Contribute

General remarks

PyScanFCS has no funding and no developer community. My personal objective is to keep PyScanFCS operational on Linux and Windows which is currently limited by the free time I have available.

An active community is very important for an open source project such as PyScanFCS. You can help this community grow (and thus help improve PyScanFCS) in numerous ways:

  1. Tell your colleagues and peers about PyScanFCS. One of them might be able to contribute to the project.

  2. If you need a new feature in PyScanFCS, publicly announce a bounty for its implementation.

  3. If your research heavily relies on FCS, please consider diverting some of your resources to the development of PyScanFCS.

  4. You don’t have to be a Python programmer to contribute. If you are familiar with reStrucuredText or LaTeX, you might be able to help out with the online documentation.

  5. Please cite: Müller et al. Methods in Molecular Biology, Humana Press, Springer, New York. 2014, 635-651. doi:10.1007/978-1-62703-649-8_29

If you are planning to contribute to PyScanFCS, please contact me via the PyScanFCS issue page on GitHub such that we may coordinate a pull request.

For documentation writers

To build this documentation, fork PyScanFCS, navigate to the docs (not doc) directory and run.

  • pip install -r requirements.txt followed by

  • sphinx-build . _build.

This will create the html documentation on your computer. Syntax warnings and errors will be displayed during the build (there should be none). After making your changes to your forked branch, create a pull request on GitHub.

If you only found a typo or wish to make text-only changes, you can also use the GitHub interface to edit the files (without testing the build step on your computer).

For developers

Running from source

It is recommended to work with virtual environments.

Windows

The easiest way to run PyScanFCS from source on Windows is Anaconda.

  • conda install matplotlib numpy pip scipy wxpython

  • pip install cython wheel simplejson sympy lmfit

  • pip install -e .  # in the root directory of the repository

Ubuntu 17.10

PyScanFCS requires wxPython >= 4.0.1 which is not available as a binary wheel on PyPI and thus must be built from the .tar.gz. Install all dependencies (https://github.com/wxWidgets/Phoenix/blob/master/README.rst):

  • pip install cython matplotlib lmfit numpy scipy sympy

  • sudo apt-get install -qq libgtk2.0 libgtk2.0-dev libwebkitgtk-dev dpkg-dev build-essential python3.6-dev libjpeg-dev libtiff-dev libsdl1.2-dev libnotify-dev freeglut3 freeglut3-dev libsm-dev libgtk-3-dev libwebkit2gtk-4.0-dev libxtst-dev libgstreamer-plugins-base1.0-dev

  • pip install wxPython  # this will take some time

  • pip install -e .  # in the root directory of the repository

Testing

PyScanFCS is tested using pytest. If you have the time, please write test methods for your code and put them in the tests directory. You may run all tests by issuing:

python setup.py test

Pull request guidelines

Please fork PyScanFCS and create a pull request (PR) introducing your changes.

  • A new PR should always be made into the master branch.

  • If a PR introduces a new functionality or fixes a bug, it should provide a test case, i.e. a new file or function in the tests directory (see here for examples). Note that currently there is no recipe for testing the graphical user interface code.

  • New code should follow the style guide for Python. Please use flake8 to check the files you changed or created.

  • New code should be documented well.

  • Make sure to update the changelog.

Windows test binaries

After each commit to the PyScanFCS repository, a binary installer is created by Appveyor. Click on a build and navigate to ARTIFACTS (upper right corner right under the running time of the build). From there you can download the Windows installer for the commit.