ggplot分组柱状图添加指定组间显著性星号标记实现方法

geom_signif和ggpubr的显著性标注函数默认适配单分组x轴结构，你的图表是双分组dodge柱状图：x轴主分类为年龄组，柱体按奖励条件做位置偏移，函数无法自动识别偏移后的柱体坐标，因此标注错位或无法正常渲染。

直接手动指定dodge后每个柱体的实际x坐标，用geom_signif的手动标注模式添加标记即可，不需要做复杂的统计映射：

• 两个年龄组的x轴基准位置：Young Adult在x=1，Older Adult在x=2
• 你设置的position_dodge宽度为0.6，4个奖励条件柱体在每个年龄组下从左到右的偏移量为-0.225、-0.075、0.075、0.225，对应实际x坐标可直接计算得到
• 仅给Young Adult组Self High Value和Self Low Value的比较对标注**，其余相邻柱体比较统一标注ns

library(tidyverse)
library(ggsignif)

viz_data_one <- tibble(
  age_group = c(rep("Young Adult", 4), rep("Older Adult", 4)),
  MemoryAccuracy = c(32.8, 28.448, 27.672, 27.075, 29.667, 28.944, 27.556, 28.889),
  upper = MemoryAccuracy + 1.76,
  lower = MemoryAccuracy - 1.76,
  reward = rep(c("Self High Value", "Self Low Value", "Other High Value", "Other Low Value"), 2)
) %>% 
  mutate(
    reward = as_factor(reward) %>% fct_relevel("Self High Value",
                                               "Self Low Value",
                                               "Other High Value",
                                               "Other Low Value"))
viz_data_one <- viz_data_one %>%
  mutate(upper = MemoryAccuracy + ifelse(age_group == "Young Adult", 1.76, 1.94),
         lower = MemoryAccuracy - ifelse(age_group == "Young Adult", 1.76, 1.94))
      
p <- viz_data_one %>% 
  ggplot(aes(x = age_group,
             y = MemoryAccuracy,
             fill = reward,
             ymin = lower,
             ymax = upper)) +
  geom_col(width = .5, position = position_dodge(.6),
           color = "black", key_glyph = "polygon") +
  geom_errorbar(width = .1, position = position_dodge(.6)) +
  # 显著性标注层
  geom_signif(
    xmin = c(0.775, 0.925, 1.075, 1.775, 1.925, 2.075),
    xmax = c(0.925, 1.075, 1.225, 1.925, 2.075, 2.225),
    y_position = c(35.5, 34.5, 34.5, 32.5, 31.5, 31.5),
    annotation = c("**", "ns", "ns", "ns", "ns", "ns"),
    tip_length = 0.01,
    size = 0.5,
    textsize = 5,
    manual = TRUE
  ) +
  scale_fill_manual(values = c("#E495A5", "#ABB065", "#39BEB1", "#ACA4E2" )) +
  labs(
    x = "Age Group",
    y = "Memory Accuracy (%)",
    fill = NULL,
    title = ""
  ) +
  theme(
    plot.margin = unit(c(1, 1, 1, 1), "cm"),
    panel.background = element_blank(),
    plot.title = element_text(size = 14, face = "bold",
                              hjust = 0.5,
                              margin = margin(b = 15)),
    axis.line = element_line(color = "black"),
    axis.title = element_text(size = 14, color = "black",
                              face = "bold"),
    axis.text = element_text(size = 18, color = "black"),
    axis.text.x = element_text(margin = margin(t = 10)),
    axis.text.y = element_text(size = 14),
    axis.title.y = element_text(margin = margin(r = 10)),
    axis.ticks.x = element_blank(),
    legend.position = c(0.90, 0.99),
    legend.background = element_rect(color = "black"),
    legend.text = element_text(size = 15),
    legend.margin = margin(t = 5, l = 5, r = 5, b = 5),
    legend.key = element_rect(color = NA, fill = NA)
  ) +
  guides(
    fill = guide_legend(
      keywidth = .5,
      keyheight = .5,
      default.unit= "cm"
    )
  )

p + expand_limits(y = 38)

• 运行前需先安装依赖：执行install.packages("ggsignif")即可
• 若标注高度不合适，直接修改y_position向量内的数值，数值越大标注位置越靠上
• tip_length控制显著性横线两端的短竖线长度，textsize控制标注文字大小，size控制横线粗细
• 若要增减比较对，同步增删xmin/xmax/y_position/annotation四个参数的元素即可，四个向量长度必须保持一致

内容的提问来源于stack exchange，提问作者Tessa