Skip to main content

Manipulate physical quantities in Python

Project description

physipy

Binder PyPI version

This python package allows you to manipulate physical quantities, basically considering in the association of a value (scalar, numpy.ndarray and more) and a physical unit (like meter or joule).

>>> from physipy.quickstart import nm, hp, c, J
>>> E_ph = hp * c / (500 * nm)
>>> print(E_ph)
3.9728916483435158e-19 kg*m**2/s**2
>>> E_ph.favunit = J
>>> print(E_ph)
3.9728916483435158e-19 J

For a quickstart, check the quickstart notebook on the homepage Get a live session at Binder

Installation

pip install physipy

Goals

  • Few LOC
  • Simple architecture, with only 2 classes (namely Dimension and Quantity)
  • High numpy compatibility
  • Human-readable syntax (fast syntax !)

Use case

  • Define scalar and arrays of physical quantities
  • Compute operation between them : add, sub, mul, div, pow, and so on
  • Display physical quantities in various “units”

Implementation approach

The implementation is pretty simple :

  • a Dimension object represents a physical dimension. For now, these dimension are based on the SI unit. It is basically a dictionary where the keys represent the base dimensions, and the values are the exponent these dimensions.
  • a Quantity object is simply the association of a value, scalar or array (or more!), and a Dimension object. Note that this Quantity classe does not sub-class numpy.ndarray (although Quantity objects are compatible with numpy's ufuncs). Most of the work is done by this class.
  • By default, a Quantity is displayed in term of SI untis. To express a Quantity in another unit, just set the "favunit", which stands for "favourite unit" of the Quantity : my_toe_length.favunit = mm.
  • Plenty of common units (ex : Watt) and constants (ex : speed of light) are packed in. Your physical quantities (my_toe_length), units (kg), and constants (kB) are all Quantity objects.

Numpy's support

Numpy is almost fully and transparently handled in physipy : basic operations, indexing, numpy functions and universal functions are handled. There are more than 150 functions implemented ! Some limitation still exist but can be can be circumvented. See the dedicated notebook : https://github.com/mocquin/physipy/blob/master/docs/notebooks/Numpy.ipynb.

Matplotlib's units support

Matplotlib allows defining a physical units interface, which can be turned on using just setup_matplotlib, all plot involving a physical quantity will automatically label the axis accordingly :

import numpy as np
import matplotlib.pyplot as plt
from physipy import s, m, units, setup_matplotlib
setup_matplotlib() # make matplotlib physipy's units aware
mm = units["mm"]   # get millimiter
ms = units["ms"]   # get millisecond

y = np.linspace(0, 30) * mm
x = np.linspace(0, 5) * s
y.favunit = mm # no need to call ax.yaxis.set_units(mm)
x.favunit = ms # no need to call ax.xaxis.set_units(ms)

fig, ax = plt.subplots()
ax.plot(x, y)

Checkout the dedicated notebook on matplotlib support.

Widgets

Some ipywidgets are provided to make your physical researches and results more interactive : Checkout the dedicated notebook on ipywidgets.

Known issues

See the dedicated notebook.

Benchmark

Benchmark results using asv are available at https://mocquin.github.io/physipy/ :

./docs/notebooks/ressources/asv_screenshot.png

See also the corresponding notebook at : https://github.com/mocquin/physipy/blob/master/docs/notebooks/Benchmarking%20with%20AirSpeedVelocity.ipynb.

About angles and units

See : https://www.bipm.org/en/CGPM/db/20/8/. Astropy's base units : https://docs.astropy.org/en/stable/units/standard_units.html#enabling-other-units

Alternative packages

A quick performance benchmark show that physipy is just as fast (or faster) than other well-known physical packages, both when computing scalars (int or float) and numpy arrays :

For a more in-depth comparison, checkout this repository (not maintenained be it should!) : https://github.com/mocquin/quantities-comparison :

There are plenty of python packages that handle physical quantities computation. Some of them are full packages while some are just plain python module. Here is a list of those I could find (approximately sorted by guessed-popularity) :

If you know another package that is not in this list yet, feel free to contribute ! Also, if you are interested in the subject of physical quantities packages in python, check this quantities-comparison repo and this talk. Also check this comparison table and this takl.

Some C/C++ alternatives :

License

This project is licensed under the MIT License - see the LICENSE.md file for details

Acknowledgment

Thumbs up to phicem and his pysics package, on which this package was highly inspired. Check it out !

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

physipy-0.2.3.tar.gz (76.5 kB view hashes)

Uploaded source

Built Distribution

physipy-0.2.3-py3-none-any.whl (52.9 kB view hashes)

Uploaded py3

Supported by

AWS AWS Cloud computing Datadog Datadog Monitoring Facebook / Instagram Facebook / Instagram PSF Sponsor Fastly Fastly CDN Google Google Object Storage and Download Analytics Huawei Huawei PSF Sponsor Microsoft Microsoft PSF Sponsor NVIDIA NVIDIA PSF Sponsor Pingdom Pingdom Monitoring Salesforce Salesforce PSF Sponsor Sentry Sentry Error logging StatusPage StatusPage Status page