Version 2.0

New Python implementation

With this release, we added a Python implementation of the Hybrid Patch Decomposition algorithm (HyPaD) which can be found in the python subfolder. The Matlab version is now located in the matlab subfolder and is the same as in the previous release.

We provide a Jupyter Notebook as an interface to the Python version in the UserFile.ipynb. It contains all instructions that you need to get started.

Please be aware that this is the first public release of the new Python version of the HyPaD algorithm. While we have tested it on several test instances it has not been tested and used as extensively as the Matlab version. Thus, you might encounter some issues that we have not yet detected. In this case, please create a new issue on GitHub and I'll take a look at it.

Version 1.1

Adjustments for generalized bound concepts

This release of the HyPaD algorithm makes use of the generalized local upper/lower bound concepts. One advantage of this change is that it simplifies the update of the lower bound sets on the patch level. What is more, the generalized bound concepts allow to potentially remove empty boxes from the enclosure. We refer to our corresponding paper¹ for more details and a comparison of the generalized and classic bound concepts.

New interface for optimization problems

The interface for the optimization problem files (e.g., those in the problems folder) has been slightly extended. HyPaD now uses the same interface as the enclosure algorithms AdEnA¹ and MOMIBB² that are also available on GitHub. This makes it easier to compare the different algorithms with each other.

MIT license

In order to make the HyPaD algorithm as accessible as possible, we decided to change the license to MIT license.

References

1. Gabriele Eichfelder and Leo Warnow, Advancements in the computation of enclosures for multi-objective optimization problems, European Journal of Operational Research, 310(1), 315-327, 2023. ↩ ↩²

2. Gabriele Eichfelder, Oliver Stein, and Leo Warnow, A solver for multiobjective mixed-integer convex and nonconvex optimization, Journal of Optimization Theory and Applications, 2022. ↩

Version 1.0

Public Release of HyPaD

This is the first release version of HyPaD - a Hybrid Patch Decomposition Algorithm to solve multi-objective mixed-integer optimization problems. It is proven to work with all test instances provided with this release.

However, since this is the first public release of the algorithm, you might encounter some issues that we have not yet detected. In this case, please create a new issue on GitHub and I'll take a look at it.