- Jan 30, 2024
-
-
Updates for DAOD_MUON1
-
- Jan 26, 2024
-
-
Move list of IDTIDE excluded auxdata in IDTIDE config
-
- Jan 25, 2024
-
-
DerivationFrameworkMuons - Fix Muon5
-
Fixing recursion issue in TruthDecayCollectionMaker Under certain circumstances, some decays will be created with loops in the particle records in the decays of BSM particles. It's a bit unusual, but when that happens there can be an infinite recursion induced in the truth decay collection maker, because the unique ID (previously: barcode) isn't set until after the decay vertex and outgoing particles are added to the record. That means if for some reason we have a vertex that loops back on itself, the particle could be added recursively and infinitely... until we get a stack overflow and crash (thanks to Serhan for the tip there). Reported by Laura Bruce. Should be fixed here.
-
- Jan 23, 2024
-
-
Introduced ResidualPullCalculator CA config
-
FTAG: Python config cleanup
-
- Jan 16, 2024
-
-
Add AntiKt4UFOCSSKJets to FTAG1
-
- Jan 10, 2024
-
-
adding in to physlite an extra container with sihit electrons
-
Thomas Strebler authored
-
- Jan 09, 2024
-
-
JetTagTools clean up
-
remove some unused trigger collections from EGAM derivations
-
Kin Yip Fung authored
-
- Jan 05, 2024
-
-
Kin Yip Fung authored
-
Adam Edward Barton authored -
-
- Dec 22, 2023
-
-
Fix METRemappingAlg for PHYS->PHYSLITE production
-
Adapting IDTRKVALID CA config for ITk (ATLIDTRKCP-515)
-
- Dec 21, 2023
-
-
Default blocks are added by default. PRW parameters are obtained from ConfigAccumulator. Add autoconfigFlags as an attribute to the ConfigAccumulator so information can be obtained in other blocks. Update the PileupReweighting block so that campaign, input files, and useDefaultConfig can be obtained from autoconfigFlags. The wrapper function that handeled this before has been removed. addDefaultAlgs function is now part of the ConfigFactory class and is called automatically when initializing an object instance. Default parameters are now obtained from ConfigAccumulator except for noSystematics flag so that this can be changed from the command line if not specified directly.
-
- Dec 20, 2023
-
-
adding DSID 411316 in HadronOrigin and HF Classifier
-
- Dec 19, 2023
-
-
DerivationFrameworkMCTruth: keep all daughters of TruthTausWithDecayParticles Hello, This MR changes the number of generations of daughter particles retained when building the TruthTausWithDecayParticles/TruthTausWithDecayVertices collections. It was previously set to 1, but this can lead to pathological situations, e.g. when the truth tau decays to kaons: if we don't know what the kaon decays to, we cannot properly compute the truth visible tau momentum. When set to -1, we retain all generations, and this seems to fix the problem. A priori, we can now recompute the various decorations of TruthTaus using TruthTausWithDecayParticles/TruthTausWithDecayVertices, such as `pt/eta/phi/m_vis`, the number of neutral/charged pions, ... We will need to rework a bit BuildTruthTaus to do this. The only variables we don't seem to get is `polarizationPhi` and `polarizationTheta` decorations, they exist for TruthTaus but not TruthTausWithDecayParticles, however these seem to always be 0. We still need to check we can rerun photon dressing for TruthTausWithDecayParticles. Marking as urgent as we'd need this for the upcoming DAOD_PHYS production that is imminent. Cheers, Bertrand
-
added back in call to CloseBy correction which was accidentally removed in MR 67455
-
- Dec 18, 2023
-
-
Fixed bug in JETM7 when running over MC
-
- Dec 15, 2023
-
-
Adding DAOD_NCB1 for cosmic ray studies in NCB group As in the title, this MR is to add the new format to run on AOD from cosmic ray runs. This is needed for the analysis we are starting in the NCB group.
-
* Fix trigger skimming for Run 3 triggers in LLP1 and Trig8. Mostly, to get legacy L1 electron triggers into the list for 2022. Not based on prescales while TriggerAPI is being updated. Thus includes more than needed. * Remove Run 2 trigger style matching from Trig8 for Run 3 * Add MuonSegments_LRT container to LLP1. Needed in future for using reprocessed Run 2 data--TBD replacing this with a muonsegment merger.
-
add xAOD::JetTrigAuxContainer to SlimmingHelper (AFT-658)
-
- Dec 14, 2023
-
-
fixing MUON5
-
- Dec 13, 2023
-
-
FTAG: Remove broken flip strategies for GN2 & Code Cleanup
-
Remove barcodes from the print statements and comments
-
- Dec 12, 2023
-
-
Use the MC::isHadron function instead of local functions
-
First attempt of the migration to CA (ATLASRECTS-7845)
-
- Dec 11, 2023
-
-
Replace checks on hard-coded PDG code values with MC:: functions
-
Remove the mentions of barcodes from the documentation/comments in physicsAnalysis
-
Use appropriate ITk TSOS Cfg name for IDTIDE
-
Make VarHolder scope local This clears most of the "variable not decorated" warnings. The VarHolder object in decorateElec() was originally locally scoped in r21. At some point during the port, it became a class member. This prevented the variables from being decorated correctly to each electron. Transitioning back to a locally-scoped variable resolves the problem.
-
- Dec 08, 2023
-
-
Trig8 jet chain and RoI list
-
Optimise JETM7 size
-
Adding small-R UFO jets with pT > 7 GeV cut to JETM3/JETM4 (incl. smart slimming list)
-
Vakhtang Tsulaia authoredRemoved unused local variable
-
- Dec 07, 2023
-
-
Cleaning of HadronOriginClassifier code. During the implementaiton of the HF classification, the code of HadronOriginClassifier was modified to find a bug in it. Most of this modifications were comments and prints to find the bug and they have been removed now that is fixed.
-
Convert r21 decorators to const Accessor for vertex fits Previously, secondary vertex decoration was done using a syntax like (*decorator)(*vertex) = value; This has been changed in two places to use a const Accessor, as this makes the logic of the decoration more obvious and might avoid some subtle bugs with decoration. This was an attempt to resolve a warning WARNING saveSecondaryVertices - missing "SecondaryVertexIndex" variable but this change does not seem to fix the problem.
-
