evalhf/constant - constants known to evalhf
|
Description
|
|
•
|
The following constants are known to evalhf:
|
Pi
|
gamma
|
Catalan
|
true
|
false
|
FLT_RADIX
|
DBL_MANT_DIG
|
DBL_DIG
|
DBL_EPSILON
|
DBL_MIN_EXP
|
DBL_MIN
|
DBL_MIN_10_EXP
|
DBL_MAX_EXP
|
DBL_MAX
|
DBL_MAX_10_EXP
|
LNMAXFLOAT
|
|
|
•
|
evalhf(Digits) returns the approximate number of digits of the hardware floating-point system in use, regardless of the actual setting of Digits. This constant takes into account the precision returned by the standard mathematical functions.
|
•
|
Inside evalhf, true has the value 1.0 and false has the value 0.0.
|
•
|
The following constants are taken from the ANSI-C definition of floating-point parameters of an IEEE-standard floating-point system.
|
•
|
FLT_RADIX: The base (or radix) of the floating-point number system.
|
•
|
DBL_MANT_DIG: The number of significant base digits.
|
•
|
DBL_DIG: The (approximate) equivalent number of significant decimal digits.
|
•
|
DBL_EPSILON: The smallest positive x such that
|
•
|
DBL_MIN_EXP: Minimum normalized exponent
|
•
|
DBL_MIN: Minimum normalized positive number
|
•
|
DBL_MIN_10_EXP: Minimum exponent of a representable power of 10
|
•
|
DBL_MAX_EXP: Maximum exponent
|
•
|
DBL_MAX: Maximum number
|
•
|
DBL_MAX_10_EXP: Maximum exponent of a representable power of 10
|
•
|
LNMAXFLOAT: A number slightly smaller than the largest number x such that evalhf(exp(x)) and evalhf(exp(-x)) both successfully compute, and such that the results of those computations can continue to be used in evalhf. On most machines, this value is set as evalhf(-ln(DBL_MIN)) truncated to 6 decimal places.
|
|
|
Download Help Document
Was this information helpful?