Skip to content

HugoDelhaye/select_scene

Repository files navigation

Scene Selection Analysis

Analysis of Super Mario Bros gameplay data to identify scenes with interesting learning patterns across human players, imitation learning models, and PPO agents.

Quick Start

Setup

This project uses airoh for task automation and reproducible workflows.

# Install dependencies (includes airoh and invoke)
pip install -r requirements.txt

# Or use airoh task for setup
invoke setup

# (Optional) Install Jupyter kernel
pip install ipykernel

To see all available tasks:

invoke --list

See .invoke-cheatsheet.md for a quick reference of all tasks.

Run Interactive Dashboard

Option 1: Web Dashboard (Recommended - Fast & Modern)

# Using airoh task
invoke dashboard

# Or directly
python run_dashboard.py
# Opens at http://localhost:8050

Option 2: Export to HTML (For Sharing)

# Export all scenes using airoh task
invoke export-html

# Export specific level or scene
invoke export-html --level=w1l1
invoke export-html --level=w1l1 --scene=0

# Or use the Python script directly
python export_dashboard.py --output-dir html_export

# Then open html_export/index.html in your browser
# Share the entire html_export/ folder with collaborators

Option 3: Jupyter Notebook (Legacy)

jupyter notebook notebook/figure_exploration.ipynb

Data Pipeline

Generate metrics from raw data:

# Using airoh task (recommended)
invoke process-data

# Or directly
python code/make_df_metrics.py

Available Tasks

See all available airoh tasks:

invoke --list

Key tasks:

  • invoke setup - Install Python dependencies
  • invoke process-data - Generate metrics from raw data
  • invoke dashboard - Launch interactive web dashboard
  • invoke export-html - Export scenes to standalone HTML
  • invoke stats - Display dataset statistics
  • invoke list-scenes - List all available scenes
  • invoke clean - Remove generated outputs

Project Structure

  • code/ - Core processing and visualization modules
  • notebook/ - Jupyter notebooks for analysis
  • sourcedata/ - Data files (parquet format)
  • run_dashboard.py - Launch web dashboard
  • export_dashboard.py - Export to HTML files

See CLAUDE.md for detailed architecture documentation.

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Contributors 3

  •  
  •  
  •