Home

TopSF

A better framework for scale factor measurement, designed with flexibility in mind.

For feature request or bug report please make an issue in this GitHub repository. Feedback is always appreciated here. Thanks!

This set of Python scripts is designed to help users set up the histograms and datacards for Higgs Combine necessary for the scale factor measurement. Previously there has not been a decent framework with enough flexibility to support any kind of configuration, from the name and location of input ntuple files, category naming, to the number of tagging and number of processes added into the measurement.

Here, instead of having everything hardcoded somewhere, buried deep in C++ code, the configurations for pretty much everything, from the name and location of input ntuple files, category naming, etc., are located in one single easy-to-read YAML file. The code presented here will have as little assumptions as possible, solely designed for scale factor measurement for any object tagger.

Features

• Supports any ROOT ntuple structure. No need to conform to any hardcoded format.
• Supports any tagger that gives one discriminator value, such as cut-based, BDT, or neural networks.
• Supports any file name format, any number of processes, and any number of tagging categories you have.
• Supports any kind of event category definition. You are not limited to pT ranges. You can define event categories in any way with any number of variables.
• Everything can be defined in one YAML file for datacard creation. No more surprises hidden deep in raw code.
• Also contains helpful HiggsCombine script to give you an idea of what you can do with the output datacard.