grafx.utils
- get_node_ids_from_type(G: GRAFX, node_type: str)
Retrieves the node IDs for of a specific type in the graph.
- Parameters:
G (
GRAFX) – The target graph.node_type (
str) – The node type to retrieve.
- Returns:
A list of node IDs that match the given type.
- Return type:
List[int]
- count_nodes_per_type(G: GRAFX, types_to_count: list | None = None)
Counts the number of nodes for each specified type in the graph.
- Parameters:
G (
GRAFX) – The target graph.types_to_count (
list, optional) – A list of node types to count. IfNone, counts all types present in the graph (default:None).
- Returns:
A dictionary with node types as keys and counts as values.
- Return type:
Dict[str, int]
- create_empty_parameters(processors, G, std=0.01)
Creates and initializes parameter tensors in a nested dictionary format from a given graph and processors. The tensors values are sampled from a normal distribution \(\mathcal{N}(0, \sigma^2)\), where the standard deviation \(\sigma\) is given by the
stdargument.- Parameters:
processors (
Mappings) – A dictionary of processors, eitherdictornn.ModuleDict, where keys are node types and values are processors.G (
GRAFX) – The graph containing nodes whose parameters are to be initialized.std (
float, optional) – Standard deviation for the parameter initialization (default:0.01).
- Returns:
A module dictionary with initialized parameters for each node type in the graph.
- Return type:
nn.ModuleDict
- permute_grafx_tensor(G_t, node_id, node_attrs=['node_types', 'rendering_orders'], id_attrs=['edge_indices'])
Permutes the node and edge attributes of a given
GRAFXTensoraccording to a given node ordering. Attributes that are not provided in thenode_attrsorid_attrsare left unchanged.- Parameters:
G_t (
GRAFXTensor) – The graph tensor to permute.node_id (
LongTensor) – The permutation index given by the node IDs.node_attrs (
List[str], optional) – List of node attributes to permute (default:["node_types", "rendering_orders"]).id_attrs (
List[str], optional) – List of attributes that contain node IDs (default:["edge_indices"]).
- Returns:
The permuted graph tensor.
- Return type: