There are a number of powerful interfaces allowing R to interface with major database types, but has anyone written in an opinionated way about which DBs are more or less natural to use with R, maybe in a comparative way?
The question is general, motivated by a specific use-case: A small organization which owns some internet-connected industrial equipment is interested in some visualization/ dashboarding, w data from many pieces of equipment, updated at high freq time intervals (data available 10-100x per second). Therefore, I have not so wide (about 150 - 200 pieces of equipment, ie variables), but really long data (~300k observations per day).
For now, the data is nice and rectangular- I could just use csvs, though I’m intrigued with the fst and feather packages. I’m worried, however, that I need to take a longer view and anticipate growth in the system, and perhaps try to do this in a more “mature” way.
Then again, maybe simpler is better, and I should avoid the overhead of a a DB data repo.
More context: I’m an intermediate R user, and love to do analysis with the many powerful tidyverse tools, and sometimes Shiny, but I know little beyond the most basic SQL commands, and almost nothing about DB admin.
Any ideas are appreciated.