Testing R code under memory restrictions

I am running some R code on large files in the 1 to 4 GB range. Some of the code draws on packages that include compiled code written in other programming languages. I would like to know if this code will run on a machine with less RAM than I have, say 3 GB. Is there a way to restrict the memory available to R and RStudio, including any non-R code called by R, for testing purposes? Either an R-level or an OS-level solution would be acceptable.

I am working on a Windows 7 machine. I also plan on doing some testing on a Ubuntu virtual machine I have installed on this computer.