IOHGNBG

IOHGNBG also enables the integration of GNBG problems into IOHprofiler standards. It is provided as a library for use in the GNBG Competition 2025, with additional details available on the competition website.

As part of the IOHprofiler ecosystem, IOHGNBG is actively developed, with ongoing updates to its features and interfaces.

Installation

The minimum supported Python version is 3.10. Install IOHGNBG via pip and git:

pip install iohgnbg

Basic Usage

The following example shows how to get a benchmark GNBG problem (2025 GECCO competition):

import iohgnbg
import ioh
import os
gnbg_problem = iohgnbg.get_problem(1)

print(gnbg_problem.meta_data)

logger = ioh.logger.Analyzer(
    root=os.getcwd(),  # Current working directory
    folder_name="AttachedLogger",  # Folder to store logs
    algorithm_name="None",  # Name of the algorithm (can be customized)
)

# Attach the logger to the created clustering problem
gnbg_problem.attach_logger(logger)

Tutorials

Explore the following Jupyter notebooks for step-by-step tutorials on using IOHGNBG:

1. Random Search Tutorial : Learn how to define a GNBG optimization problems, solve with Random Search and integrate with IOHinspector.

License

This project is licensed under a standard BSD-3 clause License. See the LICENSE file for details.

Acknowledgments

We acknowledge the foundational work on the GNBG generator proposed in the paper:

• "GNBG: A generalized and configurable benchmark generator for continuous numerical optimization" by Danial Yazdani, Mohammad Nabi Omidvar, Delaram Yazdani, Kalyanmoy Deb, and Amir H. Gandomi (arXiv:2312.07083).

The basic instances provided are from the GECCO Competition 2025. We also offer participants the possibility to use IOHprofiler to log their experiments.