how to get the neighboring elements in a numpy array with taking boundaries into account?

Question

I want to get the neighbors of the certain element in the numpy array. Lets consider following example

    a = numpy.array([0,1,2,3,4,5,6,7,8,9])
         


        

Answer 1

The method I came up with involves slicing the array and then padding as necessary to fill out-of-bounds values.

Taken from here:

def surrounding(x, idx, radius=1, fill=0):
""" 
Gets surrounding elements from a numpy array 

Parameters: 
x (ndarray of rank N): Input array
idx (N-Dimensional Index): The index at which to get surrounding elements. If None is specified for a particular axis,
    the entire axis is returned.
radius (array-like of rank N or scalar): The radius across each axis. If None is specified for a particular axis, 
    the entire axis is returned.
fill (scalar or None): The value to fill the array for indices that are out-of-bounds.
    If value is None, only the surrounding indices that are within the original array are returned.

Returns: 
ndarray: The surrounding elements at the specified index
"""

assert len(idx) == len(x.shape)

if np.isscalar(radius): radius = tuple([radius for i in range(len(x.shape))])

slices = []
paddings = []
for axis in range(len(x.shape)):
    if idx[axis] is None or radius[axis] is None:
        slices.append(slice(0, x.shape[axis]))
        paddings.append((0, 0))
        continue

    r = radius[axis]
    l = idx[axis] - r 
    r = idx[axis] + r

    pl = 0 if l > 0 else abs(l)
    pr = 0 if r < x.shape[axis] else r - x.shape[axis] + 1

    slices.append(slice(max(0, l), min(x.shape[axis], r+1)))
    paddings.append((pl, pr))

if fill is None: return x[slices]
return np.pad(x[slices], paddings, 'constant', constant_values=fill)