通过匹配两列值删除观察

Question

I have the data df . I want to delete last observations after matching two column values ie, cate=Yes ~ value=1 .

df <- data.frame(id=c(1,1,1,1,1,2,2,2,2,2,3,3,3,3,3,4,4,4,5,5,6,6,6,6,7,7,7,7,7),
       cate=c('No','Yes','Yes','No','Yes','No','Yes','Yes','Yes','No','No','No','Yes','Yes',
 'No','No','Yes','Yes','No',NA,'No','Yes','Yes','Yes','No','Yes','Yes','Yes','Yes'),
                 value=c(0,0,0,0,0,0,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,1,1,0,0))
df
   id cate value
1   1   No     0
2   1  Yes     0
3   1  Yes     0
4   1   No     0
5   1  Yes     0
6   2   No     0
7   2  Yes     1
8   2  Yes     0
9   2  Yes     0
10  2   No     0
11  3   No     0
12  3   No     0
13  3  Yes     0
14  3  Yes     0
15  3   No     0
16  4   No     0
17  4  Yes     0
18  4  Yes     0
19  5   No     0
20  5  Yes     0
21  6   No     0
22  6  Yes     1
23  6  Yes     0
24  6  Yes     0
25  7   No     0
26  7  Yes     1
27  7  Yes     1
28  7  Yes     0
29  7  Yes     0

I want to delete observations per group id after matching cate=Yes and value=1 .

Then the expected output is

   id cate value
1   1   No     0
2   1  Yes     0
3   1  Yes     0
4   1   No     0
5   1  Yes     0
6   2   No     0
7   2  Yes     1
8   3   No     0
9   3   No     0
10  3  Yes     0
11  3  Yes     0
12  3   No     0
13  4   No     0
14  4  Yes     0
15  4  Yes     0
16  5   No     0
17  5  Yes     0
18  6   No     0
19  6  Yes     1
20  7   No     0
21  7  Yes     1

Answer 1

We could group by 'id', get the cumulative sum of logical expression ( cumsum ), take the cumsum again, then filter the rows where the values are less than 2 (thus it will get the full row for some 'id' that doesn't have any match and the rows till the first match if there are)

library(dplyr)
df %>% 
  group_by(id) %>% 
  filter(cumsum(cumsum(cate == 'Yes' & value == 1))<= 1) %>%
  ungroup

Answer 2

• We can use slice to select indices from 1 to the required row, taking care of NA , so we use coalesce with n() to select all rows which does not meet our condition.

library(dplyr)

df |> group_by(id) |> 
      slice(1:coalesce(which(cate == "Yes" & value == 1)[1] , n()))

• Output

# A tibble: 21 × 3
# Groups:   id [7]
      id cate  value
   <dbl> <chr> <dbl>
 1     1 No        0
 2     1 Yes       0
 3     1 Yes       0
 4     1 No        0
 5     1 Yes       0
 6     2 No        0
 7     2 Yes       1
 8     3 No        0
 9     3 No        0
10     3 Yes       0
# … with 11 more rows