如何用ggplot2将系数图的A-J标签按类别分组展示

核心思路是先给你的数据添加分组变量，再通过分面或视觉区分实现分组展示，以下是具体步骤：

1. 给数据添加分组列 #

假设你的数据框（比如叫df）包含term（存A-J标签）、estimate（系数值）、std.error（标准误）这些绘图必需列。用dplyr的case_when给每个标签匹配对应分组：

library(dplyr)
library(ggplot2)

# 新增分组变量并固定顺序，避免绘图乱序
df <- df %>%
  mutate(
    category = case_when(
      term %in% c("A", "B", "C") ~ "Category 1",
      term %in% c("D", "E", "F") ~ "Category 2",
      term %in% c("G", "H", "I", "J") ~ "Category 3",
      TRUE ~ "Other" # 兜底处理意外标签
    ),
    term = factor(term, levels = c("A", "B", "C", "D", "E", "F", "G", "H", "I", "J")),
    category = factor(category, levels = c("Category 1", "Category 2", "Category 3"))
  )

2. 分面式分组（最接近示例图效果） #

用facet_wrap把不同分组拆成独立子图，每个子图只显示对应组的标签，视觉上分组清晰：

ggplot(df, aes(x = term, y = estimate)) +
  # 绘制带95%置信区间的系数点
  geom_pointrange(aes(ymin = estimate - 1.96*std.error, ymax = estimate + 1.96*std.error)) +
  # 添加系数为0的参考线
  geom_hline(yintercept = 0, linetype = "dashed", color = "red") +
  # 按分组分面，每个分面独立显示x轴标签
  facet_wrap(~category, scales = "free_x", ncol = 1) # ncol=1是纵向排列，改成3可横向展示
  theme_bw() +
  labs(x = "标签", y = "系数（95%置信区间）")

3. 同画布内分组（带颜色+分隔线版） #

如果不想拆分成分面，可在同一图里用颜色区分分组，再加分隔线强化边界：

ggplot(df, aes(x = term, y = estimate)) +
  geom_pointrange(aes(ymin = estimate - 1.96*std.error, ymax = estimate + 1.96*std.error, color = category)) +
  geom_hline(yintercept = 0, linetype = "dashed", color = "red") +
  # 在标签C/D、F/G之间添加分组分隔线
  geom_vline(xintercept = 3.5, linetype = "solid", color = "gray50") +
  geom_vline(xintercept = 6.5, linetype = "solid", color = "gray50") +
  theme_bw() +
  labs(x = "标签", y = "系数（95%置信区间）", color = "分组")

内容的提问来源于stack exchange，提问作者Neicooo