如何使用 ggplot、geom_line 和 geom_ribbon 在布局中绘制多个 y 轴

Question

There are different data sets as bottom.

1-1.Data Set(cidf_min.csv)

name	number	value	samples	conf	lower	upper	level
apple	1	0.056008	100	0.95	0.05458	0.059141	2
apple	2	0.048256	100	0.95	0.046363	0.059142	2
apple	3	0.042819	100	0.95	0.040164	0.059143	2
apple	4	0.038663	100	0.95	0.035155	0.059144	2
apple	5	0.035325	100	0.95	0.030146	0.059145	2

1-2.Data Set(newdf_min.csv)

name	number	value	samples	conf	lower	upper	level	max
apple	2	0.01854	100	0.95	-0.06963	0.045235	2	2

 '''code''' cidf<-read.csv("D:/cidf_min.csv") newdf<-read.csv("D:/newdf_min.csv") p_min<-ggplot(cidf, aes(x=number, y=value, group=name))+geom_line(aes(color=level))+geom_ribbon(aes(ymin=lower, ymax=upper, fill=level, group=name), alpha=0.3)+geom_text(data=newdf, aes(label=name, color=level), hjust=-0.2, vjust=0.5, size=3, show.legend=F)+coord_cartesian(xlim=c(0,max(cidf$number)*1.2))+xlab(~"Con (\×"~C[max]*")")+ylab(~"score ("*mu*"C/"*mu*"F)")+theme_bw()

2-1.Data Set(cidf_max.csv)

name	number	value	samples	conf	lower	upper	level
apple	1	0.068832	100	0.95	0.061945	0.069416	2
apple	2	0.065256	100	0.95	0.053687	0.065841	2
apple	3	0.060492	100	0.95	0.046201	0.06155	2
apple	4	0.05585	100	0.95	0.039848	0.058739	2
apple	5	0.047585	100	0.95	0.033555	0.056066	2

2-2.Data Set(newdf_max.csv)

name	number	value	samples	conf	lower	upper	level	max
apple	2	0.024221	100	0.95	-0.04546	0.076362	2	2

 '''code''' cidf<-read.csv("D:/cidf_max.csv") newdf<-read.csv("D:/newdf_max.csv") p_max<-ggplot(cidf, aes(x=number, y=value, group=name))+geom_line(aes(color=level))+geom_ribbon(aes(ymin=lower, ymax=upper, fill=level, group=name), alpha=0.3)+geom_text(data=newdf, aes(label=name, color=level), hjust=-0.2, vjust=0.5, size=3, show.legend=F)+coord_cartesian(xlim=c(0,max(cidf$number)*1.2))+xlab(~"Con (\×"~C[max]*")")+ylab(~"score ("*mu*"C/"*mu*"F)")+theme_bw()

3-1.Data Set(cidf_mean.csv)

name	number	value	samples	conf	lower	upper	level
apple	1	0.069673	100	0.95	0.069673	0.069673	2
apple	2	0.06133	100	0.95	0.057955	0.062792	2
apple	3	0.060497	100	0.95	0.046201	0.06155	2
apple	4	0.054623	100	0.95	0.044241	0.058739	2
apple	5	0.039852	100	0.95	0.031906	0.043719	2

3-2.Data Set(newdf_mean.csv)

name	number	value	samples	conf	lower	upper	level	max
apple	2	0.014323	100	0.95	-0.06793	0.045717	2	2

 '''code''' cidf<-read.csv("D:/cidf_mean.csv") newdf<-read.csv("D:/newdf_mean.csv") p_mean<-ggplot(cidf, aes(x=number, y=value, group=name))+geom_line(aes(color=level))+geom_ribbon(aes(ymin=lower, ymax=upper, fill=level, group=name), alpha=0.3)+geom_text(data=newdf, aes(label=name, color=level), hjust=-0.2, vjust=0.5, size=3, show.legend=F)+coord_cartesian(xlim=c(0,max(cidf$number)*1.2))+xlab(~"Con (\×"~C[max]*")")+ylab(~"score ("*mu*"C/"*mu*"F)")+theme_bw()

I already drew 3 plots using code of ggplot , geom_line and geom_ribbon etc.

1. I want to merge plots of p_min , p_max and p_mean .

2. p_min , p_max and p_mean must locate in y axis.

3. x axis is number(1,2,3,4,5).

Let me know how to draw plots of multiple y axis using complex variables in a layout.

Answer 1

What I did I first renamed each dataset to cidf1,cidf2, and cidf3 so that they won't mix. and the same for newdf. I added a column type that basically has info of what type of graph it is. I was not sure whether you know the tidyvesre, so I used basic R to add a column using $ operator

#Set 1
cidf1 <- read.csv("cidf_min.csv")
cidf1$type ="P_min"
newdf1<-read.csv("newdf_min.csv")
newdf1$type ="P_min"
 
#Set 2
cidf2<-read.csv("cidf_max.csv")
cidf2$type ="P_max"
newdf2<-read.csv("newdf_max.csv")
newdf2$type ="P_max"

#Set 3
cidf3<-read.csv("cidf_mean.csv")
cidf3$type ="P_mean"
newdf3<-read.csv("newdf_mean.csv")
newdf3$type ="P_mean"
  

then I combined them together:

cidf = rbind(cidf1,cidf2,cidf3)
newdf =rbind(newdf1,newdf2,newdf3)

And plotted them, setting color=type to color each line according to the dataset. I removed other things that you had in your ggplot as they were not related to your question asked.

ggplot(cidf, aes(x=X, y=value))+geom_line(aes(color=type)) +
  geom_ribbon(aes(ymin=lower, ymax=upper, fill=type), alpha=0.3) +
  xlab(~"Con (\u00D7"~C[max]*")")+ylab(~"score ("*mu*"C/"*mu*"F)")+theme_bw()

So it is very close to what you are looking for. Let me know if I misinterpret what you want to do and I will update my code