Can you slice a numpy 2D array with start and stop rules not at an index but at values greater than 1?

Question

I have turned a geotiff LiDAR elevation surface into a numpy array by the below code:

example_array = np.array(geotiff_surface)

The array is hundreds of values by hundreds of values but could be simplified into an example below:

example_array = [[ 0, 0, 0, 1, 1, 0, 0, 0]
                 [ 0, 0, 2, 2, 2, 2, 0, 0]
                 [ 0, 3, 3, 3, 3, 0, 0, 0]
                 [ 4, 4, 4, 4, 4, 4, 0, 0]
                 [ 0, 0, 0, 5, 5, 0, 0, 0]]

with the help of others I am trying to slice the 2D array to increase the values at specific locations by 1000 with the code below:

sub_array = example_array[1:-1, 1:-1]
sub_array = [sub_array > 1] += 1000

in hopes to get the below correct example array:

correct_example_array = [[ 0,  0,    0,    1,    1,   0, 0, 0]
                         [ 0,  0,    2,   1002, 1002, 2, 0, 0]
                         [ 0,  3,   1003, 1003,  3,   0, 0, 0]
                         [ 4, 1004, 1004, 1004, 1004, 4, 0, 0]
                         [ 0,  0,    0,    5,    5,   0, 0, 0]]

when in reality what my current code gives me is this array:

  incorrect_example_array = [[ 0,  0,    0,    1,    1,    0,   0, 0]
                             [ 0,  0,   1002, 1002, 1002, 1002, 0, 0]
                             [ 0, 1003, 1003, 1003, 1003,  0,   0, 0]
                             [ 4, 1004, 1004, 1004, 1004, 1004, 0, 0]
                             [ 0,   0,    0,    5,    5,    0,  0, 0]]

Is there a way to get the slice to understand to skip the zeros in the lists and start slicing [1:-1] within values greater than 1?

Answer 1

Here is a way to do it (not nice one maybe):

for i in range(example_array.shape[0]):
    if len(np.where(example_array[i]!=0)[0][1:-1])>0:
        example_array[i][np.where(example_array[i]!=0)[0][1:-1]] = example_array[i][example_array[i]!=0][1:-1]+1000
print(example_array)