GNN.__call__

GNN.__call__(graph, get_info=False)[source]

Processes a given graph through a sequence of steps: normalization, encoding, coupling, and decoding. The method applies a series of transformations to the input graph and returns a decoded graph / array along with optional processing information.

Parameters:

  • graph (JaxGraph) – The input graph to be processed.

  • get_info (bool) – A boolean indicating whether detailed processing information should be returned. Defaults to False.

Returns:

A tuple consisting of the processed decoded graph / array and an optional dictionary with detailed information about each processing step if get_info is True.

Return type:

tuple[JaxGraph | Array, dict]