Benefit of NumPy arrays over Python arrays

The need for NumPy arises when we are working with multi-dimensional arrays. The traditional array module does not support multi-dimensional arrays.

Let’s first try to create a single-dimensional array (i.e one row & multiple columns) in Python without installing NumPy Package to get a more clear picture.

Python3

from array import *


arr = array('i', [25, 16, 3])
print(arr)

Output:

array('i', [25, 16, 3])

Now, Let’s try to create a multi-dimensional array by using the array module.

Python3

from array import *


arr = array('i', [25, 16, 3], [5, 19, 28])
print(arr)

Output:

TypeError: array() takes at most 2 arguments (3 given)

We see that the array module does not support multi-dimensional array, this is where we require NumPy. NumPy supports large, multi-dimensional arrays and has a large collection of high-level math functions that can operate on those arrays.

Let’s use NumPy to create a multi-dimensional array.

Python3

from numpy import *


arr = array ([[25, 31, 3], [5, 19, 28]])
print(arr)

Output:

[[25 31  3]
 [ 5 19 28]]