raptors 0.0.2

Rust-powered, NumPy-compatible array library scaffold.

Raptors

Rust-powered, NumPy-compatible Python array library scaffolding.

Current release: 0.0.2

Getting Started

This repository contains the initial skeleton for raptors, a Rust-backed array library exposed to Python. The project is being developed by Odos Matthews ([e-mail], GitHub: eddiethedean).

Layout

• rust/: Rust crate exposing core functionality via pyo3.
• python/: Python package wrapper publishing the raptors module.
• tests/: pytest-based test suite comparing behavior to NumPy.
• benches/: Benchmark harnesses for performance tracking.
• ci/: Continuous-integration workflows and scripts.
• docs/: Documentation sources (overview, guides, references).

Consult docs/overview.md for an expanded project overview and links to the full plan.

Installation (local development)

# Build and install the Rust extension into your active virtualenv.
maturin develop

# Or build a release wheel.
maturin build --release

Quickstart

import raptors

# Construct arrays from Python iterables or helper constructors.
a = raptors.array([1.0, 2.0, 3.0])
b = raptors.ones(3)

# Run elementwise math and reductions.
c = a.add(b).scale(2.0)
total = c.sum()
average = c.mean()

# Interoperate with NumPy (requires numpy).
import numpy as np

numpy_view = raptors.to_numpy(c)
assert np.allclose(numpy_view, [4.0, 6.0, 8.0])

Continuous Integration

The .github/workflows/build-wheels.yml workflow builds and tests wheels across Ubuntu, macOS, and Windows for Python 3.9–3.12 using maturin. Successful runs upload ready-to-use wheel artifacts and exercise the Python bindings via pytest.