如何在ggplot的右上角添加包含数据的geom_text标签

Question

我有一个分组图，其中我已经用它的 x 轴值标记了峰值。但是，我想在情节的右上角有另一个标签和另一个度量，但我无法让它工作。这是一个例子。

d <- data.frame(USArrests)
nrow(d)
d$predictor <- factor(c(rep(c(1, 2), times = 25)))

label <- d %>% 
  group_by(predictor) %>%
  filter(Assault == max(Assault))
label$measure1 <- c(1.45, 5.67)
label$measure2 <- c(4.55, 6.11)

library(ggplot2)
ggplot(d, aes(x=UrbanPop, y=Assault, fill=predictor)) +
  geom_col(position=position_dodge(width = 0, preserve = "single"), width = 5) +
  geom_text(data = label, aes(label = UrbanPop)) +
  geom_text(data = label, aes(label = measure), hjust="right", vjust="top")

我希望右上角的第二个标签以红色显示“measure1 = 1.45”，然后在新行“measure2 = 4.55”，然后在其下方以绿色“measure1 = 5.67”和新行“measure2= 6.11"。显然“度量”是动态对象，所以我不想插入静态标题。非常感谢任何帮助！

Answer 1

这有点笨拙，但您可以在 data.frame 中创建完整标签，并使用\n 强制它们位于不同的行上。

d <- data.frame(USArrests)
d$predictor <- factor(c(rep(c(1, 2), times = 25)))

label <- d %>% 
  group_by(predictor) %>%
  filter(Assault == max(Assault))
label$measure1 <- c(1.45, 5.67)
label$measure2 <- c(4.55, 6.11)

label2 <- label %>%
  pivot_longer(measure1:measure2, 'measure', 'value') %>%
  mutate(label = case_when(
    predictor == 1 & measure == 'measure1' ~ paste0(measure, ' = ', value),
    predictor == 1 & measure == 'measure2' ~ paste0('\n', measure, ' = ', value),
    predictor == 2 & measure == 'measure1' ~ paste0('\n\n', measure, ' = ', value),
    predictor == 2 & measure == 'measure2' ~ paste0('\n\n\n', measure, ' = ', value)
  ))

ggplot(d, aes(x=UrbanPop, y=Assault, fill=predictor)) +
  geom_col(position=position_dodge(width = 0, preserve = "single"), width = 5) +
  geom_text(data = label, aes(label = UrbanPop)) +
  geom_text(data = label2, 
            aes(x = Inf, y = Inf, label = label, color = predictor), 
            hjust="right", vjust="top")

Answer 2

例如，使用 x 和 y = Inf。在您的示例中，分面是有意义的，因为您将标签传递给不同的数据集。

library(tidyverse)
d <- data.frame(USArrests)

d$predictor <- factor(c(rep(c(1, 2), times = 25)))

label <- d %>% 
  group_by(predictor) %>%
  filter(Assault == max(Assault))
label$measure1 <- c(1.45, 5.67)
label$measure2 <- c(4.55, 6.11)

ggplot(d, aes(x=UrbanPop, y=Assault, fill=predictor)) +
  geom_col(position=position_dodge(width = 0, preserve = "single"), width = 5) +
  geom_text(data = label, aes(label = UrbanPop)) +
  geom_text(data = label, aes(x = Inf, y = Inf, label = measure2), hjust="right", vjust="top")+
  facet_grid(~predictor)
#> Warning: position_dodge requires non-overlapping x intervals

#> Warning: position_dodge requires non-overlapping x intervals

^{由reprex package (v0.3.0) 于 2020-04-17 创建}