I have vectors that index the events A, B, C. Each event occurs at least once but can occur multiple times. For example:
c("A", "B", "C")
c("A", "A", "B", "B", "C")
c("B", "A", "C")
c("C", "B", "A")
c("A", "B", "B", "C", "A")
For each sequence, I want to identify events that occur in the following order: A-B-C. I want to index these sequences with a vector of 0/1's indicating whether or not the matches this condition. There should only be one A or C per sequence, but multiple B's between them are acceptable. When events are not in the order A-B-C, they do not match the condition. These rules should return the vectors:
c(1,1,1)
c(0,1,1,1,1)
c(0,0,0)
c(0,0,0)
c(1,1,1,1,0)
I'm not sure what general type of problem this falls under so am having trouble searching for a solution. Any suggestions would be appreciated!
Edit: open to either a base R solution applied to each vector (e.g., via lapply
) or a tidyverse solution would be more appropriate so edited my question to include both list of vectors and a data frame with input/desired output
Edit^2: added an additional test case
## sequences
s1<-c("A", "B", "C")
s2<-c("A", "A", "B", "B", "C")
s3<-c("B", "A", "C")
s4<-c("C", "B", "A")
s5<-c("A", "B", "B", "C", "A")
## make into a list
s.list <- list(s1,s2,s3,s4,s5)
## s.list as a data frame
# https://stackoverflow.com/questions/57270001/list-to-dataframe-conversion-in-r-keeping-list-indexes
s.df <- s.list %>%
purrr::map(~as_tibble(.)) %>%
dplyr::bind_rows(.id="group")
## solution...
## desired output
# as a list
s.indexed <- list(c(1,1,1),
c(0,1,1,1,1),
c(0,0,0),
c(0,0,0),
c(1,1,1,1,0)
)
s.indexed
# as a data frame
s.df <- s.df %>%
bind_cols(index = unlist(s.indexed))
s.df
This uses str_locate
to find the desired characters/events. It works for one occurrence of the event per entry. It's possible to extend this to work on any number of occurrences (by using str_locate_all
) but it would add some layers of complexity.
library(stringr)
lapply(s.list, \(x){
strng <- paste0(x, collapse="")
loc <- str_locate(strng, "AB+C")
zeros <- rep(0, nchar(strng))
if(!is.na(loc[1])){
zeros[seq(loc[1], loc[2])] <- 1
}
zeros
})
[[1]]
[1] 1 1 1
[[2]]
[1] 0 1 1 1 1
[[3]]
[1] 0 0 0
[[4]]
[1] 0 0 0
[[5]]
[1] 1 1 1 1 0