如何为Violin plot设置与填充色同色但透明度不同的轮廓？

你之前尝试的color = aes(fill = implant, alpha = 0.8)写法是错误的——aes()内部不能嵌套另一个aes()，且alpha不属于颜色映射的参数，无法这样设置。下面提供两种可行方案：

方案一（推荐，适用于ggplot2 3.3.0及以上版本） #

利用after_scale()函数直接继承填充色并调整透明度，无需手动匹配颜色：

# 定义ggplot对象
v1 <- ggplot(spread, aes(x = device, y = value, fill = implant)) +
  # 修改geom_violin的color参数，让轮廓继承填充色并设置透明度为0.8
  geom_violin(position = "identity", alpha = 0.4, 
              color = after_scale(alpha(fill, 0.8))) +
  facet_wrap(~ n, 
             labeller = as_labeller(function(x) paste0("n =", x)), 
             ncol = 7) +
  labs(x = "device", y = expression( ~ bar(x)^" *"),
       title = expression( "Spread and density of "~ bar(x)^" *" ~ "estimates for " ~ italic(n) ~ " repeat measures." ), 
       subtitle = expression (italic("Derived from 1,000 bootstrap estimates for each ") ~ italic(n))) +
  theme_gdocs() +
  theme(legend.position = "right", 
        plot.title = element_text(family = "arial", size = 14), 
        plot.subtitle = element_text(family = "sans", size = 10, margin = margin(t=-6)), 
        strip.text = element_text(family = "sans", size = 10, margin = margin(t=8, b=4))
  )

# 渲染图形
plot(v1)

说明：after_scale()会在颜色映射完成后修改颜色值，alpha(fill, 0.8)就是将填充色的透明度调整为0.8，直接作为轮廓颜色，保证两者色调完全一致。

方案二（适用于旧版ggplot2） #

手动提取填充色，添加透明度后通过scale_color_manual绑定到轮廓：

# 加载scales包（用于alpha函数）
library(scales)

# 获取implant变量对应的默认填充色
fill_colors <- scale_fill_discrete()$palette()(length(unique(spread$implant)))
# 给填充色添加0.8的透明度，作为轮廓色
outline_colors <- alpha(fill_colors, 0.8)

# 定义ggplot对象
v1 <- ggplot(spread, aes(x = device, y = value, fill = implant, color = implant)) +
  geom_violin(position = "identity", alpha = 0.4) +
  # 手动设置轮廓颜色匹配填充色（带透明度）
  scale_color_manual(values = outline_colors) +
  facet_wrap(~ n, 
             labeller = as_labeller(function(x) paste0("n =", x)), 
             ncol = 7) +
  labs(x = "device", y = expression( ~ bar(x)^" *"),
       title = expression( "Spread and density of "~ bar(x)^" *" ~ "estimates for " ~ italic(n) ~ " repeat measures." ), 
       subtitle = expression (italic("Derived from 1,000 bootstrap estimates for each ") ~ italic(n))) +
  theme_gdocs() +
  theme(legend.position = "right", 
        plot.title = element_text(family = "arial", size = 14), 
        plot.subtitle = element_text(family = "sans", size = 10, margin = margin(t=-6)), 
        strip.text = element_text(family = "sans", size = 10, margin = margin(t=8, b=4))
  )

# 渲染图形
plot(v1)

内容的提问来源于stack exchange，提问作者CBRF23