I'm attempting to create a new column smk_R from the data I have. For each ID variable, I have data of two types over time. Type 1 data are my anchors and will be kept to use for later analysis. The information in Type 0 rows is also important and should be pushed to the next Type 1 row later in time within each ID. Essentially, I am looking to see if people smoked a cigarette between two Type 1 assessments (smk=0 for no and smk=1 for yes). If they did, the next Type 1 assessment should indicate smk_R=1 even if smk=0 at that specific Type 1 assessment. Any thoughts on how to do this would be much appreciated. I don't have the variable grp in my data but if that can be created from dat1, I think I can take the max of smk within group to get smk_R.
ID<-c(5,5,5,5,5,5,5,5,5,5,5,5,5,5,9,9,9,9,9,9,9,9,9,9,9,9,9,9)
time<-c(0.16,0.35,0.72,1.17,1.19,1.19,1.65,1.99,2.2,2.37,2.78,3.57,3.88,4.12,0.29,0.35,0.79,1.17,1.29,1.29,1.75,1.96,2.27,2.57,2.78,3.57,4.88,5.12)
type<-c(0,1,0,1,0,1,0,0,0,0,0,1,1,1,0,1,0,1,0,1,0,0,0,0,0,1,1,1)
smk<-c(1,0,0,0,0,1,1,1,1,0,0,0,0,0,1,0,0,0,0,1,0,0,0,0,0,0,1,1)
grp<-c(1,1,2,2,3,3,4,4,4,4,4,4,5,6,1,1,2,2,3,3,4,4,4,4,4,4,5,6)
smk_R<-c(1,1,0,0,1,1,1,1,1,1,1,1,0,1,1,1,0,0,1,1,0,0,0,0,0,0,1,1)
dat1<-cbind.data.frame(ID,time,type,smk)
dat1
ID time type smk
1 5 0.16 0 1
2 5 0.35 1 0
3 5 0.72 0 0
4 5 1.17 1 0
5 5 1.19 0 0
6 5 1.19 1 1
7 5 1.65 0 1
8 5 1.99 0 1
9 5 2.20 0 1
10 5 2.37 0 0
11 5 2.78 0 0
12 5 3.57 1 0
13 5 3.88 1 0
14 5 4.12 1 0
15 9 0.29 0 1
16 9 0.35 1 0
17 9 0.79 0 0
18 9 1.17 1 0
19 9 1.29 0 0
20 9 1.29 1 1
21 9 1.75 0 0
22 9 1.96 0 0
23 9 2.27 0 0
24 9 2.57 0 0
25 9 2.78 0 0
26 9 3.57 1 0
27 9 4.88 1 1
28 9 5.12 1 1
dat2<-cbind.data.frame(dat1,grp,smk_R)
dat2
ID time type smk grp smk_R
1 5 0.16 0 1 1 1
2 5 0.35 1 0 1 1
3 5 0.72 0 0 2 0
4 5 1.17 1 0 2 0
5 5 1.19 0 0 3 1
6 5 1.19 1 1 3 1
7 5 1.65 0 1 4 1
8 5 1.99 0 1 4 1
9 5 2.20 0 1 4 1
10 5 2.37 0 0 4 1
11 5 2.78 0 0 4 1
12 5 3.57 1 0 4 1
13 5 3.88 1 0 5 0
14 5 4.12 1 0 6 1
15 9 0.29 0 1 1 1
16 9 0.35 1 0 1 1
17 9 0.79 0 0 2 0
18 9 1.17 1 0 2 0
19 9 1.29 0 0 3 1
20 9 1.29 1 1 3 1
21 9 1.75 0 0 4 0
22 9 1.96 0 0 4 0
23 9 2.27 0 0 4 0
24 9 2.57 0 0 4 0
25 9 2.78 0 0 4 0
26 9 3.57 1 0 4 0
27 9 4.88 1 1 5 1
28 9 5.12 1 1 6 1
The addition in your comment looks like a good approach. Then you could do (for example):
library(dplyr)
dat2 <- dat1 %>%
arrange(ID, time, type) %>%
group_by(ID) %>%
mutate(grp = cumsum(c(1, type[-n()]))) %>%
group_by(ID, grp) %>%
mutate(smk_R = max(smk))
as.data.frame(dat2)