engineeringkit

Engineering concepts made computable → inspectable → visualizable → understandable.

EngineeringKit is an educational scientific-Python toolkit following the design philosophy of MathematicsKit: explicit intermediate results, NumPy/SciPy for mature numerics, and hand-written formulas/algorithms where seeing the mechanics is the lesson. SI units are used by public APIs.

Install

pip install engineeringkit

Beam

from engineeringkit.structural_engineering import SimplySupportedBeam

beam = SimplySupportedBeam(5.0, 200e9, 8e-6)
r = beam.solve_point_load(10_000)
print(r.summary())
r.plot()

PID

from scipy import signal
from engineeringkit.control_systems import PID

plant = signal.TransferFunction([1], [1, 2, 1])
r = PID(3, 1, 0.2).simulate(plant)
r.plot()

RLC

from engineeringkit.electrical_engineering import SeriesRLC

r = SeriesRLC(10, 0.1, 100e-6).step_response()
r.plot()

Thermodynamics

from engineeringkit.thermodynamics import OttoCycle

print(OttoCycle(10).summary())

Pipe flow

from engineeringkit.fluid_engineering import reynolds_number, colebrook_friction_factor

Re = reynolds_number(998, 2, 0.05, 1.002e-3)
f = colebrook_friction_factor(Re, 1e-4)

Release scope (v0.1)

All twelve domains include executable educational models. Version 0.1 adds physical input validation, independent conservation/analytical/distribution benchmarks, public scalar API units, and twelve detailed model guides. It also corrects RLC current response, FM phase integration and unsigned modulation, and improves numerical limiting cases. See docs/source/guides/ for model assumptions and VERIFICATION.md for the actual test/build evidence.

Documentation: https://cpoli.github.io/engineeringkit/. To cite engineeringkit, use CITATION.cff (GitHub's "Cite this repository").

Inspect a numerical method

import numpy as np
from engineeringkit.integrators import integrate_ode

result = integrate_ode(lambda t, y: -y, [1.0], np.linspace(0, 2, 101), "rk4")
print(result.summary())
fig, ax = result.plot(["normalized state"])

Verify locally

python -m pip install -e '.[dev]'
MPLBACKEND=Agg python -m pytest
python -m ruff check .
python -m ruff format --check .
python -m sphinx -W --keep-going -b html docs/source docs/_build/html
python -m build

Coverage requirement

Run the suite with production statement coverage:

MPLBACKEND=Agg python -m pytest --cov=engineeringkit --cov-report=term-missing

The report enforces 100% statement coverage. Test modules are excluded from the coverage denominator; all production modules, private validation helpers, constants, units, plotting and integration code are included. No production statements are marked with coverage-ignore pragmas. This measures statement execution, not 100% branch coverage or proof of scientific correctness.

Documentation navigation

Documentation now follows MathematicsKit's domain-first architecture: Subpackages, History, Examples and API. Each domain hub links all three learning paths; grouped cards and cross-links are generated from one registry in docs/source/conf.py. Executed galleries come from examples/<domain>/, and detailed SI/assumption guides remain linked from the galleries and domain hubs. See docs/README.md.

The MathematicsKit-style Sphinx navigation connects twelve domain hubs to History, Examples and API pages. History pages explain key equations and limits; 24 plotted gallery scripts include twelve explicit benchmark demonstrations.