Chapter Contents |
Previous |
Next |

SAS/OR Software |

In Version 7, PROC NETFLOW includes another class of optimization algorithm, the Interior-Point algorithm, which can be used as an alternative to the Simplex algorithm to solve network problems. The Interior-Point algorithm often solves network problems and linear programming (LP) problems much faster than the Simplex algorithm, and it can solve problems that are very large. LP problems that have millions of variables can be solved on a PC in a couple of hours.

To use the Interior-Point algorithm, specify the INTPOINT option in the PROC NETFLOW statement. If PROC NETFLOW detects that there are no arcs and nodes in the model's data (that is, that there is no network component), then it assumes that it is dealing with an LP problem, and it automatically uses the Interior-Point algorithm to perform the optimization. Previously, these problems could be solved only by PROC LP. Data for an LP problem resembles the data for side constraints and nonarc variables that is supplied to PROC NETFLOW when a constrained network problem is solved. It is also very similar to the data that PROC LP requires.

For more information, see the NETFLOW chapter in *SAS/OR User's Guide:
Mathematical Programming*. The "Interior-Point Algorithm" section
describes how to use the algorithm and also discusses the theory on
which the algorithm is based.

Chapter Contents |
Previous |
Next |
Top of Page |

Copyright © 1999 by SAS Institute Inc., Cary, NC, USA. All rights reserved.