Introduction

SAS Optimization includes both general optimization (OPTLP, OPTMILP, OPTQP, and CLP) and network optimization (OPTNETWORK) procedures for solving various optimization problems. Optimization solvers are provided for linear (LP), mixed integer linear (MILP), quadratic (QP), conic, nonlinear (NLP), mixed integer nonlinear (MINLP), network, and black-box optimization problems. Constraint logic programming (CLP) is also supported. Two decomposition algorithms (Dantzig-Wolfe and Benders) can exploit several different types of structures in LP and MILP problems to decrease solution time. The network solver includes algorithms that solve optimization problems such as shortest path and minimum cost network flow; it also provides algorithms that analyze network structure, including connected components and cycle detection. The OPTMODEL procedure builds optimization models and calls the various solvers.

Optimization models are built in SAS, and solvers are executed to provide solutions to specific problems. The specialized procedures accept input tables that specify a problem (LP, MILP, QP, network, and CLP) and invoke the corresponding solver.

Note: When you license SAS Optimization, you also have access to SAS/OR software. For more information about SAS/OR procedures, see the SAS/OR documentation.