Help with Naïve Bayes Analysis in R

I am trying to analyse some tweets using the naive bayes method on R. However, I am having issues, below is my code. When I try and produce a text model, I keep getting told that y cannot be constant. I am not clear what this error message means or how to solve it.

Help would be much appreciated.


id_train <- sample(1:5000, 1500, replace = FALSE)
head(id_train, 10)
[1] 2638 874 3650 3740 789 553 1705 4368 4557
[10] 2828
MasterCorpus$id_numeric <- 1:ndoc(MasterCorpus)
dfmat_training <- corpus_subset(MasterCorpus, id_numeric %in% id_train) %>%

  •  dfm(remove = stopwords("en"), stem = TRUE)

dfmat_test <- corpus_subset(MasterCorpus, !id_numeric %in% id_train) %>%

  •  dfm(remove = stopwords("english"), stem = TRUE)

tmod_nb <- textmodel_nb(dfmat_training, dfmat_training$author)
Error in textmodel_nb.dfm(dfmat_training, dfmat_training$author) :
y cannot be constant

How many authors are there in training?

There are only two authors

This topic was automatically closed 21 days after the last reply. New replies are no longer allowed.

If you have a query related to it or one of the replies, start a new topic and refer back with a link.