ggplot2中多geom图层颜色统一及comparison图例移除问题

需求说明 #

已通过scale_fill_manual为stat_halfeye和geom_boxplot设置了Answerer为lemonchiffon、NotAnswerer为darkorange的填充色，需完成以下调整：

• 将geom_point和geom_path中原偏蓝色的点/线条（对应comparison=0）改为lemonchiffon
• 将原偏灰色的点/线条（对应comparison=1）改为darkorange
• 移除comparison对应的颜色图例

修改要点 #

1. 新增scale_color_manual，把comparison的两个取值映射到目标颜色
2. 在guides参数中添加color = "none"，移除comparison的图例

修改后的完整代码 #

library(ggplot2)
library(tidyquant)
library(ggdist)
library(ggthemes)

box_width <- .12
jitter_width <- .025

pj <- position_jitter(seed = 1, width = jitter_width, height = 0)


ggplot(data = df, aes(x = name, y = value, fill = name)) +
  
  # 半小提琴图
  stat_halfeye(
    aes(
      justification = ifelse(name == "Answerer", 1.2, -.2),
      side = ifelse(name == "Answerer", "left", "right")
    ),
    adjust = 0.5,
    .width = 0,
    point_colour = NA
  ) +
  
  # 箱线图
  geom_boxplot(
    width = box_width,
    alpha = 0.5
  ) +
  scale_fill_manual(values = c("Answerer" = "lemonchiffon",
                               "NotAnswerer" = "darkorange"))+
  
  # 散点
  geom_point(
    aes(x = name_num, color = comparison),
    alpha = 0.5,
    position = pj
  ) +
  
  # 连接线
  geom_path(aes(x = name_num, group = IPU_id, color = comparison),
            alpha = 0.5,
            position = pj,
  ) +
  # 新增颜色映射并移除comparison图例
  scale_color_manual(values = c("0" = "lemonchiffon", "1" = "darkorange")) +
  guides(fill = "none", color = "none")

验证数据 #

df <- structure(list(IPU_id = c(2L, 2L, 116L, 116L, 352L, 352L, 481L, 
481L, 583L, 583L, 1077L, 1077L, 1394L, 1394L, 1447L, 1447L, 1538L, 
1538L, 1982L, 1982L, 2459L, 2459L, 2729L, 2729L, 2763L, 2763L, 
2809L, 2809L, 3162L, 3162L, 3165L, 3165L, 3188L, 3188L, 3249L, 
3249L, 3454L, 3454L, 3618L, 3618L, 3780L, 3780L, 3796L, 3796L, 
3972L, 3972L, 4407L, 4407L, 4779L, 4779L), comparison = c(0L, 
0L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 0L, 0L, 1L, 1L, 1L, 1L, 1L, 
1L, 0L, 0L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 0L, 0L, 1L, 
1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 0L, 0L, 0L, 0L, 0L, 0L, 1L, 
1L), name = structure(c(1L, 2L, 1L, 2L, 1L, 2L, 1L, 2L, 1L, 2L, 
1L, 2L, 1L, 2L, 1L, 2L, 1L, 2L, 1L, 2L, 1L, 2L, 1L, 2L, 1L, 2L, 
1L, 2L, 1L, 2L, 1L, 2L, 1L, 2L, 1L, 2L, 1L, 2L, 1L, 2L, 1L, 2L, 
1L, 2L, 1L, 2L, 1L, 2L, 1L, 2L), levels = c("Answerer", "NotAnswerer"
), class = "factor"), value = c(-0.0431558441558438, 1.70782692307692, 
0.684046511627906, 0.423953124999999, 0.302268456375839, -0.309683333333333, 
0.792993421052627, -0.871427152317877, 2.55586178861788, -0.85030081300813, 
-0.241509900990097, 0.026694999999999, 0.880453020134229, -0.298657718120805, 
-0.0876920529801324, -0.250344370860927, -6.02329885057471, -8.22881976744186, 
-6.48995, 1.035525, 9.35135416666668, -2.60511643835616, 0.643900662251654, 
0.167112582781379, 0.572819078947369, -2.0765165562914, 0.462728395061728, 
-0.141359756097561, 0.76575, -1.60808333333333, -0.894208609271523, 
-0.786271523178809, 0.754248201438848, -0.160610714285716, -0.338933884297521, 
-1.21221008403362, 6.022116, -0.524599999999991, 0.425596774193548, 
0.163237012987012, 0.189037634408601, -2.02574175824177, 0.1978, 
1.16077962962963, -1.45908116883117, 0.964378289473683, 0.421305555555556, 
0.540716783216783, 0.16981683168317, -0.798830000000001), name_num = c(1.1325, 
1.8675, 1.1325, 1.8675, 1.1325, 1.8675, 1.1325, 1.8675, 1.1325, 
1.8675, 1.1325, 1.8675, 1.1325, 1.8675, 1.1325, 1.8675, 1.1325, 
1.8675, 1.1325, 1.8675, 1.1325, 1.8675, 1.1325, 1.8675, 1.1325, 
1.8675, 1.1325, 1.8675, 1.1325, 1.8675, 1.1325, 1.8675, 1.1325, 
1.8675, 1.1325, 1.8675, 1.1325, 1.8675, 1.1325, 1.8675, 1.1325, 
1.8675, 1.1325, 1.8675, 1.1325, 1.8675, 1.1325, 1.8675, 1.1325, 
1.8675)), row.names = c(NA, -50L), class = c("tbl_df", "tbl", 
"data.frame"))

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