---
id: eigenpy
version: "3.13.0"
license: BSD-2-Clause
license_treatment: permissive
maintenance: active
---

# eigenpy — Bindings between Numpy and Eigen using Boost.Python

License: permissive · Maintenance: active · Downloads: 1.3M/mo

## What it is and what it does

EigenPy is a C++ binding layer that bridges arrays and the Eigen linear algebra library via Boost.Python. It enables Python code to work directly with Eigen's high-performance matrix and tensor operations without copying data between Python and C++. The package exposes Eigen's standard decomposition routines (Cholesky, SVD, QR), geometry module, tensor support, and STL/Boost types (optional, std::pair, maps, variants), all while maintaining memory alignment for vectorization.

The package is designed for scientific and numerical computing workflows where performance matters. Researchers and engineers working with large matrices, tensor computations, or algorithms that benefit from Eigen's optimizations can use it to prototype and deploy code that seamlessly integrates Python's ease of use with C++'s computational speed. Installation is straightforward on Linux, macOS, and Windows via conda, apt, or Homebrew, with pre-built wheels for modern Python versions.

Use it for:

- Prototyping numerical algorithms in Python that leverage Eigen's optimized matrix decompositions without rewriting in C++.
- Building robotics or computer vision pipelines that need fast linear algebra operations with zero-copy overhead.
- Wrapping existing Eigen-based C++ libraries to expose them safely and efficiently to Python code.
- Performing tensor computations via Eigen::Tensor while maintaining full interoperability with Python arrays.
- Developing scientific simulations that require both Python's flexibility and C++'s performance for linear algebra kernels.

## Worth the install?

AI-flagged interpretation of the facts above — verify before relying.

EigenPy provides Python bindings between NumPy arrays and the Eigen C++ linear algebra library, enabling zero-copy memory sharing and access to Eigen's matrix decomposition and tensor operations from Python.

Yes, if you need high-performance linear algebra in Python with tight array integration and are on a supported platform (Python >=3.10, Linux/macOS/Windows). The zero-copy memory sharing and access to Eigen's decompositions make it valuable for numerical computing. Medium install friction (compiled dependencies) is offset by pre-built wheels and active maintenance. No known security issues and permissive licensing. Not necessary if you only use built-in operations or existing alternatives.

## Install

pip install eigenpy
uv add eigenpy
poetry add eigenpy

## Installing eigenpy

Before you install: Medium install friction due to compiled C++ dependencies (cmeel, cmeel-boost). Pre-built wheels cover modern Python versions (3.10–3.14) and common platforms (Linux x86_64/aarch64, macOS x86_64/arm64), making installation straightforward on supported systems. Active maintenance with recent releases.

License in practice: BSD-2-Clause is permissive and imposes minimal restrictions; you may use, modify, and distribute EigenPy in commercial or proprietary projects provided you include the license notice and disclaimer.

Quickstart:

pip install eigenpy
import eigenpy
# EigenPy enables zero-copy operations on arrays with Eigen's matrix decompositions

Requires Python >=3.10. Pre-built wheels available for Linux (glibc 2.28+), macOS 10.9+, and arm64 platforms; other systems may require building from source.

Verify before relying:

- Whether Eigen::Ref and Eigen::Tensor support are fully documented with usage examples.
- Performance characteristics and memory overhead of the zero-copy binding compared to alternatives.
- Compatibility with specific versions of cmeel and cmeel-boost beyond the stated Python requirement.

## Package facts

- License: BSD-2-Clause (permissive)
- Python support: supports_current
- Install friction: medium
- Maintenance: active
- Downloads: 1.3M/month (top 5,000 on PyPI)
- Known vulnerabilities: none known

## Tags

numpy eigen bindings, c++ linear algebra python, matrix decomposition python, eigen tensor python, numpy eigen interop, boost python bindings, cholesky svd qr python, linear-algebra, c++-bindings, numerical-computing

[View on SkillFed](https://skillfed.io/packages/eigenpy) · [View on PyPI](https://pypi.org/project/eigenpy/)