如何将 anti_join 与两个变量的不同级别一起使用？

Question

我已经尝试了几个小时，但我无法弄清楚。我有一个包含主题和条件 df1 的数据框，我想从中排除具有特定值的观察值（来自 df2 的变量“值”中小于 3）。我无法使其工作，因为我需要从 df1 中删除两个变量的不同级别的组合。

这是 df1:

df1 <- structure(list(subject = c(1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L,2L, 2L, 2L, 2L, 
                                  2L, 2L, 2L, 2L, 2L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L), 
                      condition = c("A", "A", "A", "B", "B", "B", "C", "C","C", "A", "A", 
                                    "A", "B", "B", "B", "C", "C", "C", "A", "A", "A","B", "B", "B", "C", "C", "C")), 
                 row.names = c(NA, -27L), class = c("tbl_df", "tbl", "data.frame"))

这是 df2

df2 <- structure(list(subject = c(1L, 1L, 1L, 2L, 2L, 2L, 3L, 3L, 3L,4L, 4L, 4L, 5L, 5L, 5L), 
                      condition = c("A", "B", "C", "A", "B","C", "A", "B", "C", "A", "B", "C", "A", "B", "C"), 
                      value = c(10L, 8L, 7L, 3L, 8L, 5L, 3L, 3L, 9L, 8L, 7L, 8L, 10L, 6L, 2L)), 
                 row.names = c(NA,-15L), class = c("tbl_df", "tbl", "data.frame"))

而且我想在 df1 中删除所有值小于 3 的主题和条件的组合，所以这将是最终的 df:

df3 <- structure(list(subject = c(2L, 3L, 3L, 5L), 
                      condition = c("A","A", "B", "C")), 
                 row.names = c(NA, -4L), 
                 class = c("tbl_df","tbl", "data.frame"))

到目前为止我一直这样做，但我不能再这样做了，因为我有数百行...

df3 <- df1 %>% filter(!(subject==2 & condition=="A" |
                        subject==3 & (condition=="A" | condition=="B") |
                        subject==5 & condition=="C"))

Answer 1

您 df3 的示例结果与您用来派生它的代码冲突，因此这里有一个 dplyr 解决方案，用于对 df3 的每种解释。

注意：这两种结果只有当你

...exclude observations which have a certain value (less than [or equal to] 3 in the variable "value" from df2.

所以我使用不等式 <= 3 而不是 < 3 来实现这些解决方案。

`df3`

的第 1 次解读

获取df3的版本

# A tibble: 4 x 2
  subject condition
    <int> <chr>    
1       2 A        
2       3 A        
3       3 B        
4       5 C

您在此处提供的示例结果

And I want to remove in df1 all the combinations of subject and condition with a value under 3 so this would be the final df:

df3 <- structure(list(subject = c(2L, 3L, 3L, 5L), 
                      condition = c("A","A", "B", "C")), 
                 row.names = c(NA, -4L), 
                 class = c("tbl_df","tbl", "data.frame"))

只需在 df2 上使用 filter():

library(dplyr)


# ...
# Code to generate 'df1' and 'df2'.
# ...

df3 <- df2 %>% filter(value <= 3)

`df3`

的二次解读

不过，看来您实际上想要以下版本的df3

# A tibble: 18 x 2
   subject condition
     <int> <chr>    
 1       1 A        
 2       1 A        
 3       1 A        
 4       1 B        
 5       1 B        
 6       1 B        
 7       1 C        
 8       1 C        
 9       1 C        
10       2 B        
11       2 B        
12       2 B        
13       2 C        
14       2 C        
15       2 C        
16       3 C        
17       3 C        
18       3 C

你在这里得到的：

df3 <- df1 %>% filter(!(subject==2 & condition=="A" |
                        subject==3 & (condition=="A" |condition=="B") |
                        subject==5 & condition=="C"))

在的情况下，你应该 anti_join() 你的 df1 到 df2 的 filter()ed 版本：

library(dplyr)


# ...
# Code to generate 'df1' and 'df2'.
# ...


df3 <- df1 %>%
  anti_join(df2 %>% filter(value <= 3), by = c("subject", "condition"))

如何将 anti_join 与两个变量的不同级别一起使用？

How to use anti_join with different levels of two variables?

r

anti-join

tidyverse

`df3`

`df3`