Welcome to StarQuant

中文

StarQuant is light-weighted, integrated algo-backtest/trade system/platform for individual trader, it is mainly used for future trading at present, stock and other commodity will be included in future.

Features

• decoupled module design, namely strategy, marketdatafeed, trade run by different processes, communicate by message queue(nanomsg), low-latency(30 -100 microsenconds);
• event-driven based backtest system(same as vnpy), strategy bactested in python, run directly in live；support multiple data source, i.e. data can be loaded from MongoDB, CSV file; support multiple timescale data such as tick ,1min bar, 1h bar;
• support strategy parameter optimization through multi-process or genetic algorithm;
• marketdata record;
• simulated trading(paper brokerage);
• support multiple API and accounts, autoconnect/logout/reset, working in 7*24h；support self-defined instructions, such as cancel all; support local stop orders;
• strategy can be dynamically managed, i.e. init/start/stop/edit/remove at will;
• pyQt5 based GUI interface for monitoring and manual control；
• risk manage, flow control;
• realtime message notify through wechat(itchat) ...

Architecture

C++ 11 based, client-server, event-driven, decoupled module design.

Development Environment

Manjaro（arch，Linux 4.14)，python 3.7.2，gcc 8.2, anaconda 5.2

third party softwares:

• boost 1.69
• nanomsg
• log4cplus
• yamlcpp
• libmongoc-1.0
• ...

python modules:

• pandas-datareader
• psutil
• pyyaml
• pyqt
• qdarkstyle
• tushare
• pyfolio
• itchat
• ...

How to Run

compile files in cppsrc:

$ cd cppsrc
$ mkdir build
$ cd build
$ cmake ..
$ make
$ cp StarQuant/apiserver.exe ../../

start apiserver and gui, strategy, recorder:

$ ./apiserver

$ python gui.py
$ python strategy.py
$ python recorder.py

User Guide

To be continued

Demo

Current State

Currently StarQuant is under test, v1.0-alpha version has been released.