static T
epsilon
() throw();
The function returns the difference between 1 and the smallest value greater than 1 that is representable for the type
(which is the value FLT_EPSILON
for type float).
See the related numeric_limits sample program.