If I understand you correctly, indicator(-1), indicator(-2), etc are your threshold regressors. Then you want, for example:
Dependent variable followed by list of threshold regressors:
indicator indicator(-1)
List of non-threshold regressors:
c
Threshold variable specification
1 4
This is all explained in the documentation.
Problem with lags in SETAR model
Moderators: EViews Gareth, EViews Moderator
-
EViews Rebecca
- EViews Developer
- Posts: 113
- Joined: Thu Apr 18, 2013 8:37 am
-
statistik123
- Posts: 10
- Joined: Sun Jan 24, 2016 8:34 am
Re: Problem with lags in SETAR model
Yes, you are right!
I tried it the same way like in the documentation and you did. But with the same:
I tried it the same way like in the documentation and you did. But with the same:
I got also an error: Threshold variable lags must be positive integers.
Dependent variable followed by list of threshold regressors:
indicator indicator(-1)
List of non-threshold regressors:
c
Threshold variable specification
1 4
-
EViews Rebecca
- EViews Developer
- Posts: 113
- Joined: Thu Apr 18, 2013 8:37 am
Re: Problem with lags in SETAR model
Are you saying you ran the above and got an error? What exactly did you enter?
-
statistik123
- Posts: 10
- Joined: Sun Jan 24, 2016 8:34 am
Re: Problem with lags in SETAR model
Yes, I ran the above and got errors. The strange thing is that the errors are different:
Threshold variable lags must be positive integers.
Specification leads to singular matrix in at least one sub-sample.
I enter exactly:
zewindicator zewindicator(-1)
c
1 4
...I don´t know. When this is the right equation maybe there is some mistake in the uploaded time serie. I will check it again! But good to know, that I understand the documentation right and that there is no mistake and I have to search somewhere else.
Threshold variable lags must be positive integers.
Specification leads to singular matrix in at least one sub-sample.
I enter exactly:
zewindicator zewindicator(-1)
c
1 4
...I don´t know. When this is the right equation maybe there is some mistake in the uploaded time serie. I will check it again! But good to know, that I understand the documentation right and that there is no mistake and I have to search somewhere else.
Who is online
Users browsing this forum: No registered users and 2 guests
