Numpy intersect1d 与以矩阵为元素的数组

问题描述

我有两个数组，一个形状为 (200000, 28, 28)，另一个形状为 (10000, 28, 28)，所以实际上两个以矩阵为元素的数组.现在我想计算并获取在两个数组中重叠的所有元素(以 (N, 28, 28) 形式).使用普通的 for 循环会变慢，所以我用 numpys intersect1d 方法尝试了它，但我不知道如何将它应用于这种类型的数组.

I have two arrays, one of the shape (200000, 28, 28) and the other of the shape (10000, 28, 28), so practically two arrays with matrices as elements.Now I want to count and get all the elements (in the form (N, 28, 28)), that overlap in both arrays. With normal for loops it is way to slow, so I tryied it with numpys intersect1d method, but I dont know how to apply it on this types of arrays.

推荐答案

使用来自 这个关于唯一行的问题

def intersect_along_first_axis(a, b):
    # check that casting to void will create equal size elements
    assert a.shape[1:] == b.shape[1:]
    assert a.dtype == b.dtype

    # compute dtypes
    void_dt = np.dtype((np.void, a.dtype.itemsize * np.prod(a.shape[1:])))
    orig_dt = np.dtype((a.dtype, a.shape[1:]))

    # convert to 1d void arrays
    a = np.ascontiguousarray(a)
    b = np.ascontiguousarray(b)
    a_void = a.reshape(a.shape[0], -1).view(void_dt)
    b_void = b.reshape(b.shape[0], -1).view(void_dt)

    # intersect, then convert back
    return np.intersect1d(b_void, a_void).view(orig_dt)

请注意，使用 void 对浮点数是不安全的，因为它会导致 -0 不等于 0

Note that using void is unsafe with floats, as it will cause -0 to be unequal to 0

这篇关于Numpy intersect1d 与以矩阵为元素的数组的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持！