Avoiding Floating Point Arithmetic Inaccuracies in Python

Common Python Problems - [3] Float Data Types

1. Floating Point Arithmetic Inaccuracies

Example 1

x = 1/10

y = 1/5

z = x + y

print(z)  # expected output is 0.3

Output

0.30000000000000004

Solution: Using the decimal module in Python, we can accurately represent decimal values and avoid floating point inaccuracies.

from decimal import Decimal

x = Decimal("1") / Decimal("10")

y = Decimal("1") / Decimal("5")

z = x + y

print(z)  # expected output is 0.3

Output

0.3

2. Overflow Errors

Example 1

An example of a float overflow error in Python can be seen below:

num1 = 9e309

num2 = 2

result = num1 * num2

print(result)

Output

inf

Solution: Python will raise an OverflowError when the value of a float is too large. To avoid this error, you can use the Decimal module to handle large numbers and prevent float overflow errors. For example:

num1 = Decimal("9e309")

num2 = Decimal(2)

result = num1 * num2

print(result)

Output

1.8E+310

3. Rounding Errors

Example 1

# We are trying to calculate 0.1 + 0.2

x = 1/10

y = 1/5

z = x + y

print(z)

Output

0.30000000000000004

Solution:We can use the round() function to round off the result

x = 1/10

y = 1/5

z = x + y

rounded_z = round(z, 2)

print(rounded_z)

Output

0.3

Common Python Problems - [3] Float Data Types

1. Floating Point Arithmetic Inaccuracies

2. Overflow Errors

3. Rounding Errors

Related

Common Python Problems - [2] Mutable Arguments

Common Python Problems - [1] None and Boolean Comparison

Common Python Problems

Subscribe

Subscribe to our newsletter and never miss out lastest news.