conservation-conformance

Cross-language conformance test suite for the Conservation Spectral SDK — verifies that Python, C, JavaScript, and Rust implementations produce identical spectral results.

What This Gives You

• Reference test vectors — Precomputed adjacency matrix with expected eigenvalues, eigenvectors, conservation ratios, spectral gap, and Cheeger constant
• Four language runners — Automated comparison against Python, C (gcc), JavaScript (Node), and Rust implementations
• Configurable tolerances — Eigenvalue (1e-6), eigenvector (1e-4), conservation ratio (1e-4), and scalar (1e-6) tolerances
• Pass/fail summary — Clear ✅/❌ output per implementation with per-value detail

Quick Start

# Run conformance tests against all implementations
python run_conformance.py

Output:

╔══════════════════════════════════════════════════════════════╗
║  Conservation Spectral SDK — Cross-Language Conformance     ║
╚══════════════════════════════════════════════════════════════╝

📊 CONFORMANCE SUMMARY
  Python       ✅ PASS
  C            ✅ PASS
  JS           ✅ PASS
  Rust         ✅ PASS

🎉 ALL IMPLEMENTATIONS CONFORM

How It Works

1. Loads expected_results.json — reference vectors computed by compute_reference.py
2. For each language, builds the same 5-vertex chord-progression graph
3. Runs Laplacian construction → eigendecomposition → conservation analysis
4. Compares results against reference within tolerance
5. Reports per-value agreement and overall pass/fail

Files

File	Purpose
expected_results.json	Reference eigenvalues, eigenvectors, ratios, scalars
test_vectors.json	Input adjacency matrix and attributes
compute_reference.py	Generate reference results from the Python SDK
run_conformance.py	Main conformance runner (Python, C, JS, Rust)

How It Fits

This is the conformance layer for the conservation spectral ecosystem:

• Rust: conservation-spectral
• Python: conservation-spectral-python
• TypeScript: conservation-spectral-js
• Ada: conservation-spectral-ada

All implementations must pass this suite before release.

Requirements

• Python ≥ 3.10 with numpy, scipy, and the conservation-spectral package installed
• gcc for C tests
• node for JavaScript tests
• cargo for Rust tests

License

MIT

Documentation

📚 OpenConstruct Docs