from abc import ABCMeta, abstractmethod from collections.abc import Mapping from typing import Any, ClassVar, Literal from typing_extensions import Self from numpy import ndarray from numpy.random import RandomState from scipy.sparse._csr import csr_matrix from .._typing import ArrayLike, Float, Int, MatrixLike from ..base import BaseEstimator, OutlierMixin, RegressorMixin from ._base import LinearClassifierMixin, SparseCoefMixin from ._sgd_fast import ( LossFunction, ) # Authors: Peter Prettenhofer (main author) # Mathieu Blondel (partial_fit support) # # License: BSD 3 clause LEARNING_RATE_TYPES: dict = ... PENALTY_TYPES: dict = ... DEFAULT_EPSILON: float = ... # Default value of ``epsilon`` parameter. MAX_INT = ... class _ValidationScoreCallback: def __init__( self, estimator: SGDClassifier, X_val: csr_matrix | ndarray, y_val: ndarray, sample_weight_val: ndarray, classes: None | ndarray = None, ) -> None: ... def __call__(self, coef: ndarray, intercept: float) -> Float: ... class BaseSGD(SparseCoefMixin, BaseEstimator, metaclass=ABCMeta): _parameter_constraints: ClassVar[dict] = ... def __init__( self, loss: str, *, penalty: str = "l2", alpha: float = 0.0001, C: float = 1.0, l1_ratio: float = 0.15, fit_intercept: bool = True, max_iter: int = 1000, tol: float = 1e-3, shuffle: bool = True, verbose: int = 0, epsilon: float = 0.1, random_state=None, learning_rate: str = "optimal", eta0: float = 0.0, power_t: float = 0.5, early_stopping: bool = False, validation_fraction: float = 0.1, n_iter_no_change: int = 5, warm_start: bool = False, average: bool = False, ) -> None: ... @abstractmethod def fit(self, X, y): ... def fit_binary( est: BaseEstimator, i: Int, X: MatrixLike, y: ArrayLike, alpha: Float, C: Float, learning_rate: str, max_iter: Int, pos_weight: Float, neg_weight: Float, sample_weight: ArrayLike, validation_mask: None | ArrayLike = None, random_state: RandomState | None | Int = None, ) -> tuple[ndarray, float, int]: ... class BaseSGDClassifier(LinearClassifierMixin, BaseSGD, metaclass=ABCMeta): loss_functions: ClassVar[dict] = ... _parameter_constraints: ClassVar[dict] = ... @abstractmethod def __init__( self, loss: str = "hinge", *, penalty: str = "l2", alpha: float = 0.0001, l1_ratio: float = 0.15, fit_intercept: bool = True, max_iter: int = 1000, tol: float = 1e-3, shuffle: bool = True, verbose: int = 0, epsilon=..., n_jobs=None, random_state=None, learning_rate: str = "optimal", eta0: float = 0.0, power_t: float = 0.5, early_stopping: bool = False, validation_fraction: float = 0.1, n_iter_no_change: int = 5, class_weight=None, warm_start: bool = False, average: bool = False, ) -> None: ... def partial_fit( self, X: MatrixLike, y: ArrayLike, classes: None | ArrayLike = None, sample_weight: None | ArrayLike = None, ) -> Self: ... def fit( self, X: MatrixLike, y: ArrayLike, coef_init: None | MatrixLike = None, intercept_init: None | ArrayLike = None, sample_weight: None | ArrayLike = None, ) -> Self: ... class SGDClassifier(BaseSGDClassifier): feature_names_in_: ndarray = ... n_features_in_: int = ... t_: int = ... classes_: ndarray = ... loss_function_: LossFunction = ... n_iter_: int = ... intercept_: ndarray = ... coef_: ndarray = ... _parameter_constraints: ClassVar[dict] = ... def __init__( self, loss: Literal[ "hinge", "log_loss", "log", "modified_huber", "squared_hinge", "perceptron", "squared_error", "huber", "epsilon_insensitive", "squared_epsilon_insensitive", ] = "hinge", *, penalty: None | Literal["l2", "l1", "elasticnet"] = "l2", alpha: Float = 0.0001, l1_ratio: Float = 0.15, fit_intercept: bool = True, max_iter: Int = 1000, tol: None | Float = 1e-3, shuffle: bool = True, verbose: Int = 0, epsilon: Float = ..., n_jobs: None | Int = None, random_state: RandomState | None | Int = None, learning_rate: str = "optimal", eta0: Float = 0.0, power_t: Float = 0.5, early_stopping: bool = False, validation_fraction: Float = 0.1, n_iter_no_change: Int = 5, class_weight: Mapping[str, float] | None | str = None, warm_start: bool = False, average: int | bool = False, ) -> None: ... def predict_proba(self, X: MatrixLike) -> ndarray: ... def predict_log_proba(self, X: MatrixLike | ArrayLike) -> ndarray: ... class BaseSGDRegressor(RegressorMixin, BaseSGD): loss_functions: ClassVar[dict] = ... _parameter_constraints: ClassVar[dict] = ... @abstractmethod def __init__( self, loss: str = "squared_error", *, penalty: str = "l2", alpha: float = 0.0001, l1_ratio: float = 0.15, fit_intercept: bool = True, max_iter: int = 1000, tol: float = 1e-3, shuffle: bool = True, verbose: int = 0, epsilon=..., random_state=None, learning_rate: str = "invscaling", eta0: float = 0.01, power_t: float = 0.25, early_stopping: bool = False, validation_fraction: float = 0.1, n_iter_no_change: int = 5, warm_start: bool = False, average: bool = False, ) -> None: ... def partial_fit( self, X: MatrixLike, y: ArrayLike, sample_weight: None | ArrayLike = None, ) -> Self: ... def fit( self, X: MatrixLike, y: ArrayLike, coef_init: None | ArrayLike = None, intercept_init: None | ArrayLike = None, sample_weight: None | ArrayLike = None, ) -> SGDRegressor | Self: ... def predict(self, X: MatrixLike) -> ndarray: ... class SGDRegressor(BaseSGDRegressor): feature_names_in_: ndarray = ... n_features_in_: int = ... t_: int = ... n_iter_: int = ... intercept_: ndarray = ... coef_: ndarray = ... _parameter_constraints: ClassVar[dict] = ... def __init__( self, loss: str = "squared_error", *, penalty: None | Literal["l2", "l1", "elasticnet"] = "l2", alpha: Float = 0.0001, l1_ratio: Float = 0.15, fit_intercept: bool = True, max_iter: Int = 1000, tol: None | Float = 1e-3, shuffle: bool = True, verbose: Int = 0, epsilon: Float = ..., random_state: RandomState | None | Int = None, learning_rate: str = "invscaling", eta0: Float = 0.01, power_t: Float = 0.25, early_stopping: bool = False, validation_fraction: Float = 0.1, n_iter_no_change: Int = 5, warm_start: bool = False, average: int | bool = False, ) -> None: ... class SGDOneClassSVM(BaseSGD, OutlierMixin): feature_names_in_: ndarray = ... n_features_in_: int = ... loss_function_: LossFunction = ... t_: int = ... n_iter_: int = ... offset_: ndarray = ... coef_: ndarray = ... loss_functions: ClassVar[dict] = ... _parameter_constraints: ClassVar[dict] = ... def __init__( self, nu: Float = 0.5, fit_intercept: bool = True, max_iter: Int = 1000, tol: None | Float = 1e-3, shuffle: bool = True, verbose: Int = 0, random_state: RandomState | None | Int = None, learning_rate: Literal["optimal", "constant", "invscaling", "adaptive"] = "optimal", eta0: Float = 0.0, power_t: Float = 0.5, warm_start: bool = False, average: int | bool = False, ) -> None: ... def partial_fit( self, X: MatrixLike, y: Any = None, sample_weight: None | ArrayLike = None, ) -> Self: ... def fit( self, X: MatrixLike, y: Any = None, coef_init: None | MatrixLike = None, offset_init: None | ArrayLike = None, sample_weight: None | ArrayLike = None, ) -> Self: ... def decision_function(self, X: MatrixLike) -> ndarray: ... def score_samples(self, X: MatrixLike) -> ndarray: ... def predict(self, X: MatrixLike) -> ndarray: ...