EViews.com

thank you!
incidentally, wasn't aware of c=split, that's pretty useful.

Hello, Not a huge bug, but probably worth fixing. EViews version: 14 (April 22 build) When I use MIDAS, if I set the sample with a gap in the middle, for example 2010Q3–2026Q1 excluding 2020Q1 to 2022Q2 (pandemic), I noticed the R-squared can turn negative. In the regression I was testing, I was get...

great news, thanks

much appreciated!

Could you update to the latest version to see if that fixes it?

I updated to the latest version (March 4 2026 build), but unfortunately I still have the same problem

My organization is currently blocking updates. But I'll complain and get on their case and update you. Thanks

Sorry, should have mentioned that earlier.
EViews 14, Feb. 6 2025 build.

As the title says, I'm encountering an apparent bug when trying to solve the model with stochastic simulations, if I select a bootstrap sample that is not continuous. For example, this leads to a crash where EViews closes: sample s1 2000q1 2025q4 if @year<>2020 m.stochastic(s=s1) m.solve Any help/fi...

Follow-up question: could the data description also be added fetched?

For example, it would be nice to see a description of the series added as a series attribute.

Thanks for the quick response. I understand it may be a very large download, but if there's anything that can be done to make it smaller (or download faster) in the future, that would make the feature much more helpful. For example, in my case, all these series were being downloaded from the very sa...

First: thanks to the EViews team for adding the Statistics Canada data connectivity feature. It's pretty handy to have! But, at the moment, its usefulness is severely limited by the fact that it takes a very long time for me to fetch data from it. I will share some details below about that. mode qui...

Ok, I see. Thanks for the quick response!

In Eviews 13, a line like this would work: var var_example.ls(noconst) 0 0 x1 x2 @ exog1 It would successfully create a VAR with no lags -- each endogenous variable is simply only regressed against "exog1". I perfectly understand that this is not exactly what a VAR should be - but the abil...

It seems that this is specific to the model you're using. Because I tried it on my end, and the model I was using solved successfully both times. Strange!

Well, one idea: in X-13, you can switch the model estimation method to "additive" rather than multiplicative. Then, values of zero won't be a problem (NAs are still an issue). The downside is that you may get a negative value in 2020m4 and 2020m5, but then you could overwrite that to zero.