What's Changed

Improved contributing guidelines
Reduced duplication of metadata in the Python files
Improved documentation
Refactoring mostly for shebo by @weslleyspereira in #14

Full Changelog: v2.0.0...v2.0.1

Contributors

weslleyspereira

Assets 2

11 Dec 17:45

weslleyspereira

v2.0.0

d4d0970

soogo v2.0.0

What's Changed

Add MaximizeDistance acquisition function and remove duplicate code by @Byron-Selvage in #6
Shebo by @weslleyspereira in #7
Clean fsapso by @Byron-Selvage in #9
fsapso by @Byron-Selvage in #8
Convergence tests by @Byron-Selvage in #10
Restructure acquisition methods by @weslleyspereira in #12
Restructure optimize by @weslleyspereira in #13

New Contributors

@Byron-Selvage made their first contribution in #6

Full Changelog: v1.2.2...v2.0.0

Contributors

weslleyspereira and Byron-Selvage

Assets 2

14 Jul 22:02

weslleyspereira

v1.2.2

ae38a52

soogo v1.2.2

Fix issues with PDM

Full Changelog: v1.2.1...v1.2.2

Assets 2

14 Jul 21:08

weslleyspereira

v1.2.1

bf7d57d

soogo v1.2.1

Full Changelog: v1.2.0...v1.2.1

Assets 2

19 May 23:54

weslleyspereira

v1.2.0

04119fc

soogo v1.2.0

Staying await of blackboxoptim to avoid confusion with the Julia package blackboxoptim.jl .

This release also removes unnecessary package dependencies.

Full Changelog: v1.1.0...v1.2.0

Assets 2

19 Mar 21:56

weslleyspereira

v1.1.0

58410c0

v1.1.0

BlackBoxOptim v1.1.0

Main changes:

Improve all examples in the examples folder.
Expose most of optimize.py to the main API of the package.
In optimize.py:
- Surrogates are updated with the samples from last iteration.
  This only occurs if the surrogate can be accessed outside the function scope.
- surrogate_optimization() and socemo() end execution the first time an iteration fails to acquire new points.
- Fix the local acquisition function neval in SOCEMO.
In acquisition.py:
- Use Pymoo default termination methods instead of MaximumGenerationTermination.
- WeightedAcquisition.argminscore() now returns -1 if there is no feasible point.
- WeightedAcquisition.acquire() may now return less than n points.
- ParetoFront.acquire() now can use information about the non-dominated sample points to initialize the genetic algorithm.
- ParetoFront.acquire() only choose points that are expected to be non-dominated.