统计学习作业2：scipy求有约束优化问题

Scipy是一个用于数学、科学、工程领域的常用软件包,可以处理插值、积分、优化、图像处理、常微分方程

问题

针对于有约束问题

python的scipy库给出了Sequential Least SQuares Programming (SLSQP) Algorithm。

实例

• SLSQP解决模板

对于下面的问题：

>>> from scipy.optimize import minimize
>>> import numpy as np

# 定义目标函数
>>> def rosen(x):
...     """The Rosenbrock function"""
...     return sum(100.0*(x[1:]-x[:-1]**2.0)**2.0 + (1-x[:-1])**2.0)

# 约束项
# 变量范围约束
>>> from scipy.optimize import Bounds
>>> bounds = Bounds([0, -0.5], [1.0, 2.0])
# 不等式约束
>>> ineq_cons = {'type': 'ineq',
...              'fun' : lambda x: np.array([1 - x[0] - 2*x[1],
...                                          1 - x[0]**2 - x[1],
...                                          1 - x[0]**2 + x[1]]),
# jac 是fun的梯度矩阵
...              'jac' : lambda x: np.array([[-1.0, -2.0],
...                                          [-2*x[0], -1.0],
...                                          [-2*x[0], 1.0]])}

>>> eq_cons = {'type': 'eq',
...            'fun' : lambda x: np.array([2*x[0] + x[1] - 1]),
...            'jac' : lambda x: np.array([2.0, 1.0])}

# 初始值
>>> x0 = np.array([0.5, 0])
>>> res = minimize(rosen, x0, method='SLSQP', jac=rosen_der,
...                constraints=[eq_cons, ineq_cons], options={'ftol': 1e-9, 'disp': True},
...                bounds=bounds)
# may vary
Optimization terminated successfully.    (Exit mode 0)
            Current function value: 0.342717574857755
            Iterations: 5
            Function evaluations: 6
            Gradient evaluations: 5
>>> print(res.x)

例题

# -*- coding: utf-8 -*-
"""
Created on Sun Sep 22 20:14:59 2019

@author: Mingxing Li
"""

from scipy.optimize import minimize
from scipy.optimize import Bounds
import numpy as np

def fun(x):
    return 10-x[1:]**2.0-x[:-1]**2.0

# 约束项
# 不等式约束
ineq_cons = {"type": "ineq",
             "fun": lambda x:np.array([-x[0]**2+x[1]])}
                     
# 等式约束
eq_cons = {"type": "eq",
           "fun": lambda x: np.array([x[0]+x[1]])}

# 初始值
x0 = [-1, 0]
res = minimize(fun, x0, method='SLSQP', 
                    constraints=[eq_cons, ineq_cons], options={'ftol': 1e-9, 'disp': True})


print(res)

运行结果

Optimization terminated successfully.    (Exit mode 0)
            Current function value: 8.0
            Iterations: 2
            Function evaluations: 8
            Gradient evaluations: 2
     fun: 8.0
     jac: array([ 2., -2.])
 message: 'Optimization terminated successfully.'
    nfev: 8
     nit: 2
    njev: 2
  status: 0
 success: True
       x: array([-1.,  1.])

需要注意，这个方法貌似对初值很敏感。。。