为什么 dplyr::mutate 函数给出错误答案？

如何解决为什么 dplyr::mutate 函数给出错误答案？

我不知道我的代码有什么问题。我想创建一个包含 total/sum(total 列的列。但是，它只包含每一行的值 1。 我写了这段代码：

eth_3 <- pop %>% 
  filter(ETHNICITY == 2) %>% 
  group_by(PROVINCE,ETHNICITY) %>% 
  summarise(total = sum(WEIGHT)) %>% 
  select(PROVINCE,total) %>% 
  mutate(S = total/sum(total))

得到这个结果

PROVINCE   total     S
     <int>   <dbl> <dbl>
1       11 93925.      1
2       12  2016.      1
3       13    40       1
4       14   255.      1
5       16    10       1
6       18    58.3     1

输出必须是：

   PROVINCE    total         S
      <int>    <dbl>     <dbl>
 1       11 93925.   0.968    
 2       12  2016.   0.0208   
 3       13    40    0.000412 
 4       14   255.   0.00263  
 5       16    10    0.000103 
 6       18    58.3  0.000601 
 7       19     9.67 0.0000997
 8       21    50.3  0.000519 
 9       31    34.7  0.000358 
10       32   142.   0.00147 

这是dput

structure(list(PROVINCE = c(11L,12L,13L,14L,16L,18L,19L,21L,31L,32L,33L,34L,35L,36L,52L,62L,63L,64L,74L,81L,91L),total = c(93925.4300413131,2015.98999500274,40,255.349998474121,10,58.3199987411499,9.6700000762939,50.340000152588,34.6899995803834,142.189999580384,30.0199995040892,48.5600004196165,160.789996147154,60.8100004196172,9.8800001144409,52.199997901915,21.60000038147,19.7199993133544,10.130000114441,9.8999996185303,28.1400003433227
),S = c(1,1,1)),row.names = c(NA,-21L),groups = structure(list(PROVINCE = c(11L,.rows = structure(list(1L,2L,3L,4L,5L,6L,7L,8L,9L,10L,11L,15L,17L,20L,21L),ptype = integer(0),class = c("vctrs_list_of","vctrs_vctr","list"))),class = c("tbl_df","tbl","data.frame"),.drop = TRUE),class = c("grouped_df","tbl_df","data.frame"))

解决方法

summarise 默认取消对最后一层的分组。因此，在 summarise 之后，您的数据仍按 PROVINCE 分组。在计算比例之前，您应该ungroup。

library(dplyr)

eth_3  <- pop %>% 
           filter(ETHNICITY == 2) %>% 
           group_by(PROVINCE,ETHNICITY) %>% 
           summarise(total = sum(WEIGHT)) %>% 
           select(PROVINCE,total) %>%
           ungroup %>%
           mutate(S = total/sum(total))
           #mutate(S = prop.table(total))

如果您有 dplyr > 1.0.0，您可以指定 .groups = 'drop' 而不是使用 ungroup。

pop %>% 
  filter(ETHNICITY == 2) %>% 
  group_by(PROVINCE,ETHNICITY) %>% 
  summarise(total = sum(WEIGHT),.groups = 'drop') %>% 
  select(PROVINCE,total) %>%
  mutate(S = total/sum(total))