R语言绘图，大组中有小组, 标组间差异

R语言绘图，大组中有小组

含Error bar

组件比较线段 星号

绘图效果如下：

image.png

参考数据如下：
aqy.csv

Treatment,group, dose,result
A,n,0,20413
A,n,0,15551
A,n,0,22592
A,n,2,6512
A,n,2,7082
A,n,2,5761
B,n,4,3304
B,n,4,4270
B,n,4,3432
B,n,6,1004
B,n,6,1281
B,n,6,1121
C,y,0,194638
C,y,0,187731
C,y,0,186185
C,y,2,193504
C,y,2,157346
C,y,2,192429
D,y,4,156758
D,y,4,124163
D,y,4,145171
D,y,6,87017
D,y,6,80206
D,y,6,67719

代码如下:

library(ggplot2)
library(ggpubr)

setwd("~/hss/aqy/")
data <- read.csv("aqy.csv")


ggplot(data, aes(x = factor(dose), y = result, fill = group)) +
  geom_bar(stat = "summary", fun = "mean", position = position_dodge(1),
           #width = 1.5,                     
           alpha = 0.3) + 
  # 误差线
  geom_errorbar(stat = "summary", fun.data = "mean_se", 
                position = position_dodge(width = 1), width = 0.2) +
  
  # dose 0 的比较
  geom_segment(x = 1-0.3, xend = 1+0.3,
               y = max(data$result) * 1.1, yend = max(data$result) * 1.1,
               size = 0.5) +
  geom_segment(x = 1-0.3, xend = 1-0.3,
               y = max(data$result) * 1.08, yend = max(data$result) * 1.1,
               size = 0.5) +
  geom_segment(x = 1+0.3, xend = 1+0.3,
               y = max(data$result) * 1.08, yend = max(data$result) * 1.1,
               size = 0.5) +
  stat_compare_means(aes(group = group), 
                     method = "t.test",
                     label = "p.signif",
                     label.y = max(data$result) * 1.15,
                     size = 10,
                     data = subset(data, dose == 0)) +
  
  # dose 2 的比较
  geom_segment(x = 2-0.3, xend = 2+0.3,
               y = max(data$result) * 1.0, yend = max(data$result) * 1.0,
               size = 0.5) +
  geom_segment(x = 2-0.3, xend = 2-0.3,
               y = max(data$result) * 0.98, yend = max(data$result) * 1.0,
               size = 0.5) +
  geom_segment(x = 2+0.3, xend = 2+0.3,
               y = max(data$result) * 0.98, yend = max(data$result) * 1.0,
               size = 0.5) +
  stat_compare_means(aes(group = group), 
                     method = "t.test",
                     label = "p.signif",
                     label.y = max(data$result) * 1.05,
                     size = 10,
                     data = subset(data, dose == 2)) +
  
  # dose 4 的比较
  geom_segment(x = 3-0.3, xend = 3+0.3,
               y = max(data$result) * 0.9, yend = max(data$result) * 0.9,
               size = 0.5) +
  geom_segment(x = 3-0.3, xend = 3-0.3,
               y = max(data$result) * 0.88, yend = max(data$result) * 0.9,
               size = 0.5) +
  geom_segment(x = 3+0.3, xend = 3+0.3,
               y = max(data$result) * 0.88, yend = max(data$result) * 0.9,
               size = 0.5) +
  stat_compare_means(aes(group = group), 
                     method = "t.test",
                     label = "p.signif",
                     label.y = max(data$result) * 0.95,
                     size = 10,
                     data = subset(data, dose == 4)) +
  
  # dose 6 的比较
  geom_segment(x = 4-0.3, xend = 4+0.3,
               y = max(data$result) * 0.8, yend = max(data$result) * 0.8,
               size = 0.5) +
  geom_segment(x = 4-0.3, xend = 4-0.3,
               y = max(data$result) * 0.78, yend = max(data$result) * 0.8,
               size = 0.5) +
  geom_segment(x = 4+0.3, xend = 4+0.3,
               y = max(data$result) * 0.78, yend = max(data$result) * 0.8,
               size = 0.5) +
  stat_compare_means(aes(group = group), 
                     method = "t.test",
                     label = "p.signif",
                     label.y = max(data$result) * 0.85,
                     size = 10,
                     data = subset(data, dose == 6)) +
  
  # 添加 dose 0 和 2 的 n 组之间的比较
  geom_segment(x = 1-0.3, xend = 2-0.3,
               y = max(data$result) * 0.7, yend = max(data$result) * 0.7,
               size = 0.5) +
  geom_segment(x = 1-0.3, xend = 1-0.3,
               y = max(data$result) * 0.68, yend = max(data$result) * 0.7,
               size = 0.5) +
  geom_segment(x = 2-0.3, xend = 2-0.3,
               y = max(data$result) * 0.68, yend = max(data$result) * 0.7,
               size = 0.5) +
  stat_compare_means(data = subset(data, group == "n" & dose %in% c(0, 2)),
                     aes(group = factor(dose)),
                     method = "t.test",
                     label = "p.signif",
                     label.y = max(data$result) * 0.75,
                     size = 10) +
  
  labs(x = "Dose", 
       y = "Result",
       fill = "Group") +
  theme_minimal() +
  theme(
    panel.grid = element_blank(),
    axis.line = element_line(color = "black"),
    text = element_text(color = "black"),
    axis.text = element_text(color = "black", size = 20),
    axis.title = element_text(color = "black", size = 26),
    legend.text = element_text(color = "black", size = 20), 
    legend.title = element_text(color = "black", size = 22)    
  ) +
  scale_fill_manual(values = c("n" = "#644296", "y" = "#D1352C")) +
  scale_y_continuous(expand = c(0, 0), 
                     limits = c(0, max(data$result) * 1.3))