Re: Very slow filter.cc

[akira]

Thanks for your comment, David,

From: David Bateman <address@hidden>
>What was going to be my first step was to compile three different 
>versions of octave.
>These being 2.1.50, 2.1.57 and 2.1.64, which correspond to the last 
>three testing releases. The first before NDArray's were integrated, the 
>second when NDArray's were integrated, but the filter function wasn't 
>changed to use them and the last with the filter function converted to 
>NDArray's. Getting the times of the test case you propose for these 
>three build, would really confirm your problem.

I tried octave-2.1.50 and octave-2.1.57. It took the same time as
octave-2.1.64. It means NDarray support in both filter.cc and octave
is not the slowing factor, is it right?

Also, I carefully compared filter.cc of octave-2.1.50 and
octave-2.0.17. It seems there is no difference in filtering algorithm
itself. I realized that the slow filter.cc comes from more essential
part of octave-2.1.x.

_/_/_/  _/  _/            _/            _/_/        Akira NISHIMURA
 _/    _/  _/            _/             | Dept. of Media and Cultural Studies
_/okyo _/_/niversity of _/nformation _/_/ciences address@hidden
                   http://www.rsch.tuis.ac.jp/~akira



-------------------------------------------------------------
Octave is freely available under the terms of the GNU GPL.

Octave's home on the web:  http://www.octave.org
How to fund new projects:  http://www.octave.org/funding.html
Subscription information:  http://www.octave.org/archive.html
-------------------------------------------------------------