Wjj->add(jetTDS,jj[0],jj[1]); Wjj->set_charge(1); Wjj->set_pdgId(PDG::W_plus); Wjj->set_dataType(m_dataType); m_WjjContainer->push_back(Wjj); } } } Special Utility for Neutrino:  Special Utility for Neutrino In some analyses, you want to reconstruct neutrino NeutrinoIParticle To help you in PhysicsAnalysis/AnalysisCommon/SpecialUtils candidatesFromWMass(lepton, pxMiss, pyMiss, neutrinoContainer) neutrinosFromColinearApproximation(particleA, particleB, pxMiss, pyMiss, neutrinoContainer) Objective is: each user does not have to write the same piece of code for general solutions to a subclass of problem Example: use the W mass constraint for the pz solution of the neutrino momentum for (; leptonItr != leptonItrE; ++leptonItr) { bool findNeutrino = NeutrinoUtils::candidatesFromWMass((*leptonItr), m_pxMiss, m_pyMiss, (*m_neutrinoContainer)); if (findNeutrino) { do something …} } Special Particle: CompositeParticle:  Special Particle: CompositeParticle CompositeParticle is made up of other IParticles add two particles to make a composite particle In 9.0.4 (need symlink with ParticleBaseContainer) CompositeParticle* myZ = new CompositeParticle myZ->add(myBJetContainer, oldEle ) In 10.0.0 (no need for container, but need to make a copy before adding) CompositeParticle* myZ = new CompositeParticle Electron* newEle = new Electron( oldEle ) Composite particles can Navigate to its daughters Creating Your Own Containers:  Creating Your Own Containers MuonContainer * muonContainer = new MuonContainer(); Muon * newMuon = new Muon(); newMuonset_IDTrack(…); newMuonset4Mom(HepLorentzVector(px,py,pz,e)); … muonContainer->push_back(newMuon); m_storeGaterecord(muonContainer, “containerName”); The muonContainer “owns” all the newMuon inside it: Delete muonContainer; will also delete all the newMuon Some caveats if you put particles in your container:  Some caveats if you put particles in your container PhotonContainer* photonContainer= new PhotonContainer(SG::VIEW_ELEMENTS); photonContainerpush_back(newPhoton); photonContainer does not own the newPhoton inside it Delete photonContainer; will NOT delete the newPhotons The AOD container that you retrieve from POOL owns its AOD If you create your own container of pre-selected AOD, be careful: m_userElectronContainer->push_back(*elecItr); *elecItr is pointer to an Electron AOD in original electron AOD container: that container own *elecItr: m_userElectronContainer can NOT own *elecItr. So m_userElectronContainer must be previously created with “View elements” Container ownership: SG::OWN_ELEMENTS, SG::VIEW_ELEMENTS Exercise 6: again Z mass peak:  Exercise 6: again Z mass peak This time use some of the tools you just learned 1) Use tool for combinatorics 2) Now write a selector which only selects electron with pT>20GeV and use that within the combinatorics tool Note there is the method bool goodOnes(CALLER *caller, OUT &comb, CRITERIA criteria) with CALLER = AnalysisSkeleton OUT = std::vector<const Electron*, std::allocator<const Electron*> > CRITERIA = bool (AnalysisSkeleton::*) (const std::vector<const Electron*, std::allocator<const Electron*> >&), COLL = const ElectronContainer] Selection routine must be friend of AnalysisSkeleton You need something like selectElectron(AnalysisSkeleton * self, const std::vector<const ElectronContainer::base_value_type*> &ll) More Examples:  More Examples Look in the package AnalysisExamples https://uimon.cern.ch/twiki/bin/view/Atlas/AnalysisExamples (a) Batch production of large scale ESD and AOD (b) Merging AOD files (c) Documentation (d) Single Particle Identification Example (e) Z -> 2 leptons example (f) Higgs -> 4 leptons example (g) ttbar example (h) How to use the analysis tools (i) How to do back navigation from the AOD to the ESD (j) Access to tracks (k) Interface to the interactive analysis (l) How to use ntuple and histograms in your analysis algorithms Some words on MC Truth:  Some words on MC Truth Aim to do a navigation between mother and daughters in Truth block very similar to HepMC Can exchange an AOD container from reconstruction with TruthParticle container User analysis can run transparently either on MC truth or on reconstructed data Within athena truth is stored as A HepMC::GenEvent, which contains HepMC::GenVertex HepMC::GenParticle Atlfast GenEvent contains truth from Event Generator Full simulation Truth from Event Generator + GEANT4 hits Available filters for MC Truth:  Available filters for MC Truth McVtxFilter tool Usefull for selecting particles such as Z ee, , bb.. McAODFilter.McVtxFilterTool.DecayPatterns += [ "Z[e+,e]" ] Also possible McAODFilter.McVtxFilterTool.DecayPatterns += [ "#id 6[24,5]" ] this is the decay pattern for a t->W+ b And simple ones work as well McAODFilter.McVtxFilterTool.DecayPatterns += [ "gamma" ] Selection filter easy selection of particle by flavour/type Examples (not only MCTruth, also reco objects): //Create a filter of electrons AnalysisUtils::Selection<PdgIdFilter<ParticleBase> > eleFilter; eleFilter.set_pdgId( PDG::e_minus ); // loose cut on sign: selects both e+ and e- eleFilter.set_matchSign( false ); // Create a IParticleFilter to select jets AnalysisUtils::Selection<IParticleFilter> filter; filter.set_ptMin( 60.*GeV ); std::vector<ParticleJet*> myJets; filter.getObjects( jets, myJets ); Few words on ESD :  Few words on ESD Same can be done on ESD’s Instead of using the electron you use the egammaObject from the reconstruction start with the AOD but then go to ESD via back navigation EventSelector.BackNavigation = True Advantage you have the particle Electron and can use 4-vector utilities But of course you can start directly from ESD Example in ~mwielers/maxidisk/tutorial_files/results/ZeeOnESD.cxx If we have time and you’re still keen Just have a look at it I doesn’t work…. It crashes due to incompatibility between 9.0.4 and 10.0.0 Might work with new AOD/ESD’s but not sure In principle:  In principle copy ZeeOnESD.cxx to src copy ZeeOnESD.h to UserAnalysis Add to requirements use egammaEvent egammaEvent-01-* Reconstruction Add ZeeOnESD to src/components/UserAnalysis_entries.cxx Change jobOptions theApp.TopAlg += [ "ZeeOnESD" ] ZeeOnESD = Algorithm( "ZeeOnESD" ) ZeeOnESD.egamma Container = "egammaCollection" ZeeOnESD.egammaEtCut = 20.0*GeV ZeeOnESD.egammaEtaCut = 2.5 Add ZeeOnESD to src/components/UserAnalysis_entries.cxx New AOD’s produced with release 10 available in Xxx I will try to get the Z->ee example running under 10 soon Will also give example for an analysis only on ESD level Last remark:  Last remark Results of the exercises are in ~mwielers/maxidisk/tutorial_files/result

#number presentations

Add a comment

Related presentations

Related pages

⚡Presentation "Analysis of BS, ESD and AOD data What ...

Analysis of BS, ESD and AOD data What trigger data is available offline and what data should be available? How are we going to analyse HLT data?
Read more

⚡Presentation "Full Dress Rehearsal (FDR1) studies Sarah ...

Datasets used So far: –3075,3077,3078,3079 in egamma stream fdr08_run1.000XXXX.StreamEgamma.merge.AOD.o1_r12_t1 (in /data/atlas07/sallwood/FDRtest ...
Read more

Antivirus xp 2010

... [VPS 100305-0] 4.8.1368 [ (!) ... KTF MUSIC AoD Server ... http://www.malekal.com/tutorial_MalwareBytes_AntiMalware.php
Read more

397-2-evo 3d forum - forum prfiles - Pastebin.com

397-2-evo 3d forum - forum prfiles. ... http://www.germanhs.de/aod//profile.php?mode ... http://t3-recovery.com/forum/profile.php?mode=viewprofile&u=100305.
Read more


注意事項 1. 協会会員の方(または関係者)以外は投稿することができません。 2. 名前の欄には本名、あるいは本名 ...
Read more


... aDt http://blloves.imotor.com/viewthread.php?tid=140106&extra= lNc http://www.xszjbook.com/read.php?tid=34620&ds=1 aOd http://www.juece.com.cn/news ...
Read more

Full text of "Statistical Abstract of the United States"

Full text of "Statistical Abstract of the United States" See other formats ...
Read more

Lopesito Blog about Files:So-netブログ

... o quinze capitulo 15 e 17 raquel de queiroz haynes manual renault 19 91-92 Train.Signal.Cisco.CCNP.ISCW.642-825.Training.Tutorials ... d3 AOD-v2 .1.2 ...
Read more


[357013] SfTbNbxuVQhpuIcUiMp 投稿者:paycheck advance 投稿日:2012/12/02(Sun) 14:51 This establishment featured to traditions that its ...
Read more