[高级编程技术作业-Week 11]NumPy相关习题练习

代码：

import numpy as np
from scipy.linalg import toeplitz
from scipy.linalg import svdvals
import time

n = 200
m = 500

A = np.random.normal(size = (n, m))
print(A)
B = toeplitz(range(m))
print(B)

#9.1
print('\nEx 9.1')
def f(A, B):
    w = np.linalg.eigvals(B)
    C = B - w[0] * np.eye(m)
    return np.dot(A, C)

print(A+A)
print(np.dot(A, A.T))
print(np.dot(A.T, A))
print(np.dot(A, B))
print(f(A, B))

#9.2
print('\nEx 9.2')
b = np.random.random(m)
print(np.linalg.solve(B, b))

#9.3
print('\nEx 9.3')
print(np.linalg.norm(A, ord = 'fro'))
print(np.linalg.norm(B, ord = np.inf))
u, s, vh = np.linalg.svd(B)
print(max(s), min(s))

#9.4
print('\nEx 9.4')
'''
borrowed from Wikipedia and modified
'''
def power_iteration(A, num_simulations = 100):
    # Ideally choose a random vector
    # To decrease the chance that our vector
    # Is orthogonal to the eigenvector
    # num_simulation for the maximum possible iteration count(if not converge)
    b_k = np.random.rand(A.shape[1])
    b_k1 = np.zeros(A.shape[1])
    b_k1_norm = 0
    for iter_count in range(num_simulations):
        # calculate the matrix-by-vector product Ab
        b_k1 = np.dot(A, b_k)
        # calculate the norm
        b_k2_norm = b_k1_norm #store the last norm
        b_k1_norm = np.linalg.norm(b_k1)
        # re-normalize the vector
        b_k = b_k1 / b_k1_norm
        # convergence judgement
        if(abs(b_k1_norm - b_k2_norm) < 0.000005):
            break
    return b_k1_norm, b_k, iter_count+1
    
Z = np.random.normal(size = (n, n))
start = time.clock()
val, vec, i = power_iteration(Z)
finish = time.clock()
print("Largest eigenvalue:\n", val)
print("Corresponding eigenvector:\n", vec)
print("Iteration count for convergence: (100 indicates misconvergence)\n", i)
print("Time for iterations:\n", finish - start)

#9.5
print('\nEx 9.5')
C = np.random.randint(0, 2, (n, n))
s = svdvals(B)
print(s)

#9.6
print('\nEx 9.6')
z = np.random.normal(size = 1)
print(z)
i = np.argmin(abs(A-z))
print(A[i//m][i%m])

总结：涉及的函数

NumPy：

    eye()

    zeros()

    dot()

    argmin()

    numpy.random

        random()

        normal()

        rand()

        randint()

    numpy.linalg

        eig()

        eigvals()

        solve()

        norm()

        svd()

SciPy

    scipy.linalg

        toeplitz()

        svdvals()