athena merge requestshttps://gitlab.cern.ch/atlas/athena/-/merge_requests2021-05-18T03:04:36+02:00https://gitlab.cern.ch/atlas/athena/-/merge_requests/43511Slight cleanup for b-jet trigger EDM2021-05-18T03:04:36+02:00Dan GuestSlight cleanup for b-jet trigger EDMA bit of cleanup around the b-jet trigger EDM:
- Remove some vector-per-jet (one entry per track) variables for IPxD taggers. None of what I've removed is required for training any part of the training, and I don't remember using any of...A bit of cleanup around the b-jet trigger EDM:
- Remove some vector-per-jet (one entry per track) variables for IPxD taggers. None of what I've removed is required for training any part of the training, and I don't remember using any of it for useful diagnostics in quite a few years. We're also trying to get away from using IPxD at all.
- Add the `Jvt` and the JVT inputs to the b-jets. The interplay between JVT and b-tagging in the trigger is something we're going to need to study.
Any opinions from @cpollard ?https://gitlab.cern.ch/atlas/athena/-/merge_requests/43505Add Precision Tracking to the existing Fullscan LRT Chains2021-05-15T03:05:52+02:00Benjamin Philip KerridgeAdd Precision Tracking to the existing Fullscan LRT ChainsThese changes add Precision Tracking to the existing fullscan LRT chains.
Changes:
- EDM Changes to support PT (adds 2 collections `HLT_IDTrack_FSLRT_IDTrig` and `HLT_IDTrack_FSLRT_IDTrigAux`)
- Adds Precision Tracking as part of the LRT...These changes add Precision Tracking to the existing fullscan LRT chains.
Changes:
- EDM Changes to support PT (adds 2 collections `HLT_IDTrack_FSLRT_IDTrig` and `HLT_IDTrack_FSLRT_IDTrigAux`)
- Adds Precision Tracking as part of the LRT tracking step
Jira: ATR-22502 ATR-22311https://gitlab.cern.ch/atlas/athena/-/merge_requests/43468Adding LLP tau trigger RNN chain2021-05-27T03:03:14+02:00Guillermo Nicolas HamityAdding LLP tau trigger RNN chainAdding displaced tau trigger chain discussed in ATR-21754 file options with placeholder configuration files
- configs for LLP tau RNN (network and flattening files) have been added to eos and updated in algorithm
- implementation of LLP...Adding displaced tau trigger chain discussed in ATR-21754 file options with placeholder configuration files
- configs for LLP tau RNN (network and flattening files) have been added to eos and updated in algorithm
- implementation of LLP RNN chain into menu (tau25, tau160, tau200 and ditau chains with tracktwoLLP sequence)
- chain will be used for sampleT studiesGuillermo Nicolas HamityGuillermo Nicolas Hamityhttps://gitlab.cern.ch/atlas/athena/-/merge_requests/43382Minor changes in Small R Jet algorithm in L1Calo Bitwise Offline software sim...2021-05-18T03:03:03+02:00Varsiha SothilingamMinor changes in Small R Jet algorithm in L1Calo Bitwise Offline software simulationMinor changes made to algorithm, fixing bugs in algorithm. Changes will be needed for L1Topo simulation.
Additional variables have been added to the small R, large R and tau EDM to return global coordiantes.Minor changes made to algorithm, fixing bugs in algorithm. Changes will be needed for L1Topo simulation.
Additional variables have been added to the small R, large R and tau EDM to return global coordiantes.https://gitlab.cern.ch/atlas/athena/-/merge_requests/43368Integrate nav conversion with transform2021-06-04T12:46:05+02:00Tomasz BoldIntegrate nav conversion with transformThis the first attempt to run Run2ToEun3NavConverter as a part of regular reco transfrom.
@wprzygod - there is only one minor change in the name of the output collection HLTNav_All insetead of HLTNav_all.
There is couple of questions t...This the first attempt to run Run2ToEun3NavConverter as a part of regular reco transfrom.
@wprzygod - there is only one minor change in the name of the output collection HLTNav_All insetead of HLTNav_all.
There is couple of questions to @tamartin @rbielski - is the place where it is configured now ok?
I tested it with q431 - the output AOD has JSON files in the metadata but not HLTNav objects. What have I missed?
More importantly, I do not think we want to enable this conversion yet by default. So another flag is needed?https://gitlab.cern.ch/atlas/athena/-/merge_requests/43343Sweeping !42183 from 21.3 to 21.9.
TrigT1CaloFexPerf: Adding Phase II ffex tr...2021-05-11T03:07:10+02:00Atlas NightlybuildSweeping !42183 from 21.3 to 21.9.
TrigT1CaloFexPerf: Adding Phase II ffex trigger EM algorithmTrigT1CaloFexPerf: Adding Phase II ffex trigger EM algorithm
See merge request atlas/athena!42183TrigT1CaloFexPerf: Adding Phase II ffex trigger EM algorithm
See merge request atlas/athena!42183https://gitlab.cern.ch/atlas/athena/-/merge_requests/43325Trigger: Fix xAOD::AuxSelection WARNING Selected dynamic Aux atribute not fou...2021-05-12T03:04:38+02:00Rafal Bielskirafal.bielski@cern.chTrigger: Fix xAOD::AuxSelection WARNING Selected dynamic Aux atribute not found in the registry (3rd attempt)Third attempt after !39270 and !43320 at fixing ATR-22421. This time instantiating all AuxElement Accessors for all Trigger EDM decorations at compile-time by listing them in a header file in TrigOutputHandling and including it in HLTEDM...Third attempt after !39270 and !43320 at fixing ATR-22421. This time instantiating all AuxElement Accessors for all Trigger EDM decorations at compile-time by listing them in a header file in TrigOutputHandling and including it in HLTEDMCreator.cxx.
The fixed warning is now marked as failure in check_log.py to prevent it from returning. Warning and info message count checks are also refined in two TriggerTest tests.
Performance effects:
* single-threaded compilation time of the TrigOutputHandling package: +2% (10s)
* TrigOutputHandlingLib library size: 5.8 MB -> 6.2 MB for .so, 97 MB -> 103 MB for .so.dbg
* no runtime overhead - running 3 tests before and 3 after the change, timing seemed to improve (could be due to some external factors)
----
:warning: **NOTE FOR RC:** This MR has a runtime conflict with any MR adding new decorations in the [TriggerEDMRun3 list](https://gitlab.cern.ch/atlas/athena/blob/master/Trigger/TriggerCommon/TrigEDMConfig/python/TriggerEDMRun3.py). Merging such changes and this MR together will result in broken CI despite the two MRs both having green CI on their own. I will rebase this MR after the one I'm aware of, !42764, but please let me know if I miss any others.https://gitlab.cern.ch/atlas/athena/-/merge_requests/43246TriggerJobOpts: refactor EDMVersion flag2021-05-10T03:05:40+02:00Frank WinklmeierTriggerJobOpts: refactor EDMVersion flagMove the code to discover the `EDMVersion` flag in bytestream files to
`TrigEDMConfig.Utils` and cache the result for a given bytestream file.
This ensures we are only reading the file once. This is especially
important for mixed old/new...Move the code to discover the `EDMVersion` flag in bytestream files to
`TrigEDMConfig.Utils` and cache the result for a given bytestream file.
This ensures we are only reading the file once. This is especially
important for mixed old/new job options where the flags are not being
locked and hence this function is called every time code reads the
`EDMVersion` flag.
cc @rbielskihttps://gitlab.cern.ch/atlas/athena/-/merge_requests/43050TEG Forward: Setup precise and fast forward into the e/g menu2021-05-04T03:03:03+02:00Joao Victor Da Fonseca PintoTEG Forward: Setup precise and fast forward into the e/g menuThis MR set the fast and precise steps (not used yet) into the menu. Just build the skeleton.
ping: @cjmeyer, @fernando , @dbakshig , @mschefer, @mverissi , @eegidiop
Ticket: ATR-18968This MR set the fast and precise steps (not used yet) into the menu. Just build the skeleton.
ping: @cjmeyer, @fernando , @dbakshig , @mschefer, @mverissi , @eegidiop
Ticket: ATR-18968Joao Victor Da Fonseca PintoJoao Victor Da Fonseca Pintohttps://gitlab.cern.ch/atlas/athena/-/merge_requests/43008Trigger EDM cleanup for Run 2 reprocessing2021-04-29T03:04:54+02:00Oleg KuprashTrigger EDM cleanup for Run 2 reprocessingRemove precision HLT tracking containers from the AODSLIM EDM set -> reduce event size for bulk Run-2 MC samples for physics analyses.
Tagging @bernius, @aranzazu, @sutt, @nagano for info.Remove precision HLT tracking containers from the AODSLIM EDM set -> reduce event size for bulk Run-2 MC samples for physics analyses.
Tagging @bernius, @aranzazu, @sutt, @nagano for info.https://gitlab.cern.ch/atlas/athena/-/merge_requests/42984Photons for TLA2021-05-15T03:04:03+02:00Marco MontellaPhotons for TLAThis is our first implementation of a TLA Sequence in the Egamma HLT slice that operates by making a deep copy of the HLT_FastPhoton container (xAOD::TrigPhotonContainer) produced by the FastPhoton reconstruction step, and saving the res...This is our first implementation of a TLA Sequence in the Egamma HLT slice that operates by making a deep copy of the HLT_FastPhoton container (xAOD::TrigPhotonContainer) produced by the FastPhoton reconstruction step, and saving the resulting container to the output.
The TLA Sequence is defined and configured in `Trigger/TriggerCommon/TriggerMenuMT/python/Egamma/TLAPhotonSequenceSetup.py` and is composed of:
- an algorithm which creates the deep copy of the input photons and pushes photons above a certain energy to the output container. [configured in `Trigger/TrigHypothesis/TrigEgammaHypo/python/TrigEgammaTLAPhotonFexMTConfig.py` and relevant source code in `Trigger/TrigHypothesis/TrigEgammaHypo/src/`]
- a HypoAlg and a HypoTool, which do nothing but to associate the TrigCompositeUtils::Decision* object set in the previous chain steps to the photons passing the selector implemented in the previous steps [source code in `Trigger/TrigHypothesis/TrigEgammaHypo/src/`]
At the current stage, the "TLA Step" is implemented as an additional chain step that is appended to the existing list of steps only for chains with the tag `PhotonDS` in their name. The addition is performed at line 110 and following of `Trigger/TriggerCommon/TriggerMenuMT/python/Egamma/PhotonChainConfiguration.py`.
We are currently testing our implementation through the new `HLT_PhotonDS_g35_loose_L1EM22VHI` chain, defined at line 211 of LS2_v1.py.
------ MAY 2nd EDIT ------
Following the comments on this thread and on [ATR-19317](https://its.cern.ch/jira/browse/ATR-19317), we altered the configuration of the sequence in TLAPhotonSequenceSetup.py to have the Selector (TrigEgammaTLAPhotonFex) run within the event view produced by the precision photon reconstruction step following the examples of both fastPhotonMenuSequence and precisionPhotonMenu sequence [https://acode-browser1.usatlas.bnl.gov/lxr/source/athena/Trigger/TriggerCommon/TriggerMenuMT/python/HLTMenuConfig/Egamma/PhotonMenuSequences.py#L76](https://acode-browser1.usatlas.bnl.gov/lxr/source/athena/Trigger/TriggerCommon/TriggerMenuMT/python/HLTMenuConfig/Egamma/PhotonMenuSequences.py#L76).
Currently, the test run on the PhotonDS chain fails, likely because the input collection again cannot be accessed. The full warning level log printout can be found in the attached file [log_warning_2.05.txt](/uploads/c29dc508f862fc145e0569652984965c/log_warning_2.05.txt)https://gitlab.cern.ch/atlas/athena/-/merge_requests/42917Implement the asymmetric quarter rings at Fast Calo step2021-04-27T03:02:42+02:00Joao Victor Da Fonseca PintoImplement the asymmetric quarter rings at Fast Calo stepThis MR is a part of Paulo Cesar's QT to implement the asymmetric quarter rings for calibration (and classification) future studies.
ping: @cjmeyer @fernando @dbakshig @mverissi @efurtado @eegidiopThis MR is a part of Paulo Cesar's QT to implement the asymmetric quarter rings for calibration (and classification) future studies.
ping: @cjmeyer @fernando @dbakshig @mverissi @efurtado @eegidiopPaulo Cesar Machado De Abreu FariasPaulo Cesar Machado De Abreu Fariashttps://gitlab.cern.ch/atlas/athena/-/merge_requests/42916Implement the asymmetric quarter rings at Fast Calo step2021-04-25T23:13:42+02:00Joao Victor Da Fonseca PintoImplement the asymmetric quarter rings at Fast Calo stepThis MR is a part of Paulo Cesar's QT to implement the asymmetric quarter rings for calibration (and classification) future studies.This MR is a part of Paulo Cesar's QT to implement the asymmetric quarter rings for calibration (and classification) future studies.Paulo Cesar Machado De Abreu FariasPaulo Cesar Machado De Abreu Fariashttps://gitlab.cern.ch/atlas/athena/-/merge_requests/42899Implement tools decorating L1 xAOD RoI objects with threshold patterns2021-05-03T17:08:24+02:00Rafal Bielskirafal.bielski@cern.chImplement tools decorating L1 xAOD RoI objects with threshold patternsImplement tools which decorate L1 xAOD RoI objects with threshold patterns, as discussed in ATR-23290. The decoration is a **64-bit** mask where each bit corresponds to L1 trigger threshold number of a given type and the value indicates ...Implement tools which decorate L1 xAOD RoI objects with threshold patterns, as discussed in ATR-23290. The decoration is a **64-bit** mask where each bit corresponds to L1 trigger threshold number of a given type and the value indicates whether the threshold passed or not for the given RoI object (TOB).
There is an interface class `IRoIThresholdsTool` to be used by clients to call the `decorateThresholds` method. There is also a base class `RoIThresholdsTool` taking care of the menu information retrieval, input and output data handles, and the act of adding the decoration. The concrete implementations only need to provide the `getPatterns` method which translate the information from the menu and the RoI object (TOB) into the 64-bit mask.
For `xAOD::eFexEMRoI` a new tool is created - `eFexEMRoIThresholdsTool`.
For `xAOD::MuonRoI` (MUCTPI), the existing `TrigThresholdDecisionTool` is adapted to derive from the new base class and follow the new interfaces and logic. cc @armbrust
The clients, eFex EM and MU "RoI unpacking tools" in L1Decoder, are adapted to depend on and read the new decoration for seeding the HLT chains.
Left to do in a separate MR: Implement the `getPattern` logic for `xAOD::eFexEMRoI` (`eFexEMRoIThresholdsTool`). For now it returns an all-pass mask.
NB: As there was no opposition in Jira, I opted for a 64-bit mask instead of 32-bit to be on the safe side regarding the possible number of thresholds.https://gitlab.cern.ch/atlas/athena/-/merge_requests/42801Add muon LRT physics chains with EF step2021-04-23T03:04:00+02:00John BainesAdd muon LRT physics chains with EF stepAdd 3 new Muon LRT trigger chains including a EF step:
HLT_mu24_LRT_d0loose_L1MU20
HLT_mu24_LRT_d0medium_L1MU20
HLT_mu24_LRT_d0tight_L1MU20
ATR-23252
tested with runTrigART.pyAdd 3 new Muon LRT trigger chains including a EF step:
HLT_mu24_LRT_d0loose_L1MU20
HLT_mu24_LRT_d0medium_L1MU20
HLT_mu24_LRT_d0tight_L1MU20
ATR-23252
tested with runTrigART.pyhttps://gitlab.cern.ch/atlas/athena/-/merge_requests/42764implement UE subtraction for e/g triggers for heavy ion2021-05-13T03:03:18+02:00Ran Biimplement UE subtraction for e/g triggers for heavy ionimplement UE subtraction in the precision calo sequence for e/g triggers. an `xAOD::HIEventShapeContainer` collection is produced from FS calo info, which is then used to correct the calo cells in RoIs. the correction is done by a new al...implement UE subtraction in the precision calo sequence for e/g triggers. an `xAOD::HIEventShapeContainer` collection is produced from FS calo info, which is then used to correct the calo cells in RoIs. the correction is done by a new algorithm, `HLTCaloCellCorrector`, based on HICaloCellCorectionTool [[1]] which was removed in rel 22.0.
configure triggers to use the sequences with UE-subtraction when the 'ion' tag is present in the name.
add the event shape collection to trigger edm.
[[1]]: https://acode-browser1.usatlas.bnl.gov/lxr/source/athena/Trigger/TrigHypothesis/TrigHIHypo/src/HICaloCellCorectionTool.cxx?v=21.0https://gitlab.cern.ch/atlas/athena/-/merge_requests/42692Build the skeleton of the trigger electron forward sequence2021-04-24T03:03:05+02:00Joao Victor Da Fonseca PintoBuild the skeleton of the trigger electron forward sequenceThis MR provide some ways to work with the forward electron sequence. We built only the fast step.
Not in operation for now but we are pushing this first piece now to avoid large MRs.This MR provide some ways to work with the forward electron sequence. We built only the fast step.
Not in operation for now but we are pushing this first piece now to avoid large MRs.Joao Victor Da Fonseca PintoJoao Victor Da Fonseca Pintohttps://gitlab.cern.ch/atlas/athena/-/merge_requests/42670EDMVersion flag: change default to -1, remove from TriggerFlags2021-04-20T03:02:54+02:00Rafal Bielskirafal.bielski@cern.chEDMVersion flag: change default to -1, remove from TriggerFlagsThe following changes are made:
* Remove `TriggerFlags.EDMVersion` which has been deprecated for a few months. All clients are already using `ConfigFlags.Trigger.EDMVersion`.
* Change default value of `ConfigFlags.Trigger.EDMVersion` fro...The following changes are made:
* Remove `TriggerFlags.EDMVersion` which has been deprecated for a few months. All clients are already using `ConfigFlags.Trigger.EDMVersion`.
* Change default value of `ConfigFlags.Trigger.EDMVersion` from 3 to intentionally invalid value -1 as discussed in ATR-22856.
* Adapt a few places in TriggerJobOpts and TrigEDMConfig which would treat -1 same as 1 or 2 (e.g. if they checked `EDMVersion <= 2`) to check for 1 or 2 exactly, and throw an exception if unhandled value is provided. The few "EDMVersion < 3" cases in TrigMonitoring are not modified, as the logic still holds for them and they are not crucial/central enough to implement extensive sanity checks on central flags.
Closes ATR-22856https://gitlab.cern.ch/atlas/athena/-/merge_requests/42636This MR to introduce LRT electron chains branched from https://gitlab.cern.ch...2021-04-20T03:03:41+02:00Debottam Bakshi GuptaThis MR to introduce LRT electron chains branched from https://gitlab.cern.ch/atlas/athena/-/merge_requests/42368Mentioned in ATR-23090 this MR will implement an electron chain that runs the fast track finder with a large radius tracking configuration (FTF_electronLRT) and then adds an additional d0 cut to the hypo (branched from https://gitlab.cer...Mentioned in ATR-23090 this MR will implement an electron chain that runs the fast track finder with a large radius tracking configuration (FTF_electronLRT) and then adds an additional d0 cut to the hypo (branched from https://gitlab.cern.ch/atlas/athena/-/merge_requests/42368).https://gitlab.cern.ch/atlas/athena/-/merge_requests/42612Master add EF steps to muon LRT performance chain2021-04-21T03:02:36+02:00John BainesMaster add EF steps to muon LRT performance chainAdds a chain HLT_mu6_LRT_idperf_L1MU6 to the LS2_v1 menu. This chain differs from HLT_mu6_LRT_idperf_l2lrt_L1MU6 in that it adds the EF steps of the selection i.e. EFMUSA and a LRT configuration of InDet Precision Tracking.
ATR-23252Adds a chain HLT_mu6_LRT_idperf_L1MU6 to the LS2_v1 menu. This chain differs from HLT_mu6_LRT_idperf_l2lrt_L1MU6 in that it adds the EF steps of the selection i.e. EFMUSA and a LRT configuration of InDet Precision Tracking.
ATR-23252