如何在ggplot箱线图中按period设置异常点颜色且不新增图例

直接修改geom_boxplot的映射并隐藏多余图例即可实现需求，以下是两种方案：

方案一：使用默认配色 #

ggplot(dat_bee_merged, aes(x = factor(region), y = colony_loss_proportion)) +
  geom_boxplot(aes(fill = factor(period), outlier.colour = factor(period))) +
  geom_hline(yintercept = median(dat_bee_merged$colony_loss_proportion), color = "red", linetype = "dashed", size = 1) +
  labs(x = "Climatic Regions", y = "Normalized colony loss", fill = "period") +
  ggtitle("Honeybee colony loss across different climatic regions and periods") +
  guides(color = "none") +
  theme_minimal()

关键改动说明： #

• 在geom_boxplot的aes中添加outlier.colour = factor(period)，让异常点颜色与箱线图填充色绑定
• 用guides(color = "none")隐藏异常点颜色生成的多余图例，只保留填充色的周期图例

方案二：自定义配色（可选） #

如果需要指定特定颜色，可同时设置填充色和异常点颜色为同一套自定义值：

# 自定义周期配色，替换为你需要的颜色代码或名称
custom_palette <- c("period_1" = "#2E86AB", "period_2" = "#F26430")

ggplot(dat_bee_merged, aes(x = factor(region), y = colony_loss_proportion)) +
  geom_boxplot(aes(fill = factor(period), outlier.colour = factor(period))) +
  geom_hline(yintercept = median(dat_bee_merged$colony_loss_proportion), color = "red", linetype = "dashed", size = 1) +
  labs(x = "Climatic Regions", y = "Normalized colony loss", fill = "period") +
  ggtitle("Honeybee colony loss across different climatic regions and periods") +
  scale_fill_manual(values = custom_palette) +
  scale_color_manual(values = custom_palette) +
  guides(color = "none") +
  theme_minimal()

内容的提问来源于stack exchange，提问作者sikimimi