On Wed, Feb 29, 2012 at 12:03 PM, Arno Onken wrote:
> Hi Juan,
>
>> Is there any reason to have statistics-1.1.0/combnk.m which seems a
>> slower implementation of combinatorics-1.0.9/combs.m?
>
> Thanks for doing this comparison. It looks like combnk does things in a
> MATLAB compatible way, whe
Hi Juan,
> Is there any reason to have statistics-1.1.0/combnk.m which seems a
> slower implementation of combinatorics-1.0.9/combs.m?
Thanks for doing this comparison. It looks like combnk does things in a
MATLAB compatible way, whereas combs does not. combnk handles strings
differently and app
Hi,
Is there any reason to have statistics-1.1.0/combnk.m which seems a
slower implementation of combinatorics-1.0.9/combs.m?
Maybe is a bad idea to have statistics depending on combinatorics?
Quick benchmark:
octave> t0=cputime(); combs(1:10,5); cputime()-t0, t0=cputime();
combnk(1:10,5); cput
