public static X GetMin(Function f, X x, double Lambda, double eps)
{
int k = 0;
Console.WriteLine(new String('-', 84));
Console.WriteLine("{0,55}{1,30}", "Fast Gradient Descent", "|");
Console.WriteLine(new String('-', 84) + "|");
Console.WriteLine(
"{0,3} |{1,20} |{2,10} |{3,10} |{4,20} |{5,10} |",
"k", "x", "y", "Lambda", "G", "Norma");
Console.WriteLine(
"{0}{2}{1}{1}{2}{1}",
"----|", "-----------|", "---------------------|");
while (f.Norma(x) > eps)
{
Console.WriteLine(
"{0,3} |{1,20} |{2,10:0.0000} |{3,10:0.00} |{4,20} |{5,10:0.000} |",
k, x, f.F(x), Lambda, new X(f.Fdx1(x), f.Fdx2(x)), f.Norma(x));
Lambda = Dichotomy.GetMin((lambda) => f.F(new X(x.X1 - lambda * f.Fdx1(x), x.X2 - lambda * f.Fdx2(x))), -100, 100, 0.1);
x = new X(x.X1 - Lambda * f.Fdx1(x), x.X2 - Lambda * f.Fdx2(x));
k++;
}
Console.WriteLine(new String('-', 84));
return(x);
}
public static X GetMin(Function f, X x0, double Lambda, double eps)
{
X x = x0;
X x_prev = x;
int k = 0;
X d = new X(0, 0);
double Betta = 0;
Console.WriteLine(new String('-', 84));
Console.WriteLine("{0,55}{1,30}", "Nonlinear Conjugate Dradient Method", "|");
Console.WriteLine(new String('-', 84) + "|");
Console.WriteLine(
"{0,3} |{1,20} |{2,10} |{3,10} |{4,20} |{5,10} |",
"k", "x", "y", "Lambda", "G", "Norma");
Console.WriteLine(
"{0}{2}{1}{1}{2}{1}",
"----|", "-----------|", "---------------------|");
while (f.Norma(x) > eps)
{
Console.WriteLine(
"{0,3} |{1,20} |{2,10:0.0000} |{3,10:0.00} |{4,20} |{5,10:0.000} |",
k, x, f.F(x), Lambda, new X(f.Fdx1(x), f.Fdx2(x)), f.Norma(x));
if (k == 0)
{
//5
d = new X(-f.Fdx1(x0), -f.Fdx2(x0));
}
else
{
{
Betta = Math.Pow(f.Norma(x), 2) / Math.Pow(f.Norma(x_prev), 2);
}
}
//7
d = new X(-f.Fdx1(x) + Betta * d.X1, -f.Fdx2(x) + Betta * d.X2);
//8
Lambda = Dichotomy.GetMin((lambda) => f.F(new X(x.X1 + lambda * d.X1, x.X2 + lambda * d.X2)), -100, 100, 0.1);
//9
x_prev = x;
x = new X(x.X1 + Lambda * d.X1, x.X2 + Lambda * d.X2);
//10
k++;
}
Console.WriteLine(new String('-', 84));
return(x);
}
