## how to generate an arma

For questions regarding programming in the EViews programming language.

Moderators: EViews Gareth, EViews Jason, EViews Moderator, EViews Matt

hnida
Posts: 9
Joined: Fri Jan 02, 2009 7:09 am

### how to generate an arma

Hi
I am using eviews for the first time, so my question is quite obvious i guess...
I am using eviews3.0, and i want to generate different armas, that's the syntax i use:
create u 2000
genr a=rnd (nrnd is not working on this version i guess)
GENR ARMA=0
GENR ARMA= 0.9*ARMA(-1)+a+0.8*A(-1)
When i open the variables, i have different values for the a, but always get "na" for the arma. I tried different combination and it seems like the programm is not able to calculate the term arma(-1).

startz
Non-normality and collinearity are NOT problems!
Posts: 3386
Joined: Wed Sep 17, 2008 2:25 pm

### Re: how to generate an arma

hnida wrote:Hi
I am using eviews for the first time, so my question is quite obvious i guess...
I am using eviews3.0, and i want to generate different armas, that's the syntax i use:
create u 2000
genr a=rnd (nrnd is not working on this version i guess)
GENR ARMA=0
GENR ARMA= 0.9*ARMA(-1)+a+0.8*A(-1)
When i open the variables, i have different values for the a, but always get "na" for the arma. I tried different combination and it seems like the programm is not able to calculate the term arma(-1).

For the first observation A(-1) doesn't exist. It's treated as an NA, which then propagates to the rest of the series. Try

Code: Select all

`genr a=rnd GENR ARMA=0SMPL 2 2000GENR ARMA= 0.9*ARMA(-1)+a+0.8*A(-1)`

By the way, I would be very surprised if nrnd wasn't working in version 3.

hnida
Posts: 9
Joined: Fri Jan 02, 2009 7:09 am

### Re: how to generate an arma

it's working perfectly. I thought the observations started at 0...
Tanks a lot.