TRS-SO
MODEL III
Accuracy
Single-precisioncalculations involving + , ,
*,
and /are accurate to six significant
digits; double-precision calculations involving the same operations are accurate to
16
significant digits.
The exponentiation operator
(l) (displayed as
"[")
is
single-precision.
The trigonometric and logarithmic functions are single-precision; other functions
have a precision depending on the input argument and on the function. For
example,
CDBL
returns a double-precision value; ABS returns a value with the same
precision as the input argument.
When converting from single- to double-precision, use the following technique to
avoid introduction
of
incorrect values in the extra digits
of
precision:
double-precision variable
= VAL ( STR$ ( Single-precision variable))
A/18
To Next Page
Loading...
