instead of dict[int, array], consider using a dataclass where each layer is an attribute. dataclasses can be used in numba as described here: https://github.com/numba/numba/issues/4037
possible advantage
can add more attributes, such as the identifiers of layers for which the node has at least one edge. this could potentially speed up the generation of the walks
instead of
dict[int, array], consider using a dataclass where each layer is an attribute. dataclasses can be used in numba as described here: https://github.com/numba/numba/issues/4037possible advantage