Adding a very loose configuration of the InDet trackselector for the muon trigger (mostly to include the pT>2 GeV cut (which we had implemented separately in Run 2, but is currently missing in the run 3 trigger) to reduce the CPU usage - ATR-18811).

Rename the package to `TrigMuonHypo` and drop the "MT" from all
component and files names.

Add a custom `logging` formatter for all child loggers in the
TriggerMenuMT package. Due to the deep module hierarchy, logger names
often become too long and make log file reading difficult. The custom
formatter now only shows the first and last module name, e.g.:

Py:TriggerMenuMT.HLTMenuConfig.UnconventionalTracking.FullScanLRTConfiguration
becomes
Py:TriggerMenuMT:FullScanLRTConfiguration

Also remove all custom logger names and use the module `__name__`
consistently.

Mark Sutton authored 3 years ago and Johannes Elmsheuser committed 3 years ago

This si strange, if we patched it before because of differenece, and not we run again
asking if we want to reverse the patch what does that mean ? That the same reference is used
by two different jobs, each of which gives different output ?

Savanna Shaw authored 4 years ago and Walter Lampl committed 4 years ago

A couple updates to the trigger config for the MDT drift circle on track creator and trackbuilder to bring the new JO inline with the standard configuration. Also making fixing teh MuonExtrapolator config to include the updated properties, and switching on some of the trigger monitoring.

Edward Moyse authored 4 years ago and Walter Lampl committed 4 years ago

This was mostly done by enabling wrapping (it's off by default) and using my comparison script:
https://gitlab.cern.ch/emoyse/Scripts/-/blob/master/CompareConfigs.py

Changes to match old config
- MuonTrackSummaryHelperToolCfg should use AtlasExtrap
- InDetTrackSummaryHelperToolCfg should have InDet set to true
- MooCandidateMatchingToolCfg should use MuonStraightLineExtrap
- MuonCombinedInDetDetailedTrackSelectorToolCfg uses new AtlasTrackSummaryToolCfg
- SolenoidalIntersectorCfg has same name as old config, to allow for deduplication
- misc changes to match naming of oldconfiguration
- Correct naming of MuonCombinedInDetCandidateAlgCfg and added missing LRT configuration (currently off by default)
- Add missing MuonAlignmentErrorDbAlg to MuonGeometryConfig and added missed folder for MuonAlign errors

Improvements to configuration:
- Added AtlasTrackSummaryToolCfg
- Wrapping Debug commented out (will be removed once debugging finished)
- switch MuonCreatorTool to use summarytool which knows about ID
- added some more new flags to wrapping (will eventually need to handle this all properly)
- MuonCombinedRecoConfig: Need to enable ID via Detector.Reco flags now
- Temporarily comment out MuonAlignmentErrorDbAlg (causes lots of errors in some workflows)

Turning the muon byte stream decoding on only if we aren't running on MC, and adding the RDO containers to the view data verifier.

Properly configuring the MuonLayerHoughTool to not use the truth info when running in the trigger.

Adding the inside-out reco to the muon triggers in the new JO test (ATR-20388):

- Add the inside-out reco algorithms to the muon trigger configuration, and run in a sequence only if we don't find any muons in the outside-in reco sequence
- Add the configuration for the algorithms that check for outside-in muons and merges the muon trigger containers
- Update the muon reco config to remove a couple cases of unneeded 'addPublicTools' and to use the correct track summary tool for a given use case (CombinedMuonTrackSummaryTool when looking at combined muons, and the MuonTrackSummaryTool otherwise)

Adding a muon chain with track isolation to the trigger new JO test (ATR-20388):
- Added the new JO configuration for the isolation algorithm and tool
- Added the menu configuration for the isolation step

Also updated slightly the flags for combined muons (to switch off running some things like staco muons that we don't use in the trigger)

A few fixes for combined muons in the trigger new JO test:
- Create separate flags for SA vs combined muons, so that the combined muons are configured correctly
- Remove a few cases of addPublicTool that are no longer necessary (since the tools are all private)
- Fix some a bit the trigger configuration so that the right collections are available for combined muons.

Adjusting the EF sequence configuration so we can run in both RoI and full scan mode, and adding a single muon full scan trigger to the new JO menu. (ATR-20388).

Savanna Shaw authored 4 years ago and Vakhtang Tsulaia committed 4 years ago

Refactoring the muon trigger new JO config a bit to make it easier to follow what each step is doing.

Updating the muon trigger to use the SelectionCA, which factors out a bit some of the step/hypo setup.

Applying some fixes to the trigger new JO config to be able to run muon triggers independant of the rest of the menu:
- Removing the confguration of region selector service from the general JO as this sets up the region selector for all detectors (the service is only needed for the Calo based triggers, and is already configured in the calo data access configuration, so isn't needed generally. All other detectors use the detector specific region selector tools)
- Adding the configuration of the LAr/Tile geometries to the muon trigger config (needed since we extrapolate muons through the calorimeter).

Savanna Shaw authored 4 years ago and Edward Moyse committed 4 years ago

A few updates for the muon trigger in the new JO test (mostly related to the decoding):
- Turn on the seeded (RoI based) BS decoding for the trigger (including configuration of the Region selector which is needed for the seeded decoding)
- Removing the passing of the `doTrigger` variable to the decoding alg configurations, since we have an actual flag that we use elsewhere in the muon configuration.
- Refactor the decoding in the trigger configuration so that it can be used in two different steps of the muon triggers that run in different RoIs
- Fix a bug in the muon trigger TrackParticleCnvAlg configuration that was leading to a low efficiency of MS-only muons

Adding combined muons to the trigger new JO test (ATR-20388):
- Include the relevent offline algorithms for combined muons to the trigger test
- Update slightly the MuonCombinedTagTool configuration to be able to run in the trigger
- Add empty steps to chains not running the combined muons so that combined chains will work (need the same number of steps)

Also updated the input maker for the muComb and EF MS steps to use the RoIs centered on the L2 MS tracks to try to improve the efficiency for later steps for low pT muons.

Savanna Shaw authored 4 years ago and Adam Edward Barton committed 4 years ago

Adding the L2 combined muon step to the trigger new JO test (ATR-20242). The step runs:
- A view data verifier to pick up the L2 SA muons from the previous step
- The ID fast track finding
- the muComb algorithm

Currently it seems that the combined electron+muon chains need to have the same number of steps for electrons and muons, so I split the muon chains into two type. The default chains (used in combination with other signatures) run the L2 SA and combined steps, and the msonly chains run the L2 SA and precision SA steps (and the L2 combined and preicion SA steps are only created for the relevant chains).

Updating a bit the muon trigger for the newJO test (ATR-20388):
- Updating some parts of the configuration for muon candidate builiding to be more appropriate for the trigger use case
- Setting up the ID geometry so we can extrapolate muons through the full detector
- Adding the EF MS hypo to replace the fake hypo

Adding the building of muoncandidates and xAOD muons for the SA part of the muon trigger in the new JO test. In principle this lets us run ms-only type chains in the new configuration (or will once I add a sensible hypo in the future).
- Made some changes in the configuration to various tools for trigger running
- Added configuration for alg to convert MS tracks to xADO trackParticles (temporarily in trigger config)
- Updated trigger configuration to call muon candidate and creator algs

Part of the migration in ATR-20388

Removing some data from the view data verifier because it is produced by algorithms already running in the view.