EViews.com

Hello,
I would like to know how to specify lags when doing the stepwise least squares method. I intend to run a backward stepwise least squares from a general to a specific model à la Hendry. I do not want to write down every single lagged independent variable into the "List of search regressors"-window. Is there a special way to include all lags(maximum possible number of lags) that the programme should stepwisely analyze like for instance inflation(-x) or inflation(-all) or inflation(-all lags) etc? Thank you very much in advance. Greetings

inflation(-1 to -5)

will include all lags of inflation between 1 and 5

Thank you Gareth!
Now, I have encountered another problem-After clicking the OK-button to run the Stepwise Regression, EViews shows a failure message saying: "Positive or non-negative argument to function expected". What does this mean? The options that I have chosen are backwards, unidirectional with p-value of 0.05, 1000 backward steps, no number of regressors to select and no weight series. Thanks a lot. Regards

Hard to say without having your workfile.

Hey Gareth,

is it possible to specify the lags for a differenced value? e.g. d(inflation(-1 to -5)) Because then I get an error message.

No.

If im not mistaken, and if you haven't already found a simple solution to your problem, I believe you could use: then you can use: ?