Page History

Computers can represent only a finite number of digits. It is therefore impossible to precisely represent repeating binary-representation values such as 1/3 or 1/5 with the most common floating-point representation: binary floating point.

Wiki MarkupWhen precise computation is necessary, use alternative representations that can accurately represent the values. For example, if you are performing arithmetic on decimal values and need an exact decimal rounding, represent the values in binary-coded decimal instead of using floating-point values. Another option is decimal floating-point arithmetic as specified by ANSI/IEEE 754-2007. ISO/IEC WG14 has drafted a proposal to add support for decimal floating-point arithmetic to the C language \[ [ISO/IEC DTR 24732|AA. Bibliography#ISO/IEC DTR 24732]\].unmigrated-wiki-markup

When precise computation is necessary, carefully and methodically estimate the maximum cumulative error of the computations, regardless of whether decimal or binary is used, to ensure that the resulting error is within tolerances. Consider using numerical analysis to properly understand the problem. An introduction can be found in \[ [Goldberg 1991|AA. Bibliography#Goldberg 91]\].

Noncompliant Code Example

...

ISO/IEC DTR 24732

Bibliography

...

\[[Goldberg 1991|AA. Bibliography#Goldberg 91]\] \[[IEEE 754 2006|AA. Bibliography#IEEE 754 2006]\]
[IEEE 754 2006]

...

FLP01-C. Take care in rearranging floating point expressions      05. Floating Point (FLP)      FLP03-C. Detect and handle floating point errors