Switching regression Problem
Moderators: EViews Gareth, EViews Moderator
-
- Posts: 63
- Joined: Sat Jun 06, 2015 11:48 am
Re: Switching regression Problem
I do not understand the screen you posted...
its very easy, look:
This is model in sample 1-6000
https://ibb.co/b7Bu2R
This is a forecast static output out of sample 1-12536
https://ibb.co/dxgoU6
https://ibb.co/buTwGm
This is same model with in sample 1-6000 and static forecast output out of sample 1 - 12262
https://ibb.co/d5NdU6
LOOK, THE IN SAMPLE DATASET IS SAME.
THE OUTPUT IN SAMPLE IS TOTALLY DIFFERENT.
its very easy, look:
This is model in sample 1-6000
https://ibb.co/b7Bu2R
This is a forecast static output out of sample 1-12536
https://ibb.co/dxgoU6
https://ibb.co/buTwGm
This is same model with in sample 1-6000 and static forecast output out of sample 1 - 12262
https://ibb.co/d5NdU6
LOOK, THE IN SAMPLE DATASET IS SAME.
THE OUTPUT IN SAMPLE IS TOTALLY DIFFERENT.
Last edited by nicotrader on Mon Jan 08, 2018 3:49 pm, edited 1 time in total.
-
- Posts: 63
- Joined: Sat Jun 06, 2015 11:48 am
Re: Switching regression Problem
I USE THE DIALOGUE WINDOW, I DO NOT INSERT THE CONTROLS MANUALLY.
-
- Posts: 63
- Joined: Sat Jun 06, 2015 11:48 am
Re: Switching regression Problem
sorry but the software have indisputably a bug.
-
- Non-normality and collinearity are NOT problems!
- Posts: 3775
- Joined: Wed Sep 17, 2008 2:25 pm
Re: Switching regression Problem
Gareth is trying to help track down the problem by examining the situation in a controlled way. If you are interested in solving the puzzle, cooperate by following Gareth's lead. There may or may not be a problem, but no progress is made by simply repeating your opinion that there is a bug.
-
- Fe ddaethom, fe welon, fe amcangyfrifon
- Posts: 13312
- Joined: Tue Sep 16, 2008 5:38 pm
Re: Switching regression Problem
This is the first time you've mentioned a sample of 6000. Please, we're trying to help you, but you really aren't being very clear. If you believe this is a bug, please give precise instructions on how to replicate it.
As an aside, here is a video of me trying to replicate it, and being unable to.
http://www.eviews.com/Forum_support/nic ... rader.html
Also, it is probably worth mentioning the version of EViews and build date you are using (Help->About EViews).
As an aside, here is a video of me trying to replicate it, and being unable to.
http://www.eviews.com/Forum_support/nic ... rader.html
Also, it is probably worth mentioning the version of EViews and build date you are using (Help->About EViews).
Follow us on Twitter @IHSEViews
-
- Posts: 63
- Joined: Sat Jun 06, 2015 11:48 am
Re: Switching regression Problem
I apologize but I am not able to enter the commands manually.
As you can see in the photos I run the models through the dialogs in quick estimate.
I posted the results highlighting a clear software problem.
Again, in relation to the dataset the results in-sample change.
As you can see in the photos I run the models through the dialogs in quick estimate.
I posted the results highlighting a clear software problem.
Again, in relation to the dataset the results in-sample change.
-
- Posts: 63
- Joined: Sat Jun 06, 2015 11:48 am
Re: Switching regression Problem
Sorry Gareth it does not matter which sample I use, which is 6000 or 10361, the result is that in sample output changes every time.
-
- Posts: 63
- Joined: Sat Jun 06, 2015 11:48 am
Re: Switching regression Problem
I do not understand what is not clear to you.
I posted each step and it is clear from the screenshots that the output is always different even if I use the same in sample dataset.
I posted each step and it is clear from the screenshots that the output is always different even if I use the same in sample dataset.
-
- Fe ddaethom, fe welon, fe amcangyfrifon
- Posts: 13312
- Joined: Tue Sep 16, 2008 5:38 pm
Re: Switching regression Problem
Give clear instructions indicating what you did.
Three pictures aren't enough.
Just use words to describe every thing you do between opening EViews and seeing the results you think are wrong.
Also, provide the version and build date of your copy of EViews.
Three pictures aren't enough.
Just use words to describe every thing you do between opening EViews and seeing the results you think are wrong.
Also, provide the version and build date of your copy of EViews.
Follow us on Twitter @IHSEViews
-
- Posts: 63
- Joined: Sat Jun 06, 2015 11:48 am
Re: Switching regression Problem
eviews 10 build june 26 2017 build
-
- Fe ddaethom, fe welon, fe amcangyfrifon
- Posts: 13312
- Joined: Tue Sep 16, 2008 5:38 pm
Re: Switching regression Problem
Well your first step should be to update to the latest version.
Follow us on Twitter @IHSEViews
-
- Posts: 63
- Joined: Sat Jun 06, 2015 11:48 am
Re: Switching regression Problem
There is no fix in the update ....
Gareth please
try to do some tests this way:
1- take the dataset file I sent you.
2- try to perform tests estimating the same model but on different datasets.
Ex. 1-11000 1-12000 1-12200
3- But every time you estimate a model with a different dataset removes the observations from the excel file and loads the model every time on the dataset modified by excel. I work this way
Gareth please
try to do some tests this way:
1- take the dataset file I sent you.
2- try to perform tests estimating the same model but on different datasets.
Ex. 1-11000 1-12000 1-12200
3- But every time you estimate a model with a different dataset removes the observations from the excel file and loads the model every time on the dataset modified by excel. I work this way
-
- Fe ddaethom, fe welon, fe amcangyfrifon
- Posts: 13312
- Joined: Tue Sep 16, 2008 5:38 pm
Re: Switching regression Problem
Estimating on different data sets, will, of course, give different estimates and then produce different forecasts.
I'll also point out that re-estimating the same equation on the same data will also, probably, give different estimates and then produce different forecasts, since a) you're using random starting values, and b) your estimates didn't converge.
I'll also point out that re-estimating the same equation on the same data will also, probably, give different estimates and then produce different forecasts, since a) you're using random starting values, and b) your estimates didn't converge.
Follow us on Twitter @IHSEViews
-
- Fe ddaethom, fe welon, fe amcangyfrifon
- Posts: 13312
- Joined: Tue Sep 16, 2008 5:38 pm
Re: Switching regression Problem
One of your initial posts says that you experience the error without re-estimating. Are you now saying that isn't the case?
Follow us on Twitter @IHSEViews
-
- Posts: 63
- Joined: Sat Jun 06, 2015 11:48 am
Re: Switching regression Problem
how is it possible?
if I always use the same in sample dataset ???
if I always use the same in sample dataset ???
Who is online
Users browsing this forum: No registered users and 19 guests