Contents
Run the following outside a release to get a new release containing some configurations in the src directory:
wget https://raw.githubusercontent.com/cms-ttH/ttH-TauMCGeneration/master/test/ttl_fast.sh bash ttl_fast.sh
The resulting configurations are a mashup of information from McM and the FastSim TWiki pages. Examples and templates for CRAB submission files can be found in test/crab.
The following settings should be updated in the script referenced above for following iterations, most of this information is near the top of the file:
- Era (
era) - Software release (
release) - Conditions (
globaltag) - Pile-up dataset (
premix) - The HLT menu (
hlt)
Check the beamspot settings, too!
Example configurations to copy LHE data out of Notre Dame can be generated (from the base directory of this repository):
cd test python crab/copy.py
with CRAB configuration files generated in test/crab.
We can look at either extracting the full chain in one query, or following a specific LHE dataset (recommended).
To clone workflows from McM, select a dataset to clone, i.e.:
/TTToSemilepton_TuneCUETP8M2_ttHtranche3_13TeV-powheg-pythia8/RunIISpring16MiniAODv2-premix_withHLT_80X_mcRun2_asymptotic_v14-v1/MINIAODSIM
Then visit the McM homepage, and click the “Request” item at the top of the page, and select “Output Dataset” below to paste the dataset to clone:
Searching for workflows to produce a dataset.
By selecting the link in the column “Dataset name” in the search results, all steps to reproduce the original dataset can be obtained:
Workflows to completely reproduce a dataset.
If this yields too many results, an alternative is to follow the “parent” links in DAS and look at the field “McM info” under the dataset name. Then search for the McM id by selecting “Navigation” and then filling in the field “prepid”:
Searching for the parent LHE dataset and its McM info.
Using this prepid to search for the McM setup.
Now the second to left icon 
in the search results can be used to download
the setup scripts. For the LHEGS step:
#!/bin/bash
source /afs/cern.ch/cms/cmsset_default.sh
export SCRAM_ARCH=slc6_amd64_gcc481
if [ -r CMSSW_7_1_25/src ] ; then
echo release CMSSW_7_1_25 already exists
else
scram p CMSSW CMSSW_7_1_25
fi
cd CMSSW_7_1_25/src
eval `scram runtime -sh`
export X509_USER_PROXY=$HOME/private/personal/voms_proxy.cert
curl -s --insecure https://cms-pdmv.cern.ch/mcm/public/restapi/requests/get_fragment/HIG-RunIISummer15wmLHEGS-00482 --retry 2 --create-dirs -o Configuration/GenProduction/python/HIG-RunIISummer15wmLHEGS-00482-fragment.py
[ -s Configuration/GenProduction/python/HIG-RunIISummer15wmLHEGS-00482-fragment.py ] || exit $?;
scram b
cd ../../
cmsDriver.py Configuration/GenProduction/python/HIG-RunIISummer15wmLHEGS-00482-fragment.py \
--fileout file:HIG-RunIISummer15wmLHEGS-00482.root \
--mc --eventcontent RAWSIM,LHE \
--customise SLHCUpgradeSimulations/Configuration/postLS1Customs.customisePostLS1,Configuration/DataProcessing/Utils.addMonitoring \
--datatier GEN-SIM,LHE --conditions MCRUN2_71_V1::All --beamspot Realistic50ns13TeVCollision \
--step LHE,GEN,SIM --magField 38T_PostLS1 \
--python_filename HIG-RunIISummer15wmLHEGS-00482_1_cfg.py --no_exec -n 42 || exit $? ;Repeat this for the DR step to see the following lines
cmsDriver.py step1 \
--filein "dbs:/TTToSemilepton_TuneCUETP8M2_ttHtranche3_13TeV-powheg-pythia8/RunIISummer15wmLHEGS-MCRUN2_71_V1-v1/GEN-SIM" \
--fileout file:HIG-RunIISpring16DR80-01830_step1.root \
--pileup_input "dbs:/Neutrino_E-10_gun/RunIISpring15PrePremix-PU2016_80X_mcRun2_asymptotic_v14-v2/GEN-SIM-DIGI-RAW" \
--mc --eventcontent PREMIXRAW --datatier GEN-SIM-RAW --conditions 80X_mcRun2_asymptotic_v14 \
--step DIGIPREMIX_S2,DATAMIX,L1,DIGI2RAW,HLT:25ns10e33_v2 --nThreads 4 --datamix PreMix --era Run2_2016 \
--python_filename HIG-RunIISpring16DR80-01830_1_cfg.py --no_exec \
--customise Configuration/DataProcessing/Utils.addMonitoring -n 960 || exit $? ;
cmsDriver.py step2 \
--filein file:HIG-RunIISpring16DR80-01830_step1.root \
--fileout file:HIG-RunIISpring16DR80-01830.root \
--mc --eventcontent AODSIM,DQM --runUnscheduled --datatier AODSIM,DQMIO \
--conditions 80X_mcRun2_asymptotic_v14 \
--step RAW2DIGI,RECO,EI,DQM:DQMOfflinePOGMC --nThreads 4 --era Run2_2016 \
--python_filename HIG-RunIISpring16DR80-01830_2_cfg.py --no_exec \
--customise Configuration/DataProcessing/Utils.addMonitoring -n 960 || exit $? ;Note that this step is happening in CMSSW_8_0_14.
And for the MiniAOD step (note again the used release, which is equivalent to the DR step)
cmsDriver.py step1 \
--filein "dbs:/TTToSemilepton_TuneCUETP8M2_ttHtranche3_13TeV-powheg-pythia8/RunIISpring16DR80-premix_withHLT_80X_mcRun2_asymptotic_v14-v1/AODSIM" \
--fileout file:HIG-RunIISpring16MiniAODv2-02983.root \
--mc --eventcontent MINIAODSIM --runUnscheduled --datatier MINIAODSIM \
--conditions 80X_mcRun2_asymptotic_v14 \
--step PAT --era Run2_2016 \
--python_filename HIG-RunIISpring16MiniAODv2-02983_1_cfg.py --no_exec \
--customise Configuration/DataProcessing/Utils.addMonitoring -n 960 || exit $? ;Note the following from the FastSim TWiki about runTheMatrix.py workflows:
FullSim:
- do not run harvesting (the 4th command printed by runTheMatrix.py)
- do not run the ALCA step (the 5th command printed by runTheMatrix.py)
- in the 2nd command printed by runTheMatrix.py
- replace the DIGI:pdigi_valid with DIGI (do not produce the truth collection 'trackingParticles')
- in the 3rd command
- remove ,EI,VALIDATION from the -s option
- replace --eventcontent RECOSIM,DQM with --eventcontent AODSIM
- replace --datatier GEN-SIM-DIGI-RECO,DQMIO with --datatier AODSIM
Unfortunately, these steps don't work well with the McM workflows. First, a pile-up sample needs to be produced. To set up the environment
scram p CMSSW CMSSW_8_0_20
cd CMSSW_8_0_20/src
curl -s --insecure https://cms-pdmv.cern.ch/mcm/public/restapi/requests/get_fragment/HIG-RunIISummer15wmLHEGS-00482 --retry 2 --create-dirs -o Configuration/GenProduction/python/HIG-RunIISummer15wmLHEGS-00482-fragment.py
git clone git@github.com:cms-ttH/ttH-TauMCGeneration.git ttH/TauMCGeneration
eval `scram runtime -sh`
scram b
cd ../..To directly produce AODSIM using FastSim, the following cmsDriver.py
command can be used:
cmsDriver.py Configuration/GenProduction/python/HIG-RunIISummer15wmLHEGS-00482-fragment.py \
-n 500 \
--python_filename all_fast.py \
--fileout file:all_fast.root \
--pileup_input "dbs:/Neutrino_E-10_gun/RunIISpring16FSPremix-PUSpring16_80X_mcRun2_asymptotic_2016_v3-v1/GEN-SIM-DIGI-RAW" \
--mc --eventcontent AODSIM --fast \
--customise SimGeneral/DataMixingModule/customiseForPremixingInput.customiseForPreMixingInput \
--customise ttH/TauMCGeneration/customGenFilter.customizeForGenFiltering \
--datatier AODSIM --conditions auto:run2_mc --beamspot Realistic50ns13TeVCollision \
--step LHE,GEN,SIM,RECOBEFMIX,DIGIPREMIX_S2,DATAMIX,L1,DIGI2RAW,L1Reco,RECO,HLT:@fake1 \
--datamix PreMix \
--era Run2_25ns \
--no_exec \Notice the second --customise parameter, which will set up a basic
filter after the generation step to trim the selected events.
Premixed pile-up is available from the following two datasets:
/Neutrino_E-10_gun/RunIISpring16FSPremix-PUSpring16_80X_mcRun2_asymptotic_2016_v3-v1/GEN-SIM-DIGI-RAW /Neutrino_E-10_gun/RunIISummer16FSPremix-PUMoriond17_80X_mcRun2_asymptotic_2016_TrancheIV_v4-v1/GEN-SIM-DIGI-RAW
As a final step, produce the MiniAOD:
cmsDriver.py \
-n 500 \
--python_filename maod_fast.py \
--fileout file:moad_fast.root \
--filein file:all_fast.root \
--mc --eventconcent MINIAODSIM --fast \
--datatier MINIAODSIM --conditions auto:run2_mc \
--step PAT --runUnscheduled \
--no_execSomething like the following can be added to a parameter set to filter MiniAOD:
process.load('ttH.TauMCGeneration.eventFilterMAOD_cfi')
process.evpath = cms.Path(process.ttHfilter)
process.output = cms.OutputModule(
"PoolOutputModule",
fileName=cms.untracked.string(options.outputFile),
outputCommands=cms.untracked.vstring(['keep *']),
SelectEvents=cms.untracked.PSet(SelectEvents=cms.vstring('evpath')),
dataset=cms.untracked.PSet(filterName=cms.untracked.string(''))
)