Australia
rjfhud(a)powerup.com.au
- Original Message -
From: Gústaf Steingrímsson
To: 'gretl-users(a)lists.wfu.edu'
Sent: Wednesday, October 27, 2010 12:12 AM
Subject: [Gretl-users] Simple ARIMA problem
Hi
While trying to work with ARIMA(1,1,1) in Gretl I got
Dear Gustaf and Richard,
I guess part of the problem is the estimation method. GRETL offers Exact
maximum likelihood estimation (exact MLE) and conditional MLE for ARIMA
models. Only the second is equivalent to OLS. Exact and conditional ML
parameter estimates will "tend" to coincide only in "larg
Hi
While trying to work with ARIMA(1,1,1) in Gretl I got into problems. The
problem is that having set upp the numbers (values, prediction and errors that
was produced by Gretl) in spreadsheet and using the parameters that I got I
couldnt get the same forecast number as Gretl got. I used just t