Omads

Latest version: v2408.0

Safety actively analyzes 681866 Python packages for vulnerabilities to keep your Python projects secure.

Scan your dependencies

2408.0

What's new!

- New multiobjective optimization (MOO), Pareto study, is now enabled in the search, poll, and MADS solvers
- New constrained and unconstrained MO test functions and examples are added
- New postprocessing mechanism is enabled that allows live update of results in the output files during optimization
- Dependencies are updated to use the recent version of samplersLib

Constrained MO Pareto studies solved using MADS

![constrained_MOO](https://github.com/user-attachments/assets/8906cc92-1e38-41f2-84cc-5818ae99104c)

Unconstrained MO Pareto studies solved using MADS

![unconstMOO](https://github.com/user-attachments/assets/02c376f9-3b6c-457b-b236-5ee509e9024c)

2407.0

- Add a base class for the mesh
- Add and enable a new class object 'Gmesh' that adapts the mesh granularity and anisotropy
- Improve code modularity
- Relax the version of some dependencies

2404.2

- Fix a display issue
- Relax the version requirement of some dependent libraries
- Minor edits

2404.1

- Compatible with Python 3.11
- Fix minor issues related to the dataclass package upgrade
- Upgrade the versions of some dependent libraries

OMADS-2401
New Features

- Significant improvement on solving high-dimensional problems
- Significant Improvement on dealing with narrow feasible regions using extreme, progressive and hybrid barriers
- Inclusion of the samplersLib new library in the search step that has a new active online samplers based on estimated non-parametric distributions
- Dealing with ordinal discrete and categorical parameters
- Results postprocessing and informative output logging
- Modular organized source code

2.1.0

In this release two new modules have been added; `SEARCH` and `MADS`. The following new features has been added too:
* New variable types are now supported; integer, discrete and categorical
* New search types added in the search step
* Sampling: LHC, RS and Halton
* VNS
* Improve constraints handling by using the progressive barrier and the augmented Lagrangian multipliers

1.5.0

In this release, the following new features have been provided:

- Parallel multiprocessing for the points evaluation in the poll.
- New execution command options
- User constant parameters
- Technical and code documentation
- GitHub page for the documentation

Links

Releases

© 2024 Safety CLI Cybersecurity Inc. All Rights Reserved.