Computational Notebook Submissions

Below, we outline the categories that we will accept for computational notebook submissions. We have included an innovative uses category, to be inclusive of as many creative uses of notebooks as possible, however, this year will only accept Jupyter Notebooks, R Markdown documents (including Shiny apps), and Quarto Markdown documents.

We intend to have an open line of communication during the submission process, so if you have questions or concerns about these formats, please contact us at usrsecon2023@easychair.org.

Templates

1. Reproducible Documents

There are many good reasons to publish a reproducible document with a notebook. You may want to publish a lesson or showcase a tool or a workflow. These documents must be reproducible; that is, they are executable and produce the same output when the cells are run in order from top to bottom. Singular notebooks and markdown documents will be compiled to HTML pages, but you also have the option to use JupyterBook (Jupyter), Bookdown (R Markdown), or Quarto Book to create full books. These documents may not exclusively contain text; they must include a significant amount of code. The reproducibility requirement excludes the use of interactive components used to select or manipulate data or run a computation, however, widgets used solely to display or visualize data are allowed. Reproducible documents are expected to be executed sequentially, from beginning to end, without additional interaction or input on behalf of the user.

Example use cases

  • Publish a lesson
  • Showcase a tool
  • Showcase a workflow
  • Notebook-based research/data papers
  • Instructional documents

2. Interactive Notebooks

With the exception of display and visualization widgets, interactivity generally renders notebook execution irreproducible. But just because the notebooks aren’t reproducible doesn’t mean they aren’t valuable to the research software community; they may promote accessibility and reusability of research data and code. If you are using Python and Jupyter, we particularly encourage the use of Voilà to turn any interactive notebook into a standalone dashboard or web application by hiding code cells but leaving markdown and interactive widgets. Shiny natively supports both narrative documents interspersed with Shiny components and full page applications. Some widget libraries remain interactive when compiled to HTML and will be published in the same manner as reproducible notebooks. Notebooks with widgets that don’t remain interactive must be hosted on an interactive environment such as Binder or shinyapps.io. Value comes from the interactive elements of the notebooks, are often executed non-linearly, and require interaction or input from the user.

Example use cases

  • Narrative workflow with interactive components
  • Interactive tutorial or teaching tool
  • Data analysis dashboard
  • Web applications

Examples

3. Innovative Uses of Notebooks

Notebooks continue to evolve in their capabilities and use-cases. In an attempt not to exclude innovative notebook formats, we will have an open category of submission.

Example use cases

  • New plugin capabilities
  • Literate programming packages
  • Code-generated art