<cfloat> (float.h)

Header file:

<cfloat> (float.h)

Characteristics of floating-point types

This header file describes the characteristics of floating-point types for special system and compiler implementations.

A floating-point number consists of four elements:

• A symbol (a sign): positive or negative;
• A base (a base): This means that a string of different numbers can be represented by a single number (2 is binary), 10 is decimal (decimal), and 16 is hexadecimal (hexadecimal) ...);
• An effective number (a significand): This is a series of numbers mentioned earlier, the number of digits in this series is well-known precision;
• An exponent (an exponent): it represents the offset of a valid number, affecting its value in the following way:

Floating-point value = valid number x cardinality index (with the corresponding flag (sign))(image)

Macro Constants

The table below shows the names of the different variables defined in this header file and the minimum maximum values for all implementations
(The specific implementation may have values larger or smaller as described below):

When a group of macros is prefixed with Flt_,dbl_ and Ldbl_, the flt_ starts with float,以DBL_开头的适用于double,以LDBL_开头的适用于long double;

Name	Value	Represent	Said
Flt_radix	2 or larger	Cardinality (radix)	Cardinality of all floating-point types
flt_mant_dig Span style= "font-size:15px;" >dbl_mant_dig ldbl_mant_dig	 	(Mantissa DIGits)	the precision of the effective number, the number of digits that match the number of valid digits
flt_dig dbl_dig ldbl_dig	6 or greater 10 or larger 10 or greater	Precision (DIGits)	with a decimal number can be converted to a floating point and convert it back again without changing the integer number
flt_min_exp dbl_min_exp ldbl_min_exp	 		the smallest negative integer value for generating a normalized floating-point exponent
flt_min_10_exp Span style= "font-size:15px;" >dbl_min_10_exp ldbl_min_10_exp	-37 or smaller -37 or smaller -37 or smaller	min. 10 base exponent	decimal exponential expression for generating a standard floating-point number
flt_max_exp dbl_max_exp ldbl_max_exp	 		Maximum integer value for generating a normalized floating-point exponent
flt_max_10_exp Span style= "font-size:15px;" >dbl_max_10_exp ldbl_max_10_exp	37 or greater 37 or greater	10-based exponent	decimal exponential expression for generating a standard floating-point number
flt_max dbl_max ldbl_max	1E+37 or greater 1E+37 or greater 1E+37 or larger	Maximum number	Represents the maximum limited number of floating-point numbers
flt_epsilon dbl_epsilon ldbl_epsilon	1E-5 or smaller 1E-9 or smaller 1E-9 or smaller	epsilon:)	can represent a difference of 1 and Min values greater than 1
flt_min dbl_min ldbl_min	1E-37 or smaller 1E-37 or smaller 1E-37 or smaller	minimum	represents the least limited number of floating-point numbers
flt_rounds	 	Rounding Method	rounding behavior, Possible values: -1: Indeterminate 0: rounding to 0 1: rounding 2: Positive Infinity 3: Negative infinity * applies to all floating-point types (float, Double and long double)
flt_eval_method		assignment-formatted properties, possible values: -1: Indeterminate 0: Range and precision of the assignment type 1: Assign a value of float and double as double, long double as long double 2: all are treated as long double when assigned, and other negative numbers are expressed as the definition of implementation (implementation-defined) behavior * applies to all floating-point types
Decimal_dig		Decimal precision	You can convert a decimal number to a floating-point type number, And again to the same number as the original without loss of precision

Versatility

FLT_EVAL_METHOD AndDECIMAL_DIG 是在符合C99标准的库中定义的或者只在符合C11的库中

Please see

<climits> (LIMITS.H) The size of an integer type

/*

First modified: January 30, 2016 20:32:35

*/

<cfloat> (float.h)