C Numeric

Curated hub for C numerics: math, complex arithmetic, floating-point environment, random number generation, bit utilities, and newer checked integer helpers.

The C numerics area covers much more than “math functions”. It includes floating-point classification and exceptions, complex arithmetic, pseudo-random number generation, bit utilities, type-generic dispatch, and newer checked integer arithmetic helpers.

Use this hub when your question is about numeric computation or numeric representation. For raw memory, strings, or time/date functions, start from Memory, String, or Chrono instead.

# Start Here

Real-valued math

Elementary functions, powers, rounding, classification, special values, and the core floating-point function families.

sqrt · pow · fpclassify · NAN

Complex arithmetic

Complex numbers, projections, magnitudes, trigonometric/exponential functions, and the C99 complex model.

cabs · creal · cproj · I

Floating-point environment

Exception flags, rounding direction, saved environments, and the parts of C numerics that depend on the current FP execution state.

feclearexcept · fetestexcept · rounding macros

Pseudo-random number generation

The classic `rand`/`srand` surface and its limits, suitable mostly for simple or legacy code paths.

rand · srand · RAND_MAX

Bit utilities

C23 bit operations and endian-related helpers for low-level integer representation work.

stdc_leading_zeros · endian

Type-generic and checked arithmetic

Macros that dispatch across floating/complex families and the newer checked integer operations added for overflow-aware arithmetic.

tgmath · ckd_add · ckd_sub · ckd_mul

# Quick Map

If you need to...	Start with	Why
Compute elementary functions, powers, roots, or rounding behavior	Math	The main real-valued function families, classification macros, and floating constants live here.
Work with `double complex` and related types	Complex	Complex-specific functions and constants are separate from the real-valued math surface.
Inspect or control FP exceptions and rounding mode	Floating-point environment	These APIs deal with execution-state side effects, not just numeric values.
Generate simple pseudo-random integers	Random	The C library exposes the traditional `rand`/`srand` interface here.
Use type-generic numeric macros across float/double/complex variants	Type-generic math	`<tgmath.h>` chooses the right overload-like macro expansion for you.
Write overflow-aware integer arithmetic in modern C	ckd_add, ckd_sub, ckd_mul	The checked integer helpers are the direct route when arithmetic overflow must be detected rather than ignored.
Reason about bit widths, leading/trailing zeros, or endianness	Bit manipulation	The newer bit utilities cover low-level representation-oriented integer tasks.

# Numeric Families

Family	Representative entry points	Use it for
Core math	sin, cos, pow, sqrt, fma	Real-valued mathematical computation, rounding, powers, logs, and transcendental functions.
Classification and special values	fpclassify, isfinite, isnan, NAN, INFINITY	Detect value categories and handle NaN/infinity-aware code paths.
Complex arithmetic	cabs, creal, cimag, cpow	Complex-number computation and decomposition under the C complex model.
Floating-point environment	feclearexcept, fetestexcept, feraiseexcept, feholdexcept	Exception flags, saved environments, and rounding/FP-state-sensitive execution.
Pseudo-random numbers	rand, srand, RAND_MAX	Traditional C library PRNG entry points for simple or legacy use.
Type-generic dispatch	tgmath	Calling the right math/complex family through macros instead of choosing the suffixed name manually.
Checked and bit-oriented integer helpers	ckd_add, ckd_sub, ckd_mul, stdc_leading_zeros	Overflow-aware arithmetic and low-level integer representation utilities in newer C revisions.

# Common Entry Points

Task	Go here	Notes
Absolute value / magnitude	abs, fabs, cabs	Pick integer, real floating, or complex magnitude based on the numeric domain.
Exponentials, logs, and powers	exp, log, pow	These are central math-family entry points and often pair with classification/error handling.
Roots and distance-like helpers	sqrt, cbrt, hypot	Use these instead of ad hoc formulas when precision and special cases matter.
Classify a floating-point result	fpclassify, isfinite, isnan, signbit	These are often needed around exceptional or boundary numeric cases.
Check or clear FP exceptions	feclearexcept, fetestexcept	Start here when correctness depends on FP status flags rather than only returned values.
Seed and draw from the classic PRNG	srand, rand	Useful mainly for compatibility and simple code; not a modern high-quality generator design.

# Standard Evolution

Standard	Navigation impact
C99	Major expansion: complex arithmetic, floating-point environment, type-generic math, classification macros, and many modern math facilities became part of the standard library.
C11	The numeric area gained alignment with newer language features and continued to rely on the C99 math/complex/fenv core as the main navigation backbone.
C23	Added newer bit helpers and checked integer arithmetic entry points, making `/c/numeric/` broader than just floating-point and complex APIs.

# Boundary Lines

This hub covers	Use a different hub for
Math, complex numbers, FP environment, random numbers, checked arithmetic, and bit-oriented numeric helpers.	String for text-to-number conversion and byte-string utilities, Chrono for time/date arithmetic, Language for constant-expression and conversion rules.
The C library view of numerics.	C++ numerics when you need the C++ library surface instead of the C one.