R语言ggplot2 facet分面标签如何添加化学式下标

直接给factor()的labels参数传入expression对象无法实现分面标签下标渲染，核心原因有两个：

• 因子的标签仅支持存储纯文本内容，传入的expression会被强制转换为普通字符串，plotmath语法不会被识别
• 原有expression写法将带[3]下标的内容整体包裹在引号内，即使渲染逻辑正确，引号内的内容也会被判定为普通文本，不会触发下标解析

方案1：基础ggplot2实现（无需安装额外包） #

不要提前在数据列中存储expression，改为在分面时通过labeller参数传入命名的表达式标签向量，注意expression的写法：下标部分不要放在引号内，用~表示空格。

# 定义匹配参数名的标签向量，名称对应数据中Parameter列的原始值
facet_labs <- c(
  "Alkalinity" = " Alkalinity",
  "Ca (CaCO3)" = "Ca (CaCO3)",
  "Cl(mg/l)," = "Cl (mg/l)",
  "Conductivity (mS/m)" = "Conductivity (mS/m)",
  "Cu (mg/l)" = "Cu (mg/l)",
  "F (mg/l)" = "F (mg/l)",
  "Fe (mg/l)" = "Fe (mg/l)",
  "K (mg/l" = "K (mg/l)",
  "Mg (CaCO3)" = "Mg (CaCO3)",
  " Mn(mg/l)" = "Mn (mg/l)",
  "Na (mg/l)" = "Na (mg/l)",
  "NO3" = expression(NO[3]~(mg/l)),
  "pH" = "pH",
  "SiO2" = expression(SiO[2]~(mg/l)),
  "SO4" = expression(SO[4]~(mg/l)),
  "TDS (TDS Calculated)" = "TDS (TDS Calculated)",
  "Turbidity (NTU)" = "Turbidity (NTU)",
  "Zn (mg/l)" = "Zn (mg/l)"
)

# 绘图时直接使用原始Parameter列分面，指定labeller
library(ggplot2)
ggplot(na.omit(NewDat1), aes(Year, Value, color = Location)) + 
  geom_line() + 
  geom_point() + 
  facet_wrap(~Parameter, scales = "free_y", labeller = labeller(Parameter = facet_labs))

注意：需要保证命名向量的名称和数据里Parameter列的参数值完全匹配，否则标签会映射失败，建议提前修正数据里Parameter列的笔误（比如"K (mg/l"缺少右括号、"Cl(mg/l),"多了逗号）。

方案2：ggtext包实现（语法更简单） #

如果不想记忆plotmath的expression语法，可以使用ggtext包，直接用markdown的下标语法~下标内容~写标签，配置主题让分面标题支持markdown渲染即可，写法更不容易出错。

# 首次使用先安装：install.packages("ggtext")
library(ggtext)

# 用markdown语法定义标签，不需要写expression
facet_labs_md <- c(
  "Alkalinity" = " Alkalinity",
  "Ca (CaCO3)" = "Ca (CaCO3)",
  "Cl(mg/l)," = "Cl (mg/l)",
  "Conductivity (mS/m)" = "Conductivity (mS/m)",
  "Cu (mg/l)" = "Cu (mg/l)",
  "F (mg/l)" = "F (mg/l)",
  "Fe (mg/l)" = "Fe (mg/l)",
  "K (mg/l" = "K (mg/l)",
  "Mg (CaCO3)" = "Mg (CaCO3)",
  " Mn(mg/l)" = "Mn (mg/l)",
  "Na (mg/l)" = "Na (mg/l)",
  "NO3" = "NO~3~ (mg/l)",
  "pH" = "pH",
  "SiO2" = "SiO~2~ (mg/l)",
  "SO4" = "SO~4~ (mg/l)",
  "TDS (TDS Calculated)" = "TDS (TDS Calculated)",
  "Turbidity (NTU)" = "Turbidity (NTU)",
  "Zn (mg/l)" = "Zn (mg/l)"
)

ggplot(na.omit(NewDat1), aes(Year, Value, color = Location)) + 
  geom_line() + 
  geom_point() + 
  facet_wrap(~Parameter, scales = "free_y", labeller = labeller(Parameter = facet_labs_md)) +
  # 配置分面标题使用markdown渲染
  theme(strip.text = element_markdown())

两种方案都可以正确渲染NO₃、SiO₂、SO₄这类化学符号下标，第二种方案除了下标外还支持上标、字体颜色、加粗等更丰富的格式，适合更复杂的标签需求。

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