0

I've tried to use various function in QDAP, such as word_cor, word_associate, ect..., but I continually get back an error that says "can't allocate vector memory of 1.5gb" (something along these lines).

However, I don't need the function to return all of the results of a word_cor (and the like), even at a .7 correlation. I just want the top 50 or so.

Can/should I use the top_n function in dpylr, or is there another way to dramatically cut down the returns? I've looked into sparse matrices, but I don't think that's what I need.

Alex
  • 77
  • 1
  • 10
  • What are you trying to do or understand? – Tyler Rinker Jun 17 '17 at 00:58
  • I'm trying to run a series of tests for descriptive statistics on a large volume of text; the text is organized observation-by-row, and has other data like group, time, etc... Every time I try to run a test I get that error. – Alex Jun 19 '17 at 13:02
  • I would suggest qdap is not the right tool for this. I'd make a document-term-matrix, determine the top n terms you want (using column sums after excluding stopwords) and do your distance measure on this. – Tyler Rinker Jun 21 '17 at 01:26

0 Answers0