Contributing to PyNN¶
Discussions about PyNN take place in the NeuralEnsemble Google Group.
Setting up a development environment¶
We strongly suggest you work in a virtual environment, e.g. using virtualenv or Anaconda.
In addition to the requirements listed in Installation, you will need to install:
to run tests, and:
to build the documentation.
$ git clone https://github.com/<username>/PyNN.git pyNN_dev
To work on the development version:
$ git checkout master
To keep your PyNN repository up-to-date with respect to the official repository, add it as a remote:
$ git remote add upstream https://github.com/NeuralEnsemble/PyNN.git
and then you can pull in any upstream changes:
$ git pull upstream master
To get PyNN onto your
PYTHONPATH there are many options, such as:
pip editable mode (pip install -e /path/to/PyNN)
creating a symbolic link named
pyNNfrom somewhere that is already on your
PYTHONPATH, such as the
site-packagesdirectory, to the
We try to stay fairly close to PEP8. Please note in particular:
indentation of four spaces, no tabs
single space around most operators, but no space around the ‘=’ sign when used to indicate a keyword argument or a default parameter value.
some function/method names in PyNN use
mixedCase, but these will gradually be deprecated and replaced with
lower_case_with_underscores. Any new functions or methods should use the latter.
we currently target versions 3.8+
Running the PyNN test suite requires the pytest_ package, and optionally the pytest-cov package. To run the entire test suite:
To see how well the codebase is covered by the tests, run:
$ pytest --cov=pyNN --cov-report term --cov-report html
There are currently two sorts of tests, unit tests, which aim to exercise small pieces of code such as individual functions and methods, and system tests, which aim to test that all the pieces of the system work together as expected.
If you add a new feature to PyNN, or fix a bug, you should write both unit and system tests.
Unit tests should where necessary make use of mock/fake/stub/dummy objects to
isolate the component under test as well as possible. The
module is a complete mock simulator backend that may be used for this purpose.
Except when testing a specific simulator interface, unit tests should be able to
run without a simulator installed.
System tests should be written so that they can run with any of the simulators. The suggested way to do this is to write test functions that take a simulator module as an argument, and then wrap these functions with the following decorator:
@pytest.mark.parametrize("sim", (pyNN.nest, pyNN.neuron, pyNN.brian2))
To run only the tests within a file named ‘test_electrodes’ located inside system/scenarios, use:
$ pytest test/system/scenarios/test_electrodes.py
To run a single specific test named ‘test_changing_electrode’ located within some file (and added to registry) inside system/scenarios, use:
$ pytest test/system/scenarios/test_electrodes.py::test_changing_electrode
The best way to get started with contributing code to PyNN is to fix a small
bug (bugs marked “minor” in the bug tracker) in your checkout of
the code. Once you are happy with your changes, run the test suite again to check
that you have not introduced any new bugs. If this is your first contribution
to the project, please add your name and affiliation/employer to
After committing the changes to your local repository:
$ git commit -m 'informative commit message'
first pull in any changes from the upstream repository:
$ git pull upstream master
then push to your own account on GitHub:
$ git push
Now, via the GitHub web interface, open a pull request.
PyNN documentation is generated using Sphinx.
To build the documentation in HTML format, run:
$ make html
doc subdirectory of the source tree. Many of the files contain
examples of interactive Python sessions. The validity of this code can be tested
$ make doctest
PyNN documentation is hosted at http://neuralensemble.org/docs/PyNN
Making a release¶
To make a release of PyNN requires you to have permissions to upload PyNN packages to the Python Package Index, and to upload documentation to the neuralensemble.org server. If you are interested in becoming release manager for PyNN, please contact us via the mailing list.
When you think a release is ready, run through the following checklist one last time:
do all the tests pass? This means running pytest in
test/systemand running make doctest in
doc. You should do this on at least two Linux systems – one a very recent version and one at least a year old, and on at least one version of Mac OS X. You should also do this with multiple Python versions (3.7+).
do all the example scripts generate the correct output? Run the
examples/toolsand then visually check the
.pngfiles generated in
examples/tools/Results. Again, you should do this on at least two Linux systems and one Mac OS X system.
does the documentation build without errors? You should then at least skim the generated HTML pages to check for obvious problems.
have you updated the version numbers in
have you updated the changelog?
Once you’ve confirmed all the above, create source and wheel packages using:
$ python -m build
and check that they install properly (you will find them in the
Now you should commit any changes, then tag with the release number as follows:
$ git tag x.y.z
x.y.z is the release number. You should now upload the documentation
to http://neuralensemble.org/docs/PyNN/ by updating the neuralensemble.github.io
repository (in the
If this is a final release, there are a few more steps:
if it is a major release (i.e. an
x.y.0release), create a new bug-fix branch:$ git branch x.y
upload the packages to PyPI (Do not upload development releases to PyPI):$ twine upload dist/PyNN-x.y.z*
make an announcement on the mailing list
if it is a major release, write a blog post about it with a focus on the new features and major changes
go home, take a headache pill and lie down for a while in a darkened room (-;