0
$\begingroup$

I was following the examples provided by Forecasting: principles and practice. I have modified a little bit the Holt-Winters' example:

 aust <- window(austourists,start=2009)
> fit1 <- hw(aust,seasonal="additive")
Error in ets(x, "AAA", alpha = alpha, beta = beta, gamma = gamma, damped = damped,  : 
  You've got to be joking. I need more data!
> fit2 <- hw(aust,seasonal="multiplicative")
Error in ets(x, "MAM", alpha = alpha, beta = beta, gamma = gamma, damped = damped,  : 
  You've got to be joking. I need more data!

>aust
         Qtr1     Qtr2     Qtr3     Qtr4
2009 55.55857 33.85092 42.07638 45.64229
2010 59.76678 35.19188 44.31974 47.91374

Of course I understand that the problem is that the time series is too short. But with two previous periods of time the formula has not enough. Reviewing the formula I don't know why it doesn't work with two periods of time.

$\endgroup$

1 Answer 1

4
$\begingroup$

You have 8 observations, and you are trying to fit a model with 3 smoothing parameters (level, slope and seasonal) and 5 state parameters (level, slope and 3 seasonal states). So you have one observation for each parameter, which is insufficient.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.