|
float
The term float[(p)] describes a
datatype,
or a
decimal datatype.
The flag "p" is an abbreviation for "precision" and, in this context, is synonymous to accuracy. The flag "p" is an optional parameter. The value of the flag "p" must be less or equal to 48 (p <= 48). Like all decimal datatypes, the datatype float is also ANSI compatible. The datatype float[(p)] requires either 4 or 8
bytes of
storage.
The values which can possibly be stored are decimal numbers with machine dependent ranges of values.
Usage of the datatype float[(p)]
The datatype float[(p)] is always used if rounded values cannot cause problems.
In this case more accurate datatypes should be used, because with decimal datatypes the places after the decimal point can be rounded, depending on the method of calculation used. The
ASE
Client programme
isql
for example only displays an accuracy of 6 digits after the decimal point. The following digits are being rounded up or down respectively.
See also:
real.
|
