具有大型数据集的R中的堆栈图

Question

我有一个相对较大的数据集，我正在尝试将其绘制到R中。这就是我的数据集的片段的样子

> dat
    Who      Activity Duration
1    Ch    Stationary       14
2    Ch    Stationary       18
3    Ch   Interaction        2
4    Ch    Stationary        6
5    Ch   Interaction        1
6    Ch       Display       10
7    Ch   Interaction        6
8    Ch    Stationary        5
9    Ch    Stationary       20
10   Ch    Stationary       13
11   Ch    Stationary       17
12   Co   Interaction        3
13   Co    Stationary       31
14   J    Stationary        8
15   R           OOS        1
16   R   Interaction        1
17   J           OOS        4

我正在尝试使用持续时间而不是原始值作为百分比来构建堆积的条形图。到目前为止，这是我想出的：

p7 <- ggplot() + 
  geom_bar(aes(y= dat$Duration, x= dat$Who, fill = dat$Activity), 
           stat ="identity")
p7 <- p7 + scale_fill_brewer(palette="Set3")
p7 <- p7 + xlab("Chimp ID") + ylab ("Total Time Spent (min)")
p7 <- p7 + labs(fill = "Behaviour")
p7 + theme_classic()
p7 #Palette graph with percentage values
p7 + 
  geom_bar(position = "fill", stat="identity") + 
  scale_y_continuous(labels = scales::percent_format())

但是，这仍然不会返回总计100％的小节。

Answer 1

你好索非亚，

问题在于scales::percent_format()仅使用百分号格式化数字。您需要自己计算百分比。

下面的代码为您计算出来。

library(tidyverse)

dat <-  read.delim(text="id,Who,Activity,Duration
1,Ch,Stationary,14
2,Ch,Stationary,18
3,Ch,Interaction,2
4,Ch,Stationary,6
5,Ch,Interaction,1
6,Ch,Display,10
7,Ch,Interaction,6
8,Ch,Stationary,5
9,Ch,Stationary,20
10,Ch,Stationary,13
11,Ch,Stationary,17
12,Co,Interaction,3
13,Co,Stationary,31
14,J,Stationary,8
15,R,OOS,1
16,R,Interaction,1
17,J,OOS,4", sep=",")

dat <- dat%>%
  group_by(Who)%>%
  mutate(percent = Duration/sum(Duration))

p7 <- ggplot() + 
  geom_bar(aes(y= dat$percent, x= dat$Who, fill = dat$Activity), 
           stat ="identity")
p7 <- p7 + scale_fill_brewer(palette="Set3")
p7 <- p7 + xlab("Chimp ID") + ylab ("Total Time Spent (min)")
p7 <- p7 + labs(fill = "Behaviour")
p7 + theme_classic()
p7 #Palette graph with percentage values
p7 + 
  geom_bar(position = "fill", stat="identity") + 
  scale_y_continuous(labels = scales::percent_format())