jaehyeok/nanoprocessing

Package to process CMS nanoAOD to flat ntuples

nanoprocessing

This is a package to process CMS nanoAOD to flat ntuples. It is assumed that you are running this package at KISTI.

Setup

You can set up environment, you can do the following on a KISTI machine.

source /cvmfs/cms.cern.ch/cmsset_default.sh
cd /cvmfs/cms.cern.ch/slc7_amd64_gcc700/cms/cmssw/CMSSW_10_5_0/src
cmsenv
cd -

Then, check out the package

git clone [email protected]:jaehyeok/nanoprocessing.git

Fastjet

The package uses Fastjet [1] for fat-jet clustering (to calculate MJ). So, you need to install and set it up on your machine. Or, include /cms/ldap_home/<Username>/tools/fastjet-install/bin in your PATH.

compilation

The compliation can be done by running ./complie.ch. It compiles two scripts (process_nano.cpp and skim.cpp) and copy them to your home directory so that they can be used by condor.

Run

You can run the machinery with ./process_nano.exe. It takes a 4 arguments.

$ ./process_nano.exe
 Please provide proper arguments

   ./process_nano.exe [input dir] [output dir] [process] [list of processed files]

The "list of processed files" is generated by condor/check_progress.py.
An example is

./process_nana.exe /xrootd/store/data/Run2017D/SingleMuon/NANOAOD/Nano14Dec2018-v1 root://cms-xrdr.private.lo:2094//xrd/store/user/jaehyeok/2017v4/2019_10_23/ SingleMuonRun2017D flist_outputdir_xrd_store.txt

Skim

You can run the skim

$ ./skim.exe /xrootd_user/yjeong/xrootd/nanoprocessing/[year]/JetHTRun_v7 rpvfitnbge0 [simultaneous run number] 

You can use the skim script

$ ./scripts/run_skim.sh [year]

Renormalization

Renormalization of tree branch

$ ./norm_weights.exe [input dir] [tag] [output dir] [pre-norm dir]

Slurm submit for processing

cd condor
./submit_all.py [year]

where year = 2016, 2017 or 2018. Under ~/flist diretory files that contain the nanoAOD files to be processed will be generated. It list is based on samples/samples[year].txt. In addition, slurm submit files will be generated in submit_scripts directory together with the actuall commands on the terminal. You can just copy the command and paste them in command line.

Slurm submit for renormalization

cd condor
./submit_norm.py [year]

where year = 2016, 2017 or 2018. Under ~/flist/norm diretory files that contain the nanoAOD files to be processed will be generated. In addition, slurm submit files will be generated in submit_scripts_norm directory together with the actuall commands on the terminal. You can just copy the command and paste them in command line.

Tools

Check the status of the processing by by ./check_progress.py [year]. It will prin out a table of status, and generate a list of process files in ~/flist directory.

[1] http://fastjet.fr/