EViews.com

Hello,

I get the error message: "Near singular matrix in computing the steady state probabilities." when I try the example regarding Hamiltons GNP, using reviews workfile attached: I followed all the steps from the manual and I keep receiving this error message.

The same happens with data I used to re-produce the findings from a published paper (10.1109/TPWRS.2020.3015871).
Already my covariates have been tested for multi-colinearity presence.

I am PhD student and I am using the Student Lite version.

Can anyone help with this ?

Kind regards

Hi. I've tested this on various versions of EViews (not Student Lite yet, but I will if we can't pin this down) and in all cases, it converges and produces standard errors. But before going further, I did want to ask you if you are using the seetings in that equation without change, and if so, what convergence tolerance you are using. Thanks.

Thanks for your reply.

I used 1e-8 as a tolerance convergence.
Also, you can see the settings in the photos attached. Maybe it is something that I overlooked since I am a beginner at Eviews.
If so please feel free to inform me.

Best regards

Thanks. That's useful, but I still can't replicate. Could you send me a screenshot of your output. And can you take a look at the Help/About EViews main menu item and give me information about your version and build-date. Thanks.

Thanks

The output that I get is: And the version etc. is:

Okay. Thanks. Let me see what I can figure out.

Thanks for your assistance.

Still looking into this. The problem is that while I do see the error in the Student Lite version, it does not appear in the standard EViews 12, nor in any of our standard tests. Will continue to explore. Thanks for you patience.

Hi. Found the issue. It's EVIews 12 only. A patch has been posted. You can initiate by selecting Help/EViews Update... from the main EViews menu.

Sorry for the inconvenience. Let us know if you have any additional issues.