R wrapper around OANDA FX trading V20 API

This project is an attepmt to bring research, backtest, trading, and monitoring using R wrapper around OANDA broker's HTTP API . It was motivated by a need to quickly turn research in to executable trades. There are three components:

• Research
• Trading/Backtesting
• Account Monitoring

QUICK GUIDE

Installation

1. Prerequisites: Ensure that the following packages are installed

  > install.packages(c("curl","jsonlite","devtools")) # only the package
  > install.packages(c("curl","jsonlite","devtools","shiny","shinythemes","ini")) # package + dashboard
  > install.packages(c("curl","jsonlite","devtools","shiny","shinythemes","ini","zoo")) #package + dashboard + demo backtest & toy algo trading

2. Install oanda's api

  > library(devtools)
  > install_github(repo="ltekengineering/oanda", subdir="oanda")

3. Additional features:

  foo@bar: git clone git@github.com:ltekengineering/oanda.git    

Trading API

• Implementation of the trading API in R for algorithmic trading. A complete tutorial is here.

  Using library

    > library(oanda)    

  Setup

  • Point to the practice account and authenticate with code generated on https://www1.oanda.com/demo-account/login

    > oanda.init("practice",'44667a4c1a94db7f245ea21fe2f82341-04efe2ed548b91e6bb2d7d7707668d8a')

  Time series of bid prices

  > eur_usd <- oanda.instruments.bid("EUR_USD","M1",from = "1608064200") #2020-12-15 03:30 PM
  > usd_cad <- oanda.instruments.bid("USD_CAD","S5",from="2020-12-15T15:30:00") #with a resolution of 5 seconds
  > head(usd_cad$candles)
    complete volume       time     b_o     b_h     b_l     b_c
  1     TRUE      2 1608064200 1.26891 1.26891 1.26890 1.26890
  2     TRUE      1 1608064210 1.26887 1.26887 1.26887 1.26887
  3     TRUE      1 1608064215 1.26889 1.26889 1.26889 1.26889
  4     TRUE      2 1608064230 1.26892 1.26892 1.26889 1.26889
  5     TRUE      2 1608064245 1.26891 1.26891 1.26890 1.26890
  6     TRUE      2 1608064260 1.26891 1.26891 1.26889 1.26889
  
  > libray(plotly)
  > usd_cad$candles$time <- as.POSIXct(usd_cad$candles$time,origin="1970-01-01")
  > fig <- usd_cad$candles[1:50,] %>% plot_ly(x = ~time, type="candlestick",open = ~b_o, close = ~b_c, high = ~b_h, low = ~b_l) 
  > fig <- fig %>% layout(title = "USD/CAD Candlestick Chart: Bid")
  > fig

  

  Create a market order

  > oanda.orders.createMarketOrder('101-001-3704066-001',"USD_MXN",10000) #long 10000 USD/MXN
  > oanda.orders.createMarketOrder('101-001-3704066-001',"USD_MXN",-10000) #short 10000 USD/MXN
  > oanda.orders.createMarketOrder('101-001-3704066-001',"USD_MXN",10000,priceBound=19.86263) #worst pirce of 19.86263 else kill the order
  > info <- oanda.orders.createClientExtensions("A comment","strategy_9","my_order_10000")
  > oanda.orders.createMarketOrder('101-001-3704066-001',"USD_MXN",10000,priceBound=19.86263, info)  #add customization

Shiny Dashboard

• A shiny app to monitor account activity in real time. Visit the wiki page for a detailed tutorial.

1. Go to dashboard directory

  foo@bar: cd dasboard

2. Edit token and accounts parameters in config_ppe.ini or config_prod.ini with your token and account numbers
3. Start the dashboard

  foo@bar: ln -s config_ppe.ini config.ini 
  foo@bar: Rscript --no-save dash.r        

Point to config_prod.ini when you're ready to go live

4. Go here http://localhost:8989/

Backtest & Algo Trading

• Back-testing demo scripts have been removed