[Help-gsl] Re: for those who care about siman

[Jakub Narebski]

Marco Maggi wrote:
> "Liam Healy" wrote:
>>
>> I did find that the C example worked, but I could not get the
>> equivalent computation to work from Lisp.
> 
> The algorithm is so simple that you can write the thing directly
> in Lisp in a couple of days. Start with the page at Wikipedia,
> then you can read my code which is a little clearer than the
> one in GSL but does the same things.
> 
>> If it's possible to influence the design, I'd like to
>> suggest that printing be separate from the main library function.
> 
> My opinion is that the GSL siman module itself does not follow
> the GSL coding guidelines: it should implement an interface
> like the one of the minimisation modules, with an iterate
> function that we have the responsibility of invoking again and
> again; with such an interface we could be free to inspect
> the current solution and decide what to do and what to log. 

I wonder if it would be possible to implement ASA (Adaptive Simulated
Annealing) algorithm:
  http://www.ingber.com/#ASA
  http://asa-caltech.sourceforge.net/

in GSL... I couldn't get GSL siman to work as I wanted...

-- 
Jakub Narebski
Warsaw, Poland
ShadeHawk on #git