GGplot2 在图表的两个不同部分添加中位数

Question

我正在尝试将中值绘制成单个案例研究的图表。不幸的是，我无法得到我想要的结果。这是我的代码。

Rate <- c(8, 4, 5, 5, 5, 4, 8, 7, 18, 15, 9, 8, 17, 19, 11, 11)
Phase <- c("A", "A", "A", "A", "A", "A", "A", "A", "B", "B", "B", "B", "B", "B", "B", "B")
Occasions <- c(1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16)
db4 <- data.frame(Occasions, Phase, Rate)

medians <- db4 %>% 
      group_by(Phase) %>% 
      mutate(M = median(Rate))

    ggplot(data = db4, aes(x = Occasions, y = Rate)) +
      geom_point(aes(colour = Phase)) +
      geom_line(aes(group = Phase, color = Phase)) +
      geom_hline(yintercept = medians$M, colour = 'red') +
      geom_vline(xintercept = 8.5, linetype = "dashed", color = "blue", size = 0.5) +
      xlab("Occasions") +
      ylab("Ratings")

在下面，您可以找到来自这段代码的情节。然而，两条中线目前都是红色的，并且延伸到整个地块。我想只在图表的左侧有第一条红色中线，只在图表的右侧有第二条蓝色中线。我还想为 "Baseline" 和 "Intervention"

两侧添加标题

感谢您的帮助！

Answer 1

下面是我将如何处理这个情节。我不会使用分面，因为您想将所有内容都放在一个轴上，但我也不会使用两个标题。相反，只需重新标记您的图例以明确哪些点是基线，哪些点是干预。

要固定中线并防止它们穿过整个地块，您需要使用 geom_segment 而不是 geom_hline。这也意味着您需要向 medians 数据框添加更多列以提供正确的输入（段的开始和结束坐标）。我将 color = Phase 移动到顶部 ggplot() 因此 geom-segment 继承并拆分了两个阶段。最后我让 geom_vline 使用 xintercept 的美学，所以它的位置不是硬编码的（在你的情况下可能是也可能不是理想的）。

library(tidyverse)
db4 <- data.frame(
  Rate = c(8, 4, 5, 5, 5, 4, 8, 7, 18, 15, 9, 8, 17, 19, 11, 11),
  Phase = c("A", "A", "A", "A", "A", "A", "A", "A", "B", "B", "B", "B", "B", "B", "B", "B"),
  Occasions = c(1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16)
)

medians <- db4 %>% 
  group_by(Phase) %>%
  summarise(
    y = median(Rate),
    x_start = min(Occasions),
    x_end = max(Occasions)
    )

ggplot(data = db4, aes(x = Occasions, y = Rate, color = Phase)) +
  geom_point() +
  geom_line() +
  geom_segment(
    data = medians,
    mapping = aes(x = x_start, xend = x_end, y = y, yend = y)
    ) +
  geom_vline(
    mapping = aes(xintercept = median(Occasions)),
    linetype = "dashed",
    color = "blue",
    size = 0.5
    ) +
  labs(x = "Occasions", y = "Ratings") + 
  scale_colour_hue(labels = c("Baseline", "Intervention"))

由 reprex package (v0.2.0) 创建于 2018-08-13。

Answer 2

您可以简单地使用 facet_grid 和免费的 x 刻度。

将 A 和 B 更改为 Baseline 和 Intervention 重新编码该列。

db4 <- data.frame(Occasions, Phase, Rate) %>% 
  mutate(Phase = recode(Phase, A = "Baseline", B = "Intervention"))

medians <- db4 %>% 
  group_by(Phase) %>% 
  mutate(M = median(Rate))

ggplot(data = db4, aes(x = Occasions, y = Rate)) +
  geom_point(aes(colour = Phase)) +
  geom_line(aes(group = Phase, color = Phase)) +
  geom_hline(aes(yintercept = M), colour = 'red', data = medians) +
  facet_grid(~Phase, scales = "free_x") +
  xlab("Occasions") +
  ylab("Ratings")

GGplot2 在图表的两个不同部分添加中位数

GGplot2 Add medians in two different part of the graph

plot

r

median

ggplot2