analysis/gradient_model.py

import numpy as np
from sklearn.ensemble import GradientBoostingRegressor
from sklearn.model_selection import learning_curve

from analysis.shap_model import draw_shap_beeswarm
from metrics.calculate_regression_metrics import calculate_regression_metrics
from static.config import Config
from static.new_class import Container
from static.process import grid_search, bayes_search


class GradientBoostingParams:
    @classmethod
    def get_params(cls):
        return {
            'n_estimators': [50, 100, 150],
            'learning_rate': [0.01, 0.1, 0.2],
            'max_depth': [3, 5, 7],
            'min_samples_split': [2, 5, 10],
            'min_samples_leaf': [1, 2, 4]
        }


# 梯度提升回归
def gradient_boosting_regression(container: Container):
    x_train = container.x_train
    y_train = container.y_train
    x_test = container.x_test
    y_test = container.y_test
    hyper_params_optimize = container.hyper_params_optimize
    info = {}

    gradient_boosting_regression_model = GradientBoostingRegressor(random_state=Config.RANDOM_STATE)
    params = GradientBoostingParams.get_params()

    if hyper_params_optimize == "grid_search":
        best_model = grid_search(params, gradient_boosting_regression_model, x_train, y_train)
    elif hyper_params_optimize == "bayes_search":
        best_model = bayes_search(params, gradient_boosting_regression_model, x_train, y_train)
    else:
        best_model = gradient_boosting_regression_model
        best_model.fit(x_train, y_train)

    info["参数"] = best_model.get_params()

    y_pred = best_model.predict(x_test)
    # y_pred = best_model.predict(x_test).reshape(-1, 1)
    container.set_y_pred(y_pred)

    train_sizes, train_scores, test_scores = learning_curve(best_model, x_train, y_train, cv=5)

    train_scores_mean = np.mean(train_scores, axis=1)
    train_scores_std = np.std(train_scores, axis=1)
    test_scores_mean = np.mean(test_scores, axis=1)
    test_scores_std = np.std(test_scores, axis=1)
    container.set_learning_curve_values(train_sizes, train_scores_mean, train_scores_std, test_scores_mean,
                                        test_scores_std)

    info["指标"] = calculate_regression_metrics(y_pred, y_test)

    container.set_info(info)
    container.set_status("trained")
    container.set_model(best_model)

    return container