Newsgroups: comp.ai.genetic
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news4.ner.bbnplanet.net!news3.near.net!paperboy.wellfleet.com!news-feed-1.peachnet.edu!usenet.eel.ufl.edu!news.mathworks.com!newshost.marcam.com!usc!math.ohio-state.edu!news.cyberstore.ca!skypoint.com!umn.edu!news
From: "James A. Larson" <larso171@maroon.tc.umn.edu>
Subject: Re: Sugal V2.0 User Group (one person user group)
To: thasan@mayfair.demon.co.uk
Message-ID: <24496.larso171@maroon.tc.umn.edu>
X-Minuet-Version: Minuet1.0_Beta_16
Sender: news@news.cis.umn.edu (Usenet News Administration)
Nntp-Posting-Host: dialup-3-183.gw.umn.edu
X-Popmail-Charset: English
Organization: University of Minnesota, Twin Cities
Date: Wed, 6 Sep 1995 08:22:58 GMT
Lines: 69

Hi Thasan,

I wrote Dr. Anthony Hunter a couple of times in the past 2 weeks.  

He has agreed to put a Windows executable (.exe) at his site, so that a 
person can run it without compiling first.  I presume it will be his Dejong 
functions program.  One can't model their own problem without writing their 
own evaluation function and compiling.  But at least with the Dejong 
executable, one can see what the interface looks like (excellent, 
marvelous graph), and do a lot of interesting experiments on the Dejong 
functions.

He didn't say when he would put the executable on his ftp site, but I 
presume its there now.

Per his correspondence with me, he has very recently put in the multiple 
parallel genetic operators capability that I mentioned in my previous 
posting: 

  "I also requested support for multiple parallel genetic operators (e.g. 
   where the user specifies a 20% probability of choosing crossover 
   operator 1, a 25% probability of chossing crossover operator 2, ... a 
   15%    probability of choosing mutation operator 1, a 20% probability 
   of choosing mutation operator 2, ..., and so on."  
 
This capability is in his Sugal V. 2.1 which he is still developing (I 
don't have it, by the way).  He didn't indicate when he will finish Sugal 
V. 2.1 and make it available.

>I would be interested in that too. Since my algoritms is a uni- crossover
>type and I need to change the crossover rate at run time. I am also working
>along in that line and would be glad to exchange ideas. I also welcome any
>suggestions in this regard.

It might be easy to do -- he has built-in the ability to change the 
mutation rate based on the generation counter.  He might have that 
capability too for the mutation rate and crossover rate.  But I don't know 
without taking a long look at it.  I'm methodically studying the code and 
his programming manual so as to understand the extensibility features 
better.

I'm not getting my own ftp site.  However, I plan on putting together my 
own "pseudo-ftp-site" where I organize a collection of the few items that I 
have that are of value to others.  And to resolve the problem with my 
Internet supplier about why I'm sometimes having problems mailing large 
files.  When that problem is resolved, and with my few "great works" 
organized, it will be very simple for me to email items on request.

In the meantime, he does take user contributions.  He has agreed to store 
at his ftp site a couple of text file documents I've written (1) On how the 
bit strings in the Dejong function F1 are converted to a set of real 
variables (2) How the parameter setting (SuSetParameter) works (3) A 
diagram of the pool data structure.

He also said he hopes to start a mailing list sometime this week.

A couple of other items:

  There's an error in Dejong Function F2.

  Sugal V 2.1 will have ordering problem capabilities like for the 
Traveling Salesman Problem (TSP).  Per his correspondence, it even 
graphically displays a map of the thing.

  He plans on adding multiple population support, and coevolution in Sugal 
V. 3.0 (no indication of when that will be).

Jim Larson

