Newsgroups: bionet.population-bio,comp.theory.cell-automata,comp.ai.alife
Path: cantaloupe.srv.cs.cmu.edu!fs7.ece.cmu.edu!hudson.lm.com!godot.cc.duq.edu!newsfeed.pitt.edu!gatech!howland.reston.ans.net!math.ohio-state.edu!uwm.edu!caen!news.tc.cornell.edu!travelers.mail.cornell.edu!newstand.syr.edu!galileo.cc.rochester.edu!prodigal.psych.rochester.edu!stevens
From: stevens@prodigal.psych.rochester.edu (Greg Stevens)
Subject: Simulated Environments Parameter Question
Message-ID: <1994Dec5.205450.28653@galileo.cc.rochester.edu>
Sender: news@galileo.cc.rochester.edu
Nntp-Posting-Host: prodigal.psych.rochester.edu
Organization: University of Rochester - Rochester, New York
Date: Mon, 5 Dec 94 20:54:50 GMT
Lines: 35
Xref: glinda.oz.cs.cmu.edu comp.theory.cell-automata:3064 comp.ai.alife:1427

Okay, I have a simple environment which has a world which is a matrix of
integers WORLD-SIZE by WORLD_SIZE, which at each position can either have
an agent, a bit of food or blankness.

Agents move around and when they hit a piece of food their energy gets
incremented by FOOD_VAL, otherwise every move costs them 1 energy point
(energy isa decremented every action).

If ENERGY gets above (INITIAL_ENERGY + REPROD_THRESH) then it reproduces
and the energy is lowered again to (REPROD_THRESH).

New food apears on the matrix at a rate of REGEN_RATE for each space.

Now, I have found that:    

If REGEN_RATE is too low, trhey all starve and don't reproduce.
If REGEN_RATE is too high, they al;l reproduce and there is no selection
pressure for good food-searching strategies (movenent is controlled by
the chromosome).

If INITIAL_AGENT_NUM is too high with respect to WORLD_SIZE, they
"over-graze" and starve.

If REPROD_THRESH is too big, not enough reproduce and if it is too small
too many reproduce.

Etc.

Does anyone know of any hints, other than methodical trial and error,
for discovering good combinations of values for these parameters?

Greg Stevens

stevens@prodigal.psych.rochester.edu

