Simple template for Python projects

Simple and straightforward template for your Python projects. This is for educational purposes, notably to illustrate how to package your project and to demonstrate how to use the pyproject.toml file.

You could add many tools to this template (use Typer for your CLI, pre-commit hooks, pylint, etc.).

Installation

Create a simple conda environment and install your project with pip. Update the dependencies in requirements.txt to your needs.

conda create -n myenv
conda activate myenv
conda install python pip
pip install .

If you want an editable install, with the extras dependencies:

conda create -n myenv-dev
conda activate myenv-dev
conda install python pip
pip install -e .[dev,test]

Various tips

Matplotlib style sheet

Create your own Matplotlib stylesheet in ~/.config/matplotlib/stylelib/mytheme.mplstyle and use it with plt.style.use(mytheme).

Jupyter notebooks

If you use Jupyter Notebooks, run %load_ext autoreload and %autoreload 2. Your modules will be reloaded automatically.
- If you use VSCode, you can even add them to jupyter.runStartupCommands.
To only install jupyter once: create a conda environment with jupyter notebook and jupyter lab, and install nb_conda_kernels from conda-forge: "This extension enables a Jupyter Notebook or JupyterLab application in one conda environment to access kernels for Python, R, and other languages found in other environments".

Make animations with ffmpeg

To make animations from Matplotlib plots: I prefer to save each frame and then create a video with ffmpeg, than to use Matplotlib animation utilities. ffmpeg is an extremely powerful tool, with many options, but it can be hard to use. But in most cases the following command will be enough:

ffmpeg -framerate $FRAMERATE -i ./plot_%04d.png ./animation.mp4

where you set FRAMERATE to the framerate you want, after having saved your figures to plot_0000.png, plot_0001.png, plot_0002.png, etc.

sshfs

If you want to browse files from a distant server as if they were local files, use sshfs. For me it is very useful for looking at plots and figures that were generated on the server, or to copy them locally easily.

Install the package on your machine.
Create a ~/.ssh/sshfs_config file containing the SSH configuration necessary for sshfs. By default, sshfs uses ~/.ssh/config, but it won't work if you enable ForwardX11 for example; and I prefer to keep things separate.
Run sshfs -F ~/.ssh/sshfs_config remote:/path/to/remote/dir ~/path/to/local/dir

You could create an executable scripts containing more than one of those commands if you want to mount multiple directories. You could also add this to your crontab.

Use fusermount -u ~/path/to/local/dir to unmount.

mxmpl / python-template