如何在ggplot的facet_grid中去除重复的Y轴标签？

你遇到的问题根源是代码里的scale_y_discrete(limits = d_demog$model)强制所有分面显示完整的Y轴标签列表，加上默认facet_grid的行为，就导致了重复。下面给你两种简便的解决方法：

方法一：使用ggh4x包的facet_grid2（推荐） #

ggh4x扩展包增强了ggplot的分面功能，能轻松实现每个分面只显示对应组的Y轴标签，同时保持facet_grid的紧凑布局：

1. 先安装并加载包：

install.packages("ggh4x")
library(ggh4x)

2. 修改原代码中的分面部分，把facet_grid替换为facet_grid2，并移除强制显示全部Y标签的设置：

修改后的完整代码：

ggplot(d_demog, aes(y = model, x = coef, shape = model,
               xmin = ci_lower, xmax = ci_upper)) +
  geom_errorbarh(aes(xmin = ci_lower, xmax = ci_upper), 
                 position = position_dodge(.5),
                 linewidth = .8,
                 height = 0,
                 alpha = 0.5) + 
  geom_point(size = 3.5, fill = "white", stroke = 2) +
  geom_vline(xintercept = 0, 
             linetype = "dashed", 
             color = "black",
             linewidth = .3) +
  labs(x = "Coefficient",
       y = NULL) +
  scale_x_continuous(breaks = seq(-0.5, 0.2, 0.1)) +
  coord_cartesian(xlim = c(-0.55, 0.21)) +
  scale_shape_manual(values = c("CohortA" = 23,
                                "CohortB" = 23,
                                "Female" = 21,
                                "Male" = 21,
                                "White" = 25,
                                "Hispanic" = 25,
                                "Black" = 25)) +
  # 替换为ggh4x的增强分面函数
  facet_grid2(rows = vars(group),
              scales = "free_y",
              space = "free") +
  theme_minimal(base_size = 15) +
  theme(legend.position = "none",
        panel.grid = element_blank(),
        strip.text = element_blank(),
        strip.background = element_blank())

方法二：手动调整ggplot主题（无需额外包） #

如果不想安装新包，可以通过修改theme的axis.text.y，结合分面面板索引控制标签显示：

ggplot(d_demog, aes(y = model, x = coef, shape = model,
               xmin = ci_lower, xmax = ci_upper)) +
  geom_errorbarh(aes(xmin = ci_lower, xmax = ci_upper), 
                 position = position_dodge(.5),
                 linewidth = .8,
                 height = 0,
                 alpha = 0.5) + 
  geom_point(size = 3.5, fill = "white", stroke = 2) +
  geom_vline(xintercept = 0, 
             linetype = "dashed", 
             color = "black",
             linewidth = .3) +
  labs(x = "Coefficient",
       y = NULL) +
  scale_x_continuous(breaks = seq(-0.5, 0.2, 0.1)) +
  coord_cartesian(xlim = c(-0.55, 0.21)) +
  scale_shape_manual(values = c("CohortA" = 23,
                                "CohortB" = 23,
                                "Female" = 21,
                                "Male" = 21,
                                "White" = 25,
                                "Hispanic" = 25,
                                "Black" = 25)) +
  facet_grid(rows = vars(group),
             scales = "free_y",
             space = "free") +
  theme_minimal(base_size = 15) +
  theme(legend.position = "none",
        panel.grid = element_blank(),
        strip.text = element_blank(),
        strip.background = element_blank(),
        # 只显示当前分面对应的Y轴标签，其余设为透明
        axis.text.y = element_text(
          color = ifelse(
            d_demog$model %in% split(d_demog$model, d_demog$group)[[as.character(panel_number())]],
            "black", "transparent"
          )
        ))

内容的提问来源于stack exchange，提问作者a_todd12