In this realese a first working, though still experimental and under development, parallel genetic algorithm is available. It's a hybrid one, which divides different possible rival Generations between multiple processes to develop and later again divides all Members of these rival Generations between a new batch of parallel process to have their fitness values computed.
Code tested on a simplified case of the endurance function optimisation and surprisingly it works.
What's Changed
- 39-add-docstrings-to-the-classes-in-genetic_algorithmpy by @gnypit in #45
- 20-use-manager-dict-from-multiprocessing-to-store-genome by @gnypit in #47
- manager-only-in-genetic-algorithm by @gnypit in #48
Full Changelog: v0.1.0...v0.1.1
Contributors
gnypit
Assets 2
0
Join discussion