R order geom_bar 基于一个级别答案

【问题标题】：R order geom_bar based on one levelR order geom_bar 基于一个级别
【发布时间】：2020-02-10 13:09:34
【问题描述】：

如何根据变量var 的第一级对条形图进行排序？现在订单是两个条的总和我想要它仅按value 排序，其中var 是"first"。下面是一个工作示例。

library("ggplot2")
ct <- rep(c("A", "B", "C", "D"),  2)
var <- c("first","first","first","first","second","second","second","second")
value <- c(1,5,0.5,8,2,11,0.2,0.1)

df <- data.frame(ct, var, value)
df

ggplot() +
  geom_bar(data = df
           , aes(x = reorder(ct, value)
                 , y = value
                 , fill = var
           )
           , stat = "identity"
           , position = "dodge"
           , width = .90 )

【问题讨论】：

标签： r ggplot2 geom-bar

【解决方案1】：

使用forcats::fct_reorder2()，您无需修改您的data.frame：

library(ggplot2)
library(forcats)

ggplot() +
  geom_bar(data = df
           , aes(x = forcats::fct_reorder2(ct, var=='first', value, .desc = F)
                 , y = value
                 , fill = var
           )
           , stat = "identity"
           , position = "dodge"
           , width = .90 ) + labs(x = "ct")

【讨论】：

【解决方案2】：

应该这样做。

df %>% # take the dataframe and then...
    group_by(ct) %>% # per group...
    mutate(order = sum(value * (var == "first"))) %>% # you calculate the sum of value where var is "first"
    ggplot(aes(reorder(ct, order) # ...and then plot everything.
               , value
               , fill = var)) +
    geom_bar(stat = "identity"
             , position = "dodge"
             , width = .9)

【讨论】：