二维 numpy 数组中的优化坐标选择

Question

我有一个 numpy 坐标数组。我想要 select Xmin 和 Xmax 之间以及 Ymin 和 Ymax 之间的那些。

这是我的代码：

grid = np.random.randint(2, size=81).reshape(9,9)
List_of_coordinates = np.argwhere(grid == 0)
Xmin, Xmax, Ymin, Ymax = 0,3,6,9
List_of_coordinates = List_of_coordinates[List_of_coordinates[:, 1] >= Ymin]
List_of_coordinates = List_of_coordinates[List_of_coordinates[:, 1] < Ymax]
List_of_coordinates = List_of_coordinates[List_of_coordinates[:, 0] >= Xmin]
List_of_coordinates = List_of_coordinates[List_of_coordinates[:, 0] < Xmax]

有没有更快的方法？

Answer 1

您可以再次使用argwhere。

import numpy as np
grid = np.random.randint(2, size=81).reshape(9,9)
List_of_coordinates = np.argwhere(grid == 0)
Xmin, Xmax, Ymin, Ymax = 0,3,6,9
# select coordinates between Xmin and Xmax and Ymin and Ymax
coordinates = np.argwhere(
np.logical_and(List_of_coordinates[:,0] >= Xmin, List_of_coordinates[:,0] <= Xmax) &
np.logical_and(List_of_coordinates[:,1] >= Ymin, List_of_coordinates[:,1] <= Ymax)
)

Answer 2

首先对网格进行切片，您将获得相对于切片开头的索引，然后添加切片的开头以返回绝对坐标：

List_of_coordinates = (np.argwhere(grid[Xmin:Xmax,Ymin:Ymax] == 0)
                       +np.array([Xmin,Ymin]))

例子

网格：

array([[0, 1, 0, 0, 0, 0, 0, 1, 1],
       [1, 0, 1, 0, 1, 0, 1, 0, 0],
       [0, 1, 0, 1, 0, 1, 0, 0, 1],
       [1, 1, 1, 1, 1, 1, 1, 1, 1],
       [0, 1, 1, 1, 0, 0, 1, 0, 0],
       [1, 0, 0, 0, 0, 1, 1, 0, 0],
       [0, 0, 0, 0, 0, 1, 1, 0, 1],
       [1, 1, 0, 1, 1, 1, 1, 1, 0],
       [1, 0, 1, 0, 1, 0, 1, 0, 0]])

切片：

array([[0, 1, 1],
       [1, 0, 0],
       [0, 0, 1]])

0 的坐标：

array([[0, 6],
       [1, 7],
       [1, 8],
       [2, 6],
       [2, 7]])

Answer 3

我（微）对您的解决方案、您的解决方案的优化版本和@mozway 的方法进行了基准测试（我没有包括@Mohammad 的答案，因为输出不同，最初的结果与优化后的原始结果大致相同) 具有更大的示例数据。运行在 2 核 google colab 实例上。

import numpy as np

grid = np.random.randint(2, size=(1000,1000))
Xmin, Xmax, Ymin, Ymax = 0,300,600,900

%%timeit
List_of_coordinates = np.argwhere(grid == 0)
List_of_coordinates = List_of_coordinates[List_of_coordinates[:, 1] >= Ymin]
List_of_coordinates = List_of_coordinates[List_of_coordinates[:, 1] < Ymax]
List_of_coordinates = List_of_coordinates[List_of_coordinates[:, 0] >= Xmin]
List_of_coordinates = List_of_coordinates[List_of_coordinates[:, 0] < Xmax]
List_of_coordinates

原解10 loops, best of 5: 30.1 ms per loop

%%timeit
List_of_coordinates = np.argwhere(grid == 0)
LoC = List_of_coordinates[
    (List_of_coordinates[:, 1] >= Ymin) &
    (List_of_coordinates[:, 1] < Ymax) &
    (List_of_coordinates[:, 0] >= Xmin) &
    (List_of_coordinates[:, 0] < Xmax)               
]
LoC

优化原版100 loops, best of 5: 19.4 ms per loop

%%timeit
LoC = np.argwhere(grid[Xmin:Xmax, Ymin:Ymax] == 0) + np.array([Xmin,Ymin])
LoC

@mozway 的解决方案1000 loops, best of 5: 1.69 ms per loop

二维 numpy 数组中的优化坐标选择

Optimized coordinates selection in 2D numpy array

python

optimization

numpy

sudoku