Class Activity 18

# load the necessary libraries
library(tidyverse)
library(shiny)
library(readr)
library(janitor)
library(purrr)
library(lubridate)
library(DT)
library(ggthemes)
library(rvest)
library(polite)

Writing the Server Function

The server part of your Shiny app listens for changes in the inputs and dynamically updates the outputs. Here’s how you can set it up:

  • Reactive Expression: Creates a subset of covid_data based on user inputs. This reactive expression ensures that any change in inputs automatically triggers data filtering.
  • Render Functions:
    • renderPlot: Generates and renders a plot based on the filtered data.
    • renderDataTable: Displays the filtered data as a table.

Running the App

ui <- fluidPage(
  titlePanel("Tracking Covid in Minnesota"),
  sidebarLayout(
    sidebarPanel(
      sliderInput(inputId = "months", label = "Month?", min = 1, max = 12, value = c(3,6)),
      radioButtons(inputId = "years", label = "Year?", choices = 2020:2023, selected = 2023),
      selectInput(inputId = "county", label = "County?", choices = county_names, selected = "Rice")
    ),
    mainPanel(
      plotOutput(outputId = "plot"), br(),
      DT::dataTableOutput(outputId = "table")
    )
  )
)
server <- function(input, output) {
    filtered_data <- reactive({
        subset(covid_data,
               counties %in% input$county &
               month >= input$months[1] & month <= input$months[2] & 
          year == input$years) })

    output$plot <- renderPlot({
        ggplot(filtered_data(), aes(x=dates, y=cases, color=counties)) + theme_economist_white()+
                geom_point(alpha=0.5, color = "blue") + theme(legend.position = "none") +
                    ylab("Number of Cases") + xlab("Date")})

    output$table <- DT::renderDataTable({
        filtered_data()})
        
}
app <- shinyApp(ui = ui, server = server, options = list(height = 1200))
app

(Time permitting) Additional Layout

To enhance your Shiny app, first install and load the shinythemes package to access additional aesthetic themes. Update your UI to utilize a navbarPage structure, incorporating tabPanel elements to separate the plot and data table into distinct tabs. Apply one of the themes, such as “cerulean”, to improve the visual appeal and user experience of your app. Replace the month and year selection inputs in your Shiny app with a dateRangeInput to allow users to select a specific date range for viewing COVID-19 trends.

library(shinythemes)
ui1 <- navbarPage(theme = shinytheme("cerulean"), title = "Tracking Covid in Minnesota",
  tabPanel("Plot",
    fluidPage(
      titlePanel("Covid Trends by County"),
      sidebarLayout(
        sidebarPanel(
          dateRangeInput("dateRange", 
                         "Select Date Range", 
                         start = min(covid_data$dates), 
                         end = max(covid_data$dates),
                         min = min(covid_data$dates), 
                         max = max(covid_data$dates)
                         ),
          selectInput(inputId = "dv", label = "County", choices = levels(covid_data$counties), selected = "Aitkin")
        ),
        mainPanel(
          plotOutput(outputId = "plot")
        )
      )
    )
  ),
  tabPanel("Data",
    fluidPage(
      titlePanel("Covid Data Table"),
      sidebarLayout(
        sidebarPanel(),  
        mainPanel(
          DT::dataTableOutput(outputId = "table")
        )
      )
    )
  )
)


server1 <- function(input, output) {
  filtered_data <- reactive({
    filter(covid_data, 
           counties %in% input$dv &
           dates >= input$dateRange[1] & dates <= input$dateRange[2])
  })

  output$plot <- renderPlot({
    data <- filtered_data()  
    ggplot(data, aes(x = dates, y = cases, color = counties)) +
      geom_point(alpha = 0.5, color = "blue") +
      labs(y = "Number of Cases", x = "Date") +
      theme_minimal()
  })


  output$table <- DT::renderDataTable({
    filtered_data()  
  })
}

app1 <- shinyApp(ui = ui1, server = server1, options = list(height = 1200))
app1