TMtxOptimization Class

class TMtxOptimization : public TMtxComponent;

Interfaces the optimization routines.

The component can be used to find the minimum of function of several variables.

How to use TMtxOptimization component?

Drop a TMtxOptimization component on the form.
Define the number of variables and their initial values by accessing the TMtxOptimization::VariableParameters vector.
Define any additional constant parameters by accessing TMtxOptimization::ConstantParameters vector.
Define any additional constant pointer parameter by using the TMtxOptimization::SetObjects method.
Define real function (must be of TRealFunction type).
Define optimization method by accessing the TMtxOptimization::OptimizationMethod property.
Depending on optimization method you'll have to (optionally) define the gradient calculation procedure (TMtxOptimization::GradProcedure method) or gradient/Hessian matrix calculation procedure (TMtxOptimization::GradHessProcedure method). If you don't specify the GradProcedure or GradHessProcedure then the numeric approximation will be used to calculate the gradient vector and Hessian matrix. In this case you must also specify which gradient aproximation method you will use - access the TMtxOptimization::NumericGradMethod property.
Call the TMtxOptimization::Recalculate method to find the minimum of function of several variables.

Results:

TMtxOptimization::MinValue : RealFunction, evaluated at minimum.
TMtxOptimization::VariableParameters : minimum position.
TMtxOptimization::Iterations: Number of iterations needed to reach specified (TMtxOptimization::Tolerance property) minimum precision.
TMtxOptimization::InverseHess : Inverse Hessian matrix (returned only by BFGS, ConjGrad and Marquardt methods)
TMtxOptimization::StopReason: Why did the optimization algorithm stopped ?

Examples

cpp

#include "MtxVec.h"
#include "MtxVecTools.h"
#include "Optimization.h"
#include "Units.Math387.h"

using namespace Dew;
using namespace Dew::Math;
using namespace Dew::Math::Units::Math387;

// define the real function to be minimized
static double Banana(TVec* pars, TVec* consts, const DewArray<tObject*>& objConsts)
{
    return 100*Power((*pars)[1] - Power((*pars)[0], 2), 2) + Power(1 - (*pars)[0], 2);
}

void Example()
{
    TMtxOptimization* MtxOptim = new TMtxOptimization(nullptr);
    if (MtxOptim != nullptr)
    {
        // define two variables and their initial values
        MtxOptim->VariableParameters->SetIt(false, DewArray<double>{2, -1});
        // use BFGS optimization method
        MtxOptim->OptimizationMethod = TOptMethod::optBFGS;
        // tolerance for MinValue and gradient calculation
        // additional note : since we did not define the GradProc,
        // the internal numerical gradient approximation will be used
        MtxOptim->Tolerance = 2.0e-6;
        MtxOptim->GradTolerance = 2.0e-6;
        // function to be minimized
        MtxOptim->RealFunction = Banana;
        // finally, calculate minimum
        MtxOptim->Recalculate();
    }
    delete MtxOptim;
}

Constructors

Name	Description
TMtxOptimization (2)

Destructors

Name	Description
~TMtxOptimization

Properties

Name	Type	Description
AutoUpdate	bool	Automatic recalculation.
BlockAssign	bool	Block streaming of specific properties when storing only a "template".
ConstantParameters	TVec*	set/read the additional constants used in minimized function.
Dirty	bool	Becomes true after any of the properties have changed.
EditorActive	bool	Returns True, if the component editor is displayed.
GradHessProcedure	TGradHess	Defines the gradient vector and Hessian matrix calculation routine.
GradProcedure	TGrad	Defines the gradient vector calculation routine.
GradTolerance	double	The precision for numeric gradient and/or Hessian matrix calculation.
Lambda0	double	Initial lambda step used in Marquardt optimization algorithm.
MaxIterations	int	maximum number of iterations allowed for minimum search.
Name	DewString
NumericGradMethod	TNumericGradMethod	Defines which gradient numerical approximation method will be used to to evaluate gradient.
OptimizationMethod	TOptMethod	Optimization algorithm used for minimum search.
RealFunction	TRealFunction	Defines the function to be minimized.
Reference	TReferenceList*	Stores a list of components that have to be notified, when this component is destroyed.
SoftSearch	bool	Internal line search algorithm.
StopReason	TOptStopReason	Stop reason for optimization algorithm.
Tolerance	double	Precision.
VariableParameters	TVec*	Set/read the variables in minimized function.
Verbose	tStrings*	If not nil then the optimization method uses it for logging each optimization step.

Methods

Name	Description
AllowStreaming	Returns the negatated value of BlockAssign.
Assign	Assign values of all published properties from Source.
AssignTemplate	Assign values of "template-like" published properties from Source.
EditorClass
LoadFromStream	Load the component from Src stream.
LoadTemplateFromFile	Load a template from file.
LoadTemplateFromStream	Load a template from stream.
Recalculate	Triggers TMtxOptimization recalculation.
ReferenceRemoved
Reset	Reset is called after loading the data from stream or file.
SaveTemplateToFile	Save template to file.
SaveTemplateToStream	Save template to stream.
SaveToStream	Save the component to Dst stream.
SetObjects	Define any additional Object constant parameters in the RealFunction.

Fields

Name	Type	Description
InverseHess	TMtx *	Returns the inverse of Hessian matrix.
Iterations	int	Returns number of optimization algorithm iterations.
MinValue	double	Stores value of the objective function, evaluated at minimum.
Tag	intptr_t