Time for action – calculating the exponential moving average

Given an array, the exp function calculates the exponential of each array element. For example, look at the following code:

x = np.arange(5)
print "Exp", np.exp(x)

It gives the following output:

Exp [  1.           2.71828183   7.3890561   20.08553692  54.59815003]

The linspace function takes, as parameters, a start and a stop and optionally an array size. It returns an array of evenly spaced numbers. The following is an example:

print "Linspace", np.linspace(-1, 0, 5)

This will give us the following output:

Linspace [-1.   -0.75 -0.5  -0.25  0.  ]

Let's calculate the exponential moving average for our data:

1. Now, back to the weights—calculate them with exp and linspace.

   N = int(sys.argv[1])
   weights = np.exp(np.linspace(-1., 0., N))

2. Normalize the weights. The ndarray object has a sum method that we will use.

   weights /= weights.sum()
   print "Weights", weights

   For N = 5, we get the following weights:

   Weights [ 0.11405072  0.14644403  0.18803785...