Consider using asarray rather than astype #4
Description
If you do not need to cast the array, it is faster to use asarray rather than astype since asarray will be much less expensive if not cast is needed.
In [16]: a = np.arange(1e8, dtype='float32')
In [17]: %timeit a.astype('float32')
10 loops, best of 3: 200 ms per loop
In [18]: %timeit np.asarray(a, dtype='float32')
1000000 loops, best of 3: 965 ns per loop
In [19]: %timeit a.astype('float64')
1 loops, best of 3: 270 ms per loop
In [20]: %timeit np.asarray(a, dtype='float64')
1 loops, best of 3: 269 ms per loop