MACD code

macd.jl

function macd{T}(ta::TimeArray{T,1}, fast::Int, slow::Int, signal::Int)
 
  fastma = ema(ta, fast)
  slowma = ema(ta, slow)
  mcdval = fastma - slowma
  sigval = ema(mcdval, signal)

  merge(mcdval, sigval, ["macd", "signal"])
end

Of course, you can always refactor ...

function macd{T}(ta::TimeArray{T,1}, fast::Int, slow::Int, signal::Int) 
  mcdval =  ema(ta, fast) -  ema(ta, slow)
  merge(mcdval, ema(mcdval, signal), ["macd", "signal"])
end

Let's not make it a one-liner though! Which by the way is not all that big of a jump.

macd{T}(ta::TimeArray{T,1}, fast::Int, slow::Int, signal::Int) =  merge((ema(ta, fast) -  ema(ta, slow)), ema((ema(ta, fast) -  ema(ta, slow)), signal), ["macd", "signal"])

Hmm, probably the most descriptive code to match algorithm would be this

function macd{T}(ta::TimeArray{T,1}, fast::Int, slow::Int, signal::Int) 
  mcdval = ema(ta, fast) -  ema(ta, slow)
  sigval = ema(mcdval, signal)
  merge(mcdval, sigval, ["macd", "signal"])
end