Code Development Guidelines

This document describes the coding requirements and guidelines to be followed during the development of PlasmaPy and affiliated packages.

Code written for PlasmaPy must be compatible with Python 3.7 and later.

Coding Style

TL;DR: use pre-commit

PlasmaPy has a configuration for the pre-commit framework that takes care of style mostly automatically. Install it with pip install pre-commit, then use pre-commit install within the repository.

This will cause pre-commit to download the right versions of linters we use, then run an automated style checking suite on every commit. Do note that this works better with a git add, then git commit workflow than a git commit -a workflow — that way, you can check via git diff what the automated changes actually did.

Note that the “Style linters / pre-commit (pull_request)” part of our Continuous Integration system can and will (metaphorically) shout at you if it finds you didn’t apply the linters. Also note that the linters’ output may vary with version, so, rather than apply black and isort manually, let pre-commit do the version management for you instead!

Our pre-commit suite can be found in .pre-commit-config.yaml. It includes

  • black to automatically format code and ensure a consistent code style throughout the package

  • isort to automatically sort imports.

  • nbqa to automatically apply the above to example notebooks as well.

  • a few tools for requirements.txt, yml files and the like.

PlasmaPy Code Style Guide, codified

  • PlasmaPy follows the PEP8 Style Guide for Python Code. This style choice helps ensure that the code will be consistent and readable.

    • Line lengths should be chosen to maximize the readability and elegance of the code. The maximum line length for Python code in PlasmaPy is 88 characters.

    • Docstrings and comments should generally be limited to about 72 characters.

  • During code development, use black to automatically format code and ensure a consistent code style throughout the package and isort to automatically sort imports.

  • Follow the existing coding style within a subpackage. This includes, for example, variable naming conventions.

  • Use standard abbreviations for imported packages when possible, such as import numpy as np, import matplotlib as mpl, import matplotlib.pyplot as plt, and import astropy.units as u.

  • __init__.py files for modules should not contain any significant implementation code, but it can contain a docstring describing the module and code related to importing the module. Any substantial functionality should be put into a separate file.

  • Use absolute imports, such as from plasmapy.particles import Particle, rather than relative imports such as from ..particles import Particle.

  • Use Optional[type] for type hinted keyword arguments with a default value of None.

  • There should be at least one pun per 1284 lines of code.

  • Avoid using lambda to define functions, as this notation may be unfamiliar to newcomers to Python.

Branches, commits, and pull requests

Before making any changes, it is prudent to update your local repository with the most recent changes from the development repository:

git fetch upstream

Changes to PlasmaPy should be made using branches. It is usually best to avoid making changes on your main branch so that it can be kept consistent with the upstream repository. Instead we can create a new branch for the specific feature that you would like to work on:

git branch *your-new-feature*

Descriptive branch names such as grad-shafranov or adding-eigenfunction-poetry are helpful, while vague names like edits are considered harmful. After creating your branch locally, let your fork of PlasmaPy know about it by running:

git push --set-upstream origin *your-new-feature*

It is also useful to configure git so that only the branch you are working on gets pushed to GitHub:

git config --global push.default simple

Once you have set up your fork and created a branch, you are ready to make edits to PlasmaPy. Switch to your new branch by running:

git checkout *your-new-feature*

Go ahead and modify files with your favorite text editor. Be sure to include tests and documentation with any new functionality. We recommend reading about best practices for scientific computing. PlasmaPy uses the PEP 8 style guide for Python code and the numpydoc format for docstrings to maintain consistency and readability. New contributors should not worry too much about precisely matching these styles when first submitting a pull request, GitHub Actions will check pull requests for PEP 8 compatibility, and further changes to the style can be suggested during code review.

You may periodically commit changes to your branch by running

git add filename.py
git commit -m "*brief description of changes*"

Committed changes may be pushed to the corresponding branch on your GitHub fork of PlasmaPy using

git push origin *your-new-feature*

or, more simply,

git push

Once you have completed your changes and pushed them to the branch on GitHub, you are ready to make a pull request. Go to your fork of PlasmaPy in GitHub. Select “Compare and pull request”. Add a descriptive title and some details about your changes. Then select “Create pull request”. Other contributors will then have a chance to review the code and offer contructive suggestions. You can continue to edit the pull request by changing the corresponding branch on your PlasmaPy fork on GitHub. After a pull request is merged into the code, you may delete the branch you created for that pull request.

Commit Messages

Good commit messages communicate context and intention to other developers and to our future selves. They provide insight into why we chose a particular implementation, and help us avoid past mistakes.

Suggestions on how to write a git commit message:

  • Separate subject from body with a blank line

  • Limit the subject line to 50 characters

  • Capitalize the subject line

  • Do not end the subject line with a period

  • Use the imperative mood in the subject line

  • Wrap the body at 72 characters

  • Use the body to explain what and why vs. how

Documentation

  • All public classes, methods, and functions should have docstrings using the numpydoc format.

  • Docstrings may be checked locally using pydocstyle.

  • These docstrings should include usage examples.

Warnings and Exceptions

  • Debugging can be intensely frustrating when problems arise and the associated error messages do not provide useful information on the source of the problem. Warnings and error messages must be helpful enough for new users to quickly understand any problems that arise.

  • “Errors should never pass silently.” Users should be notified when problems arise by either issuing a warning or raising an exception.

  • The exceptions raised by a method should be described in the method’s docstring. Documenting exceptions makes it easier for future developers to plan exception handling.

Units

  • Code within PlasmaPy must use SI units to minimize the chance of ambiguity, and for consistency with the recognized international standard. Physical formulae and expressions should be in base SI units.

    • Functions should not accept floats when an Astropy Quantity is expected. In particular, functions should not accept floats and make the assumption that the value will be in SI units.

    • A common convention among plasma physicists is to use electron-volts (eV) as a unit of temperature. Strictly speaking, this unit corresponds not to temperature but is rather a measure of the thermal energy per particle. Code within PlasmaPy must use the kelvin (K) as the unit of temperature to avoid unnecessary ambiguity.

  • PlasmaPy uses the astropy.units package to give physical units to values.

    • All units packages available in Python presently have some limitations, including incompatibility with some NumPy and SciPy functions. These limitations are due to issues within NumPy itself. Many of these limitations are being resolved, but require upstream fixes.

  • Dimensionless units may be used when appropriate, such as for certain numerical simulations. The conventions and normalizations should be clearly described in docstrings.

Equations and Physical Formulae

  • If a quantity has several names, then the function name should be the one that provides the most physical insight into what the quantity represents. For example, gyrofrequency indicates gyration, whereas Larmor_frequency indicates that this frequency is somehow related to someone named Larmor. Similarly, using omega_ce as a function name will make the code less readable to people who are unfamiliar with this particular notation.

  • Physical formulae should be inputted without first evaluating all of the physical constants. For example, the following line of code obscures information about the physics being represented:

>>> omega_ce = 1.76e7*(B/u.G)*u.rad/u.s   

In contrast, the following line of code shows the exact formula which makes the code much more readable.

>>> omega_ce = (e * B) / (m_e * c)       

The origins of numerical coefficients in formulae should be documented.

  • Docstrings should describe the physics associated with these quantities in ways that are understandable to students who are taking their first course in plasma physics while still being useful to experienced plasma physicists.

  • SI units that were named after a person should not be capitalized except at the beginning of a sentence.

  • Some plasma parameters depend on more than one quantity with the same units. In the following line, it is difficult to discern which is the electron temperature and which is the ion temperature.

    >>> ion_sound_speed(1e6*u.K, 2e6*u.K)  
    

    Remembering that “explicit is better than implicit”, it is more readable and less prone to errors to write:

    >>> ion_sound_speed(T_i=1e6*u.K, T_e=2e6*u.K)    
    
  • SI units that were named after a person should be lower case except at the beginning of a sentence, even if their symbol is capitalized. For example, kelvin is a unit while Kelvin was a scientist.

Angular Frequencies

Unit conversions involving angles must be treated with care. Angles are dimensionless but do have units. Angular velocity is often given in units of radians per second, though dimensionally this is equivalent to inverse seconds. Astropy will treat radians dimensionlessly when using the dimensionless_angles equivalency, but dimensionless_angles does not account for the multiplicative factor of 2*pi that is used when converting between frequency (1 / s) and angular frequency (rad / s). An explicit way to do this conversion is to set up an equivalency between cycles/s and Hz:

>>> from astropy import units as u
>>> f_ce = omega_ce.to(u.Hz, equivalencies=[(u.cy/u.s, u.Hz)])   

However, dimensionless_angles does work when dividing a velocity by an angular frequency to get a length scale:

>>> d_i = (c/omega_pi).to(u.m, equivalencies=u.dimensionless_angles())    

Examples

Examples in PlasmaPy are written as Jupyter notebooks, taking advantage of their mature ecosystems. They are located in docs/notebooks. nbsphinx takes care of executing them at documentation build time and including them in the documentation.

Please note that it is necessary to store notebooks with their outputs stripped (use the “Edit -> Clear all” option in JupyterLab and the “Cell -> All Output -> Clear” option in the “classic” Jupyter Notebook). This accomplishes two goals:

  1. helps with versioning the notebooks, as binary image data is not stored in the notebook

  2. signals nbsphinx that it should execute the notebook.

Note

In the future, verifying and running this step may be automated via a GitHub bot. Currently, reviewers should ensure that submitted notebooks have outputs stripped.

If you have an example notebook that includes packages unavailable in the documentation building environment (e.g., bokeh) or runs some heavy computation that should not be executed on every commit, keep the outputs in the notebook but store it in the repository with a preexecuted_ prefix, e.g. preexecuted_full_3d_mhd_chaotic_turbulence_simulation.ipynb.

Benchmarks

PlasmaPy has a set of asv benchmarks that monitor performance of its functionalities. This is meant to protect the package from performance regressions. The benchmarks can be viewed at benchmarks. They’re generated from results located in benchmarks-repo. Detailed instructions on writing such benchmarks can be found at asv-docs. Up-to-date instructions on running the benchmark suite will be located in the README file of benchmarks-repo.