Question

我有一个包含3个变量的长格式数据集。我正在使用ggplot2绘制两个变量和另一个变分面。我也想从每个方面绘制观察的标准误差条，但我不知道如何。有人知道吗？

这是我所拥有的图片。我想在每个方面都有标准误差条。谢谢！

编辑：这是一些示例数据和情节。

data <- data.frame(rep(c("1","2","3","4","5","6","7","8","9","10",
                         "11","12","13","14","15","16","17","18","19","20",
                         "21","22","23","24","25","26","27","28","29","30",
                         "31","32"), 2),
                   rep(c("a","b","c","d","e","f","g","h","i","j","k","l"), 32),
                   rnorm(n = 384))

colnames(data) <- c("estado","sector","VA")

ggplot(data, aes(x = estado, y = VA, col = sector)) + 
  facet_grid(.~sector) + 
  geom_point()

Answer 1

如果您想要的只是平均值＆amp;标准误差条与每个＆＃34; estado＆＃34; - ＆＃34;扇区＆＃34;组合，你可以让ggplot完成所有工作，将geom_point()行改为stat_summary()：

ggplot(data,
       aes(x = estado, y = VA, col = sector)) +
  facet_grid(. ~ sector) + 
  stat_summary(fun.data = mean_se)

有关该功能的更多详细信息，请参阅ggplot2包中的?mean_se。默认参数选项为您提供高于＆amp;的1个标准误差的平均值和范围。低于平均值。

如果您想显示原始点，只需添加geom_point()行即可。（虽然我认为这个情节对于读者而言会相当混乱，在这种情况下......）

Answer 2

也许你可以试试下面的东西？

set.seed(1)
library(dplyr)
dat = data.frame(estado = factor(rep(1:32, 2)), 
                 sector = rep(letters[1:12], 32), 
                 VA = rnorm(384))
se = function(x) {
    sd(x)/sqrt(length(x))
}
dat_sum = dat %>% group_by(estado, sector) %>%
    summarise(mu = mean(VA), se = se(VA))
dat_plot = full_join(dat, dat_sum)
ggplot(dat_plot, aes(estado, y = VA, color = sector)) + 
    geom_jitter() + 
    geom_errorbar(aes(estado, y = mu, color = sector, 
                      ymin = mu - se, ymax = mu + se)) + 
    facet_grid(.~sector)

绘制标准误差条

2 个答案: