Swing Trading Automat

An automated trading system for Binance using Python.

Project Structure

swing_trade_automat_2/
├── docs/               # Documentation and development plans
├── logs/              # Trading logs and performance plots
│   └── plots/         # Generated trading performance visualizations
├── model/             # Trained ML models
├── src/               # Source code
├── tools/             # Analysis and utility tools
├── venv/              # Python virtual environment
├── .env               # Environment variables (not in git)
└── requirements.txt   # Python dependencies

Setup

1. Clone the repository:

# Clone using HTTPS
git clone https://github.com/padak/swing_trade_automat_2.git
# or clone using SSH
git clone [email protected]:padak/swing_trade_automat_2.git

# Navigate to the project directory
cd swing_trade_automat_2

2. Create required directories:

mkdir -p logs/plots model

3. Create and activate a virtual environment:

python -m venv venv
source venv/bin/activate  # On Unix/macOS
# or
.\venv\Scripts\activate  # On Windows

4. Install dependencies:

pip install -r requirements.txt

5. Create a .env file in the root directory with your Binance API credentials:

BINANCE_API_KEY=your_api_key_here
BINANCE_API_SECRET=your_api_secret_here

# Optional trading parameters (defaults will be used if not set)
SYMBOL=TRUMPUSDC
INVEST_USD=500
PROFIT_THRESHOLD_PERCENT=0.002
SLEEP_TIME=300  # 5 minutes in seconds

# Strategy parameters (optional)
FAST_MA_PERIOD=10
SLOW_MA_PERIOD=30
RSI_PERIOD=14
RSI_OVERBOUGHT=70
RSI_OVERSOLD=35

Running the System

1. Start the trading bot:

python src/trend_detector_v2_gemini.py

2. Analyze trading performance:

python tools/analyze_trading_log.py

3. View trading performance plots:

python tools/plot_log.py

4. Inspect the trained model:

python tools/inspect_model.py

Directory Usage

• logs/: Contains trading activity logs and performance metrics
  • plots/: Generated visualizations of trading performance
• model/: Stores trained machine learning models
• src/: Main trading bot implementation
• tools/: Utility scripts for analysis and visualization
  • analyze_trading_log.py: Analyzes trading performance and suggests improvements
  • plot_log.py: Generates visual performance charts
  • inspect_model.py: Examines the trained ML model parameters

Development

Follow the incremental development plan outlined in docs/DEVELOPER_PLAN.md.

Monitoring

1. Check the trading log in logs/trading_log.csv for detailed trade history
2. View performance plots in logs/plots/trading_performance.png
3. Monitor model performance using the analysis tools in the tools/ directory

Troubleshooting

1. If you see "Model not found" errors:

   • Run the trading bot first to generate the initial model
   • Check that the model/ directory exists

2. If you see "Log file not found" errors:

   • Ensure the trading bot has been run at least once
   • Verify the logs/ directory exists

3. If plots are not generating:

   • Ensure the logs/plots/ directory exists
   • Check that matplotlib is properly installed
   • Verify there is data in the trading log file