mirror of
https://github.com/chrisnov-it/quantumbotx.git
synced 2026-07-27 18:57:47 +00:00
Reynov Christian
273 lines
11 KiB
Python
273 lines
11 KiB
Python
#!/usr/bin/env python3
|
|
"""
|
|
Diagnostic script to identify the root cause of backtesting issues
|
|
Focuses on the specific problem: 100% max drawdown and terrible performance across all strategies
|
|
"""
|
|
|
|
import sys
|
|
import os
|
|
sys.path.append(os.path.dirname(os.path.abspath(__file__)))
|
|
|
|
import pandas as pd
|
|
import yfinance as yf
|
|
from core.backtesting.engine import run_backtest as run_original_backtest
|
|
from core.backtesting.enhanced_engine import run_enhanced_backtest
|
|
from core.strategies.bollinger_squeeze import BollingerSqueezeStrategy
|
|
from datetime import datetime, timedelta
|
|
|
|
def download_test_data():
|
|
"""Download recent EURUSD data for testing"""
|
|
try:
|
|
# Download EURUSD data for the last 6 months
|
|
end_date = datetime.now()
|
|
start_date = end_date - timedelta(days=180)
|
|
|
|
# Download EURUSD data
|
|
ticker = yf.Ticker("EURUSD=X")
|
|
df = ticker.history(start=start_date, end=end_date, interval="1h")
|
|
|
|
if df.empty:
|
|
print("❌ Failed to download EURUSD data")
|
|
return None
|
|
|
|
# Standardize column names
|
|
df.columns = df.columns.str.lower()
|
|
df.reset_index(inplace=True)
|
|
|
|
print(f"✅ Downloaded {len(df)} EURUSD bars from {df['datetime'].min()} to {df['datetime'].max()}")
|
|
return df
|
|
|
|
except Exception as e:
|
|
print(f"❌ Error downloading data: {e}")
|
|
return None
|
|
|
|
def test_strategy_signals(df, strategy_class, params):
|
|
"""Test if strategy is generating reasonable signals"""
|
|
print(f"\n🔍 Testing {strategy_class.name} signal generation...")
|
|
|
|
# Create mock bot
|
|
class MockBot:
|
|
def __init__(self):
|
|
self.market_for_mt5 = "EURUSD"
|
|
self.timeframe = "H1"
|
|
self.tf_map = {}
|
|
|
|
try:
|
|
# Initialize strategy
|
|
strategy_instance = strategy_class(bot_instance=MockBot(), params=params)
|
|
|
|
# Analyze data
|
|
df_with_signals = strategy_instance.analyze_df(df.copy())
|
|
|
|
# Count signals
|
|
signal_counts = df_with_signals['signal'].value_counts()
|
|
print(f"Signal distribution: {dict(signal_counts)}")
|
|
|
|
# Check if we have reasonable signals
|
|
buy_signals = len(df_with_signals[df_with_signals['signal'] == 'BUY'])
|
|
sell_signals = len(df_with_signals[df_with_signals['signal'] == 'SELL'])
|
|
hold_signals = len(df_with_signals[df_with_signals['signal'] == 'HOLD'])
|
|
|
|
total_bars = len(df_with_signals)
|
|
|
|
print(f"BUY signals: {buy_signals} ({buy_signals/total_bars*100:.1f}%)")
|
|
print(f"SELL signals: {sell_signals} ({sell_signals/total_bars*100:.1f}%)")
|
|
print(f"HOLD signals: {hold_signals} ({hold_signals/total_bars*100:.1f}%)")
|
|
|
|
# Check for indicators
|
|
print(f"Available columns: {list(df_with_signals.columns)}")
|
|
|
|
# Check ATR
|
|
if 'ATRr_14' in df_with_signals.columns:
|
|
atr_stats = df_with_signals['ATRr_14'].describe()
|
|
print(f"ATR stats: min={atr_stats['min']:.6f}, max={atr_stats['max']:.6f}, mean={atr_stats['mean']:.6f}")
|
|
else:
|
|
print("❌ ATR indicator missing!")
|
|
|
|
return df_with_signals
|
|
|
|
except Exception as e:
|
|
print(f"❌ Error in strategy analysis: {e}")
|
|
return None
|
|
|
|
def test_backtest_calculations(df_with_signals, params):
|
|
"""Test the actual backtest calculations step by step"""
|
|
print(f"\n🧮 Testing backtest calculations...")
|
|
|
|
# Parameters for testing
|
|
risk_percent = float(params.get('lot_size', 1.0))
|
|
sl_atr_multiplier = float(params.get('sl_pips', 2.0))
|
|
tp_atr_multiplier = float(params.get('tp_pips', 4.0))
|
|
|
|
print(f"Risk: {risk_percent}%, SL: {sl_atr_multiplier}x ATR, TP: {tp_atr_multiplier}x ATR")
|
|
|
|
# Simulate a few trades manually
|
|
trades_found = 0
|
|
capital = 10000.0
|
|
initial_capital = capital
|
|
|
|
for i in range(1, min(len(df_with_signals), 100)): # Check first 100 bars
|
|
current_bar = df_with_signals.iloc[i]
|
|
signal = current_bar.get("signal", "HOLD")
|
|
|
|
if signal in ['BUY', 'SELL']:
|
|
trades_found += 1
|
|
entry_price = current_bar['close']
|
|
atr_value = current_bar.get('ATRr_14', 0)
|
|
|
|
print(f"\nTrade #{trades_found} at bar {i}:")
|
|
print(f" Signal: {signal}")
|
|
print(f" Entry price: {entry_price}")
|
|
print(f" ATR: {atr_value}")
|
|
print(f" Capital: ${capital:.2f}")
|
|
|
|
if atr_value > 0:
|
|
# Calculate position size
|
|
sl_distance = atr_value * sl_atr_multiplier
|
|
tp_distance = atr_value * tp_atr_multiplier
|
|
|
|
# For EURUSD (forex major)
|
|
contract_size = 100000
|
|
amount_to_risk = capital * (risk_percent / 100.0)
|
|
risk_in_currency_per_lot = sl_distance * contract_size
|
|
|
|
if risk_in_currency_per_lot > 0:
|
|
calculated_lot_size = amount_to_risk / risk_in_currency_per_lot
|
|
lot_size = max(0.01, min(calculated_lot_size, 10.0))
|
|
else:
|
|
lot_size = 0.01
|
|
|
|
print(f" SL distance: {sl_distance:.5f}")
|
|
print(f" TP distance: {tp_distance:.5f}")
|
|
print(f" Amount to risk: ${amount_to_risk:.2f}")
|
|
print(f" Risk per lot: ${risk_in_currency_per_lot:.2f}")
|
|
print(f" Calculated lot size: {calculated_lot_size:.4f}")
|
|
print(f" Final lot size: {lot_size:.2f}")
|
|
|
|
# Set SL/TP levels
|
|
if signal == 'BUY':
|
|
sl_price = entry_price - sl_distance
|
|
tp_price = entry_price + tp_distance
|
|
else:
|
|
sl_price = entry_price + sl_distance
|
|
tp_price = entry_price - tp_distance
|
|
|
|
print(f" SL price: {sl_price:.5f}")
|
|
print(f" TP price: {tp_price:.5f}")
|
|
|
|
# Check if prices are reasonable
|
|
if sl_price <= 0 or tp_price <= 0:
|
|
print(" ❌ Invalid SL/TP prices!")
|
|
elif abs((sl_price - entry_price) / entry_price) > 0.1:
|
|
print(" ❌ SL distance too large (>10%)!")
|
|
elif abs((tp_price - entry_price) / entry_price) > 0.2:
|
|
print(" ❌ TP distance too large (>20%)!")
|
|
else:
|
|
print(" ✅ Trade setup looks reasonable")
|
|
else:
|
|
print(" ❌ Invalid ATR value!")
|
|
|
|
if trades_found >= 5: # Limit to first 5 trades
|
|
break
|
|
|
|
print(f"\nFound {trades_found} trades in first 100 bars")
|
|
return trades_found > 0
|
|
|
|
def run_diagnostic():
|
|
"""Main diagnostic function"""
|
|
print("🚀 Starting Backtesting Diagnostic...")
|
|
print("=" * 60)
|
|
|
|
# 1. Download test data
|
|
df = download_test_data()
|
|
if df is None:
|
|
return
|
|
|
|
# 2. Test strategy parameters
|
|
params = {
|
|
'bb_length': 20,
|
|
'bb_std': 2.0,
|
|
'squeeze_window': 10,
|
|
'squeeze_factor': 0.7,
|
|
'rsi_period': 14,
|
|
'lot_size': 1.0, # 1% risk
|
|
'sl_pips': 2.0, # 2x ATR
|
|
'tp_pips': 4.0 # 4x ATR
|
|
}
|
|
|
|
print(f"Test parameters: {params}")
|
|
|
|
# 3. Test Bollinger Squeeze strategy signals
|
|
df_with_signals = test_strategy_signals(df, BollingerSqueezeStrategy, params)
|
|
if df_with_signals is None:
|
|
return
|
|
|
|
# 4. Test manual calculations
|
|
if not test_backtest_calculations(df_with_signals, params):
|
|
print("❌ No valid trades found for manual testing")
|
|
return
|
|
|
|
# 5. Run original backtest
|
|
print(f"\n🔄 Running original backtest...")
|
|
try:
|
|
original_result = run_original_backtest('bollinger_squeeze', params, df, 'EURUSD')
|
|
print(f"Original engine result:")
|
|
print(f" Total trades: {original_result.get('total_trades', 0)}")
|
|
print(f" Total profit: ${original_result.get('total_profit_usd', 0):.2f}")
|
|
print(f" Win rate: {original_result.get('win_rate_percent', 0):.1f}%")
|
|
print(f" Max drawdown: {original_result.get('max_drawdown_percent', 0):.1f}%")
|
|
except Exception as e:
|
|
print(f"❌ Original backtest failed: {e}")
|
|
original_result = None
|
|
|
|
# 6. Run enhanced backtest
|
|
print(f"\n🚀 Running enhanced backtest...")
|
|
try:
|
|
enhanced_result = run_enhanced_backtest('bollinger_squeeze', params, df, 'EURUSD')
|
|
print(f"Enhanced engine result:")
|
|
print(f" Total trades: {enhanced_result.get('total_trades', 0)}")
|
|
print(f" Gross profit: ${enhanced_result.get('total_profit_usd', 0):.2f}")
|
|
print(f" Spread costs: ${enhanced_result.get('total_spread_costs', 0):.2f}")
|
|
print(f" Net profit: ${enhanced_result.get('net_profit_after_costs', 0):.2f}")
|
|
print(f" Win rate: {enhanced_result.get('win_rate_percent', 0):.1f}%")
|
|
print(f" Max drawdown: {enhanced_result.get('max_drawdown_percent', 0):.1f}%")
|
|
|
|
# Check individual trades
|
|
if enhanced_result.get('trades'):
|
|
print(f"\nLast few trades:")
|
|
for i, trade in enumerate(enhanced_result['trades'][-3:]):
|
|
print(f" Trade {i+1}: {trade['position_type']} | Entry: {trade['entry']:.5f} | Exit: {trade['exit']:.5f} | Profit: ${trade['profit']:.2f}")
|
|
|
|
except Exception as e:
|
|
print(f"❌ Enhanced backtest failed: {e}")
|
|
enhanced_result = None
|
|
|
|
# 7. Analysis and recommendations
|
|
print(f"\n📊 DIAGNOSIS SUMMARY:")
|
|
print("=" * 60)
|
|
|
|
if original_result and enhanced_result:
|
|
if enhanced_result.get('max_drawdown_percent', 0) > 90:
|
|
print("❌ CRITICAL ISSUE: Enhanced engine shows extreme drawdown!")
|
|
print(" Possible causes:")
|
|
print(" - Position sizing too aggressive")
|
|
print(" - Spread costs too high")
|
|
print(" - SL/TP calculation errors")
|
|
print(" - Strategy generating bad signals")
|
|
elif original_result.get('max_drawdown_percent', 0) > 90:
|
|
print("❌ CRITICAL ISSUE: Original engine shows extreme drawdown!")
|
|
print(" Possible causes:")
|
|
print(" - Position sizing calculation error")
|
|
print(" - SL/TP logic bug")
|
|
print(" - Strategy overfitting")
|
|
else:
|
|
print("✅ Backtest engines working reasonably")
|
|
|
|
print("\nRecommendations:")
|
|
print("1. Check position sizing calculations")
|
|
print("2. Verify SL/TP distance calculations")
|
|
print("3. Test with more conservative parameters")
|
|
print("4. Validate strategy signal quality")
|
|
|
|
if __name__ == '__main__':
|
|
run_diagnostic() |