如何去除ggplot2图例标签周围的方框？

问题场景 #

我绘制了一张包含第一分位数和第三分位数ribbon的折线图，用ggplot2绘图时，图例标签周围出现了方框。尝试theme(legend.background=element_blank())或guides(guides_legend)覆盖等方法都没法去除。

数据集与绘图代码 #

mutate(code_range = fct_reorder2(code_range, year, order)) %>% 
  ggplot(aes(x=year, y=value, group=code_range, color=code_range)) +
  geom_ribbon(aes(ymin = min_range, 
                  ymax = max_range), fill = "grey70", alpha=0.1) +
  geom_line(size=1) +
  labs(title="Title", 
       subtitle=paste0("subtitle"),
       x="",y="",
       caption="Source") +
  scale_colour_manual(values = c("min" = "#878787", "q1" = "#B5B5B5", "median" = "#27408B", 
                                 "q3" = "#B5B5B5", "max" = "#878787")) +
  scale_fill_manual(values = c("min" = "#FFFFFF", "q1" = "#B5B5B5", "median" = "#27408B", 
                               "q3" = "#B5B5B5", "max" = "#878787")) +
  theme_opts +
  theme(legend.title = element_blank(), 
        legend.position = "bottom") +
  theme(legend.background=element_blank()) 
plot 

绘图效果 #

解决方法 #

你看到的方框是单个图例项符号的背景框，legend.background控制的是整个图例区域的背景，对它无效。需要修改legend.key的设置：

在theme()中添加legend.key = element_blank()，就能直接去除这些方框。

修改后的完整代码：

mutate(code_range = fct_reorder2(code_range, year, order)) %>% 
  ggplot(aes(x=year, y=value, group=code_range, color=code_range)) +
  geom_ribbon(aes(ymin = min_range, 
                  ymax = max_range), fill = "grey70", alpha=0.1) +
  geom_line(size=1) +
  labs(title="Title", 
       subtitle=paste0("subtitle"),
       x="",y="",
       caption="Source") +
  scale_colour_manual(values = c("min" = "#878787", "q1" = "#B5B5B5", "median" = "#27408B", 
                                 "q3" = "#B5B5B5", "max" = "#878787")) +
  scale_fill_manual(values = c("min" = "#FFFFFF", "q1" = "#B5B5B5", "median" = "#27408B", 
                               "q3" = "#B5B5B5", "max" = "#878787")) +
  theme_opts +
  theme(legend.title = element_blank(), 
        legend.position = "bottom",
        legend.background = element_blank(),
        legend.key = element_blank()) # 新增此行去除方框
plot 

如果想优化图例显示（避免color和fill图例重复），可以添加guides()设置合并图例：

guides(fill = guide_legend(override.aes = list(alpha = 0.3)), 
       color = guide_legend(override.aes = list(fill = NA)))

内容的提问来源于stack exchange，提问作者user11841097