如何解决geom_text_repel标注散点图时标签换行后溢出画布问题

我用geom_text_repel给散点图中的部分点添加标注，希望通过文本换行避免标签过长，但标签始终会超出绘图区域被截断。之前成功实现过标签正常显示的效果，但现在找不到对应代码了。

我尝试用字符串换行的方式处理，代码如下：

mobility %>%
  ggplot(aes(access, success, colour=group)) +
  geom_point() +
  geom_text_repel(data=subset(mobility, label == 1),
            aes(access, success, label=stringr::str_wrap(university, 3)),
            show.legend = FALSE, box.padding = unit(0.45, "lines"),
            vjust="inward", hjust="inward") +
  scale_y_continuous(labels = percent) +
  scale_x_continuous(labels = percent) +
  guides(color = guide_legend(nrow = 3), text='none') +
  theme_economist() +
  labs(x='Access rate', y='Success rate', colour='University type', title='Figure 3')

目前不管是否启用换行，标签都会出现截断问题。如何实现标签换行、不重叠且不超出绘图区域的效果？

数据示例：

mobility <- structure(list(university = c("Royal Veterinary College", "LSE", 
"Imperial College", "Oxford", "Cambridge", "London South Bank", 
"London Metropolitan", "University of East London", "King's College London Other Russell Group", 
"University of Durham Other Russell Group", "University ofWarwick Other Russell Group", 
"University College London Other Russell Group", "University of Bristol Other Russell Group", 
"QueenMary, University of Lon.. Other Russell Group", "University of Sheffield Other Russell Group"
), group = c("Old Universities (Pre-1992)", "Most Selective Russell", 
"Most Selective Russell", "Most Selective Russell", "Most Selective Russell", 
"Post-1992 (least selective)", "Post-1992 (least selective)", 
"Post-1992 (least selective)", "Other Russell Group", "Other Russell Group", 
"Other Russell Group", "Other Russell Group", "Other Russell Group", 
"Other Russell Group", "Other Russell Group"), access = c(0.013, 
0.046, 0.038, 0.009, 0.01, 0.257, 0.246, 0.295, 0.059, 0.012, 
0.015, 0.043, 0.01, 0.161, 0.016), success = c(0.625, 0.611, 
0.603, 0.59, 0.544, 0.18, 0.164, 0.138, 0.498, 0.489, 0.461, 
0.455, 0.423, 0.422, 0.422), label = c(1, 1, 1, 1, 1, 1, 1, 1, 
NA, NA, NA, NA, NA, NA, NA)), row.names = c(NA, 15L), class = "data.frame")

解决方案 #

核心修改点及完整代码 #

library(ggplot2)
library(ggrepel)
library(stringr)
library(ggthemes)
library(scales)

mobility %>%
  ggplot(aes(access, success, colour=group)) +
  geom_point() +
  geom_text_repel(data=subset(mobility, label == 1),
                  aes(label=str_wrap(university, 15)),
                  show.legend = FALSE,
                  box.padding = unit(0.8, "lines"),
                  point.padding = unit(0.5, "lines"),
                  max.overlaps = Inf,
                  direction = "both") +
  scale_y_continuous(labels = percent, expand = expansion(mult = 0.1)) +
  scale_x_continuous(labels = percent, expand = expansion(mult = 0.1)) +
  guides(color = guide_legend(nrow = 3)) +
  theme_economist() +
  theme(plot.margin = margin(1, 1.5, 1, 1.5, "cm")) +
  labs(x='Access rate', y='Success rate', colour='University type', title='Figure 3')

关键调整说明 #

1. 修正换行宽度：把str_wrap(university, 3)改成str_wrap(university, 15)，过窄的宽度会让标签拆成过多行，反而容易溢出，15-20的宽度更适合大学名称的换行逻辑。
2. 扩展绘图区域：在坐标轴设置中加入expand = expansion(mult = 0.1)，给绘图区四周留出额外空间，避免标签超出边界。
3. 调整标签排版参数：
   • 增大box.padding和point.padding，让标签与点、标签之间的间距更合理，减少重叠和溢出
   • 开启max.overlaps = Inf允许标签适当重叠（如果需要），direction = "both"让标签可以向任意方向调整位置
   • 移除vjust="inward", hjust="inward"，避免限制标签的排版灵活性
4. 增加图表边距：用theme(plot.margin = margin(1, 1.5, 1, 1.5, "cm"))给图表四周留出足够空间容纳标签，抵消theme_economist()的窄边距默认设置。

内容的提问来源于stack exchange，提问作者selroh18