author_path = os.path.join(args.dataset, "total", author_id)
for file_index in training_set_indexes:
classifier.add_example(os.path.join(author_path, str(file_index) + ".txt"), author_index)
# end for
# end for
# >> 8. Train model
classifier.train()
# >> 9. Test model performance
success = 0.0
count = 0.0
for author_index, author_id in enumerate((args.author1, args.author2)):
author_path = os.path.join(args.dataset, "total", author_id)
for file_index in test_set_indexes:
author_pred, _, _ = classifier.pred(os.path.join(author_path, str(file_index) + ".txt"))
if author_pred == author_index:
success += 1.0
# end if
count += 1.0
# end for
# end for
# >> 11. Save results
average_success_rate = np.append(average_success_rate, [(success / count) * 100.0])
# Delete variables
del classifier
# end for
# Log results
str(file_index) + ".txt"), 1)
# end for
# >> 8. Train model
classifier.train()
# >> 9. Test truth author
success = 0.0
count = 0.0
author_path = os.path.join(args.dataset, "total", args.author)
for file_index in test_set_indexes:
file_path = os.path.join(author_path, str(file_index) + ".txt")
# Document success rate
if not args.sentence:
author_pred, _ = classifier.pred(file_path)
if author_pred == 0:
success += 1.0
# end if
count += 1.0
else:
# Sentence success rate
nlp = spacy.load(args.lang)
doc = nlp(io.open(file_path, 'r').read())
for sentence in doc.sents:
sentence_pred, _ = classifier.pred_text(sentence.text)
if sentence_pred == 0:
success += 1.0
# end if
count += 1.0
# end for
training_text_path = os.path.join(args.dataset, text_codes[training_index] + ".txt")
training_text_author = texts_data[text_codes[training_index]]
classifier.add_example(training_text_path, training_text_author)
# end for
# Train model
print(u"Training...")
classifier.train()
# Test model performance
success = 0.0
count = 1.0
for test_index in test_set_indexes:
test_text_path = os.path.join(args.dataset, text_codes[test_index] + ".txt")
observed_author = texts_data[text_codes[test_index]]
predicted_author = classifier.pred(test_text_path)
if observed_author == predicted_author:
success += 1
# end if
count += 1
# end for
# >> 11. Save results
average_success_rate = np.append(average_success_rate, [(success / count) * 100.0])
# Delete variables
del classifier
# end for
# Log results
logging.save_results(u"Average success rate ", np.average(average_success_rate), display=True)
os.path.join(author_path,
str(file_index) + ".txt"), author_index)
# end for
# end for
# >> 8. Train model
classifier.train()
# >> 9. Test model performance
success = 0.0
count = 0.0
for author_index, author_id in enumerate(c):
author_path = os.path.join(args.dataset, "total", author_id)
for file_index in test_set_indexes:
author_pred = classifier.pred(os.path.join(
author_path,
str(file_index) + ".txt"),
show_graph=True)
if author_pred == author_index:
success += 1.0
# end if
count += 1.0
# end for
# end for
# >> 11. Save results
average_success_rate = np.append(average_success_rate,
[(success / count) * 100.0])
# Delete variables
del classifier
# end for
test_set_indexes = np.array([0])
# >> 6. Create Echo Word Classifier
classifier = RCNLPEchoWordClassifier(size=rc_size, input_scaling=rc_input_scaling, leak_rate=leaky_rate,
input_sparsity=rc_input_sparsity, converter=converter, n_classes=2,
spectral_radius=rc_spectral_radius, w_sparsity=rc_w_sparsity, w=w)
# >> 7. Add examples
for author_index, author_id in enumerate((args.author1, args.author2)):
author_path = os.path.join(args.dataset, "total", author_id)
for file_index in training_set_indexes:
classifier.add_example(os.path.join(author_path, str(file_index) + ".txt"), author_index)
# end for
# end for
# >> 8. Train model
classifier.train()
# >> 9. Test model performance
author_index = 0
author_id = args.author1
author_path = os.path.join(args.dataset, "total", author_id)
print("Testing model performances with %d text files for author from %s..." % (test_set_indexes.shape[0],
author_path))
classifier.pred(os.path.join(author_path, str(0) + ".txt"), show_graph=False, print_outputs=True)
# Delete variables
del classifier
# end for
# end if
# end for
# >> 8. Train model
classifier.train()
# >> 9. Test model performance
success = 0.0
count = 0.0
for author_index, author_id in enumerate(
(args.author1, args.author2)):
author_path = os.path.join(args.dataset, "total", author_id)
for file_index in test_set_indexes:
file_path = os.path.join(author_path,
str(file_index) + ".txt")
author_pred, _, _ = classifier.pred(os.path.join(
author_path,
str(file_index) + ".txt"),
show_graph=False)
# Success rate
if not args.sentence:
author_pred, _, _ = classifier.pred(
os.path.join(author_path,
str(file_index) + ".txt"))
if author_pred == author_index:
success += 1.0
# end if
count += 1.0
else:
# Sentence success rate
nlp = spacy.load(args.lang)
doc = nlp(io.open(file_path, 'r').read())
for sentence in doc.sents:
# >> 8. Train model
classifier.train()
# >> 9. Test model performances
success = 0.0
count = 0.0
max_prob = 0
max_cat = ''
# >> 10. Test same author
for file_index in test_set_indexes:
file_path = os.path.join(author_path, str(file_index) + ".txt")
# Doc. success rate
if not args.sentence:
author_pred, same_prob, diff_prob = classifier.pred(file_path)
same_probs = np.append(same_probs, same_prob)
if same_prob > max_prob:
max_prob = same_prob
max_cat = 'same'
# end if
if same_prob > diff_prob and same_prob > args.threshold:
success += 1.0
# end if
count += 1.0
else:
# Sentence success rate
nlp = spacy.load(args.lang)
doc = nlp(io.open(file_path, 'r').read())
for sentence in doc.sents:
author_pred, same_prob, diff_prob = classifier.pred_text(sentence.text)
# >> 7. Add authors examples
classifier.add_example(text1_path, 0)
# >> 8. Add negative examples
others_path = os.path.join(args.dataset, "total", "others")
for file_index in range(0, args.negatives):
file_path = os.path.join(args.dataset, "others", str(file_index) + ".txt")
classifier.add_example(file_path, 1)
# end for
# >> 8. Train model
classifier.train()
# Get similarity
author_pred, same_prob, diff_prob = classifier.pred(text2_path)
print("%s : %f" % (text2, same_prob))
# Save
similarity_matrix[index1, index2] = same_prob
# end if
index2 += 1
# end for
index1 += 1
# end for
plt.imshow(similarity_matrix, cmap='gray')
plt.show()
pickle.dump(similarity_matrix, open(args.matrix, 'w'))
# Get links
count_links = 0