Viz Error with Empty Clusters (pyLDAvis, Int64Index) about gsdmm

[EssbieWGT]

Circling back on this. While it negatively affects the visualization of the model by packing the clusters more tightly together, I found that making the following two changes solved the problem.

First, change the utils.py per this. Next make the following update to the workaround code:

    matrix = []
    for cluster in mgp.cluster_word_distribution:
        total = sum([occurance for word, occurance in cluster.items()])
        assert not math.isnan(total)
        # assert total > 0
        if total == 0:
            row = [(1 / len(vocabulary))] * len(vocabulary)   
        else:
            row = [cluster.get(term, 1) / total for term in vocabulary] # <--- updated from 0 to 1 
        for f in row:
            assert not isinstance(f, complex)
        matrix.append(row)
    return matrix, doc_topic_dists2, doc_lengths, vocabulary, term_counts

These two steps got the visualization working for me just fine! Will close this one out.

[ernests]

Coming back to this!
I fixed this by adding mds='mmds' as a parameter to prepare_visualization_data function.
As I understand, it is different dimension reduction algorithm and works if scikit is installed(https://github.com/bmabey/pyLDAvis/blob/master/pyLDAvis/_prepare.py#L337).

Solution from here - `bmabey/pyLDAvis#69 (comment)