geope.grape

geope.grape.Grape

Gradient/Hessian-based GRAPE optimiser for quantum gate synthesis.

Mirrors the Geope usage pattern: it is constructed from a Parameters object (the single source of truth for all configuration and the live optimisation state), while the optimiser method, its hyperparameters and max_steps are arguments of :meth:optimize.

Attributes:

Name	Type	Description
params		The bound Parameters object (also the source of the lazily-built/cached optimisation functions).
precision		Target fidelity threshold.
method		Optimiser method from the most recent :meth:optimize call ('gd', 'adam', 'nr-trm' or 'nr-rfo'); None until :meth:optimize is first called.
step_size		Transient last step size (always 0 for GRAPE).
history		Optional History logger (None unless supplied).

__dict__ class-attribute

__doc__ class-attribute

str(object='') -> str str(bytes_or_buffer[, encoding[, errors]]) -> str

Create a new string object from the given object. If encoding or errors is specified, then the object must expose a data buffer that will be decoded using the given encoding and error handler. Otherwise, returns the result of object.str() (if defined) or repr(object). encoding defaults to sys.getdefaultencoding(). errors defaults to 'strict'.

__module__ class-attribute

str(object='') -> str str(bytes_or_buffer[, encoding[, errors]]) -> str

Create a new string object from the given object. If encoding or errors is specified, then the object must expose a data buffer that will be decoded using the given encoding and error handler. Otherwise, returns the result of object.str() (if defined) or repr(object). encoding defaults to sys.getdefaultencoding(). errors defaults to 'strict'.

__weakref__ property

list of weak references to the object

__init__(params, precision=0.9999999, verbose=False, history=None)

Initialise the Grape optimiser.

Grape requires a Parameters object — the engine, initial parameters, drift, seed, initialisation spread, projective flag and param_transform are all read from it.

Parameters:

Name	Type	Description	Default
params	Parameters	A Parameters instance bundling every input the optimiser needs.	required
precision	float	Target fidelity. Defaults to 0.9999999.	0.9999999
verbose	bool	Whether to print progress. Defaults to False.	False
history	History | None	Optional History logger. When supplied, the full run trajectory is recorded into it; when None (default), no history is kept.	None

Raises:

Type	Description
TypeError	If params is not a Parameters instance.

init(init_parameters=None, drift_parameters=None, seed=None)

(Re-)initialise optimiser state.

Sets up initial parameters, drift parameters and the live state (params.parameters / params.fidelity), and records step 0 into history when one is attached.

Parameters:

Name	Type	Description	Default
init_parameters	ndarray | None	Initial parameter array. Defaults to random.	None
drift_parameters	ndarray | None	Fixed drift parameter values. Defaults to ones.	None
seed	int | Array | None	Random seed (int) or JAX key for reproducibility.	None

optimize(max_steps=100, method='nr-trm', **optimizer_kwargs)

Run the GRAPE optimisation loop.

Iterates gradient/Hessian update steps until the fidelity exceeds self.precision or max_steps is reached.

Parameters:

Name	Type	Description	Default
max_steps	int	Maximum number of optimisation steps. Defaults to 100.	100
method	str	'gd', 'adam', 'nr-trm' (default) or 'nr-rfo'.	'nr-trm'
**optimizer_kwargs	float	Method hyperparameters — learning_rate for 'gd'/'adam', delta for 'nr-trm', kappa for 'nr-rfo'.	required

Returns:

Type	Description
Parameters	The bound Parameters instance, carrying the final
Parameters	parameters (current array) and fidelity (scalar). The full
Parameters	trajectory and best_* live on grape.history when a
Parameters	History was supplied.