ggplot绘制带抖动个体点的多组均值连线散点图及误差条问题

核心思路 #

1. 用position_jitter替代手动计算抖动，避免x轴异常，同时高效实现个体点的分散展示
2. 先分组计算每个condition+group的均值和标准误
3. 给每个group在对应condition的x轴位置设置固定偏移，实现同condition内两组均值的连线

完整代码 #

library(tidyverse)

# 原始数据
df <- data.frame(   
    subject = c(1, 1, 1, 1, 2, 2, 2, 2, 3, 3, 3, 3),   
    condition = c("A", "A", "B", "B", "A", "A", "B", "B", "A", "A", "B", "B"),   
    group = c("group1", "group2", "group1", "group2",                
              "group1", "group2", "group1", "group2",                
              "group1", "group2", "group1", "group2"),   
    value = c(10, 7, 8, 9, 11, 12, 13, 2, 2, 3, 4, 5)
)

# 计算均值和标准误，同时设置每个group在condition内的x轴偏移
summary_df <- df %>%
  group_by(condition, group) %>%
  summarise(
    mean_val = mean(value),
    se_val = sd(value)/sqrt(n()),
    # 给group1和group2分别设置左右偏移，保证在同一condition内
    x_pos = as.numeric(factor(condition)) + ifelse(group == "group1", -0.2, 0.2),
    .groups = "drop"
  )

# 绘图
ggplot() +
  # 抖动后的个体数据点
  geom_point(data = df, 
             aes(x = condition, y = value, color = group),
             position = position_jitter(width = 0.3), # 控制抖动宽度，避免重叠
             alpha = 0.6) +
  # 每个condition内连接两组均值的线
  geom_line(data = summary_df,
            aes(x = x_pos, y = mean_val, group = condition),
            color = "black", linewidth = 0.8) +
  # 均值点
  geom_point(data = summary_df,
             aes(x = x_pos, y = mean_val, color = group),
             size = 3, shape = 18) +
  # 标准误误差条
  geom_errorbar(data = summary_df,
                aes(x = x_pos, ymin = mean_val - se_val, ymax = mean_val + se_val, color = group),
                width = 0.1) +
  # 重置x轴刻度，对应原始condition标签
  scale_x_continuous(breaks = 1:2, labels = c("A", "B")) +
  # 标签设置
  labs(x = "条件", y = "数值", color = "组别") +
  theme_minimal()

代码说明 #

• position_jitter(width=0.3)：自动对个体点进行横向抖动，无需手动计算坐标，解决x轴异常问题
• x_pos：给每个group在对应condition的x轴位置设置±0.2的偏移，确保同condition内两个group的均值点可以用线连接
• geom_line(group=condition)：指定按condition分组连线，实现每个condition内两组均值的连接
• 误差条和均值点使用计算好的偏移位置，和个体点的抖动范围对齐，保证图面整洁

内容的提问来源于stack exchange，提问作者Ryan Panela