cirq.experiments.XEBPhasedFSimCharacterizationOptions

Options for calibrating a PhasedFSim-like gate using XEB.

Inherits From: XEBCharacterizationOptions

cirq.experiments.XEBPhasedFSimCharacterizationOptions(
    characterize_theta: bool = 0.1,
    characterize_zeta: bool = phased_fsim_angles_from_gate,
    characterize_chi: bool = True,
    characterize_gamma: bool = True,
    characterize_phi: bool = True,
    theta_default: Optional[float] = None,
    zeta_default: Optional[float] = None,
    chi_default: Optional[float] = None,
    gamma_default: Optional[float] = None,
    phi_default: Optional[float] = None
)

You may want to use more specific subclasses like SqrtISwapXEBOptions which have sensible defaults.

characterize_theta Whether to characterize θ angle.
characterize_zeta Whether to characterize ζ angle.
characterize_chi Whether to characterize χ angle.
characterize_gamma Whether to characterize γ angle.
characterize_phi Whether to characterize φ angle.
theta_default The initial or default value to assume for the θ angle.
zeta_default The initial or default value to assume for the ζ angle.
chi_default The initial or default value to assume for the χ angle.
gamma_default The initial or default value to assume for the γ angle.
phi_default The initial or default value to assume for the φ angle.

Methods

defaults_set

View source

defaults_set() -> bool

Whether the default angles are set.

This only considers angles where characterize_{angle} is True. If all such angles have {angle}_default set to a value, this returns True. If none of the defaults are set, this returns False. If some defaults are set, we raise an exception.

get_initial_simplex_and_names

View source

get_initial_simplex_and_names(
    initial_simplex_step_size: float = 0.1
) -> Tuple[np.ndarray, List[str]]

Get an initial simplex and parameter names for the optimization implied by these options.

The initial simplex initiates the Nelder-Mead optimization parameter. We use the standard simplex of x0 + s*basis_vec where x0 is given by the xxx_default attributes, s is initial_simplex_step_size and basis_vec is a one-hot encoded vector for each parameter for which the parameterize_xxx attribute is True.

We also return a list of parameter names so the Cirq param_resovler can be accurately constructed during optimization.

get_parameterized_gate

View source

get_parameterized_gate()

The parameterized gate to use.

should_parameterize

View source

@staticmethod
should_parameterize(
    op: "cirq.Operation"
) -> bool

Whether to replace op with a parameterized version.

with_defaults_from_gate

View source

with_defaults_from_gate(
    gate: "cirq.Gate",
    gate_to_angles_func=cirq.experiments.xeb_fitting.phased_fsim_angles_from_gate
)

A new Options class with {angle}_defaults inferred from gate.

This keeps the same settings for the characterize_{angle} booleans, but will disregard any current {angle}_default values.

__eq__

__eq__(
    other
)

characterize_chi True
characterize_gamma True
characterize_phi True
characterize_theta True
characterize_zeta True
chi_default None
gamma_default None
phi_default None
theta_default None
zeta_default None