qcmet.benchmarks.UpperBoundOnVD

class qcmet.benchmarks.UpperBoundOnVD(target_circuit, mu=0.1, eta=0.95, qubits=None, seed=None, save_path=None)

Implementation of the upper bound on the variation distance (VD) metric.

This class generates a number of trap circuits with similar structure to that of a given target circuit, according to the quantum accreditation protocol (AP). The trap circuits should output zero states in the noiseless case, thereby the number of non-zero outputs is used to compute the upper bound on the VD.

__init__(target_circuit, mu=0.1, eta=0.95, qubits=None, seed=None, save_path=None)

Initialize the upper bound on the VD benchmark.

Parameters:

• target_circuit (QuantumCircuit) – A target circuit to estimate the upper bound on VD for. The target circuit must follow the restriction such that it has alternating cycles of one-qubit and two-qubit gates, and the two-qubit gates must be CZ gates.

• mu (float, optional) – The desired accuracy of the benchmark ∈ (0, 1). Defaults to 0.1.

• eta (float, optional) – The desired confidence of the benchmark ∈ (0, 1). Defaults to 0.95.

• qubits (int | List[int]) – The number of qubits as either a list of qubit indices or int specifying number of qubits. Defaults to number of qubits in target circuit.

• seed (int, optional) – Random seed to use for randomisations. Defaults to None.

• save_path (str | Path | FileManager | None, optional) – Directory path to save results. Defaults to None.

Methods

__init__(target_circuit[, mu, eta, qubits, ...])	Initialize the upper bound on the VD benchmark.
analyze()	Analyze measurements to return benchmark results.
generate_circuits()	Generate benchmark circuits, user facing.
generate_example_target_circuit(num_qubits, ...)	Randomly generate an example target circuit that satisfies the restriction.
has_plotting()	Check if _plot function is implemented in benchmark.
load_circuit_measurements(circuit_measurements)	Load measurement counts into the experiment_data DataFrame.
measurements_to_probabilities()	Convert raw measurement counts to normalized probabilities.
parse_target_circuit(target_circuit)	Parse the target circuit's structure.
plot([axes])	Plot benchmark result, user facing.
run([device, num_shots, max_circs_per_job])	Run benchmark.
save()	Save benchmark current state.
set_save_path(save_path)	Set benchmark save path if not set in class constructor.

Attributes

circuits	Gets all benchmark circuits.
experiment_data	Getter for experiment_data dataframe.
num_qubits	Number of qubits in this benchmark.

parse_target_circuit(target_circuit)

Parse the target circuit’s structure.

This method breaks down the circuit structure in terms of alternating cycles of one- and two-qubit gates. Each cycle starts with one-qubit gate(s) and ends with (possibly parallel) two-qubit gate(s). Gates up to the next non-parallel two-qubit gate becomes the next cycle. A dictionary is generated for each cycle, which tells for each qubit in this cycle whether one-qubit gate(s) or a two-qubit gate is existent on this qubit. This information is later used for constructing trap circuits.

Parameters:

target_circuit (QuantumCircuit) – The target circuit to parse.

Returns:

The list of circuit structure information, one element for each cycle. Each element is a dict: { “1q”: [bool], “2q”: [[int, int, CircuitInstruction]] }. The 1q list in dict has length equal to the number of qubits, which tells whether a one-qubit gate exists for each qubit. The 2q list has length equal to the number of parallel two-qubit gates in this cycle, and contains the qubit indices and the 2q gate itself.

Return type:

list

static generate_example_target_circuit(num_qubits, cycles, seed=None)

Randomly generate an example target circuit that satisfies the restriction.

Parameters:

• num_qubits (int) – The number of qubits.

• cycles (int) – The number of cycles.

• seed (int, optional) – Random seed. Defaults to None.

Returns:

An example target circuit.

Return type:

QuantumCircuit