ggplot2中如何将facet_wrap的Y轴分类strip移至对应分类顶部？

问题背景 #

现有数据集：

df <- data.frame(ORDER = c(1,2,3,4),
                 NAME_LABEL = c("COMPANY1", "COMPANY2", "COMPANY3", "COMPANY4"),
                 LABEL = c("ICT hardware", "Software", "Automobiles", "Travel"),
                 RD_PC = c(60, 50, 30, -10),
                 COLOR = c("POS", "POS", "POS", "NEG"))

原绘图代码：

col_fig5 <- c("#6B89BC", "#A4C76D")

test <- ggplot(df, aes(x = forcats::fct_rev(reorder(NAME_LABEL, ORDER, sum)), group = LABEL, fill = COLOR)) +
  geom_col(aes(y = RD_PC, fill = COLOR), width = 0.85) +
  scale_fill_manual(values = col_fig5) +
  ylim(c(-55, 95)) +
  coord_flip() +
  facet_wrap(~ LABEL, ncol = 1, scales = "free_y", strip.position = "left", as.table=TRUE) +
  labs(
    y = "Percentage change",
    x = "",
    fill = "",
    color = "",
    caption = ""
  ) +
  theme_classic() +
  theme(strip.text.y = element_text(face = "bold", hjust = 0, vjust = 1, size = 4),
        strip.background = element_blank(),
        panel.grid.major.x = element_line(colour="grey85", size=0.1),
        panel.grid.minor.x = element_line(colour="grey85", size=0.1),
        panel.grid.major.y = element_line(colour="grey85", size=0.1),
        axis.text.y = element_text(size=5,  hjust=1),
        axis.title.y = element_text(size=8, vjust=0.5, hjust=1),
        axis.ticks.y = element_blank(),
        axis.text.x = element_text(size=6, angle=0),
        axis.title.x = element_text(size=8),
        axis.line.x = element_line(size=0.4),
        axis.line.y = element_blank(),
        plot.caption = element_text(size=6, hjust=0),
        legend.position = "none")

需求：公司数量较多时，将每个LABEL分类的strip放置在Y轴对应分类的顶部，调整默认参数未生效，寻求可行方案。

解决方案 #

方法一：使用ggh4x包实现灵活分面标签定位 #

ggh4x扩展包提供了比原生ggplot2更灵活的分面控制，能轻松实现strip与对应Y轴分类的对齐：

1. 先安装并加载包：

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

2. 修改原绘图代码，替换facet_wrap为facet_wrap2，并调整标签对齐：

col_fig5 <- c("#6B89BC", "#A4C76D")

test <- ggplot(df, aes(x = forcats::fct_rev(reorder(NAME_LABEL, ORDER, sum)), fill = COLOR)) +
  geom_col(aes(y = RD_PC), width = 0.85) +
  scale_fill_manual(values = col_fig5) +
  ylim(c(-55, 95)) +
  coord_flip() +
  # 用facet_wrap2替换原生facet_wrap，设置strip在左侧并外置对齐
  ggh4x::facet_wrap2(~ LABEL, ncol = 1, scales = "free_y", strip.position = "left", strip.placement = "outside") +
  labs(
    y = "Percentage change",
    x = "",
    fill = "",
    color = "",
    caption = ""
  ) +
  theme_classic() +
  theme(strip.text.y.left = element_text(face = "bold", hjust = 1, vjust = 0.5, size = 4),
        strip.background = element_blank(),
        panel.grid.major.x = element_line(colour="grey85", size=0.1),
        panel.grid.minor.x = element_line(colour="grey85", size=0.1),
        axis.text.y = element_text(size=5, hjust=1),
        axis.title.y = element_blank(), # 可根据需求保留或删除Y轴标题
        axis.ticks.y = element_blank(),
        axis.text.x = element_text(size=6, angle=0),
        axis.title.x = element_text(size=8),
        axis.line.x = element_line(size=0.4),
        axis.line.y = element_blank(),
        plot.caption = element_text(size=6, hjust=0),
        legend.position = "none")

这个方法适合一个LABEL对应多个公司的场景，分面逻辑清晰，标签位置自动对齐对应分类。

方法二：手动添加标签（无需额外包） #

如果每个LABEL仅对应单个公司，可以直接用geom_text将LABEL文本添加到对应条形的左侧顶部，模拟strip效果：

col_fig5 <- c("#6B89BC", "#A4C76D")

test <- ggplot(df, aes(x = forcats::fct_rev(reorder(NAME_LABEL, ORDER, sum)), fill = COLOR)) +
  geom_col(aes(y = RD_PC), width = 0.85) +
  # 添加LABEL标签到条形左侧顶部，位置可通过y参数调整
  geom_text(aes(y = ifelse(RD_PC > 0, RD_PC * 1.1, -10), label = LABEL), 
            hjust = 1, size = 1.5, fontface = "bold") +
  scale_fill_manual(values = col_fig5) +
  ylim(c(-55, 95)) +
  coord_flip() +
  labs(
    y = "Percentage change",
    x = "",
    fill = "",
    color = "",
    caption = ""
  ) +
  theme_classic() +
  theme(panel.grid.major.x = element_line(colour="grey85", size=0.1),
        panel.grid.minor.x = element_line(colour="grey85", size=0.1),
        axis.text.y = element_text(size=5, hjust=1),
        axis.title.y = element_text(size=8, vjust=0.5, hjust=1),
        axis.ticks.y = element_blank(),
        axis.text.x = element_text(size=6, angle=0),
        axis.title.x = element_text(size=8),
        axis.line.x = element_line(size=0.4),
        axis.line.y = element_blank(),
        plot.caption = element_text(size=6, hjust=0),
        legend.position = "none")

这种方法无需依赖额外包，标签位置完全可控，适合单公司对应单LABEL的场景。

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