[galib] minimization of objective function problem

[Jose daVeiga]

Hi, 

I happen to have the same problem.  

When I tried ga.minimize() right after initializing the ga it didn’t
consider the individuals with the least score as the fittest (I am using
GARealGenome, could this be a problem with this specific class?).  

I resorted to using the inverse value for my objective as someone else
suggested...but it is a hassle and not very clear.

Any suggestions?

Thanks, 

José daVeiga

PhD Candidate  -   Architecture,  Technology
Department of Architecture and Urban  Design
University of California, Los Angeles (UCLA)
------------------------------------------------------------
2nd Year Studio Instructor      (Part-Time Lecturer ’04/’05)
Department of Architecture
Cal Poly Pomona              College of Environmental Design
------------------------------------------------------------
(310) 415-9893(C)                          jfernand at ucla.edu
http://www.bol.ucla.edu/~jfernand
------------------------------------------------------------
Principal & Managing Partner        Architecture Convolution
(310) 593-4860 (W)                        (310) 593-4862 (F)
http://www.convolution.ws                jose at convolution.ws
------------------------------------------------------------
Master of Arts in Architecture       (UCLA '03 - Technology)
Master of Architecture                  (UCLA '98 - MarchII)
Professional Degree in Architecture (Lusiada University '96)
LEEDtm Accredited Professional                         ('04)
Licensed Architect in Portugal                         ('97)
------------------------------------------------------------

-----Original Message-----
From: Anthony Liekens [mailto:a.m.l.liekens at tue.nl] 
Sent: Monday, June 06, 2005 4:47 AM
To: Paulo Almeida
Cc: galib at mit.edu
Subject: Re: [galib] minimization of objective function problem

Once you have initialized your ga, you can issue

    ga.minimize();

for minimization problems.

Anthony,-

Paulo Almeida wrote:

>Hi,
>
>I have just started using galib, and I have no experience at all, but
couldn't 
>you try doing that mathematically, like the previous poster suggested,
using 
>the symmetric or inverse value of what your function currently returns?
>
>-- Paulo
>
>
>On Monday 06 June 2005 11:23, #SAWHNEY ADITI# wrote:
>  
>
>>Hi,
>>I am using teh latest version of galib with vc+= and have the same
problem
>>as ther person below. Ie  I want to minimize my objective function.,
ie a
>>memebr of population iwth lower value of the float thast the objectiev
>>function returns should be conisdred "fitter". how can i do this?
please do
>>help.
>>
>>yours sicnerley,
>>aditi.
>>    
>>
>_______________________________________________
>galib mailing list
>galib at mit.edu
>http://mailman.mit.edu/mailman/listinfo/galib
>  
>