通过索引数组（如二维数组的 np.in1d）屏蔽 2D Numpy 数组答案

【问题标题】：Mask a 2D Numpy Array by array of indexes like np.in1d for 2d arrays通过索引数组（如二维数组的 np.in1d）屏蔽 2D Numpy 数组
【发布时间】：2015-03-17 16:33:38
【问题描述】：

    np.array(
[[0,13,0,2,0,0,0,0,0,0,0,0],
 [0,0,15,0,9,0,0,0,0,0,0,0],
 [0,0,0,0,0,18,0,0,0,0,0,0],
 [0,0,0,0,27,0,20,0,0,0,0,0],
 [0,0,0,0,0,20,0,10,0,0,0,0],
 [0,0,0,0,0,0,0,0,8,0,0,0],
 [0,0,0,0,0,0,0,14,0,14,0,0],
 [0,0,0,0,0,0,0,0,12,0,25,0],
 [0,0,0,0,0,0,0,0,0,0,0,11],
 [0,0,0,0,0,0,0,0,0,0,15,0],
 [0,0,0,0,0,0,0,0,0,0,0,7],
 [0,0,0,0,0,0,0,0,0,0,0,0]])

我正在尝试找到如何获取一个像上面这样的 numpy 数组，然后在一个高性能操作中用我想要归零的元素索引对其进行屏蔽

[0,1] [1,4] [4,7] [7,8] [8,11]

所以我剩下的就是

np.array(
[[0,0,0,2,0,0,0,0,0,0,0,0],
 [0,0,15,0,0,0,0,0,0,0,0,0],
 [0,0,0,0,0,18,0,0,0,0,0,0],
 [0,0,0,0,27,0,20,0,0,0,0,0],
 [0,0,0,0,0,20,0,0,0,0,0,0],
 [0,0,0,0,0,0,0,0,8,0,0,0],
 [0,0,0,0,0,0,0,14,0,14,0,0],
 [0,0,0,0,0,0,0,0,0,0,25,0],
 [0,0,0,0,0,0,0,0,0,0,0,0],
 [0,0,0,0,0,0,0,0,0,0,15,0],
 [0,0,0,0,0,0,0,0,0,0,0,7],
 [0,0,0,0,0,0,0,0,0,0,0,0]])

类似于 np.in1d 的功能但对于 2d 数组？我可以遍历每个元素，但数组可以非常庞大，因此矢量单操作掩码是最好的。是否可以？如果这是一个愚蠢的问题，我相信我会被告知！

【问题讨论】：

标签： python numpy

【解决方案1】：

您可以通过以下方式直接访问这些索引

indexes = [[0,1], [1,4], [4,7], [7,8], [8,11]]
indexes =zip(*indexes)
>>[(0, 1, 4, 7, 8), (1, 4, 7, 8, 11)]
a[indexes[0], indexes[1]]=0
>>
[[ 0  0  0  2  0  0  0  0  0  0  0  0]
 [ 0  0 15  0  0  0  0  0  0  0  0  0]
 [ 0  0  0  0  0 18  0  0  0  0  0  0]
 [ 0  0  0  0 27  0 20  0  0  0  0  0]
 [ 0  0  0  0  0 20  0  0  0  0  0  0]
 [ 0  0  0  0  0  0  0  0  8  0  0  0]
 [ 0  0  0  0  0  0  0 14  0 14  0  0]
 [ 0  0  0  0  0  0  0  0  0  0 25  0]
 [ 0  0  0  0  0  0  0  0  0  0  0  0]
 [ 0  0  0  0  0  0  0  0  0  0 15  0]
 [ 0  0  0  0  0  0  0  0  0  0  0  7]
 [ 0  0  0  0  0  0  0  0  0  0  0  0]]

【讨论】：

如果你把indexes在压缩后变成tuple，你可以跳过索引，即indexes = tuple(zip(*indexes)); a[indexes] = 0。

【解决方案2】：

我想你在找这个

a = np.array(
[[0,13,0,2,0,0,0,0,0,0,0,0],
 [0,0,15,0,9,0,0,0,0,0,0,0],
 [0,0,0,0,0,18,0,0,0,0,0,0],
 [0,0,0,0,27,0,20,0,0,0,0,0],
 [0,0,0,0,0,20,0,10,0,0,0,0],
 [0,0,0,0,0,0,0,0,8,0,0,0],
 [0,0,0,0,0,0,0,14,0,14,0,0],
 [0,0,0,0,0,0,0,0,12,0,25,0],
 [0,0,0,0,0,0,0,0,0,0,0,11],
 [0,0,0,0,0,0,0,0,0,0,15,0],
 [0,0,0,0,0,0,0,0,0,0,0,7],
 [0,0,0,0,0,0,0,0,0,0,0,0]])

b = np.array([[0,1],[1,4],[4,7],[7,8],[8,11]])

# get x coordinates in an array
c1 = b[:,0]
# get y coordinates in an array
c2 = b[:,1]
a[c1[:,None],c2] = 0

a 
array([[ 0,  0,  0,  2,  0,  0,  0,  0,  0,  0,  0,  0],
       [ 0,  0, 15,  0,  0,  0,  0,  0,  0,  0,  0,  0],
       [ 0,  0,  0,  0,  0, 18,  0,  0,  0,  0,  0,  0],
       [ 0,  0,  0,  0, 27,  0, 20,  0,  0,  0,  0,  0],
       [ 0,  0,  0,  0,  0, 20,  0,  0,  0,  0,  0,  0],
       [ 0,  0,  0,  0,  0,  0,  0,  0,  8,  0,  0,  0],
       [ 0,  0,  0,  0,  0,  0,  0, 14,  0, 14,  0,  0],
       [ 0,  0,  0,  0,  0,  0,  0,  0,  0,  0, 25,  0],
       [ 0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0],
       [ 0,  0,  0,  0,  0,  0,  0,  0,  0,  0, 15,  0],
       [ 0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  7],
       [ 0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0]])

【讨论】：

将b 转换为数组后，a[tuple(b.T)] = 0 可能是完成任务的最紧凑方式。