def write_topics_dictionary():
topics_dictionary = {}
counter = 1
for reuter in retrieve_all_relevant_reuters():
_update_topics_dictionary(_get_topics(reuter), topics_dictionary)
print(counter)
counter += 1
with open('topics_dictionary.json', 'w') as f:
json.dump(topics_dictionary, f, indent=4, sort_keys=True)
def write_topic_nonverted_index():
topic_nonverted_index = {}
counter = 1
for reuter in retrieve_all_relevant_reuters():
_update_topic_nonverted_index(_get_topics(reuter),
_preprocess_content(reuter), topic_nonverted_index)
print(counter)
counter += 1
with open('topic_nonverted_index.json', 'w') as f:
json.dump(topic_nonverted_index, f, indent=4, sort_keys=True)
def write_document_topic_dictionary():
document_topic_dictionary = {}
j = 1
for r in retrieve_all_relevant_reuters():
newid = _get_id(r)
topics = _get_topics(r)
document_topic_dictionary[newid] = []
for t in topics:
document_topic_dictionary[newid].append(t)
print(j)
j += 1
with open('document_topic_dictionary.json', 'w') as f:
json.dump(document_topic_dictionary, f, indent=4, sort_keys=True)
def write_targets():
with open('topics_dictionary.json', 'r') as f:
topics_dictionary = json.load(f)
k = 1
for reuter in retrieve_all_relevant_reuters():
newid = _get_id(reuter)
topics = _get_topics(reuter)
for t in topics_dictionary:
if t in topics:
topics_dictionary[t] = 1
else:
topics_dictionary[t] = 0
vector = []
for t in topics_dictionary:
vector.append(topics_dictionary[t])
with open(f'training_nn/{newid}.json', 'w') as f:
json.dump(vector, f)
print(k)
k += 1