📡 Info Diffusion Simulator

A toy model for exploring how information from a video spreads through a population — like a zeitgeist ripple moving outward from the first viewers.

🌐 Live demo →

No install, no Python — the simulator runs entirely in your browser.

Concept

Uses a SIR-style epidemiological model adapted for information spread:

• S (Susceptible): People who haven't heard the information yet
• I (Informed): People who know and are actively sharing it
• R (Recovered/Silent): People who know but have stopped talking about it

Versions

Version	Feature
v0.1	Core SIR math model, CSV output
v0.2	Stochastic Monte Carlo simulation, variance analysis
v0.3	Visualizations: S-curves, comparisons, sensitivity plots
v0.4	Interactive CLI scenario explorer with presets
v0.5	Streamlit web dashboard (local)
v0.6	Static web app (dashboard.html) — JS port of the model, runs on GitHub Pages
v0.7	Landing page (index.html) — animated SIR canvas, S-curve, and preset gallery

Quick Start

pip install -r requirements.txt
cd src

# Run the deterministic model
python sir_model.py

# Run Monte Carlo simulations
python monte_carlo.py

# Generate visualizations
python visualize.py

# Interactive explorer
python explore.py

# Local web dashboard (Streamlit)
python -m streamlit run dashboard.py

The Python source remains the reference implementation. docs/index.html is the animated landing page; docs/dashboard.html is the self-contained interactive simulator — same math, runnable anywhere a browser can open a file. To preview locally:

bash scripts/preview.sh
# then visit http://localhost:8000          (landing)
# or     http://localhost:8000/dashboard.html  (simulator)

The preview script composes .brand/ and assets/ into docs/ (mirroring what the GitHub Actions deploy does) and serves the result.

Parameters

Parameter	Description	Default
seed_viewers	Initial informed population	10
population_size	Total population	1,000,000
conversation_rate	People told per day per informed person	3
spread_probability	Chance listener cares/shares	0.15
recovery_rate	Rate at which informed become silent	0.1
days_to_simulate	Simulation duration	60

The Viral Coefficient

The viral coefficient k = conversation_rate × spread_probability determines growth:

• k > recovery_rate: Information spreads exponentially (viral!)
• k < recovery_rate: Information fizzles out
• k ≈ recovery_rate: Slow, steady diffusion

---