Shiny is R's framework for creating interactive web applications. No HTML/CSS/JavaScript required!
Building Web Apps with Shiny
1. Shiny Fundamentals
Core Concepts:
- UI (User Interface): Defines the app layout and inputs
- Server: Contains the app logic and computations
- Reactivity: Automatic updates when inputs change
# Minimal Shiny app (app.R)
library(shiny)
ui <- fluidPage(
sliderInput("num", "Choose number:", 1, 100, 50),
plotOutput("hist")
)
server <- function(input, output) {
output$hist <- renderPlot({
hist(rnorm(input$num), main = paste(input$num, "Random Values"))
})
}
shinyApp(ui, server)2. Building Interfaces
Key Inputs:
ui <- fluidPage(
# Common input types
selectInput("var", "Variable:", names(mtcars)),
numericInput("bins", "Bin count:", 10, min = 1),
dateRangeInput("dates", "Date range:"),
checkboxGroupInput("species", "Species:",
choices = unique(iris$Species)),
actionButton("go", "Update Results")
# Layout with sidebar
sidebarLayout(
sidebarPanel(inputs...),
mainPanel(outputs...)
)
)3. Reactive Programming
How Shiny Updates:
server <- function(input, output) {
# Reactive expression (cached calculation)
data <- reactive({
iris[iris$Species == input$species, ]
})
# Observer (side effects)
observeEvent(input$go, {
showNotification("Data updated!")
})
# Render outputs
output$plot <- renderPlot({
ggplot(data(), aes(Sepal.Length, Sepal.Width)) +
geom_point()
})
}4. Professional Interfaces
Enhance UX with:
ui <- navbarPage("My App",
tabPanel("Analysis",
fluidRow(
column(4, wellPanel(...)),
column(8, ...)
)),
tabPanel("About", includeMarkdown("about.md")),
# Add CSS/JavaScript
tags$head(
tags$style(HTML("
body { background-color: #f9f9f9; }
.well { border: 1px solid #ddd; }
"))
)
)5. Building Dashboards
With shinydashboard:
library(shinydashboard)
ui <- dashboardPage(
dashboardHeader(title = "Sales Dashboard"),
dashboardSidebar(
selectInput("region", "Region:", c("North", "South")),
menuItem("Overview", tabName = "overview")
),
dashboardBody(
tabItems(
tabItem("overview",
fluidRow(
valueBoxOutput("total_sales"),
box(plotOutput("trend"))
))
)
)
)6. Deploying Shiny Apps
Publishing Options:
# 1. shinyapps.io (free tier available)
rsconnect::deployApp()
# 2. RStudio Connect (enterprise)
# 3. Shiny Server (self-hosted)
# 4. Docker container
# Bundle for sharing
shiny::runUrl("https://example.com/app.zip")
0 Interaction
0 Views
0 Likes
×
