Scipy练习

Exercise 10.1: Least squares

---

题目描述

Generate matrix A belongs to Rm*n with m > n. Also generate some vector b belongs to Rm.
Now x = arg min ||Ax -b||2
Print the norm of the residual

解题思路

• 构造函数 A*x - b = Y。
• 生成正太分布随机数作为参数点。
• 运用Python 的 leastsq 函数实现最小二乘。
  
  

解题代码

import numpy as np
from scipy.optimize import leastsq    

def err(x, A, b):  
    return A * x - b

x = 1 
m = 100  
A = np.array([np.random.random() * 10 for i in range(m)])  
b = np.array([np.random.random() * 10 for i in range(m)])   
result = leastsq(err, x, args = (A, b))  
print(result[0])  

Exercise 10.2: Optimization

---

题目描述

Find the maximum of the function

                                    f(x) = sin2(x - 2)e^-x2

解题思路

• 构造函数 f(x) = sin2(x - 2)e^-x2 。
• 利用 Python 的 minimize_scalar 函数解方程 。

解题代码

import numpy as np  
from math import sin, exp 
from scipy.optimize import minimize_scalar   

def f(x):  
    return ((-1) * (sin(x - 2))**2) * exp(-1 * (x**2))  

result = minimize_scalar(f)  
print((-1) * result.fun)

Exercise 10.3: Pairwise distances

---

题目描述

 Let X be a matrix with n rows and m columns. 
 How can you compute the pairwise distances between every two rows?
 As an example application, consider n cities, 
 and we are given their coordinates in two columns. 
 Now we want a nice table that tells us for each two cities, how far they are apart.
 Again, make sure you make use of Scipy's functionality instead of writing your own routine.

解题思路

• 运用Python 的 dist.cdist 函数

解题代码

import numpy as np
import scipy.spatial.distance as dist

n, m = 6, 3
X = np.random.randint(5, 20, size = (n, m))
D = dist.cdist(X, X)
print(D)