The most recent release of R is available from the command-line and for web applications written with Shiny.

Configuring web applications

Web applications written with Shiny use Shiny Server Open Source as an application server. This will automatically take care of starting up your code when someone tries to access it with a web browser. To deploy your application, do the following:

  1. Place your application files in a directory under your site root.

  2. Edit the .silk.ini file in your site root. Add a section similar to the following.

    [app]
    type = rlang
    root = myapp
    uri = /myapp

    The .silk.ini section name should be either "app" or any descriptive name you wish for your application that starts with "app: ". This could be useful if you wish to configure multiple apps.

    root is the directory under your site root containing all your application files.

    uri is the path part of a browser URL after the hostname at which your app should be reached.

  3. Notify SAA that you have made these changes, so that Shiny Server can be configured for you and the web server can be set up to speak with it.

Reloading applications

Shiny Server looks for changes to applications and will usually reload them automatically when necessary. If it misses a change, just update the timestamp on one of your files. For example, touch myapp/server.R.

Console logs

App server logs are available in /var/log/shiny-server/. Log file names are prefixed with the app path, your username, and a timestamp, with a new file each time your app is relaunched (e.g., myApp-mynetid-20191202-142304-23813.log). When debugging, make sure to view the most recent file.

Changing server options

Certain application options for timeouts and HTTP headers can be configured by creating a .shiny_app.conf file in your application directory that specifies the new settings. See the Local App Configurations section in the Shiny Server documentation for a list of settings, and further down the page for the full description of each option.

Packages

You can install CRAN packages within your own library. When you first try to install a package, R will let you know that you cannot write to the system-wide default library and then offers to create one in your home directory, located at ~/R/. All of your installed packages will live there.

For example, you can install caret within R by typing the following:

install.packages("caret")