README.md

In palencar/TraderBot: Quantitative analysis bot.

TraderBot

Installation

Ubuntu Dependencies

sudo apt-get install -y r-base r-base-dev sqlite3 libxml2-dev libmysqlclient-dev libssl-dev libssh2-1-dev libcurl4-openssl-dev

Clone repository

Clone the git repository.

git clone https://github.com/palencar/TraderBot
cd TraderBot

Using MySQL Database

Create the database using the schema file mysql-schema.sql. Uncomment the following lines o the file config.yml. Set the values for host, database, user and password.

engine:   "mysql"
host:     "db_hostname"
database: "db_name"
user:     "db_username"
password: "db_password"

Using SQLite Database

Create the database from the schema file.

sqlite3 db.sqlite < sqlite3-schema.sql

Uncoment the following lines on the file config.yml:

engine:   "sqlite"
database: "db.sqlite"

Install Packages on R

Start a R session.

R

Install devtools Package.

install.packages("devtools")

Install TraderBot and dependencies.

library(devtools)
install_git(".")

Using

Load TraderBot

library(TraderBot)

Download Data

intraday

Download daily data from google.

saveSymbolsDaily("BVMF3")

Daily

Download last 15 days of intraday data from google.

saveSymbolsIntraday("BVMF3")

Splits and Dividends

updateAdjust("BVMF3")

Notes

The data is automatically updated only on live trading. If there's not enough data, the operation (trade, simulation or backtest) will be skipped.

Backtest

Test strategy for minSamples random parameters. The parameter limits are defined on config.yml.

computeBacktest(Symbols = "BVMF3", minSamples = 1000, timeFrame="1D")

Simulate

Test strategy using the values on the tradeParameters.csv file. Every line defines the parameters for a given time frame.

computeSimulation(Symbols = "BVMF3", timeFrame="1D")

Live Trading

Update assets on database and execute the strategy. The parameter timeFrame can be one of the following: 1D, 1H, 30M, 15M, 10M and 5M.

computeStream(Symbols = "BVMF3", timeFrames = c("5M", "10M", "15M", "30M", "1H", "1D"))

If an alert is triggered, the results ar shown on the terminal like below. The chart is written on charts directory.

[1] "Chart [RPMG3] [15M] [2018-01-03 16:40:00]: buy"
[1] "Chart [EMBR3] [30M] [2018-01-03 13:00:00]: sell"
[1] "Chart [CTKA4] [15M] [2018-01-02 14:18:00]: buy"
[1] "Chart [SPXI11] [30M] [2018-01-02 11:16:00]: sell"
  symbol timeframe            datetime alert price
1  RPMG3       15M 2018-01-03 16:40:00   buy  8.19
2  EMBR3       30M 2018-01-03 13:00:00  sell 21.02
3  CTKA4       15M 2018-01-02 14:18:00   buy  3.05
8 SPXI11       30M 2018-01-02 11:16:00  sell 90.47

It is possible to send to email with mutt or copy to aws s3 bucket using the alert options on config.yml.

Shiny App

A simple Shiny App showing charts of the wallet symbols, alerts triggered, etc.

runShinyApp()

Shell Scripts

There are some shell scripts calling various module functions on sh directory.

Live trading for all the symbols included on database.

sh/live.sh

palencar/TraderBot documentation built on April 30, 2020, 11:50 p.m.