Shiny: R Friday Forum

Jonathan Hill
2016-02-26

Everything's on GitHub!

Start here:

install.packages("shiny")
require(shiny)
runExample()

Valid examples are "01_hello", "02_text", "03_reactivity", "04_mpg", "05_sliders", "06_tabsets", "07_widgets", "08_html", "09_upload", "10_download", "11_timer"

runExample("01_hello")

Files

Basic

  • app.R
  • server.R
  • ui.R
  • global.R

Advanced

  • helpers.R
  • DESCRIPTION.txt
  • Readme.md
  • Shiny Modules
  • Database (MySQL, PostgreSQL, Redshift)
  • images, .css, .js, etc.

Basic flow

alt text

Execution

Runs once during initialization

# -- global.R -- #

library(dplyr)
df <- read.csv("file.csv")

Runs throughout the session

# -- server.R -- #

server <-
    function(input, output, session) {
        # regular computations
}

server.R

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

    output$<id> <- renderPlot({

        # user input will trigger this statement #

        x <- mtcars[ , input$<id> ]
        plot(x, y)
    })
}

Input

Must be placed within one of the following functions:

Function Output
render a shiny UI component
reactive a reactive expression
observe a reactive observer
isolate a non-reactive copy of a reactive object

Reactivity

When an input changes the server will rebuild each output that depends on it (event if that dependence is indirect). You can control that process with the following functions:

1 - Render(expr) Reactive Output

output$z <- renderText({
    input$a
})

An output automatically updates whenever an input changes

alt text

2 - Reactive(expr) to Modularize Reactions

Used to create objects that will be used in multiple outputs:

x <- reactive({ input$a })

output$y <- renderText({ x() })

output$z <- renderText({ x() })

alt text

3 - Isolate(expr) to Prevent Reactions

Shiny will only update z when b changes:

output$z <- renderText({
    paste(
        isolate(input$a),
        input$b
        )
    })

alt text

4 - Observe(event) to Trigger Code

Use observers to contain computationally intensive pieces of your server:

observeEvent(input$submit, {

    cashflow <- function(
        loanTerm = input$loan_term,
        defaultRate = input$default_rate,
        ...
    )
})

alt text

User Interface

# -- ui.R -- #

ui <- fluidPage(
    textInput("a", "")
)

ui

## <div class="container-fluid">
## <div class="form-group shiny-input-container">
## <label for="a"></label>
## <input id="a" type="text"
## class="form-control" value=""/>
## </div>
## </div>

Layout Options

  • fluidPage
  • fluidRow
  • flowLayout
  • sidebarLayout
  • splitLayout

HTML:

HTML('complete layout here')

  • verticalLayout
  • tabsetPanel
  • navlistPanel
  • navbarPage

What will this layout produce?

ui <- fluidPage(
    fluidRow(column(width = 4), 
             column(width = 4), 
             column(width = 4)
        ),
    fluidRow(column(width = 12)
        )
)

HTML Tags

h1("Header 1")
br()
p(strong("bold"))
code("code")

Header 1



bold

code

More Tags

Type [tags$] + [TAB] for a complete list.

There are over 100 available, and if you can't find what you are looking for, you can just pass the HTML as text through:

HTML("<p>Raw html</p>")

JavaScript and CSS

Short Scripts:

includeCSS('raw CSS here')

includeScript('raw JavaScript here')

Long Scripts and Images:

Place the file in your www/ subdirectory, then reference it with:

# CSS
tags$head(tags$link(rel = "stylesheet",
 type = "text/css", href = "<file name>"))

# JavaScript
tags$head(tags$script(src = "<file name>"))

# Images
img(src="<file name>")

Demo APP

runExample("06_tabsets")

Learn More

Building Tools

GitHub

Everything for this presentation can be found on my github account.

You can also check out Vadim's, too.