R语言_量化回测_Quantitative-backtest-based-on-R-software.zip

install.packages('xts') install.packages('zoo') install.packages('TTR') install.packages('quantmod') library(xts) library(zoo) library(TTR) library(quantmod) stock1<-getSymbols("601318.ss", src='yahoo',from="2013-01-01", to=Sys.Date(), auto.assign=F) class(stock1) names(stock1) head(stock1) names(stock1) = c('open', 'high', 'low', 'close', 'volume', 'adjusted') head(stock1) stock1df = as.data.frame(stock1) head(stock1df) stock1df['2019-09-09',] dateArea <- function(sDate = sDate, eDate = Sys.Date()){ if(class(sDate)=='character') sDate = as.Date(sDate) if(class(eDate)=='character') eDate = as.Date(eDate) return(paste(sDate, eDate, sep= '/')) } ma <- function(cdata, mas = c(5, 20, 60)){ if(nrow(cdata) <= 60) return(NULL) ldata <- cdata for(m in mas){ ldata <- merge(ldata, SMA(cdata, m)) } names(ldata) <- c('value', paste('ma', mas, sep = '')) return(ldata) } sDate <- as.Date('2018-01-01') eDate <- Sys.Date() cdata <- stock1[dateArea(sDate, eDate)]$close head(cdata) ldata = NULL ldata <- ma(cdata, c(5, 20, 60)) tail(ldata) plot(ldata) #多头排列买入 buypoint <- function(ldata){ idx <- which(ldata$ma5>ldata$ma20 & ldata$ma20>ldata$ma60) return(ldata[idx,]) } buypoint <- buypoint(ldata) head(buypoint) #空头排列卖出 sellpoint <- function(ldata){ idx <- which(ldata$ma5<ldata$ma20 & ldata$ma20<ldata$ma60) return(ldata[idx,]) } sellpoint <- sellpoint(ldata) head(sellpoint) bsdata <- merge(buypoint$value, sellpoint$value) head(bsdata) names(bsdata) = c('buy', 'sell') head(bsdata) signal <- function(buy, sell){ selldf <- data.frame(sell, op = as.character(rep('S', nrow(sell)))) buydf <- data.frame(buy, op = as.character(rep('B', nrow(buy)))) sdata <- rbind(buydf, selldf) sdata[order(as.Date(row.names(sdata))),] #order按索引排序 } sdata <- signal(buypoint, sellpoint) View(sdata) trade <- function(sdata, capital = 10000,fee_ratio=0.001){ amount = 0 #持有股票数 cash = capital #所持现金 ticks <- data.frame() for(i in 1:nrow(sdata)){ row <- sdata[i,] if(row$op == 'B'){ amount0<- floor((1-fee_ratio)*cash/row$value) amount <- amount + amount0 cash <- cash - amount0*row$value*(1+fee_ratio) } if(row$op =='S'){ cash <- cash+amount*row$value*(1-fee_ratio) amount = 0 } row$cash <- round(cash,2) row$amount <- amount row$asset <- round(cash+amount*row$value, 2) ticks <- rbind(ticks, row) } ticks$diff <- c(0, round(diff(ticks$asset), 2)) #由于差分，向量的长度缩小了一个，需要增加一个向量单元 rise <- ticks[intersect(which(ticks$diff >0), which(ticks$op =='S')),] fall <- ticks[intersect(which(ticks$diff <0), which(ticks$op =='S')),] return(list(ticks = ticks, rise = rise, fall = fall)) } Return <- trade(sdata) View(Return$ticks) tick <- Return$ticks head(tick) #年化收益 date<-row.names(tick) sday<-as.Date(date[1]) eday<-as.Date(date[length(date)]) eday-sday #相隔890天 res<-tick$asset res end<-as.numeric(res[length(res)]) begin<-as.numeric(res[length(1)]) yield_year<-end/begin-1 time_span_year<-890/365; yieldYear<-(yield_year+1)^(1/time_span_year)-1 #年化收益 yieldYear out <- zoo(tick[,'asset'], as.Date(row.names(tick))) out <- as.xts(out) names(out) = 'asset' #作图 par(mfrow=c(2,1)) plot(ldata[,'value']) points(buypoint[,'value'],col='red') points(sellpoint[,'value'],col='green') plot(out)