Initializing help system before first use

estimatemarginals

Purpose
Estimate better marginal values for variables and constraints for degenerate problems.
Synopsis
procedure estimatemarginals(sbvars:array(vars: set of mpvar) of real)
procedure estimatemarginals(dualslb:array(constriants: set of linctr) of real, dualsub:array(constriants: set of linctr) of real)
procedure estimatemarginals(sbvars:array(vars: set of mpvar) of real, efforlimit:integer, delta:real)
procedure estimatemarginals(dualslb:array(constriants: set of linctr) of real, dualsub:array(constriants: set of linctr) of real, effortlimit:integer)
Arguments
sbvars 
An array of reals that will be populated with the approximations for the marginal values. The approximation is carried out for the variables included in the variables set.
dualslb 
An array of reals that will be populated with the approximations for the lower bounds for the row marginal values. The approximation is carried out for the constraints in the constraints set.
dualsub 
An array of reals that will be populated with the approximations for the upper bounds for the row marginal values. The approximation is carried out for the constraints in the constraints set.
efforlimit 
Effort limit spent to approximate the effect of the move of a variable, expressed as an upper limit of simplex iterations per variable.
delta 
The size of the perturbation applied to force a movement in the variable.
Further information
1. This procedure can be used to estimate the marginal values of variables in degenerate problems. In degenerate problems, the reduced costs and row duals do not always provide a good representation of the effect on the objective when forcing a move in a variable. Also, in degenerate problems, the reduced costs and row duals may depend on the final basis found, and multiple correct alternatives might exists. This function attempts to identify better marginal values by simulating a move in the variables.
2. Prior to calling estimatemarginals, the current LP problem must have been solved to optimality and an optimal basis must be available.
3. It is important to note that the procedure provides an estimate only.
4. This procedure relies on the XPRSstrongbranch and XPRSestimaterowdualranges functions, refer to the Xpress Optimizer Reference Manual for more information.
Module
mmxprs

Parent Topic Procedures and functions

© 2001-2025 Fair Isaac Corporation. All rights reserved. This documentation is the property of Fair Isaac Corporation (“FICO”). Receipt or possession of this documentation does not convey rights to disclose, reproduce, make derivative works, use, or allow others to use it except solely for internal evaluation purposes to determine whether to purchase a license to the software described in this documentation, or as otherwise set forth in a written software license agreement between you and FICO (or a FICO affiliate). Use of this documentation and the software described in it must conform strictly to the foregoing permitted uses, and no other use is permitted.