Select Elements from NumPy Array by Index Range

In this article, we will discuss how to select an element or a subarray from a NumPy array using indexing.

Creating a NumPy Array with numpy.arange()

First, let’s create a NumPy array using numpy.arange():

import numpy as np

# Create a numpy ndarray
npArray = np.arange(1, 20, 2)
print(npArray)

Contents of the NumPy array:

[ 1  3  5  7  9 11 13 15 17 19]

Select Single Element from NumPy Array by Index

To select an element from a NumPy array, we use the [] operator:

# Syntax to access an element
# ndarray[index]

It returns the element at the specified index.

Example: Selecting an element at index 2 from npArray:

import numpy as np

# Create a numpy ndarray
npArray = np.arange(1, 20, 2)
print(npArray)

# Select an element at index 2 (Index starts from 0)
elem = npArray[2]
print('Element at index 2 is:', elem)

Output:

[ 1  3  5  7  9 11 13 15 17 19]
Element at index 2 is: 5

Select a Subarray from a NumPy Array by Index Range

We can select a subarray using the [] operator:

# Syntax to access a subarray
# ndArray[first:last]

This returns a subarray with elements from the first to last - 1 indices.

Let’s see some examples:

Example 1: Selecting a subarray with elements from index 1 to 6:

import numpy as np

# Create a numpy ndarray
npArray = np.arange(1, 20, 2)
print(npArray)

# Select elements from index 1 to 6
subArray = npArray[1:7]
print(subArray)

Output:

[ 1  3  5  7  9 11 13 15 17 19]
[ 3  5  7  9 11 13]

Example 2: Selecting elements from the beginning to index 3:

# Selecting elements from the beginning to index 3
subArray = npArray[:4]

print(subArray)

Output:

[1 3 5 7]

Example 3: Selecting elements from index 2 to the end:

# Selecting elements from index 2 to the end:
subArray = npArray[2:]

print(subArray)

Output:

[ 5  7  9 11 13 15 17 19]

NumPy Broadcasting

A subarray returned by the [] operator is just a view of the original array, meaning the data is not copied. Modifications in the subarray will be reflected in the original array.

Example to illustrate this:

import numpy as np

# Create a new array and select a subarray
npArray = np.arange(1, 20, 2)
subArray = npArray[1:7]

# Modify the subarray
subArray[1] = 220

# Observe changes in both arrays
print('Modified Sub Array:', subArray)
print('Original Array:', npArray)

Output:

Modified Sub Array: [  3 220   7   9  11  13]
Original Array: [  1   3 220   7   9  11  13  15  17  19]

In data science, this behavior is useful for handling large datasets without unnecessary copying.

Example 2:

In this example we will select a subarray and change all values in that sub array. Corresponding values in original array will also change. For example,

import numpy as np

# Create a new array and select a subarray
npArray = np.arange(1, 20, 2)
subArray = npArray[1:7]

# Modify all elements of subarray
subArray[:] = 220

# Observe changes in both arrays
print('Modified Sub Array:', subArray)
print('Original Array:', npArray)

Output:

Modified Sub Array: [220 220 220 220 220 220]
Original Array: [  1 220 220 220 220 220 220  15  17  19]

Creating a Copy of a Sub Array

If you need an independent copy of a subarray, use the .copy() method:

import numpy as np

# Create a new array and select a subarray
npArray = np.arange(1, 20, 2)
subArray = npArray[1:7]

# Fetch a copy of a subarray
subArrayCopy = npArray[1:7].copy()

# Modify the copy
subArrayCopy[:] = 220

# The original array remains unchanged
print('Modified Sub Array:', subArrayCopy)
print('Original Array:', npArray)

Output:

Modified Sub Array: [220 220 220 220 220 220]
Original Array: [ 1  3  5  7  9 11 13 15 17 19]

Here, we created a copy of the selected sub array from NumPy Array. Any changes made in it will have no effect in original array.

Complete Example

import numpy as np


# Create a numpy ndArray
npArray = np.arange(1, 20, 2)

print('Contents of numpy ndArray')
print(npArray)

print('*** Select an element by Index ***')

# Select an element at index 2 (Index starts from 0)
elem = npArray[2]

print('Element at 2nd index  : ' , elem)

print('*** Select a by sub array by Index Range ***')

# Select elements from index 1 to 6
subArray = npArray[1:7]

print('Sub Array from 1st to 6th index are :', subArray)

# Select elements from beginning to index 3
subArray = npArray[:4]

print('Sub Array from beginning to 3rd index are :', subArray)

# Select elements from 2nd index to end
subArray = npArray[2:]

print('Sub Array from 2nd index to end are :', subArray)


print('*** Sub Array is just a View not the copy ***')


npArray = np.arange(1, 20, 2)

print('Contents of Original Array : ', subArray)

# Select a sub array of elements from index 1 to 6
subArray = npArray[1:7]

print('Contents of Sub Array : ', subArray)

# Change contents of sub array
subArray[1] = 220
'''
Sub array is just a view of original array i.e. data is not copied just a view of sub array is created.
Any modification in it will be reflected in original nodArray too
'''
print('Contents of modified Sub Array : ', subArray)
print('Contents of Original Array : ', npArray)


print('*** Create a copy of Sub Array of ndArray *** ')

npArray = np.arange(1, 20, 2)
print('Contents of Original Array : ', subArray)

# Fetch a copy of sub array from index 1 to 6
subArray = npArray[1:7].copy()
print('Contents of Sub Array : ', subArray)

# Change contents of sub array
subArray[1] = 220

'''
As subArray is a copy of sub array not the view only, so changes made in it will not be reflected in main array.
'''
print('Contents of modified Sub Array : ', subArray)
print('Contents of Original Array : ', npArray)

Output:

Contents of numpy ndArray
[ 1  3  5  7  9 11 13 15 17 19]
*** Select an element by Index ***
Element at 2nd index  :  5
*** Select a by sub array by Index Range ***
Sub Array from 1st to 6th index are : [ 3  5  7  9 11 13]
Sub Array from beginning to 3rd index are : [1 3 5 7]
Sub Array from 2nd index to end are : [ 5  7  9 11 13 15 17 19]
*** Sub Array is just a View not the copy ***
Contents of Original Array :  [ 5  7  9 11 13 15 17 19]
Contents of Sub Array :  [ 3  5  7  9 11 13]
Contents of modified Sub Array :  [  3 220   7   9  11  13]
Contents of Original Array :  [  1   3 220   7   9  11  13  15  17  19]
*** Create a copy of Sub Array of ndArray *** 
Contents of Original Array :  [  3 220   7   9  11  13]
Contents of Sub Array :  [ 3  5  7  9 11 13]
Contents of modified Sub Array :  [  3 220   7   9  11  13]
Contents of Original Array :  [ 1  3  5  7  9 11 13 15 17 19]

Summary

We learned how select elements and subarrays in NumPy, and understood both the default behavior (view) and how to create independent copies when needed.