How to Convert NumPy Array of Floats into Integers

In this article, we will see how to convert the NumPy Array of Floats into Integers. We are given a NumPy array of float-type values. Our task is to convert all the float-type values of the Numpy array to their nearest array of integer values.

Input1: [1.2 4.5 9.1 6.5 8.9 2.3 1.2]
Output1: [1 4 9 6 8 2 1]
Input2: [ 1.2  3.4  5.6  7.8  9.1 11.2 14.5 16.7]
Output2: [ 1  3  5  7  9 11 14 16]

Convert NumPy Array of Floats into Integers in 1-D Array

There are some methods to convert the float values in our NumPy array to the array of nearest integer values in Python. Here are some of the methods mentioned below :

• Naive Approach
• Using numpy.astype()
• Using numpy.asarray()
• Using np.int_()

Naive Approach

By explicitly type casting we can convert each element of the original Numpy array to its closest integer value. This technique converts a float value to an integer using the int() or round() methods. Use int() if you have to obtain floor values; otherwise, use round() to remove the decimal point and round to the nearest integer.

Python3

import numpy as np

org = np.array([1.2,3.4,5.6,7.8,9.1,11.2,14.5,16.7])

#Displaying the original array
print("Original Array : ")
print(org,"\n")

#coverting array's float value to its integer value using int()
integer_array = np.array([int(i) for i in org])

#Displaying the result
print("Integer Array : ")
print(integer_array,"\n")

#coverting array's float value to its integer value using round()
round_array = np.array([round(i) for i in org])

#Displaying the result
print("Round Array : ")
print(round_array)

Output

Original Array : 
[ 1.2  3.4  5.6  7.8  9.1 11.2 14.5 16.7] 
Integer Array : 
[ 1  3  5  7  9 11 14 16] 
Round Array : 
[ 1  3  6  8  9 11 14 17]

Time complexity: O(N)
Auxiliary Space: O(N)

Using numpy.astype()

In this example, we are using numpy .astype(). By passing the desired data type in the method, we can convert it to any data type within the Numpy array. Along with numpy.astype(), round() must be used to obtain precise values. Let’s see the code now. We had added both the methods with and without round(). Now its solely up to you which you choose depending on your need.

Python3

import numpy as np

org = [1.2,3.4,5.6,7.8,9.1,11.2,14.5,16.7]
org = np.array(org)

print("Original Array : ")
print(org,"\n")

#Just eliminating the integer part
integer_array = org.astype(int)

#Displaying the result
print("Integer Array : ")
print(integer_array,"\n")

#converting for closest integer value
round_array = np.round(org).astype(int)

#Displaying the result
print("Round Array : ")
print(round_array)

Output

Original Array : 
[ 1.2  3.4  5.6  7.8  9.1 11.2 14.5 16.7] 
Integer Array : 
[ 1  3  5  7  9 11 14 16] 
Round Array : 
[ 1  3  6  8  9 11 14 17]

Time complexity: O(N)
Auxiliary Space: O(N)

Using numpy.asarray()

In this example, we are using numpy.asarray() to convert NumPy Array of floats into integers.

Python3

import numpy as np

#creating the original array
org = [1.2,4.5,9.1,6.5,8.9,2.3,1.2]
org = np.array(org)

#Displaying the original array
print("Original Array : ")
print(org,"\n")

#converting the array to its integer form
new = np.asarray(org,dtype="int")

#Displaying the result
print("New Array : ")
print(new)

Output

Original Array : 
[1.2 4.5 9.1 6.5 8.9 2.3 1.2] 
New Array : 
[1 4 9 6 8 2 1]

Time complexity: O(N)
Auxiliary Space: O(N)

Using np.int_()

In this example, we are using np.int_() to convert NumPy Array of floats into integers.

Python3

import numpy as np

#creating the original array
org = [1.2,4.5,9.1,6.7,8.9,2.8,1.7]
org = np.array(org)

#Displaying the original array
print("Original Array : ")
print(org,"\n")

#Just eliminating the integer part
integer_array = np.int_(org)

#Displaying the result
print("Integer Array : ")
print(integer_array,"\n")

#converting for closest integer value
round_array = np.round_(org)
round_array = np.int_(round_array)

#Displaying the result
print("Round Array : ")
print(round_array)

Output

Original Array : 
[1.2 4.5 9.1 6.7 8.9 2.8 1.7] 
Integer Array : 
[1 4 9 6 8 2 1] 
Round Array : 
[1 4 9 7 9 3 2]

Time complexity: O(N)
Auxiliary Space: O(N)

Python NumPy Array of Floats into Integers in 2-D Array

There are some methods to convert the float values in our numpy array to the array of nearest integer values. Here are some of the methods mentioned below :

• Naive Approach
• Using numpy.astype()
• Using numpy.asarray()
• Using np.int_()

Naive Approach.

This implementation is same as we saw in 1-D array. We are here using explicit type casting too.

Python3

import numpy as np

org = np.array([
    [1.2,2.3,3.4],
    [0.1,1.3,2.6],
    [1.5,4.5,9.2]])

#displaying the original array
print("Original Array : ")
print(org,"\n")

# Create an empty list
new = []

for i in range(org.shape[0]):
    helper = []
    for j in range(org.shape[1]):
        helper.append(int(org[i, j]))
    new.append(helper)

integer = np.array(new)

#displaying the integer array

print("Integer Array : ")
print(integer)

Output

Original Array : 
[[1.2 2.3 3.4]
 [0.1 1.3 2.6]
 [1.5 4.5 9.2]] 
Integer Array : 
[[1 2 3]
 [0 1 2]
 [1 4 9]]

Time complexity: O(N*M) , where N is no. of rows and M is no. of columns
Auxiliary Space: O(N*M) , where N is no. of rows and M is no. of columns

Using numpy.astype()

In this example, we are using numpy.astype(), similar to that of the 1-D array. Just call the method to the whole 2-D array at once. Now lets move to the code implementation.

Python3

import numpy as np

org = np.array([
    [1.2,2.3,3.4],
    [0.1,1.3,2.6],
    [1.5,4.5,9.2]])

#displaying the original array
print("Original Array : ")
print(org,"\n")

#applying .astype()
integer = org.astype(int)

#displaying the integer array

print("Integer Array : ")
print(integer)

Output

Original Array : 
[[1.2 2.3 3.4]
 [0.1 1.3 2.6]
 [1.5 4.5 9.2]] 
Integer Array : 
[[1 2 3]
 [0 1 2]
 [1 4 9]]

Time complexity: O(N*M) , where N is no. of rows and M is no. of columns
Auxiliary Space: O(N*M) , where N is no. of rows and M is no. of columns

Using numpy.asarray()

Its too has the same use case as the previous one in 1-D array. It takes whole 2-D array at once and then convert it into the desired data type which is passed through the function. Now lets move to code implementation.

Python3

import numpy as np

org = np.array([
    [1.2,2.3,3.4],
    [0.1,1.3,2.6],
    [1.5,4.5,9.2]])

#displaying the original array
print("Original Array : ")
print(org,"\n")

#applying .asarray() with dtype (data type) as integer
integer = np.asarray(org, dtype=int)

#displaying the integer array

print("Integer Array : ")
print(integer)

Output

Original Array : 
[[1.2 2.3 3.4]
 [0.1 1.3 2.6]
 [1.5 4.5 9.2]] 
Integer Array : 
[[1 2 3]
 [0 1 2]
 [1 4 9]]

Time complexity: O(N*M) , where N is no. of rows and M is no. of columns
Auxiliary Space: O(N*M) , where N is no. of rows and M is no. of columns

Using np.int_()

We had seen np.int_() implementation in 1-D array part. Its same for the 2-D array too. Now lets see the code implementation.

Python3

import numpy as np

org = np.array([
    [1.2,2.3,3.4],
    [0.1,1.3,2.6],
    [1.5,4.5,9.2]])

#displaying the original array
print("Original Array : ")
print(org,"\n")

#applying int_() to org array
integer = np.int_(org)

#displaying the integer array

print("Integer Array : ")
print(integer)

Output

Original Array : 
[[1.2 2.3 3.4]
 [0.1 1.3 2.6]
 [1.5 4.5 9.2]] 
Integer Array : 
[[1 2 3]
 [0 1 2]
 [1 4 9]]

Time complexity: O(N*M) , where N is no. of rows and M is no. of columns
Auxiliary Space: O(N*M) , where N is no. of rows and M is no. of columns