施密特正交化

Gram-Schmidt Orthogonalization

Gram-Schmidt算法

算法推导

算法实现

import numpy as np
from decimal import Decimal
import math


def Hadamard(v):
    n = len(v)
    detL = abs(np.linalg.det(v))
    detL = Decimal(str(detL))
    product = 1
    for i in range(n):
        product *= np.linalg.norm(v[i])
    product = Decimal(str(product))
    ab = detL / product
    res = math.pow(ab, 1 / n)
    return res


"""
随机生成一个v*v矩阵的优质基，
上限为N，
Hadamard下限是h
"""


def random_basis(N, v, h):
    res = np.random.randint(-N, N + 1, (v, v))
    while Hadamard(res) < h:
        res = np.random.randint(-N, N + 1, (v, v))
    print(Hadamard(res))
    return res


test_1 = np.array([[11, 0], [0, 1]])


def orthogonal(m):
    n = np.shape(m)
    # print(n)
    M = np.zeros(n)
    n = n[0]
    M[0, :] = m[0, :]
    for i in range(1, n):
        M[i, :] = m[i, :]
        for j in range(0, i):
            u_ij = np.dot(m[i, :], M[j, :]) / (np.linalg.norm(M[j, :]) ** 2)
            M[i, :] -= u_ij * M[j, :]
    print('H:'+str(Hadamard(M)))
    return M


# print(random_basis(200,3,0.8))
print(orthogonal(test_1))
print()