有没有更有效的方法在numpy中生成距离矩阵？

import numpy as np
from scipy.spatial.distance import cdist

# 生成两个集合
set1 = np.array([[1, 2], [3, 4], [5, 6]])
set2 = np.array([[7, 8], [9, 10]])

# 计算欧氏距离矩阵
dist_matrix = cdist(set1, set2, metric='euclidean')
print(dist_matrix)

import numpy as np

# 生成两个集合
set1 = np.array([[1, 2], [3, 4], [5, 6]])
set2 = np.array([[7, 8], [9, 10]])

# 计算欧氏距离矩阵
dist_matrix = np.sqrt(np.sum((set1[:, np.newaxis] - set2) ** 2, axis=2))
print(dist_matrix)

import numpy as np

# 生成两个集合
set1 = np.array([[1, 2], [3, 4], [5, 6]])
set2 = np.array([[7, 8], [9, 10]])

# 计算欧氏距离矩阵
dist_matrix = np.linalg.norm(set1[:, np.newaxis] - set2, axis=2)
print(dist_matrix)