tfgnn.homogeneous

December 14, 2023 · View on GitHub

Constructs a homogeneous GraphTensor with node features and one edge_set.

tfgnn.homogeneous(
    source: tf.Tensor,
    target: tf.Tensor,
    *,
    node_features: Optional[FieldOrFields] = None,
    edge_features: Optional[FieldOrFields] = None,
    context_features: Optional[FieldOrFields] = None,
    node_set_name: Optional[FieldName] = const.NODES,
    edge_set_name: Optional[FieldName] = const.EDGES,
    node_set_sizes: Optional[Field] = None,
    edge_set_sizes: Optional[Field] = None
) -> GraphTensor

Args
`source`	A dense Tensor with the source indices for edges
`target`	A dense Tensor with the target indices for edges
`node_features`	A Tensor or mapping from feature name to Tensor of features corresponding to graph nodes.
`edge_features`	Optional Tensor or mapping from feature name to Tensor of features corresponding to graph edges.
`context_features`	Optional Tensor or mapping from name to Tensor for the context (entire graph)
`node_set_name`	Optional name for the node set
`edge_set_name`	Optional name for the edge set
`node_set_sizes`	Optional Tensor with the number of nodes per component. If this is provided, edge_set_sizes should also be passed and it should be the same length.
`edge_set_sizes`	Optional Tensor with the number of edges per component. If this is provided, node_set_sizes should also be passed and it should be the same length.

Returns
A scalar `GraphTensor` with a single node set and edge set.

Args

Returns