How to Sort a NumPy Array

Copy Link LinkedIn Message Facebook Email X

Numpy allows for easy array sorting. The function np.sort returns a sorted copy of an array, leaving the original array unchanged. A related function is np.argsort, which returns the indices of the sorted elements. This lesson covers NumPy sort, NumPy Argsort, and additional functions you can use to sort a NumPy array.

NumPy Sort

import numpy as np

# standard sorting

a = np.array([2, 1, 4, 3])
# returns the sorted array (a copy)
sorted_array = np.sort(a)  
print(sorted_array)

array([1, 2, 3, 4])

NumPy Argsort

# sorting with indices

a = np.array([2, 1, 4, 3])
# Returns the indices that would sort an array.
sorted_indices = np.argsort(a)  

print(sorted_indices)

array([1, 0, 3, 2])

How to Sort Arrays In-Place

# in-place sorting

a = np.array([2, 1, 4, 3])
# sorts a in place, returns nothing
a.sort() 
print(a)

array([1, 2, 3, 4])

Multi-dimensional Array Sort

# multi-dimensional sorting

a = np.array([[14, 11], [12, 13]])
print(a)
print()
# Sort along the first axis
print(np.sort(a, axis=0))
print()
# Sort along the second axis
print(np.sort(a, axis=1))

[[14 11]
 [12 13]]
    
[[12 11]
 [14 13]]
    
[[11 14]
 [12 13]]

Summary: NumPy Sort

• To sort arrays using NumPy sort, use the command np.sort() to return a sorted copy of an array without altering the original.
• Use the NumPy argsort command - np.argsort() - to return indices that would sort the array.
• array.sort() sorts the array in-place, changing the original array.
• Multi-dimensional arrays can be sorted along specific axes using np.sort(a, axis=n). This does not alter the original array.
• When sorting multi-dimensional arrays, the axis parameter determines the axis along which the array is sorted. For 2D arrays, axis=0 sorts along the rows (which sorts the columns) while axis=1 sorts along the columns (which sorts the rows).