panchi

panchi is a Python-native linear algebra library designed for learning, experimentation, and visual intuition.

The goal is not performance. The goal is clarity.

Why panchi?

Most linear algebra libraries optimize for speed and abstraction. panchi optimizes for understanding.

panchi is built for students who want to see the math happen, educators who need transparent implementations, and anyone who has ever wondered what linear algebra is actually about.

Think of it as a lab, not a production engine.

Philosophy

1. Explicit over implicit – Algorithms are implemented directly, not delegated to opaque backends
2. Readable over clever – Code prioritizes clarity and educational value over terse optimizations and pythonisms
3. Mathematical over computational – Objects behave like mathematical entities with proper operator overloading
4. Visual by default – Visualization is a first-class feature, not an afterthought
5. Informative errors – Error messages guide learning by explaining what went wrong and why

Installation

pip install panchi

Requires Python 3.10+. For optional Manim-powered visualizations:

pip install panchi[manim]

A Taste

import panchi as pan
from panchi.algorithms import rref, solve

# Vectors and matrices behave like their mathematical counterparts
A = pan.Matrix([[1, 2, 3], [2, 5, 7], [0, 1, 2]])
b = pan.Vector([1, 0, 0])

# Solve Ax = b — see the status, not just the answer
result = solve(A, b)
print(result.status)    # 'unique'
print(result.solution)  # the solution vector x

# Row reduction shows every step it takes
reduction = rref(A)
print(reduction)        # full step-by-step walkthrough
print(reduction.rank)   # 3

See what a matrix does to the plane:

from panchi.visualizations import Animator2D

animator = Animator2D()
animator.animate_transform(pan.Matrix([[0, -1], [1, 0]]))  # 90° rotation

Documentation

Full documentation, user guides, and the API reference are available at https://gustavo-galvao-e-silva.github.io/panchi/

Contributing

panchi welcomes contributions that align with its educational mission. See CONTRIBUTING.md for guidelines. Thanks to all of our contributors, whose names can be found in CONTRIBUTORS.md.

License

MIT License – see LICENSE for details.

Acknowledgments

panchi is inspired by Gilbert Strang's Introduction to Linear Algebra and 3Blue1Brown's Essence of Linear Algebra — resources that make the subject visible, not just computable.