cerebro

Live Trader

This project is still in progress and is not a full version.

golang live trading framework

Introduce

This project was inspired by backtrader

python backtrader is a great project but it has disadvantage of python GIL so i want solve by golang

Installation

go get github.com/gobenpark/cerebro

Usage

🙏 Plz wait for beta version

1. first implement interface store.Store

package main

import (
	"context"
	"sort"
	"time"

	"github.com/gobenpark/proto/stock"
	"github.com/gobenpark/cerebro/container"
	"github.com/gobenpark/cerebro/order"
	"github.com/rs/zerolog/log"
	uuid "github.com/satori/go.uuid"
)

type store struct {
	name string
	uid  string
	cli  Client
}

func NewStore(name string) *store {
	cli, err := stock.NewSocketClient(context.Background(), "localhost:50051")
	if err != nil {
		panic(err)
	}

	return &store{name, uuid.NewV4().String(), cli}
}

func (s *store) LoadHistory(ctx context.Context, code string, du time.Duration) ([]container.Candle, error) {
	panic("implement me ")
}

func (s *store) LoadTick(ctx context.Context, code string) (<-chan container.Tick, error) {
	panic("implement me ")
}

func (s *store) Order(code string, ot order.OType, size int64, price float64) error {
	panic("implement me ")
}

func (s *store) Cancel(id string) error {
	panic("implement me ")
}

func (s *store) Uid() string {
	return s.uid
}

2. implement your own strategy

Next(broker,container) is incomming tick or candle data in this part you can every using indicator and buy, sell using broker

import (
	"fmt"

	"github.com/gobenpark/cerebro/broker"
	"github.com/gobenpark/cerebro/container"
	"github.com/gobenpark/cerebro/indicators"
	"github.com/gobenpark/cerebro/order"
)

type Bighands struct {
	Broker broker.Broker
	indi   indicators.Indicator
}

func (s *Bighands) Next(broker *broker.Broker, container container.Container) {
	rsi := indicators.NewRsi(14)
	rsi.Calculate(container)
	fmt.Println(rsi.Get()[0])

	fmt.Println(broker.GetCash())

	obv := indicators.NewObv()

	obv.Calculate(container)
	fmt.Println(obv.Get()[0])
	fmt.Println(container.Code())

	
	sma := indicators.NewSma(20)
	sma.Calculate(container)
	fmt.Println(sma.Get()[0])
	fmt.Println(container.Code())
}

func (s *Bighands) NotifyOrder(o *order.Order) {
	switch o.Status() {
	case order.Submitted:
		fmt.Printf("%s:%s\n", o.Code, "Submitted")
		fmt.Println(o.ExecutedAt)
	case order.Expired:
		fmt.Println("expired")
		fmt.Println(o.ExecutedAt)
	case order.Rejected:
		fmt.Println("rejected")
		fmt.Println(o.ExecutedAt)
	case order.Canceled:
		fmt.Println("canceled")
		fmt.Println(o.ExecutedAt)
	case order.Completed:
		fmt.Printf("%s:%s\n", o.Code, "Completed")
		fmt.Println(o.ExecutedAt)
		fmt.Println(o.Price)
		fmt.Println(o.Code)
		fmt.Println(o.Size)
	case order.Partial:
		fmt.Println("partial")
		fmt.Println(o.ExecutedAt)
	}
}

func (s *Bighands) NotifyTrade() {
	panic("implement me")
}

func (s *Bighands) NotifyCashValue() {
	panic("implement me")
}

func (s *Bighands) NotifyFund() {
	panic("implement me")
}

3. using cerebro !!

package main

import (
	"time"
	"github.com/gobenpark/cerebro/broker"
	"github.com/gobenpark/cerebro/cerebro"
	"github.com/gobenpark/cerebro/strategy"
)

func main() {
	bk := broker.NewBroker(100000, 0.0005)

	upbit := NewStore("binance")

	smart := &strategy.Bighands{
		Broker: bk,
	}

	cb := cerebro.NewCerebro(
		cerebro.WithBroker(bk),
		cerebro.WithStore(upbit, "KRW-MFT", "KRW-LBC"),
		cerebro.WithStrategy(smart),
		cerebro.WithResample("KRW-MFT", time.Minute*3, true),
		cerebro.WithResample("KRW-LBC", time.Minute*3, true),
		cerebro.WithLive(true),
		cerebro.WithPreload(true),
	)

	err := cb.Start()
	if err != nil {
		panic(err)
	}
}

Concepts

Live Trader have several part of trading components

1. Cerebro : the Cerebro managements all trading components and make dependency graph

2. Store components is user base implements for external real server ex) Binance , upbit , etc

3. Strategy is user base own strategy

Feature

1. Indicator
   • bollinger band
   • RSI
   • Simple Moving Average
   • On Balance Bolume

TODO

1. new feature Observer is observing price and volume for find big hands
2. new feature Signal
3. support news, etc information base trading
4. multi store one cerebro
5. feature broker slippage
6. Chart

Version

1. MAJOR version when you make incompatible API changes,
2. MINOR version when you add functionality in a backwards compatible manner, and
3. PATCH version when you make backwards compatible bug fixes. Additional labels for pre-release and build metadata are available as extensions to the MAJOR.MINOR.PATCH format.

https://semver.org

Contribute

if you have any idea and want to talk this project send me mail (qjadn0914@naver.com) or issue