generate a word-topic matrix for a new doc

[angela]

Hi,

I already trained an LDA model on a list of documents. And now I want to apply this trained model to a new document and generate a word-topic matrix. That's to say, I want the rows to be every single word in the new document, and the columns are the topics. The value in the matrix represents how likely that word is assigned to the topic.

But I don't want the model to consider one word at a time: I want the model to think about the whole new document first, and then give the values based on every word in the new document. 

I'm not sure how I can realize this in gensim. Could anyone help me with that? Thanks!