Installing R and RStudio

R is a language for statistical computing and graphics. R’s use in the data science, econometrics and marketing communities has taken off over recent years and (at a bare minimum) should be considered as an open source replacement to Stata and SPSS.

Installing R

Watch our YouTube video, in which we walk you through the setup on Windows.

Go to the R website and download the most recent installer for your operating system.

• Windows users: choose the “base” subdirectory, then proceed to the download.
• Mac users: pick the release listed under “latest release” (pick the first, if it does not work, try the second).

We strongly suggest you to install R in the directory C:\R\R-4.x.x\ rather than the default directory, C:\Program Files\R\R-4.x.x\.

Installing RStudio

RStudio provides an easy to work with interface to R, and its format should feel familiar to other software environments like Stata or SPSS.

Download and install the free version of RStudio for your operating system from here.

Verifying your Installation

Open RStudio from the start menu. After starting up, you should see the version corresponding to the one chosen on the website.

Installing additional R Packages

You may need some additional libraries to work with R (e.g., some extra code that helps you to run your statistical analyses).

To install packages, open RStudio (if not already opened in the previous step). In the console, copy and paste the following:

packages <- c("reshape2", "rmarkdown",
              "data.table", "Hmisc", "dplr",
              "stargazer", "knitr",
              "xtable","tidyverse",
              "RSQLite", "dbplyr")

install.packages(packages)

• If you are asked if you want to install packages that need compilation, type n followed by Return. Package compilation is likely to cause some errors, and you’re all good going with packages that have already been compiled (typically, these are earlier versions of the package).
• Wait until all the packages have been installed and the you are done. It may take a while, so be patient

Making R available on the command prompt

You have just installed R and RStudio, and learnt how to open RStudio from the start menu. However, for many of the applications that follow, you are required to access R directly from the command prompt. For example, this will enable you to run a series of R scripts in batch - which will significantly ease the burden of building complex data workflows.

Windows

For you to be able to use R from the command prompt, Windows users need to follow the steps below. On Mac and Linux, R is available from the command line by default.

Warning

Making R available via the PATH settings on Windows.

We need to update our PATH settings; these settings are a set of directories that Windows uses to “look up” software to startup.

• Open the settings for environment variables

  • Right-click on Computer.
  • Go to “Properties” and select the tab “Advanced System settings”.
  • Choose “Environment Variables”

• Alternatively, type “environment variable” (Dutch: omgevingsvariabelen) in your Windows 10 search menu, and press Enter.

• Select Path from the list of user variables. Choose Edit.

• Windows 7 and 8 machines: If you chose your installation directory to be C:\R\R-4.x.x\ during your installation (i.e., you did not use the default directory), copy and paste the following string without spaces at the start or end:

    `;C:\R\R-4.x.x\bin` (replace `4.x.x` by your actual version number!)
  

• Windows 10 machines:

  • Click New and paste the following string:

    C:\R\R-4.x.x\bin (replace 4.x.x by your actual version number!)

  • Click on OK as often as needed.

Warning

Making R available via the PATH settings on Mac/Linux

• Paste this command in your terminal: nano ~/.bash_profile
• Add the following two lines to it:

export R_HOME="/Library/Frameworks/R.framework/Resources"
export R_USER="/Library/Frameworks/R.framework/Resources"

Tip

Keep in mind that after you add a new directory to the PATH variable, you need to start a new command prompt/terminal session to verify whether it worked. Sometimes it may take a couple of minutes until your PATH is recognized by the terminal.

Now let’s verify whether we can open R from the command prompt

Open the command prompt/terminal and enter:

R --version

followed by pressing Return. The expected return begins with:

R version 4.x.x (20xx-xx-xx) -- "Some Funky Name"

Great job - you’ve managed to install R and configure it for use for data-intensive projects!

Making R find packages on the command prompt

You can now access R directly from the command prompt. Nevertheless, code that runs perfectly on R Studio might return Error in library(x) on the command prompt.

Why is that? Sometimes, when running R from the command line, it doesn’t find the packages that were installed in your user library paths.

Solution: Tell R where to find your user library.

Warning

Making R find your user library via the PATH settings on Windows.

• In RStudio, type .libPaths() and note the path to your user directory (typically the one that contains your user name).

• Open the settings for environment variables

  • Right-click on Computer.

  • Go to “Properties” and select the tab “Advanced System settings”.

  • Choose “Environment Variables”

• Select New and name it R_LIBS_USER. Variable value is the path (that you previously noted) to your user directory.

• Check whether .libPaths() only specifies your allocated user directory by typing .libPaths() into a new RStudio session.

  • If not, it is likely that you do not have Admin rights on your computer and R is installed elsewhere. Add another environment variable and name it R_LIBS_SITE. Variable value is the path that is listed second in the .libPaths() output.

Rather want to set R_LIBS_USER on a Mac or Linux machine? Read more here.

Verify that you can access your packages

Close all command prompts/terminals. Open one again, type R to open R and then enter:

library(x)

Note that the command library requires you to specify the package name without quotation marks (e.g., library(tidyverse), not library("tidyverse")).

Expect a return beginning with:

Attaching package: 'x'

Get an error message? Try reinstalling the package using install.packages("name_of_the_page").