r - ggplot2 - 对重叠图使用两种不同的色标

Question

我试图覆盖两个不同的图。一个是geom_boxplot，另一个geom_jitter。我希望每个都有自己的色标。但是当我添加第二个色阶时，我得到了错误

 "Scale for 'fill' is already present. Adding another scale for 'fill', 
  which will replace the existing scale."

我假设我做错了什么。任何建议将不胜感激

这是我的工作代码的粗略示例：

P <-  ggplot(dat) + 
          geom_boxplot(aes(x=ve, y=metValue, fill=metric), alpha=.35, w=0.6, notch=FALSE, na.rm = TRUE) + 
          scale_fill_manual(values=cpalette1) + 
          geom_hline(yintercept=0, colour="#DD4466", linetype = "longdash") +
          theme(legend.position="none")

P + geom_jitter(dat2, aes(x=ve, y=metValue, fill=atd), 
                size=2, shape=4, alpha = 0.4, 
                position = position_jitter(width = .03, height=0.03), na.rm = TRUE) + 
              scale_fill_manual(values=cpalette2)

dat并dat2具有相同的架构，但不同的值。

我发现了几个解决重叠图的例子，但没有一个似乎解决了这个特定问题。

Answer 1

First, made two sample data frames with the same names as in example.

dat<-data.frame(ve=rep(c("FF","GG"),times=50),
                metValue=rnorm(100),metric=rep(c("A","B","D","C"),each=25),
                atd=rep(c("HH","GG"),times=50))
dat2<-data.frame(ve=rep(c("FF","GG"),times=50),
                metValue=rnorm(100),metric=rep(c("A","B","D","C"),each=25),
                atd=rep(c("HH","GG"),times=50))

I assume that you do not need to use argument fill= in the geom_jitter() because color for shape=4 can be set also with colour= argument. Then you can use scale_colour_manual() to set your values. Instead of cpallete just used names of colors.

P <-  ggplot(dat) + 
  geom_boxplot(aes(x=ve, y=metValue, fill=metric), alpha=.35, w=0.6, notch=FALSE, na.rm = TRUE) +  
  geom_hline(yintercept=0, colour="#DD4466", linetype = "longdash") +
  scale_fill_manual(values=c("red","blue","green","yellow"))+
  theme(legend.position="none")

P + geom_jitter(data=dat2, aes(x=ve, y=metValue, colour=atd), 
                size=2, shape=4, alpha = 0.4, 
                position = position_jitter(width = .03, height=0.03), na.rm = TRUE) + 
                scale_colour_manual(values=c("red","blue"))