CouplingGraph

class CouplingGraph(graph, num_qudits=None, remote_edges=[], default_weight=1.0, default_remote_weight=100.0, edge_weights_overrides={})

Bases: Collection[tuple[int, int]]

A graph representing connections in a qudit topology.

__init__(graph, num_qudits=None, remote_edges=[], default_weight=1.0, default_remote_weight=100.0, edge_weights_overrides={})

Construct a new CouplingGraph.

Parameters:

• graph (CouplingGraphLike) – The undirected graph edges.

• num_qudits (int | None) – The number of qudits in the graph. If None, the number of qudits is inferred from the maximum seen in the edge list. (Default: None)

• remote_edges (Iterable[tuple[int, int]]) – The edges that cross QPU chip boundaries. Distributed QPUs will have remote links connect them. Notes, remote edges must specified both in graph and here. (Default: [])

• default_weight (float) – The default weight of an edge in the graph. (Default: 1.0)

• default_remote_weight (float) – The default weight of a remote edge in the graph. (Default: 100.0)

• edge_weights_overrides (Mapping[tuple[int, int], float]) – A mapping of edges to their weights. These override the defaults on a case-by-case basis. (Default: {})

Notes

The qudits are assumed to be numbered starting from 0.

Raises:

• ValueError – If num_qudits is too small for the edges in graph.

• ValueError – If num_qudits is less than zero.

• ValueError – If any edge in remote_edges is not in graph.

• ValueError – If any edge in edge_weights_overrides is not in graph.

Methods

all_pairs_shortest_path()	Calculate all pairs shortest path matrix using Floyd-Warshall.
all_to_all(num_qudits)	Return a coupling graph with all qudits connected.
get_individual_qpu_graphs()	Return a list of individual QPU graphs.
get_induced_subgraph(location)	Return the edges induced by the vertices specified in location.
get_neighbors_of(qudit)	Return the qudits adjacent to qudit.
get_qpu_connectivity()	Return the adjacency list of the QPUs.
get_qpu_to_qudit_map()	Return a mapping of QPU indices to qudit indices.
get_qudit_degrees()
get_qudit_to_qpu_map()	Return a mapping of qudit indices to QPU indices.
get_rooted_minimum_span(root)	Connect root to every other node in the graph.
get_shortest_path_tree(source)	Return shortest path from source to every node in self.
get_subgraph(location[, renumbering])	Returns the sub-coupling-graph with qudits in location.
get_subgraphs_of_size(size)	Find all sets of indices that form connected subgraphs on their own.
grid(num_rows, num_cols)	Return a coupling graph with a grid of qubits.
is_distributed()	Return true if the graph represents multiple connected QPUs.
is_embedded_in(graph)	Check if this CouplingGraph is embedded within graph.
is_fully_connected()	Return true if the graph is fully connected.
is_fully_connected_without(qudit)	Return true if the graph is fully connected without qudit.
is_linear()	Return true if the graph is linearly connected.
is_valid_coupling_graph(coupling_graph[, ...])	Return true if the coupling graph is valid.
linear(num_qudits)	Return a coupling graph with nearest-neighbor connectivity.
maximal_matching([edges_to_ignore, randomize])	Generate a random graph matching for the coupling graph.
qpu_count()	Return the number of connected QPUs.
relabel_subgraph([relabeling])	Renumber the vertices in graph according to the optionally provided relabeling dictionary, or relabel so that the vertices are in renumbered in least to greatest order and in the set {0,...,|V|-1}.
ring(num_qudits)	Return a coupling graph with ring connectivity.
star(num_qudits)	Return a coupling graph with one qudit connected to all else.