#!/usr/bin/env python3 """Backtest FINAL - Stratégie Probabiliste Optimisée BNB 1m""" import pandas as pd import numpy as np import requests import json from datetime import datetime from final_probabilistic_strategy import FinalProbabilisticStrategy def get_bnb_data(): url = "https://api.binance.com/api/v3/klines" params = {'symbol': 'BNBUSDT', 'interval': '1m', 'limit': 1000} response = requests.get(url, params=params, timeout=10) data = response.json() df = pd.DataFrame(data, columns=[ 'timestamp', 'open', 'high', 'low', 'close', 'volume', 'close_time', 'quote_volume', 'trades', 'taker_buy_base', 'taker_buy_quote', 'ignore' ]) df['timestamp'] = pd.to_datetime(df['timestamp'], unit='ms') for col in ['open', 'high', 'low', 'close', 'volume']: df[col] = df[col].astype(float) return df def backtest(df, strategy, name): capital = 100.0 trades = [] open_position = None for i in range(strategy.lookback_window + 50, len(df)): window = df.iloc[:i+1] current_price = window['close'].iloc[-1] if open_position: entry = open_position['entry_price'] leverage = open_position['leverage'] if open_position['side'] == 'LONG': pnl_pct = ((current_price - entry) / entry) * 100 * leverage else: pnl_pct = ((entry - current_price) / entry) * 100 * leverage # Limite -100% pnl_pct = max(pnl_pct, -100) # Check liquidation if pnl_pct <= -100: pnl_usd = capital * -1.0 capital += pnl_usd trades.append({ 'pnl_pct': -100, 'pnl_usd': pnl_usd, 'exit_reason': 'LIQUIDATION', 'probability': open_position['probability'], 'trend_context': open_position['trend_context'] }) open_position = None continue # Check SL/TP hit_sl = ( (open_position['side'] == 'LONG' and current_price <= open_position['stop_loss']) or (open_position['side'] == 'SHORT' and current_price >= open_position['stop_loss']) ) hit_tp = ( (open_position['side'] == 'LONG' and current_price >= open_position['take_profit']) or (open_position['side'] == 'SHORT' and current_price <= open_position['take_profit']) ) if hit_sl or hit_tp: pnl_usd = capital * (pnl_pct / 100) capital += pnl_usd trades.append({ 'pnl_pct': pnl_pct, 'pnl_usd': pnl_usd, 'exit_reason': 'TP' if hit_tp else 'SL', 'probability': open_position['probability'], 'trend_context': open_position['trend_context'] }) open_position = None if not open_position and capital > 0: signal = strategy.analyze('BNB', window) if signal: open_position = { 'side': signal.side, 'entry_price': signal.entry_price, 'stop_loss': signal.stop_loss, 'take_profit': signal.take_profit, 'leverage': signal.leverage, 'probability': signal.probability, 'trend_context': signal.trend_context } if not trades: return None winners = [t for t in trades if t['pnl_pct'] > 0] losers = [t for t in trades if t['pnl_pct'] <= 0] total_win = sum(t['pnl_usd'] for t in winners) total_loss = abs(sum(t['pnl_usd'] for t in losers)) pf = (total_win / total_loss) if total_loss > 0 else float('inf') pnl_pct = ((capital - 100) / 100) * 100 # Calibration error avg_prob = np.mean([t['probability'] for t in trades]) actual_wr = (len(winners) / len(trades)) * 100 calib_err = abs(avg_prob * 100 - actual_wr) return { 'name': name, 'trades': len(trades), 'winners': len(winners), 'losers': len(losers), 'win_rate': actual_wr, 'profit_factor': pf, 'pnl_pct': pnl_pct, 'final_capital': capital, 'calib_error': calib_err } def main(): print("="*80) print("🎯 BACKTEST FINAL - STRATÉGIE PROBABILISTE OPTIMISÉE") print("="*80) df = get_bnb_data() print(f"\n📊 Données: {len(df)} bougies BNB 1m") print(f" Période: {df['timestamp'].iloc[0]} → {df['timestamp'].iloc[-1]}") results = [] # Test 1: Standard print("\n📊 TEST 1: Standard (Lev 100x)") s1 = FinalProbabilisticStrategy(leverage=100, z_threshold=1.8, min_probability=0.50) r1 = backtest(df, s1, "Standard 100x") if r1: results.append(r1) print(f" Trades:{r1['trades']} | WR:{r1['win_rate']:.1f}% | PF:{r1['profit_factor']:.2f} | PnL:{r1['pnl_pct']:.1f}%") # Test 2: Conservateur print("\n📊 TEST 2: Conservateur (Lev 50x)") s2 = FinalProbabilisticStrategy(leverage=50, sl_pct=0.06, tp_pct=0.12, z_threshold=2.0, min_probability=0.52) r2 = backtest(df, s2, "Conservateur 50x") if r2: results.append(r2) print(f" Trades:{r2['trades']} | WR:{r2['win_rate']:.1f}% | PF:{r2['profit_factor']:.2f} | PnL:{r2['pnl_pct']:.1f}%") # Test 3: Agressif print("\n📊 TEST 3: Agressif (Lev 150x)") s3 = FinalProbabilisticStrategy(leverage=150, z_threshold=1.7, min_probability=0.48, min_ev=0.12) r3 = backtest(df, s3, "Agressif 150x") if r3: results.append(r3) print(f" Trades:{r3['trades']} | WR:{r3['win_rate']:.1f}% | PF:{r3['profit_factor']:.2f} | PnL:{r3['pnl_pct']:.1f}%") # Test 4: Sans filtre tendance print("\n📊 TEST 4: Sans Filtre Tendance (Lev 100x)") s4 = FinalProbabilisticStrategy(leverage=100, z_threshold=1.8, trend_filter_enabled=False) r4 = backtest(df, s4, "Sans Filtre 100x") if r4: results.append(r4) print(f" Trades:{r4['trades']} | WR:{r4['win_rate']:.1f}% | PF:{r4['profit_factor']:.2f} | PnL:{r4['pnl_pct']:.1f}%") # Comparaison if results: print("\n" + "="*100) print("🏆 COMPARAISON") print("="*100) print(f"\n{'Stratégie':<20} | {'Trades':<7} | {'WR%':<7} | {'PF':<7} | {'PnL%':<10} | {'CalibErr%':<10}") print("-"*100) best_pnl = -float('inf') best_name = "" for r in results: print(f"{r['name']:<20} | {r['trades']:<7} | {r['win_rate']:<7.1f} | {r['profit_factor']:<7.2f} | {r['pnl_pct']:<10.1f} | {r['calib_error']:<10.1f}") if r['pnl_pct'] > best_pnl: best_pnl = r['pnl_pct'] best_name = r['name'] print("-"*100) print(f"\n🏆 MEILLEUR: {best_name} ({best_pnl:.1f}%)") # Sauvegarder output = { 'asset': 'BNBUSDT', 'interval': '1m', 'date': datetime.now().isoformat(), 'results': results, 'best': best_name } with open('final_probabilistic_results.json', 'w') as f: json.dump(output, f, indent=2) print("\n💾 Résultats: final_probabilistic_results.json") print("\n✅ BACKTEST FINAL TERMINÉ\n") if __name__ == '__main__': main()