athena merge requestshttps://gitlab.cern.ch/atlas/athena/-/merge_requests2024-03-26T14:12:38+01:00https://gitlab.cern.ch/atlas/athena/-/merge_requests/70099Clean-up of MDTSimHit.h2024-03-26T14:12:38+01:00John Derek ChapmanClean-up of MDTSimHit.hWas previously part of !69655, which had to be reverted while other changes
are debugged. This part is still safe to go in though.
It allows deformations to be applied to an already created MDTSimHit instance.Was previously part of !69655, which had to be reverted while other changes
are debugged. This part is still safe to go in though.
It allows deformations to be applied to an already created MDTSimHit instance.https://gitlab.cern.ch/atlas/athena/-/merge_requests/69753Add algorithm to compute technical performance for Acts tracks.2024-03-26T14:12:21+01:00Goetz GayckenAdd algorithm to compute technical performance for Acts tracks.The algorithm computes particle reconstruction efficiency of particles
which are considered reconstructible, the hit efficiency and purity.
These values are computed in eta bins (or optionally in PDG ID bins),
and can be further divided...The algorithm computes particle reconstruction efficiency of particles
which are considered reconstructible, the hit efficiency and purity.
These values are computed in eta bins (or optionally in PDG ID bins),
and can be further divided by pt. The reference truth particles can be
selected.
Also added algorithm to count measurements associated to truth particles.
The output of the latter is needed for the former.https://gitlab.cern.ch/atlas/athena/-/merge_requests/70088Use APRDefaults to name the Event Data TTree/RNTuple2024-03-26T14:10:45+01:00Marcin NowakUse APRDefaults to name the Event Data TTree/RNTupleMake the AtheaPool Conversion Service use by default the prefix defined in APRDefaults as the name prefix for POOL containers. This prefix in most cases corresponds to the name of the main Event Data TTree/RNTuple.
There should be no...Make the AtheaPool Conversion Service use by default the prefix defined in APRDefaults as the name prefix for POOL containers. This prefix in most cases corresponds to the name of the main Event Data TTree/RNTuple.
There should be no change for the standard "CollectionTree" name in normal Athena jobs. But by configuring the Output.StorageTechnology.EventData="ROOTRNTUPLE" the main RNTuple will be now by default called "EventData" (or whatever is specified in APRDefaults).
These defaults can be still overridden in the usual way by setting the Prefix property of this service.https://gitlab.cern.ch/atlas/athena/-/merge_requests/70087CaloNoise: Add method to get getEffectiveSigma by IdentiiferHash2024-03-26T13:35:24+01:00Walter LamplCaloNoise: Add method to get getEffectiveSigma by IdentiiferHashCaloNoise: Add method to get getEffectiveSigma by IdentiiferHash to avoid unnecessary back-and-forth conversion of identifiersCaloNoise: Add method to get getEffectiveSigma by IdentiiferHash to avoid unnecessary back-and-forth conversion of identifiershttps://gitlab.cern.ch/atlas/athena/-/merge_requests/70094ActsTrackFinding: simplify filling of helper objects2024-03-26T13:34:08+01:00Tim AdyeActsTrackFinding: simplify filling of helper objects* `TrackingSurfaceHelper` vector can now be filled element-by-element. This simplifies (and maybe slightly speeds up) its use in `CoreStripSpacePointFormationTool`, `ProtoTrackCreationAndFitAlg`, and `TrackFindingAlg`.
* `TrackFindingMea...* `TrackingSurfaceHelper` vector can now be filled element-by-element. This simplifies (and maybe slightly speeds up) its use in `CoreStripSpacePointFormationTool`, `ProtoTrackCreationAndFitAlg`, and `TrackFindingAlg`.
* `TrackFindingMeasurements::addDetectorElements()` call doesn't need the whole `clusterContainer`, just the measurement type.
* Remove `gatherGeoIds()`, moving functionality into `TrackFindingMeasurements::addDetectorElements`
* Improve `MeasurementRangeList` API
* Avoid taking reference to temporary object when constructing `CkfBranchStopper`. This was only used by `TrackStatePrinter`.
* Rationalise some names in `TrackFindingMeasurements`: `m_measurementOffset` → `m_measurementOffsets`, `measurementOffsets()` → `measurementContainerOffsets()`, and `measurementOffsetVector()` → `measurementOffsets()`.https://gitlab.cern.ch/atlas/athena/-/merge_requests/67272Add jet fold hash2024-03-26T11:52:53+01:00Dan GuestAdd jet fold hashThis just adds a variable to flavor tagging. Nothing else should change.
We found that our flavor tagging networks were a bit overtrained on the ttbar dataset we normally use. We're considering working around this by adding a per-jet ha...This just adds a variable to flavor tagging. Nothing else should change.
We found that our flavor tagging networks were a bit overtrained on the ttbar dataset we normally use. We're considering working around this by adding a per-jet hash and training several networks, so we can avoid ever applying the tagger to a jet we trained on.
This is adding the first part: the hash. It mixes some hit and constituent info with the event number to get something pretty random.
Tagging @svanstro, @hartman, @bdong, @liganghttps://gitlab.cern.ch/atlas/athena/-/merge_requests/69461FTF GPU ITk: Add triplet confirmation2024-03-26T11:42:45+01:00Aleksandra PorebaFTF GPU ITk: Add triplet confirmationAdd a fake triplet filter based on duplicates finding:
Criteria for being a duplicate:
- the same inner and middle spacepoint
- outer spacepoints are on different layers
- pT has the same sign
- pT is equal within stddev
Additionally:
...Add a fake triplet filter based on duplicates finding:
Criteria for being a duplicate:
- the same inner and middle spacepoint
- outer spacepoints are on different layers
- pT has the same sign
- pT is equal within stddev
Additionally:
- Set FTF useGPU flag based on InnerDetector flag
- Add two more outer layers from volumes 91/71 and 90/70 as possible middle spacepoint layers
The code will be further optimized to fully utilize GPU resources
Related to EFTRACK-76
Tagging @demelianhttps://gitlab.cern.ch/atlas/athena/-/merge_requests/60065remove CaloCalTopoClusters from FTAG12024-03-26T11:05:14+01:00Ligang Xiaremove CaloCalTopoClusters from FTAG1CaloCalTopoClusters is not needed and removed from FTAG1.
@coccaroCaloCalTopoClusters is not needed and removed from FTAG1.
@coccarohttps://gitlab.cern.ch/atlas/athena/-/merge_requests/70023Do not create views at offline level but only RoIs2024-03-26T10:30:38+01:00Carlo Varnicarlo.varni@cern.chDo not create views at offline level but only RoIsThis will replace the ACTS Event View maker with a simple algorithm for creating an RoI for later use in (e.g.) the clustering
Current sequence:
![graphviz__11_](/uploads/2cc97bd7ba463b023533d57ac638c6d3/graphviz__11_.png)This will replace the ACTS Event View maker with a simple algorithm for creating an RoI for later use in (e.g.) the clustering
Current sequence:
![graphviz__11_](/uploads/2cc97bd7ba463b023533d57ac638c6d3/graphviz__11_.png)https://gitlab.cern.ch/atlas/athena/-/merge_requests/70081main-cppcheck-DataQualityUtils2024-03-26T10:29:45+01:00Shaun Roemain-cppcheck-DataQualityUtilscppcheck: unused return value, now formally ignored.cppcheck: unused return value, now formally ignored.https://gitlab.cern.ch/atlas/athena/-/merge_requests/70076DerivationFrameworkMCTruth: add read handles to ClassifyAndCalculateHFAugment...2024-03-26T10:29:09+01:00Bertrand Martin Dit LatourDerivationFrameworkMCTruth: add read handles to ClassifyAndCalculateHFAugmentationHello,
This MR is replacing `evtStore()->retrieve` with ReadHandles, to be able to run derivations in MT (ATLASG-2470).
It fixes the error:
```
19:55:08 StoreGateSvc 0 0 WARNING retrieve(con...Hello,
This MR is replacing `evtStore()->retrieve` with ReadHandles, to be able to run derivations in MT (ATLASG-2470).
It fixes the error:
```
19:55:08 StoreGateSvc 0 0 WARNING retrieve(const): No valid proxy for object AntiKt4TruthWZJets of type xAOD::JetContainer(CLID 1244316195)
19:55:08 ToolSvc.DFCommonClassifyAndCalculateHFAugmentation 0 0 ERROR could not retrieve JetContainer ' 'jetCollectionName':'AntiKt4TruthWZJets''
19:55:08 HFClassificationCommonKernel 0 0 ERROR Augmentation failed!
```
Decorators are also replaced with DecorHandles.
Cheers,
Bertrandhttps://gitlab.cern.ch/atlas/athena/-/merge_requests/70062Adding truth labels to HLT Jets2024-03-26T10:28:07+01:00Stefano FranchellucciAdding truth labels to HLT JetsAdding FTag truth labels to HLT jets in the derivation framework. This is going to be needed for tagger calibrations (`FTAG1` and `FTAG2`), for `TLA` analysis studies, and in general for validation `PHYSVAL`.
Tagging some interested peo...Adding FTag truth labels to HLT jets in the derivation framework. This is going to be needed for tagger calibrations (`FTAG1` and `FTAG2`), for `TLA` analysis studies, and in general for validation `PHYSVAL`.
Tagging some interested people @yuchou @mamerl @mmontell @dguest @bdong @ligang @lshi @lbeemstehttps://gitlab.cern.ch/atlas/athena/-/merge_requests/69515Reorganize L1Calo Monitoring2024-03-26T09:40:54+01:00Will ButtingerReorganize L1Calo MonitoringThis is the first MR in what is a major re-organization of L1Calo monitoring as discussed @ https://indico.cern.ch/event/1364745/contributions/5777902/attachments/2809852/4903876/monitoringFeb2024.pdf
The main changes here are:
* A ...This is the first MR in what is a major re-organization of L1Calo monitoring as discussed @ https://indico.cern.ch/event/1364745/contributions/5777902/attachments/2809852/4903876/monitoringFeb2024.pdf
The main changes here are:
* A new L1CaloMonitoringCfgHelper class has been created to manage the defining of all monitoring hists and trees in a coherent way, according to the new convention
* Most of the L1Calo non-legacy monitoring algs have been moved over to using this new class
* The new class allows us to auto-generate the config file needed for DataQualityConfigurations for all the histograms.
This MR is part of a larger project that is a work in progress, but given the significant work done so far in this MR I am pushing it.
Please note lots of bits of code are commented out, this is because this is indeed still a bit of a work in progress but we need to see how things perform in the wider setups.https://gitlab.cern.ch/atlas/athena/-/merge_requests/69945Prevent ONNX from spawning additional threads2024-03-26T08:53:26+01:00Joshua Falco Beirerjoshua.beirer@cern.chPrevent ONNX from spawning additional threadsPrevent ONNX from spawning additional threads in the FastCaloSim implementation.
Should fix ATLASSIM-7146 (and potentially ATLASSIM-7031 ?)
fyi @zhangr @jchapmanPrevent ONNX from spawning additional threads in the FastCaloSim implementation.
Should fix ATLASSIM-7146 (and potentially ATLASSIM-7031 ?)
fyi @zhangr @jchapmanhttps://gitlab.cern.ch/atlas/athena/-/merge_requests/69344Return error for not setting ptau for GN2v012024-03-25T18:48:14+01:00Binbin DongReturn error for not setting ptau for GN2v01print error message when using new GN2v01 tagger but no ptau value is set.
cc @fdibelloprint error message when using new GN2v01 tagger but no ptau value is set.
cc @fdibellohttps://gitlab.cern.ch/atlas/athena/-/merge_requests/69017support tau class in btagging tool2024-03-25T18:48:14+01:00Binbin Dongsupport tau class in btagging toolAdded support for the tau class in GN2v01 b-tagger in the btagging selection tool, as mentioned in AFT-709.
The default value for ptau is set (as 0.) so no change is required for the DL1/GN2v00 user case.
cc @ligang, @svanstro @dguest,...Added support for the tau class in GN2v01 b-tagger in the btagging selection tool, as mentioned in AFT-709.
The default value for ptau is set (as 0.) so no change is required for the DL1/GN2v00 user case.
cc @ligang, @svanstro @dguest, @fdibellohttps://gitlab.cern.ch/atlas/athena/-/merge_requests/70025Update to Gaudi v38r0.0012024-03-25T16:45:17+01:00Frank WinklmeierUpdate to Gaudi v38r0.001A very minor patch to our current Gaudi version. Changes compared to v38r0.000 (see [diff](https://gitlab.cern.ch/atlas/Gaudi/-/compare/v38r0.000...v38r0.001?from_project_id=6676&straight=false)):
- New `maxParallelismExtra` property in...A very minor patch to our current Gaudi version. Changes compared to v38r0.000 (see [diff](https://gitlab.cern.ch/atlas/Gaudi/-/compare/v38r0.000...v38r0.001?from_project_id=6676&straight=false)):
- New `maxParallelismExtra` property in `AvalancheSchedulerSvc` to adjust the maximum TBB parallelism ( gaudi/Gaudi!1567)
This is required for the new async I/O HltEventLoopMgr ( ATR-26285) to be usable at Point-1. The default behavior is unchanged.
cc @ssottoco @leggetthttps://gitlab.cern.ch/atlas/athena/-/merge_requests/70079main-cppcheck-InDetPhysValMonitoring2024-03-25T14:54:12+01:00Shaun Roemain-cppcheck-InDetPhysValMonitoringMove to at-declaration initialisation of members in InDetPerfPlot_Vertex, solving uninitialised members warning in the processMove to at-declaration initialisation of members in InDetPerfPlot_Vertex, solving uninitialised members warning in the processhttps://gitlab.cern.ch/atlas/athena/-/merge_requests/70065CPAlgorithms: updated experimental OverlapRemoval2024-03-25T14:51:10+01:00Baptiste Ravinabaptiste.ravina@cern.chCPAlgorithms: updated experimental OverlapRemovalNew overlap removal scheme based on the nominal-OR decision, but run also for systematically-varied kinematic selections.
This is for experimental tests only at this stage, tagging @mdubovsk @ekourlit. I tested it locally on a small sam...New overlap removal scheme based on the nominal-OR decision, but run also for systematically-varied kinematic selections.
This is for experimental tests only at this stage, tagging @mdubovsk @ekourlit. I tested it locally on a small sample, and I see the intended behaviour in one electron event: the branch `el_select_or` is set to 1, and at the same time we have `el_select_loose_NOSYS` (the nominal ID+Iso+Pt+Eta cuts) set to 0, and all but one systematic variations also set to 0. Using the previous setup (run nominal-OR only, and only on the nominal kinematic selection), this electron event is not present at all.https://gitlab.cern.ch/atlas/athena/-/merge_requests/70063Extend SiHitCollectionCnv unit tests2024-03-25T14:49:56+01:00John Derek ChapmanExtend SiHitCollectionCnv unit testsThis merge request extends the SiHitCollectionCnv unit tests to monitor the level
of compression seen in the persistent formats and check the behaviour for SiHits caused by
delta-rays (i.e. no valid truth link).This merge request extends the SiHitCollectionCnv unit tests to monitor the level
of compression seen in the persistent formats and check the behaviour for SiHits caused by
delta-rays (i.e. no valid truth link).