如何消除facet_wrap子图内部边框、缩小间距并隐藏指定子图y轴刻度与轴线

问题1：消除子图内部边框并缩小子图间距 #

你设置panel.spacing未生效，核心原因是theme_bw()自带的panel.border会覆盖部分布局逻辑，需要配合调整边框显示规则来消除内部重复边框，同时确保自定义主题优先级更高。

问题2：隐藏第2、3张子图的y轴刻度与轴线 #

可以借助ggplot2原生的after_stat()结合panel_number()，针对指定面板精准控制轴元素的显示状态，无需额外安装扩展包。

完整修正代码 #

先修正你数据中的语法错误（如向量缺逗号、数据集变量名不一致），再整合布局调整逻辑：

library("ggplot2")
library("dplyr")
library("ggprism")

# 修正并统一数据集
plan <- data.frame(
  passage = c(1, 1, 3, 3, 1, 1, 3, 3, 1, 1, 3, 3),
  value = c(0.7858260, 0.4605525, 0.7958260, 0.5505525, 4.5000000, 0.3223315, 4.1000000, 3.2700000, 4.6000000, 0.3523315, 5.6000000, 4.4200000),
  sd = c(0.7858260, 0.4605525, 0.7958260, 0.5505525, 4.5000000, 0.3223315, 4.1000000, 3.2700000, 4.6000000, 0.3523315, 5.6000000, 4.4200000),
  gravity = c("uni", "iso","uni", "iso","uni", "iso","uni", "iso","uni", "iso","uni", "iso"),
  parameter = c("conc.e", "conc.e", "conc.e", "conc.e","conc.m","conc.m","conc.m","conc.m","conc.l","conc.l","conc.l","conc.l")
)

# 自定义主题元素
no_grid <- theme(panel.grid.major = element_blank(), panel.grid.minor = element_blank())
tick <- annotation_ticks(sides = "bl", outside = FALSE, minor.length = unit(0, "pt"), tick.length = unit(0, "pt"))
no_ribbonBGX <- theme(strip.background = element_blank(), strip.placement = "outside")
titleSpacing <- theme(axis.title.x = element_text(vjust= -1),
                      axis.text.x = element_text(vjust=-0.5), axis.title.y = element_text(vjust= 3))

# 核心布局调整主题
facet_layout_fix <- theme(
  # 缩小子图间距（负数可让子图更紧凑，按需调整）
  panel.spacing = unit(-5, "pt"),
  # 移除默认面板边框，避免内部重复边框
  panel.border = element_blank(),
  # 重新绘制外侧坐标轴，保证轴的完整性
  axis.line = element_line(color = "black"),
  # 针对第2、3面板隐藏y轴元素
  axis.text.y = element_text(color = after_stat(ifelse(panel_number() %in% c(2,3), "transparent", "black"))),
  axis.ticks.y = element_line(color = after_stat(ifelse(panel_number() %in% c(2,3), "transparent", "black"))),
  axis.line.y = element_line(color = after_stat(ifelse(panel_number() %in% c(2,3), "transparent", "black")))
)

# 替换为你实际的自定义填充刻度
FillScale <- scale_fill_manual(values = c("uni" = "#619CFF", "iso" = "#F8766D"))
# 替换为你实际的y轴标签
ylab31 <- ylab("自定义Y轴标签")
# 替换为你实际的分面标签映射
facetLabel <- c("conc.e" = "E", "conc.m" = "M", "conc.l" = "L")

p1 <- ggplot(plan, aes(x = passage, y = value, fill = factor(gravity, levels = c("uni", "iso")) )) + 
  geom_bar(stat = 'identity', position = position_dodge(width = 1.2), width = 1, color = "black")  +
  geom_errorbar(aes(ymin = value - sd, ymax = value + sd), position = position_dodge(width = 1.2), width = 1) +
  facet_wrap(vars(factor(parameter, levels = c("conc.e", "conc.m", "conc.l"))), 
             ncol = 3, scales = "free_y",  
             labeller = as_labeller(facetLabel), strip.position="bottom")  +
  scale_y_continuous(limit = c(0,7.5), expand = expansion(mult = c(0,0.03))) + 
  scale_x_continuous(limit = c(0, 4), breaks = c(1,3), labels = c("Low", "High")) +
  xlab("Passage") + ylab31  + coord_cartesian(clip = "off") + tick +
  FillScale + theme_bw() + no_grid + theme(aspect.ratio = 2) + 
  facet_layout_fix + titleSpacing + no_ribbonBGX + theme(strip.text.y = element_blank())

p1

关键调整说明 #

1. 子图边框与间距：

   • 用panel.border = element_blank()移除默认面板边框，彻底消除子图之间的重复边框
   • 用axis.line = element_line(color = "black")重新绘制坐标轴，保证图表外侧轴的完整性
   • panel.spacing = unit(-5, "pt")缩小子图间距，负数数值越大，子图越紧凑（可根据视觉效果调整）

2. 隐藏指定面板的y轴：

   • panel_number()会返回当前面板的序号（从左到右、从上到下计数）
   • 通过ifelse判断面板序号，将第2、3面板的y轴文本、刻度、轴线设为透明，实现隐藏效果

内容的提问来源于stack exchange，提问作者P.Tovi