Floating-point arithmetic
This section describes how FLINT implements floating-point arithmetic outside the arb ball arithmetic (which has its own group, Ball arithmetic): the fixed-precision nfloat type and its kernels, the limb-level mp_real balls and their hypergeometric series, correctly rounded elementary functions, decimal arithmetic, double-word to quad-double expansions and balls in dfloat, and interval arithmetic with doubles. The algorithms of the elementary functions and constants themselves are described with the special functions (Elementary functions, Mathematical constants). Start with the overview, which compares all floating-point types in FLINT and says which to use.
In this section#
The FLINT team, “Floating-point arithmetic,” FLINT: Fast Library for Number Theory, documentation, version 3.7.0-dev, 2026. https://flintlib.org/doc/algorithms/floating-point/index.html (generated 2026-11-20).
@misc{flint-doc-algorithms-floating-point-index,
author = {{The FLINT team}},
title = {{Floating-point arithmetic}},
howpublished = {FLINT: Fast Library for Number Theory, documentation, version 3.7.0-dev},
year = {2026},
url = {https://flintlib.org/doc/algorithms/floating-point/index.html},
note = {Generated 2026-11-20}
}