Optimization (Object)

Represents an Optimization object.

Properties


Property AlgorithmType() As
Returns or sets the algorithm type. Currently available algorithms are gradient and simulatedAnnealing
See also:
Property Constraints() As (Read Only)
Returns the collection of optimization constraints.
Property ConvSpeed() As
Returns or sets the convergence speed for some gradients and the simulated annealing.
Property FreeParameters() As (Read Only)
Returns the collection of the free parameters.
Property MaxEvalsNb() As
Returns or sets the maximum number of model updates allowed during one run of the optimization.
Property MaxEvalsWoImprovement() As
Returns or sets the maximum number of model updates without improvement of the problem solution during one run of the optimization.
Property MaxTime() As
Returns or sets the maximum time allowed for one run of the optimization (in minutes).
Property ObjectiveParameter() As
Returns or sets the objective parameter of the optimization. This parameter can not exist (in this case the get_ method returns E_FAIL) when the optimization contains only constraints and uses Simulated Annealing, or if the optimization feature doesn't contain all information necessary to be run.
Property OptimizationType() As
Returns or sets the type of the optimization: minimum, maximum or target value searched on the objective parameter.
See also:
Property TargetValue() As (Read Only)
Returns the objective parameter target value. (used only if the optimization type is a target value search)
Property UseMaxEvalsWoImprovement() As
Returns or sets if the number of updates without improvement of the solution has to be used as a termination criterion.
Property UseMaxTime() As
Returns or sets if max time has to be used as a termination criterion.

Methods


Sub Run( iWithStopDialog)
Runs the optimization as it is defined. The stop dialog appears if argument is TRUE
Before running, a check is made to ensure that the optimization feature contains enough information to run the optimization. In the case where some information is missing, this method returns E_FAIL
WARNING : if argument is TRUE, the optimization is launched asynchronously, and you can not run several optimizations in this mode.