ggplot2 facet_grid分面图y轴标签手动重命名错位问题

facet_grid()开启scales = "free"、space = "free"做纵向分面时，每个分面会独立渲染自身的y轴刻度。此时如果给scale_y_discrete()传入无命名的标签向量，函数会按每个分面内出现的因子水平数量，从向量头部开始截取对应长度的标签做匹配，不会跨分面做全局顺序对齐，这就是标签重复、错位的根本原因。

核心逻辑是让标签和原始因子值做一一绑定，放弃依赖位置顺序的匹配方式，两种常用实现方式如下：

• 方法1：传入带命名的标签向量给scale_y_discrete()
  把12个自定义标签和你在fct_relevel()中排序好的12个原始因子水平一一对应，构造带名称的标签向量，名称填原始因子值，值填要显示的自定义标签，顺序和你设置的因子水平保持一致即可：

  y_labels <- c(
    "BNC_contr_frac_MNC" = "BNC contribution of a fraction to its MNC [%]",
    "FNC_contr_frac_MNC" = "FNC contribution of a fraction to its MNC [%]",
    "MNC_contr_frac_C" = "MNC contribution of a fraction to its C [%]",
    "BNC_contr_bothfrac_bothfrac_MNC" = "BNC contribution of both fractions to their joined MNC [%]",
    "FNC_contr_bothfrac_bothfrac_MNC" = "FNC contribution of both fractions to their joined MNC [%]",
    "MNC_contr_bothfrac_bothfrac_C" = "MNC contribution of both fractions to their joined C [%]",
    "BNC_contr_frac_bulk_C" = "BNC contribution of a fraction to bulk soil C [%]",
    "FNC_contr_frac_bulk_C" = "FNC contribution of a fraction to bulk soil C [%]",
    "MNC_contr_frac_bulk_C" = "MNC contribution of a fraction to bulk soil C [%]",
    "BNC_contr_bothfrac_bulk_C" = "BNC contribution of both joined fractions to bulk soil C [%]",
    "FNC_contr_bothfrac_bulk_C" = "FNC contribution of both joined fractions to bulk soil C [%]",
    "MNC_contr_bothfrac_bulk_C" = "MNC contribution of both joined fractions to bulk soil C [%]"
  )
  

  把原来无命名参数的scale_y_discrete()替换为传入上述命名向量的版本即可，不管分面怎么切割，标签都会按原始值精准匹配，不会错位：

  scale_y_discrete(labels = y_labels)
  

• 方法2：数据预处理阶段直接重编码因子标签
  更一劳永逸的方式是在绘图前的数据处理环节，直接把target_var列的因子水平替换为要显示的自定义标签，后续绘图不需要额外设置坐标轴标签，从数据源头上固定映射关系，完全避免分面带来的匹配错误：

  dat_AS_regr <- dat_AS_regr %>%
    mutate(
      target_var = fct_relevel(target_var,
        "BNC_contr_frac_MNC", "FNC_contr_frac_MNC", "MNC_contr_frac_C",
        "BNC_contr_bothfrac_bothfrac_MNC", "FNC_contr_bothfrac_bothfrac_MNC", 
        "MNC_contr_bothfrac_bothfrac_C","BNC_contr_frac_bulk_C","FNC_contr_frac_bulk_C", 
        "MNC_contr_frac_bulk_C","BNC_contr_bothfrac_bulk_C","FNC_contr_bothfrac_bulk_C", 
        "MNC_contr_bothfrac_bulk_C"),
      target_var = fct_recode(target_var, !!!y_labels)
    )
  

注意：只要使用了自由标尺的分面，就不要用无命名的、依赖位置顺序的标签向量，只要单个分面内没有展示全量因子水平，位置匹配逻辑就会出现错位。

内容的提问来源于stack exchange，提问作者nay