ROC++: Robust Optimization in C++

Over the last two decades, robust optimization has emerged as a popular means to address decision-making problems affected by uncertainty. This includes single-stage and multi-stage involving real-valued and/or binary decisions exogenous (decision-independent) endogenous (decision-dependent) uncertain parameters. Robust techniques rely on duality theory potentially augmented with approximations transform (semi-)infinite problem finite program, counterpart. Whereas writing down model for is usually simple task, obtaining counterpart requires expertise. To date, very few solutions are available that can facilitate modeling solution of such problems. been major impediment their being put practical use. In this paper, we propose ROC++, an open-source C++ based platform automatic optimization, applicable wide array both parameters, easy use extend. It also applies certain classes stochastic programs continuously distributed parameters Our naturally extends existing off-the-shelf deterministic platforms offers ROPy, Python interface in form callable library, ROB file format storing sharing We showcase power ROC++ several interest. help streamline researchers practitioners. comes detailed documentation its expansion. The latest version be downloaded from https://sites.google.com/usc.edu/robust-opt-cpp/ . Summary Contribution: paper “ROC++: Optimization C++” proposes new modeling, automatically reformulating, solving real- binary-valued adaptive variables. language similar one provided case state-of-the-art solvers. library. History: Accepted Ted Ralphs, Area Editor Software Tools. Funding: material upon work supported National Science Foundation under Grant No. 1763108. support gratefully acknowledged. Supplemental Material: software supports findings study within Supplementary Information ( https://pubsonline.informs.org/doi/suppl/10.1287/ijoc.2022.1209 ) or IJOC GitHub repository https://github.com/INFORMSJoC at https://dx.doi.org/10.5281/zenodo.6360996 ).