CODEWITHSUNDEEP
rforecasting
Bryan
Bryan

Reputation: 6149

Is there an easier way to extract combined fitted values and forecast values of an ARIMA in R?

Simple question -- I want to extract the fitted values and mean forecast of an ARIMA as one single time series, using R. Right now, I use this slow and cumbersome code:

x<-auto.arima(y)
z<-forecast(x, 365)

fitted<-z$fitted
mean<-z$mean
merged<-merge.zoo(fitted, mean)
merged$fitted[is.na(merged$fitted)]<-0
merged$mean[is.na(merged$mean)]<-0

finalforecast<-merged$fitted+ merged$mean

There must be an esier way to do this, but looking at the documentation, I'm drawing a blank. I really want to avoid the hassle of merging the time series, then dropping in zeroes (to fill the NAs), and then doing the addition. I know I can create my own function, but I'd be surprised if there isn't something simple that's pre-baked.

Thoughts?

Upvotes: 0

Views: 2778

Answers (1)

Rcoster
Rcoster

Reputation: 3210

Why not just c(z$fitted, z$mean)?

EDIT: If you need a ts object: ts(c(z$fitted, z$mean), start=start(y), frequency=frequency(y))

Upvotes: 3

Related Questions