8
$\begingroup$

I am working on some neural networks' benchmarks and they take a significant amount of time to finish. So, I am using multiple RStudio sessions to run them simultaneously. Is this safe? I mean is it safe from the perspective of the variables, because I ran the same script with slightly changed the parameters of the neural networks. Are the variables being affected by these sessions? Do they overlap?

e.g. if I have a counter variable does it increase from both sessions?

$\endgroup$
  • 1
    $\begingroup$ do you really need the interface ? you can run in via Rscript ... $\endgroup$ – user4581 Oct 22 '15 at 18:59
  • $\begingroup$ Its more convient to me. If there was a problem with the interface and teh shared memory I would try the Rscript. Thanks anyway! $\endgroup$ – vagelis Oct 22 '15 at 22:03
5
$\begingroup$

Open up your process explorer, do you see two copies of R studio (or R itself) running? If so, than your operating system has allocated separate memory spaces for the two copies of R studio, in which case you are perfectly safe.

In general, you have to do lots of work to share memory, and most well written programs will do their best to protect you from the risks of doing so.

$\endgroup$
  • $\begingroup$ Yeap! You're right! I see two copies of R Studio running! So as you say it must be safe and no memory is shared between them. $\endgroup$ – vagelis Oct 22 '15 at 18:57
  • 4
    $\begingroup$ @vagelis if you do so you waste lots of system resources for Rstudio GUI, better way to go would be to run multiple instances of R alone.... $\endgroup$ – Tim Oct 22 '15 at 20:19
  • $\begingroup$ @Tim Yah, I'll second that, good point. $\endgroup$ – Matthew Drury Oct 22 '15 at 20:26
  • $\begingroup$ An appropriate place for such questions would be the forums the R Studio site hosts. $\endgroup$ – Glen_b -Reinstate Monica Oct 22 '15 at 22:34

Not the answer you're looking for? Browse other questions tagged or ask your own question.