Standadise Numpy Array

Question

I am trying to standardize a numpy array. I seem to be doing something wrong as the value of some elements of the output array is incorrect. I'd appreciate any help. Please find the code below: PYTHON CODE:

from numpy import loadtxt
import numpy as np
import math
class matrix(object):
 """A matrix class."""
 def __init__(self):
 self.array_2d = np.array([[0, 1, 2], [3, 4, 5], [6, 7, 8]])
 def standardise(self):
 rows, columns = self.array_2d.shape
 temp = self.array_2d
 for j in range(columns):
 for i in range(rows):
 self.array_2d[i,j] = (temp[i, j] - min(temp[:, j])) / (max(temp[:, j]) - min(temp[:, j]))
m = matrix()
print(m.array_2d, "\n")
m.standardise()
print(m.array_2d)

• Nin17: can you add the desired output?
• slothrop: temp = self.array_2d doesn't make a copy of array_2d: it just makes the name temp point to the same array. So in your for loop, temp points to the same array that you've been changing in previous iterations of the loop, not to the original array.

Answer 1

If what you want to do is just to scale the matrix you dont have to do it in a for loop. You can do like this because Numpy is vectorized by default.

a = np.array([[0, 1, 2], [3, 4, 5], [6, 7, 8]])
scaled_a = (a - a.min(axis=1))/(a.max(axis=1) - a.min(axis=1))