Shiny 02: Intro to Shiny (cont.)

Explore additional features in Shiny App.

Welcome to our shiny app development class! Throughout this session, we will continue diving into the world of Shiny and explore its incredible potential for creating interactive web applications. Today we will use the function reactiveVal.

1 Shiny App to collect data.

The reactiveVal function is utilized to create a “reactive value” object which has special capabilities for reactive programming. It serves as a variable that allows both reading and writing of values. Whenever the value is read from a reactiveVal object, the calling reactive expression becomes dependent on it. Similarly, when the value is altered, any reactives that were previously dependent on it are notified.

Let’s make a shiny app and see how this works!

# Install and call the following packages.
library(shiny)
library(DT)

# Initialize an empty data frame to store the collected data
collected_data <- data.frame(
  Treatment = character(0),
  PlotNumber = numeric(0),
  StandCount = numeric(0),
  PlantHeight = numeric(0),
  Severity = numeric(0)
)

# Create a variable to store the selected rows
selected_rows <- reactiveVal()

# Define the UI
ui <- fluidPage(
  titlePanel("Soybean Field Data Collection"),
  
  sidebarLayout(
    sidebarPanel(
      selectInput("treatment", "Select Treatment:", c("T1", "T2", "T3")),
      numericInput("plot", "Enter Plot Number:", value = 101, min = 101, max = 304),
      numericInput("stand_count", "Enter Stand Count:", value = 0),
      numericInput("plant_height", "Enter Plant Height (in cm):", value = 0),
      numericInput("severity", "Enter Severity Rating:", value = 0),
      actionButton("submit", "Submit Data"),
      actionButton("delete", "Delete Selected Row"),
      downloadButton("downloadData", "Download Data")
    ),
    
    mainPanel(
      DTOutput("dataTable")
    )
  )
)

# Define the server
server <- function(input, output, session) {
  collected_data_reactive <- reactiveVal(NULL)
  
  observeEvent(input$submit, {
    new_entry <- data.frame(
      Treatment = input$treatment,
      PlotNumber = input$plot,
      StandCount = input$stand_count,
      PlantHeight = input$plant_height,
      Severity = input$severity
    )
    # Append the new entry to the collected data
    collected_data <- rbind(collected_data_reactive(), new_entry)
    collected_data_reactive(collected_data)
    
    # Reset input fields
    updateSelectInput(session, "treatment", selected = "T1")
    updateNumericInput(session, "plot", value = 101)
    updateNumericInput(session, "stand_count", value = 0)
    updateNumericInput(session, "plant_height", value = 0)
    updateNumericInput(session, "severity", value = 0)
  })
  
  observeEvent(input$delete, {
    # Get the selected row(s) and remove them from the collected data
    selected_rows(input$dataTable_rows_selected)
    if (length(selected_rows()) > 0) {
      collected_data <- collected_data_reactive()
      collected_data <- collected_data[-selected_rows(), ]
      collected_data_reactive(collected_data)
    }
  })
  
  output$dataTable <- renderDT({
    datatable(collected_data_reactive(), options = list(select = "multi"))
  })
  
  # Download data as a CSV file
  output$downloadData <- downloadHandler(
    filename = function() {
      "collected_data.csv"
    },
    content = function(file) {
      write.csv(collected_data_reactive(), file)
    }
  )
}

# Run the app
shinyApp(ui, server)

Now that we have successfully developed a shiny app for data collection, we can now proceed to working on another app designed for processing the collected data.

2 Shiny App to process collected data.

We will use shinyjs and ggstatsplot to generate a shiny app that will use collected data to process and generate some preliminary data viz and descriptive stats. For this you will need to use .csv file Aggressiveness_Zone.csv that was sent to your email.

# Load required libraries
library(shiny)
library(ggplot2)
library(dplyr)
library(ggstatsplot)
library(shinyjs)

# Define UI
ui <- fluidPage(
  useShinyjs(),
  titlePanel("Summary Statistics and Plots"),
  
  sidebarLayout(
    sidebarPanel(
      fileInput("file", "Choose CSV File"),
      selectInput("plotType", "Select Plot Type",
                  c("Box-violin Plot", "Box Plot", "Violin Plot")),
      
      # Add options to modify axis label and tick label size
      sliderInput("axisLabelSize", "Axis Label Size", min = 8, max = 20, value = 12),
      sliderInput("axisTickLabelSize", "Axis Tick Label Size", min = 8, max = 20, value = 10)
    ),
    
    mainPanel(
      tabsetPanel(
        tabPanel("Summary Statistics", tableOutput("summaryTable")),
        tabPanel("Plots", plotOutput("plot"))
      )
    )
  )
)

# Define server
server <- function(input, output) {
  data <- reactive({
    inFile <- input$file
    if (is.null(inFile)) return(NULL)
    read.csv(inFile$datapath)
  })
  
  # Define plot_type in the global scope
  plot_type <- reactive({
    if (!is.null(data())) {
      switch(input$plotType,
             "Box-violin Plot" = {
               ggstatsplot::ggbetweenstats(data(), x = AEZ, y = RF, messages = FALSE)
             },
             "Box Plot" = ggplot(data(), aes(x = AEZ, y = RF, fill = AEZ)) +
               geom_boxplot() +
               scale_fill_brewer(palette = "Set1") +
               theme(axis.text.x = element_text(size = input$axisTickLabelSize),
                     axis.text.y = element_text(size = input$axisTickLabelSize),
                     axis.title.x = element_text(size = input$axisLabelSize),
                     axis.title.y = element_text(size = input$axisLabelSize))
             ,
             "Violin Plot" = ggplot(data(), aes(x = AEZ, y = RF, fill = AEZ)) +
               geom_violin() +
               geom_point(aes(x = AEZ, y = RF), position = position_jitterdodge(jitter.width = 0.2, dodge.width = 0.75), size = 2) +
               stat_summary(aes(group = AEZ), fun.data = mean_cl_boot, geom = "crossbar", width = 0.5) +
               scale_fill_brewer(palette = "Set2") +
               theme(axis.text.x = element_text(size = input$axisTickLabelSize),
                     axis.text.y = element_text(size = input$axisTickLabelSize),
                     axis.title.x = element_text(size = input$axisLabelSize),
                     axis.title.y = element_text(size = input$axisLabelSize))
      )
    }
  })
  
  output$summaryTable <- renderTable({
    if (is.null(data())) return(NULL)
    summary_stats <- data() %>%
      group_by(AEZ) %>%
      summarize(
        Mean = mean(RF),
        Standard_Deviation = sd(RF),
        Standard_Error = sd(RF) / sqrt(n())
      )
    as.data.frame(summary_stats)
  })
  
  output$plot <- renderPlot({
    plot_type()
  })
}

# Run the application
shinyApp(ui, server)