home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
The Epic Interactive Encyclopedia 1998
/
Epic Interactive Encyclopedia, The - 1998 Edition (1998)(Epic Marketing).iso
/
Menufiles
/
Vin-archive
/
MessySUBS.LHA
/
MoreMESSY_Subs
/
Floating-point_notation
/
INFOTEXT
< prev
Wrap
Text File
|
1992-09-03
|
791b
|
20 lines
System of representing numbers as multiples
of the appropriate base raised to some power;
for example, the number 97.8 can be expressed
as 0.0978 x 10^3 or 9780 x 10^-2. In
computing, the number is expressed as a
decimal fraction, so 97.8 becomes 0.978 x
10^2. See fixed-point notation. In a
computer, the numbers are represented by
pairs; 97.8 = (.978, +2). The first number of
the pair is called the mantissa and the
second the exponent. The definition applies
equally to numbers expressed to a different
base (for example, binary fractions muliplied
by two raised to some power). The advantage
of floating-point notation is that very large
and very small numbers can be expressed with
a few digits. Thus (.978, +18) written out in
full would require 15 digits (978 followed by
18 zeros).