example.py

import numpy as np
from gensim import matutils
from gensim.models.ldamodel import LdaModel
from sklearn.datasets import fetch_20newsgroups
from sklearn.feature_extraction.text import CountVectorizer
from gensim.sklearn_integration.SklearnWrapperGensimLdaModel import SklearnWrapperLdaModel
## class to execute fit_predict. Will later on add to lda wrapper.
class dummy(SklearnWrapperLdaModel):
    def fit_predict(self,X):
        corpus=matutils.Sparse2Corpus(X)
        return SklearnWrapperLdaModel.fit(self,corpus)

    
rand = np.random.mtrand.RandomState(8675309)
cats = ['rec.sport.baseball', 'sci.crypt']
data = fetch_20newsgroups(subset='train',
                        categories=cats,
                        shuffle=True)

vec = CountVectorizer(min_df=10, stop_words='english')
X = vec.fit_transform(data.data)  

vocab = vec.get_feature_names() 

vec = CountVectorizer(min_df=10, stop_words='english')

X = vec.fit_transform(data.data)
#corpus=matutils.Sparse2Corpus(X,documents_columns=False)
vocab = vec.get_feature_names() #vocab to be converted to id2word 
id2word=dict([(i, s) for i, s in enumerate(vocab)])

obj=dummy(id2word=id2word,num_topics=5,passes=20)
lda=obj.fit_predict(X)
lda.print_topics()