ISmQuadraticProgramming.PerformanceInfo

Syntax

Description

The PerformanceInfo property returns information about executed optimization.

Comments

To determine parameters of linear constraints, use the ISmQuadraticProgramming.LinearConstraints property.

Example

To execute the example, add a link to the Stat system assembly.

Sub UserProc;
Var
    QP: SmQuadraticProgramming;
    i, j, Res, N: Integer;
    d, time: double;
    CF, Lb, Ub, init: Array Of Double;
    LinC1: Array Of Double;
    H, A: Array Of Double;
    Bound: ISlBoundaryRegion;
    LinCons: ISlLinearConstraints;
    LCon1: ISlLinearConstraint;
    InitP, Sol, BoundL, PerformanceObj, PerformancePoint: Array Of Double;
Begin
    QP := New SmQuadraticProgramming.Create;
    N := 7;
    CF := New Double[N];
    Lb := New Double[N];
    Ub := New Double[N];
    LinC1 := New Double[N];
    init := New double[N];
    // Definition area
    Lb[0] := -0.01; Ub[0] := 0.01;
    Lb[1] := -0.10; Ub[1] := 0.15;
    Lb[2] := -0.01; Ub[2] := 0.03;
    Lb[3] := -0.04; Ub[3] := 0.02;
    Lb[4] := -0.10; Ub[4] := 0.05;
    Lb[5] := -0.01; Ub[5] := double.PositiveInfinity;
    Lb[6] := -0.01; Ub[6] := Double.PositiveInfinity;
    // Initial approximations
    init[0] := -0.01;
    init[1] := -0.03;
    init[2] := 0.0;
    init[3] := -0.01;
    init[4] := -0.1;
    init[5] := 0.02;
    init[6] := 0.01;
    QP.InitialApproximation.AutoCreate := False;
    QP.InitialApproximation.InitValues := init;
    // Criterion function
    CF[0] := -0.02;
    CF[1] := -0.2;
    CF[2] := -0.2;
    CF[3] := -0.2;
    CF[4] := -0.2;
    CF[5] := 0.04;
    CF[6] := 0.04;

    H := New double[N, N];
    H[0, 0] := 2; H[0, 1] := 0; H[0, 2] := 0; H[0, 3] := 0; H[0, 4] := 0; H[0, 5] := 0; H[0, 6] := 0;
    H[1, 0] := 0; H[1, 1] := 2; H[1, 2] := 0; H[1, 3] := 0; H[1, 4] := 0; H[1, 5] := 0; H[1, 6] := 0;
    H[2, 0] := 0; H[2, 1] := 0; H[2, 2] := 2; H[2, 3] := 2; H[2, 4] := 0; H[2, 5] := 0; H[2, 6] := 0;
    H[3, 0] := 0; H[3, 1] := 0; H[3, 2] := 2; H[3, 3] := 2; H[3, 4] := 0; H[3, 5] := 0; H[3, 6] := 0;
    H[4, 0] := 0; H[4, 1] := 0; H[4, 2] := 0; H[4, 3] := 0; H[4, 4] := 2; H[4, 5] := 0; H[4, 6] := 0;
    H[5, 0] := 0; H[5, 1] := 0; H[5, 2] := 0; H[5, 3] := 0; H[5, 4] := 0; H[5, 5] := -2; H[5, 6] := -2;
    H[6, 0] := 0; H[6, 1] := 0; H[6, 2] := 0; H[6, 3] := 0; H[6, 4] := 0; H[6, 5] := -2; H[6, 6] := -2;
    QP.CriterionFunction := CF;
    QP.QuadraticForm := H;
    // Linear constraints
    A := New double[N, N];
    A[0, 0] := 1.00; A[0, 1] := 1.00; A[0, 2] := 1.00;
    A[0, 3] := 1.00; A[0, 4] := 1.00; A[0, 5] := 1.00; A[0, 6] := 1.00;
    A[1, 0] := 0.15; A[1, 1] := 0.04; A[1, 2] := 0.02;
    A[1, 3] := 0.04; A[1, 4] := 0.02; A[1, 5] := 0.01; A[1, 6] := 0.03;
    A[2, 0] := 0.03; A[2, 1] := 0.05; A[2, 2] := 0.08;
    A[2, 3] := 0.02; A[2, 4] := 0.06; A[2, 5] := 0.01; A[2, 6] := 0.00;
    A[3, 0] := 0.02; A[3, 1] := 0.04; A[3, 2] := 0.01;
    A[3, 3] := 0.02; A[3, 4] := 0.02; A[3, 5] := 0.00; A[3, 6] := 0.00;
    A[4, 0] := 0.02; A[4, 1] := 0.03; A[4, 2] := 0.00;
    A[4, 3] := 0.00; A[4, 4] := 0.01; A[4, 5] := 0.00; A[4, 6] := 0.00;
    A[5, 0] := 0.70; A[5, 1] := 0.75; A[5, 2] := 0.80;
    A[5, 3] := 0.75; A[5, 4] := 0.80; A[5, 5] := 0.97; A[5, 6] := 0.00;
    A[6, 0] := 0.02; A[6, 1] := 0.06; A[6, 2] := 0.08;
    A[6, 3] := 0.12; A[6, 4] := 0.02; A[6, 5] := 0.01; A[6, 6] := 0.97;

    // Area borders
    Bound := QP.Boundary;
    // Lower area border
    Bound.BoundaryLower := Lb;
    // Upper area border
    Bound.BoundaryUpper := Ub;
    // Linear constraint coefficient
    LCon1 := QP.LinearConstraints.Add;
    For i := 0 To N - 1 Do
        LinC1[i] := A[0, i];
    End For;
    LCon1.Value := LinC1;
    // Lower linear constraint
    LCon1.BoundaryLower := -0.13;
    // Upper linear constraint
    LCon1.BoundaryUpper := -0.13;
    // Linear constraint coefficient
    LCon1 := QP.LinearConstraints.Add;
    For i := 0 To N - 1 Do
        LinC1[i] := A[1,i];
    End For;
    LCon1.Value := LinC1;
    // Lower linear constraint
    LCon1.BoundaryLower := double.NegativeInfinity;
    // Upper linear constraint
    LCon1.BoundaryUpper := -0.0049;
    // Linear constraint coefficient
    LCon1 := QP.LinearConstraints.Add;
    For i := 0 To N - 1 Do
        LinC1[i] := A[2, i];
    End For;
    LCon1.Value := LinC1;

    // Lower linear constraint
    LCon1.BoundaryLower := double.NegativeInfinity;
    // Upper linear constraint
    LCon1.BoundaryUpper := -0.0064;
    // Linear constraint coefficient
    LCon1 := QP.LinearConstraints.Add;
    For i := 0 To N - 1 Do
        LinC1[i] := A[3, i];
    End For;
    LCon1.Value := LinC1;
    // Lower linear constraint
    LCon1.BoundaryLower := double.NegativeInfinity;
    // Upper linear constraint
    LCon1.BoundaryUpper := -0.0037;
    // Linear constraint coefficient
    LCon1 := QP.LinearConstraints.Add;
    For i := 0 To N - 1 Do
        LinC1[i] := A[4, i];
    End For;
    LCon1.Value := LinC1;
    // Lower linear constraint
    LCon1.BoundaryLower := double.NegativeInfinity;
    // Upper linear constraint
    LCon1.BoundaryUpper := -0.0012;
    // Linear constraint coefficient
    LCon1 := QP.LinearConstraints.Add;
    For i := 0 To N - 1 Do
        LinC1[i] := A[5, i];
    End For;
    LCon1.Value := LinC1;
    // Lower linear constraint
    LCon1.BoundaryLower := -0.0992;
    // Upper linear constraint
    LCon1.BoundaryUpper := double.PositiveInfinity;
    // Linear constraint coefficient
    LCon1 := QP.LinearConstraints.Add;
    For i := 0 To N - 1 Do
        LinC1[i] := A[6, i];
    End For;
    LCon1.Value := LinC1;
    // Lower linear constraint
    LCon1.BoundaryLower := -0.003;
    // Upper linear constraint
    LCon1.BoundaryUpper := 0.002;
    // Calculation and solution output
    Res := QP.Execute;
    time := QP.PerformanceTime / 1000;
    Debug.WriteLine("Execution time, sec " + time.ToString);
    Debug.WriteLine("Execution status: " + Res.ToString);
    Debug.WriteLine(QP.Errors);

    If res = 0 Then
        Debug.WriteLine("== Criterion function value ==");
        Debug.WriteLine(QP.OptimalFunctionValue.ToString);
        Debug.WriteLine("Number of criterion function calls: "
        + qp.PerformanceInfo.ObjFunCalls.ToString);
        Debug.WriteLine("Number of constraints calls: "
        + qp.PerformanceInfo.ConstraintsCalls.ToString);
        Debug.WriteLine("Actual number of iterations: "
        + qp.PerformanceInfo.Iterations.ToString);
        Debug.WriteLine("== Solution ==");
        Sol := QP.Solution;
        For i := 0 To Sol.Length - 1 Do
            Debug.WriteLine(i.ToString + ": " + Sol[i].ToString);
        End For;
        Debug.WriteLine("== Common constraints: ==");
        d := 0;
        BoundL := Bound.LagrangeMultiplier;
        For i := 0 To N - 1 Do //
            d := Sol[i];
            Debug.WriteLine(i.ToString + ": ");
            Debug.WriteLine(Lb[i].ToString + "<=" + d.ToString +
            "<= " + Ub[i].ToString);
            Debug.WriteLine("Lagrange multiplier: " + BoundL[i].ToString);
            Debug.WriteLine("  ");
        End For;
        Debug.WriteLine("   ");
        Debug.WriteLine("== Linear constraints ==");
        LinCons := QP.LinearConstraints;
        For i := 0 To LinCons.Count - 1 Do
            LCon1 := LinCons.Item(i);
            Debug.WriteLine(i.ToString + ": " + LCon1.BoundaryLower.ToString + " <= "
            + LCon1.Result.ToString + " <= " + LCon1.BoundaryUpper.ToString);
            Debug.WriteLine("Lagrange multiplier: " + LCon1.LagrangeMultiplier.ToString);
            Debug.WriteLine("  ");
        End For;
    End If;
    Debug.WriteLine("== Initial approximation ==");
    PerformancePoint := QP.PerformanceInfo.InitFeasiblePoint;
    For i := 0 To PerformancePoint.Length - 1 Do
        Debug.WriteLine(i.ToString + ": " + PerformancePoint[i].ToString);
    End For;
End Sub UserProc;

After executing the example the console window displays:

If there are any calculation errors.
Optimal and reference value.
Information about executed optimization.
Criterion function values.
Constraints and Lagrange multipliers.