double keyword in C designates a double-precision floating-point data type used to represent fractional numbers. It is the default data type for floating-point literals in C and typically conforms to the IEEE 754 binary64 standard.
Technical Specifications
While the C standard does not strictly mandate the underlying representation, on virtually all modern architectures,double possesses the following characteristics:
- Memory Size: 8 bytes (64 bits).
- Bit Layout (IEEE 754):
- 1 Sign bit
- 11 Exponent bits
- 52 Mantissa (significand) bits
- Precision: 15 to 17 significant decimal digits.
- Range: Approximately to .
- Standard Limits: Defined in the
<float.h>header asDBL_MINandDBL_MAX.
Syntax and Initialization
Floating-point literals without a suffix are treated asdouble by the compiler. They can be written in standard decimal notation or scientific (exponential) notation.
Format Specifiers
Input and output operations via the standard I/O library (<stdio.h>) require specific format specifiers. There is a critical distinction between reading and writing a double.
Output (printf)
When passed to variadic functions like printf, float arguments are automatically promoted to double. Therefore, the standard specifier for printing a double is %f (or %e for scientific notation, %g for the shortest representation).
scanf)
When reading input, you must pass a pointer to the exact type. You must use the %lf (long float) specifier to read into a double variable. Using %f will result in undefined behavior, as scanf will attempt to write a 4-byte float into an 8-byte double memory space.
Type Promotion and Literals
Because floating-point literals aredouble by default, assigning a literal to a standard float causes an implicit downcast, which may trigger compiler warnings regarding truncation.
float and double, the float operand is implicitly promoted to double before the operation is evaluated, yielding a double result.
Tired of Poor C Skills? Fix That With Deep Grasping!Learn More
