Options

This section documents the options components of the ensemble module.

options

Defines sklearn.ensemble options interoperability.

This module provides classes to handle scikit-learn ensemble model options for integration with Nextmv.

CLASS	DESCRIPTION
GradientBoostingRegressorOptions	Options wrapper for sklearn.ensemble.GradientBoostingRegressor.
RandomForestRegressorOptions	Options wrapper for sklearn.ensemble.RandomForestRegressor.

Variables

GRADIENT_BOOSTING_REGRESSOR_PARAMETERS List of option parameters for GradientBoostingRegressor. RANDOM_FOREST_REGRESSOR_PARAMETERS List of option parameters for RandomForestRegressor.

GRADIENT_BOOSTING_REGRESSOR_PARAMETERS module-attribute

GRADIENT_BOOSTING_REGRESSOR_PARAMETERS = [
    Option(
        name="loss",
        option_type=str,
        choices=[
            "squared_error",
            "absolute_error",
            "huber",
            "quantile",
        ],
        description="Loss function to be optimized.",
    ),
    Option(
        name="learning_rate",
        option_type=float,
        description="Learning rate shrinks the contribution of each tree by learning_rate.",
    ),
    Option(
        name="n_estimators",
        option_type=int,
        description="The number of boosting stages to perform.",
    ),
    Option(
        name="subsample",
        option_type=float,
        description="The fraction of samples to be used for fitting the individual base learners.",
    ),
    Option(
        name="criterion",
        option_type=str,
        choices=["friedman_mse", "squared_error"],
        description="The function to measure the quality of a split.",
    ),
    Option(
        name="min_samples_split",
        option_type=int,
        description="The minimum number of samples required to split an internal node.",
    ),
    Option(
        name="min_samples_leaf",
        option_type=int,
        description="The minimum number of samples required to be at a leaf node.",
    ),
    Option(
        name="min_weight_fraction_leaf",
        option_type=float,
        description="The minimum weighted fraction of the sum total of weights required to be at a leaf node.",
    ),
    Option(
        name="max_depth",
        option_type=int,
        description="Maximum depth of the individual regression estimators.",
    ),
    Option(
        name="min_impurity_decrease",
        option_type=float,
        description="A node will be split if this split induces a decrease of the impurity greater than or equal to this value.",
    ),
    Option(
        name="random_state",
        option_type=int,
        description="Controls the random seed given to each Tree estimator at each boosting iteration.",
    ),
    Option(
        name="max_features",
        option_type=int,
        description="The number of features to consider when looking for the best split.",
    ),
    Option(
        name="alpha",
        option_type=float,
        description="The alpha-quantile of the huber loss function and the quantile loss function.",
    ),
    Option(
        name="max_leaf_nodes",
        option_type=int,
        description="Grow trees with max_leaf_nodes in best-first fashion.",
    ),
    Option(
        name="warm_start",
        option_type=bool,
        description="When set to True, reuse the solution of the previous call to fit and add more estimators to the ensemble, otherwise, just erase the previous solution.",
    ),
    Option(
        name="validation_fraction",
        option_type=float,
        description="The proportion of training data to set aside as validation set for early stopping.",
    ),
    Option(
        name="n_iter_no_change",
        option_type=int,
        description="n_iter_no_change is used to decide if early stopping will be used to terminate training when validation score is not improving.",
    ),
    Option(
        name="tol",
        option_type=float,
        description="Tolerance for the early stopping.",
    ),
    Option(
        name="ccp_alpha",
        option_type=float,
        description="Complexity parameter used for Minimal Cost-Complexity Pruning.",
    ),
]

Parameters for scikit-learn's GradientBoostingRegressor.

You can import the GRADIENT_BOOSTING_REGRESSOR_PARAMETERS constant directly from ensemble:

from nextmv_sklearn.ensemble import GRADIENT_BOOSTING_REGRESSOR_PARAMETERS

This constant defines all available parameters that can be configured for the scikit-learn GradientBoostingRegressor model through Nextmv options system.

See Also

sklearn.ensemble.GradientBoostingRegressor : The scikit-learn class this configures. GradientBoostingRegressorOptions : Class that uses these parameters.

GradientBoostingRegressorOptions

GradientBoostingRegressorOptions()

Options for the sklearn.ensemble.GradientBoostingRegressor.

You can import the GradientBoostingRegressorOptions class directly from ensemble:

from nextmv_sklearn.ensemble import GradientBoostingRegressorOptions

This class provides configuration options for scikit-learn's GradientBoostingRegressor, allowing integration with Nextmv's options system. It encapsulates all parameters available for the scikit-learn implementation.

ATTRIBUTE	DESCRIPTION
params	List of Nextmv options parameters for GradientBoostingRegressor configuration. TYPE: list

Examples:

>>> from nextmv_sklearn.ensemble import GradientBoostingRegressorOptions
>>> options = GradientBoostingRegressorOptions()
>>> nextmv_options = options.to_nextmv()

Initialize GradientBoostingRegressorOptions.

Sets up the parameters from the predefined list of GradientBoostingRegressor parameters.

Source code in nextmv-scikit-learn/nextmv_sklearn/ensemble/options.py

def __init__(self):
    """Initialize GradientBoostingRegressorOptions.

    Sets up the parameters from the predefined list of GradientBoostingRegressor
    parameters.
    """
    self.params = GRADIENT_BOOSTING_REGRESSOR_PARAMETERS

params instance-attribute

params = GRADIENT_BOOSTING_REGRESSOR_PARAMETERS

to_nextmv

to_nextmv() -> Options

Converts the options to a Nextmv options object.

Transforms the internal parameters into a Nextmv options object that can be used for configuration and command-line parsing.

RETURNS	DESCRIPTION
Options	A Nextmv options object containing all parameters for the GradientBoostingRegressor.

Source code in nextmv-scikit-learn/nextmv_sklearn/ensemble/options.py

def to_nextmv(self) -> nextmv.Options:
    """Converts the options to a Nextmv options object.

    Transforms the internal parameters into a Nextmv options object that can be
    used for configuration and command-line parsing.

    Returns
    -------
    nextmv.Options
        A Nextmv options object containing all parameters for the GradientBoostingRegressor.
    """

    return nextmv.Options(*self.params)

RANDOM_FOREST_REGRESSOR_PARAMETERS module-attribute

RANDOM_FOREST_REGRESSOR_PARAMETERS = [
    Option(
        name="n_estimators",
        option_type=int,
        description="The number of trees in the forest.",
    ),
    Option(
        name="criterion",
        option_type=str,
        choices=[
            "squared_error",
            "absolute_error",
            "friedman_mse",
            "poisson",
        ],
        description="The function to measure the quality of a split.",
    ),
    Option(
        name="max_depth",
        option_type=int,
        description="The maximum depth of the tree.",
    ),
    Option(
        name="min_samples_split",
        option_type=int,
        description="The minimum number of samples required to split an internal node.",
    ),
    Option(
        name="min_samples_leaf",
        option_type=int,
        description="The minimum number of samples required to be at a leaf node.",
    ),
    Option(
        name="min_weight_fraction_leaf",
        option_type=float,
        description="The minimum weighted fraction of the sum total of weights required to be at a leaf node.",
    ),
    Option(
        name="max_features",
        option_type=int,
        description="The number of features to consider when looking for the best split.",
    ),
    Option(
        name="max_leaf_nodes",
        option_type=int,
        description="Grow trees with max_leaf_nodes in best-first fashion.",
    ),
    Option(
        name="min_impurity_decrease",
        option_type=float,
        description="A node will be split if this split induces a decrease of the impurity greater than or equal to this value.",
    ),
    Option(
        name="bootstrap",
        option_type=bool,
        description="Whether bootstrap samples are used when building trees.",
    ),
    Option(
        name="oob_score",
        option_type=bool,
        description="Whether to use out-of-bag samples to estimate the generalization score.",
    ),
    Option(
        name="n_jobs",
        option_type=int,
        description="The number of jobs to run in parallel.",
    ),
    Option(
        name="random_state",
        option_type=int,
        description="Controls both the randomness of the bootstrapping of the samples used when building trees and the sampling of the features.",
    ),
    Option(
        name="verbose",
        option_type=int,
        description="Controls the verbosity when fitting and predicting.",
    ),
    Option(
        name="warm_start",
        option_type=bool,
        description="When set to True, reuse the solution of the previous call to fit and add more estimators to the ensemble, otherwise, just erase the previous solution.",
    ),
    Option(
        name="ccp_alpha",
        option_type=float,
        description="Complexity parameter used for Minimal Cost-Complexity Pruning.",
    ),
    Option(
        name="max_samples",
        option_type=int,
        description="If bootstrap is True, the number of samples to draw from X to train each base estimator.",
    ),
    Option(
        name="monotonic_cst",
        option_type=int,
        description="Indicates the monotonicity constraint to enforce on each feature.",
    ),
]

Parameters for scikit-learn's RandomForestRegressor.

You can import the RANDOM_FOREST_REGRESSOR_PARAMETERS constant directly from ensemble:

from nextmv_sklearn.ensemble import RANDOM_FOREST_REGRESSOR_PARAMETERS

This constant defines all available parameters that can be configured for the scikit-learn RandomForestRegressor model through Nextmv options system.

See Also

sklearn.ensemble.RandomForestRegressor : The scikit-learn class this configures. RandomForestRegressorOptions : Class that uses these parameters.

RandomForestRegressorOptions

RandomForestRegressorOptions()

Options for the sklearn.ensemble.RandomForestRegressor.

You can import the RandomForestRegressorOptions class directly from ensemble:

from nextmv_sklearn.ensemble import RandomForestRegressorOptions

This class provides configuration options for scikit-learn's RandomForestRegressor, allowing integration with Nextmv's options system. It encapsulates all parameters available for the scikit-learn implementation.

ATTRIBUTE	DESCRIPTION
params	List of Nextmv options parameters for RandomForestRegressor configuration. TYPE: list

Examples:

>>> from nextmv_sklearn.ensemble import RandomForestRegressorOptions
>>> options = RandomForestRegressorOptions()
>>> nextmv_options = options.to_nextmv()

Initialize RandomForestRegressorOptions.

Sets up the parameters from the predefined list of RandomForestRegressor parameters.

Source code in nextmv-scikit-learn/nextmv_sklearn/ensemble/options.py

def __init__(self):
    """Initialize RandomForestRegressorOptions.

    Sets up the parameters from the predefined list of RandomForestRegressor
    parameters.
    """
    self.params = RANDOM_FOREST_REGRESSOR_PARAMETERS

params instance-attribute

params = RANDOM_FOREST_REGRESSOR_PARAMETERS

to_nextmv

to_nextmv() -> Options

Converts the options to a Nextmv options object.

Transforms the internal parameters into a Nextmv options object that can be used for configuration and command-line parsing.

RETURNS	DESCRIPTION
Options	A Nextmv options object containing all parameters for the RandomForestRegressor.

Source code in nextmv-scikit-learn/nextmv_sklearn/ensemble/options.py

def to_nextmv(self) -> nextmv.Options:
    """Converts the options to a Nextmv options object.

    Transforms the internal parameters into a Nextmv options object that can be
    used for configuration and command-line parsing.

    Returns
    -------
    nextmv.Options
        A Nextmv options object containing all parameters for the RandomForestRegressor.
    """

    return nextmv.Options(*self.params)