How to generate numpy arrays with random regions with rectangular shape with random size?

Question

I would like to generate a binary array (values are 1s and 0s) filled with rectangular regions. Something like this:

a = np.array([0, 1, 1, 0, 0]
             [0, 1, 1, 0, 0]
             [0, 0, 0, 0, 0]
             [0, 1, 1, 1, 0]
             [0, 1, 1, 1, 0])

The idea is that the rectangles could be generated randomly, and even intersect, like here:

b = np.array([0, 1, 1, 0, 0]
             [0, 1, 1, 1, 0]
             [0, 1, 1, 1, 0]
             [0, 1, 1, 1, 0]
             [0, 1, 1, 1, 0])

How could one accomplish this?

Answer 1

Here's a function that places ones in a random location based on the shape you set:

def fill_array(base_array, shape_to_be_inserted):
    row_loc = np.random.randint(low=0, high=(base_array.shape[0]-shape_to_be_inserted.shape[0]+1))
    col_loc = np.random.randint(low=0, high=(base_array.shape[1]-shape_to_be_inserted.shape[1]+1))

    output_array = np.copy(base_array)
    for i in range(base_array.shape[0]):
        for k in range(base_array.shape[1]):
            if i >= row_loc and i < row_loc+shape_to_be_inserted.shape[0]:
                if k >= col_loc and k < col_loc+shape_to_be_inserted.shape[1]:
                    output_array[i, k] = 1
    return output_array

And here I put the function in a loop:

empty_area = np.zeros(shape=(8, 8))
some_shape = np.zeros(shape=(3, 3))

for i in range(20):
    empty_area = fill_array(empty_area, some_shape)

I have to be honest - I found this solution very unsatisfactory and I'd like to hear how others would solve this problem.