Emergence

EES 4760/5760

Agent-Based & Individual-Based Computational Modeling

Jonathan Gilligan

Class #9: Tues. February 6 2018

Make-up class

Make-up class

Based on doodle poll, the make-up class will be held on

  • Monday Feb. 19 from 6:00–7:15 PM.
  • Place to be announced.

Team Projects

Team Projects

  • For next Tuesday: Read Chapter 10 and the ODD of the model you will work on. You will spend significant time in class working with your partner(s) to start turning the ODD into a working NetLogo model.

Emergence

Emergence

Flocking

  • Play with the model.
    • Adjust the parameters and see how they change the flocking behavior

Flocking Model Overview

  • Entities:
    • Birds: state-variables flockmates, nearest-neighbor
  • Process:
    • Each bird identifies its flockmates
    • Each bird adjusts its direction
    • Each bird moves forward one patch

Flocking Model Design Concepts

  • Emergence: Large flocks emerge from each bird acting independently, looking only at nearby birds.
  • Adaptation:
    • If the nearest-neighbor is too close, the bird separates by turning away from it.
    • Otherwise, the bird:
      1. aligns: turns toward its flockmates
      2. coheres: turns slightly toward the direction the rest of its flockmates are flying.
  • Sensing: The bird can only see a certain distance (vision)
  • Interaction:
    • Each bird interacts with its flockmates

Submodels

  • find-flockmates:
    • flockmates are all birds within vision distance
    • Alternate interactions:
      • flockmates interacts with 6 nearest birds, regardless of distance.
      • Bird only interacts with nearest member of flockmates
  • separate: Turn away from nearest-neighbor by up to max-separate-turn
  • align: Turn toward center of flockmates by up to max-align-turn
  • cohere: After aligning, turn toward average direction flockmates are flying, by up to max-cohere-turn

Observations:

  • How to measure flock formation?
count turtles with [any? flockmates]
mean [count flockmates] of turtles
mean [min [distance myself] of other turtles] of turtles
standard-deviation [heading] of turtles

Digression: Selecting Turtles

  • Selection primitives:
    • Returning agent-sets
      • n-of, min-n-of, max-n-of, other,
      • turtles-on, turtles-at, turtles-here, at-points
      • in-radius, in-cone,
      • with, with-min, with-max
    • Returning individual turtles
      • one-of, min-one-of, max-one-of
      • (may return nobody)
    • Look at Agentset category in NetLogo dictionary
  • Be careful:
    • Some primitives expect agent-sets
    • Others expect individual turtles.

Practice Selecting Turtles

  • Turn 5 turtles red:

    ask n-of 5 turtles [ set color red ]
  • Now for each of those turtles, select all the turtles within a radius of 5 and turn them green

    ask turtles with [color = red] 
    [
      ask other turtles in-radius 5 [ set color green ] 
    ]
  • Now ask each green turtle to calculate the distance to the closest red turtle

    show [
      min [distance myself] of turtles with [color = red]
      ] of turtles with [color = green]
  • Now get the average over all the green turtles of the distance to the closest red turtle

    show mean [
      min [distance myself] of turtles with [color = red]
      ] of turtles with [color = green]          

Experiments

Experiments

  • Create a Behaviorspace experiment and call it “Baseline”
    • change one parameter and see how it affects the various measures of flocking.
  • Next, duplicate “Baseline” and call it “Flock Type”
    • vary that parameter while also varying the flock-type
  • Next, duplicate “Baseline” and call it “Multiple”
    • vary more than one parameter (e.g., vision and max-cohere-turn or max-align-turn)
  • Use the analyze_behaviorspace app at https://analyze-behaviorspace.jgilligan.org/ to graph the output from your BehaviorSpace experiments.
  • Try creating a summary table, saving it to your computer, and opening it in Excel.