This package provides an implementation of the derivative-free optimization algorithm, DFO, developed by A. Conn, K. Scheinberg, L. Vicente. Using this package, the user can solve a derivative-free blackbox optimization problem with the DFO method as well as five derivative free algorithms from the scipy.optimize library. The scipy algorithms are the Nelder-Mead, Powell, SLSQP, COBYLA and BFGS algorithms.
To run a set of sample problems, the user can call the “run_test_func.py” module.
- Write your blackbox optimization function in a new Python module. For examples of such functions see blackbox_opt/test_funcs/funcs_def.py
- Write a run file with the module you wrote in the previous step imported. You can modify the run_test_func.py module and use it as your run file. Note that you should specify the algorithms(s) that you wish to solve your problems with, a starting point and suitable options.