GlobalMeanPool#

class deeplay.models.gnn.gtogmpm.GlobalMeanPool(*args, **kwargs)#

Bases: DeeplayModule

Global mean pooling layer for Graph Neural Networks.

Constraints#

  • Inputs:

    • x: torch.Tensor of shape (num_nodes, num_features)

    • batch: torch.Tensor of shape (num_nodes,)

    Inputs can be passed to the forward method as a tuple or as separate arguments.

  • Output: torch.Tensor of shape (batch_size, num_features)

Examples#

>>> # Global mean pooling layer
>>> layer = GlobalMeanPool().create()
>>> # Define input as a tuple of node features and batch
>>> x = torch.randn(10, 16)
>>> batch = torch.Tensor([0, 0, 0, 1, 1, 1, 2, 2, 2, 2]).long()
>>> out1 = layer((x, batch))
>>> # Define input as separate arguments
>>> out2 = layer(x, batch)
>>> torch.allclose(out1, out2)
True
>>> out1.shape
torch.Size([3, 16])

Methods Summary

forward(x)

Define the computation performed at every call.

Methods Documentation

forward(x)#

Define 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.