计算 R / dplyr 中不同子集的值

Question

我有一个数据集，其中有许多由个人 ID 提交的值，这些值被组织成子集。我想为每个 ID 计算一个值 = ID 分数的平均值/子集分数的平均值。我尝试了很多使用 group_by()、summarize() 和 spread() 的选项，但无法安排。

library(dplyr)

df  <- data.frame(stringsAsFactors=FALSE,
                  Subset = c("A","B","C","D","A","B","C","D","A","B","C","D"),
                  ID = c(1,2,3,4,5,3,1,5,2,3,4,1),
                  score = c(123,42,564,234,123,345,6678,87,543,121,123,55))

averages <-
  df %>%
  group_by(Subset) %>%
  summarise(mean.subs = mean(score)) %>%
  ungroup() %>%
  group_by(ID) %>%
  summarise(mean.id = mean(score) / mean.subs)

如有任何帮助，我将不胜感激。

Answer 1

我想你想用 mutate 而不是 summarize:

library(dplyr)

df %>% 
  dplyr::group_by(Subset) %>% 
  dplyr::mutate(mean.subs = mean(score)) %>% 
  dplyr::ungroup() %>% 
  dplyr::group_by(ID) %>%
  dplyr::mutate(mean.id = mean(score)) %>%
  dplyr::rowwise() %>% 
  dplyr::mutate(value = mean.id / mean.subs)

   Subset    ID score mean.subs mean.id   value
   <chr>  <dbl> <dbl>     <dbl>   <dbl>   <dbl>
 1 A          1   123      263    2285.  8.69  
 2 B          2    42      169.    292.  1.73  
 3 C          3   564     2455     343.  0.140 
 4 D          4   234      125.    178.  1.42  
 5 A          5   123      263     105   0.399 
 6 B          3   345      169.    343.  2.03  
 7 C          1  6678     2455    2285.  0.931 
 8 D          5    87      125.    105   0.838 
 9 A          2   543      263     292.  1.11  
10 B          3   121      169.    343.  2.03  
11 C          4   123     2455     178.  0.0727
12 D          1    55      125.   2285. 18.2

此外，要跨行计算，您还需要使用 rowwise。您可能还想在末尾添加另一个管道到 ungroup 以取消组合您的输出。

计算 R / dplyr 中不同子集的值

Calculating values across different subsets in R / dplyr

r

subset

dplyr