In what format do the text files in my Corpus have to be in order to use the tm package?

Asked May 12 '17 at 12:39

Active May 12 '17 at 12:39

Viewed 86 times

I am currently working with a flat text file that is literally just a large body of text copy and pasted into notepad.I designated a folder in my desktop as my corpus to use as my working directory.

    #Map a directory to the corpus location on the desktop 
    cname<-file.path("~", "Desktop", "corpus" )

    #Begin text processing 
    docs<-Corpus(DirSource(cname))
    summary(docs)

Before I can do any sort of clustering, I am getting the error from summary:

    Error in DirSource("cname/") : empty directory

Does this error involve the way the document is formatted? What is the proper way to format a flat text file in order to use the tm package?

asked May 12 '17 at 12:39

Vindication09

1

It looks like a problem with `cname`. Check that it is actually the directory you expect (and that `~` means what you think it does). – Andrew Gustar May 12 '17 at 12:44
After some research, it seems to be a bug with the latest version of cm. http://stackoverflow.com/questions/24643925/r-got-problems-in-reading-text-file – Vindication09 May 12 '17 at 20:23
That link is three years old. – Andrew Gustar May 13 '17 at 06:33

In what format do the text files in my Corpus have to be in order to use the tm package?

0 Answers0