I am trying to calculate the longest exact common substring without gaps between a string and a vector of strings in R. How do I modify stringdist to return any common string anywhere in the two compared strings and return the distance?
Reproduce data:
string1 <- "whereiam"
vec1 <- c("firstiam","twoiswhereiaminthisvec","thisisthree","fouriamhere","fivewherehere")
Attempted stringdist function tried (doesnt work for my purposes):
library(stringdist)
stringdistvec <- stringdist(string1,vec1,method="lcs")
[1] 8 14 13 11 11 #not calculating the lcs type I want
Desired result instead with explanation of matches:
#desired to work to get this result:
desired_stringdistvec <- c(3,8,1,3,5)
[1] 3 8 1 3 5
#match 1: iam (3 common substr)
#match 2: whereiam (8 common substr)
#match 3: i (one letter only)
#match 5: iam (3 common substr)
#match 6: where (5 common substr)
One approach might be to look at the transformation sequence produced by adist() and count the characters in the longest contiguous match:
trafos <- attr(adist(string1, vec1, counts = TRUE), "trafos")
sapply(gregexpr("M+", trafos), function(x) max(0, attr(x, "match.length")))
[1] 3 8 1 3 5