Navigation Menu

Skip to content

kdlong/UWVV

 
 

Repository files navigation

UWVV

Some tools for CMS analyses

UWVV is designed for analyses that use final state particles (typically leptons) to reconstruct intermediate and initial states. For example, in the H->ZZ->4l analysis, electron and muon pairs are built into Z candidates, and the Z candidates are built into Higgs candidates. It contains tools for building a full analysis flow out of CMS EDM modules, and for making flat ntuples where each row represents one initial state candidate.

It uses the CMSSW framework and expects miniAOD input. Much of the inspiration (and a little bit of the code) comes from FSA. A few tools, like the batch submission scripts, are specific to the computing infrastructure at the University of Wisconsin - Madison.

Setup

Current supported CMSSW release: CMSSW_8_0_30+

scram pro -n uwvv CMSSW CMSSW_8_0_[current version]
cd uwvv/src
cmsenv
git cms-init # do before anything else
git clone --recursive https://github.com/uwcms/UWVV.git # or git@github.com:uwcms/UWVV.git if you prefer that
cd UWVV
source recipe/setup.sh # install necessary packages
pushd ..
scram b -j 8 # compile
popd

Several fragile dependencies that are used in only some analyses are included only if the --hzzExtras or --met options are used with setup.sh. Modules that depend on the optional packages are saved in .txt files which are copied to .cc files.

To set up a python virtual environment with Rootpy, a non-bugged IPython, and some other nice things

source recipe/setupPython.sh

The first time you call setupPython.sh, it creates the virtualenv, installs the packages in it, and activates it. After that, it just activates it.

Use

To make a basic ntuple of four-lepton final state candidates, do

cmsRun Ntuplizer/test/ntuplize_cfg.py channels='zz' isMC=1 inputFiles=file:aNiceMonteCarloFile.root

For more on how to build your own analysis, see the AnalysisTools directory. For more on making ntuples, see the Ntuplizer directory.

About

Some tools for CMS analyses

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • C++ 50.3%
  • Python 43.7%
  • Shell 6.0%