Skip to content

DGBatch

DGBatch dataclass 

DGBatch(
    edge_src: Tensor,
    edge_dst: Tensor,
    edge_time: Tensor,
    edge_x: Optional[Tensor] = None,
    edge_type: Optional[Tensor] = None,
    node_x_time: Optional[Tensor] = None,
    node_x_nids: Optional[Tensor] = None,
    node_x: Optional[Tensor] = None,
    node_y_time: Optional[Tensor] = None,
    node_y_nids: Optional[Tensor] = None,
    node_y: Optional[Tensor] = None,
)

Container for a batch of events/materialized data from a DGraph.

Each DGBatch holds edge and node information for a slice of a dynamic graph, including optional dynamic node features and edge features. Hooks read and write additional attributes to the container transparently during dataloading.

Parameters:

  • edge_src (Tensor) –

    Source node indices for edges in the batch. Shape (E,).

  • edge_dst (Tensor) –

    Destination node indices for edges in the batch. Shape (E,).

  • edge_time (Tensor) –

    Timestamps of each edge event. Shape (E,).

  • edge_x (Tensor | None, default: None ) –

    Edge features for the batch. Tensor of shape (T x V x V x d_edge).

  • edge_type (Tensor | None, default: None ) –

    Type of each edge. Shape (E,)

  • node_x (Tensor | None, default: None ) –

    Dynamic node features for nodes in the batch. Tensor of shape (T x V x d_node_dynamic).

  • node_x_time (Tensor | None, default: None ) –

    Timestamps corresponding to dynamic node features.

  • node_x_nids (Tensor | None, default: None ) –

    Node IDs corresponding to dynamic node features.

  • node_y (Tensor | None, default: None ) –

    Dynamic node labels for nodes in the batch. Tensor of shape (T x V x d_node_labels).

  • node_y_time (Tensor | None, default: None ) –

    Timestamps corresponding to dynamic node labels.

  • node_y_nids (Tensor | None, default: None ) –

    Node IDs corresponding to dynamic node labels.