As a complement to the previous note, I have collected every single entity that sponsored a national R conference in France over the last decade.
The data include 9 conferences, plus a forthcoming one. For each conference, the number of sponsors varies between 8 and 46, and the number of organising sponsors varies between 1 and 4. The dataset has 201 sponsors in total: make sure to check the codebook and coding notes for details.
Listing the most frequent sponsors is fairly instructive, as it shows the mix of public (academic or governmental) and private (commercial) institutions that make R conferences possible:
|GDR Stat Santé||public||4|
The private-sector sponsors, in particular, are interesting: the full sample includes mostly statistical consulting and training small businesses, as well as a few academic publishers and some industrial research units, including some large companies working on the life sciences (biology, health and nutrition), as well as energy and transport.
If one looks at the broader spectrum of scientific disciplines and domains covered by the sponsors, (applied) mathematics, statistics and computer science are, of course, well represented, as are the aforementioned life sciences, with a special mention to INRA and agricultural sciences.
Looking at this data, I would recommend to anyone who wants to spend their working lives doing statistical computing/programming (with R) to study either that directly through a mathematics or statistics degree, or to go into biology and study lots of biostatistics, bioinformatics and computational biology.
The types of occupations associated with R come, unsurprisingly, from three main domains: higher education and scientific research, software development, and statistical consulting and training. If you like to compute stuff and/or to teach how to do it, then R is clearly made for you!
Note that, while underrepresented among the sponsors and organising committees of the conferences listed in the data, R is also widespread in the social sciences, where it tends to gradually replace other statistical software like SPSS or Stata.
Also note that the data cover only (French) R conferences and therefore excludes many other relevant ‘computational science’ conferences. As a consequence, physics and other domains are underrepresented by construction, which would likely not occur if one repeated the same exercise on conferences like SciPy or SIGGRAPH.
- First published on July 13th, 2019