SubmitSVFit

cmsrel CMSSW_9_4_4
cd CMSSW_9_4_4/src
cmsenv
git clone https://github.com/SVfit/ClassicSVfit TauAnalysis/ClassicSVfit -b fastMTT_21_06_2018
git clone https://github.com/SVfit/SVfitTF TauAnalysis/SVfitTF
git clone https://github.com/cecilecaillol/SubmitSVFit
cd $CMSSW_BASE/src
scram b -j 4

To submit jobs via condor:

First, merge the skimmed files so that the merged files have about 15000 events. You need to edit controlledMerged.py with the name of the tree in the files, with the names of the input and output directories, and with the list of directories for which the merging should be applied.

cd SubmitSVFit/tools
python controlledMerged.py

Copy the directory with the merged files to /hdfs and enable your proxy.

Then submit the jobs. You need to edit test/svFitSubmitter.py so that it uses the executable for the final state you want to run (SVFitStandAloneFSATauDM_emu_norecoil or SVFitStandAloneFSATauDM_etau_norecoil or SVFitStandAloneFSATauDM_mutau_norecoil). You also need to edit submit_et2018.sh (for example) with the location of the input and output directories, and with the exact list of directories. You can also edit this file to change which systematic shifts are computed and which are not.:

cd SubmitSVFit/test
sh submit_et2018.sh

Old instructions below:

There are a number of current versions of the stand alone svFitter included here. You can find them in ROOT/bin/SVFitStandAlone... with the names of their executables defined ROOT/bin/BuildFile.xml.

To run in interactive mode for example:

SVFitStandAloneFSA inputFile=coolInputFile.root newOutputFile=1 newFile=tmpOut.root doES=1

• inputFile = obvious
• newOutputFile = 0/1
  • 0 = update input file with svFit vars
  • 1 = output new file with original TTree and new svFit vars
• newFile = name of output file, default is newFile.root if none specified
• doES = apply energy scale adjustments providing nominal, shift UP and shift DOWN values of svFit
  • 0 = default, no shift
  • 1 = apply shifts

To submit jobs to condor:

cd test
python svFitSubmitter.py -dr -sd /hdfs/store/user/truggles/svFitTestSept03 -es=1 --jobName svFitSept03Test

• -dr = dryRun and outputs a command for you to run
• -sd = select directory, the input directory, this will produce a list of all files in that directory to run over
  you must have your files in a /hdfs directory.
• -es = apply energy scale, see above
• --jobName = applys a job name to append to your file names and output directory structure

To get your files from elsewhere to /hdfs do something like this:

mkdir /hdfs/store/user/truggles/mySubmitDir
rsync -ahP /nfs_scratch/truggles/httSept04skimMerged/*.root /hdfs/store/user/truggles/httSept04skimMerge/