openfermioncirq.VariationalStudy

class openfermioncirq.VariationalStudy(name: str, ansatz: openfermioncirq.variational.ansatz.VariationalAnsatz, objective: openfermioncirq.variational.objective.VariationalObjective, preparation_circuit: Optional[cirq.circuits.circuit.Circuit] = None, initial_state: Union[int, numpy.ndarray] = 0, target: Optional[float] = None, black_box_type: Type[openfermioncirq.variational.variational_black_box.VariationalBlackBox] = <class 'openfermioncirq.variational.variational_black_box.UnitarySimulateVariationalBlackBox'>, datadir: Optional[str] = None)

The results from optimizing a variational ansatz.

A VariationalStudy is used to facilitate optimizing the parameters of a variational ansatz. It contains methods for performing optimizations and saving and loading the results.

Example::

ansatz = SomeVariationalAnsatz() objective = SomeVariationalObjective() study = SomeVariationalStudy(‘my_study’, ansatz, objective) optimization_params = OptimizationParams(

algorithm=openfermioncirq.optimization.COBYLA, initial_guess=numpy.zeros(5))

result = study.optimize(optimization_params, identifier=’run0’) print(result.optimal_value) # prints a number print(result.params.initial_guess) # prints the initial guess used study.save() # saves the study with all results obtained so far

name

The name of the study.

circuit

The circuit of the study, which is the preparation circuit, if any, followed by the ansatz circuit.

ansatz

The ansatz being studied.

objective

The objective function of interest.

target

An optional target value one wants to achieve during optimization.

trial_results

A dictionary of OptimizationTrialResults from optimization runs of the study. Key is the identifier used to label the run.

num_params

The number of parameters in the circuit.

__init__(name: str, ansatz: openfermioncirq.variational.ansatz.VariationalAnsatz, objective: openfermioncirq.variational.objective.VariationalObjective, preparation_circuit: Optional[cirq.circuits.circuit.Circuit] = None, initial_state: Union[int, numpy.ndarray] = 0, target: Optional[float] = None, black_box_type: Type[openfermioncirq.variational.variational_black_box.VariationalBlackBox] = <class 'openfermioncirq.variational.variational_black_box.UnitarySimulateVariationalBlackBox'>, datadir: Optional[str] = None) → None

Parameters:

• name – The name of the study.
• ansatz – The ansatz to study.
• objective – The objective function.
• preparation_circuit – A circuit to apply prior to the ansatz circuit. It should use the qubits belonging to the ansatz.
• initial_state – An initial state to use if the study circuit is run on a simulator.
• target – The target value one wants to achieve during optimization.
• black_box_type – The type of VariationalBlackBox to use for optimization.
• datadir – The directory to use when saving the study. The default behavior is to use the current working directory.

Methods

__init__(name, ansatz, objective, …)	param name:The name of the study.
extend_result(identifier, …)	Extend a result by repeating the run with the same parameters.
load(name, datadir)	Load a study from disk.
optimize(optimization_params, identifier, …)	Perform an optimization run and save the results.
optimize_sweep(param_sweep, identifiers, …)	Perform multiple optimization runs and save the results.
save()	Save the study to disk.
value_of(params)	Determine the value of some parameters.

Attributes

ansatz	The ansatz associated with the study.
circuit	The preparation circuit followed by the ansatz circuit.
num_params	The number of parameters of the ansatz.
objective	The objective associated with the study.