Numpy 用于重新排序，以便将当前行的每一 n 行追加到其中

Question

我想重塑 Numpy 数组 A，不是通过附加正常的下一行，而是通过附加它之后的每一 N 行。

前任：

A = [[1  2  3  4]
[5  6  7  8]
[9  10 11 12]
[13 14 15 16]
[17 18 19 20]
[21 22 23 24]]

现在，我想从 A 构造一个大小为 2x12 的数组，如下所示

 A = [[1 2 3 4 9 10 11 12 17 18 19 10]
 [5 6 7 8 13 14 15 16 21 22 23 24]]

在这里，您可以看到当前行中的每一 2 行都附加到它并形成了新的重塑数组。

Answer 1

您可以使用

简单的索引，array.ravel()和np.vstack()：

In [37]: np.vstack((A[::2].ravel(), A[1::2].ravel()))
Out[37]: 
array([[ 1,  2,  3,  4,  9, 10, 11, 12, 17, 18, 19, 20],
       [ 5,  6,  7,  8, 13, 14, 15, 16, 21, 22, 23, 24]])

Answer 2

这种方法怎么样？

# extract alternative rows starting from 0th row (1st row)
In [18]: A[0::2]
Out[18]: 
array([[ 1,  2,  3,  4],
       [ 9, 10, 11, 12],
       [17, 18, 19, 20]])
# and then flatten to 1D array
In [19]: A[0::2].flatten()
Out[19]: array([ 1,  2,  3,  4,  9, 10, 11, 12, 17, 18, 19, 20])

# extract alternative rows starting from 2nd row
In [17]: A[1::2]
Out[17]: 
array([[ 5,  6,  7,  8],
       [13, 14, 15, 16],
       [21, 22, 23, 24]])

# and then flatten to 1D array
In [20]: A[1::2].flatten()
Out[20]: array([ 5,  6,  7,  8, 13, 14, 15, 16, 21, 22, 23, 24])

# to get 2D, just put them in `np.vstack` (in the order you want the final array)
In [21]: np.vstack((A[0::2].flatten(), A[1::2].flatten()))
Out[21]: 
array([[ 1,  2,  3,  4,  9, 10, 11, 12, 17, 18, 19, 20],
       [ 5,  6,  7,  8, 13, 14, 15, 16, 21, 22, 23, 24]])