Install either with the conda
environment file (strongly recommended) or via pip using the [dev] extras:
--This uses ssh to clone the repo, feel free to use your protocol of choice--
$ git clone [email protected]:AndrewRook/ptplot.git
$ cd ptplot
$ conda env create -f environment.yml
$ conda activate ptplot-dev
$ pip install -e .
--OR (Note that you will need to separately install nodejs)--
$ pip install -e .[dev]ptplot uses pytest, flake8, mypy, and black. All of these must
pass in order for a PR to be merged, so it's valuable to run them
yourself locally before pushing changes:
$ python -m mypy ptplot/
$ python -m pytest tests/ ptplot/
$ python -m black -l 120 ptplot/
$ python -m flake8 ptplot/ptplot's primary form of documentation is currently Jupyter
notebooks. Unfortunately, due to how plotly renders animations,
the animation notebook is quite large in size. Whenever you
work with ptplot animations in the notebook, please check the
size of the resulting notebook before committing it to the repo.
Note: When you make an update to any of the custom extensions, you may need to force-reload the notebook pages in order to clear the cache and make Jupyter look for the new JS files.
- Make sure that all changes you want have been merged to
main. - Rerun all the notebooks and make sure they still work. If the
results have changed in any way, make sure that's reflected in the
versions that are in
main. If not, rerun them and PR those updates in. - In GitHub, create a new release
based on the
mainbranch. Give it a version tag that makes sense.ptplotuses semantic versioning, with no "v" preceding the version numbers. - Go to the Actions tab for the repo and confirm that the publish action runs successfully (takes a couple of minutes).