HistFitter software framework for statistical data analysis

• Published in: The European physical journal. C, Particles and fields Particles and fields, 2015-04, Vol.75 (4), p.1-20, Article 153
• Main Authors: Baak, M., Besjes, G. J., Côté, D., Koutsman, A., Lorenz, J., Short, D.
• Format: Article
• Language: English
• Subjects: Astronomy; Astrophysics and Cosmology; Collection; Collisions (Nuclear physics); Computer programs; Construction; Data processing; Distribution (Probability theory); Elementary Particles; Hadrons; Heavy Ions; Large Hadron Collider; Measurement Science and Instrumentation; Nuclear Energy; Nuclear Physics; Particle accelerators; Physics; Physics and Astronomy; Quantum Field Theories; Quantum Field Theory; Regular Article - Experimental Physics; Searching; Software; String Theory
• ISSN: 1434-6044; 1434-6052
• DOI: 10.1140/epjc/s10052-015-3327-7

We present a software framework for statistical data analysis, called HistFitter, that has been used extensively by the ATLAS Collaboration to analyze big datasets originating from proton–proton collisions at the Large Hadron Collider at CERN. Since 2012 HistFitter has been the standard statistical tool in searches for supersymmetric particles performed by ATLAS. HistFitter is a programmable and flexible framework to build, book-keep, fit, interpret and present results of data models of nearly arbitrary complexity. Starting from an object-oriented configuration, defined by users, the framework builds probability density functions that are automatically fit to data and interpreted with statistical tests. Internally HistFitter uses the statistics packages RooStats and HistFactory. A key innovation of HistFitter is its design, which is rooted in analysis strategies of particle physics. The concepts of control, signal and validation regions are woven into its fabric. These are progressively treated with statistically rigorous built-in methods. Being capable of working with multiple models at once that describe the data, HistFitter introduces an additional level of abstraction that allows for easy bookkeeping, manipulation and testing of large collections of signal hypotheses. Finally, HistFitter provides a collection of tools to present results with publication quality style through a simple command-line interface.