更改 ggplot geom_bar 中 NA 值的标签和颜色？

Question

我有一个这样的情节：

data <- data.frame(time = c("time1", "time2", "time1"),
                   sex = c("m", "f", NA))

ggplot(data = data) +
  geom_bar(aes(x = time, y = (..count..)/sum(..count..),
               fill = sex),
           position = "fill") +
  scale_y_continuous(name = "percentation",
                     labels = scales::percent)

我正在尝试更改 NA 值的标签和颜色。我用过scale_fill_manual像添加

  scale_fill_manual(labels = c("m" = "male",
                               "f" = "female",
                               "NA" = "unknown"),
                    values = c("m" = "blue",
                               "f" = "red",
                               "NA" = "green"))

不幸的是，这不起作用，所以我想问一下，这里有人可以帮助我解决这个问题吗？是否有特殊的词或其他东西来选择 NA 值？另外，我想保留直接分配，而"m" = "male"不是仅仅通过编写"male"来控制分配。

Answer 1

NA虽然可以使用颜色/填充比例参数轻松调整值的颜色na.value = ，但指南中的标签需要更多的工作。然而，使用尺度函数参数的非常通用的函数形式，仍然可以在不修改数据的情况下轻松实现：labels

library(ggplot2)
data <- data.frame(time = c("time1", "time2", "time1"), sex = c("m", "f", NA))

ggplot(data = data) +
  geom_bar(aes(x = time, y = after_stat(prop), fill = sex), position = "fill") +
  scale_y_continuous(labels = scales::percent) +
  scale_fill_discrete(
    labels = function(breaks) {breaks[is.na(breaks)] <- "unknown"; breaks},
    na.value = "green"
  )

^{使用reprex 包(v2.0.1)创建于 2023-03-02}