Conference site ยป Proceedings

Boost-histogram: High-Performance Histograms as Objects

Henry Schreiner
Princeton University

Hans Dembinski
TU Dortmund

Shuo Liu
Sun Yat-sen University

Jim Pivarski
Princeton University



Unlike arrays and tables, histograms in Python have usually been denied their own object, and have been represented as a single operation producing several arrays. Boost-histogram is a new Python library that provides histograms that can be filled, manipulated, sliced, and projected as objects. Building on top of the Boost libraries' Histogram in C++14 provided interesting distribution and design challenges with useful solutions. This is meant to be a foundation that others can build on; in the Scikit-HEP project1, a physicist friendly front-end \textquotedbl{}Hist\textquotedbl{} and a conversion package \textquotedbl{}Aghast\textquotedbl{} are already being designed around boost-histogram.


Histogram, Analysis, Data processing, Data reduction, NumPy, Aggregation



Bibtex entry

Full text PDF