Prevent Python from rounding down to zero here?

Question

I have the following array and function:

import numpy as np
a = np.array([24, 23,  4, 52, 34, 49, 59, 18, 19])

def normalize(a):
    amin, amax = min(a), max(a)
    for i, val in enumerate(a):
        a[i] = (val-amin) / (amax-amin)
    return a

I get the following result:

array([0, 0, 0, 0, 0, 0, 1, 0, 0])

How can I prevent Python from not revealing the decimals of the zeros?

Answer 1

I guess what happens is because a.dtype is integer, so everything is converted to integer when you update a single position with a[i] =....

In general, you should avoid looping in numpy:

a = np.array([24, 23,  4, 52, 34, 49, 59, 18, 19])

def normalize(a):
    # np.min is vectorized. Python's `min` is not
    amin, amax = np.min(a), np.max(a)
    return (a-amin)/(amax-amin)

normalize(a)