add painted desert model

[Mallory3125]

Summary

This PR adds the Painted Desert Challenge example adapted from NetLogo to mesa-examples. It demonstrates emergent sorting behavior and self-organization using Mesa agents and Solara visualization.

Motive

The Painted Desert Challenge model is a valuable addition to Mesa as it:

• Illustrates emergent behavior: Shows how simple local agent rules lead to complex global patterns (sorted colored chip piles).
• Demonstrates self-organization: Agents organize their environment without central coordination, merging smaller piles into larger ones.
• Teaches decentralized systems: No single agent is in charge, yet the system as a whole accomplishes a sophisticated task.
• Showcases Mesa’s spatial capabilities: Uses OrthogonalMooreGrid with property layers for chip colors, and Solara-based visualization for dynamic spatial modeling.

Implementation

The model simulates chip collector agents (representing termites) moving on a grid with colored wood chips, sorting them into distinct piles by color.

Model Structure:

• ChipCollectorModel (painted_desert/model.py):

  • Initializes a grid with patches (cells) and a pcolor property layer for chip colors (0 = empty, 1+ = colored chips).
  • Populates the grid with chips based on density and creates agents at random positions.

• ChipCollector Agent (painted_desert/agents.py):

  • Behavior rules:
    1. find_chip: Wander until finding a chip of assigned color, pick it up, and move away.
    2. find_new_pile: Locate another chip of the same color to identify a pile.
    3. find_empty_spot: Deposit the chip near the pile in an empty adjacent cell.
    4. wiggle/get_away: Random movement functions for exploration and relocation.

• Visualization(app.py):

  • Renders agents (white when empty-handed, colored to match carried chips) and chip patches.
  • Adjustable parameters via sliders: grid size, chip density, number of agents, and number of chip colors.

Usage Examples

To run the example:

solara run examples/painted_desert/app.py

[EwoutH]

Thanks for the PR!

@quaquel could your review the use of the discrete grid in this example model?