#!/usr/bin/env python3
"""step29_compress_predictive.py — 压缩预测信号到稀缺（数据定阈值，验证预测力保持）
在四因子基础上加更多预测因子压缩信号量：
- 行业状态（sec_above/sec_ret20）
- 个股技术（bias60深度/量能/距低点）
- 资金流
逐层验证：信号量下降 且 大涨率/avg60 保持或提升
"""
import numpy as np
import pandas as pd

print("=== 加载 ===", flush=True)
panel = pd.read_pickle("/tmp/panel_12d.pkl")
panel = panel.sort_values(["code", "date"]).reset_index(drop=True)
panel["fwd_ret60"] = panel.groupby("code")["close"].transform(lambda x: x.shift(-60)/x - 1) * 100
panel["is_big"] = (panel["fwd_ret60"] >= 50).astype(int)
print("面板:", len(panel), flush=True)

base = panel.dropna(subset=["fwd_ret60"])
print("基线: 大涨率={:.2f}% avg60={:.2f}%".format(
    base["is_big"].mean()*100, base["fwd_ret60"].mean()), flush=True)

def validate(cond, label, min_n=50):
    s = base[cond]
    if len(s) < min_n:
        print("{}: n={} 样本不足".format(label, len(s)), flush=True)
        return None
    rate = s["is_big"].mean()*100
    avg = s["fwd_ret60"].mean()
    wr = (s["fwd_ret60"]>0).mean()*100
    cand = s[["code", "date"]].sort_values(["code", "date"])
    cand["prev"] = cand.groupby("code")["date"].shift(1)
    cand["gap"] = (pd.to_datetime(cand["date"]) - pd.to_datetime(cand["prev"])).dt.days
    cand = cand[(cand["prev"].isna()) | (cand["gap"] > 30)]
    cand["year"] = cand["date"].str[:4]
    n_dedup = len(cand)
    monthly = n_dedup / max(len(cand["year"].unique()),1) / 12
    print("{}: n={} 大涨率={:.2f}% avg60={:.2f}% wr={:.1f}% | 去重{} 月均{:.1f}".format(
        label, len(s), rate, avg, wr, n_dedup, monthly), flush=True)
    return {"n": n_dedup, "monthly": monthly, "rate": rate, "avg": avg, "wr": wr}

# 基础四因子
c4 = (panel["mkt_rsi"] < 41) & (panel["mcap_q"] < 0.2) & (panel["pe_q"] < 0.2) & (panel["news3"] >= 1)
print("\n=== 四因子基准 ===", flush=True)
validate(c4, "四因子(弱市+小市值+低PE+新闻)")

print("\n=== 加行业状态 ===", flush=True)
validate(c4 & (panel["sec_above"] == 0), "+行业MA20下")
validate(c4 & (panel["sec_ret20"] < 0), "+行业20日跌")
validate(c4 & (panel["sec_ret20"] < -5), "+行业20日深跌")

print("\n=== 加个股技术 ===", flush=True)
validate(c4 & (panel["bias60"] < -10), "+bias60<-10(深跌)")
validate(c4 & (panel["bias60"] < -15), "+bias60<-15")
validate(c4 & (panel["bias60"] < -20), "+bias60<-20")
validate(c4 & (panel["vol_ratio"] < 1.0), "+缩量")
validate(c4 & (panel["dist_lo20"] < 5), "+距前低<5%")

print("\n=== 组合最优方向 ===", flush=True)
# 行业弱 + 深跌
validate(c4 & (panel["sec_ret20"] < -5) & (panel["bias60"] < -15), "四因子+行业深跌+bias60<-15")
validate(c4 & (panel["sec_ret20"] < -5) & (panel["bias60"] < -10) & (panel["dist_lo20"] < 8), "四因子+行业深跌+bias60<-10+距低点<8")

print("\n=== 完成 ===", flush=True)