python - 给定中心点索引获取二维数组的子集

Question

给定一个二维数组和一个索引为 (x,y) 的特定元素，如何获得以该元素为中心的二维正方形子集 (n x n)？

只有当子集数组的大小完全在原始数组的范围内时，我才能实现它。如果特定元素靠近原始数组的边缘或角落，我会遇到问题。对于这种情况，子集数组对于原始数组之外的元素必须具有 nan 值。

Answer 1

这是我将如何做到这一点:

def fixed_size_subset(a, x, y, size):
    '''
    Gets a subset of 2D array given a x and y coordinates
    and an output size. If the slices exceed the bounds 
    of the input array, the non overlapping values
    are filled with NaNs
    ----
    a: np.array
        2D array from which to take a subset
    x, y: int. Coordinates of the center of the subset
    size: int. Size of the output array
    ----       
    Returns:
        np.array
        Subset of the input array
    '''
    o, r = np.divmod(size, 2)
    l = (x-(o+r-1)).clip(0)
    u = (y-(o+r-1)).clip(0)
    a_ = a[l: x+o+1, u:y+o+1]
    out = np.full((size, size), np.nan, dtype=a.dtype)
    out[:a_.shape[0], :a_.shape[1]] = a_
    return out

---

样本运行:

# random 2D array
a = np.random.randint(1,5,(6,6))

array([[1, 3, 2, 2, 4, 1],
       [1, 3, 1, 3, 3, 2],
       [1, 1, 4, 4, 2, 4],
       [1, 2, 3, 4, 1, 1],
       [4, 1, 4, 2, 3, 4],
       [3, 3, 2, 3, 2, 1]])

---

fixed_size_subset(a, 3, 3, 5)

array([[3., 1., 3., 3., 2.],
       [1., 4., 4., 2., 4.],
       [2., 3., 4., 1., 1.],
       [1., 4., 2., 3., 4.],
       [3., 2., 3., 2., 1.]])

让我们尝试一些示例，其中切片数组小于预期的输出大小:

fixed_size_subset(a, 4, 1, 4)

array([[ 1.,  2.,  3.,  4.],
       [ 4.,  1.,  4.,  2.],
       [ 3.,  3.,  2.,  3.],
       [nan, nan, nan, nan]])

fixed_size_subset(a, 5, 5, 3)

array([[ 3.,  4., nan],
       [ 2.,  1., nan],
       [nan, nan, nan]])

下面的也行:

fixed_size_subset(a, -1, 0, 3)

array([[ 1.,  3., nan],
       [nan, nan, nan],
       [nan, nan, nan]])