Relative Position of Words in String in R

Question

I have the following term documnet matrix and dataframe.

tdm <- c('Free', 'New', 'Limited', 'Offer')



Subject                                               Free New Limited Offer                                                    

'Free Free Free! Clear Cover with New Phone',          0   0     0      0
'Offer ! Buy New phone and get earphone at             0   0     0      0
1000. Limited Offer!'

I want to derive the following dataframe as the output

Subject                                              Free  New Limited Offer    
'Free Free Free! Clear Cover with New Phone',        1,2,3  8   NA     NA
Offer ! Buy New phone and get earphone at  1000.      NA    3   12      1,13
Limited Offer!'                                                                         

I tried the following code and got a result but this only gives me the position of the word along a string. I need the position of the words as in Hell0 - 1 new- 2.

for(i in 1:length(tdm))
{    word.locations <- 
gsub(")","",gsub("c(","",unlist(paste(gregexpr(pattern 
= tdm[i], DF$Subject))), fixed = TRUE), fixed = TRUE)
  df <- cbind(DF,word.locations)
  }
  colnames(DF) <- c("text", word)

I request someone to help.

Answer 1

Given the inputs:

tdm <- c('Free', 'New', 'Limited', 'Offer')
subject <- c("Free Free Free! Clear Cover with New Phone",
             "Offer ! Buy New phone and get earphone at 1000. Limited Offer!")

I'd do something like:

sapply(tolower(tdm), function(x) {
    lapply(strsplit(tolower(subject), "(\\s+)|(?!')(?=[[:punct:]])", perl = TRUE), 
      function(y) {
        y <- y[nzchar(y)]
        toString(grep(x, y))
      })
})
##      free      new limited offer  
## [1,] "1, 2, 3" "8" ""      ""     
## [2,] ""        "4" "12"    "1, 13"

What's going on:

• Use tolower on both the string to match against and the terms being matched.
• Use strsplit to split words and punctuation into separate items in a list element.
• Remove any empty vectors with nzchar().
• Use grep() to find the location of the matches.
• Use toString() to paste the locations together as a comma-separated string.