How to use vnpy: install, run and build quant trading apps

What vnpy is and who it suits

vnpy (VeighNa) is a Python framework for developers and traders who want to build their own quantitative trading systems. It bundles connections to many trading interfaces, a set of ready-made trading apps such as strategy backtesting, risk management and data management, and a newer vnpy.alpha module for machine-learning-based strategy research.

The README lists Chinese futures and securities markets as its main focus, alongside overseas brokers such as Interactive Brokers. It is a framework to build on, not a finished product you switch on.

• Good fit: Python developers building custom quant systems
• Good fit: researchers who want data, model training, signals and backtests in one workflow
• Poor fit: complete beginners with no Python or trading background, since the README has no beginner walkthrough
• Poor fit: anyone expecting plug-and-play trading or guaranteed returns

What you need before you install vnpy

The README states that Python 3.10 or newer is required, and its badges list Python 3.10 to 3.13. It also lists Windows, Linux and macOS as supported platforms.

Many trading gateways need a broker or data-vendor account and vendor-specific setup. Plan for that before you expect live connections to work.

• Python 3.10 or newer (3.10 to 3.13 per the README badges)
• Windows, Linux or macOS
• A broker or data account if you want to connect a specific gateway

Get the vnpy source code

The README's install instruction is to clone the repository with git. Run the command below in the folder where you want the project to live. Then follow the README and project documentation for the remaining setup steps.

Explore the trading apps and modules

vnpy is organised as a set of modules. The README describes a trader core, gateways for specific markets, trading apps, an API layer, an event engine and database adapters. You can pick the pieces your project needs.

Some useful apps described in the README include cta_strategy for CTA strategy engines, cta_backtester for backtesting with a graphical interface, risk_manager for pre-trade limits, and data_manager for viewing and importing or exporting historical data. Each app is maintained in its own linked repository.

• cta_strategy: CTA strategy engine with fine-grained control over order placement and cancellation
• cta_backtester: strategy backtesting and parameter optimisation in a graphical interface
• risk_manager: rules such as order flow control, order size and cancel counts
• data_manager: browse stored data and import or export CSV files
• web_trader: REST and Websocket web service for browser-based setups

Choose a database for your data

The README says SQLite is the default database option. It is a single-file database that needs no separate server, which suits beginners. MySQL and PostgreSQL are relational options, and QuestDB, TDengine and MongoDB are listed among the other choices.

The README notes that PostgreSQL is recommended only for experienced users.

• SQLite: default, single file, no server to configure
• MySQL: widely used open-source relational database
• PostgreSQL: feature-rich, recommended only for experienced users
• QuestDB, DolphinDB, TDengine, MongoDB: time-series and document options listed in the README

Try the vnpy.alpha research workflow

The vnpy.alpha module is aimed at multi-factor machine-learning strategy research. It has four parts: dataset for feature engineering, model for training predictions, strategy for building signal-based strategies, and lab for managing the research workflow from data to backtest.

The README points to example notebooks in examples/alpha_research. These cover Lasso, LightGBM and MLP workflows. The notebooks are the most concrete starting point the README offers for this module.

• dataset: factor features and data processing helpers
• model: Lasso, LightGBM and MLP model templates
• strategy: cross-sectional and time-series strategy types
• lab: data management, training, signal generation and backtesting