def train_model(nodes, datasets):
y = np.array(range(len(nodes)))
nodes = list(nodes)
vectorizer = DictVectorizer(sparse=True)
for i, dataset in enumerate(datasets):
g = compute_dataset(dataset)
nodes.extend(g.classes)
sys.stdout.write('\r')
sys.stdout.write(str(i+1))
sys.stdout.flush()
X = vectorizer.fit_transform([dict(node.concept_vector) for node in nodes])
y = y + [-1 for i in range(len(nodes) - len(y))]
unlabeled = []
model = LabelSpreading()
model.fit(X, y)
model.vectorizer = vectorizer
return model
def train_model(nodes, datasets):
y = np.array(range(len(nodes)))
nodes = list(nodes)
vectorizer = DictVectorizer(sparse=True)
for i, dataset in enumerate(datasets):
g = compute_dataset(dataset)
nodes.extend(g.classes)
sys.stdout.write('\r')
sys.stdout.write(str(i + 1))
sys.stdout.flush()
X = vectorizer.fit_transform([dict(node.concept_vector) for node in nodes])
y = y + [-1 for i in range(len(nodes) - len(y))]
unlabeled = []
model = LabelSpreading()
model.fit(X, y)
model.vectorizer = vectorizer
return model
