Humans communicate with each other through narratives. In order to be useful to humans, data, computations and visualisations need to be embedded into a story.
Reproducibility, reusability, and remixability are fundamental to the scientific process.
We propose to make reusability and composition a first-class citizen in computer aided research by enabling the publication of dynamic and interactive scientific narratives that can be verified, altered, reused, and cited.
Specifically, we will do this by:
-
developing a simple specification for dependencies, code execution, data/parameter inputs, and data/paper outputs. This specification will initially be designed to allow services like mybinder and everware to execute workflows;
-
implementing command-line software and Web interfaces to create, configure, and execute (locally or remotely) scientific repositories that provide this specification;
-
extending the current Jupyter-based systems to also interact with R Markdown and work within the R ecosystem more cleanly;
-
integrating our implementation with GitHub, continuous integration, and pull requests, to facilitate review, interaction, and collaboration;
-
integrating our implementation with Zenodo, to provide DOIs for citation;
-
prototyping ideas for composition of repositories;
-
engaging with the larger academic community including scientists, publishers, and librarians around executable papers and workflows and specifications for same;