如何在不填充其 k 轴的情况下填充 3D np.array 的 i、j 轴？

Question

我有一个 3-D ndarray。

>>> b = np.arange(27).reshape(3,3,3)
>>> b
array([[[ 0,  1,  2],
        [ 3,  4,  5],
        [ 6,  7,  8]],

       [[ 9, 10, 11],
        [12, 13, 14],
        [15, 16, 17]],

       [[18, 19, 20],
        [21, 22, 23],
        [24, 25, 26]]])

Numpy pad 函数returns 一个 5x5x5 数组：

>>> np.pad(b, (1, 1), constant_values=0)
array([[[ 0,  0,  0,  0,  0],
        [ 0,  0,  0,  0,  0],
        [ 0,  0,  0,  0,  0],
        [ 0,  0,  0,  0,  0],
        [ 0,  0,  0,  0,  0]],

       [[ 0,  0,  0,  0,  0],
        [ 0,  0,  1,  2,  0],
        [ 0,  3,  4,  5,  0],
        [ 0,  6,  7,  8,  0],
        [ 0,  0,  0,  0,  0]],

       [[ 0,  0,  0,  0,  0],
        [ 0,  9, 10, 11,  0],
        [ 0, 12, 13, 14,  0],
        [ 0, 15, 16, 17,  0],
        [ 0,  0,  0,  0,  0]],

       [[ 0,  0,  0,  0,  0],
        [ 0, 18, 19, 20,  0],
        [ 0, 21, 22, 23,  0],
        [ 0, 24, 25, 26,  0],
        [ 0,  0,  0,  0,  0]],

       [[ 0,  0,  0,  0,  0],
        [ 0,  0,  0,  0,  0],
        [ 0,  0,  0,  0,  0],
        [ 0,  0,  0,  0,  0],
        [ 0,  0,  0,  0,  0]]])

但是，我想要一个像这样的 5x5x3 数组：

array([[[ 0,  0,  0,  0,  0],
        [ 0,  0,  1,  2,  0],
        [ 0,  3,  4,  5,  0],
        [ 0,  6,  7,  8,  0],
        [ 0,  0,  0,  0,  0]],

       [[ 0,  0,  0,  0,  0],
        [ 0,  9, 10, 11,  0],
        [ 0, 12, 13, 14,  0],
        [ 0, 15, 16, 17,  0],
        [ 0,  0,  0,  0,  0]],

       [[ 0,  0,  0,  0,  0],
        [ 0, 18, 19, 20,  0],
        [ 0, 21, 22, 23,  0],
        [ 0, 24, 25, 26,  0],
        [ 0,  0,  0,  0,  0]]])

如何实现上述目标？

Answer 1

您可以使用 ((0,0),(1,1),(1,1)) 代替 (1,1) 来填充：

np.pad(b, ((0,0),(1,1),(1,1)), constant_values=0)

...或者仅 trim 第一项和最后一项：

np.pad(b, (1,1), constant_values=0)[1:-1]

Answer 2

可能不是最优雅的解决方案但是：

>>> np.pad(b, (1,1), constant_values=0)[1:-1]
 
array([[[ 0,  0,  0,  0,  0],
        [ 0,  0,  1,  2,  0],
        [ 0,  3,  4,  5,  0],
        [ 0,  6,  7,  8,  0],
        [ 0,  0,  0,  0,  0]],

       [[ 0,  0,  0,  0,  0],
        [ 0,  9, 10, 11,  0],
        [ 0, 12, 13, 14,  0],
        [ 0, 15, 16, 17,  0],
        [ 0,  0,  0,  0,  0]],

       [[ 0,  0,  0,  0,  0],
        [ 0, 18, 19, 20,  0],
        [ 0, 21, 22, 23,  0],
        [ 0, 24, 25, 26,  0],
        [ 0,  0,  0,  0,  0]]])

适合我。

如何在不填充其 k 轴的情况下填充 3D np.array 的 i、j 轴？

How to pad the i, j axes of a 3D np.array without padding its k axis?

python

numpy