PyPRS: A Python Software Package for Parallel Ranking and Selection Procedures

We introduce Python Parallel Ranking and Selection (PyPRS), a Python software package specifically developed to solve large-scale ranking and selection problems in parallel computing environments. The underlying parallel computing framework is Ray. PyPRS incorporates four well-known parallel procedures: the good selection procedure, the parallel adaptive survivor selection procedure, the knockout-tournament procedure, and the fixed-budget knockout-tournament procedure. The key features of PyPRS include (i) a modular structure that facilitates easy extension, enhancement, and customization of procedures; (ii) a plug-and-play functionality that enables easy applications of both built-in and custom procedures to various problems; and (iii) an intuitive graphical user interface that improves user accessibility and ease of operation. History: Accepted by Ted Ralphs, Area Editor for Software Tools. Funding: G. Jiang was supported by the National Natural Science Foundation of China [Grants 72293562 and 72171060]. Y. Zhong was supported by the National Natural Science Foundation of China [Grants 72571049 and 72101047], the Major Program of National Social Science Foundation of China [Grant 25&ZD196], and the Humanities and Social Science Fund of Ministry of Education of China [Grant 24XJA630003]. Supplemental Material: The software that supports the findings of this study is available within the paper and its Supplemental Information ( https://pubsonline.informs.org/doi/suppl/10.1287/ijoc.2024.1045 ) as well as from the IJOC GitHub software repository ( https://github.com/INFORMSJoC/2024.1045 ). The complete IJOC Software and Data Repository is available at https://informsjoc.github.io/ .