On Feb 25, 11:07 am, [EMAIL PROTECTED] wrote:
> On Mon, 25 Feb 2008, Carlo Hamalainen wrote:
> > On Mon, Feb 25, 2008 at 6:20 AM, <[EMAIL PROTECTED]> wrote:
> >> Dirty, William. I can't believe you blame this on me -- that was all
> >> Robert's fault. Anyway. I've co-opted Ajanki's framework, and have
> >> rewritten the core of the search algorithm to be iterative. And, without
> >> a single goto! I'm going to test this some more, and update my patch for
> >> #2271 in a few hours. Seems to work flawlessly, and I can color K_22 in
> >> less than 8 seconds (not impressive, no, except that previously,
> >> attempting to color K_14 exceeded recursion limit).
>
> > Are you still keen for me to write a wrapper for my C++ code? I got
> > most of the way on the weekend...
>
> While I put a fair amount of work into making the search iterative, my pride
> doesn't get in the way -- if your code is better, then it goes in.
>
> What is "better" is a bit debatable: faster is a big portion, but quality of
> code and documentation is big, too... and these days, documentation is
> getting more attemtion than code. As such, I've put as much effort into
> documenting Ajanki's code as I have into making it work better. Also, a side
> effect of my effort is that the code is now ready to be rewritten in Cython,
> so easy to maintain.
>
> Your code is probably a million times faster than mine (if we convert
> Ajanki's code to Cython, yours will probably remain faster, but less so).
> How's the documentation? Doctests?
>
> Here's what I've got so far. I'm inclined to say, I should finish
> documenting this file (2 or three more routines needed) so we can include
> this in the next release. Or, if your code is all documented & ready to go,
> we can go with that; no skin off my back.
>
> http://sage.math.washington.edu/dlx.py
I would also suggest to go that way since we can then merge the ticket
dependent on it. Once we have the correctly, but not blazingly fast
version in Sage we can always switch to the C++ version as it is
convenient for the integrators. Please post a patch with the
documented code once it is ready, so we can merge it for
2.10.3.alpha0.
Cheers,
Michael
--~--~---------~--~----~------------~-------~--~----~
To post to this group, send email to sage-devel@googlegroups.com
To unsubscribe from this group, send email to [EMAIL PROTECTED]
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---
