I recently developed an R Shiny application to filter and group data in rshiny. owever, I’ve encountered a specific issue that I haven’t been able to resolve. The problem arises when I attempt to group my dataframe based on the selection from a dropdown menu, which is linked to a variable named char_column_selector. When I group the dataframe using this dropdown selection, the lines in my plot disappear, making it difficult to visualize the data effectively.

library(shiny)
library(shinyWidgets)
library(dplyr)
library(plotly)

ui <- fluidPage(
  sidebarLayout(
    sidebarPanel(
      selectInput("char_column_selector", "Select char_ Column", choices = NULL, selected = ""),
      uiOutput("quantiles_input"),
      uiOutput("value_bins_input"),
      uiOutput("dynamic_inputs")
    ),
    mainPanel(
      h3("Dataset Overview"),
      tableOutput("table"),
      plotlyOutput("plot")
    )
  )
)


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

  data_loaded <- reactiveVal(FALSE)
  dataset <- reactiveVal(NULL)
  

  observe({
    
    data <- tibble(
      date = sample(seq.Date(as.Date('2023-01-01'), as.Date('2024-01-01'), by="day"), 100, replace = TRUE),
      id = 1:100,
      principal = round(runif(100, 1000, 10000), 2),
      amount = round(runif(100, 500, 5000), 2),
      char_age = sample(18:70, 100, replace = TRUE),
      char_gender = sample(c("Male", "Female", "Other"), 100, replace = TRUE),
      char_region = sample(c("North", "South", "East", "West"), 100, replace = TRUE),
      char_score = sample(300:850, 100, replace = TRUE)
    )
    dataset(data)
    data_loaded(TRUE)
  })
  
  
  char_columns <- reactive({
    req(data_loaded())
    names(dataset())[grepl("^char_", names(dataset()))]
  })
  
  
  observe({
    updateSelectInput(session, "char_column_selector", choices = c("", char_columns()), selected = "")
  })
  
  
  output$dynamic_inputs <- renderUI({
    req(data_loaded())
    inputs <- lapply(char_columns(), function(col) {
      if (is.numeric(dataset()[[col]])) {
        numericRangeInput(inputId = col, label = col, 
                          value = c(min(dataset()[[col]], na.rm = TRUE), max(dataset()[[col]], na.rm = TRUE)), 
                          min = min(dataset()[[col]], na.rm = TRUE), 
                          max = max(dataset()[[col]], na.rm = TRUE))
      } else {
        selectInput(inputId = col, label = col, 
                    choices = unique(dataset()[[col]]), 
                    selected = unique(dataset()[[col]]), 
                    multiple = TRUE)
      }
    })
    do.call(tagList, inputs)
  })
  
  
  filtered_data <- reactive({
    req(data_loaded())
    data <- dataset()
    
    
    for (col in char_columns()) {
      if (is.numeric(data[[col]])) {
        range_vals <- input[[col]]
        if (!is.null(range_vals)) {
          data <- data %>% filter(data[[col]] >= range_vals[1] & data[[col]] <= range_vals[2])
        }
      } else {
        selected_vals <- input[[col]]
        if (!is.null(selected_vals) && length(selected_vals) > 0) {
          data <- data %>% filter(data[[col]] %in% selected_vals)
        }
      }
    }
    data
  })
  
  value_bins <- reactive({
    data <- filtered_data()
    quantiles <- quantile(data$principal, probs = seq(0, 1, by = 0.25), na.rm = TRUE)
    quantiles
  })
  
  output$value_bins_input <- renderUI({
    quantiles_str <- paste(round(value_bins(), 2), collapse = ", ")
    textInput("value_bins_text", label = "Quantiles of Principal", value = quantiles_str)
  })
  
  output$quantiles_input <- renderUI({
    req(input$char_column_selector)
    if (input$char_column_selector != "") {
      selected_col <- filtered_data()[[input$char_column_selector]]
      
      if (is.numeric(selected_col)) {
        quantiles <- quantile(selected_col, probs = seq(0, 1, by = 0.25), na.rm = TRUE)
        quantiles_str <- paste(round(quantiles, 2), collapse = ", ")
        textInput("quantiles_text", label = paste("Quantiles of", input$char_column_selector), value = quantiles_str)
      }
    }
  })
  
  grouped_data <- reactive({
    data <- filtered_data()
    if (input$char_column_selector == "") {
      data %>% 
        group_by(date) %>% 
        summarise(principal_mean = mean(principal, na.rm = TRUE)) %>%
        mutate(date = as.Date(date)) 
    } else {
      selected_col <- input$char_column_selector
      if (is.numeric(data[[selected_col]])) {
        req(input$quantiles_text)
        quantiles <- as.numeric(strsplit(input$quantiles_text, ",\s*")[[1]])
        data <- data %>% mutate(char = cut(data[[selected_col]], breaks = quantiles, include.lowest = TRUE))
        data %>% 
          group_by(date, char) %>% 
          summarise(principal_mean = mean(principal, na.rm = TRUE)) %>%
          arrange(date) %>%
          mutate(date = as.Date(date))
      } else {
        data %>% 
          group_by(date, char = .data[[selected_col]]) %>% 
          summarise(principal_mean = mean(principal, na.rm = TRUE)) %>%
          arrange(date) %>%
          mutate(date = as.Date(date))
      }
    }
  })
  
  
  output$table <- renderTable({
    req(data_loaded())
    grouped_data()
  })
  
  output$plot <- renderPlotly({
    req(data_loaded())
    data <- grouped_data()
    
    if (input$char_column_selector == "") {
      plot_ly(data, x = ~date, y = ~principal_mean, type = 'scatter', mode = 'lines+markers')
    } else {
      plot_ly(data, x = ~date, y = ~principal_mean, type = 'scatter', mode = 'lines+markers', color = ~char) 
    }
  })
}

shinyApp(ui = ui, server = server)