How to Compare Floating-Point Numbers for Near-Equality in Python?

Floating-Point Comparison with Almost-Equality in Python

Comparing floating-point numbers for equality in Python can present challenges due to precision issues and rounding errors. To address this, it's essential to employ techniques that allow for a meaningful comparison.

One approach is to use the standard library function introduced in Python 3.5:

import math
math.isclose(a, b, rel_tol=1e-09, abs_tol=0.0)

The rel_tol parameter specifies a relative tolerance, which is a multiplier applied to the larger of the two numbers being compared. abs_tol represents an absolute tolerance applied directly to the difference between the numbers.

If the difference between a and b is less than either the relative or absolute tolerance, the two numbers are considered almost-equal and the function returns True. Otherwise, it returns False.

For Python versions prior to 3.5, an equivalent function can be defined as follows:

def isclose(a, b, rel_tol=1e-09, abs_tol=0.0):
    return abs(a - b) <= max(rel_tol * max(abs(a), abs(b)), abs_tol)

The above is the detailed content of How to Compare Floating-Point Numbers for Near-Equality in Python?. For more information, please follow other related articles on the PHP Chinese website!