filter out `n-1` consecutive rows if there are at least 3 consecutive values-CodePudding

I need to filter out n-1 consecutive rows if there are at least 3 consecutive values type=Yes. For example for id=2, there are 3 continuous Yes so for this id I only keep it up to the first value =Yes.

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),
    type = c("No", "Yes", "Yes", "DK", "Yes", "No", "Yes", "Yes", "Yes", 
              "DK", "No", "No", "Yes", "Yes", "DK", "No", "Yes", "Yes", "No", 
              NA, "No", "Yes", "Yes", "Yes", "No", "Yes", "Yes", "Yes", "Yes"
              ),
    cumlsum_Yes = c(0, 1, 2, 0, 1, 0, 1, 2, 3, 0, 0, 0, 
                    1, 2, 0, 0, 1, 2, 0, 0, 0, 1, 2, 3, 0, 1, 2, 3, 4)
  )

i.e.

   id type cumlsum_Yes
1   1   No           0
2   1  Yes           1
3   1  Yes           2
4   1   DK           0
5   1  Yes           1
6   2   No           0
7   2  Yes           1
8   2  Yes           2
9   2  Yes           3
10  2   DK           0
11  3   No           0
12  3   No           0
13  3  Yes           1
14  3  Yes           2
15  3   DK           0
16  4   No           0
17  4  Yes           1
18  4  Yes           2
19  5   No           0
20  5 <NA>           0
21  6   No           0
22  6  Yes           1
23  6  Yes           2
24  6  Yes           3
25  7   No           0
26  7  Yes           1
27  7  Yes           2
28  7  Yes           3
29  7  Yes           4

The expected result looks like this:

   id type cumlsum_Yes
1   1   No           0
2   1  Yes           1
3   1  Yes           2
4   1   DK           0
5   1  Yes           1
6   2   No           0
7   2  Yes           1
8   3   No           0
9   3   No           0
10  3  Yes           1
11  3  Yes           2
12  3   DK           0
13  4   No           0
14  4  Yes           1
15  4  Yes           2
16  5   No           0
17  5 <NA>           0
18  6   No           0
19  6  Yes           1
20  7   No           0
21  7  Yes           1

CodePudding user response：

This is ugly, but it works

library(tidyverse)

out <- data.frame(id=numeric(), type=character(), cumlsum_Yes=numeric())

for(i in 1:length(unique(df$id))) {
  
  group <- df %>% 
    filter(id == i)
  
  if(any(group$cumlsum_Yes >= 3 )) {
    group <- group[1:(first(which(group$type == "Yes"))), ]
  }
  
  out <- rbind(out, group)
  
}

out

#    id type cumlsum_Yes
# 1   1   No           0
# 2   1  Yes           1
# 3   1  Yes           2
# 4   1   DK           0
# 5   1  Yes           1
# 6   2   No           0
# 7   2  Yes           1
# 8   3   No           0
# 9   3   No           0
# 10  3  Yes           1
# 11  3  Yes           2
# 12  3   DK           0
# 13  4   No           0
# 14  4  Yes           1
# 15  4  Yes           2
# 16  5   No           0
# 17  5 <NA>           0
# 18  6   No           0
# 19  6  Yes           1
# 20  7   No           0
# 21  7  Yes           1