We welcome people who want to make contributions to Numba, big or small! Even simple documentation improvements are encouraged. If you have questions, don’t hesitate to ask them (see below).
We have a public development mailing-list that you can e-mail at firstname.lastname@example.org. If you have any questions about contributing to Numba, it is ok to ask them on this mailing-list. You can subscribe and read the archives at https://groups.google.com/a/continuum.io/forum/#!forum/numba-dev, and there is also a Gmane mirror allowing NNTP access: http://news.gmane.org/gmane.comp.python.numba.devel/
If you want to contribute, we recommend you fork our Github repository, then create a branch representing your work. When your work is ready, you should submit it as a pull request from the Github interface.
If you want, you can submit a pull request even when you haven’t finished working. This can be useful to gather feedback, or to stress your changes against the continuous integration platorm. In this case, please prepend [WIP] to your pull request’s title.
Numba has a number of dependencies (mostly numpy and llvmpy) with non-trivial build instructions. Unless you want to build those dependencies yourself, we recommend you use Conda to create a dedicated development environment and install precompiled versions of those dependencies there:
$ <path_to_miniconda>/conda create -n numbaenv python=3.4 llvmpy numpy
This installs an environment based on Python 3.4, but you can of course choose another version supported by Numba.
To activate the environment for the current shell session:
$ source <path_to_miniconda>/activate numbaenv
Those instructions are for a standard Linux shell. You may need to adapt them for other platforms.
Once the environment is activated, you have a dedicated Python with the requested dependencies:
$ python Python 3.4.1 |Continuum Analytics, Inc.| (default, May 19 2014, 13:02:41) [GCC 4.1.2 20080704 (Red Hat 4.1.2-54)] on linux Type "help", "copyright", "credits" or "license" for more information. >>> import llvm >>> llvm.__version__ '0.12.7'
For a quick development workaround, we recommend you build Numba inside its source checkout:
$ python setup.py build_ext --inplace
This assumes you have a working C compiler and runtime on your development system.
Numba is validated using a test suite comprised of various kind of tests (unit tests, functional tests). The test suite is written using the standard unittest framework.
The various test modules are inside the numba/tests directory. There are two entry points to run the test suite:
if you want to run the whole test suite (which will take a couple of minutes), call the runtests.py script; in particular, the -m flag will parallelize the test suite into several processes:
$ python runtests.py -m
if you want to run an individual test module, invoke it as a python module, for example:
$ python -m numba.tests.test_closure
Both the global test runner runtests.py and individual modules allow you to pass various options to influence test running and report. Pass -h or --help to get a glimpse at those options.
Any non-trivial change should go through a code review by one or several of the core developers. The recommended process is to submit a pull request on github.
A code review should try to assess the following criteria:
All Python code should follow PEP 8. Our C code doesn’t have a well-defined coding style (would it be nice to follow PEP 7?). Code and documentation should generally fit within 80 columns, for maximum readability with all existing tools (such as code review UIs).
The repository’s master branch is expected to be stable at all times. This translates into the fact that the test suite passes without errors on all supported platforms (see below). This also means that a pull request also needs to pass the test suite before it is merged in.
Numba is to be kept compatible with Python 2.6, 2.7, 3.3 and 3.4 under at least Linux, OS X and Windows. Also, Numpy versions 1.6 and upwards are supported.
We don’t expect invidual contributors to test those combinations themselves! Instead, we have a continuous integration platform. Part of the platform is hosted at Travis-CI. Each time you submit a pull request, a corresponding build will be started at Travis-CI and check that Numba builds and tests without any errors. You can expect this to take less than 20 minutes.
Some platforms (such as Windows) cannot be hosted by Travis-CI, and the Numba team has therefore access to a separate platform provided by Continuum, our sponsor. We hope parts of that infrastructure can be made public in the future.