r - 计算组内类别的比例

Question

我想计算组内项目的百分比。例如，有 2 组，每组包含 3 个水果。我想知道每组中水果的比例是多少(即每组加起来应该是 100%)。我可以使用下面的代码来实现这一点，但感觉太冗长了。任何人都可以提出任何改进或已经存在的功能来简化它吗？

library(tidyverse)
#some data
fruit <- rep(c("apples", "oranges", "bananas"), 
             times=c(3, 2, 5))
group <- rep(c(1, 2), times=c(35, 65))
df <- data.frame(fruit, group, stringsAsFactors=FALSE)

#get %'s for each fruit within each group
df2 <- df %>%
  #get numerator
  group_by(fruit, group) %>%
  summarise(`Total by fruit and group` = n()) %>%
  #get denominator
  left_join(df %>%
              group_by(group) %>%
              summarise(`Total by group` = n())) %>%
  #work out % as numerator/denominator *100
  mutate(`%`=`Total by fruit and group` / `Total by group`*100 ) %>%
  select(fruit, group, `%`) %>%
  arrange(group)

Answer 1

使用 dplyr 你可以这样做:

Reprex

• 代码

library(dplyr)

df %>% 
  group_by(group) %>% 
  count(fruit) %>% 
  mutate(freq = n / sum(n) * 100) %>% 
  select(-n)

• 输出

#> # A tibble: 6 x 3
#> # Groups:   group [2]
#>   group fruit    freq
#>   <dbl> <chr>   <dbl>
#> 1     1 apples   34.3
#> 2     1 bananas  42.9
#> 3     1 oranges  22.9
#> 4     2 apples   27.7
#> 5     2 bananas  53.8
#> 6     2 oranges  18.5

^{由 reprex package 于 2022-02-19 创建(v2.0.1)}