
DoubleToFix |
ptolemy.actor.lib.conversions.DoubleToFix |
This actor converts a DoubleToken to a FixToken with a specified
precision. Note that this conversion is lossy, in that the output
is an approximation of the input. The approximation can be
constructed using a variety of rounding and overflow strategies.
The precision of the output is given by the precision parameter,
which is an integer matrix of the form [m, n], where
the total number of bits in the output is m, of which
n are integer bits. The default precision is [16, 2], which means
that an output has 16 bits, of which 2 bits represent the
integer part.
The rounding strategy is defined by the rounding parameter and
defaults to nearest (or half_floor), selecting the nearest
representable value. The floor value nearer to minus infinity is used
for values half way between representable values. Other strategies
such as truncate are described under ptolemy.math.Rounding.
The overflow strategy is defined by the overflow parameter and
defaults to saturate (or clip). Out of range values are
saturated to the nearest representable value. Other strategies
such as modulo are described under ptolemy.math.Overflow.
Author(s): Bart Kienhuis, Contributor: Edward A. Lee, Ed.Willink,
Version:$Id: DoubleToFix.java,v 1.59 2006/05/08 05:28:15 neuendor Exp $
Pt.Proposed Rating:Green (pwhitake)
Pt.Accepted Rating:Green (pwhitake)
precision
The precision of the output fixed-point number, which is represented
by a 2-element integer matrix.