Merge updates to the jet energy corrections code

JESCorrections/readme.md

@@ -6,21 +6,43 @@ for the jets used in the High-Level Trigger (HLT).
 
 Instructions to set up the CMSSW area to produce JRA NTuples, and derive JESCs:
 ```
-scram project CMSSW_12_0_0_pre4
-cd CMSSW_12_0_0_pre4/src
-eval `scram runtime -sh`
-git cms-merge-topic missirol:devel_hltRun3TRK_1200pre4
+cmsrel CMSSW_12_3_0_pre4
+cd CMSSW_12_3_0_pre4/src
+cmsenv
+git clone [email protected]:sparedes/JMETriggerAnalysis.git -o sparedes -b run3_12_3_X
+git clone [email protected]:sparedes/JetMETAnalysis.git -o sparedes -b dev_run3
 
-git clone https://github.com/missirol/JetMETAnalysis.git -o missirol -b devel_hlt2
-git clone https://github.com/missirol/JMETriggerAnalysis.git -o missirol -b run3
-scram b -j 8
+# PFHC: preliminary HLT-PFHC for Run-3
+cp /afs/cern.ch/work/p/pdas/public/run3/PFHC_Run3Winter21_HLT_V3.db JMETriggerAnalysis/NTuplizers/test/
+
+scram b -j 12
 ```
 
 #### Produce JRA NTuples with HLT Jets
+First test your setup locally: 
+```bash
+# First make your way to the 'test' directory
+cd JMETriggerAnalysis/JESCorrections/test/
+cmsRun jescJRA_cfg.py  maxEvents=1000  output=test_jra_step.root 
+```
+Note: You might want to double check that what is set in the `jescJRA_cfg.py`, such as the `.root` file matches the sample you want to use, or that the `.db` file of the PFHCs is up-to-date.  
 
- * How to run the JRA step locally and with crab
- * Which samples to use, how to find them in DAS
- * What to do if a sample is not available at a Tier-2 (transfer request, Rucio rule)
+If the test runs sucessfuly, you can then inspect the output root file `test_jra_step.root`  and make sure the contents are as expected.
+
+Once all looks good you can submit your crab jobs. 
+```bash
+crab submit -c crab/my_crab_submission_file.py
+```
+Note: You might need to change the line where the `.db` file of the PFHCs is picked up so that CRAB can find the file. 
+
+Once the crab jobs are finished, the output `.root` files can be found in the Tier2 (T2) specified in the configuration file,
+and then transferred over to `eos` for the next steps.
+
+**Note** that in order to run the ntuple-making process on CRAB, there needs to be a copy of them on **a** T2. You can request to transfer them to your local T2 by requesting it through `rucio` (find more info in the [CMS Rucio Twiki](https://twiki.cern.ch/twiki/bin/viewauth/CMS/Rucio)):
+```bash
+rucio add-rule cms:/QCD_Pt-15to7000_TuneCP5_Flat_14TeV_pythia8/Run3Winter20DRMiniAOD-FlatPU0to80_110X_mcRun3_2021_realistic_v6_ext1-v1/GEN-SIM-RAW  1 T2_BE_IIHE --asynchronous  --ask-approval --lifetime 5184000
+```
+replacing the necessary dataset name, T2 name, lifetime, etc. 
 
 #### Derive Jet Energy Scale Corrections from JRA NTuples
 
@@ -37,8 +59,48 @@ The `fitJESCs` script is an example of
 a wrapper executing the various steps of the JESCs derivation
 (caveat: the script presently includes several hard-coded parameters).
 
-List of Run-3 HLT JESCs:
+##### Run on HTCondor
+The `-b` and `-j` flags can be used to run one HTCondor job for each jet collection's correction (see the submission script `test/sub_jecs.htc` for an example). 
+
+##### Sample information
+The necessary samples are large QCD simulated datasets with a flat p<sub>T</sub> over a large range.
+The `L1` correction needs to match jets from samples generated with and without pileup, so we will need to produce two JRANtuples. 
+The samples used for the latest run 3 preliminary JECs are the following:
+```
+# FlatPU
+/QCD_Pt-15to7000_TuneCP5_Flat_14TeV_pythia8/Run3Winter20DRMiniAOD-FlatPU0to80_110X_mcRun3_2021_realistic_v6_ext1-v1/GEN-SIM-RAW
+# NoPU
+QCD_Pt-15to7000_TuneCP5_Flat_14TeV_pythia8/Run3Winter20DRMiniAOD-NoPU_110X_mcRun3_2021_realistic_v6_ext1-v1/GEN-SIM-RAW
+```
+You can find them in DAS searching for:
+```
+dataset=/QCD_Pt-15to7000_TuneCP5_Flat_14TeV_pythia8/Run3Winter20DRMiniAOD-*0to80_110X_mcRun3_2021_realistic_v6_ext1-v1/GEN-SIM-RAW
+```
+You can also find similar samples doing something like:
+```
+dataset=/QCD*_Pt*/*/GEN-SIM*RAW
+```
+
+##### List of Run-3 HLT JESCs:
+
+  * `Run3Winter21_V2_MC`:
+
+    - Description: preliminary HLT JESCs for Run-3 studies
+
+    - Tag of `sparedes/JMETriggerAnalysis`: `HLT_JESCs_Run3Winter21_V2`
+
+    - Tag of `sparedes/JetMETAnalysis`: `Run3Winter21_V2`
+
+    - JRA NTuples:
+      ```
+      root://cms-xrd-global.cern.ch//store/group/phys_jetmet/saparede/runIII_hlt_jec/jra_ntuples/jescs_dec_PFHC_E2to500_noPU.root
+      root://cms-xrd-global.cern.ch//store/group/phys_jetmet/saparede/runIII_hlt_jec/jra_ntuples/jecs_dec_PFHC_E2to500_flatPU.root      
+      ```
+
+    - executable for JESCs fits (contains settings of all JESCs fits):
+      [`JESCorrections/test/fitJESCs`](https://github.com/sparedes/JMETriggerAnalysis/blob/HLT_JESCs_Run3Winter21_V2/JESCorrections/test/fitJESCs)
 
+
   * `Run3Winter20_V2_MC`:
 
     - Description: preliminary HLT JESCs for Run-3 studies
@@ -61,3 +123,5 @@ List of Run-3 HLT JESCs:
       - JRA NTuples affected by a bug in the `rho` value saved for Calo and PFCluster (AK4 and AK8) HLT jets:
         the `rho` based on PF-candidates was erronously used,
         instead of the `rho` values calculated from calo-towers and PFClusters, respectively.
+
+

JESCorrections/test/crab/sub_Run3Winter21_FlatPU.py

@@ -0,0 +1,38 @@
+from CRABClient.UserUtilities import config
+
+sample_name = 'flatPU_Run3Winter21_E2to500_take7'
+
+RAW_DSET = '/QCD_Pt15to7000_TuneCP5_14TeV-pythia8/Run3Winter21DRMiniAOD-FlatPU0to80FEVT_castor_112X_mcRun3_2021_realistic_v16-v2/GEN-SIM-DIGI-RAW'
+
+config = config()
+
+config.section_('General')
+config.General.requestName = 'jmeTriggerNTuple_'+sample_name
+config.General.transferOutputs = True
+config.General.transferLogs = False
+
+config.section_('JobType')
+config.JobType.pluginName  = 'Analysis'
+config.JobType.psetName = 'jescJRA_cfg.py'
+config.JobType.inputFiles = []
+config.JobType.pyCfgParams = ['output='+sample_name+'.root']
+config.JobType.maxMemoryMB = 2500
+config.JobType.inputFiles = ['PFHC_Run3Winter21_E2to500.db']
+config.JobType.allowUndistributedCMSSW = True
+
+config.section_('Data')
+config.Data.publication = False
+config.Data.ignoreLocality = False
+config.Data.splitting = 'Automatic'
+config.Data.inputDataset = RAW_DSET
+config.Data.outLFNDirBase = '/store/user/saparede/hlt_runIII_jesc_oct/crab_out/'+sample_name
+config.Data.unitsPerJob = 2700
+config.Data.totalUnits = -1
+
+config.section_('Site')
+config.Site.storageSite = 'T2_BE_IIHE'
+if config.Data.ignoreLocality:
+   config.Site.whitelist = ['T2_CH_CERN', 'T2_DE_*']
+
+config.section_('User')
+

JESCorrections/test/crab/sub_Run3Winter21_NoPU.py

@@ -0,0 +1,38 @@
+from CRABClient.UserUtilities import config
+
+sample_name = 'noPU_Run3Winter21_E2to500_take7'
+
+RAW_DSET = '/QCD_Pt15to7000_TuneCP5_14TeV-pythia8/Run3Winter21DRMiniAOD-NoPUFEVT_castor_112X_mcRun3_2021_realistic_v16-v2/GEN-SIM-DIGI-RAW'
+
+config = config()
+
+config.section_('General')
+config.General.requestName = 'jmeTriggerNTuple_'+sample_name
+config.General.transferOutputs = True
+config.General.transferLogs = False
+
+config.section_('JobType')
+config.JobType.pluginName  = 'Analysis'
+config.JobType.psetName = 'jescJRA_cfg.py'
+config.JobType.inputFiles = []
+config.JobType.pyCfgParams = ['output='+sample_name+'.root']
+config.JobType.maxMemoryMB = 2500
+config.JobType.inputFiles = ['PFHC_Run3Winter21_E2to500.db']
+config.JobType.allowUndistributedCMSSW = True
+
+config.section_('Data')
+config.Data.publication = False
+config.Data.ignoreLocality = False
+config.Data.splitting = 'Automatic'
+config.Data.inputDataset = RAW_DSET
+config.Data.outLFNDirBase = '/store/user/saparede/hlt_runIII_jesc_oct/crab_out/'+sample_name
+config.Data.unitsPerJob = 2700
+config.Data.totalUnits = -1
+
+config.section_('Site')
+config.Site.storageSite = 'T2_BE_IIHE'
+if config.Data.ignoreLocality:
+   config.Site.whitelist = ['T2_CH_CERN', 'T2_DE_*']
+
+config.section_('User')
+

JESCorrections/test/fitJESCs

@@ -6,24 +6,30 @@ set -euo pipefail
 ## -------------------------
 JESC_BASE=${CMSSW_BASE}/src/JMETriggerAnalysis/JESCorrections/test
 
-JRANTuple_NoPU=root://cms-xrd-global.cern.ch//store/group/phys_jetmet/saparede/runIII_hlt_jec/jra_ntuples/npv_fix_noPU/npvFix_noPU.root
-JRANTuple_PU=root://cms-xrd-global.cern.ch//store/group/phys_jetmet/saparede/runIII_hlt_jec/jra_ntuples/npv_fix_flatPU/npvFix_flatPU.root
+JRANTuple_NoPU=root://cms-xrd-global.cern.ch//store/group/phys_jetmet/saparede/runIII_hlt_jec/jra_ntuples/jescs_dec_PFHC_E2to500_noPU.root
+JRANTuple_PU=root://cms-xrd-global.cern.ch//store/group/phys_jetmet/saparede/runIII_hlt_jec/jra_ntuples/jecs_dec_PFHC_E2to500_flatPU.root
 
 ERA=Run3Winter20_V2_MC
 
+#JET_TYPES=(
+#  ak4caloHLT
+#  ak4pfclusterHLT
+#  ak4pfHLT
+#  ak4pfchsHLT
+#  ak4pfpuppiHLT
+#  ak8caloHLT
+#  ak8pfclusterHLT
+#  ak8pfHLT
+#  ak8pfchsHLT
+#  ak8pfpuppiHLT
+#)
 JET_TYPES=(
-  ak4caloHLT
-  ak4pfclusterHLT
-  ak4pfHLT
-  ak4chsHLT
-  ak4puppiHLT
-  ak8caloHLT
   ak8pfclusterHLT
   ak8pfHLT
-  ak8chsHLT
-  ak8puppiHLT
+  ak8pfchsHLT
+  ak8pfpuppiHLT
 )
-## -------------------------
+### -------------------------
 
 ## -------------------------
 ## Main
@@ -46,6 +52,8 @@ Options:
   -h, --help          Show this help message
   -o, --output-dir    Path to output directory          [Default: .]
   -n, --maxEvents     Max number of events to process   [Default: -1]
+  -b, --batch-mode    Run on batch mode (see htc scritp)[Default: true]
+  -j, --jet-type      Jet name for batch (e.g. ak4pfpuppiHLT)   
   -p, --parallelise   Parallelise by jet algorithm      [Default: false]
   --skip-L1-PFPuppi   Skip L1FastJet for PFPuppi jets   [Default: false]
   -d, --dry-run       Enable dry-run mode               [Default: false]
@@ -55,19 +63,30 @@ Options:
 maxEvents=-1
 paralleliseByAlgo=false
 skipL1PFPuppi=false
+doBatch=false
+batchJetType="dummyJetType"
 
 while [[ $# -gt 0 ]]; do
   case "$1" in
-    -h|--help) showHelpMsg=true; shift;;
+    -h|-khelp) showHelpMsg=true; shift;;
+    -j|--jet-type) batchJetType=$2; shift; shift;;
     -o|--output-dir) outputDir=$2; shift; shift;;
     -n|--maxEvents) maxEvents=$2; shift; shift;;
     -p|--parallelise) paralleliseByAlgo=true; shift;;
     --skip-L1-PFPuppi) skipL1PFPuppi=true; shift;;
     -d|--dry-run) dryRun=true; shift;;
+    -b|--batch-mode) doBatch=true; shift;;
+    -j|--jet-type) batchJetType=$2; shift; shift;;
     *) printf "%s\n" ">> ERROR -- unrecognised command-line argument: ${1}"; exit 1; shift;;
   esac
 done
 
+if [ "X$doBatch" = "Xtrue" ]
+then
+  echo "running on batch mode for jet-type $batchJetType"
+  JET_TYPES=( $batchJetType )
+fi
+
 # print help message
 if [ "X${showHelpMsg}" = "Xtrue" ]; then
   usage
@@ -100,6 +119,7 @@ runJECWorflowForOneAlgo(){
   cd ${OUTDIR}
 
   stepN=0
+  stepName=""
 
   dRgen=0.2
   if [[ ${jetType} == ak8* ]]; then dRgen=0.4; fi

JESCorrections/test/jescInspector_cfg.py

@@ -81,30 +81,30 @@
     if type(_mod) == cms.OutputModule:
        process.__delattr__(_modname)
        if opts.verbosity > 0:
-          print '> removed cms.OutputModule:', _modname
+          print('> removed cms.OutputModule:', _modname)
 
 # remove cms.EndPath objects from HLT config-dump
 for _modname in process.endpaths_():
     _mod = getattr(process, _modname)
     if type(_mod) == cms.EndPath:
        process.__delattr__(_modname)
        if opts.verbosity > 0:
-          print '> removed cms.EndPath:', _modname
+          print('> removed cms.EndPath:', _modname)
 
 # remove selected cms.Path objects from HLT config-dump
-print '-'*108
-print '{:<99} | {:<4} |'.format('cms.Path', 'keep')
-print '-'*108
+print('-'*108)
+print('{:<99} | {:<4} |'.format('cms.Path', 'keep'))
+print('-'*108)
 for _modname in sorted(process.paths_()):
     _keepPath = _modname.startswith('MC_') and ('Jets' in _modname or 'MET' in _modname or 'AK8Calo' in _modname)
     if _keepPath:
-      print '{:<99} | {:<4} |'.format(_modname, '+')
+      print('{:<99} | {:<4} |'.format(_modname, '+'))
       continue
     _mod = getattr(process, _modname)
     if type(_mod) == cms.Path:
       process.__delattr__(_modname)
-      print '{:<99} | {:<4} |'.format(_modname, '')
-print '-'*108
+      print('{:<99} | {:<4} |'.format(_modname, ''))
+print('-'*108)
 
 # remove FastTimerService
 if hasattr(process, 'FastTimerService'):
@@ -119,7 +119,7 @@
 ###
 
 ## customised JME collections
-from JMETriggerAnalysis.Common.customise_hlt import *
+from JMETriggerAnalysis.Common.customise_hlt import addPaths_MC_JMEPFCluster, addPaths_MC_JMEPFPuppi
 process = addPaths_MC_JMEPFCluster(process)
 process = addPaths_MC_JMEPFPuppi(process)