主页 / coding / 问题 / 79574696
Accepted
Daniel Lobo
Asked: 2025-04-15 16:32:47 +0800 CST

将ggplot中的图例分组为多个部分

  • 772

我想将图例分成ggplot多个部分。请考虑以下示例（摘自https://stackoverflow.com/a/68608592/15910619，已接受 @stefan 的答案）

library(ggplot2)
library(ggnewscale)

diamonds$cut = factor(diamonds$cut, levels=c("Fair","Good", "Very Good",
                                             "Premium","Ideal"))

labels <- levels(diamonds$cut)
labels <- setNames(labels, labels)
labels["Fair"] <- "Very Good"
labels["Good"] <- "Premium"

colors <- hcl(seq(15, 325, length.out = 5), 100, 65)
colors <- setNames(colors, levels(diamonds$cut))

ggplot() + 
  geom_bar(data = diamonds, aes(color, fill = cut)) + 
  scale_fill_manual(aesthetics = "fill", values = colors, labels = labels[1:2],
                    breaks = names(colors)[1:2], name = "First Group:",
                    guide = guide_legend(title.position = "left", order = 1)) +
  new_scale_fill() +
  geom_bar(data = diamonds, aes(color, fill = cut)) + 
  scale_fill_manual(aesthetics = "fill", values = colors, labels = labels[3:5],
                    breaks = names(colors)[3:5], name = "Second Group:",
                    guide = guide_legend(title.position = "left", order = 0)) +
  theme(legend.position = "bottom", 
        legend.direction = "horizontal",
        legend.key = element_rect(fill = "white"))

但是我想改变图例中各个元素的位置样式，如下面的修改后的快照所示

在此处输入图片描述

有没有什么方法可以实现图例中的这种分组ggplot？

非常感谢您的时间和关注。

3 个回答

  1. 如果您想要一条线来分隔两个图例，您可以利用ggh4xelement_part_rect包中的函数。

    library(ggplot2)
library(ggnewscale)
library(ggh4x)

    ... +  
scale_fill_manual(aesthetics = "fill", values = colors, labels = labels[3:5],
                    breaks = names(colors)[3:5], name = "Second Group:",
                    guide = guide_legend(title.position = "top", order = 0)) +
  theme(legend.position = "inside",
        legend.position.inside = c(0.18,0.88), # May need adjusting
        legend.spacing.y = unit(5,'pt'),
        legend.box.background=element_rect(color='white'),
        legend.background=element_part_rect(color='black', side='b'),
        legend.direction = "horizontal")

    在此处输入图片描述

    • 3

  2. 近似结果：

    library(ggplot2)
library(ggnewscale)

diamonds$cut = factor(diamonds$cut, levels=c("Fair","Good", "Very Good",
                                             "Premium","Ideal"))

labels <- levels(diamonds$cut)
labels <- setNames(labels, labels)
labels["Fair"] <- "Very Good"
labels["Good"] <- "Premium"

colors <- hcl(seq(15, 325, length.out = 5), 100, 65)
colors <- setNames(colors, levels(diamonds$cut))

ggplot() + 
  geom_bar(data = diamonds, aes(color, fill = cut)) + 
  scale_fill_manual(aesthetics = "fill", values = colors, labels = labels[1:2],
                    breaks = names(colors)[1:2], name = "First Group:",
                    guide = guide_legend(title.position = "top", order = 1)) +
  new_scale_fill() +
  geom_bar(data = diamonds, aes(color, fill = cut)) + 
  scale_fill_manual(aesthetics = "fill", values = colors, labels = labels[3:5],
                    breaks = names(colors)[3:5], name = "Second Group:",
                    guide = guide_legend(title.position = "top", order = 0)) +
  theme(legend.position = "inside",
        legend.position.inside = c(0.2,0.85),
        legend.spacing.y = unit(0,'pt'),
        legend.box.background  = element_rect(color = 'black', linewidth = 1),
        legend.direction = "horizontal",
        legend.key = element_rect(fill = "white"))

    在此处输入图片描述

    编辑1：

      ... +
theme(legend.position = "inside",
        legend.position.inside = c(0.2,0.85),
        legend.spacing.y = unit(0,'pt'),
        legend.background= ggh4x::element_part_rect(color='black', side='t', linewidth  = 1),
        legend.box.background  = element_rect(color = 'black', fill = "white", linewidth = 1),
        legend.direction = "horizontal",
        legend.key = element_rect(fill = "white"))

    在此处输入图片描述

    • 2
  3. Best Answer

    正如您在评论中提到的，您可以通过添加一个虚假类别来实现所需的结果。但是，为了使两个图例的宽度相同，我们必须为虚假类别使用与第二组图例中最后一个类别（即"Ideal"示例数据）相同的标签。为了避免标签显示在图例中，我使用以下ggtext包使标签透明：

    注意：原则上，同样的方法也适用于@Edward 的方法，以获得跨越图例框宽度的分隔线。

    library(ggplot2)
library(ggnewscale)
library(ggtext)

colors = c(colors, " " = "transparent")
labels = c(labels, " " = "<span style='color: transparent'>Ideal</span>")

ggplot() +
  geom_bar(data = diamonds, aes(color, fill = cut), show.legend = TRUE) +
  scale_fill_manual(
    aesthetics = "fill", 
    values = colors, 
    labels = labels[c(1:2, 6)],
    breaks = names(colors)[c(1:2, 6)], name = "First Group:",
    guide = guide_legend(title.position = "top", order = 1),
    limits = names(colors)[c(1:2, 6)]
  ) +
  new_scale_fill() +
  geom_bar(data = diamonds, aes(color, fill = cut)) +
  scale_fill_manual(
    aesthetics = "fill", values = colors, labels = labels[3:5],
    breaks = names(colors)[3:5], name = "Second Group:",
    guide = guide_legend(title.position = "top", order = 0)
  ) +
  theme(
    legend.position = "inside",
    legend.position.inside = c(.05, .95),
    legend.justification.inside = c(0, 1),
    legend.direction = "horizontal",
    legend.key = element_rect(fill = "white"),
    legend.background = element_rect(linewidth = .5, color = "black"),
    legend.box.background = element_rect(fill = "white", linewidth = NA),
    legend.spacing.y = unit(0, "pt"),
    legend.text = ggtext::element_markdown()
  )

    • 1

相关问题