Moin everybody,
initially triggered by these harmless appearing WARNINGS
RAWtoALL 09:00:19 MuonCombinedAlg.MuonCom...CombinedMuonTrackBuilder 407 4 WARNING appendSelectedTSOS:: skip duplicate measurement of type drift circle
RAWtoALL 09:00:19 MuonCombinedAlg.MuonCom...CombinedMuonTrackBuilder 407 4 WARNING appendSelectedTSOS:: skip duplicate measurement of type drift circle
RAWtoALL 09:00:19 MuonCombinedAlg.MuonCom...CombinedMuonTrackBuilder 407 4 WARNING appendSelectedTSOS:: skip duplicate measurement of type drift circle
RAWtoALL 09:00:19 MuonCombinedAlg.MuonCom...CombinedMuonTrackBuilder 407 4 WARNING appendSelectedTSOS:: skip duplicate measurement of type drift circle
RAWtoALL 09:00:19 MuonCombinedAlg.MuonCom...CombinedMuonTrackBuilder 407 4 WARNING appendSelectedTSOS:: skip duplicate measurement of type drift circle
RAWtoALL 09:00:19 MuonCombinedAlg.MuonCom...CombinedMuonTrackBuilder 407 4 WARNING appendSelectedTSOS:: skip duplicate measurement of type drift circle
RAWtoALL 09:00:19 MuonCombinedAlg.MuonCom...CombinedMuonTrackBuilder 407 4 WARNING appendSelectedTSOS:: skip duplicate measurement of type drift circle
RAWtoALL 09:00:19 MuonCombinedAlg.MuonCom...CombinedMuonTrackBuilder 407 4 WARNING appendSelectedTSOS:: skip duplicate measurement of type drift circle
RAWtoALL 09:00:19 MuonCombinedAlg.MuonCom...CombinedMuonTrackBuilder 407 4 WARNING appendSelectedTSOS:: skip duplicate measurement of type drift circle
RAWtoALL 09:00:19 MuonCombinedAlg.MuonCom...CombinedMuonTrackBuilder 407 4 WARNING appendSelectedTSOS:: skip duplicate measurement of type drift circle
I dug into the two rabbit holes: The hit sorting & counting in the MuPatTrackBuilder
and the hit recovery throughout the muon chain.
-
Hit sorting on the MS track: If two hit lists are merged, the track parameters of one list are recalculated. I.e., the extrapolator is launched to extrapolate the surface parameters In addition, the complete method of adding the hit to the list is horribly complicated :scream::scream::scream:. The sorting is now simplified drastically exploiting the centers of the associated readout element. In rare cases, it can happen that 2 cluster hits from the same layers are added to the track fit. These hits are then sorted by their respective residual pull and removed later
-
MuonChamberHoleRecoveryTool:
data22_13p6TeV spot test:
Algorithm | Time per event (22.0) | Time per event (patch) |
---|---|---|
MuPatTrackBuilder | 163 | 145 |
MuPatTrackBuilder_EMEO | 182 | 180 |
MuonCombinedMuonCandidateAlg | 143 | 115 |
MuonCombinedMuonCandidateAlg_EMEO | 70 | 70 |
MuonCombinedAlg | 151 | 130 |
MuonCombinedAlg_EMEO | 103 | 81 |
MuonCombined_LRT | 19 | 13 |
MuonInsideOutRecoAlg | 204 | 145 |
MuGrilAlg_LRT | 64 | 47 |