Solid

Why we included this project

Solid collects the classic gradient-free optimization methods into one Python library with a shared interface. Instead of pulling together separate packages for genetic algorithms, simulated annealing, particle swarm, tabu search, harmony search, and evolutionary algorithms, you subclass the algorithm you want, implement a few abstract methods, and call run() to get the best solution and its objective value. That consistency is what makes it useful: you can prototype a metaheuristic quickly or compare several approaches on the same problem without rewriting boilerplate. The implementations are deliberately basic and readable, so the library doubles as a teaching tool and is easy to modify when you need a custom variant. It suits optimization problems where gradients are unavailable or impractical, such as discrete search, scheduling, and black-box tuning.