SiD Full Reconstruction Examples

This page gives detailed descriptions of the steps necessary to run the full reconstruction using the SiD software chain.

CLIC CDR Production

This example describes the steps performed for the CLIC CDR mass production for a center-of-mass energy of 3 TeV. 60 BX worth of γγ→hadrons background is overlaid over each signal event. The incoherent pair background is not taken into account. LumiCal and BeamCal collections are ignored, since the occupancy due to the missing incoherent pair background would be largely underestimated. Three different collections of reconstructed particles are selected based on their mean cluster time (with varying cuts) to be used in a further analysis step.

The required input files are simulated event samples for the signal and the γγ→hadrons background. The software versions used are

• org.lcsim: CLIC_CDR (based on 1.18)
• slicPandora: CLIC_CDR

Background Overlay and Tracking

Pandora PFA

PFO Selection, Truth Linking and DST Creation

ILC DBD Production

This example describes the steps performed for the ILC DBD mass production for a center-of-mass energy of 1 TeV. Since single bunch crossing time stamping is assumed throughout the detector, one BX worth of γγ→hadrons background and incoherent pair background is overlaid over each signal event. The BeamCal collection is ignored due to its very high occupancy.

The required input files are simulated event samples for the signal, the incoherent pair background (pre-mixed such that one event corresponds to one BX) and the γγ→hadrons background (one event corresponds to one γγ interaction). The software versions used are

• org.lcsim: 2.2
• slicPandora: ILC_DBD (slicPandora v2012-09-04, PandoraPFA v00-09, LCIO v02-01)
• Marlin: 0116 (ILCSoft v01-16 including Marlin v01-04, LCFIPlus v00-05-02)

The default detector model is sidloi3.

Background Overlay and Tracking

The first step is to overlay the beam-related backgrounds on the signal event, digitize the tracker hits and run the track reconstruction. It is done in org.lcsim. The required steering files are

• sid_dbd_prePandora.xml
• sidloi3_trackingStrategies_default.xml

To run this step type

java -jar lcsim-<version>-bin.jar sid_dbd_prePandora.xml -DinputFile=<inputFile> -DgghadFile=<gghadFile> -DpairsFile=<pairsFile> -DtrackingStrategies=<trackingStrategies> -DoutputFile=<outputFile>

The executed drivers are:

• Report on progress - org.lcsim.job.EventMarkerDriver
• Overlay γγ→hadrons - org.lcsim.util.OverlayDriver
• Overlay incoherent pairs - org.lcsim.util.OverlayDriver
• Set up tracker barrel digitization - org.lcsim.recon.tracking.digitization.sisim.config.SiTrackerBarrelSensorSetup
• Set up tracker endcap digitization - org.lcsim.recon.tracking.digitization.sisim.config.SiTrackerEndcap2SensorSetup
• Set up vertex barrel digitization - org.lcsim.recon.tracking.digitization.sisim.config.SiVertexBarrelSensorSetup
• Set up vertex endcap digitization - org.lcsim.recon.tracking.digitization.sisim.config.SiVertexEndcapSensorSetup
• Set up forward disk digitization - org.lcsim.recon.tracking.digitization.sisim.config.SiVertexEndcapSensorSetup
• Strip hit digitization - org.lcsim.recon.tracking.digitization.sisim.config.StripDigiSetupDriver
• Pixel hit digitization - org.lcsim.recon.tracking.digitization.sisim.config.PixelDigiSetupDriver
• Tracker hit conversion - org.lcsim.fit.helicaltrack.HelicalTrackHitDriver
• Remove hits from overlapping tracker modules - org.lcsim.recon.cheater.RemoveMultipleHelicalTrackHitsCheater
• Track reconstruction - org.lcsim.recon.tracking.seedtracker.steeringwrappers.SeedTrackerWrapper
• Calculate track states for PFA - org.lcsim.recon.tracking.seedtracker.SeedTrackerTrackStateDriver
• Calculate subdetector hits for vertexing - org.lcsim.recon.util.TrackSubdetectorHitNumbersDriver
• Write output file - org.lcsim.util.loop.LCIODriver

Pandora PFA

The second step is the particle flow reconstruction which includes the calorimeter hit digitization, clustering of the calorimeter hits and formation of particle flow objects (PFO), i.e. reconstructed particles, which includes a particle identification. It is done in slicPandora, which is a front end to PandoraPFA. The required steering files are:

• sid_dbd_pandoraSettings.xml
• sidloi3_pandora.xml

To run this step type

PandoraFrontend -g sidloi3_pandora.xml -c sid_dbd_pandoraSettings.xml -i <inputFile> -o <outputFile>

Vertex Finding

The third step is the vertex finding. It is done in Marlin using LCFIPlus. The required steering file is:

• sid_dbd_vertexing.xml

A GEAR file containing geometry information is not required. The magnetic field strength of the solenoid is taken from the steering file instead.

To run this step type

Marlin sid_dbd_vertexing.xml

The input and output file names have to be modified in the sid_dbd_vertexing.xml.

Truth Linking and DST Creation

The last step of the reconstruction chain creates truth links between the Monte Carlo particles and the high level reconstructed objects like tracks, clusters and reconstructed particles. Two output files are created: a REC file containing all collections and thus the full information and a reduced file format (DST) which contains only the high level reconstructed objects and the list of Monte Carlo particles, as well as the truth links. This step is performed in org.lcsim. The required steering file is

• sid_dbd_postPandora.xml

To run this step type

java -jar lcsim-<version>-bin.jar sid_dbd_postPandora.xml -DinputFile=<inputFile> -DrecFile=<recFile> -DdstFile=<dstFile>

The executed drivers are:

• Report on progress - org.lcsim.job.EventMarkerDriver
• Create truth links - org.lcsim.recon.util.McTruthLinker
• Write REC file - org.lcsim.util.loop.LCIODriver
• Remove tracker hits not associated with a track - org.lcsim.recon.util.RemoveUnusedTrackerHits
• Prepare collections for DST writing - org.lcsim.util.LCIOFlagDriver
• Write DST file - org.lcsim.util.loop.LCIODriver

The DST files contain the following collections:

• PrimaryVertex
• PrimaryVertex_RP
• BuildUpVertex
• BuildUpVertex_RP
• BuildUpVertex_V0
• BuildUpVertex_V0_RP
• PandoraPFOCollection
• RecoMCTruthLink
• Tracks
• TrackMCTruthLink
• ReconClusters
• ClusterMCTruthLink
• MCParticlesSkimmed
• HelicalTrackHits

The DST file should contain sufficient information to analyze the events. Jet clustering, flavor tagging and creation of the final tuples used in the analysis are specific to the events analyzed and are thus not performed in the production context.