Re: [Gretl-users] degree of freedom of Ljung-Box
On Mon, 23 May 2011, Sam Sam wrote:
When testing if the residuals of ARIMA is autocorrelated, degree
of freedom of Ljung-Box Q stat. is just the number of lag, or
the number of lag minus the number of parameter in gretl?
The number of lags
http://en.wikipedia.org/wiki/Ljung%E2%80%93Box_test
Allin Cottrell