3d Numpy 数组到 2d [英] 3d Numpy array to 2d

问题描述

我有一个像这样的 3d 矩阵

I have a 3d matrix like this

arange(16).reshape((4,2,2))
array([[[ 0,  1],
        [ 2,  3]],

        [[ 4,  5],
        [ 6,  7]],

        [[ 8,  9],
        [10, 11]],

        [[12, 13],
        [14, 15]]])

并希望以网格格式将它们堆叠起来，以

and would like to stack them in grid format, ending up with

array([[ 0,  1,  4,  5],
       [ 2,  3,  6,  7],
       [ 8,  9, 12, 13],
       [10, 11, 14, 15]])

有没有一种方法可以不显式地对它们进行 hstacking(和/或 vstacking)或添加额外的维度和整形(不确定这是否可行)?

Is there a way of doing without explicitly hstacking (and/or vstacking) them or adding an extra dimension and reshaping (not sure this would work)?

谢谢，

推荐答案

In [27]: x = np.arange(16).reshape((4,2,2))

In [28]: x.reshape(2,2,2,2).swapaxes(1,2).reshape(4,-1)
Out[28]: 
array([[ 0,  1,  4,  5],
       [ 2,  3,  6,  7],
       [ 8,  9, 12, 13],
       [10, 11, 14, 15]])

<小时>

我在此处发布了更多用于将数组整形/取消整形为块的通用函数.

这篇关于3d Numpy 数组到 2d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IT屋！