def predict(self, X):
with tf.Session(graph=self.model.graph) as session:
saver = tf.train.Saver()
if (self.model.load(session, saver)):
num_epochs_trained = self.model.model_graph.cur_epoch_tensor.eval(
session)
print('EPOCHS trained: ', num_epochs_trained)
else:
return
#if len(X.shape)<4:
#X = np.array([X])
X_ = prepare_dataset(X)
y_l = list()
start = 0
end = self.model.batch_size
while end < X.shape[0]:
x = X_[start:end]
print('from {} to {}'.format(start, end))
y_pred = self.model.model_graph.predict(session, x)
y_l.append(np.array(y_pred[0]))
start = end
end += self.model.batch_size
else:
x = X_[start:]
xsize = len(x)
print('from {} to {}'.format(start, len(X_)))
p = np.zeros([self.model.batch_size - xsize] +
list(x.shape[1:]))
y_pred = self.model.model_graph.predict(
session, np.concatenate((x, p), axis=0))
y_l.append(np.array(y_pred[0][0:xsize]))
return np.vstack(y_l)
def reconst_loss(self, inputs):
''' ------------------------------------------------------------------------------
DATA PROCESSING
------------------------------------------------------------------------------ '''
inputs = utils.prepare_dataset(inputs)
return self.model.reconst_loss(inputs)
def interpolate(self, input1, input2):
input1 = utils.prepare_dataset(input1)
input2 = utils.prepare_dataset(input2)
return self.model.interpolate(input1, input2)
def encode(self, inputs):
''' ------------------------------------------------------------------------------
DATA PROCESSING
------------------------------------------------------------------------------ '''
inputs = utils.prepare_dataset(inputs)
return self.model.encode(inputs)