Odtnhj

I have a list of 196 strings in the form 2009/EPS.WCR.PL6.MAIS.0036, 2016/EPS.WCR.PL6.NORM.0077 etc. What varies is the year date and the four numbers at the end. Also there are either NORM or MAIZE. I would like to go through this list and extract these bits of information to create a some sort of distance matrix. The code I have written so far is as follow:
c(substr(df[i,3], 1, 4),substr(df[1,3], 18, 21),substr(df[i,3], 22, nchar(df[i,4]))),
where df is the list of these catagorical variables.

Where i loops through the list. Is there a nice way of getting a distance between these strings based on the bits of information that I am extracting?

Thank you in advance.

If you data has the same structure throughout then try:

 data <- c("2009/EPS.WCR.PL6.MAIS.0036", "2016/EPS.WCR.PL6.NORM.0077")
 str(data)
 substr(data, start = 1, stop = 4)
 substr(data, start = 18, stop = 21)
 substr(data, start = 23, stop = 26)

The following function uses CRAN package stringdist to compute the distances between the strings in its first argument. You can pass a method of your choice, see the help page help("stringdist").

special_dist <- function(x, method = "osa")
 y <- sub("(^[[:digit:]]+).*[[:punct:]]([[:digit:]]+$)", "\1\2", x)
 res <- sapply(y, function(z) stringdist::stringdist(z, y, method = method))
 rownames(res) <- colnames(res)
 res


x <- c("2009/EPS.WCR.PL6.MAIS.0036", "2016/EPS.WCR.PL6.NORM.0077")
special_dist(x)
# 20090036 20160077
#20090036 0 4
#20160077 4 0

special_dist(x, "jaccard")
# 20090036 20160077
#20090036 0.0000000 0.5714286
#20160077 0.5714286 0.0000000