app.R

# install.packages(c("shiny", "dplyr", "ggplot2", "DT"))

library(shiny)
library(dplyr)
library(ggplot2)
library(data.table)

ui <- fluidPage(
  titlePanel("Country Representation Rankings Explorer"),

  sidebarLayout(
    sidebarPanel(
      h4("Controls"),
      selectInput(
        inputId = "metric",
        label = "Select a Representation Metric:",
        choices = c(
          "Overall", 
          "Representation Gap", 
          "Ethnicity", 
          "Gender", 
          "Religion", 
          "Language"
        ),
        selected = "Overall"
      ),
      checkboxInput(
        inputId = "sort_desc",
        label = "Sort in descending order",
        value = TRUE
      ),
      hr(),
      sliderInput(
        inputId = "top_n",
        label = "Number of Countries to Display in Plot:",
        min = 5,
        max = 50,
        value = 10,
        step = 1
      )
    ),

    mainPanel(
      tabsetPanel(
        tabPanel(
          title = "Plot",
          plotOutput("bar_plot", height = "600px")
        ),
        tabPanel(
          title = "Data Table",
          DTOutput("data_table")
        ),
        tabPanel(
          title = "Summary",
          verbatimTextOutput("summary_text")
        )
      )
    )
  )
)

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

  # Load the data (adjust the path as necessary)
  data_raw <- reactive({
    read.csv("CountryRepresentationRankings.csv", stringsAsFactors = FALSE)
  })
  
  # Reactively prepare the data for plotting and table display
  data_prepared <- reactive({
    req(data_raw())
    # Convert to tibble for convenience
    df <- tibble::as_tibble(data_raw())
    
    # Ensure the selected metric is numeric for plotting
    # (This is typically already numeric, but we can do a safe check.)
    # Also handle potential issues with missing values.
    df <- df %>%
      mutate(
        across(all_of(input$metric), as.numeric)
      ) %>%
      filter(!is.na(.data[[input$metric]]))
    
    # Sort ascending or descending
    if (input$sort_desc) {
      df <- df %>% arrange(desc(.data[[input$metric]]))
    } else {
      df <- df %>% arrange(.data[[input$metric]])
    }
    
    # Keep only the top N rows for plotting
    df <- df %>% slice_head(n = input$top_n)
    
    df
  })

  # Render bar plot of the selected metric for top N countries
  output$bar_plot <- renderPlot({
    df <- data_prepared()
    
    # Convert Country to factor for a nice ordered plot
    df <- df %>%
      mutate(Country = factor(Country, levels = Country))
    
    ggplot(df, aes(x = Country, y = .data[[input$metric]])) +
      geom_col(fill = "steelblue") +
      coord_flip() +
      labs(
        x = "Country",
        y = paste0(input$metric, " Index"),
        title = paste("Top", input$top_n, "Countries by", input$metric)
      ) +
      theme_minimal(base_size = 14)
  })
  
  # Render data table of the entire dataset
  output$data_table <- renderDT({
    req(data_raw())
    datatable(
      data_raw(),
      options = list(pageLength = 10, autoWidth = TRUE),
      filter = "top",
      rownames = FALSE
    )
  })

  # Render summary statistics text for the selected metric
  output$summary_text <- renderPrint({
    df <- data_raw()
    req(df)
    # Convert the chosen metric to numeric if needed
    df[[input$metric]] <- suppressWarnings(as.numeric(df[[input$metric]]))
    
    valid_metric <- df[[input$metric]][!is.na(df[[input$metric]])]
    if (length(valid_metric) == 0) {
      cat("No valid numeric data for the selected metric.")
    } else {
      summary_stats <- summary(valid_metric)
      cat("Summary of", input$metric, "across all countries:\n")
      print(summary_stats)
    }
  })
}

shinyApp(ui = ui, server = server)