Chan/research/step7_nested_fx.py

"""Step 7：区间套验证。

逐层加过滤，看每一层带来多少增量：
  L0  所有 1h 分型
  L1  + 背驰（价格创新极值但 MACD 面积衰减）
  L2  + 大级别中枢位置（买贴支撑 / 卖贴压力）
  L3  L1 + L2 组合
对照组：单级别三类买卖点（滞后 16~21 根）。
"""
from __future__ import annotations

import argparse
import sys
from pathlib import Path

import numpy as np
import pandas as pd

sys.path.insert(0, str(Path(__file__).resolve().parent))

from lib.bsp_eval import baseline_stats
from lib.data import fetch_ohlcv
from lib.fx_signal import add_forward_returns, extract_fx_signals, signals_to_frame
from lib.nested_level import annotate_position, build_htf_zones

sys.path.insert(0, str(Path(__file__).resolve().parents[1]))
from chanlun import TF_DF

pd.set_option("display.width", 240)

HORIZONS = (3, 5, 10, 20, 40)


def summarize(g: pd.DataFrame, df: pd.DataFrame, label: str, min_n: int = 10) -> list[dict]:
    base = baseline_stats(df, HORIZONS).set_index("horizon")
    out = []
    for h in HORIZONS:
        col = f"ret_{h}"
        if col not in g:
            continue
        r = g[col].dropna().to_numpy()
        if len(r) < min_n:
            continue
        dirs = g.loc[g[col].notna(), "direction"].to_numpy()
        sd = r.std(ddof=1)
        out.append({
            "分组": label, "持有": h, "n": len(r),
            "收益": r.mean(), "胜率": (r > 0).mean(),
            "超额": r.mean() - float(np.mean(dirs) * base.loc[h, "base_mean_long"]),
            "t值": r.mean() / (sd / np.sqrt(len(r))) if sd else np.nan,
        })
    return out


def show(rows: list[dict]) -> None:
    if not rows:
        print("  (样本不足)")
        return
    d = pd.DataFrame(rows)
    d["收益"] = d["收益"].map(lambda v: f"{v * 100:+.2f}%")
    d["超额"] = d["超额"].map(lambda v: f"{v * 100:+.2f}%")
    d["胜率"] = d["胜率"].map(lambda v: f"{v * 100:.0f}%")
    d["t值"] = d["t值"].map(lambda v: f"{v:+.2f}")
    print(d.to_string(index=False))


def main() -> None:
    ap = argparse.ArgumentParser()
    ap.add_argument("--symbol", default="BTC/USDT:USDT")
    ap.add_argument("--ltf", default="1h", help="小级别：出分型信号")
    ap.add_argument("--htf", default="1d", help="大级别：出支撑压力中枢")
    ap.add_argument("--tol", type=float, default=0.01, help="贴近边界的相对阈值")
    args = ap.parse_args()

    df = fetch_ohlcv(args.symbol, args.ltf, 10**9)
    df_htf = fetch_ohlcv(args.symbol, args.htf, 10**9)
    print(f"[data] 小级别 {args.ltf} rows={len(df)}  大级别 {args.htf} rows={len(df_htf)}")
    print(f"       {df['date'].iloc[0]} -> {df['date'].iloc[-1]}\n")

    chan = TF_DF(df, 1, args.ltf)
    sigs = extract_fx_signals(chan, chan.dataframe)
    sig = signals_to_frame(sigs)
    print(f"[分型] 共 {len(sig)} 个（底 {int((sig.direction == 1).sum())} / "
          f"顶 {int((sig.direction == -1).sum())}）")
    print(f"[确认滞后] 中位数 {sig['lag'].median():.0f} 根  均值 {sig['lag'].mean():.2f} 根  "
          f"P90 {sig['lag'].quantile(.9):.0f} 根  最大 {sig['lag'].max()} 根")
    print(f"  —— 对照：笔 9~10 根、线段 101~121 根、三类买卖点 16~21 根\n")

    sig = add_forward_returns(sig, chan.dataframe, HORIZONS)
    zones = build_htf_zones(df_htf, args.htf)
    print(f"[大级别中枢] {args.htf} 上共 {len(zones)} 个可用中枢")
    sig = annotate_position(sig, chan.dataframe, zones, tol=args.tol)

    div = sig["is_divergence"]
    pos = sig["position_ok"].astype(bool)
    print(f"[过滤器覆盖] 背驰 {div.sum()}/{len(sig)} ({div.mean()*100:.0f}%)   "
          f"位置正确 {pos.sum()}/{len(sig)} ({pos.mean()*100:.0f}%)   "
          f"两者兼备 {(div & pos).sum()}\n")

    print("########## 逐层过滤效果 ##########")
    rows = []
    rows += summarize(sig, chan.dataframe, "L0 全部分型")
    rows += summarize(sig[div], chan.dataframe, "L1 +背驰")
    rows += summarize(sig[pos], chan.dataframe, "L2 +位置")
    rows += summarize(sig[div & pos], chan.dataframe, "L3 背驰+位置")
    show(rows)

    print("\n########## L3 多空拆分 ##########")
    both = sig[div & pos]
    rows = summarize(both[both.direction == 1], chan.dataframe, "L3 做多", min_n=5)
    rows += summarize(both[both.direction == -1], chan.dataframe, "L3 做空", min_n=5)
    show(rows)

    print("\n########## 背驰强度分层（面积比 ratio，越小背驰越强）##########")
    rows = []
    for lo, hi, name in [(0, 0.5, "ratio<0.5"), (0.5, 0.8, "0.5-0.8"),
                         (0.8, 1.0, "0.8-1.0"), (1.0, 99, "ratio>1 无背驰")]:
        g = sig[(sig.ratio >= lo) & (sig.ratio < hi)]
        rows += summarize(g, chan.dataframe, name)
    show(rows)

    print("\n########## 中枢内 vs 中枢外（对应两种玩法）##########")
    rows = summarize(sig[sig.inside_zone.astype(bool)], chan.dataframe, "中枢内做短差")
    rows += summarize(sig[sig.outside_zone.astype(bool)], chan.dataframe, "中枢外做趋势")
    show(rows)

    out_dir = Path(__file__).parent / "out"
    out_dir.mkdir(exist_ok=True)
    tag = f"{args.symbol.split('/')[0]}_{args.ltf}_{args.htf}"
    sig.to_csv(out_dir / f"step7_fx_{tag}.csv", index=False)
    print(f"\n明细已写入 {out_dir}/step7_fx_{tag}.csv")


if __name__ == "__main__":
    main()