Random interpolation between 2 numpy arrays

I am looking for a way to make a random interpolation between two numpy arrays with few “magical” Numpy operations as possible.

for example, having

a =[1,2,3] b=[4,5,6]

i would like to have as output

c = [ between 1 and 4, between 2 and 5, between 3,6] of integers.

means sort of

def mate_interpolate(arr,arr2):
takes 2 parents of same dimention
created a child array of same dimention 
on each place having the element, what value is randomly uniform between
the values of parents 

randfloats = np.random.uniform(0,1,arr.shape)
child =  arr2 - arr
child *= randfloats
child += arr
#child = arr + (arr2 - arr) * randfloats
return child

but without creating this additional array and/or with only integers.


For floats, you can just do:

rng = np.random.default_rng()
result = rng.uniform(a, b)

If you want integers, you can either truncate, or if b > a:

result = rng.integers(a, b, endpoint=True)