Page 1 of 1

Descriptive statistics when using AR(1) model

Posted: Wed Aug 25, 2010 4:22 am
by lentjeuhz
Hi

I use the following model:

dep c var1 var2 var3...... AR(1)

When running this regression for my complete sample (1998 2009) I see that Eviews has automatically adjusted the sample to 1999 2009.

To compute the descriptive statistics I first select all my variables (dep, var1, var2, var3, etc) and open them as a group. Next, I adjust my sample to 1999 2009 using the sample button and as a last step I click on view descriptive stats > common sample.

However, EViews uses a different number of observations for the regression (equation) than for descriptive statistics. The problem is solved when I estimate my regression without AR(1).

What I would like to do is compute descriptive statistics using the observations EViews used in the regression I specified.

Thank you for your help!

Lennart

Descriptive statistics when using AR(1) model

Posted: Wed Aug 25, 2010 6:58 am
by EViews Gareth
After estimating, set the sample to be

Code: Select all

If resid<>na
which is an easy way of setting it equal to just the observations used