import sys, os
import numpy as np, pandas as pd
from scipy.stats import spearmanr
import warnings; warnings.filterwarnings('ignore')

sys.path.insert(0, os.path.abspath(os.path.join(os.path.dirname(__file__), '..', '..')))
from backtesting.framework.config import STRATEGY_NAME, ACTIVE_SIGNAL_FN, ACTIVE_PARAMS, load_data

def run_test_1_1():
    print("=" * 80)
    print(f" TEST 1.1: INFORMATION COEFFICIENT (IC) ANALYSIS - {STRATEGY_NAME}")
    print("=" * 80)
    
    if ACTIVE_SIGNAL_FN is None:
        print("FAIL: No ACTIVE_SIGNAL_FN defined in config.py")
        return

    dc, spy, vf, daily_ret = load_data()
    
    raw_signal = ACTIVE_SIGNAL_FN(dc, spy, vf)
    fwd = dc[vf].pct_change(60).shift(-60)
    top_n = ACTIVE_PARAMS.get('top_n', 15)
    
    ic_vals = []
    dates = []
    
    # Calculate IC every 60 days
    for i in range(200, len(dc)-60, 60):
        sig = raw_signal.iloc[i].dropna()
        common = sig.index.intersection(fwd.iloc[i].dropna().index)
        if len(common) >= top_n:
            top_stocks = sig[common].nlargest(top_n)
            corr, _ = spearmanr(top_stocks.values, fwd.iloc[i][top_stocks.index].values)
            if not np.isnan(corr):
                ic_vals.append(corr)
                dates.append(dc.index[i])
                
    if not ic_vals:
        print("FAIL: Could not calculate IC (Not enough overlap between signal and forward returns).")
        return
        
    ic_series = pd.Series(ic_vals, index=dates)
    ic_mean = ic_series.mean()
    ic_std = ic_series.std()
    n = len(ic_series)
    t_stat = ic_mean / (ic_std / np.sqrt(n)) if ic_std > 0 else 0
    
    print(f"  N observations:  {n}")
    print(f"  IC Mean (Top {top_n}):  {ic_mean:.4f}")
    print(f"  IC Std Dev:      {ic_std:.4f}")
    print(f"  t-statistic:     {t_stat:.2f}")
    print("-" * 80)
    if t_stat > 2.0:
        print(f"  VERDICT: PASS (Statistically significant edge)")
    else:
        print(f"  VERDICT: FAIL (t-stat < 2.0)")

if __name__ == "__main__":
    run_test_1_1()