0

我有 2 个数组:
-image是一个 NxN 数组,
-indices是一个 Mx2 数组,其中最后一个维度将有效索引存储到image.

我想imageindices.

似乎numpy.add.at(image, indices, 1)应该做的伎俩,除了我不能让它执行二维索引image

image = np.zeros((5,5), dtype=np.int32)
indices = np.array([[1,1], [1,1], [3,3]])
np.add.at(image, indices, 1)
print(image)

结果:

[[0 0 0 0 0]
 [4 4 4 4 4]
 [0 0 0 0 0]
 [2 2 2 2 2]
 [0 0 0 0 0]]

期望的结果:

[[0 0 0 0 0]
 [0 2 0 0 0]
 [0 0 0 0 0]
 [0 0 0 1 0]
 [0 0 0 0 0]]
4

1 回答 1

6
In [477]: np.add.at(x,(idx[:,0],idx[:,1]), 1)                                                          
In [478]: x                                                                                            
Out[478]: 
array([[0., 0., 0., 0., 0.],
       [0., 2., 0., 0., 0.],
       [0., 0., 0., 0., 0.],
       [0., 0., 0., 1., 0.],
       [0., 0., 0., 0., 0.]])

或等效地

In [489]: np.add.at(x,tuple(idx.T), 1)                                                                 
In [490]: x                                                                                            
Out[490]: 
array([[0., 0., 0., 0., 0.],
       [0., 2., 0., 0., 0.],
       [0., 0., 0., 0., 0.],
       [0., 0., 0., 1., 0.],
       [0., 0., 0., 0., 0.]])

在哪里:

In [491]: tuple(idx.T)                                                                                 
Out[491]: (array([1, 1, 3]), array([1, 1, 3]))
In [492]: x[tuple(idx.T)]                                                                              
Out[492]: array([2., 2., 1.])
于 2019-06-08T22:59:35.590 回答