stanhf: HistFactory models in the probabilistic programming language Stan

Abstract

In collider physics, experiments are often based on counting the numbers of events in bins of a histogram. We present a new way to build and analyze statistical models that describe these experiments, based on the probabilistic programming language Stan and the HistFactory specification. A command-line tool transpiles HistFactory models into Stan code and data files. Because Stan is an imperative language, it enables richer and more detailed modeling, as modeling choices defined in the HistFactory declarative specification can be tweaked and adapted in the Stan language. Stan was constructed with automatic differentiation, allowing modern computational algorithms for sampling and optimization.

arXiv comment8 pages, see https://github.com/xhep-lab/stanhf

References in corpus (14)

• Asymptotic formulae for likelihood-based tests of new physics
• MCMC using Hamiltonian dynamics
• Rank-normalization, folding, and localization: An improved $\hat{R}$ for assessing convergence of MCMC
• Visualization in Bayesian workflow
• ROOT - A C++ Framework for Petabyte Data Storage, Statistical Analysis and Visualization
• A Future Percent-Level Measurement of the Hubble Expansion at Redshift 0.8 With Advanced LIGO
• HEPData: a repository for high energy physics data
• Evidence for $B^{+}\to K^{+}\nu \bar{\nu }$ decays
• Search for bottom-squark pair production with the ATLAS detector in final states containing Higgs bosons, $b$-jets and missing transverse momentum
• Etalumis: Bringing Probabilistic Programming to Scientific Simulators at Scale
• Reframing Jet Physics with New Computational Methods
• Probabilistic Programming with Densities in SlicStan: Efficient, Flexible and Deterministic
• The Bayes factor surface for searches for new physics
• A note on blind technique for new physics searches in particle physics

Cited by in corpus (1)

• Inferring correlated distributions: boosted top jets