gnnwr.networks module

class gnnwr.networks.STNN_SPNN(STNN_insize: int, STNN_outsize, SPNN_insize: int, SPNN_outsize, activate_func=ReLU())

Bases: Module

STNN_SPNN is a neural network with dense layers, which is used to calculate the spatial proximity of two nodes and temporal proximity of two nodes at the same time. | The each layer of STNN and SPNN is as follows: | full connection layer -> activate function

Parameters:

• STNN_insize (int) – input size of STNN(must be positive)

• STNN_outsize (int) – Output size of STNN(must be positive)

• SPNN_insize (int) – input size of SPNN(must be positive)

• SPNN_outsize (int) – Output size of SPNN(must be positive)

• activate_func (torch.nn.functional) – activate function(default: nn.ReLU())

forward(input1)

Defines the computation performed at every call.

Should be overridden by all subclasses.

Note

Although the recipe for forward pass needs to be defined within this function, one should call the Module instance afterwards instead of this since the former takes care of running the registered hooks while the latter silently ignores them.

class gnnwr.networks.STPNN(dense_layer, insize, outsize, drop_out=0.2, activate_func=ReLU(), batch_norm=False)

Bases: Module

STPNN is a neural network with dense layers, which is used to calculate the spatial and temporal proximity of two nodes. | The each layer of STPNN is as follows: | full connection layer -> batch normalization layer -> activate function -> drop out layer

Parameters:

• dense_layer (list) – a list of dense layers of Neural Network

• insize (int) – input size of Neural Network(must be positive)

• outsize (int) – Output size of Neural Network(must be positive)

• drop_out (float) – drop out rate(default: 0.2)

• activate_func (torch.nn.functional) – activate function(default: nn.ReLU())

• batch_norm (bool) – whether use batch normalization(default: False)

forward(x)

Defines the computation performed at every call.

Should be overridden by all subclasses.

Note

Although the recipe for forward pass needs to be defined within this function, one should call the Module instance afterwards instead of this since the former takes care of running the registered hooks while the latter silently ignores them.

class gnnwr.networks.SWNN(dense_layer=None, insize=-1, outsize=-1, drop_out=0.2, activate_func=PReLU(num_parameters=1), batch_norm=True)

Bases: Module

SWNN is a neural network with dense layers, which is used to calculate the spatial and temporal weight of features. | The each layer of SWNN is as follows: | full connection layer -> batch normalization layer -> activate function -> drop out layer

Parameters:

• dense_layer (list) – a list of dense layers of Neural Network

• insize (int) – input size of Neural Network(must be positive)

• outsize (int) – Output size of Neural Network(must be positive)

• drop_out (float) – drop out rate(default: 0.2)

• activate_func (torch.nn.functional) – activate function(default: nn.PReLU(init=0.1))

• batch_norm (bool) – whether use batch normalization(default: True)

forward(x)

Defines the computation performed at every call.

Should be overridden by all subclasses.

Note

Although the recipe for forward pass needs to be defined within this function, one should call the Module instance afterwards instead of this since the former takes care of running the registered hooks while the latter silently ignores them.

gnnwr.networks.default_dense_layer(insize, outsize)

generate default dense layers for neural network

Parameters:

• insize (int) – input size of neural network

• outsize (int) – output size of neural network

Returns:

dense_layer – a list of dense layers of neural network

Return type:

list

gnnwr.networks.weight_share(model, x, output_size=1)

weight_share is a function to calculate the output of neural network with weight sharing.

Parameters:

• model (torch.nn.Module) – neural network with weight sharing

• x (torch.Tensor) – input of neural network

• output_size (int) – output size of neural network

Returns:

output – output of neural network

Return type:

torch.Tensor