Why astype(uint) on np.array doesn't change type of an element of the np.array?

Question

When converting an np.array to uint8 using astype the type of an element of the array doesn't change.

>>> x = np.array([[1.0, 2.3], [1.3, 2.9]])
>>> x.astype(uint8)
array([[1, 2],
       [1, 2]], dtype=uint8)
 >>> type(x[0,0])
<type 'numpy.float64'>

Why the element is still float64 and not uint8?

Answer 1

astype returns a copy of the origin array.

Use x = x.astype(uint8) instead

Answer 2

astype returns a copy of an array, so you must assign it:

x = x.astype(uint8)