require(quantmod)
require(xts)
data2 <- read.csv("data2.csv",header=T)
data2 <- as.xts(read.zoo(data2))
 
Macd <- function(data,fast=12,slow=26,difnm=9){
 
  total <- slow+difnm#EMA26需要累積26天，MACD9需要累積9天，這裡為34天
  Dl <- (Hi(data)+Lo(data)+2*Cl(data))/4 #每日需求指數
  nEMA <- runMean(Dl,n=fast) #12日快線的平均
  mEMA <- runMean(Dl,n=slow)#26日慢線的平均
  DIF <- nEMA-mEMA #快慢線平均相減
  MACD <- runMean(DIF,n=difnm)#快慢線平均相減的九日平均
  nEMA <- nEMA[total]#能有計算結果的第一天 首日EMA12
  mEMA <- mEMA[total]#首日EMA26
  MACD <- MACD[total]#首日MACD
  DIF <- DIF[total]#首日OSC
  Date <- as.character(index(data[total,]))
  all <- data.frame(Date,MACD,DIF,stringsAsFactors = FALSE) #合併第一天的macd osc 日期 
 
  for(w in c((total+1):nrow(data))){ #從第35天開始到最後
    todayCl <- as.numeric(Dl[w])  #每一天的Dl
    nEMA=(nEMA*(fast-1)+todayCl*2)/(fast+1) #nEMA=(前一日nEMA*(n-1)＋今日收盤價×2)/(n+1)
    mEMA=(mEMA*(slow-1)+todayCl*2)/(slow+1)#mEMA=(前一日mEMA*(m-1)＋今日收盤價×2)/(m+1)
    DIF=nEMA-mEMA
    MACD=(MACD*(difnm-1)+DIF*2)/(difnm+1)
    Date <- as.character(index(data[w,]))
    all <- rbind(all,c(Date,MACD,DIF))#合併新資料到dataframe
  }
  all$Date <- as.Date(all$Date) #轉型態整理資料 以下可省略
  all$MACD <- round(as.numeric(all$MACD),2)
  all$DIF <- round(as.numeric(all$DIF),2)
  all$OSC <- all$DIF-all$MACD
  all <- as.xts(read.zoo(all))
 
}
 
macd <- Macd(data2,12,26,9)
 

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