如何在Shiny中通过pickerInput添加对应颜色的多组geom_hline

实现步骤 #

1. 预先计算每个类别的对照组均值，生成独立数据集
2. 在Server端根据用户选中的类别，动态筛选出对应的均值数据
3. 在ggplot中添加geom_hline，映射颜色与条形一致，设置为虚线样式

修改后的完整代码 #

library(shiny)
library(ggplot2)
library(dplyr)
library(shinyWidgets)

# 示例数据（匹配你提供的dput结构）
df <- structure(list(
  category = rep(c("A", "B", "C", "D"), each = 4),
  group = rep(c("对照组", "实验组1", "实验组2", "实验组3"), 4),
  value = c(10, 12, 15, 13, 8, 9, 11, 10, 15, 17, 20, 18, 7, 8, 9, 8)
), class = "data.frame", row.names = c(NA, -16L))

# 预计算每个类别的对照组均值
control_means <- df %>%
  filter(group == "对照组") %>%
  group_by(category) %>%
  summarise(mean_val = mean(value, na.rm = TRUE), .groups = "drop")

ui <- fluidPage(
  pickerInput("category_picker", "选择类别",
              choices = unique(df$category),
              multiple = TRUE,
              selected = unique(df$category)[1]),
  plotOutput("bar_plot")
)

server <- function(input, output) {
  # 动态筛选选中类别的对照组均值
  selected_means <- reactive({
    control_means %>%
      filter(category %in% input$category_picker)
  })
  
  output$bar_plot <- renderPlot({
    p <- df %>%
      filter(category %in% input$category_picker) %>%
      ggplot(aes(x = group, y = value, fill = category)) +
      geom_bar(stat = "identity", position = "dodge")
    
    # 仅当有选中类别时添加水平线
    if(nrow(selected_means()) > 0) {
      p <- p + geom_hline(
        data = selected_means(),
        aes(yintercept = mean_val, color = category),
        linetype = "dashed",
        size = 1
      ) +
      # 保持线条颜色与条形填充色一致
      scale_color_manual(values = scales::hue_pal()(length(unique(df$category))))
    }
    
    p
  })
}

shinyApp(ui, server)

关键代码说明 #

• 预计算对照组均值：用dplyr筛选对照组数据，按类别分组计算均值，避免重复计算提升性能
• 动态筛选选中项：通过reactive函数根据用户选择的类别，从预计算的均值数据集中筛选对应条目
• 添加水平线：用geom_hline映射yintercept为均值，color为类别，设置linetype = "dashed"实现虚线效果；通过scale_color_manual让线条颜色和条形填充色保持统一
• 条件显示：判断selected_means()的行数，确保只有选中类别时才加载线条

内容的提问来源于stack exchange，提问作者hks