libcity.model.road_representation

class libcity.model.road_representation.ChebConv(config, data_feature)[源代码]

基类：libcity.model.abstract_traffic_state_model.AbstractTrafficStateModel

calculate_loss(batch)[源代码]
参数

batch – dict, need key ‘node_features’, ‘node_labels’, ‘mask’

Returns:

forward(batch)[源代码]

自回归任务

参数

batch – dict, need key ‘node_features’ contains tensor shape=(N, feature_dim)

返回

N, feature_dim

返回类型

torch.tensor

predict(batch)[源代码]
参数

batch – dict, need key ‘node_features’

返回

torch.tensor

training: bool
class libcity.model.road_representation.DeepWalk(config, data_feature)[源代码]

基类：libcity.model.abstract_traffic_tradition_model.AbstractTraditionModel

run(data=None)[源代码]
参数

data – input of tradition model

返回

output of tradition model

class libcity.model.road_representation.GAT(config, data_feature)[源代码]

基类：libcity.model.abstract_traffic_state_model.AbstractTrafficStateModel

calculate_loss(batch)[源代码]
参数

batch – dict, need key ‘node_features’, ‘node_labels’, ‘mask’

Returns:

forward(batch)[源代码]

自回归任务 :param batch: dict, need key ‘node_features’ contains tensor shape=(N, feature_dim)

返回

N, feature_dim

返回类型

torch.tensor

predict(batch)[源代码]
参数

batch – dict, need key ‘node_features’

返回

torch.tensor

training: bool
class libcity.model.road_representation.GeomGCN(config, data_feature)[源代码]

基类：libcity.model.abstract_traffic_state_model.AbstractTrafficStateModel

calculate_loss(batch)[源代码]
参数

batch – dict, need key ‘node_features’, ‘node_labels’, ‘mask’

Returns:

forward(batch)[源代码]

自回归任务

参数

batch – dict, need key ‘node_features’ contains tensor shape=(N, feature_dim)

返回

N, output_classes

返回类型

torch.tensor

get_input(config, data_feature)[源代码]
predict(batch)[源代码]
参数

batch – dict, need key ‘node_features’

返回

torch.tensor

training: bool
class libcity.model.road_representation.LINE(config, data_feature)[源代码]

基类：libcity.model.abstract_traffic_state_model.AbstractTrafficStateModel

calculate_loss(batch)[源代码]

输入一个batch的数据，返回训练过程的loss，也就是需要定义一个loss函数

参数

batch (Batch) – a batch of input

返回

return training loss

返回类型

torch.tensor

forward(I, J)[源代码]
参数

  • I – origin indices of node i ; (B,)

  • J – origin indices of node j ; (B,)

返回

[u_j^T * u_i for (i,j) in zip(I, J)]; (B,) elif order == ‘second’:

[u’_j^T * v_i for (i,j) in zip(I, J)]; (B,)

返回类型

if order == ‘first’

training: bool
class libcity.model.road_representation.Node2Vec(config, data_feature)[源代码]

基类：libcity.model.abstract_traffic_tradition_model.AbstractTraditionModel

run(data=None)[源代码]
参数

data – input of tradition model

返回

output of tradition model