Types of output

Building Dashboards with shinydashboard

Png Kee Seng

Researcher

Output and render functions

In a shinyApp, there will be outputs
For each output, a pair of functions need to be called
- The first is a function in the UI
- This allows us to decide where the output should be placed
- Such an output function typically take the form:
```
<type>output(<identifying label>, ...)
```
The second code required is defined within the server function
- This will take the form:
```
output$<identifying label> <- render<type>(...)
```

Example: Sleep study

Let's first import the data set

library(tidyverse)
sleep <- read_csv("../data/Time Americans Spend Sleeping.csv")

These are the columns

$ index                   : num [1:945] 0 1 2 3 4  ...
$ Year                    : num [1:945] 2003 2004 2005 2006 2007 ...
$ Period                  : chr [1:945] "Annual" "Annual" "Annual" "Annual" ...
$ Avg hrs per day sleeping: num [1:945] 8.57 8.55 8.62 8.63  ...
$ Standard Error          : num [1:945] 0.018 0.026 0.023 0.024  ...
$ Type of Days            : chr [1:945] "All days" "All days" "All days" ...
$ Age Group               : chr [1:945] "15 years and over" "15 years and over" ...
$ Activity                : chr [1:945] "Sleeping" "Sleeping" "Sleeping" ...
$ Sex                     : chr [1:945] "Both" "Both" "Both" ...

textOutput

Let us first place textOutput in the UI

ui <- fluidPage(
  titlePanel("Sleeping habits in America"), 
  textOutput("textlabel"))

server <- function(input, output) {
}

shinyApp(ui, server)

renderText

To render the text output, place renderText in the server
We will need the output$ notation

ui <- fluidPage(
  titlePanel("Sleeping habits in America"), 
  textOutput("textlabel"))

server <- function(input, output) {
  output$textlabel <- renderText("Sleep deprivation is a serious health issue 
                                in many major cities around the world.")
}


shinyApp(ui, server)

textOutput rendered

text output demo with no input interactions.

textOutput with interactions

To allow for interactions between inputs and outputs, we will have to call the inputs using input$

input$<type>

textOutput with interactions

As an example, let's consider a text input.

ui <- fluidPage(
  titlePanel("Sleeping habits in America"), 

  textInput("textlabel", 
            "Tell me your name", 
            value = "", 
            placeholder = "Don"),

  textOutput("textoutlabel"))

server <- function(input, output) {
  output$textoutlabel <- renderText(paste0("Hello ",

                                       input$textlabel,

                                       "Sleep deprivation is a serious health issue 
                                        in many major cities around the world."))
}
shinyApp(ui, server)

textOutput with interactions

text output with interactions.

plotOutput

Place plotOutput() in the UI

ui <- fluidPage(
  titlePanel("Sleeping habits in America"), 

  plotOutput("plotlabel"))


server <- function(input, output) {
}

shinyApp(ui, server)

renderPlot

We will need to place renderPlot() in server()

ui <- fluidPage(
  titlePanel("Sleeping habits in America"), 
  plotOutput("plotlabel"))

server <- function(input, output) {

  output$plotlabel <- renderPlot({
    ggplot(sleep) + 
      geom_histogram(aes(x = `Avg hrs per day sleeping`))
    })

}

shinyApp(ui, server)

plotOutput rendered

plotoutput demo.

plotOutput with selectInput

User inputs can be used to produce different results

ui <- fluidPage(
  titlePanel("Sleeping habits in America"), 
  selectInput("selectlabel", 
              "Select an option",
              choices = c("Histogram", "Boxplot")),
  plotOutput("plotlabel"))

plotOutput with selectInput

server <- function(input, output) {
  output$plotlabel <- renderPlot({
    if (input$selectlabel == "Histogram"){

      ggplot(sleep) + 
        geom_histogram(aes(x = `Avg hrs per day sleeping`))

      }
    else if (input$selectlabel == "Boxplot"){

      ggplot(sleep) + 
        geom_boxplot(aes(x = `Avg hrs per day sleeping`))

  }
  })
}

shinyApp(ui, server)

plotOutput with selectInput

Demo of plotOutput with selectInput

plotOutput with sliderInput

Let's consider another example involving sliderInput

ui <- fluidPage(
  titlePanel("Sleeping habits in America"), 

  sliderInput("sliderlabel", 
              "Average hours of sleep",
              min = 7.5,
              max = 11,
              value = c(7.5, 11),
              step = 0.02),

  plotOutput("plotlabel"))
server <- function(input, output) {

  output$plotlabel <- renderPlot({
    filter(sleep, 
           `Avg hrs per day sleeping` >= input$sliderlabel[1],
           `Avg hrs per day sleeping` <= input$sliderlabel[2]) %>%
      ggplot() +
      geom_boxplot(aes(x=`Avg hrs per day sleeping`))
  })

}
shinyApp(ui, server)

plotOutput with sliderInput

Demo of plotoutput with sliderinput

Icons in shiny

shiny also contains icons that can be used in a shinyApp

Some icons in shiny

¹ https://fontawesome.com/icons

Icons in shiny: An example

Use icon("...")
Recall an earlier example
E.g., icon("bed")

ui <- fluidPage(
  titlePanel("Sleeping habits in America"), 
  sliderInput("sliderlabel",

              list(icon("bed"),

              "Select an option:"),
              min = 7.5,
              max = 11,
              value = c(7.5, 11),
              step = 0.02),
  plotOutput("plotlabel"))

server <- function(input, output) {
  output$plotlabel <- renderPlot({
    filter(sleep, 
           `Avg hrs per day sleeping` >= input$sliderlabel[1],
           `Avg hrs per day sleeping` <= input$sliderlabel[2]) %>%
      ggplot() +
      geom_boxplot(aes(x=`Avg hrs per day sleeping`))
  })
}

shinyApp(ui, server)

Rendered app with icon

Icon demo

Let's practice!

Building Dashboards with shinydashboard