R & Python: A Love Story


This announcement from Rstudio caught my attention: https://rstudio.com/solutions/r-and-python/

A few years ago neither Rstudio nor Jupyter were really polyglot and you could not easily use both R and Python at the same time. I consider rpy2 to be a game changer as it allowed to seamlessly have a python jupyter notebook to use R in the background whenever needed.

Rstudio also allowed to have python chunks in Rmd documents but this was a bit cumbersome. It appears that the integration has dramatically improved thanks to reticulate with builds on the C++ interface of both languages (so the data representation conversion from a language to another does not need to go through file and is much cleaner and controled). Rstudio has improved a lot and they are now even embedding (open source) jupyterlab in (closed source) Rstudio server. This answers the need that many data scientist have to navigate between both languages. Although nothing fundamentally changes in term of reproducibility, this is likely to make the Rstudio IDE become more and more popular.