Hierarchical Topics Counts

[dalematthews1]

With hierarchical topic modelling is there an easy way to get the counts for each of the topics in the hierarchy? For example I would like to produce a plot similar to visualise_hieararchy() which also shows counts of documents belonging to each topic in the hierarchy.

[dalematthews1]

Another thing which I am interested in doing; getting a list of all of the topics belonging to a hierarchical cluster.

[dalematthews1]

I think this is mostly solved. I found I can generate the counts for each topic using the 'Topics' column of the hierarchical_topics dataframe and getting the topic counts from the topic_model.get_topic_info(). I then found I can modify the dendrogram returned by visualize_hierarchy() by modifying the trace.hovertext of the plot so that the plot shows the counts when you hover. "getting a list of all topics belonging to a hierarchical cluster" - this is simply contained in the hierarchical_topics dataframe. I still think it would be a good idea to explicitly include counts in the hierarchical_topics dataframe and an option to include them as hover text in the prompt.