When passed 2 arrays of differing length, scipy.spatial.distance.hamming raises:

AttributeError: 'bool' object has no attribute 'mean'

This is due to the way the hamming method calculates distance:

u = _validate_vector(u)
v = _validate_vector(v)
return (u != v).mean()

If the arrays have different lengths a single bool is returned by (u != v), and the mean method excepts. Would it be reasonable to detect length differences explicitly and raise a more meaningful exception?