file=results_file)
print("Support: %d" % (support if support is not None else -1),
file=results_file)
report = classification_report(y_true, y_pred)
with directory.joinpath("classification_report").open(
mode='w') as report_file:
print(report, file=report_file)
cm = confusion_matrix(y_true, y_pred)
with directory.joinpath("confusion_matrix").open(mode='w') as cm_file:
np.savetxt(cm_file, cm, fmt="%d")
if __name__ == "__main__":
data = read_dataset_splits(reader=data_readers.read_question_only_data,
splits=["tiny", "train", "dev"])
X_train, y_train = prepare_data(data.train)
X_dev, y_dev = prepare_data(data.dev)
tokenizer = Tokenizer(num_words=MAX_NUM_WORDS,
oov_token="<UNK>",
split=' ',
lower=True)
tokenizer.fit_on_texts(X_train)
embeddings = getFastTextEmbeddings(tokenizer.word_index)
model = simpleRNN(embeddings, hidden_dim=200)
X_train = tokenizer.texts_to_sequences(X_train)
X_dev = tokenizer.texts_to_sequences(X_dev)
from model_utils import get_response_time_label, add_cosine_similarity, add_question_length, add_jensen_shannon
from scipy.stats import pearsonr
random.seed(Config.SEED)
def calc_correlation(data, col_name):
return pearsonr(data.train['response_time_sec'], data.train[col_name])
if __name__ == '__main__':
results = {}
data = read_dataset_splits(
reader=data_readers.read_question_and_context_data,
window_size=10,
include_question_text=True,
include_context_text=True,
include_context_speaker=False,
include_context_times=False)
data = add_jensen_shannon(data)
results['question_and_js'] = calc_correlation(data, 'jensen_shannon')
df = read_corpus(split='train')
all_words = [item for sublist in df.text for item in sublist]
for N_words in [25, 50, 100]:
top_words = [
item[0] for item in Counter(all_words).most_common(N_words)
]
data = read_dataset_splits(
reader=data_readers.read_question_and_context_data,
window_size=10,
('tfidf', TfidfTransformer()),
('clf', DummyClassifier(random_state=SEED))])
pipe.fit(train['question'], train['question_class'])
preds = pipe.predict(dev['question'])
print("Dummy Classifier: ")
report = classification_report(dev['question_class'], preds)
print(report)
cm = confusion_matrix(dev['question_class'], preds)
plot_cm(cm, title="Dummy Classifier")
return models
if __name__ == '__main__':
data = read_dataset_splits(reader=read_question_only_data)
data = add_classes(data)
models = run_baselines(data)
train, dev, test = data['train'], data['dev'], data['test']
dev_new = copy.deepcopy(dev)
pipe = Pipeline([
('vect', CountVectorizer(tokenizer=dummy_tokenizer, lowercase=False)),
('tfidf', TfidfTransformer()),
('clf', LogisticRegression(class_weight='balanced', random_state=SEED))
])
#Generate results for Logistic regression and output to CSV file.
#True class, predicted class and class probabilities.
# g = models['Logistic Regression']
# pipe.set_params(**g)
report = classification_report(y_true, y_pred)
with directory.joinpath("classification_report").open(
mode='w') as report_file:
print(report, file=report_file)
cm = confusion_matrix(y_true, y_pred)
with directory.joinpath("confusion_matrix").open(mode='w') as cm_file:
np.savetxt(cm_file, cm, fmt="%d")
if __name__ == "__main__":
data = read_dataset_splits(
reader=data_readers.read_question_and_context_data,
splits=["tiny"],
window_size=WINDOW_SIZE,
include_question_text=True,
include_context_text=True,
include_context_speaker=True,
include_context_times=False)
X_train, y_train = prepare_data(data.tiny)
X_dev, y_dev = prepare_data(data.tiny)
tokenizer = Tokenizer(num_words=MAX_NUM_WORDS,
oov_token="<UNK>",
split=' ',
lower=True)
tokenizer.fit_on_texts(X_train)
embeddings = getFastTextEmbeddings(tokenizer.word_index)
model = simpleRNN(embeddings, hidden_dim=200)