Hi Kilian,
I may have found a bug... the "Ngram creator" (characters) appears to double the corpus frequency over what's truly in the corpus. Didn't test word frequencies, as I'm only chopping up strings anyway. Version 2.11.0.
Thanks
E
Hi Kilian,
I may have found a bug... the "Ngram creator" (characters) appears to double the corpus frequency over what's truly in the corpus. Didn't test word frequencies, as I'm only chopping up strings anyway. Version 2.11.0.
Thanks
E
Hi Ergonomist,
I can not reproduce the problem with v. 2.11. In my workflow (attached) the counts are correct. What settings do you use? Can you reproduce the problem and attach a workflow?
Cheers, Kilian
Hi Kilian,
Big issue - this was right before Christmas, and I cannot remember where I used this. :-( Sorry!!! If I were to rediscover it, I'd bump this thread up again, but don't hold your breath... Thanks for the speedy reply, anyway. 8)
Cheers,
E
This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.