Behold a colab notebook that recommends a black-box optimizer for your objective function. It is powered by this repository, that assigns Elo ratings (json and html) to global derivative-free Python global optimization "strategies", using the HumpDay package.
See the article for a detailed description.
A choice of python package (such as scipy.optimize, ax-platform, hyperopt, nevergrad, optuna, platypus, pymoo, pySOT or skopt) together with a fixing of additional choices (such as selection of method, choice of parameters etc).
HumpDay contains scripts for choosing optimizers based on your own problems.
Got a good objective function, or optimization strategy? Shove it in the HumpDay package and file a pull request, or at least suggest it in the issues.
