I have a document Term matrix with nine documents:
I am running the code as below:
import pyLDAvis.gensim
topicData = pyLDAvis.gensim.prepare(ldamodel, docTermMatrix, dictionary)
pyLDAvis.display(topicData)
I am getting the below error when executing pyLDAvis.display function:
TypeError: Object of type 'complex' is not JSON serializable
Can someone guide here? What could be the reason?
I had the same problem. Following the GH issue referenced by user3411846 I found a different, simpler solution.
The complex number had come from coordinate calculation and specifying the "mds" worked.
https://github.com/bmabey/pyLDAvis/issues/69#issuecomment-311337191
So your code would be
topicData = pyLDAvis.gensim.prepare(ldamodel, docTermMatrix, dictionary, mds='mmds')
Other options for mds are here: https://pyldavis.readthedocs.io/en/latest/modules/API.html#pyLDAvis.prepare
add this to your pyldavis/utils.py (NumPyEncoder) if np.iscomplexobj(obj): return abs(obj)
class NumPyEncoder(json.JSONEncoder): def default(self, obj): if isinstance(obj, np.int64) or isinstance(obj, np.int32): return int(obj) if isinstance(obj, np.float64) or isinstance(obj, np.float32): return float(obj) if np.iscomplexobj(obj): return abs(obj) return json.JSONEncoder.default(self, obj)
#resolved my issue