What
is it? |
This
project explores an artificial society known as Sugarscape.
This society was conceived of by Joshua Epstein and Robert Axtell
(http://www.brook.edu/SUGARSCAPE).
In this society, agents (red) wander around on a landscape consisting
of two connected mounds of sugar (yellow). As the agents move
around on the landscape they metabolize some of their existing
sugar. The agents are constantly looking for new sugar, and
when they find it they add it to their resources. If the agents
ever eat all of their sugar, then they will die. If they get
enough sugar, then they can reproduce. Each agent has a random
value for its metabolism, initial quantity of sugar, and vision
level (that controls how far around it can look for more sugar).
The sugar itself can grow back at different rates. For a more
thorough description of Sugarscape, check out the Sugarscape
web site (http://www.brook.edu/SUGARSCAPE). |
How
to Use It |
Click
the SETUP button to setup the agents (red). Click the
GO button to start the simulation.
The N-AGENTS slider controls the initial number of agents.
The HATCH-THRESHOLD slider sets the sugar level at which
the agents reproduce. The CAPCITY-CONSTANT slider controls
how much sugar there is initially. The MILLI-ALPHA slider
controls how fast the sugar grows back if it is not growing
back instantly (controlled by the INSTANT-GROWBACK? slider).
The BLACK-IS and WHITE-IS sliders control the color coding
of the sugar. The BIRTH slider controls whether (1) or
not (0) the agents can give birth. The SHOW SUGAR button
colors the turtles based on how much sugar they have.
The SHOW VISION button colors the turtles based on how
great their vision is. The SHOW METABOLISM button colors
the turtles based on the value of their metabolism. |
 |
|
Things
to Notice |
Watch the
COUNT-TURTLES monitor to see how the agent population changes
over time. You can also watch the AVERAGESUGAR, AVERAGEVISION
and AVERAGEMETABOLISM monitors to see how the average values
of these turtle traits change over time. The plot window graphs
the total number of agents, as well as these average values,
over time. The values of these traits can be influenced by the
environmental conditions. Sometimes coniditions will strongly
select for longer range vision, or lower metabolism.
|
Explorations |
Under the
current setting, the agents can survive and reproduce. Is the
current Sugarscape capable of supporting an infinite number
of agents, or is there a limit? How is that limit determined?
How does the growback rate affect turtle vision and metabolism?
How are these two quantities related in the population of turtles
and within individual turtles? There are many extensions of
this model that are described in the Artificial Societies book.
Can you implement some of them? |
StarLogo
Features |
The turtles
in Sugarscape are all called from an "ask-turtles"
inside of an Observer procedure. This keeps the turtles in sync
with each other. This also keeps them in sync with the patch
procedures that are growing back the sugar.
The turtles also make use of the "leap" command. This
command is like the "jump" command, except that turtles
will not execute it unless there are no other turtles on the
patch they are trying to get to. By calling leap within the
recursive turtle procedure "relocate", it is ensured
that no two turtles stay on the same patch. |