Numpy.nonzero () returns an array, the position of non-zero elements. If the two-dimensional array is non-zero elements is described in several odd row, three-dimensional array is non-zero elements described in the first few lines of the first group of several columns.
For example as follows:
Two-dimensional array:
a = np.array([[1, 0, 3], [0, 2, 0], [0, 0, 9]]) b = np.nonzero(a) print(b)
Results: (array ([0, 0, 1, 2], dtype = int64), array ([0, 2, 1, 2], dtype = int64))
The first description line array, the second array that describes the column, in order to transform our understanding of the results:
array[0, 0, 1, 2]
array[0, 2, 1, 2]
We see the first nonzero element 1 in row 0 0, the corresponding figures for the bold:
array[0, 0, 1, 2]
array[0, 2, 1, 2]
3 is a second non-zero elements in row 2 0, corresponding to:
array[0, 0, 1, 2]
array[0, 2, 1, 2]
The third is non-zero elements 2 in a row 1, corresponding to:
array[0, 0, 1, 2]
array[0, 2, 1, 2]
9 is a fourth nonzero elements in two rows and two, corresponding to:
array[0, 0, 1, 2]
array[0, 2, 1, 2]
As another three-dimensional array of Liezi:
a = np.array([[[0,1],[2,0]],[[0,3],[4,0]],[[0,0],[5,0]]]) b = np.nonzero(a) print(b)
结果为：(array([0, 0, 1, 1, 2], dtype=int64), array([0, 1, 0, 1, 1], dtype=int64), array([1, 0, 1, 0, 0], dtype=int64))
The same deformation:
array [0, 0, 1, 1, 2] first described in several groups
array [0, 1, 0, 1, 1] Description Line
array [1, 0, 1, 0, 0] Description column
1 is the first nonzero element in a row 0 Group 0, corresponding to
array[0, 0, 1, 1, 2]
array[0, 1, 0, 1, 1]
array[1, 0, 1, 0, 0]
2 is a second non-zero elements in Row 1 0 0 group, the corresponding
array[0, 0, 1, 1, 2]
array[0, 1, 0, 1, 1]
array[1, 0, 1, 0, 0]
3 is the third non-zero elements in row 0 1 1 group, corresponding to
array[0, 0, 1, 1, 2]
array[0, 1, 0, 1, 1]
array[1, 0, 1, 0, 0]
4 is a fourth non-zero elements in row 0 1 1 group, corresponding to
array[0, 0, 1, 1, 2]
array[0, 1, 0, 1, 1]
array[1, 0, 1, 0, 0]
5 is the fifth non-zero elements in 2 groups of rows 10 columns corresponding to
array[0, 0, 1, 1, 2]
array[0, 1, 0, 1, 1]
array[1, 0, 1, 0, 0]
Higher dimension calculation similar, readers can derive their own