| """
|
| Convert 1-minute data to higher timeframes (5m, 15m, 1h, 4h, daily).
|
|
|
| This allows testing the same strategy on different timeframes to find
|
| the optimal trading interval.
|
| """
|
|
|
| import pandas as pd
|
| import os
|
| import sys
|
|
|
| sys.path.append(os.path.dirname(os.path.abspath(__file__)))
|
| import config
|
|
|
|
|
| def resample_to_timeframe(df, timeframe):
|
| """
|
| Resample 1-minute OHLCV data to a higher timeframe.
|
|
|
| Parameters
|
| ----------
|
| df : pd.DataFrame
|
| 1-minute OHLCV data with DatetimeIndex
|
| timeframe : str
|
| Pandas resample frequency: '5min', '15min', '1H', '4H', '1D'
|
|
|
| Returns
|
| -------
|
| pd.DataFrame
|
| Resampled OHLCV data
|
| """
|
|
|
| if not isinstance(df.index, pd.DatetimeIndex):
|
| df = df.set_index('timestamp')
|
|
|
|
|
| resampled = df.resample(timeframe).agg({
|
| 'open': 'first',
|
| 'high': 'max',
|
| 'low': 'min',
|
| 'close': 'last',
|
| 'volume': 'sum'
|
| }).dropna()
|
|
|
| return resampled
|
|
|
|
|
| def create_all_timeframes(source_file):
|
| """
|
| Create datasets for all timeframes from 1-minute source data.
|
|
|
| Parameters
|
| ----------
|
| source_file : str
|
| Path to 1-minute CSV file
|
|
|
| Returns
|
| -------
|
| dict
|
| Dictionary of {timeframe: output_file_path}
|
| """
|
| print("\n" + "="*70)
|
| print("CREATING MULTI-TIMEFRAME DATASETS")
|
| print("="*70)
|
|
|
|
|
| print(f"\n📥 Loading source data: {source_file}")
|
| df = pd.read_csv(source_file)
|
| df['timestamp'] = pd.to_datetime(df['timestamp'])
|
| df = df.set_index('timestamp')
|
|
|
| print(f" Loaded {len(df):,} 1-minute candles")
|
| print(f" Period: {df.index[0]} to {df.index[-1]}")
|
|
|
|
|
| timeframes = {
|
| '5min': '5-minute',
|
| '15min': '15-minute',
|
| '30min': '30-minute',
|
| '1H': '1-hour',
|
| '4H': '4-hour',
|
| '1D': 'daily'
|
| }
|
|
|
| output_files = {}
|
|
|
| for tf_code, tf_name in timeframes.items():
|
| print(f"\n🔄 Creating {tf_name} dataset...")
|
|
|
|
|
| resampled = resample_to_timeframe(df, tf_code)
|
|
|
|
|
| output_file = os.path.join(
|
| config.DATA_DIR,
|
| f"btc_{tf_code.lower()}_binance.csv"
|
| )
|
|
|
|
|
| resampled = resampled.reset_index()
|
| resampled.to_csv(output_file, index=False)
|
|
|
| output_files[tf_name] = output_file
|
|
|
| print(f" ✓ Created {len(resampled):,} {tf_name} candles")
|
| print(f" ✓ Saved to: {output_file}")
|
|
|
| return output_files
|
|
|
|
|
| def estimate_trades_per_timeframe():
|
| """Display estimated trades per day for each timeframe."""
|
| print("\n" + "="*70)
|
| print("ESTIMATED TRADING CHARACTERISTICS BY TIMEFRAME")
|
| print("="*70)
|
|
|
| print("\nBased on current parameters (MA 15/60, ~2 trades/day on 1-min):\n")
|
|
|
| timeframes = [
|
| ("1-minute", 1, "~2 trades/day", "High frequency, many signals"),
|
| ("5-minute", 5, "~8-12 trades/week", "Active intraday"),
|
| ("15-minute", 15, "~2-4 trades/week", "Swing trading"),
|
| ("30-minute", 30, "~1-2 trades/week", "Position trading"),
|
| ("1-hour", 60, "~5-10 trades/month", "Low frequency"),
|
| ("4-hour", 240, "~1-3 trades/month", "Very selective"),
|
| ("Daily", 1440, "~2-6 trades/year", "Long-term holds")
|
| ]
|
|
|
| print(f"{'Timeframe':<15} {'Candle (min)':<15} {'Frequency':<20} {'Style'}")
|
| print("-" * 70)
|
| for tf, mins, freq, style in timeframes:
|
| print(f"{tf:<15} {mins:<15} {freq:<20} {style}")
|
|
|
| print("\n💡 Hypothesis:")
|
| print(" - Longer timeframes = Less noise, fewer false signals")
|
| print(" - Better win rate but fewer opportunities")
|
| print(" - May need to adjust MA periods (15/60 → 10/30 or 20/80)")
|
|
|
|
|
| def main():
|
| """Main execution."""
|
| print("\n🌍 MULTI-TIMEFRAME DATA CREATOR")
|
|
|
|
|
| source_file = os.path.join(config.DATA_DIR, "btc_1m_binance_730d.csv")
|
|
|
| if not os.path.exists(source_file):
|
| print(f"\n✗ Source file not found: {source_file}")
|
| print(f" Please ensure you have 1-minute data available.")
|
| return
|
|
|
|
|
| output_files = create_all_timeframes(source_file)
|
|
|
|
|
| estimate_trades_per_timeframe()
|
|
|
| print("\n" + "="*70)
|
| print("✅ MULTI-TIMEFRAME DATASETS CREATED")
|
| print("="*70)
|
|
|
| print("\n📁 Created files:")
|
| for tf_name, filepath in output_files.items():
|
| print(f" - {tf_name}: {filepath}")
|
|
|
| print("\n💡 Next steps:")
|
| print(" 1. Test current parameters on each timeframe:")
|
| print(" - Update config.py to use each timeframe file")
|
| print(" - Run: python run_backtest.py")
|
| print(" - Compare results")
|
| print("\n 2. Optimize parameters per timeframe:")
|
| print(" - Run: python optimize_parameters.py (for each TF)")
|
| print(" - Find best settings for each interval")
|
| print("\n 3. Compare performance:")
|
| print(" - Which timeframe gives best Sharpe ratio?")
|
| print(" - Which has best win rate?")
|
| print(" - Which has lowest drawdown?")
|
|
|
| print("="*70 + "\n")
|
|
|
|
|
| if __name__ == "__main__":
|
| main()
|
|
|
