Ejemplo n.º 1
0
 def load_topics(self, dirname):
     self.topics = []
     for subdir in [x[0] for x in os.walk(dirname)][1:]:
         for file in os.listdir(subdir):
             if file.endswith('pkl'):
                 print("attempting... ", file)
                 lda = ScikitLda.load(subdir + "/" + file)
                 for topic in lda.topics:
                     self.topics.append(topic / topic.sum())
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

from scipy.spatial.distance import cosine
from corpora.scikit import ScikitLda
import argparse

if __name__ == "__main__":
    parser = argparse.ArgumentParser(description="Calculate distance between topics")
    parser.add_argument("-t", "--topic-model", default=None, required=True)
    args = parser.parse_args()

    lda = ScikitLda.load(args.topic_model)

    topics = []
    for topic in lda.topics:
        topics.append(topic / topic.sum())

    for i in range(len(topics)):
        for j in range(i + 1, len(topics)):
            print("Topic#{0} Topic#{1} {2}".format(i, j, cosine(topics[i], topics[j])))
from corpora.corpus import load_vraagtekst_corpus
from corpora.util import select_top
from matplotlib.pyplot import savefig
import numpy as np

if __name__ == '__main__':
    corpus = load_vraagtekst_corpus('data/preprocessedData.pkl')

    print("nSamples (docs) : {0}".format(corpus.num_samples))
    print("nFeatures(words): {0}".format(corpus.num_features))

    print("saving dictionary")
    corpus.save_dictionary('data/preprocessedData.dic')

    print("computing LDA")
    lda = ScikitLda(corpus=corpus, n_topics=10)
    lda.fit()

    print("saving LDA")
    lda.save('data/preprocessedData.lda_10.pkl')

    topicWords, topicWeightedWords = lda.topic_words()

    for topic_idx, wordsInTopic in enumerate(topicWords):
        print("Topic #{0}:".format(topic_idx))
        print(" ".join(wordsInTopic))

    topicsByOrg, orgs = topics_by_discrete_property(
        lda, corpus.metadata_frame['individu of groep'])
    averageWeights = np.average(lda.weights, axis=0)
    # get topic specificity by comparing with the average topic weights