Merged requested to merge adye/athena:adye-ckf7 into master
More debugging printout, following !60420 (merged). As before, printout is only activated with
- Split printing from
TrackFindingToolinto new tool,
SpacePointposition for unmeasured strip coordinates. If that fails (eg. SP is not on surface), uses center.
globalToLocalcomparison for input measurements. If
Trktransforms agree, then the printout in those columns is suppressed.
using ActsTrk::TrackFindingTool::traj_Type = Acts::VectorMultiTrajectory. It's simpler to specify
Reco_tf.py --CA \ --preExec 'flags.Acts.doPrintTrackStates=True;' \ --preInclude InDetConfig.ConfigurationHelpers.OnlyTrackingPreInclude,ActsInterop.ActsCIFlags.actsValidateTracksFlags \ --inputRDOFile /cvmfs/atlas-nightlies.cern.ch/repo/data/data-art/PhaseIIUpgrade/RDO/mc15_14TeV.900040.PG_singlemu_Pt100_etaFlatnp0_43.recon.RDO.e8185_s3856_r13824/RDO.29831688._000001.pool.root.1 \ --outputAODFile AOD.pool.root \ --maxEvents 20
globalToLocal comparison works for everything except for strip discs (search for
log.RAWtoALL). The issue is probably with the
Trk local coordinate transform (
localPositionStrip2D line 309), which I cobbled together by trial and error to be fairly close, but not perfect.
This is a big improvement on what was there before, so I think we should merge this to aid further debugging. Just to be clear, the comparison is only done if
ConfigFlags.Acts.doPrintTrackStates=True and even when it fails, it just shows the mismatched coordinates (no error message).