For general guidance in using RStudio, see the RStudio documentation.
RStudio memory issues
RStudio crashes when it runs out of memory. This is because memory is a finite resource, and it’s not easy to predict memory usage or exact availability. But if your data is of order of a couple of gigabytes or more, then simply putting it all into a dataframe, or doing processing on it, may mean you run out of memory. For more about memory capacity in the Analytical Platform, and how to work with larger datasets, see the memory limits section.
If RStudio crashes on startup, and you’ve identified from Grafana that it is because the memory is full, then you can fix it by clearing your RStudio session.
Once RStudio is running again, you can get a better understanding of what takes up memory by using the pryr package. To free up a bit of memory, for example when a variable points to a big dataframe, you can instead assign something a null to the variable, and then run
gc(). To free up all the user memory you can click on the ‘broom’ to clear the environment.
Clearing your RStudio session
The RStudio session is the short term ‘state’ of RStudio, including:
- which project and files are open (displayed in the RStudio interface)
- console and terminal history
- the global environment – R variables that have been set and are held in memory
The session is persisted between restarts of RStudio, to make it easier for you to pick up where you left off. However you can clear it to solve a number of problems, including if the memory is full.
To clear the RStudio session:
Close RStudio, if it is open in any window (because it continually saves its session to disk).
Open the control panel, navigate to Analytical tool, click resetting your home directory.
Select the Restart button for RStudio.
In the control panel, select Open for RStudio. It may take between one and five minutes before RStudio is available. You may need to refresh your browser for the tool to load.