def __init__(self, is_training=False, **kwargs):
super(GCONVRNN, self).__init__(**kwargs)
self._train_batch_size = int(kwargs['data'].get('batch_size'))
self._test_batch_size = int(kwargs['data'].get('test_batch_size'))
self._epoch = int(kwargs['train'].get('epoch'))
self._logstep = int(kwargs['train'].get('logstep'))
self._checkpoint_secs = int(kwargs['train'].get('checkpoint_secs'))
self._data = utils.load_dataset_dcrnn(
seq_len=self._model_kwargs.get('seq_len'),
horizon=self._model_kwargs.get('horizon'),
input_dim=self._model_kwargs.get('input_dim'),
mon_ratio=self._mon_ratio,
scaler_type=self._kwargs.get('scaler'),
is_training=is_training,
**self._data_kwargs)
for k, v in self._data.items():
if hasattr(v, 'shape'):
self._logger.info((k, v.shape))
# Build models.
scaler = self._data['scaler']
W = self._data['adj_mx']
laplacian = W / W.max()
laplacian = scipy.sparse.csr_matrix(laplacian, dtype=np.float32)
lmax = graph.lmax(laplacian)
if is_training:
self._train_model = Model(is_training=True,
laplacian=laplacian,
lmax=lmax,
batch_size=self._train_batch_size,
reuse=True,
**self._model_kwargs)
else:
self._test_model = Model(is_training=False,
laplacian=laplacian,
lmax=lmax,
batch_size=self._test_batch_size,
reuse=False,
**self._model_kwargs)
max_to_keep = self._train_kwargs.get('max_to_keep', 100)
self.saver = tf.train.Saver(tf.global_variables(),
max_to_keep=max_to_keep)
self.summary_writer = tf.summary.FileWriter(self._log_dir)
# Log model statistics.
total_trainable_parameter = utils.get_total_trainable_parameter_size()
self._logger.info('Total number of trainable parameters: {:d}'.format(
total_trainable_parameter))
for var in tf.global_variables():
self._logger.debug('{}, {}'.format(var.name, var.get_shape()))
def __init__(self, is_training=False, **kwargs):
self._kwargs = kwargs
self._alg = kwargs.get('alg')
self._data_kwargs = kwargs.get('data')
self._train_kwargs = kwargs.get('train')
self._test_kwargs = kwargs.get('test')
self._model_kwargs = kwargs.get('model')
self._base_dir = kwargs.get('base_dir')
self._epochs = self._train_kwargs.get('epochs')
# logging.
self._log_dir = get_log_dir(kwargs)
log_level = self._kwargs.get('log_level', 'INFO')
self._logger = utils.get_logger(self._log_dir,
__name__,
'info.log',
level=log_level)
self._writer = tf.summary.FileWriter(self._log_dir)
self._logger.info(kwargs)
self._mon_ratio = float(self._kwargs.get('mon_ratio'))
# Model's args
self._input_dim = int(self._model_kwargs.get('input_dim'))
self._nodes = int(self._model_kwargs.get('num_nodes'))
self._drop_out = float(self._train_kwargs.get('dropout'))
self.batch_size = int(self._data_kwargs['batch_size'])
# Test's args
self._flow_selection = self._test_kwargs.get('flow_selection')
self._run_times = self._test_kwargs.get('run_times')
# Data preparation
self._day_size = self._data_kwargs.get('day_size')
self._data = utils.load_dataset_gatlstm(
num_nodes=self._model_kwargs.get('num_nodes'),
input_dim=self._model_kwargs.get('input_dim'),
mon_ratio=self._mon_ratio,
scaler_type=self._kwargs.get('scaler'),
is_training=is_training,
**self._data_kwargs)
for k, v in self._data.items():
if hasattr(v, 'shape'):
self._logger.info((k, v.shape))
# Build models.
scaler = self._data['scaler']
if is_training:
self.model = GATLSTMModel(scaler=scaler,
batch_size=self.batch_size,
**self._model_kwargs)
else:
self.model = GATLSTMModel(scaler=scaler,
batch_size=1,
**self._model_kwargs)
# Learning rate.
max_to_keep = self._train_kwargs.get('max_to_keep', 100)
self.saver = tf.train.Saver(tf.global_variables(),
max_to_keep=max_to_keep)
self.summary_writer = tf.summary.FileWriter(self._log_dir)
# Log model statistics.
total_trainable_parameter = utils.get_total_trainable_parameter_size()
self._logger.info('Total number of trainable parameters: {:d}'.format(
total_trainable_parameter))
for var in tf.global_variables():
self._logger.debug('{}, {}'.format(var.name, var.get_shape()))
def __init__(self, args, adj_mx, **kwargs):
self._kwargs = kwargs
self._data_kwargs = kwargs.get('data')
self._model_kwargs = kwargs.get('model')
self._train_kwargs = kwargs.get('train')
# logging.
self._log_dir = self._get_log_dir(kwargs)
log_level = self._kwargs.get('log_level', 'INFO')
self._logger = utils.get_logger(self._log_dir,
__name__,
kwargs['name'] + '_info.log',
level=log_level)
self._writer = tf.summary.FileWriter(self._log_dir)
self._logger.info(kwargs)
# Data preparation
if self._data_kwargs.get('data_type') == 'npz':
self._data = utils.load_dataset(**self._data_kwargs)
elif self._data_kwargs.get('data_type') == 'csv':
self._data = utils.load_dataset_from_csv(**self._data_kwargs)
for k, v in self._data.items():
if hasattr(v, 'shape'):
self._logger.info((k, v.shape))
# Build models.
scaler = self._data['scaler']
with tf.name_scope('Train'):
with tf.variable_scope('DCRNN', reuse=False):
self._train_model = DCRNNModel(
args=args,
is_training=True,
scaler=scaler,
batch_size=self._data_kwargs['batch_size'],
adj_mx=adj_mx,
**self._model_kwargs)
with tf.name_scope('Test'):
with tf.variable_scope('DCRNN', reuse=True):
self._test_model = DCRNNModel(
args=args,
is_training=False,
scaler=scaler,
batch_size=self._data_kwargs['test_batch_size'],
adj_mx=adj_mx,
**self._model_kwargs)
# Learning rate.
self._lr = tf.get_variable('learning_rate',
shape=(),
initializer=tf.constant_initializer(0.01),
trainable=False)
self._new_lr = tf.placeholder(tf.float32,
shape=(),
name='new_learning_rate')
self._lr_update = tf.assign(self._lr, self._new_lr, name='lr_update')
# Configure optimizer
optimizer_name = self._train_kwargs.get('optimizer', 'adam').lower()
epsilon = float(self._train_kwargs.get('epsilon', 1e-3))
optimizer = tf.train.AdamOptimizer(self._lr, epsilon=epsilon)
if optimizer_name == 'sgd':
optimizer = tf.train.GradientDescentOptimizer(self._lr, )
elif optimizer_name == 'amsgrad':
optimizer = AMSGrad(self._lr, epsilon=epsilon)
# Calculate loss
output_dim = self._model_kwargs.get('output_dim')
preds = self._train_model.outputs
labels = self._train_model.labels[..., :output_dim]
null_val = 0.
self._loss_fn = masked_mae_loss(scaler, null_val)
self._mape_fn = masked_mape(scaler, null_val)
self._rmse_fn = masked_rmse_loss(scaler, null_val)
self._train_loss = self._loss_fn(preds=preds, labels=labels)
tvars = tf.trainable_variables()
grads = tf.gradients(self._train_loss, tvars)
max_grad_norm = kwargs['train'].get('max_grad_norm', 1.)
grads, _ = tf.clip_by_global_norm(grads, max_grad_norm)
global_step = tf.train.get_or_create_global_step()
self._train_op = optimizer.apply_gradients(zip(grads, tvars),
global_step=global_step,
name='train_op')
max_to_keep = self._train_kwargs.get('max_to_keep', 100)
self._epoch = 0
self._saver = tf.train.Saver(tf.global_variables(),
max_to_keep=max_to_keep)
# Log model statistics.
total_trainable_parameter = utils.get_total_trainable_parameter_size()
self._logger.info('Total number of trainable parameters: {:d}'.format(
total_trainable_parameter))
for var in tf.global_variables():
self._logger.debug('{}, {}'.format(var.name, var.get_shape()))
def __init__(self, **kwargs):
self._kwargs = kwargs
self._data_kwargs = kwargs.get('data')
self._model_kwargs = kwargs.get('model')
self._train_kwargs = kwargs.get('train')
# logging.
self._log_dir = self._get_log_dir(kwargs)
log_level = self._kwargs.get('log_level', 'INFO')
self._logger = utils.get_logger(self._log_dir,
__name__,
'info.log',
level=log_level)
self._writer = tf.summary.FileWriter(self._log_dir)
self._logger.info(kwargs)
# Data preparation
# load data set
self._data = utils.load_dataset(**self._data_kwargs)
# print(self._data.keys())
# print(len(self._data))
# print(self._data['x_train'].shape)
# print(self._data['y_train'].shape)
# print(self._data['x_val'].shape)
# print(self._data['y_val'].shape)
# print(self._data['x_test'].shape)
# print(self._data['y_test'].shape)
# exit()
# (23974, 12, 207, 2)
# (23974, 12, 207, 2)
# (3425, 12, 207, 2)
# (3425, 12, 207, 2)
# (6850, 12, 207, 2)
# (6850, 12, 207, 2)
# import our node2vec data and replace
# but how do we comply with the dimensions
# we can plant the same into every time step, but what about the same dimension
# exit()
# I think we just need to attach our node to vector here
for k, v in self._data.items():
if hasattr(v, 'shape'):
self._logger.info((k, v.shape))
# Build models.
scaler = self._data['scaler']
# scaler is the mean and standard deviation pre-computed and stored
# used to normalize data and de-normalize data
# print(scaler)
# exit()
with tf.name_scope('Train'):
# Batch size is a term used in machine learning and refers to the number of training examples utilised in one iteration.
with tf.variable_scope('DCRNN', reuse=False):
self._train_model = DCRNNModel(
is_training=True,
scaler=scaler,
batch_size=self._data_kwargs['batch_size'],
adj_matrix_file=self._data_kwargs['graph_pkl_filename'],
**self._model_kwargs)
with tf.name_scope('Test'):
with tf.variable_scope('DCRNN', reuse=True):
self._test_model = DCRNNModel(
is_training=False,
scaler=scaler,
batch_size=self._data_kwargs['test_batch_size'],
adj_matrix_file=self._data_kwargs['graph_pkl_filename'],
**self._model_kwargs)
# Learning rate.
self._lr = tf.get_variable('learning_rate',
shape=(),
initializer=tf.constant_initializer(0.01),
trainable=False)
self._new_lr = tf.placeholder(tf.float32,
shape=(),
name='new_learning_rate')
self._lr_update = tf.assign(self._lr, self._new_lr, name='lr_update')
# Configure optimizer
optimizer_name = self._train_kwargs.get('optimizer', 'adam').lower()
epsilon = float(self._train_kwargs.get('epsilon', 1e-3))
optimizer = tf.train.AdamOptimizer(self._lr, epsilon=epsilon)
if optimizer_name == 'sgd':
optimizer = tf.train.GradientDescentOptimizer(self._lr, )
elif optimizer_name == 'amsgrad':
optimizer = AMSGrad(self._lr, epsilon=epsilon)
# Calculate loss
output_dim = self._model_kwargs.get('output_dim')
# preds is a placeholder of outputs, which is a stacked tensor
preds = self._train_model.outputs
# print(preds.eval())
# exit()
# You must feed a value for placeholder tensor
labels = self._train_model.labels[..., :output_dim]
null_val = 0.
self._loss_fn = masked_mae_loss(scaler, null_val)
self._train_loss = self._loss_fn(preds=preds, labels=labels)
tvars = tf.trainable_variables()
grads = tf.gradients(self._train_loss, tvars)
max_grad_norm = kwargs['train'].get('max_grad_norm', 1.)
grads, _ = tf.clip_by_global_norm(grads, max_grad_norm)
global_step = tf.train.get_or_create_global_step()
self._train_op = optimizer.apply_gradients(zip(grads, tvars),
global_step=global_step,
name='train_op')
# print(type(self._train_op))
# exit()
max_to_keep = self._train_kwargs.get('max_to_keep', 100)
self._epoch = 0
self._saver = tf.train.Saver(tf.global_variables(),
max_to_keep=max_to_keep)
# Log model statistics.
total_trainable_parameter = utils.get_total_trainable_parameter_size()
self._logger.info('Total number of trainable parameters: {:d}'.format(
total_trainable_parameter))
for var in tf.global_variables():
self._logger.debug('{}, {}'.format(var.name, var.get_shape()))
def __init__(self, is_training=False, **kwargs):
super(DCRNNSupervisor, self).__init__(**kwargs)
self._data = utils.load_dataset_dcrnn(
seq_len=self._model_kwargs.get('seq_len'),
horizon=self._model_kwargs.get('horizon'),
input_dim=self._model_kwargs.get('input_dim'),
mon_ratio=self._mon_ratio,
scaler_type=self._kwargs.get('scaler'),
is_training=is_training,
**self._data_kwargs)
for k, v in self._data.items():
if hasattr(v, 'shape'):
self._logger.info((k, v.shape))
# Build models.
scaler = self._data['scaler']
if is_training:
self.model = DCRNNModel(scaler=scaler,
batch_size=self._data_kwargs['batch_size'],
adj_mx=self._data['adj_mx'],
**self._model_kwargs)
else:
self.model = DCRNNModel(scaler=scaler,
batch_size=1,
adj_mx=self._data['adj_mx'],
**self._model_kwargs)
# Learning rate.
self._lr = tf.get_variable('learning_rate',
shape=(),
initializer=tf.constant_initializer(0.01),
trainable=False)
self._new_lr = tf.placeholder(tf.float32,
shape=(),
name='new_learning_rate')
self._lr_update = tf.assign(self._lr, self._new_lr, name='lr_update')
# Configure optimizer
optimizer_name = self._train_kwargs.get('optimizer', 'adam').lower()
epsilon = float(self._train_kwargs.get('epsilon', 1e-3))
optimizer = tf.train.AdamOptimizer(
self._lr,
epsilon=epsilon,
)
if optimizer_name == 'sgd':
optimizer = tf.train.GradientDescentOptimizer(self._lr, )
elif optimizer_name == 'amsgrad':
optimizer = AMSGrad(self._lr, epsilon=epsilon)
# Calculate loss
output_dim = self._model_kwargs.get('output_dim')
preds = self.model.outputs
labels = self.model.labels[..., :output_dim]
null_val = 0.
self._loss_fn = masked_mse_loss(scaler, null_val)
# self._loss_fn = masked_mae_loss(scaler, null_val)
self._train_loss = self._loss_fn(preds=preds, labels=labels)
tvars = tf.trainable_variables()
grads = tf.gradients(self._train_loss, tvars)
max_grad_norm = kwargs['train'].get('max_grad_norm', 1.)
grads, _ = tf.clip_by_global_norm(grads, max_grad_norm)
global_step = tf.train.get_or_create_global_step()
self._train_op = optimizer.apply_gradients(zip(grads, tvars),
global_step=global_step,
name='train_op')
max_to_keep = self._train_kwargs.get('max_to_keep', 100)
self._epoch = 0
self._saver = tf.train.Saver(tf.global_variables(),
max_to_keep=max_to_keep)
# Log model statistics.
total_trainable_parameter = utils.get_total_trainable_parameter_size()
self._logger.info('Total number of trainable parameters: {:d}'.format(
total_trainable_parameter))
for var in tf.global_variables():
self._logger.debug('{}, {}'.format(var.name, var.get_shape()))
def __init__(self, sess, adj_mx, dataloaders, kwargs):
self._kwargs = kwargs
self._data_kwargs = kwargs.get('data')
self._model_kwargs = kwargs.get('model')
self._train_kwargs = kwargs.get('train')
self._paths_kwargs = kwargs.get('paths')
self._save_tensors = kwargs.get('tf_config').get('save_tensors', False) \
if kwargs.get('tf_config') else False
self._trace = kwargs.get('tf_config').get('trace', False) \
if kwargs.get('tf_config') else False
self._save_graph = kwargs.get('tf_config').get('save_graph', False) \
if kwargs.get('tf_config') else False
self._log_dir = self._get_log_dir(kwargs)
self._writer = tf.summary.FileWriter(self._log_dir, sess.graph) \
if self._save_graph else tf.summary.FileWriter(self._log_dir)
# Data preparation
self._data = dataloaders
# for k, v in self._data.items():
# if hasattr(v, 'shape'):
# self._kwargs.logger.info((k, v.shape))
# Build models.
scaler = self._data['scaler']
with tf.name_scope('Train'):
with tf.variable_scope('DCRNN', reuse=False):
train_batch_size = dataloaders['train_loader'].batch_size
self._train_model = DCRNNModel(is_training=True,
scaler=scaler,
batch_size=train_batch_size,
adj_mx=adj_mx,
**self._model_kwargs)
with tf.name_scope('Val'):
with tf.variable_scope('DCRNN', reuse=True):
val_batch_size = dataloaders['val_loader'].batch_size
self._val_model = DCRNNModel(is_training=False,
scaler=scaler,
batch_size=val_batch_size,
adj_mx=adj_mx,
**self._model_kwargs)
with tf.name_scope('Test'):
with tf.variable_scope('DCRNN', reuse=True):
test_batch_size = dataloaders['test_loader'].batch_size
self._test_model = DCRNNModel(is_training=False,
scaler=scaler,
batch_size=test_batch_size,
adj_mx=adj_mx,
**self._model_kwargs)
# Learning rate.
self._lr = tf.get_variable('learning_rate',
shape=(),
initializer=tf.constant_initializer(0.01),
trainable=False)
self._new_lr = tf.placeholder(tf.float32,
shape=(),
name='new_learning_rate')
self._lr_update = tf.assign(self._lr, self._new_lr, name='lr_update')
# Configure optimizer
optimizer_name = self._train_kwargs.get('optimizer', 'adam').lower()
epsilon = float(self._train_kwargs.get('epsilon', 1e-3))
optimizer = tf.train.AdamOptimizer(self._lr, epsilon=epsilon)
if optimizer_name == 'sgd':
optimizer = tf.train.GradientDescentOptimizer(self._lr, )
elif optimizer_name == 'amsgrad':
optimizer = AMSGrad(self._lr, epsilon=epsilon)
# Calculate loss
output_dim = self._model_kwargs.get('output_dim')
preds = self._train_model.outputs
labels = self._train_model.labels[..., :output_dim]
null_val = 0. if kwargs['model'].get('exclude_zeros_in_metric',
True) else np.nan
loss_func_dict = {
'mae': masked_mae_loss(scaler, null_val),
'rmse': masked_rmse_loss(scaler, null_val),
'mse': masked_mse_loss(scaler, null_val)
}
self._loss_fn = loss_func_dict.get(kwargs['train'].get(
'loss_func', 'mae'))
self._metric_fn = loss_func_dict.get(kwargs['train'].get(
'metric_func', 'mae'))
self._train_loss = self._loss_fn(preds=preds, labels=labels)
tvars = tf.trainable_variables()
grads = tf.gradients(self._train_loss, tvars)
max_grad_norm = kwargs['train'].get('max_grad_norm', 1.)
grads, _ = tf.clip_by_global_norm(grads, max_grad_norm)
self._train_op = optimizer.apply_gradients(
zip(grads, tvars),
global_step=tf.train.get_or_create_global_step(),
name='train_op')
max_to_keep = self._train_kwargs.get('max_to_keep', 100)
self._saver = tf.train.Saver(tf.global_variables(),
max_to_keep=max_to_keep)
# load model
model_filename = self._paths_kwargs.get('model_filename')
if model_filename is not None:
self._saver.restore(sess, model_filename)
self._kwargs.logger.info(
'Pretrained model was loaded from : {}'.format(model_filename))
else:
sess.run(tf.global_variables_initializer())
# Log model statistics.
total_trainable_parameter = utils.get_total_trainable_parameter_size()
self._kwargs.logger.info('Total number of trainable parameters: {:d}'.\
format(total_trainable_parameter))
for var in tf.global_variables():
self._kwargs.logger.debug('{}, {}'.format(var.name,
var.get_shape()))
def __init__(self, kwargs):
self._kwargs = kwargs
self._data_kwargs = kwargs.get('data')
self._model_kwargs = kwargs.get('model')
self._train_kwargs = kwargs.get('train')
# logging.
self._log_dir = self._get_log_dir(kwargs)
log_level = self._kwargs.get('log_level', 'INFO')
self._logger = utils.get_logger(self._log_dir,
__name__,
'info.log',
level=log_level)
self._writer = tf.summary.FileWriter(self._log_dir)
self._logger.info(kwargs)
# Data preparation
self.batch_size = self._data_kwargs.get('batch_size')
self.val_batch_size = self._data_kwargs['val_batch_size']
self.test_batch_size = 1
self.horizon = self._model_kwargs.get('horizon')
self.seq_len = self._model_kwargs.get('seq_len')
self.validation_ratio = self._data_kwargs['validation_ratio']
self.test_ratio = self._data_kwargs['test_ratio']
sensor_filename = self._data_kwargs['sensor_filename']
self.sensor_df = pd.read_csv(sensor_filename)
distance_filename = self._data_kwargs['distance_filename']
self.dist_df = pd.read_csv(distance_filename)
partition_filename = self._data_kwargs['partition_filename']
self.partition = np.genfromtxt(partition_filename,
dtype=int,
delimiter="\n",
unpack=False)
self.clusters = np.unique(self.partition)
self.max_node = max(np.bincount(self.partition))
self._model_kwargs['num_nodes'] = self.max_node
# Build models.
#scaler = self._data['scaler']
with tf.name_scope('Train'):
with tf.variable_scope('DCRNN', reuse=False):
self._train_model = DCRNNModel(
is_training=True,
batch_size=self._data_kwargs['batch_size'],
**self._model_kwargs)
with tf.name_scope('Test'):
with tf.variable_scope('DCRNN', reuse=True):
self._test_model = DCRNNModel(
is_training=False,
batch_size=self._data_kwargs['test_batch_size'],
**self._model_kwargs)
# Learning rate.
self._lr = tf.get_variable('learning_rate',
shape=(),
initializer=tf.constant_initializer(0.01),
trainable=False)
self._new_lr = tf.placeholder(tf.float32,
shape=(),
name='new_learning_rate')
self._lr_update = tf.assign(self._lr, self._new_lr, name='lr_update')
# Configure optimizer
optimizer_name = self._train_kwargs.get('optimizer', 'adam').lower()
epsilon = float(self._train_kwargs.get('epsilon', 1e-3))
optimizer = tf.train.AdamOptimizer(self._lr, epsilon=epsilon)
if optimizer_name == 'sgd':
optimizer = tf.train.GradientDescentOptimizer(self._lr, )
# Calculate loss
output_dim = self._model_kwargs.get('output_dim')
preds = self._train_model.outputs
labels = self._train_model.labels[..., :output_dim]
self.preds_test = self._test_model.outputs
self.labels_test = self._test_model.labels[..., :output_dim]
null_val = 0.
self._loss_fn = masked_mae_loss(null_val)
# self._loss_fn = masked_mae_loss(scaler, null_val)
self._train_loss = self._loss_fn(preds=preds, labels=labels)
#print('output labels', labels.shape)
self._test_loss = self._loss_fn(preds=self.preds_test,
labels=self.labels_test)
tvars = tf.trainable_variables()
grads = tf.gradients(self._train_loss, tvars)
max_grad_norm = kwargs['train'].get('max_grad_norm', 1.)
grads, _ = tf.clip_by_global_norm(grads, max_grad_norm)
global_step = tf.train.get_or_create_global_step()
self._train_op = optimizer.apply_gradients(zip(grads, tvars),
global_step=global_step,
name='train_op')
max_to_keep = self._train_kwargs.get('max_to_keep', 100)
self._epoch = 0
self._saver = tf.train.Saver(tf.global_variables(),
max_to_keep=max_to_keep)
# Log model statistics.
total_trainable_parameter = utils.get_total_trainable_parameter_size()
self._logger.info('Total number of trainable parameters: {:d}'.format(
total_trainable_parameter))
for var in tf.global_variables():
self._logger.debug('{}, {}'.format(var.name, var.get_shape()))
def __init__(self, **kwargs):
self._kwargs = kwargs
self._data_kwargs = kwargs.get('data')
self._model_kwargs = kwargs.get('model')
self._train_kwargs = kwargs.get('train')
self._test_kwargs = kwargs.get('test')
# logging.
self._log_dir = _get_log_dir(kwargs)
log_level = self._kwargs.get('log_level', 'INFO')
self._logger = utils.get_logger(self._log_dir,
__name__,
'info.log',
level=log_level)
self._writer = tf.summary.FileWriter(self._log_dir)
self._logger.info(kwargs)
self._mon_ratio = float(self._kwargs.get('mon_ratio'))
# Model's args
self._seq_len = int(self._model_kwargs.get('seq_len'))
self._horizon = int(self._model_kwargs.get('horizon'))
self._input_dim = int(self._model_kwargs.get('input_dim'))
self._nodes = int(self._model_kwargs.get('num_nodes'))
self._r = int(self._model_kwargs.get('r'))
# Test's args
self._flow_selection = self._test_kwargs.get('flow_selection')
self._run_times = self._test_kwargs.get('run_times')
self._lamda = []
self._lamda.append(self._test_kwargs.get('lamda_0'))
self._lamda.append(self._test_kwargs.get('lamda_1'))
self._lamda.append(self._test_kwargs.get('lamda_2'))
# Data preparation
self._day_size = self._data_kwargs.get('day_size')
self.data = utils.load_dataset_dcrnn_fwbw(
seq_len=self._model_kwargs.get('seq_len'),
horizon=self._model_kwargs.get('horizon'),
input_dim=self._model_kwargs.get('input_dim'),
mon_ratio=self._mon_ratio,
**self._data_kwargs)
for k, v in self.data.items():
if hasattr(v, 'shape'):
self._logger.info((k, v.shape))
# Build models.
scaler = self.data['scaler']
with tf.name_scope('Train'):
with tf.variable_scope('DCRNN', reuse=False):
self.train_model = DCRNNModel(
is_training=True,
scaler=scaler,
batch_size=self._data_kwargs['batch_size'],
adj_mx=self.data['adj_mx'],
**self._model_kwargs)
with tf.name_scope('Val'):
with tf.variable_scope('DCRNN', reuse=True):
self.val_model = DCRNNModel(
is_training=False,
scaler=scaler,
batch_size=self._data_kwargs['val_batch_size'],
adj_mx=self.data['adj_mx'],
**self._model_kwargs)
with tf.name_scope('Eval'):
with tf.variable_scope('DCRNN', reuse=True):
self.eval_model = DCRNNModel(
is_training=False,
scaler=scaler,
batch_size=self._data_kwargs['eval_batch_size'],
adj_mx=self.data['adj_mx'],
**self._model_kwargs)
with tf.name_scope('Test'):
with tf.variable_scope('DCRNN', reuse=True):
self.test_model = DCRNNModel(
is_training=False,
scaler=scaler,
batch_size=self._data_kwargs['test_batch_size'],
adj_mx=self.data['adj_mx'],
**self._model_kwargs)
# Learning rate.
self._lr = tf.get_variable('learning_rate',
shape=(),
initializer=tf.constant_initializer(0.01),
trainable=False)
self._new_lr = tf.placeholder(tf.float32,
shape=(),
name='new_learning_rate')
self._lr_update = tf.assign(self._lr, self._new_lr, name='lr_update')
# Configure optimizer
optimizer_name = self._train_kwargs.get('optimizer', 'adam').lower()
epsilon = float(self._train_kwargs.get('epsilon', 1e-3))
optimizer = tf.train.AdamOptimizer(
self._lr,
epsilon=epsilon,
)
if optimizer_name == 'sgd':
optimizer = tf.train.GradientDescentOptimizer(self._lr, )
elif optimizer_name == 'amsgrad':
optimizer = AMSGrad(self._lr, epsilon=epsilon)
# Calculate loss
output_dim = self._model_kwargs.get('output_dim')
# fw decoder
preds_fw = self.train_model.outputs_fw
labels_fw = self.train_model.labels_fw[..., :output_dim]
# bw encoder
enc_preds_bw = self.train_model.enc_outputs_bw
enc_labels_bw = self.train_model.enc_labels_bw[..., :output_dim]
null_val = 0.
self._loss_fn = masked_mse_loss(scaler, null_val)
self._train_loss_dec = self._loss_fn(preds=preds_fw, labels=labels_fw)
# backward loss
self._train_loss_enc_bw = self._loss_fn(preds=enc_preds_bw,
labels=enc_labels_bw)
self._train_loss = self._train_loss_dec + self._train_loss_enc_bw
tvars = tf.trainable_variables()
grads = tf.gradients(self._train_loss, tvars)
max_grad_norm = kwargs['train'].get('max_grad_norm', 1.)
grads, _ = tf.clip_by_global_norm(grads, max_grad_norm)
global_step = tf.train.get_or_create_global_step()
self._train_op = optimizer.apply_gradients(zip(grads, tvars),
global_step=global_step,
name='train_op')
max_to_keep = self._train_kwargs.get('max_to_keep', 100)
self._epoch = 0
self._saver = tf.train.Saver(tf.global_variables(),
max_to_keep=max_to_keep)
# Log model statistics.
total_trainable_parameter = utils.get_total_trainable_parameter_size()
self._logger.info('Total number of trainable parameters: {:d}'.format(
total_trainable_parameter))
for var in tf.global_variables():
self._logger.debug('{}, {}'.format(var.name, var.get_shape()))
def __init__(self, adj_mx, **kwargs):
self._kwargs = kwargs
self._data_kwargs = kwargs.get('data')
self._model_kwargs = kwargs.get('model')
self._train_kwargs = kwargs.get('train')
# logging.
self._log_dir = self._get_log_dir(kwargs)
log_level = self._kwargs.get('log_level', 'INFO')
self._logger = utils.get_logger(self._log_dir,
__name__,
'info.log',
level=log_level)
self._writer = tf.summary.FileWriter(self._log_dir)
self._logger.info(kwargs)
# Data preparation
self._data = utils.load_dataset(
**self._data_kwargs) # return 3 DataLoaders and 1 scaler
for k, v in self._data.items():
if hasattr(v, 'shape'):
self._logger.info((k, v.shape))
# Build models.
scaler = self._data['scaler']
with tf.name_scope(
'Train'
): # reuse、variable_scope讲解：https://www.jianshu.com/p/ab0d38725f88
with tf.variable_scope(
'DCRNN', reuse=False): # reuse==False的含义: 该作用域下创建的变量不会重用
self._train_model = DCRNNModel(
is_training=True,
scaler=scaler,
batch_size=self._data_kwargs['batch_size'],
adj_mx=adj_mx,
**self._model_kwargs)
with tf.name_scope('Test'):
with tf.variable_scope('DCRNN', reuse=True): # 测试时创建的变量可以重用
self._test_model = DCRNNModel(
is_training=False,
scaler=scaler,
batch_size=self._data_kwargs['test_batch_size'],
adj_mx=adj_mx,
**self._model_kwargs)
# Learning rate.
self._lr = tf.get_variable('learning_rate',
shape=(),
initializer=tf.constant_initializer(0.01),
trainable=False)
self._new_lr = tf.placeholder(tf.float32,
shape=(),
name='new_learning_rate')
self._lr_update = tf.assign(self._lr, self._new_lr, name='lr_update')
# Configure optimizer
optimizer_name = self._train_kwargs.get('optimizer',
'adam').lower() # 默认是'adam'
epsilon = float(self._train_kwargs.get('epsilon', 1e-3))
optimizer = tf.train.AdamOptimizer(self._lr, epsilon=epsilon)
if optimizer_name == 'sgd':
optimizer = tf.train.GradientDescentOptimizer(self._lr, )
elif optimizer_name == 'amsgrad':
optimizer = AMSGrad(self._lr, epsilon=epsilon)
# Calculate loss
output_dim = self._model_kwargs.get(
'output_dim') # output_dim在配置文件里写的1，指只预测speed这一个特征
preds = self._train_model.outputs
labels = self._train_model.labels[..., :output_dim]
null_val = 0.
self._loss_fn = masked_mae_loss(scaler, null_val)
self._train_loss = self._loss_fn(preds=preds, labels=labels)
tvars = tf.trainable_variables()
grads = tf.gradients(self._train_loss, tvars)
max_grad_norm = kwargs['train'].get('max_grad_norm', 1.)
grads, _ = tf.clip_by_global_norm(
grads, max_grad_norm
) # 在一次迭代更新中，所有权重的梯度的平方和在一个设定范围以内，这个范围就是clip_gradient.
global_step = tf.train.get_or_create_global_step()
self._train_op = optimizer.apply_gradients(zip(grads, tvars),
global_step=global_step,
name='train_op')
max_to_keep = self._train_kwargs.get('max_to_keep', 100)
self._epoch = 0
self._saver = tf.train.Saver(
tf.global_variables(),
max_to_keep=max_to_keep) # Saver将保存最近的max_to_keep个模型
# Log model statistics.
total_trainable_parameter = utils.get_total_trainable_parameter_size()
self._logger.info('Total number of trainable parameters: {:d}'.format(
total_trainable_parameter))
for var in tf.global_variables():
self._logger.debug('{}, {}'.format(var.name, var.get_shape()))
def __init__(self, **kwargs):
self._kwargs = kwargs
self._data_kwargs = kwargs.get('data')
self._train_kwargs = kwargs.get('train')
self._test_kwargs = kwargs.get('test')
self._model_kwargs = kwargs.get('model')
self._alg_name = self._kwargs.get('alg')
# logging.
self._log_dir = self._get_log_dir(kwargs)
log_level = self._kwargs.get('log_level', 'INFO')
self._logger = utils.get_logger(self._log_dir,
__name__,
'info.log',
level=log_level)
self._writer = tf.summary.FileWriter(self._log_dir)
self._logger.info(kwargs)
# Data's args
self._day_size = self._data_kwargs.get('day_size')
# Model's Args
self._rnn_units = self._model_kwargs.get('rnn_units')
self._seq_len = self._model_kwargs.get('seq_len')
self._horizon = self._model_kwargs.get('horizon')
self._input_dim = self._model_kwargs.get('input_dim')
self._input_shape = (self._seq_len, self._input_dim)
self._output_dim = self._model_kwargs.get('output_dim')
self._nodes = self._model_kwargs.get('num_nodes')
self._n_rnn_layers = self._model_kwargs.get('n_rnn_layers')
# Train's args
self._drop_out = self._train_kwargs.get('dropout')
self._epochs = self._train_kwargs.get('epochs')
self._batch_size = self._data_kwargs.get('batch_size')
# Test's args
self._run_times = self._test_kwargs.get('run_times')
self._flow_selection = self._test_kwargs.get('flow_selection')
self._test_size = self._test_kwargs.get('test_size')
self._results_path = self._test_kwargs.get('results_path')
self._mon_ratio = self._kwargs.get('mon_ratio')
# Load data
self._data = utils.load_dataset_lstm(seq_len=self._seq_len,
horizon=self._horizon,
input_dim=self._input_dim,
mon_ratio=self._mon_ratio,
test_size=self._test_size,
**self._data_kwargs)
for k, v in self._data.items():
if hasattr(v, 'shape'):
self._logger.info((k, v.shape))
scaler = self._data['scaler']
# Model
with tf.name_scope('Train'):
with tf.variable_scope('LSTM', reuse=False):
self._train_model = EncoderDecoderLSTM(
is_training=True,
scaler=self._data['scaler'],
batch_size=self._batch_size,
**self._model_kwargs)
with tf.name_scope('Val'):
with tf.variable_scope('LSTM', reuse=True):
self._val_model = EncoderDecoderLSTM(
is_training=True,
scaler=self._data['scaler'],
batch_size=self._data_kwargs['val_batch_size'],
**self._model_kwargs)
with tf.name_scope('Eval'):
with tf.variable_scope('LSTM', reuse=True):
self._eval_model = EncoderDecoderLSTM(
is_training=True,
scaler=self._data['scaler'],
batch_size=self._data_kwargs['eval_batch_size'],
**self._model_kwargs)
with tf.name_scope('Test'):
with tf.variable_scope('LSTM', reuse=True):
self._test_model = EncoderDecoderLSTM(
is_training=True,
scaler=self._data['scaler'],
batch_size=self._data_kwargs['test_batch_size'],
**self._model_kwargs)
# Learning rate
self._lr = tf.get_variable('learning_rate',
shape=(),
initializer=tf.constant_initializer(0.01),
trainable=False)
self._new_lr = tf.placeholder(tf.float32,
shape=(),
name='new_learning_rate')
self._lr_update = tf.assign(self._lr, self._new_lr, name='lr_update')
# Configure optimizer
optimizer_name = self._train_kwargs.get('optimizer', 'adam').lower()
epsilon = float(self._train_kwargs.get('epsilon', 1e-3))
optimizer = tf.train.AdamOptimizer(
self._lr,
epsilon=epsilon,
)
if optimizer_name == 'sgd':
optimizer = tf.train.GradientDescentOptimizer(self._lr, )
elif optimizer_name == 'amsgrad':
optimizer = AMSGrad(self._lr, epsilon=epsilon)
# Calculate loss
output_dim = self._model_kwargs.get('output_dim')
preds = self._train_model.outputs
labels = self._train_model.labels[..., :output_dim]
null_val = 0.
self._loss_fn = masked_mse_loss(scaler, null_val)
# self._loss_fn = masked_mae_loss(scaler, null_val)
self._train_loss = self._loss_fn(preds=preds, labels=labels)
tvars = tf.trainable_variables()
grads = tf.gradients(self._train_loss, tvars)
max_grad_norm = kwargs['train'].get('max_grad_norm', 1.)
grads, _ = tf.clip_by_global_norm(grads, max_grad_norm)
global_step = tf.train.get_or_create_global_step()
self._train_op = optimizer.apply_gradients(zip(grads, tvars),
global_step=global_step,
name='train_op')
max_to_keep = self._train_kwargs.get('max_to_keep', 100)
self._epoch = 0
self._saver = tf.train.Saver(tf.global_variables(),
max_to_keep=max_to_keep)
# Log model statistics.
total_trainable_parameter = utils.get_total_trainable_parameter_size()
self._logger.info('Total number of trainable parameters: {:d}'.format(
total_trainable_parameter))
for var in tf.global_variables():
self._logger.debug('{}, {}'.format(var.name, var.get_shape()))
