Introduction to Shiny

Session - First Shiny App

Zoë Turner

Not using the template

On a computer - first create a project
On the cloud and next step - open a new R file and save it as app.R

Type shinyapp in the R file and the use Shift + Tab
Type "Hello, world!" between the brackets after the function fluidPage
Save file!
Run the app

Application types

We used the app.r but it is possible to use (or see) two files being used for ui and server.
Early on Shiny relied on two files but it has since been updated to one file.
Which format you use depends on your app and what you prefer.

{golem}

If the app is big or is being built for production then consider looking into {golem}
{golem} uses many principles from R package development - check out nhsbsaShinyR template
It also supports modules which can be reused across complex shiny apps

No titles

Titles using titlePanel()

The template app had a title in the ui <- fluidPage() section

    # Application title
    titlePanel("Old Faithful Geyser Data"),

but this is not included using this method.

Let’s add the title to the minimal template
Change the name
Note there is a comma at the end of the line - not good practice to leave in but it works and useful for adding future code to this section

Bit more to the app

Replace the text after the function fluidPage() with:

  selectInput("dataset", label = "Built in R datasets", choices = ls("package:datasets")),
  verbatimTextOutput("summary"),
  tableOutput("table")

Save and Run

UI functions

fluidPage() sets up the visual structure of the page
selectInput() lets the user interact with the app by providing a value
verbatimTextOutput() and tableOutput() are output controls that tell Shiny where to put rendered output

ui <- fluidpage(
  selectInput("dataset", label = "Built in R datasets", choices = ls("package:datasets")),
  verbatimTextOutput("summary"),
  tableOutput("table")
)

What’s in the functions?

HTML!
Call a function outside a Shiny app (in the console) to see shiny::fluidPage()

Server functions

We have the code saying what’s possible to select (UI) but not what is going to appear (Server)

So if we add the following between the { after function(input, output, session), save and run:

  output$summary <- renderPrint({
    dataset <- get(input$dataset, "package:datasets")
    summary(dataset)
  })
  
  output$table <- renderTable({
    dataset <- get(input$dataset, "package:datasets")
    dataset
  })

Exercise (from Mastering Shiny)

Take the following code fix the error

ui <- fluidPage(
  sliderInput("x", label = "If x is", min = 1, max = 50, value = 30),
  "then x times 5 is",
  textOutput("product")
)

server <- function(input, output, session) {
  output$product <- renderText({ 
    x * 5
  })
}

shinyApp(ui, server)

How to create a new app

Remember the first slide on how to create a new app (from next step)

10:00

Solution

library(shiny)

ui <- fluidPage(
  sliderInput("x", label = "If x is", min = 1, max = 50, value = 30),
  "then x times 5 is",
  textOutput("product")
)

server <- function(input, output, session) {

  output$product <- renderText({
    input$x * 5
  })
}

shinyApp(ui, server)

Reduce duplication!

No matter what you code in, duplication will eventually cause more work:

Cause bugs as one change can be in multiple places
Can be computationally wasteful

In R code we avoid duplication using variables or functions but these don’t work or aren’t efficient in Shiny apps

Reactive expressions

Uses a function and new code structure with round and curly brackets

reactive({})

We can update the code in the shiny app by adding into the server object

# Create a reactive expression
  dataset <- reactive({
    get(input$dataset, "package:datasets")
  })

Removing the lines

dataset <- get(input$dataset, "package:datasets")

Let’s try this (it will be broken!)…

Famous error message

Error: object of type 'closure' is not subsettable
Error: cannot coerce class ‘c("reactiveExpr", "reactive", "function")’ to a data.frame

Although data is referred to it needs to be the reactive form so replace dataset with dataset()

And try running again…

Exercise from Mastering Shiny

Let’s make the following code reactive and remove the repetition

library(shiny)

ui <- fluidPage(
  sliderInput("x", "If x is", min = 1, max = 50, value = 30),
  sliderInput("y", "and y is", min = 1, max = 50, value = 5),
  "then, (x * y) is", textOutput("product"),
  "and, (x * y) + 5 is", textOutput("product_plus5"),
  "and (x * y) + 10 is", textOutput("product_plus10")
)

server <- function(input, output, session) {
  output$product <- renderText({ 
    product <- input$x * input$y
    product
  })
  output$product_plus5 <- renderText({ 
    product <- input$x * input$y
    product + 5
  })
  output$product_plus10 <- renderText({ 
    product <- input$x * input$y
    product + 10
  })
}

shinyApp(ui, server)

10:00

Solution

library(shiny)

ui <- fluidPage(
  sliderInput("x", "If x is", min = 1, max = 50, value = 30),
  sliderInput("y", "and y is", min = 1, max = 50, value = 5),
  "then, (x * y) is", textOutput("product"),
  "and, (x * y) + 5 is", textOutput("product_plus5"),
  "and (x * y) + 10 is", textOutput("product_plus10")
)

server <- function(input, output, session) {
  
  product <- reactive({input$x * input$y})
  
  output$product <- renderText({
    product()
  })
  output$product_plus5 <- renderText({
    product() + 5
  })
  output$product_plus10 <- renderText({
    product() + 10
  })
}

shinyApp(ui, server)

Cheat sheets

Cheat sheets are a great way of seeing the possible functions in a package like {shiny}

But are also built into RStudio in Help > Cheat Sheets > Web applications with Shiny

Where to find Cheat sheets

Download from the web
Are pdf and in landscape format

Exercise (from Mastering Shiny)

Take the following code and move the lines you think you’ll need to the right places in a shiny app

tableOutput("mortgage")
output$greeting <- renderText({
  paste0("Hello ", input$name)
})
numericInput("age", "How old are you?", value = NA)
textInput("name", "What's your name?")
textOutput("greeting")
output$histogram <- renderPlot({
  hist(rnorm(1000))
}, res = 96)

10:00

Solution

library(shiny)

ui <- fluidPage(
  textInput("name", "What's your name?"),
  textOutput("greeting"),
)

server <- function(input, output, session) {
  output$greeting <- renderText({
    paste0("Hello ", input$name)
  })
}

shinyApp(ui, server)

End session

Acknowledgements

Mastering Shiny by Hadley Wickham
Beginners course by Chris Beeley aimed at relatively novice R users
Shiny workshop for NHS-R Community
Shiny workshop materials on the NHS-R Community GitHub.