"""LightGBM Classifier.
This module contains the LightGBMClassifierWrapper, which is a configuration
class for the LightGBMClassifier. It provides parameter spaces for
grid search and Bayesian optimization.
"""
import logging
from typing import Any, Dict, Optional, Union
import pandas as pd
from skopt.space import Categorical, Integer, Real
from ml_grid.model_classes.lightgbm_class import LightGBMClassifier
from ml_grid.util import param_space
from ml_grid.util.global_params import global_parameters
[docs]
class LightGBMClassifierWrapper:
"""LightGBMClassifier with support for both Bayesian and non-Bayesian parameter spaces."""
def __init__(
self,
X: Optional[pd.DataFrame] = None,
y: Optional[pd.Series] = None,
parameter_space_size: Optional[str] = None,
):
"""Initializes the LightGBMClassifierWrapper.
Args:
X (Optional[pd.DataFrame]): Feature matrix for training. Defaults to None.
y (Optional[pd.Series]): Target vector for training. Defaults to None.
parameter_space_size (Optional[str]): Size of the parameter space for
optimization. Defaults to None.
"""
global_params = global_parameters
[docs]
self.algorithm_implementation = (
LightGBMClassifier()
) # Custom scikit-learn wrapper
[docs]
self.method_name = "LightGBMClassifier"
[docs]
self.parameter_vector_space = param_space.ParamSpace(parameter_space_size)
[docs]
self.parameter_space: Union[Dict[str, Any], List[Dict[str, Any]]]
if global_params.bayessearch:
self.parameter_space = {
"boosting_type": Categorical(("gbdt", "dart", "goss")),
"num_leaves": Integer(2, 100),
"learning_rate": Real(1e-5, 1e-1, prior="log-uniform"),
"n_estimators": Integer(50, 500),
"objective": Categorical(("binary",)),
"metric": Categorical(("logloss",)),
"feature_fraction": Real(0.8, 1.0),
}
else:
self.parameter_space = [
{
"boosting_type": ["gbdt", "dart", "goss"],
"num_leaves": list(
self.parameter_vector_space.param_dict.get("log_large_long", [])
),
"learning_rate": list(
self.parameter_vector_space.param_dict.get("log_small", [])
),
"n_estimators": list(
self.parameter_vector_space.param_dict.get("log_large_long", [])
),
"objective": ["binary"],
"metric": ["logloss"],
"feature_fraction": [0.8, 0.9, 1.0],
}
]
logging.getLogger("ml_grid").debug("Imported LightGBM classifier wrapper class")