Thoughts on MC samples needed for studying particle flow in ATLAS

Intro

• Want to produce some samples with the following features:
  • Calibration hits on (so can see how much of true energy is in cluster)
  • Varying pileup in steps of 10 from 0 - 80 (this is all in one sample but the different pileup regimes can be picked out by during analysis)
  • An increased fraction of events with BCID=1 (ie. which have no out of time pileup). In the MC12 bunch train we have 4 mini trains of 32 bunches each = 144 bunches and only 1 see's no OOT pileup. Want to increase this by a factor of 25 so ~20% of the sample will have no OOT pileup.
  • Want to produce a new DESD which is AOD + Calibration HITs + All calo cells for these studies
  • Everything else should be standard MC12 full sim configuration (including pileup noise suppression threshold and LAr OFCs for now)
• In the end will want various samples in this configuration - for now just want a test sample to check everything works ok.
• 50k events would give ~5k each for the pileup points and ~1k each of no OOT pileup for each pileup point (~4k each with 'normal' OOT pileup)
• In the end may want to try 25ns, 14TeV, ...
• Probably can turn off trigger in these samples to speed things up (anything else slow that can be turned off? (maybe muons? FCAL?) )
• Pileup truth is too big to have on
• What signal samples to use?

DESDM_ALLCELLS

• For studies that require calo cells we can use the new DESDM_ALLCELLS more details here here

Technical details

• From talking to John Chapman for the mu range:

You'd replace RunDependentSimData/configLumi_mc12_v1.py in the command by the following preInclude:

RunDependentSimData/configLumi_muRange.py

and add the following to preExec_h2r

preExec_h2r='userRunLumiOverride={"run":195847, "startmu":0.0, "endmu":50.0, "stepmu":10.0, "startlb":1, "timestamp":1328039085}'

(Ayana says need to start from 0.01to avoid a problem with mu=0)

• To change BCID distribution need to make a new version of:

http://acode-browser2.usatlas.bnl.gov/lxr-rel17/source/atlas/Simulation/SimulationJobOptions/share/pileup/preInclude.PileUpBunchTrains2011Config8_DigitConfig.py?v=release_17_2_3

Some tests

• Tried to run a test running (in 17.2.8.4):

DigiMReco_trf.py DataRunNumber='195847' HighPtMinbiasHitsFile=root://eosatlas//eos/atlas/atlasdatadisk/mc12_8TeV/HITS/e1119_s1469_s1471/mc12_8TeV.119996.Pythia8_A2MSTW2008LO_minbias_inelastic_high.merge.HITS.e1119_s1469_s1471_tid743322_00/HITS.743322._010000.pool.root.1 LowPtMinbiasHitsFile=root://eosatlas//eos/atlas/atlasdatadisk/mc12_8TeV/HITS/e1119_s1469_s1471/mc12_8TeV.119995.Pythia8_A2MSTW2008LO_minbias_inelastic_low.merge.HITS.e1119_s1469_s1471_tid743321_00/HITS.743321._002000.pool.root.4 autoConfiguration='everything' conditionsTag='OFLCOND-MC12-AFII-07' digiRndmSvc='AtDSFMTGenSvc' digiSeedOffset1=0 digiSeedOffset2=1 geometryVersion='ATLAS-GEO-20-00-01' inputHitsFile=../user.jboyd.123456.MadGraph_ttbargammaPt30.201302181008_Merge_AFII_100_a188_v5_TEST2.130218152938_EXT0/user.jboyd.001007.EXT0._00017.HITS.pool.root jobNumber=0 maxEvents=10 numberOfHighPtMinBias='0.045336' numberOfLowPtMinBias='39.954664' outputAODFile='AOD.pool.root' outputESDFile='ESD.pool.root' outputTAGFile='TAG.root' preExec_e2a='rec.Commissioning.set_Value_and_Lock(True);jobproperties.Beam.energy.set_Value_and_Lock(4000*Units.GeV);muonRecFlags.writeSDOs=True;jobproperties.Beam.numberOfCollisions.set_Value_and_Lock(20.0);jobproperties.Beam.bunchSpacing.set_Value_and_Lock(50)' preExec_h2r='from LArROD.LArRODFlags import larRODFlags;larRODFlags.doOFCPileupOptimization.set_Value_and_Lock(True);larRODFlags.NumberOfCollisions.set_Value_and_Lock(20);larRODFlags.UseDelta.set_Value_and_Lock(3)' preExec_r2e='from LArROD.LArRODFlags import larRODFlags;larRODFlags.doOFCPileupOptimization.set_Value_and_Lock(True);larRODFlags.NumberOfCollisions.set_Value_and_Lock(20);larRODFlags.UseDelta.set_Value_and_Lock(3);rec.Commissioning.set_Value_and_Lock(True);jobproperties.Beam.energy.set_Value_and_Lock(4000*Units.GeV);muonRecFlags.writeSDOs=True;jobproperties.Beam.numberOfCollisions.set_Value_and_Lock(20.0);jobproperties.Beam.bunchSpacing.set_Value_and_Lock(50);from CaloRec.CaloCellFlags import jobproperties;jobproperties.CaloCellFlags.doLArCellEmMisCalib=False' preInclude_h2r='Digitization/ForceUseOfPileUpTools.py,SimulationJobOptions/preInclude.PileUpBunchTrains2011Config8_DigitConfig.py,RunDependentSimData/configLumi_mc12_v1.py' tmpRDO='RDO.pool.root' skipEvents=0 triggerConfigByRun='{195847:"MCRECO:DB:TRIGGERDBMC:324,143,265"}' postInclude_h2r=FastSimulationJobTransforms/postInclude.AF2DigiSetContainerNames.py

• from https://twiki.cern.ch/twiki/bin/viewauth/AtlasProtected/PrivateSampleProduction#Interactive_AN5
• Needed to change a few things:LArRODFlags -> LArRODFlags, CaloCellFlags -> CaloCellFlags, had to remove the postExec_e2a - not sure why?
• The produced TAG file does not have the expected mu distribution (=0), but does seem to have many vertices - not sure whats going on...
• John C suggested using: 17.2.1.4.Z-MCProd not sure how to set that up!
• Now trying with:

DigiMReco_trf.py DataRunNumber='195847' HighPtMinbiasHitsFile=root://eosatlas//eos/atlas/atlasdatadisk/mc12_8TeV/HITS/e1119_s1469_s1471/mc12_8TeV.119996.Pythia8_A2MSTW2008LO_minbias_inelastic_high.merge.HITS.e1119_s1469_s1471_tid743322_00/HITS.743322._010000.pool.root.1 LowPtMinbiasHitsFile=root://eosatlas//eos/atlas/atlasdatadisk/mc12_8TeV/HITS/e1119_s1469_s1471/mc12_8TeV.119995.Pythia8_A2MSTW2008LO_minbias_inelastic_low.merge.HITS.e1119_s1469_s1471_tid743321_00/HITS.743321._002000.pool.root.4 autoConfiguration='everything' conditionsTag='OFLCOND-MC12-AFII-07' digiRndmSvc='AtDSFMTGenSvc' digiSeedOffset1=0 digiSeedOffset2=1 geometryVersion='ATLAS-GEO-20-00-01' inputHitsFile=../../user.jboyd.123456.MadGraph_ttbargammaPt30.201302181008_Merge_AFII_100_a188_v5_TEST2.130218152938_EXT0/user.jboyd.001007.EXT0._00017.HITS.pool.root jobNumber=0 maxEvents=10 numberOfHighPtMinBias='0.045336' numberOfLowPtMinBias='39.954664' outputAODFile='AOD.pool.root' outputESDFile='ESD.pool.root' outputTAGFile='TAG.root' preExec_e2a='rec.doTrigger=False;rec.Commissioning.set_Value_and_Lock(True);jobproperties.Beam.energy.set_Value_and_Lock(4000*Units.GeV);muonRecFlags.writeSDOs=True;jobproperties.Beam.numberOfCollisions.set_Value_and_Lock(20.0);jobproperties.Beam.bunchSpacing.set_Value_and_Lock(50)' preExec_h2r='from LArROD.LArRODFlags import larRODFlags;larRODFlags.doOFCPileupOptimization.set_Value_and_Lock(True);larRODFlags.NumberOfCollisions.set_Value_and_Lock(20);larRODFlags.UseDelta.set_Value_and_Lock(3);userRunLumiOverride={"run":195847, "startmu":0.01, "endmu":50.0, "stepmu":10.0, "startlb":1, "timestamp":1328039085}' preExec_r2e='rec.doTrigger=False;from LArROD.LArRODFlags import larRODFlags;larRODFlags.doOFCPileupOptimization.set_Value_and_Lock(True);larRODFlags.NumberOfCollisions.set_Value_and_Lock(20);larRODFlags.UseDelta.set_Value_and_Lock(3);rec.Commissioning.set_Value_and_Lock(True);jobproperties.Beam.energy.set_Value_and_Lock(4000*Units.GeV);muonRecFlags.writeSDOs=True;jobproperties.Beam.numberOfCollisions.set_Value_and_Lock(20.0);jobproperties.Beam.bunchSpacing.set_Value_and_Lock(50);from CaloRec.CaloCellFlags import jobproperties;jobproperties.CaloCellFlags.doLArCellEmMisCalib=False' preInclude_h2r='Digitization/ForceUseOfPileUpTools.py,SimulationJobOptions/preInclude.PileUpBunchTrains2011Config8_DigitConfig.py' tmpRDO='RDO.pool.root' skipEvents=0 triggerConfigByRun='{195847:"MCRECO:DB:TRIGGERDBMC:324,143,265"}' postInclude_h2r=FastSimulationJobTransforms/postInclude.AF2DigiSetContainerNames.py preExec='ScaleTaskLength=0.1' 

• this is trying to include: new mu distribution, trigger off, ScaleTaskLength=0.1 (for testing on 10 events)
• this gives a strange error from digi:

EFlowRec setup

• I do the following (thanks to Mark H) (in 17.2.7.5):

> export SVNROOT=$SVNUSR
> cmt co MakeCellInfo mhodgkin/MakeCellInfo
> cmt co CalCellInfoEvent mhodgkin/CalCellInfoEvent 
> cmt co eflowD3PDMaker mhodgkin/eflowD3PDMaker
> cp ~jboyd/public/eflow/eflowD3PD.py mhodgkin/eflowD3PDMaker/python
> cd mhodgkin/MakeCellInfo/cmt/
> cmt br make
> cd -
> cd mhodgkin/eflowD3PDMaker/cmt
> cmt br make
running with a topOption.py that you can copy from: ~jboyd/public/eflow/topOptions.py
and you also need: ~jboyd/public/eflow/myAlgs.py in you run directory.

• (I have verified that using the DESDM_ALLCELLS gives the same results as ESD, and that the DESD contains the calibration hits if the ESD it is made from also contains calibration hits)
• (You can make the DESD from ESD in 17.2.X.Y if you have this tag: PrimaryDPDMaker-00-04-50-01 checked out and compiled and then run the command:

>Reco_trf.py inputESDFile=ESD.pool.root  outputDESDM_ALLCELLSFile=desdm.root maxEvents=50

• I copied a DESDM from a J1 ESD (without calibration hits) with mu=12 to: ~jboyd/public/eflow/desdm.root
• Currently the D3PD only contains the eventInfo (runnumber, lumiblock etc...), the primary vertex info and the eflow stuff - will want to add more stuff in future...