Error: No tidy method for objects of class LDA_VEM§

Question

I am literally following the steps as presented in chapter 6 of the "Text Mining in R: a Tidy Approach" book. See: https://www.tidytextmining.com/topicmodeling.html

#import libraries
library(topicmodels)
library(tidytext)

#access dataset
data("AssociatedPress")

# set a seed so that the output of the model is predictable
ap_lda <- LDA(AssociatedPress, k = 2, control = list(seed = 1234))

#tidy model object
ap_topics <- tidy(ap_lda, matrix = "beta")

Gives me the following error in the terminal:

Error: No tidy method for objects of class LDA_VEM

What I should be getting, meanwhile is:

## # A tibble: 20,946 x 3
##    topic term           beta
##    <int> <chr>         <dbl>
##  1     1 aaron      1.69e-12
##  2     2 aaron      3.90e- 5
##  3     1 abandon    2.65e- 5
##  4     2 abandon    3.99e- 5
##  5     1 abandoned  1.39e- 4
##  6     2 abandoned  5.88e- 5
##  7     1 abandoning 2.45e-33
##  8     2 abandoning 2.34e- 5
##  9     1 abbott     2.13e- 6
## 10     2 abbott     2.97e- 5
## # ... with 20,936 more rows

Why am I seeing this error instead of the desired result?

What version of `topicmodels` and `tidytext` are you using? You can check with `sessionInfo()` — MrFlick, Dec 11 '18 at 16:37

score 1 · Accepted Answer · answered Dec 12 '18 at 04:01

You are not the first person to run into this problem, but it is very difficult to reproduce. In fact, I personally have never been able to reproduce the error. However, I know it is a real problem because, well... If you'd like to see other people also struggling, check out here or here or here or here.

As best as I can tell, this is likely a bug in the topicmodels package, related to S4 registration. You can avoid running into this by not saving .RData between sessions and always starting from a fresh R session when you open up work.

score 0 · Answer 2 · answered Dec 11 '18 at 16:43

0

You need to first tidy the AssociatedPress data. Like this:

#if(!require("topicmodels")) install.packages("topicmodels")
#install.packages("topicmodels")
library(topicmodels)
data("AssociatedPress",package="topicmodels")
AssociatedPress
#Getting the Terms
terms<-Terms(AssociatedPress)
head(terms)
#tidyig with the tidy function
ap_tidy<-tidy(AssociatedPress)

And then:

ap_lda<-LDA(AssociatedPress,k=2,control=list(seed=1234))
ap_topics<-tidy(ap_lda,matrix="beta")
head(ap_topics)

Which gives:

 topic term          beta
  <int> <chr>        <dbl>
1     1 aaron     1.69e-12
2     2 aaron     3.90e- 5
3     1 abandon   2.65e- 5
4     2 abandon   3.99e- 5
5     1 abandoned 1.39e- 4
6     2 abandoned 5.88e- 5

answered Dec 11 '18 at 16:43

NelsonGon

13,015
7
27
57

1

Running your code gives me the exact same error – Vasino Dec 11 '18 at 16:46
Well I'm using tidytext 0.1.9 – NelsonGon Dec 11 '18 at 16:48
Switched to your version and still works for me. – NelsonGon Dec 11 '18 at 16:54
that is profoundly confusing – Vasino Dec 11 '18 at 16:59
Is this all the code you used? – NelsonGon Dec 11 '18 at 17:00
Yes, I copy-pasted my entire script – Vasino Dec 11 '18 at 17:01
Running your code throws no error for me(Just tried it now). As cliched as it sounds,maybe try restarting your session or remove and reinstall the packages? – NelsonGon Dec 11 '18 at 17:03
Still nothing, I appreciate the suggestion though – Vasino Dec 11 '18 at 17:06

Error: No tidy method for objects of class LDA_VEM§

2 Answers2