nip.graph_isomorphism.rollout_samples.GraphIsomorphismRolloutSamples

class nip.graph_isomorphism.rollout_samples.GraphIsomorphismRolloutSamples(run_id: str, iteration: int, wandb_entity: str | None = None, wandb_project: str | None = None, silence_wandb: bool = True)

A message exchange in the graph isomorphism task.

Methods Summary

__contains__(item)
__enter__()
__exit__(exc_type, exc_value, traceback)
__getitem__(key)
__init__(run_id, iteration[, wandb_entity, ...])
__iter__()
__len__()
__repr__()	Return repr(self).
__str__()	Return str(self).
finish()	Finish the W&B run.
visualise([graph_layout_function, ...])	Visualize the rollout as a plotly graph.

Methods

__contains__(item)

__enter__()

__exit__(exc_type, exc_value, traceback)

__getitem__(key)

__init__(run_id: str, iteration: int, wandb_entity: str | None = None, wandb_project: str | None = None, silence_wandb: bool = True)

__iter__()

__len__()

__repr__()

Return repr(self).

__str__()

Return str(self).

finish()

Finish the W&B run.

visualise(graph_layout_function: callable | None = None, graph_layout_seed: int | None = None, colour_sequence: str = 'Dark24', node_text_colour: str = 'white')

Visualize the rollout as a plotly graph.

Parameters:

• graph_layout_function (callable, default=None) – A function which takes a networkx graph and returns a dictionary of node positions. Best to use a function from networkx.layout, possibly partially applied with some arguments. If None, uses networkx.spring_layout with k=4/sqrt(n), where n is the number of nodes in the graph.

• graph_layout_seed (int, default=None) – The seed to use for the graph layout function. If None, the random number generator is the RandomState instance used by numpy.random.

• colour_sequence (str, default="Dark24") – The name of the colour sequence to use to colour the nodes. Must be one of the colour sequences from plotly.express.colors.qualitative.

• node_text_colour (str, default="white") – The colour of the node labels.