qiskit_noise_learning.gate_sets.QiskitGateSet
class qiskit_noise_learning.gate_sets.QiskitGateSet(num_qubits: int | None = None, *, target: Target | None = None, qubit_subset: Iterable[int] | None = None, add_default_spam: bool = True, name: str | None = None, latex_str: str | None = None)
Bases: GateSet[QiskitGate]
A gate set whose noise is to be learned and that is specified using Qiskit objects.
Here, we are using “gate” in the context of noise learning, where our operations of interest are not typically the smallest operations that are discretely executed on a device (such as cz gates), but rather collections of such operations whose noise will be learned together as a unit. To the point, typically, in this class, gates are layers. They need not be however; partial width layers and layers with overlapping operations within are valid. We require this object to represent a set of them because in certain cases their noise needs to be learned together in order to have a consistent gauge defined between their noise models, or to allow parametrizations of their noise models to be correlated.
This object satisfies the Python mapping protocol so that, for example, gates can be extracted with dictionary syntax. The names of gates are always strings.
>>> from qiskit_noise_learning.gate_sets import QiskitGateSet
>>> from qiskit.circuit import QuantumCircuit
>>> # instantiate a new gate set on 10 qubits
>>> gate_set = QiskitGateSet(10)
>>> # the gate set comes populated with preparation and measurement gates on all 10 qubits
>>> assert len(gate_set) == 2
>>> assert "P" in gate_set and "M" in gate_setParameters
- num_qubits – How many qubits the QPU of interest has. If a
targetis provided, this field may be omitted. - target – An optional
Targetagainst which operations will be validated whenever gates are added to the gate set. Its number of qubits must matchnum_qubitsif both are present. - qubit_subset – A subset of
range(num_qubits)specifying the region of interest of the QPU. All gates added must act within this subset. By default, contains all qubits. Whenadd_default_spamisTrue, the iteration order of this argument determines the qubit ordering of the default preparation and measurement gates. - add_default_spam – Whether to initialize the gateset with gates that respectively implement state preparation (given name
"P") and state measurement (given name"M") on all qubits in the region of interest. - name – Name for this gate set. If
None,namefalls back to the class name. - latex_str – An optional LaTeX string for rendering this gate set.
__init__
__init__(num_qubits: int | None = None, *, target: Target | None = None, qubit_subset: Iterable[int] | None = None, add_default_spam: bool = True, name: str | None = None, latex_str: str | None = None)
Methods
Column 1 | Column 2 |
|---|---|
__init__([num_qubits, target, qubit_subset, ...]) | |
add_box_as_gate(box_instr, *[, name, latex_str]) | Add a Qiskit circuit instruction containing a box operation as a gate. |
add_circuit_as_gate(circuit[, qubit_idxs, ...]) | Add a quantum circuit object as a gate. |
add_gate(gate) | Add a gate to the gate set. |
add_measurement(qubit_idxs, operation_type, ...) | Add a gate to this gate set that measures specified qubits. |
add_preparation([qubit_idxs, annotations, ...]) | Add a gate to this gate set that prepares (or resets) specified qubits. |
build_new_gate([name, idle_unused, latex_str]) | Return a circuit builder whose contents will be added as a gate. |
draw() | Draw the device topology with per-gate coloring. |
get(k[,d]) | |
items() | |
keys() | |
values() |
Attributes
Column 1 | Column 2 |
|---|---|
label | A string label for use in plotter legends. |
latex_str | A LaTeX string for this gate set. |
math_label | A string label for use within latex math mode. |
model_gate_set | The model for this gate set. |
name | Name for this gate set, defaulting to the class name. |
num_qubits | The total number of qubits of the device this gateset acts on. |
qubit_subset | The indices of the subset of device qubits that all gates act on. |
target | The target of this gateset, if one exists. |
model_gate_set
Type: ModelGateSet
The model for this gate set.
target
add_gate
add_gate(gate: QiskitGate)
Add a gate to the gate set.
>>> from qiskit_noise_learning.gate_sets import QiskitGateSet, QiskitGate
>>> from qiskit.circuit import QuantumCircuit
>>> gate_set = QiskitGateSet(10)
>>> circuit = QuantumCircuit(5)
>>> circuit.cx(3, 4)
>>> gate_set.add_gate(QiskitGate("gate0", circuit, [4, 5, 7, 8, 9]))
>>> assert "gate0" in gate_setParameters
gate – The gate to add.
Raises
ValueError – If the gate acts on some qubits outside of the valid range, or this gate set has a target and some member of the gate does not comply, or if the name is already used in the gate set.
add_box_as_gate
add_box_as_gate(box_instr: CircuitInstruction, *, name: str | None = None, latex_str: str | None = None) → str
Add a Qiskit circuit instruction containing a box operation as a gate.
>>> from qiskit_noise_learning.gate_sets import QiskitGateSet, QiskitGate
>>> from qiskit.circuit import QuantumCircuit
>>> gate_set = QiskitGateSet(10)
>>> circuit = QuantumCircuit(10)
>>> with circuit.box():
... circuit.cx(3, 4)
... # use noop to indicate that only the first 7 qubits will be part of the gate,
... # otherwise it would be restricted to only qubits 3 and 4, and learned noise
... # will only be with respect to those two qubits
... circuit.noop(range(7))
>>> name = gate_set.add_box_as_gate(circuit[0])
>>> assert name in gate_setParameters
- box_instr – The circuit instruction containing a box operation.
- name – The name of the gate, or
Noneto have a name chosen for you. - latex_str – An optional LaTeX string for this gate.
Returns
The name of the added gate.
Raises
ValueError – If the provided instruction does not contain a qiskit.circuit.BoxOp operation, or if the instruction acts on non-physical qubits.
add_circuit_as_gate
add_circuit_as_gate(circuit: QuantumCircuit, qubit_idxs: Iterable[int] | None = None, *, annotations: Iterable[Annotation] | None = None, name: str | None = None, latex_str: str | None = None) → str
Add a quantum circuit object as a gate.
This method is a thin wrapper for add_gate() that constructs a QiskitGate for you.
>>> from qiskit_noise_learning.gate_sets import QiskitGateSet, QiskitGate
>>> from qiskit.circuit import QuantumCircuit
>>> gate_set = QiskitGateSet(10)
>>> circuit = QuantumCircuit(5)
>>> circuit.cx(3, 4)
>>> name = gate_set.add_circuit_as_gate(circuit, [4, 5, 7, 8, 9])
>>> assert name in gate_setThe circuit.qubits are completely irrelevant and do not, for example, represent physical qubits. Instead, the mapping dict(zip(circuit.qubits, qubit_idxs)) provides the recipe for which physical qubits each qubit in the circuit corresponds to.
Parameters
- circuit – The circuit to use as a gate.
- qubit_idxs – The physical qubits on which the circuit acts.
- annotations – The annotations that describe how to implement the circuit, or
Noneto use the default annotations ofQiskitGate. - name – The name of the gate, or
Noneto have a name chosen for you. - latex_str – An optional LaTeX string for this gate.
Returns
The name of the added gate.
build_new_gate
build_new_gate(name: str | None = None, idle_unused: bool = True, latex_str: str | None = None) → GateBuilder
Return a circuit builder whose contents will be added as a gate.
>>> from qiskit_noise_learning.gate_sets import QiskitGateSet, QiskitGate
>>> from qiskit.circuit import QuantumCircuit
>>> gate_set = QiskitGateSet(10)
>>> with gate_set.build_new_gate() as builder:
... builder.circuit.cx(4, 5)
>>> assert builder.name in gate_set
>>> assert set(gate_set[builder.name].qubit_idxs) == set(range(10))Parameters
- name – The name of the gate, or
Noneto have a name chosen for you. - idle_unused – Whether all qubits in
qubit_subsetthat are not already part of the gate will be automatically included as idling qubits. - latex_str – An optional LaTeX string for this gate.
Returns
A circuit builder.
add_measurement
add_measurement(qubit_idxs: Iterable[int] | None = None, operation_type: type[Operation] = <class 'qiskit.circuit.measure.Measure'>, *, annotations: Iterable[Annotation] | None = None, name: str | None = None, latex_str: str | None = None) → str
Add a gate to this gate set that measures specified qubits.
Parameters
- qubit_idxs – The physical qubit indices to measure.
- operation_type – The type of measurement operation to apply on each qubit.
- annotations – The annotations that describe how to implement the measurement, or
Noneto use the default annotations ofQiskitGate. - name – The name of the gate, or
Noneto have a name chosen for you. - latex_str – An optional LaTeX string for this gate.
Returns
The name of the added gate.
add_preparation
add_preparation(qubit_idxs: Iterable[int] | None = None, *, annotations: Iterable[Annotation] | None = None, name: str | None = None, latex_str: str | None = None) → str
Add a gate to this gate set that prepares (or resets) specified qubits.
Parameters
- qubit_idxs – The physical qubit indices to prepare.
- annotations – The annotations that describe how to implement the preparation, or
Noneto use the default annotations ofQiskitGate. - name – The name of the gate, or
Noneto have a name chosen for you. - latex_str – An optional LaTeX string for this gate.
Returns
The name of the added gate.