Elemental Microscopy Quickstart

This repository contains minimal configuration files to get started for an Elemental Microscopy article, using the uv python package manager.

People

This quickstart guide was developed by Colin Ophus and Georgios Varnavides.

Installation Instructions

Edit locally and push changes

Warning

First, ask one of the organization members to give you write access to the forked repo.

Warning

In what follows and the configuration files, you will see references to em-quickstart-uv, which you should replace with the repo name.

Warning

This quickstart guide assumes use of the uv python package manager. You can install uv on your platform following the instructions here. Trust us, you won't regret it.

Steps to edit locally:

• git clone repo (git clone git@github.com:msa-em/change-to-repo-name.git)
• switch to dev branch (git checkout dev)
• edit requirements.in to add your package dependencies, pinning versions as necessary
• edit myst.yml file by:
  • commenting these lines out

    jupyter: true

  • uncommenting these lines

  #  jupyter:
    #    server:
    #      url: 'http://localhost:8888'
    #      token: '512ac78f14e1141db1fac17e8b4099c1e5bc7d589518b38c'

• open two terminal windows:
  • start the jupyter server in one of the terminal windows (uvx --with-requirements=requirements.in --from=jupyter-core --with=jupyterlab jupyter lab --IdentityProvider.token=512ac78f14e1141db1fac17e8b4099c1e5bc7d589518b38c --ServerApp.allow_origin='http://localhost:3000' --port=8888)
  • start MyST in the other terminal window (uvx --from=mystmd myst start)
• edit, commit, and push to dev as per usual
  • make sure NOT to commit your must.yml changes!
• open a draft pull request into main (if one doesn't already exist) and keep pushing your changes to dev
  • this will enable live previews and checks (see below)
  • when you're ready, merge pull request into main. Note this need not be "final" -- it's best practice to merge thematic content changes together.

Note

If you don't plan on editing the notebooks, you can skip the myst.yml and jupyter lab steps above

Preview draft deployed site

The repo has two github actions to automatically deploy computational sites, for the following two case:

• Commits to main directly
• Pull requests into main

If you followed the instructions above (i.e. working off of dev and have an open pull request into main), then you should see a github-actions bot at the top of your pull request which will keep getting edited. Simply click on the Inspect link to see the curvenote staging site, and press preview to see the deployed site based on your latest commit.