BOCoDe Documentation

BOCoDe is a comprehensive Python library containing optimization benchmark problems for testing and evaluating optimization algorithms.

Note

This project is under active development.

Features

  • Diverse Benchmark Collection: Problems from various sources including Botorch, BBOB, CEC competitions, and engineering applications

  • Multiple Problem Types: Single and multi-objective, with and without constraints

  • Standardized Interface across all benchmark problems

  • Function Filtering to find problems matching specific criteria

  • Visualization Tools for understanding problem landscapes and behavior

  • Framework Integration

Citing

  1. If you use BOCoDe in your research, please cite the following paper (full report coming soon):

    @misc{yu2025bocode,

    author={Rosen Ting-Ying Yu, Advaith Narayanan, Cyril Picard, Faez Ahmed},

    title = {{BOCoDe}: Benchmarks for Optimization and Computational Design},

    year={2025},

    url={https://github.com/rosenyu304/BOCoDe} }

  2. If you use the the BOCoDe interfaces to other libraries or open source code functions (ex: BoTorch, BBOB, NEORL, MODAct, LassoBench, …), please cite them accordingly.

License

BOCoDe is MIT licensed.

Contents

Contents: