TrovBase reaction to Posit
Language agnosticism FTW
Last week, the #RStats world was taken aback by the news that RStudio is changing its name to Posit. This announcement coincides with RStudio/Posit’s move to become less focused on R and bring in more Python to the provided user workflows. The company also made a very public splash about their work on Quarto, a language agnostic report generation for communicating code results with inline code samples, graphs, and the like. How do we view this development at TrovBase?
Despite some R users' frustration, we see this as an unambiguously positive turn for Posit/RStudio. The obvious comparison to make is to Project Jupyter, an ancestor of Quarto that worked well for power users. Project Jupyter has various extensions to replicate many things that Quarto (and its parent RMarkdown) has out-of-the-box. What Jupyter has shown is the promise of a notebook with inline code and graphs to improve prototyping and tutorials (passing the market test despite dissent with Google making such notebooks easily runnable in browser via Google servers through its already popular Collab service).
Quarto promises something even more complete. For people who want automated report generation with the benefits of code and graphs embedded throughout, Quarto works well without having to spend lots of time searching for the right extension or debugging pandoc (usually). That's so much of why we are excited about it at TrovBase: we can generate qmd documents in either R or Python (even interspersed if necessary) that render as a gorgeous document for users without us having to choose great defaults for documents; we can implement the best practices that Posit is applying. This means we can get our users on board with this new workflow for creating scientific output on the same platform that helps them use best practice R & Python code for their statistical analysis.
Of course, Quarto is an emerging standard, so in the short-run we plan to give users the option to stitch together their tabulations and regressions as a Jupyter notebook, along with Google Collab integration. We anticipate there is more thinking to be done on our end for how we make it easy for users to do work with scripts first that then make it in the notebook or other generated document later. But we are very excited about this development so that TrovBase datasets can have attached reports that generate for any kind of collaborator, whether those collaborators want to use R or Python. Within a team or the greater scientific community, TrovBase will ensure scientific code is accessible and usable as a template to build with for existing or new datasets.
We love R here at TrovBase, but we know that Python has advantages of its own and reasons for its use for many. Being able to have the best of each language's features, AND the tools that have been built around them by their communities, confirms that this is the right time for building data science tools using both languages that bring standardized workflows to everyone who wants to work with data.
If you enjoyed this post, you are probably a great candidate for TrovBase. We are still accepting our early cohort of first users. Join the wait list here.
 | A guest post by
|