1

I'm trying to create an interactive pyLDAvis visualization for an LDA model I have built, and although I am able to create a static visualization, I am struggling with the following:

  1. Outputting a dynamic visual that interacts with the 'relevance metric' slider on the top right hand corner of the screen. The words per topic, the saliency measures, etc. do not change when I adjust the relevance metric.

Screenshot of the static visual as displayed on the webpage

  1. After being unable to display an interactive visual through a saved html file and on a webpage, I tried to display the visual on Spyder. I tried using the following commands:
In: pyLDAvis.display(vis_data)
Out: <IPython.core.display.HTML object>

and

In: pyLDAvis.show(vis_data)
Out: FileNotFoundError: [Errno 2] No such file or directory: 'https://cdn.jsdelivr.net/gh/bmabey/pyLDAvis@3.3.1/pyLDAvis/js/ldavis.v1.0.0.css'

Below is the code that I have used to create the model and the plot. I'd appreciate it if I could get help in creating an interactive html file or in displaying the result in an IDE (preferably Spyder). My python version is 3.8.3. Thanks!

# Calling File Cleanser on a file
fileTextTokenized = FileCleanser(mahabharatatext)
        
# Converting each word into a word-count dictionary format
dictionary = Dictionary(fileTextTokenized)
dictionary.filter_extremes(no_below = 100, no_above = 0.8)

gensim_corpus = [dictionary.doc2bow(word) for word in fileTextTokenized]
temp = dictionary[0]
id2word = dictionary.id2token

# Setting model parameters    
chunksize = 2000
passes = 20
iterations = 400
num_topics = 6

lda_model = LdaModel(
corpus=gensim_corpus,
id2word=id2word,
chunksize=chunksize,
alpha='auto',
eta='auto',
iterations=iterations,
num_topics=num_topics,
passes=passes
)

vis_data = gensimvis.prepare(lda_model, gensim_corpus, dictionary)
vis_data
pyLDAvis.display(vis_data)
pyLDAvis.save_html(vis_data, './FileModel'+ str(num_topics) +'.html')
beginnerprogrammerforever
  • 77
  • 5

1 Answers1

0

I am also trying to create an LDA vizualization and it works for me with my data. I just used your last 2 pieces of code (adding sort_topics=False) :

vis_data = pyLDAvis.gensim_models.prepare(model, corpus, dictionary, sort_topics=False)

pyLDAvis.save_html(vis_data, 'path/doc_name.html')

Once you have saved your model you can click on it and it will open in your browser and you can change the topic and the lambda value (I am using chrome).

MarionEtp
  • 73
  • 5