Numbers and maths

Numbers look like the simplest thing in any language, and most of the time they are — until 0.1 + 0.2 doesn't equal 0.3, a total comes out a penny short, or a rounding rule surprises you. Python actually gives you several different number types, each making a different trade between speed, range, and exactness, and most number bugs come from using the wrong one for the job.

This guide covers the lot: the built-in int, float, and complex; the exact Decimal and Fraction types for money and ratios; and the math, statistics, and random modules from the standard library. The recurring theme is a single question — do I need this to be exact? — and matching the answer to the right type.

Sections

• Learn — four notebooks: the numeric types and their arithmetic, how floating point really behaves (and why), exact arithmetic with Decimal and Fraction, and a tour of math, statistics, and random.
• Recipes — rounding numbers correctly, handling money with Decimal, formatting numbers for display, and the mistakes that cost a penny or a debugging hour.
• Reference — a math module reference, the number side of the format mini-language, and a numeric-types-and-operators lookup.
• Concepts — essays on how floating point works under the hood, and on choosing the right numeric type for a given job.

New to the topic? Start with Learn → Numeric types. Here for a specific task? Recipes is task-focused and Reference is for lookups.