No, the df are H-(p+q), against what your intuition tells you. The problem
is that you're thinking of the residuals as losing a df for each parameter,
but the asymptotics of the problem involve the sample autocorrelations of
the residuals. You can read the details in the original paper by Box an
David Stoffer wrote:
>
> I stand corrected. I thought I checked this a long time ago, but
> apparently not. tsdiag.Arima DOES NOT use the fact that the series it is
> testing (or diagnosing, if you will) are residuals from an ARIMA fit.
>
> I keep a list of R time series bloopers here:
> h
I stand corrected. I thought I checked this a long time ago, but apparently
not. tsdiag.Arima DOES NOT use the fact that the series it is testing (or
diagnosing, if you will) are residuals from an ARIMA fit.
I keep a list of R time series bloopers here:
http://www.stat.pitt.edu/stoffer/tsa2/R
I believe tsdiag() uses the correct degrees of freedom in applying Box.test,
but the graphic shows "lag" on the horizontal axis when it should display
"degrees of freedom".
raf.rossignol wrote:
>
> Hello,
>
> Prof Brian Ripley wrote:
>>
>> I think you are referring to its application to t
Hello,
Prof Brian Ripley wrote:
>
> I think you are referring to its application to the residuals of an
> ARMA(p, q) fit, and that is not what Box.test says it does.
>
> It is very easy to edit the code if you want to use a different degrees of
> freedom.
>
I am also new to R, but it seems to
Bear in mind that Box.test knows nothing about p or q: it says
Compute the Box-Pierce or Ljung-Box test statistic for examining
the null hypothesis of independence in a given time series.
I think you are referring to its application to the residuals of an
ARMA(p, q) fit, and that is n
Spencer Graves has discussed this recently ( in the last couple of
months ) in a thread and I think his FinTS package has a variant of the
Box.test
function that does what you need.
On Fri, May 16, 2008 at 11:45 AM, Nuno Prista wrote:
Dear colleagues,
I am new to R and statistics so ple
Dear colleagues,
I am new to R and statistics so please keep that in mind.
I have doubts on the df calculation of Ljung-Box test (Box.test). The
function seems to use always the df=lag=m and not df=m-p-q like suggested in
Ljung and Box (1978) paper (that is referenced).
Do you agree wi
8 matches
Mail list logo
