Derivative-free optimization algorithms
This package contains derivative-free optimization algorithms. With the exception of NMK, the included algorithms are even “rank-based” algorithms, which do not use the actual objective values, but only their ranks. Thus, they do not require objective values to be scalar and finite. A total order on the objective values is sufficient. For example, an objective function could return a list of values, which are then compared lexicographically. The package can only handle bound constraints explicitly. For other constraints, infeasible solutions should instead be penalized, e.g., evaluated with the worst possible function value (possibly infinity).
The package is geared to work with optimization problems as defined in the package optproblems. The whole package assumes minimization problems throughout.
The documentation is located at https://www.simonwessing.de/dfoalgos/doc/
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
|Filename, size & hash||File type||Python version||Upload date|
|dfoalgos-0.5.tar.gz (21.4 kB) View hashes||Source||None|