未来のヒントが見える？XRPの価格を追いかける特別な線「ハルMA」

#!/usr/bin/env python3
"""
Hull Moving Average戦略
遅延を減らした滑らかな移動平均線
"""
import pandas as pd
import numpy as np


def calculate_hull_ma_signals(df: pd.DataFrame, period: int = 20) -> pd.DataFrame:
    """
    Hull Moving Averageシグナルを生成
    
    Parameters:
    -----------
    df : pd.DataFrame
        OHLCVデータ
    period : int
        HMA計算期間（デフォルト: 20）
    
    Returns:
    --------
    pd.DataFrame
        シグナル列が追加されたDataFrame
    """
    df = df.copy()
    
    # WMA（加重移動平均）計算関数
    def wma(series, period):
        weights = np.arange(1, period + 1)
        return series.rolling(period).apply(lambda x: np.dot(x, weights) / weights.sum(), raw=True)
    
    # Hull Moving Average計算
    half_period = int(period / 2)
    sqrt_period = int(np.sqrt(period))
    
    df['wma_half'] = wma(df['close'], half_period)
    df['wma_full'] = wma(df['close'], period)
    df['raw_hma'] = 2 * df['wma_half'] - df['wma_full']
    df['hma'] = wma(df['raw_hma'], sqrt_period)
    
    # HMAの傾き（トレンド方向）
    df['hma_slope'] = df['hma'] - df['hma'].shift(1)
    
    # シグナル生成（HMAの方向転換）
    df['is_buy'] = (
        (df['hma_slope'] > 0) & 
        (df['hma_slope'].shift(1) <= 0) &
        (df['close'] > df['hma'])
    )
    
    df['is_sell'] = (
        (df['hma_slope'] < 0) & 
        (df['hma_slope'].shift(1) >= 0) &
        (df['close'] < df['hma'])
    )
    
    # NaN値をFalseに置換
    df['is_buy'] = df['is_buy'].fillna(False)
    df['is_sell'] = df['is_sell'].fillna(False)
    
    print(f"Hull MA: 期間={period}")
    print(f"買いシグナル数: {df['is_buy'].sum()}")
    print(f"売りシグナル数: {df['is_sell'].sum()}")
    
    return df