Description
Dataflow notebooks extend the Jupyter Notebook environment and IPython kernel to allow users to construct clear dependencies between cells, making it possible to construct a graph that details all defined dependencies between cells. In this environment, unique, persistent cell identifiers make references between cells more robust than existing solutions. Using the dependency graph, we can dynamically update upstream dependencies for a cell or select downstream cells that depend on the current one. This helps users better organize their work, allows for non-linear orderings of cells, and enables greater reproducibility and more opportunities for reuse.