EViews.com

Hello, You're correct that comparisons with the empty string can be a bit unintuitive (we actually had this discussion 9 years ago 8)). Your basic comparison expression, tab(!i,1) <> "c", is one in which an empty table cell will be treated as a missing value rather than an empty string, an...

Hello, The @rinstr function is behaving as intended. While the @rinstr function performs a right-to-left search for a matching substring, in contrast to @instr's left-to-right search, both functions return a character location for the match, which is numbered left-to-right starting at 1. The search ...

Hello,

You're correct, a severe performance loss issue is present in the stom/stomna commands for the recent versions of EViews. The next patch for EViews 14 will restore the commands' performance.

Hello,

You're correct that the DPI information wasn't being recorded correctly. The next patch to EViews 14 will include a fix for this issue.

Hello,

A fix will be included in the new EViews patch.

Hello,

It's indeed a bug. It'll be fixed in the next patch.

Hello,

Include the "title" option with the save procedure, e.g.,

Hello, Are you storing/viewing the result of @varlist or @subst via an alpha series? Alpha series have a adjustable character limit (Options -> General Options... -> Series and Alphas -> Alpha truncation) that would explain the issue you describe. However, if you simply avoid an alpha series, for ex...

Hello,

You can perform a Lasso regression as a special case using EViews' ENET estimator. The documentation has all the details, but you can easily evaluate multiple lambda values as part of the estimation.

Hello, The setmaxerrs command specifies which error will halt the program, e.g. setmaxerrs 4 stopping at the 4th error, not the number of allowed errors, e.g. setmaxerrs 4 stopping at the 5th error. A difference of only one error, but I understand how the description of setmaxerrs could be ambiguous...

Hello,

There have been updates to various model procedures, but nothing that I believe would cause such a fundamental change to add factor behavior. I haven't been able to replicate the issue you describe, would you be able to provide a workfile demonstrating the problem?

Hello,

The function you describe already exists: @firstsby. For your example, I believe "@firstsby(x, @crossid)" would produce the results you want.

Hello, Given the situation you describe, I'd consider it very unlikely that the solution changes you've observed are a consequence of the underlying solver used, e.g. Broyden vs Newton. While changing the number of equations in the model will affect the sequence of innovations used across the stocha...

You're right. That should've been fixed in the most recent patch, but the next patch will correct this issue.

Hello,

Yes, that's a small reporting error. Fixed in the next patch.