Skip to content

laboneq_applications.analysis.two_qubit_rb

Analysis for the two-qubit randomized benchmarking experiment.

The experiment is defined in laboneq_applications.experiments.

Data is acquired in SINGLE_SHOT + DISCRIMINATION mode, so each shot yields a binary 0/1 result per qubit. The per-shot bit outcomes of both qubits in a pair are combined into a joint two-qubit state ("00", "01", "10", "11"), and the probability of each joint state is computed by averaging over shots for each circuit in the sweep.

The "00" state probability is fitted with an exponential decay and the error-per-Clifford (EPC) is extracted using the two-qubit depolarising formula::

EPC = (d - 1) / d * (1 - exp(-decay_rate))

where d = 2^2 = 4 for two qubits.

The analysis returns a dictionary of the form::

{"q0_uid-q1_uid": {"epc_avg": float}}

PlotPopulationTQRBOptions

Options for the plot_population task of the two-qubit RB analysis.

Attributes:

Name Type Description
do_fitting bool

Whether to overlay the exponential fit on the P(00) data. Default: True.

analysis_workflow(result, qubit_pairs, length_cliffords, variations, options=None)

Two-qubit RB analysis workflow.

The workflow consists of the following steps:

Changed in version 27.1.0

The analysis moved out of contrib. It takes the qubit pairs as UIDs instead of quantum elements, averages single-shot data over the shots of each sequence, and plots the Clifford fidelity.

Parameters:

Name Type Description Default
result RunExperimentResults

The experiment results returned by the run_experiment task.

required
qubit_pairs list[list[str]]

List of [q0, q1] pairs that were benchmarked, passed by UID.

required
length_cliffords list[int]

List of numbers of 2-qubit Clifford gates swept.

required
variations int

Number of random circuit samples per sequence length.

required
options TuneUpAnalysisWorkflowOptions | None

Analysis workflow options.

None

Returns:

Name Type Description
WorkflowBuilder None

The builder for the analysis workflow.

calculate_qubit_population_tq_rb(qubit_pairs, result, length_cliffords, variations)

Compute joint two-qubit state populations from single-shot data.

For each shot, the per-qubit discrimination bits (0 or 1) are combined into a joint two-qubit state label ("00", "01", "10", "11"). The probability of each joint state is then obtained by averaging over shots for every circuit in the sweep.

Parameters:

Name Type Description Default
qubit_pairs list[list[str]]

List of [q0, q1] pairs, passed by UID.

required
result RunExperimentResults

The experiment results: discriminated states 0 or 1, with one row per shot and one column per sequence.

required
length_cliffords list

List of Clifford sequence lengths.

required
variations int

Number of random circuit samples per sequence length.

required

Returns:

Type Description
dict[str, dict[str, ArrayLike]]

Dictionary keyed by pair UID "q0_uid-q1_uid":

{
    pair_key: {
        "sweep_points": array,  # Clifford count per circuit
        "00": array,  # P(q0=0, q1=0) per circuit
        "01": array,  # P(q0=0, q1=1) per circuit
        "10": array,  # P(q0=1, q1=0) per circuit
        "11": array,  # P(q0=1, q1=1) per circuit
    }
}

compute_pair_epc(qubit_pairs, fit_results)

Compute the EPC for each qubit pair from the P(00) fit.

The error per two-qubit Clifford is extracted from the exponential decay of the joint "00" state population::

EPC = (d - 1) / d * (1 - exp(-decay_rate))

with d = 4 (two-qubit Hilbert space dimension).

Parameters:

Name Type Description Default
qubit_pairs list[list[str]]

List of [q0, q1] pairs, passed by UID.

required
fit_results dict[str, ModelResult]

Per-pair fit results as returned by fit_data.

required

Returns:

Type Description
dict[str, dict[str, float]]

Dictionary {"q0-q1": {"epc_avg": float}} keyed by pair UID.

exponential_decay_fit(x, data, param_hints=None)

Fit an exponential decay model to RB data.

Parameters:

Name Type Description Default
x ArrayLike

Sequence lengths (number of Cliffords).

required
data ArrayLike

Joint state population values.

required
param_hints dict | None

lmfit parameter hints dictionary.

None

Returns:

Type Description
ModelResult

The lmfit result.

fit_data(qubit_pairs, processed_data_dict, options=None)

Fit an exponential decay to P(00) for each qubit pair.

Parameters:

Name Type Description Default
qubit_pairs list[list[str]]

List of [q0, q1] pairs, passed by UID.

required
processed_data_dict dict[str, dict[str, ArrayLike]]

Per-pair joint state populations as returned by calculate_qubit_population_tq_rb.

required
options FitDataOptions | None

Fit options.

None

Returns:

Type Description
dict[str, ModelResult]

Dictionary keyed by pair UID with lmfit ModelResult values.

plot_population(qubit_pairs, processed_data_dict, fit_results, options=None)

Plot the joint two-qubit state populations for each pair.

One figure is produced per qubit pair, showing P(00) vs. number of Cliffords together with the exponential fit and extracted EPC.

Parameters:

Name Type Description Default
qubit_pairs list[list[str]]

List of [q0, q1] pairs, passed by UID.

required
processed_data_dict dict[str, dict[str, ArrayLike]]

Per-pair joint state populations.

required
fit_results dict[str, ModelResult] | None

Per-pair fit results (keyed by pair UID).

required
options PlotPopulationTQRBOptions | None

Plot options.

None

Returns:

Type Description
dict[str, Figure]

Dictionary keyed by pair UID ("q0_uid-q1_uid") with matplotlib Figure objects.