Re: An example of poor C++ performance

[Todd Neal]

John W. Eaton wrote:
> On 23-Feb-2005, Steve C. Thompson <address@hidden> wrote:
>
> | It sounds like you are suggesting that the C++ code needs to be
> | vectorized.  But isn't this a compounded hassle?
>
> We could have the octave_rand::scalar() function not check to see
> if the generator has been initialized (and require to you to ensure
> that that has happened) but that would allow you to do even more silly
> things (like try to get the next number in the sequence before the
> generator was properly initialized).
>
> The default generator in Octave is known to be slow.  We have
> considered changing it, but no decision has been made.
I looked at the source and I am not for certain but it could be the 
layer upon layer of function call and return that happens when 
generating a number.

Considering a uniform_dist,

(1) octave_rand::scalar() calls dgenunf
(2) dgenunf calls genunf
(3) genunf calls ranf
(4) ranf calls ignlgi
(5) ignlgi calls a few functions, but essentially generates the number


Todd




-------------------------------------------------------------
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
-------------------------------------------------------------