r function not return any result - r

Question

I have created the following function. R should return 1 as a result but I doesn't give any result?
phrasedis <- function(string, phrase1, phrase2,n) {
char1 <- unlist(gregexpr(phrase1,string))
char2 <- unlist(gregexpr(phrase2,string))
for ( i in 1: 2) {
for (j in 1: 2) {
pos1[i] <- sapply(strsplit(substr(text, 0,char1[i]-1),"\\s"),length)+1
pos2[j] <- sapply(strsplit(substr(text, 0,char2[j]-1),"\\s"),length)+1
dist <- pos2[i] - pos1[j]
a <- ifelse(pos1[i]>0 & pos2[j]>0 & dist>0 & dist<=6,1,0)
if (a==1) break
return(a)
}
}
}
text <- "phone rang a b c d e f z y z phone rang but no answer"
b <- sapply(text, FUN=function(str) phrasedis(str,"phone rang" , "no answer",6))
What it should do is return 1 if the distance between phone rang and no answer is less than 6 words, otherwise return 0.
Thank you very much for your help.

Answer 1

The logic of your function is wrong.
First of all, you put the return() statement inside the loop, so the loop stops always in the first iteration due to that return() statement.
Then, you don't create the vectors pos1 and pos2, so your function can't even work. The only reason you don't complain about an error, is because you have a pos1 and pos2 in your global environment probably.
But even when placing the return statement where it's supposed to go (at the end!) and creating a pos1 and pos2 vector of length 2, your function can't work because your loop is wrong.
You loop over 1 and 2, which doesn't make sense at all unless you have exact 2 matches for both phrase1 and phrase2 in string. Due to that and the fact there is only 1 match for phrase2, when j==2 the outcome of the substr(text, 0, char2[j] -1) is NA, which has an exact length of 1 so pos2[j] becomes 2. Meanwhile pos1[i] is still 1, which fulfills your condition and hence 1 is returned.
This is how you could do it:
phrasedis <- function(string, phrase1, phrase2,n) {
char1 <- gregexpr(phrase1,string)[[1]]
char2 <- gregexpr(phrase2,string)[[1]]
# -1 is returned if no match was found for either phrase
if(any(c(char1,char2) == -1)){
return(0)
}
# Calculate the end positions of the words
end1 <- char1 + attr(char1, "match.length")
#set a to 0
a <- 0
# loop over all matches in char1
for(i in seq_along(char1)){
# Find the closest match for phrase 2
thepos <- which.min(abs(char2 - end1[i]))
# get all words in between.
# Don't forget to trim white spaces before and after
inbetween <- trimws(substring(string, end1[i], char2[thepos]-1))
inbetween <- strsplit(inbetween,"\\s")[[1]]
if(length(inbetween) <= n){
a <- 1
break
}
}
return(a)
}
This is how it works:
> text <- "phone rang a b cd phone rang d e f g h i no answer"
> phrasedis(text,"phone rang" , "no answer",6)
[1] 1
> text <- " There is nothing in this text"
> phrasedis(text,"phone rang" , "no answer",6)
[1] 0
> text <- "No answer but the phone rang"
> phrasedis(text,"phone rang" , "no answer",6)
[1] 0