Actions

Spatial-Stochastic: Difference between revisions

From Santa Fe Institute Events Wiki

(Ideas for Stochasticity and Space)
 
No edit summary
Line 1: Line 1:
= Stochasticity and Space Project =  
= Stochasticity and Space Project =  


I (Chloe) remember Stochasticity and Space as the minimal set of what today's pre-dinner discussion decided to concentrate on, and as I recall it we agreed to write up what we were thinking about as explicitly as possible so that we could see exactly what we we have to weld into a project. Explicit hypotheses, links to papers, definitions of what state variables are tracked, definitions of traits w.r.t. space variables, sample code?  
==Things we Need:==
* Explicit hypotheses
* links to papers,  
* state variables  
* definitions of agent traits w.r.t. space variables,  
* forcing conditions
* sample code?  


I have a square-grid cellular automaton with agents running in Python that I can share, with the skeleton of an experimental `runner' (load an initial condition, specify the algorithms for the updates of /n/ state variables, run for specified time, save result).
Looks like we can do anything on a square grid in NetLogo, and even import GIS data.
 
==Proposed: ==
 
===State variables:===
 
* Limiting resource (which can also be in excess)(E.g., water)
* Patch/cell/region ability to conserve resource (E.g., soil organic matter)
* Patch connections (e.g., elevation -> horizontal flow)
 
===Agent characteristics:===
 
* Efficiency in capturing resource
* Efficiency in surviving low-resource conditions
* R-k strategy
* Spatial reach, for getting resource or spreading offspring
* Interactions? One kind of agent more efficient in the presence of another?
 
===Forcing conditions:===
* Input of resource, variable in space or time
* Variance of input
 
===Hypotheses:===
* Variable resource input favors different agent traits than reliable inputs with the same mean value
* Agents develop spatial variability to gather sparse resources
* Unreliable resources favor cooperative strategies (Demetrius, e.g.)
* Hubler's hypotheses: with little noise, the system minimizes dissipation and is efficient; with much noise, the system maximizes dissipation and is robust.
 
===Papers:===

Revision as of 21:59, 6 June 2012

Stochasticity and Space Project

Things we Need:

  • Explicit hypotheses
  • links to papers,
  • state variables
  • definitions of agent traits w.r.t. space variables,
  • forcing conditions
  • sample code?

Looks like we can do anything on a square grid in NetLogo, and even import GIS data.

Proposed:

State variables:

  • Limiting resource (which can also be in excess)(E.g., water)
  • Patch/cell/region ability to conserve resource (E.g., soil organic matter)
  • Patch connections (e.g., elevation -> horizontal flow)

Agent characteristics:

  • Efficiency in capturing resource
  • Efficiency in surviving low-resource conditions
  • R-k strategy
  • Spatial reach, for getting resource or spreading offspring
  • Interactions? One kind of agent more efficient in the presence of another?

Forcing conditions:

  • Input of resource, variable in space or time
  • Variance of input

Hypotheses:

  • Variable resource input favors different agent traits than reliable inputs with the same mean value
  • Agents develop spatial variability to gather sparse resources
  • Unreliable resources favor cooperative strategies (Demetrius, e.g.)
  • Hubler's hypotheses: with little noise, the system minimizes dissipation and is efficient; with much noise, the system maximizes dissipation and is robust.

Papers: