Extracting sample and observations from equation

[glockner]

Hello, I'd like to extract the sample length and number of observations from a series of equations to put into an Eviews table when running a loop through multiple equations. I've discovered that the command "[equation].@smpl" does not reference the actual sample used in estimating the equation, but the sample I set initially in the program (for example "smpl 1990m1 @last). Is there another command that returns the sample used in the regression equation? Thanks.

[EViews Gareth]

Since the observations used are often very difficult to define by a sample, you cannot retrieve them in sample form.

The easiest way to do it is to generate the residual series from the equation then set a sample of if resid01<>NA (since the residual series will only have values for observations used for estimation).

[glockner]

Thanks. This helps.