Installing Rstudio server on University Network

rstudioserver
ami

#1

I requested to the IT department of the University of Porto to install Rstudio server on our University’s network, they did it although it’s only accessible through the University’s IP (strange solution, but VPN solves this issue).
It was installed on Linux (http://rstudio.up.pt) but I’m always requesting some Linux packages to be installed, in order to have all R packages working. Before this experience with the Rstudio server, I used (and I still use) the Rstudio server AMI (Louis Aslett one; here), and it was much straightforward. So my questions are:

  1. do you have a list of Linux packages or an image that I can use to request the installation by the IT department? A complete solution, like the one from Louis, I had almost no extra packages to install.

  2. which is the minimal hardware power that I should request to compute some Bayesian structural modeling analysis with a few thousand subjects? I know that this is a highly subjective question, but which would you require based on my description?


#2

Are you having trouble installing R packages from the RStudio Server interface, @Sinval? Can you just open up RStudio Server and then install.packages() yourself?

Oops, I didn’t read. If it’s Linux packages, you could take a look at whether your department uses a module system and see whether some of the common ones needed (ncdf4 is a pretty important one for me, for example, but I imagine a C or CPP compiler would be on your list) can just be loaded in without admin help.


#3

There’s a list of common Linux packages needed for R listed here:


#4

Thanks, that’s a good answer to my first question. Now I will wait for the "processing power recommendations.


#5

Take a look at: https://support.rstudio.com/hc/en-us/articles/115002344588-Configuration-and-sizing-recommendations


#6

Thanks, very useful.