Source code for blocksnet.machine_learning.strategy.sklearn.ensemble.base_strategy

from abc import ABC
from ..base_strategy import SKLearnBaseStrategy
from sklearn.base import BaseEstimator
from blocksnet.machine_learning.strategy import BaseStrategy

class SKLearnEnsembleBaseStrategy(SKLearnBaseStrategy, ABC):

    def __init__(self, model_cls : type[BaseEstimator], estimators : list[tuple[str, BaseEstimator]], model_params : dict | None):
        super().__init__(model_cls, model_params)
        self._estimators = estimators

    @property
    def estimators(self) -> list[tuple[str, BaseEstimator]]:
        if self._estimators is None or len(self._estimators) == 0:
            raise ValueError("Estimators are not defined")
        return self._estimators
        
    @estimators.setter
    def estimators(self, estimators):
        for name, estimator in estimators:
            if not isinstance(name, str):
                raise ValueError(f"Estimator name {name} is not a valid string")
            if not isinstance(estimator, BaseEstimator):
                raise ValueError(f"Estimator {name} is not a valid estimator")
        self._estimators = estimators

    def _load_model(self, path):
        super()._load_model(path)
        self.estimators = self.model.estimators