Bootstrap using ARIMA model
Hi David,
here is my code including what i did for the tsboot:
rm(list = ls())
library(boot)
library(tseries)
library(TTR)
library(quantmod)
library(scales)
library(forecast)
library(zoo)
library(TSA)
security<-"NFLX"
startDate<-"2012-06-01"
endDate<-"2016-10-31"
qte_list<-c("AdjClose")
data=get.hist.quote(instrument = security, startDate, endDate, quote =
qte_list, provider = "yahoo" )
func.ar<- ar(logret)
func.model<-list(order = c(func.ar$order,0,0),ar=func.ar$ar)
func.res<- func.ar$resid[!is.na(func.ar$resid)]
func.res<-func.res - mean()
func<- function(logret,formula){
d = logret
return(RSI(exp(logret)))
}
func.sim<-function(res,n.sim,ran.args){
rg1<- function(n, res) sample(res, n, replace=TRUE)
ts.orig<-ran.args$ts
ts.mod<-ran.args$model
mean(ts.orig)+ts(arima.sim(model=ts.mod,n=n.sim, ran.gen=rg1,
res=as.vestor(res)))
}
myboot<-tsboot(exp(logret),func,R=500,sim="model", ran.gen=func.sim,
ran.args = List(ts=log(data[,1],model=func.sim))
Best,
Ash
On Thu, Dec 1, 2016 at 1:50 PM, Bert Gunter <bgunter.4567 at gmail.com> wrote:
Just briefly to follow up David's comment, though this is mainly about statistics and therefore off topic here... Bootstrapping time series is a subtle issue that requires familiarity with the technical details-- and maybe even current research. The tsboot() function gives you several options from which you must choose *appropriately* -- or maybe choose something else entirely. The Help doc gives you a sense of the difficulties: *************** Model based resampling is very similar to the parametric bootstrap and all simulation must be in one of the user specified functions. This avoids the complicated problem of choosing the block length but relies on an accurate model choice being made. Phase scrambling is described in Section 8.2.4 of Davison and Hinkley (1997). The types of statistic for which this method produces reasonable results is very limited and the other methods seem to do better in most situations. Other types of resampling in the frequency domain can be accomplished using the function boot with the argument sim = "parametric". **** Moral: If you don't know what you're doing, seek local expertise to help -- remote sites offering suggestions from those who aren't familiar with the details of your data and analysis goals (maybe you don't need to do this at all!) may lead you to irreproducible nonsense. Cheers, Bert Bert Gunter "The trouble with having an open mind is that people keep coming along and sticking things into it." -- Opus (aka Berkeley Breathed in his "Bloom County" comic strip ) On Thu, Dec 1, 2016 at 7:45 AM, Ashwini Patil <ash369ster at gmail.com> wrote:
Hi,
I want to implement a bootstrap method for time series.
I am taking the adj close values from yahoo for NFLX and now I need to
bootstrap these values using ARIMA model.
here is my code so far:
rm(list = ls())
library(boot)
library(tseries)
library(TTR)
library(quantmod)
library(scales)
library(forecast)
library(zoo)
library(TSA)
security<-"NFLX"
startDate<-"2012-06-01"
endDate<-"2016-10-31"
qte_list<-c("AdjClose")
data=get.hist.quote(instrument = security, startDate, endDate, quote =
qte_list, provider = "yahoo" )
logret<-diff(log(data[,1]))
fit11<-auto.arima(logret, max.order=10)
When i use auto.arima, I get an order of (0,0,0) with non-zero mean.
After
this, I tried to use tsboot function but it is not yielding any answers.
Any and all help is appreciated.
Thank you!
[[alternative HTML version deleted]]
______________________________________________ R-help at r-project.org mailing list -- To UNSUBSCRIBE and more, see https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/
posting-guide.html
and provide commented, minimal, self-contained, reproducible code.