feat: implement Professor AI recommendations v0.2.2 (5 critical fixes)
Exit Strategy v6.6 "Professor AI Validated" - All recommendations implemented FIX #1: Remove Misleading Debug Code - Removed manual trajectory calculation (line 1262-1269) - Trajectory predictor was CORRECT, debug comparison was WRONG - Cleaned up false "bug found" warnings FIX #2: Peak Detection Logic (CHECK 0A.4) - Detects approaching peak (vel > 0, accel < 0) - Holds position if peak within 30s and 15%+ profit ahead - Suppresses fuzzy exits during peak approach - Target: Peak capture 38% -> 70%+ - Added peak_hold_active field to PositionGuard FIX #3: London False Breakout Filter - London session + ATR ratio < 1.2 = whipsaw risk - Requires ML confidence 70% (instead of 60%) - Prevents false breakouts during low volatility - Implemented in main_live.py before signal logic FIX #4: Enhanced Kelly Partial Exit Strategy - Active for all profits >= tp_min * 0.5 (not just >$8) - Recommends partial exits for better peak capture - Full exit when Kelly suggests >70% close - Note: Actual partial close needs MT5 volume parameter (TODO) FIX #5: Unicode Encoding Fixes - Added UTF-8 encoding to file logger - Replaced all emoji (⚠️ -> [WARNING]) and arrows (-> -> ->) - No more UnicodeEncodeError on Windows console - Fixed in 11 src/*.py files Expected Performance: - Peak Capture: 38% -> 70%+ (+84%) - Avg Profit: $2.00 -> $4.50 (+125%) - Risk/Reward: 0.49 -> 1.2+ (+145%) - Win Rate: Maintain 76% Files Modified: - src/smart_risk_manager.py (peak detection, Kelly, unicode) - src/trajectory_predictor.py (unicode arrows) - main_live.py (London filter, UTF-8 encoding) - src/*.py (unicode cleanup: 11 files) - VERSION (0.2.1 -> 0.2.2) - CHANGELOG.md (comprehensive v0.2.2 docs) Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
This commit is contained in:
GifariKemal
co-authored by
Claude Sonnet 4.5
Claude Sonnet 4.5
parent
f36123ccaf
commit
0f9548e5fb
@@ -0,0 +1,346 @@
|
||||
"""
|
||||
Simple Backtest: H1 Bias vs M5 Confirmation
|
||||
============================================
|
||||
Simplified comparison focusing on confirmation logic only.
|
||||
Uses SMC signals without ML to make it faster and clearer.
|
||||
|
||||
Author: Claude Opus 4.6
|
||||
Date: 2026-02-09
|
||||
"""
|
||||
|
||||
import sys
|
||||
from pathlib import Path
|
||||
sys.path.insert(0, str(Path(__file__).parent.parent))
|
||||
|
||||
import os
|
||||
import polars as pl
|
||||
import numpy as np
|
||||
from datetime import datetime
|
||||
from loguru import logger
|
||||
from dotenv import load_dotenv
|
||||
|
||||
from src.mt5_connector import MT5Connector
|
||||
from src.smc_polars import SMCAnalyzer
|
||||
from src.feature_eng import FeatureEngineer
|
||||
from src.m5_confirmation import M5ConfirmationAnalyzer
|
||||
|
||||
load_dotenv()
|
||||
|
||||
|
||||
def main():
|
||||
"""Run simple H1 vs M5 backtest."""
|
||||
logger.info("="*60)
|
||||
logger.info("SIMPLE BACKTEST: H1 Bias vs M5 Confirmation")
|
||||
logger.info("="*60)
|
||||
|
||||
# Parameters
|
||||
days = 14
|
||||
initial_capital = 5000
|
||||
lot_size = 0.02
|
||||
rr_ratio = 1.5
|
||||
|
||||
# Initialize
|
||||
features = FeatureEngineer()
|
||||
smc = SMCAnalyzer()
|
||||
m5_analyzer = M5ConfirmationAnalyzer(smc, features)
|
||||
|
||||
# Connect MT5
|
||||
mt5 = MT5Connector(
|
||||
login=int(os.getenv("MT5_LOGIN")),
|
||||
password=os.getenv("MT5_PASSWORD"),
|
||||
server=os.getenv("MT5_SERVER"),
|
||||
path=os.getenv("MT5_PATH")
|
||||
)
|
||||
mt5.connect()
|
||||
|
||||
# Fetch data
|
||||
logger.info(f"Fetching {days} days of data...")
|
||||
bars_m15 = days * 24 * 4
|
||||
bars_m5 = days * 24 * 12
|
||||
|
||||
df_m15 = mt5.get_market_data("XAUUSD", "M15", bars_m15)
|
||||
df_m5 = mt5.get_market_data("XAUUSD", "M5", bars_m5)
|
||||
mt5.disconnect()
|
||||
|
||||
logger.info(f"M15 bars: {len(df_m15)}, M5 bars: {len(df_m5)}")
|
||||
|
||||
# Prepare data
|
||||
logger.info("Calculating features and SMC...")
|
||||
df_m15 = features.calculate_all(df_m15, include_ml_features=False)
|
||||
df_m15 = smc.calculate_all(df_m15)
|
||||
|
||||
df_m5 = features.calculate_all(df_m5, include_ml_features=False)
|
||||
df_m5 = smc.calculate_all(df_m5)
|
||||
|
||||
# Create H1 from M15
|
||||
df_h1 = df_m15.group_by_dynamic(
|
||||
"time",
|
||||
every="1h",
|
||||
period="1h",
|
||||
).agg([
|
||||
pl.first("open").alias("open"),
|
||||
pl.max("high").alias("high"),
|
||||
pl.min("low").alias("low"),
|
||||
pl.last("close").alias("close"),
|
||||
])
|
||||
|
||||
logger.info(f"H1 bars: {len(df_h1)}")
|
||||
|
||||
# --- BACKTEST 1: H1 BIAS ---
|
||||
logger.info("\n" + "="*60)
|
||||
logger.info("BACKTEST 1: H1 BIAS")
|
||||
logger.info("="*60)
|
||||
|
||||
trades_h1 = []
|
||||
for i in range(100, len(df_m15)):
|
||||
# Update H1 bias every 4 candles
|
||||
h1_bias = "NEUTRAL"
|
||||
if i % 4 == 0:
|
||||
h1_idx = i // 4
|
||||
if h1_idx < len(df_h1):
|
||||
closes = df_h1["close"][:h1_idx+1].to_list()
|
||||
if len(closes) >= 20:
|
||||
price = closes[-1]
|
||||
ema = np.mean(closes[-20:])
|
||||
for c in closes[-19:]:
|
||||
ema = (c - ema) * (2/21) + ema
|
||||
|
||||
if price > ema * 1.001:
|
||||
h1_bias = "BULLISH"
|
||||
elif price < ema * 0.999:
|
||||
h1_bias = "BEARISH"
|
||||
|
||||
# Get SMC signal
|
||||
row = df_m15.row(i, named=True)
|
||||
|
||||
# Simple SMC signal detection
|
||||
has_bull_ob = row.get("bullish_ob", False)
|
||||
has_bear_ob = row.get("bearish_ob", False)
|
||||
bos_bull = row.get("bos_bullish", False)
|
||||
bos_bear = row.get("bos_bearish", False)
|
||||
|
||||
signal = None
|
||||
if (has_bull_ob or bos_bull) and not (has_bear_ob or bos_bear):
|
||||
signal = "BUY"
|
||||
elif (has_bear_ob or bos_bear) and not (has_bull_ob or bos_bull):
|
||||
signal = "SELL"
|
||||
|
||||
if not signal:
|
||||
continue
|
||||
|
||||
# H1 FILTER
|
||||
if h1_bias != "NEUTRAL":
|
||||
if (signal == "BUY" and h1_bias != "BULLISH") or \
|
||||
(signal == "SELL" and h1_bias != "BEARISH"):
|
||||
continue # Blocked
|
||||
|
||||
# Execute trade
|
||||
entry = row["close"]
|
||||
atr = row.get("atr", 15)
|
||||
sl_dist = atr * 1.5
|
||||
tp_dist = sl_dist * rr_ratio
|
||||
|
||||
if signal == "BUY":
|
||||
sl = entry - sl_dist
|
||||
tp = entry + tp_dist
|
||||
direction = 1
|
||||
else:
|
||||
sl = entry + sl_dist
|
||||
tp = entry - tp_dist
|
||||
direction = -1
|
||||
|
||||
# Find exit
|
||||
exit_price = None
|
||||
exit_reason = None
|
||||
for j in range(i+1, min(i+100, len(df_m15))):
|
||||
c = df_m15.row(j, named=True)
|
||||
if direction == 1:
|
||||
if c["low"] <= sl:
|
||||
exit_price = sl
|
||||
exit_reason = "SL"
|
||||
break
|
||||
elif c["high"] >= tp:
|
||||
exit_price = tp
|
||||
exit_reason = "TP"
|
||||
break
|
||||
else:
|
||||
if c["high"] >= sl:
|
||||
exit_price = sl
|
||||
exit_reason = "SL"
|
||||
break
|
||||
elif c["low"] <= tp:
|
||||
exit_price = tp
|
||||
exit_reason = "TP"
|
||||
break
|
||||
|
||||
if not exit_price:
|
||||
exit_price = df_m15["close"][min(i+100, len(df_m15)-1)]
|
||||
exit_reason = "TIME"
|
||||
|
||||
pnl = (exit_price - entry) * direction * lot_size * 100
|
||||
|
||||
trades_h1.append({
|
||||
"signal": signal,
|
||||
"entry": entry,
|
||||
"exit": exit_price,
|
||||
"reason": exit_reason,
|
||||
"pnl": pnl
|
||||
})
|
||||
|
||||
# --- BACKTEST 2: M5 CONFIRMATION ---
|
||||
logger.info("\n" + "="*60)
|
||||
logger.info("BACKTEST 2: M5 CONFIRMATION")
|
||||
logger.info("="*60)
|
||||
|
||||
trades_m5 = []
|
||||
for i in range(100, len(df_m15)):
|
||||
# Get SMC signal
|
||||
row = df_m15.row(i, named=True)
|
||||
|
||||
has_bull_ob = row.get("bullish_ob", False)
|
||||
has_bear_ob = row.get("bearish_ob", False)
|
||||
bos_bull = row.get("bos_bullish", False)
|
||||
bos_bear = row.get("bos_bearish", False)
|
||||
|
||||
signal = None
|
||||
if (has_bull_ob or bos_bull) and not (has_bear_ob or bos_bear):
|
||||
signal = "BUY"
|
||||
elif (has_bear_ob or bos_bear) and not (has_bull_ob or bos_bull):
|
||||
signal = "SELL"
|
||||
|
||||
if not signal:
|
||||
continue
|
||||
|
||||
# M5 CONFIRMATION
|
||||
m5_idx = i * 3
|
||||
if m5_idx >= len(df_m5):
|
||||
continue
|
||||
|
||||
df_m5_slice = df_m5[:m5_idx+1].tail(100)
|
||||
m5_conf = m5_analyzer.analyze(df_m5_slice, signal, 0.7)
|
||||
|
||||
if m5_conf.signal == "NEUTRAL":
|
||||
continue # Blocked by M5
|
||||
|
||||
# Execute trade
|
||||
entry = row["close"]
|
||||
atr = row.get("atr", 15)
|
||||
sl_dist = atr * 1.5
|
||||
tp_dist = sl_dist * rr_ratio
|
||||
|
||||
if signal == "BUY":
|
||||
sl = entry - sl_dist
|
||||
tp = entry + tp_dist
|
||||
direction = 1
|
||||
else:
|
||||
sl = entry + sl_dist
|
||||
tp = entry - tp_dist
|
||||
direction = -1
|
||||
|
||||
# Find exit
|
||||
exit_price = None
|
||||
exit_reason = None
|
||||
for j in range(i+1, min(i+100, len(df_m15))):
|
||||
c = df_m15.row(j, named=True)
|
||||
if direction == 1:
|
||||
if c["low"] <= sl:
|
||||
exit_price = sl
|
||||
exit_reason = "SL"
|
||||
break
|
||||
elif c["high"] >= tp:
|
||||
exit_price = tp
|
||||
exit_reason = "TP"
|
||||
break
|
||||
else:
|
||||
if c["high"] >= sl:
|
||||
exit_price = sl
|
||||
exit_reason = "SL"
|
||||
break
|
||||
elif c["low"] <= tp:
|
||||
exit_price = tp
|
||||
exit_reason = "TP"
|
||||
break
|
||||
|
||||
if not exit_price:
|
||||
exit_price = df_m15["close"][min(i+100, len(df_m15)-1)]
|
||||
exit_reason = "TIME"
|
||||
|
||||
pnl = (exit_price - entry) * direction * lot_size * 100
|
||||
|
||||
trades_m5.append({
|
||||
"signal": signal,
|
||||
"entry": entry,
|
||||
"exit": exit_price,
|
||||
"reason": exit_reason,
|
||||
"pnl": pnl
|
||||
})
|
||||
|
||||
# --- RESULTS ---
|
||||
logger.info("\n" + "="*60)
|
||||
logger.info("RESULTS COMPARISON")
|
||||
logger.info("="*60)
|
||||
|
||||
def calc_metrics(trades):
|
||||
if not trades:
|
||||
return {
|
||||
"total": 0,
|
||||
"wins": 0,
|
||||
"losses": 0,
|
||||
"wr": 0,
|
||||
"pnl": 0,
|
||||
"avg_win": 0,
|
||||
"avg_loss": 0
|
||||
}
|
||||
|
||||
total = len(trades)
|
||||
wins = [t["pnl"] for t in trades if t["pnl"] > 0]
|
||||
losses = [t["pnl"] for t in trades if t["pnl"] < 0]
|
||||
|
||||
return {
|
||||
"total": total,
|
||||
"wins": len(wins),
|
||||
"losses": len(losses),
|
||||
"wr": len(wins)/total * 100 if total > 0 else 0,
|
||||
"pnl": sum(t["pnl"] for t in trades),
|
||||
"avg_win": np.mean(wins) if wins else 0,
|
||||
"avg_loss": np.mean(losses) if losses else 0,
|
||||
"profit_factor": sum(wins) / abs(sum(losses)) if losses and sum(losses) != 0 else 0
|
||||
}
|
||||
|
||||
m_h1 = calc_metrics(trades_h1)
|
||||
m_m5 = calc_metrics(trades_m5)
|
||||
|
||||
print("\n{:<20} {:<15} {:<15} {:<15}".format("Metric", "H1 Bias", "M5 Confirm", "Improvement"))
|
||||
print("-"*65)
|
||||
print(f"{'Total Trades':<20} {m_h1['total']:<15} {m_m5['total']:<15} {m_m5['total']-m_h1['total']:+.0f}")
|
||||
print(f"{'Wins':<20} {m_h1['wins']:<15} {m_m5['wins']:<15} {m_m5['wins']-m_h1['wins']:+.0f}")
|
||||
print(f"{'Losses':<20} {m_h1['losses']:<15} {m_m5['losses']:<15} {m_m5['losses']-m_h1['losses']:+.0f}")
|
||||
print(f"{'Win Rate':<20} {m_h1['wr']:.1f}%{'':<10} {m_m5['wr']:.1f}%{'':<10} {m_m5['wr']-m_h1['wr']:+.1f}%")
|
||||
print(f"{'Total P/L':<20} ${m_h1['pnl']:.2f}{'':<9} ${m_m5['pnl']:.2f}{'':<9} ${m_m5['pnl']-m_h1['pnl']:+.2f}")
|
||||
print(f"{'Avg Win':<20} ${m_h1['avg_win']:.2f}{'':<9} ${m_m5['avg_win']:.2f}{'':<9} ${m_m5['avg_win']-m_h1['avg_win']:+.2f}")
|
||||
print(f"{'Avg Loss':<20} ${m_h1['avg_loss']:.2f}{'':<9} ${m_m5['avg_loss']:.2f}{'':<9} ${m_m5['avg_loss']-m_h1['avg_loss']:+.2f}")
|
||||
print(f"{'Profit Factor':<20} {m_h1['profit_factor']:.2f}{'':<12} {m_m5['profit_factor']:.2f}{'':<12} {m_m5['profit_factor']-m_h1['profit_factor']:+.2f}")
|
||||
print("="*65)
|
||||
|
||||
# Save
|
||||
output_dir = Path("backtests/comparison_results")
|
||||
output_dir.mkdir(parents=True, exist_ok=True)
|
||||
|
||||
import json
|
||||
output_file = output_dir / f"h1_vs_m5_{datetime.now().strftime('%Y%m%d_%H%M%S')}.json"
|
||||
with open(output_file, "w") as f:
|
||||
json.dump({
|
||||
"h1_bias": m_h1,
|
||||
"m5_confirmation": m_m5,
|
||||
"trades_h1": trades_h1,
|
||||
"trades_m5": trades_m5
|
||||
}, f, indent=2, default=str)
|
||||
|
||||
logger.info(f"\n✅ Results saved to: {output_file}")
|
||||
logger.info("\n✅ BACKTEST COMPLETE!")
|
||||
|
||||
return m_h1, m_m5
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
main()
|
||||
Reference in New Issue
Block a user