Hi tor, 16 07 2009 kl. 17:34 +0200, skrev Christian Neumair: > the attached patch fixes qamdemod(). Its purpose is to provide a > nearest-neighbour quantizer against a pre-defined set of points (called > QAM signal space constellations).
This patch looks good to me, so I applied it. > It still features a loop which is not pre-compiled > > for k = 1:numel(y) > [n z(k)] = min(abs(y(k) - x)) > end > > but the only way that this could be vectorized is by converting the > input data to a vector, processing it and converting it back to a > matrix/multi-dimensional array again with something like > > a=y(:); > b=min(abs(repmat(a,[1 m]) - x)) > z=reshape(x,size(y); > > and I am not sure whether this is wise from a speed and memory usage > POV. Maybe this nearest-neighbour quantization loop could/should be > moved into a builtin function? I'm not a user of this package, so I'm probably asking a stupid question here, but does this constitute a performance bottle-neck for real applications? If so, then it might be sensible to a C++ version of this code. From a maintenance point-of-view, I'd rather like to avoid this though... Søren ------------------------------------------------------------------------------ Enter the BlackBerry Developer Challenge This is your chance to win up to $100,000 in prizes! For a limited time, vendors submitting new applications to BlackBerry App World(TM) will have the opportunity to enter the BlackBerry Developer Challenge. See full prize details at: http://p.sf.net/sfu/Challenge _______________________________________________ Octave-dev mailing list Octave-dev@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/octave-dev
