raise an error, when Vector.normalize() tries to normalize a "zero-vector"

[Devrim-Celik]

Test script:
Problem is, when you try to normalize a Vector Object, which only has zero value, it tries to divide
them by 0 (their current magnitude), setting the values to nan. It would be handy if a error would pop up.

from p5 import *

x = Vector(0, 0)

print(x.magnitude)
x.normalize()
print(x)
print(x.magnitude)

Expected behavior:

0.0
ValueError(...)

Actual behavior

0.0
Vector(nan, nan, nan)
nan

Errors produced (if any):

p5 version:
0.3.0a2

Python version:
Python 3.6.4

Operating System:
MacOS

[arihantparsoya]

See #38 .