r - 如何制作带有嵌套分组变量的堆叠条形图？

Question

我正在尝试制作一个包含两个变量的堆叠条形图。我想要的结果如下:

这是我的数据的第一部分。还有 220 行:

      Type      Week Stage
   <chr>    <dbl> <dbl>
 1 Captured     1     2
 2 Captured     1     1
 3 Captured     1     1
 4 Captured     1     2
 5 Captured     1     1
 6 Captured     1     3
 7 Captured     1    NA
 8 Captured     1     3
 9 Captured     1     2
10 Captured     1     1

到目前为止我还没有取得任何进展，这是我到目前为止的代码

library(data.table)
dat.m <- melt(newrstudio2, id.vars="Type")
dat.m

library(ggplot2)
ggplot(dat.m, aes(x=Type, y=value, fill=variable)) +
  geom_bar(stat="identity")

我想我需要计算每种类型每周每个阶段的观察次数？我已经尝试过长数据和宽数据，但我需要将周与类型结合起来？我不知道，我很茫然。

Answer 1

替代方式:

set.seed(123)
# sample data
my_data <- data.frame(Type = sample(c("W", "C"), 220, replace = TRUE),
                      Week = sample(paste0("Week ", 1:4), 220, replace = TRUE),
                      Stage = sample(paste0('S', 1:4), 220, replace = TRUE))
head(my_data)
library(ggplot2)

ggplot(my_data, aes(x = Type, fill = Stage)) + 
  geom_bar(aes(y = (..count..)/sum(..count..)), position = "fill") +
  facet_grid(. ~ Week, switch="both") + 
  scale_y_continuous(labels = scales::percent) + 
  ylab("Stage [%]") + 
  theme(strip.background = element_blank(),
        strip.placement = "outside",
        panel.spacing = unit(0, "lines"))