Final Project — Swarm Navigation

Setup

python -m venv .venv
source .venv/bin/activate  # Windows: .venv\Scripts\activate
pip install -r requirements.txt

Run (simulate all permutations)

python run_all.py

Outputs in results/:

• summary.csv – success/fail, reason, steps
• runs/perm_###.json – small per-permutation metadata
• traces/perm_###.json – pretty JSON trajectory (0-based indices)

Visualize saved runs (no simulation)

python visualize.py

Writes PNGs to results/figures/perm_###.png.

Can edit which pngs are created by setting PERM_SELECTION in visualize.py.

Rules enforced

• Boundary (unit circle): center-only ⇒ fail if ‖x‖ > 1 (‖x‖ == 1 allowed).
• Obstacles: outer-edge ⇒ collide if ‖x − c_obs‖ ≤ R_obs + r.
• Robot–Robot: outer-edge ⇒ collide if ‖x_i − x_j‖ ≤ 2r.

Knobs (config.py)

CONFIG = {
  "n": 5, "r": 0.05, "dt": 0.005, "vmax": 0.05, "tol": 0.005,
  "max_steps": 30000, "results_dir": "results",
  "show_progress": True, "progress_every": 5,
}

Clean re-run:

rm -rf results && python run_all.py && python visualize.py