Skip to content

release-content.md

Latest commit

 

History

History
executable file
·
50 lines (34 loc) · 1.53 KB

release-content.md

File metadata and controls

executable file
·
50 lines (34 loc) · 1.53 KB
Raw

TITLE: v0.1.0 — First Release: Exact Geometry for Python

BODY: Python bindings for Constraint Theory — snap vectors to exact Pythagorean coordinates with zero floating-point drift, powered by Rust.

Quick Example

from constraint_theory import PythagoreanManifold

# Create manifold with ~1000 valid states
manifold = PythagoreanManifold(density=200)

# Snap any direction to nearest exact Pythagorean triple
x, y, noise = manifold.snap(0.577, 0.816)
# Result: (0.6, 0.8) — exact 3/5, 4/5 ratio

Key Features

  • Zero Drift — Vectors snap to exact Pythagorean triples (3/5, 4/5, etc.), eliminating floating-point accumulation errors
  • Blazing Fast — Rust-powered KD-tree lookups in ~100ns via PyO3 bindings
  • Batch Processing — SIMD-accelerated batch operations for thousands of vectors
  • Cross-Platform Determinism — Same results on any machine, any architecture
  • Simple API — Just snap() and snap_batch() — no configuration needed

Installation

pip install constraint-theory

Or with NumPy support:

pip install constraint-theory[numpy]

Core Crate

This package provides Python bindings for the Rust implementation: constraint-theory-core

What's Next

  • Higher-dimensional generalizations (3D Pythagorean quadruples)
  • GPU implementations (CUDA, WebGPU)
  • More language bindings

See the README for full documentation and examples.