错误条在堆积条形图中的位置不正确 (r)

Question

我正在尝试根据之前回答的问题 Making stacked bar plot with specified error bar values in R 制作带有误差线的堆叠条形图

但是，我的误差线位置不正确。我尝试更改 SD 的顺序，这会相对于彼此移动误差线，但它们仍然没有与堆叠的条对齐。

x <- data.frame(Period = c("Mon","Tue","Wed","Thu","Fri","Sat","Sun"), Sample = c("Day","Day","Day","Day","Day","Day","Day"), "12AM-6AM" = c(23.14,16.43,16,22.71,36.86,87.14,110.71), "6AM-12PM" = c(16.14,20.86,18.43,16.71,15.14,14.29,16), "12PM-6PM" = c(26.86,23.71,25.57,23.43,35.29,38,30), "6PM-12AM" =c(35.86,34.14,31.71,36.43,45.57,44,27.14))

library(dplyr)
library(reshape2)

 mx <- melt(x,  id.vars=1:2)
 mx <- mx %>% group_by(Period) %>%
      mutate(pos = cumsum(value)) %>%
      ungroup() %>%
      mutate(sd = c( 5.4, 2.7, 4.7, 4.4, 8.2, 13.2, 20.7,
                    5.6, 2.3, 5.4, 5.5, 1.6, 4.1, 3.1,
                   5.2, 5.6, 5.9, 3.5, 6.3, 4.5, 6.9,
                   3.5, 6.0, 5.9, 6.2, 8.2, 9.0, 2.4
                   ),
             upper = pos + sd/2,
             lower = pos - sd/2)

    days<-c("Mon", "Tue", "Wed","Thu","Fri", "Sat", "Sun")
    ggplot(mx, aes(x=Period, y=value, fill=variable), xLabels=NA) +
      geom_bar(stat="identity") +
      facet_grid(~Sample) + geom_errorbar(aes(ymin = lower, ymax = upper),
 width = .2, col = "red") +
      theme_bw() + scale_x_discrete(limits = days) +
      xlab(" Day of the Week") + scale_fill_grey() +
      ylab ("Number of calls")

另外，我对 R 非常陌生，甚至对堆栈溢出也很陌生（这是我的第一个问题！）- 所以任何关于如何更清晰地格式化问题或更有效地在网站上搜索答案的建议，总是值得赞赏的:)

Answer 1

我认为这是因为 variable 在 y 轴上从上到下绘制，但 cumsum(value) 中的总和是假设它们从下到上绘制的。

我在求和之前对数据框进行了重新排序，然后它就起作用了：

x <- data.frame(Period = c("Mon","Tue","Wed","Thu","Fri","Sat","Sun"), Sample = c("Day","Day","Day","Day","Day","Day","Day"), "12AM-6AM" = c(23.14,16.43,16,22.71,36.86,87.14,110.71), "6AM-12PM" = c(16.14,20.86,18.43,16.71,15.14,14.29,16), "12PM-6PM" = c(26.86,23.71,25.57,23.43,35.29,38,30), "6PM-12AM" =c(35.86,34.14,31.71,36.43,45.57,44,27.14))

library(tidyverse)
library(reshape2)
#> 
#> Attaching package: 'reshape2'
#> The following object is masked from 'package:tidyr':
#> 
#>     smiths

mx <- melt(x,  id.vars=1:2)
mx <- mx %>% 
  mutate(sd = c( 5.4, 2.7, 4.7, 4.4, 8.2, 13.2, 20.7,
                 5.6, 2.3, 5.4, 5.5, 1.6, 4.1, 3.1,
                 5.2, 5.6, 5.9, 3.5, 6.3, 4.5, 6.9,
                 3.5, 6.0, 5.9, 6.2, 8.2, 9.0, 2.4)
  ) %>%
  group_by(Period) %>%
  arrange(desc(variable)) %>%
  mutate(
    pos = cumsum(value),
    upper = pos + sd/2,
    lower = pos - sd/2
  ) %>%
  ungroup()

days<-c("Mon", "Tue", "Wed","Thu","Fri", "Sat", "Sun")
ggplot(mx, aes(x=Period, y=value, fill=variable), xLabels=NA) +
  geom_bar(stat="identity") +
  facet_grid(~Sample) + geom_errorbar(aes(ymin = lower, ymax = upper),
                                      width = .2, col = "red") +
  theme_bw() + scale_x_discrete(limits = days) +
  xlab(" Day of the Week") + scale_fill_grey() +
  ylab ("Number of calls")

^{由reprex package (v0.2.1) 于 2018 年 11 月 10 日创建}