CharacterizationOptions¶

class CharacterizationOptions(sensitivity=False, parameter_sweep=False, auto_tune=False, gamma=None, beta=None, cost_qubo=None, penalty_qubo=None, constraints=None, ansatz=None)[source]¶

Bases: object

Configuration for characterize_and_validate().

All fields are optional; default-construct for a basic run with no sub-analyses. The dataclass validates field combinations at construction time (__post_init__), so misconfiguration surfaces before any API call.

Examples

>>> CharacterizationOptions(parameter_sweep=True, sensitivity=True)
>>> CharacterizationOptions(gamma=1.2, beta=0.7)

Attributes Summary

`ansatz`	Ansatz configuration dict (e.g. `{"mixer": "x", "layers": 1}`).
`auto_tune`	Request automatic penalty tuning.
`beta`	Fixed β value.
`constraints`	Constraint descriptors.
`cost_qubo`	Cost-only `BinaryOptimizationProblem` for penalty analysis.
`gamma`	Fixed γ value.
`parameter_sweep`	Request a γ/β parameter sweep.
`penalty_qubo`	Penalty-only `BinaryOptimizationProblem` for penalty analysis.
`sensitivity`	Request per-qubit sensitivity analysis.

Attributes Documentation

ansatz: dict | None = None¶

Ansatz configuration dict (e.g. {"mixer": "x", "layers": 1}).

The auto_warmstart key is reserved for the backend and rejected at construction time if supplied.

auto_tune: bool = False¶: Request automatic penalty tuning.

beta: float | None = None¶: Fixed β value. Mutually exclusive with parameter_sweep.

constraints: list | None = None¶: Constraint descriptors.

cost_qubo: BinaryOptimizationProblem | None = None¶: Cost-only BinaryOptimizationProblem for penalty analysis.

gamma: float | None = None¶: Fixed γ value. Mutually exclusive with parameter_sweep.

parameter_sweep: bool = False¶

Request a γ/β parameter sweep.

Mutually exclusive with fixed gamma / beta.

penalty_qubo: BinaryOptimizationProblem | None = None¶: Penalty-only BinaryOptimizationProblem for penalty analysis.

sensitivity: bool = False¶: Request per-qubit sensitivity analysis.