How can I change numpy array values?

Question

I can't really understand why my code isn't working, see:

import numpy as np

def round2c(x):
    for i in x:
        for j in i:
            if (j >= 0 and j<= 0.33):
                j = 0
            else:
                j = 1

    return x

list = [[0.05,0.1,0.4],[0.5,0.5,0.3]]
list2 = round2c(list)
print(list2)

a = np.array([[0.05,0.1,0.4],[0.5,0.5,0.3]])
b = round2c(a)
print(b)

It's neither working with a np.ndarray nor with a list. What am I doing wrong?

Isn't it something like `b = a.apply(lambda x: round2c(x))`? — , Aug 31 '22 at 08:42
Assigning to the loop variable `j` does not change values in the list itself. You need to assign via index — rdas, Aug 31 '22 at 08:42

score 0 · Answer 1 · answered Aug 31 '22 at 08:46

I think your problem is the assigment of variables, check the following is working.

import numpy as np

def round2c(x):
    for num1, i in enumerate(x):
        for num2, j in  enumerate(i):
            if (0 <= j<= 0.33):
                i[num2] = 0
            else:
                i[num2] = 1
        x[num1] = i
    return x

list1 = [[0.05,0.1,0.4],[0.5,0.5,0.3]]
list2 = round2c(list)
print(list2)

a = np.array([[0.05,0.1,0.4],[0.5,0.5,0.3]])
b = round2c(a)
print(b)

output:

[[0, 0, 1], [1, 1, 0]]
[[0. 0. 1.]
 [1. 1. 0.]]

zvxayr · Answer 2 · 2022-08-31T09:04:37.250

If your goal is to binarize the numbers, numpy arrays has the method digitize which achieves what your code does without creating a function yourself.

list1 = [[0.05,0.1,0.4], [0.5,0.5,0.3]]
list1 = np.array(list1)

# bins contains the numbers that separates the values in the list
# right is False since 0.33 is included in the left interval
list2 = np.digitize(list1, bins=[0.33], right=False)
print(list2)

The docs for the binarize method is in this link.

How can I change numpy array values?

2 Answers2