swarm-modeling
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RES: Interfacing GALib to Swarm


From: Marcus Vinicius Pereira Pessoa
Subject: RES: Interfacing GALib to Swarm
Date: Fri, 10 Oct 1997 08:46:34 -0300

Hi, 

I'm just a newcomer into the Objective-C (and even plain C) and into the 
Swarm's world. Having a similar 
problem of choosing a GA to use with Swarm, I also evaluated the possibility of 
porting GALib to a Swarm 
application.

But studying more carefully the Breeder Lib I thought it would worth a try to 
change it, once it provides a good 
sketch of a general GA. So I tried the approach #1 suggested by Merelo, and it 
wasn't a painful job to change 
the BinChromosome into a GroupChromossome, in which I use lattices instead of 
bytes.

I think that an outsider opinion (Merelo is the Breeder´s father) could be 
valuable.

Vinicius. 

_______________________________________________________________

Marcus Vinicius Pereira Pessôa
1º Ten.-Av.
Centro de Computação da Aeronáutica de São José dos Campos
Centro Técnico Aeroespacial
Work: 055 012 340 5808
Home: 055 012 332 8150
FAX  : 055 012 340 5817
Email: address@hidden
_______________________________________________________________


----- Mensagem original -----
De:             J.J. Merelo Guervos [SMTP:address@hidden
Enviada em:             Sexta-feira, 10 de Outubro de 1997 04:23
Para:           address@hidden
Assunto:                Interfacing GALib to Swarm

>>>>> "Rob" == Rob Kewley <address@hidden> writes:

Hi,

    Rob> While the existing Breeder-1.0 library does provide this
    Rob> capability, it restricts the user to binary representations

It does not. You can implement different representations by
1. Implementing the protocols in a different way.
2. Subclassing the Chromosome
3. Writing a new one altogether, and using the Population class


    Rob> more versatility.  GALib from MIT, provides this versatility

    Rob>        see http://lancet.mit.edu/ga/

It does so through templates. Which means that, once again, you have
to write your own class, with the restrictions that templates impose
on it (a restriction very similar to the one imposed by protocols).

But, going back to the principles, there's not much that can be done
with non-binary chromosomes that can be done with other kind of
chromosomes.

You could also try http://kal-el.ugr.es/~jmerelo/GAJS.html (in
JavaScript, probably easier to port to ObjC), and
http://kal-el.ugr.es/GAGS (once again, only binary representation).

See ya!

                        JJ

-- 
JJ Merelo                         | http://kal-el.ugr.es/htbin/jj-plex
Grupo Geneura ---- Univ. Granada  | http://kal-el.ugr.es/



                  ==================================
   Swarm-Modelling is for discussion of Simulation and Modelling techniques
   esp. using Swarm.  For list administration needs (esp. [un]subscribing),
   please send a message to <address@hidden> with "help" in the
   body of the message.
                  ==================================

<<application/ms-tnef>>


reply via email to

[Prev in Thread] Current Thread [Next in Thread]