deltafq 0.5.0

A comprehensive Python quantitative finance library

DeltaFQ

Modern Python quantitative finance framework focused on strategy research, backtesting execution, and performance visualization.

Language / 语言: 中文 | English

---

Installation

pip install deltafq

• Requires Python ≥ 3.9.
• Optional components like Plotly and TA-Lib can be installed via pip install deltafq[viz] and pip install deltafq[talib].

---

Overview

• Lightweight, modular quantitative research infrastructure covering the full pipeline: data → indicators → strategy → backtest → visualization.
• Built-in consistent signal standard (Series type) for strategy reuse and component decoupling.
• Suitable for desktop research workflows and script automation, supporting rapid validation and continuous integration.

---

Core Capabilities

• Data Access: Unified data fetching, cleaning, and validation processes.
• Indicator Library: TechnicalIndicators/SignalGenerator provide mainstream indicators and various combination methods.
• Strategy Layer: BaseStrategy abstracts strategy lifecycle for easy extension and backtest reuse.
• Backtest Execution: BacktestEngine integrates underlying execution, position management, and performance metrics.
• Performance Visualization: PerformanceReporter (bilingual) and PerformanceChart (Matplotlib / Plotly).

---

Module Architecture

deltafq/
├── data        # Data acquisition, cleaning, storage interfaces
├── indicators  # Technical indicators and factor calculations
├── strategy    # Signal generators and strategy base classes
├── backtest    # Backtest execution, performance metrics, reporting
├── charts      # Signal/performance chart components
└── trader      # Trading execution and risk control (ongoing expansion)

API Interfaces

• data: DataFetcher (fetch Yahoo Finance data using yfinance), DataCleaner, DataStorage
• indicators: TechnicalIndicators (SMA/EMA/RSI/KDJ/BOLL/OBV/MACD, etc.), TalibIndicators (optional, requires TA-Lib), FundamentalIndicators
• strategy: BaseStrategy (strategy base class), SignalGenerator (signal generation and combination)
• backtest: BacktestEngine (backtest engine), PerformanceReporter (performance reports, bilingual support)
• charts: PerformanceChart (visualization using Matplotlib/Plotly), PriceChart, SignalChart
• trader: ExecutionEngine (trading execution engine), OrderManager, PositionManager, Broker (broker API not yet integrated)

---

Quick Start (BOLL Strategy)

import deltafq as dfq

symbol = "AAPL"
fetcher = dfq.data.DataFetcher()
indicators = dfq.indicators.TechnicalIndicators()
signals = dfq.strategy.SignalGenerator()
engine = dfq.backtest.BacktestEngine(initial_capital=100_000)
reporter = dfq.backtest.PerformanceReporter()
chart = dfq.charts.PerformanceChart()

data = fetcher.fetch_data(symbol, "2023-01-01", "2023-12-31", clean=True)
bands = indicators.boll(data["Close"], period=20, std_dev=2)
signal_series = signals.boll_signals(price=data["Close"], bands=bands, method="cross_current")

trades_df, values_df = engine.run_backtest(symbol, signal_series, data["Close"], strategy_name="BOLL")

reporter.print_summary(symbol, trades_df, values_df, title=f"{symbol} BOLL Strategy", language="en")
chart.plot_backtest_charts(values_df=values_df, benchmark_close=data["Close"], title=f"{symbol} BOLL Strategy")

---

Examples & Tools

• 01_fetch_yahoo_data.py: Fetch historical data from Yahoo Finance using yfinance
• 02_compare_indicators.py: Technical indicator calculation and comparison
• 03_compare_signals.py: Multi-indicator signal generation and combination
• 04_backtest_execution.py: Single-strategy backtest execution workflow
• 05_backtest_report.py / 05_backtest_charts.py: Performance reports and chart visualization
• 06_base_strategy_demo.py: Moving average crossover strategy example based on BaseStrategy
• 07_backtest_engine_tpl.py: BacktestEngine template usage example
• 08_deltafq_template.ipynb: Complete strategy template example (BOLL strategy, two implementation methods)
• 09_multi_factor_strategy.ipynb: Multi-factor strategy example (SMA/EMA/RSI/KDJ/BOLL/OBV combination)

---

Community & Contributing

• Welcome to provide feedback and submit improvements via Issues / PRs.
• The project follows a clean code style; it's recommended to run basic lint/tests before submitting.

---

License

MIT License. See LICENSE for details.