Candle Stick

I have developed the following R package to determine candle stick patterns. You can check out more at GitHub.

CandleStickPattern

This package finds common candle stick patterns using daily data (OHLC data).

Overview

The current version package covers the following patterns:

  • Doji family (1-day pattern)
    • Doji doji()
    • Dragonfly Doji dragonfly.doji()
    • Gravestone Doji gravestone.doji()
  • Hammer family (2-day pattern)
    • Hammer/Hangman hammer()
    • Inverted Hammer/Shooting Star inverted.hammer
  • Engulfing family (2-day pattern)
    • Bullish Engulfing bullish.engulf()
    • Bearish Engulfing bearish.engulf()
  • Harami family (2-day pattern)
    • Bullish Harami bullish.engulf()
    • Bearish Harami bearish.harami()
  • Reversal family (2-day pattern)
    • Piercing line piercing.line()
    • Dark cloud Cover dark.cloud.cover()
  • Kicking family (2-day pattern)
    • Kick up kick.up()
    • Kick down kick.down()
  • Three-in-a-row (3-day pattern)
    • Three white soldiers three.white.soliders()
    • Three Black crows three.black.crows()
  • Star (3-day pattern)
    • Morning star morning.star()
    • Evening star evening.star()
  • Three Method (5-day pattern)
    • Rising three rising.three()
    • Falling three falling.three()

Moreover, it captures trends using exponential moving average (EMA):

  • EMA trends
    • up trend up.trend()
    • down trend down.trend()

Installation

To install this package, it is the easist to install devtools package first:

install.packages("devtools")

Now load the package using library()

library(devtools)

Then we can use the install_github to get the package directly:

install_github("kochiuyu/CandleStickPattern")

Finally, we load the package using library()

library(CandleStickPattern)

Usage

You can see the details of how each function works in vignette

The following demonstrates the usage of the function doji() using Microsoft (ticker: MSFT). We use quantmod package to download data:

library(quantmod)
getSymbols("MSFT", from = "2011-07-08" , to = "2011-07-19")

We first plot the candle chart first:

candleChart(MSFT,theme='white')

We can see that July 11, 12, 13, and 18 are likely to follow doji pattern. Let us see if we can capture using the function doji:

doji(MSFT)
## Warning: timezone of object (UTC) is different than current timezone ().
##             doji
## 2011-07-08 FALSE
## 2011-07-11  TRUE
## 2011-07-12  TRUE
## 2011-07-13  TRUE
## 2011-07-14 FALSE
## 2011-07-15 FALSE
## 2011-07-18  TRUE
## 2011-07-19 FALSE
Next