Function MasterMIPLinearizationCommand(n, ModelConstraint, Command, CommandData)

MasterMIPLinearizationCommand

The procedure MasterMIPLinearizationCommand allows you to retrieve or modify certain aspects of the linearization of a constraint added for linearization counter n at the individual level. The argument Command specifies which data (e.g. GetDeviation) should be retrieved or modified. The retrieved or modified value is passed through the CommandData argument.

MasterMIPLinearizationCommand(
     n,                ! (input) integer scalar value
     ModelConstraint,  ! (input) scalar value
     Command,          ! (input) element parameter into
                       !         MasterMIPLinearizationCommands
     CommandData       ! (inout) scalar value (in) or parameter (out)
     )

Arguments

n

The linearization counter as returned by MasterMIPAddLinearizations when adding this linearization.

ModelConstraint

Scalar reference to a constraint for which certain aspects of the linearization have to be retrieved or modified.

Command

Element parameter into MasterMIPLinearizationCommands that specifies which data should be retrieved or modified. Possible values are:

Command	Description
GetDeviation	Get the value of the deviation variable.
RemoveDeviation	Delete the deviation variable.
GetWeight	Get the objective coefficient of the deviation variable.
SetWeight	Set the objective coefficient of the deviation variable.
GetDeviationBound	Get the upper bound of the deviation variable.
SetDeviationBound	Get the upper bound of the deviation variable.
GetLagrangeMultiplier	Get value of the shadow price (Lagrange multiplier) of constraint for last solved NLP.

CommandData

The retrieved or modified value.

Return Value

MasterMIPLinearizationCommand has no return value.

Note

• Normally, the weight obtained with ‘GetWeight’ equals the value of the penalty multiplier, as passed to MasterMIPAddLinearizations, times the shadow price (Lagrange multiplier) of the constraint. With ‘SetWeight’ this weight can be changed.

• Note that ‘SetWeight’ can be used to create a deviation variable (slack) if the linearization does not have one. To do so the value filled in for CommandData should be unequal to 0.

• The lower bound of a deviation variable always equals 0.