如何为ggplotly双数据框绘图的均值点自定义Tooltip

问题背景 #

已基于Drug分组生成ggplot图表，创建了汇总主数据均值的mean_df数据框，将各组均值以三角形点添加至图中，并用stat_summary连接均值线。当前仅主数据的Tooltip正常显示，需为均值点添加包含自定义标签的x、y值Tooltip。

原始数据 #

Day,Drug,Sex,Y
1,A,Female,2.192306074
1,B,Male,4.551912798
1,B,Female,1.574070652
1,C,Female,-0.143946163
1,A,Male,5.144422967
1,C,Male,5.724705829
2,A,Male,2.691617258
2,B,Female,-3.0289955
2,C,Male,0.338102762
2,A,Female,-0.558581233
2,B,Female,-2.942620032
2,C,Male,1.024670497
3,A,Male,2.264980803
3,C,Female,2.103722883
3,A,Female,2.091621938
3,B,Male,1.535299922
3,B,Male,1.618399767
3,C,Female,0.136160703

数据转换与变量定义 #

# 读取数据
df <- read.delim("clipboard", sep = ",")

# 变量定义
grouping_var <- "Drug"
x <- "Day"
y <- "Y"
color <- "Sex"

原有绘图代码 #

text <- paste("X:", df[[grouping_var]], "<br>",
              "Y:", df[[y]], "<br>",
              "group:", grouping_var, "<br>",
              "color:", color)
mean_df <- df %>% group_by(get(x), get(grouping_var)) %>% summarise(Mean = mean(get(y)), .groups = "drop") %>% 
  rename(x = 1, grouping_var = 2)

p <- df %>% ggplot(aes(x = get(x), y = get(y), text = text)) +
  geom_point(aes_string(group = grouping_var, color = color)) +
  stat_summary(fun=mean, geom="line", aes_string(group = grouping_var), color = "red", lty = 2) +
  geom_point(data = mean_df, aes(x = x, y = Mean, group = grouping_var, color = grouping_var),
             pch = 2, size = 3,inherit.aes = F) + 

  xlab(x)+
  ylab(y)

ggplotly(p, tooltip = c("text"))

解决方案 #

要让均值点显示自定义Tooltip，需为mean_df创建专属的tooltip文本，并在绘制均值点时映射text美学属性，同时确保ggplotly能识别两个图层的tooltip内容。修改后的代码如下：

# 主数据的tooltip文本
text <- paste("X:", df[[grouping_var]], "<br>",
              "Y:", round(df[[y]], 3), "<br>",
              "分组:", grouping_var, "<br>",
              "性别:", color)

# 生成均值数据框并添加tooltip文本
mean_df <- df %>% 
  group_by(!!sym(x), !!sym(grouping_var)) %>% 
  summarise(Mean = mean(get(y)), .groups = "drop") %>% 
  rename(x = !!sym(x), grouping_var = !!sym(grouping_var)) %>%
  # 为均值点创建自定义tooltip
  mutate(mean_text = paste("Day:", x, "<br>",
                          "均值Y:", round(Mean, 3), "<br>",
                          "Drug分组:", grouping_var))

p <- df %>% ggplot(aes(x = get(x), y = get(y), text = text)) +
  geom_point(aes_string(group = grouping_var, color = color)) +
  stat_summary(fun=mean, geom="line", aes_string(group = grouping_var), color = "red", lty = 2) +
  # 绘制均值点时映射text属性
  geom_point(data = mean_df, aes(x = x, y = Mean, group = grouping_var, color = grouping_var, text = mean_text),
             pch = 2, size = 3, inherit.aes = F) + 

  xlab(x)+
  ylab(y)

# ggplotly自动识别两个图层的tooltip
ggplotly(p)

关键修改点 #

• 为mean_df添加mean_text字段，定义均值点的tooltip内容，包含Day、均值Y和Drug分组信息
• 在绘制均值点的geom_point中，新增text = mean_text的美学映射
• 移除ggplotly中指定的tooltip = c("text")，让其自动读取各图层的text属性
• 使用!!sym()替代get()和重命名操作，让代码更规范且避免潜在的命名冲突

内容的提问来源于stack exchange，提问作者Abbas