C Programming/float.h

The header float.h contains macros that expand to various limits and parameters of the standard floating-point types.

The macros, as defined on ISO 9899:1999 section 5.2.4.2.2, are:

• FLT_ROUNDS – indicates the rounding mode for floating-point addition, meaning:

• -1 indeterminable;
• 0 toward zero;
• 1 to nearest;
• 2 toward positive infinity;
• 3 toward negative infinity;
• other value means implementation-defined rounding behavior.

• FLT_EVAL_METHOD (C99) – determines the evaluation formats of expressions involving all floating types:

• -1 indeterminable;
• 0 evaluate all operations and constants just to the precision of the type;
• 1 evaluate operations and constants of type float and double to the range and precision of the double type;
• 2 evaluate all operations and constants to the range and precision of the long double type;
• other value means implementation-defined behavior.

• FLT_RADIX – radix of exponent representation (at least 2).

• FLT_MANT_DIG, DBL_MANT_DIG, LDBL_MANT_DIG – number of digits in the floating-point significand.

• DECIMAL_DIG (C99) – (at least 10)

• FLT_DIG, DBL_DIG, LDBL_DIG - (at least 6, 10, 10)

• FLT_MIN_EXP, DBL_MIN_EXP, LDBL_MIN_EXP

• FLT_MIN_10_EXP, DBL_MIN_10_EXP, LDBL_MIN_10_EXP, (at least -37)

• FLT_MAX_EXP, DBL_MAX_EXP, LDBL_MAX_EXP

• FLT_MAX_10_EXP, DBL_MAX_10_EXP, LDBL_MAX_10_EXP (at least +37)

• FLT_MAX, DBL_MAX, LDBL_MAX – (Maximum value, at least 1E+37)

• FLT_EPSILON, DBL_EPSILON, LDBL_EPSILON – (Machine epsilon, no greater than 1E-5, 1E-9, 1E-9)

• FLT_MIN, DBL_MIN, LDBL_MIN – (no greater than 1E-37)

External links

• float.h: floating types – Base Definitions Reference, The Single UNIX® Specification, Issue 7 from The Open Group
• C++ reference for implementation defined numeric constants inherited from C