ラーニングカーブ（Learning Curve）

from __future__ import annotations

import matplotlib.pyplot as plt
import numpy as np
from sklearn.datasets import make_regression
from sklearn.linear_model import Ridge
from sklearn.model_selection import learning_curve
from sklearn.pipeline import make_pipeline
from sklearn.preprocessing import StandardScaler


def plot_learning_curve_for_ridge() -> None:
    """訓練サンプル数ごとの RMSE を描画して保存する。"""
    features, targets = make_regression(
        n_samples=1500,
        n_features=25,
        n_informative=8,
        noise=12.0,
        random_state=42,
    )

    model = make_pipeline(StandardScaler(), Ridge(alpha=10.0))
    train_sizes, train_scores, valid_scores = learning_curve(
        estimator=model,
        X=features,
        y=targets,
        train_sizes=np.linspace(0.1, 1.0, 8),
        cv=5,
        scoring="neg_mean_squared_error",
        shuffle=True,
        random_state=42,
        n_jobs=None,
    )

    train_rmse = np.sqrt(-train_scores.mean(axis=1))
    valid_rmse = np.sqrt(-valid_scores.mean(axis=1))
    train_std = np.sqrt(train_scores.var(axis=1))
    valid_std = np.sqrt(valid_scores.var(axis=1))

    plt.figure(figsize=(7, 5))
    plt.plot(train_sizes, train_rmse, color="#1d4ed8", label="Train RMSE")
    plt.fill_between(
        train_sizes,
        train_rmse - train_std,
        train_rmse + train_std,
        alpha=0.2,
        color="#1d4ed8",
    )
    plt.plot(train_sizes, valid_rmse, color="#ea580c", label="Validation RMSE")
    plt.fill_between(
        train_sizes,
        valid_rmse - valid_std,
        valid_rmse + valid_std,
        alpha=0.2,
        color="#ea580c",
    )
    plt.xlabel("学習サンプル数")
    plt.ylabel("RMSE")
    plt.title("Learning Curve for Ridge Regression (RMSE)")
    plt.legend(loc="upper right")
    plt.grid(alpha=0.3)


plot_learning_curve_for_ridge()