RStudio

Installing R and RStudio

R and RStudio are both free, open-source software, available for all commonly used operating systems. R is developed cooperatively and noncommercially under the auspices of the Free Software Foundation; RStudio is a commercial product.

R and RStudio install in the standard manner on each of Windows, macOS, and Linux systems. System-specific instructions for installing R are given below. Regardless of your operating system, you should install R before installing RStudio.

Please read and follow these instructions carefully. Installation assistance will also be availabile from the instructor (John Fox) and teaching assistant (Allison Leanage) prior to the start of the lecture series and during office hours.

Installing R on Windows

bulletVisit the Comprehensive R Archive Network (CRAN) and select a mirror site; a list of CRAN mirrors appears at the upper left of the CRAN home page. I suggest that you use the 0-Cloud mirror, which is the first on the list. Click on the link Download R for Windows, which appears near the top of the page; then click on install R for the first time, and subsequently on Download R x.y.z for Windows (where x.y.z is the current version of R, which is R 4.1.0 at the start of the lectures series). Once it is downloaded, double-click on the R installer. You may take all of the defaults, but I suggest that you make the following modifications:

bulletInstead of installing R in the standard location, C:\Program Files\R\R-x.y.z, I suggest that you use C:\R\R-x.y.z. Again, x.y.z is the current version of R. This will allow you to install packages in the main R library without running R with administrator privileges and may avoid problems that sometimes occur when there are spaces in paths.

bulletYou may take all of the remaining defaults in the R for Windows installer.

Building Packages Under Windows, etc. (Optional)

bulletIf you wish to build packages, or use compiled C, C++, or Fortran code in R, or use the rstan package for Bayesian inference, you will have to install some additional software and properly configure your Windows system. You do not have to be able to build R packages in order to install pre-built Windows binary packages from CRAN, so these steps are generally unnecessary unless you plan to write your own packages, use compiled code, or use rstan. None of these topics are covered in the lecture series.

bulletClick on the Rtools link on the R for Windows CRAN page. Download the current version of the Rtools installer and run it. You may take all of the other defaults. An additional necessary step is to add the Rtools usr\bin subdirectory to your system path; for example, if Rtools is installed in c:\rtoolsxy (which is the standard location for version xy of Rtools), then you would add c:\rtoolsxy\usr\bin; to your system path. Type this location carefully, including the terminating semicolon -- you don't want to mess up your path.

An alternative, and possibly safer, procedure for specifying the path to Rtools is described on the Rtools webpage.

If you want to be able to build R packages outside of RStudio, also add c:\R\R-x.y.z\bin; to the path (assuming that you installed R in the location that I suggested).

bulletIf you want to be able to build PDF help files for packages, download and install the MiKTeX LaTeX system; there is also a link to MiKTeX on the Building R for Windows page. Installing MiKTeX will also allow you to create Sweave and knitr LaTeX documents in RStudio, and to compile R Markdown documents directly to PDF files.

Installing R on macOS

bulletVisit the Comprehensive R Archive Network (CRAN) and select a mirror site; a list of CRAN mirrors appears at the upper left of the CRAN home page. I suggest that you use the 0-Cloud mirror, which is the first on the list. Click on the link Download R for MacOS X, which appears near the top of the page; then click on R-x.y.z.pkg (where x.y.z is the current version of R -- R 4.1.0 at the start of the lectures series), which assumes that you are using macOS 10.11 (El Capitan) or higher. You'll also find older versions of R if you have an older version of macOS. Note: As a general matter, you're probably better off updating your macOS to the current version.

bulletTwo macOS installers are provided for the current version of R: one for Macs that use Intel chips and one for newer Macs that use the Apple M1 chip. The former installer is named R-x.y.z.pkg and the latter is named R-x.y.z-arm64.pkg. At present, I recommend that you use the R-x.y.z.pkg installer whether or not you have a Mac with an Intel chip. The Intel version will work with both kinds of Macs, and there are still some compatibility issues for certain packages associated with the M1-specific version of R.

bulletOnce it is downloaded, double-click on the R installer. You may take all of the defaults.

Building Packages Under macOS, etc. (Optional)

bulletIf you wish to build packages, or use compiled C, C++, or Fortran code in R, or use the rstan package for Bayesian inference, you must install the Apple Xcode developer tools. None of these topics is covered in the lecture series. For macOS 10.7 (Lion) or higher, you can install Xcode for free from the App Store. For earlier versions of macOS, Xcode can be installed from your system DVD or downloaded from the Apple developer website. You do not need Xcode to install pre-built macOS binary packages from CRAN, so this step is unnecessary unless you plan to write your own packages, use compiled code, or use the rstan Bayesian estimation package.

bulletSome R packages include Fortran, C, or C++ code; to build such packages, you will have to install compilers for these languages.The C and C++ compilers are included in the Apple Xcode tools, but you will have to separately download and install a Fortran compiler.

bulletIf you want to be able to build PDF help files, download and install the MacTeX LaTeX system. Installing MacTeX will also allow you to create Sweave and knitr LaTeX documents in RStudio, and to compile R Markdown documents directly to PDF files.

Installing X-Windows on macOS (Optional)

bulletSome R software (e.g., my Rcmdr package) makes use of the Tcl/Tk graphical-user-interface (GUI) builder via the tcltk package to create point-and-click interfaces and to display GUI elements such as progress bars. To use the tcltk package, which is a standard part of the R distribution, you must have the X11 windowing system installed on your Mac. Some other packages that don't use Tcl/Tk, such as the rgl package for dynamic 3D graphics, also require X11.

bulletCheck to see whether the X11 windowing system (X Windows) has already been installed on your computer. If you wish, it should do no harm to skip this step and simply go to the next step to install XQuartz.

For OS X 10.6 and 10.7, the file X11.app should appear in the Utilities folder under Applications in the finder. This application should always be installed under OS X 10.7.

For OS X 10.8 or higher, the file is named XQuartz.app and is no longer included with the operating system. XQuartz.app may also be installed in OS X 10.6 or 10.7.

Note that if you upgrade macOS, you will have to reinstall XQuartz even if you installed it previously.

You may also issue the command capabilities("X11") at the R command prompt. If the response is TRUE then X11 is installed.

bulletIf neither X11.app nor XQuartz.app is installed, install XQuartz from https://www.xquartz.org/. As mentioned, it should do no harm to install XQuartz even if you have X11 currently installed.

Installing R on Linux Systems

bulletVisit the Comprehensive R Archive Network (CRAN) and select a mirror site near you; a list of CRAN mirrors appears at the upper left of the CRAN home page. I suggest that you use the 0-Cloud mirror, which is the first on the list. Click on the link Download R for Linux, which appears near the top of the page. R is available for several Linux distributions (Debian, RedHat, SUSE, and Ubuntu); select your distribution, and proceed as directed.

bulletIf you have a Linux or Unix system that's not compatible with one of these distributions, you will have to compile R from source code; the procedure for doing so is described in the R FAQ (frequently asked questions) list.

Installing RStudio

bulletGo to the RStudio download page, select the free version of RStudio Desktop, click the Download button, and click on the link to the appropriate installer for your operating system (Windows, macOS, or Linux distro). Visit the RStudio IDE home page for more information about RStudio.

bulletOnce it is downloaded, run the RStudio installer and take all of the defaults: In Windows, double-click on the RStudio installer to start the installation; in macOS, double-click on the downloaded RStudio disk-image file, and drag the RStudio icon to the Applications folder.

bulletWhen you first run RStudio, it should detect your R installation and start the R console. To configure RStudio to your taste, select Tools > Global Options (Windows) or RStudio > Preferences (macOS) from the RStudio menus. In particular, I suggest that on the General options screen you deselect Restore .RData into workspace at startup, and set Save workspace to .RData on exit to Never.

bulletIf you encounter difficulties, consult the RStudio troubleshooting guide. or seek help from John or Allison.

Installing R Packages for the Lecture Series

bulletOnce you have installed R and RStudio, you can install additional packages required for the lecture series by typing the following command at the > command prompt in the R Console (and pressing the Enter or return key):

install.packages(c("car", "data.table", "effects", "knitr", "lme4", "rgl", "rmarkdown", "sfsmisc", "tidyverse"))

You can simply copy and paste this command from these installation instructions. Alternatively, you can install packages from the RStudio Packages tab. Be aware that, depending on the speed of your internet connection, it may take some time to download and install these packages and their dependencies.

bullet If you want to try using C++ code within R (not discussed in the lecture series), also install the Rcpp package, install.packages("Rcpp"). You'll also have to install a C++ compiler, as described in the sections above on building packages under Windows and macOS.

bullet Similarly, if you want to use the Stan Bayesian statistical software via the rstan package (not discussed in the lecture series), you'll have to install the package by the command install.packages("rstan"), and also install a C++ compiler.