Chapter Contents |
Previous |
Next |

The NLP Procedure |

**DECVAR***name_list [=numbers] [, name_list [=numbers] ...]***;**

**VAR***name_list [=numbers] [, name_list [=numbers] ...]***;**

**PARMS***name_list [=numbers] [, name_list [=numbers] ...]***;**

**PARAMETERS***name_list [=numbers] [, name_list [=numbers] ...]***;**

The DECVAR statement lists the names of the *n* > 0 decision variables and
specifies grid search and initial values for an iterative
optimization process.
The decision variables listed in the DECVAR statement cannot also be
used in the MIN, MAX, MINQUAD, MAXQUAD, LSQ, GRADIENT,
HESSIAN, JACOBIAN, CRPJAC, or NLINCON statement.

The DECVAR statement contains a list of decision variable names (not separated by commas) optionally followed by an equal sign and a list of numbers. If the number list consists of only one number, this number defines the initial value for all the decision variables listed to the left of the equal sign.

If the number list consists of more than one number, these numbers specify the grid locations for each of the decision variables listed left of the equal sign. You can use the TO and BY keywords to specify a number list for a grid search can be used. When a grid of points is specified with a DECVAR statement, PROC NLP computes the objective function value at each grid point and chooses the best (feasible) grid point as a starting point for the optimization process. The use of the BEST= option is recommended to save computing time and memory for the storing and sorting of all grid point information. Usually only feasible grid points are included in the grid search. If the specified grid contains points located outside the feasible region and you are interested in the function values at those points, it is possible to use the INFEASIBLE option to compute (and display) their function values as well.

Chapter Contents |
Previous |
Next |
Top |

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