使用ggplot结合ifelse设置颜色时修复图例显示异常问题

问题背景 #

我需要绘制分面直方图，为Low和Middle水平添加均值竖线，同时根据sig字段设置竖线颜色：当sig为"Sig"时使用对应level的颜色，不显著时使用黑色。当前代码已实现核心功能，但geom_vline中的ifelse逻辑导致图例显示异常；移除该逻辑后图例恢复正常，需要修复这个矛盾问题。

原代码：

test %>%
  ggplot(aes(x = time, fill = level, color = level)) +
  geom_histogram(aes(y = after_stat(density * nrow(d))),
                 binwidth = 10,
                 position = "identity",
                 alpha = 0.25) +
  geom_vline(aes(xintercept = mean_bylevel, 
                 color = ifelse(sig == "Sig", 
                                level, "Not")),
             show.legend = FALSE) +
  facet_wrap(~condition_new) +
  scale_fill_manual(values = c("Low" = "palegreen", 
                               "Middle" = "lightpink")) +
  scale_color_manual(values = c("Low" = "forestgreen", 
                                "Middle" = "hotpink3", 
                                "Not" = "black"))

数据集：

test <- structure(list(time = c(30, 70, 20, 80, 30, 50, 50, 50, 60, 60, 
30, 50, 30, 80, 60, 70, 60, 80, 80, 60, 70, 70, 20, 60, 60, 50, 
30, 50, 80, 60, 40, 80, 30, 60, 50, 70, 40, 20, 30, 40, 80, 30, 
70, 70, 60, 20, 40, 30, 80, 60, 70, 70, 40, 60, 40, 30, 20, 70, 
60, 70, 80, 70, 40, 30, 50, 80, 20, 60, 20, 50, 30, 60, 20, 30, 
70, 20, 60, 40, 70, 70, 70, 40, 40, 80, 60, 50, 60, 30, 60, 70, 
30, 30, 20, 60, 80, 40, 60, 40, 20, 80), level = c("Low", "Low", 
"Middle", "Low", "Low", "Middle", "Middle", "Middle", "Low", 
"Low", "Middle", "Middle", "Middle", "Low", "Middle", "Low", 
"Middle", "Middle", "Low", "Middle", "Low", "Low", "Low", "Middle", 
"Middle", "Low", "Middle", "Middle", "Low", "Middle", "Low", 
"Low", "Low", "Middle", "Middle", "Low", "Low", "Low", "Middle", 
"Middle", "Middle", "Low", "Middle", "Low", "Low", "Middle", 
"Middle", "Low", "Low", "Middle", "Low", "Middle", "Middle", 
"Low", "Low", "Middle", "Middle", "Low", "Middle", "Low", "Middle", 
"Low", "Low", "Middle", "Middle", "Low", "Low", "Middle", "Low", 
"Middle", "Middle", "Low", "Low", "Middle", "Middle", "Middle", 
"Low", "Middle", "Low", "Low", "Low", "Middle", "Middle", "Low", 
"Low", "Low", "Middle", "Middle", "Low", "Middle", "Low", "Middle", 
"Low", "Middle", "Low", "Middle", "Middle", "Low", "Low", "Low"
), mean_bylevel = c(50, 49.5, 51.2, 50.4, 50, 50.1, 49.4, 51, 
50, 49.5, 49.4, 50.1, 51, 50, 51.2, 50.4, 51.2, 50.1, 50.4, 49.4, 
50, 50, 49.5, 51, 51.2, 50.4, 49.4, 51, 49.5, 50.1, 50, 50, 50, 
49.4, 50.1, 50.4, 49.5, 50, 51.2, 51, 50.1, 50, 51, 50.4, 50, 
51.2, 49.4, 49.5, 50.4, 51.2, 50, 50.1, 49.4, 49.5, 50, 51, 51.2, 
50.4, 50.1, 50, 51, 49.5, 50, 49.4, 51.2, 50, 49.5, 49.4, 50, 
51, 50.1, 50.4, 50.4, 50.1, 49.4, 51.2, 50, 51, 49.5, 50, 50, 
51.2, 49.4, 49.5, 50.4, 50, 50.1, 51, 50.4, 51.2, 50, 51, 49.5, 
49.4, 50, 50.1, 49.4, 50, 49.5, 50.4), sig = c("Not", "Not", 
"Not", "Not", "Not", "Not", "Not", "Not", "Not", "Not", "Not", 
"Not", "Not", "Not", "Not", "Not", "Not", "Not", "Not", "Not", 
"Not", "Not", "Not", "Not", "Not", "Not", "Not", "Not", "Not", 
"Not", "Not", "Not", "Not", "Not", "Not", "Not", "Not", "Not", 
"Not", "Not", "Not", "Not", "Not", "Not", "Not", "Not", "Not", 
"Not", "Not", "Not", "Not", "Not", "Not", "Not", "Not", "Not", 
"Not", "Not", "Not", "Not", "Not", "Not", "Not", "Not", "Not", 
"Not", "Not", "Not", "Not", "Not", "Not", "Not", "Not", "Not", 
"Not", "Not", "Not", "Not", "Not", "Not", "Not", "Not", "Not", 
"Not", "Not", "Not", "Not", "Not", "Not", "Not", "Not", "Not", 
"Not", "Not", "Not", "Not", "Not", "Not", "Not", "Not", "Not"
), condition_new = c("D", 
"C", "A", "B", "A", "D", "C", "B", "D", "C", "C", "D", "B", "A", 
"A", "B", "A", "D", "B", "C", "A", "D", "C", "B", "A", "B", "C", 
"B", "C", "D", "A", "D", "D", "C", "D", "B", "C", "A", "A", "B", 
"D", "D", "B", "B", "A", "A", "C", "C", "B", "A", "D", "D", "C", 
"C", "A", "B", "A", "B", "D", "D", "B", "C", "A", "C", "A", "A", 
"C", "C", "D", "B", "D", "B", "B", "D", "C", "A", "A", "B", "C", 
"D", "A", "A", "C", "C", "B", "D", "D", "B", "B", "A", "D", "B", 
"C", "C", "A", "D", "C", "D", "C", "B")), row.names = c(NA, -100L
), class = c("tbl_df", "tbl", "data.frame"))

解决方案 #

核心思路是将颜色逻辑从aes()映射中剥离，提前生成专门的颜色变量，并通过尺度设置限定图例显示范围，避免ggplot混淆颜色映射层级。

方法1：提前处理数据生成颜色变量 #

先在数据中新增vline_color列，定义竖线的颜色规则：

test <- test %>%
  mutate(vline_color = ifelse(sig == "Sig", level, "Not"))

然后修改ggplot代码，直接映射该变量，并通过scale_color_manual的limits参数只保留level相关的图例项：

test %>%
  ggplot(aes(x = time, fill = level, color = level)) +
  geom_histogram(aes(y = after_stat(density * nrow(d))),
                 binwidth = 10,
                 position = "identity",
                 alpha = 0.25) +
  geom_vline(aes(xintercept = mean_bylevel, color = vline_color),
             show.legend = FALSE) +
  facet_wrap(~condition_new) +
  scale_fill_manual(values = c("Low" = "palegreen", 
                               "Middle" = "lightpink")) +
  scale_color_manual(
    values = c("Low" = "forestgreen", 
               "Middle" = "hotpink3", 
               "Not" = "black"),
    limits = c("Low", "Middle") # 仅显示level的图例，排除"Not"
  )

方法2：临时生成颜色变量（不修改原数据） #

如果不想改动原数据集，可以在ggplot流程中临时生成变量：

test %>%
  mutate(vline_color = ifelse(sig == "Sig", level, "Not")) %>%
  ggplot(aes(x = time, fill = level, color = level)) +
  geom_histogram(aes(y = after_stat(density * nrow(d))),
                 binwidth = 10,
                 position = "identity",
                 alpha = 0.25) +
  geom_vline(aes(xintercept = mean_bylevel, color = vline_color),
             show.legend = FALSE) +
  facet_wrap(~condition_new) +
  scale_fill_manual(values = c("Low" = "palegreen", 
                               "Middle" = "lightpink")) +
  scale_color_manual(
    values = c("Low" = "forestgreen", 
               "Middle" = "hotpink3", 
               "Not" = "black"),
    limits = c("Low", "Middle")
  )

问题原因说明 #

原代码中，在aes(color = ifelse(...))里直接生成混合了level和"Not"的向量，会让ggplot将其识别为新的离散变量，扩大颜色映射范围，即使设置show.legend = FALSE，也会干扰主图层的颜色图例显示。通过提前生成变量并限定图例范围，就能让主图层的图例恢复正常。

内容的提问来源于stack exchange，提问作者a_todd12