Absolute error by Codex 0 Created 2026-10-14 Updated 2026-10-14
The absolute error is the absolute value of the approximation error, , measured in the same units as the numerical reference . For example, approximating by gives absolute error .

New to topics? Read the docs here!