在R ggplot中为X轴（因子水平）添加间隔及替换统计星号为柱状标注

1. 修复X轴刻度问题 #

报错原因是hrs为因子类型，而scale_x_continuous仅适用于连续数值变量，两者不兼容。提供两种解决方法：

方法一：将hrs转换为数值型变量 #

先把数据中的hrs从因子转为数值，再用scale_x_continuous设置刻度：

# 取消分组并转换hrs类型
df02_EC_01 <- df02_EC_01 %>% 
  ungroup() %>% 
  mutate(hrs = as.numeric(as.character(hrs)))

# 重新绘图
plot01 <- ggplot(df02_EC_01, aes(x=hrs, y=Vol_mm3, fill=Temp, color=Temp)) +
  geom_dotplot(binaxis = "y", 
               stackdir = "centerwhole", 
               alpha = .5, 
               dotsize = .3, 
               method = "histodot", 
               binwidth = 1, 
               stackratio = 1, 
               position = "dodge",
               binpositions="all") +
  geom_boxplot(varwidth = T,  size= .5, coef=500, alpha=.5) +
  labs(title = "Decalcification Process with EDTA",
       x = "Hours",
       y = "Mineralized tissue (mm3)") +
  theme(plot.title = element_text(hjust = 0.5)) +
  stat_compare_means(aes(group=Temp), label = "p.signif") + 
  scale_x_continuous(breaks = seq(0, 72, by = 12), 
                     labels = seq(0, 72, by = 12),
                     limits = c(0,72)) + # 可选：限制X轴范围到72
  ylim(0,30)

方法二：保留因子类型，用scale_x_discrete调整刻度 #

若需保留hrs的因子属性，可指定要显示的刻度标签：

plot01 <- ggplot(df02_EC_01, aes(x=hrs, y=Vol_mm3, fill=Temp, color=Temp)) +
  geom_dotplot(binaxis = "y", 
               stackdir = "centerwhole", 
               alpha = .5, 
               dotsize = .3, 
               method = "histodot", 
               binwidth = 1, 
               stackratio = 1, 
               position = "dodge",
               binpositions="all") +
  geom_boxplot(varwidth = T,  size= .5, coef=500, alpha=.5) +
  labs(title = "Decalcification Process with EDTA",
       x = "Hours",
       y = "Mineralized tissue (mm3)") +
  theme(plot.title = element_text(hjust = 0.5)) +
  stat_compare_means(aes(group=Temp), label = "p.signif") + 
  scale_x_discrete(breaks = c("0","12","24","48","72"), # 指定显示的因子水平
                   labels = c("0","12","24","48","72")) +
  ylim(0,30)

2. 替换显著性星号为柱状连线标注 #

使用ggsignif包的geom_signif生成带连线的显著性标注，替代stat_compare_means的星号：

步骤1：安装并加载ggsignif #

install.packages("ggsignif")
library(ggsignif)

步骤2：修改绘图代码 #

用geom_signif替换stat_compare_means，指定比较组与标注位置：

# 先确保hrs为数值型（参考方法一的转换）
df02_EC_01 <- df02_EC_01 %>% 
  ungroup() %>% 
  mutate(hrs = as.numeric(as.character(hrs)))

plot01 <- ggplot(df02_EC_01, aes(x=hrs, y=Vol_mm3, fill=Temp, color=Temp)) +
  geom_dotplot(binaxis = "y", 
               stackdir = "centerwhole", 
               alpha = .5, 
               dotsize = .3, 
               method = "histodot", 
               binwidth = 1, 
               stackratio = 1, 
               position = "dodge",
               binpositions="all") +
  geom_boxplot(varwidth = T,  size= .5, coef=500, alpha=.5) +
  labs(title = "Decalcification Process with EDTA",
       x = "Hours",
       y = "Mineralized tissue (mm3)") +
  theme(plot.title = element_text(hjust = 0.5)) +
  # 生成柱状显著性标注
  geom_signif(comparisons = list(c("RT", "37C")), 
              y_position = c(28,26,24,22,20), # 每个时间点的标注高度，可按需调整
              x_position = c(0,12,24,48,72), # 对应时间点的X轴位置
              map_signif_level = TRUE, # 显示显著性等级（如***、**）
              tip_length = 0.01) +
  scale_x_continuous(breaks = seq(0, 72, by = 12), 
                     labels = seq(0, 72, by = 12)) +
  ylim(0,30)

若需显示具体p值而非等级，可调整参数：

geom_signif(comparisons = list(c("RT", "37C")), 
            y_position = c(28,26,24,22,20),
            x_position = c(0,12,24,48,72),
            map_signif_level = FALSE,
            test = "t.test", # 根据你的统计方法调整
            tip_length = 0.01)

内容的提问来源于stack exchange，提问作者Marcello