ggplot2:在boxplot后面绘制线条

时间:2017-01-06 16:21:15

标签: r plot ggplot2

我想使用geom_line 后面我的箱线图绘制一条线,我终于设法将线条绘图与箱线图结合起来。我有这个数据集,我用它来创建一个箱线图:

>head(MdataNa)

            1           2           3           4           5           6           7
1 -0.02798634 -0.05740014 -0.02643664  0.02203644  0.02366325 -0.02868668 -0.01278713
2  0.20278229  0.19960302  0.10896017  0.24215229  0.31925211  0.29928739  0.15911725
3  0.06570653  0.08658396 -0.06019098  0.01437147  0.02078022  0.13814853  0.11369999
4 -0.42805441 -0.91945721 -1.05555731 -0.90877542 -0.77493682 -0.90620917 -1.00535742
5  0.39922939  0.12347996  0.06712451  0.07419287 -0.09517628 -0.12056720 -0.40863078
6  0.52821596  0.30827515  0.29733794  0.30555717  0.31636676  0.11592717  0.16957927

我有葡萄糖浓度,应在此框图后面的一行中绘制:

# glucose curve values
require("scales")
offconc <- c(0,0.4,0.8,1.8,3.5,6.9,7.3)
offtime <- c(9,11.4,12.9,14.9,16.7,18.3,20.5)

# now we have to scale them so they fit in the (boxplot)plot
time <- rescale(offtime, to=c(1,7))
conc <- rescale(offconc, to=c(-1,1))
glucoseConc <- data.frame(time,conc)
glucoseConc2 <- melt(glucoseConc, id = "time")

然后我绘制了这些数据,但我只能在箱图的 FRONT 中绘制葡萄糖曲线,而不是在它后面,我使用了这段代码:

boxNa <- ggplot(stack(MdataNa), aes(x = ind, y = values)) +
  geom_boxplot() + 
  coord_cartesian(y = c(-1.5,1.5)) + 
  labs(list(title = "After Loess", x = "Timepoint", y = "M")) +
  geom_line(data=glucoseConc2,aes(x=time,y=value),group=1)

以上代码的输出:

graph produced by the above code

评论建议的编辑(不工作)

boxNa <- ggplot(stack(MdataNa), aes(x = ind, y = values)) +
  geom_line(data=glucoseConc2,aes(x=time,y=value),group=1) +
  geom_boxplot(data=stack(MdataNa), aes(x = ind, y = values)) + 
  coord_cartesian(y = c(-1.5,1.5)) + 
  labs(list(title = "After Loess", x = "Timepoint", y = "M")) 

这会产生以下错误:

  

错误:提供给连续刻度的离散值

我可能做错了吗?

1 个答案:

答案 0 :(得分:0)

这是一个解决方案。 我们的想法是将x轴转换为连续值:

ggplot() +
  geom_line(data=glucoseConc2,aes(x=time,y=value),group=1)+
  geom_boxplot(data=stack(MdataNA), aes(x = as.numeric(ind), y = values, group=ind)) + 
  coord_cartesian(y = c(-1.5,1.5)) + 
  labs(list(title = "After Loess", x = "Timepoint", y = "M"))+
  scale_x_continuous(breaks=1:7)