Understanding Floating‑Point Precision Issues in Python and How to Handle Them

Python has a surprising "bug": the expression 0.1 + 0.2 == 0.3 evaluates to False .

This occurs because computers store numbers in binary, and decimal fractions like 0.1, 0.2, and 0.3 become infinitely repeating binary fractions that must be truncated, leading to a loss of exact precision.

Consequently, most decimal numbers suffer from tiny errors, which are usually negligible in everyday use but become problematic when comparing floating‑point numbers for equality.

To compare floating‑point numbers safely, developers typically define an acceptable error margin (e.g., 10⁻¹⁰) and consider two numbers equal if their absolute difference is smaller than this threshold.

For domains requiring high precision, such as finance or scientific computing, Python’s built‑in decimal module can be used to control precision accurately.

When creating Decimal objects, pass the value as a string (e.g., Decimal('0.1') ) to avoid inheriting the original floating‑point error.

In practice, a simple way to avoid floating‑point pitfalls in monetary calculations is to work with the smallest currency unit (e.g., cents) as an integer, thereby eliminating decimal arithmetic altogether.