Release Notes

Jump to bottom

Martin Buist edited this page Jul 5, 2018 · 2 revisions

Release 3.1 (July 2018)

Added Simulated Annealing as an optimisation method
Added multi-level multi-start optimisation

List of optimisers now included now accessible with the same cost function

Nelder-Mead simplex
Levenberg-Marquardt (least squares)
Genetic Algorithm
Differential Evolution
Particle Swarm
Simulated Annealing

Along with Box constraints, should you need them.

Release 3.0 (August, 2014)

Introduction of Particle Swarm algorithm.
Improved GenericOptimizer that collates and provides common functionality.
Experimental support for multi-threaded execution using C++11 futures.
All unit tests and examples have been grouped into suites.

Release 2.1 (November, 2013)

Introduction of Genetic algorithm.
Introduction of Differential Evolution algorithm.

Release 2.0 (June, 2013)

Introduction of Levenberg-Marquardt algorithm.
There was a change in the interface of GenericCostFunction. The overloaded operator () now has to return a vector of residuals defined as the difference between the experimental data and your model equation.

Release 1.1 (August, 2012)

Significant code refactoring, with addition of GenericOptimizer class. Improved testing.

Release 1.0 (June, 2012)

First release of Unfit with working Nelder-Mead algorithm for minimisation problems.