python实现:
import sympy
import numpy as np
#f为要求极值的函数,x0为初始位置,max_iter为最大迭代次数,epsilon为相邻两次迭代的x改变量
def newton_x(f, x0, max_iter, epsilon):
i = 0 #记录迭代次数的变量
x0 = float(x0) #浮点数计算更快
df = sympy.diff(f, x) #定义一阶导数
d2f = sympy.diff(f, x, 2) #定义二阶导数
while i < max_iter:
gk = df.subs(x, x0)
Gk = d2f.subs(x, x0)
xnew = x0 - gk/Gk
i += 1
print('迭代第%d次:%.5f' %(i, xnew))
if abs(df.subs(x, xnew)-df.subs(x, x0)) < epsilon:
break
x0 = xnew
return xnew
#f为要求极值的函数,X0为初始位置,max_iter为最大迭代次数,epsilon为相邻两次迭代的x改变量
def newton_x0x1(f, X0, max_iter, epsilon):
i = 0 #记录迭代次数的变量
X0[0], X0[1] = float(X0[0]), float(X0[1]) #浮点数计算更快
df0 = sympy.diff(f, x0) #定义一阶导数
df1 = sympy.diff(f, x1)
d2f0 = sympy.diff(f, x0, 2) #定义二阶导数
d2f1 = sympy.diff(f, x1, 2)
df0df1 = sympy.diff(sympy.diff(f, x0), x1)
while i < max_iter:
gk = np.mat([float(df0.subs([(x0, X0[0]), (x1, X0[1])])), float(df1.subs([(x0, X0[0]), (x1, X0[1])]))]).T #梯度矩阵
Gk = np.mat([[float(d2f0.subs([(x0, X0[0]), (x1, X0[1])])), float(df0df1.subs([(x0, X0[0]), (x1, X0[1])]))],
[float(df0df1.subs([(x0, X0[0]), (x1, X0[1])])), float(d2f1.subs([(x0, X0[0]), (x1, X0[1])]))]]) #海塞矩阵
dk = -Gk.I*gk
Xnew = [X0[0] + dk[0,0], X0[1] + dk[1,0]]
i += 1
print('迭代第%d次:[%.5f, %.5f]' %(i, Xnew[0], Xnew[1]))
if abs(f.subs([(x0, Xnew[0]), (x1, Xnew[1])])-f.subs([(x0, X0[0]), (x1, X0[1])])) < epsilon:
break
X0 = Xnew
return Xnew
if __name__ == '__main__':
x = sympy.symbols("x")
x0 = sympy.symbols("x0")
x1 = sympy.symbols("x1")
result = newton_x(x**4-4*x, 10, 50000, 1e-5)
print('最佳迭代的位置:%.5f' %result)
result = newton_x0x1((x0-1)**2+(x1-1)**4, [10,10], 50000, 1e-5)
print('最佳迭代位置:[%.5f, %.5f]' %(result[0], result[1]))
C++实现:
#include#include #include const double dx = 1e-3; double f(double x) { return pow(x, 4) - 4 * x; } double df(double x) { //return 4 * pow(x, 3) - 4; return (f(x + dx) - f(x)) / dx; } double d2f(double x) { //return 12 * pow(x, 2); return (df(x + dx) - df(x)) / dx; } double f(std::vector X) { return pow(X[0] - 1, 2) + pow(X[1] - 1, 4); } double df0(std::vector X) { //return 2 * (X[0] - 1); return (f({ X[0] + dx, X[1] }) - f(X)) / dx; } double df1(std::vector X) { //return 4 * pow(X[1] - 1, 3); return (f({ X[0], X[1] + dx }) - f(X)) / dx; } double d2f0(std::vector X) { //return 2; return (df0({ X[0] + dx, X[1] }) - df0(X)) / dx; } double d2f1(std::vector X) { //return 12 * pow(X[1] - 1, 2); return (df1({ X[0] , X[1] + dx }) - df1(X)) / dx; } double df0df1(std::vector X) { //return 0; return (df1({ X[0] + dx, X[1] }) - df1(X)) / dx; } double newton_x(double x0, int max_iter, double epsilon) { int i = 0; double xnew; while (i < max_iter) { double gk = df(x0); double Gk = d2f(x0); xnew = x0 - gk/Gk; ++i; std::cout << "迭代次数:" << i << " " << x0 << std::endl; if (abs(f(xnew) - f(x0)) < epsilon) break; x0 = xnew; } return xnew; } std::vector newton_x0x1(std::vector X0, int max_iter, double epsilon) { int i = 0; std::vector Xnew; while (i < max_iter) { Eigen::Vector2f gk; gk << df0(X0), df1(X0); Eigen::Matrix2f Gk; Gk << d2f0(X0), df0df1(X0), df0df1(X0), d2f1(X0); Eigen::Vector2f dk = -Gk.inverse()*gk; Xnew = { X0[0] + dk(0), X0[1] + dk(1) }; ++i; std::cout << "迭代次数:" << i << " " << X0[0] << " " << X0[1] << std::endl; if (abs(f(Xnew) - f(X0)) < epsilon) break; X0 = Xnew; } return X0; } int main(int argc, char* argv[]) { double result = newton_x(10, 50000, 1e-5); std::cout << "最佳迭代位置:" << result << std::endl; std::vector results = newton_x0x1({ 10,10 }, 50000, 1e-5); std::cout << "最佳迭代位置:" << results[0] << " " << results[1] << std::endl; system("pause"); return EXIT_SUCCESS; }



