pykoop.QmcCenters

class QmcCenters(n_centers=100, symmetric_range=False, qmc=None, qmc_kw=None, random_state=None)

Bases: Centers

Centers generated with Quasi-Monte Carlo sampling.

Parameters:

n_centers (int) –
symmetric_range (bool) –
qmc (Callable[[...], QMCEngine] | None) –
qmc_kw (Dict[str, Any] | None) –
random_state (int | RandomState | None) –

centers_

Centers, shape (n_centers, n_features).

Type:: np.ndarray

n_centers_

Number of centers generated.

Type:: int

n_features_in_

Number of features input.

Type:: int

range_max_

Maximum value of each feature used to generate grid.

Type:: np.ndarray

range_min_

Minimum value of each feature used to generate grid.

Type:: np.ndarray

qmc_

Quasi-Monte Carlo sampler instantiated from qmc.

Type:: stats.qmc.QMCEngine

Examples

Generate centers using Latin hypercube sampling (default)

>>> qmc = pykoop.QmcCenters(n_centers=10)
>>> qmc.fit(X_msd[:, 1:])  # Remove episode feature
QmcCenters(n_centers=10)
>>> qmc.centers_
array([...])

Generate centers using a Sobol sequence

>>> qmc = pykoop.QmcCenters(n_centers=8, qmc=scipy.stats.qmc.Sobol)
>>> qmc.fit(X_msd[:, 1:])  # Remove episode feature
QmcCenters(n_centers=8, qmc=<class 'scipy.stats._qmc.Sobol'>)
>>> qmc.centers_
array([...])

__init__(n_centers=100, symmetric_range=False, qmc=None, qmc_kw=None, random_state=None)

Instantiate QmcCenters.

Parameters:

n_centers (int) – Number of centers to generate.
symmetric_range (bool) – If true, the grid range for a given feature is forced to be symmetric about zero (i.e., [-max(abs(x)), max(abs(x))]). Otherwise, the grid range is taken directly on the data (i.e., [min(x), max(x)]). Default is false.
qmc (Optional[Callable[..., stats.qmc.QMCEngine]]) –
Quasi-Monte Carlo method from scipy.stats.qmc to use. Argument is the desired subclass of scipy.stats.qmc.QMCEngine to use. Accepts the class itself, not an instance of the class. Possible values are
- scipy.stats.qmc.Sobol – Sobol sequence,
- scipy.stats.qmc.Halton – Halton sequence,
- scipy.stats.qmc.LatinHypercube – Latin hypercube sampling (LHS),
- scipy.stats.qmc.PoissonDisk – Poisson disk sampling (requires scipy v1.9.0.),
- scipy.stats.qmc.MultinomialQMC – Multinomial distribution, and
- scipy.stats.qmc.MultivariateNormalQMC – Multivariate normal distribution.
If None, defaults to Latin hypercube sampling.
qmc_kw (Optional[Dict[str, Any]]) – Additional keyword arguments passed when instantiating qmc. If seed is specified here, it takes precedence over random_state.
random_state (Union[int, np.random.RandomState, None]) – Random seed.

Return type:

None

Methods

`__init__`([n_centers, symmetric_range, qmc, ...])	Instantiate `QmcCenters`.
`fit`(X[, y])	Generate centers from data.
`get_metadata_routing`()	Get metadata routing of this object.
`get_params`([deep])	Get parameters for this estimator.
`set_params`(**params)	Set the parameters of this estimator.

fit(X, y=None)

Generate centers from data.

Parameters:

X (np.ndarray) – Data matrix.
y (Optional[np.ndarray]) – Ignored.

Returns:

Instance of itself.

Return type:

Centers

Raises:

ValueError – If any of the constructor parameters are incorrect.

get_metadata_routing()

Get metadata routing of this object.

Please check User Guide on how the routing mechanism works.

Returns:: routing – A MetadataRequest encapsulating routing information.
Return type:: MetadataRequest

get_params(deep=True)

Get parameters for this estimator.

Parameters:: deep (bool, default=True) – If True, will return the parameters for this estimator and contained subobjects that are estimators.
Returns:: params – Parameter names mapped to their values.
Return type:: dict

set_params(**params)

Set the parameters of this estimator.

The method works on simple estimators as well as on nested objects (such as Pipeline). The latter have parameters of the form <component>__<parameter> so that it’s possible to update each component of a nested object.

Parameters:: **params (dict) – Estimator parameters.
Returns:: self – Estimator instance.
Return type:: estimator instance