ggplot2：图例透明度调整、样式修改及顺序匹配问题

问题描述 #

我用ggplot2绘制了两组带geom_ribbon的geom_line数据，每组设置了不同颜色和透明度，但三个需求未实现：

• 调整图例中线条的透明度，使其匹配对应geom_line的alpha值，尝试guides(color=guide_legend(override.aes=list(alpha=c(0.7, 0.5))))但无效；
• 希望图例中的线条带有对应颜色及透明度的填充背景（如灰色线条搭配浅灰雾状背景）；
• 图例顺序需匹配设定的因子顺序（versicolor在前，setosa在后），但当前未生效。

复现代码 #

library(tidyverse)
library(ggplot2)

samples <- data.frame(
  Species_Number=c(2,1),
  Species=c("setosa", "versicolor")
)

samples <- samples %>%
  dplyr::mutate(Species = factor(Species, levels=c("versicolor", "setosa")))

df1 <- iris %>%
  dplyr::filter(Species=="setosa") %>%
  dplyr::mutate(Position=1:50) %>%
  dplyr::mutate(Species1_Sepal = Sepal.Width) %>%
  dplyr::mutate(Species1_Error = abs(Sepal.Width - mean(Sepal.Width))) %>%
  dplyr::select(Position, Species1_Sepal, Species1_Error)

df2 <- iris %>%
  dplyr::filter(Species=="versicolor") %>%
  dplyr::mutate(Position=1:50) %>%
  dplyr::mutate(Species2_Sepal = Sepal.Width) %>%
  dplyr::mutate(Species2_Error = abs(Sepal.Width - mean(Sepal.Width))) %>%
  dplyr::select(Position, Species2_Sepal, Species2_Error)

df_final <- df1 %>%
  dplyr::left_join(df2, by = c("Position" = "Position"))

Species1 <- samples$Species[samples$Species_Number==1]
Species2 <- samples$Species[samples$Species_Number==2]

colorscale <- c("#333333", "red")
names(colorscale) <- c(Species1, Species2)
LegendTitle <- "Species"

Plotv1 <- ggplot() +
  geom_line(data = df_final, aes(x=Position, y=Species1_Sepal, col = Species1), linewidth = .75, alpha = 0.7) +
  geom_ribbon(data = df_final, aes(x=Position, ymax=Species1_Sepal+Species1_Error, ymin=Species1_Sepal-Species1_Error, fill = Species1), linewidth = 1.2, alpha = 0.25) +
  geom_line(data = df_final, aes(x=Position, y=Species2_Sepal, col = Species2), linewidth = .75, alpha = 0.5) +
  geom_ribbon(data = df_final, aes(x=Position, ymax=Species2_Sepal+Species1_Error, ymin=Species2_Sepal-Species1_Error-0.2, fill = Species2), linewidth = 1.2, alpha = 0.25) +
  scale_fill_manual(values = colorscale) +
  scale_color_manual(values = colorscale) +
  guides(color=guide_legend(title=LegendTitle)) +
  guides(fill = "none")

因子顺序检查结果 #

#Check Factor Order
> str(samples)
'data.frame':   2 obs. of  2 variables:
 $ Species_Number: num  2 1
 $ Species       : Factor w/ 2 levels "versicolor","setosa": 2 1

> Species1
[1] versicolor
Levels: versicolor setosa

> Species2
[1] setosa
Levels: versicolor setosa

> names(colorscale)
[1] "versicolor" "setosa"

修复方案 #

核心问题在于数据为宽格式，导致ggplot无法统一映射美学属性，图例设置未精准关联分组。以下是修复后的代码，同时解决三个需求：

修复后代码 #

library(tidyverse)
library(ggplot2)

# 保留原数据处理逻辑
samples <- data.frame(
  Species_Number=c(2,1),
  Species=c("setosa", "versicolor")
) %>%
  mutate(Species = factor(Species, levels=c("versicolor", "setosa")))

df1 <- iris %>%
  filter(Species=="setosa") %>%
  mutate(Position=1:50) %>%
  mutate(Species1_Sepal = Sepal.Width) %>%
  mutate(Species1_Error = abs(Sepal.Width - mean(Sepal.Width))) %>%
  select(Position, Species1_Sepal, Species1_Error)

df2 <- iris %>%
  filter(Species=="versicolor") %>%
  mutate(Position=1:50) %>%
  mutate(Species2_Sepal = Sepal.Width) %>%
  mutate(Species2_Error = abs(Sepal.Width - mean(Sepal.Width))) %>%
  select(Position, Species2_Sepal, Species2_Error)

df_final <- df1 %>% left_join(df2, by = "Position")

Species1 <- samples$Species[samples$Species_Number==1]
Species2 <- samples$Species[samples$Species_Number==2]

# 转换为长格式数据，统一分组管理
df_long <- df_final %>%
  pivot_longer(
    cols = -Position,
    names_to = c("Group", ".value"),
    names_pattern = "(Species\\d)_(Sepal|Error)"
  ) %>%
  mutate(Species = case_when(
    Group == "Species1" ~ Species1,
    Group == "Species2" ~ Species2
  )) %>%
  mutate(Species = factor(Species, levels = levels(samples$Species))) %>%
  # 为每组设置对应的线条透明度
  mutate(line_alpha = ifelse(Species == "versicolor", 0.7, 0.5))

colorscale <- c("#333333", "red")
names(colorscale) <- c("versicolor", "setosa")
LegendTitle <- "Species"

Plotv2 <- ggplot(df_long, aes(x = Position)) +
  # 先绘制误差带，避免覆盖线条
  geom_ribbon(
    aes(ymax = Sepal + Error, ymin = Sepal - Error, fill = Species),
    alpha = 0.25
  ) +
  # 线条映射颜色和透明度
  geom_line(
    aes(y = Sepal, color = Species, alpha = line_alpha),
    linewidth = .75
  ) +
  # 指定颜色和填充值，强制图例顺序匹配因子层级
  scale_color_manual(
    values = colorscale,
    breaks = levels(df_long$Species),
    guide = "legend"
  ) +
  scale_fill_manual(
    values = colorscale,
    breaks = levels(df_long$Species),
    guide = "none"
  ) +
  # 自定义图例：匹配透明度、添加填充背景
  guides(
    color = guide_legend(
      title = LegendTitle,
      override.aes = list(
        alpha = c(0.7, 0.5),
        fill = alpha(colorscale, 0.2) # 填充背景透明度设为0.2
      ),
      keywidth = unit(2, "cm"),
      keyheight = unit(1, "cm")
    ),
    alpha = "none" # 隐藏单独的透明度图例
  )

print(Plotv2)

关键修复点 #

1. 长格式数据：通过pivot_longer合并两组数据，统一由Species分组，避免重复调用geom，让ggplot能自动关联所有美学属性；
2. 图例透明度：将线条透明度作为映射变量，或在override.aes中直接指定对应分组的alpha值，确保图例与绘图区线条透明度一致；
3. 图例填充背景：在override.aes中添加fill参数，用alpha()调整填充透明度，同时设置图例key尺寸让背景更清晰；
4. 图例顺序：通过breaks = levels(df_long$Species)强制图例顺序与因子层级完全匹配，解决顺序错乱问题。

内容的提问来源于stack exchange，提问作者JVGen