"""
diag_isolation.py — Isolate which agent change is causing the 12.83% failure.

Tests:
  A. exp_shared.run_backtest as-is (hazard call + float32 + rolling vol_p60 + per-day OB)
     → Expected: ~12.83%/1739  [already confirmed]

  B. Same as A but WITHOUT set_esoteric_hazard_multiplier call
     → Expected: closer to 111%? This isolates the hazard call.

  C. Same as A but WITHOUT rolling vol_p60 (use static vol_p60 always)
     → Expected: somewhere between A and B?

Goal: confirm hazard call is the dominant regressor.
"""
import sys, time
from pathlib import Path
import numpy as np
import pandas as pd
import math, gc

ROOT = Path(r"C:\Users\Lenovo\Documents\- DOLPHIN NG HD HCM TSF Predict")
sys.path.insert(0, str(ROOT / 'nautilus_dolphin'))
sys.path.insert(0, str(ROOT / 'nautilus_dolphin' / 'dvae'))

import exp_shared
from nautilus_dolphin.nautilus.proxy_boost_engine import create_d_liq_engine
from nautilus_dolphin.nautilus.adaptive_circuit_breaker import AdaptiveCircuitBreaker

print("exp_shared path:", exp_shared.__file__)
print()

# ── Shared data load ──────────────────────────────────────────────────────────
exp_shared.ensure_jit()
d = exp_shared.load_data()

def run_variant(label, use_hazard_call, use_rolling_vol):
    print(f"\n{'='*60}")
    print(f"  {label}")
    print(f"  hazard_call={use_hazard_call}  rolling_vol={use_rolling_vol}")
    print(f"{'='*60}")

    kw = exp_shared.ENGINE_KWARGS.copy()
    kw.update({'sp_maker_entry_rate': 1.0, 'sp_maker_exit_rate': 1.0, 'use_sp_slippage': False})

    acb = AdaptiveCircuitBreaker()
    acb.preload_w750(d['date_strings'])

    eng = create_d_liq_engine(**kw)
    eng.set_ob_engine(d['ob_eng'])
    eng.set_acb(acb)

    if use_hazard_call:
        eng.set_esoteric_hazard_multiplier(0.0)
        lev_after = getattr(eng, 'base_max_leverage', None)
        print(f"  After hazard call: base_max_leverage={lev_after}")
    else:
        lev_now = getattr(eng, 'base_max_leverage', None)
        print(f"  No hazard call: base_max_leverage={lev_now}")

    daily_caps, daily_pnls = [], []
    all_vols = []
    t0 = time.time()

    for i, pf in enumerate(d['parquet_files']):
        ds = pf.stem
        df = pd.read_parquet(pf)
        for c in df.columns:
            if df[c].dtype == 'float64':
                df[c] = df[c].astype('float32')
        acols = [c for c in df.columns if c not in exp_shared.META_COLS]

        if eng.ob_engine is not None:
            eng.ob_engine.preload_date(ds, d['OB_ASSETS'])

        bp = df['BTCUSDT'].values if 'BTCUSDT' in df.columns else None
        dvol = np.zeros(len(df), dtype=np.float32)
        if bp is not None:
            rets = np.diff(bp.astype('float64')) / (bp[:-1].astype('float64') + 1e-9)
            for j in range(50, len(rets)):
                v = np.std(rets[j-50:j])
                dvol[j+1] = v
                if v > 0: all_vols.append(v)

        cap_before = eng.capital

        if use_rolling_vol and len(all_vols) > 1000:
            vp60 = np.percentile(all_vols, 60)
        else:
            vp60 = d['vol_p60']

        vol_ok = np.where(dvol > 0, dvol > vp60, False)
        eng.process_day(ds, df, acols, vol_regime_ok=vol_ok)
        daily_caps.append(eng.capital)
        daily_pnls.append(eng.capital - cap_before)

        if eng.ob_engine is not None:
            eng.ob_engine._preloaded_placement.clear()
            eng.ob_engine._preloaded_signal.clear()
            eng.ob_engine._preloaded_market.clear()
            eng.ob_engine._ts_to_idx.clear()

        del df
        gc.collect()

        if (i+1) % 10 == 0 or i == 0 or i == len(d['parquet_files'])-1:
            elapsed = time.time() - t0
            print(f"  Day {i+1}/{len(d['parquet_files'])}: cap=${eng.capital:,.0f} trades={len(eng.trade_history)} ({elapsed:.0f}s)")

    tr = eng.trade_history
    n = len(tr)
    roi = (eng.capital - 25000.0) / 25000.0 * 100.0

    peak_cap, max_dd = 25000.0, 0.0
    for cap in daily_caps:
        peak_cap = max(peak_cap, cap)
        max_dd = max(max_dd, (peak_cap - cap) / peak_cap * 100.0)

    elapsed = time.time() - t0
    print(f"\n  RESULT: ROI={roi:+.2f}%  T={n}  DD={max_dd:.2f}%  ({elapsed:.0f}s)")
    return {'label': label, 'roi': roi, 'trades': n, 'dd': max_dd}


if __name__ == '__main__':
    results = []

    # Variant B: no hazard call, rolling vol (isolate hazard call effect)
    results.append(run_variant(
        "B: No hazard call, rolling vol",
        use_hazard_call=False,
        use_rolling_vol=True,
    ))

    # Variant C: hazard call, static vol (isolate rolling vol effect)
    results.append(run_variant(
        "C: Hazard call, static vol",
        use_hazard_call=True,
        use_rolling_vol=False,
    ))

    # Variant D: no hazard call, static vol (cleanest comparison to replicate style)
    results.append(run_variant(
        "D: No hazard call, static vol",
        use_hazard_call=False,
        use_rolling_vol=False,
    ))

    print(f"\n{'='*60}")
    print("  ISOLATION SUMMARY")
    print(f"{'='*60}")
    print(f"  {'Config':<45} {'ROI':>8} {'T':>6} {'DD':>7}")
    print(f"  {'-'*70}")
    print(f"  {'A: Hazard call + rolling vol (fork as-is)':<45} {'~+12.83%':>8} {'~1739':>6} {'~26.2%':>7}  [prior run]")
    for r in results:
        print(f"  {r['label']:<45} {r['roi']:>+7.2f}% {r['trades']:>6} {r['dd']:>6.2f}%")
    print(f"  {'replicate style (no hazard, float64, static)':<45} {'~+111.0%':>8} {'~1959':>6} {'~16.9%':>7}  [prior run]")
    print(f"  {'GOLD target':<45} {'+181.81%':>8} {'2155':>6} {'17.65%':>7}")