如何在ggplot2分组多值柱状图中添加累积百分比

需求说明 #

现有按Taxa分组、每组包含两个Method观测值的分组柱状图，需要为每个Taxa组添加累积百分比相关可视化（组内占比或组间累积折线），尝试通过过滤数据集并使用geom_line实现但未成功，需明确正确实现方式，包括geom_line中使用filter的方法。

原始代码与数据集 #

原始绘图代码 #

df %>% mutate(Taxa = fct_relevel(Taxa, 
            "Cnidaria ", "Echinodermata", "Mollusca", 
            "Bryozoa", "Porifera", "Annelida", 
            "Foraminifera", "Platyhelminthes", "Brachiopoda")) %>% ggplot(aes(x=Taxa,y=Observations,fill = Method)) +
  geom_bar(stat = "identity", width = 0.8 ,size = 1, position=position_dodge(), alpha = 0.9) +
  geom_text(aes(label=Observations), vjust=1.6, color="white", size=4, position=position_dodge(0.9)) +
  theme_bw() +
  theme(legend.background = element_rect(fill="grey95", size=0.5,
                                         linetype="solid", colour ="black"),
        axis.text.x = element_text(angle = 45, hjust=1, size = 10, face="bold"), 
        axis.text.y = element_text(size = 10, face="bold"),
        axis.title = element_text(size = 12, face="bold"),
        panel.grid.minor = element_blank())

数据集 #

df <- structure(list(Taxa = c("Cnidaria ", "Cnidaria ", "Echinodermata", 
"Echinodermata", "Mollusca", "Mollusca", "Bryozoa", "Bryozoa", 
"Porifera", "Porifera", "Annelida", "Annelida", "Foraminifera", 
"Foraminifera", "Platyhelminthes", "Platyhelminthes", "Brachiopoda", 
"Brachiopoda"), sum = c(26L, 26L, 19L, 19L, 17L, 17L, 15L, 15L, 
14L, 14L, 11L, 11L, 9L, 9L, 7L, 7L, 2L, 2L), Method = structure(c(2L, 
1L, 2L, 1L, 2L, 1L, 2L, 1L, 2L, 1L, 2L, 1L, 2L, 1L, 2L, 1L, 2L, 
1L), .Label = c("BGR", "GSR"), class = "factor"), Observations = c(16L, 
10L, 10L, 9L, 11L, 6L, 12L, 3L, 11L, 3L, 8L, 3L, 4L, 5L, 6L, 
1L, 2L, 0L)), class = c("tbl_df", "tbl", "data.frame"), row.names = c(NA, 
-18L))

解决方案 #

方法1：添加组内占比文本 #

在每个柱状图上方显示该Method对应Taxa组的百分比占比：

library(tidyverse)

df %>%
  mutate(Taxa = fct_relevel(Taxa, 
                            "Cnidaria ", "Echinodermata", "Mollusca", 
                            "Bryozoa", "Porifera", "Annelida", 
                            "Foraminifera", "Platyhelminthes", "Brachiopoda"),
         # 计算组内百分比，保留1位小数
         group_pct = round(Observations / sum * 100, 1)) %>%
  ggplot(aes(x=Taxa,y=Observations,fill = Method)) +
  geom_bar(stat = "identity", width = 0.8 ,size = 1, position=position_dodge(), alpha = 0.9) +
  # 原始观测值文本
  geom_text(aes(label=Observations), vjust=1.6, color="white", size=4, position=position_dodge(0.9)) +
  # 组内占比文本
  geom_text(aes(label=paste0(group_pct, "%")), vjust=-0.5, size=3.5, position=position_dodge(0.9)) +
  theme_bw() +
  theme(legend.background = element_rect(fill="grey95", size=0.5,
                                         linetype="solid", colour ="black"),
        axis.text.x = element_text(angle = 45, hjust=1, size = 10, face="bold"), 
        axis.text.y = element_text(size = 10, face="bold"),
        axis.title = element_text(size = 12, face="bold"),
        panel.grid.minor = element_blank())

方法2：添加全局累积百分比折线（双Y轴） #

绘制所有Taxa组总和的累积占比折线，使用双Y轴适配数值范围：

# 预计算累积百分比数据
cumul_data <- df %>%
  distinct(Taxa, sum) %>%
  mutate(Taxa = fct_relevel(Taxa, 
                            "Cnidaria ", "Echinodermata", "Mollusca", 
                            "Bryozoa", "Porifera", "Annelida", 
                            "Foraminifera", "Platyhelminthes", "Brachiopoda")) %>%
  arrange(Taxa) %>%
  mutate(global_total = sum(sum),
         cumulative_sum = cumsum(sum),
         cumulative_pct = round(cumulative_sum / global_total * 100, 1))

# 绘图
df %>%
  mutate(Taxa = fct_relevel(Taxa, 
                            "Cnidaria ", "Echinodermata", "Mollusca", 
                            "Bryozoa", "Porifera", "Annelida", 
                            "Foraminifera", "Platyhelminthes", "Brachiopoda")) %>%
  ggplot() +
  # 分组柱状图
  geom_bar(aes(x=Taxa,y=Observations,fill = Method),
           stat = "identity", width = 0.8 ,size = 1, position=position_dodge(), alpha = 0.9) +
  geom_text(aes(x=Taxa,y=Observations,label=Observations, group=Method),
            vjust=1.6, color="white", size=4, position=position_dodge(0.9)) +
  # 累积百分比折线（适配Y轴范围）
  geom_line(data = cumul_data, aes(x=Taxa, y=cumulative_pct * max(df$Observations)/100, group=1),
            color="darkred", size=1.2) +
  geom_text(data = cumul_data, aes(x=Taxa, y=cumulative_pct * max(df$Observations)/100, label=paste0(cumulative_pct, "%")),
            vjust=-0.5, color="darkred", size=3.5) +
  # 设置双Y轴
  scale_y_continuous(
    name = "Observations",
    sec.axis = sec_axis(~ . * 100 / max(df$Observations), name = "Cumulative Percentage (%)")
  ) +
  theme_bw() +
  theme(legend.background = element_rect(fill="grey95", size=0.5,
                                         linetype="solid", colour ="black"),
        axis.text.x = element_text(angle = 45, hjust=1, size = 10, face="bold"), 
        axis.text.y = element_text(size = 10, face="bold"),
        axis.text.y.right = element_text(color="darkred", size=10, face="bold"),
        axis.title = element_text(size = 12, face="bold"),
        axis.title.y.right = element_text(color="darkred", size=12, face="bold"),
        panel.grid.minor = element_blank())

方法3：在geom_line中使用filter筛选特定Method #

仅为指定Method绘制其观测值的累积折线：

df %>%
  mutate(Taxa = fct_relevel(Taxa, 
                            "Cnidaria ", "Echinodermata", "Mollusca", 
                            "Bryozoa", "Porifera", "Annelida", 
                            "Foraminifera", "Platyhelminthes", "Brachiopoda")) %>%
  ggplot(aes(x=Taxa,y=Observations,fill = Method)) +
  geom_bar(stat = "identity", width = 0.8 ,size = 1, position=position_dodge(), alpha = 0.9) +
  geom_text(aes(label=Observations), vjust=1.6, color="white", size=4, position=position_dodge(0.9)) +
  # 筛选GSR方法，计算累积值并绘制折线
  geom_line(data = . %>% filter(Method == "GSR") %>% arrange(Taxa) %>% mutate(cumulative = cumsum(Observations)),
            aes(x=Taxa, y=cumulative, group=1), color="darkblue", size=1) +
  geom_text(data = . %>% filter(Method == "GSR") %>% arrange(Taxa) %>% mutate(cumulative = cumsum(Observations)),
            aes(x=Taxa, y=cumulative, label=cumulative), vjust=-0.5, color="darkblue", size=3.5) +
  theme_bw() +
  theme(legend.background = element_rect(fill="grey95", size=0.5,
                                         linetype="solid", colour ="black"),
        axis.text.x = element_text(angle = 45, hjust=1, size = 10, face="bold"), 
        axis.text.y = element_text(size = 10, face="bold"),
        axis.title = element_text(size = 12, face="bold"),
        panel.grid.minor = element_blank())

注：geom_line的data参数中，用.指代上游数据集，链式调用filter筛选目标Method后计算累积值即可。

内容的提问来源于stack exchange，提问作者C. Guff