python - 快速重复(垂直)二维数组的每一半的numpy方法

Question

假设我们有以下二维数组:

In [200]: a = np.arange(8).reshape(4,2)

In [201]: a
Out[201]:
array([[0, 1],
       [2, 3],
       [4, 5],
       [6, 7]])

如何重复它的每一半，所以我有以下二维数组:

array([[0, 1],
       [2, 3],
       [0, 1],
       [2, 3],
       [4, 5],  # second half 
       [6, 7],
       [4, 5],
       [6, 7]])

我的尝试产生了错误的结果:

In [202]: np.tile(np.split(a, 2), 2).reshape(-1,2)
Out[202]:
array([[0, 1],
       [0, 1],
       [2, 3],
       [2, 3],
       [4, 5],
       [4, 5],
       [6, 7],
       [6, 7]])

Answer 1

reshape 将第一个轴分成两个，给我们一个 3D 数组，然后沿着第一个轴重复，最后 reshape 回 2D -

np.repeat(a.reshape(-1,2,2),2,axis=0).reshape(-1,2)

概括它 -

def repeat_blocks(a):
    N = a.shape[0]
    B = N//2 # Block length
    R = 2 # number of repeats
    out = np.repeat(a.reshape(N//B,B,-1),R,axis=0).reshape(N*R,-1)
    return out

示例运行 -

案例#1:

In [120]: a
Out[120]: 
array([[0, 1],
       [2, 3],
       [4, 5],
       [6, 7]])

In [121]: repeat_blocks(a)
Out[121]: 
array([[0, 1],
       [2, 3],
       [0, 1],
       [2, 3],
       [4, 5],
       [6, 7],
       [4, 5],
       [6, 7]])

案例#2:

In [123]: a
Out[123]: 
array([[ 0,  1],
       [ 2,  3],
       [ 4,  5],
       [ 6,  7],
       [ 8,  9],
       [10, 11]])

In [124]: repeat_blocks(a)
Out[124]: 
array([[ 0,  1],
       [ 2,  3],
       [ 4,  5],
       [ 0,  1],
       [ 2,  3],
       [ 4,  5],
       [ 6,  7],
       [ 8,  9],
       [10, 11],
       [ 6,  7],
       [ 8,  9],
       [10, 11]])