Class CatBoost

catboost/python-package/catboost/core.py:2530–4836

CatBoost model. Contains training, prediction and evaluation methods.

class CatBoost(_CatBoostBase):
    """
    CatBoost model. Contains training, prediction and evaluation methods.
    """

    def __init__(self, params=None):
        """
        Initialize the CatBoost.

        Parameters
        ----------
        params : dict
            Parameters for CatBoost.
            If  None, all params are set to their defaults.
            If  dict, overriding parameters present in dict.
        """
        super(CatBoost, self).__init__(params)

    def _dataset_train_eval_split(self, train_pool, params, save_eval_pool):
        """
        returns:
            train_pool, eval_pool
                eval_pool will be uninitialized if save_eval_pool is false
        """

        is_classification = self._is_classifier(params)

        return train_pool.train_eval_split(
            params.get('has_time', False),
            is_classification,
            params['eval_fraction'],
            save_eval_pool
        )

    def _prepare_train_params(self, X=None, y=None, cat_features=None, text_features=None, embedding_features=None,
                              pairs=None, graph=None, sample_weight=None, group_id=None, group_weight=None, subgroup_id=None,
                              pairs_weight=None, baseline=None, use_best_model=None, eval_set=None, verbose=None,
                              logging_level=None, plot=None, plot_file=None, column_description=None, verbose_eval=None,
                              metric_period=None, silent=None, early_stopping_rounds=None, save_snapshot=None,
                              snapshot_file=None, snapshot_interval=None, init_model=None, callbacks=None):
        params = deepcopy(self._get_canonized_params())

        if isinstance(X, FeaturesData):
            warnings.warn("FeaturesData is deprecated for using in fit function "
                          "and soon will not be supported. If you want to use FeaturesData, "
                          "please pass it to Pool initialization and use Pool in fit")

        cat_features = _process_feature_indices(cat_features, X, params, 'cat_features')
        text_features = _process_feature_indices(text_features, X, params, 'text_features')
        embedding_features = _process_feature_indices(embedding_features, X, params, 'embedding_features')

        train_pool = _build_train_pool(X, y, cat_features, text_features, embedding_features, pairs, graph,
                                       sample_weight, group_id, group_weight, subgroup_id, pairs_weight,
                                       baseline, column_description)
        if train_pool.is_empty_:
            raise CatBoostError("X is empty.")

        allow_clear_pool = not isinstance(X, Pool)