如何在Base R绘图中为不同菌株分配调色板并按cit分组着色

需求与问题 #

需要绘制散点图：

• X轴：elongation_time，Y轴：fluorescence
• 数据按strain（菌株）和cit两个维度分组
• 不同strain使用专属调色板，每个cit在对应调色板中拥有唯一颜色
• 已将不同strain数据拆分到NMmedians和cit20medians两个数据框，但设置col参数时出现问题：
  • 使用as.factor(cit20medians$cit)无法切换到目标调色板
  • 直接使用darkcols无法与cit分组一一对应

原尝试代码：

#defining my two palettes:
darkcols <- c(cit0 = "#1548E0", cit3 = "#008F00", cit6 = "#CCAD00", cit9 = "#B85C00", citMax =  "#B80000", citMin = "#8F008F") #for normalized
cols  <-    c(cit0 = "royalblue2", cit3 = "green3", cit6 = "gold1", cit9 = "darkorange2", citMax = "red2", citMin = "magenta3") #for WT

palette = cols

pdf("20citvsNMwt.pdf", width = 2, height = 1.67, pointsize = 7, useDingbats = F, bg = "white" )
# sets margin stuff:
par( mex = 0.65 ) 
par( mar = c(7,6.5,4,3) )
par( oma = c(0,0.5,1,0) )

#first plotting one strain (works as intended)
plot( NMmedians$elongation_time, NMmedians$ratio, 
      col = as.factor(NMmedians$cit),
      pch = 20,
      #cex = 0.6,
      axes = F,
      xlim = c(150,400),
      ylim = c(0,1),
      xlab = "",
      ylab = "fluorescence"
)

#now trying to plot the other strain, the col parameter is where I've been focusing my efforts:
points(cit20medians$elongation_time, cit20medians$ratio, 
 pch = 20, 
 col = list(darkcols, as.factor(cit20medians$cit))
 )

axis( 1 )
axis( 2 )
title( xlab = "elongation time", line = 4.5 )

dev.off()

解决方案 #

核心思路：利用命名调色板向量的特性，通过cit列的取值直接索引对应调色板中的颜色，确保每个cit分组匹配到正确的颜色。

修正后的代码：

# 定义两个命名调色板
darkcols <- c(cit0 = "#1548E0", cit3 = "#008F00", cit6 = "#CCAD00", cit9 = "#B85C00", citMax = "#B80000", citMin = "#8F008F") # 对应normalized菌株
cols <- c(cit0 = "royalblue2", cit3 = "green3", cit6 = "gold1", cit9 = "darkorange2", citMax = "red2", citMin = "magenta3") # 对应WT菌株

pdf("20citvsNMwt.pdf", width = 2, height = 1.67, pointsize = 7, useDingbats = F, bg = "white" )
# 设置绘图边距
par( mex = 0.65 ) 
par( mar = c(7,6.5,4,3) )
par( oma = c(0,0.5,1,0) )

# 绘制第一个菌株：通过cit值索引cols调色板
plot( NMmedians$elongation_time, NMmedians$ratio, 
      col = cols[as.character(NMmedians$cit)],  # 关键：用cit字符串匹配调色板的命名
      pch = 20,
      axes = F,
      xlim = c(150,400),
      ylim = c(0,1),
      xlab = "",
      ylab = "fluorescence"
)

# 绘制第二个菌株：通过cit值索引darkcols调色板
points(cit20medians$elongation_time, cit20medians$ratio, 
       pch = 20, 
       col = darkcols[as.character(cit20medians$cit)]  # 同样用cit字符串匹配命名调色板
)

axis( 1 )
axis( 2 )
title( xlab = "elongation time", line = 4.5 )

dev.off()

关键说明 #

1. 原代码问题：
   • as.factor(NMmedians$cit)会把cit转换为整数水平，默认调用当前全局palette的对应位置颜色，无法自动切换到指定调色板
   • 直接使用darkcols会将整个颜色向量重复填充，无法与每个数据点的cit分组一一对应
2. 修正逻辑：
   • 调色板是命名向量，每个颜色的名字与cit的取值完全对应
   • 将cit列转换为字符型（as.character()），作为索引从调色板中取出对应颜色，确保每个数据点匹配到正确的分组颜色

内容的提问来源于stack exchange，提问作者Loud