[Coin-announce] New project: RBFOpt, a library for black-box optimization

Giacomo Nannicini giacomo.n at gmail.com
Tue Jun 23 08:34:51 EDT 2015

I am happy to announce that RBFOpt is now available on COIN-OR. RBFOpt
is a library written in Python for black-box (also known as
derivative-free) optimization. It is a global solver that requires
zero-order information - just function values, not even gradients!
RBFOpt is aimed at finding the global minimum in a small number of
function evaluations, which is particularly useful when each function
evaluation requires an expensive computation, e.g. a time-consuming
computer simulation. The software is released under the Revised BSD

At the moment, RBFOpt can handle box-constrained (that is, all
variables are bounded) mixed-integer problems. Using RBFOpt in an
application is very easy: the end-user only needs to implement the
function that evaluates the objective function and provide the
variable bounds.

I encourage you to use RBFOpt and/or contribute! More information as
well as the relevant links can be found on the homepage:
Please note that the project documentation and more detailed
installation instructions are available on GitHub.

Comments, questions and feedback are welcome!


More information about the Coin-announce mailing list