# This code is part of Qiskit.
#
# (C) Copyright IBM 2017, 2018.
#
# This code is licensed under the Apache License, Version 2.0. You may
# obtain a copy of this license in the LICENSE.txt file in the root directory
# of this source tree or at http://www.apache.org/licenses/LICENSE-2.0.
#
# Any modifications or derivative works of this code must retain this
# copyright notice, and modified files need to carry a notice indicating
# that they have been altered from the originals.
"""Choose a Layout by finding the most connected subset of qubits."""
import numpy as np
import rustworkx
from qiskit.transpiler.layout import Layout
from qiskit.transpiler.basepasses import AnalysisPass
from qiskit.transpiler.exceptions import TranspilerError
from qiskit.transpiler.passes.layout import disjoint_utils
from qiskit._accelerate.dense_layout import best_subset
[docs]class DenseLayout(AnalysisPass):
"""Choose a Layout by finding the most connected subset of qubits.
This pass associates a physical qubit (int) to each virtual qubit
of the circuit (Qubit).
Note:
Even though a ``'layout'`` is not strictly a property of the DAG,
in the transpiler architecture it is best passed around between passes
by being set in ``property_set``.
"""
def __init__(self, coupling_map=None, backend_prop=None, target=None):
"""DenseLayout initializer.
Args:
coupling_map (Coupling): directed graph representing a coupling map.
backend_prop (BackendProperties): backend properties object
target (Target): A target representing the target backend.
"""
super().__init__()
self.coupling_map = coupling_map
self.backend_prop = backend_prop
self.target = target
self.adjacency_matrix = None
if target is not None:
self.coupling_map = target.build_coupling_map()
[docs] def run(self, dag):
"""Run the DenseLayout pass on `dag`.
Pick a convenient layout depending on the best matching
qubit connectivity, and set the property `layout`.
Args:
dag (DAGCircuit): DAG to find layout for.
Raises:
TranspilerError: if dag wider than self.coupling_map
"""
if self.coupling_map is None:
raise TranspilerError(
"A coupling_map or target with constrained qargs is necessary to run the pass."
)
layout_components = disjoint_utils.run_pass_over_connected_components(
dag,
self.coupling_map if self.target is None else self.target,
self._inner_run,
)
layout_mapping = {}
for component in layout_components:
layout_mapping.update(component)
layout = Layout(layout_mapping)
for qreg in dag.qregs.values():
layout.add_register(qreg)
self.property_set["layout"] = layout
def _inner_run(self, dag, coupling_map):
num_dag_qubits = len(dag.qubits)
if num_dag_qubits > coupling_map.size():
raise TranspilerError("Number of qubits greater than device.")
num_cx = 0
num_meas = 0
if self.target is not None:
num_cx = 1
num_meas = 1
else:
# Get avg number of cx and meas per qubit
ops = dag.count_ops(recurse=True)
if "cx" in ops.keys():
num_cx = ops["cx"]
if "measure" in ops.keys():
num_meas = ops["measure"]
best_sub = self._best_subset(num_dag_qubits, num_meas, num_cx, coupling_map)
layout_mapping = {
qubit: coupling_map.graph[int(best_sub[i])] for i, qubit in enumerate(dag.qubits)
}
return layout_mapping
def _best_subset(self, num_qubits, num_meas, num_cx, coupling_map):
"""Computes the qubit mapping with the best connectivity.
Args:
num_qubits (int): Number of subset qubits to consider.
Returns:
ndarray: Array of qubits to use for best connectivity mapping.
"""
from scipy.sparse import coo_matrix, csgraph
if num_qubits == 1:
return np.array([0])
if num_qubits == 0:
return []
adjacency_matrix = rustworkx.adjacency_matrix(coupling_map.graph)
reverse_index_map = {v: k for k, v in enumerate(coupling_map.graph.nodes())}
error_mat, use_error = _build_error_matrix(
coupling_map.size(),
reverse_index_map,
backend_prop=self.backend_prop,
coupling_map=self.coupling_map,
target=self.target,
)
rows, cols, best_map = best_subset(
num_qubits,
adjacency_matrix,
num_meas,
num_cx,
use_error,
coupling_map.is_symmetric,
error_mat,
)
data = [1] * len(rows)
sp_sub_graph = coo_matrix((data, (rows, cols)), shape=(num_qubits, num_qubits)).tocsr()
perm = csgraph.reverse_cuthill_mckee(sp_sub_graph)
best_map = best_map[perm]
return best_map
def _build_error_matrix(num_qubits, qubit_map, target=None, coupling_map=None, backend_prop=None):
error_mat = np.zeros((num_qubits, num_qubits))
use_error = False
if target is not None and target.qargs is not None:
for qargs in target.qargs:
# Ignore gates over 2q DenseLayout only works with 2q
if len(qargs) > 2:
continue
error = 0.0
ops = target.operation_names_for_qargs(qargs)
for op in ops:
props = target[op].get(qargs, None)
if props is not None and props.error is not None:
# Use max error rate to represent operation error
# on a qubit(s). If there is more than 1 operation available
# we don't know what will be used on the qubits eventually
# so we take the highest error operation as a proxy for
# the possible worst case.
error = max(error, props.error)
max_error = error
if any(qubit not in qubit_map for qubit in qargs):
continue
# TODO: Factor in T1 and T2 to error matrix after #7736
if len(qargs) == 1:
qubit = qubit_map[qargs[0]]
error_mat[qubit][qubit] = max_error
use_error = True
elif len(qargs) == 2:
error_mat[qubit_map[qargs[0]]][qubit_map[qargs[1]]] = max_error
use_error = True
elif backend_prop and coupling_map:
error_dict = {
tuple(gate.qubits): gate.parameters[0].value
for gate in backend_prop.gates
if len(gate.qubits) == 2
}
for edge in coupling_map.get_edges():
gate_error = error_dict.get(edge)
if gate_error is not None:
if edge[0] not in qubit_map or edge[1] not in qubit_map:
continue
error_mat[qubit_map[edge[0]]][qubit_map[edge[1]]] = gate_error
use_error = True
for index, qubit_data in enumerate(backend_prop.qubits):
if index not in qubit_map:
continue
for item in qubit_data:
if item.name == "readout_error":
mapped_index = qubit_map[index]
error_mat[mapped_index][mapped_index] = item.value
use_error = True
return error_mat, use_error