ggplot：在具有相同颜色图例的facet中订购条形图

Question

使用数据集：

library(tidyverse)
df1 <- structure(list(dept = c("A", "B", "C"), 
n = c(70908L, 50004L, 294614L)), class = c("tbl_df", "tbl", 
"data.frame"), row.names = c(NA, -3L), .Names = c("dept", "n"))

df2 <- structure(list(dept = c("A", "B", "D", "E"), 
n = c(544L, 684L, 375L, 1191L)), class = c("tbl_df", "tbl", "data.frame"), 
row.names = c(NA, -4L), .Names = c("dept", "n"))

> df1
# A tibble: 3 x 2
   dept      n
  <chr>  <int>
1     A  70908
2     B  50004
3     C 294614

> df2
# A tibble: 4 x 2
   dept     n
  <chr> <int>
1     A   544
2     B   684
3     D   375
4     E  1191

我想在具有相同填充/图例的方面创建一个有序的条形图。这是我到目前为止所能得到的：

df <- rbind(df1 %>% mutate(source = 1),
       df2 %>% mutate(source = 2))

df %>% ggplot(aes(reorder(dept, n), n, fill = dept)) +
  geom_bar(stat = "identity") +
  coord_flip() +
  facet_wrap(~source, scales = "free")

Rplot

如何使reorder功能在每个方面都有效？

Answer 1

要在两个方面中订购值，请使用它创建新列var3和reorder。在此之后，只需将scale_x_discrete中的标签从var3更改回dept。

library(tidyverse)
df <- rbind(df1 %>% mutate(source = 1),
            df2 %>% mutate(source = 2)) %>%
    mutate(var3 = paste(source, dept))
ggplot(df, aes(reorder(var3, n), n, fill = dept)) +
    geom_bar(stat = "identity") +
    coord_flip() +
    xlab("dept") +
    facet_wrap(~ source, scales = "free") +
    scale_x_discrete(breaks = df$var3, labels = df$dept)