Recursive least squares

[ivanof971]

Hello,

I would like to know if there is a way to run a recursive least squares algorithm in order to check for linear versus non-linear dependence in my variables.

Best regards,
Konstantinos

[startz]

Hello,

I would like to know if there is a way to run a recursive least squares algorithm in order to check for linear versus non-linear dependence in my variables.

Best regards,
Konstantinos

You might take a look in the manual or the help system for "recursive least squares."

[ivanof971]

Thanks for the reply.

I have already checked that. But I am interested in a time series regression with many lags, while the document says that the option is not available for ARMA models.
That's why I asked for the algorithmn in order to try to modify the procedure accordingly.

[startz]

You can put in lags of the dependent variable, but admittedly that's not going to get you all the way to ARMA errors.