flashgg

Introduction

The basic instructions to setup and run flashgg are described here and in corresponding READMEs in subdirectories of the repository.

If you get stuck or have questions, please first consult the FAQs page here.

Flashgg setup for the very low mass analysis in CMSSW_10_6_X:

Get flashgg:

export SCRAM_ARCH=slc7_amd64_gcc700
cmsrel CMSSW_10_6_29
cd CMSSW_10_6_29/src/
cmsenv
git cms-init  
git clone -b lowmass_106X https://github.com/dsperka/flashgg

cd flashgg/
git remote add myFlashGG https://github.com/elfontan/flashgg-1
git remote -v 
git fetch myFlashGG 
git merge myFlashGG/preApproval_lowMass_106X_2018UL
mkdir MicroAOD/data
cd MicroAOD/data
cp /eos/cms/store/group/phys_higgs/cmshgg/flashgg-data/MicroAOD/data/PhotonID_94X_*_RunIIFall17_3_1_0_MCv2_Lowmass_18pT18_M55_BDT.weights.xml .
cp /eos/cms/store/group/phys_higgs/cmshgg/flashgg-data/MicroAOD/data/TMVAClassification_*_SL_2016.xml .
cp -r  /eos/cms/store/group/phys_higgs/cmshgg/flashgg-data/MicroAOD/data/ParticleNetSV/ ../data/.
cp  /eos/cms/store/group/phys_higgs/cmshgg/flashgg-data/MicroAOD/data/QGL_AK4chs_94X.db ../data/.
cd -
cd ..
source flashgg/setup_flashgg.sh 
scram b -j 8

NOTE 1: If the MicroAOD/data directory is there as a symbolic link, remove it (rm data) and recreate it, copying inside the relevant files as indicated above. This is relevant in order to successfully run the crab step for MicroAOD production.

Examples to run on Run 2 2018 UL sample: MicroAOD STEP

• Manual test for microAOD production with a single file:

  cmsRun MicroAOD/test/microAODstd.py processType=[sig] datasetName=glugluh conditionsJSON=MetaData/data/MetaConditions/Era2018_UL_lowMassDiphotonAnalysis.json
      

• MicroAOD step on a dataset (note that you need a proper json file in the campaign directory):

  source /cvmfs/cms.cern.ch/crab3/crab.sh
  voms-proxy-init --rfc --voms cms -valid 192:00
  ./prepareCrabJobs.py -V v0 -C analysisLM_UL18 --meta-conditions /afs/cern.ch/work/e/elfontan/private/DiPhotonAnalysis/myFlashGG/CMSSW_10_6_8/src/flashgg/MetaData/data/MetaConditions/Era2018_UL_lowMassDiphotonAnalysis.json  -O T2_US_MIT -o /store/user/elfontan -s campaigns/TEST_M70_UL.json --mkPilot
      

  The jobs can be submitted and monitored with this kind of commands:

  echo pilot_*.py | xargs -n 1 crab-pre submit                       ## it will submit the pilot test crab jobs
  echo crabConfig_v0_GluGluHToGG_M*.py | xargs -n 1 crab-pre submit  ## submission of the signal jobs
  echo crab_v0_GluGluHToGG_M* | xargs -n 1 crab-pre status           ## to check the status of the jobs
      

  The output is located in DAS in the prod/phys03 database in the form:

  /GluGluHToGG_M70_TuneCP5_13TeV-amcatnloFXFX-pythia8/elfontan-analysisLM_UL18-v0-v0-RunIISummer20UL18MiniAODv2-106X_upgrade2018_realistic_v16_L1v1-v1-82689a8dd9f3f8660dbba0021defef47/USER
      

Examples to run on Run 2 2018 UL sample: Flashgg STEP

• First you need to create a catalogue for the newly produced dataset. From the flashgg directory, run

  fggManageSamples.py -C analysisLM_UL18 import '/GluGluHToGG_M*/elfontan-analysisLM_UL18*/USER'
  
      

  The output will be located in a subdirectory of MetaData/data with the same name of the campaign, e.g.:

  /afs/cern.ch/work/e/elfontan/private/DiPhotonAnalysis/myFlashGG/CMSSW_10_6_8/src/flashgg/MetaData/data/analysisLM_UL18/datasets.json 
      

• Then, you can prepare the configuration file to launch your jobs and produce the final miniTrees. Note that the relevant ingredients for this step are:
  • [1] the name of the campaign (and the PU profile coherent with the era under consideration: make sure to use the UL one!);
  • [2] the name of the MetaCondition file: at the moment use Era2018_UL_lowMassDiphotonAnalysis_noDiphotonBoundaries to run without any categorization;
  • [3] the name of the Systematics configuration file in the Metaconditions: flashggDiPhotonSystematics2018LM_UL_cfi.
• NOTE 1: It is crucial at this stage to choose properly the name of the process because some steps and filters (like for example the PromptFakeFilter) run only for certain processId. If you are ntuplizing a gJets sample, use for example gjets_promptfake, while if you are working with QCD, you can use for example qcd_fakefake.
• NOTE 2: The latest scales and smearing files used by the mass measurement team is not centrally available. Copy them from ~jtao/public/ForElisa/Run2018_09Sep2021_RunFineEtaR9Et_stochastic_oldFormat* to EgammaAnalysis/ElectronTools/data/ScalesSmearings.
• Technical note: copying the proxy file to the working node is not yet supported when using HTCondor as bacth system. Therefore the user must set the X509_USER_PROXY environment variable and run with the --no-copy-proxy option

  cd Systematics/test
  voms-proxy-init -voms cms --valid 168:00
  cp /tmp/MYPROXY ~/
  export X509_USER_PROXY=~/MYPROXY
  fggRunJobs.py --load analysisLM_UL18.json -d outdir_analysisLM_UL18 workspaceStd.py -n 300 -q workday --no-copy-proxy --no-use-tarball