R语言ggplot2绘制分组条形图并同时展示每组计数与占比

1. 依赖包加载 #

需要提前安装加载以下三个包：

library(ggplot2)
library(dplyr)
library(scales)

2. 数据预处理 #

首先修正数据集构造方式避免数据类型错误，同时计算每个Variable组内的占比：

# 构造数据集
Variable <- c("ABC","ABC","MNO","MNO","XYZ","XYZ")
Flag <- c("Y","N","Y","N","Y","N")
Count <- c(1234,9876,6789,3210,5500,7890)
df <- data.frame(Variable, Flag, Count)

# 按Variable分组计算组内各Flag的占比
df <- df %>% 
  group_by(Variable) %>% 
  mutate(percentage = Count/sum(Count))

3. 合并标注绘图代码 #

直接在geom_text的label参数中拼接计数和百分比即可，注意保持分组偏移宽度一致避免标签错位：

ggplot(df, aes(x = Variable, y = Count, fill = Flag)) +
  # 生成分组条形图
  geom_bar(stat = "identity", position = position_dodge(width = 0.9)) +
  # 叠加计数+百分比标注
  geom_text(aes(label = paste0(Count, "\n", percent(percentage, accuracy = 0.1))),
            position = position_dodge(width = 0.9), # 和条形偏移宽度保持一致
            colour = "white", size = 3.5, vjust = 1.2) + # 调整标签样式和位置
  labs(x = "变量", y = "计数", fill = "标识") +
  theme_bw()

参数调整说明 #

• 可修改accuracy参数调整百分比保留的小数位数，比如accuracy=1就是保留整数百分比
• 调整vjust参数可以上下移动标签位置，调大数值标签会向下移动，调小会向上移动
• 修改position_dodge(width)可以调整同组两个条形的间距，需保证条形和文本的该参数一致

内容的提问来源于stack exchange，提问作者Prabhu