ggplot2：为分面网格标签添加换行且不生成额外列

嗨，看起来你是想给ggplot分面里的长标签换行，还不想额外新增数据列对吧？我来帮你搞定这个问题～

首先先把你的数据框补全（原代码里p_m列没写完，我按照逻辑补全了内容）：

# 构建数据框
phone <- c("p", "b", "t", "d", "k", "g", "ts", "dz", "tʃ", "dʒ", "f", "v", "s", "z")
count <- c(1088, 211, 1260, 208, 2596, 1041, 15, 13, 1262, 51, 22, 9, 765, 8)
prop <- c(0.84, 0.16, 0.86, 0.14, 0.71, 0.29, 0.53, 0.46, 0.96, 0.04, 0.71, 0.29, 0.99, 0.01)
voicing <- c("unvoiced", "voiced", "unvoiced", "voiced", "unvoiced", "voiced", 
             "unvoiced", "voiced", "unvoiced", "voiced", "unvoiced", "voiced", 
             "unvoiced", "voiced")
p_m <- c("Bilabial Plosive (N=1299)", "Bilabial Plosive (N=1299)", 
         "Dental Plosive (N=1468)", "Dental Plosive (N=1468)", 
         "Velar Plosive (N=3637)", "Velar Plosive (N=3637)",
         "Alveolar Affricate (N=28)", "Alveolar Affricate (N=28)",
         "Palatal Affricate (N=1313)", "Palatal Affricate (N=1313)",
         "Labiodental Fricative (N=31)", "Labiodental Fricative (N=31)",
         "Alveolar Fricative (N=773)", "Alveolar Fricative (N=773)")

df <- data.frame(phone, count, prop, voicing, p_m)

假设你原本的绘图代码是类似这样的（分面展示每个发音部位的比例）：

library(ggplot2)

# 原始绘图，长标签会挤在一起
ggplot(df, aes(x = phone, y = prop, fill = voicing)) +
  geom_col() +
  facet_wrap(~p_m) +
  theme(axis.text.x = element_text(angle = 45, hjust = 1))

下面给你两种不需要新增数据列的换行方法：

方法一：手动插入换行符\n #

你可以直接在现有p_m列里批量添加换行，比如在(N=前拆分文本，不用新增列：

# 批量在"(N="前插入换行
df$p_m <- gsub(" \\(N=", "\n(N=", df$p_m)

# 重新绘图
ggplot(df, aes(x = phone, y = prop, fill = voicing)) +
  geom_col() +
  facet_wrap(~p_m) +
  theme(axis.text.x = element_text(angle = 45, hjust = 1),
        strip.text = element_text(size = 10)) # 可调整标签字号让换行更美观

方法二：用stringr::str_wrap自动换行 #

如果不想手动指定拆分位置，直接在分面函数里用str_wrap自动按字符宽度拆分，完全不用修改原始数据：

library(stringr)

ggplot(df, aes(x = phone, y = prop, fill = voicing)) +
  geom_col() +
  facet_wrap(~str_wrap(p_m, width = 15)) + # 指定每行最多15个字符自动换行
  theme(axis.text.x = element_text(angle = 45, hjust = 1),
        strip.text = element_text(size = 10))

这两种方法都不会生成额外的数据列，分面标签会自动变成多行，再也不会挤在一起啦～

备注：内容来源于stack exchange，提问作者Wangana