直接标签将标签放置在错误的位置 [英] Direct labels puts labels at wrong place

问题描述

我正在尝试使用 ggplot2 进行绘制，并尝试将标签放置在具有置信区域的平均轨迹上的正确位置.由于无法共享数据，因此创建了一个可重现的示例:

I'm trying to do a plot with ggplot2 and trying to put labels at the right place on mean trajectories with confidence regions. Since I cannot share my data, I created a reproducible example:

set.seed(456)
library(ggplot2)
library(directlabels)
time<-1:25
df2<-data.frame(time,value=runif(2500),group=as.factor(c(rep('Alpha',1250),rep('Beta',1250))))
ggplot(df2, aes(x=time, y=value,group=group,colour=group))+
      stat_summary(geom="ribbon", fun.data=mean_cl_normal,fun.args=list(conf.int=0.95), fill="grey")+
      stat_summary(geom="line", fun=mean,size=2,show.legend=TRUE)+
      stat_summary(geom="point", fun=mean, color="red",size=3)+
      theme(legend.position='none',axis.title=element_text(size=40),
        axis.text=element_text(size=30),panel.background = element_blank(),axis.line = 
        element_line(colour = 'black'))+
      geom_dl(aes(label = group),method=list(dl.combine('first.qp','last.qp')),color='black')

我得到以下图:

如您所见，标签是完全错误的.我正在尝试将标签放在第一个均值点旁边和/或最后一个均值点旁边.根据我的数据，结果更糟:两个标签都在较低的轨迹下并且都重叠.我还收到以下 warning :

As you can see, the labels are completely wrong. I'm trying to put the labels next to the first mean point and/or next to the last mean point. With my data, the result is worse: both labels are under the lower trajectory and both overlapped. I also get the following warning:

┌ Warning: RCall.jl: Warning in regularize.values(x, y, ties, missing(ties)) :
│   collapsing to unique 'x' values
│ Warning in regularize.values(x, y, ties, missing(ties)) :
│   collapsing to unique 'x' values
│ Warning in regularize.values(x, y, ties, missing(ties)) :
│   collapsing to unique 'x' values
│ Warning in regularize.values(x, y, ties, missing(ties)) :
│   collapsing to unique 'x' values

如何解决我的问题并将标签放在每个轨迹的第一个/最后一个平均值旁边?我试图更改 method 选项，但没有成功...由于绘制的点是平均值，带有 .points 的方法不起作用.而且，如何控制标签的大小?

How can I solve my problem and put the labels next to the first/last mean values of each trajectory? I tried to change the method option but with no success... As the plotted points are means, the methods with .points don't work. Moreover, how can I control the size of labels?

我尝试了另一个 ggplot2 code:

I tried another ggplot2code:

ggplot(df2, aes(x=time, y=value,group=group,colour=group))+
       stat_summary(geom="ribbon", fun.data=mean_cl_normal,
                fun.args=list(conf.int=0.95), fill="grey")+
       stat_summary(geom="line", fun=mean,size=2,show.legend=TRUE)+
       stat_summary(geom="point", fun=mean, color="red",size=3)+
       theme(legend.position = 'none',axis.title=element_text(size=40),
                axis.text=element_text(size=30),panel.background = 
                element_blank(),axis.line = element_line(colour = 
                'black'))+
       stat_summary(aes(label=group), fun=mean, geom="text", size=8)

我得到以下图:

如果我只能保留第一个和/或最后一个标签，我认为我已经接近最后一个情节了.

I think I'm close with this last plot, if I can keep only the first and/or last labels.

推荐答案

您使用的最后一种方法已经走在正确的轨道上.要使其正常工作，您必须过滤df，使其第一次和最后一次仅包含obs.试试这个:

You were already on the right track with your last approach. To make it work you have to filter your df to include only the obs for the the first and last time. Try this:

set.seed(456)
library(ggplot2)
library(dplyr)

time<-1:25
df2<-data.frame(time,value=runif(2500),group=as.factor(c(rep('Alpha',1250),rep('Beta',1250))))

dlabs <- df2 %>%
  group_by(group) %>% 
  arrange(time) %>% 
  filter(time %in% c(first(time), last(time)))

ggplot(df2, aes(x=time, y=value,group=group,colour=group))+
  stat_summary(geom="ribbon", fun.data=mean_cl_normal,fun.args=list(conf.int=0.95), fill="grey")+
  stat_summary(geom="line", fun=mean,size=2,show.legend=TRUE)+
  stat_summary(geom="point", fun=mean, color="red",size=3)+
  theme(legend.position='none',axis.title=element_text(size=40),
        axis.text=element_text(size=30),panel.background = element_blank(),axis.line = 
          element_line(colour = 'black'))+
  stat_summary(data = dlabs, aes(label = group, hjust = ifelse(time == first(time), 1.2, -.2)), fun = mean, geom = "text", color = "black") +
  scale_x_continuous(expand = expansion(mult = .1))

这篇关于直接标签将标签放置在错误的位置的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！