geponce-blog

[geponce-blog]

On the use of R-notebooks

Lately, I have been using R for all kind of tasks, from cleaning up large data files (i.e. 200 GB) to the visualization of spatial data. It is impressive how diverse the R-Language has become. Pretty much there’s a package for every single thing you can do with a computer. Without a doubt, major contributions by R-Studio software company have made R a more popular/diverse/complete programming language (yes, it is a programming language! ;-)). One of the most useful contributions I have experienced in the last couple of months is the use of R-notebooks. With the inclusion of R-markdown (markdown on steroids) the idea of R-notebooks is just great. In my opinion, almost all the required characteristics to develop fully-working-reproducible workflows are present (in some cases, moving data along with workflows (scripts) is still an issue if you want to keep things in the free/open source field, however, paying for some storage can solve it, e.g. Github for large files).

One of my favorite features is the concept of `code chunks/engines`. Since I commonly work with different tools, being able to have a section within my workflow (same script) where is possible to call a tool like `awk` (still such a useful tool) to make a bunch of neat text processing and then call R-functions in another chunk to plot something, then incorporate some python code for doing some special processing is just as simple as that. These days, pretty much every new project I start is through the use of an r-notebook, just starting typing about ideas, questions, data analysis, plots, etc. Then, it can evolve into a shiny application, a bookdown or a blogdown.

Cheers!