eviews student and ARDL

For technical questions regarding estimation of single equations, systems, VARs, Factor analysis and State Space Models in EViews. General econometric questions and advice should go in the Econometric Discussions forum.

Moderators: EViews Gareth, EViews Moderator

fati
Posts: 11
Joined: Mon Mar 26, 2018 10:05 am

eviews student and ARDL

Postby fati » Mon Mar 26, 2018 1:48 pm

Hello.
I've eviews10 student version.I can't use ARDL because i've 9 serials and the programme tell me " singular matrix".why ? and where is the probleme ?
Thank you for your answers

startz
Non-normality and collinearity are NOT problems!
Posts: 3289
Joined: Wed Sep 17, 2008 2:25 pm

Re: eviews student and ARDL

Postby startz » Mon Mar 26, 2018 2:11 pm

Quite likely you have perfect multicollinearity. If you post exactly what you did someone may be able to give more detailed advice

fati
Posts: 11
Joined: Mon Mar 26, 2018 10:05 am

Re: eviews student and ARDL

Postby fati » Mon Mar 26, 2018 2:43 pm

thank you.

i've 8 independent variables and one dependant.I've the periode of 15 years.My serials are a stationnarity in I(0) and I(1).
When i clic in ARDL and choose 2 lags or any other number,the programme no fonctionne and the text is "error message.....singular matrix".

excuse my english if its bad

startz
Non-normality and collinearity are NOT problems!
Posts: 3289
Joined: Wed Sep 17, 2008 2:25 pm

Re: eviews student and ARDL

Postby startz » Mon Mar 26, 2018 3:08 pm

I believe the problem is that you have too many lagged variables for the number of observations available.

fati
Posts: 11
Joined: Mon Mar 26, 2018 10:05 am

Re: eviews student and ARDL

Postby fati » Tue Mar 27, 2018 1:12 am

in this condition i must to show the existence of multicollinearity and remove it later to be able to use ARDL.This is it ?

or another sotluion ,
I will study each independent variable with the dependent variable, it would be better.

startz
Non-normality and collinearity are NOT problems!
Posts: 3289
Joined: Wed Sep 17, 2008 2:25 pm

Re: eviews student and ARDL

Postby startz » Tue Mar 27, 2018 6:07 am

No, the only real answer is to get more data.

fati
Posts: 11
Joined: Mon Mar 26, 2018 10:05 am

Re: eviews student and ARDL

Postby fati » Tue Mar 27, 2018 10:55 am

Thanks for your help.
I still have the same problem, or it is little or it is too important. In both situations I have nothing

startz
Non-normality and collinearity are NOT problems!
Posts: 3289
Joined: Wed Sep 17, 2008 2:25 pm

Re: eviews student and ARDL

Postby startz » Tue Mar 27, 2018 10:59 am

You can't have more series (lagged or otherwise) than you have data points. If you've been able to increase your sample to substantially more than 15 points and you're still having trouble, you might want to post your workfile and details of what you've done to see if someone can catch something.

fati
Posts: 11
Joined: Mon Mar 26, 2018 10:05 am

Re: eviews student and ARDL

Postby fati » Tue Mar 27, 2018 5:21 pm

I can't go beyond 15 years, the data is limited.

startz
Non-normality and collinearity are NOT problems!
Posts: 3289
Joined: Wed Sep 17, 2008 2:25 pm

Re: eviews student and ARDL

Postby startz » Tue Mar 27, 2018 5:36 pm

Then I suspect you can't have a model with 9 variables. I know that's frustrating, but regressions aren't magic.

fati
Posts: 11
Joined: Mon Mar 26, 2018 10:05 am

Re: eviews student and ARDL

Postby fati » Wed Mar 28, 2018 10:35 am

the problem is especially in the duration of 15 years and not the software.
Thanks for your help.


Return to “Estimation”

Who is online

Users browsing this forum: No registered users and 3 guests