High‑performance quantitative backtesting platform built in Rust with Python bindings and a Streamlit UI.

GlowBack provides a fast, realistic backtesting engine with data management, storage, and analytics. It includes:

• Event‑driven simulation engine with slippage/latency/commission models
• Data ingestion (CSV, Alpha Vantage, sample data)
• Arrow/Parquet columnar storage and SQLite metadata catalog
• Strategy library (4 built‑in strategies)
• Python bindings (async support)
• Streamlit UI for strategy development and analysis

Phase 0+ (Production Infrastructure) is complete. Phase 1 (Alpha) is in progress.

• Streamlit interface for data loading, strategy editing, running backtests, and result analysis

• Realistic market simulation with configurable market hours and resolution
• Multi‑asset backtesting: equities and crypto (spot) with asset-class-aware fees, market hours, and fractional quantities
• Multi‑symbol backtesting with chronological event ordering
• Performance analytics (Sharpe, Sortino, Calmar, CAGR, Max Drawdown, etc.)
• Risk analytics (VaR, CVaR, skewness, kurtosis)
• Strategy library: Buy & Hold, Moving Average Crossover, Momentum, Mean Reversion, RSI
• Storage: Arrow/Parquet with batch loading and round‑trip I/O
• Catalog: SQLite metadata with indexed queries

Prerequisites:

• Rust (latest stable)
• Python 3.8+ (for Python bindings)

# Clone
git clone <repository-url>
cd glowback

# Run tests
cargo test --workspace

# Basic usage
cargo run --example basic_usage -p gb-types

# Market simulator tests
cargo test -p gb-engine simulator

# Parquet loader tests
cargo test -p gb-data parquet

cd ui && python setup.py
# Opens http://localhost:8501

cp .env.example .env   # configure ports, API key, etc.
docker compose up --build -d

Services:

• UI: http://localhost:8501
• API: http://localhost:8000 (set GLOWBACK_API_KEY to require auth)
• Engine: http://localhost:8081 (health JSON)

All services include health checks, restart policies, and resource limits. See the Deployment Guide for production configuration.

import glowback

manager = glowback.PyDataManager()
manager.add_sample_provider()
manager.add_csv_provider("/path/to/data")
manager.add_alpha_vantage_provider("your_api_key")

bars = manager.load_data(symbol, "2023-01-01T00:00:00Z", "2023-12-31T23:59:59Z", "day")

cargo test --workspace
# 25 passed; 0 failed

In progress

• Performance benchmarking and optimization
• Additional strategies (Bollinger Bands, pairs trading)

Planned

• Advanced analytics (drawdown, factor exposure)
• Parameter sweep and walk‑forward optimization
• Expanded documentation

Contributions are welcome. Focus areas:

1. Performance optimization
2. Additional strategies and analytics
3. UI enhancements
4. Documentation and examples

Open an issue or PR with a clear description of the change.

MIT License — see LICENSE.