ggplot森林图右侧添加文本列及隐藏表头绘图行问题求助

需求1：限制绘图区域右侧边界，避免文本与CI线重叠 #

直接通过coord_cartesian()手动锁定x轴显示范围，预留右侧空间放置文本，同时精准定位文本位置：

1. 先计算合理的x轴绘图上限（比如取所有CI上限的1.2倍），确保CI线不会延伸到留白区
2. 用coord_cartesian(xlim = c(min_x, max_plot_x))限制绘图区域，不删除数据仅控制显示范围
3. 将文本的x坐标设为max_plot_x + 固定偏移量，通过hjust=0左对齐，保证和对应行严格对齐

示例代码片段：

# 计算绘图区域的x轴上下限
max_plot_x <- max(df$upper) * 1.2
min_x <- min(df$lower) * 1.1

ggplot(df, aes(y = term, x = estimate)) +
  geom_pointrange(aes(xmin = lower, xmax = upper)) +
  # 锁定绘图范围，允许文本绘制在区域外
  coord_cartesian(xlim = c(min_x, max_plot_x), clip = "off") +
  # 添加点估计+CI文本
  geom_text(aes(x = max_plot_x + 0.5, label = paste0(round(estimate,2), " (", round(lower,2), "-", round(upper,2), ")")), 
            hjust = 0, size = 3) +
  # 添加p值文本
  geom_text(aes(x = max_plot_x + 2, label = paste0("p=", round(p_value,3))), 
            hjust = 0, size = 3) +
  # 调整右侧边距，防止文本被裁剪
  theme(plot.margin = margin(5, 50, 5, 5))

需求2：分组绘图并添加无绘图区域的文本表头 #

无需拆分绘图再合并，直接在数据中插入表头行，将该行的绘图相关字段设为NA，让geom_pointrange自动跳过绘图，仅保留文本：

1. 为每个model分组添加一行表头数据，term设为分组标题，estimate/lower/upper设为NA
2. 用facet_wrap(~model, scales = "free_y")按模型分组
3. 单独设置表头文本的样式，区分于普通行

示例完整代码：

library(ggplot2)

# 构造示例数据
df <- data.frame(
  model = rep(c("Model A", "Model B"), each = 3),
  term = c("变量1", "变量2", "变量3", "变量1", "变量2", "变量3"),
  estimate = c(0.5, 0.8, -0.3, 0.6, 0.9, -0.2),
  lower = c(0.2, 0.5, -0.6, 0.3, 0.6, -0.5),
  upper = c(0.8, 1.1, 0, 0.9, 1.2, 0.1),
  p_value = c(0.01, 0.002, 0.15, 0.005, 0.001, 0.2)
)

# 添加分组表头行
header_rows <- data.frame(
  model = c("Model A", "Model B"),
  term = c("**Model A 结果**", "**Model B 结果**"),
  estimate = NA, lower = NA, upper = NA, p_value = NA
)
df_with_header <- rbind(header_rows, df)

# 计算绘图x轴范围
max_plot_x <- max(df$upper, na.rm = TRUE) * 1.2
min_x <- min(df$lower, na.rm = TRUE) * 1.1

# 绘制森林图
ggplot(df_with_header, aes(y = reorder(term, desc(term)), x = estimate)) +
  geom_pointrange(aes(xmin = lower, xmax = upper), na.rm = TRUE) +
  facet_wrap(~model, scales = "free_y", ncol = 1) +
  coord_cartesian(xlim = c(min_x, max_plot_x), clip = "off") +
  # 添加点估计+CI文本（空出行表头位置）
  geom_text(aes(x = max_plot_x + 0.4, label = ifelse(is.na(estimate), "", paste0(round(estimate,2), " (", round(lower,2), "-", round(upper,2), ")"))), 
            hjust = 0, size = 3) +
  # 添加p值文本（空出行表头位置）
  geom_text(aes(x = max_plot_x + 1.8, label = ifelse(is.na(p_value), "", paste0("p=", round(p_value,3)))), 
            hjust = 0, size = 3) +
  # 绘制自定义表头文本
  geom_text(aes(x = min_x, label = term), color = "darkred", size = 4, hjust = 0, na.rm = TRUE) +
  theme(
    plot.margin = margin(5, 60, 5, 5),
    strip.background = element_blank(),
    strip.text = element_blank() # 隐藏默认facet标题，用自定义表头替代
  )

注：na.rm=TRUE让绘图元素自动忽略NA值行，不会生成空的绘图区域；reorder(term, desc(term))调整行顺序，让表头显示在每组顶部。

内容的提问来源于stack exchange，提问作者Rolvix Patterson