1 # coding=utf-8
2
3 #共轭梯度算法求最小值
4 import numpy as np
5
6 from scipy
import optimize
7
8
9
10
11 def f(x, *
args):
12 u, v =
x
13 a, b, c, d, e, f,g,h =
args
14 return a*u**g+ b*u*v + c*v**h + d*u + e*v +
f
15
16
17 def gradf(x, *
args):
18 u, v =
x
19 a, b, c, d, e, f,g,h =
args
20 gu = g*a*u + b*v + d
# u-component of the gradient
21 gv = b*u +h*c*v + e
# v-component of the gradient
22 return np.asarray((gu, gv))
23
24
25
26
27 if __name__ ==
"__main__":
28 args = (2, 3, 7, 8, 9, 10, 2, 2)
# parameter values
29 x0 = np.asarray((0, 0))
# Initial guess.
30 res1 = optimize.fmin_cg(f, x0, fprime=gradf, args=
args)
31 #res2 = optimize.fmin(f,x0,args)
32 print(res1)
33 #print(res2)
34 print(f(res1,*args))
转载于:https://www.cnblogs.com/shizhenqiang/p/8184411.html
相关资源：目标函数极值求解的几种方法 最速下降法，你牛顿法，共轭梯度法编程实现
