如何在ggplot中标记数据异常值并识别对应站点

Hey there! Let's tackle how to mark outliers in your ggplot boxplot and identify which sites they belong to. Here's a step-by-step solution tailored to your existing code:

第一步：先识别数据中的异常值 #

First, we need to calculate the outlier thresholds for each environmental parameter (your variable column) and filter out the rows that qualify as outliers. 异常值的定义是超出Q1-1.5×IQR或Q3+1.5×IQR的数值，这也是ggplot箱线图默认的异常值判定规则。

# 加载dplyr包（如果没装先运行install.packages("dplyr")）
library(dplyr)

# 计算每个变量的四分位数、IQR和异常值阈值
outlier_thresholds <- df %>%
  group_by(variable) %>%
  summarise(
    Q1 = quantile(value, 0.25, na.rm = TRUE),
    Q3 = quantile(value, 0.75, na.rm = TRUE),
    IQR = Q3 - Q1,
    lower_bound = Q1 - 1.5 * IQR,
    upper_bound = Q3 + 1.5 * IQR
  )

# 合并阈值到原数据，筛选出异常值（假设你的站点列名为Site，记得改成实际列名）
outliers <- df %>%
  left_join(outlier_thresholds, by = "variable") %>%
  filter(value < lower_bound | value > upper_bound)

第二步：绘制带异常值标记和站点标签的箱线图 #

Now we'll update your ggplot code to highlight these outliers and label their corresponding sites. You can choose to keep the jitter or not—either way, the explicit outlier markers will stand out clearly.

不带jitter的版本 #

library(ggplot2)

ggplot(data = df, aes(x = variable, y = value, fill = variable)) +
  geom_boxplot() +
  # 用红色大点标记异常值
  geom_point(data = outliers, aes(color = "Outlier"), size = 3) +
  # 添加站点标签，放在点的上方
  geom_text(data = outliers, aes(label = Site),
            vjust = -1, hjust = 0.5, size = 3, family = "Times New Roman") +
  theme_bw() +
  labs(x = "Environmental Parameters", y = "Standardized Range", color = "") +
  theme(
    legend.position = "none",
    text = element_text(family = "Times New Roman", face = "bold", size = 12)
  )

带jitter的版本 #

ggplot(data = df, aes(x = variable, y = value, fill = variable)) +
  geom_boxplot() +
  geom_jitter(position = position_jitter(0.1), alpha = 0.6) + # 降低jitter点透明度避免遮挡
  geom_point(data = outliers, aes(color = "Outlier"), size = 3) +
  geom_text(data = outliers, aes(label = Site),
            vjust = -1, hjust = 0.5, size = 3, family = "Times New Roman") +
  theme_bw() +
  labs(x = "Environmental Parameters", y = "Standardized Range", color = "") +
  theme(
    legend.position = "none",
    text = element_text(family = "Times New Roman", face = "bold", size = 12)
  )

小提示 #

• 如果你的站点列名不是Site，一定要把代码里的Site改成你实际的列名（比如station_id之类的）。
• 调整geom_text的vjust和hjust参数可以改变标签的位置，避免重叠。
• 你也可以用geom_label代替geom_text，给标签加个背景框，更易读。

内容的提问来源于stack exchange，提问作者Joerick Paderogao