[T.3.1] Integrate Computational Notebooks with ReBenchDB

[smarr]

The goal is to have a somewhat common UI that enables easy reuse of analyses from previous projects, or experiments.

ReBenchDB will need an API to access data, as well as a library of analyses. The API should be general enough to be used by other scripts as well.

Use cases we would want to support are:

• basic analyses (speedups, overhead, scaling, performance regressions, ...)
• correlating of data (different metrics, historical data, ...)

Other than Jupyter there might be other systems that could be appropriate: https://www.samlau.me/pubs/computational-notebooks-design-space_VLHCC-2020.pdf

Ideally, the reports are easily versioned and reproducible.

Analyses, plots, and data processing code would idea be made available as a library of reusable notebook cells, that can be combined, adapted, and shared freely.