rt-core-amador

Lightweight mathematical tools implemented in pure Python. The package has no runtime dependencies and requires Python 3.10 or newer.

Scope: this is a new, self-contained implementation inspired by the requested project idea. "Amador Transform" here means the explicitly documented phase-modulated discrete Fourier transform below; it does not claim to reproduce an earlier implementation or an independently established mathematical definition.

Install

python -m pip install rt-core-amador

Quick start

from rt_core_amador import AmadorTensor, AmadorTransform

tensor = AmadorTensor([[1, 2], [3, 4]])
print((tensor * 2).to_list())

transform = AmadorTransform(phase=0.2)
coefficients = transform.forward([1, 0, -1, 0])
restored = transform.inverse(coefficients)

API

AmadorTensor

A small dense tensor backed by a flat tuple of Python floats. Construction accepts a scalar or rectangular nested list/tuple. Supports element-wise addition, subtraction, multiplication and division with another tensor of the same shape or a scalar, plus dot() for vectors and matmul() for matrices. This intentionally does not implement NumPy-style broadcasting.

AmadorTransform

forward(values) computes a DFT normalized by 1/N and applies a phase rotation exp(i * phase * k) to frequency bin k. inverse(coefficients) removes that rotation and computes the corresponding inverse. Both operations accept a non-empty one-dimensional sequence; output values are Python complex numbers.

NavierStokesOptimizer

A simple explicit 2D incompressible Navier–Stokes projection step on a rectangular periodic grid. Call step(u, v, dt) with same-sized 2D velocity grids. It returns (u_next, v_next, pressure). This is a basic educational finite-difference solver, not a production CFD package: explicit time steps must satisfy stability constraints, and boundary conditions are periodic.

KahlerManifold

Numerically estimates the Hermitian metric g[i, j] = d²φ / dz[i] d(conjugate(z[j])) from a real-valued potential φ(z). The default potential sum(abs(z_i)**2) yields the identity metric. metric_at(point) returns a matrix of complex values and norm_squared(point, tangent) evaluates the induced squared norm. Finite differences are approximate and the supplied potential should be smooth and strictly plurisubharmonic near the point.

Development

Run the tests with:

python -m unittest discover -s tests -v

Build the source and wheel distributions with:

python -m pip install --upgrade build
python -m build

PyPI publishing

The GitHub Actions workflow at .github/workflows/publish.yml builds on a GitHub Release and publishes with PyPI Trusted Publishing (OIDC). Configure a trusted publisher on PyPI for this GitHub repository, workflow file publish.yml, and the pypi environment. Create a GitHub environment named pypi before publishing. No PyPI API token is stored in GitHub secrets.