search

robjhyndman / forecast

Forecasting Functions for Time Series and Linear Models
http://pkg.robjhyndman.com/forecast
1.11k stars 341 forks source link

Problem with tbats refit #325

Open robjhyndman opened 8 years ago

robjhyndman commented 8 years ago 
x1 <- c(taylor[1:3000])
x2 <- c(taylor[500:length(taylor)])
fit <- tbats(x1, seasonal.periods=c(48,336))
fit2 <- tbats(x2, model=fit)
plot(fit)
plot(fit2)
plot(forecast(fit2))
mitchelloharawild commented 5 years ago

Reprex:

library(forecast)
x1 <- c(taylor[1:3000])
x2 <- c(taylor[500:length(taylor)])
fit <- tbats(x1, seasonal.periods=c(48,336))
fit2 <- tbats(x2, model=fit)
plot(fit)

plot(fit2)

plot(forecast(fit2))

Created on 2019-01-03 by the reprex package (v0.2.1)

robjhyndman commented 5 years ago

The problem is that the level component in fit2 is contaminated by the seasonal components. I think we should leave this for now and see if the same problem occurs after a re-implementation in fable.