AxiomPy: The Python Mathematics & Computation Engine

Author: RK Riad Khan (rkriad585)

Overview

AxiomPy is a Python mathematics engine built from first principles. It provides a unified API for linear algebra, graph analysis, number theory, automatic differentiation, optimization, signal processing, statistics, caching, constants, CLI tools, and number-theoretic novelties — all wired through a single Axiom singleton.

Features

• First-principle implementations — algorithms written from scratch.
• Linear algebra — Vector, Matrix, Tensor, SparseMatrix, decompositions.
• Graph analysis — PageRank, MST, max flow, bipartiteness, topological sort.
• Number theory — CRT, Miller-Rabin, discrete log, totient, prime sieves.
• Automatic differentiation — reverse-mode AD with sin, exp, log, tanh.
• Calculus — numerical derivatives, integration, ODE solvers (RK4, adaptive).
• Optimization — gradient descent, Newton, Nelder-Mead, L-BFGS, simulated annealing.
• Signal processing — FFT, convolution, FIR filters, spectrograms.
• Statistics — descriptive stats, hypothesis tests, distributions, Bayesian.
• Electromagnetism — Coulomb fields, potentials, magnetic fields.
• Constants — 35 built-in math/physics/astronomy constants with search.
• Cache — LRU cache with TTL, hit/miss stats, async memoize, persistence.
• CLI — axiompy eval, factors, convert, constants, stats, help, shell.
• Magic functions — roman numerals, number words, Kaprekar, Goldbach, look-and-say, Ulam spiral, Smith numbers, emirps, and more.
• ASCII visualization — line plots, histograms, bar charts, field plots.
• Clean API — from axiompy import Axiom gives access to everything.

Installation

pip install axiom-math

Requirements: Python ≥ 3.9, numpy.

Quick start

from axiompy import Axiom

# Linear algebra
M = Axiom.Matrix([[1, 2], [3, 4]])
v = Axiom.Vector([5, 6])
print(M @ v)           # Vector([17.0, 39.0])
print(M.determinant)   # -2.0

# PageRank
g = Axiom.Graph()
g.add_edge("A", "B"); g.add_edge("B", "C"); g.add_edge("C", "A")
print(Axiom.graph_analysis.pagerank(g))

# Automatic differentiation
x = Axiom.autodiff.Variable(3.0)
y = x ** 2 + 2 * x + 1
y.backward()
print(x.grad)          # 8.0 (dy/dx)

# Constants & CLI
Axiom.constants.PI           # 3.14159...
Axiom.constants.find("mass") # search by name

CLI

axiompy eval "2 * pi * 5^2"    # evaluate expression
axiompy factors 84              # prime factorization
axiompy constants mass          # search constants
axiompy convert 42              # binary, hex, roman, words
axiompy stats                   # cache & system info
axiompy help magic              # topic help
axiompy shell                   # interactive REPL

Documentation

Guides are in the docs/ directory:

Examples

Ready-to-run scripts in examples/:

uv run python examples/linear_algebra.py
uv run python examples/pagerank.py
uv run python examples/autodiff.py
uv run python examples/number_theory.py
uv run python examples/electric_field.py
uv run python examples/covariance.py
uv run python examples/ascii_plot.py

Development

git clone https://github.com/rkriad585/AxiomPy.git
cd AxiomPy
uv sync
uv build

Contributing

See docs/contributing.md. Bug reports and pull requests welcome on the GitHub repository.

License

MIT — see LICENSE.