LobsterSkimming

CMS skimming code that's been lobsterized for use on ND's computing resources

Setup dependencies

Run the python setup script

python setup.py

This script will setup a basic CMSSW release and checkout the necessary NanoAOD skimming package. The script also automatically gets the cfg and json directories from the TopEFT/topcoffea repo.

The repo is currently setup to automatically generate and use a CMSSW_10_6_19_patch2 release, but can be changed to use a different release. The simplest method would be to modify the setup.py script, re-run it and then also update the lobster_config.py config by changing the sandbox_location variable.

The CMSSW release that is used to construct the environment that actually runs the skimming job does not have to necessarily match the CMSSW release that is used to provide the dependencies for the lobster program itself. So changing what CMSSW release is used for doing the actual skimming should not have to result in a change to the CMSSW release used for your lobster environment.

Running the code

• Make sure to do a cmsenv in a CMSSW release compatible with your lobster installation

cd CMSSW_X_Y_Z/src
cmsenv

• Activate your lobster environment

source ~/.lobster/bin/activate

• Run the code

lobster process lobster_config.py

Configuring the lobster job

The main configuration option in lobster_config.py is the choice of the cfg_name variable. This should correspond to one of the cfg names from topcoffea cfg files and will determine which dataset samples should be run over. The step, tag, and ver variables are used to define the directory names of the lobster job. If testing=True, then only the step variable will be used and the output directories will instead be changed to something of the form $USER/{step}/test/lobster_test_{tstamp}, where {tstamp} will correspond to a datetime timestamp of when the lobster job was started.