def calculate_similarity(self, query_file, data_file, filename, k=None):
"""Calculate the similarity between a query file and a data file.
The results are written to a file named 'filename'."""
queries_set = doc.DocumentSet(query_file)
documents_set = doc.DocumentSet(data_file)
results = []
for query in queries_set.documents:
results.extend(self._tfidf(query, documents_set))
output.write_output_file(filename, results)
def calculate_similarity(self, query_file, data_file, filename, k=None):
"""Calculate the similarity between a query file and a data file.
The results are written to a file named 'filename'."""
queries_set = doc.DocumentSet(query_file)
documents_set = doc.DocumentSet(data_file)
results = []
for query in queries_set.documents:
results.extend(self._tfidf(query, documents_set))
output.write_output_file(filename, results)
def main():
"""Calculates the basic word overlap similarity between qrys.txt & docs.txt.
The results are written out to the file 'overlap.top'."""
query_file = 'data/qrys.txt'
data_file = 'data/docs.txt'
queries_set = doc.DocumentSet(query_file)
documents_set = doc.DocumentSet(data_file)
results = []
for query in queries_set.documents:
results.extend(_calculate_overlap(query, documents_set.inverted_index))
# Output the overlaps.
output.write_output_file('overlap.top', results)
def start_program():
account = input('Введите id аккаунта или ник ')
try:
vk.params['user_id'] = int(account)
except ValueError:
print("введен ник")
vk.params['screen_name'] = account
user_info = vk.get_user_info()
print(user_info['response'][0]['id'])
user_id = user_info['response'][0]['id']
users = vk.search_users(user_info)
result = vk.compare_friends_groups(users)
top10_users = s.find_top10(result)
top10_users_with_photos = vk.find_top3_photos(top10_users)
output = o.create_output_file(top10_users_with_photos)
db.write_db_output(user_id, output)
o.write_output_file(output)
print("Программа завершена")
def calculate_similarity(self, query_file, data_file, filename):
"""Calculate the similarity between a query file and a data file.
The results are written to a file named "filename"."""
queries_set = doc.DocumentSet(query_file)
documents_set = doc.DocumentSet(data_file)
results = []
for query in queries_set.documents:
# Compute the initial tfidfs.
initial_tfidfs = self.tf_idf._tfidf(query, documents_set)
# Select the top n_d scoring documents.
initial_tfidfs = sorted([(-s, d) for (_, d, s) in initial_tfidfs])
initial_tfidfs = [(d, -s) for (s, d) in initial_tfidfs[:self.n_d]]
selected_docs = [document for (document, _) in initial_tfidfs]
# Combine the top documents into a 'mega document'.
summed_counter = counter.Counter(query.words_counter)
for document in selected_docs:
summed_counter += document.words_counter
mega_document = doc.document_from_dict(None, dict(summed_counter))
# Select the top n_w scoring words (via tf.idf) from the megadocument.
word_scores = []
for word in sorted(list(mega_document.words_counter)):
score = self.tf_idf._document_tfidf(word, mega_document, documents_set)
word_scores.append((-score, word))
word_scores = sorted(word_scores)[:self.n_w]
word_scores = [(word, -score) for (score, word) in word_scores]
# Use these new words as the next query, and return the tf.idf scores.
new_query = doc.document_from_dict(query.id, dict(word_scores))
results.extend(self.tf_idf._tfidf(new_query, documents_set))
output.write_output_file(filename, results)
