如何用ggplot2自定义条形图填充比例并添加X轴计数标注？

1. 按指定比例绘制堆叠条形图 #

首先需要把宽格式数据转换为长格式，让ggplot能识别分组填充逻辑。用tidyr::pivot_longer将g1、g2、g3列转为分组和对应比例，再计算每个分组的实际outcome数值（比例×总outcome）：

library(tidyverse)

# 数据格式转换：宽转长 + 计算分组对应结果值
df_long <- df %>%
  pivot_longer(cols = starts_with("g"), 
               names_to = "group", 
               values_to = "proportion") %>%
  mutate(group_outcome = outcome * proportion)

# 绘制堆叠条形图
ggplot(df_long, aes(x = age_group, y = group_outcome, fill = group)) +
  geom_col() +
  labs(y = "outcome", fill = "分组")

geom_col默认使用position = "stack"，会自动按比例堆叠各分组的条形高度，完全匹配g1/g2/g3的比例要求。

2. 在X轴下方添加count值 #

用geom_text在X轴对应位置下方添加count数值，通过设置y轴负坐标让文本显示在条形下方，同时调整轴范围避免文本被截断：

ggplot(df_long, aes(x = age_group, y = group_outcome, fill = group)) +
  geom_col() +
  # 添加count文本到X轴下方
  geom_text(data = df, aes(x = age_group, y = -5, label = paste("count:", count)), 
            color = "dimgray", size = 3.5) +
  # 扩展Y轴范围，避免文本被截断
  expand_limits(y = -10) +
  labs(x = "年龄组", y = "outcome", fill = "分组") +
  # 调整X轴标签与文本的间距
  theme(axis.text.x = element_text(margin = margin(t = 8)))

完整整合代码 #

library(tidyverse)

df <- tribble(
  ~age_group, ~count, ~outcome, ~g1, ~g2, ~g3,
  "first", 10, 110, 0.3, 0.6, 0.1,
  "second", 20, 70, 0.4, 0.3, 0.3,
  "third", 80, 35, 0.25, 0.5, 0.25
)

df_long <- df %>%
  pivot_longer(cols = starts_with("g"), 
               names_to = "group", 
               values_to = "proportion") %>%
  mutate(group_outcome = outcome * proportion)

ggplot(df_long, aes(x = age_group, y = group_outcome, fill = group)) +
  geom_col() +
  geom_text(data = df, aes(x = age_group, y = -5, label = paste("count:", count)),
            color = "dimgray", size = 3.5) +
  expand_limits(y = -10) +
  labs(x = "年龄组", y = "结果值", fill = "分组") +
  theme(axis.text.x = element_text(margin = margin(t = 8)))

最终图表会呈现：每个年龄组的条形按g1/g2/g3比例堆叠填充，X轴下方对应显示该组的count数值。

内容的提问来源于stack exchange，提问作者dcsuka