BLT is a framework for analyzing bacon ntuples originally developed by @jiafulow.
The has been tested to run with CMSSW_8_0_20 with ROOT6. This will work cmslpc using scientific linux 6.
source /cvmfs/cms.cern.ch/cmsset_default.sh
export SCRAM_ARCH=slc6_amd64_gcc530
cmsrel CMSSW_8_0_20
cd CMSSW_8_0_20/src
cmsenv
These forks are sync'ed to Kevin's repositories on 2015/11/17
git clone [email protected]:NWUHEP/BaconProd.git # this is not needed for running BLT
git clone [email protected]:NWUHEP/BaconAna.git
If you plan on producing ntuples using the bacon framework, you will need to checkout some additional CMSSW dependencies,
source BaconProd/scripts/setup_prod.sh
# If copying via AFS is too slow, use scp instead:
# scp -r -C <USERNAME>@lxplus.cern.ch:/afs/cern.ch/work/k/ksung/public/Development/Run2Packages/* ./
git clone [email protected]:NWUHEP/BLT.git
scram b -j 12
cd BLT/BLTAnalysis/test
cmsRun makingBacon_MC_25ns_MINIAOD.py
cd BaconProd && git checkout master && git fetch upstream && git merge upstream/master && git push origin master && cd -
cd BaconAna && git checkout master && git fetch upstream && git merge upstream/master && git push origin master && cd -
There are two example analyzers included in the BLT repository. These produce a set of skimmed ntuples...
7 input arguments are mandatory: [input file] [no of events] [dataset] [datasetgroup] [selection] [period] [jobid]
cd BLT/BLTAnalysis/test
DemoAnalyzer Output.root 1000 DYJetsToLL_M-50 DYJetsToLL mumu 2015 0
or alternatively you can provide a text file with a list of input files,
DemoAnalyzer input.txt 1000 DYJetsToLL_M-50 DYJetsToLL mumu 2015 0
These instructions are intended for running on the Northwestern tier 3, but can be easily modified to run on any batch system that uses condor as a scheduler (e.g., cmslpc).
cd BLT/BLTAnalysis/test/condor
Edit jobify.py L49 from 'MACHINE':'cmslpc' to 'MACHINE' : 'ttgrid',
./run_all_20151119.sh
When the condor jobs are done, run the following,
python retrieve.py