用ggplot2中的横杆连接意味着答案

【问题标题】：Connect means with a crossbar in ggplot2用ggplot2中的横杆连接意味着
【发布时间】：2021-10-03 14:31:03
【问题描述】：

我有一个看起来像这样的数据框

presence = rep(c("yes","no"),each=3,2)
students = sample(seq(0,100,by=10), 12, replace = TRUE)
year = rep(c("2015","2016"),each=2,6)

data = data.frame(presence, students, year)

我这样绘制我的数据

ggplot(data, aes(as.factor(presence), students, color=factor(year))) +
  geom_jitter(size=4, alpha=0.4, shape=16,position = position_jitterdodge()) +
  stat_summary(
    aes(color = year),
    fun=mean, 
    fun.min = function(x) mean(x) - sd(x), 
    fun.max = function(x) mean(x) + sd(x),
    geom = "pointrange",  size = 0.4,
    position = position_dodge(0.8)
  )  +
  stat_summary(fun = mean, geom="line")

我真正努力做的是找到一种方法将红色的点（平均值）和蓝色的点连接起来。

除了 geom_segment 之外还有其他方法可以让我的情节看起来像这样吗？

【问题讨论】：

标签： r ggplot2 dplyr tidyverse

【解决方案1】：

这是我的尝试：

经验教训：永不放弃！

set.seed(12345)

presence = rep(c("yes","no"),each=3,2)
students = sample(seq(0,100,by=10), 12, replace = TRUE)
year = rep(c("2015","2016"),each=2,6)

data = data.frame(presence, students, year)

data1 <- data %>% 
    group_by(year, presence) %>% 
    mutate(mean = mean(students)) %>% 
    mutate(min = mean(students) - sd(students)) %>% 
    mutate(max = mean(students) + sd(students)) 


ggplot(data1, aes(as.factor(presence), students, color=factor(year))) +  
    geom_bar(alpha=0, stat = "identity", position = "dodge", color="white") +
    geom_linerange(aes(ymin = min, ymax = max), position = position_dodge(width = 0.8)) +
    #geom_point(aes(y = mean), position = position_dodge(width = 0.8)) + 
    stat_summary(fun = mean, geom="point", color= "red", position = position_dodge(width = 0.8), aes(group=factor(year))) +
    stat_summary(fun = mean, geom="line", position = position_dodge(width = 0.8), aes(group=factor(year))) +
    geom_jitter(size=4, alpha=0.4, shape=16,position = position_jitterdodge()) +
    theme_classic()

【讨论】：

【解决方案2】：

我想你用geom="line" 和group = year 再做一次stat_summary()，虽然我认为从图表中读取有点困难：

set.seed(12345)

presence = rep(c("yes","no"),each=3,2)
students = sample(seq(0,100,by=10), 12, replace = TRUE)
year = rep(c("2015","2016"),each=2,6)

data = data.frame(presence, students, year)

g = ggplot(data, aes(as.factor(presence), students, color=factor(year))) +
  geom_jitter(size=4, alpha=0.4, shape=16,position = position_jitterdodge()) +
  stat_summary(
    aes(color = year),
    fun=mean, 
    fun.min = function(x) mean(x) - sd(x), 
    fun.max = function(x) mean(x) + sd(x),
    geom = "pointrange",  size = 0.4,
    position = position_dodge(0.8)
  )  +
  stat_summary(fun = mean, geom="line")

g + stat_summary(aes(group = year),geom="line",
position=position_dodge2(0.8),fun="mean")

【讨论】：

你让我的回答蒙羞！

【解决方案3】：

您可以使用geom_line，但这需要您自己添加data。但是，我不熟悉stat_summary，因为我更喜欢自己控制流程。

所以我编辑了您的代码，看看是否可以复制它，并为您提供geom_segment 的替代方案。



data %>% 
        ggplot(
                mapping = aes(as.factor(presence), students, color=factor(year))
        )  + geom_line(
                data = data %>% group_by(year, presence) %>% summarise(students = mean(students)),
                mapping = aes(x = as.factor(presence), y = students,group = as.factor(year))
        )  + geom_point(
                data = data %>% group_by(year, presence) %>% summarise(students = mean(students)),
                mapping = aes(x = as.factor(presence), y = students,group = as.factor(year))
        ) + geom_jitter(size=4, alpha=0.4, shape=16,position = position_jitterdodge()) + geom_errorbar(
                data = data %>% group_by(year, presence) %>% summarise( sd = sd(students), students = mean(students)),
                aes(
                        ymin = students - sd,
                        ymax = students + sd
                ),
                width = 0.05
        )

【讨论】：