Numpy binary matrix - get rows and columns of True

问题:

I have a binary numpy 2D array, say,

import numpy as np
arr = np.array([
#   Col 0   Col 1  Col 2
    [False, False, True],  # Row 0
    [True, False, False],  # Row 1
    [True, True, False],  # Row 2
])

I want the row and column of each True element in the matrix:

[(0, 2), (1, 0), (2, 0), (2, 1)]

I know I can do this through iteration:

links = []
nrows, ncols = arr.shape
for i in xrange(nrows):
    for j in xrange(ncols):
        if arr[i, j]:
            links.append((i, j))

Is there a faster or more intuitive way?

回答1:

You are looking for np.argwhere -

np.argwhere(arr)

Sample run -

In [220]: arr
Out[220]: 
array([[False, False,  True],
       [ True, False, False],
       [ True,  True, False]], dtype=bool)

In [221]: np.argwhere(arr)
Out[221]: 
array([[0, 2],
       [1, 0],
       [2, 0],
       [2, 1]])