Skip to content

yunruse/Noether

Repository files navigation

Noether

523 units, 60 prefixes

PyPI

Noether is a unit-enriched Python package, akin to Wolfram Alpha or gnu units. It has a large (and expanding) catalogue of up-to-date units and constants, allowing code to be written directly in the units they are concerned with while also ensuring e.g. you don't mistakenly add an energy to a length.

Just grab Python 3.10 or later and pip install noether to run!

Development is ongoing, especially in expanding the unit catalogue and improving unit display mechanisms.

Usage

Noether can be used as a Python package or as a CLI:

$ alias noe='python -im noether'
$ noe marathon
marathon  # length, 42195 m, Race length based on Greek legend, set by convention from 1908 Summer Olympics
$ noe 23degC @ degF
73.4 °F  # temperature
$ noe 'horsepower @ dB(kW)'
-1.33418061 dB(kW)  # power, 0.73549875 kW

The CLI allows a few niceties such as slightly terser syntax, but otherwise behaves identically to Python:

$ noe 5cm @ in --value
1.9685039370078738
$ noe
>>> 5*cm @ inch
1.9685039370078738 in  # length

In addition to @ for display, you can more permanently set display units:

>>> display(inch)
>>> 5 * cm
>>> mile
mile  # length, 63360 in

Units propagate uncertainty automatically under most operations:

>>> m(5, 0.1)**3
125 ± 7.5 m**3  # volume

You can define your own units and dimensions:

>>> foo = Unit(3e11 * furlong / fortnight, 'foo')
>>> c @ foo
6.008724999284181 foo  # speed

>>> health = Dimension.new('health', 'H')
>>> apple = Unit(health, 'apple', 'a')
>>> apple / day
apple / day  # health / time, 1.1574074074074073e-05 a / s

Various conf settings allow for customisation to behaviour:

>>> conf.info_spectrum = True
>>> nm(400)
4e-07 m  # length, visible, purple

Use conf.save() to save to (by default) ~/.config/noether.toml.

Roadmap

The approximate roadmap for future releases is:

  • Support for GNU units, including exporting a file to work directly in units
  • A custom format (likely YAML) for defining the catalogue far more quickly
  • Support for numpy, matplotlib

For more, see...

Other unit packages include::

  • gnu units, a command-line tool you likely already have
  • pint, a Python package with numpy support
  • units, a simple Python package for defining your own units
  • unyt, a Python package with numpy support
  • Wolfram Alpha, a comprehensive online intelligence engine

📚 Did you know?

>>> lunation / (year % lunation)
2.7153489666011486

A lunation (about 29 days) separates one full moon from another. Every so often a thirteenth full moon occurs in a year - "a blue moon". "Once in a blue moon" is actually only every 2.71 years or so - not as rare as you'd think. Don't tell Sinatra!