Numpy dot 在標準化后返回不同的值。我有兩個函數應該返回相同的值。import numpy as npfrom sklearn.preprocessing import normalizedef foo1(x, y):    with np.errstate(invalid='ignore'):        x_norm = np.nan_to_num(x / (np.linalg.norm(x, axis=0)))    z = np.dot(x_norm, y / np.linalg.norm(y))    print(z)def foo2(x, y):    x_norm = normalize(x, axis=0)    z = np.dot(x_norm, normalize(y))    print(z)最小的可重現示例x = np.array([[1, 2, 3],              [4, 5, 6]])y = np.array([[1], [2], [3]])foo1(x, y)foo2(x, y)輸出[[0.62190562] [1.47271032]][[1.0611399 ] [2.79304638]]預期的第一個值。