EPJ Web Conf.
Volume 245, 202024th International Conference on Computing in High Energy and Nuclear Physics (CHEP 2019)
|Number of page(s)||5|
|Section||6 - Physics Analysis|
|Published online||16 November 2020|
zfit: scalable pythonic fitting
University of Zurich
* e-mail: Jonas.Eschle@cern.ch
** e-mail: email@example.com
*** e-mail: firstname.lastname@example.org
**** e-mail: email@example.com
Published online: 16 November 2020
Statistical modeling and fitting is a key element in most HEP analyses. This task is usually performed in the C++ based framework ROOT/RooFit. Recently the HEP community started shifting more to the Python language, which the tools above are only loose integrated into, and a lack of stable, native Python based toolkits became clear. We presented zfit, a project that aims at building a fitting ecosystem by providing a carefully designed, stable API and a workflow for libraries to communicate together with an implementation fully integrated into the Python ecosystem. It is built on top of one of the state-of-theart industry tools, TensorFlow, which is used the main computational backend. zfit provides data loading, extensive model building capabilities, loss creation, minimization and certain error estimation. Each part is also provided with convenient base classes built for customizability and extendability.
© The Authors, published by EDP Sciences, 2020
This is an Open Access article distributed under the terms of the Creative Commons Attribution License 4.0, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.