If we wanted to create a NumPy array with the numbers 0-1000 ... but that would be slow. Or, we could use np.add.reduce to achieve the same thing with no loop: x1 = np.array([[0,1,2],[3,4,5 ...

So, as you waltz through the world of NumPy, keep the invisible line in your mind for optimal performance. Python performance gets a bad rap compared with languages such as Java. Use these tips to ...