NanoAnalysis

HEP-EX analyzing framework for NanoAOD data in CMS experiment (in dev)

Attention: This repo. is still in update so the documentation is not complete and may be outdated. I will make a more detailed documentation after everything is prepared

Python & coffea/uproot instead of C++ & ROOT

Vectorization instead of for-loop

Brief introduction

• datasets/

  Hierarchically stores dataset info in the structure of datasets/$type/$year/$channel/$division/$version.yaml

• src/

  Containing all source code to process NanoAOD into slimmed files (e.g., *.parq) and to calculate correction and uncertainty by coffea (read *.root files into vector-like structure, e.g., NanoEvents) and awkward (a numpy-like library but more capable to process nested and variable-sized data in json-like format)

• condor/

  Use parseDatasets.py to parse filelists from *.yaml and generate corresponding job description files as *.submit. Then use integrated condor_submit.py script to submit condor jobs for processing NanoAOD files with codes in src/

• notebooks/

  Jupyter-notebboks to analyze/postprocess *.parquet coming from processed NanoAOD data. Produce the final plots here!

References

NanoAOD data tier