Skip to content

GradientBoostingRegressor

Bases: Regressor

Gradient boosting regression.

Parameters:

Name Type Description Default
number_of_trees int

The number of boosting stages to perform. Gradient boosting is fairly robust to over-fitting so a large number usually results in better performance.

 100
learning_rate float

The larger the value, the more the model is influenced by each additional tree. If the learning rate is too low, the model might underfit. If the learning rate is too high, the model might overfit.

 0.1

Raises:

Type Description
OutOfBoundsError

If number_of_trees or learning_rate are less than or equal to 0.
Source code in src/safeds/ml/classical/regression/_gradient_boosting.py 
 19
 20
 21
 22
 23
 24
 25
 26
 27
 28
 29
 30
 31
 32
 33
 34
 35
 36
 37
 38
 39
 40
 41
 42
 43
 44
 45
 46
 47
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
class GradientBoostingRegressor(Regressor):
    """
    Gradient boosting regression.

    Parameters
    ----------
    number_of_trees:
        The number of boosting stages to perform. Gradient boosting is fairly robust to over-fitting so a large
        number usually results in better performance.
    learning_rate:
        The larger the value, the more the model is influenced by each additional tree. If the learning rate is too
        low, the model might underfit. If the learning rate is too high, the model might overfit.

    Raises
    ------
    OutOfBoundsError
        If `number_of_trees` or `learning_rate` are less than or equal to 0.
    """

    def __hash__(self) -> int:
        return _structural_hash(
            Regressor.__hash__(self),
            self._target_name,
            self._feature_names,
            self._learning_rate,
            self._number_of_trees,
        )

    def __init__(self, *, number_of_trees: int = 100, learning_rate: float = 0.1) -> None:
        # Validation
        if number_of_trees < 1:
            raise OutOfBoundsError(number_of_trees, name="number_of_trees", lower_bound=ClosedBound(1))
        if learning_rate <= 0:
            raise OutOfBoundsError(learning_rate, name="learning_rate", lower_bound=OpenBound(0))

        # Hyperparameters
        self._number_of_trees = number_of_trees
        self._learning_rate = learning_rate

        # Internal state
        self._wrapped_regressor: sk_GradientBoostingRegressor | None = None
        self._feature_names: list[str] | None = None
        self._target_name: str | None = None

    @property
    def number_of_trees(self) -> int:
        """
        Get the number of trees (estimators) in the ensemble.

        Returns
        -------
        result:
            The number of trees.
        """
        return self._number_of_trees

    @property
    def learning_rate(self) -> float:
        """
        Get the learning rate.

        Returns
        -------
        result:
            The learning rate.
        """
        return self._learning_rate

    def fit(self, training_set: TabularDataset | ExperimentalTabularDataset) -> GradientBoostingRegressor:
        """
        Create a copy of this regressor and fit it with the given training data.

        This regressor is not modified.

        Parameters
        ----------
        training_set:
            The training data containing the feature and target vectors.

        Returns
        -------
        fitted_regressor:
            The fitted regressor.

        Raises
        ------
        LearningError
            If the training data contains invalid values or if the training failed.
        TypeError
            If a table is passed instead of a tabular dataset.
        NonNumericColumnError
            If the training data contains non-numerical values.
        MissingValuesColumnError
            If the training data contains missing values.
        DatasetMissesDataError
            If the training data contains no rows.
        """
        wrapped_regressor = self._get_sklearn_regressor()
        fit(wrapped_regressor, training_set)

        result = GradientBoostingRegressor(number_of_trees=self._number_of_trees, learning_rate=self._learning_rate)
        result._wrapped_regressor = wrapped_regressor
        result._feature_names = training_set.features.column_names
        result._target_name = training_set.target.name

        return result

    def predict(self, dataset: Table | ExperimentalTable | ExperimentalTabularDataset) -> TabularDataset:
        """
        Predict a target vector using a dataset containing feature vectors. The model has to be trained first.

        Parameters
        ----------
        dataset:
            The dataset containing the feature vectors.

        Returns
        -------
        table:
            A dataset containing the given feature vectors and the predicted target vector.

        Raises
        ------
        ModelNotFittedError
            If the model has not been fitted yet.
        DatasetMissesFeaturesError
            If the dataset misses feature columns.
        PredictionError
            If predicting with the given dataset failed.
        NonNumericColumnError
            If the dataset contains non-numerical values.
        MissingValuesColumnError
            If the dataset contains missing values.
        DatasetMissesDataError
            If the dataset contains no rows.
        """
        return predict(self._wrapped_regressor, dataset, self._feature_names, self._target_name)

    @property
    def is_fitted(self) -> bool:
        """Whether the regressor is fitted."""
        return self._wrapped_regressor is not None

    def _get_sklearn_regressor(self) -> RegressorMixin:
        """
        Return a new wrapped Regressor from sklearn.

        Returns
        -------
        wrapped_regressor:
            The sklearn Regressor.
        """
        from sklearn.ensemble import GradientBoostingRegressor as sk_GradientBoostingRegressor

        return sk_GradientBoostingRegressor(n_estimators=self._number_of_trees, learning_rate=self._learning_rate)

is_fitted: bool property

Whether the regressor is fitted.

learning_rate: float property

Get the learning rate.

Returns:

Name Type Description
result float

The learning rate.

number_of_trees: int property

Get the number of trees (estimators) in the ensemble.

Returns:

Name Type Description
result int

The number of trees.

fit(training_set)

Create a copy of this regressor and fit it with the given training data.

This regressor is not modified.

Parameters:

Name Type Description Default
training_set TabularDataset | ExperimentalTabularDataset

The training data containing the feature and target vectors.

 required

Returns:

Name Type Description
fitted_regressor GradientBoostingRegressor

The fitted regressor.

Raises:

Type Description
LearningError

If the training data contains invalid values or if the training failed.
TypeError

If a table is passed instead of a tabular dataset.
NonNumericColumnError

If the training data contains non-numerical values.
MissingValuesColumnError

If the training data contains missing values.
DatasetMissesDataError

If the training data contains no rows.
Source code in src/safeds/ml/classical/regression/_gradient_boosting.py 
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
def fit(self, training_set: TabularDataset | ExperimentalTabularDataset) -> GradientBoostingRegressor:
    """
    Create a copy of this regressor and fit it with the given training data.

    This regressor is not modified.

    Parameters
    ----------
    training_set:
        The training data containing the feature and target vectors.

    Returns
    -------
    fitted_regressor:
        The fitted regressor.

    Raises
    ------
    LearningError
        If the training data contains invalid values or if the training failed.
    TypeError
        If a table is passed instead of a tabular dataset.
    NonNumericColumnError
        If the training data contains non-numerical values.
    MissingValuesColumnError
        If the training data contains missing values.
    DatasetMissesDataError
        If the training data contains no rows.
    """
    wrapped_regressor = self._get_sklearn_regressor()
    fit(wrapped_regressor, training_set)

    result = GradientBoostingRegressor(number_of_trees=self._number_of_trees, learning_rate=self._learning_rate)
    result._wrapped_regressor = wrapped_regressor
    result._feature_names = training_set.features.column_names
    result._target_name = training_set.target.name

    return result

predict(dataset)

Predict a target vector using a dataset containing feature vectors. The model has to be trained first.

Parameters:

Name Type Description Default
dataset Table | ExperimentalTable | ExperimentalTabularDataset

The dataset containing the feature vectors.

 required

Returns:

Name Type Description
table TabularDataset

A dataset containing the given feature vectors and the predicted target vector.

Raises:

Type Description
ModelNotFittedError

If the model has not been fitted yet.
DatasetMissesFeaturesError

If the dataset misses feature columns.
PredictionError

If predicting with the given dataset failed.
NonNumericColumnError

If the dataset contains non-numerical values.
MissingValuesColumnError

If the dataset contains missing values.
DatasetMissesDataError

If the dataset contains no rows.
Source code in src/safeds/ml/classical/regression/_gradient_boosting.py 
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
def predict(self, dataset: Table | ExperimentalTable | ExperimentalTabularDataset) -> TabularDataset:
    """
    Predict a target vector using a dataset containing feature vectors. The model has to be trained first.

    Parameters
    ----------
    dataset:
        The dataset containing the feature vectors.

    Returns
    -------
    table:
        A dataset containing the given feature vectors and the predicted target vector.

    Raises
    ------
    ModelNotFittedError
        If the model has not been fitted yet.
    DatasetMissesFeaturesError
        If the dataset misses feature columns.
    PredictionError
        If predicting with the given dataset failed.
    NonNumericColumnError
        If the dataset contains non-numerical values.
    MissingValuesColumnError
        If the dataset contains missing values.
    DatasetMissesDataError
        If the dataset contains no rows.
    """
    return predict(self._wrapped_regressor, dataset, self._feature_names, self._target_name)