Tutorial by

Rebecca Barter

Interactivity

&

@rlbarter

rlbarter

Link to slides:

https://slides.com/rebeccabarter/shiny/fullscreen

Website:

www.rebeccabarter.com

rebeccabarter@berkeley.edu

Link to example app:

https://github.com/rlbarter/ds421-shiny-tutorial-2019

To be interactive or... not to be interactive

Tell a story from the data

Let the audience explore the data

Static visualization better

Interactive visualization better

So you want to be interactive...

Plotly

Flexdashboards

Shiny barebones

User interface (UI)

the web page that shows the app to the user

Server

the computer that powers the app

Every Shiny app has two parts

Starting a new shiny project

app.R

library(shiny)

# Define UI
ui <- fluidPage(
   


















)

# Code that takes inputs from UI and outputs plots, tables, etc
server <- function(input, output) {
   
  







}

# Run the application 
shinyApp(ui = ui, server = server)

UI:

Describes the webpage

(what goes where)

Server:

R code for generating plots and tables under the hood

Runs the app

library(shiny)

# Define UI for application that draws a histogram
ui <- fluidPage(
   
   # Application title
   titlePanel("Old Faithful Geyser Data"),
   
   # Sidebar with a slider input for number of bins 
   sidebarLayout(
      sidebarPanel(
         sliderInput("bins",
                     "Number of bins:",
                     min = 1,
                     max = 50,
                     value = 30)
      ),
      
      # Show a plot of the generated distribution
      mainPanel(
         plotOutput("distPlot")
      )
   )
)

# Define server logic required to draw a histogram
server <- function(input, output) {
   
   output$distPlot <- renderPlot({
      # generate bins based on input$bins from ui.R
      x    <- faithful[, 2] 
      bins <- seq(min(x), max(x), length.out = input$bins + 1)
      
      # draw the histogram with the specified number of bins
      hist(x, breaks = bins, col = 'darkgray', border = 'white')
   })
}

# Run the application 
shinyApp(ui = ui, server = server)

app.R

library(shiny)

# Define UI for application that draws a histogram
ui <- fluidPage(
   
   # Application title
   titlePanel("Old Faithful Geyser Data"),
   
   # Sidebar with a slider input for number of bins 
   sidebarLayout(
      sidebarPanel(
         sliderInput("bins",
                     "Number of bins:",
                     min = 1,
                     max = 50,
                     value = 30)
      ),
      
      # Show a plot of the generated distribution
      mainPanel(
         plotOutput("distPlot")
      )
   )
)

# Define server logic required to draw a histogram
server <- function(input, output) {
   
   output$distPlot <- renderPlot({
      # generate bins based on input$bins from ui.R
      x    <- faithful[, 2] 
      bins <- seq(min(x), max(x), length.out = input$bins + 1)
      
      # draw the histogram with the specified number of bins
      hist(x, breaks = bins, col = 'darkgray', border = 'white')
   })
}

# Run the application 
shinyApp(ui = ui, server = server)

Inputs

Outputs

Let's make a shiny app!

We will pretty much follow:

https://deanattali.com/blog/building-shiny-apps-tutorial/

https://daattali.com/shiny/bcl/

Reactivity

Reactive variables can only be used inside "reactive contexts"

input

Examples of reactive variables:

Examples of reactive contexts:

renderPlot()

reactive()

observe()

what is the value of y?

anything from the list

any object defined based on values from the list

> x <- 5
> y <- x + 10
> x <- 10
> y

input

Note: to access a reactive variable, it must be followed by parentheses "()"

Speeding up shiny

Make sure you are doing most computation outside of the shiny app itself