高級編程技術，第十三週

10.1

# -*- coding: UTF-8 -*-
import numpy as np
import scipy.optimize as opt


A = np.random.rand(20, 10)
b = np.random.rand(20, 1)


def err(p,A,b):
	x = p.reshape(10,1)    #沒有這一步擬合結果會有問題,因爲leastsq傳入的p會從mat變成ndarray,之後不滿足矩陣運算的操作會出現一些py自定義的結果
	return (np.dot(A,x) - b).reshape(-1)
	
p0 = np.random.rand(10,1)
para = opt.leastsq(err,p0,(A,b))
x1 = para[0]
print("the answer:",x1)
print("the norm:",np.linalg.norm(err(x1, A, b)))

10.2

import numpy as np
import scipy.optimize as opt

def f(x):
    return - np.sin(x - 2) ** 2 * np.e ** ( - x ** 2)
   
para = opt.minimize(f,0)
print(-para.fun)

10.3

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

A = np.random.randint(0,2,size=(5,5))

para = dis.pdist(A)
print(A)
print(dis.squareform(para))