The NEOS Server offers Condor for the derivative-free minimization of
inequality constrained problems in
Condor is an extension of
unconstrained optimimization by quadratic approximation method,
to constrained problems. See also the article in Math. Prog. B, vol. 92 (2002).
Condor is intended for difficult, noisy, and expensive to evaluate functions for
which derivatives are not or not easily available. It evaluates only function
Problems should be of moderate dimension (up to about 50 variables).
A typical application problem is described
Condor has interfaces to AMPL, SIF, C++, and XML. It is available in serial and
parallel versions. Source code, binaries and
documentation are available
technical report includes in particular a comparison to UOBYQA and
Condor's main interface is the XML one and its main application is the
minimization of black-box functions with no available derivatives. This AMPL
installation is only for demonstration purposes. A fully functional XML version
is available at the
Condor was developed by
Frank Vanden Berghen
This solver was implemented by
Hans Mittelmann and executes at
Using the NEOS Server for Condor
The user must submit a model in
Examples of such models can be found
The model is specified by a model file, and optionally, a data file and a
If the command file is specified it must contain the AMPL
The commands file can contain any AMPL command or set
options for Condor, by setting the variable 'condor_options'.
These are the available options.