一个条形图在条形图中放错了位置（ggplot2）

Question

我正在制作一个 2 面板条形图，但似乎存在一个问题，即单个条不合适。它是右侧面板上部的栏。它被填充为 TR 但它占据了 TL 的位置。

数据是：

DATA2=structure(list(A = structure(c(1L, 1L, 1L, 2L, 2L, 2L, 2L, 2L, 
2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 
1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 
2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 
1L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 
2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 
2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 1L, 1L, 1L, 1L, 
1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 2L, 2L, 
2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 1L, 1L, 1L, 
1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 
1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 
1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 
1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 
2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 
2L, 2L, 2L, 2L, 2L, 2L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 
1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L), .Label = c("For", "Par"), class = "factor"), 
B = structure(c(2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 1L, 1L, 1L, 
2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 
2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 1L, 1L, 2L, 2L, 2L, 
2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 
2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 1L, 1L, 
2L, 2L, 1L, 1L, 2L, 1L, 1L, 1L, 2L, 2L, 1L, 2L, 2L, 2L, 2L, 
2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 1L, 1L, 1L, 
1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 
1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 
1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 
1L, 1L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 1L, 1L, 
2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 
2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 1L, 1L, 
1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 
1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 
1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 
1L, 1L, 1L, 1L), .Label = c("TL", "TR"), class = c("ordered", 
"factor")), C = structure(c(1L, 4L, 4L, 1L, 1L, 1L, 4L, 1L, 
1L, 1L, 1L, 4L, 4L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 
1L, 1L, 1L, 1L, 4L, 1L, 5L, 1L, 4L, 4L, 1L, 1L, 1L, 1L, 1L, 
1L, 1L, 4L, 1L, 1L, 1L, 1L, 1L, 1L, 5L, 1L, 1L, 1L, 1L, 1L, 
5L, 1L, 1L, 1L, 1L, 5L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 
1L, 2L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 
1L, 4L, 4L, 1L, 1L, 1L, 1L, 1L, 3L, 4L, 1L, 1L, 1L, 1L, 1L, 
1L, 1L, 1L, 5L, 4L, 1L, 4L, 5L, 1L, 1L, 1L, 1L, 1L, 3L, 5L, 
3L, 4L, 4L, 1L, 3L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 4L, 1L, 1L, 
3L, 1L, 4L, 1L, 1L, 1L, 1L, 2L, 4L, 1L, 1L, 3L, 1L, 1L, 1L, 
5L, 4L, 1L, 4L, 5L, 1L, 1L, 1L, 1L, 4L, 2L, 1L, 4L, 1L, 1L, 
1L, 1L, 2L, 4L, 5L, 1L, 1L, 5L, 1L, 1L, 1L, 1L, 2L, 1L, 1L, 
1L, 1L, 1L, 5L, 1L, 4L, 1L, 1L, 3L, 1L, 1L, 1L, 1L, 2L, 1L, 
1L, 4L, 1L, 1L, 1L, 5L, 1L, 4L, 1L, 1L, 1L, 1L, 1L, 2L, 1L, 
4L, 4L, 1L, 3L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 
1L, 1L, 1L, 1L, 4L, 1L, 3L, 1L, 1L, 1L, 3L, 5L, 3L, 1L, 1L, 
4L, 1L, 1L, 1L, 5L, 1L, 4L), .Label = c("0-2", "2-4", "4-6", 
"6-8", "8-10"), class = c("ordered", "factor"))), row.names = c(NA, 
-240L), class = "data.frame")

剧情代码如下：

ggplot(DATA2, aes(x=C,fill=B)) +
  geom_bar(position = position_dodge(width = -0.9, preserve = "single")) +
  facet_grid(.~A) +
  theme_bw() + 
  scale_fill_manual(values = c("grey","grey40")) + 
  coord_flip()

右侧面板上部的小条 (count = 1) 放错了位置，尽管填充是正确的。有小费吗？谢谢！

Answer 1

您的数据中没有显示变量值的组合，即c("Par","TL","8-10") 和c("Par", "TR", "2-4")。当您使用geom_bar(stat = "count")（这是默认值）时，ggplot2 显然不会生成计数为 0 的这些组合。对于每个配对，条形图从上向下呈现。当其中一个计数不存在时，另一个将呈现在条形配对的顶部。

我们可以看到同样的事情是这个更简单的例子：

library(dplyr)
library(tidyr)
library(ggplot2)

df <- data.frame(A = rep(c("Group 1","Group 2"),each=3),
                 B = c("possum",rep("dog",2),rep("possum",3)))

df
#>         A      B
#> 1 Group 1 possum
#> 2 Group 1    dog
#> 3 Group 1    dog
#> 4 Group 2 possum
#> 5 Group 2 possum
#> 6 Group 2 possum

ggplot(df, aes(x=A,fill=B)) +
  geom_bar(position = position_dodge(width = -0.9, preserve = "single"))

解决方法是手动计数：

df_tally <- df %>% 
  group_by(A,B) %>% 
  tally() %>% 
  ungroup()

df_tally
#>   A       B          n
#> 1 Group 1 dog        2
#> 2 Group 1 possum     1
#> 3 Group 2 possum     3

然后我们需要添加缺少的组合：

df_tally <- complete(df_tally, A, B)

df_tally
#>   A       B          n
#> 1 Group 1 dog        2
#> 2 Group 1 possum     1
#> 3 Group 2 dog       NA
#> 4 Group 2 possum     3

现在，由于我们已经完成了计数，我们使用stat = "identity"：

ggplot(df_tally, aes(x=A, y=n, fill=B)) +
  geom_bar(stat = "identity",
           position = position_dodge(width = -0.9, preserve = "single"))
#> Warning: Removed 1 rows containing missing values (geom_bar).

在你的情况下，这看起来像：

DATA3 <- DATA2 %>% 
  group_by(A,B,C) %>% 
  tally() %>% 
  ungroup() %>% 
  complete(A,B,C)


ggplot(DATA3, aes(x=C,y=n,fill=B)) +
  geom_bar(stat="identity", 
           position = position_dodge(width = -0.9, preserve = "single")) +
  facet_grid(.~A) +
  theme_bw() + 
  scale_fill_manual(values = c("grey","grey40")) + 
  coord_flip()
#> Warning: Removed 2 rows containing missing values (geom_bar).

更新：dplyr 0.8.0

从 dplyr 0.8.0 开始，group_by 具有 .drop 参数，用于维护没有任何数据的因子组。这稍微简化了必要的代码。示例：

DATA3 <- DATA2 %>% 
  group_by(A,B,C, .drop = FALSE) %>% 
  summarise(n = n())