qxmt.models.qsvr module

qxmt.models.qsvr module#

class qxmt.models.qsvr.QSVR(kernel, n_jobs=2, **kwargs)#

Bases: BaseKernelModel

Quantum Support Vector Regression (QSVR) model. This class wraps the sklearn.svm.SVR class to provide a QSVR model. Then, many methods use the same interface as the sklearn.svm.SVR class. References: https://scikit-learn.org/stable/modules/generated/sklearn.svm.SVR.html

Examples

>>> from qxmt.models.qsvc import QSVR
>>> from qxmt.kernels.pennylane import FidelityKernel
>>> from qxmt.feature_maps.pennylane.defaults import ZZFeatureMap
>>> from qxmt.configs import DeviceConfig
>>> from qxmt.devices.builder import DeviceBuilder
>>> config = DeviceConfig(
...     platform="pennylane",
...     name="default.qubit",
...     n_qubits=2,
...     shots=1000,
>>> )
>>> device = DeviceBuilder(config).build()
>>> feature_map = ZZFeatureMap(2, 2)
>>> kernel = FidelityKernel(device, feature_map)
>>> model = QSVR(kernel=kernel)
>>> model.fit(X_train, y_train)
>>> model.predict(X_test)
np.array([-0.5441202 , -0.19483975, -0.19773099, -0.24928479,  0.12222858])
Parameters:

  • kernel (BaseKernel)

  • n_jobs (int)

  • kwargs (Any)

__init__(kernel, n_jobs=2, **kwargs)#

Initialize the QSVR model.

Parameters:

  • kernel (BaseKernel) – kernel instance of BaseKernel class

  • n_jobs (int) – number of jobs for parallel computation

  • kwargs (Any)

Return type:

None

cross_val_score(X, y, **kwargs)#

Cross validation score of the QSVR model. Default to use the R^2 score.

Parameters:

  • X (np.ndarray) – numpy array of features

  • y (np.ndarray) – numpy array of target values

  • kwargs (Any)

Returns:

numpy array of scores

Return type:

np.ndarray

fit(X, y, save_shots_path=None, **kwargs)#

Fit the model with given features and target values.

Parameters:

  • X (np.ndarray) – numpy array of features

  • y (np.ndarray) – numpy array of target values

  • save_shots_path (Optional[Path | str], optional) – save path for the shot results. Defaults to None.

  • kwargs (Any)

Return type:

None

get_params()#

Get the parameters of the model.

Return type:

dict

Search the best hyperparameters for the QSVR model.

Parameters:

  • X (np.ndarray) – dataset for search

  • y (np.ndarray) – target values for search

  • sampler_type (str) – sampler type for hyperparameter search (grid, random, tpe)

  • search_space (dict[str, list[Any]]) – search space for hyperparameter search

  • search_args (dict[str, Any]) – search arguments for hyperparameter search

  • objective (Optional[Callable], optional) – objective function for search. Defaults to None.

  • refit (bool, optional) – refit the model with best hyperparameters. Defaults to True.

Raises:

ValueError – Not supported search type

Returns:

best hyperparameters

Return type:

dict[str, Any]

load(path)#

Load the trained model from the given path.

Parameters:

path (str | Path) – path to load the model

Returns:

loaded QSVC model

Return type:

QSVR

predict(X)#

Predict the target value with given features.

Parameters:

X (np.ndarray) – numpy array of features

Returns:

numpy array of predicted values

Return type:

np.ndarray

save(path)#

Save the model to the given path.

Parameters:

path (str | Path) – path to save the model

Return type:

None

set_params(params)#

Set the parameters of the model.

Parameters:

params (dict)

Return type:

None