Adding L1REPACK:FullMC workflow to run on RAW of MC. #14635
Conversation
…ator in REPACK:Full, and clean REPACK:uGT
|
A new Pull Request was created by @rekovic for CMSSW_8_1_X. It involves the following packages: Configuration/Applications @cmsbuild, @franzoni, @davidlange6 can you please review it and eventually sign? Thanks. cms-bot commands are list here #13028 |
|
Hi We need to continue with the db geometry. What is missing? On 25 May 2016, at 16:40, rekovic <[email protected]mailto:[email protected]> wrote: Adding L1REPACK:FullMC workflow to run on RAW of MC. Available L1REPACK workflows of re-emulation of Stage2 L1T available now:
L1REPACK:FullMC workflow starts from unpacking RPC, DT, CSC, ECAL, and HCAL. In a MC-production-like workflow, use it as: cmsDriver.py reHLT --conditions auto:run2_mc -s L1REPACK:FullMC,HLT:@Fake --datatier GEN-SIM-RAW,AODSIM -n 10 --era=Run2_2016 --eventcontent RAWSIM,AODSIM --filein "root://eoscms.cern.ch//eos/cms/store/mc/RunIISpring16DR80/TTJets_TuneCUETP8M1_13TeV-amcatnloFXFX-pythia8/RAWAODSIM/PUSpring16RAWAODSIM_80X_mcRun2_asymptotic_2016_v3-v2/30000/040AFBF1-8E0B-E611-9C22-001E67E95A58.root" --processName HLT2 --python_filename reHLT.py Add a flag to see printouts of L1 menu --customise=HLTrigger/Configuration/CustomConfigs.L1T L1REPACK:Full workflow starts from unpacked TriggerPrimitives and re-emulates muon track finders, uGMT, Calo Layer1 and Layer2, and uGT at the end of the sequence. It re-packs re-emulated uGMT, CaloLayer2 and uGT in RAW data collection. In a typical HLT test use it as: cmsDriver.py RelVal --step=L1REPACK:Full --conditions=auto:run2_hlt_GRun --filein=/store/data/Run2016A/ZeroBias1/RAW/v1/000/271/336/00000/00963A5A-BF0A-E611-A657-02163E0141FB.root --custom_conditions= --fileout=RelVal_L1RePack_GRun_DATA.root --number=10 --data --datatier GEN-SIM-DIGI-RAW --eventcontent=RAW --customise=HLTrigger/Configuration/CustomConfigs.L1T --era=Run2_2016 --customise= --scenario=pp --python_filename=RelVal_L1RePack_Full_GRun_DATA.py --customise=L1Trigger/Configuration/L1Trigger_custom.customiseResetPrescalesAndMasks L1REPACK:uGT workflow starts from unpacked uGMT and CaloLayer2 outputs, re-emulates uGT, and packs the re-emulated uGT in RAW data collection. In a typcal HLT test us it as: cmsDriver.py RelVal --step=L1REPACK:uGT --conditions=auto:run2_hlt_GRun --filein=/store/data/Run2016A/ZeroBias1/RAW/v1/000/271/336/00000/00963A5A-BF0A-E611-A657-02163E0141FB.root --custom_conditions= --fileout=RelVal_L1RePack_GRun_DATA.root --number=10 --data --datatier GEN-SIM-DIGI-RAW --eventcontent=RAW --customise=HLTrigger/Configuration/CustomConfigs.L1T --era=Run2_2016 --customise= --scenario=pp --python_filename=RelVal_L1RePack_uGT_GRun_DATA.py --customise=L1Trigger/Configuration/L1Trigger_custom.customiseResetPrescalesAndMasks For L1REPACK:Full2015Data workflow is intended for 205 data, starts from unpacked TriggerPrimitives and re-emulates muon track finders, uGMT, Calo Layer1 and Layer2, and uGT at the end of the sequence. It re-packs re-emulated uGMT, CaloLayer2 and uGT in RAW data collection. Configuration for the HFTPs (1x1) which is different in 2016 wrt 2015 has already been included in the above workflow via setting ESProducer preference. However, the geometry for 2016 needs to be loaded via a cmsDriver flag --geometry flag. cmsDriver.py RelVal --step=L1REPACK:Full2015Data --conditions=auto:run2_hlt_GRun --filein=/store/data/Run2015D/MuonEG/RAW/v1/000/256/677/00000/80950A90-745D-E511-92FD-02163E011C5D.root --custom_conditions= --fileout=RelVal_L1RePack_GRun_DATA.root --number=10 --data --datatier GEN-SIM-DIGI-RAW --eventcontent=RAW --customise=HLTrigger/Configuration/CustomConfigs.L1T --era=Run2_2016 --customise= --scenario=pp --python_filename=RelVal_L1RePackFull2015Data_GRun_DATA.py --customise=L1Trigger/Configuration/L1Trigger_custom.customiseResetPrescalesAndMasks --geometry=Extended2016,Extended2016Reco You can view, comment on, or merge this pull request online at: Commit Summary
File Changes
Patch Links:
— |
|
Hmm, why is MC different from data for 2016? |
|
@Martin-Grunewald In workflow "FullMC" we are re-emulating TPs and then the L1T. |
|
@davidlange6 We need to make sure that the geometry used for HCAL TPs, loaded by --geometry=Extended2016,Extended2016Reco, is used. |
|
it should already be there in the GT for MC. Is it not?
|
|
Pull request #14635 was updated. @cmsbuild, @franzoni, @davidlange6 can you please check and sign again. |
|
please test |
|
The tests are being triggered in jenkins. |
|
Explicitly confirmed that our customizations |
|
Comparison is ready @slava77 comparisons for the following workflows were not done due to missing matrix map:
|
Adding L1REPACK:FullMC workflow to run on RAW of MC.
(Also fixed Csc InputTag for Omtf emuator in REPACK:Full, and clean REPACK:uGT.
Available L1REPACK workflows of re-emulation of Stage2 L1T available now:
L1REPACK:FullMC workflow starts from unpacking RPC, DT, CSC, ECAL, and HCAL.
It uses unpacked TriggerPrimitives for ECAL and simulates TriggerPrimitives DT TPs, CSC TPs, and HCAL TPs. It then emulates TwinMux, muon track finders (BMTF, OMTF, EMTF) uGMT, Calo Layer1, Calo Layer2, and uGT at the end of the sequence.
It re-packs re-emulated uGMT, CaloLayer2 and uGT in RAW data collection.
In a MC-production-like workflow, use it as:
Add a flag to see printouts of L1 menu
--customise=HLTrigger/Configuration/CustomConfigs.L1TL1REPACK:Full workflow starts from unpacked TriggerPrimitives and re-emulates muon track finders, uGMT, Calo Layer1 and Layer2, and uGT at the end of the sequence. It re-packs re-emulated uGMT, CaloLayer2 and uGT in RAW data collection. In a typical HLT test use it as:
L1REPACK:uGT workflow starts from unpacked uGMT and CaloLayer2 outputs, re-emulates uGT, and packs the re-emulated uGT in RAW data collection. In a typcal HLT test us it as:
For L1REPACK:Full2015Data workflow is intended for 205 data, starts from unpacked TriggerPrimitives and re-emulates muon track finders, uGMT, Calo Layer1 and Layer2, and uGT at the end of the sequence. It re-packs re-emulated uGMT, CaloLayer2 and uGT in RAW data collection.
Configuration for the HFTPs (1x1) which is different in 2016 wrt 2015 has already been included in the above workflow via setting ESProducer preference. However, the geometry for 2016 needs to be loaded via a cmsDriver flag --geometry flag.
For a typical HLT test use it as: