带面板数据的条形图

Question

我一直在努力安静地制作一个条形图，向我展示 voted

的每个类别中有多少观察值

事情是这样的。我的面板数据框中有一个“投票”变量，数字告诉我，哪个政党在 2012 年和 2014 年投票给了哪个政党。

这是一个可重现的例子

id <- c('1','1','1','2','2','2','3','3','3')
wave <- c(2012, 2013, 2014, 2012, 2013, 2014, 2012, 2013, 2014)
voted <- c(1, NA, 4, 1, NA, 2, 2, NA, 4)
mydata <- data.frame(id, wave, voted)

仅在 2012 年和 2014 年举行过选举 - 因此是国民议会。每个数字代表党的名称。

我想在条形图中显示有多少受访者投票给政党 1、2、4...

我试过以下方法：

ggplot(mydata, aes(x = as.factor(voted), y = year, fill = year)) +
  geom_bar(stat = "identity", position = "dodge")

然而，如您所见，它看起来真的很奇怪。

ggplot

我对这样的东西更感兴趣：

How it should look

有人有什么建议吗？提前致谢！

Answer 1

这是应用我在上面评论中确定的问题的说明：

In a barplot the y value is the height of the bar which I think is "voted" in your request and the x value is the categories, probably "id" and the subcategories probably "year". I don't think you want to make the y value a factor. Furthermore "year" is not a column name and you are missing a parenthesis in the ggplot call. (You should have gotten and error with that code. –

ggplot(mydata, aes(x = id, y = voted, fill = factor(wave))) +
    geom_bar(stat = "identity", position = "dodge")
#Warning message:
#Removed 3 rows containing missing values (geom_bar).

Answer 2

您在找这样的东西吗？

library(dplyr)
library(ggplot2)

mydata %>% 
  group_by(wave) %>% 
  count(voted) %>% 
  ggplot(aes(x=factor(voted), n, fill=factor(wave)))+
  geom_col(position = position_dodge())

带面板数据的条形图

Barplot with paneldata

r

ggplot2

bar-chart

panel-data