requirements.txt

# Minimal requirements for Vercel/Render deployment
# For full development dependencies, see requirements-full.txt or pyproject.toml

# FastAPI server for demo UI
fastapi==0.115.6
uvicorn[standard]==0.34.0
# Updated to version with Python 3.13 wheel support (avoids Rust compilation)
pydantic>=2.10.0,<3.0.0

# Flask for demo UI server
flask==3.1.0

# Minimal scientific computing (only numpy for market simulator)
numpy==2.4.0

# Core utilities
python-dateutil==2.9.0.post0

# WebSocket support for real-time price updates
websockets>=13.0

# Optional: QuestDB for time-series metrics logging
questdb==4.1.0

# Optional: ML dependencies for pattern detection
# To use ML features, set 'ml_enabled': true in config and run: python ml/train.py
scikit-learn>=1.5.0
joblib==1.3.2

# Optional: Production Dashboard dependencies (for Streamlit visualization)
# To install production dashboard dependencies:
# pip install streamlit plotly pandas
# 
# Production dashboards:
# - freqtrade/ui/prod_config.py - Configuration management
# - freqtrade/ui/prod_monitor.py - Real-time monitoring
# Uncomment to install:
# streamlit>=1.31.0
# plotly>=5.18.0
# pandas>=2.1.0

# Optional: DSPy for LM-based insights in external analysis script
# To use DSPy insights: pip install dspy-ai ollama
# Updated to version with Python 3.13 compatibility and newer pydantic support
dspy-ai>=3.0.0,<4.0.0
ollama==0.1.0

# Optional: Knowledge Graph for post-mortem analysis
# To use knowledge graph features: pip install networkx pyvis python-louvain
# These enable trade analysis visualization and pattern detection
networkx>=3.4.2
pyvis>=0.3.2
python-louvain>=0.16

# Optional: Full stack for local development (commented out for Vercel)
# To install full dependencies locally:
# pip install -r requirements-full.txt
# or
# pip install -e .[dev]