我如何在下面将这个 for 循环矢量化，我需要在其中将值设置为我需要舍入的范围？

Question

我有一个 np.array q，其中包含一些值，例如：[1,3,5,7] .
还有一个 np.array z。我需要舍入一些值，然后将它们用作
中的索引第三个数组'mapping'.

import numpy as np

    q = [1,3,5,7]
    z = [0,50.3,240.4,252.9,256]

    mapping = np.zeros(256)
    for i in range(len(q)):
        print(i)
        start, end = int(round(z[i])), int(round(z[i + 1]))
        mapping[start:end] = int(round(q[i]))

    print(mapping)

这里的输出是：

Answer 1

这是我的方法：

repeats = np.diff(list(np.round(z))+ [256]).astype(int)
# repeats = array([ 49, 191,  12,   3])

np.repeat(np.round(q), repeats)

输出：

array([1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
       1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
       1, 1, 1, 1, 1, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3,
       3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3,
       3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3,
       3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3,
       3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3,
       3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3,
       3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3,
       3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3,
       3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 5, 5,
       5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 7, 7, 7])

注意：这只有 255 个元素，与您预期的输出不同，因为，说实话我不太理解您的逻辑。

我如何在下面将这个 for 循环矢量化，我需要在其中将值设置为我需要舍入的范围？

How can I vectorize this for loop below, where I need to set values to a range I need to round?

python

numpy

vectorization