NAG C Library NAG C LibraryLinear programming, quadratic programming, mixed-integer linear and quadratic programming, minimization of a nonlinear function (unconstrained, bound-constrained, linearly constrained, and nonlinearly constrained), and minimization of a sum of squares (unconstrained, bound constrained, linearly constrained, and nonlinearly constrained). Sparse problems can be handled.
For problems with nonlinear constraints, a sequential QP algorithm is used. For unconstrained problems and problems with simple bounds, quasi-Newton and conjugate gradient methods are provided. The Nelder-Mead simplex method is provided for unconstrained problems. For mixed-integer programming, a branch and bound method is used. For minimizing a sum of squares, a Gauss-Newton method is used. The LP and QP routines use a numerically stable active-set strategy in which the linear constraint matrix may be dense or sparse. Problem data may be supplied in MPSX format. The nonlinear programming routine exploits sparsity.
A consistent option-setting mechanism is provided for all routines, in order to keep the basic parameter-list to a minimum, while allowing a large degree of flexibility in controlling the algorithm. The routines have the ability to print the solution, as well as various amounts of intermediate output to monitor the computation.
Service routines are provided for approximating first or second derivatives by finite differences, for checking user-supplied routines for first derivatives, and for computing a covariance matrix for nonlinear least squares problems.
The NAG C Library is available in tested, compiled form for a large number of different hardware and software computing environments.
Visit the NAG WWW site or contact:
NAG, Inc. NAG Ltd. 1400 Opus Place, Suite 200 Wilkinson House, Jordan Hill Road Downers Grove, IL 60515-5702 Oxford OX2 8DR USA England Phone: +1 (630) 971 2337 Phone: +44 (1865) 511245 Fax: +1 (630) 971 2706 Fax: +44 (1865) 310139 email: naginfo@nag.com email: infodesk@nag.co.uk NAG GmbH Schleissheimerstrasse 5 W-8046 Garching bei München Germany Phone: +49 (89) 320 7395 Fax: +49 (89) 320 7396 email: s380101@sunmail.lrz-muenchen.de IECJ - (IRIS Explorer Center Japan) Nagashima Building 2F 2-24-3 Higashi Shibuya-ku Tokyo Japan Phone: +81 3 5485 2901 Fax: +81 3 5485 2903 email: help@iris.explorer.co.jp NAG en France Espace III 62 Boulevard Frederic Arnaud 09200 Saint Girons Toulouse France Phone: +33 56104 8463 Fax: +33 56104 8489 email: nash@cict.fr
NAG C Library Manual, Mark 6, NAG Ltd., 2000.
[ Optimization Software Guide | OTC Home Page | NEOS Server | NEOS Guide ]