Skip to main contentIBM Quantum Documentation
This page is from an old version of Qiskit SDK. Go to the latest version

CouplingMap

class qiskit.transpiler.CouplingMap(couplinglist=None, description=None)

GitHub

Bases: object

Directed graph specifying fixed coupling.

Nodes correspond to physical qubits (integers) and directed edges correspond to permitted CNOT gates, with source and destination corresponding to control and target qubits, respectively.

Create coupling graph. By default, the generated coupling has no nodes.

Parameters

  • couplinglist (list or None) – An initial coupling graph, specified as an adjacency list containing couplings, e.g. [[0,1], [0,2], [1,2]]. It is required that nodes are contiguously indexed starting at 0. Missed nodes will be added as isolated nodes in the coupling map.
  • description (str) – A string to describe the coupling map.

Attributes

description

graph

distance_matrix

Return the distance matrix for the coupling map.

For any qubits where there isn’t a path available between them the value in this position of the distance matrix will be math.inf.

is_symmetric

Test if the graph is symmetric.

Return True if symmetric, False otherwise

physical_qubits

Returns a sorted list of physical_qubits


Methods

add_edge

add_edge(src, dst)

Add directed edge to coupling graph.

src (int): source physical qubit dst (int): destination physical qubit

add_physical_qubit

add_physical_qubit(physical_qubit)

Add a physical qubit to the coupling graph as a node.

physical_qubit (int): An integer representing a physical qubit.

Raises

CouplingError – if trying to add duplicate qubit

compute_distance_matrix

compute_distance_matrix()

Compute the full distance matrix on pairs of nodes.

The distance map self._dist_matrix is computed from the graph using all_pairs_shortest_path_length. This is normally handled internally by the distance_matrix attribute or the distance() method but can be called if you’re accessing the distance matrix outside of those or want to pre-generate it.

connected_components

connected_components()

Separate a CouplingMap into subgraph CouplingMap for each connected component.

The connected components of a CouplingMap are the subgraphs that are not part of any larger subgraph. For example, if you had a coupling map that looked like:

0 --> 1   4 --> 5 ---> 6 --> 7
|     |
|     |
V     V
2 --> 3

then the connected components of that graph are the subgraphs:

0 --> 1
|     |
|     |
V     V
2 --> 3

and:

4 --> 5 ---> 6 --> 7

For a connected CouplingMap object there is only a single connected component, the entire CouplingMap.

This method will return a list of CouplingMap objects, one for each connected component in this CouplingMap. The data payload of each node in the graph attribute will contain the qubit number in the original graph. This will enables mapping the qubit index in a component subgraph to the original qubit in the combined CouplingMap. For example:

from qiskit.transpiler import CouplingMap
 
cmap = CouplingMap([[0, 1], [1, 2], [2, 0], [3, 4], [4, 5], [5, 3]])
component_cmaps = cmap.connected_components()
print(component_cmaps[1].graph[0])

will print 3 as index 0 in the second component is qubit 3 in the original cmap.

Returns

A list of CouplingMap objects for each connected

components. The order of this list is deterministic but implementation specific and shouldn’t be relied upon as part of the API.

Return type

list

distance

distance(physical_qubit1, physical_qubit2)

Returns the undirected distance between physical_qubit1 and physical_qubit2.

Parameters

  • physical_qubit1 (int) – A physical qubit
  • physical_qubit2 (int) – Another physical qubit

Returns

The undirected distance

Return type

int

Raises

CouplingError – if the qubits do not exist in the CouplingMap

draw

draw()

Draws the coupling map.

This function calls the graphviz_draw() function from the rustworkx package to draw the CouplingMap object.

Returns

Drawn coupling map.

Return type

PIL.Image

from_full

classmethod from_full(num_qubits, bidirectional=True)

Return a fully connected coupling map on n qubits.

Return type

CouplingMap

from_grid

classmethod from_grid(num_rows, num_columns, bidirectional=True)

Return a coupling map of qubits connected on a grid of num_rows x num_columns.

Return type

CouplingMap

from_heavy_hex

classmethod from_heavy_hex(distance, bidirectional=True)

Return a heavy hexagon graph coupling map.

A heavy hexagon graph is described in:

https://journals.aps.org/prx/abstract/10.1103/PhysRevX.10.011022

Parameters

  • distance (int) – The code distance for the generated heavy hex graph. The value for distance can be any odd positive integer. The distance relates to the number of qubits by: n=5d22d12n = \frac{5d^2 - 2d - 1}{2} where nn is the number of qubits and dd is the distance parameter.
  • bidirectional (bool) – Whether the edges in the output coupling graph are bidirectional or not. By default this is set to True

Returns

A heavy hex coupling graph

Return type

CouplingMap

from_heavy_square

classmethod from_heavy_square(distance, bidirectional=True)

Return a heavy square graph coupling map.

A heavy square graph is described in:

https://journals.aps.org/prx/abstract/10.1103/PhysRevX.10.011022

Parameters

  • distance (int) – The code distance for the generated heavy square graph. The value for distance can be any odd positive integer. The distance relates to the number of qubits by: n=3d22dn = 3d^2 - 2d where nn is the number of qubits and dd is the distance parameter.
  • bidirectional (bool) – Whether the edges in the output coupling graph are bidirectional or not. By default this is set to True

Returns

A heavy square coupling graph

Return type

CouplingMap

from_hexagonal_lattice

classmethod from_hexagonal_lattice(rows, cols, bidirectional=True)

Return a hexagonal lattice graph coupling map.

Parameters

  • rows (int) – The number of rows to generate the graph with.
  • cols (int) – The number of columns to generate the graph with.
  • bidirectional (bool) – Whether the edges in the output coupling graph are bidirectional or not. By default this is set to True

Returns

A hexagonal lattice coupling graph

Return type

CouplingMap

from_line

classmethod from_line(num_qubits, bidirectional=True)

Return a coupling map of n qubits connected in a line.

Return type

CouplingMap

from_ring

classmethod from_ring(num_qubits, bidirectional=True)

Return a coupling map of n qubits connected to each of their neighbors in a ring.

Return type

CouplingMap

get_edges

get_edges()

Gets the list of edges in the coupling graph.

Returns

Each edge is a pair of physical qubits.

Return type

Tuple(int,int)

is_connected

is_connected()

Test if the graph is connected.

Return True if connected, False otherwise

largest_connected_component

largest_connected_component()

Return a set of qubits in the largest connected component.

make_symmetric

make_symmetric()

Convert uni-directional edges into bi-directional.

neighbors

neighbors(physical_qubit)

Return the nearest neighbors of a physical qubit.

Directionality matters, i.e. a neighbor must be reachable by going one hop in the direction of an edge.

reduce

reduce(mapping, check_if_connected=True)

Returns a reduced coupling map that corresponds to the subgraph of qubits selected in the mapping.

Parameters

  • mapping (list) – A mapping of reduced qubits to device qubits.
  • check_if_connected (bool) – if True, checks that the reduced coupling map is connected.

Returns

A reduced coupling_map for the selected qubits.

Return type

CouplingMap

Raises

CouplingError – Reduced coupling map must be connected.

shortest_undirected_path

shortest_undirected_path(physical_qubit1, physical_qubit2)

Returns the shortest undirected path between physical_qubit1 and physical_qubit2.

Parameters

  • physical_qubit1 (int) – A physical qubit
  • physical_qubit2 (int) – Another physical qubit

Returns

The shortest undirected path

Return type

List

Raises

CouplingError – When there is no path between physical_qubit1, physical_qubit2.

size

size()

Return the number of physical qubits in this graph.

Was this page helpful?
Report a bug or request content on GitHub.