Functions for manipulating floating-point numbers

ABSTRACT

Floating-point arithmetic is used for most scientific and engineering calculations and needs frequently arise for determining the precision of a given floating-point number, for detailed access to its component parts and for reconstructing it from its parts, much as for complex numbers. We give detailed definitions of three suitable functions in the hope that their wide use will enhance portability. They are first given without reference to any particular language and then made specific to Fortran.This note has been discussed by IFIP WG 2.5 on Numerical Software and has been approved by the working group, but does not constitute an official IFIP document.

REFERENCES

• 1: Brian Ford, Parameterization of the Environment for Transportable Numerical Software, ACM Transactions on Mathematical Software (TOMS), v.4 n.2, p.100-103, June 1978 [doi>10.1145/355780.355782]
• 2: W. S. Brown , S. I. Feldman, Environment parameters and basic functions for floating-point computation, Proceedings of the SIGNUM Conference on the Programming Environment for Development of Numerical Software, p.42-45, October 01-01, 1978