【解方程】scipy.optimize.solve.

最新推荐文章于 2024-03-22 19:45:00 发布

guofei9987

最新推荐文章于 2024-03-22 19:45:00 发布

阅读量1.6k

点赞数

分类专栏：数学

本文链接：https://blog.csdn.net/guofei9987/article/details/75050307

版权

数学专栏收录该内容

11 篇文章 1 订阅

订阅专栏

fsolve求方程的解

如果要求解方程：

⎧ ⎩ ⎨ ⎪ ⎪ f 1 (u 1, u 2, u 3) = 0 f 2 (u 1, u 2, u 3) = 0 f 3 (u 1, u 2, u 3) = 0

$\left \{ \begin{array}{ccc} f1(u1,u2,u3)=0\\ f2(u1,u2,u3)=0\\ f3(u1,u2,u3)=0 \end{array}\right.$

那么func这么定义：

def func(x):
  u1,u2,u3=x
  return [f1(u1,u2,u3),f2(u1,u2,u3),f3(u1,u2,u3)]

代码案例1：

import scipy.optimize as opt
import numpy as np
def f(x):
    x0,x1,x2=x
    return np.array([5*x1+3,4*x0*x0-2*np.sin(x1*x2),x1*x2-1.5])

result=opt.fsolve(f,[1,1,1])
print(result)
print(f(result))

代码案例2：

如果给了Jacobian矩阵，那么迭代速度更快
Jacobian矩阵的定义是：

⎡ ⎣ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ \partial f 1 \partial u 1 \partial f 2 \partial u 1 \partial f 3 \partial u 1 \partial f 1 \partial u 2 \partial f 2 \partial u 2 \partial f 3 \partial u 2 \partial f 1 \partial u 2 \partial f 2 \partial u 2 \partial f 3 \partial u 2 ⎤ ⎦ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥

$\left [ \begin{array}{ccc} \frac{\partial f1}{\partial u1}&\frac{\partial f1}{\partial u2}&\frac{\partial f1}{\partial u2}\\ \\ \frac{\partial f2}{\partial u1}&\frac{\partial f2}{\partial u2}&\frac{\partial f2}{\partial u2}\\ \\ \frac{\partial f3}{\partial u1}&\frac{\partial f3}{\partial u2}&\frac{\partial f3}{\partial u2} \end{array} \right ]$

import scipy.optimize as opt
import numpy as np
def obj_func(x):
    x0,x1,x2=x
    return [5*x1+3,4*x0*x0-2*np.sin(x1*x2),x1*x2-1.5]

def jacobian(x):
    x0, x1, x2 = x
    return [
        [0,5,0],
        [8*x0,-2*x2*np.cos(x1*x2),-2*x1*np.cos(x1*x2)],
        [0,x2,x1]
    ]
result=opt.fsolve(obj_func,[1,1,1],fprime=jacobian)
print(result)
print(obj_func(result))