2组条形离散变量

Question

我有一个数据框，其中包含2组的“ Y”或“ N”列，例如：

drug<-c("Y","Y","N","Y","Y","Y","N","N","N","N","N","Y","Y","Y","N","N")
   group<-c(0,0,0,0,0,0,0,0,1,1,1,1,1,1,1,1)                 
df<-data.frame(drug,group)

我想为两个组的“ Y” /“ N”设置条形图，两个组彼此相邻。 我用ggbarplot尝试了各种方法，并得到了奇怪的图

ggbarplot(my_matches, x = "group", y = "drug", 
                 color = "group", palette = c("#00AFBB", "#FC4E07"))

并尝试制作表格并将其绘制为类似barplot的

counts0<-df[which(df$group==0),]
counts1<-df[which(df$group==1),] 

grp0<-table(counts0$drug)
grp1<-table(counts1$drug)
s<- as.data.frame(t(rbind(grp0,grp1)))

barplot(s$grp0, s$grp1,beside=T)

您可以说，我是一个初学者，一直在疯狂地尝试解决这个问题。请帮忙！

Answer 1

首先，不需要将向量创建为数据帧列，并且df不是很好的变量名（有一个同名的函数）。像这样一步创建您的数据框：

mydata <- data.frame(drug  = c("Y","Y","N","Y","Y","Y","N","N","N","N","N","Y","Y","Y","N","N"),
                     group = c(0,0,0,0,0,0,0,0,1,1,1,1,1,1,1,1))

第二：如果您正在使用数据帧，则值得学习dplyr。因此，将其与ggplot2一起安装，然后加载：

library(dplyr)
library(ggplot2)

现在我们可以按组计算是/否：

mydata %>% 
  count(group, drug)

# A tibble: 4 x 3
  group drug      n
  <dbl> <fct> <int>
1     0 N         3
2     0 Y         5
3     1 N         5
4     1 Y         3

然后绘制计数与组的关系。我们需要将组转换为因子，因为组是一个分类变量：

mydata %>% 
  count(group, drug) %>% 
  mutate(group = factor(group)) %>% 
  ggplot(aes(group, n)) + 
  geom_col(aes(fill = drug))