如何将闪避点图中的成对点与 R 中的 ggplot2 连接起来？

Question

我有几个变量测量了两次，我想将它们显示在一个图表中，显示箱线图以及单个值（使用geom_dotplot ），其中来自同一变量中的两个测量值的值通过可自定义的线连接。

我找到了一个使用geom_point的解决方案（基本上生成一个新数据），但如果可能的话，我想要一个点图。

为 5 个主题和 2 个变量创建图表的简化代码

library(ggplot2)


dat <- data.frame(subject     = c(1, 2, 3, 4, 5, 1, 2, 3, 4, 5, 1, 2, 3, 4, 5, 1, 2, 3, 4, 5), 
                  measurement = c("M1", "M1", "M1", "M1", "M1", "M2", "M2", "M2", "M2", "M2",
                                  "M1", "M1", "M1", "M1", "M1", "M2", "M2", "M2", "M2", "M2"), 
                  variable    = c("Var1", "Var1", "Var1", "Var1", "Var1", "Var1", "Var1", "Var1", "Var1", "Var1",
                                  "Var2", "Var2", "Var2", "Var2", "Var2", "Var2", "Var2", "Var2", "Var2", "Var2"), 
                  value       = c(20, 15, 21, 23, 15, 22, 21, 31, 24, 17, 26, 10, 23, 22, 19, 29, 15, 25, 29, 19))


ggplot(dat, aes(variable, value, colour = measurement)) +
  geom_boxplot(outlier.shape = NA) + 
  geom_dotplot(aes(fill = measurement),
               binaxis = "y", binwidth = 0.5, stackdir = "center", 
               position = position_dodge(0.75))

在当前图表中，我希望在变量内的两个测量值之间用线连接成对的红色和青色点，如我手工绘制的 Var1 所示：

谢谢，铅

Answer 1

您可以通过使用facet_wrap并指定aes(group = subject)来实现类似的 output ，因此每条灰线都是一个单独的主题。 我提供了两个选项，一个按变量拆分，另一个按测量拆分。 第二个 plot 可能更接近你最初的想法。

library(tidyverse)

dat <- data.frame(subject     = c(1, 2, 3, 4, 5, 1, 2, 3, 4, 5, 1, 2, 3, 4, 5, 1, 2, 3, 4, 5), 
                  measurement = c("M1", "M1", "M1", "M1", "M1", "M2", "M2", "M2", "M2", "M2",
                                  "M1", "M1", "M1", "M1", "M1", "M2", "M2", "M2", "M2", "M2"), 
                  variable    = c("Var1", "Var1", "Var1", "Var1", "Var1", "Var1", "Var1", "Var1", "Var1", "Var1",
                                  "Var2", "Var2", "Var2", "Var2", "Var2", "Var2", "Var2", "Var2", "Var2", "Var2"), 
                  value       = c(20, 15, 21, 23, 15, 22, 21, 31, 24, 17, 26, 10, 23, 22, 19, 29, 15, 25, 29, 19))

# split by measurement
dat %>% 
  mutate(new = factor(interaction(measurement, variable))) %>%
  ggplot(aes(x = new, y = value, color = measurement)) + 
  geom_boxplot(outlier.shape = NA) + 
  geom_dotplot(aes(fill = measurement),
               binaxis = "y", binwidth = 0.5, stackdir = "center", 
               position = position_dodge(0.75)) +
  geom_line(aes(group = subject), color = "gray") +
  facet_wrap(~measurement, scales = "free") +
  scale_x_discrete(labels = c("Var1", "Var2")) +
  xlab("")

# split by variable
dat %>% 
  mutate(new = factor(interaction(variable, measurement))) %>%
  ggplot(aes(x = new, y = value, color = variable)) + 
  geom_boxplot(outlier.shape = NA) + 
  geom_dotplot(aes(fill = variable),
               binaxis = "y", binwidth = 0.5, stackdir = "center", 
               position = position_dodge(0.75)) +
  geom_line(aes(group = subject), color = "gray") +
  facet_wrap(~variable, scales = "free") +
  scale_x_discrete(labels = c("M1", "M2")) + 
  xlab("")