Using NetLogo’s Built-in Analysis Tools (20 min)

NetLogo provides powerful tools for data collection without programming complex analysis routines.

Monitors: Real-Time Data Tracking¶

Monitors show live values as your model runs:

; In interface, create monitor that shows:
count turtles with [carrying-food?]
; Label it: "Ants with Food"

mean [pheromone] of patches with [pheromone > 0]
; Label it: "Average Pheromone"

ticks
; Label it: "Time Steps"

Strategic monitor placement:

• Outcome variables: Your main research question measures
• Sanity checks: Total population, conservation laws
• Process indicators: Rates of change, intermediate states

Plots: Visualizing Change Over Time¶

Plots track multiple variables and show patterns:

Basic setup in Interface tab:

1. Click “Plot” button to create new plot
2. Name it (e.g., “Population Dynamics”)
3. Set X and Y axis labels and ranges
4. Add “pens” for different variables

Code to update plots:

to update-plots
  ; This automatically runs every tick
  
  ; Plot 1: Population by type
  set-current-plot "Population Dynamics"
  set-current-plot-pen "Cooperators"
  plot count turtles with [strategy = "cooperate"]
  
  set-current-plot-pen "Defectors"  
  plot count turtles with [strategy = "defect"]
  
  ; Plot 2: Average satisfaction
  set-current-plot "Satisfaction Levels"
  set-current-plot-pen "satisfaction"
  plot mean [satisfaction] of turtles
end

Histograms: Understanding Distributions¶

See how values are distributed across your population:

; In interface, create histogram
; Set pen update to:
histogram [age] of turtles

; Or show distribution of some other property:
histogram [count link-neighbors] of turtles  ; Network degree
histogram [money] of turtles                  ; Wealth distribution
histogram [satisfaction] of turtles           ; Satisfaction levels

Activity: Build Analysis Dashboard¶

Create a complete monitoring system:

1. Add 3-4 monitors tracking key variables in your model
2. Create 2 plots showing how important things change over time
3. Add 1 histogram showing distribution of agent properties
4. Run your model and observe what patterns emerge

BehaviorSpace: Systematic Experiments¶

BehaviorSpace runs multiple experiments automatically with different parameter combinations.

Setting up experiments:

1. Tools menu → BehaviorSpace
2. Click “New” to create experiment
3. Vary parameters systematically:

["tolerance" [0.1 0.2 0.3 0.4 0.5]]
["population-size" [100 200 500]]

4. Set number of repetitions (e.g., 10 runs per combination)
5. Choose which variables to measure:

count turtles with [color = red]
segregation-index
mean [satisfaction] of turtles

6. Run experiment (can take a while!)

BehaviorSpace generates spreadsheet with all results for further analysis.