Development of exoplanet
is actively happening on GitHub and we would love your
contributions. There are a few different methods of contributing to
exoplanet
and the details are discussed below.
If you run into issues, bugs, or anything else, it is very useful if you can post an issue on the GitHub repository. When you post an issue, please provide the details to reproduce the issue. For example, if you find a bug, please provide a standalone and executable snippet of code that demonstrates the issue. It’s also useful to include details about your platform and the versions of key packages that your using.
If you’re not familiar with the workflow for contributing code to a GitHub repository, an excellent place to start is the AstroPy developer docs.
After getting some familiarity with the workflow, you should fork the exoplanet repository and clone it to your local machine:
git clone https://github.com/YOURUSERNAME/exoplanet.git
cd exoplanet
git checkout -b BRANCHNAME
for some name BRANCHNAME
describing your contribution.
Then you should set up an isolated environment for development using a Conda
environment,
virtualenv, venv, or similar. If using conda
,
you can get the current development environment from the environment.yml
file:
conda env create --prefix env -f environment.yml
conda activate ./env
If you have an existing conda
environment for exoplanet
, you can update
it using:
conda env update --prefix ./env -f environment.yml --prune
If you’re using a pip
based environment, you can install the developer
dependencies as follows:
python -m pip install -U pip
python -m pip install -U .[dev]
After your environment is set up, you can install the current development
version of exoplanet
:
python -m pip install -e .
exoplanet
is mostly arranged as a typical Python project with the module
root in the src/exoplanet
directory. But there are a few directions that can
be useful before diving in:
1. Tutorials: The tutorials are written using the jupytext tool to keep the size of the repository
from blowing up. The tutorial files are saved (using jupytext) in the
docs/tutorials
directory as .py
files in the py:light
format. You
can edit these files as if they are Jupyter notebooks by using the extension:
jupyter nbextension install --py jupytext
jupyter nbextension enable --py jupytext
If you are contributing a new tutorial, you should copy one of the existing ones and try to follow roughly the same format.
2. Case studies: These are more in depth tutorials that require more computational run time. These can be found in the case studies repo and there is more information there about how to contribute.
3. Theano ops: exoplanet
comes bundled with a set of custom Theano ops
that are implemented in src/exoplanet/theano_ops
. As a user, you’ll rarely
interact with these directly and we haven’t put a lot of work into making them
user friendly, but if you are interested in diving in, here are some tips.
First, you should check out the Theano docs that describe how to develop new ops
in Python and
C/C++.
Most of the exoplanet
ops are implemented in C++ for speed and we’ve made
the design decision to separate the “science” code (which implements the actual
operation without any dependency on Theano) and “wrapper” code (which sets up
the interface). The science code is implemented as a header-only C++ library in
src/exoplanet/theano_ops/lib/include/exoplanet
and then, in most cases,
these functions are accessed via the pybind11 interface implemented in
src/exoplanet/theano_ops/driver.cpp
. Then the wrappers are implemented as
submodules in src/exoplanet/theano_ops
. A good place to start is the
KeplerOp
implemented in src/exoplanet/theano_ops/kepler.py
.
If you’re contributing a change to the code (either a new feature or bug fix), make sure that you implement at least one test that checks the behavior of your code. Then, you should run all of the unit tests before submitting a pull request using the following command:
python -m pytest -v tests
We have a pretty strict (but easy to implement!) set of style guidelines for the
codebase. For Python code, we use isort and black. The custom settings for these projects can be
found in pyproject.toml
. Before opening a pull request, you can run the
formatters as follows:
isort -rc src
black src
Or, you can use pre-commit to automatically apply the formatting whenever you commit:
python -m pip install -U pre-commit
pre-commit install
Note
Most of this build process is based on the October 2019 update to
this blog post so you should check that out if you want more info.
This section is mainly internal, but these are the steps that should be executed to produce a new release.
Update citation date and version in src/exoplanet/citations.py
.
Update changelog date in HISTORY.rst
.
3. Tag a GitHub release on both the exoplanet repository and the case studies repository