GraphBuilder

class GraphBuilder(pos, edge_index, x=None, edge_attr=False, custom_edge_func=None, loop=True, **kwargs)

Bases: object

A class that allows an easy definition of Graph instances.

Compute the edge attributes and create a new instance of the Graph class.

Parameters:

• pos (torch.Tensor or LabelTensor) – A tensor of shape (N, D) representing the positions of N points in D-dimensional space.

• edge_index (torch.Tensor) – A tensor of shape (2, E) representing the indices of the graph’s edges.

• x (torch.Tensor | LabelTensor, optional) – Optional tensor of node features of shape (N, F), where F is the number of features per node.

• edge_attr (bool) – Whether to compute the edge attributes.

• custom_edge_func (Callable, optional) – A custom function to compute edge attributes. If provided, overrides edge_attr.

• loop (bool) – Whether to include self-loops.

• kwargs – Additional keyword arguments passed to the Graph class constructor.

Returns:

A Graph instance constructed using the provided information.

Return type:

Graph

static _create_edge_attr(pos, edge_index, edge_attr, func)

Create the edge attributes based on the input parameters.

Parameters:

• pos (torch.Tensor | LabelTensor) – Positions of the points.

• edge_index (torch.Tensor) – Edge indices.

• edge_attr (bool) – Whether to compute the edge attributes.

• func (Callable) – Function to compute the edge attributes.

Raises:

ValueError – If func is not a function.

Returns:

The edge attributes.

Return type:

torch.Tensor | LabelTensor | None

static _build_edge_attr(pos, edge_index)

Default function to compute the edge attributes.

Parameters:

• pos (torch.Tensor | LabelTensor) – Positions of the points.

• edge_index (torch.Tensor) – Edge indices.

Returns:

The edge attributes.

Return type:

torch.Tensor